ABC | Volume 113, Nº4, October 2019

Original Article Şahiner et al LAA occlusion with amulet device Arq Bras Cardiol. 2019; 113(4):712-721 treatment may explain this difference in our patient series. We planned the antiplatelet regimen according to the patients’ risk of stroke and thrombus formation over the device. DAPT was administered to most patients (88.3%) at discharge in our group. In addition, we planned an extensive antiplatelet therapy for patients who had peri-device leak at follow-up. Also, five of our patients had thrombus in the LAA before the procedure. These patients underwent low-dose anticoagulant therapy for six months. As peri-device leak and presence of thrombus in the LAA before the procedure were considered risk factors for thrombus formation on the closure device, we decided to individualize the antiplatelet therapy of these patients. Moreover, the relatively small sample size may be another cause for this discrepancy. Our series had only one major periprocedural complication. The indication of LAA occlusion for this patient was preprocedural hemorrhagic CVE with an effective dosage of dabigatran. LAA occlusion was planned as thromboembolic prevention for this patient. Nonetheless, we observed a major periprocedural complication during the procedure. In this patient, the lobe hooks erupted from the LAA and damaged the pulmonary artery. These hooks are designed to allow better implantation and fixation of the device. When we reevaluated the preprocedural MDCT, we noted the close neighborhood between LAA and pulmonary artery. This close relationship resulted in pulmonary artery rupture. Although referred to urgent surgery, the patient did not survive. Previous case reports showed that postprocedural pulmonary artery rupture could be an early or delayed complication. 18,19 Most of these case reports mentioned that this complication is related to the anatomical relationship between LAA landing zone and pulmonary artery. 18,19 Halkin A. et al. 20 classified this relationship according to the contact point between LAA and pulmonary artery and they emphasized that the type 2 (proximal contact) relationship has a higher pulmonary artery rupture risk than the others. 20 We reevaluated the relationship between pulmonary artery and LAA in our patient after this study and found that it was a type 2 relationship (Figure 4). Thrombus presence in LAA is considered a contraindication for LAA occlusion. 21 In our series, we detected thrombus at the bottom of the LAA in five patients. We considered that the thrombus at distal LAA could be attached to the LAA occlusion device, with a modified technique and minimal manipulation of catheters in the left atrium. Consequently, the procedures were performed successfully with no periprocedural neurological complications. We have reported one of these cases previously. 22 Tarantini et al. 23 recently reported in their multicenter study that LAA occlusion could be safely and effectively performed in 28 patients with distal LAA thrombus. 23 In line with these findings, we also suggest that LAA occlusion could be successfully conducted in patients with distal thrombus in experienced centers. However, the procedure should be canceled if the thrombus is located at the proximal LAA. Percutaneous LAA occlusion is a complex procedure that has some periprocedural risks as we mentioned before. Consequently, preprocedural patient evaluation, patients with appropriate indications, and operator experience are very important to avoid possible complications. In our study, we demonstrated that LAA occlusion using the Amulet™ LAA occluder could be performed with high procedural success. In our series, all but one of the procedures were completed safely without complications. We did not find any clinical events directly related to AF or the LAA procedure during postprocedural follow-up. On the other hand, further large-scale randomized trials and long-term outcome data are necessary to verify the efficacy and safety of LAA occlusion using the Amulet™ LAA occluder device. Limitations This study was not designed as a randomized prospective controlled trial; consequently, it has some limitations. First, we did not have a control group to compare the effectiveness of LAA occlusion in preventing thromboembolic events. Second, our mean follow-up duration was relatively short, and long‑term outcomes of LAA occlusion cannot be inferred from our results. However, the LAA occlusion procedure and Amulet device are relatively new, and data about this device are limited. Therefore, studies like ours are still important and valuable Table 2 – Procedural and Follow-up Outcomes Procedural Outcomes Patients (n = 60) Technical Success 60 (100%) Procedural Success 59 (98.3%) Periprocedural Mortality 1 (1.6%) Periprocedural Morbidity 9 (15.0%) Major Bleeding 0 (0.0%) Minor Bleeding 6 (10.0%) Stroke 0 (0%) Systemic Embolization 0 (0%) Device Embolization 0 (0%) Pericardial Effusion 2 (3.2%) Pericardial Tamponade 1 (1.6%) Follow-up Outcomes Patients (n = 59) Mortality 2 (3.4%) Stroke/TIA 2 (3.4%) Ischemic Stroke 0 Hemorrhagic Stroke 0 TIA 2 (3.4%) Pulmonary Thromboembolism 1 (1.7%) Life Threating or Major Bleeding 1 (1.7%) Minor Bleeding 3 (5.1%) Major Findings in Follow up TEE Peridevice Leak (>3 mm) at 1 st month 2 (3.4%) Peridevice Leak (>3 mm) at 6 th month 0 Device Related Thrombus 0 Device Embolization 0 TIA: transient ischemic attack; TEE: transesophageal echocardiography. 718