Nidal Al-Huniti and Diansong Zhou
Clinical Pharmacology & Pharmacometrics, AstraZeneca, Wilmington, DE 19803, USA
Objectives: Many antidepressants failed to demonstrate superiority in double blinded clinical trials due to significant placebo response (PR) in children. The purpose of current study was to explore an appropriate placebo effect model.
Methods: PR from randomized controlled trials studying antidepressants in children (6-18 yrs) with MDD was obtained from literature. The PR described as Children’s Depression Rating Scale-Revised (CDRS-R) score was used to develop a nonlinear mixed effect model with NONMEM 7.2 and nlme package in R. PR was modeled as E0*(1-Emax*time/(ET50+time)). 1000 trial simulations were conducted to evaluate the variability of placebo effect.
Results: There were 81 CDRS-R observations from 13 clinical trials for 9 drugs. The number of patients from placebo arm was in the range of 15 to 187 in these studies and the average CDRS-R score at study entry was in the range of 52 to 64.6. Parameter estimates were similar using NONMEM and nlme. The baseline CDRS-R score was estimated to be 58.6 ± 1.1 while the maximum effect (Emax) for PR was 42% ± 2%, indicating the significant PR in the clinical trials. The ET50, time inducing a response equals to one-half of Emax, was 1.9 ± 0.3 weeks. The simulations for PR also indicated significant variability of placebo effect in MDD clinical trials.
Conclusions: The nonlinear mixed effect PR model in MDD children demonstrated significant and variable placebo response. The model can provide a useful tool for evaluating time course of placebo effect.
Reference: PAGE 22 (2013) Abstr 2678 [www.page-meeting.org/?abstract=2678]
Poster: CNS