We talk about preemption so often that we can be hesitant to post on preemption cases unless we can see that they offer something new or different. (Except for Bexis. Preemption is like his morning, afternoon, and evening coffee.) When we saw that there was a generic preemption decision in the Yaz litigation, we were intrigued. First of all, we did not know there was even generic drospirenone. Second, we have posted on several decisions from that litigation, but never on preemption. Given our assumption that all the drospirenone used by the plaintiffs was branded and our basic knowledge of the post-approval labeling history of the drug, the lack of preemption challenges should not have been too challenging. Third, given the tenor of some of the decisions from the litigation and the pretty clear law, from our perspective, barring product liability claims based on the use of generic drugs, we wondered how the decision would play out.
So we read the decision in In re: Yasmin and Yaz (Drospirenone) Marketing, Sales Practices and Products Liability Litigation (Gannon), No. 3:13-cv-10143-DRH-PMF, 2014 U.S. Dist. LEXIS 56862 (S.D. Ill. Apr. 24, 2014), with at least a little hope that the generic manufacturer’s motion for judgment on the pleadings would get a fair shake. It started well enough. Plaintiff was taking a generic version of drospirenone called Gianvi at the time of her pulmonary embolus and there were complicated and competing versions of the history of the product that raised questions as to what entity manufactured the drug she took. The court cut through the competing allegations—well pleaded allegations get credited in this posture—and saw that plaintiff sought to impose liability on the movant based on its role as a generic distributor. Id. at *8. Then, the court walked through the basic holdings of Levine, Mensing and Bartlett. As we are wont to do, we can quibble with some details of how those cases were presented, but the court was right that, for generic drugs, Mensing held all failure to warn claims preempted and Bartlett held that design defect claims preempted when they were “warning-based” or based on ineffectiveness or unreasonable dangerousness. Id. at **14-21. Along the way, the decision even brushed aside the claim that preemption for generic drugs is somehow different after the Food and Drug Administration Amendment Act of 2007. Id. at *17 n.5.
Then the decision held that the rationales of Mensing and Bartlett for preemption for claims against generic manufacturers applied equally to claims against generic distributors. Id. at **20-21. And the decision rejected that the generic distributor’s agreement with the branded manufacturer created joint liability and changed the preemption analysis. Despite the agreement, “Teva had no authority to make unilateral changes to Gianvi’s label. At most, Teva could have sought assistance from the FDA in convincing Bayer to adopt a stronger label for Yaz so it could do the same with regard to Gianvi. The same is true with regard to Teva’s ability to alter Gianvi’s design or composition.” Id. at **22-23. This is good and even implicitly recognizes preemption principles from Bartlett we expounded on here.
Then the decision seems to start with the horse before the cart in looking at the scope of applicable Illinois design defect law before deciding if plaintiff’s design defect claims were preempted. We have criticized a few decisions in the past for figuring out ways that a hypothetical claim might be parallel to federal requirements before looking at whether such a claim could be asserted within the bounds of state law. Plaintiff argued that Illinois law was different than the New Hampshire law at issue in Bartlett in that it did not place a duty on a manufacturer to sell a same product based upon its design and labeling. Analyzing Illinois case law and Bartlett, the court found that Illinois imposed strict liability for selling a drug that was not “reasonably safe,” which was what New Hampshire did. In other words, Illinois has a fairly traditional strict liability design defect claim and it should be preempted like the New Hampshire strict liability design defect claim was in Bartlett. Id. at ** 23-28. The court continued on this path in considering the argument that the option under Illinois law to use either the consumer expectation test or risk-utility test to establish that a product was unreasonably dangerous negated Bartlett preemption. The court rejected the argument based on the Drager and Fullington (on remand) decisions. Id. at **29-31. The court specifically endorses the Drager reasoning that:
PLIVA cannot be required to stop selling its product, but at the same time it is prohibited from making any changes to the product itself or the accompanying warnings. Regardless of the way in which Maryland assesses the unreasonableness of a product’s risks, if PLIVA’s metoclopramide is unreasonably unsafe, there is no apparent action that PLIVA can take in compliance with FDCA restrictions to avoid strict liability.
Id. at **30-31. All good so far.
Then three paragraphs undo everything. Like the unsatisfying ending to a TV series—Dexter comes to mind and some may say True Detective fits too—your opinion can shift right as you are expecting an opinion/show to wrap up nicely. Plaintiff’s last argument for avoiding preemption was based on an “exception” from a footnote in Bartlett. Except the footnote was just dicta—saying what that decision was not addressing—and did not recognize an exception. The Yaz court glossed over what the Supreme Court actually held in Bartlett and ignored what it had already recognized about Illinois design defect claims. This allowed the plaintiff to plead a parallel, non-preempted claim by saying her allegations tracked the purportedly endorsed “exception” from Bartlett. Confused? So were we.
Rather than characterize them, here are the three paragraphs, revealing all the analysis that went into this great leap:
In Bartlett, the Supreme Court expressly noted an exception for state law claims that parallel the federal misbranding statute. Bartlett 133 S.Ct. 2477 n.4 (“We do not address state design-defect claims that parallel the federal misbranding statute.”). The federal statute requires a manufacturer to pull a drug from the market (even though approved by the FDA) if it is “dangerous to health” even when used in accordance with the FDA-approved directions. Id.; cf. 21 U.S.C. § 352(j). This exception only applies where the plaintiffs claim is based on scientific information that was not available when the drug was approved by the FDA. Bartlett, 131 S.Ct. 2477 n.4.
In the instant case, the plaintiff alleges that “the Gianvi birth control pills, as manufactured, designed, sold, supplied and introduced into the stream of commerce by Defendants, were dangerous to health when used in the dosage, manner, or with the frequency or duration prescribed, recommended and/or suggested in its labeling, in violation of 410 ILCS § 620/3, et seq. As such, the sale of such drugs in this state was strictly prohibited” (Doc. 20 ¶ 134). The plaintiff also sets forth allegations regarding Gianvi’s allegedly dangerous design that purportedly was not available to the FDA (Doc. 20 ¶¶ 38-48). The plaintiff contends that “unlike in Bartlett, the jury will be asked to find whether new evidence concerning Gianvi that had not been made available to the FDA rendered Gianvi so dangerous as to be misbranded under the federal misbranding statute” (Doc. 28 p. 15).
Considering this argument, the Court finds that, to the extent the plaintiff’s design defect claim parallels the federal misbranding statute, it is not foreclosed by Bartlett.
Id. at **31-32.
This is all premised on a footnote in Bartlett that starts off with a sentence that makes it clear that the rest of the language in the footnote is just dicta: “We do not address state design-defect claims that parallel the federal misbranding statute.” It also ends with “Because the jury was not asked to find whether new evidence concerning sulindac that had not been made available to the FDA rendered sulindac so dangerous as to be misbranded under the federal misbranding statute, the misbranding provision is not applicable here,” which reinforces that no exception to preemption was being articulated. The dicta does say that “The misbranding statute requires a manufacturer to pull even an FDA-approved drug from the market when it is ‘dangerous to health’ even if ‘used in the dosage or manner, or with the frequency or duration prescribed, recommended, or suggested in the labeling thereof,’” but the cited portion of the misbranding statute only defines a situation when an approved drug can be “misbranded”—it does not require withdrawal of the drug. The rest of the footnote did not create an exception, either: “The parties and the Government appear to agree that a drug is misbranded under federal law only when liability is based on new and scientifically significant information that was not before the FDA,” there is nothing like a definition of an exception to the broad preemption holding in Bartlett. Indeed, the footnote itself was attached to a sentence saying “New Hampshire’s warning-based design-defect cause of action is pre-empted with respect to FDA-approved drugs sold in interstate commerce.” So, at best, it was talking about a possible exception to preemption for “warning-based design defect” claims, which were not at issue in the Yaz decision.
Which brings us back to the basic proposition that you need to identify a cognizable state law claim before you determine if it is preempted. For a court sitting in diversity, there certainly needs to be reluctance to extending state law just to get around preemption. So, what claim was plaintiff asserting here that might be subject to the purposed exception to preemption? The quoted paragraphs are part of section of the decision called “Count II Strict Products Liability – Design Defect.” As presented earlier, Illinois design defect was design-based, not warning-based. But the Bartlett language was about warning-based design defect and the referenced allegations from the plaintiff’s amended complaint seem to deal with the dangerousness of the product with and without reference to its labeling. And there is reference to misbranding under the federal FDCA and the state FDCA—410 ILCS § 620/3 refers to part of the Illinois FDCA that prohibits acts like selling misbranded drugs and §620/15(j) mimics the federal misbranding statute from the Bartlett footnote. But there is no private right of action under the federal FDCA, as recognized by the Illinois Supreme Court in Martin v. Ortho Pharmaceutical Corp., 661 N.E.2d 352, 355 (Ill. 1996), and there does not seem to be one under the Illinois FDCA either—certainly not something that could be assumed. So, what cause of action is being asserted here? A brand new cause of action for design defect failure to withdraw a misbranded drug? To quote Marty Huggins from The Campaign, “It’s a mess.” And that is even without considering Buckman preemption for whatever claim this was that fit within the made up exception to Bartlett preemption.
We do not know if the Third Circuit decision in Fosamax discussed yesterday directly addressed Illinois design defect law when it found the design defect claims under the law of 28 states were preempted under Bartlett. Regardless of whether it did, it is clear that the Yaz decision is an outlier. There is nothing particularly special about Illinois design defect law that would justify claims predicated on a generic drug being unreasonably dangerous surviving a vigorous preemption challenge. We expect other plaintiffs will try to make the same arguments about their claims. We hope other courts will do a better job with them.