Isolated surgical left atrial appendage closure: Revisiting utility and indications in a burgeoning era of percutaneous therapy

Abstract

Background and Aim:

Oral anticoagulation (AC) and percutaneous left atrial appendage (LAA) occlusion are the primary treatment modalities for stroke prevention in atrial fibrillation (AF), but there remains a subset of patients in whom these approaches present excess risk and isolated surgical LAA excision should be considered. We describe a 63‐year‐old female with AF and recurrent thromboembolic events who presented with an acute intraparenchymal hemorrhage and was found to have an intracardiac thrombus.

Methods:

Given contraindications to AC and LAA occlusion, an isolated LAA surgical excision was pursued.

Results:

She underwent successful surgical LAA excision and has since remained event‐free.

Conclusion:

It is important to recall the utility of therapies that have been previously used with success for intracardiac thrombi and still remain as viable options.

1 |. INTRODUCTION

Atrial fibrillation (AF) increases the risk of thromboembolism and stroke. Oral anticoagulation (AC) and percutaneous left atrial appendage occlusion (LAAO) with short term AC are the primary treatment modalities for stroke prevention, but there remains a subset of patients in whom these approaches present excess risk.¹ Although rarely performed, isolated surgical left atrial appendage (LAA) excision should not be forgotten as a viable alternative for patients in whom standard practices are not feasible.

2 |. CASE PRESENTATION

A 63‐year‐old woman with long‐standing persistent AF on coumadin therapy presented with acute left parieto‐occipital intraparenchymal hemorrhage (IPH) with intraventricular extension while on coumadin with an international normalized ratio (INR) of 3.0. Her medical history included prior aortoiliac saddle thrombus requiring emergent bilateral iliofemoral catheter embolectomy (INR 1.3). Echocardiography revealed hyperdynamic left ventricular function, dilated atria and significant left ventricular hypertrophy. An extensive inflammatory, autoimmune and hypercoagulable diagnostic evaluation was only remarkable for elevated beta‐2‐microglobulin, but cardiac magnetic resonance imaging was not suggestive of amyloidosis or other infiltrative processes. During hospitalization, she developed acute lower extremity pain and was found to have recurrent bilateral aortoiliac saddle thrombi requiring emergent bilateral aortoiliac embolectomy and four‐compartment fasciotomy.

Given the history of IPH, LAAO consultation was sought. Screening transesophageal echocardiography revealed a large LAA thrombus measuring 2.89 cm × 4.11 cm (Figure 1), disqualifying her candidacy for percutaneous LAA closure. Despite rebleeding risk, it was decided to retrial AC with an alternative agent given high thrombus burden and thromboembolic risk. Apixaban was started and increased to full dose within a week, but despite appropriate adherence, she developed acute subclavian artery thrombosis also requiring urgent surgical management. Faced with the challenge of minimizing thromboembolic risk given recurrent large‐vessel occlusive thromboembolism on AC, while avoiding potentially catastrophic bleeding (Figure 2), our patient was referred for surgical thrombus extraction (Figure 1) and isolated LAA excision. After careful discussion, surgery was performed with the patient in the right semi-decubitus position via left posterolateral thoracotomy through the fourth intercostal space, using TEE guidance and cardiopulmonary bypass (CPB) due to concern for LAA adhesions from prior thoracic surgeries. Cannulation of the right and left femoral artery and the vein was performed via cut‐cut down, using 19 and 25‐Fr cannulas, respectively. Considerable dissection through a densely adhered pericardium was performed to expose the LAA. 4‐0 Prolene sutures were placed in horizontal mattress fashion and the appendage was amputated. An additional layer of 4‐0 prolene running suture was placed for hemostasis. The patient was weaned from CPB without any issues and discharged on a postoperative day 6. She has remained event‐free at 1‐year, further implicating LAA thrombi as the etiology of her recurrent presentations.

FIGURE 1.

A, B, Echocardiographic evidence of large left atrial appendage (LAA) thrombus measuring 2.89 cm × 4.11 cm. C, Excised LAA

FIGURE 2.

Event timeline. AC, anticoagulation; INR, international normalized ratio; LAA, left atrial appendage

3 |. DISCUSSION

The thromboembolic risk of AF has been well elucidated, with 15% of all strokes occurring per year in the United States being attributed to AF and about 90% of these resulting from thrombus formation in the LAA.^2,3 In recent years, nonpharmacologic approaches for LAA closure have been established as an attempt to provide a therapeutic option beyond AC. The Watchman device is the only FDA‐approved LAAO device, but it is important to highlight that patients with absolute contraindications to AC were not included in clinical trials. Therefore, the clinical utility in this specific patient population remains uncertain. Also, the device does not target patients with nonvalvular AF that develop thrombi outside of the LAA; recent studies have raised concern over device‐related thrombus and association with a higher rate of stroke and embolic phenomena.⁴ Ultimately, while multiple percutaneous LAAO devices have been developed with studies showing efficacy, there is still no alternative for patients with preexisting intracardiac thrombus. Prior studies using diverse diagnostic modalities in patients with AF have identified the incidence of LAA thrombi up to 20%.^5,6 Moreover, the presence of dense echo contrast in the LAA, a precursor to thrombus formation, was similarly noted in 20% of patients with AF in the Stroke Prevention in Atrial Fibrillation III trial and was associated with threefold increased stroke risk.⁷ In general, visualization of LAA thrombi is associated with poorer outcomes. A recent study evaluating the clinical impact of LAA thrombus in a population referred for transcatheter aortic valve replacement found an incidence of LAA in 11% of patients and in 32% of patients with AF. Postprocedurally, the stroke rate was 20% in patients with known LAA thrombus compared with 3.8% in patients without intracardiac thrombi.⁸

Despite significant morbidity and mortality of thromboembolic events in this population, there are only a few cases in the literature that describe management of LAA thrombus in patients with AF and contraindications to AC. One study reported a patient with persistent atrial thrombus despite AC who underwent successful LAAO device implantation with an Amulet device. A cerebral protection device was placed beforehand, and he endured no postprocedural complications.⁹ A second report described a woman with LAA thrombus and recurrent thromboembolic events while on AC. She underwent LAAO with an Amplatzer Cardiac Plug through a no‐touch technique.¹⁰ Similarly, another report described a patient with recent episodic gastrointestinal bleeding who developed recurrent frontal embolic stroke and was found to have a significant thrombus burden in the LAA. She underwent a thoracoscopic beating‐heart approach using hoop‐less open‐ended surgical exclusion with an AtriClip device.¹¹

Our case was performed using a posterolateral thoracotomy approach, with the necessity of CPB driven by reoperation. The LAA is a very thin‐walled structure, easily injured during the dissection. Opening the left atrium may introduce air, yielding embolic stroke. Hence, CPB was used to decompress and lower the risk of injury. Furthermore, a Maze procedure was not performed given a lack of efficacy in patients with long‐standing AF and significant LA dilation.

Further studies are warranted to elucidate if techniques such as delivery system cannulation without the use of a pigtail guide or using cerebral protection systems can allow for the utilization of LAA closure devices in patients with preexisting intracardiac thrombus. For the time being, surgical excision is an accepted alternative approach. The 2016 European Society of Cardiology and the 2014 American Heart Association/American College of Cardiology Atrial Fibrillation guidelines offer a class IIb recommendation for both routine surgical LAA excision in patients undergoing cardiac surgery and for isolated excision in patients with contraindications for AC and high risk of stroke.^12–14 Similarly, the 2014 American Association of Thoracic Surgery guidelines of AF management and prevention also offer a class IIB recommended for LAA excision at the time of surgery in patients who are considered high risk for AC.¹⁵ More promising, advances in surgical technology have facilitated the development of novel LAA exclusion systems that may potentially be of clinical utility in this group of patients. A study evaluating the AtriClip device revealed that all clips were in a stable position, with no evidence of intracardiac thrombi, LAA perfusion, strokes or neurological events 36 months following device placement.⁵ While newer technology is being designed to assess minimally invasive techniques that decrease the risk of long‐term thrombosis, peridevice leak, hemodynamic impact, and neurohormonal modulation, further studies are warranted for patients with preexisting LAA thrombus.

4 |. CONCLUSION

This case highlights the complex clinical challenge of determining the ideal option for thromboembolism prevention in patients with LAA thrombi in whom AC is contraindicated. In an era of expanding percutaneous and medical treatment options, it is important to recall the utility of therapies that may seem novel in contemporary practice but had been previously used with success and remain viable options.