Cardiac Resynchronisation Therapy and AV Nodal Ablation Trial in Atrial Fibrillation Patients (CAAN-AF)

Background: Cardiac Resynchronization Therapy (CRT) is an established treatment in heart failure (HF) patients with ventricular dyssynchrony who remain in sinus rhythm. Available clinical data has shown inferior outcomes of CRT in HF patients with co-existent atrial fibrillation (AF), who comprise up to 27% of HF patients, and are over-represented in advanced HF classes. We hypothesize, based on the results of a systematic review we recently published in the Journal of the American College of Cardiology, that AV nodal ablation may improve survival, heart failure and functional outcomes in CRT recipients with co-existent AF.

Design: This study will be a multicentre, prospective, randomized controlled trial. Patients with ischemic or nonischemic cardiomyopathy heart failure (NYHA II, III or ambulatory class IV), left ventricular dysfunction (EF ≤ 35%), prolonged intraventricular conduction (QRS duration ≥ 120ms), and persistent or permanent AF will be considered for the study. Persistent AF will be defined as patients where obtaining and maintaining sinus rhythm is deemed either not worthwhile, or to be ineffective in the long term, or where both the patient and the physician accept the presence of AF, where rhythm control intervention is, by definition, no longer pursued. Permanent AF is defined as AF where sinus rhythm cannot be restored.

Eligible subjects will be randomized into one of two arms: (1) CRT-D plus AV nodal ablation ("AV nodal ablation arm [AVNA]") or (2) CRT-D alone ("rate control arm").

Enrollment: 590 subjects, with 295 subjects in the AV node ablation arm and 295 subjects in the control arm, will be enrolled. Study patients will undergo stratified randomization at ≥ 30 days after CRT implant. Participants in will sign informed consent and be screened prior to randomisation. After CRT implant, patients will have at least 30 days for optimisation of heart failure therapy, prior to randomisation.

Randomisation: A computer-generated web-based randomisation schedule will be used. Randomisation will be stratified by trial centre. Randomisation is considered the trial entry point.

Outcomes: The primary endpoint is a composite of all cause mortality and non-fatal heart failure events. Secondary endpoints include all-cause mortality, cardiovascular mortality (including classification in terms of suddenness and arrhythmic mechanism by prespecified Hinkle-Thaler criteria), non-fatal heart failure events, 6-minute walking test distance, quality of life, unplanned hospitalization, and ventricular arrhythmias requiring device therapy, inappropriate shocks, cardiovascular MRI prediction of response, percentage pacing and prediction of response to therapy, ventricular reverse remodeling.

Statistical Plan: The study is powered to find a 25% relative reduction in event rate, with sample sizes calculated assuming a two-tailed α=0.05,1-β=0.80, and 10% sample size increment allow for to drop in the event rate (AV nodal ablation arm), drop out or cross-over (feasibly, control to AVNA arm only). It is planned to perform three interim (0.25, 0.5 0.75 information fractions) and a final analysis requiring 295 patients per arm with a final P-value at ≤ 0.045; stopping rules according to the method of O'Brien and Fleming. The boundaries (z scores: ±4.332, ±2.963, ±2.359, ±2.014; and nominal P-values: 0.000015, 0.0031, 0.014, 0.044)) were derived using the statistical package PASS (V12). Outside of these defined analyses, the Data Safety Monitoring Board (DSMB) will have access to data reports and will be able to stop the trial at any time.

All analyses will be based on the intention-to-treat principle. The primary (binary) mortality-outcome will be analysed using the Cochran-Mantel-Haenszel statistic and logistic regression with pre-specified (baseline) covariates. Time-to-event analyses will be initially undertaken by the Kaplan-Meier survival analysis approach. Key secondary outcomes such as all-cause mortality, cardiovascular mortality, unplanned hospitalisation, and rates of ventricular arrhythmia episodes will be analysed using either Cox proportional hazards models or Fine and Gray competing risks regression as appropriate. Continuous secondary outcomes such as the 6-minute walking distance, Short Form 36 (SF36) scores, Minnesota Living with Heart Failure (MLWHF) score will be compared between randomised groups over time using linear mixed effects models.

Significance and Impact: The study will answer a central clinical question directly impacting the care of HF patients with AF, and will be expected to change current HF management guidelines.