Work

Improved outcome prediction in large vessel occlusion stroke: A comparison of Tmax-based hypoperfusion ratios HIR10/4 and HIR10/6

Year: 2025

Type: article

Abstract: Background The hypoperfusion intensity ratio (HIR), derived from CT perfusion (CTP), reflects collateral flow quality, but its optimal formulation remains unsettled. Objective To compare the prognostic utility of conventional HIR 10/6 (time-to-maximum [T max ] >10s / >6s) with an alternative HIR 10/4 (T max >10s / >4s) in patients undergoing endovascular thrombectomy (EVT). Methods We retrospectively analyzed consecutive large-vessel occlusion (LVO) strokes treated with EVT at our quaternary academic medical center (2015–2024). CTP was processed using RAPID software. Primary outcomes were modified Rankin Scale (mRS) at discharge and at 90 days. Secondary outcomes included National Institutes of Health Stroke Scale (NIHSS) at discharge and hemorrhagic transformation on follow-up imaging. Ordinal and linear regression models were adjusted for age, sex, vascular risk factors, intravenous thrombolysis, presentation window, and transfer status. Receiver operating characteristic (ROC) curves assessed discriminative performance. Results Among 280 patients (mean age 70 ±15 years; 49% male), mean HIR 10/6 was 0.418 and HIR 10/4 was 0.241. Each 0.1-unit increase in HIR 10/4 translated to 31% higher odds of worse mRS at discharge (adjusted odds ratio [aOR] 1.31, 95% CI 1.12–1.54; p =0.001), 24% higher odds of worse mRS at 90 days (aOR 1.24, 95% CI 1.04–1.48; p =0.016), and higher discharge NIHSS (β=0.88, 95% CI 0.05–1.71; p =0.039). HIR 10/6 showed weaker associations and was not significant for discharge NIHSS. Neither HIR metric demonstrated meaningful discrimination for hemorrhagic transformation or dichotomized outcomes (area under the curve [AUC] 0.50–0.52). Conclusions Incorporating mildly delayed tissue (T max >4s) may enhance the physiologic sensitivity of HIR. HIR 10/4 may be a potential prognostic biomarker; however, prospective validation is needed, and its standalone ability to discriminate binary outcomes remains limited.

Source: Journal of Stroke and Cerebrovascular Diseases

DOI: 10.1016/j.jstrokecerebrovasdis.2025.108504