Background: Surveillance after left atrial appendage (LAA) closure is important to assess for device success and complications. There is limited data on the role of cardiac computerized tomography angiography (CCTA) for LAA device monitoring.
Methods: We retrospectively analyzed consecutive patients with non-valvular atrial fibrillation who underwent LAA closure who had both trans-esophageal echocardiography (TEE) and CCTA performed post-LAA closure. Prospective cardiac-gated CCTA were performed with Toshiba 320-detector or Siemens 2nd generation 128-slice dual-source scanners. We assessed for residual contrast patency of the LAA by comparing linear attenuation coefficient (Hounsfield unit, HU) in the LAA and left atrium (LA) on CCTA. Residual contrast patency was categorized as due to peri-device leak (presence of ostial gap) or fabric leak (diffusion of contrast through device membrane). We also assessed for device compression, thrombus and pericardial effusion on CCTA. These were compared to TEE findings.
Results: Sixty-one patients underwent CCTA at median 93.5d (1-267) post-LAA closure (55 WATCHMAN, 4 Amulet, 2 ACP). Post-procedural TEE was done at median 93d (0-289). The median age was 77.5yrs (60-88) and median CHADS-VASc score 4 (2-8). Procedural success was 100%. On TEE, the presence of any peri-device leak was 26 (41.9%), with >3mm leak observed in 10/62 (16.1%), and >5mm observed in none. The presence of residual contrast patency in the LAA was observed in 36/61 (59%) on CCTA, of these, 29 (80.5%) had peri-device leak, and 7 (19.4%) had fabric leak. The median linear attenuation coefficient in patients with residual contrast patency was 328 (133-652) HU, and all had >100 HU in the LAA and HU attenuation ratio LA/LAA of >0.25. Median maximum device compression was 5.7% in those with peri-device leak, and 19.4% (p=0.027) for those without peri-device leak on CCTA. Among patients with peri-device leak on CCTA, 89.6% had peri-device leak on TEE, with corresponding 93.8% sensitivity, 69.9% specificity, and AUC=0.82 (95% CI 0.72-0.93).
Conclusions: CCTA is more sensitive than TEE for assessing peri-device leak after LAA closure. Greater device compression appears to be associated with lower incidence of peri-device leak.