Characterizing liquid embolic distribution and its impact on chronic subdural hematoma outcomes following middle meningeal artery embolization

BACKGROUND AND PURPOSE: The non-contrast CT (NCCT) appearance of liquid embolic material post-middle meningeal artery embolization (MMAE) for chronic subdural hematomas (cSDH) is poorly described. We aimed to categorize the spectrum of post-procedural liquid embolic distribution and determine whether embolic penetration impacts resorption.

MATERIALS AND METHODS: We retrospectively reviewed patients with cSDH >5 mm at a single institution from 2019-2024. Patients with <1 month of follow-up imaging or embolization with coils or particles were excluded. Embolic membrane penetration during angiography was recorded, and the postprocedural appearance of embolic material was classified. Regression analyses were performed to assess if specific embolic distributions were associated with greater resorption.

RESULTS: Eighty-eight patients (73.9% male; mean age 73.5±11.2 years) with 109 unilateral cSDHs were included. Membrane penetration occurred in 39 cases (35.8%). On follow-up NCCT, liquid embolic material was visualized on the outer surface of the cSDH in 85 cases (78.0%), within the cSDH in 16 cases (14.7%), and on the inner surface of the cSDH in 8 cases (7.3%). N-butylcyanoacrylate (nBCA) was more likely to penetrate the subdural membrane (55.1%) and reach deeper compartments (40.8%) compared to Onyx-18 (17.0% and 7.5%; both p<0.0001). Membrane penetration was associated with greater mean absolute reduction (1.05 cm vs. 0.66 cm, p<0.001), mean percent reduction (73.8% vs. 53.5%, p=0.01), and a faster average resorption rate (0.013 mm/day vs. 0.007 mm/day, p=0.001). Deeper compartment penetration visualized on follow-up NCCT was associated with faster resorption on univariate analysis (0.014 mm/day vs. 0.008 mm/day, p=0.01) but did not impact absolute reduction, percent reduction, or complete resorption. In regression analysis, membrane penetration was associated with faster resorption (β = 0.004 mm/day, p=0.04), greater percent reduction (β = 32.65%, p=0.001), and nearly four times the odds of complete resolution (β = 1.38, p=0.04).

CONCLUSIONS: We present a framework for characterizing the appearance of liquid embolic agents after MMAE. Visualized membrane penetration during angiography was associated with greater and faster cSDH resorption and higher odds of complete resolution. nBCA was associated with deeper compartment penetration; however, deeper compartment penetration on NCCT was not independently predictive of improved outcomes after adjustment for covariates.

ABBREVIATIONS: MMAE = Middle meningeal artery embolization; cSDH = Chronic subdural hematoma; nBCA = N-butylcyanoacrylate