FDA Orders GLP-1 NAION Warning: What the Label Change Means for the 278-Case Vision-Loss Lawsuit MDL

The FDA on November 9, 2026 ordered every GLP-1 drug, from Ozempic and Wegovy to Mounjaro and Zepbound, to carry a new warning that the class may increase the risk of non-arteritic anterior ischemic optic neuropathy (NAION), a rare optic-nerve injury that can cause sudden, permanent vision loss. For the GLP-1 NAION warning lawsuit docket, MDL No. 3163, with 278 pending federal cases before Judge Karen Spencer Marston in Philadelphia, the label change removes the single fact both sides had built around, that no United States label mentioned NAION, and reshapes the failure-to-warn, preemption, causation and limitations arguments that will decide the litigation.

The agency did not find that GLP-1 drugs cause NAION. It found a "potential increased risk," based on a pooled analysis of 126 clinical trials and a real-world Sentinel study whose result depends on which comparator drug is used. That hedged finding is precisely why the warning helps plaintiffs on some issues and helps Novo Nordisk and Eli Lilly on others. What follows sets out what the FDA did, the numbers it relied on, how the change interacts with Wyeth v. Levine and Merck v. Albrecht, what it means for claimants injured before and after November 9, and where it lands in the MDL schedule.

What did the FDA announce about GLP-1 drugs and NAION?

The FDA issued a Drug Safety Communication on Monday, November 9, 2026 requiring the application holders for all GLP-1 receptor agonists to add information about a potential increased risk of NAION to the Warnings and Precautions section of their labeling. The agency said NAION cases have been reported for every currently marketed GLP-1 product, which is why it acted on the whole class rather than on semaglutide alone.

Warnings and Precautions is the section that matters in litigation. It sits above Adverse Reactions in prescribing information and is where FDA places risks it considers clinically significant enough to change prescriber behavior. Until this month, the United States labels for semaglutide products did not mention NAION at all, which the site reported in its September and October coverage of the docket.

The communication also tells clinicians what to do. FDA advises that suspected NAION be referred promptly to an ophthalmologist, that the GLP-1 drug be considered for discontinuation if NAION is confirmed unless the benefits clearly outweigh the risks, and that cases be reported through MedWatch. Patient-facing guidance tells users to seek care right away for sudden vision loss in one eye, which often appears on waking, along with blind spots or blurred central vision.

Which GLP-1 drugs must carry the NAION warning?

Every GLP-1 receptor agonist marketed in the United States is covered. National coverage on November 9 and 10 named the following products; the requirement is class-wide, so the list is illustrative rather than exhaustive.

Product	Active ingredient	Company	Primary use
Ozempic	semaglutide	Novo Nordisk	Type 2 diabetes
Wegovy	semaglutide	Novo Nordisk	Chronic weight management
Rybelsus	oral semaglutide	Novo Nordisk	Type 2 diabetes
Victoza / Saxenda	liraglutide	Novo Nordisk (plus generics)	Diabetes / weight management
Mounjaro	tirzepatide	Eli Lilly	Type 2 diabetes
Zepbound	tirzepatide	Eli Lilly	Chronic weight management
Trulicity	dulaglutide	Eli Lilly	Type 2 diabetes
Soliqua	insulin glargine / lixisenatide	Sanofi	Type 2 diabetes
Exenatide (generic)	exenatide	Generic manufacturers	Type 2 diabetes

Reporting on the communication states that generic exenatide is included because the reference brand, Byetta, was withdrawn for reasons other than safety or effectiveness. That detail matters for the generic-drug analysis below.

Eli Lilly said the communication "does not reference any findings specific to Lilly medicines" and that it will respond to FDA promptly. Novo Nordisk said the data do not establish a cause-and-effect relationship and that it remains confident in the benefit-risk profile of its products.

How big is the NAION risk the FDA found with GLP-1 drugs?

Small in absolute terms and statistically fragile, which is the heart of the coming causation fight. Because individual trials recorded few or no NAION cases, FDA pooled 126 randomized controlled trials from GLP-1 development programs.

Data source	Population	Finding
Pooled RCT meta-analysis	126 trials; 134,253 participants (78,217 GLP-1, 56,036 comparator)	15 probable NAION events: 10 GLP-1, 5 placebo, 0 non-GLP-1 active comparators
Pooled RCT incidence	Same	6.4 events per 100,000 patient-years (GLP-1) vs 3.8 (comparator)
Sentinel cohort vs SGLT-2 inhibitors	Adults with type 2 diabetes starting therapy 2019-2024	Hazard ratio 1.19 (95% CI 0.91-1.54), not statistically significant
Sentinel cohort vs DPP-4 inhibitors	Same	Hazard ratio 1.43 (95% CI 1.02-2.01), statistically significant
Animal studies	Nonclinical data	No mechanism identified

Read together, the data show roughly two extra NAION cases per 100,000 patient-years of GLP-1 use, a confidence interval against one comparator that crosses 1.0, and a significant result against another whose lower bound sits barely above it. FDA also noted that NAION risk factors often overlap with conditions common among patients treated with GLP-1 drugs, which made the reports hard to interpret and complicates attribution. Its inquiry followed a July 2024 study in JAMA Ophthalmology that associated semaglutide with NAION.

For comparison, the European Medicines Agency's safety committee concluded on June 6, 2025 that NAION is a "very rare" side effect of semaglutide, affecting up to 1 in 10,000 patients, so United States labels now catch up to European product information by about 17 months.

How many GLP-1 vision loss lawsuits are pending in MDL 3163?

The JPML's report of November 2, 2026 lists 278 pending actions in MDL No. 3163, In re: GLP-1 RAs NAION Products Liability Litigation, every one of the 278 ever filed. The docket has grown quickly: 216 actions on September 1 and 247 on October 1. Some national coverage of the FDA action said "at least 4,000" vision-loss claims are pending; that figure belongs to the separate gastrointestinal-injury docket, MDL No. 3094, which had 4,082 pending actions on the same report. Both MDLs sit before Judge Marston in the Eastern District of Pennsylvania.

The Panel created MDL 3163 by transfer order on December 15, 2025. Defendants are Novo Nordisk and Eli Lilly entities; the plaintiffs allege that the manufacturers knew or should have known of the NAION signal and failed to warn prescribers and patients. A parallel state-court docket in New Jersey runs on its own calendar.

Does the FDA NAION warning help Ozempic vision loss lawsuits?

On some elements, yes; on others, it gives the defense new material. The table summarizes the cross-currents before the sections below explain each.

Issue	How the November 9 label change cuts	Likely beneficiary
Existence of a risk worth warning about	FDA now says a warning belongs in Warnings and Precautions	Plaintiffs
Preemption for pre-warning injuries	FDA acted on its own timeline after reviewing the same data	Contested; both sides
General causation	FDA declined to find causation; data are comparator-dependent	Defendants
Post-warning injuries	Prescribers are now warned; learned intermediary doctrine applies	Defendants
Statute of limitations	Public, nationwide notice of a possible link on a fixed date	Defendants (for late filers)
Generic GLP-1 claims	Generic makers cannot change labels on their own	Defendants

The strongest plaintiff point is simple: the regulator now agrees the risk is significant enough to warn about, and it reached that view using clinical-trial data that were in the manufacturers' hands. A failure to warn claim requires proof that the label was inadequate; a defendant arguing that no reasonable warning was required faces a federal agency that has just required one.

The strongest defense point is equally simple. FDA said "potential" risk, found no mechanism in animal data, and reported a Sentinel result that was not significant against SGLT-2 inhibitors. A warning is not an admission of causation, and FDA itself framed it that way.

Can drugmakers argue preemption after the FDA added the NAION warning?

Yes, and the label change is now the main exhibit for both sides of that argument. Preemption (Products) for brand-name drugs is governed by two Supreme Court cases. In Wyeth v. Levine, 555 U.S. 555 (2009), the Court held that state failure-to-warn claims against brand manufacturers are not preempted, because the changes-being-effected (CBE) regulation lets a manufacturer strengthen a warning on its own when it has "newly acquired information," unless there is "clear evidence" that FDA would have rejected the change. In Merck Sharp & Dohme Corp. v. Albrecht, 587 U.S. 299 (2019), the Court held that preemption is a question of law for the judge, and that the manufacturer must show it fully informed FDA of the justification for the warning and that FDA, by action carrying the force of law, told it the change was not allowed.

Plaintiffs will argue the November 9 communication defeats any clear-evidence defense: far from rejecting a NAION warning, FDA has ordered one, and the data it used, chiefly the pooled trial results, were available to the sponsors long before. They will also argue that the CBE pathway was open to Novo Nordisk at least from mid-2024, when the JAMA Ophthalmology study appeared, and certainly from June 2025, when the EMA safety committee concluded that European product information for semaglutide should list NAION.

Defendants will argue the opposite inference. FDA had the same signal for more than two years, evaluated it, and did not require a warning until November 2026; on their account that history shows the agency did not believe the evidence met the labeling standard earlier, and a unilateral warning would have been "false or misleading" under the misbranding provisions. Expect them to point to any FDA communications in 2024 and 2025 that described NAION as under evaluation. Under Albrecht, the question will turn on the precise content of the sponsor-agency exchanges, which is why preemption discovery was frontloaded.

Neither argument is new, but both are now sharper. Judge Marston will decide preemption as a matter of law on the record the parties build, and the November 9 communication, plus whatever FDA review documents underlie it, will be at the center of that record.

What about generic GLP-1 drugs?

Claims against generic manufacturers face a separate hurdle. Under PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011), generic makers must keep labels identical to the reference listed drug, so state-law claims that they should have warned differently are generally preempted. Generic liraglutide and generic exenatide are on the market; claimants who took only generics will typically find their failure-to-warn theory foreclosed against the generic maker, whatever the outcome against the brands.

Can I still sue if I lost vision before the FDA warning?

Yes. A label change does not extinguish claims for injuries that occurred before it, and those pre-warning claims are the core of MDL 3163. For a claimant who developed NAION before November 9, 2026, the label in force at the time contained no NAION warning, so the inadequacy element of a failure-to-warn claim looks stronger, not weaker, after FDA's action. The live questions for those claimants remain general causation, specific causation (whether GLP-1 use, rather than diabetes, hypertension or sleep apnea, caused their particular injury) and preemption.

Claimants injured after the new language reaches labels are differently placed. Under the learned intermediary doctrine, the manufacturer's duty runs to the prescriber; once the prescriber receives a warning that specifically names NAION and tells clinicians to consider stopping the drug if it occurs, a post-warning plaintiff must attack the adequacy of the new wording or the prescriber's conduct, both harder cases.

Can the label change itself be used as evidence at trial?

Probably, though defendants will object. Federal Rule of Evidence 407 bars evidence of subsequent remedial measures offered to prove culpability, but many federal courts have held that the rule does not reach changes a regulator required rather than ones a defendant chose. State evidence rules vary, and remanded cases will be tried under them.

Does the new warning start the statute of limitations for NAION claims?

It may, for some claimants. Most states apply a discovery rule, under which the statute of limitations begins when a plaintiff knew or reasonably should have known of the injury and its probable cause. A nationwide FDA warning on a fixed date, amplified by national news coverage, is the kind of public event defendants routinely argue put claimants on inquiry notice. Personal injury limitations periods are commonly two or three years, and some states run shorter.

For anyone diagnosed with NAION while on a GLP-1 drug, November 9, 2026 is now a date defense counsel will cite. Claimants' counsel can expect to argue that earlier publicity, such as the July 2024 study and the June 2025 European label change, did not give notice to a United States patient, while defendants may argue that those events started some clocks even earlier. The answer is state-specific, and the risk falls most on late filers, which is one reason intake typically accelerates after a federal label change.

Where does the GLP-1 NAION warning lawsuit docket stand on the MDL 3163 schedule?

Judge Marston's Case Management Order No. 12, entered July 2, 2026, grants early discovery and motion practice on two cross-cutting issues: general causation (whether GLP-1 drugs can cause NAION at all) and preemption and warning adequacy. Case Management Order No. 14, entered August 27, 2026, sets the scheduling for that work, and Rule 702 and summary-judgment briefing on the cross-cutting issues is not expected before mid-2027. No bellwether schedule has been set.

Date	Event
July 2024	JAMA Ophthalmology study associates semaglutide with NAION
June 6, 2025	EMA safety committee calls NAION a "very rare" semaglutide side effect
December 15, 2025	JPML creates MDL No. 3163, assigns Judge Marston (E.D. Pa.)
July 2, 2026	CMO No. 12 frontloads general causation and preemption
August 27, 2026	CMO No. 14 scheduling order
November 2, 2026	JPML report: 278 pending actions
November 9, 2026	FDA requires class-wide NAION warning

Because the communication arrived while preemption discovery is open, both sides can now pursue the agency record behind it: the review memos, the dates FDA first flagged the signal to sponsors, and any sponsor submissions. That record will feed both the Albrecht analysis and the general causation experts. Expect early motion practice over the scope of discovery into FDA communications and over whether the gastrointestinal docket's Daubert Challenge ruling in MDL 3094, still pending after the September 2026 hearing, signals how Judge Marston will treat comparator-dependent epidemiology.

The GLP-1 NAION warning: what it means for claimants' counsel, defendants and funders

Claimants' counsel. The warning validates the existence of a signal and weakens any clear-evidence defense for pre-warning injuries, but it does not supply causation. The FDA numbers, two extra events per 100,000 patient-years and a non-significant Sentinel comparison, are a reason to invest in ophthalmology and epidemiology experts now rather than rely on the label at Rule 702. Case selection matters: claims with documented GLP-1 use, a clean differential for other NAION risk factors, and injury before November 9 are the ones the label helps. Limitations review for every intake with an older diagnosis is the urgent administrative task.

Defendants. The communication's own caveats, especially the absence of a mechanism and the SGLT-2 result, are the defense's best general-causation evidence, and they now come from FDA rather than from company-retained experts. The preemption argument becomes historical: why the agency waited until 2026 and what sponsors told it along the way. Post-warning claims and generic-only claims are likely to be the first categories challenged.

Funders and PI firms. Docket growth after a federal label change is typical, and MDL 3163 was already adding roughly 30 actions a month. The economic value of that inventory depends on two rulings that will not come before mid-2027 at the earliest: general causation and preemption. Advertising-driven intake will draw more post-warning and late-diagnosis claimants, the categories the warning hurts most, so per-claim underwriting should distinguish by injury date.

Prescribers and patients. The practical change is clinical: sudden vision loss on a GLP-1 drug should prompt same-day ophthalmology referral, documentation, and a discussion about continuing therapy. Those records, made close in time, are also the evidence that later determines whether any claim is viable.

FAQ

What is NAION?

Non-arteritic anterior ischemic optic neuropathy is damage to the optic nerve caused by reduced blood flow, typically producing sudden, painless vision loss in one eye. FDA says the underlying cause is not well understood, and vision loss can be permanent.

When did the FDA add the NAION warning to GLP-1 drugs?

The FDA announced the requirement in a Drug Safety Communication on November 9, 2026. Manufacturers must update the Warnings and Precautions section of their labeling; revised labels will roll out as application holders submit and FDA approves the changes.

Does the FDA say Ozempic causes blindness?

No. FDA described a "potential increased risk" and said available data do not establish causation; its animal data review found no mechanism. Novo Nordisk said the data do not establish a cause-and-effect relationship.

Is there a GLP-1 vision loss settlement?

No settlement has been announced in MDL No. 3163. The court is litigating general causation and preemption first, and no bellwether trial date has been set.

How many GLP-1 vision loss lawsuits are pending in MDL 3163?

278 federal actions were pending on the JPML's November 2, 2026 report, up from 247 a month earlier, all before Judge Karen Spencer Marston in the Eastern District of Pennsylvania.

Does the warning help someone who used a generic GLP-1?

Usually not against the generic maker. Under PLIVA v. Mensing, failure-to-warn claims against generic manufacturers are generally preempted because they must copy the brand label.