Abstract Body: Background: Managing unruptured intracranial aneurysms (IAs) involves balancing procedural risks with the long-term risk of rupture. Aneurysm wall enhancement (AWE) has been associated with aneurysm instability and symptomatic presentation. However, assessing AWE typically relies on subjective two-dimensional visual inspection, which lack objectivity. This study aims to evaluate a predictive model for symptomatic status by integrating clinical, morphological, and three dimensional-AWE (3D-AWE) metrics from a large dataset of IAs.

Methods: Patients from three large academic center from the US and China, with saccular unruptured IAs underwent 3T high-resolution magnetic resonance imaging. Morphological assessment was conducted through angiographic studies. The determination of AWE was performed using a 3D-mapping pipeline that quantifies the average signal intensity of the aneurysm wall after contrast administration producing an objective metric: 3D-AWE (figure). Symptomatic status was defined by rupture, sentinel headaches, cranial nerve palsy, and focal neurological symptoms due to mass effect by the aneurysm. The optimal multivariate model for predicting symptomatic presentation of IAs was determined using an all-subsets regression approach.

Results: A total of 389 IAs were included, with 73 (19%) being symptomatic at presentation. The model using clinical and morphological aneurysm variables achieved an AUC of 0.78, with 60% sensitivity and 86% specificity in identifying symptomatic aneurysms. Adding 3D-AWE metrics improved the model's performance, increasing the AUC to 0.82, with 73% sensitivity and 78% specificity. High-risk locations (anterior communicating, posterior communicating, and basilar arteries; OR 1.87, p 0.04), size ratio (OR 1.52, p<0.001), smoking (OR 1.91, p 0.07), and positive 3D-AWE (OR 7.09, p<0.001) were identified as independent predictors of symptomatic presentation.

Conclusions: Incorporating objective 3D-AWE data enhanced the predictive performance of clinical and morphological information for detecting symptomatic IAs.