A Case-Matched Validation Study of Anatomic Severity Grade Score in Predicting Reinterventions After Endovascular Aortic Aneurysm Repair

Abstract

In 2002, the Society for Vascular Surgery (SVS) created the anatomic severity grading (ASG) score to classify abdominal aortic aneurysms (AAA). We have previously correlated high ASG score with increased technical difficulty during endovascular aneurysm repair (EVAR). The objective of our study was to identify the predictive capability and cutoff value of preoperative ASG score for reintervention after EVAR. We completed a retrospective review of AAA patients treated with elective EVAR from 2007 to 2011. Patients who had reinterventions as well as preoperative M2S 3-dimensional reconstructions were identified and then compared with a case-matched control group of patients without reintervention. ASG component scores (neck, aortic, and iliac) and total ASG scores were calculated using M2S software (M2S Inc, NH). Student t test was used to calculate P values where P < .05 is significant. An area under the receiver operating curve (AUROC) was created to identify the critical ASG score for predicting reintervention. A total of 623 AAA patients were treated with EVAR within the 5-year study period. Of those, 79 (13%) had reinterventions. Of the 79 reintervention patients, 45 had preoperative M2S 3-dimensional reconstructions available for ASG score calculation. The reintervention group (mean age 74 ± 8 years, 80% men) had a mean ASG score of 18 ± 5 (range, 8-30) and was compared with the case-matched control group of 45 EVAR patients (mean age 74 ± 7 years, 80% men) who had a mean ASG score of 13 ± 4 (range, 4-21; P < .0001). Demographics and risk factors were not significantly different between the two groups. The mean AAA diameter for all patients was 52 ± 14 mm and was not significantly different between the reintervention and non-reintervention groups. After AUROC analysis, an ASG score of 17 was highly predictive for reintervention (area, 0.8; sensitivity, 60%; specificity, 78%; positive predictive value, 73%; negative predictive value, 66%). An ASG score <17 yielded a 34% reintervention rate, whereas an ASG score >17 yielded a 73% reintervention rate (P = .0002). The lowest ASG score that yielded a 100% reintervention rate was 22. Most reinterventions fell into three categories: proximal extension cuff (17 [38%]), distal extension cuff (8 [18%]), and type II endoleak embolization (12 [27%]). Those that received proximal extensions had significantly higher mean total ASG score vs the case-matched controls (19 vs 15, P = .0005) as well as significantly higher individual component scores of mean neck score (3 vs 2, P = .047) and mean aorta score (7 vs 5, P = .004). Those that received distal extensions had significantly higher mean iliac score vs the case-matched controls (mean of 9 vs 7, P = .013), and those that required an embolization had a significantly higher mean aorta branch score compared with the case-matched controls (mean of 2 vs 1; P = .017). Preoperative total ASG score strongly predicts reintervention after EVAR. Use of a cutoff ASG value predictive of prohibitive reintervention rates can help guide the decision between endovascular vs open AAA repair.