KOHEN AVOCATS · PARIS, FRANCE
Ozempic Zepbound Eye Risk Warnings: What Can Patients Check About Federal Labeling and State-Law Claims?

Hassan Kohen is a Paris-based lawyer admitted to the Paris Bar (Avocat au Barreau de Paris). Initial strategic analysis: €80 incl. VAT, personal reply within 24 hours.

100% confidential · Legal professional privilege · No obligation

Article generated by artificial intelligence, through a process designed and supervised by the firm

Sources: U.S. court opinions and official U.S. government publications, read for this article

Maître Hassan KOHEN, avocat au Barreau de Paris
Hassan KOHEN
Attorney at the Paris Bar

Ozempic Zepbound Eye Risk Warnings: What Can Patients Check About Federal Labeling and State-Law Claims?

Speak to a lawyer 06 89 11 34 45

The U.S. Food and Drug Administration has warned of a potential increased risk of a rare eye condition, non-arteritic anterior ischemic optic neuropathy (NAION), associated with GLP-1 receptor agonists. In its drug safety communication, the FDA lists Ozempic and Zepbound among the medicines in the class and says it is requiring application holders to add information about that potential risk to the Warnings and Precautions section of their labeling. The agency describes a potential association; its communication does not determine why any particular person lost vision or whether anyone is legally responsible.

For a patient, the immediate questions are practical: what symptoms need medical attention, which label applied when the medicine was used, and what records could help explain an individual experience? For someone considering a possible claim, the answer also depends on the medicine, the medical evidence, the warning actually given, the applicable state law and the interaction between state claims and federal regulation. The discussion that follows is general in scope: it describes checks useful in comparable situations and makes no findings about the named products or their manufacturers. It uses federal rules and decisions from identified courts; it does not predict the result of an individual claim.

I. What should a patient check after an eye-risk warning?

A. Separate the FDA’s safety signal from an individual diagnosis

Put medical needs first. The FDA says NAION can cause sudden, painless reduced vision in one eye, including permanent vision loss. It lists sudden loss of vision, missing or blurred parts of the visual field, changes in color perception and reduced central sharpness among symptoms patients may notice. Its guidance tells patients to contact a health care professional right away if they experience such symptoms. A patient who has experienced NAION should discuss with a health care professional whether to continue or stop the medication; a headline alone is not a treatment plan.

The agency’s communication is broader than a notice about two brand names. It addresses GLP-1 receptor agonists as a class, notes several approved medical uses and identifies a potential increased risk of NAION. The FDA describes the condition as related to decreased blood flow to the front of the optic nerve while saying that its underlying cause is not completely understood. It also identifies factors that may be relevant, including diabetes, high blood pressure, sleep apnea, smoking, anemia and kidney disease. Those factors are reasons to take a complete history, not a basis for deciding the cause of a particular injury from a checklist.

The numbers in the communication have a specific meaning. The FDA reports an analysis of 126 randomized controlled trials, with a rate of 6.4 probable NAION events per 100,000 person-years among participants treated with a GLP-1 receptor agonist and 3.8 per 100,000 person-years in the comparator group. Person-years measure accumulated time in the study; the two rates are not an individual’s predicted chance of injury. The agency also describes a separate observational analysis in which one comparison was not statistically significant and another showed an increased risk. Reading both findings helps prevent a patient from treating a safety signal as either proof of personal causation or proof that no assessment is needed.

Keep the source and the date on which you obtained it. A screenshot of an abbreviated headline may leave out the agency’s description of the comparison group, the condition, or its qualifications. Save the complete communication and any later version you find, and note whether a document is an FDA safety communication, the product’s prescribing information, information given by a pharmacy, or advice recorded by a clinician. They serve different purposes. A communication can explain the agency’s action; the labeling and treatment records help establish what information was available at a particular point in care.

The medical record is more useful when it preserves the sequence rather than a conclusion. Record the medicine and strength, when treatment began or changed, when a symptom was first noticed, and who evaluated it. Preserve the name and address of each treating practice in your own files, the test and imaging reports, the diagnosis if one was made, and records of relevant prior eye or systemic conditions. If a date is uncertain, mark it as uncertain rather than assigning the date of a later appointment to the onset of symptoms. Distinguish what the patient remembers from what a contemporaneous record states.

A prescription bottle, pharmacy printout, package or electronic dispensing history may identify the particular product and dispensing date. A benefits statement can help locate a transaction, but it may not contain the entire warning delivered with the medicine. If there were several fills, keep the records for each one; a medicine’s label can change over time, and a later pharmacy handout is not automatically the document that accompanied an earlier prescription. Store originals safely and work from copies where possible.

For a reader without symptoms, the first check may be simpler: locate the current prescribing information and bring questions about an individual medical history to the clinician who manages the treatment. A recent warning does not, by itself, tell that reader to stop a prescribed medicine. The FDA’s guidance calls for medical evaluation of symptoms and an individualized discussion if NAION has occurred. Neither the regulatory notice nor the court decisions below substitute for that clinical decision.

B. Find the relevant federal label and distinguish it from a report

Federal labeling rules organize the information a prescriber is meant to use. 21 C.F.R. § 201.57 specifies the content and format of covered prescription-drug labeling, including the Warnings and Precautions section. Its warning provisions address clinically significant hazards and explain that a causal relationship need not have been definitely established before a warning is added where the regulation’s standard is met. That regulatory threshold for warning language is different from a court’s ultimate resolution of an individual injury claim.

Federal 21 C.F.R. § 201.56 sets general content and format requirements for the covered labeling, including that it be informative, accurate and neither promotional in tone nor false or misleading. When examining a label, look beyond a single sentence. Identify the product, the revision date if shown, the section in which a risk appears, the population described and whether the language addresses frequency, symptoms, monitoring or a particular clinical setting. Compare the complete relevant sections before concluding that two versions say the same thing.

Labeling may be updated after approval. Federal 21 C.F.R. § 314.70 describes supplements and other changes to an approved new drug application, including a changes-being-effected route for certain changes that add or strengthen specified safety information based on newly acquired information. That does not mean every new article, complaint or adverse-event report automatically required a label change. The available route, the kind of information, the product’s approval pathway and what the agency had already considered can matter.

Federal 21 C.F.R. § 314.3 defines newly acquired information for this regulatory framework. The definition can include information, data or analyses not previously submitted to the FDA, including a new analysis of previously submitted data when it reveals risks of a different type or greater severity or frequency than previously included in submissions to the agency. A useful comparison therefore identifies both the earlier information and what the later information actually added. The date on a paper is a starting point for that inquiry, not the complete answer.

Keep separate three questions: what was in the label; what information was available to the relevant application holder; and what a treating clinician communicated to the patient. The same document may help with more than one question, but it should not be assumed to answer all three. A patient handout, for example, may not reproduce the full prescribing information. A later regulatory notice may show why a label was to change without proving what an earlier prescriber read or why a particular treatment was selected.

The FDA also accepts reports of suspected problems. Its MedWatch voluntary reporting page says patients, consumers and health professionals can report observed or suspected adverse events. A report can provide information for safety monitoring. It is not a court finding, a diagnosis or a substitute for the medical records needed to understand an individual event. If submitting a report, describe what is known, identify uncertainty and retain a copy of the information supplied.

Federal 21 C.F.R. § 314.80 separately addresses postmarketing reporting of adverse drug experiences by covered applicants. These regulatory reporting obligations should not be confused with a patient’s voluntary report or with a finding that a medicine caused a particular event. A report may prompt further examination; it does not supply a comparison population or prove a causal link by itself. Keep the report’s source and timing alongside the underlying treatment documents, not in place of them.

II. How do courts examine a possible failure-to-warn claim?

A. Compare what the courts decided, and what they did not decide

Federal preemption is one issue that can arise when a state-law claim asks for a different prescription-drug warning. The question is not simply whether the FDA approved a label. Courts examine the federal framework, the information at issue, the particular proposed warning and the procedural record. The following decisions concern different medicines and different injuries. They illustrate legal questions a reader can investigate; they are not findings about the eye-risk communication described above.

In In re Fosamax, 118 F.4th 322 (3d Cir. 2024), the United States Court of Appeals for the Third Circuit decided on September 20, 2024 that the state-law claims at issue were not preempted on the record before it, vacated judgment for the manufacturer and remanded for further proceedings. The decision did not award damages or decide that a particular plaintiff’s injury was caused by the drug. In In re Taxotere, No. 2:22-md-03023-JTM-MBN (E.D. La. Jan. 16, 2026), the United States District Court, Eastern District of Louisiana, on January 16, 2026 cited In re Fosamax when describing the federal warning-label framework. The Louisiana federal court granted preemption summary judgment for two manufacturers of § 505(b)(2)-approved versions after finding no qualifying newly acquired information on that record. Together, the decisions show why the exact product, information and FDA history matter more than a general statement that all warning claims either succeed or fail.

The Third Circuit opinion concerned a brand-name manufacturer and expressly did not decide the different generic-drug regime. Its reasoning turned on the evidence before that court about the FDA’s response to a proposed warning and whether federal law prevented the state-law warning sought by the plaintiffs. A remand leaves issues for later proceedings. According to the Supreme Court docket, the Court denied a petition to review that Third Circuit decision on June 16, 2025; the docket records a procedural disposition, not a merits opinion. Someone comparing the opinion with another situation should identify the actual label proposal and agency response before assuming the same preemption analysis applies.

The Louisiana federal district court examined two products approved through the § 505(b)(2) pathway, which the opinion distinguishes from the abbreviated pathway for generic drugs. Its decision compared later information with the available earlier literature and concluded that the information identified on that record did not qualify for the asserted labeling change. It did not create a blanket rule that every case involving a generic medicine is preempted.

In Warner v. Amgen Inc., No. 25-1268 (1st Cir. Oct. 9, 2026), the United States Court of Appeals for the First Circuit decided on October 9, 2026 to affirm dismissal of the original pre-approval warning theory but reverse the denial of leave to amend a limited post-approval theory. The court allowed the proposed amendment to proceed only to the extent described in its opinion; it did not find that the amended claim would ultimately succeed. For a reader, the useful check is to separate a proposed original warning from a proposed later change and identify the information said to support each one.

The First Circuit addressed a biologic and the federal regulatory provision applicable to it; the new drug application regulation discussed above has a related but distinct context. The court evaluated studies presented in support of the requested amendment and the procedural question whether that amendment would be futile. A decision at that stage accepts or evaluates allegations under the applicable procedural standard; it is not the same as a trial finding. When reviewing any comparable opinion, note whether the court considered a motion to dismiss, a request to amend, summary judgment or a final judgment. The evidence available and the question answered can differ substantially.

These opinions come from different federal courts. The Third Circuit and First Circuit decisions have their own geographic and precedential reach, while a federal district court decision does not bind every court nationwide. State-law claims also depend on the state law that applies to the particular case. The cases are useful because they show the questions to ask about federal preemption and evidence; they cannot identify the governing state rule or establish an outcome without the relevant facts and law.

B. Assemble the evidence without assuming a legal outcome

Start with a product-and-timeline file. List each prescription, fill, dose change and relevant visit; attach the underlying records rather than relying on a reconstructed summary alone. Mark the first symptom, the first report to a clinician and the date of each examination as separate events. Preserve the medical diagnosis and the basis for it, including reports of relevant testing. If the diagnosis changed, keep both the earlier and later records so that the development of the assessment can be understood.

Different records answer different questions even when they share a date. A dispensing history may show a fill without showing whether a dose was taken; a clinician’s note may describe what was reported at a visit without identifying the exact package. Keep those distinctions visible in the timeline. If a family member is assembling the file, record whether an entry comes from the patient’s recollection, a witness, or a medical or pharmacy document. Where a record is missing, write down its likely custodian and the reason it may matter. This makes a later comparison more reliable than combining uncertain details into one confident account.

Next, create a warning file. Gather the prescribing information associated with the product, pharmacy materials actually received, instructions in the medical record and any communication from a treating professional that the patient retained. Note which version was in use at each point, if that can be established, and identify gaps rather than filling them from memory. A warning published after an event may be important context, but it does not automatically establish what was known or communicated before that event.

Then make an information file. Record the date and source of any study, agency communication or safety report you rely on, and preserve the complete document. Ask what the material reports: a suspected event, an observational association, a clinical-trial comparison, a regulator’s conclusion or a judicial finding. Check the medicine, outcome and population studied. The federal definition of newly acquired information calls for attention to what had previously been submitted to the agency and what the later information revealed. A reader normally cannot determine that entire regulatory history from a headline.

The question of medical causation requires its own file. A person may have a real injury while its cause remains disputed or uncertain. Relevant records can include prior vision history, other diagnoses, medication history, examination findings and the clinician’s explanation of alternative possibilities. Do not write a conclusion into the record as though it were a test result. Preserve the clinician’s actual assessment and note whether it addresses the individual patient, a general association, or both. The FDA’s description of overlapping risk factors makes that distinction particularly useful here.

When comparing a later label with an earlier one, save both versions in full and identify the section and revision information on each. A highlighted change is helpful as a navigation aid, but the surrounding language can qualify what changed. Keep the patient-facing material separately from the full prescribing information and note how each was obtained. If the version supplied at a particular visit cannot be located, identify that gap explicitly instead of treating the current online version as historical proof.

A claim analysis also needs the applicable state and the identity of the potentially responsible parties. A prescribing decision, pharmacy communication and product-labeling decision are different acts. Their legal significance cannot be inferred merely from a name on a package. Identify what each document shows and what remains unknown. An applicable state-law theory must be examined under the governing state’s rules, while a federal preemption defense must be evaluated against the relevant federal authorities and the specific record.

For each potentially relevant document, ask a concrete question. Does it identify the same product and strength? Was it created before or after treatment? Does it contain the full warning or only a summary? Does it describe a medical observation or state an opinion about its cause? Does an agency document report a requirement, a proposal or a completed label revision? Keeping those distinctions in a simple index makes it easier to compare records without treating every later development as proof of an earlier failure.

If there is an existing lawsuit, read the actual complaint or order before describing its status. A complaint contains allegations; an order may resolve only one motion or one party’s position. A headline about a case does not establish that a claim has been certified, that liability has been found or that an individual patient is included. The three opinions above demonstrate why the stage and scope of a decision matter: one remanded after vacating judgment, another granted summary judgment on a particular record, and another permitted a limited amendment.

Finally, keep questions of treatment, reporting and litigation distinct. A medical evaluation addresses the patient’s health. A voluntary FDA report contributes information to the agency’s monitoring system. A possible legal claim requires evidence and analysis under applicable law. Taking one step does not automatically complete either of the others. Retaining accurate copies and identifying unknowns is useful for all three without predetermining their answers.

Conclusion

The FDA’s communication identifies a potential increased risk and requires a labeling change; it does not resolve an individual patient’s diagnosis or claim. A person with the symptoms described by the agency should seek prompt medical attention. For the legal questions, preserve the treatment timeline, the relevant versions of the warning and the medical evidence, then examine the applicable state law and federal framework against the actual facts. The decisions discussed here show why both the information available and the procedural stage matter.

📄 Official circular