Risk Factors and Predictive Model for Ischemic Complications in Endovascular Treatment of Intracranial Aneurysms: Insights From a Large Patient Cohort.

OBJECTIVES

There remains a conspicuous absence of systematic analysis concerning the risk factors for the development of ischemic complications in the interventional treatment of IAs. Our study aimed to identify the risk factors for ischemic complications after the interventional treatment of IAs and to make an individualized prediction of the occurrence of ischemic complications, providing important reference guidance for clinicians.

METHODS

This study encompassed a sample of 473 patients diagnosed with intracranial aneurysms (IA) and treated at our center between February 2022 and April 2024. Ischemic complications were identified via clinical symptomatology and corroborated with diagnostic subtraction angiography (DSA), computed tomography (CT), or magnetic resonance imaging (MRI). We used a machine learning (ML) approach to screen potential variables for ischemic complications and identify correlations between them, and subsequently constructed a logistic regression model to quantify these correlations.

RESULTS

Patients were categorized based on the occurrence or absence of ischemic complications. A total of five potential factors were screened using LASSO regression, XGBoost, and Randomforest algorithms: hypertension, history of alcohol consumption, multiple IAs, rupture status, and antiplatelet agent. Multivariate analysis further disclosed that hypertension, history of alcohol consumption, ruptured aneurysms, and antiplatelet agent were independent risk factors for postoperative ischemic complications. The predictive model, derived from the multivariate regression analysis results, demonstrated robust reliability.

CONCLUSIONS

Hypertension, history of alcohol consumption, ruptured aneurysms, and antiplatelet agent as independent risk factors for ischemic complications following the interventional treatment of IAs. Accordingly, we constructed the first risk prediction model regarding ischemic complications of all IAs based on these factors, aiming to enhance prognostic judgment and treatment strategy planning.