Review

CNS pharmacokinetics
of antifungal agents
Shravan Kethireddy & David Andes†
University of Wisconsin School of Medicine and Public Health, Madison 600 Highland Ave, H4/572,
1. Fungal infections and CNS
d
te
Madison, WI 53792, USA

The goal in treatment of infections is to achieve a beneficial effect whilei

involvement
2. Factors impacting CNS drug
minimizing toxicity. It is widely recognized that the principles of pharma-
h ib
o
penetration and accumulation
3. Antifungal CNS
cokinetics and pharmacodynamics are critical to determining an adequate
r
p of the
y
dose–response relationship. There has been an increased involvement
CNS to infection from opportunistic and endemic fungi over lthe
pharmacokinetic and treatment
t last several

ric it has become

investigations
decades due to establishment of solid-organ and bone marrow transplanta-
tion as well as immunosuppression from HIV. In this tregard
4. Amphotericin B formulations

critical to define optimal dosing regimens by ansunderstanding of the

nagent to the targeted CNS
5. Flucytosine
6. Triazoles processes which govern delivery of an antifungal
i o
t is to: i) summarize published
7. Echinocandins
b u
site of involvement. The objective of this review
8. Expert opinion
t r i
experimental and clinical antifungal pharmacokinetics; and ii) examine the
relationship between CNS antifungal
i s pharmacokinetics and efficacy.

d
Examination of these studies reveal marked variability among antifungal

drelationship between CNS antifungal pharma-

drugs with regard to cerebrospinal fluid and brain parenchymal penetration.
n
cokinetics and efficacy a
Formal examination of the

suggest that brainn g are limited. The few experimental studies available

ti
parenchymal kinetics is a superior predictor of antifungal

i n
efficacy than cerebrospinal fluid concentrations.

Pr , central nervous system, cerebrospinal fluid, pharmacokinetics

Keywords: antifungal

Expert Opin. Drug Metab. Toxicol. (2007) 3(4):573-581

t d. 1. Fungal infections and CNS involvement

L
UK
Outcome following invasive fungal infections is generally poor. When these

a
pathogens breach the CNS, treatment success becomes even more dismal. The

rm
ability of antifungal drugs to achieve adequate concentrations in the tissues of the

fo
CNS is one of numerous factors that impact treatment outcome of these infections.

n
All systemic fungal pathogens have been associated with CNS involvement. Those

fI
which most commonly disseminate to the CNS include Cryptococcus neoformans,

t o Candida spp. and Aspergillus spp. (Table 1). One common factor to consider when

h
designing an optimal antimicrobial therapy is where within an end organ does the

r ig pathogen reside? For example, common bacterial pathogens, such as Streptococcus

py pneumoniae, primarily produce disease of the meninges. Similar considerations for

fungi are difficult since these pathogens do not respect the tissue compartments

C o within the CNS. Histopathologic studies have demonstrated organism invasion of

the meninges and parenchymal tissues with all fungal pathogens [1].

2. Factors impacting CNS drug penetration and accumulation

The pharmacologic goal of antimicrobial administration is to achieve adequate
concentrations at the site of infection. For most tissue sites, serum drug concentra-
tions provide an excellent surrogate of the interstitial tissue concentration where
pathogens reside for most end organs. The tissues of the CNS and eye, however, are
relatively protected tissue sites due to blood tissue barriers that limit diffusion of
many molecules. Furthermore, there are tissue efflux pumps that can impact CNS

10.1517/17425225.3.4.573 © 2007 Informa UK Ltd ISSN 1742-5225 573

CNS pharmacokinetics of antifungal agents

Table 1. Incidence of CNS involvement associated with drug from a MW of 250 – 400 Da can decrease permeation by
invasive fungal infection. 100-fold [4,5]. The upper limit of MW to allow for efficient
diffusion is 300 – 400 Da [4]. Among the available antifungal
Organism CNS involvement Mortality agents, one would expect that a small molecule such as flucy-
tosine (120 g/mol) would diffuse easily due to its
Invasive candidiasis 3 – 64% 11 – 67%
small MW (Table 2). Conversely, the large MW of the cyclic
Invasive aspergillosis 4 – 6% 80 – 90% hexapeptide echinocandin molecules would be anticipated to
Cryptococosis 67 – 84% < 1% limit penetration of drugs such as caspofungin, anidulafungin
Histoplasmosis* 5 – 20% 20 – 40% and micafungin. The MW of each of the echinocandins
exceeds 1000 g/mol. Polyene and triazole molecules are of
Coccidioidomycosis* 25% 26%
intermediate-to-large size. Although the MW of the parent
Blastomycosis‡ 40% 4.3 – 22% amphotericin B molecule is identical among the four prepara-
Zygomycosis 12% 79 – 98% tions (amphotericin B deoxycholate, liposomal amphotericin
Dematiacious (cladophialophora) 100% 71 – 74% B, amphotericin B lipid complex and amphotericin B colloidal
dispersion), the particle size of the complex varies markedly.
Data taken from [37,49,63-66]. Among the lipid preparations, the liposomal product is
*Disseminated disease.

‡Patients with HIV/AIDS.

two orders of magnitude smaller than other lipid associated
preparations. Among the triazoles, fluconazole and voricon-
azole would appear to have an advantage over either
accumulation for many drugs. In order for drugs to enter the itraconazole or posaconazole for CNS diffusion due to their
cerebrospinal fluid (CSF) and parenchyma of the brain, the lower MW (Table 2).
compounds need to traverse either the epithelial layer of A second drug property that impacts penetration into the
the choroid plexus or the cerebral endothelium. This tissue CNS is binding to plasma protein. The large size of serum
barrier is structurally different from other blood tissue barriers proteins, such as albumin, precludes penetration of protein
due to the size of the intercellular connections through which bound molecules. In the absence of inflammation, there is
drugs pass. All other tissue sites possess fenestrated cell con- very little albumin in the CSF (CSF:serum albumin ratio of
nections for which the space between cells is 100 Å, allowing 1:200). Plasma protein binding varies widely among the
ready diffusion of most pharmacologic molecules. However, to antifungal agents. The amphotericin B preparations and each of
gain access to the CNS, drugs must traverse tight junctions the echinocandins exhibit binding to albumin in excess of 95%.
which are much smaller (20 Å) and preclude diffusion of large Conversely, protein binding to flucytosine is negligible. Among
molecular weight (MW) drugs. The blood–brain barrier is one the triazole compounds, fluconazole is least impacted by
of these tissue barriers and is composed of the layer of protein binding (10%), voriconazole is intermediate (58%),
endothelium from the vessels surrounding the brain and spinal and both itraconazole and posaconazole exhibit a very high
cord. The endothelial cells are linked by tight junctions with degree of binding (> 98%). The lower protein bound drugs
only 0.02% of the capillaries possessing fenestrations [2]. The would be expected to penetrate more readily than those bound
choroid plexus makes up a second drug barrier termed the to a higher degree.
blood–CSF barrier [9]. Although the capillary membrane at A third drug property that governs permeation across the
the choroid plexus is mostly fenestrated, the ependymal cell CNS tissue barrier is molecule lipophilicity. Capillaries of the
layer of the plexus which contacts the CSF is constituted by brain are devoid of aqueous pores to facilitate aqueous
tight junctions that similarly limit large molecule transport [3]. diffusion, thus lipid diffusion becomes a critical determinant
Once a compound traverses the CNS tissue barriers, there of drug penetration. The permeability coefficient for lipid
are efflux pumps (P-glycoprotein [P-gp]) in the choroid diffusion is known as the lipid:aqueous partition or the
plexus that can impact the ability of drugs to accumulate in octanol:water partition (Log P) [6,7]. Drugs which readily
the CNS. enter the CNS compartments often possess an octanol:water
There are several physiochemical properties that impact the partition coefficient of ∼ 1. However, highly lipophilic
ability of drugs to traverse the CNS tissue barrier. These compounds also frequently demonstrate a high level of bind-
factors include: compound i) molecular size; ii) lipophilicity; ing to serum proteins. Thus, drug properties of lipophilicity
iii) plasma protein binding; iv) efflux pump affinity; and protein binding can be conflicting characteristics in regard
v) molecular charge; and vi) cerebral blood flow. The impact of to CNS penetration. Among the antifungal drugs, the two
molecular size is intuitive, the larger the molecule the less it triazole compounds, itraconazole and posaconazole, possess
would be expected to be able to traverse the tight junctions of high lipophilicity. However, as discussed, this physiochemical
the CNS tissue barrier. The diffusion coefficient is an estimate characteristic is associated with high affinity for serum
of the ability of a compound to penetrate the CNS serum albumin. The characteristics of protein binding and
barrier and is approximately proportional to the reciprocal of lipophilicity conflict in regard to CNS penetration, making
the square root of the molecular mass. Doubling the size of a pharmacokinetic predictions difficult.

574 Expert Opin. Drug Metab. Toxicol. (2007) 3(4)

 Kethireddy & Andes

Table 2. Physiochemical properties of antifungal agents.

Molecular weight Log P % Protein binding to albumin P-glycoprotein substrate

Fluconazole 309 2.17 10 +

Itraconazole 705 6.99 98 ++
Voriconazole 349 2.56 58 +
Posaconazole 700 6.1 99 +
5-FC 120 -0.89 5 -
AmB 924 (*< 0.4 µm) 0.95 > 95 -
ABLC* 1.6 – 11 um 0.95 > 95 -
L-AmB* 0.08 um 0.95 > 95 -
ABCD 0.12 – 0.14 µm
Caspofungin 1093 -2.8 98 -
Anidulafungin 1140 0.21 98 -
Micafungin 1291 -3.8 98 -

Data taken from [6,67,68].

*Particle size.

5-FC: 5-Flurocytosine; ABCD: AmB colloidal dispersion; ABLC: AmB lipid complex; AmB: Amphotericin B; L-AmB: Liposomal AmB.

Once within the CNS space, the drug must accumulate to CNS inflammation, often associated with an infectious process.
achieve therapeutic concentrations. Efflux pumps at the CNS Inflammation can disrupt the CNS blood barrier tight junctions
tissue barrier are capable of removing molecules from the CSF. and enhance the ability of pharmaceutical agents to penetrate
Substrates of P-gp, a membrane bound P-ATPase efflux pump, the CNS. Yet, the degree of inflammation is difficult to mea-
has a high affinity for lipophilic molecules. These transporters sure and accurately reproduce from study to study. Finally, it is
protect the brain from toxic xenobiotics and also decrease critical to consider the manner in which CNS kinetic data are
entry of some therapeutic drugs into the CNS. The only anti- presented. Most studies report CSF concentrations relative to
fungal compounds that serve as substrates for P-gp are drugs serum and include an estimate of penetration. It is not uncom-
from the azole class. Among the triazole drugs, itraconazole mon to find reports based on a single and simultaneous mea-
exhibits the most significant affinity for this protein [8]. surement of serum and the CSF, especially in human studies.
The understanding of these basic principles of CNS It is important to recognize the amount of time necessary for
pharmacokinetics should allow prediction of antifungal equilibration of drug in the CSF. The time to peak serum
penetration and accumulation. However, the complex and concentrations is much earlier than the time to peak CSF
often conflicting nature of several of these physiochemical concentration. With these caveats in mind, the most valid
properties can make these predictions difficult. The studies parameters of drug entry into the CSF include: i) CSF:serum
examining the reliability of these predictions for CNS concentration ratio at steady-state; and ii) CSF: serum ratio of
pharmacokinetics and treatment efficacy of antifungal drugs the area under the concentration–time curve.
are considered herein. There are several study types that allow one to understand
and predict the likelihood of adequate antifungal CNS
3. Antifungal CNS pharmacokinetic and pharmacokinetics for treatment of CNS fungal infections.
treatment investigations These investigations include both animal models and human
trials (Tables 3 and 4). The study designs include: i) measure-
Interpretation of pharmacokinetic studies in the CNS requires ment of antifungal drug concentrations in CSF or brain paren-
consideration of several important experimental variables. chyma, with or without CNS inflammation; ii) measurement
First among these is the realization that the CNS is not of CNS concentration and correlation with antifungal treat-
a homogenous pharmacologic compartment. More simply ment effect; and iii) examination of antifungal treatment effect
put, CSF concentrations do not necessarily predict brain without CNS drug concentration measurements. Several assay
parenchymal or meningeal concentrations. In fact, there is methodologies have been used measure of CNS antifungal con-
even a lack of rapid equilibration within different CSF centrations including microdialysis, postmortem tissue samples,
locations such as the lumbar cistern and ventricles. As a positron emission tomography (PET) imaging and, most
general rule, brain tissue concentrations are most often greater commonly, CSF sampling. Unfortunately, there is significant
than those detected in the CSF. A second study variable that heterogeneity in study design for antifungal CNS kinetics
markedly impacts kinetic measurements is the presence of which makes comparison across studies and drugs somewhat

Expert Opin. Drug Metab. Toxicol. (2007) 3(4) 575

CNS pharmacokinetics of antifungal agents

Table 3. CNS pharmacokinetics of antifungal agents in animal models.

CSF concentration Brain concentration Ref.

Fluconazole 42 – 84% (75%) 50 – 100% [24,26,30,34,35,69]

Itraconazole < 1% (< 1%) Hydroxy itraconazole found in one study [30,34,35]

Voriconazole 68 – 100% Detectable in guinea-pig brain [39]

Posaconazole NA NA
5-FC NA NA
AmB < 1% 3% (18%) [9,10,17]

ABLC < 1% (27%) [10]

ABCD < 1% (22%) [10]

L-AmB < 1% (3%) [10]

Caspofungin 0 10 – 20% [50]

Anidulafungin 0 9 – 15% [51]

Micafungin 0 8 – 18% [58,70]

Value in parentheses represents study in presence of CNS infection.

5-FC: 5-Flurocytosine; ABCD: AmB colloidal dispersion; ABLC: AmB lipid complex; AmB: Amphotericin B; CSF: Cerebrospinal fluid; L-AmB: Liposomal AmB;
NA: Not applicable.

difficult. The available literature for each of the antifungal a strong correlation between the maximal (peak concentration)
drugs from animal models is listed in Table 3. and total (AUC) AmB concentration in the brain tissue and
Candida burden in brain tissue at the end of therapy. This
4. Amphotericin B formulations concentration-dependent pharmacodynamic relationship is
similar to that observed for these antifungals in other infection
The CNS pharmacokinetics of amphotericin B (AmB) deoxy- sites. Human CNS pharmacokinetic evaluation of AmB and
cholate and the lipid-associated preparations have been examined ABLC are limited to a report of CSF concentrations in a small
in experimental studies using rabbit models both with and series of patients with CNS infections. Similar to animal model
without CNS infection [9,10]. Systemic administration of each studies, concentrations in CSF were either undetectable or
of the formulations did not produce measurable concentra- very low relative to serum AmB levels. Limited CSF accumula-
tions in the CSF regardless of CNS inflammation due to tion of amphotericin B led to the development of intrathecal
candida or cryptococcal meningitis. However, detectable brain therapy for several CNS fungal infections [11,12]. However,
parenchymal concentrations were observed even in the absence despite the lack of measurable CSF amphotericin B concentra-
of CNS infection. Brain concentrations in non-infected tions, there is a large clinical experience of successful use of
animals ranged 3 – 27% of those observed in rabbit serum. these products for treatment of CNS fungal infections [13-15].
The penetration of these compounds was enhanced in pres- In fact, amphotericin B remains the standard of treatment for
ence of infection two- to fourfold. The tissue concentrations certain CNS infections, such as cryptococcal meningitis [16].
observed with common dosing regimens would be expected to Similarly, there are numerous animal model treatment studies
effectively inhibit the growth of or kill invading fungi. The demonstrating efficacy with these polyene drugs [9,10,17-22].
degree of penetration in a Candida meningitis model was Taken together, these data suggest a poor correlation between
similar among AmB, AmB lipid complex (ABLC) and AmB CSF concentration and efficacy in treatment of CNS
AmB colloidal dispersion (ABCD). The penetration of LAmB fungal infections. The experimental data from Groll et al.
relative to serum concentrations was lower than the other prep- suggest that the ability of these drugs to achieve adequate brain
arations. However, the kinetics of liposomal AmB (L-AmB) parenchymal concentrations may better correlate with
are characterized by very high serum concentrations which are treatment efficacy than CSF measurements.
more than 30-fold greater than each of the other amphotericin
B formulations. Thus, the absolute concentrations of L-AmB 5. Flucytosine
in the brains of rabbits in this model were significantly higher
than the other preparation (L-AmB brain tissue concentrations Both animal and human studies have examined the CNS
3.6- to 5.2-fold greater than AmB, ABLC, ABCD). Interestingly pharmacokinetics of flucytosine. Bennett and associates accessed
these differences in brain parenchymal concentrations favoring serum and CSF concentrations in animals and humans with
the L-AmB formulation were associated with enhanced CNS mycoses. Using a bioassay to measure drug concentrations
therapeutic efficacy in this model. The investigators observed they found that CSF concentrations in humans ranged

576 Expert Opin. Drug Metab. Toxicol. (2007) 3(4)

 Kethireddy & Andes

17 – 62 µg/ml and were ∼ 74% of simultaneously determined report nearly undetectable concentrations [30,34-36]. These
serum concentrations [23]. These concentrations far exceed those findings would not be terribly surprising given the larger MW
associated with growth inhibition in vitro and would be antic- and high protein binding. However, study of brain parenchymal
ipated to be adequate for efficacy. However, studies examining concentrations demonstrate that itraconazole does accumulate
the relationship between CSF concentrations and therapeutic in this tissue space. Haynes et al. found brain parenchymal
effect are unavailable. concentrations of the microbiologically active itraconazole
metabolite (hydroxyl-itraconazole) approaching those measured
6. Triazoles in the serum of mice with Histoplasma meningitis [18].
Several CNS fungal infection models have undertaken
The available triazole antifungal agents (fluconazole, comparative CSF kinetic and efficacy studies with fluconazole
voriconazole, itraconazole and posaconazole) exhibit variable and itraconazole [30,34-36]. The kinetic results from these inves-
physiochemical characteristics and, not surprisingly, differ in tigations confirm those discussed above for which fluconazole
CNS pharmacokinetics. The pharmacokinetics of each of would appear to have an advantage based on CSF concentrations.
these compounds in the CSF and brain parenchyma has been However, despite the lack of appreciable CSF concentrations
extensively examined in animal models both with and without of itraconazole, therapeutic efficacy is similar to that observed
CNS infection. Among the triazoles, fluconazole achieves the for fluconazole in cryptococcal, Coccidioides and Histoplasma
highest concentrations in the CSF. Even in models with an intact meningitis models. For example, Sorensen and colleagues
CNS blood barrier, concentrations were in the range of 42 – 84% measured serum and CSF concentrations of fluconazole and
of those observed in serum. For example, Madu et al. itraconazole in a rabbit Coccidioides meningitis model [30].
found that CSF concentrations of fluconazole were 84.3% of Itraconazole CSF concentrations were undetectable; however,
those observed in serum following intravenous administration fluconazole concentrations exceeded the Coccidiodes spp.
of drug to healthy rabbits [24]. Achievable concentrations are MIC90. Despite this marked difference in CSF concentrations,
even greater in the presence of CNS infection. Several human there was no difference in the ability of either treatment to
CSF kinetics studies with fluconazole report similar findings. reduce the burden of organisms in the spinal cord and brain of
For example, Foulds et al. examined the fluconazole concen- these mice. It is hypothesized that the activity of itraconazole
tration in serum and CSF in patients without meningeal in this model is due to a combination of higher concentrations
inflammation [25]. CSF concentrations were in the range of in the target tissues of the brain parenchyma and the lower
52 – 62% of those in serum at steady-state. Study of brain tis- MIC values for this drug organism combination [37]. An ani-
sue fluconazole kinetics demonstrated a concentration profile mal model study of Histoplasma CNS infection also supports
that closely approximates that observed in the CSF. Using this theory [18]. Haynes et al. compared the CNS kinetics and
intracerebral microdialysis in rats, Yang et al. found that efficacy of several antifungals in a murine model. They simi-
fluconazole rapidly reaches equilibrium between the plasma larly found undetectable itraconazole concentrations in the
and brain extracellular fluid with an average brain distribution CSF but parenchymal concentrations that are higher than the
coefficient of 0.60 [26]. Thaler and colleagues examined MIC of the infecting organisms. Similar to other groups com-
fluconazole penetration by HPLC analysis in non-inflammed paring the efficacy of fluconazole and itraconazole in these
human cerebral tissue and found an average brain/plasma ratio CNS infection models, they reported no difference in efficacy
of 1.33, indicating nearly complete equilibration with in this CNS endemic fungal infection model. More impor-
serum [27]. PET scan imaging of 18F-labeled fluconazole in tantly, most comparable clinical trials have also not demon-
healthy human volunteers revealed relatively homogenous dis- strated differences in treatment efficacy for CNS fungal
tribution of the drug throughout the brain with calculated infections. Two cryptococcal meningitis trials comparing the
values in the range of 4.15 – 5.48 µg/ml [28]. In a murine his- efficacy of fluconazole and itraconazole for consolidation
toplasmosis meningitis model, Haynes et al. reported brain therapy have observed therapeutic equivalence [16,38]. However,
parenchymal fluconazole concentrations as high as 13.85 µg/ml these patient trials have not included CNS pharmacokinetic
following dose levels relevant to those used in patients [18]. evaluation. Taken together, these animal models and clinical
These concentrations are several-fold higher than that needed to trials point to discordance between triazole CSF concentrations
inhibit growth (minimum inhibitory concentration [MIC90]) and treatment efficacy similar to that observed for AmB.
of Histoplasma spp. [29]. The ability of fluconazole to achieve Although brain parenchymal concentrations were not examined
concentrations in the various CNS compartments greater than in the animal model treatment studies, one certainly wonders
the MIC90 of common fungal pathogens would suggest it if the ability of itraconazole to accumulate in this tissue
would be effective in therapy of these infections. Indeed, mul- accounts for efficacy in these models.
tiple experimental and clinical trial studies with fluconazole The newer triazole voriconazole is structurally related to
for treatment of susceptible Candida, Cryptococcal and endemic fluconazole, whereas posaconazole is similar in structure to
fungi demonstrate the successful use of this triazole [22,30-33]. itraconazole. Lutsar and colleagues studied voriconazole con-
The kinetics of itraconazole has been similarly investigated. centrations in the CSF and brain in uninfected guinea-pigs
Both animal models and human studies of CSF pharmacokinetics and in the CSF of humans with a wide range of CNS

Expert Opin. Drug Metab. Toxicol. (2007) 3(4) 577

CNS pharmacokinetics of antifungal agents

Table 4. CNS pharmacokinetics of antifungal agents in For example, Imai et al. found posaconazole therapy of CNS
humans. aspergillosis in neutropenic mice to be equivalent to AmB at
reducing brain parenchymal organism burden [46]. It will be
CSF Brain Ref.
important for future studies to include kinetic investigation in
concentration concentration
the CNS. Given the structural similarities to itraconazole, one
Fluconazole 52 – 82% 116% [18,25,31,71] would anticipate low CSF concentrations, but significant
(70 – 89%) accumulation in the brain parenchyma. Few anecdotal reports
Itraconazole (< 10%) NA [42] detail successful use of posaconazole for CNS fungal infections
[39, 40,
in humans [49].
Voriconazole (38 – 68%) 1.2 – 1.9 µg/g
42-44]

Posaconazole NA NA 7. Echinocandins
5-FC 74% NA [23]
The CNS pharmacokinetics of each of the three available
AmB (0 – 4%) NA [72] echinocandin compounds, caspofungin, micafungin and
ABLC (< 1%) NA [13] anidulafungin, has been studied in detail using a non-infected
ABCD NA NA rabbit model [50-52]. In each instance, the investigators have
reported undetectable CSF concentrations even using dose lev-
L-AmB NA NA
els far exceeding those used in current clinical regimens. However,
Caspofungin NA NA these same investigations found brain parenchymal concentrations
Anidulafungin NA NA in the range of 10 – 20% of those measured in serum. For
example, Groll et al. performed extensive tissue distribution
Micafungin NA NA
studies with anidulafungin in healthy rabbits [51]. Although a CSF
Value in parentheses represents study in presence of CNS infection. assay for anidulafungin did not identify a measurable amount
5-FC: 5-Flurocytosine; ABCD: AmB colloidal dispersion; ABLC: AmB lipid of drug, brain parenchymal concentrations ranged 0.24 – 3.9 µg/g
complex; AmB: Amphotericin B; CSF: Cerebrospinal fluid; L-AmB: Liposomal over a dose range of 0.5 – 10 mg/kg [51]. These tissue concentra-
AmB; NA: Not applicable. tions exceed the MIC90 of fungal pathogens in the echinocandin
spectrum and would be anticipated to be sufficient for treat-
fungal disease. In animals without CNS infection, CSF con- ment success. This same research group has undertaken similar
centrations were very similar to those observed in serum. investigation of the CNS kinetics of caspofungin and
Among a group of 14 patients with invasive fungal infections, micafungin and report nearly identical findings [50,52].
CSF concentrations of voriconazole were nearly half (46%) of Numerous animal model studies have examined the efficacy
those reported concomitantly in serum [39]. Brain tissue levels of each of the echinocandin drugs in CNS fungal infection
of voriconazole have also been reported in a patient with an models [53-57]. Comparative studies using these models have
intracerebral aspergillosis [40]. Drug concentrations in biopsies examined echinocandin efficacy relative to other antifungal
ranged 1.2 – 1.9 µg/g, indicating that voriconazole also accu- drugs and have demonstrated therapeutic equivalence. For
mulates in the brain parenchyma. These concentrations are example, Imai and colleagues compared the efficacy of con-
above or near the MIC90 for most target fungal pathogens [41]. ventional amphotericin, ABLC and caspofungin in a murine
Other anecdotal publications also include measurement of model of CNS aspergillosis [53]. Caspofungin prolonged
voriconazole concentration in CSF or brain tissue and report survival in > 80% of infected animals when compared to
similar findings. In humans treated for CNS aspergillosis, CSF controls [53]. Similar observations were made by Ibrahim and
concentrations have been measured in 38 – 68% of those colleagues in a murine model of zygomycosis [54]. Singh et al.
observed in plasma [42-44]. Although clinical experience with and Hussain et al. both found clinical efficacy of caspofungin
voriconazole in treatment of CNS fungal infections is small, in murine models of CNS aspergillosis and azole resistant
the limited reports suggest efficacy [41]. candida meningitis, respectively [55,56]. Each of these studies
The most recently approved triazole, posaconazole, has not did not include pharmacokinetic assays. However, in a
been evaluated as extensively with regard to CNS efficacy and recent study by Hope et al., micafungin was evaluated in a
no published or presented CNS kinetic studies are available at rabbit model of Candida meningoencephalitis. in which detailed
this time. However, the in vivo activity of posaconazole has kinetic evaluation was incorporated [58]. Similar to previous
been investigated in several animal models of CNS fungal kinetic evaluation of micafungin, CSF concentrations were
infections with Aspergillus, Cryptococcus and Phaeohyphomycosis. undetectable. However, parenchymal and meningeal concen-
In these studies, posaconazole was administered using clini- trations of micafungin exceeded the MIC90 of the infecting
cally relevant dosing regimens and compared with either AmB organisms. In fact, the micafungin concentrations in the
or another triazole. Without exception, outcome following meningeal tissues were 10-fold greater than the brain paren-
posaconazole therapy as measured by survival or fungal CNS chyma. The relationship between parenchymal micafungin
burden was equivalent or superior to the other therapies [45-48]. concentrations and reduction in CNS Candida burden in

578 Expert Opin. Drug Metab. Toxicol. (2007) 3(4)

 Kethireddy & Andes

these experiments were strong. Furthermore, the tissue However, each of the available drugs has been demonstrated to
concentrations in these infected animals were nearly 30% accumulate in the brain parenchyma at concentrations
greater than those previously observed in models without CNS exceeding the MIC90 values of most infecting pathogens.
inflammation [58]. Human CNS kinetic data with these Fewer clinical studies are available to corroborate these data.
compounds are not available. However, there are anecdotes Investigation of the relationship between CNS antifungal
and small case series suggesting clinical efficacy of caspofungin concentrations and outcome are even more uncommon.
for patients with CNS fungal infection [59-62]. However, the available data suggest a poor relationship between
antifungal CSF concentration and outcome. Conversely,
8. Expert opinion efficacy in these studies has been reasonably correlated with
brain parenchymal antifungal penetration. Additional studies
Understanding the pharmacokinetics of antimicrobial defining the relationship between antifungal concentrations in
compounds at the site of infection is important to optimize the CNS and treatment efficacy will be critical for optimal
drug choice and dosing regimen design. The CNS represents a therapy of these increasingly common infections.
critical tissue site due to poor outcome of fungal infection at
this site and variable kinetics due to the CNS blood barrier. Conflict of interest
The CSF kinetics of all available antifungals has been
characterized in the CSF and brain parenchyma in animal D Andes has received research grants from and is an adviser
models. Kinetics of these drugs differs markedly in the CSF. and speaker for Pfizer, Merck, Astellas, Schering, Elan.

Bibliography Infect. Dis. Clin. North Am. (1999) Antimicrob. Agents Chemother. (2005)
Papers of special note have been highlighted 13(3):595-611. 49(12):5092-5098.
as either of interest (•) or of considerable 8. IMBERT F, JARDIN M, 14. GALLIS HA, DREW RH,
interest (••) to readers. FERNANDEZ C: Effect of efflux PICKARD WW: Amphotericin
1. ZHANG J-R, TUOMANEN E: Molecular inhibition on brain uptake of itraconazole B: 30 years of clinical experience.
and cellular mechanisms for microbial in mice infected with Cryptococcus Rev. Infect. Dis. (1990) 12(2):308-329.
entry into the CNS. J. Neurovirol. (1999) neoformans. Drug Metab. Dispos. (2002) 15. WALSH TJ, HIEMENZ JW, SEIBEL NL:
5:591-603. 31:319-325. Amphotericin B lipid complex for invasive
2. NAU R, SORGEL F, PRANGE HW: 9. LEE JW, AMANTEA MA, FRANCIS PA fungal infections. Clin. Infect. Dis. (1998)
Pharmacokinetic optimisation of the et al.: Pharmacokinetics and safety of a 26:1383-1396.
treatment of bacterial central nervous unilamellar liposomal formulation of 16. VAN DER HORST CM, SAAG MS,
system infections. Clin. Pharmacokinet. amphotericin B in rabbits. Antimicrob. CLOUD GA: Treatment of Cryptococcal
(1998) 35(3):223-246. Agents Chemother. (1994) 38(4):713-718. meningitis associated with the acquired
• Comprehensive review of CNS •• Elegant study defining the impact of immunodeficiency syndrome. National
antimicrobial pharmacology. antifungal CNS pharmacokinetics in the Institute of Allergy and Infectious Diseases
brain parenchyma. Mycoses Study Group and AIDS Clinical
3. DE LANGE ECM DM: Consideration
in the use of cerebrspinal fluid 10. GROLL AH, GIRI N, PETRAITIS V Trials Group. N. Engl. J. Med. (1997)
pharmacokinetics to predict brain target et al.: Comparative efficacy and distribution 337(1):15-21.
concentrations in the clinical setting. of lipid formulations of amphotericin 17. PERFECT DD Jr: Comparison of
Clin. Pharmacokinet. (2002) B in experimental Candida albicans amphotericin B and N-D-ornithyl
41(10):691-703. infection of the central nervous system. amphotericin B methyl ester in
J. Infect. Dis. (2000) 182:274-282. experimental Crytococcal meningitis
4. PARDRIDGE: Blood–brain barrier drug
targeting: the future of brain drug 11. LABADIE EL, HAMILTON RH: and Candida albicans endocarditis with
development. Mol. Interven. (2003) Survival improvement in Coccidiodal pyelonephritis. Antimicrob. Agents
3(2):90-105. meningitis by high-dose intrathecal Chemother. (1985) 28(6):751-755.
amphotericin B. Arch. Intern. Med. (1986) 18. HAYNES RR, CONNOLLY PA,
5. FISCHER H, GOTTSCHLICH R,
146:2013-2018. DURKIN MM et al.: Antifungal therapy
SEELIG A: Molecular parameters
governing passive diffusion. 12. CLEMONS KV, SOBEL RA, for central nervous system histoplasmosis,
J. Membrane Biol. (1998) 165:201-211. WILLIAMS PL, STEVENS DA: using a newly developed intracranial model
Amphotericin B: 30 years of clinical of infection. J. Infect. Dis. (2002)
6. KATZUNG: Basic and Clinical
experience. Rev. Infect. Dis. (1990) 185:1830-1832.
Pharmacology. 9th Edition.
12(2):308-329. 19. CLEMONS KV, STEVENS DA:
Lange Medical Books/McGraw Hill,
New York, USA (2004). 13. WÜRTHWEIN G, GROLL AH, Comparative efficacies of four amphotericn
HEMPEL G et al.: Population B formulations against systemic murine
7. ANDES DR, CRAIG WA:
pharmacokinetics of amphotericin aspergillosis. Antimicrob. Agents Chemother.
Pharmacokinetics and pharmacodynamics
B lipid complex in neonates. (2004) 48(3):1047-1050.
of antibiotics in meningitis.

Expert Opin. Drug Metab. Toxicol. (2007) 3(4) 579

CNS pharmacokinetics of antifungal agents

20. CLEMONS KV, SOBEL RA, to fluconazole as a cause of failure during 37. LIPTON SA, HICKEY WF,
WILLIAMS PL, PAPPAGIANIS D, treatment of histoplasmosis in patients MORRIS JH, LOSCALZO J: Candidal
STEVENS DA: Efficacy of intravenous with acquired immunodeficiency disease infection in the central nervous system.
liposomal amphotericin B against syndrome. Clin. Infect. Dis. (2001) Am. J. Med. (1984) 76(1):101-108.
Coccidioidal meningitis in rabbits. 33:910-913. 38. SAAG MS, CLOUD GA, GRAYBILL JR
Antimicrob. Agents Chemother. (2002) 30. SORENSEN KN, SOBEL RA, et al.: A comparison of itraconazole versus
46(8):2420-2426. CLEMONS KV, PAPPAGIANIS D, fluconazole as maintainence therapy for
21. CLEMONS KV, ESPIRITU M, STEVENS DA, WILLIAMS PL: AIDS-associated cryptococcal meningitis.
PARMAR R, STEVENS DA: Comparative Comparison of fluconazole and Clin. Infect. Dis. (1999) 28:291-296.
efficacies of conventional amphotericn B, itraconazole in a rabbit model of 39. LUTSAR R RS, TROKE P: Voriconazole
liposomal amphotericn B, caspofungin, Coccidioidal meningitis. Antimicrob. Agents concentrations in the cerebrospinal fluid
micafungin, and voriconazole alone and Chemother. (2000) 44(6):1512-1517. and brain tissue of guinea pigs and
in combination against experimental • Important comparison of efficacy of immunocompromised patients.
murine central nervous system aspergillosis. antifungal drugs with varying CSF Clin. Infect. Dis. (2003) 37:728-732.
Antimicrob. Agents Chemother. (2005) pharmacokinetics.
40. ELTER T, SIENIAWSKI M,
49(12):4867-4875. 31. TUCKER RM, WILLIAMS PL, GOSSMANN A et al.: Voriconazole
22. JAFARI HS, SÁEZ-LLORENS X, ARATHOON EG et al.: Pharmacokinetics brain tissue levels in rhinocerebral
SEVERIEN C et al.: Effects of antifungal of fluconazole in cerebrospinal fluid and aspergillosis in a successfully treated young
therapy on inflammation, sterilization, and serum in human Coccidioidal meningitis. woman. Intern. J. Antimicrob. Agents (2006)
histology in experimental Candida albicans Antimicrob. Agents Chemother. (1988) 28(3):262-265.
meningitis. Antimicrob. Agents Chemother. 32(3):369-373.
41. JOHNSON LB, KAUFFMAN CA:
(1994) 38(1):83-89. 32. SORENSEN KN, SOBEL RA, Voriconazole: a new triazole antifungal
23. BLOCK ER, BENNETT JE: CLEMONS KV et al.: Comparative agent. Clin. Infect. Dis. (2003)
Pharmacological studies with efficacies of terbinafine and fluconazole 36:630-637.
5-fluorocytosine. Antimicrob. Agents in treatment of experimental Coccidioidal
42. VERWEIJ PE, BRINKMAN K,
Chemother. (1972) 1(6):476-482. meningitis in a rabbit model. Antimicrob.
KREMER HP, KULLBERG BJ, MEIS JF:
24. MADU A, CIOFFE C, MIAN U Agents Chemother. (2000)
Aspergillus meningitis: diagnosis by
et al.: Pharmacokinetics of fluconazole 44(11):3087-3091.
non-culture based microbiological methods
in cerebrospinal fluid and serum of 33. LORTHOLARY O, NICOLAS M, and management. J. Clin. Micro. (1999)
rabbits: validation of an animal model SOREDA S et al.: Fluconazole, with 37:1186-1189.
used to measure drug concentrations in or without dexamethosone for experimental
43. SCHWARTZ S, THIEL E: CNS-
cerebrospinal fluid. Antimicrob. Agents cryptococcosis: impact of treatment timing.
aspergillosis: are there new treatment
Chemother. (1994) 38(9):2111-2115. J. Antimicrob. Chemother. (1999)
options? Mycoses (2003) 46(Suppl. 2):8-14.
25. FOULDS G, BRENNAN DR, 43:817-824.
44. SCHWARTZ S, MILATOVIC D,
WAJSZCZUK C et al.: Fluconazole 34. SAVANI DV, PERFECT JR, COBO LM,
THIEL E: Successful treatment of
penetration into cerebrospinal fluid DURACK DT: Penetration of new azole
cerebral aspergillosis with a novel triazole
in humans. J. Clin. Pharmacol. (1988) compounds into the eye and efficacy in
(voriconazole) in a patient with acute
28(4):363-366. experimental Candida endopthalmitis.
leukemia. Br. J. Haematol. (1997)
26. YANG H, WANG Q, ELMQUIST WF: Antimicrob. Agents Chemother. (1987)
97:663-665.
Fluconazole distribution to the brain: 31(1):6-10.
45. AL-ABDELY HM, NAJVAR L,
a crossover study in freely-moving rats 35. PERFECT SD Jr, DURACK DT:
BOCANEGRA R et al.: SCH 56592,
using in vivo microdialysis. Pharm. Res. Comparison of itraconazole and
amphotericin B, or itraconazole
(1996) 13(10):1570-1575. fluconazole in treatment of Cryptococcal
therapy of experimental murine
27. THALER F, BERNARD B, TOD M meningitis and Candida pyleonephritis
cerebral phaeohyphomycosis
et al.: Fluconazole penetration in in rabbits. Antimicrob. Agents Chemother.
due to Ramichloridium obovoideum.
cerebral parenchyma in humans at steady (1986) 29(4):579-583.
Antimicrob. Agents Chemother. (2000)
state. Antimicrob. Agents Chemother. (1995) • Important comparison of efficacy of
44(5):1159-1162.
39:1154-1156. antifungal drugs with varying CSF
pharmacokinetics. 46. IMAI JK, SINGH G, CLEMONS KV,
28. FISCHMAN AJ, ALPERT NM, LIVNI E STEVENS DA: Efficacy of posaconazole
et al.: Pharmacokinetics of 18F-labeled 36. KAMBERI P, SOBEL RA,
in a murine model of central nervous
fluconazole in healthy human subjects by CLEMONS KV et al.: Comparison
system aspergillosis. Antimicrob. Agents
positron emission tomography. of itraconazole and fluconazole treatment
Chemother. (2004) 48(10):4063-4066.
Antimicrob. Agents Chemother. (1993) in a murine model of Coccidioidal
meningitis. Antimicrob. Agents 47. BARCHIESI F, SCHIMIZZI AM,
37:1270-1277.
Chemother. (2006) 51:998-1003. CASELLI F et al.: Activity of the new
29. WHEAT LJ, CONNOLLY P, antifungal triazole, posaconazole, against
• Important comparison of efficacy of
SMEDEMA M, BRIZENDINE E, Cryptococcus neoformans. J. Antimicrob.
antifungal drugs with varying CSF
HAFNER R: Emergence of resistance Chemother. (2001) 48:769-773.
pharmacokinetics.

580 Expert Opin. Drug Metab. Toxicol. (2007) 3(4)

 Kethireddy & Andes

48. GRAYBILL JR, NAJVAR LK, 56. SINGH G, IMAI J, CLEMONS KV, for the management of patients with
JOHNSON E, BOCANEGRA R, STEVENS DA: Efficacy of caspofungin blastomycosis. Infectious Diseases Society
LOEBENBERG D: Posaconazole therapy against central nervous system Aspergillus of America. Clin. Infect. Dis. (2000)
of disseminated phaeohyphomycosis fumigatus infection in mice determined by 30(4):679-683.
in a murine model. Antimicrob. Agents TaqMan PCR and CFU methods. 65. WHEAT J, SAROSI G, MCKINSEY D
Chemother. (2004) 48(6):2288-2291. Antimicrob. Agents Chemother. (2005) et al.: Practice guidelines for the
49. BAYER AS, YOSHIKAWA TT, 49(4):1369-1376. management of patients with
GALPIN JE, GUZE LB: Unusual 57. LUQUE JC, CLEMONS KV, histoplasmosis. Infectious Diseases
syndromes of coccidioidomycosis: STEVENS DA: Efficacy of micafungin Society of America. Clin. Infect. Dis.
diagnostic and therapeutic considerations; alone or in combination against systemic (2000) 30(4):688-695.
a report of 10 cases and review murine aspergillosis. Antimicrob. Agents 66. BLACK KE, BADEN LR: Fungal
of the English literature. Medicine Chemother. (2003) 47(4):1452-1455. infections of the CNS-treatment strategies
(Baltimore) (1976) 55(2):131-152. 58. HOPE WW, M.D et al.: Pharmacokinetics for the immnocompromised host.
50. GROLL AH, GULLICK BM, and pharmacodynamics of micafunign CNS Drugs (2007) 21:293-318.
PETRAITIENE R et al.: Compartmental in experimental hematogenous Candida •• Excellent review of CNS fungal infections.
pharmacokinetics of the antifungal meningoencephalitis. 46th Interscience 67. WANG EJ, LEW K, CASCIANO CN,
echinocandin caspofungin (MK-0991) Conference on Antimicrobial agents and CLEMENT RP, JOHNSON WW:
in rabbits. Antimicrob. Agents Chemother. chemotherapy. Washington DC, USA Interaction of common azole antifungals
(2001) 45(2):596-600. (2006):398. with P glycoprotein. Antimicrob. Agents
51. GROLL AH, MICKIENE D, •• Important study of relationship between Chemother. (2002) 46(1):160-165.
PETRAITIENE R et al.: Pharmacokinetic CNS kinetics and efficacy which includes
68. WISHART DS, DRUGBANK EA:
and pharmacodynamic modeling of detailed measurement of drug
A comprehensive resource for in silico
Anidulafungin (LY303366): reappraisal concentrations in various CNS
drug discovery and exploration.
of its efficacy in neutropenic animal compartments and tissues.
Nucleic Acids Res. (2006)
models of opportunistic mycoses using 59. MAERTENS J, RAAD I, PETRIKKOS G 34(Database issue):D668-D672.
optimal plasma sampling. et al.: Efficacy and safety of caspofungin
69. WALSH TJ, FOULDS G, PIZZO PA:
Antimicrob. Agents Chemother. (2001) for treatment of invasive aspergillosis
Pharmacokinetics and tissue penetration
45(10):2845-2855. in patients refractory to or intolerant
of fluconazole in rabbits. Antimicrob. Agents
52. GROLL AH, MICKIENE D, of conventional antifungal therapy.
Chemother. (1989) 33(4):467-469.
PETRAITIS V et al.: Compartmental Clin. Infect. Dis. (2004) 39:1563-1571.
70. HAJDU R, THOMPSON R,
pharmacokinetics and tissue distribution 60. KARTSONIS NA, SAAH AJ,
SUNDELOF JG et al.: Preliminary animal
of the antifungal echinocandin lipopeptide JOY LIPKA C, TAYLOR AF, SABLE CA:
pharmacokinetics of the parenteral
Micafungin (FK463) in rabbits. Salvage therapy with caspofungin for
antifungal agent MK-0991. Antimicrob.
Antimicrob. Agents Chemother. (2001) invasive aspergillosis: results from the
Agents Chemother. (1997) 41:2339-2344.
45(12):3322-3327. caspofungin compassionate use study.
J. Infect. (2005) 50:196-205. 71. CHIN T, FONG IW,
53. IMAI J, SINGH G, FERNANDEZ B,
VANDENBROUCKE A:
CLEMONS KV, STEVENS DA: 61. NATARAJAN G, LULIC-BOTICA M,
Pharmacokinetics of fluconazole in serum
Efficacy of abelcet and caspofungin, alone RONGKAVILIT C, PAPPAS A,
and cerebrospinal fluid in a patient with
or in combination, against CNS aspergillosis BEDARD M: Experience with caspofungin
AIDS and Cryptococcal meningitis.
in a murine model. J. Antimicrob. in the treatment of persistent fungemia in
Pharmacotherapy (1990) 10(4):305-307.
Chemother. (2005) 56:166-171. neonates. J. Perinatol. (2005)
25(12):770-777. 72. UTZ CJ, SHADOMY S: Antifungal
54. IBRAHIM AS, BOWMAN JC,
activity of 5-fluorocytosine as measured
AVANESSIAN V et al.: Caspofungin 62. ODIO CM, ARAYA R, PINTO LE
by disk diffusion susceptibility testing.
inhibits Rhizopus oryzae 1,3 β-D-glucan et al.: Caspofungin therapy of neonates
J. Infect. Dis. (1977) 135(6):970-974.
synthase, lowers burden in brain measured with invasive candidiasis. Pediatr. Infect.
by quantitative PCR, and improves survival Dis. J. (2004) 23(12):1093-1097.
at a low but not a high dose during murine 63. PATTERSON TF, KIRKPATRICK WR, Affiliation
disseminated zygomycosis. Antimicrob. WHITE M et al.: Invasive aspergillosis. Shravan Kethireddy & David Andes†
Agents Chemother. (2005) 49(2):721-727. †Author for correspondence
Disease spectrum, treatment practices,
55. HOSSAIN MA, REYES GH, LONG LA, and outcomes. I3 Aspergillus Study University of Wisconsin School of Medicine
MUKHERJEE PK, GHANNOUM MA: Group. Medicine (Baltimore) (2000) and Public Health, Madison 600 Highland Ave,
Efficacy of caspofungin combined with 79(4):250-260. H4/572, Madison, WI 53792, USA
amphotericin B against azole-resistant Tel: +1 608 263 1545; Fax: +1 608 263 4464;
64. CHAPMAN SW, BRADSHER RW,
Candida albicans. J. Antimicrob. Chemother. E-mail: dra@medicine.wisc.edu
CAMPBELL GD, PAPPAS PG,
(2003) 51:1427-1429. KAUFFMAN CA: Practice guidelines

Expert Opin. Drug Metab. Toxicol. (2007) 3(4) 581