Abstract 286: Prediction Of Adverse Events Following Stent Graft Repair Of Type B Aortic Dissection From Imaging Modalities

Abstract

Objectives: The prophylactic treatment of uncomplicated TBAD with TEVAR is controversial. This warrants the identification of a high-risk category of dissections that may particularly benefit from surgical intervention. The analysis of radiographic features presents a promising modality of assessing this high-risk cohort. We test the ability of aortic size and shape metrics from the literature to predict patient suitability for TEVAR from pre-operative imaging. Methods: We collected a single institutional retrospective cohort of 36 patients with TBAD who received TEVAR and had pre-operative and follow-up CTA imaging. We tested 8 aortic size and shape metrics. We segmented each patient’s aorta and true lumen from the pre-operative scan. Tortuosity, mean diameter, centerline curvature, and eccentricity were measured from the centerline. True and false lumen volumes and max diameter were calculated. The question mark angle, as previously defined by Li et al., was also measured. Univariate and multivariate logistic regression analyses were performed. Results: In the univariate analysis, pre-operative false lumen volume (OR 26.2 95% CI 2.72 to 252, p = .005), mean diameter (OR 9.6 95% CI 2.19 to 42.3, p = .003), and maximum diameter (OR 9.0 95% CI 2.12 to 38.3, p = .003) were all significantly associated with post-TEVAR outcomes. Tortuosity index (OR 2.7, 95% CI 1.14 to 6.37, p = .024) was the only significant shape parameter. In multivariate analysis, we found that pre-operative maximum diameter (OR 10.0, 95% CI 1.50 to 67.0, p < .018) is a significant predictor of TEVAR outcomes independent of shape parameters, all of which were not significant. Conclusions: False lumen volume and maximum diameter can predict the occurrence of reintervention and type I endoleak following TEVAR for uncomplicated TBAD. While size measures are effective in explaining aortic dissection behavior, current shape measures are not as effective and better methodologies must be developed.