A new study published in the journal Brain This highlights a significant advance in understanding multiple sclerosis (MS): behind a disease known for its high variability, two distinct biological subtypes appear to be emerging. Researchers arrived at this classification by combining brain MRI data and a blood marker, serum neurofilament light (sNfL), using an artificial intelligence model.
Multiple sclerosis (MS), an autoimmune and neurodegenerative disease, affects approximately 130,000 people in France and is one of the leading causes of severe non-traumatic disability in young adults. However, its progression remains difficult to predict: some patients experience rapid inflammatory flare-ups, while others undergo a more silent but destructive progression. The scientific challenge, therefore, is to better characterize the disease, beyond just its symptoms.
Two profiles: early inflammation or degeneration first
The AI model used (SuStaIn), applied to data from approximately 600 patients, distinguished two forms. In the so-called "early sNfL" profile, neurofilament levels increase early, in parallel with rapid brain lesions, suggesting a more aggressive and inflammatory disease. Conversely, the "late sNfL" profile initially shows a loss of brain volume (atrophy), a sign of an initial neurodegenerative mechanism, before blood markers rise.
This distinction aligns with long-standing clinical observations that are difficult to formalize. For neurologist Céline Louapre, it would allow us to "predict the future in a certain way," even though experts urge caution: an MRI and a blood test are not yet sufficient, at this stage, to predict an individual patient's progression. Neurofilament testing, in particular, is not yet used routinely.
Towards more personalized care and new therapeutic approaches
The potential benefit is significant, as most current treatments primarily target inflammation and reduce relapses, without truly repairing myelin or slowing neuronal degeneration. Identifying a patient's trajectory earlier could prevent wasting time on poorly suited treatments and allow for more intensive monitoring of at-risk individuals.
In parallel, research on progressive forms is also progressing in the therapeutic field: an international consortium (BRAVEinMS) has identified a candidate molecule, the BavisantThis drug, capable in preclinical models of stimulating remyelination and protecting neurons while reducing inflammation, could accelerate the arrival of phase 2 clinical trials in a disease where treatment options remain limited.
These works converge towards the same promise: to move from a medicine that mainly treats symptoms to a more refined strategy, based on the real biology of the disease, in order to offer more targeted and effective treatments according to the profile of each patient.
