OPTIMIZATION AND EVALUATION OF NANO-LIPOSOMES OF BCS CLASS III DRUG INTENDED FOR OCULAR HERPES TREATMENT USING ARTIFICIAL NEURAL NETWORK

The objective of present experimental work was to optimize processing parameters of ocular liposomes using artificial neural network (ANN) as an optimization tool. The human eye is easily accessible site for topical dosing of drugs intended for local action or in the interior of the eye. Ganciclovir (GCV) is a synthetic analogue of 2'-deoxy-guanosine used in treatment and prevention of cytomegalovirus infections and in treatment of ocular herpes infection (orphan disease). The poor bioavailability (5-9 %) of this drug has lead development of ocular liposome to improve the bioavailability and enhance therapeutic effectiveness. GCV loaded liposomes were prepared by classical rotary flask evaporation method using HSPC/DSPG as lipid part and cholesterol as vesicle stabilizer. Box-Behnken design was applied on the selected processing parameters i.e. flask speed (X1), hydration volume (X2) and hydration temperature (X3) whereas the formulation variables were kept constant. The feed forward back propagation neural network (BPNN) method was optimized using experimental data and validated for accurate prediction of dependent variables. The obtained root mean square value of the trained ANN model was 0.00004, which indicated that the optimal model was reached. The predicted data from ANN and the experimental data were compared by chi square test which showed no significant difference. The results conclude that the ANN model provides accurate prediction and should be further explored in optimization of novel pharmaceutical formulations