CABANA Misses Primary Endpoint, but Electrophysiologists See Support for Ablation

BOSTON, MA—(UPDATED) Purists will say the CABANA trial failed to show that catheter ablation improves hard outcomes in patients with A-fib, but electrophysiologists who perform the procedure see it differently.

In the intention-to-treat analysis, patients who underwent ablation did not have a lower rate of a composite of death, disabling stroke, serious bleeding, or cardiac arrest compared with patients treated with medications alone (8.0% vs 9.2%; HR 0.86; 95% CI 0.65-1.15), Douglas Packer, MD (Mayo Clinic, Rochester, MN), reported at the Heart Rhythm Society 2018 Scientific Sessions here.

More than one-quarter of patients in the medication arm crossed over and underwent ablation, however, and it was the “treatment received” and per protocol analyses that drew the most attention from electrophysiologists. In both of those, ablation had a significant advantage over drug therapy for the primary composite outcome. The “treatment received” analysis also showed reductions in all-cause mortality and a composite of death or cardiovascular hospitalization with ablation.

“We do think that ablation is an acceptable treatment strategy for treating atrial fibrillation with low adverse event rates, which surprised us quite a bit, even in the higher-risk patients,” Packer said during his presentation.

Still, not everyone with A-fib should be ablated, he said later during a press conference. “That’s just simply not the case,” he commented. “I think what [the trial] did was it identified a group of patients who are most likely to benefit and a group of patients who are most likely to improve from the standpoint of quality of life and comfort.”

I now have hard data to talk to a patient about it and I think that’s really useful with shared decision-making, which you have to do when you talk to somebody about ablation. Christine Albert

Christine Albert, MD (Brigham and Women’s Hospital, Boston), who served as a moderator at the press conference, said the CABANA findings will have an effect on her practice.

Importantly, she said, the trial showed that ablation was safe. “We knew that there was a pretty high complication rate and we were doing this for the most part [for] symptoms,” she said. “But now we can look at these data and see that there wasn’t really harm.”

There might be groups of patients who benefit more from ablation, and it’ll be important to evaluate subgroups when the results are published, she said. But in the meantime, the study “makes me a little more confident when I have that discussion about offering ablation,” she said. “I now have hard data to talk to a patient about it and I think that’s really useful with shared decision-making, which you have to do when you talk to somebody about ablation.”

She indicated that she was convinced by the “treatment received” and per protocol analyses, saying: “As a physician, I am going to look at the on-treatment [analysis] because those people actually got it.”

Mintu Turakhia, MD (Palo Alto VA Health Care System, CA), agreed: “What the message should not be is that it was a negative trial. It’s important that a trial this complicated not get oversimplified as such. I think the takeaway on the totality of evidence that we have is that ablation works.”

He acknowledged that trials typically succeed or fail based on their performance on primary endpoints, but said this was an unusual case because of the high crossover rate.

“You cannot ignore the intention-to-treat analysis, because in fact that is the gold standard for assessing a trial, but that measure becomes less stable and less robust when you start having crossovers in the study,” he said. “So you can’t look at that in isolation without taking context from the other two as-treated and per protocol analyses.”

That sentiment was shared by other electrophysiologists interviewed by TCTMD, although clinical trialists from other specialties were more circumspect.

“In general, I feel that per protocol analyses are helpful for exploratory purposes,” stroke neurologist Larry Goldstein, MD (University of Kentucky, Lexington), told TCTMD in an email. “Studies need to be primarily analyzed and interpreted based on an intention-to-treat analysis for the primary hypothesis.”

Interventional cardiologist Sunil Rao, MD (Duke University, Durham, NC), called CABANA a neutral trial when considering the intention-to-treat analysis of the primary endpoint, adding in an email to TCTMD, “Everything else is interesting, but exploratory.”

He continued on to say that “there are likely patients who benefit from and prefer ablation, and they should be offered it. The challenge now is figuring out who those patients are.”

Interventional cardiologist C. Michael Gibson, MD (Baim Institute for Clinical Research, Boston, MA), called the secondary endpoint data “provocative” and “hypothesis-generating.”

The as-treated analysis “may validate [electrophysiologists’] enthusiasm for ablation, but we need to learn much more about how the analysis was done,” he told TCTMD via Twitter, noting that such analyses are often more valuable for evaluating safety rather than efficacy. In contrast to an intention-to-treat analysis, “in an as-treated analysis, randomization is not preserved and imbalances in risk factors, demographics, and confounders may occur,” Gibson explained.

Adverse Events Low

CABANA randomized 2,204 patients with new-onset or undertreated A-fib to catheter ablation (with a primary approach of pulmonary vein isolation) or drug therapy that included either rate or rhythm control (87.2% received rhythm control). Anticoagulation was used in both groups.

Crossover was common, however, with 9.2% of patients in the ablation arm not undergoing the procedure and 27.5% of those in the drug therapy arm ultimately undergoing ablation.

In the intention-to-treat analysis, there was no difference between the trial arms in terms of the primary composite endpoint or any of its components. A secondary composite endpoint of all-cause death or cardiovascular hospitalization favored ablation (51.7% vs 58.1%; HR 0.83; 95% CI 0.74-0.93). The procedure also reduced the recurrence of A-fib (HR 0.53; 95% CI 0.46-0.61).

The “treatment received” analysis was even more positive for ablation, showing significant reductions in the primary composite endpoint (7.0% vs 10.9%; HR 0.67; 95% CI 0.50-0.89), all-cause mortality (4.4% vs 7.5%; HR 0.60; 95% CI 0.42-0.86), and death or cardiovascular hospitalization (41.2% vs 74.9%; HR 0.83; 95% CI 0.74-0.94).

Adverse events were low, Packer said. The most common observed in the ablation arm were hematoma associated with catheter insertion (2.3%) and pericardial effusion not requiring intervention (2.2%). Cardiac tamponade with perforation occurred in eight patients (0.8%). On the other side, the most frequent adverse events associated with drug therapy were hyper- or hypothyroidism (1.6%) and major proarrhythmic events (0.8%).

A Change in Approach

Packer said at the press conference that the CABANA results have changed his approach with patients. Over the past 4 or 5 years, he said, there was “real equipoise” between ablation and drug therapy that was conveyed to patients considering their options.

I think it would be inappropriate for someone to look at CABANA and say everybody needs to get ablated. Douglas Packer

Now, “it is much less common for me to say, ‘Let’s just go with the drug and see how it goes,’” Packer said, noting that he would still use that approach when the cause of the patient’s symptoms is unclear. In his discussions with patients, he said he’s more likely to say that ablation is a reasonable option and that the risks seem to be lower than previously thought.

He cautioned, however, against overinterpreting the findings.

“Not all therapies are right for all patients, and so I think you still have to back off just a little bit and think a little bit about it,” Packer said. “I think it would be inappropriate for someone to look at CABANA and say everybody needs to get ablated. I think what it says is . . . that physicians should think about it and that there are now more data . . . that suggest that it’s relevant.”

Turakhia predicted that the trial results would not slow the 15% year-over-year increase in ablations that has been seen in the United States in recent years. Combined with the recently published results of the AATAC and CASTLE-AF trials, CABANA serves as “a signal to continue to proceed,” he said, noting that there were no concerning safety signals in the study.

“The procedure has matured, and it will continue to streamline,” Turakhia said. “We will get better and smarter and faster about how we can do these cases efficiently, and I think it will be commonplace.”

Is It Time for a Sham-Controlled Trial?

The topic of a sham-controlled trial continued to come up in discussions about CABANA, but electrophysiologists interviewed by TCTMD dismissed the idea.

Turakhia said CABANA had difficulty enrolling patients even though there was less evidence of ablation’s benefits when the trial started. Considering the data accumulated at this point, “to expect that you’re going to add clarity to AF ablation with a sham study in 2018 is incredibly naive,” he said.

Moussa Mansour, MD (Massachusetts General Hospital, Boston), said such a trial would be unethical, pointing to evidence from AATAC, CASTLE-AF, and now CABANA that ablation has a positive impact.

And Bruce Lindsay, MD (Cleveland Clinic, OH), echoed that idea. “No one would feel it was ethical,” he said, “to subject people to an invasive procedure like that and then do nothing.

Debating the merits of the design, endpoints, and interpretation is important and should not be taken personally. Sunil Rao

Even if the prospects for a sham-controlled trial seem dim, Rao indicated that having such discussions was healthy for the profession. He said that “debating the merits of the design, endpoints, and interpretation is important and should not be taken personally,” adding that the situation with CABANA reminds him of what happened with the COURAGE trial.

“Both trials compared strategies of a procedure versus medical therapy, both trials challenged existing clinical paradigms, both trials had crossovers, and both trials resulted in some hand-wringing and (hopefully) self-examination,” Rao said.

“It will take some time to digest the results and put them into clinical context, but the results should inform the conversation electrophysiologists have with patients,” he continued. “I expect that many patients will not want to take antiarrhythmics long term and ablation may be reasonable for them. It will be interesting to see if the guidelines change, the number of ablation procedures decrease, the [appropriate use criteria] folks come up with guidance for atrial fibrillation ablation, and the trial results affect payment.”