Photo of Eric Alexander

Back in April, we pondered whether the new judge in the Valsartan MDL would change things for the better.  In contrast to the Zantac MDL, which was established a year later and has proceeded on a very similar contamination theory, the first several years of the Valsartan MDL saw a bunch of bad rulings on preemption, class certification, state law predictions, and other issues.  The decision that prompted our musings involved the exclusion of the plaintiff’s economic expert in a third party payor bellwether case and showed how seriously the court took its gatekeeping responsibilities as emphasized in the 2023 revisions to Fed. R. Evid. 702.  We offered our hope that the court would reexamine preemption or, “[i]f that is a bridge too far, then other holes in plaintiffs’ case—at least while it is based on the mere assumption of dangerous contamination—should present real obstacles for the Valsartan MDL plaintiffs.”  Before we cause major soft tissue injury trying to pat ourselves on the back for our prescience, last week’s decision out of the Valsartan MDL that granted summary judgments to the defendants in “what was to be the first bellwether trial”—in a case selected by the parties—was not exactly as we forecast.  In re: Valsartan, Losartan & Irbesartan Prods. Liab. Litig., No. 19-2875, 2025 WL 3131002 (D.N.J. Nov. 10, 2025) (“Roberts”), excluded plaintiff’s sole specific causation expert based on her unreliable differential diagnosis (or, more accurately, differential etiology) because she lacked a reliable basis to rule in the medication use and rule out several other risk factors for plaintiff’s decedent’s liver cancer.  To plaintiff’s credit, he conceded that the exclusion of this expert would be fatal to his case; to the expert’s credit, she conceded that some of the steps in her causation analysis were “pure speculation.”  Id. at *2 & 21.  Still, this is a major win and affirmation that the proper Rule 702 standards should not yield to the needs of the plaintiffs in an MDL.

Decisions excluding specific causation opinions are often not as useful as decisions excluding a general causation opinion because the former involve a case-specific inquiry.  When the expert purports to form her specific causation opinion based on conducting a differential diagnosis, it can be even harder to extrapolate a decision excluding that opinion.  For us, the key part of Roberts is the ruling on the unreliability of the plaintiff’s expert’s ruling in of the prescription drug use as a substantial cause of the decedent’s liver cancer.  We will focus on that.  The court’s finding that the expert’s ruling out of the decedent’s various risk factors for liver cancer was unreliable is also worth reading and, as the court noted, would have been a sufficient reason to exclude her opinion.  Id. at *27 n.54.  Neither, at least according to the court, involved deciding general causation.  Id. at *3 n.5.  (The differences between general and specific causation blur when the focus is on issues such as the level of exposure needed to cause a disease or the specificity of the disease that the exposure can cause (e.g., cancer versus liver cancer versus hepatocellular carcinoma).)  The decedent in Roberts had such strong and established risk factors for the type of liver cancer he had that we can see why the defendants picked the case as a bellwether.  We speculate that the plaintiff lawyers assumed that Alabama’s law on causation would help them advance their argument that the prescription drug use was the final straw; they clearly liked that the Roberts decedent’s cancer diagnosis came soon after the product recall and that his remaining pills were apparently shown to have much higher levels of NDMA than FDA set as the maximum.  (The reliability of that testing, which is certainly not a foregone conclusion given the history of the litigation, was also not addressed.)  Regardless of how it got selected as a bellwether, Roberts teed up some classic causation issues that should impact other plaintiffs and experts in the MDL.

Roberts started with the affirmation that the 2023 amendment to Rule 702 “clarified that the preponderance standard applies to the three reliability-based requirements added in 2000—requirements that many courts have incorrectly determined to be governed by the more permissive Rule 104(b) standard.”  Id. at *6 n.11.  Placing the burden on plaintiff as the proponent of the expert evidence helped the Roberts court reject plaintiff’s arguments that his expert should get a pass based on the relatively few epidemiologic studies directly on-point, that the court should defer a ruling on admissibility until after the jury would have already heard the expert’s testimony, that the court was usurping the jury’s role, and implicitly that the evidence plaintiff needed to get to trial in an MDL bellwether case should not be excluded.  Throughout, the court cited cases over the thirty-two years since Daubert that have emphasized the gatekeeping role, the focus on methodology, and what reliable causation opinions need to be based on—in our view, the more convincing cases.  Fundamentally, plaintiff’s expert was unable to articulate how much NDMA from Valsartan is needed to cause the kind of cancer that plaintiff’s decedent had.  Id. at *25.  We do think this is fundamentally a general causation question, but it is also a key question for a reliable differential etiology—you cannot include a particular exposure on the list of potential causes unless the relevant level of exposure is known to be a cause.  This was essentially the issue in General Elec. Co. v. Joiner, 522 U.S. 136 (1997), which involved causation opinions about cancer and workplace exposure to alleged carcinogens.  Joiner did not expressly distinguish between general and specific causation or describe the plaintiff’s experts’ approach as differential diagnosis, but the issues analyzed in Roberts are very similar.  We see this return to Daubert and its early progeny, aided by the clear allocation of burden, as a good thing.

Roberts proceeded to analyze the three types of evidence that the expert claimed supported ruling in plaintiff’s decedent’s Valsartan use as a potential cause of his specific kind of liver cancer.  First were the studies that looked at human use of Valsartan and liver cancer, of which two reported a statistically significant increased risk.  One stated that “[c]ausation cannot be inferred” from its results and the other described its results as inconclusive.  Id. at *22.  The expert tried to jack up the very low relative risk calculations from these studies with a series of unfounded assumptions about exposure levels and dose-response relationships, which the court saw as unreliable.  When we say the relative risk calculations were low, we are reminded of an expert we saw at a number of trials a long time ago.  He liked to tell the jury about a simple attributable risk calculation he did to say what percentage of the injuries in an exposed population came from the exposure as opposed to other risk factors.  For these studies, using the same calculation, 86.2% and 89.3% of the exposed population in these studies did not get liver cancer from the exposure.  This is why Daubert II and Havner required epidemiologic studies with statistically significant relative risks greater than 2.0—i.e., greater than 50% on an attributable risk exposed calculation—to establish general causation.  That may be a little far afield, but the low risks in these studies could not be reinterpreted upwards.

The expert also tried to rely on an occupational exposure study, which the Zantac MDL judge had rejected as a basis for a reliable general causation opinion.  Roberts followed the reasoning in Zantac (discussed in detail here and taking in our nod for fourth best decision of 2022 here), which included that the study’s author said it “was not designed to assess NDMA exposure through an oral medication.”  2025 WL 3131002, *23.  The expert’s last hope was to blur the lines with animal studies on cancer with NDMA exposure.  She lacked a reliable basis to extrapolate from the animals to humans or to establish a dose-response relationship tying plaintiff’s decedent’s purported NDMA exposure to a level that was associated with liver cancer in the study animals.  Id. at *24-26.  We were pleased to see the old Soldo decision—which focused on the testimony of an expert whose deposition in the case was the first one we ever took—still being cited for how to evaluate medical causation opinions that are purportedly based on animal studies.  The Roberts expert’s opinions did not come close to the standards for a reliable causation opinion.  Plaintiff also pushed that the animal studies supported a rapid onset from NDMA exposure to the development of cancer, but that was both speculative and insufficient to make up for the big holes in the expert’s causation opinion.

As we noted above, the Roberts decision also includes detailed discussion of the expert’s methodologic failures in attempting to rule out alternative causes for the plaintiff’s decedent’s liver cancer.  There were many.  However, rather than recapping those, we return to the court’s rejection of plaintiff’s argument that the court should just let the expert testify and be subjected to “vigorous cross-examination,” before the court rules on admissibility:

This Court shall do no such thing. Given this Court’s serious concerns with Dr. Siddiqui’s testimony, this would amount to a total abdication of its gatekeeping responsibility. The entire purpose of Rule 702 is to keep unreliable expert testimony from ever reaching a jury. And for good reason. How could the Court ever “unring” the proverbial bell after the jury had listened to hours of unreliable expert testimony? The prejudice to Defendants would be insurmountable.

Id. at *12 (emphasis in original).  Any federal judge who still thinks that the Federal Rules of Evidence favor admissibility of expert testimony, that the proverbial tie goes to the runner, and/or that cross can cure any expert admissibility issues better have a good answer to the question posed in Roberts.

Photo of Stephen McConnell

Council for Responsible Nutrition v. James, 2025 WL 3165673 (2d Cir. Nov. 13, 2925), is a Second Circuit decision about a New York state restriction on the sale of certain dietary supplements to minors. This blog covers the case because the court’s decision includes a disturbing preemption holding.  This particular blogger covers the case because it addresses a recent matter of particular concern that has arisen in more and more cases — the extent to which the government can and should act as our food nanny. It also helps that the CRN case is a bit strange.

To begin with, the “James” named as a defendant is New York Attorney General Letitia James, who right now has at least one case that is probably causing her more worry than this one. CRN, the plaintiff, is a trade group representing the dietary supplement industry. CRN filed the lawsuit to enjoin enforcement of section 391-00 of the New York General Law, which prohibits selling dietary supplements to anyone under age 18 if the supplement is “labeled, marketed, or otherwise represented for the purpose of achieving weight loss or muscle building.”  CRN’s theory was that the statute violated the First Amendment’s free speech clause, was unconstitutionally vague, and was preempted by federal law.  The district court denied CRN’s motion for a preliminary injunction because CRN was unlikely to prevail on the merits, could not demonstrate irreparable harm, and showed no favorable balance of equities. The district court also dismissed CRN’s claims except for the First Amendment claim. The case went up to the Second Circuit. 

The Second Circuit affirmed the district court, holding that it had not abused its discretion in concluding that the plaintiff was unlikely to win on substance, there was no irreparable harm, and the public interest would not be served by an injunction. 

The free speech issue hinged on a commercial speech analysis under the Central Hudson test. The Second Circuit held that protection of minors’ health was a substantial government interest, that the statute directly advanced that interest because youth appeared to be misusing the supplements, and the statute was not overinclusive as it was reasonable for the Legislature to focus on the product’s marketing. The Second Circuit also held that requiring retailers to age-verify that purchasers were adults did not constitute compelled speech. In sum, the statute passed intermediate scrutiny for laws restricting commercial speech.  

The Second Circuit also held that CRN’s vagueness claim fell short of the burden of showing “that no set of circumstances exists” for which the law’s application would be unambiguous.  

Now we get to the preemption ruling.  We are not pleased. Why?  We are in the land of express food preemption in section 343-1(a) of the Food, Drug, and Cosmetic Act (FDCA), which provides that “no State or political subdivision of a State may directly or indirectly establish under any authority continue in effect as to any food in interstate commerce … (5) any requirement respecting any claim of the type described in section 343(r)(1) of this title made in the label or labeling of food that is not identical to the requirement of section 343(r) of this title.”  Section 343(r)(1), in turn, prohibits health-related claims regarding a product’s nutrients, while section 343(r)(6) makes certain exceptions for dietary supplements. That is pretty strong and clear preemption.  We might even call it “plain.” (Foreshadowing.)

As the Second Circuit saw it, the issue was whether the statute’s “age restriction, by being triggered by (among other things) a health claim made on a product label, is therefore a ‘requirement respecting’ such a health claim.”  The Second Circuit answered that question with a No, and how it got there is a little wobbly. Seemingly contrary to the Riegel  discussion of “requirement” always meaning the same thing in the FDCA (552 U.S. at 324, not even mentioned in the CRN opinion) the Second Circuit held that the meaning of “requirement” in the preemption clause of the food section of the FDCA is not “plain” and is therefore subject to narrowing under the presumption against preemption – notwithstanding the Supreme Court’s Franklin decision abolishing any presumption against preemption in express preemption cases.  According to the CRN decision, Franklin’s erasure of the presumption again preemption applies only if the statutory text is “plain.”  That seems to us to be sloppy, lazy reasoning. It is as if the Second Circuit was looking for an excuse to avoid the job of statutory interpretation. This approach seems to ignore, if not outright defy, Congressional  intent. 

Then, almost unbelievably, it gets worse.  The Second Circuit proceeds to lean on Bates (which is not even an FDCA case), to hold in CRN that incidental economic effects that “do[] not require a … maker to put anything in particular on a product label” do not count as “requirements.”  

After stumbling over the express preemption issue, the Second Circuit’s perfunctory treatment of the irreparable harm and public interest elements seems inevitable. Lost sales would not be irreparable harm and the industry pecuniary interests “pale in comparison to the State’s goal of protecting youth from products that unfettered access to dietary supplements present.”  It is hard to tell what is worse – the logic or the prose.  

Photo of Michelle Yeary

Defendants know that the quickest way to level an uneven playing field is to… well… sprint. And few procedural tools reward nimble defense work quite like snap removal—the perfectly valid act of removing a case to federal court before the forum defendant is served. Plaintiffs call it gamesmanship. We call it reading the statute. Yet even among defense counsel, one deceptively simple question can cause more anxiety than a late Friday notice of deposition: When is removal actually “effected”? Unfortunately, the answer to that question depends on what circuit you are in.

Some circuits hold that removal is effective the instant a defendant files the notice of removal in federal court. No muss. No fuss. No waiting for a plaintiff to check their email. Efficient, predictable, and refreshingly literal. The kind of interpretation that warms a defense lawyer’s heart. Other circuits insist that removal becomes effective only after the defendant (1) files the notice in federal court, (2) provides plaintiff with written notice of the removal, and (3) files a copy of the removal notice with the state court. See 28 U.S.C. §1446(d). This is the procedural equivalent of “measure twice, cut once,” except that by the time you finish measuring, the plaintiff has already handed the forum defendant a summons. The split feels like a Philosophy 101 question: If a defendant files a notice of removal in federal court, but no one in state court hears it, has removal really occurred? Apparently, depending on the circuit, the answer ranges from “Absolutely yes,” to “Absolutely no,” to “Why are you asking us? We didn’t write this statute.”

And here is the reason we are drawing your attention to the divide. Snap removal is already a race, but MDLs turn it into a triathlon. Because once a case is swept into an MDL, the governing procedural law is not the law of the circuit where the case originated—it’s the law of the circuit where the MDL sits. This cannot be overstated. Once the case is transferred, the MDL judge applies the MDL circuit’s law on procedural questions like removal timing. So, as defense counsel we need to be on high alert as to which law applies.

As the defendant in Spillman v. Philips, 2025 WL 3162000 (W.D. Pa. Nov. 12, 2025) found out. The case is part of the In re: Philips Recalled CPAP, Bi-Level PAP, and Mechanical Ventilator Products Liability Litigation. Plaintiff was from Alabama. The defendants were all from Massachusetts.  Plaintiff filed in state court in Massachusetts. Before any of the forum defendants were served, the case was removed. Then, in chronological order, two defendants were served, plaintiff’s counsel received notice of the removal, and defendants filed the notice of removal with the state court.  Id. at *1.  And, before plaintiff’s motion to remand was heard the case was transferred to the MDL. So, the applicable law switched from First Circuit to Third Circuit.

The Spillman decision really does not address First Circuit law. So, we took a quick look and it does not appear that the First Circuit has addressed the specific issue of whether removal is effected at the time it is filed or only after notice to the plaintiffs and state court. However, we did find Berberian v. Gibney, 514 F.2d 790, 792 (1st Cir. 1975) that holds that “the jurisdiction of the federal court attaches as soon as the petition for removal is filed with it” (but that the state court retains concurrent jurisdiction until the notice is filed with it). Which we would argue favors removal being effective as of the federal filing. Unfortunately, the most recent decisions by courts in the Third Circuit favor removal requiring all three steps before it is effected.  Id. at *5-6. That is the law the court applied in remanding Spillman to state court.

Since our reason for this post is to advise defendants to know your circuits—plural—we also decided to poke around and see if we could give you a head start on that task. The following is not exhaustive, but we can report that there are few decisions on the specific issue as it relates to the snap removals. Adding to the First and Third Circuit law above, we found the following, which also demonstrate the divergent positions: Almonte v. Target Corporation, 462 F.Supp.3d 360, 365 (S.D.N.Y. 2020) (“In [the Second] Circuit, a federal court obtains jurisdiction when the notice of removal is filed with the federal court”); Burroughs v. Palumbo, 871 F.Supp. 870, 872 (E.D. Va. 1994) (finding federal jurisdiction attached when the removal petition was filed in federal court); Hamptom v. Union Pacific R. Co., 81 F.Supp.2d 703, 707 (E.D. Tex. 1999) (you need to complete all three steps before a removal is effected); Anthony v. Runyan, 76 F.3d 210, 214 (8th Cir. 1996) (“removal is effected when the notice of removal is filed with the state court and at no other time”). There are more decisions along these lines from other courts as well.

What does all this mean? Expect plaintiffs to push back. And expect courts to continue grappling with the split. Until the Supreme Court steps in (don’t hold your breath), inconsistency is part of the landscape. Snap removal is a tool that rewards being organized, proactive, and just a little faster than the other side. Knowing which circuit’s rule governs may be the difference between a case staying in federal court or being catapulted back into a plaintiff-friendly state venue. In short, defendants who understand the timing rules don’t just play the game. They control the tempo.

Photo of Bexis

As we’ve pointed out elsewhere, particularly in our duty to test cheat sheet, most states do not recognize any sort of separate negligence or strict liability claim for “duty to test” or “failure to test” separate and apart from the more usual sort of product liability claims involving the design and warnings of products.  So defendants looking to file motions to dismiss against testing-based counts of their opponents’ complaints could do worse than to check out our cheat sheet.

Manufacturers of FDA pre-market approved medical devices, however, have a second option.  They can also go after testing-based claims on preemption, because claims that manufacturers of such devices should have done more or different testing than the FDA considers necessary for approval are either “different from or in addition to” the FDA’s PMA criteria.  Indeed, the seminal express preemption case, Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), recognized that express preemption under 21 U.S.C. §360k(a) extends to “tort law, applied by juries under a negligence or strict-liability standard,” such as state tort claims alleging “negligence in the design, testing, inspection, distribution, labeling, marketing, and sale of [a PMA device],” id. at 320, although the dismissed testing claims were not separately discussed by the Supreme Court in Riegel.

Continue Reading PMA Preemption of Negligent Testing Claims
Photo of Bexis

Today’s guest post is by Howard Dorfman, a old friend of the Blog, who is now continuing to think important thoughts as an Adjunct Professor at Seton Hall Law School. This post concerns potential product liability implications of the FDA’s decision to release so-called “complete response letters (“CRLs”), apparently as a matter of future routine. We hadn’t written about that topic, and were more than pleased when Prof. Dorfman indicated his interest in doing so. As always, our guest posters deserve all the credit (and any blame) for their work.

**********

On September 4, the Food and Drug Administration (FDA) announced its intention to release future Complete Response Letters (CRLs) “promptly” following their being sent to drug sponsors. That day, the agency also released 89 previously unpublished CRLs issued from 2024 to the present related to pending or withdrawn New Drug Applications (NDAs) and other drug filing applications. According to the FDA, each of released CRLs details specific deficiencies in the drug filings relating to safety and efficacy identified by the agency that prevented approval of the application.

At the same time, the FDA released 89 new CRLs, adding to the more than 200 CRLs that were previously made public in July, ostensibly designed to allow third parties greater access to the reasoning behind the agency’s review process. However, the previous release of CRL letters only included rejection letters for drugs that were eventually approved. The September release represented the first time the agency had released such documents for non-approved drug submissions. Further, the FDA indicated it would henceforth release CRLs as they are issued, that is at the same time sent to the drug sponsors.

The September 4 pronouncement was not the first time FDA expressed its intentions regarding release of CRLs. On July 10, the agency announced its intention to embrace what it referred to as “radical transparency” by publishing more than 200 CRLs issued to a sponsor when it does not approve a New Drug Application, Abbreviated New Drug Application or Biologics License Application.

CRLs are official documents issued by the FDA to sponsors when the agency does not approve their submitted drug/biologic applications. It is important to recognize that a CRL is an intermediate step, not representing an outright rejection of the application. Rather, the CRL provides detailed feedback, often regarding safety, efficacy, or manufacturing issues, focusing the applicant’s attention on the application’s shortcomings and highlighting what would be needed for eventual approval.

CRLs often contain confidential and critical information regarding the feasibility of drug development, clinical trials, or manufacturing issues, including details that many companies choose not to publicly disclose. While the FDA touted its publication as a major “transparency” initiative, the 200 published letters did not reveal much to the public because they were for drugs that had already been approved. As a result, the relevant approval packages — including CRLs — were already publicly available and, in the end, only 22 of the 200 letters had not actually been published previously.

In the abstract, it is difficult to argue against greater transparency into government activities, particularly as pertaining to deliberations and decisions made that affect the availability of prescription drugs and biologics. However, the expansion of the process whereby CRLs are released has gone from limiting the release to CRLs for products that eventually obtained approval by the agency to those issued for submissions that have not yet been approved, including CRLs relating to applications that are voluntarily withdrawn or even abandoned. These latter documents had not previously been released by the agency, as it left to the discretion of the sponsor whether to communicate about rejected applications, including to indicate the reasons given for the rejection. Pharmaceutical companies generally release such information through press releases and/or in SEC-mandated filings.

It is well-established that pharmaceutical companies’ press releases can significantly increase their product liability exposure, particularly if the information can be characterized as false and misleading or failing to disclose important information. Historically, press releases have been considered a form of promotion and are subject to the same strict standards imposed by FDA regulations pertaining to promotional oversight by FDA for prescription drugs. 

The most significant issue impacting pharmaceutical companies’ liability is that immediate and unrestricted FDA release of all categories of CRLs is that the applicants have no time to react to the deficiencies that the FDA identifies. A successful regulatory response requires a thorough analysis of the FDA’s stated reasons presented for denial of marketing approval.  Such responses rarely can be developed and submitted to the agency within a relatively short period of time. In the interim, the CRL has been made public, where it can serve as an unwarranted imprimatur of FDA support for unsubstantiated causation allegations in future or current product liability litigation. The letters have the potential to serve as a source of expert opinion alleging product defect or company failure to undertake adequate clinical trials in support of their drug applications.

As recent litigation has shown, a CRL from the FDA can be a potent piece of evidence for plaintiffs’ experts in product liability litigation. By detailing what the FDA sees as deficiencies in an application, a CRL can support claims that a manufacturer failed to ensure a product was safe, effective, or properly manufactured before bringing it to market. Historically, plaintiffs’ experts have relied upon CRLs as a component of their testimony in litigation brought against a drug or biologic manufacturer.

Establishing evidence of a manufacturer’s negligence: Because CRLs frequently address deficiencies in the drug manufacturer’s manufacturing process, plaintiffs’ experts often use CRLs to make arguments that the manufacturer was aware of the product’s safety, efficacy, or quality issues and, despite this knowledge, continued to develop and sought regulatory approval. Moreover, an expert could compare the deficiencies cited in the CRL with the company’s public statements to investors in the form of press releases or in SEC-required disclosure documents or in its promotional materials and claim supposed discrepancies that in their opinion suggest negligence or fraudulent conduct.

Establishing causation: If a CRL documents a specific safety concern, identifies a problem in a clinical trial (such as criticizing its statistical analysis or sample size) or indicates the need to develop a robust Risk Evaluation and Mitigation Strategy (REMS) program, the applicant can expect to see expert reports parroting those statements as opinions that the individual or collective failures cited in the CRL were the most likely cause of the resulting injury.

Manufacturing defects: CRLs frequently cite deficiencies related to “Current Good Manufacturing Practices” (CGMPs), such as insufficient process validation or problems at a manufacturing facility. Even when such problems have been corrected, an adverse manufacturing expert witness can be expected to reference these findings to support a claim that the product was negligently made, which could have led to contamination or other harmful issues.

Labeling and “Failure to warn” claims: Labeling issues are frequent reasons for a CRL. A plaintiff-side expert will utilize labeling issues in a CRL to make claims that the manufacturer’s initial labeling was inadequate and failed to provide proper warnings about the product’s potential risks. The failure to warn allegations would no doubt follow the drug, beginning with approval process through marketing and throughout the product’s life cycle.

Of course, the defense will counter that notwithstanding statements made in a CRL, the most significant and definitive response to any allegations made by the plaintiff’s expert remains the fact that the reasons for the CRL were fixed, as demonstrated by the FDA’s eventual approval of the application. Final FDA approval, after any CRL problems were corrected, provides the most definitive evidence that the product is indeed safe and effective. Nonetheless, plaintiffs can – and do − point to the existence of a CRL to raise questions regarding the agency’s initial concerns with the product’s application and the approved product, thereby indicating that the company’s initial submission upon which the subsequent filing was based, was flawed and open the application and the approval process to scrutiny and remains relevant.

It is impossible to foresee how a trial or appellate court will view these arguments at the pretrial, trial and appellate levels – but the ongoing Fosamax litigation underscores how different courts have reached different conclusions about the same CRL.

Summary

Transparency appears to be the new “mantra” for FDA as demonstrated by the updated and expanded policies developed pertaining to the CRL process. While most of the discussion has focused on the impact on its impact on SEC disclosures and SEC litigation implications, an overlooked and equally important consideration for pharmaceutical companies is the potential impact on product liability litigation. In these circumstances, mere disclosures already part of SEC disclosure obligations may be insufficient in mitigating liability exposure for the pharmaceutical developer. As “radical” disclosure expands, companies will have to focus on existing policies and procedures in numerous aspects of the internal drug review and approval process to limit, if not to avoid, this increased risk.

Photo of Lisa Baird

We have been mulling over Loper Bright Enterprises v. Raimondo, 603 U.S. 369 (2024) and federal preemption. 

Yes, we need a life, but let’s put that aside for the moment.

In particular, we’ve been reviewing a rash of complaints where plaintiffs contend that the FDA’s decisions about whether to grant or deny premarket approval deserve no deference—and indeed should be invalidated—because the Agency made the wrong call and Loper Bright says they should be ignored.  Further, according to plaintiffs, an invalid PMA means manufacturers cannot assert a preemption defense, and they are off to the races with their tort claims faulting the FDA-required warning label, the mandated device design, and the scrutinized and approved manufacturing process. 

Loper Bright does not allow plaintiffs’ lawyers to overturn an exercise of FDA discretion in carrying out its assigned function in deciding whether to grant a PMA or PMA Supplement. Loper Bright recognizes that “[w]hen the best reading of a statute is that it delegates discretionary authority to an agency, the role of the reviewing court” is pretty much limited to “ensuring the agency has engaged in reasoned decisionmaking.”  Loper Bright also says nothing about agency policymaking or factfinding, which get very deferential review.  (See our post here.)

But more than that, these plaintiffs’ lawyers have the whole thing entirely backwards.  Loper Bright doesn’t undermine express medical device preemption, it gives manufacturers and defense lawyers a chance to re-set and expand how medical device preemption works.

Loper Bright involved the National Marine Fisheries Service and a federal law from 1976 (the Magnuson-Stevens Fishery Conservation and Management Act) that had something to do with the Atlantic herring fishery.  Something about the herring needing paid chaperones?  Or maybe the fishing boats?  (We are trying to care about the details, and failing.) 

Anyway, the National Marine Fisheries Service had issued regulations based on its interpretation of the statute, and the question was whether the Supreme Court had to defer to the agency’s interpretation of that statute or not.  Given that it is called the Supreme Court, you will not be shocked to learn that the Justices on said court decided that, while it is fine for a federal agency to exercise its discretion when Congress has authorized it to do so, interpreting federal laws is the job of the federal courts and they do not need the agency’s help with that, thank you very much.

Stated another way, Loper Bright got rid of a rule (Chevron deference) which had required courts to defer to “permissible” agency interpretations of the statutes those agencies administer, even when the reviewing court read the statute differently.  It freed federal courts from having to pay heed to an agency’s interpretation of a federal statute, and allows them to figure out the statute’s meaning on their own.

Ok, now let’s look at another federal law from 1976:  the Medical Device Amendments to the federal Food, Drug, and Cosmetic Act, and in particular the express preemption provision for medical devices, 21 U.S.C. § 360k(a). 

The first Supreme Court decision to interpret Section 360k(a) is one of our all-time bottom ten cases, Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996).  It is just a mess.  A plurality decision with no cogent rationale and conclusions that suggest the court perhaps forgot to read the statute it was interpreting.

How did Lohr get so off track?  Maybe it is because Lohr’s interpretation of 21 U.S.C. § 360k(a) was constrained by the FDA’s attempt to interpret that statute, 21 C.F.R. § 808.1(d). 

Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), tells us that: 

In Lohr, a majority of this Court interpreted the MDA’s pre-emption provision in a manner “substantially informed” by the FDA regulation set forth at 21 CFR § 808.1(d)….That regulation says that state requirements are pre-empted “only when the Food and Drug Administration has established specific counterpart regulations or there are other specific requirements applicable to a particular device.”

* * *

Informed by the regulation, we concluded that federal manufacturing and labeling requirements applicable across the board to almost all medical devices did not pre-empt the common-law claims of negligence and strict liability at issue in Lohr. 

* * *

Even though substantial-equivalence review under § 510(k) is device specific, Lohr also rejected the manufacturer’s contention that § 510(k) approval imposed device-specific “requirements.”

Riegel, 552 U.S. at 322 (emphasis added).

What Riegel is telling us is that because the Supreme Court in Lohr relied on an agency interpretation of a federal statute, it issued a cockamamie opinion.  One that said that general federal requirements aren’t “federal requirements”, and that device-specific 510(k) review doesn’t impose device-specific federal requirements.  Lohr made little sense, because the Agency’s interpretation of the “federal requirement” part of the statute made no sense.

Compare that with how the Supreme Court in Riegel handled the “state requirement” prong of 21 U.S.C. § 360k(a) after it concluded it was not constrained by the Agency’s statutory interpretation of 21 C.F.R. § 808.1(d).  (Yes, the Court has been throwing shade at deference to agency statutory interpretation for some time.)

The particular question for the Court in Riegel was whether “state requirements” include only positive enactments (statutes or regulations) or “state requirements” also encompass duties imposed tangentially, by imposing liability through common-law tort claims.

The FDA’s regulation said that “state requirements” in the federal statute only meant laws or regulations, not tort claims.  But the Court in Riegel didn’t agree, so it ignored the regulation and interpreted the statute as it wished:

All in all, we think that [21 C.F.R.] § 808.1(d)(1) can add nothing to our analysis but confusion.  Neither accepting nor rejecting the proposition that this regulation can properly be consulted to determine the statute’s meaning; and neither accepting nor rejecting the FDA’s distinction between general requirements that directly regulate and those that regulate only incidentally; the regulation fails to alter our interpretation of the text insofar as the outcome of this case is concerned.

Riegel, 552 U.S. at 329-30.

We would all have been better off if Riegel ignored all portions of the Agency’s statutory interpretation regulation, both this state requirement bit and the federal requirement bit that Lohr had mangled.

But now that Loper Bright is here, maybe federal courts can start the statutory interpretation process over again based in the text of 21 U.S.C. § 360k(a), and free from the shackles of Lohr as tainted by the misguided regulation.  Bexis certainly thinks so.

If express preemption analyses started with the language of the statute, and Riegel instead of Lohr,we could revisit the question of whether the 510(k) process imposes device-specific federal requirements (it does), particularly in its modern, more enhanced form.  And we could revisit the question of whether federal requirements are “federal requirements” even though they apply to all devices and not just a particular device (they do).

Photo of Stephen McConnell

As mentioned last week, we’re having an odd moment in our practice when virtually all our cases involve feverish battles over whether the cases belong in state or federal court.  Care to guess which side of the argument we’re adopting? (Hint: we prefer judges who will actually pay attention to dispositive motions and might even occasionally grant them.)

Speaking of odd, today’s case, Riddley v. Coopersurgical, Inc., 2025 U.S. Dist. LEXIS 215830, 2025 WL 3062996 (N.D. Texas Nov. 3, 2025), reached an odd result in an odd way. It might even be a first as far as we can tell. We hope not to see its like again. 

Riddley is yet another case in which the plaintiff sought to recover for injuries she allegedly sustained from the use of Filshie Clips, a federally regulated tubal ligation device.  The plaintiff filed her lawsuit in Texas state court against the manufacturer, the manufacturer’s parent company, and a distributor. The defendants removed the case to federal court based on diversity. After discovery closed, the plaintiff and the defendants filed motions for summary judgment. So far, so normal.

And then the magistrate-judge in Riddley sua sponte remanded an obviously diverse case at the summary judgment stage because there was no summary judgment-quality evidence of one defendant’s citizenship.  The evidence of citizenship – that the defendant in question “admitted” (in its answer) to diverse citizenship –supposedly wasn’t good enough at the summary judgment stage.  The decision brands that admission as a “stipulation between the parties” that could not establish diversity/subject matter jurisdiction.  “At the summary judgment stage particularly, the requirement is for evidence, and there is no record evidence in this case on which a reasonable jury could base a finding as to [the defendant’s] citizenship.” 

Since admissions are used routinely to establish all sorts of facts at trial, this ruling does not seem to make much sense. Nor, in light of the admission, does citizenship seem to be a “disputed” issue for purposes of Rule 56. Think about how much oxygen is expended by judges imploring the parties to agree on things. And now here, where the parties agree on something about as basic as can be, the court blows a raspberry at the parties and points them to the exit sign.

To our eyes, the magistrate-judge was standing hard on ceremony. The court emphasized how the scheduling orders directed the parties to Fifth Circuit precedents on subject matter jurisdiction and instructed the parties that any brief supporting summary judgment must be accompanied by “evidence sufficient to support a reasonable jury finding on any relevant jurisdictional facts.”  It seems as if the magistrate-judge wanted to teach the parties a lesson – a very hard lesson.  Another judge might have, after the scolding, permitted the parties to rectify the technical error/omission.  Not this magistrate-judge.  The case was remanded to state court.  Adios. 

But no matter how wrong we might think the Riddley decision is, it suggests a potential trap for the unwary.  After reading this blogpost, you are no longer unwary. 

Photo of Michelle Yeary

This post comes from the non-Butler Snow side of the blog.

Today’s case gives us two cautionary tales. First, there are many ways to make friends with a court. Dumping thousands of exhibits onto its doorstep and saying, “You figure it out,” is not one of them. Second, taking documents stamped CONFIDENTIAL or HIGHLY CONFIDENTIAL and casually using them in another case as if those words were decorative is sanction bait.

Heinrich v. Ethicon, Inc., 2025 U.S. Dist. LEXIS 216989 (D. Nev. Nov. 4, 2025), is a long-running pelvic mesh case—filed in 2013 in the MDL and remanded in 2020. A trial on statute of limitations only resulted in a plaintiff verdict. After which both sides sought additional discovery. Plaintiff’s request to re-depose the implanting surgeon was denied. However, plaintiff had a collateral state court action pending against the surgeon asserting breach of fiduciary duty and fraud. So, plaintiff took his deposition in that case. But here’s the rub—at that deposition, plaintiff showed the doctor confidential and highly confidential documents produced by defendant in the MDL. Defendant moved for sanctions.

In its motion, defendant claimed that plaintiff had violated the MDL protective order with respect to over 200 documents. In response plaintiff claimed that most of the documents had been publicly disclosed or de-designated prior to the deposition. After argument on the motion, the court ordered plaintiff to file a supplemental brief and specifically directed plaintiff not to include any additional argument and not to file any additional documents, but rather to only identify which documents she believed were not subject to the protective order and why.  Id. at *4-5.

Not only did plaintiff include additional argument in her filing, she attached 5,890 pages of exhibits and asked the court to figure out which were confidential and which were not. To nobody but plaintiff’s surprise, the court was not interested in playing “Where’s Waldo?” with confidential materials. It was plaintiff’s (really her counsel’s) job to sort through the documents. Instead, plaintiff violated the court’s order—which is never a good look—and asked the court to be an ”archaeologist” or “pig, hunting for truffles.” Id. at *6. We are all aware of the discovery strategy of burying your opponent in a mountain of paper and hoping they give up. Spoiler: judges don’t give up. They have long memories and limited patience for document dumps masquerading as compliance. So, the court struck everything except the portion of the supplement that contained a spreadsheet which attempted to show which documents had been previously disclosed. Id. at *7.

On the merits of the motion, the court found that as to twenty-five documents, plaintiff had violated the MDL protective order by using them in the state court litigation. As to those documents plaintiff “dumped” on the court exhibits, exhibit lists, hearing transcripts, and court orders from other cases and “speculated” that the documents may have been disclosed in the “hodge podge.” But again, the court was unwilling to “excavate masses of papers in search of revealing tidbits.” Id. at *18 (citation omitted). Because the MDL protective order explicitly limited the use of confidential materials to that litigation only, plaintiff violated the court order.

Nor could plaintiff demonstrate that the violation was “substantially justified.” Plaintiff easily could have put defendants on notice of the state court deposition and her intention to use the confidential documents. While the MDL protective order did not address this particular situation, it contained provisions for other situations where a party may want to use confidential information, calling for advance notice. Id. at *20-21. “Such notice would have given plaintiff a safe harbor, protected defendants, and brought about the confidentiality issue for resolution before the proverbial horse was out of the barn.” Id. at *22.

And this is most definitely why sanctions are warranted. The defendant did not find out until after the fact. The barn door was supposed to be locked, subject to notice and a chance to object—not flung open in the name of creative litigation strategy. Defendants shouldn’t be left chasing hoofprints and damage control. Courts have to reinforce that confidential actually means confidential. Here, that meant plaintiff was prohibited from using the state court deposition of the surgeon in the federal case, was prohibited from using the 25 confidential documents in any other action, including the action against the surgeon, and had to pay reasonable attorneys’ fees and expenses in connection with bringing the sanctions motion. Id. at *22-23.  

Between the data dump and the protective order violation, this case was fraught with procedural mischief and disrespect for the court. So, the morals of the cautionary tales—follow the court’s orders. The court — and your credibility — will thank you.

Photo of Bexis

Back in the bad old days of the Bone Screw litigation, we had to fight our way through a thicket of scurrilous allegations about how our clients supposedly promoted off-label use through continuing medical education seminars that the Bone Screw plaintiffs claimed were used to reward surgeons who regularly used our clients’ products with excessive speaker fees.  Back then – in the mid 1990s – the plaintiffs’ preferred avenues for asserting such allegations were state-law based:  negligence per se, fraud on the FDA, and conspiracy.  By the time that the infamous Franklin False Claims Act (“FCA”) decisions came down (United States ex rel. Franklin v. Parke-Davis,147 F. Supp.2d 39 (D. Mass. 2001), and United States  ex rel. Franklin v. Parke-Davis, 2003 WL 22048255 (D. Mass. Aug. 22, 2003)), we had won Buckman (and a lot of other things), so Bone Screw-related promotion allegations were never the subject of FCA litigation.

But the Bone Screw promotional allegations were close enough to what has been subsequently alleged ad nauseum in FCA litigation that we’ve followed similar FCA litigation ever since.  Today’s case, United States v. Gilead Sciences, Inc., 2025 WL 2627686 (E.D. Pa. Sept. 11, 2025), does not involve off-label use, but does involve allegations of kickbacks – through speaker programs and donations to charitable organizations.  We’re happy to say that the entire action was dismissed – on both sets of facts.  We’re even happier to recommend the discussion in Gilead as providing useful guidance for how pharmaceutical companies can manage both types of programs in compliance with applicable law.

Continue Reading FCA Dismissal Illustrates Pharmaceutical Promotion Done Right
Photo of Eric Hudson

We’ve blogged a lot recently about preemption and the dismissal of complaints alleging that certain over the counter products, including acne medications, sunscreens, antiperspirants, expectorants, and shampoos contain benzene.  Almost a year ago we blogged about the dismissal of an OTC case involving medicated shampoo that allowed plaintiff leave to amend. Today’s decision, Pineda v. Lake Consumer Products, Inc., 2025 WL 2698991 (E.D. Pa. Sept. 22, 2025), is a mixed bag that addresses plaintiff’s amended complaint. It’s about coal-tar shampoos, which are known to include benzene and are subject to an FDA monograph that recognizes the naturally occurring presence of benzene in coal tar. Yet, shockingly, plaintiff filed a class action claiming she would not have purchased the shampoos had she known they contained benzene.

Continue Reading OTC Preemption Letdown in the Eastern District of Pennsylvania

Drug & Device Resources

Bexis’ Book

FDA-regulated products now account for an estimated one-fifth of overall economic activity in the U.S. They have also been the focus of a litigation explosion. This timely guide covers all aspects of litigation involving drugs, medical devices, vaccines and other FDA-regulated prescription products.
Learn more about this book