Preemption

Subscribe to Preemption

Today we have another guest post by long-time friend of the blog, Dick Dean, and his colleague at Tucker, Ellis, Mike Ruttinger.  Regular readers will recall, that right after Sikkelee v. Precision Airmotive Corp., ___ F.3d ___, 2018 WL 5289702 (3d. Cir. Oct. 25, 2018), was decided, we blogged about the aspect of that decision that we thought was most directly relevant to drug/device litigation – the court’s rejection of tort claims based on failure to make reports to government agencies.  We briefly mentioned the remainder of the Third Circuit’s decision (which was actually by far the lengthier discussion), but didn’t spend much time on it.  In this post, Dick and Mike rectify that oversight.  As always, our guest bloggers deserve 100% of the credit (and any blame) for their discussion.

**********

The Third Circuit is having a bad year on preemption.  Its decision in In re Fosamax Products Liab. Lit., 852 F.3d 268 (3rd Cir. 2017), in which it held that it is for juries and not judges to determine whether there is “clear evidence” sufficient to meet the Wyeth v. Levine, 555 U.S. 555 (2009), standard for preemption in a failure-to-warn case, was accepted for review by the Supreme Court and is widely expected to be reversed.  [Editorial note – Fosamax ended up tied for the worst decision of 2017.  We hope to be rid of it in 2018.]  And now the Circuit has injected needless confusion into the test for impossibility preemption set forth in Levine’s follow-up case, PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011).  Mensing is familiar to many as the case that clarified the rule for determining when an impossible-to-resolve conflict between federal and state law preempts plaintiffs’ claims.  If the change the plaintiff seeks is one that requires prior approval or federal permission, then the claim is preempted.  Id. at 620 (“The question for ‘impossibility’ is whether the private party could independently do under federal law what state law requires of it.”) (emphasis added).  [Editorial note:  We call that the “independence principle.”]

The recent Third Circuit decision, Sikkelee v. Precision Airmotive Corp., No. 17-3006, 2018 WL 5289702, at *8 (3rd Cir. Oct. 25, 2018), is a wrongful-death case that originated from a 2005 airplane crash.  As the date suggests, it has been around for quite a while; this is the litigants’ second trip to the Third Circuit after a 2016 appeal [blogged about here] culminated in denial of a petition for certiorari and a remand for the Middle District of Pennsylvania to consider conflict preemption issues.  Specifically, the focus of Sikkelee became the design of the airplane carburetor.  The district court initially found the plaintiff’s defect claim to be barred by field preemption under the Federal Aviation Act because federal regulation of aviation is so extensive as to preempt the entire field of airplane design-related tort law.  Sikkelee v. Precision Airmotive Corp., 45 F. Supp. 3d 431 (M.D. Pa. 2014).  But the Third Circuit reversed that decision, suggesting that while field preemption does not apply, “the case law of the Supreme Court and our sister Circuits supports the application of traditional conflict preemption principles.”  Sikkelee v. Precision Airmotive Corp., 822 F.3d 680, 699 (3rd Cir. 2016).  Accordingly, the Third Circuit remanded with an opinion directing that the district court should consider the conflict preemption principles set forth in Mensing.

Given the Third Circuit’s lengthy discussion of Mensing in its 2016 opinion, what came next was quite a surprise.  The panel that issued the 2016 decision acknowledged the role that Mensing, as well as a subsequent decision, Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013), play in the conflict preemption analysis at length.  The court even honed in on the FAA’s “preapproval process for aircraft component part designs” as a key factor for the district court to consider in any conflict preemption analysis because the FAA would need to preapprove the alternate design that the plaintiff alleged as the basis for her lawsuit.  Sikkelee, 822 F.3d at 708 (“Thus, the reasoning of the Bartlett majority, 133 S. Ct. at 2473, 2480, and the consideration we must give to the FAA’s views under separation of powers principles, see Wyeth, 555 U.S. at 576-77, 129 S. Ct. 1187, lead us to conclude that the FAA’s preapproval process for aircraft component part designs must be accorded due weight under a conflict preemption analysis.”).  On remand, the district court followed the Third Circuit’s suggestion and found that the design-defect claim regarding the carburetor was indeed preempted because federal regulations required prior approval of the suggested design change.  Sikkelee v. AVCO Corporation, 268 F. Supp. 3d 660 (M.D. Pa. 2017).  But on October 25, that decision was reversed on appeal by a completely different panel of Third Circuit judges.  The new panel found no conflict preemption, applying the “clear evidence” test from Wyeth v. Levine rather than the Mensing prior approval test.  Specifically, the panel reasoned that “the nature of FAA regulations and Lycoming’s interactions with the FAA—including the changes it has made to its type certificate—demonstrate that Lycoming could have—indeed it had—adjusted its design.”  Sikkelee, 2018 WL 5289702 at *8.  For the defendant “to be entitled to an impossibility-preemption defense,” the court reasoned, “it must present ‘clear evidence that the [FAA] would not have approved a change.’”  Id.  Because it found evidence that the FAA would have permitted the change, the court held conflict preemption inapplicable.

The contrast between the Third Circuit’s two Sikkelee decisions is made only starker by the dissenting opinion filed by Judge Roth.  From the outset, Judge Roth notes that the majority erred by taking “a piecemeal approach to the Supreme Court’s impossibility preemption precedents.”  Id. at *13.  Put simply, Wyeth v. Levine cannot be read in a vacuum; for the Supreme Court’s trilogy of conflict preemption cases—Levine, Mensing, and Bartlett—to make sense, they must be read together.  This is not a novel position, but one spelled out by a wide variety of courts over the last five years.  Among the many to do so are Yates v. Ortho-McNeil-Janssen Pharmaceutical, Inc., 808 F.3d 281 (6th Cir. 2015), In re Celexa and Lexapro Marketing and Sales Practices Litigation, 779 F.3d 34 (1st Cir. 2015) (reading Wyeth and Mensing in combination), Utts v. Bristol-Myers Squibb Co., 226 F. Supp. 3d 166, 178-83 (S.D.N.Y. 2016), and—yes—the Third Circuit’s first Sikkelee decision, Sikkelee, 822 F.3d at 702-03 (reading Levine, Mensing, and Bartlett together to spell out different preemption rules for claims based on different regulatory scenarios).  The point, Judge Roth explained after reviewing the three decisions, is that:

When a manufacturer operating in a federally regulated industry has a means of altering its product independently and without prior agency approval . . . state-law claims against the manufacturer alleging a tortious failure to make those alterations ordinarily are not preempted; but, when federal regulations prohibit a manufacturer from altering its product without prior agency approval, state-law claims imposing a duty to make a different, safer product are preempted.

Sikkelee, 2018 WL 5289702 at *13.

Put another way, the fact that a defendant may have made changes in the past which were approved does not negate the fact that it still had to ask a federal agency for permission to make a change.  The fact that a party has to ask is dispositive, as the Supreme Court clarified in Mensing when it held that “[t]he question for ‘impossibility’ is whether the private party could independently do under federal law what state law requires of it.”  564 U.S. at 620.  Levine was simply the wrong framework because in Levine, it was undisputed that the brand drug-manufacturer could unilaterally do what the plaintiff alleged state law required.

In contrast with Levine, on which the majority relied, it was undisputed in Sikkelee that the design change would require prior approval.  The first Third Circuit panel expressly held that “the type certification process results in the FAA’s preapproval of particular specifications from which a manufacturer may not normally deviate without violating federal law.”  Sikkelee, 822 F.3d at 702.  The majority ignored that conclusion entirely.  Indeed, though parsing several of the applicable FAA regulations, the majority never addressed whether those regulations require prior approval of the requested change.  Instead, it short-circuited the inquiry by merely concluding that since the changes had been made subsequent to the accident, the prior approval element was meaningless.  Judge Roth, in dissent, was the only member of the panel to address the prior-approval issue and came to the same conclusion as the first Sikkelee panel—that prior approval was a necessary predicate to the design change.  And since prior approval was required, the case was more like Mensing than Levine, leaving no need to apply the clear-evidence standard.

The trilogy of Levine, Mensing, and Bartlett lay out a clear rule for conflict preemption—the same one summarized by Judge Roth.  If en banc review does not cure the Sikkelee opinion, the Third Circuit may find that Fosamax is not the last preemption decision it sends to the Supreme Court for review.

An unexpected bit of good news as we go into November. Bexis will be presenting at the Reed Smith annual client CLE program on big-deal pending cases.  In preparation, he took a look at the SCOTUSBlog page for Merck Sharp & Dohme Corp. v. Albrecht, No 17-290 (that’s the Supreme Court name for In re Fosamax in the Third Circuit).  There’s a new entry from last Friday: “Justice Alito is no longer recused in this case.”  See also, from the Supreme Court’s official docket:

Oct 26 2018 Justice Alito is no longer recused in this case.

That’s good news for our side, because as we mentioned at the time certiorari was accepted:

We do note one unfortunate aspect of the order granting review: “Justice Alito took no part in the consideration or decision of this petition.” Since Justice Alito has historically supported preemption – he wrote the dissent in Wyeth v. Levine, 555 U.S. 555 (2009) – that means we’re short a vote.

Not any more. Our understanding, based on something we heard several years ago, is that Justice Alito at one time held stock in Merck. Evidently not any longer. While we’re not counting our chickens in Albrecht before they’ve hatched, we’re not playing a man short any longer, either. So we like our side’s chances in Albrecht even better now.

For more of our Albrecht/Fosamax coverage see here, here, here, and here.

A year and a half ago we celebrated a rare prescription drug preemption win in the Philadelphia County Court of Common Pleas.  Then the decision was appealed, and we held our breath.  Preemption is never an easy sell in state courts, and Pennsylvania appellate courts are not exactly defendant friendly in prescription medical product liability cases.

Well, the wait is over and preemption won.  See Caltagirone v. Cephalon, Inc., 190 A.3d 596 (Pa. Super. 2018), allocatur denied, No. 248 EAL 2018 (Pa. Oct. 16, 2018).  We held off blogging even after the published Pa. Superior Court win, because we didn’t want to jinx anything.

Here’s what went down.

As we described before, the plaintiff had no basis for saying that anything about the drug in question (a strong opioid) was defective.  As the court pointed out, the purported wrongful death in Caltagirone was 2½ years after use of the drug at issue had ceased.  190 A.3d at 598 (plaintiff’s physician “stopped prescribing [the drug] for Mr. Caltagirone and moved him to other opioids.  About two and a half years later, on May 15, 2014, Mr. Caltagirone died.”).  Probably because the drug that killed the plaintiff’s decedent was a generic (that’s a guess on our part, but methadone has been around a long time), plaintiff nonetheless sued the manufacturer of this branded drug despite the 18-month gap.  “The overarching theme of the complaint is that even though Mr. Caltagirone died from methadone toxicity. . ., his underlying addiction was proximately caused by [defendant’s] program of promoting [the branded drug] for non-FDA approved pain management.”  Id.

In other words, rather than alleging any sort of traditional product liability “defect,” plaintiff attacked the drug because of alleged off-label promotion and off-label use – the decedent had been prescribed the drug for “migraine headaches.”  Id. at 597.

However, the defendant’s labeling forcefully warned about the drug’s addictive potential:

[The drug] carries a “Black Box” warning label, (the most serious type of FDA warning, named for the required distinctive black perimeter), advising of the risk of serious adverse health consequences from the use of [the drug], including respiratory depression, addiction, and death.  The Black Box label warns against the use of [the drug] for any condition other than cancer pain, including, specifically, migraine headaches.

Id. at 597-98.  Thus, the plaintiff’s arguments on appeal focused largely on the alleged off-label promotion of the drug – that it supposedly violated the much–discussed FDA ban on any promotion, truthful or not, of off-label uses of regulated products.

The Superior Court recognized that claim for what it was – an attempt at private enforcement of the FDA’s off-label promotion ban – facially preempted under Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001):

[Plaintiff’s] pleadings are legally insufficient. . . .  [T]he pervasive claim of [the] complaint is that [defendant’s] various derelictions, (principally, promoting sales for off-label purposes), were not approved or were in direct violation of the FDCA or its implementing regulations.

However, with narrow exceptions not asserted and not applicable here, the general rule is that there is no private right to enforce the law and regulations of the FDCA.  [citing 21 U.S.C. §337(a) and Buckman]

Because [Plaintiff’s] claims rely on asserted violations of the FDA’s “off-label” restrictions, which are pre-empted, the trial court properly sustained [defendant’s] preliminary objections.

Id. at 599-600.

Plaintiff appealed again, to the Pennsylvania Supreme Court, but on October 16, that court rejected the appeal.  We could stop holding our breath.

So we have another win for Buckman, this time against a plaintiff hell-bent on suing the manufacturer of a drug that didn’t in fact cause the complained-of injuries.  Once and for all, “off-label promotion” alone can’t be the basis of a state-law tort claim.

This post comes from the Cozen O’Connor side of the blog.

 

Today’s story is about a class action, one in which the defendant was sued for labeling its product “No Sugar Added” even though everyone involved, including the plaintiff, understood from the very start that no sugar had been added to the defendant’s product. You can probably already see where this is going.

The named plaintiff in Perez v. The Kroger Co. 2018 WL 4735701 (C.D.C. Sept. 28, 2018), alleged that the defendant, The Kroger Company, improperly labeled its 100% Apple Juice product as No Sugar Added. She hoped to represent a class that would seek financial damages for this alleged misrepresentation under California’s familiar Unfair Competition (“UCL”), False Advertising (“FAL”) and Consumer Legal Remedies (“CLRA”) laws. And she based her claims on an FDA regulation that prohibited food manufacturers from labeling a product as No Sugar Added unless, among other things, the food that the product was intended to resemble, or for which it was a substitute, “normally contains added sugar.” According to plaintiff, sugar is not normally added to apple juice (presumably because apple juice already has enough), so Kroger’s 100% apple juice product was mislabeled as No Sugar Added.

Now, as unsettling as the prospect might be of allowing a product with no sugar added to remain on the market labeled No Sugar Added, plaintiff’s claims nonetheless failed. In fact, plaintiff’s claims did not survive a motion to dismiss. Why? On this blog you should already know that answer about 40% of the time—preemption.

The Nutrition Labeling and Education Act (“NLEA”), an amendment to the FDCA, contains an express preemption clause. It is a strong one. It prohibits any state from enforcing a food labeling requirement—through, say, class action claims under the UCL, FAL and CLRA—that is “not identical to” FDA labeling regulations. Id. at *3. And plaintiff’s interpretation of the FDA’s labeling regulation was not the same as the FDA’s interpretation. The FDA interpreted it much less narrowly.

Plaintiff claimed that, under the FDA regulation, Kroger could use a No Sugar Added label only if sugar is normally added to the specific product that Kroger’s 100% apple juice product was intended to “resemble” or “substitute”—i.e., apple juice. The FDA, on the other hand, expressed a different view in a letter responding to a public interest group of some sort. According to the FDA, the comparison product need not have “the same name or the same juice content.” In the case of Kroger’s 100% apple juice product, this meant that it could be considered a “substitute” for a broader range of products than proposed by plaintiff, including “juice with added sugar, fruit-flavored soft drinks sweetened with sugar, or other sugar-sweetened beverages.” Id. at *6-7.

The district court had to decide whether to accept this FDA interpretation. To do this, the court had to determine whether, under Auer v. Robbins, 519 U.S. 452, 461–62, (1997), the FDA’s interpretation was “plainly erroneous or inconsistent with the regulation.” The court held that it was not. Rather, it found the FDA’s interpretation to be the result of a fair and considered judgment. The court gave the FDA’s interpretation deference and, accordingly, held that plaintiff’s claims were preempted. Id. at *6-7.

Interestingly, plaintiff argued that the court should not defer to the FDA’s interpretation because the FDA had staked out “nothing more than a convenient litigating position.” Id. at *6. That argument, if anything, backfired. The district court was not aware of any litigation actually involving the FDA in which its interpretation of this regulation was at issue. Rather, the FDA stated its interpretation of the regulation in response to the letter from the public advocacy group.

On the other hand, the district court noted that plaintiff’s counsel had, in fact, brought similar claims that were dismissed by other courts. With this history in mind, the court admonished plaintiff’s “counsel to tread carefully in continuing to bring these particular claims.” Id. at *7.

It would seem to go without saying that for a defendant to be liable for the purported “common-law” claim of failure to report adverse events to the FDA, there must actually be some adverse events that needed to be reported.  One would think so, but certain California breast implant plaintiffs (yes, some still exist) would beg to differ – at least they did before the recent decision in Mize v. Mentor Worldwide LLC, No. BC649083, slip op. (Cal. Super. Oct. 1, 2018).

One problem that that current breast implant litigants face that their more numerous predecessors did not is preemption.  All that earlier litigation caused the FDA to upclassify breast implants to Class III, pre-market approved devices, and PMA means preemption.

In California, that also means the filing of half-baked, failure-to-report claims that no self-respecting plaintiff would otherwise bring, as a way to allege something that gets around preemption.  Trouble is, these breast implant plaintiffs can’t even allege that the defendant didn’t report any adverse incidents.  Even what plaintiff did allege was notably speculative:

Plaintiff now has alleged, however, that if [defendant] had reported additional adverse incidents subsequent to 2000, and if the FDA had made such incidents public, and if Plaintiff’s doctors had been aware of such reports, Plaintiff’s doctors might have provided an earlier diagnosis leading to earlier surgery to remove the implants and Plaintiff’s damages . . . might have been lessened.

Slip op. at 5. That’s a lot of “what ifs” piled on top of “what ifs,” but this plaintiff couldn’t even get to that.

There weren’t any unreported adverse events.

So the plaintiff tried to make them up.

The entire questionable “causal chain” wasn’t based on any known, but unreported, events at all, but rather on allegations about how studies were conducted:

[I]t is premised on [defendant’s] failure to report adverse incidents that were not detected because of how [defendant] conducted the studies rather than on a failure to report adverse incidents that actually occurred.

Id.  Even for a liberal jurisdiction, that was just too much.  There must be something that actually wasn’t reported.

Because Plaintiff has failed to allege facts showing that [defendant] failed to report actual adverse events that in fact occurred, the failure to warn (failure to report adverse events) claim is preempted because plaintiff has failed to allege how [defendant’s] actions in conducting these studies violated federal law.

Id.

While the Mize court “adopted the reasoning” of Ebrahimi v. Mentor Worldwide LLC, 2017 WL 4128976 (C.D. Cal. Sept. 15, 2017) – a case we discussed here – Mize was really a step into fantasy beyond even Ebrahimi.  As pleaded, Ebrahimi at least involved allegations that (vaguely) alleged that events weren’t reported.  Mize didn’t.  She seems to have been alleging that the defendant was obligated to conduct studies in a way that maximized the number of reportable adverse events.  That “duty” is, of course, contrary to tort policy and medical ethics, both of which seek to reduce, not increase, product injuries.

The plaintiff in Mize also tried to allege a “manufacturing defect,” but that didn’t fare any better.  Lacking any direct evidence, plaintiff tried to rely on “allegations that supported . . . [a] 1998 Consent Decree.” Slip op. at 3.  But plaintiff’s implant wasn’t manufactured until at least two years after that decree, and the decree itself was “evidence of a promised change in practices,” so the decree could not be evidence of any defect in the device implanted in the plaintiff.  Id.  Again, simple logic seems beyond the plaintiff in Mize.

Finally, the Blog wishes to express its appreciation to Dustin Rawlin, of Tucker Ellis, and his team of, Peter Choate, Monee Hanna and Allison Burke, who not only won the case, but were thoughtful enough to send it along to us.  Keep up the good work.

If a court acknowledges that no state or federal appellate courts in the jurisdiction have addressed the question before it, we think at a minimum there also should be an acknowledgement of the Erie doctrine. Yet, in the case of Fogel v. Sorin Group USA, Inc., 2018 WL 4680022 (S.D.N.Y. Sep. 28, 2018) you get the former without the latter. Fogel is one of our least favorite types of decisions, one that claims to be a prediction of state law but instead over reaches to create new liability where it did not previously exist. That is not the job of federal courts interpreting state law.

Under the Erie doctrine, in the words of the Supreme Court:

[a] federal court in diversity is not free to engraft onto those state rules exceptions or modifications which may commend themselves to the federal court, but which have not commended themselves to the State in which the federal court sits.

Day & Zimmerman, Inc. v. Challoner, 423 U.S. 3, 4 (1975). And, not surprisingly, the Second Circuit agrees. See Runner v. New York Stock Exchange, Inc., 568 F.3d 383 (2d Cir. 2009) (“our role as a federal court sitting in diversity is not to adopt innovative theories that may distort established state law”). But Fogel disregarded Erie and then disregarded that New York has not recognized a failure to warn claim based on failure to report adverse events to the FDA.

Here are the facts. Defendant manufacturers a heart valve that underwent pre-market approval by the FDA. Fogel at *1. Plaintiff’s child underwent surgery in which her pulmonary heart valve was replaced with defendant’s valve. Two years later, the valve failed and plaintiff’s child had to undergo revision surgery during which complications occurred that caused permanent injuries. Id. at *2.

In deciding defendant’s motion to dismiss, the court found several of plaintiff’s claims were preempted. A fraud allegation that defendant concealed information from the FDA during the PMA process was preempted under Buckman as fraud-on-the-FDA. Id. at *4. Design defect was preempted under Riegel because the valve’s design was approved by the FDA and any challenge to that design would impose state law requirements that are “different from or in addition to” FDA requirements. Id. Manufacturing defect failed because plaintiff “failed to plausibly allege a manufacturing defect that violated FDA requirements.” Id. at *5. A conclusory allegation of a deviation was insufficient.

That leaves failure to warn – where the decision goes astray. As with Stengel v. Medtronic Inc., 704 F.3d 1224 (9th Cir., 2013) (en banc) and Hughes v. Boston Scientific Corp., 631 F.3d 762 (5th Cir. 2011), the Fogel court recognizes that a traditional state law failure to warn claim must be preempted under Riegel. “To the extent Plaintiff[] claim[s] that the [device’s] warning label was inadequate . . ., like the design-defect claim, must fail because the warning was approved by the FDA.” Fogel at *5. That should be the end of the story. Because “any attempt to predicate the [] claim on an alleged state law duty to warn doctors directly would have been expressly preempted.” Stengel, 704 F.3d at 1234. So let’s call failure-to-report claims what they really are – a sidestep around preemption. Frankly, they shouldn’t even get that far because what they really are are attempts at private enforcement of FDA reporting requirements which should be impliedly preempted under Buckman and 21 U.S.C. § 337(a).

That’s certainly the case in New York where a failure-to-report claim has not been recognized under state law. In fact, had the district court engaged in an Erie analysis it would have found that in other contexts New York has actually rejected state-law tort claims predicated on failure to report something to a governmental body. See Heidt v. Rome Memorial Hospital, 724 N.Y.S.2d 139, 787 (N.Y. App. Div. 2007) (“Plaintiff has cited no authority to support the proposition that a physician has a common-law duty to report actual child abuse, let alone suspected child abuse. There are good reasons for the absence of such a duty.”); Diana G-D v. Bedford Central School Dist., 932 N.Y.S.2d 316, 329 (N.Y. Sup. 2011), aff’d, 961 N.Y.S.2d 305 (N.Y. App. Div. 2013) (“there is simply no evidence that defendants’ failure to make such a report was knowingly and willful,” which was required for civil liability under child abuse reporting statute); In re Agape Litigation, 681 F. Supp.2d 352, 360-61 (S.D.N.Y. 2010) (rejecting private cause of action premised on federal reporting requirements in Bank Secrecy Act). A more fulsome discussion of these analogous cases in other states can be found in our post here.

District courts faced with an undecided state law question are not allowed to create liability where it did not previously exist. It was not the district court’s job to expand New York’s duty to warn the medical community to include the FDA.

The federal requirements require that adverse events and other reports be made to the FDA. While New York law may require manufacturers to warn the medical profession, that is not the same as a duty to report to the FDA.

Pearsall v. Medtronics, Inc., 147 F. Supp.3d 188, 201 (E.D.N.Y. 2015)(rejecting failure-to-report claims as preempted and not valid under NY law). So, under existing New York law, failure to warn physicians imposes an obligation different from and in addition to the FDA’s requirement to report adverse events. Making the claim both preempted as non-parallel and as purely FDCA-based.

 

We would be hard pressed to think of a recent judicial decision we have blasted as hard or often as the Third Circuit’s Fosamax opinion.  We deemed it the worst case of 2017.  It was bad enough that our hometown federal appellate court held that it was up to a jury whether the FDA would have rejected a stronger drug warning.  It was worse that the court held that a defendant had to prove by clear and convincing evidence that the FDA would have rejected the warning.  And it was still worse, and kind of weird, that the Third Circuit arrived at this holding in a case where the FDA actually had rejected a stronger warning.  The result of the Fosamax decision was that summary judgement in favor of the defense on preemption became a pipe dream, and ultimate victory at trial became a lot tougher.

Right from the start, we thought Fosamax was reversal bait.  The excellent certiorari briefing made us more sure of that.  Then, when SCOTUS invited the government to lob in an amicus brief, we became still surer.  Now that we have read the government’s brief, we are even surer. Lawyers from the Solicitor General’s office and the Department of Health and Human Services are on the brief.  These lawyers are typically brilliant and the positions they stake out typically command a lot of respect from courts.

When we clerked, our Judge always paid a lot of attention to government amicus briefs.  In the last SCOTUS case we worked on, the government’s position turned out to be pretty much dispositive.  We represented a father who had been tossed in the clink because he had failed to pay child support.  It was a civil contempt proceeding that ushered him into jail.  The state (South Carolina) did not provide him counsel because it wasn’t a criminal case.  We argued that however the case was styled, if imprisonment was a possibility, the right to counsel was implicated.  The other side was represented by then Penn Law professor, now Third Circuit Judge, Stephanos Bibas, who was an amazingly effective advocate for what we thought was a weak position.  We argued that counsel had to be provided and he argued not so.  Then the SG’s office filed a brief, and its answer split the difference: sometimes.  Guess what SCOTUS decided?  Sometimes won.

What do we have here with the government’s Fosamax brief?  The headline is that the government argues for reversal of Fosamax.  It argues that the Third Circuit erred in holding that, under Wyeth v. Levine, a jury must resolve the preemption defense “As a factual matter subject to a clear-and-convincing evidence standard.”  Yay. But is there a ‘sometimes’ aspect to the government’s position that makes us a bit uneasy?  Yes, there is.

As always, the statement of the Question Presented sets the tone.  Here is how the government frames the inquiry:  “Whether a state failure-to-warn claim alleging the insufficiency of brand-name drug labeling is preempted by the Federal Food, Drug, and Cosmetic Act, 21 U.S.C. 301 et seq., when the Food and Drug Administration, after the drug manufacturer provided it with the relevant scientific data, rejected the manufacturer’s application to modify its labeling to warn about the risk underlying the tort claim.”

The first thing we notice is that the government is focusing on the fact that the FDA did actually reject a proposed warning.  That rejection came in the form of a Complete Response Letter in 2009.  The central point in the government’s brief is that the meaning and effect of that Complete Response Letter presents legal questions that a judge, not jury, must resolve. It was a federal agency decision, and courts, not juries, interpret agency decisions.  Even if factual findings are necessary to interpret the federal agency decision, such factual findings are decided by the court, not the jury.

The government’s arguments on these points are compelling, and it is hard to see how reversal can be avoided.  But how useful would reversal on this ground be for the rest of us?  Whenever the FDA actually rejects an enhanced warning, there will be an agency decision, and the government’s analysis would kick in and make the issue legal, not factual.  But what if there is no rejection?  What if the defendant wants to argue that there are other circumstances, not quite reaching the level of an actual agency rejection, that demonstrate that such rejection would have been inevitable?  Are we still in fact-land, as opposed to legal-land?  The ruling in Fosamax might end up a little narrower than we’d like.

Let’s not dwell on these questions for now.  No point in being a nervous Nellie.  There are other goodies in the government’s brief.  For instance, while the Third Circuit in Fosamax speculated that possibly the FDA’s rejection was a result of dissatisfaction with the specific language proposed by the pharmaceutical company, the government smacks that position down and restores the focus to the proper place: the state of the scientific evidence. Look again at the Question Presented.  Hone in on this language:  “relevant scientific data.” The FDA looked at the data in the available literature and adverse events reports.  “No sound basis thus exists for concluding that the FDA determined in May 2009 that the data was sufficient to warrant a warning but that it rejected petitioner’s proposal because of petitioner’s proposed text.”  If SCOTUS adopts this reasoning, as it should, plaintiffs will have one less arrow in their speculative quiver to stave off summary judgment on preemption.

Even better is how the government brief eviscerates the clear-and-convincing evidence standard.  First, the brief makes the undeniable point that the Wyeth v. Levine decision simply did not address this issue.  Thus, we are thrust upon first principles.  This is a product liability case. It is a civil action for money damages, and “nothing warrants a higher standard of proof to resolve the preemption question here.”  Absent legislative specification, there is no basis to impose a higher evidentiary standard on defendants seeking to available themselves of SCOTUS preemption doctrine.  We are not entirely fond of the government’s explanation of the Wyeth “clear evidence” language, seeming to tie it into a presumption against preemption, but we are trying not to be churlish.  In any event, we are content with the government’s fundamental point that the Wyeth v. Levine  “clear evidence” formulation was not intended “as a shorthand for a formal clear-and-convincing evidentiary standard, but merely to indicate that a manufacturer asserting preemption must show that a labeling charge was not warranted under the relevant statutory and regulatory framework.”

Once the government shows that resolving the Fosamax case requires resolving a legal issue, that no factual findings are required, and that the preponderance of the evidence standard applies, the case becomes easy.  The FDA rejected the proposed warning because the data was not yet in place to justify the warning.  The Third Circuit got it wrong, the district court got it right, and the litigation should be over.

Next week, under pressure from the Drug and Device Law Lifelong Best Friend, we are participating in a “murder mystery dinner theatre” in the “conservatory” of a local cemetery.   (We didn’t know cemeteries had “conservatories.”) It is a Halloween-themed event, with costumes encouraged, and we may or may not wear our eerily-lifelike Standard Poodle mask/hood. In any event, the premise of the event is that actors are scattered among the paying audience “guests.” At some point during the cocktail hour, one of the actors will “die.” During the ensuing dinner hour, clues are revealed and everyone tries to solve the “murder” in time for dessert. We think this sounds like fun, and we like the idea of not knowing what to expect and not being able to predict the result.

But sometimes a predictable result (to the extent that preemption jurisprudence is ever predictable) is just fine. In In re Bard IVC Filters Prods Liab. Litig. (Hyde v. C. R. Bard, Inc.), 2018 WL 4356638 (D. Ariz. Sept. 12, 2018), the plaintiff was implanted with the defendant’s inferior vena cava (“IVC”) filter. Three years later, the plaintiff learned that the filter had perforated the IVC wall and had fractured. The filter was removed shortly thereafter. The plaintiff filed suit, asserting the usual panoply of product liability claims. After the court granted summary judgment for the defendant on several claims, the plaintiff’s claims for strict liability design defect and negligent design remained pending, along with a claim for negligence per se.

Under Wisconsin law, which governed the plaintiff’s substantive claims, a claim for negligence per se arises from violation of a statute, where the plaintiff can show that “(1) the harm inflicted was the type the statute was designed to prevent; (2) the person injured was within the class of persons sought to be protected; and (3) there is some expression of legislative intent that the statute become a basis for imposition of civil liability.” Hyde, 2018 WL 4356638 at *2. In her negligence per se claim, the plaintiff asserted that the defendants violated provisions of the Federal Food, Drug, and Cosmetic Act. As the court commented, “Far from containing an expression that FDA regulations are intended to form the basis of civil liability, . . . [t]he FDCA leaves no doubt that it is the Federal Government rather than private litigants who are authorized to file suit for noncompliance with the medical device provisions.” Id. (citing Buckman Co. v. Plaintiffs’ Legal Comm., 531 U.S. 341, 349 n.4 (2001)) (internal punctuation and additional citations omitted).   “Thus,” the court continued, a private litigant cannot bring a state-law claim [that] is in substance . . . a claim for violating the FDCA – that is, when the state claim would not exist if the FDCA did not exist,” because, under Buckman, such claims are impliedly preempted by the FDCA.  Id. (citations omitted). All correct, even if it conflates Buckman preemption with the plaintiff’s simple failure to state a negligence per se claim under the requirements of Wisconsin state law.

The court held that, as in Buckman, the plaintiff’s negligence per se claim was more accurately characterized as a “negligence claim based solely on violations of FDA regulations,” id., and was therefore impliedly preempted. As the court emphasized, “. . .where the plaintiff was not suing under state law for conduct that happen[ed] to violate the FDCA, but instead [was] suing solely because the conduct violate[d] the FDCA,” the claim was preempted by federal law. Id. (emphasis in original, internal punctuation and citation omitted). The court contrasted such claims to traditional tort claims like plaintiff’s negligent design claim, which arose from a duty owed under state law and which was not subject to Buckman preemption.

We like this correct, methodical, predictable decision.   We’ll let you know how the mystery thing goes.

We’ve written about a lot of Risperdal summary judgment wins. No medical causation, no warnings causation (learned intermediaries aware of risks), no alternative design, no fraud. So, when we see an opinion that overturns a plaintiff’s verdict on the grounds of (1) impossibility preemption; (2) clear evidence preemption; and (3) no evidence of general causation, we can’t help but wonder how it got to trial in the first place. So we decided to do a little digging. From our review of the case, it appears these issues were all raised at the summary judgment stage but denied. What changed before and after trial? Not the facts that support these arguments. The regulatory history hasn’t changed. The experts’ opinions haven’t changed. Yet, defendant had to go through an amateur-hour trial (we’ll tell you more about that later) and then wait over a year for these post-trial rulings granting judgment as a matter of law. Sure, better a late win then no win at all – but it certainly feels like this could have been avoided.

The case is Byrd v. Janssen Pharm, Inc., No. 1:14-cv-0820, slip op. (N.D.N.Y. Sep. 21, 2018) and, as mentioned above, involved Risperdal, an antipsychotic drug prescribed to treat serious mental conditions – schizophrenia, manic depression, and autism. Plaintiff alleged that his use of Risperdal caused him to develop abnormal breast tissue growth. The two claims that went to trial were negligent design, manufacturing, and warning defect and strict liability design, warning, and misrepresentation. Id. at 3.

The opinion methodically sets out both defendant’s arguments and plaintiff’s responses, but we’re going to jump right to the conclusions. First up was preemption. Standard plaintiff argument: defendant unilaterally should have changed its warning to include gynecomastia and was able to do it via the Changes Being Effected (“CBE”) regulations. Standard impossibility preemption defense: federal law prohibited defendant from changing the FDA-approved labeling and/or there is “clear evidence” that the FDA would have rejected the proposed labeling change. Id. at 12. The court was persuaded as to both impossibility and clear evidence. Defendant presented “clear evidence” that the FDA had rejected its request to add safety and dosing information for pediatric use of Risperdal. Id.

But the court spent most of its analysis on whether a CBE label change even was permissible under federal law. A CBE labeling change can only be made on the basis of new information concerning a serious risk. “[H]ere, the relationship between antipsychotics and [abnormal breast development] was not new information because it had been discussed in basic psychiatry textbooks for decades, and the FDA does not consider gynecomastia a serious adverse event.” Id. at 9. A “serious” adverse event is defined by federal regulations to be an event that either “resulted in inpatient hospitalization or required surgical intervention to prevent inpatient hospitalization.” Id. at 15. And, both plaintiff’s and defendant’s regulatory experts agreed that gynecomastia “would not be a serious adverse event.” Id. at 16-17. Now, plaintiff’s expert was Dr. Plunkett and she was quick to voice her personal disagreement with the FDA on this point – but that’s irrelevant (both to us and to the court). Id. at 17.

The court didn’t stop there. Defendant also argued that plaintiff had failed to satisfy his burden of proof on causation. While defendant made arguments regarding both proximate and medical causation, the court focused its attention on the latter and specifically the lack of general causation evidence. Id. at 26. Starting with Dr. Plunkett who “admitted to not being a causation expert,” but opined on it anyway – the court found her opinion unsupported by the literature. Id. None of the three pieces of literature relied on by Dr. Plunkett included a control group, so at best they were evidence of an association, not a correlation. Dr. Plunkett’s reliance on this literature demonstrated a “disregard for the difference between an association between two things and a causal relationship between those two things.” Id. at 29; see id. at 30 (“a correlation between Risperdal and gynecomastia cannot be drawn without a control group”). The fact that these studies lack a control group was likely not “new” information at trial and again begs the question why this issue is only being properly addressed post-trial.

Plaintiff’s other causation expert likewise had no support for a general causation opinion. His conclusion was that plaintiff’s gynecomastia was “secondary at least in part to prolonged use of Risperdal.” Id. at 31. But, putting aside reliance on the same literature relied on by Plunkett, the only basis plaintiff’s second expert had for his general causation opinion was his differential diagnosis. A differential diagnosis, however, “generally does not prove general causation.” Id. at 33. It assumes general causation has already been proven. Without general causation, defendant was entitled to judgement as a matter of law.

Still, the opinion continues. The remainder of the decision addressed defendant’s alternative request for a new trial based on the inappropriate conduct of plaintiff’s counsel. The court did not need to decide this issue having already found two grounds to overturn the verdict and award judgement in defendant’s favor. Based on the description of plaintiff’s trial antics, however, we can only assume that the court wanted this opportunity to admonish plaintiff’s counsel. Defendant pointed out 23 separate incidences of plaintiff’s attorney’s misconduct in front of the jury. Id. at 34. In concluding that plaintiff’s counsel’s behavior did warrant a new trial, the court relied on:

(1) Plaintiff’s counsel’s self-deprecating tone of voice and posture when referring to his lack of professional skills and/or experience, (2) his helpless tone of voice and posture when referring to the fact that he was bullied as a child, (3) his alternating innocent and defensive tones of voice in response to an admonishment by the Court, (4) the sympathetic facial expressions of the jurors following the aforementioned acts and/or accompanying comments, (5) the credulous expressions of the jurors following Plaintiff’s counsel’s acts of asserting the truth of Plaintiff’s case and/or vouching for his witnesses, and (6) the jurors’ reactions following Plaintiff’s counsel’s acts of offering his personal opinions about the evidence and/or testifying when he could not otherwise introduce evidence.

Id. at 37-38. While this behavior more than justified a new trial – it wasn’t necessary because no childish antics could overcome the fact that plaintiff had failed to prove general causation and that defendant had clear evidence to support impossibility preemption. Both of those things were true a year ago too. But better late than never.

We recently commented on Sumpter v. Allergan, Inc., 2018 WL 4335519 (E.D. Mo. Sept. 11, 2018), for several reasons – all bad – but one stands out, the ruling that used what’s essentially res ipsa loquitur to wave plaintiff through PMA preemption with a manufacturing defect.  According to Sumpter:

Plaintiffs allege that, due to manufacturing defect, the products . . . [user plaintiff] received were not the products approved by the FDA.  Even though “the precise contours of their theory of recovery have not yet been defined,” the Court holds Plaintiffs’ allegations sufficiently allege that Defendants did not adhere to FDA manufacturing requirements.

Id. at *2.  For this remarkable inference, Sumpter cites only Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), which had nothing to do with manufacturing defects as a form of parallel claim.

What Sumpter held is not the law.  Not even close.  While we recognize that manufacturing defect claims in some circumstances can thread the “narrow gap” between express and implied preemption, such allegations cannot do so by mere inference.  “Plaintiffs cannot simply incant the magic words ‘[defendant] violated FDA regulations’ in order to avoid preemption.”  Wolicki-Gables v. Arrow International, Inc., 634 F.3d 1296, 1301 (11th Cir. 2011) (citation and quotation marks omitted).

This complaint is impermissibly conclusory and vague; it does not specify the manufacturing defect; nor does it specify a causal connection between the failure of the specific manufacturing process and the specific defect in the process that caused the personal injury. Nor does the complaint tell us how the manufacturing process failed, or how it deviated from the FDA approved manufacturing process.

Funk v. Stryker Corp., 631 F.3d 777, 782 (5th Cir. 2011).

To infer an FDCA violation, and therefore an unpreempted manufacturing defect, from the mere fact of a device malfunction demands an “actual guarantee” of safety.  But “the FDA recognizes the device will not always function in accordance with specifications,” so such “additional requirement[s]” are “precluded.”  Walker v. Medtronic, Inc., 670 F.3d 569, 580 (4th Cir. 2012).  The “FDA approves the process by which a Class III device is manufactured, but it does not guarantee that every device manufactured in that process will work.”  Banner v. Cyberonics, 2010 WL 455286, at *4 (D.N.J. Feb. 4, 2010).  Therefore, “if the FDA approves a manufacturing process and the defendant-manufacturer conforms with it, a device thereby produced that nevertheless does not function as intended does not give rise to liability.”  Id.

For similar reasons, numerous courts have rejected res ipsa loquitur as a basis for inferring FDCA violations in PMA preemption cases.

[Plaintiff] relies, instead, on the doctrine of res ipsa loquitur for the proposition that full compliance would have resulted in a problem-free device.  Res ipsa loquitur does not suffice.  Res ipsa loquitur permits an inference of negligence when there can be no other explanation. . . . If negligence were the only cause of a Class III device’s failure, there would be no need for the MDA’s ongoing reporting requirements. See 21 C.F.R. § 814.84(b)(2). Plaintiff is ultimately wrong when he assumes that premarket approval guarantees the device is completely safe.

Clark v. Medtronic, Inc., 572 F. Supp.2d 1090, 1094 (D. Minn. 2008).  “Because of the general nature of the doctrine of res ipsa loquitur, a fact finder could rule in favor of [plaintiff] without finding that [defendant] violated any federal regulation.  Such a finding would impose requirements ‘different from, or in addition to,’ those in the [device’s] PMA and would thus be expressly preempted.”  Williams v. Bayer Corp., 541 S.W.3d 594, 612-13 (Mo. App. 2017).  Accord McAfee v. Medtronic, Inc., 2015 WL 3617755, at *6 (N.D. Ind. June 4, 2015) (“[t]o the extent [plaintiff] alleges that [defendant] is liable under a theory of res ipsa loquitur . . ., his claims are preempted”); Kallal v. Ciba Vision Corp., 2013 WL 328985, at *3 (N.D. Ill. Jan. 28, 2013) (preempting “argument hints at reliance on the doctrine of res ipsa loquitur”), aff’d, 779 F.3d 443 (7th Cir. 2015); Knoppel v. St. Jude Medical, Inc., 2013 WL 12116393, at *7 (C.D. Cal. Sept. 24, 2013) (“the res ipsa loquitur doctrine is barred by [preemption] because the PMA process does not demand that a product be risk free, but only that its benefits, if manufactured according to specifications, outweighs its risks”); Gross v. Stryker Corp., 858 F. Supp.2d 466, 483 (W.D. Pa. 2012) (“references to federal regulations in his negligence and res ipsa loquitur claims are too vague and general”); Desabio v. Howmedica Osteonics Corp., 817 F. Supp.2d 197, 202 (W.D.N.Y. 2011) (res ipsa claims “are based squarely on Defendants’ purported breach of state tort duties of care,” which “are precisely the type of claims Riegel held are preempted”); Cafferty v. Cayuga Medical Center, 2011 WL 541809, at *5 (N.D.N.Y. Feb. 8, 2011) (“one may not infer a manufacturing defect without creating a duty that adds to, rather than parallels, federal requirements”); Funk v. Stryker Corp., 673 F. Supp.2d 522, 531 (S.D. Tex. 2009), (because “the PMA process does not demand that an innovation be risk free . . . [i]t would follow that one may not infer a defect in the product simply because a patient encountered negative side effects in using it”), aff’d, 631 F.3d 777 (5th Cir. 2011); LeMay v. Eli Lily [sic] & Co., 960 F.Supp. 183, 186 (E.D. Wis. 1997) (parallel claims “proceed only for violations of the FDA regulations” but “res ipsa loquitur could allow recovery for negligence that Congress has preempted,” so plaintiffs “may not rely on the doctrine”).

Therefore, “[t]he fact that the [defendant’s device] allegedly failed does not itself establish a deviation from the FDA-approved standards.”  Smith v. St. Jude Medical Cardiac Rhythm Management Div., 2013 WL 1104427, at *4 (D. Md. Mar. 13, 2013).  In Burgos v. Satiety, Inc., 2013 WL 801729 (E.D.N.Y. March 5, 2013), no inference that an investigational device violated FDA regulations could be inferred from its malfunction, even though plaintiff had been offered a settlement and there were allegations of spoliation (neither of which were present in Sumpter).

Plaintiff has failed to provide any circumstantial evidence, other than the fact that Plaintiff was injured and that she was offered some money, to argue that the . . . device was actually manufactured in a way that violated the IDE.  Plaintiff has not presented any evidence that the . . . device was adulterated, or that it was unreasonably dangerous and unfit for [its] intended purpose to show that the IDE was violated.  Failure to provide any evidence in support of Plaintiff’s claim is fatal and it must be dismissed.

Id. at *7 (citations and quotation marks omitted).  See also Herrnandez v. Stryker Corp., 2015 WL 11714363, at *3 (W.D. Wash. March 13, 2015) (plaintiff could not escape preemption on “conclusory allegations” that explanted device violated FDA regulations); Ali v. Allergan USA, Inc., 2012 WL 3692396, at *11 (E.D. Va. Aug. 23, 2012) (facts that support causation from a malfunction “are not sufficient to link the [malfunction] or [plaintiff’s] injuries to any federal violation . . . in the manufacture of the device”); Anthony v. Stryker Corp., 2010 WL 1387790, at *4 (N.D. Ohio Mar. 31, 2010) (an “attempt to recast generalized deviations from ‘manufacturing performance standards’ as specific violations of federal regulations is insufficient to state a claim”); Rankin v. Boston Scientific Corp., 2010 WL 672135, at *4 (E.D. Ky. Feb. 19, 2010) (“that the [device] allegedly failed during normal use does not override the clear language of §360(a) or the Supreme Court’s ruling in Riegel that the plaintiffs’ claims are preempted by federal law”); Cenac v. Hubbell, 2009 WL 10678961, at *4 (E.D. La. July 31, 2009) (“mere fact that a device malfunctions or produces an adverse result does not plead a violation of federal law”); Delfino v Medtronic, Inc., 2018 WL 2688420, at *8 (Minn. Dist. May 18, 2018) (“the mere fact that a device malfunctioned is not evidence that the device violated any federal requirement”).

Sumpter cited nothing to support its letting the plaintiff slide under TwIqbal on the theory that a device malfunction, without more, is enough to establish not only a product defect, but an FDCA violation.  That’s two bridges too far.  One can only hope that defense counsel will be able to show the trial court the error of its ways soon, without the need for an appellate court to do it for them.