Drug-induced QTC prolongation dangerously underestimates proarrhythmic potential: lessons from terfenadine.

Background: Terfenadine's proarrhythmia prompted market with-drawal; therapeutic antihistaminic concentration is less than 1 nM, whereas IC50 of I-Kr and I-Na exceed 200 nM. Methods and Results: Rabbit hearts were perfused with terfenadine (1-10,000 nM; 10-450 minutes). A dosage of 1 nM tended to shorten action potential duration (APD(60)) (-30 +/- 30.5 ms; n = 6); 10 nM (450 minutes) significantly prolonged APD(60) (46 +/- 11 ms; n = 6), but after 1 hour washout, APD(60) further prolonged. Above 30 nM, APD(60) shortening was followed by prolongation; net effect depended on exposure time (n = 33). In the mu M range, cardiac wavelength (lambda) shortened (APD(60) shortened, conduction slowed; P < 0.05). Terfenadine induced triangulation, reverse use dependence, instability and dispersion of repolarization (TRIaD) at 1 to 1000 nM, increasing with concentration (450 minutes: 1 nM yielded 50% of hearts, 10 nM 100%) and exposure (100 nM: 10 minutes yielded 16%, 30 minutes 33%, 150 minutes 66%, 450 minutes 100%). TRIaD with APD prolongation preceded two Torsade de Pointes, with shortening seven ventricular tachycardia and five ventricular fibrillation. Terfenadine causes normally little QT(c) prolongation in patients and Food and Drug Administration records suggest that incidence of ventricular tachycardia/ventricular fibrillation exceeds Torsade de Pointes. Conclusion: For terfenadine, TRIaD predicts drug-induced proarrhythmia: with lambda prolongation, Torsade de Pointes is preferred, otherwise ventricular tachycardia/ventricular fibrillation. APD/QT(c) alone is clearly inadequate for proarrhythmia evaluation.