J Pharm Pharmaceut Sci (www.cspscanada.

org) 8(1):26-38, 2005

26
Corresponding Author: N. Udupa, Manipal College of Pharmaceuti-
cal Sciences, Manipal- 576 104, Karnataka, India.
udupa1553@yahoo.com
Formulation development, in vitro and in vivo evaluation of membrane
controlled transdermal systems of glibenclamide.
Srinivas Mutalik, Nayanabhirama Udupa
Manipal College of Pharmaceutical Sciences, Manipal, Karnataka State, India
Received 5 November 2004, Revised 12 December 2004, Accepted 15 December 2004, Published 21 January 2005
Abstract PURPOSE: The objective of the present
study was to develop the membrane controlled trans-
dermal systems of glibenclamide and to evaluate with
respect o various in vitro and in vivo parameters.
METHODS: The membrane moderated transdermal
systems were prepared using drug containing carbopol
gel as reservoir and ethyl cellulose, Eudragit RS-100,
Eudragit RL-100 and Ethylene vinyl acetate (EVA)
(2%, 9% and 19% vinyl acetate content) rate control-
ling membranes. The possible interaction between
drug and polymer was studied by IR spectroscopy,
DSC and HPTLC analysis. The formulations were
subjected to various physicochemical studies, in vitro
drug release studies and permeation studies through
mouse skin. The blood glucose reducing hypoglycemic
activity of the systems was studied in both normal and
diabetic mice. Various biochemical parameters and his-
topathological studies were carried out after treating
the diabetic mice for 6 weeks. Skin irritation tests, oral
glucose tolerance test and pharmacokinetic evaluations
were carried out in mice. RESULTS: The results sug-
gested no interaction between drug and polymer. Vari-
ations in drug release/permeation profiles among the
formulations containing different rate controlling
membranes were observed. The scanning electron
microscopic studies of EVA membranes demonstrated
no changes in the surface morphology after in vitro
skin permeation studies. The system with EVA rate
controlling membrane (with 19% vinyl acetate) was
selected for in vivo experiments. The transdermal sys-
tem produced better improvement with respect to
hypoglycemic activity, glucose tolerance test, all the
tested biochemical, histopathological and pharmacoki-
netic parameters compared to oral administration, and
exhibited negligible skin irritation. CONCLUSION:
The present study shows that membrane controlled
transdermal systems of glibenclamide exhibited better
control of hyperglycemia and more effectively
reversed the diabetes mellitus complications than oral
glibenclamide administration in mice.
INTRODUCTION
Diabetes mellitus is a chronic metabolic disorder char-
acterized by high blood glucose concentration-hyperg-
lycemia-caused by insulin deficiency, often combined
with insulin resistance (1). Glibenclamide, an impor-
tant drug of sulfonylurea class, is currently available
for treating hyperglycemia in (Non-Insulin Dependent
Diabetes Mellitus (NIDDM); but has been associated
with severe and sometimes fatal hypoglycemia and gas-
tric disturbances like nausea, vomiting, heartburn,
anorexia and increased appetite after oral therapy (2).
Since these drugs are usually intended to be taken for a
long period, patient compliance is also very important
(3). We already have reported the feasibility of applica-
tion of transdermal delivery for glibenclamide. Glib-
enclamide (molecular weight: 494 and pK
a
: 5.3) showed
favorable partition coefficients (log octanol/buffer:
0.32 0.0.07; log isopropylmyristate/buffer: 0.50
0.05) and negligible skin degradation (4, 5). In another
study, we reported the formulation and evaluation of
matrix type transdermal patches of glibenclamide (6).
It is highly accepted that membrane controlled trans-
dermal systems have the distinct advantage that the
drug release rate, which is regulated by permeation
through the rate controlling membrane, remains rela-
tively constant as long as drug loading in the reservoir
is maintained at a high level (7). Hence in the present
study, we have formulated the membrane moderated
transdermal systems of glibenclamide using carbopol
gel as drug reservoir and various polymeric rate con-
trolling membranes prepared by ethyl cellulose,
Eudragit RL-100, Eudragit RS-100, ethylene vinyl ace-
tate (containing 2%, 9% and 19% vinyl acetate) and
evaluated with respect to various in vitro parameters
(physical characteristics like thickness, drug content,
moisture content/uptake, scanning electron micros-
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
27
copy, flatness, in-vitro release/permeation kinetics,
etc) and pharmacological, biochemical and histopatho-
logical effects in vivo in mouse model.
MATERIALS AND METHODS
Ethyl cellulose (EC; with an ethoxy content of 47.5-
53.5% by weight and a viscosity of 14 cps in a 5% w/w,
80:20 toluene:ethanol solution at 25 C) was pur-
chased from SD Fine Chemicals Ltd., India. Eudragit
RL-100 (ERL) and Eudragit RS-100 (ERS) were
obtained from Rohm Pharma, Germany. Carbopol
934P NF was purchased from B.F. Goodrich, Ger-
many. Di-n-Butylphthalate was procured from Ranb-
axy Laboratories, India. Ethylene vinyl acetate (EVA)
membranes with 2% vinyl acetate (VA) content
(EVA2%; 3M CoTran 9726), 9% VA content (EVA9%;
3M CoTran 9702) and 19% VA content (EVA19%; 3M
CoTran 9715), backing layer (a polyester film lami-
nate; 3M Scotchpak Backing 1006) and release liner (a
fluropolymer coated polyester film; 3M Scotchpak
1022 Release Liner) were gift samples from 3M Phar-
maceuticals, USA. Sodium deoxycholate, anthrone,
thiourea, streptozotocin, bovine serum albumin were
purchased from Sigma Chemical Company, USA.
Polyisobutylene was purchased from Aldrich, USA.
Glibenclamide was a gift from BAL Pharma, Modi-
Mundi Pharma and Wallace Pharmaceuticals, India.
All the other chemicals used were of analytical/reagent
grade.
DEVELOPMENT OF MEMBRANE CONTROLLED
TRANSDERMAL SYSTEMS
The transdermal systems were fabricated by encapsu-
lating the drug reservoir within a shallow compart-
ment molded from a drug impermeable backing
laminate and a rate controlling membrane. EC, ERL
and ERS rate controlling membranes were prepared by
dissolving 250, 300 and 300 mg of respective polymers
in 5 ml chloroform. Di-n-Butyl phthalate (30% w/w of
polymer) was used as plasticizer. The polymeric solu-
tion was poured on the mercury surface (25 cm
2
) and
dried at room temperature. After 24 h, the films were
cut into 12 cm
2
area. EVA rate controlling membranes
were gift samples from 3M Pharmaceuticals, USA.
The reservoir (0.5% carbopol gel) of the drug was pre-
pared as per the formula given in Table 1.
Carbopol was soaked in 5 ml water and neutralized
using triethanolamine (q.s.) to form a gel. Drug in 5 ml
ethanol was added slowly to carbopol gel with con-
stant stirring.
Table 1: Reservoir of the glibenclamide membrane
controlled transdermal systems. The area of the
transdermal system was 12 cm
2.
Accurately weighed quantity of the gel (1 g) containing
drug (12 mg of glibenclamide) was placed on a sheet of
backing layer (3M Scotchpak Backing 1006) covering 3
cm x 4 cm areas. A rate controlling membrane was
placed over the gel and the edges of 3 cm x 4 cm area
were heat-sealed to obtain a leak proof device. To
ensure intimate contact of the patch to the skin, a pres-
sure sensitive adhesive, polyisobutylene (PIB), was
applied onto rate controlling membrane (3 ml; 10% w/
v in petroleum ether). A release liner (3M Scotchpak
1022 Release Liner) was placed over the adhesive
coated rate controlling membrane.
DRUG-POLYMER INTERACTION STUDIES
Infra-red (IR) spectroscopy (using IR-Spectrophotome-
ter; FTIR-8300, Shimadzu, Japan; by KBr pellet
method), differential scanning calorimetry (DSC) (Per-
kin Elmer, USA; at scanning rate of 10 C/min
between 50 and 300 C), and high performance thin
layer chromatographic (HPTLC) analysis (using
CAMAG-HPTLC system, Switzerland) were carried
out on pure substances and their physical mixtures to
search the possible interaction between glibenclamide
and carbopol (6).
IN VITRO EVALUATION OF TRANSDERMAL SYSTEMS
For drug content determination, the whole contents of
transdermal systems (n=3) were taken into a 100 ml
volumetric flask and dissolved in methanol. The solu-
tion was filtered through 0.45- membrane (Nulge
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
28
Nunc, UK) prior to drug analysis. The viscosity of
drug containing carbopol gel was determined using
Brookefield synchro electric viscometer (Brookefield
Engineering Ltd., USA). The TD bar spindle of LV-4
at 12 - gear was employed.
The in vitro drug release studies and in vitro skin per-
meation studies were carried out using USP basket
type dissolution apparatus (using 900 ml of phosphate
buffer pH 7.4 as dissolution medium) and vertical type
diffusion cells (using hairless mouse skin as membrane
barrier), respectively (6). The morphology of the EVA
rate controlling membranes before and after in vitro
skin permeation experiments was analyzed by scan-
ning electron microscopy (JEOL-JSM840A, Japan).
IN VIVO STUDIES
The animals used for in vivo experiments were adult
Swiss albino mice (6-8 weeks old) of either sex, weigh-
ing 25-30 g, from the Department of Radiobiology,
Kasturba Medical College, Manipal. The animals were
housed in polypropylene cages, 4 per cage, with free
access to standard laboratory diet (Lipton Feed, Mum-
bai, India) and water. They were kept at 251C and
45-55% relative humidity with a 12 h light/dark cycle.
The in vivo experimental protocol was approved by
the Institutional Animal Ethical Committee, Kasturba
Medical College, Manipal.
HYPOGLYCEMIC ACTIVITY IN NORMAL MICE
The hair on the backside of the mice was removed
with an electric hair clipper on the previous day of the
experiment. Following an overnight fast, mice were
divided into 3 groups (n=6). The mice were treated as
following:
Group I (Control) - 0.2 ml of 0.5% w/v sodium car-
boxymethyl cellulose (CMC); p.o.
Group II - Glibenclamide (5 mg/kg; p.o.). The oral
doses were given using a round tipped stainless steel
needle attached to 1 ml syringe and the dose of 5 mg/
kg was selected by conducting a series of experiments
with graded doses ranging between 1 to 10 mg/kg.
Group III - Applied with 2.5 cm
2
transdermal system
prepared with EVA19% rate controlling membrane,
containing 2.5 mg of drug in 0.5% carbopol gel.
At time intervals between 2-24 h after treatment (acute
study), blood was collected from orbital sinuses; blood
glucose levels were determined using Accutrend Alpha
Glucometer (Roche Diagnostics, Germany). In the
long-term study, the above treatments were adminis-
tered/applied once daily for 6 weeks. Blood glucose
levels were determined once in every 2 weeks in over
night fasted mice, 2 h after drug treatment as previ-
ously described.
INDUCTION OF DIABETES MELLITUS AND
HYPOGLYCEMIC ACTIVITY IN DIABETIC MICE
The overnight fasted mice were made diabetic by a sin-
gle intraperitoneal injection of streptozotocin (150
mg/kg; i.p.) dissolved in citrate buffer (3 mM; pH 4.5)
(6). Seven days later, mice with blood glucose levels
between 300-400 mg/dL were selected. The acute and
long-term hypoglycemic activity of the transdermal
patches was evaluated in overnight fasted diabetic mice
as described in above.
EFFECT ON GLUCOSE TOLERANCE
After an overnight fast, mice were divided into 3
groups (n=6). Control group was administered with
0.2 ml of CMC. Other 2 groups were administered
with glibenclamide (5 mg/kg; p.o.) or applied with
transdermal system as described in earlier experiments.
Two hours later, glucose was administered orally (2 g/
kg) to all the 3 groups. Blood samples were collected
just prior to and at 0.5, 1.0 and 2.0 h after the glucose
feeding and glucose level was determined. The percent-
age change in blood glucose was estimated in compari-
son with the control group. During the testing period,
food was not provided; but water was given ad libitum.
BIOCHEMICAL AND HISTOPATHOLOGICAL EVALUATION
At the end of the long-term treatment in the diabetic
mice, lipid profile (high-density lipoprotein-choles-
terol, triglycerides and total cholesterol), alanine tran-
saminase (ALT), aspertate transaminase (AST), urea
and creatinine levels were estimated in serum using
Auto-analyzer (Hitachi 911, Japan). Then the animals
were sacrificed and a part of the liver was processed for
glycogen estimation and total protein (6). Pieces of
liver, pancreas and stomach were subjected to histo-
pathological studies using haematoxylene and eosin (H
& E) staining.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
29
SKIN IRRITATION TEST (VISUAL AND
HISTOPATHOLOGICAL EVALUATION OF SKIN)
The mice were divided into 5 groups (n=6). On the
previous day of the experiment, the hair on the back-
side area of mice was removed. The animals of group I
was served as normal, without any treatment. One
group of animals (Group II, control) was applied with
marketed adhesive tape (official adhesive tape in USP).
Transdermal systems (blank, without drug and drug
loaded) were applied onto nude skin of animals of III
and IV groups. A 0.8% v/v aqueous solution of forma-
lin was applied as a standard irritant (Group V). The
animals were applied with new patch/formalin solu-
tion each day upto 7 days and finally the application
sites were graded according to a visual scoring scale,
always by the same investigator. The erythema scale
was as follows: 0, none; 1, slight; 2, well defined; 3,
moderate; and 4, scar formation. The edema scale was:
0, none; 1, slight; 2, well defined; 3, moderate; and 4,
severe. After visual evaluation of skin irritation, the
animals were sacrificed and skin samples were pro-
cessed for histological examination (6).
PHARMACOKINETIC EVALUATION
Overnight fasted mice, whose hair was previously
removed, were divided into 2 groups (n=6) and treated
as follows.
Group I - Glibenclamide (5 mg/kg; p.o.).
Group II - Applied with 2.5 cm
2
transdermal system
prepared with EVA19% rate controlling membrane,
containing 2.5 mg of drug in 0.5% carbopol gel.
Blood samples were withdrawn at different time inter-
vals from orbital sinuses using heparinized capillaries.
Plasma was separated by centrifugation using Biofuge-
13 (Heraeus Instruments, Germany) and stored in vials
at -70 C until further analysis.
ANALYSIS OF GLIBENCLAMIDE
Glibenclamide was estimated by an earlier reported
reverse phase HPLC method (27). A Shimadzu Class
VP series HPLC system with two LC-10AT pumps, a
SPD-10A variable wavelength programmable UV/Vis
detector, a SCL-10A system controller and a RP C-18
column (Luna, Phenomenex, USA; 250 mm x 4.6 mm;
particle size 5 m) was used.
The system was equipped with Class VP series version
6.12 software.
Chromatographic conditions: The mobile phase con-
sisted of 20 mM monobasic potassium dihydrogen
orthophosphate in water, which was adjusted to pH
3.5 with phosphoric acid, and acetonitrile in the pro-
portion of 60:40 v/v. The mobile phase was filtered
through 0.22 m membrane filter (Sartorius, Ger-
many). The flow rate was 1 ml/min and the column
effluent was monitored at 225 nm. The total run time
of the method was set at 20 min. The peaks were well
resolved and the retention time for glibenclamide and
glipizide (internal standard) was 9.60 and 5.96 min
respectively. No interfering peaks were observed at the
retention time of glibenclamide and glipizide.
Standard Solutions: A standard stock solution of glib-
enclamide (100 g/ml) was prepared in acetonitrile.
The calibration curve standard solutions were prepared
by adding known amount of glibenclamide (concentra-
tions: 1-20 g/ml) and glipizide, an internal standard (1
g/ml), to blank plasma.
Extraction procedure: A volume of 0.1 ml of blank
mouse plasma and 0.1 ml of 0.1 N hydrochloric acid
were mixed thoroughly. The plasma was spiked with
standard glibenclamide and glipizide solutions to yield
concentrations of 1-20 g/ml of glibenclamide and 1
g/ml of glipizide, respectively. Then the mixture was
gently shaken for 3 min and then it was added with 5
ml of benzene in a 20 ml glass tube. The tube was gen-
tly shaken using cyclomixer (Remi cyclomixer, Mum-
bai) for 5 min and centrifuged (Remi Centrifuge,
Mumbai, India) for 10 min at 3000 rpm. After centrifu-
gation, the organic phase was transferred into a conical
tube for evaporation to dryness under nitrogen. The
residue was dissolved in 0.1 ml of equilibrated mobile
phase by vortexing. An aliquot of 20 l was injected
into the chromatograph.
Calibration curve: Calibration curve was obtained by
plotting peak area ratios of glibenclamide to glipizide
(y-axis) against glibenclamide concentration (x-axis).
The pharmacokinetic parameters were calculated using
noncompartmental pharmacokinetics data analysis
software, PK Solutions 2.0.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
30
STATISTICAL ANALYSIS
The results were analyzed by Student's t-test using
Graph Pad Instat Software (Version: 1.13). Difference
below the probability level 0.05 was considered statisti-
cally significant.
RESULTS
Drug-polymer interaction studies
The IR spectral analysis of glibenclamide alone showed
that, the principal peaks were observed at wave num-
bers of 1527.50, 1157.2, 1618.2, 1714.6 and 819.7 con-
firming the purity of the drug as per established
standards (8). In the IR spectra of the physical mixture
of glibenclamide and carbopol, the major peaks of glib-
enclamide were 1527.50, 1157.2, 1618.2, 1716.5 and
819.7 wave numbers. However, some additional peaks
were observed with physical mixtures, which could be
due to the presence of polymer.
The DSC analysis (Figure 1) of pure glibenclamide
showed a sharp endotherm peak at 175.16 C corre-
sponding to its melting point.
Figure 1: DSC Thermograms. a=Carbopol; b=mixture of
Carbopol and glibenclamide; c=glibenclamide.
The DSC analysis of physical mixture of drug and car-
bopol revealed negligible change in the melting point
of glibenclamide in the presence carbopol (172.08 C
for the mixture of glibenclamide and carbopol). In
HPTLC analysis, the R
f
value of pure glibenclamide
was found to be 0.92. In the presence of polymer, the
R
f
value of the drug was unchanged and found to be
0.92.
DRUG CONTENT AND VISCOSITY
The drug content of the transdermal systems was
ranged between 99.950.12 and 99.040.13%. The vis-
cosity of reservoir of membrane controlled transder-
mal systems (0.5% carbopol gel containing
glibenclamide) was uniform among the batches and
found to be 17100150 cps.
IN VITRO DRUG RELEASE STUDIES
The results of in vitro drug release studies from trans-
dermal systems are depicted in Figure 2.
Figure 2: Cumulative percentage of glibenclamide
released in in vitro dissolution studies from reservoir
transdermal systems prepared using ethyl cellulose (EC),
Eudragit RL-100 (ERL), Eudragit RS-100 (ERS) and
ethylene vinyl acetate (EVA) copolymer rate controlling
membranes containing 2%, 9% and 19% VA (EVA2%,
EVA9% and EVA19% respectively). Each point represents
MeanSE, n=3; * significant compared to EVA2%;
Control=Suspension of glibenclamide in phosphate
buffer.
The formulations with ERL rate controlling mem-
brane exhibited the greatest cumulative percentage of
drug release value (97.459.01%) followed by ERS
(89.855.25%), EC (81.555.32 %), EVA19%
(80.028.32%), EVA9% (70.255.24%) and EVA2%
(55.657.36%) membrane containing systems at the
end of 24 h.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
31
IN VITRO SKIN PERMEATION STUDIES
The results of in vitro skin permeation of glibencla-
mide from patches are shown in Figure 3.
Figure 3: Cumulative amount of glibenclamide
permeated (mg/cm
2
) across mouse skin from reservoir
transdermal systems prepared using ethyl cellulose (EC),
Eudragit RL-100 (ERL), Eudragit RS-100 (ERS) and
ethylene vinyl acetate (EVA) copolymer rate controlling
membranes containing 2%, 9% and 19% VA (EVA2%,
EVA9% and EVA19% respectively). Each point represents
MeanSE, n=3; * significant compared to EVA2%.
The formulations (1 cm
2
) with ERL rate controlling
membrane exhibited the greatest (332.5413.36 g)
cumulative amounts of drug permeation followed by
ERS (291.5712.31 g), EC (270.2312.62 g),
EVA19% (267.2510.28 g), EVA9%
(224.5812.61g) and EVA2% (165.589.65 g) mem-
brane containing devices at the end of 24 h. The trans-
dermal systems with EVA2% rate controlling
membrane showed significantly low (p<0.05) cumula-
tive amount of drug permeation compared to other
rate controlling membrane systems.
SCANNING ELECTRON MICROSCOPY
Figure 4 shows the microstructure of different EVA
rate controlling membranes of reservoir systems before
and after the in vitro drug permeation experiments.
No considerable difference was observed in the micro-
structure of EVA films before and after in vitro perme-
ation experiments.
Figure 4: SEM photographs of EVA rate controlling
membranes A and A1= EVA2% rate controlling
membranes before and after in vitro skin permeation
studies, respectively. B and B1= EVA9% rate controlling
membranes before and after in vitro skin permeation
studies, respectively. C and C1= EVA19% rate controlling
membranes before and after in vitro skin permeation
studies, respectively.
IN VIVO STUDIES
Acute hypoglycemic activity
The results of acute hypoglycemic activity of transder-
mal system in comparison with glibenclamide (5 mg/
kg; p.o.) in both normal and diabetic mice are shown
in Table 2.
The blood glucose reducing effect was significant in
oral and transdermal patch treated animal groups upto
10 h, compared with control group (p<0.05). Glib-
enclamide (oral) produced a decrease of 39.716.81
(normal mice, p<0.05 compared to control) and
38.122.12% (diabetic mice, p<0.05 compared to dia-
betic control) in blood glucose levels at 2 h. In case of
transdermal system, the blood glucose reducing
response was gradual. A maximum blood glucose
reducing response was observed after 6 h and thereafter
remained stable upto 24 h.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
32
Table 2: Reduction in blood glucose levels after oral and
transdermal administration of glibenclamide in normal
and diabetic mice (acute study). All values are expressed
as MeanSE, n=6; CMC=Carboxymethyl cellulose; TP-
R=Reservoir transdermal system; EVA=Ethylene vinyl
acetate; GLB=Glibenclamide; * significant compared to
control (p<0.05); # significant compared to GLB
(p<0.05); significant compared to DC (Diabetic control)
(p<0.05); significant compared to GLB (p<0.05).
In orally glibenclamide treated group, the blood glu-
cose levels decreased after 6 h. The blood glucose lev-
els at the end of 24 h were only 10.585.22% and
13.254.55% in normal and diabetic mice, respec-
tively. On the other hand, the transdermal system pro-
duced significant reduction in blood glucose levels
upto 24 h compared to control (p<0.05). The
untreated group did not show any noticeable hypogly-
cemia.
LONG-TERM HYPOGLYCEMIC ACTIVITY
The results of long-term hypoglycemic activity of
transdermal system in comparison with oral glibencla-
mide in both normal and diabetic mice are shown in
Table 3. The transdermal system produced significant
(p<0.05) blood glucose reducing effect upto 6 weeks
without causing severe hypoglycemia in the initial
hours of treatment, which was observed with oral
administration.
EFFECT ON GLUCOSE TOLERANCE (GTT)
The control group showed high-elevated blood glucose
levels (p<0.05) after glucose administration
(+81.042.81, +62.053.41 and +15.014.01% at
0.5, 1.0 and 2.0 h, respectively) (Figure 5).
The hypoglycemia produced after transdermal delivery
was significantly (p<0.05) lower than the control
group. On the contrary, the orally glibenclamide
administered group showed severe hypoglycemia rang-
ing from -30.953.52 to -42.202.22% at all intervals
of the study period.
Table 3: Absolute blood glucose levels after oral and
transdermal administration of glibenclamide in long-
term study. All values are expressed as MeanSE, n=6;
CMC=Carboxymethyl cellulose; TP-R=Reservoir
transdermal system; EVA=Ethylene vinyl acetate;
GLB=Glibenclamide; * significant compared to control
(p<0.05); # significant compared to GLB (p<0.05);
significant compared to DC (Diabetic control) (p<0.05);
significant compared to GLB (p<0.05).
Figure 5: Effect on glucose tolerance after oral and
transdermal administration of glibenclamide in mice.
Each point represents MeanSE, n=6; * significant
compared to control (p<0.05); # significant compared to
GLB (p<0.05).
BIOCHEMICAL AND HISTOPATHOLOGICAL EVALUATION
The results of biochemical studies are shown in Table
4. The glycogen and total protein levels in the liver of
diabetic mice were significantly lowered compared to
normal mice (p<0.05). The oral as well as transdermal
treatment of glibenclamide significantly (p<0.05)
increased liver glycogen and total protein levels at the
end of 6 weeks. The serum lipid profile (total choles-
terol, triglycerides and high-density lipoprotein-choles-
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
33
terol), hepatic enzymes (ALT and AST), urea and
creatinine levels were significantly increased in diabetic
control mice compared to normal mice (p<0.05). The
glibenclamide treatment (both oral and transdermal)
significantly (p<0.05) reversed these changes at the
end of 6 weeks.
Table 4: Liver protein and glycogen levels and serum
lipid profile (TC, TG and HDL-C), alanine transaminase,
aspertate transaminase, urea and creatinine levels in
diabetic mice after oral and transdermal administration
of glibenclamide. All values are expressed as MeanSE,
n=6; NC=Normal control; DC=Diabetic control;
CMC=Carboxymethyl cellulose; TP-R=Reservoir
transdermal system; EVA=Ethylene vinyl acetate;
GLB=Glibenclamide; TC=Total cholesterol;
TG=Triglycerides, HDL-C= High density lipoprotein-
cholesterol, ALT=Alanine transaminase, AST=Aspertate
transaminse; # significant compared to NC (p<0.05);
*significant compared to DC (p<0.05).
The histopathological studies of liver, pancreas and
stomach from diabetic mice are presented in Table 5.
Table 5: Histopathological evaluation of liver, pancreas
and stomach from diabetic mice treated with oral and
transdermal administration of glibenclamide.
CA=Cellular atypia; Deg=Degeneration; Nec=Necrosis;
Con=Congestion; Inf=Inflammation; GLB=glibenclamide;
TP-R=Reservoir transdermal system; EVA=Ethylene vinyl
acetate. Histopathological scale: + = slight; ++ =
moderate; +++ = severe.
The liver and pancreas from untreated diabetic mouse
showed severe/moderate cellular atypia, inflammation,
necrosis, degeneration and congestion. The stomach
samples from diabetic mice showed moderate ulcer-
ation. The severe/moderate toxic manifestations
including gastric ulceration were considerably reversed
with oral but especially with transdermal administra-
tion of glibenclamide by controlling the hyperglyce-
mia.
SKIN IRRITATION TEST
The results (Table 6) showed that the prepared systems
(both blank and drug loaded) and USP adhesive tape
produced negligible erythema and edema.
Table 6: Results of skin irritation test. Visual observation
values are expressed as MeanSE, n=6; * significant
compared to formalin (p<0.05); GLB=Glibenclamide; TP-
R=Reservoir transdermal system; EVA=Ethylene vinyl
acetate; Blank=Without drug; Inf= Inflammation.
Erythema scale: 0, none; 1, slight; 2, well defined; 3,
moderate; and 4, scar formation. Edema scale: 0, none;
1, slight; 2, well defined; 3, moderate; and 4, severe.
Histopathological scale: + = slight; ++ = moderate; +++
= severe.
On the other hand, standard irritant, formalin pro-
duced severe erythema and edema. The histopathologi-
cal examination of the skin indicated that adhesive tape
and prepared patches produced mild inflammation and
edema. Formalin produced high grade of irritation,
indicated by severe inflammation and edema besides
showing discontinuity in epidermis, thin epidermis,
ulceration and hyperplasia.
PHARMACOKINETIC STUDIES
The plasma concentrations of glibenclamide after
transdermal and oral administration against time are
shown in Figure 6.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
34
Figure 6: Plasma concentration-time profile of
glibenclamide after oral and transdermal system
treatment in mice. * Significant compared to GLB-Oral
(p<0.05). Each point represents MeanSE; n=6.
Peak plasma concentration, C
max
, after oral administra-
tion was 9.130.45 g/ml and t
max
was 2.0 h. In the
case of transdermal system, the C
max
and t
max
were
6.190.14 g/ml and 12.0 h respectively. The pharma-
cokinetic parameters were calculated from the plasma
concentrations of the drug and recorded in Table 7.
DISCUSSION
In this study, membrane moderated transdermal sys-
tems containing glibenclamide were prepared using dif-
ferent rate controlling membranes. It was desired to
develop a transdermal system that allows one to pro-
vide an optimum drug release via the most appropriate
choice of rate controlling membrane and finally to
produce the desired overall constant/controlled drug
release. Ethanol (50% w/w) was incorporated in the
reservoir of the transdermal system as it significantly
enhanced the permeation rate of glibenclamide in our
earlier study (4, 5).
The pharmacokinetic parameters obtained with glib-
enclamide transdermal system were significantly differ-
ent (p<0.05) from those obtained with respective oral
glibenclamide administration, which could be due to
the rapid absorption of drugs via oral route; whereas
drug in transdermal route were slowly but continu-
ously absorbed. With respect to all in vivo experi-
ments, similar results were observed with matrix
transdermal patches in our earlier study.
Table 7: Pharmacokinetic parameters of glibenclamide
after oral and transdermal administration All values are
expressed as MeanSE, n=6; GLB=Glibenclamide; TP-
R=Reservoir transdermal system; EVA=Ethylene vinyl
acetate; C
max
=Maximum concentration; T
max
=Time of
maximum concentration; K
e
=Elimination rate constant;
AUC=Area under plasma concentration-time curve; t
1/
2
=Elimination half-life; MRT=Mean residential time. *
significant compared to oral GLB (p<0.05).
In drug-carbopol interaction studies, no distinct differ-
ence in the IR peaks and melting point (DSC analysis)
of drug in the physical mixture and R
f
values of drug
(HPTLC analysis) in the polymeric solution used in
our study indicates that the carbopol do not alter the
performance characteristics of the drug from the sys-
tems studied. All these results suggest that there is no
interaction between glibenclamide and carbopol. Good
uniformity with respect to drug content and viscosity
among the batches with all formulations was observed.
In the membrane controlled transdermal systems, drug
reservoir is encapsulated in a shallow compartment
molded from a drug impermeable backing membrane
while the drug delivery side is covered by rate control-
ling polymeric membrane. These systems have the
advantage that the drug release rate, which is con-
trolled by the rate controlling membrane, remains rela-
tively constant as long as drug loading in the reservoir
is maintained at a high level and thus provide a zero
order drug release (7). But in the in vitro dissolution
studies, the systems with ERL, ERS and EC mem-
branes did not provide a constant drug release rate and
showed high release of drug in the initial hours. This
may be due to the loss of integrity of the films by
means of solubilization of a part of the polymer by
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
35
ethanol, which was incorporated to the drug reservoir
to enhance the drug permeation rate. In addition,
direct exposure of Eudragit films to the dissolution
medium might also be responsible for the initial burst
release as they are permeable to aqueous medium (9).
The cumulative amount of drug released at the end of
24 h was depending on the hydrophilicity of the poly-
mers (ERL>ERS>EC). ERL films tend to swell more
than ERS films in aqueous medium due to the higher
concentration of hydrophilic quaternary groups. Also,
ERL membranes are more permeable to aqueous
medium than other two membranes (10). On the other
hand, the systems with EVA membranes exhibited a
constant drug release rate upto 24 h as the integrity of
these membranes was unaffected by either ethanol con-
tent of the drug reservoir or direct contact with aque-
ous medium. The release rate studies revealed that, as
the vinyl acetate content in copolymer increases, the
cumulative percentage of drug release also increases,
i.e., the membrane shows a lower resistance to the per-
meation of drug molecules. These observations are in
accordance with the earlier findings (11).
In the in vitro skin permeation studies also, the sys-
tems with ERL, ERS and EC membranes did not pro-
vide a constant drug release and showed high
permeation of drug in the initial hours. This is again
because of the ethanol content (50%) of the drug reser-
voir, which might have partly dissolved the polymeric
membranes and thereby disrupting the integrity of the
membrane (9). The ERL, ERS and EC membranes
showed loss of uniform structure after fabrication into
final system. This was confirmed by examining the
films after 24h of fabrication of final systems where the
ethanol containing gel is in direct contact with mem-
brane or after in vitro skin permeation experiments. It
was also supported by the reduced flatness of these rate
controlling membranes after preparing the final trans-
dermal devices. The films prior to fabrication into final
systems exhibited 100% flatness indicating no constric-
tion. After 24 h of fabrication, they showed reduced
flatness values (80%, 80% and 90% of flatness for ERL,
ERS and EC membranes, respectively).
These observations suggest that the films prepared by
ERL, ERS and EC lose their integral structure after
coming into contact with drug reservoir.
Hence it was thought to use those membranes, which
retain their integrity when they come in contact with
the drug reservoir. The membranes prepared by ethyl-
ene vinyl acetate copolymer (EVA) are widely used to
control the rate of drug release from many transdermal
drug delivery systems (11, 12). These membranes did
not lose their integral structure when fabricated into
final system and also after in vitro skin permeation
studies. This could be due to the ability of these films
to resist the effect of ethanol in drug reservoir. This is
also supported by the SEM studies of the EVA films
before fabricating into final system and after in vitro
skin permeation experiments, where the films main-
tained uniform and smooth surface, 100% flatness and
integrity of the structure after permeation experi-
ments. Therefore EVA rate controlling membranes are
suitable for reservoir transdermal systems in the
present study.
The cumulative amount of drug permeated at the end
of 24 h was increased as the vinyl acetate (VA) content
in the EVA rate controlling membranes was increased.
This observation is in accordance with the earlier
reports where drug permeation was increased as the
vinyl acetate content in the EVA copolymer was
increased (11,13,14). It is well recognized that it is pos-
sible to alter the permeability of EVA copolymer
membranes by varying vinyl acetate content. The
changes in the permeability have been attributed to the
alterations in the glass transition temperature and crys-
tallinity of the polymer. As the vinyl acetate content
increases, the crystallinity of the polymer decreases
rapidly and this could be the reason for high perme-
ation rate observed with EVA membrane with 19%
VA content in comparison with EVA membranes con-
taining 9% and 2% VA in the present study (15).
The cumulative amounts of glibenclamide permeated
per square centimeter of transdermal devices with
ERL, ERS and EC rate controlling membranes
through the mouse skin when plotted against time, the
permeation profiles of drug seem to follow mixed
order/apparent zero order kinetics as observed with
matrix systems (Figure 3). High amount of drug was
released in the initial hours because of increased perme-
ability of the membranes by ethanol. However, later
the drug was released depending on the concentration
and thus altering the system towards a first order pat-
tern.
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
36
The in vitro permeation profiles of formulations with
ERL, ERS and EC membranes did not fit into zero
order behavior completely and they could be best
expressed by Higuchis equation (R
2
= 0.9976 to
0.9990) (16, 17).
The data was further treated as per the following equa-
tion (18,19).
M
t
/M

=K.t
n
,
where, M
t
/M

is the fractional release of drug, M

t
is
the amount released at time t, M

is the total amount

of drug contained in the transdermal patch, t is the
release time, K is a kinetic constant and n is the diffu-
sional release exponent indicative of the operating
release mechanism. The Fickian diffusion dominated
drug release from these systems was further confirmed
by the n values (0.3897 n 0.4934) obtained by the
plots of Korsmeyer's equation. These observations
demonstrate that, if the drug release is not properly
regulated by the rate controlling membranes, the sys-
tem may exhibit diffusion predominated drug release
instead of zero order kinetics.
The reservoir systems with EVA rate controlling
membranes exhibited a zero order pattern throughout
24 h of the permeation studies (R
2
=0.9922 to 0.9995) as
the integrity of these membranes was not affected by
the drug reservoir (Figure 3). Since the concentration
of the drug in equilibrium with the inner surface of the
enclosing membrane is constant, zero order perme-
ation profile is generally expected with membrane con-
trolled systems (20). These results are in agreement
with earlier reports where membrane controlled trans-
dermal systems with EVA rate controlling membranes
showed zero order release (11,20).
The systems with ERL, ERS and EC rate controlling
membranes exhibited high drug release in the initial
hours and on the whole, the drug permeation pattern
was diffusion dominated. The membranes lost their
integrity after formulating into final systems. Hence,
further systems were developed using EVA rate con-
trolling membranes having varying VA content
(EVA2%, EVA9% and EVA19%). Among these sys-
tems, the device with EVA19% membrane exhibited
high cumulative amounts of drug permeation at the
end of 24 h. The release/permeation rate was in a con-
stant manner throughout 24 h. Thus this system was
selected for further in vivo studies.
The results of short-term blood glucose reducing effect
clearly show that the system could sustain the drug
release for a period of 24 h when compared with oral
administration where the effect declined after 6 h in
agreement with short half-life of glibenclamide (21).
Also the study clearly shows that severe hypoglycemia
associated with oral administration of glibenclamide
can be successfully overcome by membrane moderated
transdermal system. The results of long-term study
indicated that transdermal system provides optimum
blood glucose reduction upon chronic application
without producing drastic decrease in blood glucose
levels in initial time intervals. The results of GTT
show that the glucose tolerance curve was completely
inhibited in the treated groups. Transdermal route
effectively maintained the normoglycemic levels in
contrast to the oral group, which produced remarkable
hypoglycemia, an indication that a similar incident
might be prevented in diabetic patients.
In biochemical and histopathological studies, the
favourable results from glibenclamide treatment with
respect to liver glycogen, liver protein, serum lipid
profile, serum urea and serum creatinine levels could
be due to increased insulin release and peripheral
uptake of glucose after drug treatment, as explained
before
6
. The membrane moderated systems of glib-
enclamide also produced improved repair of the tissues
after diabetes induced tissue injury in comparison with
oral administration, which could be due to better con-
trol of hyperglycemia. The transdermal systems (both
drug loaded and blank) did not produce any cutaneous
reaction in skin irritation test indicating they are well
tolerated by the subjects. Similar results have been
reported by Krishna and Pandit (20)

and Kulkarni et al
(22). In the pharmacokinetic study, though the rise in
drug concentration was slower than oral administra-
tion, the drug concentration in plasma remained high
for longer period with transdermal systems. Glibencla-
mide binds to plasma proteins to the extent of 99% (6).
The prolongation of plasma half life by transdermal
systems indicate that the drug when administered by
transdermal systems will remain in the body for a
longer period and thus will exert a sustained action.
The significantly less elimination rate constants (K
e
)
and high mean residential time (MRT) values of glib-
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
37
enclamide obtained with transdermal systems further
support the sustained or slow release of drug from the
transdermal systems. Although, the C
max
was signifi-
cantly less with transdermal devices, the AUC values
were significantly high compared to oral route, which
could be due to maintenance of concentration of drug
with in the pharmacologically effective range for
longer period of time from the transdermal systems.
The significantly high AUC values observed with
transdermal devices also indicate increased bioavailabil-
ity of drug from these systems compared to oral
administration.
In the present study, with respect to the in vivo studies
conducted, membrane moderated transdermal system
of glibenclamide produced better improvement with
all the tested parameters compared to oral administra-
tion. This could be due to slow and continuous supply
of glibenclamide at a desirable rate to systemic circula-
tion by transdermal patch, which improved day-to-day
glycemic control in diabetic subjects

(23). Further, the
slow and sustained release of the drug from the trans-
dermal systems might reduce manifestations like sulfo-
nylurea receptor down regulation and the risk of
chronic hyper-insulinemia, a major risk factor for ath-
erosclerosis frequently associated with oral therapy of
glibenclamide (2,24,25). The present study shows that
membrane moderated transdermal system of glibencla-
mide exhibited better control of hyperglycemia besides
more effectively reversing the complications associated
with diabetes mellitus than oral glibenclamide adminis-
tration in mice.
In one of our earlier studies, we calculated the target
permeation rate for transdermal delivery of glibencla-
mide in man (60 kg) as 193.8 g/h based on available
pharmacokinetic data (4). In this study, the cumulative
amount of drug permeated from the system (1 cm
2
) at
the end of 24 h is 267.2510.28 g. Hence the trans-
dermal system with an area of about 18 cm
2
would be
sufficient to provide an optimum effect. But, it is well
known that human skin is less permeable compared to
mouse skin (26). However, in the view of encouraging
results obtained in mice, it can be predicted that the
required minimum effective concentration could be
achieved within an appreciable range of application
area in humans in spite of the greater barrier properties
of human skin when compared to mouse skin.
ACKNOWLEDGEMENTS
We are thankful to Council for Scientific and Indus-
trial Research (CSIR), India for providing Senior
Research Fellowship to one of the authors (Dr. Srini-
vas Mutalik). We are thankful to Ms. Chetana, Ms.
Sulochana and Mr. Subramanian for help and coopera-
tion, Dr. G.C. Jagetia (Head, Department of Radiobi-
ology, KMC, Manipal) and Dr. P. Uma Devi (Head,
Department of Research, J. N. Cancer Hospital, Bho-
pal) for providing some laboratory facilities, and Dr.
Sharath Kumar, Pathologist, District Hospital, Udupi
for helping in histopathological studies. We are
indebted to Bal Pharma, Modi-Mundi Pharma and
Wallace Pharmaceuticals, India for providing glibencla-
mide as a gift sample.
REFERENCES
[1] Nolte, M.S., Karam, J.H., Pancreatic hormones and
antidiabetic drugs, in Katzung BG (eds), Basic and
clinical pharmacology. 8th ed., Lange Medical Books/
McGraw-Hill Publishing Division, New York, pp 711-
734, 2001
[2] Davis, S.N., Granner D.K., Insulin, oral hypoglyce-
mic agents, and the pharmacotherapy of the endo-
crine pancreas, in Hardman JG: Limbird LE (eds),
The pharmacological basis of therapeutics. 9th ed.,
McGraw-Hill Co., New York, pp 1487-1517, 1996
[3] Takahshi, Y., Furuya, K., Iwata, M., Onishi, H.,
Machida, Y. and Shirotake, S., Trial for transdermal
administration of sulfonylureas. Yakugaku Zassi, 12:
1022-1027, 1997
[4] Mutalik, S and Udupa, N., Transdermal delivery of
glibenclamide and glipizide: In vitro permeation stud-
ies through mouse skin. Pharmazie, 12: 838-841, 2002
[5] Mutalik, S and Udupa, N., Effect of some penetra-
tion enhancers on the permeation of glibenclamide
and glipizide through mouse skin. Pharmazie, 12:
891-894, 2003
[6] Mutalik, S and Udupa, N., Glibenclamide transder-
mal patches: Physicochemical, pharmacodynamic,
and pharmacokinetics evaluations. J Pharm Sci, :
1577-1594, 2004
[7] Guy, R.H, Hadgraft, J., Selection of dug candidates
for transdermal drug delivery, in Guy RH: Hadgraft
J (eds), Transdermal drug delivery Developmental
Issues and Research Initiatives, Marcel Dekker, New
York, pp 59-81, 1989
[8] Moffat, A.C., Clarke's Isolation and Identification of
Drugs, 2nd ed., The Pharmaceutical Press, London,
1986
J Pharm Pharmaceut Sci (www.cspscanada.org) 8(1):26-38, 2005
38
[9] Kibbe, H.A., Hand Book of Pharmaceutical Excipi-
ents, 3rd ed., American Pharmaceutical Association
and The Pharmaceutical Press, London, 2000.
[10] Devi, K., Saisivam, S., Maria, G.R. and Deepti, P.U.,
Design and evaluation of matrix diffusion controlled
transdermal patches of verapamil hydrochloride.
Drug Dev Ind Pharm, 5: 495-503, 2003.
[11] Ocak. F. and Agabeyoglu, I., Development of a mem-
brane controlled transdermal therapeutic system con-
taining isosorbide dinitrate. Int J Pharm, 180: 177-183,
1991.
[12] Friend, D.R., Catz, P., Heller, J. and Okagaki, M.,
Transdermal delivery of levonorgestrel. V. Prepara-
tion of devices and evaluation in vitro. Pharm Res, 11:
938-944, 1989
[13] Kagayama, A., Mustafa, R., Akoho, E., Khawam, N.,
Truelove, J. and Hussain, A., Mechanism of diffusion
of compounds through ethylene vinyl acetate copoly-
mers I. Kinetics of diffusion of 1-chloro-4-nitroben-
zene, 3-4-dimethylphenol and 4-hexylresorcinol. Int J
Pharm, 18: 247-258, 1984
[14] Peterson, T., Burton, S., Englund, B. and Grosh, S.,
In vitro permeability of poly(ethylene vinyl acetate)
and microporous polyethylene membranes. Proc Int
Symp Control Release Bioactive Mater, 17: 411-412,
1990
[15] Baker, R.W., Heller, J., Materials selection for trans-
dermal drug delivery systems, in Guy RH: Hadgraft J
(eds), Transdermal drug delivery Developmental
Issues and Research Initiatives, Marcel Dekker, New
York, pp 293-311, 1989.
[16] Chien, Y.W., Transdermal therapeutic systems, in
Robinson JR: Lee VHL (eds), Controlled drug deliv-
ery fundamentals and applications, 2nd ed., Marcel
Dekker, New York, pp 524-552, 1987
[17] Higuchi, T., Mechanism of sustained action medica-
tion-Theoretical analysis of rate release of solid drugs
dispersed in solid matrices. J Pharm Sci, 52: 1145-
1149, 1963
[18] Korsmeyer, R.W., Gurny, R., Doelker, E., Buri, P.
and Peppas, N.A., Mechanism of solute release from
porous hydrophilic polymers. Int J Pharm, 15: 25-35,
1983
[19] Guyot, M. and Fawaz, F., Design and in vitro evalua-
tion of adhesive matrix for transdermal delivery of
propranolol. Int J Pharm, 204: 171-182, 2000
[20] Krishna, R. and Pandit, J.K., Transdermal delivery of
propranolol. Drug Dev Ind Pharm, 15: 2459-2465,
1994
[21] Christ,,O.E., Heptner, W. and Rupp, W., Pharmaco-
kinetics of a new highly effective sulfonylurea deriva-
tive. Acta Diabetol Lat, 1: 101-115, 1969
[22] Kulkarni, R.V., Mutalik, S. and Hiremath, D., Effect
of plasticizers on the permeability and mechanical
properties of Eudragit films for transdermal applica-
tion. Indian J Pharm Sci, 1: 28-31, 2002
[23] DCCT Research Group., The effect of intensive
treatment of diabetes on the development and pro-
gression of long-term complications in insulin depen-
dent diabetes mellitus. N Engl J Med, 329: 977-986,
1993
[24] Faber, O.K., Beck-Neilsen, H. and Binder, C., Acute
actions of sulfonylurea drugs during long-term treat-
ment of non-insulin dependent diabetes mellitus. Dia-
betes Care, 13: 26-31, 1990.
[25] Bitzen, P.O., Melander, A. and Schersten, B., Long-
term effects of glipizide on insulin secretion and
blood glucose control in patients with non-insulin
dependent diabetes mellitus. Euro J Clin Pharmacol,
42: 77-82, 1992
[26] Catz, P. and Friend, D.R., Transdermal delivery of
levonorgestrel. VIII. Effect of enhancers on rat skin,
hairless mouse skin, hairless guinea pig skin, and
human skin. Int J Pharm, 58: 93-102, 1990
[27] Emilsson, H., Sjoberg, S., Svedner, M. and Christen-
son, I., High performance liquid chromatographic
determination of glibenclamide in human plasma and
urine. J Chromatography, 383: 93-102, 1986