Stopping Beta-Blockers Did Not Go as Planned for Long-Time Users

LONDON — Heart attack survivors who were expected to be fine after going off their beta-blockers fared worse clinically compared with those who continued them, a randomized trial showed.

After years of taking these relatively inexpensive and accessible drugs, patients assigned to beta-blocker interruption in the French ABYSS trial were at a disadvantage regarding the primary composite endpoint of death, nonfatal myocardial infarction (MI), nonfatal stroke, and hospitalization for cardiovascular reasons up to 4 years later compared with the beta-blocker continuation group, according to Johanne Silvain, MD, PhD, of Sorbonne Université and Pitié-Salpêtrière Hospital in Paris.

These events occurred in 23.8% versus 21.1% of patients in the two groups, respectively (HR 1.16, 95% CI 1.01-1.33), with the withdrawal strategy failing to meet a predetermined threshold of noninferiority. Not a single endpoint component favored beta-blocker interruption over continuation for secondary prevention in this low-risk cohort:

• Death: 4.1% vs 4.0%
• MI: 2.5% vs 2.4%
• Stroke: 1.0% for both
• Hospitalization for cardiovascular causes: 18.9% vs 16.6%

The excess hospitalizations were the reason why the strategy could not reach noninferiority in this trial, Silvain said at the European Society of Cardiology annual congress. What’s more, stopping beta-blockers did not improve quality of life per the European Quality of Life-5 Dimensions questionnaire.

Results of the ABYSS trial were simultaneously published in the New England Journal of Medicine.

The negative results for beta-blocker interruption in this cohort with a median left ventricular ejection fraction (LVEF) of 60% and no recent cardiovascular events may be surprising to some, as other mounting data seemed to back suspicions that long-term beta-blockers have little added benefit in contemporary practice to make up for potential side effects such as depression and fatigue.

This spring, REDUCE-AMI investigators could not prove that beta-blocker initiation reduces death and MI rates in a randomized trial of similar patients with a preserved LVEF. Silvain’s group said the main difference between REDUCE-AMI and ABYSS was the latter’s inclusion of cardiovascular hospitalization as an assessment after beta-blocker interruption.

In any case, U.S. guidelines have already stopped recommending beta-blockers 1 year after an MI unless the patient has low LVEF, hypertension, arrhythmia, or another indication for them.

This was a big change after decades of blanket endorsements for beta-blockers in ST-segment elevation MI (STEMI) and non-STEMI, regardless of LVEF. The old thinking was based on evidence for beta-blockers dating back to the 1970s and 1980s, before the advent of now-standard therapies such as aspirin and statins.

During a press conference, Silvain recommended not stopping chronic beta-blocker treatment — which may be considered relatively low-cost, and well-tolerated if patients have been on it for years already — on the basis of the present safety signal observed in ABYSS.

Tomas Jernberg, MD, PhD, of Danderyd Hospital and the Karolinska Institutet in Stockholm, suggested hitting the brakes on any more guideline updates until more modern beta-blocker trials are reported. “The results of current trials will provide additional data on whether the use of these drugs in this population may be approaching final retirement,” he wrote in an accompanying editorial.

In particular, people with a midrange LVEF in the 40-49% range, a group that has been underrepresented in the trials so far, “may have a benefit from treatment that differs from the benefit in patients with a higher ejection fraction,” he noted.

Ongoing trials like SMART-DECISION and BETAMI-DANBLOCK are expected to report on more patients with mildly reduced LVEF who go off beta-blocker therapy.

ABYSS was conducted at 49 French centers and included 3,698 MI survivors on beta-blocker therapy. All had an LVEF of at least 40% while receiving long-term beta-blocker treatment and had no history of a cardiovascular event in the previous 6 months. Mean age was 63.5, and about 83% were men.

A previous STEMI was recorded in 63%, and 7.8% had more than one previous MI. Over 90% had undergone some sort of coronary revascularization for the index MI. The most frequently prescribed beta-blocker at baseline was bisoprolol (71.5%). Silvain’s team said that the two randomized groups were well balanced in terms of baseline characteristics and medications taken for secondary prevention.

There were a median 2.9 years between last MI and randomization. Median follow-up in ABYSS was 3 years.

In just the first 6 months, the beta-blocker interruption group already showed increases in blood pressure and heart rate. This was maintained during follow-up, Silvain said during the press conference.

Over time, crossovers occurred more frequently in the beta-blocker interruption group (8.6% vs 2.8% of the continuation group).

Silvain reported that the per-protocol analysis was ultimately consistent with interruption not meeting noninferiority criteria (HR 1.06, 59% CI 0.92-1.22).

The study authors acknowledged that the ABYSS results may be of questionable generalizability to centers outside France. Additionally, their unblinded trial design may have affected evaluation of quality of life, they cautioned.

“The open-label design is problematic when an endpoint, such as hospitalization for cardiovascular reasons, drives the results of the trial,” Jernberg warned. “Several of the beta-blockers that were prescribed to patients in the trial lack documentation regarding long-term effects after myocardial infarction. Moreover, although the beta-blocker doses may reflect current practice, they were lower than those used in some trials showing effects in patients with cardiovascular disease.”

Please enable JavaScript to view the

comments powered by Disqus.