In a significant medical breakthrough, scientists have harnessed the power of artificial intelligence to identify two entirely new biological subtypes of multiple sclerosis (MS). This landmark discovery promises to revolutionise how the complex neurological condition is monitored and treated, moving away from a one-size-fits-all approach.
How AI Unlocked New MS Classifications
The pioneering research was conducted by a team from Queen Square Analytics and University College London (UCL). Their study, published in the journal Brain on Wednesday 31 December 2025, analysed data from 634 MS patients. The scientists employed sophisticated AI algorithms to scrutinise routine brain scans alongside measurements of a key blood biomarker called serum neurofilament light chain (sNfL), which indicates nerve cell injury.
This innovative methodology revealed two distinct "biologically informed MS sub-types," which the current clinical labels of relapsing-remitting or progressive MS fail to capture. Dr Arman Eshaghi, the study's lead author from the UCL Queen Square Institute of Neurology, stated: "Using routine brain images and a blood marker of nerve-cell injury, we identified two distinct biological trajectories in multiple sclerosis."
Defining the Two New Subtypes
The first newly identified subtype has been termed 'early-sNfL'. Patients in this group exhibit high levels of the sNfL biomarker early in their disease progression. Crucially, this is coupled with visible damage to a specific brain region known as the corpus callosum, which is vital for coordinating thought, memory, and movement.
The second subtype is 'late-sNfL'. Here, patients show a later rise in the sNfL marker. This pattern is associated with early volume loss in both the cortical and deep grey matter of the brain, as detailed by the researchers.
A Step Towards Personalised MS Care
This discovery has profound implications for the future of MS management. It provides a biological explanation for why the disease follows such varied paths in different individuals, something that has long puzzled clinicians and patients alike.
Caitlin Astbury, Senior Research Communications Manager at the MS Society, highlighted the importance of the findings to The Guardian. "MS is complex and these [current] categories often don't accurately reflect what is going on in the body, which can make it difficult to treat effectively," she explained. "The more we learn about the condition, the more likely we will be able to find treatments that can stop disease progression."
Dr Eshaghi emphasised that this work is a major stride towards more personalised medicine. By understanding an individual's specific biological subtype, healthcare professionals could tailor monitoring and treatment strategies far more precisely than is possible today.
MS is a lifelong condition affecting the brain and spinal cord, where damage to the protective sheath around nerves leads to symptoms like severe fatigue, chronic pain, muscle spasms, and mobility issues. This UK-led research, blending cutting-edge AI with clinical neuroscience, offers renewed hope for the over 130,000 people living with MS across the country.
