Specialists discuss concerns over impact of semaglutide on the eye
Click Here to Manage Email Alerts
A paper recently published in JAMA Ophthalmology by Hathaway and colleagues at Harvard suggested a potential association between prescribed semaglutide and an increased risk for nonarteritic anterior ischemic optic neuropathy.
Semaglutide is a GLP-1 receptor agonist primarily used as a treatment for diabetes (Ozempic, Novo Nordisk) and more recently greenlit as a breakthrough treatment for obesity (Wegovy, Novo Nordisk). In both conditions, it has shown remarkable effectiveness, with the added benefit of lowering the risk for heart attack and stroke. Understandably, the study findings spread concern and fueled debate in the medical community and beyond.
Source: Joseph F. Rizzo III, MD
“We were motivated to do this study purely because of our experience,” Joseph F. Rizzo III, MD, principal investigator of the study, said. “I had seen a patient who had NAION, and she was on Ozempic. I was meeting with my trainees the same day, and I was telling them about this when a resident walked in and said that she had just seen a case like that in the emergency room. Later that week, I saw a third case. Three cases within 7 or 8 days were very odd, I thought. That’s when I began to develop the strategy to do a clinical study.”
Rizzo has been a pioneer in the research and treatment of blinding diseases, with a special focus on optic neuritis and nonarteritic anterior ischemic optic neuropathy (NAION). He is director of the neuro-ophthalmology service at Mass Eye and Ear, one of the country’s largest and most highly specialized services in this subspecialty.
His study was based on “the traditional clinical experience of just seeing patients, thinking about what was happening and trying to understand it,” and he sees this as an added value. However, he took care to specify that it was observational, that more research is needed and that the findings should not lead to any change in practice patterns.
“The only change should be informing the patients of a potential risk,” he said.
Best-equipped specialists to diagnose NAION
The study retrospectively evaluated a cohort of 16,827 Mass Eye and Ear patients, 710 with type 2 diabetes and 979 with overweight or obesity. In the diabetes group, 194 patients were treated with semaglutide and 516 with non-GLP-1 receptor agonist (RA) medications. In the overweight/obesity group, 361 were prescribed semaglutide and 618 non-GLP-1 RA medications. In the diabetes group, NAION occurred in 17 semaglutide users and six non-semaglutide users. In the overweight/obesity group, NAION occurred in 20 semaglutide users and three non-semaglutide users.
These findings suggested an association between semaglutide and NAION, with a hazard ratio of 4.28 and 7.64, respectively, in the two groups.
“NAION is not a common disease and is often misdiagnosed. The primary advantage of our study was that it was conducted solely in my office where we only have neuro-ophthalmologists practicing, and these are the doctors who are most well equipped to diagnose NAION. We did manual chart review to ensure the accuracy of our diagnosis,” Rizzo said.
There is no specific diagnostic code for NAION, and many cases might therefore be inaccurately diagnosed, not detected or not reported, he said.
“And the consequence of this — and we’re wrestling with this right now — is that a larger study is needed to investigate this association, but without a diagnostic code, it is very difficult to appreciate how we’re going to find those cases. That’s the challenge,” he said.
Question marks
Susan P. Mollan, MBChB, FRCOphth, professor and neuro-ophthalmology consultant at the University of Birmingham, United Kingdom, wrote an invited commentary on the study in the same issue of JAMA Ophthalmology. She wrote that being vigilant for potential complications in the ever-changing landscape of systemic therapies is a duty toward patients and that the authors “should be commended for their clinical observations and acting upon them.” However, she recommended that these results should be interpreted with caution.
“The team started by finding a few patients where they considered there may be an association and worked backward. Working backward could create a bias within their research,” she said.
One limitation of the Harvard study was that patients were on semaglutide for a varied amount of time and developed NAION at various intervals from the start of the therapy, while a shorter interval of about 6 weeks should be expected for a drug-related complication.
“If we think a drug is causative, generally speaking, there would be a similar time point in which everybody would have the complication of the drug. And if we stop the drug and restart it, does the same complication occur? This a real challenge for some eye-related drug effects, as mostly it is not feasible to retest to confirm an association, particularly where significant sight loss might occur,” Mollan said.
She also argued that diabetes is a risk factor for NAION. Therefore, NAION development in these patients may be due to their underlying risk.
“Semaglutide was rigorously studied in several randomized controlled trials worldwide, and now many millions of people take this medicine throughout the world. This is the first study to ever report an association with NAION,” she said.
Is the risk clinically significant?
In another study published this year, Wai and colleagues found a higher rate of progression of proliferative diabetic retinopathy (PDR) and risk of new-onset diabetic macular edema in patients with diabetes treated with GLP-1 RAs as compared with those treated with SGLT2 inhibitors. The data analyzed in the study were collected from the TriNetX platform, which provided access to electronic medical records of more than 120 million patients.
“Various groups have reported on patients who initiate therapy on the GLP-1 agonists with existing diabetic retinopathy and experience progression to vision-threatening complications,” according to Ehsan Rahimy, MD, principal investigator of the study and clinical assistant professor at Stanford University. “The paper by Hathaway and colleagues has now opened a second area of interest, more specifically related to semaglutide and a potentially increased risk of NAION.”
GLP-1 agonists have become popular in recent times, and based on the sheer volume of patients who are on them, an increasing number of these reactions are to be expected.
At any rate, it is important to consider or contextualize that these events are rare, especially NAION, he said.
“In both our sites, we did see increased risks of developing these events, and both our studies do show statistical significance in terms of increased risk. But given that these are rare events overall, is the increased risk really clinically significant?” he said.
He believes that the small risk of having an ocular event cannot be a reason to discourage patients from not taking what can potentially be life-changing medications that have shown remarkable results. The important thing is to better characterize and screen for patients who are at higher risk and monitor them appropriately.
Real-world data
In a letter to the editor published in the Journal of Diabetes Science and Technology, David C. Klonoff, MD, medical director of the Diabetes Research Institute at Mills-Peninsula Medical Center in San Mateo, California, noted that the findings of the Hathaway study, based on a single tertiary care university eye clinic, “might not pertain to the general population.”
To assess the risk of NAION in patients prescribed semaglutide for diabetes or obesity, he performed a series of real-world analyses based on the electronic health record and claim dataset of Atropos Health, comprising 66 million people nationally.
“We worked with their statisticians, and after controlling for confounders and doing subgroup and sensitivity analyses, we found no significant increased risk of NAION in semaglutide users in the general population who were not necessarily selected from a database of patients being seen at a tertiary referral ophthalmology clinic,” Klonoff said.
The frequency of NAION in the analyses ranged from 0.07% to 0.24%.
“Some of the studies showed a slight increase, some showed a slight decrease, but we were looking for statistical significance because if it’s not statistically significant, then I don’t think you should act upon it,” he said.
He emphasized that GLP-1 RAs have been a breakthrough discovery that has changed the lives of many patients.
“They improve blood sugar. They help people lose weight. They decrease the risk of heart disease. They decrease the risk of kidney disease. There’s some evidence that they decrease cravings in people who have addictions, and we are finding other new indications for them. It’s an amazing class of drugs,” he said.
His only concern in prescribing semaglutide would be the worsening of a preexisting ocular condition. GLP-1 manufacturers often require a formal ophthalmological examination in their clinical trials because there is a proven small risk that these drugs might cause worsening of diabetic retinopathy.
“If someone already has diabetic retinopathy or has suffered an episode of NAION, then I would not start the drug. But otherwise, in the general population, I see no relevant evidence for concerns, and these are the patients I normally see,” Klonoff said.
Mechanism unknown
Investigating the mechanism through which semaglutide could potentially cause NAION was beyond the scope of the Harvard study, and causative mechanisms for this association are difficult to define.
“GLP-1 agonists have anti-inflammatory properties, so it is unlikely that they might cause inflammation in the optic nerve,” Klonoff said.
“GLP-1 is neuroprotective, as shown by all the basic science studies, so one would expect that it should be protective of the optic nerve rather than not,” Mollan said.
Recent literature has shown efficacy of GLP-1 agonists in partially reversing the neuropathological changes related to Alzheimer’s disease and Parkinson’s disease, she said, in addition to reducing the cardiovascular risk in patients with diabetes.
“If we think that NAION is a vascular disease, we would not expect that the drug might actually cause this,” she said.
Rizzo agreed that the association of NAION with a drug that is anti-inflammatory, neuroprotective and vasculoprotective is problematic, noting that at present there is no substantive basis for explaining the link between semaglutide and NAION. He said that the pathophysiology of NAION itself is unclear, and no one mechanism has been definitively demonstrated.
“The general idea is that NAION occurs because cells begin to swell in the very tight space at the optic nerve head. The swelling would close some of the small blood vessels going into the optic nerve head and thereby potentially cause a stroke of the nerve,” Rizzo said, reflecting that the swelling of cells is likely to be somehow involved in the mechanism that links semaglutide to NAION.
In the discussion section of the Harvard paper, the authors wrote that “GLP-1 RA-induced enhanced sympathetic nervous system activity might influence optic nerve head perfusion and potentially increase the risk of NAION,” but, again, Rizzo pointed out that this study was only observational and that potential causative mechanisms need a different kind of investigation.
“We have a NIH laboratory-based science grant to study NAION, and we are beginning to look at the potential roles for the GLP-1 receptors in human tissue to answer the question of why NAION happens in some of these patients,” he said.
Characterizing patients at risk
The priority now is to characterize the phenotype of patients at risk, according to Rahimy.
“The leading theory in our field is that in those patients who have dramatic responses to the GLP-1 agonists — those who suddenly lose a lot of weight or have a sudden crash of A1c — the rapidity of change may precipitate the worsening of diabetic retinopathy. There is a precedent to that in patients who have gastric band surgery. When they quickly lose a lot of weight and their A1c goes down, we have seen a worsening of DR,” Rahimy said.
“We have seen this in diabetic retinopathy,” Mollan said. “If you tighten glycemic control very tightly very quickly, you can have a paradoxical effect where actually the eye disease gets worse for a period of time.”
In other words, rather than being a secondary effect of the drug, the worsening of DR and maybe NAION could be secondary to what the drug does in the body.
“In people who have a rapid response, we think that semaglutide potentially disrupts the regional homeostasis of the retina. We like to see slower, controlled reductions in A1c and weight loss. I think that, probably, this similarly applies to patients from an ischemic optic neuropathy standpoint,” Rahimy said.
They both agreed that patients with preexisting episodes of NAION should not be prescribed semaglutide. A history of NAION in one eye increases the risk of developing it in the fellow eye by 10% to 15%.
“If I am dealing with a patient who has had a history of NAION, that is probably the one situation I would tell the endocrinologist or internal medicine doctor to pause and have a real detailed discussion with this patient. Ultimately, that’s something for the patient and their prescribing doctor to decide, but I would do my part to inform them about the findings of the recent study,” Rahimy said.
A team approach is always beneficial, and ophthalmologists should discuss with primary care providers and endocrinologists the risks and benefits of prescribing semaglutide in patients with specific eye conditions.
“I would advise the endocrinologist or PCP to ask patients about their eye health. If a person has a known ‘disc at risk’ or optic nerve head drusen and a previous ischemic optic neuropathy, an ophthalmologist may help to talk about the risks, providing the risks vs. benefits of starting semaglutide. There are many studies reporting the health benefits of using GLP-1 RA for their systemic disease that likely outweigh not starting GLP-1 RA,” Mollan said.
Responding to patients’ concerns
Rizzo said he would personally raise concern with patients if they already have vision loss of any origin.
“I think it is only fair to the patients to inform them of the potential risk. They need to know that an attack of NAION related to semaglutide would further reduce their vision and perhaps create functional difficulties in their lives, so I would urge more caution if someone has glaucoma and has already lost vision, for instance. Of course, it’s all about how patients feel. People will make individual choices — they just need to be well informed,” he said.
Some of his patients stopped using semaglutide the day the study results were publicized, and some decided not to stop. But he has been clear with everybody.
“I tell my patients about our study. I also tell them that really understanding if there is a causal effect requires a different type of study and that there’s going to be a period of uncertainty,” Rizzo said.
GLP-1 RAs have been used in the United Kingdom since 2006, and many people have been on semaglutide for the last decade, Mollan said.
“This drug has changed the course of their lives in terms of reducing diabetes, blood pressure and heart failure, all of which cause major morbidity in type 2 diabetes and obesity,” she said.
If her patients wish to discuss the study, she uses the joint North American Neuro-Ophthalmology Society and American Academy of Ophthalmology statement that clarifies the methodology, summarizes the findings and explains the limitations of the study.
Rizzo agreed that the findings from a tertiary institution cannot be generalized to the larger population but pointed out that the patients with NAION were not there because they had some other eye condition.
“They were referred to us because they had visual loss,” he said.
Since then, he has continued to see cases and has been contacted by people with NAION from around the world.
“But, of course, that’s all anecdotal. The question is whether there’s truly an increased risk. That’s the ultimate question. We need to do larger studies to assess true risk beyond the association that we found and then ultimately try to understand cause and effect,” he said.
- References:
- Al Jarallah OJ, et al. J Pharm Bioallied Sci. 2023;doi:10.4103/jpbs.jpbs_215_23.
- Alharbi SH. Ther Adv Endocrinol Metab. 2024;doi:10.1177/20420188231222367.
- American Academy of Ophthalmology and North American Neuro-Ophthalmology society issue advice on weight loss drug and eye health. https://www.aao.org/newsroom/news-releases/detail/weight-loss-drug-and-eye-health. Published July 8, 2024. Accessed Oct. 11, 2024.
- Castellana E. Eur J Hosp Pharm. 2024;doi:10.1136/ejhpharm-2024-004333.
- Drucker DJ. Diabetes Care. 2024;doi:10.2337/dci24-0003.
- Erbil D, et al. Brain Inj. 2019;doi:10.1080/02699052.2019.1587000.
- Hathaway JT, et al. JAMA Ophthalmol. 2024;doi:10.1001/jamaophthalmol.2024.2296.
- Joo JH, et al. Ophthalmol Sci. 2024;doi:10.1016/j.xops.2024.100547.
- Kapoor I, et al. Clin Exp Ophthalmol. 2024;doi:10.1111/ceo.14445.
- Klonoff DC, et al. J Diabetes Sci Technol. 2024;doi:10.1177/19322968241268050.
- Lafferty RA, et al. Expert Opin Pharmacother. 2023;doi:10.1080/14656566.2023.2192865.
- Li X, et al. Ophthalmic Res. 2023;doi:10.1159/000528627.
- Lum F. Am J Ophthalmol. 2024;doi:10.1016/j.ajo.2024.08.037.
- Madan S, et al. Natl Med J India. 2023;doi:10.25259/NMJI_982_20.
- Martin-Gutierrez MP, et al. Eye (Lond). 2024;doi:10.1038/s41433-023-02716-4.
- Mollan SP. JAMA Ophthalmol. 2024;doi:10.1001/jamaophthalmol.2024.2514.
- Rizvi AA, et al. Diabetes Metab Syndr Obes. 2022;doi:10.2147/DMSO.S351982.
- Rizzo JF 3rd. J Neuroophthalmol. 2019;doi:10.1097/WNO.0000000000000870.
- Singh AK, et al. J Assoc Physicians India. 2024;doi:10.59556/japi.72.0621.
- Tauqeer Z, et al. Ophthalmic Epidemiol. 2024;doi:10.1080/09286586.2024.2399764.
- Uzakgider NK, et al. Beyoglu Eye J. 2024;doi:10.14744/bej.2023.09815.
- Wai KM, et al. Am J Ophthalmol. 2024;doi:10.1016/j.ajo.2024.04.010.
- For more information:
- David C. Klonoff, MD, of Diabetes Research Institute, Mills-Peninsula Medical Center, San Mateo, California, can be reached at dklonoff@diabetestechnology.org.
- Susan P. Mollan, MBChB, FRCOphth, of Birmingham Neuro-Ophthalmology, University Hospitals Birmingham NHS Foundation Trust, United Kingdom, can be reached at soozmollan@doctors.org.uk.
- Ehsan Rahimy, MD, of Byers Eye Institute, Stanford Medicine, can be reached at erahimy@gmail.com.
- Joseph F. Rizzo III, MD, of Neuro-Ophthalmology Service, Mass Eye and Ear, can be reached at joseph_rizzo@meei.harvard.edu.
Click here to read the At Issue to this Cover Story.
Collapse
Click Here to Manage Email Alerts