Photo of Eric Alexander

Implied Preemption.  Off-label promotion. TwIqbal.  They make up a core of our posts, yet we never seem to tire of them.  Maybe our readers, especially interlopers from the other side of the v., tire of reading about them, but we can often find a wrinkle in a case that merits our huzzahs or inspires a rant.  Today’s case falls into the praiseworthy category, as the court dismissed a complaint predicated on violations of the FDCA in spite of sympathetic allegations that might have carried the day with some other courts. Markland v. Insys Therapeutics, Inc., — F. Supp. 3d –, 2017 WL 4102300 (M.D. Fla. Sept. 15, 2017), involved the alleged death of a patient as a result of respiratory distress from the defendant’s sublingual spray prescription painkiller drug, which she had started the day before.  Rather than offer the typical product liability claims under Florida law, perhaps because the labeling had extensive warnings on respiratory distress, plaintiff asserted only a claim for negligent marketing.  Calling it “negligent marketing” does not really identify what duty was allegedly breached, whether state law recognizes a claim for such a breach, and such a claim would be preempted.  The allegedly actionable conduct in Markland was promoting the drug for off-label use, like the chronic back pain of plaintiff’s decedent, as opposed to the approved indication for breakthrough pain with cancer.  While we do not know the merits, there were many allegations about off-label promotion, which seem to tie to the conduct at issue in well-publicized federal and state investigations.

Defendants moved to dismiss on various grounds, the most relevant of which (for our purposes) were that there was no claim under Florida law for this conduct and it would be impliedly preempted under Buckman anyway.  These had been hot topics recently in some Florida state and federal cases we have discussed, like Mink (here and here) and Wolicki-Gables, but those dealt with PMA devices and the additional issue of express preemption.  Here, with a prescription drug marketed under an NDA approval, there is no express preemption to navigate, but the plaintiff still had to walk a narrow path to state a claim that would not be impliedly preempted.  As we have said before, we think the appropriate order of analysis here would be the determine if there was a cognizable state law claim asserted and then determine if it was preempted, but the Markland court did not separate out its analysis.  It also did not weigh in on whether the allegations here were of truthful off-label promotion that might implicate First Amendment protection.  Instead, it assumed that the off-label promotion alleged violated the FDCA’s prohibition on misbranding.  2017 WL 4102300, *6 & n.4.

The court could take this approach because the plaintiff’s claim was so squarely focused on alleged violations of the FDCA.  Since Buckman, plaintiffs tend to be a bit cagier in making it look like their claims were not predicated on violations of the FDCA or fraud on the FDA.  The Markland plaintiff, however, labeled the defendant’s alleged conduct as violating the FDCA, “federal law,” and “requirements imposed by the FDA regarding the condition that this drug should be utilized to treat cancer patients with breakthrough cancer pain.” Id. at *9.  “Hence, [the claim], while framed in the language of negligence, appears to derive from [defendant’s] alleged off-label promotion of [the drug]” and “the very concepts of off-label use and off-label marketing are born out of the FDCA.” Id.   This was well phrased, as was the later statement that “it is only because of the existence of the FDCA’s restrictions on off-label marketing that Mr. Markland claims [defendant’s] actions were improper or otherwise violated a duty.” Id.

This is the recipe for implied preemption under Buckman.  It also means there is no negligence claim under Florida law, “which bars plaintiffs from using state negligence actions to seek recovery for FDCA violations.” Id. at 10 (citing negligence per se cases).  Of course, Buckman recognized that the FDCA does not provide for a private right of action, and preempts claims with FDCA violations as “critical element[s],” which should prevent such piggybacking.  So, plaintiff’s case was done and could not be revived by amending the complaint.  In other words, there was no need for a second and third strike before judgment could be entered.  This was so despite the Court’s expression of compassion:

The Court does not question for a moment the grievous nature of Carolyn Markland’s death, nor the deep sadness Mr. Markland must face on a daily basis as a result of his wife’s untimely passing. Nonetheless, the Court must act within the bounds of the law.

Id. at *11.  This a good lesson, especially for courts sitting in diversity, that the law should not be expanded to allow for recovery by sympathetic who cannot make their case under accepted tort theories.

 

Normally, when we think of decisions relating to medical monitoring, the issue is whether a state will recognize medical monitoring for uninjured people as a separate claim or relief that can be sought under an existing theory of recovery.  Just last month, we noted that it looked like the issue had been largely resolved against allowing such claims or relief.  Sometimes, the issue is whether the plaintiff sufficiently pleads the elements of a medical monitoring claim in a jurisdiction that recognizes it.   Today’s case involves a different consideration of medical monitoring, looking at whether the requirements of Fed. R. Civ. 23 are met and a medical monitoring class can be certified.  There is quite a bit to Barraza v. C.R. Bard Inc., No. CV16-01374-PHX-DGC, 2017 WL 3976720 (D. Ariz. Sept. 11, 2017), and we are only going to focus we find most interesting.

Barraza comes from an IVC Filter MDL, which has been the source of some other decisions that drew our interest, and ultimately focused on the question of whether class certification was appropriate for eleven separate classes, each for the residents of a state that has recognized medical monitoring and had a resident proposed class representative.  Each class sought medical monitoring for people with one of seven of the defendant’s IVC filter devices in-place (after being implanted at any time) who had not brought a case alleging personal injury.  In other words, these were to be classes of uninjured plaintiffs with on-going use of the allegedly defective devices.  (Note that the proposed class definition does not expressly exclude all patients claiming current complications, but the court addresses the case as though it presents a “no injury” class, so we will too.)  Keep in mind that a common reason for rejecting medical monitoring has been that the tort system is predicating on an actual injury, giving rise to accrual of claims, damages that can be determined by somewhat predictable rules, a duty to mitigate, etc., and people who have not have an actual injury do not fit well within the existing tort system.  Someone with an actual injury, however, may be entitled to compensation for on-going medical care to minimize the progression or sequelae of the injury.  With that in mind, we turn to the evaluation of whether individual considerations or common issues predominate in trying to decide the elements of medical monitoring—as identified by the plaintiffs based on an amalgam of the law of the eleven states.

We will focus on the elements that mattered to the outcome.  While plaintiffs argued that negligent design and failure to warn could be decided on common evidence, the court disagreed.  The seven devices were designed and launched over a more than ten year period and exhibited different design features, manufacturing specifications, and testing.  Similarly, the labeling for the devices differed depending on the date and product, but seemed to address the risks that plaintiffs claim required monitoring.  “Trial of a single class representative’s claim would not suffice because the representative would have received a different filter with different warnings than many members of the class.”  Similarly, the application of affirmative defenses like assumption of the risk and contributory negligence would also turn on individual evidence about what the plaintiff and her doctor knew and did.  The court noted how some of the named plaintiffs—putative class representatives—had ignored recommendations for medical follow up and removal of the device they claim subjects them to an increased risk of harm requiring monitoring.  Thus, individual considerations in evaluating liability predominated and “the classes cannot be certified simply because Plaintiffs allegedly face a common risk and need medical monitoring.”

The related issues of whether the proposed monitoring was necessary and different from the treatment the plaintiffs would otherwise receive also turned on individual considerations.

Here, the amount of monitoring a class member would require in a normal course of her treatment and illness, without the monitoring sought in this case, is an individualized inquiry into the medical needs and ongoing course of treatment for each class member.

For instance, some named plaintiffs were already undergoing monitoring of their own doctor’s devising with different levels of compliance.

Even what law would apply to classes defined by the state of residency involved individual considerations as the state where each plaintiff’s implant surgery occurred, the state where the injury occurred, and the state where the defendant designed the products and drafted labeling could affect the law that would apply.

Put it all together and plaintiff did not come particularly close to satisfying the predominance requirement and class certification under Fed. R. Civ. P. 23(a).  (The plaintiffs also tried for certification of a 23(b)(2) class, but that was pretty much a non-starter as the relief sought—paying for monitoring—is not injunctive.)  Some of the result here is likely due to the plaintiffs’ insistence on broad classes and the selection of putative class representatives with warts, but Barraza also illustrates how class treatment of medical monitoring claims should be a long shot even when state law allows monitoring for uninjured people.

The cases we typically write about focus on events after medical products have made their way through the applicable regulatory process and been marketed in the United States.  The marketed product is used, people claim to have been injured by it, and they sue the manufacturer under various product liability theories.  Often, in the course of seeking lots of discovery about the product and manufacturer, the plaintiffs in such cases will seek documents relating to foreign regulatory submissions and interaction.  While there might be some relevance of such documents in some cases, getting and producing them tends a pretty big pain for reasons like privacy laws and translation.  Plus, there is plenty of room for confusion given different regulatory standards in the United States and elsewhere around the world.  Under the old version of Fed. R. Civ. P. 26, however, some degree of burdensome foreign discovery was pretty standard in large drug or device litigation.  We have seen some signs (like here and here) that the new version of Rule 26 will be less friendly to plaintiffs seeking foreign fishing expeditions.

Schueneman v. Arena Pharms., Inc., No. 10cv1959-CAB (BLM), 2017 WL 3118738 (S.D. Cal. July 21, 2017) (Magistrate order), aff’d 2017 WL 3587961 (S.D. Cal. Aug. 21, 2017), is a securities case based on the allegation that the defendant pharmaceutical company defrauded investors by its representations about the likelihood that its new weight loss drug would get approved by FDA.  The focus of the allegation is about tumors in a rat study that ultimately led to FDA issuing a non-approvable letter in October 2010.  (As an aside, the drug was developed with a serotonin receptor subtype mechanism of action that was supposed to avoid heart valve issues seen with other weight loss drugs, litigation over which included allegations about a signal from tumors in a rat study—although the main plaintiff expert supporting that allegation did not get past Daubert.)  The non-approval was made public at that time, although the plaintiff investors claim that they continue to have been misled for another three months.  (The decisions do not mention either, but there was a fair amount of press on the rat study in September 2010 and the drug was ultimately approved, after more research, in June 2012.)  In March 2012, the company submitted an application to the European Medicines Agency to try to market the drug in Europe.  After some back and forth, the company withdrew that application in May 2013.  (We are feeling parenthetical, so we will note that the withdrawal of the European application was also the subject of U.S. press attention, spurred on by Public Citizen, whose actions often seem coordinated with litigation efforts.)  The plaintiffs sought every document conceivably related to the European application and its withdrawal, even though the allegations in the case were about earlier representations about FDA approval.  Defendants balked and the fight was joined.

The requirement that discovery be “relevant to any party’s claim or defense and proportional to the needs of the case” allowed for a more rational analysis that we have seen in the past on such disputes.  Noting the “different standards utilized by the EMA and the FDA and the different time frames of the statements,” the court required plaintiffs to “explain how documents presented to the EMA, especially those unrelated to the Rat Study, are relevant to the issue in this case of whether Defendants ‘intentionally withheld information material to the market’s assessment of whether and when the FDA would likely approve [L]ocaserin.’”  2017 WL 3118738, *4.  Plaintiffs could not do that.  They also could not establish proportionality, a burden now assigned to the proponent of the discovery (although the posture here was a motion to compel rather than a motion for protective order).  Defendants had already produced or were producing lots of documents and estimated that a production of the documents at issue would take weeks and cost $80,000.  In the past, such a cost estimate for even marginal documents sought from a drug company would probably not be seen as an undue expense sufficient to cut off the discovery.  Now, in the age of proportionality, it is part of the puzzle:

Given the extremely broad discovery requests, including the fact that the requests are not limited to the Rat Study, the minimal relevance of the statements made after the class period in an application in the European Union involving drug standards that are different from those in the United States, and the large number of potentially responsive documents, the Court finds that requests 34-37 are not proportional to the needs of the case.

Id. Similar reasoning could be applied to disallow quite a few requests in drug and device product liability cases, particularly those on foreign regulatory discovery.

The defendant also forced plaintiffs to reveal the identity and contact information for six confidential informants relied upon for allegations in the complaint.  Plaintiffs claimed attorney work product protection, which the magistrate rejected and the district court found to be so frivolous as to merit the award of expenses under Fed. R. Civ. P. 37(a)(5). Id. at **6-7; 2017 WL 3587961, *2.  We cannot recall seeing a product liability plaintiff resist discovery on the basis for allegations in a complaint on a claim of work product protection, so we will not dwell on this holding—but the same reasoning would apply.  Rather than just sound like homers and cheer when a drug company wins discovery motions, we will say that we like it when discovery is ordered on the actual allegations in the complaint and limited on issues outside of the allegations in the complaint.

 

We have offered our view that cases seeking to impose liability based on well-known risks found with an entire class of prescription medications tend to be weak.  We think design defect claims usually are clearly preempted in this context and warnings claims will often be preempted too, even with Levine’s high “clear evidence” hurdle.  Cases about thrombotic risks with hormonal contraceptives have featured prominently in such posts, like this opus, precisely because design is not the issue and FDA has long been intimately involved with labeling of these products.

Another obvious fertile ground for preemption has been with gastrointestinal bleeding with anticoagulants, something of the therapeutic flip side to the risk of thrombosis.  First, it is a well-known issue.  Our quick PubMed searches easily got us to articles about this from the 1950s.  Second, this risk has been described in drug labels for a long time.  We easily found this as the first warning in prescription labels as early as 1998, although we suspect they had been around for a few decades by that point.  Third, this risk has been seen with every anticoagulant since there have been anticoagulants.  We have no doubt that any anticoagulant drug coming to market gets a thorough review of its bleeding risk and its labeling about that risk by FDA.  This surely includes attention to any differences in the labeling of the different anticoagulants and whether any post-approval studies or adverse events merit changes.  These facts should make it hard to articulate, let alone prove, a design defect claim that gets by Bartlett or a warning claim that gets by Levine, unless Buckman gets ignored.

We say “should,” but, in all fairness, it certainly depends on where the case is and who is deciding it.  Even in the nascent era of drug and device product liability litigation where cases should pretty much be in federal court unless they are in state court in the defendant’s true home state, the court can be all but determinative of the decisions on litigation-altering issues.  The selection of court can, in turn, depend on the selection of the MDL’s home in litigations where the lawyer advertising drums up enough cases to get the JPML’s attention.  We were going to contrast cases decided by different MDL courts overseeing product liability litigation over the bleeding risk of relatively new prescription anticoagulants.  Instead, we will be discussing one decision addressing allegations we think are pretty typical of what is getting offered up elsewhere and our dear readers can draw their own conclusions.

Fortner v. Bristol-Myers Squibb Co., No. 17cv1562 (DLC), MDL No. 2754, 2017 U.S. Dist. LEXIS 117030 (S.D.N.Y. July 26, 2017), comes out of the Eliquis MDL.  Based on the JPML’s statistics, when decided, there were 23 pending cases out of a total of 69 ever-filed cases in this relatively young MDL.  The drug was approved in 2012 with extensive warnings about the risk of bleeding.  Plaintiffs in the MDL offered various allegations about how the drug was defectively designed because it had a clotting risk, was not accompanied by a drug-specific clotting test, was not accompanied by an “antidote,” and was to be taken twice a day.  These same criticisms were offered as warnings claims, but there were no allegations that the manufacturer had received post-approval safety information triggering some alleged duty to try to change any aspect of the label through the CBE process.  The manufacturers challenged whether these allegations stated any state law claim that was not preempted and, before there was even an MDL established, dismissed a number of cases without prejudice in Utts I, which we discussed here.  After the MDL was established, the plaintiffs got another shot with amended complaints and still came up short in Utts II, this time with prejudice.  The court, in an exercise of magnanimity, invited the remaining plaintiffs to see if they could come up with complaints that stated a non-preempted claim.  That is how we get to Fortner, who alleged a variety of claims under Tennessee law based on the same allegations about the drug, manufacturers, and FDA that most of the remaining plaintiffs apparently offered.

As is often the case with pleading around statutes of limitation—complaints with dates for everything but when plaintiff’s alleged injury occurred—it looks like the fourth attempt at a complaint was modified to be vague, repeating allegations “in less detail and without identifying or appending the specific studies from which these allegations are drawn.” Id. at *7.  The Fortner court saw through this “pleading tactic” of “masking the basis for her claim”:  The complaint’s “claims do not become more plausible simply because the plaintiff has omitted from the FAC the sources upon which her conclusory factual allegations are based.” Id. at **7-8.  Well stated and clearly correct, but many courts let uncertainty work to the plaintiff’s advantage in this posture, despite TwIqbal’s requirement of factual allegations that plausibly state a claim.

The critical aspect of Fortner’s approach is that the court required the plaintiff to plead a warning claim based on “sufficient factual content to support a plausible inference that there exists newly acquired information such that the defendants could unilaterally have changed the Eliquis label to include additional warnings.” Id. at *8.  This, in turn, flowed from the court’s prior decisions holding that “post-approval failure to warn claims are preempted unless the plaintiff can plausibly allege that there existed ‘newly acquired information’ such that, pursuant to the Changes Being Effected (‘CBE’) regulation, the defendants could independently have updated the Eliquis label to include such warnings.” Id. at *5.  There is no such thing as a pre-approval warning claim—absent an allegation that the launch label resulted from fraud-on-the-FDA that side-stepped Buckman—so this is a pretty good statement of what a non-preempted prescription drug warnings claim should allege.

By contrast, under the court’s prior analysis, there is no such thing as a non-preempted post-approval design defect claim because “FDA regulations prohibit a change of the type implicated by the claim.” Id. Here, the first urged defect was twice daily dosing—which is a design issue if the plaintiff alleges the product should have been designed to deliver the effective dose by taking it once a day, for instance, and something that clearly cannot be changed without a new NDA.  The other urged defects are things we see as more labeling than design issues—lack of a drug-specific clotting test or an “antidote” to the drug that could be recommended or sold with the drug.  Even if such a test or antidote existed, it could not be sold with the drug based on anything the manufacturer could have done independent of FDA action.  In reaffirming its prior decision on the preemption of pre-approval design defect claims, the Fortner court noted that Yates was the only appellate court to address the issue and no binding authority disagrees with its analysis.

Based on a trio of preemption rulings at the pleading stage, it looks like the Eliquis MDL will be short lived.  That is not always the case with MDL proceedings based on dubious claims, where the burden of one-sided discovery and the weight of the docket tend to dictate the result more than anything approaching the merits.  In terms of issues that seem as obvious to us as preemption of pre-market prescription drug design defect—we note that “duh” and “no duh” mean the same thing, like “regardless” and “irregardless” or “flammable” and “inflammable”—it will help to have more appellate courts follow Yates.

 

Our day job has been keeping us busy, so busy with depositions, motions, delayed flights, and assorted drama that we have not posted in more than a month.  After such a long layoff, we had hoped to return with a vengeance, a “the North remembers” sort of vengeance.  Instead, we get fair market value, Current Procedural Terminology codes, and a Daubert challenge to a defense expert accountant.  Eh.  The decision is in a False Claims Act case about Medicare fraud and kickbacks, about which we care and have even spent some billable time.  But the reason we are writing about U.S. ex rel Lutz v. Berkeley Heartlab, Inc., No. 9:14-cv-00230-RMG, 2017 U.S. Dist. LEXIS 107481 (D.S.C. July 12, 2017), is because the rejection of opinions offered by the expert in this case undercuts the damages claims we see in other cases.

Lutz involves allegations that the defendants overbilled Medicare in connection with blood tests.  One part of the alleged scheme was that the defendant labs charged high processing and handling fees.  (We will ignore the issues of FCA materiality and causation that seem to flow from the absence of evidence of Medicare payment on processing and handling charges.) From what we can tell, the defendant’s expert opined that these fees were consistent with the fair market value—a defined term under the, aptly named, Stark Act—and did not represent kickbacks to referring physicians.  He offered a “charge-based methodology” to conclude that the fees were consistent with fair market value because physicians allegedly expected to recover for time spent on processing handling, not just for the actual blood draw.

“[I]t is no secret that the sticker price of services listed in physician bills and hospital chargemasters are totally unmoored from the reality of arm’s-length transactions actually taking place in the marketplace.” Id. at *12.  We think you can see where we are going with this.  In product liability cases, economic damages related to medical care, past and future, are typically based on amounts incurred.  The collateral source rule is often used offensively to create a windfall of the difference between the amount on the bills and the portion of that amount anybody ever paid or will pay.  We have called this “phantom damages.”  In FCA, RICO, and consumer protection cases, for instance, the plaintiff’s theory of price gouging or other impropriety sometimes relies on the list price of a drug or the amounts appearing on bills, regardless of what typically gets paid.  Although in the context of granting a challenge to an expert for healthcare providers that wanted the charges to matter, the Lutz decision supports the more typical defense position.

Actual transactions, not just bills, are a better indicator of fair market value because “physicians deliberately set their fees higher than the amount they either expect to receive or do in fact receive to ensure that no money is left on the table.” Id. at *16.  The court identified a number of decisions, from various types of cases, recognizing that “physicians’ billed charges do not necessarily reflect the market value of physician services.” Id. at **17-18 (citing cases).  So, the next time you have a case where the plaintiff says her damages include the total of all her medical bills and the only support she needs is her expert mouthing that “the charges were all reasonable and necessary,” think about whether Lutz and the cases it cites help you push back.  That was our return:  brief, a little bloody, and with the promise of more to come.

Earlier this week, we posted on the Ninth Circuit’s conversion of the Daubert’s gate (that the trial court should keep) into more of a swinging saloon door.  A week before the Ninth Circuit ruled that a trial court had erred in excluding unreliable causation testimony (and granting summary judgment as a result), the Third Circuit had affirmed a trial court’s exclusion of unreliable causation testimony (and grant of summary judgment as a result).  Even though we are discussing In re Zoloft (Sertraline Hydrochloride) Prods. Liab. Litig., __ F.3d __, 2017 WL 2385279 (3d Cir. 2017), second, it really is a bigger deal because it reaffirmed the end of an entire MDL.

We followed the district court’s Daubert rulings on the epidemiology and mechanism experts offered for all the plaintiffs.  We watched in amazement as the plaintiffs got to try again and still could not offer reliable expert testimony on general causation.   With our typical restraint, we applauded the court’s subsequent decision that no plaintiff could make out a case for general causation between maternal use of the drug and the cardiac birth defects claimed without the excluded experts and that was fatal to their claims.  We found that the plaintiffs, maybe because of the sympathy associated with their claimed injuries, got plenty of leeway before the court determined that there was simply no there (i.e., good science) there.  (Along the way, we saw that Pennsylvania and West Virginia state courts came to similar conclusions.)

The appeal to the Third Circuit focused on whether the biostatistician offered as a back-up expert on epidemiology was properly excluded, with plaintiffs conceding that they should have lost if he was.  Plaintiffs’ central contention was that the district court created a standard that requires general causation opinions to be “supported by replicated observational studies reporting a statistically significant association between the drug and the adverse effect.”  We think that standard, similar to Havner and Daubert II, is a fine standard, but the district court did not create or apply such a standard in knocking out the biostatistician.  Likewise, the Third Circuit declined to “state a bright-line rule” that “statistical significance is necessary to prove causality.”  (We think it is, because the Bradford Hill Criteria, which the biostatistician purported to apply, starts with an association demonstrated through epidemiologic studies.  We will try to resist arguing for the tighter standard given the result.)  The district court considered the lack of multiple statistically significant studies supporting an association to be contrary to what teratologists generally require and thus relevant to whether an opinion without such support was unreliable.  A flexible approach to evaluating the reliability of a general causation opinion was fine with the Third Circuit and its reading of the Bradford Hill Criteria.  (There is flexibility, but only when there is an association from epidemiologic studies as a predicate.  OK, we will have to try harder.)

The Third Circuit “accept[ed] that the Bradford Hill and weight of the evidence analyses are generally reliable.  We also assume that the ‘techniques’ used to implement the analysis (here, meta-analysis, trend analysis, and reanalysis) are themselves reliable.”  That assumption is dicta—which is a good thing—because the court concluded that the biostatistician did not reliably apply the methodology or techniques that he claimed to be applying.  First, he gave lip service to analyzing “multiple positive, insignificant results,” but he really just eyeballed trends.  Second, his trend analysis was based on cherry picking and inconsistent application of basic statistics principles.  Third, his meta-analysis was also result-driven, as he could not justify why he included some studies and excluded others.  Fourth, his reanalysis was done for no reason but to conclude that a published study reporting no association should have found one.  Altogether, “the fact that Dr. Jewell applied these techniques inconsistently, without explanation, to different subsets of the body of evidence raises real issues of reliability.  Conclusions drawn from such unreliable application are themselves questionable.”

The court probably could have stopped there.  It went on to detail how the biostatistician’s purported application of Bradford Hill was riddled with errors that he could not explain.  This was more than enough to conclude that the district court had not abused its discretion in excluding the expert.

Along the way, however, it noted that it may be possible to have a reliable reanalysis that draws a different conclusion than the original published study and that an expert can make unsupported assumptions in connection with doing an “informational” reanalysis.  It offered that “[t]hese inquiries are more appropriately left to the jury.”  We disagree and think the broader context has to be considered.  A plaintiff’s expert offered on the epidemiologic evidence who cannot offer a reliable opinion that there is an association between the exposure and the type of injury the plaintiff claims, let alone that there is a causal relationship, should not be talking to the jury about anything.  A plaintiff’s expert offered on the epidemiologic evidence who can offer a reliable opinion that there is a causal relationship between the exposure and the type of injury the plaintiff claims can be allowed to discuss the various analyses she did to form that opinion.  And the defense can cross-examine her on whether some of her analysis was result-driven for-litigation drivel or based on unsupported assumptions.  A jury can hear that sort of back and forth and decide what weight to give to the expert’s testimony on general causation.  However, no trial court should abrogate its gatekeeping role and let juries hear about reanalysis of published studies unless plaintiffs have reliable evidence of general causation in the first place.  I guess we prefer the opinions of the district court, which took its gatekeeping seriously, even if it let plaintiffs take a few shots at entry.

 

Parties often file motions in limine on fairly case-specific issues, building on the history of discovery and motions practice in the case.  Applying a ruling on in limines from one case to another can be a dicey proposition as potentially significant differences in the facts, law, claims and defenses asserted, and other rulings can usually be identified.  Plus, many pre-trial decisions on evidence do not last once doors get opening and evidence can be cumulative, among other reasons why judges’ minds change.  Still, we do posts on rulings on motions in limine that we guess might have some relevance to other cases our readers have.  When we do, we can be hamstrung by the limited information in these opinions on the facts, allegations, and other rulings, such as rulings on motions for summary judgment that would typically be rendered before the in limines are decided.

In In re Depakote, No. 15-CV-702-NJR-SCW, 2017 WL 2126837, *2 (S.D. Ill. May 16, 2017), we have rulings on a grab bag of motions in limine after the court issued partial summary judgment for the drug manufacturing defendant based on “preemption of label changes related to development delay” after fetal exposure of the medication.  If this summary judgment ruling sounds familiar, then you might need a hobby.  You also might have read any number of our posts on the Rheinfrank case and the ultimate affirmance of its preemption decision and defense verdict by the Sixth Circuit.  Like here, here and here.  As we said of the appellate decision, “The court held, ‘[g]iven, then, that as of 2008 the FDA did not believe the state of the data supported a developmental delay warning, it stands to reason that as of 2003, with even less data to go on, the FDA would similarly have rejected a developmental delay warning.’ Thus, Rheinfrank joins those courts that have drawn a preemptive line barring all plaintiffs who used a drug prior to an FDA insufficient evidence decision concerning the risk at issue.”  We also wondered how the case got to a jury given trial court’s preemption decision and the requirements of Ohio law.  We have similar wonderment at the instant case, which involved 2006 fetal exposure.  The decision is sparse on facts, but it is hard to imagine a viable product liability claim for the plaintiff’s injury when plaintiff cannot prove that the label in place when the drug was prescribed to the mother should have said something different about the risk of the injury in question.  Some sort of warnings claim seems to be pending, along with a claim for punitive damages.

With that background, we turn to the subset of the 28 total motions in limine that we think matter most.  First, consistent with the preemption ruling and the applicable regulation, plaintiff could not preclude the manufacturer from explaining that a proposed change to the label through the CBE regulation is still subject to FDA’s decision to accept, reject, or modify the proposed change.  2017 WL 2126837, *2.  What the CBE regulation has to do with plaintiff’s surviving warnings claim or whether it could have used (e.g., based on new evidence of safety) is unclear. Second, plaintiff’s red herring argument about whether FDA is required to do its own post-marketing studies did not preclude the defendant from presenting evidence that it followed labeling regulations.  Third, while the court allowed some speculative evidence from the plaintiff’s mother in connection with the inquiry on proximate cause for failure to warn—which should not happen unless a different developmental delay warning has been articulated—this opened the door to some amount of evidence on what she knew and how she behaved. Id. at **2-3.

As to the defendant’s motions, the post-conception labeling and regulatory communications were excluded and defendant agreed not to raise pre-conception discussions with FDA about labeling changes that were not yet approved. Id. at *3.  (With the limited information presented in the opinion, we cannot say if there was evidence that FDA rejected or discouraged whatever labeling change plaintiff was allowed to urge in the case.)  The court also excluded a 2009 FDA alert about the risk of birth defects with the drug, although plaintiff was allowed to discuss any pre-conception studies that went into the alert. Id. at *4.  Limiting warnings evidence to what existed before the prescription at issue makes sense, but it also makes sense that there needs to be a claim based on what the warnings should have said instead at that point.  Along those lines, plaintiff was not allowed to argue that the drug should have been contraindicated for use in pregnancy because plaintiff offered no expert who disclosed such an opinion. Id. (And such an opinion would have had some preemption problems, we think.)  Plaintiff was allowed, however, to offer evidence about foreign labels for the drug in place before plaintiff’s conception, holding that the manufacturer’s knowledge of these labels was relevant to its duty to warn. Id. at **7-8.  Breaking somewhat from its previously firm line on the irrelevance of the post-conception evidence to the warnings claim, the court did not foreclose the possibility that some post-conception marketing materials could be relevant if “they contain pre-2006 facts.” Id. at *6.  Again, we would think the pre-2006 facts would need to relate to whatever about the 2006 warnings for developmental delay that plaintiff was allowed to claim should have been different.

There was one last ruling that bears some discussion.  The defendant had pleaded guilty to a misdemeanor and paid a very, very large fine related to allegations of off-label promotion for use of the drug for schizophrenia and elderly dementia, which were not labeled indications. Id. at *9.  The plaintiff’s mother did not use the drug for these conditions and there was no evidence to “connect these activities with the 2006 Depakote teratogenicity warning.” Id. So, there was far too much prejudice compared to the probative value to let the jury hear about the plea or fines.  However, plaintiff was allowed to “introduce evidence regarding the off-label marketing and sales efforts . . . regarding bipolar disorder,” which was an indication added during 2006 (based on our quick look). Id.  The court saw this evidence as supporting plaintiff’s claim for punitive damages.  It seems to us that a plaintiff’s punitive damages evidence is not relevant unless it tends to show that the underlying conduct giving rise to liability was done with the requisite intent.  It does not sound like there is any connection between an alleged, non-preempted issue with the 2006 warnings for developmental delay and any off-label promotion issue, but maybe that link was just not spelled out in the opinion.

 

Charges of discovery abuse get thrown around frequently in product liability litigation.  We have not done a scientific survey, but we guess that such charges are levied against the manufacturer defendants more often than against individual plaintiffs.  For one thing, seeking burdensome discovery, and then discovery on discovery, has been in the product liability plaintiff game plan for a long time.  There also tends to be more discovery that a defendant could produce—and, therefore, be accused on wrongfully withholding—than a plaintiff could produce.  There is also the practical consideration that large manufacturers tend to have the financial wherewithal to pay fees when ordered and contingency plaintiffs do not—although the lawyers who front the money for those plaintiffs and make the decisions about how to proceed in discovery typically do.  While there are occasions where courts require plaintiffs and their lawyers to pay substantial defense costs because of bad conduct in discovery or in the litigation more broadly, an argument about how to calculate fees to be awarded for discovery abuse is something that we generally hope to avoid.  It is not quite up there with arguing about the maximum acceptable ratio of punitive to compensatory damages that can be awarded, but it still makes us a little uncomfortable.

The Supreme Court’s decision in Goodyear Tire & Rubber Co. v. Haeger, 581 U.S. __ (2017), slip op., involves a very large award of fees based on the district court’s conclusion that the manufacturer defendant in a product liability case had intentionally withheld important internal testing documents.  The plaintiffs did not learn about the documents until after they had settled, when a reference appeared in a newspaper article about another similar case.  Because the case had resolved, the late application to shift costs and fees appealed to the court’s inherent authority.  Using that authority, the court not only determined that the defendant had engaged in bad faith discovery for years, but that it should pay the plaintiff $2.7 million for all costs and fees since the initial “dishonest discovery response.” Slip op. at 3.  It specifically determined that egregious conduct by a party negates the typical requirement that fees be limited to those caused by the sanctionable conduct. Id. As a back-up in case the Court of Appeals reduced the award, the court determined that the costs and fees excluding what plaintiffs estimated they incurred in pursuing other defendants and in proving medical damages, would be $2 million. Id. at 4.  (We find the whole concept of the fees incurred in the context of a presumably contingency fee representation somewhat bizarre.  Did the plaintiffs’ lawyers actually charge more than $2.7 million in costs and fees to their clients when the proceeds of the settlement(s) were divided up?)  The Court of Appeals affirmed the full amount and the Supreme Court granted cert.

The Haeger Court started by distinguishing between sanctions that compensate and sanctions that punish.  The latter can only be awarded if the trial court provides the “procedural guarantees applicable in criminal cases, such as a ‘beyond a reasonable doubt’ standard of proof.” Id. at 6 (citation omitted).  (As an aside, we are not sure that each state requires such a standard of proof when punitive damages are offered, so maybe this Court would be strict in its evaluation of punitive damages.)  “When (as in this case) those criminal-type protections are missing, a court’s shifting of fees is limited to reimbursing the victim.” Id. Damages to reimburse must have been caused by the sanctionable conduct, not merely come after it started.

The court’s fundamental job is to determine whether a given legal fee—say, for taking a deposition or drafting a motion—would or would not have been incurred in the absence of the sanctioned conduct. The award is then the sum total of the fees that, except for the misbehavior, would not have occurred.

Id. at 7-8 (citation omitted).  The court has some leeway in making large sanction award as long as the touchstone is causation.  A plaintiff can be hit for all costs of a defending case initiated in “complete bad faith,” such as we have seen relatively recently. Id. at 8.  Sanctions can also be based on the court’s assessment of whether failure to disclose “evidence fatal to its position” affected the timing (but not amount) of settlement. Id. at 8-9.

In Haeger, the trial court did not apply a but-for causation test to its damages calculation, so it will have to do it over with the right standard (unless it determines that there was some sort of waiver).  While we do not know what the amount of the sanction will be on remand, we do have an inkling that the damages imposed for discovery misconduct will tend to be less if the Haeger standards are followed in other cases.  In a case where a fundamental lie by the plaintiff—claiming to have used the defendant’s product, claiming legal authority to initiate a suit, claiming no knowledge of the injury or its cause until shortly before bringing suit—caused a case to be brought or stick around, some bold judges can still impose significant sanctions following Haeger’s principles.

 

Earlier this week, we discussed how the presentation of the federal question of express preemption from the face of a complaint can lead to removal.  Part of why the defendant drug or device manufacturer may prefer federal court over state court is that the belief that the chances of winning on preemption are better in federal court.  On the other hand, we have described many instances where federal courts mess up their preemption analysis by presuming that state law imposes the duty that plaintiff claims does not conflict with FDA obligations or by extending state law in new directions to provide a basis for a parallel claim, Erie restraint notwithstanding.  It may be that state court judges are less likely to impose duties not recognized explicitly in higher court decisions.

Tibbe v. Ranbaxy, Inc., No. C-16o472, 2017 Ohio App. LEXIS 1139 (Ohio App. Mar. 29, 2017), is a case that stayed in state court despite the explicit claims that the defendants—the generic drug manufacturer and the non-diverse pharmacy defendant—violated the FDCA in various ways.  On its basic facts and history, the case had the hallmarks of a case pursued in disregard of controlling law.  A typical warnings claim that information in the generic drug label about the risk that allegedly befell plaintiff was insufficient should not fly post-Mensing.  A claim predicated on defrauding the FDA should not fly post-Buckman.  Claims against the pharmacy that it should be liable for nothing more than filling a prescription with the generic form of the particular antibiotic (presumably as required by plaintiff’s insurance) should not fly under Ohio law.  It should not have been enough to defeat a motion to dismiss for plaintiff to claim that discovery might help them determine if the generic drug’s label was different than the reference drug’s label.  After the Sixth Circuit’s decision in Fulgenzi v. Pliva, 711 F.3d 578 (6th Cir. 2013), discussed here, this is something that can and should be determined before suing.  Even though Fulgenzi held that there is exception to Mensing’s preemption of warnings claims for generic drugs where there has been a failure to update, the plaintiff there lost because the prescriber did not review the label.  So, you would think some pre-suit investigation into the labels of the reference drug and the generic drug and the prescriber’s practices should be done in determining if there is a good faith basis to plead a non-preempted claim.  Maybe that is just our silly defense-sided way of thinking.

Regardless, in Tibbe, the plaintiff got her discovery and the preemption issues were presented again on motion for summary judgment.  Despite plaintiff’s earlier protestations, the labels were actually the same during the relevant time periods, including the language as to the risk of the condition that plaintiff claimed to have developed from the medication, so Mensing applied and Fulgenzi did not.  2017 Ohio App. LEXIS 1139, *10.  The intermediate appellate court reviewed the grant of summary judgment de novo.  Even though the Sixth Circuit had carved out an exception for non-preempted generic drug warnings claims in Fulgenzi, later the same year it recognized in Strayhorn v. Wyeth Pharms., Inc., 737 F.3d 378, 391 (6th Cir. 2013), discussed here, that Mensing had broad application to “claims that the generic manufacturers failed to provide additional warnings beyond that which was required by federal law of the brand-name manufacturer,” no matter how the claims were couched. Id. at *18.  Faced with this law and the factual record on sameness, plaintiff came up with an argument that we do not recall seeing before.  She claimed that there was a “duty to warn consumers of the generic version of the drug that they cannot bring a state-law failure-to-warn claim when their prescriptions are filled with Ranbaxy’s generic minocycline and the labeling is that of the RLD.” Id. at *19.  In other words, the plaintiff claimed the manufacturer had to give legal advice—not just legal advice, but legal advice about Mensing that was contrary to the position plaintiff advocated.

This is where being in state court maybe helped the defendants.  The court did not have to engage in much of an analysis to see whether there was already a duty to do what the plaintiff wanted and did not even consider making up a new duty.  The duty to warn under the Ohio Product Liability Act related to “the risks associated with the product.” Id. at *20.  “There is no corresponding duty to warn a consumer of her legal rights or the prospective outcome of litigation should she decide to sue a drug manufacturer at a future point in time.  Thus, a claim based on that theory would not be available under Ohio law.” Id. at **20-21.  Any warnings claim based on actual Ohio law conflicted with federal law and was preempted.

If you have been following along for a while, then you have surely run across our posts making some combination of the following points:  1) design defect claims rarely make sense for a drug because changing the design in some material way will usually make it a different drug, 2) such design defect claims, if recognized by state law, will usually be preempted because FDA approval of a different drug cannot be assumed, and 3) courts really should analyze conflict preemption by first determining that there is an actual state law duty that has been asserted or supported (depending on the procedural posture).  One such post walked through why it took so long until a circuit court held that a design defect claim with a prescription drug was preempted.   That case, Yates, has been followed a number of times, including on motions to dismiss, but there are still some glitches.

The decision in Young v. Bristol-Myers Squibb Co., No. 4:16-CV-00108-DMB-JMV, 2017 WL 706320 (D. Miss. Feb. 22, 2017), counts as a glitch on the preemption front even though the court dismissed (without prejudice) the design defect claim and eight of the nine other claims asserted.  The plaintiff claimed to have suffered ketoacidosis and renal failure from taking a prescription diabetes medication right around the time FDA issued a Public Health Advisory about the risk of ketoacidosis for the class of medications, SGLT-2 inhibitors, to which it belonged.  Several months later, the drug’s label was revised to include warnings about ketoacidosis and urosepsis, a blood infection stemming from a urinary tract infection.  Plaintiff claimed that the inherent design of the drug, like all SGLT-2 inhibitors, created a risk of ketoacidosis.  When plaintiff sued, she asserted a wide range of claims and defendants moved to dismiss on various grounds.  We will address only some of them.

Part of our point here is that the order can matter.  We do not have the briefs, so all we can go off of here is the opinion.  After the preliminary issue of whether common law claims are subsumed by the Mississippi Product Liability Act—the four here were—the court starts off the meat of the analysis with this:  “The defendants argue that Young’s claim for defective design must fail because Young has failed to plead a feasible design alternative and because federal law preempts the design defect claims.” Id. at *5. So, what gets analyzed first? Preemption. (Remember, federal courts are supposed to try to resolve disputes on nonconstitutional grounds if they can.) In so doing, the court has to hold out as unresolved whether Mississippi law imposes the very duties that might create the conflict leading to preemption. As the court recognized at the end of its, to us, flawed preemption analysis:

If there is no state law duty, the state law cause of action must certainly fail but there can be no conflict so as to justify preemption. Put differently, the absence of a state law duty is fatal to a claim but not under the doctrine of conflict preemption.

Id. at *8.  This logic suggests that the court needs to decide first whether there is a state law duty to do what the plaintiff urges was necessary.  Because the court never determined that there was such a duty, the whole discussion of preemption seems like a bunch of dictum to us.

Continue Reading Another Court Tackles Prescription Drug Design Defect