Durability of the survival effect of cardiac resynchronization therapy by level of left ventricular functional improvement: fate of "nonresponders".

BACKGROUND: Although improvement in left ventricular (LV) function has been shown to portend superior short-term outcomes in patients with heart failure undergoing cardiac resynchronization therapy (CRT), the durability of this effect at 5 years has not been established.
OBJECTIVE: To determine the long-term outcomes of patients undergoing CRT on the basis of the degree of echocardiographic response.
METHODS: We extracted clinical data on a cohort of 880 consecutive patients undergoing the new implantation of a CRT device between September 30, 2003, and August 6, 2007. Patients with an ejection fraction (EF) ≤35% undergoing initial CRT implantation, with an available pre-CRT and follow-up echocardiogram, were included in the final cohort. On the basis of changes in LVEF, patients were categorized into "nonresponders" (change in EF ≤4%), "responders" (EF change 5%-20%), and "super-responders" (change in EF >20%). A Cox multivariate model was performed to determine the effect of response on long-term survival free of LV assist device or heart transplant.
RESULTS: A total of 526 patients met inclusion criteria, of whom 196 (37.3%) were classified as nonresponders, 236 (44.9%) as "responders," and 94 (17.9%) as "super-responders." In multivariate analysis, "super-responders" had the best survival and nonresponders the worst over a mean of follow-up of 5.3 ± 2.4 years. At 5 years, survival free of LV assist device or heart transplant among super-responders was 82%; responders, 70%; and nonresponders, 48%.
CONCLUSIONS: In patients with heart failure undergoing CRT, survival benefit is durable at 5 years of follow-up and its degree intimately tied to the level of improvement in ventricular function. The prognosis of nonresponders is exceptionally poor.