Heart Rate Variability: a possible tool for risk stratification in the Brugada syndrome

Purpose: Risk stratification in Brugada syndrome (BrS) is still uncertain matter of debate. Heart rate variability (HRV) stratifies risk well in various conditions. Given the relationship between high vagal tone and cardiac events in BrS we assessed the potential prognostic role of HRV in BrS. Methods: We collected 159 24-h ECG Holters (H-Scribe 4.33.12 by Mortara Instrument, Inc) of 94 BrS patients (84% males, mean age 43.4±11 years; 9 patients (pts) with VT/VF, 50 with a spontaneous diagnostic pattern and 44 drug-induced). For each patient the average 24-h pNN-50, RMS-SD and M-SD were computed. For 56 pts pNN-50, RMS-SD, M-SD and the percentage of time showing a spontaneous pattern were also computed every hour. Results: Among the three HRV indicators pNN-50 showed a significant discriminatory power in differentiating patients with and without VT/VF [Area Under the Curve (AUC)=0.73; 95% CI 0.62-0.81, p=0.02]. A cut-off value of pNN-50>11% corresponded with the highest average of sensitivity and specificity (spec. 66%, sens.78%, PPV 19%, NPV 97%). Mean pNN-50 in the study population was 12±11% (normal value in healthy controls from the literature 9±7%). pNN-50 was higher in VT/VF pts (18±13% vs 11±10%, p=0.03). At univariate analysis neither age, sex, presence of a spontaneous diagnostic pattern, positive family history for sudden cardiac death nor previous syncope were VT/VF predictors. Only pNN-50>11% was a significant VT/VF predictor [OR 6.8 (95% CI 1.3-34.6) p=0.02] despite a moderate inverse correlation between age and pNN-50 (R= -0.44, p<0.01). We didn't find any significant difference in term of hourly percentage of spontaneous pattern time between day and night (11.1%±22.9% vs 6.5%±19.4%, p=0.25). Nevertheless pNN-50 was found to be significantly lower during the phases showing a spontaneous pattern on Holter monitoring compared to the others (7.9%±6.7% vs 14.9%±12.9%, p<0.01). This result may explain the fact that the presence of a spontaneous diagnostic pattern was not a predictor of VT/VF [OR 0.7 (95% CI 0.2-2.7) p=0.58]. Conclusions: Heart rate variability may contribute to risk stratification in BrS pts as pNN-50>11% was the only significant predictor of VT/VF with a high negative predictive value.