Computed tomography detection and quantification of left atrial appendage residual patency as collateral finding after percutaneous closure

Background: Peridevice leaks after left atrial appendage closure (LAAC) may increase the risk of embolic stroke. This study appraises the value of a clinically indicated angio-computed tomography (CT) to assess the presence and size of LAA patency after percutaneous closure. Methods: We retrospectively analysed patients who underwent LAAC in our centre for a clinically indicated angio-CT to quantify Hounsfield units (HU) in LAA and in the left atrium (LA) and correlated them with the presence and size of LAA leaks at TEE. Results: CT scan was available in 56 patients of whom 40 also underwent TEE assessment. Any LAA leak at TEE was present in 9/40 (22.5%) patients of whom all had HU >100 in the LAA. However, HU measured in the LAA was >100 HU in 8 additional patients with no leak at TEE, leading to a sensitivity of 100% (9/9), specificity of 74.1% (23/31) and diagnostic accuracy of 80% (32/40). LAA HU or LAA/LA HU ratio did not discriminate LAA leak size at angio-CT. However, a coaptation gap >3 mm at angio-CT between device and LAA ostium was present in all cases with leak size >3 mm at TEE. Conclusions: HU > 100 in the LAA and a coaptation gap >3 mm between device and LAA ostium at angio-CT identified all LAA leaks and those >3 mm at TEE, respectively