The solubility of voriconazole in various oils (cotton seed oil, isopropyl myristate, isopropyl palmitate, castor oil, oleic acid, olive oil, soya oil, cod liver oil), surfactants (tween 80, span 80), co-surfactants (propylene glycol, PEG 200, PEG 400, plurololeique) was determined by dissolving an excess amount of voriconazole in 2 ml of each selected oils, surfactants, co-surfactants stirred at 100 rpm for 72 hours to achieve equilibrium then suspension was centrifuged using centrifuge (Bio Era, Mumbai, India) at 3000 rpm for 15 min. The supernatant was taken and filtered through 0.45 µm membrane filter (whatmann) [10, 11]. The concentration of voriconazole was determined in each oil, surfactant and co-surfactant by double beam UV/Visible spectrophotometer (Labindia, Mumbai, India) at their particular wavelength 255 nm. On the basis of result, it was found that oleic acid, isopropyl myristate and isopropyl palmitate consumed maximum amount of voriconazole thus chosen as a vehicle for microemulsion than other oils.

In order to find out concentration range of components for the existence range of microemulsion, pseudo-ternary diagram were constructed using tween 80 to propylene glycol by water titration method at ambient temperature (25°C). Weight ratios i.e. 1:1 and 2:1 were considered while constructing pseudo-ternary phase diagram. For each two phase diagram, each oil phase (oleic acid, isopropyl myristate, isopropyl palmitate) to amount of S_mix (tween 80 and propylene glycol) were mixed at ratio of 1:9, 2:8, 3:7, 4:6, 5:5, 6:4, 7:3, 8:2, 9:1 respectively. The mixture of oil, surfactant and co-surfactant at weight ratios were diluted with dropwise addition of distilled water, under moderate magnetic stirring (Remi Equipments Pvt. Ltd. Mumbai, India). After being equilibrated, the mixtures were assessed visually for clear/transparent, single phase, low viscous and flowable o/w microemulsions [12, 13]. The pseudo-ternary phase diagrams were constructed using CHEMIX (CHEMIX School version 3.5) software.

For selection of MEs from preliminary batches, the phase diagram regions were recognized. VCZMEs were prepared by addition of 0.1% w/v (1 mg/ml) voriconazole in mixtures of each oil, surfactant, co-surfactant with varying component ratio and an appropriate amount of water was added to mixture drop by drop and the microemulsion containing voriconazole was obtained by stirring the mixture using magnetic stirrer (RemiEquipmentsPvt. Ltd. Mumbai, India) at 25°C at 100 rpm [12].Various batches of microemulsions were prepared by water titration method and optimization was done in terms of thermodynamic stability/dispersibility, clarity, droplet size, viscosity studies.

The selected microemulsion formulations were imparted with mucoadhesive characteristics by using optimized modified chitosan solution. In brief, the selected formulations in their optimized S_mix ratios(1:1 and 2:1) along with optimized oil to S_mix ratios were titrated against the Mod.CH (0.25% w/v) solution in distilled water. As different composition ratio, described in Table 1. The Mod.CH (0.25% w/v)dispersion added dropwise to mixture of VCZ (0.1% w/v) containing optimized S_mix and oil ratio and stirred under magnetic stirrer (Remi Equipments Pvt. Ltd. Mumbai, India) at 25°Cat 100 rpmfor 2 hours [12].

Microemulsions are thermodynamically stable systems and formed at particular composition (concentration) of oil, surfactant, co-surfactant and water. In addition no creaming or cracking and phase separation as compared to regular emulsion have kinetic stability results ultimately phase separates, thus selected regions of microemulsions also characterized to prove thermodynamic stability. The microemulsion formulations subjected to stress tests like heating cooling cycle, centrifugation, freeze thaw cycle, those formulations passed these stress tests were subjected to dispersibility test [14, 15].

The drug content (%) of microemulsion was determined by dissolving required quantity (2 ml) of MEs (equivalent to 2 mg VCZ) in mixture of methanol/water. After suitable dilutions with methanol and distilled water, absorbance was determined using double beam UV/Visible spectrophotometer (Lab India, Mumbai, India) keeping blank microemulsion at wavelength 255 nm [15].

FTIR studies were performed todetermine the compatibility study between drug and other components of microemulsions (oil, surfactant/co-surfactant and Mod.CH) by KBR dispersion method considering % transmittance of FTIR spectrophotometer (IRAffinity1, Shimadzu, Japan). The base line correction was completed using dried potassium bromide and % transmittance analyzed at in the spectral region of 4000-400 cm^-1 using a resolution of 4 cm^-1 and 40 cm^-1 scans [13].

The droplet size, zeta potential and polydispersity index (PDI) of the MEs was determined by photon correlation spectroscopy (PCS) (which analyzes the fluctuations in light scattering due to browning motion of the particles) with a Nano sizer SZ-100 (Horiba nano size analyzer, Japan) able to measure sizes between 10-3000 nm. Light scattering was monitored at 25°C at 90°angles. The dispersed formulations were measured after dilution (1:100) i.e. 1ml formulation was dispersed in 100 ml distilled water to produce the required count rate (50-200) to enable accurate measurement. All the analysis was repeated in triplicate [16].

The viscosity of all microemulsions was determined without dilution using Fungilab viscometer (Fungilab expert series serial EXPL 3000050, code V300003, Spain) using spindle no.L3 at 100 rpm and temperature set at 25°C ± 0.5°C. The experiments were performed three times for each sample and the results are presented as mean ± SD.

The pH values of the all microemulsion samples were measured by a pH meter (Digital Systronics, Mumbai, India) at 25°C ± 0.5°C. Which was calibrated at pH 4 and pH 7 standard buffer solutions. All measurements were carried out in triplicate [17].

The electric conductivity of ME was measured with a conductivity meter (Equip-Tronics, EQ-664, Mumbai, India) equipped with inbuilt magnetic stirrer. This was performed by using conductivity cell (with a cell constant of 1.0) consisting of two platinum plates separated by desired distance and having liquid between the platinum plate acting as a conductor. All measurements were carried out in triplicate [18].

The in vitro drug release studies were carried out in the modified USP dissolution apparatus 1 (37°C ± 0.5°C) containing a two-sided open glass cylinder for 12 hours. The diffusion barrier was dialysis membrane, molecular weight cut off 12000-14000A° (Himedia, Mumbai, India) as a release barrier. A pre-soaked dialysis membrane was adapted to the terminal portion of the glass cylinder. In each case, microemulsion (2 ml i.e. VCZ 2mg/ml) was accurately introduced into the glass cylinder from the open side and this cylinder was fixed on the stirrer. The stirrer was suspended in 100 ml dissolution of simulated tear fluid (pH 7.4) medium maintained at 37°C ± 0.5°C at 50 rpm. Aliquots of samples were withdrawn at predetermined time intervals with volume replacement. The withdrawn samples were analyzed for drug content, by measuring absorbance at 255 nm in the double beam UV/Visible spectrophotometer (Lab India, Mumbai, India). Sink conditions were maintained throughout the release period. The same method was used for lyophilized powder (VOZOLE (1% w/v) i.e. 10 mg VCZ reconstituted with 1 ml sterile water for injection) prepared as 2 ml for release study. The comparative release study of drug from VCZMEs and VOZOLE (1% w/v) were carried out. Data obtained in triplicate were analyzed graphically i.e. percent cumulative release versus time graph was plotted [12, 19].

Drug permeation studies were carried out by using modified all glass Franz diffusion cell, putting the voriconazole microemulsion 0.1% w/v i.e. (1 mg/ml) on a freshly excised goat cornea. The fresh, whole eyeballs of goats were obtained from a local butcher's shop and transported to the laboratory chilled in normal saline (4°C). The cornea was then carefully excised along with 2 to 4 mm of surrounding scleral tissue and was washed with normal saline until the washing was free from protein. The excised cornea was fixed between the clamped donor and receptor compartments of an all glass modified Franz diffusion cell in such a way that its epithelial surface faced the donor compartment.The corneal area available for diffusion was 0.75 cm². The receptor compartment was filled with 10 ml freshly prepared simulated tear fluid (pH 7.4), and all air bubbles were expelled from the compartment. The whole corneal preparation procedure was completed within 1 hour after the sacrifice of goat. The aliquots (1 ml) of the formulated microemulsions were placed on the excised cornea and the opening of the donor cell was sealed with a glass cover slip to prevent evaporation. The acceptor solution was kept at 37°C ± 0.5°Cwith constant stirring using a teflon-coated magnetic stir bead. The permeation study was carried out for 4 hours and samples were withdrawn from the receptor at end of 4 hours. The withdrawn samples were analyzed for drug content by measuring absorbance at 255 nm in a double beam UV/Visible spectrophotometer (Lab India, Mumbai, India). The same method was used for marketed formulation (VOZOLE (1% w/v) i.e. 10 mg VCZ lyophilized powder reconstituted with 1 ml sterile water for injection) prepared as 1ml for permeation study. The comparative release study of drug from selected VCZMEs with VOZOLE (1% w/v) were carried out. Data obtained in triplicate [22, 23].

Permeation (%) = Amount of drug permeated in receptor/Initial amount of drug in donor ×100

Antifungal activity of formulations was checked by cup-plate method. The medium for antifungal activity was used sabouraud dextrose agar (dextrose 4 gm, mycological peptone 1 gm, agar 1.5 gm). The sabouraud dextrose agar was used in quantity of 6.5 gm and dissolved by heating in100 ml of distilled water with frequent agitation and boiled for 1 min.to dissolve the entire powder and the pH of media maintained at 5.6. Then the prepared media was autoclaved at 121°C for 15 min. The media were poured aseptically into 93 mm diameter glass petri plate and kept till the solidification. The surface of agar plates was pierced using a sterile cork borer. The Candida albicans were inoculated onto agar surface by streaking. An aliquot marketed formulation (VOZOLE (1% w/v)) from 5 µl as a test sample was prepared and placed in first 7 mm wells, An aliquot (IPMVM-11) from 5 µl as a standard sample was prepared the standard sample placed in second 7 mm wells and aliquot Mod.CH (0.25% w/v) solution from 5 µl placed in third 7 mm wells. The plates were left for 30 min and incubated at 25°C ± 1°C for 24 hours. After that zone of inhibitions were measured [24].

Rabbits were provided by the Central Animal House, Satara College of Pharmacy Satara (SCOPS), Degaon, Satara, MS, India. All experiments were conducted with the permission of Central Animal Ethical Committee, Satara College of Pharmacy Satara (SCOPS), Degaon, Satara, MS, India and were carried out in accordance with the current guidelines for the care of laboratory animals. The animals used for research study, approved by the Local Animal Ethical Committee of Shivaji University (Approval No. 2011/158).

In vivo studies were carried out on healthy rabbit eyes which inflamed with fungal keratitis. In vivo study performed for determination of VCZ concentrations in the aqueous humor were carried out on rabbit eyes; evaluation of therapy and measurement of ocular irritancy were conducted on fungal keratitis induced rabbits. For study five months old male albino white rabbits weighing 3–3.5 kg were used. Rabbits were housed in a air-conditioned room maintained at temperature (22°C ± 1°C) with an alternating 12 hour light–dark cycle and fed a standard pellet diet and water ad libitum. Rabbits were provided with artificial fluorescent light. These conditions were checked, on a daily basis, to ensure animal safety and well-being.

IPMVM-11 was selected for the in vivo study on the basis of droplet size, viscosity and better ex vivo study through goat cornea. The rabbits were divided into three groups of three rabbits each (six eyes). The rabbits in groups 1 and 2 were topically instilled with IPMVM-11 (in right eye) and marketed formulation (VOZOLE (1% w/v)) (in left eye) respectively, while group 3 (control group) received sterile physiological saline solution. Using a calibrated dropper, 50 µl of IPMVM-11 was instilled twice daily for group 1, whereas group 2 was treated with 50 µl of the VOZOLE (1% w/v) four times a day for three days before the generation of fungal keratitis and continued for three days after induction. After 24 hours the induction of fungal keratitis, the aqueous humor sampling was performed for determination of VCZ content within aqueous humor.

All values existing in the study are average of triplicate experiments for the same time points. Ex vivo permeation and in vivo profile of voriconazole microemulsions were tested statistically using one-way analysis of variance (ANOVA) followed by Dunnett’s test using graph pad prism 5 software (graph pad software inc., San Diego CA) at different level of significance. (^†Statistically significant difference at p < 0.05; ^††statistically significant difference at p < 0.01; ^†††statistically significant). The results were expressed as mean values ± SD.

The solubility study indicates the drug loading capability. Oils can solubilize the lipophilic drug in specific amount. To develop topical VCZMEs for ocular application, the solubility determined in different oils, surfactants, co-surfactants [13, 15]. On the basis of solubility study of voriconazole in various oils like oleic acid (35.74 ± 2.22mg/ml), isopropyl myristate (13.83 ± 1.19mg/ml), isopropyl palmitate (13.05 ± 1.23mg/ml), olive oil (12.65 ± 1.07 mg/ml), cotton seed oil (11.88 ± 1.03 mg/ml), soya oil (10.38 ± 1.08 mg/ml), castor oil (6.27 ± 2.91 mg/ml), liquid paraffin (3.11 ± 0.54 mg/ml).

There are several components exhibiting surfactant properties may be employed for microemulsion systems, but the choice is limited as very few surfactants are applicable for eye. Generally non-ionic surfactants are preferred. VCZ solubility in various surfactants/co-surfactants like tween 80 (25.14 ± 1.06mg/ml), propylene glycol (34.63 ± 2.13mg/ml), PEG 200 (31.58 ± 2.03 mg/ml), PEG 400 (26.73 ± 2.22 mg/ml), plurololeique (3.41 ± 0.25 mg/ml), span 80 (1.77 ± 0.21 mg/ml). The HLB value of tween 80 and propylene glycol is 15 and 11.6 respectively. As per concern o/w microemulsion usually high HLB surfactant match with low HLB co-surfactant [29]. Since the non-ionic surfactant and co-surfactant do not ionise at any extent in solution, so that tween 80 and propylene glycol were selected as an ideal surfactant and co-surfactant for the system [30].

Fig. (1). Phase diagrams of (A) O/W microemulsion region of OA, Smix 1:1 and water. (B) O/W microemulsion region of OA, Smix 2:1 and water. (C) O/W microemulsion region of IPM, Smix 1:1 and water. (D) O/W microemulsion region of IPM, Smix 2:1 and water. (E) O/W microemulsion region of IPP, Smix1:1 and water. (F) O/W microemulsion region of IPP, Smix 2:1 and water.

The results of droplet size of freshly prepared VCZME formulations are characterized which are formulated with water titration method. All ME formulations showed a mean droplet size below 250 nm which is an optimum size for ophthalmic delivery as the human eye can tolerate particles smaller than 10 µm [31]. However, according to Hoar and Schulman the droplet size range of MEs in between 5-200 nm [32].The droplet size of formulations i.e. OAVM-1 to OAVM-6 ranges from 160.1 ± 3.09 nm to 245.2 ± 2.90 nm, IPMVM-7 to IPMVM-12 ranges from 158.1 ± 2.13 nm to 203.2 ± 2.78 nm and IPPVM-13 to IPPVM-18 ranges from 152.3 ± 3.06 nm to 241.4 ± 3.24 nm (Table 2). At an optimum polymer concentration of Mod.CH (0.25% w/v) with increasing the concentration of surfactant the droplet size of each selected formulations was decreased. The droplet size of MEs reduces as increased tween 80 concentrations. These results also have support to previous published mucoadhesive chitosan-coated cationic dexamethasone microemulsion [12].

Zeta potential is an important parameter for surface characterization technique which helps in determining the potential stability and surface charge of nanodroplet system. Normally absolute large negative or positive zeta potential value is required for colloidal dispersion stability as electrostatic repulsion between particles with similar charges avoid aggregation [33]. All formulations of exhibited positive zeta potential values VCZMEs, which were OAVM-1 to OAVM-6 varies from 12.8 ± 2.13 mV to 16.7 ± 1.12 mV, IPMVM-7 to IPMVM-12varies from 10.7 ± 1.8 mV to 15.4 ± 1.60 mV, IPPVM-13 to IPPVM-18varies from 10.4 ± 1.08 mV to 14.8 ± 1.18 mV (Table 2).The positive surface potential value of VCZMEs due to presence of cationic agent such as Mod.CH (0.25% w/v) at oil/water interface moreover surfactant/co-surfactant (tween 80/propylene glycol) is non-ionic in nature.

The polydispersity index (PDI) is a marker of particle size distribution. It also indicates the stabilization of formulation. Its value ranges from 0.01 to 0.5 indicates narrow size distribution, while PDI above 0.7 indicates very broad size distribution [34].The polydispersity index of all MEs, OAVM-1 to OAVM-6 varies from 0.298 ± 0.04 to 0.461 ± 0.01, IPMVM-7 to IPMVM-12varies from 0.366 ± 0.02 to 0.453 ± 0.02, IPPVM-13 to IPPVM-18varies from 0.339 ± 0.02 to 0.441 ± 0.03 (Table 2). This indicates narrow size distribution with higher stability of MEs.

The corneal hydration level of normal mammalian cornea is between 75% and 80%. A variation in these levels shows damage to the endothelium or epithelium [23]. The selected microemulsions with optimum concentration of Mod.CH (0.25% w/v) did not show any corneal damage i.e.. they did not show any destruction to cornea. The corneal hydration value of the selected formulations (OAVM-2, IPMVM-11, IPPVM-17) and VOZOLE (1% w/v) was between 75 to 80%, which indicated that the formulations did not cause any harm to the corneal tissue. The corneal hydration level of selected VCZMEs with VOZOLE (1% w/v) reported in Table 5.

Currently used ocular drugs towards eye are associated with a number of problems complicated by the unique physiology of the eye i.e. toughest corneal barrier. Since precorneal loss of drug associated with poor bioavailability. To maintain drug in aqueous humor to enhance bioavailability of drug, this is sign of successful therapy. So it is necessary to determine the drug concentration in aqueous humor especially in case fungal keratitis which fulfils the aim of in vivo studies. VCZ concentration in aqueous humor shown in Fig. (6). Marketed formulation (VOZOLE (1% w/v)) showed the highest VCZ concentration in aqueous humor for 1 hour. However, no sustainable drug concentration obtained with VOZOLE (1% w/v). The concentration peak (C_max) of VCZ in aqueous humor was higher and sustained of 14.05 ± 0.57 µg/ml was reached at 240 min. (t_max= 4 hours) with IPMVM-11 whereas the concentration of 4.31 ± 0.58 µg/ml was reached at 60 min. (t_max= 1 hours) i.e. lowest achieved with VOZOLE (1% w/v). The sustained release of VCZ was seen to reach that peak state (t_max= 4 hours) acted as a facts for attain sustained release effect. The AUC_0→t, calculated for IPMVM-11 and VOZOLE (1% w/v) which was 28.25 µg h/ml and 4.13 µg h/ml respectively, i.e. the AUC_0→t of IPMVM-11was around 6.84-fold(p<0.01) higher than VOZOLE (1% w/v). Moreover the bioavailability of IPMVM-11 enhanced as comparatively VOZOLE (1% w/v). The possibility of drug delivery from VOZOLE (1% w/v) was lowest and the drug delivery from IPMVM-11 was as long as achieved in aqueous humor. This result suggested to longer residence time of IPMVM-11 in eye responsible for higher concentration of drug in aqueous humor.

Di Colo has shown that still in a solution form of water soluble chitosan (chitosan hydrochloride) could enhance the transcorneal penetration of ofloxacin [44]. In our present study contribution with loading of VCZ in oil phase and water soluble Mod.CH (0.25% w/v) at oil/water interphase might be responsible for enhance the residence time over cornea and might be due to transcorneal penetration occurs. Bioadhesive effect of the Mod.CH was resulted to an enhancement of the corneal permeability most probably concurrent to the polycationic nature of chitosan [1].

Fig. (6). VCZ Concentration in Aqueous Humor with Time Profile of IPMVM-11 and VOZOLE (1% w/v) in Rabbits.

Ocular irritancy and topical therapy studies were carried out for the assessment of the results obtained with healthy rabbits. In this study, the effectiveness of ME (IPMVM-11) was investigated in fungal keratitis induced rabbit eyes. Candida albicans was used to produce fungal keratitis. The fungal keratitis in all experimental rabbit groups was confirmed after 18 hours and topical therapy was initiates with IPMVM-11 and marketed formulation (VOZOLE (1% w/v)). The treatment for infected groups (group 1 and 2) were proceeding with two drops (1 drop ≈ 50 µl) at every 12 hour for a day, during 1 week. In addition, group of 3 rabbit (control group) was induced Candida albicans treated with sterile physiological saline solution, however not applied any other therapy. Images during treatment of infected rabbit eyes with Candida albicans after 0, 24, 72 and 168 hours with treatment (VOZOLE (1% w/v) and IPMVM-11) and without treatment (control) were illustrated in Fig. (7).

Fig. (7). Images during treatment of infected rabbit eyes with Candida albicans after 0, 24, 72 and 168 hrs with VOZOLE (1% w/v) and IPMVM-11.

Modified draize test performed for evaluation of rabbit eyes with representation of clinical grading system. During treatment the conjunctival redness was studied and depicted in Table 6. The control group of rabbits showed considerably higher grades than VOZOLE (1% w/v) and IPMVM-11. The clinical grading system carried out throughout 7 days study of 24, 72 and 168 hours. After 24 hours, the score of the control group was 2.96 ± 0.22 whereas score of VOZOLE (1% w/v) and IPMVM-11 was 2.44 ± 0.34 and 1.82 ± 0.47 respectively. After 72 hours, the score of the control group was 3.00 ± 0.00 whereas score of VOZOLE (1% w/v) and IPMVM-11 was 2.03 ± 0.43 and 0.93 ± 0.54 respectively. After 168 hours, the score of the control group was 3.00 ± 0.00 whereas score of VOZOLE (1% w/v) and IPMVM-11 was 1.10 ± 0.46 and 0.29 ± 0.16 respectively. The clinical grading system also depicted in Table 4.In addition IPMVM-11 applied rabbits, established lower scores than VOZOLE (1% w/v) applied rabbits in all clinical signs. This occurrence was deliberated to be due to the enhancement of residence time of drug on cornea and the concentration drug in aqueous humor was maintained higher from IPMVM-11 than VOZOLE(1% w/v).