Predictors of spontaneous reverse remodeling in mild heart failure patients with left ventricular dysfunction.

BACKGROUND There are limited data regarding factors associated with spontaneous left ventricular reverse remodeling (S-LVRR) among mildly symptomatic heart failure (HF) patients and its prognostic implications on clinical outcomes. METHODS AND RESULTS Best subsets logistic regression analysis was used to identify factors associated with S-LVRR (defined as ≥15% reduction in left ventricular end-systolic volume at 1-year of follow-up) among 612 patients treated with internal cardioverter defibrillator-only therapy in Multicenter Automatic Defibrillator Implantation Trial-Cardiac Resynchronization Therapy (MADIT-CRT) and to create a score for the prediction of S-LVRR. Cox proportional hazards regression modeling was used to assess the clinical outcome of all internal cardioverter defibrillator-only patients (n=714) with a high S-LVRR score. S-LVRR occurred in 25% of internal cardioverter defibrillator-only patients. Predictors of S-LVRR included systolic blood pressure≥140 mm Hg, serum creatinine<1.0 mg/dL, QRS 130 to 160 ms, and nonischemic cardiomyopathy. Multivariate analysis showed that each 1-point increment in S-LVRR score (range, 0-7) was associated with an 11% (P=0.019) reduction in the risk of HF or death. Treatment with cardiac resynchronization therapy was associated with a significant reduction in the risk of HF or death only among internal cardioverter defibrillator-treated patients with a low (Q1-3) S-LVRR score (hazard ratio=0.55; P<0.001), but not among those with a higher (Q4) score (hazard ratio=1.06; P=0.72). CONCLUSIONS Our data suggest that approximately one quarter of mild HF patients eligible for biventricular pacing experience S-LVRR. Combined assessment of clinical factors associated with S-LVRR can be used to identify mild HF patients with a low risk for clinical events without cardiac resynchronization therapy intervention. CLINICAL TRIAL REGISTRATION URL http://www.clinicaltrials.gov. Unique identifier: NCT00180271.