Towards a new generation of dynamic indices for the assessment of drug-induced proarrhythmic risk

Abstract

The detection of proarrhythmic side effects in many marketed drugs has prompted the search for complementary non-invasive markers of cardiotoxicity risk. Novel indices based on the study of cardiac vector dynamics have now emerged. Herein, we use quaternion methods to study alterations in the angular velocity during ventricular repolarization. The assessment was conducted in patients with Torsade de Pointes episodes when undergoing a Sotalol treatment. The algorithm includes a Principal Component Analysis to homogenize the information from three databases and, at the same time, reduce the space to three dimensions. Significant differences were found (p < 5E-4) in the angular velocity of the second half of repolarization loop (T-wave). The ratio between the maximums of the angular velocity in both T-wave halves showed a trend to 1 in the proximity of an arrhythmic event and reached sensitivity/specificity pair of 97/100 (AUC 0.99) in the comparison of healthy population with at-risk patients. The high performance of the method exceeded the expectations of the standard measurement and it becomes a promising outcome for the development of safe drug therapies.