FDA要求GLP-1药物增加NAION风险提示后的用药困境
GLP-1 Eye Warning Leaves Patients, Doctors With a Difficult Choice
FDA于11月9日宣布要求所有GLP-1类药物增加非动脉炎性前部缺血性视神经病(NAION)风险提示。作者在文中指出,现有临床试验数据与观察性研究提示GLP-1使用者发生NAION的风险略高于对照组,但无法据此判断确诊后继续用药或停药对另一只眼的影响,呼吁FDA开展针对性研究以填补这一空白。
原文基于现有数据提出了一个尚未被回答的临床问题,有助于理解患者面临的具体两难处境。
— FDA should study outcomes when patients stop, continue, or change treatment after NAION
by Henry Bair, MD, MBA
November 18, 2026 • 4 min read
When someone loses vision in one eye while taking a GLP-1 drug, the first priority is an urgent eye examination. If the diagnosis is non-arteritic anterior ischemic optic neuropathy (NAION), the question that follows is: should we stop the drug?
On November 9, FDA announced it is requiring new warnings across all products in the GLP-1 drug class, which includes semaglutide (Ozempic, Wegovy) and tirzepatide (Zepbound, Mounjaro), among others, about a possible increased risk of NAION.
NAION occurs when inadequate blood flow damages the nerve that carries visual information from the eye to the brain. It can cause sudden, painless vision loss, often noticed on waking. The loss can be permanent, and there is no proven treatment to reverse it. The condition is rare, but these medications are widely used: in a March 2026 KFF poll, about one in eight U.S. adults reported currently taking a GLP-1 drug and one in five reported ever taking one.
As an ophthalmologist, I have routinely asked patients with suspected NAION about GLP-1 use for the past 2 years. I welcome the warning. The FDA should now pair it with a plan to study what happens when patients stop, continue, or change treatment after a NAION diagnosis.
People who develop NAION in one eye are already at increased risk of developing it in the other. How GLP-1 drugs might contribute remains uncertain. One hypothesis is that rapid reductions in blood sugar could cause swelling where the optic nerve enters the eye, compromising blood flow in susceptible patients. That possibility does not tell us whether continuing the medication adds to the risk in the other eye, or whether stopping reduces it.
These medications can be difficult to give up. Some patients take them to manage diabetes that has resisted other treatment. Certain drugs in the class also reduce the risk of heart attacks or worsening kidney disease in eligible patients. Stopping requires a plan for the condition being treated. When counseling patients who develop NAION while taking a GLP-1 drug, I recommend involving the prescribing doctor before changing treatment.
Regulators around the world have responded differently to this dilemma. European guidance says to stop semaglutide if NAION is confirmed. The FDA, whose warning covers the broader drug class, advises clinicians to consider discontinuation unless the benefits clearly outweigh the risk. I support that approach, with the patient, ophthalmologist, and prescribing doctor weighing the potential harm to vision against the benefits of treatment with a GLP-1.
Across 126 clinical trials, the FDA identified 15 probable NAION cases. The rates were equivalent to roughly six cases per 100,000 patients followed for a year among those receiving GLP-1 drugs, compared with four among those receiving a placebo or another diabetes medication. Larger observational studies have also raised concern, though their estimates vary and they cannot fully separate the effects of treatment from the conditions that led patients to receive it.
But estimating the risk of developing NAION does not tell us whether a patient should continue a GLP-1 drug after NAION diagnosis. One major FDA study excluded people with previous damage to the optic nerve from inadequate blood flow to help identify new cases accurately. That was a sensible design for its purpose, but it also means the study cannot establish whether continuing treatment after a diagnosis changes the risk to the other eye.
Some research has examined second-eye involvement. An analysis presented at the American Academy of Ophthalmology annual meeting this fall found that, among patients who developed NAION in both eyes, the second eye was affected sooner in GLP-1 users. It did not establish whether stopping would have prevented that outcome. Nor do we have reliable comparative evidence that switching to another GLP-1 drug is safer.
The FDA should assess whether its existing safety surveillance can answer these questions. If those systems fall short, the agency should require manufacturers to support additional studies. Federal law gives the agency that authority for serious safety concerns when routine reporting and available surveillance are insufficient.
The work could begin in participating eye clinics, linking confirmed diagnoses to prescribing and medical records. Researchers would document which medication a patient was taking, whether it was stopped or changed, when, and why. They would follow vision in the other eye alongside overall health, including how well replacement treatment managed diabetes or other conditions. Researchers would need to account for differences in diabetes severity and visual damage that affect both treatment decisions and outcomes.
The FDA should publish the study plans, identify who is responsible, and set dates for progress updates. Funding should be disclosed and results analyzed independently of manufacturers. Because NAION is uncommon, firm answers may take years. Patients and clinicians should still be able to see what is being learned and what remains uncertain.
The patient deciding what to do about the next dose cannot wait years for better evidence. Today's decision will have to be made with what we know. We owe that patient our best judgment, and we should follow what happens afterward. The next person facing the same choice deserves more to go on.
来源:MedPage Today · medpagetoday.com