麻豆果冻传媒影音

CLEVELAND -- Whether vagal nerve stimulation improved outcomes for heart failure patients with reduced ejection fraction (HFrEF) remained unclear after ANTHEM-HFrEF was terminated early, with the pivotal trial not meeting its primary endpoint but suggesting benefit in another for which it was powered.

For the primary endpoint of cardiovascular death or heart failure hospitalizations the trial was originally powered for, the hazard ratio came out at 0.84 (95% CI 0.63-1.13, one-sided P=0.122), Marvin Konstam, MD, of Tufts Medical Center in Boston, reported at the Heart Failure Society of America (HFSA) meeting.

"You had a hazard ratio of 0.84, which is pretty impressive with that background therapy. If I was thinking through this, if that sample size had really gone up to 1,000 or even higher, you resize it, you'd have a positive trial," said session moderator Christopher O'Connor, MD, of the Inova Health System in Falls Church, Virginia.

It was an "incredibly disappointing and heartbreaking result," said HFSA late-breaking trial session discussant Nancy Sweitzer, MD, PhD, of Washington University in St. Louis.

The mechanism seems promising, she said, as the vagus nerve has a big role in parasympathetic tone and heart rate control. And the trial protocol ensured an effective "dose" by using ambulatory uptitration of dose until the heart rate variation pattern suggested autonomic engagement, she added.

The trial enrolled patients with persistent New York Heart Association class II or III symptoms despite guideline-directed medical therapy, with prior heart failure hospitalization, a left ventricular ejection fraction ≤35%, NT-proBNP of at least 800 pg/mL, and a 6-minute walk test distance of 150-450 meters. Participants were randomized 2:1 to vagal nerve stimulation with guideline-directed medical therapy or medication alone.

After the first 400 patients, the adaptive design enrolled more in blocks of 100 with prespecified futility checks to justify continued enrollment. After the 500 patients, there was also an analysis of whether the predicted probability of success was 95% or greater, in which case enrollment would stop and the existing patients would be followed out to 16 months.

However, that design might have been the trial's downfall, Sweitzer suggested.

Although the data safety monitoring committee recommended study continuation without change after the interim analysis at 500 patients, the sponsor announced it was terminating the trial.

"The company knew they hadn't had a home run yet at 500 patients; they may have at 1,000, we don't know," Sweitzer noted. "The data are completely insufficient. I think this was a clever design, but you might have shot yourself in the foot."

It was not study futility, safety concerns, or device malfunction that led to early termination, but rather "financial" reasons, according to Konstam. Serious device- or implant-related adverse events occurred in 3.5% of patients at 90 days, with a 0.3% rate of vagus nerve stimulation-related syncope or ventricular arrhythmia.

Before the database was locked, his group worked with the FDA to find an endpoint suitable for the 532 patients they had enrolled, short of the up to 1,000 the trial was designed for. They settled on a supplemental endpoint of a win ratio with a hierarchical endpoint encompassing time to cardiovascular death, the rate of overall heart failure hospitalization events, and Kansas City Cardiomyopathy Questionnaire overall summary score with a 5-point threshold for win or loss.

By that analysis, vagal nerve stimulation "won" in 85.4% of pairings for an overall win ratio of 1.25 (1.00-1.55, one-sided P=0.023).

"Given the positive findings from the win ratio, with contributions of all three components, and favorable trends with reduced sample size in the primary endpoint and components justify continued exploration of the potential value of this therapy," Konstam concluded.

"I think it's the best we could do under the circumstances. The question is, is it enough to keep going? In my estimation, it is. There'd be reason to try again and see what happens."

Comment