Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/ou_press/left-atrial-appendage-patency-and-device-related-thrombus-after-UJys4oO2il
- Publisher
- Oxford University Press
- Copyright
- Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com.
- ISSN
- 2047-2404
- DOI
- 10.1093/ehjci/jey010
- Publisher site
- See Article on Publisher Site
Abstract
Abstract Aims Transoesophageal echocardiography studies have reported frequent peri-device leaks and device-related thrombi (DRT) after percutaneous left atrial appendage (LAA) occlusion. We assessed the prevalence, characteristics and correlates of leaks and DRT on cardiac computed tomography (CT) after LAA occlusion. Methods and results Consecutive patients underwent cardiac CT before LAA occlusion to assess left atrial (LA) volume, LAA shape, and landing zone diameter. Follow-up CT was performed after >3 months to assess device implantation criteria, device leaks and DRT. CT findings were related to patient and device characteristics, as well as to outcome during follow-up. One-hundred and seventeen patients (age 74 ± 9, 37% women, CHA2DS2VASc 4.4 ± 1.3, and HASBLED 3.5 ± 1.0) were implanted with Amplatzer cardiac plug (ACP)/Amulet (71%) or Watchman (29%). LAA patency was detected in 44% on arterial phase CT images and 69% on venous phase images. The most common leak location was postero-inferior. LAA patency related to LA dilatation, left ventricular ejection fraction impairment, non-chicken wing LAA shape, large landing zone diameter, incomplete device lobe thrombosis, and disc/lobe misalignment in patients with ACP/Amulet. DRT were detected in 19 (16%), most being laminated and of antero-superior location. DRT did not relate to clinical or imaging characteristics nor to implantation criteria, but to total thrombosis of device lobe. Over a mean 13 months follow-up, stroke/transient ischaemic attack occurred in eight patients, unrelated to DRT or LAA patency. Conclusion LAA patency on CT is common after LAA occlusion, particularly on venous phase images. Leaks relate to LA/LAA size at baseline, and device malposition and incomplete thrombosis at follow-up. DRT is also quite common but poorly predicted by patient and device-related factors. computed tomography, left atrial appendage occlusion, peri-device leaks, device-related thrombus Introduction Atrial fibrillation (AF) is the most frequent cardiac arrhythmia, affecting up to 13% of patients over 80 years, and responsible for 15–20% of all ischaemic strokes.1,2 Oral anticoagulation (OAC) using either vitamin K antagonists or the more recently introduced factor II/Xa inhibitors has shown reduction in ischaemic stroke and mortality among these patients.3,4 However, the need for an alternative therapy has emerged due to the high number of patients contra indicated to OAC.5 As more than 90% of atrial thrombi are located in the left atrial appendage (LAA) in patients with non-valvular AF, LAA occlusion devices have been developed as an alternative approach to reduce the risk of stroke.6 The efficacy and safety of percutaneous LAA closure using either the Watchman device (Boston Scientific, Natick, Massachusetts) or the Amplatzer cardiac plug (ACP), and more recently introduced Amulet device (St. Jude Medical, Minneapolis, MN, USA) has been proven,7,8 and this procedure has been integrated in the guidelines for the management of patients with chronic AF.9 However, obtaining complete LAA occlusion appears to be quite challenging, with many patients still showing peri-device flow on transoesophageal echocardiography (TOE).10 In addition, a significant number of patient also develop thrombus on the atrial aspect of the device.11 Although residual leaks and device-related thrombi (DRT) have not yet been associated with an increased risk of clinical complications,12 such findings represent a major clinical management issue, often justifying the continuation of OAC therapy.7 While most studies have employed TOE, cardiac computed tomography (CT) may represent an efficient alternative method for the management of patients undergoing LAA occlusion.13 Several pilot studies suggest that it may be superior to TOE for accurate device sizing before the procedure,14 as well as for the detection of LAA patency during follow-up.15 The aim of this study was to assess the prevalence, size, and location of leaks and DRT on cardiac CT after percutaneous LAA occlusion, and to identify patient-related and device-related correlates. Methods Population and study design From June 2012 to December 2016, consecutive patients with indication for percutaneous LAA occlusion were included. The inclusion criterion was an indication for LAA occlusion according to the guidelines from the European Society of Cardiology.1 Exclusion criteria were contra-indications to iodine-enhanced CT and percutaneous LAA occlusion, and failure to obtain patient consent. At the time of inclusion, all patients underwent transthoracic echocardiography (TTE) to measure left ventricular ejection fraction (LVEF), and contrast-enhanced cardiac CT to measure the left atrial (LA) volume and assess the LAA shape and landing zone maximal diameter. An additional CT acquisition at venous phase was performed in all patients to rule out thrombus. Percutaneous LAA occlusion was performed within the week following inclusion. Follow-up cardiac CT was performed at 3 months, unless indicated earlier by potential complications. All patients with leaks or DRT on post-procedural CT were followed by TOE. OAC was re-introduced in case of thrombus or peri-device leak >5 mm on TOE.7 Additional follow-up CT studies were not part of the study protocol but when performed, results were analysed. Follow-up visits were scheduled at 1, 3, 12 months, and every 6 months afterwards and adverse events were recorded. The study was approved by the Institutional Ethics Committee, and all patients provided informed consent. Pre-procedural CT Cardiac CT studies were performed on a 64-slice dual source CT system (Siemens Definition, Siemens Medical Systems, Forchheim, Germany). Tube current was set to 120 kV in patients with body mass index (BMI) >27 and 100 kV in those with BMI <27. Acquisition was set on end-systole using prospective ECG triggering, the delay being set in percentage of the RR interval in patients in sinus rhythm, and in ms in those with arrhythmia. Images were acquired using a biphasic injection protocol: 1 mL/kg of Iomeprol 350 mg/mL (Bracco, Milan, Italy) at the rate of 5 mL/s, followed by a 1 mL/kg flush of saline at the same rate. A bolus tracking method was applied to acquire arterial phase images, the region of interest being positioned within the LA. Image reconstruction was performed with 0.75-mm slice thickness, using a B26f soft-tissue convolution kernel. Typical inplane pixel size was 0.5 × 0.5 mm. Image analysis was performed using 3mensio software (3mensio Medical Imaging BV, Bilthoven, The Netherlands). LA volume was measured using the biplane area–length method. The LAA shape was categorized as either chicken wing or non-chicken wing on a volume rendering reconstruction. The maximum landing zone dimension was measured on a cross-sectional plane at the level of the specific landing zone of ACP/Amulet and Watchman devices.7,8 LAA occlusion procedure Implanted devices consisted exclusively of ACP up to mid 2014, and of either ACP/Amulet or Watchman devices afterwards. Details regarding LAA occlusion procedures and specific features of ACP and Watchman devices have already been described in detail.7,8 Procedures were performed under general anaesthesia. All patients were on aspirin therapy that was continued during and after LAA occlusion. After femoral vein access and TOE-guided trans-septal puncture, angiography (right anterior oblique 20–30° with caudal angulation up to 20° depending on LAA shape and orientation) was done. Heparin intravenous injection was performed for an activated clotting time greater than 250 s. LA pressure was measured and fluid was delivered to the patient to reach a pressure greater than 10 mmHg to get closer to the awake LA pressure and to avoid discrepancy between intra-operative TOE and pre-operative CT measurements. The size of the device was chosen based on the maximum landing zone diameter measured on preoperative CT, this size being further confirmed during the procedure based on TOE measurements. Device oversizing conformed to the manufacturer’s instructions. The antithrombotic regimen after LAA occlusion consisted of single anti-platelet therapy, as this strategy is the current consensus in our institution.16 Follow-up CT In all patients, post-procedural CT comprised an arterial phase acquisition identical to the one performed at baseline. LA volume was measured again using the biplane area–length method. The analysis of device implantation on follow-up CT images is illustrated in Figure 1. In patients implanted with ACP/Amulet, multi-planar reformatting was used to analyse appropriate implantation criteria13 including proper tire-shape of the lobe, complete separation of the disc and the lobe, concavity of the disc, position of at least two-third of the lobe within circumflex artery (CX) level, and proper alignment between the disc and the lobe, the latter angle being measured and expressed in degree. In addition, thrombosis inside the device was categorized as either total or partial. On the atrial aspect of the device images were reviewed to look for enhancement defects suggestive of thrombus. When present, the maximal dimension and location of the defect was measured. In order to analyse LAA patency, regions of interest were drawn within the LA and the LAA to measure attenuation coefficients. LA attenuation was measured at the centre of the LA chamber. LAA attenuation was measured by drawing a region of interest within the LAA cavity, distant from the LAA borders in order not to include LAA trabeculations or surrounding fat of low attenuation, as well as distant from any device-related beam hardening artefact of higher attenuation. Complete occlusion at the arterial phase was defined as an LAA density <100 HU, and <25% of that of the LA, as previously suggested.13 Thus, any LAA exhibiting a local density ≥100 HU or ≥25% of that of the LA was considered to be patent, meaning that there was proof of residual flow within the LAA. In patients with positive LAA enhancement, leak necks were looked for on the device margins. When visible, the maximal dimension and the location of leak necks were assessed on a reconstructed plane parallel to LAA orifice. In addition, the surface of leak necks was measured and expressed as a percentage of total LAA orifice area, including the device. In case of Watchman devices, LAA patency was characterized as either due to fabric or marginal leaks. In order to further analyse LAA patency an additional acquisition at the venous phase was added to the study protocol, starting from July 2013. Such method was shown able to improve LAA contrast filling as compared with arterial phase images in the clinical setting of thrombus detection.17 Venous images were acquired at 60 s post-contrast, the imaging volume being limited to the LAA area, and the acquisition and reconstruction parameters being similar to arterial phase imaging. On these venous phase images, complete occlusion was defined as LAA density <100 HU, and <150% of that measured at the same site on arterial phase images. Figure 1 View largeDownload slide Analysis of device implantation on follow-up CT images. In patients with ACP/Amulet devices, appropriate implantation criteria comprised proper tire-shape of the lobe, complete separation of the disc and lobe, concavity of the disc, position of at least two-third of the lobe within the circumflex artery level, and the angle of alignment between the disc and the lobe. In addition, thrombosis inside the device was categorized as either total or partial. The measurement of the angle of alignment between the disc and the lobe of ACP devices is illustrated in (A). (yellow arrows in B and C) ACP devices with total and partial lobe thrombosis, respectively. (arrows in D and E) Watchman devices with total and partial thrombosis, respectively. Figure 1 View largeDownload slide Analysis of device implantation on follow-up CT images. In patients with ACP/Amulet devices, appropriate implantation criteria comprised proper tire-shape of the lobe, complete separation of the disc and lobe, concavity of the disc, position of at least two-third of the lobe within the circumflex artery level, and the angle of alignment between the disc and the lobe. In addition, thrombosis inside the device was categorized as either total or partial. The measurement of the angle of alignment between the disc and the lobe of ACP devices is illustrated in (A). (yellow arrows in B and C) ACP devices with total and partial lobe thrombosis, respectively. (arrows in D and E) Watchman devices with total and partial thrombosis, respectively. Statistical analysis The Shapiro–Wilk test of normality and D’Agostino tests for skewness and kurtosis were used to assess whether quantitative data conformed to the normal distribution. Continuous variables are expressed as mean ± SD. Categorical variables are expressed as fraction (%). Continuous variables were compared using independent-sample parametric (unpaired Student’s t-test) or non-parametric tests (Mann–Whitney) depending on data normality. Categorical variables were compared using Fisher’s exact or χ2 tests, as appropriate. The reproducibility of the measurement of the disc-lobe angle was assessed by one observer measuring twice the angle in 20 randomly selected patients. The agreement was reported by calculating the intra-class correlation coefficient and the 95% limits of agreement. All statistical tests were two-tailed. A P-value <0.05 was considered to indicate statistical significance. Analyses were performed using NCSS 8 (NCSS Statistical Software, Kaysville, UT, USA). RESULTS Baseline characteristics and LAA occlusion procedures Patient characteristics at baseline are shown in Table 1. The population studied comprised 117 patients [age 74 ± 9, 43 (37%) women]. Mean CHA2DS2VASc score was 4.4 ± 1.3, and mean HASBLED score 3.5 ± 1.0. All had non-valvular AF with contra-indication for OAC therapy. Mean LVEF was 56 ± 10%, and 14 (12%) patients showed impaired LVEF, i.e. <50%. The mean X-ray exposure for each CT study was 3.8 ± 1.6 mSv. On pre-procedural CT the LA volume was 146 ± 50 mL. LAA shape was of chicken wing type in 61 (52%). Maximal landing zone diameter was 24.5 ± 4.7 mm. LAA occlusion procedures were successfully achieved with ACP/Amulet in 83 (71%) patients, and Watchman devices in 34 (29%) patients. Acute failure to implant the device occurred in none of the patients. The average number of devices used per case was 1.07 (range 1–3). We used a total of 125 devices in these 117 patients. During the procedure, the device had to be repositioned in 34% of cases to reach optimal positioning. Procedural complications occurred in 4 (4%) patients, consisting of two major bleeding at access site, one cardiac tamponade resolved after surgery, and one device embolization detected on TOE at 1-month follow-up. Table 1 Patient characteristics at baseline n = 117 Demographics Age (years) 74 ± 9 Female gender 43 (37) BMI (kg/m2) 24 ± 5 Medical history Paroxysmal AF 42 (36) Persistent/permanent AF 75 (64) History of stroke/TIA 64 (55) Coronary artery disease 44 (38) Diabetes Mellitus 39 (34) Hypertension 108 (92) CHA2DS2VASc score 4.4 ± 1.3 HASBLED score 3.5 ± 1.0 Indication for LAA closure History of major bleeding 104 (89) High-fall risk 6 (5) Other indication 7 (6) Pre-procedural imaging LVEF on TTE (%) 56 ± 10 LVEF<50% on TTE 14 (12) LA volume on CT (mL) 146 ± 50 Chicken wing LAA shape 61 (52) Maximum landing zone diameter on CT (mm) 24.5 ± 4.7 Implanted device ACP/Amulet 83 (71) Watchman 34 (29) n = 117 Demographics Age (years) 74 ± 9 Female gender 43 (37) BMI (kg/m2) 24 ± 5 Medical history Paroxysmal AF 42 (36) Persistent/permanent AF 75 (64) History of stroke/TIA 64 (55) Coronary artery disease 44 (38) Diabetes Mellitus 39 (34) Hypertension 108 (92) CHA2DS2VASc score 4.4 ± 1.3 HASBLED score 3.5 ± 1.0 Indication for LAA closure History of major bleeding 104 (89) High-fall risk 6 (5) Other indication 7 (6) Pre-procedural imaging LVEF on TTE (%) 56 ± 10 LVEF<50% on TTE 14 (12) LA volume on CT (mL) 146 ± 50 Chicken wing LAA shape 61 (52) Maximum landing zone diameter on CT (mm) 24.5 ± 4.7 Implanted device ACP/Amulet 83 (71) Watchman 34 (29) Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; AF, atrial fibrillation; BMI, body mass index; CT, computed tomography; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; OAC, oral anticoagulant; TIA, transient ischaemic attack; TTE, transthoracic echocardiography. Table 1 Patient characteristics at baseline n = 117 Demographics Age (years) 74 ± 9 Female gender 43 (37) BMI (kg/m2) 24 ± 5 Medical history Paroxysmal AF 42 (36) Persistent/permanent AF 75 (64) History of stroke/TIA 64 (55) Coronary artery disease 44 (38) Diabetes Mellitus 39 (34) Hypertension 108 (92) CHA2DS2VASc score 4.4 ± 1.3 HASBLED score 3.5 ± 1.0 Indication for LAA closure History of major bleeding 104 (89) High-fall risk 6 (5) Other indication 7 (6) Pre-procedural imaging LVEF on TTE (%) 56 ± 10 LVEF<50% on TTE 14 (12) LA volume on CT (mL) 146 ± 50 Chicken wing LAA shape 61 (52) Maximum landing zone diameter on CT (mm) 24.5 ± 4.7 Implanted device ACP/Amulet 83 (71) Watchman 34 (29) n = 117 Demographics Age (years) 74 ± 9 Female gender 43 (37) BMI (kg/m2) 24 ± 5 Medical history Paroxysmal AF 42 (36) Persistent/permanent AF 75 (64) History of stroke/TIA 64 (55) Coronary artery disease 44 (38) Diabetes Mellitus 39 (34) Hypertension 108 (92) CHA2DS2VASc score 4.4 ± 1.3 HASBLED score 3.5 ± 1.0 Indication for LAA closure History of major bleeding 104 (89) High-fall risk 6 (5) Other indication 7 (6) Pre-procedural imaging LVEF on TTE (%) 56 ± 10 LVEF<50% on TTE 14 (12) LA volume on CT (mL) 146 ± 50 Chicken wing LAA shape 61 (52) Maximum landing zone diameter on CT (mm) 24.5 ± 4.7 Implanted device ACP/Amulet 83 (71) Watchman 34 (29) Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; AF, atrial fibrillation; BMI, body mass index; CT, computed tomography; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; OAC, oral anticoagulant; TIA, transient ischaemic attack; TTE, transthoracic echocardiography. Device findings on follow-up CT CT findings at follow-up are summarized in Table 2. Follow-up CT was performed at 3 months in all patients, except for the one who was studied at 1 month after device embolization. LA volume was 149 ± 50 mL, with no significant change as compared to pre-procedural LA volume (P = 0.95). Among patients implanted with ACP/Amulet (n = 83), most fulfilled the criteria for proper implantation, with a tire-shaped device lobe in 74 (89%) , a separation between the disc and the lobe in 63 (76%), a concavity of the disc in 72 (87%), and a position of the lobe at least two-third beyond the CX in 73 (88%). The maximum angle between the disc and the lobe was 11 ± 9°, and was found >20° in 20 (24%) patients. The reproducibility of disc-lobe angle measurements was good (intra-class correlation coefficient = 0.96), with narrow 95% limits of agreement (−2.1 to +1.9°). Out of all patients, device lobe thrombosis was found to be total in 71 (61%) patients, with no significant difference between ACP/Amulet and watchman devices (P = 0.57). Table 2 Follow-up CT findings Total population (n = 117) ACP/Amulet population (n = 83) Watchman population (n = 34) P-value Delay to follow-up CT (months) 5 ± 2 5 ± 2 5 ± 2 0.43 LA volume (mL) 149 ± 50 145 ± 53 159 ± 50 0.12 ACP implantation criteria (n = 83) Tire-shape device lobe 74 (89) Separation of disc and lobe 63 (76) Concavity of the disc 72 (87) Position of the lobe >2/3 within CX 73 (88) Angle between disc and lobe (°) 11 ± 9 Angle between disc and lobe >20° 20 (24) Total device thrombosis 71 (61) 49 (59) 22 (65) 0.57 LAA patency on arterial phase 51 (44) 36 (43) 15 (44) 0.98 LAA patency on venous phase (n = 98) 67 (69) 46 (72) 21 (62) 0.31 Thrombus on atrial aspect of the device 19 (16) 16 (19) 3 (9) 0.17 Total population (n = 117) ACP/Amulet population (n = 83) Watchman population (n = 34) P-value Delay to follow-up CT (months) 5 ± 2 5 ± 2 5 ± 2 0.43 LA volume (mL) 149 ± 50 145 ± 53 159 ± 50 0.12 ACP implantation criteria (n = 83) Tire-shape device lobe 74 (89) Separation of disc and lobe 63 (76) Concavity of the disc 72 (87) Position of the lobe >2/3 within CX 73 (88) Angle between disc and lobe (°) 11 ± 9 Angle between disc and lobe >20° 20 (24) Total device thrombosis 71 (61) 49 (59) 22 (65) 0.57 LAA patency on arterial phase 51 (44) 36 (43) 15 (44) 0.98 LAA patency on venous phase (n = 98) 67 (69) 46 (72) 21 (62) 0.31 Thrombus on atrial aspect of the device 19 (16) 16 (19) 3 (9) 0.17 Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage. Table 2 Follow-up CT findings Total population (n = 117) ACP/Amulet population (n = 83) Watchman population (n = 34) P-value Delay to follow-up CT (months) 5 ± 2 5 ± 2 5 ± 2 0.43 LA volume (mL) 149 ± 50 145 ± 53 159 ± 50 0.12 ACP implantation criteria (n = 83) Tire-shape device lobe 74 (89) Separation of disc and lobe 63 (76) Concavity of the disc 72 (87) Position of the lobe >2/3 within CX 73 (88) Angle between disc and lobe (°) 11 ± 9 Angle between disc and lobe >20° 20 (24) Total device thrombosis 71 (61) 49 (59) 22 (65) 0.57 LAA patency on arterial phase 51 (44) 36 (43) 15 (44) 0.98 LAA patency on venous phase (n = 98) 67 (69) 46 (72) 21 (62) 0.31 Thrombus on atrial aspect of the device 19 (16) 16 (19) 3 (9) 0.17 Total population (n = 117) ACP/Amulet population (n = 83) Watchman population (n = 34) P-value Delay to follow-up CT (months) 5 ± 2 5 ± 2 5 ± 2 0.43 LA volume (mL) 149 ± 50 145 ± 53 159 ± 50 0.12 ACP implantation criteria (n = 83) Tire-shape device lobe 74 (89) Separation of disc and lobe 63 (76) Concavity of the disc 72 (87) Position of the lobe >2/3 within CX 73 (88) Angle between disc and lobe (°) 11 ± 9 Angle between disc and lobe >20° 20 (24) Total device thrombosis 71 (61) 49 (59) 22 (65) 0.57 LAA patency on arterial phase 51 (44) 36 (43) 15 (44) 0.98 LAA patency on venous phase (n = 98) 67 (69) 46 (72) 21 (62) 0.31 Thrombus on atrial aspect of the device 19 (16) 16 (19) 3 (9) 0.17 Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage. LAA patency on follow-up CT LAA enhancement suggestive of leakage/patency on arterial images was found in 51 (44%) patients, with no significant difference between ACP/Amulet and Watchman devices (P = 0.57). LAA density was 297 ± 80 HU in these patients, vs. 65 ± 18 HU in others (P < 0.001). Examples of LAA patency in patients implanted with ACP and Watchman devices are shown in Figure 2. The location and size of the leak neck could be assessed in 34 patients. In these patients, the most common leak location was postero-inferior (26/34, vs. five antero-superior and three postero-superior, P < 0.001). The maximum size of the leak neck was 15 ± 5 mm, and its surface covered 11.5 ± 3.5% of the total LAA orifice area including the device. Most of the leaks amenable to sizing and localization were found in patients with ACP/Amulet (33/34), because LAA patency in patients with Watchman devices were most commonly related to fabric leaks (14/15 leaks found in the Watchman population). Follow-up CT studies comprised additional imaging at the venous phase in 98/117 (84%) patients. LAA patency was more commonly found on venous phase images than on arterial phase images (69% vs. 44% on arterial phase, P < 0.001). LAA density on venous phase images was 179 ± 43 HU in these patients, vs. 68 ± 16 HU in others (P < 0.001). All patients with positive LAA enhancement on arterial phase images also showed positive enhancement on venous phase images. A total of 26 patients showed a positive LAA enhancement on venous phase images while arterial phase images were negative, as illustrated in Figure 3. This pattern was equally observed in ACP and Watchman groups (24% vs. 18%, P = 0.16). None of these 26 leaks could be localized or measured, as the neck was never visible. Figure 2 View largeDownload slide Peri-device leaks on arterial phase CT images. (A) LAA patency related to an antero-superior leak (arrow in A) in a patient implanted with an ACP device. The measurement of the maximum dimension of the leak neck in the same patient is illustrated in (B). (C) LAA patency related to a fabric leak (arrows in C) in a patient with Watchman device and incomplete device thrombosis. No leak neck was found on device margins. Figure 2 View largeDownload slide Peri-device leaks on arterial phase CT images. (A) LAA patency related to an antero-superior leak (arrow in A) in a patient implanted with an ACP device. The measurement of the maximum dimension of the leak neck in the same patient is illustrated in (B). (C) LAA patency related to a fabric leak (arrows in C) in a patient with Watchman device and incomplete device thrombosis. No leak neck was found on device margins. Figure 3 View largeDownload slide Use of venous phase CT imaging to detect LAA patency. (A) An arterial phase image in a patient implanted with an ACP device. Lobe thrombosis is complete and measurements of attenuation coefficients within the LA and the LAA do not suggest LAA patency. (B) The corresponding CT image acquired at the venous phase. Lobe thrombosis seems complete as measurements of attenuation coefficients within the LA and the LAA do not suggest LAA patency. Figure 3 View largeDownload slide Use of venous phase CT imaging to detect LAA patency. (A) An arterial phase image in a patient implanted with an ACP device. Lobe thrombosis is complete and measurements of attenuation coefficients within the LA and the LAA do not suggest LAA patency. (B) The corresponding CT image acquired at the venous phase. Lobe thrombosis seems complete as measurements of attenuation coefficients within the LA and the LAA do not suggest LAA patency. Device-related thrombus on follow-up CT In the total population, an enhancement defect suggestive of thrombus was found on the atrial side of the device in 19 (16%) patients, with no significant difference between ACP/Amulet and Watchman populations (16/83 vs. 3/34, P = 0.17). Seventeen of these patients were also scanned at the venous phase, and all defects were still visible at this phase. The maximum dimension of the defect was 19 ± 6 mm, and the most common location was antero-superior (10/18 vs. four antero-inferior, three postero-superior, and two postero-inferior, P < 0.001). Most of the defects were laminated, except for one that was protruding in the left atrium. TOE was performed in all 19 patients showing enhancement defects on CT. Definite organized thrombus was retained in 5/19. On TOE, these consisted of one large protruding thrombus of 25 mm maximal diameter, and four laminated thrombi of 2.2 ± 0.5 mm mean thickness. OAC was introduced in all five patients and all thrombi resolved after therapy. In 14/19 TOE showed thin laminated defects of less than 1 mm thickness. Because this semiology could also indicate prominent endothelialisation, and in order to mitigate the bleeding risk in this specific population, OAC was not introduced based on such TOE findings. Examples of peri-device enhancement defects on CT after LAA occlusion are shown in Figure 4. A correlation between CT and TOE findings is illustrated in Figure 5. Figure 4 View largeDownload slide Examples of device-related thrombi after LAA occlusion. Enhancement defects suggestive of thrombus on the atrial aspect of the device are shown in eight patients (yellow arrows). (A, B, D, and F–H) ACP-related thrombi while in (C and E) Watchman device-related thrombi. Most of the thrombi are laminated except for the one in (H), which protrudes into the LA chamber. None of these patients experienced stroke during follow-up. Figure 4 View largeDownload slide Examples of device-related thrombi after LAA occlusion. Enhancement defects suggestive of thrombus on the atrial aspect of the device are shown in eight patients (yellow arrows). (A, B, D, and F–H) ACP-related thrombi while in (C and E) Watchman device-related thrombi. Most of the thrombi are laminated except for the one in (H), which protrudes into the LA chamber. None of these patients experienced stroke during follow-up. Figure 5 View largeDownload slide TOE correlates of peri-device enhancement defects on CT after LAA occlusion. (A–C) A 72-year-old man studied at CT and TOE 3 months after LAA occlusion. CT showed an anterior laminated peri-device enhancement defect (arrows in A and B) suggestive of laminated thrombus. TOE showed mild endocardial thickening in the area (arrow in C). This TOE appearance did not discriminate between laminated thrombus and locally prominent endothelialisation. In order to mitigate bleeding risk, OAC was not introduced. (D–F) A 78-year-old woman studied at CT and TOE 3 months after LAA occlusion. CT shows a large enhancement defect protruding within the LA chamber (arrows in D and E), consistent with device-related thrombus. TOE confirmed the presence of a sessile thrombus (arrow in F). OAC was introduced, leading to complete thrombus resorption without embolic complication. Figure 5 View largeDownload slide TOE correlates of peri-device enhancement defects on CT after LAA occlusion. (A–C) A 72-year-old man studied at CT and TOE 3 months after LAA occlusion. CT showed an anterior laminated peri-device enhancement defect (arrows in A and B) suggestive of laminated thrombus. TOE showed mild endocardial thickening in the area (arrow in C). This TOE appearance did not discriminate between laminated thrombus and locally prominent endothelialisation. In order to mitigate bleeding risk, OAC was not introduced. (D–F) A 78-year-old woman studied at CT and TOE 3 months after LAA occlusion. CT shows a large enhancement defect protruding within the LA chamber (arrows in D and E), consistent with device-related thrombus. TOE confirmed the presence of a sessile thrombus (arrow in F). OAC was introduced, leading to complete thrombus resorption without embolic complication. Findings on second follow-up CT An additional follow-up CT was performed in 23 patients. The minimum delay between first and second follow-up CT was 3 months, and the median delay between device closure and second follow-up CT was 10 months. Leaks were found in 18/23 patients on the first follow-up CT, and in 17/23 patients on the second follow-up CT (leak had resolved in one patient, and no novel leak had appeared). Enhancement defects suggestive of thrombus were found on the atrial aspect of the device in 5/23 patients on the first follow-up CT, and in 4/23 patients on the second follow-up CT. Between the first and the second follow-up CT, the defects had resolved in three patients, and a de novo defect had appeared in two. The two de novo defects had appeared between 3 and 6 months and between 4 and 7 months, respectively. Correlates of LAA patency The correlates of LAA patency are analysed in Table 3. The prevalence of leaks on arterial phase images did not relate to age, gender, delay to follow-up CT, or type of device. On pre-procedural CT, patients with leaks showed higher LA volume (157 ± 56 vs. 138 ± 43 mL, P = 0.04), larger maximal landing zone diameter (25.9 ± 5.1 vs. 23.3 ± 4.0, P = 0.003), and less likely chicken wing-shaped LAA (38% vs. 63% in patients with no leaks, P = 0.01). Patients with leaks at follow-up also showed slightly lower LVEF on baseline TTE (54 ± 9 vs. 58 ± 9%, P = 0.02). On follow-up CT, patients with leaks showed less likely total device thrombosis (23% vs. 91% in those with no leak, P < 0.001). Among the criteria for correct ACP/Amulet implantation, leaks were not found to relate to the tire shape of the lobe, separation of disc and lobe, concavity of the disc, or position of the lobe two-third within CX artery. However, patients with leaks showed a higher angle between the disc and the lobe (14 ± 8 vs. 9 ± 8°, P = 0.02). Last, although enhancement defects on the atrial aspect of the device were more common in patients with no leaks, the relationship did not reach statistical significance (22% peri-device defects in patients with no leaks vs. 9% in those with leaks, P = 0.08). Table 3 Correlates of LAA patency LAA patency (n = 52)a No LAA patency (n = 65)a P-value Patient characteristics Age (years) 73 ± 9 74 ± 9 0.46 Female gender 15 (29) 28 (43) 0.11 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 29 (56) 35 (54) 0.86 CHA2DS2VASc score 4.3 ± 1.3 4.4 ± 1.4 0.81 Pre-procedural imaging LVEF on TTE (%) 54 ± 9 58 ± 9 0.02 LA volume on CT (mL) 157 ± 56 138 ± 43 0.04 Chicken wing LAA shape 20 (38) 41 (63) 0.01 Maximum landing zone diameter on CT (mm) 25.9 ± 5.1 23.3 ± 4.0 0.003 Implanted device ACP/Amulet 36 (69) 47 (72) 0.66 Watchman 16 (31) 18 (28) 0.66 Delay to follow-up CT (months) 4.7 ± 2.0 4.7 ± 2.3 0.98 ACP implantation criteria (n = 83) Tire-shape device lobe 33 (92) 41 (87) 0.42 Separation of disc and lobe 31 (86) 32 (68) 0.08 Concavity of the disc 32 (89) 40 (85) 0.42 Position of the lobe >2/3 within CX 30 (83) 43 (91) 0.41 Angle between disc and lobe (°) 14 ± 8 9 ± 8 0.02 Total device thrombosis 12 (23) 59 (91) <0.001 Thrombus on atrial aspect of the device 5 (9) 14 (22) 0.08 LAA patency (n = 52)a No LAA patency (n = 65)a P-value Patient characteristics Age (years) 73 ± 9 74 ± 9 0.46 Female gender 15 (29) 28 (43) 0.11 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 29 (56) 35 (54) 0.86 CHA2DS2VASc score 4.3 ± 1.3 4.4 ± 1.4 0.81 Pre-procedural imaging LVEF on TTE (%) 54 ± 9 58 ± 9 0.02 LA volume on CT (mL) 157 ± 56 138 ± 43 0.04 Chicken wing LAA shape 20 (38) 41 (63) 0.01 Maximum landing zone diameter on CT (mm) 25.9 ± 5.1 23.3 ± 4.0 0.003 Implanted device ACP/Amulet 36 (69) 47 (72) 0.66 Watchman 16 (31) 18 (28) 0.66 Delay to follow-up CT (months) 4.7 ± 2.0 4.7 ± 2.3 0.98 ACP implantation criteria (n = 83) Tire-shape device lobe 33 (92) 41 (87) 0.42 Separation of disc and lobe 31 (86) 32 (68) 0.08 Concavity of the disc 32 (89) 40 (85) 0.42 Position of the lobe >2/3 within CX 30 (83) 43 (91) 0.41 Angle between disc and lobe (°) 14 ± 8 9 ± 8 0.02 Total device thrombosis 12 (23) 59 (91) <0.001 Thrombus on atrial aspect of the device 5 (9) 14 (22) 0.08 Values are represented using the mean ± SD and n (%). a Refers to LAA patency on arterial phase CT images. ACP, Amplatzer cardiac plug; BMI, body mass index; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; TTE, transthoracic echocardiography. Table 3 Correlates of LAA patency LAA patency (n = 52)a No LAA patency (n = 65)a P-value Patient characteristics Age (years) 73 ± 9 74 ± 9 0.46 Female gender 15 (29) 28 (43) 0.11 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 29 (56) 35 (54) 0.86 CHA2DS2VASc score 4.3 ± 1.3 4.4 ± 1.4 0.81 Pre-procedural imaging LVEF on TTE (%) 54 ± 9 58 ± 9 0.02 LA volume on CT (mL) 157 ± 56 138 ± 43 0.04 Chicken wing LAA shape 20 (38) 41 (63) 0.01 Maximum landing zone diameter on CT (mm) 25.9 ± 5.1 23.3 ± 4.0 0.003 Implanted device ACP/Amulet 36 (69) 47 (72) 0.66 Watchman 16 (31) 18 (28) 0.66 Delay to follow-up CT (months) 4.7 ± 2.0 4.7 ± 2.3 0.98 ACP implantation criteria (n = 83) Tire-shape device lobe 33 (92) 41 (87) 0.42 Separation of disc and lobe 31 (86) 32 (68) 0.08 Concavity of the disc 32 (89) 40 (85) 0.42 Position of the lobe >2/3 within CX 30 (83) 43 (91) 0.41 Angle between disc and lobe (°) 14 ± 8 9 ± 8 0.02 Total device thrombosis 12 (23) 59 (91) <0.001 Thrombus on atrial aspect of the device 5 (9) 14 (22) 0.08 LAA patency (n = 52)a No LAA patency (n = 65)a P-value Patient characteristics Age (years) 73 ± 9 74 ± 9 0.46 Female gender 15 (29) 28 (43) 0.11 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 29 (56) 35 (54) 0.86 CHA2DS2VASc score 4.3 ± 1.3 4.4 ± 1.4 0.81 Pre-procedural imaging LVEF on TTE (%) 54 ± 9 58 ± 9 0.02 LA volume on CT (mL) 157 ± 56 138 ± 43 0.04 Chicken wing LAA shape 20 (38) 41 (63) 0.01 Maximum landing zone diameter on CT (mm) 25.9 ± 5.1 23.3 ± 4.0 0.003 Implanted device ACP/Amulet 36 (69) 47 (72) 0.66 Watchman 16 (31) 18 (28) 0.66 Delay to follow-up CT (months) 4.7 ± 2.0 4.7 ± 2.3 0.98 ACP implantation criteria (n = 83) Tire-shape device lobe 33 (92) 41 (87) 0.42 Separation of disc and lobe 31 (86) 32 (68) 0.08 Concavity of the disc 32 (89) 40 (85) 0.42 Position of the lobe >2/3 within CX 30 (83) 43 (91) 0.41 Angle between disc and lobe (°) 14 ± 8 9 ± 8 0.02 Total device thrombosis 12 (23) 59 (91) <0.001 Thrombus on atrial aspect of the device 5 (9) 14 (22) 0.08 Values are represented using the mean ± SD and n (%). a Refers to LAA patency on arterial phase CT images. ACP, Amplatzer cardiac plug; BMI, body mass index; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; TTE, transthoracic echocardiography. Correlates of DRT The correlates of DRT are analysed in Table 4. The prevalence of peri-device enhancement defects did not relate to age, gender, delay to follow-up CT, type of device, nor to baseline LVEF, LA volume, landing zone diameter, or chicken wing-shaped LAA on pre-procedural CT. On follow-up CT, patients with peri-device defects showed more likely total device lobe thrombosis (84% vs. 56% in those with no thrombus, P = 0.02). Among the criteria for correct ACP implantation, enhancement defects did not relate to the tire shape of the lobe, separation of disc and lobe, concavity of the disc, position of the lobe two-third within CX artery, or angle between the disc and the lobe. Table 4 Correlates of device-related thrombus Thrombus (n = 19) No thrombus (n = 98) P-value Patient characteristics Age (years) 74 ± 11 74 ± 9 0.94 Female gender 5 (26) 38 (39) 0.31 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 10/19 (53) 54/98 (55) 0.78 CHA2DS2VASc score 4.3 ± 1.5 4.4 ± 1.3 0.94 Pre-procedural imaging LVEF on TTE (%) 57 ± 12 56 ± 9 0.86 LA volume on CT (mL) 142 ± 66 147 ± 46 0.68 Chicken wing LAA shape 7 (37) 54 (55) 0.14 Maximum landing zone diameter on CT (mm) 23.5 ± 6.4 24.7 ± 4.3 0.32 Implanted device ACP/Amulet 16 (84) 67 (68) 0.17 Watchman 3 (16) 31 (32) 0.17 Delay to follow-up CT (months) 4.4 ± 2.1 4.8 ± 2.2 0.50 ACP implantation criteria (n = 83) Tire-shape device lobe 14 (88) 60 (90) 0.55 Separation of disc and lobe 10 (63) 53 (79) 0.17 Concavity of the disc 14 (88) 58 (87) 0.57 Position of the lobe > 2/3 within CX 14 (88) 59 (88) 0.83 Angle between disc and lobe (°) 8.8 ± 8.7 12.0 ± 9.1 0.21 Total device thrombosis 16 (84) 55 (56) 0.02 LAA patency on arterial phase 5 (26) 47 (48) 0.08 LAA patency on venous phase (n = 98) 9 (56) 59 (72) 0.24 Thrombus (n = 19) No thrombus (n = 98) P-value Patient characteristics Age (years) 74 ± 11 74 ± 9 0.94 Female gender 5 (26) 38 (39) 0.31 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 10/19 (53) 54/98 (55) 0.78 CHA2DS2VASc score 4.3 ± 1.5 4.4 ± 1.3 0.94 Pre-procedural imaging LVEF on TTE (%) 57 ± 12 56 ± 9 0.86 LA volume on CT (mL) 142 ± 66 147 ± 46 0.68 Chicken wing LAA shape 7 (37) 54 (55) 0.14 Maximum landing zone diameter on CT (mm) 23.5 ± 6.4 24.7 ± 4.3 0.32 Implanted device ACP/Amulet 16 (84) 67 (68) 0.17 Watchman 3 (16) 31 (32) 0.17 Delay to follow-up CT (months) 4.4 ± 2.1 4.8 ± 2.2 0.50 ACP implantation criteria (n = 83) Tire-shape device lobe 14 (88) 60 (90) 0.55 Separation of disc and lobe 10 (63) 53 (79) 0.17 Concavity of the disc 14 (88) 58 (87) 0.57 Position of the lobe > 2/3 within CX 14 (88) 59 (88) 0.83 Angle between disc and lobe (°) 8.8 ± 8.7 12.0 ± 9.1 0.21 Total device thrombosis 16 (84) 55 (56) 0.02 LAA patency on arterial phase 5 (26) 47 (48) 0.08 LAA patency on venous phase (n = 98) 9 (56) 59 (72) 0.24 Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; BMI, body mass index; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; TTE, transthoracic echocardiography. Table 4 Correlates of device-related thrombus Thrombus (n = 19) No thrombus (n = 98) P-value Patient characteristics Age (years) 74 ± 11 74 ± 9 0.94 Female gender 5 (26) 38 (39) 0.31 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 10/19 (53) 54/98 (55) 0.78 CHA2DS2VASc score 4.3 ± 1.5 4.4 ± 1.3 0.94 Pre-procedural imaging LVEF on TTE (%) 57 ± 12 56 ± 9 0.86 LA volume on CT (mL) 142 ± 66 147 ± 46 0.68 Chicken wing LAA shape 7 (37) 54 (55) 0.14 Maximum landing zone diameter on CT (mm) 23.5 ± 6.4 24.7 ± 4.3 0.32 Implanted device ACP/Amulet 16 (84) 67 (68) 0.17 Watchman 3 (16) 31 (32) 0.17 Delay to follow-up CT (months) 4.4 ± 2.1 4.8 ± 2.2 0.50 ACP implantation criteria (n = 83) Tire-shape device lobe 14 (88) 60 (90) 0.55 Separation of disc and lobe 10 (63) 53 (79) 0.17 Concavity of the disc 14 (88) 58 (87) 0.57 Position of the lobe > 2/3 within CX 14 (88) 59 (88) 0.83 Angle between disc and lobe (°) 8.8 ± 8.7 12.0 ± 9.1 0.21 Total device thrombosis 16 (84) 55 (56) 0.02 LAA patency on arterial phase 5 (26) 47 (48) 0.08 LAA patency on venous phase (n = 98) 9 (56) 59 (72) 0.24 Thrombus (n = 19) No thrombus (n = 98) P-value Patient characteristics Age (years) 74 ± 11 74 ± 9 0.94 Female gender 5 (26) 38 (39) 0.31 BMI (kg/m2) 24 ± 4 24 ± 5 0.91 History of stroke/TIA 10/19 (53) 54/98 (55) 0.78 CHA2DS2VASc score 4.3 ± 1.5 4.4 ± 1.3 0.94 Pre-procedural imaging LVEF on TTE (%) 57 ± 12 56 ± 9 0.86 LA volume on CT (mL) 142 ± 66 147 ± 46 0.68 Chicken wing LAA shape 7 (37) 54 (55) 0.14 Maximum landing zone diameter on CT (mm) 23.5 ± 6.4 24.7 ± 4.3 0.32 Implanted device ACP/Amulet 16 (84) 67 (68) 0.17 Watchman 3 (16) 31 (32) 0.17 Delay to follow-up CT (months) 4.4 ± 2.1 4.8 ± 2.2 0.50 ACP implantation criteria (n = 83) Tire-shape device lobe 14 (88) 60 (90) 0.55 Separation of disc and lobe 10 (63) 53 (79) 0.17 Concavity of the disc 14 (88) 58 (87) 0.57 Position of the lobe > 2/3 within CX 14 (88) 59 (88) 0.83 Angle between disc and lobe (°) 8.8 ± 8.7 12.0 ± 9.1 0.21 Total device thrombosis 16 (84) 55 (56) 0.02 LAA patency on arterial phase 5 (26) 47 (48) 0.08 LAA patency on venous phase (n = 98) 9 (56) 59 (72) 0.24 Values are represented using the mean ± SD and n (%). ACP, Amplatzer cardiac plug; BMI, body mass index; CT, computed tomography; CX, circumflex artery; LAA, left atrial appendage; LVEF, left ventricular ejection fraction; TTE, transthoracic echocardiography. Patient outcomes Over a median follow-up of 13 months (Q1–Q3: 12–13 months), death occurred in four patients, related to post-operative complications after device embolization in one, heart failure in one, and non-cardiac causes in two. Stroke occurred in five patients and symptoms suggestive of transient ischaemic attack (TIA) in three. The median delay between device implant and stroke signs was 6 months. None of the eight patients with stroke or suspected TIA showed thrombus on the atrial side of the device at follow-up CT, and the prevalence of device leaks was similar to that of the remaining population (50% vs 43%, P = 0.72). All eight patients were implanted with ACP/Amulet devices, although the device type was not found significantly related to stroke/TIA occurrence (P = 0.06). Besides stroke and TIA, none of the patients in the whole studied population showed clinical signs of extra-cerebral systemic emboli. Discussion This study is to our knowledge the largest report on post-procedural CT findings in patients undergoing LAA occlusion. Its main results are that LAA patency is extremely common after LAA occlusion, detected in the majority of patients, the most common leak location being postero-inferior. Leaks are more likely found in patients with LA dilatation and LVEF impairment, non-chicken wing LAA shape, large landing zone diameter, incomplete device lobe thrombosis, as well as disc/lobe misalignment in patients with ACP/Amulet. Follow-up CT also detects a substantial number of peri-device defects suggestive of thrombus on the atrial aspect of the device, most being laminated and located on the antero-superior aspect of the LAA orifice. Such defects do not relate to clinical or imaging characteristics nor to implantation criteria, but are much more common in patients with total thrombosis within the device, hence a population different than the one presenting with device leaks. Population and procedures This study involved 117 patients whose clinical characteristics are consistent with the population usually referred for LAA occlusion.7,8 The sizing and deployment of occluding devices followed the standardized protocol recommended by the manufacturers. The rates of procedure-related complications and stroke during follow-up were consistent with prior reports.7,8 The only specificity of our patient management protocol was the use of single antiplatelet therapy after LAA occlusion, which is the current consensus applied in our institution for these patients at high-bleeding risk.18 The TOE and CT protocols for pre-procedural and follow-up assessment applied state-of-the-art methods that are consistent with prior reports, except for the addition of venous phase CT images at follow-up, which was implemented to improve the sensitivity of the method in detecting LAA patency. This approach is similar to the one previously reported for the detection of LAA thrombus, as false-positive defects of LAA enhancement are quite common on arterial phase images for haemodynamic reasons, particularly in AF patients.17 The management of patients with leaks and device-related thrombus conformed to the one suggested by other authors,7,10 patients being followed with TOE and OAC being re-introduced in case of thrombus or peri-device leak >5 mm on TOE. Peri-device leaks The detection of LAA patency was based on measurements of attenuation coefficients within the LA and the LAA. The thresholds used on arterial phase images were identical to those proposed in a recent study,13 and a similar approach was adapted to the specific contrast of venous phase images. Our results confirm that LAA patency is extremely common on arterial phase CT images (44%), this rate being higher than the one reported by large studies using TOE.19 Moreover, this prevalence is largely underestimated on arterial phase images, as LAA patency becomes present in the vast majority of patients on venous phase images (69%). One explanation for a higher sensitivity to detect LAA patency at CT is that unlike TOE, the method does not require direct visualization of the leak jet. Indeed, LAA patency is demonstrated by an LAA contrast uptake that can be the consequence of leaks that are inaccessible to TOE (sub-millimetric marginal leaks, trans-fabric leaks, and defects of endothelialisation). Of note, the size of the leaks reported in the present study (mean of 15 mm or 11.5% of LAA orifice area) only refers to the larger leaks because non-visible leak necks were obviously not measured. In addition, this size cannot be compared to the sizing of leak jets on TOE as the measurement methods differ. Interestingly, LAA patency does not seem to be related to the delay after LAA occlusion since most of the leaks were found to be stable between the first and the second post-procedural CT studies in patients who were scanned twice during follow-up. Our results indicate that device leaks are partly explained by patient-related factors, including baseline LA dilatation, LVEF impairment, and large diameter at landing zone. These three predictors are common characteristics found in patients with severely remodelled atria, the enlarged anatomy possibly making device stability more complicated to achieve. In addition, device leaks are less common when the LAA is of chicken wing shape, which might be explained by a rather tubular than conical landing zone geometry in these patients, facilitating device stability. Besides patient characteristics, device leaks are also partly explained by device-related factors. In patients implanted with ACP/Amulet device, disc-lobe misalignment seems to be the dominant cause, which is consistent with a past study.13 In addition, patients with persistent LAA patency are much more likely to show incomplete thrombosis of the device lobe, suggesting that leaks occur in patients with a lower ability to generate thrombus. In patients implanted with ACP/Amulet, the most common location of peri-device leaks is the postero-inferior region of LAA orifice, which can be explained by the complex anatomy of this area. Indeed, the stability of the device and its proper impaction into the LA wall is likely to be more challenging on the ridge between the LAA and the left inferior pulmonary vein. In patients implanted with Watchman devices, most of the leaks did not occur on the device margins but rather through the fabric, hence directly related to the absence of complete thrombosis within the device. DRT The present study shows that enhancement defects highly suggestive of thrombus are quite commonly found on CT after LAA occlusion on the atrial aspect of the device. The prevalence of such thrombi (16%) is higher than the one usually reported.20 One may argue that the high prevalence of thrombus is related to the use of single antiplatelet therapy. However, a recent study demonstrated the safety and efficacy of single antiplatelet therapy,18 with rates of device-related thrombus similar to the ones reported in the Percutaneous Closure of the Left Atrial Appendage Versus Warfarin Therapy for Prevention of Stroke in Patients With Atrial Fibrillation (PROTECT) and ASA Plavix Feasibility Study With Watchman Left Atrial Appendage Closure Technology (ASAP) trials.7,21 In addition, prior TOE studies focusing on device-related thrombus have also reported high rates of thrombus despite dual antiplatelet therapy.11,22 However, the exact nature of the enhancement defects found on CT remains equivocal. In most cases TOE findings were not able to discriminate between definite organized thrombus and locally prominent endothelialisation. Interestingly, in the sub-population that underwent a second follow-up CT, 3/5 defects had spontaneously resolved, and two de novo defects had appeared months after the procedure. Such time course would be quite unlikely for an endothelialisation process. Nevertheless, our findings outline the need for future biological research aiming at clarifying the formation of thrombus and endothelial tissue on the atrial aspect of the device after LAA occlusion. As for clinical correlates, our results indicate that CT-defined DRT are poorly predicted by patient baseline characteristics or device implantation criteria. The only imaging finding statistically related to thrombus on the atrial aspect of the device is complete thrombosis within the device. Thus, DRT and peri-device leaks seem to occur in quite different populations. This suggests a potential biological substrate explaining such differences in the response profile after LAA occlusion. Such heterogeneous response to LAA occlusion should be clarified in future studies characterizing homeostatic, fibrinolytic, and endothelial factors of thrombogenicity in chronic AF. Clinical implications In the absence clinical results relating device leaks and DRT to stroke risk,12 the management of patients with such imaging findings remains empirical, based on systematic OAC therapy in patients with thrombus, as well as in those with large leaks.7,10 In the present study, the number of events during follow-up is obviously too limited to draw any conclusion. However, it is interesting to note that none of the eight patients who experienced stroke or suspected TIA during follow-up showed a device-related thrombus, and that the prevalence of leaks in these eight patients was similar to that of the remaining population. Except for one large thrombus protruding into the LA chamber, most of the documented thrombi were laminated, hence at lower embolic risk. As for device leaks, our findings suggest that it may not expose to increased stroke risk since these patients seem to have a lower ability to generate the thrombus necessary for complete LAA occlusion. Last, regarding device types, the present study is the first able to compare with sufficient sample sizes CT imaging outcomes after LAA occlusion between ACP/Amulet and Watchman devices. Interestingly, both types show similar numbers of leaks and similar rates of thrombosis, both within the device and on its atrial aspect. Study limitations The main limitation of this study is its sample size, which particularly prevented us from analysing potential imaging predictors of stroke after LAA occlusion. Therefore, the impact of the reported image features on patient outcome remains uncertain, and their clinical management still empirical. Another important limitation is the absence of biological characterization of thrombogenicity. Indeed, biological factors may play a major role since the response profile after LAA occlusion appears to be highly heterogeneous and quite poorly predicted by clinical or imaging characteristics. The agreement between CT and TOE for the assessment of peri-device leaks and DRT cannot be assessed in the present study because only the patients with positive CT findings underwent TOE. However, the aim of the study was to assess the prevalence, size, and location of leaks and DRT on cardiac CT after percutaneous LAA occlusion, and to identify patient-related and device-related correlates. TOE data is provided to illustrate how abnormal CT findings relate to TOE semiology, but the agreement between both methods was out of the scope of the present study. Last, our results indicate that venous phase imaging is more sensitive than arterial phase imaging for the detection of LAA patency after LAA closure, and we can thus anticipate that some leaks were missed in the minority of patients that were only scanned at arterial phase (16% of the population). Conclusion LAA patency is extremely common on CT after LAA occlusion, detected in the majority of patients, the most common leak location being postero-inferior. Leaks are more likely found in patients with LA dilatation and LVEF impairment, non-chicken wing LAA shape, large landing zone diameter, incomplete device thrombosis, as well as disc/lobe misalignment in patients with ACP/Amulet devices. CT signs suggestive of device-related thrombus are also quite common, most thrombi being laminated and located on the antero-superior aspect of the LAA orifice. These do not relate to clinical or imaging characteristics nor to implantation criteria, but are much more common in patients with total thrombosis within the device, hence a population different than the one presenting with peri-device leaks. Further studies are desirable to relate such image findings to biological factors of thrombogenicity and risk of stroke. Funding The research leading to these results has received funding from Agence Nationale de la Recherche (Grants ANR-10-IAHU-04; ANR-11-EQPX-0030), and from the European Research Council (Grant Agreement ERC n°715093). Conflict of interest: None declared. References 1 Camm AJ , Lip GY , De Caterina R , Savelieva I , Atar D , Hohnloser SH et al. 2012 focused update of the esc guidelines for the management of atrial fibrillation: an update of the 2010 esc guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association . Eur Heart J 2012 ; 33 : 2719 – 47 . Google Scholar Crossref Search ADS PubMed 2 Go AS , Hylek EM , Phillips KA , Chang Y , Henault LE , Selby JV et al. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the anticoagulation and risk factors in atrial fibrillation (ATRIA) Study . JAMA 2001 ; 285 : 2370 – 5 . Google Scholar Crossref Search ADS PubMed 3 Hart RG , Pearce LA , Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation . Ann Intern Med 2007 ; 146 : 857 – 67 . Google Scholar Crossref Search ADS PubMed 4 Fuster V , Rydén LE , Cannom DS , Crijns HJ , Curtis AB , Ellenbogen KA et al. ACC/AHA/ESC 2006 guidelines for the management of patients with atrial fibrillation—executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines and the European Society of Cardiology Committee for Practice Guidelines (writing committee to revise the 2001 guidelines for the management of patients with atrial fibrillation) . Eur Heart J 2006 ; 27 : 1979 – 2030 . Google Scholar Crossref Search ADS PubMed 5 Onalan O , Lashevsky I , Hamad A , Crystal E. Nonpharmacologic stroke prevention in atrial fibrillation . Expert Rev Cardiovasc Ther 2005 ; 3 : 619 – 33 . Google Scholar Crossref Search ADS PubMed 6 Blackshear JL , Odell JA. Appendage obliteration to reduce stroke in cardiac surgical patients with atrial fibrillation . Ann Thorac Surg 1996 ; 61 : 755 – 9 . Google Scholar Crossref Search ADS PubMed 7 Reddy VY , Sievert H , Halperin J , Doshi SK , Buchbinder M , Neuzil P et al. Percutaneous left atrial appendage closure vs warfarin for atrial fibrillation: a randomized clinical trial . JAMA 2014 ; 312 : 1988 – 98 . Google Scholar Crossref Search ADS PubMed 8 Tzikas A , Shakir S , Gafoor S , Omran H , Berti S , Santoro G et al. Left atrial appendage occlusion for stroke prevention in atrial fibrillation: multicentre experience with the AMPLATZER Cardiac Plug . EuroIntervention 2016 ; 11 : 1170 – 9 . Google Scholar Crossref Search ADS PubMed 9 Meier B , Blaauw Y , Khattab AA , Lewalter T , Sievert H , Tondo C et al. EHRA/EAPCI expert consensus statement on catheter-based left atrial appendage occlusion . Europace 2014 ; 16 : 1397 – 416 . Google Scholar Crossref Search ADS PubMed 10 Holmes DR , Reddy VY , Turi ZG , Doshi SK , Sievert H , Buchbinder M et al. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial . Lancet 2009 ; 374 : 534 – 42 . Google Scholar Crossref Search ADS PubMed 11 Sedaghat A , Schrickel JW , Andrié R , Schueler R , Nickenig G , Hammerstingl C. Thrombus formation after left atrial appendage occlusion with the Amplatzer Amulet device . JACC Clin Electrophysiol 2017 ; 3 : 71 – 75 . Google Scholar Crossref Search ADS PubMed 12 Saw J , Tzikas A , Shakir S , Gafoor S , Omran H , Nielsen-Kudsk JE et al. Incidence and clinical impact of device-associated thrombus and peri-device leak following left atrial appendage closure with the Amplatzer Cardiac Plug . JACC Cardiovasc Interv 2017 ; 10 : 391 – 9 . Google Scholar Crossref Search ADS PubMed 13 Saw J , Fahmy P , DeJong P , Lempereur M , Spencer R , Tsang M et al. Cardiac CT angiography for device surveillance after endovascular left atrial appendage closure . Eur Heart J Cardiovasc Imaging 2015 ; 16 : 1198 – 206 . Google Scholar Crossref Search ADS PubMed 14 Clemente A , Avogliero F , Berti S , Paradossi U , Jamagidze G , Rezzaghi M et al. Multimodality imaging in preoperative assessment of left atrial appendage transcatheter occlusion with the Amplatzer Cardiac Plug . Eur Heart J Cardiovasc Imaging 2015 ; 16 : 1276 – 87 . Google Scholar Crossref Search ADS PubMed 15 Jaguszewski M , Manes C , Puippe G , Salzberg S , Müller M , Falk V et al. Cardiac CT and echocardiographic evaluation of peri-device flow after percutaneous left atrial appendage closure using the AMPLATZER cardiac plug device . Cathet Cardiovasc Intervent 2015 ; 85 : 306 – 12 . Google Scholar Crossref Search ADS 16 Jalal Z , Dinet ML , Combes N , Pillois X , Renou P , Sibon I et al. Percutaneous left atrial appendage closure followed by single antiplatelet therapy: short- and mid-term outcomes . Arch Cardiovasc Dis 2017 ; 110 : 242 – 9 . Google Scholar Crossref Search ADS PubMed 17 Romero J , Husain SA , Kelesidis I , Sanz J , Medina HM , Garcia MJ. Detection of left atrial appendage thrombus by cardiac computed tomography in patients with atrial fibrillation: a meta-analysis . Circ Cardiovasc Imaging 2013 ; 6 : 185 – 94 . Google Scholar Crossref Search ADS PubMed 18 Rodriguez-Gabella T , Nombela-Franco L , Regueiro A , Jiménez-Quevedo P , Champagne J , O'Hara G et al. Single antiplatelet therapy following left atrial appendage closure in patients with contraindication to anticoagulation . J Am Coll Cardiol 2016 ; 68 : 1920 – 1 . Google Scholar Crossref Search ADS PubMed 19 Viles-Gonzalez JF , Kar S , Douglas P , Dukkipati S , Feldman T , Horton R et al. The clinical impact of incomplete left atrial appendage closure with the Watchman device in patients with atrial fibrillation: a PROTECT AF (Percutaneous Closure of the Left Atrial Appendage Versus Warfarin Therapy for Prevention of Stroke in Patients With Atrial Fibrillation) substudy . J Am Coll Cardiol 2012 ; 59 : 923 – 9 . Google Scholar Crossref Search ADS PubMed 20 Lempereur M , Aminian A , Freixa X , Gafoor S , Kefer J , Tzikas A et al. Device-associated thrombus formation after left atrial appendage occlusion: a systematic review of events reported with the Watchman, the Amplatzer Cardiac Plug and the Amulet . Catheter Cardiovasc Interv 2017 ; doi:10.1002/ccd.26903. 21 Reddy VY , Möbius-Winkler S , Miller MA , Neuzil P , Schuler G , Wiebe J et al. Left atrial appendage closure with the Watchman device in patients with a contraindication for oral anticoagulation: the ASAP study (ASA Plavix Feasibility Study With Watchman Left Atrial Appendage Closure Technology) . J Am Coll Cardiol 2013 ; 61 : 2551 – 6 . Google Scholar Crossref Search ADS PubMed 22 Plicht B , Konorza TF , Kahlert P , Al-Rashid F , Kaelsch H , Jánosi RA et al. Risk factors for thrombus formation on the Amplatzer Cardiac Plug after left atrial appendage occlusion . JACC Cardiovasc Interv 2013 ; 6 : 606 – 13 . Google Scholar Crossref Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author(s) 2018. For permissions, please email: journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Journal
European Heart Journal – Cardiovascular Imaging – Oxford University Press
Published: Dec 1, 2018