Although Czimmer v. Janssen Pharmaceuticals, Inc., 2014 WL 1335511 (Pa. C.P. Philadelphia Co. Jan. 2, 2014), was decided several months ago, we only became aware of it (because of our ongoing automatic searches) the other day. While there are a number of grounds for feeling uneasy about what went on in Czimmer, we’re focusing on only a couple. To us, Czimmer is a poster child for what can go wrong when FDA regulatory “experts” spouting legal conclusions are allowed to run wild in a courtroom. It’s also a preemption case, or should have been.
In Pennsylvania, as elsewhere, it is “well-settled that an expert is not permitted to give an opinion on a question of law.” Waters v. State Employees’ Retirement Board, 955 A.2d 466, 471 n.7 (Pa. Commw. 2008). “Just because a hired expert makes a legal conclusion does not mean that a trial judge has to adopt it.” Gregg v. V-J Auto Parts, 943 A.2d 216, 223 (Pa. 2007) (barring testimony on what constituted a “substantial factor”). “[E]xpert opinions relating to the meaning of laws are not relevant evidence.” Newtown Square East, L.P. v. National Realty Corp., 38 A.3d 1018, 1026 n.10 (Pa. Commw. 2011). Therefore, an “expert opinion on a question of law is inadmissible,” 41 Valley Associates v. Board of Supervisors, 882 A.2d 5, 14 n.12 (Pa. Commw. 2005); see Browne v. Pennsylvania DOT, 843 A.2d 429, 433 (Pa. Commw. 2004) (“[l]egal opinion testimony is not admissible”), and “admission of incompetent opinion evidence which goes to the legal conclusions to be drawn by the factfinder constitutes reversible error.” Taylor v. Fardink, 331 A.2d 797, 799 (Pa. Super. 1975). Expert opinions have “no effect” on questions of law. Novak v. Jeannette District Memorial Hospital, 600 A.2d 616, 618 (Pa. Super. 1991).
Experts may not offer legal opinions because “[e]ach courtroom comes equipped with a ‘legal expert,’ called a judge, and it is his or her province alone to instruct the jury on the relevant legal standards.” Burkhart v. WMATA, 112 F.3d 1207, 1213 (D.C. Cir. 1997) (citation omitted). Precluding experts from offering legal conclusions prevents them from “usurping [the court’s] role.” Bergman v. United Service Automobile Association, 742 A.2d 1101, 1108 (Pa. Super. 1999) (excluding opinion defining “bad faith”) (footnote omitted); see March v. Downingtown Area School District, 775 A.2d 876, 880 (Pa. Commw. 2000) (affirming exclusion of expert opinion “more in the nature of argument on legal questions that were . . . for the court to decide”). Thus, a legal standard is “not an issue of fact subject to expert testimony, but [is] an issue of law for the trial judge to decide.” Commonwealth v. Possinger, 399 A.2d 1077, 1083 (Pa. Super. 1979).
This rule was ignored in Czimmer, with the predictable effect that legal questions concerning FDA regulations were improperly submitted for a lay jury to decide. The legal question involved the FDA’s classification of a drug (Topamax) “as a Category C drug in the FDA Pregnancy Category.” 2014 WL 1335511, at *2. Two FDA pregnancy categories were at issue in Czimmer, categories C and D.
According to information available on the FDA’s website, “[i]n 1979, the A, B, C, D, and X category classification that is based on the adverse events and potential benefits of a particular drug was implemented in the United States in order to standardize pregnancy labeling.” These are:
- A − Adequate, well-controlled studies in pregnant women have not shown an increased risk of fetal abnormalities.
- B − Animal studies have revealed no evidence of harm to the fetus, however, there are no adequate and well-controlled studies in pregnant women or Animal studies have shown an adverse effect, but adequate and well-controlled studies in pregnant women have failed to demonstrate a risk to the fetus.
- C − Animal studies have shown an adverse effect and there are no adequate and well-controlled studies in pregnant women; or no animal studies have been conducted and there are no adequate and well-controlled studies in pregnant women.
- D − Studies, adequate well-controlled or observational, in pregnant women have demonstrated a risk to the fetus. However, the benefits of therapy may outweigh the potential risk.
- X − Studies, adequate well-controlled or observational, in animals or pregnant women have demonstrated positive evidence of fetal abnormalities. The use of the product is contraindicated in women who are or may become pregnant.
- NA − FDA pregnancy rating not available.
In 2004, the drug was approved by the FDA under Category C. 2014 WL 1335511, at *2. Plaintiff used the drug in 2006. According to the decision, however, the plaintiff offered evidence of “pregnancy reports in 2000, 2002, and 2005 in the possession of the Defendant indicating the contrary.” Id. at *2. All this material was concededly submitted to the FDA in a timely fashion, id., so we’re not sure what the point of it was, except as an attack on the FDA’s initial classification decision. As we discuss below, attacks on FDA classification decisions in state-law should be preempted.
Here’s where the regulatory experts began running amok. Plaintiff’s regulatory expert “maintained that the Defendant was negligent in failing” to send a “Dear Doctor” letter (technically a “Dear Healthcare Provider” (DHCP) letter – but this is Philly CP) discussing these reports. Id. at *2. To us, allowing an expert to opine on whether a defendant is “negligent” is a no-no. It’s no different than the opinion on what constituted “bad faith” in Bergman or on what was a “substantial factor” in Gregg.
But compared to what follows, that was minor league.
The crux of [the expert’s] opinion dealt with the fact that although the FDA had in its possession the adverse studies/reports and did not order any labeling change, the Defendant was nonetheless obligated to inform physicians/healthcare providers of this information in order for these would be prescribers to make a knowing and intelligent decision as to whether or not to prescribe the drug to their patients.
2014 WL 1335511, at *2. That’s not even a proper negligence opinion. The FDA does not allow regulated companies to send Dear Doctor letters willy-nilly. The FDA has guidelines on when such letters may be sent.
According to the FDA, Dear Doctor letters are for “important new information” – they’re not for rehashing something that doctors already should know:
[A] DHCP letter is used to inform health care practitioners about important new information about a drug. In most cases, the new information is about an important new safety concern that could affect the decision to use a drug or require some change in behavior by health care practitioners, patients, or caregivers to reduce the potential for harm from a drug. In some cases, the new information is about how to improve the effectiveness of a drug.
Draft guidance, at 3 (emphasis added). New. New. New. New. Moreover, Dear Doctor letters should “avoid discussion of non-critical information that could obscure the more important information.” Id. at 4. The only other FDA-recognized use for a Dear Doctor letter is “to correct misinformation in advertising or other types of prescription drug promotion.” Id. at 3. The FDA’s guidelines aren’t legally binding, but they are legal in nature. The judge – not an expert – thus should have decided (after briefing by the parties) whether a Dear Doctor letter was required/permissible under the FDA’s standards. Dear Doctor letters are drug labeling, so the FDA’s standards control
The next relevant issue in Czimmer involved a consent form from an FDA clinical trial concerning the drug. Plaintiff was never enrolled in this, or any other clinical trial. 2014 WL 1335511, at *3. The FDA’s regulation of informed consent in clinical trials – and thus whether a clinical trial informed consent forms bear on anything outside such trials – is complex and detailed. See 21 C.F.R. Parts 50.1, and 56. Instead of deciding anything, however, the trial court allowed, indeed encouraged, expert witnesses to usurp its role:
The use of the consent form in a clinical trial, where people volunteer to be subject of a study is much different from what information would be contained in the labeling of a drug. . . .Defendant’s regulatory expert  testified that the consent form warnings should never appear in the labeling because they are based on entirely different criteria. It became clear to the Court based on the conflicting expert opinions that there existed a viable factual dispute as to the role the consent form played in the negligence issue. It was therefore proper and correct of the Court to admit the consent form, leaving it to the jury to decide whether or not and to what extent it contributed to the negligence of the Defendant. Keep in mind that the Defense countered [plaintiff’s expert’s] opinions through their regulatory expert  who had worked for the FDA, who opined otherwise.
2014 WL 1335511, at *3. Whether or not the FDA’s informed consent regulations for clinical trials mandate pregnancy warnings in all consent forms, regardless of what is actually known to science, or whether the forms reflect the state of the art (and thus could be relevant) is again a legal issue for the court to decide. Either the regs do, or they don’t. Conversely, this question is not a matter for experts to dispute and a jury to decide.
The third legal question we think was improperly submitted to the jury involved pregnancy Category C versus pregnancy Category D. The plaintiff’s causation case hinged on the prescriber’s statement “that she would not have prescribed the drug in the first place if there had been a warning on the label that there were studies showing potential harm to human fetuses as contained in Category D.” 2014 WL 1335511, at *3. The basic problem is that the drug wasn’t classified by the FDA as Category D. It was classified as Category C (that’s since changed since, something not apparent from the opinion). Once again, the trial court punted the issue – what was the proper classification of a drug – to expert regulatory witnesses:
Plaintiff’s expert  opined that merely because the FDA, in possession of the adverse event reports did not order a change in the labeling did not absolve the Defendant.  Defendant’s regulatory expert, stated otherwise. Thus, it was proper for the Court to leave it to the jury to decide which opinion to accept.
2014 WL 1335511, at *5.
That’s three, and this one is the worst of the bunch. Not only is whether a drug meets the FDA’s regulatory criteria (set forth above) for Category C versus Category D a legal opinion based on those criteria – it is a legal decision beyond the competence of a state court to adjudicate. What classification a regulated product falls into, whether it’s a drug’s pregnancy category or a device’s regulatory pedigree (pre-market approval, as opposed to §510k clearance), or whether it’s a combination product regulated as a “drug” or “device” – is a matter delegated solely to the discretion of the FDA.
There doesn’t seem to be a fraud on the FDA claim underlying Czimmer, since there appears to be no dispute that everything the FDA should have considered was in fact in the hands of the Agency. Otherwise, however, the claim being asserted in Czimmer shares the same pernicious effects as the claim in Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001), that warrant an application of preemption. At bottom the claim in Czimmer – the only one where there’s evidence of warning causation – has to be that the FDA got it wrong; that based on all the evidence that the FDA had before it (according to plaintiff’s experts) the drug should have been in Category D. Otherwise, the prescriber’s testimony that a Category D warning would have made a difference in the prescribing decision, 2014 WL 1335511, at *3, is immaterial because no factual basis for that statement existed.
But like a fraud on the FDA claim, a claim that the FDA made a mistake and should have placed the drug in Category D depends on the jury being allowed to accept a contrafactual assumption about the FDA: that the Agency should have done something that it, in fact, did not. The jury was permitted to find that the FDA should have taken a different regulatory action; specifically, that it should have classified the drug into Category D. This finding, of course, ignores, what the FDA actually did at the relevant time, which is classify the drug into Category C. This is a fundamental, inherent conflict between what the FDA did and what the jury found it should have done. That has to be preempted under Buckman, because federal law, reflecting an in-force decision by the FDA, is supreme over any state-law jury’s supposition that the FDA should have done something else. Even if plaintiff’s claim were not predicated on FDA having been wrong and Defendant needing to do things it could not have done by itself − and it was − the preemption principles at play here still should require exclusion of expert testimony like plaintiff offered.
Nor is this Wyeth v. Levine, 555 U.S. 555 (2009), as Czimmer held. 2014 WL 1335511, at *5. Levine turned on the defendant (supposedly) having the unilateral power under an FDA labeling regulation to change its label. There’s no similar unilateral power for a drug manufacturer to change a drug’s classification. That’s for the FDA to decide, not a manufacturer. So the situation is closer to PLIVA v. Mensing, 131 S. Ct. 2567 (2011), as the defendant argued. 2014 WL 1335511, at *5. But the conflict here is deeper than Mensing, since that case only dealt with a label change, whereas the plaintiff’s theory in Czimmer depended on the jury being told it could ignore the FDA’s basic classification of the drug, because based on all the information before it, the FDA got it wrong. Mensing is relevant, because a drug’s classification can be changed only after the FDA says so. To us, however, the case is much closer to Buckman, because Buckman shares the same type of fundamental conflict – a state-law jury cannot be allowed to ignore what the FDA actually did in the real world in favor of what that jury thinks that the FDA should have done, or what the law should be – in whatever fantasy world the plaintiff in Czimmer concocted from her experts’ unfettered legal conclusions.