Highlights
– In a post hoc exploratory analysis of the randomized LAAOS III trial, surgical left atrial appendage occlusion (LAAO) performed during cardiac surgery was associated with fewer ischemic strokes classified as cardioembolic and with fewer cortical infarcts on imaging.
– Among participants who experienced an ischemic stroke, allocation to LAAO correlated with lower early disability (better modified Rankin Scale scores at day 7 or discharge) and reduced 30-day mortality after stroke.
– These findings support the mechanistic rationale for surgical LAAO as an adjunctive stroke-prevention strategy in patients with atrial fibrillation (AF) undergoing cardiac surgery, while highlighting limitations inherent to post hoc analyses and the need for further confirmatory research.
Background: clinical context and unmet need
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and a major cause of ischemic stroke. Thromboembolism originating from the left atrial appendage (LAA) accounts for a large proportion of AF-related cerebral emboli. Anticoagulation reduces stroke risk substantially, but residual risk remains and anticoagulation is contraindicated or not tolerated in some patients.
Left atrial appendage occlusion (LAAO) can be performed surgically at the time of cardiac surgery or percutaneously as a standalone procedure. The randomized LAAOS III trial previously demonstrated that adding surgical LAAO to standard care (which frequently included oral anticoagulation) reduced stroke or systemic embolism incidence in patients with AF undergoing cardiac surgery. Understanding how LAAO modifies the mechanism and severity of strokes—beyond simply reducing overall incidence—has important implications for clinical decision-making and mechanistic understanding of embolic disease in AF.
Study design and methods
The present report is a post hoc exploratory analysis of the LAAOS III randomized clinical trial. LAAOS III enrolled patients with AF and CHA2DS2‑VASc score ≥2 who were undergoing cardiac surgery for other indications at 105 centers across 27 countries between July 2012 and October 2018.
Participants were randomized to surgical LAAO performed at the time of cardiac surgery plus standard care, versus standard care alone. Standard care typically included guideline-directed management of AF and stroke risk, including anticoagulation when appropriate.
This analysis focused on participants who experienced a first ischemic stroke during trial follow-up (median follow-up 3.8 years). Outcomes adjudicated for each stroke included: functional outcome (modified Rankin Scale, mRS) assessed at day 7 or discharge, 30-day mortality after the stroke, neuroimaging characteristics (presence of cortical infarct), and clinical adjudication of the likely stroke mechanism (with particular attention to infarcts of presumed cardioembolic origin). Analyses were conducted post hoc and are therefore exploratory; investigators performed formal adjudication and statistical comparisons such as odds ratios, risk differences, and hazard ratios with 95% confidence intervals where appropriate.
Key findings
Population and event counts
Among 4,811 participants randomized in LAAOS III, 273 experienced a first ischemic stroke during follow-up. Mean age at first ischemic stroke was 75 years (SD 7); 62% were male and 38% female. The analysis compared stroke characteristics and outcomes between the LAAO and no‑LAAO groups.
Functional outcome and early mortality
Stroke severity at 7 days or discharge, measured by mRS, favored the LAAO group. Participants allocated to LAAO had a shifted distribution toward lower (better) mRS scores compared with those randomized to no LAAO (common odds ratio for a worse mRS, 0.80; 95% CI, 0.65–0.99). In practical terms, strokes occurring after LAAO were on average less disabling in early follow-up.
Thirty‑day mortality following ischemic stroke was also lower in the LAAO group: 16.5% versus 20.1% in the no‑LAAO group, corresponding to a hazard ratio of 0.55 (95% CI, 0.31–0.97). This suggests a clinically meaningful reduction in early poststroke mortality among patients who had undergone surgical LAAO.
Imaging pattern and presumed mechanism
Neuroimaging demonstrated fewer cortical infarcts among stroke cases in the LAAO arm: 46.2% versus 61.3% in the no‑LAAO arm. The difference in proportions was −15.2% (95% CI, −26.7% to −3.7%). Cortical infarcts are commonly associated with embolic phenomena, particularly cardiogenic embolism to the cerebral cortex.
Similarly, adjudication identified a lower proportion of ischemic strokes classified as presumed cardioembolic in the LAAO group compared with no LAAO (42.9% vs 57.9%; difference −15.1%; 95% CI, −26.5% to −3.7%). These converging signals—fewer cortical infarcts and fewer events adjudicated as cardioembolic—support the hypothesis that surgical LAAO reduces cardiac‑origin embolization.
Effect sizes and statistical considerations
Point estimates indicated a favorable effect of LAAO both on stroke mechanism (fewer cardioembolic strokes) and on outcomes after stroke (better early functional status, lower 30‑day mortality). Confidence intervals for the primary contrasts reported excluded the null for several comparisons, but it is important to emphasize the exploratory, post hoc nature of these analyses and the potential for multiple testing and residual confounding despite randomization at trial entry.
Expert commentary: mechanistic interpretation and clinical relevance
These findings reinforce the biologic plausibility that the left atrial appendage is a dominant intracardiac source of thrombi in patients with AF. By occluding or excluding the LAA surgically at the time of cardiac surgery, the substrate available for thrombus formation is reduced, plausibly lowering the frequency of emboli that are large or cortical, and thereby decreasing the incidence of severe cortical strokes and early mortality.
Notably, most participants in LAAOS III continued to receive guideline‑recommended anticoagulation when appropriate, and LAAO was evaluated as an adjunct, not as a substitute, for medical therapy in this trial. Therefore, the observed reductions in cardioembolic strokes and improved stroke outcomes argue that combined strategies (surgical LAAO plus anticoagulation when tolerated) can have an additive protective effect.
From a clinical standpoint, the results strengthen the case for routine consideration of surgical LAAO in patients with AF undergoing cardiac operations for other indications, particularly when surgical exposure allows safe and effective occlusion. However, application to populations not undergoing cardiac surgery, or to percutaneous LAAO devices, cannot be assumed from these data and requires direct evaluation.
Limitations and caveats
Key limitations include the post hoc nature of the analysis and potential biases in stroke characterization: not all strokes may have had uniform imaging or comprehensive etiologic workup, and adjudication of mechanism entails clinical judgment. The timing and completeness of neuroimaging were not uniform across centers, which could influence identification of cortical versus subcortical infarcts. Additionally, because randomization occurred at trial entry for prevention of first stroke, the subset analysis of stroke cases is not a randomized comparison for stroke severity per se and is susceptible to residual confounding.
Generalizability is confined to patients with AF undergoing cardiac surgery; many trial participants continued anticoagulation after surgery, so these findings do not speak directly to anticoagulation cessation strategies after LAAO. Finally, surgical techniques for LAAO vary (excision, epicardial clip, suture closure), and the analysis does not provide granular comparison of technical approaches or their relative efficacy.
Clinical implications and research priorities
For clinicians, the analysis provides supportive evidence that surgical LAAO performed during cardiac surgery for patients with AF confers not only a reduction in stroke incidence (as shown in primary LAAOS III results) but also alters the phenotype of residual strokes toward less embolic‑appearing patterns and better early outcomes. When feasible and performed by experienced surgeons, LAAO should be part of the perioperative discussion in eligible patients with AF.
Outstanding research questions include: how these findings translate to percutaneous LAAO devices; the long‑term durability and completeness of different surgical LAAO techniques; the optimal integration of LAAO with anticoagulation (when and whether anticoagulation can be safely reduced or stopped); and whether specific subgroups derive larger absolute benefit (for example, patients with prior LAA thrombus, very high CHA2DS2‑VASc scores, or contraindications to anticoagulation).
Conclusion
Post hoc analysis of the LAAOS III randomized trial indicates that surgical LAAO at the time of cardiac surgery in patients with AF is associated with fewer ischemic strokes judged to be cardioembolic, a lower proportion of cortical infarcts on imaging, better early functional outcomes, and lower 30‑day mortality among those who had a stroke. These observations support the mechanistic role of the left atrial appendage in AF‑related embolism and reinforce consideration of surgical LAAO as an adjunctive stroke prevention strategy in appropriate surgical candidates. Clinicians should interpret these exploratory findings in the context of the trial’s randomized primary outcomes, the adjunctive role of LAAO alongside anticoagulation, and the need for further prospective studies focused on mechanism and long‑term outcomes.
Funding and clinicaltrials.gov
The LAAOS III trial is registered at ClinicalTrials.gov (Identifier: NCT01561651). Funding details and full trial funding acknowledgments are available in the primary LAAOS III publications and the trial registry.
References
1. Katsanos AH, Whitlock RP, Belley‑Côté EP, et al. Stroke Mechanism and Severity After Left Atrial Appendage Occlusion: Insights From the LAAOS III Randomized Clinical Trial. JAMA Neurol. 2025 Nov 17:e254478. doi:10.1001/jamaneurol.2025.4478 IF: 21.3 Q1 .
2. Whitlock RP, Belley‑Côté EP, et al. Left Atrial Appendage Occlusion during Cardiac Surgery to Prevent Stroke. N Engl J Med. 2021;384:2081–2091.
Note: Readers are encouraged to consult the full text of the referenced LAAOS III publications and current AF management guidelines for detailed methodology, complete safety results, and procedural recommendations.