Photo of Andrew Tauber

This post is from the non-Reed Smith side of the blog.

This blog has repeatedly lamented the tendency of MDL courts to flout federal pleading standards when assessing the sufficiency of master complaints. All too often MDL courts disregard Rule 8(a), which—as authoritatively interpreted by the Supreme Court in Twombly and Iqbal—requires plaintiffs to plead facts plausibly suggesting an entitlement to relief, and Rule 9(b), which requires that fraud be alleged with particularity. Today we report on another example of this unfortunate tendency, In re: Allergan Biocell Textured Breast Implant Product Liability Litigation, 2021 WL 1050910 (D.N.J. Mar. 19, 2021). Although the court got a number of things right and dismissed a few claims from the master complaint, it allowed all too many of the claims to proceed on frustratingly familiar grounds that effectively insulate MDL master complaints from many motions to dismiss.

As its full name indicates, In re: Allergan involves the manufacturer’s Class III textured breast implants and Class II textured tissue expanders, which were voluntarily recalled after evidence suggested that they cause a certain form of cancer at a higher rate than other textured breast implants. The plaintiffs allege that the process by which the implants were manufactured resulted in “overly aggressive and inconsistent texturing” that in turn allegedly increased the implants’ surface area and led particles to form on their surface. The plaintiffs claim that these purported conditions caused them to suffer cancer or be at a higher risk of suffering cancer. Based on those allegations, the plaintiffs assert failure-to-warn, manufacturing-defect, negligence-per-se, breach-of-warranty, misrepresentation, and consumer-fraud claims under the state laws of various states.

The manufacturer moved to dismiss the plaintiffs’ master complaint, moving to dismiss all claims on state-law grounds and moving to dismiss the claims implicating Class III devices on preemption grounds as well. The motion to dismiss on preemption grounds targeted claims involving devices that had received premarket approval (PMA) from the FDA and devices that that had been cleared by the agency under the Investigational Device Exemption. Applying unduly lenient pleading standards, the court granted the motion in small part and denied it in large part. The court justified its application of a lax pleading standard on two grounds.

First, the court refused to consider whether the master complaint adequately pleaded certain elements of the plaintiffs’ claims, precisely because those claims were asserted in an MDL master complaint. For example, the court declared that it “need not review the factual sufficiency of Plaintiffs’ negligent misrepresentation allegations,” explaining that assessing their “factual sufficiency under the potentially varying state laws of negligent misrepresentation would be both cumbersome and unrealistic at this stage, especially when individual Plaintiffs may allege separately in their Short Form Complaints [the] misrepresentations to which they each [supposedly] have been exposed.” 2021 WL 1050910, at *32. Indeed, time and again the court “decline[d] to scrutinize … at this stage” of the proceedings whether the master complaint alleged facts sufficient to establish elements of the plaintiffs’ claims. Id. at *42. This is true with respect to claims subject to Rule 8(a) as well as claims subject to Rule 9(b). Simply put, the court would not apply the otherwise applicable pleading requirements because the complaint at issue was a master complaint in an MDL. When one compares the result in the In re: Allergan MDL to the opposite result in D’Addario v. Johnson & Johnson, 2021 WL 1214896 (D.N.J. 2021), a one-off case raising similar allegations against another manufacturer’s textured breast implants, the MDL effect is cast in sharp relief.

The court’s second justification for ignoring basic federal pleading standards is in some sense even more disturbing because it would by its logic apply to all cases, not just MDLs. According to the court, “fairness compels that some leniency be afforded plaintiff[s] from the stringent Twombly/Iqbal pleading standards to allow [their manufacturing-defect] claim to proceed” because the plaintiffs “do[] not have access to” the relevant PMAs, which are confidential, and are thus supposedly unable to plead a specific federal violation, as is necessary to avoid express preemption under 21 U.S.C. § 360k(a). 2021 WL 1050910, at *13 (internal quotation marks omitted). That of course echoes the Seventh Circuit’s abominable decision in Bausch v. Stryker Corp., 630 F.3d 546 (7th Cir. 2010), which we have pilloried many times, including here and here. And, as in Bausch, the court’s analysis fails to appreciate that both Twombly and Iqbal involved situations in which the plaintiffs did not have access to information within the defendant’s control. In Twombly, the plaintiffs asserted an antitrust claim that required them to prove, and thus to allege, a conspiracy among the defendants. Despite the defendants controlling the information that would be necessary to establish a conspiracy, the Supreme Court held that the plaintiffs were required to plead facts sufficient to plausibly suggest an entitlement to relief. Iqbal confirmed that requirement, holding that a plaintiff who brought a Bivens action had to plead facts sufficient to plausibly suggest that the defendant acted with “discriminatory intent,” even though the defendant’s mental state was known only to the defendant. In short, Twombly and Iqbal hold that a plaintiff must plead facts sufficient to state a claim even where the plaintiff is at an informational disadvantage. Like Bausch before it, In re: Allergan cannot be reconciled with this Supreme Court precedent.

You know that things are not going to go well from a defense perspective when a court ignores Twombly and Iqbal. And, sure enough, there is much to dislike in In re: Allergan.

That said, the court did get some things right.

Rejecting the plaintiffs’ contrary contention, the court held that claims implicating devices used pursuant to the Investigational Device Exemption (IDE) are subject to express preemption under 21 U.S.C. § 360k(a). 2021 WL 1050910, at *8.

Another bright spot in an otherwise dreary decision is the court’s recognition that the Changes Being Effected (CBE) regulation, which allows device manufacturers to change a device label under certain circumstances without receiving prior FDA approval, “is permissive, not mandatory”—and that a state-law failure-to-warn claim that would require a manufacturer to have used the CBE process to change its label is therefore preempted under § 360k(a) because it would impose “a state-law duty that differs from or adds to the federal requirements.” 2021 WL 1050910, at *9–10 (internal quotation marks omitted). On this basis, the court dismissed the plaintiffs’ failure-to-warn claims insofar as they rested on the allegation that the manufacturer failed to revise its labeling to warn of the purportedly enhanced risk of cancer.

Satisfying too is the court’s recognition that § 360k(a) preempts claims implicating a device that had originally been classified as a Class II device but was subsequently reclassified as a Class III device and given premarket approval. As the court put it, “claims against … reclassified devices, which had the PMA approval when used by Plaintiffs, are treated no differently from the claims against the” devices that were PMA-approved from the outset. 2021 WL 1050910, at *14.

Finally, the court correctly concluded that the plaintiffs’ failure-to-warn claims were expressly preempted insofar as they rested on the manufacturer’s alleged failure “to conduct post-PMA clinical studies,” because “there is no state law duty that requires [a manufacturer] to undertake [such] studies.” 2021 WL 1050910, at *15.

Apart from that and the dismissal of a various claims under the laws of various states, the decision isn’t great, but it does provide ample fodder for this blogpost.

Different models of breast implants are at issue in the In re: Allergan MDL. They were distributed through different legal pathways. The great majority received PMA; a few others were IDE devices; yet others (empty implants used as “tissue expanders”) were cleared through the 510(k) process. As noted above, the court recognized that claims implicating IDE devices, like those implicating PMA devices, are subject to preemption under § 360k(a). Although the manufacturer did not seek dismissal of claims implicating the tissue expanders on preemption grounds, the court nevertheless addressed preemption in the context of 510(k) devices. Without any analysis beyond an indirect citation to the ill-conceived and outdated decision in Medtronic, Inc. v. Lohr, 518 U.S. 470 (1996), the court said in dicta that claims implicating 510(k) devices are not subject to preemption.  As the blog has noted before, that conclusion is dubious not only because the 510(k) today is significantly different from the process by which the Lohr device was cleared, but also given PLIVA v. Mensing, 564 U.S. 604 (2011), and Mutual Pharmaceutical Co. v. Bartlett, 570 U.S. 472 (2013), which suggest that some claims implicating 510(k) devices are at least impliedly preempted even if not expressly preempted.

While the court recognized that claims implicating PMA and IDE devices are in principle subject to preemption under § 360k(a), it took an exceedingly narrow view of when claims are preempted.

Citing a series of unpublished Ninth Circuit decisions that mechanically follow that court’s misguided decision in Stengel v. Medtronic, Inc., 704 F.3d 1224 (2013), while at the same time ignoring Mink v. Smith & Nephew, Inc., 860 F.3d 1319 (11th Cir. 2017), and other precedential appellate decisions that have rejected such claims, the court held that failure-to-warn claims predicated on a manufacturer’s alleged failure to file adverse-event reports with FDA are neither expressly nor impliedly preempted

In so concluding, the court disregarded the nature of adverse-event reports, misconstrued Restatement (Second) of Torts § 388 cmt. n, and adopted an unduly restrictive view of Buckman Co. v. Plaintiffs’ Legal Committee, 531 U.S. 341 (2001). The court failed to recognize that, as explained in Aaron v. Medtronic, Inc., 209 F. Supp. 3d 994, 1005 (S.D. Ohio 2016), adverse-event reports are not warnings. The court compounded that error by misreading the Restatement as imposing a state-law duty to submit adverse-event reports to the FDA. Comment n to § 388 says that a manufacturer can satisfy its duty to warn a product’s end-user by warning a third-party “through whom the [product] is supplied” when the manufacturer can reasonably rely on the third-party to convey the warning to the end-user, such as when the third-party has a duty to do so. In treating the FDA as a third-party intermediary for purposes of § 388, the court ignored the obvious fact that medical products such as breast implants are not supplied by the FDA, the fact that the FDA is not obligated to make adverse-event reports public, and the fact that adverse-event reports are not actively distributed to doctors even when made public. Finally, declaring itself “bound by the Third Circuit’s interpretation of the holdings in Buckman,” the court—citing no Third Circuit decision, when the only Third Circuit precedent, Sikkelee v. Precision Airmotive Corp., 907 F.3d 701, 716–17 (3d Cir. 2018), supports preemption—suggests that failure-to-warn claims predicated on a manufacturer’s alleged failure to file adverse-event reports with FDA are not preempted under Buckman because Buckman’s holding is, supposedly, limited to fraud-on-the-FDA claims. 2021 WL 1050910, at *11. That suggestion ignores the various cases that have characterized such failure-to-warn claims as fraud-on-the agency claims and cannot be reconciled with the Supreme Court’s own understanding of Buckman, which, the Court has explained, held that the FDCA preempts any “state tort-law claim based on failure to properly communicate with the FDA.” PLIVA, Inc. v. Mensing, 564 U.S. 604, 619 (2011).

The In re: Allergan court also missed the mark when analyzing whether plaintiffs’ manufacturing-defect claims are preempted. As noted at the outset, the court—ignoring Twombly and Iqbal in the name of “fairness”—excused the plaintiffs’ failure to identify a specific PMA requirement that the manufacturer allegedly violated. Moreover, it held that the plaintiffs could base their manufacturing-defect claims on the manufacturer’s alleged violation of an FDA Current Good Manufacturing Practice (CGMP). 2021 WL 1050910, at *13. There is conflicting law on that point, and the court’s conclusion is arguably contrary to Irizarry v. Abbott Laboratories, 833 F. App’x 947, 949–50 (3d Cir. 2020), in which the Third Circuit affirmed the dismissal of a manufacturing-defect claim, finding that a complaint that “d[id] not set forth the premarket approval requirements” that were allegedly violated “d[id] not plausibly allege” a parallel claim that survived preemption.” This blog has repeatedly argued that because the CGMPs are intentionally vague and designed to give manufacturers complete discretion in how they are to be implemented, a state-law claim based on an alleged CGMP violation necessarily imposes a state-law duty that is “different from, or in addition to” the federal requirements and is thus expressly preempted under 21 U.S.C. § 360k(a).

But even if a CGMP violation could in theory support a manufacturing-defect claim, it is hard to see how 21 C.F.R. § 820.30(g), the CGMP cited by the In re: Allergan court, could sustain a non-preempted claim. To start, § 820.30(g) addresses “design validation,” not manufacturing processes. Furthermore, its requirement that a manufacture conduct “testing of production units under actual or simulated use conditions” is—to the knowledge of this blogger—not found in the law of any state. If it is not, and the court cited no state’s laws to suggest otherwise, then a claim based on an alleged violation of § 820.30(g) would be both expressly preempted, because the state and federal requirements would not be “identical” (Medtronic, Inc. v. Lohr, 518 U.S. 470, 495 (1996)), and impliedly preempted, because the “existence of the[] federal enactment is a critical element in [the plaintiffs’] case.” Buckman, 531 U.S. at 353.

As if these problems weren’t enough, the court’s conclusion that the plaintiffs’ manufacturing-defect claims avoid preemption disregards the essence of their master complaint. To state a manufacturing-defect claim, one must allege and ultimately prove that the particular unit received by the plaintiff differed from its intended design or from other ostensibly identical units. The In re: Allergan plaintiffs, however, do not allege that their textured breast implants differed from their intended design or from other ostensibly identical units. On the contrary, the plaintiffs contend that every Biocell breast implant was defective because the manufacturing process supposedly resulted in the formation of particles and excessive surface areas—and even assert a class action on behalf of all recipients. Thus, plaintiffs’ real complaint is with the manufacturing process as designed and as approved by the FDA. That is to say, the plaintiffs’ manufacturing defect claims “are a frontal assault on the FDA’s decision to approve” the device’s PMA “after weighing the product’s benefits against its inherent risks.” In re Medtronic, Inc., Sprint Fidelis Leads Prod. Liab. Litig., 623 F.3d 1200, 1207 (8th Cir. 2010).

By this point it should not surprise the reader that the In re: Allergan court denied the manufacturer’s motion to dismiss the plaintiffs’ negligence-per-se claims on preemption grounds. What is particularly depressing about the court’s ruling in this regard is that it relied on In re Orthopedic Bone Screw Product Liability Litigation, 193 F.3d 781 (3d Cir. 1999), to support its conclusion. But that decision is part of the same litigation that ended in Buckman, which squarely held that 21 U.S.C. § 337(a) impliedly preempts any state-law claim for which the existence of the FDCA “is a critical element.” 531 U.S. at 353. Given that the plaintiffs’ negligence-per-se claims rest on the defendants’ alleged violation of the FDCA and its implementing regulations, Buckman plainly precludes such claims (as many but not all courts have held). The In re: Allergan court went astray when it concluded that the plaintiffs’ negligence-per-se claims were not preempted under Buckman because they “invoke the statutory violations to prove defendants’ liability for a separate underlying tort, instead of contending the violations themselves form a cause of action.” 2021 WL 1050910, at *14 (quotation marks omitted). But the mere fact that states recognize negligence actions does not save negligence-per-se claims from preemption. Rather, to avoid preemption under Buckman, “the conduct on which the claim is premised must be the type of conduct that would traditionally give rise to liability under state law—and that would give rise to liability under state law even if the FDCA had never been enacted.” Riley v. Cordis Corp., 625 F. Supp. 2d 769, 777 (D. Minn. 2009). A negligence-per-se claim predicated on a violation of the FDCA and its implementing regulations does not satisfy that test.

Despite getting most of the preemption analysis wrong, the In re: Allergan court did ultimately dismiss some of the plaintiffs’ failure-to-warn, negligence-per-se, negligent-misrepresentation, warranty, and medical-monitoring claims on state-law grounds, concluding (after providing useful multi-state surveys) that some states do not recognize the plaintiffs’ theories of liability. That is a good reminder that dismissal on state-law grounds is possible even when a court gets the preemption analysis wrong and that counsel should not rely exclusively on preemption when seeking dismissal of arguably preempted claims.

But the In re: Allergan court let most of the claims (and class allegations) through after declaring once again that it would review the sufficiency of plaintiffs’ allegations “with leniency.” 2021 WL 1050910, at *18. That’s an MDL for you.