Predictive factors of ventricular fibrillation triggered by pause-dependent torsades de pointes associated with acquired long QT interval: role of QT dispersion and left ventricular function

INTRODUCTION: Death due to acquired torsades de pointes usually is caused by ventricular fibrillation (VF), but the contributing factors to VF triggered by pause-dependent torsades de pointes are not understood. METHODS AND RESULTS: We evaluated 91 patients who fulfilled four criteria: (1) pause-dependent torsades de pointes; (2) prolonged QT interval and/or corrected QT (QTc) (>0.44 sec); (3) long-short initiation sequence; and (4) conditions known to induce pause-dependent torsades de pointes. There were 38 patients with a documented VF (group I) and 53 without VF (group II). Absolute and relative dispersions of QT and QTc were calculated based on the 12-lead standard ECG. Group I differed from group II with regard to myocardial infarction history (32% vs 13%; P = 0.035), left ventricular ejection fraction (44% +/- 14% vs 65% +/- 9%; P < 0.0001), presence of structural heart disease (100% vs 20.8%; P < 0.0001), QT mean (591 +/- 73 msec vs 514 +/- 78 msec; P < 0.0001), QTc mean (563 +/- 76 msec vs 508 +/- 90 msec; P = 0.002), absolute QT dispersion (166 +/- 56 msec vs 84 +/- 49 msec; P < 0.0001), relative QT dispersion (9.9% +/- 3.5% vs 6.3% +/- 3.2%; P < 0.0001), absolute QTc dispersion (158 +/- 57 msec vs 81 +/- 44 msec; P < 0.0001), and relative QTc dispersion (9.9% +/- 3.6% vs 6.2% +/- 3%; P < 0.0001). Multiple regression analysis showed that ejection fraction (P = 0.0001), presence of structural heart disease (P < 0.0001), and relative QTc dispersion (P = 0.038) were the only independent predictors of VF. CONCLUSION: Left ventricular function, presence of structural heart disease, and QTc relative dispersion should be evaluated carefully in patients with conditions susceptible to inducing torsades de pointes.