This year has seen a lot – and most of it would not fall into the positive category.  The pandemic; severe wildfires in the U.S. and Australia; record setting tropical storms in the Atlantic; social unrest over police brutality; explosion in Beirut.  Unfortunately, we could go on.  Even more unfortunate is the fact that all of us tend to focus on the negative.  It’s not hard to do with today’s media coverage.  But there is a flipside.  Anti-racist books are topping bestseller lists.  Animals are being adopted by the hundreds of thousands in America.  People are finally listening to their parents and washing their hands.  Drive-in movie theaters made a comeback.  John Krasinski’s “Some Good News.”  Companies from Ford and GM to Apple to your local distillery shifted production to make much needed masks, ventilators, and hand sanitizer.  And Hamilton the movie lived up to the hype.

That’s the world in general.  Here in the DDL world, our 2020 “p” word isn’t pandemic.  It’s Pradaxa, and preemption, oh and pummels.  The defense has had a lot to be happy about in the Pradaxa litigation.  We’ve been bringing you consistent wins since January (here’s a collection).  Now here is one more to add to the slate – Silverstein v. Boehringer Ingelheim Pharmaceuticals, Inc., 2020 U.S. Dist. LEXIS 188176 (S.D. Fla. Oct. 7, 2020).

It’s a lengthy decision that goes into extensive detail about the studies, analyses and science that were submitted to the FDA during and after the approval of the drug.  See id. at *36-*88.  Since the court found none of it to be “newly acquired evidence” that would support a CBE label change, we’re going to skip the details and focus on the legal conclusions.

Plaintiff suffered a significant gastrointestinal bleed while on Pradaxa, an anticoagulant, and alleged that the drug was defective because defendant did not adequately warn that blood plasma concentrations should be monitored and that certain patient characteristics, such as renal impairments and concomitant use of other drugs, could increase the risk of severe bleeding.  At the summary judgement stage, the primary issue was whether plaintiff could survive preemption because the manufacturer could have submitted a revised warning under the Changes Being Effected (“CBE”) regulations.  To make a CBE label change, however, the manufacturer would have to have “newly acquired information” – information not previously provided to the FDA – that demonstrates “risks of a different type or greater severity or frequency than previously included in the submission to FDA.”  Id. at *89-90.

We’ve already given away the ending – the court rejected the contention that the manufacturer had the evidence to support a CBE label change.  Before reaching this conclusion, the court took a slightly less defense-friendly view on the burden of proof.  Acknowledging that most other courts to examine the issue have held that plaintiffs bear the burden to prove that the manufacturer obtained newly acquired information to support a label change, this court took a more middle of the road position.  Id. at *31.

I find that Plaintiffs should bear the initial burden of identifying the specific information that they assert [the manufacturer] acquired after the FDA approved [the drug] and should have used to modify the [drug’s] label. Once Plaintiff points to this specific information, [the manufacturer] bears the burden of proving that it does not meet the definition of “newly acquired information” under the CBE regulation. This allocation of burdens avoids making [the manufacturer] prove a negative – that it acquired no new information after [the drug] was approved that would have justified a CBE modification.

Id. at *32-33.  As the lengthy recitation of evidence shows, plaintiff carried their burden of proof by pointing to all sorts of supposedly new analyses, papers, etc.  But so too did the manufacturer.  The court found several points persuasive.  Such as that the FDA had approved Pradaxa’s labeling on “19 separate occasions” without requiring the warning Plaintiff wanted and after reviewing much of the evidence plaintiff argued was newly acquired.  Id. at *84.  Also, when the FDA approved Pradaxa it was already aware of the inconclusive evidence regarding whether stroke risk went down at higher levels of plasma concentration.  So, the evidence plaintiff relied on did not reveal a risk of a different type to fall under the CBE regulations.  Id. at *95.  Even viewing the evidence in “a light most favorable to plaintiff,” id. at *98, despite preemption being a question of law, the evidence did not support a conclusion different from what was known at the time of approval.  Id.  Moreover, to support a CBE label change, the evidence also must be scientifically reliable.  “A CBE change cannot be rooted in conjecture or hypothesis.”  Id. at *20-21.  Plaintiff’s uncorroborated, unpeer-reviewed, and unreplicated sources were not sufficiently scientifically reliable to be newly acquired information.  Finally, the type of risk at issue – a major bleed – was known by the FDA and was included in the label which “conclusively shows that it was submitted to the FDA.”  Id. at *109.

Because the court concluded that none of the evidence put forth by plaintiff was “newly acquired,” it dd not have to reach the issue of whether the FDA would have rejected the proposed label change to conclude the plaintiff’s claims were preempted.  Id. at *111.  The court did, however, address defendant’s alternate argument that the Pradaxa label was adequate as a matter of law.  Plaintiff’s first mistake was relying on Dr. Plunkett for label adequacy.  Because she is not a medical doctor, “[s]he does not, and presumably cannot, opine on how a treating physician would interpret and respond to the [drug’s] label.  Id. at *113.  Even so, what plaintiff was asking the court to do is require that the manufacturer not only warn of the risk, but also the importance of and how to monitor for the risk.  Under Florida law, a manufacturer “need not warn about the specific manner in which the injury may occur.”  Id. at *114.  Because the label contained “accurate, clear, and unambiguous warnings” regarding bleeding, therapeutic monitoring, and increased risk to patients with renal impairment – it was “sufficient to educate a reasonable treating physician that Pradaxa presented” the precise risk at issue.  Id. at *115.  Therefore, the warnings were adequate, and defendant was entitled to summary judgment on all counts.

So, we know Pradaxa goes in the 2020 win column.  But, will the 2020 Halloween rare full moon, blue moon end up as just a picturesque backdrop for outdoor parties or Night of the Living Dead?  Given how the year is going . . . .

Today we have another guest post from friend-of-the-Blog, Dick Dean at Tucker Ellis.  He’s familiar with the ongoing Pradaxa litigation and is pleased with the preemption pummeling Pradaxa plaintiffs have been receiving.  Here’s his post about yet another favorable decision from the state-court Pradaxa proceedings in Connecticut.  With decisions like this, who needs snap removal?  As always, our guest posters deserve 100% of the credit (and any blame) for their work.


No commentator or blog has more precisely and persistently probed preemption and its “newly acquired information” component than has the Drug and Device Law Blog.  And the Blog has done it with poetry, or at least alliteration.  If you went right to the story and didn’t savor the headlines you missed:  “Post-Albrecht Preemption Pummels Pradaxa Plaintiffs,” followed shortly by “Post–Albrecht Preemption Persistently Pummels Pradaxa Plaintiffs.”  But what started this parade was “Plaintiff’s Pyrrhic Pradaxa ‘Victory’,” reporting on Roberto v.  Boehringer Ingelheim Pharmaceuticals, Inc., No. CPLHHDCV166068484S, 2019 WL 5068452 (Conn. Sup. Ct. Sept. 11, 2019).  The Roberto court found a labeling claim about the bleeding risk from a blood thinner preempted because there was no newly acquired information on a general bleeding risk that would have warranted a CBE label change.  But the trial court also found that there was sufficient evidence to sustain a verdict of a heightened risk of bleeding in GERD (gastroesophageal reflux disease) patients, specifically.  Indeed, the court concluded in Roberto that it was unclear whether information bearing on the GERD risk had been submitted to the FDA.  But the Pyrrhic Victory post recognized that the bigger picture was the general bleeding risk, and that there were only a few GERD cases in this litigation.  So the GERD finding was a pyrrhic victory.

But now even that plaintiff’s Pyrrhic Pradaxa victory has been put out to pasture.  No longer are there GERD outliers that escape preemption.

In Estep v. Boehringer Ingelheim Pharmaceuticals, Inc., No. X03 HHD CV17-6075319S, 2020 WL 5290777 (Conn. Sup. Ct. Aug. 25, 2020), the same court that decided Roberto considered a follow-up GERD case, which had an “expanded record” on the newly acquired information issue that had allowed the Roberto plaintiff to escape preemption.  Id. at *7.  The Estep Plaintiff likewise claimed the new information on the GERD issue was a European label and a 2013-2014 medical journal article.  The court again found that these items showed reasonable evidence of different or greater risks not stated in the U.S. label.  Id. at *8.  But that left the issue of “newness.”  The definition of “newly acquired information” uses phrases like “not previously submitted to the Agency” and “previously included in submissions to the FDA.”  21 C.F.R. §314.3(b).  Defendant argued that it had submitted data to the FDA that were equivalent to what plaintiff cited in the European label and the journal article—specifically, it relied on reports of bleeding in GERD patients.  Id. at *9.

Plaintiff in Estep argued that the information submitted to the FDA had to be “tethered to a proposed label change” in order to qualify as “previously submitted” to the agency.  Id.  That novel argument was to no avail.  The Court found that a straightforward application of the rules of statutory construction supported the defendant’s position.  It then looked closely at the information submitted to the agency to determine whether it was “equivalent to the claimed newly acquired information.”  Id. at *9.  The defendant pointed to several reports submitted to the FDA prior to the original approval in October 2010 that showed gastritis-like symptoms, defined to include GERD, were associated with an increased risk of GI bleed.  Id. at *13.  The court then cited almost an entire page of tables and charts linking bleeding risks in gastritis-like symptoms, which had been submitted.  Id.  In addition, in the “expanded record” defendant showed that it had submitted the very European label relied upon by plaintiffs to the FDA in September 2012, after approval but before a December 2013 label change and before plaintiff’s bleed in February 2015.  Id.

Estep is a significant decision.  It is the first holding that hard data submitted without a specific request for a label change can defeat a claim of newly acquired information.  Not that it should make any difference, but it’s another plaintiff propaganda point put down, and recognized for what it is, a distinction without a difference.

Estep stands in stark contrast to Evans v. Gilead Sciences, Inc., No. 20-cv-00123-DKW-KJM, 2020 WL 5189995 (D. Haw. Aug. 31, 2020), which the Blog discussed here last month.  That post noted the adverse warning holding in Evans that “…even if there was little or no ‘newly acquired information’ relevant to Evans’ claim, that did not make it impossible for Gilead to change its….label.”  Id. at *11.  The opinion in Estep provides a detailed refutation explaining why Evans was erroneous.  2020 WL 5290777 at *10-*11.  Estep cited to the FDA’s comments supporting the rule change in 2008 adding the concept of “newly acquired information.”  Those comments pointed out that allowing sponsors to change labeling for a product without limitation based on previously submitted information would undermine the FDA label approval process.  Id. at *11.

If a plaintiff is going to allege an inadequate warning because a CBE could have been made, then the CBE regulation’s newly acquired information prerequisite must be confronted—not glossed over.  Estep also provides an excellent summary of the law in regard to pleading “newly acquired information.”  Id. at *5-*6.  Once a defendant makes an initial showing of a preemption defense, then a plaintiff must specifically identify the “newly acquired information” he relies upon.  Numerous posts on this blog have treated this issue.  In addition to the “Pummels” posts, seeOnline Reviews Are Not ‘Newly Acquired Evidence,” “On Newly Acquired Evidence,” “‘No Newly Acquired Evidence’ Argument On Implied Preemption Gaining Traction,” and “Plaintiffs Have Burden To Plead Newly Acquired Evidence.”  As the Blog’s post about Evans discussed, that decision stumbled badly on this increasingly key pleading issue—ignoring many cases in the last three years directly on point while citing to a 2018 Ninth Circuit decision having nothing to do with “newly acquired information.”  2020 WL 5189995 at *10.

Estep underscores what an outlier decision Evans truly is.

Lyons v. Boehringer Ingelheim Pharmaceuticals, Inc., 2020 WL 5835125 (N.D. Ga. Sept. 29, 2020), was a wrongful death action alleging that the anticoagulant drug Pradaxa was defective and not accompanied by adequate warnings that blood plasma concentrations should be monitored and that certain patient characteristics, such as age, renal impairments, and concomitant statin use, could increase risks of severe bleeding. The design defect claim turned out to be wholly premised on the alleged failure to warn, and the failure to warn claim turned out to be wholly preempted. Thus, the Lyons case is consistent with other Pradaxa cases arriving at the same result, and represents a sound application of the SCOTUS Albrecht opinion. (We discussed these Pradaxa cases here and here, for example.)

To try to get around the preemptive effect of an FDA-approved drug label,the plaintiff in Lyons contended that the drug manufacturer could have submitted a Changes Being Effected regarding the needs for plasma monitoring and paying attention to certain patient characteristics. But the Lyons court rejected this contention because there was no “newly acquired information” supporting the new warnings and, in any event, the extensive record of information exchanges between the manufacturer and the FDA provided “clear evidence” that the FDA would have rejected the proposed new warnings.

The issues of newly acquired information and likely rejection by the FDA go together, of course. All of the supposedly new papers, analyses, emails, European labels, etc. regarding the drug’s safety either contained information the FDA already had, or were not so much new information as mere “trial balloons.” There was simply no basis furnished by the plaintiff to suggest that the FDA would have seen fit to add warnings to the label. Moreover, based on all this information, the FDA had twice rejected some of the warnings, reasoning that “independent risk factors for bleeding … do not need to be specifically in the label” because “we do not want the [label] to become too long to be useful.”

One of the more interesting aspects of Lyons is the issue about whether the judge or jury decides the issue of “newly acquired information.” In Albrecht, SCOTUS decided that it is up to the judge, not the jury, to determine whether the FDA would have rejected the proposed warning. That means summary judgment is possible on that issue. But the plaintiff in Lyons argued that, despite Albrecht, “newly acquired information” is an issue for the jury. That position would go a long way toward making summary judgment impossible – the plaintiff’s goal. Like other courts that have tackled this point, the Lyons court held that the logic of Albrecht means that whether something is “newly acquired information” must be decided by the court, since that issue is usually a predicate for the issue of likely FDA rejection. Applying this analysis, the Lyons court granted the defendant summary judgment.

Inasmuch as Lyons falls in line with other cases, perhaps its ruling does not constitute newly acquired information. But it is helpful and reassuring nonetheless. Clear evidence of judicial rationality is always welcome.

Not long ago, in our “Post-Albrecht Preemption Pummels Pradaxa Plaintiffs” post we discussed several recent favorable preemption decisions in product liability litigation involving that drug:  Ridings v. Maurice, ___ F. Supp.3d ___, 2020 WL 1264178 (Mag. W.D. Mo. March 16, 2020), Adkins v. Boehringer Ingelheim Pharmaceuticals, Inc., 2020 WL 1704646 (Conn. Super. March 13, 2020), Ridings v. Maurice, 2019 WL 8223599 (W.D. Mo. Oct. 20, 2019), and Roberto v. Boehringer Ingelheim Pharmaceuticals, Inc., 2019 WL 5068452 (Conn. Super. Sept. 11, 2019).

Those were all favorable cases, but they were all by trial-level courts, and thus subject to the vagaries of the appellate process.  How about a similar decision from the other end of the appellate process?  Thus, we present Boone v. Boehringer Ingelheim Pharmaceuticals, Inc., ___ A.3d ___, slip op., 2020 WL 2121063 (Conn. May 4, 2020).

Boone was an appeal from a trial that produced a defense verdict.  The opinion addresses several rather case/product-specific issues before it gets to the preemption topic that is near and dear to our blogging hearts.  Thus, we’ll pass over pages *2-11 of the Connecticut Supreme Court’s opinion (which resolve disputes over spoliation and rebuttal evidence).

Boone unanimously affirmed the trial court’s grant of a preemption summary judgment motion against “a design defect claim related to the defendants’ failure to develop and market a reversal agent.”  2020 WL 2121063, at *12.  This claim ran afoul of what we call the “Mensing independence principle” – “The relevant inquiry, [Mensing] held, was whether the defendants ‘‘could independently do under federal law what state law requires. . . .”  Id. at *14 (quoting PLIVA, Inc. v. Mensing, 564 U.S. 604, 620 (2011)) (emphasis supplied by Boone).

The preempted claim, which really isn’t a “design defect” since it doesn’t involve any aspect of the actual product’s design, see id. at *12 n.32, was that the defendant should not have sold the drug at all until it had also developed and obtained FDA approval for a different drug.  As the Boone court recognized, id. at *14, this claim was also preempted under Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013).  Moreover, footnote 32 is itself interesting, since usually courts steer away from deciding constitutional issues when they can avoid them.  But in Boone, “[b]ecause we conclude that the trial court properly granted the defendants’ motion for summary judgment on federal preemption grounds,” the court did not address whether a claim was stated under Connecticut state law.  Preemption was evidently an easy call in Boone.

Boone recognized that, despite a purported presumption against preemption, the logic of Mensing and Bartlett “compels” preemption.  Plaintiff was using state law to hold one drug’s FDA approval hostage to the agency also approving a second drug.  That theory self-evidently depended on the FDA acting to approve the second drug – which required preemption.  “[I]t is enough to hold that when a party cannot satisfy its state duties without the [f]ederal [g]overnment’s special permission and assistance, which is dependent on the exercise of judgment by a federal agency, that party cannot independently satisfy those state duties for [preemption] purposes.’’  Boone, 2020 WL 2121063, at *14 n.34 (quoting Mensing, 564 U.S. at 623-24) (emphasis added by Boone).

In order to cure the design defect alleged by the plaintiff, the defendants would have had to bring [the second drug] to market before the [alleged injury occurred].  Because there is no dispute that [the second drug] was not approved by the FDA until [later], the defendants could not have satisfied their alleged state law duty to the decedent without marketing an unapproved drug in violation of federal law.

Id. at *15.

Plaintiff unsuccessfully argued “that the test for preemption set forth in Mensing and Bartlett is inapplicable to present case because those cases do not involve brand-name drugs.”  Id.  No dice.  While the “different levels of control” that branded and generic manufacturers exercised over their labels “informed the [supreme] court’s analysis . . ., the nature of the underlying test remained consistent:  whether the defendant ‘‘could independently do under federal law what state law requires.”  Id. (once again quoting and emphasizing Mensing independence principle).  Boone agreed with the holding in Yates v. Ortho-McNeil-Janssen Pharmaceuticals, Inc., 808 F.3d 281 (6th Cir. 2015), that, ‘‘contrary to [the plaintiff’s] contention that the impossibility preemption in Mensing and Bartlett is limited to generic drugs, we view Levine, Mensing, and Bartlett as together stating the same test for impossibility preemption.’  Boone, 2020 WL 2121063, at *15 (quoting Yates, 808 F.3d at 296-97).

Along the way, Boone rejected the plaintiff’s argument (which readers will recognize from our previous Pradaxa preemption pummels plaintiffs post) that Merck Sharp & Dohme Corp. v. Albrecht, 139 S. Ct. 1668 (U.S. 2019), somehow confined the preemption inquiry in branded drug cases to the so-called “clear evidence” test.  Boone made clear that “clear evidence” was only a piece of the implied preemption puzzle:

The plaintiff in the present case asserts that a recent United States Supreme Court case explaining that particular standard, [Albrecht], stands for the broad proposition that impossibility preemption ‘‘only applies when a defendant can affirmatively show that it attempted to get the FDA to allow the safer alternative proposed by the plaintiff and the FDA affirmatively and officially rejected it.’’ (Footnote omitted.)  We disagree. The clear evidence standard in [Levine] applies only when a defendant seeks to prove that compliance with a state law obligation remains impossible notwithstanding its ability to act unilaterally under federal law.

2020 WL 2121063, at *13 n.33 (emphasis original).  The clear evidence test applies when drug manufacturers “could have satisfied their state law obligation to provide a label with an adequate warning by unilaterally making label amendments.”  Id. (citing FDA CBE regulation).  Because the plaintiff’s preempted theory was not something that could be addressed by a CBE label change, “Albrecht is inapposite.”  Id.

The plaintiff’s last-ditch, state-of-the-art argument against preemption in Boone also failed.  Whether or not it was “technologically feasible” to seek FDA approval of the second drug, doing so still required the FDA to act to grant approval.  That pesky Mensing independence principle sank the plaintiff once again.

Although such practical considerations may sometimes limit the options available to a manufacturer; that fact is inapposite to the question of whether marketing [the second drug] would have required the FDA’s ‘‘special permission and assistance.”  For similar reasons, we are also unpersuaded that the FDA’s subsequent approval of [the second drug] is dispositive.  The possibility that the FDA would have looked favorably on an earlier application does nothing to alter the fact that, at the time of the decedent’s death, the defendants were prevented from unilaterally marketing [the second drug] under federal law.

Boone, 2020 WL 2121063, at *15 (quoting Mensing, other citations omitted).  Boone agreed with earlier decisions (including Maurice) that had held similar claims preempted.  Id. at *15 n.38 (reaching “same conclusion” as Ridings v. Maurice, 2019 WL 4888910, at *6 (W.D. Mo. Aug. 12, 2019), and Chambers v. Boehringer Ingelheim Pharmaceuticals, Inc., 2018 WL 849081, at *13 (M.D. Ga. Jan. 2, 2018), and giving the “but see” to In re Xarelto (Rivaroxaban) Products Liability Litigation, 2017 WL 1395312, at *3 (E.D. La. April 13, 2017)).  We agree that the Xarelto decision is wrongly decided.

Unless the plaintiff wants to appeal Boone to the United States Supreme Court – go ahead make our day, after Bartlett we can imagine what the Court would do with this theory – this is the end of the appellate line.  Further, given the Connecticut Supreme Court’s reasoning in Boone, we’re cautiously optimistic that it will uphold the later Connecticut state decisions in Roberto and Adkins, which are wending their way through the appellate process.  While the arguments aren’t all identical, at least two of those plaintiffs’ major contentions (generics are different and Albrecht) bit the dust in Boone.

Finally, we are gratified to see implied impossibility preemption applied to “you should have made a different/additional drug” claims.  The same preemption rationale that Boone adopted should also be fatal to the claims we discussed here, that the drug that reduced AIDS from a death sentence to a treatable chronic condition was “defective” because the defendant didn’t make a “better” drug sooner.  Any defect claim predicated on the possible FDA approval of a different drug necessarily depends on action by the FDA, and is thus preempted under the Mensing independence principle.

Two of longest recent entries in our Post-Levine Drug/Vaccine Preemption Cheat Sheet are Pradaxa wins.  Adkins v. Boehringer Ingelheim Pharmaceuticals, Inc., slip op., 2020 WL 1704646 (Conn. Super. March 13, 2020) (#106), and Ridings v. Maurice, ___ F. Supp.3d ___, 2020 WL 1264178 (Mag. W.D. Mo. March 16, 2020) (#108).  There’s also a third recent decision, Ridings v. Maurice, 2019 WL 8223599 (W.D. Mo. Oct. 20, 2019) (which we’ll call “Ridings 0”), that was a procedural precursor to the more recent Ridings opinion.  Between them, these recent Pradaxa wins provide some useful guidance about how to approach branded prescription drug preemption motions in the new post-Albrecht, “matter of law” world.  We’ll go through them in order.

Note:  many of the factual rulings in these cases are similar (sometimes identical) to a Pradaxa decision we already blogged about last September, Roberto v. Boehringer Ingelheim Pharmaceuticals, Inc., 2019 WL 5068452 (Conn. Super. Sept. 11, 2019) (#96).  Today, we’re mining these opinions for more broadly applicable propositions, so this post won’t go into nearly the item-by-item detail that our Roberto post did.

In Ridings 0, the court had to reconsider a prior decision that had held preemption of the plaintiff’s warning claims to be a jury question due to disputed issues of fact.  2019 WL 8223599, at *1-2 (discussing Ridings v. Maurice, 2019 WL 4888910 (W.D. Mo. Aug. 12, 2019)).  The court recognized the obvious – that the Supreme Court’s decision in Merck Sharp & Dohme Corp. v. Albrecht, 139 S. Ct. 1668 (U.S. 2019), required those disputed issues to be decided by the court.  Ridings 0, 2019 WL 8223599, at *2 (“To the extent that the Court’s order mandat[es] the submission of the preemption question to the jury, it is contrary to Albrecht.”).  Thus, the court had to create a process for undertaking the preemption analysis.

For those of you who might be making similar procedural suggestions, here’s what the Ridings 0 court came up with:

  • Since “fact-finding” was involved, the decision would not be made solely on the summary judgment record.
  • Rather, the court set a preemption hearing at which “witnesses (including expert witnesses)” would be examined and “factual determinations based on credibility and persuasiveness” would be made.
  • The hearing would “require two stages of analysis,” with a shifting burden of proof.
  • “First, [plaintiff] must show that ‘newly acquired information’ existed such that [defendant] could unilaterally change its label in accordance with the CBE regulation.”
  • Second, if plaintiff established newly acquired information, the defendant “may still establish an impossibility preemption defense by presenting “clear evidence” that the FDA would have rejected the labeling change.”

2019 WL 8223599, at *2-3.

This process led the Ridings court to conclude all of the plaintiff’s warning claims were preempted.  Ridings, 2020 WL 1264178, at *23 (“all of [plaintiff’s] remaining claims are preempted by federal law”).  But before we get to the preemption analysis, here are a few more procedural steps that Ridings took:

  • “To keep the hearing focused on the actual issue in the case, the Court thus required [plaintiff] to specifically identify the warning (or warnings) that he believed should have been given by [defendant].” Id. at *4.

This requirement is significant, since plaintiffs love for their warning claims to be a constantly moving target.  Consider it a collateral benefit of preemption.

  • Because “it was not the usual practice to ask a party to prove a negative, . . . the best practice was to view the question of preemption in drug litigation as involving a shifting burden.”  Id.
  • First, the “defendant must make an initial prima facie showing of preemption for the warning it did provide.”   This step was undisputed, since “the warning label for [the drug] in effect at the time [plaintiff] was using [it] had been approved by the FDA.”  Id.
  • Once the defendant puts preemption at issue, “a plaintiff then bears the burden of coming forward with evidence sufficient to conclude that the defendant could and should have availed itself of the [CBE] regulation[] and given a different warning.”  Id.
  • If plaintiff establishes the prerequisites to application of the CBE regulation, then “the defendant then bears the burden of coming forward with ‘clear evidence’ that the FDA – even if presented with such ‘new’ evidence – nonetheless would have rejected the warning(s) advocated by the plaintiff.”  Id.
  • The hearing was conducted in accordance with “the standards associated with bench trials.”  Id.
  • Following the hearing, “the parties submitted their respective proposed findings of fact and conclusions of law.”  Id.

Ridings, 2020 WL 1264178, at *4.  So if a post-Albrecht court asks “what now” in response to a preemption motion, the Ridings litigation provides a road map for how to get from here (motion filed) to there (hopefully, dismissal).

The substantive preemption portion of Ridings is 35 pages (18 Westlaw star pages) long.  Here are some highlights that we find useful for defendants pursuing implied impossibility preemption in a post-Albrecht world.

  • In general, “under Mensing and similar cases, federal law preempts any state law claim requiring a drug manufacturer to change its labeling” subject to “limited opportunities [for] a drug manufacturer to unilaterally change a drug label that can, in some cases, effectively undercut any claim of impossibility.”  Ridings, 2020 WL 1264178, at Id. *12.
  • “[T]he determination of preemption will often require the Court to make factual determinations based on competing and contradictory evidence.”  Id. at *14 n.27.
  • “[T]he issue of ‘newly acquired information,’ [requires that] the Court consider[] whether there is adequate proof that [defendant] was in possession of such information, and that such information was reasonable evidence of an association of a serious hazard with [the drug] sufficient to require [defendant] to unilaterally revise its label to include a warning.”  Id. at *15.
  • “[N]ewly acquired information must provide reasonable evidence of a causal association of a clinically significant adverse reaction linked to a drug. A clinically significant adverse reaction has a significant impact on therapeutic decision-making, such as a risk that is potentially fatal or otherwise serious.”  Id. (regulatory citations and quotation marks omitted) (emphasis original).

This is another point that warrants emphasis.  In some situations, the requirement of a “clinically significant” risk can be of critical importance.

  • The FDA’s requirement that “there must be sufficient evidence of a causal association between the drug and the information sought to be added” enforces the Agency’s objective of preventing overwarning.  Id. (citation and quotation marks omitted) (emphasis original).
  • “[S]tudies concluding that it ‘remains unknown’ whether a drug is linked to a particular adverse reaction or risk or that ‘further studies are required to address possible clinical consequences’ do not constitute reasonable or well-grounded scientific evidence of ‘clinically significant adverse effects’ under the CBE regulation.”  Id.

That’s another ruling with broad importance, since these kind of tentative studies are frequently all plaintiffs have in prescription drug product liability litigation.

  • While “it is clear that newly acquired information may encompass a new analysis of old data . . . studies published after a plaintiff’s injury are not relevant to constitute newly acquired information.”  Id. (citations and quotation marks omitted).
  • “[A] claim that a drug label should be changed based solely on the information previously submitted to the FDA is preempted because the CBE regulation cannot be used to make a label change based on such information.  Id. at *16 (citation and quotation marks omitted).
  • “[W]arnings approved for a foreign label are not in and of themselves newly acquired evidence when they are based on consideration of substantially similar information. Foreign drug labeling is the product of different and distinct regulatory standards and decisions.”  Id. at *17.
  • [R]ejection of bad ideas is as important to the process as the confirmation and embracing of good ideas.  [A withdrawn patent] was a bad idea that simply did not pan out.  It does not constitute newly acquired evidence under the CBE regulation.”  at *18.
  • An article concluding “that there is no single . . . optimal benefit-risk [point] for all patients . . . does not establish any “risks of a different type or greater severity or frequency.”  Id. at *20 (citation, footnote, and quotation marks omitted).
  • “[P]reliminary discussions do not provide reliable evidence of new risks. They are essentially uncorroborated trial balloons.”  Id.

We think that Ridings is thus brimming with favorable points and usable quotes.  In particular, we recommend the holdings that articles concluding only “we’re not sure” and “conduct more research” and “preliminary discussions” are all too tentative to be newly acquired information, and thus insufficient to prevent preemption are important, since so much of the other side’s arguments are based on taking precisely this sort of material out of context.

Finally, Ridings took a stab at “clear evidence” as an alternative holding.  Even though the defendant hadn’t submitted to the FDA the language that plaintiff claimed was necessary to have an adequate warning, Ridings found “clear evidence” that the FDA would not have allowed it.  Ridings, 2020 WL 1264178 at *21.  As we’ve pointed out elsewhere, plaintiffs now claim that Albrecht requires actual submission and rejection of a warning proposal for “clear evidence” to exist.  Not so – at least on the well-developed regulatory record in Ridings:

[T]he Court wishes to be clear – it should not always be the case that simple inaction by the FDA in light of submitted data will always be “clear evidence” that the FDA would reject a particular warning.  In this case, however, in light of the known issues and the ongoing give-and-take between [defendant] and the FDA on these issues . . . the FDA’s continued inaction does represent clear evidence under these facts.

Id. (emphasis original).

The other major recent Pradaxa opinion is Adkins, 2020 WL 1704646, decided by the same judge who issued Roberto.  As we would expect, Adkins does a lot of quoting from and citing to Roberto.  However, Adkins also adds some independently useful points.  First, Adkins rejects the plaintiff’s argument that, by only addressing the “clear evidence” prong of impossibility preemption, Albrecht impliedly eliminated all other bases for preemption.  Not at all:

The plaintiff claims that in this context a “state law claim is preempted only when there is ‘clear evidence’ that the FDA would not have approved a change to the drug’s label.” . . .  This statement . . . is not correct.  There are actually two theories of preemption. . . .  The first, or threshold, theory addresses whether the pharmaceutical company had authority under federal law to change the label on its own to conform to the plaintiff’s state law demands.  If it did not, then it would also be impossible for the company to comply with state law.

Adkins, 2020 WL 1704646, at *2 (footnote omitted).  Plaintiff’s warning claims must navigate the “lengthy and complex” CBE regulation, of which “newly acquired information” is “the key component.”  Id. at *3 & n.4.  Albrecht didn’t have to address the requirements that the FDA wrote into its CBE regulation because the defendant in that case “conceded that the FDA’s CBE regulation would have permitted [it] to try to change the label.”  Id. at *4 (quoting Albrecht).  Further, a plaintiff cannot simply declare ex cathedra the s/he has “newly acquired information” – the “basic adversary nature of our judicial system” affords defendants the right to dispute plaintiffs’ claims of purportedly newly acquired information.  Id. at *4 n.6.

Also, after Albrecht, a “factual dispute” does not end, but only begins, the court’s inquiry:

[T]he plaintiff’s brief repeatedly asserts that the court should deny the summary judgment motion because there are factual disputes, implicitly suggesting that these factual disputes should be resolved at trials. . . .  These positions are obviously contradictory [with Albrecht].

Adkins, 2020 WL 1704646, at *5 (footnote omitted).  All aspects of preemption post-Albrecht, not just “clear evidence,” are legal questions for judges to decide.  Id. (“[I]t seems unlikely that the Supreme Court would hold that the first prong is triable to the jury while the second prong is not.  Both prongs involve complicated legal analysis,”).  Interestingly, Adkins relies in part on Ridings 0 for support of this proposition.  Id.

Adkins followed the same shifting burden of proof as previously stated in Roberto, and similar to that utilized in Ridings (although Ridings was not decided on summary judgment):

For a case, such as the present one, arising on summary judgment, our courts have traditionally applied a burden-shifting approach in which the defendant must initially make a showing of entitlement to judgment as a matter of law and the plaintiff must then show a genuine issue of material fact.  That approach applies even when the defendant seeks summary judgment on an affirmative defense.

Adkins, 2020 WL 1704646, at *6 (citations omitted).  “[T]here is no need to assign a burden of proof and the court’s only task is decide how the law applies in this context.”  Id.  The defendant rebuts any “presumption against preemption . . . by showing that the plaintiff’s proffer does not constitute newly acquired information.”  Id.

Adkins evaluates essentially the same purported “newly acquired information” as Roberto and Ridings, so we won’t repeat what we’ve already covered in discussing those cases.  Adkins did criticize plaintiff for indiscriminately claiming all sorts of things as “newly acquired information” and not adequately briefing most of them.  Id. at *6-7.  Typical plaintiff behavior in our experience.  From our non-Pradaxa-specific vantage point, the rest of Adkins is notable chiefly for coming to the same conclusion about medical literature with “further study” conclusions as previously stated in Ridings:

[T]he commentary’s rather tentative statement . . . and its conclusion that its proposal for monitoring “lends itself to further clinical trials” do not establish newly acquired information with any reliability.  The article presents a proposal rather than making a finding. . . .  Given the inconclusive nature of the commentary . . . it is simply unclear what change to the label the [article] would have permitted the defendants to make.  The court concludes that [it] does not constitute newly acquired information.

Adkins, 2020 WL 1704646, at *13 (citations and footnote omitted).

Ridings and Adkins are chock full of useful propositions and refutations of plaintiffs’ overreaching post-Albrecht procedural arguments.  Like the Gadolinium litigation we’ve discussed recently, the factual and legal weaknesses of the plaintiffs’ Pradaxa preemption positions confirm the statement we made last year about why we think Albrecht, on balance, will be quite beneficial to defendants.  “In 95% of preemption cases (at least), we think defendants have the better side of the regulatory record, thus we should win most straight-up preemption arguments.”

We were going back through some old cases the other day and came across a gem from our hometown court right here in in San Francisco.  It caught our eye because it deals with an angle of federal preemption on which we have written before and which we think is underappreciated, so we’re going to write about it again.  The issue is implied preemption, more specifically implied preemption where the plaintiff cannot produce “newly acquired information” supporting the drug warning that the plaintiff claims a drug manufacturer should have added to its labeling.  This is a pathway to implied preemption separate and apart from introducing “clear evidence” that the FDA would not have allowed the warning purportedly required by state law.

We will get more into these two pathways in a minute, but it all matters because of the FDA’s Changes Being Effected regulations and ultimately Wyeth v. Levine, where the Supreme Court held that a drug manufacturer could comply with both federal drug regulation and state-law duties to warn by unilaterally strengthening its labeling under the CBE regs.  No impossibility, no preemption.

We have always thought that the Supreme Court was far too generous in its reading of the CBE regulations and the extent to which they actually allow drug manufacturers to change their labels.  But even taking Wyeth v. Levine at face value, it was not the end of implied preemption.  Prescription drug manufacturers most commonly try to invoke implied preemption with “clear evidence” that the FDA considered and rejected the warning purportedly required by state law.  The example that most readily comes to mind is Dolin v. GlaxoSmithKline LLC, 901 F.3d 803 (7th Cir. 2018), where the Seventh Circuit held that federal law preempted the plaintiff’s state law warnings claims because the FDA had repeatedly instructed the company not to add the warning that the plaintiff wanted.

There is, however, more than one way to skin a cat.  Another avenue to implied preemption is to argue that the drug manufacturer could not have changed the drug warnings as the plaintiffs command because there is no “newly acquired information” to support the warning.  That too is grounded in the CBE regulations, which do not allow additional warnings based on information already submitted to the FDA when the drug labeling was approved.

This was the way to summary judgment late last year in a Pradaxa case in California state court.  See Pradaxa Cases, No. CJC-16-004863, 2019 WL 6043513 (Cal. Super. Ct. Nov. 8, 2019).  We have written on Pradaxa cases before—most recently Roberto v. Boehringer Ingelheim here, and another post here covers “newly acquired information” cases, too.  The California summary judgment order is the most succinct, particularly in setting forth a very helpful two-prong framework.

Like all Pradaxa plaintiffs, the California plaintiff alleged that the product—which is an anticoagulant—placed her at an increased risk of bleeding, and that the defendant manufacturer should have warned more strongly about that risk.  The court, however, found those claims to be impliedly preempted.  Here is the quote that you will want to focus on:

Warning preemption analysis is a two-prong analysis, assessed by the Court.  (See Merck Sharp & Dohme Corp. v. Albrecht (2019) 139 S. Ct. 1668, 1676. . . .)  “First, the plaintiff must show that there existed ‘newly acquired information’ such that the defendants could unilaterally change the label pursuant to the CBE regulation without FDA approval. . . .”  . . . Second, “if the plaintiff can point to the existence of ‘newly acquired information’ to support a labeling change under the CBE regulation, the burden then shifts to the manufacturer to show by ‘clear evidence’ that the FDA would not have approved the labeling change made on the basis of this newly acquired information.”

Pradaxa, 2019 WL 6043513, at *2 (some citations omitted).  We saw this two-prong formulation for the first time in Roberto, but we repeat it here because it so cleanly captures two avenues to implied preemption in one gloss.  The plaintiff has to produce “newly acquired information” to support the demanded change.  And it can’t be just any old evidence—it has to be meaningful, concrete, and not previously submitted to the FDA:

“[N]ewly acquired information” is defined as: “data, analyses, or other information not previously submitted to the [FDA], which may include (but is not limited to) data derived from new clinical studies, reports of adverse events, or new analyses of previously submitted data (e.g., meta-analyses) if the studies, events, or analyses reveal risks of a different type or greater severity or frequency than previously included in submissions to FDA.” (21 C.F.R. § 314.3(b) [emphasis supplied].)

. . . [I]t also cannot be rooted in conjecture or hypothesis. Rather, it must conclusively establish, by scientifically valid measurable and statistically significant data, that the different or increased risks are actual and real.

Id. at *3.  If the plaintiff cannot make this showing, i.e., demonstrate that the CBE regs can apply in the first instance, only then does the burden shift to the defendant to produce “clear evidence” that the FDA would not have permitted the proposed stronger warning.

From there, it was a short walk to summary judgment.  Plaintiffs produced evidence that they claimed would support a stronger warning, but with regard to all of it, the court found (repeatedly) that the manufacturer “submitted this information to the FDA before Pradaxa’s warning was initially approved.”  Id. at *4-*5.  The information therefore was not “newly acquired” and it could not support a label change under the CBE regulations.  Because the manufacturer therefore could not have complied with the alleged state law duties without violating federal law, the state law claims were preempted.

Not every case will involve the same factual and regulatory history as Pradaxa, but we still think these are very useful cases in understanding and applying implied preemption.

Various plaintiff-side consortia have taken it into their heads to sue every manufacturer of so-called “novel oral anticoagulants” because these products, gasp, can cause serious, and sometime fatal, bleeding incidents.  Fortunately, on the whole the plaintiffs haven’t done so well with these cases – losing almost all the trials – because jurors can be taught the dictionary definition of “anticoagulant.”  Thus, it was initially disturbing to us to read that a Connecticut state court recently upheld a plaintiff’s verdict in a Pradaxa case, Roberto v. Boehringer Ingelheim Pharmaceuticals, Inc., No. CPL-HHD CV16-6068484S, slip op. (Conn. Super. Sept. 11, 2019) (yeah, we sent it to Westlaw, but it hasn’t shown up there yet).

Then we read it.

Sure, the plaintiff scraped by with somewhat oddball verdict in a somewhat oddball case – the plaintiff had a history of gastroesophageal reflux disease (“GERD”) − but on two major issues, preemption and punitive damages, the plaintiff didn’t get what he wanted at all.  Thus, here’s another dictionary definition:

Pyrrhic victory n. A victory that is offset by staggering losses.  [After Pyrrhus]

American Heritage College Dictionary, at 1115 (3d ed. 2000).

On the merits of the underlying claim, plaintiff received a ridiculously high half million dollars on a warning claim that was almost certainly bogus.  He tried to get around the learned intermediary rule with a self-serving (and uncorroborated) claim that:

while waiting at the doctor’s office, . . . he saw a pamphlet on [the drug] and asked a nurse about it.  Ultimately, the plaintiff had a discussion with [the prescriber] and decided to change to Pradaxa.

Slip op. at 8.  That didn’t succeed in ousting the rule altogether – since it is still a prescription drug – but it conveniently provided plaintiff a second bite at the causation apple:

[A]ccording to the evidence, if the plaintiff had known that there was an increased risk of bleeding for a patient with a history of GERD, he would not have asked to switch to Pradaxa.  Although this testimony relied on the benefit of hindsight, the testimony was admissible and the jury was entitled to credit it.


The prescriber, of course knew all about the risks of bleeding – knowledge that usually precludes warning causation − but from his “general” (that is, “not specifically mention[ing] GERD”) description of his informed consent practices, “the jury could have reasonably inferred from it that [the prescriber] would have mentioned the increased bleed risks for patients with a history of GERD if the label had disclosed that risk.”  Id. at 10-11.  It was a “close call,” but there was “minimally sufficient evidence” to establish a basis for a jury to believe that a different warning could have made a difference.”  Id. at 11.  Significantly, plaintiff failed to prove the usual claim that we see advanced in this type of case:

In contrast, there is no similar testimony from the plaintiff that he would have rejected [the drug] if he learned that its label required blood monitoring. Thus, there is insufficient evidence to sustain the verdict on the theory that the absence of warnings about blood concentration and monitoring was a cause in fact of the plaintiff’s injury.

Id. at 12 n.5.

The plaintiff also supposedly proved medical causation – although we think the result in Roberto was simply  wrong under Connecticut law.  The bleed was caused by plaintiff’s GERD (specifically, by an ulcer) and there was no evidence that the drug caused the ulcer.  The Court held that but for causation didn’t matter; that plaintiff could recover “by showing that [the drug], even if not the original cause of the bleed, made it more severe.”  Id. at 14.  Plaintiff had expert testimony that plaintiff “at least was less likely to develop” a bleed compared to some unstated “different anticoagulant” – the drug “[m]ade it worse, exacerbated it.”  Id.

Roberto tellingly cited nothing to support this reasoning, and in fact the Connecticut Supreme Court has rejected just such “increased risk” causation testimony.  We discussed in this post how that dumbed down approach to causation is not sufficient in product liability, where the product must actually cause the risk, and remains controversial even in “lost chance” medical malpractice cases where it originated.  We cited Boone v. William W. Backus Hospital, 864 A.2d 1 (Conn. 2005), as one of the decisions refusing to permit mere “increased risk” causation in any circumstance.  Boone held:

[The plaintiff] must show . . . that the decreased chance for successful treatment more likely than not resulted from the defendant’s negligence.”  Thus, in order to satisfy the elements of a lost chance claim, the plaintiff must first prove that prior to the defendant’s alleged negligence, the decedent had a chance of survival of at least 51 percent.  Once this threshold has been met, the plaintiff must then demonstrate that the decedent had a decreased chance for successful treatment and that this decreased chance more likely than not resulted from the defendant’s negligence.  Accordingly, it is not sufficient for a lost chance plaintiff to prove merely that a defendant’s negligent conduct has deprived him or her of some chance; in Connecticut, such plaintiff must prove that the negligent conduct more likely than not affected the actual outcome.

Id. at 18 (citations and quotation marks omitted) (emphasis original).  Plaintiff’s “at least was less likely” thus shouldn’t cut it in Connecticut.  Roberto cites no testimony indicating that Boone’s more likely than not standard was satisfied, so we think the entire verdict should have fallen on causation grounds.  We’re not even sure the court in Roberto realized it was being asked to allow a Boone-barred increased risk causation theory.

So through the first 15 pages of Roberto, we were not happy at all.  Then we got to punitive damages.

The jury found that punitive damages were appropriate.  Id. at 3.  But under Connecticut law, the court decides how much to award.  Id. at 18 (citing Conn. Gen. Stat. §52-240b).  The court awarded the princely sum of one dollar, explaining that the jury had been hoodwinked, calling out the expert that did it, and finding no evidence to justify such damages.

[T]he court must state that it simply does not credit the testimony of Dr. Plunkett. . . .  Although Plunkett was entitled to review company documents . . ., the court believes that Plunkett engaged in unreliable mind-reading in concluding that the company put sales over safety.  The court views the company’s documents and emails very differently. . . .  [T]t is entirely appropriate for employees of a for-profit company . . . to consider topics such as cost and sales.  If a pharmaceutical company cannot make a profit selling a drug, the company would likely withdraw the drug, with all its attendant benefits, from the market.

Roberto, slip op. at 18-19 (footnote omitted).  The court utterly rejected the opinions of plaintiff’s expert on this issue:

This case is not one in which a company, motivated by greed, proceeded to ignore safety standards, defy government regulations, or disregard scientific literature in order to put an unreasonably dangerous or socially worthless product on the market.  On the contrary, all experts agreed that [the drug] provides significant benefits in reducing the risk of a stroke, with all its devastating consequences.  Fortunately, the plaintiff himself achieved this benefit and did not suffer a stroke.

Id. at 20 (footnote omitted).  That “[t]he FDA has never recalled the product and instead has approved the label some eighteen times . . . without more . . . should preclude an award of any significant punitive damages.  Id. at 21.  But there was more, and the court surveyed the evidence, concluding:

Unless the defendants are required to ignore all of these experts, articles, authorities, and examples, there is no basis for any significant punitive damages.  Accordingly, the court awards [punitive damages] in the amount of $1.

Id. at 22 (citations omitted).  Given this conclusion, it may have been more appropriate to grant judgment n.o.v. against the claim for punitive damages, but the $1 award is the functional equivalent – and the appellate standard of review is probably abuse of discretion.

And then we come to preemption.  As already mentioned, plaintiff pursued two warning-related claims, the plaintiff-peculiar GERD claim on which the court found sufficient evidence to support the verdict, and the blood monitoring claim typically advanced by all plaintiffs, which failed.  The preemption ruling in Roberto was similar.

First, “three days after the verdict,” Albrecht was decided, meaning that the excuse that preemption was for the “jury’s consideration” vanished.  Id. at 23 n.16.  Thus, the court had to decide the issue itself:

[T]he court views it as its obligation to decide the preemption issue in the first instance rather than merely pass it on to the appellate courts or have the parties waste resources taking an appeal that would result in a remand to this court to decide the very matter that the court can decide today.


In so doing, Roberto made a number if interesting preemption-related legal rulings.  Preemption in prescription drug cases involves a “two-pronged” test, the first requiring the plaintiff to establish that prerequisites to use of the CBE regulation, such as “newly acquired information” were satisfied, and the second being the so-called “clear evidence” directly at issue in Albrecht.

Post-FDA approval preemption analysis proceeds in two stages. . . .  “[I]f the plaintiff can point to the existence of “newly acquired information” to support a labeling change under the CBE regulation, the burden then shifts to the manufacturer to show by “clear evidence” that the FDA would not have approved the labeling change made on the basis of this newly acquired information.

Id. at 27-28 (quoting Utts v. Bristol-Myers Squibb Co., 251 F. Supp.3d 644, 661 (S.D.N.Y. 2017), aff’d, 919 F.3d 699 (2d Cir. 2019)).  See Id. at 28 n.9 (discussing burden of proof issues and concluding “it is fair to expect the plaintiff to come forward with the newly acquired information in question”).

The court (correctly, we believe) viewed Albrecht as applying to all preemption questions presented by either prong.  Thus whether “newly acquired information” existed was also a question of law.  Id. at 29-30.

The phrase “newly acquired information” is the key component of the CBE regulation. . . .  [I]f there is no newly acquired information, then the manufacturer is under no duty to change its label and related state failure to warn claims are preempted.

Id. at 25-26 (citations omitted).

Although the issues and analysis on the first prong of the preemption test, involving newly acquired information, are not identical to those involved in the second, clear evidence prong, it seems unlikely that the Supreme Court would hold that the first prong is triable to the jury while the second prong is not.  Both prongs involve complicated legal analysis.

Id. at 30 (footnote omitted).  The court also demurred on the supposed “presumption against preemption,” concluding that after Albrecht, it was “unclear whether the presumption applies in this situation,” particularly since “the FDA contemplated that the CBE regulation would be used ‘sparingly.’”  Id. at 31 (citation omitted).

Reviewing the case law, Roberto arrived at some conclusions concerning what could, and could not, be “newly acquired information”:

Information previously known to the manufacturer, but not submitted to the FDA, may constitute “newly acquired information,” provided that the information meets the other CBE requirements.  And, as the regulation suggests, “[n]ewly acquired information” can include either new data or new analyses of previously submitted data.  However, any claim that a drug label should be changed based solely on information previously submitted to the FDA is preempted because the CBE regulation cannot be used to make a label change based on such information.

Id. at 30-31 (citations and quotation marks omitted).  Further, because the FDA does “not allow a change to labeling to add a warning in the absence of reasonable evidence of an association between the product and an adverse event,” the Agency “contemplate[s] that the CBE regulation would be used sparingly.”  Id. at 32 (citation and quotation marks omitted).

Finally, the CBE regulation’s “newly acquired information” prerequisite was not predicated on a contemporaneous request for a label change:

First, the regulation itself does not require a specific request for a label change. . . .  Second, one can assume that the FDA, as a public agency, will . . . request a label change if the circumstances warrant.  Indeed, . . . the FDA has a statutory obligation to do so. Third, the defendants cannot be faulted if they exercise caution in submitting a study to the FDA even though they are not sure whether it merits a labeling change.

Id. at 54-55 (citations omitted).

As to the claims most Pradaxa plaintiffs bring – concerning some blood plasma concentrations and the monitoring of same – the court held that no “newly acquired information” existed to permit the manufacturer’s resort to the FDA’s CBE regulation, as of plaintiff’s January, 2014 claimed injury.  Roberto, slip op. at 50.  “The opinion goes through a plethora of Pradaxa-specific information, that anyone interested in the details can read, but:

In sum, after review of the numerous articles and reports identified by the parties in their briefs, the court concludes that there was no newly acquired information that would have allowed the defendants to make a label change on their own on the topics of Pradaxa blood concentration levels or blood monitoring.  Accordingly, the court finds these claims preempted.

Id. at 34-50.

But as discussed at the beginning of this post, the plaintiff had a second warning claim specific to his situation – lack of a warning about greater bleed risks to GERD sufferers like himself – indeed, that GERD claim “was the only claim for which there was sufficient evidence to prove causation in fact.”  Id. at 50.  As to this claim, the timing of the defendant’s submission to the FDA was “unclear.”  Id. at 55.  Thus, “the court cannot conclude that the GERD information was previously submitted to the FDA.”  Id. at 56.  The second prong of preemption also was not satisfied.  “[T]he defendants’ argument amounts to a second prong ‘clear evidence’ claim” that lacked either proof of what information the FDA received or a definitive FDA disapproval of a warning change.  Id. at 56-57.  Since it does not appear that the defendant asserted a “clear evidence” defense to plaintiff’s GERD claim, that claim survived preemption.

Thus, the favorable preemption holding in Roberto is applicable to the majority of Pradaxa plaintiffs asserting claims concerning blood concentration and monitoring (at least for injuries predating January, 2014).  Only those few Pradaxa plaintiffs with GERD-related claims can make use of the second, adverse ruling.  A Pyrrhic plaintiff’s victory indeed.

Finally, for those who are gluttons for punishment, Roberto rejected largely case-specific new trial arguments concerning non-FDA-approved drug dosages, misconduct by plaintiff’s counsel, and evidence of foreign GERD labeling.  Id. at 57-62.  The defense also lost a Connecticut-specific “phantom damages” argument concerning recovery of medical charges that were billed but then written off by the government.  Id. at 62-65.

Happy birthday, Bob Marley. (We mean the transcendent reggae singer, not the Maine comedian.) Now let’s get together and feel alright about another good personal jurisdiction decision, In re Pradaxa, No. CJC-16-004863 (Cal. Super. Ct. Jan. 31, 2019). The case strikes a blow against California litigation tourism. There were some awful decisions out of California on this topic in the past. Call this new decision a redemption song.

A bunch of non-California residents claimed injuries from Pradaxa, and sued a number of corporate defendants associated with the medicine’s manufacture and sale. But none of those corporate defendants was incorporated in California, nor did they own, lease, or maintain any property in the Golden State. The defendants challenged personal jurisdiction. The plaintiffs did not even argue that there was general personal jurisdiction over the defendants. The Bauman case thoroughly foreclosed that notion. Instead, the issue was whether the SCOTUS BMS decision left any room for jammin’ the out of state defendants into a California court via specific personal jurisdiction.

Remember, the plaintiffs took the medicine outside California. So what bases could the plaintiffs lively up themselves to show that their claims related to or arose out of the defendants’ contacts with California? The plaintiffs did what some other plaintiffs have done by exploiting the existence of an in-state clinical trial of the drug. If the mere existence of clinical trials does the trick, then the SCOTUS BMS case is a dead letter for pharma companies, since clinical trials often take place in big (and plaintiff-friendly) states. That would be a crazy baldhead result, given that BMS itself involved a pharma defendant. The Pradaxa court was too smart for that. It looked to the qualitative and quantitative nature of the clinical trial in California, and concluded that, in the grand scheme of things, questions relating to 32 in-state clinical trial sites in one massive clinical trial were “too attenuated to support the exercise of specific jurisdiction.” All non-resident Pradaxa plaintiffs were consequently dismissed from the California mass tort for lack of specific personal jurisdiction under BMS. Their exodus is our freedom time.

Since it is a sure thing that forum-shopping plaintiff lawyers will continue to pursue the clinical trial angle, you should pay heed to the factors the California court considered in finding the clinical trial insufficient to establish specific jurisdiction: (1) the forum state was not overrepresented in the trial, and (2) the alleged problems with the trial did not relate to the claimed inadequacies in the warnings. The plaintiffs made much of the fact that there had to be corrections made to label with respect to the adverse event reports out of the California clinical trial, but the “negligible changes in the data” could not support claims. (E.g., the hazard ratio for a life-threatening bleed went from 0.80 to 0.81.). The court was not impressed by the plaintiffs’ argument.

But we are impressed by the rigor and clarity of the court’s reasoning. If corporate defendants can earn such a good and sensible result in San Francisco, we all have cause for optimism. Hallelujah. Don’t worry about a thing. Could you be loved? Every little thing is gonna be alright. And never give up. We offer congratulations, gratitude, and a tip of the cyber hat to Eric Hudson at Butler Snow, who argued and won the motion.

By the way, speaking of congratulations, and speaking of never giving up, today is the birthday of another pop star. In fact, according to an MTV Europe poll in 2008, he is the “Best Act Ever.” We won’t tell you who he is; you’ll have to click on the link at the end. Of course, since we’re telling you to click on a link, you might have some idea what awaits you. Feeling dread? Don’t. Embrace the wonderful, sheer inanity of the Best Act Ever.

Bexis just turned 65 (on 1/25/2021) – the classic retirement age.  That’s an occasion to look back and evaluate what’s gone on over the course of an entire legal career.  So how have we done, as defense lawyers, over the course of our entire careers, at our primary job – which is to prevail for our clients in drug and medical device product liability litigation?  We’re lawyers.  We work for clients.  Thus, the bottom line, in a legal career, as in an individual matter, is:   Are our clients better off now than when we started?

Bexis started practicing law in 1982 (in the depths of a legal recession, he received only one offer from a law firm in Philadelphia, where he wanted to practice).  So, we’ll take that as a starting point.

Personal Jurisdiction

Great progress here.  For decades any of our corporate clients were subject to “general” personal jurisdiction (being sued for anything occurring anywhere) in any state where they did “continuous and substantial” business.  Beginning in 2010, we started following personal jurisdiction decisions.  Since then, general jurisdiction has largely been defanged – limited to a corporation’s principal place of business and state of incorporation.  A huge victory, Bristol-Myers Squibb Co. v. Superior Court, 137 S. Ct. 1773 (2017) (“BMS”), has raised the possibility of elimination of litigation tourism altogether, but follow-up has been mixed.

Since BMS, California, Illinois and New York mostly respect the Supreme Court’s decisions, but Pennsylvania does not.  Personal jurisdiction has also largely put a stop to multi-plaintiff, misjoined complaints as a way of defeating diversity jurisdiction and aggregating claims where plaintiffs want them.  Perhaps by this June there will be more clarity, when the Supreme Court tackles stream of commerce jurisdiction post-BMS.  But those cases do not involving forum shopping plaintiffs, so it’s hard to say.  Another opportunity, yet to be realized, could use personal jurisdiction as a basis substantially to restrict the use of nationwide class actions.  Even where we are, though, is an incomparably better place than where our clients were when Bexis’ generation of defense lawyers began their practices.


And how about preemption in tort litigation?  Back when Bexis’ generation was getting started, it essentially didn’t exist.  Thank/blame cigarette defense lawyers (something Bexis also did for a while at a prior firm) for changing that, beginning with Cipollone v. Liggett Group, Inc., 505 U.S. 504 (1992).

It didn’t start well for us.  We manned the barricades during the first wave of FDCA preemption litigation, in vaccine cases, but got clobbered.  We took it on the chin again, in Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), which was decided only a few months after Bexis had succeeded in getting about 85% of the claims in the Bone Screw litigation preempted.  Back to the drawing board.  What’s left after Lohr?  Well that claim the Bone Screw plaintiffs were pushing for “fraud (on the FDA)” should be preempted because it’s based on the FDCA, and an obscure section of that statute prohibits private enforcement.  Try implied preemption, write it up, and let’s see what happens.  Presto:  We get Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001), a few years later.

Then we get a degree of medical device preemption back in Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), with extensive preemption with respect to pre-market approved medical devices (a minority of all devices), with the boundaries still being fought over.  The pendulum swung again with a bad loss in Wyeth v. Levine, 555 U.S. 555 (2009), that left our ears ringing for a while.  But our side pulled itself off the deck to score wins in PLIVA, Inc. v. Mensing, 564 U.S. 604 (2011), Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013), and to a large extent in Merck Sharp & Dohme Corp. v. Albrecht, 139 S. Ct. 1668 (U.S. 2019).  Thus, even in prescription drug cases, there is now a lot of preemption to be had, as our drug preemption cheat sheet shows.  With some help from Congress, our side even obtained a measure of revenge for our vaccine clients in Bruesewitz v. Wyeth LLC, 562 U.S. 223 (2011),

So while preemption is still a major battleground, is now represents our clients’ most powerful defense.  When we started out, back in the 1980s, it essentially didn’t exist.  Preemption is another major accomplishment of Bexis’ generation of defense lawyers.

Class Actions

Another greatly positive development has been the virtual eviction of class actions from product liability claims involving personal injury.  If you wonder why class actions rank this high, you don’t remember what it was like in drug/device litigation in the 1980s and 1990s.  Back when Bexis’ generation started their legal careers, we had to take class actions in product liability litigation very seriously.  While there were never a lot of certifications, there were enough of them that – during the Fen-Phen and Bone Screw litigations, for example – plaintiffs would argue that there was some sort of “modern trend” favoring certification of personal injury class actions.  Some courts so held.  See In re A.H. Robins Co., 880 F.2d 709, 738 (4th Cir. 1989) (later abrogated).  Bexis remembers how relieved we were beating the class certification motion in Bone Screw, and now depressing it was when the opposite happened (at least in a state court) in Fen-Phen.  One turned into a broad defense victory, while the other produced one of the most abused settlements in mass tort history.

Then our side of the “v.” prevailed in Amchem Products, Inc. v. Windsor, 521 U.S. 591 (1997), and Ortiz v. Fibreboard Corp., 527 U.S. 815 (1999).  Since then – with a lot of blood, sweat, and good legal argument from our side – class actions (at least successful ones) largely disappeared from mass torts, as we’ve discussed before.  Check out our federal and state class action cheat sheets.  The few courts willing to certify class actions in drug and medical device cases have so far gotten shot down on appeal (in the St. Jude litigation) or threw in the towel (on Zyprexa).  Following enactment of CAFA, most class action decisions going forward, and essentially everything in mass torts, will be made by federal courts applying post Amchem/Ortiz law.

Medical monitoring, a non-personal-injury derivative of personal injury causes of action that the plaintiffs’ bar dreamt up with class actions in mind, has largely failed in recent years to produce very many successful certifications – despite lots of attempts.  Thank defense counsel for ensuring that the claimant-specific elements of medical monitoring have been preserved, even as some courts have recognized the claim.

Likewise, class actions involving purely economic losses, usually brought as adventurous applications of consumer fraud, RICO, or warranty claims, have had rough going – at least in the drug/device field.  And if the United States Supreme Court ever finds a case that it can review on the merits, we think that the abusive cy pres practice of giving supposed “class” recoveries to pro-plaintiff charities will be defeated.

As a measure of how far out of the mainstream tort class actions have become over the last couple of decades, the ALI’s Aggregate Litigation principles project, for all its pro-plaintiff leanings in other areas of the law, states quite clearly that personal injury class actions are disfavored for a variety of reasons.

There’s also a distinct trend afoot, not limited to tort cases, to tighten consideration of class action allegations.  We have the rise of ascertainability.  The old rule of no “merits” consideration during class certification is out the window.  The Supreme Court is going to consider another of our bêtes noir, the no-injury economic loss class action.  Maybe more favorable movement will occur.

To top it all off, our side has also had a good deal of success arguing against cross-jurisdictional class action tolling – that failed class actions filed in one court should not toll the statute of limitations on claims filed in a different court.  That deprives failed class actions of the one substantive benefit that they could confer upon plaintiffs (as opposed to their lawyers).

We’re still litigating economic loss and similar no-injury class actions, but overall our clients are a lot better off on the class action front now than they were when we got into this business.

Expert Witnesses

One word:  Daubert.  Back when Bexis’ generation started practice, the courts waved through just about any garbage that a plaintiff’s expert wanted to say.  See Wells v. Ortho Pharmaceutical Corp., 788 F.2d 741, 744-45 (11th Cir. 1986) (allowing testimony with no epidemiologic or other statistically significant support for the proposition that spermicide, intended to prevent conception, was defective for purportedly causing birth defects).

Then came Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U .S. 579 (1993).  At first the effect of Daubert was unclear.  It could have been interpreted as loosening the already capacious federal standard for expert certification even further.  But Bexis’ colleagues on the right side of the “v.,” with a lot of hard work and inspired argument, gained the upper hand and Daubert is now synonymous with more scientifically rigorous expert admissibility standards.  What mattered most wasn’t the language standard itself, or of Rule 702, but Daubert’s requirement that judges – not juries – act as “gatekeepers” of admissible expert testimony.  Given the amount of junk science that plaintiffs’ experts were spewing, if we could just get courts believing that they had an obligation to review things critically, we would win.

Daubert was a drug case. It was the Bendectin litigation’s lasting gift to the legal profession.  It’s taken several return trips to the Supreme Court to nail it down.  See General Electric Co. v. Joiner, 522 U.S. 136 (1997); Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999); Weisgram v. Marley Co., 528 U.S. 440 (2000).  And we may well need more, since some federal courts, particularly the Ninth Circuit, have been lax at enforcing what we think Daubert intended to accomplish.

After a while, the Daubert divide’s gotten to be like night and day.  We don’t win every case, but we win a lot more of them than before.  Nineteen years after Wells, the same court decided McClain v. Metabolife International, Inc., 401 F.3d 1233 (11th Cir. 2005), reversing and requiring judgment n.o.v. where an expert relied on little more than temporal association. That’s monumental change for the better.  These days, a couple of MDLs a year are wiped out by favorable Daubert expert exclusion decisions.

Another beneficial aspect of Daubert – stringent substantive review of expert opinions, by whatever name – is increasingly finding its way into state court decisions as well, in places like New York, Texas, New Jersey, and even Florida.  It’s not perfect, but compared to where Bexis’ baby boom cohort of defense counsel were when their careers started, this is another area where we think that, after all our laboring in the litigation vineyards, our clients are a lot better off.


Another accomplishment of Bexis’ generation of defense counsel is strengthened requirements for pleading actual facts, rather than legal conclusions, in federal court.  The blog has been all over Ashcroft v. Iqbal, 129 S. Ct. 1937 (U.S. 2009), and Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), from the beginning.  We even popularized a shorthand for what the Supreme Court did – “TwIqbal.”

TwIqbal has every right to be popular, at least among our clients and colleagues.

Before these decisions, the federal pleading standard was a joke.  Plaintiffs could survive a motion to dismiss without pleading a single actual fact, only the same boilerplate they could repeat over and over again in thousands of identical complaints, with only the names changed to encourage the greedy.

Now, under TwIqbal’s “plausibility” standard, things have gotten much better.  How much better?  Well, our TwIqbal cheat sheet includes only drug/device cases, and imposes a very high bar for adding cases – no claims must survive a motion to dismiss, no matter how good some of the rulings might be.  It now contains almost 300 cases.  In particular, TwIqbal has done an excellent job in combination with post-Riegel device preemption cases, with most (but not all) courts requiring plaintiffs to plead facts, not “magic words,” when claiming purported FDCA violations.  That’s very concrete improvement.

And it’s carried over to other areas as well.  Our TwIqbal cheat sheet includes a wealth of cases requiring plaintiffs to plead things like warning causation in learned intermediary rule cases, a deviation from intended result when a manufacturing defect is alleged, and the language of any supposed express warranty.

So from the outset of litigation – the adequacy of the initial pleadings − our clients are a lot better off than when we first got our seats at the table.

Learned Intermediary Rule

When we started out, the learned intermediary rule was already widespread, but today it is essentially universal.  Our state-by-state synopsis of learned intermediary rule cases now has something favorable from every jurisdiction in the country.  Only one state ever rejected the rule, and that has been overruled by statute.  On the other side of the ledger, Bexis alone has been involved in decisions adopting or reaffirming the rule in seven states (PA, NJ, OH, CT, KY, TX, AZ).  All told, the number of jurisdictions with high court or statutory adoption of the learned intermediary rule has reached 36.  The most populous state with less than high court adoption is Indiana, where the intermediate appellate court has followed the rule since 1979.

Beyond simply the number of states adopting the learned intermediary rule, we’ve also seen a strong trend towards its expansion in various directions. It’s expanded from drugs to medical devices. The rule has grown from being mostly about adequacy of warnings to now having its greatest use in determining whether an allegedly defective warning had any causal effect.  If the learned intermediary physician doesn’t read the warning (or knows what a “better” warming allegedly would reveal), there’s no causation.  The rule has also expanded from failure to warn claims to other claims such as consumer fraud.  The rule has also been increasingly applied to protect entities like pharmacists, in addition to product manufacturers.

While the learned intermediary rule isn’t as powerful a defense as preemption or Daubert, it is an effective way of getting rid of individual cases, since in most cases, the prescribing physician is an independent witness who does not say whatever plaintiff’s counsel wants.

So back when Bexis’ generation got started, while the learned intermediary rule was a majority rule, the other side (and its academic fellow travelers) still had ambitions of getting rid of it.  Those ambitions have been crushed.  So with the learned intermediary rule as well, we’d have to say that our clients are quite a bit better off now than when 65-year-olds began practicing law.

Prevention of Innovative Liability Theories

In the early 1980s, when Bexis’ peers graduated law school, market share liability was a major threat to burst its DES bounds and become generally accepted.  That hasn’t happened.  No state has adopted it since Hawaii in 1991, and some of the states that did so earlier, like New York, have tightly confined it to the original DES set of facts. Score one for the good guys.

The third Restatement of Torts, adopted in 1997 and published the following year, cut back on some of the loopier aspects of strict liability, including liability for unknowable risks, and failure to recall/retrofit claims.  Those claims haven’t gone much of anywhere, either, and we have been able to cut back on the loopy idea that “negligence” must be kept out of strict liability.  We’ve largely kept an independent duty to test out of the law, too.  Fraud on the FDA got preempted out of existence.

The advent of broad preemption (discussed above) has led to plaintiffs arguing for a variety of novel liability theories.  The most dangerous of these is innovator liability – imposing liability for injuries caused by (preempted) generic drugs (90% market share) on the (unpreempted) manufacturers of bio-equivalent branded products (10% market share).  While two outlier states have adopted innovator liability, the feared rush has not materialized, and the defense appears to have stemmed that tide.  Another bizarre theory is “warning” liability based on alleged failure to report adverse events to the FDA.  Since the overruling of a bad Ninth Circuit prediction of Arizona law by the Arizona Supreme Court, that theory, too, is resembling a bad idea whose time has passed.

Our desire to combat novel theories is why we have pushed Erie conservatism on the Blog since day one.  Since so much of drug/device litigation is in the federal courts, emphasizing federalist restraints on “prediction” of novel liability is one way of combating judicial activism in this area.

But we’ve hardly been universally successful.

Despite our best efforts, consumer fraud claims have become staples of our opponent’s litigation strategy, and thus banes of our existence.  When Bexis’ cohort first started practicing law, practically nobody ever encountered them.  Now, they’re routine.  However, in most instances, courts have ruled that such statutes can’t be enforced extraterritorially, outside of the state that enacted a particular statute.  That at least cuts down on the size of any attempt to aggregate claims.

Another negative on Bexis’ watch is that medical monitoring went from a legal peculiarity to, if not a majority rule, at least being allowed by a fair number of states, as our 50-state survey shows.  So we haven’t been able to stop that one either.  Something else we didn’t see much of back in the 1980s was the so-called post-sale duty to warn.  That’s proliferated quite a bit, as even the Third Restatement included it.  Fortunately, there aren’t that many post-sale duty claims in our neck of the woods, and they tend to be preempted.

We’ve waged a see-saw battle with negligence per se claims based upon alleged FDCA violations.  Many the older cases that were around when Bexis’ cohort was getting started allowed those claims without a lot of discussion, because after all the FDCA was enacted to make products safer, wasn’t it? However, the principle that the FDCA prohibits plaintiffs from privately enforcing the statute against violators, enunciated by the Supreme Court in Buckman, has helped our clients defeat those claims more often in recent years.  We’ve also pushed a variety of state-law arguments.  But negligence per se hasn’t yet gone the way of the dinosaurs, and some courts still allow such claims.

Finally, we thought we had “public nuisance” beat in the drug/device arena.  But lately, it has made something of a comeback.

All this adds up to a mixed record in beating back the various novel theories of liability that plaintiffs have invented over the years.  We’ve gotten rid of some altogether, and limited others.  But some genies have escaped from the bottle despite the best efforts of our generation of defense lawyers.

First Amendment and Off Label Use

When Bexis’ generation started, the First Amendment had just been extended to commercial speech.  The notion that the First Amendment could limit the FDA’s ability to prohibit truthful manufacturer speech about off-label uses hadn’t even been invented yet.

In the Bone Screw litigation, plaintiffs sought to make off-label use – physician employment of drugs and devices for uses beyond what FDA-regulated labeling stated – into some kind of tort.  Buckman killed that, recognizing off-label use as legal, “accepted and necessary,” and “generally accepted.”  531 U.S. at 350-51.

Now we have Sorrell v. IMS Health, Inc., 564 U.S. 552 (2011), United States v. Caronia, 703 F.3d 149 (2d Cir. 2012), and similar cases.  The FDA has started to back-peddle, and given where the Supreme Court seems to be going on strict scrutiny of content- and speaker-based restraints on commercial speech, it may just be a matter of time before the FDA’s ban falls.

When Bexis started raising this issue in the Bone Screw litigation, people chalked it up to overzealousness on behalf of his clients.  Now we have the Medical Information Working Group.  His generation of lawyers made something out of nothing here.


Now we get into our failures.  Even though we’ve had some success modifying the federal rules of procedure, electronic discovery remains mostly unfettered as a new and excessive cost to our clients.  When Bexis’ generation started out, there was no such thing.

Now electronic discovery has gotten entirely out of hand.  It’s hideously expensive, ridiculously intrusive, and almost entirely a one way street.  It often leads to satellite litigation over discovery.  Although we’ve advocated electronic discovery for defendants – to the point of having a cheat sheet − tort plaintiffs simply don’t often have large, frequently upgraded computer systems that lead to the imposition of astronomical discovery costs.  In large litigations, maybe one in a million of electronically produced documents is ever used at trial.

Everything else that our side’s been able to accomplish in limiting or streamlining discovery – routinized plaintiff questionnaires, federal-state coordination, restrictions on apex depositions, the inadvertent production doctrine, etc. – pales by contrast to the constantly metastasizing disaster that is electronic discovery.

Reducing Overall Litigation

Utter, absolute failure.  Mass tort MDLs now make up half the total federal case load.  The other side has been much more efficient in soliciting large numbers of plaintiffs to populate the ever growing number of pharmaceutical and medical device mass torts than our side has been in stopping them.  The racket that mass torts have become is so downright predictable that we parodied it years ago – but had to update that parody because in MDLs, irony is dead.

Beneath everything lies the simple fact that, since the Supreme Court’s first benighted decision in Bates v. State Bar of Arizona, 433 U.S. 350 (1977), extending First Amendment protection to lawyer advertising, the other side’s solicitation machines have become more and more effective, and there’s not a constitutional thing we can do about it.  Even when our side gets a crumb from the Supreme Court, such as Florida Bar v. Went For It, Inc., 515 U.S. 618 (1995), upholding a trivial 30-day cooling off period from personalized solicitations, the vote was only 5-4.

Mass torts are all about volume.  As we lamented in “Anatomy of a Mass Tort 2.0”:

Regardless of what is the triggering event, the creation of the mass tort itself follows – almost entirely extra-legally.


Since the courts discovered that lawyers have a constitutional right to use advertisements to solicit clients, they have done so – with a frequency that would no doubt have shocked the Founding Fathers.  Anybody who watches TV knows what we’re talking about.

And it’s not just lawyers, either.  Media specialists and third-party litigation funders also combine to run (with apologies to Marlowe) the race that launched 100,000 TV (and internet) ads.  Anyone who responds – and many do – becomes “inventory,” their claims available for sale in bulk to enterprising lawyers hoping to cash in on the next mass tort.

Thus, within hours after the triggering event, “law firm” websites will invite product users to sign on as potential clients.  Similar solicits in other forms of media follow.

This early stage is usually the point of no return.  The lawyers involved will freely buy, sell, and trade the would-be clients that have signed up – to the point where those clients often don’t know who their lawyers actually are.  But once they have clients, no matter how attenuated the relationship, lawyers cannot ethically let them go.  Regardless of the facts, the mass-tort litigation process takes on a life of its own.

As long as society tolerates virtually unlimited lawyer solicitation as a constitutional right, there’s not a lot of ways for our side to close the litigation floodgates.  So in this most fundamental of our client responsibilities – protecting them from litigation – Bexis’ generation of defense counsel has not been able to do what their clients need most.

Not that we haven’t tried; it’s just our side’s efforts to stop the onslaught of boilerplate, virtually uninvestigated filings hasn’t accomplished very much.  Lone Pine orders and plaintiff fact sheets are a handy invention, but they only channel and divert the torrent; they don’t stop it.  Nothing seems to.  Look at the list of the most popular drugs of 10 years ago.  Well over half of them have been the subject of coordinated mass torts – usually MDLs.  Are most of the widely prescribed drugs defective?  Is the FDA that incompetent.  Of course not.  Those drugs are simply the ones with the volume of sales to support the current mass tort racket.

And our side, Bexis’ generation, has been unable to stop this trend.  If everything that we do is ultimately supposed to contribute to deter future litigation against our clients, then it hasn’t worked at all.  Maybe it’s time for us to get out of the way and let the next generation of defense lawyers try their luck.

With 2020 mercifully coming to an end, it is once again time for the Drug & Device Law Blog’s top ten decisions of the year.  In keeping with COVID-19’s dominance of 2020, we present our top ten in the context of countermeasures against another social ill – the tort pandemic raging across much of the nation.  Some of the cases here will keep our clients socially distant from aggregate litigations in superspreader venues where only plaintiffs want to be.  Others resemble personal protective equipment, shielding our clients from harmful allegations in conflict with regulatory oversight.  Others are more like vaccines, inoculating our clients from dangerous liability theories.  Still others more resemble the central attribute of the effort to create COVID-19 vaccines – that science must be respected and the crackpot theories peddled by the other side would do more harm than good.

Next, our annual reminder.  Our scope of coverage is limited.  Our top (and bottom) ten lists are limited to cases involving prescription medical product liability litigation (relatively broadly defined).  Cases are decided each year that significantly impact what we do, even though not involving prescription medical products.  So, we’ll salute Rockhind v. Stevenson, 236 A.3d 630 (Md. 2020), adopting Daubert; Berry v. the City of Chicago, ___ N.E.3d ___, 2020 WL 5668974 (Ill. Sept. 24, 2020), rejecting no-injury medical monitoring; Texas Brine Company, L.L.C. v. American Arbitration Association, Inc., 955 F.3d 482 (5th Cir. 2020), in which a third straight court of appeal approved of removal before service; and Exela Pharma Sciences, LLC v. Sandoz, Inc., ___ F. Supp.3d ___, 2020 WL 5535026 (W.D.N.C. Sept. 15, 2020), preempting attempts to second guess FDA product decisions in commercial litigation.

Enough with the introductions.  Now, let’s proceed with the protective measures.

  1. Rios v. Bayer Corp., ___ N.E.3d ___, 2020 WL 2963318 (Ill. June 4, 2020).  For decades, litigation tourist plaintiffs looking for generous juries and favorable rulings have flocked to Madison, Cook, and St. Clair counties in Illinois.  After Rios it’s time to start mucking out these Augean stables.  Rios applied the Supreme Court’s BMS (2017+1) decision the way it was intended, to prevent any plaintiff from anywhere from relying on secondary or immaterial forum “contacts” that had nothing to do with their cases as a basis for personal jurisdiction.  160 out-of-state plaintiffs (suing in Madison County) advocated a “loose and spurious” form of general jurisdiction under the guise of “specific” jurisdiction.  They cobbled together supposed forum “contacts” that had nothing to do with their cases − “clinical trials,” us[ing] the state . . . for [a] physician training program,” “orchestrat[ing] a marketing campaign” – even though they weren’t in the trials and were unaffected by either the training or marketing.  In stark contrast to our worst case of 2020, Rios refused to allow “specific” jurisdiction to be based on “contacts” so general that any plaintiff from anywhere could assert then.  Without the plaintiffs themselves being prescribed, having purchased, or ever using the product in state, no “meaningful” basis for meeting the “arising from”/”relating to” standard for specific personal jurisdiction existed in Rios.  “Illinois has no particular interest in resolving claims that did not arise out of or relate to activities that occurred here.”  Further, all these “out-of-state plaintiffs could pursue their claims elsewhere.”  And since a number of the plaintiffs in fact had filed suit elsewhere, judicial economy also supported dismissal.  Cleaning up an earlier mess, Rios also expressly overruled the earlier M. (2016-8) decision which had held that clinical trials established jurisdiction even for non-resident plaintiffs who were never trial subjects.  Since Illinois has already rejected general jurisdiction by consent, that means that litigation tourists are now out of personal jurisdiction arguments in Illinois.  We rhapsodized about Rios here.
  2. This space intentionally left blank.
  3. Boone v. Boehringer Ingelheim Pharmaceuticals, Inc., 239 A.3d 1175 (Conn. 2020). The Connecticut Supreme Court unanimously affirmed the complete preemption of prescription drug design defect claims.  State judges are generally less friendly towards federal preemption of their state’s law than federal courts, so that’s an accomplishment right there.  Applying one of our favorites − the “Mensing (2011+1) independence principle” – Boone preempted a rather unusual contention that the defendant should have not have sold an FDA approved product until also getting FDA approval of a supposed “antidote.”  This purportedly common-law claim rather explicitly sought to hold hostage an FDA approved drug until a separate, later FDA action approving a different drug.  Beyond being contingent on agency action, that claim was further preempted as a “stop selling” claim under Bartlett (2013+1) because the only way to avoid liability under the plaintiff’s theory was not to sell the FDA-approved product at all, pending development of a second drug.  The court explicitly rejected plaintiff’s argument (which we see all the time) that Mensing and Bartlett should be limited to generic drugs, and instead agreed with Yates (2015+1) to hold that Supreme Court’s implied preemption analysis was broadly applicable.  Also rejected was another common claim from the other side, that Albrecht (2019+1, 2019-6) required all prescription drug implied preemption arguments to jump through the hurdles of the “clear evidence” test.  That test only applies if the FDA’s CBE regulation applies and allows unilateral warning revision.  CBE Bartlett is not a basis for changing drug design or availability.  Finally, the plaintiffs’ claim that it “pre-approval” design defect claims escape preemption was also rejected as another variant of a stop selling claim.  We went bonkers for Boone here.
  4. In re Mirena IUS Levonorgestrel-Related Products Liability Litigation (No. II), ___ F.3d ___, 2020 WL 7214264 (2d Cir. Dec. 8, 2020). Our best Daubert decision of 2020 was decided earlier this month.  Short but oh so sweet, the Second Circuit affirmed in toto In re Mirena IUS Levonorgestrel-Related Products Liability Litigation (No. II), 387 F. Supp.3d 323 (S.D.N.Y. 2019), which was our 9th best case of 2019.  The court properly held that none of the plaintiffs’ experts offered general causation opinions that could withstand any degree of scientific examination.  This result seems obvious – the condition at issue is “idiopathic intracranial hypertension,” and “idiopathic” means precisely that cause cannot be determined – but other circuits have been much more hospitable to junk science.  Not Mirena.  Plaintiffs first argued that the inquiry into their experts’ methods was too rigorous.  Mirena court affirmed that Daubert requires a “hard look” at expert opinions and that “an expert’s methodology must be reliable at every step of the way,” emphasizing the necessity of “rigorous examination.”  “[N]ot only was it appropriate for the district court to take a hard look at plaintiffs’ experts’ reports, the court was required to do so to ensure reliability.”  Then, with almost Shakespearean inconsistency, plaintiffs complained about the trial court considering their experts’ conclusions, as opposed to their methods.  Wrong again.  Mirena holds that it is entirely proper, as a Daubert factor, to consider that an expert’s conclusions are not generally accepted.  Mirena also affirms that support by published studies may be required where the expert’s methods are not otherwise reliable, and rejects differential diagnosis as the sole basis for a general causation opinion.  Affirming summary judgment, Mirena also recognizes that “there is a general causation requirement across all states.”  We marveled at Mirena here.
  5. In re Taxotere (Docetaxel) Products Liability Litigation, 966 F.3d 351 (5th Cir. 2020). The passive aggressive approach many MDL plaintiffs and their counsel take towards their obligations to complete “fact sheets” or similar disclosure documents is one to the banes of any MDL defense counsel’s existence.  Missed deadlines and incomplete answers abound and require inordinate amounts of work keeping track of deficiencies and notifying plaintiffs repeatedly of their failure to meet their obligations.  MDL deficiency proceedings drag on for many months.  That’s what happened in Taxotere.  After more than four months of inaction, the plaintiff hadn’t done a thing – no submission at all.  Only after the MDL court set a hearing did plaintiff make a woefully inadequate submission.  After a judicial sanctions warning, followed by yet another round of extensions and deficiencies, the court had finally had enough and dismissed the case with prejudice.  Plaintiff appealed and lost, bigtime, in a rare published appellate opinion on this topic.  Taxotere is a long overdue reaffirmation of the power of MDL judges to impose the ultimate sanction of dismissal with prejudice on MDL plaintiffs who only want to wait for settlement, while doing nothing.  Taxotere rejected plaintiff’s dithering, multi-factor argument in favor of a simple review of the MDL court’s discretion to dismiss – was there a “clear record” of delay, and would “lesser sanctions” suffice.  In Taxotere there was extensive delay, which in and of itself demonstrated that no lesser sanction would stop it.  No fact sheet; no case.  Dismissal sanctions had to be real and enforceable.  We touted Taxotere here.
  6. This entry is from the non-RS side of the blog. Dolin v. GlaxoSmithKline LLC, 951 F.3d 882 (7th Cir. 2020). Again?  The Seventh Circuit reaffirms its prior decision that we named as fifth best decision of 2018.  The court rejected an argument that Albrecht (2019+1, 2019-6) precluded defendants from ever having “clear evidence” that the FDA would have rejected the warning in question unless the particular defendant itself had unsuccessfully submitted the particular warning to the FDA.  Instead, the FDA’s decision to impose classwide labeling on the defendant’s drug that did not contain the language plaintiff advocated sufficed as “clear evidence” of an adverse FDA’s decision − whether or not the particular defendant being sued actually submitted a label change.  Albrecht did not repudiate the “would have” rejected standard in Levine (2009-1), but merely clarified it.  Rather than overrule or make huge changes to Levine, the principal holding in Albrecht was that preemption is a question of law.  The record showed that the FDA had “full information” from the defendant, so preemption was not defeated by the FDA deciding to act more broadly than in a product-specific manner.  Classwide labeling is formal enough to be a preemptive FDA action.  Thus, the court emphasized the parts of Albrecht we liked and minimized those parts we did not.  We (well some of us) discussed Dolin here.
  7. In re Viagra (Sildenafil Citrate) & Cialis (Tadalafil) Products Liability Litigation, 424 F. Supp.3d 781 (N.D. Cal. 2020).  These favorable Daubert decisions should spell the end of another MDL.  All of plaintiffs’ general causation experts excluded after a 4-day Daubert hearing.  A single medical journal article reporting an “association” was not enough.  Association is not causation.  Plaintiffs’ experts’ Bradford Hill analysis was result-driven and truncated, relying on a single factor – association – to the exclusion of the other eight factors.  Plaintiff’s theory was not supported by any scientist, researcher, regulatory agency, or other qualified person, save their own paid experts.  The science was all the other way.  Lots of things may be biologically “plausible,” but that doesn’t mean they have any actual scientific support.  Science worked.  An article found an association and suggested further study.  Other scientists took up the challenge, but nothing more significant than the initial association has been found.  A very good decision, particularly from a court in the Ninth Circuit.  We vouched for Viagra here.
  8. Keen v. C.R. Bard, Inc., ___ F. Supp.3d ___, 2020 WL 4873634 (E.D. Pa. Aug. 19, 2020) & Keen v. C.R. Bard, Inc., ___ F. Supp.3d ___, 2020 WL 4818801 (E.D. Pa. Aug 19, 2020). Two excellent decisions handed down the same day.  These could just as well have been issued as a single opinion, so we consider them together.  The first decision is notable for its solid reaffirmation of established Pennsylvania law precluding strict liability in prescription medical product liability litigation.  Pennsylvania appellate decisions involving prescription medical product liability litigation are unanimous in rejecting strict liability, and so are vast majority of federal cases that purport to apply Pennsylvania law.  Also cites nonbinding jury instructions.  The other parts of the opinion, denying summary judgment on other claims, are routine and pale by comparison.  The second Keen opinion concerned trial evidence, and rejected the fallacy − adopted primarily in pelvic mesh cases like Eghnayem (2017-2), and Cisson (2016-2) − that evidence of FDA §510(k) clearance is excludable as irrelevant or unduly prejudicial to plaintiffs.  This was never the law prior to Pelvic Mesh, and was created not so much from the law, but rather to force settlement.  Not anymore, since Keen held that the jury, “after taking into consideration the history of the FDA’s 510(k) clearance of the [product], could determine that [defendant] took reasonable and appropriate steps in its effort to bring [the product] to market.”  That’s the right result, since in no other area of tort law is regulatory compliance viewed as an all or nothing (preemption or inadmissibility) proposition.  We gave kudos to Keen here and here.
  9. Cavanaugh v. Stryker Corp., ___ So.3d ___, 2020 WL 5937405 (Fla. App. Oct. 7, 2020). Defense counsel managed an excellent result despite some pretty bad facts in Cavanaugh.  The key to the case was misuse of the device by medical personnel who failed to read the defendant’s warnings.  The key to the appeal, and why it made our list, is some pretty basic product liability law.  As in many states, Florida allows proof of design defect on the basis of either “consumer expectation” or “risk/utility” theories.  Defendants, particularly those that manufacture complicated products like medical devices, much prefer the latter to the former, and in Cavanaugh the District Court of Appeal agreed.  The consumer expectation test does not logically apply to a prescription-only product about which a “ordinary consumer” would not have any meaningful expectations.  These rulings should have lasting impact in Florida.  Further, since the appeal came in the context of objections to jury instructions, plaintiff never requested an instruction that the “learned intermediary” was the “consumer” for purposes of measuring consumer expectations.  We commended Cavanaugh here.
  10. (Tie) Stiens v. Bausch & Lomb Inc., ___ S.W.3d ___, 2020 WL 7266398 (Ky. App. Dec. 11, 2020). This late addition to our list reaffirms both the purely federal nature of allegations concerning truthful off-label promotion and the vitality of the state of the art defense in a case where the off-label use at issue was also the medical standard of care.  Yes, the defendant engaged in off-label promotion, but plaintiff had no evidence:  (1) of any untrue statement, or (2) that the risk was known at the time of the surgery (plaintiff’s expert admitted only extant medical literature was “not germane”).  Rather, plaintiff’s surgeon was the first to report the risk.  Unknown risks are not foreseeable, as required by the state’s product liability act.  Nor did defendant improperly fail to test off-label uses for unknown risks.  Drug manufacturers are not legally required to test their products for unknown risks of off-label uses.  Nor could plaintiff use that unknown risk gin up a claim over truthful off-label promotion.  Assuming that to be an FDCA violation, off-label promotion must be false also to support state-law liability, and the defendant neither knew nor should have known of that risk at the time of the promotion.  Following Caronia (2012+7) and Caplinger (2013+9), Stiens held that mere off-label promotion is not a tort claim under state law.  We saluted Steins here.
  11. (Tie) Ridings v. Maurice, 444 F. Supp.3d 973 (W.D. Mo. 2020). 2020 saw quite a few favorable implied preemption cases involving “newly acquired information,” and this was the best of them.  Extensive analysis, including a road map of how, procedurally, a court is to approach a preemption motion after Albrecht (2019+1, 2019-6).  In Ridings that included a bench trial at which plaintiff had to specify exactly what was supposedly wrong with the defendant’s warning.  “Newly acquired information” is separate from, and precedes “clear evidence” in evaluating prescription drug preemption issues.  Further, it is the plaintiff’s burden to put the defendant in the “clear evidence” box, so plaintiff must come forward with evidence showing that all the prerequisites of the FDA’s CBE regulation are met, not just newly acquired information but that such information involved a “serious” risk.  Good holdings that inconclusive “needs more study” medical articles, foreign warnings, and discontinued patents don’t trigger the CBE regulation.  For good measure, Ridings also found clear evidence that the FDA would not allow the warning plaintiff sought based on FDA inaction in the face of all the information being disclosed and considerable back and forth on the precise topic.  We rapturously reviewed Ridings  (here).

That’s a wrap for our top ten picks as the best drug/medical device decisions of 2020.

But wait; there’s more.  Like the 2020 hurricane season, there were a lot more than ten pro-defense decisions worth naming, so here’s our list of runners up, the next ten most favorable 2020 decisions.

Honorable Mentions:  (11) Zitney v. Wyeth LLC, ___ A.3d ___, 2020 WL 7040432 (Pa. Super. Dec. 1, 2020).  Precedential holding, in hotly contested test case, that no separate duty exists, once a prescription medical product manufacturer has provided adequate warnings with its product, to repeat those warnings in a Dear Doctor letter addressed specifically to a plaintiff’s prescribing physician (here, and here).  (12) This space intentionally left blank.  (13) Hubbard v. Bayer HealthCare Pharmaceuticals, Inc., ___ F.3d ___, 2020 WL 7550285 (11th Cir. Dec. 22, 2020).  Best learned intermediary case of 2020.  Prescriber already knew of risk and thought it too small to change anything.  Georgia does not have a heeding presumption.  Brand new, so no post yet.  (14) Almond v. Janssen Pharmaceuticals, Inc., ___ F.R.D. ___, 2020 WL 6545892 (E.D. Pa. Nov. 6, 2020).  A nationwide class action involving medical monitoring?  No, way.  The multiple conflicting state approaches to medical monitoring alone preclude any class, and justify striking the class allegations at the outset (here).  (15) Irizarry v. Abbott Laboratories, ___ F. Appx. ___, 2020 WL 6441117 (3d Cir. Nov. 3, 2020).  One of the most hotly disputed issues in PMA preemption cases is whether plaintiffs have to plead the nature of the claimed FDCA violation and causation to state a “parallel” violation claim.  In Irizarry the Third Circuit came down on the side of enforcing TwIqbal.  Would have ranked much higher if precedential (here).  (16) Polt v. Sandoz, Inc., 452 F. Supp.3d 557 (E.D. Pa. 2020).  Invokes Erie restraint to reject a novel medication guide exception to Pennsylvania’s learned intermediary rule.  Negligence per se cannot be used to create new duties not recognized by the common law (here).  (17) Henry v. Angelini Pharma, Inc., 2020 WL 1532174 (E.D. Cal. March 31, 2020).  A couple of years ago we had the bright idea that personal jurisdiction was a likely defense to innovator liability, since the defendant never sold the product that the plaintiff took, it did not commit any act in the relevant jurisdiction.  There were no cases then, but in Henry a California court agreed with our theory and dismissed the case (here).  (18) Gayle v. Pfizer, Inc., 452 F. Supp.3d 78 (S.D.N.Y. 2020).  Another excellent implied preemption “newly acquired information” decision.  A disorganized mass of some 6000 adverse event reports is not such evidence, hence preemption. (here).  (19) Conley v. St. Jude Medical, LLC, ___ F. Supp.3d ___, 2020 WL 5087889 (M.D. Pa. Aug. 28, 2020).  Dynamite PMA preemption case.  Good on parallel claims, specificity of pleading, failure to report adverse events, and applying Albrecht to medical devices (here).  (20) In re DePuy Orthopædics, Inc. ASR Hip Implant Products Liability Litigation, 953 F.3d 890 (6th Cir. 2020).  A useful appellate reminder that an overseas litigation tourist can’t also sue an overseas defendant, since overseas parties on both sides of the “v.” destroys diversity jurisdiction.  Not more highly ranked because this just doesn’t happen all that often (here).

Our 2020 collection of cases concludes with more near misses than usual – there were quite a few good post-Albrecht preemption cases this year; you can check them all out on our preemption scorecard – so here are the next tier of victories:  Webb v. Mentor Worldwide LLC, 453 F. Supp.3d 550 (N.D.N.Y. 2020) (here), D’Addario v. Johnson & Johnson, 2020 WL 3546750 (D.N.J. June 30, 2020) (here), and Smith v. Teva Pharmaceuticals United States, 437 F. Supp. 3d 1159 (S.D. Fla. Feb. 4, 2020) (here).  Finally, Markham v. Ethicon, Inc., 434 F. Supp.3d 261 (E.D. Pa. 2020), if followed elsewhere, could be a big removal deal (here).

Reviewing our prior lists of best and worst decisions, from this year’s lists, we just discussed the Mirena (2019+9) affirmance above, and mentioned last week, that our #7 best case of 2019 got reversed.  In addition, appeals are pending in Nowell (2019+13) and (we think) in Roberto (2019+20).  The others appear final, with appeals failing to change the result in Sherman (2019+6), Davis (2019+18), A.Y. (2019-3), and Painters & Allied Trades (2019-4).

We skimmed over a couple of earlier years of our top/bottom ten lists, but found nothing more than the Dolin (2018+5) re-affirmance mentioned above and the Russell (2018+17) and Hammons (2018-3) results mentioned last week.

Looking forward, we’ve already mentioned that the Supreme Court will decide, before the current term ends in June 2021, a couple more (non-drug/device) product liability personal jurisdiction cases – ones that do not litigation tourists.  Maybe the decision will provide more clarity about the “arising from”/”relating to” specific personal jurisdiction standard.  We’d like to see further review in the United States Supreme Court of both the Hammons (2020-1) and Russell (2020-2) decisions, since in both instances state highest court refused to follow Supreme Court precedent and relied instead on lone dissents, but Supreme Court review is notoriously hard to get.  Further, an appeal is also pending in the Ridings case discussed above as our number ten case.  Another possibility is that we’ll get lucky, and the Mirena plaintiffs will appeal that case to the Supreme Court.  We think Mirena would be an excellent vehicle for pro-defense resolution of some Daubert-related circuit splits.

Finally, on the administrative front, the biggest action we see on the horizon is whether the incoming administration will let stand the recent expansion of PREP Act liability to actors without government affiliations, decisions not to use covered countermeasures and creating an exclusive federal cause of action.  The FDA’s restarted attempt to address its obsolete, 1950s-vintage “intended use” regulations will also plod forward.

Legislatively, we don’t expect much that affects our product liability bailiwick.  However, that’s not necessarily a bad thing.  With Congress closely balanced, we don’t see much likelihood that legislation precluding snap removals will succeed.