Frances Brightman, Hitesh Mistry, Jonathan Swinton, Eric Fernandez, David Orrell, Christophe Chassagnole
Physiomics plc, Oxford, United Kingdom
Objectives: In the 90s and 00s, a dozen compounds were withdrawn from the market because of association with a rare but fatal heart arrhythmia known as Torsade de Pointes (TdP). The occurrence of TdP is assumed to associate with prolongation of the QT interval in the ECG, and this in turn is linked with inhibition of the hERG potassium channel. However, rejecting compounds based on hERG IC50 alone risks rejecting valuable compounds for treating disease, and the absence of hERG inhibition is not a guarantee that a compound will have no effect on the QT interval [1]. Furthermore, an association between QT prolongation and TdP risk is not necessarily valid for non-cardiac drugs [2]. These reports highlight a need to quantitatively assess the relationship between QT prolongation and other markers for TdP propensity for non-cardiac drugs. Here we present a new method for effective prediction of TdP risk categories [3], based on the IC50 values of hERG and two other cardiac ion channels
Methods: We have developed a mathematical model to predict TdP propensity that takes as input: 1) IC50s against a small number of specified ion channels that are routinely screened in early development and 2) the likely highest exposure in man. A leave-one-out cross validation was performed to assess the predictivity of the model against: 1) measuring just hERG alone versus hERG + hNav1.5 + hCav1.2 and 2) other models in the literature.
Results: We have shown that measuring hERG + hNav1.5 + hCav1.2, rather than hERG alone, improves the ability to predict TdP risk categories by a significant factor for two sets of compounds available in the literature. In addition, we have shown that our approach is more predictive than the standard approach, as well as being marginally better than the current state of the art [4].
Conclusions: Our initial modelling approach has demonstrated that it is possible to predict TdP risk categories, which correspond to the number of reported incidents of TdP, quite well. The model is also computationally far simpler and more efficient than a recently published mechanistic approach [4] and has marginally improved predictivity. A future project will assess how well activity against ion channels compares with changes in QT interval in terms of predicting TdP risk categories. Our ultimate goal is to assess whether some combination of ion-channel activity and ECG measurements is able to predict the propensity for TdP better than QT prolongation alone.
References:
[1] Lu, H. R. et al. Predicting drug-induced changes in QT interval and arrhythmias: QT-shortening drugs point to gaps in the ICHS7B Guidelines. British Journal of Pharmacology 154, 1427-1438 (2008).
[2] Ahmad, K. & Dorian, P. Drug-induced QT prolongation and proarrhythmia: an inevitable link? Europace 9, iv16-iv22 (2007).
[3] Redfern, W. et al. Relationships between preclinical cardiac electrophysiology, clinical QT interval prolongation and torsade de pointes for a broad range of drugs: evidence for a provisional safety margin in drug development. Cardiovascular Research 58, 32-45 (2003).
[4] Mirams, G. R. et al. Simulation of multiple ion channel block provides improved early prediction of compounds’ clinical torsadogenic risk. Cardiovasc Res 91, 53-61 (2011).
Reference: PAGE 22 () Abstr 2788 [www.page-meeting.org/?abstract=2788]
Poster: New Modelling Approaches