Purpose Lateral compression (LC) has recently been identified as a strong predictor of aneurysm occlusion and retreatment after endovascular treatment of intracranial aneurysms (IAs) with WEB. This study aims to compare the performance of 3 different WEB LC thresholds for the prediction of aneurysm occlusion and retreatment in a large cohort of IAs treated with WEB.

Methods We retrospectively reviewed IAs treated with WEB without adjunctive flow-diverter or coil use at our institution. Baseline patient and IA characteristics were collected. Aneurysm occlusion at follow-up was determined by 3 neurointerventionalists using the WEB occlusion scale. LC was determined by (1) measuring the average implanted WEB width at its equator in 2 orthogonal projections, and (2) subtracting it from the nominal WEB width: LC = nominal - average implanted WEB width at the equator. The patient cohort was categorized according to 3 different WEB LC thresholds: previously-derived optimal LC thresholds for complete and adequate aneurysm occlusion using receiver operating characteristic (ROC) analysis, and empiric practical LC thresholds (table 1). Univariate analyses comparing the performance of the 3 different WEB LC thresholds for the prediction of aneurysm occlusion and retreatment were conducted using Chi-Square test.

Results 193 aneurysms were included, mean maximum IA size 6.3mm, mean neck size 3.6mm, mean dome-to-neck ratio 1.5. Mean time to first follow-up 6.9 months and last follow-up 14.3 months. IA locations were 34% ACOM, 28% MCA, 9% basilar tip, 4% ICA terminus and 25% other. 40% of IAs were treated with WEB17. A significantly-higher proportion of IAs met the empiric practical LC thresholds (81.3%) than the optimal ROC LC threshold for complete occlusion (58.5%) and the optimal ROC LC threshold for adequate occlusion (65.3%,p-value<0.0001, table 2). There was no significant difference in aneurysm occlusion rates at first follow-up among IAs meeting the 3 different LC thresholds (table 2). Similarly, there was no significant difference in aneurysm occlusion rates at last follow-up among IAs meeting the 3 different LC thresholds (table 2). Further, there was no significant difference in retreatment rates among IAs meeting the 3 different LC thresholds (table 2).

Conclusion There was no significant difference between the performance of the ROC-derived and the empiric practical WEB LC thresholds for the prediction of aneurysm occlusion and retreatment in this large cohort of IAs treated with WEB. The more-inclusive empiric practical WEB LC thresholds may be more clinically-useful in determining whether a WEB device has been sized appropriately.

Disclosures J. Delgado Almandoz: 2; C; Microvention/Terumo. Y. Kayan: 2; C; Microvention/Terumo. A. Copelan: 2; C; Microvention/Terumo. A. Ertelt: None. J. Scholz: None.