It wasn’t a complete win, but the summary judgment outcome in Rheinfrank v. Abbott Laboratories, Inc., ___ F. Supp.3d ___, 2015 WL 4743056 (S.D. Ohio Aug. 10, 2015), has to put a spring in the step of the defendants as they approach trial. What’s left doesn’t strike us as a very good warnings case. Rheinfrank involved claims that the antiepileptic drug Depakote caused the minor plaintiff’s birth defects. Make no mistake about it, Depakote has a known association with such injuries. First approved in 1983, it’s been a Pregnancy Category D drug since 1988, meaning, according to FDA regulations, that:
there is positive evidence of human fetal risk based on adverse reaction data from investigational or marketing experience or studies in humans, but the potential benefits from the use of the drug in pregnant women may be acceptable despite its potential risks.
21 C.F.R. §201.57(c)(9)(i)(A)(4). Not only that, since 2003, this drug has carried a black box “teratogenicity” warning, as well as other quite explicit, and all-caps, language to the same effect. For details, see 2015 WL 4743056, at *2-3.
Plaintiff-mother had used Depakote for years, through four previous uneventful pregnancies. Id. at *1. On her fifth pregnancy, even though Depakote came with all these warnings, she continued to take it. Id. Her allegations did try to change the subject, however. In addition to claiming that the black box warning (more about that later) and all the other teratogenicity language were inadequate, she asserted that the defendants failed to warn altogether about “developmental delay.” Id. at *5.
Trouble was, the defendant had tried unsuccessfully for years to get the FDA to add a developmental delay warning, but the FDA hadn’t let them. The regulatory facts:
- In 2005, the manufacturer sent a letter to the FDA, trying to “update” the label, proposing “revised information related to teratogenicity and additional information for developmental delay” in both the warnings and patient information section. Id. at *7.
- The 2005 FDA submission included “[a]n outline of safety-related changes for teratogenicity/developmental delay” including “[n]ew information concerning the use of [Depakote] in women of childbearing potential: teratogenicity and developmental delay.” Id. The submission was supported by a “white paper” on this risk. Id.
- The 2005 submission proposed to warn about both an increased risk and “reports” of developmental delay. Id. at *8.
- “On February 7, 2006, the FDA responded . . . and stated the proposed sentence referencing developmental delay should not be incorporated into labeling.” Id.
- FDA rejected the proposed labeling because the studies on which it was based “did not adequately control for” several confounding variables. Id.
- The manufacturer tried again in 2007, writing to the same FDA employee who had handled the 2005 proposal and specifically “requesting advice” about “developmental labeling.” Id.
- The 2007 initiative included “and updated analysis” and “more compelling data” on developmental delay and Depakote. Id. It again proposed explicit warnings about “developmental delay.” Id.
- On March 3, 2008, the FDA again said no. The data were “not ripe” and only “interim.” The “sample size” was too small, and “there are too many confounding factors to believe the data is reliable.” The defendant’s adverse events – 240 of them – were also “too confounded to assess.” Id. at *9.
- The manufacturer tried a third time in April, 2009, submitting four new articles from the medical literature. Id.
- In September, 2009, the FDA told the defendant to sit tight; that the Agency had requested data directly from the main researcher “and planned to independently review the data.” Id. FDA had “concerns” with the researcher’s “methodology” and “needed time to evaluate the data.” Id.
- The manufacturer responded by asking if it could submit a “changes being effected” (“CBE” or “CBE-0” – they are the same thing) in the interim – to which the FDA responded it could, but the Agency “would not take action until reviewing the data.”
- “[O]n November 18, 2009, [the manufacturer] submitted a CBE-0 to add developmental delay warnings for Depakote.” Id.
Plaintiffs’ big problem in Rheinfrank was that all this activity – particularly the FDA’s refusals based on insufficient data – occurred after the drug had been taken, back in 2003. If the FDA thought the data lacking as late as 2009, then there was necessarily even less data to support a label change several years earlier in 2003.
This subsequent regulatory history provided “clear evidence” under Wyeth v. Levine, 555 U.S. 555 (2009), that the FDA would not have allowed a developmental delay warning at any time in 2003 that could conceivably have affected the pregnancy at issue in Rheinfrank:
Preemption is warranted because there is clear evidence the FDA would not have approved a change to the Depakote label adding a developmental delay warning prior to [plaintiff’s] injury. The Court finds the FDA’s February 2006 decision that developmental delay warnings “should not be incorporated into [Depakote] labeling” and the FDA’s 2008 belief that “the data do not provide sufficient evidence to support [Depakote] labeling changes at this time” constitute “clear evidence” that when confronted by the issue in 2003, the FDA would have rejected an attempt to add a developmental delay warning.
2015 WL 4743056, at *10 (citing Robinson v. McNeil Consumer Healthcare, 615 F.3d 861 (7th Cir. 2010); In re Fosamax, 951 F. Supp.2d 695 (D.N.J. 2013); Dobbs v. Wyeth Pharmaceuticals, 797 F. Supp. 2d 1264 (W.D.Okla.2011), Kaleta v. Abbott Laboratories, Inc. (In re Depakote), ___ F. Supp.3d ___, 2015 WL 4880655 (S.D. Ill. Feb. 20, 2015)).
What the hey…? Where did this prior Depakote opinion from last February – one that’s favorable on post-Levine preemption – come from? We’d never heard of it. Well, we have a PACER account and we’re not afraid to use it. Sure enough, this opinion exists (and now is headed for F. Supp., as is Rheinfrank). It reached the same result, only on a motion in limine in the context of the Depakote MDL. While the timing in Rheinfrank was bad for the plaintiffs, the timing in the Depakote bellwether was atrocious: try 1999, over a decade before the FDA even started to move during the above regulatory history. Depakote reached the same preemptive result:
The Court finds that there is clear evidence that the FDA would not have approved a change to the 1999 label to include a warning of developmental delay. . . . In light of the fact that the FDA rejected the developmental delay warning in 2006 because it did not find that the available scientific evidence at that time supported the addition of such a warning, it is highly unlikely that the available scientific evidence, seven years prior to that date in 1999, would have supported the addition of such a warning. Notably, the FDA did not actually approve this developmental delay language until 2011.
Depakote, 2015 WL 4880655, at *3 (emphasis original). Since all of the relevant administrative avenues for a label change (CBE, prior approval supplement, request for advice) involved the “same standard” – “reasonable evidence of an association” – the fact that the FDA rejected one, was “clear evidence” it would have likewise rejected any other approach. Id. at *4. The FDA had to balance competing concerns in deciding whether adequate evidence exists:
While it is important for a manufacturer to warn of potential side effects, it is equally important that it not overwarn because overwarning can deter potentially beneficial uses of the drug by making it seem riskier than warranted and can dilute the effectiveness of valid warnings. Therefore, warnings may only be added when there is “reasonable evidence of an association of a serious hazard with the drug.” It is technically a violation of federal law to propose a CBE that is not based on reasonable evidence.
Rheinfrank, 2015 WL 4743056, at *13 (quoting Depakote, 2015 WL 4880655, at *4) (other citations and quotation marks omitted). Courts expressing overwarning concerns? We love it.
Since the Supreme Court never bothered to define “clear evidence” in Levine, the rulings in Rheinfrank and Depakote provide another fact pattern against which courts can judge post-Levine preemption.
These cases also disposed of several legal arguments against preemption. As is common, plaintiffs retained regulatory “experts” to try to pick apart the defendant’s FDA submissions, claiming that this or that data was withheld, buried, or misinterpreted. Fortunately, these courts recognized that you can’t do that either:
Plaintiffs support their position with proffered expert testimony of four experts who have reviewed and opined on allegedly misleading information in [defendant’s] White Papers. . . . Plaintiffs’ argument that [defendant] withheld certain information or misrepresented the results of studies in its 2005 and 2007 submissions to the FDA appears to be a fraud-on-the-FDA theory, which is preempted.
Rheinfrank, 2015 WL 4743056, at *11 (citing, inter alia, Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001)). Hypothesizing that the FDA might have done something other than what it did was also “speculative”:
[A]n expert’s opinion that the FDA would have reacted differently if the submissions to the FDA in 2005 and 2007 had been supported by different evidence is speculative. . . . [W]hat [defendant] could have or should have done is immaterial because we know what [it] did. Similarly, Wyeth v. Levine provides for preemption where there is clear evidence that the FDA would have rejected a label change, and, again, we know that the FDA did reject it. . . . Testimony about what [defendant] could have and should have researched or stated to the FDA in its applications is speculative, and does not serve as the basis for a genuine issue of material fact.
Id. at *12 (citation and quotation marks omitted) (emphasis original).
Plaintiffs also tried to play the “what if” game, alleging that the defendants could/should have conducted additional research to move the science along faster. The courts rejected this attempt to rewrite history. “Plaintiffs reason that if [defendant] had funded or conducted studies that generated subsequently published data, those studies would have generated the same data sooner, and by [whatever time plaintiffs need to prove causation], the FDA would have allowed the developmental delay warnings based on that data. This reasoning is speculative.” Rheinfrank, 2015 WL 4743056, at *11. Attempting to rewrite history is also confusing and collateral. “Providing jurors with the context necessary to make sense of the developmental delay warnings would necessitate a mini-trial on the label change [defendant] tried to make later, about the evolution of developmental delay literature/data, [defendant’s] FDA submissions, whether these submissions were sufficient, and the FDA’s decisions with regard to those submissions.” Depakote, 2015 WL 4880655, at *4.
Plaintiffs’ fall-back position was that the defendant should have rewritten the boxed warnings that the FDA had ordered back in 2003. You can’t do that either, as we’ve discussed before, because the FDA controls boxed warnings:
The FDA regulations on Black Box warnings provide that: “to ensure the significance of boxed warnings in drug labeling, they are permitted in labeling only when specifically required by FDA.” 44 Fed. Reg. 37, 434, 37, 448 (June 26, 1979). . . . The parties agree that Abbott could not have unilaterally
changed the Black Box Warning.
Depakote, 2015 WL 4880655, at *5 (emphasis original).
That’s preemption, and that’s the extent of the relevant rulings in Depakote. Rheinfrank, deciding summary judgment motions, was more extensive, and addressed other interesting issues. It cut the plaintiffs a bit of a break by refusing to hold that the extensive teratogenicity warnings (including the black box) were adequate as a matter of law. “[T]he Court is not persuaded that the mere fact that the label listed Depakote as a Pregnancy Category D drug and included a Black Box warning indicates that the label was adequate as a matter of law.” Rheinfrank, 2015 WL 4743056, at *15.
One of the purported “deficiencies” that prevented summary judgment was “not stating that birth defects are greater with [Depakote] than with other AEDs.” Id. That, however, is a comparative claim, and there’s a good basis for a preemption argument there, as well. As we’ve discussed before, the FDA requires “adequate and well-controlled studies” before it will allow such claims. 21 C.F.R. §201.57(c)(3)(v). See Guenther v. Novartis Pharmaceutical Corp., 2013 WL 4648449, at *6 (M.D. Fla. Aug. 29, 2013) (comparative claim preempted absent proof of FDA-mandated level of scientific support). There’s also considerable law for the common-sense proposition that manufacturers simply have no duty to discuss the relative safety of competing products – including (insofar as relevant specifically to Rheinfrank) the Sixth Circuit interpreting Ohio law. Ackley v. Wyeth Laboratories, 919 F.2d 397, 405 (6th Cir. 1990). To all that we add the fact that Rheinfrank is a federal diversity case, as to which controlling Erie principles require federal courts to refrain from predicting novel, expansive tort duties. See, e.g., In re Darvocet, Darvon, & Propoxyphene Products Liability Litigation, 756 F.3d 917, 937 (6th Cir. 2014) (“federal courts must be cautious when making pronouncements about state law and when given a choice between an interpretation of [state] law which reasonably restricts liability, and one which greatly expands liability, we should choose the narrower and more reasonable path”). Thus, not only are these warning claims rather unattractive to a jury (given the extensive warnings that would have to be found “inadequate”), but legally they are extremely tenuous.
While Rheinfrank let those warning claims slide for the moment, it undercut them significantly. First, it held that the plaintiffs’ expert wasn’t qualified to testify on FDA issues. Rheinfrank, 2015 WL 4743056, at *16 (“testimony about what Defendants should have included in the label or what materials should have been submitted to the FDA falls outside the scope of his expertise”). Second, it eliminated plaintiffs’ weird Restatement §324A “good Samaritan” assumed-duty theory, based on the label already being comparative:
[T]he “voluntary duty” rule has no application to cases based on failure to provide adequate warnings with prescription drugs, whether grounded in negligence or strict liability, as in this case. This is so because it has become a well established rule in such cases that the manufacturer satisfies his duty to warn of dangers associated with use of the product by providing adequate warnings to the medical profession, and not the ultimate user” under the learned intermediary theory.
2015 WL 4743056, at *17 (citation and quotation marks omitted). Rheinfrank held simply that “whether it was adequate for Defendants to offer a comparative warning is a question of fact for the jury.” Id. While we think that’s wrong, given the Erie principles and FDA substantiation requirements already discussed, the ruling leaves plaintiffs with a rather narrow and unpalatable theory.
Rheinfrank also held the warning claims survived the defendants’ common-knowledge defense. Id. at *18-19. While that’s a difficult defense to establish as a matter of law, the discussion indicates additional reasons why plaintiffs have a tough row to hoe in pursuing an inadequate warning claim. We’ll leave it at that, because any readers interested in still more Depakote-specific facts can read the opinion.
Plaintiffs also escaped summary judgment on physician-related warning/causation issues thanks to rather generous interpretation of the prescribing physician’s failure-to-read testimony combined with Ohio’s heeding presumption. Id. at *24-26. Here, again, plaintiffs lost the big battle – their attempt to avoid the learned intermediary rule altogether was rejected. Plaintiffs could not use yet another Good Samaritan assumed duty theory to claim that a modification to a “patient information leaflet” created a further duty to distribute that leaflet directly to all patients:
The “voluntary duty” doctrine has no application to prescription drug failure to warn cases grounded in negligence or strict liability, because of the learned intermediary defense. . . . Informational pamphlets . . . can aid in expanding the range of communications between doctor and patient. . . . Nevertheless, a direct relationship between the manufacturer and the patient does not arise as a result of the provision of such brochures. . . . [W]e do not believe that by preparing such brochures and distributing them to physicians, a prescription drug manufacturer undertakes to render a voluntary service so as to invoke the “voluntary duty” rule . . . thereby extending the scope of its duty to warn. . . . [I]in this case, [defendant’s] voluntary expansion of its patient information leaflet in its package insert did not abrogate the role of the learned intermediary; rather, it was intended to aid the learned intermediary.
Rheinfrank, 2015 WL 4743056, at *23-24 (citations and quotation marks omitted). Once again, the most dangerous aspect of the plaintiff’s warning claim was thrown out.
Rheinfrank granted summary judgment against the plaintiffs’ design defect claim. Ohio has a statutory alternative design requirement, so they tried to argue that a different manufacturer’s different drug was an alternative design. Id. at *27. No dice.
[T]here is no evidentiary support for Plaintiffs’ contention that there were alternative drugs that could have controlled [plaintiff-mother’s epilepsy] as well as the [drugs] prescribed for her for over twenty years. . . . Merely because some AEDs are considered less teratogenic than Depakote does not mean they would have been suitable for controlling [her] seizures.
2015 WL 4743056, at *28.
Plaintiffs sloughed off (“no information responsive”) an interrogatory about who provided their physicians with inadequate “representations or warranties.” Id. at *29. Too bad for them. Their laziness bound them to answers that precluded any finding of reliance, and thus required summary judgment against negligent misrepresentation and fraud. Id. An unjust enrichment claim likewise failed. “[A] seller is not unjustly enriched simply because he derives a profit from the sale of a product.” Id. at *30 (citation and quotation marks omitted).
Finally, for basically anachronistic reasons, plaintiffs punitive damages claim survived, only with respect to “common-law” claims that have now been abolished for years in Ohio. Id. at *30-31. All punitive damages claims under the current Ohio statutory scheme were barred. The Ohio statute requires proof of fraud on the FDA as a prerequisite, which is preempted under Buckman, supra, absent the FDA confirming it was defrauded. Rheinfrank, 2015 WL 4743056, at *32 (following Monroe v. Novartis Pharmaceuticals Corp., 29 F. Supp.3d 1115 (S.D. Ohio 2014)).
Thus, while the plaintiffs in Rheinfrank get to take part of their warning claim to the jury, their allegations have been shorn of most of their more objectionable aspects. First and foremost, their developmental delay claim is preempted, depriving them of an argument that something was omitted altogether from the defendants’ warnings. FDA issues are out. Plaintiffs are stuck arguing that black box/pregnancy category D/all caps teratogenicity warnings were “inadequate,” based on prescriber causation testimony that’s rather thin. Even to get to that point, plaintiffs have to assert a factually (was anything actually compared?), legally (no basis in Ohio law), and jurisprudentially (Erie doctrine) dubious “comparative” product warning claim. Two other attempted expansions of warning duties through “Good Samaritan” assumed duty arguments failed. The learned intermediary rule survives intact. All claims other than warnings fail for one reason or another. Finally, while we don’t like punitive damages, the only such claim that survived doesn’t exist any longer under current Ohio law. All in all, some very useful precedent.