Comparison of methods for predicting dissolution and the theoretical implications of particle-size-dependent solubility.

Experimental dissolution data of cilostazol suspensions and hydrocortisone powders were simulated using either the Wang-Flanagan equation (1999. J Pharm Sci 88:731-738; 2002. J Pharm Sci 91:534-542) or the method of Johnson and coworkers (1989. Int J Pharm 51:9-17; 1993. Pharm Res 10:1308-1314; 1996. Pharm Res 13:1795-1798; 2003. Drug Dev Ind Pharm 29:833-842). Both methods were able to simulate experimental data with similar accuracy. For the method of Johnson and coworkers (1989. Int J Pharm 51:9-17; 1993. Pharm Res 10:1308-1314; 1996. Pharm Res 13:1795-1798; 2003. Drug Dev Ind Pharm 29:833-842), a single set of hydrodynamic assumptions was able to simulate both cilostazol and hydrocortisone with similar accuracy. For the Wang-Flanagan equation (1999. J Pharm Sci 88:731-738; 2002. J Pharm Sci 91:534-542), significantly different diffusion layer thicknesses gave the best simulations for cilostazol and hydrocortisone, but a single value of 38 µm provided good overall simulation of dissolution. The general computational method was enhanced to make solubility dependent on particle size, according to the Ostwald-Freundlich equation; it was also able to simulate Ostwald ripening. The enhanced computational method provided no way to explain the large increase in bioavailability of cilostazol in dogs when the drug was dosed as a nanoparticle versus micronized preparation. The method provides a computational tool for exploring theoretical implications and explaining the behavior of nanoparticles.