Long-term Impact of CRT in Mild Heart Failure
REVERSE was a prospective, randomized, double-blind, parallel-controlled multinational study designed to determine whether CRT limited the progression of HF compared with optimal medical therapy alone. The study included ACC/AHA stage C, NYHA Class I or II HF patients with QRS ≥120 ms, and LVEF ≤40% on optimal medical therapy. Patients were implanted with a CRT-device with (CRT-D) or without (CRT-P) defibrillator and randomized 2:1 to CRT ON or CRT OFF. Patients were then programmed as randomized through 12 months in North America and through 24 months in Europe. By the study design, all patients were programmed to CRT ON after the randomization period and followed for 5 years from implantation. First enrolment occurred in September 2004, and enrolment was completed in September 2006. The final 5-year follow-up was in November 2011. The present analysis is confined to subjects randomized to CRT ON, who received up to 5 years of CRT therapy.
Long-term results analysis pre-specified in the protocol included NYHA class, 6-min hall walk, QoL, echocardiographic parameters, all-cause mortality, adverse events, and HF hospitalization. The primary endpoint in the main study was the clinical composite endpoint and the second powered endpoint was reverse remodelling measured by the left ventricular end-systolic volume index (LVESVi). Quality of life was measured by the Minnesota Living with heart failure Questionnaire and the Kansas City Cardiomyopathy (KCCQ) questionnaire. During the randomized period, patients were evaluated every 6 months with NYHA class, 6-min hall walk, and QoL by medical staff unaware of randomization assignment. Thereafter, patients were evaluated annually by staff unblinded to CRT programming. Likewise, echocardiographic data were assessed by core-labs and HF-related hospitalizations and mortality data by an independent endpoint committee unaware of CRT programming for the randomized phase of the study and thereafter unblinded to programming.
To assess risk of CRT therapy and balance it to potential benefit, analysis was focused on LV lead-related adverse events requiring invasive intervention or resulting in termination of significant device function as adjudicated by the independent Adverse Event Advisory Committee over the 5 years of study follow-up.
All the patients randomized to CRT ON with data available were used in all analyses. There was no imputation of missing data. Continuous variables are reported as mean ± standard deviation. Survival analysis was performed using the Kaplan–Meier method. Information on hospitalizations was available in 369 of 419 CRT ON patients for the entire 5-year period. The other 50 patients were in a subset (17 of 35) of European centres and the Canadian centre which did not collect hospitalizations after randomized follow-up (12 or 24 months). Patients at these centres were censored from the Kaplan–Meier curve on their last date of randomized follow-up. All hospitalizations collected through randomized follow-up and all hospitalizations through 5 years in the USA were adjudicated for HF relatedness by the AEAC. In Europe after 2 years, the investigator answer to the question 'Was worsening of heart failure signs/symptoms the reason for admission or an event which occurred during the patient's hospitalisation?' was used as adjudication. One-sample t-tests were used to compute P-values, and the P-values were not adjusted for multiplicity.
Methods
REVERSE was a prospective, randomized, double-blind, parallel-controlled multinational study designed to determine whether CRT limited the progression of HF compared with optimal medical therapy alone. The study included ACC/AHA stage C, NYHA Class I or II HF patients with QRS ≥120 ms, and LVEF ≤40% on optimal medical therapy. Patients were implanted with a CRT-device with (CRT-D) or without (CRT-P) defibrillator and randomized 2:1 to CRT ON or CRT OFF. Patients were then programmed as randomized through 12 months in North America and through 24 months in Europe. By the study design, all patients were programmed to CRT ON after the randomization period and followed for 5 years from implantation. First enrolment occurred in September 2004, and enrolment was completed in September 2006. The final 5-year follow-up was in November 2011. The present analysis is confined to subjects randomized to CRT ON, who received up to 5 years of CRT therapy.
Outcomes
Long-term results analysis pre-specified in the protocol included NYHA class, 6-min hall walk, QoL, echocardiographic parameters, all-cause mortality, adverse events, and HF hospitalization. The primary endpoint in the main study was the clinical composite endpoint and the second powered endpoint was reverse remodelling measured by the left ventricular end-systolic volume index (LVESVi). Quality of life was measured by the Minnesota Living with heart failure Questionnaire and the Kansas City Cardiomyopathy (KCCQ) questionnaire. During the randomized period, patients were evaluated every 6 months with NYHA class, 6-min hall walk, and QoL by medical staff unaware of randomization assignment. Thereafter, patients were evaluated annually by staff unblinded to CRT programming. Likewise, echocardiographic data were assessed by core-labs and HF-related hospitalizations and mortality data by an independent endpoint committee unaware of CRT programming for the randomized phase of the study and thereafter unblinded to programming.
Adverse Effects
To assess risk of CRT therapy and balance it to potential benefit, analysis was focused on LV lead-related adverse events requiring invasive intervention or resulting in termination of significant device function as adjudicated by the independent Adverse Event Advisory Committee over the 5 years of study follow-up.
Statistical Methods
All the patients randomized to CRT ON with data available were used in all analyses. There was no imputation of missing data. Continuous variables are reported as mean ± standard deviation. Survival analysis was performed using the Kaplan–Meier method. Information on hospitalizations was available in 369 of 419 CRT ON patients for the entire 5-year period. The other 50 patients were in a subset (17 of 35) of European centres and the Canadian centre which did not collect hospitalizations after randomized follow-up (12 or 24 months). Patients at these centres were censored from the Kaplan–Meier curve on their last date of randomized follow-up. All hospitalizations collected through randomized follow-up and all hospitalizations through 5 years in the USA were adjudicated for HF relatedness by the AEAC. In Europe after 2 years, the investigator answer to the question 'Was worsening of heart failure signs/symptoms the reason for admission or an event which occurred during the patient's hospitalisation?' was used as adjudication. One-sample t-tests were used to compute P-values, and the P-values were not adjusted for multiplicity.