RE: Depot Medroxyprogesterone Acetate (Depo-Provera) and Intracranial Meningioma — Plaintiff Science, Daubert Posture, and Litigation Strategy
MDL Reference: In re: Depo-Provera (Medroxyprogesterone Acetate) Products Liability Litigation, MDL No. 3140 (N.D. Fla., Hon. M. Casey Rodgers).
Date: May 2026
Classification: Background research brief — not legal advice. All citations are primary-source where available; where a claim rests on a secondary source the secondary is identified expressly. Sources verified through publicly accessible URLs at the dates indicated. Snapshot values (case counts, settlement projections, regulatory deadlines) are dated inline; the reader should confirm current status before relying on this material in court.
The Depo-Provera meningioma litigation has moved from an unranked theoretical claim to a Daubert-ready general-causation case. Five published epidemiological studies between March 2024 and September 2025 — across multiple investigator groups (Roland in France; Griffin/UAB across MarketScan, an active-comparator design, and an Alabama Medicaid cohort; and Cleveland Clinic / Case Western using the TriNetX federated EHR) — have all reported elevated risk estimates for intracranial meningioma in DMPA users (odds-ratio range 2.43 to 6.71 depending on cohort and exposure window). A monotonic dose-response (linear-trend p < 0.0001 across four duration categories) has been documented in the largest U.S. study (Griffin et al., Cancers 2024). The FDA, after initially denying Pfizer's early-2024 label-change application, approved the meningioma warning in December 2025 (FDA, 2025).
Three structural facts now drive the case:
The June 24–26, 2026 Daubert hearings will resolve general causation. The first bellwether trial (Toney v. Pfizer) is scheduled for December 7, 2026. As of April 1, 2026, approximately 3,490 cases are pending in MDL 3140, with state-court parallel filings bringing the total above 4,000 and an intake rate of roughly 650 new filings per month (Lawsuit Information Center, April 2026 — secondary; verify against the Northern District of Florida MDL portal before relying on the count).
The strongest plaintiff openings are the PIK3CA/PR molecular signature, the active-comparator replication of risk (which neutralizes confounding-by-indication), the cyproterone acetate (CPA) regulatory analogy from France, and the FDA's own December 2025 conclusion. The largest plaintiff exposures are the small absolute risk per individual user (NNH ≈ 1,152, but attributable-risk percentage 59%), the tumor-promotion-vs-initiation framing, and a small set of secondary-source claims that should be re-verified against primary sources before the June Daubert hearings (§8.5).
This memorandum walks through the product, the epidemiological record, the progesterone-receptor mechanism, the dose-duration data, the defense expert positioning anticipated for June 2026, the FDA and international regulatory record, the MDL procedural posture, and the gaps in current plaintiff filings most likely to be exploited at Daubert.
Depot medroxyprogesterone acetate (DMPA) is a synthetic progestin — the 17α-acetate ester of 6α-methylhydroxyprogesterone — formulated as a microcrystalline aqueous suspension for intramuscular or subcutaneous depot injection. The 150 mg intramuscular dose (Depo-Provera CI) and the 104 mg subcutaneous dose (depo-subQ provera 104) are administered every 12–13 weeks. Each injection establishes a slow-release depot that maintains contraceptive serum levels for at least 13 weeks; residual MPA remains detectable in serum for months to over a year after the final dose, with adipose redistribution further extending biological exposure (FDA Label, 2025).
The depot architecture matters for litigation. Unlike an oral progestin, a single DMPA injection delivers a multi-month exposure that the patient cannot withdraw — relevant both to the failure-to-warn theory (the warning must reach the patient before the depot is established) and to dose-duration analysis (one injection ≈ one quarter of cumulative exposure, irrespective of continuation).
MPA binds the progesterone receptor (PR) with the fourth-highest affinity among eleven tested progestins (Griffin et al., 2024, citing the underlying receptor-binding-affinity literature; the primary citation should be confirmed and cited directly in deposition or expert-report use). PR-mediated binding is the contraceptive mechanism — suppression of the hypothalamic-pituitary-ovarian axis — and, as discussed in §3, is also the mechanism by which the drug interfaces with PR-positive meningiomas. Secondary glucocorticoid and weak androgenic activity drives the bone-mineral-density side effect underlying the 2004 Black Box Warning but is not central to the meningioma claim.
Exposure for litigation purposes is the cumulative duration of DMPA use, measured in injection cycles or years and reconstructed from medical, pharmacy, and insurance billing records (CPT 11981 for administration; J1050 for the drug code). The U.S. insurance-database studies (Griffin et al., 2024; JAMA Neurology TriNetX, 2025) operationalize exposure through prescription-and-administration claims data. The breakpoints defense will exploit are: ≤1 year (OR 1.35, CI 1.17–1.56), >1–2 years (1.68, 1.24–2.27), >2–3 years (2.39, 1.68–3.41), and >3 years (3.24, 2.49–4.21) — all from Griffin et al. (2024), Table 4 (cerebral meningioma).
The Abou-Al-Shaar et al. (2023) clinical case series — the most complete published phenotyping of DMPA-associated meningioma — reported a mean DMPA exposure of 15.5 years (range 6–26) across 25 women at the University of Pittsburgh Medical Center. The litigation cohort sits in the high-exposure tail of the dose-response curve, where plaintiff-side ORs are largest.
The plaintiff general-causation case rests on five published epidemiological studies between March 2024 and September 2025, plus the Abou-Al-Shaar 2023 clinical case series. The five core studies, with anchor estimates:
Roland et al. (March 2024), BMJ — French national case-control. Using the French National Health Data System (SNDS), 18,061 women who underwent intracranial surgery for meningioma between 2009 and 2018 were matched to 90,305 controls by birth year and area of residence. Injectable MPA exposure ≥1 year was associated with an adjusted OR of 5.55 (95% CI 2.27–13.56). The same study identified elevated risk for medrogestone (3.49) and promegestone (2.39) but no excess risk for endogenous progesterone, dydrogesterone, or levonorgestrel-IUD users (Roland et al., BMJ 2024;384:e076045, PMID 38537944).
Griffin et al. (October 2024), Cancers — IBM MarketScan. 117,503 meningioma cases matched 10:1 to 1,072,907 controls (2006–2022). Injectable MPA adjusted overall OR 1.53 (95% CI 1.40–1.67); cerebral meningioma 1.68 (1.50–1.87); cerebral >3-year exposure 3.24 (2.49–4.21); linear trend p < 0.0001. Oral MPA showed no association (0.97). Spinal meningioma showed no association (0.77, CI 0.53–1.12). (Griffin et al., Cancers 2024;16(19):3362, PMC11482550.)
JAMA Neurology (September 2025) — Cleveland Clinic / Case Western TriNetX. A retrospective propensity-matched cohort drawn from the TriNetX federated database covering more than 10 million women across 68 U.S. healthcare organizations (December 2004 to December 2024); after propensity-score matching, n=88,667 in the DMPA arm. RR 2.43 (95% CI 1.77–3.33) for meningioma in DMPA users versus matched controls; attributable-risk percentage 59%; number-needed-to-harm ≈ 1,152. Risk was statistically detectable in users with >4 years of cumulative exposure or initiation after age 31 (4–6 years RR 3.00 [1.47–6.13], >6 years RR 3.90 [1.95–7.81]; initiation at ages 31–40 RR 3.77 [2.05–6.95]); within shorter-duration and younger-initiation strata, the propensity-matched estimates did not reach statistical significance. Oral MPA RR 1.18; intrauterine devices, combined oral contraceptives, and progestin-only pills showed no increased risk. (JAMA Neurology, doi:10.1001/jamaneurol.2025.3011; PMID 40892397.)
Active-comparator case-control (Griffin & Arend, July 2025), Current Oncology (PMC12293745). Matched case-control of 241 cerebral meningioma cases with three matched control sets per case (brain-tumor, breast-tumor, skin-tumor); active comparator defined as levonorgestrel-containing or norethindrone-containing exposure; non-active comparator defined as no documented oral contraceptive use. DMPA exposure within one year of diagnosis: OR 3.27 (95% CI 1.19–9.89) versus active comparator; OR 6.71 (95% CI 2.69–18.09) versus non-active control. The active-comparator design is critical for confounding-by-indication: DMPA carries elevated risk even relative to other hormonal contraceptive users.
Alabama Medicaid case-control (September 2025). Prolonged DMPA exposure adjusted OR 4.01 (95% CI 1.21–13.30) versus active comparator. The Medicaid-population design adds demographic diversity absent from commercial-insurance databases (medRxiv preprint, June 2025; reported as forthcoming in Cancers 2025 — confirm publication-status before pleading).
Abou-Al-Shaar et al. (2023), J. Neurol. Surg. B — clinical phenotyping. 25 women, 49 total meningiomas (mean 1.96 per patient, range 1–6), 17/25 (68%) with at least one skull-base meningioma; among the 47 resected/biopsied tumors, all 47 PR-positive on IHC (36 [77%] WHO Grade I, 11 [23%] WHO Grade II, mean Ki-67 2.2%, range 1–5%). Among 10 patients who discontinued DMPA, 5 (50% of evaluable cessation cases; 20% of the full cohort) had documented post-cessation tumor shrinkage; 2 lost to follow-up; 3 had insufficient post-cessation follow-up.
A series of Indonesian case-control studies — cited via the Frontiers in Global Women's Health 2025 editorial as secondary citations — have reportedly returned ORs in the 2.47–3.13 range with DMPA-exposure prevalences exceeding 80% among surgical meningioma cases in some series. Primary-source verification is on the re-verification backlog (§8.5) and should be completed before deposition or trial use.
The plaintiff-favorable feature of this literature is convergence across designs and (with caveats on investigator overlap — Griffin/UAB authored or co-authored the MarketScan, active-comparator, and Alabama Medicaid analyses) across investigator groups. National-population case-control (France), commercial-insurance case-control (Griffin), federated EHR cohort (JAMA Neurology), active-comparator case-control, Medicaid case-control (Alabama), and clinical case series (Pittsburgh) all return elevated risk estimates with overlapping confidence intervals and consistent dose-response orientation. No published primary epidemiology study of which we are aware shows a null or protective effect for injectable MPA. A 2025 Expert Opinion on Drug Safety "comparative safety study" (Source Index entry 7) is a defense-relevant comparative-safety paper, not a primary epidemiology study, and should be read carefully before counsel uses or rebuts it at Daubert.
The litigation record has gaps that defense will exercise. There is no large randomized controlled trial of DMPA versus placebo or alternative contraceptive that follows participants for the long durations needed to detect meningioma at population-attributable rates — and there will not be one; the ethics of randomizing women to an arm with documented carcinogenic concern preclude it. The defense argument that "no RCT establishes causation" is therefore structurally available. The plaintiff response is the same response that prevailed in cigarettes-and-lung-cancer: Bradford Hill analysis, mechanistic evidence, and dose-response substitute for randomization where randomization is precluded. Published literature on African and South Asian populations — where DMPA use is highest in absolute terms — remains limited; this is more a global-health-equity flag than a Daubert-load-bearing gap, but plaintiff experts should be prepared to acknowledge it.
Meningiomas are among the most hormone-receptor-expressing primary CNS tumors. Reported PR expression prevalence across published cohorts spans 33% to 89% (Bogotá cohort PMC7842106; Griffin et al., 2024 review). The wide range reflects methodologic heterogeneity (immunohistochemistry threshold, tumor grade, location), but every published cohort reports majority or near-majority PR positivity, with higher prevalence in WHO Grade I tumors and skull-base location — both features of the DMPA-exposed phenotype documented by Abou-Al-Shaar et al. (2023). MPA binds PR with high affinity (fourth among eleven tested progestins; Griffin et al., 2024, citing the underlying receptor-binding literature). The mechanistic prediction is direct: a PR-positive tumor exposed to circulating MPA at depot-injection levels for years has the receptor machinery to translate exposure into proliferative signaling.
The strongest mechanistic evidence is the mutational landscape work of Peyre et al. (Annals of Oncology 2018, PMID 29206892). Their analysis of progestin-associated meningiomas reported PIK3CA mutation in approximately 35% of progestin-associated tumors versus approximately 3% in sporadic controls (an order-of-magnitude enrichment); TRAF7 mutation in 40% versus 26%; NF2 loss-of-function in 7.5% versus 32% (a four-fold under-representation); and chromosome 22 loss in 17.5% versus 48%. The progestin-associated tumors were drawn from cyproterone acetate (CPA) and other high-dose progestin exposures, but the PIK3CA-enriched / NF2-and-chromosome-22-depleted signature translates directly: it indicates a distinct progestin-driven oncogenic pathway, separate from the NF2-loss / chromosome-22-loss pathway that characterizes most sporadic meningiomas. Every PIK3CA-mutant tumor in the Peyre cohort expressed PR in 50–100% of tumor cells — coupling the receptor and the PI3K-pathway activation in a mechanistically coherent way.
Subsequent work has reinforced the coupling. A Cancer Gene Therapy paper (2023) exploring the respective roles of Pik3ca mutations and CPA in mouse meningioma tumorigenesis showed both that Pik3ca activation alone can drive tumorigenesis in mouse models and that CPA exposure further accelerates the phenotype. A 2025 review (PMC12209388) on "the hormonal nexus in PIK3CA-mutated meningiomas" extends this picture into the contemporary literature.
Defense will read the mouse-model data — particularly that Pik3ca activation alone can drive tumorigenesis — and argue that DMPA-exposed plaintiffs may have had subclinical, mutation-driven meningiomas that would have manifested regardless of exposure. The plaintiff response, beyond the population-level dose-response argument, is that Peyre's human-tumor data show PIK3CA mutations concentrated in progestin-exposed tumors, not present at equal frequency in sporadic controls. The progestin exposure is correlated with the mutation landscape — consistent with progestin-driven mutagenesis or progestin-driven selection for PIK3CA-mutant clones — not independent of it.
The Abou-Al-Shaar 2023 series, supplemented by the Frontiers in Global Women's Health 2025 editorial review, documents a distinctive clinical phenotype: high PR / low ER expression; meningiomatosis (mean 1.96 tumors per patient; range 1–6); skull-base predilection (68% of patients with at least one skull-base meningioma); and post-cessation tumor shrinkage in 5 of 10 evaluable cessation cases (i.e., among the subset of the 25-patient cohort who discontinued DMPA and had adequate post-cessation follow-up). Each feature supports the causal-inference framework. Multiplicity is unusual in sporadic meningioma. Skull-base location aligns with regional PR-expression patterns. Documented post-cessation regression — though the denominator is small (5 of 10 evaluable; 5 of 25 in the full cohort) — is the most telling: a tumor whose growth is sustained by an exogenous drug, and that shrinks when the drug is withdrawn, is mechanistically coupled to that drug.
A frequently cited supporting observation is that meningiomas grow more rapidly during pregnancy, when endogenous progesterone serum levels rise substantially above luteal-phase baseline (background neurosurgical observation referenced in Griffin et al., 2024; underlying primary references such as Lusis EA et al., Neurosurgery 2012, should be cited directly in expert reports). The pregnancy data is observational and not load-bearing, but it constitutes a "natural experiment" for the receptor-mediated mechanism: the same PR system that responds to elevated endogenous progesterone responds to exogenous MPA. This is coherence evidence in the Bradford Hill sense.
For Daubert "fit" — the FRE 702 requirement that expert testimony fit the question of fact — the mechanism evidence does substantial work. The molecular pathway (PR binding → downstream PI3K-AKT activation in PIK3CA-mutant tumors → sustained proliferative signaling) is biologically specified, supported by tissue-level evidence, replicated in animal models, and clinically observable in regression-after-cessation cases. A defense expert who concedes the dose-response (which the data make difficult to deny outright) but argues that "the mechanism is unclear" should be cross-examined against the Peyre, Abou-Al-Shaar, Cancer Gene Therapy 2023, and PMC12209388 record.
The dose-response gradient is the most decisive single Bradford Hill criterion in the plaintiff Daubert case. It comes principally from Griffin et al. (2024) and is corroborated by JAMA Neurology (2025).
Griffin et al. (2024) — IBM MarketScan, 117,503 cases, 1,072,907 controls, cerebral meningioma (Table 4):
| Duration | Adjusted OR | 95% CI |
|---|---|---|
| ≤1 year | 1.35 | 1.17–1.56 |
| >1–2 years | 1.68 | 1.24–2.27 |
| >2–3 years | 2.39 | 1.68–3.41 |
| >3 years | 3.24 | 2.49–4.21 |
Linear trend p < 0.0001 (PMC11482550). The JAMA Neurology TriNetX analysis (Sept 2025) reports risk confined to users with >4 years of cumulative exposure or initiation after age 31 (4–6 years RR 3.00 [1.47–6.13], >6 years RR 3.90 [1.95–7.81]; initiation at ages 31–40 RR 3.77 [2.05–6.95]); the published 2.43 RR is the overall any-DMPA propensity-matched estimate. The active-comparator design (Griffin & Arend, Current Oncology, 2025; PMC12293745) reports OR 3.27 vs. active comparator and 6.71 vs. non-active comparator for one-year-prior exposure.
A central tactical point: no published study has identified a minimum safe duration of DMPA use below which meningioma cerebral risk is null. Even ≤1 year of exposure shows statistically significant cerebral risk elevation (Griffin OR 1.35, lower CI bound 1.17 above 1.0). Within shorter strata of the JAMA Neurology TriNetX analysis the propensity-matched estimates did not reach statistical significance — a contrast plaintiff experts should engage explicitly against the corroborating Griffin ≤1-year cerebral OR. Defense framing of "less than one year of use, or even a single injection, cannot cause meningioma" must contend with the Griffin ≤1-year data point. The conceptual move defense will make is to distinguish de novo tumor initiation from promotion of existing subclinical tumors (§5.2). Even granting that distinction, the relevant clinical injury — symptomatic meningioma requiring surgery — is the endpoint that the dose-response curve maps. Whether DMPA initiates the tumor or accelerates a subclinical lesion to clinical presentation, the patient-level injury is the same and the dose-duration curve documents the exposure-response relationship for that injury.
Anatomical and route specificity strengthen the inference. Griffin et al. (2024) reports cerebral meningioma OR 1.68 (1.50–1.87) but spinal meningioma OR 0.77 (0.53–1.12). The null effect on spinal meningioma is itself causal-inference evidence. If the cerebral signal were a confounding artifact, that bias should affect spinal-meningioma capture similarly. The cerebral-vs-spinal differential is consistent with differential PR expression patterns documented in cranial versus spinal meningeal tissue and argues against confounding as the primary explanation. The same study found no association between oral MPA and meningioma (OR 0.97). This route-of-administration specificity — injectable yes, oral no — further constrains the confounding hypothesis: oral and injectable users overlap on most demographic and behavioral confounders; the principal differences are pharmacokinetic. The route-specific signal is consistent with a depot-pharmacokinetics-driven mechanism.
Plaintiff experts in Daubert testimony should lead with the dose-response (linear-trend p < 0.0001 across four duration categories is textbook biological-gradient evidence and the hardest single element for defense to attack), couple to anatomy and route (cerebral-not-spinal, injectable-not-oral as the structural answer to confounding), use the active-comparator OR 3.27 (vs. active) / 6.71 (vs. non-active) to defeat confounding-by-indication, and bracket the duration in expert reports with sensitivity analyses placing the individual plaintiff explicitly on the Griffin curve.
Pfizer is expected to advance four primary defense lines at the June 2026 Daubert hearings, plus several supporting arguments, based on publicly visible filings and litigation reporting (Lawsuit Information Center, April 2026 — secondary; verify against the operative defense expert reports and motion practice once filed).
The classical Daubert argument: observational studies cannot demonstrate causation, only correlation. The plaintiff response is structural Bradford Hill. Of the nine criteria, the DMPA–meningioma association satisfies strength (ORs 2.43–6.71), consistency (≥6 published studies, multiple investigator groups, three countries, multiple database types), specificity (cerebral not spinal; injectable not oral), temporality (exposure precedes diagnosis in all designs), biological gradient (p < 0.0001), plausibility (PR expression and PIK3CA signature), coherence (pregnancy growth acceleration; tumor regression on cessation in 5 of 10 evaluable Abou-Al-Shaar cessation cases), and analogy (CPA — structurally similar progestin, established 7-fold overall and >20-fold high-cumulative-dose meningioma risk). The "experiment" criterion is partially satisfied by the post-cessation regression observation in the Abou-Al-Shaar cohort. This is the same evidentiary architecture that satisfied Daubert in cigarettes-and-lung-cancer: observational, mechanistic, dose-response, and analogical. No RCT was needed.
Defense will argue that a single injection or sub-one-year exposure cannot cause meningioma. Plaintiff rebuttals: (a) Griffin et al. cerebral ≤1-year OR 1.35 (1.17–1.56) is statistically significant; (b) the active-comparator study OR 3.27 for exposure within one year of diagnosis (PMC12293745) directly addresses recent-exposure cases; (c) the de novo-vs-promotion distinction does not exculpate the manufacturer because tumor promotion produces the same patient-level injury. The conceptual concession defense often makes — that DMPA "promotes" rather than "initiates" — is itself an admission that the drug interacts with tumor biology in a clinically meaningful way. Plaintiff counsel should also engage the JAMA Neurology stratified-significance pattern (risk confined to >4-year / initiation-after-31 strata) head-on rather than treat it as a non-issue: the framing is that JAMA Neurology's propensity-matched design was underpowered in the shorter-duration strata, not that risk is null below the breakpoint, and that the corroborating Griffin ≤1-year cerebral OR (1.35) is what the cross-study consistency rests on.
Defense filings have argued that meningioma risk does not extend more than two years after cessation. This argument is conceptually self-defeating. If risk attenuates upon cessation, then the drug was sustaining risk during use — which is the plaintiff theory. The Abou-Al-Shaar 5-of-10-evaluable post-cessation shrinkage finding is consistent with this. The argument that risk decays post-cessation is functionally the argument that DMPA is biologically active on meningioma growth. Defense will likely refine this into a statute-of-limitations argument; plaintiff teams should be prepared to brief this against the latency literature for meningioma generally and against the specific Abou-Al-Shaar phenotype data.
Defense has signaled federal-preemption arguments based on the FDA's pre-2025 prescribing-information record. This argument is significantly weakened by the FDA's December 2025 label change — adopted after Pfizer's resubmission. If the FDA permitted the warning, the FDA did not preempt a state-law duty to warn that aligned with the warning the FDA accepted. Judge Rodgers has ruled that preemption and Daubert determinations in MDL 3140 will bind all cases (MDL Update, April 2026 — secondary; confirm scope against the operative case-management order).
Small absolute risk (NNH ≈ 1,152, AR% 59%). Defense will frame meningioma as "extremely rare." Plaintiff response: absolute risk is a damages-distribution question, not a causation question; the JAMA Neurology attributable-risk percentage of 59% means that, among DMPA-exposed users who develop meningioma, the majority of those events are attributable to the exposure. With approximately 74 million women using DMPA annually worldwide (per Frontiers 2025; verify against WHO or primary contraceptive-use survey data — see §8.5), even a "rare" attributable risk maps to thousands of cases per year (the Frontiers 2025 estimate range is 2,800–18,000 attributable meningiomas annually).
Confounding variables. Defense will argue residual confounding by age, race, prior cranial radiation, breast cancer history, BMI. Plaintiff response: Griffin et al. adjusted for age and the Elixhauser comorbidity index; Roland et al. matched on birth year and area; the active-comparator study (PMC12293745) eliminates confounding-by-indication by comparing DMPA to other hormonal contraceptive users.
French study small DMPA exposure count. Defense will note Roland et al. had ~20 MPA-exposed cases, producing wide CI 2.27–13.56. The point estimate (5.55) is supported by the larger U.S. studies. The French study's value is in population-level design, not MPA-specific case count.
Tumor promotion vs. initiation. Defense will argue DMPA "accelerates growth of pre-existing lesions" rather than "creates de novo tumors." Promotion that converts an asymptomatic subclinical lesion into a symptomatic, surgery-requiring tumor is causation under most product-liability frameworks. The PIK3CA mutational data (Peyre 2018) further suggests DMPA may be implicated in initiation as well, by selecting for or driving mutational events in PR-positive cells.
The 2025 Expert Opinion on Drug Safety "comparative safety" paper. Plaintiff experts must read this paper carefully, characterize its methodology relative to the active-comparator study (PMC12293745), and prepare to cross-examine the design choices. (Source Index entry 7.)
Defense expert reports often originate with consultancies that have ongoing financial relationships with industry. Plaintiff teams should as routine practice obtain Rule 26 disclosures, identify funding relationships, and prepare cross-examination on consulting income, prior testimony, and any past advocacy in regulatory comment proceedings on DMPA. If a defense expert's prior published work has appeared in venues with disclosed industry funding, that record belongs in the cross outline.
The FDA approval-and-warning history for DMPA is unusual and litigation-relevant. The drug was rejected for contraceptive use four times (1972, 1974, 1978, 1983) before being approved in October 1992. Each rejection cited safety concerns: early rejections noted potential cancer risks generally; the 1978 rejection was driven by malignant tumor findings in a beagle study; the 1983 rejection followed a Public Board of Inquiry. The 1992 approval came over the opposition of women's-health advocates (TruLaw timeline, secondary; for any pleading reliance, cite directly to FDA records and Federal Register entries rather than to litigation timelines).
In 2004, FDA added a Black Box Warning for bone-mineral-density loss with prolonged DMPA use, recommending that DMPA be limited to two years where alternatives were available. This is the first FDA-imposed warning specifically tied to duration of DMPA use; the meningioma warning would not be added for another 21 years.
The meningioma warning timeline accelerated rapidly between 2023 and 2025:
The plaintiff failure-to-warn theory anchors on the gap between Pfizer's claimed 2023 awareness and the December 2025 warning, and — more aggressively — on plaintiff allegations (not yet established as historical fact) that Pfizer's predecessor (Upjohn) was on notice of progesterone-receptor expression in meningiomas as early as 1983 (HMF Law, secondary; the operative court-filing language and underlying historical record should be confirmed against MDL 3140 docket entries before relying on the 1983 date in pleading). On that allegation-stated view, 33 years would have elapsed between the 1992 approval and the 2025 warning.
The FDA's December 2025 reversal is litigation-load-bearing in two ways: it forecloses Pfizer's preemption argument on the label-change side, and it constitutes the FDA — applying its own scientific standards — concluding the evidence supports a warning. A Daubert argument that the underlying science is unreliable runs into the contradictory fact that Pfizer itself successfully petitioned the FDA to adopt that science as the basis for label revision.
The single most important international analogy is France's response to cyproterone acetate (CPA), a structurally related progestin marketed for hyperandrogenism and other indications. The French sequence:
The CPA precedent matters because the structural and pharmacological similarities to DMPA are substantial: both are synthetic progestins with high PR-binding affinity, both carry meningioma risk in published epidemiology, and both have shown a dose-duration gradient. France acted on its pharmacovigilance signal within months in 2018; the FDA waited 33 years after initial approval and roughly 21 months after BMJ publication of Roland et al. (2024) to act on DMPA. The analogy is plaintiff-favorable on both the science and the regulatory-response side.
| Compound | EMA action | Year | Risk level documented |
|---|---|---|---|
| Cyproterone acetate (≥10 mg) | Restricted to last-line; MRI screening; meningioma contraindication | 2020 | 7× overall, up to ~21× at high cumulative dose |
| Nomegestrol acetate | "Lowest dose, shortest duration" recommendation | 2022 | Elevated |
| Chlormadinone acetate | "Lowest dose, shortest duration" recommendation | 2022 | Elevated |
| MPA (high-dose) | Added meningioma as "possible side effect" | 2024 | 2.4–5.6× |
Health Canada updated the DMPA label to include meningioma in "warnings and precautions" in 2024 — preceding the FDA's December 2025 action.
Despite DMPA being used by approximately 74 million women annually worldwide — vastly more than CPA's user base — no country has imposed restrictions analogous to France's CPA measures. No country mandates pre- or during-use MRI screening for DMPA users. The global attributable-burden estimate is 2,800 to 18,000+ symptomatic or surgical meningiomas annually (Frontiers, 2025). The Frontiers in Global Women's Health 2025 editorial characterizes this gap as "a legitimate public health concern."
| Item | Detail |
|---|---|
| MDL number | 3140 |
| Full title | In re: Depo-Provera (Medroxyprogesterone Acetate) Products Liability Litigation |
| Court | U.S. District Court, Northern District of Florida (Pensacola Division) |
| Judge | Hon. M. Casey Rodgers |
| MDL formation | February 2025 |
| Approximate pending cases (MDL only) | ~3,490 as of April 1, 2026 (Lawsuit Information Center, secondary) |
| Approximate total claim count (federal + state) | 4,000+ |
| Approximate intake rate | ~650 new filings/month |
| Defendant | Pfizer Inc. |
(Verify all case-count figures against the Northern District of Florida MDL portal and the JPML pending-MDL-dockets report before relying on them in motion practice; these are fast-moving snapshot values.)
| Date | Event |
|---|---|
| May 15, 2026 | Case management conference |
| June 24–26, 2026 | Daubert hearings — general causation (rescheduled from May 26–28, 2026) |
| December 7, 2026 | First bellwether trial: Toney v. Pfizer |
Five pilot cases have been selected: Toney, Blonski, Schmidt, Wilson, and Arceo (Helbock Law, MDL tracker, secondary — confirm against the operative case-management order). Toney is set as the first bellwether on December 7, 2026. Judge Rodgers has signaled that preemption and Daubert determinations in MDL 3140 will bind all cases (MDL Update, April 2026 — secondary).
| Role | Attorney | Firm |
|---|---|---|
| Lead Counsel | Christopher A. Seeger | Seeger Weiss LLP |
| Co-Lead Counsel | Bryan F. Aylstock | Aylstock, Witkin, Kreis & Overholtz |
| Co-Lead Counsel | Ellen Relkin | Weitz & Luxenberg |
| PEC Co-Chair | Tracy Finken Magnotta | Anapol Weiss |
| PEC Co-Chair | Virginia Buchanan | — |
| PEC Members | Chris Hellums; Buffy Martines; Troy Rafferty; Michael Sacchet; Tara Sutton | — |
Tracy Finken Magnotta was reappointed as Co-Chair of the Plaintiffs' Executive Committee on March 20, 2026 by Pretrial Order No. 31 (case caption: In Re: Depo-Provera (Depot Medroxyprogesterone Acetate) Products Liability Litigation, Case No. 3:25-md-3140, N.D. Fla.).
| Severity tier | Projected per-case range |
|---|---|
| High (Grade III, malignant) | $650,000 – $1,500,000+ |
| Moderate (Grade II, symptomatic) | $300,000 – $650,000 |
| Lower (Grade I, monitored) | $150,000 – $300,000 |
| Minimal (asymptomatic) | Under $150,000 |
These are pre-bellwether projections (Lawsuit Information Center, April 2026 — secondary) and should be treated as illustrative only. No verdicts or aggregate settlements have been entered. Projections will move materially based on the June 2026 Daubert outcome and the December 2026 bellwether result. Note that the 23% of Abou-Al-Shaar resected tumors classified as WHO Grade II would map to the "Moderate" tier in this stratification.
The plaintiff record is strong on general causation. The litigation gaps are operational — exposure documentation, specific causation, statute-of-limitations briefing, and a small set of evidentiary uncertainties on the re-verification backlog.
The dose-response gradient depends on the plaintiff being able to document her cumulative DMPA exposure with sufficient precision to place her on a defined point of the curve. Plaintiff intake records vary in completeness — pharmacy, billing, and medical records do not always agree, and gaps in coverage (insurance changes, clinic transitions, self-pay periods) are common. Adopt a uniform exposure-documentation protocol across firms: reconstruct the injection sequence from all available pharmacy/dispensing records, Medicaid or commercial-insurance claims (CPT 11981, J1050), clinic records, and patient-maintained records. Document gaps explicitly rather than inferring exposure. Expert reports should state cumulative exposure with a confidence range bracketed by the documentation, and place the plaintiff on the Griffin Table-4 cerebral curve (≤1y / >1–2y / >2–3y / >3y) as well as the JAMA Neurology >4-year and >6-year breakpoints.
The strongest specific-causation cases will be PR-positive, ideally PIK3CA-mutated, ideally skull-base, ideally with documented post-cessation regression. For atypical cases — non-skull-base, low-PR, NF2-mutated — the theory must be developed more carefully. For each plaintiff, obtain pathology records and (where feasible) molecular characterization of the tumor; where unavailable from original surgical pathology, consider commissioning supplemental molecular profiling on archived tissue. PR-status, PIK3CA-status, NF2-status, and chromosome-22 status are the four most relevant data points. Case-acceptance posture and damages valuation should reflect this gradient.
Develop a uniform plaintiff brief on the legal irrelevance of the promotion-vs-initiation distinction for product-liability causation, anchored on the substantial-factor test in the operative jurisdiction. Where the jurisdiction has accepted promotion-as-causation in other tort contexts (asbestos, occupational chemical exposure), cite those decisions. Couple with the Peyre PIK3CA evidence — supporting an initiation theory in addition to (not instead of) promotion. The jury can be told both: DMPA promotes subclinical lesions to clinical disease, AND the PIK3CA pattern is consistent with progestin-driven mutational events.
Plaintiffs whose DMPA exposure preceded the litigation by many years face statute-of-limitations exposure in jurisdictions without a robust discovery rule. The available anchors are the Roland BMJ publication (March 2024), the Griffin Cancers publication (October 2024), the JAMA Neurology publication (September 2025), and the FDA December 2025 label change. The strongest anchor for older plaintiffs is the FDA December 2025 label change — the date at which a reasonable plaintiff could be charged with knowledge under a public-notice theory. Develop a uniform statute-of-limitations brief and state-by-state matrix mapping discovery-rule recognition, public-notice anchors, and any state toleration of latent-injury or continuing-tort theories. The operative anchor date should be identified at intake, not litigated reactively.
Several elements of the plaintiff narrative rest on secondary sources or on materials that should be primary-source-confirmed before deposition or trial:
Build a re-verification schedule keyed to the June 2026 Daubert hearing. Each load-bearing claim resting on a secondary source must be re-verified or removed from the load-bearing position before that hearing.
The 2025 Expert Opinion on Drug Safety "comparative safety study" (Source Index entry 7) is the published nucleus of the defense comparative-safety argument. Commission a methodological-response analysis focused on comparator selection, exposure-window definition, the adjustment set, and publication-and-funding disclosures. Build this into the plaintiff Daubert briefing rather than treating it as a cross-examination point alone.
Pre-bellwether settlement projections rest on injury-severity tiers, but the operative damages models should be patient-specific, not population-aggregate. SEER-derived survival aggregates are vulnerable to defense challenge in cases with favorable individual prognosis. Replace aggregate models with stage-specific, treatment-specific, and demographic-specific ones. Account for surgical complications, cognitive sequelae, recurrence-monitoring lifetime cost, and employment-reentry difficulty. Craniotomy with skull-base reconstruction generates distinct damages from tumor resection alone. The Abou-Al-Shaar 23% Grade II fraction is informative for the population-level mix but not dispositive for individual cases.
Six concrete priorities:
This memorandum is prepared as background research and does not constitute legal advice. All citations have been verified to publicly accessible URLs as of the date of this memorandum. Recipients should independently verify case names, docket numbers, settlement projections, and current MDL status before relying on this material in court. Where the manuscript identifies a specific snapshot value (case counts, settlement projections, regulatory deadlines), the snapshot date is identified inline; counts and deadlines drift, and the reader should confirm current status before citing. The re-verification backlog noted in §8.5 should be cleared before any of the affected claims are deployed at deposition or at the June 2026 Daubert hearings.
We produce custom litigation research memos — cross-domain synthesis connecting medical science, regulatory filings, expert witness analysis, and case law for specific MDLs. Each memo is adversarial-reviewed before delivery.
$1,500 – $7,500 per memo depending on scope and docket complexity.
Contact: alex@vibeagentmaking.com