AAPS PharmSciTech (2020) 21:58

DOI: 10.1208/s12249-019-1606-6

Research Article

In Vivo Evaluation of a Novel Chitosan-Polycaprolactone Based Mucoadhesive

Gastro-Retentive Sustained Release Drug Delivery System for Milnacipran HCl

Talib Hussain,1,2 Muhammad Ijaz,3 Rahat Shamim,1 Khalid Hussain,1 Nasir Abbas,1 Amjad Hussain,1
Julia Anita Griessinger,4 and Nadeem Irfan Bukhari1,5

Received 11 September 2019; accepted 9 December 2019

Abstract. The study was aimed to develop a gastro-retentive mucoadhesive sustained release
matrix formulation for milnacipran HCl (MCN) by using the design of experiment (DoE).
The gastro-retentive swellable mucoadhesive matrix tablets were prepared by modified
solvent-based wet granulation through mixing milnacipran (MCN), chitosan low molecular
weight (CH-LM), chitosan medium molecular weight (CH-MM), and polycaprolactone
(PCL). Optimization of the formulation was carried out via DoE. Formulations were
characterized by DSC, FTIR, and in vitro drug release testing. In vitro mucoadhesive studies
were performed on rabbit’s intestinal mucosa. In vivo drug release studies were performed
on dogs. Optimized matrix formulations showed no significant interaction among the
polymers and MCN, confirmed by DSC and FTIR, and were characterized as swellable
controlled release matrix systems. The optimized formulations MOPT3 and MOPT4 showed
significantly improved adhesion time of 12 h on the gastric mucosa. Based on the in vivo
analysis, the elimination half-life of MCN was increased that proved the matrix formulation
to be sustained release DDS. The Tmax was extended from 2 to 12 ± 1.63 h for MOPT4. Cmax
of matrix was reduced to 121.60 ± 9.496 ng/ml as compared to 149.22 ± 9.942 ng/ml of
solution. The bioavailability of the matrix formulation was significantly improved as
compared to the MCN solution by 272.20 ± 48.11%. The controlled drug release and strong
mucoadhesive properties of the gastro-retentive matrix formulations suggested the potential
application of the formulations for the extended oral delivery of MCN.
KEY WORDS: milnacipran; anomalous release; canine-based controlled delivery; bioavailability;
mucoadhesion.

INTRODUCTION The gastric emptying process can vary between a few

minutes and 12 h, depending upon the pharmaceutical
A drug delivery scientist, while developing an oral formulation and physiological condition of the subject. Since
controlled drug delivery system, focuses mainly on the the majority of active pharmaceutical ingredients (APIs) are
achievement of more predictable and increased bioavailabil- absorbed in the gastric and upper part of the small intestine,
ity of drugs. Nevertheless, the designing process is barred by an inconsistency in the emptying process may lead to
certain physiological complications, such as failure to confine irregularities of the plasma levels and variations in bioavail-
and restrict the drug delivery system (DDS) within the ability. Additionally, the comparatively short transitory phase
preferred sections of the gastrointestinal (GI) tract and the of 2–3 h through the most important absorption zone can lead
extremely capricious nature of gastric emptying process that to an inadequate drug release from the DDS. Consequently,
have to be considered. targeting of a DDS in a specific region of GI tract might offer
several advantages, particularly for drugs that have an
absorption window in the upper GI tract (1). Furthermore,
1 APIs having a short half-live require frequent administration
University College of Pharmacy, University of the Punjab, Lahore,
54000, Pakistan. which escalates the probability of non-compliance by patient,
2
Institute of Pharmaceutical Sciences, University of Veterinary and therefore affecting the overall effectiveness of therapy,
Animal Sciences, Lahore, Pakistan. followed causing an inconvenience to patients (2) and also
3
Faculty of Pharmacy, COMSATS University of Islamabad, Lahore produce the shorter therapeutic action. Considering
campus, Lahore, Pakistan. sustained/controlled release system as potential candidate
4
Institute of Pharmacy, University of Innsbruck, Innsbruck, Austria. for above problems, it might not produce desired longer
5
To whom correspondence should be addressed. (e–mail: therapeutic responses if the absorption site is only in the first
nadeem_irfan@hotmail.com)

1530-9932/20/0000-0001/0 # 2020 American Association of Pharmaceutical Scientists

 58 Page 2 of 11 Hussain et al. (2020) 21:58

part of the GI tract. Therefore, such drugs are suitable the dog center of the University of Veterinary and Animal
candidates for oral gastro-retentive drug delivery systems for Sciences (UVAS), Lahore. An ethical approval was obtained
their extended absorption phase and therapeutic actions. from the animal ethics committee of the Punjab University
Milnacipran (MCN), a serotonin-norepinephrine reup- College of Pharmacy, University of the Punjab, Lahore, vide
take inhibitor used for clinical treatment of fibromyalgia, is letter no. AEC/UCP/1036/4313, dated 12 December 2015.
freely soluble in water and intestinal fluids so it is impossible
to get prolonged therapeutic effects when administered in the Preparation of Swellable Matrix Tablets
form of immediate release tablets as it requires multiple time
administration (3). The excessive (quick) release of MCN in The swellable matrix tablets of MCN (for design of
immediate release tablets is linked with unwanted effects such experiment (DoE) analysis) were prepared by modified
as abdominal irritation, diarrhea, nausea, tachycardia, palpi- solvent-based wet granulation method. Briefly, the weighed
tations, hypertension, extremes of elevated mood or feelings quantities of PCL were dissolved in chloroform/DCM. The
of happiness that may shift to a depressed or sad mood, solution of PCL in chloroform/DCM was added to the dry
unsteadiness or falling, and swelling of the mouth or throat mixture of CH, drug, and diluents followed by blending this
(4,5). These might be associated with poor compliance, so mixture to cause the granulation. The prepared granules were
there is a requirement of MCN formulation that can compressed directly to swellable mucoadhesive matrix tab-
overcome these problems and produces therapeutic effect lets. The granules were prepared using drug to polymer ratios
for approximately 24 h when administered orally (6). of 0.5/1, CH to PCL ratios of 80/20, 85/15, 90/10, 95/05,
Moreover, the high water solubility of MCN makes it difficult respectively, using both low and medium molecular weight
to incorporate the API within conventional sustained release chitosan (CH-LM and CH-MM), separately. Each blend
formulations. A novel polymeric gastro-retentive contained fixed amount of MCN (drug), lactose (diluent),
mucoadhesive matrix system containing the lipophilic and magnesium stearate (lubricant) in quantities of 100, 320,
polycaprolactone (PCL) for sustained release of MCN might and 5 mg, respectively. The data obtained from these
be developed for once daily dosage regimen. These gastro- formulations were analyzed and the optimized levels of
retentive sustained release systems ensure the availability of factors were found on basis of the desired quality attributes.
desired therapeutic drug concentration to the absorption site DoE optimized the blends with drug (100 mg) and polymers
for the duration of therapy that can minimize the possibility quantities (200 mg) using both grades of chitosan and solvents
of local or systemic toxic response (7). Therefore, in this study (DCM and chloroform).
two polymer combinations were evaluated to retard and The optimized formulations, labeled as MOPT1 to MOPT4
control the drug release. (Table I), with a reduced amount of MCN (30 mg) according
Hence, the current study was aimed at the development to the dose (weight-dependent) of the animal (dog) were
and characterization of mucoadhesive gastro-retentive prepared using above optimized blend of polymers (200 mg)
sustained release matrix system for MCN. The MCN con- and incorporating additional diluent as replacement of drug.
taining gastro-retentive mucoadhesive tablets were character- Based on superior in vitro release properties of the MOPT4
ized in vitro to determine mucoadhesion, drug release, and formulation compared to other formulations, it was selected
quality parameters. The optimized formulation was evaluated for the final in vivo pharmacokinetic study. The prepared
for the in vivo pharmacokinetics study. matrix tablets were stored in desiccators at room temperature
for further processing.
MATERIALS AND METHODS
Characterization of Swellable Matrix Tablets
Materials
The prepared swellable mucoadhesive matrix tablets
Milnacipran HCl (99.7%), polycaprolactone (40 kD) were characterized for physical parameters including weight
granular flakes, and chitosan (93% w/w) (76.19% degree of variation, friability, hardness, thickness, diameter, and content
de-acetylation) as low molecular weight and medium molec- uniformity according to USP specifications.
ular weight were purchased from Sigma-Aldrich, Pakistan.
Magnesium stearate NF, lactose of Parchem USA, acetoni- Fourier Transform Infrared Spectroscopy
trile, dichloromethane (DCM), chloroform, glacial acetic acid,
hydrochloric acid (37% v/v), monobasic potassium phos- Drug and excipients compatibility analysis was per-
phate, potassium chloride, potassium dihydrogen phosphate, formed using Fourier transform infrared spectroscopy
ammonium chloride, triethylamine, diethyl ether, orthophos- (FTIR). The scanned spectra were collected on FTIR
phoric acid, and sodium hydroxide were used from Merck, spectrophotometer (Thermo Nicolet 6700) at a resolution of
F.R Germany. EDTA tubes (K3EDTA) were from ATLAS- 4 cm−1 in between wavenumber of 400 and 4000 cm−1. The
LABOVAC Italiano, and normal saline (0.9%) from Otsuka, small amount of grinded sample was placed on the detector
Pakistan. All chemicals and reagents purchased were of and was compressed to a 12-mm disc by a hydraulic press at
analytical grade. 10 tons’ compression force for 30 s.

Animals Differential Scanning Calorimetry

For in vivo pharmacokinetics study, mixed breed, hybrid Differential scanning calorimeter (Q2000 DSC TA
strain adult male dogs weighing 12–18 kg were obtained from instrument) was used for the analysis of drug, excipients,
In Vivo Evaluation of Novel CH-PCL GRDDS for Milnacipran (2020) 21:58 58 Page 3 of 11

Table I. Optimized Formulations of MCN Based on Preliminary Factors for In Vivo Analysis on Dogs at Scaled Down MCN (30 mg)

Formulation Polymer ratios (mg) Solvents Tablet wt (mg)

CH-LM CH-MM PCL

MOPT1 10 185 05 Dichloromethane 555

MOPT2 05 185 10 Dichloromethane 555
MOPT3 10 185 05 Chloroform 555
MOPT4 05 185 10 Chloroform 555

CH-LM chitosan low molecular weight, CH-MM chitosan medium molecular weight, PCL polycaprolactone

and prepared mucoadhesive tablets. Samples (5.5 mg) of pure then stuck to the surface of mucosa by applying a little force
drug (MCN), CH-LM, CH-HM, PCL, and matrix tablets were with fingertip for almost 30 s. The glass slides having tablets
sealed in aluminum pan individually and differential scanning attached to mucosa were then placed in beaker containing
calorimetry (DSC) thermograms were recorded from 25 to 500 ml of HCl buffer, pH 1.2, being maintained at 37°C ±
250°C, a flow rate of nitrogen gas at 10 ml per min and 0.5°C to determine the adhesion time for 12 h. The buffer was
heating rate of 10°C/min was maintained during the experi- stirred at 50 rpm with a paddle type of impeller to simulate
ment. An empty aluminum pan was used as a reference. gastric conditions. The detachment time for the
mucoadhesive tablets was noted and mentioned as a
Dissolution Studies mucoadhesive time.
The in vitro mucoadhesion strength was measured in
In vitro dissolution studies for gastro-retentive matrix terms of the force required to pull out an intestinal mucosa
tablets were conducted to analyze the percentage of drug (surface area 3 mm2) from the surface of mucoadhesive tablet
released after specified time intervals, using USP dissolution with a tensiometer (LLOYD instruments, LF2276, Germany).
test apparatus II (paddle type), Erweka, Germany. Dissolu- The mucoadhesive tablets (weighing 550 ± 15 mg) were
tion was carried out using 900 ml HCl buffer, pH 1.2 (37 ± placed on the base support of the instrument previously
0.5°C), with paddles rotation set at 50 rpm. With each immersed in HCl buffer, pH 1.2, for at least 15 min. Intestinal
sampling, 5 ml aliquots were pipetted at 0.5, 1, 1.5, 2, 3, 4, 6, mucosa was fixed uniformly to upper support plate of the
8, 10, and 12 h intervals, and replaced with dissolution instrument with cyanoacrylate adhesive. The mucosa was
medium, maintained at 37 ± 0.5°C. The filtered samples, then suspended from the tensiometer spring to lower towards
diluted ten times with the dissolution media, were derivatized the tablet until it just contacted the surface of the tablet. A
using ninhydrin to form Ruhemann’s Purple and were 60-dyne force, measured by the torsion balance of the
assayed for released MCN at 570 nm using dual beam UV- instrument as a negative force, was applied to the tablet for
visible spectrophotometer (AG instruments UV-1602, 120 s, so that a complete bond formation can take place.
Shimadzu, Japan) and compared against the standard curve Then, the intestinal mucosa was raised slowly until it was
prepared of known drug solutions analyzed at same wave- completely detached from the tablet. The tensiometer value
lengths using the same method of derivatization (8). The at this point represented the adhesive bond strength between
obtained results in triplicate were averaged, standard devia- mucosa and tablet and was expressed as a positive force in
tions were calculated, and best-fit values were selected within dyne. This procedure was applied to three randomly selected
5% of the mean value. The percentage drug release was tablets from each MOPT1 to MOPT4 formulations, and the
calculated from the standard curve prepared. average was calculated with standard deviations (10).

Kinetics Analysis of Drug Release In Vivo (Pharmacokinetic) Study

The release data were analyzed using DD Solver, a For the pharmacokinetics, optimized formulation MOPT4
freeware Microsoft Excel add-in (9). The in vitro drug release (30 mg MCN) was selected based on better in vitro charac-
pattern was analyzed by fitting data to various release kinetic teristics (mucoadhesion and release parameters/kinetics). The
models including Zero order, First order, Higuchi, Hixson selected formulation MOPT4 was administered to model
Crowell, and Korsmeyer-Peppas model using the kinetic animals (dogs).
equations.
Study Design and Animal Housing
Mucoadhesive Time and Strength
The in vivo study used a balanced, single dose, parallel
The time of mucoadhesion was determined by fixing the design to compare the pharmacokinetics of MCN’s solution
mucoadhesive tablets on the surface of a freshly cut intestinal and developed formulation in two groups of dogs. One group
mucosa of rabbit. The mucosa was stretched and spread on was tested with oral solution of MCN, while another group
the flat glass slide and tied to the surface with a string crossing was tested with formulation (MOPT4).
the glass slide at three different positions. The mucoadhesive Six male hybrid strain dogs of weight between 12 and
tablets were soaked with HCl buffer, pH 1.2 for 5 min, and 18 kg were housed in the standard kennels equipped with a
 58 Page 4 of 11 Hussain et al. (2020) 21:58

well-ventilated, humidity- and temperature-controlled envi- HPLC Analysis of MCN in Blood Samples
ronment of Pet Centre, UVAS, Lahore, and fed on the
standard diet for successive 2 months. The individual dog was Preparation of Stock and Working Solutions
given a free access to water throughout the experiment with
feeding schedule approved by the veterinary nutrition MCN stock solution was prepared at a concentration of
department of UVAS, Lahore. The full protein diet along 1 mg/ml in double distilled water. Different working standards
with balanced administration of carbohydrates and bones was were prepared at concentration range of 100 ng/ml to
fed daily. At least 12 h prior to oral drug administration, the 100 μg/ml of MCN from stock solutions by serial dilutions
animals were fasted with only free water access. with double distilled water. The quantitative analysis of drug
was performed using calibration/standard curve, constructed
from prepared plasma drug standards by incorporating MCN
Administration of Drug Solution (Reference) and MOPT4 working standards in blank plasma at concentrations ranging
(Test) from 10 to 100 μg/ml. All the solutions were stored at 5 ± 3°C
till further analysis.
The 100 ml normal saline 0.9% (NS) was administered to
each dog 1 h prior to the administration of mucoadhesive Mobile Phase
tablets to build up the volume of blood of the animals in
order to maintain full health status. The oral tablets of The mobile phase comprised 720 ml phosphate buffer,
formulation (MOPT4) having dose of 2 ± 0.25 mg/kg in an pH 3.65 (prepared dissolving 1.7 g of potassium dihydrogen
unblinded, parallel design were administered at 9 AM every phosphate in 1 l of double distilled water to make 12.5 mM
day to individual dogs, without breaking the tablet, deep in solution) and 280 ml of acetonitrile. To this solution, 0.20% of
the throat with plenty of water after administration. The oral triethylamine (1 ml) was added, mixed, and adjusted pH of
solution of 30 mg/10 ml was administered with 90 ml of water solution to 3.65 using 0.1 M orthophosphoric acid. The mobile
in the same scheme as matrix tablets. phase was filtered through 0.45 μm nylon membrane filter
using syringe filtration assembly and degassed by placing in
an ultrasonic bath (12).
Blood Sampling from Animals
Instrumentation
Blood samples (≤ 5 ml) were withdrawn from a heparin-
ized catheter placed in either cephalic vein or lateral
For the HPLC analysis, Agilent Technologies 1260
saphenous vein before administration (zero time) as a blank
infinity auto sampler gradient liquid chromatography
and subsequently at predefined time intervals (1, 2, 4, 8, 12,
equipped with degasser, a pump, and UV-vis/fluorescent
18, 24, 48 h) in 5 ml K3EDTA (1.5% w/v) containing tubes.
detector was employed. Agilent HC-C18 fully end capped,
Samples at zero time were withdrawn to serve as blank
4.6 × 250 mm, 5 μm (588905-902) column placed at controlled
control for each animal.
temperature in column oven was used. The column was
The collected blood samples were immediately centri-
primarily equilibrated for at least 2 h with prepared mobile
fuged at 2500×g for 10 min by refrigerated centrifugation
phase at a flow rate of 0.5 ml/min. Separation was achieved by
(Hettich Zentrifugen, Japan) at 5 ± 3°C. The plasma layer was
isocratic solvent elution at a flow rate of 1 ml/min, while
separated from blood samples and transferred to Eppendorf
column temperature was kept ambient during the analysis.
tubes. These tubes were stored in freezer at − 18°C till
Backpressure of the system was about 2000 psi. The
extraction of MCN from plasma and analyzed for drug
separations of column were monitored at UV wavelength of
concentration.
220 nm and the injection volume was kept at 20 μl.

Statistical Data Analysis

Extraction of MCN from Samples
SPSS version 21 was used for the statistical analysis of
Frozen plasma samples of animals were thawed at the different parameters used in the study. The non-
room temperature. In a glass tube, 500 μl simple plasma parametric-based analysis of 2 sample independent test using
sample (10 μg/ml water spiked sample) or animal plasma Mann-Whitney test (2-tailed) was applied for determining the
sample was mixed with 150 μl of saturated ammonium significance of differences between different pharmacokinetic
chloride (NH4Cl, pH 9.6) on a horizontal shaker for 5 min. parameters computed by non-compartmental analysis that
To this mixture, 5 ml of diethyl ether was added and shaken were evaluated at a significance level of 0.05 (13).
on the horizontal shaker for 15 min and then centrifuged at
4 ± 0.5°C for 10 min at 2500×g. The organic layer collected RESULTS
from the samples was transferred to the conical bottom
tubes and was evaporated under nitrogen to dryness at a All the tablets prepared from the optimized gastro-
temperature of maximum 40°C. The dried residues were retentive mucoadhesive formulations via the modified
reconstituted with 1 ml of mobile phase, filtered through solvent-based wet granulation were characterized for differ-
0.45 μm nylon membrane filter using syringe filtration ent quality control parameters. The friability values of all four
assembly, and transferred to HPLC auto sampler vials for tablet formulations (0.38–0.47%) were within specified limits
analysis (11). of 1% of weight (14). Hardness of the tablets were found
In Vivo Evaluation of Novel CH-PCL GRDDS for Milnacipran (2020) 21:58 58 Page 5 of 11

between 5.8 and 6.6 kg/cm2, higher hardness was observed diffusion was excessive during the first hour of the study,
when formulated with chloroform as a solvent (MOPT3 and i.e., 25.25% for (MOPT2) as compared to 16.22% of (MOPT4),
M OPT 4). The content uniformity of drug in prepared while a sufficient API amount, i.e., ≥ 50% was released from
formulations showed the finest results of drug contents (100 the formulations in 4 h as shown in Fig. 2.
± 2%) in each tablet and low degree of weight variation Kinetic analysis was performed to interpret the release
among the tablets (1.8–3.8%) was noted. pattern as well as to determine the mechanism of drug release
from matrix system as shown in Table II. Higuchi model was
Structural Characterization Using FTIR and DSC followed by almost all the formulations that described drug
release independent of concentration and mimic zero order
The spectral analysis of MCN indicated stretching release mechanism making the systems to be designated as
vibrations of primary -NH2 at two points of expression as controlled release systems (15). The values of diffusion
3220 and 3148 cm−1, and methyl (CH3), C-H stretching exponent derived from Korsmeyer-Peppas model designated
vibrations at 2923, 2892, and 2809 cm−1. The carbonyl amide as n were found between 0.5 and 0.7 corresponding non-
(RNCO) sharp stretching bands were observed at wavelength Fickian drug release mechanism with combined diffusion and
of 1606 cm−1 and bending at 1590 cm−1. The secondary amine dissolution system that allowed both swelling and erosion of
(C-N) and C-H bending vibrations were displayed at broader the matrices.
region of 1470–1430 cm−1.
The mucoadhesive matrix tablets of MCN (MOPT1, Comparative Release Profiles of Formulations
MOPT2, MOPT3, and MOPT4) with polymers (CH-LM, CH-
MM, and PCL) scanned for FTIR spectrum displayed the The release profiles of scaled optimized formulations
hydroxyl (OH) of chitosan at 3510 cm−1 and wider bands of MOPT1–MOPT4 were compared to the reference standard
primary amine (NH2) of both MCN and CH in the region of specifications established by USP for the dissimilarity factor,
3250–3400 cm−1. The presence of amide carbonyl linkage f1, and similarity factor, f2. The extent to which the optimized
(O=C-N) of MCN was confirmed by stretching peak at formulations meet the desired release specifications was
1652 cm−1 and secondary amine (N-C) group was attributed verified. The values of f1 were found within the specified
to the stretching peaks at 1560 cm−1. These characteristics limits of 0–15 and f2 were determined within 50–100 that
conform the presence of all three polymers in blend with correspond to similarity with the reference. The best results
MCN. (f1 = 10.38, f2 = 61.28) were produced by MOPT1, consisting of
The optimized formulations were also evaluated for the CH-LM and PCL (dissolved in DCM) followed by MOPT4,
purity of the chemical substances and drug to excipient prepared using chloroform as solvent for PCL and CH-MM
compatibility using DSC. as polymer. Formulation MOPT3 was found dissimilar on the
The thermograms of the optimized formulations basis of f1 values (f1 = 15.24, f2 = 51.58) found greater than
(M OPT 1–4), as expressed in Fig. 1, indicated a well- 15.
established compatibility among the different formulation
ingredients. The shortest bending endotherm at 57.6°C Mucoadhesion
(MOPT1), 49.8°C (MOPT2), 56.7°C (MOPT3), and 58.2°C
(MOPT4) represented the PCL endotherm being shifted to The gastro-retentive mucoadhesive formulations contain-
lower temperature indicative of amorphous form of crystal- ing CH remained attached to the mucus membrane for a
line polyester due to interactions and bonding with chitosan. longer period due to covalent interactions with mucous
The compatibility of drug (MCN) in the matrix was confirmed membrane. The gastro-retentive tablets remained intact in
by the presence of sharper endothermic peaks at 153.44°C the buffer media for a prolonged time as shown in Fig. 3.
(M OPT 1), 153.14°C (M OPT 2), 153.48°C (M OPT 3), and MOPT4 remained attached to the mucous membrane for
154.30°C (MOPT4) with little enthalpy changes representing almost 12 h demonstrating excellent adhesion time on the
amorphous nature of the drug. The lesser melting point and mucous membrane.
heat of fusion than the original pure MCN values represents
the covalent and ionic interactions depicting a true blend of In Vivo Studies
MCN. The endothermic peaks near 200°C in all the
thermograms of the formulations represented the presence The adjusted dose according to weight of individual dog
of lactose in the matrix having amorphous characteristics was administered in the form of oral solution (30 mg/100 ml)
indicative of small enthalpy change of 14.13 J/g. and plasma level time curve for MCN was drawn for
The above findings indicated the possibility to formulate individual dog, and mean values were calculated. The drug
MCN with these two polymers using modified solvent-based (solution as well as MOPT4) was well tolerated by all the
granulation technique without any significant change in the animals with least appearance of adverse events (AEs),
chemical and physical properties of the polymers. during and after the course of therapy. The level of tolerance
was measured on basis of animal’s behavior and CNS effects
Release Parameters and Kinetics of Drug Release that appeared when higher dose (100 mg) was administered
(16). The primary tolerance parameters were response to
Release parameters, including percentage of drug re- environment, food, starring at attendant, abnormal eye
leased after maximum dissolution time (12 h), T25%, T50%, movements, walking overly excited and fearing of water,
and T80%, were calculated from the in vitro dissolution etc., along with AEs like loss of appetite, induction of
experiment for formulations M OPT 1–M OPT 4. The drug vomiting, and diarrhea. Pharmacokinetic parameters were
 58 Page 6 of 11 Hussain et al. (2020) 21:58

Fig. 1. Thermogram of minaciporan HCl (a), chitosan (b), PCL (c),

and formulation MOPT4 (d) indicating melting endotherms at different
temperatures

calculated based on obtained plasma level time curve data six dogs as well as for mean values using non-compartmental
after administration of oral solution of MCN to six dogs approaches opting PKSolver.
adopting non-compartmental model approach. The good
regression values obtained after employing non- Validation of the Developed HPLC Method
compartmental approach suggested valid investigational pa-
rameters based on computed R2 values. Pharmacokinetic Recovery studies of MOPT4 formulation depicted the
parameters of absorption, distribution, and elimination based method as highly accurate when evaluated on basis of
on the plasma level time curve were calculated for individual addition of three different concentrations (4, 10, and 20 μg/ml)
In Vivo Evaluation of Novel CH-PCL GRDDS for Milnacipran (2020) 21:58 58 Page 7 of 11

Fig. 2. Percentage of drug released according to time from MCN

matrix formulations MOPT1 to MOPT4 with standard deviation (n = 6)
using 900 ml HCl buffer, pH 1.2 at 37°C for 12 h

to 24 μg/ml sample and obtained values were 97.97, 101.3, and of the drug from the body. A markedly less AUC0-∞ value
99.76%, respectively. RSD values were 1.58, 1.18, and 0.65% was observed for oral solution, i.e., 945.49 ± 86.56 compared
respectively with very low level of present error. The to 1032.57 ± 99.20 ng.h.ml−1 of MOPT4. The mean residence
reproducibility and repeatability were evaluated on the basis time (MRT0-t) of MCN, after adminstration of MOPT4 was
of interday and intraday precision analysis (n = 6). Interday found as 16.25 ± 0.95 h with total AUMC0-∞ of 45,410.3 ±
precision was 3.7 ± 1.0% while intraday was evaluated as 2.56 7023.3 ng.h2.ml−1. For oral MCN solution, MRT0-∞ and
± 2.1% with RSD of 3.4 ± 1.8% and error of 1.19 ± 0.89%. AUMC0-∞ values were calculated as 8.88 ± 0.75 h and
The method was also found sensitive based on LOD and 9170.11 ± 1392.56 ng.h 2 .ml −1 , respectively. Volume of
LOQ values. The very low values of detection (LOD = distribution (Vd) and total clearance (CLT) from body were
3.08 μg/ml) and quantification (LOQ = 9.33 μg/ml) demon- found as 137.68 ± 12.037 ml and 10.74 ± 1.22 ml.h−1 for
strate the developed method as most sensitive in estimating MOPT4; for the oral solution, Vd and CLT were calculated as
minutes quantities of MCN in the plasma samples of dogs. 275.53 ± 34.81 ml and 29.05 ± 2.88 ml.h−1, respectively.

Pharmacokinetics of MCN Swellable Matrix Tablet (MOPT4) Comparative Pharmacokinetics of MCN Solution and MOPT4
and MCN Oral Solution Tablet

The maximum peak plasma concentration (Cmax) of Comparative plasma level time profiles after administra-
121.60 ± 9.50 ng/ml was achieved at the Tmax of 12 ± 1.63 h tion of MCN oral solution and formulation MOPT4 are given
for matrix tablet (MOPT4) as compared to 149.22 ± 9.94 ng/ml in Fig. 4. Concentrations of MCN at zero time of administra-
observed at Tmax of 2 h for oral solution, without any lag tion as well as beyond 24 h in case of oral solution and 48 h in
time. The elimination half-life of MCN was significantly case of test formulation administration were below the
increased when formulated in MOPT4 formulation, with an detection limit (LOQ) of used HPLC method. The MCN
average value of 8.89 ± 0.86 h compared to oral solution (6.57 from the oral solution showed rapid absorption with a
± 0.37 h), which represents a better and longer stay of MCN maximal concentration attained at 2 h of dosing and then
in the plasma. The elimination rate constant calculated for promptly eliminated from the dog plasma with no detection
MOPT4 as 0.078 ± 0.008 h−1 compared to 0.105 ± 0.00682 h−1 of drug after 24 h. By contrast, the MCN from formulation
for the oral solution showed a significantly lower elimination MOPT4 showed a slowly increase to maximal peak plasma

Table II. Drug Release Kinetics of MCN Formulations Showing Correlation Coefficient R2, Slope K, and n Values of Formulations by
Applying Different Kinetic Equations on Dissolution Profiles

Formulation Hixson Crowell Zero order First order Higuchi Korsmeyer-Peppas

Q01/3 – Qt1/3 = KHCt Qt = K0t Qt = lnQ0 − K1t Mt/M∞ = KHt1/2 Mt/M∞ = Ktn

KHC R2 K0 R2 K1 R2 KH R2 R2 n

MOPT1 0.001 0.9864 0.159 0.927 0.003 0.975 3.511 0.9619 0.995 0.676
MOPT2 0.001 0.9694 0.157 0.756 0.004 0.990 3.561 0.9900 0.990 0.512
MOPT3 0.001 0.9897 0.165 0.883 0.004 0.983 3.674 0.9735 0.991 0.620
MOPT4 0.001 0.9934 0.161 0.947 0.003 0.975 3.548 0.9486 0.996 0.718
 58 Page 8 of 11 Hussain et al. (2020) 21:58

Fig. 3. In vitro mucoadhesive capacity (expressed as detachment force,

mean ± SD, n = 3) and in vitro residence time (mucoadhesion time) in
stomach mucosa using simulated gastric fluid pH 1.2 at 37°C (mean ±
SD, n = 3) of swellable matrix tablets

concentration at 12 h and also the elimination was extended freely soluble in water and intestinal fluids so it is impossible
till sampling interval of 48 h. to get prolonged therapeutic effects when administered in the
According to Mann-Whitney test statistics, all the form of immediate release tablets (3). The swellable
parameters calculated shown in Table III were found mucoadhesive gastro-retentive matrix tablets were prepared
significantly different (P < 0.05). Higher MCN concentrations by a modified solvent-based granulation method. The char-
were found for MCN solution at various time points (h), i.e., acterization of the physical parameters of swellable matrix
at 0.5 h, 47.81 ± 7.64 ng/ml as compared to 15.56 ± 1.79 ng/ml tablet was found satisfactory. The physical appearance of the
of MOPT4 tablets (P = 0.004). The measured variations among tablets was smooth, shiny, and light brown in color. The FTIR
the six dogs after administration of MCN solution were found spectra of MCN, CH-LM, CH-MM, PCL, and different
greater than the test formulation, indicating stability of drug formulations were generated to evaluate and to identify the
release from test formulation (17). The major differences possible physical interactions between the polymers and the
were observed in absorption phase. Rapid peak plasma drug. The aromatic ring presence was indicated as
concentration was observed within 2 h after administration monosubstituted benzene and showed stretching and bending
of the solution, whereas rapid peak plasma concentration was vibrations at 691 and 732 cm−1 conforming the presence of
delayed by up to 12 h when MOPT4 tablets were administered MCN in spectrum run (19). The mucoadhesive matrix tablets
(P = 0.004). Similarly, absorption phase was steadier and of MCN (MOPT1, MOPT2, MOPT3, and MOPT4) with polymers
much prolonged in case of MOPT4 tablets as compared to (CH-LM, CH-MM, and PCL) were scanned for FTIR
solution representing sustained release of MCN from MOPT4 spectrum to confirm the existence of physical interactions
tablets (18). and absence of any chemical interactions between drug and
polymers. The peak at 1712 cm−1 was assigned to C=O group
Relative Bioavailability Pharmacokinetics existence in the N(COR)2 amide ester linkage that might be
formed by reaction of OH of PCL with NH of CH during
Relative bioavailability was calculated using total area solvent blending of the matrix (20). The existence of ether
under the curve values of test (MOPT4 matrix tablets) and the linkages of saccharide structure of chitosan (C-O-C) were
reference (solution) concentrations. The percentage relative indicated by peaks at 1050–1240 cm−1. The presence of
bioavailability was found massively increased for matrix benzene ring was confirmed by stretching and bending
tablet formulation, i.e., 272.20 ± 48.11%. The three-fold vibrations between wavenumbers of 690 to 750 cm−1 (21).
higher AUC 0-∞ of the matrix formulation (2795.80 ± The DSC thermogram of pure untreated MCN showed a
302.31 ng.h.ml−1) in comparison to the AUC0-∞ of the sharp endothermic peak at 179.85°C having heat of fusion as
solution (1027.12 ± 99.2 ng.h.ml −1 ) demonstrated the 97.96 J/g corresponding to its melting point and also
maximum increment in the oral bioavailability of MCN after demonstrated the purity of drug substance (22). The elon-
administration of MOPT4. The use of the independent sample gated stretched endotherms obtained at 90.1°C (MOPT1),
Mann-Whitney U test indicated that the distribution of 90.02°C (MOPT2), 91.7°C (MOPT3), and 90.2°C (MOPT4)
AUC0-∞ (ng.h.ml−1) represents a sustained release of MCN might be representative of ionic bonds formed with amino
from MOPT4 tablets, significantly different from the oral group of CH-LM with esters, carbonyl groups of PCL, and
solution (significance value of 0.002). MCN (23). The endotherms at 119.96°C (MOPT1), 126.2°C
(MOPT2), 122.4°C (MOPT3), and 129.24°C (MOPT4) might be
DISCUSSION due to CH-MM and CH-LM interaction and combination
with dominance of medium molecular weight component
Milnacipran (MCN), a serotonin-norepinephrine reup- (24). Release studies showed that when a higher amount of
take inhibitor used for clinical treatment of fibromyalgia, is PCL, in comparison to CH-LM was incorporated in
In Vivo Evaluation of Novel CH-PCL GRDDS for Milnacipran (2020) 21:58 58 Page 9 of 11

Fig. 4. Comparative plasma level time curve of MCN oral solution and
matrix MOPT4 tablets on dogs (n = 6) using a validated HPLC method

optimized blends, it resulted in an improved control on drug concentration and mimic zero order release mechanism
release (release retarded). Similarly, CH-MM appeared to be making the systems to be designated as controlled release
engaged as major swelling retardant polymer employing systems (15). The swelling controlled matrix system devel-
release dependent on diffusion from the matrix blends. The oped excessive pores in the matrix during the dissolution
presence of CH-LM in these formulations caused the early studies (26). These pores are typically responsible for the
relaxation of the polymeric matrix allowing dissolution media diffusion of drug from the matrix, along with polymeric
to penetrate CH-MM, which on the other hand swells initially relaxations that cause the matrix to erode and dissolve in the
to gain a specific size with porous nature. The release of drug media. Higuchi equation described the best possible way of
was found highly influenced by initial rate of swelling and drug diffusion from such matrix. The release data of the
later gradual and rapid erosion of the matrix concluding that formulations when applied to Higuchi equation showed a
a proper matrix blend of PCL and CH follows non-Fickian maximum correlation among release characteristics. Based on
diffusion type of drug release mechanism (anomalous release) this result, the release mechanism from the matrices is
(25). characterized as diffusive controlled and a near zero order
Drug diffusion from the matrices was greatly influenced drug release. This can be explained by the characteristics of
by pore formation in the matrix because of initial rapid chitosan. Chitosan due to its hydrophilic nature became
swelling and relaxation of polymeric blends. The diffusion of hydrated and swells in the presence of acidic media and
drug was the rate-limiting step that made it possible to caused the polymeric relaxation possessed diffusion of drug
characterize these formulations as diffusion-controlled sys- from the matrix (27). The slow relaxation of CH-MM in these
tems. Higuchi model was followed by almost all the formu- optimized blends caused the constant swelling rate along with
lations that described drug release independent of uniform erosion/dispersion of the whole matrix. This results

Table III. Comparative Non-compartmental Pharmacokinetics Parameters of MCN Oral Solution and MOPT4 Tablets by Non-parametric Test
Computed by PKSolver

Parameters Concentration (ng/ml) Difference

Mean ± SD (Sig. P 0.05)

MCN solution MOPT4 matrix tablets Mann-Whitney test (2-tailed)

Tlag (h) 0 0 –
Cmax (ng/ml) 149.22 ± 9.94 121.60 ± 9.496 0.004
Tmax (h) 2±0 12.00 ± 1.63 0.001
AUC0-t (ng/ml.h) 945.49 ± 86.56 2701.67 ± 276.83 0.004
AUC0-∞ (ng/ml.h) 1032.57 ± 99.2 2794.24 ± 302.31 0.004
AUMC0-∞ (ng/ml.h2) 9170.11 ± 1392.6 45,410.3 ± 7023.3 0.004
Vd (ml) 275.53 ± 34.80 137.68 ± 12.036 0.004
t1/2 (h) 6.573 ± 0.37 8.888 ± 0.864 0.004
Ke (h−1) 0.105 ± 0.007 0.078 ± 0.008 0.004
CLT (ml/h) 29.05 ± 2.88 10.74 ± 1.22 0.004
MRT0-∞ (h) 8.88 ± 0.75 16.25 ± 0.949 0.004

MCN milnacipran, AUC area under the curve, AUMC area under the moment curve, CL clearance, MRT mean residence time
 58 Page 10 of 11 Hussain et al. (2020) 21:58

to a drug release at a controlled rate and a predetermined controlled drug release and strong mucoadhesive properties
combined diffusion and dissolution parameters. Therefore, of the developed formulation suggested the potential appli-
the developed blends can be used for a control release of a cation of the gastro-retentive matrix system for the oral
API for a period of 12 h under intense acidic gastric delivery of MCN.
conditions (28).
The time a mucoadhesive system consumes before being ACKNOWLEDGMENTS
attached to the mucous membrane is lag time of adhesion and
the length it remained intact with membrane is called The authors are thankful to Pharmedic Laboratories for
mucoadhesion time. The gel-like appearances of the poly- providing analytical facilities. The authors also extend their
meric blends contributed to the strong adhesion. The results gratitude to Mr. Ahmad Aqib and Miss Sehar for cooperating
of mucoadhesive strength indicated that the excess quantity in arranging and managing the data for the research article.
of chitosan in the tablets have resulted in improved
mucoadhesion properties. After hydration of the tablets and
fastening to the mucus membrane, tablet produced a gel-like
network that interacted with mucus membrane by physical REFERENCES
entanglement along with the secondary hydrogen bonding
and van der Waals interactions. The swelling of the polymers
due to the water uptake had increased the kinesis of the 1. Malik R, Garg T, Goyal AK, Rath G. Polymeric nanofibers:
chitosan that facilitated the interaction of polymer to the targeted gastro-retentive drug delivery systems. J Drug Target.
mucus membrane. The greater hydration of CH-MM pro- 2015;23(2):109–24.
2. Kaur R, Garg T, Malik B, Gupta UD, Gupta P, Rath G, et al.
vided more effective chain flexibility for better physical Development and characterization of spray-dried porous
entanglement with mucus membrane (10). The main reason nanoaggregates for pulmonary delivery of anti-tubercular drugs.
for mucoadhesion lies in the fact that amino groups of intact Drug Delivery. 2016;23(3):872–7.
chitosans were dissociated with the positive charge that 3. Kulkarni S, Kulkarni R, Jadhav P. Tiwari A. Google Patents:
interacts with mucin chains and at the pH 1.2 exist as Controlled release pharmaceutical compositions of milnacipran;
2014.
negatively charged sialic acid (pKa 2.6) and sulfate residues 4. English C, Rey JA, Rufin C. Milnacipran (Savella), a treatment
of mucin glycoprotein. The greater amount of medium option for fibromyalgia. Pharm Ther. 2010;35(5):261.
molecular weight chitosan in the formulations was evident 5. Kyle JA, Dugan BD, Testerman KK. New drug approvals:
of this higher mucoadhesive strength due to the higher milnacipran for treatment of fibromyalgia. Ann Pharmacother.
2010;44(9):1422–9.
positive charge of dissociated amino groups (29). The
6. Ratnakar NC, Gohel MC. Formulation and eveluation of
pharmacokinetics of drugs demonstrates its absorption, milnacipran HCl controlled release osmotic tablets. Pharma
distribution, and elimination after it has been administered. Sci Monit. 2018;9(1).
The formulation development to design a dosage regimen of 7. Luo Y, Teng Z, Li Y, Wang Q. Solid lipid nanoparticles for oral
drugs can be utilized to efficiently determine steady state drug drug delivery: chitosan coating improves stability, controlled
delivery, mucoadhesion and cellular uptake. Carbohydr Polym.
concentration in plasma at specified time intervals that leads 2015;122:221–9.
to determining the optimum drug concentration at various 8. Hussain T, Shahzad MK, Hayat K, Hussain K, Bukhari NI.
time points in target tissues and fluids. The improved values Simple and sensitive colorimetric method for the determination
of Cmax and Tmax in the test formulation as compared to the of milnacipran in bulk and swellable matrix tablets. Pharm
Chem J. 2016;50(5):346–52.
oral solution represent drug sustainability of matrix
9. Zhang Y, Huo M, Zhou J, Zou A, Li W, Yao C, et al. DDSolver:
conforming a sustained release matrix (30). Sustained drug an add-in program for modeling and comparison of drug
release was also proved from the excessive increase in the dissolution profiles. AAPS J. 2010;12(3):263–71.
total area under curve (AUC0-∞) of MOPT4 formulation from 10. Luppi B, Bigucci F, Abruzzo A, Corace G, Cerchiara T, Zecchi
oral solution (31). V. Freeze-dried chitosan/pectin nasal inserts for antipsychotic
drug delivery. Eur J Pharm Biopharm. 2010;75(3):381–7.
11. Puozzo C, Filaquier C, Zorza G. Determination of milnacipran,
a serotonin and noradrenaline reuptake inhibitor, in human
CONCLUSION plasma using liquid chromatography with spectrofluorimetric
detection. J Chromatogr B Anal Technol Biomed Life Sci.
Modified solvent-based wet granulation method showed 2004;806(2):221–8.
promising results in solving the compatibility issue of water- 12. Parejiya PB, Movaliya VR, Barot BS, Modi D, Shelat PK,
Shukla A. Quantitative determination of milnacipran HCl in
soluble drug with polymer excipients. The optimized formu-
rabbit plasma by HPLC and its application to pharmacokinetics
lations developed by combining the chitosan and study. J Liq Chromatogr Relat Technol. 2014;37(1):99–111.
polycaprolactone showed appropriate water uptake and 13. Allen P, Bennett K, Heritage B (2018) SPSS statistics: a
controlled swelling properties. The diffusion of drug from practical guide with student resource access 12 months: Cengage
the mucoadhesive matrix was found purely dependent on the AU
14. USP. The United States Pharmacopeia. US Pharmacopeial
presence of pores in the system. All the gastro-retentive Convention Inc., Rockville, MD,USA: 2748–27512005
formulations developed showed improved mucoadhesion and 15. Hussain T, Saeed T, Mumtaz AM, Javaid Z, Abbas K, Awais A,
higher retention time on the mucus membrane. Pharmacoki- et al. Effect of two hydrophobic polymers on the release of
netic studies performed in vivo revealed that the developed gliclazide from their matrix tablets. Acta Pol Pharm.
2013;70(4):749–57.
formulation led to improved bioavailability of the drug. The
16. Pickering G, Macian N, Delage N, Picard P, Cardot J-M,
modified method for the development of gastro-retentive Sickout-Arondo S, et al. Milnacipran poorly modulates pain in
sustained release matrix tablets showed promising results and patients suffering from fibromyalgia: a randomized double-blind
might be applicable for the formulations of tricky drugs. The controlled study. Drug Des Devel Ther. 2018;12:2485.
In Vivo Evaluation of Novel CH-PCL GRDDS for Milnacipran (2020) 21:58 58 Page 11 of 11

17. Parejiya PB, Barot BS, Patel HK, Chorawala MR, Shelat PK, methotrexate-loaded chitosan microspheres for nasal adminis-
Shukla A. In vivo performance evaluation and establishment of tration. Drug Dev Ind Pharm. 2009;35(3):379–86.
IVIVC for osmotic pump based extended release formulation of 25. Siepmann J, Peppas N. Hydrophilic matrices for controlled drug
milnacipran HCl. Biopharm Drug Dispos. 2013;34(4):227–35. delivery: an improved mathematical model to predict the
18. Gaspar MC, Grégoire N, Sousa JJ, Pais AA, Lamarche I, Gobin resulting drug release kinetics (the “sequential layer” model).
P, et al. Pulmonary pharmacokinetics of levofloxacin in rats after Pharm Res. 2000;17(10):1290–8.
aerosolization of immediate-release chitosan or sustained- 26. Peppas NA, Narasimhan B. Mathematical models in drug
release PLGA microspheres. Eur J Pharm Sci. 2016;93:184–91. delivery: how modeling has shaped the way we design new
19. Gopal PR, Prabakar AC, Chandrashekar E, Somaiah PV. drug delivery systems. J Control Release. 2014;190:75–81.
Synthesis and characterization of process related impurities of 27. Omwancha WS, Mallipeddi R, Valle BL, Neau SH. Chitosan as
(±)-milnacipran. J Chin Chem Soc. 2013;60(6):639–44. a pore former in coated beads for colon specific drug delivery of
20. Wu C-S. A comparison of the structure, thermal properties, 5-ASA. Int J Pharm. 2013;441(1):343–51.
and biodegradability of polycaprolactone/chitosan and acrylic 28. Sahoo S, Sasmal A, Nanda R, Phani A, Nayak P. Synthesis of
a c i d g r a f t e d p o l y c a p r o l a c t o n e / c h i t o s a n . P o l y m e r. chitosan–polycaprolactone blend for control delivery of
2005;46(1):147–55. ofloxacin drug. Carbohydr Polym. 2010;79(1):106–13.
21. Abdolmohammadi S, Siyamak S, Ibrahim NA, Yunus WMZW, 29. Abruzzo A, Bigucci F, Cerchiara T, Cruciani F, Vitali B, Luppi
Rahman MZA, Azizi S, et al. Enhancement of mechanical and B. Mucoadhesive chitosan/gelatin films for buccal delivery of
thermal properties of polycaprolactone/chitosan blend by propranolol hydrochloride. Carbohydr Polym. 2012;87(1):581–8.
calcium carbonate nanoparticles. Int J Mol Sci. 30. Vemula SK, Veerareddy PR, Devadasu VR. Pharmacokinetics
2012;13(4):4508–22. of colon-specific pH and time-dependent flurbiprofen tablets.
22. Bauer M, Megret C, Lamure A, Lacabanne C, Fauran-Clavel Eur J Drug Metab Pharmacokinet. 2015;40(3):301–11.
MJ. Differential scanning calorimetry study of the interaction of 31. Tang M, Hu P, Huang S, Zheng Q, Yu H, He Y. Development of
antidepressant drugs, noradrenaline, and 5-hydroxytryptamine an extended-release formulation for apremilast and a level A in
with a membrane model. J Pharm Sci. 1990;79(10):897–901. vitro–in vivo correlation study in beagle dogs. Chem Pharm
23. Senda T, He Y, Inoue Y. Biodegradable blends of poly (ε- Bull. 2016;64(11):1607–15.
caprolactone) with α-chitin and chitosan: specific interactions,
thermal properties and crystallization behavior. Polym Int.
2002;51(1):33–9. Publisher’s Note Springer Nature remains neutral with
24. Sun Y, Cui F, Shi K, Wang J, Niu M, Ma R. The effect of regard to jurisdictional claims in published maps and institu-
chitosan molecular weight on the characteristics of spray-dried tional affiliations.