Comparisons between cross-section and long-axis-section in the quantification of aneurysmal wall enhancement of fusiform intracranial aneurysms in identifying aneurysmal symptoms

Abstract

BackgroundTo investigate the quantification of aneurysmal wall enhancement (AWE) in fusiform intracranial aneurysms (FIAs) and to compare AWE parameters based on different sections of FIAs in identifying aneurysm symptoms.MethodsConsecutive patients were prospectively recruited from February 2017 to November 2019. Aneurysm-related symptoms were defined as sentinel headache and oculomotor nerve palsy. All patients underwent high resolution magnetic resonance imaging (HR-MRI) protocol, including both pre and post-contrast imaging. CRstalk (signal intensity of FIAs' wall divided by pituitary infundibulum) was evaluated both in the cross-section (CRstalk−cross) and the long-axis section (CRstalk−long) of FIAs. Aneurysm characteristics include the maximal diameter of the cross-section (Dmax), the maximal length of the long-axis section (Lmax), location, type, and mural thrombus. The performance of parameters for differentiating symptomatic and asymptomatic FIAs was obtained and compared by a receiver operating characteristic (ROC) curve.ResultsForty-three FIAs were found in 43 patients. Eighteen (41.9%) patients who presented with aneurysmal symptoms were classified in the symptomatic group. In univariate analysis, male sex (P = 0.133), age (P = 0.013), FIAs type (P = 0.167), mural thrombus (P = 0.130), Lmax (P = 0.066), CRstalk−cross (P = 0.027), and CRstalk−long (P = 0.055) tended to be associated with aneurysmal symptoms. In the cross-section model of multivariate analysis, male (P = 0.038), age (P = 0.018), and CRstalk−cross (P = 0.048) were independently associated with aneurysmal symptoms. In the long-axis section model of multivariate analysis, male (P = 0.040), age (P = 0.010), CRstalk−long (P = 0.046), and Lmax (P = 0.019) were independently associated with aneurysmal symptoms. In the combination model of multivariate analysis, male (P = 0.027), age (P = 0.011), CRstalk−cross (P = 0.030), and Lmax (P = 0.020) were independently associated with aneurysmal symptoms. CRstalk−cross has the highest accuracy in predicting aneurysmal symptoms (AUC = 0.701). The combination of CRstalk−cross and Lmax exhibited the highest performance in discriminating symptomatic from asymptomatic FIAs (AUC = 0.780).ConclusionAneurysmal wall enhancement is associated with symptomatic FIAs. CRstalk−cross and Lmax were independent risk factors for aneurysmal symptoms. The combination of these two factors may improve the predictive performance of aneurysmal symptoms and may also help to stratify the instability of FIAs in future studies.