Drug response hysteresis in the concentration-QTc analysis of early clinical trials

According to the International Conference on Harmonisation E14 QAs R3, concentration-QTc analysis can serve as an alternative to the by-time-point analysis or intersection-union test as the primary basis for decisions to classify the QTc risk of a drug. In a recent scientific white paper on concentration-QTc analysis, a pre-specified linear mixed effect model was suggested to study a QTc prolongation effect. The model assumes a direct time-concordant relationship (direct effect) between QTc interval and drug-concentrations. In the case of lagged drug QTc effects, also called 'hysteresis', a linear direct effect model may yield a biased QTc estimate. In this work, we estimate the bias of QTc effects via simulations when hysteresis is not accounted for in the linear mixed effect model analysis. Simulations are performed to compare different methods of identifying hysteresis when two physiologically plausible QT drug mechanisms are considered. The focus of this paper is to provide hysteresis identification methods and assess the influence of hysteresis in estimating the QT prolongation for most commonly observed QT-Concentration profiles.