Identification of the optimal thresholds for the hypoperfusion intensity ratio in moyamoya disease

The authors' objective was to retrospectively compare two methods for defining the hypoperfusion intensity ratio (HIR) in moyamoya disease (MMD) by using hypoperfused volumes calculated from time to maximum of the residue function (Tmax) thresholds of 10 seconds/4 seconds and 10 seconds/6 seconds.

All hemispheres were categorized into normal, ischemic, and hemorrhagic groups. Hypoperfused volumes were calculated using Tmax thresholds of 10 seconds, 6 seconds, and 4 seconds. HIR was computed as Tmax > 10 seconds/Tmax > 4 seconds (HIR10/4) and Tmax > 10 seconds/Tmax > 6 seconds (HIR10/6). Angiographic collaterals were assessed using CT perfusion (CTP)-sourced images (CTP-sis). The 3-month clinical follow-up included primary outcomes (survival or death) and secondary outcomes (modified Rankin Scale [mRS] and Katz activities of daily living [ADL] scale scores). Multivariate logistic regression and correlation analyses were conducted.

Thirty patients (54 hemispheres) were included. Patients with poor primary outcomes exhibited higher rates of hypertension (p = 0.015), larger hypoperfused volumes, and elevated HIR10/4 and HIR10/6 (p < 0.001). The regression model with HIR10/4 outperformed that with HIR10/6 for predicting primary outcomes (Z = 2.02, p = 0.044). Both HIR10/4 and HIR10/6 correlated with mRS and ADL scores (p < 0.05). Although there was no correlation between HIR and CTP-sis when all hemispheres were included, a significant association was found between HIR10/4 and CTP-sis when hemispheres with hemorrhagic lesions were excluded (p = 0.013).

HIR10/4 exhibited a superior predictive value for primary outcomes in MMD compared to HIR10/6. Additionally, HIR10/4 showed a significant association with angiographic collaterals, particularly in ischemic MMD cases. This study suggested that HIR defined as Tmax > 10 seconds/Tmax > 4 seconds may be more suitable than Tmax > 10 seconds/Tmax > 6 seconds in MMD.