CTA-based morphological parameters predict adverse prognosis in stanford type B aortic dissection: a single-center retrospective cohort study

Abstract

To investigate the prognostic value of computed tomography angiography (CTA)-based morphological parameters for adverse aortic events in patients with acute Stanford type B aortic dissection (TBAD) and to develop an internally validated prediction model. This single-center retrospective cohort study included 292 consecutive patients with CTA-confirmed acute TBAD. Baseline clinical and CTA-derived morphological variables were collected, and the primary endpoint was a composite of adverse aortic events. Candidate predictors were evaluated using least absolute shrinkage and selection operator (LASSO) logistic regression with 10-fold cross-validation, followed by multivariable logistic regression. Model performance was assessed using discrimination, calibration, precision-recall analysis, and decision curve analysis, with internal validation based on 1,000 bootstrap resamples. During a median follow-up of 13.20 months (interquartile range, 8.15–19.52 months), 78 patients (26.7%) experienced adverse aortic events. The final LASSO-selected model included age, symptom onset-to-admission time, white blood cell count, D-dimer, maximum aortic diameter, TL/FL ratio, and false lumen status. In multivariable analysis, older age (adjusted OR = 1.043), higher white blood cell count (OR = 1.164), higher D-dimer (OR = 1.056), and larger maximum aortic diameter (OR = 1.110) were associated with higher odds of adverse events, whereas a higher TL/FL ratio was inversely associated with the outcome (OR = 0.060). Compared with partial false-lumen thrombosis, patent false lumen was associated with higher odds (OR = 3.718), whereas complete thrombosis was associated with lower odds (OR = 0.253). The apparent AUC was 0.849 (95% CI: 0.800–0.898), and the optimism-corrected AUC was 0.828 after 1,000 bootstrap resamples. The optimism-corrected calibration slope and Brier score were 0.893 and 0.143, respectively. The AUPRC was 0.708 vs. a no-skill baseline of 0.267, and decision curve analysis showed positive net benefit across a broad range of threshold probabilities. A model integrating baseline clinical and CTA-derived morphological characteristics demonstrated good internally validated discrimination and reasonable calibration for short- to medium-term adverse aortic events in acute TBAD. Independent external validation is required before broader clinical application.