IMPLEMENTING CENTRAL COMPOSITE DESIGN FOR THE DEVELOPMENT OF TACROLIMUS FILM FOR SUBLINGUAL ADMINISTRATION

Objective : The study aimed to develop fast-dissolving films (FDFs) of the immunosuppressant drug tacrolimus monohydrate for sublingual administration, employing central composite design (CCD), to improve its bioavailability. Methods: Tacrolimus: β-cyclodextrin inclusion complexes prepared earlier were formulated into FDFs. CCD was used for developing optimal film formulation with the desired characteristics. The solvent casting method was used for the preparation of films. For optimization, the independent variables selected were the concentration of hydroxy propyl methyl cellulose E5 (HPMC E5) (X 1 ) and concentration of croscarmellose sodium (CCS) (X 2 ) and the responses were disintegration time (Y 1 ) and percentage drug release at 5 min (Y 2 ). The suggested optimal films were subjected to further characterization. Results: All the formulations showed good mechanical properties. The composition of optimized FDF constituted 3.016% w/v of HPMC and 11.731%w/w of CCS and its average disintegration time was 27.28s and showed 83.13% mean drug release at 5 min. Differential Scanning calorimetry (DSC) analysis showed complete dispersion and partial conversion into the amorphous form of the drug, which was also confirmed by X-ray diffraction (XRD) studies. Scanning Electron Microscopy (SEM) revealed the smooth and porous nature of the film. Conclusion: The developed FDF may be used sublingually for delivering tacrolimus efficiently, avoiding its oral bioavailability problems.


INTRODUCTION
The sublingual route of administration has the advantage of bypassing first-pass metabolism and thereby increasing the bioavailability of drugs. Also, the sublingual mucosa is having relatively increased permeability due to the thin membrane and supply of vascular and lymphatic drainage. Additionally, reduced drug dosage to achieve similar target levels compared with oral administration may result in lower medical costs and patient compliance will be more especially in patients facing difficulty in swallowing and chewing [1,2].
The commonly used sublingual dosage forms are orodispersible films and tablets. As the fast-dissolving film (FDF) is administered through the sublingual route, rapid absorption of the drug is possible, which finally leads to the quick onset of drug action [3]. They are taken without water and due to their ability to disintegrate within a few minutes, they quickly release medication in the mouth [4,5]. Various superdisintegrants are also employed in the formulation of fast-dissolving films [6].
Tacrolimus is an immunosuppressant having poor and erratic bioavailability mainly due to low solubility, the influence of food intake and concomitant medication, and heapto first-pass metabolism. Tacrolimus has played a key role in today's high success rate of solid organ transplantation. A variety of clinical situations requiring nonoral medication delivery arise, presenting the need for reliable alternative routes of tacrolimus administration. Many clinical studies suggest the use of sublingual administration of tacrolimus producing comparable blood-drug levels with similar or even lower doses than the oral route in the lung as well as liver transplant recipients [7][8][9].
Response surface methodology (RSM) is an efficient technique for formulation optimization by efficiently exploring the relationships between the investigated factors and measured responses using the minimum number of experimental trials. Central composite design (CCD) is one of the most pertinent designs of RSM, which consists of factorial and axial points as well as one point at least at the center of the experimental region that provides properties like orthogonality and rotatability for fitting quadratic polynomials. This diversity of points is useful for providing complete knowledge of responses using the least number of experiments [10].
In the present work, an attempt was made to formulate tacrolimus into FDF formulation for sublingual application. CCD was employed to analyze the effect of the independent variables on the selected responses and to suggest the optimized production parameters for attaining a film formulation with desirable responses.

Formulation of sublingual ODFs
To increase the solubility of tacrolimus, it was converted to an inclusion complex with β-cyclodextrin as the carrier by the kneading method. The drug and the carrier were triturated with a mixed solution of 1:1 (v/v) methanol: water in a glass mortar to obtain a thin slurry. The slurry was kneaded thoroughly for around 40 min. After the evaporation of the solvents, the complexes were dried at room temperature for a period of 1 d [11].
FDF formulations loaded with β-cyclodextrin inclusion complex of tacrolimus monohydrate were formulated with HPMC E5 as the filmforming polymer, PEG 400 as the plasticizer, and CCS as the superdisintegrant [12]. Menthol was used as the flavouring agent to give a mouth refreshment feeling and saccharine sodium was used as the sweetening agent.
The films were prepared by solvent casting method. An accurately weighed quantity of HPMC E5 was soaked in distilled water for 30 min for proper dispersion of the polymer. Added PEG 400, inclusion complex of the drug and other excipients, and the final volume was adjusted with methanol. The dispersion was sonicated to remove the air-entrapped bubbles.
The casting solution (10 ml) was poured into a glass mold (area =32 cm 2 , every 2 cm 2 contains 1 mg of drug) and the solvent was allowed to evaporate by placing the glass mold in an oven for 24 h. The dried films were stored in a desiccator at room temperature.

Experimental design using CCD
CCD was used for the formulation and optimization of FDF using design expert software (version 13, Statease Inc, Minneapolis, USA) to investigate the influence of two independent variables i.e., the concentration of HPMC E5 (X1) and concentration of CCS (X2) on two responses, disintegration time (Y1) and in vitro percentage drug release at 5 min (Y2). A total of thirteen experimental runs were generated by the software, including 5 replicates of the center points, 4 axial, and 4 factorial points, and were carried out in randomized order. The different levels of variables are presented in table 1 and the actual composition of each formulation batch (F1-F13) is given in table 2.  *note: the dose of the drug-β-cyclodextrin inclusion complex was calculated to contain 1 mg of the drug in a single film of 2 cm 2 area

Physical and mechanical properties
The prepared films were evaluated for physical characteristics, weight uniformity, thickness, and folding endurance.
The prepared films were tested visually for their physical appearance and surface texture as well as for any drug precipitation and air bubble entrapment. Weight variation was calculated by weighing the film formulations in electronic analytical balance. The films of 2 × 1 cm 2 size were cut from three different positions of each batch prepared and analyzed. The film thickness was measured using a micrometer screw gauge [13]. Thickness was measured at different locations of a film and the average value was determined.
Folding endurance was measured manually by repeated folding of the films at the same place till they broke. The folding endurance value is the number of times the film is folded without breaking [14,15]. All the evaluations were done in triplicate and the results were expressed as average values and SD.

Surface pH
The surface pH of the film and the sublingual p H (6.8) should be closer so that the film dissolves quickly and be compatible. The surface pH is determined to evaluate the possible irritative effects of the formulations on the mucosae.
The film was placed in a Petri dish and was moistened with 1 ml of distilled water and kept for 30 seconds. The p H was noted after bringing the electrode of the p H meter in contact with the surface of the formulation and allowing equilibration for 1 min [16,17].

Drug content uniformity
The film was placed in a 100 ml volumetric flask and then dissolved in 50 ml methanol using a magnetic stirrer. The volume was completed with methanol and the solution was filtered and drug content was estimated by HPLC (instrument model: HPLC Agilent 1260 infinity).
The estimation of tacrolimus in the samples was done by the HPLC method according to USP. The mobile phase was constituted from different compositions of 6 mmol phosphoric acid and acetonitrile and tertiary butyl methyl ether (81: 19). The diluent used was acetonitrile and water (7:3). The column employed was of dimension 15 cm X 4.6 mm and the column temperature, 60 °C. 20 µl of injection volumes were used and the eluents were examined at 220 nm by a stream rate of 1.5 ml/minute [16,18].

In vitro disintegration time
No official guidelines are available for the determination of in vitro disintegration time of sublingual films. So, a non-official method was adopted. The disintegration test was performed by placing a film of size 2 × 1 cm 2 in a glass Petri dish containing 10 ml of phosphate buffer pH 6.8. The time required for the film to disintegrate was recorded [14,19].

In vitro drug release studies
The drug release study from the prepared FDFs was performed in beakers, each containing 100 ml of phosphate buffer at pH 6.8 [4,14]. The beakers were placed in a shaking water bath set at 37±0.5 °C and 100 rpm. Samples were removed at predetermined time intervals and filtered through a Whatman filter paper and the drug content of sample solutions was determined after suitable dilutions by HPLC. The aliquots withdrawn were replaced with fresh medium to maintain sink condition.
The determinations were conducted in triplicates. The amount of drug released at each time interval as well as the cumulative amount of drug released, was calculated as a function of time and the drug release profile graphs were constructed.

Statistical analysis
The experimental data obtained from the evaluations of all the film formulations (13 batches) were analyzed using Design Expert software. The optimum concentration of the independent variables, concentrations of HPMC E5 and CCS, was chosen based on the condition of the responses in obtaining maximum in vitro drug release at 5 min and minimum disintegration time. The response surface behavior was investigated for the response function (Y) employing the polynomial equation (1) and the generalized response surface model; Where Y is the predicted response; β0 is constant; β1 and β2 are the linear and quadratic coefficients and β11, β22, and β12 are the interaction coefficients.
The analysis of variance (ANOVA) was used to identify significant effects of factors on response regression coefficients. All significant independent variable effects (p<0.05) were included in the reduced model. To visualize the interaction effect of the variables on the responses, three-dimensional (3D) response surface plots were composed.
With the application of desirability constraints on in vitro drug release at 5 min and disintegration time, a numerical optimization analysis was then employed by Design-Expert® software to develop the optimal ODF formulation with the desired responses. The disintegration time and drug release studies were carried out on the prepared optimized formulation to verify the theoretical prediction.
The relative errors (%) between the predicted and observed values for each response were calculated [20].

Differential scanning calorimetry (DSC)
The prepared optimized formulation was subjected to DSC analysis (instrument model Q20 TA instruments) and compared with the thermogram of pure tacrolimus. Film strips weighing 10-12 mg were hermetically sealed in a standard aluminium pan and were heated at a scanning rate of 10 °C/min from room temperature to 225 °C under nitrogen gas flow.

X-ray diffraction (XRD) analysis
The roentgen graphic analyses were conducted with bruker model D8 Advance X-ray diffract meter (The samples were analyzed utilizing monochromatic Cu K-radiation at 40 mA and 40 kV within the region of 2.5° ≤ 2θ ≤ 40° at 25 °C in continuous scan mode with a step size of 0.02°. The diffractogram of the film was compared to the pure tacrolimus.

Surface morphology by scanning electron microscopy (SEM)
SEM studies of optimized FDF and that of pure tacrolimus were done with the Scanning Electron Microscope (JOEL JSM-6390; JAPAN), samples were mounted with double-sided carbon tape on aluminium stubs, and all the samples were sputtered with a thin layer of gold in auto fine coater JEOL JFC 1600 and the images were examined under vacuum at an accelerating voltage of 10 kV.

Physical and mechanical properties
FDFs should possess suitable mechanical strength to resist handling during processing, packing, and transit without being damaged.
All the films were smooth, transparent, homogeneous, and flexible.
The results of evaluations of weight variation, film thickness, folding endurance, surface p H, and drug content uniformity are presented in table 3. The weight variation was found to be acceptable as indicated by the small standard deviations. This shows the uniform distribution of the ingredients in the prepared films. All the films were of uniform thickness, in the range of 0.090 to 0.113 mm. Folding endurance is used to estimate the mechanical properties of a film and all the batches showed satisfactory flexibility. The surface p H of all the formulations was in the sublingual p H range; therefore, no possibility of irritation may be expected when administered sublingually and thus, patient acceptance will not be affected. The drug content of the ODFs was in the range of 95.71-99.12%, implying uniform distribution of the drug in the films. Data are expressed as mean±SD, n=3

In vitro disintegration time
The basic requisite of FDF is its short disintegration time in the saliva when placed in the mouth. But none of the official guidelines states the acceptable range of disintegration time. In the present study, almost all the batches have shown a disintegration time of fewer than 60 seconds, and the results obtained are presented in table 5. This response was selected as one of the independent variables, X1, for the optimization of the formulation.

In vitro drug release studies
No official compendial methods have been reported for release studies of FDF dosage forms. The literature reports various approaches, including the use of different volumes from 5 to 900 ml, different apparatus (pharmacopoeial or custom-made), and different media [19].
A fast release of the drug from the sublingual film is necessary so that the dose is not swallowed into the stomach. Moreover, since tacrolimus is a BCS class II drug, once the drug is released into the sublingual mucosa it will be rapidly absorbed and the higher permeability of the sublingual mucosa also aids the absorption process. So, the greater the drug release, the better we can expect bioavailability.
The cumulative percentage of drug release from all the prepared batches is shown in fig. 1a, 1b, and 1c. For all the batches, more than 95% of drug release was obtained in 30 min and almost 80% of drug content was released at around 5 min and this % drug release at 5 min was selected as one of the independent variables (X2) for the optimization of the formulation.
The fast drug release from the film can be attributed to the hydrophilic nature of the used polymer, HPMCE5, which absorbs water, dissolves rapidly and, thus introducing porosity into the film. The external solvent then diffuses into the film, thereby accelerating the drug dissolution [21].
Sublingual administration of tacrolimus FDF is expected to improve the drug's bioavailability by bypassing first-pass hepatic metabolism. In addition, the inclusion of the drug into β-cyclodextrin complexes before its formulation as a film enhances the drug's solubility too.

Statistical analysis
Based on CCD analysis, the effect of concentration of HPMC E5 (X1) and concentration of CCS (X2) on two responses disintegration time (Y1) and percentage drug release at 5 min (Y2) was studied and evaluated the main effects, interaction effects, and quadratic effects of the process variables on selected responses.

Fitting model to the data
The Design expert software suggested a quadratic model fitting best for both responses (disintegration time and percentage drug release at 5 min) as this model maximized adjusted R 2 and the predicted R 2 values.

ANOVA analysis
Analysis of variance (ANOVA) was performed to assess the significance of the quadratic polynomial models developed. The regression coefficients and p-values obtained are given in table 4. The small p-value (p<0.05) of the terms in the models indicated a significant influence on the response variables.

Sufficiency of models
Diagnostic plots (such as predicted vs. actual) are useful graphic tools to observe the adequacy of developed models. It shows the relationship between predicted and experimental values and describes model suitability. Table 5 shows the predicted and experimental values of disintegration time and in vitro drug release of the films and fig. 2a and 2b show the predicted vs. actual plots for all responses. The closer the points are to the diagonal line, the better the model fits the experimental data [22].

Effect of factors on responses
The 3D response surface plots were used to study the effects of independent variables on responses. These plots are known to study the interaction effects of the factors on the responses [23]. The 3D graphs for both responses are given in fig. 3a and 3b

Effect on disintegration time (Y1)
The  The reasons might be that CCS was swelled in water due to its good water uptake ability, which is beneficial to disintegration, but an excess quantity would lead to high viscosity with swelling, which would suppress the water penetration.

Effect on in vitro drug release at 5 min (Y2)
The polynomial equation proposed for the % in vitro drug release at 5 min was: 79.62-6.67895X1+1.18817X2+0.1X1X2-2.02875X1 2 -1.45375X2 2 The Model F-value was 60.31 implies that the model is significant. In this case, the significant model terms are X1, X2, X1², and X2 2 and their P-values were less than 0.05. The Lack of Fit F-value of 0.61 implies the Lack of Fit is not significant. Fig. 3b is the 3-D surface plot of the effect of the two independent variables on the cumulative % release of the drug from the films within 5 min. Response surface 3D plot reveals that in vitro drug release at 5 min (Y2) significantly decreases with the increase in HPMC (X1) concentration. On contact with the release medium, the viscous nature of the film-forming polymer (HPMC) creates thick matrix gel fluid, which governs drug release by resisting drug diffusion or matrix erosion and may lead to a delay in the drug release from the films. Another reason contributing to decreased drug release at increased polymer concentration may be the increase in the time required for wetting and dissolving the drug molecules present in the polymer matrices [15].

Selection of optimum formulation
The optimized formulation was selected with the following criteria: minimum disintegration time and the maximum cumulative drug release of the FDFs and from among the independent variables' concentration solutions suggested by the software, the following composition was selected for the preparation of the optimal formulation: 3.016% w/v of HPMC E5 (X1) and 11.731%w/w CCS (X2) with the desirability of 1. With these values, 25.79 sec for disintegration time (Y1) and 84.41% for in vitro drug release at 5 min (Y2) were the responses predicted.

Validation of optimum formulation
To validate these predicted values generated according to the results of the CCD design, the optimum formulation batches were prepared according to the optimized concentrations of the factors HPMC E5 and CCS and subjected to the disintegration and drug release tests. The data means of 27.7 sec for disintegration time and 83.13% for in vitro drug release at 5 min were obtained. These differences in the film's responses were not significantly deviated from the predicted responses of the optimized formula (table 6). From the confirmational analysis, it was shown that the above data means were within the range of 95% low and high prediction intervals. The results indicate the effective use of DOE for investigating variables' effects and optimization.

DSC analysis of optimized FDF
DSC thermograms of pure tacrolimus and optimized film are shown in fig. 4. DSC of pure tacrolimus showed a broad endothermic event attributed to dehydration happening at temperatures between 75 °C and 122 °C followed by melting of the anhydrate form with a sharp endothermic peak at 129.46 °C which corresponds to its melting point and indicating its crystalline nature. The optimized film formula showed the disappearance of the characteristic peak of tacrolimus which might be due to the homogenous dispersion of the drug in the formed film.

XRD
To obtain further information on the solid-state changes, XRD spectra were carried out on the optimized FDF formulation. The diffractograms of the drug and the optimized films are shown in fig. 5. The presence of numerous distinct peaks in the spectrum of tacrolimus indicates that the drug is present as a crystalline material.
The FDF formulation revealed a decrease in the intensity of diffraction peaks, indicating that the drug completely dispersed in the film matrix. This corresponds with the results obtained from DSC testing. Apart from the reduction of peak intensities, they were significantly broadened, which indicated that due to the inclusion complex formation, the drug was converted to a partial amorphous form.

Fig. 5: XRD pattern of a) tacrolimus monohydrate b) optimized film SEM
The SEM image of the pure drug as well as the morphology of the optimized FDF formula, is given in fig. 6a and 6b. From fig. 6a, it was found that tacrolimus occurred as cuboidal-shaped crystals, and the SEM image of the film revealed a smooth, slightly porous surface, indicating complete and uniform embedment of the drug: βcyclodextrin complexes in the polymeric film matrix. The porous nature indicates a uniform distribution of the drug complexes that help the medium to get penetrate inside the film. Very few drug crystals were found in the film, probably formed during the drying of the film. The XRD analysis suggested that the drug in the FDFS was in a partial amorphous state. SEM analysis revealed the smooth and porous nature of the film. Accordingly, it could be concluded that FDFs, optimized using CCD, may provide a novel sublingual delivery system of tacrolimus for avoiding its oral bioavailability problems.

AUTHORS CONTRIBUTIONS
All the authors have contributed equally.