390 F. Supp. 3d 1102
N.D. Cal.2018Background
- Multidistrict litigation centralized hundreds of federal suits alleging Monsanto's glyphosate-based herbicides (Roundup) cause Non‑Hodgkin’s Lymphoma (NHL); court bifurcated proceedings into general causation and specific causation phases.
- At the general causation phase plaintiffs must present at least one admissible expert opinion showing glyphosate can cause NHL at human‑relevant exposure levels; otherwise summary judgment for Monsanto follows.
- Plaintiffs primarily relied on (1) IARC’s 2015 “probably carcinogenic” hazard classification, (2) epidemiological studies (case‑control studies, pooled analyses, meta‑analyses), (3) Animal carcinogenicity bioassays, and (4) mechanistic (genotoxicity/oxidative stress) data.
- Monsanto emphasized the large Agricultural Health Study (AHS) cohort (Andreotti 2018) showing no association, and attacked recall bias, confounding (other pesticides), latency, exposure misclassification, and some experts’ methods.
- The court applied Rule 702/Daubert gatekeeping: admissibility depends on qualification, relevance (fit), and reliable methodology; district courts in Ninth Circuit give a ‘‘liberal thrust’’ favoring admission while screening out unreliable ‘‘junk science.”
Issues
| Issue | Plaintiff's Argument | Defendant's Argument | Held |
|---|---|---|---|
| Whether IARC classification alone satisfies general‑causation (human‑relevant dose) | IARC’s ‘‘probably carcinogenic’’ finding shows glyphosate is a credible carcinogenic hazard relevant to causation | IARC is a hazard identification, not a risk assessment; its conclusion doesn’t establish causation at human‑relevant exposures for civil standard | IARC hazard finding is insufficient by itself; experts must go beyond IARC to address causation at realistic human exposures |
| Admissibility of plaintiffs’ experts (Daubert/Rule 702) | Experts (Portier, Ritz, Weisenburger et al.) used Bradford‑Hill synthesis of epidemiology, animal, and mechanistic data to conclude glyphosate can cause NHL | Monsanto: experts’ methods/results are unreliable, results‑driven, suffer from logical gaps, misuse of statistics (e.g., pooling rodents), overreliance on flawed studies | Court admitted three core experts (Portier, Ritz, Weisenburger) as sufficiently reliable (with limitations); excluded some others (Neugut, Nabhan) and disallowed Portier’s novel pooled animal‑analysis method |
| Weight to give competing epidemiology (case‑control/meta‑analyses vs. AHS cohort) | Case‑control studies and meta‑analyses show modest associations and pattern of ORs >1; cohort (AHS) may suffer exposure misclassification and imputation bias | AHS (Andreotti 2018) is largest, prospective, avoids recall bias, shows no association; case‑control studies suffer recall bias, confounding, short latency | Court: epidemiologic evidence is mixed but debatable; reasonable experts may prefer case‑control/meta‑analytic signal over AHS given methodological disputes; that choice does not make expert inadmissible |
| Role of animal and mechanistic data for admissibility/fit | Animal bioassays and mechanistic (genotoxicity/oxidative stress) studies support biological plausibility and coherence in Bradford‑Hill analysis | Monsanto: animal data involve high doses and require careful extrapolation; mechanistic studies are inconsistent or flawed | Court: animal and mechanistic evidence are relevant to biological plausibility and admissible; expert analyses of these data are permissible so long as methodology is reliable (Portier’s pooling method excluded) |
Key Cases Cited
- Anderson v. Liberty Lobby, Inc., 477 U.S. 242 (summary judgment standard governs whether a reasonable jury could find by preponderance)
- Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (federal trial judge acts as gatekeeper for expert testimony under Rule 702)
- Kumho Tire Co. v. Carmichael, 526 U.S. 137 (Daubert gatekeeping applies to all expert testimony, not only pure‑science experts)
- General Electric Co. v. Joiner, 522 U.S. 136 (court may exclude expert testimony when there is too great an analytical gap between data and opinion)
- Messick v. Novartis Pharm. Corp., 747 F.3d 1193 (Ninth Circuit emphasizes liberal thrust favoring admission of expert testimony)