Yook-Hwan Noh, Hyeong-Seok Lim, Kyun-Seop Bae
Department of Clinical Pharmacology and Therapeutics, College of Medicine, University of Ulsan, Asan Medical Center, Seoul, Korea
Objectives: In general bioequivalence (BE) study is conducted in healthy volunteers with serial dense blood sampling for pharmacokinetics (PK). However, some drugs are inadequate to administer to healthy volunteers mainly for the toxicity. In this case, population modeling analysis with sparse concentration data in each patient could be an alternative option. The study was designed to explore the possible implementation of modeling method for BE test especially for drugs, the BE study of which can't be conducted in healthy volunteers for ethical reason such as cytotoxic anti-cancer agents with anastrozole as a model drug, BE study in healthy volunteers was done recently.
Methods: We evaluated the bioequivalency with AUC, and Cmax as metrics using the same concentration data sets. The plasma concentration data was obtained from a BE study conducted in the Clinical Trial Center at the Asan Medical Center (Seoul, Korea). A total of 24 healthy male Korean volunteers were enrolled and evaluated for PK analysis. Participants were randomized to receive 1 mg of the test or reference formulation, and PK parameters were calculated by noncompartmental analysis using Phoenix WinNonlin® 6.1 (Pharsight Corporation, MO, USA). The formulations were considered bioequivalent if the 90% confidence intervals of the geometric mean ratios of test to reference formulations for AUC and Cmax were within the BE limits of 0.8 to 1.25. Nonlinear mixed-effect modeling for both formulations were also conducted using NONMEM® (ICON development solutions, Dublin, Ireland) and the results were used to characterize and compare the PK. In addition, Individual AUC and Cmax were calculated using NONMEM®, which were used for the BE test for both formulaltions.
Results: In noncompartmental analysis, the 90% confidence intervals of the geometric mean ratios of test formulation to reference formulation were 0.96-1.08 for Cmax and 0.93-1.0 for AUC. In modeling analysis both formulations could be best described by a two-compartmental disposition model with lag phase in absorption and the 90% confidence intervals of the geometric mean ratios of test formulation to reference formulation were 0.96-1.04 for Cmax and 0.92-0.99 for AUClast.
Conclusions: The BE test of both test and reference formulations had similar results for PK parameters such as AUC and Cmax in both noncompartmental and modeing analysis, which suggest the possible use of modeling analysis in BE test with sparse concentration data.
Reference: PAGE 21 () Abstr 2402 [www.page-meeting.org/?abstract=2402]
Poster: Other Modelling Applications