See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/332266903

Development and Optimization of Acyclovir Loaded Mucoadhesive Microspheres

by Box -Behnken Design

Article  in  Dhaka University Journal of Pharmaceutical Sciences · April 2019

DOI: 10.3329/dujps.v18i1.41421

CITATIONS READS
18 342

5 authors, including:

James Regun Karmoker Ikramul Hasan

University of Oklahoma Health Sciences Center University of Dhaka
11 PUBLICATIONS   146 CITATIONS    19 PUBLICATIONS   94 CITATIONS   

SEE PROFILE SEE PROFILE

Nusrat Ahmed Mohammad Saifuddin

University of Dhaka Bangladesh University of Business and Technology
9 PUBLICATIONS   36 CITATIONS    13 PUBLICATIONS   29 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Formulation and In-Vitro Characterization Polymeric Microspheres DDS View project

Formulation Development View project

All content following this page was uploaded by Nusrat Ahmed on 08 April 2019.

The user has requested enhancement of the downloaded file.

 Development and Optimization of Acyclovir Loaded
Mucoadhesive Microspheres by Box – Behnken Design
James Regun Karmoker1, Ikramul Hasan2, Nusrat Ahmed1, Mohammad
Saifuddin3 and Md. Selim Reza2
1
Department of Pharmacy, School of Medicine, University of Asia Pacific, Dhaka-1205, Bangladesh
2
Department of Pharmaceutical Technology, Faculty of Pharmacy, University of Dhaka
Dhaka-1000, Bangladesh
3
Department of Mathematics and Statistics, Bangladesh University of Business and Technology
Rupnagar, Mirpur-2, Dhaka-1216, Bangladesh

(Received: August 13, 2018; Accepted: December 12, 2018; Published (Web): April 01, 2019)

ABSTRACT: The grail of the study was to design, develop and characterize sustained release mucoadhesive
microspheres of acyclovir and to optimize the drug release profile using response surface methodology by applying
Box–Behnken design (BBD) which was equipped with three levels and three factors. Microspheres were prepared
from Methocel K15M and Ethocel Standard 45 Premium using the emulsification solvent evaporation technique. The
independent factors were the amounts of Methocel K15M (X1), amount of Ethocel Standard 45 Premium (X2), and
RPM (X3). The dependent variables were cumulative percentage drug release (CDR) at 8 hour (Y1), bond strength
(Y2), and swelling at 4 hour (Y3). To understand the effects of different factor level combinations on the responses,
various response surface graphs and contour plots were prepared. Predicted values and experimental values for
optimized formulation (X1 = 600 mg, X2 = 500 mg, and X3 = 336.57) was found to be in close agreement.

Key words: Acyclovir, Box-Behnken Design, kinetic models, microspheres, mucoadhesion, optimization.

INTRODUCTION
After 1950s there has been an incessant turn body site. Microspheres constitute an important part
down in the launch of new chemical moieties, the of particulate DDS by virtue of their small size and
reasons could be many to name a few would be efficient carrier characteristics despite having
research and development costs which is escalating limitation due to their short residence time at the site
day by day, less number of organizations conducting of absorption. Therefore, by providing an intimate
pharmaceutical research, term of patent etc. It is contact of the DDS with the absorbing membranes
estimated that the whole cycle involved in the which can be achieved by coupling bioadhesion
marketing of a drug i.e., discovery, clinical phase, characteristics to microspheres would be
regulatory approval costs around $ 120 million, advantageous.2
estimated time could be around 10 years while drug The development of an effective DDS, however,
delivery systems would cost around 40%.1 Drug perpetually involves rational unification of a plethora
delivery systems (DDS) have had a mammoth impact of polymers and excipients. Optimizing the
on the healthcare system due to its precise capacity to formulation composition and the manufacturing
control the release rates or target drugs to a specific process of such a drug delivery product to embellish
Correspondence to: Md. Selim Reza the desired quality persona is, therefore, a herculean
Tel: 880-2-9661900, Ext. 8182; Fax: 880-2-9667222 task. The conventional approach of optimizing a
E-mail: selimreza@du.ac.bd
formulation or process essentially involves studying
Dhaka Univ. J. Pharm. Sci. 18(1): 00-00, 2019 (June) the influence of one variable at time (OVAT), while
2 Karmoker et al.

keeping others as stable. Using this OVAT approach, intravenously to obtain the desired therapeutic
the solution of a specific problematic property can be effect.14 ACV is almost completely unionized, has
achieved somehow, but attainment of the true maximum solubility 2.5 mg/ml at pH 7.0, soluble in
optimum composition or process is never acidic pH and is predominantly absorbed from the
guaranteed.3 In order to overcome most shortcomings upper gastrointestinal tract (GIT). ACV is currently
inherent to the traditional OVAT approach, design of marketed as capsules (200 mg), tablets (200, 400 and
experiments (DoE), the holistic approach encircling 800 mg) and topical ointment. Oral acyclovir is
the application of appropriate experimental designs typically used as 200 mg tablets, five times a day. In
coupled with the generation of mathematical addition, for immunocompromized patients with
equations and graphic outcomes, and depicting a relapsing herpes simplex infection, long-term
complete picture of variation of the response(s) as a administration of acyclovir (6 months or longer) are
function of the factor(s), is being widely practiced.4 required. The currently available conventional
Of late, for formulation optimization, the response therapy is coupled with a number of limitations such
surface methodology (RSM), using proper DoE, has as highly variable absorption and low bioavailability
become extensively used. For understanding the (10-20%) after oral administration. Moreover,
effects of formulation variables (independent factors) decrease in bioavailability was observed when dose is
and the reciprocal actions between factors on the increased. Furthermore, because the mean plasma
responses (dependent factors), Box–Behnken design half-life of the drug is 2.5 h, five times a day
(BBD) is considered as one of the appropriate RSMs. administration is required which causes compliance
In BBD, factors are usually taken at three levels and problems to patients.11,15
all the design points fall within the safe operating Therefore, the aim of the study was to develop
zone. BBD is still considered to be more proficient mucoadhesive sustained-release formulations of ACV
and most powerful than other designs such as the in the form of microspheres using BBD in
three-level full factorial design, central composite combination with a desirability function and to
design (CCD) and doehlert design, despite its poor evaluate the main effects of formulation variables on
coverage of the corner of nonlinear design space.5,6 three responses: cumulative percentage of drug
Moreover, it demands less experimental runs than release at 8 hour, bond strength and swelling at 4
three-level full factorial design and CCD, and is hour.
consequently less exclusive. To optimize the
preparation of oral controlled release delivery
MATERIAL AND METHODS
systems7 and sustained-release pellets,8 it has been
Materials. Acyclovir was supplied from Incepta
productively used.
Pharmaceuticals Ltd. (Dhaka, Bangladesh). Ethocel
Acyclovir [9-(2-hydroxyethoxymethyl) guanine]
standard 45 premium and Methocel K-15M were
(ACV), is a synthetic purine nucleoside analog
obtained from Colorcon Asia Pvt. Ltd, USA. Ethanol,
derived from guanine9,10 which is considered as the
Dichloromethane (DCM), sorbitanmonooleate (Span
first agent to be licensed for the treatment of herpes
80), heavy liquid paraffin and n-hexane were
simplex virus (HSV-1, HSV-2) infections and is the
supplied by Merck, Germany. All other chemicals
most widely used drug for infections such as
were of analytical grade.
cutaneous herpes, genital herpes, chicken pox,
Preparation of acyclovir microspheres. The
varicella zoster infections through interfering with
method of preparation of acyclovir microspheres was
DNA synthesis and inhibiting viral replication.11-13
based on emulsion solvent evaporation method using
According to the Biopharmaceutical Classification
two different polymers: Ethocel Standard 45
System (BCS), ACV falls under the BCS Class III
premium and Methocel K15M. Weighed quantities of
drug i.e. soluble with low intestinal permeability and
polymers were dissolved in 20 ml solvent mixture
needs to be administered in large doses orally or
Development and Optimization of Acyclovir Loaded Mucoadhesive 3

(DCM:Ethanol = 1:1). Then the required amount of precision of the design, BBD requires 15
acyclovir was added and dissolved by vortexing and experimental runs with three central points.5 Using
sonication until a clear solution was formed. This Design-Expert software (V. 7.0.0.1; Stat-Ease Inc.,
solution was the internal phase. For the preparation of Minneapolis, Minnesota), a total of 15 experimental
external phase, 50 ml heavy liquid paraffin runs were generated and evaluated. The major
emulsified with 1 ml Span 80 was taken in a 500 ml response factors used to evaluate the quality of the
beaker and stirred using a stirrer. The internal phase microspheres formulation, including cumulative
was then slowly poured drop by drop to the external percentage drug release (CDR) at 8 hour (Y1), bond
phase, while stirring at required rpm held by the strength (Y2), and swelling at 4 hour (Y3), were
mechanical stirrer equipped with a three-blade determined (Table 1). The selected factors with the
propeller, at room temperature. The whole system actual and coded levels according to the design are
was stirred for 3 hours. After that the microspheres represented in table 1 and 2. The results obtained for
were separated by filtration, the excess of paraffin oil each response were fitted to a quadratic polynomial
was eliminated by repeated washing (3 times) with model explained by a nonlinear Eq. (1):
n-hexane (50 ml) and finally dried overnight at room
y = β0 + β1X1 + β2X2 + β3X3 + β4X1X2 + β5X2X3 +
temperature to yield free flowing spherical β6X1X3 + β7X12 + β8X22 + β9X32 (1)
products.16-18
Box-Behnken experimental design. To explore Here, y represents the measured response, β0– β9
and optimize the main effects, quadratic effects, and are regression coefficients and X1, X2, and X3 are
interaction effects of the formulation ingredients on independent factors. By applying analysis of variance
the performance of the mucoadhesive micorspheres, a (ANOVA), lack of fit, and coefficient of
three-factor, three-level BBD was used. In such determination (R2) as a measure of goodness of fit of
settings, to determine the experimental error and the the fitted model, models were validated.

Table 1. Variables used in the Box-Behnken design.

Levels, actual (Coded)

Low (-1) Medium (0) High (+1)
Independent variables
X1: Amount of Methocel K15M (mg) 400 500 600
X2: Amount of Ethocel Standard 45P (mg) 400 450 500
X3: RPM 300 350 400
Goals
Dependent variables
Y1: CDR at 8 hour (%) Minimize
Y2: Bond strength (N/m2) Maximize
Y3: Swelling at 4 hour (%) Minimize

Optimization using the desirability function. any bias. In the current study, all the responses were
To optimize manifold responses, they should be concurrently optimized by a desirability function that
highly interconnected with each other. It is uses the numerical optimization method introduced
improbable that the values enviable to optimize the by Derringer and Suich in the Design-Expert
effect of one response will have same effect on the software (Stat-Ease Inc.). 19 Recently, for the
second response, thus a variance can occur between optimization of multiple responses, desirability
them. Hence, the most favorable compromising zone function approach was reported in several articles.5,8
must be required for each of the responses devoid of
4 Karmoker et al.

In vitro dissolution study. The dissolution replaced with an equivalent amount (10 ml) of fresh
studies of the microspheres were carried out in a type dissolution medium to maintain sink condition.
I USP dissolution test apparatus (Basket type). Collected samples were analyzed by measuring the
Weighed quantities of microspheres (equivalent to 20 absorbance through an UV spectrophotometer at 255
mg of acyclovir) from each batch were placed in 900 nm after filtration and suitable dilution to determine
ml of dissolution medium (0.1N HCl of pH 1.2) and the amount of the acyclovir released from the
were stirred at 100 rpm at 37 ± 0.5°C.9,20 A 10 ml microspheres. The percentage of drug release was
aliquot was withdrawn from the dissolution medium plotted against time. Each experiment was repeated
at pre-determined intervals of 1st hour, 2nd hour, 3rd three times. To find percentage of release, the average
hour, 4th hour, 5th hour, 6th hour, 7th hour and 8th hour. of the percentage of release was calculated for each
At each interval, the withdrawn medium was batch.

Table 2. Experimental matrix and observed responses from randomized runs in the BBD.

Run Independent variables Dependent variables

X1 X2 X3 Y1 Y2 Y3
(mean ± SD, n = 3) (mean ± SD, n = 3)
1 0 1 1 81.27±0.89 142.75 189.37±1.65
2 -1 1 0 86.77±1.01 181.92 212.48±3.56
3 -1 0 -1 88.18±1.00 112.87 215.75±1.25
4 0 0 0 85.16±0.89 167.98 200.68±1.77
5 0 0 0 84.32±0.76 159.35 198.55±4.84
6 0 -1 -1 86.22±1.02 146.07 202.99±2.43
7 1 1 0 69.15±0.90 187.23 174.02±4.05
8 1 0 1 76.31±0.90 136.11 182.58±3.68
9 0 1 -1 77.72±1.39 156.03 186.65±2.32
10 -1 -1 0 97.10±1.00 126.81 301.54±1.85
11 1 -1 0 79.31±1.12 167.98 187.21±3.69
12 -1 0 1 92.38±0.90 120.17 220.04±2.37
13 0 0 0 86.08±1.00 177.27 202.90±4.42
14 1 0 -1 73.86±1.13 146.73 181.72±1.75
15 0 -1 1 87.48±1.01 128.80 213.27±3.72

Figure 1. Determination of mucoadhesive strength by modified balance method.

Development and Optimization of Acyclovir Loaded Mucoadhesive 5

Ex-vivo mucoadhesion test. A modified balance at pre-determined intervals of 1st hour, 2nd hour, 3rd
method (Figure 1), adopted from Kyada et al. was hour, 4th hour and 5th hour and weighed after
used for determining the mucoadhesive strength.9 removing excess water with the help of filter paper.21
Fresh goat intestinal mucosa was obtained from a Then, the swelling percentage was calculated as per
local slaughterhouse and used within 2 hours of the Eq. (4):
slaughter. By removing the underlying fat and loose
tissues, the mucosal membrane was separated. The % Swelling = [(Final weight – Initial weight) × 100] /
membrane was washed with distilled water and then Initial weight (4)
cut into pieces. Two pieces of goat stomach mucosa
were pasted to the bottom of two glass vials with RESULTS AND DISCUSSION
cyanoacrylate adhesive separately. From that one In vitro dissolution study. It has been found
glass vial was fixed on the surface of the wooden (Table 2) that the rate of drug release from the
support and other piece was tied with the balance on microspheres depends on the polymeric concentration
left hand side. A glass beaker was kept on right hand and indicated that the release rate decreased with the
side. The right and left sides were balanced by adding increasing amount of polymer. The decrease in
extra weight on both the left and right hand side. release rate with increasing polymeric concentration
Approximately 100 mg of microspheres from each can be described by a decreased amount of drug
batch was placed between these glass vials containing present close to the surface and also by the fact that
goat stomach mucosa, and 3 drops of 0.1N HCl of pH with the increasing of polymer concentration the
1.2 were added on it for wetting. Then, two vials amount of uncoated drug decreases. Release data
were gently pressed to remove the presence of air. were analyzed with statistical level of significance (α
The balance was kept in this position for 5 minutes. = 0.05) and they were found statistically significant
As the quantity of microspheres was less, since in every case “p” value was found less than
microspheres formed the monolayer along the surface 0.05. Moreover, increased polymeric concentration as
of disk. Water was added slowly at 1ml/min to the well as high viscosity of the spheres also contributes
right-hand pan until the microsphere detached from to lower drug release. Here, in some cases for two
the goat stomach mucosal membrane. The water in different batches having the same polymeric
ml (1 ml equivalent to 1 gram) required to detach the concentration showed different pattern of drug
microspheres from the mucosal surface gave the release. RPM could be a probable reason for such
measure of mucoadhesive strength.9 The characteristics. When RPM is decreased, sizes of the
mucoadhesive strength was calculated using Eq. (2) spheres are increased that causes an increase in
and (3): porosity, facilitate more diffusion and finally results
in increased drug release.22,23 RPM can also
Force of adhesion (N) = [mucoadhesion strength
(gm) × 9.81] / 1000 (2) contribute in other way around on drug release. With
decreasing RPM, size is increased as a result surface
Bond strength (N/m2) = force of adhesion / surface area is decreased that slows the burst release and
area of the bottom of vial (3)
minimizes the drug release.
Swelling measurement. Microspheres were studied Ex vivo mucoadhesion test. The data for
for swelling characteristics. Approximately 100 mg mucoadhesive strength in gm (mean ± SD, n = 3) for
drug-loaded microspheres from each batch were run 1 to run 15 were found to be 7.17 ± 0.21, 9.13 ±
placed separately in a vessel containing 100 ml 0.1N 0.35, 5.67 ± 0.25, 8.43 ± 0.45, 8.00 ± 0.46, 7.33 ±
HCl of pH 1.2 and temperature maintained at 37 ± 0.35, 9.40 ± 0.36, 6.83 ± 0.25, 7.83 ± 0.35, 6.37 ±
0.5°C. The microspheres were periodically removed 0.25, 8.43 ± 0.35, 6.03 ± 0.35, 8.90 ± 0.20, 7.37 ±
6 Karmoker et al.

0.40, 6.47 ± 0.35, respectively whereas bond strength demonstrate that polymeric concentration in the
in N/m2 was reported in table 2. The range for formulation maintains a reciprocal relationship with
mucoadhesive strength of run 1 to 15 was found to be percentage swelling of that formulation. It has been
5.67 ± 0.252 to 9.40 ± 0.361 gram. Results showed previously reported that the higher the swelling of the
that at highest polymeric concentration (e.g. 1100 mg polymers, the higher is the drug release from the
for run 7), bond strength (N/m2) was maximum. The microspheres.21
order of bond strength of fifteen formulations also Box-Behnken statistical analysis. A 3-factor, 3-
depicts that Methocel K15M played an important role level BBD requires 15 experimental runs. Based on
in the matrix mucoadhesion as everywhere the experimental runs generated at different factor
concentration of Methocel K15M was higher than level combinations, a series of experiments was
corresponding Ethocel standard 45P except run 2. performed. In table 2, the experimental matrix from
Methocel K15M, one of the most widely used the randomized runs for the independent variables
mucoadhesive materials, has the capacity to generate and responses observed is exposed. Ten batches
carboxyl groups which facilitate the formation of showed % CDR (Y1) of microspheres greater than 80
hydrogen bonds with mucus. % and the range of Y1 for all batches was 69.15% -
Several studies have demonstrated that acyclovir 97.10 %. Similarly, the range for bond strength (Y2)
is absorbed paracellularly by passive diffusion in the was 187.23 – 112.87 N/m2 and the percentage
gastrointestinal tract and its poor solubility results in swelling at 4 hour (Y3) was 174.02% – 301.54%.
low oral bioavailability.15,24-26 Many in vitro everted All responses were fitted to a second quadratic
sac and in situ single-pass perfusion experiments model and the adequacy of this model was verified
showed that absorption of acyclovir in the by ANOVA, lack of fit and coefficient of
gastrointestinal tract primarily takes place at the determination (R2) tests. The results of lack of fit
upper and middle part of gastrointestinal tract. Hence, tests and ANOVA of the quadratic models for all
it is an efficient way to enhance the bioavailability of responses are revealed in table 3. Here, in the
acyclovir by introducing gastro retentive dosage form ANOVA test, the p values of F-statistic of the model
which can enable continuous drug delivery to the for responses Y1, Y2 and Y3 were 0.0003, 0.0252, and
upper part of the gastrointestinal tract. It has been 0.0495 respectively. As a consequence, from the p
reported that incorporation of mucoadhesive values for this model it can be accomplished that all
materials into microspheres significantly increase the of the responses (Y1, Y2 and Y3,) fitted the quadratic
gastrointestinal transit time of microspheres.2,9 model well (p<0.05). Moreover, the lack of fit test is
Swelling measurement. As shown in table 2, another excellent statistical parameter for checking
the microspheres from run 7 exhibited lowest better fitness of the model. It juxtaposes the residual
percentage swelling (174.02%) at 4 hour while the error to the pure error from the replicated design
highest swelling was noticed in run 10 (301.54%). points. A model with a significant lack of fit (p value
Maximum swelling for microspheres was reached at > 0.10 or smaller) lacks prediction efficiency, so a
4 hour, after which erosion or breakdown took place non-significant lack of fit value in the model is highly
since most of the formulations showed less desirable. All of the responses fitted in the quadratic
percentage swelling at the 5th hour in comparison model showed a non-significant lack of fit (p > 0.1)
with their corresponding percentage swelling at 4th except Y3, proving the adequacy of the model fit.
hour. Md et al. experienced same sort of swelling Table 4 shows a summary of the multiple regression
behavior for acyclovir in acidic medium.21 analysis of the responses for the second-order
This effect might be due to the acid solubility of quadratic model. The R2 values signify the measure
drug that might have influenced the swelling of the amount of variation around the mean explained
behavior of the microspheres. These results also by the model. In this study, the R2 values for the
Development and Optimization of Acyclovir Loaded Mucoadhesive 7

responses Y1, Y2 and Y3 were 99%, 92% and 90%, the design space, a ratio greater than 4 is desirable.
respectively. Signal-to-noise ratio, a measure of the The ratios of “adequate precision” for Y1, Y2 and Y3
range of a predicted response relative to its associated were 22.676, 7.397 and 8.508, respectively,
error, is called “adequate precision”. For navigating indicating an adequate signal.

Figure 2. Effects of X1 and X2 on the % CDR at 8 hour at the mid-level of X3: (A) contour plot and (B) its response surface.

Figure 3. Effects of X1 and X2 on the bond strength (N/m2) at the mid-level of X3: (A) contour plot and (B) its response surface.
8 Karmoker et al.

Figure 4. Effects of X1 and X2 on the % swelling at 4 hour at the mid-level of X3: (A) contour plot and (B) its response surface.

Figure 5. Overall desirability (D) as a function of X1 and X2: (A) contour plot and (B) its response surface.

Coefficient estimation and equations for in terms of coded factors where intercept is 85.19,
responses. The responses obtained at various levels coefficient value for A:X1 is -8.23, B:X2 is -4.40 and
of independent variables were subjected to multiple C:X3 is 1.43; quadratic polynomial Eq. (6) for Y2 in
regression to give quadratic polynomial Eq. (5) for Y1 terms of coded factors where intercept is 168.20,
Development and Optimization of Acyclovir Loaded Mucoadhesive 9

coefficient value for A:X1 is 12.03, B:X2 is 12.28 and coefficient value for A:X1 is -28.04, B:X2 is -17.81
C:X3 is -4.23 and quadratic polynomial Eq. (7) for Y3 and C:X3 is 2.27.
in terms of coded factors where intercept is 200.71,

Table 3. ANOVA and lack of fit tests of the quadratic model for the responses.

Response F Value Probability > F

Y1
Model 43.42 0.0003a
Lack of fit
3.33 0.2393b
Y2
Model 6.66 0.0252a
Lack of fit
1.71 0.3896b
Y3
Model
4.80 0.0495a
Lack of fit
92.62 0.0107a

a
Significance probability values (Probability > F) less than 0.05 implies that the model is significant; bNonsignificant lack of fit ( p value >
0.1) proves the adequacy of model fit

Table 4. Summary of the regression analysis of the responses.

Quadratic model R2 Adjusted R2 Predicted R2 Adequate precision SD CV (%)

Y1 0.9874 0.9646 0.8268 22.676 1.36 1.63
Y2 0.9230 0.7845 0.0652 7.397 10.7 7.11
Y3 0.8962 0.7093 -0.6507 8.508 16.27 7.95

Y1 = 85.19 – 8.23 × A – 4.40 × B + 1.43 × C + 0.043 × A × B – 0.44 × A × C

+ 0.57 × B × C – 1.30 × A2 – 0.81 × B2 – 1.21 × C2 (5)

Y2 = 168.20 + 12.03 × A + 12.28 × B – 4.23 × C – 8.96 × A × B – 4.48 × A × C

+ 1.00 × B × C – 8.33 × A2 + 6.11 × B2 – 30.90 × C2 (6)

Y3 = 200.71 - 28.04 × A - 17.81 × B + 2.27 × C + 18.97 × A × B – 0.86 × A × C

- 1.89 × B × C + 10.03 × A2 + 8.08 × B2 – 10.72 × C2 (7)

Response surface and contour plot analysis. will be influenced by the change in the level of
To further elucidate the relationship between the another factor. Here, one independent variable must
independent and dependent variables, two always be fixed as these types of plots can only
dimensional contour plots and three-dimensional express two independent variables at a time against
response surface plots of the responses across the the response.27
selected factors were constructed as shown in figures Optimization and evaluation of the optimized
2, 3 and 4. These types of plots are very useful for formulation. The independent variables were
studying the interaction effects between two factors simultaneously optimized for all three responses by
and for understanding how the effect of one factor using the desirability function after studying the
10 Karmoker et al.

effects of the dependent and independent variables on individual desirability function as the geometric
the responses. Responses Y1, Y2 and Y3 were mean by abroad grid search and practicability search
transformed into individual desirability scales d1, d2 over the domain by the Design-Expert software (Stat-
and d3, respectively. Constraints were set against all Ease Inc.), the global desirability value was
of the responses. Among the responses, Y2 was set to calculated. Figure 5 shows the response surface and
be maximized and Y1 and Y3 were set to be counter plot for the desirability function holding the
minimized. Equal weight and importance were given variable X1 and X2.
to all of the responses. Finally, by combining the

Table 5. Comparison of predicted and observed experimental values of acyclovir microspheres prepared under optimum conditions.

Responses Predicted value Observed value Residuals Bias* (%)

Y1 (%) 69.99 65.57 ± 3.15 -4.42 6.74
Y2 (N/m2) 181.18 178.81 ± 2.64 -2.37 1.33
Y3 (%) 191.29 185.10 ± 7.53 -6.19 3.34

*Bias (%) = [(predicted value – observed value) × 100] / observed value

The optimized formulation was achieved at X1 = continuous drug delivery to the upper part of the
600 mg, X2 = 500 mg and X3 = 336.57 RPM with the gastrointestinal tract, mucoadhesion could be a
corresponding desirability (D) value of 0.917. This resolution by means of applying Methocel K15M in
factor level combination predicted the responses Y1 = the formulation. Furthermore, the study successfully
69.99 (%), Y2 = 181.18 N/m2 and Y3 = 191.29 %. verified the use of the BBD combined with a
Finally, three batches of the optimized formulations desirability function for the optimization of different
were prepared to confirm the validity of the optimal responses of acyclovir loaded microspheres since the
parameters and predicted responses calculated. All of observed values were found to be in close agreement
the responses were evaluated for each optimized with the estimated values. Response surface plots and
formulation. The comparisons of predicted and contour plots were used to study the effects of
experimental results are shown in table 5. It can be different formulation variables on the responses and
seen that the experimental values were in very close the results obtained suggest that the RSM using the
agreement with the predicted values, indicating the BBD could be a suitable approach for understanding
triumph of the BBD pooled with a desirability formulation variables and for optimizing the
function for the assessment and optimization of formulation efficiently. However, this study only
acyclovir loaded microspheres formulations. investigated on in vitro and ex vivo settings. Further
study focusing on in vivo settings can be taken into
CONCLUSION consideration for better predictability of drug
delivery in a biological system.
The present study lucratively demonstrated that
using Methocel K15M and Ethocel Standard 45
Premium as polymeric materials, mucoadhesive ACKNOWLEDGEMENT
microspheres, able to promote a sustained release of Authors are grateful to their designated institute
acyclovir were produced by emulsion-solvent for providing the necessary facilities to accomplish
evaporation technique which could be a smart the research work. Incepta Pharmaceuticals Ltd,
solution to the pharmacokinetic limitation of Bangladesh is gratefully acknowledged for providing
acyclovir. The results obtained indicate that drug with the active ingredient.
release can be slowed down by increasing polymeric
concentration in the formulation. Moreover, for
Development and Optimization of Acyclovir Loaded Mucoadhesive 11

CONFLICT OF INTEREST 12. Jalón, E.G., Blanco-Prı́, M.J., Ygartua, P. and Santoyo, S.
2003. Increased efficacy of acyclovir-loaded microparticles
The authors declare that there is no conflict of
against herpes simplex virus type 1 in cell culture. Eur. J.
interests regarding the publication of this paper. Pharm. Biopharm. 56, 183-187.
13. Rokhade, A.P., Patil, S.A. and Aminabhavi, T.M. 2006.
REFERENCES Synthesis and characterization of semi-interpenetrating
polymer network microspheres of acrylamide grafted dextran
1. Tiwari, G., Tiwari, R., Sriwastawa, B., Bhati, L., Pandey, S.,
and chitosan for controlled release of acyclovir. Carbohydr.
Pandey, P. and Bannerjee, S.K. 2012. Drug delivery systems:
Polym. 67, 605-613.
an updated review. Int. J. Pharm. Investig. 2, 2-11.
2. Vasir, J.K., Tambwekar, K. and Garg, S. 2003. Bioadhesive 14. Jana, S., Sharma, R., Maiti, S. and Sen, K.K. 2016.

microspheres as a controlled drug delivery system. Int. J. Interpenetrating hydrogels of O-carboxymethyl tamarind

Pharm. 255, 13-32. gum and alginate for monitoring delivery of acyclovir. Int. J.
Biol. Macromol. 92, 1034-1039.
3. Singh, B., Kumar, R. and Ahuja, N. 2005. Optimizing drug
15. Dhaliwal, S., Jain, S., Singh, H.P. and Tiwary, A.K. 2008.
delivery systems using systematic "design of experiments"
Mucoadhesive microspheres for gastroretentive delivery of
part I: fundamental aspects. Crit. Rev. Ther. Drug Carrier.
acyclovir: in vitro and in vivo evaluation. The AAPS J. 10,
Syst. 22, 27-106.
322-330.
4. Singh, B., Bhatowa, R., Tripathi, C.B. and Kapil, R. 2011.
16. Akhter, S., Paul, S., Hasan, I., Ayon, N.J., Haider, S.S. and
Developing micro-/nanoparticulate drug delivery systems
Reza, M.S. 2013. Preparation, characterization and
using" design of experiments". Int. J. Pharm. Investig. 1, 75-
compatibility studies of naproxen loaded microspheres of
87.
cellulosic and polymethacrylic polymeric blend. Dhaka Univ.
5. Marasini, N., Yan, Y.D., Poudel, B.K., Choi, H.G., Yong,
J. Pharm. Sci. 12, 11-21.
C.S. and Kim, J.O. 2012. Development and optimization of
self-nanoemulsifying drug delivery system with enhanced 17. Ayon, N.J., Hasan, I., Islam, M.S. and Reza, M.S. 2014.

bioavailability by Box-Behnken design and desirability Preparation and characterization of gliclazide incorporated

function. J. Pharm. Sci. 101, 4584-4596. cellulosic microspheres: studies on drug release,
compatibility and micromeritics. Dhaka Univ. J. Pharm. Sci.
6. Ferreira, S.C., Bruns, R.E., Ferreira, H.S., Matos, G.D.,
13, 149-166.
David, J.M., Brandao, G.C., da Silva, E.P., Portugal, L.A.,
18. Saha, N., Hasan, I., Nazmi, M. and Reza, M.S. 2013. Design
Reis, P.S.D., Souza, A.S. and Santos, W.N.D. 2007. Box-
and development of sustained release microspheres of
Behnken design: an alternative for the optimization of
ibuprofen by emulsification solvent evaporation method
analytical methods. Anal. Chim. Acta. 597, 179-186.
using polymeric blend. Bangladesh Pharm. J. 16, 39-44.
7. Kim, M.S., Kim, J.S., You, Y.H., Park, H.J., Lee, S., Park,
19. Derringer, G. and Suich, R. 1980. Simultaneous optimization
J.S., Woo, J.S. and Hwang, S.J. 2007. Development and
of several response variables. J. Qual. Technol. 12, 214-219.
optimization of a novel oral controlled delivery system for
tamsulosin hydrochloride using response surface 20. Vinodbhai, P.K., Gohel, M.C., Parikh, R.K., Bariya, S. and
methodology. Int. J. Pharm. 341, 97-104. Suthar, R.N. 2011. Sustained release floating microspheres

8. Bodea, A. and Leucuta, S.E. 1998. Optimization of of acyclovir: formulation, optimization, characterization and

propranolol hydrochloride sustained-release pellets using in vitro evaluation. Int. J. Drug Dev. Res. 3, 242-251.

box-behnken design and desirability function. Drug Dev. Ind. 21. Md, S., Ahuja, A., Khar, R.K., Baboota, S., Chuttani, K.,
Pharm. 24, 145-155. Mishra, A.K. and Ali, J. 2011. Gastroretentive drug delivery
system of acyclovir-loaded alginate mucoadhesive
9. Kyada, C., Ranch, K. and Shah, D. 2014. Optimization of
2 microspheres: formulation and evaluation. Drug Deliv. 18,
mucoadhesive microspheres of acyclovir by applying 3 full
255-264.
factorial design. J. Drug Deliv. Sci. Tech. 24, 61-68.
22. Jalil, R. and Nixon, J.R. 1990. Microencapsulation using
10. Tao, Y., Lu, Y., Sun, Y., Gu, B., Lu, W. and Pan, J. 2009.
poly (l-lactic acid) II: preparative variables affecting
Development of mucoadhesive microspheres of acyclovir
microcapsule properties. J. Microencapsul. 7, 25–39.
with enhanced bioavailability. Int. J. Pharm. 378, 30-36.
11. Bhosale, U.V., Devi, K. and Choudhary, S. 2012. Multiunit 23. Freiberg, S. and Zhu, XX. 2004. Polymer microspheres for
floating drug delivery system of acyclovir: development, controlled drug release. Int. J. Pharm. 282, 1-18.
characterization and in vitro-in vivo evaluation of spray-dried 24. Meadows, K.C. and Dressman, J.B. 1990. Mechanism of
hollow microspheres. J. Drug Deliv. Sci. Tech. 22, 548-554. acyclovir uptake in rat jejunum. Phar. Res. 7, 299-303.
 12 Karmoker et al.

5. O’Brien, J.J. and Campoli-Richards, D.M. 1989. Acyclovir: 27. Wu, X.G., Li, G. and Gao, Y.L. 2006. Optimization of the
an updated review of its antiviral activity, pharmacokinetic preparation of nalmefene-loaded sustained-release
properties and therapeutic efficacy. Drugs. 37, 233–309. microspheres using central composite design. Chem. Pharm.
26. Park, G.B., Shao, Z. and Mitra, A.K. 1992. Acyclovir Bull. 54, 977-981.
permeation enhancement across intestinal and nasal mucosae
by bile salt-acylcarnitine mixed micelles. Pharm. Res. 9,
1262-1267.

View publication stats