[Application of genetic algorithm in formulation optimization of microemulsion for transdermal delivery]

Using naproxen as model drug, the formulation of microemulsion vehicle for transdermal delivery was optimized by genetic algorithm. The ranges of microemulsion composed of Tween 80, IPM, alcohol and water, were defined through construction of the pseudo-ternary phase diagrams. Systematic model microemulsions containing naproxen 1.12% were prepared by a three-factor, three-level center design method, and their diffusion studies via excised rabbit skin were performed. Using the quadratic regression model of steady-state permeation rate (Js) of naproxen as objective function, the consequence of center design experiment was optimized by genetic algorithm, and the formulation of microemulsion with highest permeation rate for naproxen was screened. The optimum formulation was composed of 21.41% Tween 80, 15.17% alcohol, 4.14% IPM, and 59.28% water, and the anticipated Js was 183.57 microg x cm(-2) x h(-1). The results of back substitution test showed the steady-state permeation rate of naproxen microemulsion prepared according to optimum formulation was 189.43 microg x cm(-2) x h(-1), which was higher than anticipated value. This result demonstrated optimizing formulation of microemulsion for transdermal delivery by genetic algorithm is feasible, reliable and reasonable.