Serum GFAP for stroke diagnosis in regions with limited access to brain imaging (BE FAST India)

Abstract

Introduction Despite a high burden of stroke, access to rapid brain imaging is limited in many middle- and low-income countries. Previous studies have described the astroglial protein GFAP (glial fibrillary acidic protein) as a biomarker of intracerebral hemorrhage. The aim of this study was to test the diagnostic accuracy of GFAP for ruling out intracranial hemorrhage in a prospective cohort of Indian stroke patients. Patients and methods This study was conducted in an Indian tertiary hospital (Christian Medical College, Ludhiana). Patients with symptoms suggestive of acute stroke admitted within 12 h of symptom onset were enrolled. Blood samples were collected at hospital admission. Single Molecule Array technology was used for determining serum GFAP concentrations. Results A total number of 155 patients were included (70 intracranial hemorrhage, 75 ischemic stroke, 10 stroke mimics). GFAP serum concentrations were elevated in intracranial hemorrhage patients compared to ischemic stroke patients [median (interquartile range) 2.36 µg/L (0.61–7.16) vs. 0.18 µg/L (0.11–0.38), p < 0.001]. Stroke mimics patients had a median GFAP serum level of 0.14 µg/L (0.09–0.26). GFAP values below the cut-off of 0.33 µg/L (area under the curve 0.871) ruled out intracranial hemorrhage with a negative predictive value of 89.7%, (at a sensitivity for detecting intracranial hemorrhage of 90.0%). Discussion The high negative predictive value of a GFAP test system allows ruling out patients with intracranial hemorrhage. Conclusion In settings where immediate brain imaging is not available, this would enable to implement secondary prevention (e.g., aspirin) in suspected ischemic stroke patients as soon as possible.