Optimization of the Factors Affecting the Absorption of Vardenafil from Oral Disintegrating Tablets: A Clinical Pharmacokinetic Investigation

Because of poor solubility and considerable metabolism, vardenafil (VRD) bioavailability is 15%. To overcome this obstacle, this study aimed to increase the solubility, hasten the onset of action, and mask the unpleasant taste of VRD utilizing β-cyclodextrin (β-CD) and formulation of the inclusion complex as oral disintegrating tablets (ODTs). The solubility of the obtained complexes in various ratios has been studied. A Box–Behnken design (BBD) was utilized to investigate the influence of excipients on the quality of ODTs. The solubility of VRD was improved at 1:2 drug:β-CD ratio. The formulated VRD-ODTs exhibited satisfying results regarding the hardness and disintegration time. In addition, in vivo taste masking and disintegration time showed improved results, after placing the tablets in the oral cavity of the healthy volunteers. When compared with the marketed tablets, the pharmacokinetic parameters for the optimized VRD-ODTs exhibited a significant improvement with p < 0.05 in the maximum plasma concentration and reduction in the time needed to reach this concentration. Finally, the optimized VRD-ODTs exhibited increased oral absorption of VRD and subsequent decrease in the time of onset of clinical effect and masking the unpleasant taste.


Introduction
Vardenafil (VRD) is a potent phosphodiesterase (type V) inhibitor, used for the treatment of the erectile dysfunction disease [1]. Its mechanism of action depends on inhibition of the degradation of cyclic GMP (cGMP) in the smooth muscle tissues located on the internal surface of the blood vessels that supply the corpus cavernosum of the penis. This accumulation of cGMP in the corpus cavernosum leads to the release of nitric oxide that causes dilation of the blood vessels then, the erection occurs successfully [2]. Furthermore, VRD has been utilized in patients with pulmonary hypertension due to the presence of PDE5 in the smooth muscle of the arterial wall within the lungs [1]. According to the biopharmaceutical classification system (BCS), VRD was classified as a class II (high permeability/low solubility) drug [3] that suffers from drawbacks of low bioavailability (15%) and bitter taste. It is also

Solubility Study
The effect of inclusion complexes on the solubility of VRD was evaluated according to the method reported by Higuchi and Connors [31]. Briefly, an excess of raw VRD and VRD-β-CD were added to vials containing distilled water then placed in a shaking water bath at 25 ± 0.5 • C. Samples were analyzed for VRD content at 230 nm until equilibrium is attained.

Preparation of ODTs
Direct compression method was utilized for the preparation of the suggested VRD-ODTs as displayed in Table 2. The tablet blend was compressed using 9 mm flat punches with compression force of 10 KN into 200 mg tablets using a tablet press (Erweka, GmbH, Heusenstamm, Germany).

Evaluation of the Prepared VRD-ODTs
Evaluation of VRD-ODTs was performed on the tablets of all batches considering the visual inspection, weight and content uniformity, thickness, hardness and friability according to the Pharmacopeial requirements.

In Vitro Disintegration of VRD-ODTs
VRD-ODTs (6 tablets/batch) were placed in the baskets of USP disintegration apparatus (Pharmatest, PT-DT7, Hainburg, Germany). The apparatus run utilizing distilled water as the immersion fluid at 37 ± 0.5 • C. The tablets were observed, and the time taken for complete disintegration of all tablets was determined.

In Vitro Dissolution of VRD-ODTs
USP dissolution apparatus II (paddle method) of Erweka GmbH, (Heusenstamm, Germany) was used in the dissolution of VRD from the ODTs. The study was performed with 900 mL distilled water at 50 rpm and equilibrated at 37 ± 0.5 • . Samples of 5 mL were withdrawn at the predetermined time intervals 5,10,15,20,30,45, and 60 min and replaced with a fresh preheated medium for each time point, then analyzed by UV spectrophotometer at 230 nm. The experiment was performed three times for each formula and the mean values of the cumulative % release of VRD after 60 min were determined.

VRD-ODTs Formulation Data Analysis by BBD
Hardness (Y 1 ) and disintegration time (Y 2 ) were analyzed using the experimental design software. Significance of the analysis was set for any factor at p < 0.05. The optimized VRD-ODT formulation suggested was prepared and checked for the hardness and disintegration time results. The observed values were compared with the predicted ones and the residuals were calculated.

In Vivo Taste Masking and Disintegration Time Evaluation
A single-blind study was intended for disintegration time and the taste masking tests in the buccal cavity of six healthy human volunteers. The study was performed and approved at the Egyptian Research and Development Company (ERDC), Cairo, Egypt on 30 August 2017 with Ethical Approval Code (Verd-p 0566/449). The human subjects were asked to rate the bitter taste of the optimized formula utilizing a scale of 0-3. When the score ≤ 1, the taste was acceptable while if the score >1, indicates the tablet is bitter and not acceptable [14]. Also, the disintegration time of the tablet in the oral cavity was recorded.

Pharmacokinetic Parameters Evaluation
An open-label, single dose, randomized, one-period, parallel design comprising fourteen days of screening preceding 24 h study periods was used. The participants were administered a buccal 10-mg dose of VRD from the optimized formulation tablet (test). While, the marketed tablets (reference) were administered the same dose orally with water. The study was carried out at the Egyptian Research and Development Company (ERDC), Cairo, Egypt. ERDC Research Ethical Committee had formally approved the study design protocol on 30 August 2017 with Ethical Approval Code (Verd-p 0566/449). The study was accomplished in agreement with the Declaration of Helsinki and the International Conference on "Harmonization of Good Clinical Practices".

Population and Sampling
Healthy male volunteers (25-43 years of age) at the time of screening were selected for the study. The selected subjects signed written informed consent, were willing to participate in this clinical trial, and to comply with the study requirements. Complete medical history, laboratory analysis, and physical examination were performed for the selected candidates to ensure their eligibility for participation. Subjects were divided into two groups (6 each). The first group was administered the optimized formulation while the second one was given the marketed VRD tablets. Blood samples (5 mL) were collected at 0, 0.16, 0.25, 0.5, 0.75, 1, 1.25, 1.5, 2, 2.5, 3, 4, 6, 8, 12, and 24 h in heparinized tubes. The tubes were centrifuged at 3500 rpm for 10 min (Centurion, West Sussex, UK) and the separated serum was stored at −80 • C until analysis.

Chromatographic Conditions
VRD detection in human plasma was conducted using a high-performance liquid chromatography coupled with MS/MS method (HPLC-MS/MS) that was developed at the ERDC according to the reported method with slight modification [32]. Validation of the method was based on the FDA Bio-analytical Method Validation Guidelines 2003. Assay linearity was verified for VRD at a concentration range of 3-350 ng/mL with regression coefficients (R 2 ) of 0.996 and 0.994 for VRD. The lower limits of quantification were 3 ng/mL for VRD. The HPLC-MS/MS system consisted of HPLC, Agilent series 1200 (Agilent Technologies Deutschland GmbH, Waldbronn, Germany), used with a Triple Quad G1311A quaternary pump equipped with 6400 Series Triple Quadrupole LC/MS detector and mass hunter software. Chromatography was performed (75% acetonitrile: 25% buffer (ammonium formate 20 mM + 0.2% (v/v) formic acid in water) as mobile phase at a flow rate of 0.35 ml/min and a reversed phase column Intersil ODS-3 (4.6 mm × 50 cm, dp 5µm) at 25 • C. Sildenafil was selected as an internal standard (IS).

Pharmacokinetic Analysis
A noncompartmental analysis of the pharmacokinetic parameters was achieved by unpaired t-test (two-tailed) using Kinetica™ software (Version 4; Thermo Fisher Scientific, Waltham, MA, USA, 2005). Any significant difference in drug plasma concentration between the two groups was assessed with two-way ANOVA followed by Sidak's multicomparison test using GraphPad Prism 6 (GraphPad Software, Inc., San Diego, CA, USA, 2012). Results were considered significant at p < 0.05. The peak plasma concentration achieved by the drug (C max ), the time after administration of a drug when the maximum plasma concentration is reached (t max ), the area under curve (AUC), elimination rate constant (K el ) and mean drug residence time (MRT) was calculated to allow the relative bioavailability [(AUCformulation/AUCtablets) × 100] to be determined.

Results and Discussion
VRD-ODTs were developed to deliver the drug rapidly. A pre-formulation study involving complexation of VRD with β-CD was carried out. A direct compression method was used for the formulation of 15 formulae of VRD-ODTs according to BBD. All the prepared formulations were evaluated for weight uniformity, thickness, friability, hardness, content uniformity, and in vitro disintegration as well as in vitro dissolution. The results of all experiments were used to correlate the independent variables that constitute the combination of excipients of tablets with the dependent variables that represent the quality parameters of the ODTs. BBD utilized these relations to statistically optimize the process to produce VRD-ODTs with maximum hardness and minimum disintegration time. Finally, the obtained optimized formulation was evaluated in vivo on healthy human volunteers compared with the marketed Levitra tablets. The details of the results and their discussion are given in the following sections.

Saturation Solubility Studies of the Prepared Complexes
The data represented in Figure 1 revealed that the solubility of raw VRD was 0.13 mg/mL. While the solubility of VRD in solid dispersion using β-CD in different molar ratios 1:1, 1:1.5, and 1:2 was increased to reach 13.7 mg/mL for VRD-β-CD 1:2 molar ratio. This finding confirms the formation of a complex with β-CD and improves VRD solubility which in a good agreement with the previously reported results of carvedilol [33], piroxicam [34], ketoconazole [35], gliclazide [36], zafirlukast [37], and aripiprazole [38].

Development of VRD-ODTs
Development of the formulation in the present study was mainly based on three factors namely mannitol percentage as a diluent, Explotab percentage as a superdisintegrant, and Avicel percentage as a binder. Various ratios of each component combinations were used to get oral disintegrating tablets with good quality attributes. All formulations were suggested by BBD as described in Table  1.

Evaluation of VRD-ODTs
All batches of VRD-ODTs were evaluated and results are shown in Table 3. The tablet weight variation of the prepared batches was less than 2%, in accordance with USP requirements. The tablets formulations met the European Pharmacopeia limits for disintegration of oral disintegration tablets (<3 min). Friability, hardness and thickness of the prepared tablets were in the acceptable limits as indicated in Table 3. Table 3. Characteristics of VRD-ODTs, data expressed as mean ± SD (n = 10).

Development of VRD-ODTs
Development of the formulation in the present study was mainly based on three factors namely mannitol percentage as a diluent, Explotab percentage as a superdisintegrant, and Avicel percentage as a binder. Various ratios of each component combinations were used to get oral disintegrating tablets with good quality attributes. All formulations were suggested by BBD as described in Table 1.

Evaluation of VRD-ODTs
All batches of VRD-ODTs were evaluated and results are shown in Table 3. The tablet weight variation of the prepared batches was less than 2%, in accordance with USP requirements. The tablets formulations met the European Pharmacopeia limits for disintegration of oral disintegration tablets (<3 min). Friability, hardness and thickness of the prepared tablets were in the acceptable limits as indicated in Table 3.
In addition, drug content was in the range of 95.43-101.7%. To avoid delaying the disintegration of ODTs, the hardness usually planned to be lower than the conventional tablets. The hardness is an essential factor which affects the disintegration and dissolution times that influence bioavailability [39]. Some challenges encountered in the formulation and production of ODTs such as the disintegration time and mechanical strength. The ideal ODTs should have concise disintegration time that is usually about one min or less which require low mechanical strength. The disintegration time is directly proportional with the mechanical strength of the tablets. Therefore, it is essential to have a good compromise between mechanical strength and disintegration time [9].     Table 4 shows the estimate effect of each factor, F-ratio and p-values for Y1 and Y2 from two-way analysis of variance (ANOVA). From the obtained analysis, Explotab percentage (X2) had no significant effect on the hardness of tablets (Y1) but had a significant antagonistic effect on the disintegration time of the tablets (Y2) with a P-value of 0.0001. Also, it was found that mannitol percentage (X1) and Avicel percentage (X3) had significant synergistic effects on the hardness (Y1) with p-values of 0.0009 and 0.0001, respectively. In addition, the interaction term X1X2 showed a significant synergistic effect on the hardness with a p-value of 0.0035. The R-Squared statistic indicates that the model as fitted explains 97.647% and 82.457% of the variability in hardness and disintegration time, respectively.   Table 4 shows the estimate effect of each factor, F-ratio and p-values for Y 1 and Y 2 from two-way analysis of variance (ANOVA). From the obtained analysis, Explotab percentage (X 2 ) had no significant effect on the hardness of tablets (Y 1 ) but had a significant antagonistic effect on the disintegration time of the tablets (Y 2 ) with a P-value of 0.0001. Also, it was found that mannitol percentage (X 1 ) and Avicel percentage (X 3 ) had significant synergistic effects on the hardness (Y 1 ) with p-values of 0.0009 and 0.0001, respectively. In addition, the interaction term X 1 X 2 showed a significant synergistic effect on the hardness with a p-value of 0.0035. The R-Squared statistic indicates that the model as fitted explains 97.647% and 82.457% of the variability in hardness and disintegration time, respectively.

Mathematical Modeling of the Data
Mathematical modelling for Y 1 and Y 2 of VRD-ODTs were generated after analysis of the data using the Statgraphics ® software (Equations (1) and (2)).
Equations (1) and (2) reflect the quantitative effect of formulation factors; mannitol % (X 1 ), Explotab % (X 2 ), and Avicel % (X 3 ) and their interactions on the responses; the hardness (Y 1 ) and the in vitro disintegration time (Y 2 ). Figure 3, 2D Pareto charts, showed the effect of X 1 -X 3 and their interactions on Y 1 and Y 2 . Figure 3 displayed that X 1 and X 3 have significant synergistic effects on Y 1 , and the interaction between X 1 and X 2 have also a significant synergistic effect on Y 1 . While X 2 has no significant effect on Y 1 . Figure 4, 3D response surface plots, confirmed the understanding of the influence of each factor on Y 1 , Similar findings were reported in the literature for the effect of excipients on the prepared tablet properties [40][41][42][43][44][45][46].

Mathematical Modeling of the Data
Mathematical modelling for Y1 and Y2 of VRD-ODTs were generated after analysis of the data using the Statgraphics ® software (Equations (1) and (2)).
Equations (1) and (2) reflect the quantitative effect of formulation factors; mannitol % (X1), Explotab % (X2), and Avicel % (X3) and their interactions on the responses; the hardness (Y1) and the in vitro disintegration time (Y2). Figure 3, 2D Pareto charts, showed the effect of X1-X3 and their interactions on Y1 and Y2. Figure 3 displayed that X1 and X3 have significant synergistic effects on Y1, and the interaction between X1 and X2 have also a significant synergistic effect on Y1. While X2 has no significant effect on Y1. Figure 5, 3D response surface plots, confirmed the understanding of the influence of each factor on Y1, Similar findings were reported in the literature for the effect of excipients on the prepared tablet properties [40][41][42][43][44][45][46]. Also, Figure 3 displayed that X2 has a significant antagonistic effect on Y2. This finding means that increasing X2 level will lead to reduction in the disintegration time. Other studies have investigated the effect of Explotab on tablet disintegration times [43,47,48]. Both X1 and X3 have no significant effect on Y2. Also, 3D response surface plots ( Figure 4) were graphically established utilizing the software.  Also, Figure 3 displayed that X 2 has a significant antagonistic effect on Y 2 . This finding means that increasing X 2 level will lead to reduction in the disintegration time. Other studies have investigated the effect of Explotab on tablet disintegration times [43,47,48]. Both X 1 and X 3 have no significant effect on Y 2 . Also, 3D response surface plots ( Figure 4) were graphically established utilizing the software.

Prediction of the Optimized Formulation
The optimized formulation of VRD-ODTs was predicted upon the data analysis with desirability function equal to 0.883603 over the indicated region. The optimum combination of factors was 38.52% of Mannitol concentration, 9.99% of Explotab concentration, and 25% of Avicel concentration. The observed, predicted and residual values for Y1 were 68.45 N, 65.90 N and 2.55 N, respectively while, for Y2 were 51.77 s, 49.78 s and 1.99 s, respectively. This finding supports the mathematical experimental design in maximizing Y1 and minimizing Y2 that fulfills the Pharmacopoeial requirements via the direct compression method.

In Vivo Taste Masking and Disintegration Time
The results of the in vivo taste masking test were listed in Table 5. The scores of the six volunteers were equal or less the one which indicate the formulation has an acceptable taste masking effect. The mean result of the in vivo disintegration time was 62.33 s which is acceptable according to European Pharmacopeia (<3 min).

Pharmacokinetic Parameters Evaluation
The VRD plasma concentration time profiles from the optimized formulation of VRD-ODT and the marketed Levitra tablet (Bayer AG, Leverkusen, Germany) are represented in Figure 5. The values of Cmax, tmax and AUC(0-24), t1/2, Kel, and MRT for VRD from these formulations, are summarized in Table 6. The results indicated that optimized VRD-ODTs bioavailability (F) compared with the

Prediction of the Optimized Formulation
The optimized formulation of VRD-ODTs was predicted upon the data analysis with desirability function equal to 0.883603 over the indicated region. The optimum combination of factors was 38.52% of Mannitol concentration, 9.99% of Explotab concentration, and 25% of Avicel concentration. The observed, predicted and residual values for Y 1 were 68.45 N, 65.90 N and 2.55 N, respectively while, for Y 2 were 51.77 s, 49.78 s and 1.99 s, respectively. This finding supports the mathematical experimental design in maximizing Y 1 and minimizing Y 2 that fulfills the Pharmacopoeial requirements via the direct compression method.

In Vivo Taste Masking and Disintegration Time
The results of the in vivo taste masking test were listed in Table 5. The scores of the six volunteers were equal or less the one which indicate the formulation has an acceptable taste masking effect. The mean result of the in vivo disintegration time was 62.33 s which is acceptable according to European Pharmacopeia (<3 min).

Pharmacokinetic Parameters Evaluation
The VRD plasma concentration time profiles from the optimized formulation of VRD-ODT and the marketed Levitra tablet (Bayer AG, Leverkusen, Germany) are represented in Figure 5. The values of C max , t max and AUC (0-24) , t 1/2 , K el , and MRT for VRD from these formulations, are summarized in Table 6. The results indicated that optimized VRD-ODTs bioavailability (F) compared with the marketed tablet was 125.445%. This data indicated that ODT's improved the bioavailability of VRD over the marketed tablets. The oral absorption of VRD from ODTs was obviously higher when compared with the marketed tablets which was obvious from the value of C max that increased significantly from 12.29 ng/mL for the marketed tablet to 18.19 ng/mL (for the optimized VRD-ODTs. marketed tablet was 125.445%. This data indicated that ODT's improved the bioavailability of VRD over the marketed tablets. The oral absorption of VRD from ODTs was obviously higher when compared with the marketed tablets which was obvious from the value of Cmax that increased significantly from 12.29 ng/mL for the marketed tablet to 18.19 ng/mL (for the optimized VRD-ODTs. In addition, the tmax of optimized VRD-ODTs shortened to 1 h when compared with tmax of 2 h for the marketed tablets which indicated that the onset of action of VRD from optimized ODTs was accelerated in comparison with the marketed tablets. The analysis of variance showed that there were significant differences among the samples (p < 0.05) taken at 0.5, 0.75, 1, and 1.25 h from the two groups of volunteers indicating the significant improvement achieved by the ODTs. Complexation of VRD with β-CD followed by its formulation as taste masked ODTs enhance both the solubility and dissolution rate that reflects on the availability of VRD ready for absorption [49]. Accordingly, optimized VRD-ODTs is a promising approach for improved VRD bioavailability and absorption rate.  Table 6. Pharmacokinetic parameters ± SD of VRD following the administration of a single oral dose (10 mg) of either the VRD marketed tablet, or the optimized VRD-ODTs.

Pharmacokinetic Parameter VRD Marketed Tablet Optimized VRD-ODTs
Cmax (ng/mL) 12   In addition, the t max of optimized VRD-ODTs shortened to 1 h when compared with t max of 2 h for the marketed tablets which indicated that the onset of action of VRD from optimized ODTs was accelerated in comparison with the marketed tablets. The analysis of variance showed that there were significant differences among the samples (p < 0.05) taken at 0.5, 0.75, 1, and 1.25 h from the two groups of volunteers indicating the significant improvement achieved by the ODTs. Complexation of VRD with β-CD followed by its formulation as taste masked ODTs enhance both the solubility and dissolution rate that reflects on the availability of VRD ready for absorption [49]. Accordingly, optimized VRD-ODTs is a promising approach for improved VRD bioavailability and absorption rate.

Conclusions
The present study proves that the inclusion complex of VRD with β-CD increased its aqueous solubility in a 1:2 molar ratio and masked its bitter taste. The optimized formulation compromises between the hardness and the disintegration time. The studied tablets' excipients showed the varied impact to fulfill all the required characteristics of ODTs. However, incorporation of an optimized concentrations, via using the design of the experiment, of these excipients achieved the best balance and therefore produced VRD-ODTs with superior characteristics. The developed tablet offers a taste-masked, satisfactory hardness, and short disintegration time for the rapid release of the drug. The pharmacokinetic data point out to the improved bioavailability of VRD over the marketed tablets. In addition, in vivo data found that the oral absorption of VRD from ODTs was obviously higher than that from the marketed tablets. Moreover, the t max was shortened to 1 h in comparison with 2 h for the marketed tablets which indicated the rapidity of onset of action of VRD and hence improved patient efficacy and satisfaction.