Brugada syndrome risk scores: what we've learned and what's next

TL;DR

This paper reviews progress in predicting sudden cardiac death risk in Brugada Syndrome and suggests future directions for improving risk scores.

Contribution

The paper synthesizes existing risk scores and proposes new approaches like AI and polygenic risk scores for better risk stratification.

Findings

Current risk scores for Brugada Syndrome have limited external validation and modest predictive accuracy.

Structured risk models have improved intermediate-risk patient stratification despite guideline limitations.

Future innovations include AI on ECGs, wearable tech, and polygenic scores to enhance risk prediction.

Abstract

Brugada Syndrome (BrS) is a rare but clinically significant inherited arrhythmia disorder characterized by a type 1 ECG pattern and an increased risk of sudden cardiac death (SCD). Since its first description in 1992, BrS has been the subject of intensive investigation, yet risk stratification remains one of its greatest challenges. While survivors of cardiac arrest and patients with documented ventricular fibrillation (VF) are clear candidates for implantable cardioverter-defibrillators (ICDs), predicting risk in asymptomatic or intermediate-risk individuals is less straightforward. Over the past two decades, multiple risk scores have been developed—including the Sieira, Shanghai, BRUGADA-RISK, and PAT—each integrating combinations of clinical, ECG, electrophysiological study (EPS), and genetic data. Performance metrics vary, with C-statistics ranging from 0.70 to 0.82 in derivation cohorts, but external validation has often been limited. Importantly, current ESC and AHA/ACC guidelines only endorse syncope and EPS inducibility as validated predictors, reflecting the cautious stance of expert panels in the face of heterogeneous data. Nonetheless, the emergence of structured risk models has improved our ability to stratify intermediate-risk patients and stimulated further innovation. Looking ahead, opportunities lie in integrating artificial intelligence applied to raw ECG waveforms, wearable technology for dynamic monitoring, advanced cardiac imaging biomarkers, and polygenic risk scores. Multinational collaboration and federated learning will be essential to overcome statistical fragility and ensure global applicability. Ultimately, BrS risk scores should be considered decision-support tools that enrich but do not replace clinical judgment. Shared decision-making remains central, particularly in asymptomatic patients where ICD implantation is not a clear-cut choice.