Improving the Detection of Subtle IKr-Inhibition

Abstract

QT prolongation is an incomplete measure of drug-induced changes in repolarization. In this study, we investigated a novel, automatic ECG technique for describing ventricular repolarization morphology and we compared these results to corrected QT (QTc) prolongation for identifying ECGs of healthy individuals on moxifloxacin. We analysed data from the US FDA ECG Warehouse involving 160 standard ECGs from 40 healthy subjects enrolled in a randomized, parallel, placebo-controlled, 'thorough QT' study. Computerized ECG analysis included a series of scalar and vectorial parameters describing duration of repolarization segments and T-wave/loop morphology including its symmetry, amplitude and shape. Binary logistic models for the identification of moxifloxacin-induced abnormalities of the repolarization were developed. Moxifloxacin induced significant changes in several ECG parameters including QT and QT apex and early repolarization duration (ERD)(30)(%), T-wave amplitude and slopes of the ascending and descending arm of the T-wave. The logistic model based only on T-wave morphology parameters outperformed the model based on QTc interval for identifying the presence of moxifloxacin. Combining information about repolarization interval duration with T-wave morphology significantly improved the detection of presence of moxifloxacin (p < 0.01). The increased sensitivity of our novel ECG method contributes to a >40% reduction in the sample size required to detect significant QTc prolongation induced by moxifloxacin. Repolarization morphology is significantly altered by moxifloxacin. The computerized ECG technique provides a novel method for quantifying morphological changes of repolarization segment. Our new parameters reflecting the morphology of the T-wave outperformed QTc measurements when identifying moxifloxacin-induced blockade of the outward rapid components of the delayed rectifier repolarizing potassium current (I(Kr)). These data indicate that the analysis of T-wave morphology could play a role in the assessment of drug toxicity.