The results of a long-awaited study of catheter ablation for atrial fibrillation has doctors wondering whether CABANA is the new ORBITA. The CABANA trial, started nearly ten years ago, is the first randomized controlled trial to measure outcomes for patients with atrial fibrillation undergoing ablation versus drug therapy.
Atrial fibrillation (AF) is a heart condition in which the upper chambers of the heart beat irregularly. AF can cause shortness of breath, fatigue, and nausea in the short term, but can also lead to serious complications such as cardiac arrest, stroke, and death. In a catheter ablation procedure, the doctor uses a catheter to create scars on the heart tissue, which eliminates the erratic electrical signals and “resets” the heart rhythm. This procedure has been used for twenty years, and has grown to a $4.5 billion industry, but without a randomized trial, some cardiologists have wondered whether the benefits of catheter ablation are just placebo.
The CABANA trial sought to test this for the first time, by randomizing more than 2,200 patients with AF to either ablation or drug treatment and looked at cardiovascular outcomes after five years. The results of the CABANA trial were presented at the annual meeting of the Heart Rhythm Society (HRS) last week. The trial found no significant difference between the ablation group and drug group for the primary endpoints (all-cause mortality, serious stroke or bleeding, or cardiac arrest). Does this mean ablation is not an effective treatment?
This is where it gets tricky. For one, ablation was helpful in reducing certain secondary endpoints, such as cardiovascular death and hospitalization and AF recurrence. Another important wrinkle in the study is that there was significant crossover between the two study groups – 101 patients who were assigned to ablation switched to drug therapy (some people could not afford the procedure) and 303 patients who were assigned to the drug study arm switched to ablation.
When looking just at those who received the intervention (known as per protocol analysis), those who received ablation had better outcomes in the primary endpoints. Some electrophysiologists took this as proof that ablation is an effective treatment. But others, such as cardiologist Dr. John Mandrola, pointed out that relying on per protocol analysis is sacrificing the randomization element, and may introduce bias because those patients who choose ablation are likely different from those remained in the drug arm.
Dr. Milton Packer, cardiologist at Baylor Health, concurred.“Why would you design a randomized trial if you are going to analyze it in a way that is inconsistent with randomization? We do not perform or accept ‘per-protocol’ analyses in trials of drug therapy,” he said, quoted in Forbes.
Further, the trial was designed knowing there would be a high level of crossover, and the trial leaders declared from the onset that they would rely on the primary endpoint results from the randomized groups. But when the study results came out negative, the trial leaders at the conference instead presented alternative analyses based on the per protocol results (what Packer refers to as “alternative facts”).
1. This was our research question. This how we tried to answer it. Unfortunately, no benefit.
2. Here’s some confounded data that makes ablation look good (didn’t feel like adjusting anything yet).
3. You guys go out there and keep doing what you want!
— Boback Ziaeian (@boback) May 10, 2018
While we still don’t have proof that ablation works better than drugs at reducing major outcomes, it might be the case that ablation helps some patients for whom initial drug therapy isn’t effective (such as the patients who switched over from the drug arm to the ablation arm). Also, many electrophysiologists piped up on Twitter to say that the primary endpoints were not why they personally offered patients ablation – it was more about improving quality of life and preventing AF recurrence.
The positive effects that doctors observe from ablation might be due to bias or the placebo effect. We won’t know until we do a placebo-controlled trial like ORBITA for ablation. Despite some doctors’ insistence that we shouldn’t do a sham-controlled ablation trial yet, we think it is not just helpful but necessary if we want to keep doing this procedure.
As Dr. Rita Redberg said at the Lown Conference: “People say, ‘It’s unethical to do a sham controlled trial.’ I think it’s unethical not to!”