Early 3D Growth in Uncomplicated Type B Aortic Dissection is Associated with Long-Term Outcomes

Abstract

ObjectiveLate adverse events (LAE) are common among initially uncomplicated type B aortic dissection (uTBAD), however, identifying those patients at highest risk of LAE remains a significant challenge. Early false lumen (FL) growth has been suggested to increase risk, but confident determination of growth is often hampered by error in 2D clinical measurements. Semi-automated 3D mapping of aortic growth, such as by vascular deformation mapping (VDM), can potentially overcome this limitation using CT angiograms (CTA). We hypothesized that FL growth in the early pre-dissection phase by VDM can accurately predict LAEs. MethodsWe performed a two-centre retrospective study of uTBAD patients, with paired CTAs in the acute (1-14 days) and subacute/early chronic (1-6 months) periods. VDM analysis was used to map 3D growth. Standard clinical CT measures (i.e., aortic diameters, tear characteristics) were also collected. Multivariate analysis was conducted using a decision tree and Cox proportional hazards model. LAEs were defined as aneurysmal FL (>55mm); rapid growth (>5mm within 6 months); aorta-specific mortality, rupture, or re-dissection. Results107 (69% male) initially uTBAD patients met inclusion criteria with a median follow-up of 7.3 (IQR 4.7-9.9) years. LAEs occurred in 72 patients (67%) at 2.5 (IQR 0.7-4.8) years after the initial event. A multivariate decision tree model identified VDM growth (>2.1 mm) and baseline diameter (>42.7 mm) as optimal predictors of LAEs (AUC-ROC = 0.94), achieving an 87% accuracy (sensitivity of 93%, specificity of 76%) after leave-one-out validation. Guideline reported high-risk features were not significantly different between groups. ConclusionEarly growth of the FL in uTBAD was the best tested indicator for LAEs and improves upon the current gold-standard of baseline diameter in selecting patients for early prophylactic TEVAR.