Heart HealthResearch PaperPaywall

Left Atrial Appendage Closure vs Anticoagulation: Why Pooled Trial Data Can Mislead

A JACC evidence review argues the major LAAC trials asked different questions, so pooling them misleads. Anticoagulation stays first-line for most AF patients.

Wednesday, November 18, 2026 1 view
Published in J Am Coll Cardiol
cardiac catheterization lab with interventional cardiologists performing a procedure beside a fluoroscopy monitor showing the heart

Summary

Atrial fibrillation raises stroke risk, and most patients take long-term blood thinners. Left atrial appendage closure (LAAC) is a device procedure that seals off a small pouch in the heart where clots tend to form. It has been tested against anticoagulants in randomized trials enrolling more than 7,000 people. This JACC Evidence Review argues that the newest trials (OPTION, CLOSURE-AF, CHAMPION-AF) should not be lumped into one pooled result, because they enrolled different patients, used different comparators, devices, drug regimens and endpoints, and followed patients for different lengths of time. The authors conclude that anticoagulation remains the preferred first-line strategy for most eligible patients. LAAC is a reasonable option when long-term anticoagulation is unsafe or unsustainable, though trial evidence in that group is limited. Because the two approaches carry different kinds of risk, the choice should rest on shared decision-making. This summary is based on the abstract only.

0:00--:--

Detailed Summary

Stroke prevention in atrial fibrillation (AF) is a cornerstone of cardiovascular healthspan, since AF becomes more common with age and stroke is a leading cause of disability. Anticoagulation is standard, but bleeding risk and adherence problems have driven interest in left atrial appendage closure (LAAC), a device that seals the pouch where most AF-related clots originate.

This JACC Evidence Review, a narrative expert analysis rather than a new trial, examines the major randomized trials of LAAC versus anticoagulation, which together enrolled more than 7,000 participants. The authors treat each trial, including the recent OPTION, CLOSURE-AF and CHAMPION-AF, as a distinct experiment and propose a phenotype-based framework for interpreting them.

Their central result is that these trials differ materially in patient phenotype, comparator (warfarin versus direct oral anticoagulants), device generation, antithrombotic regimen after the procedure, endpoint construction and follow-up. Study-level meta-analysis can show the broad direction of effect for a device class, but a pooled estimate does not map to any single patient group or treatment decision. The authors also flag methodological problems: bleeding, the supposed advantage of LAAC, was included in composite endpoints tested for noninferiority; noninferiority margins varied; some analyses may undermine noninferiority interpretation; and some trials excluded periprocedural bleeding from primary safety endpoints. Current practice also avoids routine aspirin alongside stable anticoagulation, which lowers bleeding in anticoagulated patients compared with the control arms of the trials.

The authors conclude that anticoagulation remains first-line for most eligible patients. LAAC is reasonable when long-term anticoagulation is unsafe or unsustainable, provided procedural risk, feasibility of adjunct antithrombotic therapy, surveillance needs and competing mortality are weighed.

Caveats: this summary rests on the abstract only, so trial-level numbers and the details of the framework are not covered. Randomized evidence in the population where LAAC is most justified is limited.

Key Findings

  • The major RCTs of LAAC versus anticoagulation enrolled over 7,000 participants but asked different clinical questions, so they are not one coherent evidence base.
  • Trials differed in patient phenotype, comparator, device generation, postprocedure antithrombotic regimen, endpoints and follow-up, limiting the meaning of a pooled estimate.
  • Bleeding inside noninferiority composites, varying margins, and excluded periprocedural bleeding complicate interpretation of safety claims.
  • Anticoagulation remains preferred first-line for most eligible patients. Current standard of care drops concomitant aspirin in stable patients on therapeutic anticoagulation, so today's anticoagulated patients carry lower bleeding risk than the trials' control arms did.
  • LAAC is reasonable when long-term anticoagulation is unsafe or unsustainable, but RCT evidence in that group is limited.

Methodology

This is a JACC Evidence Review, an expert narrative appraisal of existing randomized trials (OPTION, CLOSURE-AF, CHAMPION-AF and others), not a new trial or formal meta-analysis. It evaluates each trial's design, comparator, endpoints and noninferiority framework and proposes a phenotype-based interpretive framework.

Study Limitations

This summary is based on the abstract only, so trial-specific data and the full framework are unavailable. The review is an expert interpretation rather than new pooled analysis, and the authors disclose extensive industry ties, including device makers. RCT evidence for patients who cannot take anticoagulation is limited.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: