Hypertrophic cardiomyopathy: electrical abnormalities detected by the extended-length ECG and their relation to syncope.

BACKGROUND Ventricular repolarization abnormalities can represent a trigger for lethal arrhythmias in hypertrophic cardiomyopathy (HCM). We sought to assess whether multiparametric computerized surface ECG analysis identifies repolarization abnormalities in HCM patients, and whether this approach allows identification of patients with syncope. METHODS In 28 HCM patients and 102 healthy subjects (14 and 51 males, mean age 44 +/- 15 and 41 +/- 14 years, respectively), 8-lead ECG (I, II, V1-V6) was recorded for 5 min, acquired in digital format and analyzed. Heart-rate corrected QT (QTc) and T wave complexity index (TWCc), QT dispersion, activation-recovery interval (ARI) and its dispersion, signal duration in the terminal portion of the filtered QRS at 25 Hz (LAS(25 Hz)) were analyzed among other parameters. RESULTS Compared to healthy subjects, HCM patients exhibited longer QRS, filtered QRS, QTc and QTd, greater TWCc, minor ARId and LA(25 Hz). QRS duration and maximal septum thickness were linearly correlated (r=0.231 p<0.001). ARId shortening depended on ARI shortening in lead V1 (241 +/- 51 vs. 287 +/- 45, HCM vs. healthy subjects, p<0.0001) and lengthening in V6 (257 +/- 42 vs. 209 +/- 34, HCM vs. healthy subjects, p<0.0001). Significant factors for syncope at Wilks' stepwise discriminant analysis were TWCc, QRSd and LAS(25 Hz) (F=14.394, 10.098 and 9.226, respectively) with 92.3% positive predictive accuracy. CONCLUSIONS In HCM, longer QRS and QT intervals are consequences of increased left ventricular mass, while ARI seems to reflect myocardial activation rather than inhomogeneity of recovery. The simultaneous evaluation of TWC, QRSd and LAS(25 Hz), unable by itself to hold a predictive value, yielded high accuracy in predicting cardiogenic syncope.