BACKGROUND: Device-related thrombosis (DRT) following left atrial appendage occlusion is a rare but feared complication. The diagnostic value of cardiac compute tomography (CT) for detection of DRT is unknown. This study sought to evaluate the clinical value of cardiac CT for detection of DRT using transesophageal echocardiography (TEE) as the reference standard and to provide insights into the causes, natural history, and risk of DRT.

METHODS: We reviewed 301 consecutive patients undergoing left atrial appendage occlusion at Aarhus University Hospital, Denmark, between 2010 and 2017. Of these, 248 patients had cardiac CT and TEE imaging available at 8-week follow-up; 139 had complete 12-month imaging. A blinded investigator analyzed all images. On TEE, an echo-dense mass attached to the device was defined as DRT. Cardiac CT was analyzed for presence of hypoattenuated thickening (HAT) on the device, which was subclassified as low grade or high grade. High-grade HAT was considered as definite DRT.

RESULTS: At 8 weeks, TEE detected 5 (2%) cases with DRT; and cardiac CT 6 (2.4%) cases with high-grade HAT. At 12 months, both TEE and cardiac CT detected 2 (1.4%) cases with DRT or high-grade HAT, respectively. Cardiac CT demonstrated low-grade HAT in 9 (3.6%) cases at 8 weeks; and 13 cases (9.4%) at 12-months. High-grade HAT/DRT was associated with thromboembolism in 2 cases, whereas low-grade HAT was not related to embolic events. Low-grade HAT resolved spontaneously over time.

CONCLUSIONS: Cardiac CT seems equally good as TEE for detection of DRT. In addition, cardiac CT demonstrates cases with low-grade HAT, not visualized by TEE. The clinical significance hereof requires further investigation.