Dermatological and Transdermal Formulations

ISBN: 0-8247-9889-9
This book is printed on acid-free paper.
Headquarters
Marcel Dekker, Inc.
270 Madison Avenue, New York, NY 10016
tel: 212-696-9000; fax: 212-685-4540
Eastern Hemisphere Distribution
Marcel Dekker, AG
Hutgasse 4, Postfach 812, CH-4001 Basel, Switzerland
tel: 41-61-261-8482; fax: 41-61-261-8896
World Wide Web

http://www.dekker.com
The publisher offers discounts on this book when ordered in bulk quantities. For more information,
write to Special Sales/Professional Marketing at the headquarters address above.
Copyright 䉷 2002 by Marcel Dekker, Inc. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, microfilming, and recording, or by any infor-
mation storage and retrieval system, without permission in writing from the publisher.
Current printing (last digit):

10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

Preface
The past two decades have witnessed brilliant discoveries regarding the structure and func-
tions of the stratum corneum.
—Albert Kligman, 2000
An immense amount of research has been carried out over the past two decades on
the micromorphology of the skin, in particular of the stratum corneum, and the
important role that this organ plays in the maintenance of human life. It has also
been nearly two decades since the publication of Brian Barry’s book Dermatological
Formulations—Percutaneous Absorption. This book remains one of the most widely
and frequently cited references in the field of skin transport and also has been used
extensively as an introduction to the complexities surrounding the theory and de-
velopment of topical pharmaceutical products.
The introduction and subsequent success of transdermal therapeutic systems
have advanced our understanding of the structure of the skin and the mechanisms
of transport through the barrier membrane. In addition, technological developments
in molecular biology and pharmacology have led to an increased understanding of
the biochemistry of skin diseases. The result is the introduction of new therapeutic
strategies that use both existing and new chemical entities to treat skin diseases. This
volume serves as a useful addition to the literature in the dermatopharmaceutics field.
The rational treatment of skin diseases, based on the biochemical mechanisms
underlying the pathology, is discussed in Chapter 2. For example, vitamin D3 deriv-
atives, such as calcipotriol and calcitriol, have recently been introduced as topical
therapeutic modalities for the treatment of psoriasis. This is a result of the finding
iii
iv Preface
that some compounds of this type possess a high binding affinity to specific cellular
receptors and are potent regulators of cell differentiation and inhibitors of cell pro-
liferation in human keratinocytes. Cosmetic scientists have long known the epidermal
advantages of another vitamin, retinol (vitamin A). Deficiency of this vitamin has
been implicated in squamous metaplasia and keratinization of epithelial tissue, and
several derivatives have been synthesized and evaluated for their effects in such
diseases as acne, psoriasis, and hyperkeratosis. The results have been somewhat
variable; however, the recent identification of several receptor proteins for retinoic
acid should lead to the development of more potent analogs with fewer side effects.
The ability to enhance skin penetration and permeation has been the subject of
considerable research over the past two decades and is reviewed in Chapter 6. The
science of penetration enhancement has expanded considerably over the past few
years, and it is now possible to increase drug delivery across the skin using both
chemical and physical means. Various synthetic (e.g., SEPA威 and Azone威) and nat-
ural (e.g., terpenes) compounds have proved useful in this respect. Moreover, there
is evidence that the skin penetration of large molecules such as insulin can be in-
creased using physical methods of enhancement, such as iontophoresis.
The use of the skin as a drug delivery route for both topical and systemic
therapy is covered in Chapter 7. Transdermal drug delivery using patches or semi-
solid formulations is now a reality with products available for travel sickness, hy-
pertension, angina, postmenopausal symptoms, male hypogonadism, pain, inflam-
mation, and smoking cessation. Problems of irritation are being overcome with the
development of skin-compatible materials, such as some of the newer pressure-sen-
sitive adhesives. The success of such systems has been achieved only by means of
a greater understanding of the physical and biochemical nature of the permeation
routes through the skin, especially in relation to the intercellular lipid lamellae of
the stratum corneum (as discussed in Chapters 1, 3, and 4). In addition, the methods
of studying percutaneous absorption, both in vivo and in vitro, have become more
standardized thanks to the efforts of the American Association of Pharmaceutical
Scientists (AAPS), the U.S. Food and Drug Administration (FDA), and other indus-
trial and regulatory bodies. Chapter 5 provides a complete description of the AAPS/
FDA guidelines for such experimentation, together with a full evaluation of the in
vivo tape-stripping procedure. This chapter also gives a description of the use of
cultured skin membranes for the study of irritation and other toxic responses to
materials applied to the skin.
The formulation of dermatological vehicles has become more innovative with
the introduction of many new excipient materials and the development of delivery
systems made up of vesicles such as liposomes and niosomes. These are discussed
in Chapters 6 and 7. As an example, the feasibility of using supersaturated solutions
as vehicles for improving dermal drug delivery has been established. This type of
strategy will undoubtedly reduce the amount of drug necessary for a therapeutic
effect, which should result in fewer local side effects and a lower incidence of
unwanted systemic effects. The use of liposomes and niosomes in cosmetic formu-
lation is reputed to impart beneficial properties to such products as moisturizers.
Although these types of vesicles can be useful in the targeting of pharmaceutical
agents to epidermal sites of action, their usefulness in transporting drugs across the
skin to systemic sites has yet to be fully established. The usefulness of multivariate
optimization in the scale-up of dermatological dosage forms is discussed in Chapter
Preface v
9. Chapter 8 covers bioequivalence of dermal and transdermal systems. Safety con-

siderations are outlined in Chapters 10 and 11.
This book will be useful to pharmacy students and practitioners and cosmetic
and veterinary scientists. It may also prove useful to toxicologists working in the
field of risk assessment and dermatologists requiring a deeper understanding of the
mechanisms of drug transport through the skin. The chapters have been authored by
international experts in their fields and provide a comprehensive review of current
dermatopharmaceutics.
I acknowledge, and am extremely grateful for, the hard work and infinite pa-
tience shown by the contributors to this volume. I would also like to acknowledge
the unreserved help provided by my colleague Dr. Keith Brain. Finally, since the
written word cannot fully express my love and gratitude to Peggy for her support
and encouragement during many hours of word processing and reference hunting,
this absentee husband will find other ways.
Kenneth A. Walters
Contents
Preface iii
Contributors ix
1. The Structure and Function of Skin 1

Kenneth A. Walters and Michael S. Roberts
2. Common Skin Disorders and Their Topical Treatment 41

C. Colin Long
3. Basic Mathematical Principles in Skin Permeation 61

Adam C. Watkinson and Keith R. Brain
4. Skin Transport 89
Michael S. Roberts, Sheree Elizabeth Cross, and Mark A. Pellett
5. Methods for Studying Percutaneous Absorption 197

Keith R. Brain, Kenneth A. Walters, and Adam C. Watkinson
6. Formulation Strategies for Modulating Skin Permeation 271

Adrian F. Davis, Robert J. Gyurik, Jonathan Hadgraft,
Mark A. Pellett, and Kenneth A. Walters
vii
viii Contents
7. Dermatological Formulation and Transdermal Systems 319

Kenneth A. Walters and Keith R. Brain
8. Bioavailability and Bioequivalence of Dermatological

Formulations 401
Christian Surber and Adrian F. Davis
9. Scale-up of Dermatological Dosage Forms: A Case for

Multivariate Optimization and Product Homogeneity 499
Orest Olejnik and Bruce A. Firestone
10. Safety Considerations for Dermal and Transdermal Formulations 511

Peter J. Dykes and Anthony D. Pearse
11. Transdermal Delivery and Cutaneous Reactions 529

Jagdish Singh and Howard I. Maibach
Index 549
Contributors
Keith R. Brain, Ph.D. An-eX Analytical Services Ltd., Cardiff, Wales
Sheree Elizabeth Cross, Ph.D. Department of Medicine, University of Queens-

land, Princess Alexandra Hospital, Brisbane, Queensland, Australia
Adrian F. Davis, Ph.D. GlaxoSmithKline Consumer Healthcare, Weybridge, Kent,

England
Peter J. Dykes, B.Sc., Ph.D. Cutest (Skin Toxicity Testing Company), Cardiff,
Wales
Bruce A. Firestone, Ph.D. Allergan, Inc., Irvine, California
Robert J. Gyurik MacroChem Corporation, Lexington, Massachusetts
Jonathan Hadgraft, D.Sc., F.R.S.C. Medway Sciences, University of Greenwich,

Chatham Maritime, Kent, England
C. Colin Long, F.R.C.P. Department of Dermatology, Cardiff and Vale NHS Trust,
Cardiff, Wales
Howard I. Maibach, M.D. Department of Dermatology, University of California

School of Medicine, San Francisco, California
ix
x Contributors
Orest Olejnik, Ph.D., M.R.Pharm.S. Allergan, Inc., Irvine, California
Anthony D. Pearse, M.Sc., M.I.Biol., C.Biol., F.I.Sc.T. Department of Dermatol-

ogy, University of Wales College of Medicine, and Cutest (Skin Toxicity Testing
Company), Cardiff, Wales
Mark A. Pellett, Ph.D., M.R.Pharm.S. Whitehall International, Havant, England
Michael S. Roberts, Ph.D., D.Sc. Department of Medicine, University of Queens-

land, Princess Alexandra Hospital, Brisbane, Queensland, Australia
Jagdish Singh Department of Pharmaceutical Sciences, College of Pharmacy,

North Dakota State University, Fargo, North Dakota
Christian Surber, Ph.D., Priv.-Doz. Dr. Institute of Hospital-Pharmacy, University

Clinics, Kantonsspital Basel, Basel, Switzerland
Kenneth A. Walters, F.I.Biol., Ph.D. An-eX Analytical Services Ltd., Cardiff,

Wales
Adam C. Watkinson, Ph.D., M.B.A.* An-eX Analytical Services Ltd., Cardiff,

Wales
*Current affiliation: Strakan Pharmaceuticals Ltd., Galashiels, Scotland.

1
The Structure and Function of Skin
KENNETH A. WALTERS
An-eX Analytical Services Ltd., Cardiff, Wales
MICHAEL S. ROBERTS
University of Queensland, Princess Alexandra Hospital, Brisbane,
Queensland, Australia
I. INTRODUCTION
The skin is the largest organ of the body, accounting for more than 10% of body
mass, and the one that enables the body to interact most intimately with its environ-
ment. Figure 1 shows a diagrammatic illustration of the skin. In essence, the skin
consists of four layers: the stratum corneum (nonviable epidermis), the remaining
layers of the epidermis (viable epidermis), dermis, and subcutaneous tissues. There
are also several associated appendages: hair follicles, sweat ducts, apocrine glands,
and nails. Many of the functions of the skin can be classified as essential to survival
of the body bulk of mammals and humans in a relatively hostile environment. In a
general context, these functions may be classified as protective, maintaining homeo-
stasis, or sensing. The importance of the protective and homeostatic role of the skin
is illustrated in one context by its barrier property. This allows the survival of humans
in an environment of variable temperature; water content (humidity and bathing);
and the presence of environmental dangers, such as chemicals, bacteria, allergens,
fungi, and radiation. In a second context, the skin is a major organ for maintaining
the homeostasis of the body, especially in terms of its composition, heat regulation,
blood pressure control, and excretory roles. It has been argued that the basal meta-
bolic rate of animals differing in size should be scaled to the surface area of the
body to maintain a constant temperature through the skin’s thermoregulatory control
(1). Third, the skin is a major sensory organ in terms of sensing environmental
influences, such as heat, pressure, pain, allergen, and microorganism entry. Finally,
the skin is an organ that is in a continual state of regeneration and repair. To fulfill
1
2 Walters and Roberts
Figure 1 Components of the epidermis and dermis of human skin.
each of these functions, the skin must be tough, robust, and flexible, with effective
communication between each of its intrinsic components.
Many agents are applied to the skin either deliberately or accidentally, with
either beneficial or deleterious outcomes. The use of topical products was evident in
ancient times, and there are reports of systemic benefits of topical anti-infective and
hormonal agents in the 1940s. Modern transdermal patch technology was introduced
in the late 1970s. The main interests in dermal absorption assessment are in the
application of compounds to the skin (a) for local effects in dermatology (e.g., cor-
ticosteroids for dermatitis); (b) for transport through the skin for systemic effects
(e.g., nicotine patches for smoking cessation); (c) for surface effects (e.g., sunscreens,
cosmetics, and anti-infectives); (d) to target deeper tissues (e.g., nonsteroidal anti-
inflammatory agents [NSAIDs] for muscle inflammation); and (e) unwanted absorp-
tion (e.g., solvents in the workplace, agricultural chemicals, or allergens). Figure 2
summarizes these processes and sites of effect of compounds applied to the skin.
The skin became popular as a potential site for systemic drug delivery because it
>
Figure 2 (A) Structure of the skin and processes of percutaneous absorption and trans-
dermal delivery. Absorption can occur through sweat ducts (1), intercellular regions of the
stratum corneum (2), and through the hair follicles (3). (B) Dermal absorption, sites of action
and toxicity.
Structure and Function of Skin 3
was thought to (a) avoid the problems of stomach emptying, pH effects, and enzyme
deactivation associated with gastrointestinal passage; (b) to avoid hepatic first-pass
metabolism; and (c) to enable control of input, as exemplified by termination of
delivery through removal of the device. In practice, as discussed later in this book,
delivery of solutes through the skin is associated with various difficulties, such as
(a) the variability in percutaneous absorption owing to site, disease, age, and species
differences; (b) the skin’s ‘‘first-pass’’ metabolic effect; (c) the reservoir capacity of
the skin; (d) irritation and other toxicity caused by topical products; (e) heterogeneity
and inducibility of the skin in both turnover and metabolism; (f) inadequate definition
of bioequivalence criteria; and (g) an incomplete understanding of the technologies
that may be used to facilitate or retard percutaneous absorption. However, the con-
trolled delivery of solutes through the skin continues to be of interest, with the further
development of technologies, such as chemical penetration enhancement, sonopho-
resis, transferosomes, and electroporation. The extent to which these are translated
into practice will be defined by time.
II. GROSS STRUCTURE AND FUNCTION OF THE SKIN

Whereas Figure 1 provides an overview of the gross structure of the skin, Figure 3
represents the skin components in terms of the various functions they perform. It
needs to be emphasized that the protection, homeostatic, and sensing functions of
the skin are both overlapping and integrated. For instance, barrier properties to a
chemical entity involves resistance to its entry (barrier provided by stratum corneum),
metabolism for that proportion of entity bypassing the stratum corneum (in viable
epidermis), sensing of and attention to damage caused by entry (inflammatory me-
diator release in epidermis, with involvement of dermis), and removal of entity from
site by dermal blood supply and distribution into those body organs specifically
responsible for elimination of the entity by metabolism (liver) and excretion (kidney).
Heat regulation occurs through the use of the subcutaneous fat pad, physiological
regulation of blood flow to effect, for instance, heat loss by vasodilation, and cooling
by perspiration. We now consider the structure and functions provided by each skin
component in some detail.
A. The Epidermis
The epidermis performs a number of functions, as shown in Figure 3, one of the
most important being the generation of the stratum corneum, as described later. The
stratum corneum is the heterogeneous outermost layer of the epidermis and is ap-
proximately 10–20 ␮m thick. It is nonviable epidermis and consists, in a given cross-
section, of 15–25 flattened, stacked, hexagonal, and cornified cells embedded in a
mortar of intercellular lipid. Each cell is approximately 40 ␮m in diameter and 0.5
␮m thick. The thickness varies, however, and may be a magnitude of order larger
in areas such as the palms of the hand and soles of the feet, areas of the body
associated with frequent direct and substantial physical interaction with the physical
environment. Not surprisingly, the absorption of solutes, such as methyl salicylate,
is slower through these regions than through the skin of other parts of the body. The
stratum corneum barrier properties may be partly related to its very high density (1.4
g/cm3 in the dry state), its low hydration of 15–20%, compared with the usual 70%
Figure 3 Skin components and their function.
for the body, and its low surface area for solute transport (it is now recognized that
most solutes enter the body through the less than 0.1-␮m–wide intercellular regions
of the stratum corneum). Each stratum corneum cell is composed mainly of insoluble
bundled keratins (⬃70%) and lipid (⬃20%) encased in a cell envelope, accounting
for about 5% of the stratum corneum weight. The intercellular region consists mainly
of lipids and desmosomes for corneocyte cohesion, as described later. The barrier
function is further facilitated by the continuous desquamation of this horny layer
with a total turnover of the stratum corneum occurring once every 2–3 weeks. Ac-
cordingly, very lipophilic agents, such as sunscreens and substances binding to the
horny layer (e.g., hexachlorophane), may be less well absorbed into the body than
would be indicated by the initial partitioning of the agents into the horny layer from
an applied vehicle. The stratum corneum also functions as a barrier to prevent the
loss of internal body components, particularly water, to the external environment. It
is estimated that the efficiency of this barrier is such that water loss from ‘‘insensible
perspiration’’ is restricted to 0.5 ␮L/cm2/h⫺1, or 250 mL of water per day for a normal
adult. Disorders of epithelization, such as psoriasis, lead to a faster skin turnover,
sometimes being reduced to 2–4 days, with improper stratum corneum barrier func-
tion formation.
The cells of the stratum corneum originate in the viable epidermis and undergo
many morphological changes before desquamation. Thus the epidermis consists of
several cell strata at varying levels of differentiation (Fig. 4). The origins of the cells
of the epidermis lie in the basal lamina between the dermis and viable epidermis. In
this layer there are melanocytes, Langerhans cells, Merkel cells, and two major
keratinic cell types: the first functioning as stem cells having the capacity to divide
and produce new cells; the second serving to anchor the epidermis to the basement
Figure 4 Epidermal differentiation: major events include extrusion of lamellar bodies, loss
of nucleus, and increasing amount of keratin in the stratum corneum. The diagram is not to
scale and only a few cells are shown for clarity.
membrane (2). The basement membrane is 50–70 nm thick and consists of two
layers—the lamina densa and lamina lucida—which comprise mainly proteins, such
as type IV collagen, laminin, nidogen, and fibronectin. Type IV collagen is respon-
sible for the mechanical stability of the basement membrane, whereas laminin and
fibronectin are involved with the attachment between the basement membrane and
the basal keratinocytes.
The cells of the basal lamina are attached to the basement membrane by hemi-
desmosomes, which are found on the ventral surface of basal keratinocytes (3).
Hemidesmosomes appear to comprise three distinct protein groups: two of which are
bullous pemphigoid antigens (BPAG1 and BPAG2), and the other epithelial cell-
specific integrins (4–6). BPAG1 is associated with the organization of the cytoskel-
etal structure and forms a link between the hemidesmosome structure and the keratin
intermediate filaments. The integrins are transmembrane receptors that mediate at-
tachment between the cell and the extracellular matrix. Human epidermal basal cells
contain integrins ␣2␤1, ␣3␤1, and ␣6␤4. Integrin ␣6␤4 and BPAG2 appear to be the
major hemidesmosomal protein contributors to the anchoring of the keratinocyte,
spanning from the keratin intermediate filament, through the lamina lucida, to the
lamina densa of the basement membrane (7). In the lamina densa, these membrane-
spanning proteins interact with the protein laminin-5 which, in turn, is linked to
collagen VII, the major constituent of the anchoring fibrils within the dermal matrix.
It has also been suggested that both BPAG2 and integrin ␣6␤4 mediate in the signal
transductions required for hemidesmosome formation (8) and cell differentiation and
proliferation. Integrin ␣3␤1 is associated with actin and may be linked with laminin-
5. Epidermal wounding results in an up-regulation of these proteins that appears to
be involved with cell motility and spreading. The importance of maintaining a secure
link between the basal lamina cells and the basement membrane is obvious, and the
absence of this connection results in chronic blistering diseases such as pemphigus
and epidermolysis bullosa.
In addition to hemidesmosome cell–matrix binding, another site for adhesion
of the cells of the epidermal basal layer and the basal membrane is the adherens
junction (9). The adherens junction expresses a protein profile different from des-
mosomes and hemidesmosomes (10,11) and contains talin, vinculin, and cadherins,
and with the possible participation of type XIII collagen (12). Whereas the desmo-
somes and hemidesmosomes are linked to cytoplasmic keratin, the adherens junctions
are linked to cytoplasmic actin microfilaments.
The cohesiveness of, and the communication between, the viable epidermal
cells, the cell–cell interaction, is maintained in a fashion similar to the cell–matrix
connection, except that desmosomes replace hemidesmosomes. Adherens junctions
are also located at keratinocyte–keratinocyte borders (13,14). At the desmosomal
junction there are two transmembrane glycoprotein types: desmogleins and desmo-
collins (each of three subtype desmosome-specific cadherins), which are associated
with the cytoplasmic plaque proteins, desmoplakin, plectin, periplakin, envoplakin,
and plakoglobin (Fig. 5), and provide a linkage to keratin intermediate filaments. On
the other hand, in the adherens junction, classic cadherins act as transmembrane
glycoproteins and these are linked to the actin filaments by ␣-, ␤-, and ␥-catenins
(15,16). Thus, in the epidermis, the desmosomes are responsible for interconnecting
individual cell keratin cytoskeletal structures, thereby creating a tissue very resistant
to shearing forces.
Figure 5 Epidermal cell cohesion and communication is provided by desmosomes and

adherens junctions. Different functions are attributed to these junctions. See text for details.
The importance of the calcium ion in cell regulation and intercellular com-
munication has been known for some time (17). It is not surprising, therefore, that
the formation of desmosomes and hemidesmosomes appears to be induced by Ca2⫹
and mediated by protein kinase C (PKC) activation (18). The presence of Ca2⫹
activates the metabolism of inositol phospholipids, resulting in the generation of
diacylglycerol and inositol-1,4,5-triphosphate. The diacyglycerol subsequently acti-
vates protein kinase C, which plays an important role in keratinocyte differentiation
(19), and the inositol-1,4,5-triphosphate generates further Ca2⫹ influx into the cyto-
plasm. The epidermal distribution and function of Ca2⫹, and several other physio-
logical elements, has been recently reviewed (20). As discussed later, Ca2⫹ also plays
a role in proteolysis and desquamation.
Another cell type found in the epidermal basal layer is the Langerhans cell.
Although it has been proposed that these dendritic cells plays a role in the control
of the proliferation of keratinocytes (21), they have since become recognized as the
prominent antigen-presenting cells of the skin immune system (22,23). As such, their
main function appears to be to pick up contact allergens in the skin and present these
agents to T lymphocytes in the skin-draining lymph nodes; thus, they play an im-
portant role in contact sensitization. Cell surface moieties on the Langerhans cells
are modified and the cells increase in size following topical application of hapten.
The activated cells migrate from the epidermis to the dermis and from there to the
regional lymph nodes where they sensitize T cells. The ability of Langerhans cells
to migrate from bone marrow, localize in a specific region of the epidermis, and
further migrate when activated, suggests that there is some mechanism for accu-
mulation in the epidermis, adhesion to keratinocytes and the basement membrane,
and for disruption of the adhesive bond. Migration into the epidermis may be me-
diated by granulocyte–macrophage colony-stimulating factor (GM-CSF), tumor ne-
crosis factor-␣ (TNF-␣), interleukin-6 (IL-6), transforming growth factor-␤ (TGF-
␤), chemotactic cytokines, such as monocyte chemotactic protein (MCP), and
cutaneous lymphocyte-associated antigen (CLAA) (24). The adhesive bonds within
the epidermis appear to be formed by interaction of Langerhans cells with extracel-
lular matrix proteins, such as fibronectin and laminin, through ␤1-integrins (25).
Detachment of Langerhans cells from keratinocytes and the basement membrane
following skin sensitization may be mediated by epidermal cytokines, including GM-
CSF and TNF-␣ (26), whereas cell maturation, which occurs during transit to the
local lymph nodes, may be mediated by GM-CSF. Recently, it was shown that IL-
1␤ and TNF-␣ acted directly on Langerhans cells to reduce adhesion to keratinocytes
and the basement membrane by down-regulating the binding protein E-cadherin (27).
Melanocytes are a further functional cell type of the epidermal basal layer and
are also present in hair and eyes. The main function of these cells is to produce
melanins, high molecular weight polymers of indole quinone, which affect pigmen-
tation of the skin, hair, and eyes (28,29). Melanin is produced in the melanosomes:
membrane-bound organelles that are transferred to keratinocytes, probably by a pro-
cess involving phagocytosis (30), to provide a uniform distribution of pigmentation.
Intracellular movement of melanosomes is possibly mediated by actin and microtu-
bules (31). Visible pigmentation is dependent not only on the number, shape, and
size of the melanosomes, but also on the chemical nature of the melanin. Hair color
is governed by melanocytes which reside in the hair bulbs within the dermis (32).
Melanosomes are transferred to the growing hair shaft. The major function of skin
pigmentation is to provide protection against harmful environmental effects, such as
ultraviolet (UV) radiation, especially in the proliferating basal layers where the mu-
tagenic effects of this type of insult have particularly serious implications. Melano-
cytes remain attached to the basal layer and are thought to exist in a nonproliferative
state when in contact with undifferentiated keratinocytes. Recent studies have indi-
cated, however, that melanocytes can proliferate if they are separated from the basal
layer and surrounded by differentiated keratinocytes (33).
Regulation of melanogenesis is a complex process (34) involving some 80
genetic loci. It is mutations of these genes that lead to pathological states, such as
albinism, vitiligo, and Waardenberg’s syndrome. The initial substrate for melanin is
tyrosine, which is hydroxylated to dihydroxyphenylalanine (DOPA), and from there
may be processed through several routes to produce either the eumelanins (5,6-
dihydroxyindole melanin and 5,6-dihydroxyindole-2-carboxylic acid melanin) or
pheomelanins (Fig. 6). Eumelanins are brown–black, whereas pheomelanins are yel-
low–red. It is thought that interactions between the tridecapeptide ␣-melanocyte-
Figure 6 Routes of synthesis of the melanins: tyrosine is converted to dihydroxyphenyl-

alanine (DOPA) by tyrosine hydroxylase and is subsequently converted to DOPAquinone by
DOPA oxidase. Subsequent processing is by several different pathways to yield either dihy-
droxyindole melanin, dihydroxyindole-2-carboxylic acid melanin or pheomelanin. (From Ref.
34.)
stimulating hormone (␣-MSH) and agouti signal protein are responsible for govern-
ing the type of melanogenesis pathway followed and that, in conditions in which
␣-MSH dominates, eumelanins are produced. UV radiation appears to increase pro-
duction of the precursor hormone proopiomelanocortin, which increases ␣-MSH pro-
duction, resulting in increased levels of eumelanins (35,36).
The final type of cell found in the basal layer of the stratum corneum is the
Merkel cell. These cells can be distinguished from the keratinocytes by their clear
cytoplasm and lack of tonofilaments. The cells are closely associated with nerve
endings, present on the other side of the basement membrane, which suggests they
function as sensory receptors of the nervous system. Although histochemical evi-
dence demonstrating the presence of acetylcholinesterase suggests a sensory role for
Merkel cells, there has been no direct evidence for the release of neurotransmitters.
Indeed, acetylcholinesterases have been found in keratinocytes (37). Despite this lack
of confirmation, most researchers in the field agree that Merkel cells play a role (a)
in the mechanosensory system; (b) in trophic action on peripheral nerve fibers; (c)
in stimulating and maintaining proliferation and keratinocytes; and (d) in release of
bioactive substances to subepidermal structures (38,39).
B. The Dermis
The dermis, a critical component of the body, not only provides the nutrative, im-
mune, and other support systems for the epidermis, through a thin papillary layer
adjacent to the epidermis, but also plays a role in temperature, pressure, and pain
regulation. The main structural component of the dermis is referred to as a coarse
reticular layer. The dermis is about 0.1–0.5 cm thick and consists of collagenous
fibers (70%), providing a scaffold of support and cushioning, and elastic connective
tissue, providing elasticity, in a semigel matrix of mucopolysaccharides. In general,
the dermis has a sparse cell population. The main cells present are the fibroblasts,
which produce the connective tissue components of collagen, laminin, fibronectin,
and vitronectin; mast cells, which are involved in the immune and inflammatory
responses; and melanocytes involved in the production of the pigment melanin.
Contained within the dermis is an extensive vascular network providing for the
skin nutrition, repair, and immune responses and, for the rest of the body, heat
exchange, immune response, and thermal regulation. The blood flow rate to the skin
is about 0.05 mL min⫺1 cc⫺3 of skin, providing a vascular exchange area equivalent
to that of the skin surface area. Skin blood vessels derive from those in the subcu-
taneous tissues, with an arterial network supplying the papillary layer, the hair fol-
licles, the sweat and apocrine glands, the subcutaneous area, as well as the dermis
itself. These arteries feed into arterioles, capillaries, venules, and, thence, into veins.
Of particular importance in this vascular network is the presence of arteriovenous
anastomoses at all levels in the skin. These arteriovenous anastomoses, which allow
a direct shunting of up to 60% of the skin blood flow between the arteries and veins,
thereby avoiding the fine capillary network, are critical to the skin’s functions of
heat regulation and blood vessel control. Blood flow changes are most evident in
the skin in relation to various physiological responses and include psychological
effects, such as shock (‘‘draining of color from the skin’’) and embarrassment
(‘‘blushing’’); temperature effects; and physiological responses to exercise, hemor-
rhage, and alcohol consumption.
The lymphatic system is an important component of the skin in regulating its

interstitial pressure, mobilization of defense mechanisms, and in waste removal. It
exists as a dense, flat meshwork in the papillary layers of the dermis and extends
into the deeper regions of the dermis. Cross and Roberts (40) have shown that
whereas blood flow determines the clearance of small solutes, such as water and
lidocaine, lymphatic flow is an important determinant in the dermal removal of larger
solutes, such as interferon. Also present in the dermis are a number of different types
of nerve fibers supplying the skin, including those for pressure, pain, and
temperature.
C. The Subcutis
The deepest layer of the skin is the subcutaneous tissue or hypodermis. The hypo-
dermis acts as a heat insulator, a shock absorber, and an energy storage region. This
layer is a network of fat cells arranged in lobules and linked to the dermis by
interconnecting collagen and elastin fibers. As well as fat cells (possibly 50% of the
body’s fat), the other main cells in the hypodermis are fibroblasts and macrophages.
One of the major roles of the hypodermis is to carry the vascular and neural systems
for the skin. It also anchors the skin to underlying muscle. Fibroblasts and adipocytes
can be stimulated by the accumulation of interstitial and lymphatic fluid within the
skin and subcutaneous tissue (41).
D. Skin Appendages
There are four skin appendages: the hair follicles with their associated sebaceous
glands, eccrine sweat glands, apocrine sweat glands, and the nails. Each appendage
has a different function as outlined in Table 1. The hair follicles are distributed across
the entire skin surface with the exception of the soles of the feet, the palms of the
hand and the lips. A smooth muscle, the erector pilorum, attaches the follicle to the
dermal tissue and enables hair to stand up in response to fear. Each follicle is as-
sociated with a sebaceous gland that varies in size from 200 to 2000 ␮m in diameter.
The sebum secreted by this gland (Table 2), consisting of triglycerides, free fatty
acids, and waxes, protects and lubricates the skin as well as maintaining a pH of
about 5. The fractional area for these is slightly more than 1/1000 of the total skin
surface (see Table 1). Also described in Table 1 are the eccrine or sweat glands and
apocrine glands, accounting for about two-thirds and one-third of all glands, respec-
tively. The eccrine glands are epidermal structures that are simple, coiled tubes aris-
ing from a coiled ball, of approximately 100 ␮m in diameter, located in the lower
dermis. It secretes a dilute salt solution with a pH of about 5, this secretion being
stimulated by temperature-controlling determinants, such as exercise and high en-
vironmental temperature, as well as emotional stress through the autonomic (sym-
pathetic) nervous system (see Table 1). These glands have a total surface area of
about 1/10,000 of the total body surface. The apocrine glands are limited to specific
body regions and are also coiled tubes. These glands are about ten times the size of
the eccrine ducts, extend as low as the subcutaneous tissues and are paired with hair
follicles.
In many respects the nail may be considered as vestigial in humans. However,
some manipulative and protection function can be ascribed. Certainly nail plate com-
position, layers of flattened keratinized cells fused into a dense, but somewhat elastic
Structure and Function of Skin
Table 1 Appendages Associated with the Skin
Appendage
Hair follicle and sebaceous
Parameter gland Eccrine gland Apocrine gland Nails
Function Protection (hair) and lubrication Cooling Vestigal secondary sex Protection
(sebum) gland?
Distribution Most of the body Most of the body Axillae, nipples, anogenital Ends of fingers
and toes
Average/cm2 57–100 100–200 Variable —
Fractional area 2.7 ⫻ 103 10⫺4 Variable —
Secretions Sebum Sweat (dilute ‘‘Milk’’ protein, lipoproteins, Nil

saline) lipid
Secretions stimulated by Heat (minor) Heat, cholinergic Heat —
Biochemical innervation — Cholinergic Cholinergic (?) —

of gland response
Control Hormonal Sympathetic Sympathetic nerves —

nerves
13
Table 2 Lipid Composition of Human Sebum
Lipid Lumen of glanda Skin surfacea
Squalene 15 15
Wax esters 25 25
Cholesteryl esters 2 2
Triglycerides 57 42
Fatty acids 0 15b
Cholesterol 1 1
a
Expressed as weight%.
b
Most abundant fatty acid in human sebum is C16:1.
Source: Courtesy of P. Wertz.
mass, will afford some protection to the highly sensitive terminal phalanx. The cells
of the nail plate originate in the nail matrix and grow distally at a rate of about 0.1
mm/day. In the keratinization process the cells undergo shape and other changes,
similar to those experienced by the epidermal cells forming the stratum corneum.
This is not surprising because the nail matrix basement membrane shows many
biochemical similarities to the epidermal basement membrane (42). Thus, the ex-
pression of integrins ␣2␤1 and ␣3␤1 within the nail matrix basement membrane zone
is indicative of a highly proliferative tissue. The structure of the keratinized layers
is very tightly knit but, unlike the stratum corneum, no exfoliation of cells occurs.
Given that it is a cornified epithelial structure, the chemical composition of the nail
plate is not remarkable, and there are many similarities to that of the hair (Table 3)
(43,44). Thus, the major components are highly folded keratin proteins (containing
many disulfide linkages) with small amounts (0.1–1.0%) of lipid, the latter presum-
ably located in the intercellular spaces. The principal plasticizer of the nail plate is
water, which is normally present at a concentration of 7–12%.
The nail plate comprises two major layers (the dorsal and intermediate layer)
with, possibly, a third layer adjacent to the nail bed (45,46). The dorsal nail plate is
harder and thinner than the intermediate plate, suggesting that there are differences
in the chemical composition of the two layers, which further suggests that applied
drugs may possess differing partitioning tendencies between the layers. The latter is
a particularly important consideration for the topical treatment of fungal infections
of the nail (onychomycoses) (47), and mechanisms for enhancing solubility and
diffusivity of drugs within these layers have been suggested (48).
III. DEVELOPMENT OF THE STRATUM CORNEUM

It is the structure of the stratum corneum that enables terrestrial animals to exist in
a nonaquatic environment without desiccation. The ability to control both the loss
of water and the influx of potentially harmful chemicals and microorganisms is the
result of the evolution of a unique mixture of protein and lipid materials that col-
lectively form this coherent membrane composed of morphologically distinct do-
mains. These domains are principally proteinaceous, the keratinocytes; or lipophilic,
the intercellular spaces. Because of the lack of nuclei in the keratinocytes, the stratum
corneum is often considered to be a dead tissue, but this is certainly a simplistic
Table 3 Amino Acid Composition of Nail, Hair, and Stratum Corneum
Amino acid Nail Hair Stratum corneum
Lysine 3.1a 2.5a 4.2a

Histidine 1.0 0.9 1.5
Arginine 6.4 6.5 3.9
Aspartic acid 7.0 5.4 7.9
Threonine 6.1 7.6 3.0
Serine 11.3 12.2 13.6
Glutamic acid 13.6 12.2 12.6
Proline 5.9 8.4 3.0
Glycine 7.9 5.8 24.5
Alanine 5.5 4.3 4.4
Valine 4.2 5.5 3.0
Methionine 0.7 0.5 1.1
Isoleucine 2.7 2.3 2.7
Leucine 8.3 6.1 6.9
Tyrosine 3.2 2.2 3.4
Phenylalanine 2.5 1.7 3.2
Half-cysteine 10.6 15.9 1.2
Sulfur 3.2%b 4.5%b 1.4%b
a
Expressed as residues per 100 residues.
b
Expressed as % dry weight.
Source: Ref. 43.
description and many events occur that indicate that ‘‘dead’’ does not necessarily
indicate a lack of response.
A. Epidermal Differentiation
The development of the stratum corneum from the keratinocytes of the basal layer
involves several steps of cell differentiation, which has resulted in a structure-based
classification of the layers above the basal layer (the stratum basale). Thus the cells
progress from the stratum basale through the stratum spinosum, the stratum granu-
losum, the stratum lucidium, to the stratum corneum (49). Cell turnover, from stratum
basale to stratum corneum, is estimated to be on the order of 21 days.
The exact mechanism whereby keratinocytes in the basal layer are stimulated
to initiate the differentiation process is not fully understood. Protein kinase C and
several keratinocyte-derived cytokines may play a regulatory role in the differenti-
ation process (50,51). Thus, for example, interleukin-1 (IL-1) stimulates the produc-
tion of other cytokines, in both autocrine and paracrine fashion, which act to induce
proliferation and chemotaxis (Fig. 7). These inducing cytokines include GM-GSF,
transforming growth factor-␣ (TGF-␣), nerve growth factor, amphiregulin, IL-6, and
IL-8. On the other hand, transforming growth factor-␤ (TGF-␤), the production of
which is also initiated by IL-1, suppresses keratinocyte growth, but stimulates ker-
atinocyte migration, the latter possibly as a result of modulation of hemidesmosomal
and desmosomal integrins. It has also been suggested that urokinase-type plasmin-
ogen activator (uPA) may activate growth factors and stimulate epidermal prolifer-
Figure 7 Cytokine involvement in keratinocyte proliferation. Secretion of interleukin-1

(IL-1) leads to the autocrine and paracrine stimulation of the release of other cytokines,
including interleukins-6 and -8 (IL-6 and IL-8), granulocyte–macrophage colony-stimulating
factor (GM-CSF) and transforming growth factor-␣ (TGF-␣), all of which induce epidermal
proliferation. Transforming growth factor-␤ (TGF-␤) and tumor necrosis factor-␣ (TNF-␣),
both of which suppress proliferation, are also released. TGF-␤ can induce keratinocyte mi-
gration. (From Ref. 51.)
ation (52). Generation and activation of the serine protease, plasmin, from plasmin-
ogen is induced by uPA and the activated plasmin instigates localized extracellular
proteolysis of cell surface adherent proteins and eventual disruption of the hemides-
mosomes. The increasing understanding of the role of cytokines in the maintenance
of epidermal homeostasis has stimulated research into the possibility of using such
compounds as active principles in cosmetic products (53).
The stratum spinosum (prickle cell layer), which lies immediately above the
basal layer, consists of several layers of cells that are connected to each other and
to the stratum basale cells by desmosomes and contain prominent keratin tonofila-
ments. The cells of the stratum spinosum have a larger cytoplasm than those of the
stratum basale. Within the cytoplasm are numerous organelles and filaments. It is
clear that the ␣-keratins of the stratum spinosum are somewhat different from those
found in the stratum basale (54), indicating that, although mitosis has ceased and a
phase of terminal differentiation has been initiated, the cell still maintains a capacity
to alter the transcriptional expression of its genes. In the outer cell layers of the
stratum spinosum, intracellular membrane-coating granules (100–300 nm in diam-
eter) appear within the cytosol, marking the transition between the stratum spinosum
and stratum granulosum.
Although further keratin differentiation occurs in the stratum granulosum
(55,56), new keratin synthesis stops. The most characteristic features of this layer
are the presence of many intracellular keratohyalin granules and membrane-coating
granules, the assembly of the latter appearing to take place in the endoplasmic re-
ticulum and Golgi regions (57,58). Within these granules lamellar subunits arranged
in parallel stacks are observed. These are believed to be the precursors of the inter-
cellular lipid lamellae of the stratum corneum (57,59). Also present in the lamellar
granules are hydrolytic enzymes, the most important of which is stratum corneum
chymotryptic enzyme (SCCE). SCCE is a serine protease that, because of its ability
to locate at desmosomal regions in the intercellular space, has been implicated in
the desquamation process (60–62). In the outermost layers of the stratum granulosum
the lamellar granules migrate to the apical plasma membrane where they fuse and
eventually extrude their contents into the intercellular space (57). At this stage in
the differentiation process, as a result of the release of selective lysing enzymes, the
keratinocytes lose their nuclei and other cytoplasmic organelles, become flattened
and compacted to form the stratum lucidum, which eventually forms the stratum
corneum. The extrusion of the contents of lamellar granules is a fundamental re-
quirement for the formation of the epidermal permeability barrier (63,64), and dis-
turbances in this process have been implicated in various dermatological disorders
(65).
The entire process of epidermal terminal differentiation is geared toward the
generation of the specific chemical morphology of the stratum corneum. Thus, the
end products of this process are the intracellular protein matrix and the intercellular
lipid lamellae.
B. Cornified Cell Envelope

The cornified cell envelope (about 15 nm thick and 10% mass of the stratum cor-
neum) is the outermost layer of a corneocyte, which consists mainly of tightly bun-
dled keratin filaments aligned parallel to the main face of the corneocyte. The en-
velope consists of both protein and lipid components. The protein envelope (⬃10
nm thick) is a covalent, cross-linking of several proteins as a result of actions by
sulfhydryl oxidases and transglutaminases; whereas the lipid envelope (⬃5 nm thick)
are lipid attached covalently to the protein envelope. Sulfhydryl oxidases and trans-
glutaminases lead to the formation of disulfide and isopeptide bonds, respectively
(66). It has been suggested that cross-linking by the creation of N-(␥-glutamyl) lysine
isodipeptide bonds formed by epidermal transglutaminases is a reaction possibly
mediated by cholesterol sulfate (67). The envelope lies adjacent to the interior surface
of the plasma membrane. In addition to the predominant protein loricrin, several
other envelope precursor proteins have been identified including cystatin-␣ (68),
cornifin-␣ (69), elafin, and filaggrin (70). The predominance of the structural proteins
in the cornified envelope are as follows: involucrin (65 kDa; 2–5%), loricin (26 kDa;
80%), small proline-rich proteins (a family of 11–14 closely related proteins, in-
cluding cornifins and pancornulins, 6–26 kDa; 3–5%), and cystalin A or keratolinin
(12 kDa; 2–5%). There are also a range of proteins with an expression of less than
1%, including elafin, profilaggrin, keratin intermediate filaments, desmoplakin I and
II, S100 proteins, and annexin I (also called lipocortin I) (66).
Formation of the envelope is believed to occur in two stages. In the first stage
soluble proteins, such as involucrin and cystatin-␣, form a scaffold to which other
insoluble precursors, including loricrin, are added in the latter stage. Thus, the cor-
nified envelope is formed by the sequential deposition of consecutively expressed
proteins starting with the fixation of involucrin as a scaffold on the intracellular
surface of the plasma membrane in a calcium- and phospholipid-dependent manner.
It is cross-linked to desmoplakin and envoplakin and also covalently bound to ␻-
hydroxyceramides. Other proteins then reinforce the envelope by attaching, including
loricin and small proline-rich proteins (66). The cross-linked protein complex of the
corneocyte envelope is very insoluble and chemically resistant. Cornified cell en-
velopes are also present in the hair follicle and nail matrix but, although morpho-
logically similar, the pattern and types of precursor are slightly different from those
of the epidermis (71).
It is currently proposed that the corneocyte protein envelope plays an important
role in the structural assembly of the intercellular lipid lamellae of the stratum cor-
neum. The work of Downing and colleagues (72–75) has demonstrated that the
corneocyte possesses a chemically bound lipid envelope comprised of N-␻-hydroxy-
ceramides that are ester-linked to the numerous glutamate side chains provided by
the ␤-sheet conformation of involucrin in the envelope protein matrix (74) (Fig. 8).
Recent molecular modeling of the human involucrin molecule has suggested that a
conventional ␣-helical conformation could also provide the requisite number of glu-
tamate side chains for ester linkage (76). The lipids of the cornified cell envelope
are resistant to extraction by chloroform–methanol mixtures, but can be extracted
following alkaline hydrolysis. This lipid envelope may provide the framework for
the generation of the intercellular lipid lamellae. Inhibition of the formation of N-␻-
hydroxyceramides may be achieved using aminobenzotriazole, an inhibitor of type
4 cytochrome P450. Behne et al. (77) demonstrated that, in the absence of N-␻-
hydroxyceramides, the stratum corneum intercellular lipid lamellae were abnormal
and permeability barrier function was disrupted. These data provide direct evidence
supporting the important roles of N-␻-hydroxyceramides in epidermal barrier ho-
meostasis and corneocyte lipid envelope formation.
Figure 8 The corneocyte protein envelope plays an important role in the structural assem-
bly of the intercellular lipid lamellae of the stratum corneum. The corneocyte possesses a
chemically bound lipid envelope comprised of N-␻-hydroxyceramides that are ester-linked to
the numerous glutamate side chains provided by the ␤-sheet conformation of involucrin in
the envelope protein matrix. (From Ref. 74.)
C. Stratum Corneum Proteins

It has been suggested that the keratinization process contributes to the barrier func-
tion of the stratum corneum by making the corneocytes practically insoluble in most
diffusing solutes (78). The poor stratum corneum protein solubility partly results
from the extensive cross-linking of both the cell envelope and intracellular proteins.
The majority of the intracellular protein in the stratum corneum is composed of
keratin filaments—which are cross-linked by intermolecular disulfide bridges (79,80)
—and the components of the cornified cell envelope. In the terminal stages of dif-
ferentiation, the keratinocytes contain keratin intermediate filaments (keratins 1 and
10, which are derived from keratins 5 and 14 present in basal keratinocytes) together
with several other proteins, including involucrin, loricrin, and profilaggrin (81–83).
Loricrin and involucrin are major components of the cornified cell envelope; whereas
profilaggrin is implicated in both the alignment of the keratin filaments (84) and
epidermal flexibility; the latter based on the water-holding capacity of the constituent
amino acids (85). Profilaggrin is a large, highly phosphorylated protein, consisting
of multiple (10–12) filaggrin units with N- and C-terminal domains, which first
appears in keratohyalin granules in the stratum granulosum. The profilaggrin mole-

cule is processed in a calcium-dependent manner by dephosphorylation and prote-
olysis into individual filaggrin molecules that serve to aggregate keratin filaments.
The term filaggrin (filament aggregating protein) was used to name the keratin matrix
proteins. The interaction between filaggrin and keratin is believed to be ionic (86).
Evidence of a role for profilaggrin as a calcium binder in epidermal differentiation
has also been presented (87). It is important to recognize that filaggrin does not exist
beyond the lower layers of the stratum corneum. It is completely proteolyzed into
constituent amino acids about 2–3 days after its formation from profilaggrin (88).
D. The Intercellular Lipids

The composition of the stratum corneum intercellular lipids is unique in biological
systems (Table 4). These lipids exist as a continuous lipid phase; occupying about
20% of the stratum corneum volume, and arranged in multiple lamellar structures.
A remarkable feature is the lack of phospholipids and the preponderance of choles-
terol (27%) and ceramides (41%), together with free fatty acids (9%), cholesteryl
esters (10%), and cholesteryl sulfate (2%) (89). This composition varies with body
site. There are distinct alterations in the distribution of lipid type during the course
of epidermal differentiation (Fig. 9). Phospholipids, which dominate in the basal
layer, are converted to glucosylceramides and, subsequently, to ceramides and free
fatty acids, and are virtually absent in the outer layers of the stratum corneum. Levels
of sphingolipids and free fatty acids increase in the latter stages of terminal
differentiation.
Eight classes of ceramides (designated ceramides 1–8) have been isolated and
identified in human stratum corneum (90,91). Ceramide classification is arbitrarily
based on polarity, with ceramide 1 being the least polar. Their structures are based
on sphingolipids (Fig. 10), and they contain long-chain bases, such as sphinganine
Table 4 Lipid Content of the Stratum Corneum

Intercellular Space
Lipid % (w/w) mol %
Cholesterol esters 10.0 7.5a

Cholesterol 26.9 33.4
Cholesterol sulfate 1.9 2.0
Total cholesterol derivatives 38.8 42.9
Ceramide 1 3.2 1.6
Ceramide 2 8.9 6.6
Ceramide 3 4.9 3.5
Ceramide 4 6.1 4.2
Ceramide 5 5.7 5.0
Ceramide 6 12.3 8.6
Total ceramides 41.1 29.5
Fatty acids 9.1 17.0a
Others 11.1 10.6b
a
Based on C16 alkyl chain.
b
Based on MW of 500.
Figure 9 There are distinct alterations in the distribution of lipid type during the course
of epidermal differentiation. Polar phospholipids, which dominate in the basal layer, are vir-
tually absent in the outer layers of the stratum corneum. Levels of ceramides and neutral
lipids increase in the latter stages of terminal differentiation.
and 4-hydroxysphinganine, N-acetylated by different fatty acids. Because of its

unique structure, ceramide 1, an acylceramide, may function as a stabilizer of the
intercellular lipid lamellae. This molecule consists of a sphingosine base, with an
amide-linked ␻-hydroxyacid and a nonhydroxyacid ester linked to the ␻-hydroxyl
group (75). It is possible that ceramide 1 acts as a ‘‘molecular rivet’’ in the inter-
cellular lipid lamellae of the stratum corneum. There is strong evidence to indicate
that the intercellular lipid lamellae are further stabilized by the chemical links be-
tween the long-chain ceramides and glutamate residues on the corneocyte protein
envelope (74). Extracellular calcium may also be involved in the formation and
Figure 10 Ceramides of the human stratum corneum intercellular space.

stabilization of the lamellae (92). Wertz (93) and Nemes and Steinhardt (66) have
suggested that the long-chain ceramides constituting the lipid envelope and attached
covalently to the protein envelope function in a ‘‘Velcro-like’’ fashion by interdigi-
tating with the intercellular lipids, allowing the structural integrity of the lipid la-
mellae to be maintained.
The functions of the individual ceramide type are not fully understood, and the
knowledge that has been accumulated is based mainly on examination of barrier
function, lipid content, and lipid distribution in diseased skin. For example, acylcer-
amides, isolated from acne comedones and the skin of patients with acne, contained
higher proportions of saturated and monounsaturated C16 and C18 fatty acids, and less
linoleate than those isolated from control subjects (94). Distribution of free fatty
acids showed a similar pattern. Reduction in dietary linoleate in experimental animals
results in epidermal hyperproliferation and impaired skin barrier function (95). In
patients with acne, epidermal hyperproliferation produces a keratinous follicular plug
that results in the formation of a comedone. These observations suggested a potential
role of ceramide 1 as an essential constituent of the skin barrier and, possibly, as a
mediator of epidermal proliferation (96). Also, the distribution and amount of cer-
amide types in psoriatic scales is different from that in normal skin (97), but the
significance of this anomaly is unknown. Similarly, ceramide content was reduced
in the stratum corneum of patients with atopic dermatitis (98).
In many biological membranes cholesterol acts as a stabilizer and reduces the
mobility of the alkyl chains. The exact function of cholesterol and cholesterol esters
in the stratum corneum intercellular lamellae are unknown, although it is likely that
cholesterol acts to reduce fluidity of the ceramide alkyl chains. Cholesterol and cer-
amides are present at almost equimolar proportions throughout the stratum corneum.
Norlén et al. (99) obtained values of 37%mol for ceramides and 32%mol for cho-
lesterol using human forearm skin (interestingly, the molar distribution of cholesterol
esters and free fatty acids was also similar at 15%mol and 16%mol, respectively).
This finding supports the suggestion that cholesterol and ceramide may interact on
a molecular one-to-one basis in the stratum corneum intercellular lamellae (100).
There is strong evidence that cholesterol interacts with phospholipids to form one-
to-one molar complexes involving hydrogen bonding of the 3-␤-hydroxyl of choles-
terol with the glyceryl oxygen at the 2 position of the phospholipid (101). It is
possible that a similar type of binding occurs between ceramides and cholesterol
within skin lipids.
The exact functions of cholesterol esters within the stratum corneum lamellae
are also elusive. It is theoretically possible that cholesterol esters may span adjacent
bilayers and serve as additional stabilizing moieties. Similarly, the role of fatty acids
is unclear. The recent work of Norlén and colleagues (102) has indicated that the
free fatty acids of the stratum corneum are composed entirely of saturated long-chain
acids, the majority of which are lignoceric acid (C24, 39%mol) and hexacosanoic
acid (C26, 23%mol). The authors extracted lipid from the deeper layers of the stratum
corneum and concluded that the sometimes reported presence of shorter-chain sat-
urated and unsaturated fatty acids in the outer layers of the stratum corneum is the
result of contamination from sebaceous gland lipid and the environment.
Overall, the intercellular lipid lamellae appear to be highly structured, very
stable, and constitute a highly effective barrier to chemical penetration and perme-
ation. However, the exact structure and physical state of the stratum corneum inter-
cellular lipid lamellae are not known. Forslind and colleagues (103,104) proposed a
domain mosaic model in which the long-chain ceramides are in a crystalline state,
whereas short-chain and unsaturated free fatty acids are in the liquid state. The model
proposes that large crystalline domains are surrounded by thin liquid crystalline chan-
nels and suggests that any water present in the region is associated with the liquid
crystalline phase or the corneocytes. Considerable information on lipid structure
within the stratum corneum has been generated by Bouwstra and colleagues (105–
108) using small-angle X-ray diffraction and transmission electron microscopic tech-
niques. These and earlier studies have shown that the lipid lamellae of the stratum
corneum are orientated parallel to the corneocyte surface and have repeat distances
of approximately 6.0–6.4 and 13.2–13.4 nm. Bouwstra et al. (107) have proposed
that the broad band represents regions where ceramide moieties are partly inter-
digitating, and the narrow band represents regions of full interdigitation.
In a more recent study on lipid packing (109), the Leiden group have evaluated
lipid organization of the stratum corneum using electron diffraction. Whereas wide-
angle X-ray diffraction techniques were able to demonstrate lattice spacings that were
consistent with orthorhombic (crystalline) packing of the lipids (reflections at 0.415
and 0.375 nm), they cannot confirm the presence or absence of hexagonal (gel)
packing, where only the 0.415 nm reflection occurs. On the other hand, electron
diffraction technology can distinguish between orthorhombic and hexagonal packing.
In this elegant study the authors found that, although the majority of lipids in the
intercellular space were present in the crystalline state, there were some lipids ex-
isting in the gel state that has a slightly looser hexagonal packing arrangement in
the outer layers of the stratum corneum. It was suggested that the existence of the
gel phase represents the influence of contaminating sebaceous lipid in this region,
but it is tempting to speculate that the alteration in lipid states in these outer layers
is somehow related to the process of desquamation. Fenske et al. (110) showed a
similar lateral packing in model membrane systems made of stratum corneum lipids.
E. Desquamation
The mechanisms underlying the desquamation of stratum corneum cells are not fully
understood. Suzuki et al. (111) suggested that, through the action of two types of
serine protease, the degradation of desmosomes leads to desquamation. Certainly
there has to be proteolysis of any intercellular adhesive structures between the ter-
minal keratinocytes. Egelrud’s group have suggested that desquamation may be reg-
ulated by the extent of activation of protease precursors and changes in the pH of
the stratum corneum intercellular space (112–114). Tape strips of the outer layers of
human stratum corneum contained precursors and active forms of both stratum cor-
neum chymotryptic enzyme and stratum corneum tryptic enzyme (113). Although
both enzymes possessed maximum activity at pH 8.0, considerable activity was re-
tained at pH 5.5 (the pH of the skin surface).
Other proteins that may play a role in desquamation include cathepsin D, a
protease active in the acid range (115), desquamin (116), and stratum corneum ge-
latinase (117).
Relevant to the discussion on desquamation is the role of corneodesmosomes
or corneosomes, a description for homogeneously electron-dense desmosomes in the
intercellular region. Much emphasis has been placed on the protein corneodesomosin,
which is located in the extracellular part of the desmosomes and adjacent parts of
the cornified cell envelope. It has been suggested that this protein is continuously
degraded, providing an explanation for the gradient of increased corneocyte cohe-
siveness from the skin surface toward deeper layers (118). It has been postulated
that cell cohesion is lost through proteolytic degradation, which may be inhibited by
calcium ions.
Scaly skin diseases may sometimes be a consequence of a disrupted desqua-
mation process. Desquamation is associated with a conversion of cholesterol sulfate
to cholesterol (119). Interestingly, X-linked ichthyosis, a scaly disease characterized
by a disrupted desquamation process, is identified with a lack of the enzyme cho-
lesterol sulfatase (120). More recent work (121) has shown that hyperkeratosis at-
tributable to desmosomes is associated with an increased content of cholesterol sul-
fate in patients with X-linked ichthyosis. It is apparent that cholesterol sulfate retards
desquamation by acting as a serine protease inhibitor.
IV. EPIDERMAL REPAIR MECHANISMS

A. The Effects of Hydration
Hydration of the stratum corneum can lead to profound changes in its barrier prop-
erties (122). The mechanisms involved in the hydration response are not full defined,
although it is likely that it is the result of a combination of water-induced swelling
of the corneocytes and some form of water-induced expansion of the intercellular
lipid lamellae. In the normal state, the stratum corneum holds between 15 and 20%
(dry weight) water, most of which appears to be associated with intracellular keratin
(123). Stratum corneum water can be increased up to about 400% (dry weight)
following excessive soaking. Swelling of corneocytes is possibly due to increased
uptake of water, which then interacts with keratin to expand the spatial orientation
of the protein. The observation that the corneocytes of the nail plate and hair do not
swell to the same extent as those of the stratum corneum following excessive hy-
dration indicates that the degree of interaction between water and keratin is a function
of the positioning and stability of disulfide bonds in the peptide (44). Thus, where
the ␣-helix keratin filaments are loosely packed and more flexible, as in the stratum
corneum keratinocytes, there is a greater ability to alter conformation to accommo-
date water.
The site of interaction of water with intercellular lipid lamellae is less clear.
Wide-angle X-ray diffraction studies indicated that no bilayer swelling occurred with
hydration (124). This suggests that water molecules are not absorbed between the
lamellar regions. Bouwstra’s group has used freeze–fracture electron microscopy and
other techniques to examine fully hydrated human stratum corneum (125,126). By
using an elegant technique of sample preparation (126), which involved multiple
folding of the sample, visualization of the interface between the stratum corneum
and the hydration medium was possible. As would be expected for fully hydrated
stratum corneum the observed corneocytes were swollen, with pools of water ap-
parently displacing and separating keratin filaments. Distinct water domains (pools)
were observed in the intercellular spaces of hydrated stratum corneum. Also observed
in the intercellular spaces were vesicle-like lipid structures, which suggested that
lipids were extracted, presumably from the lamellae.
It has been suggested that the vesicle-like structures observed by van Hal et
al. (126) may depict the lacunae that result from desmosomal degradation (127).
Lacunae are discontinuous microdomains located in the extracellular space in the
middle to outer layers of the stratum corneum (128). During hydration the lacunae
provide an obvious site for water pooling and, during prolonged exposure to water,
lateral expansion of the lacunae occurs through polar head regions of the intercellular
lipids (Fig. 11) (127). Although the expansion of the individual lacunae may lead to
a continuous lacunar system, this process does not appear to disrupt the lipid la-
mellae. Menon and Elias (127) have proposed that the continuous lacunal system
may represent a putative ‘‘poor’’ pathway through the stratum corneum.
It is well recognized that natural moisturizing factor (NMF) can make up to
10% of the corneocyte dry weight and, as humectants, these materials can sorb water
extensively. There appears to be an absence of NMF in severe, dry flaking skin in
both psoriasis and ichthyosis vulgaris. Rawlings et al. (88) have pointed out that the
amino acids to which filaggrin is proteolyzed are themselves precursors for the nat-
ural moisturizing factor. Glutamine is converted to the potent humectant, pyrrolidone
carboxylic acid, a major component of NMF, whereas histidine is converted to uro-
canic acid. Interestingly, filaggrin is converted to NMF only when the water activity
is between 0.70 and 0.95, filaggrin being stable at higher water activities and pro-
teolysis being impeded by low water activity. Hence, under occlusive conditions the
stratum corneum NMF level decreases to close to zero, and all corneocytes contain
filaggrin. The result of this homeostatic mechanism is that the skin has prevented
itself from being ‘‘overhydrated.’’
In conclusion, the current observations suggest that stratum corneum hydration
does not lead to an overall decrease in intercellular lipid order and only small
amounts of water are present in the intercellular polar head group regions (89).
Therefore, it is tempting to revisit a possible mechanism by which hydration pro-
motes percutaneous absorption, which has been raised in an earlier review (122). In
that model, swelling of the keratin is akin to the ‘‘bricks’’ becoming swollen in the
‘‘bricks-and-mortar’’ model of the stratum corneum, with a loosening of the inter-
cellular lipid ‘‘mortar.’’ The overall effect should be an increase in the mobility of
the chains and in permeability, without an effect on the lipid ordering.
B. Chemical Damage
When the stratum corneum is perturbed, several localized biochemical events occur
that result in rapid reconstitution of barrier function (64,129–137). Thus, in extreme
cases of stratum corneum damage, such as acetone-induced delipidation (129–131)
or tape-stripping (137), there appears to be a biphasic pattern of recovery: a rapid
phase of repair, followed by a slower phase of normalization. The initial rapid phase
of barrier recovery involves the expeditious secretion of preformed lamellar bodies
from the granular cells into the intercellular space (64), an increase in epidermal
cholesterol and fatty acid synthesis (134,135), and accelerated production and secre-
tion, into the intercellular space, of new lamellar bodies. The subsequent and slower
phase of barrier repair involves an increase in ceramide synthesis (135) and an in-
crease in DNA synthesis (136) leading to epidermal hyperplasia. A similar response
to barrier perturbation occurs following treatment of the skin with sodium dodecyl
sulfate (SDS) (137), but the magnitude of the response depends on the severity of
Figure 11 During hydration the lacunae formed by degrading desmosomes provide an

obvious site for water pooling and, during prolonged exposure to water, lateral expansion of
the lacunae occurs through polar-head regions of the intercellular lipids. Although the expan-
sion of the individual lacunae may lead to a continuous lacunar system, this process does not
appear to disrupt the lipid lamellae. Menon and Elias (127) have proposed that the continuous
lacunal system may represent a putative aqueous ‘‘pore’’ pathway through the stratum
corneum.
the induced perturbation. It is remarkable that the initial perturbation, which occurs
in the outermost layers of the stratum corneum, can rapidly stimulate biochemical
events in the stratum granulosum and lower levels of the epidermis.
Although the exact mechanisms stimulating these events are unknown, there is
some indication that a change in the rate of transepidermal water loss (TEWL) in-
duced by barrier alterations, may play a role (131). This increase in TEWL may lead
to focal changes in the concentration of certain ions in the outer epidermis. In the
normal state, the epidermis possesses a Ca2⫹ ion gradient such that there is more
Ca2⫹ in the outer layers than the inner (138). Following barrier disruption the Ca2⫹
gradient is lost. The presence of higher levels of intracellular Ca2⫹ in the outer
epidermis is believed to block lamellar body secretion (139,140), and reduced levels
will stimulate secretion. In addition, K⫹ may play a role in this homeostatic mech-
anism and may also influence barrier repair independently of Ca2⫹ (141). Thus,
although there are still many uncertainties concerning the biochemistry of barrier
repair, there is much evidence that suggests the role of ion concentration and the
induction of lipid-producing enzymes; such as 3-hydroxy-3-methylglutaryl coenzyme
A and serine palmitoyl transferase (142).
Perturbation of barrier function sometimes, but not always, also induces an
inflammatory response that results in irritation. It is important to appreciate that
irritation is used to describe skin reactions that can range from a mild and transient
erythema or itch, to serious vesiculation (see Chaps. 12 and 13). Whereas the insults
of solvent delipidation and tape-stripping of the stratum corneum result in barrier
repair and epidermal hyperplasia, they do not necessarily lead to an irritant reaction.
On the other hand, application of SDS almost always results in an irritant response
(143,144). Although solvent delipidation and tape-stripping of the stratum corneum
both physically remove the intercellular lipid lamellae, which results in considerable
increases in TEWL, SDS intercalates with the lamellae and increases fluidity in this
region (145), resulting in an increase in TEWL. Furthermore, although other surface-
active agents, such as sodium laurate and polysorbates, can increase TEWL to levels
similar to SDS, the resultant irritation is much less and, in some cases, not signifi-
cantly different from untreated skin (146). It follows that irritation subsequent to
exposure to SDS must be a result of factors other than an increase in water transport
and the stimulation of lipogenesis.
That surface-active agents can cause skin irritation is well established and has
been so for many years (147). Also, whereas ionic surfactants can cause severe
irritation, nonionic surfactants are considered virtually nonirritant in normal use
(148,149). Thus, much of the research on surfactant-induced skin irritation has in-
volved studies on SDS. The collective data suggest that SDS can interact with both
lipid and protein structures in the stratum corneum. Interaction with lipids will in-
crease lipid fluidity and thereby enhance skin permeability. This alone, however,
apart from increasing its own permeation, will not account for the irritation caused
by SDS. Although SDS can penetrate into the corneocyte and interact with the protein
structure such that ␣-keratin is uncoiled (150), it is difficult to relate this aspect to
an irritant response. A more likely explanation for the irritation induced by SDS is
its capacity to stimulate keratinocyte production of inflammatory mediators such as
IL-1 and PGE2 (151). Whether this induction is secondary to some interaction be-
tween SDS and the corneocyte cell membrane is unknown.
There is a range of mechanisms by which solvents may affect skin permeability,

as proposed by Menon et al. (152). Suhonen et al. (89) recently reviewed chemical
enhancement in terms of stratum corneum alterations. They suggested that both hy-
dration and temperature effects by transitions were involving the hydrocarbon chains
of the stratum corneum lipid components. They suggested that enhancer actions could
be located either in the lipid region near the polar head group or between the hy-
drophobic tails. Extraction of lipids would also lead to an increase in disorder be-
cause there is more space (free volume) in which the hydrocarbon chains can move
(Fig. 12). Suhonen et al. (89) suggest that ethanol may act by displacing bound water
molecules at the lipid headgroup–membrane interface region with a resulting in-
creased interdigitation of the hydrocarbon chains. Ethanol extracts lipids from the
stratum corneum only at high concentrations, with a resulting greater free volume
for the lipid chains, as shown by infrared spectroscopy.
Keratolytics are becoming increasing used to promote permeability of the hu-
man epidermis, especially the nail plate, to topical agents. Walters et al. (153) have
previously reported that agents that act as accelerants on the epidermis may not
necessarily do so with the nail plate. Many enhancers have an effect on intercellular
lipids, which constitute less than 1% of the nail weight. Quintanar-Guerrero et al.
(154) suggest that keratolytic agents may facilitate antimycotic penetration through
the nail plate by pore formation. This effect, observed using scanning electron mi-
croscopy, was most pronounced for papain, followed by salicylic acid, and then urea.
Effects on deeper regions of the nail required a combination of papain and salicylic
acid. In general, keratolytics have a limited effect on skin permeability, emphasizing
the role of the intercellular lipids as a barrier to skin penetration by solutes. It has
been suggested that other agents that affect stratum corneum protein structures (e.g.,
propylene glycol, ethanol, and dimethyl sulfoxide), create a reversible conformation
change in the keratin protein from an ␣-helix to a ␤-sheet as a consequence of a
replacement of water that is bound to polar protein side chains. Dithiothreitol, a
disulfide-reducing agent, has also been suggested to enhance hydrophilic solute pen-
etration by an altered protein conformation to the ␤-sheet as a consequence of the
appearance of free thiols.
C. Biochemical Abnormalities
There are a large number of diseases that can affect epidermal barrier function, and
it is beyond the scope of this chapter to consider any of these in great depth. Some
of the diseases affect the formation of the corneocyte (‘‘broken brick syndrome’’)
whereas other affect the intercellular lipid (‘‘weak mortar syndrome’’). For instance,
Nemes and Steinhardt (66) refer to more than ten different diseases involving genes
that encode keratin intermediate filaments, including Unna–Thost disease and tylosis.
Other genetic diseases are related to defects in the genes associated with the structural
proteins of the cornified envelope or transglutaminases. For instance, a genetic defect
in TGM1, the gene that encodes the transglutaminase I enzyme, leads to a life-
threatening disease: lamellar ichthyosis (66).
A reduction of the effective intercellular lipid barrier properties can lead to
deficiencies ranging from dry skin (depletion of lipids owing to excessive use of
detergents), to hyperproliferation and abnormal scaling. Causes include essential fatty
acid deficiency, abnormal intercellular deposition of various lipids, accumulation of
Figure 12 Creation of free space (free volume) in the intercellular lipid lamellae of the
stratum corneum allows a greater mobility of the hydrocarbon chains that may result in
enhanced diffusivity. This could be induced by enhancer shape [e.g., oleic acid and lauroca-
pram (Azone)] or by electrostatic headgroup interactions.
cholesterol sulfate in X-linked ichthyosis, genetic defects of lipid metabolism (e.g.,

Refsum’s disease and Sjogren–Larsson syndrome resulting from phytanoyl-CoA hy-
droxylase and fatty aldehyde deficiencies, respectively).
V. CONCLUDING REMARKS
The aim of this chapter has been to introduce the reader to the basic morphology
and function of skin, to outline the stages in the development of the barrier layer, to
define the chemical makeup of the stratum corneum, and to illustrate repair mech-
anisms following barrier disruption. What is evident is that the skin is more than
another simple biological barrier membrane, into and through which therapeutic
agents can be delivered. Rather, the skin should be viewed as an extremely selective
semipermeable membrane overlying a powerful immune system ready to react to
any given insult. Although it is difficult to resolve the latter, without pharmacological
intervention, the problem of membrane permeability may be approached by several
diverse strategies. It is important to fully understand the mechanisms of percutaneous
absorption and the means by which skin permeation is quantified. The remaining
chapters in this book develop the concept of the skin, both as a therapeutic target
and as a portal for drug delivery to the systemic sites. Formulation development and
scale-up are addressed together with a comprehensive evaluation of bioequivalence
for dermatological and transdermal dosage forms. Finally, consideration is given to
adverse cutaneous reactions and safety aspects of formulations applied to the skin.
REFERENCES
1. Nevill AM. (1994) The need to scale for differences in body size and mass: and
explanation of Klieber’s 0.75 mass exponent. Am Physiol Soc pp 2870–2873.
2. Lavker RM, Sun T. (1982) Heterogeneity in epidermal basal keratinocytes: morpho-
logical and functional correlations. Science 215:1239–1241.
3. Borradori L, Sonnenberg A. (1999) Structure and function of hemidesmosomes: more
than simple adhesion complexes. J Invest Dermatol 112:411–418.
4. Sawamura D, Li K, Chu M–L, Uitto J. (1991) Human bullous pemphigoid antigen
(BPAG1): amino acid sequences deduced from cloned cDNAs predict biologically im-
portant peptide segments and protein domains. J Biol Chem 266:17784–17790.
5. Li K, Tamai K, Tan EML, Uitto J. (1993) Cloning of type XVII collagen. Comple-
mentary and genomic DNA sequences of mouse 180-kDa bullous pemphigoid antigen
(BPAG2) predict an interrupted collagenous domain, a transmembrane segment, and
unusual features in the 5⬘-end of the gene and the 3⬘-untranslated region of mRNA. J
Biol Chem 268:8825–8834.
6. Stepp MA, Spurr–Michaud S, Tisdale A, Elwell J, Gipson IK. (1990) ␣6␤4 Integrin
heterodimer is a component of hemidesmosomes. Proc Natl Acad Sci USA 87:8970–
8974.
7. Burgeson RE, Christiano AM. (1997) The dermal–epidermal junction. Curr Opin Cell
Biol 9:651–658.
8. Dowling J, Yu QC, Fuchs E. (1996) ␤4 Integrin is required for hemidesmosome for-
mation, cell adhesion and cell survival. J Cell Biol 134:559–572.
9. Kaiser HW, Ness W, Jungblut I, Briggaman RA, Kreysel HW, O’Keefe EJ. (1993)
Adherens junctions: demonstration in human epidermis. J Invest Dermatol 100:180–
185.
10. Kaiser HW, Ness W, Offers M, O’Keefe EJ, Kreysel HW. (1993) Talin: adherens
junction protein is localized at the epidermal–dermal interface in skin. J Invest Der-
matol 101:789–793.
11. Amagai M. (1995) Adhesion molecules. I: Keratinocyte–keratinocyte interactions; cad-
herins and pemphigus. J Invest Dermatol 104:146–152.
12. Peltonen S, Hentula M, Hägg P, Ylä–Outinen H, Tuukkanen J, Lakkakorpi J, Rehn M,
Pihlajaniemi T, Peltonen J. (1999) A novel component of epidermal cell–matrix and
cell–cell contacts: transmembrane protein type XII collagen. J Invest Dermatol 113:
635–642.
13. O’Keefe E, Briggaman RA, Herman B. (1987) Calcium-induced assembly of adherens
junctions in keratinocytes. J Cell Biol 105:807–817.
14. Haftek M, Hansen MU, Kaiser HW, Kreysel HW, Schmitt D. (1996) Interkeratinocyte
adherens junctions: immunocytochemical visualization of cell–cell junctional struc-
tures, distinct from desmosomes, in human epidermis. J Invest Dermatol 106:498–504.
15. Buxton RS, Magee AI. (1992) Structure and interactions of desmosomal and other
cadherins. Semin Cell Biol 3:157–167.
16. Tsukita S, Tsukita S, Nagafuchi A. (1990) The undercoat of adherence junctions: a key
specialized structure in organogenesis and carcinogenesis. Cell Struct Funct 15:7–12.
17. Petersen OH. (1999) Waves of excitement: calcium signals inside cells. Biologist 46:
227–230.
18. Kitajima Y, Aoyama Y, Seishima M. (1999) Transmembrane signaling for adhesive
regulation of desmosomes and hemidesmosomes, and for cell–cell detachment induced
by pemphigus IgG in cultured keratinocytes: involvement of protein kinase C. J Invest
Dermatol Symp Proc 4:137–144.
19. Yuspa SH. (1994) The pathogenesis of squamous cell cancer: lessons learned from
studies of skin carcinogenesis. Cancer Res 54:1178–1189.
20. Warner RR. (2000) The distribution and function of physiological elements in skin. In:
Lodén M, Maibach HI, eds. Dry Skin and Moisturizers, Boca Raton: FL, CRC Press,
pp 89–108.
21. Potten CS, Allen TD. (1976) A model implicating the Langerhans cell in keratinocyte
proliferation control. Differentiation 5:43–47.
22. Aiba S, Katz SI. (1990) Phenotypic and functional characteristics of in vivo activated
Langerhans cells. J Immunol 145:2791–2796.
23. Teunissen MBM. (1992) Dynamic nature and function of epidermal Langerhans cells
in vivo and in vitro: a review, with emphasis on human Langerhans cells. Histochem
J 24:697–716.
24. Nakamura K, Saitoh A, Yasaka N, Furue M, Tamaki K. (1999) Molecular mechanisms
involved in the migration of epidermal dendritic cells in the skin. J Invest Dermatol
Symp Proc 4:169–172.
25. Staquet M–J, Levarlet B, Dezutter–Dambuyant C, Schmitt D. (1992) Human epidermal
Langerhans cells express ␤1 integrins that mediate their adhesion to laminin and fibro-
nectin. J Invest Dermatol 99:12S–14S.
26. Kimber I, Cumberbatch M. (1992) Stimulation of Langerhans cell migration by tumour
necrosis factor ␣ (TNF-␣). J Invest Dermatol 99:48S–50S.
27. Jakob T, Udey MC. (1998) Regulation of E–cadherin-mediated adhesion in Langerhans
cell-like dendritic cells by inflammatory mediators that mobilize Langerhans cells in
vivo. J Immunol 160:4067–4073.
28. Jimbow K, Fitzpatrick TB, Quevedo WC. (1986) Formation, chemical composition and
functions of melanin pigments in mammals. In: Matolsky A, ed. Biology of the Integ-
ument. Vol. 2. New York: Springer-Verlag, pp 278–296.
29. Jimbow K, Lee SK, King MG, Hara H, Chen H, Dakour J, Marusyk H. (1993) Melanin
pigments and melanosomal proteins as differentiation markers unique to normal and
neoplastic melanocytes. J Invest Dermatol 100:259S–268S.
30. Prota G. (1996) Melanins and melanogenesis. Cosmet Toiletr 111:43–51.
31. Tuma MC, Gelfand VI. (1999) Molecular mechanisms of pigment transport in mela-
nophores. Pigment Cell Res 12:283–294.
32. Ortonne J–P, Prota G. (1993) Hair melanins and hair color: ultrastructural and bio-
chemical aspects. J Invest Dermatol 101:82S–89S.
33. Valyi–Nagy IT, Hirka G, Jensen PJ, Shih I–M, Juhasz I, Herlyn M. (1993) Undiffer-
entiated keratinocytes control growth, morphology, and antigen expression of normal
melanocytes through cell–cell contact. Lab Invest 69:152–159.
34. Hearing VJ (1999). Biochemical control of melanogenesis and melanosomal organi-
zation. J Invest Dermatol Symp Proc 4:24–28.
35. Chakraborty A, Slominski A, Ermak G, Hwang J, Pawelek J. (1995) Ultraviolet B and
melanocyte-stimulating hormone (MSH) stimulate mRNA production for ␣MSH re-
ceptors and proopiomelanocortin-derived peptides in mouse melanoma cells and trans-
formed keratinocytes. J Invest Dermatol 105:655–659.
36. Suzuki I, Tada A, Ollmann MM, Barsh GS, Im S, Lamoreux ML, Hearing VJ, Nordlund
JJ, Abdel–Malek ZA. (1997) Agouti signaling protein inhibits melanogenesis and the
response of human melanocytes to ␣-melanotropin. J Invest Dermatol 108:838–842.
37. Grando SA, Kist DA, Qi M, Dahl MV. (1993) Human keratinocytes synthesize, secrete,
and degrade acetylcholine. J Invest Dermatol 101:32–36.
38. Tachibana T. (1995) The Merkel cell: recent findings and unresolved problems. Arch
Histol Cytol 58:379–396.
39. Halata Z. (1993) Sensory innervation of the hairy skin (light- and electronmicroscope
study). J Invest Dermatol 101:75S–81S.
40. Cross SE, Roberts MS. (1993) Subcutaneous absorption kinetics of interferon and other
solutes. J Pharm Pharmacol 45:606–609.
41. Szuba A, Rockson SG. (1997) Lymphedema: anatomy, physiology and pathogenesis.
Vasc Med 2:321–326.
42. Cameli N, Picardo M, Perrin C. (1994) Expression of integrins in human nail matrix.
Br J Dermatol 130:583–588.
43. Baden HP, Goldsmith LA, Fleming B. (1973) A comparative study of the physico-
chemical properties of human keratinized tissues. Biochim Biophys Acta 322:269–278.
44. Gniadecka M, Nielsen OF, Christensen DH, Wulf HC. (1998) Structure of water, pro-
teins, and lipids in intact human skin, hair, and nail. J Invest Dermatol 110:393–398.
45. Dawber RPR. (1980) The ultrastructure and growth of human nails. Arch Dermatol
Res 269:197–204.
46. Jemec GBE, Serup J. (1989) Ultrasound structure of the human nail plate. Arch Der-
matol 125:643–646.
47. Kobayashi Y, Miyamoto M, Sugibayashi K, Morimoto Y. (1999) Drug permeation
through the three layers of the human nail plate. J Pharm Pharmacol 51:271–278.
48. Wang JCT, Sun Y. (1999) Human nail and its topical treatment: brief review of current
research and development of topical antifungal drug delivery for onychomycosis treat-
ment. J Cosmet Sci 50:71–75.
49. Eckert EL. (1989) Structure, function, and differentiation of the keratinocyte. Physiol
Rev 69:1316–1346.
50. Dinarello CA. (2000) Interleukin-1 receptors and signal transduction. In: Kydonieus
AF, Wille JJ, eds. Biochemical Modulation of Skin Reactions. Boca Raton, FL: CRC
Press, pp 173–187.
51. Kondo S. (1999) The roles of keratinocyte-derived cytokines in the epidermis and their
possible responses to UVA-irradiation. J Invest Dermatol Symp Proc 4:177–183.
52. Jensen PJ, Lavker RM. (1999) Urokinase is a positive regulator of epidermal prolif-
eration in vivo. J Invest Dermatol 112:240–244.
53. Petersen R–D, Reinhold W, Tyborczyk J. (1997) Cytokines in cosmetology. Cosmet
Toiletr 112:65–70.
54. Skerrow D, Skerrow CJ. (1983) Tonofilament differentiation in human epidermis: iso-
lation and polypeptide chain composition of keratinocyte subpopulations. Exp Cell Res
143:27–35.
55. Baden HP. (1979) Keratinization in the epidermis. Pharm Ther 7:393–411.
56. Hunter I, Skerrow D. (1989) Relationship of the final ␣-keratin compositions in dy-
skeratoses to those along the normal keratinization pathway: aetiological and diagnostic
significance. Br J Dermatol 120:363–369.
57. Landmann L. (1988) The epidermal permeability barrier. Anat Embryol 178:1–13.
58. Madison KC, Howard EJ. (1996) Ceramides are transported through the Golgi appa-
ratus in human keratinocytes in vitro. J Invest Dermatol 106:1030–1035.
59. Wertz PW, Downing DT. (1982) Glycolipids in mammalian epidermis: structure and
function in the water barrier. Science 217:1261–1262.
60. Egelrud T, Lundström A. (1991) A chymotrypsin-like proteinase that may be involved
in desquamation in plantar stratum corneum. Arch Dermatol Res 283:108–112.
61. Egelrud T, Régnier M, Sondell B, Shroot B, Schmidt R. (1993) Expression of stratum
corneum chymotryptic enzyme in reconstructed human epidermis and its suppression
by retinoic acid. Acta Derm Venereol 73:181–184.
62. Sondell B, Thornell L–E, Egelrud T. (1995) Evidence that stratum corneum chymo-
tryptic enzyme is transported to the stratum corneum extracellular space via lamellar
bodies. J Invest Dermatol 104:819–823.
63. Wertz PW, Downing DT, Freinkel RK, Traczyk TN. (1984) Sphingolipids of the stratum
corneum and lamellar granules of fetal rat epidermis. J Invest Dermatol 83:193–195.
64. Menon GK, Feingold KR, Elias PM. (1992) Lamellar body secretory response to barrier
disruption. J Invest Dermatol 98:279–289.
65. Fartasch M, Bassukas ID, Diepgen TL. (1992) Disturbed extruding mechanism of la-
mellar bodies in dry non-eczematous skin of atopics. Br J Dermatol 127:221–227.
66. Nemes Z, Steinert PM. (1999) Bricks and mortar of the epidermal barrier. Exp Mol
Med 31:5–19.
67. Kawabe S, Ikuta T, Ohba M, Chida K, Ueda E, Yamanishi K, Kuroki T. (1998) Cho-
lesterol sulfate activates transcription of transglutaminase 1 gene in normal human
keratinocytes. J Invest Dermatol 111:1098–1102.
68. Takahashi M, Tezuka T, Katunuma N. (1992) Phosphorylated cystatin alpha is a natural
substrate of epidermal transglutaminase for formation of skin cornified envelope. FEBS
Lett 308:79–82.
69. Fujimoto W, Nakanishi G, Arata J, Jetten AM. (1997) Differential expression of human
cornifin ␣ and ␤ in squamous differentiating epithelial tissues and several skin lesions.
J Invest Dermatol 108:200–204.
70. Steinert PM, Marekov LN. (1995) The proteins elafin, filaggrin, keratin intermediate
filaments, loricrin, and small proline-rich proteins 1 and 2 are isodipeptide cross-linked
components of the human epidermal cornified cell envelope. J Biol Chem 270:17702–
17711.
71. Baden HP, Kvedar JC. (1993) Epithelial cornified envelope precursors are in the hair
follicle and nail. J Invest Dermatol 101:72S–74S.
72. Swartzendruber DC, Wertz PW, Madison KC, Downing DT. (1987) Evidence that the
corneocyte has a chemically bound lipid envelope. J Invest Dermatol 88:709–713.
73. Downing DT. (1992) Lipid and protein structures in the permeability barrier of mam-
malian epidermis. J Lipid Res 33:301–313.
74. Lazo ND, Meine JG, Downing DT. (1995) Lipids are covalently attached to rigid
corneocyte protein envelopes existing predominantly as ␤-sheets: a solid-state nuclear
magnetic resonance study. J Invest Dermatol 105:296–300.
75. Wertz PW, Downing DT. (1989) Stratum corneum: biological and biochemical consid-
erations. In: Hadgraft J, Guy RH, eds. Transdermal Drug Delivery. New York: Marcel
Dekker, pp 1–22.
76. Downing DT, Lazo ND. (2000) Lipid and protein structures in the permeability barrier.
In: Lodén M, Maibach HI, eds. Dry Skin and Moisturizers. Boca Raton, FL: CRC
Press, pp 39–44.
77. Behne M, Uchida Y, Seki T, Ortiz de Montellano P, Elias PM, Holleran WM. (2000)
omega-Hydroxyceramides are required for corneocyte lipid envelope (CLE) formation
and normal epidermal permeability barrier function. J Invest Dermatol 114:185–192.
78. Kurihara–Bergstrom T, Good WR. (1987) Skin development and permeability. J Con-
trolled Release 6:51–58.
79. Sun T–T, Green H. (1978) Keratin filaments of cultured human epidermal cells. J Biol
Chem 253:2053–2060.
80. Steinert PM, North ACT, Parry DAD. (1994) Structural features of keratin intermediate
filaments. J Invest Dermatol 103:19S–24S.
81. Steven AC, Bisher ME, Roop DR, Steinert PM. (1990) Biosynthetic pathways of fi-
laggrin and loricrin—two major proteins expressed by terminally differentiated mouse
keratinocytes. J Struct Biol 104:150–162.
82. Yaffe MB, Murthy S, Eckert RL. (1993) Evidence that involucrin is a covalently linked
constituent of the highly purified cultured keratinocyte cornified envelopes. J Invest
Dermatol 100:3–9.
83. Robinson NA, LaCelle PT, Eckert RL. (1996) Involucrin is a covalently crosslinked
constituent of highy purified epidermal corneocytes: evidence for a common pattern of
involucrin crosslinking in vivo and in vitro. J Invest Dermatol 107:101–107.
84. Dale BA, Holbrook KA, Steinert PM. (1978) Assembly of stratum corneum basic
protein and keratin filaments in macrofibrils. Nature 276:729–731.
85. Scott IR, Harding CR, Barrett JG. (1982) Histidine-rich protein of the keratohyalin
granules: source of free amino acids, urocanic acid and pyrolidone carboxylic acid in
the stratum corneum. Biochim Biophys Acta 719:110–117.
86. Mack JW, Stevens AC, Steinert PM. (1993) The mechanism of interaction of filaggrin
with intermediate filaments. The ionic zipper hypothesis. J Mol Biol 232:50–66.
87. Markova NG, Marekov LN, Chipev CC, Gan S–Q, Idler WW, Steinert PM. (1993)
Profilaggrin is a major epidermal calcium-binding protein. Mol Cell Biol 13:613–625.
88. Rawlings AV, Scott IR, Harding CR, Bowser PA. (1994) Stratum corneum moisturis-
ation at the molecular level. J Invest Dermatol 103:731–740.
89. Suhonen TM, Bouwstra JA, Urti A. (1999) Chemical enhancement of percutaneous
absorption in relation to stratum corneum structural alterations. J Controlled Release
59:149–161.
90. Wertz WP, Miethke MC, Long SA, Strauss JS, Downing DT. (1985) The composition
of the ceramides from human stratum corneum and from comedones. J Invest Dermatol
84:410–412.
91. Stewart ME, Downing DT. (1999) A new 6-hydroxy-4-sphingenine-containing cer-
amide in human skin. J Lipid Res 40:1434.
92. Abraham W, Wertz PW, Landmann L, Downing DT. (1987) Stratum corneum lipid
liposomes: calcium-induced transformation into lamellar sheets. J Invest Dermatol 88:
212–214.
93. Wertz P. (1997) Integral lipids of hair and stratum corneum. Experientia 78:227–237.
94. Perisho K, Wertz PW, Madison KC, Stewart ME, Downing DT. (1988) Fatty acids of
acylceramides from comedones and from the skin surface of acne patients and control
subjects. J Invest Dermatol 90:350–353.
95. Holman RT. (1968) Essential fatty acid deficiency. Prog Chem Fats Other Lipids 9:
275–348.
96. Jung E, Griner RD, Mann–Blakeney R, Bollag WB. (1998) A potential role for cer-
amide in the regulation of mouse epidermal keratinocyte proliferation and differentia-
tion. J Invest Dermatol 110:318–323.
97. Motta S, Monti M, Sesana S, Caputo R, Carelli S, Ghidoni R. (1993) Ceramide com-
position of the psoriatic scale. Biochim Biophys Acta 1182:147–151.
98. Imokawa G, Kuno H, Kawai M. (1991) Stratum corneum lipids serve as a bound-water
modulator. J Invest Dermatol 96:845–851.
99. Norlén L, Nicander I, Rozell BL, Ollmar S, Forslind B. (1999) Inter- and intra-indi-
vidual differences in human stratum corneum lipid content related to physical param-
eters of skin barrier function in vivo. J Invest Dermatol 112:72–77.
100. Brain KR, Walters KA. (1993) Molecular modeling of skin permeation enhancement
by chemical agents. In: Walters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration
Enhancement. New York: Marcel Dekker, pp 389–416.
101. Finean JB. (1990) Interaction between cholesterol and phospholipid in hydrated bilay-
ers. Chem Phys Lipids 54:147–156.
102. Norlén L, Nicander I, Lundsjö A, Cronholm T, Forslind B. (1998) A new HPLC-based
method for the quantitative analysis of inner stratum corneum lipids with special ref-
erence to the free fatty acid fraction. Arch Dermatol Res 290:508–516.
103. Forslind B. (1994) A domain mosaic model of the skin barrier. Acta Derm Venereol
(Stockh) 74:1–6.
104. Engström S, Forslind B, Engblom J. (1996) Lipid polymorphism—a key to understand-
ing of skin penetration. In: Brain KR, James VJ, Walters KA, eds. Prediction of Per-
cutaneous Penetration. Vol. 4b. Cardiff: STS Publishing, pp 163–166.
105. Bouwstra JA, Cheng K, Gooris GS, Weerheim A, Ponec M. (1996) The role of cer-
amides 1 and 2 in the stratum corneum lipid organisation. Biochim Biophys Acta 1300:
177–186.
106. Bouwstra JA, Gooris GS, Cheng K, Weerheim A, Bras W, Ponec M. (1996) Phase
behavior of isolated skin lipids. J Lipid Res 37:999–1011.
107. Bouwstra JA, Gooris GS, Dubbelaar FER, Weerheim A, IJzerman AP, Ponec M. (1998)
Role of ceramide 1 in the molecular organization of the stratum corneum lipids. J Lipid
Res 39:186–196.
108. van der Meulen J, van den Bergh BAI, Mulder AA, Mommaas AM, Bouwstra JA,
Koerten HK. (1996) The use of vibratome sections for the ruthenium tetroxide protocol:
a key for optimal visualization of epidermal lipid bilayers of the entire human stratum
corneum in transmission electron microscopy. J Microstruct 184:67–70.
109. Pilgram GSK, Engelsma–van Pelt AM, Bouwstra JA, Koerten HK. (1999) Electron
diffraction provides new information on human stratum corneum lipid organization
studied in relation to depth and temperature. J Invest Dermatol 113:403–409.
110. Fensky DB, Thewalt JL, Bloom N, Kitson N. (1994) Models of stratum corneum
intercellular membranes: 3H NMR of macroscopically oriented multilayers. Biophys J
67:1562–1573.
111. Suzuki Y, Nomura J, Koyama J, Horii I. (1984) The role of proteases in stratum cor-
neum: involvement in stratum corneum desquamation. Arch Dermatol Res 286:249–
253.
112. Brattsand M, Egelrud T. (1999) Purification, molecular cloning, and expression of a
human stratum corneum trypsin-like serine protease with possible function in desqua-
mation. J Biol Chem 274:30033–30040.
113. Ekholm IE, Brattsand M, Egelrud T. (2000) Stratum corneum tryptic enzyme in normal
epidermis: a missing link in the desquamation process? J Invest Dermatol 114:56–63.
114. Egelrud T. (2000) Desquamation. In: Lodén M, Maibach HI, eds. Dry Skin and Mois-
turizers. Boca Raton, FL: CRC Press, pp 109–117.
115. Horikoshi T, Chen S–H, Rajaraman S, Brysk H, Brysk MM. (1998) Involvement of
cathepsin D in the desquamation of human stratum corneum. J Invest Dermatol 110:
547.
116. Brysk MM, Bell T, Brysk H, Selvanayagum P, Rajaraman S. (1994) Enzymic activity
of desquamin. Exp Cell Res 214:22.
117. Watkinson A. (1998) Stratum corneum gelatinase: a novel late differentiation, epidermal
cystein protease. J Invest Dermatol 110:539.
118. Lundstrom A, Serre G, Haftek M, Egelrud T. (1994) Evidence for a role of corneo-
desmosin, a protein which may serve to modify desmosomes during cornification, in
stratum corneum cell cohesion and desquamation. Arch Dermatol Res 286:369–375.
119. Long SA, Wertz PW, Strauss JS, Downing DT. (1985) Human stratum corneum polar
lipids and desquamation. Arch Dermatol Res 277:284–287.
120. Koppe JG, Marinkovic–Ilsen A, Rijken Y, DeGroot WP, Jobsis AC. (1978) X-linked
ichthyosis. A sulfatase deficiency. Arch Dis Child 53:803–806.
121. Sato J, Denda M, Nakanishi J, Nomura J, Koyama J. (1998) Cholesterol sulfate inhibits
proteases that are involved in desquamation of stratum corneum. J Invest Dermatol
111:189–193.
122. Roberts MS, Walker M. (1993) Water—the most natural penetration enhancer. In: Wal-
ters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration Enhancement. New York:
Marcel Dekker, pp 1–30.
123. Vavosour I, Kitson N, MacKay A. (1998) What’s water got to do with it? A nuclear
magnetic resonance study of molecular motion in pig stratum corneum. J Invest Der-
matol Symp Proc 3:101–104.
124. Bouwstra JA, Gooris GS, Salomons–de Vries, van der Spek JA, Bras W. (1992) Struc-
ture of human stratum corneum as a function of temperature and hydration: a wide-
angle X-ray diffraction study. Int J Pharm 84:205–216.
125. Hofland HEJ, Bouwstra JA, Boddé HE, Spies F, Nagelkerke JF, Cullander C, Junginger
HE. (1995) Interactions between nonionic surfactant vesicles and human skin in vitro:
freeze fracture electron microscopy and confocal laser scanning microscopy. J Lipo-
some Res 5:241–263.
126. van Hal DA, Jeremiasse E, Junginger HE, Spies F, Bouwstra JA. (1996) Structure of
fully hydrated human stratum corneum: a freeze–fracture electron microscopy study. J
Invest Dermatol 106:89–95.
127. Menon GK, Elias PM. (1997) Morphologic basis for a pore-pathway in mammalian
stratum corneum. Skin Pharmacol 10:235–246.
128. Hou SYE, Mitra AK, White SH, Menon GK, Ghadially R, Elias PM. (1991) Membrane
structures in normal and essential fatty acid-deficient stratum corneum: characterization
by ruthenium tetroxide staining and X-ray diffraction. J Invest Dermatol 96:215–223.
129. Feingold KR. (1991) The regulation and role of epidermal lipid synthesis. Adv Lipid
Res 24:57–82.
130. Proksch E, Holleran WM, Menon GK, Elias PM, Feingold KR. (1993) Barrier function
regulates epidermal lipid and DNA synthesis. Br J Dermatol 128:473–482.
131. Grubauer G, Elias PM, Feingold KR. (1989). Transepidermal water loss: the signal for
recovery of barrier structure and function. J Lipid Res 30:323–334.
132. Ghadially R, Brown BE, Sequeira–Martin SM, Feingold KR, Elias PM. (1995) The
aged epidermal permeability barrier: structural, functional, and lipid biochemical ab-
normalities in humans and a senescent murine model. J Clin Invest 95:2281–2290.
133. Menon GK, Feingold KR, Moser AH, Brown BE, Elias PM. (1985) De novo sterol-
ogenesis in the skin: fate and function of newly synthesized lipids. J Lipid Res 26:
418–427.
134. Grubauer G, Feingold KR, Elias PM. (1987) The relationship of epidermal lipogenesis
to cutaneous barrier function. J Lipid Res 28:746–752.
135. Holleran WM, Feingold KR, Mao–Qiang M, Gao WN, Lee JM, Elias PM. (1991)
Regulation of epidermal sphingolipid synthesis by permeability barrier function. J Lipid
Res 32:1151–1158.
136. Proksch E, Feingold KR, Mao–Qiang M, Elias PM. (1991) Barrier function regulates
epidermal DNA synthesis. J Clin Invest 87:1668–1673.
137. Taljebini M, Warren R, Mao–Qiang M, Lane E, Elias PM, Feingold KR. (1996) Cu-
taneous permeability barrier repair following various types of insults: kinetics and
effects of occlusion. Skin Pharmacol 9:111–119.
138. Menon GK, Grayson S, Elias PM. (1985) Ionic calcium reservoirs in mammalian ep-
idermis: ultrastructural localization by ion-capture cytochemistry. J Invest Dermatol 84:
508–512.
139. Menon GK, Elias PM, Lee SH, Feingold KR. (1992) Localization of calcium in murine
epidermis following disruption and repair of the permeability barrier. Cell Tissue Res
270:503–512.
140. Lee SH, Elias PM, Feingold KR, Mauro T. (1994) A role for ions in barrier recovery
after acute perturbation. J Invest Dermatol 102:976–979.
141. Lee SH, Elias PM, Proksch E, Menon GK, Mao–Qiang M, Feingold KR. (1992) Cal-
cium and potassium are important regulators of barrier homeostasis in murine epider-
mis. J Clin Invest 89:530–538.
142. Elias PM, Feingold KR. (1992) Lipids and the epidermal water barrier: metabolism,
regulation, and pathophysiology. Semin Dermatol 11:176–182.
143. Imokawa G. (1980) Comparative study on the mechanism of irritation by sulfate and
phosphate type of anionic surfactants. J Soc Cosmet Chem 31:45–66.
144. Patil SM, Singh P, Maibach HI. (1994) Cumulative irritancy in man to sodium lauryl
sulfate: the overlap phenomenon. Int J Pharm 110:147–154.
145. Downing DT, Abraham W, Wegner BK, Willman KW, Marshall JL. (1993) Partition
of sodium dodecyl sulfate into stratum corneum lipid liposomes. Arch Dermatol Res
285:151–157.
146. van der Valk PGM, Nater JP, Bleumink E. (1984) Skin irritancy of surfactants as
assessed by water vapor loss measurements. J Invest Dermatol 82:291–293.
147. Prottey C. (1978) The molecular basis of skin irritation. In: Breuer MM, ed. Cosmetic
Science. Vol. 1. New York: Academic Press, pp 275–349.
148. Singer EJ, Pittz EP. (1985) Interaction of surfactants with epidermal tissues: biochem-
ical and toxicological aspects. In: Rieger MM, ed. Surfactants in Cosmetics. New York:
Marcel Dekker, pp 133–194.
149. Zhou J, Mark R, Stoudemayer T, Sakr A, Lichtin JL, Gabriel KL. (1991) The value
of multiple instrumental and clinical methods, repeated patch applications, and daily
evaluations for assessing stratum corneum changes induced by surfactants. J Soc Cos-
met Chem 42:105–128.
150. Imokawa G, Sumura K, Katsumi M. (1975) Study on skin roughness caused by sur-
factants. II Correlation between protein denaturation and skin roughness. J Am Oil
Chem Soc 52:484–489.
151. Cohen C, Dossou G, Rougier A, Roguet R. (1991) Measurement of inflammatory
mediators produced by human keratinocytes in vitro: a predictive assessment of cuta-
neous irritation. Toxicol In Vitro 5:407–410.
152. Menon GK, Lee SH, Roberts MS. (1998) Ultrastructural effects of some solvents and
vehicles on the stratum corneum and other skin components: evidence for an ‘‘extended
mosaic-partitioning model of the skin barrier.’’ In: Roberts MS, Walters KA, eds. Der-
mal Absorption and Toxicity Assessment. New York: Marcel Dekker, pp 727–751.
153. Walters KA, Flynn GL, Marvel JR. (1985) Physicochemical characterization of the
human nail: solvent effects on the permeation of homologous alcohols. J Pharm Phar-
macol 37:771–775.
154. Quintanar–Guerrero D, Ganem–Quintanar A, Tapia–Olguin P, Kalia YN, Buri P.
(1998) The effect of keratolytic agents on the permeability of three imidazole anti-
mycotic drugs through the human nail. Drug Dev Ind Pharm 24:685–690.
2
Common Skin Disorders and Their
Topical Treatment
C. COLIN LONG
Cardiff and Vale NHS Trust, Cardiff, Wales
I. INTRODUCTION
Dermatologists have the advantage over other clinicians in that their patients present
with diseases that are usually visible and accessible. This means that the majority of
skin diseases may be treated topically with treatment delivered directly to the desired
site of action, thereby avoiding, or at least we hope attenuating, the potential for
systemic side effects. Systemic treatment may be needed if the skin disease is severe,
recalcitrant, or fails to respond to topical therapy. The aim of this chapter is to
provide a brief synopsis of common skin disorders and their current topical treatment
for nonspecialists. A glossary of common dermatological terms has been included at
the end of this chapter (see Sec. XII).
II. PSORIASIS
A. Introduction
Psoriasis is a chronic inflammatory skin disease of unknown etiology that affects
between 1 and 3% of the population. There is increased proliferation of the epidermis
with infiltration of inflammatory cells within the dermis and epidermis, coupled with
dilation of the upper dermal capillaries. Psoriasis tends to run in families and may
be associated with certain HLA phenotypes; individuals with first-degree relatives
with psoriasis are more likely to develop the disease themselves. Abnormalities of
arachidonic acid metabolism have been demonstrated within psoriatic plaques, and
it is possible that arachidonic acid and its metabolites may be intimately involved
in the psoriatic process. An increase in the levels of prostaglandins (PGs) may cause
41
42 Long
vasodilation and erythema, and leukotrienes, such as 5,12-dihydroxy-6,14-cis-8,10-

trans-eicosatetraenoic acid (LT-B4) and 12-hydroxy-5,8,14-cis-10-trans-eicosatetrae-
noic acid (12-HETE), as well as interleukin (IL)-8, and the complement product C5a
des arg may cause neutrophil accumulation (1–4). Raised levels of calmodulin, a
cellular receptor protein for calcium, have been demonstrated in psoriatic lesions.
The calcium–calmodulin complex may influence cell proliferation in psoriasis by
modulating the activities of phospholipase-A2 (which releases arachidonic acid from
cell membranes) and phosphodiesterase. Drugs such as anthralin (dithranol) and cy-
closporine, which are beneficial in psoriasis, are calmodulin antagonists. Intracellular
cAMP levels are decreased within psoriatic lesions and drugs that decrease cAMP
levels, such as ␤-blockers or lithium, may worsen psoriasis. Conversely, drugs such
as benoxaprofen (now withdrawn) which elevate cAMP levels may improve
psoriasis.
The role of T lymphocytes in psoriasis is unclear. However, T-helper lympho-
cytes form a major part of the dermal inflammatory cell infiltrate, and thus the effect
of cyclosporine in psoriasis may be partly due to its anti–T-helper cell activity.
B. Clinical Patterns
Several clinical patterns in psoriasis are recognized. The most common is chronic
plaque psoriasis in which there are erythematous plaques of psoriasis with an over-
lying silvery scale usually affecting the elbows, knees, and at times, the scalp and
lower back (Fig. 1). Guttate psoriasis, which may be precipitated by a streptococcal
infection of the throat, is characterized by numerous small, scaling erythematous
plaques on the trunk and limbs. Psoriasis may also affect the flexures and may cause
a glazed erythematous appearance similar to that seen in seborrheic eczema.
Erythrodermic psoriasis is characterized by severe erythema affecting the whole
of the patient’s skin. This may develop following deterioration of the patient’s pso-
riasis or be precipitated by use of potent topical or systemic steroids. There may be
associated systemic symptoms, and the patient is at risk from hypothermia owing to
excessive heat loss, dehydration, and cardiac failure.
Figure 1 A large plaque of psoriasis on the trunk: There is erythema with overlying silvery
scale.
Treatment of Skin Disorders 43
Generalized pustular psoriasis is another rare, but extremely serious, type of

psoriasis in which the patient has widespread areas of erythema with overlying sheets
of sterile pustules. Localized pustular psoriasis may also affect the palms and soles.
The nails may be involved in psoriasis, and patients may develop pitting or
lifting of the nail plate from the nail bed (onycholysis). Patients may also present
with gross thickening of the nails owing to subungual hyperkeratosis.
C. Topical Therapy
1. Emollients
Emollients act by blocking transepidermal water loss and help to soften and soothe
the skin. In psoriasis, they help reduce scaling and may make the skin more com-
fortable. Patients should be encouraged to use an emollient bath oil or shower gel
when bathing and to apply emollients when other treatments (see later discussion)
have been washed off. There are numerous emollients available, and it is important
that the patient tries several until they find one that suits them best. Emollients are
particularly beneficial in patients with erythrodermic or pustular psoriasis who are
unable to tolerate other more ‘‘active’’ forms of topical therapy.
2. Coal Tar
Many preparations of coal tar are available, and there are differences between the
composition of coal tar obtained from different sources. Coal tar may be formulated
in different vehicles to produce creams, gels, and pastes. The mode of action is
unclear, but it is believed to have an antimitotic effect on proliferative cells (5). Tar
is also a photosensitizer and may be used in combination with ultraviolet light (UV)B
radiation therapy in the Goeckerman regimen. Crude coal tar preparations have the
disadvantages of mess and smell, but do not cause systemic toxicity. Tar preparations
may be applied to normal skin without ill-effect, but may be irritant to the skin of
the face, flexures, and genitals, and can cause folliculitis. It is doubtful whether
concentrations of crude coal tar greater than 10% are of any additional benefit. Coal
tar solutions (in alcohol) are cleaner and more cosmetically acceptable, but they are
less potent and less effective.
3. Anthralin (Dithranol)
Anthralin [known abroad as dithranol], a synthetic derivative of chrysarobin (an
extract of tree bark) is the most effective topical antipsoriatic treatment. Similar to
tar, its mechanism of action is unknown, but anthralin is known to act as an anti-
mitotic agent by reducing DNA synthesis (6), and it also inhibits the enzyme glucose-
6-phosphate dehydrogenase. Anthralin is active only in its reduced form and thus
must be combined with an antioxidant. In Lassar’s paste, anthralin is combined with
salicylic acid, which acts as a kerotolytic agent (and will help remove scale), and in
Dithrocream, ascorbic acid is present as an antioxidant. Anthralin may cause irrita-
tion, burning, and staining of the skin, and may also stain clothing. It is important
to initiate therapy with a weak concentration (such as 0.1%) and then increase the
concentration as the patient’s tolerance increases. In Ingram’s method of applying
anthralin the patient soaks in a warm bath containing coal tar solution (1:800) and,
after drying, is exposed to UVB radiation. A paste of the agent is then applied to
the lesions. This procedure is repeated daily. In the hospital anthralin is frequently
44 Long
left on for several hours, but short-contact therapy, in which it is washed off 30–60
min after application, may be more convenient for outpatients.
4. Calcipotriol [Calcipotriene]
Calcipotriol [known in the United States as calcipotriene] is a vitamin D analogue
that has benefit in psoriasis. Vitamin D analogues reverse the increased proliferation
and other changes seen in psoriatic skin, and this may be through intracellular vi-
tamin D receptors known to be present in epidermal keratinocytes, Langerhans cells,
T lymphocytes, and macrophages (7). Vitamin D analogues may also affect the in-
flammatory cell infiltrate. Calcipotriene (ointment or cream) has the advantage of
being cosmetically more acceptable to patients than tar or anthralin (dithranol) prep-
arations, but may cause irritant dermatitis in some patients. Patients are limited to a
maximum of 100 g/week because there is a potential risk of hypercalcemia and
hypercalciuria.
5. Topical Steroids
Topical steroids have anti-inflammatory, immunosuppressive, antimitotic, and vaso-
constrictive effects on the skin. Topical steroids are effective in reducing the inflam-
matory changes seen in psoriasis, but there is a risk of precipitating widespread
erythroderma, particularly if potent preparations are used, or if topical steroids are
suddenly withdrawn. For these reasons, topical steroids are usually avoided in the
general treatment of psoriasis. However, weak topical steroids are useful in treating
areas such as the scalp, face, and flexures, where other treatments such as tar, an-
thralin, or calcipotriene are likely to cause irritation.
6. Treatment of Scalp Psoriasis
Scalp psoriasis, similar to psoriasis elsewhere, will respond to a similar range of
topical therapy although the presence of hair makes treatment more difficult. Appli-
cation is messy. For thick plaques on the scalp topical therapies include oil of cade,
‘‘ung. cocois,’’ and combinations of coal tar and salicylic acid (such as 6% coal tar
and 3% salicylic acid), all of which are effective at removing scale and settling
inflammation. These preparations may be applied to the scalp and left on overnight
before being washed out in the morning with a tar-based shampoo. Patients should
be warned to use old pillowcases or towels, or to wear a showercap to protect
bedding. Other more cosmetically acceptable preparations include calcipotriene scalp
solution (Dovonex), 0.025% fluocinolone acetonide gel (Synalar gel), and 0.1% be-
tamethasone valerate (Betnovate scalp application). Various tar-based shampoos can
be used to reduce mild inflammation and scaling.
III. ECZEMA
A. Classification of Eczema
Eczema may be considered as either endogenous or exogenous. The terms eczema
and dermatitis are synonomous, although dermatitis is sometimes used to imply that
the eczema has been caused by an external agent (exogenous). Endogenous eczemas
include atopic eczema, seborrheic eczema, discoid eczema, pompholyx, and varicose
eczema. Exogenous eczemas include both irritant and allergic contact dermatitis as
well as photodermatitis (caused by the interaction of light and chemicals absorbed

by the skin).
Acute eczema presents as a pruritic erythematous confluent papular rash with
an ill-defined border. There may be vesicles present, and if these rupture there may
be exudation and ‘‘weeping.’’ Chronic eczema tends to be erythematous, scaly, and
is less likely to be vesicular. There may also be some degree of lichenification and
fissuring.
B. Atopic Eczema (Dermatitis)

The etiology of atopic eczema is unknown. Patients have increased levels of serum
IgE and some have precipitating antibodies to environmental allergens, including
foods and inhaled materials. Many patients will have a positive response to intra-
cutaneous challenge with pollen, house dust mite, cat fur, and fish antigens. However,
the significance of these positive reactions is unclear. Patients with atopic eczema
have reduced numbers of circulating T-suppressor cells which are responsible for
modulating immunoglobulin-producing B lymphocytes. Low levels of the unsatu-
rated fatty acids ␥-linoleic and dihomo-␥-linolenic acid have been reported (8).
In most patients there is a family history of eczema or of other atopic diseases,
such as asthma or allergic rhinitis. Atopic eczema usually presents during infancy
and, often, may resolve during childhood, whereas in others it may persist into adult
life. Atopic eczema usually affects the face, wrists, and the flexural aspects of the
elbows and knees (Fig. 2). There may be some involvement of the trunk, and the
rash may become generalized. The eczema may be complicated by bacterial infec-
tion, and there is evidence to suggest that many exacerbations of atopic eczema may
be due to occult infection with Staphylococcus aureus. Eczematous skin is also more
prone to infections with wart viruses, molluscum contagiosum, and herpesviruses.
Patients with atopic dermatitis may develop a widespread and potentially fatal rash,
eczema herpeticum, following the development of herpes simplex or following con-
tact with individuals affected with herpes simplex.
Figure 2 Excoriated atopic eczema in the flexural aspects of the elbow: There is licheni-
fication with increased prominence of the skin markings.
46 Long
C. Contact Dermatitis
Irritant contact dermatitis is caused by (usually repeated) exposure to chemical or
mechanical trauma. Some individuals seem more prone than others. In allergic con-
tact dermatitis the sensitizing agent (antigen) crosses the stratum corneum to reach
the epidermal Langerhans cells. The antigen is processed by the Langerhans cells
and presented to circulating T lymphocytes with subsequent development of a clone
of T lymphocytes with a specific memory for that particular antigen. In an individual
sensitized to a particular antigen, repeat exposure to that antigen will result in mi-
gration of the sensitized T lymphocytes to the site of exposure, with initiation of the
inflammatory eczematous process. Both irritant and allergic contact dermatitis usu-
ally start at the site(s) of initial or more frequent contact, but may spread to involve
other areas.
D. Treatment of Eczema
In exogenous irritant or allergic contact dermatitis the mainstay of treatment is to
identify the precipitating agent and to avoid it if at all possible. Otherwise the same
general principles apply to the treatment of all forms of eczema. If there is a pos-
sibility that the eczema may be infected, skin swabs should be submitted for bacterial
culture and sensitivity and, if appropriate, an antibiotic such as flucloxacillin should
be prescribed. For acute, particularly wet and weeping eczema, astringent solutions
such as potassium permanganate (1:10,000) are indicated. The involved area, such
as hands or feet, can be placed in a bowl of the solution or, alternatively, wet gauze
swabs may be applied directly to the skin. The majority of patients with eczema
have a chronic, dry scaling, rash. These patients should be advised to use emollients
frequently. They should also avoid soap and use soap substitutes, such as emulsifying
ointment or Diprobase, wherever possible. Emollient bath oil and gels should be
used when bathing, and ointments and creams applied to the skin after bathing. If
these simple measures fail to settle the eczema, a topical corticosteroid may be
necessary. Topical corticosteroids have an anti-inflammatory, immunosuppressive,
antimitotic, and vasoconstrictive action on the skin. These actions are mediated by
a nuclear receptor for hydrocortisone to which other steroids also bind (9). The
stronger the affinity of the steroid for the receptor the more potent the steroid. The
anti-inflammatory action of the corticosteroids depends on the induction of peptides,
known as lipocortins, which antagonize the actions of phospholipase-A2 which acts
to release arachidonic acid from membrance phospholipids. Other effects of corti-
costeroids include lysosomal and cellular membrane stabilization, a reduction in the
number of epidermal Langerhans cells, and modulation of the migration of inflam-
matory cells.
Topical corticosteroids are classified as being mild (e.g., hydrocortisone 1%),
moderately potent (e.g., clobetasone butyrate 0.05%), potent (e.g., betamethasone
valerate 0.1% or hydrocortisone butyrate 0.1%), or as very potent (e.g., clobetasol
propionate 0.05%).
Creams are suitable for moist or weeping areas of eczema, whereas ointments
should be used for dry, scaly, or lichenified areas. Local side effects of topical steroids
include masking or worsening of infection (especially fungal infections), thinning of
the skin, induction of striae, bruising and telangectasia, on aggravation of rosacea.
Use of more potent topical steroids may result in pituitary–adrenal axis suppression,
iatrogenic Cushing’s syndrome, and stunted growth.
In general, one should aim to use a steroid of sufficient potency to control the
eczema, and then aim to reduce the potency of the topical steroid preparation as the
eczematous rash improves. There is a tendency for the eczematous rash to rebound
when treatment is stopped or the potency of the topical steroid is decreased. The
aim should be for the patient to use the least potent topical corticosteroid that will
control the rash and, preferably, to use simple emollients only. It is important that
patients are prescribed adequate quantities of topical therapy and that these are ap-
plied regularly. There is doubt as to whether topical steroids need to be applied twice
daily and some topical steroids such as fluticasone propionate 0.05% (Cutivate) or
mometasone furoate 0.1% (Elocon) are claimed to be effective when applied once
daily. Patients may apply moisturizers ad lib in between applications of cortico-
steroids.
Coal tar preparations may be helpful with reducing the pruritis of eczema and
may be particularly valuable in lichenified and localized eczema, such as lichen
simplex. Occlusive bandaging, such as Viscopaste, Coltapaste, or Ichthopaste band-
ages, may be of benefit, particularly if the eczema is excoriated. The bandages are
soothing and prevent further excoriation. Some patients, however, may develop sen-
sitivity to the preservatives found in these bandages.
Infection with S. aureus can frequently exacerbate eczema and patients may
benefit from the use of a combined topical steroid antibiotic preparation such as
Fucibet. However, topical antibiotics may cause sensitization, and, in recent years,
the problem of sensitization to topically applied corticosteroids themselves has also
been recognized, and should be considered if the eczema fails to settle.
E. Seborrheic Eczema
In infants, seborrheic eczema may present as greasy adherent scale on the scalp
(cradle cap). Seborrheic eczema in adults principally affects the greasier areas of the
body, including the scalp, eyebrows, eyelids, nasal–labial areas, and chin. In young
men, it may also affect the presternal area and upper back and, in the elderly, may
involve the flexures and may become generalized. Individuals who are immunosup-
pressed (including those with human immunodeficiency virus; HIV) are more prone
to develop seborrheic eczema. It is possible that, in affected individuals, the com-
mensal yeast-like microorganism Pityrosporum ovale has become pathogenic and
provokes an inflammatory response (10). Cradle cap in infants may be treated with
olive oil or arachis (peanut) oil. In adults, weak topical steroids, with or without
azoles such as miconazole, clotrimazole, or ketoconazole, may be used. Ketoconazole
shampoo, used two or three times weekly as a liquid soap to wash the affected areas,
can also be helpful. Topical 8% lithium succinate cream has also been of use.
IV. ACNE
A. Introduction
Acne is one of the most common and distressing of skin diseases commonly present
during adolescence and usually (but not always) resolves in early adult life. Seventy
percent of the population develop acne, but only a relatively small proportion seek
48 Long
Figure 3 Nodular acne of the face and neck: There are pustules, papules, nodules, and
cysts present. There is a high risk of scarring without treatment.
medical attention. Several variants of acne are recognized, including infantile acne,
which occurs on the face during the first few months and usually settles spontane-
ously, and occupational acne, resulting from exposure to oil, coal tar, chlorinated
hydrocarbons, or insecticides. Acne may be precipitated or exacerbated by certain
combined oral contraceptive pills or by androgenic hormones.
Acne vulgaris commonly affects the face, chest, and upper back, and usually
presents during puberty. The clinical features include an increased rate of sebum
secretion, comedones, papules, and pustules (Fig. 3). Severe acne may be compli-
cated by atrophic or nodular keloid-type scars or by the formation of chronic nodules
and cysts (Fig. 4).
Patients with acne tend to have a higher sebum excretion rate than others, and
there is a degree of correlation between the sebum secretion rate and the severity of
the acne (11,12). Circulating androgens stimulate the sebaceous glands with resulting
hypertrophy and increased sebum secretion. Furthermore, there is abnormal keratin-
Figure 4 Pitted scars on the chest following acne.

ization of the epithelium lining the hair follicle, which may lead to obstruction of
the follicle with resulting comedone (blackhead) formation. Propionibacterium ac-
nes, a gram-positive commensal bacterium, proliferates within the obstructed hair
follicle, and may break down the lipid esters of sebum to liberate potentially irritating
fatty acids (13). Eventual rupture of the wall of the obstructed follicle and the release
of fatty acids into the surrounding dermis will result in an inflammatory response.
B. Treatment of Acne
The aims of treatment are to reduce the bacterial population of the hair follicles; to
encourage the shedding of comedones; to reduce the rate of sebum production; and
to reduce the degree of inflammation. Topical therapy is appropriate for mild-to-
moderate acne, but more severe forms of acne, in which there is a risk of scarring,
will require systemic therapy. Skin cleansers such as Phisomed or Hibiscrub are of
some value. Benzoyl peroxide reduces comedone formation, as well as reducing the
population of P. acnes, and may also have an anti-inflammatory effect. Benzoyl
peroxide cream may be applied twice daily at an initial concentration of 2.5% and
increased to 5 or 10% as tolerated. Benzoyl peroxide can have an irritant effect and
may also bleach both hair and clothing.
1. Topical Antibiotics
Topical erythromycin, clindamycin, and tetracycline are all effective in acne (14–
17). These antibiotics reduce the population of P. acnes and Staph. epidermidis, and
may have a separate anti-inflammatory action. The advantage of topical antibiotics
is the reduction in the risk of potential systemic side effects, and this is particularly
true with topical clindamycin. Topical tetracyclines may cause some yellow staining
of clothing and fluoresce under ultraviolet radiation. It is also possible that they may
exacerbate the problem of bacterial antibiotic resistance.
2. Topical Retinoids
Topical retinoids, including tretinoin and isotretinoin, act by decreasing epidermal
proliferation and reducing the abnormal keratinization process in the hair follicle.
This prevents new comedones forming and softens and removes existing comedones.
There is also a reduction in the level of P. acnes within the hair follicle. Although
topical retinoids are of particular value in severe acne, when there are numerous
comedones present, they are also effective in other forms of mild to moderate acne.
Initially, there may be some increased irritation and pain, but this usually settles with
use. Tretinoin cream is preferable to tretinoin gel for those with dry or fair skin.
3. Azelaic Acid
Azelaic acid, a dicarboxylic acid, produced by the yeast Pityrosporum, has been of
benefit in mild-to-moderate acne (18). The mode of action is unknown, but similar
to other acne treatments, may normalize follicular keratinization (possibly by reduc-
ing filagrin formation) and may reduce the population of P. acnes. Twenty percent
azelaic acid cream is well tolerated but may cause some local irritation.
V. ROSACEA
Rosacea is a chronic inflammatory skin disorder, affecting the face, which causes
persistent erythema associated with telangectasia and papules. It most commonly
50 Long
Figure 5 Rosacea with erythema, telangiectasia, and papules: There is also a bulbous
enlargement of the nose (rhinophyma).
affects the forehead, nose, cheeks, and chin. Characteristically small pustules and
papules arise on a background of erythema and telangectasia (Fig. 5). The patients
may also complain of flushing in response to trivial stimuli. The rash may resemble
that seen in acne, but rosacea usually affects an older-aged group and is not char-
acterized by comedones. Persistent inflammation of the nose may result in rhino-
phyma (an irregular bulbous enlargement of the nose characterized by prominent hair
follicles). Over 30% of patients with rosacea may also suffer with conjunctivitis and
blepharitis. The etiology of rosacea is unknown, although the Demodex folliculorum
mite is present in increased numbers (19).
It is possible that rosacea is the result of repeated environmental trauma (cold
wind, ultraviolet radiation, and heat) which damage the upper dermal collagen and
vasculature. Rosacea-like symptoms may be precipitated by the prolonged use of
even moderately potent topical steroids on the face. Topical metronidazole gel
(0.75%) applied twice daily is effective therapy for most individuals with rosacea.
VI. LICHEN PLANUS

Lichen planus is an inflammatory skin disorder, of unknown etiology, characterized
by the presence of pruritic violaceous papules. Common sites are the flexural aspects
of the wrists and forearms, but the rash may also affect the trunk and limbs (Fig.
6). In approximately 30% of patients there is mucosal involvement with a reticulate
rash affecting the inside of the mouth. The nails may be affected in 10% of patients.
Lichen planus may be associated with other autoimmune diseases, including
vitiligo and myasthenia gravis. The etiology is unknown but the deposition of IgM
at the dermoepidermal junction, coupled with a dense inflammation in the upper
dermis, suggest an autoimmune process. Lichen planus-like rashes may be precipi-
tated by various drugs, including thiazide diuretics, gold, tetracyclines, and para-
aminosalicylic acid (PAS). The disease is usually self-limiting, but topical cortico-
steroids may be helpful.
Figure 6 Multiple flat-topped violaceous papules of lichen planus on the flexural aspects
of the wrist.
VII. PITYRIASIS ROSEA

Pityriasis rosea is an erythematous scaling rash, of unknown etiology, thought to be
secondary to infection with an, as yet unidentified, virus. Patients develop a solitary
erythematous scaling patch some 2–3 cm in diameter. After a few days, other smaller
plaques develop on the trunk. The individual lesions tend to be oval and their lon-
gitudinal axis run parallel to the lines of the ribs. The rash may be associated with
mild pruritus and can last for several weeks before resolving spontaneously. Topical
steroids may be of benefit.
VIII. SOLAR KERATOSES

Solar (actinic) keratoses present as scaling hyperkeratotic plaques or papules on skin
exposed to light. They are most commonly seen on elderly subjects with fair skin
who have had high levels of ultraviolet exposure over many years. They may be
associated with other signs of photodamage such as yellowing, coarsening, and wrin-
kling of the skin. Individuals with large numbers of solar keratoses are at increased
risk of developing nonmelanoma skin cancer. Histologically, the lesions show epi-
dermal thickening, with abnormal epidermal differentiation and scaling. Solar kera-
toses are common, and up to 20% of individuals over the age of 60 are affected.
Some solar keratoses may resolve spontaneously, but a small proportion may develop
into squamous cell carcinoma. Solar keratoses may respond to topical 5% 5-fluoro-
uracil cream (Efudix) an antimetabolite that inhibits DNA synthesis. This should be
applied once or twice daily for 10–14 days, and may cause the lesions and the
surrounding skin to become sore and inflamed. Alternatively, the 5% 5-fluorouracil
cream can be applied twice daily for 2 days each week and the frequency increased
until the level of the patient’s tolerance is reached, but it is probable that some degree
of irritation is necessary for the treatment to be effective. The accompanying irritation
and soreness may be treated with topical corticosteroids. Topical 5-fluorouracil has
also been claimed to be effective for superficial basal cell carcinoma and for Bowen’s
52 Long
disease. Topical tretinoin 0.05% and isotretinoin 0.1% have also been used to treat
multiple solar keratoses.
IX. SUNSCREENS
Sunscreens are preparations that filter out or reflect harmful ultraviolet radiation.
Various disorders may be precipitated or aggravated by exposure to ultraviolet light.
These include polymorphic light eruption, Hutchinson’s summer prurigo, cutaneous
porphyria, rosacea, and lupus erythematosus. Exposure to ultraviolet radiation is also
a major risk factor for both malignant melanoma and nonmelanoma skin cancer as
well as photoaging. Recurrent attacks of herpes simplex may also be precipitated by
ultraviolet exposure. Sunscreens differ greatly in their protective capacity and some
of the newer sunscreens also offer some degree of protection against ultraviolet A.
X. ULCERS
Ulcers may be seen at any body site, but are most commonly seen on the legs,
probably caused by a combination of trauma and impaired circulation. The venous
return of blood from the legs depends on efficient working of the calf muscles to
act as a pump coupled with the action of valves in the deep veins preventing the
reflux of blood. Damage to the valves of the deep veins following deep vein throm-
bosis may be associated with pregnancy. Injury or immobilization may also lead to
valvular incompetence.
The most common types of ulcer are venous ulcers caused by leaking valves
in the deep veins, resulting in venous hypertension, and edema of the subcutaneous
tissue. An extravascular accumulation of fibrinous material leaked from dermal blood
vessels results in a fibrous cuff around the capillaries that prevents diffusion of
oxygen and other nutrients through the blood vessel wall as well as causing fibrosis,
and sclerosis of the dermal capillaries. Venous ulcers are more common in women
than men and result from inadequate provision of nutrients and oxygen to the skin.
Venous ulcers are most commonly seen on the medial aspect of the lower leg usually
above the medial malleolus (inner aspect of the ankle) (Fig. 7). Large ulcers may
encircle the leg.
On examination, the lower leg and ankle and foot may be edematous, there
may be prominent varicose veins present. Leakage of blood into the skin may cause
deposition of hemosiderin. The ulcers may heal spontaneously or may become
chronic and indolent. The ulcers may be complicated by infection, bleeding or by
eczema. Varicose eczema, surrounding the ulcer, is common, and in many patients,
it may be due to allergic contact hypersensitivity to medicaments, such as neomycin
clioquinal (Vioform), lanolin, or ethylene diamine, used in treating the leg ulcer or
eczema. Rarely, in long-standing ulcers, malignant change with the development of
squamous cell carcinoma may occur, and patients with long-standing ulcers may
become anemic. It is important to try and improve venous return (drainage) by
1. Elevation of the legs for regular periods
2. Compression bandaging with elasticated stockings or bandages (which
should be graduated so that pressure is greatest at the ankle and least at
the top of the dressing)
Figure 7 A large venous ulcer on the medial (inner) aspect of the leg.
3. Regular exercise
4. Weight reduction
Exudate and slough should be removed and ulcers may be cleaned with normal
saline, sodium hypochloride solution, Eusol, or 5% hydrogen peroxide. Topical an-
tiseptics, such as povidone iodine or potassium permanganate, may help reduce the
bacterial load. Crust may be loosened by the application of saline- or potassium
permanganate-soaked dressings. Local surgical debridement may be necessary for a
thick eschar. Hydrogen peroxide cream 1.5%, streptokinase/streptodornase solution
(Varidase) may also be useful in helping remove thick slough.
Numerous dressings are available to treat leg ulcers, including nonadherent
gauze, paraffin gauze, silver sulfadiazine (Flamazine), and absorbent hydrocolloid
dressings. Any ulcerated area of skin will rapidly become colonized with numerous
different bacteria, and these should be treated only if they are causing local cellulitis.
The main function of leg ulcer dressings should be to provide comfort, protection
for the wound, and an optimum local environment to allow reepithelization. Hydro-
colloid dressings have the advantage that they may be left in situ for several days
before changing. They have the disadvantages of an unpleasant odor and an unpleas-
ant fluid may collect beneath the dressing. Surrounding (varicose) eczema may be
treated with topical corticosteroids.
Ischemic leg ulcers are due to reduced circulation secondary to atherosclerosis,
vasculitis, or other causes of arterial obstruction. Arterial ulcers tend to be more
sharply defined and painful than venous ulcers. In the leg, they are more common
on the anterior aspect of the shin, rather than the medial aspect of the ankle. Com-
pression bandaging will impair arterial blood supply further and thus it is important
that arterial pulses are examined and if necessary Doppler ultrasound or other in-
vestigations are performed so that arterial ulcers are not incorrectly treated with
compression. Ulcers may also occur in diabetes, and vasculitis, or may be due to
pyoderma gangrenosum, as well as secondary to infection, trauma, or malignant
disease.
Decubitus ulcers (pressure sores) are the result of localized ischemia secondary
to prolonged pressure in patients who are immobile. They are most common over
54 Long
the sacrum ischial tuberosities, heels, occiput, shoulders, and elbows. The ulcers are
often deep and sloughy. Neuropathic ulcers result from decreased cutaneous inner-
vation, resulting in diminished sensation, and are most commonly seen on the feet.
Leg ulcers may also be due to sickle cell disease, idopathic thrombocytopenic pur-
pura, tuberculosis, syphillis, and deep fungal infections.
XI. INFECTION OF THE SKIN

A. Yeast Infections
1. Pityriasis Versicolor
This disease is caused by Pityrosporum orbiculare, a gram-positive yeast-like mi-
croorganism that is usually a skin commensal. In some individuals, the organism can
become pathogenic. Pityrosporum versicolor often affects young adults, causing
brown scaly macules on the trunk and sometimes on the limbs. Carboxylic acids
released by the organisms inhibit melanogenesis and thus the affected areas may
appear relatively pale following exposure to sunlight. The infection may be dem-
onstrated by microscopy of skin scrapings suitably treated with potassium hydroxide
solution.
Most patients will respond to a topical imidazole drug, such as miconazole,
clotrimazole, or econazole creams, applied once daily for 6 weeks. Ketoconazole
shampoo can also be used to wash the affected areas daily for 4–5 days. Other
effective topical treatments include Whitfield’s ointment (6% benzoic acid and 3%
salicylic acid in emulsifying ointment), selenium sulfide shampoo, and 20% sodium
thiosulfate solution.
2. Candida
The yeast Candida albicans may cause vulvovaginitis in women, especially during
pregnancy, in those taking oral contraceptives, or those who are receiving systemic
antibiotics for acne. It may also cause stomatitis in infants, and may exacerbate
intertrigo in the body folds of obese individuals and the napkin area during infancy.
The nail plate may also be infected, and the organism may cause chronic paronychia
in those involved with wet-work such as bar workers or housewives. Topical treat-
ments with imidazole creams is often effective, although more serious infections may
require systemic therapy.
B. Dermatophyte Infections
Dermatophyte infection (ringworm) is restricted to invasion of the stratum corneum,
nails, and hair. The dermatophytes, Trichophyton, Epidermophyton, and Microsporum
species may infect humans. Microsporum species are usually acquired from infected
cats or dogs (M. canis) and are a frequent cause of tinea capitis (ringworm affecting
the head) in children. Infections from farm or other animals tend to cause more
vigorous inflammation than those from other sources. The infection may be diag-
nosed from microscopy of skin, nail, or hair treated with potassium hydroxide. Al-
ternatively, the fungus may be cultured.
Tinea corporis (ringworm affecting the skin of the trunk or limbs) often presents
as a pruritic, annular, erythematous, scaling plaque, which may resemble a patch of
eczema or psoriasis, but is often solitary. Tinea cruris (ringworm affecting the groin)
presents as a well-demarcated pruritic erythematous scaling rash affecting the groins.
The rash may extend onto the thigh and genitalia. Trichophyton rubrum and Epi-
dermophyton floccosum are the most common causative fungi. Tinea pedis (ringworm
affecting the feet) may affect the skin of the toe web spaces, sole, or may extend
onto the sides and dorsal aspect of the feet. Trichophyton rubrum, T. mentagrophytes;
and E. floccosum are the most common causative organisms.
Tinea manuum is a chronic form of ringworm affecting the hands (frequently
only one palm will be affected); T. rubrum is the most commonly identified organism.
Tinea unguium (ringworm affecting the nail plate and nail bed) may be caused by
T. rubrum, T. mentagrophytes, or E. floccosum. Affected nails are often thickened
and have a yellowish discoloration. Onycholysis (separation of the nail plate from
the nail bed) may also be seen.
Tinea incognito is the term used to describe dermatophyte infections treated
inappropriately with topical corticosteroids, which suppress the inflammatory re-
sponse, but allow the fungus to proliferate.
Topical imidazole creams (e.g., miconazole, econozole, or clotrimazole), which
interfere with ergosterol synthesis and thereby impair fungal cell wall permeability,
when used twice daily for 2 weeks, are adequate for most limited areas of fungal
infection. The more recently introduced topical allylamine terbinafine is also very
effective (20). More extensive infections, or involvement of the nail or scalp will
require systemic therapy.
C. Bacterial Infections
Various acute bacterial infections may affect the skin. These include impetigo, ery-
sipelas, cellulitis, furuncles, carbuncles, anthrax, diphtheria, and various mycobac-
terial infections, including tuberculosis and leprosy. Of these only impetigo (in which
a small area is infected) or furuncles, which are both caused by Staphylococcus
aureus, are amenable to topical treatment. In impetigo, which is more common in
young children, an inflamed erythematous area with a yellow crust may develop on
exposed skin. Local treatment with antibiotic washes, such as Phisomed or Hibiscrub,
and topical mupirocin or fucidic acid (Fucidin) ointment may be sufficient. More
extensive areas, larger than a few centimetres in diameter, will require treatment with
systemic antibiotics.
1. Antibiotics
Furuncles are hair follicles infected with S. aureus and present as yellow-headed
pustules. They are commonly seen on the back of the neck in men or in patients
treated with ointments or tar (particularly if the skin has been occluded). Extensive
areas of folliculitis (furuncles) will require systemic floxacillin (flucoxacillin), but
solitary or isolated lesions will respond to topical mupirocin or sodium fusidate
(Fucidin). Mupirocin (pseudomonic acid) interferes with bacterial protein synthesis,
has the advantage of no cross-resistance with other antibiotics and is available only
as a topical preparation. It is effective against both staphylococci and streptococci
and may be used in the treatment of folliculitis, infected eczema, and as prophylaxis
against nasal carriage of staphylococci. Fucidic acid inhibits bacterial protein syn-
thesis and is particularly effective against staphylococcal skin infections. Topical
56 Long
application may lead to a hypersensitivity reaction. Metronidazole inhibits DNA syn-

thesis and is active against anaerobic bacteria and protozoa, and it can be used
topically in the treatment of rosacea. It has also been used to reduce the smell of
infected, sloughy ulcers.
2. Antiseptics
Antiseptics have bactericidal activity and may be used as cleansing agents on the
skin, as an adjunct to antibiotic therapy, or to prevent secondary bacterial infection.
Povidone iodine is a powerful bactericide and may also be effective against viruses.
It is commonly used preoperatively to minimize the chance of sepsis, and it may
also be of help in the treatment of leg ulcers and herpetic lesions. Potassium per-
manganate 1:8,000 or 1:10,000 aqueous solution is an effective antiseptic that may
be used as a soak in the treatment of infected or weeping eczema. Potassium per-
manganate also acts as an astringent which helps ‘‘dry up’’ exudative or weeping
rashes. Chlorhexidine is an effective skin disinfectant and is available in a wide
range of preparations.
D. Viral Infections
1. Herpes Simplex
Herpes simplex type I commonly causes herpetic lesions on the face and oropharynx
whereas herpes simplex type 2 affects the genitalia. Following a primary infection,
the virus may become latent within nerve ganglia and recurrent episodes may occur
periodically. In children, primary herpes simplex type 1 infection presents as an acute
gingivostomatitis with vesicles (which may subsequently ulcerate) scattered on the
lips and buccal mucosa. The infection is accompanied by malaise, headache, and
fever. Primary herpes simplex type 2 infections are usually sexually transmitted and
cause multiple painful genital or perianal blisters and ulcers. The virus can also be
inoculated to other areas, or other individuals during contact sports such as rugby or
wrestling. Recurrent infections occur more or less at the same site each time, and
are often precipitated by ultraviolet stress, respiratory tract infections, or menstrua-
tion. There may be a preceeding discomfort or tingling followed within a few hours
by the development of erythema and vesicle formation. The episodes are usually
self-limiting. A crust often develops within 48 h and the lesions resolves within 5 or
6 days. Some patients with recurrent herpes simplex develop a widespread rash
known as erythema multiforme.
Topical antiseptics or antibiotics may prevent secondary bacterial infections,
and topical 5-idoxuridine lotion or acyclovir cream applied five times daily at the
first sign of discomfort may be effective in reducing the length and severity of attacks
(21). Following entry into herpes-infected cells, acyclovir is phosphorylated to the
active compound, acyclovir triphosphate, by herpesvirus-coded thymidine kinase.
Acyclovir triphosphate inhibits herpes-specific DNA polymerase, thereby preventing
further viral synthesis (22).
Herpes zoster (shingles) is caused by a reactivation of the varicellar–zoster
virus which causes childhood chickenpox. Following an attack of chickenpox, the
virus may remain dormant in a sensory root ganglion and may become reactivated
at a later date, often many years later. Shingles is more common in the elderly, those
with lymphoma, AIDS, or other causes of immunosuppression. The attacks often
start with unilateral paresthesiae or pain and there may be accompanying systemic
upset and fever. Vesicular lesions similar to those seen in varicella (chickenpox)
develop within the course of the cutaneous nerve. Approximately 25% of patients
develop a distressing postherpetic neuralgia that may persist for months, or some-
times years, after the rash has settled. Although topical therapy has little role in
treating acute herpes zoster, topical antiseptics, such as Betadine paint may prevent
secondary bacterial infection. Recently, a counterirritant cream containing capsaicin
at a concentration of 0.75% (Axain) has been introduced for the treatment of post-
herpetic neuralgia, and this should be applied three or four times daily after the
herpetic lesions have healed (23).
2. Viral Warts
Viral warts are caused by infection with one of the many papillomaviruses, and are
spread by direct contact from infected individuals or possibly from shed skin on
changing room floors. Warts present as horny nodules in which small black throm-
bosed capillaries may be seen. Wart infections may persist for many months and
even years, but most resolve spontaneously. Topical therapies, including salicylic
acid (10–50%), lactic acid (4–20%), podophyllin (up to 15%), or glutaralderhyde
(10%), may be effective. Topical podophyllin (15%) in compound benzoin tincture
may be applied weekly to external genital warts and should be washed off 6 h after
application. The preparation is irritant and care should be taken not to apply the
paint to nonaffected skin. Severe toxicity has been reported to be caused by treatment
of extensive lesions. Podophyllotoxin 0.5% may be applied to genital warts for 3-
consecutive days and repeated weekly for up to 5 weeks if necessary.
3. Molluscum Contagiosum
Molluscum contagiosum is a common viral infection, caused by a poxvirus, and is
common in childhood. Individuals with atopic eczema are particularly prone to in-
fection. Small pink umbilicated papules occur on the skin of the trunk and limbs.
The lesions usually resolve spontaneously within a few months but may be treated
by curettage, cryotherapy, or topical salicylic acid preparations as used for viral
warts. An alternative is to squeeze the papule with a forceps so as to express the
contents and to then apply silver nitrate, phenol, or iodine.
E. Scabies
Scabies is an infestation with the mite Sarcoptes scabei var. hominis, which is spread
by close skin-to-skin contact. The female mite burrows into the human stratum cor-
neum and lays her eggs. Affected individuals become sensitized to the mite and to
their waste products. Individuals who are immunosuppressed may develop extensive
infestation. Scabies presents as a severe generalized pruritus, affecting the trunk and
limbs, which is often worse at night. The patient may have a generalized eczematous
rash which may become impetiginized, and small linear streaks of scabies burrows
may be seen, in particular on the flexural aspect of the wrists, on the palm, on the
sides of the fingers, and on the soles of the feet. In men, papular lesions on the
genitalia are common. Norwegian scabies is characterized by infestation with large
numbers of mites and is most commonly seen in immunosuppressed patients, such
as those with leukemia or HIV infection. Patients develop crusted water lesions on
58 Long
the hands and feet which contain hundreds of mites. This form of scabies is highly
contagious owing to the large number of mites present.
Benzyl benzoate (25% lotion), 10% sulfur in yellow soft paraffin, lindane (1%
lotion), malathion (0.5% lotion), and permethrin (5% cream), all are effective. How-
ever, lindane should be avoided during pregnancy or in nursing mothers. It is im-
portant that all members of an affected household and any person having close
physical contact should be treated whether symptomatic or not. Treatment should be
applied to the whole body and in infants and younger children this should include
the scalp, neck, face, and ears. It is important that all areas, particularly the finger
and toe web spaces, and the genital areas are adequately treated. Two ‘‘coats’’ should
be applied (using a 2-in. paint brush for lotions if possible) to ensure all areas are
treated. The treatment should be washed off after 24 h and all clothing changed. The
most common causes of treatment failure are failure to treat all members of a house-
hold or failure to treat all areas of the body. It is important to warn patients that the
pruritus associated with infestation may persist for several days after treatment, and
topical steroids applied after successful treatment of the mites can help to settle the
pruritus.
F. Insect Bites
Some individuals seem particularly sensitive to bites from flea and other insects.
They develop pruritic papules and sometimes blister at the site of injury. Topical
corticosteroids may be of benefit, and any potential animal source should be treated.
GLOSSARY OF DERMATOLOGICAL TERMS

Abscess A localized collection of pus greater than 1 cm in diameter
Angioedema A transient acute swelling caused by tissue edema
Annular Ring-like
Arcuate Forming part of a circle
Atrophy Thinning of the skin owing to reduction in the thickness of the
epidermis, dermis, or subcutis
Bulla A larger fluid-containing lesion of more than 1 cm in diameter
Circinate Circular
Comedone A plug of keratin and sebum obstructing the follicle
Ecchymosis Larger purpura greater than 1 cm in diameter
Erosion Complete or partial loss of the epidermis that will heal without
scarring
Excoriation Ulcer or erosion produced by physical trauma
Fissure A split through the surface of the skin
Hematoma A deep swelling in the skin secondary to hemorrhage
Lichenification A thickening of the skin with increased prominence of skin
markings
Macule A flat area of discolored skin
Nodule A firm raised lesion larger than 1 cm in diameter
Nummular Coin-like or discoid
Papilloma A pendulant mass projecting from the skin
Papule A small, firm, raised lesion of less than 0.5 cm in diameter
Petechiae A small macule caused by presence of blood in the skin

Plaque A circumscribed elevated area of skin, usually larger than 1 cm in
diameter
Purpura Larger macules or papules caused by the collection of blood in the
skin
Pustule A small visible collection of pus
Reticulate Net-like
Scale Small flake arising from the horny layer (stratum corneum) or crust
of adherent dry blood or tissue fluid
Scar The permanent replacement of normal structures of the skin with
fibrous tissues secondary to healing
Stria A linear streak of atrophic skin
Telangectasia Dilated small cutaneous blood vessel
Tumor Large nodule
Ulcer An area of complete loss of the epidermis and sometimes of the
dermis and subcutis; ulcers heal to leave a scar
Vesicle A fluid-containing lesion of less than 1 cm in diameter
Weal An elevated off-white area of skin often with surrounding erythema
(redness), which is usually transient
BIBLIOGRAPHY
For further information and good pictorial examples of the skin conditions described
in this chapter the reader is referred to the following compendia.
Champion RH, Burton JL, Ebling FJG, eds. Textbook of Dermatology. Oxford: Blackwell
Scientific, 1992.
Lawrence CM, Cox NH. Physical Signs in Dermatology. London: Wolfe, 1993.
du Vivier A. Atlas of Clinical Dermatology. 2nd ed. London: Gower Medical, 1993.
REFERENCES
1. Brain S, Camp R, Dowd P, Kobza-Black A, and Greaves M. The release of leukotriene
B4 like material in biologically active amounts from the lesional skin of patients with
psoriasis. J Invest Dermatol 83:70–73, 1984.
2. Woolard PM. Stereochemical differences between 12-hydroxy-5,8,10,14-eicosatetrae-
noic acid in latelets and psoriatic lesions. Biochem Biophys Res Commun 136:169–76,
1986.
3. Gearing AJ, Fincham NJ, Bird CR, Wadhwa M, Meager A, Cartwright JE, Camp RD.
Cytokines in skin lesions of psoriasis. Cytokine 2:68–75, 1990.
4. Schröder J–M, Christophers E. Identification of C5a des arg and an anionic neutrophil
activating peptide (ANAP) in psoriatic scales. J Invest Dermatol 87:53–58, 1986.
5. Griffiths WAD, Wilkinson JD. Topical therapy. In: Champion RH, Burton J, Ebling FJG,
eds. Textbook of Dermatology. Oxford: Blackwell Scientific, pp 3054–3055, 1992.
6. Griffiths WAD, Wilkinson JD. Topical therapy. In: Champion RH, Burton J, Ebling FJG,
eds. Textbook of Dermatology. Oxford: Blackwell Scientific, pp 3055–3056, 1992.
7. Milde P, Hauser U, Simon T, Mall G, Ernst V, Haussler MR, Frosch P, Rautertberg EW.
Expression of 1,25-dihydroxy-vitamin D3 receptors in normal and psoriatic skin. J Invest
Dermatol 97:230–239, 1991.
60 Long
8. Hansen AE. Serum lipid changes and therapeutic effects of various oils in infantile
eczema. Proc Soc Exp Biol Med 31:160–161, 1933.
9. Marks R. Topical corticosteroids: treatment of eczematous disorders. In: Munson PL,
ed. Principles of Pharmacology: Basic Concepts and Clinical Applications. New York:
Chapman & Hall, pp 1223–1226, 1995.
10. Hay R, Roberts SOB, and Mackenzie DWR, Mycology. In: Champion RH, Burton J,
Ebling FJG, eds., Textbook of Dermatology. Oxford: Blackwell Scientific, pp 1176–
1179, 1992.
11. Burton JL, Schuster S. The relationship between seborrhoea and acne vulgaris. Br J
Dermatol 84:600–601, 1971.
12. Cunliffe WJ, Schuster S. Pathogenesis of acne. Lancet 1:685–687, 1969.
13. Ebling FJG, Cunliffe WJ. Disorders of the sebaceous glands. In: Champion RH, Burton
J, Ebling FJG, eds., Textbook of Dermatology. Oxford: Blackwell Scientific, pp 1699–
1744, 1992.
14. Dobson RL, Belknap BS. Topical erythromycin solution in acne: results of a multi-clinic
trial. J Am Acad Dermatol 3:478–482, 1980.
15. Feucht CL, Allen BS, Chalker BK, Smith JG. Topical erythromycin with zinc in acne:
a double-blind controlled study. J Am Acad Dermatol 3:483–491, 1980.
16. Gloor M, Kraft H, Franke M. Effectiveness of topically applied antibiotics on anaerobic
bacteria in the pilosebaceous duct. Dermatologica 157:96–104, 1984.
17. Stoughton RB. Topical antibiotics for acne vulgaris. Arch Dermatol 115:486–489, 1979.
18. Nazzaro–Porro M, Passi S, Picardo M. Beneficial effect of 15% azelaic acid cream on
acne vulgaris. Br J Dermatol 109:371–374, 1983.
19. Spickett SG. Aetiology of rosacea. Br Med J 1:1625–1626, 1962.
20. Ryder NS. Terbinafine: mode of action and properties of squalene epoxidase inhibition.
Br J Dermatol 126(suppl 39):2–7, 1992.
21. Fiddian AP, Yeo JM, Stubbings R, Dean D. Successful treatment of herpes labialis with
topical acyclovir. Br Med J 286:1699–1701, 1983.
22. Motley RJ. Viral skin disease and its treatment. In: Munson PL, ed. Principles of Phar-
macology: Basic Concepts and Clinical Applications. New York: Chapman & Hall, pp
1255–1257, 1995.
23. Westerman RA, Roberts RG, Kotzmann RR, Westerman DA, Delaney C, Widdop RE,
Carter BE. Effects of topical capsaicin on normal skin and affected dermatomes in herpes
zoster. Clin Exp Neurol 25:71–84, 1988.
3
Basic Mathematical Principles in
Skin Permeation
ADAM C. WATKINSON* and KEITH R. BRAIN

I. INTRODUCTION
sound knowledge of the underlying mathematical principles of membrane transport
is essential if we are to expand our understanding of how membrane barriers fulfill
their function and how we can alter their properties to our advantage. The subject
of the mathematics of diffusion are enough to fill entire books (1), but in this chapter
we have attempted to pick out those mathematical solutions and descriptions that are
both commonly used and most appropriate in the field of percutaneous absorption.
It is the purpose of this work to attempt to present these equations in a manner that
will enable readers to apply them to real numbers generated in their laboratories.
At its simplest and most ideal a membrane can be described as a homogeneous
slab of an inert material, with a finite and uniform thickness. This is a convenient
theoretical picture and, although it is somewhat removed from the reality of such
complex biological membranes as the stratum corneum, it is a logical model with
which to begin when attempting to construct any sort of mathematical treatise of the
process of membrane permeation.
Much of the early mathematics relating to transmembrane diffusion had its
origins in the theoretical description of heat transfer and conductance. Indeed, the
most basic of the diffusion equations, Fick’s first law, has its roots here.
61
62 Watkinson and Brain
II. MOLECULAR DIFFUSION

The transport of molecules across any membrane, including the skin, occurs by the
process of passive diffusion. Molecular motion within an isolated system is indis-
criminate: molecules are said to follow ‘‘random walks’’ and, therefore, are not sub-
ject to a net force in any particular direction. Initially, this may seem anomalous
when considering the process of diffusion. If molecules within a membrane move
randomly why can we always predict that some proportion of them will traverse
such a barrier?
A simple way to explain this phenomenon is by considering the diffusion of a
dye in water. We begin with the situation shown in Figure 1a in which a planar
boundary separates two sides (A and B) of a container, one containing a solution of
dye in water and the other containing pure water. Next we remove this boundary
without creating any disturbance at the interface of the two liquids. On a molecular
level the dye molecules move randomly and during this process some fraction of
them will cross over from side A to side B of the container. The same fraction of
water molecules will cross over from side B to side A. However, as there are more
dye molecules in side A the random nature of their motion dictates that there will
be a net movement of dye from side A to side B of the container. The same is true
for the water molecules moving from B to A. There will follow a transfer of mole-
cules down the concentration gradients (see Fig. 1b) between the two sides of the
container until the dye is thoroughly mixed with the water to produce a homogeneous
solution (see Fig. 1c). A more succinct way of expressing this phenomenon is to say
that diffusion of both species has occurred in the direction of decreasing concentra-
tion (activity) of that species.
Figure 1 Representation of dye diffusion.

Mathematical Principles in Skin Permeation 63
The foregoing series of diffusional events occurs because the system in question
is not at equilibrium, and the laws of thermodynamics dictate that it must move
toward such a state. This irreversible tendency toward the achievement of a lower-
energy state arises as the result of increased entropy within the system and can be
expressed in terms of a net decrease in the Gibbs free energy of the system where,
under isothermal conditions,
⌬G = ⌬H ⫺ T⌬S
where ⌬G represents the free energy change, ⌬S is the decreasing entropy of the
system, T the temperature (in this case constant), and ⌬H the change in enthalpy.
The process of diffusion is driven primarily by the increase in entropy associated
with movement toward a more disordered (mixed) system, but in nonideal cases
(which are very common) is accompanied by some change in enthalpy.
In the example discussed thus far we have examined the interdiffusion of two
species (i.e., the movement of two mobile phases into each other). The purpose of
this chapter is to examine the diffusion of one mobile phase (the permeant) into a
second stationary phase (the membrane). Here, we may view the membrane as a
fixed plane of reference and consider only the flux of the permeant into it. To un-
derstand this in terms of the bidiffusional process, described earlier, we can consider
the void space and other mobile molecules within a membrane to be the second
diffusing substance.
A. Fick’s First Law of Diffusion

In transport, the flow (or flux, Ji in mol cm⫺2 s⫺1) is related to the velocity of
molecular movement (v in cm s⫺1) and the concentration (Ci in mol cm⫺3) of the
molecules in motion, Eq. (1).
Ji = Civ (1)
A fundamental principle of irreversible thermodynamics is that ‘‘the flow, at
any point in the system, at any instant, is proportional to the appropriate potential
gradient.’’ This was outlined briefly in the consideration of the foregoing dye ex-
periment, and it can be expressed mathematically for a species i as shown in Eq. (2)
where ⭸␮i /⭸xi is the gradient and Li is some proportionality constant.
Ji = ⫺Li 冉冊
⭸␮i
⭸x
(2)
Equation 2 is a general form of Fick’s first law of diffusion.

If we assume constant temperature and pressure then we can write Eq. (3) and
begin to see how the concentration gradient (⭸Ci /⭸x) will determine flow.
冉冊冉冊冉冊
⭸␮i
⭸x
=
⭸␮i
⭸Ci
⭸Ci
⭸x
(3)
From classic thermodynamics (the reader is referred to any physical chemistry text-
book) we have Eq. (4)
␮ = ␮ 0i ⫹ RT ln ␥i Ci (4)
and
⭸␮i
⭸Ci
= RT 冋 1 ⫹ Ci 冉冊册冒
⭸ ln ␥i
⭸Ci
Ci (5)
Substitution of Eq. (5) and Eq. (3) into Eq. (2) gives
Ji = ⫺
L iRT
Ci
冋 1 ⫹ Ci 冉冊册冉冊
⭸ ln Ci
⭸Ci
⭸Ci
⭸x
(6)
The general expression for the force Fi , acting on a molecule is given by Eq. (7).
Fi = ⫺N A⫺1 冉冊
⭸␮ i
⭸x
(7)
Rearrangement and substitution of this into Eq. (2) gives,

Ji = ⫺NA L i Fi = Civ (8)
which, on further rearrangement yields,
Ci
Li = (9)
NA fi
where fi = v/Fi and is defined as the frictional coefficient.
Substitution of Eq. (9) into Eq. (6) produces
Ji = ⫺
RT
NA fi 冋1 ⫹ Ci 冉冊册冉冊
⭸ ln ␥i
⭸Ci
⭸Ci
⭸x
(10)
that is,
Ji = ⫺Di 冉冊
⭸Ci
⭸x
(11)
where
Di =
RT
NA fi 冋 1 ⫹ Ci 冉冊册⭸ ln ␥i
⭸Ci
(12)
Equation 11 is the classic form of Fick’s first law of diffusion.

As pointed out in the discussion of the dye experiment, Eq. (11) shows that
diffusion will cease when the concentration gradient is zero (i.e., when the concen-
tration in the system is uniform). Note, that Di , the diffusion coefficient, is a function
of RT/NA which is equal to the molecular kinetic energy (kBT ) of the system. Also,
although there is a dependency of Di on concentration owing to solute–solute inter-
actions, it is, in practice, small, as the effect of concentration decreases with pro-
gressively dilute solutions (as Ci → 0, Di → RT/NA).
B. Fick’s Second Law of Diffusion

Fick’s second law relates the rate of change in concentration with time at a given
point in a system to the rate of change in concentration gradient at that point. Under
non–steady-state conditions we must consider the principle of conservation of mass
to describe the transport taking place. Consider a thin section of cross-sectional area
A and thickness ⌬x (see Fig. 1) that has a concentration C at position x and time t.
The amount of diffusing substance (moving from left to right) that enters the slab
per unit time is Jin A, where Jin is the flux and, therefore, the increase in concentration
inside the section owing to this influx of material (which has a volume A⌬x) is given
by Eq. (13).
dC Jin A Jin
= = (13)
dt A⌬x ⌬x
However, material is also leaving this section with a finite flux that we will call Jout
and the resulting change in concentration can be expressed as shown in Eq. (14).
dC Jout A Jout
=⫺ =⫺ (14)
dt A⌬x ⌬x
The difference between Eqs. (13) and (14) is equal to the net rate of change of
concentration in the section,
dC Jin ⫺ Jout
= (15)
dt ⌬x
If we now use Fick’s first law [see Eq. (12)] to describe these fluxes, we can now
write Eq. (16).
dCin dCout
Jin ⫺ Jout = ⫺D ⫹D
dx dx
= ⫺D
dCin
dx
⫹D
d
dx
冉冉冊冊
Cin ⫹
dCin
dx
⌬x
d2C
= D⌬x (16)
dx 2
Substituting this into Eq. (15) yields Fick’s second law of diffusion [Eq. (17)].
⭸C ⭸2C
=D (17)
⭸t ⭸x 2
Equation 17 contains partial derivatives because C is a function of both x and t.
C. Solutions to Fick’s Laws

Fick’s laws are of more applicability if we can specify certain parameters or bound-
aries within which to apply them. These boundary conditions allow us to be more
exacting in defining the problem with which we are dealing and further enable us
to solve some interesting diffusion problems. There are many published solutions to
these laws of diffusion that cover a wide variety of different diffusional problems.
It is impossible to cover such a plethora of solutions in a single chapter, but the
solutions presented in the following section are some of those more widely applicable
in the laboratory. It is important that the reader is familiar with the need that boundary
conditions be obeyed experimentally, or at least, be aware of those conditions used
in the derivation of any model they choose to use. This will enable them to use these
solutions to their full extent; but at the same time, to appreciate their limitations.
1. Diffusion Through a Homogeneous Membrane with a Constant

Activity Difference and a Constant Diffusion Coefficient
Of the solutions determined for Fick’s second law, this situation is possibly the
closest to that used experimentally for the determination of diffusional phenomena.
The mathematical boundary conditions imposed are those of a well-designed diffu-
sion experiment when a permeant is at a high, fixed activity on one side of an inert
homogeneous membrane through which it diffuses into a sink on the other side.
Before the start of the experiment the membrane is entirely devoid of permeant.
This solution was demonstrated by Daynes (2) and Barrer (3) for which the
diffusive flow begins at the high-concentration side (the donor side) of the membrane
where C = C0 and x = h at all time t. There is no diffusant material within the
membrane before ingress of the permeant being modeled, implying that at t = 0 we
have C = 0 for all x. Diffusion occurs in the direction of decreasing x toward the
opposite side of the membrane where x = 0 and C = 0 (sink receptor phase) for all
time t. In this model the diffusion coefficient of the permeant is set as a constant D,
and the concentration C, of material at any point x, within the membrane can be
calculated as a function of time using Eq. (18).
冘冉冊冉冊
⬁
C0x 2 C0 n␲ x ⫺Dn 2␲ 2t
C= ⫹ cos(n␲)sin exp (18)
h ␲ n=1 n h h2
In the literature the form of diffusion equations, such as Eq. (18) is often simplified
by normalizing the concentration and distance variables relative to their maxima. In
Eq. (18), this involves normalizing x (distance) relative to h (membrane thickness)
where ␹ = x/h and C (concentration in the membrane at any point x) relative to C0
(concentration in the outer layer of the membrane at x = h) where u = C/C0. It is
also the norm to introduce the term ␶ = Dt/h2. These simplifications yield Eq. (19)
in which we can see that, for any value of ␶, if x = 0 (i.e., at the distal side of the
membrane), we obtain u = 0.
冘
⬁
2 1
u=␹⫹ cos(n␲)sin(n␲␹)exp(⫺n2␲ 2␶) (19)
␲ n=1 n
If we now take a value of ␶ = 0.1 and generate a plot of u (normalized concentration)

against ␹ (normalized distance) that Eq. (19) produces, we obtain the graph depicted
in Figure 2.
Although Eqs. (18) and (19) are often presented in these forms (4,5), they are
more easily understood if rewritten using slightly different boundary conditions.
The boundary conditions used in the derivation of Eq. (18) state that diffusion
occurs in the direction of decreasing x, resulting in the graphic form of this equation
shown in Figure 2. This mathematical quirk makes the application of Eq. (18) to
real situations a little difficult to visualize. To make the situation more ‘‘palatable,’’
we can simply reverse the form of Eq. (19), making the values of ␹ increase from
zero to unity by expressing the function as in Eq. (20). This situation is shown in
Figure 3 and has obvious conceptual advantages over the graph produced by Eq. 19
(i.e., we normally think of diffusion occurring in the direction of positively increas-
ing, rather than negatively decreasing, distance into a membrane.
Figure 2 The concentration–depth profile of a permeant within a membrane, as described

by Eq. (19) where ␶ = 0.1.
冘
⬁
2 1
u=1⫺␹⫹ cos(n␲)sin[n␲ (1 ⫺ ␹)]exp(⫺n2␲ 2␶) (20)
␲ n=1 n
Equation (20) is often presented in the literature (6,7) in the form shown in Eq. (21).
冘
⬁
2 (⫺1)n
u=1⫺␹⫹ sin[n␲ (1 ⫺ ␹)]exp(⫺n2␲ 2␶) (21)
␲ n=1 n
The difference between Eqs. (20) and (21) arises because the two equations use
different methods of making the sign of the alternate terms in the summation flip
from positive to negative [Eq. (22)].
Figure 3 The concentration–depth profile of a permeant within a membrane described by

Eq. (20) where ␶ = 0.1.
冘冘
⬁ ⬁
(⫺1)n 1
= cos(n␲) (22)
n=1 n n=1 n
It is clear from Eq. (21) that as t → ⬁ the exponential term will approach zero thus
reducing Eq. (21) to the simple form of Eq. (23).
u=1⫺␹ (23)
This is obviously a linear function and represents the concentration gradient that
occurs within a membrane once diffusion has reached steady state. Figure 4 depicts
Eq. (21) for increasing values of ␶ (i.e., it shows the pattern of buildup of a penetrant
within a membrane with increasing time; namely, the imperfection in the smoothness
of the some of the graphs is simply due to the relatively low number of points
calculated. As steady state is reached the distribution pattern becomes linear, as
represented by Eq. (23).
Attempts have been made to utilize expressions such as Eq. (21) to analyze
data acquired by the skin-stripping technique (7). This method of analysis has been
taken a step further by separately modeling concentration gradients across the stratum
corneum and viable epidermis (8).
Although useful, equations such as those discussed in the foregoing, are of
limited practical use for interpreting permeation data, as they describe the concen-
tration within a membrane at any time point t, and at any position x, within that
membrane. A more useful solution that yields the cumulative mass Q, of permeant
that passes through a unit area of a membrane in a time t is provided by the following
mathematical steps. By differentiating Eq. (21) relative to x, we obtain an expression
that describes the instantaneous concentration gradient. The flux dM/dt is then de-
termined at x = h and subsequent integration (between t = 0 and t = t) of this
expression produces Eq. (24), which describes the increase in Q relative to t.
冋冘冉冊册
⬁
Dt 1 2 (⫺1)n ⫺Dn2␲ 2t
Q = C0h ⫺ ⫺ 2 exp (24)
h2 6 ␲ n=1 n2 h2
Figure 4 Representation of a concentration–depth profile as predicted by Eq. (21) for

different values of ␶.
A graphic representation of Eq. (24) (Fig. 5) depicts the time-dependent nature

of the total mass transfer Q, through the membrane. As t → ⬁, the exponential term
in Eq. (24) tends to zero; therefore, Eq. (24) approximates to the line described by
Eq. (25) which, on rearrangement, becomes Eq. (26). Figure 5 shows the line de-
scribed by Eqs. (25) and (26) in relation to the full diffusion curve described by Eq.
(24).
Q = C0h 冋册
Dt
h2
⫺
1
6
(25)
Q=
DC0
h
冋册
t⫺
h2
6D
(26)
If we put Q = 0 into Eq. (26) we can solve for t and this yields the value of the
time axis intercept known as the lag time (tlag) as described by Eq. (27), which relates
it inversely to the diffusion coefficient and directly to the diffusional pathlength. The
use of this extrapolation for the calculation of diffusion coefficients is commonplace,
although when skin is concerned, the pathlength is unknown, making calculation of
absolute values of D difficult.
h2
tlag = (27)
6D
If we differentiate Eq. (26) relative to time we obtain Eq. (28), possibly the most
well-known form of Fick’s first law of diffusion, that describes the flux J, at steady
state. Note the link between the steady-state form of this law [Eq. (28)] and the more
general form that we derived earlier [Eq. (11)].
dQ DC0
=J= (28)
dt h
Figures 6 and 7 demonstrate the effect that changes in the values of the diffusion
Figure 5 Plot of Eq. (24) and Eqs. (25) and (26) representing diffusion through a mem-
brane with increasing time.
Figure 6 Effect of altering the diffusion coefficient on theoretical permeation profiles gen-
erated by Eq. (24) (h = 0.03, C0 = 100).
coefficient and the diffusional pathlength have on such plots. It is clear that a reduced
diffusion coefficient results in slower mass transfer, as does an increased diffusional
pathlength.
It is often impractical to use the forms of Eqs. (25), (26), and (28) as shown
because they include a term, C0 (the concentration of permeant in the outer layer of
the membrane), that is extremely difficult to measure. We can replace the value of
C0 with a term that links it to the concentration in the vehicle Cv through the partition
coefficient K, as described by Eq. (29) which rearranges to Eq. (30). The partition
Figure 7 Effect of altering the diffusional pathlength on theoretical permeation profiles

generated by Eq. (24) (D = 0.00001, C0 = 100).
coefficient is simply a measure of the relative affinity that a diffusant has for the
two media involved (i.e., the membrane and the vehicle above it).
C0
K= (29)
Cv
C0 = KCv (30)
Substitution of Eq. (30) into Eq. (28), for example, produces Eq. (31), which is of
more practical use, as it links flux to the concentration of the permeant in the vehicle.
dQ DKCv
=J= (31)
dt h
We have already examined one derivation of the steady-state form of Fick’s
first law but we can approach the solution from a slightly different angle. Consider
a plane sheet, of thickness h, for which surfaces, x = 0, x = h, are maintained at
constant concentrations C0 and Ch, respectively. If the diffusion process has reached
steady state, thus producing a constant concentration gradient across it at all points,
provided that D is constant, using Fick’s second law of diffusion [see Eq. (17)] we
obtain the situation described in Eq. (32)
dC d 2C
=D =0 (32)
dt dx 2
(i.e., the rate of change of the concentration gradient across the membrane is zero).
This situation is known as steady state. If we now integrate Eq. (32) relative to x
once, we obtain Eq. (33).
dC
D = cons tan t (33)
dx
Equation 33 shows us that the concentration across the membrane reduces linearly
with distance (here from C0 to Ch) and, therefore, that the rate of transfer of diffusant
(the flux, J) is the same at all positions within the membrane. This rate is given by
Fick’s first law [see Eq. (11)] and we can now write
dC D(C0 ⫺ Ch)
J=D = (34)
dx h
If we know the membrane thickness h and the concentrations C0 and Ch, then we
can measure D from the determination of the flux J. Indeed, it is often assumed (by
the use of sink conditions in the receptor solution below the membrane) that the
value of Ch, the concentration at the inner surface of the membrane is zero. Thus,
the problem is simplified further and Eq. (34) reduces to Eq. (35), which is identical
with Eq. (28) derived earlier.
DC0
J= (35)
h
However, as it is difficult to measure the concentration C0 (the amount of diffusant
in the outermost layer of the membrane) we perform the same trick as earlier and
use the partition coefficient and the vehicle concentration to give us Eq. (31).
Frequently, particularly in biological membranes, there is a practical difficulty

in measuring the diffusional pathlength. Because of this, and that information con-
cerning the individual effects of changes in K and D is often not required, a com-
posite parameter is usually used to replace these values in Eq. (26). The permeability
coefficient P, is thus defined as P = KD/h, and this simplifies Eq. (31) further to
give Eq. (36):
J = PCv (36)
Equation (36) is perhaps the most basic and frequently used expression in the routine
assessment of membrane permeability. However, it should always be noted that the
principles on which this equation are based stipulate that the donor concentration is
constant and that the diffusion has reached steady state. In practice this either means
using a saturated donor solution in the presence of excess permeant, or that the
change in donor concentration during the course of the experiment is negligible.
There are also problems associated with the assumption of steady state and the
assessment of when it has been attained.
The use of the steady-state method for the assessment of permeability coeffi-
cients and lag times has been questioned (9,10). It is often difficult to assess when
this period has been reached and, by using only this steady-state data, we lose out
on that data collected at time points before this region. One way to partially solve
this is to utilize Eq. (24) and fit it to the entire set of diffusion data using a nonlinear
curve-fitting software package. If the values of C0 (the permeant concentration in the
outer layer of the membrane) in Eq. (24) are replaced by the term KCv (as described
earlier) we obtain Eq. (37).
冋冘冉冊
⬁
Dt 1 2 (⫺1)n ⫺Dn 2␲ 2t
Q = KCv h 2 ⫺ ⫺ 2 exp (37)
h 6 ␲ n=1 n2 h2
The equation, as shown, is of limited use, as there are three unknown parameters,
but if we replace Kh and D/h2 with P1 and P2, respectively, we obtain Eq. (38) with
just two variables. By fitting P1 and P2 we can obtain a value of the permeability
coefficient as P1P2 = KhD/h2 = KD/h, which is equal to the permeability coefficient.
冋冘册
⬁
1 2 (⫺1)n
Q = P1Cv P2t ⫺ ⫺ 2 exp(⫺P2n 2␲ 2t) (38a)
6 ␲ n=1 n2
Equations (37) and (38a) have been used extensively for the calculation of
diffusional parameters (e.g., Ref. 11). A similar use of a non–steady-state approach
(12) has been used where the partition coefficients were experimentally measured,
and the diffusion coefficient and pathlength estimated using a fitting routine.
It is also possible to use a short-time approximation derived by integrating a
Fourier transformation (see Eq. 37) that yields an expression valid at small t (see
Eq. 38b) (13,14).
log 冋册冋册
Q
t 3/2
= log
8KCv
h2␲ 1/2
⫹
3
2
log D ⫺
h2
9.2Dt
(38b)
By using Eq. (38b) we can construct a plot of log(Q/t 3/2) against 1/t that has
a gradient of ⫺h2/9.2D (from which D can be calculated) and an intercept on the y-
axis that gives an estimate of the remaining parameters. The short-time method is
valid up to approximately 2.7 times the lag time. The major drawback of this method
is the requirement for excellent analytical sensitivity to collect good data at very
short time periods. Note that, as with all the other equations derived in this section,
this method is valid only if the diffusion coefficient is constant.
Equation (38b) is identical with that derived by Hadgraft (6), who presented it
in a slightly different form [Eq. (39)] where C0 = KCv as outlined earlier in Eq. (30).
Q=
8C0 D 3/2t 3/2
h3␲ 1/2
exp 冉冊
⫺h2
4Dt
(39)
2. Diffusion into a Homogeneous Membrane with a Constant Donor

Activity, Constant Diffusion Coefficient, and an Impermeable
Distal Side
A further solution to Fick’s second law that has essentially the same boundary con-
ditions as those discussed in the derivation of Eq. (18), except the distal side of the
membrane is an impermeable surface, has recently been used extensively for the
analysis of absorption data collected by attenuated total-reflectance Fourier transform
infra red spectroscopy (ATR–FTIR).
The methodology of ATR–FTIR systems has been described by various authors
in diffusion studies on polymers (15–19), semisolids (20), glycerogelatin films (21),
and stratum corneum (22). Briefly, a membrane is sandwiched between an imper-
meable zinc selenide ATR crystal, and a reservoir of permeant that provides a con-
stant concentration in the upper surface of the membrane. The membrane is initially
devoid of permeant. As diffusion into the membrane occurs, there will be a buildup
of permeant at the impermeable membrane–crystal interface until saturation is
reached at a concentration C0 (the solubility of the penetrant in the membrane). An
analytical solution describing the buildup of permeant at this interface with time can
be obtained, using Fick’s second law and the relevant initial and boundary conditions
(1), and is given as Eq. (40)
冘冉冊
⬁
C 4 (⫺1)n ⫺D(2n ⫹ 1)2␲ 2t
=1⫺ exp (40)
C0 ␲ n=0 2n ⫹ 1 4h2
where C is the diffusant concentration at the interface, t is time, D is the diffusion
coefficient of the permeant, and h is the membrane thickness. There will be an initial
period during which permeant concentration at the interface increases, followed by
an exponential rise to a plateau that represents the saturation of the membrane with
permeant. Plots of C against time are given in Figures 8 and 9 for different values
of D and h, respectively.
For large values of time, the n = 0 term in Eq. (40) predominates, and the
long-time approximation of Eq. (40) is given by Eq. (41).
ln 冋冉
␲
4
1⫺
C
C0
冊册冉冊
=
␲ 2D
4h2
t (41)
Equation 41 may be fitted to experimental data in a plot of ln[(␲/4)(1 ⫺ C/C0)]

against time. The diffusion coefficient D, may then be obtained from the gradient of
this plot and the film thickness h.
It is possible, assuming the Beer–Lambert law applies, to replace the concen-
tration terms in Eq. (40) with experimental absorbance values to give Eq. (42) in
Figure 8 Increase in concentration at the distal face of a membrane supported on an

impermeable surface as described by Eq. (40) (effect of altering D, h = 0.03).
which A = area of penetrant peak (at time t) of IR absorbance relating to the per-
meant, and A0 = area of penetrant peak corresponding to the situation where the
membrane is saturated (in the plateau region of the curve).
冘冉冊
⬁
A 4 (⫺1)n ⫺D(2n ⫹ 1)2␲ 2t
=1⫺ exp (42)
A0 ␲ n=0 2n⫹ 1 4h2
Experimental values of penetrant peak areas against time are fitted using Eq.
(42). Values of D/h2 and A0 are allowed to vary until a best fit is achieved as measured
Figure 9 Increase in concentration at the distal face of a membrane supported on an

impermeable surface, as described by Eq. (40) (effect of altering h, D = 0.00001).
by minimization of ␹ 2. This technique has great potential for the deconvolution of

partitioning and diffusional phenomena in the synergistic action of certain penetration
enhancers (22,23). A limitation of this technique is that the diffusant under investi-
gation must have an absorption band in its IR spectrum that can be distinguished
from the IR absorption spectrum of the test film. Compounds containing cyano and
azo groups are particularly useful in this respect.
3. Diffusion Out of a Precharged Homogeneous Membrane, With a
Constant Diffusion Coefficient, With and Without a Reservoir
This scenario is applicable to controlled-release devices that may include transdermal
patches, as well as oral and subcutaneous devices. Equations describing release pro-
files from devices that use a rate-controlling membrane have been derived by Had-
graft (6) for two situations: first, when a rate-controlling membrane is initially full
of material, but no reservoir is present and, second, when an infinite reservoir exists.
The solution derived by Hadgraft (6) describing the burst effect release into a
receiver fluid from a membrane for the situation without a reservoir is given in Eq.
(43) (Mt = amount of drug released at time t, M⬁ = total amount of drug contained
in the membrane at t = 0, the other parameters are as already described.
冋冉冘冊冉冊册
⬁
8 1 (2n ⫺ 1)2␲ 2Dt
Mt = M⬁ 1⫺ 2 exp ⫺ (43)
␲ n=1 (2n ⫺ 1)2 4h2
A second, but identical, form of this equation is often quoted (e.g., Ref. 1) and
follows as Eq. (44). The only difference between Eqs. (43) and (44) is in the way
the summation term is calculated. In Eq. (43) the sum runs from unity upward,
whereas in Eq. (44) it runs from zero. this has the effect of producing a slightly
different premultiplier and exponential term.
冋冉冘冊冉冊册
⬁
8 1 (2n ⫹ 1)2␲ 2Dt
Mt = M⬁ 1⫺ 2 exp ⫺ (44)
␲ n=0 (2n ⫹ 1)2 4h2
Note, Eq. (43) is the full solution for the burst effect without a reservoir and may
be applied over the complete time range of an experiment.
Hadgraft (6) also simplified this expression to give both short [see Eq. (45)]
and long [see Eq. (46)] time approximations (clearly, at long times the amount of
diffusant released is equal to that initially in the membrane).
冉冊
1/2
Dt
Mt = 2M⬁ (45)
␲h2
Mt = M⬁ (46)
Equation (47) (6) is the solution describing the buildup of diffusant in the receptor
compartment for the case in which the membrane was initially charged with permeant
and there was a reservoir present. Here, M⬁ is not a finite quantity of drug, for it
represents the content of the unchanging reservoir.
冋冘冉冊册
⬁
Dt 1 2 1 ⫺n 2␲ 2Dt
Mt = M⬁ 2 ⫹ ⫺ 2 exp (47)
h 3 ␲ n=1 n2 h2
Again, short- and long-term approximations can be made. The short-term approxi-
mation of Eq. (47) is identical with Eq. (45), whereas to derive the long-term ap-
Figure 10 Diffusion from a presaturated slab with and without a reservoir, together with
the long-time approximation as in Eq. (48) (D = 0.00001, h = 0.03, M⬁ = 100).
proximation, we simply note that large t will make the negative exponential term in
Eq. (47) tend to zero. This solution is given in Eq. (48).
Mt = M⬁ 冉Dt
h2
⫹
1
3 冊 (48)
The depletion of a membrane that is not in contact with a reservoir of permeant

[see Eq. (42)] is clearly demonstrated in the graphic representation of Eqs. (43) and
(47) (Fig. 10). Note that the long-time approximation of Eq. (48) produces a straight-
line with a y-axis intercept equal to one-third of the value of M⬁ used, and has a
gradient equal to DM⬁ /h2, as would be predicted from its form.
A point that we have not considered so far is that many of the solutions ex-
amined contain summation expressions and that the value of n used within these will
affect the result achieved. This is clearly demonstrated in Figure 11, in which n in
Eq. (43) has been made equal to 10, 20, and 500. The most dramatic differences are
seen at low values of t. This serves to demonstrate that the researcher should assess
the effect of differing values of n on the result before applying any of the equations
presented herein. The value that is used will alter the computational time required,
depending on the simulation software being used. It is suggested that n = 100 is a
reasonable approximation in most cases.
D. Complex Barriers
1. Diffusion Through Laminates
There are numerous examples for which a membrane is not a single homogeneous
system and, indeed, may consist of several barriers. For example, the skin can be
viewed at different levels of complexity that, as they are probed, reveal more and
more barriers in series with one another. At its simplest the skin might be viewed
as a single homogeneous slab, but the structure can also be viewed as three barriers
Figure 11 The effect of changing the value of n in the summation term of Eq. (43) (D =
0.00001, h = 0.03, M⬁ = 1).
in series (i.e., the stratum corneum, epidermis, and dermis). Furthermore, the stratum
corneum contains bilayer structures that might also be considered to exhibit the
properties of barriers in series. The simplest starting point in this argument is to
assume that the constituent membranes that make up the laminate are themselves
isotropic and that each layer ‘‘i’’ of the laminate contributes a diffusional resistance
Ri to the overall resistance of the membrane (24). R is obviously inversely propor-
tional to the diffusivity within the layer and the partition coefficient relative to ex-
ternal phases. The shorter the pathlength, the lower the resistance will be; thus,
hi 1
Ri = = (49)
Di K i Pi
where Pi is equal to the permeability coefficient, as defined earlier. The overall
resistance of the laminate system (Rtotal ) is equal to the sum of the individual resis-
tances of the n membranes that make up the laminate, giving,
冘
i=n
h1 h2 hn
Rtotal = Ri = ⫹ ⫹ ⭈⭈⭈ ⫹ (50)
i=1 D1K1 D2K2 Dn Kn
Importantly, the partition coefficients in Eq. (50) relate the concentration of permeant
in the i th phase of the laminate to the concentration in the initial phase and not the
(i ⫺ 1)th phase.
In the skin we may, as alluded to earlier, consider it to be a trilaminate of
stratum corneum, epidermis, and dermis. By using Eq. (50) to describe its resistance
we would therefore arrive at Eq. (51).
hsc he hd
Rskin = ⫹ ⫹ (51)
Dsc Ksc De Ke Dd Kd
If we now take the view that if the stratum corneum is the rate-limiting barrier, then
the resistance of the other two layers becomes negligible and we obtain Eq. (52),
which is not very surprising because we already know that for a single isotropic
membrane we have P = KD/h.
hsc
Rskin = (52)
Dsc Ksc
This approach to diffusion in laminates can be used to look at the situation
when we have an isotropic lipophilic membrane sandwiched between two aqueous
phases (for example, a donor and receptor phase in a Franz diffusion experiment).
Even though it is usual practice to vigorously stir receptor solutions, unstirred aque-
ous layers will form at stationary surfaces in both donor and receptor compartments
(the thickness of these layers being related to the efficiency and rate of stirring). This
situation approximates a trilaminate system of aqueous–lipophilic–aqueous barriers.
hence, we can write Eq. (53) where Raq1, is the resistance of the stationary aqueous
barrier above the membrane, Rlip the resistance of the membrane itself and Raq2 the
resistance of the stationary aqueous layer beneath the membrane.
haq1 hlip haq2
Rtotal = Raq1 ⫹ Rlip ⫹ Raq2 = ⫹ ⫹ (53)
Daq1 Dlip K Daq2
Only one partition coefficient K, appears in Eq. (53) because the partition coefficient
between the bulk aqueous regions and the unstirred diffusion layers is considered to
be unity. Thus, K is the partition coefficient between the bulk aqueous phase and the
membrane.
This type of system has been rigorously analyzed (4,25,26) and, at steady state,
under zero-order boundary conditions, yields Eq. (54).
dQ
dt
= 冋 KDm Daq
hm Daq ⫹ (haq1 ⫹ haq2)KDm 册 (Cv ⫺ Cr) = Ptotal ⌬C (54)
In Eq. (54) we have dQ/dt = steady-state flux; K = membrane/water partition coef-

ficient; subscripts m and aq refer to the membrane and aqueous layers, respectively;
h = thickness of layer, as indicated by subscript, Cv = vehicle or donor phase con-
centration; and Cr = receptor phase concentration. It is clear that in most cases the
term hm Daq >> (haq1 ⫹ haq2)KDm leading to a simplification of Eq. (54) to the com-
monly used equation for mass transfer at steady state; that is, we end up with dQ/
dt = KDm ⌬C/hm. However, with a highly lipophilic penetrant, or for extremely thin
membranes, this empiricism may not hold. Furthermore, if both of these scenarios
occur and we also have large values of Dm that are similar in magnitude to Daq the
permeability coefficient will be described by Eq. (55), and the diffusion process will
be almost completely controlled by the aqueous boundary layers.
Daq
Ptotal = (55)
(haq1 ⫹ haq2)
In our model we have seen that, as the permeant partition coefficient rises, the
diffusion process becomes more and more under the control of the aqueous layers.
The structure of skin can be approximated to a laminated structure in which the
layers increase in hydrophilicity as the skin is permeated. Thus, in skin our ‘‘receptor
phase’’ aqueous boundary layer is replaced by the aqueous epidermal region below
the stratum corneum. Hence, Eq. (55) can be applied and we observe a change from
lipoidal membrane diffusion control to aqueous diffusion control on increasing the

lipophilicity of the permeant. This is a simple demonstration of the situation in which
the stratum corneum is no longer the major barrier to percutaneous penetration. This
phenomenon has been demonstrated by Roberts et al. (27), who showed that on
increasing the lipophilicity of a homologous series of phenolic compounds there was
a permeability dependency on partition coefficient (or lipophilicity) only until aque-
ous diffusion layer control took over.
By using the definition of lag time (tlag = h2/6D) already derived, we can define
the lag time for the situation in which we have a very thin membrane (infinitely thin
membrane) as shown in Eq. (56).
(haq1 ⫹ haq2)2
tlag = (56)
6Daq
Equation (57) is the result we obtain for a very thick membrane when K is
large. If the diffusion layers are of equal thickness we will obtain the expression
shown in Eq. (58).
冘
hm haq1 haq2 K
tlag = (57)
haq Daq
hm haq K
tlag = (58)
2Daq
The use of the laminate principle can be taken further if it is assumed that the
diffusional process occurs as a series of point-to-point movements of the diffusing
species. Most of the foregoing mathematics view permeability as a rate process that
contains contributions from both an equilibrium and a nonequilibrium step, but the
assumption of instantaneous partitioning of a solute is not always valid and a treat-
ment based on Eyring’s absolute reaction rate theory attempts to account for this
(28–30). If the flow of molecules through a membrane is viewed as a series of
successive molecular jumps of length ␭ from one energy minimum to another, the
whole diffusion process can be seen in terms of an energy profile (Fig. 12).
Thus, a third factor is introduced into the mathematical approach to diffusion
(i.e., the activation energy ⌬Ga required to prevent spontaneous partitioning from
occurring). If Ci is the concentration (molecules per milliliter) at the i th position in
the membrane, then the amount of material in a 1-cm2 cross-section and length ␭i
(the distance between equilibria mazima) is Ci ␭i , and the velocity of forward diffu-
sion is
vf = ki Ci ␭i (59)
where ki = rate constant for crossing barrier i.

Similarly the rate of backward diffusion over the barrier i will be
vb = ki⫹1Ci⫹1␭i⫹1 (60)
If it is assumed that ki = ki⫹1 = k and ␭i = ␭i⫹1 = ␭ then the net rate of diffusion, or
flux is
Figure 12 (a) A schematic representation of membrane interfaces and the phases involved
in solute transfer; (b) possible potential energy profile for a solute molecule diffusing through
a membrane.
J = ki Ci ␭i ⫺ k⬘i⫹1Ci⫹1␭i⫹1 (61)
⫺k␭(Ci⫹1 ⫺ Ci)
J = k␭(Ci ⫺ Ci⫹1) =
␭
k␭2dC
J= (62)
dt
where the diffusion coefficient, D = k␭2.

The diffusional process can be viewed as five resistances in series, two at the
interfaces, one in the membrane, and two from the bulk solvent phase on either side
of the membrane. If it is now assumed that diffusion in the bulk solvent phase is
much faster than in the membrane phase or at the interfaces, and that the energy
barriers within the membrane are the same height as are those on either side of it,
the following equation can be derived for the permeability coefficient P [for an
excellent account of this derivation see Ref. 30]:
␭ksm km
P= (63)
(2km ⫹ mkm)
where ␭ = mean jump distances, m = No. of jumps in the membrane, ksm, kms, km =
rate constants of adsorption and desorption at the interfaces, and diffusion in the
membrane, respectively.
However, the ratio ksm /kms can be defined as the partition coefficient K, giving
1/P = 2/(ksm ␭ ) ⫹ m/(km ␭K) (64)

As m␭ = ⌬x, the membrane thickness, and D = k␭ , then2
1 2␭ ⌬x
= ⫹ (65)
P Dsm Dm K
So, if diffusion in the membrane is the rate-limiting step (km << kms), then
Dm K
P= (66)
⌬x
(i.e., the same equation as produced by Fick’s first law) and, if the slowest step is
diffusion through the interface, them km >> kms, so
Dsm
P= (67)
2␭
that is, the permeability constant will be independent of the partition coefficient and
the membrane thickness. If this is true, then the rate of diffusion will be controlled
by the nature of the interface and penetrant relative to each other. This can be
expressed thermodynamically by considering the expansion of the term ksm:
Dsm ksm␭
P= = (68)
2␭ 2
and since ksm = (kT/h)e ⫺⌬Gsm /RT, then
P = (␭kT/2h)e ⫺⌬Gsm /RT (69)
where ⌬Gsm is the free energy of activation necessary for crossing the interface,
which will be dependent on the physicochemical relation between penetrant and
barrier. Thus, a compound that interacts in a way such that the value of ⌬Gsm is
reduced, will increase the rate of permeation (i.e., enhance penetration). The same
is true if the membrane is the rate-limiting medium (i.e., ⌬Gm becomes rate-deter-
mining) and the creation of free volumes within the acyl chain region by an enhancer
may reduce ⌬Gm.
It seems likely that it is a combination of these effects that is responsible for
the enhancing capacity of many compounds.
2. Diffusion Through Shunt Routes (Barriers in Parallel)
Many membranes can be viewed as containing more than one distinct route through
which diffusion can occur. The skin may be thought of as at least possessing the
potential to exhibit such parallel routes, in that its structure is pierced by numerous
appendages, such as hair follicles and sweat ducts. Indeed, although an issue of some
debate, it has been suggested that such routes are important in both passive (31–33)
and impassive (34,35) diffusion processes.
The mathematical problem of defining parallel routes of diffusion is less com-
plex than for the laminated structures we examined in the foregoing. If two diffu-
sional pathways through a single membrane are truly parallel, then the total flux
through that membrane is simply the sum of the individual fluxes through the parallel
routes (36). Thus, for unit area, the total flux (at any time) through n parallel routes
is defined by Eq. (70).
Jtotal = Ja ⫹ Jb ⫹ ⭈⭈⭈ ⫹ Jn (70)

Hence, for any situation (whether steady state or otherwise), we can simply sum the
expressions describing flux through each route to gain the overall flux. The simplest
example would be steady-state flux through two parallel routes (A and B) in which
the diffusion coefficients are DA and DB, the pathlengths hA and hB, and the partition
coefficients KA and KB. This situation is described by Eq. (71), in which the driving
concentration is C0 in the common donor solution, and PA and PB are the permeability
coefficients for the two routes. Note that the flux expression is for unit area.
Jtotal = JA ⫹ JB =
KA DA C0 KB DB C0
hA
⫹
hB
= C0
hA
冋
KA DA KB DB
⫹
hB
册
= C0[PA ⫹ PB] (71)
The situation is more complex if diffusion has not yet reached steady state, but the
total flux is still just the sum of the flux through the two routes. Equation (72)
demonstrates the increasing degree of complexity that this situation yields for the
same situation as that described in Eq. (71), but before a steady state has been
reached.
再冋冘冉冊册冎
⬁
DA t 1 2 (⫺1)n ⫺DA n2␲ 2t
Q= KA C0hA 2 ⫺ ⫺ 2 exp
hA 6 ␲ n2 h 2A
再冋冘冉冊册冎
n=1
⬁
DB t 1 2 (⫺1)n ⫺DB n 2␲ 2t
⫹ KB C0hB 2 ⫺ ⫺ 2 exp (72)
hB 6 ␲ n=1 n2 h 2B
In a manner analogous to that used earlier we can derive the long-time approximation
of Eq. (72) as shown in Eq. (73) and use this to define an expression for the lag
time [Eq. (74)].
Q= 再
KA C0 hA 冋
DA t
2 ⫺
hA
1
6
册冎再
⫹ KB C0 hB 冋 DB t
2 ⫺
hB
1
6
册冎 (73)
hA hB (KA hA ⫹ KB hB)
tlag = (74)
6(KA DA hB ⫹ KB DB hA)
If the two diffusional pathlengths for routes A and B are the same then Eq. (74)
reduces to Eq. (75).
h2(KA ⫹ KB)
tlag = (75)
5(KA DA ⫹ KB DB)
The shape of a plot of amount permeated versus time will obviously be dependent
on the relative ease by which a permeant can pass through the parallel routes. If one
of the routes is significantly more permeable than the other, there will be an initial
period where diffusion appears regular (at which time flux is occurring by a single,
more permeable, route only) before an increase in flux occurs as material starts
issuing from the second route. This increase in flux will be followed by a period in
which steady state is re-established (37).
An interesting example of parallel routes of diffusion are the pores that exist
in some membranes. If these are angled or tortuous, then there is an obvious length-
ening of the distance that a permeant must travel (i.e., the diffusional pathlength h
is increased). This will manifest itself in all the parameters that are dependent on
the membrane thickness, which is usually taken as the diffusional pathlength. To
deal with this situation a variable, referred to as the tortuosity factor (␶) is introduced
as a premultiplier to the membrane thickness (i.e., the new, increased, diffusional
pathlength, H = ␶h).
3. Barriers with Dispersed Phases
The theory dealing with diffusion through heterogeneous systems has been addressed
by Barrer (38) in great detail for numerous types of systems. Many of these complex
systems are beyond the scope of this chapter so we will consider only those simple
cases in which the dispersion is uniform. A more comprehensive review of complex
systems (if particle size and shape are irregular and the dispersion nonuniform) is
provided by Barrer (38) and Higuchi and Higuchi (39).
Possibly the simplest case to deal with is that in which the dispersed phase is
diffusionally impervious and inert. The effect of such a dispersed phase (the filler)
will be to partially obstruct the movement of diffusants through the permeable phase
(the matrix). This will result in an increased diffusional pathlength which is addressed
mathematically by the introduction of the tortuosity factor, ␶. The addition of filler
will also reduce the amount of space within the membrane through which diffusion
can occur. This will be related to the relative amounts (volume fractions) of matrix
(␾1) and filler (␾2) that are present in the system where we have ␾1 = 1 ⫺ ␾2. In a
random dispersion the relative cross-sectional areas of the phases are equal to their
relative volume fraction. The tortuosity factor is applied to the membrane thickness
h, such that the new, increased, diffusional pathlength H equates to the term ␶h.
Hence, for a diffusional system at steady state we can use Eq. (31) and obtain
Eq. (76).
dQ KC0 D ␾1 KC0 D(1 ⫺ ␾2)
=J= = (76)
dt h ␶ H
It is clear that the lag time will reflect the longer pathlength, but it is not affected
by the volume of filler present, as demonstrated by Eq. (77).
h2␶ 2 H 2
tlag = = (77)
6D 6D
Slightly more complex situations arise if the filler has adsorptive properties.
This situation has been examined by numerous authors (39–42). Irrespective of how
the permeant molecules are adsorbed onto the filler, the steady-state flux for a stan-
dard zero-order process is still given by Eq. (76) (because the filler is not involved
in any absorption process). Hence, it is only the lag time that is affected and length-
ened by such a process. If the amount of permeant adsorbed is linearly related to
the concentration, then the lag time is described by Eq. (78) where ␬ is the adsorptive
capacity of the filler.
␶ 2h2
tlag = (1 ⫺ ␬␾2) (78)
6D
If the situation arises where the concentration is particularly high then the filler can
adsorb only a finite amount of permeant and the steady state is still represented by
Eq. (76), but the expression for the lag time approximates to Eq. (79). Note that Eq.
(77) is an approximate solution only.
tlag = 冉
1
4
⫺
␬␾2
2KC0
冊 ␶ 2 h2
D
(79)
A greater complexity arises if the system that we are dealing with contains
filler particles that are themselves permeable. An emulsion is a good example of
such a system where diffusant can pass through both the filler and its supporting
matrix. Higuchi and Higuchi (39) derived an expression [Eq. (80)] for the total
permeability of a model system of this nature, in which it was assumed that the
dispersed phase was spherical.
2P 21(1 ⫺ ␾2) ⫹ P1P2(1 ⫹ 2␾2) ⫺ GP1[(P2 ⫺ P1)/(2P1 ⫺ P2)]2(2P1 ⫹ P2)(1 ⫺ ␾1)

Ptotal =
P1(2 ⫹ ␾2) ⫹ P2(1 ⫺ ␾2) ⫺ G[(P2 ⫺ P1)/(2P1 ⫺ P2)]2(2P1 ⫹ P2)(1 ⫺ ␾1)
(80)
III. VARIABLE BOUNDARY CONDITIONS

In deriving the diffusion equations that we have examined thus far, certain parameters
were fixed to simplify the mathematics. These fixed parameters are referred to as
boundary conditions. In an ideal world it would be feasible to conduct a permeation
experiment in which all the boundaries of a system remain constant, except time and
mass transported. However, in reality, situations in which these boundaries are not
fixed (so-called moving boundary situations) are commonplace. Examples of these
moving boundaries include the evaporation of volatile formulation components from
a gel or cream as it is applied and during its time of use (which may lead to con-
centration of the permeant in the vehicle), the swelling or contraction of a delivery
device, alterations in the barrier function of the skin itself, owing to the absorption
of formulation excipients or some form of occlusion. Changes of this type may
include increases or decreases in both the diffusion and partition coefficient of the
permeant under investigation.
A. Variations in the Thickness of Applied Medicaments

The mathematics of such situations is complex, but there are partial and full solutions
to some of these issues in the literature. Guy and Hadgraft (43) have addressed the
issue of the dependence of flux from an applied medicament on its thickness and
derived short and long-time approximate solutions [Eqs. (81) and (82), respectively].
Mt =
8M⬁␭pD 3/2 t 3/2
␲ 1/2h3(K ⫹ ␭p 1/2)冉冊
exp
⫺h2
4Dt
(81)
Mt = M⬁ 冉
1 ⫺ exp 冉冊冊
⫺␭pDt
Kh2
(82)
Flynn and colleagues (44) examined this situation and arrived at a solution that
describes the situation over the full time period [Eq. (83)]. Using such expressions
it is possible to understand how such changes in an applied formation may affect
diffusion through the skin.
Mt 兹DmK
=
再冎
M⬁ (hv ⫹ hm K)s 3/2
冉冑冑
sinh 冑 s
Dv
hv cosh
s
Dm
hm ⫹ K
Dm
Dv
sinh 冑s
Dm
hm cosh 冑冊
s
Dv
hv
⭈
冉冑冑K 冑 Dm
Dv
cosh
s
Dv
hv cosh
s
Dm
hm ⫹ sinh 冑
s
Dv
hv sinh 冑冊
s
Dm
hm
(83)
The case for which the diffusion coefficient is some function of either time,
depth, or concentration, may well describe the situation in biological and indeed
synthetic membranes more accurately than that for which its value is assumed con-
stant. The movement of a diffusant, and perhaps some constituents of the vehicle in
which it is formulated, into a membrane could quite easily change the nature of the
membrane itself. One can visualize a situation in skin, for example, where the ab-
sorption of penetration enhancers over the period of an experiment may change the
value of D. Indeed, the movement of solvent molecules into the stratum corneum
may also affect the partitioning behavior of the diffusant. These phenomena will not
occur instantaneously, thus making the diffusional and partitioning steps of the pen-
etration process variables that are time-dependent. These are complicated situations
that are beyond the scope of this chapter. The interested reader is referred to Crank
and Park (45) and Crank (1) for further reading.
The possibility of there being a depth dependency of rates of diffusion through
stratum corneum has been addressed (7). If the skin is treated with a penetration
enhancer, there will be a concentration gradient of that enhancer across the skin (as
a consequence of Fick’s second law of diffusion). If the enhancer has the same effect
per lipid molecule at all depths; then it follows that the degree of enhancement will
reduce as the skin is penetrated farther. If, for the sake of argument, the enhancer
action arises from increasing the diffusion coefficient of a penetrant, we will see a
consequent reduction in its value with depth, owing to the nature of the enhancer
distribution.
B. Effect of Temperature on Diffusion Through Membranes

Temperature does not tend to affect partitioning phenomena, phase volumes, or mem-
brane thicknesses (or diffusional pathlengths) to a great extent. However, in certain
membranes containing structures that are subject to phase changes on heating or
cooling, there may be a concurrent effect on the diffusion. For example, the flux of
water through human stratum corneum increases with temperature and this increase
has been linked to an increasingly fluid (or disordered) environment within the stra-
tum corneum lipid bilayers (46). The diffusion coefficient itself may be temperature-
dependent, and the nature of this dependence can be empirically expressed in a form
analogous to the Arrhenius equations for reaction kinetics.
D = D0 exp 冋册
⫺Ea
RT
(84)
In Eq. (84) D0 represents the diffusion coefficient at infinite temperature and

can be calculated by back-extrapolating a plot of ln D versus (1/T ) to the y-intercept

[Eq. (85)].
Ea
ln D = ln D0 ⫺ (85)
RT
Ea is an activation energy term and is determined by the nature of the barrier. Al-
though Eq. (85) is empirical, it appears to hold in the temperature range over which
permeability is generally measured. This relation holds for permeability coefficients
where D and D0 are replaced in Eqs. (86) and (87) with P and P0, respectively. If a
membrane contains constituents that are subject to temperature-dependent phase tran-
sitions (as outlined earlier) it is unlikely that these equations will not hold in the
temperature region in which the transition occurs.
P = P0 exp 冋册
⫺Ea
RT
(86)
Ea
ln P = ln P0 ⫺ (87)
RT
IV. CONCLUDING REMARKS

As stated in the Introduction, the purpose of this chapter was to increase the acces-
sibility of the mathematical approach to the interpretation of diffusion processes and
thereby to promote a clearer understanding of better ways to use data in this area of
research. We would like to think that the preceding pages have demonstrated the
utility with which mathematics and the basic principles of diffusion can be used in
examining some of the processes involved in percutaneous penetration.
REFERENCES
1. Crank J. The Mathematics of Diffusion. 2nd ed. London: Oxford University Press, 1975.
2. Daynes HA. The process of diffusion through a rubber membrane. Proc R Soc Lond A
97:286–307, 1920.
3. Barrer RM. Permeation, diffusion and solution of gases. Trans Faraday Soc 35:628–
643, 1939.
4. Flynn GL, Carpenter OS, Yalkowsky SH. Total mathematical resolution of diffusion
layer control of barrier flux. J Pharm Sci 61:312–314, 1972.
5. Flynn GL. Dermal diffusion and delivery principles. In: Swarbrick J, Boylan JC, eds.
Encyclopedia of Pharmaceutical Technology. Vol. 3. New York: Marcel Dekker, pp.
457–503, 1991.
6. Hadgraft J. Calculations of drug release rates from controlled release devices. The slab.
Int J Pharm 2:177–194, 1979.
7. Watkinson AC, Bunge AL, Hadgraft J. Computer simulation of penetrant concentration–
depth profiles in the stratum corneum. Int J Pharm 87:175–182, 1992.
8. Yamashita F, Koyama Y, Sezaki H, Hashida M. Estimation of a concentration profile of
acyclovir in the skin after topical administration. Int J Pharm 89:199–206, 1993.
9. Shah JC. Analysis of permeation data: evaluation of the lag time method. Int J Pharm
90:161–169, 1993.
10. Potts RO, Guy RH. Drug transport across the skin and the attainment of steady-state
flux. Proc Int Symp Control Release Bioact Mat 21:162–163, 1994.
11. Lafforgue C, Eynard I, Falson F, Watkinson A, Hadgraft J. Percutaneous absorption of

methyl nicotinate. Int J Pharm 121:89–93, 1995.
12. Parry GE, Bunge AL, Silcox GD, Pershing DW, Pershing DW. Percutaneous absorption
of benzoic acid across human skin. I. In-vitro experiments and mathematical modelling.
Pharm Res 7:230–236, 1990.
13. Rogers WA, Buritz RS, Alpert D. Diffusion coefficient, solubility, and permeability for
helium in glass. J Appl Phys 25:868–875, 1954.
14. Short PM, Abbs ET, Rhodes CT. Effect of non-ionic surfactants on the transport of
testosterone across a cellulose acetate membrane. J Pharm Sci 59:995–998, 1970.
15. Brandt H. Determination of diffusion specific parameters by means of IR–ATR spec-
troscopy. Exp Tech Phys 33:423–431, 1985.
16. Brandt H, Hemmelmann K. On the evidence of non-linear sorption at the surface poly-
ethylene/ATR–element for the diffusion of ethylacetate in polystyrene films. Exp Tech
Phys 35:349–358, 1987.
17. Hemmelmann K, Brandt H. Investigation of diffusion and sorption properties of poly-
ethylene films for different liquids by means of IR–ATR spectroscopy. Exp Tech Phys
34:439–446, 1986.
18. Van Alsten JG, Lustig SR. Polymer mutual diffusion measurements using infrared ATR
spectroscopy. Macromolecules 25:5069–5073, 1992.
19. Potts RO, Doh L, Venkatraman, S, Farinas KC. Characterisation of solute diffusion in
a polymer using ATR–FTIR spectroscopy. Macromolecules 27:5220–5222, 1994.
20. Wurster DE, Buraphacheep V, Patel JM. The determination of diffusion coefficients in
semisolids by Fourier transform infrared (FT–IR) spectroscopy. Pharm Res 10:616–620,
1993.
21. ˜ AM, Watkinson AC, Brain KR, Hadgraft J, Armstrong NA. Use of ATR–FTIR
Tralhao
spectroscopy to study the diffusion of ethanol through glycerogelatine films. Pharm Res
12:572–575, 1995.
22. Harrison JE, Watkinson AC, Green DM, Hadgraft J, Brain KR. The relative effect of
Azone and Transcutol on permeant diffusivity and solubility in human stratum corneum.
Pharm Res 13:542–546, 1996.
23. Watkinson AC, Brain KR, Hadgraft J. The deconvolution of diffusion and partition
coefficients in permeability studies. Proc Int Symp Controlled Release Bioact Mat 21:
160–161, 1994.
24. Flynn GL, Yalkowsky SH, Roseman TJ. Mass transport phenomena and models: theo-
retical concepts. J Pharm Sci 63:479–510, 1974.
25. Flynn GL, Yalkowsky SH. Correlation and prediction of mass transport across mem-
branes. I. Influence of alkyl chain length on flux determining properties of barrier and
diffusant. J Pharm Sci 61:838–852, 1972.
26. Stehle RG, Higuchi WI. Diffusional model for transport rate studies across membranes.
J Pharm Sci 56:1367–1368, 1967.
27. Roberts MS, Anderson RA, Swarbrick J. Permeability of human epidermis to phenolic
compounds. J Pharm Pharmacol 29:677–683, 1978.
28. Zwolinski BJ, Eyring H, Reese CE. Diffusion and membrane permeability. J Phys Chem
53:1426–1453, 1949.
29. Danielli JF. The theory of penetration of a thin membrane-derivation of a fundamental
equation. In: Davson H, Danielli JF. The Permeability of Natural Membranes. 2nd ed.
London: Cambridge University Press, pp. 324–335, 1952.
30. Lakshminarayanaiah N. Transport Phenomena in Membranes. New York: Academic
Press, pp. 119–126, 1969.
31. Schaefer H, Watts F, Brod J, Illel B. Follicular penetration. In: Scott RC, Guy RH,
Hadgraft J, eds. Prediction of Percutaneous Penetration. Vol. 1. London: IBC Technical
Services, pp. 163–173, 1989.
32. Lauer AC, Lieb LM, Ramachandran C, Flynn GL, Weiner ND. Transfollicular drug
delivery. Pharm Res 12:179–186, 1995.
33. Hueber F, Wepierre J, Schaefer H. Role of transfollicular routes in percutaneous ab-
sorption of hydrocortisone and testosterone: in vivo study in the hairless rat. Int J Pharm
5:99–107, 1992.
34. Jadoul A, Hanchard C, Thysman S, Preat V. Quantification and localisation of fentanyl
and trh delivered by iontophoresis in the skin. Int J Pharm 120:221–228, 1995.
35. Craane van Hinsberg WHM, Verhoef JC, Bax LJ, Junginger HE, Bodde HE. Role of
appendages in skin resistance and iontophoretic peptide flux: human versus snake skin.
Pharm Res 12:1506–1512, 1995.
36. Scheuplein RJ, Analysis of solubility data for the case of parallel diffusion pathways.
Biophys J 6:1–17, 1966.
37. Scheuplein RJ, Blank IH, Brauner GJ, MacFarlane DJ. Percutaneous absorption of ste-
roids. J Invest Dermatol 52:63–70, 1969.
38. Barrer RM. Diffusion and permeation in heterogeneous media. In: Crank J, Park GS,
eds. Diffusion in Polymers. London: Academic Press, pp. 165–217, 1968.
39. Higuchi WI, Higuchi T. Theoretical analysis of diffusional movement through hetero-
geneous barriers. J Am Pharm Assoc Sci Ed 49:598–606, 1960.
40. Finger KF, Lemberger AP, Higuchi T, Busse LW, Wurster DE. Investigation and devel-
opment of protective ointment IV. The influence of active fillers on the permeability of
semisolids. J Am Pharm Assoc Sci Ed 49:569–573, 1960.
41. Flynn GL, Roseman TJ. Membrane diffusion II. Influence of physical adsorption on
molecular flux through heterogeneous dimethypolysiloxane barriers. J Pharm Sci 60:
1788–1796, 1970.
42. Paul DR, Kemp DR. The diffusion time lag in polymer membranes containing adsorptive
fillers. J Polymer Sci Polymer Symp 41. Transport Phenomena Through Polymer Films,
Kumins CA, ed. pp. 79–93, 1973.
43. Guy RH, Hadgraft J. A theoretical description relating skin penetration to the thickness
of the applied medicament. Int J Pharm 6:321–332, 1980.
44. Addicks WJ, Flynn GL, Weiner N, Curl R. A mathematical model to describe drug
release from thin topical applications. Int J Pharm 56:243–248, 1989.
45. Crank J, Park GS. Diffusion in Polymers. London: Academic Press, 1968.
46. Potts RO, Francoeur ML. Lipid biophysics of water loss through the skin. Proc Natl
Acad Sci USA 87:1–3, 1990.
4
Skin Transport
MICHAEL S. ROBERTS and SHEREE ELIZABETH CROSS

University of Queensland, Princess Alexandra Hospital, Brisbane,
Queensland, Australia
MARK A. PELLETT
Whitehall International, Havant, England
I. INTRODUCTION
The skin is a tissue that separates the internal living organism from the external
environment. It has a complex structure and performs many physiological functions
such as metabolism, synthesis, temperature regulation, and excretion. The outermost
layer of this organ, the stratum corneum (SC), is considered to be the main barrier
to the percutaneous absorption of exogenous materials. The skin barrier is important
in the maintenance of water within the body and in protection of the body from the
ingress of compounds, particularly important from an occupational viewpoint for
workers in the cosmetic and agrochemical industries (1).
A. Skin as a Delivery Mode

Examples of products targeted to the surface of the skin for protection are shown in
Figure 1. In contrast, percutaneous absorption of pharmaceuticals for either systemic
or local (appendageal, epidermal, and lower tissue) delivery is a desirable process
(see Fig. 1), and can be attained by the combination of appropriate solute properties
for skin transport with appropriate dosage form design (e.g., patches, gels, creams,
ointments) (Fig. 2). Compounds have been applied to the skin for many centuries
(2) and, indeed, drugs in the form of plant or animal extracts have been applied for
the relief of a variety of local disorders. In recent years, systemic delivery through
the transdermal route has led to the development and successful marketing of various
pharmaceuticals in a patch form (e.g., scopolamine, nitroglycerin, clonidine, estra-
89
90 Roberts et al.
Figure 1 Prevention and management of various conditions using the skin as a delivery
site.
diol, testosterone, timolol, fentanyl, and nicotine). The number of compounds being
screened for potential transdermal application is increasing.
B. Advantages of Transdermal Delivery

In contrast to the traditional oral route, first-pass metabolism is minimized, which
can often limit the tolerability and efficacy of many orally and parenterally delivered
drugs. Furthermore, some drugs degrade in the acidic environment of the stomach,
and other drugs, such as NSAIDs, can cause gastrointestinal bleeding or irritation.
The mixing of drugs with food in the stomach, and the pulsed, often erratic delivery
of drugs to the intestine leads to variability in the plasma concentration–time profiles
achieved for many drugs. The transdermal route provides a more-controlled, non-
invasive method of delivery, with the added advantage of being able to cease ab-
sorption in the event of an overdose or other problems. Furthermore, patient com-
pliance may be improved because of the reduced frequency of administration for
short half-life medications or avoidance of the trauma associated with parenteral
therapy.
C. Disadvantages of Transdermal Delivery

As with the other routes of drug delivery, transport across the skin is also associated
with several disadvantages, the main drawback being that not all compounds are
suitable candidates. A number of physicochemical parameters have been identified
that influence the diffusion process, and variations in permeation rates can occur
between individuals, different races, and between the old and young (3). Furthermore,
diseased skin, as well as the extent of the disease, can also affect permeation rates
Skin Transport 91
Figure 2 Examples of dosage forms used in topical delivery.

92 Roberts et al.
(3). The metabolic enzymes in the skin can also pose a problem, and some drugs
are almost completely metabolized before they reach the cutaneous vasculature (4).
Another problem that can arise, which is sometimes overlooked, is that some drugs
can be broken down before penetration through the SC by the bacteria that live on
the skin surface (4).
D. The Stratum Corneum Barrier

The barrier properties of the SC are now recognized as the major rate-limiting step
in the diffusion process of a drug permeating across skin (Table 1). However, as
pointed out by Scheuplein and Blank (5), other skin components can contribute to
the overall barrier resistance, especially for lipophilic solutes. Scott et al. (6) showed
that the permeability to water in vivo and in vitro increased after mild, superficial
epidermal alterations: suction blister top removal > adhesive tape stripping > sand-
paper abrasion > scalpel blade. After each alteration, the epidermis regenerated in a
distinct, biphasic manner. In the rapid first phase, the permeability decreased with
the development of a scab. In the second phase, there was a return to normal per-
meability, with a gradual thickening of the SC. Schaefer et al. (7) showed that in
psoriatic skin, the epidermal and dermal concentrations of radiolabeled triamcinolone
acetonide were three to ten times higher than in normal skin. A similar increase was
reported when the SC was removed by stripping before application.
The structure of the SC, as discussed in an earlier chapter, has been likened to
‘‘bricks-and-mortar’’ (Fig. 3), where the bricks are the component cells, or corneo-
cytes, and the mortar is the intercellular lipids (8). The membrane is interrupted only
by appendages such as hair follicles and sweat glands. However, it is still considered
to be a predominantly dual-compartment system composed of a matrix of corneo-
cytes tightly packed with keratin, surrounded by a complex array of lipids arranged
in bilayers (9–11).
Transport across the SC is largely a passive process, and thus the physico-
chemical properties of a permeant are an important determinant of its ability to
penetrate and diffuse across the membrane. There are generally considered to be
three routes by which compounds can diffuse across the SC: intercellular, transcel-
lular, and transappendageal (see Fig. 3). Evidence for and against these routes will
be discussed in more detail in the next section (Sec. II) of this chapter. Once it has
penetrated through the epidermis, a compound may be carried away by the dermal
blood supply or be transported to deeper tissues (see Fig. 3). Therefore, owing to
the structure of the skin, the desired physicochemical properties of a permeant are
dependent on the route taken to traverse the SC.
E. Sebum as a Barrier
The surface of the skin is the first point of contact for a topically applied formulation.
Under normal circumstances, this is covered by a 0.4- to 10-␮m irregular and dis-
continuous layer of sebum, sweat, bacteria, and dead cells (12–16). The presence of
this layer is considered to have a negligible effect on percutaneous absorption, as it
allows polar and nonpolar materials to penetrate (12,17–19). Furthermore, no cor-
relation has been found between the hydration state of the SC and the removal of
the sebum layer by swabbing with solvents, the total amount of sebhorreic lipids,
or their composition (13,14,20). Therefore, the contribution of these endogenous
Skin Transport 93
Table 1 Historical Development of the SC Barrier as a Major Rate-Limiting Step in

Drug Permeation Across Skin
Year Development Ref.
1853 Determined that the epidermis was more impermeable than the dermis 253
1877 Theorized that intact skin of humans is totally impermeable to all 254
substances
1904 Determined that the skin is more permeable to lipid soluble 255
substances than water or electrolytes
1909 Discovered that penetration through ichthyotic skin was similar to 256
that of healthy skin not less than, as was expected
1919 Determined that mustard gas penetrated into the outer layer of the 257
skin readily, but was unable to rapidly penetrate further
1924 A theorized electrical barrier between the stratum corneum and the 258
malpighian layer that reduced the permeability of the skin to ions
1930 Outlined the significance of lipid solubility in skin permeability 259
1939 The various layers of the skin can be exposed by stripping with 260
adhesive cellophane tape
1945 Suggested that the stratum lucidum was the skin’s barrier layer to the 261
penetration of both ions and uncharged molecules
1945 Suggested that the entire stratum corneum was responsible for the 262
high DC and AC resistance in the skin
1951 Separation of the epidermis from the dermis used for the first time to 33
determine the differences between their permeabilities
1951 Determined the horny layer is the barrier to diffusion of water 33
through the skin
1953 Determined that the permeability of the skin to water remained 263
unchanged until the lowest lying layers of the stratum corneum
was removed, indicating this region must contain the rate-limiting
step
1953 Conceded that the stratum corneum is uniformly impermeable to 264
water penetration regardless of distance from the surface
1954 The stratum corneum was still thought to be a porous membrane 265
through which ions and large molecules could freely permeate
1962 Determined the outer layer of the stratum corneum greatly impedes 266
penetration of substances, the concentrations decrease exponentially
with distance from the surface
1964 Techniques involving drying and staining of skin samples before 267
microscopy alter the appearance and barrier function of the skin
Source: Ref. 5.
surface materials to skin transport processes is effectively discounted and will not
be discussed further in this chapter.
As stated earlier, the skin is an important barrier to the ingress of undesirable
compounds and a potential drug delivery route for therapeutically useful compounds.
Therefore, it is important to understand how molecules traverse the skin and how
these processes can be influenced to enhance permeation. One aim of this chapter is
to define the current understanding of the processes involved in the transport of
solutes through the skin from their application site through their eventual diffusion
94 Roberts et al.
Figure 3 Schematic representation of the processes contributing to the permeability of a

solute through the skin to the bloodstream or underlying tissues.
into the systemic blood supply or into deeper tissues. Figure 4 shows the concepts
addressed in this chapter. Earlier reviews of the literature listed in Table 2 may be
used to provide a more substantial reference list of the historical developments in
percutaneous absorption from pharmaceutical preparations.
II. TRANSPORT PATHWAYS THROUGH THE STRATUM CORNEUM

As mentioned previously and shown in Figure 3, there are three pathways postulated
for the diffusion of solutes through the SC: transcellular, intercellular (paracellular),
and transappendageal. The following sections describe the nature of the transcellular
and intercellular pathways as they relate to skin transport and examine the experi-
mental and theoretical evidence for their existence. Transappendageal transport is
examined as a separate section toward the end of this chapter. The transport of solutes
through the nail plate is also considered later.
A. Transcellular Pathway
It was originally believed that transcellular diffusion mechanisms dominated over
the intercellular and transappendageal routes during the passage of solutes through
the SC (21). However, transport by the transcellular route would involve the repeated
partitioning of the molecule between lipophilic and hydrophilic compartments, in-
cluding the almost impenetrable corneocyte intracellular matrix of keratin and ker-
atohyaline. Scheuplein further suggested that polar and nonpolar solutes permeate
Skin Transport 95
Figure 4 Concepts addressed in this chapter.

96 Roberts et al.
Table 2 Early Reviews on Percutaneous Absorption
Year Author Year Author
1950s 1970s
1954 Rothman (265) 1971 Idson (128)
Katz and Poulsen (281)
Scheuplein and Blank (5)
1960s 1972 Katz and Poulsen (282)
1961 Wagner (268) 1973 Katz (283)
Poulsen (284)
1974 Anderson and Roberts (285)
1962 Barr (269) 1974 Schaefer (286)
1963 Malkinson and Rothman (270) 1975 Idson (287)
1964 Kligman (267) 1977 Dugard (288)
Malkinson (271) Higuchi (289)
Tregear (272) Malkinson and Gehlmann (290)
Welles and Lubowe (273) Webster and Maibach (291)
1965 Stoughton (274) 1978 Scheuplein (22,292)
Vinson et al. (275)
1966 Reiss (276) 1979 Flynn (293)
Tregear (12)
Vickers (277) 1980s
1969 Barrett (278) 1980 Scheuplein (294)
Blank and Scheuplein (279) 1983 Barry (126)
Winkleman (280)
the SC by different mechanisms. The polar solutes were thought to diffuse through
a high-energy pathway involving immobilized water near the outer surface of keratin
filaments. In contrast, the lipid-soluble solutes diffused through a nonpolar (intersti-
tial) lipid pathway (5,22). Our analysis of Scheuplein’s data and our own phenol
data (23,24), suggested that all solutes were transported through a lipid pathway and
ameliorated through the effects of an unstirred water (viable epidermal) layer, as
evidenced by a decrease in the energy of activation for permeation. Although the
lipid route was thought to be transcellular, evidence for its location was not defined.
Scheuplein also recognized that the dermis contributed to the resistance of the more
lipophilic solutes. Most experimental evidence now suggests that transport through
the SC is by the intercellular route.
B. Intercellular Pathway
The intercellular SC spaces were initially dismissed as a potentially significant dif-
fusion pathway because of the small volume they occupy (5). However, the physical
structure of the intercellular lipids was thought to be a significant factor in the barrier
properties of the skin (25). Tracer studies (26,27) provided evidence that the inter-
cellular lipid, and not the corneocyte protein, was the main epidermal permeability
barrier. Chandrasekaran and Shaw (28) also concluded that the lipid barrier domi=
nated. Theoretical evidence, presented by Albery and Hadgraft in 1979 (29), suggested
that the tortuous intercellular diffusional pathway around keratinocytes was the
Skin Transport 97
preferred route of penetration through the SC, rather than a drug diffusing through the
keratinized cells (transcellular route). However, it should be recognized that although
theoretical considerations favor this route, there are difficulties in designing appropriate
diffusion studies to confirm that this route is the predominant pathway (30).
Most evidence for the existence of the intercellular lipid transport pathway
comes from the microscopic organizational structure of the lipid bilayers, the ob-
served histological localization of applied substances within these bilayers following
topical application, and the effects of delipidization of the bilayers by appropriate
solvents. Histochemical studies have shown that the intracellular spaces of the SC
are devoid of lipid (31,32) and that because lipid present in other regions is highly
nonpolar, there is no structure suitable to form a lipid diffusional matrix around the
intracellular keratin filaments. The intercellular volume fraction is also much larger
than originally estimated (33), and experimental evidence using precipitation of per-
cutaneously applied compounds has led to visualization of permeation through in-
tercellular pathways (34).
In 1991, Boddé et al. (35) visualized the diffusion of mercuric chloride through
dermatomed human skin samples by using ammonium sulfide vapor to precipitate
the compound within the sample and subsequent transmission electron microscopy.
Their results indicated that the intercellular route of transport through the SC pre-
dominated; however, after longer transport times, the apical corneocytes tended to
take up the compound, leading to an apparent bimodal distribution. There was mer-
cury both inside and outside the cells in the apical region of the SC (35), whereas
in the medial and proximal region the mercury was located intercellularly. This led
to the suggestion of the presence of two types of cells: apical corneocytes that tended
to take up mercuric ions relatively easily, and medial and proximal corneocytes, that
were less capable of doing so. It has been suggested that, in the corneocytes, the
desmosomes may serve to channel material into the cell, especially in the squamous
region where the desmosomes are beginning to disintegrate. Hence, the cellular lipid
envelopes are leaky, suggesting a reservoir function for the apical zone of the SC.
Elias et al. (36) examined the penetration of [3H]water and [14C]salicylic acid
across the same tissue samples and tried to correlate diffusion with the thickness,
number of cell layers, and lipid composition of leg and abdominal skin. They found
that differences in the thickness and number of cell layers in the SC were insufficient
to account for differences observed in percutaneous transport across the leg and
abdomen, and that total lipid concentration may be the critical factor governing skin
permeability.
The lipid lamellae in the SC play a key role in the barrier function of the skin.
The major lipids are ceramides, cholesterol, and free fatty acids (37,38). Figure 5
shows a diagrammatic representation of the lamellae, dimensions of lipid arrange-
ment in an individual lamella, and the types of lipids in an intercellular lipid bilayer.
In reality, the lipid composition and arrangement is much more complex. For in-
stance, the lipids present in epidermis could be further classified as phospholipids,
monohexosylceramides, ceramides, cholesterol, cholesterol esters, cholesterol sulfate,
triglycerides, and fatty acids (39). Furthermore, at least six subclasses of ceramides
have been described in pig SC (37,40).
Bouwstra et al. (41), in a summary of the X-ray analysis of the SC and its
lipids, concluded that the lamella is the fundamental structure in intercellular do-
mains. An individual lamella is about 13 nm (130 Å) in width and comprises two
98 Roberts et al.
Figure 5 Diagrammatic representation of the structure of the stratum corneum showing

(A) the bricks and mortar model of its gross structure; (B) the intercellular bilayers; (C) the
spatial organization of lipids within the bilayers; (D) the location of polar and lipid domains;
(E) the presence of proteic and desmosomal structures within the lipid bilayers.
Skin Transport 99
or three lipid bilayers (see Fig. 5). Although 13.4-nm (134-Å) lamella is dominant
in human SC, an occasional 6.4-nm (64-Å) lamella was observed. In mouse SC, the
dominant lamella is 13.1 nm (131 Å), with an occasional 6.0-nm (60-Å) one. The
actual organization of (pig) SC lipids is as two lamellar phases with periodicities of
approximately 6 and 13 nm, respectively (40). At the molecular level, Bouwstra et
al. (40) suggest that the short periodicity phase is composed of only one bilayer,
akin to phospholipid membranes. In contrast, the long periodicity phase consists of
two broad and one narrow low–electron-dense regions. It is suggested that the two
broad regions are formed by partly interdigitating ceramides, with long-chain fatty
acids of approximately 24–26 carbon atoms, whereas the narrow low–electron-den-
sity region is formed by fully interdigitating ceramides, with a short free fatty acid
chain of approximately 16–18 carbon atoms.
C. Lipid and Polar Pathways Through the Intercellular Lipids

Both diffusional and morphometric data have been presented to support lipid and
polar pathways through SC lipids. Southwell and Barry (42) used penetration-en-
hancing solvents to modify the different diffusional routes through the SC and the
partitioning of drugs into these pathways. Steady-state fluxes were measured in vitro
for polar methanol, nonpolar octanol, and an intermediate compound, caffeine, se-
lected as model penetrants through human SC conditioned on both sides with water
or the two accelerants. They concluded that 2-pyrrolidone enhances permeation
through the polar route of the skin by increasing the diffusivity, but reduced nonpolar
route transport. Whereas dimethylformamide (DMF) promotes polar route absorption
by raising diffusivity and partitioning, but reduces nonpolar absorption by decreasing
both parameters. Blank and McAuliffe (43) also suggested the presence of polar and
nonpolar pathways in the SC through different routes, on the basis of selective sol-
vent effects on the permeability constants for tritiated water (a polar molecule) and
for benzene (a relatively nonpolar molecule). Several investigators recognized the
presence of a polar pathway and, through modeling, showed that for lipophilic sol-
utes, such as steroids (44) and ␤-blockers (45), the contribution of the polar pathway
was negligible.
Kim et al. (46) elaborated on an in vitro model for skin permeation in which
penetration could occur across the main barrier, the SC by one of two parallel path-
ways: the lipoidal pathway and the pore pathway, with this barrier existing in series
with an epidermal–dermal porous barrier. According to this model very lipophilic
molecules are rate-limited by the epidermal–dermal barrier, as described previously
by Scheuplein and Blank (5). Extremely polar permeants are rate-limited by the pore
pathway of the SC with its limiting permeability coefficient, whereas permeants with
intermediate polarity are transported by the lipoidal pathway and exhibit a lipophil-
icity–dependent permeability coefficient. Such a model has also been proposed by
Cooper and Kasting (47) and is discussed by Roberts and Walters (3).
Matsuzaki et al. (48) found that the permeability of very polar solutes through
model SC membrane systems was almost constant and similar to that of potassium
ions, whereas, for the more lipophilic solutes, permeability increased with solute
lipophilicity. This data suggests, therefore, that solutes may be transported through
both a polar and nonpolar pathway through the intercellular region. Peck et al.
(49,50) examined the in vitro passive transport of urea, mannitol, sucrose, and raf-
100 Roberts et al.
finose across intact and ethanol-treated human epidermal membranes. From the rel-
ative permeabilities of these four solutes and hindered diffusion theory, effective
pore radii estimates for intact and ethanol-treated human epidermal membrane were
between 1.5 and 2.5 nm (15 and 25 Å) and 1.5 and 2.0 nm (15 and 20 Å), respec-
tively. Further studies on the temperature dependence of human epidermal membrane
permeability with urea, mannitol, tetraethylammonium ion, and corticosterone
strongly support the existence of a porous permeation pathway. Interestingly, the
radii estimated is similar to that determined by iontophoretic studies (51,52) for the
pore size range attributed for small solutes to transport through the polar intercellular
lipid pathway. Sznitowska et al. (53) examined the percutaneous penetration of
baclofen, a model zwitterion, in vitro using human cadaver skin with various solvent
pretreatments. They concluded that the polar pathway might be intercellular and
comprises the aqueous regions surrounded by polar lipids. Finally, Menon and Elias
(54) applied hydrophilic and hydrophobic tracers to murine skin in vivo under basal
conditions or after permeation enhancement with occlusion, vehicle enhancers, a lipid
synthesis inhibitor, sonophoresis, and iontophoresis. Using ruthenium and microwave
postfixation methods, tracers were found localized in discrete lacuna domains in the
extracellular lamellar membrane system, regardless of their polarity or the enhance-
ment method. Although extracellular lacunar domains were interpreted as being a
potential pore pathway for penetration of polar and nonpolar molecules across the
SC, the continuity of such a pathway is unclear. Figure 6 is our interpretation of the
possible polar and lipid pathways for intercellular transport.
III. PRINCIPLES OF SKIN TRANSPORT

The process of percutaneous absorption involves several individual transport pro-
cesses, some of which occur in series and others in parallel (see Fig. 3). The two
key determinants for a solute crossing a membrane are solubility and diffusivity. The
relative solubility of a solute in two phases determines its partition coefficient and,
therefore, the likelihood of the solute being taken up into the SC from a vehicle.
Also, solubility will determine whether a solute is likely to be desorbed from the SC
into deeper layers. The diffusivity is a measure of the speed at which a solute crosses
a given barrier and is affected by binding, viscosity of the environment, and the
tortuosity of the path.
In the first step of the transport process, molecules must be in solution in the
vehicle to partition from the vehicle into the lipids in the outermost part of the SC;
they must then diffuse through it; partition back out of the SC and into the viable
epidermis. Next, molecules diffuse through the viable epidermis and papillary dermis.
At the capillary plexus a high percentage of molecules are transferred into the cir-
culating blood and a lower percentage diffuses into deeper tissues (see Fig. 3). To
predict the penetration of a given solute it is necessary (a) to define the skin barrier
in terms of a mechanistic model, and (b) to relate transport to a physical property
of the solute, such as its organic solvent–water partition coefficient. Scheuplein and
Blank (5) suggested that an appropriate skin model is a multilayer barrier consisting
of the SC (10 ␮m) (S1), the viable epidermis (100 ␮m) (S2), and the upper papillary
layer of the dermis (100–200 ␮m) (Fig. 7). We now develop a steady-state model
for skin transport consistent with this model and based on the theoretical consider-
ations presented in Chapter 3.
Skin Transport 101
Figure 6 Partitioning and diffusion processes involved in solute penetration through the
stratum corneum. (Adapted from Refs. 95 and 123.)
A. Stratum Corneum–Vehicle Partition Coefficients

We first consider the partitioning between the SC and vehicle. The previous chapter
recognized that the chemical potential gradient across a membrane is a major deter-
minant of flux (J), the amount of solute passing through a unit area of membrane in
unit time. The chemical potential of a solute in a phase is also a major determinant
for its partitioning into another adjacent phase. In an ideal solution, the chemical
potential of a solute ␮i is defined by the standard chemical potential state ␮ 0i for that
solute, and its activity ai (defined as the product of its activity coefficient ␥i and
concentration Ci , expressed as a mole fraction; i.e., ai = ␥i Ci), the gas constant R,
and absolute temperature T:
␮i = ␮ 0i ⫹ RT ln ai = ␮ 0i ⫹ RT ln ␥i Ci (1)
The partitioning of a solute between the SC and vehicle is defined by the chemical
potential difference between the solute in the SC ␮sc and that in the vehicle ␮v for
which chemical potential is defined by Eq. (1) for each phase. At equilibrium, the
chemical potential of the solute in the two phases is equal (i.e., ␮sc = ␮v) such
that (55)
102 Roberts et al.
Figure 7 Diagrammatic representation of steady-state concentrations of a solute in the skin

after topical application. (From Ref. 5.)
␮ 0sc ⫹ RT ln ␥scCsc = ␮ 0v ⫹ RT ln ␥vCv (2)
Rearranging and defining the ratios of ␥iCi as a SC–vehicle partition coefficient

K asc⫺v , based on activities, yields
a
K sc⫺v =
asc ␥scCsc
av
=
␥vCv
= exp 冉 RT 冊
␮ 0sc ⫺ ␮ v0
(3)
However, we could also define an SC–vehicle partition coefficient Ksc⫺v based on

concentrations.
Skin Transport 103
Ksc⫺v =
Csc ␥v asc ␥v
= =
Cv ␥sc av ␥sc
exp 冉
␮ sc
0
RT
冊
⫺ ␮ v0
(4)
Ksc⫺v could also be defined in terms of the solubility of the solute in the SC (Ssc)
and vehicle (Sv):
Ssc
Ksc⫺v = (5)
Sv
Solubility, in this chapter, unless otherwise specified, is defined by a phase
being saturated with solute such that the chemical potential of the dissolved solute
␮i,saturated = ␮pure. As a number of studies have applied various approaches used to
determine solubility in the prediction of percutaneous penetration flux, we now con-
sider the prediction of solubility using these approaches in further detail.
1. Conventions Used for Prediction of Solubility, Partition Coefficient,
and Flux
It should be emphasized that the prediction method is dependent on the definition
of the standard state ␮ 0i . Two conventions are widely used in defining the standard
state of a solute (56). In our experience it is often convenient to use convention 1
and established regular solution theory, to derive solubilities of solutes in different
vehicles. This recognizes that, as an approximation, solutes with the same fractional
solubility for different solvents have the same activity. Convention 2 is easiest to
apply over a range of concentrations.
In the first, or Raoult’s law, convention 1, the standard state is the pure sub-
stance as a liquid ␮ 0pure. Hence, ␮ 0sc = ␮ 0v = ␮ 0pure and expressing Ci in mole fractions,
␥sc → 1, ␥v → 1 as Xi → 1. Further, applying Eq. (1) and noting that ␮i,saturated =
␮pure in saturated systems, it is evident that the (thermodynamic) activity of the solute
in the SC (asc) is equal to that in the vehicle (av) and that of a pure liquid solute
(apure); that is, asc = av = apure. Accordingly, noting from Eq. (1) that the mole fraction
solubility (Xi) is related to activity (ai) and the activity coefficient (␥i) by:
ai
Xi = (6)
␥i
then
Xsc ␥v
Ksc⫺v = = (7)
Xv ␥sc
By definition, ␥i = 1 for an ideal binary liquid mixture. An ideal mixture re-
quires that (a) both phases are mutually soluble in all proportions, and (b) the partial
vapor pressure of a given component (pi) is directly related to its vapor pressure as
a pure liquid (p0i ) by its mole fraction in the mixture (Xi), as defined by Raoult’s
law. The extent which ␥i deviates from unity can be considered as a measure of
deviation from Raoult’s law, which is defined in terms of the partial vapor pressure
of a solute in solution ( pi), that of the pure component (p0i ) and Xi (i.e., p =
p0i ␥i Xi). Barry et al. (57) shows an example of such a deviation for the vapor pressure
of benzyl alcohol plotted against its mole fraction.
The second convention is to use the infinitely dilute solution of the solute in
a given phase as the standard state where ␥ * sc → 1 and ␥ v* → 1 as Xi and Ci → 0.
104 Roberts et al.
Note that, following the convention of Davis et al. (58), the activity coefficient from
convention 2 is given an asterisk to distinguish it from the convention 1 or Raoult’s
law activity coefficient. The activity coefficient from the second convention ␥ * i has
also been referred to as Henry’s law activity coefficient, on the basis that this law
defines a constant activity coefficient of 1 with varying low concentrations and a
␥ i* of less than 1 is, therefore, a measure of deviation from this law. Hence, under
ideal conditions in convention 2, ␥ sc* → 1 and ␥ * v → 1, and K sc⫺v = Ksc⫺v. It is to
a
be emphasized that, in contrast to convention 1, the activities of solute based on this

convention are not equal in saturated solutions (i.e., asc ≠ av ≠ apure) but are related
to the standard chemical potentials, as shown in Eq. (3) in which the standard state
is as ␥ i* → 1, Ci → 0. The activity is also asterisked to distinguish it from activity
as defined by convention 1.
2. Prediction of Solubility and Partition Coefficients with the First
(Raoult’s) Convention
We now examine the estimation of solubility of a solute in a phase based on the
first (Raoult’s) convention. Solubility and partitioning can be described in terms of
the energy required to convert from the solid solute to a molecular form, the energy
of dissolution in a vehicle, and the energy of dissolution in the SC (Fig. 8A). As
many solutes used in topical delivery are solids, it is necessary to express solubilities
and partition coefficients in terms of the activity of the pure solid (ssolid), also referred
to as its ideal solubility (X 0i ). This ideal solubility varies with the nature of the solute
crystal and is related to the energy associated with the formation of the pure liquid
form by melting of the crystals at a melting point (Tm) (see Fig. 8A). X 0i is a function
of the molar heat of fusion (⌬Hf ), melting point (Tm), gas constant (R), room tem-
perature (T ), and (⌬Cp), the difference in heat capacity of the crystalline and molten
states (59):
ln X 0i =
⫺⌬Hf
RT
冋
1
T
⫺
1
Tm
册 ⫹
⌬Cp
R
冋
Tm ⫺ T
T
⫺ ln
Tm
T
册 (8)
Yalkowsky and Valvani (60) have pointed out that the last term can usually be
ignored without any significant loss of accuracy so that
ln X 0i ⬵
⫺⌬Hf
RT 冋 1
T
⫺
1
册
Tm
(9)
They further note that, as the free energy of fusion is 0 at the melting point, Eq. (9)
could be expressed in terms of the entropy of fusion (⌬Sf = ⌬Hf /Tm). Given entropies
of fusion are relatively constant across solutes, Grant and Higuchi (61) have sug-
gested Eq. (8) be written in terms of the conventional centigrade temperature (T)
and melting point (MP) as follows:
log X 0i ⬵ ⫺0.0099[MP ⫺ 25] (10)
In a nonideal solution, ␥i < 1 as a result of solute–solvent interactions. The
dissolution of a solute in a solvent is characterized by the following processes: (a)
dispersion forces associated the transfer of solute molecules from its solution, the
formation of a cavity in the solvent to accommodate the solute molecules, and the
reorientation of solvent molecules around the solute molecules in the cavity; (b)
Skin Transport 105
Figure 8A Processes involved in the dissolution of crystalline solute and the related
energies.
dipole–dipole and induced dipole–dipole interactions between solute and solvent

molecules; and (c) H-bonds between solute and solvent (see Fig. 8A).
In the commonly used Hildebrand solubility parameter approach, it is assumed
that nonideality is solely due to dispersion forces. Defining the solubility parameter
(␦i) as the energy required to move a molecule from its solution, ␦i = (⌬Hv ⫺ RT/
Vi); where ⌬Hv is the heat of vaporization and Vi is the molar volume of the solute.
106 Roberts et al.
Figure 8B Solute–vehicle and stratum corneum interactions promoting partitioning into

the stratum corneum.
Summing the energies required to remove a molecule of solute, to make a cavity in

a solvent and reorientate solvent molecules around solute yields the convention 1 or
Raoult’s law activity coefficient (␥i) for a solute in the solvent.
Vi␾ I2
ln ␥i = [␦i ⫺ ␦I ]2 (11)
RT
where Vi is the molar volume of the solute, ␦i and ␦I are the solubility parameters
for the solute and solvent, ␾I is the volume fraction of the solvent (␾I → 1 for dilute
solutions). In a mixed solvent, ␦I = ␾A␦A ⫹ ␾B␦B ⫹ . . . . Hence, combining Eqs. (6),
(9), and (11), the mole fraction solubility of a solute Xi is given by
ln Xi = ln X i0 ⫺ ln ␥i = ⫺
⌬Hf
R
冋
1
T
⫺
1
册
Tm
⫺ [␦i ⫺ ␦I ]2
Vi␾ I2
RT
(12)
Hence, the solute solubility Xi is enhanced either by the solute’s lower-melting point
[first term in right hand side of Eq. (11) becomes zero when the solute is a liquid;
Skin Transport 107
i.e., Tm ⱕ T ] or by choosing a solvent with a solubility parameter close to that of

the solute so that (␦i ⫺ ␦I )2 is minimized.
We now use the foregoing concepts to derive the SC–vehicle partition coeffi-
cient (Ksc⫺v). Substituting Eq. (11) into Eq. (7) yields for the SC and for a vehicle
(62):
Vi␾ 2v Vi␾ 2sc
ln Ksc⫺v = ln ␥v ⫺ ln ␥sc = [␦i ⫺ ␦v ]2 ⫺ [␦i ⫺ ␦sc]2 (13)
RT RT
If low concentrations of solute are present in both the vehicle and the SC, ␾v → 1
and ␾sc → 1 so that Eq. (12) reduces to
Vi
ln Ksc⫺v = ([␦i ⫺ ␦v]2 ⫺ [␦i ⫺ ␦sc]2) (14)
RT
Table 3 shows calculated SC–vehicle partition coefficients derived for theophylline
by Sloan et al. (62) based on a SC solubility parameter of 10 cal1/2 cm3.2. Hence,
isopropyl myristate would appear to be the best vehicle for optimizing the partition-
ing of theophylline into the SC. Sloan et al. (62), have also pointed out that flux of
solutes through the membrane is enhanced when the solubility parameter of the
vehicle or cosolvent mixture is close to that of the SC.
Kadir et al. (63–65) studied the human skin permeability coefficients of theo-
phylline and adenosine from alkanoic acid solutions and found certain acids pro-
moted penetration through vehicle effects (‘‘push’’ effect) whereas propionic acid
enhances the penetration of theophylline and adenosine by promoting their solubility
in the skin–propionic acid medium (‘‘pull’’ effect). The push effect could be esti-
mated by the solubility parameter approach. A push effect is equivalent to an increase
in Ksc⫺v by an increase in the (␦i ⫺ ␦v)2 term in Eq. (14), whereas a pull effect is an
increase in Ksc owing to a decrease in the (␦i ⫺ ␦sc)2 term in Eq. (14) (see Fig. 8B).
Attempts have been made to modify the Hildebrand solubility parameter by
including, in addition to the dispersion component (d), the polar component (p) and
hydrogen-bonding interactions, to give the so-called three-dimensional solubility pa-
rameter ␦ 3D
i = (␦ di )2 ⫹ (␦ pi )2 ⫹ (␦ Hi )2 (66,67). Groning and Braun (68) used this con-
Table 3 Vehicle Solubility Parameters, Molar Volumes, and Estimates of Theophylline

Stratum Corneum–Vehicle Partition Coefficientsa
Molar Solubility Experimental

volume parameter mole fraction
Vehicle (cm3/mol) (cal/cm3)1/2 log Ksc⫺v solubility ⫻ 103
Isopropyl myristate 317.0 8.5 1.12 0.109

Octanol 157.5 10.3 ⫺0.18 1.49
Dimethylformamide 77.0 12.1 ⫺0.98 14.8
Propylene glycol 73.5 14.8 ⫺1.22 3.30
Ethylene glycol 56.0 16.1 ⫺0.92 2.30
Formamide 39.7 17.9 ⫺0.06 0.346
a
Based on it having a solubility parameter of 14.0 (cal/cm3)1/2 and a molar volume of 110 cm3/mol, the
stratum corneum having a solubility parameter of 10 (cal/cm3)1/2
Source: Ref. 62.
108 Roberts et al.
cept to describe the permeation of solutes through the skin. Ruelle et al. (69) have
suggested that both the usual solubility parameter and the three-dimensional (hydro-
gen bond) solubility parameter are often inappropriate to account for hydrogen bond-
ing, as there is an exothermic reaction with the formation of a solute–solvent hy-
drogen bond. They suggest that stability constants defining interactions between
solute and solvent and between themselves may be more appropriate. Ando et al.
(70) assumed that dispersion forces applied to a nonpolar lipid pathway, whereas
ion–dipole interactions with keratin applied to a polar pathway. The solvatochromic
approach introduced to percutaneous absorption by Roberts et al. (71) and Abraham
et al. (72) allows dispersion, dipolar, and hydrogen bonding to be included as separate
terms. The solubility of a liquid solute can be expressed in terms of the dispersion,
dipole, and hydrogen interactions, using a linear free energy approach:
Vi
⫺RT ln Xi ⫹ const = ⌬Gi = ⌬G di ⫹ ⌬G ip ⫹ ⌬G Hi = A␦ 2i
100
⫹ B␲ i*␲ *
I ⫹ C␣I ␤i ⫹ D␣i ␤I (15)
where constants A, B, C, and D are determined by regression; Vi is the molar volume
of the solute; ␦ 2i Vi /100 is the energy associated with creating a cavity for a solute
molecule of molar volume V * i , ␲ I*, and ␲ i* are the dipole solvatochromic parameters
for the solvent and solute, respectively; and ␣i, ␣i, ␤I, and ␤i are the solvatochromic
parameters for hydrogen-bonding–donating ability of solvent and solute and hydro-
gen-bonding–accepting ability of the solvent and solute, respectively. Hence, for a
solid solute, applying Eqs. (5), (10), and (15):
V2
ln Xi = const ⫹ A␦ 2i ⫹ B␲ *
i ␲*
I ⫹ C␤2 ⫹ D␣2 ⫹ 0.0099(mp ⫺ 25) (16)
100
Yalkowsky et al. (73) applied Eq. (16) in the estimation of the solubility of 185
solutes in water and obtained (r 2 = 0.977):
ln Xw = 0.86 ⫺ 0.062Vi ⫹ 4.9␤i ⫺ 0.0099(mp ⫺ 25) (17)
According to Eq. (5), ln Ksc⫺v is simply defined by ln Ssc ⫺ ln Sv and, hence, is of
the same form as given in Eq. (16). Abraham et al. (74) showed that for 613 solutes
the solvatochromic regression (r = 0.9974) for log Koct was
log Koct = 0.088 ⫹ 0.562R ⫺ 11.054␲ H ⫹ 0.034␣ ⫺ 3.460␤ ⫹ 3.814Vx (18)
where R is an excess molar refraction, ␲ the solute dipolarity/polarizability, ␣ and
H
␤ the effective solute hydrogen bond acidity and basicity, and VX the characteristic
volume of McGowan. It is apparent that log Koct is dominated by solute hydrogen
basicity favoring distribution into water, and solute size favoring distribution into
octanol. Yalkowsky et al. (73) argue that the octanol–water partition coefficient
method (discussed in the next section) is superior for the estimation of water solu-
bility, as it is two orders of magnitude larger and achieves the same fit with fewer
variables.
3. Prediction from Octanol–Water Partition Coefficients
The use of the logarithm of octanol–water partition coefficients (log Koct) for the
prediction of biological activity through structure–activity relations originates from
Skin Transport 109
the work of Hansch (75). Log Koct has also been used in the evaluation of SC–water
partition coefficients (76). Roberts et al. (77) reported the following regression for
45 solutes (r 2 = 0.839):
log Ksc⫺v = 0.57 log Koct ⫺ 0.1 (19)
In theory, an estimate of the solubility of solutes in SC Ssc should be possible by
using Eq. (19) in an extended and rearranged form of Eq. (7), providing the solubility
of solutes in water Xw can be defined:
log Xsc = log Ksc⫺w ⫹ log Xw = 0.57 log Koct ⫺ 0.1 ⫹ log Xw (20)
Yalkowsky and Valvani (60) have suggested that the solubility in water Xw can be
estimated using Eqs. (6) and (7) when the activity coefficient for electrolytes in water
␥w can be estimated:
ln Xw = ln X 0i ⫺ ln ␥w = ⫺
⌬Hf
R 冋
1
T
⫺
1
Tm 册 ⫺ ln ␥w (21)
Rearranging Eq. (7):

log ␥w = log Koct ⫺ log ␥oct (22)
Yalkowsky and Valvani (60) followed an approach similar to the treatment of
regular solutions by Hildebrand and Scott (59) to deduce that, as the adhesive inter-
actions between octanol and a solute approximately equals the sum of the cohesive
interactions in octanol and in the solute, ␥oct ⬃ 1 for most solutes. Accordingly,
log ␥w ⬇ log Koct (23)
On substitution into Eq. (21)
log Xw = log X i0 ⫺ log ␥w = ⫺

⌬Hf
2.303R 冋
1
T
⫺
1
Tm 册 ⫺ log Koct (24)
Further, by combining Eq. (10) and (24) and assuming rigid solutes, Yalkowsky and
Valvani (60) obtained the following regression for 155 solutes (r 2 = 0.979):
log Sw = ⫺1.05 log Koct ⫺ 0.0012MP ⫹ 0.87 (25)
Hence, an expression for the solubility of solutes in SC (Xsc ) is
log Xsc = log Ksc⫺w ⫹ log Xw = ⫺0.48 log Koct ⫺ 0.0012MP ⫹ 0.77 (26)
Roberts et al. (24) also observed a decrease in the estimated maximum flux of phe-
nols with octanol water partition coefficient and suggested that it may reflect that
the more polar phenols had lower molar volumes. Importantly, Eq. (26) suggests that
highest solubilities in the SC will be seen for the lowest-melting point solutes.
4. Use of Group Contributions to Estimate SC–Water Partition
Coefficients and Make Deductions about the Naute of the Barrier
Group contributions are now widely used in the estimation of octanol–water partition
coefficients, the approach attributed to Hansch and co-workers (75). Recognizing
that the free energy for transfer of a solute from water to SC (⌬Gsc⫺w) is related to
its SC–water partition coefficient (Ksc⫺w), Scheuplein and Blank (5) assumed that
110 Roberts et al.
(⌬Gsc⫺w) was expressed additively by the individual groups in a solute and applied
it to their series of alcohols. Written in terms of transfer from water to SC (5):
⫺CH2
⌬Gsc⫺w = ⌬G sc⫺w
OH
⫹ n⌬G sc⫺w = ⫺RT ln Ksc⫺w (27)
⫺1
They reported a ⌬G ⫺CHsc⫺w of ⫺460 cal mol
2
at 25⬚C, and noted that it was, in absolute
terms, lower than for the transfer of — CH2 into olive oil (⫺740 cal mol⫺1). Roberts
et al. (78) reported group contributions for water to SC and water to octanol as
follows (in cal mol⫺1 at 25⬚C) — OH (670, 710), — CH2 (⫺410, ⫺710), — Cl (810,
1210), — Br (960, 1550), — NO2 (510, 740), and — COOCH3 (240, 680). Although
the hydroxyl group is similar for the transfer from water to octanol, the nonpolar
contributions are almost half. Roberts et al. (78), in further recognizing the essential
temperature independence of Ksc (temperature range 12.6–34.5⬚C), suggested that
the partition was entropy-driven. In further analyses, they compared the relative
enthalpies and entropies of solute transfer from water into the SC and into various
crystalline states of lecithin. It was concluded that the hydrophobic phase of the SC,
with which the solutes are most associated, might be considered to be in a liquid
crystalline state or more polar. Anderson’s group (79) reported that, at 37⬚C, the
partition coefficients into untreated and delipidized SC were similar. They reported
a similar free-energy group contribution for the transfer of the — OH and — COOCH3
group from water to the protein domain of the SC (580, 160 cal mol⫺1 at 25⬚C).
New free-energy group transfers were (in cal mol⫺1 at 25⬚C): — CONH2 (660);
— CON(CH3)2 (160); and — COOH (30). They also reported free-energy group trans-
fers for — CH2 and — COOCH3 from water into octanol similar to Roberts et al.
(78), but a somewhat higher value for the — OH group (2350 cal mol⫺1). Davis et
al. (58) report a range of literature values for — OH substituents on an aromatic ring
that are all less than 870 cal mol⫺1 (mean 610 calmol⫺1) for transfer from water into
octanol and are comparable with those reported for transfer from water to SC. The
SC–water partition coefficients are equilibrium values, which will also occur
throughout the SC during the permeation process. The differences in the group con-
tributions for a SC protein domain and transport barrier is consistent with the par-
titioning involving binding to protein sites, the exact site of which remains ill-defined
and may include keratin within the cells and desmosomes in the intercellular space.
Interestingly, the solubility parameter for a keratin fragment [11.8 (cal/cm3)0.5] (68)
is similar to that for butanol [11.18 (cal/cm3)0.5], advocated by Roberts et al. (24,78)
as being of a polarity similar to the apparent partition coefficient domain of the SC.
It should be emphasized that group contributions for partition coefficients differ from
those for permeation, which are considered later.
A model of partitioning into the protein domains in the intercellular region and
into keratinocytes during the permeation process is illustrated in Figure 9. It is ap-
parent that the major effect of partitioning is to act as a buffer in the transport process
and, as we will see later, this buffering will decrease the diffusivity and increase the
lag time of solutes traversing the SC. As a consequence, the apparent permeability
coefficient for a solute through the SC can be shown to depend on both the affinity
of solutes for lipids in the diffusion pathway and on binding to other sites [see Eqs.
(36) and (37)]. In summary, SC–water partition coefficients are useful in helping
define the fraction of unbound solute and its ability to diffuse down the SC inter-
cellular lipid pathway [see Fig. 9; discussed later in Eq. (32)].
Skin Transport 111
Figure 9 Partition and diffusive processes in solute transport through the stratum corneum,
assessing an intercellular lipid pathway.
B. Steady-State Flux
As derived in Chapter 3 and presented previously (3), the steady-state solute flux
through the SC per unit area of application J sssc is determined by the concentrations
of solute immediately below the outside Csc(o) and inside Csc(i) the SC:
Dsc
ss
J sc = (Csc(o) ⫺ Csc(i)) (28)
hsc
where Dsc is the effective diffusion coefficient in the SC, with a diffusion path length
hsc. It should be noted that Dsc is defined both by diffusion of free solute and the
instantaneous partitioning into immobile sites in the diffusion path. Chandrasekaran
et al. (80,81) proposed a dual sorption model using this concept to describe the
uptake of drugs by skin. A simple form of their expression can be derived ignoring
the differences in the apparent volumes for partitioning and diffusion as well as the
rates of distribution [Later work will show that these assumptions are questionable
with significance for certain situations (82)]. Under linear binding and ‘‘instantane-
ous’’ equilibrium conditions, the unbound concentration of solute in SC lipids
(Csc,u) is related to that partitioned and bound to other SC components (Csc,b) by a
partition coefficient (K sc sc
b/u); that is, K b/u = Csc,b /Csc,u. If only unbound solute diffuses,
the change in unbound concentration in the SC ⭸Csc,u /⭸t is given by:
112 Roberts et al.
⭸Csc,u ⭸2Csc,u ⭸Csc,b ⭸2Csc,u ⭸Csc,u

= Dsc,u ⫺ = Dsc,u ⫺ K sc (29)
⭸t ⭸x ⭸t ⭸x 2 ⭸t
2 b/u
Rearranging
⭸Csc,total ⭸Csc,u ⭸2Csc,u
= (1 ⫹ K sc
b/u) = Dsc,u (30)
⭸t ⭸t ⭸x 2
since Csc,total = Csc,u ⫹ Csc,b.
Further recognizing that total concentration Csc in the SC is related to fu,sc the
fraction unbound in the SC by Cu,sc = fu,scCsc and that fu,sc = 1/(1 ⫹ K sc
b/u), Eq. (30)
can be expressed as
⭸Csc,u ⭸ 2Csc,u ⭸ 2Csc,u
= fu,sc Dsc,u = D (31)
⭸t ⭸x 2 ⭸x 2
sc
since Dsc = fu,sc Dsc,u. Hence, the measured Dsc will reflect a reduction in the free-
diffusion coefficient as a result of binding to immobile components in the diffusion
path. If we assume that the unbound concentration is in the internal environment of
the intercellular lipid bilayer, then
Cu,sc Cu,sc Cw Klipid⫺w
fu,sc = = = = Klipid⫺sc (32)
Csc Cw Csc Ksc⫺w
We now include the SC lipid pathway–vehicle partition coefficient K sc
lipid⫺v , following
the earlier assumption that the solute in this bilayer is unbound and that the lipid
bilayer is the diffusion path:
sc C sc
lipid
K lipid⫺v = (33)
Cv
Substituting Eq. (33) into Eq. (28), recognizing that the transport pathway is the lipid
bilayer, therefore, accounting for partitioning differences, yields
ss
J sc =
K sc
lipid⫺v Dsc
hsc 冉Cv ⫺
sc
K lipid⫺ve
sc
K lipid⫺v 冊冉
C ssve = k psc Cv ⫺
K sc
lipid⫺ve
sc
K lipid⫺v 冊
ss
C ve
= k sc
p 冉 Cv ⫺
1
Kve⫺v
ss
C ve 冊 (34)
where C ssve is the concentration of the solute in the viable epidermis, and K lipid⫺ve
sc
is
the SC lipid–viable epidermis partition coefficient, and Kve⫺v is the viable epidermis–
vehicle partition coefficient. Similar expressions could be derived for flux through
the viable epidermis with the dermis as an adjacent phase as shown in Figure 7. An
equation similar in form to Eq. (34) has also been used to describe the flux through
SC into the receptor phase of an in vitro penetration study (83). As stated earlier,
transport through the SC may occur through various pathways. In earlier work (3,83),
we have recognized that k sc p is a composite parameter and may be more properly
expressed as, for instance,
p = k p,lipid ⫹ k p,polar ⫹ k p,appendages
k sc sc sc sc
(35)
sc sc sc
where k ,k
p,lipid , and k
p,polar are the component SC permeability coefficients
p,appendageal
for lipid, polar, and appendageal pathways. Kasting et al. (84) have discussed the
Skin Transport 113
relative magnitudes of k sc sc sc
p,lipid and k p,polar. They suggest that, whereas k p,lipid may vary
from 0.3 ⫻ 10 to 13,000 ⫻ 10 cm/h, k p,polar varies from 0.1 ⫻ 10 to 1 ⫻ 10⫺5
⫺5 ⫺5 sc ⫺5
cm/h. Thus, k sc
p,polar becomes important only for very low permeability solutes. The
appendageal component is added to Eq. (35) for completeness, given the later dis-
cussion on potential transport through this pathway. For most solutes, penetration
appears to occur through the lipid pathway, with a permeability coefficient k sc p,lipid. If
we assume that diffusion occurs through the lipid pathway and, as stated earlier, that
the distribution volumes are the same, then the partition coefficient of relevance will
appear to be that from the vehicle into this pathway (Klipid⫺v) and not Ksc⫺v [Eq. (36)]
sc
sc Dsc K lipid⫺v fu,sc Dsc,u K sc sc
lipid⫺v K lipid⫺w Dsc,u
k p,lipid = K sc
lipid⫺v = = (36)
hsc hsc Ksc⫺w hsc
so that when the vehicle is water, then Dsc is an apparent value when the partitioning
is assumed to be determined by Ksc⫺v; that is,
冉冊
2
sc K sc
lipid⫺w Dsc,u D app
sc
k = Ksc⫺w
p = Ksc⫺w (37)
Ksc⫺w hsc hsc
where the apparent diffusion coefficient (D asc) is given by
冉冊
2
app K sc
lipid⫺w
D sc = Dsc,u (38)
Ksc⫺w
It should be reemphasized that this analysis is based on the binding sites being
present in the diffusion pathway. In reality, the distribution volumes for lipids and
other binding sites differ, and more complex expressions are appropriate (82). Nev-
ertheless, these expressions show that both the SC–vehicle partitioning and diffusion
down an exclusively lipid pathway affect the observed permeability coefficients. The
diffusion coefficient Dsc,u = kB T/6␲␩r where kB is the Boltzmann’s constant, T is the
temperature, ␩ is the viscosity of the pathway, and r is the radius of the diffusing
solute.
Several other heterogeneous skin permeability models have been described.
Albery and Hadgraft (29) assumed that impermeable corneocytes were embedded in
a permeable, homogeneous lipid phase. In Tojo’s model (85), both the lipid and
corneocytes phases are permeable, but with a partition coefficient between them.
Heisig et al. (86) have suggested that the heterogeneity of the SC precludes an
analytical solution. He used a ‘‘brick-and-mortar’’ model of the SC (ten layers with
corneocytes 30-␮m wide and 1-␮m thick and a lipid channel of 0.1 ␮m, as shown
in Fig. 5) and concluded that the long lag times and very small human SC perme-
abilities can be predicted only for a highly staggered corneocyte geometry and only
when the corneocytes are 1000 times less permeable than the lipid phase. Plewig
and Marples (87) observed that SC was 15–20 layers of flat cells that are thin
squames with a thickness of approximately 0.5 ␮m and a width of 30–40 ␮m.
A key issue is then ‘‘what is the polarity of the lipid bilayer environment’’ (see
Figs. 5 and 6) in which the solute is diffusing. Anderson and Raykar (88) observed
similar group contributions for polar, hydrogen-bonding substituents from permea-
bility and octanol–water data and suggested, as a consequence, that ‘‘the barrier
microenvironment resembles that of a hydrogen-bonding solvent.’’ It is important to
recognize that this barrier phase is more lipophilic than defined by Ksc⫺v , which
114 Roberts et al.
defines an average polarity of the lipid and ‘‘proteinaceous’’ phases involved in

partitioning.
The diagrammatic representation of the steady-state concentration–distance
profile in each skin region consistent with this model is shown in Figure 7A. It is
evident that there is a favorable distribution of the solute into the SC from the
vehicle, as defined by Ksc⫺v being more than 1. Furthermore, the steady-state con-
centration of solute in each phase declines with distance in accordance with the
concentration difference between the boundaries of the phase. It is also evident that
the concentration of solute at the inner face of the SC (Csc(i)) is not 0, arising from
a significant viable epidermal resistance or poor perfusion. When there is no resis-
tance and the viable epidermal phase approaches a perfect sink [Csc(i) = 0], as rep-
resented in Figure 10A, Eq. (34) reduces to
a
Ksc⫺v D sc Cv
J sssc = = k sc
p Cv (39)
hsc
1. Steady-State Function According to Convention 1
The model described here is also based on the assumption that the viable epidermis
approaches a perfect sink (see Fig. 10A). Noting av = ␥vCv , and applying convention
1 for standard state, Eq. (7) could be substituted into Eq. (39) to give the widely
quoted expression derived by Higuchi (18):
a
Ksc⫺v D sc a
Cv D sc␥v Cv D asc av
J sssc = = = (40)
hss hsc␥sc hsc␥sc
Equation (40) has commonly been quoted as the basis for an identical flux of
a solute from different saturated solutions through membranes. This basis is readily
seen for saturated systems by using convention 1 and equilibrium conditions, in that
if the vehicle is in equilibrium with pure solute then the activity of the solute in the
vehicle is the same as that for the pure solute. It is apparent, however, that an
interaction between the vehicle and the membrane could affect either the solubility
in the membrane (defined by ␥sc in convention 1) or Dsc . Higuchi (18) pointed out
that solutes with low-activity coefficients had low-escaping tendencies from the ve-
hicle, and thus, low rates of penetration of solutes through the skin. He reported the
following limiting activity coefficients of the nerve gas sarin in various solvents:
water 14, diethylene glycol 2.4, isoamyl alcohol 1.07, and benzyl alcohol 0.446.
Higuchi also recognized that phenol is less toxic when dissolved in vehicles in which
it has a high affinity (also expressed as: low activity or high solubility).
2. Steady-State Flux According to Convention 2
Roberts and Anderson (89) suggested that the expression for J sssc in terms of activities
is given by
D asc K sc⫺v
a
a*
v
J sssc = (41)
hsc
The maximum flux is given by Eq. (42) if (␥ sc * ⬵ ␥* v ⬵ 1) and saturated solutions
v = ␥ *S
are used (i.e., a* v v and Ksc⫺v = Ssc /Sv (Eq. (3) and Eq. (5))):
ss D asc Ssc D asc Ksc⫺v Sv

J sc = = (42)
hsc hsc
Skin Transport 115
Figure 10 (A) Sink conditions for flux of solute through the stratum corneum; (B) com-
parison of diffusional resistance of dermis and epidermis for aqueous solutions of alcohols.
(From Ref. 5.)
Where Ksc⫺v is the SC–vehicle partition coefficient, Ssc is the solubility of the solute
in the SC, and Sv is the solubility of the solute in the vehicle.
3. Steady-State Flux Through SC Under Nonsink Conditions
The steady-state flux of solute through the SC (J sssc) is defined by Eq. (39) when
Csc(i) = 0. When there is a significant concentration at the viable epidermis interface
(Figures 7 and 10A), a lower flux (given by Eq. (34)) is observed. Equation (34) can
also be rearranged to obtain an equation similar in form to that given in Eq. (39) (90):
J ssskin = k⬙C
p v (43)
where k⬙p is the effective permeability constant for the system, as defined not only
by the usual SC permeability coefficient k sc
p , but also, by the viable epidermis per-
116 Roberts et al.
meability coefficient k ve p , the permeability coefficient associated with the transfer from
the vehicle to the SC (not shown here), and an effective removal permeability co-
efficient k rp , reflecting blood perfusion in vivo or receptor sampling in vitro and
deeper tissue penetration:
1 1 1 1
= sc ⫹ ve ⫹ r (44)
k⬙p k p kp kp
Given that resistance to transport through a phase is defined by the reciprocal
of the permeability coefficient through that phase, Eq. (44) is simply stating that the
total resistance of the skin is given by the sum of its individual component resis-
tances, a similar expression is shown in Eq. (45) [Eqs. (49)–(53) in Chapter 3 express
this concept in mathematical terms]. Scheuplein (91) showed that the SC was the
major resistance to the skin permeation of water using such an expression for the
overall resistance of the skin Rskin [see Eq. (51) in Chap. 3].
Rskin = Rsc ⫹ Rve ⫹ RD = 9.1 ⫻ 106 ⫹ 6.3 ⫻ 103
⫹ 6.3 ⫻ 103 = 9.1 ⫻ 106 s cm⫺1 (45)
where Rsc is the resistance of the SC, Rve the resistance of viable epidermis, and RD
the resistance of dermis. As shown in Figure 10B, the dermal resistance contributes
significantly to the overall skin resistance for the longer-chain alcohols.
C. Solute Structure–Transport Relations

1. Aqueous Solutions
Most solute structure–transport studies have used permeability coefficients of solutes
determined using excised human epidermis or animal skin. Various solute structure–
epidermal permeability relations have been reported over nearly three decades (Table
4). The logarithm of octanol–water partition coefficient log Koct is often used to
define solute structure–transport relations as it is a relevant physicochemical prop-
erty, which can be readily determined experimentally. log Koct can also be estimated
by a fragment addition approach. A reduced form of Eq. (44) expresses the relative
importance of the polar and lipid pathways in SC penetration, together with the
resistance of the aqueous diffusion layer:
冋册
⫺1
1 1
k⬙p = ⫹ aq (46)
k sc
p,lipid ⫹ k p,polar
sc
kp
In early work, such as that for the phenolic compounds (24), regressions be-
tween log kp,sc and log Koct were made, recognizing the aqueous boundary layer
defined in Eq. (46), but not adequately recognizing the potential high dependence
on solute size (Fig. 11A). This work is predicated on a linear free energy relation
between log Ksc⫺water and log Ksolvent⫺water, as illustrated by the linear relation between
log Ksc⫺w and log Koct given in Eq. (19) (see Fig. 11B). If D asc /hsc is relatively constant
for the series of solutes chosen, then according to Eq. (39), log k sc p should be directly
related to log Koct when there is no evidence of either a polar pathway or another
barrier in series.
As discussed earlier, there is evidence for a polar pathway with a defined radius
for transport for very polar solutes. Transport through this pathway is expected to
Skin Transport
Table 4 Historical Development of Skin Transport–Structure Relations
Approach Relationship Ref.
1970s: Simplistic kp related to alcohol carbon chain length and Kamyl caproate Scheuplein (5,21,157)
kp phenylboronic acids, alcohols, and steroids related to Koctanol, MW, Lien and Tong (295)
and molar refraction
Flux related to Kmineral oil–water Michaels et al. (8)
kp and Ksc⫺v phenols, aromatic/aliphatic alcohols, and steroids related Roberts group
to Koctanol, MW, and H bond number (23,24,76,78)
1980s Maximum flux related to dipole moment, assuming polar and nonpolar Ando et al. (70)
pathways
Maximum flux related to MW and melting point Kasting et al. (92)
Dermal clearance rate related to Koctanol Siddiqui et al. (44)
kp related to partition coefficients Various including (88,296)
1990s: Selected multivariate
SAR of large kp datasets
Koctanol and MW log kp = ⫺2.74 ⫹ 0.71 log Koct ⫺ 0.0061MW, r 2 = 0.67, n = 93 Potts and Guy (94)
Fragmental analysis log kp = ⫺2.76 ⫹ 0.24 (C *) ⫺ 0.47 (aromatic rings) ⫹ 0.46 (halide) Pugh and Hadgraft (297)
⫺ 1.27 (amine) ⫺ 0.64 (nonaromatic) ⫺ 1.24 (steroid) ⫺ 0.47
(OH) ⫺ 0.325 (O) ⫺ 0.36 (amide), r 2 = 0.68, n = 90
MW and H bonds log kp = ⫺2.170 ⫹ 0.07(log Poct)2 ⫹ 0.835 log Koct ⫺ 0.265 Lien and Gao (298)
Hn ⫹ 1.844 log MW, r 2 = 0.956, n = ????
Solvatochromic or similar log kp = ⫺1.29 ⫺ 1.72⌺␣2 ⫺ 3.932⌺␤2 ⫹ 0.026Vx, r 2 = 0.94, n = 37 Potts and Guy (299)
approaches log kp = ⫺0.51 ⫺ 0.59␲2 ⫺ 0.63⌺␣2 ⫺ 3.42⌺␤2 ⫹ 1.8Vx, r 2 = 0.96, Roberts et al. (71)
n = 46
log kp = ⫺5.24 ⫹ 0.44R2 ⫺ 0.41␲2 ⫺ 1.63⌺␣2 ⫺ 3.28␤2 ⫹ 2.01Vx, Abraham (300,301)
r 2 = 0.96, n = 47
117
118 Roberts et al.
Figure 11 (A) The relation between log k sc p and log Koct for phenols; (B) the relation
between log Ksc⫺water and log Koct for phenols; (C) theoretical prediction of overall log k⬙p
versus log Koct for a solute with MW of 300 and the contribution of various diffusion pathways
to this transport; (D) the relation between activation energies (Ea) for phenols (●) and alcohols
(䡩). (From Refs. 78[A]; 24[B].)
be characterized by a k sc p,polar [upper limit 0.15 cm/h (84)] independent of log Koct.
⫺5
Kasting et al. (84) also refer to an aqueous boundary layer with k aq p = 1 ⫻ 10
cm/h. Figure 11C shows a plot of the predicted log permeability coefficient versus
log octanol–water, based on these considerations. It is apparent that a sigmoidal
curve results, reflecting transport by the polar pathway at low log Koct, a linear portion
reflecting log k sc
p,lipid versus log Koct, and a limiting aqueous boundary layer perme-
ability at high log Koct. This curvature is consistent with the phenols’ results shown
in Figure 11A. Roberts et al. (23) interpreted the sigmoidal decrease in the activation
energy for permeation for both alcohols and phenols with increasing log Koct (see
Fig. 11D) as additional evidence of the aqueous boundary layer effect they used to
explain the curvature of the plot of log k sc p versus log Koct at high log Koct values (see
Fig. 11A). Kasting et al. (92) suggested that log Dsc may be related to solute molec-
ular weight by Eq. (47):
Skin Transport 119
Dlipid = Do (MW)⫺b (47)

in which b should not be 1/3 to 1/2 as assumed for liquid diffusion by Scheuplein
and Blank (5), but >3 consistent with diffusion in polymer membranes and lipid
bilayers. Anderson and Raykar (88) reported that for the combined sets of methyl-
substituted phenols and 21 esters of hydrocortisone, the following relation was found:
log kp = 0.83 log Koct ⫺ 4.4 log MW ⫹ 6.4.
Kasting et al. (92) found that their data sets could be equally well described
by a free volume model, in which MV is the molecular volume of the solute and ␤
is a constant:
Dlipid = Do exp(⫺␤MV) (48)
The importance of molecular size as determinant of SC permeability coefficients has
also been recognized by Flynn (93). An extended plot of this data is shown in Figure
12. It is apparent that the study temperature is also a key determinant of kp. The line
divides the data into two. Some anomalies are readily apparent. First, most of the
high-permeability solutes above the line were studied at less than 30⬚C and those
below it at greater than 30⬚C. However, those above the line are low MW and have
low hydrogen-bonding capabilities (i.e., one or zero groups), whereas those below
the line have a MW > 300 and three or more hydrogen-bonding groups. The solutes
with two hydrogen-bonding groups are equally distributed.
Potts and Guy (94) suggested that the sigmoidal relations could be linearized
when the proper dependence of diffusivity on solute molecular size, as described by
the free volume model (see Eq. (48)], is derived. The combination of octanol–water
Figure 12 Confounding effects of solute size (datapoints approximately proportional to

molecular weight), number of solute hydrogen-bonding groups (where 䉭䊱 = 0, 〫⽧ = 1, ▫ 䡲
= 2, and 䡩 ● = 3) and study temperature (filled datapoints [䊱⽧䡲 ●] t ⱖ 30⬚C and open data-
points [䉭〫▫䡩] t < 30⬚C). The arbitrary line divides the dataset in two. (From Ref. 93.)
120 Roberts et al.
partition coefficients and solute size as determinants in skin transport is the basis for
perhaps the most widely quoted relation in this area—that of Potts and Guy (94).
They related log k sc
p (cm/h) to log Koct and molecular weight MW (a likely determinant
of Dsc) for 97 solutes and obtained the relation:
p = 0.71 log Koct ⫺ 0.0061MW ⫺ 2.74

log k sc r 2 = 0.69 (49)
If the slope of log Koct reflects the partitioning from water into SC lipid and
that for MW reflects the size effect in diffusion, the slope of 0.71 suggests that the
polarity in the intercellular diffusion pathway is less than octanol, but greater than
that observed with partitioning [see Eq. (19)]. An alternative analysis is to express
k sc
p in terms of groups of fragments from which solutes can be constructed. This
analysis recognizes that some different fragments, such as noncarbonyl carbon atoms
C# may promote absorption by enhancing partitioning, whereas others, such as amine,
hydroxyl, amide, and ‘‘O’’ groups slow diffusion by hydrogen bonding effects. Other
groups such as aromatic rings, nonaromatic rings, and steroids probably slow dif-
fusion by steric effects. Pugh et al. (95) reported the following relation for Flynn’s
dataset (n = 97):
p = ⫺2.76 ⫹ 0.241 (C ) ⫺ 0.470 (aromatic rings) ⫹ 0.460 (halide)

log k sc #
⫺ 1.27 (amine) ⫺ 0.644 (nonaromatic rings) ⫺ 1.24 (steroid)

⫺ 0.477 (OH) ⫺ 0.325 (‘‘O’’) ⫺ 0.356 (amide) r 2 = 0.68 (50)
By using an approach similar to that described earlier for SC–water partition
coefficients [see Eq. (27)], group contributions for the free energy for functional
group transfer during permeation can be estimated (88). Relatively high free energies
for permeation are required for polar groups ( — CONH2, 3.05 kcal/mol; — COOCH3,
1.25 kcal/mol; — COOH, 1.95 kcal/mol; — OH, 2.45 kcal/mol). The results of some
other sophisticated model-based structure–transport analyses are given in Table 4.
Barratt (96) used the Flynn dataset (93) and classified them into steroids (A), other
active molecules (B), and (C) the remainder. He reported a high correlation for
log k sc
p with log Koct, MP, and MV for steroids and small molecules, but not for the
others. Barratt also found that 90% of the variability was explained by the relation
(n = 60, group C plus ‘‘nonhydrocortisone group A’’; r 2 = 0.904):
p = ⫺0.00933MV ⫹ 0.82 log Koct ⫺ 0.00387MP ⫺ 2.355

log k sc (51)
The 1.5 order of magnitude higher value for the 12 hydrocortisone permeability
coefficients excluded from the analysis probably reflects that their permeability co-
efficients were measured at 37⬚C, whereas phenols, alcohols, and some other solutes
were measured at 25⬚C. This emphasizes the need to ensure that comparable ex-
perimental conditions have been employed in aggregating data from different
sources. Table 5 shows a listing of the data used in the analyses of Barratt and in
work by our group.
Solute structure–transport relations have also been studied by representing the
transport process in terms of models, such as parallel pathways of transport, and
multiple phases in series (71). Liron and Cohen (97) used regular solution theory
(discussed in Fig. 8) to show that the porcine skin permeability coefficient of pure
unbranched alkanoic acids (C2–C7) reached a maximum in the solubility parameter
range of 9.7–10 cal1/2 cm3/2. Groning and Braun (68), using O-acylglucosylceramide
Skin Transport 121
as a model for the intercellular lipoid matrix, showed that the steady-state flux of
three groups of solutes could be related to their three-dimensional solubility param-
eter differences with the solubility parameter of that intracellular lipid ([␦solute ⫺
␦O-acylglucosylceramide]2) (Fig. 13).
An alternative process is to recognize that transport is a series of steps, the
first being partitioning into the SC and the second diffusion through the SC. There-
fore, it is desirable to separate out the partition and diffusion components of k sc
p to
better understand these determinants of epidermal transport. The aqueous SC–water
partition coefficient Ksc⫺v was related to the octanol–water partition coefficient Koct
for 45 solutes by Eq. (19). As discussed previously, with Eq. (49), there may be a
need to assume a higher slope consistent with the more lipophilic transport pathway
than the partitioning environment. By using Eq. (19) in an attempt to remove the
influence of partitioning, Pugh et al. (98) showed for polyfunctional solutes (N = 53)
that
D asc
log = ⫺1.50 ⫺ 0.91␣ ⫺ 1.58␤ ⫺ 0.003MW r 2 = 0.94 (52)
h
where ␣ and ␤ are the H-bond donor ability (acidity) and the H-bond acceptor ability
(basicity) of the solute. Hence, diffusivity of a solute in the SC is both a function
of the hydrogen bonding of a solute and its size. The SC barrier was shown to be a
predominantly H-bond donor rather than acceptor with ␣sc:␤sc = 0.6:0.4. Also, log Dsc
is related to the number of hydrogen-bonding groups on a molecule in a nonlinear
manner and is suggestive of an adsorption isotherm, being maximal for small, non–
hydrogen-bonding molecules and reaching a low minimum with about four hydro-
gen-bonding groups (98). Comparing Eqs. (49) and (50), it is interesting to speculate
whether the melting point term found empirically in Eq. (49) is, in fact, a proxy
measure of hydrogen bonding, because hydrogen bond donor ability, and other in-
termolecular forces and molecular symmetry, are predictors of melting point (99).
2. Other Vehicles
The number of studies on structure–transport relations of solutes from other vehicles
is more limited. One of the earliest studies reported is that of Blank (100) in which
it was shown that whereas the permeability coefficient of alcohols through human
skin from saline increased with the number of carbon atoms, the permeability co-
efficients from nonaqueous vehicles decreased (Fig. 14A). Similar results have been
shown for phenolic solutes (see Figs. 11A and 14B). Roberts (101) attempted to
predict the observed relations using data for the epidermal permeability from aqueous
solutions and the estimated permeability coefficients. Arachis oil–solvent partition
coefficients were measured for a number of phenolic compounds using water–
ethanol combinations as solvents (see Fig. 14C). Noting that Ksc⫺v is related to Koct
for this series of solutes by Eq. (19) and the partition coefficient between arachis oil
and water Koil–water for these solutes are defined by
log Koil–water = 0.98 log Koct ⫺ 0.81 (53)
the apparent SC–arachis oil partition coefficient can be predicted by a suitable sub-
stitution of Eq. (53) into Eq. (19):
122
Table 5 Flynns’ Dataset for Skin Permeability Transport Relations
log PC log P Mpt Temp Barratt log mv log

Chemical (cm/h) (log Koct) (⬚C) (⬚C) Receptor a,b group Ksc (A3) Kbhex ␣ ␤ ␲* RCf Ref.
Aldosterone ⫺5.52 1.08 164 26 PEN/STR A 0.83 313.80 * 0.40 1.9 3.47 4.00 302
Amobarbital ⫺2.64 1.96 158 32 Buff 7.4 B 204.40 303
Atropine ⫺5.07 1.81 192 30 Ring 6.5 B 266.10 1.81 * * * * 8
Barbital ⫺3.95 0.65 192 32 Buff 7.4 B 155.60 303
Benzyl alcohol ⫺2.22 1.10 25 a a
C 0.61 89.24 ⫺0.62 0.33 0.5 0.87 1.36 21,5
4-Bromophenol ⫺1.44 2.59 68 25 DDH2O C 91.95 ⫺0.20 0.67 0.2 1.17 1.28 24,76
2,3-Butanediol ⫺4.40 ⫺0.92 25 30 a
C 89.64 21,5
Butanoic acid ⫺3.00 0.79 25 25 a
C 0.18 81.56 ⫺0.96 0.60 0.45 0.62 1.64 21,5
n-Butanol ⫺2.60 0.88 25 25 0.9% NaCl C 0.40 81.95 ⫺0.70 0.37 0.48 0.42 1.38 21,5
Butan-2-one ⫺2.35 0.28 25 30 a
C 75.20 ⫺0.25 0.00 0.51 0.70 0.93 21,5
Butobarbital ⫺3.71 1.65 127 32 Buff 7.4 B 188.10 303
4-Chlorocresol ⫺1.26 3.10 48 25 DDH2O C 103.90 0.36 0.65 0.22 1.02 1.29 24,76
2-Chlorophenol ⫺1.48 2.15 25 25 DDH2O C 87.90 24,76
4-Chlorophenol ⫺1.44 2.39 45 25 DDH2O C 88.17 ⫺0.12 0.67 0.2 1.08 1.28 24,76
Chloroxylenol ⫺1.28 3.39 116 25 DDH2O C 119.30 1.08 0.64 0.21 0.96 1.26 24,76
Chlorpheniramine ⫺2.66 3.81 25 30 Ring 6.5 B 246.20 3.39 * * * * 8
Codeine ⫺4.31 0.89 145 37 CPB 7.4 B 254.30 0.89 * * * * 304
Cortexolone ⫺4.13 2.52 208 26 PEN/STR A 1.36 317.00 ⫺1.00 0.35 1.57 3.45 3.36 302
Cortexone ⫺3.35 2.88 138 26 PEN/STR A 1.57 309.80 0.48 0.15 1.13 3.39 2.26 302
Corticosterone ⫺4.22 1.94 183 26 PEN/STR A 1.23 316.50 ⫺1.62 0.40 1.63 3.43 3.51 302
Cortisone ⫺5.00 1.42 228 26 PEN/STR A 0.93 320.80 ⫺0.55 0.35 1.84 3.50 3.38 302
o-Cresol ⫺1.80 1.95 34 25 DDH2O C 1.03 88.42 0.25 0.52 0.3 0.86 1.26 24,76
m-Cresol ⫺1.82 1.96 25 25 DDH2O C 1.03 88.97 ⫺0.35 0.57 0.34 0.88 1.40 24,76
Roberts et al.
p-Cresol ⫺1.75 1.95 34 25 DDH2O C 1.03 88.87 ⫺0.19 0.57 0.31 0.87 1.34 24,76
n-Decanol ⫺1.10 4.00 25 a a
C 178.30 * 0.37 0.48 0.42 1.38 21,5
2,4-Dichlorophenol ⫺1.22 3.08 60 25 DDH2O C 102.40 24,76
Diethylcarbamazine ⫺3.89 ⫺0.31 49 30 Ring 6.5 B 195.50 8
Digitoxin ⫺4.89 1.86* 240 30 Ring 6.5 B 682.60 8
Skin Transport
Ephedrine ⫺2.22 1.03 39 30 Ring 6.5 B 156.00 8
b-Estradoil ⫺3.52 2.69 179 26 PEN/STR A 1.66 255.30 0.88 0.95 3.30 302
Estriol ⫺4.40 2.47 282 26 PEN/STR A 1.36 262.40 1.40 1.22 3.36 302
Estrone ⫺2.44 2.76 254 26 PEN/STR A 1.66 249.50 0.56 0.91 3.10 302
Ethanol ⫺3.10 ⫺0.31 25 25 0.9% NaCl C ⫺0.31 50.72 ⫺2.10 0.37 0.48 0.42 1.38 21,5
2-Ethoxyethanol ⫺3.60 ⫺0.54 25 30 a
C 90.21 * 0.30 0.83 0.50 1.92 21,5
Ethyl benzene 0.08 3.15 25 25 DDH2O C 97.78 3.00 0.00 0.15 0.51 0.28 305
Ethyl ether ⫺1.80 0.83 25 30 a
C 83.31 0.60 0.00 0.45 0.25 0.82 21,5
4-Ethyl phenol ⫺1.46 2.40 45 25 DDH2O C 104.60 0.23 0.55 0.36 0.90 1.41 24,76
Etorphine ⫺2.44 1.86 215 37 Tris 7.4 B 368.20 1.86 * * * * 306
Fentanyl ⫺2.25 4.37 84 37 CPB 7.4 B 314.70 4.37 * * * * 304,307
Fentanyl (2) ⫺2.00 4.37 84 30 Ring 6.5 B 314.70 4.37 * * * * 3
Fluocinonide ⫺2.77 3.19 311 37 Succ 4 B 412.00 $
Heptanoic acid ⫺1.70 2.50 25 25 a
C 1.78 129.30 0.45 0.60 0.45 0.60 1.64 21,5
n-Heptanol ⫺1.50 2.72 25 25 0.9% NaCl C 1.48 130.10 1.01 0.37 0.48 0.42 1.38 21,5
Hexanoic acid ⫺1.85 1.90 25 25 a
C 1.08 113.80 0.24 0.60 0.45 0.60 1.64 21,5
n-Hexanol ⫺1.89 2.03 25 25 0.9% NaCl C 1.00 112.80 0.45 0.37 0.48 0.42 1.38 21,5
Hydrocortisone (2) (HC) ⫺3.93 1.53 214 32 Buff 7.4 A 0.85 326.00 ⫺2.04 0.70 1.87 3.49 4.36 303
HC Dimethylsuccinamate ⫺4.17 2.03 223 37 Succ 4 A 437.50 79,309
HC Hemipimelate ⫺2.75 3.26 112 37 Succ 4 A 449.20 79,309
HC Hemisuccinate ⫺3.20 2.11 171 37 Succ 4 A 401.80 79,309
HC Hexanoate ⫺1.75 4.48 152 37 Succ 4 A 424.00 79,309
HC 6-OH-hexanoate ⫺3.04 2.79 144 37 Succ 4 A 1.58 432.10 —* —* —* 79,309
HC Octanoate ⫺1.21 5.49 115 37 Succ 4 A 455.90 79,309
HC Pimelamate ⫺3.05 2.31 185 37 Succ 4 A 452.70 79,309
HC Propionate ⫺2.47 3.00 196 37 Succ 4 A 375.80 79,309
HC Succinamate ⫺4.59 1.43 227 37 Succ 4 A 405.30 79,309
Hydromorphone ⫺4.82 1.25 267 37 CPB 7.4 B 250.70 1.25 * * * * 304
Hydroxypregnenolone ⫺3.22 3.00 150 26 PEN/STR A 317.60 302
17-Hydroxyprogesterone ⫺3.22 2.74 220 26 PEN/STR A 1.60 311.60 0.40 0.25 1.31 3.35 2.73 302
Isoquinoline ⫺1.78 2.03 28 32 NaOH 7.4 C 96.32 303
⫺2.04 ⫺0.52
123
Me-4-hydroxy benzoate 1.96 128 25 PBS 7.4 C 111.90 0.69 0.45 1.37 1.76 24,76
124
Table 5 Continued
log PC log P Mpt Temp Barratt log mv log

Chemical (cm/h) (log Koct) (⬚C) (⬚C) Receptor a,b group Ksc (A3) Kbhex ␣ ␤ ␲* RCf Ref.
Meperidine ⫺2.43 2.72 25 37 CPB 7.4 B 224.60 2.72 * * * * 304

Methanol ⫺3.30 ⫺0.77 25 25 0.9% NaCl C ⫺0.22 33.21 ⫺2.80 0.43 0.47 0.44 1.44 21,5
Methyl HC Succinate ⫺3.68 2.58 143 37 Succ 4 A 418.00 79,309
Methyl HC Pimelate ⫺2.27 3.70 142 37 Succ 4 A 465.30 79,309
Morphine ⫺5.03 0.62 200 37 CPB 7.4 B 237.70 79,309
2-Naphthol ⫺1.55 2.84 123 25 DDH2O C 1.52 106.20 0.30 0.61 0.4 1.08 1.56 24,76
Naproxen ⫺3.40 3.18* 155 a
PBS 7.4 B 188.50 3.18 * * * * 309
Nicotine ⫺1.71 1.17 25 32 NaOH 9.2 B 150.80 303
Nitroglycerin ⫺1.96 2.00 25 30 Ring 6.5 C 135.20 8
3-Nitrophenol ⫺2.25 2.00 98 25 DDH2O C 91.32 1.23 0.79 0.23 1.57 1.50 24,76
4-Nitrophenol ⫺2.25 1.96 115 25 DDH2O C 91.53 ⫺2.15 0.82 0.26 1.72 1.59 24,76
NDELA ⫺5.22 a
25 a a
C 114.60 21,5
n-Nonanol ⫺1.22 3.62 25 25 0.9% NaCl C 159.90 * 0.37 0.48 0.42 1.38 21,5
Octanoic acid ⫺1.60 3.00 25 25 a
C 2.15 146.40 0.66 0.60 0.45 0.60 1.64 21,5
Roberts et al.
n-Octanol ⫺1.28 2.97 25 25 0.9% NaCl C 3.00 144.50 * 0.37 0.48 0.42 1.38 21,5
Ouabain ⫺6.11 a
190 25 Ring 6.5 B 490.30 8
Pentanoic acid ⫺2.70 1.30 25 25 a
C 0.48 98.75 ⫺0.92 0.60 0.45 0.60 1.64 21,5
n-Pentanol ⫺2.22 1.56 25 25 0.9% NaCl C 0.70 96.99 ⫺0.40 0.37 0.48 0.42 1.38 21,5
Phenobarbital ⫺3.34 1.47 178 32 Buff 7.4 B 180.50 303
Skin Transport
Phenol ⫺2.09 1.46 42 25 DDH2O C 0.73 73.12 ⫺0.82 0.60 0.3 0.89 1.36 24,76
Pregnenolone ⫺2.82 3.13 192 26 PEN/STR A 1.70 310.10 3.77 0.32 1.18 3.29 2.59 302
Progesterone ⫺2.82 3.77 130 26 PEN/STR A 2.02 304.30 3.77 0.00 1.14 3.29 2.08 302
n-Propanol ⫺2.85 0.25 25 25 0.9% NaCl C 0.30 65.51 ⫺1.52 0.37 0.48 0.42 1.38 21,5
Resorcinol ⫺3.62 0.80 113 25 DDH2O C 0.25 79.80 1.10 0.58 1.00 24,76
Salicyclic acid ⫺2.20 2.26 160 32 Buff 7.4 C 94.75 303
Scopolamine ⫺4.30 1.24 25 30 Ring 6.5 B 263.90 1.24 * * * * 8
Styrene ⫺0.19 2.95 25 a a
C 87.66 0.44 0.00 0.16 0.65 0.29 310
Sucrose ⫺5.28 ⫺2.25 187 30 Succ 4 C 267.40 $
Sufentanyl ⫺1.92 4.59 97 37 CPB 7.4 B 346.40 304,307
Testosterone ⫺3.40 3.31 155 26 PEN/STR A 1.36 270.20 3.31 0.32 1.19 2.59 2.60 302
Thymol ⫺1.28 3.34 51 25 DDH2O C 3.30 135.80 1.62 10.0 0.52 0.44 1.51 24,76
a a
Toluene 0.00 2.75 25 C 82.84 2.89 0.00 0.14 0.52 0.26 310
2,4,6-Trichlorophenol ⫺1.23 3.69 66 25 DDH2O C 117.30 24,76
Water ⫺3.30 ⫺1.38 25 25 Water C 16.75 21,5
3,4-Xylenol ⫺1.44 2.35 68 25 DDH2O C 1.28 104.60 0.28 0.56 0.39 0.86 1.47 24,76
a
Data not stated or unavailable
b
Succ, succinimate phosphate buffer; PEN/STR, aqueous solution of penicillin and streptomycin; Ring, Ringer’s buffer; CPB, citrate–phosphate buffer; Buff, buffer
(specific constituents not stated); DDH2O, distilled water.
125
Figure 13 Dependence of transepidermal flux (Jss, mol cm⫺2 h⫺1) on the difference in
solute (␦solute) and o-acylglucosylceramide (␦o-acylglucosylceramide) solubility parameter [␦solute ⫺
␦o-acylglucosylceramide] for (A) steroids; (B) a mixture of various drugs; and (C) analgesics. (From
Ref. 68.)
Skin Transport 127
Figure 14 (A) Effect of alkyl chain length on the permeability of alcohols through human
skin from olive oil (䉭), isopropyl palmitate (●), and saline (䡩); (B) epidermal permeability
coefficients (kp) and octanol–water partition coefficients (Koct) for several phenols from an
arachis oil vehicle and an aqueous receptor (䡩), ethanol 47% in water vehicle and receptor
(䡲), ethanol 95% in water vehicle and receptor (▫), and arachis oil in vehicle and receptor
(●) at 25⬚C; (C) aqueous alcohol partition coefficients (K) versus octanol–water partition
coefficients (Koct) for a group of phenols. Solvents used were water (䡩), ethanol 47% in water
(䉭), and ethanol 95% in water (▫); (D) relation between permeability coefficients for the
group of phenols and the relative humidity generated by the receptor phase used in perme-
ability studies. (From Refs. 100[A]; 101[B–D].)
128 Roberts et al.
log Ksc⫺arachisoil = log 冉

Csc Cw
冊
Cw Carachisoil
= log Ksc⫺water ⫺ log Ksc⫺oil
= ⫺0.41 log Koct ⫹ 0.71 (54)
Thus, a negative slope is predicted. The slope of the observed relation for an arachis
oil vehicle to aqueous receptor ⫺0.34 was of a magnitude comparable with that
predicted of ⫺0.41. A lower observed slope would be consistent with a higher li-
pophilicity in the SC diffusion pathway as discussed earlier [see Eq. (34)]. A com-
parison of predictions for other vehicles with observed values shown in Table 6
shows that observed and predicted slopes are similar in magnitude.
It should be recognized that linear relations between logarithms of the partition
coefficients of solutes in different solvents with the logarithm of the octanol–water
partition coefficient have been shown for many polar and semipolar solvents (75).
Leo et al. (75) point out that for nonpolar solvents, the relation with the logarithm
of the octanol–water partition coefficient is poor. A satisfactory correlation is ob-
tained when a hydrogen-bonding constant is added to this relation.
Hagedorn–Leweke and Lippold (102) quantified the transdermal permeabilities
and maximum fluxes of various sunscreens and antimicrobial compounds applied as
saturated solutions in a propylene glycol–water mixture applied to human skin in
vivo. A linear relation was found between the logarithms of permeability coefficients
of the penetrants and their corresponding octanol–vehicle partition coefficients. The
slope of 0.38 reported in their relation may be explained as being much less than
unity as a consequence of the SC being more ‘‘polar’’ than octanol, as deduced from
the aqueous partition studies [see Eq. (19)]. An additional reduction in the slope is
also expected through the cosolvency effect of propylene glycol. A slope of 0.32
was obtained in a later study on the uptake of homologous esters of nicotinic acid
by the skin of healthy volunteers (103).
3. Hydration
It should be emphasized that this approach can be used to predict only slopes and
not permeability coefficients, the absolute magnitude of which are also influenced
by vehicle effects on membrane properties, such as hydration. Indeed, as shown in
Figure 14D, the permeability coefficients in the different vehicle systems used vary
by two orders of magnitude and appear to be related to water content, as defined by
the relative humidity for the receptor solutions used. Scheuplein and Blank (5) had
previously reported that the diffusion of alcohols differs by 100-fold between hy-
drated and dehydrated SC. The role of skin hydration in percutaneous absorption has
been reviewed (104). The mechanism of water enhancement of skin permeability is
considered in a later section on vehicle–skin interactions. In their overview of the
effects of hydration on solute penetration, Roberts and Walker (104) noted that the
reported results were equivocal, with some studies reporting increases of up to ten-
fold for some substances, and others showing a very small effect. They commented
that the major effect of hydration may be on solubility in skin lipids, citing the
greater enhancement of pure glycol salicylate relative to methyl or ethyl salicylate
by hydration using the human in vivo data of Wurster and Kramer (105). A similar
enhancement was shown for methyl ethyl ketone (106). Figures 15A and B sum-
marize these in vivo results. Occlusion has been reported to have a greater effect on
Skin Transport 129
Table 6 Predicted and Observed Slopes of log Permeability

Coefficients and Partition Coefficients Versus log Octanol–Water
Partition Coefficients for a Series of Phenolic Compounds
Slope
log k sc⫺v
p vs. log Koct
Vehicle log Koil⫺v vs. log Koct Observed Predicted
Water a,b 0.57c 0.6 0.57

Arachis oila 0.98 ⫺0.34 ⫺0.41
Ethanol 47%b 0.51 ⫺0.11 0.06
Ethanol 95%b 0.18 ⫺0.21 0.39
Arachis oilb 0.98 ⫺0.25 ⫺0.41
a
Aqueous receptor.
b
Receptor same composition as donor vehicle.
c
Stratum corneum–water.
the percutaneous absorption of the more lipophilic radiolabeled tracer steroids in

vivo than for nonoccluded conditions.
In their recent summary of the effects of occlusion on percutaneous absorption,
Bucks and Maibach (107) noted that occlusion increases the normal water content
of the SC from between 5 and 15% to 50%, the temperature from 32⬚C to as much
as 37⬚C, and the skin pH from 5.6 to 6.7. Hence, a number of mechanisms may be
associated with humidity-induced penetration changes. Interpretation of studies in
vitro is further complicated by the use of hairless mouse and human epidermal
membranes to examine hydration effects. Bond and Barry (108) have shown that,
after treatment with saline, the permeability coefficient of 5-fluorouracil through hair-
less mouse skin sharply increased in permeability after approximately 50 h of hy-
dration, suggesting that the SC had started to disrupt, whereas the flux through human
abdominal skin remained unchanged.
4. Maximal Flux
By definition, if neither the vehicle nor the solute affects the membrane, the same
maximum flux will be observed for a solute from a range of vehicles irrespective of
the range of concentrations [Eq. (55)]. Hence, as Ksc = Ssc /Sv , an increase in Sv for
a given vehicle is counterbalanced by a reduction in Ksc, giving a constant flux.
Figure 6A shows an example of a constant flux for a solute through an inert mem-
brane from a number of saturated solutions in different vehicles, which show no
apparent interaction with the membrane. Hence, the maximal flux is defined by the
solubility of the solute in the SC (Ssc), which could also be expressed as the product
of the partition coefficient (Ksc) and the solubility in the vehicle (Sv ).
max D asc Ssc D sc

a
Ksc Sv
J sc = = = k psc Sv (55)
hsc hsc
Consistent with earlier derivations, D asc is the apparent diffusivity, Ssc is the solubility
of the solute in the SC, k sc
p is the epidermal permeability coefficient, and Sv is the
solubility in the vehicle. Implicit in Eq. (55) and the assumption that Ksc⫺v is inde-
130 Roberts et al.
Figure 15 (A) Cumulative urinary salicylate excretion data showing the influence of hy-
dration on the percutaneous absorption rate of glycol salicylate, hydrated system (●) and
dehydrated system (䡩); (B) expired air concentration data for the elimination of methyl ethyl
ketone showing the influence of hydration on percutaneous absorption rate: hydrated system
(▫); normal system (䡩); and dehydrated system (䉭); (C) percutaneous absorption of four
steroids in humans as a function of penetrant octanol–water partition coefficient and occlusion.
(From Refs. 105[A]; 68[B]; 107[C].)
Skin Transport 131
Figure 15 Continued
pendent of concentration for any vehicle, is that the flux (Jsc) for a given solute
concentration (Cv) in a given vehicle [see Eq. (43)] can be predicted from J max sc and
Sv; that is, Jsc = J max
sc Cv /S v. It is emphasized that such a prediction assumes that neither
the solute nor the vehicle has affected D asc or Sc. By using the data in Table 3 from
Sloan et al. (62), Zatz (109) estimated a theoretical Ksc Sv (i.e., =Ssc) and thence
J max
sc /(Ksc Sv). In theory, this latter ratio should be constant [see Eq. (55)]. In practice,
the ratio for the various vehicles used to apply theophylline to hairless mouse skin
were isopropyl myristate 50, octanol 490, dimethyl formamide 4.7, propylene glycol
3.2, ethylene glycol 2.2, and formamide 2.7. Evidently, the vehicles had caused
varying changes in skin permeability. Twist and Zatz (110) and Barry et al. (111)
have previously shown that the respective constant steady-state flux for methylpar-
aben through polydimethylsiloxane membrane and the bioavailability for the topical
steroid desonide, as measured by vasoconstrictor response, were independent of for-
mation when applied as saturated formulations.
Equation (55) suggests that the maximal flux of a solute through SC J max sc can
be enhanced by three mechanisms: (a) increasing the diffusivity of a solute in the
SC; (b) affecting the partitioning between the SC lipids and other SC components,
or (c) by increasing its solubility in SC lipid components. Tojo et al. (112) showed
that the permeation steady-state rate of progesterone and its hydroxy substituents
across the intact skin and stripped skin of the hairless mouse was approximately
proportional to the solubility of drugs in the SC or in the viable skin, respectively.
Note that Eq. (55) is a reduced form of a more general expression, analogous
to Eq. (34), which recognizes the potential effects of the viable epidermal resistance:
J max
sc p Sv (1 ⫺ fSve )
= k sc (56)
ss
where fSve is the solute concentration in the viable epidermis C , expressed as a
ve
fraction of its solubility in the viable epidermis Sve (i.e., fSve = C ssve /Sve).
132 Roberts et al.
One means of increasing its solubility beyond normal saturation is to produce

a supersaturated state by processes such as cooling, evaporation of vehicle, solvent
additions, and change of pH. Pellet et al. (113) showed that supersaturation of piroxi-
cam in a propylene glycol–water cosolvent vehicle resulted in higher (supersatu-
rated) concentrations of solute in the SC with a resultant higher flux through the SC.
Schwarb et al. (114) have recently reported similar results with fluocinonide. Had-
graft (115) has recently reviewed issues associated with using supersaturated systems
in transdermal delivery.
Maximal flux is both solute- and vehicle-dependent and can be predicted by a
number of approaches. One approach is to apply the solubility parameter approach
and estimate the solubility of a solute in SC lipids applying Eq. (12). Assuming a
similar mole heat of fusion for solutes and an ‘‘ideal’’ solution, the logarithm of the
mole fraction solubility is linearly related to the reciprocal of the melting point [see
Eq. (9)]. Several studies have applied this relation to percutaneous absorption, in-
cluding Guy and Hadgraft (116) (Fig. 16) and Kai et al. (45). An implicit assumption
of a constant diffusivity is being made here in relating the logarithm of the maximum
flux to the reciprocal of the melting point.
Because maximal flux is both solute- and vehicle-dependent, extending Eqs.
(5) and (19) under ideal conditions (i.e., ␥ *
sc = ␥ *
v = 1) in an aqueous system, log Ssc
⯝ 0.59 log Soct. Substituting into Eq. (55) yields
D asc
log J max
sc = log ⫹ 0.59 log Soct (57)
hsc
where Ssc is the solubility of the solute in the octanol. Thus, it is expected that there
will be a parabolic relation in anticipation of solutes with a polarity similar to 0.59
log Soct, which will be most soluble in the SC (3). Figure 17A shows that Yano and
co-workers’ (117) in vivo data for nonsteroidal anti-inflammatory solutes has a max-
imum log Poct at about 2.5. Consistent with Eq. (55), Tojo et al. (112) also observed
that the solubility of progesterone and its hydroxyl derivatives in the SC increased
with the lipophilicity of the penetrant, and they reported that the diffusivity of these
solutes across the SC and viable skin appeared to be independent of their polarity.
Kasting et al. (92) related log J max
sc to log Soct and molecular volume for 35
compounds. By using ANOVA, they showed that log Soct accounted for 53% of the
observed variance, and molecular volume accounted for another 21% (i.e., r 2 = 0.74
for a multivariate regression). In 15 healthy volunteers, Le and Lippold (103) studied
and estimated the uptake of homologous esters of nicotinic acid by the skin. Per-
meabilities and maximum fluxes J max sc were determined from the concentration de-
crease of the aqueous solutions after fixed time periods. Although no clear depen-
dence was observed between the maximum flux Jmax and the octanol solubility Soct
>
Figure 16 (A) Steady-state flux of hydrocortisone through polydimethylsiloxane membrane
from saturated solutions in various vehicles that themselves are not sorbed to any significant
extent by the membrane (mean ⫾ SD); (B) an inverse relation between drug flux at steady
state through excised human skin and penetrant melting point (MP) (dg, digitoxin; ou, oua-
bain; es, estradiol; at, atropine; ch, chlorpheniramine; fn, fentanyl; sc, scopolamine; ng, nitro-
glycerin; dc, diethylcarbamazine; ep, ephedrine). (From Refs. 68[A]; 116[B].)
Skin Transport 133
134 Roberts et al.
Figure 17 (A) In vivo percutaneous penetration of a series of salicylates (䡩) and of other
nonsteroidal anti-inflammatory drugs (●) plotted as a function of log Koct; (B) in vivo per-
cutaneous absorption of phenols in humans under occluded (▫) and protected (䡲) conditions
as a function of penetrant log Koct. (From Refs. 117[A]; 68;324[B].)
Skin Transport 135
of the esters, a linear relation was found between log J max sc ⫹ (1 ⫺ 0.32)log Koct
versus log Soct. The maximum fluxes of transdermally absorbed sunscreens and other
solutes were also estimated from the disappearance of the solutes from saturated
solutions of a propylene glycol–water mixture which was applied to human skin in
vivo (102). The maximum fluxes were then related to the octanol–water partition
coefficients and octanol solubilities for the compounds.
sc = ⫺0.18
Hinz et al. (118) have suggested that the significant parabolic (log J max
⫹ 1.35 log Koct ⫺ 0.30 [log Koct]2) and bilinear (log Jmax = ⫺0.17 ⫹ 1.08 log P ⫺
1.95 [log(␤ ⭈10 log Koct ⫹ 1)]) dependencies obtained may reflect a change in the
rate-limiting transport step (for compounds of high log P) from diffusion across the
SC to partitioning at the SC–viable epidermis interface. Bucks and Maibach (107)
also observed a parabolic relation for phenols in vivo using deposited solids (see
Fig. 17B). Hostynek and Magee (119) concluded in their analysis of human in vivo
skin absorption data for 28 solutes that the maximum flux was increased by
occlusion.
Roberts and Sloan (120) modified the Potts and Guy equation [see Eq. (49)]
to apply to more lipophilic and more polar vehicles than skin. Maximum fluxes for
39 prodrugs from saturated solutions in isopropyl myristate were best described by
the relation:
log J max
sc = ⫺0.193 ⫹ 0.525 log Sipm
⫹ (1 ⫺ 0.525)log Sw ⫺ 0.00364MW r 2 = 0.945 (58)
where Sipm is the solubility in isopropyl myristate, Sw is the solubility in aqueous pH
4.0 buffer and MW is molecular weight. The significant difference to the earlier
expressions derived to date is the inclusion of the water solubility term, with a slope
suggesting that it is almost as important as lipid solubility in predicting flux. A similar
model was used to analyze the maximum flux data of Kasting et al. (92) from
propylene glycol (PG), but using propylene glycol solubility SPG instead of Sw:
log J max
sc = ⫺1.673 ⫹ 0.599 log Sipm
⫹ (1 ⫺ 0.599)log SPG ⫺ 0.00595MW r 2 = 0.852 (59)
This work, using solubilities in polar and nonpolar solvents as predictors of flux,
suggests that the bipolar nature of the SC lipids needs to be recognized in modeling
maximum fluxes from different vehicles.
To date, our analysis has been limited to the prediction of maximum fluxes for
solutes on the basis that, consistent with the theoretical considerations defined by
Raoult’s or convention 1, the activity of a saturated system from any vehicle should
be identical and equal to that for a pure liquid compound. In reality, solutes and
vehicles interact with the skin affecting both solubility and diffusivity. Deviations in
maximum fluxes between vehicles or in the ratios of a flux through epidermis and
an inert membrane between vehicles (121) is evidence of either a drug–skin or a
vehicle–skin interaction as discussed in the following section.
IV. FACTORS AFFECTING SKIN FLUX

The key determinants of epidermal flux are solute concentration in the SC (Csc), the
effective diffusivity in the SC, and the potential buildup of solute concentrations in
136 Roberts et al.
the viable epidermis [see Eq. (28) and Fig. 7]. A maximum flux is attained, therefore,
at the solubility of the solute in the SC [see Eq. (55)], recognizing that solubility
may include the thermodynamically unstable potential supersaturation. The concen-
trations of solute in the SC may be related to those in the vehicle by a partition
coefficient. Our analysis also shows that the apparent diffusivity is a function of both
the diffusivity of unbound solute down the intercellular lipid pathway as well as the
fraction of solute unbound in this pathway. Finally, solutes may be transported in
the vapor phase as has been shown for the alcohols (5) and for a homologous series
of acetate esters (122). In the latter, the vapor pressure of the pure solutes decreased
as the alkyl chain was increased. The observed SC permeation rate decreased with
the decrease in vapor pressure. We now consider factors affecting each determinant.
A. Solute Concentration in Vehicle

Equation (39) suggests that J sssc should be linearly related to the concentration of
solute in the vehicle Cv , up to the solute saturation solubility in the vehicle. There-
after, at higher solute concentrations, a suspension exists and the solute flux is the
maximal flux, which has been discussed earlier. Hence, Barry et al. (57) showed that
benzyl alcohol vapor flux was linearly related to benzyl alcohol activity, suggesting
that percutaneous absorption is controlled by thermodynamic activity when the ve-
hicle has no effect on the SC barrier. It may be important to recognize that, if a
solute activity is defined as fractional solubility (as implied by convention 1), then
the flux from different vehicles will be the same for all fractional concentrations.
Flux is not necessarily linear with fractional concentrations, as illustrated by the
deviations from Raoult’s law for benzyl alcohol vapor concentration versus mole
fraction (57).
If, on the other hand, solute activities are those measured (as implied by
convention 2), for a given concentration, the highest flux will be seen from the
vehicle in which the solute is least soluble with identical fluxes being apparent when
both vehicles are saturated. These deductions are based on the assumption that the
effects of the two vehicles on the skin are the same and there is no nonlinearity in
flux versus Cv profiles.
A nonlinearity in flux–convention profiles may also arise if the concentration
of the solute used is sufficiently high to affect the integrity of the SC barrier, or if
nonsink conditions preclude the attainment of equilibrium during the course of the
experiment (90). Figure 18 shows three examples of nonlinearity. In Figure 18A a
positive deviation from linearity can be shown to arise as a consequence of solute
effects on SC permeability by comparison with a flux through an inert membrane.
The proportionality of flux to vehicle concentration through the inert membrane is
evidence that the effects do not arise from alterations in the activity coefficients of
>
Figure 18 (A) Penetration flux of phenol through rat skin (䡩) and polyethylene film (●)
at 37⬚C for various concentrations of phenol in water; (B) Fluxes of octylsalicylate at various
concentrations: (left-hand axis: polyethylene membrane (䡩) and nylon membrane filter (䉭);
(right-hand axis) human epidermis (䊱) and dialysis membrane (●); (C) penetration flux of
pentanol from an olive oil vehicle through human skin (䡩) and polyethylene film (●). (From
Refs. 101[A,C]; 325[B].)
138 Roberts et al.
the solutes in the vehicle. The permeability coefficient is relatively constant; but
increases abruptly at about 2% phenol as a result of several changes (123). Figure
18B shows a negative deviation from a linear flux versus concentration relation
through membranes other than nylon. The negative deviation here arises from octyl
salicylate self-association in the vehicle at high concentrations and is accounted for
in Eq. (40) by a reduction in Ksc⫺v as a result of a decreased ␥v [see Eq. (7)]. The
positive deviation from linearity for the nylon membrane suggests that octyl salic-
ylate has increased flux by plasticization or other effects on the membrane. Figure
18C shows a negative deviation in flux from a linear flux concentration relation as
a result of an effect such as dehydration and a reduction in Dsc , arising from the
high concentration of solute in the vehicle.
Twist and Zatz (124) reported a parabolic relation between flux and solute
concentration for methylparaben and propylparaben through polydimethylsiloxane
membrane from 1-propanol. They proposed that the propanol vehicle is sorbed by
the membrane and creates an environment (‘‘clusters’’) in which the paraben can
dissolve. The resultant paraben membrane concentration and flux is higher than if
the propanol was not present in the membrane. At high paraben concentrations, the
propanol activity in the vehicle is reduced: less partitions into the membrane. The
paraben solubility and flux therefore decreases. Another nonlinearity that may arise
is the nonlinear binding of components to SC. Bronaugh and Congdon (125) showed
that hair dye binding to human epidermis could be described by a Scatchard plot,
and that permeability values followed the rank order of dye permeability and par-
alleled the partition coefficients only when the binding sites were saturated. Wurster
(122) has reported the adsorption isotherm for sarin’s uptake on p-dioxane-condi-
tioned callous tissue.
B. Drug–Vehicle Interactions
Figure 19A shows that the penetration flux of phenol decreases through both rat skin
and polyethylene owing to a higher affinity of dimethyl sulfoxide (DMSO) than for
water, even though DMSO is a very strong penetration enhancer. A similar profile
is observed for glycerol, an agent that has less effect on the epidermis (see Fig. 19B)
(123). The effect of the DMSO and glycerol relative to water is simply a reduction
in Ksc⫺v owing to a greater solubility [see Eq. (42)] or low-activity coefficient owing
to the high affinity [see Eq. (40)] of phenol for these vehicles than for water. Indeed
when the logarithm of the penetration flux is plotted against the percentage glycerol,
a linear relation is observed (see Fig. 19C) consistent with the relation (126).
log Js = log Js water ⫹ (log Js (glycerol) ⫺ log Js (water))(1 ⫺ fg) (60)
Where (Js) is the penetration flux for a given binary composition, Js (glycerol) and
Js (water) are the penetration fluxes of phenol from glycerol and water vehicles, re-
spectively, and fg is the fraction of glycerol in the glycerol–water vehicle.
Vehicles may also affect drug release by a diffusion limitation in the vehicle
with a range of expressions being presented (see Chap. 3). Other effects such as
vehicle evaporation, dissolution kinetics, solvent flux through stratum corneum, and
changes in vehicle composition with time are dealt with elsewhere in this book and
the literature (83,126,127). In the present context, the shape of the cumulative amount
versus time profile is often indicative of whether flux is membrane-limited as dis-
Skin Transport 139
Figure 19 (A) The relation between flux of phenol through polyethylene film and excised
skin, and the percentage DMSO in the vehicle; (B,C) penetration flux of phenol through
excised skin from a 5% w/v (▫) and a 2% w/v (䡩) aqueous glycerol solution and through a
polyethylene membrane from a 2% w/v (●) aqueous glycerol solution. (From Refs. 330[A];
101[B,C].)
140 Roberts et al.
Figure 19 Continued
cussed to date or limited by diffusion in the vehicle. In general, when uptake is

limited by diffusion in the vehicle and the percentage of solute released from the
vehicle is less than 30%, an approximate form of the diffusion equation from a slab
into a perfect sink can be used. For a solute dissolved in the vehicle, the amount
released per unit area (Mt) over time t can be related to the diffusivity of the solute
in the vehicle Dv and the initial concentration of solute in the vehicle C0 by:
冉冊
1/2
Dv t
Mt ⬵ 2C0 (61)
␲
The derivation of the expression for a suspension of drug with an apparent concen-
tration Ctotal (total amount of dissolved plus undissolved solid per unit volume) and
a solubility in the vehicle Sv , described originally by Higuchi, has been reported by
many authors, including Barry (126). When Ctotal >> Sv , Mt is given by
Mt ⬵ (2Ctotal Sv Dv t)1/2 (62)
Hence, a square root relation for the amount of solute released with time is expected,
whether or not the vehicle is saturated with solute. The range of vehicles used in
topical applications is discussed in greater detail in the following chapters.
C. Water Enhancement of Permeability

The state of hydration of the normal skin has been ranked next to the nature of the
penetrating molecule as the most important factor in the rate of percutaneous passage
of any substance (128). However, the reported effectiveness of hydration on skin
penetration appears equivocal (104). It is well recognized that the SC swells contin-
uously on immersion in water, absorbing as much as ten times the dry weight. The
water is bound within the intracellular keratin. Permeability increases rapidly initially
and then slows down to a steady-state diffusion. Alonso et al. (129,130) used spin-
Skin Transport 141
label electron-spin resonance technique to monitor the effect of hydration on the

molecular dynamics of lipids at C-5, C-12, and C-16 positions of the alkyl chain.
They found that an increased hydration of neonatal SC led to an increase in mem-
brane fluidity near the membrane–water interface region and less so in the deeper
hydrophobic core. Solid-state nuclear magnetic resonance (NMR) has also been used
to show a greater mobility of the skin components in the presence of water (131).
D. Other Vehicles–Skin Interactions

Vehicles or their components can interact with SC lipids to enhance skin permea-
bility. Two books have been published relatively recently on this subject (132,133).
In addition, Williams and Barry (134) and Davies et al. (see Chap. 6) have reviewed
this area. Williams and Barry (134) identified the potential modes of actions of
accelerants by the lipid–protein-partitioning (LPP) theory, which summarizes the
mechanisms of action as the following:
1. Disruption of the intercellular bilayer lipid structure
2. Interaction with the intracellular proteins of the SC
3. Improvement of partitioning of a drug, coenhancer, or cosolvent into the
SC
Menon et al. (124) extended the LPP model to recognize
4. Disruption of the corneocyte envelope by caustic agents such as 7% phenol
and hydrocarbons
5. Effects on proteic junctions, such as desmosomes, involved in squamae
cohesion
In this chapter we have introduced a potential sixth, but as yet unexplored mecha-
nism, namely,
6. Alteration of the partitioning between SC components and the lipid in the
diffusion pathway
A diagrammatic illustration of the mechanisms of vehicles and their components on
the skin is shown in Figure 20 and Table 7. Polar channels have been suggested to
be formed by the actions of a number of lipophilic enhancers (e.g., terpenes) as an
additional mechanism of action to their effect of causing disruption of intercellular
lipid bilayers (135). Cornwell and Barry (135) determined the conductivity of human
skin in vitro before and after treatment with various enhancers. Significant increases
in the conductivity following treatment suggested that new polar channels were being
opened up in the SC, which is considered the major barrier to ion transport through
human skin (136). Cornwell and Barry (135) were also able to show a correlation
between the observed increases in conductivity following enhancer treatment and the
flux of the polar, nonelectrolyte 5-fluorouracil (5-FU), suggesting the polar channels
created allowed the passage of both ions and 5-FU. It was concluded from these
studies that because differential scanning calorimetry (DSC) had demonstrated that
terpene enhancers disrupt the intercellular lipids in the SC (137), the most likely site
of pathway formation is through the lipid bilayers. The creation of polar channels
through the SC has also been suggested by Francoeur et al. (138) following the
application of oleic acid when, contrary to pH-partitioning theory, piroxicam flux
142 Roberts et al.
Figure 20 Diagrammatic representation of the possible mechanisms of action of skin pen-

etration enhancers. (Adapted from Ref. 123.)
increased with the proportion of ionized drug. Fourier transform infrared (FTIR)
studies by Ongpipattanakul et al. (139) later suggested that polar channels were
formed by the lateral phase separation between oleic acid and indigenous lipids.
Enhancement of the skin permeability of ionized salicylic acid by DMSO (140),
another enhancer shown by DSC to cause intercellular lipid disruption (141), has
also been suggested to occur by the creation of polar channels in the intercellular
lipids (137). The effect of surfactants on skin permeation has been reviewed (109).
As discussed earlier, one mechanism for nonionic surfactants is an increase in av
Skin Transport 143
Table 7 The Effect of Vehicles on Structures of the Stratum Corneum
Solvent/Vehicle Effect
Water Increase in water localization

Occulsive agents and swelling of the
increasing intracellular keratin region
hydration of corneocytes, creating
polar pathways
Caustic solvents Breaking of desmosome

Acids junctions and separation
Phenols of corneocytes loosening
stratum corneum, together
with the disruption of
intracellular keratin
organization
[see Eq. (40)] and a ‘‘push’’ mechanism (see Fig. 8B) (65). Ionic surfactants affect
Dsc with peak effects for a given series of surfactants often occurring at C12 or C14.
Several methods have been used to assess the effects of enhancers on mem-
brane permeability. A low flux or permeability constant through the skin from a
given vehicle does not necessarily imply that enhancement has occurred. The ratio
of fluxes through skin and an inert membrane from a given vehicle is independent
of the activity of solute in the vehicle (121) and defines the ratio of the permeability
coefficients through the skin and the membrane. Table 8 shows the results obtained
for phenol from different vehicles. Also shown are the partition coefficients between
light liquid paraffin and the vehicle. It is apparent that the moderately polar hydro-
gen-bonding vehicles, and known penetration enhancers, of dimethyl formamide and
dimethyl sulfoxide have a high affinity for phenol, as is evident by their low, light-
liquid paraffin–vehicle partition coefficients and low permeability coefficients
through the inert membrane. Their permeability ratios between skin and the inert
membrane show, however, that they have markedly enhanced skin permeability.
144 Roberts et al.
Table 8 Effects of Vehicles on the Penetration of Phenol Through Polyethylene and

Excised Skin (37⬚)
Permeability
constant (N = 5)
(cm min⫺1 ⫻ 103) Partition coefficient: Permeability
light liquid ratio skin:
Vehicle Polyethylene Skin paraffin–vehicle polyethylenea
Light liquid paraffin 0.5 1.0 1.0 2

Water 0.044 0.19 0.12 4
Arachis oil 0.014 0.029 0.034b 2
Glycerol 0.005 0.010 0.015 2
Ethanol 0.004 0.016 0.030 4
Dimethyl formamide 0.002 0.022 0.008 11
Dimethyl sulfoxide 0.001 0.018 0.003 18
Estimated s.d. = ⫾20%.

a
b
Light liquid paraffin–water partition coefficient/arachis oil–water partition coefficient.
Thus, it is apparent in this instance that drug–vehicle interactions strong outweigh

vehicle–skin interactions
An alternative approach to the assessment of vehicle–skin interactions is to
compare the fluxes of solutes from saturated solutions, as these should be identical
unless the vehicle has affected the skin (D asc or Ssc) (see Sec. III.A). However, care
should be taken with this approach as some solutes, such as phenol, denature the
skin in certain vehicles at high concentrations. This is an example of a drug–skin
interaction outweighing a drug–vehicle interaction. Figure 21 shows the relation
between the maximum flux for benzophenone through epidermal and high-density
polyethylene membranes and the solubility parameters of the vehicles used. It is
evident that maximal fluxes occur for the epidermal membrane with ethanol (␦v =
14.9 cal cm⫺3) as a vehicle and for the high-density polyethylene membrane with
isopropyl myristate (␦v = 8 cal cm⫺3) or C12 –C15 benzoate alcohols (␦v = 7.6 cal
cm⫺3) as vehicles. The major effects of the solvents appear to be diffusivity changes,
and it is apparent that some solvents enhance skin permeability and others polyeth-
ylene permeability.
A third method is the assessment of the skin penetration flux for a solute or
penetration enhancer through the skin before and after application of the solute or
enhancer. Hence, the ratio of phenol permeability coefficients at high concentrations
from vehicles can be compared with that from a low concentration before and after
treatment to give a damage ratio and an irreversible damage ratio, respectively (121).
Scheuplein and Ross (142) had previously shown that pretreatment of the skin with
ethanol increased the permeability of pure butanol. Williams and Barry (137) com-
pared the drug permeability before and after terpene treatment to assess the effects
of terpenes on skin permeability.
A fourth method is to use a technique (or techniques), which allows an inde-
pendent assessment of solubility and diffusion effects. Harrison et al. (143) compared
the effects of the enhancers Azone (see Chap. 6) and the solvent Transcutol (diethyl-
eneglycol monoethyl ether) on changes in the diffusivity and solubility of a model
Skin Transport 145
Figure 21 Relation between benzophenone membrane flux and vehicle solubility parameter
(␦v) for (A) epidermis and (B) polyethylene. (From Ref. 330.)
permeant (4-cyanophenol) in human SC using attenuated total reflectance Fourier

transform infrared (ATR-FTIR) spectroscopy. They suggested that Azone acts by
reducing the diffusional resistance of the SC, whereas Transcutol increases the sol-
ubility of the penetrant in the SC barrier (Fig. 22). Zhao and Singh (144) investigated
the mechanism(s) of percutaneous absorption enhancement of propranolol hydro-
chloride across porcine epidermis by terpenes (e.g., menthone and limonene) in com-
bination with ethanol using both Franz diffusion cells and by determining the par-
titioning of propranolol hydrochloride into powdered SC from control and enhancer
146 Roberts et al.
Figure 22 The effect of transcutol pretreatment on the in vitro diffusion of cyanophenol

(CN) through the stratum corneum. (From Ref. 331.)
solutions. In a more recent review, Bach and Lippold (145) highlighted the quanti-
fication of enhancing effects on drug penetration that is possible by an indirect de-
termination through the measurement of pharmacodynamic response. They suggested
that the enhancing effects may be determined from the activity–response lines ob-
tained with and without enhancer, respectively.
A range of methods have been used to define the mechanisms by which vehicles
affect skin permeability. These include fluxes, partition coefficients, various spectro-
scopic techniques (146), and differential-scanning calorimetry. Lee and Tojo (147)
used differential-scanning calorimetry to show that the skin-enhancing effect of vi-
tamin C is through its effects on skin hydration and a ‘‘solubilizing action on the
protein domain of the SC.’’ Enhancement may also be by lipid extraction (123).
Goates and Knutson (148), in examining the influence of alcohol chain length on
mannitol permeation in human skin, used FTIR spectroscopy to show that SC lipid
conformation and mobility was unaffected, but that there was evidence of a lipid
extraction altered SC protein conformation.
E. Drug–Skin Interactions
To date, drug–skin interactions have been examined mainly in terms of the solubility
of a drug in the SC and the diffusivity of the drug in SC lipids. Drug–skin inter-
actions are also of interest from two other viewpoints: substantivity and corrosivity.
Skin Transport 147
Substantivity is a measure of the binding of solutes to binding sites in the SC

in terms of showing a resistance to being washed off or removed. This requirement
is particularly desirable for certain cosmetics, such as sunscreens. Hagedorn–Leweke
and Lippold (102) reported that the affinity of ten nonionic compounds, including
sunscreens, antioxidants, antimicrobial compounds, and a repellent to animal keratin
and human callus, was linear with concentration and that the keratin affinity was
directly related to their octanol–vehicle partition coefficients. They suggested that
genuine substantivity, associated with specific adsorption, therefore, does not seem
to occur for these solutes. As discussed earlier, saturable nonlinear binding has been
reported for hair dyes, which have an intended affinity for keratin. Often there is no
apparent relation between skin permeation and SC–water partition coefficients, un-
less the binding sites are saturated (125,149). Dressler (150) has recently reviewed
the percutaneous absorption of hair dyes. Triclosan (2,4,4⬘-trichloro-2⬘-hydroxydi-
phenyl ether), a nonionic, broad-spectrum, antimicrobial agent present in many per-
sonal care products (deodorant soaps, underarm deodorants, shower gels, and health
care personnel handwashes) shows a moderate degree of substantivity to the skin,
leading to a remnant antimicrobial effect in many products (151). Early studies on
skin binding have been summarized (126).
Substantivity is also important in skin toxicology. Islam et al. (152) mapped
the SC substantivity of chloroform in terms of exposure time and depth of penetration
into the SC. Eight minutes was required for the steady-state gradient to be estab-
lished, and substantivity was affected by evaporation. Attempts have also been made
to determine the adsorption of surfactants by the human horny layers in vivo (153).
Often the adsorption process onto keratin may take some time. For instance,
omoconazole nitrate, a topical antifungal agent, required 10 or more days to reach
equilibrium in the skin (154). Nickel and cobalt are also highly adsorbed to human
SC (155). Tape-stripping may be an appropriate method to study substantivity. This
method has been used to show that the amounts of lindane that were recovered in
tape-strippings taken at 6 h (representative of SC content) were substantially greater
than in the remainder of the skin, for both an acetone solution and a formulation
(156). Desquamation rates may also be important and, for the scalp, ranges from 8
days under normal circumstances to within 3–4 days in pityriasis and dandruff con-
ditions not associated with erythema (150).
Several solutes can affect skin permeability, and many of these effects are most
evident for pure solutes. The mechanisms by which many of these solutes affect the
skin are similar to those outlined in the previous section. Of particular interest,
however, has the the corrosivity of solutes, as these have obvious occupational health
implications. In our early work, we hypothesized that there was a threshold molar
concentration at which phenols altered skin permeability through a caustic effect
(24). As a consequence, certain phenols (e.g., phenol and the cresols) had a sufficient
skin solubility to be damaging, whereas the more nonpolar phenols did not, owing
to solubility limitations (Fig. 23A). Scheuplein and Blank (157) also showed that the
greatest extent of irreversible damage with a series of pure alcohols occurred with
methanol (see Fig. 23B). The use of pure liquid solutions may involve a solubilizing
component. Barry et al. (57,158) found that, at comparable thermodynamic activities,
liquid fluxes were often tenfold higher than vapor fluxes for model penetrants (benzyl
alcohol, benzaldehyde, aniline, anisole, and 2-phenylethanol) applied in model ve-
hicles (butanol, butyl acetate, isophorone, isopropyl myristate, propylene carbonate,
148 Roberts et al.
Figure 23 (A) Aqueous solubilities (●) and aqueous threshold concentrations for damage
(䡩) for various phenols and their octanol–water partition coefficients (Koct); (B) permeation
rates (Js) of the alcohols as pure solvents (●) and as solutions (䡩) through the epidermis. The
curves cross each other near a value of log Js = 1.6, units for Js = ␮mol cm⫺2 h⫺1. (From
Refs. 101[A]; 157[B].)
Skin Transport 149
toluene, n-heptane, and water). They suggested that the differences were reflected
by the partition coefficients and the amount of penetrant entering the SC membrane.
They also suggested that whereas liquid fluxes were membrane-controlled, an inter-
facial effect may account for a low vapor permeation.
Various authors have shown that skin irritation can be related to the solute’s
pKa. Berner et al. (159) showed that, for a homologous series of benzoic acid deriv-
atives, which permeate through human skin at comparable rates, skin irritation and
pKa were correlated for pKas less than or equal to 4. For basic permeants, skin
irritation in vivo increases with increasing pKa (160).
Quantitative structure–activity relations (QSARs) have been derived relating
skin corrosivity data of organic acids, bases, and phenols to their log (octanol–water
partition coefficient), molecular volume, melting point, and pKa (161,162). It is ap-
parent that these relations reflect permeability limitations, such as those defined by
octanol–water partitioning, molecular volume, and melting point (discussed earlier
in Sec. III.C.) flux, together with intrinsic acidity of the solutes as defined by pKa.
F. Drug–Vehicle–Skin Interactions
Barry (126) has considered several models concerned with percutaneous absorption
from binary solutions.
1. The ideal case when neither SC solubility nor diffusivity was affected by
either the vehicle or the solute
2. When a vehicle or drug leads to an increase in solute solubility in the SC
(the ‘‘pull’’ effect, see Fig. 8B)
3. When an added cosolvent reduced the partitioning of a solute into the skin
4. When alterations occur in the diffusivity of the SC
First, we consider the ideal case when neither SC solubility nor diffusivity was
affected by either the vehicle or the solute. Maximum flux in this instance is, as
described earlier, at the maximum solubility of the solute in a given vehicle, provided
release from a given vehicle is not diffusion-limited. A greater effect may be achieved
by adding other solutes with the same action to the solvent system. A combination
of three corticosteroids exhibiting independent solubility, partitioning, and diffusion
behavior resulted in a higher total steroid concentration in solution than was possible
for any steroid alone, with evidence of greater in vivo human vasoconstriction than
observed for the individual steroids (163).
In the second case (126) a vehicle or drug leads to an increase in solute sol-
ubility in the SC, the ‘‘pull’’ effect described earlier in Figure 8B. For instance, Kadir
et al. (65) reported that addition of paraffin oil to a propionic acid solution increased
the flux of either theophylline or adenosine through enhancing the flux of propionic
acid into the skin, and promoted the partitioning of the purine solutes in the modified
skin barrier (‘‘pull’’ effect). Similar effects can be achieved for adenosine using
binary vehicles of hexanoic acid and propionic acid or isopropyl myristate and pro-
pionic acid (64) and for theophylline (63). Harrison et al. (143) have shown that
transcutol enhances cyanophenol’s solubility in SC lipids. From binding studies, it
was suggested that the enhancement in the permeability coefficient of tamoxifen by
5% w/v menthol and thymol in a 50% ethanol solution was, at least partly due, to
improvement in the partitioning of the drug to the SC (164). Indeed, menthol also
150 Roberts et al.
enhanced the skin permeation of testosterone eightfold by forming eutectic mixtures

with testosterone, cholesteryl oleate, and ceramides (165), thereby, increasing the
solubility of testosterone in the SC lipids. A further 2.8-fold increase in the flux of
testosterone resulted from a corresponding increase in the solubility of testosterone
in an aqueous ethanol vehicle. The enhancing effects of 1-methyl-, 1-hexyl-, and 1-
lauryl-2-pyrrolidone on the transdermal penetration of 5-fluorouracil, triamcinolone
acetonide, indomethacin, and flurbiprofen were also suggested to be by increasing
the solubility of penetrants in the SC (166).
It appears that supersaturation of a solute in a vehicle is accompanied by su-
persaturation in the SC (113). The percutaneous absorption of nifedipine was greatly
enhanced from binary solvent systems of acetone and propylene glycol or isopropyl
myristate, relative to either saturated nifedipine solution in propylene glycol or iso-
propyl myristate alone (167). Given the effectiveness of a polymer additive, it is
possible that this system leads to enhanced absorption by facilitating supersaturated
solutions, a concept discussed in greater depth in Chapter 6.
It is possible that an added agent to a vehicle will partition into the skin and
produce a polarity of the skin that will result in a reduced partitioning of a solute
into the skin and a reduced flux. We are not aware of any specific examples of this
effect.
Finally, there are various agents, that can either promote or retard skin pene-
tration. Agents promoting diffusivity were discussed in the earlier section on vehi-
cle–skin interactions. There are also several penetration retarders, which have been
identified, including substances such as the Azone analogue N-0915, for which the
mode of action is suggested to be by increasing the order of SC lipids (168). Binary
cosolvents consisting of isopropyl myristate and short-chain alkanols, such as ethanol
(EtOH), isopropanol (iPrOH), and tertiary butanol (tBtOH), in particular a 2:8 com-
bination, produced a marked synergistic enhancement of BZ flux from the mesylate
salt, whereas a retarding effect was noticed for permeation of the benztropine base
(169). Kim et al. (170) reported that S,S-dimethyl-N-(benzenesulfonyl) iminosulfu-
rane; S,S-dimethyl-N-(2-methoxycarbonylbenzene-sulfonyl) iminosulfurane; and S,S-
dimethyl-N-(4-chlorobenzenesulfonyl) iminosulfurane significantly decreased the
permeation of hydrocortisone through hairless mouse skin and may be acting as
retardants.
G. Non–Steady-State Solutions
The mathematics for the non–steady-state condition is more complex owing to the
need to solve the second-order diffusion equation with various boundary conditions,
reflecting the system used (e.g., finite dose, infinite dose, viable epidermal, or clear-
ance limitations, and so on). Many of the solutions for various conditions are pre-
sented in our recent work (83) (see also Chap. 3). Even more complex solutions
arise when the diffusional processes in each phase are considered simultaneously. In
practical, conceptual terms, the most important consequence of the diffusion process
is to impose a lag time on the appearance of a solute at one edge of the membrane
after application at the other edge. Figure 24A shows typical profiles for the uptake
into, accumulation in, and transport of a solute through a membrane after application
of a constant concentration. The Laplace and analytical solutions are presented in
our earlier work (83) and are not reproduced here as they require appropriate non-
Skin Transport 151
Figure 24 (A) Sorption and permeation curves for a simple membrane showing the total
quantities of solute entering the membrane (Qin), exiting (Qout), and accumulating within it
(Qmem); (B) a compartmental representation of the SC as suggested by Zatz. (From Refs. 22[A];
171[B].)
linear regression techniques to fit relevant data. In contrast, the steady-state expres-
sions are straightforward and can be solved by linear regression, noting the concerns
raised in the previous chapter. The expressions for the steady-state portions of the
profiles for the amount entering the membrane Q ssin(t), amount leaving the membrane
Q ssout (t), and the amount remaining in the membrane Q ssmem(t) at different times from
a constant donor concentration are
Q ssin(t) = k pssCv (t ⫹ 2 lag) (63)
Q ssout (t) = k pscCv (t ⫺ lag) (64)
152 Roberts et al.
Ksc⫺v Cv hsc
Q ssmem(t) = (65)
2
where lag = h2sc /6Dsc [see Eq. (27); Chap. 3]. More complex expressions for lag
times, taking into account the resistance of deeper layers and clearance limitations,
have been given (90). An alternative approach is to use a compartmental model to
represent skin penetration kinetics. Many of these models have been recently re-
viewed (83). Zatz (171) presented a compartmental representation of the SC, five
compartments being the diffusion path, and another five being binding sites in the
diffusion path. They suggested that binding affected lag time, but not steady-state
flux.
In the next section, we interrelate the principles of skin transport to pharma-
cokinetic considerations. The important variable for this purpose is the absorption
rate or flux of solutes, defined as J = dQ/dt.
V. SKIN PHARMACOKINETICS
Pharmacokinetics is the time course of drugs in the body or in individual tissues
after input into the body. Relevant to transport of drugs through the skin, solutes are
normally applied to the skin for local or for systemic effects. The desirable require-
ments for the two effects are different. Systemic effects are usually best achieved by
the skin providing minimal resistance, binding, and local metabolism of solutes. In
contrast, local effects are best achieved by relatively high cutaneous concentrations,
with desirably minimal spillover to the systemic circulation so that the body load of
the drug is low or barely detectable. Therefore, we will consider both systemic and
cutaneous pharmacokinetics in this analysis. Given that a number of mathematical
models used to describe various aspects of percutaneous penetration in terms of the
underlying physical processes and representation of those processes by diffusion and
compartmental models have recently been reviewed (83), our emphasis will be placed
on approximate forms useful for interpreting the effects of vehicles and solute struc-
ture on systemic and cutaneous pharmacokinetics after topical application. Our earlier
work reported model solutions and showed cumulative amount and flux time profiles
for a range of boundary conditions and situations, as well as considering topics such
as physiological pharmacokinetic models, pharmacodynamics, deconvolution, and
methods of pharmacokinetic analysis. To minimize confusion associated with the
differing notation used in various published papers, our own earlier work, and the
previous chapter, we have adopted a convention of representing J as flux per unit
area, and the permeability coefficient as k ip . We first consider quasi–steady-state
solutions for systemic and cutaneous pharmacokinetics with a constant flux from a
vehicle and with depletion of solute in the vehicle. We then consider flux in terms
of its determinants so that the role of vehicle and solute structure on pharmacoki-
netics can be related to the physicochemical properties of the solute and the vehicle.
Finally, we consider both biological and physicochemical factors reported to affect
the determinants of flux and the resultant pharmacokinetics.
A. Pharmacokinetic Principles
The concentration C(t) of a solute at time t at any site is defined by the input flux
to the site (J(t)) and the transfer function (tr(t)) for this site:
Skin Transport 153
C(t) = J(t)*tr(t) (66)

where * is the symbol for convolution and tr(t) is the convention–time profile at the
site following the clearance of a unit impulse input to that site. In principle, Eq. (66)
can be applied to define plasma concentrations (Cp ), concentrations at sites of phar-
macological action, and solute concentrations in different parts of the skin. Hence,
concentrations in the viable epidermis (Cve ) are given by the flux of the solute to the
viable epidermis (Jve) and the transfer function for the viable epidermis (tr(t)ve) into
the dermis, deeper tissues, and skin blood circulation:
Cve (t) = Jve (t)*tr(t)ve (67)
Care needs to be exercised in attempting to extrapolate in vitro data into the
in vivo situation using Eqs. (66) and (67). These equations assume that the processes
determining concentration are independent of each other. The steady-state solute flux
through the SC per unit area of application (J sssc ), for instance, is determined by the
concentrations of solute immediately below the outside (Csc(o)) and inside (Csc(i)) of
the SC [see Eq. (28)]. As shown in Figure 7, a significant viable epidermal resistance
or poor perfusion will increase Csc(l ) and lead to a reduction in J sssc , the flux through
the SC. In these circumstances, appropriate models of skin flux defined by such
processes should be used. It may be most appropriate to recognize such processes,
for instance, by using a non–steady-state solution (90) corresponding to Eq. (43),
an expression for J ssskin , that recognizes its component permeability constants.
Systemic plasma concentrations of a solute after topical application (C topical p )
are then appropriately defined by the convolution of the skin flux–time profile for
the solute through the skin to the systemic circulation (J circskin (t)) and the plasma con-
centration–time profile after intravenous administration of a unit dose (C ivp (t)) (i.e.,
the transfer function for the whole body):
C topical
p
circ
(t) = J skin (t)*C piv(t) (68)
B. Systemic Pharmacokinetics
When the solute concentration in the vehicle (Cv ) can be assumed to be constant
(i.e., no depletion in the vehicle at the skin surface with time), the flux for the skin
system per unit area (Jskin(t)) becomes constant (or at steady-state) after a lag time
lag (see Fig. 24a). Jskin(t) can be expressed in terms of an effective permeability
constant for the system (k⬙),
p defined by Eq. (43) when t > lag.
Roberts and Walters (3) have pointed out that if there is significant skin me-
tabolism or irreversible adsorption, the in vivo flux will be reduced, as defined by
the cutaneous availability F defined by the ratio of topical and systemic areas under
the plasma concentration–time curves adjusted for dose differences. Hence, Eq. (43)
can also be rewritten as
Jskin = Fk⬙C
p v t > lag (69)
Anissimov and Roberts (90) have commented on the limitations associated with Eq.
(69). In most studies, an ‘‘instantaneous’’ transfer of a solute from the dermal blood
supply to the systemic circulation is assumed so that I (t) can be assumed to be a
value of unity. A range of models have been used to describe plasma concentrations
after intravenous administration and include a single exponential (also referred to as
154 Roberts et al.
one-compartmental), biexponential (also referred to as two-compartmental), and

physiological models (Fig. 25A). The plasma concentration–time profile after intra-
venous administration (C ivp ) for the simplest single-exponential model is given by:
doseiv
C piv = exp(⫺kel t) (70)
Vbody
In usual pharmacokinetic terms, a constant flux is the equivalent of a constant
rate of administration, such as an infusion rate. Applying Eqs. (69) and (70),
C topical
p for the topical absorption of a solute with a constant flux (Jskin) of a product
applied over an area (A) for a period of time (T ), after which it is removed, can be
再
described by Eq. (71) (3,83):
0
t < lag
Jskin A
(1 ⫺ exp[⫺kel (t ⫺ lag)]) t < lag ⫹ T
Cp = Clbody (71)
Jskin A t ⱖ lag ⫹ T
(1 ⫺ exp[⫺kel T])exp[⫺kel (t ⫺ T ⫺ lag)]
Clbody
Figure 26 shows the extent to which this quasi–steady-state absorption model plasma
concentration–time profile after topical application corresponds to an absorption
model, defined by the diffusion equation, only when the lag time is small. The
greatest deviation between the quasi–steady-state model and the diffusion-predicted
model occurs at the termination of the topical application. Hence, although Eq. (71)
is most appropriate for iontophoretic delivery with its very small lag times (172), it
should be used cautiously in describing passive percutaneous absorption kinetics.
Under steady-state conditions Eq. (71) reduces to Eq. (72) (3,84):
Jskin A
C ssp = (72)
Clbody
Hadgraft and Wolff (173) used Eq. (72) to show that it was possible to predict
nitroglycerin plasma levels from in vitro patch-release data. Equation (72) can be
rearranged to estimate the flux rate that is desirable for a transdermal delivery system
to achieve a desired steady-state plasma concentration (174):
F ⫻ dosing rate = Jskin A = C ssp Clbody
= target blood concentration ⫻ clearance (73)
Hence, for clonidine with a clearance of 3.1 mL/min kg⫺1 and a target blood
concentration of 0.5 ng/mL, a dosing rate of 0.156 mg/day is needed for a 70-kg
person (174). Table 9 gives some examples for other drugs of interest.
Usually, plasma concentration–time profiles of solutes applied to skin are also
modified by the significant depletion of solute in the topical product with time. In
the simplest case, referred to by Riegelman (175), topical absorption through the
skin is assumed to be first-order, with a rate constant ka . Plasma concentrations are
then described by the first part of Eq. (71) up until the product is removed at time
T, if at all. The plasma concentrations after product removal are then described by
the second part of Eq. (71).
Skin Transport 155
Figure 25 (A) Pharmacokinetic compartment and physiological models of the body; (B)
diagrammatic representation of the pharmacokinetic processes involved in cutaneous per-
meation.
156 Roberts et al.
Figure 26 Plasma concentration–time profile after topical patch application and termina-
tion at 10 h. Solid line represents modeling to concentration with the diffusion model, and
the dashed line shows approximation by the compartmental approach [see Eq. (71)]: (A) a
lag time of 1 h; (B) a lag of 0.2 h. The upper curves of both (A) and (B) involve a half-life
of 2 h and the lower curves a half-life of 20 h.
再
0 t < lag
ka Fdose
(exp[⫺ka (t ⫺ lag)] ⫺ exp[⫺kel (t ⫺ lag)]) lag < t < lag ⫹ T
Cp = Vbody (kel ⫺ ka)
ka Fdose
(exp[⫺ka T ] ⫺ exp[⫺kel T ])exp[⫺kel (t ⫺ T ⫺ lag)] t ⱖ lag ⫹ T
Vbody (kel ⫺ ka)
(74)
A simplified form of Eq. (74) arises over very long times when absorption is much
slower than elimination (i.e., ka << kel ), as occurs when slow-releasing patches are
applied to the skin. Hence, during the application phase, the plasma concentration is
given by:
Skin Transport
Table 9 Effective Plasma Concentrations, Epidermal Permeability Coefficients, Clearances, and Physicochemical Data Used to Predict Required
Solute Transdermal Flux (Js.a) [from Eq. (16)] for Passive Topical Delivery Systems
Estimated Js.a
Plasma level required t1/2 MP
Solute (␮g/L) Clbody (L/h) (␮g/h) (h) MW (⬚C) log Koct
Clonidine 0.2–2.0 13 2.6–26 6–20 230 140 1.77

Estradiol 0.04–0.06 615–790 24.6–47.4 0.05 272 176 2.69
Fentanyl 1 27–75 27–75 3–12 337 83 4.37
Isosorbide dinitrate 22 1.22 26.9 105 236 68 1.31
Nicotine 10–30 77.7 77.7–2231 2 162 ⬃80 1.17
Nitroglycerin 1.2–11 13.5 16.2–148.5 0.04 227 13.5 1.62
Scopolamine 0.04 67.2 2.69 2.9 303 59 1.23
Testosterone 10–100 20.8 208–2080 288 153 3.31
Timolol 15 30.7 460.5 4.1 316 72 2.46
Triprolidine 5–15 43.7 218.5–655.5 2–6 278 60 3.92
157
158 Roberts et al.
Cp =
再 ka Fdose
Vbody (kel ⫺ ka)
0
(exp[⫺ka (t ⫺ lag)])
t < lag
t ⱖ lag
(75)
Hence, the logarithm of the plasma concentration should be linearly related to the
time after application over very long times. The urinary excretion rate is an alter-
native pharmacokinetic representation amount of drug in the body with time. Beckett
et al. (176) showed that time to peak urinary excretion rate was longer and the actual
peak height lower for transdermal applications than for the oral route. A comparison
of the loglinear profiles for the oral and transdermal products shows that the terminal
phase for the topical product has a half-life of 8.4 h whereas that for the oral dose
is 3.3 h (175) (Fig. 27A). This behavior is consistent with the terminal pharmaco-
kinetic phase being controlled by the rate of topical absorption (i.e., ka << kel ). Such
a relation may also be expected with a 7-day patch of timolol, which has an elimi-
nation half-life of 4.1 h (3). Consistent with this prediction, McCrea et al. (177)
reported an apparent linear decline in plasma timolol concentrations during the ap-
plication of a patch designed to release 50% of the timolol over 7 days.
The interpretation of ka in terms of the underlying physicochemical properties
of the system is not straightforward. In principle, if the vehicle is assumed to be
homogeneous, ka may be equivalent to the permeability constant k⬙p divided by the
thickness of the vehicle film on the skin. However, this approximation applies only
when the lag time is very small relative to 1/ka (i.e., ka lag << 1). In other cases, a
quasi–steady-state assumption is inappropriate, with deviations most likely to be
seen on cessation of dosing, as shown for constant input (see Fig. 27). Modeling of
the plasma concentration–time profiles using model representation of the processes
is more appropriate in these circumstances. Solution of expressions for finite vehicle
applications show that, in general, an exponential absorption process exists under
pseudo–steady-state conditions (i.e., t > lag) when a diffusion model is used to
describe transport across the SC (82), the exponent (equivalent to ka) being a complex
function of vehicle thickness, SC–vehicle partition coefficient, SC diffusion time,
and other variables. An alternative approach is to use a compartmental representation
of the skin (116,178–181), a multiple compartmental representation of the body, on
a combination thereof. The pharmacokinetics of hyoscine (scopolamine) in the body
after topical application has been described in terms of urinary excretion rate and a
two-compartmental representation of the body (81). It is apparent in Figure 27B that
the urinary excretion–time profile has four phases: (a) an absorption phase to peak,
(b) the absorption nose associated with two-compartment kinetics, (c) a loglinear
decline while the topical application continues, consistent with Eq. (75) and (d) a
loglinear decline due to hyoscine elimination kinetics from the body as defined by
Eq. (71) for t > application time ⫹ various lags. Plasma levels after multiple dosing
have also been described with such a model (116). Berner (182) has described the
pharmacokinetics of drug delivery from transdermal controlled-release devices con-
sisting of a membrane plus a reservoir or a monolithic slab.
When the pharmacokinetics associated with skin absorption are uncertain, it
may be more appropriate to use some form of deconvolution. One of the earliest
forms used in pharmacokinetics is the Wagner–Nelson model based on the assump-
tion that the disposition of a solute after intravenous administration can be described
by a single exponential. Birmingham et al. (183) used such a model to suggest that
Skin Transport 159
Figure 27 (A) Excretion rate following oral (▫) and topical (䡩) application of norphedrine
HCl. Subtraction from the extrapolation line represents a process involved in drug absorption
and excretion (●); (B) excretion rate profile of in vivo transdermal hyoscine in humans show-
ing a comparison of theory (solid line) and in vivo data (points) after a single application.
(From Refs. 176[A]; 332[A].)
160 Roberts et al.
salicylic acid has an apparent first-order absorption rate when applied topically to
rabbits in vivo.
The clinical effect (pharmacodynamic response) associated with topical deliv-
ery is commonly examined in terms of the subsequent relations between pharma-
codynamic response and plasma concentrations of solutes. The Emax pharmacody-
namic model, defined by Eq. (76) (184) has, for instance, been used to relate
postexercise heart rate and plasma timolol concentrations Cp achieved after weekly
application of a timolol patch (177).
Emax ⫺ Cp
E = Eo ⫺ (76)
IC50 ⫹ Cp
where E is the suppression of the postexercise heart rate, Eo is the baseline exercise
heart rate before patch application, Emax is the maximal suppression for timolol, and
IC50 is the Cp corresponding to 50% of Emax. Dermatological corticosteroid products
are usually assessed for their clinical potency by skin blanching, as defined by a
vasoconstrictor assay, and is related to the dose absorbed, as defined by dose dura-
tion. Singh et al. (185) have recently attempted to validate the vasoconstrictor assay
dose–response relation using an area under the effect curve for each dose duration
and the Emax model [see Eq. (76)]. They found that model fits to all individual subject
dose–response data were unacceptable for all dermatological corticosteroid products
tested, but that population dose–responses were adequately described by the Emax
model.
C. Cutaneous Pharmacokinetics
In cutaneous pharmacokinetics, the goal is to target a given region of the skin as
shown in Figure 25B. The first approach has been to estimate skin target concentra-
tions based on a knowledge of the flux to the site and clearance from the site.
Siddiqui et al. (186) first applied this approach using methotrexate. Later studies
applied the approach with steroids (44) and phenols (101). An extension of this
approach is the use of skin target-site free-drug concentration (C*) estimated from
in vitro flux data to predict topical in vivo efficacy. Such a model has been applied
into examining critical factors that influence topical bioavailability and bioequiva-
lence (187) and in interpreting the activity of solutes such as acyclovir (ACV) in the
treatment of cutaneous herpes simplex virus-type 1 (HSV-1) infections (188). In work
to date, the target site has presumed to be the basal cell layer of the epidermis. Patel
et al. (189) have applied the C* concept in predicting the topical antiviral efficacies
of ACV formulations for the treatment of cutaneous HSV-1 infections, using a hair-
less mouse model. They found that, over a wide range of efficacies, the predictions
based on C* (estimated from the experimental in vitro fluxes) were in good agree-
ment with in vivo antiviral efficacies measured at the end of a 4-day–treatment
protocol. The physical model involved validating a ‘‘three-tiered’’ model for finite-
dose drug uptake and transport in skin with experimentally determined input param-
eters (partition coefficient K, and steady-state permeability coefficients P, for the SC,
viable epidermis, and dermis) (189).
Values of the steady-state unbound (or ‘‘free’’) concentration C* ss at different
sites in the skin (see Fig. 25B) are related to the total concentration at that site Css
by the fraction of solute unbound at that site f* u , and is defined by the input flux to
Skin Transport 161
that site and removal clearance from that site. An expression for the steady-state
concentration at a given site Css, such as the receptor solution in an in vitro situation,
has been defined (83). Hence, substituting C* ss = f*C
u ss into that expression, C* ss is
given by Eq. (77).
f*k⬘C
u p v
C*
ss = f*Css = (77)
(Cl*/A) ⫹ k⬘(K*/K
u
p m)
where k⬘p is the apparent permeability coefficient to the site; C is the concentration
of solute in the vehicle; Cl* is the clearance from the site divided by the area of
application A; k⬘ is the SC–site partition coefficient (K* = Csc /Css); and Km is the
SC–vehicle partition coefficient (Km = Csc /Cv). Siddiqui et al. (44) and Roberts (101)
used in vitro skin permeability coefficients for steroids and phenols together with in
vivo dermal clearances to estimate C* ss . The different estimated concentrations C*
ss
for phenols and steroids shown in Figure 28A, reflect the different magnitudes in
the clearance per unit area (see Fig. 28B), and the apparent permeability coefficient
corrected for partition coefficient effects.
D. Clearance
As shown in Figures 4 and 25B, solutes may either be carried away by the local
blood supply on entering the dermis or transported into deeper tissues by perfusion
or diffusion. Lymphatic transport is significant for the larger molecular weight solutes
(190). Microdialysis studies have shown significant penetration of solutes into deeper
tissues of human subjects after topical application (191,192). Muller et al. (193) have
found that when a diclofenac foam (5%) was administered epicutaneously at the
thigh (80 mg/200 cm2 twice daily for a period of 7 days) of healthy volunteers,
significantly higher skeletal muscle concentrations of diclofenac (219.68 ⫾ 66.36
ng/mL) were found compared with that found in plasma (18.75 ⫾ 4.97 ng/mL).
We have examined the disposition of a series of solutes assuming a compart-
mental representation for each tissue (194) (Fig. 29A). In general, the localized
targeting of solutes to deeper tissues is seen most readily (a) at early times after
application when there are negligible tissue concentrations as a consequence of re-
circulation (see Fig. 29B); (b) when there is a high body clearance for the solute;
and (c) when vasoactive agents are used appropriately (194). The steady-state con-
centration in each tissue C iss is described by
ki⬘p Cv A
C iss = (78)
Cl i ⫹ kpi⬘ A(K i/Km)
The free concentration in each tissue is described by Eq. 77. According to Eq. 78,
C iss can be increased by one of two methods. The most usual is to reduce the clear-
ance from the tissue and higher tissues by the local blood supply using vasoconstric-
tion or removal of solute (Fig. 30A). The second approach is to facilitate transport
to deeper tissues by using the local blood supply (195) (see Fig. 30A). We have
shown higher tissue levels at early times with methyl salicylate, consistent with the
increase in blood flow induced by methyl salicylate in human cutaneous vessels
(196). Figure 30B illustrates the effect of vasoconstriction on the loss of lidocaine
from a dermal cell relative to anesthetized blood flow and no blood flow (197). The
corresponding tissue levels of lidocaine at various depths of application after 24 h
162 Roberts et al.
Figure 28 (A) Epidermal concentration (Css)/vehicle concentration (Cv) ratio for phenols
(䡩) and steroids (●) and their octanol–water partition coefficients (Koct); (B) clearance of
steroids from dermal diffusion cells against their octanol–water partition coefficients (Koct) for
sacrificed (●) and anesthetized (䡩) rats. (From Refs. 101[A]; 44[B].)
Skin Transport 163
Figure 29 (A) Tissue deposition of dermally applied solute assuming compartmental rep-
resentation; (B) salicylic acid concentration in different tissues and in the plasma after the
application of an aqueous salicylic acid solution to the dermis. (From Ref. 194.)
164 Roberts et al.
Figure 30 (A) Diagrammatic representation of the effects of dermal blood flow on the
clearance of topically applied solutes; (B) effect of phenylephrine on the dermal clearance of
lidocaine applied in aqueous solution to rats; (C) effect of phenylephrine on the fraction of
dermally applied lidocaine penetrating into deeper tissues in rats; (D) distribution of hydro-
cortisone in human skin in vitro (䡩) and in vivo (●). (From Refs. 197[B,C]; 127[D].)
Skin Transport 165
Figure 30 Continued
under such conditions are shown in Figure 30C. In Figure 30D, the distribution of
hydrocortisone after application to human skin in vitro and in vivo is shown. An
apparent logarithmic profile is observed in vivo consistent with a constant removal
by the blood supply through the depth of the dermis. In contrast, the in vitro profile
plateaus, as would be expected if there was a clearance limitation (see Fig. 30D).
Protein binding of a solute may also affect tissue levels as shown in Figure 31
(190). Plasma protein binding will facilitate removal into the blood supply (see Fig.
166 Roberts et al.
Figure 31 (A) Diagrammatic representation of the effect of protein binding on the diffusion
and clearance of topically applied solutes; (B) effect of protein binding in perfusate on the
tissue distribution of diclofenac applied dermally to the perfused hindlimb of rats. (From Ref.
198.)
Skin Transport 167
31B), whereas tissue binding will impede binding into lower tissues in a manner
similar to the way binding can slow SC diffusion. We have recently described a
physiological disposition model for solute disposition below a topical site (198). The
model predicts that the half-life for elimination of a solute in an underlying tissue
is related to the volume of plasma in a tissue (Vp) with an extravascular water volume
(VE), the perfusate flow rate (Qp), perfusate protein binding ( fup), and tissue protein
binding ( fuT). An approximate solution applying in most cases is Eq. (79).
0.693VD 0.693 fuP VTE
t0.5el = ⯝ (79)
QP fuT QP
Hence, the retention of solutes in tissues after topical application is dependent
on the relative magnitude of binding in blood ( fup) and tissue binding ( fuT) as well
as on the tissue blood flow Qp. Interestingly, diazepam and diclofenac are highly
bound to underlying topical tissues (198) and, according to Eq. (79) are more likely
to be retained in these tissues than the more nonbound solutes.
E. Skin Metabolism
Metabolism of solutes in the SC and epidermis are also clearance mechanisms, which
can affect skin permeabilities and resultant pharmacological effects. The metabolism
of solutes by skin enzymes has recently been reviewed (4). In relation to skin trans-
port, difficulties may arise from the as yet undefined anatomical distribution of me-
tabolizing enzymes, both in the various layers of the skin and appendages and the
variable activity that may arise from processing skin for permeation studies. For
instance, the activity of some enzymes is reduced by the process of heating used to
separate the dermis from the epidermis (199). Some of the key enzymes involved in
skin metabolism include aryl hydrocarbon hydroxylase, deethylases, hydrolases,
monoxygenases, esterases, peptidases, and aminopeptidases. This skin enzyme activ-
ity can vary among species and may be induced. A major outcome of these enzymatic
activities is the skin first-pass effect whereby a significant proportion of the solute
is metabolized between application to the skin and diffusing to its site of action in
regions of the skin or into the systemic circulation. Nakashima et al. (200) used
intravenous and transdermal ointment administration of nitroglycerin to estimate that
the fraction of nitroglycerin avoiding this first-pass was 0.68–0.76 and was com-
parable with values reported in rhesus monkeys (0.80–0.84). A higher skin first-pass
effect has been reported for methyl salicylate, for which the first-pass availability in
both humans (190) and rats (201) is very low. It has been suggested that the new
retinoid, tazarotene, is superior to those used orally because of its limited percuta-
neous penetration as well as its rapid esterase metabolism in the skin to a more
water-soluble active metabolite tazarotenic acid. The latter has a resultant systemic
absorption of between 1 (normal) and 5% (psoriasis) on repeated applications
(202,203).
Bronaugh et al. (204) have also reviewed some aspects of cutaneous metabo-
lism during in vitro percutaneous absorption. From the perspective of skin transport,
skin metabolism can be adequately modeled only by using a two-phase model. One
of the first studies in this area was that of Ando et al. (205). Higuchi’s group have
since then reported in several papers on the effects of skin metabolism on solute
transport (206–210). In a later study, the influence of low levels of ethanol on the
168 Roberts et al.
simultaneous diffusion and metabolism of ␤-estradiol with several enzyme distri-

bution models was determined (211). The best model was that for which the enzyme
activity resided totally in the epidermis and near the basal layer of the epidermis.
Liu et al. (212) reported there would be less metabolism and that a much smaller
amount of the transdermal metabolite would be taken up by the blood capillary owing
to the shorter dermis pathlength for permeants in vivo than in vitro, when using
dermatomed split-thickness skin.
Recently, Sugibayashi et al. (213) reported the effect of enzyme distribution in
skin on the simultaneous transport and metabolism of ethyl nicotinate in hairless rat
skin after its topical application. Gysler et al. (214) studied the skin penetration and
metabolism of topical glucocorticoids in reconstructed epidermis and in excised hu-
man skin. The influence of enzyme distribution on skin permeation was also studied
(215). Species differences can also be important (216).
Seko et al. (217) used pretreatment with an esterase inhibitor, diisopropyl fluo-
rophosphate, to study the penetration of propyl and butyl paraben across Wistar rat
skin in vitro. A two-layer skin diffusion model predicted an increasing metabolic rate
and decreased the lag time for penetration of both the parent and metabolite.
F. Stratum Corneum Reservoir

Malkinson and Ferguson (218) originally proposed an SC reservoir based on their
studies with radiolabeled corticosteroids. Several other workers (including Vickers,
Barry and Woodford, MacKenzie and Atkinson, Wickrema, Shima, and others) have
provided evidence for the reservoir effect—these studies have been summarized by
Barry (126). In essence, occlusion for some time after topical application of a steroid
promotes absorption of the steroid retained in the horny layer, with vasoconstriction
occurring as a result. Barry (126) suggests that the phenomenon is probably a con-
sequence of the high solubility and normal low diffusivity of the steroids in the SC.
Hence, occlusion and accelerants may enhance the sorption of steroids into the SC.
Because of its low intrinsic diffusivity, the steroid remains relatively trapped in the
SC, once these are removed. On reocclusion or application of an accelerant, diffu-
sivity is promoted and the steroid is absorbed. This same effect may be possible for
certain solutes by altering dermal blood flow. Hence, high, reservoir-like concentra-
tions are observed in various tissues in the presence of a vasoconstrictor (see Fig.
30C): remove the vasoconstrictor and the reservoir effect is lost.
The assessment of SC concentrations in tape-stripped samples is becoming
increasingly advocated in topical bioequivalence assessment (219,220). It can be seen
from Figure 7 that the amount of solute in the SC at steady-state M sssc is given by
Eq. 80.
ss
M sc =
Ksc⫺v hsc
2
冉
Cv ⫺
Kve⫺v
冊
C ssve
⬇
Ksc⫺v Cv hsc
2
Css
ve → 0
(80)
Hence, assuming C ssve → 0 and using Eqs. (63), (64), and (65), results in Eq. (81).
ss a
M sc D sc
J sssc = 2 (81)
2hsc
Hence, the steady-state flux is directly proportional to the amount of solute in the
SC, providing that D asc [see Eq. (38)] is constant. The observed flux for percutaneous
Skin Transport 169
absorption of small solutes, varying 40-fold, was very similar to that predicted from
amounts present in the SC after 30 min of contact (221). This relation also exists
for variable application time, application dose, vehicle, and anatomical site (222).
G. In Vivo–In Vitro Correlations

Franz (223) conducted one of the first validations of in vitro human skin absorption
by correlating the percentage of dose absorbed in vitro and in vivo. Despite the range
of data sources a good correspondance was evident. An important component of this
study was the application of a finite dose of solutes in vitro in mimicking in vivo
results. He highlighted the importance of regional variability in skin absorption and
desquamation, pointing out that, with an average of one cell layer of SC lost per
day, any material that had not diffused beyond that layer in a day in vivo would be
lost.
There appears to be an excellent correlation between in vitro human epidermal
penetration flux and in vivo flux, as deduced by residual drug and pharmacokinetic
methods, when both groups of studies have been conducted by a single group of
investigators (as shown in Fig. 32A for 18 drugs by Shaw’s group). The extent of
correlation between in vitro and in vivo transport when cutaneous metabolism is
involved is not quite so impressive. For instance, esterase activity in vivo results in
an almost complete first-pass metabolism of methyl salicylate in human skin, much
greater than is seen in vitro (224) (see Fig. 32B). Overall, there is an increasing
emphasis on the combined use of in vitro cadaver skin, in vivo animal pharmaco-
kinetics, in vivo human pharmacokinetics, and in vitro–in vivo pharmacology and
microbiology in dermal bioequivalence assessments (225).
H. Species Differences
There is a substantive and often contradictory literature on species differences in
percutaneous absorption. Walters and Roberts (226), who summarized the variation
in lipid content and SC thickness among the various species, reported the results
from a number of studies in which differences in percutaneous penetration among
species have been compared. Figure 33A shows the results obtained for the penetra-
tion of paraquat and water permeability coefficients through the skin from various
species (227).
I. Other Determinants of Percutaneous Absorption

Roberts and Walters (3) have summarized the effects of age, race, temperature, re-
peated applications, disease, and body site on percutaneous absorption. Further de-
tails are given in Chapter 5. In general, temperature increases absorption and ab-
sorption from different sites is in the order scrotum > scalp > forehead > postauricular
> abdomen > arm > palm > plantar. Figure 33B shows the variation of absorption
with site as reported for salicylate excretion after topical methyl salicylate application
(224).
VI. TRANSAPPENDAGEAL TRANSPORT

The role of appendages in skin transport has been controversial and remains so. The
earliest evidence to support the existence of the transfollicular route of transdermal
170 Roberts et al.
Figure 32 (A) Correlation between in vivo and in vitro transdermal drug flux: each da-
tapoint represents a different drug, dashed line indicates perfect correlation between in vivo
and in vitro transdermal drug flux; (B) fraction of total concentration of topically applied
methylsalicylate determined in microdialysate or in vitro diffusion cell studies present as
salicylate following application of a 20% methylsalicylate formulation. (From Refs. 332[A];
119[B].)
Skin Transport 171
Figure 33 (A) Effect of species on the in vitro absorption of water and paraquat through
excised skin; (B) the influence of skin application site on topical bioavailability determined
from cumulative urinary recovery of salicylate following application of 5 g Metsal (methyl-
salicylate)/50 cm2 for 10 h in human volunteers. (From Refs. 227[A]; 224[B].)
172 Roberts et al.
transport, back in the 1940s, were largely qualitative histological studies based on
stain and dye localization within the appendages (Table 10). Table 11 outlines some
of the literature supporting the existence of follicular penetration of topically applied
solutes. Scheuplein (91) suggested that appendageal route dominates transport early
before the lag time is reached for transcellular transport. However, at longer times,
transcellular transport dominates (Fig. 34). Although it remains generally accepted
Table 10 Historical Evidence for the Existence of Transfollicular Penetration Following

Topical Application
Evidence Skin Ref.
Preferential staining of hair follicles following Guinea pig, Mackee et al. (261)
topical application of iron, bismuth, human
sulfonamides, and dyes in a number of
different vehicles
Changes in pharmacological response to Human Shelley and Melton
epinephrine and histamine applied in (311)
propylene glycol observed with changes in
follicular density
Follicular deposition of vitamin A observed by Guinea pig Montagna (232)
quantitative fluorescent microscopy
following application in various solvents
14
C-Labeled pesticide absorption and urinary Human Maibach et al. (236)
excretion increased over follicle-rich areas
such as the scalp and forehead, follicular
route ‘‘possibly’’ contributing
Trichlorocarbanalide compound deposition in Guinea pig Black et al. (312)
follicles and sebaceous glands seen to vary
with vehicle
[3H]hydrocortisone from hydroalcoholic Rat Illel and Schaefer (313)
vehicle penetrates normal skin 50-fold faster
than follicle-free skin. Retention also 20 to
30-fold higher in normal skin
Particle size dependency of follicular Rat Schaefer et al. (314)
penetration, optimum 5 ␮m
Greater concentrations of hydrocortisone and Rat Hueber et al. (229)
testosterone observed in epidermis and
dermis of normal skin, particularly at the
depth of sebaceous glands, compared with
follicle-free skin. In vivo effect less
pronounced than in vitro
Flux and absorption of caffeine, niflumic acid, Rat Illel et al. (315)
and p-aminobenzoate threefold slower in
follicle-free skin
Particle size dependency of follicular Mouse, Rolland et al. (230)
penetration, optimum 5 ␮m. Targeting of human
the antiacne drug adapalene into follicles is
achieved using 5-␮m microspheres as
particulate carriers
Skin Transport 173
Table 11 Summary of Studies Examining Formulation Effects on Transfollicular Transport
Evidence Skin Ref.
Deposition of vitamin A into the follicular Guinea pig Montagna (232)

duct and sebaceous glands was seen
within 10 min of application in ethanol or
chloroform vehicles, compared with much
slower penetration from oleic acid or
petrolatum.
Penetration of a trichlorocarbanalide Guinea pig Rutherford and Black
compound into follicles and sebaceous (316)
glands seen after application in a nonsoap
sodium alkoyl isothionate detergent,
compared with a soapy vehicle, which
resulted in penetration into the stratum
corneum.
Within 2 h of application of [3H]estradiol in Rat Bidmon et al. (317)
DMSO, ethylene glycol, or sesame oil
vehicles, radioactivity could be detected
in the epidermis, follicles, and sebaceous
glands. At 24 h hair follicles and
sebaceous glands still retained high
activity—possible depot function.
Increased deposition of fluorescent beads Rat Schaefer et al. (314)
into follicles following application in
lipoidal vehicles.
Liposomal entrapment of calcein, melanin, Histocultured mouse Li et al. (319,320)
and DNA allowed delivery into follicles skin
compared with control aqueous solutions.
Migration of topically applied steroids Rat Hueber et al. (321)
through the follicular duct or
accumulation in the sebaceous glands
varied with the polarity of the steroid and
the lipophilicity of the vehicle.
Vehicles favoring the transfollicular Rat Bamba and Wepierre (235)
penetration of pyridostigmine included
ethanol, DMSO, and propylene glycol.
50% ethanol, glyceryl dilaurate-based Hamster ears Lieb et al. (322)
nonionic liposomes and egg
phosphatidylcholine-based liposomes all
achieved appreciable deposits of
cimetidine into the pilosebaceous unit,
although possible ion-pairing of
cimetidine to phospholipids reduced its
pharmacological activity in these
formulations.
Nonionic liposome formulations were Hamster ears Niemiec et al. (323)
superior to phospholipid-based
formulations for delivery of interferon-␣
and cyclosporine to follicles. Both
formulations were far superior to aqueous
solutions.
174 Roberts et al.
Figure 34 Suggested domination of transappendageal transport during the earlier stages of

percutaneous penetration. (From Ref. 91.)
that the intercellular route may dominate during the steady-state penetration of com-
pounds, it has been argued that the skin appendages (hair follicles, pilosebaceous
and eccrine glands) may offer an alternative pathway for a diffusing molecule.
A. Hair Follicles and Sweat Ducts

Possible routes of penetration through hair follicles could involve the hair fiber itself,
through the outer root sheath of the hair into the viable cells of the follicle, or through
the air-filled canal and into the sebaceous gland. In addition, the release of sebum
by the sebaceous glands may provide a lipoidal pathway that may influence absorp-
tion by this route (228). The route for the sweat duct may involve diffusion through
either the lumen or walls to below the epidermis and through the thin ring of kera-
Skin Transport 175
tinized cells. Dense capillary networks closely envelop the bases of both the hair
follicles and sweat ducts, providing access to the circulation for most molecules
reaching these regions. Hueber et al. (229) used the observation that a higher res-
ervoir and permeability barrier function in appendage-free (scar) SC than in normal
SC, as supporting evidence for a significant contribution of the appendageal route
to overall skin transport.
There are estimated to be close to 500–1000 pilosebaceous units/square cen-
timeter of skin on areas such as the face and scalp, each with an orifice with a
diameter of 50–100 ␮m and 4 ⫻ 10⫺5 cm2 surface area. These orifices represent
0.1% of the surface area of the skin in low-density areas and up to 10% in high-
density areas, such as those on the face and scalp. The openings lead down to an
epithelial surface which does not have a protective SC, and exists only from the
ostia of the sebaceous gland upward to the skin surface (Fig. 35). These character-
istics have been used to selectively target drugs into the hair follicles and sebaceous
glands. Given that the exposed surface area of appendages is much higher than that
of the openings used in earlier evaluations and that the current intercellular transport
also has a restricted area for transport, the role of the follicle as a pathway for
transdermal delivery is being reconsidered. In the hair follicle, for example, the outer
Figure 35 Diagrammatic representation of the postulated effect of molecular size on the

penetration of solutes into hair follicles. (From Ref. 231.)
176 Roberts et al.
root sheath (see Fig. 35) is thought to be of greatest importance for drug delivery,
as this layer is continuous with the epidermis and is indistinguishable from it, which
potentially allows for increased surface area for absorption beneath the skin’s surface.
In addition, there is increasing demand for localized drug delivery to the hair follicle
itself, particularly for the treatment of dermatological disorders such as acne, alo-
pecia, areata, and androgenetic hair loss.
One of the most important determinants of targeted follicular transport is the
particle size of applied materials. By using fluorescent microspheres, Rolland et al.
(230) showed that the degree of penetration into the human hair follicle was inversely
related to particle size. The optimum size at which microspheres selectively entered
follicles was 3–10 ␮m; below this size, particles were also seen to be distributed in
the superficial layers of the SC (see Fig. 35). The depth of penetration into the follicle
was also determined by size, with beads of 5–7 ␮m reaching the deeper parts of the
upper follicle, though rarely penetrating the superficial SC, and those between 9 and
10 ␮m were observed to concentrate only around the opening of the follicles, but
not inside them. No beads larger than 1 ␮m were observed to penetrate as deep as
the hair bulb of the follicles.
Recent studies using fluorescence-labeled oligonucleotides and dextrans applied
to fresh human scalp skin confirmed earlier findings that follicular penetration was
determined by size and also charge (231). These studies also identified that the
primary anatomical structures for the pathway(s) of intrafollicular delivery of these
molecules were along the junction of the inner root sheath and outer root sheath (see
Fig. 35). Although this pattern of distribution was particularly evident with oligomers
formulated with cationic lipids, the molecular features that allow a selected agent to
move into and through this region await definition. In the same study, it was also
noticed that rhodamine-labeled dextran (3000 MW) applied in a hydroalcoholic for-
mulation (40% ethanol) was present in the center of the hair shaft as well as within
the follicle. It was speculated that this region of the hair shaft may be more amor-
phous relative to keratin content compared with the rest of the hair; therefore, it may
be more permeable to certain agents, although whether entry occurred by diffusion
across the hair shaft or down the cut end of the hair was unclear.
The concept that vehicle and formulation significantly influence the rate of
drug localization within hair follicles, following the application of vitamin A in
various vehicles to guinea pigs back, was noticed in 1954 by Montagna (232). Re-
views covering formulation effects, in particular the use of liposomes, to optimize
transfollicular delivery can be found (233,234). Some of the literature in this area
pointing to the favoritism for lipophilic vehicles in follicular targeting is outlined in
Table 11.
It can be seen from Table 11 that alcoholic vehicles are among those tending
to increase transfollicular penetration. Bamba and Wepierre (235) speculated that, as
ethanol is primarily a lipid solvent, as well as increasing the fluidity of lipid areas
within the SC and extending the hydrophobic domain between polar head groups, it
was also acting on the sebum within the follicles and allowing the more rapid mi-
gration of solute in the sebaceous glands, thereby making the transfollicular pathway
predominant in the initial stages of absorption. DMSO is thought to act on normal
SC by creating a ‘‘solvent pathway’’ through the skin or fluidizing the lipids (140).
However, in the studies of Bamba and Wepierre (235) the concentration was too
weak to have this enhancing effect through the whole epidermis, and it was suggested
Skin Transport 177
that its solvent properties would favor pilosebaceous migration by incorporating the
drug in the sebum.
The contribution of transappendageal transport to systemic clearance, rather
than local deposition, was considered by Maibach et al. (236) following the topical
application of radiolabeled pesticides to human volunteers. A greater urinary recovery
was noted after application to follicle-rich areas, such as the forehead and scalp, than
after application to less hairy areas such as the forearm. The authors concluded that
transfollicular transport could not be ruled out as a contributing factor to the observed
differences. However, studies examining the effect of increasing hair folicle density
on percutaneous absorption (237) failed to show any correlation with the amount of
solute absorbed, suggesting that the follicles’ overall contribution to transdermal
delivery is negligible.
In the recent review by Lauer (238), it was concluded that the contribution of
the pilosebaceous unit to localized and percutaneous absorption may have been un-
derestimated in the past and that a more detailed understanding of formulation fac-
tors, such as drug and vehicle physicochemical properties and particle size, may
allow optimization of follicular delivery. The potential clinical significance of the
ability to selectively deliver drugs to follicles for the treatment of associated der-
matological disorders warrants the pursuit of this area of transdermal research.
B. Permeation of the Nail Plate

The nature of the barrier to topical drug delivery of the nail plate as an extension
of the skin has been given some consideration, especially relative to the potential
for local treatment of fungal infections. The nail plate itself is produced mainly by
differentiation of cells in the nail matrix, and it comprises three horizontal layers: a
thin dorsal lamina, the thicker intermediate lamina, and a ventral layer from the nail
bed (Fig. 36). It contains significant amounts of phospholipid, mainly in the dorsal
and intermediate layers, which contribute to its flexibility, and it has essentially the
same keratin analysis as hair. The keratins in hair and nail are classified as ‘‘hard’’
trichocyte keratins, unlike the ‘‘soft’’ epithelial keratins found throughout the skin.
Figure 36 Diagrammatic representation of the structures of the nail plate.

178 Roberts et al.
Studies by Walters et al. (239) indicated that there is a marked difference

between the permeability characteristics of the nail plate and epidermis. These ob-
served differences have been largely attributed to the relative amounts of lipid and
protein regulation within the structures and the possible differences in the physico-
chemical nature of the respective phases. The SC contains nearly 10% lipid, pre-
dominantly intercellular, whereas lipid levels in the nail plate are near 1%, which
combined with lower water levels of about only 10% (240) affords the nail plate
much different barrier properties to external penetrants than those of the SC. The
composition of the nail plate suggests that it would be comparatively less sensitive
than the SC to the effects of penetration enhancers that produce their effects by
delipidization or fluidization of intercellular lipids. Studies using DMSO have found
this to be true (241–245), together with the observation that the nail plate is inca-
pable of absorbing much of the applied DMSO (246). The nail plate is permeable
to dilute aqueous solutions of a series of low molecular weight homologous alcohols
(239); however, it possesses a unique ability to restrict increasingly the diffusive
passage with increases in alkyl chain length. It was suggested that the nail plate
possessed a highly ‘‘polar’’ penetration route that is capable of excluding permeants
on the basis of their hydrophobicities. Interestingly, the applied concentration of
alcohols was a determinant of their penetration velocities, with pure liquid forms of
the alcohols giving a fivefold decrease in permeation (247).
The existence of a minor ‘‘lipid’’ pathway through the nail matrix, which could
become rate-controlling for hydrophobic solutes, has been suggested based on the
significant decrease in the permeation of the hydrophobic entity n-decanol following
delipidization of the nail plate by chloroform/methanol (247). Increases in nail plate
absorption of the antifungal amorolfine following pretreatment with DMSO have
been demonstrated (248); however, a decrease in the absorption of methanol and
hexanol applied with DMSO was also noted (247). Overall it can be seen that the
permeation characteristics of the nail plate do not correspond to those of the SC and
that the effects of skin penetration enhancers, such as DMSO, cannot be extrapolated
to the nail plate. The targeting of drugs into this area requires the further understand-
ing of its barrier properties to develop enhancers that specifically interfere with the
more keratinous matrix of the nail plate.
Mertin and Lippold (249) have recently examined the role of octanol–water
partition coefficients and water solubility on the permeability and maximum flux of
a homologous series of nicotinic acid esters through the nail plate and a bovine hoof
keratin membrane model. They noticed that the permeability coefficient of the nail
plate as well as the hoof membrane did not increase with increasing partition coef-
ficient (range 7 to >51,000) or lipophilicity of the ester, indicating that these barriers
behaved similar to hydrophilic gel membranes, rather than lipophilic partition mem-
branes, as seen in the SC. Further penetration studies with acetaminophen (parace-
tamol) and phenacetin showed that maximum flux was first a function of drug sol-
ubility in water or in the swollen keratin. Mertin and Lippold were also able to
predict the maximum flux of ten antimycotics through the nail plate on the basis of
their penetration rates through the hoof membrane and their water solubilities (250).
One of the major problems with the application of traditional formulations,
such as creams or solutions of polar vehicles, to the nail plate is the ease with which
they are wiped or washed off the nails. This results in the need for at least once
daily application, but patient compliance over the several months that are usually
Skin Transport 179
required for treatment of fungal infections, such as onychomycosis, decreases with

the length of therapy. Consequently, lipophilic vehicles, especially nail lacquers, have
been developed because of their better adhesion. Pittrof et al. (251) used a porcine
hoof model membrane to optimize a lacquer formulation of amorolfine for drug
release, stability, and ease of application and showed that the drug was able to
penetrate the horn barrier with approximately 1.8% of the applied dose available
under the horn after a 7-day penetration period. More recently, Mertin and Lippold
(252) used the hoof model and the compound chloramphenicol to show that the
course of penetration through the nail plate from a lacquer vehicle followed first-
order kinetics after a lag time of 400 h. Penetration was seen to be initially mem-
brane-controlled and later matrix-controlled as the drug concentration in the lacquer
decreased.
VII. CONCLUDING REMARKS

As seen in this chapter, the skin is an important barrier to the ingress of both ther-
apeutic and potentially toxic compounds. We have attempted to present a review of
the current understanding of the factors that affect the ability of a solute to traverse
the barrier and how these processes can be influenced to enhance penetration. The
increasing availability of data from both in vitro and in vivo studies will, we hope,
make the elucidation of the finer details of these transport processes and the pre-
dictability of skin permeation kinetics, through mathematical modeling and interpre-
tation of structure–activity relations and local physiology, much easier in the years
to come. Until such time, however, we hope that the concepts and studies summa-
rized in this chapter will help bring to many an appreciation of the complex nature
of skin transport processes and the amount of work that has been involved in bringing
us to the level of understanding we have today.
ACKNOWLEDGMENTS
The authors wish to acknowledge the financial support of the National Health and
Medical Research Council of Australia and the continuing encouragement of the
Lions Medical Research and PAH Research Foundations. In addition, one of us
(MSR) also thanks Jonathan Hadgraft, John Pugh, and Karen Milne for our often
lively discussions on which convention is best applied in using chemical potential
and activity as determinants of skin transport. The authors would like to thank Brett
MacFarlane for help in preparation of the manuscript.
REFERENCES
1. Halkier–Sorensen L. Occupational skin disease. In: Roberts MS, Walters KA, eds.
Dermal Absorption and Toxicity Assessment. New York: Marcel Dekker, pp 127–154,
1998.
2. Cross SE. Topical therapeutic agents used in wound care. In: Roberts MS, Walters KA,
eds. Dermal Absorption and Toxicity Assessment. New York: Marcel Dekker, pp 415–
442, 1998.
3. Roberts MS, Walters KA. The relationship between function and barrier function of
the skin. In: Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity Assess-
ment. New York: Marcel Dekker, pp 1–42, 1998.
180 Roberts et al.
4. Hotchkiss SAM. Dermal metabolism. In: Roberts MS, Walters KA, eds. Dermal Ab-
sorption and Toxicity Assessment. New York: Marcel Dekker, pp 43–101, 1998.
5. Scheuplein RJ, Blank IH. Permeability of the skin. Physiol Rev 51:702–747, 1971.
6. Scott RC, Dugard PH, Doss AW. Permeability of abnormal rat skin. J Invest Dermatol
86:201–207, 1986.
7. Schaefer H, Zesch A, Stuttgen G. Penetration, permeation, and absorption of triamcin-
olone acetonide in normal and psoriatic skin. Arch Dermatol Res 258:241–249, 1977.
8. Michaels AS, Chandrasekaran SK, Shaw JE. Drug permeation through human skin.
Theory and in vitro experimental measurement. Am Inst Chem Eng J 21:985–996,
1975.
9. Elias PM, Menon GK. Structural and lipid biochemical correlates of the epidermal
permeability barrier. Adv Lipid Res 24:1–26, 1991.
10. Elias PM, Bonar L, Grayson S, Baden HP. X-ray diffraction analysis of stratum cor-
neum membrane couplets. J Invest Dermatol 80:213–214, 1983.
11. Williams ML, Elias PM. The extracellular matrix of stratum corneum: role of lipids in
normal and physiological function. 3:95–122, 1987.
12. Tregear RT. The permeability of mammalian skin to ions. J Invest Dermatol 46:16–
23, 1966.
13. Kligman AM. The use of sebum. Br J Dermatol 75:307–319, 1963.
14. Kligman A, Christophers E. Preparation of isolated sheets of human stratum corneum.
Arch Dermatol 88:70–73, 1963.
15. Greisemer RD. J Soc Cosmet Chem 11:79, 1960.
16. Miescher G, Schuonberg A. Bull Schweiz Akad Med Wiss 1:101, 1944.
17. Tregear RT. THe permeability of skin to albumin, dextrans and polyvinyl pyrrolidone.
J Invest Dermatol 46:24–27, 1966.
18. Higuchi TJ. Physical chemical analysis of percutaneous absorption process from creams
and ointments. J Soc Cosmet Chem 11:85–87, 1960.
19. Rothman S. J Soc Cosmet Chem 6:193, 1955.
20. Gloor M, Willebrandt U, Thomer G, et al. Water content of the horny layer and skin
surface lipids. Arch Dermatol Res 268:221–223, 1980.
21. Scheuplein RJ. Mechanism of percutaneous adsorption. I. Routes of penetration and
the influence of solubility. J Invest Dermatol 45:334–346, 1965.
22. Scheuplein RJ. Permeability of the skin, a review of major concepts. Curr Prob Der-
matol 7:172–186, 1978.
23. Roberts MS, Anderson RA, Swarbrick J, Moore DE. The percutaneous absorption of
phenolic compounds: the mechanism of diffusion across the stratum corneum. J Pharm
Pharmacol 30:486–490, 1978.
25. Sweeney TM, Downing DT. The role of lipids in the epidermal barrier to water dif-
fusion. J Invest Dermatol 55:135–140, 1970.
26. Squier CA. The permeability of keratinised and nonkeratinised oral apithelium to horse-
radish peroxidase. J Ultrastruct Res 43:160–177, 1973.
27. Elias PM, Friend DS. The permeability barrier in mammalian epidermis. J Cell Biol
65:180–191, 1975.
28. Chandrasekaran SK, Shaw JE. Factors affecting the percutaneous absorption of drugs.
Curr Probl Dermatol 7:142–155, 1975.
29. Albery WJ, Hadgraft J. Percutaneous absorption: theoretical description. J Pharm Phar-
macol 31:129–39, 1979.
30. Bunge AL, Guy RH, Hadgraft J. The determination of a diffusional pathlength through
the stratum corneum. Int J Pharm 188:121–124, 1999.
Skin Transport 181
31. Elias PM, Goerke J, Friend D. Mammalian epidermal barrier layer lipids: composition
and influence on structure. J Invest Dermatol 69:535–546, 1977.
32. Elias PM, Brown BE, Fritsch P, Goerke J, Grey GM, White RJ. Localisation and
composition of lipids in neonatal mouse stratum granulosum and stratum corneum. J
Invest Dermatol 73:339–348, 1979.
33. Berenson GS, Burch GE. Studies on diffusion of water through dead human skin. Am
J Trop Med 31:843–853, 1951.
34. Niemanic MK, Elias PM. In situ precipitation in novel cytochemical technique for
visualisation of permeability pathways in mammalian stratum corneum. J Histochem
Cytochem 28:573–578, 1980.
35. Boddé HE, van den Brink I, Koerten HK, de Haan FHN. Visualisation of in vitro
percutaneous penetration of mercuric chloride; transport through intercellular space
versus cellular uptake through desmosomes. J Control Release 15:227–236, 1991.
36. Elias PM, Cooper ER, Korc A, Brown BE. Percutaneous transport in relation to stratum
corneum structure and lipid composition. J Invest Dermatol 76:297–301, 1981.
37. Bouwstra JA, Dubbelaar FE, Gooris GS, Weerheim AM, Ponec M. The role of cer-
amide composition in the lipid organisation of the skin barrier. Biochim Biophys Acta
1419:127–136, 1999.
38. Elias PM, Mak V, Thornfield C, Feingold KR. Interference with stratum corneum lipid
biogenesis. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption: Drugs–Cos-
metics–Mechanisms–Methodology. 3rd ed. New York: Marcel Dekker, pp 3–55, 1999.
39. Wertz PW. Epidermal lipids. Semin Dermatol 11:106–113, 1992.
40. Bouwstra JA, Gooris GS, Dubbelaar FE, Weerheim AM, Ijzerman AP, Ponec M. Role
of ceramide 1 in the molecular organization of the stratum corneum lipids. J Lipid Res
39:186–96, 1998.
41. Bouwstra JA, Thewalt J, Gooris GS, Kitson N. A model membrane approach to the
epidermal permeability barrier: an X-ray diffraction study. Biochemistry, 36:7717–
7725, 1997.
42. Southwell D, Barry BW. Penetration enhancers for human skin: mode of action of
2-pyrrolidone and dimethylformamide on partition and diffusion of model compounds
water, n-alcohols, and caffeine. J Invest Dermatol 80:507–514, 1983.
43. Blank IH, McAuliffe DJ. Penetration of benzene through human skin. J Invest Dermatol
85:522–526, 1985.
44. Siddiqui O, Roberts MS, Polack AE. Percutaneous absorption of steroids: relative con-
tributions of epidermal penetration and dermal clearance. J Pharmacokinet Biopharm
17:405–424, 1989.
45. Kai T, Isami T, Kobata K, Kurosaki Y, Nakayama T, Kimura T. Keratinised epithelial
transport of ␤-blocking agents. 1. Relationship between physicochemical properties of
drugs and the flux across rat skin and hamster cheek pouch. Chem Pharm Bull 40:
2498–2504, 1992.
46. Kim YH, Ghanem AH, Higuchi WI. Model studies of epidermal permeability. Semin
Dermatol 11:145–56, 1992.
47. Cooper ER, Kasting G. Transport across epithelial membranes. J Controlled Release
6:23–35, 1987.
48. Matsuzaki K, Imaoka T, Asano M, Miyajima K. Development of a model membrane
system using stratum corneum lipids for estimation of drug skin permeability. Chem
Pharm Bull (Tokyo) 41:575–579, 1993.
49. Peck KD, Ghanem AH, Higuchi WI. Hindered diffusion of polar molecules through
and effective pore radii estimates of intact and ethanol treated human epidermal mem-
brane. Pharm Res 11:1306–1314, 1994.
50. Peck KD, Ghanem AH, Higuchi WI. The effect of temperature upon the permeation
of polar and ionic solutes through human epidermal membrane. J Pharm Sci 84:975–
982, 1995.
182 Roberts et al.
51. Yoshida NH, Roberts MS. Solute molecular size and transdermal iontophoresis across
excised human skin. J Controlled Release 25:177–195, 1993.
52. Lai PM, Roberts MS. An analysis of solute structure–human epidermal transport re-
lationships in epidermal iontophoresis using the ionic mobility: pore model. J Con-
trolled Release 58:323–333, 1999.
53. Sznitowska M, Janicki S, Williams AC. Intracellular or intercellular localization of the
polar pathway of penetration across stratum corneum. J Pharm Sci 87:1109–1114,
1998.
54. Menon GK, Elias PM. Morphologic basis for a pore-pathway in mammalian stratum
corneum. Skin Pharmacol 10:235–246, 1997.
55. Florence AT, Attwood D. Physicochemical Principles of Pharmacy. 2nd ed. New York:
Macmillan, pp 158–160, 1988.
56. Florence AT, Attwood D. Physicochemical Principles of Pharmacy. 2nd ed. New York:
Macmillan, pp 48–56, 1988.
57. Barry BW, Harrison SM, Dugard PH. Correlation of thermodynamic activity and va-
pour diffusion through human skin for the model compound benzyl alcohol. J Pharm
Pharmacol 37:84–90, 1985.
58. Davis SS, Higuchi T, Rytting JH. Determination of thermodynamics of functional
groups in solutions of drug molecules. In: Bean HS, Beckett AH, Carless JR, eds.
Advances in Pharmaceutical Sciences. Vol 4. London: Academic Press, pp 73–261,
1974.
59. Hildebrand JH, Scott RL. Regular Solutions. Englewood Cliffs, NJ: Prentice Hall,
1962.
60. Yalkowski SH, Vlavani SC. Solubility and partitioning I. Solubility of nonelectrolytes
in water. J Pharm Sci 69:912–922, 1980.
61. Grant DJW, Higuchi T. Solubility Behaviour of Organic Compounds. New York: Wiley,
pp 24, 1990.
62. Sloan KB, Koch SAM, Siver KG, Flowers FP. Use of solubility parameters of drug
and vehicle to predict fluxes through skin. J Invest Dermatol 87:244–252, 1986.
63. Kadir R, Stempler D, Liron Z, Cohen S. Delivery of theophylline into excised human
skin from alkanoic acid solutions: a ‘‘push–pull’’ mechanism. J Pharm Sci 76:774–
779, 1987.
64. Kadir R, Stempler D, Liron Z, Cohen S. Penetration of adenosine into excised human
skin from binary vehicles: the enhancement factor. J Pharm Sci 77:409–413, 1988.
65. Kadir R, Stempler D, Liron Z, Cohen S. Penetration of theophylline and adenosine
into excised human skin from binary and ternary vehicles: effect of a nonionic surfac-
tant. J Pharm Sci 78:149–153, 1989.
66. Bagley EB, Nelson TP, Scigliano JM. Three dimensional solubility parameters and
their relationship to internal pressure measurements in polar and hydrogen bonding
solvents. Paint Res Inst Proc 43:35–42, 1971.
67. Martin A, Newberger J, Adjei A. Extended Hildebrand solubility approach: solubility
of theophylline in binary polar solvents. J Pharm Sci 69:487–491, 1980.
68. Groning R, Braun FJ. Three dimensional solubility parameters and their use in char-
acterising the permeation of drugs through the skin. Pharmazie 51:337–341, 1996.
69. Ruelle P, Rey–Merment C, Buchmann M, Nam–Tran H, Kesselring UW, Huyskens
PL. A new predictive equation for the solubility of drugs based on the thermodynamics
of mobile disorder. Pharm Res 8:840–850, 1991.
70. Ando HY, Schultz TW, Schnaare RL, Sugita ET. Percutaneous absorption: a new phys-
icochemical predictive model for maximum human in vivo penetration rates. J Pharm
Sci 73:461–467, 1984.
71. Roberts MS, Pugh WJ, Hadgraft J, Watkinson AC. Epidermal permeability—penetrant
structure relationships. 1. An analysis of methods of predicting penetration of mono-
functional solutes from aqueous solutions. Int J Pharm 126:219–233, 1995.
Skin Transport 183
72. Abraham MH, Chadha HS, Mitchell RC. Hydrogen bonding part 36. Determination of
blood brain distribution using octanol–water partition co-efficient. Drug Des Discov
13:123–131, 1995.
73. Yalkowsky SH, Pinal R, Banjeree S. Water solubility: a critique of the solvatochromic
approach. J Pharm Sci 77:74–77, 1988.
74. Abraham MH, Chadna HS, Whiting GS, Mitchell RC. Hydrogen bonding part 32. An
analysis of water–octanol and water alkane partitioning and the delta log P parameter
of Seiler. J Pharm Sci 83:1085–1100, 1994.
75. Leo A, Hansch C, Elkins D. Partition coefficients and their uses. Chem Rev 71:525–
516, 1971.
77. Roberts MS, Pugh WJ, Hadgraft J. Epidermal permeability: penetrant structure rela-
tionships. 2. The effect of H-bonding groups in penetrants on their diffusion through
the stratum corneum. Int J Pharm 132:23–32, 1996.
78. Roberts MS, Anderson RA, Moore DE, Swarbrick J. The distribution of nonelectrolytes
between human stratum corneum and water. Aust J Pharm Sci 6:77–82, 1977.
79. Raykar PV, Fung MC, Anderson BD. The role of protein and lipid domain in the uptake
of solutes by human stratum corneum. Pharm Res 5:140–150, 1988.
80. Chandrasekaran SK, Michaels AS, Campbell PS, and Shaw JE. Scopolamine permea-
tion through human skin in vitro. AIChE J 22:828–832, 1976.
81. Chandrasekaran SK, Bayne W, Shaw JE. Pharmacokinetics of drug permeation through
human skin. J Pharm Sci 67:1370–1374, 1978.
82. Anissimov Y, Roberts MS. Diffusion modelling of percutaneous kinetics: 2. Finite
vehicle volume and solvent deposited solids. J Pharm Sci (submitted for publication).
83. Roberts MS, Anissimov Y, Gonsalvez R. Mathematical models in percutaneous ab-
sorption. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption: Drugs–Cos-
metics–Mechanisms–Methodology. 3rd ed. New York: Marcel Dekker, pp 3–55, 1999.
84. Kasting GB, Smith RL, Anderson BD. Prodrugs for dermal delivery: solubility, mo-
lecular size and functional group effects. In: Sloane KB, ed. Prodrugs. New York:
Marcel Dekker, pp 117–161, 1992.
85. Tojo K. Random brick model for drug transport across stratum corneum. J Pharm Sci
76:889–891, 1987.
86. Heisig M, Lieckfeldt R, Wittum G, Mazurkevich G, Lee G. Nonsteady-state descrip-
tions of drug permeation through stratum corneum I. The basic brick and mortar model.
Pharm Res 13:421–426, 1996.
87. Plewig G, Marples RR. Regional differences in cell sizes of the human stratum cor-
neum. Part 1. J Invest Dermatol 54:13–18, 1970.
88. Anderson BD, Raykar PV. Solute structure–permeability relationships in human stra-
tum corneum. J Invest Dermatol 93:280–286, 1989.
89. Roberts MS, Anderson RA. The percutaneous absorption of phenolic compounds: the
effect of vehicles on the penetration of phenol. J Pharm Pharmacol 27:599–605, 1975.
90. Anissimov Y, Roberts MS. Diffusion modelling of percutaneous absorption kinetics: 1
Effects of flow rate, receptor sampling rate, viable epidermal resistance and epidermal
metabolism for a constant donor concentration. J Pharm Sci 88:1201–1209, 1999.
91. Scheuplein RJ. Properties of the skin as a membrane. Adv Biol Skin 22:125–152, 1972.
92. Kasting GB, Smith RL, Cooper ER. Effect of lipid solubility and molecular size on
percutaneous absorption. Pharmacol Skin 1:138–153, 1987.
93. Flynn GL. Physicochemical determinants of skin absorption. In: Gerrity TR, Henry CJ,
eds. Principles of Route to Route Extrapolation for Risk Assessment. Elsevier Science,
pp 93–127, 1990.
94. Potts RO, Guy RH. Predicting skin permeability. Pharm Res 9:663–669, 1992.
184 Roberts et al.
95. Pugh WJ, Hadgraft J, Roberts MS. Physicochemical determinants of stratum corneum
permeation. In: Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity As-
sessment. New York: Marcel Dekker, pp 245–268, 1998.
96. Barratt MD. Quantitative structure activity relationships for skin permeability. Toxicol
In Vitro 9:27–37, 1995.
97. Liron Z, Cohen S. Percutaneous absorption of alkanoic acids II: Application of regular
solution theory. J Pharm Sci 73:538–542, 1984.
98. Pugh WJ, Roberts MS, Hadgraft J. Epidermal permeability–penetrant structure rela-
tionships. Part 3. Effect of hydrogen bonding interactions and molecular size on dif-
fusion across the stratum corneum. Int J Pharm 138:149–165, 1996.
99. Dearden JC. The QSAR prediction of melting point, a property of environmental rel-
evance. Sci Total Environ 109–110:59–68, 1991.
100. Blank IH. Penetration of low-molecular weight alcohols into the skin. I. The effect of
concentration of alcohol and type of vehicle. J Invest Dermatol 43:425–420, 1964.
101. Roberts MS. Structure–permeability considerations in percutaneous absorption. In:
Scott RC, Guy RH, Hadgraft J, Bodde HE, eds. Prediction of Percutaneous Penetration
—Methods, Measurement and Modelling. Vol 2. IBC Technical Services, pp 210–228,
1991.
102. Hagedorn–Leweke U, Lippold BC. Absorption of sunscreens and other compounds
through human skin in vivo: derivation of a method to predict maximum fluxes. Pharm
Res 12:1354–1360, 1995.
103. Le VH, Lippold BC. Influence of physicochemical properties of homologous esters of
nicotinic acid on skin permeability and maximum flux. Int J Pharm 124:285–292, 1995.
104. Roberts MS, Walker M. Water: the most natural penetration enhancer. In: Walters KA,
Hadgraf J, eds. Pharmaceutical Skin Enhancement. New York: Marcel Dekker, pp 1–
30, 1993.
105. Wurster DE, Kramer SF. Investigations of some factors influencing percutaneous ab-
sorption. J Pharm Sci 50:288–293, 1961.
106. Munies R, Wurster DE. Factors influencing percutaneous absorption of methyl ethyl
ketone. J Pharm Sci 54:554–556, 1964.
107. Bucks D, Maibach H. Occlusion does not uniformly enhance penetration in vivo. In:
Bronaugh RL, Maibach H, eds. Percutaneous Absorption: Drugs–Cosmetics–Mecha-
nisms–Methodology, 3rd ed. New York: Marcel Dekker, pp 81–105, 1999.
108. Bond JR, Barry BW. Hairless mouse skin is limited as a model for assessing the effects
of penetration enhancers in human skin. J Invest Dermatol 90:810–813, 1988.
109. Zatz J. Modification of skin permeation by surface active agents. In: Zatz J, ed. Skin
Permeation Fundamentals and Application. Wheaton, IL: Allured Publishing, pp 149–
162, 1993.
110. Twist JN, Zatz JL. Influence of solvents on paraben permeation through idealised skin
model membranes. J Soc Cosmet Chem 37:429–444, 1986.
111. Barry B, Furand O, Woodford R, Ulshagen K, Hogstad G. Control of the bioavailability
of a topical steroid; comparison of desonide creams 0.05% and 0.1% by vasoconstrictor
studies and clinical trials. Clin Exp Dermatol 12:406–409, 1987.
112. Tojo K, Chiang CC, Chien YW. Drug permeation across the skin: effect of penetrant
hydrophilicity. J Pharm Sci 76:123–126, 1987.
113. Pellet MA, Roberts MS, Hadgraft J. Supersaturated solutions evaluated with an in vitro
stratum corneum tape stripping technique. Int J Pharm 151:91–98, 1997.
114. Schwarb FP, Imanidis G, Smith EW, Haigh JM, and Surber C. Effect of concentration
and degree of saturation of topical fluocinonide formulations on in vitro membrane
transport and in vivo availability on human skin. Pharm Res 16:909–915, 1999.
115. Hadgraft J. Passive enhancement strategies in topical and transdermal drug delivery.
Int J Pharm 184:1–6, 1999.
Skin Transport 185
116. Guy RH, Hadgraft J. Physicochemical interpretation of the pharmacokinetics of per-

cutaneous absorption. J Pharmacokinet Biopharm 11:189–203, 1983.
117. Yano T, Nakagawa A, Masayoshi T, Noda K. Skin permeability of nonsteroidal anti-
inflammatory drugs in man. Life Sci 39:1043–1050, 1986.
118. Hinz RS, Lorence CR, Hodson CD, Hansch C, Hall LL, Guy RH. Percutaneous pen-
etration of para-substituted phenols in vitro. Fundam Appl Toxicol 17:575–583, 1991.
119. Hostynek JJ, Magee PS. Modelling in vivo human skin absorption. Quant Struct Activ
Relation 16:473–479, 1997.
120. Roberts WJ, Sloan KB. Correlation of aqueous and lipid solubilities with flux for
prodrugs of 5-fluorouracil, theophylline, and 6-mercaptopurine: a Potts–Guy approach.
J Pharm Sci 88:515–522, 1999.
121. Roberts MS, Anderson RA. The percutaneous absorption of phenolic compounds: the
effect of vehicles on the penetration of phenol. J Pharm Pharmacol 27:599–605, 1975.
122. Wurster DE. Some physical–chemical factors influencing percutaneous absorption from
dermatologicals. Curr Probl Dermatol 7:156–171, 1978.
123. Menon GK, Lee SH, Roberts MS. Ultrastructural effects of some solvents on the stra-
tum corneum and other skin components: evidence for an extended mosaic partitioning
model of the skin barrier. In: Roberts MS, Walters KA, eds. Dermal Absorption and
Toxicity Assessment. New York: Marcel Dekker, pp 727–751, 1998.
124. Twist JN, Zatz JL. A model for alcohol-enhanced permeation through polydimethyl-
siloxane membranes. J Pharm Sci 79:28–31, 1990.
125. Bronaugh RL, Congdon ER. Percutaneous absorption of hair dyes: correlations with
partition coefficients. J Invest Dermatol 83:124–127, 1984.
126. Barry BW. Dermatological Formulations: Percutaneous Absorption. New York: Marcel
Dekker, 1983.
127. Schaefer H, Rieldelmeier TE. Skin Barrier. Basel: Karger, pp 153–212, 1996.
128. Idson B. Hydration and percutaneous absorption. Curr Probl Dermatol 7:132–141,
1978.
129. Alonso A, Meirelles NC, Tabak M. Effect of hydration upon the fluidity of intercellular
membranes of stratum corneum: an EPR study. Biochim Biophys Acta 1237:6–15,
1995.
130. Alonso A, Meirelles NC, Yushmanov VE, Tabak M. Water increases the fluidity of
intercellular membranes of stratum corneum: correlation with water permeability, elas-
tic, and electrical resistance properties. J Invest Dermatol 106:1058–63, 1996.
131. Wiedmann TS. Influence of hydration on epidermal tissue. J Pharm Sci 77:1037–1041,
1988.
132. Walters KA, Hadgraft J. Pharmaceutical Skin Enhancement. New York: Marcel Dekker,
1993.
133. Smith EW, Maibach HI. Percutaneous Penetration Enhancers. Boca Raton, FL: CRC
Press, 1995.
134. Williams AC, Barry BW. Chemical penetration enhancement: possibilities and prob-
lems. In: Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity Assessment.
New York: Marcel Dekker, pp 297–312, 1989.
135. Cornwell PA, Barry BW. The routes of penetration of ions and 5-fluorouracil across
human skin and the mechanisms of terpene skin penetration enhancers. Int J Pharm
94:189–94, 1993.
136. Yamamoto T, Yamamoto Y. Electrical properties of the epidermal stratum corneum.
Med Biol Eng 14:151–158, 1976.
137. Williams AC, Barry BW. Terpenes and the lipid–protein-partitioning theory of skin
penetration enhancement. Pharm Res 8:17–24, 1991.
138. Francoeur ML, Golden GM, Potts RO. Oleic acid: its effects on stratum corneum in
relation to (trans)dermal drug delivery. Pharm Res 7:621–627, 1991.
186 Roberts et al.
139. Ongpipattanakul B, Burnette RR, Potts RO, and Francoeur ML. Evidence that oleic
acid exists in a separate phase within stratum corneum lipids. Pharm Res 8:350–354,
1991.
140. Cooper ER. In: Solution Behaviour of Surfactants: Theoretical and Applied Aspects.
New York: Plenum, pp 1505–1516, 1982.
141. Goodman, Barry B. Action of penetration enhancers on human stratum corneum as
assessed by differential scanning calorimetry. In: Bronaugh RL, Maibach HI, eds. Per-
cutaneous Absorption, Mechanisms–Methodology–Drug Delivery. 2nd ed. New York:
142. Scheuplein RJ, Ross L. Effect of surfactants and solvents on the permeability of the
epidermis. J Soc Cosmet Chem 21:853–873, 1970.
143. Harrison JE, Watkins AC, Green DM, Hadgraft J, Brain K. The relative effect of Azone
and Transcutol on permeant diffusivity and solubility in human stratum corneum.
Pharm Res 13:542–6, 1996.
144. Zhao K, Singh J. In vitro percutaneous absorption enhancement of propranolol hydro-
chloride through epidermis by terpenes/ethanol. J Controlled Release 62:359–366,
1999.
145. Bach M, Lippold BC. Percutaneous penetration enhancement and its quantification.
Eur J Pharm Biopharm 46:1–13, 1998.
146. Potts RO, Guy RH. Mechanisms of Transdermal Drug Delivery. New York: Marcel
Dekker, 1997.
147. Lee AR, Tojo K. Characterization of skin permeation of vitamin C: theoretical analysis
of penetration profiles and differential scanning calorimetry study. Chem Pharm Bull
(Tokyo) 46:174–177, 1998.
148. Goates CY, Knutson K. Enhanced permeation of polar compounds through human
epidermis. I. Permeability and membrane structural changes in the presence of short
chain alcohols. Biochim Biophys Acta 1195:169–179, 1994.
149. Wolfram LJ, Maibach HI. Percutaneous absorption of hair dyes. Arch Dermatol Res
277:235–241, 1985.
150. Dressler W. Percutaneous absorption of hair dyes. In: Roberts MS, Walters KA, eds.
1998.
151. Bhargava HN, Leonard PA. Triclosan: Applications and safety. Am J Infect. Control
24:209–218, 1996.
152. Islam MS, Zhao L, McDougal JN, Flynn GL. Uptake of chloroform by skin during
short exposures to contaminated water. Risk Anal 15:343–352, 1995.
153. Imokawa G, Mishima Y. Cumulative effect of surfactants on cutaneous horny layers:
adsorption onto human keratin layers in vivo. Contact Dermatitis 5:357–366, 1979.
154. Hashiguchi T, Ryu A, Itoyama T. Uchida K, Yamaguchi H. Study of the effective dose
of topical antifungal agent. Omoconazole nitrate, on the basis of percutaneous phar-
macokinetics in guinea pigs and mice. J Pharm Pharmacol 49:757–761, 1997.
155. Fullerton A, Hoelgaard A. Binding of nickel to human epidermis in vitro. Br J Dermatol
119:675–682, 1988.
156. Dick IP, Blain PG, Williams FM. The percutaneous absorption and skin distribution of
lindane in man. II. In vitro studies. Hum Exp Toxicol 16:652–657, 1997.
157. Scheuplein RJ, Blank IH. Mechanisms of percutaneous absorption. IV Penetration of
nonelectrolytes (alcohols) from aqueous solutions and from pure liquids. J Invest Der-
matol 60:286–296, 1973.
158. Barry BW, Harrison SM, Dugard PH. Vapour and liquid diffusion of model penetrants
through human skin; correlation with thermodynamic activity. J Pharm Pharmacol 37:
226–236, 1985.
Skin Transport 187
159. Berner B, Wilson DR, Guy RH, Mazzenga GC, Clarke FH, Maibach HI. The relation-
ship of pKa and acute skin irritation in man. Pharm Res 5:660–663, 1998.
160. Nangia A, Andersen PH, Berner B, Maibach HI. High dissociation constants (pKa) of
basic permeants are associated with in vivo skin irritation in man. Contact dermatitis
34:237–242, 1996.
161. Barratt MD. Quantitative structure activity relationships (QSAR’s) for skin corrosivity
of organic acids, bases and phenols: principal components and neural network analysis
of extended datasets. Toxic In Vitro 10:85–94, 1996.
162. Barratt MD. Quantitative structure activity relationships for skin corrosivity of organic
acids, bases and phenols. Toxicol Lett 75:169–176, 1995.
163. Poulsen BJ, Chowhan ZT, Pritchard R, Katz M. The use of mixtures of topical steroids
as a mechanism for improving total drug bioavailability: a preliminary report. Curr
Probl Dermatol 7:107–120, 1978.
164. Gao S, Singh J. In vitro percutaneous absorption enhancement of a lipophilic drug
tamoxifen by terpenes. J Controlled Release 51:193–199, 1998.
165. Kaplun–Frischoff Y, Touitou E. Testosterone skin penetration enhancement by menthol
through formation of eutectic with drug and interaction with skin lipids. J Pharm Sci
86:1394–1399, 1997.
166. Sasaki H, Kojima M, Mori Y, Nakamura J, and Shibasaki J. Enhancing effect of pyr-
rolidone derivatives on transdermal penetration of 5-fluorouracil, triamcinolone ace-
tonide, indomethacin and flurbiprofen. J Pharm Sci 80:533–538, 1991.
167. Kondo S, Yamanaka C, Sugimoto I. Enhancement of transdermal delivery by superflu-
ous thermodynamic potential. III. Percutaneous absorption of nifedipine in rats. J Phar-
macobiodyn 10:643–749, 1987.
168. Hadgraft J, Peck J, Williams DG, Pugh WJ, Allan G. Mechanism of action of skin
penetration enhancers/retarders: Azone and analogues. Int J Pharm 141:17–25, 1996.
169. Gorukanti SR, Li L, Kim KH. Transdermal delivery of antiparkinsonian agent, benz-
tropine. I. Effect of vehicles on skin permeation. Int J Pharm 192:159–172, 1999.
170. Kim N, El-Khalili M, Henary MM, Strekowski L, Michniak BB. Percutaneous pene-
tration enhancement activity of aromatic S,S-dimethyliminosulfurane. Int J Pharm 187:
219–229, 1999.
171. Zatz J. Simulation studies of skin permeation. J Soc Cosmet Chem 43:37–48, 1992.
172. Singh P, Roberts MS, Maibach HI. Modelling of plasma levels of drugs following
transdermal iontophoresis. J Controlled Release 33:293–298, 1995.
173. Hadgraft J, Wolff M. In vitro/In vivo correlations in transdermal drug delivery. In:
Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity Assessment. New York:
174. Cleary GW. Transdermal delivery systems: a medical rationale. In: Shah VP, Maibach
HI, eds. Topical Drug Bioavailability, Bioequivalence and Penetration. New York: Ple-
num Press, pp 17–68, 1993.
175. Riegelman S. Pharmacokinetics. Pharmacokinetic factors affecting epidermal penetra-
tion and percutaneous absorption. Clin Pharmacol Ther 16:873–883, 1974.
176. Beckett AH, Gorrod JW, Taylor DC. Comparison of oral and percutaneous routes in
man for the systemic absorption of ‘‘ephedrines.’’ J Pharm Pharmacol 24(suppl):65–
70, 1972.
177. McCrea JB, Vlasses PH, Franz TJ, Zeoli L. Transdermal timolol: ␤ blockade and
plasma concentrations after application for 48 hours and 7 days. Pharmacotherapy, 10:
289–293, 1990.
178. Kubota K. A compartment model for percutaneous drug absorption. J Pharm Sci 80:
502–504, 1991.
179. Kubota K, Maibach HI. A compartment model for percutaneous absorption: compati-
bility of lag time and steady-state flux with diffusion model. J Pharm Sci 81:863–865,
1992.
188 Roberts et al.
180. Guy RH, Hadgraft J, Maibach HI. A pharmacokinetic model for percutaneous absorp-
tion. J Int Pharm 11:119–129, 1982.
181. Wallace SM, Barnett G. Pharmacokinetic analysis of percutaneous absorption: evidence
of parallel penetration pathways for methotrexate. J Pharmacokinet Biopharm 6:315–
325, 1978.
182. Berner B. Pharmacokinetics of transdermal drug delivery. J Pharm Sci 74:718–721,
1985.
183. Birmingham BK, Greene DS, Rhodes CT. Percutaneous absorption of salicylic acid in
rabbits. Drug Dev Ind Pharm 5:29–40, 1979.
184. Holford NHG, Scheiner LB. Understanding the dose–effect relationship: clinical ap-
plication of pharmacokinetic–pharmacodynamic models. Clin Pharmacokinet 6:429–
453, 1981.
185. Singh GJ, Adams WP, Lesko LJ, Shah VP, Molzon JA, Williams RL, and Pershing
LK. Development of in vivo bioequivalence methodology for dermatologic corticoste-
roids based on pharmacokinetic modelling. Clin Pharmacol Ther 66:346–357, 1999.
186. Siddiqui O, Roberts MS, Polack AE. Topical absorption of methotrexate: role of dermal
transport. Int J Pharm 27:193–203, 1985.
187. Borsadia S, Ghanem AH, Seta Y, Higuchi WI, Flynn GL, Behl CR, and Shah VP.
Factors to be considered in the evaluation of bioavailability and bioequivalence of
topical formulations. Skin Pharmacol 5:129–145, 1992.
188. Mehta SC, Afouna MI, Ghanem AH, Higuchi WI, Kern ER. Relationship of skin target
site free drug concentration (C *) to the in vivo efficacy: an extensive evaluation of the
predictive value of the C * concept using acyclovir as a model drug. J Pharm Sci 86:
797–801, 1997.
189. Patel PJ, Ghanem AH, Higuchi WI, Srinivasan V, Kern ER. Correlation of in vivo
topical efficacies with in vitro predictions using acyclovir formulations in the treatment
of cutaneous HSV-1 infections in hairless mice: an evaluation of the predictive value
of the C * concept. Antiviral Res 29:279–286, 1996.
190. Cross SE, Roberts MS. Subcutaneous absorption kinetics and local tissue distribution
of interferon and other solutes. J Pharm Pharmacol 45:606–609, 1993.
191. Cross SE, Anderson C, Roberts MS. Topical penetration of commercial salicylate esters
and salts using human isolated skin and clinical microdialysis studies. Br J Clin Phar-
macol 46:29–35, 1998.
192. Tegeder I, Muth–Selbach U, Lotsch J, Rusing G, Oelkers R, Brune K, Meller S, Kelm
GR, Sorgel F, Geisslinger G. Application of microdialysis for the determination of
muscle and subcutaneous tissue concentrations after oral and topical ibuprofen admin-
istration. Clin Pharmacol Ther 65:357–368, 1999.
193. Muller M, Rastelli C, Ferri P, Jansen B, Breiteneder H, Eichler HG. Transdermal pen-
etration of diclofenac after multiple epicutaneous administration. J Rheumatol 25:
1833–1836, 1998.
194. Singh P, Maibach HI, Roberts MS. Site of effects. In: Roberts MS, Walters KA, eds.
1998.
195. McNeill SC, Potts RO, Francoeur ML. Local enhanced topical delivery (LETD) of
drugs: does it truly exist? Pharm Res 9:1422–1427, 1992.
196. Cross SE, Megwa SA, Benson HAE, Roberts MS. Self promotion of deep tissue pen-
etration and distribution of methylsalicylate after topical application. Pharm Res 16:
427–433, 1999.
197. Singh P, Roberts MS. Effects of vasoconstriction on dermal pharmacokinetics and local
tissue distribution of compounds. J Pharm Sci 83:783–791, 1994.
198. Roberts MS, Cross SE. A physiological pharmacokinetic model for solute disposition
in tissues below a topical application site. Pharm Res 16:1394–1400, 1999.
Skin Transport 189
199. Wester RC, Christoffel J, Hartway T, Poblete N, Maibach HI, Forsell J. Human cadaver
skin viability for in vitro percutaneous absorption: storage and detrimental effects of
heat separation and freezing. Pharm Res 15:82–84, 1998.
200. Nakashima E, Noonan PK, Benet LZ. Transdermal bioavailability and first-pass skin
metabolism: a preliminary evaluation with nitroglycerine. J Pharmacokinet Biopharm
15:423–437, 1987.
201. Megwa, SA, Benson HA, Roberts MS. Percutaneous absorption of salicylates from
some commercially available topical products containing methyl salicylate or salicylic
salts in rats. J Pharm Pharmacol 47:891–896, 1995.
202. Marks R. Pharmacokinetics and safety review of tazarotene. J Am Acad Dermatol 39:
S134–S138, 1998.
203. Tang-Liu DD, Matsumoto RM, Usansky JI. Clinical pharmacokinetics and drug me-
tabolism of tazarotene: a novel topical treatment for acne and psoriasis. Clin Pharma-
cokinet 37:273–287, 1999.
204. Bronaugh RL, Kraeling EK, Yourick JJ, Hood HL. Cutaneous metabolism during in
vitro percutaneous absorption. In: Bronaugh RL, Maibach HI, eds. Percutaneous Ab-
sorption: Drugs–Cosmetics–Mechanisms–Methodology, 3rd ed. New York: Marcel
Dekker, pp 57–64, 1999.
205. Ando HY, Ho NF, Higuchi WI. Skin as an active metabolizing barrier I: Theoretical
analysis of topical bioavailability. J Pharm Sci 66:1525–1528, 1977.
206. Yu CD, Fox JL, Ho NF, Higuchi WI. Physical model evaluation of topical prodrug
delivery—simultaneous transport and bioconversion of vidarabine-5⬘-valerate I: phys-
ical model development. J Pharm Sci 68:1341–1346, 1979.
207. Yu CD, Fox JL, Ho NF, Higuchi WI. Physical model evaluation of topical prodrug
delivery—simultaneous transport and bioconversion of vidarabine-5⬙-valerate II: pa-
rameter determinations. J Pharm Sci 68:1347–1357, 1979.
208. Yu CD, Higuchi WI, Ho NF, Fox GL, Flynn GL. Physical model evaluation of topical
prodrug delivery—simultaneous transport and bioconversion of vidarabine-5⬙-valerate
III: permeability differences of vidarabine and n-pentanol in components of hairless
mouse skin. J Pharm Sci 69:770–772, 1980.
209. Yu CD, Fox JL, Higuchi WI, Ho NF. Physical model evaluation of prodrug delivery
—simultaneous transport and bioconversion of vidarabine-5⬙-valerate IV: distribution
of esterase and deaminase enzymes in hairless mouse skin. J Pharm Sci 69:772–775,
1980.
210. Yu CD, Gordon NA, Fox JL, Higuchi WI, Fox NF. Physical model evaluation of topical
prodrug delivery—simultaneous transport and bioconversion of vidarabine-5⬙-valerate
V: mechanistic analysis of influence of nonhomogeneous enzyme distributions in hair-
less mouse skin. J Pharm Sci 69:775–780, 1980.
211. Liu P, Higuchi WI, Song WQ, Kurihara–Bergstrom T, Good WR. Quantitative evalu-
ation of ethanol effects on diffusion and metabolism of beta-estradiol in hairless mouse
skin. Pharm Res 8:865–872, 1991.
212. Liu P, Higuchi WI, Ghanem AH, Good WR. Transport of beta-estradiol in freshly
excised human skin in vitro: diffusion and metabolism in each skin layer. Pharm Res
11:1777–1784, 1994.
213. Sugibayashi K, Hayashi T, Morimoto Y. Simultaneous transport and metabolism of
ethyl nicotinate in hairless rat skin after its topical application: the effect of enzyme
distribution in the skin. J Controlled Release 62:201–208, 1999.
214. Gysler A, Kleuser B, Sippl W, Lange K, Holtje HD, Korting HC. Skin penetration and
metabolism of topical gluticocorticoids in reconstructed epidermis and in excised hu-
man skin. Pharm Res 16:1386–1391, 1999.
215. Hatanaka T, Rittirod T, Katayama K, Koizumi T. Influence of enzyme distribution and
diffusion on permeation profile of prodrug through viable skin: theoretical aspects for
190 Roberts et al.
several steady-state fluxes in two transport directions. Biol Pharm Bull 22:623–626,
1999.
216. Rittirod T, Hatanaka T, Uraki A, Hino K, Katayama K, Koizumi T. Species differences
in simultaneous transport and metabolism of ethyl nicotinate in skin. Int J Pharm 178:
161–169, 1999.
217. Seko N, Bando H, Lim CW, Yamashita F, Hashida M. Theoretical analysis of the effect
of cutaneous metabolism on skin permeation of parabens based on a two-layer skin
diffusion/metabolism model. Biol Bull Pharm 22:281–287, 1999.
218. Malkinson FD, Ferguson EH. Percutaneous absorption of hydrocortisone-4-14C in two
human subjects. J Invest Dermatol 25:281–283, 1955.
219. Shah VP, Flynn GL, Yacobi A, Maibach HI, Bon C, Fleischer NM, Franz TJ, Kaplan
SA, Kawamoto J, Lesco LJ, Marty JP, Pershing LK, Schaefer H, Sequeira JA, Shri-
vastava SP, Wilkin J, Williams RL. Bioequivalence of topical dermatological dosage
forms—methods of evaluating bioequivalence. AAPS/FDA Workshop on: Bioequiva-
lence of Topical Dermatological Dosage Forms—Methods of Evaluating Bioequiva-
lence. Sept 4–6, 1996. Bethesda, MD. Skin Pharmacol Appl Skin Physiol 11:117–124,
1998.
SA, Kawamoto J, Lesco LJ, Marty JP, Pershing LK, Schaefer H, Sequeira JA, Shri-
vastava SP, Wilkin J, Williams RL. Bioequivalence of topical dermatological dosage
forms—methods of evaluating bioequivalence. Pharm Res 15:167–171, 1998.
221. Rougier A, Rallis M, Krien P, Lotte C. In vivo percutaneous absorption: a key role for
stratum corneum partitioning. Arch Dermatol Res 282:498–505, 1990.
222. Rougier A, Dupuis D, Lotte C, Maibach HI. Stripping method for measuring percu-
taneous absorption in vivo. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorp-
tion: Drugs–Cosmetics–Mechanisms–Methodology. 3rd ed. New York: Marcel Dekker,
pp 375–394, 1999.
223. Franz TJ. Percutaneous absorption on the relevance of in vitro data. J Invest Dermatol
64:190–195, 1975.
224. Roberts MS, Favretto WA, Mayer A, Reckmann M, Wongseelasote T. Topical bio-
availability of methyl salicylate. Aust NZ J Med 12:303–305, 1982.
225. Jamoulle JC, Schaeffer H. Cutaneous bioavailability, bioequivalence and percutaneous
absorption: in vivo methods, problems and pitfalls. In: Shah VP, Maibach HI, eds.
Topical Drug Bioavailability, Bioequivalence and Penetration. New York: Plenum
Press, pp 129–153, 1973.
226. Walters KA, Roberts MS. Veterinary applications of skin penetration enhancers. In:
Walters KA, Hadgraft J, eds. Pharmaceutical Skin Enhancement. New York: Marcel
Dekker, pp 345–364, 1993.
227. Scott RC, Walker M, Dugard PH. A comparison of the in vitro permeability properties
of human and some laboratory animal skins. Int J Cosmet 8:189–194, 1986.
228. Ebling FJG, Hale PA, Randall VA. Hormones and hair growth. In: Physiology, Bio-
chemistry and Molecular Biology of the Skin. Vol 1. New York: Oxford University
Press, pp 660–698, 1991.
229. Hueber F, Wepierre J, Schaefer H. Role of transepidermal and transfollicular routes in
percutaneous absorption of hydrocortisone and testosterone: in vivo study in the hair-
less rat. Skin Pharmacol 5:99–107, 1992.
230. Rolland A, Wagner N, Chatelus A, Shroot B, Schaefer H. Site-specific drug delivery
to pilosebaceous structures using polymeric microspheres. Pharm Res 10:1738–1744,
1993.
231. Lieb LM, Liimatta AP, Bryan RN, Brown BD, Krueger GG. Description of the intra-
follicular delivery of large molecular weight molecules to follicles of human scalp in
vitro. J Pharm Sci 86:1022–1029, 1997.
Skin Transport 191
232. Montagna W. Penetration and local effects of vitamin A on the skin of the guinea pig.
Proc Exp Biol Med 86:668–672, 1954.
234. Lauer AC, Ramachandran C, Lieb LM, Niemiec S, Weiner ND. Targeted delivery of
the pilosebaceous unit via liposomes. Adv Drug Deliv Rev 18:311–324, 1996.
235. Bamba FL, Wepierre J. Role of the appendageal pathway in the percutaneous absorp-
tion of pyridostigmine bromide in various vehicles. Eur J Drug Metab Pharmacokinet
18:339–348, 1993.
236. Maibach, HI, Feldman RJ, Milby TH, Serat WF. Regional variation in percutaneous
penetration in man. Pesticides. Arch Environ Health 23:208–211, 1971.
237. Rougier A, Lotte C, Maibach HI. In vivo percutaneous penetration of some organic
compounds related to anatomic site in humans: predictive assessment by the stripping
method. J Pharm Sci 76:451–454, 1987.
238. Lauer AC. Percutaneous drug delivery to the hair follicle. In: Bronaugh RL, Maibach
HI, eds. Percutaneous Absorption: Drugs–Cosmetics–Mechanisms–Methodology, 3rd
ed. New York: Marcel Dekker, pp 449–427, 1999.
239. Walters KA, Flynn G, Marvel JR. Physicochemical characterisation of the human nail:
permeation pattern for water and the homologous alcohols and differences with respect
to the stratum corneum. J Pharm Pharmacol 35:28–33, 1983.
240. Baden HP, Goldsmith LA, Flemming B. A comparative study of the physicochemical
properties of human keratinized tissues. Biochim Biophys Acta 322:269–278, 1973.
241. Chandrasekaran SK, Campbell PS, Michaels AS. Effect of dimethyl sulphoxide on drug
permeation through human skin. AIChE J 23:810–816, 1977.
242. Embery G, Dougard PH. The isolation of dimethyl sulfoxide soluble components from
human epidermal preparations: a possible mechanism of action of dimethyl sulfoxide
in effecting percutaneous migration phenomena. J Invest Dermatol 57:308–311, 1971.
243. Allenby AC, Creasey NH, Edington JAG, Fletcher JA, Schock C. Mechanism of action
of accelerants of skin penetration. Br J Dermatol 81(suppl 4):47–55, 1969.
244. MacGregor WS. The chemical and physical properties of DMSO. Ann NY Acad Sci
141:3–12, 1967.
245. Montes LF, Day JL, Wand CJ, Kennedy L. Ultrastructural changes in the horny layer
following local application of dimethyl sulfoxide. J Invest Dermatol 48:184–196, 1967.
246. Kligman A. Topical pharmacology and toxicology of dimethyl sulfoxide. JAMA 193:
796–804, 1965.
247. Walters KA, Flynn GL, Marvel JR. Physicochemical characterisation of the human
nail: solvent effects on the permeation of homologous alcohols. J Pharm Pharmacol
37:771–775, 1985.
248. Franz TJ. Absorption of amorolfine through human nail. Dermatology 184(suppl 1):
18–20, 1992.
249. Mertin D, Lippold BC. In vivo permeability of the human nail and of a keratin mem-
brane from bovine hooves: influence of the partition coefficient octanol/water and the
water solubility of drugs on their permeability and maximum flux. J Pharm Pharmacol
49:30–34, 1997.
250. Mertin D, Lippold BC. In vivo permeability of the human nail and of keratin membrane
from bovine hooves: prediction of the penetration rate of antimycotics through the nail
plate and their efficacy. J Pharm Pharmacol 49:866–872, 1997.
251. Pittrof F, Gerhards J, Erni W, Klecak G. Loceryl nail lacquer—realization of a new
galenical approach to onchomycosis therapy. Clin Exp Dermatol 17(suppl 1):26–28,
1992.
252. Mertin D, Lippold BC. In vivo permeability of the human nail and of a keratin mem-
brane from bovine hooves: penetration of chloramphenicol from lipophilic vehicles and
a nail lacquer. J Pharm Pharmacol 49:241–245, 1997.
192 Roberts et al.
253. Homalle A. Experiences physiologiques sur l’absorption par la tegument externe chez
l’homme dans le bain. Union Med 7:462–463, 1853.
254. Fleischer R. Untersuchungen uber das resorptionsvermaogen der menschlichen haut.
Habilitationschrift 81, 1877.
255. Schwenkenbecker A. Das absorptionsverniogen der haut. Arch Anat Physiol 121–165,
1904.
256. Oppenheim M. Beitrage zin frage der hautabsorption mit besonderer beruchtigung der
enkrankten haut. Arch Dermatol U Syph 93:85–106, 1908.
257. Smith HW, Clawes HA, Marshall EK. Mustard gas. IV. The mechanism of absorption
by the skin. J Pharmacol 13:1–30, 1919.
258. Rein H. Experimental electroendosomotic studies on living human skin. Z Biol 81:
125–140, 1924.
259. Collander R, Burland H. Permeability in chara ceratophylla. II. Permeability to non-
electrolytes. Acta Botan Fenn 11:1–114, 1930.
260. Wolf J. Die innere strucktur der zellen des stratum desquamans der menchlichen epi-
dermis. Z Mikroskop Anat Forsch 46:170–202, 1939.
261. MacKee GM, Sulzberger MB, Herrmann F, Baer RL. Histological studies on percu-
taneous penetration with special reference to the effect of vehicles. J Invest Dermatol
6:43–61, 1945.
262. Rosendal T. Concluding studies on the conducting properties of human skin to AC.
Acta Physiol Scand 9:39–49, 1945.
263. Blank IH. Factors which influence the water content of the stratum corneum. J Invest
Dermatol 18:433–440, 1953.
264. Blank IH. Further observations on factors which influence the water content of the
straum corneum. J Invest Dermatol 21:259–269, 1953.
265. Rothman S. Physiology and Biochemistry of the Skin. Chicago: University Press, pp
26–59, 1954.
266. Blank IH, Gould E. Study of mechanisms which impede the penetration of anionic
surfactants into skin. J Invest Dermatol 37:311–315, 1962.
267. Kligman A. The biology of the stratum corneum. In: Montagna W, Lobitz WC, eds.
The Epidermis. New York: Academic, pp 387–433, 1964.
268. Wagner JG. Biopharmaceutics: absorption aspects. J Pharm Sci 50:359–363, 1961.
269. Barr MJ. Percutaneous absorption. J Pharm Sci 51:395–409, 1962.
270. Malkinson FD, Rothman S. Percutaneous absorption. In: Jadassohn J, ed. Handbuch
der Haut- und Geschlecht-skrankherten, Normale und Pathologische Physiologie der
Haut. Berlin: Springer, 1:90–156, 1963.
271. Malkinson FD. Permeability of the stratum corneum. In: Montagna W, Lobitz WC,
eds. The Epidermis. New York: Academic, pp 435–452, 1964.
272. Tregear RT. The permeability of the skin to molecules of widely differing properties.
In: Rook A, Champion RH, eds. Progress In Biological Sciences in Relation to Der-
matology 2nd ed. London: Cambridge University Press, pp 275–281, 1964.
273. Welles FV, Lubowe II. Cosmetics and the Skin. New York: Van Nostrand Reinhold,
1964.
274. Stoughton RB. Penetration absorption. Toxicol Appl Pharmacol 7(suppl 2):1–6, 1965.
275. Vinson LJ, Masurat T, Singer EJ. Basic Studies in Percutaneous Absorption: Final
Comprehensive Report (no. 10). Edgewood, MD: Army Chemical Center, 1965.
276. Reiss F. Therapeutics: percutaneous absorption, a critical and historic review. Am J
Med Sci 252:588–602, 1966.
277. Vickers CFH. In: McKenna RMB, ed. Modern Trends in Dermatology. Vol. 3. London:
Butterworths, pp 84, 1966.
278. Barrett CW. J Soc Cosmet Chem 20:487, 1969.
Skin Transport 193
279. Blank IH, Scheuplein RJ. Transport into and within the skin. Br J Dermatol 81(suppl
4):4–10, 1969.
280. Winklemann RK. The relationship of structure of the epidermis to percutaneous ab-
sorption. Br J Dermatol 81(suppl 4):11–22, 1969.
281. Katz M, Poulsen JB. Absorption of drugs through the skin. In: Brodic EE, Gillette JR,
eds. Concepts in Biochemical Pharmacology. Berlin: Springer-Verlag, 1971.
282. Katz M, Poulsen JB. Corticoid, vehicle and skin interaction in percutaneous absorption.
J Soc Cosmet Chem 23:565–590, 1972.
283. Katz M. In: Ariens EJ, ed. Drug Design. Vol 4. New York: Academic Press, pp 93–
148, 1973.
284. Poulsen BJ. Design of topical drug products: Biopharmaceutics. In: Ariens EJ, ed. Drug
Design, Vol 4. New York: Academic Press, pp 149–190, 1973.
285. Anderson RA, Roberts MS. Absorption of drugs through the skin. Aust J Pharm Sci
3:75–80, 1974.
286. Schaefer H. Skin penetration of drugs. Z Hautkr 49:719, 1974.
287. Idson B. Percutaneous absorption. J Pharm Sci 64:901–924, 1975.
288. Dugard PH. Skin permeability theory in relation to measurements of percutaneous
absorption in toxicology. In: Marzulli FN, Maibach HI, eds. Dermatotoxicology and
Pharmacology. Washington: Hemisphere, pp 525–550, 1977.
289. Higuchi T. In: Roche B, ed. Design of Biopharmaceutical Properties Through Prodrugs
and Analogues. Washington: American Pharmaceutical Association, pp 409, 1977.
290. Malkinson FD, Guhlmann L. In: Drill VA, Lazar P, eds. Cutaneous Toxicology. New
York: Academic Press, pp 63, 1977.
291. Wester RC, Maibach HI. Percutaneous absorption in man and animals: a perspective.
In: Drill VA, Lazar P, eds. Cutaneous Toxicity. New York: Academic Press, 1977.
292. Scheuplein RJ. Skin permeation. In: Jarrett A, ed. The Physiology and Pathophysiology
of the Skin. Vol 5. New York: Academic Press, pp 1659–1752, 1978.
293. Flynn GL. Topical drug absorption and topical pharmaceutical systems. In: Banker GS,
Rhodes CT, eds. Modern Pharmaceutics. New York: Marcel Dekker, pp 263, 1979.
294. Scheuplein RJ. Percutaneous absorption: theoretical aspects. In: Mauvais–Jarvis PM,
et al., eds. Percutaneous Absorption of Steroids. New York: Academic Press, pp 1–17,
1980.
295. Lien EJ, Tong GL. Physicochemical properties and percutaneous absorption of drugs.
J Soc Cosmet Chem 24:371–384, 1973.
296. Ackermann C, Flynn GL, Smith WM. Ether–water partitioning and permeability
through nude mouse skin in vitro II. Hydrocortisone n-alkyl esters, alkanols and hy-
drophilic compounds. Int J Pharm 36:67–71, 1987.
297. Pugh WJ, Hadgraft J. Ab initio prediction of human skin permeability coefficients. Int
J Pharm 103:163–178, 1994.
298. Lien EL, Gao H. QSAR analysis of skin permeability of various drugs in man as
compared to in vivo and in vitro studies in rodents. Pharm Res 12:583–587, 1995.
299. Potts RO, Guy RH. A predictive algorithm for skin permeability: the effects of molec-
ular size and hydrogen bond activity. Pharm Res 12:1628–1633, 1995.
300. Abraham MH, Chandha HS, Mitchell RC. The factors that influence skin penetration
of solutes. J Pharm Pharmacol 47:8–16, 1995.
301. Abraham MH, Martins F, Mitchell RC. Algorithms for skin permeability using hydro-
gen bond descriptors: the problem of steroids. J Pharm Pharmacol 49:858–865, 1997.
302. Scheuplein RJ, Blank IH, Brauner GJ, MacFarlane DJ. Percutaneous absorption of
steroids. J Invest Dermatol 52:63–70, 1969.
303. Hadgraft J, Ridout G. Development of model membranes for percutaneous absorption
measurements. I. Isopropyl myristate. Int J Pharm 39:149–156, 1987.
194 Roberts et al.
304. Roy SD, Flynn GL. Transdermal delivery of narcotic analgesics: comparative perme-
abilities of narcotic analgesics through human cadavar skin. Pharm Res 6:825–832,
1989.
305. Dutkiewicz T, Tyras H. A study of the skin absorption of ethylenebenzene in man. Br
J Ind Med 24:330–332, 1967.
306. Jolicoeur LM, Nassiri MR, Shipman C, Choi H, Flynn GL. Etorphine is an opiate
analgesic physicochemically suited to transdermal delivery. Pharm Res 9:963–965,
1992.
307. Roy SD, Flynn GL. Transdermal delivery of narcotic analgesics: pH, anatomical and
subject influences on cutaneous permeability of fentanyl and sufentanil. Pharm Res 7:
842–847, 1990.
308. Anderson BD, Higuchi WI, Raykar PV. Heterogeneity effects on permeability–partition
coefficient relationships in human stratum corneum. Pharm Res 5:566–573, 1988.
309. Cowhan ZT, Pritchard R. Effects of surfactants on percutaneous absorption of naproxen
I: comparisons of rabbit, rat and human excised skin. J Pharm Sci 67:1272–1274, 1978.
310. Dutkiewicz T, Tyras H. Skin absorption of toluene, styrene and xylene by man. Br J
Ind Med 25:243, 1968.
311. Shelley WB, Melton FM. Factors affecting the acceleration of histamine through nor-
mal intact skin. J Invest Dermatol 13:61–71, 1949.
312. Black JG, Howes D, Rutherford T. Skin deposition and penetration of trichlorocarban-
ilide. Toxicology 3:253–264, 1975.
313. Illel B, Schaefer H. Transfollicular percutaneous absorption: skin model for quantitative
studies. Acta Dermatol Venereol 68:427–430, 1988.
Hadgraft J, eds. Prediction of Percutaneous Penetration, Methods, Measurements, Mod-
elling. London: IBC Technical Services, pp 163–173, 1990.
315. Illel B, Schaefer H, Wepierre J, Doucet O. Follicles play an important role in percu-
taneous absorption. J Pharm Sci 80:424–427, 1991.
316. Rutherford T, Black JG. The use of autoradiography to study the localization of ger-
micide in the skin. Br J Dermatol 81:75–87, 1969.
317. Bidmon HJ, Pitts JD, Solomon HF, Bondi JV, Stumpf WE. Estradiol distribution and
penetration in rat skin after topical application, studied by high resolution autoradiog-
raphy. Histochemistry 95:43–54, 1990.
318. Li L, Margolis LB, Lishko VK, Hoffman RM. Product-delivering liposomes specifi-
cally target hair follicles in histocultured intact skin. In Vitro Cell Dev Biol 28A:679–
681, 1992.
319. Li L, Lishko V, Hoffman RM. Liposome targeting of high molecular weight DNA to
the hair follicles of histocultured skin: a model for gene therapy of the hair growth
process. In Vitro Cell Dev Biol Anim 29A:258–260, 1993.
320. Li L, Lishko VK, Hoffman RM. Liposomes can specifically target entrapped melanin
to hair follicles in histocultured skin. In Vitro Cell Dev Biol 29A(3 pt 1):192–194,
1993.
321. Hueber F, Schaefer H, Wepierre J. Effect of vehicle on transdermal and transfollicular
absorption of four steroids of different lipophilicity. In: Brain KR, James VJ, Walters
KA, eds. Prediction of Percutaneous Penetration. Cardiff: STS Publishing, pp 264–
271, 1993.
322. Lieb LM, Flynn G, Weiner N. Follicular (pilosebaceous unit) deposition and pharma-
cological behaviour of cimetidine as a function of formulation. Pharm Res 11:1419–
1423, 1994.
323. Niemiec SM, Ramachandran C, Weiner N. Influence of nonionic liposomal composition
on topical delivery of peptide drugs into pilosebaceous units: an in vivo study using
the hamster ear model. Pharm Res 12:1184–1188, 1995.
Skin Transport 195
324. Wurster DE, Munies R. Investigation of some factors influencing percutaneous ab-
sorption. III. Absorption of methyl ethyl ketone. J Pharm Sci 54:1281–1284, 1965.
325. Cross SE, Pugh WJ, Hadgraft J, Roberts MS. Probing the effect of vehicles on topical
delivery: understanding the basic relationship between solvent and solute penetration
using silicone membranes. Pharm Res (in press).
326. Bucks DAW. Prediction of percutaneous absorption. PhD dissertation. University of
California, San Francisco, 1980.
327. Bucks D, Guy R, Maibach H. Effect of occlusion. In: Bronaugh RL, Maibach H, eds.
In Vitro Percutaneous Absorption: Principles, Fundamental and Applications. Boston:
CRC Press, pp 85–114, 1991.
328. Jiang R, Roberts, MS, Prankerd R, Benson HA. Percutaneous absorption of sunscreen
agents from liquid paraffin: self-association of octyl salicylate and effects on skin flux.
J Pharm Sci 86:791–796, 1997.
329. Roberts MS. Percutaneous absorption of phenolic compounds. PhD dissertation. Uni-
versity of Sydney, 1976.
330. Jiang R, Benson HA, Cross SE, Roberts MS. In vitro human epidermal and polyeth-
ylene membrane penetration and retention of the sunscreen benzophenone-3 from a
range of solvents. Pharm Res 15:1863–1868, 1998.
331. Hadgraft J. Recent developments in topical and transdermal delivery. Eur J Metab
Pharmacokinet 21:165–173, 1996.
332. Shaw JE, Prevo ME, Amkraut AA. Testing of controlled-release transdermal dosage
forms. Product development and clinical trials. Arch Dermatol 123:1548–1556, 1987.
5
Methods for Studying
Percutaneous Absorption
KEITH R. BRAIN, KENNETH A. WALTERS, and ADAM C. WATKINSON*

I. INTRODUCTION
There is an increasing demand for data describing the rate, degree, and route of
penetration of compounds across human skin. First, there is a requirement to optimize
the delivery of dermatological drugs into various skin strata for maximum therapeutic
effect. Second, the transdermal and topical routes have become popular alternatives
to more traditional methods of drug delivery. A third stimulus has been the toxico-
logical and risk assessment implications of the everyday use of a wide range of
potentially harmful materials in the agrochemical, chemical, cosmetic, household,
and pharmaceutical sectors. This has been driven largely by regulatory and safety
bodies and a perceived need for improved data on the permeability of the skin to
xenobiotics. For example, the U.S. Environmental Protection Agency (EPA) is cur-
rently addressing the issue of the dermal absorption testing of 80 compounds des-
ignated by the Occupational Safety and Health Administration (OHSA) and the In-
teragency Testing Committee (ITC) as worthy of particular interest (1,2).
Although drug delivery and toxicological considerations are perhaps the most
important factors, it is also clear that increasing the overall database on percutaneous
permeation will enhance our understanding of the mechanistics of this process. The
need for relevant data, produced under reproducible and reliable conditions, has led
to an increase in both the development and the standardization of in vitro and in
vivo test procedures. There have been numerous recommendations on in vitro and
197
198 Brain et al.
Table 1 Skin Permeation Guidelines and Protocols
Organization
Date or author Type Ref.
1987 FDA/AAPS Guideline 3

1989 FDA/AAPS Guideline 261
1991 Bronaugh and Collier Protocol 406
1993 ECETOC Protocol 5
1993 COLIPA Protocol 6
1996 EPA Protocol 1
1996 ECVAM Protocol 4
in vivo methodologies, and many of these have been collated as guidelines by both
regulatory bodies and committees of interested parties. Perhaps the most widely
known of these guidelines are those produced following Food and Drug Adminis-
tration and American Association of Pharmaceutical Science (FDA/AAPS) workshop
on the performance of in vitro skin penetration studies (3). However, more recent
publications also contain useful information. For example, see the European Centre
for the Validation of Alternative Methods (ECVAM) workshop report on Methods
for Assessing Percutaneous Absorption (4); the documents from the European Centre
for Ecotoxicology and Toxicology of Chemicals (5); and those from the European
Cosmetic Toiletry and Perfumery Association (6) (Table 1).
II. IN VITRO METHODOLOGY

In vitro techniques to assess skin penetration and permeation are used extensively
in industry and academia. Although at present, in vitro permeation data are not a
requirement of regulatory bodies, there is an increasing trend for such data to be
submitted, either alone or together with in vivo data. In some ways, in vitro tech-
niques have advantages over in vivo testing. For example, permeation through the
skin is measured directly in vitro, for which sampling is carried out immediately
below the skin surface. This contrasts with most in vivo methods, which rely on the
measurement of systemic (or at least nonlocal) levels of permeant. Some form of in
vitro diffusion cell experiment is often the most appropriate method for assessment
of percutaneous penetration in a developmental drug-delivery (transdermal or topical)
program or in a dermal toxicology screen.
The following consideration of in vitro determination of dermal absorption and
distribution combines published guidelines and the scientific literature with our own
practical experience, and aims to provide the reader with an appreciation of the
experimental choices and limitations.
A. In Vitro Skin Diffusion Cells

Most common methods for evaluation of in vitro skin penetration use diffusion cells,
and there is a plethora of literature on the satisfactory performance of such experi-
ments. The major advantage of in vitro investigations is that the experimental con-
Studying Percutaneous Absorption 199
ditions can be controlled precisely, such that the only variables are the skin and the
test material. Although a potential disadvantage is that little information on the me-
tabolism, distribution, and effects of blood flow on permeation can be obtained, it
has been reported that such procedures were more effective than several other meth-
ods for the assessment of differential delivery of hydrocortisone from commercial
formulations (7).
It is essential to consider the ultimate use of generated data when developing
experimental protocols. Routinely, simple mathematical models, which are based on
certain assumptions or boundary conditions, are applied to experimental data. The
most commonly used solutions to diffusion equations that are applied to the in vitro
situation make the following assumptions:
1. The receptor phase is a perfect sink.
2. Depletion of the donor phase is negligible.
3. The membrane is a homogeneous slab.
None of these assumptions is wholly true in practice, and the potential signif-
icance of these imperfections must not be overlooked. Careful experimental design
can be used to achieve a close approximation to reality, and the following section
discusses how this can be achieved in practice.
1. Diffusion Cell Design
In vitro systems range in complexity from a simple two-compartment ‘‘static’’ dif-
fusion cell (8) to multijacketed ‘‘flow-through’’ cells (9) (Fig. 1). Construction ma-
terials must be inert, and glass is most common, although Teflon and stainless steel
(10) are also used. Excised skin is always mounted as a barrier between a donor
chamber and a receptor chamber, and the amount of compound permeating from the
donor to the receptor side is determined as a function of time. Efficient mixing of
the receptor phase (and sometimes the donor phase) is essential, and sample removal
should be simple. Neither of these processes should interfere with diffusion of the
permeant. Comprehensive reviews on diffusion cell design are available (11–13).
Continuous agitation of the receptor medium, sampling from the bulk liquid rather
than the side arm, and accurate replenishment after sampling, are important practical
considerations. It is essential that air bubbles are not introduced below the membrane
during sampling.
Static diffusion cells are usually of the upright (‘‘Franz’’) or side-by-side type,
with receptor chamber volumes of about 2–10 mL and surface areas of exposed
membranes of near 0.2–2 cm2. Cell dimensions should be accurately measured, and
precise values should be used in subsequent calculations, with due attention to an-
alyte dilution resulting from sampling and replenishment. The main difference in the
application of these two static cell types is that side-by-side cells can be used for
the measurement of permeation from one stirred solution, through a membrane, and
into another stirred solution. This is of particular advantage when examining flux
from saturated solutions in the presence of excess solid if accumulation of solid on
the membrane surface must be prevented. This type of cell can also be modified to
allow the absorption of permeants in the vapor phase. For example, volatile material
may be retained in a small depression in the donor chamber so that the membrane
is exposed to only the permeant in the gaseous state. Upright cells are particularly
useful for studying absorption from semisolid formulations spread on the membrane
200
surface and are optimal for simulating in vivo performance. The donor compartments
can be capped to provide occlusive conditions, or left open, according to the objec-
tives of the particular study.
Flow-through cells can be useful when the permeant has a very low solubility
in the receptor medium, and designs are continuously improving (14). Sink condi-
tions are maximized as the fluid is continually replaced using a suitable pump (at a
rate of about 1.5 mL/h) (15). However, the dilution produced by the continuous flow
can raise problems with analytical sensitivity, particularly if the permeation is low.
Flow-through and static systems have produced equivalent results (16,17). Auto-
mated flow-through systems can allow unattended sampling, and commercial systems
are available. However, these have significant cost implications and are generally
limited to small numbers of cells (⬃12). For example, Moody (18) has described an
automated in vitro dermal absorption (AIVDA) system using only 0.07 cm2 of skin
per cell, held by autosampler vial inserts an autosampler carousel, with samples taken
robotically. Although such methods are undoubtedly elegant, adequate validation of
automated methods is essential.
Standard upright static diffusion cells, which offer a simple, low-cost, and very
versatile system, can be employed on a large scale (⬃144 cells) and adapted to meet
the particular requirements of a wide range of studies.
To summarize, a well-designed skin diffusion cell should
1. Be inert
2. Be robust and easy to handle
3. Allow the use of membranes of different thicknesses
4. Provide thorough mixing of the receptor chamber contents
5. Ensure intimate contact between membrane and receptor phase
6. Be maintainable at constant temperature
7. Have precisely calibrated volumes and diffusional areas
8. Maintain membrane integrity
9. Provide easy sampling and replenishment of receptor phase
10. Be available at reasonable cost
2. Receptor Chamber and Medium
Receptor chamber dimensions are constrained by the conflicting requirements of
guaranteeing that the receptor phase can act as a sink, while ensuring that sample
dilution does not preclude analysis. A large receptor volume may ensure sink con-
ditions, but will reduce analytical sensitivity unless large samples can be taken and
subsequently concentrated. Concentration of permeant in an aqueous receptor phase
may be possible by lyophilization, or by techniques such as solid-phase extraction.
The ideal receptor phase provides an accurate simulation of the conditions
pertaining to in vivo permeation of the test compound. As a general rule the con-
<
Figure 1 Basic diffusion cell designs: static horizontal cells may be jacketed (as in the
Franz-type) or unjacketed (and temperature-controlled using a water bath or heating block).
Flow-through cells usually have a small receptor chamber to maximize mixing. Side-by-side
cells are used mainly for solution vehicles.
202 Brain et al.
centration of the permeant in the receptor fluid should not be allowed to exceed
approximately 10% of saturation solubility (3). Excessive receptor-phase concentra-
tion can lead to a decrease in the rate of absorption, which may result in an under-
estimate of bioavailability. The most commonly used receptor fluid is pH 7.4 phos-
phate-buffered saline (PBS), although this is not always the most appropriate
material. It has been postulated that if a compound has a water solubility of less
than about 10 ␮g/mL, then a wholly aqueous receptor phase is unsuitable, and the
addition of solubilizers becomes necessary (19).
Receptor fluids described in the literature range from water alone to isotonic
phosphate buffers containing albumin and preservatives. Albumin increases solubility
of the permeant (20), whereas preservatives inhibit microbial growth in the receptor
fluid. Microbial growth can produce problems by partitioning of the permeant into,
or metabolism of the permeant by, the microbes. One particularly useful fluid is 25%
(v/v) aqueous ethanol, which provides a reasonable ‘‘sink’’ for many permeants,
while removing the need for other antimicrobial constituents. Other examples of
modified receptor phases include 1.5–20% Volpo N20 (see later), rabbit serum, 3%
bovine serum albumin, 50% aqueous methanol, 1.5–6% Triton-X100, and 6%
Poloxamer 188 (15). It has been suggested that there is a need for formal protocols
to determine the suitability of receptor-phase composition (21).
It is important to recognize the possibility that solubilizers may interfere with
the barrier function of the skin itself. Bronaugh (19) examined several commonly
used receptor phases and made several recommendations. One of the most useful
receptor phases was an aqueous solution containing 6% Volpo N20 (Croda Inc.), a
nonionic polyethylene glycol(PEG)-20-oleyl ether surfactant. This did not influence
the flux of either water or urea across rat skin, when compared with normal saline,
suggesting that Volpo N20 did not disrupt the barrier function of rat skin to hydro-
philic compounds. Rat skin usually responds to penetration enhancers to a greater
degree than human skin and thus one would predict that Volpo N20 should have a
negligible effect on the flux of hydrophilic compounds across human skin. Com-
mercial Volpo N20 is a mixture of PEG-oleyl derivatives that averages at PEG-20,
and it has been our experience, and that of others (22), that this diversity of com-
ponents can produce complications in sample analysis.
The problem of very lipophilic permeants has been addressed by the use of
nonaqueous (23) and nonliquid receptor media. Sheets of silicone rubber (0.02-in.
thick) were used to collect pesticides with low water solubilities, which were then
desorbed from the rubber with an appropriate solvent and subsequently analyzed
(24). Other solutions have included the use of flowing gaseous receptor phases for
volatile permeants (25).
It is also important to appreciate that the pH of an aqueous buffered receptor
solution may markedly affect the apparent ‘‘flux’’ of a permeating weakly ionizable
compound. The pH of the hydrophilic viable epidermal layers may be ‘‘altered’’
by the receptor solution, and this can theoretically result in modulation of the par-
titioning tendencies of ionizable species. This was well illustrated experimentally by
Kou et al. (26), who determined the permeation of a weak acid and a weak base
(nicardipine) through human skin into receptor solutions of varying pH (Fig. 2).
Quite clearly the data show a receiver fluid pH dependency on flux, and the authors
caution against indiscriminate use of nonphysiological receptor pHs in diffusion
experiments.
Figure 2 The effect of receptor fluid pH on the flux of a weak acid (pKa 2.60) and a weak
base (pKa 5.68) across human skin in vitro: skin permeation of the weak acid increased with
increasing pH, whereas that for the base decreased. (From Ref. 26.)
3. Selection, Variation, and Preparation of Skin Membranes

A major potential variant in the design of in vitro skin diffusion experiments is the
nature of the skin membrane. Animal skins are widely used as substitutes for human
skin (27–29), primarily owing to difficulties in obtaining human tissue. The suit-
ability of animal skin as a model for human skin is discussed later in this chapter
(see Sec. III.A) and elsewhere (30).
a. Intra- and Intersubject Variability. Few studies have addressed the issues
of intra- and intersample variability in human skin permeability, because of the few
sufficiently large datasets available for accurate statistical evaluation. There are at
least two issues in this area that need to be addressed. The first is the degree and
nature of the distribution patterns of skin permeabilities in vitro. This was addressed
to some extent by Southwell et al. (31), who investigated both in vitro and in vivo
variation in the permeability of human skin between different specimens (interspec-
imen) and the same specimens (intraspecimen). From the permeation characteristics
for a series of compounds, they concluded that in vitro interspecimen variation was
66 ⫾ 25% and intraspecimen variation 43 ⫾ 25%. The pattern in vivo was similar,
although the overall level of variation was somewhat smaller. This analysis has been
extended by assessing the statistical distribution of human skin permeabilities. Kast-
ings et al. (32) and Cornwell and Barry (33) concluded that there was evidence that
the permeability of human skin in vitro is lognormally, rather than normally distrib-
204 Brain et al.
uted. Williams et al. (34) examined the permeation of 5-fluorouracil (644 determi-
nations from 71 specimens) and estradiol (221 determinations from 28 specimens)
through human abdominal skin. Here, where site variability was excluded, the data
were lognormally distributed. A lognormal distribution implies that the use of normal
gaussian statistics is inappropriate, and that use of geometric (rather than arithmetic)
means should be considered.
b. Age and Sex Differences. The questions of how age and sex affect the
permeability of human skin have been rather poorly addressed to date. Some studies
have concluded that, in vitro in humans, there was no discernible dependence of skin
permeability on age, sex, or storage conditions (34–39). The literature on the effect
of age and sex on percutaneous absorption in other species, for which more data is
available, is confused. For example, it was reported (40) that the dermal absorption
of certain marker compounds was lower in older rats, but that, for different com-
pounds, the reverse was true (41) or indeed, that age did not influence dermal ab-
sorption in rats (42). Hairless mouse skin permeation generally increases with age,
corresponding to the single-hair cycle, but decreases with a return to the hairless
state (43,44).
The effect of age on percutaneous absorption has been examined in vivo in
humans, with variable results. It was postulated (45) that reduced hydration levels
and lipid content of older skin may be responsible for a demonstrated reduction in
skin permeability if the permeants were hydrophilic (no reduction was seen for model
hydrophobic compounds) (Table 2). The reduced absorption of benzoic acid dem-
onstrated in the elderly (46) was in line with this suggestion, but not the reduction
in absorption of testosterone (lipophilic) (47), or lack of change in the absorption of
methyl nicotinate (more hydrophilic) (48), with age. A number of potential physio-
logical changes that may be responsible for age-related alterations in skin perme-
ability have been suggested. These include a noted increase in the size of individual
stratum corneum corneocytes throughout life, increased dehydration of the outer
layers of the stratum corneum with age, decreased epidermal turnover, and decreased
microvascular clearance (49). The issue of age-related variability, however, is far
Table 2 Age-Related Differences in Percutaneous Absorption
Applied dose permeated over

7 days (%)b
Permeant log K a 22–40 yr 65 yr
Testosterone 3.32 19.0 ⫾ 4.4 16.6 ⫾ 2.5

Estradiol 2.49 7.1 ⫾ 1.1 5.4 ⫾ 0.4
Hydrocortisone 1.61 1.5 ⫾ 0.6 0.54 ⫾ 0.15
Benzoic acid 1.83 36.2 ⫾ 4.6 19.5 ⫾ 1.6
Acetylsalicylic acid 1.26 31.2 ⫾ 7.3 13.6 ⫾ 1.9
Caffeine 0.01 48.2 ⫾ 4.1 25.2 ⫾ 4.8
a
Octanol/water partition coefficient.
b
Compounds (4 ␮g/cm2) were applied in 20 ␮L acetone to the ventral forearm
(n = 3–8).
Source: Ref. 45.
from resolved. If the variability of in vivo data is actually slightly lower than that
of in vitro data (31) then the small increase in variability in vitro may make any
such differences indiscernible.
c. Racial Differences. Several authors have documented differences between
the permeabilities of skin based on its racial origin. White skin is slightly more
permeable than black skin (50–55), which correlates with observations that black
skin has both more cell layers within the stratum corneum (SC) (56) and a higher
lipid content (57). A recent study (52) of white, Hispanic, black, and Asian skin
ranked them in order of permeability to methyl nicotinate as black < Asian < white
< Hispanic. It has been reported (58,59) that the corneocytes of black, white, and
Oriental skin are of a similar size, but that there are differences in spontaneous
desquamation. It has also been reported (60,61) that there may be differences between
black and white skin in microvascular reactivity, following dermal application of
vasoactive agents, but this is possibly due to subjective, rather than objective, mea-
surement (48). No differences in the permeation of water through black or white
skin in vitro were observed (35) and, similarly, there was no racial difference in the
in vivo percutaneous absorption of diflorasone diacetate (62). More recently, Lotte
et al. (63) have determined the penetration and permeation of several compounds
into (skin-stripping; see Sec. II.B.1) and through (24-h urinary excretion) Asian,
black, and white skin. There were no statistical differences in penetration or per-
meation of benzoic acid, caffeine, or acetylsalicylic acid among the races (Table 3).
These equivocal findings highlight the necessity for further systematic research in
this area.
d. Storage Conditions. It is unclear whether the proposed lognormal distri-
bution in vitro is an experimental artifact caused by excision and isolation from the
body, or by storage conditions before use (e.g., freezing). Some authors concluded
that freezing had no measurable effect on permeability (36,37). Yazdanian (64) re-
Table 3 Race-Related Differences in Percutaneous Absorption
Amount of permeant recovered

(nmol/cm2)
Stratum corneum
Permeant Race Urine at 24 h at 30 mina
Benzoic acid Caucasian 9.0 ⫾ 1.5 6.8 ⫾ 1.0

Black 6.4 ⫾ 0.9 6.1 ⫾ 1.0
Asian 9.7 ⫾ 1.2 8.1 ⫾ 1.5
Caffeine Caucasian 5.9 ⫾ 0.6 5.5 ⫾ 0.6
Black 4.5 ⫾ 1.0 5.8 ⫾ 1.0
Asian 5.2 ⫾ 0.8 6.1 ⫾ 0.9
Acetylsalicylic acid Caucasian 6.2 ⫾ 1.9 11.9 ⫾ 1.9
Black 4.7 ⫾ 0.9 9.0 ⫾ 1.7
Asian 5.4 ⫾ 1.7 10.1 ⫾ 1.7
a
Amount in stratum corneum determined by tape-stripping (n = 6–9).
Source: Ref. 63.
206 Brain et al.
ported an effect for cattle, although there was no general pattern of differences be-
tween frozen and fresh skin. Wester et al. (65) have cautioned against the use of
frozen, stored human skin for studies in which cutaneous metabolism may be a
contributing factor. It is, however, important to appreciate that the state of hydration
of the tissue before freezing may influence subsequent permeation characteristics
(66). As a general rule tissues should not be hydrated when placed into frozen
storage.
e. Anatomical Site Variations. Perhaps the clearest data available on varia-
tion in skin permeability deals with anatomical site-to-site variation (46,67–75). Site-
to-site variation of skin permeability has been examined using the tape-stripping
method (71,72) and correlated with corneocyte diameter (46) and, hence, diffusional
pathlength. Although, in practice, skin permeation of compounds follows a different
pattern in different skin regions, it is generally agreed that some body sites (the head
and genital region) are uniformly more permeable than others (extremities). For
example, transepidermal water loss and skin permeation of benzoic acid, caffeine,
and acetylsalicylic acid decreased in the order forehead > postauricular > abdomen
> arm (74). Similarly, the permeation of hydrocortisone decreased in the order scro-
tum > jaw > forehead > scalp > back > forearm > palm (Fig. 3) (67,6), and abdomen
was more permeable to methyl salicylate than either the arms or feet (75).
For in vitro permeation studies, it is advantageous for donor skin to be from
anatomical sites relevant to the objectives of the study, although this may be logis-
tically impossible. Because it is difficult to reach any definitive conclusions about
the variance and distribution of skin permeabilities, the best advice is that experi-
menters should be aware of the possibilities and test their data appropriately. An
adequate number of replicates, with skin from different donors or sites spread evenly
throughout all test groups, should be used.
f. Membrane Preparation. Different methods can be used to prepare human
skin. The membrane is one of the following:
1. Full-thickness skin, incorporating the SC, viable epidermis, and dermis
2. Dermatomed skin, in which the lower dermis has been removed
3. Epidermal membranes, comprising the viable epidermis and the SC (pre-
pared by heat separation)
4. SC alone (prepared from step 3 by enzyme treatment)
The most suitable type of tissue is dependent on the nature of the permeant.
The environment of skin in vivo differs somewhat from that in vitro. In vivo the
continuously perfused subcutaneous vasculature, which penetrates the dermis to a
significant degree, can rapidly remove permeants reaching the epidermal–dermal
junction. These vessels, if still present, are not perfused in simple in vitro models.
In vitro the relatively aqueous environment of the dermis will inhibit the penetration
of lipophilic compounds, whereas in vivo this barrier is circumvented by the capillary
bed. Hence, the use of dermatomed, epidermal, or SC membranes is more appropriate
for particularly lipophilic permeants.
Other considerations may justify the use of epidermal membranes, even where
the dermis does not present an artificial barrier to a permeant. For example, if a
study involves an assessment of the skin content of permeant, it is much easier to
extract or solubilize epidermal membranes, or SC, than full-thickness skin. Con-
Figure 3 Regional variation in the in vivo absorption of hydrocortisone through human

skin: absorption was normalized to that through the ventral forearm. Hydrocortisone was
applied in acetone to a marked site. (From Ref. 67.)
versely, if an experiment includes tape-stripping, it may be easier to perform on full-

thickness skin, rather than epidermal membranes. However, the latter have been
successfully used to correlate in vivo and in vitro tape-stripping (76).
The preparation of epidermal membranes and SC is time-consuming, and the
necessary processing increases the possibility of damage to the skin membrane. Care-
ful consideration of the most appropriate type of skin preparation is required, and
this should address the physicochemical nature of the penetrating species, the data
required, tissue availability, and the time scales involved. With animal skin, full-
thickness membranes are usually used, because it is difficult to isolate intact epider-
mis or SC owing to the presence of numerous hair follicles, which may also com-
promise dermatomed tissue.
Dermatomed skin can be prepared in vitro as follows: most of the subcutaneous
fat is removed, leaving only a small quantity in place; the skin is placed, dermal
side down, onto a metal plate; the residual fat adheres to the metal; a thin sheet of
plastic is placed over the SC to protect it before a second metal plate is applied; the
two plates are clamped together and cooled to ⫺20⬚C until completely frozen (about
208 Brain et al.
1–3 h); on removal from the freezer, the upper plate is gently warmed to ease
removal from the SC; the plastic sheet is removed and a dermatome used to remove
strips of skin of the desired thickness (usually ranging from ⬃200–600 ␮m). Care
must be taken in the preparation of dermatomed skin to ensure that damage to hair
follicles is minimized because, if these are severed, erroneously high penetration
figures result. The use of ‘‘hairless’ varieties of animal and relatively ‘‘hair-free’’
types of human skin may reduce, but not eliminate, this problem.
For human skin, the separation of the dermis from the epidermis (SC and viable
epidermis) is a relatively simple technique (35,77). First, the subcutaneous fat is
removed by blunt dissection. The full-thickness skin membrane is then totally im-
mersed in water at 60⬚C for about 45 s. Following removal from the water, the skin
is pinned, dermal side down, to a dissecting board and the epidermis gently peeled
back using a pair of blunt curved forceps. The epidermal membrane can next be
floated onto warm water and taken up on a support membrane (membrane or paper
filter). It is then ready to be mounted in a diffusion cell. Alternatively, a microwave
technique has been proposed for the separation of epidermis from the dermis (78).
To isolate the SC from epidermal membranes, the latter are placed in trypsin solution
(0.0001%), incubated at 37⬚C for 12 h, rubbed (with a cotton bud) to remove the
epidermal cells, rinsed in distilled water, and air-dried on a surface from which they
can be easily removed (79).
4. Permeant and Application Technique
The manner in which a substance is applied to the skin surface can be a major
determinant of its subsequent absorption. Several factors must be considered in se-
lecting a suitable application procedure, including the nature of the vehicle, the
permeant concentration, the amount of vehicle applied, the mechanism of application,
the exposure time, and the method for removing an applied vehicle (if required).
Many of these issues may be intrinsic to the purpose of the study. For example, risk
assessment involving the study of the skin penetration of an ingredient in a cosmetic
or agrochemical formulation should be performed with the material in the formula-
tion as it is marketed or used, and with a regimen that mimics, as closely as possible,
the ‘‘in use’’ situation (e.g., 80,81).
a. Application Method. There are two basic approaches to applying sub-
stances to the skin. Infinite-dose techniques involve application of sufficient permeant
to make any changes in donor concentration during the experimental timeframe,
caused by diffusion or evaporation, negligible (i.e., the dose is effectively infinite).
This is desirable if the experimental objectives include calculation of diffusional
parameters, such as permeability coefficients; or for investigation of mechanisms of
penetration enhancement. Finite dose techniques (82), designed to model in-use con-
ditions, involve application of a dose that may show marked depletion during an
experiment. Depletion occurs where the proportion of permeant entering the mem-
brane is large, relative to the amount applied. Alternatively, the permeant may be
removed from the skin surface during, for example, the simulation of a rinsing or
washing procedure. With finite dosing the permeation profile may exhibit the char-
acteristic plateauing effect that accompanies donor depletion (Fig. 4). The finite dose
technique may involve application of permeants or enhancers in small volumes of
volatile solvents (e.g., acetone or ethanol). This allows assessment of the gross effects
Figure 4 Sample cumulative permeation patterns following finite- and infinite-dosing reg-
imens: with infinite dose, permeation normally reaches a steady-state flux region, from which
it is possible to calculate permeability coefficients and diffusional lag times. In finite dosing
the permeation profile normally exhibits a plateau effect as a result of donor depletion.
of enhancers, but results are more difficult to interpret mechanistically. Direct com-
parisons of finite- and infinite-dose applications are relatively rare, but the predicted
effects have been investigated (83,84).
b. Dose Level. There are conflicting reports on the effect of dose level on
the degree of permeation through skin. For example, increased dosage did not pro-
duce proportional increases in flux of ibuprofen or flurbiprofen across human skin
in vitro when deposited as a thin film from acetone (85). Three concentrations of
each drug were applied in only 50 ␮L, and the authors concluded that as the acetone
evaporated from the skin surface the thermodynamic activity of the drugs increased
until saturation was reached, at which concentration maximum flux would be ex-
pected. When all of the solvent had evaporated, leaving a plug of solid drug with
poor dissolution properties, the flux always dropped. The effect of volume on the
permeation of minoxidil from ethanol was linear between 10 and 50 ␮L (86), but it
was concluded that it was the total drug loading, rather than the application volume,
that was important. Note that this is only true for a volatile vehicle that evaporates
rapidly from the skin surface. In practice, the actual skin permeation of volatile
compounds, such as dimethylnitrosamine (Fig. 5) and 2-phenoxyethanol, is signifi-
cantly reduced by evaporation (80,87). Other literature covers the effects of formu-
lation, application time, dose variation, and occlusion on ibuprofen permeation (88),
210 Brain et al.
Figure 5 Permeation profile for dimethylnitrosamine (DMN) across human skin in vitro:
DMN was applied at finite dose levels in an oil-in-water emulsion vehicle. Note that per-
meation was significantly reduced by evaporation following a 6-h exposure. (From Ref. 80.)
the effect of application formulation (89), theories of finite dosing (90,91), thickness
of the application vehicle (92,93), and contact time effects (94).
There are several published recommendations on both the expression of dose
levels and the specific quantities involved. The FDA/AAPS guidelines (3) suggested
a universal application weight of approximately 5 mg/cm2 of formulation. COLIPA
(6) proposed 5 ␮L/cm2 for liquid formulations, and 2 mg/cm2 for semisolid formu-
lations (or 5 mg/cm2 if these are being compared with a liquid). Distribution of
accurately measured amounts of such small quantities of semisolid formulations as
an even film over the surface of skin membranes presents considerable practical
challenges (93). Pretreatment of the skin before mounting in a diffusion cell may be
easier than treating premounted membranes. Semisolid materials can be applied and
spread with a small preweighed formulation-coated spatula. The precise weight ap-
plied is determined by difference, and all test materials should be applied by the
same operator. Evaluation of spray-on formulations raises even more complications.
Practical evaluation of the effects of enhancers is best made by their incorporation
into appropriate formulations, rather than the more common procedure of pretreat-
ment of the skin, which does not model the in-use scenario.
The issue of dosage levels is of particular interest, although there is a dearth
of appropriate nonclinical data concerning dose–response curves for topical formu-
lations. Numerous topical formulations have shown clinical equivalence despite con-
taining different concentrations of the same drug (e.g., 95; see also Chap. 8). This
can be rationalized if each of the formulations were saturated with drug because,
provided that the excipients had no effect on the skin, similar flux would be predicted
from each of them (96; see also Chap. 6). Alternatively, if all of the formulations
were very poorly bioavailable, then increases in drug content may not have any
significant effect on clinical efficacy. The dosage form, level, and exposure time
must be carefully considered to ensure that a study will produce appropriate data,
particularly in the view of any regulatory bodies involved in subsequent inter-
pretation.
c. Use of Radiolabeled Permeants. The use of radiolabeled permeants is
attractive, as this greatly simplifies sample analysis, particularly for mass balance
experiments. Guidelines for the choice and use of radiolabels in permeation exper-
iments are provided by COLIPA (6); although not all problems are addressed. It is
essential that radiolabeled material is homogeneously mixed with nonlabeled per-
meant. Where a radiolabel is used to evaluate a formulated product it is prudent to
incorporate the radiolabel at the same stage of manufacture as the cold permeant.
This may be possible for formulations for which exact constituents and method of
manufacture can be reproduced. A less rigorous alternative is to manufacture a small
quantity of labeled formulation and blend this with a larger quantity of cold for-
mulation. If the test formulation is a marketed product, for which exact manufactur-
ing process and ingredients are unknown, incorporation of label in these ways is
impossible, limiting the options to spiking the market formulation with radiolabel
dissolved in the most appropriate solvent. A test for homogeneity of distribution of
the label in the final formulation is always essential. It must be appreciated that in
a complex, formulation containing discrete phases, homogeneous distribution at a
(sub)microscopic level may not be achievable by spiking and that solvent used to
incorporate a label may alter the performance of a formulation.
The greatest problem in the use of radiolabels is the lack of specificity. It is
essential to ensure that the integrity of the permeant is maintained during the ex-
periment and that counts attributed to permeation of the target molecule are not due
to impurities, degradation, exchange, or metabolism. This can be achieved by chro-
matographic separation of the analyte before counting, although this may remove the
major advantage of the radiolabeled approach.
5. The Permeation Experiment
a. Duration. Although some workers have extended the duration of experi-
ments to 120 h (85), it is recommended that they be restricted to 24 (6) or 48 h (4).
It has been suggested that 48 h may be too short to establish a steady-state flux from
an infinite dose, leading to misinterpretation of the data (97), and the ECVAM (4)
report suggests experiments may be extended to 72 h (in the presence of antimicro-
bial agents) in such instances. From electrical resistance measurements and perme-
ation parameters (Table 4), human epidermal membranes were shown to retain in-
tegrity for up to 5 days (98), provided they were supported on suitable non–
rate-limiting membranes (e.g., Millipore GSWP filters). In the absence of the filter
membrane support, tissue integrity was compromised by the physical stress accom-
panying sample withdrawal and skin washing. Investigators should, however, be
aware of possible barrier degradation over extended time frames.
b. Sample Interval. Sample intervals should be of an appropriate frequency
to allow realistic assessment of such parameters as lag time and steady-state, or
pseudo–steady-state, flux (if possible). For a compound with unknown permeation
212 Brain et al.
Table 4 Human Epidermal Membrane Integrity During Prolonged Experiments Based on

Tissue Electrical Resistivity and Permeation of Mannitol
Unsupported membrane Supported membrane

Resistance Permeability Resistance Permeability
Time (h) (KOhm⭈cm2) (⫻108 cm/s) (KOhm⭈cm2) (⫻108 cm/s)
25 44 0.48 42 —
50 35 — 40 0.64
75 27 0.94 38 0.78
100 23 — 35 0.70
125 — — 37 0.76
150 — — 37 0.80
Source: Ref. 98.
characteristics, samples should ideally be taken at 2-h intervals for the duration of
the experiment. Early samples (1–4 h) may be important in identifying compromised
cells showing anomalously high early permeability.
c. Number of Replicates. Given the high intra- and intersubject variability
in human skin permeability (see Sec. II.A.3.a), a large number of replicates for each
dosage regimen is recommended. The most widely quoted recommendation for num-
bers of replicates in in vitro studies on human skin is 12 (3), and comparisons should
be matched. Fewer replicates may be employed, if cost, time, or skin availability is
a problem, provided that the limitations of replicate reduction are recognized. The
permeability characteristics of laboratory animal skin are, in general, more uniform
than those of human skin, and fewer replicates may be successfully employed.
d. Temperature. In vitro skin diffusion experiments are normally conducted
with a skin temperature of 32⬚C (the in vivo value). This is achieved by maintaining
the receptor solutions at 37⬚C, either by immersing cells in a water bath or by using
cell jackets perfused at the correct temperature. An apparent nonlinear dependence
of skin permeability on temperature was demonstrated for flurbiprofen (99). The skin
accumulation of flurbiprofen decreased with a rise of temperature in an in vivo
experiment, but not in an equivalent in vitro experiment, suggesting participation of
increased blood flow in vivo. Over the range of 22–90⬚C the in vitro permeability
coefficient for water through SC rose in a sigmoidal manner by a factor of about 70
(100). Similar increases in skin permeability have been shown for sodium lauryl
sulfate, nickel chloride (101), and ketoprofen (102). This temperature dependence is
probably a function of lipid fluidity. Skin lipids undergo major phase transitions
between 40⬚ and 70⬚C (103), which may explain the sigmoidal pattern.
e. Skin Integrity. Skin integrity can be addressed in a qualitative manner by
simple visual examination of specimens or, more quantitatively, by measurement of
transepidermal water loss (104), or by the flux of marker compounds, such as tritiated
water (32,35,84,105) or sucrose (80,81,106). The generally accepted permeability
coefficient for water diffusion through human skin is 1.5 ⫻ 10⫺3 cm/h, or less,
although an upper limit of 2.5 ⫻ 10⫺3 cm/h has also been used (35). Samples show-
ing particularly high permeability are often rejected as outliers with questionable
integrity, but may actually represent the real population spread if their distribution
is indeed lognormal (see earlier discussion).
f. Permeant Analysis. The quality of the data derived from any experiment
is ultimately dependent on the integrity of the analytical method employed, and all
aspects of the analytical procedure should be included in the overall experimental
design (107). The ideal analytical procedure provides accurate assessment of both
the quantity and nature of the material present at a given time, and the detection
limit of the method must be capable of producing data of practical significance.
Preliminary prediction from existing data or physicochemical modeling can give
‘‘ballpark’’ estimates of requirements, and potential routes of degradation and me-
tabolism should be taken into consideration. When no permeation is detected, this
should be reported as ‘‘less than’’ the detection limit and not ‘‘zero.’’ As replication
and multiple time point sampling are common, analysis should not be unnecessarily
complex, although the integrity of the method must be beyond doubt. High-perfor-
mance liquid chromatography (HPLC) is particularly useful as relatively large (⬃200
␮L) aqueous samples can often be handled without preliminary processing or con-
centration, although the inherent properties of the permeant can limit detection. Ul-
traviolet detection is commonly used, provided that the permeant has significant
absorption, and interference from extracted skin or formulation components can
cause problems, particularly at low wavelengths. Native fluorescence is less subject
to interference than absorption, but is also a less common phenomenon. Derivati-
zation may be necessary for detection, but is generally an undesirable complication
that increases both cost and variability. Investigations of permeation of commercial
materials consisting of a range of related compounds pose particular problems (108).
Interference can arise from leaching of material from the skin or components of the
test material or components of the diffusion system. The magnitude and composition
of skin leachate depends on the type of skin membrane and the time frame. It may
interfere directly with detection, or it may bind permeant, thereby making it un-
available to detection. These effects can be adequately investigated only by real-time
comparison with appropriate controls, including permeation from placebo vehicles,
and by confirming recovery of permeant from spiked samples.
Determination of permeants in tissue samples necessitates some form of ex-
traction or a solubilization process. When radiolabeled permeants are used, tissue
samples are routinely taken up in commercially available solubilizers. Such aggres-
sive products are often not applicable in other cases for which more traditional
extraction methods are required. Recovery of permeant from SC tape strips is less
demanding and can often be accomplished by vortexing and sonicating with rela-
tively small volumes of solvent.
6. Assessment of In Vitro Skin Metabolism
Metabolism of a xenobiotic is a detoxification and elimination process involving
formation of molecules that are more hydrophilic and easily excreted. It can be
summarized as a two-phase process. The first stage is the exposure or addition of
functional groups to form a primary product by, for example, the production of free
hydroxyl, carboxyl, or amino groups. The second step usually involves conjugation
with a polar molecule (e.g., glucuronic acid) to form hydrophilic compounds that
are readily excreted (109). It is widely established that there is potential for biotrans-
214 Brain et al.
formation of molecules within the skin (110,111). The nature of skin enzymes differs
quantitatively and qualitatively from those in the liver (112). In general, the activities
of many metabolic processes are much lower in skin than in liver (111,113,114)
although certain enzymes, such as N-acetytransferases and those involved in reduc-
tive processes, have demonstrated fairly high activity (115,116) (Table 5).
The specific activities (percentage of hepatic) of several phase-1 and phase-2
enzyme–substrate systems in skin have been assessed. Phase-1 conversions ranged
from 27% (for the cytochrome P450–7–pentoxyresorufin system) to less than 1%
(for the cytochrome P450–coumarin system). Phase-2 examples include the gluta-
thione S-transferase–cis-stilbene oxide system (49% of hepatic activity) and the
UDP-glucuronyltransferase–3-hydroxybenzo[a]pyrene system (0.6–2% of hepatic
activity) (117). Induction of phase-1 enzymes in skin has been studied using topically
applied classic enzyme inducers, such as polycyclic aromatic hydrocarbons (PAHs).
Many cutaneous enzymes respond substantially (e.g., cytochrome P450 metabolism
of retinoic acid is 500–1000 times higher in the presence of a PAH inducer).
Most metabolic investigations on skin have employed epidermal homogenates
(118–121) or epidermal cell cultures (122). These are useful for the study of enzyme
activity per se, but have little predictive value for the in vivo situation in which
permeants may not contact cellular systems (123). A more appropriate in vitro model
uses metabolically active intact skin mounted in a diffusion cell under conditions
that maintain viability. Bronaugh (123) emphasizes the importance of using fresh
skin. Full-thickness skin must be dermatomed (⬃ 200 ␮m) to ensure an adequate
supply of nutrients and oxygen (9). The receptor fluid is usually HEPES-buffered
Hanks’ balanced salt solution or Dulbecco’s modified PBS (124).
Table 5 Comparisons Between Specific Activities of Cutaneous

Enzymes and Hepatic Enzymes
Cutaneous
specific activity
Enzyme system Substrate (% hepatic)
Cytochrome P-450s Aldrin 0.4–2.0

Aminopyrine 1.0
Diphenyloxazole 2.0–3.0
Ethylmorphine 0.5
7-Pentoxyresorufin 20–27
Epoxide Hydrolases cis-Stilbene oxide 9–11
trans-Stilbene oxide 24–25
Styrene oxide 6.0
Glutathione transferases cis-Stilbene oxide 49
Styrene oxide 14
Glucuronosyltransferases Bilirubin 3.0
1-Naphthol 2–50
Sulfotransferases 1-Naphthol 10
Acetyltransferases p-Aminobenzoic acid 18
2-Aminofluorene 15
Source: Ref. 111.

These techniques have been used to investigate in vitro permeation and me-
tabolism of several compounds (114), including estradiol and testosterone (124);
acetyl ethyl tetramethyltetralin and butylated hydroxytoluene (113); benzoic acid, p-
aminobenzoic acid, and ethylaminobenzoate (125); benzo[a]pyrene and 7-ethoxy-
coumarin (126); azo dyes (127); butachlor (128); atrazine (129), and retinyl palmitate
(130). The use of cultured skin for metabolic work has also been investigated
(131,132).
B. Other In Vitro Methods for Studying Percutaneous Absorption

1. Skin-Stripping
Skin-stripping with adhesive tape is commonly used in vivo, and also in vitro. Tape-
stripping experiments are performed as follows: (a) a permeant is applied to the skin
surface for a fixed time period; (b) permeant remaining on the skin surface is re-
moved (where possible) by wiping or washing; (c) a succession of SC layers are
removed by sequential tape strips; (d) the permeant contents of the tape strips are
determined. These experiments evaluate how the concentration of a permeant applied
to the skin surface changes with depth into the SC. The shape of this concentration–
depth profile will be time-dependent and vary with permeant, according to the ra-
pidity, degree, and nature of uptake by the SC.
If steady-state diffusion is achieved, then the distribution of a permeant across
a homogeneous membrane should be represented by a linear decline from the outside
to the inside (133). Before steady state, the distribution of a permeant across a
homogeneous membrane will be represented by a series of concave curves converg-
ing to a straight-line as time progresses (133). Hence, in a tape-stripping experiment,
assuming that the SC is homogeneous and that the depth profiled with each tape
strip is the same, a linear concentration–strip number profile in the steady-state
region of any diffusion experiment would be expected. In practice, neither of these
assumptions generally appears to be correct, resulting in nonlinear profiles (Fig. 6).
A linear relation between the mass of SC removed and the number of tape
strips performed was reported in vivo (134,135) using 0.6-cm–diameter circles of
tape (Transpore; 3M; St. Paul, MN). However, Schaefer et al. (136) found an ex-
ponential relation between strip number and mass of skin removed in vivo and this
has been reproduced (76). An increase in structural integrity of SC with increasing
depth has been suggested (137,138), which has been correlated with increasing lipid
content (139) and modeled mathematically (140). Such data support a nonlinear
relation between strip number and mass of skin removed (141). If the external cells
of the horny layer are less well bound, then they will obviously be more easily
removed than deeper layers. Alternative techniques have also been applied to as-
sessment of the amount of SC removed by repetitive stripping and both strip protein
content analysis and spectrophotometric methods (141) showed nonlinear depen-
dencies on strip number.
If the membrane is not homogeneous, then distribution of a permeant within
the skin at steady state will not be a linear function of depth. The majority of both
in vivo and in vitro reports show the distribution of compounds within the skin to
be related to strip number in a nonlinear fashion. There is usually an approximate
exponential decay in the amount of permeant from the outside to the inside of the
SC [e.g., erythromycin (142), lanolin (143), fentanyl (144), alniditan (145), and hair-
216 Brain et al.
Figure 6 Typical pattern of permeant content in tape strips removed from the stratum
corneum following a fixed period of exposure. Note that most of the permeant is found in
the first few strips.
dyes (146)] (Fig. 7). However, ‘‘steady state’’ may not have been attained or the
nonlinear dependence of concentration on strip number may be a result of nonlinear
removal of SC.
There is little published data on the use and validation of in vivo and in vitro
skin-stripping techniques. Pershing et al. (134) claimed that although stripping skin
in vitro produced the same net loss in weight of SC (approximately 1 mg/cm2 over
ten strips), the pattern of skin removal was different in vitro from that in vivo. They
reported a linear relation between tape strip number and weight of SC removed in
vivo, but a nonlinear pattern in vitro when proportionately more was removed in the
first five tape strips than in vivo. In contrast, Trebilcock et al. (76) found similar
nonlinear relations between strip number and mass of SC removed in vivo and in
vitro and concluded that there was no significant difference between in vitro and in
vivo tape stripping for assessing skin distribution after percutaneous penetration.
The reasons for these disparate results are unclear, although a linear correlation
can be closely approximated by an exponential one, depending on the decay constant
(147). It is also probable that the particular tape used and precise experimental pro-
tocol, will affect SC removal. The tape used can affect the resultant concentration–
depth profile (148), as can the time of application of a particular vehicle (149).
Monitoring changes in SC concentration–depth profiles as a function of time requires
appropriate controls conducted for each time point so that results can be normalized.
Figure 7 Aciclovir content in tape strips removed from human stratum corneum following
24-h exposure to Zovirax Cream (5% aciclovir): The cream was applied to skin in vitro at a
dose of 5 mg/cm2 formulation. Data are mean ⫾ SE (n = 12).
Much remains to be achieved in skin tape-stripping technology (150), and method

validation will need to be rigorous.
2. Attenuated Total Reflectance–Fourier Transform Infrared
(ATR–FTIR) Spectroscopy
One of the most important uses of data describing diffusion through skin pertains to
the analysis of mechanisms of passive and enhanced transport. Successful solutions
to such mechanistic questions require that diffusion profiles are deconvoluted to
allow separate evaluation of the contributions arising from the alternative phenomena
of diffusion and solubility or partitioning. Attempts (151) to deconvolute these pro-
cesses in permeation through both synthetic and biological membranes by using
diffusion cells, achieved limited success. The problem of deconvoluting diffusion–
cell-derived data arises because of the need to demonstrate the attainment of steady
state. This is a prerequisite to calculation of flux and other permeation parameters,
but the common practice of application of linear regression to subjectively selected
data has been criticized (97). Similarly, assessment of lag times (hence, diffusion
coefficients) must also be questioned.
The ATR–FTIR technique is fairly straightforward: (a) a membrane is placed
in direct contact with the surface of a ZnSe ATR crystal, mounted on a spectrometer;
(b) a shallow trough is placed on top of the membrane; (c) the trough is sealed to
218 Brain et al.
the membrane (e.g., with petroleum jelly); (d) the solvent–solute system under study
is placed in the trough; (e) the spectrometer is linked to a computer equipped with
the appropriate software and FTIR spectra taken at the crystal–membrane interface
over a time period. The IR beam penetrates the membrane to a depth of only about
2–3 ␮m; therefore, only the interfacial region is probed. As permeant enters the
interfacial region there is an increase in the IR peak areas associated with the pen-
etrating species. The major limitation is that the permeant must have an IR absorption
in the transparent region of the membrane (e.g., compounds containing cyano- and
azido-groups or deuterated compounds). The data obtained is the concentration–time
profile at the crystal–membrane interface, which builds up at a rate related to the
diffusion coefficient and gradually plateaus at a level related to the solubility of the
permeant in the membrane under study. These two parameters can be estimated by
a nonlinear curve fit to the data by using a solution to Ficks’ laws of diffusion.
ATR–FTIR spectroscopy was originally used to study diffusion through simple
homogeneous synthetic systems, such as silicone membranes (152,153). The tech-
nique was applied to examination of diffusion into semisolids (154) and to the in-
vestigation of ethanol diffusion in glycerogelatin films (155). In conjunction with
bulk transport techniques, it was used to show that values of diffusion coefficients
(in synthetic membranes) calculated using the two methods were similar (156). The
methodology has been specifically refined (157,158) to permit its use with human
SC (159), and it has also been applied to the investigation of morphological differ-
ences between the upper and lower layers of the SC (160,161). Data from regular
diffusion cells and ATR–FTIR spectroscopy showed a high correlation (160,161)
and has been used to predict diffusional pathlengths in SC.
3. Isolated Perfused Tissue Models

Perfused tissue models use excised regions of skin—complete with their associated
microvasculature—immediately after sacrifice, and with continuous perfusion [e.g.,
with Krebs–Ringer buffer, glucose, and albumin aerated with oxygen and carbon
dioxide (162). Several variations on the theme involve both different species and
areas of skin. Isolated perfused porcine skin flaps (163–165) (Fig. 8), bovine udders
(166,167), and rabbit ears (168,169) have been used. The perfused pig-ear model
uses an isolated ear perfused with oxygenated blood from the same pig (170). These
techniques permit the investigation of the effect of local blood circulation on the
accumulation and removal of topically applied materials.
4. Artificial (Cultured) Skin

The technology behind the construction of human skin equivalents (artificial skin)
is derived predominantly from research into the treatment of burns. A classification
and evaluation of the numerous different types of human skin equivalents concluded
that the technique is limited to the reconstitution of the epidermis with a SC (171).
Such models have been used to investigate both cutaneous metabolic events
(131,132,172) and dermal irritation (173), with varying degrees of success. The use
of artificially cultured skin in permeation experiments is still of limited application
because the methodology is both expensive and not clearly predictive of in vivo
results (174–179). In vitro skin equivalents; per se, are generally approximately ten
times more permeable than human skin (180). On the other hand, reconstructed
epidermis transplanted onto a nude athymic mouse had a permeability similar to

normal human skin after 1 month (181).
5. Autoradiography
Autoradiographic techniques have been used extensively to visualize and quantify
penetration through, and distribution within, animal (182–187) and human (405)
skin. The general procedure is as follows: (a) a radiolabeled permeant is applied to
the skin surface; (b) after a suitable time, a skin sample is excised and microtomed
perpendicular or parallel to the surface to produce a section of the skin; (c) the
section is placed in contact with a photographic emulsion; (d) exposure to radiation
produces a latent image showing the pattern of distribution of the radiolabel within
the sample; (e) photographic development reveals the image; (f) evaluation of the
intensity of the image allows qualitative and quantitative evaluations of routes and
degrees of permeation. The technique is particularly useful in the assessment of
attempts to target drugs to specific regions of the skin. For example, compounds that
act against acne are targeted at the pilosebaceous unit (186).
6. Laser-Scanning Confocal Microscopy (LSCM)
A major problem with fluorescence visualization using a conventional microscope is
that the images are often blurred owing to light originating from out-of-focus sources
above and below the region of interest. LSCM allows clear visualization of fluores-
cent probes within tissue samples (188) because it collects light from only a small
in-focus volume so that contributions to the image from fluorescence above and
below the in-focus point are reduced. A confocal image is constructed using light
from a common plane within a structure to form an optical section. LSCM has been
used to examine skin samples that are sufficiently transparent, do not scatter light
strongly, and are relatively free from autofluorescence. There are no fixing processes
involved in sample preparation; therefore, artifact formation can be largely avoided.
Specimens can be rapidly prepared and examined, and permeation of fluorescent
probes can be monitored kinetically in the same sample (188). Optical-sectioning
methods have been used to examine the permeation routes of compounds across the
cornea (189,190), the skin (191–195), the buccal epithelium (196–198), and the
nasal mucosa (199). There are, however, some limitations to the technique. The
images produced are not absolutely real-time (1-4 s are needed to collect an image,
and several images must be averaged to gain a good quality picture), the laser ex-
citation wavelengths available will not excite all fluorophores, and problems of pho-
tobleaching and photochemistry can occur. Photobleaching has been utilized as a tool
to examine diffusion within controlled-release devices (200) and gels (201). A region
of fluorescent molecules within a gel is photobleached using a high-power laser and,
as unbleached molecules move into the bleached region, the fluorescence is restored.
The rate at which this restoration of fluorescence occurs is related to the diffusion
coefficient of the molecule in the matrix. The latter technique may have potential in
the study of diffusion in SC.
7. Model Membranes
A range of synthetic artificial membranes have been evaluated as models for per-
cutaneous penetration. These may offer the possibility of providing a less variable
system for formulation evaluation and for investigation of permeation phenomena.
220 Brain et al.
Studying Percutaneous Absorption
Figure 8 The isolated perfused porcine skin flap: (A) The surgical procedure involves raising a skin flap on the ventral surface and creating a
221
tube. The superficial epigastric artery is cannulated and the tubed flap is transferred to the perfusion chamber. (B) The perfusion chamber used to
maintain viability of the isolated skin flap: The chamber is temperature-, humidity-, and pressure-regulated. The perfusate is a modified Krebs–Ringer
bicarbonate buffer (pH 7.4) containing albumin and glucose. (Courtesy of Dr. J.E. Riviere.)
222 Brain et al.
The penetration of lipophilic compounds through solid, supported liquid membranes

containing a range of simple lipids (202,203), more complex mixtures (204), and
membranes containing Azone (205) has been examined using a rotating diffusion
cell (206,207). Liquid membranes are frequently unstable to excipients, severely
restricting their application to formulation evaluation. Polymers, such as polydi-
methylsiloxane (PDMS, Silastic) offer a nonporous, hydrophobic, relatively inert,
and reproducible barrier, which has been used to evaluate factors such as donor
concentration on permeation. For example, Pellett et al. (208) demonstrated enhanced
penetration from supersaturated solutions containing piroxicam. Combinations of dif-
ferent polymers, such as PDMS, poly-(2-hydroxyethylmethacrylate; pHEMA) and
cellulose acetate, have been used to try to model the combination of hydrophilic and
lipophilic domains present within the SC (209,210). Trilamellar cellulose acetate–
silicone (equilibrated with IPM) laminates have been used to compare delivery of
piroxicam from gel and cream formulations (211). A silicone–cellulose ester multi-
laminate was reported to reproduce SC permeation for four formulations containing
methyl nicotinate (212). A six-layer composite membrane, containing dodecanol in
a collodion matrix, was unusual in that it included a protein phase (210). Although
composite systems have been defined that may provide data with a good correlation
with actual SC values in specific cases, it appears that, in general, each system is
applicable only for comparison of the permeation of a small range of related com-
pounds. Such systems have an application in quality control of existing products,
under carefully defined conditions.
Lamellar sheets, which are more representative of the structure in the lipid
domains of the SC, can be constructed, by fusing liposome vesicles prepared from
a mixture of SC lipids, onto a support membrane (213). A good correlation (r =
0.880) with guinea pig SC was reported for the permeation of a range of compounds,
including nonsteroidal anti-inflammatory agents (NSAIDs), cyclobarbitol, and lig-
nocaine (214), and the effects of several penetration enhancers were reproduced
(215). Comparison of the permeation of certain NSAIDs through several different
model membrane systems has been compared (216). Structured lipid matrices pre-
pared from cholesterol, water, and fatty acids have been proposed as suitable models
for investigations of SC barrier function and enhancer interactions (217), and have
been used to examine the permeation of 5-fluorouracil and estradiol (218). Lecithin
organogels formed by the addition of a critical quantity of water to a solution of
pure lecithin in an appropriate organic solvent (e.g., cyclooctane) contain bicontin-
uous intercalated networks of aqueous and organic phases that may have some struc-
tural similarity to the possible pathways within the SC (219). The diffusion of na-
proxen, diclofenac, and ibuprofen through the lecithin organogels was representative
of their relative rate of permeation through full-thickness human skin (Fig. 9). Com-
parison was also made between the permeation of ibuprofen, from eight commercial
formulations, through full-thickness human skin, 700-␮m–thick Silastic membranes,
and a cyclooctane–lecithin–water organogel (220). Permeation of ibuprofen through
human skin was higher from gel systems than from creams, but neither synthetic
model was capable of accurately reproducing formulation performance through skin.
Permeation modulation, using laurocapracm (Azone), was compared in three model
membrane systems; liposome-coated membranes, solid-supported liquid membranes,
and a lecithin organogel (221). Although the data were equivocal, there were indi-
Figure 9 Relation between the flux of various nonsteroidal anti-inflammatory agents across
human skin in vitro and across a lecithin–cyclooctane organogel held between cellulose nitrate
membranes. (From Ref. 219.)
cations that the organogels provided a more structurally representative model of the
skin barrier than the other two membrane systems.
III. IN VIVO SKIN PERMEATION METHODS

This section discusses the methods used for the investigation of skin penetration and
percutaneous absorption in vivo. Many of the methods are based on animal models
that are often used in the early stages in product development (i.e., before the avail-
ability of extensive toxicological data). Although the relevance of animal models to
the human situation has not been fully evaluated and is often questioned (27–
30,222,223), there is little doubt that well-planned investigations using laboratory
animals can provide useful information for future developmental work. However, the
use of animals in testing is becoming less acceptable in certain sectors, and this is
reflected in the 6th Amendment to the Cosmetics Directive (93/95/EEC), which in-
cludes a future ban on the use of animals in testing cosmetic products and their
ingredients. In the development of topical or transdermal pharmaceutical products
containing new chemical entities, full-scale clinical studies must be performed. The
development of generic dermatological or transdermal products ultimately requires
proof of bioequivalence with the innovator product. In either event, it is prudent to
have available prototype products optimized for drug delivery using human skin in
vitro or laboratory animal–human in vivo determinations.
A. Animal Models for In Vivo Percutaneous Penetration Studies

It is undeniable that the most appropriate data relating to percutaneous penetration
is that generated in vivo in human volunteers (224). However, it is not always eth-
ically or practically possible to conduct pharmacokinetic studies in humans, and
animal models, therefore, are often used as a substitute. Numerous species have been
used in such percutaneous absorption studies (Table 6).
224 Brain et al.
Table 6 Some Animal Models Used for In Vivo Percutaneous Absorption Studies
Species Ref.
Mouse 223, 234, 270, 368, 375, 391

Rat 18, 41, 62, 183, 223, 243, 246, 247, 250, 265, 266, 271, 324, 335, 336, 379,
391, 393
Guinea pig 223, 237, 267, 369, 390, 391, 393
Rabbit 223, 237, 243, 247
Dog 234, 397
Pig 185, 223, 234, 243, 393
Monkey 41, 69, 223, 246, 247, 249, 268, 272, 381, 403
1. Species Variation in Skin Structure

Most species in which skin permeability has been investigated are mammals, and
their skin is macroscopically separated into three layers, the stratum corneum, the
viable epidermis, and the dermis. The most important considerations in terms of
barrier function are differences in the sebum, SC, and follicles. Small laboratory
animals, such as rats, mice, and rabbits, lack sweat glands, but have more hair
follicles than human skin. For example, guinea pig skin contains about 4,000–5,000
follicles/cm2, rat skin about 8,000 follicles/cm2, and rabbit skin more than 10,000
follicles/cm2, whereas human skin contains about 6 follicles/cm2 (225,226).
There is little data on skin permeation in larger animal species. However, as it
has been estimated that there are some 2,000 follicles/cm2 in cattle and up to 10,000
follicles/cm2 in some skin regions of Merino sheep, it is likely that the follicular
route for skin permeation is significant in these species.
Furthermore, many chemicals can bind strongly to the keratin of hair and wool
and, when this occurs, it reduces the amount of applied chemical reaching the SC
and, hence, the amount available for absorption. It is important, therefore, that this
factor is taken into consideration in experimental design for comparisons between
shaved and chemically depilated skin sites within the same animal species.
The chemical nature of the intercellular spaces of the SC has been well estab-
lished in human, pig, and several laboratory animals (227–229). Those mammalian
species for which SC lipids have been studied extensively show that there is not a
great variation in general type among different species. The lipids present are pre-
dominantly ceramides, cholesterol (plus cholesteryl esters), and free fatty acids.
There are some interspecies differences in the quantity of lipid(s) present, and this
may have considerable relevance in differences in skin permeation (230).
The SC thickness in most animal species is between 15 and 30 ␮m but, in
general, this tends to increase with animal size. Thus, the thickness of the SC of rats
is about 20 ␮m, whereas that of pig and human are about 30 ␮m. Although the
morphological structure of the SC shows reasonable consistency among species, there
are some infrequent deviations. In sheep, for example, the distal layers of the SC
separate from the basal layer in a disorganized manner (231) and become embedded
in the surface sebum layer.
In the context of evolutionary development, it is perhaps not surprizing that
larger animals generally share more anatomical and physiological features with hu-
mans than smaller laboratory animals (232). This is also true for the integument;
therefore, it is predictable that the pig and rhesus monkey (233,234) are preferable
to such species as the hairless mouse in the prediction of human skin penetration
characteristics. If skin permeability properties of different species were related to the
body size and average physiological life time, skin permeability would be expected
to be in the order: mouse > rat > guinea pig > rabbit > monkey > dog > goat >
sheep > pig > cattle > human. As stated previously, SC thickness increases with
animal size, whereas the lipid content decreases with size. The lower amounts of SC
lipid are reflected in the observation that the permeability of many un-ionized and
lipid-soluble solutes through skin appears to be lower in the larger animal species.
2. Ranking in Skin Permeability
Investigations on the usefulness, and predictability of animal models have been of
two types: (a) those in which percutaneous absorption of one or more permeants is
measured in several species; and (b) those in which absorption of one or more
chemicals is compared between the experimental animal and human. If the data from
animal models are to be used for extrapolation to human, it is important that the
investigators appreciate the differences in the behavior of the animal model and
human tissue. The skin permeability of the animal tissue model must be the same
as that of the human, or easily related to it by a constant ratio (4). Furthermore, any
response to permeation modulators (physical or chemical), formulation excipients,
and occlusion must mimic the response in humans (235,236).
Norgaard (237) reported that the species ranking of skin permeability rates for
cobalt ions was rabbit > guinea pig > human. Tregear (238) ranked the species
permeabilities for metal ions as rabbit > rat > guinea pig > human, whereas that
given by McCreesh (239) (for two ill-defined organophosphorus solutes) was rabbit,
rat > guinea pig > cat, goat, monkey > dog > pig. Scott et al. (240) reported the
permeability coefficients of the dicationic herbicide paraquat in a range of species.
The observed permeabilities for paraquat relative to water are shown in Table 7.
Two features of paraquat permeability are readily apparent: (a) human skin is much
Table 7 The Observed Permeabilities for Paraquat and Water Through the Skin of
Various Species as Measured Using In Vitro Techniques
Permeability coefficient
(cm2/h ⫻ 10⫺5) Animal/human
ratio for
Species Type Water Paraquat paraquat
Man 92.97 0.73 —

Rat Wistar Alpk/AP 103.09 26.68 40
Hairless 103.08 35.51 50
Nude 151.72 35.34 50
Mouse Alpk/AP 143.75 97.16 135
Hairless 350.70 1066.39 1460
Guinea pig Dunkin–Hartley 442.09 195.63 270
Rabbit NZ White 252.61 79.92 110
Source: Ref. 240.

226 Brain et al.
less permeable to paraquat than any other species examined, and (b) the hairless
mouse is particularly susceptible to paraquat penetration. Given that the skin of most
of the laboratory species studied lack sweat glands, but contain more hair follicles
than humans, these different permeability coefficients may reflect differences in fol-
licular transport for this permanently ionized compound. Unfortunately, data on the
penetration of ions through larger species, such as sheep and cattle, appears to be
lacking, which precludes a more complete analysis.
Durrheim et al. (241) and Huq et al. (242) reported that the permeability of
alcohols and phenols through hairless mouse and human skin was similar. The aque-
ous concentration of phenol required to damage the human epidermis, rat, and hair-
less mouse skin was also similar at about 2%. However, the effects of hydration on
the permeability of hairless mouse and human skin differed markedly.
Table 8 summarizes some of the in vivo data reported on the penetration of
several un-ionized solutes through the skin of a variety of species (243–245). In
general, the magnitude of difference in skin permeability between the species is less
than fivefold, with a rank order of rabbit > rat > pig > monkey > human. More
recently, Moody et al. showed that absorption of the insecticide lindane was similar
in rats and rhesus monkeys (246) and also suggested that animal models for dermal
absorption of phenoxy herbicides may be useful in predicting human dermal ab-
sorption (247).
Sato et al. (230) have investigated species difference in the percutaneous ab-
sorption of nicorandil, using hairless rat, guinea pig, hairless mouse, dog, pig, and
human, and have attempted to relate this to the amount of surface lipid in these
species. As part of this study the influence of the penetration enhancers Azone and
isopropylmyristate on nicorandil permeation was also investigated. There was a clear
difference between the amounts of lipid extracted by acetone from the skin of smaller
laboratory animals (hairless mouse: 212 ␮g/cm2; hairless rat: 273 ␮g/cm2; and guinea
pig: 225 ␮g/cm2) and that from man (60.5 ␮g/cm2). This difference was reflected in
the permeability studies. Both permeation and permeation enhancement of nicorandil
Table 8 The Observed Permeabilities for Several Compounds Through the Skin of
Various Species, Including Humans, Measured Using In Vivo Techniques
Applied dose absorbed (%)

Permeant Human Pig Monkey Rabbit Rat
Acetylcysteine 2.4 6.0 — 2.0 3.5

Butter yellow 21.6 41.9 — 100.0 48.2
Caffeine 47.6 32.4 — 69.2 53.1
Cortisone 3.4 4.1 — 30.3 24.7
DDT 10.4 43.4 1.5 46.3 —
Haloprogin 11.0 19.7 — 113.0 95.8
Lindane 9.3 37.6 16.0 51.2 —
Malathion 8.2 15.5 19.3 64.6 —
Parathion 9.7 14.5 30.3 97.5 —
Testosterone 13.2 29.4 — 69.6 47.4
Source: Ref. 30.

was much greater through the skin of the hairless rat, hairless mouse, and guinea
pig than through human or pig skin. These data reflect the significance of SC lipids
for both the permeation and penetration enhancement process. Similarly, Roberts and
Mueller (248) have compared the in vitro flux of glyceryl trinitrate across hairless
mouse, pig (Yucatan), and human skin, and concluded that mouse skin was an un-
acceptable model for the prediction of human skin permeation behavior.
It is apparent, from the overall data available, that the preferred animal models
for human skin in vivo are the rhesus monkey (223,249) and pig (234). It has been
reported that permeability across hairless rat skin is similar to that of humans (250)
(for certain permeants) although the more general suitability of this species has been
questioned (251). There are certainly differences between the behavior of penetration
enhancers on the skin of hairless rats and humans (252).
To conclude, several general observations can be made on the use of animal
skin in percutaneous absorption studies:
1. Animal skin with high follicular density is poorly representative of human

skin (225,258).
2. Rat and rabbit do not give reliable estimations of human penetration
(243,253–255).
3. Pig and rhesus monkey reasonably approximate absorption of several com-
pounds in humans (223,234,238,243,256,257).
4. Shaving or depilation of hairy skin may alter the barrier function (259,260).
B. Experimental Procedures
This section is based on the recommendations and guidance offered following the
FDA–AAPS workshop (Washington, 1989) on in vivo percutaneous penetration
(261) and the ECVAM workshop report (4).
1. General Principles and Procedures

Application procedures for in vivo studies are broadly similar, with slight variations
depending on the situation. Permeants are applied to a designated area of skin, in
their pure form (262), as a solvent-deposited solid (244,263,264), in solution (265–
267), in a formulation (266,268–274), as a vapor (275), or applied to a piece of
material that is then placed on the skin surface (276). Following application of the
test permeant, various techniques may be used to monitor percutaneous absorption.
There are five approaches to permeant sampling (261): (a) surface residue; (b) in the
skin; (c) in the venous blood draining the application site; (d) in the systemic cir-
culation; and (e) in the excreta.
The most commonly used in vivo method is the assay of urinary and fecal
excretion over time (224) following topical dosing of a small amount of radiolabeled
permeant. The method is useful for assessing total absorption, is noninvasive, and
can be performed in humans. However, the resulting data must be corrected for
incomplete elimination by measurement of radiolabel in excretions following intra-
venous administration (263). It is also important that the collection period is suffi-
ciently long to include all excreted material (82) and that a full mass balance is
conducted (277–279). The application site must be protected with a chamber that
can be either occlusive or protective.
228 Brain et al.
At the end of the administration period, the chamber is removed and assayed
for residual adhering permeant. The skin site is then washed and the washings an-
alyzed. A new chamber is again applied to the application site and the process re-
peated. The skin may also be tape-stripped at this time to determine drug residence
in the SC. These techniques are simple to perform and urinary elimination can be a
useful mirror of the plasma concentration if the clearance of the penetrant is rapid.
Blood sampling has been used widely to determine the systemic availability of
transdermal products (280–285) (Fig. 10). However, for locally acting topical prep-
arations, blood sampling is of little use because, usually, blood (but not tissue) levels
will be very low (286). It is also questionable whether systemic levels are clinically
relevant. This issue was addressed by Singh and Roberts (287–289), who showed
that (in an anesthetized rat) topical coadministration of a vasoconstrictor (phenyl-
ephrine) with other permeants significantly increased the uptake of solutes by the
local tissues (Fig. 11). This effect was probably due to reduced local blood flow and
therefore reduced uptake of the permeants into the systemic circulation (290). In the
absence of vasoconstrictors, direct penetration of solutes into local tissue is more
evident after short periods because recirculation of permeated substances returning
to the dermal vasculature from the systemic blood supply tends to dominate at later
periods (287). An isolated perfused limb model (291,292) has been used to remove
the contribution of the systemic circulation and, thereby, permit examination of local
vasculature uptake and tissue distribution of permeants (293).
2. Skin Grafting
Recent developments in animal modeling for skin absorption have involved the use
of skin-grafting techniques (234,294–299). Human normal or cultured skin is grafted
onto congenitally athymic (nude) laboratory rodents (usually mouse or rat) and the
area of graft used to determine in vivo absorption. To avoid rejection of the graft
the host is immunosuppressed before and after surgery, using specific therapy, such
as the cyclic polypeptide cyclosporine (300,301). The surgical technique is straight-
forward: full-thickness sections of host skin are removed, together with the under-
lying fat layer, and split-thickness human graft skin or supported cultured skin is
attached to this site using sutures or surgical tape. It was estimated, based on surface
electrical capacitance measurements, that in grafted cultured skin, a stable skin tissue
developed by 8–12 weeks following surgery (296).
Transepidermal water loss and tritiated water permeability through human skin
grafted onto mice are similar to those for human skin in situ (181,302). Similarly,
the absorption of several compounds of diverse physicochemical properties across
grafted skin showed good correlation with human in vivo data (29,234,294,303).
A refinement of this approach involves the surgical preparation of a skin sand-
wich flap on an independent vascular supply (304–306). This procedure results in
the host animal (usually athymic nude rat) supporting grafted human skin. The sur-
gical procedures are fully described (304). There are three stages in the procedure.
Briefly, split-thickness skin is grafted to the subcutaneous surface of skin on the
abdomen of the rat to create a sandwich. Graft and host skin are then sutured in
place in the normal anatomical position. Two weeks later, the sandwich is lifted,
together with the associated vasculature, on three sides to create a sandwich flap,
and the exposed wound is covered with a syngenic split-thickness rat skin graft.
After a further week, the sandwich flap is relocated to the rat dorsal surface, but the
Figure 10 (A) Mean plasma levels of estradiol following application of two different
transdermal systems to healthy women. (B) Serum drug levels (mean ⫾ SE) before and after
application of various concentrations of testosterone to the scrotum of hypogonadal men.
(From: A, Ref. 407; B, Ref. 40.)
230 Brain et al.
Figure 11 Topical coadministration of the vasoconstrictor, phenylephrine, to anesthetized

rats significantly increased the concentration of various permeants in the underlying dermis,
compared with that in the contralateral tissue. (Ref. 289; courtesy MS Roberts.)
vasculature remains intact and accessible. The sandwich flap was healed and is ready
for use, 2 weeks after the final surgical procedure. Thus, the flap is viable and has
functional skin on both sides: rat skin on one side and human skin on the other.
Because both skin types have a common vasculature absorption across host or graft,
skin can be assessed (Fig. 12). The ability to sample and analyze blood exiting the
flap before dilution in the systemic pool minimizes the effect of postabsorption met-
abolic activity and considerably enhances sensitivity.
The permeation of benzoic acid (304,306) and estradiol (305) has been mea-
sured using the skin sandwich flap. Although this technique can provide useful data,
particularly in comparing local and systemic effects, there are limitations. The pro-
cedure is time-consuming and costly; graft survival rate is low, and there is a limited
experimental duration because of the anesthetic requirements. Furthermore, as for
standard grafts, animals with surgically produced skin flaps must be treated system-
ically with immunosuppressive drugs, and the effect of this treatment on skin barrier
function is unclear (261).
3. Residual Analysis
Techniques that assess absorption by residual analysis or difference monitor the dis-
appearance of material from the surface of the skin. Such methods are often referred
to as noninvasive because they do not involve surgical procedures or physical sam-
Figure 12 The grafted skin sandwich flap: The surgical procedure is described in the text.
The vein draining the flap is cannulated to facilitate collection of blood samples. Percutaneous
absorption can be determined through both host and grafted skin, as both are served by a
common vasculature. (From Ref. 304.)
pling of body tissues or fluids. Two general approaches have been described: single-
point (307) and continuous monitoring of permeant uptake (308). In the single-point
method, the test chemical is applied for a fixed time, and the residual formulation is
subsequently removed from the skin surface and analyzed. The technique requires
only small quantities of permeant, making the use of radiolabel both cost-effective
and ethically feasible. The main disadvantage of the single-point method is that it
provides only one assay per site per application. If absorption kinetics are required,
multiple application sites are necessary. Validity of the method is enhanced if it can
be demonstrated that the test chemical is not removed from the skin surface by means
other than percutaneous penetration (e.g., through contact with clothing). The amount
permeated is calculated as the difference between the amount applied and the amount
recovered and, if penetration and permeation are low, there may be a large error
associated with the result.
There are various techniques for application of materials. The simplest method
is to apply the test material to a defined area of skin as a solvent-deposited solid.
232 Brain et al.
Once the solvent has evaporated, then the area can be covered with a protective
material that can be either occlusive or nonocclusive. This approach can also be used
for semisolid formulations such as creams or gels. The application of materials as
solutions requires a means of restricting the material to the defined area of skin. This
can be achieved using wells attached to the skin with acrylic or silicone adhesive.
Commercially available devices, such as the Susten Skin Depot or the Bronaugh In
Vivo Ring (Fig. 13) may be modified to provide adequate restrictive devices.
Collection of donor–residual concentration data over time allows kinetic anal-
ysis of permeation. Following application of the test chemical to the skin surface,
disappearance is monitored, either spectroscopically or using radioisotope techniques
(308–311). The technique is very simple and can be performed easily using human
test subjects. For infrared spectroscopic techniques the test permeants must have
distinctive infrared absorbencies in the transparent region of the skin spectrum. Higo
et al. (310) used ATR–FTIR to follow the permeation of 4-cyanophenol across hu-
man SC. The compound was applied for periods up to 3 h in solution in propylene
glycol or propylene glycol containing oleic acid. At the end of the exposure period,
an ATR–FTIR spectrum of the treated surface was recorded. The SC was then re-
moved sequentially by tape-stripping, and an ATR–FTIR spectrum recorded follow-
ing each tape strip. The presence of cyanophenol at various depths in the SC was
determined and quantitatively validated by 14C-spiking (Fig. 14). The data demon-
strated that ATR–FTIR could be used as a technique to quantitate skin levels of
permeants and also as an in vivo method for evaluating skin permeation and pene-
tration enhancement in humans.
Figure 13 Examples of systems used to determine percutaneous absorption in vivo: The

Bronaugh in vivo ring is a protective device that may be used in studies for which evaporation
is not a factor. The Susten skin depot is particularly useful for determining the dermal ab-
sorption of volatile materials. (Courtesy Crown Glass Company, Somerville, NJ.)
Studying Percutaneous Absorption
Figure 14 The use of ATR–FTIR to determine the permeation of 4-cyanophenol across human SC: the compound was applied for periods
up to 3 h in a solution of propylene glycol (open circles) or propylene glycol containing 5% (v/v) oleic acid (closed circles). The SC was
removed sequentially by tape-stripping, and an ATR–FTIR spectrum was recorded following each tape strip. Application periods were (A) 1
h, (B) 2 h, and (C) 3 h (From Ref. 310.)
233
234 Brain et al.
Spectroscopic techniques have also been used widely in skin lipid biophysics
(312) and to assess the water content of skin in vivo and in vitro (313–315). Other
spectroscopic methods that have been used in the investigation of the properties of
skin include fluorescence (316), impedance (317), and neutron (318) spectroscopy.
Optothermal-imaging techniques (such as optothermal transient emission radi-
ometry; OTTER) have been used to monitor disappearance of material from the skin
surface (Fig. 15). The technique is nondestructive and noninvasive and has been used
to characterize skin condition (319,320), to follow the disappearance of sunscreens
and other materials from the skin surface (321,322), and to monitor emulsion break-
down on the skin (323).
4. Skin-Stripping In Vivo
In Section III.B.1, the basics of tape-stripping and its use in vitro were outlined, and
here, the in vivo applications are discussed. The in vivo technique is based on the
reservoir principle (324–328). It is hypothesized that if a compound is applied to
the skin for a limited time (e.g., 0.5 h) and then removed, the amount of drug in the
upper layers of the SC will be predictive of the overall bioavailability of the com-
pound. It follows that determination of the SC content of a permeating material
following a short-term application will predict in vivo bioavailability from a corre-
sponding administration protocol. Data obtained in studies of this type have shown
reasonable predictability for several compounds (Fig. 16).
The use of in vivo skin-stripping in dermatopharmacokinetic evaluation was
discussed at an AAPS–FDA workshop on bioequivalence of topical dermatological
dosage forms (Bethesda, USA, September 1996). Although opinion was divided, it
was concluded that SC skin-stripping ‘‘may provide meaningful information for com-
parative evaluation of topical dosage forms’’ (150). Furthermore, it was established
that a combination of dermatopharmacokinetic and pharmacodynamic data may pro-
vide sufficient proof of bioequivalence ‘‘in lieu of clinical trials.’’ However, much
remains to be validated in skin-stripping protocols.
An outline protocol for skin-stripping bioequivalence studies has been sug-
gested (150). The basic protocol has two phases: uptake and elimination.
Uptake:
1. Test and reference drug products are applied concurrently at multiple sites.
2. After exposure for a suitable time (determined by a pilot study), excess
drug is removed by wiping three times with tissue or cotton swab.
3. The adhesive tape is applied with uniform pressure. The first strip is dis-
carded (skin surface material). Repeat if necessary to remove excess sur-
face material.
4. Collect nine successive tape strips from the same site. If necessary collect
more than nine strips.
5. Repeat the procedure for each site at designated time intervals.
6. Extract the drug from the combined tape strips for each time point and site
and determine the content of drug using an appropriate validated analytical
method.
7. Express the data as amount of drug per cubic centimeter of tape.
Figure 15 The use of optothermal transient emission radiometry (OTTER) to monitor

disappearance of propylene glycol (PG) and dipropylene glycol (DPG) from the skin surface
in vivo: disappearance is shown as, (A) the relative concentrations of the two permeants
remaining on the skin surface with time (mean values from three sites); (B) disappearance of
DPG from three sites on the ventral forearm. (Ref. 322.)
236 Brain et al.
Figure 16 Correlation between the amount of compound absorbed through the skin, as
determined from total body analysis or total excretion over 4 days, and the amount of com-
pound recovered from stratum corneum tape strips following 30-min exposure. (A) Data
obtained using hairless rat (200 nmol/cm2 applied to dorsal skin). (B) Data obtained using
human skin (1000 nmol/cm2 applied to abdominal skin). (From: A, Ref. 410; B, Ref. 411.)
Elimination:
1. As for the foregoing steps 1–3.
2. After a predetermined time interval (e.g., 1, 3, 5, and 21 h postdrug re-
moval) perform foregoing steps 4–7.
The results may then be expressed as shown in Fig. 17 that plots the amount
of drug recovered from the tape strips against time. Uptake and elimination phases
are observed and ‘‘bioavailability’’ may be predicted from the AUC.
There are several sources of variability in such studies, all of which must be
considered in standard-operating procedures. The major causes of concern are vari-
ability in the following:
1. Drug application procedure
2. Type of tape
3. Size of tape
4. Pressure applied by investigator
5. Duration of application of pressure
6. Drug removal procedure
7. Drug extraction procedure
8. Analytical methods
9. Temperature
10. Relative humidity
11. Skin type
12. Skin surface uniformity
Nonetheless, following further validation, the technique will have several ad-
vantages. For example, basic pharmacokinetic parameters such as AUC, Cmax, Tmax,
and half-life (329) may be approximated from the data obtained. In addition, the
approach could be applicable to all types of topical preparation.
Pershing et al. (134) validated an in vivo skin-stripping protocol by correlating
the SC strip data obtained for betametasone dipropionate with a skin-blanching bio-
assay experiment. Brief details of the stripping protocol were as follows: 180 mg of
formulation were placed in a 1.2-cm–diameter Hilltop chamber (Hilltop Research
Inc., Cincinnati, OH) and attached to the volar aspect of the forearm, with an exposed
skin area of 1.13 cm2 (i.e., a dose of 159 mg/cm2). A maximum of three chambers
per forearm were used, each being 2 cm apart, at least 6 cm above the wrist and 6
cm below the elbow. At the end of a 24-h period the chamber was removed, residual
formulation on the skin surface wiped off, and the skin wiped with three separate
dry cotton applicators. After air drying for 2 min, the skin was stripped ten times
using 0.6-cm–diameter disks of Transpore tape. The amount of SC removed by each
tape strip was assessed by weighing each tape before and after stripping. The relation
between tape strip number and weight of SC removed was linear (r = 0.996), with
an average of about 30 ␮g of skin removed per strip. Drug was extracted from the
tape strips by vortexing in 200 ␮L of acetonitrile, followed by centrifugation and
analysis of the supernatant by HPLC. Skin blanching was assessed at 1, 24, and 48
h following removal of formulations applied under occlusion for 24 h. The corre-
lation between the amount of betamethasone dipropionate in skin and the skin-
blanching score was good (r = 0.994) (Fig. 18), although the skin-blanching scores
were not entirely objective because they were assessed visually, rather than with a
238 Brain et al.
Figure 17 The use of stratum corneum tape-stripping to determine topical bioavailability

or bioequivalence: (A) theoretical profile illustrating uptake and elimination phases, ‘‘bio-
availability’’ may be predicted from the AUC; (B) experimental data for betamethasone di-
propionate. (From Ref. 409.)
Figure 18 Correlation between the amount of betamethasone dipropionate in skin (deter-

mined by tape-stripping) and skin-blanching score (mean ⫾ SE: n = 5) for some commercial
formulations. (From Ref. 134.)
chromameter. Nonetheless, differences in responses between formulations (cream and

ointment) and manufacturers could be discerned with both the pharmacokinetic and
the pharmacodynamic techniques.
5. Microdialysis
A further technique that is currently being evaluated for the measurement of in vivo
percutaneous absorption is microdialysis (330–338). Cutaneous microdialysis de-
scribes the in vivo measurement of substances in the extracellular space of the dermis
by using a perfused dialysis membrane system. The technique involves insertion of
microdialysis tubing below the skin surface into the dermis and back out again (Fig.
19). The dialysis tube is then perfused with a receiver fluid that removes permeant
from the local area. The amount of permeant removed by the receiver medium is
then related to the total amount absorbed. Disadvantages of microdialysis include
small sample size, requirement of sensitive analytical methods, and the invasive
nature of the technique. Detection of lipophilic or highly protein-bound substances
are problematic. Nonetheless, cutaneous microdialysis may have a role in the as-
sessment of dermal toxicity and in the development of an understanding of the
pathogenesis of dermatological disease.
6. Pharmacodynamic Measurements
The use of pharmacodynamic changes (i.e., bioassays), can provide a method for
assessing local absorption. The most widely used technique is the vasoconstriction
240 Brain et al.
Figure 19 Schematic illustrating the technique of cutaneous microdialysis: the amount of

substance migrating across the epidermis and into the extracellular space of the dermis is
estimated from the concentration of permeant appearing in the perfusion fluid. (From Ref.
412; courtesy M.S. Roberts.)
assay (or skin-blanching assay) used for measuring the dermal absorption of gluco-
corticosteroids (134,339–341). The basis of the assay is the correlation between the
vasoconstriction produced by the compounds and their overall clinical effect. Al-
though such methods are inexpensive, internally consistent, and qualitatively reliable
(4), objective methods for quantifying the response are essential. Assessment of color
change can be quantitatively determined using chromameters (342,343). Pharmaco-
dynamic measurements are fully discussed in Section V of Chapter 8.
7. Clinical Evaluation
Clinical efficacy testing will always be necessary to demonstrate the effectiveness of
any new transdermal or topical therapy. For example, an extensive review (344) of
estrogen replacement therapy that addressed the key pharmacokinetic and metabolic
differences between oral and transdermal forms, with particular emphasis on their
effects on gonadotropins, hemostasis and coagulation, lipid metabolism, hepatobiliary
function, and bone, concluded that the transdermal route was a viable alternative to
the oral route of administration. Lin et al. (345) described the development of a
transdermal nicotine delivery device and generated clinical evaluation to demonstrate
the efficacy of their product. This approach was also used by Shah et al. (346) in
the development of a transdermal verapamil system.
The clinical assessment of transdermally delivered substances is often coupled
with the generation of pharmacokinetic data. However, it is more difficult to generate
pharmacokinetic data during evaluation of the clinical efficacy of dermatological
products, such as antifungal preparations, or locally acting topical products, such as
nonsteroidal anti-inflammatory agents (e.g., ibuprofen and diclofenac). Furthermore,

in the latter example, the demonstration of efficacy is difficult because the indication
is often self-resolving, and there is a significant placebo effect. Although there are
many studies that describe the bioavailability and efficacy of topical NSAIDs
(88,347–362), there still exists some scepticism over their therapeutic value, espe-
cially in comparison with simple oral dosing. Recently, several articles have ad-
dressed this issue by examining ‘‘local enhanced topical drug delivery’’ and con-
cluded that, for the compounds studied, this does indeed occur (288,289,363,364).
IV. IN VIVO–IN VITRO COMPARISONS

It is important to consider the ability of in vitro skin penetration techniques to predict
skin permeation in vivo. Unless it can be demonstrated that in vitro skin penetration
data is reasonably similar to absorption across skin in situ, there is little value in
obtaining this data. It is also beneficial to compare in vitro skin penetration data with
clinical observations (see, e.g., Chap. 6, Sec. II.B.3), although the information avail-
able is sparse.
A. Correlation Between In Vitro and In Vivo Measurements

A review of the literature reveals that, in most cases, extrapolation between in vitro
and in vivo human skin permeation is reasonably accurate (8,18,82,264,365–386).
If, however, a permeating compound undergoes extensive dermal metabolism in vivo,
then this may not be reflected by valid in vitro results unless freshly excised skin is
used (378). Overall, however, the ranking of prototype formulations or the evaluation
of the skin permeation of homologous compounds achieved using in vitro methods
will be a reflection of the in vivo scenario.
Many such studies have been performed using small laboratory animals, and
their relevance to human skin has always been questionable, especially in terms of
risk assessment. However, in terms of system validation, the comparison of in vitro
and in vivo skin penetration data obtained using animal models may be useful. A
comprehensive study of in vitro–in vivo relations for animal skin was carried out
(371) using rat tissue. Some of the data from this set is shown in Table 9 together
with other in vivo–in vitro data for different animal models.
Scott et al. (371) determined the percutaneous absorption in vivo and the skin
permeation in vitro of eight pesticides (ICI A through ICI G) varying in molecular
weight from 187.3 to 416.3, in water solubility from 0.01 to 880 mg/mL, and in log
PO/W (log octanol–water partition coefficient) from 2.1 to > 5.0. In this study com-
parisons were also made between the amount of test chemical remaining on or in
the skin after in vivo and in vitro experimentation. In general, despite a difference
in rinsing procedure between the in vivo and in vitro experiments, there was good
agreement between the amount of test chemical that could be removed from the
surface of the skin at the end of the experiments. Similarly, in most cases, the amount
of compound recovered from the skin was similar. For all test compounds there was
qualitative and quantitative agreement in ranking of absorption between the in vitro
and in vivo data (see Table 9). Usually, in vitro absorption values were higher than
in vivo values. Nonetheless, for this dataset, the predictability of the in vivo results
based on in vitro data, measured by linear regression over all time points, was rea-
sonable (r 2 = 88.5%) and at later time points was good (r2 = 95.5%).
242 Brain et al.
Table 9 Comparison Between Some In Vitro and In Vivo Penetration Data Using Rat,
Mouse, and Monkey Skin
Species Test compound In vivoa In vitroa Ratiob Ref.
Rat Anthracene 18 24 1.3 379

Benzyl acetate 2 4 2.0 380
Cypermethrin 1 1.7–2.7 1.7–2.7 370
ICI A 12 24 0.5 371
ICI B 86 76 0.9 371
ICI C 23 22 1.0 371
ICI D 56 79 1.4 371
ICI E 4 0.8 371
ICI F 32 78 2.4 371
ICI G 2 2 1.0 371
ICI H 14 9 0.6 371
Mouse Permethrin 26 18 0.7 368
DDT 25 27 1.1 368
2,4-D 44 28 0.7 368
Fenvalerate 19 20 1.0 368
Monkey 4-Amino-2-nitrophenol 64 48 0.8 381
2,4-Dinitrochlorobenzene 52 48 0.9 381
p-Nitroaniline 76 62 0.8 381
2-Nitro-p-phenylene diamine 30 30 1.0 381
Nitrobenzene 4.2 6.2 1.5 381
a
Data are expressed as % applied dose absorbed.
b
In vitro–in vivo.
In a similar study, Surber et al. (369) examined the in vitro–in vivo relation
in the penetration and permeation of structurally related phenols and steroids across
hairless guinea pig skin. The para-substituted phenols varied in log PO/W from 0.3
to 3.5 and in water solubility from 0.3 to 12.6 mg/mL. The steroids varied in log
PO/W from 1.5 to 3.9 and in water solubility from 0.003 to 0.39 mg/mL. Penetration
into, and permeation through, the skin were determined following application of the
test compound in either water or isopropyl myristate. The data are given in Table
10 and shown in Figure 20. As is usual for such studies, a parabolic relation between
permeation and test compound lipophilicity was observed for both the in vivo and
in vitro experiments. For the phenol analogues, the in vitro measurement overesti-
mated the values obtained in vivo, but this was reversed for the steroids.
Kasting et al. (377) determined the in vitro and in vivo rat skin permeation of
three lipophilic capsaicin-like vanilloids, varying in log PO/W from 3.74 to 8.02. The
degree of correlation between the data sets was dependent on the nature of the
receptor fluid in the in vitro experiment and the time over which comparisons were
made (Table 11). At the 24-h timepoint, there was closer agreement for the lipophilic
permeants when the receptor fluid was designed to accommodate such permeants.
Mostly, however, at the 72-h timepoint oleth 20-containing receptor fluids tended to
overestimate the in vivo absorption. The authors rationalized the time-related differ-
ences between in vivo and in vitro absorption by proposing that they were the con-
Table 10 Comparison Between Some In Vitro and In Vivo Skin Permeation and
Penetration Data Using Guinea Pig Skin
Permeation (through skin) Penetration (into skin)

In vitro In vivo In vitro In vivo
Test compound (␮g/cm2) (␮g/cm2) (␮g/g) (␮g/g)
A.
4-Acetamidophenol 23.9 17.3 15.5 7.6
4-Cyanophenol 408.6 317.2 154.3 122.7
4-Iodophenol 242.6 156.7 103.6 90.5
4-Pentyloxyphenol 60.8 89.9 34.1 83.2
Hydrocortisone 1.4 2.2 1.1 1.0
Estradiol 0.8 3.6 0.9 1.1
Testosterone 4.0 5.3 1.3 1.6
Progesterone 0.6 1.1 0.3 0.5
B.
4-Acetamidophenol 33.1 24.7 7.3 6.9
4-Cyanophenol 2221.8 1373.4 345.6 237.6
4-Iodophenol 537.3 280.0 113.7 84.2
4-Pentyloxyphenol 79.6 56.8 94.3 79.5
Hydrocortisone 2.0 8.6 2.0 15.3
Estradiol 4.0 13.9 9.2 64.7
Testosterone 9.3 41.3 16.7 104.2
Progesterone 2.4 15.6 7.4 64.8
Compounds were applied as saturated solution in either water (A) or isopropyl myristate (B).
Source: Ref. 369.
sequence of an observed gradual decline in the in vivo absorption rates, not observed
in vitro, possibly occurring as a result of skin turnover.
The most comprehensive studies on in vitro–in vivo correlations using human
skin are those of Franz (8,82). In these experiments conditions of application of the
compound to the skin were as similar as possible in the two protocols. The data
obtained in an initial study (8) are summarized in Table 12. There was a rank order
correlation between the two sets of data (p = 0.01) which was remarkable, given
that the in vivo data was obtained using forearm skin (263) and the in vitro data
using abdominal skin, in different laboratories. Furthermore, the correlation was con-
siderably better for those compounds that exhibited greater penetration. From the
data given in Table 12, it is obvious that for poorly permeating compounds (< 1.0%
absorption in vivo), the in vitro system significantly overestimates absorption. This
may be rationalized to some extent by a consideration of the differences in the
dynamics of the two systems. In vivo there is an average loss, through desquamation,
of about one cell layer of SC per day and it can be assumed that, for slow penetrants,
a significant proportion of the applied dose of permeant will be lost in this process.
Because of a lack of desquamation in vitro this loss does not occur, and the per-
meating compound will remain in the SC over a longer period and thus continue to
be available for diffusion into deeper epidermal layers.
244 Brain et al.
Figure 20 The in vitro–in vivo relation for the permeation of structurally related phenols
across hairless guinea pig skin: permeation was determined following application of the test
compound in either (A) water or (B) isopropyl myristate. A parabolic relation between per-
meation and lipophilicity is observed for both the in vivo and in vitro experiments. As can
be seen, here, the in vitro measurements overestimate the values obtained in vivo. (From Ref.
369.)
In a follow-up study (82), the compounds that showed the greatest difference
between in vivo and in vitro absorption (nicotinic acid, hippuric acid, thiourea, and
caffeine) were reevaluated, with a slightly different protocol. In this instance, ab-
dominal skin was used for both in vivo and in vitro application, and in vivo urine
was collected until the permeant levels had reached background levels. In addition
the area of application was washed 24 h following application, which should mini-
mize the differences that may have been caused by desquamation. The data obtained
in the follow-up study are shown in Table 13. The modified protocol in this follow-
up experiment produced a much-improved correlation between in vivo and in vitro
Table 11 Influence of In Vitro Receptor Fluid Composition on the Correlation Between

In Vitro and In Vivo Skin Permeation of a Series of Vanilloids
Test permeant
Vanillylnonanamide Olvanil NE-21610
Log P 3.74 8.02 7.16
In vivoa (24 h) 4.5 1.4 0.85

In vivoa (72 h) 12.7 3.6 2.01
In vitro
receptor fluidb %a Rc %a R %a R
0.5% BSA (24 h) 2.3 0.51 0.9 0.64 0.14 0.16

0.5% BSA (72 h) 6.8 0.54 3.4 0.94 0.45 0.22
1% Oleth 20 (24 h) 3.2 0.71 1.5 1.07 0.38 0.45
1% Oleth 20 (72 h) 11.6 0.91 6.7 1.86 1.70 0.85
6% Oleth 20 (24 h) 3.7 0.82 1.5 1.07 0.53 0.62
6% Oleth 20 (72 h) 14.4 1.13 7.9 2.19 2.63 1.31
a
% Applied dose Penetrated.
b
Dulbecco’s phosphate buffered saline (pH 7.4) containing preservative and the additives noted.
c
In vitro/in vivo ratio.
Source: Ref. 377.
Table 12 Comparison of In Vitro Human Skin Penetration Data with

Published In Vivo Values
Absorption Absorption In vitro/in vivo

Compound in vivo (%) in vitro (%) ratio
Hippuric acid 0.2 1.2 6.0

Nicotinic acid 0.3 3.3 11.0
Thiourea 0.9 3.4 3.8
Chloramphenicol 2.0 2.9 1.5
Phenol 4.4 10.9 2.5
Urea 6.0 11.1 1.9
Nicotinamide 11.1 28.8 2.6
Acetylsalicylic acid 21.8 40.5 1.9
Salicylic acid 22.8 12.0 0.5
Benzoic acid 42.6 44.9 1.1
Caffeine 47.6 9.0 0.2
Dinitrochlorobenzene 53.1 27.5 0.5
Source: Ref. 263.

246 Brain et al.
Table 13 Comparison of In Vitro and In Vivo Human Skin

Penetration Data Obtained Under Controlled Conditions

Hippuric acid 1.0 1.3 1.3

Nicotinic acid 0.3 2.3 7.7
Thiourea 3.7 4.6 1.2
Caffeine 22.1 24.1 1.1
Source: Ref. 82.
results. In most cases, quantitative agreement was excellent. The difference observed
for nicotinic acid is attributable to the observation that only 15% of an intravenous
dose of this compound is excreted in the urine and, when this is factored, the value
for in vivo absorption becomes 2.1%, virtually identical with that obtained in vitro.
It is further evident from the data given in Tables 12 and 13 that a slight alteration
in protocols caused a significant alteration in the in vivo data for each compound,
although the in vitro data (except for caffeine) was comparatively unchanged.
Bronaugh and Franz (365) studied in vivo–in vitro comparisons in the human
skin permeation of benzoic acid, caffeine, and testosterone from a petrolatum vehicle.
Good agreement was obtained (Table 14).
In summary, comparison between in vivo and in vitro data for compounds
penetrating and permeating human skin have shown that a high correlation exists
between the two sets of data. In most cases, for poor penetrants, the in vitro data
has slightly overestimated the values obtained in vivo. However, it is important to
appreciate that, as pointed out by Franz (82) and Kligman (387), when discrepancies
occur between in vivo and in vitro data, it is more likely that the in vivo data have
been measured inaccurately. It can be concluded that, provided the experimentation
is performed using suitably designed protocols and that the in vivo ‘‘in-use’’ con-
ditions are mimicked as closely as possible, valuable and realistic data on skin pen-
etration can be obtained using human skin in vitro.
Table 14 Comparison of In Vitro and In Vivo Human Skin

Penetration Data for Compounds Applied in Petrolatum

Benzoic acid 60.6 46.5 0.8

Caffeine 40.6 40.6 1.0
Testosterone 49.5 39.4 0.8
Source: Ref. 365.

B. Correlation Between In Vitro Data and Therapeutic Activity

It is important to realize the limitations of in vitro measurements and to understand
that data derived from such experiments should not be overinterpreted. Although, in
vitro measurements have undoubtedly generated, and will continue to generate, very
useful information for formulation development, and optimization (386), hazard and
risk assessment (388), and for bioequivalency testing (see Chap. 8), they are not as
useful when attempting to predict onset of therapeutic action. As discussed earlier,
neither physiological nor pharmacodynamic responses occur in vitro; thus, for ex-
ample, a formulation applied to the skin in vivo may be affected by sweat or sebum
secretion that cannot be mimicked in vitro. Similarly, a permeating compound that
has the property of vasodilation, will increase blood flow to, and increase permeant
clearance from, the skin in vivo, but will have no such effect in vitro. In vitro, and
indeed in vivo, quantitative measurements of skin penetration, permeation, and ab-
sorption lag times are limited by analytical sensitivity. Pharmacodynamic and ther-
apeutic responses are limited by receptor occupancy. Thus, whereas it may take
several hours for a permeating compound to reach levels of analytical detection, it
may take only a few minutes for sufficient molecules to permeate the skin, attach to
receptors, and elicit the pharmacodynamic response.
Nonetheless, there is some evidence that in vitro measurements of skin per-
meability can qualitatively rank the clinical response. Some examples of this are
discussed in Chapter 6 (see Sec. II.B.3). Recently, Foldvari et al. (389) have dem-
onstrated that in vitro measurements of prostaglandin-E1 permeation across human
skin can predict vasodilatory activity. Although the in vitro lag time was 5–10 h
(depending on formulation), peak blood flow in vivo, as determined using laser
Doppler flowmetry, was attained within 30 min. Nonetheless, formulation ranking of
vasodilatory efficacy was predictable by in vitro skin permeation measurement.
There is no doubt that measurement of skin permeation both in vitro and in vivo
can play a major role in the optimization of formulations for dermal and transdermal
drug delivery. The good correlation between in vitro and in vivo measurements adds
support to the usefulness of the former, more cost-effective, option. In addition to
the pharmaceutical considerations, regulatory concern about potential dermal ab-
sorption leading to possible local or systemic effects, is now a major concern for the
cosmetic and chemical industries. For products in these sectors to meet increasingly
stringent requirements, conclusive evidence for their efficacy and safety is required.
With the recent loss of the option to conduct in vivo animal testing on cosmetics
and their ingredients, and the ethical and cost constraints of human in vivo studies,
one of the most viable remaining approaches for assessing dermal absorption is that
of in vitro human skin permeation evaluation. Although in many respects these tech-
niques may appear straightforward, it is important to appreciate that a generalized
protocol may not provide an appropriate framework for a particular study, that it is
essential that studies are well-designed to fit their specific purpose, that they are
competently executed by experienced scientists, and that the resultant data is inter-
preted appropriately taking into account all of the relevant factors discussed in the
foregoing.
248 Brain et al.
REFERENCES
1. Federal Register. Requests for proposals for enforceable consent agreements; dermal
absorption rate testing of eighty OHSA chemicals; solicitation of interested parties; text
of protocol. April 3, 61(65); 1996.
2. Walker JD, Whittaker C, McDougal JN. Role of the TSCA interagency testing com-
mittee in meeting the U.S. government data needs: designating chemicals for percu-
taneous absorption rate testing. In: Marzulli FN, Maibach HI, eds. Dermatotoxicology,
5th ed. Washington, DC: Taylor & Francis, pp 371–381, 1996.
3. Skelly JP, Shah VP, Maibach HI, Guy RH, Wester RC, Flynn G, Yacobi A. FDA and
AAPS report of the workshop on principles and practices of in vitro percutaneous
penetration studies: relevance to bioavailability and bioequivalence. Pharm Res 4:265–
267, 1987.
4. Howes D, Guy R, Hadgraft J, Heylings J, Hoeck F, Maibach H, Marty J–P, Merk H,
Parra J, Rekkas D, Rondelli I, Schaefer H, Täuber U, Verbiese N. Methods for assessing
percutaneous absorption, ECVAM Workshop Report 13. ALTA 24:81–106, 1996.
5. ECETOC Monograph No. 20 Percutaneous absorption. Brussels: European Centre for
Ecotoxicology and Toxicology of Chemicals, 1993; 80 pp.
6. COLIPA (The European Cosmetic, Toiletry and Perfumery Association). Cosmetic in-
gredients: guidelines for percutaneous absorption/penetration, 1993.
7. Lehman PA, Agrawa IN, Franz TJ, Miller KJ, Franz S. Comparison of topical hydro-
cortisone products: percutaneous absorption vs. tape stripping vs. vasoconstriction vs.
membrane rate of release. Pharm Res 13:S310, 1996.
8. Franz TJ. On the relevance of in vitro data. J Invest Dermatol 64:190–195, 1975.
9. Bronaugh RL, Stewart RF. Methods for in vitro percutaneous absorption studies. IV:
The flow through diffusion cell. J Pharm Sci 74:64–67, 1985.
10. Walters KA, Flynn GL, Marvel JR. Physicochemical characterization of the human
nail: I pressure sealed apparatus for measuring nail plate permeabilities. J Invest Der-
matol 76:76–79, 1981.
11. Flynn GL, Smith EW. Membrane diffusion. I. Design and testing of a new multifeatured
diffusion cell. J Pharm Sci 60:1713–1717, 1971.
12. Nugent FJ, Wood JA. Methods for the study of percutaneous absorption. Can J Pharm
Sci 15:1–7, 1980.
13. Gummer C. The skin penetration cell: a design update. Int J Pharm 40:101–104, 1987.
14. Tanojo H, Roemelé PEH, van Veen GH, Stieltjes H, Junginger HE, Boddé HE. New
design of a flow-through permeation cell for studying in vitro permeation studies across
biological membranes. J Controlled Release 45:41–47, 1997.
15. Bronaugh RL. Methods for in vitro percutaneous absorption. In: Marzulli FN, Maibach
HI, eds. Dermatotoxicology, 5th ed. Washington, DC: Taylor & Francis, pp 317–324,
1996.
16. Hughes MF, Shrivasta SP, Fisher HL, Hall LL. Comparative in vitro percutaneous
absorption of p-substituted phenols through rat skin using static and flow-through dif-
fusion systems. Toxicol In Vitro 7:221–227, 1993.
17. Clowes HM, Scott RC, Heylings JR. Skin absorption: flow-through or static diffusion
cells. Toxicol In Vitro 8:827–830, 1994.
18. Moody RP, Ritter L. An automated in vitro dermal absorption procedure. II. Compar-
ative in vivo and in vitro dermal absorption of the herbicide fenoxaprop-ethyl (HOE-
33171) in rats. Toxicol In Vitro 6:53–59, 1992.
19. Bronaugh RL. In vitro methods for the percutaneous absorption of pesticides. In:
Honeycutt R, Zweig G, Ragsdale NN, eds. Dermal Exposure to Pesticide Use. Wash-
ington, DC: American Chemical Society, pp 33–41, 1985.
20. Dick IP, Blain PG, Williams FM. Improved in vitro skin absorption for lipophilic
compounds following the addition of albumin to the receptor fluid in flow-through
cells. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration.
Vol. 4b. Cardiff: STS Publishing, pp 267–270, 1996.
21. Sclafani J, Liu P, Hansen E, Cettina MG, Nightingale J. A protocol for the assessment
of receiver solution additive-induced skin permeability changes. An example with
gamma-cyclodextrin. Int J Pharm 124:213–217, 1995.
22. Jones SP, Greenway MJ, Orr NA. The influence of receptor fluid on in vitro percuta-
neous penetration. Int J Pharm 53:43–46, 1989.
23. Smith EW, Haigh JM. In vitro diffusion cell design and validation, Acta Pharm Nord
4:171–178, 1992.
24. Ehntholt DJ, Cerundolo DL, Bodek I, Schwope AD, Royer MD, Nielsen AP. A test
method for the evaluation of protective glove materials used in agricultural pesticide
operations. Am Ind Hyg Assoc J 51:462–468, 1990.
25. Leinster P, Bonsall JL, Evans MJ, Lloyd GA, Miller B, Rackham M. The development
of a standard test method for determining permeation of liquid chemicals through
protective clothing materials. Ann Occup Hyg 30:381–395, 1986.
26. Kou JH, Roy SD, Du J, Fujiki J. Effect of receiver fluid pH on in vitro skin flux of
weakly ionizable drugs. Pharm Res 10:986–990, 1993.
27. Wester RC. Relevance of animal models for percutaneous absorption. Int J Pharm 7:
99–110, 1980.
28. Bronaugh RL, Stewart RF, Congdon ER. Methods for in vitro percutaneous absorption
studies. II: Animal models for human skin. Toxicol Appl Pharmacol 62:481–488, 1982.
29. Reifenrath W. Evaluation of animal models for predicting skin penetration in man.
Fundam Appl Toxicol 4:S224–S230, 1984.
30. Walters KA, Roberts MS. Veterinary applications of skin penetration enhancers. In:
Walters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration Enhancement. New
York: Marcel Dekker, pp 345–364, 1993.
31. Southwell JD, Barry BW, Woodford R. Variations in permeability of human skin within
and between specimens. Int J Pharm 18:299–309, 1984.
32. Kasting GB, Filloon TG, Francis WR, Meredith MP. Improving the sensitivity of in
vitro skin penetration experiments. Pharm Res 11:1747–1754, 1994.
33. Cornwell PA, Barry BW. Effect of penetration enhancer treatment on the statistical
distribution of human skin permeabilities. Int J Pharm 117:101–112, 1995.
34. Williams AC, Cornwell PA, Barry BW. On the non-gaussian distribution of human skin
permeabilities. Int J Pharm 86:69–77, 1992.
35. Bronaugh RL, Stewart RF, Simon M. Methods for in vitro percutaneous absorption
studies VII: Use of excised human skin. J Pharm Sci 75:1094–1097, 1986.
36. Kasting GB, Bowman LA. Electrical analysis of fresh excised human skin: a compar-
ison with frozen skin. Pharm Res 7:1141–1146, 1990.
37. Harrison SM, Barry BW, Dugard PH. Effects of freezing on human skin permeability.
J Pharm Pharmacol 36:261–262, 1984.
38. Marzulli FN, Maibach HI. Permeability and reactivity of skin as related to age. J Soc
Cosmet Chem 35:95–102, 1984.
39. Roskos KV, Bircher AJ, Maibach HI, Guy RH. Pharmacodynamic measurements of
methyl nicotinate percutaneous absorption: the effect of aging on microcirculation. Br
J Dermatol 122:165–171, 1990.
40. Banks YB, Brewster DW, Birnbaum LS. Age-related changes in dermal absorption
of 2,3,7,8-tetrachlorodibenzo-p-dioxin and 2,3,4,7,8-pentachlorodibenzofuran. Fundam
Appl Toxicol 15:163–173, 1990.
41. Hall LL, Fisher HL, Sumler MR, Hughes MF, Shah PV. Age-related percutaneous
penetration of 2-sec-butyl-4,6-dinitrophenol (dinoseb) in rats. Fundam Appl Toxicol
19:258–267, 1992.
250 Brain et al.
42. Dick IP, Scott RC. The influence of different strains and age on in vitro rat skin
permeability to water and mannitol. Pharm Res 9:884–887, 1992.
43. Behl CR, Flynn GL, Linn EE, Smith WM. Percutaneous absorption of coritcosteroids:
age, site, and skin-sectioning influences on rates of permeation of hairless mouse skin
by hydrocortisone. J Pharm Sci 73:1287–1290, 1984.
44. Behl CR, Flynn GL, Kurihara T, Smith WM, Bellantone NH, Gatmaitan O, Higuchi
WI, Ho NFH, Pierson CL. Age and anatomical site influences on alkanol permeation
of skin of the male hairless mouse. J Soc Cosmet Chem 35:237–252, 1984.
45. Roskos KV, Maibach HI, Guy RH. The effect of ageing on percutaneous absorption
in man. J Pharm Biopharm 17:617–630, 1989.
46. Rougier A. Percutaneous absorption–transepidermal water loss relationship in man in
vivo. In: Scott RC, Guy RH, Hadgraft J, Boddé HE, eds. Prediction of Percutaneous
Penetration. Vol. 2. London: IBC Technical Services, pp 60–72, 1991.
47. Roskos KV, Guy RH, Maibach HI. Percutaneous absorption in the aged. Dermatol Clin
4:455–465, 1986.
48. Guy RH, Tur E, Bjerke S, Maibach HI. Are there age and racial differences to methyl
nicotinate-induced vasodilation in human skin? J Am Acad Dermatol 12:1001–1006,
1983.
49. Roskos KV, Maibach HI. Percutaneous absorption and age: implications for therapy.
Drugs Aging 2:432–449, 1992.
50. Wedig JH, Maibach HI. Percutaneous penetration of dipyrithione in man: effect of skin
color (race). J Am Acad Dermatol 5:433–438, 1981.
51. Berardesca E, Maibach HI. Racial differences in pharmacodynamic response to nico-
tinates in vivo in human skin: black and white. Acta Derm Venereol 70:63–66, 1990.
52. Leopold CS, Maibach HI. Effect of lipophilic vehicles on in vivo skin penetration of
methyl nicotinate in different races. Int J Pharm 139:161–167, 1996.
53. Weigand DA, Gaylor JR. Irritant reaction in negro and Caucasian skin. South Med J
67:548–551, 1974.
54. Kompaore F, Tsuruta H. In vivo differences between Asian, black and white in the
stratum corneum barrier function. Int Arch Occup Environ Health 65:S223–S225,
1993.
55. Kompaore F, Marty J–P, Dupont C. In vivo evaluation of the stratum corneum barrier
function in blacks, Caucasians and Asians with two noninvasive methods. Skin Phar-
macol 6:200–207, 1993.
56. Weigand DA, Haygood C, Gaylor JR. Cell layers and density of negro and Caucasian
SC. J Invest Dermatol 62:563–568, 1974.
57. Rienertson RP, Wheatley VR. Studies on the chemical composition of human epidermal
lipids. J Invest Dermatol 32:49–59, 1959.
58. Rougier A, Lotte C, Corcuff P, Maibach HI. Relationship between skin permeability
and corneocyte size according to anatomic site, age, and sex in man. J Soc Cosmet
Chem 39:15–26, 1988.
59. Corcuff P, Lotte C, Rougier A, Maibach HI. Racial differences in corneocytes. A com-
parison between black, white and Oriental skin. Acta Derm Venereol 71:146–148,
1991.
60. Berardesca E, Maibach HI. Racial differences in sodium lauryl sulphate induced cu-
taneous irritation: black and white. Contact Dermatitis 18:65–70, 1988.
61. Berardesca E, Maibach HI. Cutaneous reactive hyperaemia: racial differences induced
by corticoid application. Br J Dermatol 120:787–794, 1989.
62. Wickrema Sinha AJ, Shaw SR, Weber DJ. Percutaneous absorption and excretion of
tritium-labeled diflorasone diacetate, a new topical corticosteroid in the rat, monkey
and man. J Invest Dermatol 71:372–377, 1978.
63. Lotte C, Wester RC, Rougier A, Maibach HI. Racial differences in the in vivo percu-
taneous absorption of some organic compounds: a comparison between black, Cauca-
sian and Asian subjects. Arch Dermatol Res 284:456–459, 1993.
64. Yazdanian M. The effect of freezing on cattle skin permeability. Int J Pharm 103:93–
96, 1994.
65. Wester RC, Christoffel J, Hartway T, Poblete N, Maibach HI, Forsell J. Human cadaver
skin viability for in vitro percutaneous absorption: storage and detrimental effects of
heat-separation and freezing. Pharm Res 15:82–84, 1998.
66. Fares HM, Zatz JL. Dual-probe method for assessing skin barrier integrity: effect of
storage conditions on permeability of micro-Yucatan pig skin. J Soc Cosmet Chem 48:
175–186, 1997.
67. Feldman RJ, Maibach HI. Regional variation in the percutaneous penetration of 14C
cortisol in man. J Invest Dermatol 48:181–183, 1967.
68. Maibach HI, Feldman RJ, Mitby TH, Serat WF. Regional variation in percutaneous
penetration in man: pesticides. Arch Environ Health 23:208–211, 1971.
69. Wester RC, Noonan PK, Maibach HI. Variations in percutaneous absorption of testos-
terone in the rhesus monkey due to anatomic site of application and frequency of
application. Arch Dermatol Res 267:229–235, 1980.
70. Wester RC, Maibach HI. Cutaneous pharmacokinetics: 10 steps to percutaneous ab-
sorption. Drug Metab Rev 14:169–205, 1983.
71. Rougier A, Roguet R. Regional variation in percutaneous absorption in man: measure-
ment by the tape stripping method. Arch Dermatol Res 278:465–469, 1986.
72. Rougier A. In vivo percutaneous penetration of some organic compounds related to
anatomic site in humans: stripping method. J Pharm Sci 76:451–454, 1987.
73. Tur E, Maibach HI, Guy RH. Percutaneous penetration of methyl nicotinate at three
anatomical sites: evidence for an appendageal contribution to transport? Skin Phar-
macol 4:230–234, 1991.
74. Lotte CA, Wilson DR, Maibach HI. In vivo relationship of transepidermal water loss
and percutaneous absorption in man: effect of anatomical site. Arch Dermatol Res 279:
351–356, 1987.
75. Roberts MS, Favretto WA, Meyer A, Reckmann M, Wongseelashote T. Topical bio-
availability of methyl salicylate. Aust NZ J Med 12:303–305, 1982.
76. Trebilcock KL, Heylings JR, Wilks MF. In vitro tape stripping as a model for in vivo
skin stripping. Toxicol In Vitro 8:665–667, 1994.
77. Scheuplein RJ. Mechanism of percutaneous absorption I. Routes of penetration and
influence of solubility. J Invest Dermatol 45:334–346, 1965.
78. Kumar S, Behl CR, Patel SB, Malick AW. A simple microwave technique for the
separation of epidermis and dermis in skin uptake studies. Pharm Res 6:740–741, 1989.
79. Boddé HE, Ponec M, Ijzerman AP, Hoogstraate AJ, Salomons MAI, Bouwstra JA. In
vitro analysis of QSAR in wanted and unwanted effects of azacycloheptanones as
transdermal penetration enhancers. In: Walters KA, Hadgraft J, eds. Pharmaceutical
Skin Penetration Enhancement. New York: Marcel Dekker, pp 199–214, 1993.
80. Brain KR, Walters KA, James VJ, Dressler WE, Howes D, Kelling CK, Moloney SJ,
Gettings SD. Percutaneous penetration of dimethylnitrosamine through human skin in
vitro: application from cosmetic vehicles. Food Chem Toxicol 33:315–322, 1995.
81. Walters KA, Brain KR, Howes D, James VJ, Kraus AL, Teetsel NM, Toulon M, Wat-
kinson AC, Gettings SD. Percutaneous penetration of octyl salicylate from represen-
tative sunscreen formulations through human skin in vitro. Food Chem Toxicol 35:
1219–1225, 1997.
82. Franz TJ. The finite dose technique as a valid in vitro model for the study of percu-
taneous absorption in man. Curr Probl Dermatol 7:58–68, 1978.
252 Brain et al.
83. Franz TJ, Lehman PA, Franz SF, North–Root H, Demetrulias JL, Kelling CK, Moloney
SJ, Gettings SD. Percutaneous penetration of N-nitrosodiethanolamine through human
skin (in vitro): comparison of finite and infinite dose applications from cosmetic ve-
hicles. Fundam Appl Toxicol 21:213–221, 1993.
84. Walters KA, Brain KR, Dressler WE, Green DM, Howes D, James VJ, Kelling CK,
Watkinson AC, Gettings SD. Percutaneous penetration of N-nitroso-N-methyldodecyl-
amine through human skin in vitro: application from cosmetic vehicles. Food Chem
Toxicol 35:705–712, 1997.
85. Akhter SA, Barry BW. Absorption through human skin of ibuprofen and flurbiprofen;
effect of dose variation, deposited drug films, occlusion and the penetration enhancer
N-methyl-2-pyrrolidone. J Pharm Pharmacol 37:27–37, 1985.
86. Tata S, Flynn GL, Weiner ND. Penetration of minoxidil from ethanol/propylene glycol
solutions: effect of application volume and occlusion. J Pharm Sci 84:688–691, 1995.
87. Roper CS, Howes D, Blain PG, Williams FM. Percutaneous penetration of 2-phenoxy-
ethanol through rat and human skin. Food Chem Toxicol 35:1009–1016, 1997.
88. Treffel P, Gabard B. Ibuprofen epidermal levels after topical application in vitro: effect
of formulation, application time, dose variation and occlusion. Br J Dermatol 129:286–
291, 1993.
89. Hughes MF, Shrivastava SP, Sumler MR, Edwards BC, Goodwin JH, Shah PV, Fisher
HL, Hall HL. Dermal absorption of chemicals: effect of application of chemicals as a
solid, aqueous paste, suspension, or in volatile vehicle. J Toxicol Environ Health 37:
57–71, 1992.
90. Kubota K, Yamada T. Finite dose percutaneous drug absorption: theory and its appli-
cation to in vitro timolol permeation. J Pharm Sci 79:1015–1019, 1990.
91. Seta AH, Higuchi WI, Borsadia S, Behl CR, Malick AW. Physical model approach to
understanding finite dose transport and uptake of hydrocortisone in hairless guinea pig
skin. Int J Pharm 81:89–99, 1992.
92. Walker M, Chambers LA, Holingsbee DA, Hadgraft J. Significance of vehicle thickness
to skin penetration of halcinonide. Int J Pharm 70:167–172, 1991.
93. Addicks WJ, Flynn GL, Weiner N, Chiang C–M. Drug transport from thin applications
of topical dosage forms: development of methodology. Pharm Res 5:377–382, 1988.
94. Ferry JJ, Shepard JH, Szpunar GL. Relationship between contact time of applied dose
and percutaneous absorption of minoxidil from a topical solution. J Pharm Sci 79:483–
486, 1990.
95. Peltonen L, Solberg VM. A comparison of 0.1% hydrocortisone cream (Milliderm
0.1%; AL) and 1% hydrocortisone cream (hydrocortisone; 1% Orion) in patients with
non-infected eczemas. Curr Ther Res Clin Exp 35:78–82, 1984.
96. Davis AF, Hadgraft J. The use of supersaturation in topical drug delivery. In: Scott
RC, Guy RH, Hadgraft J, Boddé HE, eds. Prediction of Percutaneous Penetration. Vol.
2. London: IBC Technical Services, pp 279–287, 1991.
97. Potts RO, Guy RH. Drug transport across the skin and the attainment of steady-state
fluxes. Proc Int Symp Controlled Release Bioact Mater 21:162–163, 1994.
98. Peck KD, Ghanem AH, Higuchi WI, Srinivasan V. Improved stability of the human
epidermal membrane during successive permeability experiments. Int J Pharm 98:141–
147, 1993.
99. Sasaki H, Nakamura J, Shibasaki J, Ishino Y, Miyasato K, Ashizawa T. Effect of skin
surface temperature on transdermal absorption of flurbiprofen from a cataplasm. Chem
Pharm Bull 35:4883–4890, 1987.
100. Potts RO, Francoeur ML. Lipid biophysics of water loss through the skin. Proc Natl
Acad Sci USA 87:1–3, 1990.
101. Emilson A, Lindberg M, Forslind B. The temperature effect of in vitro penetration of
sodium lauryl sulfate and nickel chloride through human skin. Acta Derm Venereol 73:
203–207, 1993.
102. Ohara N, Takayama K, Machida Y, Nagai T. Combined effect of d-limonene and tem-
perature on the skin permeation of ketoprofen. Int J Pharm 105:31–38, 1994.
103. Bouwstra JA, Gooris GS, Salomons–deVries MA, van der Spek JA, Bras W. Structure
of human SC as a function of temperature and hydration: a wide-angle X-ray diffraction
study. Int J Pharm 84:205–216, 1992.
104. Benech–Kieffer F, Wegrich P, and Schaefer H. Transepidermal water loss as an integrity
test for skin barrier function in vitro: assay standardization. In: Brain KR, James VJ,
Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5a. Cardiff: STS Pub-
lishing, p 56, 1997.
105. Hood HL, Wickett RR, Bronaugh RL. In vitro percutaneous absorption of the fragrance
ingredient musk xylol. Food Chem Toxicol 34:483–488, 1996.
106. Pendlington RU, Sanders DJ, Cooper KJ, Howes D, Lovell WW. The use of sucrose
as a standard penetrant in in vitro percutaneous penetration experiments. In: Brain KR,
James VJ, Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5a. Cardiff:
STS Publishing, p 55, 1997.
107. Brain KR. Adopting an appropriate analytical approach. In: Scott RC, Guy RH, Had-
graft J, Boddé HE, eds. Prediction of Percutaneous Penetration. Vol. 2. London: IBC
Technical Services, pp 229–237, 1991.
108. Watkinson AC, Green DM, Brain KR, James VJ, Walters KA, Azri–Meehan S, Dressler
W. Nonoxynols: skin penetration of a series of homologues. In: Brain KR, James VJ,
lishing, p 22, 1997.
109. Gibson GG, Skett P. Introduction to Drug Metabolism. 2nd ed. London: Chapman &
Hall, 1994.
110. Merk HF, Jugert FK, Frankenburg S. Biotransformations in the skin. In: Marzulli FN,
Maibach HI, eds. Dermatotoxicology, 5th ed. Washington: Taylor & Francis, pp 61–
73, 1996.
111. Hotchkiss SAM. Dermal metabolism. In: Roberts MS, Walters KA, eds. Dermal Ab-
112. Sharma R. Xenobiotic metabolizing enzymes in skin. In: Brain KR, James VJ, Walters
KA, eds. Prediction of Percutaneous Penetration. Vol. 4b. Cardiff: STS Publishing, pp
14–18, 1996.
113. Bronaugh RL, Stewart RF, Storm JE. Extent of cutaneous metabolism during percu-
taneous absorption of xenobiotics. Toxicol Appl Pharmacol 99:534–543, 1989.
114. Bronaugh RL, Kraelling MEK. In vitro skin metabolism. In: Brain KR, James VJ,
Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 4b. Cardiff: STS Pub-
lishing, pp 19–21, 1996.
115. Kawakubo Y, Manabe S, Yamazoe Y, Nishikawa T, and Kato R. Properties of cutaneous
acetyltransferase catalysing N- and O-acetylation of carcinogenic arylamines and N-
hydroxyarylamine. Biochem Pharmacol 37:265–270, 1988.
116. Fuchs J, Mehlorn RJ, Packer L. Free radical reduction mechanism in mouse epidermis
skin homogenates. J Invest Dermatol 93:633–640, 1989.
117. Hotchkiss SAM. Skin as a xenobiotic metabolising organ. Prog Drug Metab 13:217–
262, 1992.
118. Andersson P, Edsbäcker S, Ryerfeldt Å, Von Bahr C. In vitro biotransformation of
glucocorticoids in liver and skin homogenate fraction from man, rat, and hairless
mouse. J Steroid Biochem 16:787–795, 1982.
119. Cheung YW, Li Wan Po A, Irwin WJ. Cutaneous biotransformation as a parameter in
the modulation of the activity of topical corticosteroids. Int J Pharm 26:175–189, 1985.
120. Kulkarni AP, Nelson JL, Radulovic LL. Partial purification and some biochemical prop-
erties of neonatal rat cutaneous glutathione S-transferases. Comp Biochem Physiol 87B:
1005–1009, 1987.
254 Brain et al.
121. Moss T, Howes D, Blain PG, Williams FM. Characteristics of sulphotransferases in

human skin. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous
Penetration. Vol. 4b. Cardiff: STS Publishing, pp 307–310, 1996.
122. Coomes MW, Norling AH, Pohl RJ, Müller D, Fouts JR. Foreign compound metabo-
lism by isolated skin cells from the hairless mouse. J Pharmacol Exp Ther 225:770–
777, 1983.
123. Bronaugh RL. Methods for in vitro skin metabolism studies. In: Marzulli FN, Maibach
HI, eds. Dermatotoxicology, 5th ed. Washington: Taylor & Francis, pp 383–388 (1996).
124. Collier SW, Sheikh NM, Sakr A, Lichtin JL, Stewart RF, Bronaugh RL. Maintenance
of skin viability during in vitro percutaneous absorption/metabolism studies. Toxicol
Appl Pharmacol 99:522–533, 1989.
125. Nathan D, Sakr A, Lichtin JL, Bronaugh RL. In vitro skin absorption and metabolism
of benzoic acid, p-aminobenzoic acid, and benzocaine in the hairless guinea pig. Pharm
Res 7:1147–1151, 1990.
126. Storm JE, Collier SW, Stewart RF, Bronaugh RL. Metabolism of xenobiotics during
percutaneous penetration: role of absorption rate and cutaneous enzyme activity. Fun-
dam Appl Toxicol 15:132–141, 1993.
127. Collier SW, Storm JE, Bronaugh RL. Reduction of azo dyes during in vitro percuta-
neous absorption. Toxicol Appl Pharmacol 118:73–79, 1993.
128. Ademola JI, Wester RC, Maibach HI. Absorption and metabolism of 2-chloro-2,6-
diethyl-N-(butoxymethyl)acetanilide (butachlor) in human skin in vitro. Toxicol Appl
Pharmacol 121:78–86, 1993.
129. Ademola JI, Sedik LE, Wester RC, Maibach HI. In vitro percutaneous absorption and
metabolism in man of 2-chloro-4-ethylamino-6-isopropylamine-s-triazine (atrazine).
Arch Toxicol 67:85–91, 1993.
130. Boehnlein J, Sakr S, Lichtin JL, Bronaugh RL. Metabolism of retinyl palmitate to
retinol (vitamin A) in skin during percutaneous absorption. Pharm Res 11:1155–1159,
1994.
131. Slivka SR. Testosterone metabolism in an in vitro skin model. Cell Biol Toxicol 8:
267–276, 1992.
132. Slivka SR, Landeen LK, Zeigler F, Zimber MP, Bartel RI. Characterisation, barrier
function, and drug metabolism of an in vitro skin model. J Invest Dermatol 100:40–
46, 1993.
133. Crank J. The Mathematics of Diffusion. 2nd ed. Oxford: Oxford University Press, 1975.
134. Pershing LK, Silver BS, Krueger GG, Shah VP, Skelly JP. Feasibility of measuring the
bioavailability of topical betamethasone dipropionate in commercial formulations using
drug content in skin and a skin blanching bioassay. Pharm Res 9:45–51, 1992.
135. Pershing LK, Corlett J, Jorgensen C. In vivo pharmacokinetics and pharmacodynamics
of topical ketoconazole and miconazole in human SC. Antimicrob Agents Chemother
38:90–95, 1994.
136. Schaefer H, Stuttgen C, Zesch A, Schalla W, Gazith J. Quantitative determination of
percutaneous absorption of radiolabelled drugs in vitro and in vivo by human skin.
Curr Prob Dermatol 7:80–94, 1978.
137. King CS, Barton PS, Nicholls S, Marks R. The change in properties of the SC as a
function of depth. Br J Dermatol 100:165–172, 1979.
138. Guy RH, Mak VHW, Kai T, Bommannan D, Potts RO. Percutaneous penetration en-
hancers: mode of action. In: Scott RC, Guy RH, Hadgraft J, eds. Prediction of Per-
cutaneous Penetration. Vol. 1. London: IBC Technical Services, pp 213–223, 1990.
139. Bommannan D, Potts RO, Guy RH. Examination of the SC barrier function in vivo
by infrared spectroscopy. J Invest Dermatol 95:403–408, 1990.
140. Watkinson AC, Bunge AL, Hadgraft J, Niak A. Computer simulation of penetrant–
depth profiles in the SC. Int J Pharm 87:175–182, 1992.
141. Marttin E, Neelissen–Subnel MTA, De Haan FHN, and Boddé HE. A critical compar-
ison of methods to quantify SC removed by tape stripping. Skin Pharmacol 9:69–77,
1996.
142. van Hoogdalem EJ. Assay of erythromycin in tape strips of human SC and some
preliminary results in man. Skin Pharmacol 5:124–128, 1992.
143. Clark EW. Short-term penetration of lanolin into human SC. J Soc Cosmet Chem 43:
219–227, 1992.
144. Jadoul A, Hanchard C, Thysman S, Préat V. Quantification and localisation of fentanyl
and TRH delivered by iontophoresis in the skin. Int J Pharm 120:221–228, 1995.
145. Jadoul A, Mesens J, Caers W, de Beukelaar F, Crabbé R, Préat V. Transdermal per-
meation of alniditan by iontophoresis: in vitro optimisation and human pharmacokinetic
data. Pharm Res 13:1348–1353, 1996.
146. Azri–Meehan S, Grabarz R, Dressler WE. Hair dyes: aspects of risk assessment. In:
Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol.
4b. Cardiff: STS Publishing, pp 22–25, 1996.
147. Auton TR. Skin stripping and science: a mechanistic interpretation using mathematical
modelling of skin deposition as a predictor of total absorption. In: Scott RC, Guy RH,
Hadgraft J, eds. Prediction of Percutaneous Penetration. Vol. 2. London: IBC Technical
Services, pp 558–576, 1992.
148. Tsai J–C, Weiner ND, Flynn GL, Ferry J. Properties of adhesive tapes used for SC
stripping. Int J Pharm 72:227–231, 1991.
149. Tsai J–C, Cappel MJ, Weiner ND, Flynn GL, Ferry J. Solvent effects on the harvesting
of SC from hairless mouse skin through adhesive tape stripping in vitro. Int J Pharm
68:127–133, 1991.
SA, Kawamoto J, Lesko LJ, Marty J–P, Pershing LK, Schaefer H, Sequeira JA, Shri-
vastava SP, Wilkin J, Williams RL. AAPS/FDA workshop report: bioequivalence of
topical dermatological dosage forms—methods of evaluation of bioequivalence. Pharm
Res 15:167–171, 1998.
151. Göpferich A, Lee G. Measurement of drug diffusivity in SC membranes and a poly-
acrylate matrix. Int J Pharm 71:245–253, 1991.
152. Brandt H. Determination of diffusion specific parameters by means of IR-ATR spec-
troscopy. Exp Tech Phys 33:423–431, 1985.
153. Hemmelmann K, Brandt H. Investigation of diffusion and sorption properties of poly-
ethylene films for different liquids by means of IR–ATR spectroscopy. Exp Tech Phys
34:439–446, 1986.
154. Wurster DE, Buraphacheep V, Patel JM. The determination of diffusion coefficients in
semisolids by Fourier transform infrared (FTIR) spectroscopy. Pharm Res 10:616–620,
1993.
155. Tralhão AM, Watkinson AC, Brain KR, Hadgraft J, Armstrong NA. Use of ATR-FTIR
spectroscopy to study the diffusion of ethanol through glycerogelatin films. Pharm Res
12:1–4, 1995.
156. Farinas KC, Doh L, Venkatraman S, Potts RO. Characterisation of solute diffusion in
a polymer using ATR-FTIR spectroscopy. Macromolecules 27:5220–5222, 1994.
157. Watkinson AC, Hadgraft J, Walters KA, Brain KR. Measurement of diffusional param-
eters in membranes using ATR-FTIR spectroscopy. Int J Cosmet Sci 16:199–210, 1994.
158. Watkinson AC, Brain K, Hadgraft J. The deconvolution of diffusion and partition co-
efficients in permeability studies. Proc Int Symp Controlled Release Bioact Mater 21:
160–161, 1994.
159. Harrison J, Watkinson AC, Green DM, Brain KR, Hadgraft J. The relative effect of
Azone and Transcutol on diffusivity and solubility in human stratum corneum. Pharm
Res 13:542–546, 1996.
256 Brain et al.
160. Pellett MA, Watkinson AC, Hadgraft J, Brain KR. Comparison of permeability data
from traditional diffusion cells and ATR-FTIR spectroscopy. Part I—synthetic mem-
branes. Int J Pharm 154:205–215, 1997.
161. Pellett MA, Watkinson AC, Hadgraft J, Brain KR. Comparison of permeability data
from traditional diffusion cells and ATR-FTIR spectroscopy. Part II—determination of
diffusional pathlengths in synthetic membranes and human stratum corneum. Int J
Pharm 154:217–227, 1997.
162. Williams PL, Carver MP, Riviere JE. A physiologically relevant pharmacokinetic model
of xenobiotic percutaneous absorption utilizing the isolated perfused porcine skin flap.
J Pharm Sci 79:305–311, 1990.
163. Williams PL, Riviere JE. Definition of a physiologic pharmacokinetic model of cuta-
neous drug distribution using the isolated perfused porcine skin flap. J Pharm Sci 78:
550–555, 1989.
164. Riviere JE. Isolated perfused porcine skin flap. In: Marzulli FN, Maibach HI, eds.
Dermatotoxicology. 5th ed. Washington: Taylor & Francis, pp 337–351, 1996.
165. Riviere JE, Monteiro–Riviere N, Williams PL. Isolated perfused porcine skin flap as
an in vitro model for predicting transdermal kinetics. Eur J Pharm Biopharm 41:152–
162, 1995.
166. Keitzmann M, Löscher W, Arens D, Maaß D, Lubach D. The isolated perfused bovine
udder as an in vitro model of percutaneous absorption: skin viability and percutaneous
absorption of dexamethasone, benzoyl peroxide and etofenamate. J Pharmacol Toxicol
Methods 30:75–84, 1993.
167. Keitzmann M, Blume B, Unger B. Percutaneous absorption of betamethasone dipro-
pionate from different formulations using the isolated perfused bovine udder. In: Brain
KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 4b. Car-
diff: STS Publishing, pp 262–266, 1996.
168. Henrikus B, Kampffmeyer HG. Metabolism of ethyl 4-amino-benzoate in the isolated
single-pass perfused rabbit ear. Skin Pharmacol 6:246–252, 1993.
169. Mura P, Nassini C, Valoti M, Santoni G, Corti P. The single-pass perfused rabbit ear
as a model for studying percutaneous absorption of clonazepam. III. Influence of ve-
hicle composition on drug permeation. Eur J Pharm Biopharm 40:90–95, 1994.
170. De Lange J, Van Eck P, Elliott GR, De Kort WLAM, Wolthuis OL. The isolated blood-
perfused pig ear: an inexpensive and animal-saving model for skin penetration studies.
J Pharmacol Toxicol Methods 27:71–77, 1992.
171. Prunieras M. Human skin equivalents: the state of the art. In: Brain KR, James VJ,
lishing, pp 419–427, 1993.
172. Ernesti AM, Swiderek M, Gay R. Absorption and metabolism of topically applied
testosterone in an organotypic skin culture. Skin Pharmacol 5:146–153, 1992.
173. Bell E, Parenteau N, Gay R, Nolte C, Kemp P, Bilvo P, Ekstein B, Johnson E. The
living skin equivalent: its manufacture, its organotypic properties and its response to
irritants. Toxicol In Vitro 5:591–596, 1991.
174. Regnier M, Caron D, Reichert U, and Schaefer H. Reconstructed human epidermis: a
model to study in vitro barrier function of the skin. Skin Pharmacol 5:49–56, 1992.
175. Chambin O, Teillaud E, Mikler C. Evaluation of Testskin LSE-100 as a model for
human skin: in vitro percutaneous absorption studies. In: Brain KR, James VJ, Walters
458–462, 1993.
176. Kennedy AH, Golden GM, Gay CL, Guy RH, Francoeur ML, Mak VHW. Stratum
corneum lipids of human epidermal keratinocyte air–liquid cultures: implications for
barrier function. Pharm Res 13:1162–1166, 1996.
177. Lotte C, Hinz RS, Rougier A, Guy RH. A reconstructed skin model (Testskin LSE)
for permeation studies. In: Brain KR, James VJ, Walters KA, eds. Perspectives in
Percutaneous Penetration. Vol. 5a. Cardiff: STS Publishing, p 61, 1997.
178. Lotte C, Zanini M, Patouillet C, Cabaillot AM, Messager A, Cottin M, Leclaire J.
Comparison of different commercial reconstructed skin or epidermis models: permea-
tion studies. In: Brain KR, James VJ, Walters KA, eds. Perspectives in Percutaneous
Penetration. Vol. 5a. Cardiff: STS Publishing, p. 62, 1997.
179. Marty P, Faure C, Laroche F, Farenc C. Assessment of human skins obtained by in
vitro culture as membrane models for cutaneous permeation tests. In: Brain KR, James
VJ, Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5a. Cardiff: STS
Publishing, p. 64, 1997.
180. Dusser I, Noel–Hudson MS, Wepierre J. Improvement of epidermal differentiation and
barrier function in reconstructed human epidermis cultured in insert at air–liquid in-
terface. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous Pene-
tration. Vol. 4b. Cardiff: STS Publishing, pp. 315–318, 1996.
181. Higounec I, Demarchez M, Regnier M, Schmidt R, Ponec M, Schroot B. Improvement
of epidermal differentiation and barrier function in reconstructed human skin after
grafting onto athymic nude mice. Arch Dermatol Res 286:1–8, 1994.
182. Bidman HJ, Pitts JD, Solomon HF, Bondil JV, Stumpf WE. Estradiol distribution and
penetration in rat skin after topical application, studied by high resolution autoradiog-
raphy. Histochemistry 95:43–54, 1990.
183. Fabin B, Touitou E. Localization of lipophilic molecules penetrating rat skin in vivo
by quantitative autoradiography. Int J Pharm 74:59–65, 1991.
184. Ritchel WA, Panchagnula R, Stemmer K, Ashraf M. Development of an intracutaneous
depot for drugs. Binding, drug accumulation and retention studies and mechanism of
depot. Skin Pharmacol 4:235–245, 1991.
185. Conti L, Ramis J, Mis R, Forn J, Vilaro S, Reina M, Vilageliu J, Basi N. Percutaneous
absorption and skin distribution of 14C-flutrimazole in mini-pigs. Drug Res 42:847–
853, 1992.
186. Vingler PF, Bague H, Pruche F, Kermicic M. Direct quantitative digital autoradiography
of testosterone metabolites in the pilosebaceous unit—an environmentally advanta-
geous trace radioactive technology. Steroids 58:429–438, 1993.
187. Chu ID, Bronaugh RL, Tryphonas L. Skin reservoir formation and bioavailability of
dermally administered chemicals in hairless guinea pigs. Food Chem Toxicol 34:267–
276, 1996.
188. Cullander C. Confocal microscopy in the study of skin permeation: utility and limi-
tations. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetra-
tion. Vol. 4b. Cardiff: STS Publishing, pp 5–6, 1996.
189. Rojanasakul Y, Paddock SW, Robinson JR. Confocal laser scanning microscopic ex-
amination of transport pathways and barriers of some peptides across the cornea. Int J
Pharm 61:163–172, 1990.
190. Imbert D, Cullander C. Assessment of cornea viability by confocal laser scanning
microscopy and MTT assay. Cornea 16:666–674, 1997.
191. Cullander C, Guy RH. Visualising the pathway of iontophoretic current flow in real
time with laser-scanning confocal microscopy and the vibrating probe electrode. In:
Scott RC, Guy RH, Hadgraft J, Boddé HE, eds. Prediction of Percutaneous Penetration.
Vol. 2. London: IBC Technical Services, pp. 229–237, 1991.
192. Simoneti O, Hoogstraate AJ, Bialik V, Kempenaar JA, Schrijvers A, Boddé HE, Ponec
M. Visualisation of diffusion pathways through the SC of native and of in vitro recon-
structed epidermis by confocal laser scanning microscopy. Arch Dermatol Res 287:
465–473, 1995.
258 Brain et al.
193. Prausnitz MR, Gimm JA, Guy RH, Langer R, Weaver JC, Cullander C. Imaging regions
of transport across human stratum corneum during high-voltage and low-voltage ex-
posure. J Pharm Sci 85:1363–1370, 1996.
194. Boderke P, Merkle HP, Cullander C, Ponec M, Boddé HE. Localization of aminopep-
tidase activity in freshly excised human skin: direct visualization by confocal laser
scanning microscopy. J Invest Dermatol 108:83–86, 1997.
195. Hakkarainen M, Jaaskelainen I, Suhonen M, Paronen P, Urtii A, Monkonen J. In vitro
penetration of liposomes into the skin studied by confocal microscopy. In: Brain KR,
James VJ, Walters KA, eds. Prediction in Percutaneous Penetration. Vol. 4b. Cardiff:
STS Publishing, pp 42–44, 1996.
196. Hoogstraate AJ, Cullander C, Nagelkerke JF, Senel S, Verhoef JC, Junginger HE,
Boddé HE. Diffusion rates and transport pathways of FITC-labelled model compounds
through buccal epithelium. Pharm Res 11:83–89, 1994.
197. Hoogstraate AJ, Cullander C, Nagelkerke JF, Spies F, Verhoef JC, Schrijvers AHG,
Junginger HE, Boddé HE. A novel in situ model for continuous observation of transient
drug concentration gradients across buccal epithelium at the microscopical level. J
Controlled Release 39:71–78, 1996.
198. Hoogstraate AJ, Senel S, Cullander C, Verhoef JC, Junginger HE, Boddé HE. Effects
of bile salts on transport rates and routes of FITC-labelled compounds across porcine
buccal epithelium in vitro. J Controlled Release 40:211–221, 1996.
199. Marttin E, Verhoef JC, Cullander C, Romeijn SG, Nagelkerke JF, Merkus FWHM.
Confocal laser scanning microscopic visualization of the transport of dextrans after
nasal administration to rats: effects of absorption enhancers. Pharm Res 14:631–637,
1997.
200. Melia CD, Cutts LS, Adler J, Davies MC, Hibberd S, Rajabi–Siahboomi AR, Bowtell
R. Visualising and measuring dynamic events inside controlled release matrix systems.
Proc Int Symp Controlled Release Bioact Mater 22:32–33, 1995.
201. Cutts LS, Roberts PA, Adler J, Davies MC, and Melia CD. The measurement of dif-
fusion coefficients in gels using the confocal laser scanning microscope. In: Hadgraft
J, Kellaway IW, Parr GD, eds. Proceedings of the 3rd UKaps Conference, p 54, 1994.
measurements. I. Isopropyl myristate. Int J Pharm 39:149–156, 1987.
measurements. I. Dipalmitoyl phosphatidylcholine, linoleic acid and tetradecane. Int J
Pharm 42:97–104, 1988.
204. Firestone BA, Guy RH. Prediction of dermal absorption. In: Alternative Methods in
Toxicology. New York: Mary Ann Liebert, pp 517–536, 1985.
205. Hadgraft J, Walters KA, Wotton PK. Facilitated transport of sodium salicylate across
an artificial lipid membrane by Azone. J Pharm Pharmacol 37:725–727, 1985.
206. Houk J, Guy RH. Membrane models for skin penetration studies. Chem Rev 88:455–
471, 1988.
207. Hadgraft J, Guy RH. Synthetic membranes as biological models. In: Ganderton D,
Jones T, eds. Advances in Pharmaceutical Sciences, Vol. 6. London: Academic Press,
pp 43–64, 1992.
208. Pellett MA, Watkinson AC, Hadgraft J, Brain KR. An ATR-FTIR investigation of the
interactions between vehicles and a synthetic membrane. Proc Int Symp Controlled
Release Bioact Mater 21:439–440, 1994.
209. Neubert R, Bendas C, Wohlrab W, Glenau B, Furst W. A multilayer membrane system
for modelling drug penetration into skin. Int J Pharm 75:89–94, 1991.
210. Neubert R, Wohlrab W, Bendas C. Modelling of drug penetration into human skin
using a multilayer membrane system. Skin Pharmacol 8:119–129, 1995.
211. Faustino EPR, Cabral Marques HM, Morais JA, Hadgraft J. Comparison of different
models for the in vitro release of topical dosage forms. In: Brain KR, James VJ, Walters
294–298, 1996.
212. Nastruzzi C, Esposito E, Pastesini C, Gambari R, Menegatti E. Comparative study on
the release kinetics of methyl nicotinate from topical formulations. Int J Pharm 90:43–
50, 1993.
213. Abraham W, Downing DT. Preparation of model membranes for skin permeability
studies using stratum corneum lipids. J Invest Dermatol 93:809–813, 1989.
214. Matsuzaki K, Imaoka T, Asano M, Miyajima K. Development of a model membrane
system using stratum corneum lipids for estimation of drug skin permeability. Chem
Pharm Bull 41:575–579, 1993.
215. Miyajima K, Tanikawa S, Asano M, Matsuzaki K. Effects of absorption enhancers and
lipid composition on drug permeability through the model membrane using stratum
corneum lipids. Chem Pharm Bull 42:1345–1347, 1994.
216. Little CJ, Brain KR. Development of stratum corneum model membrane systems and
comparison of the penetration rates of certain NSAIDs. In: Brain KR, James VJ, Wal-
ters KA, eds. Prediction in Percutaneous Penetration. Vol. 4b. Cardiff: STS Publishing,
pp 290–293, 1996.
217. Moghimi HR, Williams AC, Barry BW. A lamellar matrix model for stratum corneum
intercellular lipids. I. Characterisation and comparison with stratum corneum intercel-
lular structure. Int J Pharm 131:103–115, 1996.
218. Moghimi HR, Williams AC, Barry BW. A lamellar matrix model for stratum corneum
intercellular lipids. I. Effect of geometry of the stratum corneum on permeation of
model drugs 5-fluorouracil and oestradiol. Int J Pharm 131:117–129, 1996.
219. Little CJ, Brain KR. Stratum corneum models based on lecithin organogels. In: Brain
KR, James VJ, Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5b.
Cardiff: STS Publishing, pp 147–148, 1998.
220. Little CJ, Brain KR. Comparison of some alternative model membrane systems for the
estimation of the permeation of ibuprofen from commercial formulations through hu-
man skin. In: Brain KR, James VJ, Walters KA, eds. Perspectives in Percutaneous
221. Little CJ, Brain KR. Investigation of chemical modulation of membrane permeation in
different model membrane systems. In: Brain KR, James VJ, Walters KA, eds. Per-
spectives in Percutaneous Penetration. Vol. 5b. Cardiff: STS Publishing, pp 151–152,
1998.
222. Pershing LK, Kreuger GG. New animal models for bioavailability studies. Skin Phar-
macol 1:57–69, 1987.
223. Wester RC, Maibach HI. Animal models for percutaneous absorption. In: Shah VP,
Maibach HI, eds. Topical Drug Bioavailability, Bioequivalence, and Penetration. New
York: Plenum Press, pp 333–349, 1993.
224. Guy RH, Hadgraft J, Hinz RS, Roskos KV, Bucks DAW. In vivo evaluations of trans-
dermal drug delivery. In: Chien YW, ed. Transdermal Controlled Systemic Medications.
New York: Marcel Dekker, pp 179–224, 1987.
225. Scott RC, Corrigan MA, Smith F, Mason H. The influence of skin structure on per-
meability: an intersite and interspecies comparison with hydrophilic penetrants. J Invest
Dermatol 96:921–925, 1991.
226. Scott RC. Personal communication.
227. Gray GM, White RJ. Glycosphingolipids and ceramides in human and pig epidermis.
228. Wertz PW, Downing DT. Ceramides of pig epidermis: structure determination. J Lipid
Res 24:759–765, 1983.
260 Brain et al.
229. Wertz PW, Downing DT. Glucosylceramides of pig epidermis: structure determination.
J Lipid Res 24:1135–1139, 1983.
230. Sato K, Sugibayashi K, Morimoto Y. Species difference in percutaneous absorption of
nicorandil. J Pharm Sci 80:104–107, 1991.
231. Lloyd DH, Dick WDB, McEwan–Jenkinson D. The effects of some surface sampling
procedures on the stratum corneum of bovine skin. Res Vet Sci 26:250–252, 1979.
232. Davies B, Morris T. Physiological parameters in laboratory animals and humans. Pharm
Res 10:1093–1095, 1993.
233. Wester RC, Noonan PK. Relevance of animal models for percutaneous absorption. Int
J Pharm 7:99–110, 1980.
234. Reifenrath WG, Chellquist EM, Shipwash EA, Jederberg WW, Kreuger GG. Percuta-
neous penetration in the hairless dog, weanling pig and grafted athymic nude mouse:
evaluation of models for predicting skin penetration in man. Br J Dermatol 111(suppl
27):123–135, 1984.
235. Bond JR, Barry BW. Hairless mouse skin is limited as a model for assessing the effects
of penetration enhancers in human skin. J Invest Dermatol 90:810–813, 1988.
236. Choi H–K, Flynn GL, Amidon GL. Transdermal delivery of bioactive peptides: the
effect of n-decylmethylsulphoxide, pH, and inhibitors on enkaphalin metabolism and
transport. Pharm Res 7:1099–1106, 1990.
237. Norgaard O. Investigations with radiolabeled nickel, cobalt and sodium on the resorp-
tion through the skin in rabbits, guinea pigs and man. Acta Derm Venereol 34:440,
1957.
238. Tregear RT. Physical Functions of Skin. New York: Academic Press, 1966.
239. McCreesh AH. Percutaneous toxicity. Toxicol Appl Pharmacol 7(suppl 2):20–26, 1965.
240. Scott RC, Walker M, Dugard PH. A comparison of the in vitro permeability properties
of human and some laboratory animal skins. Int J Cosmet Sci 8:189–194, 1986.
241. Durrheim H, Flynn GL, Higuchi WI, Behl CR. Permeation of hairless mouse skin I:
experimental methods and comparison with human epidermal permeation by alkanols.
J Pharm Sci 69:781–786, 1980.
242. Huq AS, Ho NF, Husari N, Flynn GL, Jetzer WE, Condie L. Permeation of water
contaminative phenols through hairless mouse skin. Arch Environ Contam Toxicol 15:
557–566, 1986.
243. Barteck MJ, La Budde JA, Maibach HI. Skin permeability in vivo: comparison in rat,
rabbit, pig and human. J Invest Dermatol 58:114–123, 1972.
244. Feldman RJ, Maibach HI. Percutaneous penetration of some pesticides and herbicides
in man. Toxicol Appl Pharmacol 28:126–132, 1974.
245. Barteck MJ, La Budde JA. Percutaneous absorption in vitro. In: Maibach HI, ed. An-
imal Models in Dermatology. New York: Churchill–Livingstone, pp 103–112, 1975.
246. Moody RP, Ritter L. Dermal absorption of the insecticide lindane in rats and rhesus
monkeys: effect of anatomical site. J Toxicol Environ Health 28:161–169, 1989.
247. Moody RP, Franklin CA, Ritter L, Maibach HI. Dermal absorption of the phenoxy
herbicides 2,4-D, 2-4-D amine, 2,4-D isooctyl and 2,4,5-T in rabbits, rats, rhesus mon-
key and humans: a cross-species comparison. J Toxicol Environ Health 29:237–245,
1990.
248. Roberts ME, Mueller KR. Comparisons of in vitro nitroglycerin (TNG) flux across
Yucatan pig, hairless mouse, and human skins. Pharm Res 7:673–676, 1990.
249. Bucks DAW, Hinz RS, Sarason R, Maibach HI, Guy RH. In vivo percutaneous ab-
sorption of chemicals—a multiple dose study in rhesus monkeys. Food Chem Toxicol
28:129–132, 1990.
250. Rougier A. The hairless rat: a relevant animal model to predict in vivo percutaneous
absorption in humans. J Invest Dermatol 88:577–581, 1987.
251. Morimoto Y, Hatanaka T, Sugibayashi K, Omiya H. Prediction of skin permeability of

drugs: comparison of human and hairless rat skin. J Pharm Pharmacol 44:634–639,
1992.
252. Sugibayashi K, Nakagaki D, Hatanaka E, Hatanaka T, Inoue N, Kusumi S, Kobayashi
M, Kimura M, Morimoto Y. Differences in enhancing effect of 1-menthol, ethanol and
their combination between hairless rat and human skin. Int J Pharm 113:189–197,
1995.
253. Chowhan ZT, Pritchard R. Effect of surfactants on percutaneous absorption of naproxen
I: comparisons of rabbit, rat and human excised skin. J Pharm Sci 67:1272–1274, 1978.
254. Niazy EM. Differences in penetration-enhancing effect of Azone through excised rab-
bit, rat, hairless mouse, guinea-pig and human skins. Int J Pharm 130:225–230, 1996.
255. Hirvonen J, Rytting JH, Paronen P, Urtti A. Dodecyl N,N-dimethylamino acetate and
Azone enhance drug penetration across human, snake, and rabbit skin. Pharm Res 8:
933–937, 1991.
256. Marzulli FN, Brown DWC, Maibach HI. Techniques for studying skin penetration.
Toxicol Appl Pharmacol 3(suppl):76–83, 1969.
257. Galey WR, Lansdale HK, Nacht S. The in vitro permeability of skin and buccal mucosa
to selected drugs and tritiated water. J Invest Dermatol 67:713–717, 1976.
258. Smith FM, Scott RC, Foster JR. An interspecies comparison of skin structure (hair
follicle area as determined by a novel technique) and skin permeability. In: Galli CL,
Hensby CN, Marinovich M, eds. Skin Pharmacology and Toxicology. New York: Ple-
num Publishing, pp 299–305, 1990.
259. Scott RC. In vitro absorption through damaged skin. In: Bronaugh RL, Maibach HI,
eds. In Vitro Percutaneous Absorption, Principles, Fundamentals and Applications.
Boca Raton: CRC Press, pp 129–135, 1991.
86:201–207, 1986.
261. Shah VP, Flynn GL, Guy RH, Maibach HI, Schaefer H, Skelly JP, Wester RC, Yacobi
A. In vivo percutaneous penetration/absorption. Int J Pharm 74:1–8, 1991.
262. Opdam JJG, Kimmel JPF, Krüse J, Meuling WJA. The rate of percutaneous absorption
in human volunteers; dermal exposure to 1,1,1-tichloroethane. In: Brain KR, James VJ,
lishing, pp 199–202, 1996.
263. Feldman RJ, Maibach HI. Absorption of some organic compounds through the skin in
man. J Invest Dermatol 54:399–404, 1970.
264. Wester RC, Maibach HI, Melendres J Sedik L, Knaak J, Wang R. In vivo and in vitro
percutaneous absorption and skin evaporation of isofenphos in man. Fundam Appl
Toxicol 19:521–526, 1992.
265. Vollmer U, Müller BW, Wilffert B, Peters T. An improved model for studies on trans-
dermal drug absorption in vivo in rats. J Pharm Pharmacol 45:242–245, 1993.
266. Franz JM, Gaillard A, Maibach HI, Schweitzer A. Percutaneous absorption of griseo-
fulvin and proquazone in the rat and in isolated human skin. Arch Dermatol Res 271:
275–282, 1981.
267. Bogen KT, Colston BW, Machicao LK. Dermal absorption of dilute aqueous chloro-
form, trichlorethylene, and tetrachloroethylene in hairless guinea pigs. Fundam Appl
Toxicol 18:30–39, 1992.
268. Yu D, Sanders LM, Davidson GWR, Marvin MJ, Ling T. Percutaneous absorption of
nicardipine and ketorolac in rhesus monkeys. Pharm Res 5:457–462, 1988.
269. Leonardi GR, Maia Campos PMBG. Influence of glycolic acid as a component of
different formulations on skin penetration by vitamin A palmitate. J Cosmet Sci 49:
23–32, 1998.
262 Brain et al.
270. Chatterjee DJ, Li WY, Koda RT. Effect of vehicles and penetration enhancers on the
in vitro and in vivo percutaneous absorption of methotrexate and edatrexate through
hairless mouse skin. Pharm Res 14:1058–1065, 1997.
271. Megwa SA, Benson HAE, Roberts MS. Percutaneous absorption of salicylates from
some commercially available topical products containing methyl salicylate or salicylate
salts in rats. J Pharm Pharmacol 47:891–896, 1995.
272. Wrzesinski CL, Feeney WP, Feely WF, Crouch LS. Dermal penetration of 4⬙-(epi-
methylamino)-4⬙-deoxyavermectin-b1a benzoate in the rhesus monkey. Food Chem Tox-
icol 35:1085–1089, 1997.
273. Davis DAP, Kraus AL, Thompson GA, Olerich M, Odio MR. Percutaneous absorption
of salicylic acid after repeated (14-day) in vivo administration to normal, acnegenic or
aged human skin. J Pharm Sci 86:896–899, 1997.
274. Bonsall JL, Goose J. The safety evaluation of bendiocarb, a residual insecticide for
vector control. In Van Heemstra–Lequin EAH, van Sittert NJ, eds. Biological Moni-
toring of Workers Manufacturing, Formulating and Applying Pesticides. Amsterdam:
Elsevier, p 45, 1986.
275. McDougal JN, Jepson GW, Clewell HJ, Gargas ML, Andersen ME. Dermal absorption
of organic chemical vapors in rats and humans. Fundam Appl Toxicol 14:299–308,
1990.
276. Newton M, Norris LA. Potential exposure of humans to 2,4,5-T and TCDD in the
Oregon coast region. Fundam Appl Toxicol 1:339–445, 1981.
277. Bucks DAW, McMaster JR, Maibach HI, Guy RH. Bioavailability of topically admin-
istered steroids: a ‘‘mass balance’’ technique. J Invest Dermatol 91:29–33, 1988.
278. Bucks DAW, Maibach HI, Guy RH. Mass balance and dose accountability in percu-
taneous absorption studies: development of a nonocclusive application system. Pharm
Res 5:313–315, 1988.
279. Bucks DAW, Guy RH, Maibach HI. Percutaneous penetration and mass balance ac-
countability—technique and implications for dermatology. J Toxicol Cutan Ocul Tox-
icol 8:439–451, 1989.
280. Noonan PK, Gonzalez MA, Ruggirello D, Tomlinson J, Babcock–Atkinson E, Ray M,
Golub A, Cohen A. Relative bioavailability of a new transdermal nitroglycerin delivery
system. J Pharm Sci 75:688–691, 1986.
281. Hadgraft J, Hill S, Humpel M, Johnston LR, Lever LR, Marks R, Murphy TM, Rapier
C. Investigations on the percutaneous absorption of the antidepressant rolipram in vitro
and in vivo. Pharm Res 7:1307–1312, 1990.
282. Le Roux Y, Borg ML, Sibille M, Thebault J, Renoux A, Douin MJ, Djebbar F, Dain
MP. Bioavailability study of Menorest, a new estrogen transdermal delivery system,
compared with a transdermal reservoir system. Clin Pharmacokinet 10:172–178, 1995.
283. Gorsline J, Gupta SK, Dye D, Rolf CM. Steady-state pharmacokinetics and dose re-
lationship of nicotine delivered from Nicoderm (nicotine transdermal system). J Clin
Pharmacol 33:161–168, 1993.
284. Brocks DR, Meikle AW, Boike SC, Mazer NA, Zariffa N, Audet PR, Jorasky DK.
Pharmacokinetics of testosterone in hypogonadal men after transdermal delivery: influ-
ence of dose. J Clin Pharmacol 36:732–739, 1995.
285. Meikle AW, Arver S, Dobs AS, Sanders SW, Rajaram L, Mazer NA. Pharmacokinetics
and metabolism of a permeation-enhanced testosterone transdermal system in hypo-
gonadal men: influence of application site—a clinical research center study. J Clin
Endocrinol Metab 81:1832–1840, 1996.
286. Berner G, Engels B, Vogtle–Junkert U. Percutaneous ibuprofen therapy with Trauma-
Dolgit gel: bioequivalence studies. Drug Exp Clin Res 15:556–564, 1990.
287. Singh P, Roberts MS. Dermal and underlying tissue pharmacokinetics of salicylic acid
after topical application. J Pharmacokinet Biopharm 21:337–373, 1993.
288. Singh P, Roberts MS. Skin permeability and local tissue concentrations of nonsteroidal
anti-inflammatory drugs after topical application. J Pharmacol Exp Ther 268:144–151,
1994.
289. Singh P, Roberts MS. Effects of vasoconstriction on dermal pharmacokinetics and local
tissue distribution of compounds. J Pharm Sci 83:783–791, 1994.
290. Benson HAE, Schild PN, Cross SE, Roberts MS. Investigation of the effect of vaso-
active agents on local skin and underlying tissue blood flow. In: Brain KR, James VJ,
Walters KA, eds. Prediction in Percutaneous Penetration. Vol. 4b. Cardiff: STS Pub-
lishing, pp 258–261, 1996.
291. Cross SE, Wu Z, Roberts MS. Effect of perfusion flow rate on the tissue uptake of
solutes after dermal application using the rat isolated perfused limb preparation. J
Pharm Pharmacol 46:844–850, 1994.
292. Wu Z, Rivory LP, Roberts MS. Physiological pharmacokinetics of solutes in perfused
rat hindlimb: characterisation of the physiology with changing perfusate flow, protein
content and temperature using statistical moments analysis. J Pharm Biopharm 21:653–
688, 1993.
293. Roberts MS, Cross SE, Wu Z. Deep tissue penetration after dermal application observed
in the isolated perfused limb. In: Brain KR, James VJ, Walters KA, eds. Prediction in
Percutaneous Penetration. Vol. 4b. Cardiff: STS Publishing, pp 251–254, 1996.
294. Scott RC, Rhodes C. The permeability of grafted human transplant skin in athymic
mice. J Pharm Pharmacol 40:128–129, 1987.
295. Klain GJ. Dermal penetration and systemic distribution of 14C-labeled vitamin E in
human skin grafted athymic nude nice. Int J Vitam Nutr Res 59:333–337, 1989.
296. Boyce ST, Supp AP, Harriger MD, Pickens WL, Wickett RR, and Hoath SB. Surface
electrical capacitance as a noninvasive index of epidermal barrier in cultured skin
substitutes in athymic mice. J Invest Dermatol 107:82–87, 1996.
297. Boyce ST. Cultured skin substitutes grafted to athymic mice. In: Brain KR, James VJ,
lishing, pp 13, 1997.
298. Lane AT, Scott GA, Day KH. Development of human fetal skin transplanted to the
nude mouse. J Invest Dermatol 93:787–791, 1989.
299. Riviere JE. Grafted skin and skin flaps. In: Shah VP, Maibach HI, eds. Topical Drug
Bioavailability, Bioequivalence, and Penetration. New York: Plenum Press, pp 209–
221, 1993.
300. Manning DD, Reed ND, Shaffer CF. Maintenance of skin xenographs of widely di-
vergent phylogenetic origin on congenitally athymic (nude) mouse. J Exp Med 38:
488–494, 1973.
301. Biren CA, Barr RJ, McCullough JL, Black KS, Hewitt CW. Prolonged viability of
human skin xenographs in rats by cyclosporine. J Invest Dermatol 86:611–614, 1986.
302. Higounenc I, Spies F, Boddé H, Schaefer H, Demarchez M, Shroot B, Ponec M. Lipid
composition and barrier function of human skin after grafting onto athymic nude mice.
Skin Pharmacol 7:167–175, 1994.
303. Petersen RV, Kislalioglu MS, Liang WQ, Fang SM, Emam M, Dickman S. The athymic
nude mouse grafted with human skin as a model for evaluating the safety and effec-
tiveness of radiolabeled cosmetic ingredients. J Soc Cosmet Chem 37:249–265, 1986.
304. Wojciechowski Z, Pershing LK, Huether S, Leonard L, Burton SA, Higuchi WI, Krue-
ger GG. An experimental skin sandwich flap on an independent vascular supply for
the study of percutaneous absorption. J Invest Dermatol 88:439–446, 1987.
305. Pershing LK, Lambert LD, Knutson K. Mechanism of ethanol-enhanced estradiol per-
meation across human skin in vivo. Pharm Res 7:170–175, 1990.
306. Silcox GD, Parry GE, Bunge AL, Pershing LK, Pershing DW. Percutaneous absorption
of benzoic acid across human skin II: prediction of an in vivo, skin flap system using
in vitro parameters. Pharm Res 7:352–358, 1990.
264 Brain et al.
307. Yano T, Nakagawa A, Tsuji M, Noda K. Skin permeability of various non-steroidal

anti-inflammatory drugs in man. Life Sci 39:1043–1050, 1986.
308. Guy RH, Bucks DAW, McMaster JR, Villaflor DA, Roskos KV, Hinz RS, Maibach HI.
Kinetics of drug absorption across human skin in vivo. Developments in methodology.
In: Maibach HI, Schaefer H, eds. Pharmacology and the Skin. Vol. 1. Skin Pharma-
cokinetics. Basel: Karger, pp 70–76, 1987.
309. Mak VHW, Potts RO, Guy RH. Percutaneous penetration enhancement in vivo mea-
sured by attenuated total reflectance infrared spectroscopy. Pharm Res 7:835–841,
1990.
310. Higo N, Naik A, Bommannan DB, Potts RO, Guy RH. Validity of reflectance infra-
red spectroscopy. Pharm Res 10:1500–1506, 1993.
311. Sennhenn B, Giese K, Plamann K, Harendt N, Kölmel K. In vivo evaluation of the
penetration of topically applied drugs into human skin by spectroscopic methods. Skin
Pharmacol 6:152–160, 1993.
312. Naik A, Guy RH. Infrared spectroscopic and differential scanning calorimetric inves-
tigations of the stratum corneum barrier function. In: Potts RO, Guy RH, eds. Mech-
anisms of Transdermal Drug Delivery. New York: Marcel Dekker, pp 87–162, 1997.
313. Potts RO, Guzek DB, Harris RR, McKie JE. A noninvasive, in vivo technique to
quantitatively measure water concentration of the stratum corneum using attenuated
total-reflectance infrared spectroscopy. Arch Dermatol Res 277:489–495, 1985.
314. Martin KA. Direct measurement of moisture in skin by NIR spectroscopy. J Soc Cos-
met Chem 44:249–261, 1993.
315. Gilard V, Martino R, Malet–Martino M, Riviere M, Gournay A, Navarro R. Measure-
ment of total water and bound water contents in human stratum corneum by in vitro
proton nuclear magnetic resonance spectroscopy. Int J Cosmet Sci 20:117–125, 1998.
316. Pechtold LARM, Abraham W, Potts RO. Characterization of stratum corneum barrier
properties using fluorescence spectroscopy. In: Potts RO, Guy RH, eds. Mechanisms
of Transdermal Drug Delivery. New York: Marcel Dekker, pp 199–213, 1997.
317. Burnette RR, DeNuzzio JD. Impedance spectroscopy: applications to human skin. In:
Potts RO, Guy RH, eds. Mechanisms of Transdermal Drug Delivery. New York: Marcel
Dekker, pp 215–230, 1997.
318. Watkinson AC, Hadgraft J, Street PR, Richards RW. Neutrons, surfaces, and skin. In:
Potts RO, Guy RH, eds. Mechanisms of Transdermal Drug Delivery. New York: Marcel
Dekker, pp 231–265, 1997.
319. Bindra RMS, Eccleston GM, Imhof RE, Birch DJS. Opto-thermal in vivo monitoring
of human skin condition. In: Scott RC, Guy RH, Hadgraft J, Boddé HE, eds. Prediction
of Percutaneous Penetration. Vol. 2. Cardiff: STS Publishing, pp 628–635, 1991.
320. Petersen EM. The hydrating effect of a cream and white petrolatum measured by
optothermal infrared spectrometry in vivo. Acta Derm Venereol 71:373–376, 1991.
321. Imhof RE, Whitters CJ, Birch DJS. Opto-thermal in vivo monitoring of sunscreens on
skin. Phys Med Biol 35:95–102, 1990.
322. Cowen JA, Imhof RE. In vivo surface disappearance measurement of applied chemicals
using opto-thermal transient emission radiometry (OTTER). In: Brain KR, James VJ,
Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5b. Cardiff: STS Pub-
lishing, pp 94–97, 1998.
323. Imhof RE, Whitters CJ, Birch DJS. Opto-thermal in vivo monitoring of structural
breakdown of an emulsion sunscreen on skin. Clin Mater 5:272–278, 1990.
324. Dupuis D, Rougier A, Roguet R, Lotte C, Kalopissis G. In vivo relationship between
horny layer reservoir effect and percutaneous absorption in human and rat. J Invest
Dermatol 82:353–358, 1984.
325. Tojo K, Lee AC. A method for predicting steady-state rate of skin penetration in vivo.
326. Rougier A. An original predictive method in vivo percutaneous absorption studies. J

Soc Cosmet Chem 38:397–417, 1987.
327. Rougier A, Dupuis D, Lotte C. Stripping method for measuring percutaneous absorp-
tion in vivo. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption: Mecha-
nisms, Methodology, Drug Delivery, 2nd ed. New York: Marcel Dekker, pp 415–434,
1989.
328. Rougier A. Predictive measurement of in vivo percutaneous penetration. In: Scott RC,
Guy RH, Hadgraft J, eds. Prediction of Percutaneous Penetration. Vol. 1. London: IBC
Technical Services, pp 252–262, 1990.
329. Meibohm B, Derendorf H. Basic concepts of pharmacokinetics/pharmacodynamic (PK/
PD) modelling. Int J Clin Pharm Ther 35:401–413, 1997.
330. Anderson C, Andersson T, Molander M. Ethanol absorption across human skin mea-
sured by in vivo microdialysis technique. Acta Derm Venereol 71:389–393, 1991.
331. Anderson C, Andersson T, Wårdell K. Changes in skin circulation after insertion of a
microdialysis probe visualized by laser Doppler perfusion imaging. J Invest Dermatol
102:807–811, 1994.
332. Anderson C, Andersson T, Boman A. Cutaneous microdialysis for human in vivo der-
mal absorption studies. In: Roberts MS, Walters KA, eds. Dermal Absorption and
Toxicology Assessment. New York: Marcel Dekker, pp 231–244, 1998.
333. Ault JM, Lunte CE, Meltzer NM, and Riley CM. Microdialysis as a dermal sampling
technique for percutaneous absorption. In: Brain KR, James VJ, Walters KA, eds.
Prediction of Percutaneous Penetration. Vol. 3b. Cardiff: STS Publishing, pp 44–48,
1993.
334. Ault JM, Riley CM, Meltzer NM, Lunte CE. Dermal microdialysis sampling in vivo.
Pharm Res 11:1631–1639, 1994.
335. Matsuyama K, Nakashima M, Ichikawa M, Yona T, Saoh S, Groto S. In vivo micro-
dialysis for the transdermal absorption of valproate in rats. Biol Pharm Bull 17:1395–
1398, 1994.
336. Matsuyama K, Nakashima M, Nakaboh Y, Ichikawa M, Yona T, Saoh S. Application
of in vivo microdialysis to transdermal absorption of methotrexate in rats. Pharm Res
11:686–688, 1994.
337. Muller M, Schmid R, Wagner O, Vonosten B, Shayganfar H, Eichler HG. In vivo
characterization of transdermal drug transport by microdialysis. J Controlled Release
37:49–57, 1995.
338. Hegeman I, Forstinger C, Partsch C, Lagler I, Krotz S, Wolf K. Microdialysis in cu-
taneous pharmacology: kinetic analysis of transdermally delivered nicotine. J Invest
Dermatol 104:839–843, 1995.
339. Pepler AF, Morrison JC. The influence of vehicle composition on the vasoconstrictor
activity of betamethasone 17-benzoate. Br J Dermatol 85:171–176, 1971.
340. Barry BW, Woodford R. Proprietary hydrocortisone creams. Vasoconstrictor activities
and bioavailabilities of six preparations. Br J Dermatol 95:423–425, 1976.
341. Merver E. Comparison of the blanching activities of Dermovate, Betnovate, and
Eumovate creams and ointments. Int J Pharm 41:63–66, 1988.
342. Queille–Roussel C, Poncet M, Schaefer H. Quantification of skin colour changes in-
duced by topical corticosteroid preparations using the Minolta chromameter. Br J Der-
matol 124:264–270, 1991.
343. Chan SV, Li Wan Po A. Quantitative skin blanching assay of corticosteroid creams
using tristimulus colour analysis. J Pharm Pharmacol 44:371–378, 1992.
344. Corson SL. A decade of experience with transdermal estrogen replacement therapy:
overview of key pharmacologic and clinical findings. Int J Fertil 38:79–91, 1993.
345. Lin S, Ho H, Chien YW. Development of a new nicotine transdermal delivery system:
in vitro kinetics studies and clinical pharmacokinetic evaluations in two ethnic groups.
J Controlled Release 26:175–193, 1993.
266 Brain et al.
346. Shah HS, Tojo K, Huang YC, Chien YW. Development and clinical evaluation of a
verapamil transdermal delivery system. J Controlled Release 22:125–131, 1992.
347. Bialik W, Walters KA, Brain KR, Hadgraft J. Some factors affecting the in vitro pen-
etration of ibuprofen through human skin. Int J Pharm 92:219–223, 1993.
348. Monteiro–Riviere NA, Inman AO, Riviere JE, McNeill SC, Francoeur ML. Topical
penetration of piroxicam is dependent on the distribution of the local cutaneous vas-
culature. Pharm Res 10:1326–1331, 1993.
349. Treffel P, Gabard B. Feasibility of measuring the bioavailability of topical ibuprofen
in commercial formulations using drug content in epidermis and a methyl nicotinate
skin inflammation assay. Skin Pharmacol 6:268–275, 1993.
350. Seth PL. Percutaneous absorption of ibuprofen from different formulations. Compar-
ative study with gel, hydrophilic ointment and emulsion cream. Arzneimittelforschung
43:919–921, 1993.
351. Marks R, Dykes P. Plasma and cutaneous drug levels after topical application of
piroxicam gel: a study in healthy volunteers. Skin Pharmacol 7:340–344, 1994.
352. Francoeur ML, Monteiro–Riviere NA, Riviere JE. Piroxicam: evidence for local de-
livery following topical application. Eur J Pharm Biopharm 41:175–183, 1995.
353. Roy SD, Manoukian E. Transdermal delivery of ketorolac tromethamine: permeation
enhancement, device design, and pharmacokinetics in healthy humans. J Pharm Sci 84:
1190–1196, 1995.
354. Taburet AM, Singlas E, Glass RC, Thomas F, Leutenegger E. Pharmacokinetic com-
parison of oral and local action transcutaneous flurbiprofen in healthy volunteers. J
Clin Pharm Ther 20:101–107, 1995.
355. Shah AK, Wei G, Lanman RC, Bhargava VO, Weir SJ. Percutaneous absorption of
ketoprofen from different anatomical sites in man. Pharm Res 13:168–172, 1996.
356. Bouchier–Hayes TA, Rotman H, Darekar BS. Comparison of the efficacy and tolera-
bility of diclofenac gel (Voltarol Emulgel) and felbinac gel (Traxam) in the treatment
of soft tissue injuries. Br J Clin Pract 44:319–320, 1990.
357. Memeo A, Garofoli F, Peretti G. Evaluation of the efficacy and tolerability of a new
topical formulation of flurbiprofen in acute soft tissue injuries. Drug Invest 4:441–449,
1992.
358. Hosie G, Bird H. The topical NSAID felbinac versus oral NSAIDs: a critical review.
Eur J Rheumatol Inflamm 14:21–28, 1994.
359. Campbell J, Dunn T. Evaluation of topical ibuprofen cream in the treatment of acute
ankle sprains. J Accident Emerg Med 11:178–182, 1994.
360. Moore RA, Tramèr MR, Carroll D, Wiffen PJ, McQuay HJ. Quantitative systematic
review of topically applied non-steroidal anti-inflammatory drugs. Br Med J 316:333,
1998.
361. Arendtnielsen L, Drewes AM, Svendsen L, Brennum J. Quantitative assessment of
joint pain following treatment of rheumatoid arthritis with ibuprofen cream. Scand J
Rheumatol 23:334–337, 1994.
362. Dickson DJ. A double-blind evaluation of topical piroxicam gel with oral ibuprofen in
osteoarthritis of the knee. Curr Ther Res 49:199–207, 1991.
363. McNiell SC, Potts RO, Francoeur ML. Local enhanced topical delivery (LETD) of
drugs: does it truly exist? Pharm Res 9:1422–1427, 1992.
364. Benson HAE, Megwa SA, Roberts MS. Tissue penetration of salicylate following top-
ical application as methyl salicylate and amine salts. In: Brain KR, James VJ, Walters
255–257, 1996.
365. Bronaugh RL, Franz TJ. Vehicle effects on percutaneous absorption: in vivo and in
vitro comparisons with human skin. Br J Dermatol 115:1–11, 1986.
366. Scott RC. Percutaneous absorption, in vivo:in vitro comparisons. Pharmacol Skin 1:
103–110, 1987.
367. Al-Khamis K, Davis SS, Hadgraft J. In vitro–in vivo correlations for the percutaneous
absorption of salicylates. Int J Pharm 40:111–118, 1987.
368. Grissom RE, Brownie C, Guthrie FE. In vivo and in vitro dermal penetration of li-
pophilic and hydrophilic pesticides in mice. Bull Environ Contam Toxicol 38:917–924,
1987.
369. Surber C, Wilhelm K–P, Maibach HI. In vitro and in vivo percutaneous absorption of
structurally related phenol and steroid analogs. Eur J Pharm Biopharm 39:244–248,
1993.
370. Scott RC, Ramsey JD. Comparison of the in vivo and in vitro percutaneous absorption
of a lipophilic molecule (Cypermethrin, a pyrethroid insecticide). J Invest Dermatol
89:142–146, 1987.
371. Scott RC, Batten PL, Clowes HM, Jones BK, Ramsey JD. Further validation of an in
vitro method to reduce the need for in vivo studies measuring the absorption of chem-
icals through the skin. Fundam Appl Toxicol 19:484–492, 1992.
372. Ceschin–Roques CG. Hänel H, Pruja–Bougaret SM, Lagarde I, Vandermander J,
Michel G. Ciclopiroxolamine cream 1%: in vitro and in vivo penetration into the
stratum corneum. Skin Pharmacol 4:95–99, 1991.
373. Reifenrath WG, Hawkins GS, Kurtz MS. Percutaneous penetration and skin retention
of topically applied compounds: an in vitro–in vitro study. J Pharm Sci 80:526–532,
1991.
374. Lien EJ, Gao H. QSAR analysis of skin permeability of various drugs in man as
compared to in vivo and in vitro studies in rodents. Pharm Res 12:583–587, 1995.
375. Li B, Birt DF. In vivo and in vitro percutaneous absorption of cancer preventative
flavonoid apigenin in different vehicles in mouse skin. Pharm Res 13:1710–1715, 1996.
376. Macpherson SE, Barton CN, Bronaugh RL. Use of in vitro skin penetration data and
a physiologically based model to predict in vivo blood levels of benzoic acid. Toxicol
Appl Pharmacol 140:436–443, 1996.
377. Kasting GB, Francis WR, Bowman LA, Kinnett GO. Percutaneous absorption of
vanilloids: in vivo and in vitro studies. J Pharm Sci 86:142–146, 1997.
378. Bronaugh RL. Methods for in vitro metabolism studies. Toxicol Methods 5:275–281,
1995.
379. Yang JJ, Roy TA, Neil W, Krueger AJ, Mackerer CR. Percutaneous and oral absorption
of chlorinated paraffins in the rat. Toxicol Ind Health 3:405–412, 1987.
380. Hotchkiss SAM, Chidgey MAJ, Rose S, Caldwell J. Percutaneous absorption of benzyl
acetate through rat skin in vitro. 1. Validation of an in vitro model against in vivo data.
Food Chem Toxicol 28:443–447, 1990.
381. Bronaugh RL, Maibach HI. Percutaneous absorption of nitroaromatic compounds: in
vivo and in vitro studies in the human and monkey. J Invest Dermatol 84:180–183,
1985.
382. Cline JF, Nichols LD, Soybel JG, Brown LR. Transdermal diffusion: in vivo versus in
vitro comparison. Proc Int Symp Controlled Release Bioact Mater 14:182–183, 1987.
383. Sato K, Sugibayashi K, Morimoto Y, Omiya H, Enomoto N. Prediction of the in vitro
human skin permeability of nicorandil from animal data. J Pharm Pharmacol 41:379–
383, 1989.
384. Altenburger R, Rohr UD, Kissel T. Explanation for the plasma level fluctuations re-
sulting from the transdermal liquid reservoir system delivering ␤-estradiol. Proc Int
Symp Controlled Release Bioact Mater 24:695–696, 1997.
385. Chien T–Y, Gong S–J. Transdermal delivery of levonorgestrel: pilot clinical studies
on a matrix-type once-a-week transdermal (monophasic) contraceptive delivery system.
Proc Int Symp Controlled Release Bioact Mater 24:715–716, 1997.
268 Brain et al.
386. Rohr UD, Altenburger R, Kissel T. Pharmacokinetics of the transdermal reservoir mem-
brane system delivering ␤-estradiol: In vitro/in vivo correlation. Pharm Res 15:877–
882, 1998.
387. Kligman AM. A biological brief on percutaneous absorption. Drug Dev Ind Pharm 9:
521–560, 1983.
388. Gettings SD, Howes D, Walters KA. Experimental design considerations and use of in
vitro skin penetration data in cosmetic risk assessment. In: Roberts MS, Walters KA,
eds. Dermal Absorption and Toxicity Assessment. New York: Marcel Dekker, pp 459–
487, 1998.
389. Foldvari M, Oguejiofor CJN, Wilson TW, Afridi SK, Kudel TA. Transcutaneous deliv-
ery of prostaglandin E1: in vitro and laser doppler flowmetry study. J Pharm Sci 87:
721–725, 1998.
390. Bucks DAW, Marty JP, Maibach HI. Percutaneous absorption of malathion in the
guinea pig: effect of repeated topical application. Food Chem Toxicol 23:919–922,
1985.
391. Catz P, Friend DR. Transdermal delivery of levonorgestrel. VIII. Effect of enhancers
on rat skin, hairless mouse skin, hairless guinea pig skin, and human skin. Int J Pharm
58:93–102, 1990.
392. Chambin O, Vincent CM, Heuber F, Teillaud E, Pourcelot Y, Marty JP. Validation of
pig ear skin as a model for in vitro percutaneous absorption studies. In: Brain KR,
James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 4b. Cardiff:
STS Publishing, pp 271–274, 1996.
393. Chow C, Chow AYK, Downie RH, Buttar HS. Percutaneous absorption of hexachlo-
rophene in rats, guinea pigs and pigs. Toxicology 9:147–154, 1978.
394. Dick IP, Scott RC. Pig ear skin as an in vitro model for human skin permeability. J
Pharm Pharmacol 44:640–645, 1992.
395. Fujii M, Shiozawa K, Hemi T, Yamanouchi S, Susuki H, Yamashita N, Matsumoto M.
Skin permeation of indomethacin from gels formed by fatty acid esters and phospho-
lipids. Int J Pharm 137:117–124, 1996.
396. Howes D, Black JG. Comparative percutaneous absorption of pyrithiones. Toxicology
5:209–220, 1975.
397. Hunziger N, Feldmann RJ, Maibach HI. Animal models of percutaneous penetration:
comparison between Mexican hairless dogs and man. Dermatologica 156:79–88, 1978.
398. Imoto H, Zhou Z, Stinchcomb AL, Flynn GL. Transdermal prodrug concepts—per-
meation of buprenorphine and its alkyl esters through hairless mouse skin and influence
of vehicles. Biol Pharm Bull 19:263–267, 1996.
399. Itoh T, Xia J, Magavi R, Nishihata T, Rytting JH. Use of shed snake skin as a model
membrane for in vitro percutaneous penetration studies: comparison with human skin.
Pharm Res 7:1042–1047, 1990.
400. Moody RP, Nadeau B, Chu I. In vitro dermal absorption of N,N-diethyl-m-toluamide
(DEET) in rat, guinea pig and human skin. In Vitro Toxicol J Mol Cell Toxicol 8:263–
275, 1995.
401. Rigg PC, Barry BW. Shed snake skin and hairless mouse skin as model membranes
for human skin during permeation studies. J Invest Dermatol 94:235–240, 1990.
402. Sloan KB, Beall HD, Weimar WR, Villanueva R. The effect of receptor phase com-
position on the permeability of hairless mouse skin in diffusion cell experiments. Int
J Pharm 73:97–104, 1991.
403. Wester RC, Maibach HI, Bucks DAW, McMaster J, Mobayen M, Sarason R, Moore
A. Percutaneous absorption and skin decontamination of PCBs—in vitro studies with
human skin and in vivo studies in the rhesus monkey. J Toxicol Environ Health 31:
235–246, 1990.
404. Yu HY, Liao HM. Triamcinolone permeation from different liposome formulations
through rat skin in vitro. Int J Pharm 126:1–7, 1996.
405. Minter HJ, Shaw A, Howes D, Pendlington RU. The use of microautoradiography to
show the distribution of substances topically applied to skin. In: Brain KR, James VJ,
Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5b. Cardiff: STS Pub-
lishing, pp 102–103, 1998.
406. Bronaugh RL, Collier SW. Protocol for in vitro percutaneous absorption studies. In:
Bronaugh RL, Maibach HI, eds. In Vitro Percutaneous Absorption: Principles, Fun-
damentals, and Applications. Boca Raton: CRC Press, pp 237–241, 1991.
407. Le Roux Y, Borg ML, Sibille M, Thebault J, Renoux A, Douin MJ, Djebbar F, Dain
MP. Bioavailability study of Menorest, a new estrogen transdermal delivery system,
compared with a transdermal reservoir system. Clin Pharmacokinet 10:172–178, 1995.
408. Findlay JC, Place VA, Snyder PJ. Transdermal delivery of testosterone. Clin Endocrinol
Metab 64:266–268, 1987.
409. Shah VP, Pershing LK. The stripping technique to assess bioequivalence of topically
applied formulations. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percu-
taneous Penetration. Vol. 3b. Cardiff: STS Publishing, pp 473–476, 1993.
410. Rougier A, Dupuis D, Lotte C, Roguet R, Schaefer H. In vivo correlation between
stratum corneum reservoir function and percutaneous absorption. J Invest Dermatol 81:
275–278, 1983.
411. Rougier A, Lotte C, Maibach HI. In vivo percutaneous penetration of some organic
compounds related to anatomic site in humans: predictive assessment by the stripping
method. J Pharm Sci 76:451–454, 1987.
412. Roberts MS. Recent developments in skin penetration: role of physiology and solute
structure. In: Brain KR, James VJ, Walters KA, eds. Perspectives in Percutaneous
6
Formulation Strategies for Modulating
Skin Permeation
ADRIAN F. DAVIS
SmithKline Beecham, Weybridge, Surrey, England
ROBERT J. GYURIK
MacroChem Corporation, Lexington, Massachusetts
JONATHAN HADGRAFT
University of Greenwich, Chatham Maritime, Kent, England
MARK A. PELLETT
Whitehall International, Havant, England
KENNETH A. WALTERS
I. INTRODUCTION
Because of the barrier properties of the stratum corneum and, also, depending on
the physicochemical properties of the drug, transport from simple vehicles will often
be insufficient to achieve therapeutic drug concentrations at the site of action. The
therapeutic target may be the skin, or the local or distal subcutaneous tissues, de-
pending on the intent for local, regional, or systemic therapy. Simple vehicles are
defined as those in which the drug is at or close to saturation solubility, and neither
the vehicle nor the drug has any interaction with the stratum corneum to reduce its
barrier function. From this definition drug–vehicle interactions in subsaturated or
saturated solutions will not be reviewed. In these simple systems, depending on the
saturated solubility of the drug in each vehicle, large differences in skin penetration
271
272 Davis et al.
can occur between vehicles that are at a fixed drug concentration. Vehicles in which
the drug is at or near saturation will show enhanced drug penetration, compared with
those in which the drug is subsaturated (1–3).
It is often necessary to increase the amount and rate of dermal or transdermal
drug delivery to achieve the required therapeutic drug levels. In these instances skin
penetration enhancement strategies may be evaluated. The modeling of skin pene-
tration and permeation as a diffusional process allows consideration of the mecha-
nisms by which enhancement may be achieved. In its simplest form, the steady-state
flux of drug per unit area across the skin can be expressed by Fick’s first law of
diffusion:
D⌬Csc
J= (1)
L
where D is the diffusion coefficient of the drug in the stratum corneum and ⌬Csc is
the drug concentration difference across the stratum corneum; L is the apparent
thickness of the stratum corneum, accounting for tortuosity. In reality, the concen-
tration in the outermost layers of the stratum corneum is very much greater than that
in the innermost, Co >> Ci and
DCo
J= (2)
L
which is equivalent to the more usual form of this equation
KDCf
J= (3)
L
where Cf is the drug concentration in solution in the vehicle, and K is the partition
coefficient of the drug between the stratum corneum and the formulation.
It is apparent, from Eq. (2), that the physicochemical determinants that can be
manipulated in an attempt to increase skin penetration are D and Co (in the following
discussion the difference between Co(sat) and Co(supersat) will be outlined). Classic chem-
ical enhancers, such as SEPA, laurocapram (Azone), oleic acid, dimethyl sulfoxide
(DMSO), propylene glycol, fatty acid derivatives, and others are defined as vehicle
components that enter the stratum corneum and increase drug diffusivity within the
barrier membrane (D), and increase vehicle drug solubility–partitioning (Co(sat)), or
both.
Most investigations of penetration enhancement concern the evaluation of ma-
terials, such as oleic acid, that penetrate into the stratum corneum, interact with
intercellular barrier lipids, and alter D. Structure–activity studies with this class of
enhancer indicate that polar head groups linked to long alkyl chains are required.
Although chemical enhancers have been used successfully to increase skin penetra-
tion, in many instances these materials also possess some potential irritancy (4). A
typical example is sodium dodecyl sulfate (SDS).
One mechanism whereby Co(sat) may be increased is enhancer-induced modifi-
cation of the polarity of the skin, such that drug-saturated solubility in the skin is
increased. Here, it is beneficial to use solubility parameters as a measure of polarity,
because this value is known for many compounds and has been approximated also
for the stratum corneum (5,6). Coenhancer systems, for example fatty acids with
Formulation Strategies for Skin Permeation 273
propylene glycol (7), are particularly effective, and this is probably due to a complex
synergy in which each individual enhancer promotes the delivery of its partner as
well as that of the drug.
Equation (2) is a simplification of the exact physicochemical parameters that
control flux. To be more exact, the driving force for diffusion is the chemical poten-
tial gradient. Higuchi was the first to apply the more rigorous solution [Eq. (4)] to
the process of percutaneous penetration (8).
Da
J= (4)
L(␥sc)
where a is the thermodynamic activity of the drug in the vehicle (and, assuming
equilibrium, also in the outer layer of the stratum corneum), and ␥sc is the drug
activity coefficient in the stratum corneum. As a/␥sc is equivalent to Co, Eq. (4)
corresponds to Eq. (2). From this, Higuchi predicted the potential of supersaturation
(a compound having a thermodynamic activity greater than unity, for which 1 rep-
resents the compounds activity in saturated solution) to increase percutaneous pen-
etration. Supersaturation is an enhancement mechanism that is specific to the indi-
vidual drug. Thus, there is no overall reduction in skin barrier properties and less
potential for irritancy of excipients. The effect is best described as an increase in
‘‘push’’ of the drug into the stratum corneum. Supersaturation causes an increase in
drug solubility in the stratum corneum, Co(supersat) beyond and independent of saturated
solubility. From this, enhancer and coenhancer systems affecting Co(sat) or D, and
supersaturation, affecting Co(supersat), should work independently (possibly in synergy)
and be capable of multiplicative increases in penetration needed to achieve thera-
peutic levels of some drugs.
This chapter will concentrate on recent developments in the areas of chemical
penetration enhancement, supersaturated systems, and vesicles. For the chemical
penetration enhancers, only those molecules specifically designed to act as skin pen-
etration enhancers will be discussed. Other chemical enhancers and physical mech-
anisms for enhancement, including formation of prodrugs, iontophoresis, electropor-
ation, and ultrasound, have been fully discussed elsewhere (9–11), and they will not
be included here.
II. CHEMICAL ENHANCERS

A. Azone (1-Dodecylazacycloheptan-2-one)
Azone (1-dodecylazacycloheptan-2-one; laurocapram) and its analogues have prob-
ably been one of the most investigated groups of penetration enhancers. Despite the
first references to Azone as a penetration enhancer appearing in the literature as long
ago as the early 1980s, it is perhaps surprising that there has been no significant
commercialization. This is a reflection of the manner in which regulatory authorities
apply the same stringent licensing constraints to excipients as they do to new phar-
maceutical actives. The extensive research into this group of penetration enhancers
has, however, provided a clearer understanding of the mechanisms of skin penetration
modulation, and structural activity relations. Many different experiments have been
conducted that can be broken down into a number of categories:
274 Davis et al.
1. Interactions with model membranes

2. In vitro experiments
3. In vivo experiments
4. Synergy.
Representative publications will be highlighted to provide a general overview
of what is currently known about Azone and its analogues. The information can then
be applied to an understanding of other penetration enhancers.
The physicochemical properties of Azone are given in Table 1. Figure 1 pro-
vides structural details of some representative analogues that have been synthesized.
1. Interactions with Model Membranes
One of the difficulties in skin penetration studies is attempting to analyze the data
and obtain a mechanistic description of the diffusion process. The inherent variability
in tissue samples and that the skin is a heterogeneous barrier are the principal reasons
for this. To simplify the understanding of the mechanisms of action of Azone, various
experiments have been conducted in which simple-structured lipids have been used
to simulate the complex bilayer lipid structures that are found in the intercellular
regions of the stratum corneum. The structured lipids have been simple monolayers
or vesicles composed of materials such as dipalmitoyl phosphatidylcholine (DPPC)
or mixed lipids representative of those found in the skin. The simplest experiments
that have been conducted involve the incorporation of the Azone analogues into
multilamellar vesicles composed of DPPC and then measuring the ability of the
modulator to alter the phase-transition temperature of the vesicle. This can be
achieved by a light-scattering method (12) or by using differential-scanning calorim-
etry (13). Azone lowers the phase transition temperature and creates regions within
the liposome that are more fluid. It is postulated that the mechanism of action of
Azone in skin penetration enhancement is that it inserts itself into the structured
lipids of the stratum corneum where it creates a more fluid environment. A diffusing
drug, therefore, will encounter a reduced microviscosity, and penetration will be
accelerated. The method of insertion has been probed using monolayers in an au-
tomated Langmuir trough (14). Azone appears to expand the monolayer and force
apart the DPPC molecules more than would be anticipated from ideal mixing. The
experiments also allow an estimation of the area per molecule occupied by Azone.
Molecular graphics analysis of the end-on area per molecule suggest that Azone
Table 1 Some Physicochemical Properties of Azone
Empirical formula C18H35NO

Molecular weight 281.49
Log octanol/water partition coefficienta 7.8
Water solubilitya 7 ⫻ 10⫺7 g/L
Melting point ⫺7⬚C
Viscosity 45.2 cp
Refractive index 1.4701
Specific gravity 0.912
Surface tension 32.65 dyne/cm
a
Estimated values (ACD Inc. software Toronto, Canada).
Figure 1 Structure of Azone and some analogues.
exists in the monolayer with its seven-membered ring structure in an orientation that
is approximately perpendicular to the alkyl chains that are parallel to the DPPC
chains (15). A representation of this is shown in Figure 2.
Further research in which Azone was incorporated into more complex lipid
mixtures (more representative of those found in the stratum corneum) indicates that
Azone reduces the condensation state of the lipids, which corresponds to increased
fluidity (16). X-ray-scattering and differential-scanning calorimetry (DSC) studies on
similar complex lipid mixtures show that Azone reduces the phase-transition tem-
perature of the lipids and that it was intercalated into the lipid structures, which
induced lateral swelling. These findings could be explained on the basis of the shape
of Azone, which has a large interfacial area and head group compared with its alkyl
276 Davis et al.
Figure 2 The expected orientation of Azone (left-hand two molecules) and N-0915 (right-
hand two molecules).
chain volume (17). X-ray diffraction has also been used to examine the interaction
of Azone with partly neutralized fatty acids in water (18). Azone transformed the
original reversed hexagonal phase into a reversed micellar phase. This is consistent
with similar observations between Azone and soybean lecithin–water or monoolein–
water systems (19).
Figure 2 also shows the Azone analogue N-0915. Monolayer experiments on
this compound show that its presence in a DPPC monolayer causes condensation of
the lipids (14). Additionally incorporation into DPPC liposomes produces a slight
increase in the phase-transition temperature (20). This is indicative of a stabilization
of the lipid structure; hence, an increase in microviscosity. The results were explained
in terms of the ability of N-0915 to hydrogen-bond with adjacent lipid molecules,
causing the condensation. Included in the publication (20) was in vitro data on the
penetration of metronidazole and the mosquito repellent diethyl-m-toluamide through
human skin; N-0915 retarded their penetration.
Transfer studies have also been conducted on model solid-supported liquid
membranes, such isopropyl myristate (IPM) incorporated into a cellulose acetate
membrane. The presence of Azone in the IPM facilitated the flux of sodium salicylate
across the membrane. This was attributed to the ability of the nitrogen atom in Azone
to protonate at the pH used and to ion-pair with the salicylate anion (21). However,
alternative explanations have also been produced (22) in which a model of the ex-
traction coefficients using pKa, of weak acids was developed to account for the
observed variation in partition.
2. In Vitro Experiments
Many in vitro experiments have been conducted on the effects of Azone and its
analogues. These can be divided into animal and human studies.
a. Animal Studies. One of the major difficulties in conducting animal ex-
periments is to find a species that can be used to predict effects in human tissue.
The indication from the model membrane studies is that it is the ability of Azone to
intercalate into the lipids of the stratum corneum that is important. The precise nature
of the skin lipids will be important, and it has been documented that these are very
species-dependent. The results from animal experiments, therefore, have to be taken
with some caution, but could be indicative of what may be expected in human skin.
Numerous species have been examined and some comparisons determined. For ex-
ample the effect of a number of enhancers, including Azone, on the penetration of
hydrocortisone through the skin of hairless mouse (HM), hairless rat (HR), and hu-
man cadaver (HC) has been measured (23). The enhancement ratios found for Azone
were HM: 18.0; HR: 13.1; HC: 5.5, showing that, for hydrocortisone, at least, the
effects are far more pronounced in rodent skin. Shed snakeskin has also been a
popular choice for enhancer experiments. Experiments on the transport of acetamin-
ophen and ibuprofen through shed snake skin showed that the addition of Azone
increased the amount of acetaminophen, but not the amount of ibuprofen (24). The
acetaminophen studies showed that the mechanism of enhancement was an increase
in partition coefficient K, with no significant increase in the diffusion coefficient D.
For the ibuprofen formulation, Azone provided a slight increase in D, which was
offset by a decrease in K. These results highlight the importance of partitioning in
penetration enhancement. Because Azone can sometimes have a detrimental effect
on partitioning of the drug into the skin is often ignored.
A species comparison between snake skin, rabbit skin, and human skin has
also been conducted (25). Model drugs, indomethacin, 5-fluorouracil, and propran-
olol were chosen. The relative permeation improvements in human skin, snakeskin,
and rabbit skin were 10–20-fold, 5–50-fold, and 20–120-fold, respectively. Azone
significantly increased the permeation of both the lipophilic and hydrophilic com-
pounds. Rabbit skin was considered a poor model for human skin in vitro, whereas
snake skin was closer to human skin in terms of transdermal permeability. The
relative enhancement of several compounds through snake skin was dependent on
the lipophilicity and the molecular size of the permeant, with a greater permeability
increase for more hydrophilic and larger-molecular size permeants (26).
Hairless mouse skin is often used in skin permeation studies, and a species
comparison has been conducted using dihydroergotamine as the permeant. In this
study, Azone increased the flux in the order rabbit > human > rat > guinea pig >
hairless mouse (27). Guinea pig skin has also been used in a systematic study of
Azone and some analogues (1-geranyl and 1-farnesylazacycloheptan-2-one). Seven
penetrants with varying lipophilicities were examined. Large enhancement ratios
were found for drugs having octanol–water partition coefficients close to unity (28).
Hairless mouse skin has been used in a range of studies examining the enhancing
activity of Azone and its analogues (23,29–39).
b. Human Studies. Sophisticated biophysical techniques have been used to
probe the mechanism of action of Azone in human skin. For example, deuterium
278 Davis et al.
nuclear magnetic resonance (NMR) has shown that a mixture of Azone and propylene
glycol causes structural disorder of the intercellular lamellar lipid structures (40).
DSC has been used to examine the effect of Azone and its analogues on lowering
the phase-transition temperatures of human skin lipids (41). X-ray-scattering exper-
iments have shown that treatment of human stratum corneum with Azone, and its
analogues having alkyl chains of more than six carbons in length, results in disor-
dering of the lamellae. These results are similar to those obtained by thermal analysis
(42,43). Electron-spin resonance (ESR) has shown that Azone increased both the
fluidity and polarity of the environment close to the 5-doxyl stearic acid spin label
(44,45). These were similar to the results found for a similar ESR experiment con-
ducted with the enhancer oleic acid, suggesting a related mechanism of action (46).
Fourier transform infrared (FTIR) spectroscopy has indicated that perdeuterated
Azone (47) distributes homogeneously within the stratum corneum lipids, in which
it induces fluidity (48). This contrasts with the mechanism of action of oleic acid
that appears to phase-separate and form liquid pools (49). Clearly, there are differ-
ences in the mode of action: the sodium salt of heparin has its flux enhanced by
Azone, whereas oleic acid is ineffective as an enhancer for this permeant (50). FTIR
has also been used to show two discrete types of enhancer mechanism. The presence
of Azone increases the diffusion coefficient of the model permeant 4-cyanophenol,
whereas the enhancer Transcutol has no effect on D but increases the solubility of
the permeant in the skin (51).
In a structure–activity study, several permeants (alkyl anilines and phenyl al-
kanols) were examined. The log octanol–water partition coefficients (log Koct) varied
from approximately ⫺1 to ⫹4. The efficiency of Azone depended on the lipophilicity
of the permeant (see foregoing results with snake skin). The effect was dependent
on the concentration of Azone used. At 1%, it acted on compounds the log Koct of
which was < 1; however, at 5% the Azone threshold was increased to a log Koct of
2.7 (52). Compounds that are hydrophilic (e.g., methotrexate) also have their per-
meation increased by Azone (53). Because Azone influences skin lipids, the meth-
otrexate must transfer through or across the lipid domains, suggesting that hydro-
philic molecules do not permeate through a unique polar route.
Azone and its analogues (see Fig. 1) have been investigated in vitro using
human skin and the model permeant metronidazole (20). The results are shown in
Fig. 3. Enhancement ratios calculated at 40 min are Azone, 6.7; N-0539, 6.4; N-
0253, 3.4; N-0721, 1.4; N-0131, 1.1; and N-0915, 0.2. The sulfur analogue of Azone
(N-0721) is significantly less active than Azone, and the short hydrocarbon chain in
N-0131 renders it ineffective. The phase transition temperature of multilamellar
DPPC liposomes is lowered by the presence of the enhancers in rank order of their
enhancing abilities, except N-0915, which increases Tm. This can be explained by
alteration of the lateral bonding within the stratum corneum lipid lamella. A model
based on molecular graphics of Azone and N-0915 and their hydrogen-bonding ca-
pacities to cerebrosides has been proposed. It is possible that models such as this
can be used in the future to design more specific and active penetration enhancers.
One of the problems of penetration enhancers is that they tend to be nonspecific
in their action. A study on Azone and sodium lauryl sulfate (SLS) showed that the
presence of Azone in the skin enhanced the subsequent absorption of the SLS
(54,55). This has implications in the use of penetration enhancers and precautions
Figure 3 A comparison of the effects of different enhancers on the permeation of metro-

nidazole through human full-thickness skin.
may be necessary to avoid enhanced irritancy caused by subsequent contact of the

treated skin with simple household products, such as detergents.
3. In Vivo Experiments
There have been fewer in vivo experiments conducted using Azone particularly from
a mechanistic viewpoint. In part this results from the difficulty in interpreting in vivo
data. When blood levels are measured, the variability in the data is increased owing
to the additional clearance kinetics. In general, the in vivo determinations correlate
with what has been found in vitro. A review article (56) considers the clinical ex-
periments in which Azone had been examined, and little additional information has
appeared in the literature since then. Some representative publications include the
following.
Methotrexate permeation has been examined in vitro and in vivo in hairless
mice, and increased blood levels were found using Azone as an enhancer (57). This
280 Davis et al.
species has also been used in a study on acyclovir and its ability to reduce the herpes
simplex virus-1 (HSV-1) (58), on nitroglycerin absorption (59) and methotrexate
(60). In vitro and in vivo experiments have also been conducted using rat skin with
the permeant bromhexine. Azone enhanced permeation both in vitro and in vivo
(61). A novel dopaminergic D2 agonist, S(⫺)-2-(N-propyl-N-2-thienylethylamino)-5-
hydroxytetralin, was also tested, using a rat model. Azone was effective both in vitro
and in vivo (62). Similarly, the ␤-adenergic–blocking agent pindolol has been ex-
amined in rabbits, which comparable results (63). Nitrendipine patches containing
Azone have been examined in humans, with encouraging results. There was no
apparent irritancy problem in either humans or rabbits (64). Various experiments
have been conducted on the topical corticosteroid triamcinolone acetonide (TA).
Tritiated steroid was used to quantitate the amount of drug absorbed. In the presence
of Azone, higher amounts of TA were absorbed (65).
Several investigations have been conducted into the distribution of Azone in
the skin and its subsequent elimination and metabolism (66–70). Owing to the high
lipophilicity of Azone (estimated log Koct = 7.8) it rapidly penetrates into the lipids
of the stratum corneum. Further transit into the viable tissue is hindered by the very
poor water solubility of the enhancer.
4. Synergy
The predominant mechanism of action of Azone is to increase the diffusion coeffi-
cient of the permeant in the stratum corneum. A simple evaluation of Fick’s first law
of diffusion [see Eq. (3)] shows that enhancers should be capable of acting on the
diffusion coefficient in the skin as well as the solubility of the permeant in the skin
lipids. The foregoing experiments (51) showed that Azone influenced D whereas
Transcutol altered the solubility of the model permeant cyanophenol. Fick’s law
demonstrates that, if both enhancement strategies are used, there should be a mul-
tiplicative effect (i.e., synergy should be seen).
Many publications indicate the use of Azone in combination with propylene
glycol. It is probable that synergistic effects have inadvertently been masked. When
relatively systematic studies have been conducted, synergism has been demonstrated.
For example metronidazole penetration is enhanced both by Azone and by propylene
glycol alone: together synergism is found (71). A similar effect is seen for combi-
nations of Azone and Transcutol for the permeant prostaglandin E2 (72).
Recent interest in peptide and oligonucleotide delivery into and through the
skin has promoted interest in physical methods of penetration enhancement such as
iontophoresis, electroporation, and ultrasound (sonophoresis). Theoretically, syner-
gism between physical and chemical enhancement should be possible. In the limited
studies reported in the literature, it is difficult to produce guidelines in formulation
strategies. Azone appears to act synergistically with iontophoresis in the delivery of
metoprolol (73). However, there was no apparent synergistic effect for sotalol (74).
In vivo impedance spectroscopy has shown that an Azone–propylene glycol mixture
had a profound effect on the postiontophoretic skin impedance, considerably ampli-
fying the effect of current passage (75).
Fewer experiments have been conducted with ultrasound, but synergy has been
reported for the combined effect of 150 kHz ultrasound, at 111 mW/cm2 intensity,
and Azone. Aminopyrine was the permeant used in the study with excised hairless
rat skin (76). Clearly, more work is indicated to determine the precise mechanism
of synergy and how it is influenced by the physicochemical properties of the

penetrant.
B. SEPA (2-n-Nonyl-1,3-dioxolane)
SEPA is a registered tradename and an acronym for ‘‘soft enhancement of percuta-
neous absorption.’’ SEPA represents a patented class of compounds containing var-
ious substituents (77). The ring system may be dioxane or dioxolane, and the side
chains may vary. SEPA 0009 is 2-n-nonyl-1,3-dioxolane (Fig. 4)—the congener most
widely studied—and is the dioxolane for which the most safety data has been com-
piled. ‘‘SEPA’’ is used in this section to represent SEPA 0009. Because SEPA has
only carbon, hydrogen, and oxygen, minimal central nervous system (CNS) or other
untoward pharmacological activity is not expected, and none has been observed in
safety tests. It is also metabolically benign, being metabolized and readily excreted.
These safety properties are the reason for the ‘‘soft enhancer’’ designation.
1. Safety
SEPA has undergone extensive safety testing in animals, including testing for acute,
chronic, and mutagenic effects, with no problems arising. In addition, human safety
studies have been conducted as follows:
Irritation studies: 73 volunteers
Sensitization studies: 358 volunteers
Phototoxicity study: 30 volunteers
Photoallergenicity studies: 156 volunteers
Phase II and III studies: more than 3500 patients
2. Efficacy
SEPA’s broad-spectrum activity is best demonstrated when compared with other en-
hancers. Studies were performed in a conventional vertical Franz-type cell using
neonatal porcine or human transplantation skin. The formulations studied were stan-
dard hydroalcoholic gels, with drug and enhancer present in equivalent amounts. For
equivalent formulations that contained the lipophilic drug papaverine, SEPA was
approximately twofold more effective than zone over a 24-h period (Fig. 5). Azone,
although less effective than SEPA, does show useful transdermal applications for
this drug. SEPA also showed good penetration enhancement for estradiol (Fig. 6);
although, in this example, Azone showed little enhancement over the control without
an enhancer. SEPA was also a more effective enhancer than Azone for ketoprofen
(Fig. 7). Therefore, SEPA displayed a broad-spectrum enhancement that was not
totally dependent on the molecule being delivered. Both polar and nonpolar drugs
are strongly enhanced (78,79).
These example results were generated using in vitro testing methods. It is
absolutely crucial to have a transdermal in vitro testing method in place that is
Figure 4 Structure of SEPA.

282 Davis et al.
Figure 5 A comparison of the effects of SEPA and Azone on the percutaneous flux of
papaverine.
estradiol.
ketoprofen.
predictive of the in vivo condition. The following are two examples of in vitro–in
vivo correlation studies that have been carried out using SEPA.
3. In Vitro–In Vivo Correlation
a. SEPA Minoxidil Solutions. Figure 8a shows the results obtained after a
series of optimization studies had been performed using neonatal porcine skin. The
10% SEPA formulation shown had been selected earlier through a series of studies
to be the best in vitro performer (drug-delivery performance criteria were selected
from factors including peak flux, time to peak flux, and total drug delivery) from an
array of prototype formulations. As described earlier, the porcine skin model gives
results comparable with those of human skin in vitro models. Figure 8a shows how
this chosen optimized candidate performs in vitro versus the appropriate unenhanced
control. Figure 8b shows the results obtained after minoxidil (Rogaine) solutions
were applied twice daily to the scalp of human volunteers. These solutions were of
similar composition, except for the amount of enhancer, SEPA, added to the for-
mulation. Serum minoxidil levels were measured at appropriate time points. Figure
8b shows plasma minoxidil values expressed as nanograms per milliliter (ng/mL) of
serum. The serum profile clearly shows enhancement above the Rogaine reference
treatment; moreover, the optimal SEPA concentration is clearly seen to confirm the
choice of the 10% formulation, as forecast by the earlier in vitro optimization studies.
b. SEPA Ibuprofen Gels. Figure 9a shows the results obtained after a series
of optimization studies had been performed in human skin with an ibuprofen-con-
taining gel. Similar to the minoxidil approach, the 10% SEPA formulation containing
5% ibuprofen had earlier been selected from a series of studies to be the best in
vitro performer from a group of ibuprofen formulations. Figure 9a shows how this
optimized candidate performed in vitro compared with several commercially ob-
284 Davis et al.
Figure 8 A comparison of the in vitro and in vivo effects of SEPA: (a) in vitro minoxidil
permeation enhancement (n = 6, porcine skin); (b) in vivo serum levels of minoxidil following
topical application (n = 21, human volunteers).
tained products; much higher flux and cumulative drug permeation were measured
for this one than for any of the other formulations. Figure 9b shows the results
obtained after a phase 1-2 clinical study on volunteers with the same composition
gel. The delayed onset muscle soreness (DOMS) model for exercise-induced pain in
the triceps muscle was employed. Forty-eight hours after heavy exertion of the tri-
ceps, 2.5 g of a hydroalcoholic gel, either pooled placebo (placebo with and without
Figure 9 A comparison of the in vitro and in vivo effects of SEPA: (a) in vitro ibuprofen
permeation enhancement (n = 6, human skin); (b) subject global evaluation of pain relief
following exercise (severe pain group).
SEPA) or treatment, was applied as a single dose. Significant clinical alleviation of

pain (asterisks; p < 0.05) in a higher percentage of the treatment group than that of
the placebo (score 0–4; change of 1 unit is considered clinically significant) was
seen in the group initially reporting severe pain. No significant difference was seen
for the 5% ibuprofen hydroalcoholic gel without SEPA compared with the placebo.
Most importantly, statistically significant differences were seen as early as 30 min
after application of the gel, and the effect lasted for 5 h. This early onset of action
and persistence of activity with SEPA formulations is noteworthy. This muscle sore-
ness model, however, is acknowledged to be a difficult model to interpret because
the induced pain is spontaneously resolving during the testing phase.
286 Davis et al.
4. Mode of Action
There are three major areas that are currently under investigation to explain or predict
penetration enhancement results. These areas are generally thought to be barrier
function modification of the stratum corneum, partitioning effects between the for-
mulation and the stratum corneum, and thermodynamic activity of the drug in the
formulation.
There are several changes to the stratum corneum that may influence transder-
mal penetration. These include alteration of the cellular and intercellular lipid com-
position (e.g., by increasing fluidization of lipids, such as with SEPA and other
lipophilic enhancers); by removing lipids, such as with dimethyl sulfoxide or ethanol;
effects on intercellular organization and cohesion; disruption or reordering of the
water structure; and alteration of stratum corneum proteins.
Some workers have suggested that there may be two pathways through the
stratum corneum barrier: one lipophilic, the other hydrophilic. The lipoidal pore
pathway has been proposed on the molecular scale based on observations of the
effect of alkyl-pyrrolidones on the hairless mouse skin permeation of steroids (80).
Large lipoidal ‘‘pores,’’ in a gross morphological sense, have also been described
with lipophilic terpene enhancers postulated to form pools, possibly adjacent to the
stratum corneum intercellular lipid lamellae (81). Apparent aqueous pores have also
been shown after hyperhydration of the stratum corneum (82). Engstrom et al. (83)
have suggested a mechanism of action that describes ‘‘Azone holes,’’ complex struc-
tures that combine both lipophilic and hydrophilic conduits. The phenomenon of
disparate pore formation appears logical from a physicochemical perspective, as any
reduction in anisotropy to produce a hydrophilic pore must concomitantly produce
the corresponding lipophilic pore, akin to a partitioning effect. It may also be the
best explanation for the broad-spectrum action of SEPA, which may provide the best
combination of the two proposed actions: membrane fluidization anisotropy and dis-
ruptive reordering to form conduits that are both aqueous and lipoidal.
The thermodynamic activity of any transdermal delivery system, whether a
liquid, a semisolid, or a patch-type system, is of large consequence to the percuta-
neous absorption of the drug and the enhancer. This will be discussed in Sec. III,
particularly relative to supersaturated solutions.
However, there is a more fundamental underlying mechanism that needs to be
discussed, an aspect of dermal and transdermal delivery that cannot be overly em-
phasized and is the facet least explored. The underlying physics, however, are well
characterized, based on fundamental physical laws, and have been utilized in the
past by other disciplines, such as polymer science, the science of films, film for-
mation, and coating technologies.
As a background for this discussion, recall that Fick’s second law of diffusion
[see Eq. (5)] describes the way in which concentration changes over time for any
region of a solution. It does not, however, apply to non–steady-state conditions,
concentrated solutions, or conditions of partial miscibility.
⭸c
= D (⭸2c/⭸x2) (5)
⭸t
The Fickian diffusion coefficient is here defined in terms of a concentration
gradient. However, the actual driving force for the flux of any component in a more
complex system is not the concentration gradient, but the chemical potential
gradient.
Di = Bi (␦␮i/␦xi) = 0 (6)
Equation (6) describes how the chemical potential (␮i) of a nonideal solution can be
related to the diffusion coefficient Di at constant pressure and temperature. Bi is a
mobility term that is always positive. This relation is well known in the physico-
chemical field of film forming, specifically in coacervation technology (84), and it
pertains to partially miscible systems, such as those that can be obtained with hy-
droalcoholic solutions, gels, and emulsions.
Figure 10 shows a hypothetical, but typical example of a partially miscible
system containing three components that include a lipophilic phase (which may also
contain an enhancer), a hydrophilic phase (aqueous), and a homogenizing phase (e.g.,
alcohol). As alcohol is allowed to evaporate, such as happens after placement of a
hydroalcoholic gel on the skin, line AB in the phase diagram is traversed (for ref-
erence, composition A in the phase diagram comprises 60% homogenizing phase,
Figure 10 Phase diagram of a partially miscible system.

288 Davis et al.
30% lipophilic phase, and 10% hydrophilic phase). As the homogenizing phase either
evaporates or is absorbed, next indicated is composition X, the point of the phase
boundary condition. After crossing this point into the miscibility gap, phase separa-
tion, or phase inversion occurs, usually very rapidly. At this time the migration of
solution components proceeds from an area of low concentration to high concentra-
tion (an almost pure one of each phase).
Migration of a particle (of a solvent or a solute) from low concentration to
high concentration seems contrary to Fick’s second law (but recall that Fick’s law
is only for ideal conditions. The enhancer can also be driven into the stratum corneum
according to this physics. When a solute (drug) starts out dissolved in the homoge-
neous phase (at compositional point A; see Fig. 10), and it subsequently finds itself
in the miscibility gap after separation into two phases, it can now be postulated to
be in two distinct solutions, supersaturated in one or both. If it prefers one phase
thermodynamically, it may (and probably will) not have time to equilibrate into the
preferred phase before the phases separate (kinetics overrule). Now, the chemical
potential for the drug (and enhancer) realized may be of temporarily very high mag-
nitude. The solute (or permeation enhancer, or both) may now relieve this potential
energy by partitioning into the stratum corneum membrane (or by giving off heat,
crystallizing, etc.). If the formulation can produce a chemical potential gradient at
the moment of interfacial skin contact, and preserve the gradient for the appropriate
time fame (assuming the barrier is opened by the appropriate enhancer system), then
a formulation can drive high amounts of drug into the stratum corneum. This is also
true for concomitantly driving molecules of the enhancer into the stratum corneum,
preferably in the same time frame, so that enhancement may be optimal.
This effect may be termed an energetic epidermal injection, or hyperflux, con-
dition. The effect is largely to be measured empirically, because of the complexities,
and each drug must have its formulation optimized for amount of drug delivered
and the time course of delivery. Another way to say this is that the escaping tendency
(fugacity) of any formulation component (drug or enhancer) will change over time
and can be maximized for each component following physical principles (tending
over time to always decrease enthalpy and increase entropy).
Figure 11 shows the complexity arising after evaporation of even a very simple
emulsion (85). After evaporation of the decane and water, different compositions and
complex structures may appear. Prediction of fluxes from such structures becomes
very difficult indeed. The structures and forms that appear depend on the balance
between thermodynamic and kinetic considerations as time progresses after dosage.
Formed structures that are low-energy (low-enthalpy), such as the lamellar liquid
crystal depicted in Figure 11 may lower the fugacity of any single subsumed com-
ponent. On the other hand, if phase-inverted systems appear, then the kinetics may
be such that the phase (or the solute, or both) may be injected into the stratum
corneum with ease.
Three postulations can be summarized as follows: (1) a metastable state can
be achieved by combining solvents of limited miscibility; (2) when the homogenizing
solvent evaporates, situations can be produced in which the gradients of solute con-
centration and gradients in the chemical potential are of different sign; and (3) in
this state, the diffusion coefficient, defined by Fick’s law, will be negative. Thus, a
component can be induced to flow from a region of low concentration to a region
of high concentration within the vehicle or within the stratum corneum (101).
Figure 11 Phase diagram illustrating the complexities of emulsion evaporation.
C. Miscellaneous Chemical Enhancers

In addition to SEPA and Azone, a few other specifically designed enhancing mole-
cules have been evaluated. These include 1-[2-(decylthio)ethyl]azacyclopentan-2-one
(HPE-101), 4-decyloxazolid-2-one (Dermac SR-38), and dodecyl-N,N-dimethyl-
amino isopropionate (DDAIP, NexACT 88) (Fig. 12).
1. HPE-101
HPE-101 is believed to have a mechanism of action similar to that of Azone and is
also very sensitive to the vehicle of application (86,87). Thus, although the enhancer
significantly increased the urinary excretion of indomethacin following topical ap-
plication to hairless mice, it was dependent on the application vehicle (87). When
Figure 12 Structures of specifically designed skin penetration enhancers.

290 Davis et al.
the enhancer was applied in solution in polar solvents, such as dipropylene glycol,
triethylene glycol, diethylene glycol, glycerin, water, or triethanolamine, enhance-
ment ratios varied between 1.5- and approximately 67-fold. However, when applied
in solution in more lipophilic solvents, such as ethanol, isopropanol, oleyl alcohol,
isopropyl myristate, or hexylene glycol, no enhancement was observed (Table 2).
These data stress the importance of optimization of the delivery vehicle, not only
for the drug, but also for the enhancer. Combinations of HPE-101 with cyclodextrins
appear to be useful means to improve drug permeation across the skin (88). All of
the available data, however, have been obtained using hairless mice or other small
laboratory animal skin. Small laboratory animals, especially the hairless mouse, can
be uniquely sensitive to skin penetration enhancement and, as yet, the effectiveness
of HPE-101 on human skin has not been reported.
2. Dermac SR-38
Dermac SR-38 is one of a series of oxazolidinones, cyclic urethane compounds,
evaluated as transdermal enhancers (89). The compound was designed to mimic
natural skin lipids (such as ceramides), to be nonirritating, and to be rapidly cleared
from the systemic circulation following absorption (90). In animal and human safety
studies, Dermac SR-38 demonstrated a good skin tolerance (no observed irritancy
Table 2 HPE-101 Enhancement

Ratios for Indomethacin Absorption in
Hairless Mouse (Based on Urinary
Excretion)—Vehicle (Solvent) Effects
Vehicle ER a
Dipropylene glycol 21.1

Triethylene glycol 67.0
Hexylene glycol 1.54
Diethylene glycol 15.2
1,3-Butylene glycol 9.47
Trimethylene glycol 21.3
Glycerin 8.19
Water 3.09
Oleyl alcohol 0.86
Isopropanol 0.96
Ethanol 0.78
Silicone 2.47
Peppermint oil 1.08
Olive oil 0.78
Isopropyl myristate 0.80
Triethanolamine 5.88
a
ER, enhancement ratio/urinary excretion of
indomethacin when applied topically with
HPE-101 divided by urinary excretion of in-
domethacin when applied topically without
HPE-101.
Source: Ref. 87.
or sensitization at levels of 1–10% (w/w); moderate to severe irritation in rabbit at

100%); and a low degree of acute toxicity (LD50(rat oral) > 5.0 g/kg). The compound
was evaluated for its ability to enhance the human skin permeation of diverse drugs
from dermal and transdermal delivery systems (Table 3). The data for minoxidil
clearly indicate an enhancer concentration-dependent effect for permeation enhance-
ment. Dermac SR-38 enhances the skin retention of both retinoic acid, when applied
in Retin A cream, and dihydroxyacetone, when applied in a hydrophilic cream (91).
3. NexACT 88
NexACT 88 (dodecyl-N,N-dimethylaminoisopropionate; DDAIP) is one of a series
of dimethylamino alkanoates, reported to be biodegradable, which were developed
as potential nontoxic skin permeation enhancers (92). Much of the early work was
carried out using shed snake skin and, with this model, most of these compounds
were equal to, or more active than, Azone (25,92). Studies using human skin indi-
cated that dodecyl-N,N-dimethylaminoacetate (DDAA) was a more effective en-
hancer of absorption of propranolol hydrochloride and sotalol than was Azone
(74,93). Structural optimization of the compounds led to the identification of the lead
candidate dodecyl-N,N-dimethylaminoisopropionate (94,95) that appeared to be more
effective than DDAA. Mechanism of action studies indicated that the distribution of
DDAIP in stratum corneum lipids was somewhat different from that of DDAA (96),
suggesting that other interactions were contributing to the penetration enhancement
effect. It is possible that, in addition to its effect on stratum corneum lipids, DDAIP
may interact with keratin and potentially increase stratum corneum hydration.
Table 3 Dermac SR-38 Enhancement Ratios for Drug Permeation

Across Human Skin
Dermac SR-38
Drug Vehicle (wt%) ERa
Lidocaine Cream 2.5 3.2

Prilocaine Cream 2.5 4.1
Diclofenac Gel 2.5 2.3
Hydrocortisone Cream 1.0 3.0
Indomethacin Cream 1.0 1.8
Minoxidil Solution 2.0 2.6
Solution 5.0 7.3
Solution 10.0 10.4
Alprazolam Patch 5.4 1.6
Diltiazem HCl Cream 3.0 2.4
Isosorbide dinitrate Patch 8.0 3.0
Cream 1.4 1.6
Morphine sulfate Patch 3.0 1.8
Progesterone Patch 5.0 5.1
Nifedipine Cream 5.0 2.6
a
ER, enhancement ratio/in vitro skin permeation of drug when applied with Dermac
SR-38 divided by skin permeation of drug when applied without Dermac SR-38.
Source: Ref. 90.
292 Davis et al.
Preliminary animal toxicity studies carried out using DDAA and NexACT 88
indicated that these compounds had low toxicity (92), but may be mildly irritating
to rabbit skin at high concentrations. The overall safety and efficacy data suggest
that the dimethylamino alkanoates may be useful as transdermal penetration enhanc-
ers. More human skin data would be beneficial, and it is important to appreciate that,
in common with other skin penetration enhancers, reduction to practice in a clinical
situation is a prime requisite for determining efficacy.
4. Others
Other compounds have been identified and have undergone preliminary evaluation
as potential skin-penetration enhancers. The data are, however, very limited and the
candidate enhancers are mentioned here solely for completeness. The biodegradable
fatty acid esters of N-(2-hydroxyethyl)-2-pyrrolidone (decyl and oleyl) were synthe-
sized and evaluated for enhancer activity using hairless mouse skin (97). Permeation
of hydrocortisone was enhanced twofold. The activity of n-pentyl-N-acetylprolinate
as a skin permeation enhancer has been determined using human skin (98).
III. SUPERSATURATED SYSTEMS

The effect of formulation on drug bioavailability is much greater in topical drug
delivery than in any other route of administration. Components of the vehicle, in-
cluding chemical penetration enhancers, can interact with the drug and the skin to
influence both the rate and extent of absorption. However, increasing the number of
excipients in the vehicle will inevitably lead to an increase in the potential of the
vehicle to induce some form of dermatotoxicity (irritancy, sensitization, etc.). Su-
persaturated systems provide an enhancer strategy with the capability of low-dose
therapeutic efficacy and improved control of percutaneous absorption. Furthermore,
these systems can be formulated with a reduced number of excipients and, therefore,
may lead to a reduction of dermatotoxic events.
A. Physicochemistry of Supersaturated Systems

1. Supersaturation
James (99) defined a solution as a molecular dispersion of a solute in a solvent.
Under equilibrium conditions, a solution is capable of a continuous variation in
composition within certain limits. At the lower limit the solution is represented by
the pure solvent and, at equilibrium, the upper limit occurs when the solvent is
incapable of dissolving any more solute and a second phase of undissolved solute is
present. A solution in equilibrium with undissolved solute is known as a saturated
solution. In saturated solutions the thermodynamic activity of the solute is equal to
that of the pure solute, and this value is arbitrarily taken as unity. Thus, supersatu-
rated solutions are those in which the thermodynamic activity of drug in solution is
greater than unity. Supersaturated systems may be produced using the dependence
of solubility on temperature, pH, and solvent composition, and by utilizing changes
in these factors and by other methods, to be described.
2. Crystal Growth and Effects of Additives
Solutions that contain concentrations of solute in excess of the saturated solubility
will undergo phase change to form a suspension of solid that is in equilibrium with
its saturated solution. This process is known as recrystallization. Excess solute in

solution (supersaturation), however, will not necessarily always cause crystallization
to occur. In addition to supersaturation, crystallization requires formation of nuclei
of critical size, followed by crystal growth around the nuclei. At high levels of
supersaturation solutions are labile and nucleation and crystal growth occur sponta-
neously. At lower levels of supersaturation a metastable zone may be produced as
shown in Figure 13.
Additives are materials that interfere with the processes of nucleation and crys-
tal growth and act to stabilize supersaturated solutions. The mechanism of action of
these materials is uncertain, but they may work by inhibition of bulk diffusion,
inhibition of crystal surface diffusion, or by poisoning crystal growth sites. Inhibition
of bulk diffusion by an increase in bulk viscosity, seems unlikely to be important in
topical creams, gels, and ointments, but may be significant in the viscous adhesive
layer of transdermal patches and in stratum corneum. The crystal growth model of
Cabrera and Frank (100) suggests that surface migration of molecules adsorbed onto
a crystal is required before incorporation into the crystal growth sites can occur. Put
simply, molecules setting down on a flat surface of a crystal have only a single
binding surface and are likely to redissolve. At step-and-kink (corner) sites there are
two and three surfaces for bonding; thus, redissolution is much less likely, enabling
the molecule to be incorporated into the crystal structure. Thus some additives, for
example, polymers, bind to the surface of the crystal to inhibit the surface diffusion
of alighting molecules to the step-and-kink sites and allow redissolution to occur.
Figure 14 shows the effects of polymer additives on the rate of crystal growth from
an eightfold supersaturated solution of hydrocortisone acetate. All polymers studied
showed some effect, with the 1% hydroxypropyl methylcellulose (HPMC) system
Figure 13 Diagram showing the critical degree of supersaturation and the different stability
states of supersaturated systems. (From Ref. 124).
294 Davis et al.
Figure 14 The effects of various polymer additives on the rate of crystal growth from an
eightfold supersaturated solution of hydrocortisone acetate. All of the polymers studied
showed some effect, with the 1% hydroxypropyl methylcellulose system showing no change
over 3 days compared with a 30-min half-life in the no-polymer control. (From Ref. 109.)
showing no change over 3 days compared with the 30-min half-life of the untreated
system.
Kondo et al. (101) studied the effect of a wide range of polymers on plasma
levels of nifedipine following application of supersaturated solutions to rat skin. All
polymer systems generated much higher plasma levels than control systems. For
example, polyvinylpyrrolidone systems gave 40-fold increases in plasma levels of
nifedipine, and CAP systems demonstrated 75-fold increases. Recently, Ikeda et al.
(102) demonstrated superior effects of hydrophobically modified HPMC over HPMC
in inhibiting recrystallization of indomethacin in a topical gel. Other authors have
also published in this area (103–105). It is clear that for many compounds high
levels of supersaturation can be stabilized using polymeric systems. The ability of
additives, particularly polymers, to stabilize supersaturated solutions over useful time
periods is the key to the practical use of supersaturation in topical drug delivery.
Thus, a major area for further research is structure–activity of polymer antinucleant
additives.
Other compounds that are similar in structure to the crystal molecule may work
by a poisoning mechanism (i.e., partial blocking of step-and-kink sites) to inhibit
crystal growth. To our knowledge, however, these have not yet been used in topical
drug delivery systems.
B. Supersaturation Solutions to Enhance Penetration

1. Methods to Form Supersaturated Solutions
Supersaturated systems designed to enhance percutaneous penetration have been pro-
duced mainly by utilizing changes in the dependence of drug solubility on temper-
ature and solvent composition. Some work has been conducted on supersaturated
solutions formed from amorphous materials produced by mechanical processes
(106,107). Earlier reviews that are available on enhancement of percutaneous pen-

etration contain sections on supersaturation (108–112). Only data from solid disper-
sion systems and cosolvent systems that address key issues on supersaturation will
be reviewed here; thus, these methods are described only briefly.
a. Solid (Molecular) Dispersion Systems, Coprecipitates. Solid dispersion
systems were first produced by Chiou and Riegelman (113) in an attempt to increase
the dissolution rate of poorly water-soluble drugs. Such systems are made either by
melting drug and carrier (often a polymer) together or by dissolving them in a
common volatile solvent, with subsequent removal of solvent by heat. This process
is similar to that currently used to prepare thin-film transdermal patches, and this
will be discussed later. Depending on the conditions, dissolution of solid dispersions
will often result in generation of supersaturated states. These supersaturated states
will often be maintained for a considerable period by the antinucleant and anticrystal
growth effects of the polymeric carrier (103,114–117); therefore, it is not surprising
that such systems have been investigated for their ability to increase percutaneous
absorption (118–121) [see also, Japanese patents: J6 3093-714-A, J6 3093-715-A,
J6 3307-818-A, J6 3307-819-A, and J6 3297-320-A].
b. Mixed Cosolvent Systems. Saturation solubility plots often show an ex-
ponential increase with cosolvent composition, depending on the relative polarities
of the solute and binary cosolvent system (Fig. 15; curve B⬘B). A basic property of
these systems is that by mixing suitable solute–cosolvent solutions, subsaturated,
saturated, and supersaturated solutions can be formed. Figure 15 shows schematically
that mixing system B (saturated solute in 100% solvent b) with system A (no solute
in 100% solvent a) will result in systems C (subsaturated), D (saturated), E and F
(both supersaturated), depending on the ratio of A to B. In practice, systems A and
B may themselves be mixed, and system B need not necessarily be saturated. The
degree of saturation is calculated by dividing the resulting concentration after mixing
Figure 15 The mixing of system B (saturated solute in 100% solvent b) with system A
(no solute in 100% solvent a) will result in systems C (subsaturated), D (saturated), and E
and F (both supersaturated), depending on the ratio of A to B used. In practice, systems A
and B may themselves be mixed, and system B need not necessarily be saturated. (From Ref.
122.)
296 Davis et al.
with the experimentally determined value of the saturated solubility at that same
composition. Thus, supersaturated solutions of known ‘‘thermodynamic activity’’ and
drug concentration can be produced readily using a simple-mixing process. This
process is similar to the premix process currently used to manufacture topical semi-
solids. Because of their versatility to form supersaturated solutions at specific drug
concentrations and degrees of saturation, mixed cosolvent systems are being used
increasingly as research tools (122–126).
2. Membrane Transport from Supersaturated Solutions

It is clear that high degrees of supersaturation can be stabilized by the use of additives
and, theoretically, this should lead to enhancement of membrane transport that is
linearly proportional to the degree of saturation. Figure 16 shows a simple physical
model (8) that illustrates thermodynamic activity gradients in the skin after topical
application of saturated and threefold supersaturated systems. At the outer surface of
the stratum corneum activity in the vehicle (avehicle) is equal to activity in the stratum
corneum (asc); thus, following application of a threefold supersaturated solution, the
outer layer of the stratum corneum is also threefold supersaturated. A key question,
especially if the applied supersaturated state is stabilized with an antinucleant addi-
tive, is what happens to the physical state of the drug within the stratum corneum?
The two extreme possibilities are shown in Figure 16. In case 1, immediate precip-
itation of the drug occurs in the outer surface of the stratum corneum, the thermo-
dynamic activity gradient is as shown by the path A–B–C, and there is no significant
increase in transport over a saturated system. In case 2, no precipitation occurs. There
is a linear drop in thermodynamic activity over the barrier along path A–C, with
supersaturation being maintained over two-thirds of the barrier (this depends on the
initial degree of supersaturation), and the increase in membrane transport is linearly
proportional to the degree of saturation. In the next section membrane transport from
saturated and supersaturated systems will be reviewed.
Figure 16 Diagram of a simple physical model (after Higuchi) that illustrates the theoret-
ical thermodynamic activity gradients in the skin after topical application of supersaturated
systems. If immediate precipitation of the drug occurs in the outer layer of the stratum cor-
neum, the thermodynamic activity gradient is as shown by the path A–B–C. If no precipitation
occurs, there is a linear drop in the thermodynamic activity over the barrier along path A–C.
a. Cellulose Porous Membranes. The transport of hydrocortisone across po-

rous cellulose membranes from supersaturated solutions formed from polyvinyl py-
rolidane (PVP) coprecipitates has been studied by several groups (118–120). In these
systems PVP worked as both a carrier, to form the coprecipitate dispersion system,
and as an additive, to prevent crystal growth. Flux was linearly proportional to the
degree of supersaturation up to approximately 12-fold supersaturation (118). A sim-
ilar experiment (119) demonstrated that at early time points, up to approximately 6
h, flux was broadly proportional to the degree of saturation (14-fold supersaturation
generated a 10-times increase in flux). At later times, and at higher degrees of su-
persaturation, crystallization was seen in the donor phase, and flux no longer in-
creased linearly with degree of supersaturation. Figure 17 (120) shows that for up
to sixfold supersaturation of hydrocortisone, the experimental data are a good fit with
the theoretical line drawn from the origin through point Cs–Js, the flux from a
saturated solution. At higher degrees of supersaturation, rapid crystallization was seen
in the donor phase. Merkle (120) also studied transport of benzodiazepines across
porous cellulose membranes from supersaturated solutions formed from PVP copre-
cipitates. Figure 18 demonstrates that for nitrazepam, and up to 10-fold supersatu-
ration, the experimental data are a good fit with the theoretical point Cs–Js. Similar
results were reported for clonazepam (14-fold increased activity–flux) and flunitra-
zepam (5-fold increased activity–flux). Higuchi and Farrar (127) reported flux of
digoxin from supersaturated solutions formed from PVP coprecipitates of up to 20
times that from saturated solutions, provided crystallization was prevented by inclu-
sion of a further additive.
In summary, for a variety of compounds, large increases in flux across cellulose
porous membranes can be observed with increasing degrees of supersaturation, pro-
vided crystallization within the donor vehicle can be prevented by use of additives.
Figure 17 Transport of hydrocortisone–alcohol across cellulose membrane: for up to a

sixfold supersaturation, the experimental data are a good fit with the theoretical line drawn
from the origin through point Cs–Js, the flux from a saturated solution. At higher degrees of
supersaturation, rapid crystallization was seen in the donor. (From Ref. 120.)
298 Davis et al.
Figure 18 Transport of nitrazepam across cellulose membrane: for up to a tenfold super-

saturation, the experimental data are a good fit with the theoretical line drawn from the origin
through point Cs–Js, the flux from a saturated solution. (From Ref. 120.)
However, the mechanism of transport through porous membranes is not fully rep-
resentative of membranes in which transport is dependent on partitioning into and
diffusion through the membrane and in which efficient exclusion of additives from
the membrane may occur.
b. Silicone Membranes. Silastic (polydimethylsiloxane; silicon rubber) has
been widely used as a model membrane for the comparison of relative penetration
rates in transport studies (128,129), and has been validated to show response to
changes in thermodynamic activity (122,128). Davis and Hadgraft (122) used mixed
cosolvent systems to form supersaturated solutions of hydrocortisone acetate, stabi-
lized with 1% hydroxypropylmethyl cellulose as shown in Figure 14. Figure 19 (122)
shows that, for systems up to eightfold supersaturated, flux across Silastic is linearly
proportional to the degree of saturation. Pellett et al. (124) studied the transport of
piroxicam across Silastic from supersaturated solutions in propylene glycol–water,
stabilized with hydroxypropylmethyl cellulose. Owing to precipitation problems in
the donor phase, an upper limit of four times supersaturation was evaluated and
donor solutions were renewed every 12 h. Figure 20 (124) shows that, for systems
up to fourfold supersaturated, flux across silastic is linearly proportional to the degree
of saturation. Megrab et al. (126) studied the transport of estradiol across Silastic
from supersaturated solutions in propylene glycol–water, stabilized with providone
(PVP-2). Flux across Silastic from saturated solutions of estradiol increased with
increasing propylene glycol content owing to the effect of propylene glycol on in-
creasing estradiol-saturated solubility within the membrane. For this reason, flux
from supersaturated systems was normalized by flux from the corresponding satu-
rated vehicle. Figure 21 (126) shows that only at low degrees of supersaturation (up
to fourfold) is normalized flux linearly proportional to the degree of saturation.
Throughout these studies no apparent instability was seen in the donor phase, even
at higher degrees of supersaturation. Thus, the possibility exists for precipitation
Figure 19 Transport of hydrocortisone acetate across a Silastic membrane: for up to an

eightfold supersaturation, flux is linearly proportional to the degree of saturation. (From Ref.
122.)
within the Silastic membrane. Similar results have been obtained with supersaturated
solutions of diclofenac (130).
c. Human Skin In Vitro. In two studies (124,126), comparisons of flux across
silastic and human skin were made. Pellett et al. (124) found, as with Silastic, that
flux of piroxicam across full-thickness human skin in vitro was linearly proportional
to the degree of supersaturation. Figure 22 (126) shows that for up to approximately
12-fold saturation systems, flux of estradiol across epidermal membranes was linearly
proportional to the degree of supersaturation. From these data it appears that super-
saturated states may be more stable within stratum corneum than in a Silastic
membrane.
d. Synergy Between Supersaturation and Chemical Enhancers. There is an-
ecdotal evidence for synergy between supersaturation and chemical penetration en-
hancement. For example, the patent literature describes several results that suggest
enhanced percutaneous penetration in systems containing a drug in solution, a vol-
atile solvent, a residual component (often a potential lipophilic enhancer), and an
antinucleant additive. For example, Kondo (101) described supersaturated formula-
tions of nifedipine based on acetone (volatile) and propylene glycol plus isopropyl
myristate (nonvolatile) solvents. The degree of saturation in the residual phase after
evaporation was calculated at 12-fold, yet up to 75-fold increases in flux were ob-
served, presumably owing to the known coenhancer effects of propylene glyco–
isopropyl myristate mixtures.
Earlier, it was suggested that the increase in the drug-saturated solubility in the
stratum corneum and supersaturated drug concentration in the stratum corneum were
independent variables. Megrab et al. (126) measured uptake (solubility) of estradiol
in stratum corneum from saturated and supersaturated solutions in various propylene
300 Davis et al.
Figure 20 Transport of piroxicam across Silastic membrane: (a) diffusion profiles; (b) flux
as a function of degree of supersaturation. For up to fourfold supersaturation the flux across
Silastic is linearly proportional to the degree of saturation. (From Ref. 124.)
Figure 21 Transport of estradiol across a Silastic membrane: Only at low degrees of su-
persaturation, up to four, is the flux ratio (flux normalized by flux from the saturated vehicle)
linearly proportional to the degree of saturation. Throughout these studies no apparent insta-
bility was seen in the donor phase, even at higher degrees of supersaturation, and it is likely
that crystallization occurred in the Silastic membrane. (From Ref. 126.)
glycol–water vehicles. As propylene glycol content in saturated solutions increased,

so did uptake of estradiol. As degree of supersaturation increased, so also did uptake
ratio (normalized for increase in uptake from the corresponding saturated vehicle).
Figure 23 shows clearly the synergy between enhancers working to increase drug
solubility in the stratum corneum and supersaturation. Pellett et al. (131) studied the
Figure 22 Transport of estradiol across human epidermal membrane. Up to an approxi-

mately 12-fold supersaturation, flux across epidermal membranes is linearly proportional to
the degree of saturation. (From Ref. 126.)
302 Davis et al.
Figure 23 Uptake (solubility) of estradiol into human epidermal membrane from three
vehicles (20, 30, and 40% propylene glycol; PG) at saturation and at 18-, 11-, and 4-fold
supersaturation, respectively. The theoretical uptake values for the supersaturated systems
were obtained by multiplying the uptake from the saturated system by the theoretical degree
of supersaturation. The figure shows the potential for synergy between enhancers working to
saturated solubility in the stratum corneum and supersaturation. (From Ref. 126.)
synergistic interaction between oleic acid, which is thought to enhance skin perme-
ation by increasing drug diffusivity in the stratum corneum, and supersaturation in
human skin in vitro. Penetration from saturated, oleic acid-pretreated skin, sixfold
supersaturated, and oleic acid-pretreated skin with sixfold supersaturated flurbiprofen
systems, is shown in Figure 24. The results demonstrate the potential for synergy
between enhancers working to increase drug diffusivity in the stratum corneum and
supersaturation.
e. Conclusions. Provided supersaturated solutions can be stabilized in the

donor phase, there is considerable evidence that, as predicted (8), they have the
potential to deliver a proportional increase in membrane transport. Given the mech-
anisms of increased transport (i.e., supersaturation of the membrane), it seems rea-
sonable to suggest that physical stability within the membrane is also important.
There is some reassuring evidence from the systematic studies reviewed here, that
gross crystallization does not occur within the stratum corneum, and that the stratum
corneum, in fact, may be a more superior membrane in this respect than Silastic
membrane. The viscosity of the stratum corneum lipids or natural antinucleant agents
may play a role. Recently, Pellett et al. (125) have demonstrated that application of
supersaturated solutions of piroxicam in vitro to human skin gave rise to linearly
proportionally higher drug levels in tape-stripped stratum corneum and viable epi-
dermal–dermal compartments. This is an encouraging result that supports the poten-
tial usefulness of supersaturated systems for the improvement of topical therapy. In
vivo studies in humans, using techniques such as microdialysis, are now required to
confirm these in vitro results. The studies of Megrab and Pellett et al. (126,131),
describing synergy between supersaturation and chemical enhancers, generated con-
Figure 24 Penetration across human stratum corneum from saturated oleic acid (OE)-
enhanced (pretreated skin), sixfold supersaturated, and OE with sixfold supersaturated flur-
biprofen systems: the potential for synergy between enhancers working to increase drug dif-
fusivity in the stratum corneum and supersaturation is shown. (From Ref. 131.)
fidence that future developments in this area will provide greatly increased skin
penetration and permeation and improve dermal and transdermal therapy.
C. Inadvertent Supersaturation During Manufacture

Potts and Cleary (132) described how the film-coating technology used in the pro-
duction of nonreservoir transdermal patches could lead to supersaturation of drug
within the drug-containing adhesive layer. In the manufacture of these patches the
drug and adhesive polymer are dissolved in a solvent that is subsequently evaporated
by heat (the process used to produce a solid dispersion system). As has been reviewed
here, this may have dramatic and beneficial effects on drug transport (and therapy),
at least initially. However, drug release and transport from transdermal patches man-
ufactured by solvent evaporation film-coating techniques may decrease with time if
the supersaturated drug recrystallizes (132–134). There is little experimental infor-
mation in this area. Kubota (135) compared skin permeation of betamethasone ben-
zoate from freshly made silicone adhesive patches, a suspension, and a saturated
solution. Permeation from the patch was approximately 2.5 times greater than either
304 Davis et al.
cream or saturated control, and the authors suggest supersaturation as one possible
mechanism.
Heat-premix processes are commonly used in the manufacture of gels, creams,
and ointments. Cooling and mixing a premix of drug in solution with miscible non-
solvents (e.g., water) can be expected to result in formation of supersaturated sys-
tems, at least initially. Henmi (136) described a heat process that gave rise to a
supersaturated indomethacin gel, which subsequently showed reduced release and
skin permeation with time.
Potentially large increases in skin penetration and possibly therapeutic effect
can occur from inadvertent supersaturation during manufacture. Of special concern
is that during the clinical development program, during which freshly made formu-
lations are used, high skin penetration (and efficacy) data may be generated, and this
may not be matched by the stored commercial product. This puts a great responsi-
bility on the use of stability and bioequivalence techniques to ensure quality and
efficacy on storage. It is important to consider this potential problem in dermatolog-
ical and transdermal development programs.
Finally, it is also feasible that supersaturated conditions may inadvertently occur
during actual therapeutic use of transdermal patch systems. Many transdermal sys-
tems are occlusive and will accumulate water as a result of physiological transepi-
dermal water loss. This may produce high activity states for the drug and thereby
enhance penetration into and permeation across the stratum corneum.
IV. VESICLES
A. Liposomes
Mezei first suggested that liposomes could be useful drug-carrier systems for the
local treatment of skin diseases (137,138). The suggestion was based on drug dis-
position data obtained following topical application of the steroid triaminolone ace-
tonide incorporated in phospholipid liposomes formulated as lotions or gels. Encap-
sulation of triaminolone acetonide into liposomes resulted in a vehicle-dependent
4.5- to 4.9-fold increase in the amount of drug recovered from the epidermis. Sim-
ilarly, topical application of liposomal preparations of hydrocortisone resulted in
higher concentrations of the steroid in skin strata, compared with application of the
drug in an ointment formulation (139,140). Continuing work by Mezei’s group gen-
erated further data for other dermatological drugs, such as econazole, minoxidil, and
retinoic acid (141–143). Overall, the work of Mezei suggested that application of
the dermatological drugs in liposomal form, when compared with the conventional
formulations, led to increased drug concentration in the skin and subcutaneous tissues
and decreased biodisposition in plasma and remote sites (Table 4).
These encouraging early observations were followed by several confirmatory
research and clinical investigations, most notably those of Weiner’s group at the
University of Michigan (144–148) and Korting’s group at Ludwig–Maximilians
University in Munich (149–151). Many other studies have indicated the potential of
phospholipid liposomes to increase the skin content of topically applied drugs. Toui-
tou et al. (152,153) evaluated the penetration of xanthines (dyphylline and caffeine)
in liposomal and nonliposomal formulations and found that delivery to the skin was
markedly influenced by the formulation in which the liposomes were applied. Thus,
Table 4 Effect of Liposomal Encapsulation on Drug Disposition Following

Topical Application
Triamcinolone Retinoic
Site acetonide Econazole Minoxidil acid
Epidermis 443a 872 465 261

Dermis 482 265 451 283
Subcutaneous tissue 72 206 1036 44
Plasma 47 50 55 56
Liver 93 63 76 124
Heart 89 97 85 91
a
Drug concentration as a percentage of control (the control value was obtained from non-
liposomal formulations).
Source: Refs. 137–141.
liposomes containing dyphylline were formulated into a polyethylene glycol oint-

ment, a carbomer gel, and a polyethylene glycol base containing oleic acid and
diglycol. The largest skin permeation was found for the latter formulation, and the
authors suggested that the slower permeation from the polyethylene glycol ointment
may be due to higher skin partitioning and retention of the drug. Enhanced skin
retention of caffeine was also observed following application in liposomal prepara-
tions (153). Phospholipid liposome encapsulation also enhances skin retention of the
topically applied local anesthetics lidocaine and tetracaine (154,155), cyclosporine
(156), interferon (157,158), and T4 endonuclease V (159). Reports on the latter two
compounds are particularly interesting, for they indicate the potential usefulness of
liposomes in localization of biologically active proteins and peptides within the skin.
Liposomes have also been prepared from lipid mixtures similar in composition
to the stratum corneum intercellular lipid (146,160–162). The ‘‘skin’’ lipids (fully
described in Chap. 1) usually comprise ceramides, derived from bovine brain; cho-
lesterol, cholesterol sulfate, and a fatty acid. When using the rationale that liposomes
formed from skin lipids would interact more favorably with the stratum corneum
intercellular lipid regions than phospholipid-based liposomes, Weiner’s group (146)
tested this hypothesis in the HSV-1 guinea pig model. They demonstrated that skin
lipid liposomes containing interferon were more effective at reducing lesion score
than were interferon-loaded phospholipid-based liposomes, and they generated in-
creased levels of the drug in the deeper skin layers (157). Fresta and Puglisi (161)
evaluated the effect of incorporation of hydrocortisone, triamcinolone, and beta-
methasone into phospholipid and skin lipid liposomes on in vitro human skin dis-
tribution of the steroid, and then compared this with conventional ointment formu-
lation. The data clearly showed that, although the formulation appeared to have no
effect on the stratum corneum content of the three steroids, liposomal formulation
increased the epidermal and dermal concentrations. Furthermore, skin lipid liposomes
increased the concentration of drug in deeper skin strata over that resulting from
phospholipid liposome encapsulation. The authors confirmed the in vitro data by
demonstrating in vivo a greater blanching effect (vasoconstriction) following appli-
cation of the steroid in skin lipid liposomes. Finally, the authors investigated the
systemic distribution of the steroids following topical application on guinea pigs.
306 Davis et al.
Quite clearly (Table 5) liposomal incorporation reduced systemic disposition when

compared with the ointment formulation, although there appeared to be little or no
difference between the liposomal formulations.
B. Niosomes
Niosomes are another vesicle system that has been investigated for potential modi-
fication of skin permeation (163,164). Niosomes comprise nonionic surfactants, such
as polyoxyethylene alkyl ethers, and may be prepared as single or multilamellar
vesicles. Surfactants of this type enhance skin permeation (165), and this is likely
to play a role in any modification of permeation using these vehicles. The effect of
nonionic surfactant vesicles on the skin permeation of estradiol was dependent on
the physical state of the niosome (164). Thus, whereas niosomes prepared from
polyoxyethylene(3)stearyl ether and existing in the gel state did not increase estradiol
permeation, those prepared from polyoxyethylene(3)lauryl ether and polyoxyeth-
ylene(10)oleyl ether—both existing as liquid crystalline vesicles—significantly en-
hanced transport. Further experiments, in which the skin was pretreated with un-
loaded niosomes indicated that the enhanced transport of estradiol from drug-loaded
vesicles was not wholly a result of surfactant-induced penetration enhancement. The
authors postulated that niosomes fused at the surface of the stratum corneum, gen-
erating high local concentrations of estradiol that resulted in increased thermody-
namic activity of the permeant in the upper layers of the stratum corneum.
C. Mode of Action
The precise mode of interaction between lipid vesicles and skin remains unclear.
Certainly, despite earlier claims (159), there is considerable doubt about the ability
of whole vesicles to permeate intact stratum corneum. Most evidence suggests that
vesicles can penetrate the outer cell layers of the stratum corneum (166–168) where
desmosomal linkages have become disrupted and, presumably, the keratinocytes are
less tightly bound and surrounded by a mixture of intercellular lipid and sebum.
However, continuing diffusion of vesicles through the approximately 60-nm inter-
cellular space of the deeper layers of the stratum corneum seems unlikely. Current
thinking suggests that lipid vesicles fuse with endogenous lipid, either on the surface
or in the outermost layers of the stratum corneum (168,169). The fusion is followed
Table 5 Effect of Liposomal Encapsulation on Triamcinolone

Disposition Following Topical Application
Phospholipid Skin lipid

Site Ointment liposome liposome
Skin 1.36a 1.75 1.87

Subcutaneous tissue 3.73 2.22 2.05
Blood 0.04 0.02 0.02
Liver 0.19 0.13 0.09
Heart 0.09 0.09 0.09
a
Drug concentration: mg/g tissue for skin, ␮g/g tissue for all other sites.
Source: Ref. 161.
by structural changes in the deeper layers of the stratum corneum, as evidenced by

freeze–fracture electron microscopy and small angle X-ray–scattering techniques.
These structural changes are presumed to be the result of intercellular diffusion of
vesicle lipid components (not intact vesicles) to the deeper layers, interaction with
and disruption of endogenous lipid lamellae. It is simple to postulate that this inter-
action or disruption of lipid lamellae will lead to an increase in skin permeation
rates, but this does not explain the observed increase in skin retention of permeants,
as described earlier. Apparent increased skin retention may be an artifact from ex-
ogenous lipid depot formation on the skin surface. On the other hand, formations of
lipid aggregates, possibly comprising mixtures of endogenous and exogenous lipid
(170,171), which have been observed in deeper layers of the stratum corneum (164),
may provide a reservoir for topically applied drugs.
D. Transfersomes
Despite the evidence to the contrary, Cevc (172) suggested that it was possible for
whole vesicles to cross intact stratum corneum. The basic premise for this hypothesis
was the driving force provided by the osmotic gradient between the outer and inner
layers of the stratum corneum and the development of specific mixes of lipids to
form modified liposomes, termed transfersomes. The requirement for the osmotic
gradient to be maintained suggests that transfersomes will not function in occlusive
conditions, and careful formulation is necessary. Because of their unique structure
(described as a mix of phosphatidylcholine, sodium cholate, and ethanol) (173),
transfersomes are reputed to be very flexible vesicles and capable of transporting
their contents through the tortuous intercellular route of the stratum corneum. Thus,
dermal application of local anesthetics in transfersomes was ‘‘nearly as swift in action
as an injection’’ (173), and the application of the corticosteroids, triamcinolone ace-
tonide, dexamethasone, and hydrocortisone, encapsulated in transfersomes, resulted
in more reliable site-specificity for the drug (174).
E. Follicular Delivery
Lieb et al. (175,176) proposed that liposomes may be useful for targeting drugs to
skin follicles for the treatment of diseases, such as acne and alopecia. Their initial
experiments, using the hamster ear pilosebaceous unit, demonstrated that carboxy-
fluorescein, incorporated into phospholipid liposomes, was more efficiently targeted
to follicles than when formulated as a simple aqueous solution, a propylene glycol
(5%) solution, or a sodium dodecyl sulfate (0.05%) solution (175). However, most
of the carboxyfluorescein was located in the epidermis. In later experiments, appli-
cation of cimetidine, incorporated in phospholipid and nonionic liposomes, was com-
pared with its application in a 50% alcohol solution (176), and generated data that
was similarly equivocal. In this case, although small amounts of drug were located
within the pilosebaceous unit, most was located on the surface or within the stratum
corneum (determined by tape-stripping). Nonetheless, the data showed that the li-
posomal formulations were considerably more effective at delivering the drug to the
stratum corneum and the follicles than was the alcoholic solution. Interestingly, the
phospholipid liposomes delivered approximately twice as much drug to these com-
partments as the nonionic liposomes.
308 Davis et al.
More recently, this group has demonstrated that nonionic liposomes were ca-
pable of delivering the macromolecules cyclosporine and interferon-alfa into pilo-
sebaceous units (177). Three nonionic and one phospholipid liposome were evalu-
ated. The most effective vesicle comprised glyceryl dilaurate (57% by weight),
cholesterol (12%), and polyoxyethylene(10)stearyl ether (28%), which facilitated the
follicular deposition of both hydrophilic and hydrophobic drugs. The authors spec-
ulated that the low-melting point of glyceryl dilaurate (30⬚C) may have resulted in
fluidization of the liposome bilayers following application of the formulation, and
this led to partial release of polyoxyethylene(10)stearyl ether (a known skin pene-
tration enhancer) (165), which further led to an enhancement of drug deposition.
Given these studies, the authors investigated the liposomal delivery of plasmid DNA
to the skin (178). Their results suggested that nonionic–cationic liposomes (com-
prising glyceryl dilaurate, cholesterol polyoxyethylene(10)stearyl ether, and 1,2-di-
oleoyloxy-3-(trimethylammonio)-propane) could deliver expression plasmid DNA to
perifollicular cells and mediate transient transfection in vivo. These data show prom-
ise for future topical gene therapy for a number of dermatological diseases caused
by abnormal regulation of soluble cytokines.
Other recent data on liposome-mediated drug delivery to follicular regions has
been contradictory. Whereas, Bernard et al. (179) demonstrated that phospholipid
liposomes were useful for targeting delivery of an antiandrogen to the sebaceous
gland, Tshan et al. (180) found no enhancement of follicular deposition of isotretinoin
when applied in liposomal formulation. Overall, the number and variety of constit-
uents from which lipid vesicles have been constructed, the large number of permeants
and formulations studied, and the diversity of experimental methods render it difficult
to establish any systematic ground rules. It is clear, however, that although there is
considerable evidence for improved drug delivery characteristics, much remains to
be achieved in this field.
In this chapter several strategies for enhancing skin penetration have been reviewed.
Quite clearly, there are many occasions when the ability to increase the rate and
extent of dermal and transdermal drug delivery would be beneficial to therapy. This
is reflected in the considerable research effort in this area. In recent years, much of
this effort has focused on the development of chemical penetration enhancers, the
use of supersaturated systems, the use of vesicular systems, and physical methods,
such as iontophoresis and ultrasound. Limitations of space have precluded a discus-
sion of physical methods of enhancement. However, further and more extensive
information on physical and chemical enhancement of skin penetration can be found
(9–11,181–183).
REFERENCES
1. Poulsen BJ, Young E, Coquilla V, Katz M. Effect of vehicle composition on the in
vitro release of fluocinolone acetonide and its acetate ester. J Pharm Sci 57:928–993,
1972.
2. Ostrenga J, Steinmetz C, Poulsen BJ. Significance of vehicle composition I: relationship
between topical vehicle composition, skin permeability and clinical efficacy. J Pharm
Sci 60:1175–1179, 1971.
3. Lippold BC, Schneemann H. The influence of vehicles on the local bioavailability of

betamethasone-17-benzoate from solution and suspension-type ointments. Int J Pharm
22:31–43, 1984.
4. Guy RH. Current status and future prospects of transdermal drug delivery. Pharm Res
13:1765–1769, 1996.
5. Vaughan CD. Using solubility parameters in cosmetics formulations. J Soc Cosmet
Chem 36:319–333, 1985.
6. Sloan KB, Koch SAM, Siver KG, Flowers FP. Use of solubility parameters of drug
and vehicle to predict flux through skin. J Invest Dermatol 87:244–252, 1986.
7. Cooper ER, Merritt BW, Smith RL. Effect of fatty acids and alcohols on the penetration
of acyclovir across human skin in vitro. J Pharm Sci 74:688–689, 1985.
8. Higuchi T. Physical chemical analysis of percutaneous absorption process from creams
9. Walters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration Enhancement. New
York: Marcel Dekker, 1993.
10. Smith EW, Maibach HI, eds. Percutaneous Penetration Enhancers. Boca Raton: CRC
Press, 1995.
11. Prausnitz MR. The effects of electric current applied to skin: a review for transdermal
drug delivery. Adv Drug Deliv Rev 18:395–425, 1996.
12. Beastall JC, Hadgraft J, Washington C. Mechanism of action of Azone as a percuta-
neous penetration enhancer—lipid bilayer fluidity and transition–temperature effects.
Int J Pharm 43:207–213, 1988.
13. Rolland A, Brzokewicz A, Shroot B, Jamoulle JC. Effect of penetration enhancers on
the phase-transition of multilamellar liposomes of dipalmitolyphosphatidylcholine—a
study by differential scanning calorimetry. Int J Pharm 76:217–224, 1991.
14. Lewis D, Hadgraft J. Mixed monolayers of dipalmitoylphosphatidylcholine with Azone
or oleic-acid at the air–water interface. Int J Pharm 65:211–218, 1990.
15. Hoogstraate AJ, Verhoef J, Brussee J, IJzerman AP, Spies F, Boddé HE. Kinetics,
ultrastructural aspects and molecular modeling of transdermal peptide flux enhancement
by N-alkylazacycloheptanones. Int J Pharm 76:37–47, 1991.
16. Schuckler F, Lee G. The influence of Azone on monomolecular films of some stratum
corneum lipids. Int J Pharm 70:173–186, 1991.
17. Schuckler F, Bouwstra JA, Gooris GS, Lee G. An X-ray–diffraction study of some
model stratum corneum lipids containing Azone and dodecyl-L-pyroglutamate. J Con-
18. Engblom J, Engstrom S, Fontell K. The effect of the skin penetration enhancer Azone
on fatty acid–sodium soap–water mixtures. J Controlled Release 33:299–305, 1995.
19. Engblom J, Engstrom S. Azone and the formation of reversed monocontinuous and
bicontinuous lipid–water phases. Int J Pharm 98:173–179, 1993.
20. Hadgraft J, Peck J, Williams DG, Pugh WJ, Allan G. Mechanisms of action of skin
penetration enhancers and retarders—Azone and analogs. Int J Pharm 141:17–25,
1996.
21. Hadgraft J, Walters KA, Wotton PK. Facilitated transport of sodium-salicylate across
an artificial lipid-membrane by Azone. J Pharm Pharmacol 37:725–727, 1985.
22. Irwin WJ, Smith JC. Extraction coefficients and facilitated transport—the effect of
absorption enhancers. Int J Pharm 76:151–159, 1991.
23. Fuhrman LC, Michniak BB, Behl CR, Malick AW. Effect of novel penetration en-
hancers on the transdermal delivery of hydrocortisone: an in vitro species comparison.
J Controlled Release 45:199–206, 1997.
24. Bhatt PP, Rytting JH, Topp EM. Influence of Azone and lauryl alcohol on the transport
of acetaminophen and ibuprofen through shed snake skin. Int J Pharm 72:219–226,
1991.
310 Davis et al.
25. Hirvonen J, Rytting JH, Paronen P, Urtti A. Dodecyl N,N-dimethylamino acetate and
Azone enhance drug penetration across human, snake, and rabbit skin. Pharm Res 8:
933–937, 1991.
26. Itoh T, Wasinger L, Turunen TM, Rytting JH. Effects of transdermal penetration en-
hancers on the permeability of shed snakeskin. Pharm Res 9:1168–1172, 1992.
27. Niazy EM. Differences in penetration-enhancing effect of Azone through excised rab-
bit, rat, hairless mouse, guinea-pig and human skins. Int J Pharm 130:225–230, 1996.
28. Okamoto H, Hashida M, Sezaki H. Effect of 1-alkyl- or 1-alkenylazacycloalkanone
derivatives on the penetration of drugs with different lipophilicites through guinea pig
skin. J Pharm Sci 80:39–45, 1991.
29. Fincher TK, Yoo SD, Player MR, Sowell JW, Michniak BB. In-vitro evaluation of a
series of N-dodecanoyl-L-amino acid methyl-esters as dermal penetration enhancers. J
Pharm Sci 85:920–923, 1996.
30. Godwin DA, Michniak BB, Player MR, Sowell JW. Transdermal and dermal enhancing
activity of pyrrolidinones in hairless mouse skin. Int J Pharm 155:241–250, 1997.
31. Michniak BB, Chapman JM, Seyda KL. Facilitated transport of 2 model steroids by
esters and amides of clofibric acid. J Pharm Sci 82:214–219, 1993.
32. Michniak BB, Player MR, Chapman JM, Sowell JW. In vitro evaluation of a series of
Azone analogs as dermal penetration enhancers 1. Int J Pharm 91:85–93, 1993.
33. Michniak BB, Player MR, Fuhrman LC, Christensen CA, Chapman JM, Sowell JW.
In-vitro evaluation of a series of Azone analogs as dermal penetration enhancers 2.
(Thio)amides. Int J Pharm 94:203–210, 1993.
34. Michniak BB, Player MR, Chapman JM, Sowell JW. Azone analogs as penetration
enhancers—effect of different vehicles on hydrocortisone acetate skin permeation and
retention. J Controlled Release 32:147–154, 1994.
35. Michniak BB, Player MR, Fuhrman LC, Christensen CA, Chapman JM, Sowell JW.
In-vitro evaluation of a series of Azone analogs as dermal penetration enhancers 3.
Acyclic amides. Int J Pharm 110:231–239, 1994.
36. Michniak BB, Player MR, Godwin DA, Phillips CA, Sowell JW. In-vitro evaluation
of a series of Azone analogs as dermal penetration enhancers 4. Amines. Int J Pharm
116:201–209, 1995.
37. Phillips CA, Michniak BB. Topical application of Azone analogs to hairless mouse
skin—a histopathological study. Int J Pharm 125:63–71, 1995.
38. Phillips CA, Michniak BB. Transdermal delivery of drugs with differing lipophilicities
using Azone analogs as dermal penetration enhancers. J Pharm Sci 84:1427–1433,
1995.
39. Michniak BB, Player MR, Sowell JW. Synthesis and in-vitro transdermal penetration
enhancing activity of lactam N-acetic acid-esters. J Pharm Sci 85:150–154, 1996.
40. Bezema FR, Marttin E, Roemele PEH, Brussee J, Bodde HE, Degroot HJM. H-2 NMR
evidence for dynamic disorder in human skin induced by the penetration enhancer
Azone. Spectrochim Acta [Pt A] Mol Biomol Spectr 52:785–791, 1996.
41. Bouwstra JA, Peschier LJC, Brussee J, Bodde HE. Effect of N-alkyl-azocycloheptan-
2-ones including Azone on the thermal-behavior of human stratum corneum. Int J
Pharm 52:47–54, 1989.
42. Bouwstra JA, Devries MA, Gooris GS, Bras W, Brussee J, Ponec M. Thermodynamic
and structural aspects of the skin barrier. J Controlled Release 15:209–220, 1991.
43. Bouwstra JA, Gooris GS, Brussee J, Salomonsdevries MA, Bras W. The influence of
alkyl-azones on the ordering of the lamellae in human stratum corneum. Int J Pharm
79:141–148, 1992.
44. Quan DY, Maibach HI. An electron-spin-resonance study. I. Effect of Azone on 5-
doxyl stearic acid-labeled human stratum corneum. Int J Pharm 104:61–72, 1994.
45. Quan DY, Cooke RA, Maibach HI. An electron-spin-resonance study of human epi-
dermal lipids using 5-doxyl stearic acid. J Controlled Release 36:235–241, 1995.
46. Gay CL, Murphy TM, Hadgraft J, Kellaway IW, Evans JC, Rowlands CC. An electron-
spin resonance study of skin penetration enhancers. Int J Pharm 49:39–45, 1989.
47. Groundwater PW, Hadgraft J, Harrison JE, Watkinson AC. The synthesis of perdeu-
terated Azone (D(35)-1-dodecylhexahydro-2h-azepin-2-one). J Label Comp Radio-
pharm 34:1047–1053, 1994.
48. Harrison JE, Groundwater PW, Brain KR, Hadgraft J. Azone induced fluidity in human
stratum corneum—a Fourier–transform infrared spectroscopy investigation using the
perdeuterated analog. J Controlled Release 41:283–290, 1996.
49. Ongpipattanakul B, Burnette RR, Potts RO, Francoeur ML. Evidence that oleic acid
exists in a separate phase within stratum corneum lipids. Pharm Res 8:350–354, 1991.
50. Bonina FP, Montenegro L. Penetration enhancer effects on in vitro percutaneous ab-
sorption of heparin sodium salt. Int J Pharm 82:171–177, 1992.
51. Harrison JE, Watkinson AC, Green DM, Hadgraft J, Brain KR. The relative effect of
Azone and Transcutol on permeant diffusivity and solubility in human stratum cor-
neum. Pharm Res 13:542–546, 1996.
52. Diez–Sales O, Watkinson AC, Herraez–Dominguez M, Javaloyes C, Hadgraft J. A
mechanistic investigation of the in vitro human skin permeation enhancing effect of
Azone. Int J Pharm 129:33–40, 1996.
53. Brain KR, Hadgraft J, Lewis D, Allan G. The influence of Azone on the percutaneous
absorption of methotrexate. Int J Pharm 71:R9–R11, 1991.
54. Patil S, Szolarplatzer C, Maibach HI. Effect of topical laurocapram (Azone) on the in
vitro percutaneous permeation of sodium lauryl sulfate using human skin. J Invest
Dermatol 106:947, 1996.
55. Szolarplatzer C, Patil S, Maibach HI. Effect of topical laurocapram (Azone) on the in
vitro percutaneous permeation of sodium lauryl sulfate using human skin. Acta Derm
Venereol 76:182–185, 1996.
56. Hadgraft J, Williams DG, Allan G. Azone mechanisms of action and clinical effect. In:
Walters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration Enhancement. New York:
57. Chatterjee DJ, Li WY, Koda RT. Effect of vehicles and penetration enhancers on the
in vitro and in vivo percutaneous absorption of methotrexate and edatrexate through
hairless mouse skin. Pharm Res 14:1058–1065, 1997.
58. Gonsho A, Imanidis G, Vogt P, Kern ER, Tsuge H, Su MH, Choi SH, Higuchi WI.
Controlled (trans)dermal delivery of an antiviral agent (acyclovir) 1. An in vivo animal
model for efficacy evaluation in cutaneous HSV-1 infections. Int J Pharm 65:183–194,
1990.
59. Higo N, Hinz RS, Lau DTW, Benet LZ, Guy RH. Cutaneous metabolism of nitro-
glycerin in vitro 2. Effects of skin condition and penetration enhancement. Pharm Res
9:303–306, 1992.
60. Hwang GC, Lin AY, Chen W, Sharpe RJ. Development and optimization of a metho-
trexate topical formulation. Drug Dev Ind Pharm 21:1941–1952, 1995.
61. Ogiso T, Iwaki M, Tsuji S. Percutaneous absorption of bromhexine in rats. Chem Pharm
Bull 39:1609–1611, 1991.
62. Swart PJ, Weide WL, Dezeeuw RA. In vitro penetration of the dopamine-D2 agonist
N-0923 with and without Azone. Int J Pharm 87:67–72, 1992.
63. Ogiso T, Iwaki M, Tanino T, Oue H. Percutaneous absorption of pindolol and phar-
macokinetic analysis of the plasma concentration. J Pharmacobiodyn 15:347–352,
1992.
64. Ruan LP, Liang BW, Tao JZ, Yin CH. Transdermal absorption of nitrendipine from
adhesive patches. J Controlled Release 20:231–236, 1992.
312 Davis et al.
65. Wiechers JW, Drenth BFH, Jonkman JHG, Dezeeuw RA. Percutaneous absorption of
triamcinolone acetonide from creams with and without Azone in humans In vivo. Int
J Pharm 66:53–62, 1990.
66. Wiechers JW. Absorption, distribution, metabolism, and excretion of the cutaneous
penetration enhancer Azone. Pharm Weekblad Sci Ed 12:116–118, 1990.
67. Wiechers JW, Drenth BFH, Adolfsen FAW, Prins L, Dezeeuw RA. Disposition and
metabolic profiling of the penetration enhancer Azone 1. In vivo studies—urinary pro-
files of hamster, rat, monkey, and man. Pharm Res 7:496–499, 1990.
68. Wiechers JW, Drenth BFH, Jonkman JHG, Dezeeuw RA. Percutaneous absorption and
elimination of the penetration enhancer Azone in humans. Pharm Res 4:519–523, 1987.
69. Wiechers JW, Drenth BFH, Jonkman JHG, Dezeeuw RA. Percutaneous absorption,
metabolism, and elimination of the penetration enhancer Azone in humans after pro-
longed application under occlusion. Int J Pharm 47:43–49, 1988.
70. Wiechers JW, Drenth BFH, Jonkman JHG, Dezeeuw RA. Percutaneous absorption,
metabolic profiling, and excretion of the penetration enhancer Azone after multiple
dosing of an Azone-containing triamcinolone acetonide cream in humans. J Pharm Sci
79:111–115, 1990.
71. Wotton PK, Mollgaard B, Hadgraft J, Hoelgaard A. Vehicle effect on topical drug
delivery 3. Effect of Azone on the cutaneous permeation of metronidazole and pro-
pylene glycol. Int J Pharm 24:19–26, 1985.
72. Watkinson AC, Hadgraft J, Bye A. Aspects of the transdermal delivery of prostaglan-
dins. Int J Pharm 74:229–236, 1991.
73. Ganga S, Ramarao P, Singh J. Effect of Azone on the iontophoretic transdermal deliv-
ery of metoprolol tartrate through human epidermis in vitro. J Controlled Release 42:
57–64, 1996.
74. Hirvonen J, Kontturi K, Murtomaki L, Paronen P, Urtti A. Transdermal iontophoresis
of sotalol and salicylate—the effect of skin charge and penetration enhancers. J Con-
75. Kalia YN, Guy RH. Interaction between penetration enhancers and iontophoresis: effect
on human skin impedance in vivo. J Controlled Release 44:33–42, 1997.
76. Ueda H, Isshiki R, Ogihara M, Sugibayashi K, Morimoto Y. Combined effect of ultra-
sound and chemical enhancers on the skin permeation of aminopyrine. Int J Pharm
143:37–45, 1996.
77. Samour CM, Daskalakis S. U.S. Patent 4,861,764, 1989.
78. Gyurik RJ, et al. SEPA penetration enhancement of econazole in human skin. In: Brain
KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 4b. Car-
diff: STS Publishing, pp 124–126, 1996.
79. Gauthier ER, Gyurik RJ, Krauser SF, Pittz EP, Samour CM. SEPA absorption enhance-
ment of polar and non-polar drugs. In: Brain KR, James VJ, Walters KA, eds. Per-
spectives in Percutaneous Penetration. Vol. 5a. Cardiff: STS Publishing, p 79, 1997.
80. Yoneto K, Ghanem A–H, Higuchi WI, Peck KD, Li SK. Mechanistic studies of the 1-
alkyl-2-pyrroloidones as skin permeation enhancers. J Pharm Sci 84:312–317, 1995.
81. Cornwell PA, Barry BW, Bouwstra JA, and Gooris GS. Modes of action of terpene
penetration enhancers in human skin; differential scanning calorimetry, small-angle X-
ray diffraction and enhancer uptake studies. Int J Pharm 127:9–26, 1996.
82. van Hal DA, Jeremiasse E, Junginger HE, Spies F, Bouwstra JA. Structure of fully
hydrated human stratum corneum: a freeze–fracture electron microscope study. J Invest
Dermatol 106:89–95, 1996.
83. Engström S, Forslind B, Engblom J. Lipid polymorphism—a key to an understanding
of skin penetration. In: Brain KR, James VJ, Walters KA, eds. Prediction of Percuta-
neous Penetration. Vol. 4b. Cardiff: STS Publishing, pp 163–166, 1996.
84. Strathman H. Production of microporous media by phase inversion processes. In: Lloyd
DR, ed. Materials Science of Synthetic Membranes. ACS Symp Ser 269:165–195,
1996.
85. Langlois B, Friberg S. Evaporation from a complex emulsion system. J Soc Cosmet
Chem 44:23–34, 1993.
86. Yano T, Higo N, Furukawa K, Tsuji M, Noda K, Otagiri M. Evaluation of a new
penetration enhancer 1-[2-(decylthio)ethyl]azacyclopentan-2-one (HPE-101). J Phar-
macobiodyn 15:527–533, 1992.
87. Yano T, Higo N, Fukuda K, Tsuji M, Noda K, Otagiri M. Further evaluation of a new
penetration enhancer, HPE-101. J Pharm Pharmacol 45:775–778, 1992.
88. Adachi H, Irie T, Uekama K, Manako T, Yano T, Saita M. Combination effects of O-
carboxymethyl-O-ethyl-␤-cyclodextrin and penetration enhancer HPE-101 on transder-
mal delivery of prostaglandin E1 in hairless mice. Eur J Pharm Sci 1:117–123, 1993.
89. Rajadhyaksha VJ. Oxalodinone penetration enhancing compounds. U.S. Patent
4,960,771, 1990.
90. Pfister WR, Rajadhyaksha VJ. Oxazolidinones: a new class of cyclic urethane trans-
dermal enhancer (CUTE). Proc Int Symp Controlled Release Bioact Mater 24:709–
710, 1997.
91. Pfister WR, Rajadhyaksha VJ. Oxazolidinones: a new class of cyclic urethane trans-
dermal enhancer (CUTE). Pharm Res 12:S280, 1995.
92. Wong O, Huntington A, Nishihata T, Rytting JH. New alkyl N,N-dialkyl-substituted
amino acetates as transdermal penetration enhancers. Pharm Res 6:286–295, 1989.
93. Wongpayapkul L, Chow D. Comparative evaluation of various enhancers on the trans-
dermal permeation of propranolol hydrochloride. Pharm Res 8:S140, 1991.
94. Büyüktimkin S, Büyüktimkin N, Rytting JH. Synthesis and enhancing effect of dodecyl
2-(N,N-dimethylamino)-propionate (DDAIP) on the transepidermal delivery of indo-
methacin, clonidine, and hydrocortisone. Pharm Res 10:1632–1637, 1993.
95. Turunen TM, Büyüktimkin S, Büyüktimkin N, Urtti A, Paronen P, Rytting JH. En-
hanced delivery of 5-fluorouracil through shed snake skin by two new transdermal
penetration enhancers. Int J Pharm 92:89–95, 1993.
96. Turunen TM, Urtti A, Paronen P, Audus KL, Rytting JH. Effect of some penetration
enhancers on epithelial membrane lipid domains: evidence from fluorescence spectros-
copy studies. Pharm Res 11:288–294, 1994.
97. Lambert WJ, Kudla RJ, Holland JM, Curry JT. A biodegradable transdermal penetration
enhancer based on N-(2-hydroxyethyl)-2-pyrrolidone I. Synthesis and characterization.
Int J Pharm 95:181–192, 1993.
98. Harris WT, Tenjarla SN, Holbrook JM, Smith J, Mead C, Entrekin J. N-Pentyl N-
acetylprolinate. A new skin penetration enhancer. J Pharm Sci 84:640–642, 1995.
99. James KC. Solubility and Related Properties. New York: Marcel Dekker, 1986.
100. Frank FC. The influence of dislocations on crystal growth. Discuss Faraday Soc 5:48–
54, 1949.
101. Kondo S, Yamanaka C, Sugimoto I. Enhancement of transdermal delivery by superflu-
ous thermodynamic potential III. Percutaneous absorption of nifedipine in rats. J Phar-
macobiodyn 10:743–749, 1987.
102. Ikeda K, Saitoh I, Oguma T, Takagishi Y. Effect of hydrophobically modified hydroxy-
propylmethyl cellulose on recrystallization from supersaturated solutions of indometh-
acin. Chem Pharm Bull 42:2320–2326, 1994.
103. Hasagawa A, Taguchi M, Suzuki R, Miyata T, Nakagawa H, Sugimoto I. Supersatu-
ration mechanisms of drugs from solid dispersions with enteric coating agents. Chem
Pharm Bull 36:4941–4950, 1988.
104. Pearson A. Inhibition of a growth in a model pharmaceutical semisolid. PhD disser-
tation, Strathclyde University, England No BRD-97084.
314 Davis et al.
105. Yeoh TY. Use of polymer additives to inhibit phenytoin nucleation and crystal growth
from supersaturated aqueous solutions. PhD dissertation, Purdue University, West La-
fayette, IN, No. DA9513096.
106. Morita M, Horita S. Effect of crystallinity on the percutaneous absorption of cortico-
steroids II. Chemical activity and biological activity. Chem Pharm Bull 33:2091–2097,
1985.
107. Szeman J, Ueda H, Szejtli J, Fenyvesi W, Watanabe Y, Machida Y, Nagai T. Enhanced
percutaneous absorption of homogenised tolnaftate/␤ –cyclodextrin polymer ground
mixture. Drug Design Deliv 1:325–332, 1987.
108. Lippold BC. How to optimize drug penetration through the skin. Pharm Acta Helv 67:
294–300, 1992.
109. Davis AF, Hadgraft J. Supersaturated solutions as topical drug delivery systems. In:
110. Fischer W. Trends in transdermal drug delivery. In: Junginger HE, ed. Pharmaceutical
Technology: Drug Targeting and Delivery. London: Ellis Horwood, pp 190–202, 1992.
111. Zatz JL. Enhancing skin penetration of actives with the vehicle. Cosmet Toiletr 109:
27–36, 1994.
112. Pefile S, Smith EW. Transdermal drug delivery vehicle design and formulation. S Afr
J Sci 93:147–150, 1997.
113. Chiou WL, Riegelmann S. Pharmaceutical applications of solid dispersion systems. J
Pharm Sci 60:1281–1301, 1971.
114. Hasegawa A, Nakagawa H, Sugimoto I. Application of solid dispersions of spray–dried
nifedipine with enteric-coating agents to prepare a sustained-release dosage form. Chem
Pharm Bull 33:1615–1619, 1985.
115. Hasegawa A, Kawamura H, Nakagawa H, Sugimoto I. Physical properties of solid
dispersions of poorly water soluble drugs with enteric coating agents. Chem Pharm
Bull 33:3429–3435, 1985.
116. Corrigan OI, Farvar MA, Higuchi WI. Drug membrane transport enhancement using
high energy drug polvinylpyrrolidone (PVP) coprecipitates. Int J Pharm 5:229–238,
1980.
117. Merkle HP. Drug polvinylpyrrolidone co-precipitates: kinetics of drug release and for-
mation of supersaturated solutions. In: Digenis GA, Ansell J, eds. Proceedings Inter-
national Symposium Povidone. Lexington: University of Kentucky, pp 202–216, 1983.
118. Norman FH. In: Roche EB, ed. Design of Biopharmaceutical Properties Through Pro-
drugs and Analogs, Alpha Series. pp 198, 1997.
119. Corrigan OI. Drug–polvinylpyrrolidone coprecipitates. In: Smith EW, Maibach HI, eds.
Percutaneous Penetration Enhancers. Boca Raton: CRC Press, pp 221–232, 1995.
120. Merkle HP. Transdermal delivery systems. Methods Findings Exp Clin Pharmacol 11:
135–153, 1989.
121. Dittgen M, Bombor R. Process for the production of a pharmaceutical preparation.
German patent DD 217 989-A1, 1985.
122. Davis AF, Hadgraft J. Effect of supersaturation on membrane transport: 1. Hydrocor-
tisone acetate. Int J Pharm 76:1–8, 1991.
123. Pellet MA, Davis AF, Hadgraft J. Effect of supersaturation on membrane transport: 2.
Piroxicam. Int J Pharm 111:1–6, 1994.
124. Pellet MA, Castellano S, Hadgraft J, Davis AF. The penetration of supersaturated so-
lutions of piroxicam across silicone membrane and human skin in vitro. J Controlled
Release 46:205–214, 1997.
125. Pellet MA, Roberts MS, Hadgraft J. Supersaturated solutions evaluated with an in vivo
stratum corneum tape stripping technique. Int J Pharm 151:91–98, 1997.
126. Megrab NA, Williams AC, Barry BW. Oestradiol permeation through human skin and
Silastic membrane: effects of propylene glycol and supersaturation. J Controlled Re-
lease 36:277–294, 1995.
127. Higuchi WI, Farrar MA. Drug membrane transport enhancement using high energy
drug povidone coprecipitates. In: Digenis GA, Ansell J, eds. Proceedings International
Symposium Povidone, Lexington: University of Kentucky, pp 71, 1983.
128. Flynn GL, Smith RW. Membrane diffusion III. Influence of solvent composition and
permeant solubility on membrane transport. J Pharm Sci 61:61–66, 1972.
129. Tanaka S, Takanashima Y, Murayama H, Tsuchiya S. Studies on drug release from
ointments: V. Release of hydrocortisone butyrate from topical dosage forms to silicone
rubber. Int J Pharm 27:29–38, 1985.
130. Pellett MA. Unpublished observations.
131. Pellett MA, Watkinson AC, Brain KR, Hadgraft J. Diffusion of flurbiprofen across
human stratum corneum using synergistic methods of enhancement. In: Brain KR,
James VJ, Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 5a. Cardiff:
STS Publishing, p 86, 1997.
132. Potts RO, Cleary GW. Transdermal delivery: useful paradigms. J Drug Targeting 3:
247–251, 1995.
133. Ma X, Taw J, Chiang C–M. Inhibition of crystallization of steroid drugs in transdermal
patches. Proc Int Symp Controlled Release Bioact Mater 22:712–713, 1995.
134. Ma X, Taw J, Chiang C–M. Control of drug crystallization in transdermal matrix
system. Int J Pharm 143:115–119, 1996.
135. Kubota K, Sznitowska M, Maibach H. Percutaneous permeation of betamethasone 17-
valerate from different vehicles. Int J Pharm 96:105–110, 1993.
136. Henmi T, Fujii M, Kikuchi K, Matsumoto M. Application of an oily gel formed by
hydrogenated soybean phospholipids as a percutaneous absorption-type enhancer.
Chem Pharm Bull 42:651–655, 1994.
137. Mezei M, Gulasekharam V. Liposomes—a selective drug delivery system for the top-
ical route of administration: lotion dosage forms. Life Sci 26:1473–1477, 1980.
138. Mezei M, Gulasekharam V. Liposomes—a selective drug delivery system for the top-
ical route of administration: a gel dosage form. J Pharm Pharmacol 34:473–474, 1982.
139. Wohlrab W, Lasch J. Penetration kinetics of liposomal hydrocortisone in human skin.
Dermatologica 174:18–22, 1987.
140. Kim MK, Chung SJ, Lee MH, Cho AR, Shim CK. Targeted and sustained delivery of
hydrocortisone to normal and stratum corneum-removed skin without enhanced skin
absorption using a liposome gel. J Controlled Release 46:243–251, 1997.
141. Mezei M. Liposomes as a skin delivery system. In: Breimer DD, Speiser P, eds. Topics
in Pharmaceutical Sciences. Amsterdam: Elsevier, pp 345–358, 1985.
142. Mezei M. Liposomal drug delivery system for the topical route of administration. In:
Tipnis HP, ed. Controlled Release Dosage Forms. Bombay: Bombay College of Phar-
macy, pp 47–57, 1988.
143. Foong WC, Harsanyi BB, Mezei M. Biodisposition and histological evaluation of top-
ically applied retinoic acid in liposomal, cream and gel dosage form. In: Hanin I, Pepeu
G, eds. Phospholipids—Biochemical, Pharmaceutical and Analytical Considerations.
New York: Plenum, pp 279–282, 1990.
144. Ganesan MG, Weiner ND, Flynn GL, Ho NFH. Influence of liposomal drug entrapment
on percutaneous absorption. Int J Pharm 20:139–154, 1984.
145. Ho NFH, Ganesan MG, Weiner ND, Flynn GL. Mechanisms of topical delivery of
liposomally entrapped drugs. J Controlled Release 2:61–65, 1985.
146. Egbaria K, Weiner N. Liposomes as a topical drug delivery system. Adv Drug Deliv
Rev 5:287–300, 1990.
316 Davis et al.
147. Egbaria K, Weiner N. Topical delivery of liposomally encapsulated ingredients eval-

uated by in vitro diffusion studies. In: Braun–Falco O, Korting HC, Maibach HI, eds.
Liposome Dermatics. Berlin: Springer-Verlag, pp 172–181, 1992.
148. du Plessis J, Ramachandran C, Weiner N, Müller DG. The influence of particle size of
liposomes on the deposition of drug into skin. Int J Pharm 103:277–282, 1994.
149. Korting HC, Zienicke H, Schäfer–Korting M, Braun–Falco O. Liposome encapsulation
improves efficacy of betamethasone dipropionate in atopic eczema, but not in psoriasis
vulgaris. Eur J Clin Pharmacol 29:349–352, 1991.
150. Schäfer–Korting M, Korting HC, Ponce–Pöschl E. Liposomal tretinoin for uncompli-
cated acne vulgaris. Clin Invest 72:1086–1091, 1994.
151. Schmid MH, Korting HC. Therapeutic progress with topical liposome drugs for skin
disease. Adv Drug Deliv Rev 18:335–342, 1996.
152. Touitou E, Shaco–Ezra N, Dayan N, Jushynski M, Rafaeloff R, Azoury R. Dyphylline
liposomes for delivery to the skin. J Pharm Sci 81:131–134, 1992.
153. Touitou E, Levi–Schaffer F, Dayan N, Alhaique F, Riccieri F. Modulation of caffeine
skin delivery by carrier design: liposomes versus permeation enhancers. Int J Pharm
103:131–136, 1994.
154. Foldvari M, Gesztes A, Mezei M. Dermal drug delivery by liposome encapsulation:
clinical and electron microscopic studies. J Microencap 7:479–489, 1990.
155. Foldvari M. In vitro cutaneous and percutaneous delivery and in vivo efficacy of tetra-
caine from liposomal and conventional vehicles. Pharm Res 11:1593–1598, 1994.
156. Egbaria K, Ramachandran C, Weiner N. Topical delivery of cyclosporin: evaluation of
various formulations using in vitro diffusion studies in hairless mouse skin. Skin Phar-
macol 3:21–28, 1990.
157. Egbaria K, Ramachandran C, Kittayanond D, Weiner N. Topical delivery of liposomal
IFN. Antimicrob Agents Chemother 34:107–110, 1990.
158. Short SM, Rubas W, Paasch BD, Mrsny RJ. Transport of biologically active interferon-
gamma across human skin in vitro. Pharm Res 12:1140–1145, 1995.
159. Yarosh D, Bucana C, Cox P, Alas L, Kibitel J, Kripke M. Localization of liposomes
containing a DNA repair enzyme in murine skin. J Invest Dermatol 103:461–468, 1994.
160. William A, Wertz PW, Landmann L, Downing DT. Preparation of liposomes from
stratum corneum lipids. J Invest Dermatol 87:582–584, 1986.
161. Fresta M, Puglisi G. Corticosteroid dermal delivery with skin–lipid liposomes. J Con-
162. Gray GM, White RJ. Epidermal lipid liposomes: a novel nonphospholipid membrane
system. Biochem Soc Trans 7:1129–1131, 1979.
163. Hofland HEJ, van der Geest R, Boddé HE, Junginger HE, Bouwstra JA. Estradiol
permeation from nonionic surfactant vesicles through human stratum corneum in vitro.
Pharm Res 11:659–664, 1994.
164. Schreier H, Bouwstra J. Liposomes and niosomes as topical drug carriers: dermal and
transdermal drug delivery. J Controlled Release 30:1–15, 1994.
165. French EJ, Pouton CW, Walters KA. Mechanisms and prediction of nonionic surfactant
effects on skin permeability. In: Walters KA, Hadgraft J, eds. Pharmaceutical Skin
Penetration Enhancement. New York: Marcel Dekker, pp 113–143, 1993.
166. Lasch J, Laub R, Wohlrab W. How deep do intact liposomes penetrate into human
skin? J Controlled Release 18:55–58, 1991.
167. Schubert R, Joos M, Deicher M, Magerle R, Lasch J. Destabilization of egg lecithin
liposomes on the skin after topical application measured by perturbed ␥␥ angular cor-
relation spectroscopy (PAC) with 111ln. Biochim Biophys Acta 1150:162–164, 1993.
168. Hofland H, Bouwstra JA, Boddé HE, Spies F, Junginger HE. Interactions between
liposomes and human stratum corneum in vitro: freeze fracture electron microscopial
visualization and small angle X-ray scattering studies. Br J Dermatol 132:853–866,

1995.
169. Bouwstra JA, Hofland HEJ, Spies F, Gooris GS, Junginger HE. Changes in the structure
of the human stratum corneum induced by liposomes. In: Braun–Falco O, Korting HC,
Maibach HI, eds. Liposome Dermatics. Berlin: Springer-Verlag, pp 121–136, 1992.
170. Blume A, Jansen M, Ghyczy M, Gareiss J. Interation of phospholipid liposomes with
lipid model mixtures for stratum corneum. Int J Pharm 99:219–228, 1993.
171. Korting HC, Stolz W, Schmid MH, Maierhofer G. Interaction of liposomes with human
epidermis reconstructed in vitro. Br J Dermatol 132:571–579, 1995.
172. Cevc G, Blume G. Lipid vesicles penetrate into intact skin owing to the transdermal
osmotic gradients and hydration force. Biochim Biophys Acta 1104:226–232, 1991.
173. Planas ME, Gonzalez P, Rodriguez L, Sanchez S, Cevc G. Noninvasive percutaneous
induction of topical analgesia by a new type of drug carrier, and prolongation of local
pain insensitivity by anesthetic liposomes. Anesth Analg 75:615–621, 1992.
174. Cevc G, Blume G, Schätzlein. A. Transferosomes-mediated transepidermal delivery
improves the regio-specificity and biological activity of corticosteroids in vivo. J Con-
175. Lieb LM, Ramachandran C, Egbaria K, Weiner N. Topical delivery enhancement with
multilamellar liposomes into pilosebaceous units: I. In vitro evaluation using fluores-
cent techniques with the hamster ear model. J Invest Dermatol 99:108–113, 1992.
176. Lieb LM, Flynn G, Weiner N. Follicular (pilosebaceous unit) deposition and pharma-
cological behavior of cimetidine as a function of formulation. Pharm Res 11:1419–
1423, 1994.
177. Niemiec SM, Ramachandran C, Weiner N. Influence of nonionic liposomal composition
on topical delivery of peptide drugs into pilosebaceous units: an in vivo study using
the hamster ear model. Pharm Res 12:1184–1188, 1995.
178. Niemiec SM, Latta JM, Ramachandran C, Weiner N, Roessler BJ. Perifollicular trans-
genic expression of human interleukin-1 receptor antagonist protein following topical
application of novel liposome–plasmid DNA formulations in vivo. J Pharm Sci 86:
701–708, 1997.
179. Bernard E, Dubois J–L, Wepierre J. Importance of sebaceous glands in cutaneous
penetration of an antiandrogen: target effect of liposomes. J Pharm Sci 86:573–578,
1997.
180. Tshan T, Steffen H, Supersaxo A. Sebaceous gland deposition of isotretinoin after
topical application: an in vitro study using human facial skin. Skin Pharmacol 10:126–
134, 1997.
181. Hsieh DS, ed. Drug Permeation Enhancement—Theory and Applications. New York:
Marcel Dekker, 1994.
182. Braun–Falco O, Korting HC, Maibach HI, eds. Liposome Dermatics. Berlin: Springer-
Verlag, 1992.
183. Lai PM, Roberts MS. Iontophoresis. In: Roberts MS, Walters KA, eds. Dermal Ab-
7
Dermatological Formulation and
Transdermal Systems
KENNETH A. WALTERS and KEITH R. BRAIN

I. INTRODUCTION
Over the past few decades there have been many advances in our understanding of
the physicochemical properties of both formulation systems and their ingredients.
These have led to the ability to develop physically, chemically, and biologically
stable products. There has also been a significant increase in our knowledge of the
properties of skin and the processes that control skin permeation. The ground rules
for skin permeation were laid down by Scheuplein and Blank in the late 1960s and
early 1970s (1), and these have been updated on a reasonably regular basis (2–8).
We have learned, for example, that the permeation of compounds across intact skin
is controlled fundamentally by the stratum corneum, and it is the chemical compo-
sition and morphology of this layer that usually determines the rate and extent of
absorption (9,10). Similarly, we have discovered how to modify this barrier, by
chemical or physical means and, thereby, alter the rate of diffusion of many per-
meating molecules (11,12).
A basic deficiency, however, in the application of our understanding of the
barrier properties of the skin to dermatological and transdermal therapy is that this
knowledge has largely been generated by investigations on normal, rather than path-
ological, skin. The relevance of such information to diseased skin, for which per-
meation characteristics are probably significantly altered, has yet to be fully estab-
lished. There is some data on transport across skin that has been artificially damaged
(13–18), and limited information on permeation through diseased skin has been
obtained in the clinic (19–22).
In modern-day pharmaceutical practice, therapeutic compounds are applied to
the skin for dermatological (within the skin), local (regional), and for transdermal
319
320 Walters and Brain
(systemic) delivery. Whatever the target site or organ, it is usually a prerequisite that
the drug crosses the outermost layer of the skin, the stratum corneum. However, a
major function of the stratum corneum is to provide a protective barrier to the ingress
of xenobiotics and to control the rate of water loss from the body: Evolution has
generated a robust and durable barrier that fulfills its biological function throughout
an individual’s lifetime. A basic, yet thorough, understanding of the structure and
transport properties of this membrane is essential to the rational development of
topical dosage forms, and this has been provided elsewhere in this volume.
II. DRUG CANDIDATE SELECTION

Although it may appear to be a simple task to select lead compounds for pharma-
ceutical product development, based on therapeutic rationale and compound safety
and efficacy, the practicalities of this procedure are somewhat more complex. For
the most part, therapeutic efficacy is dependent on the ability of a compound to cross
biological barriers, travel to the target site, and interact with specific receptors. How-
ever, as pointed out and excellently reviewed (3), it is often more appropriate in
dermatological therapy to select compounds based on their inability to breach rele-
vant biological barriers. Because the site of action may be the skin surface, the
stratum corneum, the viable epidermis, the appendages, the dermis, or the local
subcutaneous tissues, the rules of candidate selection will vary. For the purposes of
this discussion it will be assumed that the therapeutic rationale for dermal drug
delivery has been established and that a series of compounds with appropriate phar-
macological activity have been identified. It will also be assumed that each com-
pound within the series possesses equivalent chemical and physical stability. In other
words, drug candidate selection need only be based on the ability to deliver the
compound to its site of action.
Earlier chapters in this volume have taught that the primary requirement for a
compound to penetrate the skin is the ability to leave the delivery system and enter
the stratum corneum. Furthermore, this characteristic is dependent on the stratum
corneum–vehicle partition coefficient of the compound, for which the octanol–water
partition coefficient is often used as a surrogate. Whereas it is immediately apparent
that a high value for this parameter will favor delivery into the stratum corneum, it
will not favor movement into the more hydrophilic regions of the viable epidermis.
Furthermore, the rate of diffusion through the stratum corneum and lower layers of
the skin is linked to the molecular volume of the permeant. It is evident that a
compound with a high octanol–water partition coefficient and a relatively high mo-
lecular volume will possess a high affinity for the stratum corneum (i.e., be substan-
tive to the stratum corneum). This principle is used extensively in the design of
sunscreen agents, for which it is not uncommon to add a medium-length or branched-
chain alkyl to the UV-absorbing molecule to increase residence time in the skin and
reduce systemic uptake.
The use of the skin as a route of delivery into the systemic circulation was
neither commercially nor scientifically exploited until the 1950s, when ointments
containing agents such as nitroglycerin and salicylates were developed. Angina could
be controlled for several hours by applying an ointment containing 2% nitroglycerin
(23). Similarly, topical salicylates could be absorbed through the skin into arthritic
joints. More recently, nonsteroidal anti-inflammatory agents, such as ibuprofen and
Drug Formulation and Transdermal Systems 321
ketoprofen, and hormonal steroids, such as estradiol and testosterone, have been
developed and marketed in semisolid preparations. A major problem with transdermal
semisolid preparations, however, is that of control. Drug concentrations in plasma
or duration of action, are not reliably predictable for several reasons, including the
amount and area of application and dosage frequency.
The specific advantages of transdermal therapy have been fully discussed else-
where (24). Briefly, transdermal devices are easy to apply, can remain in place for
up to 7 days (depending on the system), and are easily removed following, or during,
therapy. The reduced-dosing frequency, and the production of controllable and sus-
tained plasma levels, tend to minimize the risk of undesirable side effects sometimes
observed after oral delivery. The avoidance of extensive hepatic first-pass metabolism
is a further advantage. The major limitation to transdermal drug delivery is the
intrinsic barrier property of the skin. Although marketed patch-type transdermal de-
livery systems are available for only a limited number of drugs (e.g., scopolamine,
nitroglycerin, clonidine, estradiol, fentanyl, testosterone, and nicotine), several other
candidates are at various stages of development. Many of the drugs under investi-
gation do not intrinsically possess any significant ability to cross the skin; therefore,
ways must be found to improve their transdermal delivery. This could be achieved
by the use of prodrugs designed such that they are more rapidly absorbed than the
parent compound, yet are metabolized to the active species before receptor site oc-
cupancy (25,26). Physical methods, such as iontophoresis (27), electroporation (28),
and sonophoresis (29) have proved experimentally useful for increasing the skin
permeation of several compounds. Alternatively, the barrier may be modulated using
thermodynamic strategies, or chemically modified to reduce diffusive resistance by
the use of penetration enhancers, both of which are discussed in Chapter 6. Such
developmental strategies will increase the number of candidate drugs for transdermal
delivery in the future.
III. PREFORMULATION AND FORMULATION

Preformulation encompasses those studies that should be carried out before the com-
mencement of formulation development. The primary goal of the preformulation
process is to permit the rational development of stable, safe, efficacious dosage
forms, and it is concerned mainly with the characterization of the physicochemical
properties of the drug substance. At the preformulation stage, the final route of drug
administration is usually undecided; therefore, any protocols must be able to cover
all required aspects. The preformulation study has several distinct phases (Table 1).
A detailed description of all of the studies that form part of the preformulation stage
are given elsewhere (30) and will not be considered here. The only important aspect
of preformulation that is specific to dermatological and transdermal formulation con-
cerns drug delivery characteristics. These have been fully discussed elsewhere in this
volume (see Chapters 4 and 5).
A. Formulation of Dermatological Products

The selection of formulation type for dermatological products is usually influenced
by the nature of the skin lesion and the opinion of the medical practitioner. Kitson
and Maddin (31) elegantly stated: ‘‘It is idle to pretend that the therapy for skin
Table 1 Preformulation Tasks Listed in

Approximate Chronologic Order
Preformulation Task
1. General description of the compound

2. Calorimetry
3. Polymorphism
4. Hygroscopicity
5. Analytical development
6. Intrinsic stability
7. Solubility and partitioning characteristics
8. Drug delivery characteristics
diseases, as currently practiced, has its origins in science.’’ To this day a practicing
dermatologist would prefer to apply a ‘‘wet’’ formulation (ranging from simple tap-
water to complex emulsion formulations, with or without drug) to a wet lesion and
a ‘‘dry’’ formulation (e.g., petrolatum) to a dry lesion. The preparation of such for-
mulations as poultices and pastes is extemporaneous, and it is unlikely that the
industrial pharmaceutical formulator will be required to develop products of this type.
Solutions and powders lack staying power (retention time) on the skin and can afford
only transient relief. In modern-day pharmaceutical practice, semisolid formulations
are the preferred vehicles for dermatological therapy because they remain in situ and
deliver the drug over extended time periods. In most cases, therefore, the developed
formulation will be an ointment, emulsion, or gel. Typical constituents for these types
of formulations are shown in Table 2.
1. Ointments
An ointment is classified as any semisolid containing fatty material and intended for
external application (United States Pharmacopeia; USP). There are four types of
ointment base, and these are listed in the USP as hydrocarbon base, absorption base,
water-removable base, and water-soluble base. Only the hydrocarbon bases are com-
pletely anhydrous. The anhydrous hydrocarbon bases, which contain straight or
branched hydrocarbons with chain lengths ranging from C16 to C30, which may also
contain cyclic alkanes, are used principally in nonmedicated form. A typical for-
mulation contains fluid hydrocarbons (mineral oils and liquid paraffins) mixed with
a longer alkyl chain, higher-melting point, hydrocarbons (white and yellow soft par-
affin and petroleum jelly). The difference between white and yellow soft paraffin is
simply that the white version has been bleached. Hard paraffin and microcrystalline
waxes are similar to the soft paraffins, except that they contain no liquid components.
These anhydrous mixtures tend to produce formulations that are greasy and unplea-
sant to use, but the addition of solid components, such as microcrystalline cellulose,
can reduce the greasiness. Improved skin feel can also be attained by the incorpo-
ration of silicone materials, such as polydimethylsiloxane oil or dimethicones. Sili-
cones are often used in barrier formulations that are designed to protect the skin
against water-soluble irritants.
Although the nonmedicated anhydrous ointments are extremely useful as emol-
lients, their value as topical drug delivery systems is limited by the relative insolu-
Table 2 Constituents of Semisolid Formulations
Function Sample ingredients
Polymeric Gums Acrylic acids

thickeners Acacia Carbomers
Alginates Polycarbophil
Carageenan Colloidal solids
Chitosan Silica
Collagen Clays
Tragacanth Microcrystalline cellulose
Xanthan Hydrogels
Celluloses Polyvinyl alcohol
Sodium carboxymethyl Polyvinylpyrrolidone
Hydroxyethyl Thermoreversible polymers
Hydroxypropyl Poloxamers
Hydroxypropylmethyl
Oil phase Mineral oil Isopropyl myristate
White soft paraffin Isopropyl palmitate
Yellow soft paraffin Castor oil
Beeswax Canola oil
Stearyl alcohol Cottonseed oil
Cetyl alcohol Jojoba oil
Cetostearyl alcohol Arachis (Peanut) oil
Stearic acid Lanolin (and derivatives)
Oleic acid Silicone oils
Surfactants Nonionic Anionic
Sorbitan esters Sodium dodecyl sulfate
Polysorbates Cationic
Polyoxyethylene alkyl ethers Cetrimide
Polyoxyethylene alkyl esters Benzalkonium chloride
Polyoxyethylene aryl ethers
Glycerol esters
Cholesterol
Solvents Polar Polyethylene glycols
Water Propylene carbonate
Propylene glycol Triacetin
Glycerol Nonpolar
Sorbitol Isopropyl alcohol
Ethanol Medium-chain triglycerides
Industrial methylated spirit
Preservatives Antimicrobial Antioxidants
Benzalkonium chloride ␣-Tocopherol
Benzoic acid Ascorbic acid
Benzyl alcohol Ascorbyl palmitate
Bronopol Butylated hydroxyanisole
Chlorhexidine Butylated hydroxytoluene
Chlorocresol Sodium ascorbate
Imidazolidinyl urea Sodium metabisulfite
Paraben esters Chelating agents
Phenol Citric acid
Phenoxyethanol Edetic acid
Potassium sorbate
Sorbic acid
pH adjusters Diethanolamine Sodium hydroxide
Lactic acid Sodium phosphate
Monoethanolamine Triethanolamine
323
bility of many drugs in hydrocarbons and silicone oils. It is possible to increase drug
solubility within a formulation by incorporation of hydrocarbon-miscible solvents,
such as isopropylmyristate or propylene glycol, into the ointment. Although increas-
ing the solubility of a drug within a formulation may often decrease the release rate,
it does not necessarily decrease the therapeutic effect. It is well accepted that simple
determination of release rates from formulations may not be predictive of drug bio-
availability. For example, when formulated in a simple white petrolatum–mineral oil
ointment, the release rate of betamethasone dipropionate was considerably higher
than when the drug was formulated at the same concentration (0.05%) in an aug-
mented, and more clinically effective, ointment that contained propylene glycol (Fig.
1) (32). It is also important to appreciate that various grades of petrolatum are com-
mercially available, and that the physical properties of these materials will vary
depending on the source and refining process. Even slight variations in physical
properties of the constituents of an ointment may have substantial effects on drug
release behavior (33).
The preparation of ointment formulations may appear to be a simple matter of
heating all of the constituents to a temperature higher than the melting point of all
Figure 1 Release rates of betamethasone dipropionate from ointments into various receptor
fluids. (1) 5% hexane in acetonitrile; (2) octanol; (3) acetonitrile; (4) 60% acetonitrile in
water; (5) 95% ethanol. (From Ref. 32.)
of the excipients and then cooling with constant mixing. The reality, however, is that
the process is somewhat more complex and requires careful control over various
parameters, particularly the cooling rate. Rapid cooling, for example, creates stiffer
formulations in which there are numerous small crystallites, whereas a slow-cooling
rate results in the formation of fewer, but larger, crystallites and a more fluid product.
Further information on temperature effects and ointment phase behavior are available
(34–36).
2. Gels
The common characteristic of all gels is that they contain continuous structures that
provide solid-like properties (4). Depending on their constituents, gels may be clear
or opaque, and be polar, hydroalcoholic, or nonpolar. The simplest gels comprise
water thickened with either natural gums (e.g., tragacanth, guar, or xanthan), semi-
synthetic materials (e.g., methylcellulose, carboxymethylcellulose, or hydroxyethyl-
cellulose), synthetic materials (e.g., carbomer–carboxyvinyl polymer), or clays (e.g.,
silicates or hectorite). Gel viscosity is generally a function of the amount and mo-
lecular weight of the added thickener.
There are a variety of semisynthetic celluloses in use as thickeners in gel
formulations. These include methylcellulose (MC), carboxymethylcellulose (CMC),
hydroxyethylcellulose (HEC), hydroxypropylcellulose (HPC), and hydroxypropyl-
methylcellulose (HPMC). These celluloses can be obtained in diverse molecular
weight grades, and the higher molecular weight compounds are used at 1–5% (w/
w). In the development of prototype gel formulations, it is useful to evaluate a variety
of different types of cellulose. For example, if clarity of the gel is a major require-
ment, HPMC is preferable to MC. It is also important to appreciate that some cel-
luloses may exhibit specific incompatibilities with other potential formulation ingre-
dients. For example, HEC is incompatible with several salts, and MC and HPC are
incompatible with parabens. This latter incompatibility limits the choice of preser-
vative for gel formulations that are based on MC and HPC. Finally, the presence of
oxidative materials (e.g., peroxides, or other ingredients containing peroxide resi-
dues) in formulations gelled with celluloses should be avoided because oxidative
degradation of the polymer chains may cause a rapid decrease in formulation vis-
cosity (37).
As the branched-chain polysaccharide gums, such as tragacanth, pectin, car-
rageenan, and guar, are of naturally occurring plant origin, they can have widely
varying physical properties, depending on their source. They are usually incorporated
into formulations at concentrations between 0.5 and 10%, contingent on the required
viscosity. Viscosity may be enhanced synergistically by the addition of inorganic
suspending agents, such as magnesium aluminum silicate. Tragacanth, a mixture of
water-insoluble and water-soluble polysaccharides, is negatively charged in aqueous
solution and, therefore, incompatible with many preservatives when formulated at a
pH of 7 or higher. Similarly, xanthan gum, which is produced by bacterial fermen-
tation, is incompatible with some preservatives. Alginic acid is a hydrophilic colloidal
carbohydrate obtained from seaweed and the sodium salt, sodium alginate, is used
at 5–10% as a gelling agent. Film gels may be obtained by incorporation of small
amounts of soluble calcium salts (e.g., tartrate on citrate). Many gums are ineffective
in hydroalcoholic gels containing more than 5% alcohol.
The natural clay thickeners (e.g., bentonite and magnesium aluminium silicate)
are useful for thickening aqueous gels containing cosolvents, such as ethanol, iso-
propanol, glycerin, and propylene glycol. These materials possess a lamellar structure
that can be extensively hydrated. The flat surfaces of bentonite are negatively
charged, whereas the edges are positively charged. These clays swell in the presence
of water because of hydration of the cations and electrostatic repulsion between the
negatively charged faces. Thixotropic gels form at high concentrations at which the
clay particles combine in a flocculated structure in which the edge of one particle is
attracted to the face of another. The rheological properties of these clay dispersions,
therefore, are particularly sensitive to the presence of salts. Bentonite, a native col-
loidal hydrated aluminium silicate (mainly montmorillonite), can precipitate under
acidic conditions, and formulations must be at pH 6 or higher. A synthetic clay
(colloidal silicon dioxide) is also useful for thickening both aqueous and nonpolar
gels. The concentration of clay usually required to thicken formulations is 2–10%.
By far the most extensively employed gelling agents in the pharmaceutical and
cosmetic industries are the carboxyvinyl polymers known as carbomers. These are
synthetic high molecular weight polymers of acrylic acid, cross-linked with either
allylsucrose or allyl ethers of pentaerythritol. Pharmaceutical grades of these car-
bomers are available (e.g., Carbopol 981NF; B. F. Goodrich Performance Materials).
In the dry state, a carbomer molecule is tightly coiled, but when dispersed in water
the molecule begins to hydrate and partially uncoil, exposing free acidic moieties.
To attain maximum thickening effect the carbomer molecule must be fully uncoiled,
and this can be achieved by one of two mechanisms (Fig. 2). The most common
method is to convert the acidic molecule to a salt, by the addition of an appropriate
neutralizing agent. For formulations containing aqueous or polar solvent, carbomer
gellation can be induced by the addition of simple inorganic bases, such as sodium
or potassium hydroxide. Less polar or nonpolar solvent systems may be neutralized
with amines, such as triethanolamine or diethanolamine, or a number of alternative
amine bases (e.g., diisopropanolamine, aminomethyl propanol, tetrahydroxypropyl
ethylenediamine, and tromethamine) may be employed. Neutralization ionizes the
carbomer molecule, generating negative charges along the polymer backbone, and
the resultant electrostatic repulsion creates an extended three-dimensional structure.
Care must be taken not to under- or overneutralize the formulation, as this will result
in viscosity or thixotropic changes (38). Overneutralization will reduce viscosity, as
the excess base cations screen the carboxy groups and reduce electrostatic repulsion.
Hydrated molecules of carbomer may also be uncoiled in aqueous systems by the
addition of 10–20% of hydroxyl donors, such as a nonionic surfactant or a polyol,
that are able to hydrogen-bond with the polymer. Maximum thickening will not be
as instantaneous using this mechanism, as it is with base neutralization, and may
take several hours. Heating accelerates the process, but the system should not be
heated to more than 70⬚C. Because they are synthetic, carbomer bases vary little
from lot-to-lot, although batch-to-batch differences in mean molecular weight may
result in variations in the rheological characteristics of aqueous dispersions (39).
3. Emulsions
The most common emulsions used in dermatological therapy are creams. These are
two-phase preparations in which one phase (the dispersed or internal phase) is finely
dispersed in the other (the continuous or external phase). The dispersed phase can
Figure 2 A tightly coiled carbomer molecule will (a) hydrate and swell when dispersed in
water. (b) The molecule will completely uncoil to achieve maximum thickening when it is
converted from the acid form to the salt form on neutralization.
have either a hydrophobic-based (oil-in-water creams; O/W), or be aqueous based

(water-in-oil creams; W/O). Whether a cream is O/W or W/O depends on the prop-
erties of the system used to stabilize the interface between the phases. Because there
are two incompatible phases in close conjunction, the physical stability of creams is
always tenuous, although it may be maximized by the judicious selection of an
appropriate emulsion-stabilizing system. In most pharmaceutical emulsions, the sta-
bilizing systems comprise either surfactants (ionic or nonionic), polymers (nonionic
polymers, polyelectrolytes, or biopolymers), or mixtures of these. The most com-
monly used surfactant systems are sodium alkyl sulfates (anionic), alkylammonium
halides (cationic), and polyoxyethylene alkyl ethers or polysorbates (nonionic). These
are often used alone, or in conjunction with nonionic polymerics, such as polyvinyl
alcohol or poloxamer block copolymers, or polyelectrolytes, such as polyacrylic–
polymethacrylic acids.
The physicochemical principles underlying emulsion formulation and stabili-
zation are extremely complex and will not be covered in depth here. The interested
reader is referred to the volume edited by Sjöblom (40). Briefly, an emulsion is

formed when two immiscible liquids (usually, oil and water) are mechanically agi-
tated. When this occurs in the absence of any form of interfacial stabilization during
agitation, both liquids will form droplets that rapidly flocculate and coalesce into
two phases on standing. Flocculation is the term used to describe the close accu-
mulation of two or more droplets of dispersed phase without loss of the interfacial
film, and it is largely the result of van der Waals attraction. The flocculated droplets
may then coalesce into one large droplet, with the loss of the interfacial film. In
practice, there is a brief period when one of the phases becomes the continuous
phase because the droplets of this liquid coalesce more rapidly than the droplets of
the other. Physical stability of an emulsion is determined by the ability of an additive
to counteract the van der Waals attractions, thereby reducing flocculation and coa-
lescence of the dispersed phase. This may be achieved in two ways: an increase in
the viscosity of the continuous phase, which will reduce the rate of droplet move-
ment, or the establishment of an energy barrier between the droplets, or both. Al-
though increasing the viscosity of the continuous phase will reduce the rate at which
droplets flocculate, in pharmaceutical shelf-life terms, a stable system can be gen-
erated in this way only if the continuous phase is gelled and the droplet diameter is
smaller than 0.1 ␮m.
In pharmaceutical emulsions it is more common to develop stability using the
energy barrier technique and to complement this stabilization, if necessary, by in-
creasing the viscosity of the continuous phase. The basis of the energy barrier is that
droplets experience repulsion when they approach each other. Repulsion can be gen-
erated either electrostatically, by the establishment of an electric double layer on the
droplet surface, or sterically, by adsorbed nonionic surfactant or polymeric material.
Electrostatic repulsion is provided by ionic surfactants that, when adsorbed at the
oil–water interface, orient such that the polar ionic group enters the water. Some of
the surfactant counterion (e.g., the sodium ion of sodium dodecyl sulfate) will sep-
arate from the surface and form a diffuse cloud surrounding the droplet. This diffuse
cloud, together with the surface charge from the surfactant, forms the electric double
layer, and electrostatic repulsion occurs when two similarly charged droplets ap-
proach each other. For obvious reasons, this method of emulsion stabilization is
appropriate only for O/W formulations. In addition, it is important to appreciate that
emulsions stabilized by electrostatic repulsion are extremely sensitive to the presence
of additional electrolytes, which will disrupt the electrical double layer.
Steric repulsion may be produced using nonionic surfactants or polymers, such
as polyvinyl alcohol or poloxamers. The specific distribution of the polyethoxylated
nonionic surfactants and block copolymers (Fig. 3a and b) results in the formation
of a thick hydrophilic shell of polyoxyethylene chains around the droplet. Repulsion
is then afforded by both mixing interaction (osmotic repulsion) and entropic inter-
action (volume restriction), the latter as a result of a loss of configurational entropy
of the polyoxyethylene chains when there is significant overlap. For polymeric ma-
terials without definitive hydrophobic and hydrophilic regions, the adsorption energy
is critical to generation of steric repulsion. The adsorption energy must not be so
low that there is no polymer adsorption, nor so high that there is complete polymer
adsorption to the droplet. In either of these cases, there will be none of the loops,
or tails, (see Fig. 3c) that are essential to steric repulsion. Polymeric steric repulsion
is usually achieved using block copolymers, such as poloxamers, which consist of
Figure 3 Oil-in-water emulsions may be stabilized by (a) nonionic surfactants, (b) polox-
amer block copolymers, or (c) polymeric materials. The hydrophilic chains produce repulsion
by mixing interaction (osmotic) or volume restriction (entropic).
linked polyoxyethylene and polyoxypropylene chains. More recently, polyacrylic

acid polymers linked to hydrophobic chains (Pemulen; B. F. Goodrich) have been
used as primary emulsification systems in O/W formulations and are listed in the
USP as carbomer 1342. These materials form very stable emulsions because the
polyacrylic acid chain, anchored to the oil droplet by the alkyl methacrylate moieties,
considerably increases the surface charge on the oil droplet, forming a strong elec-
trical barrier at the interface. Furthermore, emulsion stability is enhanced by an
increase in the viscosity of the continuous phase.
Usually, for O/W emulsions, and always, for W/O emulsions, it is necessary
to select an emulsification system based on surfactants. Although ionic surfactants
can be used only for O/W emulsions, nonionic surfactants may be used for both
O/W and W/O formulations. Although, at first glance, the choice of surfactant system
appears limitless (there are hundreds to select from), there are some basic guidelines
to aid the formulator. As the first priority, use of pharmaceutically approved surfac-
tants (or those having a Drug Master File in place with the FDA) will save a con-
siderable amount of regulatory justification. Raw material suppliers will frequently
provide guidelines, although these will obviously be biased toward the use of their
products. However, it should be appreciated that suppliers have considerable expe-
rience in the applications and uses of their products, and they are a very useful
resource.
An approximate guide to emulsion formulation is provided by the hydrophilic–
lipophilic balance (HLB) system that generates an arbitrary number (usually between
0 and 20) that is assigned to a particular surfactant. The HLB value of a polyoxy-
ethylene-based nonionic surfactant may be derived from:
mol% hydrophobic group
HLB =
5
and the HLB of a polyhydric alcohol fatty acid ester (e.g., glyceryl monostearate)
may be derived from
HLB = 20 冉冊
1⫺S
A
where S is the saponification number of the ester and A the acid number of the fatty
acid. When it is not possible to obtain a saponification number (e.g., lanolin deriv-
atives), the HLB can be calculated from
(E ⫹ P)
HLB =
5
where E is the weight percent (wt%) of the polyoxyethylene chain and P the wt%
of the polyhydric alcohol group in the molecule.
From the foregoing equations it is apparent that hydrophilic surfactants have
high HLB values, and lipophilic surfactants have low HLB values. It is generally
recognized that surfactants with HLB values between 4 and 6 are W/O emulsifiers
and those with HLB values between 8 and 18 are O/W emulsifiers. It is also generally
recognized, although poorly understood, that mixtures of surfactants create more
stable emulsions than the individual surfactants. The overall HLB of a surfactant
mixture (HLBM) can be calculated from
HLBM = f HLBA ⫹ (1 ⫺ f )HLBB

where f is the weight fraction of surfactant A. The required emulsifier HLB values
for several oils and waxes are given in Table 3. Importantly, the HLB system can
be used only as an approximation in emulsion design, and stability of an emulsion
cannot be guaranteed by the use of an emulsifier mix with an appropriate HLB value.
For example, creaming of an emulsion, a typical physical stability problem, is much
more dependent on the viscosity of the continuous phase than the characteristics of
the interfacial film.
Mixtures of surfactants create more stable emulsions than the individual sur-
factants. A reasonable and coherent explanation for this is given by the gel network
theory (41,42). Briefly, this theory relates the consistencies and stabilities of O/W
creams to the presence or absence of viscoelastic gel networks in the continuous
phase. These networks form when there is an amount of mixed emulsifier, in excess
of that required to stabilize the interfacial film, that can interact with the aqueous
continuous phase. In its simplest form, a cream consists of oil, water, and mixed
emulsifier. Official emulsifying waxes may be cationic (cationic emulsifying wax
BPC; a mixture of cetostearyl alcohol and cetyltrimethylammonium bromide, 9:1),
anionic (emulsifying wax BP; a mixture of cetostearyl alcohol and sodium dodecyl
sulfate, 9:1) or nonionic (nonionic emulsifying wax BPC; a mixture of cetostearyl
alcohol and cetomacrogol, 4:1). The theory dictates that when the cream is formu-
lated it is composed of at least four phases (Fig. 4):
1. Bulk water
2. A dispersed oil phase
3. A crystalline hydrate
4. A crystalline gel phase composed of bilayers of surfactant and fatty alcohol
separated by layers of interlamellar-fixed water
An examination of ternary systems (containing emulsifying wax and water, but
no oil phase) by X-ray diffraction indicated that, for all emulsifying systems, addition
of water caused swelling of the interlamellar spaces. Ionic emulsifying systems pos-
sess a greater capacity to swell than nonionic systems. Swelling in ionic systems is
an electrostatic phenomenon, whereas in nonionic systems, it is due to hydration of
the polyoxyethylene chains and is limited by the length of this chain. The gel network
Table 3 HLB Values for Several Oils

and Waxes
Emulsion type
Constituent O/W W/O
Liquid paraffin 12 4
Hard paraffin 10 4
Stearic acid 16 —
Beeswax 12 5
Castor oil 14 —
Cottonseed oil 9 —
Figure 4 The gel network theory suggests that when a cream is formulated it is composed
of four phases: bulk water, a dispersed oil phase, a crystalline hydrate, and a crystalline gel
phase composed of bilayers of surfactant and fatty alcohol separated by layers of interlamellar-
fixed water. (Courtesy of Dr. G. M. Eccleston.)
theory also offers an explanation for the observation that nonionic O/W creams
thicken on storage. This change is related to additional gelation in the continuous
phase as a result of slow hydration of the polyoxyethylene chains of the surfactant,
which reduces the amount of free water in the formulation.
A relatively stable emulsion formulation may be prepared from a simple four-
component mixture: oil, water, surfactant, and fatty amphiphile. In practice, however,
things are never that straightforward. In addition to the four principle components,
a pharmaceutical emulsion formulation will also contain a drug, and is likely to
contain a cosolvent for the drug, a viscosity enhancer, a microbiological preservative
system, a pH adjusting–stabilizing buffer, and an antioxidant system. All of these
additional components are required so that the formulation is capable of delivering
the correct amount of drug to the therapeutic application site from a formulation that
is free from microbial contamination and is essentially physically unchanged from
the day of manufacture.
B. Preservation of Semisolid Formulations

All pharmaceutical semisolid formulations that are not sterilized unit-dose products
can support the growth of microorganisms. Preservatives are ingredients that prevent
or retard microbial growth and protect formulations from spoilage. The use of pre-
servatives is required to prevent product damage caused by microorganisms during
manufacture, storage, and inadvertent contamination by the patient during use. Sim-
ilarly, preservatives serve to protect consumers from possible infection from contam-
inated products. The characteristics of an ideal preservative system are shown in
Table 4. No single preservative meets all of these characteristics for all formulations
(43), and it is often necessary to use a system containing a combination of individual
preservatives. It is also important to appreciate that preservatives are intrinsically
toxic materials so that a balance must be achieved between antimicrobial efficacy
and dermal toxicity.
The most commonly used preservatives in pharmaceutical products are the
parabens (alkyl esters of p-hydroxybenzoic acid, such as methyl and propyl paraben).
These compounds are highly effective against both gram-positive bacteria and fungi
at low concentrations (e.g., 0.1–0.3% parabens combinations provide effective pres-
ervation of most emulsions). Because of their widespread use, the toxicological pro-
files of the parabens have been extensively researched, and the safety in use of the
lower esters (methyl, ethyl, propyl, and butyl) has been established (45). Other pre-
servatives widely used in topical pharmaceutical formulations include benzoic acid,
sorbic acid, benzyl alcohol, phenoxyethanol, chlorocresol, benzalkonium chloride,
and cetrimide. Each has particular advantages and disadvantages, which makes com-
bination preservatives particularly effective. For example, although methyl paraben
is highly active against gram-positive bacteria and moderately active against yeasts
and molds, it is only weakly active against gram-negative bacteria. However, a com-
bination of methyl paraben with phenoxyethanol provides a preservative system that
is also highly active against gram-negative species.
The acid preservatives (benzoic and sorbic acids) are active only as free acids
and formulations containing these preservatives must be buffered to acid pH values
(pH < 5). A list of pharmaceutical preservatives useful in topical formulations is
given in Table 5, together with their microbiological and physicochemical properties.
More information on preservatives and preservative systems may be found in the
British Pharmaceutical Codex (The Pharmaceutical Press) and in the excellent text
by Orth (43).
Table 4 Characteristics for an Ideal Preservative
1. Effective at low concentrations against a wide spectrum of microbes

2. Soluble in the formulation at the required concentration
3. Nontoxic and nonsensitizing to the consumer at in-use concentrations
4. Compatible with other formulation components
5. No physical effect on formulation characteristics
6. Stable over a wide range of pH and temperature
7. Inexpensive
Source: Ref. 44.

Table 5 Microbiological and Physicochemical Properties of Selected Preservatives
Antimicrobial activitya
In-use pH
Preservative Gram⫹ Gram⫺ Molds Yeasts Conc (%) rangeb O/Wc
Benzoic acid 1 2 3 3 0.1 2–5 3–6

Sorbic acid 2 2 2 1 0.2 <6.5 3.5
Phenoxyethanol 2 1 3 3 1.0 Wide —
Methyl paraben 1 3 2 2 0.4–0.8 3.0–9.5 7.5
Propyl paraben 1 3 2 2 0.4–0.8 3.0–9.5 80
Butyl paraben 1 3 2 2 0.4–0.8 3.0–9.5 280
Chlorocresol 1 2 3 3 0.1 <8.5 117–190
Benzalkonium Cl 1 2 3 2 0.01–0.25 4–10 <1
Cetrimide 1 2 3 2 0.01–0.1 4–10 <1
a
1, highly active; 2, moderately active; 3, weakly active.
b
Optimal pH range for activity.
c
Oil–water partition coefficient.
Interestingly, despite their widespread use and excellent safety profile, there is
increasing interest in the potential systemic exposure to preservatives following ap-
plication in pharmaceutical and cosmetic products. Skin penetration data are not
available in the literature for many preservatives, and much of the publicly available
data has been obtained under conditions inappropriate for risk assessment. By far
the most available data concern the parabens.
Although most studies on skin permeation of parabens have evaluated one or
two of the homologous series, recently Dal Pozzo and Pastori [46] reported on the
in vitro human skin permeation of six parabens (methyl, ethyl, propyl, butyl, hexyl,
and octyl esters). The permeants were applied to abdominal epidermal membranes,
mounted in diffusion chambers, either as solid compounds (deposited in acetone) or
as saturated solutions in various vehicles, including three typical emulsion formu-
lations (two O/W and one W/O). Following application of the unformulated pure
substance, maximum flux decreased with increasing lipophilicity, from 65.0 ␮g/cm2
h⫺1 for methyl paraben to 13.7 ␮g/cm2 h⫺1 for hexyl paraben. Similarly, when applied
as saturated aqueous solutions, the maximum flux decreased with increasing lipo-
philicity. In the latter case, however, when flux was normalized to the vehicle con-
centration of the permeant, the permeability coefficient increased with increasing
lipophilicity. Addition of 50% propylene glycol or 20% polyethylene glycol 400 to
the aqueous solutions did not alter the profile of permeation, although the perme-
ability coefficients were somewhat reduced. On the other hand, when the parabens
were dissolved in liquid paraffin, the relation between permeability coefficient and
lipophilicity appeared parabolic, maximum flux occurred for the butyl ester, and the
highest permeability coefficient occurred for the propyl ester. These results are con-
sistent with theory and demonstrate that the application vehicle can significantly
affect skin permeability characteristics of compounds.
In emulsion systems the existence of two distinct phases (oil and water) results
in distribution of the parabens according to their physicochemical characteristics, and
this can be influenced by the presence of other ingredients, such as cosolvents and
surfactants. Furthermore, these excipients may also affect skin barrier properties. It
was observed that permeation of the parabens from two O/W emulsions was higher
than expected, based on the data from simple vehicles, and that permeation from the
O/W emulsions was higher than that from the W/O emulsion (46). These results
were rationalized on the basis of permeant release from the formulation, and it was
assumed that the external lipid phase of the W/O emulsion retained parabens.
It is difficult, however, to relate this data to conditions of actual consumer
exposure because, in the DalPozzo and Pastori study (46), the formulations were
applied at infinite dose (which effectively generated occlusive conditions) and at an
artificially high permeant concentration (0.7% w/w). When low finite doses are ap-
plied to the skin surface, the vehicle is continually changing. Whether the vehicle is
O/W or W/O, the water content will be released by the shear forces generated by
application to the skin, most of the water will evaporate, and this should tend to
reduce the skin penetration of these preservatives.
It is theoretically possible to reduce skin absorption of the parabens by for-
mulation manipulation. This may be achieved by altering the distribution pattern of
the preservative within the formulation or by complexation. Care must be taken,
however, to ensure that any formulation modification does not interfere with the
antimicrobial activity of the preservative system (47). One such modification in-
volved the complexation of methyl paraben with 2-hydroxypropyl-␤-cyclodextrin
(48). Although the aqueous solubility of methyl paraben was increased considerably
in the presence of the cyclodextrin, the percutaneous penetration of the preservative
through hairless mouse skin in vivo over 24 h was reduced by 66%. Even if this
system clearly had benefit in terms of potential reduction in systemic absorption of
the preservative, there was no indication of any possible alteration in preservative
efficacy. Incorporation of butyl paraben into liposomes prepared from phosphatidyl-
choline–cholesterol–dicetyl–phosphate had no effect on preservative efficacy, al-
though the antimicrobial effect was proportional to the free, and not the total, con-
centration of the preservative (49). A further study by these authors demonstrated
that incorporation of butyl paraben in some liposomal systems had little effect on
the permeation of the preservative across guinea pig skin in vivo or in vitro (50),
although incorporation of increasing amounts of lipid into the liposome tended to
decrease the percutaneous absorption.
Although there is considerable evidence that parabens can penetrate the skin,
permeation and systemic availability of intact compounds are likely to be consider-
ably reduced by transcutaneous and systemic metabolism. Furthermore, because these
preservatives are present at concentrations of 0.1–0.2% w/w in topical pharmaceu-
tical formulations, in use, dermal exposure to these compounds is relatively low. In
the cosmetic industry there is a trend toward preservative-free and self-preserving
formulations (51). However, before taking this route, the pharmaceutical formulator
must consider the potential implications on efficacy and safety of the product.
C. Drug Release from Semisolid Formulations and SUPAC-SS

Determination of the ability of a semisolid formulation to release a drug, and the
pattern and rate of this release, are important aspects of formulation development
and optimization. However, it is also important to appreciate that the data generated
must not be over interpreted. Release studies normally involve measurement of drug
diffusion out of a mass of formulation into a receiving medium that is separated

from the formulation by a synthetic membrane (52,53). A detailed study of the data
from this type of system can generate invaluable data concerning the physical state
of the drug in the formulation. For example, examination of the early experimental
models and their refined updates that were derived to describe drug release from
semisolids, reveals that release patterns are dependent on whether the drug is present
as a solution or suspension within the formulation (54–56). These subtle differences,
together with variations in the rate of release, may be used to determine such param-
eters as drug diffusivity within the formulation matrix, the particle size of suspended
drug, and the absolute solubility of a drug within a complex formulation (Fig. 5)
(57). However, it is generally agreed that drug release-rate data cannot predict skin
permeation or bioavailability. Application of a formulation to the skin is an ephemeral
situation. The formulation will undergo considerable shearing forces, solvents will
evaporate, and excipients may interact with the skin and modulate bioavailability.
Nonetheless, release-rate determinations are important for purposes other than for-
mulation development and characterization.
The FDA has issued a guidance document (SUPAC-SS Nonsterile Semisolid
Dosage Forms, U.S. Department of Health and Human Services, Food and Drug
Figure 5 Illustration of the use of release rates from semisolid preparations to determine
drug solubility within a formulation. Data shows the release rates of benzocaine from pro-
pylene glycol/water gels as a function of drug concentration in the formulation. (From Ref.
85.)
Administration, Center for Drug Evaluation and Research, May 1997) that recom-
mends the use of in vitro drug-release testing in the scale-up and postapproval
changes for semisolids (SUPAC-SS). The FDA intends to promote the use of this
test as a quality assurance tool to monitor minor differences in formulation compo-
sition or changes in manufacturing sites, but not yet as a routine batch-to-batch
quality control test. Thus, although the agency is suggesting in vitro release rate data
for level 2 and level 3 changes in formulation components and composition, such
data are not required for a level 1 change. In the former case, the in vitro release
rate of the new or modified formulation should be compared with a recent batch of
the original formulation, and the 90% confidence limit should fall within the limits
of 75–133%. Similarly, in vitro release testing is suggested for level 2 changes in
manufacturing equipment, processes, and scale-up, and level 3 changes in manufac-
turing site. Recently, the use of in vitro testing as a quality assurance tool has been
questioned, especially for a hydrophilic formulation containing the highly water-
soluble drug, ammonium lactate (58), for which the method was insufficiently spe-
cific to differentiate between small differences in drug loading or minor composi-
tional and processing changes.
D. Formulation of Transdermal Products

This section concentrates on those aspects of development specific to transdermal
delivery systems.
1. Pharmacokinetic Modeling
To assess the feasibility of the transdermal route for a specific therapeutic agent,
several pharmacokinetic factors, including rate of absorption and elimination, must
be considered. Guy and Hadgraft (59) developed an early kinetic model for topical
drug delivery. Various rate constants were described including release rate from the
device (first- or zero-order); back diffusion from the stratum corneum to the patch
(usually insignificant); diffusion rates of the compound across the stratum corneum
and viable epidermis; reverse rate constant (enabling prediction of the stratum cor-
neum–viable epidermis partition coefficient) and the plasma clearance rate constant.
This model has been used to predict the plasma concentration of several transder-
mally administered drugs and has shown remarkable similarities between predicted
and actual plasma concentration profiles, suggesting that the model is useful for
evaluating the feasibility of potential transdermal drug delivery candidates.
Although this model provides predictive quantitative data based mainly on
physicochemical and pharmacokinetic parameters, other aspects of the percutaneous
absorption process are more difficult to predict. Of most concern is the potential for
metabolic degradation of the drug, but only limited data are available on quantitative
aspects of cutaneous metabolism. Modeling is possible, by estimating metabolic rate
constants and varying the possible residence time of the drug in the skin. More
lipophilic drugs, which have a relatively long residence time in the skin, will be
more exposed to cutaneous metabolic enzymes and be less bioavailable than their
hydrophilic counterparts.
More recent models for prediction of skin permeation and pharmacokinetics
have become increasingly elaborate and take into account such parameters as the
hydrogen-bonding capability of the permeant and potential vehicle effects (60).
2. Biopharmaceutical Considerations
A fundamental consideration in the development of transdermal therapeutic systems
is whether the dermal delivery route can provide the requisite bioavailability for drug
effectiveness. This is ultimately determined by the skin-penetration rate of the drug,
the potential for metabolism during permeation across the skin, and the biological
half-life of the drug. Penetration rates may be modified, if necessary, by the use of
penetration enhancers, but drug metabolism and plasma clearance cannot be influ-
enced by any simple means. Although prediction of skin penetration and bioavail-
ability of drugs from transdermal therapeutic systems has been reasonably accurate,
there is no doubt that testing of formulated patches in vitro and in vivo will continue
to be the most accurate means of evaluating their usefulness.
A wide variety of experimental approaches have been developed for in vitro
drug permeability determination through skin, and guidelines have been established
to rationalize this aspect of pharmaceutical development (see Chapter 5). In the early
stages of product development, skin penetration rates from prototype vehicles and
patches are usually determined in vitro using simple diffusion cells and skin from a
variety of animals. Although, the use of in vitro systems provides little quantitative
information on the transcutaneous metabolism of candidate drugs, a major advantage
is that experimental conditions can be controlled precisely so that the only variables
are in the prototype formulations. In the latter stages of product development, when
quantitative skin permeation data is required, human skin should be the membrane
of choice for use in in vitro systems. Although methods are available to improve the
sensitivity of in vitro skin penetration measurements (61), it is essential, at this stage,
to ensure that account is taken of the inherent variability in human skin permeation.
Factorial design and artificial neural networks have been used in the optimi-
zation of transdermal drug delivery formulations using in vitro skin permeation tech-
niques (62–64). For example, Kandimalla et al. (63) optimized a vehicle for the
transdermal delivery of melatonin using the response surface method and artificial
neural networks. Briefly, three solvents (water, ethanol, and propylene glycol) were
examined either as single solvents or binary and ternary mixtures. Measurements of
skin flux, lag time, and solubility were made for ten vehicles and compared with
values predicted from both a response surface generated from a quartic model and
an artificial neural network employing a two-layered back-propagation network with
all ten design points in the hidden layer. Predictability of flux using both statistical
techniques was good (Table 6), suggesting that such models may be useful in pre-
liminary formulation optimization.
A major drawback of transdermal delivery systems is the potential for localized
irritant and allergic cutaneous reactions. At the earlier stages of formulation devel-
opment, it is, therefore, important to evaluate both drugs and excipients for their
potential to cause irritation and sensitization (see Chapters 10 and 11). This is true
for all transdermal systems, but especially for those that may stay in place for pro-
longed periods. The degree of primary and chronic irritation, and the potential to
cause contact allergy, photoirritation, and photoallergy should be determined. Nor-
mally, the drug and excipients are initially separately evaluated for contact irritation
and sensitization in animal models before evaluation in human subjects. It must,
however, be emphasized that animal data are often not predictive of the human
situation. Evaluation of skin irritation and delayed contact hypersensitivity should
Table 6 Experimental Versus Predicted Fluxa of Melatonin
Application Experimental ANN Prediction RSM Prediction

vehicleb (␮g/cm2 h⫺1)c (␮g/cm2 h⫺1) (␮g/cm2 h⫺1)
W:E:P (20:60:20) 11.32 ⫾ 0.86 12.17 12.73

W:E (40:60) 10.89 ⫾ 1.36 11.83 11.76
W:P (40:60) 7.54 ⫾ 1.39 6.75 7.50
a
Flux was predicted on the basis of artificial neural networks (ANN) or response surface
methods (RSM).
b
W, water; E, ethanol; P, propylene glycol.
c
Data are means ⫾ standard deviation (n = 3) and represent flux through rat dorsal skin.
Source: Ref. 63.
always be carried out using the final and complete formulation in human volunteers.
Fortunately, most of the observed skin reactions to transdermal systems are transient
and mild and disappear within hours of patch removal.
3. Design Considerations
All patch-type transdermal delivery systems developed to date can be described by
three basic design principles: drug in adhesive, drug in matrix (usually polymeric),
and drug in reservoir (Fig. 6). In the latter the reservoir is separated from the skin
by a rate-controlling membrane. Although there are many differences in the design
of transdermal delivery systems, several features are common to all systems includ-
ing the release liner, the pressure-sensitive adhesive and the backing layer, all of
which must be compatible for a successful product. For example, if a system is
designed in such a way that the drug is intimately mixed with adhesive, or diffuses
from a reservoir through the adhesive, the potential for interaction between drug and
adhesive, which can lead to either a reduction of adhesive effectiveness, or the for-
mation of a new chemical species, must be fully assessed. Similarly, residual mono-
mers, catalysts, plasticizers, and resins may react to give new chemical species.
Additionally, the excipients, including enhancers, or their reaction products, may
interfere with adhesive systems. Incompatibilities between the adhesive system and
other formulation excipients, although undesirable, may not necessarily be impeding
and designs in which the adhesive is remote from the drug delivery area of the
system may be developed (see Fig. 6d). There are three critical considerations in the
selection of a particular system: adhesion to skin, compatibility with skin, and phys-
ical or chemical stability of total formulation and components.
All devices are secured to the skin by a skin-compatible pressure-sensitive
adhesive. These adhesives, usually based on silicones, acrylates, or polyisobutylene,
can be evaluated by shear-testing and assessment of rheological parameters. Standard
rheological tests include creep compliance (which measures the ability of the adhe-
sive to flow into surface irregularities), elastic index (which determines the extent of
stretch or deformation as a function of load and time) and recovery following de-
formation. Skin-adhesion performance is based on several properties, such as initial
and long-term adhesion, lift, and residue. The adhesive must be soft enough to ensure
initial adhesion, yet have sufficient cohesive strength to remove cleanly, leaving no
residue. Because premature lift will interfere with drug delivery, the cohesive and
Figure 6 Typical transdermal drug delivery system designs.
adhesive properties must be carefully balanced and maintained over the period of
intended application. This can be evaluated only by wear-testing, in which a placebo
patch is applied to skin.
Skin adhesion is affected by shape, conformability, and occlusivity and round
patches tend to be more secure than those of sharply angled geometry. If the patch
is able to conform to the skin contours it resists lifting and buckling with movement.
The presence of water may affect adhesive properties; therefore, the occlusivity of
the system must be taken into consideration. Occlusion for prolonged periods can
lead to excessive hydration and problems associated with microbial growth that may
increase the possibility of irritation or sensitization to the various system components.
The backing material and release liner can be fabricated from a variety of
materials, including polyvinylchloride, polyethylene, polypropylene, ethylene vinyl
acetate, and aluminium foil. The principal requirement is that they are impervious
to the drug and other formulation excipients. The most useful backing materials are
those that conform with the skin and provide sufficient resistance to transepidermal
water loss to allow some hydration of the stratum corneum, yet maintain a healthy
subpatch environment. The release liner must be easily separated from the adhesive
layer without lifting off any of the adhesive to which it is bound. Liners are usually
films or coated papers, and silicone release coatings are used with acrylate- and
rubber-based adhesive systems, whereas fluorocarbon coatings are used with silicone
adhesives (65).
4. Drug and Enhancer Incorporation
The three principal methods for incorporating the active species into a transdermal
system have led to the loose classification of patches as membrane, matrix, or drug-
in-adhesive types. It is, however, quite possible to combine the main types of patch;
for example, by placing a membrane over a matrix, or using a drug-in-adhesive in
combination with a membrane–matrix device to deliver an initial bolus dose.
Membrane patches contain a delivery rate-controlling membrane between the
drug reservoir and the skin. Microporous membranes, which control drug flux by the
size and tortuosity of pores in the membrane, or dense polymeric membranes,
through which the drug permeates by dissolution and diffusion, may be used. Several
materials can be used as rate-controlling membranes (e.g., ethylene–vinyl acetate
copolymers, silicones, high-density polyethylene, polyester elastomers, and poly-
acrylonitrile). Ideally, the membrane should be permeable only to the drug and en-
hancer (if present) and should retain other formulation excipients. Membranes have
also been designed such that they allow differential permeation of an enhancer and
drug (66–68). This type of membrane, sometimes designated as a one-way mem-
brane, is useful when the drug is present in the adhesive and the enhancer is for-
mulated in a reservoir.
Asymmetric polymeric membranes have also been evaluated for use in trans-
dermal delivery systems (69). Asymmetric poly(4-methyl-1-pentene) membranes,
fabricated using a dry–wet inversion method, were used to control the delivery of
nitroglycerin. The release rates of nitroglycerin were strongly influenced by the na-
ture and amount of the nonsolvent (butanol) used, together with the solvent (cyclo-
hexane), in the casting process. This is, perhaps, not surprising, as increasing the
amount of nonsolvent increases the porosity of the cast membrane. The concept of
fine-tuning delivery of a drug through a given membrane by subtle adjustment of
the porosity creates some exciting new possibilities in transdermal technology (70).
In all marketed membrane-controlled transdermal systems, the rate-controlling
membrane is fabricated from synthetic polymeric materials. Thacharodi and Rao (71)
evaluated the potential of two biopolymers (fetal calf skin collagen and chitosan) in
membrane systems for delivery of nifedipine. Chitosan (deacetylated chitin) is a
widely distributed major constituent of the shells of marine shellfish. It was con-
cluded that the permeability of both biopolymers could be readily adjusted by altering
the fabrication method or cross-linking and, because these polymers were biocom-
patible, they were more suitable for use as rate-controlling membranes in transdermal
systems.
A variety of materials can be used in the drug reservoir, ranging from simple
formulations (such as mineral oil), to complex formulations (such as aqueous–al-
coholic solutions and gels, with or without various cosolvents, and polymeric ma-
terials). A definite requirement for a reservoir system is that it can permit zero-order
release of the drug over the delivery period. Essentially, this requires that the res-
ervoir material is saturated with the drug over the period of product application,
which can be achieved by formulating the drug as a suspension.
The second type of transdermal system is the matrix design, in which the drug
is uniformly dispersed in a polymeric matrix, through which it diffuses to the skin
surface. Here, the polymeric matrix, which may comprise silicone elastomers, poly-
urethanes, polyvinyl alcohol, polyvinylpyrrolidones, and such, may be considered
the drug reservoir. Several steps are involved in the drug delivery process: principally
dissociation of drug molecules from the crystal lattice, solubilization of the drug in
the polymer matrix, and diffusion of drug molecules through the matrix to the surface
of the skin. Many variables may affect the dissolution and diffusion rates, making it
particularly difficult, but not impossible, to predict release rates from experimental
or prototype formulations (72). For a drug to be released from a polymeric matrix
under zero-order kinetics, the drug must be maintained at saturation in the fluid phase
of the matrix, and the diffusion rate of the drug within the matrix must be much
greater than the diffusion rate in the skin.
Several methods can be used to alter the release rate of a drug or an enhancer
from a polymeric matrix, and some of these are illustrated by a study on release of
several drugs from silicone matrices (73). Silicone medical-grade elastomers (poly-
dimethylsiloxanes) are flexible, lipophilic polymers, with excellent compatibility with
biological tissues, that can be coformulated with hydrophilic excipients, such as
glycerol, and inert fillers, such as titanium dioxide, to alter release kinetics. Increasing
the amount of glycerol in the matrix increased the release rate of indomethacin,
propranolol, testosterone, and progesterone, whereas incorporation of inert fillers
(titanium dioxide or barium sulfate) tended to reduce the release rate. Hydrophilic
drug-release rates from polydimethylsiloxane matrices were also increased by up to
three orders of magnitude using polydimethylsiloxane–polyethylene oxide graft co-
polymers (74). These data demonstrate that release rates can be modulated to achieve
a desired profile by simple formulation modification.
Perhaps the simplest form of transdermal drug delivery device, which is now
most commonly employed, is the drug-in-adhesive system. This involves formulating
the drug, and enhancer if present, in an adhesive mixture that is subsequently coated
onto a backing membrane, such as a polyester film, to produce an adhesive tape.
This simplicity is, however, deceptive and several factors, involving potential inter-
action between drug or enhancer and the adhesive, need to be considered. These can
involve chemical interactions resulting in interference with adhesive performance,
breakdown of the active species, or formation of new chemical entities. Additionally,
the physicochemical characteristics of the drug and adhesive system may provide
very different release rates for hydrophilic and hydrophobic drugs: for example,
silicone adhesives are typically lipophilic, which limits solubility of hydrophilic
drugs within the adhesive matrix.
Incorporation of other excipients, such as skin permeation enhancers, into a

drug-in-adhesive system may alter drug release rates and adhesive properties. For
example, addition of 1% urea to a polyacrylate pressure-sensitive adhesive resulted
in loss of adhesion and skin contact could not be maintained over the required period
(75). A strategy to reduce the influence of drug and enhancer on adhesive properties
was to design a system in which there was no contact between these constituents
and the adhesive, by limiting the adhesive to a boundary laminate that surrounded
a drug–enhancer-releasing layer. One disadvantage of this type of system, however,
is that the drug–enhancer-releasing layer may not remain in intimate contact with
the skin. If high levels of liquid skin penetration enhancers are incorporated into
drug-in-adhesive transdermal patches, there is likely to be a loss in cohesiveness,
resulting in patch slipping and skin residues following patch removal. Cohesive
strength, can be increased by high levels of cross-linking in acrylate adhesives, but
this may alter both long-term bonding and drug release rates. These problems may
be overcome by use of grafted copolymer adhesives, such as ARcare ETA Adhesive
Systems (76), for which reinforcement is achieved mainly through phase separation
of the side chain within the continuous polymer network. A variety of side chains
are available and enhancer concentrations up to 30% can be incorporated without
seriously affecting the adhesive properties. This work has been limited to fatty acid
ester type enhancers and application to other enhancer types remains to be estab-
lished. It may also be possible to maintain adhesive properties in the presence of
skin penetration enhancers by using different molecular weight blends of acrylic
copolymers (77,78).
When enhancers are incorporated into transdermal systems it is important to
appreciate that it is a fundamental requirement that the enhancer, as well as the drug,
is released by the adhesive. Furthermore, it is probable that the presence of the
enhancer may increase ski permeation of other formulation excipients and that this
may have an influence on local toxicity. Much remains to be done in the field of
enhancer incorporation into transdermal drug delivery systems and it is encouraging
to observe the increasing efforts of adhesive manufacturers in this sphere.
5. System Manufacture and Testing
The manufacturing processes for reservoir, matrix, and drug-in-adhesive transdermal
systems are, to a large extent, similar. All involve the following stages: preparing
the drug; mixing the drug (with other excipients and penetration enhancers, if re-
quired) with the reservoir, matrix, or adhesive; casting into films and drying (or
molding and curing); laminating with other structural components (e.g., backing
layer, rate-controlling membrane, and release liner); die-cutting; and finally, pack-
aging. Casting and lamination are the most critical steps in the manufacturing pro-
cess: tensions and pressures must be carefully controlled to provide a wrinkle-free
laminate that ensures reproducible adhesive-coating thickness and uniform drug con-
tent (79).
As with all controlled-release delivery systems, final product checks include
content uniformity, release-rate determination, and physical testing. Content-unifor-
mity evaluation involves removing a random sample of patches from a batch and
assaying the amount of drug present. Of the several methods available for determin-
ing drug release rates from controlled-release formulations, the U.S. Pharmaceutical
Manufacturers Association (PMA) Committee (80) has recommended three: the
‘‘paddle over disk’’ (which is identical with the USP paddle dissolution apparatus,
except that the transdermal system is attached to a disk or cell resting at the bottom
of the vessel that contains medium at 32⬚C); the ‘‘cylinder-modified USP basket’’
(which is similar to the USP basket method, except that the system is attached to
the surface of a hollow cylinder immersed in medium at 32⬚C), and the ‘‘recipro-
cating disk’’ (in which patches attached to holders are oscillated in small volumes
of medium, allowing the apparatus to be useful for systems delivering low concen-
trations of drug). Researchers at the FDA have developed a modified paddle proce-
dure (essentially the ‘‘paddle over disk’’ method) for determining drug release from
transdermal systems (81). One problem with the original method was the mode of
maintaining the patch in position in the dissolution beaker, and a device to improve
and maintain placement of the patch was subsequently suggested (82). Although the
industry is moving rapidly in method development, much remains to be established
in the field of pharmacopeal standards for transdermal drug delivery systems.
IV. REVIEW OF CLINICAL USE OF TRANSDERMAL SYSTEMS

Despite the barriers to systemic delivery of drugs through intact human skin, there
have been several successful transdermal products. The remainder of this chapter
describes recent clinical studies on several of these products, including those deliv-
ering nicotine, glyceryl trinitrate, estrogens, androgens, fentanyl, clonidine, and non-
steroidal anti-inflammatory agents (NSAIDs).
A. Nicotine
1. Smoking Cessation
Data on efficacy, safety, and pharmacoeconomics of transdermal (TD) nicotine ther-
apy for smoking cessation up to 1994 were reviewed (86). TD nicotine more than
doubled success rates of smoking cessation in motivated subjects smoking 10–15,
or more, cigarettes a day. Application site reactions (erythema or burning ⱕ16%,
transient itch ⱕ50%) caused discontinuation in 10%, or fewer of subjects. Sleep
disturbance due to nocturnal nicotine absorption occurred in less than 13% of sub-
jects. Attempts to develop a nicotine transdermal system (TDS), with reduced adverse
skin reactions, used in vitro human cadaver skin permeation to demonstrate a per-
meation profile comparable with those from Habitrol and Nicoderm, and a clinical
study compared systemic bioavailability and pharmacokinetic profiles (87). Imme-
diate effects of TD nicotine on sleep architecture, snoring, and sleep-ordered breath-
ing in 20 nonsmoking subjects with a history of habitual snoring receiving placebo
or a nicotine TD system delivering 11 mg/24 h were reported (88). Patches were
applied at 6 PM and removed after 12 h. Mean nicotine level was nondetectable with
placebo and 7.8 ⫾ 2.3 ng/mL with active therapy. Nicotine significantly decreased
total sleep time by 33 min, sleep efficiency from 89.7 to 83.5%, and rapid eye
movement (REM) sleep from 18.8 to 15.1%. Although sleep index was unchanged,
mean snoring intensity decreased by 1.1 dB (p = 0.01) with nicotine. Nausea (50%)
and vomiting (20%) were predominant side effects.
Concurrent administration of the nicotine antagonist mecamylamine and nico-
tine TD was evaluated (89) in a trial with 48 healthy smokers. Nicotine TD (6–8
weeks) plus oral mecamylamine (2.5–5 mg twice a day for 5 weeks) was compared
with nicotine TD plus placebo. Mecamylamine treatment began 2 weeks before

smoking cessation. At 7 weeks with mecamylamine, the continuous abstinence rate
tripled (50%) and point abstinence doubled (58%) compared with that for placebo.
Continuous abstinence was threefold higher for mecamylamine at 6 months and
ninefold higher at 12 months. It was concluded that agonist–antagonist therapy may
substantially improve smoking cessation. The pharmacokinetic and pharmacody-
namic interactions between TD mecamylamine and intravenous nicotine were inves-
tigated (90). Cigarette smokers received TD mecamylamine (6 mg/24 h) and placebo
patches for 7 days each. On day 5, subjects received a combined infusion of deu-
terium-labeled nicotine and cotinine; the disposition kinetics of nicotine and cotinine,
and cardiovascular and plasma catecholamine responses to nicotine were measured.
Fifty percent of subjects were studied under alkaline urine conditions and 50% under
acidic urine conditions. Steady-state plasma mecamylamine concentrations were dou-
ble (12.2 vs. 6.3 ng/mL), consistent with lower renal clearance (2.1 vs. 5.8 mL/min
kg⫺1) during alkaline compared with acidic urine conditions. Mecamylamine did not
significantly affect clearance of nicotine or cotinine, but did significantly reduce the
volume of distribution and inhibited the epinephrine-releasing effects of nicotine. It
was concluded that mecamylamine had little effect on nicotine clearance and was
not expected to affect steady-state levels during TD nicotine dosing. Reduction of
the volume of distribution of nicotine by mecamylamine suggested that part of the
antagonism of nicotinic CNS effects by mecamylamine may be due to a pharma-
cokinetic interaction, probably decreased binding to nicotine receptors, or transport
of nicotine into the brain, and may decrease potential adverse cardiovascular effects
of coadministered nicotine.
Stable isotope-labeled nicotine was infused simultaneously with TD nicotine
to determine absolute bioavailability and absorption kinetics from TD systems of
different designs (91). Rapid intravenous infusion of nicotine slowed absorption of
TD nicotine, probably by constricting dermal blood vessels, thereby limiting per-
cutaneous absorption. A study evaluated safety and tolerability of a 44-mg/day dose
for smoking cessation in subjects who smoked 20 or more, cigarettes per day (92).
Smokers received 44 mg/day of TD nicotine for 4 weeks, followed by 4 weeks at
22 mg/day. Thirty-eight of forty patients (95%) completed the initial 4 weeks, and
36 (90%) completed the entire study. Confirmed smoking cessation rates were 65%
(4 weeks) and 55% (8 weeks), and self-reported smoking cessation at 3 months was
50%. Sleep complaints were reported by 33% of subjects during the 44-mg phase.
It was concluded that 44 mg/24 h nicotine patch therapy in heavy smokers was safe,
tolerable, and without significant adverse events. Efficacy and safety of 22- and 44-
mg doses of TD nicotine paired with minimal, individual, or group counseling
to improve smoking cessation rates were compared in an 8-week clinical trial using
daily cigarette smokers (15 or fewer cigarettes per day for ⱖ1 year) and
random assignment to dose and counseling conditions (93). Four weeks of 22- or
44-mg transdermal nicotine followed by 4 weeks of dosage reduction (2 weeks of
22 mg followed by 2 weeks of 11 mg). Smoking cessation rates for TD doses and
levels of counseling did not differ significantly 8 or 26 weeks postquitting date. In
those receiving minimal contact, 44 mg produced greater abstinence (68%) at 4
weeks than 22 mg (45%). Participants with minimal-contact adjuvant treatment were
less abstinent at 4 weeks than those receiving individual or group counseling (56 vs.
67%). Although a 44-mg dose decreased desire to smoke more than a 22-mg dose,
this effect was unrelated to success in quitting smoking. The 44-mg dose produced
significantly higher frequencies of nausea (28%), vomiting (10%), and erythema,
with edema at the patch site (30%) than did the 22-mg dose (10, 2, and 13%,
respectively). It was concluded that there was no general, sustained, benefit of ini-
tiating TD nicotine therapy with a 44-mg dose or providing intense adjuvant ces-
sation treatment.
The effects of using nicotine gum with nicotine TD were evaluated in healthy
subjects (374) at their work-setting in a 1-year trial (94): 149 subjects received
nicotine TD plus nicotine gum; 150 nicotine TD plus placebo gum; and 75 placebo
TD plus placebo gum. Treatment duration was 12 weeks with a 16-h 15-mg TDS,
then 10 mg (6 weeks) and 5 mg (8 weeks). Gum use was unrestricted during the
first 6 months, with recommendation to use at least four pieces a day. It was con-
cluded that adding nicotine gum use to nicotine TD use in subjects smoking 10, or
more, cigarettes a day increased abstinence rates. Efficacy of TD nicotine as an
adjunct to advice and support in patients attending hospital was investigated (95) in
234 in- and outpatients with smoking-related respiratory or cardiovascular disease
(ages 18–75 years) advised to stop smoking. Advice was reinforced by a Smoking
Cessation Counselor initially and at 2, 4, 8, and 12 weeks, supplemented by a 24-h
TDS in adjusted doses over the study period. Patients, who no longer smoked at 12
weeks, were followed up at 26 and 52 weeks and then self-reported abstinence was
validated. Of patients receiving TD nicotine 24/115 (21%) were verified as non-
smokers at 12, 26, and 52 weeks, compared with 17 of 119 (14%) of the placebo
group. Cessation was related to increasing age and lower Fagerstrom score. Minor
skin reactions and nausea were more frequent in the nicotine group (47 vs. 34% and
12 vs. 3%,, respectively), but severe skin reactions were rare (about 5%).
Nicotine concentrations in gastric juice, saliva, and plasma were monitored
after Nicorette TD systems (15 mg/16 h) were applied to seven healthy volunteers
(96). Nicotine concentration was highest in gastric juice, followed by saliva then
plasma which suggested ion-trapping of nicotine base in acidic gastric juice.
The incremental cost-effectiveness (based on physician time and retail cost of
nicotine TD and benefits, based on quality-adjusted life years; QALYs), of addition
of nicotine TD to smoking cessation counseling was investigated in men and women
smokers (25–69 years of age) receiving primary care (97). TD use produced one
additional lifetime quitter at a cost of 7,332 dollars. Incremental cost-effectiveness
of nicotine TD by age group ranged from 4,390 to 10,943 dollars per QALY for
men and from 4,955 to 6,983 dollars per QALY for women. A clinical strategy
involving limiting prescription renewals to patients successfully abstaining for the
first 2 weeks improved cost-effectiveness by 25%. These data provided support for
routine use of nicotine TD as an adjunct to smoking cessation counseling and for
health insurance coverage of TD nicotine therapy. Metanalysis estimated cost-effec-
tiveness of nicotine TD as an adjunct to brief physician counseling during routine
office visits (98). Depending on age, average costs per year of life saved ranged
from 965 to 1,585 dollars for men and from 1,634 to 2,360 dollars for women.
Incremental costs per year of life saved ranged from 1,796 to 2,949 dollars for men
and from 3,040 to 4,391 dollars for women. It was concluded that the nicotine TD
is cost-effective and less costly per year of life saved than other widely accepted
medical practices and that physicians and third-party payers should recommend the
nicotine patch to patients who wish to stop smoking.
Assessment of the use and effectiveness of free nicotine TD among Medicaid

and uninsured smokers concluded that barriers of access to effective treatment for
smoking cessation need to be eliminated (99). The majority (90%) of participants
found the TD system useful, 14% were abstinent for 6 months or more, and there
was no support for inappropriate use. The cost-effectiveness, for the National Health
Service (NHS) in the United Kingdom, of allowing general practitioners to prescribe
TD nicotine patches for up to 12 weeks was evaluated (100) using data from a
randomized, placebo-controlled efficacy trial of nicotine TD and a survey of asso-
ciated resource use in 30 general practitioner’s (GP) surgeries in England. The health
benefit of TD nicotine treatment was calculated in the number of life years that
would be saved by stopping smoking at various ages, and used an abstinence–
contingent treatment model to calculate incremental cost per life year saved by GP
counseling with nicotine-patch treatment over GP counseling alone. If the NHS al-
lowed prescription of nicotine TD for up to 12 weeks, the incremental cost per life
year saved would be 398 pounds per person younger than 35 years; 345 pounds for
ages 35–44 years; 432 pounds for ages 45–54 years; and 785 pounds for ages 55–
65 years.
Cardiovascular effects of smoking, effects of nicotine without tobacco smoke,
and available data on cardiovascular risk during nicotine replacement therapy were
reviewed (101). Although nicotine gum and TD are approved for over-the-counter
sale, and smokers with cardiovascular disease are advised to seek physician coun-
seling before using nicotine products, information on product safety in such patients
is not readily available. Nicotine can contribute to cardiovascular disease, by he-
modynamic consequences of sympathetic neural stimulation and systemic catechol-
amine release, but there are many other potential cardiovascular toxins in cigarette
smoke. Doses of nicotine obtained by cigarette smoking usually exceed those deliv-
ered TD, and cardiovascular effects of nicotine are generally more intense when
delivered rapidly by cigarette smoking, rather than more slowly by TD nicotine or
nicotine gum. As the dose–cardiovascular response for nicotine is flat, the effects of
cigarette smoking in conjunction with replacement therapy are similar to those of
cigarette smoking alone. Although cigarette smoking increases blood coagulability,
a major cardiovascular risk factor, TD nicotine does not appear to.
Wide variations in levels of nicotine (and cotinine) have been observed after
nasal and transdermal delivery. Sources of individual variability in nicotine and co-
tinine plasma levels after use of replacement systems or cigarette smoking were
evaluated (102). Cigarette smokers received four treatments of 5-days duration each,
including (a) cigarette smoking (16 cigarettes a day); (b) TD nicotine (15 mg a day);
(c) nicotine nasal spray, (24⫻1-mg doses a day); (d) placebo nicotine nasal spray
(24 doses a day). Disposition kinetics were determined by infusion of deuterium-
labeled nicotine and cotinine. There was considerable individual variation in daily
dose of nicotine absorbed (nasal spray ⫻ 5, TD ⫻ 2–3) and in plasma nicotine and
cotinine levels. Plasma nicotine levels were determined predominantly by nicotine
clearance, whereas cotinine levels were determined most strongly by nicotine dose
and, to a lesser extent, by clearance of cotinine and fractional conversion of nicotine
to cotinine. To compensate for individual differences in clearance, individualization
of dosing based on therapeutic monitoring and comparison with nicotine or cotinine
levels during cigarette smoking before treatment may be required to optimize nicotine
therapy.
Determinants of variability and the utility of baseline plasma concentrations as

predictors of concentrations during TD treatment was evaluated using data from
smoking cessation (n = 466) and pharmacokinetic studies (n = 12) (103). Plasma
concentrations of nicotine and cotinine were highly variable. Indirect estimates of
plasma clearance (baseline plasma concentration per number of cigarettes per day)
together with other factors accounted for 33% or less, variability during TD treatment
in the smoking cessation study. In contrast, 75–99% was accounted for by direct
measurements of plasma clearances and systemic doses of nicotine in the pharma-
cokinetic study. It was concluded that plasma concentrations of nicotine and cotinine
during TD nicotine treatment are poorly predicted by clinical history or baseline
plasma concentrations.
The frequency of adverse effects associated with use of TD nicotine was es-
timated by metanalysis of data from 47 reports of 35 clinical trials (104). Few adverse
cardiovascular outcomes were reported, and no excess of these outcomes was de-
tected among patients assigned to TD use. Incidence of minor adverse effects (es-
pecially sleep disturbances, nausea or vomiting, localized skin irritation, and respi-
ratory symptoms) was clearly elevated among TD groups. Incidence of nausea or
vomiting appeared lowest when the patch dose was tapered.
Application of urine and serum nicotine and cotinine excretion rates for as-
sessment of nicotine replacement in light, moderate, and heavy smokers undergoing
TD therapy was investigated (105,106). Subjects were stratified as light (10–15 cig-
arettes per day), moderate (16–30 cigarettes per day), or heavy (more than 30 cig-
arettes per day) smokers and assigned to a daily 24-h–TD system delivering a dose
of 0, 11, 22, or 44 mg. Steady-state urinary excretion rates of nicotine and cotinine
were attained in 2 and 3 days, respectively, at all doses, independent of smoking
rate. Significant underreplacement occurred with the 11-mg/day dose, particularly in
moderate and heavy smokers (<50%), and at 22 mg/day, nicotine replacement was
still less than 100% in most subjects. Only a dose of 44 mg/day provided mean
replacement exceeding 100%, regardless of baseline smoking rate. Steady-state
plasma concentrations of nicotine and cotinine were attained in 1 and 3 days, re-
spectively, at all doses, independent of baseline smoking rate.
Individualization of dosage is desirable and plasma cotinine levels at steady
state (>3 days of therapy) can be used to calculate percentage replacement using
baseline levels. Attempts were made to identify variables associated with long-term
smoking cessation following hospitalization (107). Patients were assigned to (a) min-
imal care (MC; brief physician-delivered motivational message); (b) counseling plus
active nicotine patch (CAP; motivational message, 6-week supply of nicotine TD,
and extended bedside and telephone counseling); and (c) counseling plus placebo
patch (CPP). At 6 months, abstinence rates were 4.9, 6.5, and 9.7% for MC, CPP,
and CAP treatments, respectively, but not significantly different. In another trial (108)
308 smokers were randomly allocated to (a) 3-g dextrose tablets and 15-mg nicotine
TD; (b) dextrose and placebo patch; (c) placebo tablets and nicotine TD; (d) placebo
tablets and placebo patch. The proportion of smokers abstinent in each group was
49% (dextrose plus active patch); 44% (dextrose plus placebo patch); 36% (placebo
tablet plus active patch); and 30% (placebo tablet plus placebo patch). The difference
between dextrose and placebo tablets (13%) was significant, but that between active
and placebo patches (6%) was not.
Efficacy of using nicotine TD for five months in combination with a nicotine

nasal spray for 1 year was evaluated (109) in 237 smokers (22 to 66-years old)
receiving either nicotine TD (5 months) with nicotine nasal spray for 1 year (n =
118) or nicotine TD with placebo spray (n = 119). TD treatment comprised 15-mg
nicotine (months 1–3), 10-mg (month 4), and 5-mg (month 5). The combination of
use of nicotine TD for 5 months with a nicotine nasal spray for 1 year was a more
effective method of smoking cessation than use of a patch only. Sustained abstinence
rates for the patch and nasal-spray group and the patch-only group were 51 versus
35% at 6 weeks, 37 versus 25% at 3 months, 31 versus 16% at 6 months, 27 versus
11% at 12 months, and 16 versus 9% at 6 years.
Long-term smoking-cessation efficacy of varying doses of TD nicotine was
evaluated 4–5 years after quitting among patients enrolled in the Transdermal Nic-
otine Study Group investigation (110). Self-reported continuous quit rate for patients
assigned 21 mg (20.2%) was significantly higher than that for patients assigned 14
mg (10.4%), 7 mg (11.8%), or placebo (7.4%). Relapse rates among treatment con-
ditions were similar to those 1-year after cessation.
Plasma cotinine replacement levels of 56 outpatient smokers using a 21-mg/
day TDS (Nicoderm CQ) were reported (111). Cotinine replacement was 35–232%
(mean 107%; median 90.5%). Baseline cotinine level, previous quitting attempts,
gender, and Fagerstrom Tolerance Questionnaire scores were significantly correlated
with the percentage of cotinine replacement. Baseline cotinine level plus gender was
the most powerful predictor combination. Predictors and timing of adverse experi-
ences during TD nicotine therapy were investigated (112). Intervention consisted of
brief behavioral counseling, a booklet containing smoking cessation advice, instruc-
tions on patch use, and a 12-week course of decreasing TD nicotine. Most adverse
experiences were mild. Sleep problems occurred in 48% and usually started on the
day of smoking cessation. Application site reactions occurred in 34%, most fre-
quently after 6 days of therapy. Significant predictors of sleep problems were female
gender and successfully quitting smoking. Predictors of application site reactions
were psoriasis or eczema, other skin conditions, age younger than 40 years, female
gender, and trade or university education level. Substantially increased nicotine in-
take during therapy, compared with baseline smoking, occurred in 8% who smoked
concurrently, and 4% of those who did not. It was concluded that sleep disturbance
during therapy was primarily associated with tobacco withdrawal, rather than excess
nicotine from TD treatment. Comparison between nicotine gum, TD, nasal spray,
and inhaler was made with assessments at quit date and after 1, 4, and 12 weeks
(113). Products did not differ in their effects on withdrawal discomfort, urge to
smoke, or rate of abstinence. Continuous validated 12-week abstinence rates were
20% (gum), 21% (TD), 24% (spray), and 24% (inhaler). Compliance with recom-
mended treatment use was high for TD, low for gum, and very low for spray and
inhaler.
A quantitative gas chromatography–mass spectrometry (GC–MS) method for
simultaneous determination of total and free trans-3⬘-hydroxycotinine (THOC) and
cotinine (COT) in human urine during nicotine transdermal therapy was developed
(114). Results from six consecutive 24-h urine collections in 71 subjects, who used
daily TD nicotine doses of 11, 22, and 44 mg, showed that free THOC was 76% of
total THOC, and free COT was 48% of total COT. The effect of 24-h nicotine patches
in smoking cessation was evaluated among over-the-counter customers in Denmark
(115). A total of 522 customers who smoked ten or more cigarettes per day were
randomized to either nicotine or placebo patches; 24-h patches were offered for a 3-
month period. Those smoking 20 or more cigarettes per day started with patches of
21 mg/day. Those smoking less started with 14-mg/day patches, and all participants
were gradually reduced to 7-mg/day patches. There was a significant increase in
smoking cessation rates but, after 8 weeks of follow-up, only among smokers who
used 21-mg/day patches. No significant differences in smoking cessation rates were
seen among smokers who started with low-dose nicotine or placebo patches. The
Collaborative European Antismoking Evaluation (CEASE) was a multicenter, ran-
domized, double-blind placebo-controlled smoking cessation study, the objectives of
which were to determine whether higher dosage and longer duration of nicotine patch
therapy increased the success rate (116). Thirty-six chest clinics enrolled 3575 smok-
ers, who were allocated placebo or either standard- or higher-dose nicotine TD (15
and 25 mg daily) each given for 8 or 22 weeks, with adjunctive moderately intensive
support. The 12-month sustained success rates were: 25-mg TD for 22 weeks, 15.4%;
25-mg TD for 8 weeks, 15.9%; 15-mg TD for 22 weeks, 13.7%; 15-mg TD for 8
weeks, 11.7%; and placebo, 9.9%.
Nicotine plasma levels and safety of nicotine TD in smokers undergoing situ-
ations suspected to result in increased nicotine plasma levels were assessed (117).
Effects of increasing nicotine intake through sequential administration of nicotine
TD (day 2), TD followed by consumption of nicotine gum (day 3), and TD followed
by gum consumption and cigarette smoking (day 4) were examined; nicotine plasma
levels increased transiently after addition of each nicotine source. Mean AUCs (0–
24 h) for nicotine were 453 ⫾ 120 ng h⫺1 mL⫺1 (day 2); 489 ⫾ 143 ng h⫺1 mL⫺1
(day 3); and 485 ⫾ 143 ng h⫺1 mL⫺1 (day 4). A second study evaluated effects of
physical exercise on kinetics and safety of two types of nicotine transdermal device.
Mean delivered dose was higher with Nicoderm than Habitrol, and the products were
not bioequivalent. During a 20-min–exercise period, nicotine plasma levels increased
by 13 ⫾ 9% for Nicoderm and 30 ⫾ 20% for Habitrol. After exercise, subjects
taking Habitrol had a higher incidence of adverse events compared with baseline
values, but safety profiles remained acceptable. It was concluded that both super-
imposed nicotine sources and physical exertion resulted in short-lived plasma nico-
tine elevations and temporarily increased nicotine pharmacodynamic parameters, but
without increased risk.
Short-term effects of TD nicotine replacement in pregnancy were examined
(118). After customary smoking cessation efforts had failed, six prenatal patients (28
to 37-weeks gestation) who smoked one to two packs per day were admitted for a
period of 21 h. Maternal and fetal assessments, including vital signs, biophysical
profile, and electronic fetal monitoring, amniotic fluid index, and umbilical artery
Doppler examinations were made, and salivary levels of cotinine and nicotine levels
were determined. There were no measurable differences in fetal or maternal well-
being. During TD use salivary nicotine levels increased to 19.0 ⫾ 13.5 ␮g/L at 480
min (consistent with levels in nonpregnant adults).
Surprisingly, salivary cotinine concentrations were much lower (⬃50 ␮g/L)
than those in smoking nonpregnant adults and varied little over the period that the
patch was worn. Weight changes in subjects receiving variable doses of TD nicotine
replacement were assessed in 70 subjects receiving placebo or to 11-, 22-, or 44-mg/
day doses of TD nicotine and 1 week inpatient treatment with outpatient follow-up
through 1 year (119). The study included 1 week of intensive inpatient treatment
with active TD therapy for a further 7 weeks. Counseling sessions were provided
weekly during patch therapy, with long-term follow-up at 3, 6, 9, and 12 months.
Forty-two subjects were confirmed as nonsmokers at all weekly visits during TD
therapy, and their 8-week weight change from baseline (3.0 ⫾ 2.0 kg) was negatively
correlated with the percentage of cotinine replacement (r = ⫺0.38, p = 0.012) and
positively correlated with baseline weight and age. Men had higher 8-week weight
gain (4.0 ⫾ 1.8 kg) than women (2.1 ⫾ 1.7 kg). This suggested that higher replace-
ment levels of nicotine may delay postcessation weight gain in both men and women,
but did not identify predictive factors.
TD nicotine use, nicotine and cotinine levels, and fetal effects were investigated
in pregnant cigarette smokers aged 18 years or older, whose fetuses were beyond
24-weeks gestational age (120). Serial measurements of mother and fetus were made
at baseline while the mother was smoking, while abstaining from smoking, and while
using TD nicotine therapy for 4 days in the hospital. Nonpregnant women smokers
of similar age were used as comparators. No evidence of fetal compromise was seen
while nicotine patch therapy was administered. Morning serum cotinine levels were
significantly higher in nonpregnant than in pregnant subjects, but afternoon levels
were not significantly different. Steady-state urinary levels of nicotine and cotinine
were also not significantly different between pregnant versus nonpregnant patients.
On inpatient days 2, 3, and 4 for women not smoking, but wearing nicotine TD, the
morning fetal heart rates were significantly reduced relative to baseline when subjects
were smoking.
Abuse liability and dependence potential of nicotine gum, TD, spray, and in-
haler were compared in 504 male and female smokers (121). No significant differ-
ences between products in terms of satisfaction or subjective dependence, except at
week 15 when no patch users rated themselves as dependent. Continued use of
nicotine replacement at week 15 was related to rate of delivery of nicotine: 2% for
patch, 7% for gum and inhaler, and 10% for spray. Cessation of nicotine replacement
between weeks 12 and 15 was not accompanied by withdrawal discomfort or in-
creased frequency of urges to smoke. It was concluded that abuse liability was low
for all products.
2. Treatment of Ulcerative Colitis
Ulcerative colitis is largely a disease of nonsmokers and anecdotal reports have
suggested that smoking and nicotine may improve symptoms (122). Patients with
active ulcerative colitis were treated with either nicotine TD or placebo patches for
6 weeks. All patients had been taking mesalamine, and some were also receiving
low doses of glucocorticoids. These medications were continued during the study.
Incremental doses of nicotine were used and most patients tolerated 15–25 mg/24
h: 17 of 35 patients in the nicotine group had complete remissions, compared with
9 of 37 patients in the placebo group. Patients in the nicotine group had greater
improvement in global clinical and histological grades of colitis, lower stool fre-
quency, less abdominal pain, and less fecal urgency. More of the nicotine group had
minor side effects (23 vs. 11 in placebo group), and withdrawals owing to ineffective
therapy were more common in the placebo group (3 vs. 8).
The value of TD nicotine for maintenance of remission was studied in 80
patients with ulcerative colitis in remission, using either TD nicotine or placebo
patches for 6 months (123). Incremental doses were given for 3 weeks to achieve a
maintenance dose (most tolerated 15 mg for 16 h daily). All patients were taking
mesalamine at study entry, but this was stopped when maintenance nicotine doses
were achieved. Twenty-two patients in the nicotine group were prematurely with-
drawn from the study, 14 because of relapse and 8 for other reasons, including side
effects and protocol violations. In the placebo group, 20 patients were withdrawn
prematurely, 17 owing to relapse and 3 for other reasons. Among patients using 15-
mg–nicotine patches, serum nicotine and cotinine concentrations were lower than
expected, which may have reflected poor compliance. Side effects were reported by
35 patients, 21 in the nicotine group and 14 in the placebo group. TD nicotine alone
was no better than placebo in maintaining remission of ulcerative colitis, and early
withdrawal because of side effects was more common in the nicotine group.
Nicotine alone was compared with prednisolone in 61 patients with active
ulcerative colitis treated with either nicotine TD or 15 mg of prednisolone for 6
weeks (124). Incremental nicotine doses were given for the first 9 days. Of the 43
patients who completed the trial, 6 of 19 in the nicotine group achieved full sig-
moidoscopic remission, compared with 14 of 24 with prednisolone. In those who
completed this study, nicotine alone appeared to be of only very modest benefit in
acute colitis.
Use of TD nicotine in mildly to moderately active ulcerative colitis was in-
vestigated (125) in 64 nonsmoking patients with mildly to moderately active ulcer-
ative colitis despite the use of medication. These were stratified (on the basis of
smoking history, extent of disease, and concomitant therapy) and assigned to daily
treatment with TD nicotine (n = 31) at highest-tolerated dose (11 mg for 1 week and
then ⱕ22 mg for 3 weeks) or placebo (n = 33). At 4 weeks, 39% of those who
received nicotine showed clinical improvement compared with 9% who received
placebo. Four patients receiving nicotine discontinued therapy because of side ef-
fects. At week 4, the nicotine group had trough serum concentrations of 12.3 ⫾ 8.4
ng/mL (nicotine) and 192 ⫾ 95 ng/mL (cotinine). It was concluded that transdermal
nicotine at ⱕ22 mg/d for 4 weeks was effective in controlling the clinical manifes-
tations of mildly to moderately active ulcerative colitis.
A pilot trial of nicotine TD as an alternative to corticosteroids in ulcerative
colitis was reported (126). In ten patients with mild-to-moderate clinical relapses of
ulcerative colitis during mesalamine treatment and with a previous history of poorly
tolerated steroids, TD nicotine (15 mg daily) was added for 4 weeks. Clinical re-
mission was achieved in seven patients and persisted for up to 3 months after nicotine
withdrawal.
A second study (127) investigated long-term effects. Patients with mild-to-
moderate clinical relapses of left-sided ulcerative colitis during maintenance treat-
ment with mesalamine were allocated additional treatment with either TD nicotine
or prednisone for 5 weeks. The first consecutive 15 patients in each group with
clinical and endoscopic signs of remission were followed-up for 6 months, while
continuing mesalamine maintenance treatment. Relapses of active colitis were ob-
served in 20% of patients formerly treated with nicotine and 60% of patients in the
prednisone group, and relapses occurred earlier in the latter group. As patients with
mild-to-moderate active colitis treated with mesalamine plus TD nicotine appeared
to suffer fewer relapses than patients treated with mesalamine plus oral prednisone
a long-term follow-up was carried out (128). Thirty patients with remission of distal
colitis were monitored for 12 months and relapsed patients retreated in a crossover
manner. Recurrences were observed in 14 of 15 patients initially treated with steroids
and in 7 of 15 subjects who received TD nicotine.
3. Other Indications
Individuals with major depression have a high frequency of cigarette smoking, and
TD nicotine can produce short-term improvement in mood. The effects of nicotine
patches (17.5 mg) on 12 nonmedicated outpatients with major depression were stud-
ied over 4 continuous days (129). Two patients dropped out of the study because of
nausea and vomiting. There was significant improvement in depression after day 2
of TD nicotine and patients relapsed 3 or 4 days after the final nicotine dose. Al-
though nicotine TDS produced short-term improvement of depression, with minor
side effects, nicotine TD was not recommended for clinical use in depression because
of the high health risks of nicotine. It was concluded that analogues might be de-
veloped that can improve depression without major risks.
The therapeutic response to nicotine TD was investigated in patients with Tou-
rette’s syndrome (130). Twenty patients (17 children and adolescents, 3 adults) were
studied following application of two patches (2 ⫻ 7 mg/24 h). There was a broad
range in individual response, but each patch application produced a significant re-
duction in the Yale Global Tic Severity Scale scores, for an average duration of
approximately 1–2 weeks. This suggested that TD nicotine could be an effective
adjunct to neuroleptic therapy of Tourette’s syndrome. Nicotine gum and nicotine
TD were used to reduce motor and vocal tics of children (age 8 years or older;
weight ⱖ25 kg), adolescents, and adults (131). Reduction of tics was seen during
chewing of nicotine gum, but improvement lasted no longer than 1 h after chewing.
With nicotine TD, motor and vocal tics were reduced 45% over baseline in 85% of
35 subjects within 30 min to 3 h after patch application. Relief of symptoms with a
single 7-mg patch, left on the skin for 24 h, persisted for variable periods up to 120
days. Application of a second patch for 24 h when symptoms returned resulted in
similar reduction in tic severity and frequency, which persisted an average of 13 ⫾
3 days.
Short-term nicotine injections have improved attentional performance in pa-
tients with Alzheimer’s disease (AD), but little is known about prolonged effects of
nicotine. A study evaluated clinical and neuropsychological effects of extended TD
nicotine application in AD subjects over a 4-week period (132). Patients were treated
with nicotine TD (Nicotrol) for 16 h/day at the following doses: 5 mg/day (week 1),
10 mg/day (weeks 2 and 3), and 5 mg/day (week 4). Nicotine significantly improved
attentional performance, with a significant reduction in errors of omission, which
continued throughout nicotine administration, and variability of reaction time for
correct responses was also significantly reduced. Nicotine did not improve perfor-
mance on other tests measuring motor and memory function.
4. Poisoning with Nicotine Patches
To evaluate potential adverse effects from inadvertent exposure, three marketed TD
nicotine products: Nicoderm (drug reservoir and rate-controlling membrane); Nico-
tinell (nicotine solution dispersed in cotton gauze between layers of adhesive); and
Niconil (nicotine in gel matrix), were administered topically and orally to dogs (133).
Topical nicotine doses were 1–2 mg/kg 24 h⫺1 for all products, with plasma con-
centrations 43 ng/mL, or less. Of 12 topical exposures (with Nicotinell and Niconil)

2 were associated with clinical signs (excess salivation or emesis). Oral doses (2.8
mg/kg [one patch] to 13.4 mg/kg [two patches] over 25–57 h), were two- to ninefold
higher than the oral doses reported to produce severe toxicity in children, and the
higher dose was within the known lethal range for dogs. Oral dosing of Nicotinell
and Niconil (two patches per dog) produced vomiting in 2 of 12 exposures. No
clinical signs were observed with either topical or oral dosing of Nicoderm. Char-
acteristics and outcomes of U.S. poisoning cases, involving dermal human applica-
tion of multiple nicotine TD systems, were evaluated (134). Nine cases of dermal
exposure to 2–20 nicotine TD systems resulted from intentional misuse or suicide
attempts and included concomitant exposure to other drugs in 7 of 9 cases. Mean
age was 45 years, and 7 of 9 patients were female. All suffered medical complica-
tions, including seizures, other CNS changes, cardiovascular effects, and respiratory
failure, but plasma nicotine–cotinine levels did not correlate with the severity of
illness. Eight patients were hospitalized, but all recovered.
B. Glyceryl Trinitrate and Related Drugs

Measurement of plasma concentrations of glyceryl trinitrate (GTN) is very difficult
owing to the unusual pharmacokinetics, with very rapid disappearance from plasma,
and large intraindividual and interindividual variations (135). There is extensive first-
pass hepatic extraction after oral administration and plasma levels are often unde-
tectable. With controlled-release TD GTN systems, plasma concentrations can be
maintained over 24 h, but with fluctuations and important intra- and interindividual
variability. After administration of GTN by any route, glyceryl dinitrates and mono-
nitrates are found in plasma. Bioavailability and main pharmacokinetic parameters
of the GTN metabolites 1,2- and 1,3-glyceryl dinitrate (1,2-GDN and 1,3-GDN) were
determined following application of two types of GTN TDS in healthy volunteers
(136). For 1,2-GDN, AUCs(0-Tlast) were 23.77 h ng⫺1 mL⫺1 (patch A) and 27.83 h
ng⫺1 mL⫺1 (patch B). Peak plasma levels were 2.45 ng/mL (at 6.4 h) and 2.93 ng/
mL (at 8.3 h), respectively. For 1,3-GDN AUCs(0-Tlast) were 3.32 h ng⫺1 mL⫺1 (patch
A) and 3.81 h, ng⫺1 mL⫺1 (patch B). Peak plasma levels were 0.35 ng/mL (at 6.4 h)
and 0.41 ng/mL (at 7.9 h), respectively. Statistical comparison showed bioequiva-
lence between these TD systems for the metabolites investigated. Typical side effects
observed after nitrate therapy also occurred.
The efficacy of lisinopril, TD GTN, and their combination in improving sur-
vival and ventricular function after acute myocardial infarction (AMI) was assessed
(137). A total of 19,394 patients were randomized from 200 coronary care units in
Italy and assigned 6 weeks of oral lisinopril (5-mg initial dose and then 10 mg daily),
or open control, as well as nitrates (intravenous for the first 24 h followed by TD
GTN 10 mg/day) or open control. Lisinopril, started within 24 h after AMI symptoms
began, produced significant reductions in overall mortality and in the combined out-
come measure of mortality and severe ventricular dysfunction. Administration of TD
GTN did not show any independent effect on the same outcome measures. Systematic
combined administration of lisinopril and GTN produced significant reductions in
overall mortality and in the combined endpoint. The effects of a GTN TDS on
platelet aggregation was examined in eight normal volunteers (138). A significant
effect of TD GTN on platelet aggregation was demonstrated in the presence and
absence of iloprost, but the clinical significance of the antiplatelet effect of TD GTN
remained unknown.
The anti-inflammatory and analgesic effects of TD GTN was studied in 21
patients with mild to moderate leg varicose veins who underwent vein sclerotherapy
in both legs (139). The vein in one leg was treated every 8 h with GTN and compared
with a placebo ointment applied to the vein of the other leg. Inflammation signs
were observed in all cases 15 min after first application. Intensity of inflammation
signs were 26% in GTN-treated veins and 61.5% in placebo-treated veins. One hour
later only 63% of cases in the GTN group, but all cases in the placebo group, showed
signs of thrombophlebitis. All veins in the GTN group were free of signs of throm-
bophlebitis in fewer than 48 h, whereas, of the placebo group, 45% required more
than 48 h.
Intermittent TD GTN therapy with a 10- to 12-h–patch-free period each day
has documented clinical benefits. The antianginal and anti-ischemic effects of three
dose levels of TD GTN applied for 12 h daily for 30 days and the development of
tolerance and rebound were assessed (140). There was a significant increase in tread-
mill walking time to moderate angina in each GTN patch group, compared with
placebo, at time points up to 12 h throughout the 30-day period. Secondary efficacy
parameters supported the primary efficacy results,and there was no evidence of tol-
erance or rebound. Transdermal GTN is widely used to treat angina pectoris, but
development of tolerance is a major problem (141). The effects of short (5 h) and
prolonged (3 days) exposure to transdermal GTN patches on the development of
tolerance in terms of hemodynamics and vascular reactivity in the conscious rabbit
were, therefore, investigated. It was concluded that in the rabbit, prolonged exposure
to clinical GTN patches caused hemodynamic compensation and baroreflex resetting,
but no evidence of vascular reactivity tolerance.
The efficacy of adding transdermal GTN or oral N-acetylcysteine, or both, to
conventional medical therapy was examined (142) in a trial of 200 patients with
unstable angina, followed-up for 4 months. Death, myocardial infarction, or refrac-
tory angina requiring revascularization occurred in 31% of patients receiving GTN,
42% of those receiving N-acetylcysteine, 13% of those receiving GTN plus N-ace-
tylcysteine, and 39% of those receiving placebo. There was higher probability of no
treatment failure when receiving both GTN and N-acetylcysteine than with placebo,
N-acetylcysteine, or GTN alone. However, combination of GTN and N-acetylcysteine
was associated with a high incidence of side effects (35%), mainly intolerable
headache.
The relation between tolerance development, counterregulatory responses, and
arterial vasodilating effects were studied in 20 patients with stable angina pectoris
who were exercise tested before, after 2 h, and 24 h of nitrate patch treatment (143).
Effects observed after 2 h of treatment on exercise duration, ST-segment depression,
blood pressure, and heart rate were usually lost by 24 h, although effects on arterial
pulse curves persisted after 24 h, with a mean change from baseline of 29%, com-
pared with 33% at 2 h. After 24 h, a significant decrease in hematocrit and an increase
in body weight were observed. Hematocrit changes correlated with loss of clinical
efficacy. It was concluded that clinical nitrate tolerance may be observed despite
maintenance of arterial vasodilating effects, and that tolerance is more related to
plasma volume expansion as a counterregulatory mechanism.
A subsequent study (144) examined whether GTN TDS treatment for 24 h

could induce local cutaneous changes that impaired drug delivery and clinical effi-
cacy. Twenty angina patients were exercise-tested after 2 and 24 h of treatment and
2 h after device renewal. The TDS was either renewed at a new skin location or on
the previous application site. Clinical efficacy, the effect seen on plethysmography,
and GTN plasma concentrations, all tended to increase after TDS renewal, regardless
of the application site indicating that cutaneous changes of clinical importance were
not demonstrated. The effect of GTN in patients with chronic heart failure (CHF)
treated with angiotensin-converting enzyme (ACE) inhibition was evaluated (145).
High-dose (50–100 mg) TD GTN or placebo were given daily for 12 h in 29 patients
with CHF. Exercise time (4 h after patch application) showed progressive improve-
ment during GTN administration. GTN decreased left ventricular end-diastolic and
end-systolic dimensions and augmented LV fractional shortening.
Clinical efficacy of, and patient tolerance to, sustained-release GTN TDS in
the treatment of Raynaud’s phenomenon was reported (146). Patients had primary
Raynaud’s disease or Raynaud’s phenomenon secondary to systemic sclerosis. GTN
TD (0.2 mg/h) was effective in reducing both number and severity of Raynaud’s
attacks, but headaches led to withdrawal of 8 patients and occurred in about 80% of
remaining patients. A clinical study in 20 patients with shoulder pain syndrome
caused by supraspinatus tendinitis was conducted to determine whether TD GTN has
analgesic action in this condition (147). One 5-mg GTN (Nitroplast) patch (or pla-
cebo patch) was applied per day over 3 days in the most painful area. Follow-up
showed a significant decrease in intensity of pain at 24 and 48 h in the GTN group,
but no change in the placebo group. It was concluded that GTN could be a useful
approach to management of this common condition and other tendon musculoskeletal
disorders.
The 24-h effect of TD GTN on splanchnic hemodynamics in nine patients with
biopsy-proved liver cirrhosis was evaluated (148). GTN tape, capable of releasing
15 mg of drug in 24 h, was applied to chest skin at 7 AM of the second day. After
GTN application, the mean portal blood velocity and flow significantly decreased by
18 and 22%, whereas superior mesenteric artery velocity decreased and resistance
indices increased. This indicated that GTN, from a transdermal long-acting system,
significantly influenced portal hemodynamics in liver cirrhosis, and the use of GTN
was proposed for long-term clinical studies to test efficacy in preventing gastroin-
testinal bleeding.
The role of TD GTN, as a source of exogenous nitric oxide, in the management
of primary dysmenorrhea was investigated (149) in a multinational study. Eighty-
eight patients from six countries were evaluated during three menstrual cycles while
receiving GTN (0.1 mg/h) or placebo patches. The data indicated that TD GTN, as
a source of exogenous nitric oxide, was useful as a modulator of uterine contractility
and represented a new and mechanistically different therapeutic alternative for man-
agement of primary dysmenorrhea.
The maternal and fetal cardiovascular effects of TD GTN compared with ri-
todrine for acute tocolysis (150) were studied in 60 women in preterm labor. At
doses required for acute tocolysis, TD GTN had minimal effects on maternal pulse,
blood pressure (BP), or fetal heart rate, and significantly fewer adverse cardiovascular
effects than intravenous ritodrine. It was concluded that TD GTN may be a safer
treatment for women in preterm labor.
The efficacy of TD GTN and intravenous ritodrine as tocolytics was also eval-
uated in an international study (151). A total of 245 women with preterm labor and
intact membranes between 24- and 36-weeks gestation were randomized to TD GTN
(10 to 20 mg patch) or intravenous ritodrine. GTN and ritodrine prolonged gestation
by 74% to 37 weeks. There was no significant difference in the proportion of women
receiving GTN or ritodrine who delivered within the specified days from study entry
or weeks of gestation, and no serious maternal side effects were reported for either.
It was concluded that there was no overall difference between GTN and ritodrine in
the acute tocolysis of preterm labor, but there was a suggested advantage of GTN
over ritodrine in reducing preterm delivery rate. Maternal side effect profile and
treatment discontinuation rates were fewer for GTN, suggesting it was the safer
alternative.
The nitric oxide (NO) donor morpholinosydnonmine has been reported to in-
hibit insulin release in isolated pancreatic islets; accordingly, the effect of TD GTN,
an alternative NO donor, on glucose-stimulated insulin release was studied in healthy,
young, male volunteers (152). Oral glucose tolerance tests were performed in the
presence of placebo or TD GTN (⬃0.4 mg/h of GTN) in the same patients, with a
2-week intertest interval. Glucose-stimulated maximum increases in plasma insulin
immunoreactivity were 36.3 ⫾ 5 and 78.8 ⫾ 6.1 mU/mL in the presence of active
and placebo patches, respectively, although both fasting and postload blood glucose
levels were equal. Active patches significantly decreased blood pressure with a mar-
ginal increase in heart rate. It was concluded that inhibition of glucose-stimulated
insulin release by TD GTN without causing hyperglycemia may be a novel com-
ponent of the antianginal action mechanism of nitrates.
Benzoxazinones are a potent new class of organic nitrates used in cardiovas-
cular therapy that have a coronary vascular selectivity greater than that of GTN and
isosorbide dinitrate. The ability of these new derivatives to reach therapeutic steady-
state plasma concentrations after TD administration was investigated in vitro using
human skin (153). Two members of this class: sinitrodil (ITF 296) and ITF 1129
were compared with GTN, isosorbide dinitrate, and nicorandil at two concentrations
(0.08% w/v and saturated solutions). Sinitrodil was considered a good candidate for
transdermal administration.
C. Estrogens
Numerous clinical studies have compared the effects of TD and alternative delivery
strategies for steroid hormones on a range of factors and, also, comparisons have
been made between reservoir and matrix type TDS, as well as topical gels. The most
common use of TD steroids is in hormone replacement therapy (HRT) in women.
The reduction in estrogen production in menopause may cause hot flashes, sweating,
mood and sleep disturbances, fatigue, and urogenital dysfunction. The effectiveness
of estrogen-based HRT in ameliorating these symptoms, and in preventing long-term
effects, such as osteoporosis, is well established (154). Comparative trials indicated
that 625 ␮g of oral, conjugated estrogens, 20 ␮g oral ethinyl estradiol and 50 ␮g
TD estradiol had equivalent efficacy in relief of mild-to-moderate menopausal symp-
toms and prevention of bone mineral loss. Concomitant progestogen therapy is usu-
ally included, if the uterus is intact, to protect against endometrial hyperplasia and
carcinoma. Addition of progestogen maintains and may enhance bone-conserving
effects of estrogen, and continuous regimens appear to reduce incidence of irregular

menses. Adverse reactions are predominantly local skin irritation with TDS (14%)
and systemic effects common to most forms of HRT, including breast tenderness,
flushing, headache, and irregular bleeding, occurring in 2% or fewer of patients.
Cost–benefit and cost–effectiveness of HRT in treatment of menopausal symptoms
suggested that conjugated estrogens and TD estradiol compared well with alternative
therapies, such as veralipride and Chinese medicines.
A combined TDS estradiol–norethisterone system was designed to deliver both
estradiol and norethisterone at a constant rate for up to 4 days (155). TD norethis-
terone did not appear to alter the potentially beneficial effects of TD estradiol on
total cholesterol, low-density lipoprotein (LDL) or triglyceride levels, or metabolic
parameters of bone resorption or vaginal cytology. Protection from the effects of
unopposed estradiol was achieved by sequential treatment with TD estradiol–nor-
ethisterone for 2 weeks of each 28-day cycle, and most patients experienced a regular
vaginal bleeding pattern with this regimen. Menopausal symptoms were improved
to a similar extent during estradiol-only and combined estradiol–norethisterone
phases. The system was well accepted in clinical trials and generally well tolerated,
the most common adverse effect being local irritation. Replacement therapy in 12
amenorrheic adolescents with gonadal dysgenesis treated with TD estradiol 100 ␮g
(Estraderm TTS-100) twice weekly for 3 weeks, plus medroxyprogesterone acetate
(MPA; Provera) for the last 11 days, following an interval of 1 week, was reported
(156). No significant changes were recorded in FSH, LH, estradiol-17␤, and PRL
serum levels, but significant decreases of TC values and atheromatic indices 1 (TC/
HDL) and 2 (LDL/HDL), significant increase in apolipoproteins-A1, and beneficial
effect on bone mass were seen at the end of treatment.
The efficacy and acceptability of a TD norethisterone device was assessed for
6 months in 18 patients of confirmed menopausal status (157). Therapy was contin-
uous application of Estraderm TTS 50 for 28 days, with additional application of 2-
norethisterone acetate patches for the last 12 days, repeated for six cycles. There was
significant improvement in hot flashes and sweating, and withdrawal vaginal bleeding
was established at regular intervals in 9 of 15 who completed 6 months of therapy.
Histological examination of endometrial biopsies showed secretory activity in 56%
of samples after 6 months, but no evidence of hyperplasia, premalignant, or malig-
nant changes in the remaining biopsies. Efficacy and overall acceptability of 100-
and 200-␮g twice-weekly doses of Estraderm TTS in the treatment of severe PMS
were compared (158).
Women with severe PMS received Estraderm TTS continuously with either
esterone 10 mg or MPA 5 mg, from day 17 to day 26 of each cycle. There was no
difference in change in total-ESAmax between the Estraderm 100-␮g and 200-␮g
groups, but there was a greater dropout rate and incidence of side effects attributed
to estrogen in the higher-dosage group. Mean estradiol levels were 300 (100 ␮g)
and 573 (200 ␮g) pmol/L, and 100 ␮g suppressed midluteal progesterone from a
mean of 35.5 to 3.4. It was concluded that the lower-dose therapy was as effective
in reducing symptom levels in severe PMS and was better tolerated.
Efficacy and tolerability of Menorest 50 was compared with Estraderm TTS
50 in treatment of postmenopausal symptoms in 205 women with moderate to severe
vasomotor symptoms (159). After a 4-week treatment-free period, each woman re-
ceived a cyclic regimen (25 days of a 4-week cycle) of Menorest 50 (matrix-type)
or Estraderm TTS 50 (reservoir-type) twice weekly for 12 weeks. Oral progestin was
also given for 10 days each cycle. Significant reduction in hot flashes was observed
in each group compared with baseline. There were no significant differences in mean
plasma estradiol levels and mean estradiol/estrone ratio (>1.0) in both groups after
10 weeks. Menorest 50 showed better local tolerability than Estraderm TTS 50.
Effects of daily intrauterine release of 20 ␮g of levonorgestrel by an intrauterine
device on climacteric symptoms, bleeding pattern, and endometrial histological fea-
tures in postmenopausal women receiving transdermal estrogen replacement therapy
was evaluated in 40 parous postmenopausal women over a period of 1 year (160).
Twenty women receiving a continuous TD daily dose of 50 ␮g of estradiol had a
levonorgestrel-releasing intrauterine contraceptive device inserted, whereas the con-
trol group (n = 20) received a continuous oral dose of 2 mg of estradiol valerate and
1 mg of norethisterone acetate daily. Both treatment regimens effectively relieved
climacteric symptoms.
The effect of TD estrogen replacement therapy on lipoprotein (Lp␣) and other
plasma lipoproteins was studied (161) in 30 women who had undergone a total
abdominal hysterectomy and bilateral salpingo-oophorectomy for benign gynecolog-
ical conditions treated with 1.5 mg of 17␤-estradiol gel applied daily for 12 consec-
utive months. Plasma lipoproteins were measured before treatment and at 6- and 12-
month intervals. There was significant reduction in Lp␣ levels during the first 6
months of treatment, with median values falling from 7.87 to 6.16 mg/dL, but during
the second 6 months, median concentration increased to 9.38 mg/dL. Significant
reductions in apoprotein A-I, apoprotein B, HDL-C, and HDL(3)-C were present
after 6 months, but at study completion these values were no different than those at
baseline. By avoiding the ‘‘first-pass’’ effect, this method of delivery did not appear
to produce the sustained changes in lipoproteins seen with oral treatment. After
menopause the hemostatic balance shifts toward a latent hypercoagulable state and
the effects on hemostasis of HRT with TD estradiol and oral sequential MPA were
evaluated (162). The balance between procoagulant factors and inhibitors were stud-
ied in 255 women in physiological menopause for 1–5 years, allocated to 1 year of
treatment with cyclic TD E2 (50 ␮g/day for 21 days) plus MPA (10 mg/day from
days 10 to 21), continuous TD E2 (50 ␮g/day for 28 days) plus MPA (10 mg/day
from days 14 to 25), or placebo. Continuous treatment gave significantly lower final
values of fibrinogen, factor VII, antithrombin III, protein S, and heparin cofactor II
than placebo.
Effects of HRT on bone mineral density (BMD) and disease activity in post-
menopausal women with rheumatoid arthritis (RA) were investigated in 62 patients
with RA, 22 taking placebo and 40 receiving HRT (TD estradiol patches twice
weekly for 48 weeks plus norithisterone tablets when clinically indicated) (163).
Fifty-nine percent of placebo and 78% of HRT groups completed 48 weeks. At entry,
BMD values in the lumbar spine and femoral neck were similar to those in matched
controls, whereas at the distal radius, BMD was significantly reduced to about 50%
of control values. In the HRT group, spinal BMD increased significantly by ⫹0.94%
at 48 weeks, but BMD at femoral neck and distal radius did not change in either
group. In the HRT group, there was significant improvement in well-being and ar-
ticular index. Because older women often experience side effects with conventional
HRT, a low-dose preparation (Estraderm 25) was compared with conventional HRT
(Estraderm 50) in patients with bone loss (164). A total of 196 women were studied
over 1 or 2 years, with 80 reaching 3 years of treatment. In the lumbar spine, BMD
increased maximally in year 1 in all groups, and the gain was maintained after 3
years. Only 3.9% of patients were nonresponders at this site after 3 years. Mean
changes after 3 years were 8.1 ⫾ 6.8% for Estraderm 25 and 9.0 ⫾ 8.3% for Es-
traderm 50. At the femoral neck, 10.4% of patients were nonresponders after 3 years,
and changes were significant only in Estraderm 25 in women older than 67 years,
and Estraderm 50 in those younger than 67 years. BMD change over 3 years at the
lumbar spine and femoral neck correlated with menopausal age. Use of Estraderm
50 was not associated with a greater response of bone mass and there was no evi-
dence of increasing BMD response as estradiol dosage per kilogram of body weight
increased. It was reported that oral and transdermal 17␤-estradiol provided similar
benefits in clinical studies (165). The lowest effective doses were 0.625 mg/day for
conjugated estrogens, 2 mg/day for oral 17␤-estradiol, 1.5 ␮g/day for 17␤-estradiol
gel, and 50 g/day of 17␤-estradiol TDS.
Decrease in incidence of osteoporotic fractures was achieved only when the
duration of HRT exceeded 7 years. The responses of various biochemical markers
for bone turnover to TD estradiol were measured in 11 postmenopausal women over
24 weeks (166) and compared with the within-subject variability of markers in 11
untreated healthy postmenopausal women. Mean decrease in markers of bone for-
mation ranged from 19% for procollagen type I C-terminal propeptide to 40% for
procollagen type I N-terminal propeptide (PINP). The mean decrease in markers of
bone resorption ranged from 10% for tartrate-resistant acid phosphatase (TRAP), to
67% for C-terminal cross-linked telopeptide. The ability to detect a response differed
between markers and was not dependent on the magnitude of response to therapy.
The highest number of responders were found using PINP (9 of 11) and osteocalcin
(9 of 11), and free deoxypyridinoline (8 of 11) and total deoxypyridinoline (7 of 11).
Lumbar spine BMD defined four patients as responders.
The comparative effects on BMD in routine clinical practice use of tibolone
and estrogen (unopposed or combined with cyclic progestogen) in postmenopausal
women who had not previously received estrogen or other menopausal therapy were
assessed (167). BMD was measured in the spine and hip at 12-month intervals over
3 years in 82 postmenopausal women referred for climacteric therapy. Thirty-five
women received tibolone, 24 TD estradiol alone, and 12 conjugated equine estrogens
together with cyclic progestogen; 11 received no therapy other than calcium. Spinal
BMD increased significantly in those taking tibolone over 3 years. In those receiving
conjugated equine estrogens and cyclic progestogen, spinal BMD also increased sig-
nificantly over years 1 and 2, but not year 3. Although spinal BMD rose over 3 years
in women treated with TD estradiol alone, this was not significant. No significant
change in BMD of spine or hip was observed in the control group.
A significant difference in increase of spinal BMD between treatment groups
was observed at 2 years in favor of those taking tibolone or conjugated equine
estrogens, compared with TD estradiol. The most common side effect and reason for
discontinuation with Norplant use is bleeding disturbance and, therefore, a 6-week
application of a patch releasing 100 ␮g/day estradiol was investigated as a method
of reducing this problem (168). Of 98 Norplant users, 34 had normal, and 64 ab-
normal, bleeding patterns. Estradiol (33) or placebo (31) TDS were randomly used
to treat patients with abnormal bleeding. Although there was clinical improvement
in the estradiol group, this was not significant.
Changes in serum lipoproteins, apoproteins, and coagulation factors, induced

in postmenopausal women treated by the Gynaderm TDS (designed to deliver 50 ␮g
of 17␤-estradiol per day) were studied over 6 months in 53 hysterectomized, healthy,
postmenopausal women (169). One patch was applied twice weekly. There were no
significant changes in levels of total cholesterol, triglycerides, HDL, or LDL, a sig-
nificant rise in apolipoprotein A-I (apo A-I) level at 3 months was not sustained after
6 months, and there was a significant drop in apo A-II level after 6 months. Changes
in apo B and Lp␣ were not significant. There were significant falls in levels of
antithrombin III and protein S, and a significant rise in factor VII. Changes in levels
of fibrinogen and protein C were not significant. TD estradiol administration caused
minimal changes in lipoprotein metabolism and the statistically significant changes
in the thrombophilia profile parallel those observed with oral HRT.
A clinical study evaluated the effect of HRT on plasma lipoproteins and Lp␣
profile in 42 menopausal women with primary hypercholesterolemia (total cholesterol
>240 mg/dL) (170). Patients were randomly assigned to (a) TD estradiol, 50 ␮g ⫹
medroxyprogesterone, 10 mg/day for 12 days; (b) conjugated equine estrogens, 0.625
mg/day ⫹ MPA 10 mg/day for 12 days; (c) no treatment. Total cholesterol and LDL
cholesterol significantly decreased after 6 months in both treated groups in compar-
ison with untreated women, but HDL cholesterol and triglycerides showed minimal
changes.
Comparison was made between TD estradiol (0.05 mg/day) and oral noreth-
isterone acetate (2.5 mg/day) administered for 12 days every 2 or 3 months to patients
whose menopause had begun at least 4 years earlier (171). Study duration was two
long cycles in each group within 7–10 months. Efficacy (group E/NA = 94/92%),
and systemic tolerability (95/97%) were good and continuation of spaced-out treat-
ment accepted by 88%(E) and 87%(NA). Major skin reactions occurred in 7%(E)
and 4%(NA). Progestin-associated withdrawal bleedings occurred in 61% of patients;
mean duration 4.3 ⫾ 1.9/4.8 ⫾ 1.6 days, and breakthrough bleeding requiring
sonographic or histological work-up in 8%(E) and 13%(NA).
Effects of a HRT regimen of continuous estrogen and interrupted progestogen,
administered transdermally, on the endometria of 15 healthy postmenopausal women,
and the pattern of bleeding and relief of menopausal symptoms were investigated in
a volunteer pilot study of up to 6-months duration involving weekly application of
an estrogen-only TDS releasing 50 ␮g estradiol per day interspersed with a combined
estrogen and progestogen TDS releasing 50 ␮g estradiol and 250 ␮g norethisterone
acetate per day for 3 days (172). Transvaginal ultrasound measurements of endo-
metrial thickness and endometrial biopsies were performed in month 3 at the end of
both the estrogen-only phase of treatment and combined estrogen–progestogen
phase. Treatment provided relief of hot flashes and, by month 6, 71% women who
completed treatment had no vaginal bleeding. No endometrial hyperplasia or atypical
changes were observed in biopsies, and ultrasound measurements demonstrated a
thin endometrium. Reduced immunostaining for Ki67 was observed in endometrium
from the combined phase of treatment compared with estrogen-only phase, consistent
with progestogenic-antagonism of proliferation. Exposure to progestogen did not
suppress steroid receptors, as similar immunostaining was observed in both treatment
phases.
The tolerability, adhesion, and efficacy of the matrix-type estradiol TDS, Oes-
clim 50, were compared with those of Estraderm TTS 50 (a reservoir-type system)
in a multicenter clinical trial (173). The patches were applied twice weekly for 24
days of each 28-day cycle, over 4 cycles. Oral progestogen was taken by nonhys-
terectomized patients for the last 12 days of estrogen therapy in each cycle. In the
Oesclim 50 group, 4.2% of applications caused a local skin reaction, compared with
9.5% in the Estraderm TTS 50 group. Both were well tolerated, although 7 patients
in the Oesclim 50 group, and 12 in the Estraderm TTS 50 group, discontinued owing
to adverse events. There was no significant difference between the percentages of
patients with signs of hyperestrogenism (20.3% in Oesclim and 20.0% in Estraderm
TTS 50 group). Adhesion was significantly better for Oesclim 50 (6.0% detached)
than for Estraderm TTS 50 (11.3% detached). Greater adhesion of Oesclim 50 was
particularly apparent, with three times fewer Oesclim 50 systems becoming detached
during a shower or bath. Each treatment produced significant and comparable im-
provements in vasomotor symptoms, other menopausal symptoms, and gynecological
assessments. A near-maximal effect on vasomotor symptoms was observed after ap-
proximately 1 month of treatment, and this was maintained for the entire treatment
period.
The local skin tolerability of Oesclim 50 was also compared with that of Es-
traderm TTS 50 (174). In the first study, the modified Draize–Shelanski–Jordan
method of sensitization was used to compare cutaneous tolerability of repeated ap-
plications of patches in 24 healthy postmenopausal women. This indicated no sen-
sitizing potential or induction of allergic reactions. The second study was a multi-
center clinical trial involving 283 healthy menopausal women over 4 months. In this
study, 4.2% of applications in the Oesclim group provoked reactions, compared with
9.5% in the Estraderm group, and 25.9% treated with Oesclim and 39.9% receiving
Estraderm experienced one or more reactions. Redness and itching were the most
frequent reactions in both groups. Durations of reactions were significantly shorter
in the Oesclim group, with higher percentage of durations of less than 1 h and lower
percentage of durations less than 48 h. No reactions in the Oesclim group led to
premature removal, compared with 11 in the Estraderm group. One patient in the
Oesclim group discontinued treatment because of an application site reaction, but
seven in the Estraderm group discontinued.
A comparison between efficacy and safety of two sizes of Lyrelle (matrix type)
and Estraderm TTS 50 (reservoir type) TDS was made in 394 hysterectomized post-
menopausal women in a multicenter trial (175). A significant decrease in mean num-
ber of hot flashes per day was observed in all groups from the end of cycle 1,
reaching 90% at the end of cycle 7. There was no significant difference between
Lyrelle 50 and Estraderm at any time point for any parameter, although between-
group differences for Lyrelle 80 and Estraderm occurred in cycles 1–3 in favor of
Lyrelle 80. A similar effect on blood lipid levels was observed in all groups.
The efficacy, bleeding patterns, and safety of continuous TD and sequential TD
progestogen therapy were compared with those of oral progestogen therapy in post-
menopausal women receiving TD estrogen (176). In a 1-year (13 treatment periods,
28 days each), study, 774 postmenopausal women received 50 ␮g/day of continuous
TD estradiol with either continuous or sequential TD norethisterone acetate (NETA)
in daily doses of 170 or 350 ␮g in a single TDS or sequential oral progestogen (1
mg norethisterone [NET] or 20 mg dydrogesterone per day). The average number
of hot flashes per day decreased from prestudy by over 90%, and this reduction was
unaffected by different progestogen regimes. With sequential progestogen, the bleed-
ing incidence and number of bleeding days did not change over the course of the
study, but were lower in the low-dose TD progestogen group. With continuous pro-
gestogen, the incidence of bleeding decreased in both low- and high-dose groups,
from 35 and 45% in treatment period 1, to 25 and 15%, respectively, at the end of
treatment. Adverse event incidence was similar in both groups, with 23–36% re-
porting events possibly or probably related to HRT (excluding vaginal bleeding).
Lipoprotein-␣ was reduced in all but the oral progestogen group. It was concluded
that continuous and sequential TD estrogen–progestogen treatments with estradiol–
NETA are effective and safe alternatives to continuous TD estrogen and oral se-
quential progestogen for treatment of menopausal symptoms. Continuous TD therapy
with estradiol–NETA may be more acceptable for most patients (i.e., those who wish
to avoid monthly bleeds), whereas the sequential regimen may be preferable when
monthly bleeding may be appropriate.
The safety and efficacy of TD estrogen replacement therapy in liver-trans-
planted menopausal women was investigated (177). Thirty-two menopausal women
who had undergone liver transplantation at least 6 months earlier, received TD es-
tradiol replacement therapy in combination with progestin (Estracomb Ciba, 50 ␮g/
24 h, 250 ␮g/24 h) if the uterus was intact, or estradiol alone (Estraderm Ciba, 50
␮g). Liver function and hemostatic parameters were measured at 0, 3, and 6 months
and gynecological transvaginal ultrasound (TVS) performed at 0 and 6 months. Ef-
ficacy of hormonal treatment was assessed from serum concentrations of estradiol,
estrone, FSH, LH, and SHBG, by measuring endometrial thickness with TVS and
recording changes in subjective climacteric symptoms at 0 and 6 months. Safety was
assessed by measuring liver enzyme activity, liver synthesis functions, and coagu-
lation factors. Therapy did not impair any liver parameters measured, no thrombotic
effect was detected, and hormonal effects of the regimen were verifiable biochemi-
cally, clinically, and by TVS.
Effects of continuous TD estradiol, with or without sequential oral MPA, on
serum lipids and lipoproteins in menopausal women were investigated in 62 healthy
menopausal women (178). Group A included 38 hysterectomized women treated with
continuous TD estradiol only (50 ␮g daily). Group B included 24 menopausal
women, with an intact uterus, treated with TD estradiol (50 ␮g daily) and MPA (10
mg daily for first 12 days of each calendar month). Serum lipids and lipoproteins
were reviewed after 6 months. In group A there was a small reduction in total
cholesterol (⫺5.5%) and slight lowering in LDL-cholesterol (⫺5.7%). In group B,
there were no significant changes in total cholesterol and LDL-cholesterol. HDL-
cholesterol levels did not change significantly with unopposed TD estradiol or ad-
ditional sequential MPA. Serum triglyceride concentrations decreased significantly
in both groups (⫺13.9 and ⫺13.4%, respectively). Serum lipid changes did not differ
between groups. A multicenter trial (179) compared incidence of amenorrhea in 54
postmenopausal women (mean age, 54.9 ⫾ 0.6 years) who underwent six 4-week
cycles of continuous HRT combining progestin–nomegestrol acetate 2.5 mg/day with
one of three estrogens: percutaneous 17␤-estradiol gel (1.5 mg/day, group G), TD
17␤-estradiol patch (50 ␮g/day, group P), or oral estradiol valerate (2 mg/day, group
O). The rate of amenorrhea varied significantly according to type of estrogen prep-
aration (calculated cycle-by-cycle, rates were 67–83% [group G], 25–56% [group
P], and 53–61% [group O]). Overall rates of persistent amenorrhea were not different
between groups for cycles 1 through 3, but for cycles 4 through 6, significantly more
women in groups G and O (67 and 46%, respectively) experienced amenorrhea than
did those in group P (12%). Amenorrhea rates for the entire six-cycle period were
78% for group G, 48% for group P, and 60% for group O, although these differences
were not statistically significant. Differences in rates could not be attributed to en-
dometrial atrophy, because endometrial thickness did not differ significantly among
groups. Calculated as a function of the number of women included in the trial, the
percentage of amenorrheic women was highest with group G, although findings were
similar for group O. Two 11-week, placebo-controlled studies (180) compared the
Climara 7-day matrix patch (at two dose levels) with 625 ␮g/day oral conjugated
equine estrogen, found that both the 50- and 100-␮g/day estradiol patches had a
positive effect on climacteric symptoms. Tolerance was good and similar for both.
Studies of skin irritation and adhesion revealed that the 7-day patch was well tol-
erated and that, although irritation was similar to that associated with Estraderm,
adhesion was superior. Absorption of estradiol was higher and more consistent from
buttock than abdomen, suggesting that choice of application site may require further
investigation.
Efficacy, safety, and tolerability of an estradiol gel (1.0 mg of estradiol daily;
Divigel/Sandrena) in HRT of postmenopausal women were compared with those of
an estradiol TDS (50 ␮g/24 h estradiol, Estraderm TTS) over 12 months with 120
postmenopausal women (181). Dydrogesterone tablets (Terolut), 10-mg daily for the
first 12 days of every month, were used as the progestogen component of therapy.
Twenty-five women without HRT served as reference group for BMD measurements.
Both treatment regimens were equally effective in alleviating climacteric symptoms,
preserving BMD, and were equally safe. A trend toward heavier bleeding was de-
tected in patients treated with the estradiol TD. A nonsignificant decrease of total
cholesterol and triglyceride, but no change in high-density lipoprotein cholesterol
was observed in both groups.
Acceptability of treatment was higher in the gel (96.4%) than patch group
(90.7%). Only two (3.3%) women using the gel complained of skin irritation,
whereas 28 patients (46.7%) using the patch reported this effect. Two doses of TD
estradiol gel (Divigel/Sandrena) plus oral sequential MPA were compared with oral
estradiol valerate plus oral sequential MPA (Divina/Dilena) in postmenopausal
women with climacteric complaints or already using HRT in a 2-year comparative
study (182). Groups received either (a) 1 g of gel containing 1 mg estradiol for 3
months plus 20 mg of oral MPA during last 14 days; (b) 2 g of gel containing 2 mg
of estradiol for 21 days plus 10 mg oral MPA during the last 14 days; (c) 2 mg of
estradiol valerate tablets for 3 weeks plus 10 mg of oral MPA during the last 10
days. With each preparation, climacteric complaints were significantly reduced, good
bleeding control was obtained, BMD was maintained, and bone turnover was re-
duced. Lipid parameters showed no unfavorable changes. Continuation rates were
similar in all groups, with 74% of patients completing the first year, and 94% of
patients who elected to continue completing the second year. Tolerability of gel was
good, with only 1.7% of patients discontinuing because of skin irritation.
Estradiol and estrone concentrations and bioavailability were compared after a
single dose and at a steady state during oral estradiol valerate, TD estradiol gel, and
TD estradiol TDS treatments (183). In study A, 12 healthy postmenopausal women
received 1.5 mg of estradiol as a TD gel or a 2-mg estradiol valerate tablet daily for
14 days. In study B, 15 postmenopausal women were treated for 18 days with 1.5-
mg estradiol gel or a TD system releasing estradiol 50 ␮g/24 h (replaced every 72

h). Tablet and transdermal gel yielded similar serum estradiol profiles with a peak
concentration 4–5 h after administration. The TDS gave relatively stable estradiol
levels during the mid-third of the wearing time, whereas much lower levels were
observed at the beginning and end. There was no difference in fluctuation of peak-
and-trough estradiol levels between gel (56 or 67%) and tablet (54%), but fluctuation
was greater with the TDS (89%). Bioavailability of estradiol from the gel was 61%,
compared with the tablet, and 109%, compared with the TDS. Gel was not bioequiv-
alent with tablet or TD and individual dose adjustments may be needed when chang-
ing administration forms.
The effect of HRT on the conjunctiva in postmenopausal women was the sub-
ject of a clinical study in 11 postmenopausal women receiving TD estradiol or TD
estradiol plus MPA for 4 months (184). Significant increases in serum estradiol levels,
vaginal maturation value, and mild cytological maturation changes in conjunctival
epithelium were observed. The changes were significant and the data support the
view that HRT induces cytological maturation changes in conjunctival epithelium in
postmenopausal women.
A case of sensitization to estrogen was reported in a patient receiving Estraderm
(185). A 40-year-old woman suffered from cyclic skin disorders and at each menses
developed pruritus and erythematous papulovesicular lesions over the members and
trunk. Prick and patch tests with alcoholic solutions of estrone alone and serum tests
for antiethinyl–estradiol antibodies and antiprogesterone antibodies were positive. In
cases of progesterone sensitization, treatment of choice is estrogen inhibition of ovu-
lation, whereas for estrogen sensitization, antiestrogen treatment appears more effec-
tive. Bilateral ovariectomy may be required in difficult cases.
The history, current clinical practice, choice of methods, and number of pre-
scriptions and sales of HRT in the United States were reviewed (186). The percentage
of women currently utilizing HRT was greater in women aged 40–60 (35%), but
fell with ages older than 65 (15%), and declined further in women older than 80
(7%). News media and physicians were the largest source of information on HRT,
and obstetricians and gynecologists were predominant prescribers. Women who spon-
taneously developed menopause early and younger women undergoing castration
were more likely to take TD estrogen, but 86% of U.S. prescriptions were for oral
estrogens (with conjugated equine estrogens [70%] the market leaders). Of the pre-
scriptions for women with a uterus 50% were for combined continuous estrogen and
progestogen, whereas 42% contained cyclic estrogen and progestogen. Although use
of HRT by postmenopausal women in the United States increased, the percentage of
current users remained lower than anticipated, in spite of widespread media and
educational efforts of benefits. TD estrogens were used more commonly in women
in the early postmenopausal period, whereas in women with a uterus, most U.S.
physicians prescribed combined estrogen plus progestogen, but used oral, rather than
TD, estrogen.
Bioavailability, pharmacokinetics, and tolerability of two matrix transdermal
delivery systems providing 50 ␮g/24 h of estradiol were compared in 20 healthy
postmenopausal women (187). Menorest (3–4 days suggested use) and Climara (7
days suggested use) were compared at steady-state in two 14-day treatment periods
separated by a 4-week washout, with plasma estradiol monitored in the second week
of each treatment. No differences between treatments relative to AUC, C(max), C(min),
C(average), or fluctuations in plasma estradiol. T(max) was significantly shorter for Men-
orest than for Climara, and C(max) and C(min) were significantly higher for the second
Menorest patch than for the first. Three cases of erythema with Menorest and a total
of 21 skin reactions in 15 subjects with Climara were reported. Systemic tolerability
was similar between treatments with 8 estrogen-related adverse events in 8 subjects
with Menorest and 13 events in 10 subjects with Climara. It was concluded that
although the bioavailability of estradiol from these TDS was similar, the products
were not bioequivalent because T(max) was significantly shorter for Menorest than for
Climara. A comparison of continuous combined TD delivery of estradiol–norethin-
drone acetate and estradiol alone for menopause was carried out to determine whether
a continuous estradiol–norethindrone acetate TD delivery system reduced incidence
of endometrial hyperplasia in postmenopausal women more than that of TD estradiol
alone (188). A total of 625 postmenopausal women were assigned to one of four
treatments: TD estradiol 50 ␮g/day, or TD estradiol–norethindrone acetate, with 50-
␮g estradiol and 140, 250, or 400 ␮g/day of norethindrone acetate. Endometrial
hyperplasia was found in 37.9% in the estradiol-alone group versus 0.8%, 1%, and
1.1% in the estradiol–norethindrone acetate 50–140, 50–250, and 50–400 groups,
respectively. Uterine bleeding was less frequent in the estradiol–norethindrone ace-
tate 50–140 group. The estradiol–norethindrone acetate combination TDS showed
skin tolerance comparable with that of estradiol alone.
Bioavailability of two 100-␮g daily 17␤-estradiol TDS (once-a-week matrix
patch and twice-a-week reservoir patch) was compared in healthy postmenopausal
women (189) in a two-period, crossover study with two 8-day treatment periods
separated by a minimum 7-day washout. Subjects were assigned to either (a) matrix
patch applied to abdomen and worn for 7 consecutive days, or (b) reservoir patch
applied to abdomen and worn for 4 days, followed immediately by a second reservoir
patch worn for 3 days. Three-hours after patch application serum estradiol levels
were significantly higher than levels at time of patch application. After 12 h, mean
serum estradiol level in women with matrix patches was 98.20 ⫾ 44.97 pg/mL,
significantly higher than in women with the reservoir patch (62.20 ⫾ 16.21 pg/mL).
An AUC (0–168 h) with the matrix patch was also higher than for reservoir patch.
Left ventricular heart function and its response to long-term estrogen replace-
ment therapy was assessed in 30 postmenopausal women, 20 of whom had modest-
to-severe hot flashes and 10 of whom had never had them (190). Continuous TD
estradiol was given to women with surgically induced menopause, and a combination
of TD estradiol and sequential MPA to those with spontaneous menopause. Although
HRT significantly improved heart function in healthy postmenopausal women, there
appeared to be some minor differences in response between those with flashes and
nonflashers.
Effects of TD estradiol on serum triglycerides in menopausal women with
preexisting mild-to-moderate hypertriglyceridemia were evaluated (191). Forty-four
women (posthysterectomy and maintained on 50-␮g unopposed estradiol for 6
months) were divided into those with normal baseline triglyceride concentrations
(0.4–2 mmol/L) and those with raised baseline readings (>2–4 mmol/L). Significant
reductions in serum triglyceride concentrations occurred in both groups (⫺9.6 and
⫺17%, respectively). TD estradiol therapy may be a useful treatment option in men-
opausal women with preexisting hypertirglyceridemia.
A summary of the tolerability and safety of Oesclim, which was developed

with the objective of providing improved local skin tolerability and adhesion, while
minimizing hyperestrogenic effects, has been published (192). In a comparative clin-
ical trial, Oesclim resulted in fewer than half as many application site reactions as
Estraderm TTS (4.3 vs. 9.5%), and the duration of reactions was significantly lower
in the Oesclim group. Oesclim was well tolerated in all clinical trials and reported
to have a estrogen-specific tolerability comparable to Estraderm TTS.
Low-dose Oesclim (25 ␮g/day) was associated with reduction in hyperestro-
genic side effects compared with higher doses. In a study of long-term Oesclim
therapy, 79% of patients wished to continue therapy after 1 year, and in a follow-up
study, 79.8% wished to continue at the end of 3 years.
Efficacy and safety of three dosages of Oesclim, delivering 0.025, 0.050, or
0.100 mg 17␤-estradiol per 24 h, in treatment of moderate to severe vasomotor
symptoms was evaluated in a multicenter trial (193). A total of 196 highly symp-
tomatic menopausal women received 12 weeks of continuous unopposed treatment
with one of the three dosages of Oesclim or a matching placebo patch. Reduction
in frequency of moderate-to-severe vasomotor symptoms was statistically significant
compared with placebo from week 2 onward in the Oesclim 50 and 100 groups, and
from week 3 onward in the Oesclim 25 group. Symptom severity was also reduced.
Estrogen-related adverse events were less frequent in the Oesclim 25 group.
Significant differences in estradiol bioavailability were reported from two sim-
ilarly labeled estradiol matrix TDS (Alora and Evorel) (194). The fluctuation index
produced by Evorel was significantly higher than that with Alora (135 vs. 76%) and
the estradiol baseline-corrected AUC was significantly lower for Evorel than Alora
(1870.6 vs. 2871.8 pg h⫺1 mL⫺1). Efficacy, safety and compliance with Climara 50
and Climara 100 were evaluated in 100 women (195). Reductions in weekly fre-
quency of hot flashes were 58.6% in the Climara 50 group and 72.1% in the Climara
100 group. Of 64 patients with sleep disturbances, 51 reported some improvement,
10 had no advantage, and 3 were worse. Incidence of temporary minor side effects
was dose-related (38 vs. 70%).
A multicenter study assessed the efficacy, safety, and tolerability of a low-dose
(0.0375 mg/day) estradiol matrix TDS for the treatment of moderate-to-severe post-
menopausal hot flashes in healthy women (196). Estradiol matrix or matching pla-
cebo patches were administered over three 4-week–treatment cycles to 257 patients
(130 estradiol, 127 placebo). Assessments of treatment effectiveness significantly
favored the estradiol patch over placebo. This lowest available dose estradiol TD
provided significant relief from moderate-to-severe postmenopausal hot flashes and
was well tolerated. A 3-year study enrolled 277 early postmenopausal women to
examine the efficacy of a matrix 17␤-estradiol TDS, at three dosages (25, 50, and
75 ␮g/day) combined with sequential oral dydrogesterone 20 mg/day, in preventing
bone loss (197). At 2 years, difference from placebo in percentage change from
baseline of L1-4 lumbar spine BMD was 4.7 ⫾ 0.7% (25 ␮g/day), 7.3 ⫾ 0.7% (50
␮g/day), and 8.7 ⫾ 0.7% (75 ␮g/day). There were also significant increases in fem-
oral neck, trochanter, and total hip BMD with all doses of estradiol, compared with
placebo. Most patients receiving estradiol also had a significant gain (>2.08%) in
lumbar spine bone mass and clinically significant and dose-related decreases in total
serum osteocalcin, serum bone alkaline phosphatase, and urinary C-telopeptide, with
all three markers of bone turnover returning to premenopausal levels.
The effects of four doses (0.025, 0.05, 0.06, and 0.1 mg/day) of a 7-day TD
17␤-estradiol delivery system on bone loss in postmenopausal women were evaluated
in a multicenter study (198). At 24 months, doses of 0.025, 0.05, 0.06, and 0.1 mg/
day resulted in mean increases in BMD of the lumbar spine of 2.37, 4.09, 3.28, and
4.70%, respectively, and increased BMD of the total hip by 0.26, 2.85, 3.05, and
2.03%, respectively. All increases were significantly greater than placebo. Consistent
and significant improvements in biochemical markers of bone turnover were also
noted in all treatment groups.
Two estradiol TDS that released 25 or 37.5 ␮g/day were compared with a
placebo patch on 156 patients in natural or surgical menopause suffering from at
least five hot flashes per day, treated continuously for 12 weeks, without progestin
opposition (199). ‘‘Responders’’ (patients with fewer than three hot flashes per day
at end of treatment), were 82 and 90% with 25 or 37.5 ␮g/day, respectively, both
significantly more than placebo (44%).
Efficacy and tolerability of a matrix patch delivering estradiol at doses of 0.05
and 0.10 mg/day (Estraderm MX 50, 100) in treatment of moderate to severe post-
menopausal symptoms was compared (200). A total of 254 postmenopausal women
received 0.10, 0.05 mg, or placebo for 12 weeks continuously. TDS were applied
twice weekly to the buttocks with each patient wearing two patches simultaneously.
Patches containing 0.10 and 0.05 mg estradiol were superior to placebo in reducing
hot flashes per 24 h after 4, 8, and 12 weeks of treatment. For all other efficacy
parameters studied, both dosage strengths were superior to placebo at all time points.
It was concluded that this matrix patch offered an effective and well-tolerated dosage
form and may be particularly suitable for women who experience local sensitivity
to alcohol-containing systems.
The effect of the administration route and cigarette smoking on plasma estrogen
levels during HRT was evaluated in 14 healthy postmenopausal women (6 smokers
and 8 nonsmokers) (201). All patients randomly received cyclic therapy with estra-
diol and norethisterone orally or TD, each for 6 months. Plasma levels of estrone,
estradiol, and estrone sulfate, all were 40–70% lower in smokers than nonsmokers
when HRT was given orally. Oral dosing caused higher extradiol/estradiol sulfate
and estrone/estradiol sulfate ratios compared with TD therapy in smokers (40.2 vs.
7.0; and 3.2 vs. 0.8, respectively).
Pharmacokinetics of Fem7 (an estradiol matrix-type TDS, applied once weekly)
was investigated in 36 healthy postmenopausal women at doses of 25, 50, 75, and
100 ␮g/24 h (202). Maximum plasma estradiol and estrone concentrations occurred
14–20 h after patch application, remained within the therapeutic range until removal,
and returning to baseline within 12 h. Plasma estradiol concentrations increased in
a dose-dependent manner for all dose levels, and plasma estrone increased for the
three highest doses. Treatment was well tolerated at all dose levels and no severe
adverse reactions were reported.
The efficacy of two strengths of TD estradiol matrix with daily oral doses of
conjugated equine estrogens in reducing the frequency of moderate-to-severe hot
flashes in postmenopausal women was evaluated (203). An estradiol TDS (Alora
0.05 or 0.1 mg/day) administered twice weekly or oral doses of conjugated equine
estrogens (CEE 0.625 or 1.25 mg) administered daily were given to 321 highly
symptomatic postmenopausal women for 12 weeks. Results indicated no significant
differences at any time point in mean frequency or mean percentage reduction in
frequency of moderate-to-severe hot flashes between patients given Alora 0.1 mg/
day or CEE 1.25 mg/day, or between the Alora 0.05 mg/day and CEE 0.625 mg/day
groups by week 12. There were no serious or unexpected adverse events with the
TDS, and local skin tolerability was excellent. Other estrogenic effects were com-
parable between TD and oral administration groups, except for lower incidence of
bleeding in women receiving the lower transdermal dose.
The effect of estradiol TD in postmenopausal women with confirmed pollaki-
uria and urinary incontinence was investigated in ten women using Estraderm TTS
2 mg for 8 weeks (204). In seven cases, severity of urinary incontinence was ‘‘very
effective’’ in three cases, ‘‘improved’’ in two, ‘‘slightly improved’’ in one, and ‘‘no
change’’ in one.
Postmenopausal women (especially those older than 60 years) prefer HRT that
avoids cyclical uterine bleeding and continuous combined HRT regimens were pri-
marily introduced to avoid bleeding and increase compliance. In a multicenter study,
136 women at least 2 years postmenopausal, with mild-to-moderate menopausal
symptoms, received either Estragest TTS 0.125/25 (delivering 0.125 mg norethister-
one acetate [NETA] and 25 ␮g estradiol per day) or placebo for 6 months (205).
After 4, 12, and 24 weeks the Kupperman index was significantly lower in the
Estragest group, and the severity of vaginal dryness and dyspareunia at 12 and 24
weeks was also reduced. The proportion of superficial cells increased significantly
in the Estragest, but not the placebo group. The percentage of patients reporting
amenorrhea with Estragest ranged from 80 (month 2) to 87% (month 6). In a second
multicenter study lasting 1 year, 441 postmenopausal women received one of three
continuous combined HRT regimens: group A, Estragest TTS 0.125/25; group B,
TDS delivering estradiol 50 ␮g and NETA 0.25 mg/day, group C, oral tablets con-
taining 2 mg estradiol and NETA 1 mg/day. During treatment cycles 4–6, amenor-
rhea was achieved in 73% (group A), 47% (group B), and 66% (group C). During
treatment cycles 10–12, proportions increased to 86% (A), 65% (B), and 79% (C).
Bleeding patterns in groups A and C were not significantly different, but superior to
those in group B. It was concluded that Estragest TTS 0.125/25 was effective in
treatment of mild-to-moderate menopausal symptoms and urogenital complaints, in-
duced a high rate of amenorrhea and provided good endometrial protection.
Effects of three commonly prescribed estrogen replacement therapies (oral con-
jugated equine estrogens [CEE; n = 37], oral micronized estradiol [ME; n = 25], and
TD estradiol [TE; n = 24]) on the concentrations of blood sex hormone-binding
globulin (SHBG), estradiol, and estrone were studied (206). Increases in SHBG con-
centrations were 100, 45, and 12% for subjects receiving CEE, ME, and TE regi-
mens, respectively. Decreases in the percentage estradiol not bound to protein and
increases in the percentage of estradiol bound to SHBG correlated with therapy-
mediated changes in concentrations of this protein.
Systemic bioavailability and plasma profiles of 17␤-estradiol after application
of three matrix patches: Menorest, Tradelia, and Estraderm MX, claiming to deliver
50 ␮g/day were evaluated (207). All patches were each worn randomly by 21 post-
menopausal women volunteers over 96 h, separated by an at least a 7-day washout
period. Tmax (32 h) was the only pharmacokinetic parameter identical for all patches.
Menorest produced the highest estradiol bioavailability, judged by the AUC(0–96 h) =
3967.8 ⫾ 1651.8 pg/mL h⫺1, C(average) = 41.3 ⫾ 21.3 pg/mL, C(min) = 36.8 ⫾ 8.6 pg/
mL. Tradelia (AUC(0–96 h) = 3737.9 ⫾ 1637.6 pg/mL h⫺1, C(average) = 38.9 ⫾ 17.0 pg/
mL, and C(min) = 33.8 ⫾ 26.7 pg/mL) was not significantly less than Menorest.
Estraderm MX showed the lowest estradiol profiles (AUC(0–96 h) = 3192.1 ⫾ 1646.0
pg/mL h⫺1, C(max) = 38.9 ⫾ 25.1 pg/mL, C(average) = 33.2 ⫾ 17.1 pg/mL). Menorest
showed the smallest fluctuation over the entire test period, similar to Estraderm MX,
whereas Tradelia showed the highest fluctuation and the highest C(max) = 48.0 ⫾ 20.3
pg/mL. When estradiol baseline levels, before patch application, were individually
subtracted from the subsequent estradiol level, Estraderm MX was not bioequivalent
to Menorest. A circadian curve pattern of estradiol plasma level was observed for
all patches, and in the evening, higher plasma levels were always detected. Individual
comparison of AUC(0–96 h) for each patch showed large interindividual variability
(2000–8000 pg/mL h⫺1) for all patches, but relatively small individual variability.
Women with high estradiol bioavailability (high-responders) maintained high bio-
availability with all patches, and women identified as low- and medium-responders
remained the same, regardless of the applied patch. Side effects were approximately
equal in all patches, with a maximum after 72 h.
Two long-term multicenter studies compared the efficacy on climacteric symp-
toms of a new active matrix estradiol TDS (CAS 50-28-2) with a reference reservoir
patch (both releasing 50 ␮g/day) (208). One group received the matrix patch and
the other the reservoir patch in 4-week cycles, with twice-weekly application of
patches for 3 weeks, followed by 1-week washout. Progestin opposition was with
MPA; 5 mg/day orally in the last 11 days of patch application in a German study
and with 10 mg/day in the last 12 days of patch application in an Italian study. Each
study was divided into two parts: (a) with three 4-week cycles and (b) for ten 4-
week cycles. In the German study both patches quickly relieved climacteric symp-
toms during the first 3 weeks of application, as shown by rapid decrease of the
Kupperman Index. At the end of part 1, 91% (matrix) and 96% (reservoir) group
reported relief from climacteric symptoms; at the end of part 2, these were 98% and
95%, respectively. Both patches were systemically fairly well tolerated and only 4.5
(matrix) and 3.9% (reservoir) discontinued owing to adverse reactions. Relative to
local skin reactions, the matrix patch was significantly better tolerated and adhesion
was better. In the Italian study both patches relieved climacteric symptoms during
the first 3 weeks of application. At the end of part 1 both patches relieved 95% of
patients and at the end of part 2, 100% of patients were relieved. Patches were
systemically equally fairly well tolerated with premature discontinuations for sys-
temic adverse drug reactions in 5.0% (matrix) and 3.9% (reservoir) groups. As in
the German study, matrix patches were significantly better tolerated.
The effect of short-term HRT with estradiol and norethisterone on the phar-
macokinetics of phenazone was investigated in ten women at least 6 weeks after
ovariectomy (209). Each patient received TD estradiol (4 mg, every 3 days) for a
period of 18 days, and subsequently oral norethisterone (5 mg/day) for 10 days, with
an interval of 1 day. Pharmacokinetic studies were performed before estradiol ad-
ministration, on the last day of estradiol treatment and on the last day of norethis-
terone administration. Short-term administration of estradiol did not modify the phar-
macokinetics of phenazone.
Percutaneous absorption of progesterone in postmenopausal women treated by
application of progesterone cream to the skin was evaluated in six postmenopausal
women over a 4-week period (210). Transdermal estradiol, 0.05 mg, was applied 2
days before first application of progesterone (30 mg/d) and continued throughout the
study, with patches changed twice weekly. Progesterone cream was applied once a
day for 2 weeks. On days 15–29 progesterone cream was applied twice daily (60
mg/d). Serum 17␤-estradiol and progesterone were measured over 24 h on day 1
and at weekly intervals for the study duration. Individual serum 17␤-estradiol con-
centrations ranged from 40 to 64 pg/mL, but intraindividual concentrations remained
constant. Serum progesterone concentrations were 1.6–3.3 ng/mL. After 2 weeks of
percutaneous dosing, progesterone concentrations were sustained for at least 8 h and
were consistent within an individual. An increase in progesterone concentration oc-
curred after 4 weeks, compared with 2 weeks. Individually, a significant correlation
was seen between absorption of 17␤-estradiol and progesterone.
Serum and urinary hormone levels following short- and long-term administra-
tion of two regimens of progesterone cream in postmenopausal women were eval-
uated (211) in a multiple-dose study using 24 healthy postmenopausal women. Sub-
jects were allocated to progesterone cream, 40 mg daily, or 20 mg twice daily, for
42 days. Serum progesterone was measured on days 1 and 42 before the morning
dose, and at 2, 4, 6, 12, and 24 h after the morning dose. Serum FSH, estradiol,
testosterone, and urinary pregnanediol-3-glucuronide were also measured on days 1
and 42. The mean progesterone concentration rose at each sampling time between
days 1 and 42 and there was evidence of a rise in pregnanediol-3-glucuronide over
the study course. There were no changes in FSH, estradiol, or testosterone and no
differences were detected between the regimens.
The use of TD progesterone cream for vasomotor symptoms and postmeno-
pausal bone loss was also investigated (212). One hundred two healthy women,
within 5 years of menopause, were assigned to TD progesterone cream or placebo.
Subjects were instructed to apply 0.25 teaspoon of cream (containing 20 mg pro-
gesterone or placebo) to the skin daily. Subjects received daily multivitamins and
1200 mg of calcium; symptoms were reviewed ever 4 months. In the treatment group,
69% and 55% in the placebo group initially complained of vasomotor symptoms.
Improvement or resolution of these symptoms was noted in 83% of treatment sub-
jects and 19% of placebo subjects. However, the number of women showing a BMD
gain of more than 1.2% did not differ significantly.
In addition to the widespread use of HRT TD, steroid delivery has also been
investigated as a means of contraception. A once-a-week (monophasic) contraceptive
TDS was designed to simultaneously deliver a low-dose combination of levonorges-
trel (LNG) and 17␤-estradiol (E2) for fertility regulation in females (213). In vitro
permeation studies using human cadaver skin indicated 6.0 ⫾ 0.9 ␮g/day cm⫺2 of
LNG and 2.9 ⫾ 0.5 ␮g/day cm⫺2 of E2 could be delivered. A 7-day dermal toxicity
study on six rabbits indicated minimal potential to cause skin irritation, and histo-
pathological examination revealed only mild-to-moderate inflammation. A phase 1
bioavailability–dose proportionality clinical study, consisting of pretreatment, treat-
ment, and posttreatment cycles, was conducted on fertile Chinese women. During
the pretreatment cycle, 48 subjects were given placebo patches to study wearability
(including skin irritation and adhesion tests). During the treatment cycle, each subject
in the test groups received weekly application of 1 (A), 2 (B), or 3 (C) 10-cm2
patches, and group D received daily 150 ␮g of LNG and 35 ␮g of ethynyl estradiol,
orally. The wearability study indicated patches were very well accepted. Residual
assay of used TDS indicated delivery of LNG and E2 at rates of approximately 5.0
␮g/cm2 day⫺1 and 4.0 ␮g/cm2, day⫺1, respectively, during the treatment cycle. Ra-
dioimmunoassay (RIA) of serum samples demonstrated therapeutically effective se-

rum concentration of LNG and serum profiles of progesterone, luteinizing hormone
(LH), and follicle-stimulating hormone (FHS) also indicated that ovulation inhibition
occurred in most of the subjects wearing TDS. No subject became pregnant, and
posttreatment hormonal profiles indicated that most returned to their normal men-
strual cycle on termination of patch use. A second evaluation of TD delivery of
steroids for contraception has been reported (214). A TDS changed weekly and de-
livering both E(2) and LNG at daily dosages of 3.8 ⫾ 0.8 and 2.9 ⫾ 0.7 ␮g/cm2
day⫺1, respectively, showed ovulation suppression. An alternative progestin (ST
1435) penetrated skin when formulated in acetylated lanolin or an hydroalcoholic
gel and produced ovulation suppression at a dose of 2 mg/day in a small number of
cycles. For contraceptive purposes TD systems must be perfectly adhesive, well
tolerated locally, and achieve nearly 100% efficacy. These targets are very challeng-
ing, although the potential advantages are so high that the concept deserves further
development.
A pilot clinical study was carried out to evaluate the cognitive and neuroen-
docrine response to estrogen administration for postmenopausal women with Alz-
heimer’s disease (AD) (215). Twelve women with probable AD of mild-to-moderate
severity completed the study. During an 8-week treatment period, 6 received 0.05
mg/day dosage of 17␤-estradiol by a patch and the remainder a placebo patch. Sig-
nificant effects of estrogen treatment were observed on attention and verbal memory.
In women treated with estrogen, verbal memory enhancement was positively corre-
lated with estradiol plasma levels and negatively correlated with plasma concentra-
tions of insulin-like growth factor-binding protein-3 (IGFBP-3). This suggested that
estrogen replacement may enhance cognition for postmenopausal women with AD.
A trial assessed the effectiveness and tolerability of transdermal estrogen in
men with hot flashes after hormonal therapy for prostate cancer (216). Twelve men
with moderate to severe hot flashes received either low-dose (0.05 mg) or high-dose
(0.10 mg) estrogen patches applied twice weekly for 4 weeks. After a 4-week wash-
out, each patient received the alternative dose for 4 weeks. There was a significant
reduction in overall severity of hot flashes seen with both low- and high-dose estro-
gen patches, but a significant reduction in daily frequency of hot flashes was seen
only at the high dose. Eighty-three percent reported either mild, moderate, or major
improvement in symptoms with either low- or high-dose patch. Mild, painless breast
swelling or nipple tenderness was noted in 17 and 42% of the men treated with low-
and high-dose estrogen, respectively. FSH levels decreased significantly at both
doses. Estradiol levels increased from 12.1 to 16.4 (low-dose) and 26.9 (high-dose)
pg/mL, but there was no significant change in serum testosterone or LH levels.
D. Androgens
An important aim in treating male hypogonadism is restoration of physiological
concentrations of testosterone and metabolites. New methods for testosterone deliv-
ery that have provided increased options for men requiring hormonal replacement
therapy have been reviewed (217). Intramuscular administration of testosterone is
associated with early, high serum levels followed by a gradual decline over the
dosing interval. The TDS now available as alternatives include Testoderm (applied
to the scrotum) and Androderm (applied to nonscrotal skin). Most patients achieved
normal serum testosterone levels with circadian variation and normal estradiol levels.
Serum LH levels generally decreased, but not to suppressed levels, and Testoderm
use leads to an increase in plasma dihydrotestosterone (DHT). Clinical response in
mood, energy level, and sexual function were improved with both systems and were
generally comparable with intramuscular injection. There were no clinically signifi-
cant changes in laboratory parameters, including prostatic specific antigen (PSA),
and prostate size did not increase above normal. Skin reactions were common and
may require discontinuation of therapy. Patients with inadequate scrotal size may not
achieve satisfactory results with Testoderm. Although patches are more expensive
than intramuscular (IM) injections, they require less frequent office visits and both
transscrotal and transdermal systems offer a good alternative for hypogonadal men
who do not desire fertility during the treatment period.
Scrotal testosterone patches can produce normal serum levels mimicking di-
urnal variations. This was followed in hypogonadal men treated transdermally for
up to 10 years (218). Eleven men (age 35.9 ⫾ 9.8 years) at start of study were
treated with transscrotal patches (Testoderm) because of primary (n = 4) or secondary
(n = 7) hypogonadism. Clinical examinations were performed every 3 months during
the first 5 years and every 6 months thereafter. On daily application of one patch,
testosterone levels rose from 5.3 ⫾ 1.3 to 16.7 ⫾ 2.6 nmol/L at month 3 and re-
mained in the normal range throughout treatment. Serum DHT rose from 1.3 ⫾ 0.4
to 3.9 ⫾ 1.4 nmol/L and estradiol from 52.3 ⫾ 9.3 to 71.3 ⫾ 9.6 pmol/L and
remained stable. Patients reported no local side effects apart from occasional itching.
No relevant changes occurred in clinical chemistry and hemoglobin and erythrocyte
counts remained normal. Bone density increased slightly from 113.6 ⫾ 5.4 to 129.7
⫾ 9.3 mg/cm3. In the nine patients who were younger than 50 years prostate volumes
showed a small, but insignificant, increase from 16.8 ⫾ 1.5 to 18.8 ⫾ 2.1 mL during
therapy. In two older patients, prostate volume remained constant or decreased
slightly during therapy. Prostate-specific antigen levels were constantly low in all
patients.
The possibility of immediate adverse effects of short-term testosterone admin-
istration to older men with low bioavailable testosterone, especially on the symptoms
of benign prostate hyperplasia, was investigated (219). A 9-week intervention with
either intramuscular testosterone enanthate (200 mg every 3 weeks), TD testosterone
(two 2.5-mg patches per day), or neither, was followed by a 9-week observation
period. Twenty-seven men (age 74 ⫾ 3 years) with no medical conditions known to
affect bone turnover and with total testosterone levels less than 350 ng/dL or bio-
available testosterone levels less than 128 ng/dL were included. All men receiving
testosterone treatment increased levels above their own baseline, but only six of nine
men receiving TD testosterone achieved bioavailable testosterone levels in the nor-
mal range for young men. No side effects were reported using intramuscular delivery,
but five of nine men using TD testosterone developed a rash.
As part of a phase III multicenter study, pharmacokinetics and metabolism of
a permeation-enhanced testosterone transdermal system and the influence of the ap-
plication site were investigated in 34 hypogonadal men (21–65 years of age) (220).
After an 8-week androgen washout period, two patches were applied to the back for
24 h. Serum concentrations of total testosterone (T), bioavailable testosterone (BT),
DHT, and estradiol [E(2)] increased from hypogonadal levels into normal physio-
logical ranges and declined to baseline levels within 24 h after system removal. Peak
concentrations occurred about 8 h after application for T and BT and at 13 h for

DHT and E(2). Estimated half-lives were: T, 1.29 ⫾ 0.71 h; BT, 1.21 ⫾ 0.75 h;
DHT, 2.83 ⫾ 0.97 h; and E(2), 3.53 ⫾ 1.93 h. The influence of application site was
evaluated by applying two patches for 24 h to the abdomen, back, chest, shin, thigh,
or upper arm. Hormone profiles were qualitatively similar at each site, but CSS values
were significantly different. Based on BT levels, rank ordering of sites was back >
thigh > upper arm > abdomen > chest > shin. DHT/T and E(2)/T ratios showed
negligible site variation.
Further investigations of hormone levels, pharmacokinetics, clinical response,
and safety of a permeation-enhanced testosterone transdermal system (TD) in the
treatment of hypogonadal men were reported (221). This was a multicenter study
with four consecutive periods: period I (3 weeks), evaluation of current androgen
therapy (primarily testosterone enanthate injections (mean dose 229 mg; mean in-
terval 26 d); period II (8 weeks), androgen washout; period III (3–4 weeks), single-
dose pharmacokinetic studies of TD systems; period IV (12 months), efficacy, safety,
and steady-state pharmacokinetic evaluation of TD systems (5 mg/day nominal de-
livery rate of testosterone). Thirty-seven hypogonadal men 21–65 years old enrolled,
34 entered periods III and IV and 29 (9 primary, 20 secondary hypogonadism) com-
pleted the study. Four patients withdrew because of adverse events. Measurements
included morning serum levels of total testosterone (T); bioavailable testosterone
(BT), DHT, and E(2) levels; circadian pattern of T profiles and 24-h time-averaged
T level; LH levels in patients with primary hypogonadism; and reduction of hypo-
gonadal symptoms. Safety assessments included skin tolerability, prostate parameters,
lipid profile, and systemic parameters. Twelve months of TD therapy normalized
morning serum T levels in 93% of patients, and produced more than 80% normali-
zation of BT, DHT, and E(2) levels. The TD system mimicked the circadian variation
in T levels seen in healthy young men and normalized 24-h time–average T levels
in 86% of patients. LH was suppressed in eight of nine men with primary hypogo-
nadism, and normalized in five of these. Subjective symptoms of hypogonadism,
including decreased libido and fatigue, showed improvement after 2–4 weeks of
treatment in most patients. Most adverse events were local skin reactions. Prostate
assessments showed a lower prostate-specific antigen level during TD therapy com-
pared with IM injections (0.66 vs. 1.00 ␮g/L), but prostate size did not differ sig-
nificantly between the treatment regimens.
Weight loss associated with human immunodeficiency virus (HIV) infection is
multifactorial in its pathogenesis, but it was speculated that hypogonadism contrib-
uted to depletion of lean tissue and muscle dysfunction (222). Effects of testosterone
replacement, using Androderm, on lean body mass, body weight, muscle strength,
health-related quality of life, and HIV disease markers were evaluated. Testosterone
replacement in HIV-infected men with low testosterone levels was considered safe
and associated with a 1.35-kg gain in lean body mass, a significantly greater reduc-
tion in fat mass than achieved with placebo treatment, and an increased red cell
count and improvement in role limitation owing to emotional problems. It was con-
cluded that further studies were required to assess whether such supplementation can
produce clinically meaningful changes in muscle function and disease outcome in
HIV-infected men.
Markedly decreased serum androgen levels occur in women with acquired im-
munodeficiency syndrome (AIDS) and may be a contributing factor to the wasting
syndrome. A pilot study of the effects of androgen replacement therapy was con-
ducted to determine efficacy in terms of change in serum testosterone, safety param-
eters, and tolerability, and to investigate testosterone effects on weight, body com-
position, quality of life, and functional indexes (223). Fifty-three ambulatory women
with AIDS wasting syndrome, free of new opportunistic infection within 6 weeks of
study initiation and with serum levels of free testosterone less than normal reference
range, were enrolled. Subjects weighed 92 ⫾ 2% of ideal body weight, and had lost
17 ⫾ 1% of their maximum weight. Subjects received two placebo patches (PP),
one active and one placebo patch (AP), or two active patches (AA) applied twice
weekly to the abdomen for 12 weeks. Nominal delivery rates were 150 and 300 ␮g/
day, respectively, for AP and AA groups. Serum free testosterone levels increased
significantly from 1.2 ⫾ 0.2 to 5.9 ⫾ 0.8 pg/mL (AP) and from 1.9 ⫾ 0.4 to 12.4
⫾ 1.6 pg/mL (AA). Testosterone administration was generally well tolerated locally
and systemically, with no adverse trends in hirsutism scores, lipid profiles, or liver
function tests. Improved social functioning and pain score were observed in AP-
versus PP-treated patients. These data suggested that testosterone administration may
improve the status of women with AIDS wasting.
The literature on androgen replacement for erectile dysfunction was evaluated
by metanalysis (224). Study inclusion criteria were testosterone given as the only
therapy for erectile dysfunction and a clearly stated definition of response for eval-
uating treatment. Sixteen of 73 articles published between 1966 and 1998 were in-
cluded. Overall response rate was 57%, and patients with primary versus secondary
testicular failure had a response rate of 64% versus 44%. Intramuscular and oral
methods of delivery were equivalent (response rates 51.3 and 53.2%, respectively)
but response to TD therapy was significantly different (80.9%).
It is acknowledged that women may experience symptoms secondary to an-
drogen deficiency, and there is substantial evidence that prudent androgen replace-
ment can be effective in relieving both physical and psychological symptoms of
androgen insufficiency (225). Testosterone replacement for women is now available
in a variety of formulations. It appears to be safe, with the caveat that doses are
restricted to the ‘‘therapeutic’’ window for androgen replacement in women, such
that the beneficial effects on well-being and quality of life are achieved without
incurring undesirable virilizing side effects.
For treatment of adult hypogonadal men, nightly 24-h application of the An-
droderm testosterone TDS (5 mg/day) has been demonstrated to be effective by a
series of clinical pharmacokinetic studies (226). For treatment of adolescent males,
physiological replacement can be approximated by modifying the dose and duration
of Androderm application to mimic patterns of nocturnal testosterone secretion ob-
served during puberty. A clinical audit was reported on the acceptability and efficacy
as a treatment for hypogonadism of the first transdermal testosterone therapy avail-
able in the United Kingdom (Andropatch), compared with existing androgen replace-
ment options (227). Serum testosterone and questionnaire data on treatment efficacy,
side effects, therapy preference, sexual dysfunction, and partner’s attitudes to therapy
were obtained from 50 hypogonadal men prescribed long-term testosterone replace-
ment. Eighty percent returned analyzable questionnaires and, of these, 84% experi-
enced adverse effects with TD therapy, usually dermatological problems. Twenty-
two percent elected to continue with TD therapy, 72% returned to depot, and 5% to
oral therapy. The reservoir patches were judged to be too large, uncomfortable, vi-
sually obtrustive and noisy; thus, the pharmacokinetic advantages were largely out-
weighed by low patient acceptability. The pharmacokinetics of three testosterone-
containing TDS were evaluated in healthy male volunteers (228). Type 1 and 2 were
nonscrotal membrane patches differing in adhesive type. Six subjects were treated
with low-dose testosterone type 1, high-dose testosterone type 1, and low-dose tes-
tosterone type 2. To eliminate the influence of endogenous serum testosterone, se-
cretion was suppressed by the GnRH antagonist cetrorelix. Physiological testosterone
levels were achieved during the 24-h application period. Maximal serum levels were
achieved after 4 h with both types, and both enabled a physiological circadian profile
to be achieved.
Effects of TD testosterone replacement therapy using a permeation-enhanced
system on plasma lipolytic enzymes (hepatic and lipoprotein lipase), LDL, and HDL
subfraction concentrations were assessed (229). Ten patients with primary testicular
failure were started on Testoderm therapy and evaluated before and after 3 months
of treatment. Serum testosterone level increased to within the normal range in all
subjects, whereas serum DHT increased to supranormal values. Plasma hepatic lipase
(HL) activity increased after testosterone replacement (24.7 ⫾ 7.5 vs. 29.2 ⫾ 8.3
␮mol free fatty acid released per hour) and the increase in HL correlated with the
increase in DHT. No significant change was seen in the HDL2 subfraction but HDL3
decreased after treatment (0.93 ⫾ 0.17 vs. 0.79 ⫾ 0.14 mmol/L). Whether these
changes adversely influence the cardiovascular risk in the long term remains to be
determined.
Local skin reactions at application site are common adverse events with trans-
dermal testosterone and an open-label, controlled pilot study evaluated whether top-
ical pretreatment with triamcinolone acetonide, 0.1% cream, reduced the incidence
or severity of chronic skin irritation in health volunteers (230). At all assessment
points, more subjects had lower cumulative scores with pretreatment than without
pretreatment.
E. Fentanyl
Transdermal fentanyl has been used widely in the United States since it was approved
in 1990 (231). The first clinical report of TD fentanyl therapy in cancer pain involved
five patients. Pain relief was established with intravenous fentanyl and a TDS se-
lected to deliver the same hourly dose while the intravenous infusion was tapered
over 6 h. The TDS was changed every 24 h for a total of 3–156 days. The initial
study demonstrated steady-state plasma levels were linearly related to the fentanyl
dose. A multicenter trial was conducted in 39 patients. The TD fentanyl dose was
established from a conversion table based on the dose of oral immediate-release
morphine required to control pain. The fentanyl patches were changed every 72 h
and immediate-release morphine used on an as-required basis for incidental pain.
This trial further demonstrated that patients could be converted from oral morphine
to an equianalgesic dose of transdermal fentanyl and that pain relief could be main-
tained for a lengthy time period on an outpatient basis.
The analgesic, pharmacokinetic, and clinical respiratory effects of 72-h appli-
cation of two TD fentanyl patch sizes in patients undergoing abdominal hysterectomy
were evaluated (232). Fentanyl TDS, releasing 50 ␮g/h (TF-50) or 75 ␮g/h (TF-75)
fentanyl or placebo patches were applied to 120 women 2 h before abdominal hys-
terectomy under general anesthesia. All patients had postoperative access to supple-
mental morphine using patient-controlled pumps. VAS pain scores, supplementary
analgesia, fentanyl plasma concentration, continuous hemoglobin saturation, respi-
ratory pattern, and adverse effects data were collected. Visual analogue scale (VAS)
pain scores and supplemental morphine use significantly decreased in the TF-75
group in the postanesthesia care unit, and for both TF-50 and TF-75 groups for 8–
48 h postoperatively. Between 5 and 36 h, the TD fentanyl groups had significantly
increased abnormal respiratory patterns, including apneic episodes and episodes of
slow respiratory rate, and a significantly increased requirement for oxygen supple-
mentation. Nine patients in the TD fentanyl groups were withdrawn because of severe
respiratory depression, but none in the placebo group. Although fentanyl plasma
concentration were higher in the TF-75 than in TF-50 group, differences were not
significant. Fentanyl plasma concentration decreased significantly 48 h after patch
application. Although good analgesia was the result of this combination therapy, it
was associated with a high incidence of respiratory depression requiring intensive
monitoring, oxygen supplementation, patch removal in about 11% of patients, and
opioid reversal with naloxone in about 8% of patients.
It is generally recommended that patients should be titrated with a short-acting
narcotic to control their cancer pain before they are converted to a fentanyl TDS.
However, immediate TD fentanyl therapy in patients with uncontrolled cancer pain
with direct titration of fentanyl TDS according to clinical necessity on a day-to-day
basis, with the availability of morphine solution for rescue medication has been
evaluated (233). The major objective was simplification of therapy. On average,
sufficient pain control was reached within 48 h of the start of TD fentanyl, and VAS
values at all follow-up times were significantly lower than pretreatment values. Mean
fentanyl doses were 70 ␮g/h (week 1), 98 ␮g/h (week 2), 107 ␮g/h (week 3) and
116 ␮g/h (week 4). Mean morphine doses as rescue medication steadily decreased
from 11 mg/day in week 1 to 3 mg/day in week 4 of treatment, although differences
were not significant. It was concluded that titration with a short-acting narcotic before
conversion to TD fentanyl was not necessary, provided the patients were well
monitored.
A comparison of TD fentanyl with placebo for postoperative analgesia in or-
thopedic surgery was reported (234). Forty adult patients had general anesthesia with
propofol, isoflurane in nitrous oxide, and oxygen, and small bolus of alfentanil or
sufentanil. Preoperatively, one group received TD fentanyl (75 ␮g/h) for 72 h, and
the second group a placebo patch. Morphine was given postoperatively as necessary.
Eleven patients receiving fentanyl needed morphine, compared with 19 in the placebo
group, and mean morphine dose was significantly lower in the fentanyl group. One
fentanyl group patient had decreased oxygen saturation and intense sedation, neces-
sitating administration of naloxone. The mean maximum plasma fentanyl concentra-
tion was 1.63 ng/mL.
An alternative fentanyl TDS was evaluated (235). Male adult surgical patients
received 650 or 750 ␮g of fentanyl intravenously as part of the induction of anes-
thesia, and plasma fentanyl concentrations were measured over the following 24 h.
On the first postoperative day a fentanyl TD was placed on the upper torso for 24
h and then removed. Plasma fentanyl concentrations were measured for 72 h after
device application and each patient’s clearance and unit disposition function were
determined. During the 72 h after application of the TDS, the amount of fentanyl
absorbed and the absorption rate were also determined. Residual fentanyl in the
transdermal device was measured to determine absolute bioavailability. Of 14 sub-
jects receiving TD fentanyl, 3 had clinically significant fentanyl toxicity, mandating
early removal of the TDS. In the remaining subjects the fentanyl concentration from
12 to 24 h varied over a 20-fold range (0.34–6.75 ng/mL). In subjects wearing the
device for 24 h, terminal half-life of fentanyl after device removal was 16 h. Bio-
availability of transdermally administered fentanyl was 63 ⫾ 35% and rate of fen-
tanyl absorption from 12 to 2 h in subjects still wearing the device ranged from 10
to 230 ␮g/h. In two subjects, the rate within the first 6 h briefly exceeded 300 ␮g/
h and both demonstrated fentanyl toxicity, requiring early device removal. It was
concluded that the Cygnus transdermal fentanyl device produced more highly vari-
able plasma fentanyl concentrations than those reported for the Duragesic device,
which is contraindicated for postoperative analgesia.
The use of transdermal fentanyl was evaluated in ten patients aged 9–16 years
with sickle cell pain crisis (236) who received TD fentanyl at 25 (n = 7) or 50 (n =
3) ␮g/h and morphine use was >2.5 mg/h. Average TD fentanyl dose was 0.77 ⫾
0.37 ␮g/kg h⫺1 on day 1 and 1.17 ⫾ 0.46 ␮g/kg h⫺1 on day 2. The fentanyl con-
centration at 24 and 48 h was 0.60 ⫾ 0.31 and 1.18 ⫾ 0.44 ng/mL, respectively.
There was a significant relation between dose and fentanyl concentration and no
difference in any clinical-monitoring parameter between day 1 and day 2, although
seven of ten patients reported subjective improvement in pain control. No adverse
effects were noted, but it was concluded that improved understanding of dose–effect
relations for TD fentanyl in children and adolescents was necessary before adequate
pain control could be achieved by this route.
At the start of TD fentanyl treatment, depot accumulation of the drug within
skin results in a significant delay (17–48 h) before achievement of maximum plasma
concentration (237). However, concomitant use of short-acting morphine maintained
pain relief during the titration period, and supplementary medication decreased with
duration of TD fentanyl treatment. Patient preference for TD fentanyl was indicated
by the number of patient requests for continued use at the end of the study (up to
95%). Although postoperative TD fentanyl is contraindicated, supplementary patient-
controlled analgesia was significantly reduced in patients receiving 75 ␮g/h TD fen-
tanyl, compared with placebo. Some patients with previously uncontrolled pain be-
came completely pain-free. The most common adverse events during TD fentanyl
therapy included vomiting, nausea, and constipation. Most serious adverse event was
hypoventilation, which occurred more frequently in postoperative (4%) than in cancer
patients (2%). In surgical patients, fentanyl-associated respiratory events usually oc-
curred within 24 h of patch application, although there were isolated reports of late
onset (up to 36 h).
Respiratory depression was observed in a 2-year-old boy unintentionally ex-
posed to a fentanyl TDS (238). He was found unresponsive after sleeping in bed
with his grandmother and, after intubation and ventilation, a fentanyl TDS was dis-
covered on his back. Removal of the device and treatment with naloxone resolved
symptoms. This was the first reported case of secondary exposure to a fentanyl patch
causing clinically significant respiratory depression in the pediatric population, and
it emphasized a new hazard.
A multicenter evaluation of TD fentanyl for cancer pain relief in 53 patients
who required 45 mg or more of oral morphine daily was reported (239). After 1-
week stabilization on oral morphine, patients were transferred to an appropriate dose

of TD fentanyl (25, 50, 75, or 100 ␮g/h) administered by patch every 3 days. TD
fentanyl was titrated to pain relief and patients were followed for up to 3 months.
Mean duration of TD fentanyl use was 58 ⫾ 32 days. Mean daily morphine dose
during the last 2 days of stabilization was 189 ⫾ 20 mg, and mean initial fentanyl
patch dose was 58 ⫾ 6 ␮g/h. The mean daily morphine dose taken as required for
breakthrough pain at study completion was 35 mg. Mean final fentanyl dosage at
study completion was 169 ⫾ 29 ␮g/h. Pain relief was rated as good or excellent by
82% of patients during the treatment period and 63% preferred TD fentanyl to their
last analgesic therapy. Side effects related to the patch were nausea (13%), vomiting
(8%), skin rash (8%), and drowsiness (4%). Thirty percent of patients reported ad-
verse experiences related to the patch, and 17% had to be discontinued.
Pain-related treatment satisfaction, patient-perceived side effects, functioning,
and well-being in patients with advanced cancer receiving TD fentanyl (Duragesic)
or sustained-release oral morphine (MS Contin or Oramorph SR) were compared
(240). Patients were more satisfied with TD fentanyl than sustained-release oral mor-
phine. Patients receiving TD fentanyl were more satisfied overall with their pain
medication and also reported a significantly lower frequency and effect of a pain
medication’s side effects. Measures of pain intensity, sleep adequacy, and symptoms
demonstrated no significant differences between treatment groups.
Forty-eight patients with noncancer neuropathic pain, who had participated in
a randomized controlled trial with intravenous fentanyl infusions, next received pro-
longed TD fentanyl (241). Eighteen patients stopped prematurely owing to inade-
quate pain relief, side effects, or both. Pain relief among the remaining 30 patients
completing the 12-week–dose titration protocol was substantial in 13 and moderate
in 5, and quality of life improved by 23%. Psychological dependence or induction
of depression was not observed, and tolerance emerged in only 1 patient. There was
a significant positive correlation between pain relief obtained with intravenous and
prolonged TD fentanyl. It was concluded that long-term TD fentanyl may be effective
in noncancer neuropathic pain without clinically significant management problems.
Testing with intravenous fentanyl may assist in selecting neuropathic pain patients
who can benefit from treatment by the transdermal route.
Disposition of fentanyl in dogs after intravenous and TD administration was
investigated (242). Each of six clinically normal beagles received intravenous fen-
tanyl (50 ␮g/kg body weight) and TD fentanyl (50 ␮g/h). TD fentanyl produced
average steady-state concentrations of 1.6 ng/mL. Actual delivery rate of fentanyl
was 27.5–99.6% of the theoretical rate. Mean elimination half-life after patch re-
moval was 1.39 h. It was concluded that TD fentanyl had clinically useful potential
as an analgesic in dogs. Plasma fentanyl concentrations were obtained in six intact,
mixed-breed adult dogs (two males, four females), weighing 19.9 ⫾ 3.4 kg, after
application of three sizes of fentanyl TDS delivering 50, 75, or 100 ␮g/h (243).
Results are summarized in Table 7.
TD fentanyl was evaluated in children with cancer pain (244) and measures of
analgesia, side effects, and skin changes obtained for a minimum of two doses (6
treatment days). Treatment was well tolerated and 10 of 11 patients continued the
treatment. Time to peak plasma concentration ranged from 18 to more than 66 h in
patients using a 25-␮g/h patch. Compared with published data from adults, mean
Table 7 Comparison Between Three Transdermal Fentanyl Patches in Dogs
Fentanyl delivery Plasma fentanyla Total AUC Elimination

(␮g/h) (ng/mL) (ng/h mL⫺1) half-life (app) (h)
50 0.7 ⫾ 0.2 46 ⫾ 12.2 3.6 ⫾ 1.2

75 1.4 ⫾ 0.5 101.2 ⫾ 41.4 3.4 ⫾ 2.7
100 1.2 ⫾ 0.5 80.4 ⫾ 38.3 2.5 ⫾ 2.0
a
Levels between 24 and 72 h.
Source: Ref. 243.
clearance and volume of distribution of transdermal fentanyl were equal, but vari-
ability was less.
It has been claimed that fentanyl patches are less suitable for patients who are
elderly or terminally ill and dying. However, a retrospective survey of 205 cancer
patients who died within hospital-based home care in Norrkoping, Sweden between
January 1997 and June 1998 reported that 34 patients used fentanyl TD, and in 30
it was possible to evaluate analgesic efficacy. Estimated efficacy was good (93%) or
moderate (88%) in patients younger or older than 65 years of age (245).
Constipation and use of laxatives were investigated in patients with chronic
cancer pain treated with oral morphine and TD fentanyl (246). Of 46 patients, treated
with slow-release morphine, 30–1000 mg/day for 6 days, 39 were switched to TD
fentanyl (0.6–9.6 mg/day) with a conversion ratio of 100:1. Median fentanyl doses
increased from 1.2 to 3.0 mg/day throughout the 30-day TD treatment period. Mean
pain intensity decreased slightly after conversion, although the number of patients
with breakthrough pain or who requiring immediate-release morphine as a rescue
medication was higher with TD fentanyl. The number of patients with bowel move-
ments did not change after the opioid switch, but the number of patients taking
laxatives was significantly reduced from 78–87% of patients per treatment day (mor-
phine) to 22–48% (fentanyl). Constipation is an almost universal side-effect of pro-
longed opioid analgesia, and resulting discomfort can be more severe than the pain
itself, leading to reduction of analgesic use and consequently increased pain (247).
Because of higher lipophilicity, fentanyl penetrates the blood–brain barrier more
easily and lower doses are possible; thus, comparatively less opioid is available in
the gastrointestinal tract to block local receptors. The cost of treating constipation
among terminally ill cancer patients receiving TD fentanyl and 12-h sustained-release
morphine in Ontario, Canada was examined (248). Cost of managing constipation
in a patient receiving TD fentanyl and 12-h sustained-release morphine was estimated
at 31.77 and 52.76 dollars, respectively, during the first 2 weeks of treatment. When
the acquisition costs of opioids are included, the two-week cost of managing a patient
with TD fentanyl and 12-h sustained-release morphine was 123.24 and 119.70 dol-
lars, respectively. It was concluded that acquisition costs alone should not dictate
treatment, and care should always be tailored to the needs and preferences of indi-
vidual patients.
The issues involved in switching opioids to transdermal fentanyl in a clinical
setting have been evaluated (249). Case records of patients treated with TD fentanyl
were retrospectively examined and conversion ratios calculated. Opioid therapy was
switched to TD fentanyl during inpatient treatment for 53 patients and during out-
patient treatment for 11 patients. Before conversion patients were treated with slow-
release morphine (48%), immediate-release morphine (17%), buprenorphine (11%),
tramadol (11%), levomethadone (5%), tilidine/naloxone (5%), and piritramid (3%).
Reasons for opioid rotation included inadequate pain relief (33%), patients’ wish to
reduce oral medication (20%), gastrointestinal side effects (31%), vomiting (13%),
constipation (19%), and dysphagia (27%). Reduction of side effects was reported by
10 of 19 patients. In 12 of 21 patients, when medication was switched because of
inadequate pain relief, reduction in pain intensity was reported. It was concluded
that conversion to TD therapy may readjust the balance between opioid analgesia
and side effects.
F. Anti-inflammatories
A multicenter study of comparative efficacy, tolerability, and acceptability of a flur-
biprofen local-action transcutaneous (LAT) patch (40 mg b.d.) and piroxicam gel (3
cm, 0.5% q.d.s), was conducted in general practice in the United Kingdom in 137
men and women with soft-tissue rheumatism of the shoulder or elbow tag, epicon-
dylitis, tendinitis, bursitis, or adhesive capsulitis (250). Patients received one therapy
for 4 days before crossing over for a further 4 days, followed by 6 days of their
preferred therapy. Assessment of severity of pain, tenderness, and overall clinical
condition was carried out at baseline and at 4, 8, and 14 days. There was a significant
reduction in severity of pain in favor of flurbiprofen LAT (42 vs. 26%). Eligible
dataset analysis revealed significant differences, in favor of flurbiprofen LAT, in
severity of lesion tenderness and overall change in clinical condition. Superior effi-
cacy was also indicated at the end of the crossover phase when 69% chose to con-
tinue treatment with flurbiprofen LAT. There were also significant differences in
favor of flurbiprofen LAT in assessments for night pain, quality of sleep, and patients’
overall opinion of treatment.
G. Clonidine
Pharmacokinetic and pharmacodynamic properties and safety, of a TD clonidine
system (M-5041T) were evaluated after single and repeated applications (251). In
the single-application study, one patch delivering 4, 6, or 8 mg was applied for 3
days to eight healthy subjects. In a repeated-application study, A (0–72 h), B (72–
144 h), and C (144–216 h) TD systems delivering 6 mg were applied in seven
healthy subjects. In the single-application study, plasma clonidine level, Cmax and
AUC increased in a dose-dependent manner, but not significantly. The blood pressure
(BP)-lowering effect of 8 mg was greater than that of 4 and 6 mg. Adverse effects
were reported, but did not cause withdrawal. In the repeated-application study,
plasma clonidine increased up to 48 h after application of patch A, and remained
stable until removal of patch C. Cmax and AUC did not differ significantly. During
an active period, BP decreased significantly during treatment, but BP at midnight
did not change significantly. Mild erythema and systemic adverse effects were re-
ported. In a second study (252), TD clonidine (M-5041T) was compared with oral
clonidine (Catapres TTS). One TDS containing 6 mg of clonidine was applied on
the right chest for 3 days or one tablet of Catapres TTS (0.075 mg) was given orally
every 12 h for 3 days in eight healthy subjects. Plasma clonidine concentration
increased gradually after application of TDS and decreased gradually after removal,
whereas it increased rapidly then decreased rapidly after each dose of Catapres TTS.
Elimination half-life of clonidine after patch removal was significantly greater than
after the final dose of Catapres TTS. There was no significant difference in Cmax,
AUC, or BP-lowering between the treatments. Adverse symptoms occurred more
frequently during Catapres TTS therapy and were observed when plasma clonidine
concentration was relatively high.
Clinical pharmacological evaluation of a clonidine TDS was undertaken in a
variety of phase I volunteer studies to assess potential effects of site of application,
pharmacokinetic linearity, and absolute bioavailability (253). Delivery of clonidine
from Catapres TTS was influenced by application site. The order of permeability
was chest > upper arm > upper thigh. Pharmacokinetics of clonidine following TD
delivery were linear with slight nonlinearity between devices of differing area. Ab-
solute bioavailability of clonidine from Catapres TTS was approximately 60% and
calculated in vivo absorption rate (4.32 ⫾ 1.68 ␮g/h) was in good agreement with
claimed performance (0.1 mg/day).
Premedication with oral and TD clonidine in postoperative sympatholysis was
investigated in 61 patients undergoing elective major noncardiac surgery (254). The
treatment group were premedicated with a clonidine TDS (0.2 mg/d), applied the
night before surgery and left in place for 72 h, and were given 0.3-mg oral clonidine
60–90 min before surgery. Clonidine reduced enflurane requirements, intraoperative
tachycardia, and myocardial ischemia (1 of 28 clonidine patients vs. 5 of 24 placebo).
However, postoperatively, the heart rate decreased for only the first 5 h, and incidence
of postoperative myocardial ischemia (6 of 28 clonidine vs. 5 of 26 placebo) did not
differ. Clonidine significantly reduced plasma levels of epinephrine and norepineph-
rine on the first postoperative morning. It was suggested that larger doses of clonidine
should be investigated for their ability to decrease postoperative tachycardia and
myocardial ischemia. Surgical trauma induces diffuse sympathoadrenal activation
that contributes to perioperative cardiovascular complications in high-risk patients.
Regional anesthetic and analgesic techniques can attenuate this ‘‘stress response’’
and reduce rates of adverse perioperative events, but their postoperative use is lo-
gistically difficult and expensive. Hence, use of transdermal clonidine to blunt the
stress response throughout the perioperative period was evaluated (255). Forty pa-
tients scheduled for major upper abdominal surgery were entered in a clinical trial.
Patients received either clonidine (0.2 mg orally plus clonidine TTS-3 patch the
evening before surgery plus 0.3 mg orally on call to operating room) or matched
oral and transdermal placebo. Preoperative transdermal (plus oral) clonidine admin-
istration resulted in therapeutic plasma clonidine concentrations over the periopera-
tive period (1.54 ⫾ 0.07 ␮g/mL) and reduced preoperative epinephrine and nor-
epinephrine levels by 65%. Plasma catecholamines increased in both groups after
surgery, but were markedly lower over the postoperative period in patients receiving
clonidine (who also had reduced frequency of postoperative hypertension).
A study was performed to determine if a lower than previously reported oral–
transdermal clonidine regimen could reduce postoperative morphine requirements
without producing systemic side effects (256). Twenty-nine healthy females under-
going elective abdominal hysterectomy received preoperative oral clonidine, 4–6 ␮g/
kg plus a 7 cm2 clonidine TDS (0.2 mg/24 h), or a placebo tablet and patch. Low-
dose clonidine had no potentiating effect on morphine analgesia and postoperative
morphine use, VAS pain scores and morphine plasma levels were similar between
the groups. The clonidine group experienced a significantly higher incidence of in-
traoperative and postoperative hypotension and bradycardia than the control group,
but no differences were noted in incidence of nonhemodynamic side effects.
Clinical and pharmacological indications suggested reduction of noradrenergic
tone occurs in cluster headache, during both active and remission periods, but that
sharp fluctuations of the sympathetic system may trigger the attacks (257). It was
postulated that continuous administration of low-dose clonidine could be beneficial
in the active phase by antagonizing variations in noradrenergic tone. After a run-in
week, TD clonidine (5–7.5 mg) was administered for 1 week to 13 patients suffering
from episodic or chronic cluster headache. During clonidine treatment, mean weekly
frequency of attacks reduced from 17.7 ⫾ 7.0 to 8.7 ⫾ 6.6, pain intensity of attacks
(VAS) from 98.0 ⫾ 7.2 to 41.1 ⫾ 36.1 mm, and duration from 59.3 ⫾ 21.9 to 34.3
⫾ 24.6 min. This strongly suggested that TD clonidine may be effective in the
preventive treatment of cluster headache.
To test a previous clinical observation that approximately 25% of patients with
painful diabetic neuropathy appeared responsive to clonidine, a formal clinical trial
of TD clonidine was conducted (258). In stage 1, 41 patients completed a random-
ized, three-period crossover comparison of TD clonidine (titrated from 0.1 to 0.3
mg/day) to placebo patches. Twelve apparent responders from stage 1 were entered
into the second stage, consisting of an additional four double-blind, randomized, 1-
week treatment periods with TD clonidine and placebo. Stage 1 showed that, for the
total group, pain intensity differed little between clonidine and placebo. In stage 2,
however, the 12 apparent responders from stage 1 had 20% less pain with clonidine
than placebo (95% CI: 4–35% pain reduction), confirming that their pain was re-
sponsive to clonidine. Post hoc analysis suggested patients describing their pain as
sharp and shooting were more likely to respond to clonidine. These results support
the hypothesis that there is a subset of patients with painful diabetic neuropathy who
benefit from systemic clonidine administration and also demonstrate the value of an
enriched enrollment technique in analgesic trials.
Efficacy and tolerability of TD clonidine in inner-city African-American and
Hispanic-American patients with essential hypertension was evaluated in multiple
community-based primary care centers in a 12-week trial (259). Patients with dia-
stolic BP higher than 90 mmHg were given TD clonidine (0.1 or 0.2 mg daily). The
drug was titrated after 1 month if diastolic BP exceeded 90 mmHg. At 12 weeks,
change in blood pressure, adverse effects, and patient satisfaction were assessed.
Transdermal clonidine significantly lowered BP by 15.7/12.8 ⫾ 18.1/9.6 mmHg, and
heart rate by 3 ⫾ 9 beats per minute. There were no differences in BP reduction
according to race, ethnicity, gender, or age. Most common adverse effects were
pruritus or discomfort at the patch site, dizziness, dry mouth, and fatigue (11%
discontinued owing to adverse effects). Sixty-seven percent of patients reported that
TD clonidine was more convenient than oral therapy.
H. Miscellaneous
Intraocular pressure (IOP)-lowering effects of a TD system containing pilocarpine
that was designed to avoid common side effects in glaucoma treatment with con-
ventional eye drops were reported (260). Two patches, each containing 30 mg of
pilocarpine or placebo, were applied to the supraclavicular skin of 24 patients. IOP

was recorded before (22.7 ⫾ 5.8 mmHg) and at ⫹12, 16, and 20 h after application.
Pilocarpine TD did not significantly reduce IOP although there were detectable
plasma levels of pilocarpine at 12 (2.9 ng/mL) and 20 h (1.3 ng/mL) after ad-
ministration.
The therapeutic effects of selective cholinergic replacement using oral xano-
meline, an M1/M4 receptor agonist, were assessed in a multicenter study of 343
patients with Alzheimer’s disease (261). Improvement in ADAS-Cog provided clin-
ical evidence of involvement of the M1 muscarinic receptor in cognition, and the
favorable effects of xanomeline on disturbing behavior suggested a novel approach
for treatment of noncognitive symptoms. Although adverse effects (mainly gastro-
intestinal) associated with the oral formulation appeared to limit its use, a large-scale
study investigating the safety and efficacy of transdermal xanomeline is underway.
The effects of TD scopolamine on pulmonary function, symptoms, and bron-
chial hyperresponsiveness to methacholine were reported (262). The first study eval-
uated therapeutic effects of a single scopolamine TD system (Scopoderm TTS, 2.5
cm2) on forced expiratory volume in 1 s (FEV1), reversibility, peak expiratory flow
(PEF), and symptoms in ten patients with reversible airways disease. The drug was
adequately taken up into the systemic circulation, but no significant clinical effects,
nor correlations between scopolamine levels and outcome parameters, were observed.
Because of the possibility of subtherapeutic doses, a second study with two scopol-
amine TD systems was carried out in ten patients with bronchial hyperresponsiveness
to methacholine. Blood and urine concentrations of free scopolamine were doubled
compared with the first study, but there were still no significant effects on FEV1,
PEF, symptoms, and bronchial hyperresponsiveness, although most of the patients
now reported adverse side effects. It was concluded that TD administration of sco-
polamine was not clinically useful in asthma and chronic obstructive pulmonary
disease.
Efficiency of TD scopolamine in prophylaxis of postoperative nausea and vom-
iting (PONV) after otoplasty was evaluated in a post hoc assessment of data (263).
Of 50 otoplasty patients 25 received a scopolamine patch before general anesthesia.
The placebo group received intravenous atropine (10 ␮g/kg) during induction. Sco-
polamine-treated patients suffered more from moderate preoperative bradycardia (8
of 25) than atropine-treated patients (1 of 25). After unilateral otoplasty, no scopol-
amine-treated patients, but 50% of atropine-treated patients, suffered from PONV.
After bilateral otoplasty, respective incidences were 39 and 81%. After unilateral
otoplasty no patient needed droperidol, but after bilateral otoplasty, 12 of 19 atropine-
treated and 4 of 18 scopolamine-treated patients needed droperidol. It was concluded
that TD scopolamine offers effective prophylaxis against PONV, but does not protect
from bradycardia in otoplasty. The effects of TD scopolamine (1.5 mg) and oral
cyclizine (50 mg) on postural sway, optokinetic nystagmus, and circularvection in
humans were investigated (264). Neither scopolamine nor cyclizine (at doses used
for relief of motion sickness) had a significant suppressive effect on these aspects of
visual–vestibular interaction.
A pilot clinical trial of TD administration of selegiline in HIV-positive patients
was performed to obtain preliminary data on safety, tolerability, and effect on HIV-
associated cognitive impairment (265). Both selegiline and placebo were well tol-
erated and improvements favoring the selegiline group were suggested on single tests
of verbal memory and motor–psychomotor performance.
TD patches of diltiazem hydrochloride were formulated using ethyl cellulose
and polyvinylpyrrolidone. Pharmacodynamic and pharmacokinetic performance of
diltiazem hydrochloride after TD administration was compared with that of oral
administration in male New Zealand albino rabbits (266). Pharmacokinetic param-
eters were significantly different between transdermal and oral administration. Ter-
minal elimination half-life of TD-delivered diltiazem was similar to that of oral
administration. Skin irritation studies indicated no recognizable changes after patch
removal. It was concluded that relative bioavailability of diltiazem hydrochloride and
therapeutic activity were fivefold higher after TD than after oral administration.
Transdermal histamine has been successfully used for amelioration of symptoms of
both relapsing–remitting and progressive multiple sclerosis (267). Of the 55 patients
using TD histamine cream 67% had improvements in one or more areas (extremity
strength, balance, bladder control, fatigue, activities of daily living, and cognitive
functioning) sustained for up to 3 months, and 33% had improvements in three or
more areas.
The occlusive properties of a range of hydrocolloid patches on the penetration
of triamcinolone acetonide, and hydration, were assessed in vivo using visual as-
sessment (268). There was a significant difference in rates of hydration of patches
containing either NaCMC 39%, or pectin 39%, although changes in hydrocolloid
composition did not significantly alter the blanching response.
The feasibility of use of water-activated, pH-controlled silicone reservoir de-
vices for TD administration was investigated using timolol maleate (269). Two ti-
molol patches were applied to the arm of 12 volunteers for 81 h and absorption
compared with that from an oral tablet formulation (Blocanol). In vivo plasma levels
were also compared with those predicted by kinetic simulations. Both steady-state
timolol concentrations in plasma and duration could be controlled with water-acti-
vated, pH-controlled patches, although considerable, interindividual variability in TD
absorption occurred owing to the high fractional skin control in timolol delivery.
Timolol patches were well tolerated, skin irritation was mild, and after removal of
the patches, skin changes were practically reversed in 24 h. Preclinical tests were
also reported (270) including release of timolol from patches and timolol permeation
across human cadaver skin.
Systemic effects, local tolerance, and effectiveness on the penis of topical gel
formulations containing alprostadil (prostaglandin E1) plus 5% of the penetration
enhancer SEPA versus SEPA alone (placebo) were evaluated in men with erectile
dysfunction (271). Application of prostaglandin E1 gel correlated positively with
erectile response (67–75% of patients had an erection compared with 17% of
controls.
A soft, stretchy adhesive patch (10 ⫻ 14 cm) containing 5% lidocaine (700
mg) has been used for topical treatment of pain associated with postherpetic neuralgia
(272). Systemic absorption was reported as minimal in healthy subjects and patients
with postherpetic neuralgia (3% of dose). In clinical trials (12 h–24 days), treatment
resulted in a significant reduction in pain intensity and increased pain relief compared
with vehicle patch.
Transdermal drug delivery systems are increasingly popular, yet few data exist
on emergency medical outcomes after exposures. A retrospective study (273) using
data collected through a Regional Poison Information System identified 61 cases

over a 5-year period. Routes were dermal (48), oral (10), combined oral and dermal
(1), parenteral use of gel residue (1), and combined oral and parenteral (1). Most
exposures (72%) were managed by home telephone consultation only. Eleven of 17
patients (18%) evaluated in health care facilities were admitted, including 8 (13%)
to intensive care units. Hospital admission correlated statistically with clonidine,
fentanyl, oral exposure, and drug abuse. Clonidine exposure also correlated statisti-
cally with intensive care admission. One fatality was recorded; all other patients
recovered uneventfully.
In this chapter we have described some of the considerations that are important in
the design and development of pharmaceutical products intended for application to
the skin. In addition, we have described some recent clinical observations of trans-
dermal therapeutic systems. Space limitations dictated that we could not provide an
exhaustive review of all factors, and the reader will appreciate that preformulation,
scale-up, and safety are not covered. However, these aspects are fully covered either
elsewhere in this volume or in some of the recent excellent and fully recommended
texts (7,8,83). We have attempted to share our knowledge of the formulation types
that are applied to the skin, together with their properties, problems, and clinical
usefulness. We hope that our comments will provide the novice formulator with some
insights borne of experience, and the experienced formulator with some novel in-
sights in the field of dermatological and transdermal products development and
their use.
REFERENCES
1. Scheuplein RJ, Blank IH. Permeability of the skin. Physiol Rev 1971; 51:702–747.
2. Idson B. Percutaneous absorption. J Pharm Sci 1975; 64:901–924.
3. Flynn GL. (1996). Cutaneous and transdermal delivery: processes and systems of de-
livery. In: Banker GS, Rhodes CT, eds. Modern Pharmaceutics, 3rd ed. New York:
Marcel Dekker. 1996; 239–298.
4. Barry BW. Dermatological Formulatins, Percutaneous Absorption. New York: Marcel
Dekker. 1983.
5. Shah VP, Maibach HI, eds. Topical Drug Bioavailability, Bioequivalence, and Penetra-
tion. New York: Plenum Press. 1993.
6. Schaefer H, Redelmeir TE. Skin Barrier, Principles of Percutaneous Absorption. Basel:
Karger, 1996.
7. Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity Assessment. New York:
Marcel Dekker. 1998.
8. Bronaugh RL, Maibach HI, eds. Percutaneous Absorption, Drugs, Cosmetics, Mecha-
nisms, Methodology. 3rd ed. New York: Marcel Dekker. 1999.
9. Elias PM. Lipids and the epidermal permeability barrier. Arch Dermatol Res 1981;
270:95–117.
10. Raykar PV, Fung MC, Anderson BD. The role of protein and lipid domains in the
uptake of solutes by human stratum corneum. Pharm Res 1988; 5:140–150.
11. Walters KA, Hadgraft J, eds. Pharmaceutical Skin Penetration Enhancement. New York:
12. Smith EW, Maibach HI, eds. Percutaneous Penetration Enhancers. Boca Raton: CRC
Press, 1995.
13. Behl CR, Flynn GL, Barrett M, Walters KA, Linn EE, Mohamed Z, Kurihara T, Ho
NFH, Higuchi WI, Pierson CL. Permeability of thermally damaged skin II: immediate
influences of branding at 60⬚C on hairless mouse skin permeability. Burns 1981; 7:
389–399.
14. Flynn GL, Behl CR, Walters KA, Gatmaitan OG, Wittkowsky A, Kurihara T, Ho NFH,
Higuchi WI, Pierson CL. Permeability of thermally damaged skin III: influence of
scalding temperature on mass transfer of water and n-alkanols across hairless mouse
skin. Burns 1982; 8:47–58.
15. Bronaugh RL, Stewart RF. Methods for in vitro percutaneous absorption studies. V.
Permeation through damaged skin. J Pharm Sci 1985; 74:1062–1066.
1986; 86:201–207.
17. Scott RC. In vitro absorption through damaged skin. In: Bronaugh RL, Maibach HI,
eds. In Vitro Percutaneous Absorption, Principles, Fundamentals and Applications.
Boca Radon: CRC Press. 1991; 129–135.
18. Bond JR, Barry BW. Damaging effect of acetone on the permeability barrier of hairless
mouse skin compared with that of human skin. Int J Pharm 1988; 41:91–93.
19. van der Valk PGM, Nater JP, Bleumink E. Vulnerability of the skin to surfactants in
different groups of eczema patients and controls as measured by water vapour loss.
Clin Exp Dermatol 1985; 10:98–103.
20. Turpeinen M, Salo OP, Leisti S. Effect of percutaneous absorption of hydrocortisone
on adrenocortical responsiveness in infants with severe skin disease. Br J Dermatol
1986; 115:475–484.
21. Turpeinen M. Absorption of hydrocortisone from the skin reservoir in atopic dermatitis.
Br J Dermatol 1991; 124:358–360.
22. Fartasch M, Diepgen TL. The barrier function in atopic dry skin. Acta Derm Venereol
1992; 176(suppl):26–31.
23. Reichek N, Goldstein RE, Redwood DR. Sustained effects of nitroglycerin ointment
in patients with angina pectoris. Circulation 1974; 50:348–352.
24. Cleary GW. Transdermal delivery systems: a medical rationale. In: Shah VP, Maibach
HI, eds. Topical Drug Bioavailability, Bioequivalence, and Penetration. New York:
Plenum Press, 1993; 17–68.
25. Anderson BD. Prodrugs and their topical use, In: Shah VP, Maibach HI, eds. Topical
Drug Bioavailability, Bioequivalence, and Penetration. New York: Plenum Press. 1993;
69–89.
26. Ahmed S, Imai T, Otagiri M. Stereoselective hydrolysis and penetration of propranolol
prodrugs: in vitro evaluation using hairless mouse skin. J Pharm Sci 1995; 84:877–
883.
27. Lai PM, Roberts MS. Iontophoresis. In: Roberts MS, Walters KA, eds. Dermal Ab-
sorption and Toxicity Assessment. New York: Marcel Dekker. 1998; 371–414.
28. Prausnitz MR. The effects of electric current applied to skin: a review for transdermal
drug delivery. Adv Drug Deliv Rev 1996; 18:395–425.
29. McElnay JC, Benson HAE, Hadgraft J, Murphy TM. (1993). The use of ultrasound in
skin penetration enhancement. In: Walters KA, Hadgraft J, eds. Pharmaceutical Skin
Penetration Enhancement. New York: Marcel Dekker. 1993; 293–309.
30. Steele GW. 2001. Preformulation. In: Gibson M, ed. Pharmaceutical Preformulation
and Formulation. Interpharm Press. Interpharm Book.
31. Kitson N, Maddin S. Drugs used for skin diseases. In: Roberts MS, Walters KA, eds.
Dermal Absorption and Toxicity Assessment. New York: Marcel Dekker. 1998; 313–
326.
32. Zatz JL, Varsano J, Shah VP. In vitro release of betamethasone dipropionate from
petrolatum-based ointments. Pharm Dev Technol 1996; 1:293–298.
33. Kneczke M, Landersjö L, Lundgren P, Führer C. In vitro release of salicylic acid from
two different qualities of white petrolatum. Acta Pharm Suec 1986; 23:193–204.
34. Osborne DW. Phase behavior characterization of propylene glycol, white petrolatum,
surfactant ointments. Drug Dev Ind Pharm 1992; 18:1883–1894.
35. Osborne DW. Phase behavior characterization of ointments containing lanolin or a
lanolin substitute. Drug Dev Ind Pharm 1993; 19:1283–1302.
36. Pena LE, Lee BL, Stearn JF. Structural rheology of a model ointment. Pharm Res
1994; 11:875–881.
37. Dahl T, He G–X, Samuels G. Effect of hydrogen peroxide on the viscosity of a hy-
droxyethylcellulose-based gel. 1998; Pharm Res 15:1137–1140.
38. Planas MD, Rodriguez FG, Dominguez MH. The influence of neutralizer concentration
on the rheological behaviour of a 0.1% Carbopol hydrogel. Pharmazie 1992; 47:351–
355.
39. Pérez–Marcos B, Martı́nez–Pacheco R, Gömez–Amoza, Souto C, Concheiro A, Rowe
RC. Interlot variability of carbomer 934. Int J Pharm 1993; 100:207–212.
40. Sjöblom J, ed. Emulsions and Emulsion Stability. New York: Marcel Dekker. 1996.
41. Eccleston GM. Multiple phase oil-in-water emulsions. J Soc Cosmet Chem 1990; 41:
1–22.
42. Eccleston GM. Functions of mixed emulsifiers and emulsifying waxes in dermatolog-
ical lotions and creams. Colloids Surfaces 1997; 123:169–182.
43. Orth DS. Handbook of Cosmetic Microbiology. New York: Marcel Dekker, 1993.
44. Takruri H, Anger CB. (1989). Preservation of dispersed systems. In: Lieberman HA,
Rieger MM, Banker GS, eds. Pharmaceutical Dosage Forms: Disperse Systems. Vol.
2. New York: Marcel Dekker. 1989; 73–114.
45. Cosmetic Ingredient Review. Final report on the safety of methyl, ethyl, propyl and
butyl paraben. JACT 1984; 3(5).
46. Dal Pozzo A, Pastori N. Percutaneous absorption of parabens from cosmetic formu-
lations. Int J Cosmet Sci 1996; 18:57–66.
47. Kurup TRR, Wan LSC, Chan LW. Interaction of preservatives with macromolecules.
Part II. Cellulose derivatives. Pharm Acta Helv 1995; 70:187–193.
48. Tanaka M, Iwata Y, Kouzuki Y, Taniguchi K, Matsuda H, Arima H, Tsuchiya S. Effect
of 2-hydroxypropyl-␤-cyclodextrin on percutaneous absorption of methyl paraben. J
Pharm Pharmacol 1995; 47:897–900.
49. Komatsu H, Higaki K, Okamoto H, Miyagawa K, Hashida M, Sezaki H. Preservative
activity and in vivo percutaneous penetration of butyl paraben entrapped in liposomes.
Chem Pharm Bull 1986; 34:3415–3422.
50. Komatsu H, Okamoto H, Miyagawa K, Hashida M, Sezaki H. Percutaneous absorption
of butyl paraben from liposomes in vitro. Chem Pharm Bull 1986; 34:3423–3430.
51. Kabara JJ, Orth DS, eds. Preservative-Free and Self-Preserving Cosmetics and Drugs
—Principles and Practice. New York: Marcel Dekker. 1997.
52. Shah VP, Elkins J, Hanus J, Noorizadeh C, Skelly JP. In vitro release of hydrocortisone
from topical preparations and automated procedure. Pharm Res 1991; 8:55–59.
53. Chattaraj SC, Kanfer I. Release of acyclovir from semisolid dosage forms: a semiau-
tomated procedure using a simple Plexiglass flow-through cell. Int J Pharm 1995; 125:
215–222.
54. Higuch T. Physical chemical analysis of percutaneous absorption process from creams
and ointments. J Soc Cosmet Chem 1960; 11:85–97.
55. Higuchi T. Rate of release of medicaments from ointment bases containing drugs in
suspension. J Pharm Sci 1961; 50:874–875.
56. Bunge AL. Release rates from topical formulations containing drugs in suspension. J
Control Release 1998; 52:141–148.
57. Flynn GL, Caetano PA, Pillai RS, Abriola LM, Amidon GE. Recent perspectives in
percutaneous penetration: in vitro drug release and its relationship to drug delivery. In:
Brain KR, Walters KA, eds. Perspectives in Percutaneous Penetration. Vol. 6b. Cardiff,
Wales: STS Publishing. 1999 (in press).
58. Kril MB, Parab PV, Genier SE, DiNunzio JE, Alessi D. Potential problems encountered
with SUPAC-SS and the in vitro release testing of ammonium lactate cream. Pharm
Technol 1999; March:164–174.
59. Guy RH, Hadgraft J. The prediction of plasma levels of drugs following transdermal
application. J Control Release 1985; 1:177–182.
60. Pugh WJ, Hadgraft J, Roberts MS. Physicochemical determinants of stratum corneum
permeation. In: Roberts MS, Walters KA, eds. Dermal Absorption and Toxicity As-
sessment. New York: Marcel Dekker. 1998; 245–268.
61. Kasting GB, Filloon TG, Francis WR, Meredith MP. Improving the sensitivity of in
vitro skin penetration experiments. Pharm Res 1994; 11:1747–1754.
62. Giannakou SA, Dallas PP, Rekkas DM, Choulis NH. Development and in vitro eval-
uation of nimodipine transdermal formulations using factorial design. Pharm Dev Tech-
nol 1998; 3:517–525.
63. Kandimalla KK, Kanikkannan N, Singh M. Optimization of a vehicle mixture for the
transdermal delivery of melatonin using artificial neural networks and response surface
method. J Control Release 1999; 61:71—82.
64. Takayama K, Takahara J, Fujikawa M, Ichikawa H, Nagai T. Formula optimization
based on artificial neural networks in transdermal drug delivery. J Control Release
1999; 62:161–170.
65. Marecki NM. Design considerations in transdermal drug delivery systems. Proceedings
10th Pharmaceutical Technology Conference. 1987; 311–318.
66. Yuk SH, Lee SJ, Okano T, Berner B, Kim SW. One-way membrane for transdermal
drug delivery systems. I. Membrane preparation and characterization. Int J Pharm 1991;
77:221–229.
67. Yuk SH, Lee SJ, Okano T, Berner B, Kim SW. One-way membrane for transdermal
drug delivery systems. II. System optimization. Int J Pharm 1991; 77:231–237.
68. Okabe H, Suzuki E, Saitoh T, Takayama K, Nagai T. Development of novel transdermal
system containing d-limonene and ethanol as absorption enhancers. J Control Release
1994; 32:243–247.
69. Wang D–M, Lin F–C, Chen L–Y, Lai J–Y. Application of asymmetric TPX membranes
to transdermal delivery of nitroglycerin. J Control Release 1998; 50:187–195.
70. Lai J–Y, Lin F–C, Wu T–T, Wang D–M. On the formation of macrovoids in PMMA
membranes. J Membr Sci 1999; 155:31–43.
71. Thacharodi D, Rao KP. Rate-controlling biopolymer membranes as transdermal deliv-
ery systems for nifedipine: development and in vitro evaluations. Biomaterials 1996;
17:1307–1311.
72. Iordanskii AL, Feldstein MM, Markin VS, Hadgraft J, Plate NA. Modeling of the drug
delivery from a hydrophilic transdermal therapeutic system across polymer membrane.
Eur J Pharm Biopharm 2000; 49:287–293.
73. Pfister WR, Sheeran MA, Watters DE, Sweet RP, Walters P. Methods for altering release
of progesterone, testosterone, propranolol, and indomethacin from silicone matrices:
effects of cosolvents and inert fillers. Proc Int Symp Control Release Bioact Mater
1987; 14:223–224.
74. Ulman KL, Gornowicz GA, Larson KR, Lee C–L. Drug permeability of modified
silicone polymers. I. Silicone-organic block copolymers. J Control Release 1989; 10:
251–260.
75. Hille T. Technological aspects of penetration enhancers in transdermal systems. In:

York: Marcel Dekker. 1993; 335–343.
76. Adhesives Research, Inc., Transdermal Technical Brief, Enhancer-Tolerant Adhesives.
77. Ko CU, Wilking SL, Birdsall J. Pressure sensitive adhesive property optimizations for
the transdermal drug delivery systems. Pharm Res 1995; 12:S-143.
78. te Hennepe HJC. Solution acrylic polymers for medical applications. National Starch
& Chemical B.V. 1993.
79. Jenkins AW. Developing the Fematrix transdermal patch. Pharm J 1995; 255:179–181.
80. PMA Committee Report. Transdermal Drug Delivery Systems. Pharmacop Forum
1986; 12:1798–1807.
81. Shah VP, Tymes NW, Yamamoto LA, Skelly JP. In vitro dissolution profile of trans-
dermal nitroglycerin patches using paddle method. Int J Pharm 1986; 32:243–250.
82. Man M, Chang C, Lee PH, Broman CT, Cleary GW. New improved paddle method
for determining the in vitro drug release profiles of transdermal delivery systems. J
Control Release 1993; 27:59–68.
83. Roy SD. Preformulation aspects of transdermal drug delivery systems, In: Ghosh TK,
Pfister WR, Sum SI, eds. Transdermal and Topical Drug Delivery Systems. Buffalo
Grove: Interpharm Press, 1997; 139–166.
84. Gibson M, ed. Pharmaceutical Preformulation and Formulation. Englewood, IHS
Health Group, 2001.
85. Caetano PA, Flynn GL, Farinha AR, Toscano CF, Campos RC. The in vitro release
test as a means to obtain the solubility and diffusivity of drugs in semisolids. Proc Int
Symp Control Release Bioact Mater 1999; 26:375–376.
86. Gourlay S. The pros and cons of transdermal nicotine therapy. Med J Aust 1994; 160:
152.
87. Lin SS, et al. Transdermal nicotine delivery systems—multiinstitutional cooperative
bioequivalence studies. Drug Dev Ind Pharm 1993; 19:2765–2793.
88. Davila DG, et al. Acute effects of transdermal nicotine on sleep architecture, snoring,
and sleep-disordered breathing in nonsmokers. Am J Respir Crit Care Med 1994; 150:
469–474.
89. Rose JE, et al. Mecamylamine combined with nicotine skin patch facilitates smoking
cessation beyond nicotine patch treatment alone. Clin Pharmacol Ther 1994; 56:86–
99.
90. Zevin S, Jacob P III, Benowitz NI. Nicotine–mecamylamine interactions. Clin Phar-
macol Ther 2000; 68:58–66.
91. Benowitz NL. Clinical-pharmacology of transdermal nicotine. Eur J Pharm Biopharm
1995; 41:168–174.
92. Fredrickson PA, et al. High-dose transdermal nicotine therapy for heavy smokers—
safety, tolerability and measurement of nicotine and cotinine levels. Psychopharma-
cology 1995; 122:215–222.
93. Jorenby DE, et al. Varying nicotine patch dose and type of smoking cessation coun-
seling. JAMA 1995; 274:1347–1352.
94. Kornitzer M, et al. Combined use of nicotine patch and gum in smoking cessation—
a placebo-controlled clinical-trial. Prev Med 1995; 24:41–47.
95. Campbell IA, Prescott RJ, TjederBurton SM. Transdermal nicotine plus support in
patients attending hospital with smoking-related diseases: a placebo-controlled study.
Respir Med 1996; 90:47–51.
96. Lindell G, Lunell E, Graffner H. Transdermally administered nicotine accumulates in
gastric juice. Eur J Clin Pharmacol 1996; 51:315–318.
97. Fiscella K, Franks P. Cost-effectiveness of the transdermal nicotine patch as an adjunct
to physicians’ smoking cessation counseling. JAMA 1996; 275:1247–1251.
98. Wasley MA, et al. The cost-effectiveness of the nicotine transdermal patch for smoking
cessation. Prev Med 1997; 26:264–270.
99. Jaen CR, et al. Patterns of use of a free nicotine patch program for Medicaid and
uninsured patients. J Natl Med Assoc 1997; 89:325–328.
100. Stapleton JA, Lowin A, Russell MAH. Prescription of transdermal nicotine patches for
smoking cessation in general practice: evaluation of cost-effectiveness. Lancet 1999;
354:210–215.
101. Benowitz NL, Gourlay SG. Cardiovascular toxicity of nicotine: implications for nic-
otine replacement therapy. J Am Coll Cardiol 1997; 29:1422–1431.
102. Benowitz NL, Zevin S, Jacob P. Sources of variability in nicotine and cotinine levels
with use of nicotine nasal spray, transdermal nicotine, and cigarette smoking. Br J Clin
Pharmacol 1997; 43:259–267.
103. Gourlay SG, et al. Determinants of plasma concentrations of nicotine and cotinine
during cigarette smoking and transdermal nicotine treatment. Eur J Clin Pharmacol
1997; 51:407–414.
104. Greenland S, Satterfield MH, Lanes SF. A meta-analysis to assess the incidence of
adverse effects associated with the transdermal nicotine patch. Drug Safety 1998; 18:
297–308.
105. Lawson GM, et al. Application of urine nicotine and cotinine excretion rates to as-
sessment of nicotine replacement in light, moderate, and heavy smokers undergoing
transdermal therapy. J Clin Pharmacol 1998; 38:510–516.
106. Lawson GM, et al. Application of serum nicotine and plasma cotinine concentrations
to assessment of nicotine replacement in light, moderate, and heavy smokers under-
going transdermal therapy. J Clin Pharmacol 1998; 38:502–509.
107. Lewis SF, et al. Transdermal nicotine replacement for hospitalized patients: a random-
ized clinical trial. Prev Med 1998; 27:296–303.
108. West R, Willis N. Double-blind placebo controlled trial of dextrose tablets and nicotine
patch in smoking cessation. Psychopharmacology 1998; 136:201–204.
109. Blondal T, et al. Nicotine nasal spray with nicotine patch for smoking cessation: ran-
domised trial with six year follow up. Br Med J 1999; 318:285–289.
110. Daughton DM, et al. The smoking cessation efficacy of varying doses of nicotine patch
delivery systems 4 to 5 years post-quit day. Prev Med 1999; 28:113–118.
111. Gariti P, et al. Cotinine replacement levels for a 21 mg/day transdermal nicotine patch
in an outpatient treatment setting. Drug Alcohol Depend 1999; 54:111–116.
112. Gourlay SG, et al. Predictors and timing of adverse experiences during transdermal
nicotine therapy. Drug Safety 1999; 20:545–555.
113. Hajek P, et al. Randomized comparative trial of nicotine polacrilex, a transdermal patch,
nasal spray, and an inhaler. Arch Intern Med 1999; 159:2033–2038.
114. Ji AJ, et al. A new gas chromatography–mass spectrometry method for simultaneous
determination of total and free trans-3⬘-hydroxycotinine and cotinine in the urine of
subjects receiving transdermal nicotine. Clin Chem 1999; 45:85–91.
115. Sonderskov J, et al. Nicotine patches in smoking cessation: a randomized trial among
over-the-counter customers in Denmark. Ugeskr Laeger 1999; 161:593–597.
116. Tonnesen P, et al. Higher dosage nicotine patches increase one-year smoking cessation
rates: results from the European CEASE trial. Eur Respir J 1999; 13:238–246.
117. Homsy W, et al. Plasma levels of nicotine and safety of smokers wearing transdermal
delivery systems during multiple simultaneous intake of nicotine and during exercise.
J Clin Pharmacol 1997; 37:728–736.
118. Wright LN, et al. Transdermal nicotine replacement in pregnancy: maternal pharma-
cokinetics and fetal effects. Am J Obstet Gynecol 1997; 176:1090–1094.
119. Dale LC, et al. Weight change after smoking cessation using variable doses of trans-
dermal nicotine replacement. J Gen Intern Med 1998; 13:9–15.
120. Ogburn PL, Jr, et al. Nicotine patch use in pregnant smokers: nicotine and cotinine
levels and fetal effects. Am J Obstet Gynecol 1999; 181:736–743.
121. West R, et al. A comparison of the abuse liability and dependence potential of nicotine
patch, gum, spray and inhaler. Psychopharmacology 2000; 149:198–202.
122. Pullan RD, et al. Transdermal nicotine for active ulcerative-colitis. N Engl J Med 1994;
330:811–815.
123. Thomas GAO, et al. Transdermal nicotine as maintenance therapy for ulcerative-colitis.
N Engl J Med 1995; 332:988–992.
124. Thomas GAO, et al. Transdermal nicotine compared with oral prednisolone therapy
for active ulcerative colitis. Eur J Gastroenterol Hepatol 1996; 8:769–776.
125. Sandborn WJ, et al. Transdermal nicotine for mildly to moderately active ulcerative
colitis—a randomized, double-blind, placebo-controlled trial. Ann Intern Med 1997;
126:364.
126. Guslandi M, Tittobello A. Pilot trial of nicotine patches as an alternative to corticoste-
roids in ulcerative colitis. J Gastroenterol 1996; 31:627–629.
127. Guslandi M, Tittobello A. Outcome of ulcerative colitis after treatment with transdermal
nicotine. Eur J Gastroenterol Hepatol 1998; 10:513–515.
128. Guslandi M. Long-term effects of a single course of nicotine treatment in acute ulcer-
ative colitis: remission maintenance in a 12-month follow-up study. Int J Colorectal
Dis 1999; 14:261–262.
129. SalinPascual RJ, et al. Antidepressant effect of transdermal nicotine patches in non-
smoking patients with major depression. J Clin Psychiatry 1996; 57:387–389.
130. Shytle RD, et al. Transdermal nicotine for Tourette’s syndrome. Drug Dev Res 1996;
38:290–298.
131. Silver AA, Shytle RD, Sanberg PR. Clinical experience with transdermal nicotine patch
in Tourette syndrome. CNS Spectr 1999; 4(2):68–76.
132. White HK, Levin ED. Four-week nicotine skin patch treatment effects on cognitive
performance in Alzheimer’s disease. Psychopharmacology 1999; 143:158–165.
133. Matsushima D, Prevo ME, Gorsline J. Absorption and adverse-effects following topical
and oral-administration of 3 transdermal nicotine products to dogs. J Pharm Sci 1995;
84:365–369.
134. Woolf A, et al. Self-poisoning among adults using multiple transdermal nicotine
patches. J Toxicol Clin Toxicol 1996; 34:691–698.
135. Bogaert MG. Clinical pharmacokinetics of nitrates. Cardiovasc Drugs Ther 1994; 8:
693–699.
136. Hutt V, et al. Bioequivalence evaluation of the metabolites 1,2-glyceryl and 1,3-glyceryl
dinitrate of 2 different glyceryl trinitrate patches after 12-h usage in healthy-volunteers.
ArzneimitteForschung 1994; 44:1317–1321.
137. Devita C, et al. Gissi-3—effects of lisinopril and transdermal glyceryl trinitrate singly
and together on 6-week mortality and ventricular function after acute myocardial in-
farction. Lancet 1994; 343:1115–1122.
138. Andrews R, et al. Inhibition of platelet-aggregation by transdermal glyceryl trinitrate.
Br Heart J 1994; 72:575–579.
139. Berrazueta JR, et al. Local transdermal glyceryl trinitrate has an antiinflammatory ac-
tion on thrombophlebitis induced by sclerosis of leg varicose veins. Angiology 1994;
45:347–351.
140. Parker JO, et al. Intermittent transdermal nitroglycerin therapy in angina-pectoris—
clinically effective without tolerance or rebound. Circulation 1995; 91:1368–1374.
141. Serone AP, Angus JA, Wright CE. Baroreffex resetting but no vascular tolerance in
response to transdermal glyceryl trinitrate in conscious rabbits. Br J Pharmacol 1996;
118:93–104.
142. Ardissino D, et al. Effect of transdermal nitroglycerin or N-acetylcysteine, or both, in

the long-term treatment of unstable angina pectoris. J Am Coll Cardiol 1997; 29:941–
947.
143. Klemsdal TO, Mundal HH, Gjesdal K. Mechanisms of tolerance during treatment with
nitroglycerin patches for 24 h. Eur J Clin Pharmacol 1996; 51:227–230.
144. Klemsdal TO, et al. Transdermal nitroglycerin: clinical and pharmacokinetic conse-
quences of renewing the patch and the application site. Eur J Clin Pharmacol 1997;
52:379–381.
145. Elkayam U, et al. Double-blind, placebo-controlled study to evaluate the effect of
organic nitrates in patients with chronic heart failure treated with angiotensin-convert-
ing enzyme inhibition. Circulation 1999; 99:2652–2657.
146. Teh LS, et al. Sustained-release transdermal glyceryl trinitrate patches as a treatment
for primary and secondary raynauds-phenomenon. Br J Rheumatol 1995; 34:636–641.
147. Berrazueta JR, et al. Successful treatment of shoulder pain syndrome due to supra-
spinatus tendinitis with transdermal nitroglycerin. A double blind study. Pain 1996; 66:
63–67.
148. Zoli M, et al. Transdermal nitroglycerin in cirrhosis. A 24-hour echo-Doppler study of
splanchnic hemodynamics. J Hepatol 1996; 25:498–503.
149. Moya RA, et al. Transdermal glyceryl trinitrate in the management of primary dys-
menorrhea. Int J Gynecol Obstet 2000; 69:113–118.
150. Black RS, et al. Maternal and fetal cardiovascular effects of transdermal glyceryl tri-
nitrate and intravenous ritodrine. Obstet Gynecol 1999; 94:572–576.
151. Lees CC, et al. Glyceryl trinitrate and ritodrine in tocolysis: an international multicenter
randomized study. Obstet Gynecol 1999; 94:403–408.
152. Kovacs P, et al. Effect of transdermal nitroglycerin on glucose-stimulated insulin release
in healthy male volunteers. Eur J Clin Invest 2000; 30:41–44.
153. Minghetti P, et al. In vitro skin permeation of sinitrodil, a member of a new class of
nitrovasodilator drugs. Eur J Pharm Sci 1999; 7:231–236.
154. Whittington R, Faulds D. Hormone replacement therapy. 1. Pharmacoeconomic ap-
praisal of its therapeutic use in menopausal symptoms and urogenital estrogen defi-
ciency. Pharmacoeconomics 1994; 5:419–445.
155. Wiseman LR, McTavish D. Transdermal estradiol norethisterone—a review of its phar-
macological properties and clinical use in postmenopausal women. Drugs Aging 1994;
4:238–256.
156. Creatsas G, et al. Transdermal estradiol plus oral medroxyprogesterone acetate replace-
ment therapy in primary amenorrheic adolescents—clinical, hormonal and metabolic
aspects. Maturitas 1994; 18:105–114.
157. Normantaylor JQ, et al. Hormone replacement therapy by the transdermal administra-
tion of estradiol and norethisterone. Maturitas 1994; 18:221–228.
158. Smith RNJ, et al. A randomized comparison over 8 months of 100 mu-g and 200 mu-
g twice weekly doses of transdermal estradiol in the treatment of severe premenstrual-
syndrome. Br J Obstet Gynaecol 1995; 102:475–484.
159. Pornel B. Efficacy and safety of Menorest in two positive-controlled studies. Eur J
Obstet Gynecol Reprod Biol 1996; 64(SS):S35–S37.
160. Raudaskoski TH, et al. Transdermal estrogen with a levonorgestrel-releasing intrauter-
ine-device for climacteric complaints—clinical and endometrial responses. Am J Obstet
Gynecol 1995; 172(1 pt1):114–119.
161. Haines CJ, et al. The effect of percutaneous oestrogen replacement therapy on Lp(a)
and other lipoproteins. Maturitas 1995; 22:219–225.
162. Crosignani PG, Cortellaro M, Boschetti C. Effects on haemostasis of hormone replace-
ment therapy with transdermal estradiol and oral sequential medroxyprogesterone ac-
etate: a 1-year, double-blind, placebo-controlled study. Thrombo Haemost 1996; 75:

476–480.
163. Macdonald AG, et al. Effects of hormone replacement therapy in rheumatoid-arthritis
—a double-blind placebo-controlled study. Ann Rheum Dis 1994; 53:54–57.
164. Evans SF, Davie MWJ. Low and conventional dose transdermal oestradiol are equally
effective at preventing bone loss in spine and femur at all post-menopausal ages. Clin
Endocrinol 1996; 44:79–84.
165. Roux C. Estrogen therapy in postmenopausal osteoporosis what we know and what we
don’t. Rev Rhum 1997; 64:402–409.
166. Hannon R, et al. Response of biochemical markers of bone turnover to hormone re-
placement therapy: impact of biological variability. J Bone Miner Res 1998; 13:1124–
1133.
167. Prelevic GM, et al. Comparative effects on bone mineral density of tibolone, transder-
mal estrogen and oral estrogen/progestogen therapy in postmenopausal women. Gy-
necol Endocrinol 1996; 10:413–420.
168. Boonkasemsanti W, et al. The effect of transdermal oestradiol on bleeding pattern,
hormonal profiles and sex steroid receptor distribution in the endometrium of Norplant
users. Hum Reprod 1996; 11(S2):115–123.
169. Habiba M, et al. Thrombophilia and lipid profile in post-menopausal women using a
new transdermal oestradiol patch. Eur J Obstet Gynecol Reprod Biol 1996; 66:165–
168.
170. Perrone G, et al. Effect of oral and transdermal hormone replacement therapy on lipid
profile and Lp(a) level in menopausal women with hypercholesterolemia. Int J Fertil
Menopausal Stud 1996; 41:509–515.
171. Rabe T, et al. Spacing-out of progestin—efficacy, tolerability and compliance of two
regimens for hormonal replacement in the late postmenopause. Gynecol Endocrinol
1997; 11:383–392.
172. Cameron ST, et al. Continuous transdermal oestrogen and interrupted progestogen as
a novel bleed-free regimen of hormone replacement therapy for postmenopausal
women. Br J Obstet Gynaecol 1997; 104:1184–1190.
173. Rozenbaum H, et al. Comparison of two estradiol transdermal systems (Oesclim 50
and Estraderm TTS 50). 2. Local skin tolerability. Maturitas 1996; 25:175–185.
174. Rozenbaum H, et al. Comparison of two estradiol transdermal systems (Oesclim 50
and Estraderm TTS 50). 1. Tolerability, adhesion and efficacy. Maturitas 1996; 25:161–
173.
175. AlAzzawi F, et al. A randomised study to compare the efficacy and safety of a new 17
beta-oestradiol transdermal matrix patch with Estraderm TTS 50 in hysterectomised
postmenopausal women. Br J Clin Prac 1997; 51:20.
176. Rozenberg S, Ylikorkala O, Arrenbrecht S. Comparison of continuous and sequential
transdermal progestogen with sequential oral progestogen in postmenopausal women
using continuous transdermal estrogen: vasomotor symptoms, bleeding patterns, and
serum lipids. Int J Fertil Womens Med 1997; 42(S2):376–387.
177. Appelberg J et al. Safety and efficacy of transdermal estradiol replacement therapy in
postmenopausal liver transplanted women—a preliminary report. Acta Obstet Gynecol
Scand 1998; 77:660–664.
178. Bhathena RK, et al. The influence of transdermal oestradiol replacement therapy and
medroxyprogesterone acetate on serum lipids and lipoproteins. Br J Clin Pharmacol
1998; 45:170–172.
179. Blanc B, et al. Continuous hormone replacement therapy for menopause combining
nomegestrol acetate and gel, patch, or oral estrogen: a comparison of amenorrhea rates.
Clin Ther 1998; 20:901–912.
180. Notelovitz M. Clinical experience with a 7-day estrogen patch: principles and practice.
Gynecol Endocrinol 1998; 12:249–258.
181. Hirvonen E, et al. Effects of transdermal oestrogen therapy in postmenopausal women:
a comparative study of an oestradiol gel and an oestradiol delivering patch. Br J Obstet
Gynaecol 1997; 104(S16):26–31.
182. Hirvonen E, et al. Transdermal oestradiol gel in the treatment of the climacterium: a
comparison with oral therapy. Br J Obstet Gynaecol 1997; 104(S16):19–25.
183. Jarvinen A, Nykanen S, Paasiniemi L. Absorption and bioavailability of oestradiol from
a gel, a patch and a tablet. Maturitas 1999; 32:103–113.
184. Vavilis D, et al. The effect of transdermal estradiol on the conjunctiva in postmeno-
pausal women. Eur J Obstet Gynecol Reprod Biol 1997; 72:93–96.
185. Coustou D, et al. Estrogen dermatitis. Ann Dermatol Venereol 1998; 125:505–508.
186. Carr BR. HRT management: the American experience. Eur J Obstet Gynecol Reprod
Biol 1996; 64(SS):S17–S20.
187. Andersson TLG, et al. Bioavailability of estradiol from two matrix transdermal delivery
systems: menorest and Climara. Maturitas 2000; 34:57–64.
188. Archer DF, et al. A randomized comparison of continuous combined transdermal de-
livery of estradiol–norethindrone acetate and estradiol alone for menopause. Obstet
Gynecol 1999; 94:498–503.
189. Baracat E, et al. Comparative bioavailability study of two 100-␮g daily 17-beta-estra-
diol transdermal delivery systems: once-a-week matrix patch and twice-a-week reser-
voir patch in healthy postmenopausal women. Curr Ther Res Clin Exp 1999; 60:129–
137.
190. Beljic T, et al. Effect of estrogen replacement therapy on cardiac function in postmen-
opausal women with and without flushes. Gynecol Endocrinol 1999; 13:104–112.
191. Bhathena RK, Anklesaria BS, Ganatra AM. The treatment of hypertriglyceridaemia in
menopausal women with transdermal oestradiol therapy. Br J Obstet Gynaecol 1999;
106:980–982.
192. Brackman F. Oesclim: summary of tolerability and safety. Maturitas 1999; 33(suppl
1):83–88.
193. Utian WH, et al. Efficacy and safety of low, standard, and high dosages of an estradiol
transdermal system (Esclim) compared with placebo on vasomotor symptoms in highly
symptomatic menopausal patients. Am J Obstet Gynecol 1999; 181:71–79.
194. Buch AB, et al. Significant differences in estradiol bioavailability from two similarly
labelled estradiol matrix transdermal systems. Climacteric 1999; 2:248–253.
195. Caserta R, et al. Efficacy and safety of a weak transdermal estradiol drug delivery
system: a pharmacokinetic comparison of two different dosage of formulations. G Ital
Ostet Ginecol 2000; 22:279–283.
196. Cohen L, et al. Low-dose 17-beta estradiol matrix transdermal system in the treatment
of moderate-to-severe hot flushes in postmenopausal women. Curr Ther Res Clin Exp
1999; 60:534–547.
197. Cooper C, et al. Matrix delivery transdermal 17beta-estradiol for the prevention of
bone loss in postmenopausal women. Osteopor Int 1999; 9:358–366.
198. Weiss SR, et al. A randomized controlled trial of four doses of transdermal estradiol
for preventing postmenopausal bone loss. Obstet Gynecol 1999; 94:330–336.
199. De Aloysio D, et al. Efficacy on climacteric symptoms and safety of low dose estradiol
transdermal matrix patches—a randomized, double-blind placebo-controlled study.
Arzneimittelforschung 2000; 50:293–300.
200. De Vrijer B, et al. Efficacy and tolerability of a new estradiol delivering matrix patch
(Estraderm MX) in postmenopausal women. Maturitas 2000; 34:47–55.
201. Geisler J, et al. Plasma oestrogen fractions in postmenopausal women receiving hor-
mone replacement therapy: influence of route of administration and cigarette smoking.
J Endocrinol 1999; 162:265–270.
202. Geyer D, et al. Pharmacokinetics of Fem7, a once-weekly, transdermal oestrogen re-

placement system in healthy, postmenopausal women. Gynecol Obstet Invest 1999; 48:
1–6.
203. Good WR, et al. Comparison of Alora estradiol matrix transdermal delivery system
with oral conjugated equine estrogen therapy in relieving menopausal symptoms. Cli-
macteric 1999; 2:29–36.
204. Matsumoto S, et al. A study of the clinical effect of estradiol transdermal therapeutic
system alone on pollakisuria and urinary incontinence in postmenopausal woman. Jpn
J Urol 2000; 91:501–505.
205. Mattsson L–A. Clinical experience with continuous combined transdermal hormone
replacement therapy. J Menopause 1999; 6:25–29.
206. Nachtigall LE, et al. Serum estradiol-binding profiles in postmenopausal women un-
dergoing three common estrogen replacement therapies: associations with sex hormone-
binding globulin, estradiol, and estrone levels. Menopause 2000; 7:243–250.
207. Rohr UD, Nauert C, Stehle B. 17beta-Estradiol delivered by three different matrix
patches 50 ␮g/day. A three way cross-over study in 21 postmenopausal women. Ma-
turitas 1999; 33:45–58.
208. Rovati LC, et al. Efficacy on climacteric symptoms of a new estradiol transdermal
patch with active matrix in comparison with a reference reservoir patch: two long-term
randomized multicenter parallel-group studies. Arzneimittelforschung 1999; 49:933–
943.
209. Wojcicki J, et al. Effects of estradiol and norethisterone replacement therapy on the
pharmacokinetics of phenazone. Med Sci Mon 1999; 5:754–757.
210. Burry KA, et al. Percutaneous absorption of progesterone in postmenopausal women
treated with transdermal estrogen. Am J Obstet Gynecol 1999; 180:1504–1511.
211. Carey BJ, et al. A study to evaluate serum and urinary hormone levels following short
and long term administration of two regimens of progesterone cream in postmenopausal
women. Br J Obstet Gynaecol 2000; 107:722–726.
212. Leonetti HB, Longo S, Anasti JN. Transdermal progesterone cream for vasomotor
symptoms and postmenopausal bone loss. Obstet Gynecol 1999; 225–228.
213. Chien TY, et al. Transdermal contraceptive delivery system—preclinical development
and clinical-assessment. Drug Dev Ind Pharm 1994; 20:633–664.
214. Sitrukware R. Transdermal application of steroid-hormones for contraception. J Steroid
Biochem Mol Biol 1995; 53:247–251.
215. Asthana S, et al. Cognitive and neuroendocrine response to transdermal estrogen in
postmenopausal women with Alzheimer’s disease: results of a placebo-controlled, dou-
ble-blind, pilot study. Psychoneuroendocrinology 1999; 24:657–677.
216. Gerber GS, et al. Transdermal estrogen in the treatment of hot flushes in men with
prostate cancer. Urology 2000; 55:97–101.
217. Cofrancesco J, Dobs AS. Transdermal testosterone delivery systems. Endocrinologist
1996; 6:207–213.
218. Behre HM, et al. Long-term substitution therapy of hypogonadal men with transscrotal
testosterone over 7–10 years. Clin Endocrinol 1999; 50:629–635.
219. Kenny AM, Prestwood KM, Raisz LG. Short-term effects of intramuscular and trans-
dermal testosterone on bone turnover, prostate symptoms, cholesterol, and hematocrit
in men over age 70 with low testosterone levels. Endocr Res 2000; 26:153–168.
220. Meikle AW, et al. Pharmaockinetics and metabolism of a permeation-enhanced testos-
terone transdermal system in hypogonadal men: influence of application site—a clinical
research center study. J Clin Endocrinol Metab 1996; 81:1832–1840.
221. Arver S, et al. Long-term efficacy and safety of a permeation-enhanced testosterone
transdermal system in hypogonadal men. Clin Endocrinol 1997; 47:727–737.
222. Bhasin S, et al. Effects of testosterone replacement with a nongenital, transdermal

system, Androderm, in human immunodeficiency virus-infected men with low testos-
terone levels. J Clin Endocrinol Metab 1998; 83:3155–3162.
223. Miller K, et al. Transdermal testosterone administration in women with acquired im-
munodeficiency syndrome wasting: a pilot study. J Clin Endocrinol Metab 1998; 83:
2717–2725.
224. Jain P, Rademaker AW, McVary KT. Testosterone supplementation for erectile dys-
function: results of a meta-analysis. J Urol 2000; 164:371–375.
225. Davis SR. The therapeutic use of androgens in women. J Steroid Biochem Mol Biol
1999; 69:177–184.
226. Mazer NA. New clinical applications of transdermal testosterone delivery in men and
women. J Control Release 2000; 65:303–315.
227. Parker S, Armitage M. Experience with transdermal testosterone replacement therapy
for hypogonadal men. Clin Endocrinol 1999; 50:57–62.
228. Rolf C, et al. Pharmacokinetics of new testosterone transdermal therapeutic systems in
gonadotropin-releasing hormone antagonist-suppressed normal men. Exp Clin Endo-
crinol Diabetes 1999; 107:63–69.
229. Tan KCB, Shiu SWM, Kung AWC. Alterations in hepatic lipase and lipoprotein sub-
fractions with transdermal testosterone replacement therapy. Clin Endocrinol 1999; 51:
765–769.
230. Wilson DE, et al. Use of topical corticosteroid pretreatment to reduce the incidence
and severity of skin reactions associated with testosterone transdermal therapy. Clin
Ther 1998; 20:299–306.
231. Simmonds MA. Transdermal fentanyl—clinical development in the United States. An-
ticancer Drugs 1995; 6(S3):35–38.
232. Sandler AN. Transdermal fentanyl for pain relief—guidelines for appropriate use. CNS
Drugs 1997; 7:442–451.
233. Korte W, Morant R. Transdermal fentanyl in uncontrolled cancer pain—titration on a
day-to-day basis as a procedure for safe and effective dose-finding—a pilot-study in
20 patients. Supp Care Cancer 1994; 2:123–127.
234. Vanbastelaere M, Rolly G, Abdullah NM. Postoperative analgesia and plasma levels
after transdermal fentanyl for orthopedic surgery—double-blind comparison with pla-
cebo. J Clin Anesth 1995; 7:26–30.
235. Fiset P, et al. Biopharmaceutics of a new transdermal fentanyl device. Anesthesiology
1995; 83:459–469.
236. Christensen ML, et al. Transdermal fentanyl administration in children and adolescents
with sickle cell pain crisis. J Pediatr Hematol Oncol 1996; 18:372–376.
237. Jeal W, Benfield P. Transdermal fentanyl—a review of its pharmacological properties
and therapeutic efficacy in pain control. Drugs 1997; 53:109–138.
238. Hardwick WE, King WD, Palmisano PA. Respiratory depression in a child uninten-
tionally exposed to transdermal fentanyl patch. South Med J 1997; 90:962–964.
239. Sloan PA, Moulin DE, Hays H. A clinical evaluation of transdermal therapeutic system
fentanyl for the treatment of cancer pain. J Pain Sympt Manage 1998; 16:102–111.
240. Payne R, et al. Quality of life and cancer pain: satisfaction and side effects with
transdermal fentanyl versus oral morphone. J Clin Oncol 1998; 16:1588–1593.
241. Dellemijn PLI, vanDuijn H, Vanneste JAL. Prolonged treatment with transdermal fen-
tanyl in neuropathic pain. J Pain Sympt Manage 1998; 16:220–229.
242. Kyles AE, Papich M, Hardie EM. Disposition of transdermally administered fentanyl
in dogs. Am J Vet Res 1996; 57:715–719.
243. Egger CM, et al. Comparison of plasma fentanyl concentrations by using three trans-
dermal fentanyl patch sizes in dogs. Vet Surg 1998; 27:159–166.
244. Collins JJ, et al. Transdermal fentanyl in children with cancer pain: feasibility, tolera-
bility, and pharmacokinetic correlates. J Pediatr 1999; 134:319–323.
245. Jakobsson M, Strang P. Fentanyl patches for the treatment of pain in dying cancer
patients. Anticancer Res 1999; 19:4441–4442.
246. Radbruch L et al. Constipation and the use of laxatives: a comparison between trans-
dermal fentanyl and oral morphine. Palliat Med 2000; 14:111–119.
247. Haazen L, et al. The constipation-inducing potential of morphine and transdermal fen-
tanyl. Eur J Pain 1999; 3(suppl A):9–15.
248. Flynn TN, Guest JF. Cost of managing constipation among terminally ill cancer patients
treated with transdermal fentanyl and 12 hour sustained-release morphine in Canada.
J Med Econ 1999; 2:15–32.
249. Elsner F, et al. Switching opioids to transdermal fentanyl in a clinical setting. Schmerz
1999; 13:273–278.
250. Ritchie LD. A clinical evaluation of flurbiprofen LAT and piroxicam gel: a multicentre
study in general practice. Clin Rheumatol 1996; 15:243–247.
251. Fujimura A, et al. Pharmacokinetics and pharmacodynamics of a new transdermal clon-
idine, M-5041t, in healthy-subjects. J Clin Pharmacol 1993; 33:1192–1200.
252. Fujimura A, et al. Comparison of the pharmacokinetics, pharmacodynamics, and safety
of oral (Catapres) and transdermal (M-5041t) Clonidine in healthy subjects. J Clin
Pharmacol 1994; 34:260–265.
253. Toon S, Phase-I pharmacokinetic assessment of the clonidine transdermal therapeutic
system. Eur J Pharm Biopharm 1995; 41:184–188.
254. Ellis JE, et al. Premedication with oral and transdermal clonidine provides safe and
efficacious postoperative sympatholysis. Anesth Analg 1994; 79:1133–1140.
255. Dorman T, et al. Effects of clonidine on prolonged postoperative sympathetic response.
Crit Care Med 1997; 25:1147–1152.
256. Owen MD, et al. Postoperative analgesia using a low-dose, oral–transdermal clonidine
combination: lack of clinical efficacy. J Clin Anesth 1997; 9:8–14.
257. Dandrea G, et al. Efficacy of transdermal clonidine in short-term treatment of cluster
headache—a pilot-study. Cephalalgia 1995; 15:430–433.
258. Byassmith MG, et al. Transdermal clonidine compared to placebo in painful diabetic
neuropathy using a 2-stage enriched enrollment design. Pain 1995; 60:267–274.
259. Dias VC, et al. Clinical experience with transdermal clonidine in African-American
and Hispanic-American patients with hypertension: evaluation from a 12-week pro-
spective, open-label clinical trial in community-based clinics. Am J Ther 1999; 6:19–
24.
260. Dinslage S, et al. A new transdermal delivery system for pilocarpine in glaucoma
treatment. Ger J Ophthalmol 1996; 5:275–280.
261. Bodick NC, et al. The selective muscarinic agonist xanomeline improves both the
cognitive deficits and behavioral symptoms of Alzheimer disease. Alzheimer Dis Assoc
Disord 1997; 11(S4):S16–S22.
262. Douma WR, et al. Effects of transdermal scopolamine on pulmonary function, symp-
toms and bronchial hyperresponsiveness to methacholine. Eur J Pharm Sci 1997; 5:
327–334.
263. Honkavaara P, Pyykko I. Effects of atropine and scopolamine on bradycardia and
emetic symptoms in otoplasty. Laryngoscope 1999; 109:108–112.
264. Gowans J, et al. The effects of scopolamine and cyclizine on visual–vestibular inter-
action in humans. J Vestib Res Equilibr Orient 2000; 10:87–92.
265. Sacktor N, et al. Transdermal selegiline in HIV-associated cognitive impairment: pilot,
placebo-controlled study. Neurology 2000; 54:233–235.
266. Rama RP, Diwan PV. Comparative in-vivo evaluation of diltiazem hydrochloride fol-
lowing oral and transdermal administration in rabbits. Indian J Pharmacol 1999; 31:
294–298.
267. Gillson G, et al. Transdermal histamine in multiple sclerosis: part one—clinical ex-
perience. Altern Med Rev 1999; 4:424–428.
268. Martin GP, et al. The influence of hydrocolloid patch composition on the bioavailability
of triamcinolone acetonide in humans. Drug Dev Ind Pharm, 2000; 26:35–43.
269. Sutinen R, et al. Water-activated, pH-controlled patch in transdermal administration of
timolol II. Drug absorption and skin irritation. Eur J Pharm Sci 2000; 11:25–31.
270. Sutinen R, Paronen P, Urtti A. Water-activated, pH-controlled patch in transdermal
administration of timolol I. Preclinical tests. Eur J Pharm Sci 2000; 11:19–24.
271. McVary KT, et al. Topical prostaglandin E1 SEPA gel for the treatment of erectile
dysfunction. J Urol 1999; 162:726–731.
272. Comer AM, Lamb HM. Lidocaine patch 5%. Drugs 59:245–249.
273. Roberge RJ, Krenzelok EP, Mrvos R. Transdermal drug delivery system exposure out-
comes. J Emerg Med 18:147–151.
8
Bioavailability and Bioequivalence of
Dermatological Formulations
CHRISTIAN SURBER
Kantonsspital Basel, Basel, Switzerland
ADRIAN F. DAVIS
GlaxoSmithKline Consumer Healthcare, Weybridge, Kent, England
I. INTRODUCTION
Official guidelines and current textbooks define bioavailability as the rate and extent
to which the active ingredient or therapeutic moiety is absorbed from the drug prod-
uct and becomes available at the site of action. From this, bioequivalence is defined
as the absence of a significant difference in bioavailability between pharmaceutical
equivalents or pharmaceutical alternatives, the latter being dosage forms in which
the chemical form, dosage form type, or strength of the therapeutic moiety, differ.
The absolute bioavailability is defined as the extent (here, the rate is usually
considered unimportant) to which the active ingredient or therapeutic moiety be-
comes available in the organism from a dosage form in comparison with an intra-
venously administered standard form, which is taken to be 100% bioavailable. The
relative bioavailability is defined as the rate and extent to which the active ingredient
or therapeutic moiety becomes available in the organism from a dosage form, com-
pared with a standard form administered by the same route.
After the early description of the physiological availability of vitamins in 1945
(1), erratic reports in the early 1960s (2–5) provided the first indications of the
potential for bioavailability and bioequivalence problems with multisource drug
products. Variations in absorption of active ingredients from different formulations
were soon recognized as a potential health hazard when episodes of drug toxicity
were reported following changes to the excipients in the pharmaceutical delivery
401
402 Surber and Davis
system or changes in manufacturing procedures (6). Major advances in analytical

technology in the late 1960s and early 1970s led to further in-depth investigations
involving the bioavailability of marketed drug products, and reports of bioinequiv-
alence included a series of drugs such as digoxin (7), phenytoin (8), chloramphenicol
(9), and many others. Although early legislation covering quality, safety, and efficacy
did not recognize or directly address the issues of bioavailability and bioequivalence,
in the last two decades, the bioavailability and bioequivalance of drug products have
emerged as important national and international regulatory and scientific issues.
With the growth of the worldwide generic pharmaceutical industry, bioequiv-
alence has added another dimension to the issue of quality of drug products. To be
interchangeable with the innovator drug product, a generic drug product must be not
only pharmaceutically equivalent or alternative, but also bioequivalent. Thus, bio-
equivalence plays an important role in assuring the therapeutic quality of the mul-
tisource drug formulations that compose most drug products.
The literature contains many examples of how the composition and manufac-
ture of the finished dosage form can alter the effectiveness of the drug. This is
particularly true with topical therapy, for which vehicles have profound effects on
percutaneous absorption and may also cause vehicle-related local effects. Product
efficacy and safety can depend on how much of the drug is ultimately absorbed from
its formulation and how rapidly this occurs. Thus, the two considerations—the extent
of absorption of the drug from its formulation, and the rate at which it is absorbed
—form the basis of bioavailability and bioequivalence testing and are the predictors
of therapeutic performance and therapeutic equivalence.
This chapter, which is divided into four main sections, will first discuss current
concepts of bioavailability and bioequivalence of dermal and, to a minor extent,
transdermal dosage forms. Against this background, the following two sections will
discuss related information on model systems to measure bioavailability and on fac-
tors influencing the rate and extent of absorption from topical products. In a final
section, bioequivalence, therapeutic equivalence, and control of therapy is discussed.
II. BIOAVAILABILITY AND BIOEQUIVALENCE OF TOPICAL

DOSAGE FORMS
Application of the term bioavailability to topical dosage forms requires, first, a care-
ful definition of the term topical dosage form and, second, a specific adaptation of
the general definition of bioavailability to the special case of topical dosage form.
Topical (Greek: topos = local) dosage forms include ophthalmic, nasal, and
otic preparations; mouth, throat, and pulmonary preparations; gastrointestinal, and
anorectal preparations; urogenital preparations; and dermatological preparations.
These preparations act at target sites that are identical or close to the site of appli-
cation and can be external (e.g., skin and eyes) or internal (e.g., lungs and stomach).
Instillation into a joint or an intradermal or subcutaneous injection may also be
considered—following the Greek root—topical or local applications. The motive for
topical or local delivery is the direct application of drug to the target site to maximize
efficacy, while minimizing systemic absorption, to improve safety. This chapter con-
siders only delivery systems that are placed against the skin.
Delivery systems that are placed against the skin deliver drugs to (a) the local
tissue immediately beneath the application site, (b) deeper regions in the vicinity of,
Bioavailability and Bioequivalence 403
but still somewhat remote from, the application site, and (c) the systemic circulation.
The objective of the latter is to mediate pharmacological changes at a site totally
removed from the application site. The following comprehensive definitions of the
different modes of delivery to the skin have been proposed by Flynn and Weiner
(10).
A. Modes of Delivery to the Skin
1. Topical Delivery
Topical delivery can be defined as the application of a drug-containing formulation
to the skin to directly treat cutaneous disorders (e.g., acne) or the cutaneous mani-
festations of a general disease (e.g., psoriasis), with the intent of confining the phar-
macological or other effect of the drug to the surface of the skin or within the skin.
Although systemic absorption may be unavoidable, it is always unwelcome. Semi-
solid formulations, in all their diversity, dominate the systems for topical delivery,
but foams, sprays, medicated powders, solution, and even medicated adhesive sys-
tems are in use.
2. Regional Delivery
Regional delivery, in contrast, involves the application of a drug to the skin for the
purpose of treating diseases or alleviating disease symptoms in deep tissues beneath
the application. Here, the intent is to effect or accent pharmacological actions of the
drug within musculature, vasculature, joints, and other, beneath and around the site
of application. A selectivity of action over that achieved by systemic administration
is sought. Regional activity requires percutaneous absorption and deposition, one is
depending on backleakage of drug from the venous drainage of the application site.
At best, backdiffusion would be an inefficient process; consequently, substantial sys-
temic uptake, although unwelcome, is unavoidable. Nevertheless, regional concen-
trations are thought to be higher than can be achieved by systemic administration at
the same total body exposure to the drug. The focusing of drugs into tissues in this
manner has been difficult to prove unequivocally, and thus considerable scepticism
exists concerning the validity of regional therapy. Regional delivery is accomplished
with traditional ointments and creams as well as large adhesive patches, plasters,
poultices and cataplasms.
3. Transdermal Delivery
Transdermal delivery involves the application of a drug to the skin to treat systemic
disease and is aimed at achieving systemically active levels of the drug. Although
such traditional dosage forms as ointments can be employed in this kind of therapy
(e.g., nitroglycerin ointments), adhesive systems of precisely defined size are the
rule. Here, percutaneous absorption with appreciable systemic drug accumulation is
absolutely essential. Ideally, there would be no local accumulation of drug, but such
accumulation is unavoidable. The drug is forced through the relatively small diffu-
sional window defined by the contact area of the patch. Consequently, high and
potentially irritating or sensitizing concentrations of a drug in the viable tissues
underlying the patch are preordained by the nature of the delivery process.
B. Sampled Matrix and Site of Action

For agents that are incorporated into transdermal delivery systems to treat disease of
systemic origin, the straightforward concept and measurement of bioavailability, as
used for oral drug administration, can be applied. As defined previously, bioavail-
ability is expected to be correlated with the concentration–time profile of drug in
the ‘‘biophase,’’ which includes the site of drug action. When one considers, for
example, bioavailability from an oral dosage form, it is generally accepted that drug
concentration in the blood (e.g., areas under the blood or plasma concentration–time
[AUC] profile) will adequately reflect the time course of drug presence within the
biophase; therefore, this can be used to evaluate bioavailability. Thus, two drug
products, including transdermal delivery systems, are considered to be bioequivalent
if they yield comparable bioavailability-based plasma concentration–time profiles
when administered to the same individuals under similar dosage conditions.
As an aside, the assumption that is then made, is that equivalent pharmacoki-
netics indicate equivalent therapeutics. Or, more precisely, that observed differences
found in pharmacokinetic parameters between formulations under controlled study
conditions and in a group of selected healthy volunteers are predictive of negligible
differences in performance in the clinical setting. There has been relatively little
discussion on this key assumption (11–13), which is addressed in a later section of
this chapter considering issues particular to topical and regional delivery.
Application of the term bioavailability to topical and regional delivery prepa-
rations where the drug is intended to act locally (e.g., skin or muscle) is more
complex. Determination of the relative bioavailability raises the first question: Is it
relative to the same drug administered by a different route, usually the oral route, or
is it relative to the same drug delivered from a standard topical or regional delivery
preparation? Both comparisons may be important depending on the route of admin-
istration by which efficacy and safety of the drug has been established.
The second question is: From which matrix should the drug be sampled: stra-
tum corneum, epidermis, dermis, subcutaneous tissue, appendages, muscle, or blood?
Although it is the norm for oral and transdermal dosage forms, the determination of
drug concentration in the blood cannot often be employed routinely with topical
formulations because circulating levels of drug are usually too low to be analyzed
by conventional techniques. Additionally, it can be argued that the relevance of any
serum (plasma) concentration–time curve of a topical agent is questionable, because
the curve reflects the amount of drug after the active moiety has left the site of
action.
According to the definition of bioavailability, drug determinations are intended
to reflect the rate and extent to which the active ingredient or active moiety becomes
available at the site of action. A series of methods are available to quantify or localize
drugs in the skin and muscle (14–17). However, even with the most sophisticated
methods currently available a straightforward answer is not found. Usually, several
bioanalytical techniques are often combined to deliver a more comprehesive picture
of the local situation. Development of new and improvement of existing bioanalytical
techniques to measure drugs in the skin and muscle is an active area of research
that, similar to the effect of the introduction of techniques to measure drug levels in
blood in the 1960–1970s, will transform understanding and lead to improvements
in biopharmaceutical and therapeutic quality of topical agents.
Another complicating circumstance is that our current knowledge of the target

site of many dermatological diseases is still limited. In viral, bacterial, or fungal
infections of the skin or soft tissue, both the germ and the localization of the germ
are identified: the target site is known. However for most diseases that require local
treatment, neither the target site nor the cause of the diseases, which might give some
insight into the site of action, are known.
As a consequence, the bioavailability of a topically applied drug or bioequiv-
alence of two topical dosage forms may be determined by several means, specifically,
(a) pharmaceutical parameters, (b) biopharmaceutical parameters, or (c) therapeutic
parameters, or a combination of these.
C. The Different Equivalencies

1. Pharmaceutical Equivalence
Pharmaceutical parameters describe the influence of a dosage form on a drug. This,
for example, may be the stability of a drug or the release rate of a drug from the
dosage form. Some regulatory agencies, therefore, have defined pharmaceutical
equivalence as medicinal products that contain the same amount of the same active
substance(s) in the same or similar dosage forms and that meet the same or com-
parable pharmaceutical standards. The dilemma is, for example, that two transdermal
delivery systems, containing different amounts of the same drug, that are delivered
to the body with the same kinetics under identical conditions (i.e., they are bio-
equivalent), are by definition, pharmaceutically inequivalent. This issue becomes
even more complex when semisolid dosage forms, in all their diversity, are com-
pared. This leads to the definition of pharmaceutical alternatives that are medicinal
products that contain the same therapeutic moiety, but differ in the salt or ester of
that moiety, or in dosage form type or strength.
2. Biopharmaceutical Equivalence
Pharmacokinetic parameters describe the activity of the body on the course of dis-
position of the drug. Pharmacokinetics denotes characterization of drug concentration
C, in a biological matrix as a function of time. Concentration in the sampled matrix,
most often plasma, is denoted Cm(t ), emphasizing the time dependence of the pa-
rameter. Concentration at the site of drug action (effector–receptor) Ce , may be
estimated by mathematical modeling. Pharmacokinetic parameters lead to the defi-
nitions of bioavailability and bioequivalence as described. Bioequivalence tests for
pharmaceutical products (tablets, capsules, and transdermal delivery patches) that are
based on pharmacokinetic parameters are largely standardized and corresponding
statistical evaluation methods and acceptance criteria are defined (18,19). For dosage
forms intended for topical and regional dermatological treatment, determination of
relevant pharmacokinetic parameters and their acceptance criteria is still being de-
bated (19).
It may happen that a well-designed generic product appears to be biopharma-
ceutically superior to an old innovator product. This case applies particularly to
products for topical and regional dermatological treatment, as current standards of
absolute bioavailability are often in the range of 1–5% of dose applied, such that
there is considerable scope for increase in bioavailability. The continuous improve-
ment of an innovator product by its manufacturer may also lead to an augmented,
suprabioavailable, product. In such a case, the competent generic or innovator man-

ufacturer should not be punished by the verdict bioinequivalent. Regulatory author-
ities intend to make biopharmaceutical improvements worthwhile. Adaption of the
strength, however, may be necessary to make the dosage unit bioequivalent to ex-
isting ones. Furthermore, the suprabioavailable product may become a new reference
product, indicating that marketing a product is a process that requires continuous
improvement of the biopharmaceutical quality (20).
Pharmacodynamic parameters describe the activity of the drug on biological
processes that may be only indirectly and poorly related to the therapeutic effect.
Measurement of effects on a physiological or pathophysiological process as a func-
tion of time after administration of two different products may serve as the basis for
bioequivalence assessment, in a manner analogous to that based on pharmacokinetic
data. Effect parameters do not necessarily avoid problems inherent in use of con-
centration parameters for assessment of bioavailability. Effect data, although per-
ceived as closely allied to efficacy, will usually be less relevant than drug concen-
tration data. This is because efficacy and pharmacological concentration–effect
relations are likely to be different, yet are related by drug concentration. Furthermore,
effect measurement may be so imprecise that statistical criteria can be met only with
large numbers of subjects. However, the effects of various drug classes are amenable
to accurate, precise, and reproducible quantification (e.g., the miotic response of the
pupil elicited by chlorpromazine). For dermatological products, only the vasocon-
strictor test has so far played an important role (21–23).
Pharmacodynamics can be an alternative to pharmacokinetics in bioavailability
assessment, or an adjunct to substantiate the relevance of pharmacokinetic measures.
Understanding the concentration–effect relation is important to predict the limitations
of either pharmacokinetic or pharmacodynamic assessments. Thus, pharmacokinetics
and pharmacodynamics are complementary approaches. Furthermore, integrated
pharmacodynamic–pharmacokinetic methods can provide a comprehensive descrip-
tion of the relations among dose, concentration, effect, and time. Recent guidelines
on bioequivalence testing of topical corticosteroids exemplify this integrated ap-
proach (21). As with bioequivalence tests that are based solely on pharmacokinetics,
assumptions are needed (e.g., that single-dose assessment is predictive of multidose
treatment [steady state], and that studies with healthy subjects are predictive of ther-
apeutic response in patients).
3. Therapeutic Equivalence
Some regulatory agencies consider a medicinal product therapeutically equivalent
with another product if it contains the same active substance or therapeutic moiety
and, when administered to the same individuals, shows the same efficacy and toxicity
as that product, for which efficacy and safety have been established. The safety
requirements exclude medicinal products with essentially identical therapeutic out-
comes, but with different therapeutic moieties.
For some oral drugs, the concentration–effect relation is well known, and phar-
macotherapeutic experience with the active substance is extensive and well described.
In these cases, therapeutic studies may not be necessary. For topical drug products
these requirements are generally not met. If therapeutic studies are necessary, they
should prove therapeutic (clinical) efficacy (the term efficacy is restricted to the
desired therapeutic endpoint: what is the physician trying to achieve for the patient?)
in the dosage scheme recommended for the indications claimed, and they should
compare frequency and severity of side effects with a standard treatment or with a
comparably approved drug product.
D. Statistical Considerations
The standard experimental design for bioequivalence comparisons between two oral
drugs is a crossover design that permits comparisons with far greater precision than
a parallel group study with the same number of subjects, or equally precise com-
parisons using fewer subjects.
For topical drug formulations, an experimental design that corresponds to a
crossover is a bilateral-paired comparison, in which both formulations are tested in
the same subject at the same time, on contralateral body sites. With simultaneous
testing of topical preparations, it is possible to test more than two sites on the same
volunteer or patient at the same time. By doing so, it is also possible to test more
than two formulations simultaneously. Alternatively, the formulations being studied
could each be applied at multiple sites on the same subject. This within-subject
replication would further reduce statistical error, with the consequence of greater
precision of results or a reduced sample size requirement. Advantages of the bilateral-
paired comparison design relative to crossover designs for oral products may be
partly offset by the apparently greater variability in absorption of topical drugs in
comparison with oral drugs, (19). A bioequivalence study of topical agents, therefore,
may require a larger sample size than a study of oral agents, with their less variable
rates.
When bioequivalence studies are performed for oral drug formulations, there
is an explicit attempt to select healthy volunteers within a narrow age range to
minimize interindividual variability to focus on variation between formulations. The
issue of whether observations in such a group can be generalized to patients in whom
the drug might be used has not been fully resolved. Similarly, for topical prepara-
tions, it is not clear whether the results of tests carried out on healthy skin can be
generalized to patients with diseased skin. This issue of the interaction between the
formulations and the absorption site has received little study (24).
In conducting statistical tests of bioequivalence, it was the practice, until the
early 1980s, to analyze the results using a standard statistical method, such as a t-
test or analysis of variance (ANOVA). These methods are designed to test the alter-
native hypothesis of a difference between the agents against the null hypothesis of
no difference. These methods are not applicable to bioequivalence studies and do
not validly test equivalence (25–27); they do not address the study hypothesis that
such studies should address, which is that two formulations are bioequivalent. Hauck
and Anderson (26) reasoned as follows: The test for formulations in the analysis of
variance is specifically a test of the null hypothesis that the average bioavailabilities
are equal, against the alternative that they differ. If the study is sufficiently sensitive,
it is possible to detect significant differences of little clinical meaning. Alternatively,
if the study is sufficiently insensitive, large differences will not be detected; the
evidence will not be sufficiently strong to accept the alternative hypothesis. They
believed that when one has a hypothesis to demonstrate, in this case the hypothesis
of bioequivalence, the logic of hypothesis testing requires that the hypothesis be
demonstrated to be the alternative hypothesis. They proposed a new decision rule,
starting by interchanging the null hypothesis and the alternative hypothesis in con-
nection with preset lower and upper bioequivalence limits, ␪1 and ␪ 2 , respectively.
For the multiplicative model, with 0 < ␪1 < 1 < ␪ 2 , the hypotheses are as follows:
Null hypothesis (H0): There is no bioequivalence (bioinequivalence). H01:

␮ test /␮ ref ⱕ ␪1 or H02: ␮ test /␮ ref ⱖ ␪1.
Alternative hypothesis (H1): There is bioequivalence. H1: ␪1 < ␮ test /␮ ref < ␪ 2.
For the additive model the corresponding hypotheses have to be formulated in

terms of differences (28). Hypothesis reversal implies that ␣ and ␤, the type I error
and the type II error, respectively, are also interchanged. The type I error (␣) now
is the consumer risk of incorrectly concluding bioinequivalence. Consequently, sta-
tistical power (1 ⫺ ␤) is now the probability of correctly concluding bioequivalence.
There is considerable experience with the design and analysis of bioequivalence
studies for oral formulations (29). Principles and practices for oral products may be
applied to topical products. In particular, the paired comparison study for topical
products is a direct analogue of the crossover study that is standard for oral products.
Advancements in statistical methods over the last decade do not seem to have be-
come generally adopted in the topical arena.
III. MODEL SYSTEMS

A. In Vitro Model Systems
In the last decades much effort has been directed toward the development of in vitro
model systems for the biopharmaceutical evaluation of topical formulations. Aca-
demia, industry, and regulatory authorities have developed active cooperation to stan-
dardize these model systems (30). Because of the diversity of the in vitro model
systems available, a classification of the models is useful, as it is helpful to draw a
clear distinction between two main model groups: the drug-release models, and the
drug-absorption models. These models can be used, respectively, for quality assur-
ance and for formulation development of topical dermatological drug products.
To study the quality of topical drug products, the release rate of the drug from
the formulation should not be influenced by the matrix (membrane–liquid receptor)
into which the drug diffuses. A plot of the amount of drug released from a semisolid
formulation per unit area (␮g/cm2) against the square root of time (t 1/2) may yield a
straight line, the slope of which represents the release rate (31,32). This release rate
is formulation-specific and can be used to monitor batch-to-batch uniformity.
In formulation development it is useful to have in vitro models that will predict
the relative effect of formulation changes on transport across membranes. In contrast
to drug-release models, the drug-absorption models should give linear plots with
time, indicating that transport is controlled by the membrane. By using artificial
membranes, the drug-absorption model is of use only in predicting drug–vehicle
interactions (i.e., the thermodynamic activity of the drug in the formulation). Models
with artificial membranes are not predictive for vehicle–membrane interactions (e.g.,
effects of chemical enhancers).
Figure 1 gives an overview of the in vitro model systems.
Figure 1 Classification of in vitro models for the evaluation of topical formulations. (Ex-
panded from Ref. 33.)
1. Drug-Release Models
A simple, reliable, reproducible, relevant, and generally accepted in vitro method to
assess release of drug from topical dermatological products would be highly valuable
for the same reasons that such methods have proved valuable in the development
and batch-to-batch quality control of solid oral dosage forms. For this reason, the
Food and Drug Administration (FDA), Division of Bioequivalence in the Office of
Generic Drugs, and the Division of Anti-Infective Drug Products in the Office of
Drug Evaluation II, recommends in an Interim Guidance (34) the development of in
vitro release methods for topical dermatological products to accompany in vivo phar-
macodynamic and pharmacokinetic data. This guidance (July 1, 1992, final guidance
June, 1995) provides detailed recommendations to pharmaceutical sponsors on meth-
ods to document bioequivalence of topical corticosteroids.
The two major categories of topical products, transdermal patches and der-
matological (topical) products, pose different issues for the development of an in
vitro release method.
a. Drug-Release Methods for Transdermal Patches. At present, the United
States Pharmacopeia (USP ) has identified three different apparatus–procedures for
determining the release rate of transdermal patches (35). These procedures are com-
plex and not universally applicable to the different types of patches presently on the
market. Therefore, they have developed a simple method that can be employed to
determine in vitro release profiles of all marketed transdermal patches. The method
employs a watchglass–patch–Teflon mesh sandwich assembly and the paddle
method (USP apparatus 2) to determine the release rate of the product, as a quality
control procedure to assure batch-to-batch uniformity. The method is applicable to
all brands, shapes, and strengths of nitroglycerin patches (36) and gives a release
profile that is comparable with that obtained with other more complicated and cum-
bersome compendial methods (37). The watchglass–patch–Teflon sandwich method
has also been applicable to the release test method for scopolamine, clonidine, and
estradiol patches. It also yields release data comparable with those obtained from
other methods (38).
b. Drug-Release Methods for Semisolid Formulations. Among the first ex-
periments performed on the release of drugs from ointments were those developed
by Reddish (39) and Christoff (40). The Reddish procedure involved the measure-
ment of the antiseptic properties of ointments containing various drugs. The exper-
iments were performed by applying ointment to agar that had been inoculated with
Staphylococcus aureus. After incubation, the agar plates were assessed to determine
the presence of a zone of inhibition. Thus, a crude measure of the ointment’s ability
to liberate a drug was achieved. A variation of this method was introduced (41) in
which the minimum time required for an ointment to cause inhibition of bacterial
growth was used as an index of ointment efficacy.
Two procedures, one in which the drug product is placed in direct contact with
the receptor phase (42–45), and the other in which the drug product and the receptor
medium are separated by a membrane (46–52), have been used. In the first case, the
possibility of product mixing and sloughing into the receptor media exists, making
the release rate determination difficult.
Recently, researchers at the FDA have developed a simple method to determine
the release of corticosteroids from cream, ointment, and lotion preparations. The
method uses a commercially available static diffusion cell system and a synthetic
cellulose acetate membrane. The use of a commercially available synthetic membrane
obviates the problems associated with the preparation and variability of skin mem-
branes. In a recent study (32), hydrocortisone cream was placed in the donor chamber
on the synthetic membrane and the aqueous buffer was used as the receptor medium.
The release profile of hydrocortisone was determined over a 6-h interval and plotted
as the amount released versus t 1/2. Hydrocortisone from different batches of the same
manufacturer showed similar release rates, but these differed from those of hydro-
cortisone from another manufacturer (53). When the procedure was used with creams
containing sparingly soluble corticosteroids, use of a hydroalcoholic medium as a
receptor phase was essential to increase the drug solubility to maintain sink condi-
tions. Betamethasone valerate cream from two manufacturers showed significantly
different release rate profiles under such experimental conditions (54). When em-
ploying a hydroalcoholic medium as a receptor phase, there is a question of back-
diffusion of alcoholic medium through the synthetic membrane and alteration of the
integrity of the cream preparation. With betamethasone valerate cream, for which
ethanol–water (6:4) was used as the receptor phase, mere traces of alcohol were
detected in the cream (donor chamber) at the end of a 6-h study. Microscopic ex-
amination of the cream before, and after, the in vitro release experiment showed no
differences (55). Thus, the concentration of ethanol had a negligible effect on cream
integrity (test criteria not presented!). Hydrocortisone ointment showed a much
slower rate of release when compared with the cream (32).
c. Conclusion. The objective in development of drug-release models is to
minimize the effect of the membrane on in vitro release. Systems without mem-
branes, which are feasible only if the polarity of the vehicle and the receptor fluid
are very different, are not widely used. Today, systems with membranes are more
common. The membranes used may vary from relatively wide-mesh mechanical
supports (56), to microporous membranes (57). Corbo et al. (57) and Shah et al.
(32,53–55) provide some excellent examples of the development and validation of
these methods. Guy and Hadgraft (58) have argued the case for flexibility in the
design of these models, especially in the composition of the receptor phase, to allow
the test to be optimized for drugs of different physicochemistry, especially aqueous
solubility. Most methods use the static diffusion cell system (Franz-type cell). This
cell has been automated for ease of use and to improve reproducibility (32,59). Very
recently the ‘‘Enhancer Cell,’’ a modification of the USP tablet-dissolution apparatus,
has been evaluated for in vitro release testing of topical products (60).
The importance of measuring the in vitro release rates of drugs from topical
preparations and the usefulness of drug release data is further discussed elsewhere
(33,61–64).
2. Drug Absorption Models
Drug absorption models are useful in the design and development of formulations.
Some of these methods are described in more detail in Chapter 5.
a. Synthetic (Artificial) Membrane Models. Synthetic membrane models do
not measure rates of transport equivalent to human skin, nor do they predict the
interaction of vehicle components to modify skin permeability (e.g., effects of chem-
ical enhancers). Silastic (polydimethylsiloxane; silicon rubber) has been widely used
as a hydrophobic membrane to model diffusion relevant to transport across human
skin. For example, Flynn and Smith (65) found a linear relation between the per-
centage saturation, thus, thermodynamic activity, of p-aminoacetophenone and its
steady-state transport across Silastic. Also, Tanaka et al. (66) measured the transport
of hydrocortisone butyrate propionate from various vehicles under open conditions.
Transport was directly related to drug solubility; hence, the thermodynamic activity,
in the vehicle. Because Silastic models may show the t 1/2 plots, validation of the in
vitro model is required to ensure that transport is rate-controlled by thermodynamic
activity only. Davis and Hadgraft (67) have described the development and validation
of such a method for use in design of supersaturated solutions of hydrocortisone
acetate. Hadgraft and Ridout (68) developed novel artificial lipid membranes to
mimic the stratum corneum barrier. However, transport rates are approximately 100
times less than across stratum corneum. Other systems, which may be suitable only
for use in the design of formulations to optimize thermodynamic activity, include
transport across cell-cultured epidermal membranes (69,70). However, these systems
may, more readily than Silastic, also respond to effects of vehicle components on
the barrier, which may complicate interpretation of the results.
b. Skin Membrane Models. Skin membrane models, both human and animal
skin, are widely used in the development of topical dermatological formulations. The
consensus of the FDA/AAPS-sponsored Workshop on Principles and Practices of In
Vitro Percutaneous Penetration Studies (30) was that human skin is the preferred
membrane, although the value of studies on animal skin in the formulation devel-
opment stage was recognized. Human skin is notorious for its high level of barrier
variability (71).
3. In Vitro–In Vivo Correlations
Several authors have described the correlation of in vitro release or drug absorption
studies with pharmacological effects, such as the onset time of nicotinate erythema
(72–75) and vasoconstriction of corticosteroids (54,66,76,77). Furthermore, the per-
meability of artificial membranes was also correlated with pharmacokinetic param-
eters, such as the maximum plasma concentration and the amount of drug excreted
in urine for salicylic acid (47), indomethacin (78), and nitroglycerin (79).
In vitro release will be influenced by several factors, including bulk viscosity
and microviscosity of the vehicle, polarity of the vehicle relative to the receptor
phase, hydrodynamic pressure and, also, the thermodynamic activity of the drug in
the vehicle. Of these factors only the latter affects permeation in vivo. Thus, and
considering only simple vehicles without enhancers, in vitro-release studies have the
potential only to predict in vivo permeation when thermodynamic activity dominates
the release process. Figure 2 gives an example in which release rates correlate with
the amount of drug that penetrated into the stratum corneum and to a pharmacolog-
ical response (vasoconstriction) (77).
Also, as shown in Figure 3, Shah et al. (54) found a similar correlation with
betamethasone valerate creams. However, no correlation was found in an earlier
Figure 2 Correlation between pharmacodynamic (PD; vasoconstriction at 16 h), pharma-

cokinetic (PK; in vivo; ␮g/cm2 in the stratum corneum at 24 h), and in vitro release char-
acteristics (PK; in vitro mg/cm2 released in 6 h) of two hydrocortisone creams. (From Ref.
77.)
Figure 3 Correlation between pharmacodynamic response (PD; vasoconstriction at 16 h)

and in vitro release rate into ethanol/water (6/4) (PK; in vitro mg/cm2 released in 6 h) for
betamethasone valerate cream. (From Ref. 54.)
study on hydrocortisone in which an ointment showed a much slower rate of release

compared with cream (32). It is likely that the microviscosity of the ointment caused
diffusion through the ointment to rate-limit in vitro release.
Less attention has been given to the correlation of in vitro synthetic membrane
absorption models to in vivo absorption, although by definition, these should be
better correlated with in vivo absorption in that they respond to drug thermodynamic
activity within the vehicle. However, both in vitro release and synthetic membrane
absorption models will fail to predict the effect of chemical enhancers on permeation
in vivo.
Of all of the in vitro methods, transport across human skin has the highest
potential to be predictive of in vivo permeation in humans. How useful the prediction
is has been addressed by Franz (80). He compared the permeability of 12 organic
compounds across excised skin with that previously found in vivo (81). Franz found
a rank order in vitro–in vivo correlation that is reasonable considering the substantial
differences in experimental conditions between the studies. A review of more recent
data, comparing in vitro–in vivo human skin data under controlled conditions is
urgently required to reevaluate this correlation.
B. In Vivo Model Systems

A wide variety of in vivo techniques have been developed to study kinetic, dynamic,
and therapeutic (clinical) parameters of topical drug application. In the following
section some techniques to sample drug(s) in the biological matrix and to measure
drug effects in the skin will be reviewed and discussed. Owing to the large amount
of data available, only a selection of information is presented. Although studies in
animal models are useful in the development of topical formulations, the variation
in percutaneous absorption between animals and humans requires that bioequivalence

is confirmed in the latter.
1. Kinetic Models
a. Skin-Stripping. Tape-stripping is a technique that is useful in dermatolog-
ical research for selectively and, at times, exhaustively removing the skin’s outermost
layer, the stratum corneum. Typically an adhesive tape is pressed onto the test site
of the skin and is subsequently abruptly removed. The application and removal pro-
cedure may be repeated 10 to more than 100 times (82,83). Such stripped skin has
been used as standardized injury in wound-healing research. The technique has been
adapted for studying epidermal growth kinetics (84–88), and it may also be useful
as a diagnostic tool in occupational dermatology to assess the quality of the horny
layer (89,90).
The observation that the skin may serve as a reservoir for chemicals was orig-
inally noted in 1955 (91). The localization of this reservoir within the stratum cor-
neum was demonstrated for corticosteroids (92) and confirmed by others (93–96).
The introduction of the tape-stripping method to further investigate the reservoir and
barrier functions of the skin was a significant expansion of this experimental tool in
skin research (97–99). Some of the following questions are approached.
Differences in the permeability of intact and fully stripped skin have provided
information on the diffusional resistance of the skin (100). It has been recognized
that complete removal of the horny layer after 30–40 strippings was not possible
(101), and after local application of hydrocortisone to stripped skin, a significant
barrier function still exists (102,103).
The presumption that factors which improve percutaneous absorption also re-
sult in an increase in stratum corneum reservoir (92,96) made the tape-stripping
method a possible approach for selecting or comparing vehicles for topical drugs.
Data from tape-stripping experiments were, therefore, related to (a) chemical pene-
tration into skin, (b) chemical permeation through skin, (c) chemical elimination from
the skin, (d) pharmacodynamic parameters, and (e) clinical parameters.
In in vitro and in vivo investigations with the tape-stripping technique, drug
penetration into the stratum corneum, as determined by quantification of radiolabeled
chemical on the tape after being removed from the skin, was clearly vehicle-depen-
dent (97,104–106). The tape-stripping technique has been validated and standardized
(107–110). This stripping method determines the concentration of a chemical in the
stratum corneum at the end of a short application period (30 min). A linear relation
was found between the stratum corneum reservoir content and in vivo percutaneous
absorption (total amount of drug permeated in 4 days) using the standard urinary
excretion method (81,111,112). These investigations also showed that, for a variety
of simple pharmaceutical vehicles, the percutaneous absorption of benzoic acid is
vehicle-dependent and can be predicted from the amount of the drug within the
stratum corneum 30 min after application. Once validated, the major advantages of
the tape-stripping protocol are the elimination of urinary and fecal excretion to de-
termine absorption, and the applicability to nonradiolabeled determination of per-
cutaneous absorption because the skin strippings contain adequate chemical concen-
trations for nonlabeled assay methodology. Despite that the assay provides reliable
prediction of total absorption for a group of selected compounds, mechanistic inter-
pretations are still rare. From the data of Rougier and co-workers, Auton (113) re-
cently presented the first mathematical approaches that may help explain some of
the foregoing observations.
The tape-stripping technique has also been used to analyze biological activity,
thus taking into account binding, decomposition, and metabolism of a given drug
(82,114–119). Pershing and co-workers (120–123) simultaneously compared a skin-
blanching bioassay with drug content in human stratum corneum following topical
application of commercial 0.05% betamethasone dipropionate formulations (phar-
maceutical complex vehicles). The rank order of the betamethasone dipropionate
formulation potency is similar between the visual skin-blanching assay, the tape-
stripping, and the ‘‘a scale’’ on the chromameter. The rank correlation between tape-
stripping method and the skin-blanching response was moderate to good.
Owing to a series of still unanswered questions and the potential of the tape-
stripping technique a variety of parameters, such as the properties of adhesive tapes,
vehicle effects on the harvesting of stratum corneum, and tape application and re-
moval procedures, are currently under systematic investigation (120,124–126). Re-
cently a Draft Guidance for Industry Document has been published by the FDA
(126a). This guidance, available on the World Wide Web (FDA, CDER) has caused
a heated debate (126b).
b. Skin Biopsy. The most invasive—but still practicable—method to access
skin compartments is the excision of skin tissue. In contrast to the other methods
described, the punch and the shave biopsies allow a direct ingression into the com-
partment of interest. After removal (optional) of the stratum corneum from skin with
an appropriate technique (tape-stripping, glue) the punch biopsy will represent parts
of subcutaneous tissues, dermis, and epidermis, and the shave biopsy will represent
mainly epidermis and some dermis. Parts of the stratum corneum may remain on
the epidermis, depending on the method used for stratum corneum removal. Sub-
cutaneous tissue can mechanically be divided from the dermis, and dermis can be
separated from the epidermis by heat (127). Human skin samples larger than 100
mg are difficult to obtain, and the usual amount that can be harvested is less than
50 mg.
Surber and Laugier (128,129) compared the acitretin concentration (synthetic
retinoid) in human skin after oral and topical dosing using three different skin-
sampling techniques: punch biopsy, shave biopsy, and suction blister. All three skin-
sampling techniques have been used by various investigators to quantitate drugs and
xenobiotics in the skin. Each technique gives access to distinctive skin compartments
(Fig. 4).
Despite that therapeutic acitretin concentrations were found after topical and
oral treatment, no beneficial effects in psoriasis, or other disorders of keratinization,
are observed following topical administration of this drug. Similar observations have
been made with methotrexate and cyclosporine (130,131). One may postulate that
drug concentration at a particular site within the skin following both routes of ad-
ministration could be different owing to the direction of the drug concentration gra-
dient. This hypothesis has also been postulated and illustrated by Parry et al. (118),
comparing the clinical efficacy of topical and oral acyclovir. Model predictions and
in vivo data agree that the topical administration of acyclovir results in a much
greater total epidermal acyclovir concentration than after oral administration. How-
ever, mathematical modeling of the acyclovir concentration gradient through the
Figure 4 Schematic diagram of intact skin (left) and the skin specimens taken using three
sampling techniques: BS, blister skin; BF, blister fluid.
epidermis revealed that the drug concentration at the target site of the herpes simplex
infection, the basal epidermis, was two to three times less after topical administration
than after oral administration. Furthermore, one may postulate that drug metabolism
could be different, depending on the route of administration. Data supporting this
hypothesis are still incomplete.
Despite skilled experimenters, sophisticated sampling techniques, and instru-
mentation, the information gained from tissue sampling is probably only an estimate
of the chemical distribution within the skin. Because of possible interlaminate drug
contamination, accurate and specific information on drug localization within a par-
ticular skin compartment following both routes of administration is not obtained by
these methods. The following two techniques have been used to obtain skin sections,
parallel to the outer skin surface to a depth of about 300–600 ␮m.
Semiautomated skin-sectioning technique. With this approach the skin tissue
from a punch biopsy is placed on a cryomicrotome table and 10- to 40-␮m sections
are cut parallel to the skin surface. This technique is excellently summarized in the
form of a standard operating procedure (SOP) (132,133). When the cylinder-shaped
punch biopsy is placed, dermis-side down, on a microtome chuck maintained at
⫺17⬚C it is essential that the stratum corneum surface is parallel to the cutting plane.
This can be accomplished when several metal rings of a total thickness slightly
greater than that of the skin sample are placed around the tissue, and the rings are
filled with embedding medium. Subsequently, a flat, precooled surface (glass slide)
is pressed against the stratum corneum, resulting in a flat outer surface. The whole
assembly is then frozen in approximately 10 s and can be stored. The technique,
however, does not take into account the dermal papillae, and a clear separation of
histologically distinct compartments is not possible. The method has predominately
been used to characterize the pharmacokinetic and metabolic behavior of topically
applied drugs (15,104).
With excised human skin and tissue grafted to athymic mice, the in vitro and
in vivo concentration profiles of salicylic acid and salicylate esters were obtained for
the outer several hundred microns of the skin (15). The results show significant
differences in the extent of enzymatic cleavage and distribution of metabolites be-
tween in vitro and in vivo studies. The data also suggest that in vitro results may
overestimate metabolism because of increased enzymatic activity or decreased cap-

illary removal (Fig. 5).
Manual skin-sectioning technique. With the development of the skin sandwich
flap model (134), Pershing and co-workers (117) also proposed a new manual skin-
sectioning technique. With this technique, the profile of a test compound’s disposition
in skin after topical administration is examined. The technique requires the use of
radiolabeled compound and fresh skin, without freezing. The sectioning technique
uses cyanoacrylic cement, Scotch tape, glass slides, 115-␮m–thick glass coverslips,
a thin razor blade, and a fresh 2-mm–skin punch biopsy. Briefly (Fig. 6), a 2-mm–
skin punch biopsy is sectioned in 115-␮m–thick sections starting from the stratum
corneum (see Fig. 6a) completely through the biopsy, by dipping the stratum corneum
end of the biopsy into cyanoacrylate adhesive (step 1) and placing the biopsy on a
microscope slide between two 115-␮m–thick microscope coverslips that are held in
place on the microscope slide with tape (step 2). A single-edged razor blade is used
to shave the stratum corneum side of the punch from the remainder of the biopsy.
The remaining biopsy (see Fig. 6b–d) is then dipped again in the cyanoacrylate
adhesive in the original orientation (stratum corneum face down) and placed in a
new location on the microscope slide (see Fig. 6b). The remainder of the skin biopsy
is sectioned similarly, producing multiple 115-␮m sections on an individual micro-
scope slide (step 3). After complete sectioning of the skin biopsy, the various skin
wafers on the microscope slide are individually digested with tissue solublizer and
submitted to liquid scintillation counting for drug quantification and, subsequently,
a concentration gradient and disposition profile of the test compound in the skin is
reconstructed. The authors also report that freezing the 2-mm biopsy results in re-
distribution of the drug from the stratum corneum to the middle sections (400- to
600-␮m–depth) of the skin biopsy, thereby portraying an inaccurate distribution
profile in the skin. Unfortunately, this statement is not further documented. Even
though the redistribution phenomenon was described only for the 2-mm biopsy, it
may also be relevant for the 4- to 6-mm biopsy. The statement makes a careful
reevaluation of the skin preparation procedure necessary.
c. Drug in Blood and in Excreta. In vivo percutaneous absorption may also
be determined by indirect method of measuring radioactivity in excreta after topical
application of a labeled compound. The compound, usually labeled with 14C or 3H,
is applied and the total amount of radioactivity excreted in urine or urine plus feces
is determined. The amount of radioactivity retained in the body or excreted by some
other route not assayed (e.g., CO2 or sweat) is corrected for by determining the
amount of radioactivity excreted after parenteral administration. This final amount
of radioactivity is then expressed as the percentage of applied dose that was absorbed
(81,111,112). The limitation of determining percutaneous absorption from urinary or
fecal radioactivity, or both, is that the method does not account for metabolism by
the skin or other organs. The radioactivity in excreta is a mixture of parent compound
and metabolites. This type of information is useful in defining the total disposition
of the applied topical dose. Plasma radioactivity can also be measured and percu-
taneous absorption determined by the ratio of the AUC following topical and par-
enteral administration (135). The same limitations discussed for excreta also apply
here.
A comparative example of three methods (136) was performed using [14C]-
nitroglycerin in the rhesus monkey. Topical bioavailability estimated from urinary
Figure 5 The concentration profile of (A) the diester, (B) monoester, and (C) salicylate,
as a function of the depth within human skin either (a) grafted to athymic mice, or (b) in
vitro. Results were obtained after 24 h of topical drug application. (From Ref. 15.)
418
Figure 6 Manual sectioning technique of skin biopsies (see text for details). (From Ref.
117.)
excretion was 72.7 ⫾ 5.8%. This was similar to the 77.2 ⫾ 6.7% estimated from
the plasma total radioactivity AUC. The estimates from urinary 14C and plasma 14C
were in agreement. The absolute bioavailablity estimated from the plasma’s un-
changed nitroglycerin AUCs was 56.6 ⫾ 2.5%. The difference in estimate between
that of the absolute bioavailability (56.6%) and that of 14C (72.7–77.2%) is the
percentage of compound metabolized in the skin during absorption. For nitroglycerin
this is approximately 20% (137).
Nitroglycerin is an example of a drug with a high absolute bioavailability.
However, most topically applied drugs have low absolute bioavailability. These data
demonstrate absolute bioavailability of topical steroids and retinoids ranging from 1
to 4% (91,103,111,138–142). Although not reviewed here, other series of com-
pounds, including azole antifungals, antivirals, and vitamin D3 derivatives also have
absolute topical bioavailabilities in this same range. Where bioavailability is low
there is potential for increased absorption owing to biological or pharmaceutical
reasons (see Sec. IV).
2. Dynamic Models
a. Quantification of Skin Color Changes. Skin color is predominantly deter-
mined by pigments such as hemoglobin, melanin, bilirubin, and carotene. Those
skin–color-determining components can be altered significantly by radiation (UV or
infrared light) and several substances (drugs or irritants). The quantification of ex-
perimentally induced color changes is a widely used method in dermatological and
cosmetic research, for the magnitude of the color response can be used as an indicator
of skin properties (integrity of the skin barrier or sensitivity), drug properties (con-
centration and bioavailability), vehicle properties (formulation and enhancer), pro-
tection properties (sun screen, and such) (143,144).
Until recently, assessment of skin color was performed by visual grading, based
on semiquantitative scales. Visual grading, however, needs a panel of trained exper-
imenters. Thus, only a limited number of research centers were able to generate
reproducible data. Despite that the human eye is highly sensitive in the discrimination
between colors close to each other, it has also been postulated that the human brain
is unable to memorize colors. As a consequence, the human eye is inadequate to
make comparisons at different times or at distant sites. To improve color readings,
standardized methods have been developed to express color in a more scientific way.
In tristimulus reflectance colorimetry, the reflected spectrum is analyzed at three
selected wavelengths, simulating the human color perception system. The three mea-
sured values can be converted to different color-indexing systems (Yxy system, Mun-
sell system, and the L* a* b* system) (145). Skin color measurements based on the
L* a* b* system proposed by the Commission Internationale de l’Eclairage (CIE)
has become an accepted color-indexing system (144–146). Color is expressed in a
three-dimensional coordinate system: L* represents brightness (varying between
white and black), a* represents hue and color on the green–red axis, while b* rep-
resents hue and color along the blue–yellow axis. Today the Minolta Chromameter
(Minolta Camera Co., Ltd., Osaka, Japan) is the most widely used instrument.
Erythema and tanning of the skin after photoprovocation with UVB were as-
sessed using the L* and a* parameters (146). Seitz et al. (147) used the a* parameter
to quantify the erythema, whereas the b* parameter was used for quantification of
the tanning. A positive relation between the a* color parameter and the perfusion of
the skin microcirculation in the quantification of the methyl nicotinate-induced ery-
thema was also detected (148). Wilhelm and Surber (149) challenged the skin with
different sodium lauryl sulfate concentrations. The irritation was evaluated with re-
flectance colorimetry, transepidermal water loss measurement, laser Doppler flow
measurement, and visual scores. All evaluation methods detected a significant sodium
lauryl sulfate dose dependency (dose–response). Moreover, a strong positive corre-
lation was detected between the different evaluation techniques. A similar positive
correlation was detected between transepidermal water loss measurements and the
a* parameter in the evaluation of surfactant-induced stratum corneum irritation
(150,151). Quantifing the erythema by means of skin color measurements (a* param-
eter) seems to be the most realiable method (152–154). Other approaches (e.g., laser
Doppler or transepidermal water loss) produce results with greater variability than
does the chromameter. The simultaneous use of these approaches is justified because
information from a different depth of the skin is collected.
Use of tristimulus color analysis for the quantification of skin color changes
induced by topical cortiosteroid prepartions has stimulated new research in bio-
screening for measuring topical corticosteroids potency (154a–c). For further infor-
mation see Guidance; Topical Dermatological Corticosteroids: In Vivo Bioequiva-
lence (21).
b. Skin Blood Flow (Laser Doppler Velocimetry). Laser Doppler velocimetry
(LDV) is a noninvasive, optical procedure allowing continuous, real-time monitoring
of skin blood flow. The measurement principle of the laser Doppler instrument is
based on the Doppler shift when reflected from moving particles. A low-energy
monochromatic laser beam (633 or 780 nm) is guided to the skin. The light illu-
minates a hemisphere with a diameter of 1 mm. The light is reflected by static
structures (not Doppler-shifted) and by moving red blood cells (Doppler-shifted).
The magnitude and frequency distribution of the backscattered light is related to the
number and the velocity of the red blood cells moving in the illuminated volume
(flux). The flux is measured in relative and dimensionless arbitrary units. The small
size of the illuminated volume causes a moderate-to-low reproducibility when the
probe has to be repositioned because of inhomogeneities in the microvascular ar-
chitecture. To decrease the spatial variability, the probe can be positioned above the
skin surface to increase the evaluated area and volume. Recently, the laser Doppler
technique has been considerably improved. Instead of a single-point measurement (a
major drawback of conventional laser Doppler flowmetry), readings are made with
either integrating probes, which use the signals from various adjacent measuring
volumes, or they are carried out over a scanned skin surface (up to 5 ⫻ 7 cm) to
build up a laser Doppler image (155–157). The new scanning laser Doppler tech-
nique is very useful in patch testing and psoriasis studies, because both the degree
and the area of increased blood flow can be determined.
The laser Doppler technique is particularly suitable for the assessment of local
pharmacodynamic changes induced by topical administration of vasoactive chemi-
cals. The technique has been extensively used to examine the vasoresponse in hu-
mans to topically applied esters of nicotinic acid (which are potent, local vasodila-
tors). Furthermore, these experiments have been analyzed to provide (a) insights to
percutaneous absorption process itself and (b) an understanding of the the effect of
formulation on the topical bioavailability of an applied drug (Fig. 7) (158–161).
c. Rhino Mouse Model. Adult rhino mice are hairless mutants that carry the
rhino gene, a recessive allele of the hairless gene (hrrh/hrrh). At birth, rhino mice are
indistinguishable from their nonmutant littermates. Shortly afterward, and before the
end of their first hair growth cycle, a defect in catagen results in irreversible hair
loss. The pilary canals widen, accumulate keratin, and undergo a transformation into
Figure 7 Percentage inhibition of the inflammatory reaction (LDV) induced by methyl

nicotinate in indomethacin gel-pretreated skin: no significant difference was observed between
formulations 1, 2, and 4, although urea (gel 2) did appear to have a greater enhancing effect
than ethanol (gel 3; p < 0.05). (From Ref. 159.)
horn-filled utriculi that resemble the human noninflamed microcomedone. Comedo-

lytic effects of various antiacne agents in the rhino mouse skin were first reported
(162,163) with use of qualitative methods. Later, a standardized histological and
image analysis method was introduced (164) to characterize and quantify the cuta-
neous morphological effects of antiacne agents. The following microscopic param-
eters can be assessed from skin slices obtained from skin biopsies before and after
treatment (165) (Fig. 8): on each open epidermal comedo, the largest diameter of
the utricle d, and the diameter taken at half depth D, can be used to calculate a
comedo profile (r = d/D). This ratio gives a measure of the morphological aspects
Figure 8 Effect of treatment with topical commercial preparations of retinoic acid: number
of (a) epidermal comedones; (b) comedo profile; and (c) epidermal thickness were measured
by image analysis methods (165). Animals were treated with (1) placebo, (2) Aberel gel,
0.025%; (3) Aberel cream, 0.05%; (4) Aberel lotion, 0.03%; (5) Retin A cream, 0.05%; and
(6) Retacnyl cream 0.025%. 100% control values were 69 ⫾ 1 epidermal comedones per
centimeter length of stratum corneum, comedo profile 0.6 ⫾ 0.02; epidermal thickness 21.0
⫾ 0.6 ␮m.
of the comedo. The surface of epidermis from an intercomedo area, the length of
the corresponding basal layer, and the number of epidermal comedones are further
parameters. These factors can be used for statistical analysis.
Commercially available topical preparations of retinoic acid, used clinically in
the treatment of acne, presented similar comedolytic activities (number of epidermal
comedones and comedo profile). The only significant differences observed among
the Aberel formulations, Retin A, and Retacnyl were related to the degree of epi-
dermal thickening. From the foregoing data, it would appear that comedolytic and
epidermal thickening activities of retinoic acid can be dissociated by formulation
changes. In consequence, differences in retinoic acid-induced skin irritation in for-
mulations such as Retin A and Retacnyl may be anticipated.
Possible systemic effects of the treatment were not discussed. Repeated appli-
cation (3 ⫻ 5 ⫻ 50 ␮L) of formulation to the body surface of the rhino mouse of
30–40% may produce systemic drug levels. Inadequate housing and animal behavior
may result in ingestion of formulations.
d. UVA-Induced Neutrophil Infiltration. A variety of human and animal
models have been developed for assessing the anti-inflammatory activity of topical
corticosteroids. Vasoconstriction is a rapid and popular screening technique, although
skin sites can vary in sensitivity to the assay (166). Human studies are troublesome
for both technical and ethical reasons: individual responses are highly variable, and
the tests are burdensome.
Based on an observation by Gilchrest et al. (167) that 50 J/cm2 of long-UV–
radiation (UVA) to human skin resulted in a modest infiltrate of neutrophils Wood-
bury et al. (168) and Kligman (169) irradiated hairless albino mouse skin with a
single large dose (195 J/cm2) of UVA. This produced a massive influx of neutrophils
into the dermis. This dramatic and quantifiable event formed the basis of a new
model for assaying topical corticosteroids.
Fifteen name brand corticosteroids (100 ␮L each) were evenly applied to dorsal
trunk skin immediately postirradiation and again 12 h later. Biopsies were taken at
24 h after irradiation when the density of neutrophils was at peak. The number of
neutrophils in the dermis was counted in ten fields per specimen (blinded) (Fig. 9)
(168). Further refinement and validation of the test (169) confirmed the initial study.
The data from this model are very promising and further investigation on sen-
sitivity and the influence of dosing regimens are necessary to prove the validity of
the assay.
C. Therapeutic Models and Clinical Trials

It is assumed that therapeutic models and clinical trials are the least accurate, least
sensitive, and least reproducible of the general approaches for measuring bioavail-
ability or bioequivalence. In general, they are not acceptable for formulations in-
tended for systemic availability if analytical methods are available to measure sys-
temic drug concentrations. However, clinical trials may be the only means to
determine the bioavailability or bioequivalence of dosage forms intended to deliver
the active moiety locally. Examples include topical preparations for the skin, eye,
and mucous membranes; oral dosage forms not intended to be absorbed (e.g., antacid,
radiopaque medium); and bronchodilators administered by inhalation if onset and
duration of pharmacological activity are defined (Federal Register 54:28941, 1989).
Figure 9 Comparison of name-brand and generic steroids: average number of neutrophils

per 1000⫻ field: (A) UVA only; (B) generic desoximethasone, 0.05%; (C) generic desoxi-
methasone, 0.25%; (D) name-brand desoximethasone, 0.05%; (E) name-brand desoximetha-
sone, 0.25%. Statistical analysis is shown in the inset. (From Ref. 169.)
In the following section some background into approaches that have been suggested
to evaluate bioavailability or bioequivalence for topical products for the skin are
discussed.
1. Grading of Disease States
Diseases that can be assessed and graded by objective means are rare in medicine.
Thus, the definition of numerical severity indices to grade a disease is helpful and
permits one to follow the course of a disease and to assess a particular treatment or
treatment modality of the disease. Particularly in dermatology, scores and severity
scales are widely used to grade a disease and to follow its course. A crucial part of
any assessment is the choice of what is to be measured to assess each patient’s
response.
The most primitive form of measurement is the grouping of individual obser-
vations into qualitative classes. Such a nominal scale is used in all studies in which
the patient’s response is classified as ‘‘success’’ or ‘‘failure.’’
If the measurement classes can be arranged in order of magnitude (e.g., the
overall grade of acne severity) (170), then one has an ordinal scale. In such a scale
each observation can be placed only on one class and the classes can be arranged
in order. Although ordinal measurements are often expressed as numbers, arithmetical
manipulation may not truly summarize the response measured. That is to say, the
mean score may be a poor way of summarizing responses measured on an improve-
ment-rating scale; the median, a nonparametric measure of location, is usually better.
Quantitative measurements of lesion area, lesion thickness, or other properties,
as determined by bioengineering techniques, duration to clearing, and such, add
greatly to the discriminatory power of a trial, and an attempt should always be made
to use them.
Frequently, the rating scales in dermatology are used only by their author, who
assumes that the scale measurement correlates with the responses under investigation.
In some areas of dermatological research important steps have been taken to develop
a consensus on the most meaningful rating scales (e.g., for atopic dermatitis: Con-
sensus Report of the European Task Force on Atopic Dermatitis; 171).
Lesion counts are a well-established method for assessing response in acne. A
potential bioequivalence study design may require a double-blind, 12-week, multi-
center, parallel study, including brand, generic, and vehicle in patients with mild-to-
moderate acne (grade 2–3). The primary endpoint will be lesion counts (inflamma-
tory and noninflammatory) and a global evaluation by the investigator. Patients are
evaluated at baseline and at 1, 2, 4, 8, and 12 weeks. Irritation and sensitization rates
will also be determined (F-D-C Reports. T&G-11, April 2, 1990; Topical antifungal
and acne drug product bioequivalence workshop, Nov. 16, 1989, FDA, Rockville,
MD). In several other studies the area, diameter, or thickness of lesion or the amount
of skin involved, expressed as an ‘‘intensity score’’ or as a percentage of body area,
have been used (172–174). Complicating factors include the need to instruct the
patients thoroughly on the use of the drug; assessment of patient compliance (e.g.,
by periodically weighing the residues in the container); the influence of menstrual
cycle in the course of acne; and the high placebo rate associated with acne treatment.
Published studies using photographic assessment are few, given the obvious
dependence of dermatology on visual skills for assessment. Despite that the accuracy
of photographic reproduction has been found unsatisfactory, presumably because the
many variables are not easily standardized, attempts have been made to find a con-
sensus (175,176).
In 1978 Fredriksson and Pettersson defined a ‘‘Psoriasis Area and Severity
Index’’ (PASI) and successfully assessed clinical efficacy of three different dosing
schedules of etretinate (177). The PASI score is based on severity and area of pso-
riatic lesions and is calculated by grading clinical symptoms (Eq. 1).
head trunk
PASI = 0.1(Eh ⫹ Ih ⫹ Dh )Ah ⫹ 0.1(Et ⫹ It ⫹ Dt )At⫹

upper extremities lower extremities
0.1(Eu ⫹ Iu ⫹ Du )Au ⫹ 0.1(El ⫹ Il ⫹ Dl )Al (1)
For calculation of this, the four main body areas were assessed: the head (h),
the trunk (t), the upper extremities (u), and the lower extremities (l), corresponding
to 10, 20, 30, and 40% of the total body area, respectively. The area of psoriatic
involvement of these four main areas (Ah , At , Au , Al ) was given a numerical value:
0 = no involvement; 1 = <10%; 2 = 10–30%; 3 = 30–50%; 4 = 50–70%; 5 = 70–
90%; and 6 = 90–100%. To evaluate the severity of the psoriatic lesions three target
symptoms; namely, erythema (E ), infiltration (I ), and desquamation (D) were as-
sessed according to a scale 0–4, where 0 means a complete lack of cutaneous in-
volvement, and 4 represents the severest possible involvement.
This index was further modified by adding the grading of the clinical symptom
‘‘pustulosis’’ and by reducing the grading scale to 0–3 (178,179). These values can
then be used for calculating a remission index (RI) giving the percentage improve-
ment in the psoriasis at the time TX , compared with the starting time TO (Eq. 2)
(180).
PASI TO ⫺ PASI TX
RI(%) = ⫻ 100 (2)
PASI TO
Despite that the PASI is easy to calculate, data collection remains inconvenient.
The main difficulty is the grading of the severity of the clinical symptom and ques-
tions of reproducibility may arise. For example, the grading of the ‘‘erythema’’ de-
pends highly on the grading of a previous observation. A score 2 will become 3 in
the case of an aggravation and will become 1 in case of an improvement. However,
it is impossible to say that a decline from score 3 to 2 depicts the same improvement
as a decline from score 2 to 1. The problem of grading becomes even more complex
when a scoring system is used in a multicenter study. The main problem of scoring
systems in dermatology is the mixture of clincal parameters of very different nature
that are more or less related and that entities of a very different nature are added
and multiplied, therefore, may be questioned. The localization and extent of a lesion
is easily assessed; however, the handicapping nature of the lesion remains in the
dark. Therefore, it is of utmost importance that the scoring system of clinical signs
is supplemented by patient judgments (e.g., pruritus, sleep loss, pain, or other). Per-
centages of change for an individual or a group expressed as numeric values from
clinical-scoring systems, therefore, should be supplemented by patient judgments
and, whenever possible, by additional means (colorimetry, histology, photography,
or other). An excellent example is the study design of Kimbrough–Green et al. (181)
for which the topical vehicle-controlled tretinoin treatment of melasma in black pa-
tients was studied. The melasma area and severity index (MASI) was supplemented
by colorimetric, photographic, and histological means.
Despite the problems inherent in scoring (scoring systems are usually devel-
oped for a particular disease), these systems, when standardized, are valuable in
assessing clinical signs more accurately than simple global evaluation. In some cases
the scoring systems have been used to study the influence of pharmaceutical strat-
egies (vehicle effect) by evaluating the pain-producing or pain-releasing effect [pain
questionnaire (182); pinprick test (183)].
2. Allergic Contact Reaction

The induction of allergic contact dermatitis lesions in hypersensitive subjects (hu-
mans and animals) are replicas of the natural pathology (184). Therefore, multiple
simultaneous experimental and therapeutic comparisons are possible in a double-
blind, randomized manner.
In a recent study, the immunosuppressive effect of tacrolimus (FK 506), cy-
closporine, dexamethasone, and clobetasol on dinitrofluorobenzene-induced contact
dermatitis was tested in the domestic pig (Table 1) (185). Test sites were treated 0.5
and 6 h after challenge with test solutions of drug vehicle. Efficacy of each prepa-
ration was determined clinically for intensity, extent of erythema, and consistency.
In addition, skin changes were biophysically characterized by measuring microvas-
cular perfusion and skin color changes.
The data clearly show that the topical administration of the macrolide antibiotic
tacrolimus (FK 506) and the corticosteroids have an immunosuppressive effect. Even
though a weak dose–response relation is seen with FK 506 and clobetasole, the
differences are not statistically significant over a dose range of 10.
Table 1 Influence of Topical Treatment with Drug and Placebo Formulations on the
Allergic Contact Dermatitis Test Sites in Domestic Pigs
Efficacy of drug formulation (% inhibition vs. placebo)

Microvascular
Drug Clinical evaluation Skin color perfusion
FK 506, 0.4% 72 77 64
FK 506, 0.13% 66 68 51
FK 506, 0.04% 75 51 40
Cyclosporine, 10% 9 8 4
Dexamethasone, 1.2% 31 38 25
Clobetasole, 1.2% 53 87 44
Clobetasole, 0.13% 44 76 37
Source: Ref. 185.
One may conclude that numerous intraindividual, anti-inflammatory activity

comparisons performed on a natural, controlled inflammatory disease using bio-
physical techniques is suitable both to determine drug delivery to the target sites in
the skin and to assess conventional therapeutic schedules. However, it seems unlikely
that minor changes in formulation will be detectable by these means.
In purely clinical situations, the influence of the vehicle and its ingredients on
the allergic contact dermatitis reaction has been observed (186). Several reports de-
scribe contact sensitization to topical acyclovir products (187–189). Extensive patch
testing revealed that the influence of the vehicle and its ingredients may be respon-
sible for the skin reaction. Baes et al. (190) describe two patients who had positive
patch test reactions to Zovirax (acyclovir) cream, but were negative to all its con-
stituents as well as to Zovirax ointment (paraffin as base). This phenomenon, which
is defined as compound allergy and refers to contact allergy to a preparation with
negative patch tests to all its constituents, is still rarely observed. However, the
observations made so far illustrate the necessity of patch-testing patients with sus-
pected contact allergy to their own topical products, as well as to their constituents,
to avoid missing compound allergy.
3. Infectious Conditions
a. Viral Infections. The topical delivery of antiviral compounds formulated
in various preparations has been studied in several cutaneous herpes models.
Eight different vehicles that were appropriate to deliver drug to the herpes
simplex virus-2 (HSV-2)-infected mucocutaneous surface of the vagina were studied
in mice and guinea pigs that were infected intravaginally with HSV-2. Efficacy of
the formulations were determined using survivor increase, mean survival time in-
crease, and reduction in mean vaginal lesion (Fig. 10).
To halt the viral replication and lateral spread in the basal layer during the
prodromal stage, various formulations were used to deliver interferon to cutaneous
herpes simplex 1 (HSV-1) lesions (192). The virus was inoculated with a spring-
loaded vaccination instrument. To quantitate the effects of formulation factors on
therapeutic effect a scoring system was used on a daily basis for up to 10 days. From
Figure 10 Comparison of effects of vehicles on 1% 2-acetylpyridine-semicarbazone

against HSV-2 infections of mice. (From Ref. 191.)
the vehicles used (solution, emulsions, multilamellar and large unilamellar lipo-
somes) only liposomes prepared by the hydration–rehydration method were effec-
tive. The authors hypothesize that the dehydration and subsequent rehydration of the
liposomes facilitate partitioning of interferon into liposomal bilayers, where the drug
is positioned for transfer into the lipid compartment of the stratum corneum. They
further note that liposomes do not appear to function as permeation enhancers, but
seem to provide the physicochemical environment needed for transfer of interferon
into the skin. This report describes a therapeutic model in which the influence of a
particular preparation technique (dehydration–rehydration) of a topical semisolid for-
mulation was observed.
A novel method based on the measurement of free drug concentration (C*), or
thermodynamic activity of drug at the skin target site, has been developed (193,194)
to assess bioavailability and to predict efficacy for antivirals and other dermatological
formulations. To measure C* following the administration of a topical formulation
entailed, first, establishing a correlation between the steady-state dermal drug flux
and an elicited efficacy. This was accomplished using a novel animal model in which
hairless mice were infected in a small spot at a lumbar skin area with cutaneous
HSV-1 (195,196). This induced, 3-days postinoculation, a narrow band of skin lesion
development along the peripheral neural path toward the spinal cord. Taking advan-
tage of this unique pattern of lesion development, an antiviral agent, such as acy-
clovir, was applied to a laurocapram (Azone)-pretreated skin area, dorsal to the virus
inoculation site and in the predicted path of lesion development, to limit the lesion
development. Five days after virus inoculation, the lesion development was scored
for each mouse and two different antiviral efficacy parameters were assessed sepa-
rately: (a) ‘‘topical’’ (local) efficacy, measured the antiviral activity of acyclovir de-
livery topically to the local skin area directly under the drug application site; and
(b) ‘‘systemic’’ efficacy, measured the antiviral activity of acyclovir delivery through
systemic circulation to the target site, presumably the epidermal basal layer (197).
To quantify drug flux, a transdermal delivery system was developed in this animal
model and the amount of acyclovir delivered to each infected animal could be con-
trolled during the time period of drug therapy through a rate-controlling membrane.
The actual (experimental) flux was determined at the end of each in vivo experiment
by carrying out an extraction of the residual acyclovir in the transdermal delivery
system. This extraction assay served to validate the expected (theoretical) flux or,
alternatively, provided the bounds of uncertainty to the drug flux in the particular
experiment. The results clearly showed a quantitative relation between the antiviral
efficacy and the experimental flux of acyclovir obtained from in vivo experiments.
Topical efficacy increased with increasing acyclovir flux in the range of 10–100 ␮g/
cm2 day⫺1. Given the relatively high precision of topical efficacy results, it is believed
that the quantitative nature of this animal model should be valuable in screening of
new antiviral agents for topical treatment of cutaneous herpesvirus infections and
the optimization of topical formulations. Two factors may limit the applicability of
this elegant approach to other classes of dermatological formulations. First, the drug
was delivered from a transdermal delivery system at a constant rate over several
days (with semisolid formulations, formulation application and formulation changes
are difficult to control) and, second, in the foregoing approach, the target for acy-
clovir (cutaneous HSV-1) and the localization of the target (epidermal basal layer)
is known. Unfortunately, the target sites for many dermatological agents are still
unknown.
b. Fungal Infections. In March 1990, the Division of Anti-infective Drug
Products and the Division of Bioequivalence published a draft guidance for the
performance of a bioequivalence study for topical antifungal products (198). During
the course of developing a new antifungal drug, a minimum of two well-controlled
clinical trials are required to show that the product is superior to the vehicle control.
For a generic product, a single, randomized double-blind, three-way parallel group
study has been proposed, comparing the test drug with the innovator, as well as with
the vehicle used for the generic product. The vehicle control helps eliminate inves-
tigator bias and permits a better measure of the effectiveness of the generic product.
If the reference product is approved for tinea pedis, the patients should have the
disease, as confirmed by clinical examination and mycological testing. If the clinical
efficacy is confirmed, then the generic product may be used for all of the conditions
of use specified for the innovator product. The evaluation is to be conducted 4 weeks
after starting treatment, at the time treatment is discontinued, and 2 weeks later. This
last evaluation is the most important in determining bioequivalence. Assuming the
test and reference products have identical success rates of 50%, it is estimated that
at least 108 evaluable patients will be needed for the test and reference products,
and 60 patients will suffice for the vehicle dose, as it should be easier to show a
difference between active product and vehicle than between active and active prod-
uct. If the success rate of the test drug is different from the reference drug (but still
within the equivalence region), more patients will be needed.
If the reference drug is approved for tinea pedis, the patients enrolled should
be infected with tinea pedis, proved by clinical examination and positive mycological
testing. Patients may be entered into the study and randomized on the basis of clinical
presentation and a KOH examination. Cultures should be taken, and only those
patients with clinical signs of tinea pedis, a positive KOH, and a recovered derma-
tophyte will be evaluated. Clinical and mycological examination (KOH and culture)
should be performed at the end of a 4-week treatment period and at 2-weeks
posttreatment.
The evaluation of therapeutic equivalence will be based on the percentage of
patients in each group who have a clinical cure and a mycological cure. A clinical
cure means that the signs and symptoms of the disease have cleared. Both the KOH
examination and culture must convert to negative to qualify as a mycological cure.
Each of the clinical and mycological outcomes with the reference drug must be
statistically comparable with those of the test drug. In addition, the cure rates for the
test and for the reference drugs must each be statistically superior to that of the
vehicle control for each measure of outcome for the study to be a valid evaluation
of the test product. Although these comparisons should be evaluated at the end of
treatment and at the 2-week follow-up visits, primary weight will be given to the 2-
week follow-up evaluation in determining whether bioequivalence has been
established.
A well-documented example of the foregoing types of studies that are appro-
priate in evaluating new formulations (199) involves bioequivalence studies of a
lotion and a cream formulation of the antifungal drug cyclopirox. During the devel-
opment of the products, in vitro studies were carried out using both human cadaver
and domestic pig skin treated with two formulations (cream vs. lotion). In vivo
studies included an animal model, during which guinea pigs were inoculated dorsally
with Trichophyton (tinea[sic])mentagrophytes and after 3 weeks, the resulting lesions
were treated with the lotion, cream, and the two vehicles. This test was followed by
a human model, where four sites on the forearms of human volunteers were inocu-
lated with T. mentagrophytes. Each site was treated with the lotion, cream, or the
two vehicles. After 15 days of treatment the sites were examined clinically and
mycologically. Finally, a multicenter, double-blind, parallel-group clinical trial was
conducted comparing the lotion with the vehicle in patients with plantar, interdigital,
or vesicular tinea pedis. Assessment was based on criteria that basically furnished
the base of the FDA guidance (see foregoing).
The efficacy of cyclopirox (ciclopirox olamine) lotion 1% was clearly manifest
in patients with interdigital tinea pedis. The plantar form, however, is difficult to
manage and usually requires a longer duration of therapy than was provided in the
trial.
c. Bacterial Infections. Topical and systemic antibiotic therapy is common
in dermatology. Apart from considerations on the development of antibiotic resistance
(200) or the development of allergic contact dermatitis (201), it is difficult to find a
rationale for a particular drug administration route in some diseases. In several com-
parative studies on antibiotic therapy of impetigo (Staphylococcus aureus or Strep-
tococcus pyogenes), the superiority of topical or systemic treatment regimens could
not be found (202,203). In a rare trial, in which topical or oral tetracycline was
allotted to acne patients at random, Schwanitz et al. (204) found no significant dif-
ferences between the two routes of therapy. Since 1982, when Eady and associates
(205) in an exhaustive review, felt that the question of topical versus oral acne
therapy remains open, no new information can be added to prove the superiority of
topical or systemic antibiotic treatment regimens.
4. Local Anesthesia
Topically induced anesthesia is of great value in clinical practice, particularly in
pediatric patients. The need for the development of an effective topical anesthetic
preparation prompted researchers to try various approaches.
The pharmaceutically unique eutectic mixture of the local anesthetics lidocaine
and prilocaine provides dermal anesthesia–analgesia following topical application.
The term eutectic refers to the phenomenon whereby the melting point of the two
anesthetics is lower after mixing than that of either anesthetic alone. When lidocaine
and prilocaine are mixed at 25⬚C, an oily liquid is formed that can be emulsified in
water (1:1). Thus, eutectic lidocaine–prilocaine cream is an emulsion, with the mix-
ture of local anesthetics as the dispersed phase and water as the continuous phase
(Broberg/Evers, 1981; EU patent no. 0,002,425). Whereas the cationic form of a
local anesthetic is responsible for blocking neurotransmission, it is the uncharged
(base) form that diffuses into the tissues after topical administration. Because skin
penetration is facilitated by the presence of an aqueous solvent, absorption of the
anesthetic mixture is enhanced by the high water content. Unlike other topical local
anesthetics, which are generally limited to use on mucosal membranes, eutectic li-
docaine–prilocaine cream achieves effective dermal anesthesia or analgesia follow-
ing application to intact skin under an occlusive dressing. The principal indications
in which eutectic lidocaine–prilocaine cream have been successfully studied are the
management of pain associated with venipuncture, intravenous cannulation, curettage
of molluscum contagiosum lesions, and split-skin graft harvesting (206).
The standardized pinprick test (183,207) has usually been used to determine
effectiveness of topical anesthetic preparations. Kushala and Zatz (208) successfully
applied a new method to assess the anesthetic activity of various topical lidocaine
formulations. A commercially available electrometric device to determine tooth pulp
vitality can also be used to stimulate the nerves within the skin. The device delivers
low-current electric stimulation in the form of bursts of pulses. Contact of the probe
with the skin initiates a series of low-voltage bursts; successive bursts are of incre-
mentally higher voltage (ⱖ 300 V). The voltage is transposed into arbitrary units
from 0 to 80. The digital instrument reading is proportional to voltage. To apply this
technique to humans, the subjects have to be instructed in the testing technique.
Briefly, under supervision, the subject places the device against the skin of the fore-
arm. The subject removes the probe from contact with the skin when a sensation,
usually described as a mild buzz, is first felt. When the subject is comfortable with
handling the device, the consistency of their response is evaluated. Subjects exhib-
iting inconsistent or nonreproducible results are rejected. This technique may be
considered to be a promising alternative to the conventional pinprick-test (Fig. 11).
The comparison of the mean effect values, determined electrometrically,
yielded one group of inactive formulations (two placebos) and two overlapping
groups of active lidocaine formulations. Statistical analyses indicate that all the li-
Figure 11 Mean effect for one lidocaine cream, five lidocaine suspension formulations,
and two placebo formulations: C14-T, tetradecyltrimethylammonium bromide; SLS, sodium
lauryl sulfate; C18-T, octadecyltrimethylammonium chloride. Bars indicate SEM. (From Ref.
209.)
docaine-containing suspensions performed at least as well as the plain formulation,

which contained no surfactant. Changing the pH also had no significant effect on
the in vivo performance. The pharmacodynamic data were compared with in vitro
permeation data using human cadaver skin. The rank order for the suspension for-
mulations was the same as that for steady-state permeation in the in vitro experi-
ments. Application of the cream formulation produced greater effect in vivo than
was anticipated from the in vitro flux values (209).
Pain associated with herpes zoster and postherpatic neuralgia has been difficult
to manage. None of the many suggested medical, anesthestic, physical, or surgical
methods have been widely accepted. Topical applications of analgesic and neurolep-
tic agents formulated in a variety of different vehicles, have recently been reported
with greater frequency. Their efficacy, usually determined with a pain questionnaire,
has, however, varied widely (210).
IV. FACTORS INFLUENCING THE RATE AND EXTENT OF DRUG

ABSORPTION FROM TOPICAL PRODUCTS
Factors from the following three areas may potentially influence the rate and extent
of drug absorption from topical products: (a) the skin conditions, (b) the composition
of the topical vehicle, and (c) the application procedures or application conditions,
because these affect the interaction of the topical vehicle with the skin. These have
been reviewed to identify the primary factors responsible for variation in bioavail-
ability, bioinequivalence, and therapeutic inequivalence of topical products.
A. Effect of Skin Conditions
1. Skin Microflora
The skin surface supports a microbial population that has the potential to carry out
biotransformations of topically applied therapeutic agents. At present, there appears
little in vivo evidence to suggest that the microbial transformation of compounds
applied topically for percutaneous absorption have any greater significance than the
metabolic action of the skin itself (211–215).
2. Skin pH
The pH of topical vehicles affects the extent of dissociation of ionizable drug mol-
ecules and, thus, their thermodynamic activity, partitioning, and skin penetration
(216–218). Normal human skin has a surface pH of 4–6. A pH gradient exists within
the skin (83). The influence of the pH of the skin surface or in the skin on penetration
of an ionizable xenobiotic has not been directly studied. The use of pH-sensitive (or
temperature-sensitive) delivery systems (e.g., liposomes) has been suggested and may
also offer new therapeutic approaches in dermatology (219).
3. Skin Surface Lipids

The skin possesses sebaceous glands that secrete a mixture of lipids that form an
irregular 0.4- to 4-␮m–thick film on the skin surface (220–223). Recently, Cheng
et al. (224) investigated the role of skin surface lipids on lidocaine permeation from
pressure-sensitive adhesive (PSA) tapes and found that the skin lipids can dissolve
solid drug in the PSA tape to decrease its thermodynamic activity and thus drug
permeation.
4. Temperature
Temperature changes on or in the skin are always accompanied by other physiolog-
ical reactions, such as increased blood flow, or increased moisture content of the
horny layer. These factors themselves can contribute to higher percutaneous absorp-
tion. Furthermore, increase in temperature increases drug solubility in both vehicle
and stratum corneum and increases diffusivity, both of which will lead to a further
increase in percutaneous absorption (225,226). In clinical practice, temperature may
be of minor importance, but in occupational dermatology, temperature may be an
important stimulus (227,228).
5. Blood Flow
The transepidermal resorption process feeding into the cutaneous microcirculation
brings compounds into the underlying tissues or the systemic circulation. Early stud-
ies with thyroxine and steroids (229,230) showed that the cutaneous vasculature
apparently does not function as an infinite sink that transports all topically applied
drugs to the systemic circulation. Part of the drug accumulates in the superficial
dermis and diffuses into deeper parts of the skin (231–233). Under these conditions,
cutaneous blood flow can modify the concentration levels and the accumulation of
substances in the dermis or deeper parts of the skin.
Vasoactive drugs (topical and systemic), or blood flow decrease obtained by
ligation, can modulate the transdermal delivery of drugs (234–240).
6. Diseased and Damaged Skin

Despite the limited data available, the belief that damaged skin enables chemicals
to enter the body unhindered is certainly false. The data suggest that differences in
absorption will exist for different drugs and for different disease conditions (16,241–
250). In patients with erythroderma—a most severe form of cutaneous inflammation
—percutaneous absorption of hydrocortisone ranged from 4 to 19%, compared with
approximately 1% in normal skin (24). In psoriasis, definite formulation effects were
observed (251). With treatment, a skin condition usually changes and thus affects
drug delivery into the skin (252). Damage to the skin by tape-stripping of the stratum
corneum leads to an approximately twofold increase in the permeation of hydrocor-
tisone (103).
7. Influence of Anatomical Site
The influence of skin site on the extent of absorption of compounds has been reported
(139,253–257). For example, the classic study of Feldmann (139) shows anatomical
variation in the percutaneous absorption of hydrocortisone of over two orders of
magnitude (foot arch 0.17%, forearm 1.04%, jaw angle 12.25%, scrotum 36.2%
absorption). These studies are based mainly on the determination of a drug concen-
tration in an accessible body fluid, usually urine. Pharmacodynamic studies, however,
reveal an unequivocal picture. Hansen et al. (258) examined the response of vol-
unteers to the application of nitroglycerin ointment at three body sites. Of the three
sites studied (midforehead, left lower anterior chest, and medial left ankle), the fore-
head site uniformly produced the greatest response in terms of magnitude and time
of onset of changes in systolic blood pressure and the incidence of subjective adverse
effects. Moe and Amstrong (259) studied the influence of skin site on bioavailability
of nitroglycerin using pharmacokinetic and pharmacodynamic parameters. Nitro-
glycerin ointment was administered to patients, who had severe congestive heart
failure, in a randomized fashion to three skin sites: arm, chest, or thigh. Hemody-
namic parameters and the arterial nitroglycerin concentrations were measured. No
significant differences among the sites were observed in mean arterial pressure, left
ventricular-filling pressure, right arterial pressure, or nitroglycerin concentration. This
information confirms earlier pharmacokinetic data from Noonan and Wester (260),
who studied percutaneous absorption of [14C]⫹ nitroglycerin through a uniform area
of the chest, arm, inner thigh, and postauricular region of the rhesus monkey. No
difference in percentage absorption was observed among these skin sites. However,
compared with many topical agents, nitroglycerin is well absorbed, and this will tend
to reduce differences in extent of absorption between skin sites. A recent report on
the percutaneous absorption of ketoprofen in healthy volunteers showed that absorp-
tion was similar when applied to either the back or the arm, but was lower when
applied to the knee (261).
In the management of psoriasis with topical glucocorticosteroids, any given
steroid formulation commonly results in a distinct grading of response of the pso-
riasis lesions, based on region. The dorsa of the feet and hands, elbows, and knees
respond poorly, but better than the palms and soles, which rarely respond. Lesions
on the upper thighs respond better than lesions on the lower legs; lesions on the
chest better than those on the upper arms; and those on the face best of all (262).
Thus, some of the quantitative measurements of permeation through various regions

correlate well with the clinical observations of regional response to topical gluco-
corticosteroids. Nevertheless, other explanations may be possible. The number of
‘‘receptors’’ or sensitivity of receptors at the target site may be the main determinant
of the pharmacodynamic response.
Overall, variation between and within anatomical sites is likely to be the most
important ‘‘skin condition’’ factor influencing extent of absorption. Bioequivalence
studies are almost always conducted on a single skin site, often the forearm. If we
assume bioequivalence is established, the large variation in extent of absorption
between the skin site used in the bioequivalence model and the clinical setting gives
rise to potential for therapeutic inequivalance, unless the change in the extent of
absorption with the skin site is the same for all products.
8. Influence of Appendages
Evidence to support or refute the hypothesis that appendageal pathways are signifi-
cant can be identified. The arguments against—which are based on the surface area
—are persuasively illustrated by historic examination and by various numerical cal-
culations. On the other hand, studies using (a) autoradiography, (b) an animal model
that totally lacks appendageal shunts, and (c) an alternative data analytical approach,
all point to the significant participation of the appendages in the transport (particu-
larly, the non–steady-state portion) of molecules across the skin. The importance of
these observations and the clinical relevance (e.g., multiple application) remains
vague. To truly ascertain the significance of follicular delivery, appropriate models
and quantitative methods must still be developed (220,263–290). Others have made
the case for direct drug targeting of the pilosebaceous glands and have developed
special formulations to optimize this delivery route (269,291).
9. Effect of Skin Metabolism
Far from being a passive membrane for the passage of drugs, the skin has significant
metabolic activity. As pointed out in several reviews, metabolic activity spans a broad
range of oxidative, reductive, hydrolytic, and conjugative reactions (292,293), mak-
ing the skin a source of extrahepatic metabolism of many xenobiotics and topically
applied drugs. The transport and metabolism of drugs in the skin places two kinetic
events in competition. Results of experimental (294) and theoretical (295–298) in-
vestigations have suggested that diffusional and metabolic processes in the skin are
intimately related, with one often having a profound effect on the other (15,299).
As the physicochemical properties of drugs themselves are not always optimal
for their delivery into or through skin, topical administration is not always possible.
An elegant approach for overcoming these problems is to make a transient derivative
—a prodrug of the drug—which imparts to the drug the desired transient change in
its physicochemical properties (300). Once the prodrug has overcome the skin barrier,
it is metabolized to the parent compound of established safety and efficacy.
Although the prodrug concept involves the chemical modification of a known
pharmacologically active compound into an inactive or less active bioreversible form
that becomes metabolically active within the skin layers, the softdrug concept in-
volves metabolic deactivation of an active compound within the viable layers or in
the systemic circulation (301).
10. Effect of Age

Despite much research into the mechanism of cutaneous aging and the identification
of significant age-associated biological and biophysical changes within the skin, the
question, ‘‘how does aging affect percutaneous absorption in vivo?’’ remains partly
unanswered. Some data suggest that the diminished surface lipid content of old
‘‘skin’’ implies a diminished dissolution medium for compounds administered topi-
cally. It is reasonable to speculate that this physiological change most severely affects
those permeants for which lipid solubility is low (302). Biological effect is generally
decreased in the aged individual (303); therefore, pharmacodynamic parameters, sug-
gesting reduced effect or penetration, have to be used with care. Skin permeability
is greater in premature (or newborn) infants (304). Although there is certainly a
potential for age-related effects to result in bioinequivalence in the premature or
newborn for two formulations that appear equivalent in healthy adults, there is no
direct evidence of any such occurrences (305).
11. Effect of Race
Racial differences in physicochemical properties of the skin, in in vivo percutaneous
absorption and to various chemical stimuli have been reported (306–309); however,
the data is not unequivocal. Except for one study (310), that indicates racial differ-
ences in nitroglycerin absorption after transdermal application, one may state within
the limitations of the experiments that, if any differences exist, they are below the
limit of detection of the currently established methods. Currently, no information is
available on the responsiveness of a specific disease to topical therapy in different
races.
12. Difference Between Subjects
Large differences exist in percutaneous absorption between subjects (140) as a result
of drug–vehicle–skin interaction. This can result in subject–formulation interactions,
such that the bioavailability of a test product relative to that of a reference product
may be consistently different between individuals (311). The resulting issues of
switchability and prescribability have been discussed elsewhere (29,312).
B. The Vehicle Effect

A vehicle can be defined either by its structural matrix (e.g., ointment, cream, gel,
liposome, or other) or by its excipients (e.g., petrolatum, propylene glycol, isopro-
pylmyristate, water, and the like). When discussing vehicle effects it is important to
clearly distinguish between these two terms; for example, the data from Table 2
compares the potency of various 0.5% betamethasone dipropionate products (313).
It is uncertain which contribution of matrix’ or excipient’s effects are responsible for
the differences between formulations, although the excipient’s effects will usually
overide.
1. Structural Matrix
The question is often asked whether a cream, a gel, an ointment, or a liposome is
better to deliver a drug to the skin, to promote the drug’s penetration and therapeutic
effect. Certainly, cosmetic aspects of the delivery system, a result of the structural
matrix, may have an influence on compliance and, they therefore, are of clinical
Table 2 Comparative Potency of Various 0.5%

Betamethasone Dipropionate Products
Potency group
Drug Vehicle type (U.S.)
Diprolene Ointment I
Diprolene AF Optimized cream I
Diprosone Ointment II
Diprosone Cream III
Diprosone Lotion V
Source: Ref. 313.
relevance. Moreover, ointments or heavy creams may occlude the skin and increase
drug penetration. However drug delivery to the skin is controlled by the vehicle
excipients, as these effect partitioning into and diffusion through the stratum cor-
neum, as described later. It remains speculative whether one can assign particular
structural features of a vehicle to a specific effect. Furthermore, it is difficult to
formulate a vehicle with the same ingredients and different structural features (314–
316).
2. Formulation Excipients
The effects of formulation excipients on the rate and extent of drug absorption are
greater with topical drug delivery than with any other route of drug administration.
For example, comparing alternative formulations of the same drug, differences in
extent of penetration of 10- to 50-fold and higher have been reported (317–320). To
put this into perspective, 50–100% (up to onefold) differences in extent of absorption
by the oral route are rare (7–9).
The potential for large differences in the extent of absorption between topical
formulations is due to the complex interactions between the drug, the vehicle, and
the skin, which control partitioning into and diffusion through the stratum corneum
barrier. However, it is the low extent of drug absorption under ‘‘normal’’ conditions,
often in the range of 1–5% of dose applied, that allows the potential of these inter-
actions to be realized as large differences in extent of absorption. Thus, one may
consider the low extent of absorption—low absolute bioavailability—as being a
significant factor leading to the potential for topical bioinequivalence and therapeutic
inequivalence. The case for the use of rational dosing—reduced drug concentration
in topical formulations—is developed in this chapter.
Literature describing the effect of formulation on the rate and extent of per-
cutaneous absorption is enormous. Here, we will only try to explain, on the basis of
a simple physical model, how formulation may influence the various parameters that
control the percutaneous absorption process and, thereby, affect bioavailability, bio-
equivalence, therapeutic equivalence, and potentially result in bioinequivalence and
therapeutic inequivalence. For this latter point, which will be reviewed in more detail
in Sec. IV, we will need to consider how formulations that have been declared as
bioequivalent (i.e., in a volunteer study), may interact differently in the true clinical
setting, for which variation in skin conditions, particularly anatomical site, and in
dosage requirements of individual patients may exist.
In the following discussion, only transepidermal transport (i.e., through the

stratum corneum) as distinct from transappendageal transport, will be considered.
Relatively little is known about the effects of the vehicle on absorption by the ap-
pendages (see Sec. IV.A.8), although some of the factors discussed are also likely
to be relevant to this route (321).
a. The Higuchi Physical Model. Figure 12 describes percutaneous absorption
as a passive diffusional process through parallel membranes (321). By passive, it is
meant that the absorption process is not energetically coupled to any biological
process. Higuchi (321) considered two main examples of this general model: one,
in which the rate-controlling step is in transport across the skin; the second, in which
the rate-controlling step is in release from the topically applied formulation, as dis-
cussed in Sec. III.A.1. Figure 12 (right side) describes the concentration profile of
the drug across the skin for the common case when the rate-controlling step is in
transport across the skin and the barrier is in the stratum corneum.
The various cases of the Higuchi model, and particularly the subcases of case
1, are shown schematically in Fig. 13. Although it is useful to understand the logic
of the entire Higuchi model, only the most common case (1.1.), when the barrier to
absorption is in the stratum corneum, will be considered in detail. Case 2, in which
the rate-controlling step is in the formulation, is relevant to only topical therapy
when the skin barrier function is severely damaged; for example, severe burns. Case
1.2., for which the rate-limiting step is within the viable epidermis–dermis, applies
only when the permeant is extremely hydrophobic (322,323).
b. Drug Structure–Permeation Relation. The stratum corneum barrier has
been modeled as a bricks-and-mortar structure (324) of corneocyte bricks held to-
Figure 12 Schematic of percutaneous absorption as a diffusional process across parallel

membranes. The right-hand side shows changes in drug concentration when the stratum cor-
neum is the rate-limiting barrier. The driving force for diffusion is the drug concentration in
the outer layer of the stratum corneum (Cv ⫻ Pc).
Bioavailability and Bioequivalence
Figure 13 The hierarchical structure of the Higuchi physical model of percutaneous absorption. (From Ref. 321.)
439
gether with an extracellular lipid mortar. Most agents are believed to permeate the
stratum corneum by the extracellular lipid route. The intrinsic permeation of a drug
through the stratum corneum lipids is dependent on several physicochemical param-
eters (325), and it is useful to classify these into those factors that affect solubility
in the stratum corneum lipids, such as partition coefficient and drug-melting point,
and those that affect diffusivity through the barrier, such as molecular volume and
hydrogen-bonding potential (326). Small, low-melting–point, hydrophobic com-
pounds, such as nicotinates, salicylates, and nitroglycerin are, thus, relatively well
absorbed; accordingly, they are less likely to suffer from problems of bioavailability
and bioequivalence. Large, crystalline compounds, such as corticosteroids and reti-
noic acid derivatives, even though such structures are generally hydrophobic, are
relatively poorly absorbed to the extent of 1–5% of dose applied; consequently,
problems of bioavailabilty and bioequivalence may, and do, occur (327). Large, crys-
talline polar compounds (e.g., peptides) are predicted to be poorly absorbed and
prone to problems of bioavailability.
c. Interactions Between Drug, Vehicle, and Skin. Equation (3) describes

those factors that control percutaneous absorption, based on case 1.1. (see Fig. 13)
under conditions of nondepletion of donor compartment and sink conditions in the
receptor compartment.
␦ Q C v ⫻ Pc ⫻ D
= (3)
␦t L
where, ␦ Q /␦ t is the steady-state rate of percutaneous absorption per unit area of skin;
C v , the drug concentration in solution in the vehicle; Pc , the partition coefficient of
the drug between the vehicle and the stratum corneum; D, the diffusion coefficient
of the drug in the stratum corneum; and L , the apparent thickness of the stratum
corneum. For convenience, these interactions can be considered to fall into two areas,
drug–vehicle (thermodynamic) effects and vehicle–stratum corneum (penetration en-
hancer) effects. In Eq. (3) the term C v is an approximation for [C v (t = 0) ⫺ C v (t =
t)] that is used under conditions of nondepletion–infinite dose where C v (t0 ⫺ t )
approaches zero.
Thermodynamic effect. The driving force for diffusion is the thermodynamic
activity of the drug in the vehicle which, in turn, is related to C v and to the activity
coefficient of the drug in the vehicle. Thermodynamics being an illusive discipline,
it is best to consider C v and Pc (which is proportional to the reciprocal of drug
solubility in the vehicle) as being the factors that control drug–vehicle interactions.
The higher the value of the product of C v ⫻ Pc , the higher the thermodynamic
activity and the higher the drug permeation.
Effect of drug concentration in solution. From Eq. (3), in a single vehicle,
thus, under conditions of constant Pc and D, change in C v is predicted to produce a
corresponding linear change in permeation. For example, in Fig. 14, Davis and Had-
graft (67) showed a linear increase in hydrocortisone acetate flux across a Silastic
membrane as drug concentration in a single vehicle of propylene–water (56:44) was
increased up to saturation at 0.02%. Many similar results have been reported, in-
cluding the early classic studies of Poulsen et al. (328,329) and more recently
(65,318,330).
Figure 14 Transport of hydrocortisone acetate from a single propylene glycol–water (56:

44) vehicle across Silastic membrane. Demonstration of response to Cv . (From Ref. 67.)
Effect of vehicle partition coefficient. Relatively few studies have systemati-

cally investigated the effect of vehicle partition coefficient for a fixed C v , although
Pc is equally as important as C v in determining the driving force for diffusion, the
drug concentration in the outer layer of the stratum corneum. Saturated solubility in
the vehicle, and thus Pc , may vary over several orders of magnitude (331). Davis
and Hadgraft (67) showed a linear increase in hydrocortisone acetate flux across a
Silastic membrane as a partition coefficient (expressed as reciprocal of saturated
solubility) was increased for a fixed drug concentration of 0.02% (Fig. 15).
Flynn and Smith (65) found that, for a fixed concentration of p-aminoaceto-
phenone, transport across a Silastic membrane increased linearly with increase in
partition coefficient (reduction in propylene glycol fraction) up to the point at which
saturation of the vehicle occurred. A further increase in partition coefficient (by a
reduction in the propylene glycol fraction) resulted in formation of drug suspensions
that showed no further increase in transport. It may be possible, for a fixed C v , to
reduce the partition coefficient, but to still maintain a drug in solution, such that
supersaturated solutions are formed (67,317). Figure 16 shows that transport of
0.02% hydrocortisone acetate across a Silastic membrane, from saturated to eightfold
supersaturated solutions, is linearly dependent on the degree of saturation, or at a
fixed value of C v , dependent on the partition coefficient of the vehicle. In terms of
bioavailability and bioequivalence, for a fixed value of C v , differences in thermo-
dynamic activity (otherwise expressed as partition coefficient or solubility of the
drug in the vehicle) may result in large differences in the extent of absorption and
thus bioinequivalence.
Saturated solutions. Cskin is the driving force for the diffusional process of
percutaneous absorption and
Figure 15 Transport of hydrocortisone acetate, 0.02% w/w, from various propylene glycol–
water vehicles across Silastic membranes. Response to vehicle partition coefficient expressed
as reciprocal saturated solubility. (From Ref. 67.)
Figure 16 Transport of 0.02% w/w hydrocortisone acetate from supersaturated vehicles

across Silastic membrane. Response to degree of supersaturation. (From Ref. 67.)
Cskin = C v ⫻ Pc (4)
Under stable, equilibrium conditions, flux will be at a maximum when the outer
layer of the skin is saturated and, by definition, this will occur when the vehicle is
also saturated with solute. From Eq. (4), it is apparent that there is an inverse relation
between the saturated solubility of the solute in the vehicle and the partition coef-
ficient between that vehicle and the skin, such that for all saturated systems, their
product is a constant: that is, the saturated solubility of the solute in the stratum
corneum (328).
C v(saturated) ⫻ Pc = constant (5)
Equations (4) and (5) are the basis for the statement commonly found in the
literature that all saturated vehicles of the same permeant will give the same flux,
independently of concentration, provided that the vehicle components do not alter
the barrier function of the skin and that significant depletion of the permeant from
the vehicle does not occur.
Many authors have published experimental data to support this argument
(65,67,328–330,332,333). Figure 17 shows that transport of methylparaben through
Silastic membranes is the same for all saturated solutions despite a 95-fold difference
in drug concentration in solution.
Figure 18 shows data (333) on the vasoconstrictor response of betamethasone
benzoate from five vehicles of differing ratios of mineral oil:myglyol (85:15; 60:40;
40:60; 20:80; 10:90, respectively). Addition of myglyol (a triglyceride) increases
saturated solubility of betamethasone benzoate and decreases Pc . In Fig. 18, open
symbols represent subsaturated solutions and filled symbols the single, saturated
solution and suspension systems for each of the five vehicles. This data allows an
excellent summary of the importance of drug–vehicle interactions. First, in relation
to the potential for bioinequivalence, for a fixed C v , vasoconstriction is strongly
Figure 17 Transport of methyl paraben across a Silastic membrane from simple water–
glycol-saturated solutions. Transport is dependent on thermodynamic activity, otherwise ex-
pressed as Cv ⫻ Pc. (From Ref. 65.)
Figure 18 Relative vasoconstrictor response from subsaturated solutions (open symbols)

and saturated solution and suspension (filled symbols) formulations of betamethasone ben-
zoate in mineral oil–myglyol. As an example, at 100 mg/100 g, formulations 10:90 and
60:40 give vasconstriction scores of 103.4 (→1) and 94.7 (→2) units, respectively and would
be considered bioequivalent. However, 15 mg/100 g and 3 mg/100 g vasoconstriction scores
are different at, respectively, 100.5 (→3) versus 52.1 (→4) and 78.8 (→5) versus 12.7 (→6),
and at these lower strengths the formulations would be considered bioinequivalent. (From
Ref. 333.)
dependent on the vehicle (Pc ), as discussed previously. For example, any vertical
line (fixed C v ) drawn between the 1- and 10-unit doses shows a large variation in
vasoconstrictor response depending on the vehicle composition.
The response to increase in C v is complex, again depending on the vehicle. In
any single vehicle, an increase in C v results in an increase in vasoconstriction up to
saturation (this strongly suggests that the limit of vasoconstrictor response is con-
trolled by drug absorption and not saturation of the pharmacodynamic response in
these systems). In suspension, a further increase in drug concentration beyond the
saturated concentration does not lead to any further increase in penetration or va-
soconstrictor response in a single vehicle. Across all the vehicles, saturated solutions
(the first filled symbol in each series) give essentially the same response. Thus, a
dose–response to increase in concentration will not always occur, as has been pre-
viously reported (327). Pharmacokinetic and pharmacodynamic control of dose–
response is discussed further in Sec. V.
Bioequivalence is usually based on single-strength comparisons. Thus, assum-
ing bioequivalance at a single strength, a further issue is one of the potential for
bioequivalence—therapeutic inequivalence when topical products available in a
range of concentrations have different dose–response relations because of formula-
tion differences, as illustrated in Fig. 18.
Vehicle–stratum corneum (penetration enhancer) effects. In reality, many ve-

hicle components do interact with the stratum corneum, and some—penetration en-
hancers—are included in formulations specifically for this purpose. Figure 19 shows
data (334) on the in vitro permeation of water across epidermal membranes from a
variety of vehicles. To rule out water–vehicle thermodynamic factors, these authors
used half-saturated water solutions. The horizontal line shows the theoretical flux,
assuming no interaction of the vehicle with the stratum corneum. Although flux from
many vehicles is close to the line, others are above the line, indicating effects of the
vehicle on the stratum corneum barrier because of the study’s control of thermody-
namic activity.
Depending on the mechanism of interaction of the specific enhancer, the pa-
rameters Pc (through change in drug solubility in the stratum corneum), or Dc (by
fluidizing the stratum corneum lipid barrier), or both (335), are affected. The effect
of enhancers is most marked with polar compounds, and this is likely owing to the
effect of enhancers that reduce the hydrophobicity of stratum corneum lipids, thereby
increasing solubility of these compounds. For example, Yamane et al. (320) studied
the effects of pretreatment with terpenes and oleic acid–propylene glycol on the in
vitro permeation of fluorouracil across the epidermal membrane. As shown in Fig.
20, up to approximately 100 times enhancement of penetration was observed.
In terms of bioavailability and bioequivalence, at a fixed thermodynamic ac-
tivity, differences in the enhancer activity between vehicles may result in large dif-
Figure 19 Transport of water across human stratum corneum from a series of miscible
vehicles in which the water was at half saturation. The horizontal line shows the estimated
transport of water from water at half saturation. Vehicles with water transport significantly
above the line (> ⫻2) are considered to have interacted with the stratum corneum to reduce
the barrier function. (From Ref. 334.)
Figure 20 Effect of 12-h pretreatment on enhancement of transport of 5-fluorouracil across

epidermal membranes in vitro. (From Ref. 320.)
ferences in permeation and, thus, bioinequivalence. When differences in both drug–

vehicle interaction and enhancer activity exist between vehicles, large differences in
bioavailability of up to 100-fold can occur (317).
Summary of interactions between drug, vehicle, and skin. For a given drug
concentration, drug–vehicle thermodynamic effects and vehicle–skin enhancers ef-
fects, both separately and together, may cause large differences in bioavailability, up
to about 10–50-fold and beyond, and result in gross bioinequivalence. Although
these interactions between the drug, vehicle, and skin give rise to the potential for
large differences in permeation and bioavailability, it is the low absolute bioavaila-
bility, put another way, the relatively high concentrations, in many current topical
formulations that allow this potential to be realized.
C. Influence of Application Conditions

1. Rubbing
The data available are insufficient to confirm or deny the assumption that rubbing
has an effect on drug uptake into skin (336,337).
2. Occlusion
Occlusion prevents loss of moisture from the skin and from aqueous delivery vehicles
to the atmosphere. This trapped moisture, either endogenous or exogenous, exten-
sively hydrates the stratum corneum, causing it to swell appreciably (338). The in-
creased clinical efficacy of topical drugs caused by covering the site of application
with water-impermeable material is a long-known fact (339–341). The enhanced

pharmacodynamic effect of topical corticosteroids under occlusion was demonstrated
by vasoconstriction studies (342,343). Occlusion has also been reported to increase
the percutaneous absorption of various topically applied compounds. However, from
the data available, one may suppose that occlusion does not necessarily increase
percutaneous absorption of all substances, particularly the permeation of hydrophilic
compounds. Recent studies also indicate that the extent of percutaneous absorption
may depend on the occlusive system used (138,344,345). Occlusion, per se, can
cause local skin irritation that, again, may lead to increased drug absorption. In
addition, Vickers (92) suggested a definite correlation between skin occlusion and
the extent of the drug reservoir formation in the stratum corneum after percutaneous
absorption.
3. Loss of Vehicle
Despite its importance, the problem of loss of vehicle from an application site or the
translocation of an applied dose from treated to untreated sites is rarely discussed.
A single report presents a physiologically based mathematical model that de-
scribes removal processes on the surface of the skin, including the effects of washing
and desquamation, and rubbing off onto clothing. This model was applied to human
in vitro and in vivo percutaneous absorption data of an aqueous formulation of the
herbicide fluazifop-butyl (346).
Marples and Kligman (475) report that the spread of a drug from the site of
application to other areas of skin may give rise to false interpretations in paired
comparison tests of efficacy. For example, in infected dermatoses treated with topical
neomycin, the placebo-treated site improved clinically and there was almost as much
reduction in the Staphylococcus aureus population. These clinical observations were
visualized using tetracycline hydrochloride containing fluorescent ointments, creams,
lotions, and tinctures from various sites of applications (347). The topically applied
medications did not remain confined to sites of initial application. The anatomical
sites of initial application determined the pattern of transfer. Certain patterns of
transfer were characteristic among groups of patients and for specific patients. The
vehicle of a topical preparation greatly influenced the extent and sites of transfer.
Ointments and creams were transferred considerably more readily than lotions and
tinctures.
4. Metamorphosis (Change) of the Vehicle
In experimental and clinical situations, most dermatological vehicles (structural ma-
trix and excipients) undergo considerable changes after they are removed from the
primary container and applied to the skin. The initial structural matrix of the vehicle
may change owing to rubbing (thixotropic effects) or as evaporation of volatile in-
gredients increases viscosity. These events can offer benefits that enhance the per-
formance of vehicle(s) and drug(s). Rubbing certain vehicles onto the skin may have
emulsifying effects and evaporation of certain ingredients may produce the desired
cooling effect (348). Furthermore, evaporation from topical formulations is a required
quality of repellents (349–351).
Evaporation of solvents may, depending on the polarities of the volatile sol-
vent(s) and the drug, reduce or increase the solubility of the drug in the residual
phase and, thereby, the drug thermodynamic activity and drug permeation (66).
Loss of volatile solvents, such as ethanol and isopropyl alcohol, because they
are better solvents than the residual phase, often results in an increase in thermo-
dynamic activity until saturation is achieved, which normally sets the maximum
thermodynamic activity. At this time, the drug is anticipated to precipitate, where-
upon its flux should remain constant as set by the driving force of the saturated
solution. However, the drug may supersaturate and flux levels significantly greater
those set by a saturated solution may be obtained (352,353). Clearly, in such for-
mulations, there are large differences in drug penetraton between initial and residual
states and thus also under occluded and open conditions. Especially for these types
of vehicles (alcoholic gels and lotions), biopharmaceutical studies in volunteers
should not be conducted under occlusion unless this is also used in clinical practice.
Furthermore, uptake of water from the skin, brought about by occlusion of micro-
emulsion vehicles, may result in supersaturation (354).
Cheng et al. (224) have shown that skin-surface lipids may mix with vehicles
to change drug solubility at the vehicle–skin surface–lipid interface and thus influ-
ence permeation.
Because of the complexity of theoretical models and experimental setups, only
limited data is available on the metamorphosis of the vehicle (98,355–358).
5. Factors Influencing Dose
In most medical and toxicological investigations, the dose administered is precisely
defined. Unfortunately, this is not always possible when a chemical is applied to the
skin. Because the dose is a reference value and thus of utmost importance in bio-
availability and bioequivalency considerations, the issue of dose is thoroughly dis-
cussed. Dermal exposures (amount of drug absorbed) are defined in terms of dose
(amount of drug applied), surface area, time of exposure, and frequency of
application.
a. Dose Prescribed. The dose prescribed in a clinical sense includes amount
(e.g., number or quantity [weight]) of the dosage form, the strength, and an instruc-
tion on the application modalities. The quantities prescribed are usually governed by
the total surface areas to be treated, the application schedule, and the likely duration
of treatment. The total dose prescribed includes a special consideration relative to
possible toxicity (e.g., local or systemic effects of topical steroids).
Textbooks and guidelines in clinical dermatology give various recommenda-
tions (359–365). Polano (359) recommends 10 g of cream or ointment per appli-
cation per day as the minimum amount feasible for whole-body application (⬃0.6
mg/cm2 day⫺1). Schlagel et al. (360) found 12 g/day (⬃0.7 mg/cm2 day⫺1) the min-
imum amount necessary, and ‘‘liberal’’ applications of emollients may entail using
more than 100 g/day (⬃6 mg/cm2 day⫺1). Griffiths et al. (361) (Table 3) give the
most practical information by presenting recommendations for a 7-day treatment
period.
b. Dose Applied.
Dose applied in clinical situations. The dose applied in clinical situations
reflects the compliance of the patients with the dose prescribed or recommended.
Lynfield et al. (366) determined the amount of dosage form that is applied in ‘‘clin-
ical’’ practice by both ‘‘patient’’ and by health care professionals according to the
following instruction: ‘‘Put this on thinly, covering. . . .’’ (Table 4). The data clearly
Table 3 Dosing Recommendations for a 7-Day Treatment
Site To use sparingly To use liberally Lotions
Whole-body 100 g 250–500 g 500 mL

(⬃0.8 mg/cm2, day) (⬃2–4 mg/cm2, day) (⬃4 mg/cm2, day)
Localized disease 15–30 g 50–100 g 25–100 mL
Source: Ref. 361.
reflect the variation that is present in clinical practice. This table’s data confirm
results of others (367).
The sun protection factor (SPF) is the ratio that estimates the protective efficacy
of a sunscreen against sunburn. The generally accepted methods that are used to
determine the SPF of a sunscreen require that the amount of formulation applied is
1.5 mg/cm2 (DIN-Norm) or 2 mg/cm2 or 2 ␮L/cm2 (FDA), forming a homogeneous
film of a defined thickness (20 ␮m) (368). The applied thickness of a sunscreen is
important for the degree of photoprotection (Beer’s law). In a field test, Bech–
Thomsen and Wulff (369) found that the amount of the applied sunscreen was on
average 0.5 mg/cm2. A sunbather’s application of sunscreen is, therefore, probably
inadequate to obtain the sun protection factor assigned to the preparation. Previously,
in 1985, Stenberg’s data indicated that the sun protection factor under ad libitum
conditions was only 50% of what would be achieved using 2 mg/cm2 (476).
Another important aspect of applying a semisolid formulation on to the skin
has been addressed (370). In this report recent observations were noted that although
the frequency and amount of sunscreen may be adequate, the application technique
may be inadequate. The investigators observed that the pattern of coverage was often
incomplete and was dependant on the region treated (Fig. 21).
Dose applied in experimental situations. The dose applied in experimental in
vitro or in vivo percutaneous absorption studies may be expressed as mass per unit
area of the neat substance or vehicle. According to the recommendations of the
Topical Therapeutic Products Workshop (March 26–28, 1990; Crystal City, VA) 1–
3 mg or ␮L of formulation per square centimeter should be applied to the skin,
usually with inunction, corresponding to films of 10–30 ␮m in thickness (film thick-
Table 4 Amount of Dosage Form Applied May Be Influenced by Formulation

(Consistency), the Dispenser, and Person Applying the Dosage Form
Weight
Formulation Dispenser (g) Application g/m2 mg/cm2
O/W cream Tube 30 Self 7.1 ⫾ 3.1 >0.7

O/W cream Jar 450 Self 17.0 ⫾ 17.4 >1.7
O/W cream Jar 450 Nurse 9.1 ⫾ 3.2 >0.9
Ointment Tube 30 Self 13.6 ⫾ 8.9 >1.4
Liquid Squeeze bottle 210 Self 9.4 ⫾ 5.5 >0.9
Lotion (liquid) Squeeze bottle 210 Self 12.8 ⫾ 12.7 >1.3
Source: Ref. 366.

Figure 21 Sunscreen application pattern: anatomical regions: 1, hairline; 2, forehead; 3,

ears; 4, periorbital; 5, nose; 6, cheeks; 7, nasolabial; and 8, perioral. Density of shading
indicates coverage with sunscreen: dark (e.g., 8), good coverage; light gray (e.g., 1), incom-
plete coverage. (From Ref. 370.)
ness is an estimated value calculated from the dose volume and the estimated body
surface). This guidance reflects usage in the clinical situation. The skin should be
left open to the atmospheric conditions if this mimics the clinical use situation (371).
The amount of dosage form applied in experimental human in vivo studies to
investigate percutaneous absorption are in general considerably higher than the
amount recommended by Shah et al. (371) and Griffiths (361) or for that found by
Lynfield and Schechter (366) (Table 5).
Table 5 Examples for Dose Applied in Human

In Vivo Studies
Formulation Dose (vehicle)/cm2 Ref.
Solution 6.6 ␮L/cm2 372

Solution 10 ␮L/cm2 142
Ointment 5 mg/cm2 373
Gel, cream, ointment 8.9 (2.2) mg/cm2 374
Cream 5 mg/cm2 375
c. Dosing Technique. Historically, the most frequently used in vitro method

for studying absorption is the so-called infinite-dose technique. The skin is mounted
as a barrier between two well-stirred, fluid-filled chambers. The compound under
investigation is added to the solution on one side (donor) of the (skin) membrane,
and absorption is assessed by serially sampling and assaying the concentration in the
bathing solution (receptor) on the other side. Generally, the amount of solute that
permeates during the course of an experiment is small relative to the total amount
available, and there is no appreciable reduction in solute concentration in the donor
solution. In effect, it appears to be infinite.
Several objections can be raised to the use of the infinite-dose technique as a
predictive model for living skin. The most obvious, and certainly that of greatest
significance, is the use of fully hydrated skin; both sides of the (skin) membrane are
bathed by a donor and receptor solution. Furthermore, changes in composition due
to loss of volatile components may not occur so readily under conditions of infinite
dose. Equally blatant is that under clinical or ‘‘use’’ conditions, the amounts of
material applied are of the order of a few milligrams (of vehicle) per square centi-
meter skin.
To avoid the shortcomings of the previously described approach, the so-called
finite-dose technique was introduced (376–378). The (skin) membrane is mounted
in specially constructed diffusion chambers. The dermis is bathed by thermostatically
controlled receptor medium that can be replaced at various time intervals or contin-
uously. A few milligrams of test vehicle are applied to the skin and can be throughly
rubbed into it. The amount of applied material can be determined exactly. In contrast
with the infinite-dose technique, as the absorption proceeds, the concentration of
drug on the surface of the skin in a finite-dose experiment is depleted and flux falls
during the experiment.
It is generally assumed that the finite-dose technique realistically mimics the
in vivo situation. There are several publications addressing the finite-dose situation
on topical therapy (64,296,379–384). The finite-dose situation differs from the in-
finite-dose situation (e.g., most transdermal drug-delivery devices) in that transient
effects (non–steady-state ones) may be much more important and, in fact, may dom-
inate the bioavailability issues.
Despite that in vitro percutaneous absorption data, determined by the finite-
dose technique (377), suggest that the in vitro experiments readily mimic in vivo
percutaneous absorption as determined by the Feldmann–Maibach protocol (81,111),
there is an observation that may raise new questions.
As documented by many authors (103,111,138–142,245,385–391), the abso-
lute bioavailability of topical corticosteroids and many other topical compounds,
including retinoids, vitamin D derivatives, antifungals, antiacne, and antivirals, is
generally in the range of a few percent of the dose applied. Provided that the ex-
perimental conditions permit, a mass balance following termination of the experiment
[a quality control step that should be made with all experiments (138,392)], one
usually observes that the dose recovered on the surface is in the range of 60–95%
of the dose applied. Given this observation, one may argue that, under these ex-
perimental conditions, the delivery system (vehicle) is far from being exhausted
(depleted) and even under clinical, hence, less-controlled conditions, the dose
(amount of drug) that can be recovered after an adequate time interval is high.
Therefore, one may conclude that both in experimental and clinical situations one is
actually closer to an infinite-dose situation than to that of a finite-dose. This obser-

vation asks for a revision of older statements or a refined definition of finite or
infinite-dose situations. To summarize, there are three dosing technique situations
(Table 6), with the clinical situation for most drug being in category 2. For the few,
well-absorbed molecules, such as nitroglycerin, phenols, and salicylate and nicotinate
esters, the clinical situation is in category 3.
The infinite-dose of most compounds in clinical usage is another way of stating
the problem of low absolute bioavailability, or put another way, the use of excessively
high drug concentrations in most topical formulations. This brings us back to the
proposal for rational dosing—reduced drug concentration—in topical formulations
(see Sec. IV.B.2.c).
The use of infinite dose leads to the potential for bioinequivalence as discussed.
Furthermore, use of infinite doses complicates the efforts to meet regulatory require-
ments of showing dose–response relations (21,22,123,393). For further discussion
see Sec. V.
d. Variation of (a Single) Dose. A single dose can be varied using four gen-
eral approaches (123):
1. Concentration method: varying the drug concentration in the topical prod-

uct for a constant time on a constant surface area.
2. Film-thickness method: varying the surface area of application of the same
volume or weight and the same concentration test formulation for the same
duration of time.
3. Surface method: varying the surface area and volume of formulation ap-
plied with the same concentration formulation for the same duration of
time.
4. Duration method: varying the duration of application of the same concen-
tration formulation to the same surface area of the skin.
Table 6 Dosing Situations in Topical Drug Application
Dosing Bioavailability
Category technique Loading Comment (estimate)
1 Infinite dose High-loading/cm2 Two-chambered cells in Less than 5%

(>10 mg to in vitro experiments (typically 1–2%)
several 100 mg)
2 Infinite dose In-use loading/cm2 Poorly absorbed Less than 5%
(0.5–5 mg) molecules in clinical (typically 1–2%)
situations, (e.g.,
corticosteroids)
3 Finite dose In-use loading/cm2 Well-absorbed More than 10%
(0.5–5 mg) molecules in clinical (typically 25–50%)
situations (e.g.,
nitroglycerin)
Sources: Categories 2 and 3, Refs. 361, 366.

In the following paragraph the four methods are presented and referenced.
Whenever possible some recent examples are given of a (pharmaco)kinetic (K),
(pharmaco)dynamic (D), or clinical (C) investigation.
Concentration method
(K). For many compounds, in vitro and in vivo, the relation between dose
(concentration) and the dose absorbed is roughly linear over a broad
range (75,112,394–396, and others) provided that the compound is in
solution (354).
(K). Application of various concentrations of betamethasone dipropionate
(0.020, 0.040, 0.050, and 0.063%) for 6 h was associated with a statis-
tically significant (p < 0.05) linear increase in drug uptake into treated
human stratum corneum up to the 0.050% concentration. Increasing the
drug concentration applied beyond 0.050 up to 0.063%, however, did
not further increase drug uptake (p > 0.05) (123).
(D). The increasing drug concentration (betamethasone dipropionate: 0.020,
0.040, 0.050, and 0.063%) was not associated with a corresponding in-
crease in pharmacodynamic response over the 24-h observation period.
Plotting the maximal visual skin-blanching response, as a function of the
drug concentration applied, demonstrated that the maximal pharmaco-
dynamic response to this corticosteroid occurs at strengths substantially
lower than the strength marketed for clinical use (123).
(D). A similar disparity between drug concentration applied and pharmaco-
dynamic response was previously observed with the use of topical cor-
ticosteroid formulations (374,397–399). Several authors used the dilu-
tion method to create a dose–response reaction. This, however, has the
inherent danger of altering the physicochemical parameters of the drug
in the vehicle, particularly the thermodynamic activity of the drug, which
will influence drug partitioning and permeation (400).
(C). Double-blind studies directly comparing clinical effect of the different
concentrations of the same steroid in the same vehicle are rare. One
clinical study comparing 2.5% with 1.0% hydrocortisone cream in ec-
zema revealed no difference between the two concentrations (313).
Film-thickness method
(K). In carefully controlled in vitro percutaneous absorption experiments with
halcinonide, Walker et al. (401) observed, at clinically low levels of
application (i.e., less than 5 mg/cm2 of a cream), a dose-dependent low
rate of permeation. With applications of 5 mg/cm2 and higher (increasing
vehicle thickness) the rate of permeation appeared constant, although the
total amount permeating the membrane was dependent on the dose.
(D). Jackson et al. (402) studied five different marketed betamethasone val-
erate 0.1% creams and six different marketed triamcinolone acetonide
0.1% creams, in two groups of 12 subjects each. The subjects received
five 10-␮L portions of each cream spread over different skin surface
areas to yield doses of 20, 10, 2, 1, 0.6 ␮L/cm2. The area of application
was encompassed by a Plexiglas ring open to the air. The creams re-
mained in place for 6 h. Statistically significant differences were found
among creams containing both drugs, suggesting a lack of equivalence.

The authors report a diminution of vasoconstriction with increased area
of application. In addition, particularly for the triamcinolone acetonide
creams, the largest differences among the six creams were at the lowest
‘‘dilution’’ (largest application area). This suggests that testing of topicals
should not be confined to a single surface area or dilution.
(D). Pershing et al. (123) performed a similar experiment. Under unoccluded
conditions a constant volume of a 0.05% betamethasone diproprionate
cream formulation (10 mg) was applied to an increasing surface area of
the skin (0.5, 2.0, 3.8, and 5.1 cm2) yielding doses of 20, 5, 2.63, and
1.96 mg of cream per square centimeter, or 100, 25, 13, and 10 ␮g
betamethasone diproprionate per square centimeter forming a film thick-
ness of 200, 50, 26, and 20 ␮m, respectively. Visual skin blanching
responses and objective measurements using the Minolta-Chromameter
did not reveal significant differences (p > 0.05) between the various film
thicknesses.
(D). The independence of film thickness applied on the resulting maximal
pharmacodynamic response in the Pershing study (123) differs from
those in pharmacodynamic observations by Stoughton (22) with the same
drug in a different cream formulation. Stoughton found that there are
differences in vasoconstriction between doses of 1–4 mg/cm2 but not
those higher, to as much as 50 mg/cm2. This was true for three different
potent or midpotency topical glucocorticosteroids (Lidex cream, 0.05%;
Kenalog cream, 0.1%; Diprosone cream, 0.05%). Differences between
the Stoughton and the Pershing studies with the same drug likely reflect
the influence of one or more of the following parameters: (a) vehicle
composition (with propylene glycol in the Pershing ‘‘cream’’ versus no
propylene glycol in Stoughton ‘‘cream’’); (b) the efficiency with which
the doses were spread over the skin (370); (c) the duration of drug ap-
plication (6 vs. 16 h, respectively); (d) the time at which the pharma-
codynamic response was measured (8 vs. 18 h, respectively); and (e) the
method used to calculate the composite or average skin-blanching re-
sponse in the population studied.
(D). Barry and Woodford (403) showed that there was no difference in va-
soconstriction between 6 and 16 mg/cm2 when betamethasone valerate
formulation was used, whereas Magnus et al. (404) reported a propor-
tionate increase in vasoconstriction when 3.2–9.8 mg/cm2 (1.6–4.8
mg/49 mm2) formulation (betamethasone-17-valerate 0.1%) was applied,
but no greater activity at doses higher than 9.8 mg/cm2 (4.8 mg/49 mm2).
(C). There are no controlled studies available on the effect of film thickness
applied on the resulting clinical outcome. However, certain dosage
forms, such as pastes, cataplasms, and the like, are often thickly applied
to the skin.
Surface area method
(K). Unoccluded application of 10, 40, 80, or 100 ␮L of a 0.05% betametha-
sone diproprionate cream formulation to 0.5, 2.0, 3.8, or 5.1 cm2 on
human ventral forearm skin produced a constant film thickness of 200
␮m over successively larger surface areas. Increasing the skin surface

area from 0.5 to 5.1 cm2, to which a constant film thickness is applied,
decreased the mean amount of drug uptake per 1-cm2 area twofold, but
was not significantly different (p > 0.05). These data suggest that stratum
corneum uptake of betamethasone dipropionate is independent of the
surface area of the skin treated with constant film thickness (123).
(D). Increasing volumes over successively larger skin surface area maintain-
ing a constant film thickness, did not result in any significant change in
the corticosteroid pharmacodynamic response at various time points of
more than a 24-h period (123).
(C). The surface area of skin treated with a drug may require special consid-
eration relative to possible toxicity (e.g., local or systemic effects of
topical steroids). Frequency of application (see next section) and the
treated surface-area/body-weight relation are important factors (405).
Duration method
(K). Increasing the unoccluded duration of 10 mg of 0.05% betamethasone

dipropionate cream over the same skin surface area from 0.5 to 16 h, in
vivo in humans, resulted in a linear increase of drug uptake into the
human stratum corneum up to 2 h. Increasing the duration beyond 2–6
and 16 h did not significantly alter the drug concentration in the tape-
stripped skin sample (p ⱖ 0.05). The time required for maximal drug
uptake into human stratum corneum averaged close to 6 h (123).
(K). Increasing the unoccluded duration of 30 ␮L of an ethanolic solution of
hydrocortisone (200 nmol/cm2) over the same skin surface area, from
0.5 to 6 h, in vivo in rats, resulted in no increase of drug uptake into
the stratum corneum (126).
(K). Two oil-in-water creams and two gels, containing 10 or 5% ibuprofen,
respectively, were applied for 0.5, 1, and 2 h on excised human skin.
The application time had no influence on the epidermal drug concentra-
tion, whereas the two gel formulations produced concentrations approx-
imately twice those obtained with the emulsions (406).
(K). In a human in vivo study, 1 mL of a minoxidil solution was applied
twice daily over 150 cm2 of bald scalp to each subject for 6 days. Un-
absorbed drug was washed off the scalp after 1, 2, 4, and 11.5 h of
contact time in each of four treatments. The extent of minoxidil absorp-
tion, expressed as steady-state urinary excretion of unchanged minoxidil,
minoxidil glucuronide, or the sum of these, increased in a disproportion-
ate manner with increase in contact time of the drug on the scalp. Rel-
ative to the amount absorbed after a contact time of 11.5 h, absorption
was approximately 50% complete by 1 h and <75% complete by 4 h.
This suggests that minoxidil absorption from the vehicle into skin occurs
rapidly relative to diffusion through skin (372).
(D). Increasing the duration of drug application from 0.5 to 2, and to 6 h,
using the same concentration of betamethasone dipropionate, produced
similar visual skin-blanching response profiles over time, suggesting that
even more brief durations of application may be necessary with potent
corticosteroids to achieve a reliable dose–response relation (123). These

observations confirm data from Stoughton (393) and others (407).
(D). Stoughton and Wullich (393) (Table 7) assessed bioavailability over dif-
ferent time periods of exposure when 0.05% clobetasol propionate cream
(Temovate) (class I) was applied and left on for periods of 0.5, 1.0, 1.5,
or 16.0 h and subsequently washed off. Maximal responses were
achieved after 1.5 h of exposure, but there was no significant difference
in intensity of vasoconstriction between 1.0, 1.5, and 16.0 h of exposure
before washing the sites. Exposures to 0.05% clobetasol propionate
cream for 0.5 h were not significantly different from 16 h to that of
0.05% fluocinonide cream (Lidex) (class II), but exposures to 0.05%
clobetasol propionate cream for 1.0, 1.5, and 16.0 h, all resulted in sig-
nificant increases in vasoconstriction responses compared with fluo-
cinonide cream applied and left on for 16 h.
These observations (123,393), and results from other investigators
(71,374,407), suggest that the basic methodology of the Stoughton–
McKenzie vasoconstrictor assay needs additional validation.
(C). Dithranol is estabished as a highly effective treatment for psoriasis (e.g.,
Lassar’s paste) (408,409). The disadvantages of its use are the staining
of clothing, linen, and skin, irritation, and long application time. Schaefer
and co-workers (410–413) were able to show through their laboratory
and clinical work that short-duration therapy (e.g., 10 min; ‘‘minute’’-
therapy) with higher dithranol concentrations (1–3%) presented all the
advantages of dithranol and avoided most of its previous drawbacks. This
modification of treatment has made the use of dithranol at home gen-
erally possible. This type of therapy requires a high degree of com-
pliance.
(C). Jaeger (414) showed, in a double-blind, placebo-controlled study in pa-
tients with histopathologically verified psoriasis, that the application of
betamethasone dipropionate, 0.05% for 3–5 min, combined with occlu-
sion for 20 min was as effective as long-term corticoid treatment (415).
(C). A trial of three treatment schedules, consisting of 1% ␥-benzene hex-
achloride (GBH) lotion, applied head to toe, and left on the body for 2,
6, or 12–24 h was conducted on an island of more than 2000 persons,
approximately 70% of whom had clinical evidence of scabies. Exami-
nation at 1 month after therapy showed that both the 6-h and 12- to 24-
Table 7 Vasoconstrictor Assay in 30 Subjects
Clobetasole propionate Fluocinonide

(Temovate) (Lidex) Placebo
Name U.S. class I U.S. class II cream
Time 0.5 h 1.0 h 1.5 h 16 h 16 h 16 h

Total scores 43 58 68 67 30 4
Source: Ref. 393.

h–cure rate was high (96 and 98%). There was a significantly lower cure
rate in the 2-h group, in which only 82% were cured (416).
e. Multiple Doses. The dose can also be defined in terms of frequency of
application. Intermittent therapy can be one, two, or more exposures per day. Pro-
longation of the dosing interval up to several days is also possible and may help
avoid adverse drug effects in cases for whom long-term application is necessary.
Most percutaneous absorption studies have employed a single administration of the
compound under investigation; however, the relevant clinical (and toxicological) sit-
uations usually involve multiple contacts of drug (xenobiotic) with the same skin
site. Despite this obvious relevance of topical pharmacokinetics following multiple
topical application, there has been only limited investigation of the subject.
(K). The in vivo percutaneous absorption of hydrocortisone through the skin
of the ventral forearm of the rhesus monkey, quantified by measuring
14
C in aliquots of urine over 5 days, was no different when 13.3 ␮g/cm2
was applied as a single dose or when 13.3 ␮g/cm2 was applied three
times, totalling 40 ␮g/cm2. However, when 40 ␮g/cm2 was applied as a
single dose, absorption was substantially increased over that of 13.3 ␮g/
cm2 applied either once or three times. Additionally, when the skin was
washed between applications to remove previously applied material in
the three-application experiment, there was a statistically significant in-
crease over not washing the skin (417). The study of Wester et al. (417)
has recently been repeated in humans. When a single dose (13.3 ␮g/cm2)
was tripled (40 ␮g/cm2) the amount delivered through the skin increased
by nearly three times. Three serial doses, with and without soap-and-
water washing between the doses, increased percutaneous absorption re-
markably (142) (Fig. 22).
(K). With a similar experimental protocol (417), the short- and long-term (8
days) administration of hydrocortisone in acetone or in an emulsion oint-
ment base was studied in the rhesus monkey (418). Absorption signifi-
cantly increased during long-term administration, whether applied in an
acetone or in an emulsion ointment base. A placebo study in which only
an acetone vehicle was applied for a long period followed by [14C]hy-
drocortisone application, showed no enhanced absorption. It was sug-
gested that long-term application of hydrocortisone alters the skin barrier,
resulting in enhanced absorption. After short- and long-term (8 days)
administration of a methylprednisolone aceponate ointment (Advantan),
percutaneous absorption, assessed by cumulative urinary excretion of
14
C-labeled substances (<1% of dose applied), were not statistically dif-
ferent, suggesting that repeated application of the vehicle does not
change the barrier and reservoir functions of human skin (373,419).
Azone cream was topically dosed on the ventral forearm of humans for
21 consecutive days. On days 1, 8, and 15, the Azone cream contained
14
C-labeled Azone (375). The skin application site was washed with soap
and water after each 24-h dosing. Percutaneous absorption, determined
by urinary radioactivity excretion after repeated application (8 days),
nearly doubled (p < 0.002), but stayed the same after continued repeated
application (15 days). It is concluded, that repeated application of Azone
Figure 22 Percutaneous absorption of [14C]hydrocortisone through the ventral forearm of

rhesus monkeys and humans. (From Refs. 142,417.)
results in an initial self-absorption enhancement, and steady-state per-

cutaneous absorption is established after this change.
(K). Percutaneous absorption of diflucortolone-21-valerate in an ointment
base, with and without salicylic acid, was determined following a large-
area skin treatment (20 g ointment twice a day for 8 days) in two groups
of healthy volunteers. No differences were found either in percutaneous
absorption of diflucortolone-21-valerate, or in effects on the pituitary–
adrenal axis between the two treatment groups, suggesting that the ad-
dition of salicylic acid to the vehicle did not significantly alter the skin
barrier function during the 8-day treatment (420).
Despite some in-depth investigations a clear answer pertaining to
the effect of repeated application on percutaneous absorption is still un-
available. A variety of differences of experimental conditions make com-
parison of the data difficult. Because of the therapeutic and toxicological
relevance of the repeated-dose situation, resolution of the problem is
urgently needed. From a large amount of clinical and experimental evi-
dence, one may conclude that the vehicle influences the skin barrier and
skin reservoir function. It is obvious that pretreatment of the skin with
vehicle alone or reapplication of vehicle to a drug-treated area may also
influence drug absorption (421,422).
(K/C). The concentration of iododeoxyuridine (antiviral) found in the stratum
corneum of guinea pig skin by tape-stripping at different time points
after single and multiple topical doses of the drug, was studied. With
each dosing frequency, the cumulative amount of drug in the stratum

corneum correlated with the strength of the test formulation and with
efficacy in the animal. For each of the three formulations, increasing the
number of daily doses from one up to three led to progressive increases
in cumulative stratum corneum iododeoxyuridine levels and clinical ef-
ficacy. An increase in the number of daily applications to four had little
effect on drug efficacy and was associated with a plateau in stratum
corneum iododeoxyuridine levels (82).
(C). In a double-blind multicenter, placebo-controlled study in patients with
psoriasis and in patients with atopic dermatitis, Fredriksson et al. (423)
showed that the application of halcinonide 0.1% once daily was equally
as effective as the cream applied three times daily. However, the onset
of action was more rapid when the cream was applied three times daily.
With a similar study design Sudilovsky et al. (424), under the same
clinical conditions, showed that a once-daily regimen can be an effective
treatment; however, the three-times daily regimen was superior overall,
and the authors recommend this regimen as treatment of choice in severe
psoriasis.
f. Intermittent Dosing
(C). Actinic keratoses often occur on the sun-exposed skin of adults. A stan-
dard treatment for multiple actinic keratoses is topical 5-fluorouracil ap-
plied twice daily for 2–4 weeks until maximal inflammation is achieved.
Although effective, extreme local irritation makes it unacceptable to
many patients. Pearlman (425) showed that weekly ‘‘pulse’’ dosing of
topical 5-fluorouracil on mutiple facial actinic keratoses produced the
same benefit, with much less local irritation, than a conventional daily-
dosing schedule. In view of its efficacy, relative comfort, lower cost, and
simplicity of use, weekly pulse dosing may be preferable.
(C). Adverse drug effects have also been used to study the influence of dif-
ferent application schedules. Several authors found a correlation between
the therapeutic efficacy and the skin-thinning effect of topical cortico-
steroids (426–430). Potent corticosteroids produce marked skin-thinning,
whereas weak corticosteroids induce only a slight reduction of skin
thickness. To avoid steroid-induced dermal thinning and other adverse
effects, various strategies of therapy have been developed (431). Various
studies show that a topical treatment with glucocorticosteroids two or
three times daily has an efficacy similar to once-daily treatment
(423,424,432). Similarly, the skin-thinning effect of a clobetasol-17-pro-
pionate treatment (41 days) is identical when the drug is administered
once daily or at an interval of 72 h (433). This was also shown for
fluprednidene acetate, fluocinolone acetonide, and betamethasone-17,21,-
dipropionate (434,435). Prolongation of the dosing interval up to 8 days
diminished the skin-thinning effect, but the effect could not be totally
avoided (433,436). Results from various studies in psoriatic patients sug-
gest that intermittent pulse dosing with potent corticosteroids can offer
an efficacious method for extended maintenance therapy with fewer ad-
verse effects (437).
D. Conclusion
From this review of effects of skin conditions, vehicle effects and their interactions,
and consideration of dose, we can identify certain primary factors that influence
bioavailability, bioequivalence, and therapeutic equivalence of topical products.
1. Bioavailability: Under most clinical and experimental conditions, the use
of infinite (and hypertherapeutic) doses leads to low absolute bioavailabil-
ity, often in the range 1–5% of dose applied.
2. Bioequivalence and bioinequivalence: The low, absolute bioavailability of
topical products significantly increases the likelihood of bioinequivalence
when formulation differences lead to differences in drug partitioning into,
or diffusion through, the stratum corneum. From this, a rational approach
is to reduce the dose, which in the clinical situation, requires reduction of
drug concentration. Low-concentration topical products with absolute bio-
availabilities in the target range of 50–100%, yet bioequivalent to current
products, are suggested as a future standard. This is discussed in the final
section of this chapter.
3. Bioequivalence and therapeutic equivalence: Even when bioequivalence
has been established in a volunteer model (e.g., vasoconstrictor assay),
there is the potential for therapeutic inequivalence (11–13). One concern
is that the allowable limits for bioequivalence (e.g., ⫾20%) may lead to
therapeutic difference in compounds with a steep dose–response curve
(438). Also, there is the possibility of therapeutic inequivalence owing to
differences in absorption between the volunteer model and clinical situation
(13); for example, because of differences in absorption between anatomical
sites and on diseased skin, as discussed in Sec. III. Finally, there is the
complex issue of dose–response, when topical products are available in a
range of strengths. We have seen that bioequivalence at one strength is no
guarantee of bioequivalence or therapeutic equivalence at another. These
points are discussed further in Sec. V.
V. BIOAVAILABILITY, BIOEQUIVALENCE, AND

THERAPEUTIC EQUIVALENCE
This final section will briefly review standards of bioequivalence in currently mar-
keted topical dermatological products using glucocorticosteroids as the example.
Moreover, as a result of initiatives to introduce generic drug products, therapeutic
equivalence has been defined on the basis of bioequivalence plus pharmaceutical
equivalence (439). Thus, this section will also examine the proposition that bio-
equivalence predicts therapeutic equivalence.
To accomplish this, one needs to consider the dose–response relation and the
sources of variation that lead to inter- and intraindividual differences.
A. ‘‘Biological’’ Variation
Interindividual differences (‘‘biological’’ variation) exist in virtually all steps of drugs
interacting with the human body. Application of pharmacokinetic (absorption–dis-
tribution–metabolism–elimination; ADME) or pharmacodynamic (concentration–
effect) principles (440,441) to pharmaceutical drug development or to applied ther-

apeutics for patient care, requires an understanding of these processes and their
variation.
In general, pharmacokinetic principles relate specifically to changes of drug
concentration in the body relative to time. Specifically, there are two pharmacokinetic
‘‘dose’’ relations:
1. Between the dose applied and the dose absorbed: this relation is particu-
larly important in topical therapy, but is rarely considered. The two are
related by the extent of the dose absorbed: the absolute bioavailability.
2. Between the dose absorbed and the drug concentration at the target site:
the two are related by the pharmacokinetic factors that govern distribution:
metabolism and elimination.
The pharmacodynamic relationship is between the drug concentration at the
target site and some cellular response, the classic biological concentration–effect
relation.
An understanding of the concept and an appreciation of the merits of these
approaches can be formed by focusing on the basic tenet of clinical pharmacology:
a dose (concentration)–response relation. Irrespective of the shape (relation) of these
curves, there are four parameters: slope, potency, maximal effect, and variation, that
define such a curve (Fig. 23).
Although the slope, maximal effect, and potency parameters are commonly
considered an inherent property of a drug molecule, it is well recognized that all
three parameters (slope, maximal effect, and potency) are greatly influenced by the
processes governing the absorption, distribution, metabolism, and excretion of the
drug. Variation also exists in the pharmacodynamic process owing to differences in
biology. Levy (443,444) has provided strong evidence that variation in the phar-
macodynamic process is at least as important as that in pharmacokinetics. It is ac-
cepted that these pharmacokinetic and pharmacodynamic processes contribute sig-
nificantly toward the observed variablility or so-called biological variation commonly
seen in dose (concentration)–effect curves.
Figure 23 The log dose–effect relation with its four characterized variables. (From Ref.
442.)
Inter- and intraindividual variations in response occur because of both con-

trollable and uncontrollable factors. Inappropriate compliance (445) and placebo ef-
fects (446–449) are primary contributers to the latter. There are two controllable
factors:
1. Control of variation in extent of absorption from the vehicle. This relates
to bioavailability and bioequivalence in both the volunteer and clinical
setting.
2. Individualization of dosing to adjust for variation in distribution, metabo-
lism, excretion, and concentration–effect relationships (dose titration).
Sampling site or drug quantitation procedures may also be a source of varia-
bility. Furthermore, pharmacodynamic measurements are often difficult and prone to
errors, and the extent of variability may also vary from drug to drug.
B. Variation and Dose (Concentration)–Effect Relationship

Various mathematical models have been developed to describe the concentration–
effect relation, the most widely used being the sigmoid Emax model as shown in Eq.
(6).
Emax ⫻ C n
E= (6)
EC n50 ⫹ C n
where
E is the observed effect
Emax is the maximum effect
EC50 is the concentration (dose) that elicits 50% of the maximal effect
C is drug concentration (dose)
n is the Hill exponent that governs the slope effect
Figure 24 shows a schematic of the sigmoid Emax response for a drug having both
wanted and unwanted effects over the dose range of interest.
Levy (438) has described how the same change in dose, for example, a 50%
difference in bioavailability between two products, has a relatively greater effect on
response in the lower range of the response curve than at the top. For example, in
Fig. 24 the same change in dose D results in differences in wanted responses R1 and
R2. Change in dose D can also result in relatively greater changes in unwanted effects
than in wanted effects: R3 versus R2. Olson et al. (450) have illustrated how the
steepness (sigmoidicity) of the concentration–response curve is important in deter-
mining the effect of a given variation on the extent of absorption on drug response.
Selection of optimum dose and limits on criteria for bioequivalence close to this,
ideally, should be based on dose–effect relations. Such data now exists for relatively
few systemic drugs, but this should increase through a focus on development of
surrogate markers for clinical response (451).
C. Variation in Extent of Absorption

In Sec. IV, factors affecting the rate and extent of percutaneous absorption were
reviewed. Figure 25 shows examples of variation in the extent of absorption owing
to variation among subjects, skin site, skin condition, and formulation. The low
Figure 24 Schematic of sigmoidal Emax relation: effect of change in dose D on response

R depending on position on concentration–effect curve. (From Ref. 438.)
extent of absorption of many topical products, including corticosteroids and retinoids,

in the range of 1–5% of the dose applied, is a major contributory factor that allows
the potential for variation caused by these factors to be translated into bioinequiva-
lence and therapeutic inequivalence. Intuitively, if the extent of absorption is low,
there is a potential for variation and, thus, for bioinequivalence and therapeutic in-
equivalence. In the following Secs. C.1 and C.2 the consequences of variation in the
extent of absorption on bioequivalence and therapeutic equivalence are reviewed.
1. Bioavailability and Bioequivalence of Topical Glucocorticosteroids
After reviewing the theoretical potential for bioinequivalence in Sec. II, what are the
actual standards of bioequivalence in currently marketed topical dermatological prod-
ucts? A brief review of glucocorticosteroids is of interest, as the bioequivalence of
these compounds has been relatively widely studied. In these studies the major source
of variation in absorption is between formulations.
Early studies gave the first indications for the potential for bioinequivalence of
corticosteroids formulated into standard pharmacopeial cream and ointment bases
(452–454). Following the development of topical formulations of hydrocortisone,
triamcinolone acetonide, fluocinolone acetonide, and other synthetic corticosteroids,
Barry and Woodford conducted a series of studies throughout the mid-1970s to com-
pare the bioavailability of different formulations of the same corticosteroids (403,
455–458). In summary, these studies show many examples of 50–100% differences
in bioavailability of the same corticosteroid at the same strength in the same vehicle
type. In 1987, Stoughton published results on a series of proprietary topical gluco-
corticoids (459) showing significant differences in bioavailablility between commer-
cial cream and ointment formulations containing the same active substance at the
same concentration. In the worst case, triamcinolone acetonide 0.1%, vasoconstrictor
response differed 2.5-fold within the creams and almost 4-fold (see Fig. 25D) within
Figure 25 (A) Variation in percentage of total urinary excretion of hydrocortisone from

the forearm site in 18 individuals; (B) variation in percentage of total urinary excretion of
hydrocortisone applied to different skin sites in humans; (C) effect of forearm skin condition
on urinary excretion of hydrocortisone; (D) bioequivalence of triamcinolone acetonide, 0.1%,
in various ointment formulations.
the ointments. Shah et al. (460), Jackson et al. (402), and Olsen (461) have since
confirmed bioinequivalence within commercial topical steroid formulations contain-
ing the same active substance at the same concentration.
All of the foregoing studies used the vasoconstrictor method to measure bio-
equivalence. Relatively few data are currently available for corticosteroids deter-
mining bioequivalence using drug level measurements in vivo (120,121). Recently,
Agrawal et al. (462) compared the bioavailability of hydrocortisone from seven com-
mercial products using a standard in vitro technique. Total permeation over 48 h
ranged from 3.0 ␮g (1% cream) to 0.13 ␮g (1% ointment), with large differences
between 1% cream formulations, depending on manufacturer. This group, under sim-

ilar conditions, compared the bioavailability of hydroquinone from topical over-the-
counter (OTC) and prescription products (463a). The extent of absorption varied
tenfold, with some of the 2% OTC products delivering more drug than the 4%
prescription products.
From this review, it appears that bioinequivalence is common within topical
dermatological products. However, this is to be anticipated given the low extent of
absorption and the potential for extremely large variation between formulations in
drug partitioning into and diffusion through the skin.
2. Bioequivalence and Therapeutic Equivalence
Even if we assume that bioequivalence is established between two or more topical
dermatological formulations in volunteers, there is still potential for therapeutic in-
equivalence. In this context, variation is in the extent of absorption between skin
sites (e.g., forearm and face) and skin conditions (e.g., open and occluded, healthy
or diseased) in the volunteer experimental versus the clinical setting.
a. Potential for Bioequivalence but Difference in Efficacy. Figure 26 illus-
trates the hypothetical case in which, even though two formulations are bioequiv-
alent, variation in extent of absorption occurs between the bioavailability model and
the clinical setting. The curves represent wanted and unwanted effects (see Fig. 24)
and, for simplicity, similar shapes for the effect curves in both volunteers and the
clinical setting have been assumed.
Data to exemplify the hypothetical cases in Fig. 26 are still rare (313). How-
ever, as outlined previously (see Sec. III), it is likely that there are changes in extent
of absorption in volunteers and patients owing to change in skin site (forearm in
volunteer studies versus any skin site in patients) and change in skin condition
(healthy vs. diseased skin, and occluded conditions in many experimental situations
vs. the open or occluded conditions in clinical situations). The following examples,
comparing the vasoconstrictor activity of betamethasone valerate in cream and oint-
ment formulations under occluded or open test situations illustrates the hypothetical
cases (Table 8).
Case A. If Betnovate and Betnovate N,C were tested in volunteers under open
conditions, they would be approximately bioequivalent, whereas un-
der occluded conditions in patients Betnovate is supraavailable.
Case B. If Betnovate and Benovate A,C,N were tested in volunteers under
occluded conditions, they would be approximately bioequivalent,
whereas under open conditions in patients Benovate-A would be
inferior.
Thus, two or more formulations that are bioequivalent in volunteers may give dif-
ferent clinical beneficial responses.
b. Effect of Finite and Infinite Dose: Control of Overdosing. Occlusion may
give rise to modest increases in percutaneous absorption and thus changes in efficacy.
Also of clinical importance is that gross increase in the extent of absorption between
the bioavailability model and clinical setting may occur and will result in overdosing
and, possibly, severe adverse effects. For example, Aalto–Korte (24) has described
the absorption of systemically active doses of hydrocortisone following topical ap-
Figure 26 Pictogrammed hypothetical cases in which two formulations are bioequivalent

in volunteers, but variation in extent of absorption in a therapeutic setting may result in the
potential for therapeutic inequivalence in patients. In case 1 no difference in extent of ab-
sorption in volunteers and patients for both products is observed, resulting in therapeutic
equivalence (comparable with Synalar formulations; 455). In case 2 significant change in
extent of absorption in patients is observed for both products, also resulting in therapeutic
equivalence (comparable with Propaderm formulations; 403). In case 3 significant change in
the extent of absorption in patients is observed for only one of the formulations resulting in
therapeutic inequivalence (comparable with Betnovate cream and ointment formulations,
403,455).
Table 8 Cumulative Vasoconstrictor Values for Various Betamethasone Valerate Formulations

(Betamethasone) Under Occluded and Open Conditions
Case A Case B
Formulation Occluded Open Ratio Formulation Occluded Open Ratio
(creams) conditions conditions (%) (ointments) conditions conditions (%)
— — — — Betnovate-A 2240 1230 54.9

Betnovate-C 1210 1220 100.0 Betnovate-C 2170 74.7
Betnovate-N 1300 1130 86.9 Betnovate-N 2140 1630 76.2
Betnovate 1670 1060 63.6 Betnovate 2080 1870 89.90
Sources: Case A, 403; case B, 455.
plication to patients with erythroderma. As outlined in detail in Sec. IV.C.5.c, under

experimental conditions, the delivery system (vehicle) is generally far from being
exhausted (depleted) and even under clinical, hence, less controlled, conditions the
dose (amount of drug) that can be recovered after an adequate time interval is high.
Therefore, one may conclude that in both experimental and clinical situations, one
is actually closer to an infinite dose situation than to a finite dose situation.
Figure 27 shows a schematic of the pharmacokinetic–pharmacodynamic dose
(concentration)–response, and variation in this, from a high, infinite, dose and a low,
finite, dose. On average, the two doses are bioequivalent, which requires the use of
an enhanced delivery system to increase bioavailability from the finite dose as de-
scribed later. With the infinite dose, variation in skin diffusivity causes the dose
absorbed to vary into that range where significant—and unjustified—unwanted ef-
fects may occur. For the finite dose, control by the dose on the upper limit of
Figure 27 Schematic showing the effect of finite and infinite dose situations.
absorption limits the potential for unwanted effects. Use of finite–dose-enhanced

delivery systems are described further in the following.
Among individuals, for a given skin site and skin condition, there is approxi-
mately a 10-fold variation. For any given individual, there is approximately a 50-
fold variation in the dose absorbed between skin sites. Variation in skin condition
leads to further variation (see Fig. 25A–C). Such differences in the extent of ab-
sorption between formulations, depending on skin site and condition, will also lead
to disproportionate increases in unwanted effect. This situation is probably common
in topical therapy, particularly for compounds such as corticosteroids that are applied
to body sites of differing permeabilties.
c. Conventional High-Concentration and Enhanced Low-Concentration For-
mulations. The overemphasis of drug concentration in topical vehicle design has
been mentioned, as has the fact that many current topical formulations are poorly
bioavailable; that is, they contain high concentrations of active agent compared with
the fractional amount absorbed and required. One may therefore define these for-
mulations as conventional high-concentration formulations. The application of these
formulations may create an infinite-dose situation (see Table 6) on the skin.
In Sec. IV.B.2.c, the potential of pharmaceutical techniques, such as thermo-
dynamic activity and chemical enhancement, to increase percutaneous absorption are
outlined. In the following section, arguments for increasing the bioavailability—
extent of absorption—of topical products using such techniques, with concurrent
reduction of drug concentration in these formulations will be developed. In contrast
to the aforementioned conventional high-concentration formulations, a new type of
formulation will be defined as enhanced low-concentration formulations. In Fig. 28
the concept of the enhanced low-concentration formulations to control variation in
the extent of absorption is developed.
As discussed earlier and illustrated in Figs. 25–28, one usually has little or no
control over the dose absorbed with the current conventional high-concentration for-
mulations. This fact is also pictographed in Fig. 28B. With current conventional high-
concentration formulations, the bioavailability of topical dermatological agents are
generally low (1–5%), leading to a potential for 20- to 100-fold increase in absorp-
tion, depending on skin permeability. Increasing bioavailability by pharmaceutical
>
Figure 28 Schematic presentations of the effect of enhanced low-concentration formula-
tions: (A) Chemical enhancers or supersaturation are pharmaceutical means to enhance drug
absorption through the skin. Chemical enhancers act by fluidizing the stratum corneum lipid
barrier (pictogrammed as opening the funnel), supersaturation acts by increasing the ther-
modynamic activity of the drug in the vehicle (pictogrammed as arrows). (B) Variation in
skin site and skin condition, including disease and disease stage, as well as the individual,
influence topical bioavailability. Bioavailability of topical dermatological agents are generally
low, leading to a potential for dramatic increase in absorption depending on skin permeability.
(C) Increasing bioavailability by pharmaceutical means (e.g., supersaturation) does not elim-
inate variation owing to skin site, skin condition, including disease and disease stage, or the
individual. (D) Variation in bioavailability may be reduced by reducing drug concentration in
the delivery system and, at the same time, increasing the delivery efficiency (e.g., supersat-
uration) from the delivery system.
means (e.g., chemical enhancers or supersaturation) is possible (see Fig. 28A); how-
ever it does not account for the different permeability of the skin (see Fig. 28C) and,
therefore, does not avoid differences in bioavailability (high variation). To avoid this
source of variation, one may reduce drug concentration in the delivery system and
at the same time increase delivery efficiency from the delivery system (see Fig. 28D).
Reduction of the drug concentration in the delivery system reduces the potential for
excessive delivery. This maneuver is important for individuals with highly permeable
skin (newborns), for body sites with highly permeable skin (intertrigo), or for indi-
viduals with impaired barrier function. Increasing the delivery efficiency of the de-
livery system increases drug delivery in individuals with less permeable skin (adults),
increases drug delivery in skin sites with low skin permeability (forearm) or with
increased barrier function. The skillful combination of these two strategies—reduc-
tion of the drug concentration and increased delivery efficiency—therefore, will most
likely ensure equivalent therapy and lead to a drug delivery system with highly
increased bioavailability, preferably in the range of 50–100%. By definition, the
variation in dose absorbed, depending on skin permeability, will decrease dramati-
cally. Pharmaceutical technologies, such as supersaturation or chemical enhancers,
are required to deliver low doses to the required bioavailability standards of 50–
100%, and further work to apply these to enhancement of low-concentration for-
mulations is required.
Marks et al. (464) have established that these enhanced low-concentration for-
mulations are as efficacious as conventional high-concentration formulations (Fig.
29). In study AI/AII the vasoconstrictor responses of the 1% hydrocortisone acetate
(HA) and 0.02% eightfold-saturated HA are not significantly different from each
other, yet each of these is significantly different from the control treatments (p <
0.05, Wilcoxon-matched pairs-signed ranks test). Ideally, these comparisons should
be made under nonoccluded conditions because occlusion will increase absolute bio-
availability. Nonoccluded vasoconstrictor studies on HA are not yet feasible owing
to the relatively weak response of HA in the model. However, these results dem-
onstrate bioequivalence of the 1% HA and 0.02% eightfold-saturated HA formula-
tions under the conditions of this study. In study BI/BII—the surfactant-induced
erythema test—the significance level was at p < 0.05 using the Wilcoxon-matched
pairs-signed ranks test.
From our recent in vitro membrane permeation experiments (463b), it was
concluded that drug flux can be increased supraproportionally with increasing donor
concentration (drug [super]saturation [proportional]) beyond what would be antici-
pated based on ideal donor concentrations and partition coefficient considerations
alone. These findings could not be confirmed in an in vivo investigation, probably
owing to additional vehicle effects (e.g., enhancement, irritation, or drug-binding)
that must be expected and could have altered the integrity of the stratum corneum
and, thereby, topical bioavailability.
D. Variation in the Distribution, Metabolism, and Excretion

Process and Concentration–Effect Response: Individualization
of Dosing
If there is variation in pharmacokinetic (distribution, metabolism, and excretion pro-
cess) or pharmacodynamic response (concentration–effect response) among volun-
Figure 29 Bioequivalence of 0.02% w/w supersaturated hydrocortisone acetate (HA) with

conventional 1% w/w HA cream: study AI/AII, blanching-test in volunteers; study BI/BII,
surfactant-induced erythema in volunteers. Treatments: 1, untreated; 2, 0.02% w/w supersat-
urated HA in gel; 3, gel base; 4, 1% w/w HA in cream; 5, cream base. (From Ref. 52.)
teers or patients, then one may titrate to that specific dose of drug that best suits an
individual patient. For example, in concentration-controlled trials (CCTs), the dose
is varied between patients to obtain a specific tissue, usually plasma, concentration.
CCTs address the issue of variation in pharmacokinetic factors. Levy et al. (444)
have discussed the importance of variation in pharmacodynamic factors and have
proposed the strategy of varying the dose to obtain a specific pharmacological end-
point, the so-called effect-controlled trials (ECT). They provide convincing evidence
from the literature that, contrary to current paradigm, variation in pharmacodynamic
factors are the major source of variation in therapy. Thus, there is a growing belief
that the current, fixed-dose, paradigm is outdated, and that one should be developing
new dosing strategies for dose titration and improved health care (444,465,466).
Whether this is possible with our current armamentarium of topical dermatological
products will now be discussed.
1. Dose Titration with Conventional High-Concentration Formulations

Despite some single experiments in which dose–response is observed; for example,
(a) pharmacokinetically with hydrocortisone by varying the concentration of the dose
applied (142,417); (b) pharmacodynamically with corticosteroids by varying the dose
absorbed by duration of dosing (123); and (c) clinically with 5-fluorouracil by vary-
ing the dose applied by frequency (intermitting dosing) (425). Generally, no such
relations with topical dermatological products are observed. Again, using the corti-
costeroids as a well-studied example, there are many reports of lack of, or poor,
dose–response (393,397,399,467–472).
For an increase in dose applied to result in an increase in response, both the
dose applied–dose absorbed relation (pharmacokinetic) and the dose absorbed–re-
sponse relation (pharmacodynamic) should be linear or approximately so. Thus, when
the response from increasing the dose applied is flat, this can either be due to phar-
macokinetic control (dose applied–dose absorbed is flat) or pharmacodynamic con-
trol (dose absorbed–response relation is flat), or both. Pharmacokinetic versus phar-
macodynamic control of flat topical dose–response has been the subject of some
discussion (333,460,470,473). Variation in pharmacodynamic dose–response is wide
among subjects (466), and this will confound the use of pharmacodynamic param-
eters to predict pharmacokinetics, including bioavailability and bioequivalence. As
part of a program to improve the sensitivity of the vasoconstrictor test, the recent
Guidance Topical Dermatologic Corticosteroids: In Vivo Bioequivalence (21) pro-
poses the screening of subjects into detectors and hypo- and hyperresponders. De-
tectors, subjects showing linear pharmacodynamic response to dose absorbed, are
selected for bioavailability comparisons.
Barry et al. (470), in a study designed to show pharmacokinetic control, have
shown no difference between the vasoconstrictor response and clinical antipsoriatic
response to 0.05 and 0.1% desonide creams. Both concentrations of desonide were
saturated in the vehicle and, thus, as described in Sec. IV, were predicted to give
rise to the same extent of drug penetration. Because there is no increase in dose
absorbed as dose applied is increased, there can be no difference in vasoconstriction
or clinical activity. What is not fully appreciated is the importance of lack of deple-
tion in the flat relation between dose applied and dose absorbed. Saturated vehicles
saturate the skin and thus the initial rate of absorption is the same. Only because
they are infinite doses do these vehicles remain essentially saturated and thus con-
tinue to give the same rate of drug absorption. To achieve a pseudo-dose–response
(increase in dose absorbed with dose applied) with current high-concentration for-
mulations it is necessary to increase the thermodynamic activity of the drug in the
vehicle as concentration is increased. For example, if the highest drug concentration
in a given vehicle is just saturated, then lower concentrations in this same vehicle
will give a dose–response (333) (see Sec. IV, Fig. 17 and discussion). Although
increase in thermodynamic activity can be used to generate a pseudo-dose–response
with high-concentration formulations, it is not (theoretically) possible to provide a
dose–response to increase in drug dose varied by the amount of product per skin
area (52). Again this is because of lack of depletion. The key to provision of a robust
dose–response between dose applied and dose absorbed is depletion or, put another
way, the use of rational, low but therapeutic, doses with optimized delivery to give
bioavailability in the range of 50–100%.
2. Dose Titration with Enhanced Low-Concentration Formulations
On the basis of simple pharmacokinetic modeling, low, yet rational, doses formulated
in enhanced low-concentration topical delivery systems, are predicted to produce a
strong relation between dose applied and dose absorbed (52). Although this is almost
self-evident, there are, currently, few in vivo experimental studies to support this
prediction. Figure 30 shows experimental data in volunteers for the irritant response
(on the forehead) to increasing concentrations of a sixfold supersaturated topical
retinoic acid system. Despite significant between-subject variation in response, in-
cluding the presence of hyporesponders, there was a clear dose (concentration)–
applied response indicative of a strong dose-applied–dose-absorbed relation from
this type of enhanced low-concentration formulation (52,474).
VI. CONCLUSIONS
Standards in topical bioavailability, bioequivalence, and therapeutic equivalence,
both in methods and protocols for testing, and also in biopharmaceutical parameters
of current products, are some 10–20 years behind those standards that exist in other
(e.g., oral) therapies.
Developments in methods and protocols for bioavailability and bioequivalence,
including new statistical evaluations, growing knowledge in design of studies, and
the extraordinary range of analytical assays, provoked the revision of many older
regulatory guidances. Significant attempts have been made by the authorities of many
countries to eliminate the vagueness in older guidelines concerning the investigation
of bioavailability and bioequivalence. The various recommendations emerged from
a series of international symposias (e.g., 2nd EUFEPS Nürnberg Conference: Sym-
posium on Quality and Interchangeability of Topical Products for Local Action, De-
cember 8–9, 1995, Nuremberg, Germany; or FDA/AAPS, October, 1996, Washing-
ton DC) help to aim for a coherent body of terminology on, and understanding of,
bioavailability and bioequivalence that can be used in international harmonization
on this subject.
Nevertheless, numerous issues and areas of research still require further in-
depth discussion and experimentation. The vasoconstrictor test has been utilized as
a tool to assess bioavailability and bioequivalence of topical corticosteroids for sev-
eral decades. However, work is still needed to establish the mechanism of action to
determine to what extent skin blanching is actually related to the therapeutic use of
such drug products. The vasoconstrictor test has also provoked a heated debate on
the possibilities and procedures to show dose–response relation with topical corti-
costeroids to meet some of the regulatory requirements. Topical corticosteroids rep-
resent just one class of dermatological therapeutics that can be judged by biophar-
maceutical means; however, for most other dermatological remedies, less-developed
or no bioassays exist.
Therefore, clinical trials appear to be the only means of assessing bioavaila-
bility–bioequivalence of many topical dosage forms in the foreseeable future. Es-
tablishment of bioavailability–bioequivalence criteria tailored to specific diseases, to
specific groups of chemicals with the same indication, or to special patient popula-
tions are currently being discussed.
Topical bioavailability of most currently marketed formulations is low, in the
region of 1–5% of dose applied. From this, given the profound effect that formu-
lation can have on drug partitioning into and diffusion through the skin, bioinequiv-
alence can be anticipated and, from the data available, seems widespread. Even when
bioequivalence is established in volunteers, the known occurrence of large variations
in absorption between skin sites and depending on skin condition, give little assur-
ance of therapeutic equivalence in the clinical situation. Major improvements in the
Figure 30 (Bottom) Dose (concentration)–response of forehead irritancy in 30 volunteers to

increasing doses of enhanced low-concentration retinoic acid formulation. (From Ref. 52.)
quality of therapy are predicted as a result of individualization of dosing. However,

a dose–response from topical dermatological products is poor, confounded by phar-
macodynamic variation, but controlled by the pharmacokinetics of current, high-
concentration, formulations. Low-concentration, pharmaceutically optimized formu-
lations of 50–100% bioavailability seem, from a therapeutic perspective, to set future
standards.
Many scientific and regulatory questions remain to be answered. The artful
therapies and dosage forms of the past have definitely fallen into disuse and have
been replaced by experimentally tested and proved therapeutics housed in more el-
egant delivery vehicles. Nevertheless, strong threads of ancient art in today’s topical
treatment remain, and intuition and feel, trial-and-error, still surrogate for science in
topical product development and dermatological practice.
REFERENCES
1. Melnick D, Hochberg M, Oser BL. Physiological availability of vitamins. I. The
human bioassay technique. J Nutr 30:67–79, 1945.
2. Morrison AB, Chapman DG, Campbell JA. Further studies on the relation between
in vitro disintegration time of tablets and the urinary excretion rates of riboflavin. J
Am Pharm Assoc Sci Ed 48:634–637, 1959.
3. Morrison AB, Campbell JA. The relationship between physiological availability of
salicylates and riboflavin and in vitro disintegration time of enteric coated tablets. J
Am Pharm Assoc Sci Ed 49:473–478, 1960.
4. Levy G. Comparison of dissolution and absorption rates of different commercial as-
pirin tablets. J Pharm Sci 50:388–392, 1961.
5. Middleton EJ, Davies JM, Morrison AB. Relationship between rate of dissolution,
disintegration time, and physiological availability of riboflavin in sugar-coated tablets.
J Pharm Sci 53:1378–1380, 1964.
6. Tyrer JH, Eadie MJ, Sutherland JM, Hooper WD. Outbreak of anticonvulsant toxicity
in an Australian city. Br Med J 4:271–273, 1970.
7. Danon A, Horowitz J, Ben–Zvi Z. An outbreak of digoxin intoxication. Clin Phar-
macol Ther 21:643–646, 1977.
8. Melikian AP, Straughn AB, Slywka GWA, Whyatt PL, Meyer MC. Bioavailability of
eleven phenytoin products. J Pharmacokinet Biopharm 5:133–146, 1977.
9. Glazko AJ, Kinkel AW, Alegnani WC, Holmes EL. An evaluation of absorption char-
acteristics of different chloramphenicol preparations in normal human subjects. Clin
Pharmacol 9:472–483, 1968.
10. Flynn GL, Weiner ND. Topical and transdermal delivery—provinces of realism. In:
Gurny R, Teubner A, eds. Dermal and Transdermal Drug Delivery. Stuttgart: Wissen-
schaftliche Verlagsgesellschaft, pp 33–65, 1993.
11. Somberg JC. Bioequivalence or therapeutic equivalence. J Clin Pharmacol 26:1, 1986.
12. Dettelbach HR. A time to speak out on bioequivalence and therapeutic equivalence.
J Clin Pharmacol 26:307–308, 1986.
13. Lamy PP. Generic equivalents: issues and concerns. J Clin Pharmacol 26:309–316,
1986.
14. Dupuis D, Rougier A, Roguet R, Lotte C, Kalopissis G. In vivo relationship between
horny layer reservoir effect and percutaneous absorption in human and rat. J Invest
Dermatol 82:353–356, 1984.
15. Guzek DB, Kennedy AH, McNeill SC, Wakshull E, Potts RO. Transdermal drug
transport and metabolism. I. Comparison of in vitro and in vivo results. Pharm Res
6:33–39, 1989.
16. Schalla W, Jamoulle J–C, Schaefer H. Localization of compounds in different skin

layers and its use as an indicator of percutaneous absorption. In: Bronaugh RL, Mai-
bach HI, eds. Percutaneous Absorption. 2nd ed. New York: Marcel Dekker, pp 283–
312, 1989.
17. Surber C. Drug concentration in the skin. In: Maibach HI, ed. Dermatologic Research
Techniques. Boca Raton: CRC Press, pp 151–178, 1996.
18. Metzler CM. Statistical criteria. In: Welling PG, Tse FLS, Dighe SV, eds. Pharma-
ceutical Bioequivalence. New York: Marcel Dekker, pp 35–66, 1991.
19. Hauck WW. Bioequivalence studies of topical preparations: statistical considerations.
Int J Dermatol 31(suppl 1):29–33, 1992.
20. Rauws AG. Bioequivalence: a European community regulatory perspective. In: Well-
ing PG, Tse FLS, Dighe SV, eds. Pharmaceutical Bioequivalence. New York: Marcel
Dekker, pp 419–441, 1991.
21. FDA, Center for Drug Evaluation and Research. GUIDANCE: Topical Dermatologic
Corticosteroids: In Vivo Bioequivalence. Rockville, MD: Division of Bioequivalence,
1995.
22. Stoughton RB. Vasoconstrictor assay: specific applications. In: Maibach HI, Surber
C, eds. Topical Corticosteroids. Basel: S Karger, pp 42–53, 1992.
23. Smith EW, Meyer E, Haigh JM. Accuracy and reproducibility of the multiple-reading
skin blanching assay. In: Maibach HI, Surber C, eds. Topical Corticosteroids. Basel:
S Karger, pp 65–73, 1992.
24. Aalto–Korte K. Quantifying systemic absorption of topical hydrocortisone in eryth-
roderma. Br J Dermatol 133:403–408, 1995.
25. Anderson S, Hauck WW. A new procedure for testing equivalence in comparative
bioavailability and other clinical trials. Commun Stat Theor Method 12:2663–2692,
1983.
26. Hauck WW, Anderson S. A new statistical procedure for testing equivalence in two-
group comparative bioavailability trials. J Pharmacokinet Biopharm 12:83–91, 1984.
27. Schuirmann DJ. A comparison of the two one-sided tests procedure and the power
approach for assessing the equivalence of average bioavailability. J Pharmacokinet
Biopharm 15:657–680, 1987.
28. Steinijans VW, Hauschke D, Jonkman JHG. Controversies in bioequivalence studies.
Clin Pharmacokinet 22:247–253, 1992.
29. Hauck WW, Anderson S. Types of bioequivalence and related statistical considera-
tions. Int J Clin Pharmacol Ther Toxicol 30:181–187, 1992.
30. Skelly JP, Shah VP, Maibach HI, Guy RH, Wester RC, Flynn GL, Yacobi A. FDA
and AAPS report of the workshop on principles and practices of in vitro percutaneous
penetration studies: relevance to bioavailability and bioequivalence. Pharm Res 4:
265–267, 1987.
31. Higuchi WI. Analysis of data on the medicament release from ointments. J Pharm Sci
51:802–804, 1962.
32. Shah VP, Elkins J, Hanus J, Noorizadeh C, Skelly JP. In vitro release of hydrocorti-
sone from topical preparations and automated procedure. Pharm Res 8:55–59, 1991.
33. Neubert R, Wohlrab W. In vitro methods for biopharmaceutical evaluation of topical
formulations. Acta Pharm Technol 36:197–206, 1990.
34. FDA, Center for Drug Evaluation and Research. Interim Guidance. Topical Cortico-
steroids: In Vivo Bioequivalence and In Vitro Release Methods. Rockville, MD: Di-
vision of Bioequivalence, 1992.
35. United States Pharmacopeia. Vol. XXII. Rockville, MD: United States Pharmacopeial
Convention, 1990.
36. Shah VP, Tymes NW, Yamamoto LA, Skelly JP. In vitro dissolution profile of trans-
dermal nitroglycerin patches using paddle method. Int J Pharm 32:243–250, 1986.
37. Shah VP, Tymes NW, Skelly JP. Comparative in vitro release profiles of marketed
nitroglycerin patches by different dissolution methods. J Control Release 7:79–86,
1988.
38. Shah VP, Tymes NW, Skelly JP. In vitro release profiles of clonidine transdermal
therapeutic systems and scopolamine transdermal patches. Pharm Res 6:346–351,
1989.
39. Reddish G, Wales H. Antiseptic action of USP and NF ointments. J Am Pharm Assoc
18:576–578, 1929.
40. Christoff K. Einfluss der oberflächenaktiven Substanzen OAS auf die strukturrheo-
logischen Eigenschaften der Salbengrundlagen und die Abgabeschwindigkeit von Arz-
neimitteln. Pharmazie 22:251–253, 1967.
41. Clark G, Davies G. A new method for measuring diffusion of antiseptics from oint-
ment bases. Br J Dermatol 1:521–525, 1949.
42. Unna P. Zur frage der salbengrundlagen. Dtsch Med Wochenschr 51:197–199, 1926.
43. Poulsen BJ, Young E, Coquilla V, Katz M. Effect of topical vehicle composition on
the in vitro release of fluocinolone acetonide and its acetate ester. J Pharm Sci 57:
928–933, 1968.
44. Loth H, Holla–Benninger A. Untersuchungen der arzneiliberation aus Salben. 1. Mit-
teilung: entwicklung eines in vitro Liberationmodells. Pharm Ind 40:256–261, 1978.
45. De Vos A, Vervoort L, Kinget R. Release of indomethacin from transparent oil–waters
gels. J Pharm Sci 83:641–643, 1994.
46. Bottari F, Di Colo G, Nannipieri E, Saettone MF, Serafini MF. Influence of drug
concentration on in vitro relase of salicylic acid from ointment bases. J Pharm Sci
63:1779–1783, 1974.
47. Delonca H, Chanal JL, Maillols H, Ghebbi R. Méthode d’étude in vitro des prépar-
ations cutanées semi-solides. Comparisons in vitro–in vivo. Pharm Acta Helv 52:51–
64, 1977.
48. Guy RH, Fleming R. The estimation of diffusion coefficients using the rotating dif-
fusion cell. Int J Pharm 3:143–149, 1979.
49. Touitou E, Abed L. The permeation behavior of several membrane with potential use
in the design of transdermal devices. Pharm Acta Helv 60:193–198, 1985.
50. Hadgraft J, Walters KA, Wotton PK. Facilitated percutaneous absorption: a compar-
ison and evaluation of two in vitro models. Int J Pharm 32:257–263, 1986.
51. Velissaratou AS, Papaioannou G. In vitro release of chlorpheniramine maleate from
ointment bases. Int J Pharm 52:83–86, 1989.
52. Davis A. Novel formulations for improved therapy. PhD dissertation, University of
Wales, College of Cardiff, 1995.
53. Shah VP, Elkins J, Lam SY, Skelly JP. Determination of in vitro drug release from
hydrocortisone creams. Int J Dermatol 53:53–59, 1989.
54. Shah VP, Elkins J, Skelly JP. Relationship between in vivo skin blanching and in
vitro release rate for betamethasone valerate creams. J Pharm Sci 81:104–106, 1992.
55. Shah VP, Skelly JP. Practical considerations in developing a quality control (in vitro
release) procedure for topical drug products. In: Shah VP, Maibach HI, eds. Topical
Drug Bioavailability, Bioequivalence, and Penetration. New York: Plenum Press, pp
107–117, 1993.
56. Behme RJ, Kensler TT, Brooke D. A new technique for determining in vitro release
rates of drugs from creams. J Pharm Sci 71:1303–1305, 1982.
57. Corbo M, Schultz TW, Wong GK, Van Buskirk GA. Development and validation of
in vitro release testing methods for semisolid formulations. Pharm Technol (Sept):
112–128, 1993.
58. Guy RH, Hadgraft J. On the determination of drug release rates from topical dosage
forms. Int J Pharm 60:R1–R3, 1990.
59. Rolland A, Demichelis G, Jamoulle J–C, Shroot B. Influence of formulation, receptor

fluid, and occlusion, on in vitro drug release from topical dosage forms, using an
automated flow-through diffusion cell. Pharm Res 9:82–86, 1992.
60. Fares HM, Zatz JL. Measurement of drug release from topical gels using two types
of apparatus. Pharm Technol (Jan):54–56, 1995.
61. Flynn GL. Comparisons between in vitro techniques. Acta Pharm Suec 20:54–55,
1983.
62. Poulsen BJ, Flynn GL. In vitro methods used to study dermal delivery and percuta-
neous absorption. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption. New
63. Barry BW. Dermatological Formulations. New York: Marcel Dekker, 1983.
64. Addicks W, Weiner N, Flynn G, Curl R, Topp E. Topical drug delivery from thin
applications: Theoretical predictions and experimental results. Pharm Res 7:1048–
1054, 1990.
65. Flynn GL, Smith RW. Membrane diffusion III: Influence of solvent composition and
permeant solubility on membrane transport. J Pharm Sci 61:61–66, 1972.
66. Tanaka S, Takashima Y, Murayama H, Tsuchiya S. Studies on release from ointments:
V. Release of hydrocortisone butyrate propionate from topical dosage forms to silicone
rubber. Int J Pharm 27:29–38, 1985.
67. Davis AF, Hadgraft J. Effect of supersaturation on membrane transport: 1. Hydrocor-
tisone acetate. Int J Pharm 76:1–8, 1991.
measurements. II. Dipalmitoyl phosphatidylcholine, linoleic acid and tetradecane. Int
J Pharm 42:97–104, 1988.
69. Buchmann S. Entwicklung eines in vitro absorptionmodelles auf der basis von zell-
kulturen. Basel: University of Basel, 1987.
70. Steinsträsser I. The organized HaCaT cell culture sheet: a model approach to study
epidermal peptide drug metabolism. Zürich: ETH Zürich, 1994.
71. Williams RL. Bioequivalence of topical corticosteroids: a regulatory perspective.
Semin Dermatol 31(suppl 1):2–5, 1992.
72. Beastall J, Guy RH, Hadgraft J, Wilding I. The influence of urea on percutaneous
absorption. Pharm Res 3:294–297, 1986.
73. Lippold BC, Teubner A. Biopharmazeutische qualität von Arzneiformen, insbesondere
für lokale Anwendung, abgeleitet aus Wirkungsmessungen. Pharm Ind 43:71–73,
1981.
74. Lippold BC, Teubner A. Einfluss verschiedener salbengrundlagen auf die Wirkung
von Nicotinsäurebenzylester in lösungssalben. Pharm Ind 43:1123–1133, 1981.
75. Lippold BC, Reimann H. Wirkungsbeeinflussung bei lösungssalben durch Vehikel am
beispiel von Methylnicotinat. Acta Pharm Technol 3:136–142, 1989.
76. Katz M, Poulsen BJ. Corticoid, vehicle, and skin interaction in percutaneous absorp-
tion. J Soc Cosmet Chem 23:565–590, 1972.
77. Caron D, Queille–Roussel C, Shah VP, Schaefer H. Correlation between the drug
penetration and the blanching effect of topically applied hydrocortisone creams in
human beings. J Am Acad Dermatol 23:458–462, 1990.
78. Kaiho F, Nomura H, Makabe E, Kato Y. Percutaneous absorption of indomethacin
from mixtures of fatty alcohol and propylene glycol (FAPG base) through rat skin:
effects of fatty acid added to FAPG base. Chem Pharm Bull 35:2928–2934, 1987.
79. Tojo K. Concentration profile in plasma after transdermal drug delivery. Int J
Pharm 43:201–205, 1988.
80. Franz TJ. Percutaneous absorption. On the relevance of in vitro data. J Invest Der-
matol 64:190–195, 1975.
81. Feldmann RJ, Maibach HI. Absorption of some organic compounds through the skin
in man. J Invest Dermatol 54:399–404, 1970.
82. Sheth NV, McKeough MB, Spruance SL. Measurement of stratum corneum drug
reservoir to predict the therapeutic efficacy of topical iododeoxyuridine for herpes
simplex. J Invest Dermatol 89:598–602, 1987.
83. Öhman H, Vahlquist A. In vivo studies concerning a pH gradient in human stratum
corneum and upper epidermis. Acta Derm Venereol (Stockh) 74:375–379, 1994.
84. Erikson G, Lamke L. Regeneration of human epidermal surface and water barrier
function after stripping. Acta Derm Venereol (Stockh) 51:169–178, 1971.
85. Wilhelm D, Elsner P, Maibach HI. Standardized trauma (tape-stripping) in human
vulvar and forearm skin. Acta Derm Venereol (Stockh) 71:123–126, 1991.
86. Downes AM, Matoltsy AG, Sweeney TM. Rate of turnover of the stratum corneum
in hairless mice. J Invest Dermatol 49:400–405, 1967.
87. Pinkus H. Examination of the epidermis by the strip method of removing horny layers.
I. Observations on thickness of the horny layer, and on mitotic activity after stripping.
88. Pinkus H. Examination of the epidermis by the strip method of removing horny layers.
II. Biometric data on regeneration of human epidermis. J Invest Dermatol 16:431–
447, 1951.
89. Piérard GE, Piérard–Franchimont C, Saint–Léger D, Kligman AM. Squamometry: the
assessment of xerosis by colorimetry of D-Squame adhesive discs. J Soc Cosmet Chem
47:297–305, 1992.
90. Schatz H, Kligman AM, Manning S, Stoudemayer T. Quantification of dry (xerotic)
skin by image analysis of scales removed by adhesive discs (D-Squame). J Soc Cosmet
Chem 44:53–63, 1993.
91. Malkinson FD, Ferguson EH. Percutaneous absorption of hydrocortisone-4-14C in two
human subjects. J Invest Dermatol 25:281–285, 1955.
92. Vickers CFH. Existence of reservoir in the stratum corneum. Arch Dermatol 88:20–
23, 1963.
93. Stoughton RB. Dimethylsulfoxide (DMSO) induction of a steroid reservoir in human
skin. Arch Dermatol 91:657–660, 1965.
94. Carr RD, Wieland RG. Corticosteroid reservoir in the stratum corneum. Arch Der-
matol 94:81–84, 1966.
95. Carr RD, Tarnowski WM. The corticosteroid reservoir. Arch Dermatol 94:639–642,
1966.
96. Munro DD. The relationship between percutaneous absorption and stratum corneum
retention. Br J Dermatol 81(suppl 4):92–97, 1969.
97. Lücker P, Nowak H, Stüttgen G, Werner G. Penetrationskinetik eines Tritium-
markierten 9Alpha-fluor-16 methylen-prednisolonesters nach epicutaner Applikation
beim Menschen. Arzneimittelforschung 18:27–29, 1968.
98. Tsai J–C, Cappel MJ, Flynn GL, Weiner ND, Kreuter J, Ferry J. Drug and vehicle
deposition from topical applications: use of in vitro mass balance technique with
minoxidil solutions. J Pharm Sci 81:736–743, 1992.
99. Tojo K, Lee AC. A method for predicting steady-state rate of skin penetration in vivo.
100. Moon KC, Wester RC, Maibach HI. Diseased skin models in the hairless guinea pig:
in vivo percutaneous absorption. Dermatologica 180:8–121, 1990.
101. Holoyo–Tomoka MT, Kligman AM. Does cellophane tape stripping remove the horny
layer? Arch Dermatol 106:767–768, 1972.
102. Malkinson FD. Studies on the percutaneous absorption of 14C-labelled steroids by use
of glass-flow cell. J Invest Dermatol 31:19–28, 1958.
103. Feldmann RJ, Maibach HI. Penetration of 14C hydrocortisone through normal skin:
effect of stripping and occlusion. Arch Dermatol 91:661–666, 1965.
104. Zesch A, Schaefer H. Penetrationskinetik von radiomarkierten Hydrocortisone aus
verschiedenartigen salbengrundlagen in die Menschliche haut. Arch Dermatol Forsch
252:245–256, 1975.
105. Zesch A, Schaefer H, Hoffmann W. Barriere und reservoirfunktion der einzelnen
hornschichtlagen der Menschlichen haut für lokal aufgetragene Arzneimittel. Arch
Dermatol Forsch 246:103–107, 1973.
106. Zesch A. Reservoirfunktion der hornschicht. In: Klaschka F, ed. Stratum corneum:
Struktur und Funktion. Berlin: Grosse Verlag, pp 63–76, 1981.
107. Dupuis D, Rougier A, Roguet R, Lotte C. The measurement of the stratum corneum
reservoir: a simple method to predict the influence of vehicles on in vivo percutaneous
absorption. Br J Dermatol 115:233–238, 1986.
108. Rougier A, Lotte C, Maibach HI. The hairless rat: a relevant animal model to predict
in vivo percutaneous absorption in humans? J Invest Dermatol 88:577–581, 1987.
109. Rougier A, Lotte C, Dupuis D. An original predictive method for in vivo percutaneous
absorption studies. J Soc Cosmet Chem 38:397–417, 1987.
110. Rougier A, Rallis M, Krien P, Lotte C. In vivo percutaneous absorption: a key role
for stratum corneum/vehicle partitioning. Arch Dermatol Res 282:498–505, 1990.
111. Feldmann RJ, Maibach HI. Percutaneous penetration of steroids in man. J Invest
Dermatol 52:89–94, 1969.
112. Feldmann RJ, Maibach HI. Percutaneous penetration of some pesticides and herbi-
cides in man. Toxicol Appl Pharmacol 28:126–132, 1974.
113. Auton TR. Skin stripping and science: a mechanistic interpretation using mathematical
modelling of skin deposition as a predictor of total absorption. In: Scott RC, Guy RH,
Hadgraft J, Boddé HE, eds. Prediction of Percutaneous Penetration. Vol. 2. London:
IBC Technical Services, pp 558–576, 1990.
114. Knight AG. The activity of various topical griseofulvin preparations and the appear-
ance of oral griseofulvin in the stratum corneum. Br J Dermatol 91:49–55, 1974.
115. Dittmar W, Jovic N. Laboratory technique alternative to in vivo experiments for study-
ing the liberation, penetration and fungicidal action of topical antimycotic agents in
the skin, including ciclopirox olamine. Mykosen 30:326–342, 1986.
116. Pershing LK, Corlett J, Jorgensen C. In vivo pharmacokinetics and pharmacodynamics
of topical ketoconazole and miconazole in human stratum corneum. Antimicrob
Agents Chemother 38:90–95, 1994.
117. Pershing LK, Krueger GG. Human skin sandwich flap model for percutaneous ab-
sorption. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption. New York:
118. Parry GE, Dunn P, Shah VP, Pershing LK. Acyclovir bioavailability in human skin.
119. Michel C, Purmann T, Mentrup E, Seiller E, Kreuter J. Effect of liposomes on per-
cutaneous penetration of lipophilic materials. Int J Pharm 84:93–105, 1992.
120. Pershing LK, Silver BS, Krueger GG, Shah VP, Skelly JP. Feasibility of measuring
the bioavailability of topical betamethasone dipropionate in commercial formulations
using drug in skin and a skin blanching bioassay. Pharm Res 9:45–51, 1992.
121. Pershing LK, Lambert LD, Shah VP, Lam SY. Variability and correlation of chro-
mameter and tape-stripping methods with the visual skin blanching assay in quanti-
tative assessment of topical 0.05% betamethasone dipropionate bioavailability in hu-
mans. Int J Pharm 86:201–210, 1992.
122. Pershing LK, Corlett JL, Lambert LD, Poncelet CE. Circadian activity of topical
0.05% betamethasone dipropionate in human skin in vivo. J Invest Dermatol 102:
734–739, 1994.
123. Pershing LK, Lambert L, Wright ED, Shah VP, Williams RL. Topical 0.05% beta-
methasone dipropionate: pharmacokinetic and pharmacodynamic dose–response stud-
ies in humans. Arch Dermatol 130:740–747, 1994.
124. Tsai JC, Cappel MJ, Weiner ND, Flynn GL, Ferry J. Solvent effects on the harvesting
of stratum corneum from hairless mouse skin through adhesive tape stripping in vitro.
Int J Pharm 68:127–133, 1991.
125. Tsai JC, Weiner ND, Flynn GL, Ferry J. Properties of adhesive tapes used for stratum
corneum stripping. Int J Pharm 72:227–231, 1991.
126. Henn U, Surber C, Schweitzer A, Bieli E. D-Squame adhesive tapes for standardized
stratum corneum stripping. In: Brain KR, James VJ, Walters KA, eds. Prediction of
Percutaneous Penetration. Vol. 3b. Cardiff: STS Publishing, pp 477–481, 1993.
126a. FDA, Center for Drug Evaluation and Research. Draft-Guidance for Industry: Topical
Dermatological Drug Product NDAs and ANDAs—In Vivo Bioavailability, Bioequiv-
alence, In Vitro Release, and Associated Studies. Rockville, MD: Division of Bio-
equivalence, 1998.
126b. Surber C, Schwarb FP, Smith EW. Tape-stripping technique. In: Bronaugh RL, Mai-
bach HI, eds. Percutaneous Absorption: Drugs—Cosmetics—Mechanisms—Meth-
odology, 3rd ed. New York: Marcel Dekker, pp 395–409, 1999.
127. Surber C, Wilhelm KP, Hori M, Maibach HI, Guy RH. Optimization of topical ther-
apy: partitioning of drugs into stratum corneum. Pharm Res 7:1320–1324, 1990.
128. Surber C, Wilhelm KP, Berman D, Maibach HI. In vivo skin penetration of acitretin
in volunteers using three different sampling techniques. Pharm Res 10:1291–1294,
1993.
129. Laugier J–P, Surber C, Bun H, Geiger J–M, Wilhelm K–P, Durand A, Maibach HI.
Determination of acitretin in the skin, in the suction blister, and in plasma of human
volunteers after multiple oral dosing. J Pharm Sci 83:623–628, 1994.
130. Fry L, McMinn RMH. Topical methotrexate in psoriasis. Arch Dermatol 96:483–488,
1967.
131. Surber C, Itin P, Büchner S. Clinical controversy on the effect of topical ciclosporin
[sic] A: what is the target organ. Dermatology 185:242–245, 1992.
132. Schaefer H, Lamaud E. Standardization of experimental models. In: Shroot B, Schae-
fer H, eds. Skin Pharmacokinetics. Basel: Karger, pp 77–80, 1987.
133. Schaefer H, Stüttgen G, Zesch A, Schalla W, Gazith J. Quantitative determination of
percutaneous absorption of radiolabeled drugs in vitro and in vivo by human skin.
Curr Probl Dermatol 7:80–94, 1978.
134. Wojciechowski Z, Pershing LK, Huether S, Leonard L, Burton SA, Higuchi WI,
Krueger GG. An experimental skin sandwich flap on an independent vascular supply
for the study of percutaneous absorption. J Invest Dermatol 88:439–446, 1987.
135. Wester RC, Noonan PK. Topical bioavailability of a potential anti-acne agent (SC-
23110) as determined by cumulative excretion and areas under plasma concentration
time curves. J Invest Dermatol 70:92–94, 1978.
136. Wester RC, Noonan PK, Smeach S, Kosobud L. Pharmacokinetics and bioavailability
of intravenous and topical nitroglycerin in the rhesus monkey: estimate of percuta-
neous first-pass metabolism. J Pharm Sci 72:745–748, 1983.
137. Wester RC, Maibach HI. Cutaneous pharmacokinetics: 10 steps to percutaneous ab-
sorption. Drug Metab Rev 14:169–205, 1983.
138. Bucks DAW, McMaster JMA, Maibach HI, Guy RH. Bioavailability of topically ad-
ministered steroids: a ‘‘mass balance’’ technique. J Invest Dermatol 91:29–33, 1988.
139. Feldmann RJ, Maibach HI. Regional variation in percutaneous penetration of 14C
cortisol in man. J Invest Dermatol 48:181–183, 1967.
140. Maibach HI. In vivo percutaneous penetration of corticoids in man and unresolved
problems in their efficacy. Dermatologica 152(suppl 1):11–25, 1976.
141. Wester RC, Maibach HI. Relationship of topical dose and percutaneous absorption in
rhesus monkey and man. J Invest Dermatol 67:518–520, 1976.
142. Melendres JL, Bucks DAW, Camel E, Wester RC, Maibach HI. In vivo percutaneous
absorption of hydrocortisone: multiple-application dosing in man. Pharmacol Res 9:
1164–1167, 1992.
143. Wheeler DA, Moyler DA, Thirkettle JT. Instrumental colour assessment: some prac-
tical experience. J Soc Cosmet Chem 27:15–43, 1976.
144. Agache P, Girardot I, Bernengo JC. Optical properties of the skin. In: Leveque JL,
ed. Cutaneous Investigation in Health and Disease: Noninvasive Methods and Instru-
mentation. New York: Marcel Dekker, pp 241–274, 1989.
145. Weatherall IL, Coombs BD. Skin color measurements in terms of CIELAB color space
values. J Invest Dermatol 99:468–473, 1992.
146. el-Gammal S, Hoffmann K, Steiert P, Grassmüller J, Dirschka T, Altmeyer P. Objec-
tive assessment of intra- and inter-individual skin colour variability: an analysis of
human skin reaction to sun and UVB. In: Marks R, Plewig G, eds. The Environmental
Threat to the Skin. London: Martin Dunitz, pp 99–116, 1992.
147. Seitz JC, Whitemore CG. Measurement of erythema and tanning responses in human
skin using a tri-stimulus colorimeter. Dermatologica 177:70–75, 1988.
148. Clarys P, Buchet E, Barel AO. Evaluation of different topical vasodilatory products
with non-invasive techniques. In: Scott RC, Guy RH, Hadgraft J, Boddé HE, eds.
Prediction of Percutaneous Penetration. Vol. 2. London: IBC Technical Services, pp
46–59, 1991.
149. Wilhelm K–P, Surber C, Maibach HI. Quantification of sodium lauryl sulfate irritant
dermatitis in man: comparison of four techniques: skin color reflectance, transepider-
mal water loss, laser Doppler flow measurement and visual scores. Arch Dermatol
Res 281:293–295, 1989.
150. Queille–Roussel C, Duteil L, Padilla J–M, Poncet M, Czernielewski J. Objective
assessemnt of topical anti-inflammatory drug activity on experimentally induced
nickel contact dermatitis: comparison between visual scoring, colorimetry, laser Dopp-
ler velocimetry and transepidermal water loss. Skin Pharmacol 3:248–255, 1990.
151. Zhou J, Mark R, Stoudemayer T, Skar A, Lichten JL, Gabriel KL. The value of
multiple instrumental and clinical methods, repeated patch applications, and daily
evaluations for assessing stratum corneum changes induced by surfactants. J Soc
Cosmet Chem 42:105–128, 1991.
152. Lathi A, Kopola H, Harila A, Mullylä A, Hannuksela M. Assessment of skin erythema
by eye, laser Doppler flowmeter, spectroradiometer, two-channel erythema meter and
Minolta chromameter. Arch Dermatol Res 285:278–282, 1993.
153. Treffel P, Gabard B, Bieli E. Relationship between the in vitro diclofenac epidermal
level and the in vivo anti-inflammatory efficacy in the methylnicotinate test. In: Brain
KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 3b.
154. Clarys P, Barel AO, Gabard B. Objective evaluation of the skin blanching induced
by corticosteroids. Influence of corticosteroid concentration and body region. In: Brain
KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol. 3b.
154a. Schwarb FP, Smith EW, Haigh JM, Surber C. Analysis of chromameter results ob-
tained from corticosteroid-induced skin blanching assay: comparison of visual and
chromameter data. Eur J Pharmacol Biopharmacol 47:261–267, 1999.
154b. Demana PH, Smith EW, Walker RB, Haigh JM, Kanfer I. Evaluation of the proposed
FDA pilot dose–response methodology for topical corticosteroid bioequivalence test-
ing. Pharm Res 14:303–309, 1997.
154c. Smith EW, Walker RB, Haigh JM. Analysis of chromameter results obtained from
corticosteroid-induced blanching. I. Manipulation of data. Pharm Res 15:280–285,
1998.
155. Troilius A, Wardell K, Bornmyr S, Nilson GE, Ljunggren B. Evaluation of port wine
stain perfusion by laser Doppler imaging and thermography before and after argon
laser treatment. Acta Derm Venereol (Stockh) 72:6–10, 1992.
156. Quinn AG, McLelland J, Essex T, Farr PM. Quantification of contact allergic inflam-
mation: a comparison of existing methods with a scanning laser Doppler velocimeter.
Acta Derm Venereol (Stockh) 73:21–25, 1993.
157. Speight EL, Essex TJH, Farr PM. The study of plaques of psoriasis using a scanning
laser Doppler velocimeter. Br J Dermatol 128:519–524, 1993.
158. Guy RH, Tur E, Bugatto B, Gaebel C, Sheiner LB, Maibach HI. Pharmacodynamic
measurements of methyl nicotinate percutaneous absorption. Pharm Res 1:76–81,
1984.
159. Poelman MC, Piot B, Guyon M, Leveque JL. Assessment of topical non-steroidal
anti-inflammatory drugs. J Pharm Pharmacol 41:720–722, 1989.
160. Hori M, Ohtsuka S, Sunami M, Guy RH, Maibach HI. Cutaneous pharmacodynamics
of transdermally delivered isosorbide dinitrate. Pharm Res 7:1298–1301, 1990.
161. Treffel P, Gabard B. Feasibility of measuring the bioavailability of topical ibuprofen
in commercial formulations using drug content in epidermis and methyl nicotinate
skin inflammation assay. Skin Pharmacol 129:286–275, 1993.
162. Van Scott EJ. Experimental animal integumental models for screening potential der-
matologic drugs. In: Montagna W, Van Scott EJ, Stoughton RB, eds. Pharmacology
and the Skin. Advances in Biology of the Skin. New York: Appleton Century Croft,
pp 523–533, 1972.
163. Kligman LH, Kligman AM. The effect on rhino mouse skin of agents which influence
keratinization and exfoliation. J Invest Dermatol 73:354–358, 1979.
164. Mezick JA, Bhatia MC, Shea LM. Anti-acne activity of retinoids in the rhino mouse.
In: Maibach HI, Lowe NJ, eds. Models in Dermatology. Basel: S Karger, pp 59–63,
1985.
165. Bouclier M, Chatelus A, Ferracin J, Delain C, Shroot B, Hensby CN. Quantification
of epidermal histological changes induced by topical retinoids and CD271 in the rhino
mouse model using a standarized image analysis technique. Skin Pharmacol 4:65–73,
1991.
166. Meyer E, Smith EW, Haigh JM. Sensitivity of different areas of the flexor aspect of
the human forearm to corticosteroid-induced skin blanching. Br J Dermatol 127:379–
381, 1992.
167. Gilchrest BA, Soter NA, Hawks JLM. Histologic changes associated with ultraviolet-
A induced erythema in normal skin. J Am Acad Dermatol 9:213–219, 1983.
168. Woodbury RA, Kligman LH, Woodbury MJ, Kligman AM. Rapid assay of anti-in-
flammatory activity of topical corticosteroids by inhibition of a UVA-induced neutro-
phil inflitration in hairless mouse skin. I. The assay and its sensitivity. Acta Derm
Venereol (Stockh) 74:15–17, 1994.
169. Kligman LH. Rapid assay of anti-inflammatory activity of topical corticosteroids by
inhibition of a UVA-induced neutrophil infiltration in hairless mouse skin. II. Assess-
ment of name brand versus generic potency. Acta Derm Venereol (Stockh) 74:18–19,
1994.
170. Michaelsson G, Juhlin L, Vahlquist A. Effects of oral zinc and vitamin A in acne.
Arch Dermatol 133:31–136, 1977.
171. Consensus Report of the European Task Force on Atopic Dermatits. Severity scoring
of atopic dermatitis: the SCORAD Index. Dermatology 186:23–31, 1993.
172. Marks R, Barton SP, Shuttleworth D, Finlay AY. Assessment of disease progress in
psoriasis. Arch Dermatol 125:235–240, 1989.
173. Berardesca E, Vignoli GP, Farinelli N, Vignini M, Distante F, Rabbiosi G. Non-
invasive evaluation of topical calicipotriol versus clobetasol in the treatment of pso-
riasis. Acta Derm Venereol (Stockh) 74:302–304, 1994.
174. Ramsay B, Lawrence AM. Measurement of involved surface area in patients with
poriasis. Br J Dermatol 124:565–570, 1991.
175. Larnier C, Ortonne J–P, Venot A, Faivre B, Béani J–C, Thomas P, Brown TC,
Sendagorta E. Evaluation of cutaneous photodamage using a photographic scale. Br
J Dermatol 130:167–173, 1995.
176. Seitz JC, Spencer TS. Clinical research photography. In: Maibach HI, Lowe NJ, eds.
Models in Dermatology. Basel: S Karger, pp 63–70, 1989.
177. Fredriksson T, Pettersson U. Severe psoriasis—oral therapy with a new retinoid. Der-
matologica 157:238–244, 1978.
178. Wolska H, Jablonska S, Bounameaux Y. Etretinate in severe psoriasis. Results of a
double-blind study and maintenance therapy in pustilar psoriasis. J Am Acad Dermatol
9:883–889, 1983.
179. Geiger J–M, Czarnetzki BM. Acitretin (Ro 10-1670): overall evaluation of clinical
studies. Dermatologica 176:182–190, 1988.
180. Altmeyer PJ, Matthes U, Pawlak F, Hoffmann K, Frosch PJ, Ruppert P, Wassilew
SW, Horn T, Kreysel HW, Lutz G, Barth J, Rietzschel I, Joshi RK. Antipsoriatic effect
of fumaric acid derivatives. J Am Acad Dermatol 30:977–981, 1994.
181. Kimbrough–Green CK, Griffith CEM, Finkel LJ, Hamilton TA, Bulengo–Ransby SM,
Ellis CN, Voorhees JJ. Topical retinoic acid (tretinoin) for melasma in black patients.
Arch Dermatol 130:727–733, 1994.
182. Surber C, Lüdin E, Flückiger A, Dubach UC, Ziegler WH. Pain assessment after
intramuscular injection. Arzneimittelforschung 44:1389–1394, 1994.
183. Santos DJ, Juneja M, Bridenbaugh PO. A device for uniform testing of sensory neural
blockade during regional anesthesia. Anesth Analg 66:581–582, 1987.
184. Funk JO. Horizons in pharmacologic intervention in allergic contact dermatitis. J Am
Acad Dermatol 31:999–1014, 1994.
185. Meingasser JG, Stütz A. Immunosuppressive macrolides of the type FK 506: a novel
class of topical agents for treatment of skin diseases? J Invest Dermatol 98:851–855,
1992.
186. Hannuksela M. Propylene glycol promotes allergic patch test reactions. Boll Dermatol
Allergol Prof 2:40–44, 1987.
187. Camarasa JG, Serra–Baldrich E. Allergic contact dermatitis from acyclovir. Contact
Dermatitis 19:235–236, 1988.
188. Goh CL. Compound allergy to Spectraban 15 lotion and Zovirax cream. Contact
Dermatitis 22:61–62, 1990.
189. Gola M, Francalanci S, Brusi C, Lombardi P, Sertoli A. Contact sensitization to acy-
clovir. Contact Dermatitis 20:394–395, 1990.
190. Baes H, Van Hecke E. Contact dermatits from Zovirax cream. Contact Dermatits 23:
200–201, 1990.
191. Sidwell RW, Huffman JH, Schafer TW, Shipman C. Influence of vehicle on topical
efficacy of 2-acetylpyridine thiosemicarbazone and related derivates on in vivo type
2 herpes simplex virus infections. Chemotherapy 36:58–69, 1990.
192. Weiner N, Williams N, Birch G, Ramachandran C, Shipman C, Flynn G. Topical
delivery of liposomally encapsulated interferon evaluated in a cutaneous herpes guinea
pig model. Antimicrob Agents Chemother 33:1217–1221, 1989.
193. Imanidis G, Song W, Lee PH, Su MH, Kern ER, Higuchi WI. Estimation of skin
target site acyclovir concentrations following controlled (trans)dermal drug delivery
in topical and systemic treatment of cutaneous HSV-1 infections in hairless mice.

Pharm Res 11:1035–1041, 1994.
194. Lee PH, Su MH, Kern ER, Higuchi WI. Novel animal model for evaluating topical
efficacy of antiviral agents: flux versus efficacy correlations in acyclovir treatment of
cutaneous herpes simplex virus type 1 (HSV-1) infections in hairless mice. Pharm
Res 9:979–989, 1992.
195. Gonsho A, Imanidis G, Vogt P, Kern ER, Tsuge H, Su MH, Choi SH, Higuchi WI.
Controlled (trans)dermal delivery of an antiviral agent (acyclovir). I: An in vivo an-
imal model for efficacy evaluation in cutaneous HSV-1 infections. Int J Pharm 65:
183–194, 1990.
196. Lee PH, Su MH, Ghanem AH, Inamori T, Kern ER, Higuchi WI. An application of
the C* concept in predicting the topical efficacy of finite dose acyclovir in the treat-
ment of cutaneous HSV-1 infections in hairless mice. Int J Pharm 93:139–152, 1993.
197. Price RW. Neurobiology of human herpes virus infections. Crit Rev Clin Neurobiol
2:61–123, 1986.
198. FDA, Center of Drug Evaluation and Research. Draft guidance for the performance
of a bioequivalence study for topical antifungal products. Rockville, MD: Division of
Bioequivalence.
199. Aly R, Maibach HI. Ciclopirox olamine lotion 1%: bioequivalence to ciclopirox ol-
amine cream 1% and clinical efficacy in tinea pedis. Clin Ther 11:290–303, 1989.
200. Eady EA, Cove JH. Topical antibiotic therapy; current status and future prospects.
Drugs Exp Clin Res 16:423–433, 1990.
201. Held JL, Kalb RE, Ruszkowski AM, DeLeo V. Allergic contact dermatitis from bac-
itracin. J Am Acad Dermatol 17:592–594, 1987.
202. Villiger JW, Robertson WD, Kanji K. A comparison of the new topical antibiotic
mupirocin (Bactroban) with oral antibiotics in the treatment of skin infections in
general practice. Curr Med Res Opin 10:339–345, 1986.
203. Mertz PM, Marshall DA, Eaglestein WH. Topical mupirocin treatment of impetigo is
equal to oral erythromycin therapy. Arch Dermatol 125:1069–1073, 1989.
204. Schwanitz HJ, Macher E. Internal versus topical tetracycline therapy in acne. Z Haut-
krankh 59:1515–1521, 1984.
205. Eady EA, Holland KT, Cunliffe WJ. Should topical antibiotics be used for the treat-
ment of acne vulagris? Br J Dermatol 107:235–246, 1982.
206. Buckley MM, Benfiled P. Eutectic lidocaine/prilocaine cream. Drugs 46:126–151,
1993.
207. Foldvari M. In vitro cutaneous and percutaneous delivery and in vivo efficacy of
tetracaine from liposomal and conventional vehicles. Pharm Res 11:1593–1598, 1994.
208. Kushla GP, Zatz JL. Evaluation of noninvasive method for monitoring percutaneous
absorption of lidocaine in vivo. Pharm Res 7:1033–1037, 1990.
209. Kushla GP, Zatz JL, Mills OH, Berger RS. Noninvasive assessment of anesthetic
activity of topical lidocaine formulations. J Pharm Sci 82:1118–1122, 1993.
210. King RB. Topical aspirin in chloroform and the relief of pain due to herpes zoster
and postherapeutic neuralgia. Arch Neurol 50:1046–1053, 1993.
211. Brookes FL, Hugo WB, Denyer SP. Transformation of betamethasone-17-valerate by
skin microflora. J Pharm Pharmacol 34:61P, 1982.
212. Abu Shamat MS, Beckett AH. Glyceryl trinitrate: metabolism by intestinal flora. J
Pharm Pharmacol 35:71P, 1983.
213. Denyer SP, Hugo WB, O’Brien M. Metabolism of glyeryl trinitrate by skin staphy-
lococci. J Pharm Pharmacol 36:61P, 1984.
214. Denyer SP, Guy RH, Hadgraft J, Hugo WB. The microbial degradation of topically
applied drugs. Int J Pharm 26:89–97, 1985.
215. Martin RJ, Denyer SP, Hadgraft J. Skin metabolism of topically applied compounds.
Int J Pharm 39:23–32, 1987.
216. Shaw JE, Chandrasekaran SK. Controlled topical delivery of drugs for systemic ac-
tion. Drug Metab Rev 8:223–233, 1978.
217. Samuelov Y, Donbrow M, Friedman M. Effect of pH on salicylic acid permeation
through ethyl cellulose–PEG 4000 films. J Pharm Pharmacol 31:120–121, 1979.
218. Jack L, Cameron BD, Scott RC, Hadgraft J. In vitro percutaneous absorption of
salicylic acid; effect of vehicle pH. In: Scott RC, Guy RH, Hadgraft J, Boddé HE,
eds. Prediction of Percutaneous Penetration. Vol. 2. London: IBC Technical Services,
pp 515–518, 1991.
219. Özer AY, Farivar M, Hincal AA. Temperature- and pH-sensitive liposomes. Eur J
Pharm Biopharm 39:97–101, 1993.
220. Tregear RT. Physical Functions of Skin. London: Academic Press, 1966.
221. Wheatley VR. Proc Sci Soc Toilet Goods Assoc 39:25, 1963.
222. Wheatley VR, Flech P, Esoda EGJ, Coon WM, Mandol R. Studies of the chemical
composition of the horny layer lipids. J Invest Dermatol 43:395–405, 1964.
223. Wilkenson DI. Variability in composition of surface lipids. J Invest Dermatol 54:339–
343, 1969.
224. Cheng Y–H, Hosoya O, Sugibayashi K, Morimoto Y. Effect of skin surface lipid on
skin permeation of lidocaine from pressure sensitive adhesives. Biol Pharm Bull 17:
1640–1644, 1994.
225. Cummings EG. Temperature and concentration effects on penetration of N-octyamine
through human skin in situ. J Invest Dermatol 53:64–70, 1969.
226. Katz M, Poulsen BJ. Absorption of drugs through skin. In: Brodie BB, Gillette JR,
eds. Handbook of Experimental Pharmacology, 28. New York: Springer, pp 103–174,
1971.
227. Rothenborg HW, Menné T, Sjølin KE. Temperature dependent primary irritant der-
matitis from lemon perfume. Contact Dermatitis 3:37–48, 1977.
228. Emilson A, Lindberg M, Forslind B. The temperature effect on in vitro penetration
of sodium lauryl sulfate and nickel chloride through human skin. Acta Derm Venereol
(Stockh) 73:203–207, 1993.
229. James M, Marty J–P, Wepierre J. Diffusion localisée des substances absorbées par
voie percutanée. CR Acad Sci (Paris) 278(Ser. D):2063–2066, 1974.
230. James M, Wepierre J. Absorption percutanée de la thyroxine 125I et de la triiodothy-
ronine 125I chez le rat in vivo. Ann Pharm Fr 32:633–640, 1974.
231. Guy RH, Maibach HI. Drug delivery to local subcutaneous structures following top-
ical administration. J Pharm Sci 72:1375–1380, 1983.
232. Marty J–P, Guy RH, Maibach HI. Percutaneous penetration as a method of delivery
to muscle and other tissue. In: Bronaugh RL, Maibach HI, eds. Percutaneous Ab-
sorption: Mechanism—Methodology—Drug Delivery. New York: Marcel Dekker, pp
511–529, 1989.
233. Singh P, Roberts M. Deep tissue penetration of bases and steroids after dermal ap-
plication in rat. J Pharm Pharmacol 46:956–964, 1994.
234. Benowitz NL, Jacob P, Olsson P, Johansson CJ. Intraveneous nicotine retards trans-
dermal absorption of nicotine: evidence of blood flow-limited percutaneous absorp-
tion. Clin Pharmacol Ther 52:223–230, 1992.
235. Benowitz NL, Jacob P, Jones RT, Rosenberg J. Interindividual variability in metab-
olism and cardiovascular effects of nicotine in man. J Pharmacol Exp Ther 221:368–
372, 1982.
236. Monteiro–Riviere NA. Altered epidermal morphology secondary to lidocaine ionto-
phoresis: in vivo and in vitro studies in porcine skin. Fundam Appl Toxicol 15:174–
185, 1990.
237. Riviere JE, Sage BH, Monteiro–Riviere NA. Transdermal lidocaine ionophoresis in
isolated perfused porcine skin. J Toxicol Cutan Ocul Toxicol 8:493–504, 1990.
238. Auclair F, Besnard M, Dupont C, Wepierre J. Importance of blood flow to the local
distribution of drugs after percutaneous absorption in the bipediculated dorsal flap of
the hairless rat. Skin Pharmacol 4:1–8, 1991.
239. Nakashima E, Noonan PK, Benet LZ. Transdermal bioavailability and first-pass skin
matabolism: a preliminary evaluation with nitroglycerin. J Pharmacokinet Biopharm
15:423–437, 1987.
240. Upton RN. Regional pharmacokinetics. I. Physiological and physicochemical basis.
Biopharm Drug Dispos 11:647–662, 1990.
241. Turpeinen M, Mashkilleyson N, Bjorksten F, Salo OP. Percutaneous absorption of
hydrocortisone during exacerbation and remission of atopic dermatitis in adults. Acta
Derm Venereol (Stockh) 68:331–335, 1988.
242. Turpeinen M. Influence of age and severity of dermatitis on percutaneous absorption
of hydrocortisone in children. Br J Dermatol 118:517–522, 1988.
243. Turpeinen M. Absorption of hydrocortisone from the skin reservoir in atopic derma-
titis. Br J Dermatol 124:358–360, 1991.
244. Turpeinen M, Lehtokoski–Lehtiniemi E, Leisti S, Salo OP. Percutaneous absorption
during and after the acute phase of dermatitis in children. Pediatr Dermatol 5:276–
279, 1988.
245. Wester RC, Bucks DAW, Maibach HI. Percutaneous absorption of hydrocortisone in
psoriatic patients and normal volunteers. J Am Acad Dermatol 8:645–647, 1983.
246. Wester RC, Mobayen M, Ryatt K, Bucks D, Maibach HI. In vivo percutaneous ab-
sorption of dithranol on psoriatic and normal volunteers. In: Farber E, ed. Psoriasis.
Amsterdam: Elsevier, pp 429, 1987.
247. Schaefer H, Zesch A, Stüttgen G. Skin Permeability. Berlin: Springer, 1982.
248. Erlanger M, Martz G, Ott F, Storck H, Rider J, Kessler S. Cutaneous absorption and
urinary excretion of 6-14C-5-fluorouracil following application in an ointment to
healthy and diseased human skin. Dermatologica (suppl 1):7–14, 1970.
249. Foreman M, Clanchan I, Kelly I. The diffusion of nandrolone through occluded and
non-occluded human skin. J Pharm Pharmacol 30:152–157, 1978.
250. Shani J. Skin penetration of minerals in psoriasis and guinea pig bathing in hypertonic
salt solutions. Pharmacol Res Commun 17:501–511, 1985.
251. Wang JCT, Patel BG, Ehmann CW, Lowe N. The release and percutaneous permeation
of anthralin products, using clinically involved and uninvolved psoriatic skin. J Am
Acad Dermatol 16:812–821, 1987.
252. Juhlin L, Hägglund G, Evers H. Absorption of lidocaine and prilocaine after appli-
cation of eutectic mixture of local anesthetics (EMLA) on normal and diseased skin.
Acta Derm Venereol (Stockh) 69:18–22, 1989.
253. Maibach HI, Feldmann RJ, Milby TH, Serat WF. Regional variation in percutaneous
penetration in man. Arch Environ Health 23:208–211, 1971.
254. Elias PM, Cooper ER, Korc A, Brown BE. Percutaneous transport in relation to
stratum corneum structure and lipid composition. J Invest Dermatol 76:297–301,
1981.
255. Scott RC, Corrigan MA, Smith F, Mason H. The influence of skin structure on per-
meability: an intersite and interspecies comparison with hydrophilic penetrants. J In-
vest Dermatol 96:921–925, 1991.
256. Ebihara A, Fujimura A, Ohashi K, Shiga T, Kumagai Y, Nakashima H, Kotegawa T.
Influence of application site of a new transdermal clonidine, M-5041T, on its phar-
makokinetics and pharmacodynamics in healthy subjects. J Clin Pharmacol 33:1188–
1191, 1993.
257. Hopkins K, Aarons L, Rowland M. Absorption of clonidine from transdermal thera-

peutic system when applied to different body sites. In: Weber MA, Mathias CJ, eds.
Mild Hypertension. Darmstadt: Steinkopf. pp 143–147, 1984.
258. Hansen MS, Woods SL, Willis RE. Relative effectiveness of nitroglycerin ointment
according to site of application. Heart Lung 8:716–720, 1979.
259. Moe G, Armstrong PW. Influence of skin site on bioavailability of nitroglycerin oint-
ment in congestive heart failure. Am J Med 81:765–770, 1986.
260. Noonan PK, Wester RC. Percutaneous absorption of nitroglycerin. J Pharm Sci 69:
385–386, 1980.
261. Shah AK, Wei G, Lamman RC, Bhargava VO, Weir SJ. Percutaneous absorption of
ketoprofen from different anatomical sites in man. Pharm Res 13:168–172, 1996.
262. Stoughton RB. Percutaneous absorption of drugs. Annu Rev Pharmacol Toxicol 29:
55–69, 1989.
263. Scheuplein RJ. Mechanism of percutaneous absorption. II. Transient diffusion and the
relative importance of various routes of skin penetration. J Invest Dermatol 48:79–
88, 1967.
264. Albery WJ, Hadgraft J. Percutaneous absorption: in vivo experiments. J Pharm Phar-
macol 31:140–147, 1979.
265. Pratzel HG. Iontophorese. Berlin: Springer, 1987.
266. Mackee GM, Sulzberger MB, Herrmann F, Bare RL. Histologic studies on percuta-
neous penetration with special reference to the effects of vehicles. J Invest Dermatol
6:43–61, 1945.
267. Tregear RT. Relative penetrability of hair follicles and epidermis. J Physiol 156:307–
313, 1961.
268. Grasso P. Some aspects of the role of skin appendages in percutaneous absorption. J
Soc Cosmet Chem 22:523–534, 1971.
Hadgraft J, eds. Prediction of Percutaneous Penetration. Vol. 1. London: IBC Tech-
nical Services, pp 163–173, 1990.
271. Scheuplein RJ. Mechanism of percutaneous absorption. J Invest Dermatol 45:334–
346, 1965.
272. Scheuplein RJ, Blank IH, Brauner GJ, MacFarlane DJ. Percutaneous absorption of
steroids. J Invest Dermatol 52:63–70, 1969.
273. Shelly WB, Melton FM. Studies on absorption through normal human skin. FASEB
6:199–200, 1947.
274. Behl CR, Bellantone NH, Flynn GL. Influence of age on percutaneous absorption of
drug substances. In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption: Mech-
anism—Methodology—Drug Delivery. New York: Marcel Dekker, pp 183–212,
1985.
275. Suzuki M, Asaba K, Komatsu H, Mochizuka M. Autoradiographic study on percu-
taneous absorption of several oils useful for cosmetics. J Soc Cosmet Chem 29:265–
282, 1978.
276. Nicolau G, Baughman RA, Tonelli A, McWilliams W, Schiltz J, Yacobi A. Deposition
of viprostol (a synthetic PGE2 vasodilator) in the skin following topical administration
to laboratory animals. Xenobiotica 17:1113–1120, 1987.
277. Borelli, Metzger M. Fluorescenzmikroskopische untersuchungen über die perkutane
penetration fluorescierender stoffe. Hautarzt 8:261–266, 1957.
278. Kao J, Hall J, Helman G. In vitro percutaneous absorption in mouse skin: influence
of skin appendages. Toxicol Appl Pharmacol 94:93–103, 1988.
279. Wallace SM, Barnett G. Pharmacokinetic analysis of percutaneous absorption: evi-

dence of parallel penetration pathways for methotrexate. J Pharmacokinet Biopharm
6:315–325, 1978.
280. Keister JC, Kasting GB. The use of transient diffusion to investigate transport path-
ways through skin. J Control Release 4:111–117, 1986.
281. Corroller M, Didry JR, Siou G, Wepierre J. Sebaceous accumulation of linoleic acid
following topical application in the hairless rat its mathematical treatment. In: Shroot
B, Schaefer H, eds. Skin Pharmacokinetics. Basel: Karger, pp 111–120, 1987.
282. Illel B, Schaefer H, Wepierre J, Doucet O. Follicles play an important role in per-
cutaneous absorption. J Pharm Sci 80:424–427, 1991.
283. Illel B, Schaefer H. Transfollicular percutaneous absorption. Acta Derm Venereol
(Stockh) 68:427–457, 1988.
284. Behl CR, Wittkowsky A, Barett M, Pierson CL, Flynn GL. Technique for preparing
appendage-free skin (scar) on hairless mouse. J Pharm Sci 70:835–837, 1981.
285. Hueber F, Schaefer H, Wepierre J. Role of transepidermal and transfollicular routes
in percutaneous absorption of steroids: in vitro studies on human skin. Skin Pharmacol
7:237–244, 1994.
286. Hueber F, Besnard M, Schaefer H, Wepierre J. Percutaneous absorption of estradiol
and progesterone in normal and appendage-free skin of the hairless rat: lack of im-
portance of nutritional blood flow. Skin Pharmacol 7:245–256, 1994.
287. Hueber F, Wepierre J, Schaefer H. Role of transepidermal and transfollicular routes
in percutaneous absorption of hydrocortisone and testosterone: in vivo study on hair-
less rat. Skin Pharmacol 5:99–107, 1992.
288. Mukherji E, Millenbaugh NJ, Au J. Percutaneous absorption of 2⬘,3⬘-dideoxyinosine
in rats. Pharm Res 11:809–815, 1994.
289. Bamba FL, Wepierre J. Role of appendageal pathway in the percutaneous absorption
of pyridostigmine bromide in various vehicles. Eur J Drug Metab Pharmacokinet 18:
339–348, 1993.
290. Tur E, Maibach HI, Guy RI. Percutaneous penetration of methyl nicotinate at three
anatomic sites: evidence for appendageal contribution to transport? Skin Pharmacol
4:230–234, 1991.
291. Rolland A, Wagner N, Chatelus A, Shroot B, Schaefer H. Site-specific drug delivery
to pilosebaceous structures using polymeric microspheres. Pharm Res 10:1738–1744,
1993.
292. Chan SY, Po ALW. Prodrugs for dermal delivery. Int J Pharm 55:1–16, 1989.
293. Kao J, Carver MP. Cutaneous metabolism of xenobiotics. Drug Metab Rev 22:363–
410, 1990.
294. Lodén M. The in vitro hydrolysis of di-isopropyl fluorophosphate during penetration
through human full-thickness skin and isolated epidermis. J Invest Dermatol 85:335–
339, 1985.
295. Potts RO, Guy RH. A predictive algorithm for skin permeability: the effects of mo-
lecular size and hydrogen bond activity. Pharm Res 12:1628–1633, 1995.
296. Hadgraft J. Theoretical aspects of metabolism in the epidermis. Int J Pharm 4:229–
239, 1980.
297. Guy RH, Hadgraft J, Maibach HI. A pharmacokinetic model for percutaneous ab-
sorption. Int J Pharm 11:119–129, 1982.
298. Tojo K, Valia KH, Chotani G, Chien YW. Long-term permeation kinetics of estradiol:
(IV) a theroretical approach to the simultaneous skin permeation and bioconversion
of estradiol. Drug Dev Ind Pharm 11:1175–1193, 1985.
299. Potts RO, McNeill SC, Desbonnet CR, Wakshull E. Transdermal drug transport and
metabolism: II. The role of competing kinetic events. Pharm Res 6:119–124, 1989.
300. Sloan KB. Prodrugs: Topical and Ocular Drug Delivery. New York: Marcel Dekker,
1992.
301. Bodor N, Kaminski JJ, Selk S. Soft drugs. I. Labile quaternary ammonium salts as
soft antibacterials. J Med Chem 23:469–474, 1980.
302. Roskos KV, Maibach HI, Guy RH. The effect of aging on percutaneous absorption
in man. J Pharmacokinet Biopharm 17:617–630, 1989.
303. Harvell JD, Maibach HI. Percutaneous absorption and inflammation in aged skin: a
review. J Am Acad Dermatol 31:1015–1021, 1994.
304. Micalis S, Bhatt RH, Distefano S, Caltabiano L, Cook B, Fischer J, Solomon LM,
West DP. Evaluation of transdermal theophylline pharmacokinetics in neonates. Phar-
macotherapy 13:386–390, 1993.
305. Barrett DA, Rutter N. Transdermal delivery and the premature neonate. Crit Rev Ther
Drug Carrier Syst 11:1–30, 1994.
306. Weigand DA, Gaylor JR. Irritant reaction in Negro and Caucasian skin. South Med
J 67:548–551, 1974.
307. Johnson LC, Corah NL. Racial differences in skin resistance. Science 139:766–767,
1960.
308. Reinertson RP, Wheatley VR. Studies on the chemical composition of human epider-
mal lipids. J Invest Dermatol 32:49–59, 1959.
309. Berardesca E, Maibach HI. Racial differences in sodium lauryl sulphate induced cu-
taneous irritation: black and white. Contact Dermatitis 17:12–17, 1987.
310. Williams RL, Thakker KM, John V, Lin ET, Gee WL, Benet LZ. Nitroglycerin ab-
sorption from transdermal systems: formulation effects and metabolic concentrations.
Pharm Res 8:744–749, 1991.
311. Levy G. The clay feet of bioequivalence testing. J Pharm Pharmacol 47:975–977,
1995.
312. Hauck WW, Anderson S. Measuring switchability and prescribability: when is average
bioequivalence sufficient? J Pharmacokinet Biopharm 22:551–564, 1994.
313. Cornell RC, Stoughton RB. Correlation of the vasoconstriction assay and clinical
activity in psoriasis. Arch Dermatol 121:63–67, 1985.
314. Ferreira LAM, Seiller M, Grossiord JL, Marty JP, Wepierre J. Vehicle influence on
in vitro release and percutaneous absorption of glucose: role of W/O/W multiple
emulsion. Proc Int Symp Control Release Bioact Mater 21:453–454, 1994.
315. Ferreira LAM, Doucet J, Seiller M, Grossiord JL, Marty JP, Wepierre J. In vitro
percutaneous absorption of metronidazole and glucose: comparison of O/W, W/O/W
and W/O systems. Int J Pharm 121:169–179, 1995.
316. Friedman DI, Schwarz JS, Weisspapir M. Sub-micron emulsion vehicle for enhanced
transdermal delivery of steroidal antiinflammatory drugs. J Pharm Sci 84:324–329,
1995.
317. Kondo S, Yamanaka C, Sugimoto I. Enhancement of transdermal delivery by super-
fluous thermodynamic potential. III. Percutaneous absorption of nifedipine in rats. J
Pharmacobiodyn 10:743–749, 1987.
318. Megrab NA, Williams AC, Barry BW. Oestradiol permeation through human skin and
Silastic membrane: effects of propylene glycol and supersaturation. J Control Release
36:277–294, 1995.
319. Sato K, Sugibayashi K, Morimoto Y. Effect of mode of action of aliphatic esters on
the in-vitro skin penetration of nicorandil. Int J Pharm 43:31–40, 1988.
320. Yamane MA, Williams AC, Barry BW. Effects of terpenes and oleic acid as skin
penetration enhancers towards 5-fluorouracil as assessed with time; permeation, par-
titioning and differential scanning calorimetry. Int J Pharm 116:237–251, 1995.
321. Higuchi T. Physical chemical analysis of percutaneous absorption process from creams
322. Reifenrath WG, Hawkins GS, Kurtz MS. Percutaneous penetration and skin retention
of topically applied compounds: an in vitro–in vivo study. J Pharm Sci 80:526–532,
1991.
323. Surber C, Wilhelm K–P, Maibach HI. In vitro skin pharmacokinetics of acitretin:
percutaneous absorption studies in intact and modified skin of three different species.
J Pharm Pharmacol 43:836–840, 1991.
324. Elias PM. Epidermal lipids, membranes, and keratinization. Int J Dermatol 2:1–19,
1981.
325. Potts RO, Guy RH. Predicting skin permeability. Pharm Res 9:663–669, 1992.
326. Pugh WJ, Roberts MS, Hadgraft J, Degim T. Relationship between H-bonding of
penetrants to stratum corneum lipids and diffusion within the stratum corneum. In:
Brain KR, James VJ, Walters KA, eds. Prediction of Percutaneous Penetration. Vol.
4b. Cardiff: STS Publishing, pp 48–51, 1996.
327. Stoughton RB. Are generic formulations equivalent to trade name topical glucocor-
ticoids? Arch Dermatol 123:1312–1314, 1987.
328. Poulsen BJ, Young E, Coquilla V, Katz M. Effect of topical vehicle composition on
the in vitro release of fluocinolone acetonide and its acetate ester. J Pharm Sci 57:
928–933, 1968.
329. Ostrenga J, Steinmetz C, Poulsen B. Significance of vehicle composition I: relation-
ship between topical vehicle composition, skin penetrability, and clinical efficacy. J
Pharm Sci 60:1175–1179, 1971.
330. Twist JN, Zatz JL. Influence of solvents on paraben permeation through idealized skin
model membranes. J Soc Cosmet Chem 37:429–444, 1986.
331. Yalkowsky SH, Valvani SC, Amidon GL. Solubility of nonelectrolytes in polar sol-
vents IV: nonpolar drugs in mixed solvents. J Pharm Sci 65:1488–1494, 1976.
332. Hadgraft J, Hadgraft JW, Sarkany I. The effect of glycerol on percutaneous absorption
of methyl nicotinate. Br J Dermatol 87:30–36, 1972.
333. Lippold BC, Schneemann H. The influence of vehicles on the local bioavailability of
betamethasone-17-benzoate from solution and suspension-type ointments. Int J Pharm
22:31–43, 1984.
334. Dugard PH, Scott RC. A method of predicting percutaneous absorption rates from
vehicle to vehicle: an experimental assessment. Int J Pharm 28:219–227, 1986.
335. Goodman M, Barry BW. Lipid–protein-partitioning (LPP) theory of skin enhancer
activity: finite dose technique. Int J Pharm 57:29–40, 1989.
336. Lamaud E, Schalla W, Schäfer H. Effect of repeated rubbing on the in vivo penetration
of hydrocortisone. In: Schäfer H, Schalla W, Shroot B, eds. Advances in Skin Phar-
macology and Skin Pharmacokinetics; Nice: Centre International de Recherche Der-
matologiques. pp 23–24, 1986.
337. McMaster J, Maibach HI, Wester RC, Bucks DAW. Does rubbing enhance in vivo
dermal absorption? In: Bronaugh RL, Maibach HI, eds. Percutaneous Absorption:
Mechanism—Methodoloy—Drug Delivery. New York: Marcel Dekker, pp 359–361,
1985.
338. Idson B. Percutaneous absorption. J Pharm Sci 64:901–924, 1975.
339. Garb J. Nevus verrucosus unilateris cured with podophyllin ointment. Arch Dermatol
81:606–609, 1960.
340. Scholtz JR. Topical therapy of psoriasis with fluocinolone acetonide. Arch Dermatol
84:1029–1030, 1961.
341. Sulzberger MB, Witten VH. Thin pliable plastic films in topical dermatologic therapy.
Arch Dermatol 84:1027–1028, 1961.
342. McKenzie AW. Percutaneous absorption of steroids. Arch Dermatol 86:91–94, 1962.
343. McKenzie AW, Stoughton RB. Method for comparing percutaneous absorption of
steroids. Arch Dermatol 86:608–610, 1962.
344. Guy RH, Bucks DAW, McMaster JR, Villaflor DA, Roskos KV, Hinz RS, Maibach
HI. Kinetics of drug absorption across human skin in vivo: developments in meth-
odology. Skin Pharmacol 1:70–76, 1988.
345. Kim HO, Wester RC, McMaster JA, Bucks DAW, Maibach HI. Skin absorption from
patch test systems. Contact Dermatitis 17:178–180, 1987.
346. Auton TR, Westhead DR, Woollen BH, Scott RC, Wilks MF. A physiologically based
mathematical model of dermal absorption in man. Hum Exp Toxicol 13:51–60, 1994.
347. Johnson R, Nusbaum BP, Horwitz SN, Frost P. Transfer of topically applied tetra-
cycline in various vehicles. Arch Dermatol 119:660–663, 1983.
348. Haupenthal H, Heilmann P, Moll F. Kühlsalben. Krankenhauspharm 15:671–674,
1994.
349. Reifenrath WG, Robinson PB. In vitro skin evaporation and penetration characteristics
of mosquito repellents. J Pharm Sci 71:1014–1018, 1982.
350. Hawkins GS, Reifenrath WG. Influence of skin source, penetration cell fluid, and
partition coefficient on in vitro skin penetration. J Pharm Sci 75:378–381, 1986.
351. Reifenrath WG, Spencer TS. Evaporation and penetration from skin. In: Bronaugh
RL, Maibach HI, eds. Percutaneous Absorption: Mechanism—Methodolgy—Drug
Delivery. New York: Marcel Dekker, pp 313–334, 1989.
352. Coldman MF, Poulsen BJ, Higuchi T. Enhancement of percutaneous absorption by
the use of volatile:nonvolatile systems as vehicles. J Pharm Sci 58:1098–1102, 1969.
353. Chiang C–M, Flynn GL, Weiner ND, Szpunar GJ. Bioavailability assessment of top-
ical delivery systems: effect of vehicle evaporation upon in vitro delivery of minoxidil
from solution formulations. Int J Pharm 55:229–236, 1989.
354. Kemken J, Ziegler A, Müller BW. Influence of supersaturation on the pharmacody-
namic effect of bupranolol after dermal administration using microemulsions as ve-
hicle. Pharm Res 9:554–558, 1992.
355. de Carvalho M, Falson–Rieg F, Eynard I, Rojas J, Lafforgue C, Hadgraft J. Changes
in vehicle composition during skin permeation studies. In: Brain KR, James VJ, Wal-
ters KA, eds. Prediction of Percutaneous Penetration. Vol. 3b. Cardiff: STS Publishing,
pp 251–254, 1993.
356. Guy RH, Hadgraft J. Percutaneous absorption kinetics of topically applied agents
liable to surface loss. J Soc Cosmet Chem 35:103–113, 1984.
357. Flynn GL. Topical drug absorption and topical pharmaceutical systems. In: Banker
GS, Rhodes CT, eds. Modern Pharmaceutics. New York: Marcel Dekker, pp 263–
325, 1990.
358. Williams PL, Brooks JD, Inman AO, Monteiro–Riviere NA, Riviere JE. Determina-
tion of physicochemical properties of phenol, p-nitrophenol, acetone and ethanol rel-
evant to quantitating their percutaneous absorption in porcine skin. Res Commun
Chem Pathol Pharmacol 83:61–75, 1994.
359. Polano MK. Topical Skin Therapeutics. Edinburgh: Churchill Livingstone, 1984.
360. Schlagel CA, Sanborn EC. The weights of topical preparations required for total and
partial body inuction. J Invest Dermatol 42:253–256, 1964.
361. Griffiths WAD, Wilkinson JD. Topical therapy. In: Rook A, Wilkinson DS, Ebling
FJG, Champion RH, Burton JL, eds. Textbook of Dermatology. London: Blackwell
Scientific, pp 3037–3084, 1992.
362. Habif TP. Clinical Dermatology: A Color Guide to Diagnosis and Therapy. St. Louis:
CV Mosby, 1990.
363. Long CC. The rule of hand: 4 hand areas = 2 FTU 1 g. Arch Dermatol 128:1129–
1130, 1992.
364. Lutz UL, Weirich E. Untersuchungen zur hautflächen-Dosis von Dermatopharma-
kotherapeutika. Z Hautkr 50:753–769, 1975.
365. Willis I, Kligman AM. Evaluation of sunscreen by human assay. J Soc Cosmet Chem
20:639–651, 1969.
366. Lynfield YL, Schechter S. Choosing and using a vehicle. J Am Acad Dermatol 10:
56–59, 1984.
367. Kaidberg KH, Kligman AM. Laboratory methods for appraising the efficacy of sun-
screens. J Soc Cosmet Chem 29:525–536, 1978.
368. FDA. Sunscreen drug products for over-the counter human use. Fed Reg 43:38206–
38269, 1978.
369. Bech–Thomsen N, Wulff HC. Sunbather’s application of sunscreen is probably in-
adequate to obtain the sun protection factor assigned to the preparation. Photodermatol
Photoimmunol Photomed 9:242–244, 1993.
370. Loesch H, Kaplan DL. Pitfalls in sunscreen application. Arch Dermatol 130:665–666,
1994.
371. Shah VP, Behl CR, Flynn GL, Higuchi WI, Schaefer H. Principles and criteria in the
development and optimization of topical therapeutic products. J Pharm Sci 81:1051–
1054, 1992.
372. Ferry JJ, Shepard JH, Szpunar GJ. Relationship between contact time of applied dose
and percutaneous absorption of minoxidil from topical solution. J Pharm Sci 79:483–
486, 1990.
373. Täuber U, Matthes H. Percutaneous absorption of methylprednisolone aceponate after
single and multiple dermal application as ointment in male volunteers. Arzneimittelf-
forschung 42:1122–1124, 1992.
374. Gao HY, Li Wan Po A. Topical formulations of fluocinolone acetonide. Are creams,
gels and ointments bioequivalent and does dilution affect activity? Eur J Clin Phar-
macol 46:71–75, 1994.
375. Wester RC, Melendres J, Sedik L, Maibach HI. Percutaneous absorption of Azone
following single and multiple doses to human volunteers. J Pharm Sci 83:124–125,
1994.
376. Carslaw HS, Jaeger JC. Conduction of Heat in Solids. London: Oxford University
Press, 1959.
377. Franz TJ. Percutaneous absorption. On the relevance of in vitro data. J Invest Der-
matol 64:190–195, 1975.
378. Franz TJ. The finite dose techniques as a valid in vitro model for the study of per-
cutaneous absorption in man. Curr Probl Dermatol 7:58–68, 1978.
379. Fox JL, Yu C–D, Higuchi WI, Ho NFH. General physical model for simultaneous
diffusion and metabolism in biological membranes. The computational approach for
the steady-state case. Int J Pharmaceut 2:41–57, 1979.
380. Okamoto H, Yamashita F, Saito K, Hashida M. Analysis of drug penetration through
the skin by two-layer skin model. Pharm Res 6:931–937, 1989.
381. Addicks WJ, Flynn GL, Weiner N, Curl RA. A mathematical model to describe drug
release from thin topical application. Int J Pharm 56:243–248, 1989.
382. Kubota K, Yamada T. Finite dose percutaneous absorption: theory and its application
to in vitro timolol permeation. J Pharm Sci 79:1015–1019, 1990.
383. Kubota K. Finite dose percutaneous drug absorption: a BASIC program for the so-
lution of the diffusion equation. Comput Biomed Res 24:196–207, 1991.
384. Seta Y, Ghanem AH, Higuchi WI, Borsadia S, Behl CR, Malick AW. Physical model
approach to understanding finite dose transport and uptake of hydrocortisone in hair-
less guinea pig skin. Int J Pharm 81:89–99, 1992.
385. Malkinson FD, Ferguson EH. Percutaneous absorption of hydrocortisone-4-C14 in two
subjects. J Invest Dermatol 25:281–283, 1956.
386. Feldmann RJ, Maibach HI. Percutaneous penetration of hydrocortisone with urea.
Arch Dermatol 109:58–59, 1974.
387. Britz MB, Maibach HI, Anjo DM. Human percutaneous absorption of hydrocortisone:
the vulva. Arch Dermatol Res 267:313–316, 1980.
388. Bucks DAW, Maibach HI, Guy RH. Percutaneous absortion of steroids: effect of
repeated application. J Pharm Sci 74:1337–1339, 1985.
389. Maibach HI, Feldmann RJ. Percutaneous penetration in man: comparison of some
steroids. Clin Res 14:270, 1966.
390. Franz TJ, Lehman PA. Percutaneous absorption of fluocinolone in man: assessment
of relative bioequivalence of 0.05% Lidex ointment and solution. Pharm Res 7(suppl
194):PDD 7331, 1990.
391. Wickrema SAJ, Shaw SR, Weber DJ. Percutaneous absorption and excretion of trit-
ium-labeled diflorasone diacetate, a new topical corticosteroid in the rat, monkey and
man. J Invest Dermatol 71:372–377, 1978.
392. Bucks DAW, Maibach HI, Guy RH. Mass balance and dose accountability in percu-
taneous absorption studies: development of a nonocclusive application system. Pharm
Res 5:313–314, 1988.
393. Stoughton RB, Wullich K. Relation of application time to bioactivity of a potent
topical glucocorticosteroid formulation. J Am Acad Dermatol 22:1038–1041, 1990.
394. Maibach HI, Feldmann R. Effect of applied concentration on percutaneous absorption
in man. J Invest Dermatol 52:382, 1969.
395. Roberts MS, Horlock E. Effect of repeated skin application on percutaneous absorp-
tion of salicylic acid. J Pharm Sci 67:1685–1687, 1978.
396. Lippold BC, Reimann H. Wirkungsbeeinflussung bei lösungssalben durch Vehikel am
beispiel von Methylnicotinat. Acta Pharm Technol 35:128–135, 1989.
397. Stoughton RB, Wullich K. The same glucocorticoid in brand-name products. Arch
Dermatol 125:1509–1511, 1989.
398. Guin JD, Wallis MS, Wall MS, Lehmann PA, Franz TJ. Quantitative vasoconstrictor
assay for topical corticosteroids: the puzzling case of fluocinolone acetonide. J Am
Acad Dermatol 29:197–202, 1993.
399. Gibson JR, Hough JE, Marks PM. Effect of concentration on the clinical pharmacol-
ogy of corticosteroid ointment formulations. In: Maibach HI, Surber C, eds. Topical
Corticosteroids. Basel: S Karger, pp 74–92, 1992.
400. Gibson JR, Kirsch J, Darley CR, Burke CA. An attempt to evaluate the relative clinical
potencies of various diluted and undiluted proprietary corticosteroid preparations. Clin
Exp Dermatol 8:489–493, 1983.
401. Walker M, Chambers LA, Hollingsbee DA, Hadgraft J. Significance of vehicle thick-
ness to skin penetration of halcinonide. Int J Pharm 70:167–172, 1991.
402. Jackson DB, Thompson C, McCormack JR, Guin JD. Bioequivalence (bioavailability)
of generic topical corticosteroids. J Am Acad Dermatol 20:791–796, 1989.
403. Barry BW, Woodford R. Comparative bioavailability of proprietary topical cortico-
steroid preparations; vasoconstrictor assays on thirty creams and gels. Br J Dermatol
91:323–338, 1974.
404. Magnus AD, Haigh JM, Kanfer I. Assessment of some variables affecting the blanch-
ing activity of betamethasone-17-valerate cream. Dermatologica 160:321–327, 1980.
405. Zesch A. Zur Bioäquivalenz und Bioverfügbarkeit topisch applizierter Arzneimittel.
Pharm Ind 50:746–749, 1988.
406. Treffel P, Gabard B. Ibuprofen epidermal levels after topical application in vitro: effect
of formulation, application time, dose variation and occlusion. Br J Dermatol 127:
286–291, 1993.
407. Claris P. Non-invasive Biophysical Measurements on the Skin: The Use of Some
Bioengineering Methods for the Assessment of Percutaneous Penetration. Brussels:
Vrije Universiteit Brussel; 1994.
408. Unna PG. Cigolin als heilmittel der Psoriasis. Dermatol Wochenschr 6:116–137; 150–
163; 175–183, 1916.
409. Farber FM, Harris DR. Hospital treatment of psoriasis. A modified anthralin therapy.
Arch Dermatol 101:381–389, 1970.
410. Schaefer H, Farber EM, Goldberg L, Schalla W. Limited application period for dithra-
nol in psoriasis. Br J Dermatol 102:571–573, 1980.
411. Schalla W, Bauer E, Wesendahl C, Goldberg L, Farber EM, Schaefer H. Arch Der-
matol Res 267:203, 1980.
412. Runne U, Kunze J. Short-duration (‘‘minutes’’) therapy with dithranol for psoriasis:
a new out-patient regimen. Br J Dermatol 106:135–139, 1982.
413. Schaefer H, Schalla W, Shroot B. Anthralin—facts, trends and unresolved problems.
In: Hornstein OP, Schnyder UW, Schönfeld J, eds. Neue Entwicklungen in der Der-
matologie. Berlin: Springer, pp 82–91, 1984.
414. Jaeger L. Psoriasis treatment with betamethasone dipropionate using short-term ap-
plication and short-term occlusion. Acta Derm Venereol (Stockh) 66:84–87, 1986.
415. Leibsohn E. Comparison of diflorasone diacetate and betamethasone dipropionate
ointment in the treatment of psoriasis. J Int Med Res 10:22–27, 1982.
416. Taplin D, Rivera A, Walker JG, Roth WI, Reno D, Meinking T. A comparative trial
of three treatment schedules for the eradication of scabies. J Am Acad Dermatol 9:
550–554, 1983.
417. Wester RC, Noonan PK, Maibach HI. Frequency of application on percutaneous ab-
sorption of hydrocortisone. Arch Dermatol 113:620–622, 1977.
418. Wester RC, Noonan PK, Maibach HI. Percutaneous absorption of hydrocortisone in-
creases with long-term administration. Arch Dermatol 116:186–188, 1980.
419. Täuber U. Skin pharmacokinetics of the new topical glucocorticosteroid MPA. In:
Scott RC, Guy RH, Hadgraft J, eds. Prediction of Percutaneous Penetration. Vol. 1.
London: IBC Technical Services, pp 37–48, 1990.
420. Täuber U, Weiss C, Matthes H. Does salicylic acid increase the percutaneous absorp-
tion of diflucortolone-21-valerate. Skin Pharmacol 6:276–281, 1993.
421. Tsai JC, Flynn GL, Weiner N, Ferry JJ. Influence of application time and formulation
reapplication on the delivery of minoxidil through hairless mouse skin as measured
in Franz diffusion cells. Skin Pharmacol 7:270–277, 1994.
422. Sherertz EF, Sloan KB, McTiernan RG. Effect of skin pretreatment with vehicle alone
or drug in vehicle on flux of a subsequently applied drug: results of hairless mouse
skin and diffusion cell studies. J Invest Dermatol 89:249–252, 1987.
423. Fredriksson T, Lassus A, Bleeker J. Treatment of psoriasis and atopic dermatitis with
halcinonide cream applied once and three times daily. Br J Dermatol 102:575–577,
1980.
424. Sudilovsky A, Muir JG, Bococo FC. A comparison of single and multiple applications
of halcinonide cream. Int J Dermatol 20:609–613, 1981.
425. Pearlman DL. Weekly pulse dosing: effective and comfortable topical 5-fluorouracil
treatment of multiple facial actinic keratoses. J Am Acad Dermatol 25:665–667, 1991.
426. Frosch PJ, Behrenbeck EM, Frosch K, Macher E. The Duhring chamber assay for
corticosteroid atrophy. Br J Dermatol 104:57–65, 1981.
427. Lehmann P, Zheng P, Lavker RM, Kligman A. Corticosteroid atrophy in human skin.
A study by light, scanning, and transmission electron microscopy. J Invest Dermatol
81:169–176, 1983.
428. Lavker RM, Schechter NM, Lazarus GS. Effects of topical corticosteroids on human
dermis. Br J Dermatol 115(suppl 31):101–107, 1986.
429. Woodbury R, Kligman AM. The hairless mouse model for assaying the atrophogen-
icity of topical corticosteroids. Acta Derm Venereol (Stockh) 72:403–406, 1989.
430. Lubach D, Kietzmann M. Anti-inflammatory effect and thinning of the skin: models
for testing topically applied corticosteroids on human skin. In: Maibach HI, Surber
C, eds. Topical Corticosteroids. Basel: S Karger, pp 26–41, 1992.
431. Pflugshaupt C. Diskontinuierliche topische Kortikoidtherapie. Zbl Haut Geschlecht-
skrankh 148:1229–1367, 1983.
432. English JS, Bunker CB, Ruthven K, Dowd PM, Greaves MW. A double-blind com-
parison of efficacy of betamethasone dipropionate cream twice daily versus once daily
in the treatment of steroid responsive dermatoses. Clin Exp Dermatol 14:32–34, 1989.
433. Lubach D, Rath J, Kietzmann M. Steroid-induced dermal thinning: discontinuous
application of clobetasol-17-propionate ointment. Dermatology 185:44–48, 1992.
434. Lubach D, Bensmann A, Bornemann U. Steroid-induced dermal atrophy. Dermato-
logica 179:67–72, 1989.
435. Wozel G, Barth J. Fluocinolonacetonid und Psoriasis—kontinuierliche und alterni-
erende Therapie. Dermatol Monatsschr 172:620–623, 1986.
436. Lubach D, Rath J, Kietzmann M. Alteration of skin thickness induced by initial con-
tinuous followed by intermittent topical application of clobetasol. Dermatology 190:
51–55, 1995.
437. Katz HI, Hien NT, Prawer SE, Scott JC, Grivna EM. Betamethasone dipropionate in
optimized vehicle: intermittent pulse dosing for extended maintenance treatment of
psoriasis. Arch Dermatol 123:1308–1311, 1987.
438. Levy G. Bioavailability, clinical effectiveness, and public interest. Pharmacology 8:
33–43, 1972.
439. Shah VP, Van Zwieten–Boot B. Characterisation of topical products for local action
—quality, efficacy and safety. Symposium on Quality and Interchangeability of Top-
ical Products for Local Action, Dec 8; Nürnberg: European Federation for Pharma-
ceutical Sciences, 1995.
440. Yacobi A, Skelly JP, Shah V, Benet LZ. Integration of Pharmacokinetics, Pharmaco-
dynamics, and Toxicokinetics in Rational Drug Development. New York: Plenum
Press, 1993.
441. Peck CC, Collins JM. First time in man studies; a regulatory perspective—art and
science of phase I trials. J Clin Pharmacol 30:218–222, 1990.
442. Narang PK, Bianchine JR. Phase I studies: initial evaluation of drug toxicity, dose
response, and efficacy. In: Max M, Portenoy R, Laska E, eds. Advances in Pain
Research and Therapy: 18, New York: Raven Press, pp 9–31, 1991.
443. Levy G. Concentration-controlled versus concentration-defined clinical trials. Clin
Pharmacol Ther 53:387–388, 1993.
444. Levy G, Ebling WF, Forrest A. Concentration- or effect-controlled clinical trials with
sparse data. Clin Pharmacol Ther 56:1–8, 1994.
445. Urquart J. How much compliance is enough? Pharm Res 13:10–11, 1996.
446. Meyer UA, Kindli R. Plazebos und nozebos. Ther Umsch 46:544–554, 1989.
447. Pearce JMS. The placebo enigma. Q J Med 88:215–220, 1995.
448. Ernst E, Resch KL. Concept of true and perceived placebo effects. Br Med J 311:
551–553, 1995.
449. Golomb BA. Paradox of placebo effect. Nature 375:530, 1995.
450. Olson SC, Eldon MA, Toothaker RD, Ferry JJ, Colburn WA. Controversy. II: bio-
equivalence as an indicator of therapeutic equivalence: modeling the theoretic influ-
ence of bioinequivalence on single-dose drug effect. J Clin Pharmacol 27:342–345,
1987.
451. Lee JW, Hulse JD, Colburn WA. Surrogate biochemical markers: precise measurement
for strategic drug and biologics development. J Clin Pharmacol 35:464–470, 1995.
452. Barrett CW, Hadgraft JW. The effect of particle size and the vehicle on the percuta-
neous absorption of fluocinolone acetonide. Br J Dermatol 77:576–578, 1965.
453. Sarkany I, Hadgraft JW, Caron GA, Barrett CW. The role of vehicles in the percu-
taneous absorption of corticosteroids. Br J Dermatol 77:569–575, 1965.
454. Christie GA, Moore–Robinson M. Vehicle assessment—methodology and results. Br
J Dermatol 82(suppl. 6):93–98, 1970.
455. Barry BW. Comparative bio-availability and activity of proprietary topical corticoste-
roid preparations: vasoconstrictor assays on thirty-one ointments. Br J Dermatol 93:
563–567, 1975.
456. Barry BW. Proprietary hydrocortisone creams, vasoconstrictor activities and bio-avail-
abilities of six preparations. Br J Dermatol 95:423–425, 1976.
457. Barry BW, Woodford R. Vasoconstrictor activities and bioavailabilities of seven pro-
prietary corticosteroid creams assessed using a non-occluded multiple dosage regimen;
clinical implications. Br J Dermatol 97:555–560, 1977.
458. Woodford R, Barry BW. Activity and bioavailability of amcinolone and triamcinolone
acetonide in experimental and proprietary topical preparations. Curr Ther Res 21:
877–886, 1977.
459. Stoughton RB, Cornell RC. Review of super-potent topical corticosteroids. Semin
Dermatol 6:72–76, 1987.
460. Shah VP, Peck CC, Skelly JP. ‘‘Vasoconstriction’’—skin blanching-assay for gluco-
corticosteroids—a critique. Arch Dermatol 125:1558–1561, 1989.
461. Olsen E. A double-blind controlled comparison of generic and trade-name topical
steroids using the vasoconstriction assay. Arch Dermatol 127:197–201, 1991.
462. Agrawal N, Lehman PA, Franz TJ. Bioavailability of hydrocortisone from seven top-
ical formulations using the in vitro finite dose method. Pharm Res 12:S-277, 1995.
463a. Lehman PA, Franz TJ. Bioavailability of hydroquinone from six topical formulations
using the in vitro finite dose method. Pharm Res 9:S-233, 1992.
463b. Schwarb FP, Imanidis G, Smith EW, Haigh JM, Surber C. Effect of concentration
and degree of saturation of topical fluocinonide formulations on in vitro membrane
transport and in vivo availability on human skin. Pharm Res 16:917–923, 1999
464. Marks R, Dykes PJ, Gordon J, Halan G, Davis A. Percutaneous penetration from a
low dose supersaturated hydrocortisone acetate formulation. Presented at the British
Society of Investigative Dermatology Annual Meeting, Sheffield, 1992.
465. Sanathanan LP, Peck CC, Temple RT, Lieberman R, Pledger G. Randomization, PK-
controlled dosing, and titration: an integrated approach for designing clinical trials.
Drug Inform J 25:425–431, 1991.
466. Levy G. Patient-oriented pharmaceutical research: focus on the individual. Pharm Res
12:943–944, 1995.
467. Stoughton RB. Penetration of drugs through skin. Dermatologica 152(suppl 1):27–
36, 1976.
468. Blackwelder WC. Proving the null hypothesis in clinical trials. Controlled Clin Trials
3:345–353, 1982.
469. Gibson JR, Kirsch JM, Darley CR, Harvey SG, Burke CA, Hanson ME. An assess-
ment of the relationship between vasoconstrictor assay findings, clinical efficacy and
skin thinning effects of a variety of undiluted and diluted corticosteroid preparations.
Br J Dermatol 111(suppl 27):204–212, 1984.
470. Barry BW, Fyrand O, Woodford R, Ulshagen K, Hogstad G. Control of bioavailability
of a topical steroid; comparison of desonide creams 0.05% and 0.1% by vasocon-
strictor studies and clinical trials. Clin Exp Dermatol 12:406–409, 1987.
471. Martin GP. The effect of dose under a hydrocolloid patch on the human bioavailability
of a topical corticosteroid. J Pharm Pharmacol 41:129P, 1989.
472. Kecskés A, Jahn P, Wendt H, Lange L, Kleine Kuhlmann R. Dose–response relation-
ship of topically applied methylprednisolone aceponate (MPA) in healthy volunteers.
Eur J Clin Pharmacol 43:157–159, 1992.
473. Gibson JR. Effect of concentration on the clinical potency of corticosteroid ointment
formulations. Skin Pharmacol 1:214–222, 1987.
474. Surber C, Davis A, Rufli T. Factors influencing topical bioavailability: New concept
of topical dosing. Dermatology 191:173, 1995.
475. Marples and Kligman. 1973. Rec No 26240
476. Stenberg C, Larko O. Sunscreen application and its importance for the sun protection
factor. Arch Dermatol 121:1400–1402, 1985.
9
Scale-up of Dermatological Dosage
Forms: A Case for Multivariate
Optimization and
Product Homogeneity
OREST OLEJNIK and BRUCE A. FIRESTONE

Allergan, Inc., Irvine, California
I. INTRODUCTION
The objective of any drug-delivery system is to deliver in vivo the active drug moiety
to the target tissue or receptor site. Whether the delivery system is simple or so-
phisticated, it is the responsibility of the formulator, and other product development
support functions, to design a drug-delivery system capable of consistently achieving
the desired pharmacokinetic–pharmacodynamic profile as an outcome of formulation
and manufacturing process development and optimization. To achieve a robust for-
mulation and reproducible product manufacturing in which product quality specifi-
cations are consistently met is not always an easy and straightforward exercise.
Driven by formulation obstacles, processing complexities, an ever-increasing regu-
latory environment, health care cost-containment measures, and other pharmaco-
economic challenges can certainly influence the approach to and outcome of both
product and process development. To achieve rapid and successful development of
a drug-delivery system, there is an increasing focus on experimental design tech-
niques, product scale-up planning, and process technology transfer activities. Addi-
tionally, and often considered the bane of many projects, is cost. This financial aspect
is mentioned because there is a growing requirement within the pharmaceutical in-
dustry in specifying target values for manufacturing costs as early as possible within
the life cycle of a project that could constrain the final production process. A
499
500 Olejnik and Firestone
knowledge of the effect of cost, estimated or otherwise, ensures not only that the
criteria for efficacy, safety, and quality of a product are achieved, but also that the
economic and environmental implications in product manufacturing are understood
and deemed acceptable. It further augments the necessity in the implementation of
well-planned, systematic methods, and the definitive need for understanding the per-
formance characteristics and the critical aspects of manufacture of a product, partic-
ularly in its scale-up and transition to the production environment.
To address all these factors leading up to a successfully developed and man-
ufacturable drug-delivery product that meets predefined criteria or target values
would be a major undertaking. Although not dealt with here, their influence in any
drug development project should not be viewed as insignificant. One aspect is im-
portant to stress, which centers on multivariate optimization. This technique, when
used in experimental design, offers valuable information in understanding and im-
proving formulation and process development. To the reader the application of design
optimization techniques may be well understood and considered commonplace; how-
ever, it may be surprising to learn that these enabling techniques are infrequently
applied within the pharmaceutical industry. Although available information from a
survey on 68 pharmaceutical companies was somewhat dated, 14% consistently uti-
lized optimization methods, with 35% reporting occasional use (1). This translated
into 51% of those companies who, it must be surmized, have adopted some other
iterative or time-consuming trial-by-error system in developing their drug-delivery
systems. Product quality and cost under these circumstances must surely be nega-
tively affected. Either way, these approaches are inefficient and can cause problems.
Beyond phase I, attention to scale-up and process technology transfer becomes crit-
ical to the extent that reliance and implementation of multivariate optimization tech-
niques is considered a necessity, as well as the practice of good science. It is, there-
fore, considered meaningful in discussing multivariate optimization, that is,
experimental design methods with the intent not only to maintain an awareness of,
but also to promote their utilization, especially when the intent is to conduct process
scale-up studies.
In the broad context, the practices in the processing of pharmaceutical dosage
forms have been reviewed elsewhere (2). Additionally, a workshop summary report
on the scale-up of liquid and semisolid disperse systems provides useful guidance
(3). The focus, in the second part of this chapter, is on a particular aspect of scale-
up—namely, product homogeneity which, if not adequately understood and con-
trolled, can have a catastrophic effect on product quality and performance. Homo-
geneity of topical products is a subject that has received scant discussion in the
pharmaceutical literature compared with solid oral dosage forms. Because of the
complex nature of many topical products as being multicomponent semisolid for-
mulations, there is rising awareness of the importance and challenges in obtaining
acceptable product homogeneity. This is particularly true for topical products that
are difficult to mix (viscous systems), involve dispersed solids or emulsified liquids,
or contain low concentrations of potent active substances. Achieving acceptable prod-
uct homogeneity places prime importance on choice of manufacturing equipment,
manufacturing process design, validation, and appropriate product stability monitor-
ing to assure consistent product quality and performance. From this perspective the
mixing operations, mixing equipment, and mixing parameters are highlighted. A final
Scale-Up of Dermatological Dosage Forms 501
section deals with the verification of bulk homogeneity and content uniformity of
topical products.
II. STUDY OPTIMIZATION BY EXPERIMENTAL DESIGN

Application of multivariate optimization techniques allows a structured and efficient
way in finding the set of conditions that either cause or prevent a product or process
from performing in an optimal manner. From the numerous methods (e.g., evolu-
tionary operation [EVOP], or Taguchi methods) available to the formulation and
processing scientist, experimental design and simplex optimization are the most fre-
quently used (4–7). Regardless of the selected method, the process of optimization
is based on a mathematical approach that searches for the conditions from several
variables, n of a function [f (x1, x2, x3, ⭈ ⭈ ⭈ xn)] that optimize f(x). The simplest ap-
proach, a univariate search, is taking a single variable or factor at a time to determine
an optimal solution to the problem being analyzed. An example would be in the
study of the effect of varying temperature on a formulation at a constant pH. Un-
fortunately, this procedure is inherently flawed, for it does not consider any of the
interdependencies that may exist among the variables. For pharmaceutical formula-
tions interdependencies between variables occur and are generally the rule, rather
than the exception; accordingly; a multivariate optimization approach is preferred
(8).
A. Factorial Design
When the number of variables are limited, generally to five or fewer, a two-level
factorial design can provide a significant amount of information from only a few
experiments. Such a design allows one to measure the influence of several variables
simultaneously. This particular factorial design consists of all possible combinations
of two levels of each of the variables. If two variables are involved, the full factorial
design involves 22 experiments (i.e., a total of 4). The expression for the full factorial
design is given as
N = Lk
where N is the number of experiments, L is the variable level, and k is the number
of variables.
A two-level, three-variable design requires 23 or 8 experiments. These eight
experiments can be represented in four different ways, in terms of:
Plus (⫹) and minus (⫺) notations

High and low levels of the variables
Treatment combinations
Vertices of a cube
To better illustrate this, designate the three variables as X, Y, and Z; and use
the notation approach, in which (⫺) signifies the low level of a variable and (⫹)
signifies the high level of a variable, the eight experimental trials can then be rep-
resented as
Trial no. Variable X Variable Y Variable Z
1 ⫺ ⫺ ⫺
2 ⫹ ⫺ ⫺
3 ⫺ ⫹ ⫺
4 ⫹ ⫹ ⫺
5 ⫺ ⫺ ⫹
6 ⫹ ⫺ ⫹
7 ⫺ ⫹ ⫹
8 ⫹ ⫹ ⫹
The second approach, using ‘‘high’’ and ‘‘low’’ descriptors for each of the variables,
can be represented as follows:
Trial no. Level of X Level of Y Level of Z
1 Low Low Low

2 High Low Low
3 Low High Low
4 High High Low
5 Low Low High
6 High Low High
7 Low High High
8 High High High
Interpreting these high and low or (⫹) and (⫺) descriptors obviously requires
definition of the conditions under which the specific variable is being evaluated. For
example, in studying the homogeneity effect of a product under two different mixing
times (60 and 30 min), 60 min is then the high or (⫹) level of the variable of interest
(i.e., mixing time) and 30 min is the low or (⫺) level of mixing time.
The eight experimental trials, as presented, cover all the combinations of a two-
level, three-variable design by which these combinations can be shown in the form
of a factorial block:
1 z
Trial 1 Trial 5
1 (1) z
1
Trial 2 Trial 6
x x xz
Trial 3 Trial 7
1 y yz
y
Trial 4 Trial 8
x xy xyz
In the absence of a letter, or if the number 1 is present, this refers to the low level
of a variable. If a letter is designated, then this represents the high level of the
variable (e.g., x represents the high level of variable X). When interpreting this
factorial block, consider the treatment combination xyz. This xyz block is in the x-
row, the y-row, and the z-column. Consequently, the test sample is required to be
prepared at the high level of X, the high level of Y, the high level of Z, which is
trial number 8.
The fourth, is a geometric approach in representing the 23 factorial design. A
three-dimensional schematic is used, in which the vertices of the cubical form rep-
resent the experimental trials as follows:
Each of the four ways described in the foregoing offers its own merits, but
selection is frequently based on preference. For example, in the scale-up of a for-
mulation the mixing of components to achieve product homogeneity may involve
the following variables: temperature, mixing time, and excipient supplier. It is also
assumed that the study is to determine the effects on some response of interest: here,
product homogeneity, of two levels of temperature (40⬚ and 20⬚C), two levels of
mixing time (30 and 60 min), and two different suppliers (supplier S1 and supplier
S2). Accepting the designations for temperature, mixing, and supplier as variables X,
Y, and Z, respectively, then the designations of the three variables can be described
as follows:
Variable X (temperature)
1. 20⬚C is designated as the low, or (⫺), or (1) level of variable X.
2. 40⬚C is designated as the high, or (⫹), or (x) level of variable X.
Variable Y (time)
1. 30 min is designated as the low, or (⫺), or (1) level of variable Y.
2. 60 min is designated as the high, or (⫹), or ( y) level of variable Y.
Variable Z (supplier)
1. Supplier S1 is designated as the low, or (⫺), or (1) level of variable
Z.
2. Supplier S2 is designated as the high, or (⫹), or (z) level of variable
Z.
In taking the previous example of block xyz, experimental trial 8 can be de-
scribed by any of the four following designations:
Variable
X Y Z
High High High

⫹ ⫹ ⫹
x y z
40⬚C 60 min Supplier S2
Similarly, all of the eight factorial blocks (treatment combinations), previously de-
scribed, can be defined in this manner. By taking an individual sample at each of
the eight treatment combinations any observed difference between result 2 and result
1 would be assigned to the high level of X versus the low level of X, assuming the
absence of any other experimental variables. On this basis, the same would be true
for samples 4 vs. 3, 6 vs. 5, and 8 vs. 7. Here the effect of variable X would have
been measured a total of four times. By theoretical computation, this number can be
shown to be a number necessary in making a valid decision. Additional and more
detailed information are provided in other texts that deal with complete factorial
designs, to which the reader is referred (9,10).
B. Fractional Factorial Design

From the factorial equation given earlier in this text, it is obvious that as the number
of variables increases, the number of experiments increases exponentially. Fractional
factorial designs, a subset of full factorials, promotes efficiencies in conducting the
relevant experimental studies through exploitation of any inherent redundancies that
may exist. On a priori grounds some factor effects may be insignificant, with only
the estimated main effects and some lower-order interactions evaluated. When high-
order interactions occur, their effects are negligible and can be ignored because they
are unlikely to bias the results. The mathematical expression for fractional factorial
design is given by:
N = Lk⫺f
r
where r is the resolution code of a fractional factorial design; f is the fraction of the
full factorial; N, L, and k are the terms presented previously. For example, a half-
fractional ( f = 1), two-level, five-factor design would be expressed as 25⫺1 resulting
in a total of 16 experimental trials. If the fraction is one-quarter ( f = 2) then 25⫺2
trials are required (i.e., eight experiments, or all of the experiments of the full design
are conducted. The limitation to this type of design for testing purposes is that there
is a loss in the degrees of freedom. A calculation of three-way and higher-order
interaction effects would not be possible owing to insufficient degrees of freedom.
A half-factorial design for five factors at two levels was used by Schwartz et
al. (11) in the optimization of an existing solid dosage form. In this study, three
additional levels were selected, resulting in a total of 27 experimental trials. Data
from the study were subjected to computed regression analysis to determine the fit
to a second-order model. Despite limitations in restricting the model to second-order
polynomials, the resulting equations were good predictors in subsequent computer-
ized experimentation. Associated with computer-generated graphs, such as those of
a given response variable as a function of an independent variable, allows an op-

portunity to rapidly simulate situations from which the formulator can gain an insight
into the outcome of the process being investigated or optimized. Several steps are
listed and defined, as necessary, to achieve a successful optimization program.
1. System selection
2. Variable selection
a. Independent
b. Dependent
3. Experimental trial
4. Statistical and regression analysis of data
5. Set specifications before feasibility search
6. Select constraints for grid search
7. Evaluate search results
8. Generate and evaluate
a. Derivative plots
b. Contour plots
9. Implement and monitor optimization decision(s)
This approach was effectively used in troubleshooting a formulation at the
production scale (12). Moreover, it emphasized the value in utilizing preproduction
optimization data in defining and facilitating the improvements necessary to achieve
a product that consistently meets the quality specifications.
Experimental design studies can evolve into significantly larger exercises, as
would be evident with a four-variable, five-level, full-factorial design resulting in a
need to conduct 625 experiments. However, the goals before optimization must be
clear and understood to avoid unnecessary experimentation. In the preceding design,
the goal would not be to find the optimal settings for the variables but, rather, to
gain an understanding of the system (formulation or process). For different phases
of a project, experimental design would provide the wherewithal in guiding the
project to an increasing chance of success and optimization.
Although our proceeding discussion on homogeneity centers on those param-
eters and factors that affect the uniform dispersion of excipient and the activity of a
topical product, it is understood that any studies that are conducted must be based
on a foundation of good experimental design. Only then can one achieve a solid
understanding and a capability for optimization in the formulation and process con-
trol of a pharmaceutical product.
III. HOMOGENEITY
A. Product Physicochemical Properties
Important to the assurance of product homogeneity is the design of manufacturing
processes that are capable of producing a well-mixed, homogeneous product at the
bulk stage, and that this design is maintained throughout product transfer, filling,
transport, and shelf-life. This requires a thorough understanding of those physico-
chemical properties of the product that drive the choice of equipment and the design
of process unit operations that compose the overall manufacturing process. For top-
ical products the following properties are often considered:
Viscosity (rheology)
Shear sensitivity
Settling and Sedimentation rates
Solubility and dissolution of the active components
Stability of components (actives, antioxidants, etc.)
Particle, droplet, or globule size
Heating or cooling requirements
pH
Other physicochemical properties (e.g., crystal morphology, solvation, poly-
morphism of active component)
B. Mixing Operations
The design of bulk–product-mixing operations should consider the type of mixing
that should be applied to the product to obtain the desired outcome. A diverse range
of mixing operations are commonly employed during manufacture of topical products
including the following:
Blending of multicomponent systems
Dispersing of solids into liquids
Emulsifying one liquid into another
Promotion of heat transfer between the product and vessel wall
Each of these operations requires suitably designed mixing equipment and the
application of the appropriate type of mixing under a specified set of mixing param-
eters (rates of component addition, mixing speeds, temperatures, mixing times, etc.).
Mixing can be thought of in terms of the applicable mixing scale required for a
particular process. The distinction between macroscale mixing versus microscale
mixing has been made for the purposes of discussion as now defined (13):
1. Macroscale mixing: Mixing that accomplishes adequate product flow in all
regions of the mixing vessel both side-to-side and top-to-bottom to prevent
stratification and ensure macroscopic homogeneity.
2. Microscale mixing: Mixing that accomplishes separation or dispersion of
individual components (particles, droplets, etc.) to obtain the desired par-
ticle size distribution.
Macroscale and microscale mixing can be considered pumping and shearing
processes, respectively. It follows from this discussion that virtually all multicom-
ponent products will require adequate macroscale mixing (pumping) to be rendered
homogeneous. However, macroscale mixing, while necessary, may not be sufficient
for some products. Dispersed systems may, in addition, require microscale mixing
(shearing) to render the desired particle size distribution.
C. Mixing Equipment
A diverse array of mixing equipment is available to use for liquids and semisolids
in topical product manufacture. Generally, a cylindrical tank is employed that is
equipped with one impeller agitator or more to accomplish the desired mixing op-
eration. The following types of impellers are commonly employed.:
Paddles: Two-bladed and four-bladed paddles mounted on a vertical shaft

and operated at low speeds are sometimes used for simpler mix-
ing or heat transfer operations. The blades may be pitched to
provide axial-flow patterns with low-applied shear. Scraper-
blades that sweep the vessel surface are used for viscous prod-
ucts to prevent surface build-up and to facilitate heat transfer.
Propellers: Three- or four-bladed propellers are used for high speed, axial
mixing of low-viscosity liquids. Generally achieves good pump-
ing action.
Turbines: Used for high-speed mixing of liquids over a range of viscosities
with blades that can be straight or curved, pitched or vertical.
Impeller diameter is generally smaller than for paddles. Com-
bines both pumping and shear mixing.
Dispersers: High-speed mixers using a sawblade or ‘‘Cowles’’ disk impeller
for the dispersion of solids into liquids. Accomplishes good wet-
ting and deagglomeration of solid particles through the applied
shear.
Rotorstators: Mixer in which a high-speed impeller (rotor) is oriented in close
proximity to a stationary housing (stator) leaving a narrow
‘‘gap’’ in-between. Designed for high-shear mixing applications
including emulsification of liquids and dispersion of solids.
A vessel equipped with one or more of the mixers just listed may be required
to provide the needed balance between macroscale mixing (pumping) and microscale
mixing (shear).
D. Mixing Parameters
Once the mixing equipment has been chosen to provide the appropriate type of
mixing for a particular product, the determination of optimal mixing parameters
(rates of component addition, mixing speeds or tip speeds, temperatures, mixing
times, and such) needs to be made to ensure that a homogeneous product is consis-
tently obtained. Because undermixing or overmixing can have undesirable effects on
the product, such as stratification, clumping of solids, incorrect particle size, or vis-
cosity loss through excessive shearing, process development studies should be con-
ducted to establish the optimal mixing conditions, including definition of all param-
eter ranges. This activity may also be required for in-process subparts that may be
produced, depending on their physicochemical properties and the criticality of the
mixing operation. The mixing parameters, often under worst-case conditions, are then
formally demonstrated during process validation.
From a well-known mixer power consumption law, the following equation re-
lates the power P consumed as a function of the mixer speed N, and the impeller
diameter D where K is a constant (14):
P = KN 3D 5
This equation can be further partitioned into terms describing pumping and shearing
as follows:
P = (K1N 2D 2)(K2ND 3)
shear pumping
The foregoing equation shows that a change in mixer speed N has a strong influence
on the shear component of mixing, whereas a change in impeller diameter D affects
the pumping component most strongly.
The considerations of product physicochemical properties, type of mixing re-
quired (pumping or shearing), equipment selection, and specific mixing parameters
(speed, time, temperature, etc.) should guide the development of optimal mixing unit
operations. This will ultimately lead to a manufacturing process that consistently
produces homogeneous product that can be easily scaled-up and validated.
E. Verification of Bulk Homogeneity

As part of process development and validation, the homogeneity of the bulk product
must be established after final mix (to establish final mixing conditions) by suitable
sampling of the bulk and subsequent analytical testing.
Several sampling methods are available to establish the bulk product homo-
geneity after final mix. These techniques include
1. Sampling product from various locations within the tank with a suitable
sampling device (i.e., sampling thief)
2. Sampling product from one or more valves (e.g., discharge valve) as a
function of mixing time
3. Sampling from a transfer line at the beginning, middle, and end of batch
transfer to a holding tank
Careful consideration must be given to the sampling method(s) chosen to es-
tablish bulk product homogeneity. The methods must produce samples that are rep-
resentative of actual product at that stage in its manufacture and should not introduce
any artifacts or perturbations that could be falsely interpreted. For example, it may
not be appropriate to sample from various locations within the tank if to do so means
opening a sealed tank and exposing the product to undesirable conditions (e.g., light
or oxygen). Use of a sampling thief should be done only if the device is totally inert
to the product (no drug binding, pH alteration, or particle adsorption). In fact more
than one sampling technique may be applied especially for difficult-to-mix products
(high viscosity) when stratification (settling) may occur over time, or for products
containing small quantities of a finely divided drug.
F. Content Uniformity Testing of Topical Products

Testing of topical product homogeneity should also be performed over the shelf-life
in the final container–closure system, as is often done for solid oral dosage forms.
This information is important in establishing that the product remains homogeneous
from beginning to end of the filling operation, after prolonged storage times under
various storage conditions, and that phase separation leading to variations in dose
uniformity does not occur.
Inherent in this discussion is deciding what is the relevant ‘‘scale of scrutiny’’
of the product to be tested for dose or content uniformity and how individual product
containers should be sampled. A most reasonable scale of scrutiny for dosage forms
is based on the size of a single dose, which is easily applied for dosage forms that
are discrete (e.g., tablets, capsules, and suppositories). The same definition may be
applied to topical products, as discussed by Train (15,16), but becomes harder to
quantitatively define because of the bulk nature of the product in the final container–
closure system and the imprecise nature of product application in the hands of the
patient. Clearly, what is needed for content uniformity testing of topical products is
a standardization of dose uniformity.
Hersey and Cook discussed the standardization of content uniformity for topical
products (17), although formal standards have not yet been adopted by the phar-
macopeia for topicals as they have for solids. Orr et al. defined a scale of scrutiny
for ointments based on a ‘‘minimum ointment thickness’’ necessary for a therapeutic
effect and a ‘‘minimum discrete area’’ of topical treatment that can be considered as
a separate entity from surrounding areas of skin (18). Although they recognized that
a minimum thickness of ointment film cannot be exactly measured, they reasoned
that it would fall in to the range of 10–100 ␮m. They also suggested that the
minimum discrete area for an ointment layer 50-␮m thick should not be greater than
1 mm2.
In actuality, the minimum discrete area should be defined on the basis of the
product’s clinical indication, as greatly differing areas may be involved whether one
using a product to treat acne, skin rash, or psoriasis. Estimates of total affected areas
and the minimum film thickness for a given product will permit the definition of the
appropriate scale of scrutiny. Provided that this dose level is practical from an ana-
lytical perspective (weighing or assay sensitivity, etc.), the single-dose quantity is
suitably defined for the purposes of content uniformity testing. This sampling quan-
tity can then be used to establish content uniformity for product from multiple lo-
cations within a single tube (closure-end, middle, and seal-end of expelled product
or center-versus-side wall product after cutting tubes open) as a function of product
age and storage conditions.
If we take into account that all the foregoing elements and requirements can
lead to a well-defined and controlled product, achieving and maintaining the ho-
mogeneous state of a topical, semisolid formulation through scale-up and beyond is
without question a desirable goal.
IV. CONCLUSIONS
In the real world, demands for the rapid and successful development of pharmaceu-
tical formulations and product-manufacturing processes require the application of
enabling tools, such as those of good experimental designs. Unfortunately, the tools
and techniques available to the industrial pharmaceutical scientist are not always
effectively applied or, indeed, considered. From a regulatory perspective, discussions
on the guidance for chemistry, manufacturing, and control changes have emphasized
the need for sound scientific and engineering practices during any scale-up or process
change, which certainly underscores that effective and meaningful scientific practices
are followed (19). Through an awareness of experimental design, a solid foundation
on which the actual scale-up and process manufacture of products can be accom-
plished. A further emphasis placed on product homogeneity adds and expands the
ability of the pharmaceutical scientist in achieving a successful outcome in the con-
trol and consistent manufacture of the final product.
REFERENCES
1. Shangraw RF, Demarest DA. A survey of current industrial practices in the formulation
and manufacture of tablets and capsules. Pharm Technol 17:32, 1993.
2. Lachman L, Lieberman HA, Kanig JL. The Theory and Practice of Industrial Pharmacy.
3rd ed. Philadelphia: Lea & Febiger, 1986.
3. Van Buskirk GA, Shah VP, Adair D, et al. Workshop III report: scale-up of liquid and
semisolid disperse systems. Pharm Res 11:1216–1220, 1994.
4. Fonner DE Jr, Buck JR, Banker GS. Mathematical optimization techniques in drug
product design and process analysis. J Pharm Sci 59:1587, 1970.
5. Schwartz JB. Optimization techniques in product formulation. J Soc Cosmet Chem 32:
287, 1981.
6. Wehrle P, Korner D, Benita S. Sequential statistical optimization of a positively-charged
submicron emulsion of miconazole. Pharm Dev Technol 1:97, 1996.
7. Taguchi G. Robust technology design. Mech Eng 115:60, 1993.
8. Bohidar NR, Restaino FA, Schwartz JB. Selecting key parameters in pharmaceutical
formulations by principal component analysis. J Pharm Sci 64:966, 1975.
9. Anderson VL, McLean RA. Design of Experiments: a Realistic Approach. New York:
10. Anderson RL. Practical Statistics for Analytical Chemists. New York: Van Nostrand
Reinhold, 1987.
11. Schwartz JB, Flamholz JR, Press RH. Computer optimization of pharmaceutical for-
mulations I: general procedure. J Pharm Sci 62:1165, 1973.
12. Schwartz JB, Flamholz JR, Press RH. Computer optimization of pharmaceutical for-
mulations II: application in trouble shooting. J Pharm Sci 62:1518, 1973.
13. Scott RR. A practical guide to equipment selection and operating techniques. In: Lie-
berman HA, Rieger MM, Banker GS, eds. Pharmaceutical Dosage Forms: Disperse
Systems, Vol. 2. New York: Marcel Dekker, pp 1–71, 1989.
14. McCabe WL, Smith JC, Harriott P. Unit of Operations of Chemical Engineering, 5th
ed. New York: McGraw–Hill, 1993.
15. Train D. Pharm J 185:129, 1960.
16. Hersey JA. Drug Dev Commun 1:119–132, 1974–1975.
17. Hersey JA, Cook PC. J Pharm Pharmacol 26:126–133, 1974.
18. Orr NA, Smith EA, Smith JF. Int J Pharm Technol Prod Manuf 1:4–10, 1980.
19. Lucisano LJ, Franz RM. Pharm Technol 19:30, 1995.
10
Safety Considerations for Dermal
and Transdermal Formulations
PETER J. DYKES
Cutest (Skin Toxicity Testing Company), Cardiff, Wales
ANTHONY D. PEARSE
University of Wales College of Medicine and Cutest (Skin Toxicity Testing
Company), Cardiff, Wales
I. INTRODUCTION
Unlike our ancestors, we now live in a complex environment full of a variety of
synthetic molecules. On a day-to-day basis contact is made with a wide range of
compounds and almost everything that comes into contact with the skin has the
potential for inducing adverse cutaneous reactions. A brief review of the dermato-
logical journals dealing with contact dermatitis will reveal that someone, somewhere
has reacted to something that they have been in contact with for short or long periods
of time. Contact dermatitis reactions may be immediate or delayed, chronic or acute,
irritant or allergic. An additional cofactor in the development of cutaneous reactions
may be ultraviolet radiation (UVR). UVR has the capacity to energize molecules,
and in this photoactive state, these molecules can produce phototoxic (photoirritant)
and photosensitive (photoallergic) reactions. With topical formulations the cause of
adverse reactions may be the vehicle and not the drug itself. Similarly with trans-
dermal delivery systems the adhesives used to produce intimate contact with the skin
may be a source of cutaneous reactivity. An additional problem with transdermal
delivery systems is that the physical process of removal from the skin can induce
mechanical trauma to the skin surface leading to erythema and edema. A further
complication may be that the mechanical trauma leads to altered barrier function of
the skin which, in turn, could enhance the percutaneous penetration of materials at
or near the skin surface. Whatever the mechanism, adverse cutaneous reactions are
511
512 Dykes and Pearse
a frequent problem with materials that come into contact with the skin. The resulting
contact dermatitis can be a source of considerable discomfort and inconvenience and
limit the usefulness of a topical formulation or transdermal delivery system.
II. TYPES OF CONTACT DERMATITIS

Some materials that come into contact with the skin penetrate the stratum corneum
and cause cell death or injury. This effect is visualized as a local inflammatory skin
reaction that is characterized primarily by erythema and edema; an irritant dermatitis.
These types of reactions may be further classified into acute toxic contact dermatitis
and irritant contact dermatitis. Acute toxic contact dermatitis may be provoked by a
single or, sometimes, repeated application of strongly toxic materials. The association
with the material and development of a reaction is usually quite obvious. These types
of reaction occur after contact with materials such as acids, alkalis, solvents, and
cleansers. This type of reaction is only rarely associated with topical application of
a drug or cosmetic product.
In contrast with acute toxic contact dermatitis, irritant contact dermatitis occurs
after repeated exposure. It is a localized, superficial, nonimmunologically based re-
action. Many substances, including topical pharmaceutical agents, cause irritant con-
tact dermatitis after repeated application. Often there is little indication of a reaction
after the initial contact. However, after repeated exposure, the skin becomes ery-
thematous, with drying and cracking of the skin surface. In more severe reactions
edema may occur with papule and vesicle formation. Reactions are usually localized
to the site of contact and, unlike allergic reactions, usually diminish fairly quickly
after the stimulus has been removed. Irritant contact dermatitis will develop in all
individuals given sufficient time and intensity of exposure. Reactions develop more
quickly in situations where the material is held in intimate contact with the skin and
in conditions where the barrier function of the skin is altered (e.g., in flexural creases
or under occlusion with polyethylene).
Some materials penetrate the skin and stimulate an immune response. Any
future contact results in an inflammatory immune reaction that can be intense, with
erythema, edema, and vesiculation, possibly spreading beyond the original point of
contact: an allergic contact dermatitis. Two main types of allergic contact dermatitis
are involved in adverse reactions to topically applied materials: immediate-type and
delayed-type hypersensitivity. Immediate type hypersensitivity (type I) is the result
of antibody–allergen interaction occurring in the skin, the reaction that develops
being known as allergic contact urticaria. Delayed-type hypersensitivity (type IV) is
the result of cell-mediated immunity and is the most frequently reported side effect
of topical drugs.
Some materials require ultraviolet radiation (UVR) to activate them before they
are able to elicit adverse reactions. These photosensitive reactions can be phototoxic
(photoirritant; i.e., nonimmunological) or photosensitization (photoallergic) reactions.
In phototoxic reactions the chemical or drug is altered by UVR such that it becomes
an irritant and causes inflammatory changes in the skin. In photosensitization reac-
tions a reactive species is formed that is an allergen, or one that is capable of reacting
with another moiety present in the skin to form an allergen. This can then stimulate
the immune system and produce an allergic response such as a type I immediate-,
or type IV delayed-hypersensitivity, reaction.
Safety Considerations for Dermal Formulations 513
To minimize the risks of an excessive number of cutaneous reactions in the

marketplace, pharmaceutical companies and cosmetic and toiletry manufacturers
have developed a series of safety tests to which products are subjected before launch.
Some of these tests are used for historical or legislative reasons and are not neces-
sarily the best scientifically. Many involve experimental animals. However, percu-
taneous penetration is an all-important determinant of a product’s potential for con-
tact dermatitis and, as human skin is unique because of its epidermal thickness and
barrier function, poor correlation between animal data and human experience is often
seen. As the trend is toward using fewer animal-based test procedures, for both
scientific and moral reasons, and because we believe that the most valid data is
gained from the species for which the treatment is designed, only human safety tests
will be discussed in this chapter.
III. PRETEST INFORMATION

Before any human testing is undertaken, every possible risk to the subjects must be
assessed. This evaluation must be based on all known information, be it animal or
in vitro based, that is present in the literature. To do otherwise would be considered
unethical. The major requirements before any substances can be topically applied to
humans are that they are not acute poisons or corrosive agents, and that there is no
evidence that they are potent sensitizers. Much of this information can be obtained
without the use of animal experiments. Various in vitro systems exist that will in-
dicate whether a new chemical is an acute poison or is corrosive. For example, cell
culture systems are widely used for screening (1). A system based on an epidermal
slice technique has also been used successfully as a method for identifying strong
irritants and corrosive materials (2). A more complex cell culture system, the skin
equivalent, has also been proposed for use as a screening procedure for topically
applied materials (3). In this system a three-dimensional reconstruction of human
skin is prepared, using human epidermal cells layered on top of a collagen lattice
containing fibroblasts. A ‘‘stratum corneum-like’’ structure is formed in this system
and materials, including those not normally suitable for conventional cell culture
procedures, can be applied directly to this surface. The release of inflammatory me-
diators, such as prostaglandins, from the dermal surface, thought to be central to the
development of cutaneous irritancy in vivo, can subsequently be monitored.
Once satisfactory preliminary safety data has been obtained using these or
similar in vitro systems, then it is possible to progress directly to human volunteer
studies to determine the potential for contact dermatitis. The sequence of testing
usually follows the route of irritancy tests followed by sensitization tests followed
by photosensitivity tests, if deemed necessary. This is a logical approach because
any product that is an irritant will probably not be developed further and the later
tests are more complex and costly to perform. Any tendency toward irritancy would
also cause problems in interpreting the results of other tests.
IV. IRRITANCY TESTS

Ethically, any human test procedure must be approached with the safety of the sub-
jects as a priority. To this end, procedures have been developed that minimize the
risk of serious cutaneous reactions (4,5). These procedures use a stepwise approach
of increasing times of exposure and bioengineering methods to minimize the danger

and discomfort to the subjects involved and to minimize the potential risk of irre-
versible skin damage. In most situations, however, the product being tested is not
novel and is related to previously studied formulations. In this situation, the irritant
potential can be determined by simple patch tests, such as the 48-h patch test or the
2- to 3-week cumulative irritancy test.
A. 48-Hour Irritancy Tests

The 48-h irritancy test consists of two sequential 24-h applications under occlusion
to the same site. The break in the application is to premit assessment and possible
termination of the application, should unexpected severe reactions occur. Assess-
ments are made at the end of the 48-h period, and again at 72 h. The latter assessment
time serves to check that any reactions are subsiding satisfactorily and to catch any
late reactions that may develop only several hours after the removal of the patch. It
may be appropriate to include both negative and positive controls in this type of
test, particularly if experience of human testing with the class of product is minimal.
Negative controls usually consist of the empty application chamber or the adhesive
base without active agent. Positive controls may be known irritants, such as 0.5%
sodium lauryl sulfate. It may also be desirable to compare results with existing or
marketed products at this stage, so that if some reactions occur the relation to what
is currently used is clear. An important variable, which is not possible to exclude
from this type of test, is intraindividual variability in terms of cutaneous reactivity.
This may be due to an individual’s skin type, age, or hormonal status (6). To com-
pensate for this, a reasonable number of individuals, usually 25–30, must be used
in this type of study.
The 48-h irritancy test is sometimes referred to as a primary irritancy test. This
is based on the definition that a primary irritant produces a reaction on first contact,
and a secondary irritant produces a reaction only after repeated exposure (7). This
is confusing because, depending on the concentration, a product can be both a pri-
mary and secondary irritant. Irritant potential is always dependent on conditions of
exposure and is never absolute. Prolonged application of something apparently as
innocuous as water can produce a dermatitis, given the right circumstances.
The interpretation of the results of the 48-h irritancy test is important. A neg-
ative result does not necessarily mean that the product is safe to use. Many irritants
reveal themselves only after repeated application. As many products are, by virtue
of their function, intended to be repeatedly applied to the skin, this short-term type
of test is patently inadequate as a safety test. The 48-h test does however serve as
a simple screening device and this is its primary function. It will indicate those
products with a high potential for causing cutaneous irritancy. Several products can
be tested at the same time, and information can be obtained quickly during product
development.
B. Cumulative Irritancy Tests

Cumulative irritancy assays permit a more-sophisticated ranking of products and
enable the comparison and classification of the weaker types of irritants, into which
category most topical formulations and transdermal systems fall. There are several
variations on the cumulative irritancy assay (8–10). The main variations are the type
of application patch, the number of subjects, and the duration of the study. The
central feature of these assays is that the product is repeatedly applied to the same
site such that the exposure is exaggerated. In addition, the site is occluded, such that
the barrier function of the skin is compromised, allowing greater penetration of
potential irritants. By virtue of this exaggerated exposure, it is hoped that the infor-
mation gained from a relatively small panel of individuals (usually 25–30 subjects)
will predict the likely behavior under conditions of everyday consumer use.
The cumulative irritancy assay normally consists of either a 14- or 21-day
application period. There is still some debate over whether the shorter time period
is sufficient to detect very weak irritants, and the decision between the two time
periods is usually made according to the product being tested and its proposed usage.
Thus, for products that are in prolonged daily contact with the skin (e.g., dermal and
transdermal therapeutic systems), the longer test procedure may be desirable to give
the necessary reassurance of product safety.
Another variable in the cumulative assay is the number of applications over
the test period. Some protocols call for daily applications and assessments (Monday
to Friday, but not weekends), others for application and assessment every 2 to 3 days
(Monday, Wednesday, and Friday). In both protocols the subjects continue to wear
the patches over the weekend. There is very little to choose between the two sched-
ules because both give continuous contact over the test period. The advantage of the
daily regimen is that if severe reactions develop quickly, then the application can be
stopped and, therefore the subject is exposed to minimal risk. However, this is not
likely with the types of products generally tested, especially if they have been sub-
jected to a 48-h test previously. The more frequent patch application and removal
schedule has the distinct disadvantage that trauma to the skin surface by repeatedly
applying and removing adhesive dressings may lead to severe ‘‘plaster’’ reactions
that mean that the subject is no longer able to continue with the study. This is
obviously more likely to happen in the 21-day protocol.
The nature of the application patch is also a variable in this assay. Many of
the original assays used gauze pads (Webril pads) covered with impermeable tape
such as Blenderm. We prefer an alternative system, 12-mm aluminum Finn chambers
on Scanpore tape. The advantages of this system are as follows: Scanpore tape is
not occlusive and minimizes the ‘‘plaster’’ reactions seen with occlusive tapes, such
as Blenderm; the chambers have a fixed volume and when used with filter papers
can be used for liquids; an impression of the raised rim is clearly seen on removal
of the patch, giving the reassurance that occlusion was continuous and that the sub-
ject had complied with the conditions of the study. The chambers are available
commercially as strips of chambers at fixed intervals that are perforated to allow
subdivision. A possible disadvantage of the Finn chamber system is that the area of
application (12-mm diameter = 113 mm2) is less than the gauze pad. However,
although the area of application is important in terms of the intensity of the response
(11), above a certain level (approximately 100 mm2) this effect is minimal.
With transdermal systems the patch itself is occlusive, and by its nature will
produce cutaneous reactions such as folliculitis. An additional problem when testing
transdermal systems is the pharmacological load on the subject. Topical formulations
for treating skin diseases are designed to deliver relatively small amounts of drug
locally. In contrast, transdermal systems are designed to deliver large doses syste-
matically through a relatively small area of skin. The amount absorbed from a topical
agent in a 12-mm–Finn chamber is usually negligible as far as the wellbeing of the

subject is concerned. In contrast, a transdermal system will deliver a large dose over
a short time period, and this must be taken into account when designing the study.
Thus, a transdermal estrogen product is ideally tested for irritancy in postmenopausal
women. In this situation the study almost becomes a phase I clinical study and pre-
and poststudy medical screening is advisable. An alternative approach is to reduce
the size of the transdermal patch, such that the pharmacological dose carries negli-
gible risk, but the area is such that the irritant properties can be assessed.
Evaluation and interpretation of any irritant response also varies according to
the test procedure. Grading scales for erythema vary from the simplest 0–4 scale
(none, mild, moderate, severe), to more complex scales with half-point divisions.
The scale we used (Table 1) has evolved over several years experience in assessing
these types of reaction. The irritancy potential of a product is generally assessed by
the number and severity of the reactions occurring during the study, in particular the
number of grade 2 or higher reactions. A cumulative index of irritancy can be de-
rived, and this can be related to previous assays and assays on marketed or competitor
products. An alternative approach is to determine the time at which half of the
subjects have reacted to a product in a procedure analogous to the LD50 determination
(8,12). A range of concentrations needs to be tested and the proportion of subjects
reacting to each concentration is plotted against the time of reaction and, by a curve-
fitting procedure, the time when half the subjects have shown a reaction is recorded.
A concentration at which half of the subjects react in this type of test can be deter-
mined and absolute comparisons made.
C. Facial-Stinging Test
Products may pass tests such as the cumulative irritancy test, but still cause problems
for the consumer. Disagreeable sensations may occur, particularly when products are
applied to sensitive areas such as the face. A not infrequent complaint is that of
stinging or burning or itching after application. Signs of irritation, such as erythema
or scaliness, may not necessarily occur in this situation. Part of the reason for such
sensations is that facial skin is very permeable and has a rich nerve supply. It is also
exposed to weathering and a constant bombardment of cosmetics and cleansers.
These may all contribute to increased sensitivity of this area. A method to assess the
stinging capacity of topical materials has been described (13). Subjects who are
sensitive to the stinging sensation of lactic acid are selected as panelists for testing
Table 1 Human Patch Test-Grading Scale
0 No reaction
0.5 Slight, patchy erythema
1 Slight uniform erythema
2 Moderate, uniform erythema
3 Strong erythema
4 Strong erythema, spreading outside patch
5 Strong erythema, spreading outside patch with either swelling or vesiculation
6 Severe reaction with erosion
new products. Their subjective responses are noted, and a cumulative score is derived
that is considered indicative of the potential for stinging in the general population.
V. SENSITIZATION TESTS
Allergic reactions to topically applied materials are much less common than irritant
reactions. Many sensitizers will produce allergy only in a small percentage of the
population, possibly less than 1%. As such, it is difficult to devise a test that will
accurately reflect the incidence of sensitization when used by the population at large.
The need for predictive tests was recognized more than 30 yeas ago, and the essential
features of such tests were established (14). More recent examples of sensitization
protocols indicate only small variations on the original procedures (15–17).
Essentially all protocols are based on the guinea pig assays developed to predict
allergens and the original human studies by Draize et al. (18). Thus, there is an
induction period, during which repeated exposure to the product occurs. This is
followed 1–2 weeks later by a challenge to a previously unexposed site to determine
sensitization. The test site for the induction phase is usually the back, although the
upper arm is favored in some laboratories. The challenge should take place on pre-
viously unexposed skin, either at a site on the back, well removed from the original
test site, or preferably on the upper arm. A control application of either vehicle
without the active ingredient or a transdermal patch without drug is essential in
interpreting any reactions that occur.
A. Induction Phase
The induction phase varies according to the protocol, but essentially consists of
multiple applications, under occlusion, that are either continuous, as in the cumula-
tive irritancy protocol, or intermittent (24-h on–24-h off). Typical examples are as
follows:
1. An induction period of 42 days consisting of 21 ⫻ 48-h applications, each
to a fresh site, as used for a transdermal nicotine patch (16).
2. An induction period of 21 days consisting of 9 ⫻ 24-h applications, 24-h
on–24-h off, applications to the same site (17).
3. An induction period of 15 days consisting of 6 ⫻ 48–72-h applications
to the same site (i.e., continuous application with inspection and reap-
plication).
We favor the latter variation because it gives continuous exposure to the same
site and is probably equivalent, in terms of the immunological challenge, to an
intermittent exposure over a longer period or to the longer exposure to fresh sites.
Continuous exposure to the same site under occlusion maximizes the percutaneous
penetration of any potential allergen. Intermittent exposure allows for barrier recov-
ery and, therefore, may not be as severe a provocation.
A major variation to this induction procedure is the use of chemical agents to
damage the skin before patch testing. The most common departure is the use of
sodium lauryl sulfate in the maximization procedure (19). Briefly, an area of skin is
treated with a 5% sodium lauryl sulfate solution for 24 h to induce a moderate
inflammatory response. This is followed by a 48-h application of the test material to
the same site. The sequence of 24-h irritant and 48-h test material application is
repeated five times to the same site. This procedure is designed to maximize pene-
tration of potential allergens. However, severe cutaneous reactions can develop with
this approach, and this may be unacceptable to volunteer subjects. In addition the
possibility of interaction between any sodium lauryl sulfate that has penetrated the
skin and the test material cannot be ruled out. Such an interaction may alter the
immunogenic potential of the test material.
B. Challenge Phase
In the challenge phase there is less variation in protocols. The main differences are
the use of a single 48-h challenge or two consecutive 48-h challenges (back-to-back).
The latter is claimed to be more sensitive in terms of detecting weak allergens (20).
What is perhaps more critical is the use of appropriate controls and the time of
assessment. Controls consisting of vehicle or transdermal system without active drug
must be tested alongside the product of interest to interpret any reactions. At least
two observations should be made, one assessment within 15–30 min after patch
removal, and a second assessment 48 h later to determine late reactions. Sensitization
challenges are usually carried out under occlusion and must be at sites remote from
the initial induction site (e.g., the upper arm if the back has been used in the induction
phase). The only other variation is the maximization procedure, which again uses
pretreatment of the challenge site with sodium lauryl sulfate solutions.
C. Population Size
The number of subjects used in sensitization procedures varies from 25, in the max-
imization test, to 200 in the Draize test. The sample size must be large enough to
give confidence, in terms of predicting what will happen in general use, but not so
large that adoption of the procedure becomes expensive and time-consuming. The
mathematics of extrapolating from a small test population to many consumers has
been discussed (21). Briefly, if there are no reactions in a panel of 200 subjects, then
as many as 15 out of every 1000 of the general population could react (95% con-
fidence limits). For 100 and 50 subjects the figures are 30 and 58 out of every 1000
of the population. The possibility that the sensitization test may not predict a potential
sensitizer may result from a variety of factors, such as skin site, variation in the
frequency of application during the induction phase, poor release from the vehicle,
and the dose applied. Clearly, to reduce the level of potential reactors in the general
population to fewer than 1:1000 would require an enormous number of volunteers.
The compromise is to use 100 or 200 subjects, which gives a reasonable level of
confidence in the test. The choice of 100 or 200 subjects rests, to a certain extent,
on the type of product. A cosmetic product that may be a reformulation or that does
not contain any new chemical entities may require only 100 subjects. It may be more
prudent to test a new pharmaceutical agent in a panel of 200 subjects. Any possible
regulatory requirement of countries where the product is to be sold must also be
taken into account.
D. Dosage
Undoubtedly the induction and elicitation of an contact sensitization reaction is dose-
dependent. The higher the concentration, the more likely it is that individuals who
are predisposed will become sensitized. With final formulations the test concentration
is predetermined. When a range of therapeutic doses is being considered, the highest
dose likely to be used clinically should be tested. With raw ingredients, the highest
nonirritating concentration may be tested, depending on the likely exposure. As with
irritancy, the area over which an allergen is applied is important. With a potent
sensitizer, such as 2,4-dinitrochlorobenzene (DNCB), higher dose–response curves
were obtained for a constant dose per unit area when the area of application was
increased from 8 to 80 mm2 (22). However, in a separate study with DNCB, at a
similar dose per unit area, no difference was found between areas of 180 and 710
mm2, neither in terms of the number of subjects sensitized nor in the challenge dose–
response curves (23). Thus, with the potent sensitizer DNCB, the critical surface
threshold area appears to be between 80 and 180 mm2.
VI. PHOTOSENSITIVITY
It is possible that topically applied or systemically administered chemicals or drugs
that have little or no potential to promote an irritant or allergic reaction in the skin,
may do so in the presence of sunlight. Although photosensitive reactions may be
rare, relative to irritant or sensitization reactions, the development of such reactions
can lead to withdrawal from the market (e.g., the antirheumatic drug, benoxaprofen)
(24).
A. Sunlight
Electromagnetic radiation is classified according to wavelength ranges. Electromag-
netic radiation emitted by the sun ranges from very short wavelength cosmic rays to
very long wavelength radio waves. Short-wavelength, high-energy forms of radiation
such as gamma and x-rays, are termed ionizing radiation, whereas longer lower-
energy radiation (more than 800 nm) is nonionizing and increases molecular motion,
giving a thermal effect. Wavelengths between 200 and 800 nm are capable of causing
chemical changes if absorbed by molecules in the skin. Ultraviolet radiation falls
into this category and is between 100 and 400 nm. Fortunately for us, no wavelengths
shorter than about 288 nm reach the earth’s surface, because these are filtered out
by the earth’s atmospheric ozone layer. Damage to this important protective layer by
man-made chemicals such as chlorofluorocarbons (CFCs) is currently under review
and the subject of much discussion. UVR is subdivided into UVC (100–280 nm),
UVB (280–320 nm), and UVA (320–400 nm). UVA can be further subdivided into
UVA II (320–340 nm) and UVA I (340–400 nm) (25) (Table 2).
Table 2 Nonionizing Radiation
UVC 100–280 nma

UVB 280–320 nm
UVA 320–400 nm
UVA II 320–340 nm
UVA I 340–400 nm
Visible 400–800 nm
a
Does not reach the earth’s surface.
Although sunlight that reaches the earth is essential to most living organisms
and has many beneficial effects for humans, it is UVB and UVA that are responsible
for most of the cutaneous photobiological events (26). These include DNA and RNA
damage, inhibition of protein synthesis, damage to liposomal and cellular mem-
branes, mutagenic and carcinogenic effects and erythematous responses. Chronic UV
exposure is also responsible for dermal connective tissue injury (elastosis) leading
to the changes known as photoaging (27).
B. UV-Induced Erythema
It is important to remember that sunlight is a continuous spectrum and that it is in
its entirety that it causes erythema or sunburn in exposed skin. The paler the skin
the greater the erythematous response for any given dose of UVR. Thus, skin type-
I burns far more readily than skin type-VI (Table 3).
UVB is primarily responsible for most of the erythema seen in human skin
following excessive sun exposure. This response is apparent approximately 6 h fol-
lowing exposure and is maximal at 20–24 h. In contrast to UVB, UVA can induce
an immediate erythema that usually diminishes within 2 h of exposure (28). It can
also give a delayed erythematous response that reaches a peak at 6 h. UVA contrib-
utes to about 10–20% of sunburn and passes through window glass, in contrast to
UVB, which is essentially blocked by window glass. UVA can also induce an im-
mediate pigment darkening response (IPD) (29). This reaction appears within seconds
of exposure and fades within a few minutes. The pigment darkening is due to a
photochemical reaction involving the oxidation of a low molecular weight form of
melanin in melanosomes. The response is clearest in skin type-III, -IV, or -V indi-
viduals who have higher levels of melanin. Prolonged exposure to UVA can lead to
persistent pigment darkening (PPD) or tanning of the skin (30).
The effects described so far are the results of UVR itself; however, photosen-
sitivity reactions can occur when an exogenous or endogenous chemical (chromo-
phore) absorbs UVR or visible light. These reactions may be either phototoxic (pho-
toirritant) or photosensitive (photoallergic).
C. Phototoxicity Reactions
Phototoxicity refers to skin irritation that is produced through the interaction of
chemical substances and radiant energy in the ultraviolet and visible ranges. Pho-
totoxic or photoirritant effects are immediate and nonimmunological. When testing
for the phototoxic potential of topically applied chemicals, the output from the re-
quired radiation source is UVA only. This is best accomplished by using a suitable
Table 3 Photo–Skin Types
Skin type I Always burns and never tans

Skin type II Always burns and tans with difficulty
Skin type III Often burns and tans moderately
Skin type IV Burns minimally and tans easily
Skin type V Rarely burns and tans profusely
Skin type VI Insensitive, never burns, deeply pigmented
filtered solar simulator. The clinical identification of phototoxic reactions in humans

relies on both morphology and the clinical evidence or suspicion of the presence of
possible phototoxic chemicals. Phototoxicity usually causes erythema or even bullae,
increased skin temperature and pruritus and is followed by hyperpigmentation that,
in some instances, may be long-lasting.
A range of phototoxic agents have been reported in the literature, many of
which are naturally occurring plant substances, such as the furocoumarins, including
the psoralens (31). Psoralens occur in a variety of plants, such as parsley, celery, and
citrus fruits. A plant-induced phototoxic reaction is known as phytophotodermatitis.
Perfumes that contain bergapton, which is a component of bergamot oil, a well
known photoirritant, can be phototoxic. Other phototoxic agents include coal-tar
derivatives, pyrene, anthracene, and fluoranthene (32). The cardiac antiarrhythmic
agent amiodarone has produced phototoxic effects (33), as have quinoline antima-
larial drugs (34). Tetracyclines, including demethylchlortetracycline, doxycycline,
and chlortetracycline may also prove phototoxic when taken orally (35). The thiazide
diuretics have also exhibited a phototoxic potential in experiments involving cardi-
ovascular patients with hypertension and heart failure (36). Some nonsteroidal anti-
inflammatory drugs (NSAIDs) have a phototoxic potential, both when administered
orally or topically (37). This is particularly true of propionic acid-related NSAIDs,
that produce a unique wheal-and-flare response (38,39).
Most of the information gained on the underlying mechanisms of phototoxic
reactions has been derived from animal or in vitro models. These have included
Candida albicans (40,41), photohemolysis of red blood cells, and isolated normal
human fibroblasts (42). Such in vitro models are, in our opinion, unreliable, as they
lack the complexity of the whole animal. Recently, a human skin equivalent model
has been used in an attempt to predict the phototoxic potential of topically applied
personal care products (43). Known irritants and photoirritants were applied to the
epidermal side of the ‘‘skin’’ for up to 24 h and then exposed to 2.9 J/cm2 of UVA.
The MTT reduction assay was used to assess cytotoxicity. The authors claim an 83%
prediction with known phototoxic agents using this method. However, false-negative
results were seen with well-known phototoxic agents such as amiodarone hydro-
chloride, 6-methycoumarin, bithionol, and piroxicam. Methods such as these may
become more accurate predictors if they rely on more than one marker of cell damage
for the prediction of chemical irritancy or, in this case, photoirritancy. It is our belief
that the most reliable results are obtained when using the whole animal as a model.
If the goal is to predict phototoxicity in humans, then the preferable test method
should also use, when possible, human subjects.
D. Phototoxicity Assays
The development of procedures for assessing both the phototoxic (and photoallergic)
potential of chemicals has, in general, been carried out in laboratory animals. A wide
range of animal species have been used, including rabbits, mice, guinea pigs, squirrel
monkeys, opossums, and swine. All methods require the skin to be irradiated with
UVA following application of the test substance. However the measurable end result
is not always similar to that obtained from human experiments. Increases in guinea
pig (44) or mouse ear thickness (45,46) has been used to quantify phototoxic re-
sponses. Dermatitis and the increase in weight of the mouse tail has also been used
(47). However, hairless mice and albino guinea pigs have been used where simple
erythema was the toxicological endpoint for assessing a phototoxic response (48).
The results obtained from such experiments may be false-negative or false-positive
when extrapolated to humans.
Kaidbey and Kligman (49) suggested a method for identifying potential topical
phototoxic agents in humans, although the title of this publication misguidedly refers
to photosensitizing, rather than phototoxic agents. Human testing is ethical as only
small areas of skin are irradiated and clinical experience of phototoxic reactions
indicate that when the stimulus is removed the dermatitis subsides. As with all clin-
ical trials the informed consent of the subject and ethical committee approval must
be obtained. The method we recommended is based on that of Kaidbey and Kligman
and is suitable for assessing the phototoxic potential of topically applied drugs, chem-
icals, transdermal, and skin care products.
In the protocol for phototoxicity testing, a test panel consisting of a minimum
of 12 healthy white adults with untanned back skin is required. As the dermatitis of
a phototoxic agent can be produced in almost every subject, given sufficient expo-
sure, it is not necessary to employ a large test panel. Two sets of test products and
appropriate controls (usually vehicle or base alone) are applied occlusively to the
midback area (one set on the left and one set on the right). In our experience 12-
mm aluminum Finn chambers on Scanpore tape are suitable for product application.
They are occlusive and ensure optimal product contact with the skin. An empty
chamber should also be fixed to each side of the back. Six hours after administration
of the materials and chambers, these should be removed and one set of applications
only irradiated immediately with UVA. The chemical-to-skin contact time is rela-
tively short compared with, for example, the 24 h needed in a vasoconstriction assay.
If longer times are used (e.g., 24 h), light exposure will often produce negative
results. The test sites should be irradiated with 20 J/cm2 of UVA. The radiation source
should ideally be a xenon arc lamp solar simulator.
Such a system could typically consist of a 1000-W ozone-free xenon arc lamp,
the output from which is filtered with a Schott WG 345 filter of 2-mm thickness.
This filter will block all erythemogenic UVC and UVB wavelengths below 320 nm.
In addition unwanted longer wavelength visible and infrared radiation can be re-
moved using a combination of a suitably coated dichroic mirror, water filter, and
UG11 filter. If this is not accomplished then the subjects may feel heat and or pain
from the irradiations. It is often convenient to deliver the UVA radiation to the skin
surface using an 8-mm–diameter liquid light guide. A broad-spectrum thermopile
should be used to measure the output of energy from the solar simulator, typically
expressed in milliwatt per square centimeter (mW/cm2). The thermopile should be
calibrated against a known standard from, for example, The National Physics Lab-
oratory in the United Kingdom. To calculate the time of irradiation necessary to
administer a dose of 20 J/cm2 to the skin, the formula in Table 4 should be used.
Skin assessments should be made immediately following irradiation and at 24
and 48 h after photoexposure. The grading system in Table 5 is suitable for recording
any cutaneous reactions. The irradiated sites should be compared in each subject
with the nonirradiated sites. If the response in any one subject at the irradiated site
is greater than that seen at the nonirradiated site, then that product or chemical is
deemed phototoxic. Phototoxicity is relatively easy to detect and therefore prevent;
Table 4 Calculation of Irradiation Times
mJ/cm2
t(s) =
mW/cm2
Therefore, if a reading of 200 mW/cm2 is obtained
from the thermopile and a dose of 20 J/cm2 (20,000
mJ/cm2) is required then:
20,000
t(s) =
200
and
t(s) = 100 s
however, phototoxic reactions may sometimes mask or contribute to photoallergic

reactions.
E. Photoallergic Reactions
Photoallergy (or photosensitization) refers to a dermatitis that is produced through
the interaction of chemical substances and radiant energy, in the UVR and visible
wavelengths, to produce an allergen. The chemical substance may be orally ingested
or topically applied (photocontact allergy). Unlike phototoxic reactions, photoallergic
reactions are often delayed, immunological, and less dose-dependent. Photoallergic
and particularly photocontact photoallergic reactions are relatively uncommon when
compared with phototoxic reactions. Clinically, photocontact allergic reactions pro-
duce a dermatitis that resembles allergic contact dermatitis, appearing as an acute
dermatitis affecting primarily, but not exclusively, light-exposed skin. A characteristic
histological feature of photocontact allergy is a dense perivascular round cell infiltrate
in the dermis, which helps distinguish this dermatitis from a phototoxic reaction (50).
A second and rare type of photoallergic reaction is solar urticaria (51). This occurs
after only brief exposure to light and is characterized by an immediate urticarial
wheal-and-flare reaction within minutes of exposure. The reaction usually subsides
within 1–2 h, is associated with degranulation of mast cells at the site of exposure,
and the release of neutrophil chemotactic factors and histamine into venous blood
near the reaction sites. Photoallergic reactions, when they occur, may apparently be
Table 5 Phototoxicity Grading Scale
Grade Cutaneous rection
0 No reaction
1 Mild erythema, possibly with scaling
2 Moderate or strong erythema
3 Moderate or strong erythema with a papular response
4 As grade 3, but with definite edema
5 Vesicular or bullous eruption
triggered by irradiation alone (in the absence of known sensitizers) or may be due
to exogenous chemicals and UVR. Photoallergy, whether photocontact allergic der-
matitis (delayed hypersensitivity: type IV reaction) or solar urticaria (immediate hy-
persensitivity: type I reaction), is an acquired reactivity dependent on cell-mediated
hypersensitivity or antigen–antibody interaction (52).
Test procedures designed to identify potential photosensitizing chemicals were
developed in response to the outbreak of reactions caused by the use of antibacterial
halogenated salicylanilides in the early 1960s (53). A minority of affected individuals
developed a persistent photodermatitis that lasted several years despite the avoidance
of contact with photosensitizing phenolic compounds (54). It therefore became clear
that there was a requirement for a laboratory test to detect potential photosensitizing
agents and avoid such situations.
Several photocontact sensitizers have been identified, including coumarins and
coumarin derivatives, musk ambrette, fentichlor, bromochlorosalicylanilide, chloro-
2-phenylphenol, and benzocaine (55–57). Certain sunscreens have also been reported
to produce photocontact allergic dermatitis, most notably para-aminobenzoic acid
and derivatives, benzophenone 3, mexenone, and cinnamates (58–60). Several es-
sential oils have also produced photoallergic reactions (e.g., sandalwood oil; 61).
Also, 8-methoxypsoralen, which is used in PUVA (psoralen ⫹ UVA), can also be
photoallergic as well as phototoxic (62).
F. Photoallergenicity Assays
Test procedures to identify potential photosensitizers have received less attention than
methods designed to detect either ordinary contact sensitizers or phototoxic chemi-
cals. Landsteiner and Chase (63) demonstrated that low molecular weight haptens
can produce contact dermatitis in guinea pigs. They also observed that allergic con-
tact dermatitis could be conferred on immunologically naive guinea pigs by the
passive transfer of mononuclear cells from nonsensitized animals. Furthermore
guinea pigs develop edema and erythema after contact with topically applied sen-
sitizers and, to some extent, develop a response similar to the clinical response in
humans. These observations became the cornerstone of photocontact allergic der-
matitis research over many years and led to the guinea pig becoming the most
commonly used animal in photoallergy studies (64,65). On the other hand, mouse
ear swelling is claimed to be a more sensitive model (66), but is even further removed
from the human response than that of the guinea pig. To induce contact photosen-
sitivity in any animal it has to be repeatedly exposed to the test molecule in the
presence of UVR. For this induction phase a broad-spectrum source is necessary,
which should include UVB as well as UVA. The period of induction should be
similar to that for testing contact sensitizers, followed by a rest period and then a
challenge on a previously untested site with the test chemical and UVA alone.
The photomaximization test for the prediction of photosensitizers is conducted
on humans and is similar in design to a sensitization test, but with the addition of
exaggerated UVR exposure of both the chemical and the skin to which it is applied.
In an ideal world, this type of test would be carried out on a large number of subjects
(>100) to more accurately predict the incidence of photosensitization reactions in the
population at large. From a practical point of view this is not possible because of
the demanding nature of the protocol; therefore, a test panel of 26 is normally rec-
ommended. The method described here is based on that of Kaidbey and Kligman
(67) and is suitable for identifying topical photocontact sensitizers.
The photomaximization test is a 6-week study and is divided into an induction
phase and an elicitation phase. A solar simulator, as described previously, is an ideal
source of UVR. During the induction phase a 1-mm–WG320 filter and a 2-mm–
UG11 filter should be used, allowing both UVA and UVB (290–400 nm) to reach
the subject’s skin. For the elicitation phase, a 2-mm–WG345 and a 1-mm–UG11
filter should be used to allow only UVA (320–400 nm) to reach the skin. The test
chemical together with the vehicle control is applied occlusively to the midback of
each subject using 12-mm aluminum Finn chambers on Scanpore tape. Twenty-four
hours later, the patches are removed and the test sites wiped clean and allowed to
air-dry for approximately 30 min. Each site is then exposed to twice the subjects
minimal erythema dose (MED) of solar-simulated UVR. The sites are then left un-
covered and exposed to the air for approximately 48 h. This procedure of application
and irradiation is repeated such that each subject has six applications and irradiations
over a 6-wk period. The sites are evaluated 24 h after each irradiation and, if the
reaction becomes severe such that further application and irradiation is undesirable,
then application of the material and subsequent irradiation is carried out at an ad-
jacent site.
Following completion of the induction phase, there is a 2-wk rest period, with
no applications or irradiations. Approximately 14 days after the end of the induction
phase, two sets of test materials are applied to previously untreated sites on the
midback, again using 12-mm–Finn chambers. Twenty-four hours later the patches
are removed, the skin wiped dry with a gauze swab, and one set of applications
irradiated with only 4 J/cm2 of UVA. The sites are then evaluated at 24, 48, and 72
h after the elicitation irradiation. The grading system for all irradiations (both in-
duction and elicitation) can be seen in Table 6. If one or more subjects develops a
Table 6 Grading System for Photoallergy Test: Induction and Elicitation Phase
0 No reaction.
1 Reaction readily visible, but mild unless letter grade appended (see E and F
grades). Mild reactions include weak, but definite erythema, and weak
superficial skin responses such as glazing, cracking, or peeling.
2 Definite papular response (appended E, F, or S if appropriate).
3 Definite edema (appended E, F, or S if appropriate).
4 Definite edema and papules (appended E, F, or S if appropriate).
5 Vesicular–bullous eruption (appended E, F, or S if appropriate).
E Strong erythema at patch site.
F Strong effects on superficial layers of the skin including fissures, a film of
dried serous exudate, small petechial erosions, or scabs.
S Reaction spreading beyond test site.
I Itching.
B/S Burning or stinging.
(Applications must be either terminated or moved to an adjacent nonirradiated site if a reaction score of
2 or higher occurs).
Descriptive letter designations may be added to the numerical score if experienced at the test site.
(Any other signs or symptoms (e.g., wheal-and-flare responses) may be described separately).
reaction at an irradiated site during the elicitation phase that is greater than the
corresponding unirradiated site, then that chemical is considered to be a photosen-
sitizer. In practical terms there are usually many or no reactors in a test panel making
the decision as to whether a product is a photoallergen relatively easy.
REFERENCES
1. Duffy PA, Flint OP. In vitro dermal irritancy tests. In: Atterwill C, Steele C, eds. In
Vitro Methods in Toxicology. London: Cambridge University Press, p 279, 1987.
2. Oliver GJA, Pemberton MA. An in vitro epidermal slice technique for identifying chem-
icals with potential for severe cutaneous effects. Food Chem Toxicol 23:229, 1985.
3. Dykes PJ, Edwards MJ, O’Donovan MR, Merrett V, Morgan HE, Marks R. In vitro
reconstruction of human skin: the use of skin equivalents as potential indicators of
cutaneous toxicity. Toxicol In Vitro 5:1, 1991.
4. Basketter DA, Whittle E, Griffiths HA, York M. The identification and classification of
skin irritation hazard by a human patch test. Food Chem Toxicol 32:769, 1994.
5. Dykes PJ, Black DR, York M, Dickens AD, Marks R. A stepwise procedure for evalu-
ating irritant materials in normal volunteer subjects. Toxicol In Vitro 14:204, 1995.
6. Patil S, Maibach H. Effect of age and sex on the elicitation of irritant contact dermatitis.
Contact Dermatitis 30:257, 1994.
7. Shelanski HA. Experiences with and considerations of the human patch test method. J
Soc Cosmet Chem 2:324–331, 1951.
8. Kligman AM, Wooding WM. A method for the measurement and evaluation of irritants
in human skin. J Invest Dermatol 49:78–94, 1967.
9. Phillips L, Steinberg M, Maibach HI, Akers WA. A comparison of the rabbit and human
skin response to certain irritants. Toxicol Appl Pharmacol 21:369, 1972.
10. Carabello FB. The design and interpretation of human skin irritation studies. J Toxicol
Cutan Ocul Toxicol 4:61–71, 1985.
11. Dykes PJ, Hill S, Marks R. The effect of area of application on the intensity of response
to a cutaneous irritant. Br J Dermatol 125:330, 1991.
12. Bahmer FA. In vivo assessments of irritants made simple and subject friendly. Contact
Dermatitis 33:210, 1995.
13. Frosch PJ, Kligman AM. A method for appraising the stinging capacity of topically
applied substances. J Soc Cosmet Chem 28:197, 1977.
14. Kligman AM. The identification of contact allergens by human assay. J Invest Dermatol
47:369–374, 1966.
15. Robinson MK, Stotts J, Danneman PJ, Nusair TL. A risk assessment procedure for
allergic contact sensitization. Food Chem Toxicol 27:479, 1989.
16. Jordan WP. Clinical evaluation of the contact sensitization potential of a transdermal
nicotine system (Nicoderm). J Fam Pract 34:709, 1992.
17. Gerberick GF, Robinson MK, Stotts J. An approach to allergic contact sensitization risk
assessment of new chemicals and product ingredients. Am J Contact Dermatitis 4:205,
1993.
18. Draize JH, Woodard G, Calvery HD. Methods for the study of the irritation and toxicity
of substances applied topically to the skin and mucous membranes. J Pharmacol Exp
Ther 83:377–390, 1944.
19. Kligman AM. The identification of contact allergens by human assay. III The maximi-
sation test: a procedure for screening and rating contact sensitizers. J Invest Dermatol
47:393–409, 1966.
20. Jordan WP. 24-, 48-, and 48/48-hour patch tests. Contact Dermatitis 6:151, 1980.
21. Henderson CR, Riley EC. Certain statistical considerations in patch testing. J Invest
Dermatol 6:227–232, 1945.
22. Rees JL, Friedmann PS, Matthews JNS. The influence of area of application on sensi-
tization by dinitrochlorobenzene. Br J Dermatol 122:29, 1990.
23. White SI, Friedmann PS, Moss C, Simpson JM. The effect of altering area of application
and dose per unit area on sensitization by DNCB. Br J Dermatol 115:663–668, 1986.
24. Allen B. Benoxaprofen and the skin. Br J Dermatol 109:361–364, 1983.
25. Fitzpatrick TB. Ultraviolet induced pigmentary changes: benefits and hazards. In:
Honigsmann H, Stingl G, eds. Therapeutic Photomedicine. Basel: Karger, pp 25–38,
1986.
26. Epstein JH. Photomedicine. In: Smith KC, ed. The Science of Photobiology, 2nd ed.
Boca Raton: CRC Press, 1988.
27. Pathak MA. Sunscreens: topical and systemic approaches for protection of human skin
against harmful effects of solar radiation. J Am Acad Dermatol 1:285, 1982.
28. Whitman GB, Leach EE, Deleo VA, Harber LC. Comparative response to UVA radiation
in humans and guinea pigs. In: Urbach F, Gange RW, eds. The Biological Effects of
UVA Radiation. New York: Praeger, pp 79–86, 1986.
29. Kaidbey KH, Barnes A. Determination of UVA protection factors by means of immediate
pigment darkening in normal skin. J Am Acad Dermatol 25:262, 1991.
30. Pathak MA. Activation of the melanocyte system by ultraviolet radiation and cell trans-
formation. Ann NY Acad Sci 453:328, 1985.
31. Pathak MA. Phytophotodermatitis. In: Pathak MA, Harber L, Seiji M, Kukita A, eds.
Sunlight and Man: Normal and Abnormal Photobiologic Responses. Tokyo: University
of Tokyo Press, pp 495–513, 1974.
32. Kochevar I, Armstrong RB, Einbinder J, Walther RR, Harber L. Coal tar phototoxicity:
active compounds and action spectra. Photochem Photobiol 36:65–69, 1982.
33. Rappersberger K, Honigsmann H, Ortel B, Tanew A, Konrad K, Wolff K. Photosensi-
tivity and hyperpigmentation in amiodarone treated patients: incidence, time course and
recovery. J Invest Dermatol 93:201–209, 1989.
34. Ljunggren B, Winestrand L. Phototoxic properties of quinine and quinidine: two quin-
oline methanol isomers. Photodermatology 5:133–138, 1988.
35. Bjellerup M, Ljunggren B. Photohaemolytic potency of tetracyclines. J Invest Dermatol
84:262–264, 1985.
36. Diffey BL, Langtry J. Phototoxic potential of thiazide diuretics in normal subjects. Arch
Dermatol 125:1355–1358, 1989.
37. Stern RS. Phototoxic reactions to piroxicam and other nonsteroidal anti-inflammatory
agents. N Engl J Med 309:186–187, 1983.
38. Diffey BL, Daymond JJ, Fairgreaves H. Phototoxic reactions to piroxicam, naproxen,
and tiaprofenic acid. Br J Rhematol 22:239–242, 1983.
39. Kaidbey K, Mitchell F. Photosensitizing potential of certain non-steroidal antiinflam-
matory agents. Arch Dermatol 125:783–786, 1989.
40. Daniels F. A simple microbiological method for demonstrating phototoxic compounds.
41. Mitchell JC. Psoralen type phototoxicity of tetramethylthiur-ammonosulphide for Can-
dida albicans: not for man or mouse. J Invest Dermatol 56:340, 1971.
42. Lock SO, Friend JV. Interaction of ultraviolet light, chemicals and cultured mammalian
cells: photobiological reactions of halogenated antiseptics, drugs and dyes. Int J Cosmet
Sci 5:39–49, 1983.
43. Edwards SM, Donelly TA, Sayre RM, Rheims LA. Quantitative in vitro assessment of
phototoxicity using a human skin model, Skin2. Photodermatol Photoimmunol Photomed
10:111, 1994.
44. Stott CW, Stasse J, Bonomo R, Campbell AH. Evaluation of the phototoxic potential of
topically applied agents using longwave ultraviolet light. J Invest Dermatol 55:335–338,
1970.
45. Cole C, Sambuco C, Forbes P, Davies R. Response to ultraviolet radiation: ear swelling
in hairless mice. Photodermatology 1:114–118, 1984.
46. Gerberick G, Ryan C. A predictive mouse ear-swelling model for investigating topical
phototoxicity. Food Chem Toxicol 27:813–819, 1989.
47. Ljunggren B, Muller H. Phototoxic reaction to chlorpromazine as studied with the quan-
titative mouse tail technique. Acta Derm Venereol 56:373, 1976.
48. Forbes PD, Urbach F, Davies RE. Phototoxicity testing of fragrance materials. Food
Cosmet Toxicol 15:55–60, 1977.
49. Kaidbey KH, Kligman AM. Identification of topical photosensitizing agents in humans.
J Invest Dermatol 70:149, 1978.
50. Epstein JH. Photoallergy and photoimmunology. In: Stone J, ed. Dermatologic Immu-
nology and Allergy. St Louis: CV Mosby, 1985.
51. Sams WM. Solar urticaria: studies of the active serum factor. J Allergy Clin Immunol
45:295, 1970.
52. Epstein JH. Phototoxicity and photoallergy in man. J Am Acad Dermatol 8:141, 1983.
53. Wilkinson DS. Photodermatitis due to tetrachlorosalicylanilide. Br J Dermatol 73:213,
1961.
54. Smith SZ, Epstein JH, Photocontact dermatitis to halogenated salicylanilides and related
compounds. Arch Dermatol 113:1372, 1977.
55. Kaidbey KH, Kligman AM. Photosensitization by coumarin derivatives: structure activ-
ity relationships. Arch Dermatol 117:258, 1981.
56. Raugi GJ, Storrs FJ, Larsen WG. Photoallergic contact dermatitis to mens perfume.
57. Epstein JH. Photoallergy: a review. Arch Dermatol 106:741, 1972.
58. Thune P. Contact and photocontact allergy to sunscreens. Photodermatology 1:5, 1984.
59. Szczurko C, Dompmartin A, Michel M, Moreau A, Leroy D. Photocontact allergy to
oxybenzone: ten years of experience. Photodermatol Photomed 10:144, 1994.
60. Parry EJ, Bilsland D, Morley WN. Photocontact allergy to 4-tert-butyl-4⬘-methoxy-di-
benzoylmethane (Parsol 1789). Contact Dermatitis 32:251, 1995.
61. Starke JC. Photoallergy to sandlewood oil. Arch Dermatol 96:62, 1967.
62. Fulton JE, Willis I. Photoallergy to methoxsalen. Arch Dermatol 98:455, 1968.
63. Landsteiner K, Chase MW. Experiments on transfer of cutaneous sensitivity to simple
compounds. Proc Soc Exp Biol Med 49:688, 1942.
64. Harber LC, DeLeo VA. Guinea pigs and mice as effective models for the prediction of
immunologically mediated contact photosensitivity. In: Maibach HI, Lowe NJ, eds. Mod-
els in Dermatology, Dermatopharmacology and Toxicology. New York: Karger, 1985.
65. Harber LC, Targovnik SE, Baer RW. Contact photosensitivity patterns to halogenated
salicylanilides in man and guinea pigs. Arch Dermatol 96:646, 1967.
66. Maguire HC, Kaidbey KH. Experimental photoallergic contact dermatitis: a mouse
model. J Invest Dermatol 73:147–152, 1982.
67. Kaidbey KH, Kligman AM. Photomaximization test for identifying photoallergic contact
sensitizers. Contact Dermatitis 6:161, 1980.
11
Transdermal Delivery and
Cutaneous Reactions
JAGDISH SINGH
North Dakota State University, Fargo, North Dakota
HOWARD I. MAIBACH
University of California School of Medicine, San Francisco, California
I. INTRODUCTION
Drugs and excipients have different sensitization capacities for inducing contact al-
lergy. The risk of skin reactions produced by chemicals depends on their inherent
allergenicity and ability to penetrate into the normal skin or damaged skin. As fully
described in earlier chapters, the penetration of chemicals into the skin depends on
skin condition, anatomical site, chemical characteristics of the substance, lipid sol-
ubility and concentration of the chemical. Penetration is also influenced by external
factors, especially solvents, surface-active agents, alkalies, moisture, temperature,
extreme dehydration, and mechanical effects. The length of time that a substance
contacts the skin is of great importance. Skin irritation influences the cells of the
skin and results in an increased sensitization risk. Such cell damage can be produced
by variety of chemicals or by mechanical means.
Irritation is the nonimmunological evocation of normal or exaggerated reaction
in a tissue by application of a stimulus. Irritation may be subjective or objective.
Subjective irritation refers to transient pruritus, stinging, burning, or related sensa-
tions without subsequent visible inflammation (e.g., alcohol on an open wound). A
chemical substance that evokes inflammation on initial exposure is called an acute
(primary) irritant but, on repeated exposure to an identical site, will cause cumulative
irritation. In the past, soaps, cosmetic materials, and pest control chemicals have
529
530 Singh and Maibach
been recognized as potential sources of cutaneous irritation. More recently it has

been recognized that a multitude of occupational and environmental factors, such as
organic dyes and solvents or industrial waste material contribute to this topical skin
disorder.
The most common reaction consists of a local inflammatory response charac-
terized by erythema or edema, or a corrosive reaction characterized by local tissue
destruction or necrosis. Other reactions, sometimes referred to as irritation, do not
display visible signs of inflammation. Subtle increases in epidermal thickness, with-
out visible or histological inflammation, may be produced by a variety of substances
usually thought to be nonirritating (1).
The occlusive nature of many transdermal delivery systems provides an ideal
model for inducing sensitization. Potential allergens include the adhesive, the dif-
fusion membrane, the solvent, the enhancer, and the drug. Allergic contact dermatitis
with redness, swelling, and sometimes vesiculation, is the most overt presentation
of skin sensitivity from transdermal delivery systems. The reaction is usually local-
ized to the site of application of the current patch, but may also occur at the sites
of previous applications, the flare-up reaction (2). Spread of the eczematous reaction
to sites not associated with the application of patches may occur (3). Urticaria and
angioedema are rare allergic reactions to transdermal therapeutic systems (4).
This chapter deals with the skin reactions caused by topical drug delivery
systems.
II. PREDICTIVE TESTING

A. Irritation
There is not yet an adequately validated in vitro model available to predict skin
irritation of topical chemicals. Details of the current state of development of these
potentially useful assays were summarized (5). The standard method to forecast skin
irritation is by so-called predictive tests on humans or animals. The most widely
used test for predicting potential skin irritants to humans, using animal models, was
published by Draize et al. (6), and has been refined by many groups (7). The test
was initially designed to classify chemicals that cause primary (acute) irritation.
However, in designing a test that would eliminate false-negative reactions (type 2
errors), Draize permitted the introduction of a significant number of false-positive
reactions (type 1 errors). The rabbit Draize test, properly performed and interpreted
by experienced scientists, still remains valuable.
Transepidermal water loss (TEWL) is a well-accepted method for quantifying
alterations in stratum corneum function (8), and it provides a robust method for
assessing stratum corneum damage. Irritation tends to reduce the efficiency of the
stratum corneum barrier function and may result in an increase in TEWL. This is
sometimes associated with a decrease in skin water content (9). Hence, measurement
of skin capacitance or skin hydration (10) may also be used to assess irritation
(11,12). Measurements of carbon dioxide emission from human skin can also be
used to determine the degree of irritation (13). Rates of carbon dioxide emission
from irritated skin increase roughly in proportion to the degree of irritation (14).
Transdermal Delivery and Cutaneous Reactions 531
Four techniques (skin color reflectance, TEWL, laser Doppler flow (LDF) mea-
surement, and visual scores) have been compared for their ability to quantify sodium
lauryl sulfate irritant dermatitis in humans (9). The study concluded that, although
TEWL measurements may be an accurate and sensitive method in evaluating skin
irritation when stratum corneum damage is present, color reflectance measurements
may be a useful complimentary tool in the evaluation of skin damage. Detailed
documentation on these bioengineering tools can be found in recent text books (15–
17).
B. Allergic Contact Dermatitis

Allergic contact dermatitis testing is widely performed, with both human subjects
and laboratory animals, to determine the irritant potential of various chemicals. The
oldest of these assays is the Draize guinea pig test. The Draize test with animal
models requires careful planning and performance. Buehler (18) and Magnusson and
Kligman (19) used five chemicals (benzocaine, formalin, monobenzyl ether of hy-
droquinone, potassium chromate, and tetrachlorosalicylanilide) to compare sensiti-
zation rate by the Draize test, closed patch test, and guinea pig maximization test
(GPM test). The percentage of sensitized animals was about 5% with the Draize test,
38% with the closed patch test, and 61% with the GPM test. This provides a rough
estimate of the relative capacity of the three techniques to identify contact sensitizers.
Marzulli et al. (20) tested eight compounds using various modifications of the
Draize guinea pig and human sensitization techniques (21). Skin sensitization was
observed both in humans and guinea pigs with p-phenylenediamine and dinitro-
chlorobenzene, and in humans, but not in guinea pigs, with neomycin, benzocaine,
hexachlorophene, furacin, and a mixture of methyl and propyl parabens. The authors
stated that a negative result with guinea pigs provide an insufficient basis for con-
cluding that a human is not likely to be sensitized by a substance.
The GPM test, the human maximization test,and the Draize repeat insult patch
test (22) have been used for extensive comparisons of contact sensitizers in guinea
pigs and humans (19). Contact sensitizers were rated in the five grades such that
weak (I), mild (II), moderate (III), strong (IV), and extensive (V) corresponded to
0–8%, 9–28%, 29–64%, 65–80%, and 81–100% sensitized. The results are given
in Table 1. Neither technique produced sensitization with very weak allergens, such
as hexachlorophene and lanolin. Substances that did not sensitize humans, such as
aluminum chloride, sodium lauryl sulfate, and polysorbate 80, did not sensitize the
guinea pigs either. Quantitative structure activity relations (QSAR) analysis provides
a powerful tool for predicting not only sensitization potential, but also how to define
appropriate testing parameters (23–26).
Guinea pig testing constitutes the first step in evaluating the allergenicity of
new compounds or products. There is a reasonable degree of correspondence between
the results obtained with the GPM test and the human maximization test. These two
tests rate the allergenicity of common human sensitizers in a similar fashion. Sub-
stances that sensitize in the human test also do so in the animal test (27). Detailed
discussions of animal and human sensitization assays and interpretation can be found
(28,29). Skin reactions from topical drug delivery systems, including chemicals, met-
als, and textiles, have been extensively investigated and may be found in series of
publications (30–73).
Table 1 Comparative Sensitization in Humans and Guinea Pigs by the Guinea Pig
Maximization Test and Human Maximization Test from the Following Chemicals
Guinea pig Human

maximization test maximization test
Chemicals Positive (%) Grade Positive (%) Grade
Aluminum chloride 0 I 0 I
Apresoline 80 IV 100 IV
Atabrine 90 V 78 IV
Benzocaine 28 II 22 II
Formalin 80 IV 72 IV
Hexachlorophene 0 I 0 I
Lanolin 0 I 0 I
Malathion 54 III 100 V
Mercaptobenzothiazole 40 III 38 III
Mercuric chloride 32 III 92 V
Monobenzyl ether of hydroquinone 50 III 92 V
Neomycin 72 IV 28 II
Nickel sulfate 55 III 48 III
Penicillin G 100 V 67 IV
Polysorbate 80 0 I 0 I
Potassium dichromate 75 IV 100 V
Sodium lauryl sulfate 0 I 0 I
Sulfathiazole 36 III 4 I
Streptomycin 72 IV 80 IV
Tetrachlorosalicylanilide 72 IV 88 IV
Turpentine 64 III 72 IV
Vioform 20 II 0 I
Source: Ref. 8.
III. REACTIONS TO DRUG DELIVERY SYSTEMS

A. Transdermal Therapeutic Systems
Transdermal drug delivery systems for systemic effect are feasible for small, potent,
and lipophilic drug molecules (74–76). Transdermal drug delivery systems are pres-
ently marketed in the United States for seven drugs (estradiol, clonidine, nitroglyc-
erin, scopolamine, nicotine, fentanyl, and testosterone), and others are under devel-
opment. As the drug is the most frequently identified allergen, human subjects can
be patch-tested with the drug at an appropriate concentration in a suitable vehicle
(Table 2). Skin irritation at the application site is the most common adverse effect
accompanying the use of transdermal therapeutic systems, occurring in as many as
5–24% of women (77–89). Although generally mild and transient, it appears to be
the most common reason for discontinuation of treatment during published efficacy
and tolerability studies (90). The following are adverse reactions related to the use
of commercially available transdermal drug delivery systems.
Table 2 Concentration of Drug in Appropriate Vehicle Used in Patch Tests
Drug Vehicle Concentration (%) Ref.
Clonidine Petrolatum 1 76
Petrolatum 9 77
Estradiol Petrolatum 5 75
Fentanyla
Nicotine Water 10 84
Nitroglycerin Petrolatum 2 79,80
Water 0.02 81–83
Scopolamine Petrolatum 1.8 4
Water 0.25 78
Testosterone Petrolatum 5
a
To be determined.
1. Estradiol
Estradiol is available as transdermal therapeutic systems, licensed for hormone re-
placement in postmenopausal women. After assessing data from those trials involving
more than 100 patients (and up to 15,194 patients), the reported incidence of skin
reactions to the transdermal therapeutic system was between 5 and 35% (91–96).
Most reactions consisted of mild erythema or pruritus at the application site, which
generally resolved after system removal. However, a small percentage of cases have
been of sufficient severity to cause patients to discontinue treatment. Erkkola et al.
(92) noted that 8.8% of patients withdrew from transdermal estradiol therapy because
of skin irritation, although the number of patients withdrawing from treatment for
this reason in other studies has been less than 5% (91,93,95–97). The most common
adverse effect observed using transdermal estradiol was local irritation at the appli-
cation site (98,99).
Similar results have been found in long-term (1-year) studies. Nachtigall (100)
reported skin irritation in 14% of 138 patients receiving transdermal estradiol ther-
apy; 3% of patients discontinued treatment for this reason. Randall (101) reported
on 29 patients, 10% withdrawing because of skin irritation.
Unpublished tolerability data involving 11,562 patients using estradiol trans-
dermal therapeutic systems have shown a comparable incidence of dermatological
adverse experiences. Treatment was either cyclic or continuous and, in some cases,
included concomitant oral administration of progestogen. Duration of treatment var-
ied from 8 to 52 weeks. Pooled results showed that, on average, the incidence of
local skin reactions was 14.2%. Skin reactions were the most commonly reported
adverse experience, accounting for 47% of all reported adverse experiences. These
reactions caused 6.3% of the patients, on average, to withdraw from treatment (data
on file, Ciba Geigy).
Several studies have specifically investigated the effects of the estradiol trans-
dermal therapeutic system on skin. In many cases, the cutaneous adverse effects
reported have been overcome by changing application site. Allergic contact dermatitis
has been induced by the components of the patch, as well as from the estrogen. The
components of the patch, such as adhesive (102), hydroxypropyl cellulose (103),
enhancer, such as alcohol, present in the reservoir (102,104); as well as the estrogen
(102,105,106) have been shown to cause contact dermatitis, but this is uncommon
(107,108).
2. Clonidine
Clonidine is a centrally acting ␣-agonist used primarily as an antihypertensive agent.
A common adverse effect associated with transdermally administered clonidine is
the development of local skin reactions to the clonidine preparation. Reports of such
dermatological reactions range in incidence between 5 and 42% (109–112). These
reactions vary in severity from mild erythema and pruritus to vesiculation and in-
flammatory infiltration of the skin beneath the transdermal patch. Rarely, develop-
ment of a generalized maculopapular rash has also been reported to occur following
transdermal clonidine therapy. The majority of the skin reactions requiring discon-
tinuation of therapy are mediated by a delayed-type IV hypersensitivity reaction
(allergic contact dermatitis), which can be confirmed with patch testing using com-
ponents of the clonidine transdermal device. In most of these patients the allergic
reaction is due to clonidine specifically, whereas in other patients, a specific com-
ponent of the transdermal system (polyisobutylene) functions as the allergen
(79,109).
There is an effect of race and gender on the irritation rates from clonidine patch
systems. For example, occlusive transfermal clonidine patch systems show sensiti-
zation rates of 34% in white women, 18% in white men, 14% in black women, and
8% in black men (113). Itchiness under the patch and contact dermatitis were re-
ported from clonidine transdermal patches (114). The long-term safety and efficacy
of transdermal clonidine was evaluated in 102 patients for over 5 years. Transient
local side effects occurred, mainly between weeks 4–26; thereafter, the incidence
clearly diminished and adverse events did not cause any withdrawal related to skin
reactions from 1 year up to 5 years. Overall the long-term transdermal clonidine
treatment was highly accepted and was well tolerated by the patients (115). It is
important to point out that predictive patch testing for allergic contact dermatitis
potential requires special techniques, not only for clonidine systems but for trans-
dermal systems in general (79,116).
3. Scopolamine
Scopolamine, a belladonna alkaloid indicated for nausea and vomiting associated
with motion, radiotherapy, anesthesia, and surgery, was the first drug approved for
use as a transdermal patch-type delivery system. There are three reports of allergic
contact dermatitis to scopolamine. In a group of 164 sailors, 10% developed allergic
contact dermatitis after 1.5–15 months of use (117). Patch testing with 1.8% sco-
polamine in petrolatum (2) or 0.25% in water (80) has been used to confirm allergic
contact dermatitis. Patch testing with structurally related alkaloids has failed to dem-
onstrate cross-sensitivity, indicating the specific nature of the antigenic site for sco-
polamine (2,118).
4. Nitroglycerin
Nitroglycerin is an organic nitrate used for the prevention and treatment of angina
pectoris caused by coronary artery disease. Erythema under the nitroglycerin trans-
dermal patch is frequent and represents the capacity of nitroglycerin to cause vaso-
dilation. Rubefaction at the margins of covered skin, noticed with a similar frequency
in placebo and active nitrate patches, is indicative of mild irritation (119). Irritant
reddening disappeared spontaneously within a few hours. A more severe reaction,
localized to the site of nitroglycerin and subsequently placebo patches, has been
described (120).
Allergic contact dermatitis to nitroglycerin, both in ointments and patch-type
transdermal drug delivery systems, has been reported (81,83,85,121,122). Some de-
livery systems have employed acrylate adhesives, which have been implicated as the
allergen in several of these cases (123).
5. Nicotine
The pharmacological side effects of transdermally absorbed nicotine have assumed
greater significance following recent research in alkaloid delivery through the skin.
Percutaneous administration of nicotine may reduce the craving experienced during
abstinence from cigarette smoking and, thereby, serve as useful supplementation
regimen during the behavioral modification process (124). Percutaneous nicotine ad-
ministration induces predominant sudorific and rubiform responses in the skin that
may be accompanied by subtle pyloerection, hyperalgesia, and pruritus (125).
The most common adverse effects of nicotine patches are cutaneous, charac-
terized by itching (16–29% of patients), erythema (7–25%), and edema (2–7%).
Poor cutaneous tolerability is a significant problem, resulting in withdrawal of the
treatment in 2–5% of patients (126,127). Bircher et al. (86) investigated 14 volun-
teers with a history of adverse skin reactions to nicotine transdermal therapeutic
systems. Five of 14 demonstrated contact sensitivity to 10% aqueous nicotine solu-
tion. Irritant reactions in 9 individuals were due to occlusion. The safety, tolerability,
and efficacy of transdermal nicotine patch was studied in 80 patients who smoked.
Side effects, such as itch and local erythema, were reported in 4 patients (128).
6. Testosterone
Testosterone transdermal therapeutic systems are used in the treatment of hypo-
gonadism in men. One system is designed to be applied to the scrotum and requires
changing daily. Three male subjects of nine reported transient pruritus with the pla-
cebo patch; however, none reported this with the use of testosterone transdermal
therapeutic systems (129). An alternative system for application to glabrous skin was
recently been commercialized in the United States. The package insert lists blister
development in 11.5% of the phase I–III clinical study population (130).
7. Fentanyl
Fentanyl is a narcotic analgesic used for medication before surgical procedures. Ad-
verse effects on skin (erythema) have been reported (131,132). The physicochemical
properties and adverse effects of transdermally administered fentanyl have been de-
scribed. Dermatological reactions to the fentanyl patch are generally transient and
mild (133).
B. Iontophoresis
Iontophoresis increases the penetration of ionized substances into or through a tissue
by application of an electric field (76,134–137). Iontophoresis has the potential to
overcome many limitations associated with conventional transdermal systems and

could be feasible for ionic, hydrophilic, and higher molecular weight drugs. Skin
irritation, however, has been reported following iontophoresis, but extensive toxi-
cological studies are still required (138). Such studies are underway in our laboratory
(139).
1. Barrier Properties and Skin Reactions
Skin irritation and stratum corneum integrity following 1- and 4-h saline iontopho-
resis in human subjects were evaluated using several response measurements:
1. A visual scoring system
2. Transepidermal water loss
3. Skin capacitance
4. Skin color
5. Skin temperature
Saline iontophoresis for 1 or 4 h did not produce significant changes in skin
water loss and skin water content, suggesting that skin function was unaffected by
transcutaneous electrical stimulation. However, the occurrence of transient changes
in skin structure (papules) was observed (unpublished data).
Papules are observed following iontophoresis, indicating that the electrical cur-
rent occasionally induced short-term, transient changes in skin. Direct effects of
electrical stimulation on vascular permeability were reported. For example, macro-
molecular capillary leakage was demonstrated following stimulation of the hamster
cheek pouch and the rabbit tibia with direct current of 5–50 ␮A for 30–160 min
(140). Several types of sweat retention (miliaria) have been described. Iontophoresis
produces miliaria rubra with distilled water after 10 min of current delivery at 0.5
mA/cm2 (141). The same study showed that vesicles had a different aspect than those
we have observed, in that they were uniformly scattered and their walls were fragile
enough to be rubbed off with a towel. Also, physiological saline did not produce
vesicles under the same conditions.
The effect of 4-h saline iontophoresis at the current density of 0.2 mA/cm2 was
investigated on skin barrier function and irritation in four ethnic groups (whites,
Hispanics, blacks, and Asians) (142). The results suggested that iontophoresis was
well tolerated in all four groups, and that skin barrier function, as determined by
TEWL and skin capacitance measurements, was not irreversibly affected by ionto-
phoresis in any group. There was no significant difference (p > 0.05) in skin tem-
perature, compared with baseline at all observation points in the ethnic groups. No
edema was observed in any group. However erythema was higher than the baseline
owing to iontophoresis in all the four groups (Table 3). The subjects also demon-
strated papules. The highest number of subjects exhibiting papules were in the Asian
group followed by Hispanics, whites, and blacks (142). The results of skin reactions
to iontophoresis in four ethnic groups are given in Table 4. Details of differences in
ethnic skin can be found in Berardesca and Maibach (143).
Solvents remove intercellular lipids resulting in cutaneous barrier disruption.
In a study on the effect of alcohol, acetone, and electrode gel swabbing and ionto-
phoresis on skin irritation, the skin integrity was not affected. However, erythema
and papules were observed, but these were virtually resolved 24 h after patch removal
(144).
Table 3 Draize Erythema Scoresa from Iontophoresis in

Ethnic Groups
Ethnic group Active site Control site
White 8E1 (3E1); 2E2 2E1

Black 9E1 (3E1); 1E2 2E1
Hispanic 3E1 (4E1); 7E2 2E1
Asian 4E1; 6E2 1E1
a
Entries are frequencies of subjects experiencing the level of the
erythema (E). Entries are listed such that for example: 8E1 (3E1)
indicates eight subjects developed erythema score of ‘‘1’’ which was
not resolved in three subjects 1440 min after patch removal. Ery-
thema score was not significantly different (p > 0.05) among ethnic
groups.
Source: Ref. 142.
2. Sensation and Itching

The range of sensations evoked by transcutaneous electrical stimulation have varied
from tactile (touch, vibration, or other) to pricking pain and itch. Thermal sensations,
however, have rarely been reported (145–152). A high-voltage low-current transcu-
taneous electrical stimulating device was tested for prickle sensation in 162 subjects.
The initial sensation experienced by subjects was prickle (153).
Itching is felt in certain subjects during and after iontophoresis. The way that
itch is signaled to the central nervous system (CNS) remains incompletely under-
Table 4 Skin Reactionsa to 4-h Iontophoresis: Influence of

Ethnic Group
Observation time
Patch
Ethnic group removal 60 min 1440 min
White PA 1p
AA 3p AA 2p AA 1p
Black AA 3p AA 3p AA 1c
Hispanic AA 8p AA 7p AA 3c
AC 1p
Asian AA 8p AA 10p AA 6c
AC 5p AC 5p
a
Entries are frequencies of subjects (n = 10) experiencing papules (p)
and papules in dry state [i.e., crust (c)]. Entries are listed as active
anode, AA; active cathode, AC; passive anode, PA; and passive cath-
ode, PC: for example, PA 1p at patch removal immediately after
iontophoresis indicates one subject developed papules at the passive
anode site at observation time immediately after patch removal and
AA 1c at 1440 min (24 h) indicates one subject still had papules in
dry state (crust) 1440 min after iontophoresis at the active anode site.
Source: Ref. 142.
stood. A general theory proposes that the whole spectrum of cutaneous sensations is
signaled by differences in the patterns of activity; hence, any particular neuron can
signal a variety of sensory modalities. However, the finding that high-frequency,
electrical stimulation of large myelinated axons in the peripheral nerves of conscious
humans consistently evokes painless sensations argues against such models (154).
An alternative view proposes that an individual neuron transmits a specific type of
sensory information. Itch sensation evoked by percutaneous microiontophoresis of
histamine on hairy human skin was studied (155). Iontophoresis of histamine evoked
sensations of itch in human subjects; therefore, itch sensation may be implicated
owing to release of histamine during iontophoresis.
Under proper conditions, touch sensations, such as thumping, vibration, and
pulsing, can be elicited by electrical stimulation of hairy skin (156–158). In contrast
with hairy skin, the threshold sensations on glabrous skin are touch instead of pruritus
(159). With electrode paste used on six subjects as a conductive medium, half re-
ported painful sensation, similar to acid burn, but with saline they felt itch (160).
Changes in electrode size may alter the quality of sensation from itch to pain (161).
Recent advances in defining C-fiber function is described (162).
3. Burns
Burns occur if the patient uses excessive stimulation with small-area electrodes or
if the interface between the skin and electrode is dry (163). Shealy and Maurer (164)
demonstrated that the electrode surface area must be more than 4 cm2 for a 500-␮s
85-mA–pulse, 185-pps stimulus. The heat produced must be less than 250 mcal/cm2
s⫺1 to avoid localized burns. Burton (165) described another type of injury, micro-
punctate burns. The explanation was that current flow is not distributed over a wide
surface area, but is concentrated in small punctate areas (usually hair follicles). Be-
cause of the concentration of the large volume of current in small areas, current
density is high, resulting in skin burns. These micropunctate skin burns represent
true thermal damage to the skin, but Burton (165) feels that they are of little clinical
significance, in themselves, as with allergic reactions, simple discontinuation of use
of the electrodes permit recovery.
4. Virus Activation
In humans, one isolated case of an outbreak of molluscum contagiosum, a DNA
virus of the pox group, at the site of hydrocortisone iontophoresis has been described
(166). It appears that the phenomenon is more related to the drug being delivered
than to iontophoretic mode of delivery.
C. Electroporation
Electroporation involves alteration of lipid bilayers when transient and pulsed electric
fields lead to the reversible formation of nonlamellar lipid phases: a pore. Iontopho-
resis utilizes existing pathways, such as hair follicles or sweat glands. These sweat
glands and hair follicles comprise only about 0.1% of the total skin surface area.
Thus, a high-charge density occurs around sweat glands and hair follicles, which
may potentially lead to localized skin irritation. In electroporation, the other 99.9%
of the skin’s surface area is reversibly altered using a brief pulse of electricity. As a
consequence, the current density is distributed more uniformly across the surface;
thus, potential for irritation may be reduced (167).
The effect of current and voltage on pig skin was evaluated under conditions
of iontophoresis and electroporation (167). Pigs were treated with either an ionto-
phoresis or an electroporation protocol. The study evaluated irritation, not drug de-
livery. Current densities used were in the range from 0 to 10 mA/cm2, and the applied
voltage ranged from 0 to 1000 V. The potential was a single pulse followed by 30
min of iontophoresis. Irritation was measured at 0 and 4 h after treatment. Skin
biopsies were taken for histological examination. Irritation was measured by the
visual scoring system of Draize et al. (6). Use of conventional iontophoresis, when
there was no applied voltage pulse and current density was near 0.2 mA/cm2, resulted
in no significant difference from the no-pulse values in either of these measures. The
skin response was measured in terms of erythema at the anode and the cathode.
Again iontophoresis produced a value not significantly different from that of a pulse
plus iontophoresis at both the cathode and the anode. These results showed that a
pulse voltage of up to 1000 V had no effect on erythema or edema. Erythema and
edema are equivalent for iontophoresis and electroporation. Thus, one can conclude
that electroporation under these test conditions produced no measurable damage to
skin or tissue.
IV. CONCLUSIONS
Adequate evaluation of irritation potential of chemical substances depends on a thor-
ough understanding of the variables influencing the irritant response. Guinea pig
testing and the local lymph node assays constitute a first step in evaluating the
allergenicity of new compounds or products. With the traditional Draize test, potent
irritants can be detected. Substances that irritate in humans also do so in some
animals. More sensitive animal tests will identify weak irritants. The comparative
sensitivity of these various tests is still under examination. There is a reasonable
degree of correlation between the GPM test and the human maximization test.
Transdermal therapeutic systems have proved to be a useful adjunct for ad-
ministration of systemic medications. Their potential for future applications seems
excellent. However, the systems carry a risk of either irritant or allergic skin sensi-
tivity. Avoidance of reapplication of patches directly over the previous site, should
help minimize the incidence and severity of such irritation. Keep in view that with
the delayed onset of some allergic reactions, safety data based on short-term expe-
rience should be considered with caution.
There is increasing interest in the use of iontophoresis. Such therapy may re-
quire long-term delivery and the extended wearing of delivery systems. Irritations in
such patients may be greater than those found with the more brief applications for
which iontophoresis is most widely used today. There is no doubt that iontophoresis
can be a safe and effective method of drug delivery by the innovative application of
modern electronics and material science; however, extensive skin toxicological stud-
ies are warranted. Alternatively, electroporation followed by iontophoresis can be
used to lower the skin irritation. With electroporation, new pathways are created. As
a consequence, there is more even distribution of charge; hence, there may be a
lower potential for irritation.
REFERENCES
1. Sarkany I, Gaylarde PM. Thickening of guinea pig epidermis due to application of
commonly used ointment bases. Trans St John’s Hosp Dermatol Soc 59:241, 1973.
2. Trozak DJ. Delayed hypersensitivity to scopolamine delivered by a transdermal device.
J Am Acad Dermatol 13:247–251, 1985.
3. Weickel R, Frosch PJ. Kontaktallergie auf Glycerotrinitrat (Nitroderm TTS). Hautarzt
37:511, 1986.
4. Chandraratna PAN, O’Dell RE. Allergic reactions to nitroglycerin ointment: a report
of five cases. Curr Ther Res 25:481, 1979.
5. Rougier A, Goldberg A, Maibach HI, eds. In Vitro Skin Toxicology: Irritation, Pho-
totoxicity and Sensitization. New York: Mary Ann Liebert, 1994.
6. Draize JH, Woodard G, Calvery HO. Methods for the study of irritation and toxicity
of substances applied topically to the skin and mucous membranes. J Pharmacol Exp
Ther 83:377–390, 1944.
7. Patrick E, Maibach HI. Predictive assays: animal and man, and in vitro and in vivo.
In: Textbook of Contact Dermatitis. Berlin: Springer-Verlag, p 105, 1995.
8. Idson DR. In vivo measurement of transepidermal water loss. J Soc Cosmet Chem 29:
573, 1978.
9. Wilhelm KP, Surber C, Maibach HI. Quantification of sodium lauryl sulfate irritant
dermatitis in man: comparison of four techniques: skin color reflectance, transepidermal
water loss, laser Doppler flow measurement and visual scores. Arch Dermatol Res 281:
1–3, 1989.
10. Tagami H, Ohi H, Iwatsuki K, Kanamaru Y, Yamada M, Ichijo B. Evaluation of the
skin surface hydration in vivo by electrical measurement. J Invest Dermatol 75:500–
507, 1980.
11. Thiele FAJ, Malten KE. Evaluation of skin damage. Br J Dermatol 89:373, 1973.
12. Serban GP, Henry SM, Cotty VF, Marcus AD. In vivo evaluation of skin lotions by
electrical capacitance 1. The effect of several lotions on the progression of damage
and healing after repeated insult with sodium lauryl sulfate. J Soc Cosmet Chem 32:
407–419, 1981.
13. Malten KE, Thiele FAJ. Evaluation of skin damage. II. Water loss and carbon dioxide
release measurements related to skin resistance measurements. Br J Dermatol 89:565,
1973.
14. Thiele FAJ. Measurements on the Surface of the Skin. Nijmegen, Netherlands: Drekkeij
van Mammeren, p 81, 1974.
15. Elsner P, Berardesca E, Maibach HI, eds. Bioengineering of the Skin: Water and the
Stratum Corneum. Boca Raton: CRC Press, 1994.
16. Berardesca E, Elsner P, Maibach HI, eds. Bioengineering of the Skin: Cutaneous Blood
Flow and Erythema. Boca Raton: CRC Press, 1994.
17. Berardesca E, Elsner P, Wilhelm K, Maibach HI, eds. Bioengineering of the Skin:
Methods and Instrumentation. Boca Raton: CRC Press, 1995.
18. Buehler EV. Delayed contact hypersensitivity in the guinea pig. Arch Dermatol 91:
171–177, 1965.
19. Magnusson B, Kligman AM. The identification of contact allergens by animal assay.
The guinea pig maximization test. J Invest Dermatol 52:268–276, 1969.
20. Marzulli FN, Carson TR, Maibach HI. Delayed contact hypersensitivity studies in man
and animals. Proc Joint Conf Cosmet Sci, p 107, 1968.
21. Draize JH. Dermal toxicity. Appraisal of the safety of chemicals in foods, drugs and
cosmetics. The Association of Food and Drug Officials of the United States. Texas
State Department of Health, p 46, 1959.
22. Kligman AM. The identification of contact allergens by human assay. III. The maxi-
mization test: a procedure for screening and rating contact sensitizers. J Invest Dermatol
47:393–409, 1966.
23. Lepoittevin J–P, Benezra C, Sigman CC, Bagheri D, Fraginals R, Maibach HI. Mo-
lecular aspects of allergic contact dermatitis. In: Textbook of Contact Dermatitis. Berlin:
Springer-Verlag, p 105, 1995.
24. Menne T, Flyvholm MA, Maibach HI. Prevention of allergic contact sensitization. In:
Vos JG, Younes M, Smith E, eds. Allergic Hypersensitivities Induced by Chemicals.
Boca Raton: CRC Press, p 287, 1996.
25. Marzulli FN, Maibach HI. Allergic contact dermatitis. In: Marzulli FN, Maibach HI,
eds. Dermatotoxicology. 5th ed. Washington: Hemisphere, p 143–146, 1996.
26. Marzulli FN, Maibach HI. Test methods for allergic contact dermatitis in humans. In:
Marzulli FN, Maibach HI, eds. Dermatotoxicology. 5th ed. Washington: Hemisphere,
pp 477–483, 1996.
27. Magnusson B, Kligman AM. Usefulness of guinea pig tests for detection of contact
sensitizers. In: Marzulli FN, Maibach HI, eds. Advances in Modern Toxicology. Vol 4.
Dermatotoxicology and Pharmacology. New York: John Wiley & Sons, p 551, 1977.
28. Marzulli FN, Maibach HI, eds. Dermatotoxicology. 5th ed. Washington: Hemisphere,
1996.
29. Anderson R, Maibach HI, eds. Guinea Pig Sensitization Assays. New York: Karger,
1987.
30. Welthriend S, Kwangsukstith C, Maibach HI. Contact urticaria from cucumber pickle
and strawberry. Contact Dermatitis 32:173–174, 1995.
31. Hardy MP, Maibach HI. Contact urticaria syndrome from sorbitan sesquioleate in a
corticosteroid ointment. Contact Dermatitis 32:114, 1995.
32. Maibach HI. Contact urticaria syndrome from mold on salami casing. Contact Der-
matitis 32:120–121, 1995.
33. West I, Maibach HI. Contact urticaria syndrome from multiple cosmetic components.
34. Elsner P, Maibach HI, eds. Irritant Dermatitis, New Clinical and Experimental Aspects.
Current Problems in Dermatology. Vol. 23. Basel: Karger, 1995.
35. Holness DL, Nethercott JR, Adams RM, Belsito D, Deleo V, Emmett EA, Fowler J,
Fisher AA, Larsen WG, Maibach HI, Marks J, Reitschel RL, Rosenthal LE, Schorr
WF, Storrs FJ, Taylor JS. Concomitant positive patch test results with standard screen-
ing tray in North America 1985–1989. Contact Dermatitis 32:289–292, 1995.
36. Maibach HI. Dermatologic vehicles: science and art. Cutis 55(4S):4, 1995.
37. Funk JO, Dromgoole SH, Maibach HI. Sunscreen intolerance: contact sensitization,
photocontact sensitization, and irritancy of sunscreen agents. Dermatol Clin 13:473–
481, 1995.
38. Hatch KL, Maibach HI. Textile dermatitis: an update (1) resins, additives and fibers.
Contact Dermatitis 32:319–326, 1995.
39. Hardy M, Maibach HI. Contact urticaria syndrome from sorbitan sesquioleate in a
corticosteroid ointment. Contact Dermatitis 32:114, 1995.
40. Lee CH, Maibach HI. The sodium lauryl sulfate model: an overview. Contact Der-
matitis 33:1–7, 1995.
41. Hostynek JJ, Lauerma AI, Magee PS, Maibach HI. A local lymph-node assay validation
study of a structure–activity relationship model for contact allergens. Arch Dermatol
Res 287:567–571, 1995.
42. Hui X, Wester RC, Magee PS, Maibach HI. Partitioning of chemicals from water into
powdered human stratum corneum (callus): a model study. In Vitro Toxicol 8:159–
167, 1995.
43. Tur E, Aviram G, Zeltser D, Brenner S, Maibach HI. Regional variations of human
skin blood flow response to histamine. In: Exogenous Dermatology, Advances in Skin-
Related Allergology, Bioengineering, Pharmacology and Toxicology. Basel: Karger, p
59, 1995.
44. Ale SI, Maibach HI. Clinical relevance in allergic contact dermatitis, an algorithmic
approach. Occup Environ Dermatoses 43:119–121, 1995.
45. Liden C, Maibach HI, Wahlberg JE. Skin. In: Metal Toxicology. New York: Academic,
p 447, 1995.
46. Kwangsukstith C, Maibach HI. Contact urticaria from polyurethane membrane hy-
poallergenic gloves. Contact Dermatitis 33:200–201, 1995.
47. Patil S, Harvell J, Maibach HI. In vitro skin irritation assays: relevance to human skin.
In: Gali CL, Goldberg AM, Marinovich M, eds. Modulation of Cellular Responses in
Toxicity. Berlin: Springer-Verlag, p 283, 1995.
48. Effendy I, Maibach HI. Surfactants and experimental irritant contact dermatitis. Contact
Dermatitis 33:217–225, 1995.
49. Lauerma AI, Aioi A, Maibach HI. Topical cis-urocanic acid suppresses both induction
and elicitation of contact hypersensitivity in BALB/C mice. Acta Derm Venereol 75:
272–275, 1995.
50. Andersen PH, Maibach HI. Skin irritation in man: a comparative bioengineering study
using improved reflectance spectroscopy. Contact Dermatitis 33:315–322, 1995.
51. Tur E, Eshkol Z, Brenner S, Maibach HI. Cumulative effect of subthreshold concen-
trations of irritants in humans. J Contact Dermatitis 6:216–220, 1995.
52. van der Valk PGM, Maibach HI, eds. The Irritant Contact Dermatitis Syndrome. Boca
Raton: CRC Press, 1995.
53. Ophaswongse S, Maibach HI. Allergic contact cheilitis. Contact Dermatitis 33:365–
370, 1995.
54. Effendy I, Loeffler H, Maibach HI. Baseline transepidermal water loss in patients with
acute and healed irritant contact dermatitis. Contact Dermatitis 33:371–374, 1995.
55. Kwangsukstith C, Maibach HI. Effect of age and sex on the induction and elicitation
of allergic contact dermatitis. Contact Dermatitis 33:289–298, 1995.
56. Frosch PJ, Pilz B, Andersen KE, Burrows D. Carnarasa JG, Dooms–Goossens A, Du-
combs G, Fuchs T, Hannuksela M, Lachapelle JM, Lahti A, Maibach HI, Menne T,
Rycroft RJG, Shaw S, Wahlberg JE, White IR, Wilkinson JD. Patch testing with fra-
grances: results of multicenter study of the European Environmental and Contact Der-
matitis Research group with 48 frequently used constituents of perfumes. Contact Der-
matitis 33:333–342, 1995.
57. Maibach HI. Possible cosmetic dermatitis due to mercaptobenzothiazole. Contact Der-
matitis 34:72, 1996.
58. Patil S, Singh P, Maibach HI. Radial spread of sodium lauryl sulfate after topical
application. Pharm Res 12:2018–2023, 1995.
59. Amin S, Maibach HI. Cosmetic intolerance syndrome: pathophysiology and manage-
ment. Cosmet Dermatol 9:34, 1996.
60. Smith JD, Odom RB, Maibach HI. Contact urticaria from cobalt chloride. Arch Der-
matol 111:1610, 1995.
61. Effendy I, Kwangsukstith C, Lee JY, Maibach HI. Functional changes in human stratum
corneum induced by topical glycolic acid: comparisons with all-trans-retinoic acid.
Acta Derm Venereol 75:455, 1995.
62. Nangia A, Andersen PH, Berner B, Maibach HI. High dissociation constants (pKa) of
basic permeants are associated with in vivo skin irritation in man. Contact Dermatitis
34:237–242, 1996.
63. Maibach HI. Acne necrotica varioliformis (P. acnes folliculitis); resolution with iso-
tretinoin. Eur J Dermatol 6:153, 1996.
64. Mitchell JC, Maibach HI. Difficulties to investigate dermatitis from plants, a practical
approach to office-based diagnosis. Dermatosan 44:29, 1996.
65. Effendy I, Weltfriend S, Patil S, Maibach HI. Differential irritant skin responses to
topical retinoic acid and sodium lauryl sulfate: alone and crossover design. Br J Der-
matol 134:424, 1996.
66. Flyvholm MA, Menne T, Maibach HI. Skin allergy: exposure and dose–response re-
lationships. In: Vos JG, Younes M, Smith E, eds. Allergic Hypersensitivities Induced
by Chemicals. Boca Raton: CRC Press, p 261, 1996.
67. Weltfriend S, Bason M, Lammintausta K, Maibach HI. Irritant dermatitis (irritation).
In: Marzulli FN, Maibach HI, eds. Dermatotoxicology. 5th ed. Washington: Hemi-
sphere, pp 87–118, 1996.
68. Menné T, Veien N, Maibach HI. Systemic contact-type dermatitis. In: Marzulli FN,
Maibach HI, eds. Dermatotoxicology. 5th ed. Washington: Hemisphere, pp 161–175,
1996.
69. Patil SM, Hogan DJ, Maibach HI. Transdermal drug delivery systems: adverse der-
matologic reactions. In: Marzulli FN, Maibach HI, eds. Dermatotoxicology. 5th ed.
Washington: Hemisphere, pp 389–396, 1996.
70. Amin S, Lahti A, Maibach HI. Contact urticaria and the contact urticaria syndrome
(immediate contact reactions). In: Marzulli FN, Maibach HI, eds. Dermatotoxicology.
5th ed. Washington: Hemisphere, pp 485–503, 1996.
71. Amin S, Lauerma AI, Maibach HI. Diagnostic tests in dermatology: patch and pho-
topatch testing and contact urticaria. In: Marzulli FN, Maibach HI, eds. Dermatotoxi-
cology. 5th ed. Washington: Hemisphere, pp 505–513, 1996.
72. Marzulli FN, Maibach HI. Photoirritation (phototoxicity) testing in humans. In: Mar-
zulli FN, Maibach HI, eds. Dermatotoxicology. 5th ed. Washington: Hemisphere, pp
531–533, 1996.
73. Toro JR, Engasser PG, Maibach HI. Cosmetic reactions. In: Marzulli FN, Maibach HI,
eds. Dermatotoxicology. 5th ed. Washington: Hemisphere, pp 607–642, 1996.
74. Guy RH, Hadgraft J, Bucks DAW. Transdermal drug delivery and cutaneous metabo-
lism. Xenobiotica 17:325–343, 1987.
75. Roberts MS, Singh J, Yoshida N, Currie KI. Iontophoretic transport of selected solutes
through human epidermis. In: Scott RC, Hadgraft J, Guy RH, eds. Prediction of Per-
cutaneous Penetration. Vol. 1. London: IBC Technical Services, pp 231–241, 1990.
76. Singh S, Singh J. Transdermal drug delivery by passive diffusion and iontophoresis: a
review. Med Res Rev 13:569, 1993.
77. Nater JP, de Groot AC, eds. Unwanted Effects of Cosmetics and Drugs Used in Der-
matology. Amsterdam: Elsevier, 1985.
78. Grattan CEH, Kennedy CTC. Allergic contact dermatitis to transdermal clonidine. Con-
tact Dermatitis 12:225–226, 1985.
79. Maibach HI. Oral substitution in patients sensitized by transdermal clonidine treatment.
Contact Dermatitis 16:1–8, 1987.
80. van der Willigen AH, Oranje AP, Stolze E, van Joost T. Delayed hypersensitivity to
scopolamine in transdermal therapeutic systems. J Am Acad Dematol 18:146–147,
1988.
81. Harari Z, Sommer I, Knobel B. Multifocal contact dermatitis to Nitroderm TTS 5 with
extensive postinflammatory hypermelanosis. Dermatologica 174:249–252, 1987.
82. Zugerman C, Zheutlin T, Giacobetti R. Allergic contact dermatitis secondary to nitro-
glycerin in Nitro-Bid ointment. Contact Dermatitis 5:27, 1979.
83. Rosenfeld AS, White WB. Allergic contact dermatitis secondary to transdermal nitro-
glycerin. Am Heart J 108:1061, 1984.
84. Sausker WF, Frederick FD. Allergic contact dermatitis secondary to topical nitroglyc-
erin. JAMA 239:1743–1744, 1978.
85. Topaz O, Abraham D. Severe allergic contact dermatitis secondary to nitroglycerin in

a transdermal therapeutic system. Ann Allergy 59:365–366, 1987.
86. Bircher AJ, Havald H, Rufli T. Adverse reaction to nicotine in a transdermal therapeutic
system. Contact Dermatitis 25:230, 1991.
87. Sitruk–Ware R. Estrogen therapy during menopause: practical treatment recommen-
dations. Drugs 39:203–217, 1990.
88. Utian WH. Transdermal oestradiol: a recent advance in oestrogen therapy. Drugs 36:
383–386, 1988.
89. Youngkin EQ. Estrogen replacement therapy and the Estraderm transdermal system.
Nurse Pract 15:19, 1990.
90. Balfour JA, Heel RC. Transdermal estradiol: a review of its pharmacodynamic and
pharmacokinetic properties, and therapeutic efficacy in the treatment of the menopausal
complaints. Drugs 40:561–582, 1990.
91. Buvat J, Buvat–Herbaut M, Desmons F, Cuvelier C, Vasseur C. Les patchs à l’estradiol.
Un progrès dans le traitement de la ménopause. Gynécologie 40:53–59, 1989.
92. Erkkola R, Holma P, Jarvi T, Nummi S, Punnonen R. Transdermal oestrogen replace-
ment therapy in a Finnish population. Maturitas 13:275–281, 1991.
93. Grall JY. Estrogenothérapie substitutive par voie transdermique: sur quels critères
choisir la posologie initale? Tempo Méd 410(suppl):1–7, 1990.
94. Janaud A. Estraderm TTS dans le traitement des troubles liés à la ménopause: résultats
d’une étude long terme chez 324 femmes. Gaz Méd 97:1–7, 1990.
95. Kerzel C, Keller PJ. Ein transdermales therapeutisches System zur hormonellen Ther-
apie klimakterischer ausfallserscheinungen. Multizentrische Studie. Geb Frauenheilkd
47:565–568, 1987.
96. Mück AO. Transdermales therapeutisches System zur physiologischen Östrogensubsti-
tution. Therapie 40:41–51, 1990.
97. Rabe T, Mück AO, Runnebaum B. Östrogensubstitution in der postmenopause mit
Östradiol durch ein transdermales therapeutisches System In: Rietbrock, ed. Die Haut
als Transportorgan für Arzneistoffe. Darmstadt: Steinkopff Verlag, pp 81–92, 1990.
98. Balfour JA, McTavish D. Transdermal estradiol: a review of its pharmacological profile,
and therapeutic potential in the prevention of postmenopausal osteoporosis. Drug Aging
2:487–507, 1992.
99. Gordon SF, Thompson KA, Ruoff GE, Imig JR, Lane PJ, Schwenker CE. Efficacy and
safety of a seven-day transdermal drug delivery system: comparison with conjugated
estrogens and placebo. The Transdermal Estradiol Patch Study Group. Int J Fertil Stud
40:126, 1995.
100. Nachtigall LE. Longer-term study of Estraderm TTS. In: Birdwood GF, ed. Transdermal
Estrogen Replacement for Menopausal Women. Toronto: Hans Huber, pp 25–28, 1998.
101. Randall S. Clinical experience with transdermal hormone replacement therapy. In: Bird-
wood, ed. Transdermal Estrogen Replacement for Menopausal Women. Toronto: Hans
Huber, pp 42–44, 1988.
102. Fisher AA. Contact dermatitis: highlights from the 1987 meeting of the American
Academy of Dermatology, San Antonio, Texas. Cutis 41:87, 1988.
103. Schwartz BK, Clendenning WE. Allergic contact dermatitis from hydroxypropyl cel-
lulose in a transdermal estradiol patch. Contact Dermatitis 18:106, 1988.
104. Ducros B, Bonnin JP, Navaranne A, Colomb D. Eczema due to contact with ethanol
in oestradiol transdermal patch (Estraderm TTS 50). Nouv Dermatol 8:21, 1989.
105. Carmichael AJ, Foulds IS. Allergic contact dermatitis to oestradiol in oestrogen patches.
106. McBurney EI, Noel SB, Collins JH. Contact dermatitis to transdermal estradiol system.
J Am Acad Dermatol 20:503–510, 1989.
107. Davis GF, Winter L. Cumulative irritation study of placebo estrogen patches. Curr Ther
Res 42:712–719, 1987.
108. Hogan DJ, Maibach HI. Adverse dermatologic reactions to transdermal drug delivery
systems. J Am Acad Dermatol 22:811–814, 1990.
109. Groth H, Vetter H, Kneusel J. Transdermal clonidine in essential hypertension: prob-
lems during long term treatment. In: Weber MA, Drayer JIM, Kolloch R, eds. Low
Dose Oral and Transdermal Clonidine Therapy of Hypertension. Darmstadt: Steinkopff,
pp 60–65, 1985.
110. Weber MA, Drayer JIM. Clinical experience with rate controlled delivery of antihy-
pertensive therapy by a transdermal system. Am Heart J 108:231–236, 1984.
111. Popli S, Stroka G, Ing TS. Transdermal clonidine for hypertensive patients. Clin Ther
5:624–628, 1983.
112. Boekhorst JC. Allergic contact dermatitis with transdermal clonidine. Lancet 2:1031–
1032, 1983.
113. Boehringer Ingelheim. Catapress-TTS (clonidine) package insert. Ridgefield, CT, 1987.
114. Corazza M, Montovani L, Virgili A, Strumia R. Allergic contact dermatitis from a
clonidine transdermal delivery system. Contact Dermatitis 32:246, 1995.
115. Breidthardt J, Schumacher H, Mehlburger L. Long-term (5 year) experience with trans-
dermal clonidine in the treatment of mild to moderate hypertension. Clin Autonom Res
3:385–390, 1993.
116. Maibach HI. Clonidine transdermal delivery system: cutaneous toxicity studies. In:
Weber MA, Drayer JIM, Kolloch R, eds. Low Dose Oral and Transdermal Clonidine
Therapy of Hypertension. Darmstadt: Steinkopff, pp 55–59, 1985.
117. Gordon CR, Shupak A, Doweck I, Spitzer O. Allergic contact dermatitis caused by
transdermal hyoscine. Br Med J 298:122–131, 1989.
118. van der Willigen AH, de Graaf YP, van Joost T. Periocular dermatitis from atropine.
119. Muller P, Imhof PR, Burkart F, Chu LC, Gerardin A. Human pharmacological studies
of a new transdermal system containing nitroglycerin. Eur J Clin Pharmacol 22:473–
480, 1982.
120. Letendre PW, Barr C, Wilkens K. Adverse dermatologic reaction to transdermal nitro-
glycerin. Drug Int Clin Pharm 18:69–70, 1984.
121. Fischer RG, Tyler M. Severe contact dermatitis due to nitroglycerin patches. South
Med J 78:1523–1524, 1985.
122. Hendricks AA, Dec GW. Contact dermatitis due to nitroglycerin ointment. Arch Der-
matol 115:853–855, 1979.
123. Fisher AA. Dermatitis due to transdermal therapeutic systems. Cutis 34:526, 1984.
124. Russell MAH, Sutherland G, Costello J. Nicotine replacement: the role of blood nic-
otine levels, their rate of change, and nicotine tolerance, In: Pomerleau OF, Pomerleau
CS, eds. Nicotine Replacement: A Critical Evaluation. New York: Liss, p 63, 1988.
125. Smith EW, Smith KA, Maibach HI, Andersson P–O, Cleary G, Wilson D. The local
side effects of transdermally absorbed nicotine. Skin Pharmacol 5:69–76, 1992.
126. Abelin T, Buchler A, Muller P, Besanen K, Imhof PR. Controlled trial of transdermal
nicotine patch in tobacco withdrawal. Lancet 1:7, 1989.
127. Russell MAH, Stapleton JA, Feyerabend C, Wiseman SM, Gustavson G. Targeting
heavy smokers in general practice: randomised controlled trial of transdermal nitotine
patches. Br Med J 306:1308, 1993.
128. Martin PD, Robinson GM. The safety, tolerability and efficacy of transdermal nicotine
(Nicotinell TTS) in initially hospitalised patients. NZ Med J 108:6, 1995.
129. Bals–Pratsch M, Yoon YD, Knuth UA, Meschlag E. Transdermal testosterone substi-
tution therapy for male hypogonadism. Lancet 2:943–946, 1986.
130. SmithKline Beecham. Androderm Package Insert.
131. Miser AW, Narang PK, Dothage JA, Young RC, Sindelar W, Miser JS. Transdermal
fentanyl for pain control in patients with cancer. Pain 37:15–21, 1989.
132. Gourley GK, Kowalsky SR, Plummer JL. The efficacy of transdermal fentanyl in the
treatment of postoperative pain: a double blind comparison of fentanyl and placebo
systems. Pain 40:21–28, 1990.
133. Calis KA, Kohler DR, Corso DM. Transdermally administered fentanyl for pain man-
agement. Clin Pharm 11:22–36, 1992.
134. Singh J, Roberts MS. Transdermal delivery of drugs by iontophoresis: a review. Drug
Des Deliv 4:1–12, 1989.
135. Singh J. Effect of ionization on in-vitro iontophoretic and passive transport of thyro-
tropin-releasing hormone through human epidermis. Pharm Sci 1:111, 1995.
136. Singh J, Roberts MS. Iontophoretic transport of amphoteric solutes through human
epidermis: p-aminobenzoic acid and amphotericin. Pharm Sci 1:223, 1995.
137. Singh J, Bhatia KS. Topical iontophoretic drug delivery: pathways, principles, factors,
and skin irritation. Med Res Rev 16:285–296, 1996.
138. Singh J, Maibach HI. Topical iontophoretic drug delivery in vivo: historical develop-
ment, devices and future perspectives. Dermatology 187:235, 1993.
139. Camel E, O’Connell M, Sage B, Gross M, Maibach H. The effect of saline iontopho-
resis on skin integrity in human volunteers. Fundam Appl Toxicol 32:168–178, 1996.
140. Nannmark U, Buch F, Albrektsson T. Vascular reactions during electrical stimulation.
Vital microscopy of the hamster cheek pouch and the rabbit tibia. Acta Orthop Scand
56:52–56, 1985.
141. Shelley WB, Horvath PN, Weidman FD, Pillsbury DM. Experimental miliaria in man,
I. Production of sweat retention anidrosis and vesicles by means of iontophoresis. J
Invest Dermatol 11:275, 1949.
142. Singh J, Gross M, O’Connell M, Sage B, Maibach HI. Effect of iontophoresis in
different ethnic groups’ skin function. Proc Int Symp Control Release Bioact Mater
21:365–366, 1994.
143. Berardesca E, Maibach HI. Racial differences in skin pathophysiology. J Am Acad
Dermatol 34:667, 1996.
144. Singh J, Gross M, O’Connell M, Sage B, Maibach HI. Effect of skin preparation and
iontophoresis on skin functions and irritation in human volunteers. Pharm Res 11:S101,
1994.
145. Bishop GF. Responses to electrical stimulation of single sensory units of skin. J Neu-
rophysiol 6:361, 1963.
146. Grimnes S. Electrovibration, cutaneous sensation of microampere current. Acta Physiol
Scand 118:19, 1983.
147. Laitenen LV, Eriksson AT. Electrical stimulation in the measurement of cutaneous sen-
sibility. Pain 22:139, 1985.
148. Mason JL, Mackay NAM. Pain sensations associated with electrocutaneous stimulation.
IEE Trans Biomed Eng 23:405–409, 1976.
149. Reilly JP, Larkin WD. Electrocutaneous stimulation with high voltage capacitive dis-
charges. IEE Trans Biomed Eng 30:631, 1983.
150. Saunders FA. Information transmission across the skin: high resolution tactile sensory
aids for the deaf and the blind. Int J Neurosci 19:21–28, 1983.
151. Sigel H. Price threshold stimulation with square-wave current: a new measure of skin
sensibility. Yale J Biol Med 26:145, 1953.
152. Tuckett RP. Itch evoked by electrical stimulation of the skin. J Invest Dermatol 79:
368–373, 1982.
153. Gransworthy RK, Gully RL, Kenins P, Westerman RA. Transcutaneous electrical stim-
ulation and the sensation of prickle. J Neurophysiol 59:1116–1127, 1988.
154. Collins WF, Nulsen FE, Randt CT. Relation of peripheral fiber size and sensation in
man. Arch Neurol 3:381, 1973.
155. Westerman RA, Margerl W, Handwercker HO, Szolcsányi J. Itch sensation evoked by
percutaneous microintophoresis of histamine on human hairy skin. Proc Aust Physiol
Pharmacol Soc 16:79, 1985.
156. Gibson RH. Electrical stimulation of pain and touch. In: Kenshalo DR, ed. The Skin
Senses. Springfield, IL: Charles C Thomas, pp 223–261, 1968.
157. Saunders FA. Recommended procedures for electrocutaneous displays. In: Hambrecht
FT, Reswick JB, eds. Functional Electrical Stimulation. New York: Marcel Dekker, p
303, 1977.
158. Sachs RM, Miller JD, Grant KW. Perceived magnitude of multiple electrocutaneous
pulses. Percept Psychophys 28:255, 1980.
159. Hahn JF. Cutaneous vibratory thresholds for square wave electrical pulses. Science
127:879–880, 1958.
160. Notermans SLH. Measurement of the pain threshold determined by electrical stimu-
lation and its clinical application. I. Method and factors possibly influencing the pain
threshold. Neurology 16:1071–1086, 1966.
161. Pfeiffer EA. Electrical stimulation of sensory nerves with skin electrodes for research,
diagnosis, communication and behavioral conditioning: a survey. Med Biol Eng 6:637–
651, 1968.
162. Yosipovitch G, Szolar C, Hui X, Maibach HI. Effect of topically applied menthol on
thermal, pain and itch sensation and biophysical properties of the skin. Arch Dermatol
Res 288:245, 1996.
163. Fisher AA. Dermatitis associated with transcutaneous electrical nerve stimulation. Cutis
21:24–47, 1978.
164. Shealy CN, Maurer D. Transcutaneous nerve stimulation for control of pain. Surg
Neurol 2:45, 1974.
165. Burton C. Transcutaneous electrical nerve stimulation to relieve pain. Postgrad Med
59:105, 1976.
166. Plutchik R, Bender H. Electrocutaneous pain thresholds in humans to low frequency
square-wave pulses. J Psychol 62:151–154, 1966.
167. Potts RO. Transdermal peptide delivery using electroporation. In: Proceedings of the
Third TDS Technology Symposium: Polymers and Peptides in Transdermal Delivery.
Tokyo: Nichon Toshi Center, pp 47–67, 1993.
Index
Absolute bioavailability, 401, 452 Allergic contact dermatitis, 426, 512, 531
Absorption base, 322 influence of vehicle, 427
Acetaminophen, 178, 277 to nitroglycerin, 535
Acetylsalicylic acid, 205, 206 Allergic contact hypersensitivity, 52
Acitretin, 415 Allergic reactions, 362, 517
Acne, 47, 48, 49, 176, 307 Alleviation of pain, 285
Acne treatment, 49 Alniditan, 215
Acne vulgaris, 48 Alopecia, 176, 307
Acquired immunodeficiency syndrome Alprostadil (prostaglandin E), 385
(AIDS), 56, 374 Aluminum chloride, 531
Actinic keratoses, 459 Alzheimer’s disease, 384
Activation energy, 86, 118 American Association of Pharmaceutical
Activity coefficient, 101, 114, 136, 273, 440 Scientists, 198
Acute toxic contact dermatitis, 512 Aminobenzoic acid, 215
Acyclovir, 56, 217, 280, 415, 429 Aminomethyl propanol, 326
Adapalene, 172 Ammonium lactate, 337
Adherens junction, 7 Amorolfine, 178, 179
Adhesive acrylates, 339 Analytical method integrity, 213
Adhesive performance, 342 Analytical method interference, 213
Adhesive polyisobutylene, 339 Anatomical site variations, 169, 206, 374
Adhesive silicones, 339 Androgen deficiency, women, 375
Adhesive wear-testing, 340 Androgenic hair loss, 176
Adsorptive capacity, 83 Androgens, 372
Adverse cutaneous reactions, 511 Angioedema, 530
Age Angiotensin-converting enzyme, 356
and percutaneous absorption, 436 Animal models, 223
physiological changes, 204 extrapolation to human, 225
Alcoholic gels, 448 Animal skin
Alcoholic lotions, 448 preparation, 207
Alkyl pyrrolidones, 286 suitability, 203
549
550 Index
Anthracene, 521 [Azone]

Anthralin, 42–43 orientation, 276
Anthrax, 55 physicochemical properties, 274
Antiacne agents, 422, 451 shape, 275
Antifungals, 451 species comparison, 277
Antihypertension, 381 structure, 275
Anti-inflammatories, 381 sulfur analogue, 278
Antimicrobial efficacy, 333 synergy with propylene glycol, 280
Antiseptics, 56
Antivirals, 451 Backing material requirements, 341
Apical corneocytes, 97 Bacterial infections, 430
Apocrine glands, 12 Barrier properties and skin reactions,
Apoproteins, 361 iontophoresis, 536
Apparent diffusion coefficient, 113 Barriers with dispersed phases, 83
Apparent diffusivity, 129, 136 Basal lamina, 6–7, 15
Appendage surface area, 175 Bentonite, 326
Appendageal pathways, 112 Benzalkonium chloride, 333
Appendageal shunts, 435 Benzo[a]pyrene, 215
Appendage-free stratum corneum, 175 Benzocaine, 336, 524, 531–532
Appendages, 435 Benzoic acid, 215, 333
Apresoline, 532 Benzophenone, 144, 524
Aqueous boundary layer, 116–118 Benzophenone flux, 145
Arachis oil-solvent partition coefficients, 121 Benzoxazinones, 357
Arrhenius equations, 85 Benzoyl peroxide, 49
Artificial neural network, 338 Benzyl alcohol, 333
Artificial skin, 218 Bergamot oil, 521
Aryl hydrocarbon hydroxylase, 167 Betamethasone, 305
Asymmetric polymeric membranes, 341 Betamethasone benzoate, 303, 443
Atabrine, 532 Betamethasone dipropionate, 237, 324, 415,
Athymic mouse, 219, 228 453
Atopic dermatitis, 425, 459 Betamethasone valerate, 410–412, 467
Atopic eczema, 44–45 Binary cosolvent systems, 295
Atrazine, 215 Binding to immobile components, 112
Atropine, 384 Binding to protein sites, 110
Attenuated total-reflectance Fourier transform Bioavailability
spectroscopy, 73, 217 dermatological formulations, 401
Autofluorescence, 219 drug absorption models, 411
Automated in vitro dermal absorption, 201 dynamic models, 419
Autoradiography, 219 effects of blood flow, 433
Azelaic acid, 49 effects of formulation excipients, 437
Azone, 30, 142–145, 222, 226, 272, 291, effects of skin microflora, 433
428 effects of skin pH, 433
analogues, 150, 273, 277 effects of skin surface lipids, 433
animal studies, 277 effects of temperature, 433
clinical studies, 279 influence of skin application site, 171
distribution in stratum corneum, 278 model systems, 408
estimated logP, 280 site of action, 404
human studies, 277 statistical considerations, 407
in vivo studies, 279 and therapeutic equivalence, 460
interaction with model membranes, 274 therapeutic models, 423
lipid disorder, 278 topical glucocorticosteroids, 463
mechanism of action, 274 Biochemical abnormalities, 29
Index 551
Bioequivalence Cellulose acetate, 222, 276

dermatological formulations, 401 Cellulose membranes, 297
drug absorption models, 411 Ceramide classification, 20
dynamic models, 419 Ceramide functions, 23
kinetic models, 414 Ceramides, 19–22, 97, 150, 224, 290
model systems, 408 Cetomacrogol, 331
statistical considerations, 407 Cetostearyl alcohol, 331
therapeutic equivalence, 460 Cetyltrimethyl ammonium bromide, 331–333
therapeutic models, 423 Challenge phase, 518
of topical corticosteroids, 409 Chemical damage, 26
topical antifungal products, 429 Chemical enhancers, 29, 272–273
topical glucocorticosteroids, 463 miscellaneous, 289
Biological variation, 460 synergy with supersaturated systems, 299
Biopharmaceutical considerations for Chemical potential gradient, 101, 273, 287
transdermal delivery systems, 338 Chemotactic cytokines, 9
Biopharmaceutical equivalence, 405 Chitosan, 341
Bipolar nature of the stratum corneum lipids, Chloramphenicol, 179
135 Chlorhexidine, 56
Bithionol, 521 Chloro-2-phenylphenol, 524
Block copolymers, 327 Chlorocresol, 333
Boltzmann’s constant, 113 Chlorofluorocarbons, 519
Boundary laminate, 343
Cholesterol, 23
Bovine hoof model, 178
Cholesterol esters functions, 23
Bovine udders, 218
Chronic irritation, 338
Bowen’s disease, 51
Chronic plaque psoriasis, 42
Brick and mortar model of the stratum
Chymotryptic enzyme, 17
corneum, 113
Cimetidine, 307
Bromhexidine, 280
Bromochlorosalicylanilide, 524 Cinnamates, 524
Bronaugh in vivo ring, 232 Circadian variation in plasma testosterone,
Bullous pemphigoid antigens, 7 373
Buprenorphine, 381 Clearance kinetics, 161
Burst effect, 75 dermal blood flow, 164
Butachlor, 215 vasoactive agents, 161
Butyl paraben, 334 Clearance limitations, 152
Clindamycin, 49
Clinical efficacy, 406
Caffeine, 205–206, 226, 244, 304
Clinical efficacy testing, 240
Calcipotriene, 44
Calcipotriol, 44 Clinical use of transdermal systems, 344
Calcium, 8, 18, 28 Clobetasol-17-propionate, 459
Calcium citrate, 325 Clonazepam, 297
Calcium tartrate, 325 Clonidine, 154–157, 532–534
Calculation of diffusional parameters, 72 diabetic neuropathy, 383
Candida albicans, 54, 521 Clonidine transdermal pharmacokinetics, 381
Candida nail, 54 Clotrimazole, 47, 54–55
Capillary bed, 206 Coagulation factors, 361
Capsaicin, 57, 242 Coal tar, 47
Carbomer, 325–330 derivatives, 521
Carbon dioxide emission, 530 in psoriasis, 43
Carboxymethylcellulose, 325 Coalescence, 328
Carboxyvinyl polymer, 325 Coenhancer effects, 299
Carbuncles, 55 Coenhancer systems, 272
Cellulitis, 55 Cognitive impairment, 384
552 Index
Cognitive response to estrogen, 372 Cutaneous reactions, transdermal delivery,

COLIPA, 211 529
Collodion matrix, 222 Cyclic urethanes, 290
Colloidal silicon dioxide, 326 Cyclobarbitol, 222
Comedolytic activity, 423 Cyclodextrin, 290
Comedolytic effects, 422 Cyclodextrins, 335
Complex barriers, 76 Cyclopirox, 430
Compound allergy, 427 Cyclosporine, 42, 173, 228, 305–308, 415
Concentration-depth profile of a permeant Cypermethrin, 242
within a membrane, 67, 215 Cytochrome P450, 214
Conductivity of human skin, 141 Cytokine involvement in keratinocyte
Confocal microscopy, 219 proliferation, 16
Constant diffusion coefficient, 66
Contact allergy, 338 Damage ratio, 144
Contact dermatitis, 46, 512 Damaged skin, 434
Contact hypersensitivity, 338 DDT, 226, 242
Contact photosensitivity, induction, 524 Deconvoluted diffusion profiles, 217
Contact sensitization, 9 Decubitus ulcers, 53
Contact sensitizers, 531 4-decyloxazolid-2-one, 289
Content uniformity, 508 1-[2-(decylthio)ethyl]azacyclopentan-2-one,
Contraceptive transdermal delivery, 371 289
Controlled-release devices, 75 Deethylase, 167
Coprecipitate dispersion systems, 297 Delayed contact hypersensitivity, 338
Coprecipitates, 295 Delayed onset muscle soreness, 284
Delayed-type hypersensitivity, 512
Corneocyte cohesion, 5
Demodex folliculorum, 50
Corneodesmosomes, 24
Depletion of solute in topical products, 154
Cornified cell envelope, 17
Dermac SR-38, 289–290
Correlation between in vitro and in vivo
structure, 289
transport, 169
Dermal anesthesia, 431
Corrosive reaction, 530
Dermal bioequivalence assessments, 169
Corrosivity of solutes, 147
Dermal formulations, safety considerations,
Corticosteroids, 160 511
Cosmetics Directive, 6th Amendment, 223 Dermal irritation, 218
Coumarins, 524 Dermal metabolism, 241
Creams, 326 Dermal toxicity, 239
Creep compliance, 339 Dermal-epidermal separation, 208
Crystal growth Dermatitis, 44
effects of additives, 293 Dermatological disease target site, 405
model, 293 Dermatological formulation, 319
process of nucleation, 293 bioavailability, 401
Crystalline states of lecithin, 110 bioequivalence, 401
Cultured skin, 218, 228 Dermatological product formulation, 321
Cumulative index of irritancy, 516 Dermatomed skin
Cumulative irritancy tests, 514 preparation, 207
Cumulative irritation, 529 thickness, 208
Cushing’s syndrome, 47 Dermatopharmacokinetics, 234
Cutaneous enzymes, 167 Dermatophyte infections, 54
Cutaneous metabolism, 167, 206, 214 Dermatotoxicity, 292
Cutaneous microdialysis, 239 Dermis, 11
Cutaneous pharmacokinetics, 152, 160 components, 2
Cutaneous porphyria, 52 diffusional resistance, 115
Index 553
Desmosomal degradation, 26 Dinitrochlorobenzene, 519, 531

Desmosomes, 7, 97, 110, 141–143 306 2,4-dinitrochlorobenzene, 519
Desonide, 472 Dipalmitoyl phosphatidylcholine, 274
Desquamation, 24, 147, 205, 243 Diphtheria, 55
Desquamation mechanisms, 24 Dipole-dipole interactions, 105
Determinants of epidermal flux, 121, 135 Discoid eczema, 44
Determinants of percutaneous absorption, Diseased skin, 90, 434
169 Disparate pore formation, 286
Dexamethasone, 307 Disruption of intercellular lipid bilayers, 141
Diabetic neuropathy, 383 Disruption of the corneocyte envelope, 141
Diazepam, 167 Dissolution kinetics, 138
Diclofenac, 161, 166–167, 222, 241 Distribution volumes, 113
Diethanolamine, 326 Dithranol, 43
Diethyl toluamide, 276 Dodecyl-N,N-dimethylamino isopropionate,
Diethyleneglycol monoethyl ether, 144 (see 289
also Transcutol) Dose applied in clinical practice, 448
Differential permeation, 341 Dose applied in experimental situations, 449
Diffusion cell Dose titration, 471
receptor chamber, 201 Dose-effect curves, 210, 461
receptor medium, 201 Dosing recommendations, 449
Diffusion cell design, 199–200 Dosing techniques, 451
reviews, 199 Draize erythema scores from iontophoresis,
Diffusion coefficient, 70, 82, 111, 272, 277, 537
287 Draize test, 517, 530
Diffusion in the vehicle, 140 Draize-Shelanski-Jordan sensitization, 362
Diffusion into a homogeneous membrane, 73 Droperidol, 384
Diffusion into semisolids, 218 Drug
Diffusion out of a precharged homogeneous in adhesive, 339
membrane, 75 candidate selection, 320
Diffusion path length, 111 in matrix, 339
Diffusion pathway, 110–113 in reservoir, 339
Diffusion profiles, deconvoluted, 217 Drug diffusivity, increase, 272
Diffusion through a homogeneous Drug incorporation in TDS, 341
membrane, 66 Drug release, 335, 409
Diffusion through laminates, 76 diffusion cell system, 410
Diffusion through shunt routes, 81 from formulations, 324
Diffusional pathlength, 70–72, 82, 85 methods for semisolid formulations, 410
Diffusional pathlength prediction, 218 methods for transdermal patches, 409
Diffusion-predicted model, 154 models, 409
Diflorasone diacetate, 205 testing in quality assurance, 337
Digoxin, 297 Drug reservoir materials, 342
Dihydroergotamine, 277 Drug solubility increase, 324
Dihydrotestosterone, 373 Drug structure, permeation relation, 438
Dihydroxyacetone, 291 Drug-containing adhesive layer, 303
Diisopropanolamine, 326 Drug-in-adhesive system, 342
Diltiazem hydrochloride, 385 Drug-skin interactions, 144–146
Dimethicone, 322 Drug-vehicle interactions, 138, 144, 271
Dimethyl formamide, 99, 131, 143 Drug-vehicle-skin interactions, 149
Dimethyl sulfoxide, 138, 143, 176–178, 272, Dual sorption model, 111
286 Duration of dose, benzene hexachloride, 456
Dimethylamino alkanoates, 291 Duration of dose, betamethasone
Dimethylnitrosamine, 209 dipropionate, 455
554 Index
Duration of dose, clobetasol propionate, 456 [Enhancement ratios]

Duration of dose, dithranol, 456 lidocaine, 291
Duration of dose, hydrocortisone, 455 minoxidil, 291
Duration of dose, ibuprofen, 455 morphine sulfate, 291
Duration of dose, minoxidil, 455 nifedipine, 291
Dye permeability, 138 prilocaine, 291
Dyphylline, 304 progesterone, 291
Enhancer incorporation in TDS, 341
Eccrine glands, 12 Enhancer synergy, 273
Econazole, 54–55, 304 Enhancing effect of vitamin C, 146
Eczema, 44, 55, 349 Environmental effects, 1
Eczema classification, 44 Environmental Protection Agency, 197
Eczema treatment, 46 Epidermal
Effect of gender on the irritation rates, 534 cell cultures, 214
Effect of hydration, 128 cell strata, 6
Effect of race on irritation rates, 534 cohesiveness, 7
Effect of TD estrogen replacement therapy comedones, 422
on lipoprotein, 359 differentiation, 15
growth kinetics, 414
Effect of temperature diffusion through
homogenates, 214
membranes, 85
membranes, 206
Effective diffusivity in the stratum corneum,
repair mechanisms, 25
135
resistance, 114
Effective permeability constant, 115
Epidermis, 4
Elastic index, 339
components, 2
Elastosis, 520
diffusional resistance, 115
Electric double layer, 328
functions, 4
Electromagnetic radiation, 519
regenerated, 92
Electroporation, 321, 538
Epidermophyton floccosum, 55
Electroporation effect of current and voltage,
Erectile dysfunction, 375, 385
539
Eryings’ absolute reaction rate theory, 79
Electrostatic headgroup interactions, 30
Erysipelas, 55
Electrostatic repulsion, 328
Erythema, 426, 530–535
EMLA, 431
Erythema grading, 426
Emollients psoriasis, 43
Erythrodermic psoriasis, 42
Emulsifying wax, 331
Erythromycin, 49, 215
Emulsion evaporation, phase diagram, 289 Escaping tendency, 114, 288
Emulsion physical stability, 328 Estimation octanol-water partition
Emulsion stabilization, 327–329 coefficients, 109
Emulsions, 326 Estradiol, 157, 168, 204, 215, 222, 229–230,
Energetic epidermal injection, 288 243, 281, 298, 306, 321, 365, 532–
Energy barrier technique, 328 533
Energy of dissolution in the stratum 17␤-estradiol fertility regulation in females,
corneum, 104 371
Enhanced low-concentration formulations, Estrogen replacement therapy, 240
468 safety, 363
Enhanced skin permeability, 141 Estrogens, 357
Enhancement ratios Ethanol-treated human epidermal
alprazolam, 291 membranes, 100
diclofenac, 291 7-ethoxycoumarin, 215
hydrocortisone, 291 Ethyl paraben, 334
indomethacin, 290–291 Ethylaminobenzoate, 215
isosorbide dinitrate, 291 EUFEPS, 473
Index 555
European Centre for Ecotoxicology and Follicle contribution to transdermal delivery,

Toxicology of Chemicals, 198 177
European Centre for the Validation of Follicle density, 177
Alternative Methods, 198, 211, 227 Follicle-free skin, 172
European Cosmetic Toiletry and Perfumery Follicle-rich areas, 177
Association, 198 Follicles, depth of penetration, 176
Eutectic mixtures, 150 Follicular delivery, 307, 435
Evolutionary operation, 501 Follicular density, 172
Extent of absorption, 462 Follicular density, animal skin, 224–227
Extracellular lipid, 440 Follicular disposition, 308
Follicular penetration, 172
particle size dependency, 172
Facial distribution of sunscreen, 450 Follicular surface area, 175
Facial-stinging test, 516 Follicular transport, 226
Factorial block, 503 Folliculitis, 55, 515
Factorial design in process optimization, 501 Food and Drug Administration, 198, 227,
Factors affecting absorption from topical 234, 336, 344, 409, 415, 449, 473
products, 432 Formalin, 531
Factors affecting skin flux, 135 Formation of the cornified envelope, 18
FDA (see Food and Drug Administration) Formulation, 319
FDA/AAPS-sponsored workshop, 412 biopharmaceutical considerations, 338
Fentanyl, 157, 215, 376, 532–535 design considerations, 339
Fentanyl and respiratory depression, 378 drug solubility, 336
Fentichlor, 524 pH, 333
Fertility regulation, 371 preservation, 333
Fick’s first law of diffusion, 61–63, 65, 69, strategies, 271
218, 272 48-hour irritancy tests, 514
Fick’s laws solutions, 65 Fragmental analysis, 117
Fick’s second law of diffusion, 64–66, 71– Franz cell, 199, 411
73, 85, 218 Free energies for permeation, 120
Film coating technology, 303 Free-energy group contribution, 110
Film-thickness and permeation rates, 453 Freezing, effect on permeability, 205
Finite dose, 451, 465 Fresh versus frozen skin, 206
technique, 158, 208 Fucidic acid, 55
Finite flux, 65 Fungal infections, 429
Finite versus infinite dose, 209 Furuncles, 55
Finn chambers, 515, 522–525
First-pass metabolism, 169 Gel formulation preservatives, 325
Flocculation, 328 Gel network theory, 331–332
Flow through diffusion cell, 199–201 Gels, 325
Flunitrazepam, 297 Gene therapy, 308
Fluocinolone acetonide, 463 Gibbs free energy, 63
Fluocinonide, 132 Glaucoma, 383
Fluoranthene, 521 Glossary of dermatological terms, 58
5-fluorouracil, 51, 129, 141, 150, 204, 222, Glucosylceramides, 20
277, 446, 471 Glutaraldehyde, 57
Flurbiprofen, 150, 209–212, 302, 381 Glycerogelatin films, 218
Flux from different vehicles, 136 Glyceryl dilaurate, 308
Flux from saturated solutions, 144 Glyceryl monostearate, 330
Flux prediction, 103 Glyceryl trinitrate, 227
Flux solubility parameter, 107 liver cirrhosis, 356
Flynn’s dataset, 120–122 local cutaneous changes, 356
556 Index
[Glyceryl trinitrate] Homogenization

primary dysmenorrhea, 356 mixing equipment, 506
Raynaud’s phenomenon, 356 mixing operations, 506
pharmacokinetics, 354 mixing parameters, 507
Goeckerman regimen, 43 Hormone replacement therapy, 358
Grading of disease states, 424 and bone mineral density, 359
Grading scales for erythema, 516 and plasma lipoproteins, 361
Grading system for photoallergy, 525 HPE-101, 289
Granulocyte-macrophage colony-stimulating structure, 289
factor, 9 Human epidermal membrane integrity, 212
Group contributions, 109, 113 Human graft skin, 228
Guar gum, 325 Human immunodeficiency virus (see HIV)
Gums, 325 Human skin equivalents, 218
Guttate psoriasis, 42 Humidity-induced penetration changes, 129
Hydrated stratum corneum, 25
Hydration, 25, 92, 128–130, 146, 226, 341
Hair amino acid composition, 15
Hydrocarbon base, 322
Hair dye, binding to human epidermis, 138
Hydrocortisone, 165, 199, 206, 243, 277,
Hair dyes, 215
297, 305–307, 410–411, 434, 441,
Hair follicle penetration, 176
463, 471
Hair follicles, 12, 81, 173–175, 207, 224–
Hydrocortisone acetate, 298, 411, 441, 470
226
Hydrocortisone distribution, 165
Hair growth cycle, 421 Hydrogen bonding, 105–108, 113, 276–278,
Hair keratin, 177, 224 337, 440
Hair shaft, 176 accepting ability, 108
Hairless mouse, 226 capabilities, 119
skin permeability, 204 constant, 128
Halogenated salicylanilides, 524 donating ability, 108
H-bond acceptor ability, 121 interactions, 107
H-bond donor ability, 121 vehicles, 143
Heat of fusion, 132 Hydrogen peroxide, 53
Heat premix processes, 304 Hydrophile-lipophile balance, 330
Hectorite, 325 Hydroquinone, 465
Hemidesmosomes, 7 Hydroxyethylcellulose, 325
Henry’s law, 104 Hydroxypropyl cellulose, 533
Heparin, 278 Hydroxypropyl-␤-cyclodextrin, 335
Herpes simplex, 45, 52–56, 160, 280, 416, Hyperflux, 288
427 Hyperkeratotic plaques, 51
Herpes zoster, 56, 432 Hypodermis, 12
Heterogeneous skin permeability models, 113 Hypogonadism, 372
Hexachlorophene, 6, 532
Hexyl paraben, 334 Ibuprofen, 209, 222, 241, 277, 283–285, 320
High-permeability solutes, 119 Ichthyosis, 31
Higuchi Physical Model, 438 Ichthyosis vulgaris, 26
Hildebrand solubility parameter, 105–107 Ideal solubility, 104
Historical development skin transport- Ideal solution, 132
structure relations, 117 5-idoxuridine, 56
HIV, 47, 57 Iloprost, 355
HIV-positive patients, 384 Immediate type hypersensitivity, 512
HLA phenotypes, 41 Immunosuppression, 56, 426
HLB system, 330 Impetigo, 55
Homogeneity, product physicochemical In use conditions, 246
properties, 505 In vitro measurements, limitations, 247
Index 557
In vitro methodology Intermittent dosing, clobetasol-17-propionate,

advantages, 198 459
disadvantages, 199 Intermittent transdermal GTN therapy, 355
use of data, 199 Inter-subject variability, 203
In vitro model systems, 408 Intracellular keratin, 97, 140–143
In vitro percutaneous absorption, 451 Intrafollicular delivery, 176
In vitro release Intra-subject variability, 203
factors, 412 Intrauterine device, 359
validation of, 411 In-use modelling, 208
In vitro-in vivo correlations, 169–170, 241, Ion pairing, 276
412, 453 Ion-dipole interactions, 108
In vivo animal models, 223 Ionic surfactants, 143, 330
In vivo bioavailability, 234 Iontophoresis, 280, 321, 535
In vivo experimental procedures, 227 Iontophoresis
In vivo model systems, 413 adverse reactions, 537
In vivo skin permeation, 223 burns, 538
In vivo studies effect of current and voltage, 539
application procedures, 227 effect of race, 536
permeant sampling, 227 virus activation, 538
Indomethacin, 150, 277, 289–294, 342, 412 Irreversible damage ratio, 144
Indomethacin gel, 304 Irritancy tests, 513
Induction phase of allergic reactions, 517 Irritant contact dermatitis, 46, 512
Infantile acne, 48 Irritation, 28, 218, 281, 291, 338, 529
Infectious conditions, 427 predictive testing, 530
Infinite dose, 451, 465 Ischemic leg ulcers, 53
technique, 208 Isolated perfused limb model, 228
Infinite versus finite dose, 209 Isolated perfused porcine skin flap, 221
Inflammatory mediators, 513 Isolated perfused tissue models, 218
Influence of application conditions, 446 Isopropyl myristate, 131–135, 144, 147–150,
Influence of occlusion, 446 226, 242, 276, 290, 299
Influence of rubbing, 446 Isosorbide dinitrate, 157, 357
Influence of the vehicle, allergic contact Isotretinoin, 49–52, 308
dermatitis, 427 Itching, iontophoresis, 537
Ingram’s method, 43
Inner root sheath, 176 Keratin, 5, 94, 110
Insect bites, 58 affinity, 147
Integrins, 7 filaments, 25
Interactions between drug, vehicle, and skin, hair and wool, 224
440 intermediate filament, 7
Interagency Testing Committee, 197 Keratinization process, 19
Intercellular cohesion, 286 Keratinocytes, 110
Intercellular diffusion pathway, 120 Keratohyalin granules, 20
Intercellular lipid disruption, 142 Keratolytics, 29
Intercellular lipid pathway, 96, 136, 174 Ketoconazole, 47, 54
Intercellular lipid transport pathway Ketoprofen, 212, 281, 321
evidence, 97 Kinetic models, 337
Intercellular lipids, 20, 272, 536
Intercellular space, 110 Lacuna domains, 100
Intercellular volume fraction, 97 Lacunae, 26, 27
Interferon, 305–308, 428 Lag time, 113, 152–153, 172, 179, 211, 217
Interleukins, 9, 15, 42 Lamella width, 97
Intermittent dosing, 5-fluorouracil, 459 Lamellar granules, 17
558 Index
Lamina densa, 7 [Liposomes]

Lamina lucida, 7 minoxidil, 304
Langerhans cells, 6–8, 46 nonionic, 307
Langmuir trough, 274 plasmid DNA, 308
Lanolin, 215, 532 retinoic acid, 304
Laser Doppler flowmetry, 247, 420 skin lipids, 305
Laser scanning confocal microscopy, 219 tetracaine, 305
Lassar’s paste, 43 triamcinolone, 305
Laurocapram, 222, 272–273, 428 (see also triamcinolone acetonide, 304
Azone) Liquid crystalline state, 110
Lecithin organogels, 222 Liquid membranes, 222
Leukotrienes, 42 Lisinopril, 354
Levels of differentiation, 6 Local blood circulation, 218
Levomethadone, 381 Local enhanced drug delivery, 241
Levonorgestrel, 359 Local metabolism of solutes, 152
Levonorgestrel fertility regulation in females, Local skin reactions, 376
371 Localized drug delivery, 176
Lichen planus, 50 Log-linear profiles for transdermal products,
Lidocaine, 161, 222, 230, 305, 385, 431 158
Limonene, 145 Long-term smoking-cessation efficacy, 349
Lindane, 58, 147, 226 Loss of vehicle, 447
Linear free energy, 108, 116 Low-dose therapeutic activity, 292
Lipid Lupus erythematosus, 52
composition of human sebum, 14 Lymphatic system, 12
condensation state, 275 Lymphatic transport, 161
content of aged skin, 204
distribution in diseased skin, 23 Macrolide antibiotic, 426
fluidization, 286 Macroscale mixing, 506
lamellae, 97 Magnesium aluminum silicate, 325
monolayers, 274 Malathion, 58, 226, 532
packing, 24 Manual skin-sectioning technique, 417
pathways, 99, 116, 174 Mass balance, 211, 227, 451
stabilization, 276 Mathematical models, 199
synthesis inhibitor, 100 Maximal flux, 129, 132, 149
vesicles, mode of action, 306 Maximum flux antimicrobial compounds,
Lipid-protein-partitioning (LPP) theory, 141 128
Lipoidal membrane diffusion, 79 Maximum flux sunscreens, 128, 135
Lipophilic enhancers, 141 Mecamylamine, 344
Liposomal entrapment, 173 Medial corneocytes, 97
Liposomes, 142, 222, 278, 304, 335, 428 Melanin, 9, 419, 520
betamethasone, 305 Melanins synthesis, 10
caffeine, 304 Melanocytes, 6–9
cimetidine, 307 Melanogenesis, 9
cyclosporine, 305, 308 Membrane fluidization, 286
dyphylline, 304 Membrane interfaces, 80
econazole, 304 Membrane transport mathematical principles,
enhanced skin retention, 305 61
follicular delivery, 307 Menthone, 145
hydrocortisone, 305 Merkel cells, 6, 11
interferon, 305, 308 Mesalamine, 352
isotretinoin, 308 Metabolically active intact skin, 214
lidocaine, 305 Metabolism, 167
Index 559
Methods used to assess the effects of Nail keratins, 177

enhancers on membrane permeability, Nail lacquers, 179
143 Nail matrix, 14, 177
Methotrexate, 160, 278–280, 415 Nail plate, 14, 177
Methyl paraben, 138, 333, 443 delipidization, 178
Methyl salicylate, 161 lag time, 179
Methylcellulose, 325 maximum flux, 178
Metronidazole, 50, 56, 276–280 permeation, 177
Mexenone, 524 Nails psoriasis, 43
Miconazole, 47, 54–55 Naloxone, 381
Microcrystalline wax, 322 Naproxen, 222
Microdialysis, 161, 239 National Health Service (UK), 347
disadvantages, 239 National Physics Laboratory, 522
Microemulsion vehicles, 448 Natural clay thickeners, 326
Microscale mixing, 506 Natural moisturizing factor, 26
Microvascular reactivity, 205 Neomycin, 532
Microviscosity, 274 Neuroendocrine response to estrogen, 372
Miliaria, 536 Neuropathic pain, 379
Minimal erythema dose, 525 Neutralizing agents, 326
Minolta Chromameter, 420 NexACT 88, 289–291
Minoxidil, 209, 283, 291, 304 animal toxicity studies, 292
Mixed cosolvent systems, 295 structure, 289
Mixer power consumption law, 507 Nicardipine, 202
Model intercellular lipoid matrix, 121 Nicorandil, 226, 357
Model membranes, 219 Nicotine, 157, 240, 344, 532–535
Model of partitioning into the protein gum, 346
domains, 110 poisoning, 353
Modes of delivery to the skin, 403 transdermal system, 344
Molecular diffusion, 62 treatment of Alzheimer’s disease, 353
Molecular dispersion systems, 295 treatment of Tourette’s syndrome, 353
Molecular graphics, 274 treatment of ulcerative colitis, 351
Molecular size as determinant of stratum Nifedipine, 299
corneum permeability coefficients, 119 supersaturated solutions, 294
Molecular volume, 320 Niosomes, 306
Molluscum contagiosum, 45, 57 Nitrazepam, 297
Monobenzyl ether of hydroquinone, 531 Nitrendipine, 280
Montmorillonite, 326 Nitroglycerin, 157, 167, 280, 320, 412, 419,
Morphine, 376, 382 434, 532–534
Multilayer barrier, 100 2-n-nonyl-1,3-dioxolane, see SEPA
Multiple dose Nodular acne, 48
diflucortolone-21-valerate, 458 Nonideal solutions, 287
halcinonide, 459 Nonionic liposomes, 307
hydrocortisone, 457 Nonionic surfactants, 142, 306, 330
iododeoxyuridine, 458 Nonlinearity in flux-convention profiles, 136
Multivariate optimization, 499 Nonsink conditions, 136
Mupirocin, 55 steady-state flux, 115
Musk ambrette, 524 Non-steady-state condition, 150
Myasthenia gravis, 50 Nonsteroidal antiinflammatory agents, 132,
222, 241, 320, 521
N-0915, 275–278 Nonsteroidal antiinflammatory drugs,
Nail, 12 percutaneous penetration, 134
Nail amino acid composition, 15 Norwegian scabies, 57
560 Index
n-pentyl-N-acetylprolinate, 292 Paronychia, 54

Nucleation, 293 Partially miscible systems, 287
Nude mouse, 219 Partition coefficient, 70, 72, 77–78, 82 100–
Number of applications, 515 104, 110–113, 143, 160, 272, 277, 440
prediction, 103
Occlusion, 128–129, 135, 168, 201, 225– Partitioning at the SC-viable epidermis
227, 232, 307, 340 interface, 135
Occlusive effects, 335 Passive diffusion, 62
Occulsive agents, 143 Peak urinary excretion rate, 158
Occupational acne, 48 Penetration enhancement, 271
Occupational Safety and Health Penetration enhancers, 85, 138–143, 178,
Administration, 197 226, 343, 445
Octanol-vehicle partition coefficients, 128, Penetration retardation, 150, 276
147 Penetration-enhancing solvents, 99
Octanol-water partition coefficient, 108, 116, Percutaneous absorption
121, 162, 277, 320 age variability, 204
for phenols, 127 from binary solutions, 149
Octyl paraben, 334 determination, 197
Octyl salicylate self-association, 138 of hair dyes, 147
Oil phase, 323 in vivo, 223
Oil-in-water creams, 327 of nifedipine, 150
Ointment base, 322 race variability, 205
Ointment formulation preparation, 324 sex variability, 204
Ointments, 322 Percutaneous penetration salicylates, 134
Oleic acid, 30, 141–142, 272, 302–305, 445 Perfused pig ear model, 218
Oleyl alcohol, 290 Perfused tissue models, 218
Oligonucleotide delivery, 280 Permeability
Oligonucleotides, 176 of alcohols, 127
Omoconazole nitrate, 147 coefficient, 72–77, 99, 107, 112–120,
Onycholysis, 43, 55 128–129, 138, 143, 158, 160–161,
Onychomycosis, 14, 179 208, 334
Optical sectioning methods, 219 of phenols, 127
Optothermal transient emission radiometry, predictions for other vehicles, 128
234 of very polar solutes, 99
Osmotic gradients, 307 Permeable facial skin, 516
Osmotic repulsion, 328 Permeant application
Outer root sheath, 176 amount, 208
Overmixing, 507 contact time effects, 210
Oxazolidinones, 290 dose level, 209
recommended dose levels, 210
Pain management, 376 removal of applied dose, 208
p-aminoacetophenone, 411 technique, 208
p-aminosalicylic acid, 50 Permeation experiments
Papaverine, 281 duration, 211
Papillary dermis, 100 lag time, 211
Para-aminobenzoic acid, 524 number of replicates, 212
Paraben permeability coefficients, 334 permeant analysis, 213
Parabens, 325, 333 recovery of permeant, 213
Paracetamol, 178 replicate reduction, 212
Parallel pathways, 81, 120 sample interval, 211
Paraquat, 169, 225 skin integrity, 212
Parathion, 226 steady state flux, 211
Index 561
[Permeation experiments] Phototoxicity assays, 521

temperature, 212 Physicochemical properties of a permeant,
tissue integrity, 211 92
tissue sample analysis, 213 Physiological disposition model, 167
Permeation in diseased skin, 319 Pigment darkening response, 520
Permeation patterns, cumulative, 209 Pigmentation, 9
Permethrin, 58, 242 Pilocarpine, intraocular pressure, 383
Perturbation of barrier function, 28 Pilosebaceous gland, 219, 307
Petrolatum, 246, 324 Pindolol, 280
Pharmaceutical equivalence, 405 Piritramid, 381
Pharmaceutical Manufacturers Association, Piroxicam, 132, 222, 298, 521
343 Piroxicam flux, 141
Pharmacodynamic measurements, in vivo, Pityriasis rosea, 51
239 Pityriasis versicolor, 54
Pharmacodynamic parameters, 406 Pityrosporum orbiculare, 54
Pharmacokinetic compartment and Placebo effects, 241
physiology models, 155 Plasma concentration profiles, 154, 337
Pharmacokinetic modelling, 337 Plasma concentration-time profile after
Pharmacokinetic principles, 152 topical patch application, 156
Pharmacokinetics, 405 Plasma levels after multiple dosing, 158
cutaneous, 160 Plasma lipolytic enzymes, 376
of clonidine, 157 Plasma protein binding, 165
of estradiol, 157 Plasmid DNA, 308
of fentanyl, 157 Podophyllin, 57
of isosorbide dinitrate, 157 Polar head group, 176
of nicotine, 157 Polar pathways, 116, 141
of nitroglygerine, 157 through the intercellular lipids, 99
of scopolamine, 157, 158 Poloxamer, 188, 202, 327
of testosterone, 157 Poly-2-hydroxyethylmethacrylate, 222
of timolol, 157
Polyacrylic acid polymers, 330
of triprolidine, 157
Polydimethylsiloxane, 222, 322
Phase boundary conditions, 288
Polydimethylsiloxane membrane, 131, 138
pH-controlled silicone reservoir, 385
Polyethylene glycol-20-oleyl ether, 202
Phenacetin, 178
Polymeric thickeners, 323
Phenolic permeability coefficient, 121
Polyoxyethylene alkyl ethers, 306
Phenols, percutaneous penetration, 134
Polyoxyethylene(10)oleyl ether, 306
Phenols, permeability coefficients, 127
Polyoxyethylene(10)stearyl ether, 308
Phenoxyethanol, 333
2-phenoxyethanol, 209 Polyoxyethylene(3)lauryl ether, 306
Phenylephrine, 228 Polyoxyethylene(3)stearyl ether, 306
Phospholipid liposomes, 304 Polysaccharide gums, 325
Photoaging, 520 Polysorbate 80, 531
Photoallergenicity, 281, 338, 511, 523 Pompholyx, 44
Photoallergenicity assays, 524 Population size for skin toxicity testing, 518
Photobleaching, 219 Porcine skin flap, 218
Photodermatitis, 45 Pore pathway, 99
Photoirritancy, 338, 512, 520 Pores, 82
Photomaximization test, 524 Postherpatic neuralgia, 57, 432
Photosensitivity, 511–512, 519 Postoperative pain, 377
Phototoxic agents, identification, 522 Potassium chromate, 531
Phototoxic reactions, animal, 521 Potassium permanganate, 56
Phototoxic reactions, in vitro models, 521 Potts and Guy equation, 119, 135
Phototoxicity, 281, 520 p-phenylenediamine, 531
562 Index
Prediction of efficacy, 160 Race and percutaneous absorption, 205, 436

Prediction of solubility, 103 Radiolabeled permeants, 211
Prednisolone, 352 homogeneity, 211
Preformulation, 321–322 incorporation, 211
Preservation of semisolid formulations, 333 lack of specificity, 211
Preservative antimicrobial properties, 334 Random walks, 62
Preservative physicochemical properties, 334 Raoult’s law, 103, 135–136
Preservatives, 323–325, 333 deviation, 103
skin penetration, 334 Rate of absorption, 337
Pressure sensitive adhesive, 339, 533 Rate of elimination, 337
Prilocaine, 431 Rate-controlling biopolymers, 341
Primary irritation, 338 Rate-controlling membrane, 341
Process optimization, 501 Rating scales in dermatology, 425
fractional factorial design, 504 Receptor chamber volume, 199
Processing of pharmaceutical dosage forms, Receptor medium
500 gaseous, 202
Prodrugs, 321, 435 ideal, 201
Product homogeneity, 499 microbial growth, 202
Profilaggrin, 19 nonaqueous, 202
Progesterone, 132, 243, 342 non-liquid, 202
cream, 371 pH, 202
Progestogen, 357 Reconstitution of barrier function, 26
Propionibacterium acnes, 49 Regional delivery, 403
Regional variability in skin absorption, 169
Propranolol, 277, 342
Regional variability in skin desquamation,
Propranolol hydrochloride, 291
169
Propyl paraben, 138, 333
Regular solution theory, 120
Prostaglandin E1, 247
Relative bioavailability, 401
Prostaglandin E2, 280
Release from the topically applied
Prostaglandins, 513
formulations, 336, 438
Protein binding, 165–166
Release kinetics, 342
Proximal corneocytes, 97 Release liner requirements, 341
Psoralen, 521, 524 Reservoir effect, 168
Psoriasis, 26, 41, 55, 349, 415, 425, 434, Residual analysis, 230
459, 472 Respiratory depression, fentanyl, 378
arachidonic acid metabolism, 41 Retinoic acid, 214, 291, 304, 422, 474
clinical patterns, 42 derivatives, 440
nails, 43 Retinoids, 451, 463
pustular, 43 Retinyl palmitate, 215
Psoriasis Severity Index (PASI), 425 Rhino mouse model, 421
Psoriatic skin, 92 Ringworm, 54
Pull effect, 106–107, 149 Risk assessment, 197, 208, 241, 247
Push effect, 106–107 Risk subjects, 513
Pustular psoriasis, 43 Rogaine, 283
PUVA, 524 Role of appendages in skin transport, 169
PVP coprecipitates, 297 Rosacea, 46–52, 56
Pyrene, 521 Rotating diffusion cell, 222
2-pyrrolidone, 99
Safety
Quantitative structure-activity relations, 149 dermal formulations, 511
Quasi-steady-state absorption model, 154 estrogen replacement therapy, 363
Quinoline antimalarial drugs, 521 transdermal formulations, 511
Index 563
Salicylate esters, 320, 416 Skin

Salicylic acid, 230, 412, 416 adhesion factors affecting, 340
Sandalwood oil, 524 appendages, 12
Sarcoptes scabei var. hominis, 57 biopsy, 415
Saturated systems, 114 blanching, 160, 237, 240, 415
Saturation, degree of, 295 blood flow, 11, 420, 433
SC lipid-viable epidermis partition blood vessels, 11
coefficient, 112 capacitance measurements, 536
Scabies, 57 color changes, 419
Scale-up of dermatological dosage forms, diffusion cells, 198
499 disorders, 41
Scalp psoriasis, 44 enzyme activity, 167, 214
Scanpore tape, 515, 522–525 enzyme distribution, 168
Scopolamine, 157, 532–534 enzyme induction, 214
SC-vehicle partition coefficient, 115 enzyme nature of, 214
SC-vehicle partition coefficients, calculated, equivalents, 218
107 first-pass effect, 167
SC-water partition coefficients, 110 flap, porcine, 218
Sebaceous glands, 172–176, 308 follicles, 307
Sebhorreic lipids, 92 function, 1
Seborrheic eczema, 44, 47 grafting, 228
Sebum, 12, 48, 92, 174–176, 247, 306 gross structure, 4
Sebum as a barrier, 92 infection, 54
Selegiline, 384 bacterial, 55
Semi-automated skin-sectioning technique, irritation, 149
416 and pKa, 149
Semisolid formulation constituents, 323 lipid, 5
Sensation, iontophoresis, 537 biophysics, 234
Sensitization, 281, 291 liposomes, 305
capacity, 529 membrane
humans versus guinea pigs, 532 models, 409, 412
tests, 517 preparation, 203–206
SEPA, 273, 281–286, 385 selection, 203
efficacy, 281 variation, 203
ibuprofen gels, 283 metabolism, 167, 435
in vitro-in vivo correlation, 283 assessment in vitro, 213
minoxidil solutions, 283 effects on skin transport, 167
mode of action, 286 microflora, 433
safety, 281 penetration enhancers, mechanisms of
structure, 281 action, 142
Serum lipoproteins, 361 penetration from prototype formulations,
Serum minoxidil levels, 283 338
Sesame oil, 173 permeability
Shingles, 56 age variability, 204
Side-by-side diffusion cell, 199 antimicrobials, 128
Silicone elastomers, 342 effect of age, 436
Silicone membranes, 218, 298 effect of freezing, 205
Silver sulfadiazine, 53 effect of race, 436
Sinitrodil, 357 effect of vehicle, 436
Site of action, 320, 404 enhancement, 271
Site of application, variation, 382 log normal, 203
Site variations, 434 modulation, 271
564 Index
[Skin, permeability] Solubility in the vehicle, 129

race variability, 205 Solubility in the viable epidermis, 131
ranking, 225 Solubility parameter, 106, 121, 145, 272
sex variability, 204 Solubility parameter for a keratin fragment,
site variability, 206 110
statistical distribution, 203 Solubility prediction, 104
sunscreens, 128 Solute concentration in the stratum corneum,
testosterone, 150 135
vehicle effects, 271 Solute concentration in vehicle, 136
permeation Solute effects on stratum corneum
of butyl paraben, 168 permeability, 136
guidelines and protocols, 198 Solute structure-transport relations, 116–120
large animals, 224 Solute-solvent interactions, 104
mathematical principles, 61 Solute-vehicle and stratum corneum
of preservatives, 334 interactions, 105
of propyl paraben, 168 Solvatochromic approaches, 108, 117
pH, 129, 433 Solvent delipidation, 28
pharmacokinetics, 152 Solvent effects on bioavailability, 445
polarity, modulation, 272 Solvent evaporation during film coating, 303
protection, 89 Solvent flux through stratum corneum, 138
reactions to clonidine, 534 Solvent pathway, 176
reservoir, 414 Sonophoresis, 321
sandwich flap, 228 Sorbic acid, 333
limitations, 230 Sotalol, 291
model, 417 Species differences in percutaneous
sectioning, 417 absorption, 169
site variability, 462 Species effect on in vitro absorption, 171
storage conditions, 205 Species ranking, 225
stripping, 68, 205, 213–215, 234, 414 Species variation
advantages, 237
skin lipids, 224
amount of stratum corneum removed,
skin structure, 224
237
stratum corneum lipids, 224
protocol, 234
stratum corneum thickness, 224
sources of variability, 237
Sphingolipids, 20
validation, 237
Squamae cohesion, 141
structure, species variation, 224
Squamous cell carcinoma, 51–52
surface lipids, 433, 448
Stable emulsion formulation, 332
thinning effect of topical corticosteroids,
459 State of hydration, 140
toxicology, 147 Static diffusion cell, 199
transport, 89 Steady-state concentration-distance profile,
transport principles, 100 114
turnover, 243 Steady-state concentrations after topical
water content, 234 application, 102
Smoking cessation, 344 Steady-state flux, 111
Sodium fusidate, 55 for methylparaben, 131
Sodium lauryl sulfate, 26, 212, 272, 278, Steady-state flux progesterone, 131
307, 328–331, 420, 517, 531–532 Stem cells, 6
Sodium salicylate, 276 Steric repulsion, 328
Solar keratoses, 51 Stratum corneum, 15–17
Solar urticaria, 523 amino acid composition, 15
Solid dispersion systems, 295 barrier, 92
Solubility in skin lipids, 128 historical development, 93
Index 565
[Stratum corneum] [Supersaturated systems]

cell, 5 stabilization, 293
chymotryptic enzyme, 24 synergy with chemical enhancers, 299
chymotryptic enzyme, 24 Supersaturation, 132–136, 150, 222, 442,
density, 4 448, 470
development, 14 Surface area of application, 454
hydration, 4, 25 Surface-active agents, 28, 323
intercellular lipids, composition, 20 Surfactant system choice, 330
isolation, 208 Surfactant systems, 327
proteins, 19 Surfactants
reservoir, 168, 414 adsorption of, 147
thickness, 4 effect on skin permeation, 142
transport pathways, 94 Susten skin depot, 232
vehicle partition coefficient, 101, 158, Sweat ducts, 81, 174, 226
161, 320 Sympathoadrenal activation, 382
water, 25 Synthetic membrane models, 409–411
Stratum granulosum, 15–17 Systemic pharmacokinetics, 153
Stratum lucidium, 15
Stratum spinosum, 15–17 T lymphocytes, 9, 42–46
Streptomycin, 532 Tacrolimus, 426
Structural changes in the vehicle, 447 Taguchi methods, 501
Structure-activity studies, 108, 272, 278 Tape stripping, 92, 147, 168, 206–207, 213–
Structured lipid matrices, 222 215, 228–232, 302, 307, 414
Subcutaneous vasculature, 206 amount of SC removed, 215
Subcutis, 12 distribution profiles, 215
Substantivity, 147, 320 mass of SC removed, 216
Sucrose permeability, 212 validation, 216
Sun protection factor, 449 Target site of dermatological diseases, 405
Sunlight, 519 Targeted follicular transport, 176
Sunscreens, 6, 52, 147, 234, 320, 449, 524 Tazarotene, 167
SUPAC-SS, 335 Tazarotenic acid, 167
Supersaturated solutions, 288–292 TD Nicotine
Supersaturated systems Alzheimer’s disease, 353
crystal growth, 292 depression, 353
effects of additives, 292 Tourette’s syndrome, 353
effects of coenhancers, 299 Telangectasia, 50
effects on human skin, 299 Temperature dependence, 100
estradiol, 298 Terpenes, 141, 144–145, 286, 445
flurbiprofen, 302 Testosterone, 157, 204, 215, 226–229, 243,
flux from, 297 321, 342, 372–373, 532–535
formation, 294 Tetracaine, 305
generation, 295 Tetrachlorosalicylanilide, 531
hydrocortisone acetate, 298 Tetracycline, 49–50 431, 521
in the stratum corneum, 296 Tetrahydroxypropyl ethylenediamine, 326
inadvertent during manufacture, 303 Theophylline, 149
indomethacin, 304 Therapeutic action, onset, 247
isopropyl myristate, 299 Therapeutic activity, correlation with in vitro
membrane transport, 296 data, 247
nifedipine, 299 Therapeutic equivalence, 406, 460–465
penetration enhancement, 294 Thermodynamic activity, 103, 136, 147, 209,
physicochemistry, 292 273, 286–296, 411, 428, 440
piroxicam, 298 Thermodynamic activity gradients, model,
silicone membranes, 298 296
566 Index
Thermodynamic activity of drug at the skin Transdermal estradiol-norethisterone, 358

target site, 428 Transdermal estrogen, 516
Thiazide diuretics, 50, 521 in men, 372
Thickness of applied medicaments, 84 Transdermal feasibility, 337
Thin-film transdermal patches, 295 Transdermal fentanyl, 376
Thixotropic gels, 326 children, 378
Threshold concentrations for skin damage, constipation, 380
148 neuropathic pain, 379
Tibolone, 360 Transdermal formulations, safety
Tilidine, 381 considerations, 511
Timolol, 157 Transdermal nicotine, 240
Timolol maleate, 385 Transdermal patches, 75
Tinea corporis, 54 in vitro release profiles, 410
Tinea cruris, 55 Transdermal product formulation, 337
Tinea incognito, 55 Transdermal scopolamine, 384
Tinea manuum, 55 Transdermal system manufacture, 343
Tinea pedis, 55, 430 Transdermal system matrices, 342
Tinea unguium, 55 Transdermal system testing, 343
Tissue deposition of dermally applied Transdermal therapeutic systems, reactions to
compounds, 163 drug delivery systems, 532
Tissue integrity, 211 Transepidermal water loss, 28, 206, 212,
Topical antibiotic therapy, 49, 430 228, 304, 530, 536
Topical antifungal agents, 429 Transfersomes, 307
Topical antipsoriatic treatment, 43 dexamethasone, 307
Topical antiviral efficacy, 160 hydrocortisone, 307
Topical bioavailability, 160 mode of action, 307
Topical bioequivalence, 160, 168 triamcinolone acetonide, 307
Topical corticosteroids, 46 Transfollicular penetration
Topical delivery dosage forms, 91 evidence, 172
Topical drug delivery effects of formulation increase, 176
excipients, 437 Transfollicular route, 169
Topical gene therapy, 308 Transforming growth factor-␣, 15
Topical local anesthesia, 431 Transforming growth factor-␤, 9, 15
Tretinoin, 49, 52
Topical mupirocin, 55
Triamcinolone, 305
Topical product half-life, 158
Triamcinolone acetonide, 92, 150, 280, 304,
Topical retinoids, 49
307, 385, 463
Topical steroids psoriasis, 44
Trichlorocarbanalide, 173
Topical tetracycline, 431
Trichocyte keratins, 177
Topical therapy psoriasis, 43
Trichophyton mentagrophyte, 430
Tortuosity factor, 83
Trichophyton rubrum, 55
Total mass transfer, 69
Triclosan, 147
Tragacanth gum, 325
Triethanolamine, 326
Tramadol, 381
Trilamellar laminates, 222
Transcellular pathway, 94 Triprolidine, 157
Transcutol, 144–145, 149, 278–280 Tristimulus reflectance colorimetry, 420
Transdermal delivery advantages, 90 Tritiated water permeability, 212
Transdermal delivery cutaneous reactions, Triton X-100, 202
529 Tromethamine, 326
Transdermal delivery disadvantages, 90 Tumor necrosis factor, 9
Transdermal delivery systems safety, 362 Type IV collagen, 7
Transdermal drug delivery system designs,
340 Ulcerative colitis, 351
Index 567
Ulcerative colitis Venous ulcers, 52

mesalamine, 352 Veralipride, 358
prednisolone, 352 Verification, bulk homogeneity, 508
Ulcers, 52 Very low permeability solutes, 113
Ultrasound, 280 Vesicle lipid components, diffusion in
Ultraviolet radiation, 52, 511, 519 stratum corneum, 307
Undermixing, 507 Vesicles, 304
United Kingdom National Health Service, mode of action, 306
347 nonionic surfactants, 306
United States Pharmacopeia, 409 Viable epidermal resistance, 131, 153
Unstirred diffusion layers, 78 Viable epidermis, 6, 100
Urinary excretion rate, 158 Viable epidermis-vehicle partition coefficient,
Urticaria, 530 112
UVA-induced neutrophil infiltration, 423 Viral infections, 56, 427
UVB radiation, 43 Viral warts, 57
UV-induced erythema, 520 Viscoelastic gel networks, 331
Viscosity binding sites, 113
Vanilloids, 242 Vitamin A, 176
Vapor phase absorption, 199 Vitamin D derivatives, 44, 451
Variable application time, 169 Vitiligo, 50
Variable boundary conditions, 84 Volpo N20, 202, 242
Variation in distribution, 470 Vulvovaginitis, 54
Variation in dose-effect relationship, 452,
Wagner-Nelson model, 158
462
Wart viruses, 45
Variation in excretion, 470
Water content of the stratum corneum, 129
Variation in metabolism, 470
Water enhancement of permeability, 140
Variations in levels of nicotine after
Water loss, 6
transdermal delivery, 347
Water permeability, 212, 228
Varicose eczema, 44, 52
Water-in-oil creams, 327
Vasoactive agents, 161
Water-removable base, 322
Vasoconstriction, 131, 149, 161, 168, 228,
Water-soluble base, 322
240, 305, 423, 443, 454, 464
Webril pads, 515
Vasoconstrictor assay, 160
White soft paraffin, 322
Vasodilation, 247 Whitfield’s ointment, 54
Vehicle effect on bioavailability, 436 Wool keratin, 224
Vehicle evaporation, 138
Vehicle metamorphosis, 447 Xanomeline, 384
Vehicle partition coefficient, 441 Xanthan gum, 325
Vehicle structural matrix, 436 X-ray analysis of the stratum corneum, 97
Vehicles, drug release, 138
Vehicles, effect on the stratum corneum, 143 Yeast infection, 54
Vehicles-skin interactions, 141 Yellow soft paraffin, 322

Dermatological and Transdermal Formulations

Uploaded by

Copyright:

Available Formats

Dermatological and Transdermal Formulations

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Dermatological and Transdermal Formulations

Uploaded by

Copyright:

Available Formats

ISBN: 0-8247-9889-9

This book is printed on acid-free paper.

World Wide Web

Copyright 䉷 2002 by Marcel Dekker, Inc. All Rights Reserved.

Current printing (last digit):

PRINTED IN THE UNITED STATES OF AMERICA

9. Chapter 8 covers bioequivalence of dermal and transdermal systems. Safety con-

1. The Structure and Function of Skin 1

2. Common Skin Disorders and Their Topical Treatment 41

3. Basic Mathematical Principles in Skin Permeation 61

5. Methods for Studying Percutaneous Absorption 197

6. Formulation Strategies for Modulating Skin Permeation 271

7. Dermatological Formulation and Transdermal Systems 319

8. Bioavailability and Bioequivalence of Dermatological

9. Scale-up of Dermatological Dosage Forms: A Case for

10. Safety Considerations for Dermal and Transdermal Formulations 511

11. Transdermal Delivery and Cutaneous Reactions 529

Keith R. Brain, Ph.D. An-eX Analytical Services Ltd., Cardiff, Wales

Sheree Elizabeth Cross, Ph.D. Department of Medicine, University of Queens-

Adrian F. Davis, Ph.D. GlaxoSmithKline Consumer Healthcare, Weybridge, Kent,

Bruce A. Firestone, Ph.D. Allergan, Inc., Irvine, California

Robert J. Gyurik MacroChem Corporation, Lexington, Massachusetts

Jonathan Hadgraft, D.Sc., F.R.S.C. Medway Sciences, University of Greenwich,

Howard I. Maibach, M.D. Department of Dermatology, University of California

Orest Olejnik, Ph.D., M.R.Pharm.S. Allergan, Inc., Irvine, California

Anthony D. Pearse, M.Sc., M.I.Biol., C.Biol., F.I.Sc.T. Department of Dermatol-

Mark A. Pellett, Ph.D., M.R.Pharm.S. Whitehall International, Havant, England

Michael S. Roberts, Ph.D., D.Sc. Department of Medicine, University of Queens-

Jagdish Singh Department of Pharmaceutical Sciences, College of Pharmacy,

Christian Surber, Ph.D., Priv.-Doz. Dr. Institute of Hospital-Pharmacy, University

Kenneth A. Walters, F.I.Biol., Ph.D. An-eX Analytical Services Ltd., Cardiff,

Adam C. Watkinson, Ph.D., M.B.A.* An-eX Analytical Services Ltd., Cardiff,

*Current affiliation: Strakan Pharmaceuticals Ltd., Galashiels, Scotland.

Figure 1 Components of the epidermis and dermis of human skin.

II. GROSS STRUCTURE AND FUNCTION OF THE SKIN

Figure 3 Skin components and their function.

Figure 5 Epidermal cell cohesion and communication is provided by desmosomes and

Figure 6 Routes of synthesis of the melanins: tyrosine is converted to dihydroxyphenyl-

The lymphatic system is an important component of the skin in regulating its

Fractional area 2.7 ⫻ 103 10⫺4 Variable —

Secretions Sebum Sweat (dilute ‘‘Milk’’ protein, lipoproteins, Nil

Biochemical innervation — Cholinergic Cholinergic (?) —

Control Hormonal Sympathetic Sympathetic nerves —

Table 2 Lipid Composition of Human Sebum

Lipid Lumen of glanda Skin surfacea

III. DEVELOPMENT OF THE STRATUM CORNEUM

Table 3 Amino Acid Composition of Nail, Hair, and Stratum Corneum

Amino acid Nail Hair Stratum corneum

Lysine 3.1a 2.5a 4.2a

Figure 7 Cytokine involvement in keratinocyte proliferation. Secretion of interleukin-1

B. Cornified Cell Envelope

C. Stratum Corneum Proteins

appears in keratohyalin granules in the stratum granulosum. The profilaggrin mole-

D. The Intercellular Lipids

Table 4 Lipid Content of the Stratum Corneum

Lipid % (w/w) mol %

Cholesterol esters 10.0 7.5a

and 4-hydroxysphinganine, N-acetylated by different fatty acids. Because of its

Figure 10 Ceramides of the human stratum corneum intercellular space.