Long-Term Outcomes Following Early Discontinuation of Oral Anticoagulation After Left Atrial Appendage Ligation in Patients with Atrial Fibrillation

N. J.. Goel, Y. Zhao, L. A.. Gillinov, W. Lutfi, K. M. Lawrence, W. Y. Szeto, C. R. Brown, N. D. Desai
University of Pennsylvania, Philadelphia, Pennsylvania

Purpose: Surgical left atrial appendage ligation (LAAL) has been shown to reduce long-term stroke risk in patients with atrial fibrillation (AF). However, the potential benefits or harms of discontinuing long-term oral anticoagulation meant to prevent cardioembolic stroke from atrial fibrillation after surgical LAAL has not been well-studied.

Methods: This study included Medicare Part D beneficiaries undergoing CABG or valve surgery from 2016 to 2019 diagnosed with atrial fibrillation on oral anticoagulation (OAC) prior to surgery. Data on prescription drug fills were available through at least the calendar year of index surgery, and patients were followed through the entire study period by Medicare Part A inpatient claims data. Patients with early discontinuation of OAC were those that did not fill an OAC prescription at any point 60 days following surgical discharge. Propensity score matching by 21 patient, hospital, and operative covariates was used to match cohorts for comparison based on use or non-use of concomitant LAAL during index surgery and OAC continuation or early discontinuation after surgery. Outcomes of interest included mortality, stroke, systemic arterial embolism (SAE), and major bleeding events (gastrointestinal bleeding, intracranial bleeding, and others) over three years.

Results: At total of 27,789 patients undergoing CABG or valve surgery with preoperative AF on OAC were identified. Of those, 56.8% (15,796/27,789) underwent LAAL. The rate of early OAC discontinuation was similar among LAAL and non-LAAL patients (18.0% vs 19.0%). In the unmatched cohort, patients with early OAC discontinuation were more likely to have undergone CABG and less likely to have undergone valve surgery, but CHADS2VASC score was similar (4.4 vs 4.4, p=0.98) (Table 1).

On propensity-matched comparison in patients with LAAL, OAC continuation was associated with improved risk of mortality, stroke, and SAE at three years compared with early discontinuation (absolute risk reduction [95%CI] = 2.3% [0.2%-4.4%], p=0.036). However, risk was similar when major bleeding was also considered (ARR=0.8% [-1.6% - 3.2%], p=0.52) (Figure 1 A,B). On propensity-matched comparison between LAAL patients that discontinued OAC early (“LAAL only”) and non-LAAL patients that continued OAC (“AC only”), “LAAL only” patients had lower risk of mortality, stroke or SAE at three years (ARR=2.9% [0.7%-5.1%], p=0.011) and even more so when major bleeding was considered (ARR=5.1% [2.5%-7.8%], p< 0.001) (Figure 1 C,D). This occurred despite frequent early bleeding events in patients with early OAC discontinuation, which likely prompted early OAC discontinuation in many patients.

Conclusion: Among patients with preoperative AF taking OAC, 18% discontinue OAC early after surgery. For these patients, LAAL and early OAC discontinuation offers significantly better protection from long-term mortality, stroke, and major bleeding compared with OAC continuation without LAAL. Nonetheless, for LAAL patients, OAC continuation does marginally improve long-term stroke risk.

Identify the source of the funding for this research project: This study was in part supported by the National Institutes of Health [grant number 5T32HL007843-27].