Retrograde transaortic percutaneous retrieval of a left-ventricularly embolized LAmbre occluder followed by same-session redo left atrial appendage closure: a case report Authors: Dengke Ou, Jie Song, Mingyang Tang Published: 2026 ABSTRACT AbstractBackgroundLeft atrial appendage closure (LAAC) is an established stroke-prevention strategy for selected patients with atrial fibrillation who are unsuitable for long-term oral anticoagulation. Device embolization is rare but potentially life-threatening, particularly when the occluder migrates into the left ventricle, where it may interfere with the mitral or aortic valve apparatus or provoke ventricular arrhythmia.Case summaryA 68-year-old woman with permanent atrial fibrillation, recurrent gross haematuria during anticoagulant therapy, chronic heart failure, and a high thromboembolic risk underwent LAAC under transoesophageal echocardiographic guidance. Pre-procedural imaging showed a left atrial appendage ostium of 30 mm, a landing-zone diameter of 28 mm, and a depth of 32 mm. During the first deployment attempt with a 28-/34-mm LAmbre occluder, the device showed insufficient anchoring and was not released. During attempted recapture and exchange, the partially deployed device embolized across the mitral valve into the left ventricular cavity, with a short self-terminating episode of ventricular tachycardia. Because the patient remained haemodynamically stable and real-time imaging showed no valvular entrapment or structural injury, retrograde transaortic percutaneous retrieval was performed using a 14-Fr steerable sheath and grasping forceps. After successful retrieval and immediate confirmation of valve integrity, a larger 32-/36-mm LAmbre occluder was implanted in the same session with stable position and no residual peri-device leak.DiscussionThis case illustrated that, in carefully selected haemodynamically stable patients without valvular entrapment, left ventricular outflow tract obstruction, or structural injury, retrograde transaortic percutaneous retrieval of a left-ventricularly embolized LAAC device may be feasible. The same-session redo LAAC can be considered after successful retrieval when valve integrity is confirmed and the left atrial appendage anatomy remains suitable. Device embolization is a rare but serious complication of left atrial appendage closure requiring prompt recognition. Device embolization is a rare but serious complication of left atrial appendage closure requiring prompt recognition. In haemodynamically stable patients without structural injury, percutaneous retrieval is a feasible first-line strategy that may avoid emergent surgery. In haemodynamically stable patients without structural injury, percutaneous retrieval is a feasible first-line strategy that may avoid emergent surgery. Repeat occlusion should be preceded by careful reassessment of anatomy and device stability, with post-procedural antithrombotic therapy and imaging follow-up being essential. Repeat occlusion should be preceded by careful reassessment of anatomy and device stability, with post-procedural antithrombotic therapy and imaging follow-up being essential. Introduction Left atrial appendage closure (LAAC) is an established stroke prevention strategy for selected patients with atrial fibrillation (AF) who have contraindications to, or difficulty maintaining, long-term oral anticoagulation. Although generally safe, device embolization remains a rare but serious complication. Most embolized devices are identified in the left atrium or aorta, whereas migration into the left ventricle (LV) is particularly concerning because of the potential for mitral or aortic valve injury, left ventricular outflow tract (LVOT) obstruction, and ventricular arrhythmias. We report a case of spontaneous embolization of a partially deployed LAmbre occluder into the LV, successfully treated by retrograde transaortic percutaneous retrieval followed by same-session redo LAAC. Summary figure Case presentation A 68-year-old woman with permanent atrial fibrillation presented with recurrent gross haematuria during anticoagulant therapy. She had a clinical diagnosis of chronic heart failure but no signs of acute decompensated heart failure at presentation. Cardiac auscultation revealed no obvious murmur. Transthoracic echocardiography demonstrated a preserved left ventricular ejection fraction of 62% and no significant structural abnormality apart from left atrial enlargement. The corrected CHA2DS2-VASc score was 4, and the corrected HAS-BLED score was 2. Because of recurrent clinically relevant bleeding during anticoagulant therapy and the need for stroke prevention, percutaneous LAAC was selected after clinical evaluation and shared decision-making. The procedure was performed under general anaesthesia with continuous transoesophageal echocardiographic (TEE) and fluoroscopic guidance. Venous access was obtained via the right femoral vein. After successful transseptal puncture, intravenous heparin 100 U/kg was administered, followed by an additional 1000 U hourly to maintain an activated clotting time of 250–300 s. A 6-Fr INFINITI pigtail catheter was advanced into the left atrial appendage (LAA) for angiography. Pre-procedural TEE and angiography showed a complex LAA anatomy. The maximal LAA ostial diameter was 30 mm, the landing-zone diameter was approximately 28 mm, and the LAA depth was 32 mm. The ostial diameter and landing-zone diameter were measured at different anatomical levels. A Lifetech LAAC delivery sheath (10-Fr, SL10F45*30-900, 900 mm) was advanced into the left atrium. A 28-/34-mm LAmbre occluder was initially selected because the 28-mm umbrella corresponded to the measured landing zone, whereas the 34-mm cover was expected to seal the 30-mm ostium. Before device release, the position, anchoring, size, and seal (PASS) criteria were assessed, including PASS. During the first deployment attempt, the device showed suboptimal anchoring and insufficient stability and therefore did not fully satisfy the anchoring and sizing components of the PASS criteria. The device was not released, and an exchange to a larger occluder was planned. Before the partially deployed device was completely recaptured into the delivery sheath, the occluder embolized across the mitral valve into the mid-LV cavity. A brief episode of ventricular tachycardia lasting approximately 5 s occurred and self-terminated without haemodynamic compromise. The embolized device was stabilized in the LV, and cardiothoracic surgery was placed on standby. The mechanism of embolization was considered multifactorial rather than attributable to a single cause. Possible contributing factors included inadequate anchoring, insufficient device compression, device-anatomy mismatch, imperfect coaxiality, and dynamic LAA motion during partial recapture. Because LAA contractile force was not directly measured, the role of LAA motion should be interpreted cautiously. Quantitative compression of the initial device could not be reliably calculated retrospectively from the available fluoroscopic images. Given the patient's haemodynamic stability and the absence of structural injury on real-time imaging, percutaneous retrieval was pursued. A retrograde transfemoral arterial approach was established. An Okcurve steerable sheath (Lifetech, 14-Fr, LT-SF-14-80-50, 800 mm) was advanced across the aortic valve into the LV under fluoroscopic and TEE guidance while maintaining an activated clotting time of 250–300 s. Continuous invasive arterial pressure monitoring, electrocardiographic monitoring, fluoroscopy, and real-time TEE guidance were maintained throughout retrieval. No transient LVOT obstruction, mitral valve entrapment, new mitral regurgitation, new aortic regurgitation, or pericardial effusion was observed. Cerebral embolic protection was not used because no thrombus was visualized on the embolized device, therapeutic intraprocedural anticoagulation was maintained, and urgent retrieval was prioritized to minimize LV manipulation time. Using disposable grasping forceps (JHY-FG-23-160-A3, 1600 mm), the hub/hub-frame junction was centrally grasped in a hub-first orientation. The risk of aortic and mitral valve injury was minimized by maintaining coaxial alignment, grasping the central hub/hub-frame junction rather than the device edge, avoiding guidewire or device entanglement with the valve apparatus, and gently recapturing the device into the 14-Fr steerable sheath before en bloc removal through the femoral artery. Immediate TEE after retrieval confirmed no new aortic or mitral regurgitation, no LVOT obstruction, and no pericardial effusion. Following successful retrieval and confirmation of valvular integrity, the same-session redo LAAC was performed without changing the transseptal puncture site or the planned landing zone. Same-session redo LAAC was considered reasonable because the original indication for LAAC remained unchanged, the transseptal access remained available, and the LAA anatomy remained suitable for closure. A larger 32-/36-mm LAmbre occluder was implanted. Final angiography and TEE confirmed stable device position, adequate sealing, and no residual peri-device leak (Figure 1). Summary of device embolization, retrieval, and final occlusion. (A) Pre-procedural transoesophageal echocardiographic assessment of the Left atrial appendage closure. (B) Embolized LAmbre occluder in the left ventricle. (C) Retrograde transaortic percutaneous retrieval with progressive recapture of the embolized device into a 14-Fr steerable sheath. (D) Successful same-session redo left atrial appendage closure with stable device position and no residual peri-device leak. The patient was discharged on apixaban 2.5 mg twice daily for 6 months as an individualized regimen considering the patient's thromboembolic risk and prior recurrent gross haematuria during anticoagulant therapy. At the 6-month follow-up TEE, the device position remained stable, with no device-related thrombus or residual peri-device leak. Antithrombotic therapy was subsequently de-escalated to aspirin 100 mg once daily. During 6 months of clinical follow-up, no recurrent gross haematuria, stroke, systemic embolism, rehospitalization, or device-related complication occurred. Discussion Device embolization after LAAC is uncommon but represents one of the most serious procedural complications.1,2 Left ventricular embolization is especially concerning because an embolized device may interfere with the mitral apparatus, injure the aortic valve during migration or retrieval, obstruct the LVOT, or trigger ventricular arrhythmias. In the present case, the device embolized before formal release, crossed the mitral valve, and lodged in the LV cavity, accompanied by a short self-terminating episode of ventricular tachycardia. Prompt recognition and immediate heart-team coordination were essential.3 The present case also emphasizes the importance of systematic device assessment before release. The LAmbre device consists of a distal umbrella for anchoring and a proximal cover for ostial sealing.4 Therefore, measurements at both the ostium and the landing zone are relevant for sizing. In this case, the 28-/34-mm device was initially selected because the 28-mm umbrella corresponded to the landing-zone diameter and the 34-mm cover was expected to seal the 30-mm ostium. However, the first device did not fully satisfy the anchoring and sizing components of the PASS criteria, and it was not released. The embolization that occurred during attempted recapture was likely multifactorial, involving insufficient anchoring, device-anatomy mismatch, suboptimal coaxiality, and dynamic LAA movement. This mechanism should be interpreted cautiously because objective measurement of device compression and LAA motion was not available retrospectively. Management of LAAC device embolization should be individualized according to haemodynamic status, device location, evidence of valvular or vascular injury, and local surgical and interventional expertise.1 Surgery remains necessary when the device is entrapped in the valve apparatus, causes haemodynamic instability, cannot be safely captured, or is associated with structural injury. In contrast, when the patient is stable, and imaging shows no valve entrapment, LVOT obstruction, or structural injury, percutaneous retrieval may be considered as a first rescue strategy with surgical backup available. Several technical points may have contributed to successful retrieval in this case. First, TEE and fluoroscopy were used continuously to monitor the relationship between the device, LVOT, mitral apparatus, and aortic valve. Second, the steerable sheath was advanced retrogradely across the aortic valve to provide coaxial support. Third, the hub/hub-frame junction was grasped centrally rather than at the device edge, which may reduce the risk of leaflet entrapment and uncontrolled deformation. Fourth, the occluder was gently recaptured into the 14-Fr sheath before en bloc removal. Immediate post-retrieval TEE was important to confirm the absence of new mitral or aortic regurgitation. Same-session redo LAAC after retrieval is not routine and should not be generalized to all embolization events.5 In our patient, redo closure was performed only after successful retrieval and confirmation of mitral and aortic valve integrity, absence of pericardial effusion, and preserved haemodynamic stability. The indication for LAAC remained strong because the patient had AF with thromboembolic risk and recurrent gross haematuria during anticoagulant therapy. The transseptal access remained suitable, and the LAA anatomy remained amenable to closure. After upsizing to a 32-/36-mm LAmbre occluder, final imaging confirmed stable device position and complete sealing. Post-procedural antithrombotic therapy after LAAC should be individualized according to device type, bleeding risk, thromboembolic risk, and follow-up imaging findings.6 In this patient, short-term low-dose apixaban was selected to reduce the risk of device-related thrombus during endothelialization while considering the history of gross haematuria. After follow-up, TEE confirmed no device-related thrombus and no residual leak; therapy was de-escalated to aspirin. Patient’s perspective The patient expressed relief that the complication was managed without the need for open-heart surgery and was satisfied that the intended LAA closure was completed successfully during the same procedure. Conclusion This case suggests that, in carefully selected haemodynamically stable patients without valvular entrapment, LVOT obstruction, or structural injury, retrograde transaortic percutaneous retrieval of a left-ventricularly embolized LAAC device may be feasible. The same-session redo LAAC can be considered after successful retrieval when mitral and aortic valve integrity is confirmed, the original indication for LAAC remains unchanged, and the LAA anatomy remains suitable for stable closure. Supplementary Material