RealTalk MS – Episode 438: The Discovery of 2 MS Subtypes with Dr. Arman Eshaghi

Date: January 19, 2026
Host: Jon Strum
Guest: Dr. Arman Eshaghi, Principal Investigator, University College London

Episode Overview

In this episode, Jon Strum explores a major breakthrough in how multiple sclerosis (MS) may be classified and understood. Recent research has identified two new biological subtypes of MS, challenging the traditional symptom-based categories. Dr. Arman Eshaghi, the study’s principal investigator, joins Jon to explain how artificial intelligence (AI) and novel biomarkers are driving this advancement, and what it could mean for MS care, treatment, and even prevention.

Key Discussion Points & Insights

Setting the Stage: Why MS Varies So Widely

• MS diversity: The episode opens with Jon acknowledging the broad variability in MS progression and symptoms among patients, noting:
  • "One of the more confusing aspects of MS is that it can present differently from one person to the next." [00:01]
• Breakthrough context: A University College London research team may have uncovered a reason for this variability by identifying two unique MS subtypes.

Digging Into the Science: EBV and MS

• EBV as a necessary factor: Jon summarizes recent research establishing Epstein Barr Virus (EBV) as a necessary, but not sole, factor for developing MS.
  • "It's estimated that about 95% of the world's adult population has been infected by the Epstein Barr virus. And for most people, EBV appears to simply lie dormant in our bodies." [02:00]
  • Yet only a small percentage develop MS.
• Gene-viral interaction: A study suggests that in individuals carrying the HLA DR-15 haplotype, EBV-infected B cells present myelin proteins, triggering autoimmune attack by T-cells—a possible mechanistic link to MS.

New MS Subtypes: Dr. Eshaghi’s Groundbreaking Research

The Problem with Current Classification

• Old labels fall short: Dr. Eshaghi notes that current MS categories (relapsing-remitting, secondary progressive, primary progressive) are based on symptoms, not biology:
  • "It's been known that these sorts of categories have been outdated. They don't necessarily correspond with what we call biological changes or pathophysiological changes." [10:52]
• Toward biological classification: The goal is to move beyond symptoms and use biological markers to better understand and treat MS.

The Tools: AI and Novel Biomarkers

• Machine learning (SUSTAIN): Their team used an unsupervised machine learning tool (SUSTAIN) to analyze MRI scans and blood biomarkers (serum neurofilament light chain, or sNFL).
  • "We asked computers if we could cluster, group, or profile people with MS in similar groups that have similar trajectories on the data, or biomarkers." [12:13]
• What is sNFL?: sNFL is a "smoke detector for the brain"—a protein that leaks into the blood when nerve cells are damaged.

The Discovery: Two Biological Subtypes

• Subtype 1: Early sNFL
  • Characteristics: High sNFL levels early on, pronounced MRI activity (especially in the corpus callosum).
  • Implication: Indicates a "more active disease," likely more aggressive and with rapid lesion development.
  • Quote: "Patients who had early sNFL elevations...means patients who probably have a more active disease." [14:39]
• Subtype 2: Late sNFL
  • Characteristics: Low initial sNFL, but quiet shrinkage of deep gray matter (especially thalamus) detectable by MRI before sNFL rises.
  • Implication: More subtle, "silent" progression—less visible but still significant.
  • Quote: "The silent part of disease, it has been coined differently by different people. Some people call it smoldering disease." [21:08]

The Power of AI: Uncovering Hidden Patterns

• Beyond expert detection: AI can identify subtle changes and patterns across large datasets that human doctors might miss, especially in early disease stages:
  • "These are changes that...cannot be detected by even expert radiologists that require quantification by AI tools." [16:52]
• Objective clustering: The use of unsupervised models reduces human bias, focusing more on what the data reveals than preconceived categories.

Practical Implications

Prognosis & Personalized Medicine

• Predicting outcomes: Early sNFL subtype patients were found to be 144% more likely to develop new lesions.
• Treatment selection: Understanding a patient’s subtype could inform whether to pursue high-efficacy treatments from the start.
  • Quote: "If someone is labeled at early sNFL, they're likely to be more active. So...you will think of potentially prescribing higher efficacy, stronger treatments." [23:38]
• Reducing trial-and-error: Integrating these insights could help neurologists find the best therapy on day one, rather than cycling through multiple drugs.
  • Quote: "That's the hope...to feed into the conversations in the future that people living with MS and the doctors, nursing team, and also people who care for people with MS can have to improve that decision making." [25:37]

Future of MS Research and Clinical Trials

• Precision clinical trials: Trials could be designed with better patient matching, expediting discovery of targeted therapies.
  • "If we are able to better profile the underlying biology, you can think about precision clinical trials...that would target that specific biology." [26:18]
• Longer-term adoption: Widespread clinical integration of this approach may be a decade away, though earlier adoption in trials is likely.
  • Quote: "Clinical care is probably, you know, we're thinking about probably a decade of work before we can get [there]." [27:13]

Notable Quotes & Memorable Moments

• "We knew some patients are more active than others. But...the novelty and what is innovative in this study is really we bring this together...to get a more holistic understanding of disease."
  — Dr. Arman Eshaghi [18:51]
• "It's very mysterious. We're starting to better understand it in our study...the silent progression and the slow progression of disease usually requires 5 to 10 years of follow up."
  — Dr. Eshaghi describing the late sNFL subtype [21:08]
• "It may be a decade out. But I think you've just given us a peek into what the future of MS care is going to look like."
  — Jon Strum [27:29]

Key Segment Timestamps

• 00:01 – 05:15: Background on EBV, HLA, and recent key MS research.
• 10:28: Introduction of Dr. Arman Eshaghi.
• 10:31 – 12:13: Limitations of current MS categories; need for biological classification.
• 12:13 – 15:22: How AI and biomarkers identify new MS subtypes.
• 15:44 – 18:23: Explanation of the SUSTAIN machine learning tool.
• 18:23 – 21:08: Early sNFL vs. late sNFL subtypes and implications.
• 23:20 – 27:29: How subtypes could guide personalized therapy and influence future research and clinical care.

Conclusion

Dr. Eshaghi’s work represents a transformative step in MS research—moving from traditional, symptom-driven diagnoses to biological classifications powered by artificial intelligence and new biomarkers. This has the potential to profoundly change both treatment selection and the design of future therapies, inching closer to truly personalized MS care.