A Groundbreaking Shift in MS Understanding

Multiple sclerosis (MS) has traditionally been categorized into distinct subtypes like "relapsing" and "progressive." However, a groundbreaking study published on August 20, 2025, in *Nature Medicine*, spearheaded by the Medical Center—University of Freiburg and the University of Oxford, is turning this understanding on its head.

This international research, built on the expansive NO.MS cohort data from Novartis, reveals that MS should be viewed as a dynamic continuum rather than a static set of categories.

AI Unveils Four Key Dimensions of MS

Instead of adhering to rigid phenotype classifications, the study introduces an AI-based model that identifies four crucial dimensions to track MS progression: physical disability, brain damage, clinical relapses, and silent inflammatory activity. This fresh perspective could revolutionize the way we diagnose and treat MS patients, and might even extend to other diseases.

Professor Heinz Wiendl, Medical Director of the Department of Neurology and Neurophysiology at Freiburg, asserts, "Our findings clearly demonstrate that MS cannot be simply categorized as relapsing or progressive; it is a continuous disease process with distinct transitions."

Dynamic Disease Progression Revealed

This research analyzes data from over 8,000 patients and more than 35,000 MRI scans from various studies, including the NO.MS and Roche Ocrelizumab cohorts.

The AI model outlines MS as a series of states with probabilities guiding transitions. Early, less severe states progress through inflammatory intermediates to more advanced, irreversible stages. Notably, the model indicates that progressing directly to severe stages, without any prior inflammatory signs, is almost impossible—highlighting the significance of silent, symptom-free inflammation.

Empowering Personalized Treatment

The old classification system often restricts access to effective treatments due to its adherence to strict subtype definitions. The new model encourages personalized risk assessments, independent of the diagnosed subtype.

Wiendl emphasizes, "Rather than categorizing patients, we should quantify their condition and monitor it dynamically." This is especially critical for patients with active yet clinically silent inflammatory activity, emphasizing the necessity for timely treatment decisions.

Beyond MS: A New Era in Disease Classification

This innovative state-based model isn't just a leap forward for MS research; Prof. Lutz Hein, Dean of the Faculty of Medicine at Freiburg, notes that its application could extend across various diseases in neurology and beyond.

He insists that the focus must shift from rigid categories to flexible, data-driven disease states.

Prof. Peter Berlit, Secretary General of the German Society of Neurology, adds, "The next crucial step is to implement these individualized risk assessments in clinical practice while gathering prospective data."