Predicting adverse events after thoracic endovascular aortic repair for patients with type B aortic dissection

Abstract

The potential of adverse events (AEs) after thoracic endovascular aortic repair (TEVAR) in patients with type B aortic dissection (TBAD) has been reported. To avoid the occurrence of AEs, it is important to recognize high-risk population for prevention in advance. The data of 261 patients with TBAD who received TEVAR between June 2017 and June 2021 at our medical center were retrospectively reviewed. After the implementation of exclusion criteria, 172 patients were finally included, and after 2.8 years (range from 1 day to 5.8 years) of follow up, they were divided into AEs (n = 41) and non-AEs (n = 131) groups. We identified the predictors of AEs, and a prediction model was constructed to calculate the specific risk of postoperative AEs at 1, 2, and 3 years, and to stratify patients into high-risk (n = 78) and low-risk (n = 94) group. The prediction model included seven predictors: Age > 75 years, Lower extremity malperfusion (LEM), NT-proBNP > 330 pg/ml, None distal tear, the ratio between the diameter of the ascending aorta and descending aorta (A/D ratio) > 1.2, the ratio of the area of the false lumen to the total aorta (FL ratio) > 64%, and acute TEVAR, which exhibited excellent predictive accuracy performance and discriminatory ability with C statistic of 82.3% (95% CI 77.3-89.2%). The prediction model was contributed to identify high-risk patients of postoperative AEs, which may serve to achievement of personalized treatment and follow-up plans for patients.

Keywords: Adverse events; Computed tomography angiography; Prediction model; Thoracic endovascular aortic repair; Type B aortic dissection.

Figure 1

A consort diagram of the study population. TBAD, type B aortic dissection; TEVAR: thoracic endovascular aortic dissection; CTA: computed tomography angiography; AEs: adverse events.

Figure 2

The risk score, AEs-free survival, and incidence characteristics of patients in the study. (A) The distributed characteristics (risk score, AEs-free survival time and status) and heatmap of the patients with TBAD underwent TEVAR. The dotted lines indicated the optimal cut-off value between the low- and high-risk group; (B) Kaplan–Meier curves stratified by optimal cut-off value of the score on the prediction model with the confidence limits as a colored shaded area; (C) The cumulative incidence curves of AEs using cumulative incidence method in each group.

Figure 3

Assessment of the predictive ability of the nomogram. (A) Time-dependent ROC curves for evaluating the model’s discrimination performance of the 1-,2- and 3-year AEs-free survival outcome; area under the curves (AUC) was 0.872, 0.874 and 0.848, respectively. (B) Calibration curves for the prediction model. The curves depict the calibration of the nomogram in terms of agreement between predicted risks and actual outcomes of AEs. The number of bootstraps that were used was 1000. The x and y axes represent the predicted risk and actual outcome, respectively. The black dotted line indicates perfect prediction by an ideal model.

Figure 4

Nomogram for predicting 1-, 2- and 3-year AEs-free survival of patients with TBAD following TEVAR.

Figure 5

An overview of our study.