Objectives: To assess the benefits and harms of cardiac resynchronization with (CRT-D) and compared to an ICD alone, CRT without a defibrillator (CRT-P) compared with optimal medical therapy and CRT-D compared with CRT-P in patients with an EF ≤35% and a QRS duration ≥120 ms. We also sought to assess predictors of response to CRT-D and CRT-P.
Data Sources: We searched MEDLINE, Embase®, and the Cochrane Central Register of Controlled Trials (CENTRAL) from January 1, 1995, as this is the date of first article reporting use of CRT through October 20th, 2014.
Review Methods: Paired investigators independently screened search results to assess eligibility. Investigators abstracted data sequentially and assessed risk of bias independently. Investigators graded the strength of evidence as a group.
Results: CRT-D was found to be effective in reducing heart failure hospitalizations, inducing ventricular reverse remodeling, improving quality of life, and increasing six-minute hall walk distances compared to an ICD alone with a high strength of evidence. In a meta-analysis of minimally symptomatic patients, CRT-D reduced LVESV (ml) (mean difference −22.55, 95% CI −40.66 to −9.56). This analysis was comprised primarily on NYHA class II patients; therefore, the applicability to NYHA class I patients is unclear. In a meta-analysis of patients with advanced heart failure (NYHA class III–IV), CRT-D improved quality of life scores (as measured by the Minnesota Living with Heart Failure Questionnaire) (mean difference −10.91, 95% CI −12.03 to −7.27) compared to an ICD alone. CRT-P was found to be effective in improving all-cause survival and reducing heart failure hospitalizations compared to optimal medal therapy alone with a moderate level of evidence. CRT-P was also found to induce reverse ventricular remodeling and increase six-minute hall walk distances compared to optimal medical therapy alone. These findings were primarily noted in NYHA class III–IV patients. The applicability of these findings to NYHA class I–II patients is unclear. Determining predictors of response to CRT was limited by the likely presence of reporting bias. Nevertheless, a left bundle branch (LBBB) morphology, non-ischemic cardiomyopathy (NICM), and female gender were generally associated with improved outcomes following CRT-D. Sinus rhythm (as compared to atrial fibrillation) and a wider QRS duration were associated with improved outcomes following CRT-D albeit with a lower strength of evidence. There is insufficient evidence to determine predictors of outcomes in patients undergoing CRT-P. There is insufficient evidence to determine the effectiveness of CRT-D versus CRT-P. Compared to CRT-P, device infection was slightly more common in patients receiving CRT-D.
Conclusions: There is convincing evidence that CRT-D is effective with regard to improvements in multiple clinical outcomes compared to an ICD alone in patients with an LVEF≤35% and a QRS duration ≥120ms. Similarly, there is convincing evidence that CRT-P is effective in improving multiple clinical endpoints compared to optimal medical therapy alone in the same population. The certainty of these findings varies based on NYHA class. Female gender, LBBB, a wider QRS duration, sinus rhythm, and non-ischemic cardiomyopathy are associated with improved outcomes following CRT although the likely presence of reporting bias qualifies these results. More data are needed for several questions including the efficacy of CRT in patients with a non-LBBB morphology or atrial fibrillation and the comparison of outcomes in patients receiving a CRT-D vs. CRT-P device.