We understand that we write a lot about federal preemption. You might even be rolling your eyes at yet another post on this most powerful of defenses, but we just can’t help ourselves.  Federal regulation runs deep in the drug and medical device world, and the possibility that federal law might preempt state-law claims is an ever-present possibility.  To be sure, there is no shortage of litigation alleging that drug and device companies have violated state law duties, so we don’t want to overstate matters.  Regardless, federal preemption remains the bright shiny object from which we cannot avert our gaze.

Federal preemption demands our attention also because it is so often misunderstood and/or misapplied. We have said at times that some judges are hostile to preemption, and that might be true for some of the obviously result-oriented opinions that we review.  It is more likely that courts confront the many branches of preemption and the spider web of cases applying preemption in its various flavors, and they just get it wrong.  God only knows that the authorities are not always clear.

This is a long lead-in to a recent case that got preemption wrong on so many levels that we’re not completely sure what to make of it. In Tryan v. Ulthera, Inc., No. 2:17-cv-02036, 2018 U.S. Dist. LEXIS 140111 (E.D. Cal. Aug. 17, 2018), the plaintiffs sued the manufacturer of a medical device that uses ultrasound to tighten skin and reduce wrinkles on the face, neck, and chest.  The FDA cleared the device for marketing through the 510(k) premarket notification process, which is different from the more rigorous premarket approval (or “PMA”) process that some other devices go through.  The plaintiffs’ beef was that the manufacturer represented that the device was “FDA approved” (as opposed to “FDA cleared”) in violation of FDA regulations. Id. at **2-4.

The case seems fairly straightforward. Plaintiffs based their claims on an alleged violation of the Food Drug & Cosmetic Act (“FDCA”) and its regulations.  Moreover, the plaintiffs did not allege any complication with their procedures, or even that the device did not work exactly as it was supposed to.  Instead, they alleged that had they known the device was not FDA approved, and that it was merely FDA cleared, they would have paid less for the therapy or considered alternative treatments. Id. at *4.

Whatever you think of those allegations and the likelihood that any consumer would draw any distinction between “FDA cleared” and “FDA approved,” these plaintiffs were suing because of a violation of the FDCA, which runs directly into implied preemption under Buckman and 28 U.S.C. § 337(a).  Whether the device was FDA “approved,” “cleared,” or whatever, without the FDCA, there would be no FDA action at all. Buckman is the Supreme Court case that held that state-law claims conflicting with the FDCA are impliedly preempted, and Section 337 mandates that only the federal government can enforce the FDCA.  Whenever a purported state-law cause of action includes as a “critical” element some sort of FDCA violation, it runs afoul of § 337(a) and is therefore impliedly preempted under Buckman. Buckman involved a 510(k) device, so device categories do not affect implied preemption.

The manufacturer therefore justifiable moved to dismiss on implied preemption, and in denying the motion, the district court got off on the wrong foot, and it got worse from there. The district court began its discussion with the express preemption clause of the Medical Device Amendments, which expressly preempts any state law requirement which is “different from or in addition to” federal requirements. Id. at *10.  This is a valid legal principle as far as it goes, but it is completely irrelevant to this case:  The defendant did not assert that federal law expressly preempted the plaintiffs’ claims.  The defendant asserted implied preemption, under which federal law preempts state law where the state law conflicts with federal law or stands as an obstacle.

You might ask what difference that would make? Well, it makes a big difference, because in its next breath, the district court invoked the questionable and often-misunderstood “parallel claim” exception to preemption.  As the “exception” goes, because the MDA’s express preemption clause preempts state-law requirements that are “different from or in addition to” federal requirements, claims based on state-law duties that “parallel” federal requirements are not preempted.

We can debate whether the “parallel claim” exception is a good idea or whether federal authorities actually support its existence. One thing, however, is certain:

There is no parallel claim exception to implied preemption.

The district court’s introduction of express preemption at the outset of its opinion was therefore no harmless error. It started the opinion down a legal path from which the opinion would never recover, as the court forced the plaintiffs’ allegations into an “exception” that does not exist.  For example, the district court noted with approval that the plaintiffs alleged that California law “incorporates and mirrors the FDCA.” Id. at *11.  In the context of implied preemption, that observation alone should have been the death knell of this case because it acknowledged in plain terms that the plaintiffs were suing to enforce the FDCA.  We know from Buckman and section 337(a) that such actions are not allowed.  The “mirroring” of California law matters only in food cases, to which a statutory exception to section 337(a) applies.  That exception does not apply to medical devices (or drugs), which we detailed here.

Unfortunately, the district court misapprehended this point again and again. It noted the “linkage” between state and federal law and cited the plaintiffs’ use of federal “regulatory and statutory violations” as the predicates for their state-law claims. Id. The court noted that California statutes adopt federal regulations as state-law standards. Id. The court observed that California law prohibits false and misleading advertising.  And why is it misleading to say a Class II medical device cleared through the premarket notification process is “FDA approved”?  Because such a statement allegedly violates federal regulations.  All paths lead back to the plaintiffs suing to enforce the FDCA, which is the epitome of implied preemption.

The district court also wrongly cited federal law in support of its opinion. It relied on Medtronic Inc. v. Lohr, 518 U.S. 470 (1996), to describe the FDA’s 510(k) process. Id. at **5-6.  The court certainly is not alone in citing Lohr on this point, but as we have explained here and here, Lohr understates and marginalizes the 510(k) process’ focus on safety and efficacy and was outdated even before it was published.  The district court also stated that “FDCA regulations appear to directly preclude preemption,” but once again the cited regulation defines the scope of the Medical Device Amendment’s express preemption provision.  That provision did not apply, nor did the defendants argue that it should. Id. at *13.

Finally, the court observed that 510(k) clearance does not preempt “state statutes of general application.” Id. at **13-14.  Of course the mere fact of 510(k) clearance does not create a blanket of preemption over “statutes of general application,” and we highly doubt the defendant argued that it did.  However, when the plaintiffs is claiming that a medical device manufacturer violation a state statute of any kind because its conduct violated the FDCA, that’s implied preemption, and that is exactly what the plaintiffs alleged here.

In the end, we have a class action alleging that “FDA approved” and “FDA cleared” are different enough to consumers that the law should require awarding damages to patients who allege no injury and for whom the device may have worked exactly as advertised. We are not so sure, nor are we sure that this order will survive an appeal.

It took a while for courts to catch on that implied preemption in drug cases depends on whether the plaintiffs can present “newly acquired evidence” of a relevant risk, but the argument seems to be gaining some traction. The first case to recognize the “newly acquired evidence” argument was the First Circuit’s Marcus v. Forest Pharmaceuticals opinion in 2015, which we covered here.  Since then, we have covered the topic at some length in connection with orders coming out of the Eliquis MDL in New York, which you can revisit here, here, here, and here.  These posts report on a series of orders in which the district judge dismissed multiple cases because federal law impliedly preempted the plaintiffs’ claims.

A district judge has now issued a similar dismissal ruling in a prescription drug case in the Northern District of Alabama, McGee v. Boehringer Ingelheim Pharm., Inc., No. 4:15-cv-2082, 2018 WL 1399237 (N.D. Ala. Mar. 20, 2018).  Here is the gist of it.  We all understand that Wyeth v. Levine opened the anti-preemption door by recognizing that an innovator drug manufacturer could sometimes change its label without the FDA’s pre-approval through the Changed Being Effected (or “CBE”) process.  Because that allowed the manufacturer, under some circumstances, to change its label to accommodate state law without running afoul of federal law, implied preemption did not necessarily apply.  Then came PLIVA v. Mensing, which held that federal law impliedly preempted state-law failure-to-warn claims against generic drug manufactures because generic manufacturers cannot use the CBE process, and therefore cannot change their labels without pre-approval.  Because generic manufacturers cannot change their labeling to comply with state law without violating federal law, state law claims must give way.

In McGee, the district court joined others in recognizing that even an innovator manufacturer cannot use the CBE process unless there is “newly acquired evidence” of a causal association between a risk and the drug.  In that event, the innovators cannot change their drug labeling without the FDA’s pre-approval and implied preemption applies.  Id. at **3-4.  “So, if a plaintiff can allege the existence of newly-acquired information that supports a labeling change under the CBE regulation, and if the manufacturer subsequently fails to show by ‘clear evidence’ that the FDA would not have approved a change to the label, then a failure-to-warn claim will survive the manufacturer’s preemption defense.” Id. at *4.

In other words, before we even get to Wyeth v. Levine’s “clear evidence standard,” the plaintiff has to plead facts establishing that the manufacturer had newly acquired information and that the new information provides reasonable evidence of a causal association between the alleged risk and the drug.

The plaintiff in McGee failed.  He alleged that the product manufacturer failed to warn adequately regarding the risk of diabetic ketoacidosis, or “DKA.” McGee, 2018 WL 1399237, at *1.  But the complaint was ambiguous on what the manufacturer knew about that condition and when the manufacturer knew it. Id. It all came down to timing.  The plaintiff alleged that there were reports of DKA before the FDA approved the defendant’s drug, but reports that pre-date drug approval obviously are not “newly acquired” information.  In addition, any allegation that the manufacturer should have alerted the FDA about the DKA risk before approval is a fraud-on-the-FDA claim preempted under Buckman. Id. at *4.

The plaintiff also alleged that the FDA issued a warning with reference to adverse events after the plaintiff allegedly experienced DKA, but “common sense dictates that any information [the manufacturer] obtained after January 17, 2015, when Mr. McGee suffered DKA, is irrelevant to Mr. McGee’s claims because [the manufacturer] could not have used that information to change its label and prevent Mr. McGee’s injury.” Id.

In theory, the plaintiff could have stated a claim with reference to information newly acquired after approval but before he used the product.  “But Mr. McGee lumps together claims and facts from before, during, and after this brief time period” and “[w]ithout clarification, the court cannot tell when Mr. McGee alleges [the manufacturer] had what information.” Id.  Alas, the court granted leave to amend.  But the district court in the Eliquis cases gave the plaintiffs additional chances to plead too, and they were unable to come up with the facts.

The plaintiff in McGee may have similar difficulty.  We know from Eliquis that dumbing down the complaint and making the allegations more vague will not work, and the district court in McGee has already faulted this plaintiff for his imprecise pleadings in any event.  The court had particular distaste for allegations that were “nothing but a useless set of legal conclusions.” Id. at *5.  The court continued:

Worse, the complaint spews these legal conclusion with few supporting facts that would assist [the defendant] in understanding what the company did wrong. This type of “shotgun” pleading fails to plead a cognizable cause of action.  If Mr. McGee chooses to file an amended complaint, he must eliminate the unnecessary and unhelpful “shotgun” assertions and claims.

Id. (citations omitted). These are pretty strong words, but nothing beyond what any court should expect under Rule 8 and TwIqbal.  The takeaway is that the “newly acquired evidence” argument is well-grounded in the law and supported now by substantial precedent.  Moreover, plaintiffs will not always have the facts to get around it.

The Seventh Circuit taught us recently that the letter “A” is a powerful thing. Of course, we already knew that a well-placed A can convert the ordinary (“typical”) into the extraordinary (“atypical”), the melodic (“tonal”) into the dissonant (“atonal”), and the virtuous (“moral”) into the indifferent (“amoral”).  Adding a single A to a Scrabble board can result in a seriously high-value word, provided you can manage to place it on a triple word score appended to a word that already has high value letters, like H or Y, or if you are really lucky, X.

For today, A stands for “Abbreviated”—as in Abbreviated New Drug Application, or ANDA. It’s important because the Seventh Circuit held in Guilbeau v. Pfizer, Inc., No. 17-2056, 2018 WL 476343 (7th Cir. Jan. 19, 2018), that implied preemption of a failure-to-warn claim under Pliva v. Mensing depends not on a drug’s colloquial description as “generic” or “brand name,” but rather on the nature of the drug’s approval process.  If a drug is approved through an ANDA, federal regulation of drug labeling preempts state-law failure to warn claims—even if the drug is technically the “Reference Listed Drug.”

This ruling makes perfect sense, and here is how it works. In Guilbeau, the plaintiffs alleged that manufacturers of a testosterone replacement therapy product called Depo-T failed adequately to warn regarding cardiovascular events. Id. at *1.  Depo-T, however, had a slightly unusual regulatory history.  When a manufacturer applies for drug approval under an ANDA, it has to demonstrate that the drug has the same active ingredients, effects, and labeling as a predecessor drug that the FDA has already approved. Id. at *2.  The predecessor drug is called the Reference Listed Drug. Id.

The Reference Listed Drug is usually the innovator product, approved through the NDA process and often referred to as a “brand-name” drug. The ANDA drug is often called “generic.”  If, however, an original innovator drug has been discontinued, the FDA can designate the remaining market-leading drug to take its place as the Reference Listed Drug. Id. Something like that (but not exactly) happened to Depo-T.  Because of unique circumstances surrounding the timing of the drug’s approval, the FDA classified Depo-T as an ANDA-approved drug, but the product also was designated the Reference Listed Drug for its kind of testosterone injection. Id. at *4.

So which is it? “Generic” or “brand name”?  It turns out those terms are neither useful nor dispositive when applying implied preemption.  Or, as the Seventh Circuit observed, “These colloquial terms are not quite precise enough for our purposes in this case.” Id.  The relevant distinction is NDA-approved versus ANDA-approved.  That distinction determines whether a drug manufacturer can unilaterally update its labeling through the changes being effected (or “CBE”) process and thus potentially lose the protection of implied preemption recognized in Pliva v. Mensing.  The Seventh Circuit explained it this way:

[D]espite potentially confusing references to brand-name and generic drugs—recall that the relevant FDA terms are NDA-approved and ANDA-approved—Mensing itself unambiguously refers to the lines drawn in the drug approval process as determining access to the CBE regulation. Mensing concludes that while NDA holders may use the CBE regulation to add warnings, ANDA holders (like [the manufacturer] here) may not. . . .

Despite their occasional use of these terms, the Supreme Court, Congress, and the FDA all agree that the meaningful difference is found in approval process classifications, not shorthand terms like brand-name and generic.

Id. at *6. The key point for the Seventh Circuit was that the CBE regulation is unavailable for products approved through the ANDA process.  And that goes for every ANDA product, even if it is the Reference Listed Drug—like Depo-T.  Because ANDA product manufacturers cannot change their labels unilaterally, it is impossible to change a label in response to state tort law without violating the federal duty of “sameness” that applies to all ANDA products.  That is Pliva v. Mensing implied preemption. Id. at **4-5.

The Guilbeau plaintiffs tried to avoid preemption by arguing that, because Depo-T was itself the Reference Listed Drug, the manufacturer’s duty was to have labeling the same as its own labeling.  According to the plaintiffs, that means the manufacturer could change the label all it wanted to and it would still be the “same” as itself. Id. at *5.  We appreciate the metaphysical nature of this argument:  How can something ever be different from itself?  But the answer is pretty easy.  The duty of “sameness” means that ANDA holders must match the labeling for the Reference Listed Drug as approved by the FDA, “not whatever the RLD’s manufacturer currently thinks would be best.”  Id. at *7.  In other words, “The statutes, the regulations, and the Mensing opinion do not draw the distinction plaintiffs advocate:  a difference in abilities and responsibilities between RLD ANDA holders and other ANDA holders.” Id.

In the end, ANDA drugs that are also Reference Listed Drugs are subject to a “duty of sameness indistinguishable from that of all other ANDA drugs.” Id. at *8.  They all must show that their labeling is the same as the approved labeling. Id. Thus, because Depo-T’s manufacturer “may not unilaterally change the FDA-approved language on Depo-T’s label,” a lawsuit under state law that seeks damages for the manufacturer’s failure to do so is preempted by federal law. Id.

One more important note: The Seventh Circuit held that the plaintiffs were not entitled to conduct further discovery.  The district court decided this case on a motion to dismiss, and because “preemption is a legal question for determination by courts . . . , discovery of facts may not be as vital to this inquiry as it could be to others.” Id. at *10.  With recent opinions from the Third Circuit and Ninth Circuit treating preemption as a factual issue on which discovery might be appropriate (discussed here  and here), the Seventh Circuit’s holding on discovery is most welcome.  Preemption can and should be decided on the pleadings in many instances, including the circumstances presented here.  The district court in Guilbeau correctly did exactly that, and the Seventh Circuit gave the order an A grade.

Today is the birthday of Gilbert O’Sullivan, who scored a hit back in 1972 with “Alone Again, Naturally,” the saddest song we can think of this side of Albinoni’s Adagio.  That is fitting, given our postscript.

 

December 1 is also the birthday of Sarah Silverman and Bette Midler, two women who consistently bring smiles, so we’ll discuss a good case out of the Eleventh Circuit, though probably not good enough to crack our upcoming top ten list.  In Tsavaris v. Pfizer, Inc., 2017 WL 5593488 (11th Cir. Nov. 21, 2017), the plaintiff claimed that she developed breast cancer after ingesting a generic version of a hormone replacement drug.  She sued both the brand and generic manufacturers.  We don’t know what happened regarding the brand manufacturer (they should certainly have secured a dismissal), but we know that the district court dismissed the claims against the generic manufacturer on preemption grounds well-established in the Mensing and Bartlett SCOTUS decisions and the Guarino Eleventh Circuit decision.  The district court entered final judgment for the defendant.

 

The plaintiff filed an amended complaint against the generic manufacturer, asserting that she would not have been harmed had the manufacturer not “failed in its federally mandated duty” under 21 U.S.C. section 355 to notify the FDA of certain scientific studies relevant to the cancer risk.  The district court denied the amendment on both procedural and substantive grounds, and the Eleventh Circuit affirmed, holding that the district court had not abused its discretion.  The procedural grounds were enough to bar the amendment: the plaintiff identified no newly-discovered evidence or manifest error of law, as required by Federal Rule of Civil Procedure 59(e).  That is all well and good.

 

But it is the substantive decision that is more interesting to us.  The plaintiff was complaining of a violation of a federal reporting duty owed to a federal agency, not to her.  More specifically, she premised her complaint on a provision of the federal Drug Price Competition and Patent Term Restoration Act that requires companies to submit “data relating to clinical experience and other data or information … about the safety, effectiveness, or labeling of its drug” to the FDA.  Because the plaintiff was seeking “to enforce a duty owed to a federal agency and her cause of action would not exist in the absence of that duty, her proposed second amended complaint is preempted.”

 

The Tsavaris decision was not selected by the Eleventh Circuit for publication in the Federal Reporter, perhaps because this was a pro se appeal.  But the proposed amended complaint was drafted “with the assistance of counsel,” and is fairly typical of the way some plaintiff lawyers seek to evade preemption.  Thus, the Eleventh Circuit’s finding that the amendment would have been futile provides some comfort to practitioners on the right side of the v.

*******************************************************

We mourn the passing earlier this week of beloved Reed Smith colleague Chuck Pagliotti.  Chuck was what all of us in the legal profession should aspire to be – a problem solver.  Chuck was a genius when it came to fixing messes we made on our computers, iPhones, or anything else that could occasionally baffle us.  He always did so with patience and a smile.  We shared a love for anything relating to Star Wars.  Chuck was a sweet, gentle soul. Our grief exceeds our poor power of expression.

 

 

The Ninth Circuit filed a preemption opinion the other day that should help prevent the “foodification” of medical device litigation. That made-up word refers to the wasteful food-related class action litigation that has somewhat thrived in California.  The template is familiar:  Opportunistic plaintiffs’ lawyers find some aspect of food labeling that they claim is misleading—e.g., that a product is “wholesome” or “natural”—and they recruit a plaintiff to say that he or she would not have purchased the food or would have paid less for it had truth been told.  We wrote more than two years ago about a class action claiming that Hershey chocolates made misleading “healthy diet claims.”  Who would possibly rely on “healthy diet claims” for chocolate treats?  That class action obviously had no legs, but others have gotten by.

Attempts to extend this business model to medical devices have failed, and one reason is federal preemption. That is our takeaway from DeBons v. Globus Medical, Inc., No. 14-5645, 2016 WL 4363171 (9th Cir. Aug. 16, 2016), where the Ninth Circuit affirmed the district court’s order dismissing a putative class action involving bone putty, an allograft tissue product that the FDA regulates as a medical device. See the district court’s dismissal order, DeBons v. Globus Medical, Inc., No. 2:13-cv-08518, 2014 WL 12495351 (C.D. Cal. Aug. 8, 2014).  Taking a page from the food playbook, the plaintiff claimed that the device was “not in fact FDA approved,” which we take to mean that product was “cleared” through the 510k process rather than formally “approved.”  He alleged therefore that “had made it known that the product was not FDA approved,” he would not have paid for it. Id. at *1.  He asserted a raft of consumer fraud claims, including a claim under California’s broadly written Unfair Competition Law, and purported to represent a class of other bone putty patients. Id. Bear in mind that the plaintiff alleged no surgical complication; the device evidently worked just fine.  This was strictly a play for a price refund.

Here is where it gets interesting. The main reason why medical device class actions are so rare these days is that medical treatment unavoidably presents individualized issues, making class treatment impossible.  Another reason is Section 337(a) of the FDCA, which gives the FDA exclusive authority to enforce the Act’s provisions.  This provision impliedly preempts any private action right of action filed because a device manufacturer’s conduct allegedly violates the FDCA. See Buckman Co. v. Plaintiffs’ Legal Comm., 531 U.S. 341 (2001).  When combined with the Medical Device Amendments’ express preemption provision, would-be plaintiffs face the “narrow gap” between implied preemption and express preemption on which we often comment (such as here). On the one hand, the MDA’s express preemption provision bars states from enforcing requirements “different from or in addition to” federal requirements.  21 U.S.C. § 360k(a).  On the other hand, private litigants cannot enforce the FDCA.  A plaintiff therefore “must be suing for conduct that violates the FDCA (or else his claim is expressly preempted . . . ), but the plaintiff must not be suing because the conduct violates the FDCA (such a claim would be impliedly preempted under Buckman).” DeBons, 2016 WL 43463171, at *1 (quoting Perez v. Nidek Co., Ltd., 711 F.3d 1109, 1120 (9th Cir. 2013)).

Continue Reading Another Reason Why Medical Device Class Actions Don’t Work