Left atrial appendage occlusion performed at the time of other heart surgery reduces the risk for stroke by about one third in high-risk patients with atrial fibrillation (AF), according to results of the Left Atrial Appendage Occlusion Study III (LAAOS III).
At 3.8 years’ follow-up, the primary endpoint of ischemic stroke or systemic embolism occurred in 4.8% of patients randomly assigned to LAAO and 7.0% of those with no occlusion. This translated into a 33% relative risk reduction (hazard ratio [HR], 0.67; 95% CI, 0.53 – 0.85; P = .001).
In a landmark analysis, the effect was present early on but was more pronounced after the first 30 days, reducing the relative risk by 42% (HR, 0.58; 95% CI, 0.42 – 0.80), the researchers report.
The reduction in ongoing stroke risk was on top of oral anticoagulation (OAC) and consistent across all subgroups, Richard Whitlock, MD, PhD, professor of surgery, McMaster University, Hamilton, Ontario, Canada, reported in a late-breaking trial session May 15 at the American College of Cardiology (ACC) 2021 Scientific Session.
The procedure was safe and added, on average, just 6 minutes to cardiopulmonary bypass time, according to the results, simultaneously published in the New England Journal of Medicine.
“Any patient who comes to the operating room who fits the profile of a LAAOS III patient — so has atrial fibrillation and an elevated stroke risk based on their CHA2DS2-VASc score — the appendage should come off,” he told theheart.org | Medscape Cardiology.
Commenting during the formal discussion, panelist Michael J. Mack, MD, Baylor Health Care System, Houston, Texas, said, “This is potentially a game-changing, practice-changing study” but asked if there are any patients who shouldn’t undergo LAAO, such as those with heart failure (HF).
Whitlock said about 10% to 15% of patients coming for heart surgery have a history of AF and “as surgeons, you do need to individualize therapy. If you have a very frail patient, have concerns about tissue quality, you really need to think about how you would occlude the left atrial appendage or if you would occlude.”
Reassuringly, he noted, the data show no increase in HF hospitalizations and a beneficial effect on stroke among patients with HF and those with low ejection fractions, below 50%.
Observational data on surgical occlusion have been inconsistent, and current guidelines offer a weak recommendation in patients with AF who have a contraindication to long-term anticoagulation. This is the first study to definitively prove that ischemic stroke is reduced by managing the left atrial appendage, he told theheart.org | Medscape Cardiology.
“The previous percutaneous trials failed to demonstrate that; they demonstrated noninferiority but it was driven primarily by the avoidance of hemorrhagic events or strokes through taking patients off oral anticoagulation,” he said.
The results should translate into a class I guideline recommendation, he added. “This opens up a new paradigm of treatment for atrial fibrillation and stroke prevention in that it is really the first study that has looked at the additive effects of managing the left atrial appendage in addition to oral anticoagulation, and it’s protective on top of oral anticoagulation. That is a paradigm shift.”
In an accompanying editorial, Richard L. Page, MD, University of Vermont in Burlington, said the trial provides no insight on the possible benefit of surgical occlusion in patients unable to receive anticoagulation or with a lower CHA2DS2-VASc score, but he agreed a class I recommendation is likely for the population studied.
“I hope and anticipate that the results of this paper will strengthen the guideline indications for surgical left atrial appendage occlusion and will increase the number of cardiac surgeons who routinely perform this add-on procedure,” he said. “While many already perform this procedure, cardiac surgeons should now feel more comfortable that surgical left atrial appendage occlusion is indicated and supported by high-quality randomized data.”
Unfortunately, LAAO III does not answer the question of whether patients can come off anticoagulation, but it does show surgical occlusion provides added protection from strokes, which can be huge with atrial fibrillation, Whitlock said.
“I spoke with a patient today who is an active 66-year-old individual on a DOAC, and his stroke risk has been further reduced by 30% to 40%, so he was ecstatic to hear the results,” Whitlock said. “I think it’s peace of mind.”
Global, Nonindustry Effort
LAAOS III investigators at 105 centers in 27 countries enrolled 4811 patients undergoing cardiac surgery (mean age, 71 years; 68% male) who had a CHA2DS2-VASc score of 2 or more.
In all, 4770 were randomly assigned to no LAAO or occlusion via the preferred technique of amputation with suture closure of the stump as well as stapler occlusion, or epicardial device closure with the AtriClip (AtriCure) or TigerPaw (Maquet Medical). The treating team, researchers, and patients were blinded to assignment.
Patients were followed every 6 months with a validated stroke questionnaire. The trial was stopped early by the data safety monitoring board after the second interim analysis.
The mean CHA2DS2-VASc score was 4.2, one-third of patients had permanent AF, 9% had a history of stroke, and more than two-thirds underwent a valve procedure, which makes LAAOS III unique, as many previous AF trials excluded valvular AF, Whitlock pointed out.
Operative outcomes in the LAAO and no LAAO groups were as follows:
Bypass time: mean, 119 minutes vs 113 minutes
Cross-clamp time: mean, 86 minutes vs 82 minutes
Chest tube output: median, 520 mL vs 500 mL
Reoperation for bleeding: both, 4.0%
Prolonged hospitalization due to HF: 5 vs 14 events
30-day mortality: 3.7% vs 4.0%.
The primary safety outcome of HF hospitalization at 3.8 years occurred in 7.7% of patients with LAAO and 6.8% without occlusion (HR, 1.13; 95% CI, 0.92 – 1.40), despite concerns that taking off the appendage could worsen HF risk by impairing renal clearance of salt and water.
“There’s observational data on either side of the fence, so it was an important endpoint that people were concerned about,” Whitlock told Medscape. “We had a data collection firm dedicated to admission for heart failure to really tease that out and, in the end, we saw no adverse effect.”
Although rates of ischemic stroke at 3.8 years were lower with LAAO than without (4.2% vs 6.6%; HR, 0.62; 95% CI, 0.48 – 0.80), there was no difference in systemic embolism (0.3% for both) or death (22.6% vs 22.5%).
In LAAOS III, less than 2% of the deaths were attributed to stroke, which is consistent with large stroke registries, Whitlock said. “Stroke is not what causes people with atrial fibrillation to die; it’s actually the progression on to heart failure.”
The positive effect on stroke was consistent across all subgroups, including sex, age, rheumatic heart disease, type of OAC at baseline, CHA2DS2-VASc score (≤4 vs >4), type of surgery, history of heart failure or hypertension, and prior stroke/transient ischemic attack/systemic embolism.
Panelist Anne B. Curtis, MD, University of Buffalo, New York, expressed surprise that about half of patients at baseline were not receiving anticoagulation and questioned whether event rates varied among those who did and didn’t stay on OAC.
Whitlock noted that OAC is often underused in AF and that analyses showed the effects were consistent whether patients were on or off anticoagulants.
Panelist Valentin Fuster, MD, PhD, Mount Sinai Hospital, New York City, said LAAOS III is a “fantastic study, particularly taking into account 27 countries, this is not easy.”
He suggested, however, that the 7% stroke rate in the control group is “rather high” when so many patients are receiving OAC. Whitlock replied that the annual event rate in trials such as RE-LY and AVERROES among similar patients was about 1.5% to 2.5%, “so I think our results are consistent with other trials.”
The study was sponsored by the Population Health Research Institute, McMaster University. Whitlock and Fuster reported no relevant disclosures. Curtis reported consultant fees/honoraria from Abbott, Janssen, Medtronic, Milestone Pharmaceuticals, and Sanofi Aventis and data safety monitoring board participation for Medtronic.
American College of Cardiology (ACC) 2021 Scientific Session. Abstract 402-12. Presented May 15, 2021.