Predictive Model Evaluating Risk of Hemorrhage in Intracranial Aneurysms: Analysis from Prospectively Collected HEAT Trial Database

Abstract

We analyzed the data of patients enrolled in the Hydrogel Endovascular Aneurysm Treatment (HEAT) trial to develop and validate a model to predict the risk of aneurysmal hemorrhage. Analysis included data from 600 patients enrolled for the HEAT trial and included single saccular aneurysms of 3-14 mm size. Baseline characteristics were compared between patients with ruptured and unruptured aneurysms. Regression analysis was performed in the training set to identify significant risk factors and was validated in the validation dataset. The complete dataset was used to formulate a scoring model in which positive and negative predictors were assigned 1 and-1 points, respectively. Data from 593 patients were analyzed in which 169 (28.5%) patients had ruptured aneurysms. The training (n= 297) and validation dataset (n= 296) had a comparable proportion of ruptured aneurysms (29.3% and 27.7%). Dome-to-neck ratio >2.5 (odds ratio [OR] 3.66), irregular shape (OR 3.79), daughter sac (OR 5.89), and anterior and posterior communicating artery locations (OR 3.32 and 3.56, respectively) had a higher rupture rate. Use of aspirin was associated with lower risk of hemorrhage (OR 0.16). The area under the curve from the receiver operating curve analysis was 0.88, 0.87, and 0.87 in the training, validation, and combined data set, respectively. The scoring model created a score of-1 to 2, yielding an of aneurysmal hemorrhage probability from 1.5% (score-1) to 70% (score 2). This prospective study identifies dome-to-neck ratio >2.5, irregular shape, presence of daughter sac, absence of aspirin use, and aneurysm location at anterior communicating and posterior communicating artery as factors associated with increased risk of hemorrhagic presentation in small- to medium-sized intracranial aneurysms. Our model provides an estimate of rupture risk based on the presence or absence of these factors.