Assessing the Risk of CNS Demyelination Associated with TNF-α Inhibitors: Insights from a Large US Health Network

Highlight

This nationwide study evaluated the association between tumor necrosis factor-α (TNF-α) inhibitors and central nervous system (CNS) demyelinating events using a large electronic health record database from 70 U.S. healthcare organizations. Among 283,760 patients exposed to TNF-α inhibitors and an equal number of matched controls, a modest but statistically significant increase in risk of demyelinating disease or optic neuritis was observed, particularly within 1 to 5 years of exposure. The absolute incidence remained very low (<1%), and risk varied by specific TNF-α agent.

Study Background

Tumor necrosis factor-α inhibitors have transformed treatment approaches for autoimmune diseases such as rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, and inflammatory bowel disease. Despite their efficacy, concerns persist about the potential of these biologics to unmask or induce CNS demyelinating conditions, including multiple sclerosis (MS) and optic neuritis. CNS demyelination is a debilitating process characterized by the loss of myelin sheath around nerve fibers, leading to neurological disability. Although case reports and smaller studies have suggested such a link, definitive evidence on the magnitude and timing of this risk in real-world clinical practice remains limited. This study leverages a large, nationwide U.S. electronic health record network to quantify the risk and delineate differences among TNF-α inhibitors.

Study Design

This investigation was a retrospective propensity score-matched cohort study analyzing electronic health records from the TriNetX network, encompassing 70 U.S. healthcare organizations with data spanning 2010 to 2025. The study population included 283,760 patients with newly diagnosed autoimmune diseases initiating TNF-α inhibitor therapy, matched 1:1 with patients diagnosed with similar autoimmune conditions but not exposed to TNF-α inhibitors. Patients with pre-existing CNS demyelination were excluded. Key covariates such as demographics and comorbidities—including vitamin D deficiency, obesity, thyroid disease, mononucleosis, and smoking—were matched to minimize confounding. The primary outcomes were incidence and hazard ratios (HRs) of CNS demyelinating disease and optic neuritis evaluated at three intervals: 1 year post-treatment initiation, 1 to 3 years, and 3 to 5 years.

Key Findings

Overall, TNF-α inhibitor exposure was associated with a modest but statistically significant increase in combined risk of demyelinating disease or optic neuritis during the 1-3 year (HR 1.237; 95% CI, 1.081-1.415) and 3-5 year periods (HR 1.201; 95% CI, 1.003-1.438). However, the absolute incidence of these adverse events was very low across all timepoints, remaining under 1% even at 5 years.

Drug-specific analyses revealed heterogeneity among TNF-α inhibitors:

• Infliximab: Showed the most pronounced increased risk over 5 years with an incidence of 0.531% versus 0.244% in matched controls (HR 1.983), representing nearly double the risk of demyelinating events.
• Adalimumab: Exhibited a statistically significant association specifically with optic neuritis (HR 1.278; 95% CI, 1.013-1.613), but not with demyelinating disease as a whole or the combined outcome.
• Etanercept, certolizumab, and golimumab: Did not demonstrate statistically significant increases in risk for the combined demyelination/optic neuritis endpoint.

These findings indicate that certain TNF-α inhibitors may carry differential risks related to CNS demyelination, suggesting possible distinct biological or immunological effects within this drug class.

Expert Commentary

The modest elevation in risk is consistent with prior smaller studies and post-marketing surveillance reports, reinforcing vigilance in clinical practice, especially for patients with risk factors for demyelinating diseases. Given the very low absolute incidence, the benefits of TNF-α inhibitors in controlling autoimmune disease activity generally outweigh these risks. The observed variation among agents may reflect pharmacokinetic differences, molecular structure, or mechanisms of immune modulation, but the precise underlying biology remains to be elucidated.

Limitations inherent to retrospective analyses include residual confounding despite propensity matching and potential misclassification of outcomes based on electronic health record coding. Additionally, the study does not address whether TNF-α inhibitors directly cause demyelination or unmask subclinical disease. Prospective studies and mechanistic investigations are warranted.

Conclusion

This comprehensive large-scale study demonstrates that TNF-α inhibitor therapy is associated with a small but statistically significant increased risk of CNS demyelinating events and optic neuritis, with variation seen across different agents. Despite this signal, the absolute risk remains very low, supporting the continued use of these agents with informed clinical monitoring. Physicians should maintain heightened awareness particularly for patients presenting with new neurological symptoms. Future research should focus on mechanistic pathways and risk stratification to optimize individualized patient care.

Funding and Disclosures

The original study reported proprietary or commercial disclosures stated in its footnotes. No direct funding details are specified here.

References

1. Muayad J, Mendes J, Mendes F, et al. Tumor Necrosis Factor-α Inhibition and the Unmasking of CNS Demyelination across a Large United States Health Research Network. Ophthalmology. 2026;133(7):886-894. PMID:41720406.
2. Sabat R, Georgescu I, Schreiner B. Role of TNF-α in chronic autoimmune diseases and TNF-α inhibitors. Zeitschrift für Rheumatologie. 2017;76(6):463-471.
3. Perry A, Buckley C, Smirnov I. TNF-α antagonist-induced demyelination: An update. Neurology. 2020;94(7):e740-e744.
4. Gajofatto A, Benedetti MD. Treatment strategies for multiple sclerosis: When to start, when to change, when to stop? World Journal of Clinical Cases. 2015;3(7):545-555.