Abstract Number: LBA5 Predictors of Occlusion After Flow Diversion of Internal Carotid Artery Aneurysms: A Pooled Analysis

Abstract

Introduction Flow diverters (FDs) have demonstrated a safe and effective profile for the treatment of intracranial aneurysms with complex morphologies and variable anatomic locations. However, aneurysmal persistence after flow diversion still presents in up to 25% of treated aneurysms. Herein, we aimed to perform a pooled analysis of two large studies (SCENT [Surpass Intracranial Aneurysm Embolization System Pivotal Trial to Treat Large or Giant Neck Wide Neck Aneurysms] and SESSIA [Safety and Efficacy of the Surpass Streamline for Intracranial Aneurysms]) to evaluate the predictors of occlusion at final follow‐up after treatment with a 72–96 wire device. Methods We pooled all data from the SCENT (prospective) and SESSIA (retrospective) studies that included patients treated with the Surpass Streamline. From the studies, a subset of patients with unruptured saccular aneurysms located in the internal carotid artery (ICA) up to its terminus were included. The authors collected baseline demographics, comorbidities, aneurysm dimensions, and procedural and follow‐up characteristics for each patient. For our analysis, the outcome was complete aneurysm occlusion, as reported by the studies. We performed a multivariable logistic regression to identify the predictors of complete occlusion. Next, we performed a mediation analysis framework to identify the causal relationship of the predictors with the outcome. Finally, we calculated the predicted probability of occlusion for the continuous predictors. Results A total of 348 patients with 348 aneurysms were included, 394 devices were implanted (1.13 per patient). Median age was 61 [22‐85] years, and 83.6% were females. Hypertension was the most common comorbidity (57%) followed by hyperlipidemia (36%). The ICA paraophthalmic segment was the most common location (45%), followed by the petrocavernous (29%) and supraclinoid (25%). Median aneurysm size was 7.5 [1‐29] mm, and neck size was 4.3 [1‐23] mm. At the final follow‐up, the complete occlusion rate was 73% (255/348). After adjusting for confounders and accounting for collinearity, the multivariable analysis identified aneurysm size (OR 0.89; 95% CI 0.85‐0.93;p< .001), procedural technical events (OR 0.31; 95% CI 0.15‐0.65;p = .002), and first device length (OR 0.98; 95% CI 0.96‐1.00;p = .036) as predictors of complete occlusion. Age had a non‐significant direct effect on complete occlusion (p = .091) but a significant indirect effect mediated through aneurysm size (p< .001) and technical events (p< .02). Using our model, the predicted probability of occlusion is≥ 75% in aneurysms measuring < 10 mm when no technical events are encountered. However, when technical events are encountered, the probability decreases to 50–75%. Conclusions Successful aneurysm occlusion after flow diversion is associated with aneurysm size, procedural technical events, and FD length. Age does not directly affect occlusion, but its influence is mediated through the aneurysm size and technical events. Therefore, a priori knowledge of the patient and aneurysm characteristics might guide FD selection to favor the best treatment outcomes.