SPINCOMS Biomarker Study

To determine if biomarker-based CSF testing is reliably detecting differences between patients with Multiple Sclerosis (MS), different MS-subtypes, and other central nervous system (CNS) diseases. This study will also look to identify biomarkers that could be used for the prediction, at the time of diagnosis, of the future disease clinical course and response to therapy. The SOMAscan assay will be used for CSF samples analysis.

Using machine learning, the investigators have developed from SOMAScan: 1. A molecular diagnostic test that differentiates MS from other inflammatory and non-inflammatory central nervous system (CNS) diseases (area under receiver-operator characteristic curve-AUROC of 0.98); 2. A molecular test that differentiates relapsing-remitting MS from progressive MS variants (AUROC of 0.91); and 3. A molecular test that predicts future rates of disability progression, concordance coefficient of 0.425 (p\<0.001). Because these results are derived from a single research center (NIAID/NDS), it is imperative to determine their performance in real clinical practice settings as a necessary step for their potential regulatory approval. Consequently, his application has 2 specific aims: AIM 1. To independently validate afore-mentioned CSF-biomarker-based tests for their clinical value within the multicenter Spinal fluid Consortium for MS (SPINCOMS). In Aim 1, each of the 3 defined tests will be validated in 100 new SPINCOMS patients. To validate the prognostic test, 100 MS patients with CSF collected at least 3 years ago will be evaluated at follow-up examination with standardized clinical outcomes. CSF will be analyzed blinded using pre-defined statistical models. AIM 2. To explore whether collected CSF-biomarkers point towards pathogenic heterogeneity that may predict patient-specific efficacy for different disease-modifying treatments (DMTs) or identify pathogenic mechanisms not targeted by current DMTs. In Aim 2, clustering analysis will assess pathogenic heterogeneity and explore potential predictors of response to therapy.