International Journal of PharmTech ResearchCODEN (USA): IJPRIF ISSN : 0974-4304

Vol.2, No.1, pp 656-667, Jan-Mar 2010

A NOVEL EFFERVESCENT BIOADHESIVE VAGINALTABLET OF KETOCONAZOLE: FORMULATION AND IN-

VITRO EVALUATION

Patel Geeta M.* and Patel Anita P.

*Department of Pharmaceutics & Pharmaceutical Technology,S. K. Patel College of Pharmaceutical Education and Research, Ganpat University,

Kherva-382711, Gujarat, India

*Corresponding Author: geekhappy2002@yahoo.co.inTel.: +91-1762-286082; Fax: +91-1762-286082

Abstract: Purpose: To develop and evaluate effervescent bioadhesive vaginal tablet of ketoconazole. Methods: Thevaginal tablets were prepared by direct compression. HPMC K4M, HPMC K15M, HPC, Sodium CMC, Chitosan,Sodium Alginate, Methyl Cellulose and Cabopol 941 were used as bioadhesive polymers. Effervescent was incorporatedinto the formulations as a disintegration agent. The amount of polymer blends and effervescent mixture was optimizedusing 32 full factorial design. The swellings and in-vitro release were studied. The ex-vivo mucoadhesion wasdetermined by self developed modified mucoadhesion assembly. The ex-vivo residence test was carried out by modifiedUSP dissolution test apparatus. Anti-fungal activity of the effervescent bioadhesive tablet of ketoconazole wasdetermined in comparison to Candid®-V3 tablet and Candid®-V gel. Results: A good sustained effect and a moderatebioadhesion were (0.088N to 0.267N) obtained with the tablet containing HPMC K4M: Chitosan (1:1) and effervescentmixture (3:1). Ex-vivo mucoadhesion time of all the formulations was in the range of 8 to 24 h. The effervescentketoconazole tablet showed significantly higher (p<0.05) in-vitro antifungal activity as compare to Candid®-V3 tabletand Candid®-V gel. Conclusion: Our study may provide a potential vaginal formulation of ketoconazole against Candidaalbicans.Key words: Bioadhesion, vaginal drug delivery, mucoadhesive polymers, effervescent, C. albicans

INTRODUCTIONIn recent years vaginal bioadhesive preparations havebeen developed as a new type of controlled-releaseform for the treatment of both topical and systemicdiseases. For drugs which are susceptible to gut orhepatic metabolism or which cause GI side effects,vaginal bioadhesive delivery may offer a number ofadvantages over the other routes of administration. Thegreatest advantage of such dosage forms is thepossibility of maintaining them in the vagina forextended periods of time including daytime andnighttime, thereby enabling lower dosing frequencies.1

Vaginal candidiasis is a common condition and up to75% of all women suffer at least one episode of thisinfection during their lifetime. Candida albicans is themost important cause of vaginal candidiasis,

accounting for over 80% of the infection. Mostpatients with Candida vaginitis respond to topicaltreatment with nystatin or imidazoles.2 Ketoconazole(KTZ) is an imidazole derivative antifungal agentdeveloped for treatment of human mycotic infectionsand plays an essential role in antifungalchemotherapy.3 It is a weak base with limited watersolubility.4 In the literature, some additives, such assodium bicarbonate and citric acid, were added intoformulations to improve the disintegration and thedissolution of bioadhesive vaginal tablets becausethere is rather low moisture content in vagina undernormal physiological conditions.5

The aim of this study was to prepare a newbioadhesive effervescent vaginal tablet formulation of

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 657

ketoconazole against C. albicans. Effervescent addedinto the formulations as disintegration agent would beexpected to increase the dissolution of ketoconazole.Hence, the present work was envisaged to develop anovel and aesthetic bioadhesive vaginal deliverysystem with improved efficacy. The followingbioadhesive polymers were screened to develop avaginal delivery system: HPMC K4M, HPMC K15M,HPC, Sodium CMC, Chitosan, Sodium Alginate,Methyl Cellulose and Carbopol 941.The performancesof these bioadhesive polymers were evaluated by twoparameters, the swelling behavior and the bioadhesivestrength. For the various drug-free formulations, theeffect of effervescent on polymers’ bioadhesivecharacteristics was investigated. On the basis of thesedata, suitable polymers were selected to prepare thebioadhesive effervescent vaginal tablets ofketoconazole. Swelling, bioadhesive properties anddrug release of the tablets with different proportions ofbioadhesive polymer and effervescent in formulationswere conducted. Anti-fungal activity of theeffervescent bioadhesive tablet of ketoconazole wasdetermined in comparison to Candid®-V3 tablet andCandid®-V gel.

MATERIALS AND METHODSMaterialsKetoconazole was kindly gifted by RanbaxyLaboratories Limited (Mumbai, India). Carbopol 941was kindly gifted by Corel Pharmaceutical Ltd.(Ahmedabad, India). HPMC K4M, HPMC K15M andHPC were received as gift samples from Zydus-CadilaHealthcare Ltd. (Ahmedabad, India). Sodium CMCwas kindly gifted by Mann Pharmaceuticals Pvt. Ltd.(Mehsana, India). Chitosan was obtained as a giftsample from Troikka Pharmaceuticals Ltd.(Ahmedabad, India). Sodium Alginate, MethylCellulose and MCC were purchased from S.D. FineChemicals Ltd. (Mumbai, India). All other ingredientswere of laboratory grades. Candid®-V3 tablet andCandid®-V gel were purchased from local market.

Preliminary screening of ketoconazole effervescentvaginal tabletDifferent mucoadhesive polymers were screened forketoconazole vaginal tablet. The formulations wereprepared by direct compression technique. Therequired quantity of polymers, ketoconazole,effervescent (sodium bicarbonate and citric acid at themole ratio of 3:1) and the other formulationingredients were compressed using single punch tabletmachine (Cadmack, Ahmedabad, India). The tabletmould was especially designed using stainless steeland the formulated tablets had an average weight 800± 1.5 mg, 20.7 ± 0.2 mm height, 8.5 ± 0.3 mm widthand 5.27 ± 0.057 mm thickness.

Swelling studyThe swelling behavior of tablet described as the waterabsorbing capacity. Drug-free tablets were weighedindividually (W0) and placed separately in 2% agar gelplates and incubated at 37 ± 1°C. At regular timeintervals up to 4 hours, the tablets were removed fromthe petri dish and wiped with filter paper carefully.The swollen tablet was then reweighed (Wt) and the %swelling were calculated using the followingformula:6,7

% Swelling = {(Wt – W0)/ W0} × 100…..(1)

Where, Wt is the weight of the tablet at time t and W0is the initial weight of tablet. The swelling wascalculated and then plotted as a function of time. Theslope of the linear plots was taken as the swelling rate.

Optimization using 32 full factorial designA statistical model incorporating interactive andpolynomial term was used to evaluate the response:

Y = b0 + b1X1+b2X2 + b12X1X2 + b11X12 + b22X2

Where, Y is the dependent variables, b0 is thearithmetic mean response of the nine runs, and b1 is theestimated coefficient for the factor X1. The maineffects (X1 and X2) represent the average result ofchanging one factor at a time from its low to highvalue. The interaction terms (X1X2) show how theresponse changes when two factors are simultaneouslychanged. The polynomial terms (X1

2 and X22) are

included to investigate non-linearity.A 32 full factorial design was adapted to optimize thevariables. In this design 2 factors were evaluated, eachat 3 levels, and experimental trials were performed atall 9 possible combinations. Polymer ratio (X1) andamount of effervescent (X2) were selected asindependent variables. Mucoadhesive strength, %swelling, Q8 and t80 were selected as dependentvariables (response; Y). The preparation andevaluation method for tablets and amount ofKetoconazole were kept constant for all the trials. Thefull factorial design lay out , coded values for polymerratio(X1) and amount of effervescent mixture(X2), andcomposition of factorial batches A1 to A9 are shownin Table I and Table II respectively.

Physicotechnical parameters of effervescentmucoadhesive tabletThe uniformity of weights of all tablets wasdetermined by using sartorious balance (Model CP-224 S). The hardness of the tablets was determined bydiametral compression using a dial type hardness tester(Model no 1101, Shivani Scientific India). Tabletthickness was measured using vernier calipers.8

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 658

Ex-vivo mucoadhesion studySeveral types of mucosa including rat intestine, pigoral, bovine sublingual, cow vaginal mucosa 9,10, havebeen used as model biological tissues for theevaluation of bioadhesion. A modified self developedforce detachment devise was used to measure theminimum detachment force (Figure I). A piece of ratintestine (2.0 cm x 1.0 cm) removed from newlysacrificed rat was adhered to a piece of glass, whichwas fixed on a plank and the plank was assembledwith a little crown block. After hydrating the ratintestine with distilled water, the tablet was broughtinto contact with the rat intestine by applying littleforce for minute. After the initial contact, the tabletwas encircled by a thread which fastened a light plasticbeaker through the crown block. Next, water wasdropped into the beaker at a speed of 3.0 ml/min usingperistaltic pump until the tablet and rat intestine werepulled apart by the gravity of water. The beakercontaining water was weighed and the minimumdetachment force was calculated accordingly. Theexperiments were performed in triplicate and averagevalues with standard deviation (SD) were reported.The study was approved by Shree S. K. Patel Collegeof Pharmaceutical Education and Research,Institutional Animal Ethics Committee and theapproval document no. was SKPCPER-IAEC/2008/10.

Ex-vivo mucoadhesion timeThe Ex-vivo residence time was studied using a locallymodified USP paddle apparatus (Figure 1, Dissolutiontest apparatus type-I). The dissolution medium (500 mlcitrate buffer pH 4.0) was maintained at 37 °C. Asegment of rat intestine, 2.5 cm long, was glued to thesurface of a glass slab, vertically attached to thepaddle. The mucoadhesive tablet was hydrated fromone surface using 15 ml phosphate buffer and then thehydrated surface was brought into contact with themucosal membrane. The glass slide was verticallyfixed to the paddle and allows rotating at 50 rpm. Thetime required for complete detachment of the tabletfrom the mucosal surface was recorded (mean oftriplicate determinations).11

In-vitro ketoconazole release studyThe release rate of ketoconazole effervescentmucoadhesive vaginal tablet (n=3) was determinedusing The United States Pharmacopoeia (USP) XXIVdissolution testing apparatus I (basket method) using500ml of citrate buffer pH 4.0 as a dissolutionmedium. The tablet was placed in a settling basket toprevent the tablet from floating.12 The rate of stirringwas 50 rpm and the medium temperature wasmaintained at 37 ± 0.5 ºC. A sample (10 ml) of thesolution was withdrawn from the dissolution apparatushourly and the samples were replaced with fresh

dissolution medium. The samples were filtered througha filter and diluted to a suitable concentration withcitrate buffer pH 4.0. Absorbance of these solutionswas measured at 221 nm using a Shimadzu UV-1601UV/Vis double beam spectrophotometer. Cumulativepercentage of drug release was calculated using theequation obtained from a standard curve.

Statistical analysisTests for significant differences between means wereperformed by Student’s t-test or one-way ANOVA byusing the software Sigma State version 10. Differenceswere considered significant at P < 0.05 level.

The surface response plotThe surface response plots were drawn using Sigmaplot software version 10 (Jandel Scientific, San Rafael,CA).

Kinetic modeling of drug releaseCurve fitting was performed using Microsoft Excel2007 version. The dissolution data were fitted tofollowing eqation. Release exponent ‘n’ was calculated(N. A. Peppas et al., 1985 and Korsemeyer R. W. etal., 1993).13

Mt / M = Ktn

Where, Mt / M is the fraction of the drug released attime t, k is the kinetic constant of the system, and n isthe exponent characteristic of the mode transport. Therelease exponent takes various values depending upondifferent geometries. For the drug release from acylindrical or a flat swellable polymer, if n approachesto 0.89, the release mechanism could be Case-IItransport and if n is close to 0.45, the releasemechanism can be Fickian. On the other hand if 0.45 <n < 0.89, non-Fickian transport could be obtained.

In- vitro antifungal studyIn-vitro antifungal study was performed againstCandida albicans in Sabouraud’s agar medium by thecup plate method.14 The cups cut in the inoculatedsolidified media were filled with different formulationsusing sterilized syringes. The marketed tablet(Candid®-V3) was crushed into a powder and dissolvedin 2 mL of sterilized water was applied using sterilizedsyringe. The marketed gel (Candid®-V gel) wasapplied using the sterilized syringe. The developedketoconazole effervescent mucoadhesive vaginal tabletwas swelled in 2 mL of sterile water applied into thecups. The covered petri plates were incubated at 22°Cin the BOD incubator for 48 hours. The zone ofinhibition was measured at the end of 48 hours.

Accelerated stability study of best batchIn order to determine the change in in-vitro releaseprofile and mucoadhesive strength on storage,

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 659

accelerated stability study of batch A4 was carried outat 40 ºC in a humidity chamber having 75% RH for 3months.15 After 3 month samples were withdrawn andevaluated for change in physical appearance, in-vitrodrug release pattern and mucoadhesive strength.

RESULTS AND DISCUSSIONPreliminary screeningSwelling studySwelling is important for the assessment of adhesion.Shortly after swelling, adhesion does occur, but with aweak bond formed. To develop maximum adhesionstrength, an optimum water concentration was neededfor polymer particles.It was observed that the order of swelling rate wasCarbopol 941 > Chitosan > Sod.CMC > HPMC K4M> HPMC K15M > Sod.alginate > MC in drug freeformulations. According to the comparison of thecorresponding swelling profiles of formulationswith/without effervescent, it could be seen that theeffervescent resulted in a marked increase in swellingrate. Furthermore, most tablets with 100 mgeffervescent showed a higher swelling capacity thantablets without effervescent. It was observed that theswelling rate was developed as effervescent applied toformulation, increased with increasing amount ofeffervescent. The phenomenon of swelling increasingcould be explained by the good disintegration effect ofeffervescent, which made tablets increase in volumeand construct porous channels on surface and inside oftablets. The porous channels increased the area ofcontacting between polymer particles and water so thatthe polymers could be hydrated more easily.

In-vitro dissolution studyThe dissolution of tablets were carried out in citratebuffer pH 4.0. In all the formulations, the burst releaseof ketoconazole was observed within 2 hrs. andgradually increasing upto 8-10 hrs. Formulations (F1to F8) contained single mucoadhesive polymer.Formulations F1 and F2 (carbopol 941 and chitosan)swelled rapidly with aqueous fluid and formed rigidstructure and provided a stronger sustained releaseeffect on the drug release. Formulations F2, F3, F4 andF6 ( HPMC K4M, HPMC K15M, HPC andSod.alginate) were more erodible resulting in lessprolong release. Formulations F5 and F7 ( Sod. CMCand MC) swelled rapidly when in contact with aqueousfluid to form gel which prevented fast disintegration ofthe tablet and formation of prolong release gel.Formulations F9 and F14 containing Carbopol 941 ledto gradual swelling. The drug release was in acontrolled manner for more than 8 hrs. The percentagedrug release after 8 hrs was 72.36% and 67.11 %respectively. Formulations F11, F13 and F16 showsrapid disintegration , more erodibility and completedrug release within 5-7 hrs. Formulations F10 and F12

shows good drug release, more than 90% of drugwithin 8-10 hrs.

Ex-vivo mucoadhesion studyMucoadhesive strength of formulation F1 to F16 wasdetermined using self developed force detachmentmethod (Figure 1). The mucoadhesive strength wasobserved within the range of 0.181 to 0.393 N. Basedon results obtained from dissolution study formulationF10 and F12 both show good release profile.Furthermore, formulation F10 (0.258N) show slightmore adhesion property compare to formulation F12(0.247N). So on the basis of mucoadhesive strengthand dissolution release studies, formulation F10(Chitosan and HPMC K4M) was considered a goodcandidate for development of a ketoconazoleeffervescent mucoadhesive vaginal tablet.

Optimization using 32 full factorial designThe number of experiments required for these studiesare dependent on the number of independent variablesselected. The response (Yi) is measured for each trial.In order to investigate factors systematically, afactorial design was employed in the presentinvestigation. On the basis of the preliminary trials a 32

full factorial design was employed to study the effectof independent variables i.e. Polymer ratio (X1) andamount of effervescent (X2) on dependent variablesmucoadhesive strength, % swelling, Q8 and t80.

Swelling study of factorial design batchesComparing with drug-free tablets employing the sameamount of polymer and effervescent, all ketoconazoleeffervescent mucoadhesive vaginal tablets showedlower swelling rates, which is related with the poorsolubility of ketoconazole.

Ex-vivo mucoadhesion studyIn general, the swelling state of polymer contributes toits bioadhesive behavior. It was observed that theswelling rate was developed as effervescent applied toformulation, increased with increasing amount ofeffervescent; however, the effervescent led to asignificant drop in adhesive strength. The influences ofeffervescent on swelling and mucoadhesion wereopposite, mainly due to the tiny bubbles created byeffervescent. These tiny bubbles depressed themucosa-polymer interaction, resulting in a decrease inthe mucoadhesive strength.16

The minimum adhesion strength (0.088N) wasobserved in formulation A3, which could be due to thelower ratio of HPMC K4M: Chitosan and the highercontent of effervescent. On the contrary, with anincrease in HPMC K4M: Chitosan ratio and decreasein effervescent, the maximum adhesion strength(0.267N) was obtained for formulation A7.

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 660

Ex-vivo mucoadhesion timeThe time for the tablet to detach from the rat intestinewas recorded as the mucoadhesion time. The increasein concentration of polymer blend in series fromformulation A1 to A9, showed a gradual increase inmucoadhesion time. The factorial batches A1 to A3showed mucoadhesion time 8-10 hours, batches A4 toA6 showed 12-15 hours and batches A7 to A9 showed24 hours.

In- vitro drug release study of factorial designbatchesThe release rate of ketoconazole from effervescentmucoadhesive vaginal tablet was described as afunction of time as shown in Figure 2. In all theformulations, the burst release of ketoconazole wasobserved within first 2 hrs, and then graduallyincreased up to 8-12 hrs. For the polymer mixture ofHPMC K4M and Chitosan, more drug release could beseen as decreasing HPMC K4M and Chitosan andincreasing amount of effervescent mixture

Full factorial design batchesThe mucoadhesive strength, % swelling, Q8 and t80 forthe nine batches showed wide variation. The resultsclearly indicate that all the dependent variables arestrongly dependent on the selected independentvariables as they show a wide variation among the ninebatches (A1 to A9).Drug content for ketoconazole was carried out bymeasuring the absorbance of samples at 221 nm usingShimadzu UV-1601 UV/Vis double beamspectrophotometer and comparing the content from acalibration curve prepared with standard Ketoconazolein the same medium. The total amount ofKetoconazole present in each tablet was found to be inthe range of 97.5 to 99%.

Statistical analysis of factorial design batchesThe statistical analysis of the factorial design batcheswas performed by multiple linear regression analysiscarried out in Microsoft Excel 2007. The data clearlyindicate that the values of mucoadhesive strength, %swelling, t80, and Q8 (% drug release after 8 hrs.) arestrongly dependent on the independent variables. Thefitted equations (full and reduced) relating theresponses mucoadhesive strength, % swelling, Q8 andt80 to the transformed factors are shown in Table 3.The polynomial equations can be used to drawconclusions after considering the magnitude ofcoefficient and mathematical sign it carries (i.e.positive or negative). Table 8 shows the results of theanalysis of variance (ANOVA), which was performedto identify insignificant factors.17

The high value of correlation coefficient formucoadhesive strength, % swelling, Q8 and t80indicates good fit i.e., good agreement betweendependent and independent variables. The equations

may be used to obtain estimates of the responses assmall error of variance was noticed in the replicates.The significant test for regression coefficients wasperformed by applying student F test. A coefficient issignificant if the calculated F value is greater than thecritical value of F.

Full and reduced model for mucoadhesive strengthThe significant level of coefficient b12 was found to bep = 0.07 hence it was omitted from full model togenerate reduced model. The results of statisticalanalysis are shown in Table III. The coefficients b1, b2,b11 and b22 were found to be significant at p < 0.05;hence they were retained in reduced model. Thereduced model was tested in portions to determinewhether the coefficient b12 contribute significantinformation for the prediction of mucoadhesivestrength or not. The results for testing the model inportion are shown in Table IV. The critical value of Ffor α = 0.05 is equal to 10.13 (DF = 1, 3). Since thecalculated value (F = 5.0) is less than critical value (F= 10.13), it may be concluded that the interaction termb12 do not contribute significantly to the prediction ofmucoadhesive strength and therefore can be omittedfrom the full model. The results of multiple linearregression analysis (reduced model) revealed that, onincreasing the concentration of polymer an increase inmucoadhesive strength was observed; the coefficientsb1 and b2 bear positive and a negative sign respectively.When high concentration of polymer was used, higheradhesion was expected in the tables. While in case ofamount of effervescent, on increasing amount adecrease in mucoadhesive strength was observed. It isobvious that the presence of low amount ofeffervescent adhesion is facilitated. The fitted equationfor full and reduced model relating the response isgiven below.Full model (mucoadhesive strength)Y = 0.221+ 0.0412X1 - 0.054X2- 0.011X1X2 - 0.029X1

2

- 0.027X22

Reduced modelY = 0.221+ 0.0411X1 - 0.055X2 - 0.029X1

2 - 0.027X22

Full and reduced model for % swellingThe significance level of coefficients b12 and b22 wasfound to be p = 0.27 and p = 0.11 respectively, hencethey were omitted from the full model to generate thereduced model. The coefficients b1, b2 and b11 werefound to be significant at P < 0.05, hence they wereretained in the reduced model. The reduced model wastested in portions to determine whether the coefficientsb12 and b22, contribute significant information for theprediction of t80 or not. The results for testing themodel in portions are depicted in Table IV. The criticalvalue of F for α = 0.05 is equal to 9.55 (DF = 2, 3).Since the calculated value (F = 3.366) is less than thecritical value (F = 9.55), it may be concluded that theinteraction term and polynomial terms do not

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 661

contribute significantly to the prediction of %swelling. Hence, conclusions can be drawnconsidering the magnitude of the coefficient and themathematical sign (positive or negative) it carries.An increase in the concentration of polymer leads to anincrease in % swelling is observed; both thecoefficients b1 and b2 bear a positive sign. The amountof effervescent is increased the swelling increases aswell as concentration of polymer increases theswelling increases. But both having opposite action ondrug release i.e. as the amount of effervescent isincreased the drug release increases while in case ofpolymer concentration as the concentration increasesthe drug release decreases. Hence the fitted equationfor full and reduced model relating the response wasgiven below.Full model (% swelling)Y = 20.3+ 3.083X1 + 1.608X2- 0.313X1X2 – 1.15X1

2 –0.735X2

2

Reduced model (% swelling)Y = 19.81+ 3.083X1 + 1.608X2– 1.15X1

2

Full and reduced model for Q8The significant level of coefficient b12, b11 and b22 wasfound to be p = 0.16, p = 0.08 and p = 0.92respectively, hence it was omitted from full model togenerate reduced model. The coefficients b1and b2were found to be significant at p < 0.05; hence theywere retained in reduced model. The results for testingthe model in portion are shown in Table 8. The criticalvalue of F for α = 0.05 is equal to 9.28 (DF = 3, 3).Since the calculated value (F = 3.286) is less thancritical value (F = 9.28), it may be concluded that theinteraction term and polynomial terms b12, b11 and b22do not contribute significantly to the prediction of Q8and therefore can be omitted from the full model. Thefitted equation for full and reduced model relating theresponse was given below.Full model (Q8)Y = 81.99 – 22.541X1 + 4.431X2+ 2.84X1X2 –5.435X1

2 + 0.235X22

Reduced model (Q8)Y = 78.52 – 22.541X1 + 4.432X2

Full and reduced model for t80The significant level of coefficient b22 was found to bep = 0.63 hence it was omitted from full model togenerate reduced model. The coefficients b1, b2, b12 andb11 were found to be significant at p < 0.05; hence theywere retained in reduced model. The critical value of Ffor α = 0.05 is equal to 10.13 (DF = 1, 3). Since thecalculated value (F = 0.278) is less than critical value(F = 10.13), it may be concluded that the polynomialterm b22 do not contribute significantly to theprediction of t80 and therefore can be omitted from thefull model. The fitted equation for full and reducedmodel relating the response was given below. Fullmodel (t80)

Y = 7.7+ 2.9X1 – 0.66X2- 0.397X1X2 + 1.03X12 +

0.04X22

Reduced model (t80)Y = 7.7+ 2.9X1 – 0.66X2- 0.397X1X2 + 1.03X1

2

The response surface plotFor drawing the conclusions, response surface plot wasused. Figure 3 to Figure 6 shows the plot of polymerratio (X1) and amount of effervescent (X2) versusmucoadhesive strength, % swelling, Q8 and t80respectively. The plots were drawn using Sigma plotsoftware (Jandel Scientific, San Rafael, CA).The plots demonstrate that X1 and X2 affect themucoadhesive strength, % swelling, Q8 and t80. BatchA4 was selected as best batch. It was arbitrarilydecided to select a batch of tablets that gives moderatemucoadhesive strength and drug release in a controlmanner. The final selection is done after consideringsome aspects such as drug release profile, ex-vivoretention time and t80.A checkpoint batch A10 was prepared at X1 = - 0.3level and X2 = 0.7. From the reduced model, it isexpected that the value of mucoadhesive strength ofthe checkpoint batch should be 0.154 N; the value of% swelling should be 19.91, the value of Q8 of thecheckpoint batch should be 97.40 % and the value oft80 of the checkpoint batch should be 6.96 hrs. The dataindicates that the results are as expected. Thus, we canconclude that the statistical model is mathematicallyvalid.

Kinetic analysis of dissolution dataThe in- vitro release data obtained were fitted tokorsmeyer peppas kinetic model. In the entire batchesexponent ‘n’ was in between 0.45 and 0.89, sopredominant drug release mechanism is non-Fickian(anomalous) transport.

Ex-vivo mucoadhesion timeThe increase in concentration of polymer ratio in seriesfrom formulation A1 to A9, showed a gradual rise inmucoadhesion time. The factorial batches A1 to A3showed mucoadhesion time 10 hours, batches A4 toA6 showed 15 hours and batches A7 to A9 showed 24hours.

Anti fungal studyThe ketoconazole effervescent mucoadhesive vaginaltablet had better antimicrobial activity as comparedwith the marketed formulations (Candid V3

®- tabletand Candid V®-gel). Mucoadhesive polymers of thetablet had prolonged drug release and provided bettercontact with the wells cut in the plate, while theCandid V3 suspension dried up as water was notavailable in the wells for prolonged time to allowdiffusion of drug molecule(s). The zone of inhibitionwas measured at the end of 48 hours. The effervescentketoconazole tablet showed significantly higher

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 662

(p<0.05) in-vitro antifungal activity as compare toCandid®-V3 tablet and Candid®-V gel. The results ofantifungal studies are reported in Figure 7.

Selection of best batchIt was arbitrarily decided to select a batch of tabletsthat gives good mucoadhesive strength. The finalselection is done after considering some aspects suchas drug release profile, ex-vivo retention time and t80.On the basis of mucoadhesive strength and dissolutionrelease studies A4 comprising HPMC K4M: chitosan(1:1 ratio) and amount of effervescent (50mg) wasconsidered a good candidate. The aim of study was,tablet should release more than 90% drug within 8-10hrs and tablet should have satisfactory adhesivestrength. Batch A4 shows good release profile whichexactly fit in our objective and also shows goodadhesive strength (0.252 N) which was sufficient toretain the tablet in vagina for more than 8 hrs.

Accelerated stability study of optimized batchKetoconazole effervescent mucoadhesive vaginaltablets were subjected to accelerated stability studies inaluminum pack as aluminum strip is considered thebest protecting packaging material but in the presentstudy simulation was made using aluminum foilpouch. As the dosage form is formulated formucoadhesive vaginal drug delivery, no change shouldoccur in its mucoadhesive strength and drugdissolution profile. The effect of aging was studied foroptimized formulation. The storage condition wasoptimized at 40 ± 0.5°C and 75 ± 5 % RH for 3months.

The data reveals no mark change in hardness, drugcontent and in vitro release. A slight reduction inresidence time was noticed but no much obviousalteration. The nearly constant ‘n’ value reveals thatthe release behavior from dual componentmucoadhesive system is not affected by the storagecondition.

CONCLUSIONThe results of the study reveal that with the developedformulations, the drug release and mucoadhesionproperties of mucoadhesive vaginal tablets can becontrolled by changing the polymer type, polymerconcentration and effervescent content. Ex-vivoretention studies justified the prolong retention of thetablet inside the vaginal tract. Consequently, themucoadhesive form of the drug would increase thetime of contact with the vaginal mucosa and thus itstherapeutic effect.The results of 32 full factorial design revealed that thepolymer ratio and amount of effervescent significantlyaffect the responses, mucoadhesive strength, %swelling, Q8 and t80. Full and reduced models werederived for the prediction of the response variable, Y.Based on result of multiple linear regression analysis,it was concluded that satisfactory mucoadhesion andgood drug release profile of tablet could be obtainedwhen X1 kept optimum level and X2 kept at low level.The effervescent ketoconazole tablet showedsignificantly higher (p<0.05) in-vitro antifungalactivity as compare to Candid®-V3 tablet and Candid®-V gel.

TABLE I : FULL FACTORIAL DESIGN LAYOUTVariable level in coded formBatch code

X1 X2

A1 -1 1A2 -1 0A3 -1 1A4 0 -1A5 0 0A6 0 1A7 1 -1A8 1 0A9 1 1

Check point - 0.3 0.7Coded value HPMC K4M: Chitosan

X1Amount of

effervescentX2

-1 1:0.5 500 1:1 751 1:1.5 100

Check point 1:0.8 92.5

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 663

TABLE II : COMPOSITION OF 32 FULL FACTORIAL DESIGN BATCHES

Quantity (mg)Ingredients

A1 A2 A3 A4 A5 A6 A7 A8 A9

Ketoconazole 400 400 400 400 400 400 400 400 400

HPMC K4M 50 50 50 50 50 50 50 50 50

Chitosan 25 25 25 50 50 50 75 75 75

MCC 252.5 227.5 202.5 227.5 202.5 177.5 202.5 177.5 152.5

Sod. bicarbonate 37.5 66.3 75 37.5 66.3 75 37.5 66.3 75

Citric acid 12.5 18.7 25 12.5 18.7 25 12.5 18.7 25

Mg. stearate 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5 7.5

Talc 15 15 15 15 15 15 15 15 15

TABLE III: SUMMARY OF RESULTS OF REGRESSION ANALYSIS

Coefficients for mucoadhesive strengthModelb0 b1 b2 b12 b11 b22 R2

FM 0.221 0.0412 -0.054 -0.011 -0.029 -0.027 0.9936RM 0.221 0.0411 -0.055 - -0.029 -0.027 0.9781

Coefficients for % swelling (4hrs.)b0 b1 b2 b12 b11 b22 R2

FM 20.3 3.083 1.608 -0.313 -1.15 -0.735 0.9915RM 19.81 3.083 1.608 - -1.15 - 0.9725

Coefficients for Q8

b0 b1 b2 b12 b11 b22 R2

FM 81.99 -22.541 4.431 2.84 -5.435 0.235 0.9915RM 78.52 -22.541 4.432 - - - 0.9637

Coefficients for t80

b0 b1 b2 b12 b11 b22 R2

FM 7.7 2.9 -0.66 -0.397 1.03 0.04 0.9994RM 7.72 2.9 -0.66 -0.397 1.03 - 0.9993FM indicates full model and RM indicates reduced model

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 664

Figure-1: The scheme of the device used in the mucoadhesion studie

TABLE IV: CALCULATION FOR TESTING THE MODEL INPORTION (ANOVA)For mucoadhesive strengthRegression DF SS MS F R2

FM 5 0.0311 0.0062 93.59 0.9936 Fcal=5.0RM 4 0.0306 0.0076 44.75 0.9781 Ftab=10.13

Error DF=(1,3)FM 3 0.0002 0.0001 - -RM 4 0.0007 0.0002 - -

For % swellingRegression DF SS MS F R2

FM 5 76.6782 15.3356 70.19 0.9915 Fcal=3.366RM 3 75.2071 25.0690 58.94 0.9725 Ftab=9.55Error DF=(2,3)FM 3 0.6554 0.2185 - -RM 5 2.1265 0.4253 - -For Q8

Regression DF SS MS F R2

FM 5 3258.0497 651.6099 70.25 0.9915 Fcal=3.286RM 2 3166.5984 1583.2992 79.65 0.9637 Ftab=9.28Error DF=(3,3)FM 3 27.8253 9.2751 - -RM 6 119.2766 19.8794 - -For t80

Regression DF SS MS F R2

FM 5 55.8306 11.1661 968.86 0.9993 Fcal=0.278RM 4 55.8274 13.9569 1477.89 0.9993 Ftab=10.13Error DF=(1,3)FM 3 0.0346 0.0115 - -RM 4 0.0378 0.0094 - -

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 665

Figure-2: Release profile of ketoconazole from factorial design batches (A1 to A9)

Figure-3: Response surface plot for % swelling

Figure-4: Response surface plot for Q8

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 666

Figure-5: Response surface plot for Q8

Figure-6: Response surface plot for t80

Figure-7: Zone of inhibition (mm) for KEMV vaginal tablet, marketed vaginal Candid®-V3 tablet and Candid®-Vgel. Error bar represent SD (n=3), KEMV tablet- ketoconazole effervescent mucoadhesive vaginal tablet

Patel Geeta M.et al /Int.J. PharmTech Res.2010,2(1) 667

REFERENCES

1. Ahuja A., Khar R., Ali J. Mucoadhesive drugdelivery systems. Drug Dev. Ind. Pharm.,1997, 23, 489–415.

2. Richardson M.D., Warnock D.W. FungalInfection-Diagnosis and Management,Blackwell Scientific Publications: London,1993.

3. Tawfique K., Danesment T.K., Warnock D.W.Clinical pharmacokinetics of ketoconazoleClin. Pharmacol., 1988, 14, 13–14.

4. Esclusa-D´ıaz M.T, Guimaraens-Mendez M,Perez-Marcos M.B, Vila-Jato J.L, Torres-Labandeira J.J. Characterization and in vitrodissolution behaviour of ketoconazole/and 2-hydroxypropyl-cyclodextrin inclusioncompounds. Int. J. Pharm., 1996, 143,203–210.

5. Karasulu H.Y., Taneri F., Sanal E., Guneri T.,Ertan G. Sustained release bioadhesiveeffervescent ketoconazole microcapsulestabletted for vaginal delivery. J.Microencapsul., 2002, 19,357–362.

6. Kast C.E., Valenta C., Leopold M., Bernkop-Schnurch A. Design and in vitro evaluation ofa novel bioadhesive vaginal drug deliverysystem for clotrimazole. J. Control. Rel., 2002,81,347–354.

7. Karasulu H.Y., Hilmioglu S., Metin D.Y,Guneri T. et al. Efficacy of a newketoconazole bioadhesive vaginal tablet onCandida albicans. II Farmaco, 2004, 59, 163–167.

8. Lachman. L., Lieberman Kinig. J.L. TheTheory and Practice of Industrial Pharmacy,4th edition, Varghese Publishing House:Bombay, 1999.

9. Gurny R., Meyer J.M., Peppas N.A.Bioadhesive intraoral release systems: design,

testing and analysis. Biomaterials, 1984,5:336–340.

10. Gursoy A., Sohtorik I., Uyanik N., PeppasN.A. Bioadhesive controlled release systemsfor vaginal delivery. STP Pharma., 1989, 5,886–892.

11. Noha Adel, Nafee, Fatma, Ahmed Ismail,Nabila Ahmed, Boraie Lobna, MohamedMortada. Mucoadhesive delivery systems. I.Evaluation of mucoadhesive polymers forbuccal tablet formulation. Drug. Dev. Ind.Pharm., 2004, 30(9), 985-993.

12. Tang X., Wang L. A novel ketoconazolebioadhesive effervescent tablet for vaginaldelivery: Design, in vitro and ‘in vivo’evaluation; International Journal ofPharmaceutics, 2008,350,181–187.

13. Korsmeyer R.W., Gurny R., Doelker E., BuriP., Peppas N.A. Mechanisms of solute releasefrom porous hydrophilic polymers. Int. J.Pharm., 1983, 15, 25-35.

14. Alam M. A., Ahmad F. J., Khan Z. I., Khar R.K, Ali. M. Development and evaluation ofacid-buffering bioadhesive vaginal tablet formixed vaginal infections. AAPSPharmSciTech, 2007, 8 (4), E1-E8.

15. Swamy P.A., Areefulla S.H., Shrisand S.B.,Gandra S., Prashanth B. Orodispersible tabletsof meloxicam using superdisintegrant blendsfor improved efficiency. Ind J Pharm Sci.,2007, 69(6), 836-840.

16. Bottenberg P., Cleymaet R., De Muynck C.,Remon J.P., Coomans D., Michotte Y., SlopD. Development and testing of bioadhesive,fluoride-containing slow-release tablets fororal use. J. Pharm. Pharmacol., 1991, 43,457–464.

17. Endenhall W., Sincich T. Multiple regressionsIn: A Second Course in Business Statistics,Regression Analysis. 3rd ed., CA: DellenPublishing Co: San Francisco, 1989.

*****