A heuristic approach is used to identify the mathematical model best describing the release of a drug from a single particle in a polydisperse system. The method uses only experimental measures of particle size distribution and total release from the dispersion as a function of time. The routine is based on the observation that in the first 40% of total drug release the volume-surface diameter of the distribution, dvs, can be used as a single diameter representing the entire population to estimate model dependant release parameters. This property was tested in four different single particle release models having distributions of varying standard deviation; narrow, intermediate, and wide. In its application to drug imbedded in a microsphere, the particle size distribution of the system was measured and the volume-surface diameter determined. Release from the population was measured under sink conditions. The first 40% was fit (SAS) using dvs and assuming different mechanisms to be operating; release parameters characteristic of the assumed single particle mechanism were thus generated. Since only one of the assumed mechanisms should be operating, when the parameter estimates for each are used over a larger time frame (beyond 40%) in conjunction with the effects expected for individual particles of known size via measured distribution, the best single particle model should be that one giving the best overall fit. For norgestrel in a serum albumin microsphere, the single particle model best describing release was matrix diffusion of drug from a particle surrounded by a hydrodynamic layer.
ASJC Scopus subject areas