- Veterans with type 2 diabetes had a higher risk for nonarteritic anterior ischemic optic neuropathy (NAION) after starting semaglutide versus SGLT2 inhibitors.
- Absolute risk of the potentially blinding eye condition was very low over a 7.5-year maximum follow-up.
- The FDA is reviewing evidence about NAION and semaglutide but has not issued any warnings.
Risk for a rare, potentially blinding eye condition was low, but it was significantly higher in diabetes patients who started semaglutide (Ozempic) versus SGLT2 inhibitors, an observational study of U.S. veterans showed.
Among 102,361 type 2 diabetes patients, semaglutide initiators had more than a twofold higher risk for new-onset nonarteritic anterior ischemic optic neuropathy (NAION) compared with SGLT2 inhibitor initiators over a median 2.1-year follow-up (HR 2.33, 95% CI 1.54-3.54, P<0.001), reported Jennifer Lee, MD, PhD, of the VA Palo Alto Health Care System in California, and colleagues.
But absolute risk remained low, they wrote in JAMA Ophthalmology. Overlap-weighted cumulative risks -- a measure used to balance differences between patient groups -- were 0.29% for those on semaglutide and 0.13% for those on SGLT2 inhibitors.
"Although [semaglutide is] effective for glycemic control, weight loss, and cardiovascular risk reduction, NAION has emerged as a serious rare adverse event," Lee's group noted. Clinicians should incorporate the risk of NAION into their conversations with patients while weighing semaglutide's substantial cardiometabolic benefits, they advised.
"Consistent with recent commentaries, clinicians should encourage prompt evaluation of visual symptoms and consider ocular risk factors (e.g., optic nerve disease, prior NAION)," the team added. "Ophthalmologists should identify semaglutide use in patients with NAION, who should inform their prescribing clinicians."
While rare, NAION is one of the most common causes of acute optic nerve injury in individuals over age 50. It presents as sudden vision loss from lack of blood flow to the optic nerve, earning it the nickname of "optic nerve stroke."
"The biological mechanism linking GLP-1 receptor agonists to NAION remains unclear, including hypotension, volume depletion from gastrointestinal adverse effects, rapid glycemic improvement with transient microvascular dysregulation, and impaired vascular autoregulation at the optic nerve head," the researchers added.
The signal was first flagged in 2024 by researchers who found a 4.28-fold higher NAION risk in diabetic semaglutide users compared with those not on a GLP-1 drug. That study also reported a 7.64-fold higher NAION risk in users of semaglutide for overweight or obesity. Since then, additional research highlighting this link has included a review of over 3 million patients, a case series, and a review of FDA Adverse Event Reporting System (FAERS) data.
In 2025, the European Medicines Agency recommended updating semaglutide labels to include NAION as a "very rare" side effect (affecting up to one in 10,000 users). While the FDA has not yet issued a formal warning, its Sentinel System is actively investigating the link.
Other research has since tied GLP-1 receptor agonists to other eye conditions, including age-related macular degeneration and diabetic retinopathy. Conversely, some data have shown a benefit; one recent study reported a lower risk of legal blindness among diabetes patients using GLP-1 agents.
"To address limitations of heterogeneous data sources, design quality, low NAION event rate, and short follow-up, we emulated a target trial within the Veterans Health Administration, U.S.'s largest integrated health care system from 2018 through 2025," wrote Lee's group.
The researchers used pharmacy dispensing data about veterans with type 2 diabetes on metformin who started semaglutide or an SGLT2 inhibitor between March 2018 and March 2025. Nearly all SGLT2 inhibitor users received empagliflozin (Jardiance). Veterans were followed until first occurrence of NAION, study end, or a censoring event like death, giant cell arteritis, or traumatic optic nerve injury.
Semaglutide users tended to be younger (59.2 vs 64.6), had a higher body mass index (38.8 vs 33.6), and had lower HbA1c levels (7% vs 7.3%). They were also more likely to be female (17.2% vs 6.3%) and Black (21.6% vs 17.8%).
Over a 7.5-year maximum follow-up, 173 NAION events occurred: 30 among 11,478 semaglutide users (incidence rate of 123 per 100,000 person-years) and 143 events among 90,883 SGLT2 inhibitor users (67 per 100,000 person-years).
The analysis did not adjust for dose or duration of medication use, but those factors should be considered in future studies, Lee's group suggested.
Disclosures
The study received support from the VA Cooperative Studies Program.
Lee reported no disclosures. Co-authors reported relationships with TeCure Consulting, Novartis, Alnylam, AstraZeneca, Biodesix, Celgene, Janssen, Novartis, the University of Utah, and Western Institute for Veteran Research.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.