Subscribe to eTOC

New Biomarker Blood Test Might Flag TBI Not Evident on CT Scans

Article In Brief

Elevated levels of GFAP, a marker of axonal injury and traumatic brain injury, may help triage patients with more subtle injury undetected on CT, a new analysis suggests.


No caption available.

A simple blood test could potentially help triage and evaluate patients in the emergency department with suspected traumatic brain injury (TBI) and normal CT findings, according to an August 23 report in The Lancet Neurology.

The blood test, which is under development, measures the concentration of glial fibrillary acidic protein (GFAP), a protein considered to be a specific biomarker of astrocyte injury, which tends to be elevated with TBI.

Some patients with suspected TBI with normal CT findings show pathology on MRI, the authors of the study noted. So they wanted to assess whether the GFAP test could have the “discriminative ability to identify MRI abnormalities in patients with normal CT findings.”

The multi-center study included 450 suspected TBI patients who had normal CT scans and underwent testing with GFAP and MRI. MRI detected numerous head injuries missed by CT, and the analysis found that patients with the highest concentrations of GFAP were far more likely to have an MRI that indicated TBI.

“This study builds upon a lot of prior work and excitement about the promise of biomarkers in the treatment of traumatic brain injury,” said lead author Geoffrey Manley, MD, PhD, professor and vice chair of neurosurgery at University of California, San Francisco.

In the case of GFAP, “the biomarker appears to be more sensitive than a head CT for diagnosing traumatic brain injury,” he said.

The biomarker test, under development by Abbott and still in clinical testing, is conducted using a hand-held blood analyzer that delivers results in about 15 minutes, Dr. Manley said.

Another biomarker test for mild TBI (concussion) was given the go-ahead by the US Food and Drug Administration (FDA) last year, but it is more of a traditional lab test with results taking several hours. That test, marketed by Banyan Biomarkers, measures two biomarkers, GFAP and UCH-L1 and is intended for use in patients being considered for initial CT scanning.

Patients with a suspected head injury are typically given a CT scan to rule out brain swelling and bleeding. If the scan appears abnormal, further evaluation with MRI may be done. Dr. Manley said it is not unusual for patients to be sent home from the emergency department thinking they have nothing to worry about when they in fact have mild TBI.


Plasma GFAP concentration using a point-of-care platform (A) shows fair discrimination (AUC 0.777) between patients with MRI-positive findings and patients with MRI-negative findings in patients with CT-negative traumatic brain injury. GFAP at 9-16 hours post injury (B) shows improved discrimination to good—area under the curve (AUC) 0.852—suggesting temporal effects of GFAP release or accumulation in the systemic circulation after traumatic brain injury and time for optimal or repeat measurement.

Dr. Manley said that he does not believe that GFAP or any other biomarker will be the singular tool for diagnosing TBI. He expects that like with heart attack, where troponin testing is one component of evaluation, TBI biomarkers will be among several diagnostic tools used to arrive at a firm diagnosis.

“I see this as being a triage tool,” he said, but cautioned that the study included only adult patients who got to the emergency room because of a car accident or a fall. He said more research is needed before he could say whether the test “is ready for use in kids with concussions.”

A Subset Analysis

The latest report on the GFAP biomarker test stems from a study known as TRACK-TBI (Transforming Research and Clinical Knowledge in Traumatic Brain Injury), a large prospective study being conducted at 18 Level 1 trauma centers in the US. A sub-analysis was done of 450 of the participants, most of whom were classified as having mild TBI due to a traffic accident or fall. They all had normal findings on CT scan (and a Glasgow Coma Scale score 13-15) and consented to the blood test within 24 hours of their injury and an MRI scan seven to 18 days post-injury. The study also included orthopedic trauma patients and healthy controls for comparison.

Of the 450 suspected TBI patients with normal CT scans, 120 (27 percent) had MRIs indicative of TBI. When the researchers compared MRI findings to GFAP levels, they found that GFAP concentrations were higher in patients with positive MRI scans (median plasma GFAP concentration of 414.4 pg/mL) compared with those with negative MRI scans (median plasma GFAP concentration of 74.0 pg/mL).


“I see this as being a triage tool, but the study included only adult patients who got to the emergency room because of a car accident or a fall. More research is needed before I could say whether the test is ready for use in kids with concussions.”—DR. GEOFFREY MANLEY


“The new study on GFAP was well designed and informative, but I think biomarkers are quite a ways off from being a gold standard test.”—DR. MEERYO CHOE

Of the 90 patients with GFAP concentrations in the lowest quintile, only seven (8 percent) had abnormal MRI findings, while 58 (64 percent) of those with GFAP concentrations in the highest quartile had abnormal MRI findings, the study said. GFAP levels also correlated with MRI lesion types, the researchers said.

“These results indicate that the diagnostic utility of GFAP might extend beyond CT-visible pathology and might help to identify patients with more subtle injury,” the researchers said.

The study was funded by the National Institute of Neurological Disorders and Stroke and the US Department of Defense. Dr. Manley said the portable biomarker test might also one day prove useful for detecting TBI on the battlefield.

One of the limitations of the study was the lack of longer-term follow-up of patients to determine, for instance, whether those with high GFAP levels had more neurocognitive symptoms. There may have also been selection bias in the sub-analysis since it included only those TRACK-TBI participants who had MRI.

Better Diagnostics Are Needed

An editorial that accompanied the study in The Lancet Neurology underscored the need for better diagnostic tools for TBI.

“The ability to accurately diagnose the severity of traumatic brain injury (TBI) to establish the consequent need for treatment still eludes modern emergency care systems,” wrote Fiona Lecky, MD, PhD, of the University of Sheffield. She said that while a CT scan is good at detecting patients in need of neurosurgery or neurocritical care for life-threatening injury, “a CT brain scan is negative on more than 90 percent of patients with TBI symptoms, leading to uncertainty as to the extent of TBI and likely need for further treatment.”

Dr. Lecky said the findings that GFAP concentration at nine to 16 hours post-injury discriminated well between patients with MRI-positive findings and MRI-negative findings were positive.

“These results are promising,” she wrote, “but, as the investigators acknowledge, are some distance from suggesting that GFAP assays should become part of current emergency department management of patients with TBI and negative CT scans. The clinical significance of the positive MRI findings is unknown; data on patient symptom severity at 2 weeks or later follow-up were not reported in the study. If the frequencies of disabling concussion symptoms do not differ in the MRI-positive and MRI-negative cohorts during follow-up, it is hard to argue that either GFAP or the MRI scans are providing key information to guide further management.”

She noted that nearly half of the study participants with “negative CT scans did not have MRI as part of their follow-up, and because their characteristics are not reported, it is difficult to assess the potential for selection bias.”

In addition, she said it would be time consuming and expensive to do an MRI on every suspected TBI patient who had a normal CT.

Dr. Lecky said a biomarker test could be useful for sorting out which suspected TBI patients need head imaging and/or medical follow up, but she cautioned that the biomarker test needs to be further investigated before it is ready for clinical use.

Ramon Diaz-Arrastia, MD, PhD, FAAN, Presidential Professor of Neurology at the Perelman School of Medicine at the University of Pennsylvania and director of the Penn Clinical TBI Center, said it would be beneficial to have a biomarker test that could help eliminate unnecessary CT scans (and radiation exposure). But he said adopting a new routine screening test into emergency room workflow would take planning and a change in procedures.

“To be honest, I am not sure that US emergency room physicians would stop using CT scans,” in part because of medical-legal concerns, said Dr. Diaz-Arrastia, who was a coauthor of the current study. He said the biomarker test, if it becomes available, might be more quickly adopted in settings such as concussion clinics and urgent-care centers.

Dr. Diaz-Arrastia said he often sees patients who come in with neurocognitive issues long after they've had a minor head injury, some having been told that since their CT was normal there was no need for concern. He said the gaps in diagnosing mild TBI mean that important care instructions and follow-up assessments are not getting done.

“If you have a normal CT scan, all that means is you don't need a neurosurgeon, it doesn't always mean you're OK,” he said.

Meeryo Choe, MD, assistant clinical professor in UCLA's division of pediatric neurology, said, “The new study on GFAP was well designed and informative, but I think biomarkers are quite a ways off from being a gold standard test.”

She said future research needs to address whether elevated levels of GFAP correlate with longer recovery and worse outcomes. She said seeing how GFAP levels change over time after an injury might also be revealing.

“Eventually what we want to be able to do is correlate the test to the question ‘What does the patient look like clinically?’” said Dr. Choe, associate director of UCLA's Steve Tisch BrainSPORT Program.

She added that the ideal timing and setting for such a test is also a question. She noted that in the study, which drew from emergency department patients, nine to 16 hours post injury seemed to be ideal for the test, but that timing occurred because many of the participants ended up hospitalized for their injuries.

“The majority of patients with mild TBI will go to their primary care provider,” she said.


Drs. Manley, Diaz-Arrastia, and Lecky had no disclosures.

Link Up for More Information

• Yue JK, Yuh EL, Korley FK, et al; for the TRACK-TBI Investigators. Association between plasma GFAP concentrations and MRI abnormalities in patients with CT-negative traumatic brain injury in the TRACK-TBI cohort: A prospective multicentre study Lancet Neurol 2019; Epub 2019 Aug 23.
    • Lecky F. Comment: Should plasma GFAP guide the management of patients with traumatic brain injury and a negative CT scan Lancet Neurol 2019; Epub 2019 Aug 23.