Introduction: Oscillations in blood flow over the cardiac cycle are relevant to endothelial function and vascular health. But, as yet the relationship between pulsatile flow characteristics and intracranial aneurysm (IA) growth is unclear. We analyzed time-based hemodynamic indices within unruptured IAs with known growth or stable trajectories. We hypothesize that combining pulsatile parameters can yield a predictive model for IA growth. Methods: Unruptured IA cases with at least 3 imaging studies were identified from UCLA Medical Center electronic medical records. Cases with an increase in diameter of at least 10% were assigned to the growing group and all other cases were assigned to the stable group. 16 stable and 20 growing cases from 2018-2020 qualified for the study. For each IA image study, computational fluid dynamic (CFD) flow simulation was performed and pulsatility index (PI) and wall shear stress PI (WSSPI) were recorded at 5 locations: inflow artery, IA neck, body, dome, and outflow artery. Values were compared with MANOVA. A stepwise logistic regression with growth as the outcome variable and WSSPI and PI at all locations as covariates was then performed. Results: PI was significantly higher in growing IA at the body (0.81±0.44 vs. 1.76±1.18, p=0.006) and dome (1.22±0.76 vs. 1.91±1.20 p=0.04). WSSPI was significantly higher in growing IA at the inflow artery (0.68±0.05 vs. 0.87±0.36, p=0.049), body (1.07±0.22 vs. 1.55±0.73, p=0.02), and dome (1.15±0.30 vs. 1.64±0.70, p=0.01). Logistic regression yielded a predictive model with 88.2% accuracy and AUC of 0.944: logit(growth) = 2.035 х PI_ body + 14.004 х WSSPI_ inflow + 4.263 х WSSPI_ body -17.342. Conclusions: Based on a finding of elevated pulsatility in growing IA, we propose a model to predict subsequent IA growth. This should be further tested in larger CFD studies. Upon confirmation, the model may be helpful in guiding clinical decisions.
Read full abstract