It’s written somewhat strangely – weak at the beginning, strong at the end – but the Department of Justice’s amicus curiae brief , filed yesterday in Mutual Pharmaceutical Co. v. Bartlett, No. 12-142 (U.S. pending), eventually gets to the right place. That is, even the generally anti-preemption political leadership of this FDA and DoJ took the only institutionally logical position that they could take and told the Supreme Court that, yes, state tort claims asserting that an FDA-approved product should have been removed from the market altogether were impliedly preempted.
Because we like good news (and still can use some after Stengel and Weeks), we flip right to the back of the DoJ’s Bartlett brief. The last substantive paragraph states:
Federal law would preempt a pure defective-drug-design claim that required a jury to second-guess FDA’s safety determination, without any further need to find the existence of new and scientifically significant evidence that rendered the product misbranded under federal law.
DoJ Bartlett br. at 34.
Let’s look at this. First of all, what’s this “pure” design defect business? As we said above, the DoJ brief starts out weakly, trying to turn Bartlett into something it wasn’t in order to avoid what it was. Instead of what really went on at trial (the trial court allowing plaintiffs’ counsel to ask the jury to flip a bird at the FDA), the DoJ first portrayed the verdict as based primarily on an “unreasonably dangerous even with warnings” theory under Restatement (Second) of Torts §402A, comment k (1965). DoJ br. at 14-18. The primary purpose of that argument was to reduce Bartlett to a “nothing to see here” misapplication of Mensing.
But it wasn’t, and the Court wouldn’t have taken such a case. Rather, the trial court construed New Hampshire law as allowing a design defect claim based on pure risk utility factors without any need for an alternative design, and the First Circuit converted that theory on appeal to the “remove-from-the-market” claim that the Supreme Court actually granted certiorari to review. So eventually, DoJ had to come to grips with what is actually at issue: that the trial court told the jury it could − and plaintiff’s counsel told the jury it should − find that the FDA should not have approved a drug that it did approve:
[The District Court] instructed the jury that it . . . was free to give FDA’s approval “as much or as little weight as you think it deserves. . . .” [Plaintiff’s counsel] “argued that sulindac was a “needless and useless drug. . . .” Counsel asserted that FDA merely “rubber stamped” the ANDA . . ., and that “no evidence (showed) that FDA has ever done a focused analysis on (s)ulindac” to justify its “be(ing) on the market.” FDA, counsel asserted, lacks “expertise in risk⁄benefit assessment” and “impos(es) a significant risk to . . . the safety of the public” because it is unable to perform its “mission to monitor drug safety for any class of drugs.”
DoJ br. at 10-11 (quoting trial record). Given this view of the law and the evidence, it’s hardly surprising that the Bartlett jury, applying only state law, told the FDA to pound sand and decided that an FDA-approved drug simply shouldn’t be sold at all. That’s the “pure” theory that DoJ addresses in the most interesting part of its brief.
As the first block quote above indicates, the FDA supports preemption – and does so without any use of “generic” as an adjective. That's important. The FDA states that remove-from-the-market claims are preempted as to all drugs (and presumably all other products it regulates):
Brand-name and generic drugs should be treated the same for purposes of design-defect claims. . . . [T]he active ingredient in a generic drug must be the same as that in the brand-name drug. The generic manufacturer may not alter that ingredient without prior FDA approval. But the same is true of the manufacturer of a brand-name drug. for both, any change that created a new active ingredient would require prior FDA approval. . . . [P]reemption in the generic-drug context, just as in the brand-name-drug context, should not be defeated on the rationale that the manufacturer could always comply with state law by declining to provide the very drug whose availability Congress sought to provide.
DoJ br. at 30-31 (citations and footnote omitted) (emphasis added). We’ve received the same bone-headed response from an appellate court in a branded case, see Wimbush v. Wyeth, 619 F.3d 632, 645 (6th Cir. 2010) (our #1 worst caseof 2010), so we know that allowing state juries to order federally approved drugs off the market is just as preemptive a conflict (and just as absurd a theory) in branded as in generic cases.
We note that while the DoJ to some extent addresses a straw man – the form of design defect described in Restatement (Third) of Torts, Products Liability §6(c) (1998), which it concedes has not been accepted to “any significant degree” (DoJ br. at 20) – this reasoning is also relevant to the even more extreme so-called “categorical liability” design defect claims that the Third Statement rejected, which are precisely what the plaintiff was permitted to pursue in Bartlett. Since none of these now mostly discredited meat-axe design-related theories includes an caveat approaching the “new and scientifically significant evidence that rendered the product misbranded under federal law” qualifier added by DoJ, the government’s position is effectively that all remove-from-the-market theories asserted in tort cases against FDA-approved products are preempted.
That’s as it should be.
We particularly recommend DoJ’s discussion of why failure to withdraw claims inherently conflict and interfere with the FDA’s parallel post-marketing processes – because plaintiffs (including in Bartlett) have made contrafactual statements about the FDA’s post-marketing authority, leading some courts into error as to “pharmacovigilence”:
Congress has further charged FDA with monitoring post-marketing drug safety. A manufacturer must maintain extensive clinical records and make numerous reports to FDA. . . . Those duties apply to manufacturers of generic drugs as well as the brand-name drug. . . .
The Act provides that FDA shall withdraw approval of a drug if, inter alia, it finds that the drug is not safe for the uses identified at the time of the drug’s approval or is not effective as claimed for those uses. Approval may be withdrawn only following procedures that afford the manufacturer due process and the opportunity for a hearing. Since 2007, FDA may also require (not merely request) labeling changes.
FDA’s ongoing risk-benefit analysis will sometimes take into account the availability of more effective or less risky alternatives . . . . FDA did not request withdrawal of sulindac. . . .
In the face of this elaborate regulatory regime instituted to safeguard the national market and protect consumers throughout the United States, and the extensive commitment of public and private resources to those ends, it would be inconsistent with the FDCA to conclude that a manufacturer must abandon a market it has been approved by FDA to enter in order to avoid violating a duty recognized by a jury under state tort law that deems its product unsafe.
By requiring a jury independently to balance the health risks and benefits of FDA-approved uses of a drug and to determine if the drug is “unreasonably dangerous” for those uses, a state with a pure design-defect product-liability law would force the jury to second-guess FDA’s safety determination, which balances the drug’s therapeutic risks and benefits for its labeled uses. Such ad-hoc reconsiderations on a state-by-state and lawsuit-by-lawsuit basis would undermine FDA’s drug-safety determinations, which are made based on sound scientific judgments by an expert federal agency with appropriate access to pertinent safety data, and the assurance that FDA’s approval provides for all participants in the market.
DoJ br. at 26-28 (various stuff omitted).
DoJ also distinguished the heck out of Wyeth v. Levine, 555 U.S. 555 (2009), which is probably the worst prescription drug decision of all time (at least as to preemption):
- Levine didn’t adopt any “general no-preemption rule.” Rather, Levine (like any other case) depends on “context.” DoJ br. at 31.
- Levine involved “newly acquired information undermining its drug’s safety” that the defendant had “failed to strengthen its labeling as specifically contemplated by FDA’s CBE regulation. Id.
- Levine found no obstacle preemption in “FDA’s labeling approvals, and the mere possibility that FDA would disapprove a manufacturer’s subsequent enhanced warning.” Id. at 32.
- Levine “does not categorically extend to all claims,” and in particular does not extend to design claims, because “because a manufacturer cannot unilaterally alter the design (unlike the labeling) of a drug.” Id.
- Unlike the FDA’s position in Levine, FDA drug approval as a “an expert judgment that the drug’s therapeutic benefits outweigh its risks” is a “longstanding interpretation” of the FDCA. Id. at 32-33.
- Levine involved “traditional” warning-based litigation, whereas design-based claims have been “rare” and “unusual.” Id. at 33.
- The claims at issue in Bartlett fit within Levine’s “recogni[tion] that some state-law claims might well frustrate the achievement of congressional objectives.” Id. (quoting Levine, 555 U.S. at 581).
So the good news is that, after some hemming and hawing, the DOJ did the right thing in Bartlett.
The bad news is the hemming and hawing. Whether a state-law jury should be allowed to say “no” to the marketing of a product after the FDA has said “yes” is hardly “difficult and close.” DoJ br. at 12. This type of absolute and utter conflict between state and federal law is a no-brainer, since 1913, when the Court decided in McDermott v. Wisconsin, 228 U.S. 115 (1913), that a state couldn’t ban FDA-approved dairy products.
Moreover, it was egregious and unnecessary for DoJ to advocate exceptions for Levine branded labeling claims or Riegel parallel claims, br. at 23-24, since as the brief admits, “neither of the foregoing theories is available here.” Id. at 24. Even worse is DoJ’s backhanded endorsement of a “state-law duty not to market the drug in the same circumstances,” parallel to federal “misbranding” law as not preempted. Id. Not only was (1) no such claim raised in Bartlett, and (2) “parallel” claims analysis isn’t an implied preemption concept, but generally there is no state-law cause of action for failure to recall, as we discussed here and here, among other places. We hope the Supreme Court won’t have time for such irrelevant nuances when it decides Bartlett, but anything that gives comfort to bizarre new tort theories we can’t stomach.