Using non-hypothesized based approaches for biomarker development

Current protein biomarkers are only moderately predictive in identifying individuals with mild traumatic brain injury or concussion. Therefore, more accurate diagnostic markers are needed for sport-related concussion (SRC). A case-control study of athletes who provided blood samples and were diagnosed with a concussion or were a matched non-concussed control within the NCAA-DoD Concussion Assessment, Research and Education (CARE) Consortium. The blood was collected within 48 hours of injury. Athletes with concussion were divided into 6 hours post-injury (0-6 hours post-injury) and after 6 hours post-injury (7 to 48 hours post-injury) groups. We applied a highly multiplexed proteomic technique that used DNA aptamers assay to target 1,305 proteins in plasma samples from athletes with and without a SRC. We identified 338 plasma proteins significantly differed in abundance in concussed athletes compared to non-concussed athletes. The top 20 most differentially abundant proteins discriminated concussed athletes from non-concussed athletes with an area under the curve (AUC) of 0.954 (95% confidence interval 0.922 to 0.986). Specifically, after 6 hours of injury, the individual AUC of plasma erythrocyte membrane protein band 4.1 (EPB41) was 0.956 and 0.875 for alpha synuclein (SNCA).

 

Learning Objectives