We tend to favor a “glass half full” outlook.  We are preternaturally sunshiny during our daily “how was your day” calls from the 86-year-old Drug and Device Law Dowager Countess.  (We have not mentioned, for example, that our aging dog has begun sleeping most of the day and barking most of the night, resulting in an exponential escalation of the coffee intake required for even marginal function.)  On a larger scale, while we have not been immune to sadness and frustration over choices lost during the past thirteen months, we have tried to concentrate on the continued health of everyone we love and on our ability to work from home – and to work at all – during a time of hardship for so many.

But we don’t discharge our obligation to the readers of this blog by Pollyanna-esque avoidance of bad decisions in favor of reporting only on the good.  In that vein, today’s decision has full complements of both.  Nolen v. C.R. Bard, Inc., et al., 2021 WL 1264539 (M.D. Tenn. Apr. 6, 2021) is an IVC (inferior vena cava) filter case.  The plaintiff suffered from deep vein thrombosis (“DVT”), a condition that increased the likelihood of a pulmonary embolism (a blood clot that travels to the lungs).  His physician implanted the defendant’s IVC filter to mitigate this risk.  Sometime after the implant surgery, the doctor died – much more on the implications of that in a moment.

Years passed, and the plaintiff had no symptoms suggesting that anything was wrong with the IVC filter.  Eventually, his cardiologist recommended testing to evaluate whether the filter should be removed.  He ordered a CT scan, which revealed that the filter had tilted and that some of its legs were protruding through the IVC.  The cardiologist attempted to remove the filter but did not succeed, and the filter remains implanted today.   The plaintiff’s suit alleged that he was at greater risk of future adverse events and that he will be required to undergo “recurrent, expensive imaging to detect any dangerous worsening of the situation.”  Nolen, 2021 WL 1264539 at *2.  He asserted the usual product liability claims, and the defendant moved for summary judgment.   For purposes of today’s post, we will concentrate on the court’s decisions on the punitive damages claim (good) and the failure-to-warn claim (very, very bad).

Failure to Warn

In Tennessee, like everywhere else, a plaintiff can prevail on a failure-to-warn claim only by proving both that the defendant did not provide adequate warnings and that the inadequate warning proximately caused his alleged injuries.  In a case involving a prescription medical device, in which the duty to warn is framed by the learned intermediary doctrine, Tennessee law provides that a plaintiff proves proximate causation by producing “evidence that the omitted warning would have altered the doctor’s actions and that the change in the doctor’s actions would have averted the [plaintiff’s] injury.”  Id. at *5 (internal punctuation and citation omitted).   Causation is an element of the claim, and the plaintiff has the burden of proving it – it is not the defendant’s burden to prove the absence of a causal link.  The inquiry on summary judgment, then, is whether a plaintiff has adduced any evidence that a different or stronger warning would have caused the doctor to change his behavior in a manner that would have prevented the plaintiff’s injury. In a “normal” case, that evidence comes from the doctor’s testimony.  In the absence of any such testimony, if the judge is applying the standard correctly, the defendant is entitled to summary judgment on the failure-to-warn claim.

But this was not a “normal” case.   As we mentioned, the plaintiff’s implanting physician was deceased, and he died without being deposed.  Tough luck for the plaintiff, right?  As we have discussed in previous posts, here and here, for example, there is a whole body of case law (including a case from Tennessee) holding that, in the absence of prescriber testimony, a plaintiff can’t prevail on a failure-to-warn claim because he can’t satisfy the causation element of the claim.  So, check and mate, right?

Nope.  The plaintiff’s expert opined that “no reasonable person” would have used the filter if the defendant had provided adequate warnings.   Now, proving causation is not subject to a “reasonable person” standard.  The correct standard, as the court itself articulated, is entirely subjective:  would a different warning have affected the plaintiff’s own doctor’s behavior.  But the court didn’t care.  It held,

Obviously, it would have been preferable to have [the implanter]’s own testimony regarding why he chose the filter that he did and what he would have done in response to more information.  However, [the] testimony about what a reasonable physician would have done is potentially persuasive evidence that [the doctor] would have acted differently.  There is, moreover, nothing in the record to suggest that [the doctor] . . . would depart from the ordinary concerns depicted by [the expert].”  Id. at *5.   The court concluded, “[The plaintiff’s] strategy,” in the face of the implanter’s death, “of recreating the options [the implanter] faced and setting forth evidence of how a reasonable physician would have reacted is an acceptable course of action and sufficient to defeat summary judgment in this instance.

Id. at *6.  Do we even have to tell you how much we hate this?  By allowing an expert’s “reasonable person” speculation to create a fact issue in the complete absence of record evidence, the court applied the incorrect standard and shifted the burden of proving the causation element of the failure-to-warn claim, all in one fell swoop.  We can hope only that judges facing similar cases are as incredulous as we are and that this absurd reasoning spreads no further.

Punitive Damages

But here’s the good part.   The defendant moved for summary judgment on the plaintiff’s punitive damages claim on basis of a “safe harbor” in Tennessee’s punitive damages statute, which bars awards of punitive damages for injuries caused by drugs and devices that were “manufactured and labeled in relevant and material respects in accordance with the terms of an approval or license” issued by FDA.  Id  at *7, citing Tenn. Code Ann. § 29-39-104(d)(1).  The filter was a Class II medical device, cleared for marketing through FDA’s § 510(k) process.  As most of you know, the 510(k) process does not involve the full premarket approval (“PMA”) process to which Class III devices are subject; instead, it clears devices for marketing when they are “substantially similar” to predicate devices.  The plaintiff argued that § 510(k) clearance of the filter did not constitute an “approval or license” issued by FDA.   This issue – the admissibility and weight of evidence that a Class II device was “cleared” – has huge implications for all of us who defend medical device manufacturers.  We most recently blogged about it here, in a post about a case that made our “Best of 2020” list.

And just as in that case, the Nolen court got it right, holding,

[I]t seems clear that authorization to market a product pursuant to the § 510(k) process is, in fact, a form of approval.  Indeed, as [the defendant] points out, the U.S. Supreme Court has referred to § 510(k) approval in its opinions.”  Id. at *8 (citations to Riegel and Buckman omitted).  The court underscored this conclusion with a careful discussion of statutory construction, emphasizing, “The only way that the General Assembly could have intended to use “approval” narrowly here would have been if that legislative body had specifically had the distinction between PMA and lesser forms of FDA review in mind.  But if the General Assembly was that far “into the weeds” of federal pharmaceutical law, why did it not more clearly indicate that it was drawing such a distinction?  Why did it not just say “premarket approval” or cite the relevant subsections as opposed to merely saying “approval” and citing generally to the [FDCA] as a whole?

Id. at *9.  The court concluded that “a defense to punitive damages . . . can be satisfied by establishing that the device at issue was “approved” by the FDA through the §510(k) process.”  Id. 

The defendant had one more hurdle to surmount, because the same statutory section that establishes a safe harbor for “approved devices” includes an exception stating that the defense does not apply to a device when the manufacturer withheld material information from the FDA at any time before the plaintiff’s injury.  The defendant argued that this provision was preempted by federal law because, under Buckman, policing of fraud on the FDA is reserved for the FDA.  The court held that the defendant was partly right:  under Sixth Circuit law, federal law preempts a court’s finding of fraud on the FDA; however, “a state statute [can] include a distinction based on wrongdoing before the FDA as long as it relie[s] solely on the FDA’s own finding of wrongdoing to determine whether the provision applie[s].”  Id., at *10 (emphasis in original).

In other words, “a provision like Tenn. Code Ann. § 29-39-104(d)(2) could permit punitive damages in a case arising out of an FDA-approved device based on the fact that the FDA itself made a determination that manufacturer withheld information, but the same provision could not allow punitive damages based solely on the court’s independent finding that such withholding had occurred.”  Id. (internal punctuation and citation omitted, emphasis in original).   Because the FDA had made no such determination before Nolen’s injury, the exception did not apply, the safe harbor did, and the court granted summary judgment for the defendant on the punitive damages claim.  And this holding is every bit as good as the failure-to-warn holding is bad.

As for us, we look forward with excitement to our second shot next week and, a couple weeks later, to hugging the Dowager Countess on Mothers’ Day.  The sun’ll come out tomorrow.  In the meantime, stay safe out there.

Delaware is having something of a moment in the sun.  Although the state’s license plates have long announced it as “The First State,” that refers to being the first to ratify the U.S. Constitution.  It is the second smallest in size and sixth smallest in population of the current fifty states.  The casual peruser of a map of the U.S. might miss it.  Even its state fish, the weakfish, is, well, weak.  (Disclaimer:  A few of us grew up and/or live in Delaware’s neighboring states.)  Now, though, Delaware is home to the President of the United States.  It also continues to be the preferred state of incorporation for publically traded companies, with about 300 of the Fortune 500 incorporated there.  That has led to a bit of a surge in product liability litigation in Delaware state courts.  Why now?  Well, the tightening of the rules on personal jurisdiction over the last several years means that many plaintiff lawyers have to choose between federal court, and possibly a slow MDL, or state court in Delaware, where the defendant is incorporated and/or based.  Before Bauman, it did not seem that Delaware was a popular choice for the plaintiff lawyers, certainly if proceeding in the state court where the plaintiff lived or where the plaintiff lawyer liked to bring many litigation tourism cases was an option.  Now, Delaware is getting more cases and Delaware courts are increasingly deciding cases under the law of another state, somewhat like a federal court sitting in diversity and hopefully practicing some Erie restraint.

In Pope v. Astrazeneca AB, No. N20C-06-116 FAR, 2021 Del. Super. LEXIS 281 (Del. Super. Ct. Apr. 5, 2021), we have just such a case.  The case involves a claim of Fournier’s gangrene, a nasty perineal infection, in connection with the use of a particular diabetes medication.  We have previously posted on decisions from the MDL and from the same Delaware state court about this same medication.  Like this case, the prior Delaware case even involved Texas law, which almost invariably means consideration of the second most famous “007,” § 82.007 of the Texas Code, which provides a presumption of adequacy for FDA-approved drug labels (like a license to kill warnings claims).  We have posted quite a bit on this provision, like here and here, and we can say it has been a key provision in some pretty effective tort reform.  There was a time when the suggestion that a Texan would chose to sue a Delaware drug company in its backyard would have drawn a hearty guffaw.  Times have changed.

In Pope, the plaintiffs brought claims relating to warnings, design, negligence, warranty, loss of consortium, and punitive damages, the defendants moved to dismiss, and the court addressed everything in a no-nonsense and sensible fashion.  The wrinkle in this case was that the plaintiff—the one who took the medication—claimed to have been prescribed it for weight loss as well as his diabetes.  The latter was on-label, but the former was off.  This might matter under § 82.007 because it provides an exception to the presumption of adequacy that would otherwise have applied if “the manufacturer promoted the product for a use not approved by the FDA.”  Id. at *4.  We note that the court did not get caught up in questions about how to deal with FDA approval—a fact often omitted from complaints and necessary for the presumption, which is arguably a defense—on a motion to dismiss.  Instead, it identified that the exception required pleading not just of promotion and reliance on it, but of a causal relationship between the off-label promotion and the plaintiff’s injuries.  Id. at *5.  Here, plaintiff admitted to being prescribed the medication on-label, but with the hope that it would also help him lose weight.  Common sense indicates that the alleged off-label promotion could not be the cause of plaintiff’s injuries in that situation.  Texas federal courts had addressed this in at least two cases, one of which we discussed here.  With its own Delaware version of Erie restraint (Rehoboth restraint or Appoquinimink restraint?), the court held that the exception did not apply:

Pope was using Farxiga for both an indicated and non-indicated use.  And Plaintiffs cannot establish, nor have they even pled, that Pope’s injuries were caused by his off-label use of Farxiga for weight loss, as opposed to his FDA approved use of Farxiga to treat his Type 2 diabetes.

Id. at *8.  The presumption of adequacy held and the warnings claim was dismissed.

Next up was the design defect claim.  As the court had held previously, an adequate warning defeats a design defect claim under Texas law.  Id. at *9.  As an added measure, the court looked at whether plaintiffs could carry their burden of showing a safer alternative design.  Noting that a manufacturer cannot change the design of an approved drug, “[t]o establish that a safer alternative design exists would invoke a state duty to change the design of the drug,” which meant the design claim was preempted.  Plaintiffs, however, attempted to parse the design claim into pre- and post-approval halves, the former of which they claimed was not preempted.  Relying on one of our favorites, Yates, the court rejected plaintiffs’ argument:

In order to find for Plaintiffs on such a claim, the Court would be required to engage in multiple levels of supposition. For instance, the Court would have to speculate about whether a pre-approval design change would have fixed the problem; whether the FDA would have approved such changes; whether Pope’s physician would have recommended and prescribed the differently designed drug, if it was FDA approved; and whether Pope would have taken the drug after such changes and processes were executed. The Court would be required to make unfounded inferences as to each of these questions to find for Plaintiffs, and thus is unable to countenance Plaintiffs’ pre-approval design defect argument.

Id. at *11.  Therefore, the design claim was also dismissed.

As to plaintiffs’ negligence claims, the court did not fall for the argument that merely asserted negligence is enough.  With the exception of negligent failure to test, the negligence claim was really duplicative of the warnings and design claims already dismissed.  Id. at *12.  The same went for express and implied warranty.  Id. at **13-14.  As to failure to test, this is where we have a gripe.  The statement that “[c]ourts in Texas have recognized an independent cause of action based on negligent failure to test” was based on the Murthy decision panned here for, among other problems, not making a proper Erie prediction.  Another case following Murthy (and discussed in not-so-glowing terms here) was cited, but is clearly part of a minority.  In any event, the assertions supporting a purported failure to tests were conclusory boilerplate and held to be inadequate.  Id. at *13.  Because this part of the dismissal was without prejudice, plaintiffs will get another shot at trying to plead negligent failure to test if they choose.  Should they try again, we hope the court will take a deeper dive into whether Texas really has adopted such a claim, which is logically just another version of failure to warn.

We got our second anti-Covid jab this week, so we’ll celebrate by discussing a COVID/PREP Act case. The PREP Act (PREP stands for “Public Readiness and Emergency Preparedness”) declaration of immunity is limited to “covered countermeasures” obtained either through agreement with the federal government or otherwise in response to the COVID-19 pandemic. The purpose of PREP Act immunity was to incentivize the supply of goods needed to combat scourges such as the coronavirus.

In Avicolli v. BJ’s Wholesale Club, Inc., 2021 U.S. Dist. LEXIS 67096 (E.D. Pa. April 7, 2021), the covered countermeasure was hand sanitizer. Remember when hand sanitizer was hard to get? It was like liquid gold during the early days of the pandemic. Someone who had ample stores of hand sanitizer, toilet paper, and bread yeast was a Covid tycoon. We do not miss those days. Anyway, the plaintiff in Avicolli at some point saw fit to ingest the hand sanitizer and was injured. (Where oh where could the plaintiff have gotten that dumb idea?) The plaintiff filed a lawsuit against the manufacturer, distributor, and retailer for selling a defective product, failing to warn that the product contained methanol (wood alcohol), and negligence. The defendant retailer moved to dismiss the complaint on the ground of PREP act immunity.

The key issue in Avicolli was whether the product fell within the scope of PREP Act immunity. The court held that the retailer defendant, a store from which the plaintiff had obtained the hand sanitizer, had not established immunity as a matter of law, since the plaintiff’s pleading did not exclude the possibility that the plaintiff acquired the hand sanitizer before the pandemic began, and thus not in response to the pandemic.

The ruling went the plaintiff’s way because the plaintiff had not made the gargantuan mistake of pleading that the retailer obtained the hand sanitizer after the pandemic began. This case is yet another example of how, when it comes to pleading a case, sometimes less is more. But the truth will come out eventually, and if it proves to be the case that the hand sanitizer was purchased after Covid started, the retailer should prevail on summary judgment. It is all a matter of timing.

We cannot help but wonder if the Avicolli court’s rather crabbed interpretation of immunity is consistent with the goals of the PREP Act. What if the retailer obtained the hand sanitizer before the pandemic but sold it afterwards? Isn’t the selling what matters here? Isn’t the selling the thing to be encouraged?  We’re not aware of that much caselaw on this point (see our prior posts here and here). It seems to us that the Avicolli court was very worried about affording too much immunity, and not worried at all, and certainly not worried enough, about real world incentives.

The California state court consolidated Incretin litigation has closely followed its MDL counterpart.  We reported on the trial court’s preemption-based dismissal here.  But, after reporting on the Ninth Circuit’s undoing of the federal dismissal, we opted not to explore the California Court of Appeals decision essentially doing the same thing at the state level.  See Rotondo v. Amylin Pharmaceuticals, Inc., 2018 WL 5800780 (Cal. Ct. App. Nov. 16, 2018).  Suffice it to say that the state court Incretin plaintiffs were provided the same opportunity to poke around for “newly acquired evidence” that would support an unpreempted warning based claim.

But, as with the federal court decisions, we are now in a post-Albrecht world.  So, if at first you don’t succeed . . . .

Here we are two and a half years later, and the Incretin defendants again moved to dismiss the action, and once again they succeeded.  And, once again the state court followed the federal court’s lead using the same one-two punch – Daubert and preemption.  In re Byetta Cases, Case No. JCCP 4574, slip op. (CA. Super. Apr. 6, 2021).  It’s also a lengthy decision and so today you are getting the preemption portion with Daubert to follow.

The court started by addressing the conflict between the state appellate ruling in Rotondo and the Supreme Court ruling in Albrecht – Albrecht controls.  And as pointed out yesterday, that’s generally a good thing for defendants.

[T]he issue of FDA preemption presents a question of law for decision by the Court and that in the course of doing so, the Court is authorized to decide all predicate factual questions necessary to the resolution of the ultimate legal question of whether or not “impossibility” preemption applies in a given case.

Slip op. at 24.  Not wanting to completely ignore Rotondo, the Incretin court offered to conduct a bench trial before deciding the preemption issue, but the parties stipulated that all the necessary evidence to address the issue was before the court and no trial was needed.  Slip op. at 24.  So, having satisfied any obligation that still existed under Rotondo, the court moved on to those “predicate factual questions.”

First up, what “new safety evidence” should the FDA have considered before deciding the drug’s labeling need not include the risk of pancreatic cancer?  While preemption is considered an affirmative defense, it was plaintiffs arguing more information needed to be sent to the FDA.  So, plaintiffs bore the burden of identifying the allegedly missing information and also demonstrating how and why such information “would have made a difference in the FDA’s deliberative process.”  Slip op. at 25.

The court enumerates seventeen categories or types of data plaintiffs contended were “new safety evidence.”  Slip op. at 27-36.  As noted yesterday, plaintiffs “nitpicked” practically everything that was ever submitted to the FDA.  The state court’s conclusion was that the “information” (i) was outdated and unreliable, slip op. at 27; (ii) didn’t support plaintiffs’ position as to the risk of pancreatic cancer, slip op. at 28, 31; (iii) was in fact reported to the FDA, slip op. at 29, 34-35; (iv) was litigation-driven science that the court found unreliable, slip op. at 30; (v) was publicly available and therefore available to the FDA, slip op. at 33; or (vi) involved a different drug, slip op. at 34.

Plaintiffs’ nitpicking was viewed by the court as “trying to get the Court to decide the science.”  Slip op. at 30.  That’s not the court’s job.  Nor is it the court’s job to second-guess FDA submissions.  But, like the federal court, the state court took a once-bitten approach and stayed away from Buckman this time around too.  Instead, after

[S]eriously consider[ing] each piece of plaintiffs’ proposed “new safety evidence,” the Court makes the predicate finding of one of the “brute facts” needed to resolve this legal question:  plaintiffs have failed to show that defendants failed to provide the FDA in a timely fashion with one or more piece of additional safety evidence which might have a tendency to change the FDA’s behavior.

Slip op. at 37.

The next predicate fact that needed deciding was whether the defendants have met their burden as to “clear evidence” that a unilateral label change would have been rejected by the FDA.  Tied up in that question is the threshold question of what constitutes “agency action carrying the force of law.”  Id.  In other words, what could the court consider as proof of the FDA’s position to support that it would have rejected the adding of a warning.  Relying on the concurrence in Albrecht, the court noted that the FDA’s failure to act may be an “official action” disapproving of a warning.  Slip op. at 38.

The court’s primary focus was on a 2014 article published in the New England Journal of Medicine (“NEJM”) by four FDA officials concluding that upon examination of the data regarding a “pancreatic safety signal,” the FDA believed that the product labeling “adequately reflected” current knowledge.  Slip op. at 38-39.  For several reasons the court concluded that the article was “not necessarily” the formal position of the FDA and did not bind the agency.  Id. at 40.  However, the article was “highly relevant” as evidence of the bases for the FDA’s “consistent” exercise of its labeling authority, “which does constitute official agency action.”  Id.

The FDA’s consistent evaluation and reevaluation of a product, coupled with its obligation to raise new safety information with the manufacturer when such information independently comes to the agency’s attention, constitute “official action” for purposes of preemption.  The FDA has treated with the subject of suitable labels for incretin-based mimetic products repeatedly over the last five-plus years.  Its consistent declination of addition of a pancreatic cancer warning when it is otherwise officially approving such label changes does show “agency action” under its delegated authority sufficient to support application of the Supremacy Clause.

Id. at 40-41.

Between the FDA’s inaction in requiring a warning while making labeling decisions and “[n]o recall, recanting, or questioning of the NEJM article” from any corner, defendants proved by “clear evidence” that the FDA would have rejected the warning put forth by plaintiffs.  Therefore, it was impossible for defendants to comply with both federal and state requirements.  Id. at 41.

Stay tuned for a future post on the Daubert rulings.

When last we tuned into the In re Incretin-Based Therapies (“Incretin”) multi-district litigation, the Ninth Circuit had just undone a preemption-based dismissal – but only on procedural grounds.  As we discussed, here, the Ninth Circuit avoided the merits, but ruled that the MDL court had erred in “rel[ying] on Buckman [Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001)] to circumscribe discovery.”  In re Incretin-Based Therapies Products Liability Litigation, 721 F. Appx. 580, 582 (9th Cir. 2017).  The Ninth Circuit thus allowed the Incretin plaintiffs to rummage through the defendant’s submissions to the FDA in search of “newly acquired information” that might have justified an unpreempted warning based claim, notwithstanding FDA approval of all the labels of the relevant products, including the omission of warning about a purported pancreatic cancer risk.  Id. at 583.  As we discussed, the Ninth Circuit relied on Wyeth v. Levine, 555 U.S. 555 (2009), for the “clear evidence test” that defined what was relevant, holding that:

Uncertainty about whether the FDA considered the “new safety information” and whether it would have altered the FDA’s conclusion establishes that a disputed issue of material fact should have prevented entry of summary judgment on the defendants’ preemption claim.

721 F. Appx. at 584.  The decision also relied on Stengel v. Medtronic Inc., 704 F.3d 1224, 1233 (9th Cir. 2013) (en banc), to limit the scope of preemption under Buckman.  721 F. Appx. at 583.

So Incretin was reversed and remanded for additional discovery on FDA’s repeated determinations that these drugs (Byetta, Januvia, Janumet, and Victoza) – which are mainstays in treatment of diabetes – had not been shown to cause pancreatic cancer.  In the interim, the United States Supreme Court reformulated Levine in Merck Sharp & Dohme Corp. v. Albrecht, 139 S. Ct. 1668 (2019), and the Arizona Supreme Court spanked the Ninth Circuit for improperly expanding that state’s law in Stengel.  See Conklin v. Medtronic, Inc., 431 P.3d 571, 578 (Ariz. 2018) (“We disagree with Stengel”).

If at first, you don’t succeed….

After three years, and who knows how many more millions of dollars and thousands of hours expended, the Incretin defendants tried again to get the action dismissed – and once again succeeded.  The lengthy opinion, In re Incretin-Based Therapies Products Liability Litigation, ___ F. Supp.3d ___, 2021 WL 880316 (S.D. Cal. March 9, 2021), unloaded the same one-two punch on the plaintiffs that we blogged about the last couple of weeks:  preemption and Daubert.  We’re splitting this one up, too, and covering the Incretin preemption rulings today.

As for preemption, Incretin is one more reason why we think that, on the whole, Albrecht is a big win for the defense side.  As you read this post, think about how easy it would have simply to dismiss a lot of these issues as “fact questions” prior to Albrecht.  As regular readers of this Blog know, and as the Incretin court pointed out, “in [Albrecht], the Supreme Court clarified that courts should treat the question of whether clear evidence is met ‘not as a matter of fact for a jury, but as a matter of law for the judge to decide.’”  2021 WL 880316, at *3 (quoting Albrecht, 139 S. Ct. at 1679).

Albrecht held that when conducting a preemption analysis, a judge may have to resolve contested brute facts, such as whether a manufacturer submitted all material information to the FDA because such factual questions are part and parcel of the broader legal question and do not warrant submission to a jury.  Thus, the Court will determine whether the FDA’s regulatory scheme and state law failure-to-warn requirements “irreconcilably conflict” and resolve issues regarding materiality along the way.

2021 WL 880316, at *4 (citation omitted).

The fundamental reason that the Incretin defendants won is why we have preemption in the first place.  The FDA has been all over the purported link between these diabetes drugs and pancreatic cancer since day one.  Time and time again, the Agency as evaluated the relevant science, and every time it has concluded that this asserted “risk” is fictitious.  In finding FDA’s actions preemptive, Incretin weighed in on many of the major post-Albrecht controversies.  Plaintiffs got blasted with both barrels – their claims were preempted both for lack of any “newly acquired information” that would allow a “changes being effected” (“CBE”) label change in the first instance, and because there was “clear evidence” that the FDA would have rejected such a label change if it had been permissible.

We start with the FDA’s prerequisites to the availability of CBE label changes.

[W]hile a manufacturer bears responsibility for maintaining the adequacy of its labels for as long as the drug is on the market, it “cannot propose a change that is not based on reasonable evidence.”  There must be sufficient evidence of a causal association between the drug and the information sought to be added.

Incretin, 2021 WL 880316, at *3 (Albrecht citations omitted).

Opposing preemption, the Incretin Plaintiffs opened with their by-now-familiar assertion that “Albrecht limited preemption to cases where the manufacturer has proposed a label change.”  Id. at *4.  Not so in the least.  While a manufacturer-generated label supplement certainly makes preemption easier, the fact remains that “this general principle cannot be applied in every case” because “manufacturers cannot propose a change that is not based on reasonable evidence.”  Id.  Moreover, Albrecht specifically declined to decide any “question of disapproval method.”  Id. (Albrecht citations, in both instances omitted).  Because the Incretin defendants, unlike the defendant in Albrecht, “maintain that the CBE regulations did not permit them to change their labeling,” the issues Albrecht resolved were “distinguishable.”  Id.  Thus, “the Court decline[d] to find that Albrecht forecloses preemption merely because there was no CBE or other label change request in this case.”  Id. (citations omitted).  Down goes one frequent plaintiff-side argument.

Thus, Incretin “agree[d] with Defendants that whether federal and state laws irreconcilably conflict entails the threshold inquiry of whether there is ‘newly acquired information’ to support a CBE submission.”  2021 WL 880316, at *5 (citing Albrecht).  Only “if the answer is yes” to that question, does it become necessary to consider the “clear evidence” issues addressed by Albrecht.  Id.  “As such, the preemption framework provides two potential avenues by which Defendants may establish that federal and state laws irreconcilably conflict.”  Id.

With respect to each of the four drugs at issue, Incretin concluded, first, that the requisite “newly acquired information” did not exist.  Much of this discussion is intensely product specific, so we will simply hit the legal highlights.  Plaintiffs’ basic problem throughout was that the FDA has been “investigating reports of possible increased risk” of various pancreatic conditions, including pancreatic cancer, since 2009.  2021 WL 880316, at *5-6.  Here’s what the FDA concluded:

  • 2009: “[L]ittle inference for risk is appreciated from review of spontaneous reports of pancreatic cancer” and “a causal association between exposure to one of these agents and pancreatic cancer is indeterminate.”  Id. at *5.
  • 2013: “FDA also stated that it had not concluded that incretin mimetics cause or contribute to the development of pancreatic cancer, and advised health care professionals to continue following the prescribing recommendations in the drug labeling.”  Id.
  • 2013: “FDA official reaffirmed that adverse event data was “less suitable for detecting relatively more common events with long latency periods” such as pancreatic cancer,” and “incretin-based drugs were not associated with “[o]vert pancreatic toxicity or pancreatic neoplasms.”  Id. at *6.
  • 2014: Four FDA officials published an article in the New England Journal of Medicine (“NEJM”) on the results of “a comprehensive evaluation” of “more than 250 toxicology studies, as well as a review of clinical safety databases and the results of cardiovascular outcome trials.”   They “concluded that ‘assertions concerning a causal association between incretin-based drugs and pancreatitis or pancreatic cancer . . . are inconsistent with the current data.’”  Id. (citing NEJM article).
  • 2014: FDA denied a citizen petition raising the same pancreatic cancer risk claims.  “[I]ts review [of] 49 cases recovered from FAERS [a voluntary adverse event reporting database] . . ., found no new evidence regarding the risk of pancreatic carcinoma . . . that would support any changes to the current approved labeling.”  Id.
  • 2014: FDA concluded that as to “[r]isk for pancreatic cancer,” “animal, observational, and clinical trial data reviewed by FDA to date have not supported a causal association.”  Id. at *7.  A new drug was therefore approved “without requiring a pancreatic cancer warning.”  Id. at *6.
  • 2017: FDA determined that “data generated” in a new study “do not appear to substantively alter the original FDA conclusions regarding the lack of sufficient information to conclusively determine whether long term exposure to [the drugs at issue] increase the risk of pancreatic cancer.”  Id.
  • 2019 & 2020: In connection with still more drug approvals, “for the last several years, the FDA has approved new incretin-based medications without requiring a reference to pancreatic cancer in their labels.”  Id. at *7.
  • Through 2020: “Despite their close monitoring and comprehensive evaluation of the pancreatic safety of incretin-based therapies, the FDA has never required Defendants to include a pancreatic cancer warning in their drug labeling.”  Id.

As for the plaintiffs’ purported “newly acquired information,” applying the FDA’s definition (see 21 C.F.R. §314.3), Incretin concluded that some of it was not “new,” other items didn’t qualify as “information,” and none of it qualified as both.  Plaintiffs, for their part, nitpicked practically everything that was ever submitted to the FDA.  Plaintiffs relied on foreign regulatory activity.  Id. at *8-9.  However, this material was “preliminary,” “did not rely on information different” than what the FDA had, and ultimately reached the same conclusion as the FDA – being “that there is not enough evidence at this time to confirm a link between incretin-based therapies and pancreatic cancer.”  Id. at *9.  Thus, this material failed on the facts.  It also failed on the law:  “Foreign drug labeling is the product of different and distinct regulatory standards and decisions” so that “warnings approved for a foreign label are not in and of themselves newly acquired evidence.”  Id. (quoting Ridings v. Maurice, 444 F. Supp.3d 973, 994 (W.D. Mo. 2020)).

Plaintiffs also criticized studies submitted, or not submitted, to the FDA.  The minutiae in their nitpicking was dramatic, and frankly was exactly the kind of second-guessing of FDA submissions that the Supreme Court intended to stop in Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001).  But because of the Ninth Circuit’s prior opinion, the once-bitten Incretin court stayed away from Buckman entirely, not citing that case at all, except to describe the prior Ninth Circuit opinion.  2021 WL 880316, at *4.

How’s this for nitpicking?  Plaintiffs raised:  (1) supposed exclusion of studies where the reason for the exclusion was shared with the FDA, id. at *9; (2) excluding adverse events that were, in fact, reported, id.; (3) supposed omission of studies about entirely different drugs, id.; (4) omission of “nonclinical studies” in mice and baboons, id., at *9-10; (5) claimed omissions with “no explanation as to the data’s source, context, or meaning,” id. at *10; (6) their own expert’s “re-analysis” of certain “images” that was “generated in preparation for litigation” and was “unsupported by any published research,” id.; (7) supposedly “misleading” summaries of studies submitted to the FDA that in fact had the “same” number of relevant events while analyzing “different data sets,” id. at *11; (8) purported “cherry-picking” in a published study when, in fact “all” of the information was “available for FDA scrutiny,” id.; (9) alleged deficiencies in “study protocols [that] were submitted to and approved by the FDA,” id.; (10) omission of a “tumor” event that was, in fact, “benign,” id.; (11) more summaries, despite their “context . . . [being] plainly stated in the document,” id.; (12) animal studies without any expert explanation of why they were significant, and that made “similar observations” to other studies that had been “found insufficient for a causal link[] by the FDA,” id. at *12; (13) a “post-hoc secondary . . . analysis” in which the “data would not be sufficient to reach any meaningful conclusions,” id.; (14) adverse events in animal studies that studied different organs and thus “assessed data different from [the drug’s] pancreatic safety,” id.; (15) supposed “burying” of data from a study that was not only “disclosed” to the FDA by “publicly posted” on its website, id. at *13; and (16) criticism of a cancer “background rate” from a study involving a different drug, id.

In sum, the Incretin plaintiffs plumbed the depths of the defendants’ FDA submissions (in direct violation of Buckman’s warning that “[a]pplicants would then have an incentive to submit a deluge of information that the [FDA] neither wants nor needs, resulting in additional burdens on the FDA’s evaluation of an application,” 531 U.S. at 351) and raised an olio of irrelevant, immaterial, and often simply false criticisms of those submissions.  Pre-Albrecht, the sheer volume of minutiae might have led a court to throw up its hands and exclaim “fact question” – but no longer.  The debunking of such pseudoscientific flotsam and jetsam is precisely the sort of factual analysis that Albrecht requires courts to undertake themselves with respect to preemption motions.  Incretin concluded, as to newly acquired information:

[T]he FDA’s CBE regulations are designed to ensure that only scientifically justified information is provided in the labeling for an approved product.  The regulations reflect the FDA’s cautious approach to drug labeling, recognizing that exaggeration of risk, or inclusion of speculative or hypothetical risks, could discourage appropriate use of a beneficial drug or decrease the usefulness and accessibility of important information by diluting or obscuring it. . . .  [T]he FDA, through its own evaluation and armed with information from Defendants and other sources, considered the specific issue raised by Plaintiffs in this case:  the pancreatic safety of incretin mimetics.  At no point in its years-long monitoring of these drugs did the FDA require Defendants or any other manufacturer of incretin-based therapies to add a pancreatic cancer warning to its labels.

Id. (Albrecht citations and quotation marks omitted).  Instead, the FDA did quite the opposite.  Its “finding of an indeterminate causal link between pancreatic cancer and incretin mimetics is not reasonable evidence of a causal association.”  Id. at *14.  “[I]nconclusive” data cannot be considered “newly acquired information” that would allow a unilateral label change.  Id.  “[N]one of the purported new safety information reflects well-grounded scientific evidence of causal association that would have made the risk of pancreatic cancer apparent to Defendants.”  Id.  Therefore, all plaintiffs’ claims were preempted.

But Incretin didn’t stop there.  It opted for a belt-and-suspenders approach, also concluding that the record established “clear evidence” that, in any event, the FDA wouldn’t have permitted the label change plaintiffs demanded.  2021 WL 880316, at *14-17.  Once again, Incretin is significant in rejecting the other side’s arguments that sought to use Albrecht to roll back preemption.

And here’s an important preemption practice point:  The absence of any ”newly acquired information” (step one) also had effect of establishing the obverse proposition in step two (clear evidence) – that under Albrecht, “[d]efendants fully informed the FDA of the justifications for a pancreatic cancer warning.”  Id. at *14.  If nothing “new” of any significance was withheld, then necessarily the FDA was “fully” informed.

Plaintiffs also argued that the Incretin defendants couldn’t point to “any agency action carrying force of law.”  2021 WL 880316, at *14.  They lost again.  Everything that the FDA did was “within the scope of the authority Congress has lawfully delegated.”  Id.  In particular, the Agency’s “[a]ssessment and formal rejection of the citizen petition . . . communicated the FDA’s official position that assertions concerning a causal association between pancreatic cancer and incretin-based drugs were inconsistent with the current data.”  Id.  So was the NEJM article:  “Four FDA officials authored the Assessment,” it was “identified as coming from [CDER],” it was published in “the top medical journal in the world,” and it “lack[ed] the disclaimer required when publications of FDA employees do not necessarily reflect the opinions of the agency.”  Id.

Albrecht did not purport to limit the means that FDA could legitimately use to communicate rejection of a litigation-inspired warning change.  Id. at *15.  While, according to FDA regulations, statements by FDA employees “do[] not necessarily represent the formal position of FDA,” id., (quoting 21 C.F.R. § 10.85(k)), “by its own text, th[at] regulation contemplates that other “employee-written statements” can be “representative of the FDA’s formal position.”  Id.  Given the number of FDA authors, the prestige of the publication, and the lack of a disclaimer, the NEJM article was sufficient to qualify as a “formal” agency position.  Id. (“the Court finds that the 2014 [NEJM article] amounts to ‘other agency action carrying the force of law.’” (quoting Albrecht)).

Nor could plaintiffs claim that the FDA’s rejection of a scientific basis for a pancreatic cancer risk wasn’t sufficiently “communicated” to the defendants.  The NEJM article “specifically communicated that the FDA believed that current knowledge is adequately reflected in the product information or labeling.”  Id. at *16 (citation and quotation marks omitted).  Even plaintiffs’ own expert testified that “[i]t would be a little absurd for the FDA to allow the addition of a pancreatic cancer warning after its robust and comprehensive evaluation of that specific risk did not yield evidence of causal association.”  Id. (citation and quotation marks omitted).

Further, standard language that the FDA would “continue to monitor” the issue did not deprive its conclusion of sufficient finality.  The FDA always “monitors” evolving science.  This was another plaintiff-side argument that proved too much, and would essentially eliminate preemption.

[A]s a matter of routine practice, FDA continuously monitors every medication for new or evolving information as long as a drug is on the market.”  The Supreme Court acknowledged the same in [Levine] when it noted that the CBE regulation accounts for the fact that risk information accumulates over time and that the same data may take on a different meaning in light of subsequent developments. . . .  As such, the Court rejects Plaintiffs’ contention that “the FDA’s rejection of the petition did not inform the manufacturers of incretin mimetics that any proposed pancreatic cancer warning would be rejected.

2021 WL 880316, at *17 (citations and quotation marks omitted).

Incretin also invoked 21 U.S.C. §355(o)(4)(a), conferring on the FDA “authority to mandate a label change if it learns of new safety information that should be included in the labeling of a drug.”  Id.  For that reason “[t]he Court also disagree[d]s with Plaintiffs’ claim that FDA inaction can never support a preemption finding.”  Id.  Given the FDA’s obligation to order label changes on its own volition when “new safety information” warrants, “the FDA’s silence on this issue is highly relevant to its preemption analysis.”  The record showed sustained FDA interest in the purported risk at suit – as well as the Agency’s continued approval of labeling that did not contain the sort of cancer warnings the these plaintiffs demanded.

Put another way, the Court cannot simply ignore the FDA’s demonstrated commitment to actively and continuously monitoring the pancreatic safety of incretin mimetics. . . .  Given the specific attention the agency has given to the very matter at issue in this litigation, the Court finds the FDA’s approval of new incretin-based drugs, as well as its continued approval of other label changes to the medications at issue − without requiring a pancreatic cancer reference in the label − only further supports the Court’s preemption findings.

Id. at *17 (citing Ridings).

Incretin thus stands as a comprehensive rejection of the other side’s attempts to use Albrecht as a means of rolling back implied preemption in the prescription drug arena.  It confirms that two routes to preemption exist – lack of newly acquired information and “clear evidence.”  The opinion enforces FDA standards for such information, rejecting information no different than what the FDA has, “inconclusive” information, and a wide variety of minutiae that plaintiffs’ Buckman-ignoring, second-guessing of the defendants’ FDA submissions dredged up.  On clear evidence, Incretin strongly rejects the argument that preemption only exists where the defendant itself submitted an unsuccessful label change, recognizes that FDA rejection of citizen petitions is a formal FDA action with preemptive effect under Albrecht, and holds that FDA “inaction” in situations where the FDA is actively monitoring the relevant science, is in and of itself preemptive.  On both, Incretin recognizes that the absence of “newly acquired information” supporting a CBE label change also establishes that the FDA was “fully informed” under Albrecht‘s reformulated “clear evidence” test.

We’ll address the Daubert rulings in Incretin – excluding all seven of the P-side experts that the defendants challenged – in a future blogpost.

Sometimes a case has one bright shining point floating in a sea of not-so-good ones, or a stand-out result lost among outcomes that should have gone the other way.  That describes Isaac v. C.R. Bard, Inc., No. A-19-CV-895, 2021 WL 1177882 (W.D. Tex. Mar. 29, 2021), a case remanded from the Bard IVC filter MDL.  As in other filter cases, the plaintiff in Isaac alleged complications following the implantation of a filter following a medical emergency.  The defendants’ motion for summary judgment resulted in the dismissal of some claims, but not most, and the order covers a lot of ground.  For our part, we found three parts of the order particularly interesting—the parts dealing with warnings-based claims, design defect, and the presumption of non-liability under Texas law for federally approved or licensed products, such as medical devices.

First, the highlight.  As we explained just last week, a plaintiff asserting a warnings-based claim bears the burden of proving both that the defendant’s warnings were inadequate and that the inadequate warning proximately caused her alleged injuries.  In the drug and medical device world, that means the plaintiff bears the burden of proving that a different or stronger warning would have altered her doctor’s decision to prescribe the drug or device.  That is the rule in the Fifth Circuit, too:

The Fifth Circuit has explained that causation in the context of the learned-intermediary doctrine entails two distinct factual predicates:  (1) that the doctor would have read or encountered the adequate warning, and (2) that the adequate warning would have altered his treatment decision for, or risk-related disclosures to, the patient.

Id. at *5.  We are not so sure about that very last part—that a plaintiff can meet the second factual predicate with evidence that the physician would have altered his or her risk-related disclosures.  But for now, we will focus on the first part—that the doctor would have read or encountered the adequate warning.  In Isaac, no one took the implanting physician’s deposition, so there was no evidence whether the physician would have seen a different warning, let alone what he or she would have done had the warning been stronger or different.  Id.  Importantly, it was no substitute to have an expert say that a “reasonable implanting physician” would have changed his or her treating decisions.  As the court observed, “the learned-intermediary analysis focuses on the actions of the treating physician, not the opinion of an expert witness.”  Id. (emphasis added).

The remainder of the order?  Well, not so good.  On design defect, the defendants invoked comment k to Restatement (Second) of Torts § 402A.  As our faithful readers know, comment k precludes design defect claims for “unavoidably unsafe products,” the rationale being that many products will always carry risks and thus cannot be considered “defective” so long as they come with adequate warnings.  The court in Isaac rejected comment k because, according to the court, comment k has never been extended to medical implants, let alone devices cleared under section 510(k).  Id at *6.

The court here is drawing two false distinctions to come to the wrong conclusion.  First, there is no justifiable basis for applying comment k to drugs, but not prescription medical devices.  Both exist to treat injury and illness, and both carry treatment risks that cannot be completely eliminated.  That is why federal law allows their use only when prescribed by medical doctors and other qualified medical professionals.  Prescription-only products are “unavoidably unsafe” under any measure, and comment k should apply to both prescription drugs and devices.  Second, the regulatory pathway to marketing—here the FDA’s 510(k) premarket notification process—is irrelevant to whether a product is worthy of comment k’s protection.  The standard is whether a product’s use involves unavoidable dangers or risks, and prescription-only medical devices have known risks, without regard to how the FDA permitted the device’s marketing.  This ruling is yet another example of courts incorrectly marginalizing the 510(k) process and wrongly considering the process to be unrelated to product safety.

The court also rejected the defendants’ argument that the plaintiff could not prove that a safer alternate design existed, which was required under the applicable Texas law.  Id. at *8-*7.  The filter used to treat the plaintiff was a filter designed to be retrievable—that is to say, a physician could opt to remove the filter based on the individual patient’s condition and needs.  As a safer alternative, the plaintiff offered only filters designed to be implanted permanently, which are different products.  That should have been enough for the court to rule that the plaintiff had not offered an safer alternative design.  The court, however, found a triable issue of fact, mainly because the defendant had marketed the filter as both permanent and retrievable and also because the MDL had made a similar ruling in a prior case.  Id. at *7.

Which leads us to a third part of the order that piqued our interest.  Section 82.008(c) of the Texas Civil Practices and Remedies Code establishes a rebuttable presumption that a manufacturer is not liable for any injury to a plaintiff caused by a product if:  (1) the product was subject to pre-market licensing or approval by the federal government; (2) the manufacturer complied with all the applicable procedures and requirements; and (3) the federal government approved or licensed the product for sale after full consideration of the product’s risks and benefits.  Id. at *7.

Because the FDA cleared the device—i.e., licensed it for marketing—through the 510(k) process, the defendant was entitled to Section 82.008(c)’s presumption of non-liability.  The court, however, rejected the presumption based on the misperception that the 510(k) process does not focus on safety.  Id. at *8.  Other courts have made this mistake, and nearly all of them cite Medtronic v. Lohr, 518 U.S. 470, 478 (1996), which was outdated virtually since the day it was published.  As we have explained in detail, the 510(k) process very much focuses on product safety.  The defendant in Isaac should have received the benefit of Texas’ presumption.  Given the prevalence of preemption in design defect cases involving Class III medical devices, this holding has the effect of nullifying the Texas compliance presumption in medical device cases.  Either the claim is preempted, and the presumption is not necessary, or the presumption itself falls away.  Courts, particularly federal courts, should not be interpreting state statutes out of existence.

The court made other rulings.  It found that the plaintiff’s expert had raised a triable issue of fact on medical causation and future damages (Id. at *3-*5); it ruled that the plaintiff had presented evidence of negligent manufacturing through an expert stating that the plaintiff’s filter did not meet “performance specifications” (Id. at *7); it declined to apply Texas’ presumption of non-liability for products that comply with federal safety standards or regulations (Id. at *9); and the court applied Texas’ punitive damages law (which allows punitive damages) over Arizona law (which does not), ruling that the state where the plaintiff resided, was treated, and experienced her alleged injury (Texas) had a more significant relationship with the case than the state where the defendant resided and the product was designed and manufactured (Id. at *10).

So there you have it.  A beacon of light on the warnings claims, and not much to commend after that.  You might even call it a mixed bag (we made it all the way to the end before using that phrase).

Today we report on a recent decision dismissing manufacturing-defect, warranty, and failure-to-warn claims arising from an allegedly defective breast implant. Although the decision, D’Addario v. Johnson & Johnson, 2021 WL 1214896 (D.N.J. 2021), does not stray far from the beaten path, it covers ground worth revisiting. The decision is a useful (if cursory) reminder that plaintiffs challenging the safety or effectiveness of Class III medical devices with premarket approval may not rely on the res ipsa loquitur doctrine to satisfy their pleading obligations. It is also a good reminder that dismissal on pleading grounds is possible even when a court gets the preemption analysis wrong.

Plaintiffs sue when they are injured by a medical product (and sometimes when they are not). But injury by itself does not state a claim. To state a claim, plaintiffs must allege that they were injured because the product was defective in some regard. Recognizing that federal law generally preempts design-defect claims implicating medical devices with premarket approval, plaintiffs typically claim that their injury was caused by a manufacturing defect that the manufacturer failed to warn of.

To state a manufacturing-defect claim, a plaintiff must allege that the particular unit they received deviated from its intended design or from other ostensibly identical units, and that this deviation caused the plaintiff’s injury. Thus, in federal court, where Twombly and Iqbal reign, and in state courts that enforce fact-pleading standards, a plaintiff attempting to state a manufacturing-defect claim must allege facts that, if true, would show that their injury was caused by the particular unit’s deviation from its intended design.

Frequently, however, plaintiffs asserting manufacturing-defect claims do not allege any facts suggesting that that the particular unit they received deviated from its intended design. Instead, they rely on the res ipsa loquitur doctrine to evade their pleading obligation. When applicable, the doctrine allows a court to infer tortious conduct from the mere fact that an accident has occurred. Plaintiffs often maintain that they would not have injured by an FDA-approved product unless something went wrong during the manufacture of the product. In other words, plaintiffs often argue that one can infer from their injury alone that the unit they received suffered from a manufacturing defect.

The D’Addario court rightly rejected that argument. Indirectly quoting a recent Ninth Circuit decision in another breast-implant case, the court stated that plaintiffs “cannot simply demonstrate a defect or a malfunction and rely on res ipsa loquitur” to state a manufacturing-defect claim. 2021 WL 1214896, at *4 (ultimately quoting Nunn v. Mentor Worldwide, LLC, — Fed. App’x —-, 2021 WL 406304, at *2 (9th Cir. 2021)). As the Ninth Circuit had previously explained in another case, “the FDA’s premarket approval of the process by which a Class III device is manufactured ‘does not guarantee that every device manufactured in that process will work.’” Weber v. Allergan, Inc., 940 F.3d 1106, 1111 (9th Cir. 2019) (quoting Banner v. Cyberonics, Inc., 2010 WL 455286, at *4 (D.N.J. 2010)). On the contrary, “the FDA performs a cost-benefit analysis” when reviewing PMA applications and “may approve devices” and their manufacturing processes despite “knowing that [the devices] sometimes will fail.” Id. at 1111–12. Therefore, because a malfunction might occur even if the manufacturer adheres to the FDA-approved manufacturing process, one cannot infer a manufacturing defect from the mere fact that a Class III medical device has malfunctioned. Thus, a plaintiff cannot rely on the res ipsa doctrine to avoid having to plead facts that, if true, would establish a particular unit’s deviation from its intended design.

Evidently recognizing that their manufacturing-defect claims would be preempted unless based on the violation of a federal requirement applicable to the breast implant, the D’Addario plaintiffs, like plaintiffs in many medical-device cases, alleged that the implant’s manufacturer had violated the FDA’s Current Good Manufacturing Practices (CGMPs). In a misstep, the court stated, without any analysis, that purported CGMP violations “potentially set forth a state-law claim based on a violation of federal law that would fall within the ‘parallel claim’ exception to preemption.” 2021 WL 1214896, at *4 (emphasis omitted). While some other courts have also so held, yet other courts have rightly rejected that proposition. Although a future blog post will offer an additional ground on which CGMP-based claims should be rejected, the courts that have held alleged CGMP violations insufficient to support a manufacturing-defect claim that escapes preemption under 21 U.S.C. § 360k(a) as interpreted in Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), have generally done so on the ground that the CGMPs are, by design, “intentionally vague and open-ended.” Ilarraza v. Medtronic, Inc., 677 F. Supp. 2d 582, 588 (E.D.N.Y. 2009). Because the CGMPs “are inherently flexible” and “require manufacturers to develop their own quality-system controls,” these courts hold that any tort claim predicated on an alleged CGMP violation would inevitably be enforcing a state-law requirement “different from, or in addition to,” the federal requirements applicable to the device. In re Medtronic, Inc. Sprint Fidelis Leads Prod. Liab. Litig., 592 F. Supp. 2d 1147, 1157 (D. Minn. 2009).

As noted, the D’Addario was prepared to accept that manufacturing-defect claims predicated on alleged CGMP violation might escape preemption. Nonetheless, the court the court dismissed the plaintiffs’ claims on pleading grounds. Although the court’s reasoning is somewhat opaque, it seems to have found that the plaintiffs’ “laundry list of alleged CGMP violations” was, without more, insufficient to state a claim because plaintiffs did not tie the purported violations to “a defect in the manufacture of the specific” implant they received. 2021 WL 1214896, at *4. Their allegation that the defendants’ manufacturing process resulted in allegedly harmful matter being left on the implant’s surface was also insufficient, the court found, because plaintiffs failed to allege how, if at all, that process deviated from the process that had been approved by the FDA. Id. at *5.

As many other plaintiffs have done since the Ninth Circuit’s misguided and since discredited decision in Stengel v. Medtronic Inc., 704 F.3d 1224 (9th Cir. 2013), the D’Addario plaintiffs asserted a failure-to-warn claim based on the defendants’ purported failure to submit adverse-event reports to the FDA. The court did not reject the claim as preempted on its face, as it should have, but dismissed the claim nevertheless. Without holding that failure-to-warn claims can never be predicated on an alleged failure to file adverse-event reports, the court found that the plaintiffs had failed to state a cognizable failure-to-warn claim because their particular claim was based not on the failure to report known adverse events but on the failure to report adverse events that the plaintiffs alleged would have been discovered if the defendants had conducted the post-approval studies that they were supposedly required to conduct. That, the court said, was impermissibly speculative.

In Knapp v. Zoetis Inc., 2021 U.S. Dist. 63783 (E.D. Va. March 31, 2021), the plaintiff alleged that administration of an equine antibiotic caused his horse, Boomer, to experience “persistent lameness” and permanent damage to the “musculature in his neck.” Boomer was not okay. His condition was far from stable.

The plaintiff claimed that the manufacturer knew of similar negative reactions yet failed to supply proper warnings. The plaintiff’s complaint included causes of action for negligence, failure to warn, defective design and manufacture, and breaches of various express and implied warranties. The complaint did not specify under which law the negligence was brought, cited Virginia law for two of the warranty claims, and pointed to New Jersey for the rest of the causes of action. The claims under New Jersey law were styled as a class action.

Boomer and his owner lived in Virginia, while the defendant manufacturer was a citizen of New Jersey. (Quick: Name a famous equestrian from New Jersey.) The defendant filed a motion to dismiss the claims under New Jersey law. It also moved to strike the class action.

The federal court, sitting in Virginia, applied Virginia choice of law rules. Virginia applies the lex loci delicti – the law of the place of the wrong. In personal injury tort cases, the applicable law is typically where the plaintiff (or in this case, his four-legged companion) first became ill. Warranty claims can smell like both tort and contract claims, but either way, the state where the injury occurs supplies the operative substantive law. In Knapp, poor Boomer ingested the drug and suffered injuries in Virginia. In the race between Virginia law and New Jersey law, Virginia law is leading by several lengths as we round the last turn.

The plaintiff cited a New Jersey case that considered a claim by a plaintiff who alleged an injury in Virginia caused by a drug manufactured by a New Jersey company. But that was a horse of a different color. The New Jersey court held that New Jersey law governed the claim for punitive damages. That ruling was based on New Jersey’s “significant relationship” choice of law rule, which is different from Virginia law. Moreover, even that New Jersey case ended up applying Virginia law to the substantive claims arising from the injury in Virginia.

Again, everything of importance in the Knapp case happened in Virginia, so Virginia’s law applied, not the law of New Jersey, where the defendant had its principal place of business. Giddy up!

What is especially interesting about the Knapp case is that the plaintiff’s attempt to convert his individual dispute into a nationwide class action under a different state’s law failed, with the action cut back to an individual one under the law of the plaintiff’s state. Choice of law galloped straight into the issue of standing.

The plaintiff argued that a class action under New Jersey law would be the “best way to achieve an efficient manner of uniform redress.” Whoa! Efficiency was really a code word for ponying up big class action fees for plaintiff attorneys. That would be a lot of hay. But such specious efficiency cannot vault over fundamental rules of standing.

The Knapp court held that the plaintiff lacked standing to assert any action under New Jersey law because the use of the drug and the injury did not happen in New Jersey. The plaintiff did not allege an injury in fact occurring under New Jersey law. Consequently, the plaintiff lacked standing to bring a case on behalf of either himself or others under New Jersey law. The plaintiff’s lack of standing under New Jersey prompted the Knapp court to say Nay to the class action alleged under New Jersey law. As far as we are concerned, that is the main point of the Knapp case.

That is not to say that the state court rulings are devoid of interest. The Knapp court considered the negligence, implied warranty, and express warranty claims under Virginia law. The court concluded that the plaintiff might possibly be able to make out a claim for breach of an express warranty. But the negligence and implied warranty claims flunked because the complaint did not identify any specific defect in the drug at issue. All the complaint did was say there was a defect. The plaintiff contended that he did not need to identify a specific defect That was not a whinnying – sorry, winning – argument. The Eastern District of Virginia has already rejected the argument that a plaintiff need not identify a product’s defects at the pleading stage.

In the end, the plaintiff was left only with a single express warranty claim under the law of his residence, Virginia. From the plaintiff lawyer’s perspective, seeing a wide-scale class action cut down to such a small sliver must have seemed like a nightmare. At a minimum, the result must have produced a long face.

(Answer: Jessica Springsteen, daughter of The Boss. The location of Casa Springsteen supplies a clue: Colts Neck Township.)

This blogger took last week off.  Not just from blogging, but from the confines of home and office – which are still one and the same.  During those five days of much needed sun and sea, we had a friend stay at the house to watch the dog.  Our dog/house sitter cleared out before we got home, in an abundance of pandemic caution.  Arriving home with that feeling of exhaustion only a day of travel can provide, we discovered that the door leading from our garage into the house was locked.  We were happy that our friend secured our home but dismayed on several accounts.  We never lock that door.  Hence, we do not carry a key to it.  More troublesome was that we only had with us our car key fob since we assumed we would enter through the garage.  So, we had no key to the front door and no key to the back door.  What we did have was a 100-pound Labrador-Husky mix going wild knowing we were right outside.  If only he knew how to unlock the door.  Fortunately, we still have one member of the family young enough and small enough to fit through a window to the basement that we were able to open by popping out a window fan (note to self to fix security weak spot).  So, with the overly excited large dog licking her face, father and brother holding her ankles, and mother uselessly doubled over in laughter, said small teen managed to gain our access to home sweet home.  The front and back doors may have been blocked, but we found an alternate route in (and learned a lesson about carrying spare keys).  The same cannot be said for plaintiff in Carrozza v. CVS Pharmacy, — F.3d –, 2021 WL 1206542 (1st Cir. Mar. 31, 2021).

Plaintiff was prescribed an antibiotic.  But when the pharmacist went to fill it, he received a computer notice that the plaintiff was allergic to the drug.  Further investigation revealed conflicting information including statements by plaintiff himself that he was not allergic and prior filled prescriptions for the same drug.  According to pharmacy policy, it is up to the pharmacist to exercise his individual judgment given all of the information in deciding whether to dispense the drug.  In this case, the pharmacist did fill the prescription.  After ingesting the medication, plaintiff alleges he suffered an allergic reaction including permanent ocular damage.  Id. at *1.

The district court granted the pharmacy’s motion for summary judgment and plaintiff appealed.  The circuit court took up three general issues: (1) was removal appropriate; (2) was plaintiff’s expert properly excluded; (3) was summary judgment warranted.  The answer to each question was an emphatic “yes.”

After plaintiff filed suit in state court in Massachusetts, defendant removed the case based on diversity jurisdiction.  Plaintiff was a Massachusetts resident and defendant a citizen of Rhode Island.  Id. at *3.  While plaintiff tried to take issue with defendant’s citizenship, a far-flung argument, his primary challenge to removal was that the amount in controversy did not exceed the statutorily required $75,000.  In a pre-suit demand letter, plaintiff’s counsel made an initial demand for $650,000, which was met by an offer from defendant of $5000.  Id.  Plaintiff argued that the value of the claim for removal purposes should have been the defendant’s offer, not his demand.  The court pointed to several cases finding demand letters to be an appropriate basis for the amount in controversy.  Id. at *4 & n4.  Further, the court questioned the plaintiff’s sudden about-face “by admitting at best error, and at worst deceit, regarding the scope and merits of his claim.”  Id.

The court next moved to plaintiff’s expert(s).   First, plaintiff asked the court to admit as evidence an affidavit from his treating physician based on a Massachusetts state law hearsay exception for certain medical records.  The court denied the request because the rule did not apply in federal court and advised plaintiff to adhere to the federal rules regarding expert disclosures.  Id. at *5.  A few months later, plaintiff submitted an affidavit from a retained expert who purported to opine on the pharmacist’s standard of care and medical causation.  At his deposition, however, plaintiff’s expert testified that he did not know the applicable standard of care nor did he have any firsthand knowledge of plaintiff’s treatment or injury.  He relied solely on the previously excluded treater affidavit.  Id.

Seeing that the front door was about to be slammed shut, plaintiff headed for the back door.  Following the deposition, plaintiff moved to admit the treater affidavit under Federal Rules of Evidence 803(6) or 803(4) as a record of regularly conducted activity or a statement for purposes of medical diagnosis or treatment.  Since it was neither, the court called it what it really was – an untimely and therefore inadmissible expert report.  Id.

Still hoping to find an unlocked window, plaintiff moved for leave to depose the treating physician.  The court again denied plaintiff’s request citing that plaintiff had not disclosed the treater as an expert and had had ample time to depose him as a fact witness but never did so.  Id.  Plaintiff’s “good cause” argument for re-opening discovery was “that there was no issue with his [retained] expert up until transcript of his deposition testimony was introduced by [defendant] with [summary judgment] moving papers.”  Id. at *6.  So, the expert’s opinion was “valid” right up to the point he was asked to support it?  That’s grounds for excluding the expert, not for allowing plaintiff a third attempt to get in the opinion of his treating physician.  The circuit court noted that starting with plaintiff’s first failed attempt to admit the treater affidavit, plaintiff seemingly knew of the importance of the treater’s testimony to his case, but repeatedly ignored the district court’s instructions regarding the proper steps to admit expert testimony.  Id.  The district court did not abuse its discretion at any point regarding the affidavit or deposition of plaintiff’s treating physician.  Id.

Not surprisingly, defendant moved to dismiss plaintiff’s retained expert based on his testimony that he did know the applicable standard of care, was unfamiliar with plaintiff’s alleged condition, had no knowledge of the cause of plaintiff’s condition, or in fact whether plaintiff had any injuries at all.  With no support for his “opinions,” the district court properly excluded his testimony.  Id.

That leaves defendant’s summary judgment motion.  Plaintiff brought three claims against the pharmacy – strict liability (breach of warranty), negligence, and unfair competition.  Going in reverse order, the court found summary judgment was appropriate on the unfair competition claim because plaintiff did not allege or prove any “established” form of unfairness as required by the statute.  Id. at *12.

Plaintiff’s negligence and warranty claims presented issues of first impression under Massachusetts law.  The district court dismissed plaintiff’s negligence claim holding that it failed due to a lack of admissible expert testimony on standard of care and causation.  Id. at *8.  While no Massachusetts law deals explicitly with the need for expert testimony pertaining to a pharmacist’s professional judgment, there is ample support for that requirement in other professional negligence cases.  Id. at *9.  As in those other contexts, “the appropriate standard of care with respect to a pharmacist’s dispensation of prescriptions is . . . not normally within the experience of lay persons.”  Id.  This is not a case where a layman can infer from the facts whether defendant acted negligently.

Plaintiff’s last cause of action was a strict products liability claim premised on an allegation that defendant was liable for distributing an unreasonably dangerous product to plaintiff.  Id. at *10.  But Massachusetts does not recognize strict liability in tort apart from breach of warranty under the UCC.  Id. at *11.  A claim for breach of warranty only applies to a contract for a sale of goods.  Pharmacists, however, also provide professional services.  Therefore, the circuit court had to address another first impression issue – whether the predominant purpose of a pharmacist dispending prescription medications is the provision of a service or a provision of goods.  It decided the district court was correct in holding it was the former.  The court found support in Massachusetts’s definition of “dispensing” as “including the utilization of the professional judgment of the pharmacist.”  Id.  Afterall, pharmacists have stringent education and professional licensing requirements and act more as an extension of the prescribing doctor.  Because the provision of services predominates over the transaction in goods, defendant was entitled to summary judgment on a breach of warranty claim.

The First Circuit doubled checked every door and window plaintiff tried, but they were all locked up tight by the district court.  Not even a sliver to slide through.

We understand the value of branching out and trying new things.  Recently, a close companion prevailed upon us to try calamari (thumbs down).   We signed up for a Zoom acting class during the pandemic.  And we are seriously considering a “new” hair color, to take advantage of the natural phenomenon that has occurred during the past thirteen months.  But sometimes it is satisfying to do the same thing over and over again.  One year, we returned five times to see a relatively obscure Broadway show.  We order green curry every single time we eat Thai food.   And we will happily report on every good “warnings causation” decision that comes down the pike.   (Witness this post and this post and this post and this post.)

As most of you know, a plaintiff asserting a failure-to-warn claim bears the burden of proving both that the defendant’s warnings were inadequate and that the inadequate warning proximately caused her alleged injuries.  In a prescription drug or medical device case, in most jurisdictions governed by the learned intermediary rule, a plaintiff satisfies her burden of proving causation by adducing evidence that a different or stronger warning would have altered her doctor’s decision to prescribe the drug or device.  In the absence of such evidence, if the judge applies the standard correctly, the defendant is entitled to summary judgment.  (There are wrinkles:  some jurisdictions muddy this standard by accepting evidence that the stronger warning would have caused the doctor to warn the plaintiff differently, stopping short of requiring evidence that the warning would have changed the doctor’s decision to prescribe the product.)

The exceptions are jurisdictions that apply a “heeding presumption.”  In its strongest form, the heeding presumption creates a rebuttable presumption that a doctor would have read and heeded an adequate warning.  Once the plaintiff demonstrates that the warning was inadequate, the presumption kicks in and the burden shifts to the defendant to adduce evidence that rebuts the presumption.  Needless to say, we hate heeding presumptions.  We think it skates close to a due process violation to shift the burden of proving an essential element of a substantive claim on the strength of a presumption based on sheer speculation.  Today’s case, Zember v. Ethicon, Inc., 2021 U.S. Dist. LEXIS 52878 (E.D. Wisc. Mar. 22, 2021), turned on whether or not the plaintiff was entitled to a heeding presumption and on how that affected the “warnings causation” analysis.

In Zember, a pelvic mesh case remanded from the mesh MDL, the plaintiff was implanted with the defendant’s mesh device to treat her stress urinary incontinence.  On her master complaint, she checked off eighteen (eighteen!) claims she was pursuing against the defendant, including all manner of product liability, warranty, and fraud-based claims.  The defendant moved for summary judgment on many of the claims.  The plaintiff responded only to the arguments addressing the strict liability failure-to-warn claim.  (The court granted the unopposed portions of the motion.)

With respect to the failure-to-warn claim, the defendant did not argue that the defendant’s warnings were inadequate, skipping to the causation element of the claim.  As a threshold matter, the court acknowledged that, while Wisconsin had not formally adopted the learned intermediary doctrine, the Seventh Circuit had predicted that the Wisconsin Supreme Court would adopt the doctrine.  “Thus,” the court stated, “[we] will proceed as though the learned intermediary doctrine applies under Wisconsin law.”  Zember, 2021 U.S. Dist. LEXIS 52878 at *9 n.4.   As the court described the relevant test, a Wisconsin plaintiff must establish causation by proving that a proper warning would have caused the doctor to alter his behavior and would have prevented the plaintiff’s injury.  Id. at *11.

The defendant argued that the plaintiff had adduced no evidence that her doctor would not have implanted the mesh device if a stronger warning had accompanied the device.  In opposition, the plaintiff first invoked the “neglectful mesh plaintiff’s lament,” arguing that she hadn’t had the “opportunity” to depose her implanting physician because he had “been out of the country.”  The court responded with a figurative eye-roll, noting that the case had been pending since 2012 and that it “came from the MDL court with specific instructions that discovery not be permitted to continue.”  Id. at *11-12.

Next, the plaintiff argued that the court should apply a heeding presumption; in other words, a presumption that the doctor would have heeded an adequate warning and would have changed his prescribing decision based on the warning.  Id. at *12.  The court explained that Wisconsin courts had not spoken to the applicability of a heeding presumption to failure-to-warn claims involving medical devices but that the Seventh Circuit had expressed “serious doubt” that Wisconsin would adopt the presumption in this context (and that Wisconsin had rejected the presumption in other contexts).  The court concluded,

The Court will not apply a presumption that [the doctor] would have heeded an adequate warning if he had been provided one.  Plaintiffs cannot make a showing of causation as they have not procured any testimony from [the doctor] that he either relied on the warning provided or would have acted differently if given an adequate warning.  Thus, Plaintiffs have not raised a genuine dispute as to causation, and the Court will award summary judgment to Defendants” on the failure-to-warn claim.

Id. at *13.  We love this decision.  We are delighted that the court added weight to the “no heeding presumption in Wisconsin” conclusion.  And we are pleased that the court understood and applied the correct burden of proof to the causation element of the failure-to-warn claim.   We never get enough of courts getting this stuff right, and we hope we can bring you more examples.  In the meantime, stay safe out there (and try the green curry).