As our post-Levine preemption cheat sheet demonstrates, Mensing/Bartlett preemption is breathing down the necks of all prescription drug design defect claims. Recent cases finding preemption of design defect claims due to the need for FDA pre-approval of “major” or “moderate” design changes (basically, anything that could be causal in a product liability lawsuit) include: Yates v. Ortho-McNeil Pharmaceuticals, Inc., 808 F.3d 281, 298 (6th Cir. 2015); Young v. Bristol-Myers Squibb Co., 2017 WL 706320, at *5 (N.D. Miss. Feb. 22, 2017); Utts v. Bristol-Myers Squibb Co., ___ F. Supp.3d ___, 2016 WL 7429449, at *12 (S.D.N.Y. Dec. 23, 2016); Brazil v. Janssen Research & Development LLC, 196 F. Supp.3d 1351, 1364 (N.D. Ga. 2016); Fleming v. Janssen Pharmaceuticals, Inc., 186 F. Supp.3d 826, 832-34 (W.D. Tenn. 2016); Batoh v. McNeil-PPC, Inc., 167 F. Supp.3d 296, 320-22 (D. Conn. 2016) (OTC drug); Barcal v. EMD Serono, Inc., 2016 WL 1086028, at *4 (N.D. Ala. March 21, 2016); Rheinfrank v. Abbott Laboratories, Inc., 137 F. Supp.3d 1035, 1040-41 (S.D. Ohio 2015); Trahan v. Sandoz, Inc., No. 3:13-CV-350-J-34MCR, 2015 WL 2365502, at *6 (M.D. Fla. March 26, 2015).

We don’t expect the other side just to sit idly by and watch their design defect claims get washed away by a preemptive deluge, and they haven’t. To counter preemptive FDA-pre approval design requirements, they’ve conjured up the idea of a “pre-approval” design defect.

What the heck is that, you ask? Well, since preemption depends on the regulatory requirement to get FDA approval for any design change that could affect product safety, these cockamamie claims try to change the time-line – targeting the design as it stood before the drug was first submitted to the FDA in the first place. Since way back when, the prospective NDA holder could have chosen to submit some different molecule to the FDA, plaintiffs claim that the failure to do that was a “design defect.” That is, they contend the drug was defectively designed before it could ever legally be produced commercially.

Got it?

Thankfully, this “pre-approval” defect concept hasn’t done all that well, even as a matter of preemption. In Yates the Sixth Circuit, the first appellate court to pass on such a claim, accurately rejected it as another variant of a preempted claim that the defendant should never have sold its product.

In contending that defendants’ pre-approval duty would have resulted in a [product] with a different formulation, [plaintiff] essentially argues that defendants should never have sold the FDA-approved formulation of [their drug] in the first place. We reject this never-start selling rationale for the same reasons the Supreme Court in Bartlett rejected the stop-selling rationale of the First Circuit.

808 F.3d at 800; accord Utts, 2016 WL 7429449, at *11; Brazil, 196 F. Supp.3d at 1364 (subjecting pre-approval design defect claims for preemption for similar reasons.

However, a few recent cases from Fifth Circuit turf have let pre-approval design defect claims escape preemption. See In re Xarelto (Rivaroxaban) Products Liability Litigation, 2017 WL 1395312, at *3 (E.D. La. Apr. 13, 2017) (“Louisiana law imposes a duty on all manufacturers to consider feasible, alternative designs. . . . Federal law does not prevent a drug manufacturer from complying with this state-imposed duty before seeking FDA approval.”) (following Guidry v. Janssen Pharmaceuticals, Inc., 206 F. Supp.3d 1187, 1206-97 (E.D. La. 2016)); see also Young v. Bristol-Myers Squibb Co., 2017 WL 706320, at *8 (N.D. Miss. Feb. 22, 2017) (“there is no conflict between [plaintiff’s] pre-approval theory and the defendants’ federal law duties”) (also following Guidry).

We, of course, think Yates nailed it on preemption – any common-law claim, the result of which would be a jury finding that an FDA-approved product design should never have been sold, is a stop-selling claim barred by Mensing/Bartlett. The FDA determines what products may be marketed, not individual juries misled by reptile-minded plaintiffs’ lawyers.

But this post isn’t about that – it is not another defense of preemption. Another thing Yates had to say about pre-approval design defect claims was:

[Plaintiff’s] argument regarding defendants’ pre-approval duty is too attenuated. To imagine such a pre-approval duty exists, we would have to speculate that had defendants designed [the drug] differently, the FDA would have approved the alternate design. Next, we would have to assume that [plaintiff] would have selected this [hypothetical product]. Further yet, we would have to suppose that this alternate design would not have caused [plaintiff’s injuries]. This is several steps too far. Even if New York law requires defendants to produce and market a different design, the ultimate availability to [plaintiff] is contingent upon whether the FDA would approve the alternate design in the first place.

808 F.3d at 299. Thus, the Sixth Circuit was “unable to conceive of any coherent pre-approval duty that defendants would have owed to [plaintiff] when it was developing” the product. Id. at 300. See also Young, 2017 WL 706320, at *8 (“the parties have not argued whether Mississippi law recognizes a pre-approval claim, and the Court does not reach the issue”).

The reason that Yates (and, apparently, Young) was “unable to conceive of” a state law pre-approval duty is because such duties do not exist. Design defects under Restatement (Second) of Torts §402A (1965), do not suffer from the “attenuation”/”speculation” problem identified in Yates because §402A is limited to products that are defective at sale. “The rule stated in this Section applies only where the product is, at the time it leaves the seller’s hands, in a condition . . . which will be unreasonably dangerous” to the ultimate consumer. Restatement (Second) of Torts § 402A, comment g (1965).

In the Third Restatement, this time in the black letter, rather than the comments, all “categories of product defect” are likewise determined “at the time of sale or distribution.” Restatement (Third) of Torts, Products Liability §2 (1998). The comments reinforce this view. “[F]or the liability system to be fair and efficient, the balancing of risks and benefits in judging product design and marketing must be done in light of the knowledge of risks and risk-avoidance techniques reasonably attainable at the time of distribution.” Id., comment a. “[T]he plaintiff must prove that such a reasonable alternative was, or reasonably could have been, available at time of sale or distribution.” Id. comment c. Similarly the black letter of Restatement Third §6(b), specifically applicable to prescription medical products, expressly measured defectiveness – including design defect, to the extent allowed at all − “at the time of sale or other distribution.”

Statutory product liability schemes are generally similar to the common law stated in the Restatements with respect to when defectiveness is measured. Since both Xarelto and Guidry (on which it almost exclusively relied) are from Louisiana, we looked up the equivalent provision of the Louisiana Product Liability Act, which for design defects provides:

A product is unreasonably dangerous in design if, at the time the product left its manufacturer’s control: (1) There existed an alternative design for the product that was capable of preventing the claimant’s damage. . . .

La. Stat. Ann. §9:2800.56 (emphasis added). Thus, under Louisiana law, an available alternative design must exist “at the time the product left its manufacturer’s control.” E.g., Reynolds v. Bordelon, 172 So. 3d 607, 614 (La. 2015) (“the plaintiff was first required to show an alternative design for the [product] existed at the time it left [defendant’s] control”); Roman v. Western Manufacturing, Inc., 691 F.3d 686, 700–01 (5th Cir. 2012) (“the statute required [plaintiff] to prove (i) that an alternative design existed at the time [defendant] manufactured the [product]”) (applying Louisiana law).

Thus, quite apart from preemption, there is no common-law claim for a product that became defective at some time – years, perhaps decades, before the product itself was sold – when the design was first submitted to a government regulator like the FDA. This essentially universal common-law requirement exists, as the Restatement Third discussed, to ensure that defendants are judged by the state of the art existing at the time of manufacture, not some other time way later, or presumably way earlier (although we suspect that the Restatements’ drafters were, like the court in Yates, “unable to conceive of” something as bizarre as a purported duty to redesign a product years before it had ever been sold to anyone).

In order to avoid this result, Guidry was forced to ignore the express terms of the Louisiana statute – “at the time the product left its manufacturer’s control.” Contrary to what the Louisiana legislature had mandated, Guidry replaced “product” with “design” – specifically “chemical composition”:

Defective design claims are supposedly preempted because the drug manufacturer loses control to alter the chemical composition of the drug once the FDA approves it. Application of the defendants’ preemption theory necessarily entails that the drug “leaves the manufacturer’s control” when the FDA approves it, not when it is sold to consumers. Consequently, the “unreasonably dangerous” analysis in the defective design context necessarily occurs pre-FDA approval (the only period in which the drug manufacturer has control over the drug’s design).

206 F. Supp.3d at 1208. As a regulatory matter, that proposition is simply false. A manufacturer still has “control” of product design. It can file what’s called a “supplement” to its NDA at any time to change a design. However, the FDA gets to evaluate the supplement first, before it can go into effect – and that triggers preemption.

But for present purposes, note how Guidry put the rabbit in the hat. It truncated its quotation of the statute – starting just after the legislature’s operative term “the product.” The LPLA even defines “product”:

(3) “Product” means a corporeal movable that is manufactured for placement into trade or commerce, including a product that forms a component part of or that is subsequently incorporated into another product or an immovable.

La. Stat. Ann. §9:2800.53(3). A “product” is thus “corporeal” – it is not merely its “design,” nor is it just its “chemical composition.” Guidry never acknowledges this statutory definition. “Corporeal” nowhere appears in that extremely long opinion.

Thus Guidry conveniently omitted what the legislature in fact enacted. Only by substituting “design”/“chemical composition” for “product” as defined by the LPLA could Guidry advance to its next remarkable proposition: that in the case of all FDA-approved products, Louisiana’s (or presumably some other state’s) “’unreasonably dangerous’ analysis in the defective design context necessarily occurs pre-FDA approval.” 206 F. Supp.3d at 1208 (emphasis added).

“Necessarily”? Come on, now.

What does Guidry cite for this remarkable proposition that juries must determine the defectiveness of an FDA-approved design at a point before it was ever approved by the FDA – that is, many years before it was ever used by any plaintiff?

Nothing at all. Zilch. Not a single statute. Not a single case. Guidry made it up.

As we have stated many times before, for a federal court to invent new state law, expanding liability where the state’s courts and lawmakers have not gone, is a serious violation of federalism under Erie v. Tompkins. “As always, in conducting [an Erie] inquiry our task is ‘to predict state law, not to create or modify it’ − that is, we are ‘to apply existing Louisiana law, not to adopt innovative theories for the state.’” Holden v. Connex-Metalna Management Consulting GmbH, 302 F.3d 358, 365 (5th Cir. 2002) (quoting United Parcel Service, Inc. v. Weben Industries, Inc., 794 F.2d 1005, 1008 (5th Cir. 1986)). As the en banc Fifth Circuit explained thirty years ago:

As a federal court, it is not for us to adopt innovative theories of state law, but simply to apply that law as it currently exists. . . . We are emphatically not permitted to do merely what we think best; we must do that which we think the [state’s] Supreme Court would deem best. Finally, under Erie we cannot skirt the clear import of state decisional law solely because the result is harsh.

Jackson v. Johns-Manville Sales Corp., 781 F.2d 394, 397 (5th Cir. 1986) (en banc) (citations and quotation marks omitted), overruled in part on other grounds, Salve Regina College v. Russell, 499 U.S. 225 (1991) (rejecting appellate deference to in-state district court Erie predictions).

As Jackson “emphatically” held, “under Erie we cannot skirt the clear import of state decisional law solely because the result is harsh.” That’s exactly where Guidry erred. Before mangling Louisiana’s statutory defect-at-sale requirement, Guidry complained about a harsh result:

The Court first notes that, if it finds the plaintiff’s defective design claim is preempted, even under a pre-FDA approval theory, the result is that a Louisiana plaintiff can never bring a defective design claim against a drug manufacturer. . . . And no federal remedy exists either. . . . As a result, if the defendants’ preemption argument prevails, Louisiana plaintiffs will have no remedy against a drug manufacturer for a defect in a drug’s design.

206 F. Supp. at 1206-07 (Levine quotation omitted) (emphasis original). Thus, the Erie error in Guidry is no accident. The decision deliberately flouted the law to avoid a result it didn’t like. For the hundredth time we’ll say, strange things happen in tort preemption cases. In Guidry, that strangeness was the invention out of whole cloth of a novel “pre-approval” design defect that is flatly inconsistent with Louisiana’s defect-at-sale requrement for design defects.

Not only that, but the contortions that this novel idea of “pre-approval” design defects require of the well-established defect-at-sale requirement (both statutory and common-law) will have unintended consequences. The LPLA, similarly to the common-law as stated in the Second and Third Restatements, imposes the exact same “at the time the product left its manufacturer’s control” limit on warning defects. La. Stat. Ann. §9:280.57(A). Warnings are FDA approved, too, so if defect analysis “necessarily occurs,” Guidry, 206 F. Supp.3d at 1208, prior to FDA approval, consider whether a plaintiff should be able to “claim[] that the defendants intentionally concealed or downplayed the seriousness and likelihood of these adverse side effects” during post-approval promotional activities. Id. at 1199. Guidry found nothing amiss with the warning claims, id., but unless we are dealing with a “heads plaintiffs win; tails defendants lose” situation, the defect at sale requirement – whether statutory or common-law – has to run from the same date for both warning and design claims. And on warning claims, that absurdly early date favors defendants.

In sum – forgetting about preemption − the notion of a “pre-approval” design defect is a non-starter under state law, and rightly so. The legal requirement that a product be defective at the time of distribution is a bedrock product liability principle in all states, and as just discussed, that requirement is utterly incompatible with plaintiffs’ new defect theory that pushes the defect analysis earlier by years if not decades. This novel theory is being asserted solely to avoid the FDA-pre-approval trigger for preemption first recognized in Wyeth v. Levine, 555 U.S. 555 (2009). But federal courts under Erie are not supposed to make up new state law just because Levine’s preemption analysis happens to require broad preemption in design defect cases.