Prehospital Detection of Large Vessel Occlusion and Intracerebral Hemorrhage Using a Dual-Biomarker Point-of-Care Test.
Stroke (Hoboken, N.J.) · 2026-03-02 · Method & validation
Abstract
BACKGROUND: Timely identification of stroke subtype is critical for triage and treatment. Existing prehospital stroke scales have limited diagnostic accuracy, especially in differentiating large vessel occlusion (LVO) from intracerebral hemorrhage (ICH). The LVOne point-of-care test measures d-dimer and GFAP (glial fibrillary acidic protein), biomarkers associated with thrombotic and hemorrhagic stroke, enabling biologically informed triage. METHODS: We validated a new version of the LVOne assay using plasma samples from 210 suspected stroke patients presenting within 6 hours of symptom onset, enrolled in the TIME (Testing of Identification Markers for Stroke) prospective observational cohort. This article reports a retrospective (secondary) laboratory analysis of these prospectively collected samples and clinical data. Samples were collected before neuroimaging or thrombolysis. LVOne version 1 combined d-dimer positivity, GFAP negativity, and Field Assessment Stroke Triage for Emergency Destination scoring for LVO detection only; version 2 uses improved GFAP measurement and a clinical decision rule enabling simultaneous diagnosis of LVO and ICH. Diagnostic performance (sensitivity, specificity, positive predictive value, negative predictive value) was calculated for each version. Receiver operating characteristic analyses and biomarker correlations with Field Assessment Stroke Triage for Emergency Destination were also performed. RESULTS: Both LVOne versions achieved high accuracy for LVO detection (sensitivity, 0.75; specificity, 0.92; negative predictive value, 0.97). Version 2 improved ICH specificity (0.99 versus 0.90) and positive predictive value (0.83 versus 0.23) versus version 1. Biomarker levels correlated with Field Assessment Stroke Triage for Emergency Destination scores but with limited explanatory power (GFAP R 2=0.077; d-dimer R 2=0.047), suggesting added information beyond clinical severity scales. In a specimen-matrix comparison (n=25), qualitative GFAP/d-dimer classification was concordant across plasma, venous whole blood, and capillary whole blood. CONCLUSIONS: The updated LVOne assay enables simultaneous biomarker-based identification of both LVO and ICH in suspected stroke patients. Its rapid, low-complexity format supports prehospital use and may improve triage, reduce time to treatment, and optimize outcomes. Prospective field validation is warranted. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04292600.
Abstract © the authors or publisher, via PubMed. Reuse and removal
Tags
Decision support tools · Point-of-care tests · Triage instruments
Related articles
- Integrating GFAP Testing With Clinical Scores for Prehospital Large Vessel Occlusion Detection in Patients With Acute Stroke (DETECT LVO). (Stroke, 2026)
- Evaluation of the prehospital diagnostic accuracy of a novel point-of-care test for NT-proBNP, D-dimer and H-FABP and large-vessel occlusion risk assessment (LVOCheck-EVA): a protocol for a multicenter prospective observational study in patients suspected of having a stroke. (Front Neurol, 2025)
- Novel Prehospital Prediction Model of Large Vessel Occlusion Using Artificial Neural Network. (Front Aging Neurosci, 2018)
- Optimal Transport Destination for Ischemic Stroke Patients With Unknown Vessel Status: Use of Prehospital Triage Scores. (Stroke, 2017)
- Validation of the Ultimate Stroke Scale (USS): A standardized tool for automated large vessel occlusion screening. (Interv Neuroradiol, 2025)
- Machine learning versus conventional methods for prehospital detection of stroke due to large vessel occlusion or intracranial haemorrhage. (Eur Stroke J, 2026)