Left atrial appendage occlusion vs standard of care in high stroke risk atrial fibrillation patients ineligible for anticoagulation: COMPARE-LAAO

TL;DR

A clinical trial comparing left atrial appendage occlusion to standard care in atrial fibrillation patients who cannot take blood thinners was stopped early due to slow enrollment.

Contribution

The paper highlights the challenges in conducting randomized trials for LAAO in patients ineligible for anticoagulation, suggesting global feasibility issues.

Findings

The trial was terminated early due to slow patient enrollment and loss of clinical equipoise.

There were fewer stroke events in the LAAO group compared to standard care, but results were not statistically significant.

The study suggests that conducting such trials globally may be infeasible, impacting potential reimbursement decisions.

Abstract

The left atrial appendage is the dominant source of cardioembolic stroke in patients with atrial fibrillation (AF). Contemporary guidelines recommend considering left atrial appendage occlusion (LAAO) in AF patients contraindicated to oral anticoagulation therapy (OAC), but randomized controlled trial (RCT) data for this subpopulation are lacking.

COMPARE LAAO was designed as an event-driven, multicenter, prospective, randomized, open, blinded endpoint (PROBE) trial that randomized AF patients with an increased thromboembolic risk and a contraindication to OAC 2:1 to LAAO or standard-of-care (SOC). The co-primary endpoints comprised 1) time to first occurrence of ischemic/hemorrhagic/undetermined stroke and 2) time to first occurrence of all-cause stroke/TIA/SE. The trial aimed to enroll 609 patients.

After randomization of 69 patients, the trial was terminated prematurely by the sponsor due to a slow inclusion rate. Results are discussed briefly without formal statistical testing. All-cause stroke occurred in 7/48 and 2/21 patients randomized to LAAO and SOC, respectively. According to the as-treated principle, all-cause stroke occurred in 5/41 and 4/28 patients treated with LAAO and SOC. The composite of all-cause stroke/TIA/SE occurred in 10/48 and 4/21 patients randomized to and 8/41 and 6/28 patients treated with LAAO and SOC.

Insufficient statistical power of COMPARE LAAO impedes drawing any conclusions. Among other factors, the loss of perceived clinical equipoise among physicians proved problematic for successful trial completion. Conducting an RCT on LAAO vs SOC in OAC-ineligible patients appears infeasible globally, which threatens to preclude reimbursement in the Netherlands for these patients that have no proven alternative.

The online version of this article (10.1007/s12471-025-02005-7) contains supplementary material, which is available to authorized users.

Figures