Men and Women Remodel Differently in Chronic Systolic Heart Failure: Analyses from MIRACLE and MIRACLE-ICD

Abstract

Background: Predictors of ventricular remodeling are well established in hypertensive heart disease and transmural infarction. Less is known about clinically relevant predictors of remodeling in chronic systolic heart failure. We tested the hypothesis that specific baseline characteristics would predict progressive left ventricular enlargement in a large cohort of patients with chronic systolic heart failure. Methods: MIRACLE (Multicenter InSync Randomized Clinical Evaluation) and MIRACLE-ICD were prospective, double-blind, randomized clinical trials designed to test the efficacy and safety of CRT and combined CRT/ICD therapy in outpatients with chronic systolic heart failure. We combined these studies into a single cohort since the inclusion criteria, mode of CRT, and timing of endpoint assessments were similar. Out of 1008 randomized participants, 776 underwent quantitative echocardiography at baseline and at six months. We used univariate and stepwise multivariable linear regression to determine which baseline characteristics predicted an increase or decrease in left ventricular end diastolic volume (LVEDV) over the subsequent six months. Results: LVEDV increased in 308 (40%) and decreased in 468 (60%) patients. In multivariable models, male gender, severity of mitral regurgitation, and plasma BNP levels were independent predictors of ventricular enlargement after adjustment for baseline LVEDV and randomization to CRT (all p<0.01; see Table for predicted changes in LVEDV). CRT was an independent predictor of a decrease in LVEDV, with greater reduction among patients with longer baseline QRS. In our final model, age and beta-blocker use did not independently predict change in LVEDV. Conclusion: Male gender, mitral regurgitation, and plasma BNP are independent predictors of left ventricular enlargement in chronic systolic heart failure. These findings emphasize the importance of volume load, caused by regurgitant valve lesions or manifested by increased BNP, in mediating ventricular remodeling. Moreover, they demonstrate that men and women may remodel differently, with men showing a higher likelihood of progressive LV enlargement. Further studies are indicated to determine genetic and gender-specific modifiers of ventricular remodeling. Tabled 1 Predicted Change in LVEDV (mL) at Six Months (CRT OFF) Gender MR index Plasma BNP (pg/mL) Male Female 0.1 0.3 100 500 1000 4.5 −19 −0.2 9.1 −6.3 1.1 4.2 Open table in a new tab