CHADS2 and CHA2DS2-VASc scores to predict morbidity and mortality in heart failure patients candidates to cardiac resynchronization therapy

Aims CHADS(2) and CHA(2)DS(2)-VASc scores are pivotal in assessing the risk of stroke in atrial fibrillation patients, and were recently proved to predict hospitalizations and mortality in specific clinical settings. Aim of this study was to evaluate whether these scores could predict clinical outcomes [first hospitalization for heart failure (HF) and a combined event of HF hospitalization and death for any cause] in patients candidates to cardiac resynchronization therapy and implantable defibrillator (CRT-D). Methods and results In a retrospective multicentre Italian study, we enrolled 559 consecutive HF patients candidates to CRT-D, and we grouped them in three pre-specified risk classes: low (CHADS(2)/CHA(2)DS(2)-VASc 1-2), moderate (CHADS(2)/CHA(2)DS(2)-VASc 3-4), and high (CHADS(2) 5-6/CHA(2)DS(2)-VASc 5-8). All patients underwent regular follow-up at implanting centres every 6 months; data collection was extended till the 72th month of follow-up. At a median FU of 30 months, 143 patients (25.4%) were hospitalized for HF and 110 (19.5%) died. Event-free survival analysis showed a significant difference according to baseline CHADS(2) and CHA(2)DS(2)-VASc scores (Log-Rank for HF P < 0.001 for CHADS(2) and CHA(2)DS(2)-VASc; Log-Rank for combined end-point P = 0.001 for CHADS(2), P < 0.001 for CHA(2)DS(2)-VASc). At multivariate analysis, independent predictors of endpoints were: previous atrial fibrillation (AF) or AF at implant, NYHA class, QRS duration and the CHA(2)DS(2)-VASc score (for HF hospitalization P = 0.013; for the combined event, P = 0.007), while the CHADS(2) score was not independently associated with either the end-points. Conclusion In CRT-D patients, pre-implant CHA(2)DS(2)-VASc score is an independent predictor of major clinical events at 30-month follow-up.