It’s been a little slow over the holidays so we’ll go back to something that happened last year. On December 16, 2014 – almost three whole weeks ago – the Solicitor General filed his brief responding the Supreme Court’s request for the government’s views in the case called Teva Pharmaceuticals USA, Inc., v. Superior Court Of California, Orange County, No. 13-956. Here’s a link to the case’s SCOTUSBlog page. The prior published opinion in this case, Teva Pharmaceuticals USA, Inc. v. Superior Court, 158 Cal. Rptr.3d 150 (Cal. App. 2013) (which we’ll call “Pikerie” because that’s the plaintiff’s name), involved preemption in the generic drug context. Here’s our description of Pikerie from the blog’s generic preemption scorecard:
Overruling of demurrer affirmed. A duty to update claim is not preempted. It is not impossible to update warnings that the FDA has ordered changed, it is required. Following Fulgenzi. A “Dear Doctor” letter claim is not preempted, to the extent that the communication followed FDA-approved updates in the labeling.
Long-time readers might recall that the last time the SG’s office was invited to weigh on the FDCA and preemption of product liability litigation, in the Stengel case, it produced a ridiculously cramped version of PMA preemption that sought to emasculate Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), and would have allowed virtually every claim to escape preemption. We reported on that here. Equally bizarrely, after explicitly stating that “every” court of appeals since Riegel had (in his view) gotten Riegel wrong, the SG in Stengel nonetheless recommended that certiorari be denied – hardly the a position taken by a litigant confident that the Court would agree with its position.
Thus, we know that this administration is at least as hostile to FDCA-related preemption as the prior administration was supportive. Given that “duty to update” cases are really lousy cases for plaintiffs for a host of reasons unrelated to preemption (that’s why they’ve never been brought before), we’d have to think long and hard before doing something that would give this same crew a chance to take shots as preemption under Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001). But the generic industry seems to have concluded it’s worth it, and their lawyers aren’t dumb.
The questions presented by the Pikerie petition are:
- Whether a drug manufacturer’s state-law duty to warn of its generic drug’s risks, which is consistent with the federal regulatory obligation to update the generic drug’s labeling to match that of its brand-name counterpart, is preempted by 21 U.S.C. 337(a)’s grant of exclusive authority to the United States to enforce, and to restrain violations of, the FDCA.
- Whether the FDCA renders it impossible to comply with, and thus impliedly preempts, a generic drug manufacturer’s state tort-law duty to warn by communicating such updates to healthcare professionals using Dear Health Care Provider letters.
See Brief for the United States as Amicus Curiae, 2014 WL 7169712, at *1 (U.S. filed Dec. 16, 2014) (also available here)
As in Stengel, the SG concludes that certiorari should be denied, but then takes the position that there is no preemption. Surprise, surprise, surprise. The rationale for denying certiorari is case specific and doesn’t interest us much. It has to do with the interlocutory procedural posture of the Pikerie case. SG brief, 2014 WL 7169712, at *9-13. Even there, however, the SG did not resist taking a shot at Buckman:
In particular, petitioners err in contending that Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001), broadly bars tort actions under state law that parallel duties under the FDCA and implementing regulations.
SG brief, 2014 WL 7169712, at *12. The supporting citations are only from Wyeth v. Levine, 555 U.S. 555 (2009), which of course had nothing whatsoever to do with “parallel” claims or private attempts to enforce the FDCA.
Given this administration’s history – going all the way back to the FDA’s largely empty preemption “search and destroy” mission of 2009-11, it is not surprising that they would seek to make Buckman as much of a dead letter as they would do to Riegel or Mensing. But actually it doesn’t, at least not quite. The SG starts with the questionable proposition that fraud is not “traditional state tort law” whereas failure to warn is – even where both are alleged to have taken place in the context of the FDCA. Id. at *14-15. Last time we looked fraud was a state-law claim. So were the Buckman fraud on the FDA claims. They were brought under Restatement (Second) of Torts §310 (1965). We know; some of us were there.
Briefly, the SG offers an even more strained excuse for limiting Buckman to its facts: that the claim in Buckman had to be “solely” federal “because the defendant [an FDA consultant] was not the manufacturer of the devices and therefore did not have a manufacturer’s duty to warn.” 2014 WL 7169712, at *14. Really? The SG wisely doesn’t try explaining this statement, which if taken at face value would mean that even fraud on the FDA claims aren’t preempted if brought against manufacturers. That, of course, would put the SG at cross purposes with everything in Buckman.
This extreme position indicates that Buckman’s statement that the fraud existed “solely by virtue of the FDCA,” 531 U.S. at 353, simply can’t mean what the SG claims. The Supreme Court wasn’t talking about fraud (which has always been state law), but about the opportunity for fraud, which was solely federal. Only the FDCA required the submissions that plaintiffs in Buckman alleged were fraudulently made, which the Court discussed at some length. Id. at 349-48. Preemption existed because of the exclusively federal theater into which the broader common-law fraud claim purported to intrude. That scenario created the conflict, both with the FDA’s prosecutorial discretion and with possibly burying the FDA in needless paper. Id. at 349-51.
Similarly, there is no “duty to update” under state law. There is only a general duty to provide adequate warnings. Some of those, such as the FDA’s requirement that generic warnings be updated (FDA guidance says within 30 days, see SG brief, 2014 WL 7169712, at *5 n.1) after innovator changes, or providing DHCP/”Dear Doctor” letters, are every bit as dependent on FDA labeling requirements as “fraud on the FDA” was dependent on FDA submission requirements in Buckman.
And even this SG seemed to recognize that, as the “it’s not really FDA enforcement” was followed immediately by this:
The FDCA and FDA’s implementing regulations can, of course, limit the actions a drug manufacturer can take, and state-law duties that conflict with such federal limitations would be preempted. A generic drug manufacturer, for instance, cannot currently change its labeling to identify a new risk if such a change would depart from the labeling of the relevant brand-name counterpart. But such limits do not suggest that a state-law duty to warn is preempted merely because a drug manufacturer could satisfy that duty only by taking actions that comport with federal law.
SG brief, 2014 WL 7169712, at *15 (citation omitted) (emphasis added). So a “conflict” would still be preempted.
But then it’s back to Levine again, which is puzzling. Levine is not a parallel claim case. The Levine majority cited Buckman only once, in a footnote, about the presumption against preemption, which isn’t at issue in Pikerie either. The SG goes on to make the remarkable statement that “if petitioners’ expansive reading of Buckman were correct, Buckman presumably would have barred the duty-to-warn claims in [Levine].” 2014 WL 7169712, at *17.
Wow, that’s a big “presume.” Buckman preceded Levine. If that argument actually existed, one would have expected the innovator manufacturers to have made it in Levine – and the dissent. Nobody did, as the argument, which the SG does not explain, would be nonsensical. There were plenty of reasons why the duty-to-contraindicate claim in Levine should have been preempted. The dissent enunciated them. Buckman impingement on exclusive federal enforcement wasn’t one of them.
What Buckman did say, in summarizing its holding, was this:
In sum, were plaintiffs to maintain their fraud-on-the-agency claims here, they would not be relying on traditional state tort law which had predated the federal enactments in questions. On the contrary, the existence of these federal enactments is a critical element in their case.
531 U.S. at 353 (emphasis added). One doesn’t find that statement anywhere in the SG’s brief. Cue Gomer again.
Buckman preemption, insofar as based on exclusive enforcement, as opposed to direct conflict (that an in-force FDA approval could be ignored as fraudulent), is based on the federal enactment being a “critical element.” By definition, the underlying violation, whatever it is, is a critical element of a negligence per se claim. Otherwise, it’s back to simple negligence. Buckman’s “critical element” bottom line can’t be squared with what the SG said about negligence per se:
But “a complaint alleging a violation of a federal statute as an element of a state cause of action” is still a state-law, rather than a federal-law, claim, and, as discussed, the state-law duty to warn here exists independently of federal law. California’s per se negligence doctrine, in turn, merely establishes a rebuttable presumption of negligence. Proof of a violation of federal law thus is neither necessary nor sufficient to establish liability under state law.
SG brief, 2014 WL 7169712, at *17 (citations omitted).
So, is the FDCA violation in Pikerie only an “element,” but not a “critical” one, of negligence per se? Doubtful, because otherwise the “per se” aspect of the claim has vanished. If the FDA says 30 days is sufficient, then it’s hard to say how negligence per se can be anything different. That’s pretty “critical.” Sure, the claim can fairly be described as a “state” cause of action, but so did the Supreme Court preempt “state-law causes of action” in Buckman. 531 U.S. at 346-47. That doesn’t move the ball an inch. Is it the peculiarity of California law allowing only a “presumption,” and not a finding of liability? If so, there’s another failure to update from Pennsylvania right behind Pikerie, that doesn’t have that quirk. So we don’t think so either. Nope, to us the bottom line is that Buckman’s “critical element” analysis cannot be squared with any FDCA-based negligence per se, and so the SG brief ignores that language altogether.
To us, that’s the point that most deserves review: “critical element” is a very different standard than “solely by virtue.” Where we have the SG arguing for one (which would allow even fraud on the FDA to lie against a manufacturer) and ignoring the other altogether. A parallel claim case of some sort, whether involving generic drugs or PMA medical devices, would seem to be the proper vehicle for review. We’d just rather that this vehicle have more of a conflict element to it than failure to update cases do, at least than they do at the pleadings stage. We can think of any number of PMA-related parallel claim cases that read FDA regulations wrong that would present better atmospherics for a Buckman review.
The SG also argues that failure to update doesn’t impinge on the FDA’s enforcement discretion. 2014 WL 7169712, at *18. Once again we find citations to Levine, which wasn’t an enforcement case at all. The SG’s brief contains nothing indicating that the FDA has ever thought that failures to update were serious enough to prosecute. Interestingly, Levine did have something to say about nonexistent FDA prosecutions when raised by a defendant: “the very idea that the FDA would bring an enforcement action . . . is difficult to accept − neither [defendant] nor the United States has identified a case in which the FDA has done so.” 555 U.S. at 570. The United States, having changed its preemption tune, still hasn’t identified any case of actual enforcement action.
The SG has less to say about DHCP/”Dear Doctor” letters. He distinguishes PLIVA v. Mensing, 131 S. Ct. 2567 (2011). “Mensing did not address the issue here; whether a DHCP letter could be used by a generic manufacturer to communicate warnings already present in the relevant brand-name labeling.” The emphasis is original this time, and it demonstrates to us the chief reason why “Dear Doctor” letter claims are preempted. Reiterating information “already present” on labeling isn’t what such letters are for. Such a letter exists to convey “new” information, not information already on labels. “[A] DHCP letter can be an appropriate way to bring new information to the attention of medical professionals.” SG brief, 2014 WL 7169712, at *20 (quoting government brief in Mensing); see also id. at *5 (DHCP letters used to distribute “clinically important new information”) (citing FDA guidance). Information that’s “already present” is simply not “new.” Otherwise DHCP letters are reduced to reiterating label changes, their importance vanishes, and the health care profession faces an avalanche of redundant letters from manufacturers afraid they’ll be sued if they don’t send them. Nor would a new DHCP letter, not sent by the relevant innovator manufacturer, fit within the requirement of sameness. See 21 C.F.R. §201.100(d)(1) (all generic labeling must be “the same in language and emphasis”).
As to both failure to update and DHCP claims, the SG argues that, even though prior courts have erred in reaching decisions contrary to the government’s views, such errors don’t warrant certiorari. 2014 WL 7169712, at *18-19, 22. Shades of Stengel.
Finally, a couple of other interesting things we noticed pertaining to the FDA’s pending regulatory initiative to eliminate generic preemption that we thought we’d pass along. First, the discussion of Mensing on pages 4-5, characterizing the decision twice as having “defer[red]” to the FDA’s regulatory interpretations, offers a preview of how the government will attempt to defend the FDA’s current regulatory attempt. The argument will be that the existing exclusion of generic drugs from the CBE regulation was only by virtue of FDA regulation, not the statutory mandate of “sameness,” therefore the FDA can change that by amending its regulations. Good luck. Second, and more useful, the government’s brief also states: “FDA’s proposal is not retroactive and would not apply to pending failure-to-warn claims.” 2014 WL 7169712, at *24. Because we’re sure that some plaintiff will make the contrary argument should the proposal ever go into effect, it’s nice to have this in black and white.