Models of torsades de pointes: effects of FPL64176, DPI201106, dofetilide, and chromanol 293B in isolated rabbit and guinea pig hearts.

For studying the torsades de pointes (TdP) liability of a compound, most high and medium throughput methods use surrogate markers such as HERG inhibition and QT prolongation. In this study, we have tested whether isolated hearts may be modified to allow TdP to be the direct readout. Isolated spontaneously beating rabbit and guinea pig hearts were perfused according to the Langendorff method in hypokalemic (2.1 mM) solution. The in vitro lead II ECG equivalent and the incidence of TdP were monitored for 1 h. In addition, heart rate, QTc, Tp-Te, short-term variability (STV), time to arrhythmia, and time to TdP were also analyzed. FPL64176, a calcium channel activator; and DPI201106, a sodium channel inactivation inhibitor, produced TdP in isolated rabbit and guinea pig hearts in a concentration dependent manner; guinea pig hearts were 3- to 5-fold more sensitive than rabbit hearts. Both compounds also increased QTc and STV. In contrast, dofetilide, an IKr inhibitor, produced no (or a low incidence of) TdP in both species, in spite of prolongation of QTc intervals. Chromanol 293B, an IKs inhibitor, did not produce TdP in rabbit hearts but elicited TdP concentration dependently in guinea pig hearts even though the compound had no effect on QTc intervals. IKs inhibition appears to be more likely to produce TdP in isolated guinea pig hearts than IKr inhibition. Chromanol 293B did not produce TdP in rabbit hearts presumably due to a low level of IKs channels in the heart. TdP produced in this study was consistent with the notion that its production was a consequence of reduced repolarization reserve, thereby causing rhythmic abnormalities. This isolated, perfused, and spontaneously beating rabbit and guinea pig heart preparation in hypokalemic medium may be useful as a preclinical test model for studying proarrhythmic liability of compounds in new drug development.