Predictive Modeling of Endograft Limb Occlusion After Endovascular Aneurysm Repair: A Propensity Score Matching Analysis.

Background: Endovascular aneurysm repair (EVAR) is widely used to treat abdominal aortic aneurysm (AAA). However, it carries the risk of complications that often require further interventions. Endograft limb occlusion (ELO), though relatively rare, significantly affects patient outcomes. This study investigates factors influencing and predicting ELO risk after EVAR.

Methods: All patients who underwent EVAR for AAA between 2013 and 2023 at the First Affiliated Hospital of Guangxi Medical University were examined. Patients were followed up and categorized into ELO and non-ELO groups. The groups were made comparable using propensity score matching. A prediction model was constructed using Lasso regression analysis.

Results: The study included 425 AAA patients. After propensity score matching, the analysis involved 23 patients with ELO and 69 patients without ELO. All patients with the stent positioned distally in the external iliac artery were exclusively in the ELO group, indicating that the stent position variable perfectly predicted the outcome within our matched cohort. A predictive model was constructed using Lasso regression analysis, incorporating three variables: double iliac sign (DIS), surgical approaches, and iliac artery stent oversizing. Validation of the predictive model using the Hosmer-Lemeshow test demonstrated its excellent predictive capability, achieving an area under the curve of 0.91.

Conclusion: DIS, open surgical approaches, and iliac artery stent oversizing (>50%) emerged as independent risk factors for ELO occurrence in AAA patients undergoing EVAR. The position of the stent in the external iliac artery independently predicts ELO and directly indicates a high-risk situation. Our predictive model aids in risk stratification and clinical decision-making, thereby improving patient outcomes.