鶹ý

CABANA Misses Primary But May Spur Afib Ablation Anyway

— Per protocol, ITT analyses differed on advantage for hard clinical endpoints

MedpageToday

This article is a collaboration between 鶹ý and:

BOSTON -- Does atrial fibrillation ablation do more than improve symptoms and quality of life compared with medication alone? The landmark CABANA trial answered that question, with most electrophysiologists taking it as an affirmative.

The primary intent-to-treat analysis did not show a significant impact on the composite of death, disabling stroke, serious bleeding, or cardiac arrest (HR 0.86, P=0.303) or on all-cause mortality alone (HR 0.85, P=0.377) compared with drug therapy, Douglas Packer, MD, of the Mayo Clinic in Rochester, Minnesota, reported here at the Heart Rhythm Society meeting.

However, a per-protocol analysis censoring the 9.2% of ablation patients who didn't get the procedure and the 27.5% of the drug group that crossed over to ablation showed substantial advantages to the procedure:

  • 33% relative reduction in the primary composite endpoint (7.0% vs 10.9%, HR 0.67, P=0.006)
  • 40% relative reduction in all-cause mortality (4.4% vs 7.5%, HR 0.60, P=0.005)

Both analyses showed a significant 17% relative reduction in death or cardiovascular hospitalization (HR 0.83, P=0.002).

Clinical Implications

As to whether they would call the trial negative based on its failure to meet the primary intent-to-treat endpoint or positive based on the per protocol analysis, electrophysiologists contacted by 鶹ý at the meeting uniformly gave a tempered version of the latter.

"My patients want to know, 'If I receive the following therapy, what's going to happen, Prystowsky?'" said study discussant Eric Prystowsky, MD, of St. Vincent Hospital and Health Center in Indianapolis. "'Am I going to be better, worse, am I going to have side effects?' They're not really interested in the 30% of people in a trial that crossed over and never got it."

While advocating not creeping beyond the trial criteria and taking a measured, close look at the full data once published -- hinting that he reviewed the investigators' paper for an undisclosed journal -- Prystowsky suggested the results are sufficient to show the two approaches to Afib management are at least equivalent.

"At the very least, remove the dotted line for ablation as first line in the guidelines," he urged.

Bruce Lindsay, MD, of the Cleveland Clinic, agreed in commenting on the trial: "In terms of testing the hypothesis, really what I would want to focus on is ... the people who actually got the treatment they were supposed to get."

"I've been part of a group that writes guidelines for Afib, and what we say is that the major reason to do ablation is to improve quality of life," he told 鶹ý. Now with CABANA, "we will be able to better advise patients based on our interpretation of the data from the trial. I think it's a major contribution to our literature."

Lindsay praised CABANA as a "very inclusive" study, such that its patients were representative of a large number of those who undergo ablation.

The trial included 2,204 patients ages 65 and older (or younger with a least one stroke risk factor) treated in 10 countries for new onset or undertreated paroxysmal, persistent, or long-standing persistent Afib who "warrant therapy" and were eligible for ablation and two or more rhythm or rate control drugs. Participants were randomized to ablation, primarily focused on pulmonary vein isolation, or to drug therapy alone with rate or rhythm control medications. Both groups received anticoagulation.

While CABANA started out as a mortality trial, the primary endpoint switched to a composite at the suggestion of the Data Safety Monitoring Board due to slower-than-desired enrollment and event rates.

It's probably as definitive for hard outcomes as it's going to get, with even larger randomized trials unlikely, session co-chair Fred Kusumoto, MD, of the Mayo Clinic in Jacksonville, Florida, told 鶹ý. He called the findings suggestive, particularly for certain subgroups, but not necessarily disappointing.

"Obviously, you'd like that original intention-to-treat analysis to be positive," he said, "but when you look at the on-treatment analysis it's pretty impressive. I don't think any of us want to get too ahead of this and suggest ablation for all. I don't think anyone should be either too excited or too disappointed with this analysis."

However, Milton Packer, MD, a heart failure specialist who sharply criticized the previous CASTLE-AF trial in a blog for 鶹ý, again had strong words for electrophysiologists, noting that intent-to-treat analyses are the standard in cardiology for a reason.

"The EP community is engaging in a classic example of self-deception," he commented. "They should look at the evidence objectively without regard for self-interest. The CABANA trial missed not only on its original primary endpoint but also on its new primary endpoint.

"Therefore, the trial failed to yield any reliable evidence that ablation is better than no ablation with respect to important clinical outcomes," he continued. "If the EP community wants to violate the rules of unbiased clinical trial design and analysis, they can certainly do so. But by doing so, they will be giving up their scientific credibility."

Sanjay Kaul, MD, of Cedars-Sinai Medical Center in Los Angeles, also argued against the intent-to-treat analyses driving opinion: “These are appropriate for sensitivity analyses to demonstrate robustness of the primary outcomes using the ITT framework, but not actionable for regulatory or clinical decision making on their own. I did not find in the whether per protocol analysis was prespecified, which makes it even more egregious.”

Douglas Packer agreed that it was necessary to present the intent-to-treat data. "But when you have this number of crossovers... on-treatment analyses are required," he told attendees.

On the other hand, Kaul pointed out, “The sample size estimation made allowance in the power projections for up to 25 to 30% of patients randomized to the drug arm to cross over to ablation at some point during follow-up. So, the 27% crossover in the drug arm did not have a material impact on the power of the trial, another fallacious criticism heard today from the EP community!”

Subgroups

Even “in a classical sense,” there were some suggestions of benefit, Douglas Packer argued. In the Forest plot for the intent-to-treat primary endpoint analysis, those under age 65 did significantly better with ablation, as did those in minority groups, and a trend for patients with baseline class II or higher heart failure or cardiomyopathy.

In the same subgroup analyses by treatment actually received, these advantages were even greater, whereas the trend went further in the wrong direction for patients 75 and older (HR 1.54). Those with a history of congestive heart failure had a significant 49% relative risk reduction with ablation and those with New York Heart Association class II or worse heart failure had a significant 41% relative advantage with the procedure.

In the trial, 15% of patients had heart failure, a group for which the much smaller CASTLE-AF trial showed lower risk of death from any cause or hospitalization for worsening heart failure if treated with ablation versus medication.

Douglas Packer suggested at the press conference that the findings were most likely to influence practice for patients with heart failure and underlying risk. While CASTLE-AF said it with 300 patients, CABANA reinforced the message: "Turns out 2,200 patients say basically the same thing."

He said that in his practice now, "it is much less common for me to say 'Let's just go with the drug and see how it goes.'"

Christine Albert, MD, MPH, of Brigham and Women's Hospital in Boston, also a session moderator, agreed that the findings will inform how she selectively speaks with patients for ablation and the confidence she has when doing so. Key to that was the "low harm, we didn't know that," she said, citing especially the incredibly low stroke rate.

Only three transient ischemic attacks occurred out of 1,006 ablation-treated patients (0.3%). There were a couple of vascular events and eight cases of tamponade, but no atrial esophageal fistulas.

Disclosures

The trial was funded by the NIH, St. Jude Medical, Biosense Webster, Johnson & Johnson, Medtronic, and Boston Scientific.

Packer disclosed relationships with Abbott, Abiomed, Aperture Diagnostics, Biosense Webster, Boston Scientific, CardioFocus, CardioInsight Technologies, Endosense, Hansen Medical, Johnson & Johnson, Mediasphere Medical, Medtronic, Robertson Foundation, SIEMENS, Spectrum Dynamics, St. Jude Medical, and Thermedical, with royalty income from Wiley-Blackwell, Blackwell Publishing and St. Jude Medical.

Lindsay disclosed an unrelated strategic advisory board role for Medtronic.

Primary Source

Heart Rhythm Society meeting

Packer DL, et al "Catheter ablation vs antiarrhythmic drug therapy for atrial fibrillation: The results Of the Cabana multicenter international randomized clinical trial"