In the aftermath of Levine, with its generous interpretation of the CBE regulation and its novel “clear evidence” standard, we wondered how long it would be until we saw a court holding that a failure to warn claim with a branded prescription drug was preempted. Courts were chilled for a while, but eventually the right sort of cases found their way to judges who understood preemption. Now, we have a pretty big list of decisions finding preemption of such claims, along with decisions exhibiting supportive reasoning. We are not yet at the point where preemption of failure to warn claims with branded prescription drugs—for a long time, the core claim in the biggest litigations in our bailiwick—is no longer news. Preemption is still the exception—limited to cases with a strong regulatory history of FDA rejecting the warning plaintiff wanted—rather than the rule, particularly when it comes to favorable appellate decisions.
Rheinfrank v. Abbott Labs., Inc., __ Fed. Appx. __, 2017 WL 680349 (6th Cir. Feb. 21, 2017), is another favorable appellate decision on preemption. You may recognize the name—especially if you are a blog aficionado—from our prior posts on the case. We posted on partial summary judgment being granted as to part of the failure to warn claims being offered—on preemption—and the punitive damages claim—on lack of proof of relevant FDA fraud to meet the exception under the Ohio Product Liability Act provision generally precluding punitives for FDA-approved drugs. We posted on the expansion of the preemption ruling on motion to reconsider to include design defect. (These garnered an honorable mention in our list of the best decisions of 2015.) We even posted on motions in limine rulings. Even with all of those posts, a brief recap of the facts might help. The minor plaintiff’s mother took the prescription anti-seizure medication at issue for fifteen years, including through four pregnancies, before she became pregnant with plaintiff. She kept taking the medication at issue, along with another anti-seizure medication she had been taking, through the birth of plaintiff, who was diagnosed with “physical deformities and cognitive disabilities, including Fetal Valproate Syndrome.” 2017 WL 680349, *1. The label for the medication at issue had long featured a black box warning and other warnings about birth defects, focusing on neural tube defects like spina bifida and discouraging use during pregnancy unless use of the medications “are clearly shown to be essential in the management of their seizures.” Id. at *2. Over the course of seven years after plaintiff’s birth, FDA refused the manufacturer’s repeated efforts to revise the label to address developmental delays in offspring based on data from a study that was ultimately published in the New England Journal of Medicine. Id. at **2-4. A revision of the labeling was ultimately submitted by CBE and accepted by FDA in 2011. Id. at *4. The prescriber back in 2003 and 2004 testified that she was aware of the black box warning on birth defects, would have relayed it to plaintiff, and would not have relied on other materials outside the label. Id. at *2.
Somehow, on this record, the plaintiff got to trial. Under the logic of “all’s well that ends well,” we will limit our rant on this point. After all, we have discussed other birth defect cases that got to trial despite obvious issues, resulted in big verdicts, and got affirmed on appeal. Rheinfrank proceeded to trial under the portion of the strict liability failure to warn claim that was not preempted, a strict liability claim for failure to confirm to representations, the portion of a common law negligent failure to warn claim that was not preempted, and a common law negligent design claim. Among the reasons why the two failure to warn claims should not have seen a court are that 1) Ohio law requires the allegedly inadequate warning to relate to the injury plaintiff claims, 2) claims relating to developmental delays (including as part of Fetal Valproate Syndrome) were preempted, and 3) the prescriber was aware of black box warnings about really serious birth defects and the recommendation against prescription during pregnancy in most situations. It is hard to see how plaintiff mustered evidence of proximate cause—that is, that a proposed (non-preempted) alternative warning as to a risk of an injury the plaintiff had (based on evidence that existed when the prescription was written) would have changed the prescriber’s decision to prescribe—to survive summary judgment. Based on the jury instructions that plaintiff proposed at trial, it seems like a broader discussion of risks and the impact of different warnings about risks was permitted than maybe should have been, which is often a reason why failure to warn claims get past summary judgment. Given that the prescriber denied reliance on any representations outside the label, it is hard to see how that claim got to the jury. As for the negligent design claim, it is hard to see how the same reasoning for preempting the strict liability design claim would not have applied or how a design of the drug—without being a different drug—that lacked the same birth defect risk could have been offered. Anyway, the trial judge may have known what was coming, because the jury listened to the just about the best plaintiff could offer and returned a defense verdict on all counts after two weeks.
On appeal, the plaintiff had a number of gripes. Among them was that her regulatory expert—one Dr. Parisian, whose regulatory experience with drugs has been the subject of considerable commentary—was precluded (kindof) from opining that the drug had greater birth defect risks than other drugs and that her medical experts were precluded from giving a range of regulatory opinions—even though they were allowed to opine on whether the drug’s label accurately reflected the state of the science at the time. We will not discuss these other topics except to say that plaintiff clearly got quite a bit of latitude to present his case and the preemption rulings appear not to have excluded all the evidence that they might have. That lets us focus on the Sixth Circuit’s de novo consideration and affirmance—although without a recommendation for publication—of the ruling on preemption of failure to warn claims related to the risk of developmental delay. It looks like plaintiff did not appeal as to the preemption of the strict liability design defect claim or the application of the OPLA’s punitive damages provisions, which should tell you something (and Yates v. Ortho, 808 F.3d 281 (6th Cir. 2015), undoubtedly had something to do with the former).
“Because the FDA twice refused Abbott’s attempts to strengthen [the drug’s] label, based on its own review of the evidence that the drug adversely affected the development of children exposed to it in utero, Rheinfrank’s failure-to-warn claim is preempted by federal drug labeling law.” Id. at *13. Looking at the entire history of the labeling, it was arguably three rejections of a labeling change like the plaintiff wanted, but who is counting? This was clear evidence that FDA would not have allowed the label to change before the prescription to plaintiff, based on far less evidence of a risk. “Together Wyeth and Mensing therefore indicate that ‘a court cannot order a drug company to place on a label a warning if there is ‘clear evidence’ that the FDA would not approve it.’” Id. (citing Robinson v. McNeil Consumer Healthcare, 615 F.3d 861, 873 (7th Cir. 2010) (an OTC drug warnings preemption case)). Part of the manufacturer’s evidence was the denial of a Prior Approval Supplement, which negates the typical argument that there is no conflict with federal law when a Changes Being Effected supplement could have been submitted by the manufacturer. “As the Court has since clarified [after Levine], given the FDA’s authority to rescind any unilateral CBE, all that [the manufacturer] need have done—and did do here—is show that ‘the FDA would have rescinded any change in the label,’ a showing that does not appear to exclude the kind of informal communications from FDA higher-ups that [the manufacturer].” Id. at *14 (citing Mensing, 564 U.S. at 624 n.8).
An important preemption holding in Rheinfrank has to do with the timing of the FDA’s rejections of the relevant proposed label change as against the plaintiff’s injuries. The court held, “[g]iven, then, that as of 2008 the FDA did not believe the state of the data supported a developmental delay warning, it stands to reason that as of 2003, with even less data to go on, the FDA would similarly have rejected a developmental delay warning.” Id. Thus, Rheinfrank joins those courts that have drawn a preemptive line barring all plaintiffs who used a drug prior to an FDA insufficient evidence decision concerning the risk at issue. We discussed those cases here.
The court also revealed an appreciation that the facts of Levine, where FDA’s attention to the risk at issue was considered merely “passing,” may not have broad application. “Here, by contrast, there is considerable evidence in the record showing that [the manufacturer] followed up on new data surrounding developmental delay and [the drug], that [the manufacturer] contacted the FDA multiple times to propose modifications to reflect that data, and that each time officials from the FDA unit responsible for reviewing those modifications unequivocally stated that they were inappropriate at the time.” Id. This level of attention provides evidence “clear enough to satisfy Wyeth’s standard” and “make clear that the FDA would have rejected the proposed change to [the drug’s] label.” Id. The court even rejected plaintiff’s attempt to inject a failure to test theory into the preemption analysis—that there should not be preemption because maybe the FDA would have done something different if the defendant had done extra studies and they had reported a risk of developmental delay. Relying on Mensing, this was “too conjectural to defeat preemption.” Id. at *15. And, with that, plaintiff’s last argument to reverse the well-thought out preemption decision of the district court ended.