番茄社区

LONDON -- The surprising mortality risk from central sleep apnea treatment in systolic heart failure (HF) appears real and more from sudden death than from chronic worsening, the SERVE-HF trial researchers suggested.

The initial trial results rocked the respiratory and cardiology fields with its finding of not only no benefit (hazard ratio 1.13, P=0.10 for the composite of all-cause death, life-saving cardiovascular intervention, or unplanned HF hospitalization), but also significant mortality risk from adaptive servo-ventilation when were released in May.

All-cause mortality was elevated 28% and cardiovascular mortality 34% with the intervention compared with controls, prompting a label change for a in heart failure with ejection fraction of 45% or less.

Now reported here at the European Society of Cardiology (ESC) meeting and simultaneously online in the New England Journal of Medicine are clarifying that this effect was "unlikely to be due to chance."

"We're looking at sudden death in these patients," of Imperial College London, said at an ESC press conference where he presented the full results.

"There's no difference in progressive heart failure mortality, no difference in heart failure hospitalization," he said. "If you look at patients who died without preceding hospitalizations, which is presumably sudden, the hazard ratio is 2.5 and the curves separate almost immediately and during the day and night. So I think we have to take this seriously."

Mechanisms

"We have got two main theories," Cowie said. "It does seem in subgroup analyses the more you have the worse you do, so it does seem there's something about Cheyne-Stokes we don't understand."

"Maybe it's , as suggested almost in jest a few years ago , or maybe it's positive airway pressure," he added.

It's possible that slowing, stopping, and then speeding up pattern of breathing during sleep, which is common in reduced ejection fraction HF (25% to 40%) but rare outside it, actually is beneficial in HF by "resting respiratory muscles, for example, protecting from alkalosis, and also allowing intrinsic peak in terms of keeping lung volume up and increasing oxygenation," Cowie speculated.

Another theory is applying positive air pressure to a poor ventricle is harmful, he added, "although there's not very much evidence that it is a toxic therapy." Adaptive servo-ventilation noninvasively delivers servo-controlled inspiratory pressure support on top of expiratory positive airway pressure.

"We're drilling down, and I think it's going to be more a sudden death, arrhythmia, autonomic nervous system type of thing," Cowie added.

More information on timing of death and whether it occurred during adaptive servo-ventilation might provide some clues on mechanism, , of Ohio State University Wexner Medical Center in Columbus, and , of University of Pennsylvania in Philadelphia, suggested in an accompanying editorial.

"This is an important question to be addressed," they wrote. A new technique to abolish Cheyne-Stokes respiration that is not based on positive pressure therapy (i.e., phrenic-nerve stimulation) has been developed and is in clinical trial. If Cheyne-Stokes respiration is beneficial, then this strategy may also lead to increased cardiovascular mortality. In contrast, if positive airway pressure is the offending treatment, then this new therapy might provide a benefit."

Generalizability

When asked whether adaptive-servo-ventilation should be avoided in other central sleep apneas, Cowie cautioned against extrapolating.

The kind of low ejection fraction HF patients who were included in the 1,325-patient trial were very sick, with a physiology "very different" from those with preserved ejection fraction, he said.

And obstructive sleep apnea is a "totally different thing" compared with central sleep apnea, and the findings shouldn't be generalized there either, he added.

The trial included only patients with an apnea-hypopnea index (AHI) of 15 or more apnea or hypopnea events per hour and a predominance of central events.

"Some physicians are going to have to change their practice because they went ahead of the trial results," he said. "Don't extrapolate from the study results."

, of McMaster University in Hamilton, Ontario, and a moderator at the ESC briefing, agreed that the findings were "very worrisome" but also emphasized not extrapolating the results to obstructive or non-systolic HF.

"Until we have proof of benefit, people should not even use it," he said, although supporting ongoing trials given that there were a moderate number of events in SERVE-HF and that mortality risk was an unexpected, secondary finding in a negative trial.

Another limitation was that relatively few women (10%) were included in the trial, said Mary Norine Walsh, MD, of St. Vincent Heart Center in Indianapolis and vice-president of the American College of Cardiology. Subgroup analyses tended to show greater risk among women with the intervention.

"One of the most important takeaways from this trial is the fact that we have to look at outcomes, not just surrogate endpoints, because look what was uncovered," she told 番茄社区.

Comment