Yesterday we regaled you with the first half of the decision in Cornett v. Johnson & Johnson, ___ A.2d ___, 2010 WL 2867811 (N.J. Super. A.D. July 23, 2010), concerning the court’s decision to apply the law of the plaintiff’s residence to the statute of limitations in a prescription drug case. We explained how that was a big deal, given the number of out-of-state plaintiffs who have flocked to New Jersey in recent years to take advantage of the state’s reputation for liberality in tort matters – a reputation that, as we pointed out before, isn’t entirely deserved.
Today we tackle the other half of the Cornett decision – concerning preemption. Preemption loomed large in Cornett because the product involved, a drug eluting coronary stent, is a Class III premarket approved (“PMA”) medical device. That means, of course, to anyone who follows this blog with any regularity at all, that the plaintiffs’ personal injury claims ware subject to express preemption under Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), and – depending on the allegation – maybe implied preemption under Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001), as well.
There are two ways of looking at this aspect of Cornett, since the Appellate Division found most, but not all, of the claims preempted – whereas the Law Division of the Middlesex vicinage (that Jerseyspeak for “trial court in Middlesex County”) had found the plaintiffs’ entire cases (the ruling applied to several dozen suits) preempted.
On one hand, there’s the glass-is-half-full approach: that getting an affirmance of a dismissal on the pleadings that knocked out large portions of these cases is a big step forward, given the notable liberality that many prior Appellate Division panels have shown towards expansive tort liability. See Smith v. Alza Corp., 948 A.2d 686 (N.J. Super. A.D. 2008); Sinclair v. Merck & Co., 913 A.2d 832 (N.J. Super. A.D. 2007), rev’d, 948 A.2d 587 (N.J. 2008); International Union of Operating Engineers Local No. 68 Welfare Fund v. Merck & Co., 894 A.2d 1136 (N.J. Super. A.D. 2006), rev’d, 929 A.2d 1076 (N.J. 2007); Rowe v. Hoffmann-La Roche Inc., 892 A.2d 694, (N.J. Super. A.D. 2006), rev’d, 917 A.2d 767 (N.J. Mar 29, 2007); In re Lead Paint Litigation, 2005 WL 1994172 (N.J. Super. A.D. Aug 17, 2005), rev’d, 924 A.2d 484 (N.J. Jun 15, 2007); James v. Arms Technology, Inc., 820 A.2d 27 (N.J. Super. App. Div. 2003) (overruled by Lead Paint); Staub v. Eastman Kodak Co., 726 A.2d 955 (N.J. Super. App. Div. 1999); Delgozzo v. Kenny, 628 A.2d 1080 (N.J. Super. App. Div. 1993).
On the other hand, the glass-is-half-empty approach laments that any of the plaintiffs’ allegations escaped dismissal and laments especially that New Jersey has not adopted Twombly/Iqbal and still allows vague, anything goes pleadings.
Both approaches probably have merit.
The preemption facts were not exactly easy from a defense perspective. First, the stent was used in an off-label manner – it was not approved for use in diabetic patients and at least the lead plaintiff of the group was a diabetic. Plaintiffs made by-now-familiar allegations of illegal off-label promotion. Second, the FDA issued a warning letter asserting violations of Good Manufacturing Practices (“GMPs”) that resulted in “nonconforming” stents being sold. Third, plaintiffs alleged that the defendants had not complied with some of the post-marketing surveillance conditions of the PMA, and had thus did not reveal an increased risk of the condition (thrombosis) that the lead plaintiff, at least, claimed to have. Cornett, 2010 WL 2867811, at *2-4.
Faced with Riegel and Buckman preemption, the plaintiffs made two major arguments. First, they argued that, since there was an off-label use, there wasn’t any preemption at all – their theory being that the FDA neither considered nor approved the off-label use. Second, they tried to ram all their claims through the narrow preemption exception for “parallel” violation claims.
In a ruling that is Exhibit A for the glass-is-half-full approach, the court rejected the off-label use = no preemption argument. Initially the court pointed out, as many other courts have, that off-label use is perfectly legitimate under the Medical Device Amendments:
Two decades after it enacted the MDA, Congress clarified that the off-label use of devices was not illegal per se, by denying the FDA any power “to limit or interfere with the authority of a health care practitioner to prescribe or administer any legally marketed device to a patient for any condition or disease within a legitimate health care practitioner-patient relationship.” 21 U.S.C.A. §396. It also provided a safe harbor for manufacturers of Class III devices to “disseminate” to health-care providers peer-reviewed articles or “reference publications” “concerning the safety, effectiveness, or benefit of a use not described in the approved labeling[.]” 21 U.S.C.A. §§360aaa, 360aaa-1.
Cornett, 2010 WL 2867811, at *11.
Simply because an off-label use is not itself FDA approved does not mean that the PMA process, and other aspects of the FDCA, do not regulate off-label use.
While plaintiffs are correct that the FDA does not literally “approve” off-label uses, because approval would axiomatically make them label uses, they are incorrect to argue that the absence of affirmative approval mandates the finding that no federal requirements exist. Federal requirements for off-label use manifestly exist, in the form of the safe harbor for their promotion, and the obligation to seek a supplemental PMA to add warnings to the label for off-label uses even when the manufacturer has no desire to promote them. Furthermore, plaintiffs’ argument presumes off-label use to be inherently suspect or substandard, which Congress rejected by enacting 21 U.S.C.A. §396 to protect off-label use.
Cornett, 2010 WL 2867811, at *17 (Buckman citation omitted). Indeed, Riegel itself was an off-label use case. Id. at *20. Thus, “we reject plaintiffs’ argument that . . . §360k(a) should not preempt any of their claims against defendants arising out of . . . off-label use.” Id.
Cornett joins the still small but growing list of cases rejecting the plaintiff-side argument that off-label use is incompatible with PMA preemption. See Wolicki-Gables v. Arrow International, Inc., 641 F. Supp.2d 1270, 1283-84 (M.D. Fla. 2009); Riley v. Cordis Corp., 625 F. Supp.2d 769, 777-79 (D. Minn. 2009); McMullen v. Medtronic, Inc., 2004 WL 2538642, at *4 & n.2 (S.D. Ind. Sept. 16, 2004), aff’d, 421 F.3d 482 (7th Cir. 2005).
The court’s categorical rejection of the plaintiffs’ categorical argument that off-label use precluded PMA preemption thus counts as a big thumbs up.
The Cornett decision also accurately assessed the preemptive impact of the FDCA’s express prohibition against private assertion of violation claims:
As for enforcement, all “proceedings for enforcement” of the FDCA must by brought “by and in the name of the United States.” 21 U.S.C.A. §337(a). The statute leaves “no doubt” that it prohibits private actions in which the only wrongdoing is violation of a federal law or standard. That prohibition requires a plaintiff to characterize his or her claims about a device as violations of state law, which 21 U.S.C.A. §360k or 21 U.S.C.A. §396 may then preempt.
Cornett, 2010 WL 2867811, at *11 (citation omitted). Thus, tort claims that “exist solely by virtue of the FDCA” are preempted because “they [are] in effect no more than per se claims for violation of a federal requirement.” Id. at *15. Truly “parallel” state-law claims do not need the alleged FDCA violation to exist:
[A] claim describes a “traditional” state law cause of action not simply because it contains required elements for the manufacturer’s conduct beyond the violation of a federal requirement, but rather because it would provide the required elements of a state cause of action even with no reference to federal requirements as the measure of the reasonableness or wrongfulness of the manufacturer’s conduct. . . . Regardless of how the plaintiff styles a state claim, if it is a claim that could not be articulated but for the existence of a federal requirement that was allegedly violated, it is functionally equivalent to a claim that is grounded solely on the federal violation, and is therefore impliedly preempted.
Id. at *16 (citations omitted, emphasis added). Thus, to avoid preemption a claimed “parallel” violation claim must navigate between Scylla and Charybdis. The claim must both parallel the federal requirements (and nothing more) while also being a traditional state-law claim not dependent upon the claimed FDCA violation:
[T]o avoid the statutory preemption of state requirements that vary from federal requirements, a state claim concerning PMA devices must be parallel to the federal requirements, meaning that the state requirements it embodies must not impose any duty on the manufacturer beyond satisfying the federal requirements. However, to avoid the implied preemption of claims outside the traditional areas of state regulation, a claim concerning PMA devices must also represent a traditional state cause of action that would impose a duty on the manufacturer even if there were no federal requirements at all to be referenced, much less violated.
Id. (emphasis added).
The court also found claims concerning the defendants’ alleged failure to comply with post-approval surveillance and study could be preempted. PMA imposes post-marketing duties to report and investigate adverse reactions, but affirmative label changes require additional FDA approval:
However, if such information indicates the need for an enhanced warning, the manufacturer must seek FDA approval before adding it to the label. Manufacturers have a safe harbor for promoting an off-label use, namely, distributing certain clinical study reports to physicians, as long as they also seek FDA approval for it.
2010 WL 2867811, at *16.
So the Cornett court got the big concepts right. But the devil’s in the details – or as some have said, “what the big print giveth, the fine print taketh away.”
Turning to the plaintiffs’ particular claims, Cornett acknowledged that PMA approval, and thus express preemption, “encompasses the device’s design, manufacturing methods, and label, including the information to be given to patients.” 2010 WL 2867811, at *16. The court held that there was nothing “parallel” about the plaintiffs’ design defect claim:
Because [New Jersey law] provides a different standard for the adequacy of the device’s design than the federal requirements, plaintiffs’ design defect claim is not “parallel” to them and is thus squarely within Riegel’s preemption holding.
Id. at *17. That was easy enough.
To avoid their manufacturing defect claim suffering the same fate, “plaintiffs limit their allegations . . . to what the FDA warning letter identifies as potential deviations.” Id. That concession prompted the court to hold that a proper “parallel” claim was pleaded, with the “manufacturing defect” being the violations of GMPs stated in the FDA’s letter. Id. As our device preemption scorecard indicates, that’s not all that unusual a ruling in and of itself.
But in federal court, that’s not the end of it. No, we can use Twiqbal to go after this sort of claim, either for vagueness (failure to plead what’s been violated), non-parallelness (to gin up a violation, the plaintiff interprets the regulation differently than the FDA), or irrelevance (the violation did not involve the plaintiff’s device). But unfortunately, Cornett is in state, not federal court. So the plaintiffs skate – at least for now – because it’s enough to “incorporate the FDA warning letter’s assertion of manufacturing defects,” and there’s no requirement to “directly connect the injury to the violations named in the FDA warning letter”; “could have” causation is enough. 2010 WL 2867811, at *18 & n.6.
The drip, drip, drip as the water empties from the glass reminds us, once again, why we’re such fans of Twiqbal. Well, at least the plaintiffs’ manufacturing claims are now limited to what’s in the FDA warning letter.
With respect to warning claims, there’s preemption – and maybe failure to state a claim – with respect to any and all pure failures to add more warnings to the label. Not only does Riegel preempt such claims, but the court indicates that they fail under the presumption of adequacy that New Jersey law affords to all FDA-approved warnings. “[A]bsent deliberate concealment or nondisclosure . . . compliance with FDA standards should be virtually dispositive.” 2010 WL 2867811, at *19 (quoting Perez v. Wyeth Laboratories, 734 A.2d 1245, 1259 (1999)).
Okay – but not okay. A “nondisclosure” claim, Cornett holds, isn’t preempted. “[W]hen the claim about the failure to warn for approved uses was combined with allegations of nondisclosure, it became a claim within a traditional area of state regulation that would have existed even in the absence of federal requirements.” 2010 WL 2867811, at *19. That’s peculiar, because, while citing Buckman, the court seemingly forgets that Buckman postdates the Perez language. Buckman, as we all know, holds that fraud on the FDA claims are preempted.
Thus, after Buckman, the Perez language is incomplete. The question “nondisclosure to whom?” becomes key, because non-disclosure to the FDA is just another way of saying fraud on the FDA. So, to whom? Unfortunately, here’s where Cornett comes up short, since the allegations sound to us like a warmed over fraud on the FDA claim:
While defendants supposedly withheld [post-marketing studies] from the FDA and the advisory panel, they allegedly distributed altered versions of them to physicians, to promote both approved and off-label uses with misrepresentations. . . According to plaintiffs’ complaint, submission of the studies as federal law required would have caused the FDA to improve the label, and any heightened warnings concerning approved uses would also have increased the safety of off-label uses.
2010 WL 2867811, at *19.
If it walks like a duck and quacks like a duck, chances are it’s a duck. Plaintiffs allege a claim of “deliberate nondisclosure,” id., to the FDA, and the causation element of their claim is expressly based upon the contention that, but for the nondisclosure, the FDA would have done something other than what it in fact has done. That has all the attributes of a fraud on the FDA claim, and it’s preempted under Buckman, regardless of what Perez might have said a couple of years earlier. No two ways about it, then, in this respect the Cornett decision blew it.
We’re back to half empty again
There’s also, apparently, a second failure to warn claim based upon promotion of off-label use. The court’s reasoning on that is essentially:
(1) New Jersey law limits warning claims to a product’s “intended purpose”;
(2) Off-label promotion places the use among a product’s “intended” uses;
(3) Off-label promotion is illegal, so a claim based upon off-label promotion would “parallel” FDA requirements;
(4) The violation did not have to involve fraud on the FDA, since “there were other ways in which promotion” could violate FDA regulations;
(5) Those “other ways” were “affirmatively represent[ation]” about “mortality differences” and “failure to include adequate warnings about the off-label use they were promoting.”
2010 WL 2867811, at *20.
Hmmm…. We’ve suggested before that false promotion of off-label use might survive as a “parallel” claim. But, while Cornett dances around the edges, there’s apparently no allegation that any “affirmative representation” was actually false. Otherwise, we think the court would have said so.
Instead of any allegation of actual falsity, we get, “[u]nder the liberal standard of review for a motion to dismiss, we conclude that plaintiffs’ allegation of off-label promotion without adequate warnings was a sufficient pleading.” 2010 WL 2867811, at *21. So, for lack of Twiqbal in New Jersey, the Cornett plaintiffs escape (at least for now) with an ill-defined “failure to warn by non-disclosure during off-label promotion” claim.
Cornett finishes up by addressing a few miscellaneous claims. It follows what is the majority view on express warranty – that is, “liability for statements in the FDA-approved label and other documentation is preempted,” but claimed warranties arising from “any statement that the FDA did not approve or mandate, is not preempted.” 2010 WL 2867811, at *21. We don’t have a problem with that, but what are the alleged express warranties here? Dunno. All we get is a “to the extent that” ruling that doesn’t really decide anything. Id. Since (as our cheat sheet demonstrates) one of the most frequent uses of Twiqbal is to require that the substance of claimed express warranties be pleaded, we’re once again left lamenting the difference between federal and state pleading requirements.
Implied warranty doesn’t exist under New Jersey law as a separate claim in product liability actions. 2010 WL 2867811, at *21.
Punitive damages are preempted under McDarby v. Merck & Co., 949 A.2d 223 (N.J. Super. A.D. 2008).
So it’s hard to say whether, with respect to preemption, we should consider Cornett to be half full or half empty. The design defect claim is gone, and the warning and manufacturing claims are trimmed back to specific and hard-to-prove violation related claims. Upon further development, one warning theory will almost surely be exposed as a preempted fraud on the FDA claim. From experience, we also doubt there’s any “there there” as to express warranty either. And no punies.
But one thing’s clear. The litigation’s going to be more than twice as large as it needed to be. What happens next is the expensive part, discovery – all over claims that, had they been in federal court, were in one way or another inadequately pleaded under Twombly/Iqbal.