And it’s the start of another season of too much eating, drinking, shopping, and socializing. Sometimes it feels like a year’s worth cramped into a little over a month. And, for some of us (OK, me), we’re entering this week still a little groggy and foggy. So, we’ll admit to looking for something fairly straightforward and on the shorter side to post about as we clear our heads and shake off the haze. That’s not to say Mitchell v. Boehringer Ingelheim Pharma, Inc., 2017 U.S. Dist. LEXIS 192498 (W.D. Tenn. Nov. 21, 2017) isn’t noteworthy, it’s just not overly complicated. Sometimes that’s what you need to get the ball rolling again.

Plaintiff in Mitchell alleged that she suffered diabetic ketoacidosis (“DKA”) as a result of using the prescription drug Jardiance to treat her type-2 diabetes. Id. at *3. The drug, one of a class of drugs known as SGLT-2 inhibitors, was approved by the FDA in August 2014 and plaintiff began using it in February 2015. In May 2015, the FDA issued a safety alert warning about the risk of DKA with this class of drugs. The alert was based on a review of adverse event reports collected from March 2013 to June 2014 – before Jardiance was approved. Id. at *3-4.

Products liability claims in Tennessee are governed by the Tennessee Products Liability Act. Under the TPLA, drug manufacturers have a duty to minimize the risk or dangers from their products and they can discharge that duty by providing adequate warnings and instructions. Id. at *7. Plaintiff in Mitchell alleged defendant failed to satisfy that duty because it did not warn about DKA at the time the drug was approved and did not amend the warning via the FDA’s Changes Being Effected (“CBE”) process at any time before plaintiff was injured. Id. at *4.

We see CBE and we immediately start looking for a preemption discussion. Sure enough, there is one. As discussed here many times before, the CBE process allows a manufacturer to add or strengthen a warning based on “newly acquired information” without first obtaining FDA approval. Id. at *9-10. Where a CBE label change is an option for a manufacturer, the Supreme Court has held that failure to warn claims are not preempted. See Wyeth v. Levine, 555 U.S. 555 (2009).

But what about when a CBE is not an option? In Mitchell, the court broke the plaintiff’s failure to warn claim into two pieces – a challenge to the original labeling and a challenge to warning as it was at the time plaintiff used the drug. The court found the former preempted because defendant could not have changed the original FDA approved labeling. The data plaintiff relied on to support her allegation on this claim was pre-launch data. In other words, data that was known to the FDA at the time they approved the drug’s warnings – not newly acquired. “[T]he FDA is the exclusive judge of safety and efficacy based on information available at the commencement of marketing.” Id. at *11 (citing  In re Celexa & Lexapro Mktg. & Sales Practices Litig., 779 F.3d 34, 41 (1st Cir. 2015).

Similarly, plaintiff cannot base her failure to warn claim challenging defendant’s failure to amend the warning after the drug was introduced on the market on information that existed pre-FDA approval. Id. at *13. So plaintiff cannot rely on the adverse events collected before June 2014. Plaintiff’s allegations must be based on information that was not known by the FDA at the time of approval. In her complaint, plaintiff alleged that additional DKA adverse events were reported in 2015. Defendant, citing federal regulations, argued that while later in time, the additional reports did not provide any “new” information that would support a CBE. The court determined that was an issue better suited to a motion for summary judgment and that plaintiff had done enough to withstand a motion to dismiss. Id. at *15. Not a win yet, but the door is certainly still open.

The court made a few other rulings as well. Based on the learned intermediary doctrine, plaintiff cannot bring a failure to warn claim premised on failing to warn plaintiff directly. She is limited to a claim for failure to warn her prescriber. Id. at *17-18. And plaintiff must plead warning inadequacies with specificity. The court determined her allegations as to DKA were sufficient. But, plaintiff failed to state a claim to the extent she was looking to recover for “other related health complications.” Id. at *22. Plaintiff’s complaint neither specified the alleged complications nor how the drug’s warnings were inadequate as to those complications. So, as to unspecified complications, the failure to warn claim was dismissed.

In the end, plaintiff is left with a narrow failure to warn claim and the hurdle of proving proximate cause.

Can you recall what you were doing back in March of this year? To be more precise the day before St. Patty’s Day and the day after the Ides. No? Well, apparently the defendants in the Risperdal and Invega Products Liability Cases pending in California state court were celebrating but they forgot to invite us to the party. We just learned about the very nice preemption decision entered by the trial court in that litigation. Since it’s never too late to celebrate a preemption victory, here are the highlights.

Before we get into the case specifics, the court’s general preemption analysis merits a minute of our time. ‘[I]f the tort plaintiff’s failure to warn theory was already tested by FDA action or inaction or would have required the use of a label on a prescription drug which the FDA would have prohibited,” the claim is preempted. Risperdal and Invega Product Liability Cases, 2017 WL 4100102 at *5 (Cal. Super. Mar. 16, 2017). What remains for private plaintiffs to challenge in a tort setting, therefore, is the adequacy of the label or the reasonableness of the manufacturer in updating the label based on new or additional information not already considered by the FDA. Id.

 In discussing whether information was presented to the FDA for consideration, of course, fraud-on-the-FDA and Buckman enter the discussion. The court nicely summarized the public policy reasons supporting Buckman preemption:

The understandable concern is that allowing a private right of action for “fraud on the FDA” would embroil the agency’s staff, particularly its scientific staff, in court litigation to the derogation of their performance of their primary duties. This would result both from the consumption of time in litigation activity and from a concern that their routine duties, decision-making processes and public communications would all have to be vetted with litigation avoidance in mind.

Id. at *6.

Next the court establishes that the Wyeth v. Levine “clear evidence” standard does not require indisputable evidence, but rather defendants must establish impossibility preemption by “clear and convincing evidence.” Id. And finally, the court found that preemption is a question of law, not fact and therefore an issue for the court, not the jury. To the extent deciding preemption requires the court to rule on a factual dispute regarding what the FDA would do, “the dispute is one regarding a legislative fact, not an adjudicative fact. Thus it presents a legal question for judicial resolution.” Id. at *7.

With that as background, we turn to the specific facts of the case. Risperdal is an anti-psychotic medication. Id. at *1. The Risperdal label that plaintiffs claim is inadequate was modified in 2006 to include language about prolactin elevation and the reported rate of gynecomastia (enlargement of male breasts) among Risperdal users. Id. at *3. Plaintiffs allege that defendants failed to adequately warn of this risk because the labeling did not include the “true” rate of gynecomastia and should have included an instruction to physicians to monitor blood prolactin levels. Id. at *1. Defendants moved for summary judgment on preemption grounds in 5 cases involving plaintiffs from 4 different states. Id.

With respect to the rate of gynecomastia set forth in the label, the FDA specifically approved the pooling of data from 18 studies to arrive at the rate that was used in the labeling. Id. The FDA re-examined the label in 2007-2008 and concluded that it required no labeling changes regarding gynecomastia or prolactin elevations. Id. Plaintiffs argue that the pooled average was lower and that defendants should have disclosed the higher incident rates observed in 2 of the 18 studies. Id. at *8. Because those 2 studies were among those considered by the FDA during the approval process, “the FDA’s position is clear [as to] how information regarding the 18 studies should be described in the label.” Id. This portion of plaintiffs’ claim is therefore preempted.

As to plaintiffs’ allegation that defendants failed to adequately warn about monitoring prolactin levels, plaintiffs rely on “Table 21.” This is a table containing an analysis of 5 studies purportedly showing a statistically significant association between elevated prolactin levels and gynecomastia. Id. The table was in a draft of a journal article but not in the final publication and therefore not submitted to the FDA during the approval process. The studies reviewed in the table, however, were among those considered by the FDA. Id.

 Plaintiffs argued that based on the information in the table, defendants should have independently changed their label under the Changes Being Effected (“CBE”) regulations. However, to implement a CBE label change without prior FDA approval, the change must be based on “newly acquired information” which is defined as

data, analyses, or other information not previously submitted to the agency, which may include (but are not limited to) data derived from new clinical studies, reports of adverse events, or new analyses of previously submitted data (e.g., meta-analyses) if the studies, events or analyses reveal risks of a different type or greater severity or frequency than previously included in submissions to FDA.

Id. at *9. Table 21 doesn’t fit that definition. It may be a different analysis of the data, but it did not show a different type or greater severity or frequency of the risk of gynecomastia which is the focus of the CBE regulation. To the extent plaintiffs also tried to rely on their litigation experts’ analysis of the data, the court point out that that was a tactic tried and rejected in several other litigations as a means of avoiding preemption. Id. So, plaintiffs’ claims are also preempted because the data they rely on to suggest defendants could have changed the label is not newly acquired and could not serve as the basis for an independent label change.

Not only could defendants not have changed the label on their own, the court found clear evidence that the FDA also wouldn’t have approved it.  In 2012, one plaintiffs’ counsel submitted a Citizen’s Petition to the FDA alleging that the Risperdal label did not adequately address elevated prolactin levels, the need to monitor for elevated prolactin levels, or the rates of gynecomastia. Id. at *4. Essentially the same allegations raised in the litigation. In its response denying the petition, the FDA stated that it was commonly known that Risperdal increases prolactin and that gynecomastia is one of the manifestations of increased prolactin. Id. at *5. Based on that the court concluded:

This Court is persuaded that these reasons articulated by the FDA in response to the very claims alleged here provide the kind of “clear evidence” of “legislative fact” which the U.S. Supreme Court requires before a court can hold that impossibility preemption applies. By any standard, there is “clear evidence” that Plaintiffs’ entire theory of label inadequacy focused on prolactin levels was not only considered and rejected by the FDA but also rests on information (and allegations) known to the FDA and the medical community. The FDA’s review of the 18 clinical studies—which form the underlying data of any theory that Plaintiffs posit—both pre-approval and in subsequent reviews, and its subsequent inaction, seem to be the definite upshot of a conscious FDA choice on information before the agency. It is not this Court’s job to revisit a decision made by the FDA.

Id. at *11.

It may not be the court’s job to revisit FDA decisions, but it is certainly this blog’s job to visit strong preemption decisions like this one. So, just a reminder that if you get a good decision – forward it along. Don’t make us wait 6 months to join the party.

We’ve got a copy of the Reply Brief filed by defendant Wyeth in the Wyeth v. Levine case. Here are the highlights as we see them (we’re trying to stay away from the Phenergan or Wyeth specific stuff):

  • Conflict preemption operates by the direct force of the constitution. Express congressional intent to preempt thus is not an issue. It’s enough that Congress intended to have drug labeling decisions made by the FDA. Contrary state-law claims interfere with the accomplishment of that objective, and are preempted by the Supremacy Clause. Neither the presence, nor absence, of expressly preemptive language makes a difference to a finding of actual conflict.
  • This is a failure to contraindicate case. It’s a lot different from the 40+ year old drug product liability cases plaintiffs assert as “history.” With respect to labeled uses, the FDA is required by law to balance the risks against the benefits of all labeled indications. As a matter of law, state law cannot obstruct federal regulation of drugs by imposing a different outcome to this balancing decision from the one reached by the FDA.
  • As a matter of fact, the FDA did address the risks and benefits of the particular method of drug delivery at issue in this case. It was required to do so by law. The agency told the defendant to use “identical” labeling to that which the jury held was inadequate under state law (our one foray into Phenergan-specific stuff).
  • Because plaintiffs and their amici cannot win on the facts of the case, they try to make it into something it is not. There are no allegations of concealment from the agency or of failure to make a label change in response to newly emerging risk information. Here the FDA had all the necessary information before it, and made a decision. Plaintiffs told the jury explicitly that they could, and should, ignore the FDA’s determination. The court need not reach scenarios that are not before it.
  • The FDCA has always broadly prohibited unilateral changes to agency approved labeling, especially where methods of use are involved. This was not, and could not have been, a CBE case.
  • The FDA’s interpretation of the CBE, supported by contemporaneous administrative history, as an administrative construction of the agency’s own regulation, is entitled to considerable (Auer) deference. Thus the CBE regulation was always intended to apply solely to new, emergent information. There is not factual basis for the CBE regulation, as properly interpreted, to apply to this case.
  • The FDA regularly consults with manufacturers about possible CBE submissions. It’s quite effective in making its wishes know. Therefore, it’s hardly surprising that there aren’t many enforcement actions based upon improper CBE submissions.
  • The FDA’s “association with a serious hazard” standard (§ 201.80(e)) does not address when CBE submissions are to be filed, only the nature of the information that can be included. To interpret it in the broad fashion plaintiff does would render the CBE regulation itself a nullity.
  • The argument that defendants can simply pay judgments and continue doing what they are doing is meritless. That’s another form of special pleading for tort claims that the Supreme Court has repeatedly rejected, most recently in Riegel, every time it has held that tort law is premised on legal duties and is therefore preempted.
  • The argument that the defendant must prove that a CBE submission would be unsuccessful to qualify for preemption would involve courts and juries in adjudicating hypothetical, contrafactual situations, which the Court has disapproved in analogous situations (including Buckman).
  • There are excellent discussions of Sprietsma, Bates, and Lohr as involving situations where the responsible agency was excused from making specific judgments that balanced safety and risk. The NDA process is federal safety review. This will be quite useful in other cases.
  • That devices, but not drugs, do not have express preemption language is an accident of history, as Congress didn’t get into the habit of including express preemption clauses in statutes until after the original FDCA was enacted.
  • There’s also an excellent discussion of why the legislative history of the 1933 version of the FDCA isn’t relevant to the version that finally passed five years later, since the 1933 bill did not provide for FDA pre-approval of drugs and their labeling. This will also is useful in any case where the plaintiff makes the argument.
  • It’s irrelevant what precise authority the FDA had before 2007 to order post-approval labeling changes, since that isn’t what this case is about. The FDA told the defendant exactly how to address this particular risk at the time, and did so on full information.
  • Our pick for best quote: “No matter its purposes or benefits, tort law cannot be said to complement federal regulation when it imposes a state-law duty that contradicts manufacturers’ federal obligations and FDA’s labeling approval judgments.” Reply brief at 27.

In short, there’s a lot of good stuff here for those of us laboring in the preemption vineyards.

We’ve now had a chance to review the new “change being effected” (or “CBE”) rule finalized by the FDA last week. (Here’s a link.) Here are the cites: 73 FR 49603, 2008 WL 3874230. For drugs, the rule is 21 C.F.R. Sec. 314.70(c)(6).

This is what we found.

First, the FDA made one change from the draft rule that it proposed in January to the final rule that it adopted last week. The draft rule said that “newly acquired information” included data “derived from new clinical studies, reports of adverse events of a different type or greater severity or frequency than previously included in submissions to FDA, or new analyses of previously submitted data.”

The final rule changes the location of the italicized words. In the final rule, “newly acquired information” includes data “derived from new clinical studies, reports of adverse events, or new analyses of previously submitted data (e.g., meta-analyses) if the studies, events or analyses reveal risks of a different type or greater severity or frequency than previously included in submissions to the FDA.

This change clarifies that the (what the heck is that? an adjectival phrase?) applies to all of the types of data, rather than just the adverse event reports.

We found some stuff we like in the FDA’s responses to the comments submitted about the draft rule. At page 6, you’ll see the FDA’s statement that, “A sponsor’s submission or FDA’s acceptance of a CBE supplement does not necessarily mean that a drug product actually has caused any particular adverse event or type of adverse event.” That should go without saying, but we’re glad to see it said nonetheless.

One commenter suggested that the FDA should distinguish situations in which a drug company “must” warn of a risk from situations in which a sponsor “may” warn. The FDA rejected that suggestion because the FDA-approved warning on a drug is both a floor and a ceiling on the risk information that should be provided: “FDA has previously stated and reiterates here that it ‘interprets the Act to establish both a ‘floor’ and a ‘ceiling’, such that additional disclosures of risk information can expose a manufacturer to liability under the act if the additional statement is unsubstantiated or otherwise misleading.'” Id. at 7.

As the FDA (and the Supreme Court) has said before, “FDA does not believe that the absence of an express preemption provision with respect to drugs affects the application of the doctrine of implied preemption.” Id. at 8.

The FDA devotes a couple of paragraphs to explaining the “rule of construction” found in the recently enacted Food and Drug Administration Amendments Act of 2007. Among other things, “the rule of construction operates to preserve Federal labeling obligations only in the face of an argument that ‘this paragraph’ . . . ‘affects’ [the drug manufacturer’s] responsibilities.” Since the rule of construction is limited to that one paragraph, it cannot be read to apply to the FDAAA (or the FDCA) as a whole.

The FDA rejected one comment’s proposal that the agency “provide a comprehensive, written response to every CBE supplement submitted to the agency by a sponsor, describing FDA’s ground for approval, disapproval, or, as the case may be, request for modification to the submitted CBE supplement.” Id. at 18. The FDA’s position may disappoint those who favor Catherine Sharkey’s “agency deference” model of preemption, but it does save the agency a lot of work.

Finally, the FDA repeats that the new CBE rule simply codifies what had been the agency’s longstanding CBE practice (id. at 22) and repeats the FDA’s now oft-stated views on preemption. Id. at 23. And, in a sentence that may occasionally prove useful in litigation, the FDA writes that “at least when a sponsor did not meet the standard to change its labeling through a CBE supplement under this rule to include the warning a plaintiff alleges should have been added to labeling, State law liability that is premised on a failure to warn is preempted.” Id. at 24.

We liked this rule when it was proposed back in January, and we still like it today.

A confession: There wasn’t supposed to be a Part II to this post.

When we posted our earlier users guide to the Levine defense amicus briefs, we thought we had included everything – we really did.

Shows what we know.

It seems our “good ol’ boys” approach to collecting the defense amicus briefs had a few holes in it. We called up some folks we thought were likely to have copies and asked them to send what they had to us. We checked the Supreme Court’s docket, too, but we should have known it was: (1) not that current, and (2) not 100% accurate. So we missed two – there were also amicus briefs filed by DRI (we should have known those guys wouldn’t miss this party) and by the Generic Pharmaceutical Association. We missed DRI’s appearance because the docket had them listed with the plaintiff-side counsel (we do hope they’ll get that fixed – that’s not where the “The Voice Of The Defense Bar” belongs). We missed the GPA (great acronym) because they’re new kids on the block, and because pioneer and generic pharma don’t talk to each other enough about this kind of thing.

Anyway, we’ve got ‘em now, so here goes:

Part II

We’ll do DRI’s brief first because … well … they’re DRI, and because Dan Troy wrote it – and he’s one of our preemption heroes.

But heroism aside, what does DRI have to say that’s interesting and different?

It’s not new, but DRI does highlight plaintiff’s counsel’s closing argument tirade against the FDA in the very first paragraph of the summary of argument. We observed both in Part I and in our initial post on Levine that plaintiff’s explicit attack on the FDA is one of the best things the defense side has going for it in this case. It’s really hard for the other side to tell the jury it should disregard the Agency, and then to argue before the United States Supreme Court, with a straight face, that there’s no conflict between the verdict the plaintiff obtained and the Agency’s actions.

Good facts make good law.

We’ll skip over DRI’s lengthy description of the FDA’s New Drug Application process – not because there’s anything wrong with it (it’s quite good, actually, and contains a minimum of filler between the relevant citations), but because at least four of the briefs we’ve already read have something similar in them. We do particularly like the quote from 50 Fed. Reg. at 7470 about the effects of “substantive changes in labeling.” If you ever find yourself needing to write something similar in your own cases, remember that this is here.

Plagiarize. Let no other lawyer’s work evade your eyes (which is, in and of itself, plagiarized from Tom Lehrer).

One part of DRI’s administrative argument that is distinctive is the relatively lengthy discussion of the enhancement of the FDA’s powers by the recent FDAAA. DRI br. at 12-14. This goes to show that there’s a right way (congressional action) and a wrong way (complaining in state-law litigation) to deal with perceived deficiencies in FDA regulations. The 2007 FDAAA might not be directly relevant to regulatory actions that occurred in the 1990s and before, but it sure as heck will be relevant to new litigation that comes in the door – so we’re happy to have the analysis.

The most notable “new” argument that DRI brings to the table is the detailed discussion of why and how the FDA is inherently better suited to making society-wide safety judgments about the risks and benefits of the products it regulates. Other defense briefs have this point, but none with the thoroughness of DRI. DRI’s argument starts with judicial deference to administrative decision-making:

Congressional design requires federal courts – never mind state law and state lay juries – to accept and defer to FDA’s expert judgment and exercise of discretion about such difficult issues. Determining reliable scientific data is not the judicial function. Congress vested that responsibility in the FDA and [courts] will not preempt its presumed expertise. Indeed, the Court has observed that it is enough for us that the expert agency . . . has determined that such regulation is desirable for the public health, for we are hardly qualified to second-guess the Secretary’s medical judgment.

DRI br. at 17-18 (all sorts of citations and quotations omitted).

Using the judicial deference cases as sort of an extended a fortiori argument, DRI turns to the heart of the FDA’s administrative advantage – the point, already accepted by seven justices in Riegel, that state-law juries are “poor substitutes” for the FDA at making drug safety decisions that affect everyone in our society:

A dramatic gulf separates the competence of FDA from that of state jurors who, absent preemption, are called on to “police” prescription drugs and second-guess federal regulators. Laypersons applying state law cannot substitute for FDA, the expert agency congressionally delegated the role of meticulously balancing nuanced and sometimes competing nationwide goals. FDA’s regulation of prescription drugs is, as shown above, governed by a sweeping network of federal law individually applied to each drug. A lay jury lacks the expertise and the broader perspective of FDA and thus cannot adequately adjudicate individual patient risks in the context of population benefits – quintessentially the kind of issue FDA must confront every day. Without preemption, however, state juries are asked to do just that. See, e.g., Pet. Br. 23 (“‘Thank God we don’t rely on the FDA to . . . make the safe[ty] decision. You will make the decision.’”) (quoting JA 211).

DRI br. at 19-20. That’s the DRI’s main preemption argument in a nutshell – beginning with a paraphrase of the Supreme Court’s conclusion about juries vs. the FDA in Riegel and ending with plaintiff’s counsel’s anti-FDA tirade to the jury in the Levine trial.

Pretty powerful stuff, we think.

Good enough to win? We hope.

The next portion of DRI’s brief is devoted to analysis, much of which we haven’t seen before, that supports the propositions stated in its opening paragraph. It makes great reading for defense-side fellow-travelers like us.

  • The complexity of the FDA’s decision-making. DRI br. at 20.
  • The limitations of the tort system in dealing with unavoidably unsafe products. Id. at 20-21.
  • The limited perspective of juries. Id. at 21.
  • Research on juror tendencies to overrate low-probability risks. Id.
  • Justice Breyer’s remarks about the FDA versus jurors during the Kent argument, and similar statements. Id. at 21-22. We blogged about his little speech before.
  • Juror inability to engage in the sort of tradeoffs that the FDA is called upon to make. Id. at 23.

DRI then does something else that we were hoping to see – it puts before the Court the preemption issues presented by the SSRI-suicide litigation. DRI br. at 23-26. We’ve mentioned before that, as excellent as the Levine facts are for the defense preemption position, the SSRI cases arguably present even better facts for preemption – “because of the sheer magnitude of more than a decade of intense FDA involvement, and the Agency’s repeated conclusions (through the present day) that the plaintiffs’ warning proposals lack scientific support.”

We’re delighted that DRI took the time to explain this other preemption situation to the Court.

For good measure, DRI also discusses (br. at 26-27) the Dowhal litigation (we’ve addressed that here) – another specific situation where preemption was extremely appropriate, given the amount of consideration that the FDA had given to the relative risks of the smoking cessation devices at issue.

DRI ends its brief with policy: tort liability leading to overwarning and defensive labeling (br. at 28-30); tort liability as a disincentive to innovation (id. at 31); and tort liability as an incentive to remove valuable products from the market (at least in this country) (id. at 31-33), and the tort tax that all this litigation exacts from the health care system. Id. at 33.

The other Levine amicus brief we overlooked before was filed by the GPA – on behalf of the generic drug manufacturers. The mere fact that we did so is worrisome to us. As GPA points out, over 60% of all drug prescriptions are generic these days. GPA br. at 1.

But we don’t represent generics.

Why not?

Conflicts of interest, real or potential. The problem is that in so many areas (other than product liability), the pioneer drug manufacturers (our clients) and the generics are engaged in almost constant litigation with each other. Groups on the opposite side of that much litigation don’t tend to use lawyers who also represent the other side. There are all kinds of real or potential conflicts of interest that prevent cross-representation, and thus cross-fertilization of legal thinking. In fact, just last weekend, we jokingly referred to the pioneers and the generics in the same breath as the Hatfields and the McCoys.

But on an issue like preemption in product liability litigation, the pioneer and the generic manufacturers are almost entirely on the same side. Colacicco was an example of that – where both types of manufacturers were sued in the same case. We’ve fretted before that, because the generics are relative newcomers to product liability litigation, their lawyers might lag a bit on the learning curve on an issue like preemption, where the lawyers representing pioneer manufacturers have had to think about the issue so much more – simply because their clients have been sued so much more frequently.

Indeed, one reason we launched this blog was to create a place where all defense lawyers could exchange ideas on a wide range of defense issues – with no conflict issues to complicate that exchange, because everything here’s both public and free.

(On the other hand, you get what you pay for.)

All of that is basically a lead in to this: We’re pleased and relieved to see the generics – represented by a top drawer firm – jump fully into the fray on preemption. (Although, as we noted before, one generic did nice work in Gaeta v. Perrigo and another in Mensing v. Wyeth.) We hope to see the generics around as amici more often.

So what does GPA have to say?

What would you expect?

GPA mostly explains how preemption works in the somewhat different statutory framework that applies to generic drugs. We’d already seen some of this in Colacicco, but the Third Circuit never reached any issue peculiar to the generic manufacturer’s situation.

In this brief, the generics give it to the plaintiffs with both barrels.

GPA sees preemption in the generic drug context as “even stronger” than with the pioneer manufacturers that we represent. GPA br. at 2.

We wouldn’t go that far, but we understand why GPA would.

Why? Well, here’s what GPA argues:

  • Generic manufacturers don’t have control over their own labels. By statute, generic labels must be the “same” as the label used by the original pioneer manufacturer of the drug. 21 U.S.C. §355(j)(2)(A)(v). GPA br. at 5, 9.
  • Because of the statutory “sameness” requirement, the FDA does not allow generics to change their labels without pre-approval for any reason – there’s no “changes being effected” exception for generic products. E.g., 73 Fed. Reg. 2848, 2849 n.1 (FDA Jan. 16, 2008). GPA br. at 6, 11.
  • The FDA considered, and rejected, providing CBE procedures for generic drugs. GPA br. at 9-10. Interesting – we didn’t know that.
  • Even when the pioneer manufacturer uses the CBE exception to change its labeling without prior FDA approval, a generic manufacturer may not deviate from the last FDA-approved labeling. E.g., 57 Fed. Reg. 17950, 17961 (FDA Apr. 28, 1992). GPA br. at 6, 10.
  • It is therefore impossible for an generic manufacturer to comply with both a state law determination that its warning is inadequate and the federal “sameness” mandate. 21 C.F.R. §314.150(b)(10). GPA br. at 7, 11-12.

We give GPA a 4.0.

We’re now much better informed about the generic manufacturers’ preemption arguments than we were before. We’ll try to keep our eyes open for anything new and interesting that’s relevant to generic drug product liability litigation. And we invite attorneys who represent generics to alert us to this type of information.

Remember: If nobody tells us, we can’t share.

So that’s it. Now we’re sure we’ve looked through everything that our side has filed in Levine.

But if we’re wrong, tell us.

We posted Wyeth’s principal brief in Wyeth v. Levine as soon as we got it to ensure our readers’ “first on the web” priority (much like we did with today’s Vioxx decisions). That was a “sight unseen” post, something we don’t generally like doing. We’d much rather give you our take on things – we figure that’s why most of our visitors come here in the first place, since they can get pure “news” elsewhere – but sometimes things can’t be helped.

Anyway, that was then; this is now. Now we’ve read it. We’re impressed with the job Wyeth’s brief does defending the preemption principle in a critical case. And we still think, like we did before, that the case itself is an excellent bridgehead for establishing the $64,000 question – preemption, or not – in the prescription drug product liability field.

We won’t rehash our prior post, where we discussed the favorable facts of Levine in depth, but we want to emphasize one key point – this isn’t just a failure to warn case, this is a failure to contraindicate case.

That makes a big difference, and it makes the preemption argument all the much stronger than it would be in “just” a warning case.

The plaintiff’s position in Levine is not that the warning of the risk (gangrene and amputation) of the use (intravenous (“IV”) injection) wasn’t explicit or detailed enough. No, her position is much more extreme. That position is that nobody should be allowed to use this drug for this FDA-approved purpose – ever. The plaintiff argued at trial, quite explicitly, that the FDA goofed:

Dr. Harold Green, another of respondent’s witnesses, also criticized FDA’s approval of Phenergan’s labeling, testifying that, in his opinion, “somebody at the FDA made an administrative error and approved it.” JA 82. In his view, the drug should not have been approved for intravenous administration. JA 79-80. Respondent’s FDA expert similarly “disagree[d] with FDA’s conclusions” to approve Phenergan’s labeling. JA 98. . . . During closing arguments, respondent’s counsel. . .invited the jury to override FDA’s labeling approval decision: “Thank God we don’t rely on the FDA to. . .make the safe[ty] decision. You will make the decision.” Id. “The FDA doesn’t make the decision, you do.” JA 212.

Wyeth Pr. Br. at 22-23. That’s some heavy-duty FDA-bashing. It’s facts like this that make the defense preemption argument much easier than it might otherwise be. That’s what we want in what has become the industry’s “point of the spear” case.

The alternative argument, of course, is that facts like this make anything the Supreme Court might say distinguishable. We don’t think the distinctions are all that great in terms of whether there should be preemption or not, but no matter. We’ll happily litigate any claimed distinctions once we get the core proposition – that FDA labeling decisions preempt state product liability claims – established.

That the plaintiff in Levine explicitly made (and was allowed to make) the John C. Calhoun argument that a jury applying state law can nullify supreme federal law just makes our side’s “Job 1” that much easier.

How much easier?

Well, it’s not every preemption brief that can start off with an “impossibility,” rather than an “obstacle,” argument in the lead position. Wyeth’s brief does. See Wyeth Pr. Br. at pp. 29-40.

Whoa, whoa, whoa…. Slow down guys. Cool it with the inside baseball jargon already.

OK, what we’re talking about here is that there are two main flavors of implied (that is, not by virtue of statutory language) conflict preemption: The first kind arises when, in the words of innumerable (that means so many that we won’t bother with citations) Supreme Court decisions, “compliance with both federal and state regulations is a physical impossibility.” That’s yes/no or can/can’t preemption. The feds say one thing and the state says something else, and doing one means you can’t do the other. It’s the best kind of conflict preemption to assert because the conflict is both easily apparent and complete.

It’s what happens when counsel, in closing arguments, tells the jury that they should give the FDA the old Bronx cheer.

The second form of preemption occurs, in the words of equally numerous Supreme Court decisions, “when state law stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” That’s “obstacle” preemption. It’s broader, but does not involve as palpable a degree of conflict. It’s also the type of implied preemption most commonly invoked in product liability cases.

Does impossibility work in Levine? Well, the FDA, after evaluating IV injection of Phenergan, approved the drug for that use – not once but several times. It approved labeling including that use and addressing the risks at issue in Levine. As Wyeth points out in its brief, FDA approval is “conditioned upon” the manufacturer using the labeling approved by the FDA “exactly as directed” by the Agency. Br. at 30 (quoting 21 C.F.R. §314.105(b)). The plaintiff’s verdict is based upon the theory that the FDA approved use should be contraindicated – that is, never allowed under any circumstances.

It’s rather difficult to comply with “yes, you must” and “no, you can’t” at the same time. So we’d say the impossibility argument works pretty well. But we’re biased, of course. That, and the opposing argument would seem to have a hard time passing the red face test.

Unless “yes” can be interpreted as “maybe.”

Which brings us to all plaintiffs’ favorite toy in preemption cases – the “changes being effected” (“CBE”) regulation, §314.70(c)(6)(iii), that allows manufacturers to make certain changes to their warnings without prior FDA approval. Just what those changes are is one of the key issues in the Levine case. The Vermont Supreme Court held that anything that arguably “strengthens” a warning is permitted by the CBE regulation, even if the information has been known for decades and already brought to the FDA’s attention. Wyeth argues for a narrower interpretation, that to qualify for this exception to prior FDA labeling approval the information must be something that’s “new.” Wyeth Pr. Br. at 31-32.

The argument that the CBE exception can’t be expanded so far that it swallows the ostensible rule requiring prior FDA approval of labeling is a strong one. As we’ve discussed before, all of the regulatory history supports limiting the exception to novel information that the FDA has not previously had occasion to review. The Agency itself not only supports this interpretation, but is in the midst of rulemaking that would bring the regulation’s language into line with longstanding regulatory intent. See 73 Fed. Reg. 2848 (FDA Jan. 16, 2008). As we pointed out at the time, the effect of this regulatory change would take the plaintiffs’ favorite toy away from them. In Levine, plaintiff’s reading of the CBE regulation would put manufacturers under a state-law duty to change FDA-approved labels unilaterally, even if the Agency’s approval had rejected that change and no new facts had arisen in the interim. Wyeth, of course, makes all of these points in its brief. Br. at 34-40. Not only that, but Wyeth did a lot more digging in the Federal Register than we did, in our post, so we highly recommend this part of the brief to anyone confronted with a similar CBE-based anti-preemption argument.

Anyway, Wyeth has an additional arrow in its quiver – the Supreme Court’s holding earlier this year in Riegel v. Medtronic, Inc., rejecting the contention that manufacturers had broad discretion to make unilateral post-approval changes to FDA approved devices. 128 S. Ct. 999, 1007 (2008) (“the FDA requires a device that has received premarket approval to be made with almost no deviations from the specifications in its approval application”). That same quote was point number two in our own long Riegel post, a point that Wyeth drives home in its brief:

In Riegel, the Court held that once a Class III medical device receives premarket approval, the FDCA (as amended by the Medical Device Amendments of 1976 (MDA)) forbids the manufacturer to make any change without FDA permission, including a change to the approved labeling, that would “affect safety or effectiveness.” Id. at 1005. Any change must be submitted by supplemental application for FDA approval before implementation. Id. The Court thus held that FDA’s detailed, individualized review of the safety and effectiveness of each Class III medical device imposed a federal-law “requirement” that approved devices be made “with almost no deviations from the specifications in its approval application,” and preempted conflicting state-law requirements applicable to the device. Id. at 1006-1007.

Wyeth Pr. Br. at 31. After all, as Wyeth points out, medical devices are governed by a substantively identical CBE regulation. Id. at 32. See also id. at n.13 (describing how the medical device approval process – found preemptive in Riegel – was generally modeled after the approval process for prescription drugs).

In our own Riegel post, we called the kind of Riegel-based arguments that we’re seeing in Wyeth’s brief “spillover effects.” There’s another one on page 33 – Wyeth’s argument that “[n]o less than a state statute or regulation, this common-law claim is preempted if it imposed a duty that conflicts with federal law.” The first case cited, once again, is Riegel, for its holding that “common-law liability is premised on the existence of a legal duty, and a tort judgment therefore establishes that the defendant has violated a state-law obligation.” 129 S. Ct. at 1008.

After finishing with its preemption by impossibility argument, Wyeth goes on to make the more conventional preemption by obstruction argument that prevailed in Geier v. American Honda Motor Co., 529 U.S. 861 (2000), and Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001) – and that we’ve seen be so successful in, inter alia, the SSRI warning cases. See Wyeth Pr. Br. at 40-51. This part of Wyeth’s brief makes the argument – now indelibly connected with Riegel (see the block quote in our post on that case), that allowing juries to disregard FDA drug approval decisions, puts such determinations in the hands of a legal system inherently incapable of properly balancing the benefits that drugs offer to society at large:

Vermont, by contrast, seeks to alter that balance by substituting the judgment of lay juries that focus on individual patients’ injuries on a case-by-case basis, effectively disregarding the countervailing benefits of the drug to the public as a whole. See Riegel, 128 S. Ct. at 1008. The Vermont judgment thus frustrates both Congress’s objective of having an expert agency serve as the ultimate regulator of the labeled conditions of use for which a drug is approved and FDA’s specific fulfillment of that objective.

Wyeth Pr. Br. at 40; see id. at 46. “The determination whether a drug is generally recognized as safe and effective. . .necessarily implicates complex chemical and pharmacological considerations” and is “peculiarly suited to initial determination by the FDA.” Id. at 43 (quoting Weinberger v. Bentex Pharmaceuticals, Inc., 412 U.S. 645, 653-654 (1973)).

Following a lengthy discussion (which we will leave to our readers, since it’s case specific) of everything the FDA did to evaluate the IV administration of Phenergan, Wyeth Pr. Br. at 43-44, the defense argument takes on frontally the lower court’s FDA as “minimum standards” rationale.

The regulatory scheme that Congress established in the FDCA and that FDA implemented in the specific case of Phenergan serves competing goals: to protect the public from unreasonable risk of harm, while ensuring the availability of beneficial treatments, all of which carry a certain degree of risk. And an approved drug’s labeling must provide sufficient instructions for safe and effective use, while avoiding limitations that foreclose beneficial use of a drug in an effort to avoid all risk of human error. For these reasons, as the United States has explained, “FDA interprets the [FDCA] to establish both a “floor” and a “ceiling”’ with respect to drug labeling,” and “FDA’s approval of labeling for a new drug reflects FDA’s expert judgment that the labeling strikes the appropriate balance.”

Wyeth Pr. Br. at 45 (quoting 2007 amicus brief of the United States in support of granting certiorari; other citations and quotation marks omitted). The FDA’s administrative balancing of competing objectives is thus equivalent to the kind of administrative process that supported preemption of no-airbag claims in Geier. Wyeth Pr. Br. at 47-48.

The final portion of Wyeth’s brief deals with an otherwise rather obscure (sufficiently so that we’ve only just touched on it before here and here) point about an uncodified preemption/savings clause (§202 of the bill) dating from the 1962 Amendments to the FDCA. It’s more important in Levine because the lower court held that the “direct and positive conflict” language in that section somehow eliminated the aforementioned “obstacle” form of implied conflict preemption.

In a couple of pages (pp. 52-54), Wyeth slices, dices, and purées that argument: (1) citing precedent (Geier) holding the express and implied preemption work independently; (2) demonstrating that the “direct and positive” language is simply a phrase the Supreme Court has used to describe conflict preemption generally (although there are more recent cases for this proposition – see Swift & Co. v. Wickham, 382 U.S. 111, 132 & n.3 (1965); Farmers Educational and Coop Union v. WDAY, 360 U.S. 525, 540-41 (1959)); and (3) establishing that all this language was intended to accomplish was to prevent any claim of preemption of the field.

Whew! We finally reached the end of it.

Anyway, it’s obvious to us that Wyeth has done the entire industry a favor while carrying the flag for our side in the Levine matter. We’re crossing our fingers in hopes that good facts, combined with good arguments (and good lawyering), will produce good law.

We’ve had an unusual number of plaintiffs’ side blogs link to our Riegel posts over the past several days. While we’re flattered that the other side thinks we’re worth noticing, the traffic from these links has produced quite a few comments trashing either the Supreme Court (especially Justice Scalia), the FDA, or both. We’ve returned fire, but since it’s our blog we’ve decided to elevate our comments to a small post. We’ve got basically two things to say:

About the Supreme Court:

Plaintiffs’ lawyers need to look in the mirror. The Riegel case involved a product put to a contraindicated use (calcified arteries) and then misused (overinflated) on top of that. Yet it was the manufacturer that was sued. Why? It was a suit that should never have been brought (at least against Medtronic), but the plaintiffs’ side appealed it all the way to the Supreme Court. That’s spelled h-u-b-r-i-s.

And then there’s the whole mass tort industry, which we parodied in Anatomy of a Mass Tort. Nowadays, if there’s recall, a warning letter, or practically any other other adverse information of any sort about a prescription drug or medical device, within a week there’ve been ten class actions filed and a massive plaintiff solicitation campaign going on. It’s so predictable and blatantly extortionate that it would be humorous, if it weren’t for all the time, effort, and money defendants are forced to spend defending against the mass tort of the week.

The tort system has gotten out of control. For one thing, it’s tried to usurp the role of the FDA. It’s not surprising to see courts looking for ways to rein it in.

The plaintiff’s side should stop criticizing Justice Scalia as if he were some sort of traitor. Plaintiffs never had Justice Scalia on preemption, but he’s no knee jerk vote for our side either. He’s kicked us in the teeth regularly on punitive damages. Justice Scalia has been a tort preemptionist of longstanding. He wasn’t what caused the plaintiff’s side to lose Riegel.

Again, a little mirror viewing is in order. As recently as Lohr – as recently as Bates – five justices voted against (most) preemption. The plaintiffs’ side lost four of those five in Riegel. That should say something. Plaintiffs appealed a lousy case all the way to the Supreme Court. They made some bad and extreme arguments, some of which been rejected by the Court several times before. But most importantly, the plaintiffs were trying to seize for themselves the FDA’s power to determine how devices (and drugs, but that’s Levine) are designed and what their labels would say. It’s not suprising that the FDA would oppose that.

The plaintiffs’ side should take a long hard look at the facts in Levine. Quite frankly, you’re doing it again – bad facts, bad arguments, and trying to usurp the FDA’s authority over drug labeling.

About the FDA:

We’ve had several comments calling the FDA a “joke” or something similar and mentioning (even attaching excerpts from) the recent advisory committee report commissioned by the FDA that is critical of a number (quite a few, actually) of the Agency’s shortcomings.

It’s easy to criticize, but it’s a lot harder to try to make things better. Perhaps it’s just the way the other side thinks – since second-guessing and criticizing defendants is the plaintiffs’ stock in trade – but we can’t think of anything the other side has ever actually done to try to improve the operation of the FDA. The Agency is no joke. Like it or not, it’s the entity that Congress has charged with overseeing the safety of drugs, medical devices, food, vaccines, and a lot of other stuff that keeps us alive and functioning.

The FDA only becomes a “joke” if Congress refuses to adequately fund it and otherwise fails to provide it with the authority and tools it needs to do the job – and if the administration fails to provide the leadership and will for the Agency to do its job (compare FEMA under the last two administrations). Even during the love-in that the plaintiff’s side had with the FDA under former Commissioner Kessler (amazingly, a Republican appointee), they did nothing to strengthen the Agency. All of the recent legislation that has provided additional funding sources and statutory authority for the Agency – FDAMA, user fees, and FDAAA – have been pushed by industry and either ignored, or actively sabotaged, by the other side of the “v”.

Industry even agreed to tax itself for the benefit of the FDA (something we’ve never seen lawyers do – the bar vehemently opposes any taxes on legal fees), but still the plaintiffs’ side carps. We’ve seen their hired experts come into court and call the FDA’s user fee program some sort of conflict of interest. The committee report says the FDA needs more money. That’s true. Well where’s that money going to come from? Industry has stepped up to the plate; the other side contents itself with throwing bottles from the stands.

We told you when we posted about it last week (here and here) that the FDA’s new proposed CBE regulation would create a firestorm.

The organization formerly known as ATLA was the first to react, but the Democrats in Congress were not far behind.

Here’s a link to today’s Star-Ledger reporting on a letter several leading Democrats have sent to the FDA objecting to the proposed rule.

We haven’t heard the last of this, we promise you, and if you want to read more, there’s a brand new synopsis of the FDA’s proposal at the FDA Law Blog.

The Wyeth v. Levine case is such a big deal that you can bet we’ll be discussing it a lot. For one thing, it has the potential to obliterate entire mass tort litigations at a stroke. Since we’re mass tort kind of guys, we’re keeping a mental hit list of where we’d strike first after a favorable outcome – but that’s not something we’d give away here – not now, anyway.

On the other hand, it’s not the millennium either. Levine won’t end modern prescription drug product liability litigation as we know it. But a favorable preemption ruling in Levine could well bring about a good, and long overdue, pruning. We see it – at a minimum – as eliminating a host of claims involving situations where the FDA has authoritatively spoken on a particular risk involving a particular drug. Several MDLs could well go bye-bye as a result.

Then again, we could lose. The Court could go the other way and restore the pre-Buckman world where guys like us, defending pharmaceutical companies, would gaze wistfully at the medical device folks (which also includes guys like us) and wish that we had a preemption defense, too.

Big issues raised in the Levine briefing to date, which means in the briefing on the certiorari (typically abbreviated “cert.” – referring to an obsolescent name for discretionary appellate review) petition – include:

  • Are FDA-approved labels a “minimum standard” or something more than that? We’ve already had plenty to say on this, here specifically in connection with Levine, and here more generally.
  • Is the defendant free to use a CBE (“changes being effected”) supplement to make a unilateral change that strengthens or adds a warning without prior FDA approval where the FDA already addressed that particular labeling issue? We’ve already had plenty to say on this issue, here.
  • Does the presumption against preemption apply in a conflict preemption case? We’ve already had plenty to say on this issue, here.
  • Does “direct and positive conflict” language in an uncodified part of the 1962 amendments to the FDCA oust implied “obstacle” preemption? We’ve touched on this issue here, here, and here, but haven’t devoted a lot of time to it yet. We will, but not today.
  • What degree of deference should be given to the FDA’s recent statements about preemption? We’ve already had plenty to say on this issue here.

And if you don’t want to take our word for it, all of the briefs on the cert. petition in Levine are available through the auspices of the really excellent SCOTUSblog – which is probably the legal blog we use more than any other in our actual, paying work. You can find links to all the Levine briefs right here (Levine is the second case on the list).

One other thing before we get to the point of this post. Contrary to the impression we had when the Supreme Court granted cert. last week, it’s not yet 100% sure that the Court will hear the Levine case this term. There were 6 cert. grants on the 1/18/08 order, and there are only two open slots for oral argument. For details, see the ever-informative SCOTUSblog. We think it’s likely that Levine will be one of those two because: (1) it is one of the two longest-pending cases on the Supreme Court’s docket (there’s no mystery to figuring that out; just look at the order for the lowest docket numbers after the “06-“); and (2) the Court already has two other FDCA-related preemption cases on its docket (if you don’t already know that, why are you reading this blog?), and they’ll probably want to decide them as a package. If the Court had wanted to wait, it would have followed the SG’s recommendation and simply held the Levine petition.

Getting Levine heard this term is – like everything else about the case – a big deal. Why? Politics. The legal positions taken by the SG and the FDA reflect the broader views of the administration that’s in office. We’d have to say that preemption in prescription drug and medical device tort litigation has fared pretty well with the current crew. But we read the same polls everybody else does, and we know that, for any case that gets held over for the next Supreme Court term, there’s a pretty good chance that a new administration’s political appointees will have a different take on the preemption issue.

So we’ll see what happens.

Now to the point – and yes, we do have one. There’s an old saying that “hard cases make bad law.” We don’t know who said it (somebody named “Maxim”?), but we think it’s true. So a logical first place to start when approaching something as controversial as the Levine case is with the facts. Specifically, how good are the facts for implied conflict preemption?

Well, first of all, Levine doesn’t involve any of the real preemption battleground drugs – it’s not an SSRI suicide case, nor does it involve Elidel, Prempro, Vioxx, Nexium, or any other drug that has produced multiple preemption decisions. See our Prescription Drug Preemption Scorecard for a complete list. That means Levine presents a playing field that’s relatively unimpeded by prior precedent.

That’s not to say that the facts are particularly recent – they’re anything but that. The drug in Levine is Phenergan, something that’s been on the market since 1955, and which has several uses. You may even have taken it as the active ingredient in some cough syrups. We have. It tastes rather revolting (at least it used to), but it will knock out a cough.

But that’s not the only use for Phenergan. It’s also used to treat really, really bad nausea of the sort that can land somebody in the hospital with dehydration. In these severe cases, it’s not enough just to take the stuff orally like we did with cough syrup. For faster relief, Phenergan can be given intravenously.

Beyond that, there are a couple of ways to administer a drug intravenously. There’s the “drip” method, where the fluid goes into your body at whatever rate your body wants to absorb it, and there’s the “push” method, which is used in the more serious cases when it’s necessary to get more of the stuff into you faster. No rocket science there.

Levine involves a medical decision to administer Phenergan by IV push, so this particular prescriber must have thought that the plaintiff-patient absolutely, positively needed the stuff.

Why do we say that? Well, like any prescription drug, Phenergan has inherent risks. In Levine, that risk was quite severe indeed. We don’t know why, and for purposes of Levine we don’t have to know, but if Phenergan gets into an artery – as opposed to a vein – it can cause gangrene.

That’s some risk – and that’s apparently what happened to the plaintiff in Levine. Somehow (and the briefs aren’t saying), the IV needle for the plaintiff’s Phenergan ended up poked into her artery rather than into the intended vein. As a result (we’re assuming medical causation for purposes of this discussion; we don’t know how much it’s disputed), the plaintiff lost her forearm to gangrene. That’s big damages, if there’s liability.

Gangrene is not only a serious risk of Phenergan but – more importantly for preemption purposes – it is also a known one. There was an FDA-approved warning about gangrene on the Phenergan that the plaintiff took. Naturally, the plaintiff claims that this label wasn’t enough, since that’s what plaintiffs have to do. If the drug’s label is adequate, the plaintiff loses. Plaintiffs don’t like to lose any more than we do.

First good fact for the defendant: The warning in this case was anything but wishy-washy. It mentioned the risks of “arterial” exposure in several places – including in boldface, all caps type: “INADVERTENT INTRA-ARTERIAL INJECTION CAN RESULT IN GANGRENE OF THE AFFECTED EXTREMITY.” It expressly told the treating doctor (under the learned intermediary rule all prescription drug warnings are addressed to MDs, not to patients) to consider whether an IV push method of injection was appropriate at all, given the risk. The warning advised “extreme care” not to get any Phenergan into the patient’s arteries. It stated that if anything went wrong and that did happen, it was “likely” that serious injury, specifically, “gangrene requiring amputation,” would result:

Due to the close proximity of arteries and veins in the areas most commonly used for intravenous injection, extreme care should be exercised to avoid perivascular extravasation [that’s “missing the vein and putting the drug in the surrounding tissue” for you non-doctors] or inadvertent intra-arterial [that’s “into the artery”] injection. Reports compatible with [that means “that could involve”] inadvertent intra-arterial injection of Phenergan Injection, usually in conjunction with other drugs intended for intravenous use, suggest that pain, severe chemical irritation, severe spasm of distal [that means “relatively far away”] vessels, and resultant gangrene requiring amputation are likely under such circumstances.

The label in question mentioned “gangrene” no less than four times.

In short, the defendant gave a warning that: (1) expressly mentioned the particular risk (gangrene), (2) described how serious it was (“likely” if the drug is introduced into the artery), (3) told doctors how to avoid the risk (use IV-drip in preference and be very careful), and (4) used a satisfactory format (all caps and bold). There are a lot of places – but apparently not the Peoples’ Republic of Vermont – where the Levine case would have ended with a summary judgment ruling that this warning was adequate as a matter of state law. You can find a complete list of those cases on pages 2.04-31 to -35 (there are a lot of them) in Bexis’ book.

We’ll tell you right up front that all this description comes from the defendant’s opening papers that sought (now, successfully) to get the Supreme Court to take the case. Come on, we’re defense lawyers. Do you think we’d take the facts from the plaintiff’s point of view? If you want a plaintiff-side slant, maybe you’ll find it here or here. Or maybe not. You’re not likely to hear about facts from a lawyer (us included), unless they help his/her side.

Anyway, we’d have to say that the Phenergan warning is pretty darn good – adequate as a matter of law in a lot of jurisdictions. So how does the plaintiff claim it’s inadequate if it specifically mentions the risk and tells the prescriber how to avoid it?

Good question.

The answer to that question is where preemption comes in. The plaintiff doesn’t really find fault with what the label says about the risks of IV-push administration of Phenergan. What she claims is that the defendant must go further and say don’t use IV-push at all – that the risks are so awful (and gangrene is surely that) that the added benefit of getting more of the drug into the body faster just isn’t worth it – ever. In medical/legal/regulatory jargon, that kind of a “don’t do it at all” warning is called a “contraindication.”

As you can tell, a contraindication is a very serious kind of warning. For that reason, the FDA has strict standards that have to be met before a contraindication can be approved. They’re found in 21 C.F.R. §201.57(d). That regulation requires that a contraindication must relate to a “known hazard” (the risk in Levine seems to meet that) and due to that hazard “the risk of use clearly outweighs any possible benefit.” (There’s where a contraindication doesn’t cut the regulatory mustard, in the opinion of the FDA, if not the plaintiff.)

The reason for preemption is that, in the case of Phenergan, the FDA had looked at all this itself. Gangrene was a known risk, first seen in 1967. The risk was reported to the FDA, and the resultant warning about that risk went through several versions over the years. By 1981 it said that, due to various risks, the IV-drip method was “usually preferable.” During that 14 year period before 1987, the FDA never once sought a contraindication, despite the severe nature of the risk.

There was a general reformatting of prescription drug labeling in the late 1980s, and Phenergan was part of that. In particular, the FDA sought a number of label changes relating to “recognition and management of unintended intra-arterial injection.” One of these was to enhance what the label said about IV-drip. The FDA wanted the label changed to state that the drip (as opposed to the push) method “may enhance the possibility of detecting arterial placement” – that is, the doctor could tell more easily if there was a potentially dangerous screw up. The approved label change also pointed out that the drip method “results in delivery of a lower concentration of any arteriolar irritant” (it’s words like “arteriolar” – and “perivascular extravasation” – that show why we have a learned intermediary rule). Thus doctors (who know these big words) understood that, if there actually is a screw up, and only drip is used, less Phenergan (the “arteriolar irratant”) gets where it doesn’t belong.

Again, that’s as far as the FDA ever went. It never suggested that IV-push was so risky that it should be contraindicated.

Second good fact for the defendant. The defendant manufacturer then attempted to make more changes to the Phenergan label that the FDA rejected. It made submissions in 1988 and 1989. Their exact nature isn’t particularly important, since they were rejected, but they included both inadvertent intra-arterial injection and other stuff not involved in the Levine case. This time, however, the FDA said “no.” That “no” occurred in 1997, about three years before the incident that involved this plaintiff and nearly a decade after the revised labeling was submitted.

We expect somebody to ask why the FDA took so long – the entire term of Bush I and the first Clinton term. It’s a good question, but we have no idea what the answer is. Maybe Dr. David Kessler, who ran the FDA for most of that time, does.

All we know is that, when the FDA did act, it said “no.”

Anyway, by letter dated February 21, 1997, the FDA told the manufacturer to ditch the new (or by that point, not so new) labeling changes – at least the part about intra-arterial injection: “f. Inadvertent Intra-arterial Injection – Retain verbiage in current label.” (This is in the appendix to the defendant’s papers, 160-163a for those of you keeping score at home.)

Well, when the FDA says “no,” that’s what a manufacturer does – if it likes to keep being able to sell its product. Thus, the new labeling was withdrawn and the old label apparently resubmitted (we’re not FDA regulatory experts, and we’re not sure why it was necessary to do that (perhaps other changes went through), but that seems to be what happened). The result was a second letter from the FDA dated September 18, 1998, reiterating that the manufacturer was not to change the label: “The final printed labeling (FPL) for the package insert must be identical to the draft package insert submitted May 8, 1998.” (Again, see the defendant’s appendix at 164-166a.)

Looking at those facts, we’d have to say that Levine presents a preemption case that we’re quite comfortable with. The issue of gangrene and intravenous administration is plainly not anything “new” that would justify use of a “change being effected” supplement – either under the FDA’s understanding of the CBE exception at the time that process was created, or under the Agency’s currently proposed revision to the CBE regulation.

Not only that, but Levine is a case where you can argue that the FDA said “no,” but the plaintiff is trying to get the common law to say “yes.” We like that kind of case because it’s a simple, powerful argument. The plaintiffs will undoubtedly try to nitpick exactly what label changes were before the FDA, but “retain” and “identical” are rather difficult words to get around. It’s like we tell our kids – “what part of ‘no’ don’t you understand?”

And there’s more. The plaintiff’s proposition in Levine is that the common law should essentially prohibit a medical procedure – by declaring that IV-push should be contraindicated. This is not just a matter of telling doctors to do something differently, but of telling them not to do it at all. What the plaintiff makes is a risk/benefit claim, not the usual inadequate warning claim. This sort of claim is considerably more extreme than those seen in most prescription drug product liability cases, and it plays right into the FDA’s greatest strength – it’s always been the Agency’s job to make risk/benefit decisions of this sort. Accordingly we think (hope?) it will prove more difficult for the Supreme Court to endorse such a claim than might be the case if a run-of-the-mill inadequate warning (e.g., “you should have stated the risk more clearly”) claim were at issue.

Finally, in 2006 the FDA set out six categories of failure to warn claims that it thought were preempted by the FDCA. We think that at least two of them are potentially relevant:

  • claims that a manufacturer failed to warn not including contraindications or warnings that are not supported by evidence that meets the FDA’s scientific standards; and
  • claims that a manufacturer failed to warn by not including a statement in its labeling or in advertising that the FDA had prohibited the manufacturer from stating.

71 Fed. Reg. 3922, 3936 (FDA Jan. 24, 2006) (emphasis added).

Whether the first exception applies depends upon facts that we don’t have, but it certainly doesn’t appear from what the FDA concluded that the “risk clearly outweigh benefit” standard of §201.57(d) is met. The second category of FDA-recognized preemption clearly applies, since the FDA stated explicitly that the manufacturer was to continue using the label at issue without any changes.

Another thing we like about the Levine facts is what they apparently don’t contain. For that, we do look at the plaintiff’s papers.

Reading the plaintiff’s brief against certiorari, we find no claim that the defendant misstated or understated the risk in its label. That’s rare in a drug labeling case, and quite good for the defendant. While that’s not really surprising, given the plaintiff’s contraindication argument, it’s one less bunch of adverse facts to have to deal with. Going further, there’s no suggestion that the defendant ever hid anything from the FDA – and plaintiffs absolutely love to call defendants criminals any time they possibly can. While we don’t think misrepresentations to the FDA should matter under Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001), as a matter of atmospherics they usually do.

We’ll go through the legal arguments in Levine soon enough – probably after the top-side merits briefs are filed – but looking at the case purely from a factual perspective, we have to say that we’re encouraged by what we see.