quote:

Core evidence from pivotal trials
The foundational RCTs for the Watchman device (PROTECT AF and PREVAIL) compared LAAO to warfarin:

Ischemic stroke: Numerically higher with the device. In the patient-level meta-analysis of the two trials (5-year follow-up), the hazard ratio for ischemic stroke/systemic embolism was approximately 1.7 (favoring warfarin; not always statistically significant depending on the analysis and whether procedure-related events were excluded).
Overall stroke/systemic embolism rates were similar because lower hemorrhagic stroke rates with LAAO offset the ischemic excess.
Mortality benefits (when seen) were driven more by reductions in hemorrhagic events and non-procedure-related bleeding than by superior ischemic stroke prevention.
These trials used warfarin (not modern DOACs) as the comparator and had notable procedural complications in the early experience.

More recent RCTs

CLOSURE-AF (NEJM 2026; high-risk older patients, mean CHA₂DS₂-VASc ~5.2, mostly DOAC-based medical therapy): LAAO failed to meet noninferiority for the composite of stroke, systemic embolism, major bleeding, or cardiovascular/unexplained death. Event rates favored medical therapy (≈16.8 vs 13.3 per 100 patient-years). Stroke rates were essentially identical between arms. This independent (non-industry-funded) trial is particularly relevant to Mandrola’s point.
OPTION (post-AF ablation patients): LAAO was noninferior to continued oral anticoagulation for a composite efficacy endpoint (death/stroke/systemic embolism) and superior for non-procedure-related bleeding.
CHAMPION-AF: LAAO met noninferiority versus NOACs for cardiovascular death/stroke/systemic embolism and was superior for non-procedure-related bleeding. Ischemic stroke rates were numerically higher with the device in some breakdowns.

Meta-analyses of RCTs
Pooled analyses of percutaneous LAAO versus oral anticoagulation generally show:

Similar or non-significantly higher rates of ischemic stroke/systemic embolism with the device. Lower rates of hemorrhagic stroke and non-procedure-related major bleeding. Mixed or modest effects on all-cause/cardiovascular mortality (some analyses favor LAAO, driven by bleeding reduction; others show no difference). Overall stroke rates usually comparable.

Absolute ischemic stroke rates after successful LAAO in contemporary practice are low (often ~1–2% per year), but they do not consistently beat well-managed anticoagulation.