Monika Twarogowska, Geraldine Ayral, Pauline Traynard, Claude Magnard, Jonathan Chauvin
Lixoft
Objectives: QT interval is the period when ventricles start to contract to when they finish. But when this period becomes too long or too short, the ventricular polarization and depolarization become uncoordinated leading to irregular beating of the heart, that is arrhythmia. An example is torsade de pointes. Medications can lengthen the QT interval and in the past several drugs have been withdrawn from the market because of the connection with sudden cardiac deaths. However, sometimes the benefits of a drug are higher than the risks. This is why currently all new drugs have to undergo a study to assess their potential effect on the QT interval.
In 2015, the International Council for Harmonisation (ICH) E14 guidance agreed to use a model-based study of concentration-QT data as a primary analysis in the proarrhythmic risk assessment. Currently recommended guidelines follow the methodologies presented in the “Scientific white paper on concentration-QTc modelling” by Ch. Garnett et al.[1]. This work presents the implementation of the C-QT analysis in MonolixSuite following these recommendations. Results, obtained on data of a class III antiarrhythmic agent Dofetilide, show the modelling workflow and calculations of the standard criteria for a drug-induced QTc interval prolongation.
Methods:
- The dataset[2]: PK measurements and ECG recordings in a randomized, double-blind, 2-period crossover clinical trial; single dose of Dofetilide or placebo.
- Heart rate corrected QT (QTc) intervals were obtained with the Fridericia’s formula (the standard adopted by the FDA). Derived baseline-adjusted QTc interval (ΔQTc) and baseline-adjusted placebo-corrected QTc interval (ΔΔQTc) were used as model variables.
- The pre-specified linear mixed effect models, given in the FDA guidelines and in [1], depend on the concentration and include fixed effect parameters, influence of baseline, treatment and diurnal effects.
- Monolix was used to implement models and estimate the parameters.
- Risk assessment simulations – two-sided 90% confidence interval for ΔΔQTc – were performed in Simulx.
Results: Monolix was used to fit the Dofetilide PK data with a two-compartments model with a time delay for the absorption and the baseline-adjusted QTc with the pre-specified linear mixed effect model. A joint modelling approach, where both QTc and PK models parameters are estimated simultaneously, was compared with the FDA recommended approach, where the C-QTc model uses directly the observed time-matched PK data. Both methods yielded very close estimates. The estimated parameters and their standard errors were used to calculate the estimated mean QTc change from the baseline and the 90% confidence interval for a large range of Cmax concentrations. For Cmax concentrations above 0.5 ng/mL, the estimated mean QTc change from the baseline exceeded the 10 ms safety threshold. The advantage of the joint PK-QTc model is that the QT interval prolongation can be studied for different doses and designs (instead of different Cmax values). For instance, in the case of a 125 mg q12h regimen, 88% of patients had the QT prolongation longer than 10 ms.
Conclusions: The presented work shows step-by-step how the FDA recommended workflow for C-QTc analysis can be implemented in Monolix. The procedure is generic and can be applied to other compounds in the same way.
References:
[1] Garnett C. et al. Scientific white paper on concentration-QTc modeling. J Pharmacokinet Pharmacodyn. 2018 Jun; 45(3):383-397
[2] Johannesen L. et al. Differentiating Drug-Induced Multichannel Block on the Electrocardiogram: Randomized Study of Dofetilide, Quinidine, Ranolazine, and Verapamil. Clin Pharmacol Ther. 2014 Nov; 96(5):549-58
Reference: PAGE 29 (2021) Abstr 9617 [www.page-meeting.org/?abstract=9617]
Poster: Methodology - Estimation Methods