Risk score system for late aortic events in patients with uncomplicated type B aortic dissection

Akihito Matsushita,Minoru Tabata,Wahei Mihara,Takeshi Shimamoto,Tatsuhiko Komiya,Shuichiro Takanashi,Tetsuya Tobaru,Tatsuya Nakao,Sunao Nakamura,Yasunori Sato

DOI: https://doi.org/10.1016/j.jtcvs.2019.06.019

Abstract:Objective: Several risk factors for late aortic events in patients with uncomplicated type B aortic dissection (UTBAD) have been reported; however, they remain controversial. We developed and validated a new risk prediction score system for late aortic events in patients with UTBAD. Methods: We reviewed 187 consecutive patients diagnosed with UTBAD from 2004 to 2017 at 2 centers (derivation cohort) and 219 consecutive patients diagnosed with UTBAD from 2012 to 2016 in 4 other centers (validation cohort). We explored predictors of late aortic events using Fine-Gray generalization of the proportional hazards model, then developed a risk prediction score model and determined the test reliability using time-dependent receiver operating characteristic analyses. Finally, we validated the model using external multicenter data. Results: The risk prediction score system was developed using the following independent predictors: initial aortic diameter of ≥40 mm (2 points), false lumen diameter larger than true lumen diameter (2 points), ulcer-like projection (1 point), and age ≥70 years (1 point). Receiver operating characteristic analysis showed that a cut-off total additive score of 2 points. In the validation cohort, the low-risk group (score, 0-1 point) demonstrated lower 1- and 3-year incidence rates of late aortic events than the high-risk group (score, 2-6 points) (0.9% vs 32.5% and 0.9% vs 47.1%, respectively; P < .0001). Conclusions: We developed a simple risk prediction score system for late aortic events in patients with UTBAD. High-risk patients can be identified using our model, and they should be closely monitored and considered for interventions at the appropriate timing.