EBM Review Series: Diagnostic Performance of GFAP, UCH-L1, and MAP-2 Within 30 and 60 Minutes of Traumatic Brain Injury
VACEP Evidence-Based Medicine for General Emergency Physicians Series
Authors: Paige Darrow, DO PGY-3, Martin D. Klinkhammer, MD, MPH, FACEP | Eastern Virginia Medical School
Reviewer: Sam Reed, MD PGY-2, Moira Smith, MD, MPH | UVA Health
The VACEP Evidence-Based Medicine Review Series allows Virginia emergency medicine residents and attendings to share and analyze a recent peer-reviewed clinical study. You can also read the full article, “Diagnostic Performance of GFAP, UCH-L1, and MAP-2 Within 30 and 60 Minutes of Traumatic Brain Injury” from the Journal of the American Medical Association, Volume 7 Number 9.
THE CASE
You are working in a lower resource emergency department. You receive a radio call from EMS regarding a patient found by the side of the road in his car. It appeared that his vehicle rolled to a stop after running off the road through some shrubbery. There was no airbag deployment, and he was found asleep at the scene. Medics report he smells of alcohol and his current GCS is 9. They would like to bring the patient to your ED, but are asking if you would like to divert the patient to the trauma center 1.5 hours in the opposite direction.
What role might biomarkers play in the care of the patient?
BACKGROUND
Traumatic brain injury (TBI) is a major cause of morbidity and mortality in the US. There were 214,110 TBI-related hospitalizations and 69,473 TBI-related deaths in 2021. It is a diagnosis with a wide spectrum of presentations and a range of workups, clinical decision-making tools, imaging including computed tomography (CT), and management.
Biomarkers now stand as a potential new tool in the evaluation of this condition. Biomarkers have a significant role in current neuroscience research, even more so since the 2018 FDA approval of the first blood test utilizing glial fibrillary acidic protein (GFAP) and ubiquitin C-terminal hydrolase (UCH-L1). The approved use is determining which mild traumatic brain injuries do not require a head CT; however, the test assay time of 3 hours made it impractical for clinical application. There are newer assays with 15–20 minute turnaround times, offering a new opportunity for biomarker utilization. Major biomarkers studied include GFAP, UCH-L1, glial specific binding protein (S100B), and microtubule associated protein-2 (MAP-2).
Some biomarkers are already in use in clinical practice in Europe. One study analyzed S100B and its use to screen for TBI and help differentiate which lower risk patients may not require a CT (Unden, 2010). This meta-analysis noted that low serum S100B with a cut-off of 0.10 ug/L had a sensitivity of 97% (0.91-0.99), specificity of 40% (0.30-0.51), and a negative predictive value (NPV) of 99% (0.98-100) for predicting clinically significant head CT findings. Among study patients, utilizing the test at this threshold would lead to a theoretical reduction in head CT use by 32%. The authors’ conclusion was that S100B can successfully screen for significant head injury if utilized in minor head injuries within 3 hours. Using serum S100B for this indication is currently not approved in the US but is used in Europe.
Another study assessed the use of UCH-L1 and GFAP and their role in screening for the need for head CT in lower risk TBI patients (Bazarian, 2018). This study was a prospective, multicenter observational trial of adults with head injury and GCS ≥ 9. It looked at 1959 patients, 125 of whom had significant CT-detected injuries. All patients received a head CT and had to have blood work drawn within 12 hours of presentation. Utilizing both UCH-L1 and GFAP with cutoffs of 22 pg/ml and 327 pg/ml respectively, led to a sensitivity of 97.6% (CI 93.1-99.5) a specificity of 36.4% (CI 34.2-38.7) and a NPV 99.6% (98.7-99.9) for any injury on head CT. Finally, the markers were 100% accurate in excluding neurosurgically managed lesions. Both of these studies support the use of biomarkers to screen for clinically significant injury in TBI patients. A key next step for biomarker use in TBI patients is to compare their use to existing clinical decision rules that also seek to obviate the need for head CT.
A prospective cohort study was performed at a level 1 trauma center evaluating the Canadian CT head rule (CCHR), New Orleans Criteria (NOC) and National Emergency X-Radiography Utilization Study II (NEXUS II) as implemented in real time and compared them against GFAP and UCH-L1 biomarkers in patients with GCS of 13 or greater and meeting other criteria such as systolic blood pressure (SBP) of 100 mmHg or greater (Papa, 2022). The NOC and CCHR rules had superior sensitivity in detecting CT lesions over the biomarkers alone. However, when combined with a biomarker every clinical decision rule performance was improved. Additionally, the study assessed physician comfort with the rules and many physicians felt better about decisions made when a lab value was incorporated. Overall, the above study suggests a potential utility of biomarker studies in the care of TBI patients in the emergency department.
STUDY SUMMARY
This leads us to an article that assessed the utility of serum biomarker assays in TBI management almost immediately after ED presentation (Papa, 2024). The goal of the study was to compare how different biomarkers–- GFAP, UCH-L1, and MAP2–- measured within 30-60 minutes of emergency department presentation performed in identifying intracranial lesions on CT as well as clinically important outcomes.
This was a cohort study of a separate multicenter double-blinded randomized clinical trial of pre-hospital tranexamic acid (TXA) held across 20 medical centers and 39 emergency medical systems in the US and Canada (Rowell, 2020). Using the same patients, this study drew blood samples from all patients enrolled in the TXA study. That trial was randomized into three groups, with two treatment groups receiving TXA, but it was previously demonstrated that TXA has no effect on the TBI-related biomarkers so all three groups were included in this cohort study (Hoefer, 2024). The labs were drawn either within 30 or 60 minutes from ED arrival and within 4 hours of initial injury. The analysis included 563 patients at 60 minutes and 375 at 30 mins. Patients included in the study were hemodynamically stable, >15 years old, had an initial Glasgow Coma Scale (GCS) range of 3-12 and at least one reactive pupil. The lab samples were frozen and held at a facility until sent to a central lab where GFAP, UCH-L1 and MAP-2 levels were measured via ELIZA. The outcome measures for the study included the primary outcome of intracranial lesions on head CTs performed on arrival to the hospital. Secondary outcome was the need for neurosurgical intervention within 24 hours of injury. Tertiary outcome was a combined endpoint of clinically important outcomes (CIEO) including death, intubation for greater than 7 days, or neurosurgical intervention. Receiver operating curves (ROC) were created with each biomarker to explore its ability to predict each of the study outcomes.
About the EBM Review Series
This is a literature review series started by the University of Virginia’s Josh Easter, MD, MSc, a VACEP board member working to connect the academic community in Virginia. We invite each residency in Virginia (and D.C.) to create a faculty/resident team to submit and review articles. Sign up to submit one.
Goals
Provide a brief monthly synopsis of a high yield article germane to the practice of emergency medicine for distribution to all VACEP members
Provide an opportunity for a peer reviewed publication and invited presentation for faculty and trainees
Foster an academic community focused on evidenced based medicine for emergency medicine residency programs in the region
STRENGTHS and LIMITATIONS:
Strengths:
Strengths of the study are that it is one of the first biomarker studies to look predominantly at a very sick patient population, and the labs being drawn in the hyperacute setting (within 60 minutes of ED arrival). While the lack of greater than 98% sensitivity may limit current biomarker use as a rule out test (the lower limit of the confidence interval would indicate it may miss upwards of 10% of significant findings on CT) this study did show that the biomarkers currently being studied do generally rise in the hyperacute setting. This indicates that it may have clinical utility in select situations, such as prognosticating patient risk of clinically significant outcomes.
Weaknesses:
Weaknesses of this particular study include time-to-biomarker getting drawn, male-to-female ratio of the study (only a quarter of patients were female), and the GSC of enrolled patients all being 3-12. Ideally these biomarkers would be best implemented in the higher GCS patients in whom they could potentially be utilized to make a decision on whether to get a head CT, whereas lower GCS patients typically indicate a need for a scan either way.
RESULTS:
The main result of the study was that GFAP individually outperformed all other biomarkers, it had the strongest independent association for estimating outcomes of the study. In the AUCs for identifying CT lesions at 30 and 60 minutes, GFAP demonstrated best performance for samples at 30 minutes with AUC 0.88 (95% CI, 0.85-0.92), sensitivity 98.1% (95% CI, 0.94-0.99) and specificity 34.4% (95% CI 0.27-0.42). Similar results were identified in samples at 60 minutes. This was followed by MAP-2 and UCH-L1, combinations of biomarkers were not better than GFAP alone. A GFAP level of 30 pg/ml (which is the FDA approved threshold) resulted in a NPV of 95.7% (95% CI 88.8-98.6). Elevated serum GFAP also had the highest association for needing a neurosurgical intervention within 24 hours of injury and strongest association of CIEO within 7 days of injury with an AUC of 0.89 (95% CI, 0.85-0.93).
CASE CONCLUSIONS:
Concentrations of GFAP, UCH-L1, and MAP-2 measured within 30 and 60 minutes of injury were significantly associated with presence of lesions on CT scan, degree of injury on scan, need for neurosurgical intervention within 24 hours, and a CIEO within 7 days of injury. Overall, GFAP had the strongest association of the biomarkers and the GFAP thresholds identified offer a foundation for future assays.
Utilizing biomarkers in situations where GCS is limited may help in assessing severity of injuries. GFAP and/or SB100 may have a future role in risk stratification for patients early after injury including level of care, early interventions, and monitoring. GFAP offers the biggest potential for a point of care (POC) test, thus potentially being useful in guiding triage, military, and prehospital decisions.
TAKE-HOME DISCUSSION:
Biomarkers could potentially offer a way to strengthen our clinical decision rules when managing TBIs and trauma. While they may ultimately help cut down on unnecessary CTs, this study showed that GFAP, UCH-L1, and MAP-2 are not ready for that application due to somewhat limited sensitivity and a high rate of false positives comparable to clinical decision rules (SB100 was not used in this study). However, if used as a POC test in the field, the study supports consideration of their use in strengthening triage ability and escalation of care after head injury. They could potentially help guide EMS transport to more critical centers and lead to better escalation of care and interventions for severe TBI patients. Of the biomarkers currently studied, GFAP and SB100 have the most robust data supporting their future research.
Back to the radio call, given the patient’s GCS of 9, even though alcohol is the suspected cause, due to the lack of clarity you decided to divert the patient 1.5 hours down the road. The patient began to awaken further en route to the trauma center, had a negative head CT, and was ultimately discharged 6 hours later.
REFERENCES
Undén J, Romner B. Can low serum levels of S100B predict normal CT findings after minor head injury in adults?: an evidence-based review and meta-analysis. J Head Trauma Rehabil. 2010 Jul-Aug;25(4):228-40. doi: 10.1097/HTR.0b013e3181e57e22. PMID: 20611042.
Bazarian JJ, Biberthaler P, Welch RD, Lewis LM, Barzo P, Bogner-Flatz V, Gunnar Brolinson P, Büki A, Chen JY, Christenson RH, Hack D, Huff JS, Johar S, Jordan JD, Leidel BA, Lindner T, Ludington E, Okonkwo DO, Ornato J, Peacock WF, Schmidt K, Tyndall JA, Vossough A, Jagoda AS. Serum GFAP and UCH-L1 for prediction of absence of intracranial injuries on head CT (ALERT-TBI): a multicentre observational study. Lancet Neurol. 2018 Sep;17(9):782-789. doi: 10.1016/S1474-4422(18)30231-X. Epub 2018 Jul 24. PMID: 30054151.
Papa L, Ladde JG, O'Brien JF, Thundiyil JG, Tesar J, Leech S, Cassidy DD, Roa J, Hunter C, Miller S, Baker S, Parrish GA, Davison J, Van Dillen C, Ralls GA, Briscoe J, Falk JL, Weber K, Giordano PA. Evaluation of Glial and Neuronal Blood Biomarkers Compared With Clinical Decision Rules in Assessing the Need for Computed Tomography in Patients With Mild Traumatic Brain Injury. JAMA Netw Open. 2022 Mar 1;5(3):e221302. doi: 10.1001/jamanetworkopen.2022.1302. PMID: 35285924; PMCID: PMC9907341.
Papa L, McKinley WI, Valadka AB, Newman ZC, Nordgren RK, Pramuka PE, Barbosa CE, Brito AMP, Loss LJ, Tinoco-Garcia L, Hinson HE, Schreiber MA, Rowell SE. Diagnostic Performance of GFAP, UCH-L1, and MAP-2 Within 30 and 60 Minutes of Traumatic Brain Injury. JAMA Netw Open. 2024 Sep 3;7(9):e2431115. doi: 10.1001/jamanetworkopen.2024.31115. PMID: 39230905; PMCID: PMC11375473.
Rowell SE, Meier EN, McKnight B, Kannas D, May S, Sheehan K, Bulger EM, Idris AH, Christenson J, Morrison LJ, Frascone RJ, Bosarge PL, Colella MR, Johannigman J, Cotton BA, Callum J, McMullan J, Dries DJ, Tibbs B, Richmond NJ, Weisfeldt ML, Tallon JM, Garrett JS, Zielinski MD, Aufderheide TP, Gandhi RR, Schlamp R, Robinson BRH, Jui J, Klein L, Rizoli S, Gamber M, Fleming M, Hwang J, Vincent LE, Williams C, Hendrickson A, Simonson R, Klotz P, Sopko G, Witham W, Ferrara M, Schreiber MA. Effect of Out-of-Hospital Tranexamic Acid vs Placebo on 6-Month Functional Neurologic Outcomes in Patients With Moderate or Severe Traumatic Brain Injury. JAMA. 2020 Sep 8;324(10):961-974. doi: 10.1001/jama.2020.8958. Erratum in: JAMA. 2020 Oct 27;324(16):1683. doi: 10.1001/jama.2020.20143. PMID: 32897344; PMCID: PMC7489866.
Hoefer LE, Benjamin AJ, Polcari AM, Schreiber MA, Zakrison TL, Rowell SE. TXA does not affect levels of TBI-related biomarkers in blunt TBI with ICH: A secondary analysis of the prehospital TXA for TBI trial. J Trauma Acute Care Surg. 2024 Jan 1;96(1):94-100. doi: 10.1097/TA.0000000000004130. Epub 2023 Oct 9. PMID: 37807179; PMCID: PMC10840876.
