Left atrial appendage occlusion — percutaneous device closure of the left atrial appendage for stroke prevention in nonvalvular atrial fibrillation
Atrial fibrillation quintuples the risk of ischemic stroke, and roughly 90% of left atrial thrombi in nonvalvular AFib form in the left atrial appendage (LAA) — a trabeculated, blind-ending pouch where blood stagnates during fibrillatory contraction. Oral anticoagulation is the default stroke-prevention therapy, but a substantial minority of patients cannot tolerate it long-term: prior major bleeding, falls risk, poor compliance, or genuine contraindication. For these patients, mechanically sealing the LAA removes the thrombus source without requiring systemic anticoagulation.
The LAA is a remnant of the embryonic left atrium: a narrow-necked, multilobed pouch with a highly trabeculated (pectinate muscle) internal surface. During normal sinus rhythm its contractile emptying velocity clears blood efficiently. In atrial fibrillation, coordinated contraction is lost — the appendage becomes a low-flow, stagnant reservoir where the Virchow triad (stasis, endothelial injury, hypercoagulability) converges.
Transesophageal echocardiography (TEE) studies and surgical/autopsy correlations consistently localize the large majority of nonvalvular AFib-associated left atrial thrombi specifically within the LAA rather than the atrial body — this is the anatomical rationale for treating the appendage as a discrete, mechanically excludable stroke source rather than anticoagulating the whole circulation indefinitely.
CHA₂DS₂-VASc scores annual stroke risk from Congestive heart failure, Hypertension, Age ≥75 (2 points), Diabetes, prior Stroke/TIA/thromboembolism (2 points), Vascular disease, Age 65–74, and female Sex category. A score of 0 (men) or 1 (women) is considered low risk; ≥2–3 crosses the guideline threshold recommending anticoagulation.
HAS-BLED scores bleeding risk from Hypertension, Abnormal renal/hepatic function, prior Stroke, Bleeding history/predisposition, Labile INR, Elderly (>65), and Drug/alcohol use. A score ≥3 flags elevated bleeding risk — it is not an automatic contraindication to anticoagulation, but it identifies patients where the risk-benefit calculus is closest, and where reversible bleeding risk factors should be corrected first.
LAAO candidacy sits at the intersection: elevated CHA₂DS₂-VASc (stroke risk justifies intervention) combined with a clinical contraindication to durable anticoagulation — prior life-threatening bleed, recurrent falls with head trauma risk, occupational bleeding hazard, or genuine intolerance/non-adherence despite optimized dosing.
The pivotal randomized trials — PROTECT AF (2009) and PREVAIL (2014) for the Watchman device — established non-inferiority of LAAO to warfarin for stroke prevention in eligible nonvalvular AFib patients, with a superior long-term bleeding profile after the peri-procedural period. The 2019 AHA/ACC/HRS focused update and the 2020 ESC AFib guideline both give LAAO a class of recommendation for patients with an elevated stroke risk and a formal contraindication to long-term anticoagulation.
Contraindications to LAAO itself include intracardiac thrombus already present in the LA/LAA, an LAA anatomy unsuitable for available device sizes, and inability to tolerate even short-term post-procedural antithrombotic therapy.
LAAO is not a first-line alternative to anticoagulation for anticoagulation-eligible patients — trial data support it specifically for patients who cannot durably tolerate oral anticoagulation. Shared decision-making, formalized in a dedicated decision aid, is a guideline-embedded requirement before proceeding.
The left atrial appendage cannot be reached from the venous side of the circulation without first crossing the interatrial septum. Transseptal puncture — a technique borrowed from decades of electrophysiology practice — creates a controlled, single-use passage from the right atrium into the left atrium, through which the entire LAAO delivery system is subsequently advanced under continuous fluoroscopic and echocardiographic guidance.
A steerable sheath and dilator are advanced from the femoral vein through the inferior vena cava into the right atrium. A Brockenbrough needle, curved to seat against the fossa ovalis — the thin, membranous, embryologically fused segment of the septum — is advanced under fluoroscopic and TEE/intracardiac echo (ICE) guidance until the characteristic “tenting” of the septum toward the left atrium is visualized.
Once position is confirmed in two imaging planes, the needle is advanced through the septum with a controlled pop, left atrial pressure waveform and blood return confirm entry, and the dilator/sheath are advanced over a guidewire into the left atrium. Puncture site location matters clinically: a puncture that is too posterior/inferior gives poor coaxial alignment with the appendage; anterosuperior punctures can complicate device sheath steering.
Once left atrial access is secured, the delivery sheath is exchanged over a stiff wire positioned deep in a pulmonary vein for stability, then withdrawn and re-steered toward the LAA ostium. Selective LAA angiography — contrast injected directly into the appendage via a pigtail catheter — outlines the appendage’s lobar anatomy (typically classified as chicken-wing, windsock, cactus, or cauliflower morphology), its ostial diameter, and its landing zone depth.
This angiographic road map is fused with pre-procedural CT and intra-procedural TEE measurements to finalize device sizing before the occluder itself is advanced.
LAAO is performed under simultaneous fluoroscopy and echocardiography (TEE under general anesthesia, or increasingly ICE from within the right atrium under conscious sedation). Fluoroscopy provides real-time catheter/device silhouette and contrast opacification; echo provides direct tissue-contact information — septal tenting, appendage wall proximity, device compression, and color Doppler flow around the device edge — that fluoroscopy alone cannot show.
Standard TEE views for LAAO (0°, 45°, 90°, 135°) sweep through the appendage to capture its maximum ostial diameter and depth, since appendage anatomy is rarely circular and single-plane measurement can substantially under- or over-estimate true dimensions.
Because the LAA ostium is typically oval rather than circular, TEE and cardiac CT are used together: CT provides 3D volumetric reconstruction of the true ostial ellipse, while live TEE confirms dynamic dimensions and guides the catheter in real time during the procedure itself.
The occluder is a self-expanding nitinol frame covered with a permeable or semi-permeable membrane, compressed into a delivery catheter and released at the appendage ostium, where it expands to conform to the landing zone and is anchored by circumferential fixation barbs. Correct sizing — oversizing the device relative to the measured ostium — is the single most important technical determinant of a durable, leak-free seal.
Device selection begins well before the procedure with careful measurement of the LAA ostial diameter (the plane where the appendage meets the atrial body) and landing zone depth (how far into the appendage the device can be seated) across multiple TEE angles and/or a gated cardiac CT reconstruction. Because the ostium is elliptical, the maximum diameter across all imaging planes is generally used to select device size, erring toward the larger measurement to ensure adequate compression.
Manufacturer sizing charts map a measured ostium diameter to a specific device size, targeting 8–30% oversizing (the design that permits enough radial force for a stable, leak-resistant seal without such vast oversizing that the device cannot fully expand within a shallow landing zone).
The compressed device travels through the delivery sheath to the LAA landing zone and is advanced out of the sheath incrementally. As it exits confinement, the nitinol frame — pre-shaped in its expanded geometry and constrained only by the delivery catheter — self-expands to its programmed diameter, driven by the shape-memory property of the nickel-titanium alloy returning to its heat-set austenitic form at body temperature.
Radially-oriented fixation barbs around the device’s widest circumference engage the surrounding appendage tissue, preventing migration once seated. The device remains attached to the delivery cable at this point, allowing it to be recaptured (resheathed) and repositioned or resized if imaging shows a suboptimal position before final release.
Before final release, operators confirm four criteria, commonly abbreviated PASS:
• Position — the device sits at or just distal to the ostial plane, not protruding into the atrial body nor recessed too deep into the appendage • Anchor — gentle tug testing (“tug test”) confirms the device does not dislodge and barbs are engaged • Size — device compression falls within the manufacturer-specified range (typically 10–30% compression of nominal diameter); under-compression risks embolization, over-compression risks incomplete tissue apposition • Seal — color Doppler around the entire device perimeter shows no significant residual flow into the appendage
If any PASS criterion fails, the device can be recaptured into the sheath and either repositioned in the same landing zone or exchanged for a different size — a key safety feature that distinguishes modern self-expanding occluders from earlier non-recapturable designs.
A device that looks well-seated on fluoroscopy alone can still leave a residual channel of blood flow around its edge — a peri-device leak — that preserves some risk of appendage thrombus formation. Before the delivery cable is detached and the device is permanently released, operators must actively interrogate the entire device-tissue interface with contrast angiography and echocardiographic color Doppler.
A small pigtail or multipurpose catheter positioned near the device injects a limited bolus of iodinated contrast toward the appendage. In a well-sealed device, contrast should stain the atrial-facing surface of the occluder but fail to opacify the appendage cavity beyond it — persistent filling of the appendage past the device is a direct sign of incomplete occlusion, prompting further evaluation before proceeding.
Angiography is fast and provides an immediate global impression, but it is less sensitive than echocardiography for detecting small, eccentric peri-device leak jets, so it is always paired with — not substituted for — TEE color Doppler assessment.
The device perimeter is swept circumferentially in multiple TEE angles (0°–135° in 45° increments, at minimum) with color Doppler active, specifically looking for a jet of flow squeezing between the device rim and the adjacent appendage or atrial wall. Any detected jet is measured at its narrowest (vena contracta) width, and graded:
• No leak / trivial (<1 mm) — considered a complete seal • Small (1–5 mm) — generally accepted; residual thrombus risk is low and most trial protocols permit device release • Moderate–large (>5 mm) — associated with higher subsequent thromboembolic risk in registry data; operators typically reposition, resize, or in select cases deploy a second device to reduce the leak before release
Peri-device leak is distinct from device-related thrombus (clot forming on the atrial-facing surface of the device itself) — both are surveilled at follow-up imaging, but leak assessment at implant time is specifically about residual flow past the seal.
Because the device remains mechanically tethered to the delivery cable throughout PASS and leak assessment, an unsatisfactory result at this stage is not a failure — it is an expected decision branch. The occluder can be resheathed (recaptured) back into the delivery catheter, withdrawn, and either re-deployed slightly deeper/shallower in the same landing zone or exchanged for a differently sized device, then re-imaged.
Only once position, anchoring, compression, and seal are all judged adequate is the delivery cable unscrewed from the device, permanently releasing it. A final angiographic and echocardiographic pass confirms stable position immediately after release, before the delivery system and sheath are withdrawn and the transseptal puncture site (in the septum) is left to heal without closure, as with standard transseptal electrophysiology procedures.
Registry data consistently show that moderate-to-large peri-device leak at implant, or detected at follow-up imaging, correlates with a measurably higher rate of subsequent ischemic stroke or systemic embolism — which is why leak grading is treated as a hard gate before release, not a cosmetic imaging finding.
A freshly implanted occluder is a foreign surface exposed to flowing blood — until the device is biologically incorporated into the atrial wall, it carries its own transient thrombogenic risk. The final phase of LAAO therefore hands off from mechanical occlusion to a pharmacologic bridge: a defined window of anticoagulation or antiplatelet therapy that covers the device until neoendocardial tissue — endothelial cells migrating across the disc surface — renders it biologically inert.
The major trial protocols established the modern bridging framework, though real-world practice has since diversified toward antiplatelet-only regimens in higher bleeding-risk patients:
• PROTECT AF / PREVAIL (Watchman): warfarin plus low-dose aspirin for the first ~45 days post-implant, with a TEE performed at 45 days — if the device shows a seal with no significant leak or thrombus, warfarin is discontinued and dual antiplatelet therapy (aspirin plus clopidogrel) continues to 6 months, transitioning to aspirin monotherapy indefinitely thereafter • ASAP protocol: developed specifically for patients who cannot take any anticoagulation at all — dual antiplatelet therapy alone from implant, demonstrating that even without an anticoagulation bridge, acceptable thrombus and stroke rates can be achieved in appropriately selected patients • Contemporary DOAC-based bridges: many modern protocols substitute a direct oral anticoagulant for warfarin during the initial 45-day window, avoiding INR monitoring while retaining the pharmacologic principle
Endothelialization is the process by which the device’s atrial-facing membrane becomes covered by a continuous monolayer of endothelial cells, migrating inward from the surrounding native atrial endocardium and proliferating across the device surface, typically over a fibrin/connective tissue scaffold that first organizes on the membrane in the initial days after implant.
Once a confluent endothelial layer covers the device, the surface is no longer a bare foreign material exposed to the coagulation cascade — it behaves, hemostatically, like native endocardium. This is the biological basis for eventually discontinuing systemic anticoagulation: the thrombogenic stimulus of an exposed synthetic surface has been resolved by the body’s own tissue growth, not by ongoing drug suppression of clotting.
Animal and limited human histopathology data suggest endothelial coverage generally progresses substantially within the first 45 days and continues to mature over the following months, which is why the standard follow-up TEE is scheduled at that interval before the anticoagulation bridge is lifted.
After the 45-day checkpoint confirms an adequate seal and absence of device-related thrombus, most patients transition off anticoagulation entirely, remaining on aspirin (with or without a short additional course of clopidogrel) long-term. This is the central clinical value proposition of LAAO: a finite, front-loaded period of antithrombotic exposure in exchange for indefinite freedom from the bleeding risk of chronic oral anticoagulation.
Long-term registry and randomized follow-up (including 5-year PREVAIL/PROTECT AF data) show stroke and systemic embolism rates comparable to warfarin, with a durable reduction in major bleeding events after the peri-procedural period. Device-related thrombus, seen in roughly 3–4% of patients on follow-up imaging, is managed with resumption of anticoagulation until resolution and does not necessarily indicate device failure, but it remains the key surveillance target of interval imaging in the first year.
The therapeutic logic of LAAO is a trade in time: accept a short, defined window of antithrombotic drug exposure and procedural risk, in exchange for converting a lifelong bleeding-risk trade-off into a one-time mechanical solution — durable so long as the seal holds and the appendage stays excluded from systemic circulation.