Dr Denis Roy

"This trial was designed to show the potential benefits of maintaining sinus rhythm on cardiovascular mortality, but we did not do that," he noted. Consequently, "a routine strategy of rhythm control cannot be advocated in heart-failure patients with AF. Rate control is a simpler strategy, with less cardioversion, fewer hospitalizations," he noted.

Discussant of the AF/CHF trial, Dr Rodney Falk (Harvard Medical School, Boston, MA), said: "The AF/CHF trial results have provided us with another extremely important step forward in the understanding and treatment of this common arrhythmia. We may feel a collective twinge of disappointment, but I suspect that it is probably overridden by a sigh of relief. We can now manage the common clinical situation of coexisting AF and systolic heart failure by rate control alone, without concern that we may worsen the clinical condition or even hasten the patient's demise."

Falk pointed out, however, that the type of patient enrolled in AF/CHF was relatively stable, and rhythm control may yet have a part to play in heart-failure patients in whom the onset of AF is associated with a clear-cut deterioration in clinical status.

We can now manage the common clinical situation of coexisting AF and systolic heart failure by rate control alone, without concern.

Roy explained that no previous trial looking at rhythm vs rate control in AF has included patients with heart failure. But up to 40% of heart-failure patients have AF, and the arrhythmia is thought to be an independent predictor of mortality, he said.

AF/CHF, a prospective, open-label, multicenter trial, enrolled patients with left ventricular ejection fraction (LVEF) of 35% or less and symptoms of CHF (NYHA class 2-4). Asymptomatic patients could be enrolled, however, if they had a prior hospitalization for CHF or an LVEF <25%. In addition to the CHF criteria, patients had to have a history of significant AF—defined as either one episode lasting more than six hours within the past six months or one episode of shorter duration but with prior electrical cardioversion.

The 1376 patients were randomly allocated to rhythm or rate control between May 2001 and June 2005 at 123 sites in North America, South America, and Europe, with a minimum follow-up of two years. More than two-thirds of patients were enrolled having a persistent type of AF, and more than 50% had been previously hospitalized for AF or CHF.

Electrical cardioversion was performed within six weeks of randomization in patients in the rhythm-control group who did not convert to sinus rhythm after antiarrhythmic therapy. If necessary, a second cardioversion was recommended within three months of enrollment. Amiodarone was the initial drug of choice, and sotalol and dofetilide were used in selected cases. Pacemakers were recommended to control bradycardia and to allow continued antiarrhythmic drug administration. Patients refractory to antiarrhythmic drug therapy could be referred for additional nonpharmacologic therapies such as atrial catheter ablation.

Those in the rate-control arm received titrated doses of beta blockers and digitalis, or both, and pacemaker and AV-node ablation if needed to achieve target heart rate. Patients in both groups received optimal heart-failure management with ACE inhibitors, beta blockers, and anticoagulation therapy. The percentage of patients treated with pacemaker, ICD, and catheter ablation was similar between the two groups, Roy noted.

The primary end point of the study was cardiovascular mortality. Prespecified secondary end points included total mortality, worsening CHF, stroke, hospitalization, quality of life, and cost of therapy.

Intention-to-treat analysis revealed no difference in the primary end point between the two groups. Cardiovascular death occurred in 182 (26.7%) of the patients in the rhythm-control group compared with 175 (25.2%) in the rate-control arm (hazard ratio 1.058, p=0.59).

Total mortality, worsening CHF, and stroke were similar between the two groups, as was the composite end point of cardiovascular death, worsening CHF, and stroke. Cardiovascular mortality was also similar between the two arms in a number of prespecified subgroups.

During the course of the study, 21% of patients crossed over from rhythm to rate control, primarily because of the inability to maintain sinus rhythm. Meanwhile 10% crossed from the rate-control arm to rhythm control, primarily because of worsening heart failure.

Hospitalization was higher in the rhythm group (46% vs 39% in the rate group at one year; p=0.0063), mainly due to hospitalization for AF and bradyarrhythmias (8.5% vs 4.9%, p=0.0074), Roy said. As expected, cardioversions were also much higher in the rhythm-control group (39% vs 8%).

"We could not demonstrate any benefit with the more complex rhythm strategy, so adequate rate control may be sufficient and appropriate in these patients," Roy said.

Chair of the AHA Scientific Sessions program committee, Dr Gordon Tomaselli (Johns Hopkins University, Baltimore, MD), said: "This is an extension of the AFFIRM results in some ways." AFFIRM was reported five years ago and showed that in patients with persistent AF, the strategy of restoring and maintaining sinus rhythm had no clear advantage over the strategy of controlling the ventricular rate and allowing AF to persist.

Falk said that although this is probably the end of attempted cardioversion in systolic heart failure with AF in a "stable patient"—one that fit this trial's criterion and who had adequate rate control both at rest and at low-level exercise—this was not necessarily the case in all patients.

For example, this is almost certainly not the end of this approach in patients in whom the onset of AF is associated with a clear-cut deterioration in clinical status, he noted. "We don't know yet how many patients were screened and deemed ineligible or how many symptomatic patients were not referred for study inclusion. Such information is critical in order to fully place any study in the clearest clinical perspective."

Roy said data on quality of life and costs in AF/CHF, which were also secondary end points, will be presented at a future date

If restoration of sinus rhythm does improve well-being in these patients, there still may be an argument in its favor.

Falk said: "Quality of life was not addressed in this presentation. In AFFIRM, there was no difference in quality of life between those assigned to rhythm control and those assigned to remain in AF [rate control], but in studies assessing quality of life from the perspective of the patient's actual rhythm rather than their assigned group (such as SAFE-T), subjects in sinus rhythm generally had an improved quality of life."

"If restoration of sinus rhythm does improve well-being in these patients, there still may be an argument in its favor. Thus the quality-of-life data, analyzed both by assigned treatment group and by achieved rhythm, will be of great interest," he noted.

Roy said an economic analysis of AF/CHF is also being prepared. A cost analysis from the AFFIRM study, reported in 2004, showed that rhythm control was consistently more expensive than rate control, approximately $5000 more per patient.