Photo of Stephen McConnell

Every once in a while in this space we summarize law review articles.  In the course of doing so, we typically pat ourselves on the back by announcing that we read such articles so that you don’t have to.  That is not true with the article we are discussing today, Goldberg, Gramling, & O’Rourke, “A Prescription for Pharmaceutical Preemption,” 20 Rutgers J. L. & Public Policy 209 (2023).  The article is on such an important topic (it feels as if we blog about preemption at least twice per week), and it is written so clearly and concisely that you will be cheating yourself if you do not read it.   (We know one of the authors and have tremendous respect for his intellect, but we do not think our admitted bias accounts for our high regard for the article.)

The article describes the inherent tension between FDA regulation, which necessarily takes a macro view of risks and benefits, and jury trials, which focus on the harm alleged by one unfortunate person.  Preemption ensures that the latter does not wreck the former.  Threats of such wreckage abound. Juries lack scientific expertise.  Just as bad is the fact that juries have “no line of sight to the people who benefit from the drug, and cannot assess the potential negative impact that changing the warnings to address the plaintiff’s situation in the case may have on others.”  The faux morality play produced by the plaintiff lawyer at trial might shower that lawyer and the client with dollars, but it also showers the medical/scientific landscape with uncertainty and inconsistency. 

The article does a good job of countering the plaintiff lawyer fable that FDA regulation is inadequate.  We take a tour of FDA history and get a schooling as to how rigorous the drug (and drug label) approval process is.  Here is compelling proof that the FDA does not merely wave Big Pharma drug applications by on their way to the marketplace: the FDA has approved on average only “38 new drugs per year over the past decade and only 12% of drugs entering clinical trials are ultimately approved.”

The focus of the article is on label warnings, and the article reminds us that the hierarchy of label information is designed to prevent over-warning, which can be just as much of a problem as any alleged under-warning.  The goal for each label is to “portray the drug’s safety profile with accuracy, balance, and brevity.”  But we know that plaintiff lawyers are trying to win a case.  They care not a fig for accuracy, balance, and brevity.  Unfortunately, even a halfway clever plaintiff lawyer can always find something missing from a label, whether it be a failure to dramatize a risk (not saying the risk “is really, really bad”), or, in fact, saying anything less than “Don’t you dare take this horrible drug!”

When we are considering preemption in the context of branded prescription drugs, more often than not the central issue is whether the brand manufacturer learned some additional risk information after FDA approval, and should have added warnings via the Changes Being Effected (CBE) provision in the FDA regulations.  The article goes through the key SCOTUS cases in this area — Levine, Mensing, and Albrecht. (We feel a little bad for Bartlett.  Why was it omitted from this jolly party?) To be sure, Albrecht gets star billing.  It is the latest and greatest SCOTUS pronouncement on branded drug conflict preemption.  Interestingly, SCOTUS took note in Albrecht that the FDA, through the Solicitor General, had “filed a brief in support of preemption in the case to guard against over-warning.”  Albrecht is so important because it recognizes that preemption is fundamentally a legal question.  As the article explains, “[t]he issue of preemption does not lend itself to a battle of the experts, where lay juries can be improperly influenced into overturning an agency’s scientific determinations.”

In a section of the article titled “Preemption: A Demanding Yet Attainable Defense,” we learn how courts have applied Albrecht.  In the Gibbons case (which we praised here), the Second Circuit developed a burden shifting approach.  The first inquiry is whether the plaintiff met the burden of demonstrating that the CBE process was available to the defendant.  More specifically, did the plaintiff show that there was “newly acquired information” of a “causal association” between the drug and the alleged injury?  Only once the plaintiff meets this burden does “the burden shift to the manufacturer to assert a valid preemption defense, namely by showing that the FDA was fully informed of any such new information and that there is clear evidence the FDA would not have approved the labeling change.”  Per Albrecht, these are questions for judges, not juries.

Plaintiffs frequently get tripped up by the threshold question of whether there was any “newly acquired” evidence that should have prompted a CBE.  Courts have held that draft studies, mouse studies, and adverse events do not constitute “newly acquired” information revealing risks of a different type or severity or frequency than those previously included in submissions to the FDA.  Faced with their inability to satisfy their initial burden of showing newly acquired evidence, some plaintiff have argued that Albrecht actually shifted the burden to defendants to show the non-existence of newly acquired evidence.  But most courts reject this plaintiff argument, reasoning that it would be unfair to make the manufacturer “prove a negative that it acquired no new information” to justify a CBE modification.

If the plaintiff does manage to muster up some newly acquired risk evidence, then the issue is whether there is “clear evidence” that, even with the newly acquired information, the FDA would not have approved the labeling change proposed by the plaintiff (assuming the plaintiff ever flat-out states what the warning should have been).  At this point, the article illuminates the bounds of this inquiry by reviewing what happened in Albrecht after the SCOTUS opinion.  The district court in Albrecht ultimately held that the defendant had “fully informed the FDA of the justifications for its proposed warning” and that “the FDA’s rejection was predicated on insufficient evidence of a causal link” between the drug and the alleged injuries. It helped that the FDA’s amicus brief specifically said that the FDA was fully informed of the relevant medical risk.  We should all be so lucky to have the FDA weigh in on preemption.  The frustrating thing is that we know if the FDA did lob in its two cents, it would almost always say that the label was appropriate, that the agency was fully informed, and that the plaintiff’s proposed warning is nutty.

Albrecht and its progeny contain some goodies that we defense hacks can use as we fight the preemption wars.  For example, most courts reject the plaintiff “formalistic approach” of requiring “final agency action rejecting the specific wording at issue” to support preemption.  If the plaintiff approach prevailed, companies would have an incentive to litter the FDA’s inbox with proposed warnings based on thin or bad science, and that scenario would benefit neither the FDA nor the public.  Some plaintiffs argue that the FDA’s denial of a Citizen Petition requesting a label enhancement is not enough to show that the FDA would have rejected the same request from the manufacturer.  Courts usually do not buy that argument.  Science is science, and the mouthpiece for the science does not change anything.  As the article tells us, the “citizen petition process is a formal one and provides the ‘official administrative record for an FDA decision’ Albrecht finds to be a proper basis for preemption.”

Near its end, the article returns to first principles.  There is a legitimate health care reason not to slap on additional warnings: “the FDA’s standard for requiring a warning label is ‘different from that imposed by state tort law’ and ‘the manner in which state law tort principles drive the labeling of products as a general matter.’” The FDA carefully calibrates how to “provide accurate information to medical professionals and patients without unduly discouraging the use of the product.”   The FDA approved label represents a balance, not a floor.  It is bad science casually to upend that balance and flip CBEs around like nickels.

This brief summary does not quite do justice to the brief article.  We see too many academic articles that seem biased toward speculation, soft-headedness, and, finally and inevitably, liability.  By contrast, the Rutgers article contains sober legal analysis, clear thinking, and realism.