Clin Appl Thromb Hemost. 2026 Jan-Dec;32:10760296261489115. doi: 10.1177/10760296261489115. Epub 2026 Sep 10.
ABSTRACT
BackgroundCirculating microRNAs may predict stroke outcomes, but unaccounted sampling time can distort their prognostic value. We assessed whether plasma miR-424-5p interpreted with the elapsed time from stroke onset to blood draw predicted early neurological deterioration (END) and 90-day disability after acute ischemic stroke.MethodsThis prospective single-center observational cohort enrolled consecutive adults with imaging-confirmed acute ischemic stroke in China from June 2024 to June 2025. Time zero was the first research plasma draw with contemporaneous baseline NIHSS assessment. Primary analyses evaluated post-baseline END within 72 hours and 90-day disability in patients with valid repeat sampling 18-30 hours after time zero. Baseline miR-424-5p and 24-hour change normalized to the actual interval between the two blood draws were tested using multivariable models adjusted for pretreatment/time-zero covariates and onset-to-draw windows, with bootstrap internal validation.ResultsAmong 368 patients, 73 developed END (19.8%). The repeat-sample cohort included 337 patients, of whom 149 had 90-day disability (44.2%). Higher baseline miR-424-5p independently predicted END (adjusted OR per 1 SD, 1.57; 95% CI, 1.16-2.13; p=0.003). For 90-day disability, normalized 24-hour miR-424-5p change, but not baseline level, remained independently associated with outcome (adjusted OR per 1 SD, 1.69; 95% CI, 1.24-2.31; p=0.001). Within this development dataset, biomarker addition yielded apparent AUC changes from 0.79 to 0.83 for END and from 0.80 to 0.84 for disability.ConclusionsTime-anchored baseline miR-424-5p predicted post-baseline END, whereas its normalized 24-hour trajectory predicted 90-day disability. External validation, including comparison with 24-hour neurological reassessment, is needed.
PMID:42721098 | DOI:10.1177/10760296261489115