The closure of the LAA has been established as an alternative to anticoagulation in the prevention of thromboembolism for patients with atrial fibrillation (AF). Existing percutaneous endovascular devices, notably the WATCHMAN™ (Boston Scientific), have shown a risk of thrombosis, negatively impacting patient prognosis. Recent studies, including that of Dukkipati et al., estimate an annual thrombosis incidence of 7.2% with percutaneous LAA closure devices, and an increased stroke rate associated with this complication. 

However, isolated surgical closure or resection of the LAA using devices such as vascular staplers or exclusion systems like the AtriClip™ (AtriCure™) has not clearly demonstrated this complication. This may be attributed to the epicardial application of these devices, meaning they lack direct contact with blood flow. Another potential reason for the absence of this association between epicardially applied devices and thrombosis could be the scarcity of rigorous studies that employ imaging follow-up. 

This study sought to determine the actual incidence, prognosis, and factors associated with thrombogenesis following surgical LAA occlusion. Patient data were analyzed for those who underwent two types of isolated LAA surgical closure (either resection using the Powered ECHELON™ vascular stapler, Ethicon ENDOSurger™, or exclusion using the AtriClip™ system, AtriCure™) between July 2014 and March 2020 at a single center. A total of 239 consecutive AF patients underwent minimally invasive surgical LAA occlusion (184 resection cases and 55 clipping cases). On postoperative day 2, electrocardiogram-synchronized contrast-enhanced computed tomography (CT) was performed in 223 cases (93.3%), while transesophageal echocardiography (TEE) follow-up was conducted in 16 cases where CT was contraindicated. Acute postoperative thrombus was detected on the closed stump in 35 cases (14.7%): 29 cases (15.8%) in the resection group and 6 cases (10.9%) in the clipping group. No significant difference was found between the groups, nor were significant predictors of acute-phase thrombosis identified. Thromboembolism occurred in 4 patients before the postoperative imaging follow-up, but no thrombi were found in these patients on postoperative day 2 CT. Three months after the initial CT, thrombi were no longer detected in 34 of 35 patients (97.1%). 

The authors conclude that thrombosis can occur following surgical LAA occlusion. Although its clinical significance remains unclear, it may be reasonable to continue anticoagulation therapy until the absence of thrombosis is confirmed, barring any contraindications. 

COMMENTARY:

Before delving deeper into this study, I would like to clarify some concepts to better understand our current position. AF is believed to be responsible for at least one-third of ischemic strokes. Most of these events are thromboembolic in origin, arising from the LAA. Although oral anticoagulation has proven effective and safe in preventing these events, it has limitations, including bleeding risk and lack of adherence, especially among patients treated with vitamin K antagonists (VKAs). Thanks to the LAAOS III (The Left Atrial Appendage Occlusion Study), published in 2021, we know that surgical LAA closure provides protection against ischemic strokes and systemic embolism in AF patients. Since thromboembolic risk is also elevated in patients undergoing scheduled cardiac surgery, it is incumbent upon us to close the LAA during cardiac surgery procedures using various techniques, such as the traditional continuous suture or automated devices like staplers and exclusion systems like AtriClip™. Recent studies suggest better outcomes with the AtriClip™ system, achieving a closure success rate of 96%, defined as a residual stump less than 1 cm and no contrast beyond the occlusion device on CT 12 months post-LAA closure. 

Additionally, other studies suggest discontinuing anticoagulation once effective LAA closure is confirmed, though this should particularly apply to younger patients meeting the CHA2DS2-VASc < 2 score criterion. LAAOS III study patients with a CHA2DS2-VASc score > 4 are recommended class IB anticoagulation, regardless of LAA status. Therefore, discontinuing anticoagulation is a complex issue that cannot be adequately summarized here. 

For years, interventional cardiologists have used percutaneous devices like the WATCHMAN™ (Boston Scientific™) in AF patients with contraindications or poor anticoagulation control as a stroke prevention method. The problem is that these percutaneous devices have shown a thrombosis rate close to 10% annually, likely due to their endovascular design, which maintains constant blood contact in atria lacking effective transport function, favoring thrombogenesis. 

Similarly, as an alternative to these endovascular devices, cardiac surgeons are gaining experience with AF surgery through minimally invasive procedures, allowing for isolated LAA closure using the aforementioned automatic closure devices. Moreover, through a relatively simple thoracoscopic approach, we can achieve LAA closure with automatic devices like the AtriClip™, simplifying the surgical procedure. This option presents itself as an alternative to the percutaneous endovascular devices used by interventionists, which are associated with a higher thrombosis rate and thus an increased stroke risk. 

The most relevant findings from Inoue et al., evaluating over 6 years of isolated surgical LAA resection outcomes, include: 

• Acute-phase thrombus formation with epicardial closure or exclusion devices is higher than expected, approximately 15%. 

• The positive news is that maintaining good anticoagulation during the first 3 months led to the disappearance of these thrombi, with no patient experiencing a stroke. 

Although the risk of thrombus formation was thought to be low due to the absence of a residual foreign body in the bloodstream, in comparison with percutaneous devices, the frequency of acute-phase thrombosis was unexpectedly high. These results contrast with other publications where neither CT nor TEE detected thrombosis at the AtriClip™ closure level. One possibility is that in those studies, the imaging protocol was not as rigorous as in this study, which may have allowed thrombosis cases to go undetected. 

Effective LAA exclusion requires no residual pocket at the base or closure line. This has been identified as one of the most common mechanisms of failure after automatic stapler excision. Ensuring that the LAA is well closed requires multi-view TEE, which is challenging and requires experience. Additionally, the anatomical variability of the LAA is well-documented, with trabeculations and lobes posing significant challenges. However, this issue primarily affects interventional devices with components that must adapt to LAA morphology. Surgical devices, acting on the antral line connecting the LAA to the left atrium, are less sensitive to morphological variations in achieving technical success. Untreated AF is a hypercoagulable state, and a residual LAA stump poses a thromboembolic risk, leaving the patient vulnerable. Therefore, evidence of thrombus formation on the closure line in at least 1 in 10 patients should alert us to the importance of identifying these patients through meticulous echocardiographic imaging review. If thrombi are detected, proper follow-up and confirmation of adequate anticoagulation are crucial. 

No significant difference in thrombosis rate was observed between the two LAA surgical closure techniques, although a trend toward a slightly higher thrombus formation rate was noted in the automatic stapler resection group. This may be due to exclusion systems like the Atriclip™ not affecting the endocardium as they do not damage the layers of the atrial wall, while resection may cause some damage when the stapler blade penetrates the atrial wall. 

The main limitation of this study is its retrospective nature, based on single-institution cases. Future studies should include multiple institutions and established protocols for prospective surgeries, including investigations to determine optimal postoperative anticoagulation. 

In light of these excellent results, I would like to take this opportunity to encourage and motivate cardiac surgeons to initiate programs for isolated LAA closure surgery using thoracoscopic approaches with automatic devices like the AtriClip™ at their hospitals. This epicardial LAA closure technique is beginning to demonstrate more favorable outcomes than percutaneous endovascular closure in terms of short- and especially long-term thrombosis rates. The procedure is technically simple, with an extremely low complication rate, and in the event of bleeding, surgeons themselves can resolve it. Many AF patients struggling to maintain anticoagulation could benefit from this technique. All that remains is for hospitals to offer this option and, above all, for cardiologists to become aware of this emerging alternative. 

REFERENCE:

Inoue T, Takahashi H, Kurahashi K, Yoshimoto A, Suematsu Y. Incidence of Acute Thrombosis After Surgical Left Atrial Appendage Occlusion for Atrial Fibrillation. Ann Thorac Surg. 2024 Jun;117(6):1172-1176. doi: 10.1016/j.athoracsur.2024.02.012.