Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922

Contents lists available at SciVerse ScienceDirect

Colloids and Surfaces B: Biointerfaces

journal homepage:www.elsevier.com/locate/colsurfb

Preformulation studies of mucoadhesive tablets for carbamazepine sublingual

administration
∗
Luana Perioli , Cinzia Pagano
Dipartimento di Chimica e Tecnologia del Farmaco, Università degli Studi di Perugia, Via del Liceo 1, Perugia 06123, Italy

article info abstract

Article history: The purpose of this research work was the realization of a bi-layered mucoadhesive dosage form intended for carbamazepine
Received 3 February 2012 sublingual administration and planned in order to obtain a unidirectional drug release and diffusion only across buccal mucosa
Received in revised form 1 October 2012 avoiding the liberation in the buccal environment.
Accepted 3 October 2012 Bi-layered tablets were constituted by an impermeable ethyl cellulose backing layer and a mucoad-hesive layer. The latter
Available online 10 October 2012
was composed by a blend of a semisynthetic polymer, as hydroxyethyl cellulose or hydroxypropyl methylcellulose, and a
®
synthetic polymer as, Carbopol , physically mixed in different ratios. The active ingredient carbamazepine was
Keywords:
Carbamazepine
homogeneously dispersed in the mucoadhesive layer. The prepared formulations were carefully characterized by thickness,
Mucoadhesion friability, swelling index, matrix erosion, ex vivo and in vivo mucoadhesive force and time, moreover patient acceptability
Polymers was evaluated as well.
Buccal tablets
® ®
Release Tablets constituted by Carbopol :hydroxypropyl methylcellulose (25%:75%) and Carbopol : hydroxyethyl cellulose
(75%:25%) showed the best properties and for this reason were submitted to in vitro release studies.

Both tablet groups gave good results in terms of ex vivo and in vivo bioadhesive force and time giving a sustained release.

© 2012 Elsevier B.V. All rights reserved.

1. Introduction
Carbamazepine (CARBA) is an antiepileptic and psychotropic drug
mainly used in the treatment of secondarily generalized tonic-clonic and
partial seizures, trigeminal neuralgia and in the prophylaxis of unresponsive
lithium maniac depression (bipolar disorder) [1]. According to Essential
Medicines WHO Model List, Core List, CARBA is labeled as class II drug of
the Biopharmaceutics Classification System (BCS) [2,3] because of its poor
solubility. This aspect is important as it affects drug absorption that results
slow and irregular through the GI tract reaching a lag time of 4–8 h [4,5]; the
main consequence of this is the variable CARBA bioavailability.
CARBA is currently administered by oral route and because its low water
◦ administered and could be rapidly removed if required. However, a series of
solubility (170 g/ml at 25 C) high and frequent doses are required in order to
obtain the therapeutic effect. For exam-ple, the suggested initial dose for
epilepsy therapy is 100–200 mg once or twice daily. It is gradually increased
by increments of 100–200 mg every 2 weeks to an usual maintenance dose of
0.8–1.2 g daily, up to 2 g daily may be occasionally necessary. Usu-ally,
CARBA must be administered 2–4 times daily [1] however,
∗ as: (i) a close contact between the mucosa and the dosage form, (ii) a high
Corresponding author. Tel.: +39 075 5855133; fax: +39 075 5855163. E-mail address:
luanaper@unipg.it (L. Perioli).
this aspect represents a problem for the maintenance of therapeu-tic plasmatic
levels as well as for patient compliance, in particular for pediatric people.
The design and development of formulations for buccal admin-istration
could be an interesting approach in CARBA therapeutic regimen
improvement. In fact, CARBA administration by buccal route was estimated
by many researchers to be a suitable approach to obtain a more efficient
systemic absorption on such drugs [6,7].
Oral mucosa represents an interesting drug administration site as it offers
several advantages over both injectable and oral meth-ods. Because of the
high vascularization of this site, drugs absorbed through the oral mucosa can
enter directly into the systemic circu-lation bypassing the GI tract and first-
pass metabolism in the liver. Moreover buccal devices can be easily
factors could limit the drug absorption across the oral mucosa, this include:
exposure to salivary flow, low absorption area, production of shearing forces
due to tongue movements, swallowing as well as short residence time [8,9].
Since 1980s a great interest was oriented on buccal formula-tions realized
by using mucoadhesive polymers [10,11]; in fact, these systems can offer
numerous advantages in comparison to non-adhesive solid oral dosage forms
drug concentration

0927-7765/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.colsurfb.2012.10.001
916 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922
maintenance at the absorption surface for a prolonged period of time, (iii) the
dosage form immobilization in a specific part of the oral mucosa, (iv) drug
protection from the environment [8,12].
The oral mucosa can be distinguished in four potential regions suitable for
drug delivery: buccal, sublingual, palatal, and gingival [13]. Among them, the
sublingual region can represent an inter-esting administration site as it has
some characteristics useful to improve drug bioavailability in the case of
active ingredients presenting unsatisfactory biopharmaceutical properties.
Thin and highly permeable membrane of sublingual tissue is a perfect tar-get
if a prompt onset is desired. Considerable surface area and high blood flow at
this region provide a means for rapid access to the sys-temic circulation.
Moreover, sublingual region is easily accessible, and generally well accepted
from the patient [14].
The present research work deals with the realization and char-acterization
of simple bi-layered mucoadhesive tablets intended for CARBA sublingual
administration.
2. Experimental
2.1. Materials
®
Hydroxyethyl cellulose HEC (Natrosol 250 HX PHARM) viscos-ity
3400–5000 mPa sec in water solution 1% (w/v), hydroxypropyl
®
methylcellulose HPMC (Benecel MP 844 PHARM) viscosity 29,000–
® Ex vivo mucoadhesion time studies were performed (in tripli-cate) after
43,000 mPa sec in water solution 2% (w/v) and ethylcel-lulose EC (Aqualon
-T10) were purchased from Aqualon Hercules Inc., (Wilmington, Delaware,
® ®
U.S.A.). Carbopol (Carbomer 940) viscosity 68,500 mPa sec in water
dispersion 0.5% (w/v) was pur-chased from Galeno (Firenze, Italy).
Carbamazepine was purchased from Sigma–Aldrich GmbH (Steinheim,
Germany). Deionized water was obtained by reverse osmosis process with a
MilliQ system (Mil-lipore, Roma, Italy). Other chemicals were of reagent
grade and were used without further purification.
Simulated salivary fluid (SSF) at pH 6.4 was prepared according to
Farmacopea Ufficiale Italiana (F.U. XII Ed.). One liter of SSF had the
following composition: Na2 HPO4 (1.79 g) KH2 PO4 (1.36 g), NaCl (7.02 g)
deionized water (1000 ml).
2.2. Tablet manufacturing
Bilayered tablets were prepared in 2 stages using a 12-mm-diameter die,
by a manual tableting press (PerkinElmer, Cambridge, UK) using a
3
compression force of 1.5 × 10 kg for a total time of 30 s. Each tablet layer
was prepared starting from a single powder (in the case of backing layer) or a
blend, gently mixed by mortar and pestle in the case of the mucoadhesive
layer.
Firstly the backing layer was prepared by EC (100 mg) compres-sion for
15 s, then the punch was lifted and 100 mg of the second mucoadhesive layer
blend, mixed to 50 mg of CARBA, was added to the first preformed layer and
compressed for 15 s, obtaining the bi-layered tablet. Tablet thickness was
measured (n = 10) by a micrometer (Borletti, Cremona, Italy).
Drug free tablets for in vivo mucoadhesion studies have been also
prepared following the same procedure described above (with-out CARBA
addition in the mucoadhesive layer).
2.3. Physical characterization: friability
Friability was determined according to F.U. (XII Ed.) by sub-mitting 20
previously weighed tablets to falling shocks for 4 min in a friabilator (Erweka
TA 200), set at 25 rpm/min. After 4 min, the tablets were reweighed and the
percentage friability was calculated.
2.4. Swelling studies
Tablet swelling properties (hydration %) and matrix erosion or dissolution
(DS), were measured following a procedure previously reported [15,16]. Each
tablet was weighed (W1) and put into a petri plate (9 cm diameter) containing
◦
SSF (F.U. XII Ed.) then ther-mostated at 37.0 ± 0.1 C in a ventilate heater
(Orbital Incubator, Sanyo Gallenkamp, Japan). The experiments were
performed in triplicate and data were calculated using Eqs. (1) and (2):
= W2 ×
% of Hydration (W 2 − W 1) 100 (1)
= W1 ×
W1−W3
DS 100 (2)
where W1 is the dry tablet weight, W2 is the tablet weight after immersion in
SSF for predetermined time intervals (0.5, 1, 2, 4, 8 and 12 h). W2 was
measured after that SSF excess removing from the tablets using filter paper.
◦
W3 is the weight of the swollen tablet measured after its drying at 60 C for
24 h and then placed for 48 h in a desiccator containing CaCl 2 . This
experiment was performed in triplicate and expressed as mean ± SD (error
bars are not reported in the graphs in order to allow a clear analysis of each
profile).
2.5. Ex vivo mucoadhesion time
tablet application on freshly cut porcine buccal mucosa. Pig buccal mucosa,
obtained from Large White pigs, weighing ∼165–175 Kg, was furnished by
Veterinary Service of ASL N.1 Città di Castello (Umbria, Italy) and used
within 12 h from pig sacrifice. The porcine buccal tissues (2 cm × 2 cm) were
fixed in the internal side of a beaker by cyanoacrylate glue. A side of each
tablet was wetted with 50 L of SSF and then attached to the porcine buccal
tissue by applying a light force (∼0.5 N), by a fingertip, for 20 s. Then, the
◦
beaker was filled with 1800 mL of SSF and kept at 37.0 ± 0.1 C; after 2 min
a stirring rate of 150 rpm was applied to simulate the exposure to salivary
flow in the buccal cavity. Tablet behavior and mucoadhesive time were
monitored until that complete detach-ment or matrix erosion occurred.
2.6. Ex vivo mucoadhesion force
The ex vivo adhesion force was assessed by a dynamome-ter (Didatronic,
Italy) using the abovementioned porcine buccal mucosa. The tablets (baking
layer) were attached on a support, con-nected to the dynamometer, using
cyanoacrylate glue. A piece of porcine buccal mucosa (2 cm × 2 cm) was
glued onto a support and kept in a vessel placed in a thermostatic bath at 37.0
◦
± 0.1 C. The free side of the tablet was wetted with 50 L of SSF and attached
to porcine buccal tissues by applying a light force with a fingertip for 20 s.
◦
The vessel was filled with SSF and kept at 37.0 ± 0.1 C. The measurements
started after 2 min. The maximum adhesive forces were the average of three
measurements (n = 3) and the confidence interval was determined at a 0.05
significance level.
2.7. In vivo evaluation of organoleptic properties and
acceptability
The followed procedures were in accordance with the
ethical standards of the responsible Committee on human
experimenta-tion (regional) and with the Helsinki Declaration.
The studies were performed (in triplicate) by applying drug
free tablets on sublin-gual mucosa of 3 healthy volunteers in
order to assess the residence time, the organoleptic
characteristics, the fragment loss, the sali-vary level variation,
and the possible production of irritation or
 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922
pain. Each tablet was attached to the sublingual mucosa by applying a light
force with a fingertip for 20 s.
2.8. Stability studies
CARBA bi-layered tablets were put into a glass vial, then placed at 40 ± 2
◦
C and 75% RH ± 5% and then examined after 1, 3 and 6 months.
2.9. In vitro release studies
F.U. (XII Ed.) standard paddle apparatus, properly modified, was used to
evaluate drug release by using SSF added by 1% of sodium lauryl sulfate
◦ polymers and polymers with different characteris-tics, as cellulose derivatives,
(SLS) as dissolution medium [17–19] thermostated at 37.0 ± 0.1
dissolution test was performed by two different methods: A and B.
2.9.1. Method A [20]
The tablet backing layer was attached by cyanoacrylate glue at a height of
7.0 cm to the inside of the glass dissolution vessel then filled with 1000 ml
◦ The bi-layered tablets were prepared by direct compression (DC) in two
the dissolution medium maintained at 37.0 ± 0.1 C. Paddle rotation was set
to 150 rpm.
2.9.2. Method B [21]
The tablet backing layer was fixed by cyanoacrylate glue to a support and
positioned on the bottom of the vessel then filled with 1000 ml of the
◦
dissolution medium maintained at 37.0 ± 0.1 C. Pad-dle rotation was set to
50 rpm.
In both cases four-milliliter samples were collected at pre-determined time
intervals and replaced with an equal volume of fresh SSF. The samples,
collected at established times, were fil-tered through a cellulose membrane
(Filter paper Whatman 41, Whatman GmbH, Dassel, Germany) and drug
concentration was determined by a UV-vis spectrophotometer (Agilent mod.
8453). Drug calibration curve in SSF added by SLS 1% was previously pre-
pared ( max = 285.0 nm, r = 0.9998). Tests were made in triplicate and results
were registered as an average ± SD.
3. Results and discussions
3.1. Tablet preparation and characterization
The design of mucoadhesive buccal tablets must take into consideration
some important points in order to meet patient compliance. The main
requirements are: low size, flexibility and compatibility with mucosal surface,
pain free, good taste, easy application and easy removal (if necessary).
On the basis of these characteristics, bi-layered mucoadhesive buccal
tablets intended for sublingual mucosa application have been designed and
developed. These formulations were planned in order to promote drug release
only from the side in contact to the sublingual mucosal surface avoiding the
release in sali-vary fluids. For this reason tablets were provided for two
layers: a mucoadhesive layer, constituted by a simple blend of two mucoad-
hesive polymers, in which the drug is dispersed, and a backing layer (Fig. 1).
In particular, the mucoadhesive layer (100 mg) was prepared by mixing a
semisynthetic polymer (cellulose deriva-tive) with a synthetic polymer
(polyacrylic derivative) combined to CARBA (50 mg), whereas the backing
layer was realized by EC (100 mg). The cellulose derivative combination to a
polyacrylic derivative comes from the necessity to produce a formulation
showing good mucoadhesion performances and able to guarantee, at the same
time, a high patient acceptability. It is noteworthy in fact, that polyacrylic
derivatives alone possess better bioadhesive properties than cellulose
derivatives as the high carboxylic groups
917
Fig. 1. Schematic representation of swellable by-layered tablet.
concentration makes them able to adhere strongly to mucosal sur-face
provoking sometimes pain. The employment of a mixture of polyacrylic
C. The can mitigate the strong interaction with mucosal surface obtaining a good
mucoadhesion and avoid-ing pain for the patient. Among the synthetic
® ®
polymers available, Carbomer 940 (Carbopol ) was the polyacrylic
derivative chosen for tablet preparation while HEC and HPMC were selected
among the cellulose derivative polymers; nine kinds of tablets of different
compositions have been prepared (Table 1).
different steps. DC represents an efficient and easier process, if compared to
others, because it involves only dry blend-ing and compaction of drug and
polymers. Additionally, DC is a cheap process, requires low processing times,
low labor costs, few manufacturing steps, low number of equipments, low
validation and reduced power consumption, moreover, heat and moisture are
avoided. Although DC simplicity and cost-effectiveness, this process is highly
influenced by powder characteristics thus, the selection of suitable excipients
is one of the crucial steps that could make a success in DC processing. In fact,
powder properties can influence greatly blend compaction behavior and tablet
perform-ances thus, the excipients chosen for tablets preparation by DC must
be manageable and machinable.
In regard to the prepared tablets (Table 1), tablets 1–6 were the
formulations showing high manageability and machinability while in the case
of tablets 8 and 9 (Table 1) some problems were encoun-tered. In particular,
tablet 8 was characterized by sticking while 9 underwent to capping during its
ejection from the die, for these reasons, these two tablets were discarder from
the study. Pow-ders behavior during compression can be explained analyzing
the compositions reported in Table 1. Polymers blend displaying high
manageability and machinability were those prepared by HPMC alone (tablet
®
1) or mixed to Carbopol (tablets 2, 3 and 4). In fact, t is noteworthy, that
HPMC is an excipient largely used in DC due to its high compactability and
good flow properties while the bad tablet-ing properties in the case of
polymeric mixtures used for tablets 8 and 9 could be attributable to HEC poor
flow properties.
The polymeric composition must be chosen in order to obtain final tablets
with possessing mechanical properties able to confer resistance to capping,
abrasion or breakage under storage condi-tions, transport and handling before
usage. Thus, this aspect was evaluated by submitting tablets 1–7 to friability
test. The weight loss resulted lower than 1% (0–0.17%) for all tablets meaning
that all the polymeric blends used confer good compactness, mechani-cal
resistance and low friability to final tablets. This is an important aspect as
tablets must keep their original shape and characteris-tics during their
packaging, shelf-life until their removal from the blister and the application in
the buccal cavity.
The adhesion of a mucoadhesive formulation to mucosal surface is the
results of a series of events involving the mucoadhesive poly-mers the mucin,
the main component of the gelled layer (mucus) covering the mucosa. Once in
contact with salivary fluids, the mucoadhesive layer swells generating a gelled
When a polymer is immersed into a fluid, water
layer (Fig. 2).
penetrates into the
918 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922
Table 1

Tablet compositions.

Tablet
a Mucoadhesive layer Backing layer EC (mg) Thickness (mm ± SD) (n = 10, ˛ = 0.05)
® HEC (mg) HPMC (mg)
Carbopol 940 (mg)
1 0 0 100 100 1.503 ± 0.51
2 25 0 75 100 1.393 ± 0.51
3 50 0 50 100 1.380 ± 0.88
4 75 0 25 100 1.353 ± 0.51
5 100 0 0 100 1.333 ± 0.33
6 75 25 0 100 1.373 ± 0.22
7 50 50 0 100 1.380 ± 0.00
8 25 75 0 100 Not measured
9 0 100 0 100 Not measured
a
For each tablet the mucoadhesive layer (100 mg) was added by 50 mg of CARBA

free spaces of the surface among the macromolecular chains. As water content
increases, the polymer becomes hydrated forming a tridimensional network in
which water molecules are entrapped, generating a gel-like structure. The
swelling capacity is very impor-tant as it increases the polymer chain
conformational mobility enhancing the mucoadhesive force due to a
concomitant increase in chain diffusivity and degree of entanglement with the
possibility to expose a great number of hydrophilic groups to mucosal sur-
face. Moreover, polymer swelling degree plays an important role in
conditioning drug release kinetic. For these reasons the swelling capacity,
expressed as hydration percentage, was determined for each tablet.
hydrophilicity compared to HPMC [22]. The low hydration capac-ity
®
observed for tablet 5 is ascribable to Carbopol (100%) low water affinity
that makes it unable to bind a high amount of water molecules. Regarding
tablets 2, 3 and 4 the obtained results reveal that the increase of Carbopol
®
®
content in their composi-tion (Carbopol :HPMC ratios of 1:3, 1:1 and 3:1
tablets 2, 3 and 4 respectively) does not affect significantly the swelling
®
capacity. In the case of tablet 6 the prevalence of Carbopol on HEC (75% vs
25%) is responsible for the low hydration percentage. During the first hour
HEC chains relaxation takes place, with consequent tablet swelling, as water
penetrates into the matrix.

The highest swelling was obtained for tablets 1 and 7 that hydrated
rapidly reaching 48% and 53% respectively after 1 h and 86% and 91%
respectively after 8 h (Fig. 3). In the case of tablets 2, 3, 4 and 6 similar
hydration profiles were registered with an hydration % of ∼30% after 1 h and
∼87% after 8 h. Tablet 5 showed a reduced swelling capacity reaching 72%
after 8 h. These results could be explained considering that polymer swelling
is influenced by (a) the initial wetting processes involving tablet surface, (b)
the amount and type of polymer in tablet composition.
The wetting process is strictly connected to tablet excipient nature. As
hydrophilic polymers show high water affinity, thus polymer chains enter in
close contact with water molecules. This means that the prevalence of a
cellulose derivative (HPMC or HEC) in tablet composition will make the
formulation more hydrophilic with enhanced wetting properties.
In regard to point b, the behavior of tablets 1 and 7 could be explained
considering that cellulose derivatives HPMC (tablet 1) and HEC (tablet 7),
are able to establish numerous bonds with water molecules generating rapidly
a gelled structure because of the large number of hydrophilic groups in their
structure. Tablet 7 (containing 50% of HEC) swelling capacity prevails
against tablet 1 (containing 100% of HPMC), despite HEC shows the higher
From the analysis of hydration curves obtained for each tablet, two
different phases could be distinguished (Fig. 3). Error bars have not been
indicated for graph clarity as these may be confusing since many profiles
overlap. The first phase is represented by 0–8 h range, in which the gradual
water content increase is evident for all the formulations. Then, in the range
8–12 h (second phase), a reduced hydration percentage is registered, justified
by tablet fragment lost. The latter was assessed by the erosion matrix assays
(DS) evaluated along 12 h. During the first 8 h all the tablets show similar
behav-iors, then DS% values increase as fragment loss occurred (Fig. 4). Also
in this case error bars have not been indicated for graph clarity as these may
be confusing since many profiles overlap. The high-est value was measured
for tablet 5 which, after 12 h, reached a DS percentage of 8.77%. These
observations are in according to the low hydration % registered for tablet 5.
®
The low Carbopol water affinity in fact, induces a slow and limited swelling
combined to low water penetration with high tendency to give fragments.
Moreover,

Fig. 2. Picture of the by-layered tablet after swelling in SSF. The backing layer is unaltered
while the mucoadhesive layer hydrated generating a gelled structure. In order to distinguish the Fig. 3. Hydration % vs time determined for all tablets in SSF at 37.0 ± 0.1 ◦ C at estab-lished
mucoadhesive layer it was colored in methylene blue to the mucoadhesive polymers mixture. times (0.5, 1, 2, 4, 8, 12 and 24 h).(n = 3, error bars are not reported for a
clear graph analysis.)
 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922 919

attributable to HPMC presence. In fact, HPMC hydrophilic character favors

◦
Fig. 4. Tablet DS% vs time determined for all tablets in SSF at 37.0 ± 0.1 C at estab-lished
times (0.5, 1, 2, 4, 8, 12 and 24 h). (n = 3, error bars are not reported for a clear graph analysis.)
®
Carbopol hydration generates a thick and viscous gel that eas-ily undergoes
to fragmentation because of hydrodynamic pressure effect, exerted on the
polymer network by water. Polymer swelling is an important process involved
in tablet adhesion to mucosal surface. In fact, this water-activated process
produces polymer swelling and improves the consolidation step that increases
the polymer chains mobility facilitating the interpenetration with the mucin
chains of the mucus layer [23].
The bioadhesion capacity of the employed polymers is attributable to the
presence of amino and carboxylic groups in their structure that make them
able to form hydrogen bonds with the mucosal components [23]. Tablet
ability to interact with buccal mucosa was firstly evaluated by ex vivo
mucoadhesion assays with the aim to measure the adhesion force and time of
tablets 1–7 (Table 2). Obtained data revealed that all tablets require a short
time to bind mucosal surface and that are able to adhere for a prolonged time
®
(Table 2). Tablet 6 (Carbopol :HEC 75%:25%) and 7 (Carbopol :HEC
® mucilage which limits tablet bioadhesion. Tablet hydration in fact, generates a
50%:50%) showed the lower mucoadhesion time in comparison to other
tablets and this can be explained by HEC presence in both tablets 6 and 7
compositions. As mentioned before, tablet swelling behavior influences the
mucoad-hesion performances. Tablet 1 (HPMC 100%) required less swelling
® 100% for tablets 4 and 5 respectively) is responsible for the high binding to
time (15 minutes) in comparison to tablets 2 (Carbopol :HPMC 25%:75%)
® ®
and 3 (Carbopol :HPMC 50%:50%, about 3 h), while tablet 4 (Carbopol
® that, associated to carboxylic groups ion-ization, is responsible for the gradual
:HPMC 75%:25%) and 5 (Carbopol 100%) didn’t swell during all the period
of contact with mucosa. Tablets 3 and 7 displayed the highest time (after 71 h
and >10 h respectively) in comparison to the remaining formulations that
were discarded from the successive studies. All the tablets maintained the bi-
layered structure for the entire experiment.
tablet swelling in a short time after the contact with the SSF because of the
high water affinity, HPMC chains bind a high number of water molecules
thus, a small amount of chemical groups remain available to interact with
®
mucin. In regard to Carbopol :HPMC blend, the 1:3 ratio (tablet 2) confers
high mucoadhesion capacity as HPMC hydrophilicity allows an easy water
®
penetration into the matrix promoting the relaxation of Carbopol chains then
avail-able to come in a very close contact with the mucus layer. Using the
®
same ratio of Carbopol blended to HPMC or HEC (tablets 3 and 7
respectively) the mucoadhesion force decreases as tablet swelling capacity is
reduced because only a limited number of groups are available for binding.
The same observation could be done for tablet 6 for which the low amount of
HEC is responsible for a limited swelling. In the case tablet 4, in which
®
Carbopol (75%) content prevails, the presence of HPMC (25%) is important
because water molecules are able to penetrate and to promote a gradual chain
enlargement. In this way numerous carboxylic groups are available to interact
with mucin chains. This promotes a close and strong contact with the mucosal
surface, as demonstrated from ex vivo mucoadhesion force data (Table 2) that
could be respon-sible for pain and irritation. In the case of tablet 5 the lack of
a cellulose derivative reduces the water access in the inner space of the
mucoadhesive layer bringing to a reduced interaction with the mucosal
surface. The characterization of tablet mucoadhesion capacity was completed
by in vivo assays having the objective to evaluate patient acceptability and
compliance. In particular, the investigated parameters were: (1) bioadhesion
capacity, (2) adhe-sion time, (3) swelling capacity and patient acceptability,
(4) pain or irritation, (5) taste, (6) fragment lost. These studies are very
important in the choice of the best formulations to submit to the fol-lowing in
vitro release studies. With this aim drug free tablets were prepared and applied
on the sublingual mucosa of three healthy volunteers. Formulation behavior
was evaluated in the elapsed time from tablet administration to the total
detachment. During this period the volunteers were able to eat and drink.
Tablet 1 did not show adhesion to mucosal surface (Table 2) and this can be
due to the high and quick water absorption due to HPMC pres-ence (as
observed in the hydration studies, Table 2) forming a slippery non-adhesive
weak gel which could be eas-ily removed from the mucosal surface. Tablets 4
and 5 showed a high adherence to mucosa, however it was so strong to
provoke pain and irritation after few minutes from the application. In both
®
cases the high percentage of Carbopol in the mucoadhesive layer (75% and
®
mucosal surface. In fact, Carbopol shows a high water absorption capacity
relaxation of the polymeric chains. In these conditions the carboxylic groups
are able to enter in close contact with the mucosal surface establishing
hydrogen

From ex vivo mucoadhesion force data analysis some consider-ations can

be drawn. For tablet 1 a low force value was registered,

Table 2
Ex vivo mucoadhesion times and forces and in vivo performances.

Ex vivo In vivo
Tablet Mucoadhesion Swelling Mucoadesion Bilayer integrity Mucoadhesion Pain after Fragment loss Taste
time (h) force (N ± SD) time (h) application
(n = 3; ˛ = 0.05)
1 >73 After 15 min 0.70 ± 0.12 Yes No adhesion – — —
2 >73 After 4 h 2.20 ± 0.27 Yes 17 No No Good
3 >73 After 71 h 1.66 ± 0.11 Yes No adhesion – – –
a
4 >73 no 5 Yes – Yes – –
5 >73 no 1.03 ± 0.45 Yes a Yes – –
–
6 44–46 After 3 h 1.66 ± 0.26 Yes 12 No – –
7 44–46 >10 h 1.73 ± 0.56 Yes No adhesion – – –
a The mucoadhesion time, fragment loss and taste have not been estimated as the table have been removed after application because painful.
920 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922

bonds with the mucus layer. In the case of tablet 4, HPMC (25%) improves
®
the mucoadhesion capacity in comparison to Carbopol alone as, because of
its lower water absorption capacity, it is able to promote a gradual swelling
encouraging the interaction with mucin chains. Good adhesion for long time
was observed for tablets 2 and 6 that were kept in the mouth for all the tests
without pain for the volunteers.

During the manufacturing process, as well as the storage con-ditions, the

dosage form is exposed to physical agents, such as humidity, that could
provoke changes of the shape and layers adherence that compromise the
efficiency of the dosage form. This aspect is also important as tablet
constituents are mucoad-hesive polymers that, in presence of water, become
sticky. Thus, In order to evaluate tablet susceptibility to humidity, stability
tests were executed following the indications proposed from the Euro-pean
Agency for the Evaluation of Medicinal Products [24]. For solid dosage
◦
forms, the guidelines prescribe to keep the samples at 40 ± 2 C and 75 ± 5%
RH for 6 months (accelerated tests). Each sample (in triplicate) was put into a
◦
glass vial and then placed at constant temperature and humidity (40 ± 2 C
and 75% RH ± 5% respectively). After 1 month each tablet was accurately
examined; tablets 1, 3 and 7 showed a partial layer separation (7 > 3 > 1),
tablets 4 and 5 became sticky, while tablets 2 and 6 resulted unmodi-fied and
◦
for this reason were kept at 40 ± 2 C and 75% RH ± 5% for further 5
months. At the end of this period these tablets resulted unchanged.

From the characterization studies performed, resulted that tablets 2 and 6

were the best candidate for CARBA sublingual administration as they adhere
very well on sublingual mucosa, for a long time without pain or irritation for
the patient. Thus, these two formulations have been submitted to further
studies and the in vitro drug release was investigated.
◦
Fig. 5. In vitro release profiles of tablets 2 and 6 in SSF added by SLS 1% at 37.0 ± 0.1 C (n =
3, mean ± SD). (A) Dissolution performed positioning the tablet on internal vessel surface at 7
cm from the bottom (Method A), (B) dissolution performed posi-tioning the tablet on vessel
bottom (Method B).

3.2. In vitro release studies

As official assays for mucoadhesive sublingual tablets are not reported,
this study was carried out by using the paddle apparatus (F.U. XII Ed.),
properly modified. In particular, in order to make the assay bio-relevant, two
different methods were followed in dissolu-tion tests considering two possible
tablet positions in the vessel as reported from Miyazaki et al. (Method A) and
Remunán˜-López et al. (Method B) [20,21]. The dissolution test was
performed (in tripli-cate) following the methods A and B described in method
section. In fact, the swelling capacity plays an important role in drug release,
connected to the formed gel properties as well as the dissolution method used.
Regarding the release profiles of tablets 2 and 6 submitted to method A
(Fig. 5A) it results that the amount of CARBA released from tablet 2 is higher
in respect to tablet 6 for the entire test. Specifically, after 8 h tablet 2 releases
35% of drug vs 17%, reaching 75% vs 68% after 24 h. The different release is
® that the round bottom of a dissolution vessel, in combi-nation to stirring
attributable to the different Carbopol amount in the two tested formulations.
®
The high amount of Carbopol in tablet 6 (75%), in fact, once in contact with
the dissolution medium, creates a viscous gel that delays and
reduces drug diffusion, moreover, the low HEC amount is respon-sible for a
limited formulation swelling. In the case of tablet 2, HPMC (75%)
predominance confers high water affinity to the for-mulation allowing an easy
water penetration into the matrix; the polymer network results enough
enlarged to promote a rapid and high drug diffusion through the gelled layer.
Also in the case of the release profiles obtained following method B, CARBA
release from tablet 2 results better than tablet 6 (Fig. 5B), however the total
amount of released drug from both formulations is lower if com-pared to the
results coming from method A. These differences can be ascribed to (i) the
different hydrodynamics in dissolution test-ing, mainly attributed to the
influence exercised changing paddle rotation speeds and (ii) location of the
formulation in the dissolu-tion vessel [25,26].
In regard to the first aspect, many studies performed in this field revealed
devices, results in a varied flow dynamics and surface area within the vessel
inducing a high test variability [27]. Moreover, it was assessed that different
fluid velocities exist at dif-ferent points in the dissolution vessel meaning that
the amount

Table 3
Ritger and Peppas’s kinetic mathematical model and first order kinetics model fitting.
Tablet n
Mt /M∞ = Kt

n = 1 (zero n = 0.9 n = 0.8 n = 0.7 n = 0.6 n = 0.5 n = 0.5

order) release 0–60%
Higuchi
2 y = 3.096x + 3.451 y = 4.218x + 1.896 y = 5.812x − 0.101 y = 8.025x − 2.522 y = 11.120x − 5.538 y = 15.502x − 9.422 y = 14.626x − 8.156
r = 0.983 r = 0.991 r = 0.994 r = 0.995 r = 0.993 r = 0.985 r = 0.981
6 y = 2.443x − 0.883 y = 3.333x − 2.058 y = 4.550x − 3.446 y = 6.220x − 5.122 y = 8.528x − 7.202 y = 11.755x − 9.874 y = 10.592x − 8.147
r = 0.988 r = 0.985 r = 0.979 r = 0.970 r = 0.957 r = 0.940 r = 0.955
 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922
of released drug from the same formulation changes in relation to its position
in the vessel. Some studies revealed that a ‘dead zone’ forms at the bottom of
the vessel where the agitation rate is minimum leading to low dissolution
rates [28].
Concerning data coming from the dissolution tests (methods A and B), it
is possible to assert that in the case of the latter the low paddle speed (50 rpm)
and the formulation location at the bot-tom of the vessel are responsible for
the low amount of CARBA released from both tablets 2 and 6. When the
tablet is placed at 7.0 cm from the bottom of the vessel (method A) and the
paddle speed is increased to 150 rpm, the tablet is localized in a region in
which the dissolution fluid velocity is high resulting more solicited. In these
conditions tablet swelling results faster as well as drug release.
Taking into account that this kind of tablet must be applied on subligual
mucosa and that in this region the formulation can receive different
solicitations (as salivary flow and shearing forces due to tongue movements)
method A appears to be more predictable of physiological conditions than
method B.
3.3. In vitro release mathematical model
In vitro release data were submitted to statistical investigation in order to
know the kinetic mechanism involved in CARBA release from both tablets 2
and 6. The projected tablets are classified as swellable matrix drug delivery
systems as they are constituted from a polymeric mixture able to hydrate in
presence of water. In these systems the drug is released from the matrix by
diffusion, which is controlled from the polymer network. In order to inves-
tigate what mechanism is involved in CARBA release, data of drug
dissolution coming from tablets 2 and 6 were fitted to Ritger and Peppas
n Therapeutic Drug Monitoring, third ed., Applied Therapeutics, Vancouver, 1992, pp. 1–29.
kinetics mathematical model Mt /M∞ = Kt [29], applied to swellable
matrices. The release mechanism can be controlled by water penetration rate,
responsible for hydration and drug diffu-sion, and polymeric chain relaxation
time. Diffusional exponent value of one (n = 1) means that drug release
occurs as an appar-ent zero-order mechanism when is time dependent while
value of n = 0.5 means that release is controlled by a pure Fickian diffusion
mechanism. A n value between 0.5 and 1 indicates an anomalous mechanism
(not Fickian) meaning that both liquid penetration rate and polymeric chain
relaxation rate control the drug release.
Analyzing the results (Table 3) it is recognized that the best fitting for
tablet 2 was obtained for n = 0.7, meaning that drug is released by an
anomalous mechanism. For tablet 6 the best fitting was obtained for n = 1
thus, in this case drug release follows a zero order kinetic.
As in tablet 2 the main constituent is HPMC (75%), a polymer with high
water affinity, drug release is mainly due to diffusion mechanisms. In regard
to tablet 6 the main factor governing CARBA release is the relaxation rate of
®
polymeric chains. In fact, the high Carbopol (75%) amount reduces tablet
affinity to water thus, in the case of tablet 6 the relaxation rate is the main
factor governing CARBA release.
4. Conclusions
Mucoadhesive bi-layered tablets, constituted by an imperme-able backing
layer and a mucoadhesive layer, intended for CARBA sublingual
administration have been prepared and characterized. The best formulations in
terms of swelling capacity, mucoadhe-sion performances and patient
® Guidance for industry dissolution testing of immediate release solid oral dosage forms,
acceptability were those having Carbopol :HPMC 25%:75%
®
Carbopol :HEC 75%:25% as mucoad-hesive layer. The in vitro release
studies, performed by two different methods, showed that the highest drug
release in 24 h
921
was obtained from tablet which mucoadhesive layer compositions
®
Carbopol :HPMC 25%:75%.
This buccal formulation could represent an interesting approach in the
treatment of chronic disease as it is able to promote a sus-tained CARBA
release useful to reduce daily administrations in comparison to conventional
formulations. Moreover, the devel-oped tablet is very manageable and easy to
apply, favoring patient compliance.
Conflict of interest
The authors declare that they have no personal financial or non-financial
conflicts of interest in this research.
Acknowledgments
Authors are very much grateful to Dr. Renato Innamorati from ASL N. 1
(Umbria, Italy), for providing porcine mucosa and to Mr. Marco Marani for
technical assistance.
References
[1] K. Parfitt, Martindale, The complete drug reference, 32th ed., Pharmaceutical Press,
London, 1999.
[2] M. Lindenberg, S. Kopp, J.B. Dressman, Classification of orally administered drugs on the
World Health Organization model list of the essential medicines according to the
biopharmaceutics classification system, Eur. J. Pharm. Bio-pharm. 58 (2004) 265–278.
[3] J.M. Custodio, C.Y. Wu, L.Z. Benet, Predicting drug disposition, absorp-
tion/elimination/transporter interplay and the role of food on drug absorption, Adv. Drug
Deliv. Rev. 60 (2008) 717–733.
[4] R. Levy, A.J. Wilensky, G.D. Anderson, Carbamazepine, valproic acid, phe-nobarbital, and
ethosuximide, in: W.E. Evans, J.J. Jusko (Eds.), Applied Pharmacokinetics. Principles of
[5] L.S. Koester, J.B. Bertuol, K.R. Groch, C.R. Xavier, R. Moellerke, P. Mayorga, T. Dalla
Costa, V.L. Bassani, Bioavailability of carbamazepine: -cyclodextrin com-plex in beagle
dogs from hydroxypropylmethylcellulose matrix tablets, Eur. J. Pharm. Sci. 22 (2004)
201–207.
[6] L.I. Giannola, De V. Caro, G. Giandalia, M.G. Siragusa, M. D’Angelo, L. Lo Muzio, G.
Campisi, Transbuccal tablets of carbamazepine: formulation, release and absorption
pattern, Int. J. Immunopathol. Pharmacol. 18 (2005) 21–31.
[7] G.D. Andersona, R.P. Saneto, Current oral and non-oral routes of antiepileptic drug
delivery, Adv. Drug Deliv. Rev. 64 (2012) 911–918.
[8] I.W. Kellaway, G. Ponchel, D. Duchêne, Oral mucosal drug delivery, in: M.J. Rathbone, J.
Hadgraft, M.S. Roberts (Eds.), Modified Release Drug Delivery Tech-nology, Marcel
Dekker Inc., New York, 2003, pp. 349–350.
[9] S. Rossi, G. Sandri, C.M. Caramella, Buccal drug delivery: a challenge already won? Drug
Discov. Today Technol. 2 (2005) 59–65.
[10] N. Salamat-Miller, M. Chittchang, T.P. Johnston, The use of mucoadhesive poly-mers in
buccal drug delivery, Adv. Drug Deliv. Rev. 57 (2005) 1666–1691.
[11] G.S. Asane, S.A. Nirmal, K.B. Rasal, A.A. Naik, M.S.M.Y.M. Rao, Polymers for
mucoadhesive drug delivery system: a current status, Drug Dev. Ind. Pharm. 34 (2008)
1246–1266.
[12] Y. Sudhakar, K. Kuotsu, A.K. Bandyopadhyay, Buccal bioadhesive drug delivery
– a promising option for orally less efficient drugs, J. Control. Release 114 (2006) 15–40.
[13] R. Birudaraj, R. Mahalingam, X. Li, R.J. Bhaskara, Advances in buccal drug deliv-ery,
Crit. Rev. Ther. Drug Carrier Syst. 22 (2005) 295–330.
[14] V.F. Patel, F. Liu, M.B. Brown, Advances in oral transmucosal drug delivery, J. Control.
Release 153 (2011) 106–116.
[15] L. Perioli, V. Ambrogi, D. Rubini, S. Giovagnoli, P. Blasi, C. Rossi, Novel mucoad-hesive
topical formulation containing metronidazole for the treatment of periodontal diseases, J.
Control. Release 95 (2004) 521–533.
[16] L. Perioli, V. Ambrogi, C. Pagano, S. Scuota, C. Rossi, FG90 chitosan as a new polymer
for metronidazole mucoadhesive tablets for vaginal administration, Int. J. Pharm. 377
(2009) 120–127.
[17] V.P. Shah, J.J. Konecny, R. Everett, C. McCullough, A.C. Nooryzadeh, J.P. Skelly, In vitro
dissolution profile of water-insoluble drug dosage forms in presence of surfactants, Pharm.
Res. 6 (1989) 612–618.
[18] V.P. Shah, A. Noory, C. Noory, B. McCullough, S. Clarke, R. Everett, H. Naviasky, B.N.
Srinivasan, D. Fortman, J.P. Skelly, In vitro dissolution of sparingly water soluble drug
dosage forms, Int. J. Pharm. 125 (1995) 99–106.
[19] Food and Drug Administration Center for Drug Evaluation and Research (CDER),
or
(August) 1997.
[20] S. Miyazaki, N. Kawasaki, T. Nakamura, M. Iwatsu, T. Hayashi, W.M. Hou, D. Attwood,
Oral mucosal bioadhesive tablets of pectin and HPMC: in vitro and in vivo evaluation, Int.
J. Pharm. 204 (2000) 127–132.
922 L. Perioli, C. Pagano / Colloids and Surfaces B: Biointerfaces 102 (2013) 915–922
[21] C. Remunán˜-López, A. Portero, J.L. Vila-Jato, M.J. Alonso, Design and evalua-tion of
chitosan/ethylcellulose mucoadhesive bilayered devices for buccal drug delivery, J.
Control. Release 55 (1998) 143–152.
[22] D.Y. Qiu, Y. Chen, L. Liu, G.G.Z. Zhang, Developing Solid Oral Dosage Forms:
Pharmaceutical Theory and Practice, Academic Press, New York, 2009.
[23] Y. Machida, T. Nagai, Bioadhesive preparation as topical dosage forms, in: E. Mathiowitz,
D.E. Chickering III, C.M. Lehr (Eds.), Bioadhesive Drug Delivery Systems, Marcel
Dekker Inc., New York, 1999, pp. 646–647.
[24] EMEA, Guideline on stability testing: stability testing of existing active substances and
related finished products, CPMP/QWP/122/02, rev 1, 2003.
[25] L.G. McCarthy, G. Bradley, J.C. Sexton, O.I. Corrigan, A.M. Healy, Computational fluid
dynamics modeling of the paddle dissolution apparatus: agitation rate,
mixing patterns, and fluid velocities, AAPS PharmSciTech 5 (2004) 1–10, Article 31,
http://www.aapspharmscitech.org
[26] A.M. Healy, L.G. McCarthy, K.M. Gallagher, O.I. Corrigan, Sensitivity of dissolu-tion rate
to location in the paddle dissolution apparatus, J. Pharm. Pharmacol. 54 (2002) 441–444.
[27] S.A. Qureshi, J. Shabnam, Cause of high variability in drug dissolution testing and its
impact on setting tolerances, Eur. J. Pharm. Sci. 12 (2001) 271–276.
[28] T. Mirza, Y. Joshi, Q. Liu, R. Vivilecchia, Evaluation of dissolution hydrodynamics in the
TM
USP, Peak and flat-bottom vessels using different solubility drugs, Dissol. Technol. 12
(2005) 11–16.
[29] P. Ritger, N.A. Peppas, A simple equation for description of solute release. II. Fickian and
anomalous release from swellable devices, J. Control. Release 5 (1987) 37–42.