The Crossroads of Neuroinflammation and Biomarkers in Multiple Sclerosis: A Systematic Review

The management of multiple sclerosis (MS) is shifting from a phenotype-based framework toward a biologically driven precision medicine model, as conventional magnetic resonance imaging (MRI) inadequately captures smoldering inflammation and progression independent of relapse activity (PIRA). This systematic review aimed to synthesize current evidence on the diagnostic and prognostic utility of fluid biomarkers in distinguishing acute inflammatory injury from chronic neurodegeneration. A comprehensive search of Web of Science, PubMed, and Scopus (January 2020–September 2025) identified 28 eligible studies including 7775 participants (6365 MS patients and 1410 controls). Biomarkers derived from serum, plasma, cerebrospinal fluid (CSF), and stool were evaluated in relation to clinical disability measured using the Expanded Disability Status Scale (EDSS) and magnetic resonance imaging (MRI) outcomes. Neurofilament light chain (NfL) consistently predicted acute inflammatory activity, gadolinium-enhancing lesions, and relapse-associated worsening, but levels were reduced by high-efficacy therapies and did not reliably predict PIRA. In contrast, glial fibrillary acidic protein (GFAP) was associated with astrogliosis, disability progression, and retinal thinning, even in patients with low inflammatory activity. Additional CSF, metabolic, and immunologic markers correlated with neurodegeneration and disease severity. Nevertheless, broader clinical use will require greater assay standardization, improved consistency across cohorts, and validation in prospective longitudinal studies. These findings compel a shift toward a multi-biomarker model to guide personalized therapeutic strategies and develop targeted neuroprotective treatments for progressive multiple sclerosis.