Photo of Bexis

One of the more peculiar things about Merck Sharp & Dohme Corp. v. Albrecht, 139 S. Ct. 1668 (U.S. 2019) (“Albrecht”), is the almost off-hand way that the majority (made up mostly of justices that have opposed preemption in closer cases) wandered away from the procedural preemption issues that the Court was actually deciding (whether the jury has any role in preemption decisions and whether a “clear and convincing” evidence standard applied), to redefine what had been known (after Wyeth v. Levine, 555 U.S. 555 (2009) (“Levine”)) as the “clear evidence” standard.  They did so even after expressly cautioning up front that “The question of disapproval ‘method’ is not now before us.”  Albrecht, 139 S. Ct. at 1679.  So we now have this:

[I]n [Levine], we confronted that question in the context of a particular set of circumstances. . . .  In a case like [Levine], showing that federal law prohibited the drug manufacturer from adding a warning that would satisfy state law requires the drug manufacturer to show that it fully informed the FDA of the justifications for the warning required by state law and that the FDA, in turn, informed the drug manufacturer that the FDA would not approve changing the drug’s label to include that warning.

Id. at 1687 (to avoid confusion, the Blog’s naming conventions use plaintiff names for important cases when the defendant’s name is common; the Supreme Court uses whichever is first in the caption).  At the same time, the Court also described the relevant test much more simply − whether “the relevant federal and state laws ‘irreconcilably conflic[t].’”  Id. at 1679

There are quite a few issues with Albrecht’s longer “definition” – if that’s what it is.  First, in light of the Court’s reminder about Levine’s “particular set of circumstances,” what is a case “like” Levine?  Plaintiffs trying to say that every fact pattern is “like” Levine have failed in the early going.  As previously mentioned, the Tenth Circuit in Cerveny v. Aventis, Inc., ___ F. Appx. ___, 2019 WL 3763441 (10th Cir. Aug. 9, 2019), rejected the plaintiffs’ argument that only a “manufacturer’s” actions could be considered under this definition with the retort that a case involving a third-party citizen petition inducing FDA action was not “like” Levine.

[Plaintiffs] contend that Albrecht “dictates that only labeling changes sought by the manufacturer can lead to preemption  . . .  But we note that Albrecht prefaced its requirement that “[the drug manufacturer] fully informed the FDA . . .” as applying “[i]n a case like [Levine], and noted that “in [Levine, [the Court] confronted [preemption] in the context of a particular set of circumstances.”  And here, [plaintiffs’] case involves a particular set of circumstances with a key difference. . . .  Here, [defendant] argues a different ground . . . − the FDA’s unequivocally having rejected [a] citizen petition advocating for the warning that [plaintiffs]now assert.  We see nothing in [Levine] or excluding [defendant] from justifying preemption on this basis.

Id. at *3 n.9 (citations omitted).

A second court, in North Dakota, recently rejected similar plaintiff-side arguments as “nonsensical” for similar reasons:[Plaintiff’s]

[Plaintiff] argues that, under [Albrecht], the FDA’s decision not to change a medication’s labeling triggers [Levine] preemption only if the manufacturer itself proposed the change. . . .  This Court does not read [Albrecht] so narrowly.  [Plaintiff’s] argument ignores [Albrecht’s] explicit warning that “[t]he question of disapproval ‘method’ is not now before us.”  . . .  [Albrecht] did not involve agency “actions” or “methods” similar to this case. . . .  [Albrecht] certainly did not decide whether other agency actions, such as the denial of a Citizen Petition, would be sufficient. . . .  Where the FDA has already rejected a proposed label or warning change, as it did in this case, there is no reason why the “impossibility” of a subsequent change would depend on whether the earlier change was proposed by the manufacturer as opposed to a third party or the FDA itself.  Regardless of who submitted the proposed warning or labeling change, the FDA has already decided that the relevant evidence and policies do not meet the standard to justify a change.

State v. Purdue Pharma L.P., 2019 WL 3776653, at *2-3 (N.D. Dist. July 22, 2019).

Another issue (among several) stemming from Albrecht’s definitional dictum, as to which no subsequent precedent yet exists, is what does it mean for “the drug manufacturer to show that it fully informed the FDA of the justifications” for not using the warning that the plaintiff now wants.  The rest of this post is devoted to that question.

In light of the Albrecht Court’s longer definition – as opposed to its basic “irreconcilable conflict” statement – we expect plaintiffs to make arguments that squarely implicate the Court’s prior decision in Buckman Co. v. Plaintiffs Legal Committee, 531 U.S. 341 (2001).  Buckman employed conflict preemption to bar agency fraud claims against FDA-regulated products, largely for institutional reasons:  (1) the FDCA explicitly prohibits private parties from enforcing the FDCA, id. at 349 n.4; (2) states have not “traditionally” been in the business of “policing” the adequacy of regulated entities’ submissions to the FDCA, id. at 347; and (3) FDA is “empowered to require additional necessary information,” id. at 349; and (4) state-law second-guessing the sufficiency of submissions to the FDA:

would also cause applicants to fear that their disclosures to the FDA, although deemed appropriate by the Administration, will later be judged insufficient in state court.  Applicants would then have an incentive to submit a deluge of information that the Administration neither wants nor needs, resulting in additional burdens on the FDA’s evaluation of an application.

Id. at 351.

Allegations based on “counterfactual” claims involving what the FDA might have done with different, supposedly better, information have always (except for the Third Circuit decision reversed in Albrecht) been recognized as preempted:

Effectively, actions for fraud on the FDA would allow individual juries to undertake a counterfactual FDA review, and conclude that the FDA would not have approved the [product].  However, Congress allocated the FDA responsibility to design and manage a process which would result in approval of the safest and most effective medical devices possible.  Congress also assigned the FDA the responsibility to approve or disapprove of applications to market medical devices.  Furthermore, we agree that permitting the “searching inquiry” of FDA internal procedures and personnel runs counter to both congressional intent and sound policy.  Hence, we hold that to allow common-law suits for fraud on the FDA to go forward would permit juries to reach a different conclusion than the FDA did in approving the device in question, thereby imposing a different requirement than that required by the federal requirements.

Kemp v. Medtronic, Inc., 231 F.3d 216, 235-36 (6th Cir. 2000) (pre-Buckman decision reaching same result).  See Lefaivre v. KV Pharmaceutical Co., 636 F.3d 935, 943 (8th Cir. 2011) (claims that “depend upon speculation as to the FDA’s behavior in a counterfactual situation” “second-guess[] the FDA’s decisionmaking [and] overburden[] its personnel”) (quoting Buckman, 531 U.S. at 354 (concurring opinion)).  There is “no support for the theory that it is appropriate for a court . . . to determine preemptively how a federal administrative agency will interpret and enforce its own regulations.”  Sandoz Pharmaceuticals Corp. v. Richardson-Vicks, Inc., 902 F.2d 222, 231 (3d Cir. 1990).

“Allowing juries to perform their own risk-utility analysis and second-guess the [agency’s] conclusion would disrupt the expert balancing underlying the federal scheme.”  Farina v. Nokia Inc., 625 F.3d 97, 126 (3d Cir. 2010) (applying Buckman outside FDCA context).  As a result, Buckman in particular, and the FDCA scheme in general, have almost universally been held to preclude state-law plaintiffs from looking behind FDA actions and making claims that the FDA’s basis for acting was, for one reason or another, insufficient.  Surprisingly, neither the majority nor any of the concurring opinions in Albrecht even mention Buckman – another sign that Albrecht’s dictum was not well thought out.

For example, in Riley v. Cordis Corp., 625 F. Supp.2d 769 (D. Minn. 2009), the plaintiff contended that “preemption only applies when the FDA has regulated a device with full knowledge of that device’s safety and efficacy.”  Id. at 779.  After pointing out that not a word in the Riegel v. Medtronic, Inc., 552 U.S. 312 (2008), express preemption case supported such a contention, the court held that preemption does not “turn[] on the nature or extent of the information made available to the FDA at the time it approved a device,” and relied on Buckman:

[Plaintiff’s] argument that his claims are not preempted because the FDA was fraudulently deprived of the knowledge it needed to make an informed decision about the [product] is just a thinly veiled way of claiming that the FDA would not have approved the [product] in the absence of fraud − a claim that is squarely foreclosed by Buckman.

Id. (citation omitted).  See In re Baycol Products Litigation, 532 F. Supp.2d 1029, 1053 (D. Minn. 2007) (evidence “offered only to show that the FDA was misled, or that information was intentionally concealed from the FDA . . . must be excluded” under Buckman).

A prescription drug plaintiff’s claim that “if [defendant] had funded or conducted studies that generated subsequently published data, those studies would have generated the same data sooner,” and the FDA would not have rejected the warning they sought, was dismissed as both “speculative” and preempted in Rheinfrank v. Abbott Laboratories, Inc., 119 F. Supp.3d 749 (S.D. Ohio 2015), aff’d, 680 F. Appx. 369 (6th Cir. 2017).

Plaintiffs support their position with proffered expert testimony of four experts who have reviewed and opined on allegedly misleading information in Abbott’s [various] . . . correspondences with the FDA.  Plaintiffs’ argument . . . appears to be a fraud-on-the-FDA theory, which is preempted. . . .  Regardless, an expert’s opinion that the FDA would have reacted differently if the submissions to the FDA . . . had been supported by different evidence is speculative. . . .  Testimony about what [defendant] could have and should have researched or stated to the FDA in its applications is speculative, and does not serve as the basis for a genuine issue of material fact.  Thus . . ., the Court finds that Plaintiffs’ fraud-on-the-FDA theory is either preempted or based on speculation.

Id. at 767-68 (citations omitted).

On appeal, the Sixth Circuit determined that “[t]he district court’s preemption analysis was correct.”  Rheinfrank v. Abbott Laboratories, Inc., 680 F. Appx. 369, 371 (6th Cir. 2017).

[Plaintiff] nevertheless contends that, even if the FDA signaled that it would not have approved a stronger . . . warning on [the product’s] label, a finding of “clear evidence” would still be precluded, either because [defendant] misrepresented the state of the evidence to the FDA in its communications about its proposed labeling changes, or because there was evidence about [the product’s] effects . . . that [defendant] should have known but failed to obtain.  Each of these arguments, however, is unavailing.  First, even if the evidence that [defendant] submitted was as misleading or incomplete . . ., the evidence in the record establishes that the FDA undertook its own review of the relevant empirical literature. . . .  Thus, whatever may have been [defendant’s] omissions or misrepresentations in its communications with the agency, the FDA clearly explained that, in its own view of the data, the . . .  warning . . . was not yet warranted. . . .  [F]urther . . . refus[al] to fund or conduct studies . . . is too conjectural to defeat preemption. . . .  [Plaintiff] asks this court to accept[] a series of speculations as to what the FDA could have done with different evidence that [defendant] might have collected if it had run its own studies.

Id. at 387-88 (citations omitted).  See Willis v. Abbott Laboratories, 2017 WL 5988215, at *8 (W.D. Ky. Dec. 1, 2017) (following Rheinfrank; “any speculation as to what the FDA may have done had [defendant] performed certain testing remains just that: speculation.  Therefore, the Court rejects this argument.”); In re Depakote, 87 F. Supp.3d 916, 922 (S.D. Ill. 2015) (claim that defendant “deliberately omitted” information from FDA submissions was “unavailing” to prevent preemption).

Rheinfrank relied on In re Fosamax (Alendronate Sodium) Products Liability Litigation, 2014 WL 1266994 (D.N.J. March 26, 2014), which just happens to be the same decision reversed by the Third Circuit decision that was in turn unanimously overturned by the Supreme Court in Albrecht.  While Buckman issues fell out of the Fosamax litigation by the time it reached the Supreme Court (because the plaintiffs did not appeal that issue), their claim attacking the sufficiency of the defendant’s FDA submissions about the warning in question was held preempted:

[W]hile Plaintiffs were able to find an expert to agree with their contention that [defendant] should have acted differently . . ., they did not set forth any evidence which suggests that the FDA thought the same. . . .  Moreover, to the extent that Plaintiffs claim that [defendant] withheld information from the FDA and clear evidence does not exist as to whether the FDA, if fully informed, would have rejected a stronger label, this does not defeat Defendant’s showing that it is entitled to judgment . . . on preemption grounds. . . .  Plaintiffs’ contention appears to be a fraud-on-the-FDA theory which was rejected by the Supreme Court in Buckman.

Id. at *16-17, vacated, 852 F.3d 268 (3d Cir. 2017), reversed, 139 S. Ct. 1668 (2019).  See Id. at *9 (“what [defendant] could have or should have done is immaterial because we know what [it] did”).

In Perez v. Nidek Co., 657 F. Supp.2d 1156 (S.D. Cal. 2009), the plaintiffs attacked the off-label use of certain devices as “adulteration.”  The court held such claims preempted because they “require the Court to make determinations regarding” the product that “should be decided by the FDA in the first instance.”  Id. at 1166.  The Ninth Circuit affirmed because such claims “conflict[ed] with the FDCA’s enforcement scheme.”  Perez v. Nidek Co., 711 F.3d 1109, 1119 (9th Cir. 2013).  “Although citizens may petition the FDA to take administrative action . . ., private enforcement of the statute is barred.”  Id.

The FDA knew about the allegations that the [product] was being used for unapproved . . .  use and took steps to address the allegations . . ., but it did not take final action against the defendants . . . .  All of these matters rest within the enforcement authority of the FDA, not this Court.

Id. at 1120 (quoting, in part, district court decision).

Also, in In re Trasylol Products Liability Litigation, 763 F. Supp.2d 1312 (S.D. Fla. 2010), the court applied Buckman to the plaintiffs’ allegations about the sufficiency of the information provided to the FDA in connection with a prescription drug:

The concerns expressed by the Supreme Court in Buckman hold true not only where there is a separate fraud-on-the-FDA claim but also where a plaintiff seeks to prove fraud on the FDA in order to bring a traditional state-law torts suit.  If the Court were to find fraud-on-the-FDA when the FDA itself has not made such a finding, the Court would be intruding upon the FDA’s right to police itself and second-guessing what the FDA would have done had it received the information that was allegedly withheld from it by the defendant-company.

Id. at 1325.

Claims are preempted under Buckman whenever they “requires a fact finder to make a determination under state law that federal law leaves exclusively to the FDA.”  Zimmerman v. Novartis Pharmaceuticals Corp., 889 F. Supp. 2d 757, 777 (D. Md. 2012).

Allowing state fact finders to second-guess the very decisions that federal law leaves entirely to the agency presents the same inter-branch-meddling concerns that animated Buckman.  As in Buckman, allowing punitive damages liability here would require applicants to submit a “deluge” of unnecessary information during the approval process, which in turn, delays the approval of new drugs.


This issue has also arisen repeatedly in medical device cases enforcing the “narrow gap” between express preemption under Riegel and implied preemption under Buckman.  Plaintiffs may not mount “attacks on the risk/benefit analysis that led the FDA to approve” a product.  In re Medtronic, Inc., Sprint Fidelis Leads Products Liability Litigation, 623 F.3d 1200, 1206 (8th Cir. 2010).  Claims that “depend on speculation that the FDA would have taken any particular regulatory action in response to violation of the regulations at issue, as in Buckman” are preempted.  Hughes v. Boston Scientific Corp., 631 F.3d 762, 775 (5th Cir. 2011).  It is “impermissibl[e to] allow the jury to question the FDA’s decision to approve the device for the market.”  Id. at 776 n.12.  “A challenge to the integrity of FDA’s premarket approval of the Product, however, is itself preempted under Buckman.”  McNeil-Williams v. DePuy Orthopaedics, Inc., 384 F. Supp.3d 570, 578 (E.D.N.C. 2019).  “[A] valid parallel claim cannot challenge the process itself or the requirements imposed by the FDA pursuant to that process.”  Raab v. Smith & Nephew, Inc., 150 F. Supp.3d 671, 687 (S.D.W. Va. 2015).

The thrust of [plaintiffs’] argument is that the FDA could not have granted premarket approval to the [product] because it did not have sufficient data to do so.  However, the sufficiency of the data submitted to the FDA . . . does not govern the scope of the [FDA’s] approval.  Whether a product is FDA-approved is determined by the language in the approval letter, not by the application documents submitted to it for review.  The FDA, not litigants, is entrusted with the responsibility to police the sufficiency of the evidence to support a PMA approval.

Duggan v. Medtronic, Inc., 840 F. Supp.2d 466, 471-72 (D. Mass. 2012) (citing Buckman; other citations omitted).

Plaintiff’s claim of “fraudulent concealment” depends on speculation that the FDA would have taken [a] particular regulatory action in response to violation of the regulations at issue, as in Buckman.  Indeed, Plaintiff argues that “had the FDA classified [defendant’s] recall as Class 1, it would have likely imposed more stringent notice requirements, and expanded the class of recipients of the notice.” . . . Under Buckman, such a claim is preempted.

Bush v. Thoratec Corp., 837 F. Supp.2d 603, 608–09 (E.D. La. 2011) (citation and quotation marks omitted).  See Reed v. St. Jude Medical, 2018 WL 4293146, at *4 (Mag. D. Minn. July 24, 2018) (“foreclos[ing] Plaintiffs’ arguments . . . that the FDA did not adequately study [certain subgroups] as part of the PMA”), adopted, 2018 WL 4251872 (D. Minn. Sept. 6, 2018); Rand v. Smith & Nephew, Inc., 2017 WL 8229320, at *4 (C.D. Cal. April 5, 2017) (plaintiff may not challenge the adequacy of [defendant’s] effort in proving [the product’s] . . . included in the [approval] application”); Gross v. Stryker Corp., 858 F. Supp.2d 466, 487 (W.D. Pa. 2012) (“[t]he application submitted to the FDA does not facilitate this Court’s determination of what the FDA actually approved”); Lewkut v. Stryker Corp., 724 F. Supp.2d 648, 655 (S.D. Tex. 2010) (“What was submitted in the FDA application has little bearing on this Court’s assessment of what was ultimately approved.”); Webster v. Pacesetter, Inc., 259 F. Supp. 2d 27, 37 (D.D.C. 2003) (“what was told to the FDA cannot support a tort claim, and more importantly, plaintiffs’ approach would be nothing more than an invitation for the jury to speculate about what . . . the FDA . . . might do if the facts were different”); Bagumyan v. Medtronic, Inc., 2010 WL 4009891, at *11 (Cal. App. Oct. 14, 2010) (unpublished) (“There is no authority for the proposition that a court may reexamine the [approval] process to evaluate whether the FDA expressly considered the particular feature of the device that is claimed to be defective.”).

Thus Buckman bars state-law plaintiffs from any “collateral attack on the validity of [a federal agency’s] decision” because “whether federal regulatory bodies fulfilled their duties with respect to the entities they regulate is ‘inherently federal in character.’”  Bader Farms, Inc. v. Monsanto Co., 2017 WL 633815, at *3 (E.D. Mo. Feb. 16, 2017) (quoting Buckman).  Could the Supreme Court sweep all of this precedent away?  As a matter of raw power, probably (although Congress might react).  But as a matter of jurisprudence, we think that if the Court intended to effect a legal change of that magnitude, such change would be supported by a lot more legal and policy consideration than occurred in Albrecht.  The Supreme Court, no more than Congress, cannot be viewed as “hid[ing] elephants in mouseholes” – particularly when preemption is involved.  Puerto Rico v. Franklin California Tax-Free Trust, 136 S. Ct. 1938, 1947 (2016).

So what do we think Albrecht meant by its reformulation of Levine’s “clear evidence” basis for preemption?  Well, particularly in light of what the Court actually held in Albrecht – that preemption is a legal matter to be decided by courts, not juries, including any ancillary factual issues that may arise, 139 S. Ct. at 1680  − we think that it is incumbent upon a defendant asserting that a particular common-law claim runs afoul of the FDA’s regulatory position concerning the product at issue to create a record showing that the FDA had sufficient information before it to come to that position under the relevant standards set by the FDCA.  The Court was legitimately concerned that defendants might have done the administrative equivalent of “tanking” requests for additional warnings.

Then, on the record created by the defendant, the trial court “simply” answers the preemption question, id. at 1679 − concluding yes or no based on that record.  Should a party, or the court, wish to go beyond this determination, it can ask the FDA itself, either through a citizen petition (which produces a “formal” agency decision of the sort Albrecht requires, id., see 21 U.S.C. §355(q)(1)(F); 21 C.F.R. §10.30(i)), or by inviting the FDA to present its views via an amicus curiae brief.  See Kisor v. Wilkie, 139 S. Ct. 2400, 2418 n.6 (2019) (an agency appearing as amicus is “not a party to the litigation, so “there [is] simply no reason to suspect that the interpretation [does] not reflect the agency’s fair and considered judgment”); 21 C.F.R. §10.30 (setting out formal procedures for resolution of citizen petitions, including Federal Register publication).

What we don’t think Albrecht means that plaintiffs can oppose preemption with arguments that the defendant should have done this, that, or the other thing differently in its submissions to the FDA.  That kind of second-guessing the bases for FDA decisions runs headlong into the policies that the Court enunciated in Buckman.  State common-law tort litigation is not, and should not be allowed to become, a basis for collateral attack on preemptive federal decision-making with lesser proof than applicable to legitimate administrative challenges (such as the “arbitrary and capricious” standard) under the Administrative Procedures Act.  Albrecht’s emphasis on agency action “within the scope of its congressionally delegated authority” demands no less.  139 S. Ct. at 1679.