Shaonan Wang, Muriel Boulton, Akash Khandelwal, Roberta Bursi
Pharmacometrics , Global Biometrics, Global Innovation, Grünenthal GmbH, Aachen, Germany
Objectives: Axomadol is a central analgesic, antinociceptive agent that was developed for the management of acute and chronic pain of moderate to severe intensity. In this dose finding trial, the primary endpoint was a numerical rating scale (NRS) assessing patient’s pain on a scale from 0=”no pain” to 10=”pain as bad as you can imagine”. The objective of the study was to explore the relationship between Axomadol concentrations and NRS pain scores assessed on a daily basis via different modelling approaches.
Methods:Data from a randomized, placebo-controlled, double-blind, phase IIb, parallel-arm study assessing the analgesic efficacy of three dose levels of Axomadol (50 mg, 75 mg, 125 mg bid) in patients with chronic hip and/or knee-joint osteoarthritis were modelled. A population PK model was previously developed [1]. The NRS pain scores could be considered as continuous or categorical data and were modelled in NONMEM7.2 [2] by three different approaches: a continuous model assuming normal distribution, a logit model and a truncated generalized Poisson model [3]. In all three approaches, a time/placebo effect component characterized by an exponential decay function as well as a drug effect component were included.
Results: All three approaches could be successfully implemented in NONMEM and adequately described the data. Models with a treatment-specific time/placebo effect fitted the data better than models using shared time/placebo effect. An Emax function best described the drug effect component. All three approaches were capable of predicting the mean NRS pain scores, as the predicted mean values agreed with the observed ones.
Conclusions: In this study, we have shown that i) Axomadol exposure drove the efficacy as reflected in the reduction of the NRS pain scores; ii) The logit and truncated generalized Poison models could be used to model NRS pain scores from a real dose finding clinical trial.
References:
[1] Grünenthal internal report. 2006
[2] Bauer, R.J., NONMEM User’s Guide Introduction to NONMEM7.2.0, 2011, Ellicott City, MD, USA: Icon Development Solutions.
[3] Plan EL, Elshoff JP, Stockis A, Sargentini-Maier ML, Karlsson MO. Likert pain score modeling: a Markov integer model and an autoregressive continuous model. Clin Pharmacol Ther. 2012; 91(5): 820-8
Reference: PAGE 23 () Abstr 3277 [www.page-meeting.org/?abstract=3277]
Poster: Drug/Disease modeling - CNS