Validation of the HERMES-24 Score for Outcome Prediction Post Large Vessel Occlusion Treatment in Later Time Window

Scritto il 03/07/2025
da Koji Tanaka

Neurology. 2025 Aug 12;105(3):e213796. doi: 10.1212/WNL.0000000000213796. Epub 2025 Jul 3.

ABSTRACT

BACKGROUND AND OBJECTIVES: The Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials (HERMES)-24 score is highly predictive of outcomes after anterior circulation large vessel occlusion (LVO) treatment, irrespective of intervention in the early time window. Recent evidence has further broadened the eligibility of endovascular therapy (EVT) to patients with late presentation or unwitnessed onset including those with stroke-on-awakening. We aimed to investigate the prediction ability of the HERMES-24 score in patients with anterior circulation LVO and small ischemic core presenting in the late time window from last seen normal.

METHODS: Data are from the Analysis of Pooled Data from Randomized Studies of Thrombectomy More Than 6 Hours After Last Known Well collaboration, a patient-level meta-analysis of 6 randomized trials of EVT beyond 6 hours after last known well, with an enrollment period from September 2014 to March 2019. Patients who were also part of the HERMES collaboration data set were excluded from the analyses. The HERMES-24 score was calculated as the sum of the patient's age/10 and NIH Stroke Scale (NIHSS) score at 24 hours after randomization. The predictive ability of the score for a 90-day outcome (modified Rankin Scale [mRS] scores ≤2 and ≤3, ordinal mRS score, and mortality) was investigated.

RESULTS: Among 435 patients (48.5% men, median age 71 years), the median onset-to-randomization time was 654 (interquartile range 516-849) minutes and the median baseline NIHSS score was 16 (interquartile range 13-21). The HERMES-24 score was predictive of 90-day mRS scores ≤2 and ≤3, ordinal mRS score, and mortality in both the EVT arm (n = 223, c-statistic [95% CI] 0.917 [0.875-0.944], 0.895 [0.853-0.938], 0.820 [0.745-0.891], and 0.849 [0.776-0.922], respectively) and the control arm (n = 212, c-statistic [95% CI] 0.921 [0.872-0.969], 0.879 [0.827-0.930], 0.805 [0.746-0.852], and 0.805 [0.738-0.871], respectively).

DISCUSSION: The HERMES-24 score was highly predictive of 90-day outcome among patients with stroke due to LVO and small ischemic core for those presenting in the late time window, irrespective of intervention. This score must be further validated in a real-world clinical setting if it is applicable to all patients with LVO admitted in late time windows.

PMID:40609063 | DOI:10.1212/WNL.0000000000213796