Not-In-Trial Simulation: Predicting cardiovascular risk from clinical trial data
Anne Chain (1), Jeanne Dieleman (2), Meindert Danhof (1), Miriam CJM Sturkenboom (2, 3), Oscar Della Pasqua (1, 4)
(1) Division of Pharmacology, Leiden/Amsterdam Center for Drug Research, Leiden University, The Netherlands; (2) Department of Medical Informatics, Erasmus University Medical Center, The Netherlands; (3) Department of Epidemiology, Erasmus University Medical Centre, The Netherlands; (4) GlaxoSmithKline, Clinical Pharmacology/Modelling and Simulation, United Kingdom
Objectives: The objective of this study is to better translate clinical findings to real life situations and resolve the discrepancies in pre- vs. post-market estimates of cardiovascular risk associated with QTc interval prolongation. Based on clinical trial simulation scenarios we demonstrate how to assess risk of not-in-trial patients, i.e., those ineligible due to inclusion/exclusion criteria.
Methods: In contrast to the long-established assumptions for the assessment of QTc interval prolongation (i.e., QTc = baseline + circadian rhythm + drug effect), a new mechanism-based tool is developed using the approach QTc (real life population) = current clinical model + effects of concomitant medications + effects of co-morbidity conditions. d,l-sotalol is used as a paradigm compound to assess the effects of co-medications and co-morbidities in the Rotterdam Study cohort as reference population. The additional effects are evaluated by calculating the absolute differences in QTc prolongation between taking d,l-sotalol alone and in conjunction with co-medications and comorbidities. Then the final distribution of QTc values associated with all causal factors is simulated and compared non-parametrically with the observed QTc distribution.
Results: Using the well established clinical model, simulations of the drug effect on the reference cohort showed that it is insufficient to describe the high observed QTc values. However, calculations of the absolute differences in QTc prolongation between taking d,l-sotalol alone and in conjunction with co-medications and comorbidities revealed that the additional causal factors provide additive effects. Final risk assessment is achieved by combining all the causal factors in a single simulation where the distribution of the observed QTc values is confirmed to fall within the simulated distribution.
Conclusions: QT-prolongation has become the second most common cause for post-market drug withdrawal. This situation strongly suggests that the current approach to clinical evaluation of cardiovascular risk continues to lack the predictive power required to translate findings in a clinical setting to real life situations. Our work combines all relevant causal factors that affect QTc prolongation in an integrated PKPD model, thereby enabling better estimation of the true risk-benefit ratio and possibly mitigating future drug-withdrawal due to cardiovascular safety.
