We’ve been aware of the other side attempting to construct a First Amendment counter-argument to our preemption defense for some time. It first cropped up in a 2015 360 article by a P-side thought leader (no, not necessarily an oxymoron) soon echoed by a comment to one of our posts on the Amarin First Amendment decision. DDL, and we addressed it here. Late the same year, we addressed it again, this time in response to an ATLA article, and set forth our views in more detail.

In a nutshell, the plaintiffs’ position is that, assuming truthful promotional speech about drugs is constitutionally protected (as we agree), our clients purportedly have a state-law tort “duty” to ignore the FDA and put on our labels whatever any particular plaintiff claims is “truthful” in any given case, even if we would otherwise be required to obtain FDA pre-approval. The First Amendment’s protection, these articles have argued, dissolves impossibility preemption because the First Amendment eliminates any FDA “prior restraint” on protected speech.

This argument has some rather glaring flaws, starting with the other side’s necessary concession that the speech in question is constitutionally protected, which in and of itself would preclude state-law tort liability under New York Times Co. v. Sullivan, 376 U.S. 254 (1964). Further, an FDA pre-approval requirement, designed to make sure that only truthful speech appears on drug labeling (which is quite broadly defined), would seem to be a reasonable time, place and manner requirement – given its public health-based rationale.

That was where things stood the end of 2015. Not much happened (from our perspective) for almost two years. We thought the primary target of any argument the First Amendment defeats impossibility preemption would be generic preemption. After all, generic plaintiffs don’t have much more to lose, so why not give this a flyer? But, nothing. Nor did we see the argument being made anywhere else − until a few weeks ago.

In their infinite wisdom, the other side decided to debut their First Amendment argument against preemption in the Depakote litigation. It first raised its ugly head in Swanson v. Abbott Laboratories, 2017 WL 5903362 (S.D. Ohio Nov. 28, 2017). Previous rulings in Depakote cases had found preemption of certain warning claims under the Levine “clear evidence” standard, and also preempted design defect claims under the Mensing/Bartlett impossibility rationale. See Rheinfrank v. Abbott Laboratories, Inc., 680 F. Appx. 369, 385-88 (6th Cir. 2017); Rheinfrank v. Abbott Laboratories, Inc., 137 F. Supp.3d 1035, 1040-41 (S.D. Ohio 2015); Rheinfrank v. Abbott Laboratories, Inc., 119 F. Supp.3d 749, 766-70 (S.D. Ohio 2015), aff’d, 680 Fed. Appx. 369 (6th Cir. 2017); In re Depakote, 87 F. Supp.3d 916, 922-23 (S.D. Ill. 2015).

The same preemption arguments were made – and prevailed – supporting summary judgment in Swanson. 2017 WL 5903362, at *7-8. The new wrinkle was plaintiffs arguing “that Defendants have not met their burden in proving impossibility because the First Amendment protects their right to communicate non-FDA approved information through other means.” Id. at *8. The court found plaintiffs’ reliance on off-label promotion cases “irrelevant” to a product liability case involving no such thing:

Such cases, however, are not applicable in the instant matter. . . . [T]he ability to promote off-label uses is irrelevant as to whether the FDA’s rejection of a labeling change for an on-label use constitutes clear evidence of conflict with state law.

Id. “Labeling” includes Dear Doctor letters. Id. Therefore, “[t]o hold that Defendants First Amendment right prevents preemption is contrary to case law finding preemption based on FDA regulations.” Id. (pointing out that Mensing and Bartlett also involved FDA-approved labeling).

The same defendant also obtained summary judgment in Willis v. Abbott Laboratories, 2017 WL 5988215 (W.D. Ky. Dec. 1, 2017). Again, preemption was a big part of the win. The defendant’s showing of “clear evidence” in Willis was no different than that which had prevailed both in the district court and on appeal in Rheinfrank. Willis, 2017 WL 5988215, at *4. Plaintiff threw the First Amendment hail Mary – arguing that, “because [defendant] has a First Amendment right to speak in a truthful and non-misleading manner about Depakote, it cannot claim that the FDA’s rejection of its label changes made it ‘impossible’ for it to comply with” state law. Id. at *5. Again, the argument fell woefully short:

The plaintiffs cite to no standard the Court should look to in evaluating whether [defendant] has a First Amendment right to include a warning about developmental delay in the Depakote label or whether the FDA’s rules and practices regarding drug labeling unconstitutionally infringe upon that right. . . . Instead, they cite to a line of cases which pertain exclusively to the marketing, as opposed to labeling, of drugs and seek to apply them to the present case. . . . The Court finds each of these cases, as well as the other out-of-circuit district court cases cited, to be distinguishable and unpersuasive. The plaintiffs essentially ask the Court to find that the FDA’s authority to reject proposed label changes regarding drug safety amounts to an unconstitutional suppression of a drug manufacturer’s speech. None of the cases cited by the plaintiffs support such a broad proposition, and understandably so, given the recognition courts have long given to the role of the FDA in regulating the labeling of drugs.

Id. (citations omitted). Likewise, the FDA’s control of “labeling” extended to the other means (such as “Dear Doctor” letters) that plaintiffs claimed the defendant should have employed. Id. The First Amendment simply didn’t force the defendant to speak in the manner that plaintiffs demanded.

“Comparatively,” the Supreme Court preemption precedent on which the defendant relied (chiefly Mensing/Bartlett) “spoke directly on the FDA’s power to regulate what speech appears in a drug’s ‘label’ and when that power takes preemptive effect over what is required by state tort law.” Id. at *6. Thus, the application of the First Amendment to FDA administrative actions was not analogous to preemption in product liability litigation:

The plaintiffs’ argument would require the Court to cast aside both cases and the scheme of preemption that they endorse. The Court is not persuaded that the First Amendment requires it to do so. Therefore, the Court rejects the plaintiffs’ First Amendment argument.

Id.

We find it somewhat odd to see the plaintiffs’ First Amendment argument trotted out for the first time in a branded drug preemption case. Compared to generic plaintiffs, those in Depakote aren’t nearly as devoid of alternative arguments. In any event, our dismissive opinion of that argument is only reinforced by these decisions. First Amendment principles are implicated by absolute FDA bans applicable to drug marketing. Free speech is not impinged in nearly the same way by the FDA having a say over what goes into drug labeling. Indeed, to the extent that plaintiffs invoke the First Amendment to pursue some state-law duty that would force the defendants to speak by changing their labels, they would be standing free speech on its head.

There is more to Swanson and Willis than just their preemption rulings, however. In Swanson, plaintiffs tried nitpicking which alleged birth defect risks extended past the first trimester, as opposed to which did not. That argument was quickly shut down. “Plaintiffs’ claim that [defendant] failed to warn that Depakote use’s risk continued during pregnancy beyond the first trimester is too intertwined with the [preempted] developmental delay claim to remain.” 2017 WL 5903362, at *10. As far as other products being “safer,” they simply weren’t an option in Swanson:

[I]n 1996, Plaintiff did not have other bipolar medication options that would have worked for her. . . . Plaintiffs’ treating physician during her hospitalizations, testified that no other bipolar medications available in 1996 were viable options to treat [her].

Id. Thus, “no dispute of material fact remain[ed] for the jury to decide.” Id. Given how bad the facts were for the plaintiff in Swanson, we’re surprised (and pleased) that the other side didn’t just drop the case.

Willis is a substantially longer decision than Swanson. Additional rulings in Willis put no stock in the peculiar “clear evidence” preemption standards propounded in In re Fosamax, 852 F.3d 268 (3d Cir. 2017). Willis, 2017 WL 5988215, at *6. “[T]he Sixth Circuit has already determined that the evidence in Rheinfrank, which is identical to the evidence in this case, met the ‘clear evidence’ standard. Therefore, whatever ‘clear’ means, it has been met.” Id. The court accepted the defendant’s regulatory evidence that the FDA’s rejection of proposed label changes extended to autism, id. at *7, and rejected a variety of evidentiary attacks on that evidence. Id. at *8-9. Plaintiffs’ attempt to construct a “design defect” claim out of “the drug’s indications” also failed:

[T]he lack of contraindications for the drug would still not be considered a part of its design but rather its “warnings and instructions.” Thus, the Court will not let the plaintiffs’ claim proceed under a theory that the failure to contraindicate Depakote made the drug’s design defective.

Id. at *10.

Daubert motions in Willis only knocked out one of the plaintiffs’ three experts, id. at *12-15, but those causation motions did succeed in significantly limiting the scope of the injuries at issue – by removing from consideration the only birth defect that the minor plaintiff allegedly suffered. Id. at *15. Thus, plaintiffs remaining warning-based claims – for negligent misrepresentation and fraud failed for lack of cognizable injury. Id. at *16 (plaintiff “cannot assert a claim for negligent misrepresentation, as he suffered no cognizable injury that was allegedly caused by the misrepresentations”), *17 (same result for fraud).