For many years, multiple sclerosis (MS) has been defined primarily by its symptoms rather than its underlying biological mechanisms. However, a new study challenges this approach, providing evidence that MS may develop through two distinct biological pathways. This discovery has the potential to change how doctors approach diagnosis, prognosis, and treatment strategies for the disease.

The Traditional View of MS

In clinical practice, multiple sclerosis is usually considered a single disease with a wide range of manifestations. Physicians focus on inflammatory processes, MRI changes, and neurological symptoms to guide therapy selection. Nevertheless, managing the disease remains a complex task for many patients: treatments effective for some may not work for others, and the progression of the disease often follows an unpredictable course.

New Research: Biological Patterns

Differences among patients have traditionally been seen as part of MS’s complexity. However, a study published in the journal Brain suggests that these differences may reflect fundamental biological distinctions. Instead of endless variability within a single diagnosis, researchers have proposed that the various manifestations of the disease are linked to different biological patterns.

To test this hypothesis, scientists from University College London and Queen Square Analytics decided to look beyond symptoms and clinical labels, focusing on biological signals of brain damage in MS. Their goal was not just to track disease activity, but to determine whether hidden patterns of degeneration could reveal distinct biological pathways of disease development.

Research Methods

The team combined two data sources:

• Blood analysis for neurofilament light chain (sNfL) — a protein released when nerve cells are damaged, widely used as a marker of disease activity.
• MRI scanning, which allows researchers to monitor how structural degeneration spreads through the brain over time.

Instead of analyzing each dataset separately, the researchers used a machine learning model called SuStaIn (Subtype and Stage Inference), developed at UCL. This model can identify subtle disease patterns and track their progression, enabling the team to test whether MS follows a single biological trajectory or something more complex.

Two Biological Subtypes of MS

Analysis of combined MRI and biomarker data from 634 patients revealed an unexpected pattern: rather than a single disease spectrum, two distinct structural patterns emerged. Patients clustered into separate groups, each reflecting a different pathway of neurodegeneration.

• One subtype was characterized by early damage concentrated in the brain’s cortex.
• The second subtype showed degeneration primarily in the brain’s white matter.

Although both patterns ultimately led to MS symptoms, the location and spread of damage differed significantly between the groups.

Impact on Disease Course and Treatment

The two subtypes differed not only in anatomy but also in the rate of disease progression. One group experienced slower, gradual structural decline, while the other showed rapid neurodegeneration. This confirms that MS can develop along more than one biological scenario.

This biological division helps explain why patients with similar diagnoses often have very different outcomes. If MS develops through multiple biological pathways, prognosis may depend not only on symptom severity but also on the specific disease trajectory in each patient. This could change how early signs of the disease are interpreted and how long-term risks are assessed.

Implications for Personalized Medicine

The same pattern is seen in treatment response: therapies that slow progression in one subtype may be less effective in the other. This explains why some patients respond well to certain medications while others do not.

Since changes in brain structure and blood biomarkers often appear before obvious clinical deterioration, researchers believe that this approach to subtyping could help doctors predict disease worsening earlier than is possible by symptoms alone.

For science, this concept opens new directions for studying disease mechanisms that were difficult to unravel within a single diagnosis. If future studies confirm these findings, it could bring MS treatment closer to a more personalized, biologically targeted model.

Next Steps

Currently, this work is still at the research stage and is not yet intended for clinical diagnosis or therapy selection. The next step will be to expand the study to larger and more diverse patient groups to confirm whether these biological patterns persist in real-world settings.