Interpretable machine learning analysis for predicting the instability of the distal false lumen after thoracic endovascular aortic repair in type B aortic dissection

The instability of the distal false lumen (FL) is a serious complication after thoracic endovascular aortic repair (TEVAR) for type B aortic dissection (TBAD). This study aimed to construct a prediction model to predict the risk of instability of the distal FL after TEVAR for TBAD at an early stage. We retrospectively collected the demographic, clinical, laboratory, and imaging characteristic information of patients with TBAD who received TEVAR between July 2014 and August 2023. The least absolute shrinkage and selection operator (LASSO) and multiple logistic regression were used to screen predictive variables. Then, four machine learning models were constructed using the screened features, and these models were validated using bootstrap re-samples. The area under the receiver operating characteristic (ROC) curve (AUC), decision curve analysis (DCA), calibration curve analysis, and the area under the Precision-Recall (PR) curve (PRAUC) were used to compare the performance of the models. SHapley Additive ex Planations (SHAP) was used to explain the best model. The analysis ultimately included 202 TBAD patients who received TEVAR, among whom 34 (16.83%) patients experienced instability of the distal FL. The predictors in the final model included age, the length of residual FL (LRFL), and preoperative lymphocyte percentage. The Random Forest (RF) model demonstrated the best predictive performance, with AUC and PRAUC of 0.923 (95% CI: 0.861-0.974) and 0.801 (95% CI: 0.655-0.914) respectively. SHAP analysis indicated that the anatomical index of LRFL was the most crucial feature in the prediction model. The RF model constructed in this study can effectively predict the instability of the distal FL after TEVAR in patients with TBAD, which may help clinicians develop personalized treatment and postoperative monitoring plans based on the characteristics of patients to improve the prognosis of TBAD. The unstable expansion of the distal false lumen is an important cause of poor prognosis after thoracic endovascular aortic repair (TEVAR) in patients with Stanford type B aortic dissection (TBAD). Early identification of high-risk patients is crucial for optimizing treatment and follow-up strategies. The prediction model constructed in this study can help clinicians evaluate the risk of instability of distal false lumen in patients in the early stage after TEVAR, and it has a certain degree of interpretability.