On June 4, 2026, the Judicial Panel on Multidistrict Litigation centralized federal claims against Regeneron Pharmaceuticals, Inc., Sanofi-Aventis U.S. LLC, and Genzyme Corporation as MDL No. 3180 in the District of New Jersey, before Judge Zahid N. Quraishi. Fifteen transferred actions and seven potential tag-along cases alleged that Dupixent caused or accelerated cutaneous T-cell lymphoma.
Each plaintiff retains an individual claim rather than membership in a certified class. Limitations, prescribing history, and damages stay plaintiff-specific, and no settlement binds anyone who has not agreed to it, which separates this posture from aggregate litigation structures that resolve claims collectively.
This analysis covers the procedural record, the causation dispute, the medical evidence that dispute requires, preemption and limitations defenses, corporate exposure, and claim-screening standards.
Where Dupixent Lymphoma Litigation Stands in 2026
The Panel consolidated the actions under 28 U.S.C. § 1407 in In re: Dupixent (Dupilumab) Products Liability Litigation (2026). Movants sought the Northern District of Georgia, which held four of the fifteen actions. Defendants supported centralization but asked for the Southern District of New York.
New Jersey prevailed on witness and evidence geography. Sanofi's principal place of business sits in the district, and Regeneron, headquartered in nearby Tarrytown, New York, maintains corporate offices there.
The Panel identified three common questions: whether the scientific literature shows a causal link between Dupixent and CTCL, when defendants should have learned of it, and whether their warnings were adequate.
The first-filed action was Richardson v. Regeneron Pharmaceuticals, Inc., et al. (2025), No. 3:25-cv-01125, a wrongful death claim filed October 1, 2025, in the Middle District of Tennessee. The docket had grown to 26 pending actions by the Panel's July 1, 2026 statistics report.
One scope question remains open and matters for future filings. Defendants asked the Panel to limit the proceeding to CTCL claims, while movants argued the literature also links Dupixent to peripheral T-cell lymphoma and anaplastic large cell lymphoma, subtypes similar enough that expert testimony would overlap. The Panel declined to rule, noting that no plaintiff before it alleged a non-CTCL T-cell lymphoma, and left expansion to the conditional transfer process.
Anaplastic large cell lymphoma has been litigated before as a product-attributed malignancy, and the BIA-ALCL claims show how that kind of causation record gets built.
No bellwether selection, expert-admissibility ruling, or settlement has occurred.
The Causation Dispute at the Center of the Litigation
Multiple peer-reviewed cohorts report elevated CTCL rates among treated patients. The fight is over what produced that pattern.
Unmasking holds that CTCL was present from the start, mislabeled as eczema, and became visible once Dupixent cleared the inflammation covering it. Acceleration holds that blockade of the interleukin-4 receptor drove an existing malignant clone to proliferate faster, while de novo promotion, the hardest theory to prove, holds that the drug created the malignancy.
Defendants will build their causation case on unmasking, and a 530-patient cohort from University Medical Center Utrecht supplies the evidence. Among 14 patients (2.6%) who developed clinical suspicion of CTCL during treatment, review of pretreatment biopsies showed 3 had pre-existing mycosis fungoides misdiagnosed as atopic dermatitis, in one instance for nearly 14 years.
That study also described a benign, reversible lymphoid reaction that mimics mycosis fungoides clinically but differs histologically. Where a plaintiff's diagnosis rests on thin pathology, defendants will argue the finding was this reaction and not lymphoma at all.
The mechanistic case for acceleration rests on an immunotranscriptomic analysis in the Journal of Allergy and Clinical Immunology, which found a proportional reporting ratio of 30.0 (95% CI 25.0 to 35.9) for CTCL in FDA adverse event data. Its authors concluded that "dupilumab might be causing an unmasking or progression of CTCL via the same mechanism through which it improves atopic dermatitis: IL13 receptor blockade, which leads to increased IL13 in the local milieu, driving CTCL stimulation and progression."
Two database cohorts put numbers on the association. One retrospective cohort found an odds ratio of 4.10 (95% CI 2.055 to 8.192), with no elevated risk for any other cutaneous or lymphoid malignancy. A TriNetX cohort matching 19,612 treated patients put the relative risk at 4.59 (95% CI 2.459 to 8.567).
The study that most complicates a pure unmasking defense involves asthma patients. An asthma cohort study in the European Respiratory Journal reported a hazard ratio of 4.58 (95% CI 1.82 to 11.53) for T-cell and natural killer cell lymphomas against inhaled corticosteroid comparators. Asthma patients lack the eczematous skin presentation that obscures CTCL, so misdiagnosis is a weaker explanation in that population, though the journal has since published four comments on the paper.
Defendants will also press a JAAD response letter arguing that severity confounding inflated the odds ratio, because Dupixent treats severe disease while comparators likely had milder disease. Each of these studies frames its finding as an association rather than proof of causation, and EORTC consensus recommendations advise only that the drug should be avoided where mycosis fungoides or Sézary syndrome is suspected or confirmed.
What the Causation Question Demands from the Medical Record
Timing controls causation, which narrows the record set to a specific list. An expert works through differential etiology: confirm the diagnosis, establish that the drug can cause the disease in general, then rule out the other explanations available for this plaintiff.
Dermatopathology reports and the underlying slides carry the most weight. Early mycosis fungoides overlaps with eczema, so a pretreatment biopsy read as chronic spongiotic dermatitis may, on expert re-review under current criteria, already show malignant features, which would place biological onset before exposure. Those slides usually remain with the original pathology laboratory rather than the treating hospital, so obtaining them runs through compelling third-party custodians rather than a standard records authorization.
Whether that re-review must be blinded is contested. In Adams v. Laboratory Corp. of America, 760 F.3d 1322 (11th Cir. 2014), a district court excluded a pathologist's slide-review opinion under Rule 702 partly because she had not used a blinded protocol, citing professional society guidelines.
The Eleventh Circuit reversed, calling that reasoning manifestly erroneous and treating the objection as a question of weight for the jury. Blinding still strengthens an opinion, and defendants will attack its absence, but that attack is not a reliable path to exclusion.
A published systematic review of reported cases found that post-treatment biopsies showed greater cell density and a predominantly lichenoid pattern, against more varied patterns and lower density in pretreatment specimens.
Clonality testing carries less weight than its reputation suggests. Among the cases in that review, T-cell receptor gene rearrangement studies came back negative in 60% and equivocal in 20%, which the authors flagged as a diagnostic pitfall. Clonality results rarely date the malignant clone on their own, and missing pretreatment tissue compounds that limitation rather than creating it.
The remaining records anchor the timeline and rule out competing explanations:
- Immunohistochemistry markers (CD2, CD3, CD4, CD5, CD7) across every biopsy show when an aberrant phenotype first appeared, read alongside prior treatment history for atopic dermatitis.
- Pharmacy and specialty pharmacy records fix the exposure window. Claims databases estimate exposure from days supply, and one analysis documented days-supply inconsistencies for biologics dosed at intervals longer than four weeks, so claims-derived duration warrants cross-validation against dispensing logs and prescriber notes.
Everything reduces to one comparison: the dates of the first abnormal biopsy, the first clonal result, and the formal diagnosis, set against the date of the first fill. Where the earliest objective evidence of CTCL predates that fill, the claim fails at the specific-causation stage.
Threshold Legal Issues Shaping Early Rulings
Defendants will raise preemption before reaching causation, and the current label leaves them little to work with. Impossibility preemption requires a showing that compliance with both federal labeling rules and state tort duties was impossible.
Under Wyeth v. Levine, 555 U.S. 555 (2009), that defense fails absent clear evidence the FDA would have rejected the warning. Merck Sharp & Dohme Corp. v. Albrecht, 587 U.S. 299 (2019), defined that standard as a showing that the manufacturer fully informed the agency and the agency refused, and made the question one of law for the judge.
The changes-being-effected pathway under 21 C.F.R. § 314.70 permits a manufacturer to strengthen a warning on filing, without waiting for approval. The current prescribing information, revised April 2026, contains no lymphoma, CTCL, or malignancy warning in any section. Without an agency rejection of such a supplement, the defense is currently unavailable.
Each side claims the federal safety signal. Plaintiffs point to the FDA's quarterly FAERS signals list, published March 31, 2025, which identified CTCL as a potential signal and stated the agency was evaluating the need for regulatory action. Defendants answer that an open review producing no mandated label change reflects unsettled causation evidence.
Warning claims then run through the learned intermediary doctrine, under which a manufacturer's duty is owed to the prescribing physician rather than the patient. New York's formulation in Martin v. Hacker, 83 N.Y.2d 1 (1993), is representative. Prescriber testimony that the physician would have prescribed Dupixent regardless can defeat causation at summary judgment even where the label was inadequate.
Limitations turns on discovery-rule variation, because many plaintiffs were diagnosed before the 2024 literature appeared. Multi-element states start the clock when the causal connection becomes reasonably knowable. New York is stricter: under In re New York County DES Litigation, 89 N.Y.2d 506 (1997), discovery of the injury means discovery of the diagnosis, and later discovery of an alleged drug cause does not control.
Label Silence and Corporate Exposure for Sanofi and Regeneron
Regeneron's annual report for the year ended December 31, 2025 states that Sanofi collaboration revenue, most of which is attributable to its share of Dupixent profits, represented 41% of total revenues, against 32% in 2024, and that the company is substantially dependent on that revenue.
The same filing's legal matters note describes proceedings involving EYLEA and Praluent and reports that accruals for loss contingencies were not material as of December 31, 2025. It does not identify a Dupixent lymphoma proceeding, though the annual report was filed before the Panel created MDL 3180.
Both disclosures bear on the labeling decision. A warning added through the changes-being-effected pathway would weaken the failure-to-warn theory going forward while confirming notice for the exposure period already at issue, and it would attach to the product supplying the largest single share of one defendant's revenue.
Screening Dupixent Claims Across a Growing Inventory
Screening criteria follow from the causation and limitations analysis, because the facts that defeat a claim at summary judgment defeat it at intake. Four findings ordinarily disqualify a claim: no documented Dupixent use, no pathology-confirmed CTCL diagnosis, a confirmed diagnosis predating the first dose, or an expired limitations period. Borderline limitations questions belong in attorney review rather than automatic decline.
A viable claim shows documented use with a start date, a biopsy-confirmed subtype, a diagnosis postdating the first dose, and a prior atopic dermatitis diagnosis establishing clinical context. Several factors then separate stronger profiles from weaker ones:
- Subtype: mycosis fungoides and Sézary syndrome account for the cases collected in the systematic review discussed above, and they align with the MDL's current scope, while non-cutaneous T-cell lymphomas remain outside it pending conditional transfer.
- Age at diagnosis: the TriNetX analysis placed 54.5% of post-treatment CTCL cases in patients over 60, and the pharmacovigilance analysis treated the signal among patients aged 21 to 45 as evidence against pure unmasking, so a younger claimant is not weaker on age alone.
- Latency: most cases in the retrospective cohort were diagnosed more than a year after use, so a longer interval is not disqualifying. Short-latency claims face the strongest unmasking argument.
- Progression pattern: initial skin improvement followed by rapid worsening fits the acceleration theory.
- Disease burden at diagnosis: in that same systematic review, 10 of 18 stageable cases presented at stage III or IV, which bears on damages and on the diagnostic-delay narrative.
Consistent disposition coding across an inventory serves an evidentiary purpose beyond efficiency, since screening decisions made on shifting criteria are harder to defend if challenged.
Practice Implications for MDL 3180 Counsel
Plaintiff counsel should establish early whether archived pretreatment tissue exists at all, because its availability, not its content, decides whether onset can be dated. Defense counsel will work the diagnosis itself, arguing the pathology never confirmed lymphoma, and will take prescriber depositions early enough to lock in a prescribe-anyway answer.
The first Rule 702 ruling will price this litigation. Until it issues, both sides face the same practical demand: per-plaintiff limitations analysis under the law of nine states, and disciplined reviewing claimant medical files long before bellwether selection. For a comparable early-stage pharmaceutical causation fight, see the NEC litigation.
