The International Serious Adverse Events Consortium (iSAEC) phenotype standardization project for drug-induced torsades de pointes

Serious adverse drug reactions (ADRs) represent an important problem for clinicians and for the drug development process. Since these are rare, case series for the analysis of risk are generally accrued across multiple institutions. Thus, a key step in this process is the development of standard phenotype definitions. To facilitate this, the International Severe Adverse Event Consortium (iSAEC) has initiated a phenotype standardization project to improve case ascertainment of different types of serious ADRs.1 For each ADR phenotype, a group of investigators with expertise in relevant disciplines (clinical and basic science; regulatory affairs) was convened to develop, in a pragmatic fashion, a standardized case definition with the goal of enhancing the recruitment and usability of data sets for future genomic analysis. This manuscript describes the consensus phenotype definition for the relatively rare but potentially life-threatening ADR of drug-induced torsades de pointes (DITdP), the syndrome of polymorphic ventricular tachycardia (VT) associated with QT-interval prolongation and T-wave abnormalities. Torsades de pointes (TdP) may be self-limited arrhythmia or progress to ventricular fibrillation (VF) and cardiac arrest. It is most often caused by the use of QT-prolonging anti-arrhythmic drugs (e.g. quinidine, sotalol, dofetilide, and ibutilide),2,3 but can also occur with a wide variety of non-cardiovascular drugs including antibiotics/anti-infectives (erythromycin, clarithromycin, pentamidine), antipsychotics (thioridazine, haloperidol), and antihistamines (terfenadine, astemizole).2,3 A comprehensive list of drugs with a known or potential risk for TdP is maintained at www.QTdrugs.org. Approximately 1–5% of patients treated with QT-prolonging anti-arrhythmic drugs will develop TdP.2,3 The overall incidence of DITdP with ‘non-cardiovascular’ drugs, however, seems much smaller and has not been accurately determined due in part to the fact that most cases reported are not well characterized or are mainly derived from epidemiological and post-marketing surveillance studies4 and therefore have …