Systematic evaluation of dose proportionality studies in clinical pharmacokinetics

An understanding of dose proportionality is essential in drug development, and the results are of great clinical importance for predicting the effects of dose adjustments. However, little consensus exists with regard to study design and analysis. The aim of this paper was to produce a detailed profile of the information on dose proportionality studies in the last 10 years and to provide a foundation for reflection and debate on future priorities. A total of 147 publications comprising 156 studies were analyzed. The typical dose proportionality study enrolled 20 to 30 subjects and randomly allocated them into 3 to 4 dose levels to investigate pharmacokinetic behaviors within a dose ratio range of 2-6. The most common design was the crossover experiment (52.6%), and evaluating dose-adjusted pharmacokinetic parameters followed by hypothesis testing (43%) was the most frequent statistical approach. However, the alternative crossover design and equivalence criterion based on the power model represented only 4% and 8% of studies, respectively. The power model as a recommendable empirical relationship to assess dose proportionality was applied in 25 (16%) studies. This research suggests that the alternative crossover design and power model statistical method should be attracting more attention in order to obtain more information in studies with limited subjects.