Identify associations between plasma biomarkers of CNS injury and clinical phenotype in multiple sclerosis (MS) patients from an underrepresented minority population.

Plasma samples were obtained from the UIC Neuroimmunology Biobank. 53 patients with MS provided informed consent. 66% of donors identified as Black or African American and 19% as White, Hispanic American. Plasma NFL and GFAP levels were measured using the ultra-sensitive Simoa assay (Quanterix, Billerica MA). Clinical data, including history and neurological examination, were abstracted from the clinical visit notes. The Functional System Score (FSS) and Expanded Disability Status Scale (EDSS) were calculated for each patient.  Clinical phenotyping was performed by abstraction of signs and symptoms from the notes and then mapping the terms in a neurological examination ontology (Hier and Brint, 2020). 

Abstraction of the signs and symptoms from the 53 MS donors yielded 77 unique phenotypic features.  Factor analysis was used to reduce the 77 phenotypic features into seven factors (brainstem, cognitive, cerebellar, long-tract, motor, sensory, and visual). African American patients had higher motor factor scores and cerebellar FSS than White MS patients. For all MS patients, NFL levels correlated positively with EDSS, cerebellar FSS, brainstem FSS, cognitive factor, cerebellar factor, and motor factor. GFAP levels correlated positively with EDSS, cerebellar FSS, mental FSS, and cognitive factor. Age-adjusted NFL and GFAP levels correlated most significantly with cerebellar and cognitive factors, respectively (P<0.01, Pearson r, two-tailed).