Perivascular fat attenuation on computed tomography angiography is associated with aneurysm recurrence after endovascular coil embolization

Vascular inflammation contributes to aneurysm wall instability and impaired healing after endovascular treatment. We evaluated whether perivascular adipose tissue attenuation around the ipsilateral cervical internal carotid artery on pretreatment computed tomography angiography predicts angiographic recurrence after coil embolization. This single-center retrospective cohort study included 303 adults treated with coil embolization without stent assistance for anterior-circulation aneurysms between January 2020 and December 2023. Perivascular adipose tissue was segmented around the ipsilateral C1 segment using a -190 to -30 HU threshold and measured independently by two blinded neuroradiologists. Recurrence was defined as new intra-aneurysmal flow on follow-up imaging confirmed by digital subtraction angiography. Logistic regression with candidate variables pre-selected at P < 0.05, Cox regression, receiver operating characteristic analysis, calibration, bootstrap internal validation, decision-curve analysis, side-specific PVAT ratios, and contralateral C1 measurements were used to assess the association of PVAT with recurrence. During a median follow-up of 23.50 months, 36 patients (11.90%) developed recurrence. Recurrent aneurysms had larger maximum diameter (6.82 ± 1.68 vs. 5.70 ± 1.99 mm), wider neck (3.38 ± 0.83 vs. 2.56 ± 0.86 mm), and higher ipsilateral perivascular adipose tissue attenuation (-59.50 ± 6.30 vs. -70.00 ± 6.60 HU; all P < 0.01). Perivascular adipose tissue remained independently associated with recurrence in logistic regression (odds ratio 1.30, 95% confidence interval 1.17–1.45) and Cox regression (hazard ratio 1.18, 95% confidence interval 1.11–1.26), and the LASSO-penalized optimism-corrected AUC was 0.83. Adding perivascular adipose tissue improved the apparent AUC from 0.73 to 0.86. Ipsilateral cervical internal carotid artery perivascular adipose tissue attenuation on routine computed tomography angiography is associated with angiographic recurrence after coil embolization in this single-center cohort. This exploratory association may improve risk stratification when combined with standard morphological predictors, but external validation is required before clinical use.