Several risk factors for development of a potentially fatal ventricular arrhythmia, torsade de pointes, have been observed, including female gender. However, in most investigations, only few torsade events were included and/or rarely were postdose heart rate corrected QT (QTc) measurements included, as a surrogate of drug exposure. We developed a multivariate logistic regression model using data from 22,214 patients (33% women) with 84 torsade events (56% women) to evaluate the relationship between risk factors for torsade using data from four anti-arrhythmic drug development programs. Before model development, we evaluated different QT/QTc postdose metrics (average, maximum, etc.) to determine which QT metric should be included into the model. The developed multivariate model showed that, after accounting for known risk factors for torsade and postdose QTc, that female gender remained a significant risk factor for torsade.
© 2017 American Society for Clinical Pharmacology and Therapeutics.