SHORT DESCRIPTION A non-invasive blood biomarker technology that analyzes plasma cfDNA methylation to diagnose multiple sclerosis, classify its subtypes, and predict disease progression.
NU Tech ID NU 2025-030
IP STATUS
US Patent Pending (Joint with University of Pittsburgh)
DEVELOPMENT STAGE
TRL-4: Prototype Validated in Lab: Key functions have been demonstrated in controlled laboratory settings using clinical datasets.
BACKGROUND Multiple sclerosis (MS) is a chronic autoimmune disease caused by inflammatory demyelination of nerve fibers that results in damage to the central nervous system. MS is a highly heterogenous disease with diverse clinical courses, posing significant challenges for early diagnosis, subtype classification, and long-term prognostic prediction. Most individuals are initially diagnosed with Relapsing-Remitting MS (RRMS), defined by discrete episodes of acute neurological symptoms followed by partial or full recovery. Some individuals will subsequently transition to Secondary Progressive MS (SPMS) and a small subpopulation of patients develop Primary Progressive MS (PPMS) from the onset. Patients with progressive MS (PMS), including both PPMS and SPMS, experience worsening neurological impairment without discrete periods of relapse or remission. Timely diagnosis and subclassification is key for developing appropriate treatment plans, which may include disease-modifying therapies (DMTs) that have been approved to treat RRMS. Current diagnostic methods rely on magnetic resonance imaging and invasive cerebrospinal fluid (CSF) analysis, which can be costly, difficult to access, and uncomfortable for patients. Existing blood biomarkers do not fully capture the disease’s complex pathology, resulting in suboptimal treatment decisions. There remains a pressing need for non-invasive diagnostic tools to enable rapid and accurate MS classification and prognosis for personalized patient monitoring and treatment.
ABSTRACT Northwestern and University of Pittsburgh researchers have developed a highly accurate molecular diagnostic platform that uses cell-free DNA (cfDNA) methylation profiles from patient plasma to identify current disease state and predict disease outcome. This technology utilizes whole-genome bisulfite sequencing (WGBS) of plasma cfDNA to identify specific epigenetic signatures that distinguish MS patients from healthy individuals, differentiate clinical subtypes, and stratify patients by disease severity. The system incorporates computational deconvolution using reference methylation atlases to perform tissue-of-origin (TOO) analysis, estimating the cellular sources of circulating cfDNA. Supervised machine learning classifiers trained on these epigenetic signatures and TOO features categorize disease states, while a linear mixed-effects model calculates a methylation-based prognostic risk score (MBPRS) from baseline methylation at specific prognostic regions to predict future disability progression. Initial validation of the diagnostic platform using clinical datasets shows superior performance compared to existing biomarkers such as neurofilament light chain (NfL) and glial fibrillary acidic protein (GFAP). The approach provides a single, comprehensive assay to non-invasively and simultaneously diagnose MS, stratify patients by severity, and accurately forecast long-term outcomes.
APPLICATIONS
ADVANTAGES
PUBLICATIONS
KEYWORDS cfDNA methylation, multiple sclerosis, noninvasive diagnosis, blood biomarkers, machine learning, prognosis, neurodegeneration