Background: EBV is a necessary but not sufficient factor in the pathophysiology of multiple sclerosis (MS). EBV antibodies to the nuclear antigen (EBNA1) and viral capsid antigen (VCA) rise rapidly prior to MS disease manifestations, and their absence has clinical utility with a high negative predictive value. It remains unclear whether EBV levels act as prognostic, monitoring, or pharmacodynamic/response biomarkers. Substantial literature on this topic exists but has not been systematically reviewed. We hypothesized that EBV levels against EBNA1 and VCA are potential prognostic and monitoring biomarkers in MS, and that patient population, MS clinical phenotype, and EBV assay method may play important roles in explaining variation among study outcomes.

Methods: We systematically searched PubMed and EMBASE from inception to April 1, 2022. After removal of duplicates, records were screened by abstract. Remaining full-text articles were reviewed. Clinical and MRI data were extracted from full-text articles for comparison and synthesis.

Results: Searches yielded 696 unique results; 285 were reviewed in full, and 36 met criteria for data extraction. Heterogeneity in sample population, clinical outcome measures, assay methods and statistical analyses precluded a meta-analysis. EBV levels were not consistently associated with clinical disease markers including conversion from CIS to RRMS, neurological disability, or disease phenotype. Studies using repeated-measures design suggest that EBNA1 levels may temporarily reflect inflammatory disease activity as assessed by gadolinium-enhancing Magnetic Resonance Imaging (MRI) lesions. Limited data also suggest a decrease in EBV levels following initiation of certain disease-modifying therapies.

Conclusion: Heterogeneous methodology limited generalization and meta-analysis. EBV antibody levels are unlikely to represent prognostic biomarkers in MS. The areas of highest ongoing promise relate to diagnostic exclusion and pharmacodynamic/disease response. Use of EBV antibodies as biomarkers in clinical practice remains additionally limited by lack of methodological precision, reliability, and validation.