On The Assessment of Similarity of Drug Dissolution Profiles—A Simulation Study

Abstract

In vitro dissolution has been recognized as an important element in drug development to assure the quality of drug products. It determines the rate and extent of the drug release and could be used as a surrogate for bioequivalence (BE) trials under certain conditions. Although methods for assessing the similarity of dissolution profiles between two drug products are available in the literature, most of them do not provide strong statistical and/or scientific justifications. This paper examines the properties of some commonly used methods and compares the results via a simulation study under the situations where the dissolution profile can be approximated by a linear or quadratic relationship over time. The similarity factor stated in the United States Food and Drug Administration (FDA) guideline (1) and the time series approach proposed by Chow (2) are also included. The results show that Chow’s method is very sensitive to the variation in the amount dissolved. Both methods are more relaxed and have a lower probability of declaring dissimilarity than the others.