We have been monitoring litigation involving tenofovir-based HIV medication for some time now. We reported a few weeks ago on oral argument in the California Court of Appeal, where the parties debated a novel “duty to innovate” under California law. We also gave you our view on the 2019 order that many say kickstarted these cases, Holley v. Gilead Sciences, Inc., 379 F. Supp. 3d 809 (N.D. Cal. 2019). Although couched in various terms, the plaintiffs’ core allegation in these cases is that the defendant should have marketed another drug instead of the one that the plaintiffs used, or at least that the defendant should have marketed the alternative drug earlier. We blogged on similar cases here, here, and here.
As we explained here, “hindsight” claims like these should not be allowed. A claim that the defendant “never should have sold” a drug should be preempted, the same as “stop-selling” claims, since state law should not be allowed to penalize the introduction of a drug to market when federal law expressly authorized it. A plaintiff also should not be allowed to hold out a different product as an alternative design. One drug is not an alternative design for a different drug any more than a motorcycle is an alternative design for an automobile. You get the idea.
The Holley case has now proceeded to summary judgment, and the result is similarly unsatisfying for the defense. To summarize, the drugs at issue are life-saving antiretroviral drugs used to treat patients with HIV. The defendant has developed multiple drugs used to treat or prevent infection with the AIDS virus, including several containing tenofovir disoproxil fumarate (“TDF”) as an active ingredient. The plaintiffs in Holley claim that the defendants should have offered a different drug instead, one based on tenofovir alafenamide (“TAF”), which the defendant had under development and brought to market about 10 years later. Both drugs remain on the market, and physicians can and do prescribe either, depending on their patients’ needs.
On summary judgment, the district court ruled again that federal law does not preempt the plaintiffs’ claims. We still disagree with this conclusion and believe the district court’s characterization of the plaintiffs’ claims only reinforces our view: “[Plaintiffs] allege that the TDF drugs they consumed were defective because [Defendant] failed to consider TAF as a safer alternative and that, instead of seeking approval for the TDF drugs, [Defendant] should have offered a different formulation for FDA approval.” Holley v. Gilead Sciences, Inc., No. 18-cv-06972, 2023 WL 6390598, at *3 (N.D. Cal. Sept. 28, 2023).
Taking the district court’s own gloss at face value, these are claims that the defendant never should have marketed TDF drugs and should have sold TAF drugs instead. This is a “never should have sold” claim that is preempted under such authorities as Yates v. Ortho-McNeil-Janssen Pharmaceuticals, Inc., 808 F.3d 281 (6th Cir. 2015), and the Supreme Court’s opinion in Mutual Pharmaceutical Co. v. Bartlett, 133 S. Ct. 2466, 2471 (2013). It makes no difference in our view that the FDA ultimately approved TAF-based drugs. It is undisputed that, at the time the FDA approved the first TDF-based drug, the plaintiff’s proposed “alternative” was not approved, and would not be approved for another 10+ years.
The district court made numerous other rulings as well, and we will note a few here. First, the court ruled that strict liability design defect claims are not allowed under the laws of several states, including the only state for which that was contested, Alabama. Id. at *3-*4.
Second, for the states that apply a risk-utility test, the court denied summary judgment on design defect claims, even though the plaintiffs’ experts had not offered opinions weighing the risk and benefits of the TDF drugs that the plaintiffs used. Instead, the plaintiffs’ experts had offered opinions that TDF was associated with kidney injury and that TAF drugs were safer. Id. at *4-*5. We question whether these opinions were sufficient for the claims to survive summary judgment, mainly because a risk-benefit analysis should weigh particular product’s risks and benefits—not the product’s risks weighed against the risks of a different product. Tylenol is harder on the liver than ibuprofen, but that does not mean Tylenol is defective, or say anything about how Tylenol’s liver risk weighs against its benefits. A comparison of TDF against a different drug is similarly uninformative, and unprecedented.
Third, the district court ruled that the plaintiffs’ “post approval” failure to warn claim was preempted, but that its “pre-approval” failure to warn claim was not. On the former, the court granted summary judgment, noting that federal law preempted a “post-approval” failure to warn claim unless the plaintiff could produce “newly acquired information.” Such information would have allowed the defendant unilaterally to change its product label under the FDA’s “changes being effected” rules, thus avoiding preemption. The plaintiffs’ submission of internal studies, case reports, and scientific literature did not show “newly acquired” information—the FDA was already aware of it all. Id. at *7.
The court, however, denied summary judgment on “pre-approval” failure to warn, which is not really a failure to warn claim at all. It is just a claim that the defendant should have proposed to the FDA a different warning than what the FDA approved. When looking at it that way, we cannot reconcile an order finding that “post-approval” failure to warn is preempted with an order finding that “pre-approval” failure to warn is not. The plaintiffs’ failure to identify “newly acquired information” should result in preemption of all warnings-based claims. And, if there was material information existing at the time of approval (i.e., not newly acquired) that the FDA needed and that the defendant withheld, that would be fraud on the FDA, which is preempted under Buckman in any event.
The district court made other rulings, but these are the highlights (or lowlights). These cases are interesting because of the novel and, in our view, tenuous basis on which they are being brought. Hindsight is 20/20.