Validation of Plasma Biomarkers and Stroke Scales for Rapid Prehospital Detection of Large-Vessel Occlusion Ischemic Stroke

Stroke remains a leading cause of morbidity and mortality worldwide, with large-vessel occlusion (LVO) ischemic stroke representing one of the most severe forms. LVO strokes account for a substantial portion of post-stroke disabilities and deaths, underscoring the critical need for rapid and accurate diagnosis. The advent of endovascular thrombectomy (EVT) has significantly improved outcomes for patients with LVO strokes. However, the effective deployment of EVT is contingent upon timely identification and triage of patients to EVT-capable centers, which are often limited in number and geographically dispersed. The prehospital setting presents a unique challenge in stroke care. Emergency medical services play a crucial role in the initial assessment and triage of stroke patients, but existing stroke severity scales, such as the National Institutes of Health Stroke Scale and others like the FAST-ED and RACE scales, have limitations in sensitivity and specificity. These scales often produce false negatives in LVO stroke patients and false positives in individuals with stroke mimics or hemorrhagic stroke. Consequently, the delay in accurately identifying LVO strokes in the field can lead to significant time losses, adversely affecting patient outcomes. Recognizing these challenges, researchers led by Professor Joshua Bernstock from Harvard Medical School and Brigham and Women’s Hospital aimed to enhance the diagnostic accuracy of prehospital stroke assessments. They proposed a novel approach that combines blood-based biomarkers specifically glial fibrillary acidic protein (GFAP) and d-dimer—with established stroke severity scales. GFAP is elevated in response to brain injury, particularly in hemorrhagic strokes, while d-dimer is associated with thrombotic events. The integration of these biomarkers with clinical scales aims to improve the identification of LVO strokes, thereby facilitating timely and appropriate triage to EVT-capable centers. The new research is now published in Stroke Vascular and Interventional Neurology. 

The researchers designed the TIME (Testing for Identification Markers of Stroke) trial as a prospective observational diagnostic accuracy study. Conducted at Brandon Regional Hospital in Florida from May 2021 to August 2022, the trial aimed to validate the use of plasma biomarkers, glial fibrillary acidic protein (GFAP) and d-dimer, in combination with established prehospital stroke scales for the accurate detection of LVO strokes. The trial included patients identified by ambulance crews with suspected strokes within 18 hours of symptom onset. Exclusion criteria included prior thrombolytic therapy, inability to provide a blood sample, symptom onset beyond 18 hours, or participation in concurrent clinical trials. Blood samples were collected and analyzed for plasma GFAP and d-dimer levels, using previously established cutoff values of 213 pg/mL and 600 ng/mL, respectively. These biomarkers were combined with prehospital stroke scales, such as the Field Assessment Stroke Triage for Emergency Destination (FAST-ED) and the Rapid Arterial Occlusion Evaluation (RACE) scale. The clinical data collected included demographics, clinical characteristics, laboratory results, and imaging findings. Patients were categorized into LVO ischemic stroke, non-LVO ischemic stroke, hemorrhagic stroke, transient ischemic attack, and stroke mimics based on computed tomography or magnetic resonance angiography confirmation. Out of 382 patients recruited, 323 were included in the final analysis, distributed as follows: LVO ischemic stroke, non-LVO ischemic stroke, hemorrhagic stroke, TIA, and stroke mimics. Combining plasma GFAP and d-dimer levels with prehospital stroke scales demonstrated significant diagnostic performance. The specificity of detecting LVO strokes was 94%, and sensitivity was 71%. These figures improved in a subanalysis focusing on patients presenting within 6 hours of symptom onset, achieving a specificity of 93% and a sensitivity of 81%. Importantly, this approach successfully ruled out all cases of hemorrhagic stroke, which is crucial for avoiding inappropriate treatment pathways.

The authors showed that combination of blood biomarkers and stroke scales yielded the highest diagnostic performance when GFAP and d-dimer were integrated with the FAST-ED and RACE scales. For instance, using the FAST-ED scale in combination with the biomarkers resulted in a specificity of 94% and sensitivity of 71% for LVO detection. Similarly, combining biomarkers with the RACE scale also produced high specificity and sensitivity. A key finding from the subgroup analysis was the enhanced performance within the critical early time window. For patients presenting within 6 hours from symptom onset, the combination of biomarkers and the FAST-ED scale achieved a specificity of 93% and sensitivity of 81%. This improvement highlights the importance of early detection and intervention, as delays in treatment can significantly impact patient outcomes.  The researchers’ analysis also revealed that plasma d-dimer levels were significantly higher in patients with LVO strokes compared to non-LVO suspected stroke patients. GFAP levels were markedly increased in patients with hemorrhagic stroke compared to those with nonhemorrhagic strokes. These findings underscore the potential of these biomarkers to differentiate between stroke types effectively.

The integration of GFAP and d-dimer with stroke severity scales has significant clinical implications. It enables more accurate prehospital identification of LVO strokes, facilitating timely triage to EVT-capable centers and reducing delays in treatment. The high specificity in ruling out hemorrhagic strokes also ensures that patients receive appropriate care pathways, avoiding unnecessary and potentially harmful interventions.

In conclusion Professor Joshua Bernstock and colleagues demonstrated that  combination of GFAP and d-dimer with stroke scales such as FAST-ED and RACE significantly improves the sensitivity and specificity of LVO stroke detection. This level of diagnostic precision is crucial for ensuring that patients with LVO strokes are accurately identified and directed to appropriate treatment centers promptly. One of the most significant findings is the potential to reduce delays in treatment. Early and accurate identification of LVO strokes can expedite the routing of patients to endovascular thrombectomy (EVT)-capable centers, thus reducing the time from stroke onset to treatment. The study highlights that for every 15-minute reduction in time to EVT, there is a corresponding 3.9% decrease in long-term disability. Therefore, the ability to quickly and accurately identify LVO strokes in the prehospital setting has profound implications for patient outcomes. By facilitating timely triage and treatment, the study’s approach has the potential to improve functional outcomes and reduce mortality among stroke patients. The validated biomarkers and scales provide a robust method for distinguishing between different types of strokes, including hemorrhagic strokes, which require different management strategies. This ensures that patients receive the most appropriate care quickly, which is critical in the context of stroke, where “time is brain.” Moreover, the study paves the way for the development of point-of-care diagnostic tools that can be used in ambulances and emergency departments. Portable devices capable of measuring GFAP and d-dimer levels rapidly could revolutionize prehospital stroke care, providing EMS teams with the data needed to make critical decisions on the spot.