Semaglutide Linked to Age-Related Macular Degeneration Risk

 

• Older diabetic patients on two GLP-1 drugs showed an increased risk of neovascular age-related macular degeneration in an observational cohort study.
• The absolute risk was small, however, reaching 0.2% at 3 years among GLP-1 medication users and 0.1% among non-users.
• Semaglutide represented the bulk of GLP-1 receptor agonists used in the study, hence the need for further investigation in this drug class.

Older adults taking GLP-1 receptor agonists, primarily semaglutide (Rybelsus, Ozempic, Wegovy), had a small uptick in their risk of developing neovascular age-related macular degeneration (nAMD), according to a retrospective, population-based cohort study from Canada.

Patients taking GLP-1 agonists for type 2 diabetes for at least 6 months had an excess risk of nAMD compared with matched non-users over 3 years of follow-up (adjusted HR 2.21, 95% CI 1.65-2.96), reported Reut Shor, MD, of the University of Toronto, and colleagues in JAMA Ophthalmologyopens in a new tab or window.

"The longer the exposure, the higher the risk seemed to be," Shor told MedPage Today. "This was definitely surprising, especially given the growing enthusiasm for GLP-1 receptor agonists for their cardiovascular and metabolic benefitsopens in a new tab or window."

While the absolute risks of nAMD were small in the study -- 0.2% versus 0.1% in non-users -- "it's important to remember that nAMD is a condition with profound implications for vision and quality of life," said Shor.

A growing literature nevertheless points to potential ocular safety concernsopens in a new tab or window with the GLP-1 receptor agonist drug class. Why there is a higher risk remains unknown, however.

While a rapid systemic reduction in glucose, including in the retina, may induce a temporary hypoxic state after use of GLP-1 agonists, the study authors reasoned that a prolonged increase in risk is more likely related to a mechanism of increased chemokine CXCL12 levels triggering vascular endothelial growth factor (VEGF) production.

"The adage that 'there is no such thing as a free lunch' remains true. While certainly not outweighing the good these medications offer, prescribing physicians need to keep in mind the real and serious ocular adverse events that may occur," according to Brian VanderBeek, MD, MPH, of the University of Pennsylvania Perelman School of Medicine in Philadelphia.

In an adjoining editorialopens in a new tab or window, he cited the early clinical trials that tied GLP-1 receptor agonists to diabetic retinal disease.

"This seemed counterintuitive, since nearly all other diabetic endpoints improved with these medications. Researchers then theorized that these results were not specific to GLP-1 RA [receptor agonist] use, but rather an effect of a disproportionate number of patients experiencing [diabetic retinal disease] paradoxical worsening after rapid blood glucose control -- an effect that was not expected to occur in the placebo group," VanderBeek wrote.

"Additional ocular concerns have been raised with recent reports of an increased occurrence of nonarteritic ischemic optic neuropathyopens in a new tab or window (NAION) associated with GLP-1 RA use,opens in a new tab or window" he added. "Again, the pathophysiologic mechanism is not clear, but it may be related to lower optic nerve head perfusion due to autonomic vascular dynamics and overall lower systemic blood pressure induced by GLP-1 RA use."

The editorialist stressed the need to assess whether other GLP-1 drugs can show a more favorable benefit-risk profileopens in a new tab or window, while the researchers cautioned the findings should not be generalized to specific brands -- in the study, semaglutide users accounted for 97.5% of the GLP-1 agonist use, the small remainder having used the discontinued lixisenatide (Adlyxin).

For their study, Shor and colleagues relied on data from Ontario's health system for 2020 to 2023. Individuals with diabetes ages 66 and older exposed to GLP-1 receptor agonists for 6 months or more were matched 1:2 to unexposed peers, resulting in a study cohort of 139,002 patients.

Altogether, the mean age was 66 years across the two groups, and 47% were women. The average duration of diabetes was just over 6 years, and 64% had hypertension.

According to the results, risk for nAMD was higher in those who had used the GLP-1 drugs for more than 30 months (adjusted HR 3.62, 95% CI 2.56-5.13) or 18 to 30 months (adjusted HR 2.26, 95% CI 1.49-3.45) versus those who had used them for 6 to 18 months (adjusted HR 0.94, 95% CI 0.53-1.65).

"Although not impossible, it seems unlikely that a person would progress from no AMD [age-related macular degeneration] to nAMD within 3 years. This would suggest there was an already at-risk group of patients with non-neovascular AMD within the cohort," commented VanderBeek.

"Future work should focus on controlling for the baseline level of AMD to determine if it is, in fact, this subgroup driving the results seen," he urged. "Next, work needs to be done to determine if this is an adverse effect only in patients with diabetes or if those using GLP-1 RAs for weight management or other indications are similarly at risk."

Shor's group acknowledged that the present study could not stratify risk by the different GLP-1 drugs and it did not account for factors such as dose, route of administration, and frequency of administration.

Disclosures

Shor reported research funding from the Silver Target Foundation. Co-authors reported relationships with the PSI Foundation, Fighting Blindness Canada, Alcon, Apellis, AbbVie, Bayer, Bausch, Roche, and the Silver Target Foundation.

VanderBeek reported consulting with EyePoint Pharmaceuticals and receipt of institutional research grants from Research to Prevent Blindness and the Paul and Evanina Mackall Foundation.

Primary Source

JAMA Ophthalmology

Source Reference: opens in a new tab or windowShor R, et al "Glucagon-like peptide-1 receptor agonists and risk of neovascular age-related macular degeneration" JAMA Ophthalmol 2025; DOI: 10.1001/jamaophthalmol.2025.1455.

Secondary Source

JAMA Ophthalmology

Source Reference: opens in a new tab or windowVanderBeek BL "Should we be concerned about glucagon-like peptide-1 receptor agonists?" JAMA Ophthalmol 2025; DOI: 10.1001/jamaophthalmol.2025.1599.

https://www.medpagetoday.com/ophthalmology/generalophthalmology/115919