Size ratio for clinical assessment of intracranial aneurysm rupture risk

Objectives: We previously used three-dimensional (3D) volumetric analysis to identify a novel intracranial aneurysm (IA) morphological metric, aneurysm-to-parent vessel size ratio (SR), which strongly correlated with aneurysm rupture. However, complex 3D analysis is not easily obtained, and ubiquitous IA risk assessment is traditionally performed with two-dimensional (2D) imaging, typically with size being the sole considered morphometric. Because only easily applicable 2D measurements will be of clinical value, we sought to investigate the correlation of SR determined from 2D angiography with IA rupture status. Methods: SR and traditional aspect ratio (AR) and aneurysm size parameters were measured in a retrospective cohort of 38 IA cases (16 ruptured) with 2D rotational angiographic images. These parameters were analysed for correlation with IA rupture status. Student's t-test or Wilcoxon rank-sum test was used for normally or non-normally distributed data respectively. Logistic regression was performed for independently statistically significant parameters to generate an effect size estimate (odds ratio). Area-under-the-curve (AUC) calculated from the receiver-operating-characteristic curve was additionally obtained for each index to describe differentiating capabilities. Results: Only SR achieved statistical significance (p=0.05) in Wilcoxon rank-sum test. Logistic regression generated an SR odds ratio of 3.52 (p=0.04; 95% confidence interval: 1.035-11.938) for every doubling of SR value. The AUC value of SR (0.688) was higher than that of AR (0.642) and size (0.585). Conclusions: SR had the strongest correlation with IA rupture and was demonstrated to be a valuable parameter in 2D, where it can be easily obtained from angiographic images. When eventually evaluated in a prospective data set, SR may prove to be an important tool for aneurysm rupture-risk assessment.