Claire H. Li (1,2), Eric A. Sherer (1,2,3), Lionel D. Lewis(4), Robert R. Bies(1,2)
(1) Indiana University School of Medicine, Indianapolis, IN, USA (2) Indiana Clinical and Translational Sciences Institute (CTSI), Indianapolis, IN, USA (3) Roudebush Veterans Affairs Medical Center, Indianapolis, IN, USA (4) Dartmouth Medical School and Dartmouth Hitchcock Medical Center, Hanover, NH, USA
Objectives: Food effects can have a significant impact on the bioavailability of oral anticancer drugs. The objective of this study was to assess inter-individual variability and inter-occasion variability in drug exposure utilizing a proposed oncology clinical trial design for detecting known food effects on clearance for abiraterone, a drug with a large food effect, and nilotinib, a drug with a smaller food effect.
Methods: Clinical trials of abiraterone and nilotinib were simulated based on a proposed oncology clinical trial design. Hypothetical patients, whose pharmacokinetic characteristics mimic those of the actual, cancer patient population, were given simulated doses of 1000 mg abiraterone once daily or 300-400 mg of nilotinib twice daily and virtual blood samples were taken at week 1, week 4, month 2, and month 3. Simulations of clinical trials with sample sizes of 20, 50 and 70 patients were replicated 100 times each using inter-occasion variability levels (exclusive of the food effect) of 10%, 25%, and 40%. Food effects on clearance were incorporated in the pharmacokinetic parameters and nonlinear mixed effect modeling (NONMEM) was used to evaluate the food effect on pharmacokinetic parameters, inter-individual variability, and inter-occasion variability. The model performance was evaluated by goodness-of–fit plots as well as the bias and precision of the parameters and variability.
Results: The percent bias and precision on population clearance in both abiraterone and nilotinib trials ranged within ± 20%. A trend of underestimate the reference values were showed in both treatments. A significant improvement in accuracy as well as variance was found in the individual clearance estimates with only ± 5% deviation. Percent bias and precision of Inter-individual variability were approximately 30% for abiraterone and 10-20% for nilotinib. Both trials showed a noticeable decrease in the inter-occasion variability bias and precision as the number of patients or IOV increased. Overall, besides nilotinib trials with 70 patients and 40% inter-occasion variability , greater than 80% of trials in either abiraterone or nilotinib treatment were able to detect food effect in a statistical significant level (p-value < 0.05, df=1).
Conclusions:Â The simulated trials suggest that this proposed oncology trial design can be powered to identify inter-occasion variability and inter-individual variability within individuals in the presence of a potential food effect, which impacts the overall drug exposure in cancer patients treated with abiraterone or nilotinib, under different levels of variability.
Reference: PAGE 21 (2012) Abstr 2559 [www.page-meeting.org/?abstract=2559]
Poster: Oncology