Photo of Bexis

This post is not from the Dechert side of the blog.

The United States Supreme Court has said it – the test for implied preemption under 21 U.S.C. §337(a) (the FDCA’s no-private-enforcement provision) is whether the purported state-law cause of action would exist even in the absence of the FDCA/FDA: Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341, 353 (2001) (preemption applies to “claims exist solely by virtue of the FDCA disclosure requirements” and to all claims where “existence of these federal enactments is a critical element”).  So have federal courts of appeals.

If the claim would not exist in the absence of the FDCA, it is impliedly preempted. In other words the conduct on which the claim is premised must be the type of conduct that would traditionally give rise to liability under state law − and that would give rise to liability under state law even if the FDCA had never been enacted.

Loreto v. Procter & Gamble Co., 515 F. Appx. 576, 579 (6th Cir. 2013) (citations and quotation marks omitted). Accord Caplinger v. Medtronic, Inc., 784 F.3d 1335, 1339 (10th Cir. 2015) (“§337(a) preempts any state tort claim that exists ‘solely by virtue’ of an FDCA violation”); Perez v. Nidek Co., 711 F.3d 1109, 1119 (9th Cir. 2013) (preempting a “fraud by omission claim [that] exists solely by virtue of the FDCA  requirements”) (citation and quotation marks omitted); Lofton v. McNeil Consumer & Specialty Pharmaceuticals, 672 F.3d 372, 379 (5th Cir. 2012) (following Buckman; “tort claims are impermissible if they existing solely by virtue of the FDCA disclosure requirements”).

Nonetheless, courts persist, in generic drug product liability litigation, to ignore the relevant preemption test when it comes to “duty to update” claims. In such cases, if there were no FDCA, and thus no FDA, then the FDA couldn’t be telling anybody to update anything to conform to anything else.  The latest example of this logic-defying trend is the New Jersey Supreme Court’s recent decision in In re Reglan Litigation, ___ A.3d ___, 2016 WL 4420637 (N.J. Aug. 22, 2016) (we’ll call it “NJ Reglan” since there are various other Reglan-named decisions out there).

Yes, it is an arguable violation of the FDCA’s (Hatch-Waxman) “sameness” requirement for a generic manufacturer not to update its label to match changes approved by the FDA to the “reference” drug (usually but not always the innovator) labeling, although there were apparently no deadlines for doing so at the time of the first disputed label update.  Id. at *14 (citing 2000 FDA guidance document – having no legal effect – recommending that generic manufacturers should “routinely monitor” the FDA’s website for label updates and include them “at the very earliest time possible”).

But forget Buckman for a moment.  What kind of vague, squishy standard is that?

It’s certainly not one that state law has any business adopting.  First of all, state negligence per se law has never replaced the general tort standard of care with anything, such as an FDA guidance, that lacks force of law.  E.g., Restatement (Third), Products Liability §4 (1998); Restatement (Second) of Torts §§285-288C (1965) (both limited to statutes, regulations, ordinances, etc that were legally enforceable).  We discussed this state-law based limitation on negligence per se in detail here, and Bexis’ book (see §4.02[3][b]) cites a raft of precedent for this proposition.  We don’t know of any state that imposes liability for the violation of something – such as FDA guidance documents – that even the entity that wrote it couldn’t legally enforce.  As stated in footnote 1 of the 2000 guidance document:

This guidance document represents the Agency’s current thinking on changes in labeling of approved abbreviated new drug applications (ANDAs) following revisions in the RLD’s labeling.  It does not create or confer any rights for or on any person and does not operate to bind FDA or the public.

U.S. Dep’t of HHS, FDA, “Guidance for Industry: Revising ANDA Labeling Following Revision of the RLD Labeling, at 4, n.1 (May 2000) (emphasis added).

That’s not all.  In a development not even mentioned in NJ Reglan opinion, Congress ditched the FDA’s loosey-goosey “go look for it and do it as fast as you can” updating standard right in the middle of the time period at issue in this litigation.  NJ Reglan involved “nearly 1000 individual lawsuits against over fifty brand-name and generic manufacturers.”  2016 WL 4420637, at *2.  The claimed failures to update generic labeling ranged between six months and 4½ years.  Id. at *3.  Allegedly these failures to update began in 2004 and (as to some defendants) lasted until 2009.  Id. at *15.

But in 2007, Congress amended the FDCA and enacted 21 U.S.C. §355(o)(4), concerning “labeling changes requested by [FDA].”  This provision imposed an actual notice requirement on the FDA.  “If [FDA] becomes aware of new safety information that [it] believes should be included in the labeling of the drug, [FDA] shall promptly notify the responsible person.”  Id. §355(o)(4)(A).  After receiving such notice, the “responsible person” must then file a supplement updating the labeling “within 30 days.”  Id. §355(o)(4)(B)(1).

So as of 2007, that’s what the FDCA actually required for this type of agency-driven label update – no more “watch our website” advice with no force of law.

Thus, not only did the pre-2007 FDA guidance document lack force of law, but Congress then rejected the procedure that the FDA had described in the 2000 guidance.  In what was actually legally enforceable (after 2007), Congress required actual notice to the responsible person before the 30-day obligation to update labeling could begin.  Thus, the situation in NJ Reglan was unlike any negligence per se-style substitution of regulatory standards for a state-law standard of care that we’ve ever seen.  The pre-2007 guidance-document standard not only had no force of law, but was found wanting and affirmatively repudiated by Congress when it chose to impose a legally enforceable label updating standard.

Desperate to avoid generic preemption after PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011), and Mutual Pharmaceutical Co. v. Bartlett, 133 S. Ct. 2466 (2013), the plaintiffs in NJ Reglan alleged that various generic makers of metoclopramide didn’t update their labels after the FDA approved changes to the “reference” (innovator or branded) label for Reglan in 2005 and again in 2009.  NJ Reglan, 2016 WL 4420637, at *2.  These changes strengthened warnings against long-term use of the drug (whether innovator or generic).  Id.

As discussed, prior to the 2007 enactment of §355(o)(4), the FDA was not in the habit of giving actual “notice” of labeling changes affecting generic drugs.  The NJ Reglan opinion relates:

Generic manufacturers have been given the means to learn of brand-name-labeling updates.  The Office of Generic Drugs in the Office of Pharmaceutical Science, Center for Drug Evaluation and Research, at the FDA has directed generic manufacturers to “routinely monitor the Labeling Review Branch Homepage … for information on changes in labeling.” The Office of Generic Drugs “[p]lace[s] monthly updates of approved labeling changes” for brand-name drugs with approved generic counterparts “on the Labeling Review Branch Homepage.” Ibid. “All approved labeling for [brand-name drugs] is [also] available from Freedom of Information Staff” at the FDA.

NJ Reglan, 2016 WL 4420637, at *8 (citation to FDA guidance document and footnote omitted) (emphasis added).  Thus, instead the affirmative notice required by Congress in 2007, the FDA was apparently putting label “updates” on its homepage and expecting (via “guidance” without force of law) generics to come looking for them.  Mere “means to learn” isn’t “notice” to anybody.

Without discussing either a federal- or state-law basis for doing so, the unanimous New Jersey Supreme Court readily found a “violation” (of exactly what, we’re not sure) on which to hang what it treated as a “common-law” claim (again, of what sort, we’re not sure – the opinion never mentions negligence per se).  The opinion took the presumption against preemption from Wyeth v. Levine, 555 U.S. 555 (2009) (now dead in express preemption cases) (2016 WL 4420637, at *9-10), stirred in the “New Jersey’s strong public policy of ensuring that manufacturers attach adequate warnings and instructions to prescription drugs” (derived from a statute originally enacted to restrict liability) (id. at *10), and baked until a violation popped up:

Here, defendant generic manufacturers of metoclopramide tablets did not conform their labeling to that of the brand-name drug and therefore were in violation of the FDCA’s sameness requirement. . . . No law prevented defendants from giving the same warnings that appeared on the labeling of the brand-name drug − the warnings that plaintiffs contend the [N.J. product liability statute] required. Defendants did not have to violate federal law to comply with state law.

Id. at 11.

Ironically, NJ Reglan declared that “[t]his case drives home the point made in [Levine] that the FDA does not have the resources to monitor the labeling of thousands of drugs after they are marketed,” id., when the asserted “violation” involved various generic manufacturers allegedly not ferreting out label changes under an informal FDA procedure rejected by Congress in favor of requiring affirmative FDA notice.  Nothing in NJ Reglan indicates that, either before or after 2007, the FDA ever gave such affirmative notice of label changes to any of the defendants, as §355(o)(4) contemplates.  The FDA itself couldn’t enforce that guidance, so how can state law do so?

NJ Reglan finished off preemption by rejecting the applicability of the Buckman principle discussed at the outset of this post.  “The present case is different from Buckman because, here, the ‘critical element’ to plaintiffs’ claims is not defendants’ violation of the FDCA, but defendants’ failure to give adequate warnings about the prolonged use of metoclopramide.”  2016 WL 4420637, at *12.

If only that were so.  Instead, the so-called “failure to give warnings” extends solely, and no further than, the alleged failure to “update” in accordance with the 2000 FDA guidance document:

Although generic drug labeling is required to be the same as that of the brand name under federal law, defendant generic manufacturers, apparently, did not update their labeling “at the very earliest time possible” in accordance with the directive of the [FDA]. . . . To be sure, to avoid a clash with Mensing and Hatch-Waxman, plaintiffs may not contend that defendant generic manufacturers had a duty to provide warnings beyond those that the FDA approved for the brand name.

Id. at *14-15.  So, even though NJ Reglan asserts that “Plaintiffs’ claims do not ‘exist solely by virtue of’ a federal regulatory scheme” under Buckman, id. at *15, the result is indistinguishable – the New Jersey cause of action for failure to update was defined by the FDA’s informal pre-2007 guidance (even after 2007??), enabled by that same federal duty (even though the FDA itself couldn’t enforce it), and extends no further than the federal duty.

If it quacks like a duck….

We can only repeat something we’ve said many times before, strange things happen – particularly in state court – when preemption and personal injury cross paths.  And this one really is strange, if not utterly singular:  a state-law duty to update predicated on a federal guidance document not only without force of law ab initio, but then expressly repudiated by Congress in favor of a legally-binding standard that, as far as we can tell from the opinion, no defendant ever violated because the required FDA notice was never given.

The ultimate outcome, though, probably doesn’t matter much. This litigation is just another New Jersey litigation make work project.  The court relied heavily, id. at *13-14 on Fulgenzi v. PLIVA, Inc., 711 F.3d 578 (6th Cir. 2013) (discussed here), another Reglan-related case, but guess what?  The plaintiff lost in Fulgenzi (discussed here) because the prescribing physician never read, let alone relied on, the allegedly inadequate generic warnings:

The undisputed facts in the record establish that plaintiff’s physicians did not ever read, let alone rely on, [defendant’s] inadequate 2004 warning. As set forth above, proximate cause is an essential element that plaintiff must establish in order to prevail on her Ohio failure-to-warn claim. . . .  [A]ny presumption in favor of proximate cause resulting from its inadequate warning is rebutted by the fact that plaintiff’s prescribing physicians cannot say that they ever read [defendant’s] package insert for metoclopramide − adequate or otherwise. Indeed, the inadequacy of a warning cannot be the proximate cause of a plaintiff’s injuries if the user of the product failed to read the warnings accompanying the product. Even if such a warning were adequate, it could not prevent the harm if the user did not read the warning.  Consequently, an inadequate warning in a prescription package insert cannot be the proximate cause of a resulting injury if the physician did not read the insert prior to prescribing the medication.

Fulgenzi v. PLIVA, Inc., 140 F. Supp.3d 637, 649-50 (N.D. Ohio 2015) (citations and quotation marks omitted).  That’s the law just about everywhere, as we discuss here (50-state survey of failure-to-read-warning cases).  Physicians not reading the warnings at issue is a particularly strong obstacle to failure to update claims involving generic drugs – because in most cases doctors don’t prescribe the generic product at all, rather the pharmacy substitutes a generic for an innovator product, or selects one of several available generics without any input from the prescriber.  Doctors won’t be reading labels of drugs that they didn’t actually prescribe.

There are several other reasons why duty to update claims are, frankly, lousy claims for plaintiffs to pursue. We listed them here.  They range from medical causation, to risks being well-known in the medical community, to product identification.  We’ve never seen a plaintiff assert a failure-to-update claim except where forced to by the force majeure of the Supremacy Clause.  Even pleading the elements of such a case – particularly causation − will be a challenge going forward.  See, e.g., Brinkley v. Pfizer, Inc., 772 F.3d 1133, 1138 (8th Cir. 2014); Strayhorn v. Wyeth Pharmaceuticals, 737 F.3d 378, 399-400 (6th Cir. 2013) (both dismissing update-related claims for failure to plead causation).

So these “nearly 1,000” cases will no doubt amount to busy work for lawyers – at tremendous expense to the defendants and to the justice system as a whole. We doubt that even 5% of them (less than 50) will survive summary judgment.  We further doubt that most of these cases will even get that far.  Instead, we predict that they will either be dismissed (if they can’t pass the red-face test on causation) or else will be settled for some portion of whatever nuisance value that NJ Reglan has conferred upon them.