A blood test may have the ability to rapidly distinguish haemorrhagic strokes (brain bleeds) from ischaemic (clot-caused) strokes, even before patients with stroke symptoms reach the emergency room, according to a preliminary study set to be presented at the 2025 International Stroke Conference (ISC; 5–7 February, Los Angeles, USA).
“It is crucial to differentiate these two types of stroke, because they need opposite treatments,” said lead study author Love-Preet Kalra (RKH Hospital Klinikum Ludwigsburg, Ludwigsburg, Germany). “In ischaemic stroke, you need to open the blocked blood vessel with clot-busting drugs or physically remove the clot. In contrast, in a bleeding stroke, you need to lower increased blood pressure and give medication to reverse the effects of certain blood-thinning drugs.”
In the present study, researchers examined whether blood levels of glial fibrillary acidic protein (GFAP) could be useful for quickly diagnosing stroke types. GFAP is a protein specific to the brain that is released into the bloodstream when brain cells are damaged or destroyed. It is already used in assessing traumatic brain injuries (TBIs).
The study included 353 people (average age, 75 years; 47% women) reaching the emergency room within six hours of the onset of stroke symptoms. People were excluded from the analysis if they had previously had a brain tumour, or if they had experienced a stroke or TBI within the past three months. Brain imaging diagnosed intracerebral haemorrhage in 76 patients, clot-caused strokes in 258 patients, and conditions mimicking stroke—such as seizure or migraine—in 19 patients. Blood was drawn before patients reached the emergency room at RKH Hospital Klinikum Ludwigsburg and tested for GFAP levels using a portable blood analyser, with GFAP levels being compared between groups based on diagnoses at hospital discharge.
In a parallel study published in 2024, Kalra and colleagues found that GFAP levels could rapidly distinguish who had a bleeding stroke among unresponsive patients. In this study, researchers evaluated whether levels of GFAP could differentiate between haemorrhagic and ischaemic strokes, as well as conditions that mimic a stroke. This assessment was conducted using blood samples collected by the emergency medical services ambulance team before patients arrived at the hospital.
The analysis found that GFAP levels were:
- Almost seven times higher in patients with bleeding stroke than in those with clot-caused stroke (208pg/mL vs 30pg/mL)
- More than four times higher in patients with bleeding stroke than in those with stroke mimics (208pg/mL vs 48pg/mL)
- Able to rule out bleeding stroke when below 30pg/mL in patients with moderate-to-severe neurological deficits
- Able to predict which patients had a bleeding stroke with 90–95% accuracy when age-based cutoff points were used—these age-based groups were evenly distributed: below 72 years, between 72 and 83 years, and above 83 years, with the cutoff for those below 72 years being very low
- Higher in bleeding stroke patients taking blood thinners than those not on blood-thinning medications
“I was personally surprised by the extremely elevated GFAP values in blood thinner-associated bleeding stroke and the fact that, in moderately or severely affected acute stroke patients, bleeding stroke could be excluded in all cases which showed a GFAP lower than 30pg/mL,” Kalra stated.
If larger studies confirm the results, Kalra believes early GFAP measurements could change how people with stroke symptoms are treated.
“Treatment to lower blood pressure and reverse blood-thinning medications could be performed in the prehospital setting, leading to a huge change in clinical practice,” she commented. “In the future, even blood thinners or clot-busting treatment might be applied before people reach the hospital.”
According to the researchers, a limitation of this test is that a centrifugation step—which involves separating the components of blood—is currently needed. In addition, GFAP increases with age, creating a grey area in which small bleeding strokes might not be identified or may be mistaken for ischaemic strokes in elderly patients.
“This study reveals that levels of GFAP, a marker for brain injury, are higher in patients with brain haemorrhages compared to those who have strokes caused by blood clots,” said American Heart Association (AHA) expert volunteer Louise McCullough (McGovern Medical School at UTHealth Houston, Houston, USA), who was not involved in this study. “This finding suggests that GFAP could serve as a useful prehospital test for assessing brain injuries. However, the study had a relatively small sample size and, for the test to be effective, both the patient’s blood and the GFAP test must be available as a ‘point-of-care’ test in the field. Currently, most ambulances and emergency medical services do not have access to this blood test.”
