Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM)

To compare two standard treatment strategies for atrial fibrillation: ventricular rate control and anticoagulation vs. rhythm control and anticoagulation. BACKGROUND: Atrial fibrillation is an extremely common and increasingly prevalent cardiac arrhythmia, particularly in the elderly, and is an important risk factor for stroke. Management of atrial fibrillation remains controversial, and although antiarrhythmic drugs are widely used for this condition, clinical studies to support their use are meager. Management of atrial fibrillation has at least three components: restoration and maintenance of sinus rhythm; heart rate control when maintenance of sinus rhythm or when cardioversion is not attempted or impossible; and anticoagulation. The first component of management uses antiarrhythmic drugs and the second uses a different group of antiarrhythmic drugs and catheter ablation. The third is anticoagulant therapy for patients in whom normal sinus rhythm cannot be maintained or in whom cardioversion is not attempted. The initiative was developed by staff of the Clinical Trials Branch and the NHLBI Working Group on Atrial Fibrillation which met in Bethesda in April, 1993. The initiative was given concept clearance by the February 1994 National Heart, Lung, and Blood Advisory Council. The Request for Proposals was released in May 1994. DESIGN NARRATIVE: A randomized multicenter trial. The trial enrolled only patients with atrial fibrillation who were at high risk for stroke, that is, over 65 years of age or less than 65 and with one or more other risk factors for stroke such as systemic hypertension, diabetes mellitus, congestive heart failure, transient ischemic attack, prior cerebral vascular accident. High risk patients were treated with the anticoagulant warfarin. Cardioversion (electrical or pharmacologic) might have been attempted before randomization, but if it was unsuccessful, the patient was excluded from further consideration for randomization. Normal sinus rhythm must have persisted for one hour or greater after cardioversion to qualify as successful cardioversion. Patients were randomly assigned to treatment groups which included maintenance of sinus rhythm or heart rate control. Both treatment groups had two steps. In the maintenance of sinus rhythm group, the choice of drugs was left to the primary treating physician, to be chosen from amiodarone, sotalol, propafenone, flecainide, quinidine, moricizine, disopyramide, procainamide, and combinations of these drugs. Atrioventricular nodal blocking drugs were also administered unless contraindicated. The major substudy for AFFIRM randomized initial drug choice among amiodarone, sotalol, and class I drugs. Prior drugs which were ineffective or poorly tolerated were not repeated. There were various drug exclusions depending on the patient's condition. Patients in the maintenance of sinus rhythm group had multiple cardioversions as needed. If there was treatment failure or intolerance after two or more pharmacologic trials, patients were considered for innovative therapy in Step II. In Step II, two maintenance doses of amiodarone were included, a low dose of 100 to 200 mg/day and a normal dose of 300 to 400 mg/day. Each dose of amiodarone was considered to be a single drug trial, so that patients who received treatment with amiodarone at both dosage levels were considered to have had two drug trials. It was not mandatory that Step II therapies be applied in any individual patient. The following innovative Step II therapies were approved for use in the study: (1) ablation of an atrial focus in patients with type I atrial flutter, if it was clinically documented that the atrial flutter led to atrial fibrillation; (2) atrial pacing alone, with or without documented bradycardia; (3) atrial pacing and antiarrhythmic drugs, with either single site or multiple site atrial pacing; and (4) surgical maze or atrial isolation procedures at selected centers. Catheter-based ablative procedures, such as those attempting to mimic the maze procedure were not approved in this study. Implanted atrial cardioverter defibrillators were also not approved. All therapy was periodically reviewed and subject to modification by the Steering Committee with concurrence by the DSMB and the NHLBI. In the event that sinus rhythm was not maintainable by any treatment, patients crossed over to rate control and anticoagulation. The heart rate control arm used heart rate as the therapeutic target, rather than dose of medications. Drug dosage was adjusted to achieve target heart rates. During atrial fibrillation, heart rate was assessed both at rest and during activity at each clinic visit. The pharmacologic therapies approved for use in this arm included: beta blockers, verapamil, diltiazem, digoxin, or combinations of these drugs . When Step I pharmacologic therapies failed after two or more drug trials, the treating physician could select an approved Step II innovative therapy. The two innovative therapies approved for use with the heart rate control arm were: (1) atrioventricular node modification by catheter ablation, with or without placement of a pacemaker, with or without continued drugs to slow atrioventricular node conduction, and (2) total atrioventricular junctional ablation and placement of a pacemaker. The primary endpoint by which the two strategies were compared was total mortality, analyzed by intention-to-treat. Secondary endpoints were composite end points (total mortality, disabling intracranial bleed, stroke, disabling anoxic encephalopathy, cardiac arrest, major noncentral nervous system bleed, cost of therapy, and quality of life. Follow-up was a minimum of two years and an average of 3.5 years. Recruitment and intervention extended from November 1995 through October 1999 with 4,060 patients enrolled by 213 sites.