2010 - Berlin - Germany

PAGE 2010: Methodology- Model evaluation
Julie Grenier

Population Pharmacokinetic and Pharmacodynamic Meta Analysis of Zenvia: Modeling of QT Prolongation

J. Grenier(1), JR. Lavigne(1), LE. Pope(2)

(1) Clinical Pharmacology Department, Celerion, Montrťal, Qc, Canada; (2) Avanir Pharmaceuticals, Aliso Viejo, CA, USA.

Objectives: Zenvia is a combination product of Dextromethorphan (DM) and Quinidine (Q) where Q is used at a subtherapeutic level as an inhibitor of DM metabolism by CYP2D6 enzymes. The objective of this study was to establish the relationship between the predicted plasma concentrations of Q, DM, and DX and the changes in QT intervals from baseline at 3 dose levels (including supratherapeutic).

Methods: Results from 2 thorough QT studies were pooled and included 82 subjects. A total of 7446 QT measurements were used for baseline assessment and 11,382 QT measurements were used for QT prolongation evaluation. Based on a previously developed population PK model1, individually predicted Q, DM and DX concentrations were used for QT prolongation modeling. The first step was to model all baseline and placebo QT values, using actual time. Once the most appropriate model was selected, individual baseline QT interval parameters were fixed for the model discrimination of QT prolongation. Linear, Emax and sigmoidal Emax models were tested with and without a time-delay for Q, DM and DX separately. Additive PD models were also tested. The best PD model selected was then used to fit both baseline and QT prolongation at the same time. Standardized visual predictive checks (SVPC) were evaluated for the internal validation of the final population PD model.2 

Results: The model that best fitted the baseline/placebo QT intervals used individual correction factor for RR intervals. Diurnal variations were modeled using a truncated Fourier series with 3 oscillators. Between-day variations were also fitted with a truncated Fourier series but with 2 oscillators and fitted periods. Placebo was not found to be different from baseline.3 A sigmoidal Emax model dependant on Q concentrations only was the model that best described the observed QT prolongation. The best fit was obtained by fitting the baseline parameters at the same time. No covariate was deemed significantly correlated with any of the PD parameters.

Conclusions: The wide range of available Q concentrations allowed for the characterization of a sigmoidal model relating the prolongation of QT interval to Q concentrations. Although no definitive conclusion can be made about the potential of DM or DX to induce QT interval prolongation, the plasma Q concentrations are sufficient to explain the observed drug-induced QT interval prolongation following Zenvia administration.

(1) J Grenier, JR Lavigne, L Pope. Population Pharmacokinetic Meta Analysis: Inhibition by Quinidine of the First-Pass and Systemic Metabolism of Dextromethorphan to Dextrorphan.  Abstract submitted for the PAGE meeting in Berlin, June 2010.
(2) Wang D, Standardized Visual Predictive Check in Model Evaluation - Methodology and Applications.  PAGE 18 (2009) Abstr 1501 [www.page-meeting.org/?abstract=1501]
(3) Piotrovsky V, Pharmacokinetic-Pharmacodynamic Modeling in the Data Analysis and Interpretation of Drug-induced QT/QTc prolongation, AAPS J, 2005; 7(3), E609-624.

Reference: PAGE 19 (2010) Abstr 1758 [www.page-meeting.org/?abstract=1758]
Poster: Methodology- Model evaluation
Click to open PDF poster/presentation (click to open)