Abstract
Objective: Mechanical thrombectomy for acute large vessel occlusion (LVO) is currently widely performed. However, rescue treatment (RT), such as percutaneous transluminal angioplasty (PTA) and stenting, is occasionally required, particularly in the case of atherothrombotic brain infarction (ATBI) or dissection. As RT requires higher levels of therapeutic skills and additional devices, early prediction of its performance and preparation are important. We retrospectively investigated the pre-therapeutic factors for predicting the necessity of RT.
Methods: We reviewed 149 consecutive patients who underwent mechanical thrombectomy for acute LVO between April 2014 and December 2019. Eight patients were excluded because of missing clinical data. RT was performed when severe stenosis was observed in occluded vessels or proximal to them during mechanical thrombectomy. We investigated pre-therapeutic neurological, laboratory, and radiological findings in the 141 remaining patients, and compared them between RT and non-RT groups.
Results: RT was performed on 23 of the 141 patients. We found four pre-therapeutic factors with significantly different rates between RT/non-RT as follows: (1) Atrial fibrillation 8.7%/71.1% (p <0.001), (2) diabetes mellitus 39.1%/19.5% (p = 0.04), (3) susceptibility vessel sign (SVS) by T2-weighted imaging 17.4%/66.1% (p <0.001), and (4) tapered occlusion by magnetic resonance angiography (MRA) 47.8%/11.9% (p <0.001). The plasma level of brain natriuretic peptide (BNP) was also significantly different between the two groups. When the BNP level was less than 70 pg/mL, the sensitivity for being in the RT group was 86.9% and the specificity was 83.5%.
Conclusion: Pre-therapeutic findings, such as diabetes mellitus, tapered occlusion, absence of atrial fibrillation, negative SVS, and BNP level less than 70 pg/mL, are predictors of RT in mechanical thrombectomy.
