What is the current evidence for GLP1 receptor agonists as disease modifying agents in multiple sclerosis, beyond metabolic and weight effects?

The evidence for GLP-1 receptor agonists (GLP-1 RAs) as disease-modifying agents in multiple sclerosis (MS) is currently preclinical-dominant, with mechanistic data strongly supporting direct immunomodulatory and neuroprotective actions, while clinical data remain preliminary and confounded by metabolic effects.

🔬 Preclinical Evidence: Direct Mechanisms Beyond Metabolism

In the experimental autoimmune encephalomyelitis (EAE) model, GLP-1 RAs consistently reduce disease severity through mechanisms independent of weight loss:

· Peripheral immunomodulation: Liraglutide reduces EAE clinical scores without affecting body weight, specifically by suppressing pathogenic Th1 cell expansion in the spleen.
· Central neuroprotection: Semaglutide activates the PI3K/Akt/GSK-3β pathway, attenuating demyelination and promoting remyelination via CREB/BDNF, while also reducing oxidative stress and neuroinflammation (NF-κB, TNF-α).
· Microglial modulation: GLP-1 RAs suppress microglial activation, inhibit the NLRP3 inflammasome, and activate AMPK/SIRT1 signaling.
· Dual compartment action: NLY01 (a long-acting GLP-1 RA) delays EAE onset and reduces relapses by suppressing peripheral immune cell activation and preventing their trafficking into the CNS.

A systematic review of 15 studies (8 preclinical, 4 human) confirmed that GLP-1 RAs reduce clinical severity across EAE and cuprizone models through these diverse mechanisms.

🏥 Clinical Evidence: Signals Without Definitive Disease Modification

Human data show metabolic benefits but lack evidence of neurological disease modification:

· Dulaglutide trial (RRMS): Improved weight, BMI, and glucose tolerance, with exploratory trends in walking speed and manual dexterity, but no significant changes in plasma neurofilament light chain (NfL) or brain MRI volumetric measures.
· Observational cohorts: GLP-1 RA use was associated with BMI reduction and vitamin D augmentation without EDSS or relapse rate worsening, but no signal of disability improvement.
· Preliminary progression data: A study correlating GLP-1 agonist use with reduced risk of MS progression (via Kurtzke Functional Systems Scores) suggests potential, but is not definitive.

No randomized controlled trials have yet demonstrated disease-modifying efficacy; the first PROSPERO-registered systematic review notes that human evidence shows "metabolic benefit without neurological harm," but randomized trials are "urgently needed".

⚖️ The Metabolic Confounding Problem

A critical interpretive challenge is that most GLP-1 RA effects on MS-relevant outcomes could be secondary to metabolic improvement:

· Weight loss and improved insulin sensitivity may reduce systemic inflammation and comorbidity burden, indirectly benefiting MS.
· However, preclinical EAE studies show efficacy without weight change (e.g., liraglutide), arguing against a purely metabolic mechanism.
· The dulaglutide trial improved metabolic parameters but did not modify NfL or MRI volumetry, suggesting that metabolic benefits alone do not translate to neuroaxonal protection.

💎 Expert Perspective for Your Systematic Review

For your PROSPERO-registered review, the evidence supports a plausible direct disease-modifying mechanism (anti-inflammatory, remyelinating, neuroprotective) that is not fully explained by metabolic effects. However, clinical proof of disease modification is currently absent. Your meta-analysis should:

1. Stratify by mechanism: Separate studies reporting weight-neutral vs. weight-dependent effects.
2. Prioritize biomarkers: NfL and MRI volumetric outcomes are more disease-specific than EDSS or relapse rate in this context.
3. Acknowledge risk of bias: Most preclinical studies were high risk of bias; human studies moderate to high.

The current evidence justifies cautious optimism for a direct mechanism, but definitive disease-modifying efficacy awaits adequately powered RCTs with neuroaxonal biomarkers as primary endpoints.