Aug. 22, 2018
The number of individuals with atrial fibrillation (AF) — both in the United States and worldwide — continues to grow. Contributing factors include the aging of the population and the prevalence of known risk factors for the development of AF.
AF is problematic not only because of the symptoms experienced by many patients, but because of the risk of thromboembolic events such as stroke, the development of tachycardia-induced cardiomyopathy secondary to inadequate rate control and uncertainty regarding optimal rhythm management if that is the management choice.
The development of pulmonary vein isolation (PVI) procedures provide an additional tool in the pursuit of sinus rhythm, and the use of these procedures has become widespread over the past two decades despite lack of evidence-based validation.
The recently completed Catheter Ablation Versus Anti-Arrhythmic Drug Therapy for Atrial Fibrillation (CABANA) trial was designed to help answer questions about the relative efficacy of ablative versus drug therapy in decreasing the primary endpoint, which is the composite of death, disabling stroke, serious bleeding or cardiac arrest in individuals with AF. Secondary endpoints included all-cause mortality; mortality or cardiovascular hospitalization; mortality, stroke or hospitalization for heart failure or an ischemic event; cardiovascular death; freedom from recurrent AF; and composite adverse events. Additionally, medical costs and quality of life in each arm were evaluated.
For the CABANA trial, 2,204 patients were enrolled at 126 study sites in the U.S., Australia, Canada, China, Czech Republic, Germany, Italy, South Korea, Russian Federation and the United Kingdom. CABANA aimed to study individuals over the age of 65 or less than age 65 with one or more risk factors for stroke, two or more episodes of paroxysmal atrial fibrillation (PAF) or one episode of persistent AF in the prior six months, and who are suitable for drug- (rhythm or rate control) or catheter-based treatment.
Patients were randomized in a 1-to-1 fashion into PVI (1,108 patients) or drug therapy (1,096 patients). The specific pharmacologic agents utilized were left to the discretion of the treating provider, although providers were encouraged to follow ACC/ESC/AHA treatment guidelines.
The mean patient age was 67.5 years, and 37 percent were female. The mean duration of follow-up was 48 months. All patients were to receive anticoagulation based on current guidelines.
Douglas L. Packer, M.D., former director of Heart Rhythm Services at Mayo Clinic in Rochester, Minnesota, and past president of the Heart Rhythm Society, was the principal investigator. Dr. Packer presented the trial results at the Heart Rhythm Society Annual Scientific Sessions held in Boston in May 2018:
Primary endpoint based on ITT analysis
The primary endpoint (the composite of death, disabling stroke, serious bleeding or cardiac arrest) based on ITT analysis at 48 months for PVI versus drug therapy was 8 percent versus 9.2 percent (HR 0.86, 95 percent CI 0.65-1.15; P = 0.303 NS).
Recurrence of atrial fibrillation reduced
AF recurrence was reduced by 47 percent based on ITT analysis at 48 months for PVI versus drug therapy (HR 0.53, 95 percent CI 0.46-0.61; P < 0.0001).
Primary and secondary outcomes by treatment
The primary endpoint based on actual treatment received for PVI versus drug therapy was 7.0 percent versus 10.9 percent (HR 0.67, 95 percent CI 0.50-0.89; P = 0.006).
- The primary endpoint (the composite of death, disabling stroke, serious bleeding or cardiac arrest) based on intention-to-treat (ITT) analysis at 48 months for PVI vs. drug therapy was 8 percent vs. 9.2 percent (HR 0.86, 95 percent CI 0.65-1.15; P = 0.303 NS).
- However, 9.2 percent of patients randomized to PVI did not undergo ablation, while 27.5 percent of patients randomized to drug therapy actually crossed over to PVI. By ITT analyses, the endpoint of mortality or cardiovascular hospitalization showed a 17 percent reduction, and the recurrence of AF was reduced by 47 percent.
- The primary endpoint based on actual-treatment-received analysis for PVI vs. drug therapy was 7.0 percent vs. 10.9 percent (HR 0.67, 95 percent CI 0.50-0.89; P = 0.006).
Dr. Packer notes that CABANA is a complicated study, so it is important to avoid over-interpretation of the CABANA data. ITT analysis preserves the randomization and if appropriately designed and powered will factor out selection bias; it is generally regarded as the gold standard measure in clinical trials. However, where there are a significant number of crossovers, or patients don't receive the randomized therapy, ITT analysis may fail to clearly establish anticipated outcome. In certain interventional trials, on-treatment analysis may better reflect actual treatment outcomes.
"You can't benefit from a therapy if you don't receive the therapy," said Dr. Packer. The data indicated a 33 percent reduction in the primary endpoint and a 40 percent reduction in mortality in patients who actually received ablation compared to those individuals who received drug treatment. Furthermore, even in the ITT analysis, the two secondary endpoints of total mortality or cardiovascular hospitalization and recurrent atrial fibrillation were significantly better in the ablation arm.
Dr. Packer also noted that by design the drug treatment arm included a heterogeneous group of rate control or anti-arrhythmic drugs selected by the patient's primary cardiologist. It is unclear whether or not the choice of treatment drug affected outcomes or adverse events.
Dr. Packer pointed out during his presentation that one of the findings of the trial is that ablation is a safe option for individuals with symptomatic atrial fibrillation. "There was a low rate of adverse events, even in the higher-risk patients who underwent ablation," said Dr. Packer. The most common findings were vascular injury at catheter insertion sites (2.3 percent) and cardiac perforation (0.8 percent) in the PVI arm, and thyroid abnormalities (1.6 percent) and proarrhythmic events (0.8 percent) in the drug therapy arm.
In the CABANA trial, ablation did not produce a significant reduction in either the primary endpoint or the secondary endpoint of all-cause mortality, according to an ITT analysis. The results were affected by crossovers in both directions and lower-than-expected event rates. Ablation did significantly reduce the secondary endpoint of mortality or cardiovascular hospitalization by 17 percent compared to drug therapy. There also was a significant 47 percent reduction in recurrent AF with ablation compared to drug therapy.
As such, CABANA demonstrated that ablation is an acceptable treatment strategy with low procedural risks, even in treating AF in patients with higher underlying risk. The challenge is to identify those individuals most likely to benefit from ablation. For the most part, these will be the patients with symptomatic AF. "Further subgroup analysis of trial data will provide information about the economics of each approach and quality-of-life measures, which will help us better identify those individuals most likely to benefit from ablative therapy," says Dr. Packer.