Acute canine model for drug-induced Torsades de Pointes in drug safety evaluation-influences of anesthesia and validation with quinidine and astemizole.

An acute in vivo model for drug-induced torsades de pointes (TdP) for use in safety evaluation of drugs was developed using dogs with acute complete atrioventricular (AV) block. In order to study the effects of anesthetic agents on the inducibility of TdP, arrhythmias were induced by programmed electrical stimulation (PES) before and after cumulative intravenous administration of quinidine under anesthesia with sodium pentobarbital, halothane, or isoflurane. Both prolongation of the QTc and the incidence of TdP were greatest in dogs anesthetized with halothane and were smallest in those given pentobarbital, suggesting that halothane is the most suitable anesthetic for this TdP model. To further validate this model, astemizole was administered intravenously to other dogs under halothane anesthesia. Astemizole at 0.3 mg/kg caused slight prolongation of the QT interval but did not induce any arrhythmias. At 1 mg/kg, however, TdP were induced in 5 of 10 animals and in an additional 2 animals at 3 mg/kg. Single and multiple ectopic beats preceded the induction of TdP, and the ectopic beats were observed in a dose-dependent manner. The plasma concentrations of quinidine in dogs with TdP were equivalent to or less than quinidine levels in humans with TdP, while those of astemizole were higher in dogs. In conclusion, this acute canine model of TdP with halothane anesthesia, complete AV block, PES, and simultaneous measurements of plasma drug concentration would be valuable for assessing the risk of drugs, especially I(Kr) blockers, to induce TdP in humans.