Atrial Fibrillation and Congestive Heart Failure - AF-CHF - Presented at AHA 2007


The goal of the trial was to evaluate rhythm control with rate control among patients with heart failure and atrial fibrillation.

Study Design

Patients Enrolled: 1376
Mean Follow Up: Mean follow-up 37 months
Mean Patient Age: Mean age 67 years
Female: 18

Patient Populations:

New York Heart Association (NYHA) class II to IV CHF and LVEF ≤35% or NYHA class I with prior hospitalization for heart failure or EF ≤25%; and a documented clinically significant episode of atrial fibrillation within the past 6 months.

Primary Endpoints:

Cardiovascular mortality

Secondary Endpoints:

Total mortality, stroke, worsening heart failure, composite of CV death, stroke or worsening heart failure

Drug/Procedures Used:

Patients were randomized to rhythm control (n = 682) or rate control (n = 694). Rhythm control included use of electrical cardioversion combined with antiarrhythmic drugs, including amiodarone as first line therapy and dofetilide and sotalol if needed, and additional non-pharmacological therapy in resistant patients. Rate control included use of beta-blockers, digoxin or pacemaker and AV node ablation if necessary. Patients were to receive optimal heart failure therapy and anticoagulation.

Principal Findings:

At baseline, 31% of patients had NYHA class III or IV heart failure. Mean LVEF was 27%. Atrial fibrillation was paroxysmal in 31% of patients and persistent in 69%. By trial design, rhythm control was predominantly done with amiodarone (82%) with less use of sotalol (1.8%) and dofetilide (0.4%) in the rhythm control cohort. In the rate control group, beta-blockers were used in 88% of patients and digoxin in 75%. Crossover from rhythm to rate control occurred in 21% of the rhythm group and from rate to rhythm control in 10% of the rate group.

There was no difference in the primary endpoint of cardiovascular death between the groups (26.7% of the rhythm control group vs. 25.2% of the rate control group, hazard ratio [HR] 1.06, 95% CI 0.86-1.30, p = 0.59). There was also no difference in total mortality (31.8% vs. 32.9%, p = 0.73), stroke (2.6% vs. 3.6%, p = 0.32), worsening heart failure (27.6% vs. 30.8%, p = 0.17) or the composite of CV death, stroke, or worsening CHF (42.7% vs. 45.8%, p = 0.20) for rhythm control vs. rate control, respectively. In the rhythm control group, 39% had cardioversion compared with 8% of the rate control group (p = 0.0001). Bradyarrhythmias were more common in the rhythm control group (8.5% vs. 4.9%, p = 0.007).


Among patients with heart failure and atrial fibrillation, use of rhythm control was not associated with differences in cardiovascular mortality compared with rate control through a mean follow-up of 3 years.

Results of the present study are similar to those of the AFFIRM trial, which also showed no impact on mortality with rhythm control compared with rate control for management of atrial fibrillation. Atrial fibrillation has adverse hemodynamic effects, due in part to an excessive ventricular rate, irregularity of ventricular response, and loss of atrial contraction. These adverse hemodynamic effects could potentially have an unduly negative influence in patients with CHF. Conversely, restoring sinus rhythm can improve cardiac output, exercise capacity, and maximal oxygen consumption. Despite these potential benefits with rhythm control, no impact was observed on clinical events, even worsening heart failure.


Presented by Dr. Denis Roy at the American Heart Association Annual Scientific Session, Orlando, FL, November 2007.

Keywords: Stroke, Digoxin, Heart Conduction System, Electric Countershock, Cardiac Output, Phenethylamines, Oxygen Consumption, Heart Failure, Sotalol, Bradycardia, Sulfonamides, Atrioventricular Node

< Back to Listings