Statistical aspects of bioequivalence assessment in the pharmaceutical industry

摘要

Since the early 1990's, average bioequivalence studies have served as the international standard for demonstrating that two formulations of drug product will provide the same therapeutic benefit and safety profile when used in the marketplace. Population (PBE) and Individual (IBE) bioequivalence have been the subject of intense international debate since methods for their assessment were proposed in the late 1980's. Guidance has been proposed by the Food and Drug Administration of the United States government for the implementation of these techniques in the pioneer and generic pharmaceutical industries. As of the present time, no consensus among regulators, academia, and industry has been established. The need for more stringent population and individual bioequivalence has not been demonstrated, and it is known that the criteria proposed by FDA are actually less stringent under certain conditions. The properties of method-of-moments and restricted maximum likelihood modelling in replicate designs will be explored in Chapter 2, and the application of these techniques in the assessment of average bioequivalence will be considered. Individual and population bioequivalence criteria in replicate cross-over designs will be explored in Chapters 3 and 4, respectively, and retrospective data analysis will be used to characterise the properties and behaviour of the metrics. Simulation experiments will be conducted in Chapter 5 to address questions arising from the retrospective data analyses in Chapters 2 through 4. Additionally, simulation will be used to explore of a potential phenomenon known as 'bio-creep' - that is the transitivity of individual bioequivalence in practice. Another bioequivalence problem is then considered to conclude the thesis; that of compaxing rate and extent of exposure between differing ethnic groups as described in ICH-E5 (1998). The properties of the population bioequivalence metric and an alternative metric will be characterised in small and large samples from parallel group studies. Inference will be illustrated using data from a recent submission and simulation studies.