The day before this post went live there was a retirement ceremony for Judge Jon P. McCalla, the federal district judge for whom we clerked three decades ago.  After serving as an Army officer in the Vietnam War, Judge McCalla got his J.D. from Vanderbilt and clerked for Judge Bailey Brown of the Western District of Tennessee in a clunky federal building in downtown Memphis.  Less than twenty years later, he took over the same chambers where he had clerked.  A few years later, by the time a brash young clerk from the east coast started working for him, he had a sage Senior Judge Odell Horton down the hall and a fast-rising Chief Judge Gibbons upstairs.  In terms of our future practice, this was a very different era for large-scale drug and device product liability litigation.  Daubert was barely two years old, Lohr was about to be argued before the Supreme Court, and cases like Buckman and Levine were years away.  The plaintiffs’ bar was not nearly as organized or well-funded back then, and its advertising and case accumulation campaigns were not nearly as effective.  It would take quite some time before the mainstay legal theories and defenses were nearly as defined as we would like to think they are now.

Over the years, as we settled into our specialty and wrote for this Blog on the side, some things back in that Memphis courthouse changed and some stayed the same.  Judge Gibbons headed up to the Sixth Circuit in Cincinnati.  The federal building was renamed to honor Judge Horton after his death, with his name replacing that of an anti-desegregation congressman.  Judge McCalla stayed, rising to Chief Judge of the district and later taking on the mantle of sage Senior Judge himself.  We know that he always put in the time needed to address every motion and be prepared for every argument no matter what the case was about.  As for us, we have had plenty of drug and device cases in federal courts around Tennessee and elsewhere in the Sixth Circuit, as well as a few that went up the Sixth Circuit.  Honestly, we are not sure why there have been so many in those courts compared to others.

In In re: Glucagon-Like Peptide-1 Receptor Agonists (GLP-1 RAs) Prods. Liab. Litig., 24-md-3094, 2025 WL 2394047 (E.D. Pa. Aug. 15, 2025) (“GLP-1”), an MDL court in the Eastern District of Pennsylvania (within the Third Circuit, of course) was confronted with a motion to dismiss twelve counts in the very long master complaint.  We (particularly in its more common plural usage) have spent plenty of time litigating cases in this court and others within the same circuit.  We have also spent plenty of time spilling virtual ink about the tendency of MDL courts to lighten the federal pleading standards when it comes to master complaints, even though there is no hint of an MDL exception in the Federal Rules of Civil Procedure or TwIqbal.  See here and here for two MDL decisions from within the Third Circuit that made the bottom ten list for 2023.

One very unfortunate recurring issue in many of these decisions is that master complaints typically offer purported causes of action under all the states’ laws without differentiation, and, when these are challenged, the MDL courts punt on doing state-by-state analyses.  It should be pretty obvious that certain factual allegations may support a claim under one state’s laws, but not another state’s laws.  For instance, some states do not recognize strict product liability or punitive damages, whereas other states may have product liability acts that abrogate common law and set forth specific criteria for the statutory cause of action.  When a court does not engage in state-by-state analysis in this context, which would need to include appropriate Erie predictions and deference, the result favors plaintiffs.  So too in GLP-1, where the court denied portions of the motion to dismiss as long as a cause of action was stated under the law of “some states.”  See, e.g., id. at *17.  To prevail, the defendants had to show “such allegations fail . . . as a matter of every state’s laws.”  See, e.g., id. at *16.  From our perspective, the court ruling on a motion to dismiss a master complaint should not be scared to break up the ruling by state and, as needed, by drug, alleged injury, time of use, or other differentiators.  The parties can be directed to do the briefing in a way that makes it easier for the court to issue such a detailed decision.  It would make no sense, for instance, to issue a blanket denial of a motion to dismiss as to a purported claim for innovator liability because only 96% of the states do not have it.  In addition, as we have said many times (like here and here), outside of field preemption, a proper preemption analysis is almost always going to require determining what, if any, state law claims have been stated before going on to analyze preemption.  Whether because of the courts’ convenience or something else, the cumulative result is that MDL master complaints get more of a pass than they should.

We too will pass on most of the issues raised in the motion to dismiss in GLP-1 and focus on the issue of the preemption of the design defect claims being asserted.  From our perspective, this is a settled issue for branded prescription drugs like the ones at issue in GLP-1.  We do not have a problem with an MDL court explaining why these claims are preempted regardless of which particular plaintiff might asserts them.  GLP-1 did the expected walkthrough of Bartlett, Mensing, and Buckman, and rejected plaintiffs’ stale argument that Bartlett’s holding in favor of impossibility preemption was limited to generic drugs.  Id. at *24-26.  Noting three circuit decisions that had rejected plaintiff’s same argument and the regulations that preclude a unilateral change to the “qualitative or quantitative formulation of the drug product” approved under either an NDA or ANDA, GLP-1 came out the right way.  Id. at *25-26.  One of those circuit decisions was Yates, authored by the aforementioned Judge Gibbons, and what we think is the leading case on the issue.  (It also took the spot as our top decision of 2015.)

Plaintiffs offered three purported exceptions to the conflict preemption of their “pure” design defect claims, none of which made sense in light of Bartlett and the three circuit decisions GLP-1 had already cited.  The first, labeled “parallel misbranding,” was the only one where plaintiffs could cite a circuit decision that supposedly supported their argument.  We think the GLP-1 court was too generous in its consideration of this non-exception, ultimately rejecting it but allowing plaintiffs to take another shot at pleading it.  Id. at *27-30.  The “parallel claim” concept comes from express preemption, and even a state law claim that is found to be parallel to a federal requirement can be held to be impliedly preempted.  In the context of implied preemption, like the defendants argued in GLP-1, a design defect claim based on alleged “misbranding” under the FDCA would be preempted under Buckman and not be immune to Bartlett preemption.  The Darvocet case that the plaintiffs cited, discussed here, did not adopt a parallel misbranding exception.  Id. at *28.  As GLP-1 noted, no appellate court has.  Id.  Indeed, GLP-1 did not either, because the plaintiffs “have not provided legal or factual support for the Court to apply a parallel misbranding exception in this case.”  Id. at *29.  The fact that the Sixth Circuit, a year after Darvocet, issued the Yates decision without a hint of any such exception.  Allegations about misbranding are rife in drug and device complaints, especially MDL master complaints, yet no appellate courts have adopted the urged exception in eleven years since Darvocet.  With that record, we do not understand why repleading “to the extent [plaintiffs] can in good faith claim that the parallel misbranding exception applies” was authorized.  Id. at *30.  Certainly, some plaintiff will cite the foregoing quote as an endorsement of such an exception existing when it does not.

The remaining urged exceptions were dispensed without granting leave to amend.  A “marketing-based claims” exception would not have touched a “pure” design defect claim.  Liability based on the formulation of the drugs and liability based on marketing are separate issues.  Id. at *29.  Plaintiffs also argued that a design defect claim based on an allegation of a safer alternative design would not be preempted.  Of course, some states require proof of a reasonably feasible alternative design that would have avoided the claimed injury as part of their design defect cause of action, so nothing extra would be required for the made-up exception for some of the plaintiffs.  Even without examining state law—which would certainly have made sense—it should be fairly obvious that the logic of Bartlett does not accommodate such an exception.  Manufacturers of an approved NDA or ANDA drug cannot change unilaterally to any alternative design, no matter how much better it might be.  Id. at *30.  Although not discussed in this portion of GLP-1, Yates had examined the same issue in the context of a plaintiff suing under New York law, which required proof of an alternative feasible design.  It squarely rejected any exception to preemption of the design defect claim:

To imagine such a pre-approval duty exists, we would have to speculate that had defendants designed Ortho Evra differently, the FDA would have approved the alternate design. Next, we would have to assume that Yates would have selected this method of birth control.  Further yet, we would have to suppose that this alternate design would not have caused Yates to suffer a stroke. This is several steps too far.  Even if New York law requires defendants to produce and market a different design, the ultimate availability to Yates is contingent upon whether the FDA would approve the alternate design in the first place.

Yates v. Ortho-McNeil-Janssen Pharms., Inc., 898 F.3d 281, 299 (6th Cir. 2015).  Since Yates, many cases have preempted design defect claims asserted under state law requiring such proof.  Although GLP-1 did not let plaintiffs amend to try to assert a non-existent alternative design exception to avoid conflict preemption on their design defect claims, a little more work upfront would have been nice to shut down the inevitable re-treading that will be coming later.