Risk stratification of ventricular fibrillation in Brugada syndrome using noninvasive scoring methods

Risk stratification for ventricular fibrillation (VF) in patients with Brugada syndrome (BrS) remains controversial.

The purpose of this study was to construct a novel prediction model for VF risk in BrS patients using noninvasive parameters.

A total of 143 Japanese BrS patients with VF (n = 35) and without VF (n = 108) were retrospectively enrolled. We built a logistic regression model predicting VF occurrence and evaluated it by cross-validation.

Frequencies of history of syncope and spontaneous type 1 ECG, r-J interval in V1, QRS duration in V6, and LAS40, Tpeak-Tend dispersion, and max T-wave alternans were significantly associated with VF occurrence in univariate analyses. The history of syncope, r-J interval in V1, QRS duration in V6, and Tpeak-Tend dispersion were identified as independent predictors by multivariate logistic regression analysis. The predictive model was constructed using all these parameters with good discrimination of VF occurrence (area under the curve 0.869 with 97.1% sensitivity and 65.7% specificity). The area under the curve based on leave-one-out cross-validation was 0.845, with 97.1% sensitivity and 63.0% specificity suggesting good performance of the model. Retrospective survival analysis revealed that the cumulative VF event rate was significantly higher in patients at high risk than in those with low risk using the log rank test (P = 2.97 × 10(-8)). Notably, no BrS patient below the cutoff value developed a subsequent VF event.

This novel prediction method may effectively assesses VF risk in BrS patients, especially when determining implantable cardioverter-defibrillator placement for asymptomatic BrS patients.