CHICAGO – Rate and rhythm control proved equally effective for treatment of new-onset post–cardiac surgery atrial fibrillation in a randomized trial that was far and away the largest ever to examine the best way to address this common and costly arrhythmia, Dr. A. Marc Gillinov said at the annual meeting of the American College of Cardiology.
Thus, either strategy is acceptable. That being said, rate control gets the edge as the initial treatment strategy because it avoids the considerable toxicities accompanying amiodarone for rhythm control, most of which arise only after patients have been discharged from the hospital. In contrast, when rate control doesn’t work, it becomes evident while the patient is still in the hospital, according to Dr. Gillinov, a cardiothoracic surgeon at the Cleveland Clinic .
Dr. Marc Gillinov
Atrial fibrillation (AF) is the most common complication of cardiac surgery, with an incidence variously reported at 20%-50%. It results in lengthier hospital stays, greater cost of care, and increased risks of mortality, stroke, heart failure, and infection. Postoperative AF adds an estimated $1 billion per year to health care costs in the United States.
While current ACC/AHA/Heart Rhythm Society joint guidelines recommend rate control with a beta-blocker as first-line therapy for patients with this postoperative complication, with a class I, level-of-evidence A rating, upon closer inspection the evidence cited mainly involves extrapolation from studies looking at how to prevent postoperative AF. Because no persuasive evidence existed as to how best to treat this common and economically and medically costly condition, Dr. Gillinov and his coinvestigators in the National Institutes of Health–funded Cardiothoracic Surgical Trials Network carried out a randomized trial 10-fold larger than anything prior.
The 23-site study included 2,109 patients enrolled prior to cardiac surgery, of whom 40% underwent isolated coronary artery bypass grafting (CABG) while the other 60% had valve surgery, either alone or with CABG. These proportions reflect current cardiac surgery treatment patterns nationally. Overall, 33% of the cardiac surgery patients experienced postoperative AF. The incidence was 28% in patients who underwent isolated CABG but rose with increasing surgical complexity to nearly 50% in patients who had combined CABG and valve operations. The average time to onset of postoperative AF was 2.4 days.
A total of 523 patients with postoperative AF were randomized to rate or rhythm control. Rate control most often entailed use of a beta-blocker, while amiodarone was prescribed for rhythm control.
The primary endpoint in the trial was a measure of health care resource utilization: total days in hospital during a 60-day period starting from the time of randomization. This endpoint was a draw: a median of 5.1 days with rate control and 5.0 days with rhythm control.
At hospital discharge, 89.9% of patients in the rate control group and 93.5% in the rhythm control group had a stable heart rhythm without AF. From discharge to 60 days, 84.2% of patients in the rate control group and a similar 86.9% of the rhythm control group remained free of AF.
Rates of serious adverse events were similar in the two groups: 24.8 per 100 patient-months in the rate control arm and 26.4 per 100 patient-months in the rhythm control arm. Three patients in the rate control arm died during the 60-day study period, and two died in the rhythm control group.
Of note, roughly one-quarter of patients in each study arm crossed over to the other arm. In the rate control group, this was typically due to drug ineffectiveness, while in the rhythm control arm the switch was most often made in response to amiodarone side effects.
Roughly 43% of patients in each group were placed on anticoagulation with warfarin for 60 days according to study protocol, which called for such action if a patient remained in AF 48 hours after randomization.
There were five strokes, one case of transient ischemic attack, and four noncerebral thromboembolisms. Also, 21 bleeding events occurred, 17 of which were classified as serious; 90% of the bleeding events happened in patients on warfarin.
“I found the results very striking and very reassuring,” said discussant Hugh G. Calkins. “To me, the clinical message is clearly that rate control is the preference.”
It was troubling, however, to see that 10 thromboembolic events occurred in 523 patients over the course of just 60 days. “Should we be anticoagulating these postsurgical atrial fibrillation patients a lot more frequently?” asked Dr. Calkins, professor of medicine and of pediatrics and director of the cardiac arrhythmia service at Johns Hopkins University, Baltimore.
Dr. Gillinov replied that he and his colleagues in the Cardiothoracic Surgical Trials Network consider that to be the key remaining question regarding postoperative AF. They are now planning a clinical trial aimed at finding the optimal balance between stroke protection via anticoagulation and bleeding risk.