Case Information
UNITED STATES DISTRICT COURT DISTRICT OF MASSACHUSETTS )
ELISSA M. WARNER, as Personal )
Representative of the Estate of Decedent, )
Lucas Joseph Warner, )
)
Plaintiff, )
) No. 1:24-cv-10632-JEK v. )
)
AMGEN INC. and AMGEN USA INC., )
)
Defendants. ) )
MEMORANDUM AND ORDER ON DEFENDANTS’ MOTION TO DISMISS KOBICK, J.
This is a wrongful death action brought by plaintiff Elissa Warner, as the personal representative of the Estate of Lucas Warner, against defendants Amgen Inc. and Amgen USA Inc. (collectively “Amgen”). Lucas, who was Elissa and William Warner’s son, suffered from an arteriovenous malformation, which caused him to experience severe and frequent headaches throughout his life. Lucas tragically passed away in 2020 at the age of 25. He died two years after taking a single injection of an Amgen-manufactured drug, Aimovig, that had been prescribed to treat his migraines. Shortly after his injection with Aimovig in June 2018, Lucas experienced a massive seizure, the consequences of which ultimately led to his death. Aimovig had been approved by the Food and Drug Administration (“FDA”) in May 2018, one month before Lucas’s injection with the drug.
Warner contends that Amgen’s labeling for Aimovig failed to provide adequate warnings that individuals with a history of seizures and cerebrovascular disease or surgery, like Lucas, were excluded from the FDA’s clinical testing for Aimovig. After removing the case to this Court, Amgen now moves to dismiss Warner’s complaint for failure to state a claim. Amgen contends that Warner’s wrongful death claim is preempted by the Federal Food, Drug, and Cosmetic Act, 21 U.S.C. § 301 et seq. , because it would have been impossible to comply with federal regulations governing prescription drug labeling and any state-law duty to provide warnings beyond those already included in Aimovig’s label.
Agreeing with Amgen, the Court will grant the motion to dismiss and deny Warner’s request for leave to amend the complaint. Under First Circuit precedent, Warner’s claims are preempted to the extent they allege that Amgen had a state-law duty to submit a stronger warning label to the FDA in its pre-approval review of Aimovig. And even accounting for the studies that Warner submits in connection with her request for leave to amend, Amgen has established that no newly acquired information came to light in the month between Aimovig’s approval and Lucas’s injection that would have permitted Amgen to unilaterally change Aimovig’s label under the FDA’s “changes being effected” regulation. The Warners suffered an immeasurable loss in Lucas’s passing. The law does not, however, afford them a remedy against Amgen on a claim that the warnings in Aimovig’s label were inadequate under Massachusetts law.
BACKGROUND
I. Statutory and Regulatory Framework.
The Federal Food, Drug, and Cosmetic Act (“FDCA”) requires drug manufacturers to obtain approval from the FDA before marketing or selling a new pharmaceutical drug in interstate commerce. See 21 U.S.C. § 355(a). As part of the approval process, a manufacturer must submit to the FDA a new drug application (“NDA”) that contains comprehensive information about the drug’s safety and effectiveness, the drug’s performance in clinical investigations and nonclinical studies, the composition of the drug, and the process for manufacturing the drug. See id. § 355(b)(1)(A); 21 C.F.R. § 314.50; Mut. Pharm. Co., Inc. v. Bartlett , 570 U.S. 472, 476 (2013). The manufacturer must also submit the proposed label for the drug, 21 U.S.C. § 355(b)(1)(A)(vi), including the text of the label and “annotations to the information in the summary and technical sections of the NDA that support the inclusion of each statement in the labeling,” 21 C.F.R. § 314.50(c)(2)(i). The label must, among other things, describe the drug’s dosage and administration, identify contraindications where the drug’s risks would its outweigh therapeutic benefit, warn about clinically significant risks or safety hazards, and provide a list of the most frequently occurring adverse events associated with the drug. Id. § 201.57(a); see Merck Sharp & Dohme Corp. v. Albrecht , 587 U.S. 299, 304 (2019).
The FDA will approve a drug to be marketed and sold only if there is “substantial evidence” that the drug is safe for use and “will have the effect it purports or is represented to have under the conditions of use prescribed, recommended, or suggested in the proposed labeling.” 21 U.S.C. § 355(d). The FDA must endorse “the exact text in the proposed label” before a manufacturer can secure approval of the drug. Wyeth v. Levine , 555 U.S. 555, 568 (2009); see 21 C.F.R. § 314.105(b). This regulatory scheme is intended not only to ensure that a label includes all necessary warnings, but also to “prevent overwarning, which may deter appropriate use of medical products, or overshadow more important warnings.” In re Zofran (Ondansetron) Prods. Liab. Litig. , 57 F.4th 327, 330 (1st Cir. 2023) (quotation marks and citation omitted).
Once a prescription drug is approved, the manufacturer must use the FDA-approved label when marketing and distributing the drug. See In re Celexa and Lexapro Mktg. & Sales Pracs. Litig. , 779 F.3d 34, 36 (1st Cir. 2015). If a manufacturer wishes to change the label, the “default rule” is that it must first obtain the FDA’s approval of any changes. at 37 (citing 21 C.F.R. § 314.70(b)(2)(v)(A)). But this rule has an exception: Under the FDA’s “changes being effected” (“CBE”) regulation, 21 C.F.R. § 314.70(c)(6)(iii), a manufacturer can “unilaterally amend a label and seek after-the-fact approval from the FDA” in certain circumstances, Zofran , 57 F.4th at 331. The CBE regulation permits a manufacturer to alter the label “to reflect newly acquired information” when the revision would “add or strengthen a contraindication, warning, precaution, or adverse reaction” for which there is “evidence of a causal association.” 21 C.F.R. § 314.70(c)(6)(iii)(A). Information is “newly acquired” if it consists of “data, analyses, or other information not previously submitted to the Agency, which may include (but is 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. §§ 314.3(b), 601.12(f)(6). Manufacturers may only propose label changes that are based on “reasonable evidence,” and the FDA can reject changes made pursuant to the CBE regulation after the manufacturer has made them. Albrecht , 587 U.S. at 315 (citing 21 C.F.R. §§ 314.70(c)(6)(iii)(A), (c)(7)).
II. Aimovig’s Approval and Labeling.
Aimovig, which is marketed by Amgen, is approved for the preventive treatment of migraine headaches in adults. ECF 13-1, at 1. It operates by blocking calcitonin gene-related peptide (“CGRP”) receptors. Id. CGRP is “a potent vasodilator”—that is, it widens blood vessels—and increased levels of CGRP are associated with migraine attacks. ECF 43-1, at 15. By interrupting the interaction between CGRP and CGRP receptors and thus inhibiting the function of CGRP, Aimovig decreases the frequency of migraines.
The FDA approved Aimovig on May 17, 2018. ECF 1-3. As part of its approval process, the agency solicited and prepared comprehensive reviews assessing Aimovig’s efficacy and safety. ECF 43-1 (Clinical Review), 43-2 (Non-Clinical Reviews), 43-3 (Other Reviews); see also ECF 1-3, at 3 (Aimovig FDA approval package listing all reviews conducted by the FDA). In the Clinical Review, the FDA observed that “[t]he safety profile of [Aimovig] was characterized in three pivotal trials and one dose-ranging trial.” ECF 43-1, at 16. The FDA thoroughly reviewed each of these trials—including their designs, protocols, inclusion and exclusion criteria, and results—before approving Aimovig. id. at 29-110.
Certain individuals, including those who had been diagnosed with a seizure disorder or cerebrovascular disease, were excluded from Aimovig’s clinical trials based on their medical histories and conditions. See id. at 16, 32-33, 78-79, 99; ECF 1-1, ¶¶ 25-26; ECF 33-6, Final Phase 3 Study Protocol, at 28-30. The Aimovig label approved by the FDA lists some of the categories of people who were excluded from the clinical studies, including patients who had experienced a “medication overuse headache,” heart attack, stroke, transient ischemic attack (that is, a brief interruption of blood flow to the brain), unstable angina, coronary artery bypass surgery, or other revascularization procedure “within 12 months prior to screening.” ECF 13-1, at 7, 9, 11. But Aimovig’s FDA-approved label does not state that people who had, over the course of their lifetimes, been diagnosed with a seizure disorder or cerebrovascular disease were excluded from the trials. See id. ; ECF 1-1, ¶¶ 25-26, 30.
The FDA’s reviews evaluated the known and theoretical risks and safety concerns posed by Aimovig. See generally ECF 43-1, 43-2, 43-3. As part of this process, the FDA’s Division of Cardiovascular and Renal Products (“DCRP”) was “asked to review the world’s literature and render an opinion regarding the strength and quality of published evidence concerning a theoretical risk” that blocking CGRP might cause “worsened ischemia” (i.e., less blood flow to the body’s tissues). ECF 43-3, DCRP Report, at 2 (quotation marks omitted). The DCRP found that although “CGRP is a potent microvascular vasodilator,” it is “one of multiple redundant control mechanisms of regional and tissue specific blood flow.” “Therefore,” the DCRP concluded, “if CGRP plays a role in coronary flow regulation, there are redundant systems to override inhibition of this modulator—both in healthy and diseased animals.” Id. The DCRP could not locate “information to suggest that the same is not true in the cerebral . . . circulatio[n].” Id.
The FDA’s Clinical Review analyzed Aimovig’s safety in a variety of populations. See ECF 43-1, at 127-213. The Clinical Review noted that another review submitted to the FDA, which focused on cardiovascular and cerebrovascular disorders, “revealed no potential safety concerns in regards to toxicity associated with [Aimovig] use,” though it observed that the “review is extremely limited as the population studied was primarily young and healthy.” Id. at 213 (comma omitted). The fact that patients with a variety of disorders, including seizure disorders and major neurological disorders, were excluded from Aimovig’s clinical trials “may limit the generalizability of the safety data to the larger population when considering that postmarketing use will be much less restrictive,” the Clinical Review explained. Id. at 131. In the Clinical Review’s “Labeling Recommendations,” the reviewer recommended including the following warning: “[T]here is a theoretical concern about the use of [Aimovig] in patients with major cardiovascular or cerebrovascular disease and in patients with high risk for major cardiovascular or cerebrovascular disease. Patients who have experienced a stroke, myocardial infarction [i.e., heart attack], transient ischemic attack, unstable angina or had coronary artery bypass surgery within the last year should not use [Aimovig].” at 215 (emphasis added). There is no evidence that Amgen or any reviewers within the FDA proposed a label warning against the use of Aimovig by individuals who had ever, in their lifetimes , been diagnosed with a seizure disorder or cerebrovascular disease, or a label warning that such individuals had been excluded from Aimovig’s clinical trials.
III. Factual Background.
The following facts, which are drawn from the complaint, are accepted as true for the purposes of this motion. At six years old, Lucas was diagnosed with an arteriovenous malformation (“AVM”) following an unrelenting headache. ECF 1-1, ¶ 9. The AVM, which doctors deemed inoperably large, caused Lucas to suffer debilitating headaches throughout his youth. Id. ¶¶ 10-12. Attempts to treat the AVM were, at best, only temporarily successful. Id. ¶¶ 15-20. Lucas nevertheless attained academic success, and he matriculated at American University in 2014, where he studied economics. Id. ¶¶ 13-14, 21, 62.
After a meeting in June 2018, Dr. Stephanie Wrobel-Goldberg, who was unaware that individuals with a history of seizures or cerebrovascular disease or surgery had been excluded from Aimovig’s clinical trials, prescribed Lucas Aimovig. Id. ¶¶ 28-29. On June 17, 2018—exactly one month after Aimovig received FDA approval—Lucas took his first and only injection of Aimovig. Id. ¶ 31. He had just finished his fourth year of part-time study at American University. Id. ¶ 62. Two days later, Lucas suffered a seizure and was found near death by a friend. Id. ¶¶ 32-35. Lucas was admitted to the intensive care unit at Boston Children’s Hospital (“BCH”), where doctors determined that he was suffering from multiple organ failure, significant brain damage, and continued seizure activity. Id. ¶¶ 36-40. Doctors placed Lucas into a pentobarbital coma in an attempt to halt the seizures. Id. ¶ 41. After performing multiple tests to ascertain the cause of his ailment, doctors concluded that Aimovig had crossed Lucas’s blood-brain barrier and caused the seizures. Id. ¶¶ 42-43.
Between June 29 and July 3, 2018, doctors performed three plasmapheresis treatments to remove Aimovig from Lucas’s brain and blood. ¶¶ 44-45. The treatments were effective, but by then Lucas had already suffered significant irreversible brain damage, including the near-total loss of his short-term memory. Id. ¶¶ 46-49. Lucas also suffered from mesial temporal sclerosis, which caused frequent seizures. Id. ¶ 47. On July 7, 2018, Lucas was transferred to BCH’s neurology floor, where doctors placed a gastric tube in his stomach because he was unable to swallow. Id. ¶¶ 48-50. On July 31, 2018, Lucas was transferred to a rehabilitation center, where he received daily physical, speech, and occupational therapy. Id. ¶¶ 54-56. Lucas returned home approximately one month later, where he required around-the-clock care because he received nutrition through his gastric tube, was unable to eat or drink, and could not be left alone. Id. ¶¶ 57- 60. Lucas never returned to American University.
Beginning in January 2020, Lucas began to experience renewed seizure activity due to the brain damage caused by Aimovig. Id. ¶ 65. After he completed a week-long seizure study at BCH, doctors concluded that they could not reverse the damage caused by Aimovig. Id. ¶¶ 67-68. Lucas’s cognition and health declined as the year progressed, and he was admitted to BCH on July 6, 2020. Id. ¶¶ 66, 69-70. Lucas remained at BCH until he passed away on November 17, 2020. Id. ¶ 71. His death was attributed to complications from Aimovig. Id. ¶ 72; ECF 1-1, Ex. B. IV. Procedural History.
Warner filed this lawsuit against Amgen and Novartis Institutes for BioMedical Research, Inc. (“NIBR”), in Middlesex Superior Court on November 17, 2023. ECF 1-1. The complaint asserts a single failure-to-warn wrongful death claim under M.G.L. c. 229, § 2. ¶¶ 73-97. In March 2024, Amgen removed the case to this Court on the basis of diversity jurisdiction. ECF 1. Amgen and NIBR subsequently filed separate motions to dismiss. ECF 12, 14. Warner voluntarily dismissed NIBR as a defendant and opposed Amgen’s motion. ECF 31, 33. At the hearing on Amgen’s motion to dismiss, and in response to Amgen’s preemption defense, Warner requested leave to file an amended complaint to make further allegations that Amgen possessed “newly acquired information” that would have permitted it to unilaterally strengthen the warning on Aimovig’s label via the CBE regulation in the month between Aimovig’s approval and Lucas’s injection. ECF 40, at 26-28. With leave of court, the parties submitted supplemental briefing addressing this issue. ECF 42, 43.
STANDARD OF REVIEW
In evaluating a motion to dismiss under Federal Rule of Civil Procedure 12(b)(6), the Court must determine “‘whether, construing the well-pleaded facts of the complaint in the light most favorable to the plaintif[f], the complaint states a claim for which relief can be granted.’” Cortés- Ramos v. Martin-Morales , 956 F.3d 36, 41 (1st Cir. 2020) (quoting Ocasio-Hernández v. Fortuño- Burset , 640 F.3d 1, 7 (1st Cir. 2011)). The complaint must allege “a plausible entitlement to relief.” Bell Atl. Corp. v. Twombly , 550 U.S. 544, 559 (2007). “A claim has facial plausibility when the plaintiff pleads factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged.” Ashcroft v. Iqbal , 556 U.S. 662, 678 (2009). “While legal conclusions can provide the framework of a complaint, they must be supported by factual allegations.” at 679.
DISCUSSION
Warner seeks to hold Amgen liable for Lucas’s death through a wrongful death claim under M.G.L. c. 229, § 2. To prevail on the claim, she must establish that Amgen was negligent. See Correa v. Schoeck , 479 Mass. 686, 693 (2018). “The elements of a negligence claim are that ‘the defendant owed the plaintiff a duty of reasonable care, that the defendant breached this duty, that damage resulted, and that there was a causal relation between the breach of the duty and the damage.’” Id. (quoting Jupin v. Kask , 447 Mass. 141, 146 (2006)). Warner’s theory of negligence is that Amgen had a duty under Massachusetts law to include warnings in Aimovig’s label (1) that individuals who had ever been diagnosed with a seizure disorder or cerebrovascular disease were excluded from Aimovig’s clinical testing, and (2) that, by blocking CGRP receptors, Aimovig poses cerebral risks. See ECF 1-1, ¶¶ 92-94; ECF 33, at 1, 11; ECF 42, at 2.
Assuming, without deciding, that Massachusetts law does require such warnings, Warner’s wrongful death claim is nonetheless preempted by the FDCA under Supreme Court and First Circuit precedent. Insofar as Warner seeks to hold Amgen liable for failing to propose a stronger warning label to the FDA before Aimovig’s approval, her claim is preempted because the FDA considered the exclusion criteria used in Aimovig’s clinical trials and the theoretical risks associated with CGRP antagonism. To the extent Warner seeks to hold Amgen liable for failing to strengthen its warning label after obtaining FDA approval, her claim is also preempted because there is no reasonable basis to find that Amgen possessed newly acquired information of a causal association between Aimovig and an adverse event or risk, such that it could have unilaterally strengthened the label warnings via the CBE regulation in the month after the drug’s approval. I. Impossibility Preemption.
Federal preemption doctrine derives from the Supremacy Clause, which provides that federal law “shall be the supreme Law of the Land; . . . any Thing in the Constitution or Laws of any State to the Contrary notwithstanding.” U.S. Const. art. VI, cl. 2. Federal law can preempt state law expressly or by implication. See Gade v. Nat’l Solid Wastes Mgmt. Ass’n , 505 U.S. 88, 98 (1992). There are two species of implied preemption: (1) impossibility preemption, where compliance with the requirements of state and federal law is a “physical impossibility,” and (2) obstacle preemption, where a state law “stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress.” Arizona v. United States , 567 U.S. 387, 399 (2012); see Wyeth , 555 U.S. at 563-64. Amgen raises an impossibility preemption defense, contending that it would have been impossible for it to comply with both the labeling requirements of the FDCA and the duty to warn that Warner claims is required by Massachusetts law.
Impossibility preemption is a “demanding defense.” Wyeth , 555 U.S. at 573. Like all forms of preemption, it is guided by “two cornerstones” of preemption jurisprudence. Id. at 565. “First, ‘the purpose of Congress is the ultimate touchstone in every pre-emption case.’” (quoting Medtronic, Inc. v. Lohr , 518 U.S. 470, 485 (1996)). Second, the presumption against preemption applies, such that the party asserting preemption must overcome the presumption that state regulation “can constitutionally coexist with federal regulation.” Hillsborough County, Florida v. Automated Med. Lab’ys., Inc. , 471 U.S. 707, 716 (1985); see Wyeth , 555 U.S. at 565. Amgen bears the burden of establishing that compliance with state and federal law was impossible. See Albrecht , 587 U.S. at 313; Wyeth , 555 U.S. at 573.
The Supreme Court and First Circuit have elaborated on the contours of impossibility preemption in the context of pharmaceutical regulation. In Wyeth v. Levine , the Supreme Court “undertook a careful review of the history of federal regulation of drugs and drug labeling,” and “found nothing within that history to indicate that the FDA’s power to approve or to disapprove labeling changes, by itself, pre-empts state law.” Albrecht , 587 U.S. at 311 (citing Wyeth , 555 U.S. at 565-68). Wyeth instead found that “through many amendments to the FDCA and to FDA regulations, it has remained a central premise of federal drug regulation that the manufacturer bears responsibility for the content of its label at all times.” 555 U.S. at 570-71. More specifically, a manufacturer is “charged both with crafting an adequate label and with ensuring that its warnings remain adequate as long as the drug is on the market.” at 571. This duty is often codified and enforced through state law. See Albrecht , 587 U.S. at 311-12.
Warner contends that Amgen had a state-law duty to ensure that the labeling it proposed to the FDA before Aimovig’s approval warned that people with a history of seizures and cerebrovascular disease were excluded from clinical trials, and that Amgen also had a state-law duty to strengthen its warning label after obtaining FDA approval. Under either her pre-approval or post-approval theory, Amgen “would need to change [Aimovig’s] label in order to avoid liability under state law.” Celexa , 779 F.3d at 40. But manufacturers do not have the final say on prescription drug labeling. The FDA must approve the exact text of a drug label, Wyeth , 555 U.S. at 568, and federal law generally requires manufacturers to use the FDA-approved label when marketing their drugs, Celexa , 779 F.3d at 36, 37. Unless the CBE pathway is available, a brand- name drug manufacturer must receive the FDA’s blessing before changing a drug label after approval. See id. at 37. Amgen contends that these federal-law strictures prevented it from altering Aimovig’s label to reflect the warnings that Warner maintains are required by Massachusetts law. PLIVA, Inc. v. Mensing , 564 U.S. 604, 620 (2011) (when a private actor cannot “independently do under federal law what state law requires of it,” state law is preempted). Because the analysis of Amgen’s preemption defense differs for Warner’s pre-approval theory and post-approval theory, the Court will address each in turn.
II. Preemption of Warner’s Pre-Approval Claim.
Warner first contends that Amgen is liable for failing to submit, before Aimovig’s approval, a warning label that listed all of the patient populations excluded from Aimovig’s clinical trials. ECF 1-1, ¶¶ 92-93. It is undisputed that, during the FDA’s pre-approval review of Aimovig, the FDA knew that individuals who had ever been diagnosed with a seizure disorder or cerebrovascular disease were excluded from Aimovig’s clinical trials. And it is further undisputed that, despite having this knowledge, the FDA approved a label that did not warn consumers that people who had experienced seizures or cerebrovascular disease during their lifetimes were excluded from Aimovig’s clinical trials. Warner nevertheless contends that it was not impossible for Amgen to comply with both the FDCA and state law before Aimovig’s approval, because Amgen could have proposed a label disclosing the full list of clinical trial exclusion criteria to the FDA, and the FDA might have approved such a label.
Warner’s argument finds support in several district court decisions. In Holley v. Gilead Sciences, Inc. , the court concluded that federal law does not “prevent a drug manufacturer from submitting a different warning label to the FDA prior to initial approval of a drug.” 379 F. Supp. 3d 809, 826 (N.D. Cal. 2019). The court pointed out that the FDA’s regulations governing the submission of proposed labeling establish a floor, not a ceiling, and thus “do not conflict with any state-law duties regarding adequacy of warnings.” Id. (citing 21 U.S.C. § 355(b)(1)(F); 21 C.F.R. § 314.50(c)(2)(i)). Because it is possible for a drug manufacturer to submit a stronger proposed label to the FDA under those regulations, the court held, compliance with state and federal law was not an impossibility. Similarly, in Gaetano v. Gilead Sciences, Inc. , the court explained that because the FDCA and its implementing regulations do not limit what labeling a drug manufacturer can propose to the FDA, state-law failure-to-warn claims can “fill that gap.” 529 F. Supp. 3d 333, 345 (D.N.J. 2021) (“There are no such constraints upon a manufacturer when it is choosing what labeling to propose to the FDA; only after a label is approved is the manufacturer’s discretion constrained by federal law.”); see also In re Actos (Pioglitazone) Prods. Liab. Litig. , No. 12-cv-00064-RFD, 2014 WL 60298, at *7 (W.D. La. Jan. 7, 2014) (“Although defendants are correct in stating the labeling language must not deviate from that which was approved by the FDA, defendants still possessed the ability to implement stronger warning language into labeling, by submitting stronger warning language for FDA approval.”).
First Circuit precedent, however, forecloses this line of argument. In Celexa , the First Circuit held that the FDA is “the exclusive judge of safety and efficacy based on information available at the commencement of marketing, while allowing the states to reach contrary conclusions when new information not considered by the FDA develops.” 779 F.3d at 41 (emphasis added). The First Circuit reasoned that, in Wyeth , the Supreme Court “effectively reserve[d] the launch of new drugs to the expertise of the FDA, but then preserve[d] a wide scope for the states in requiring manufacturers to respond to information not considered by the FDA.” Thus, under Celexa , the dispositive inquiry for preemption challenges to pre-approval claims is not whether the manufacturer could have proposed a label with the warning required by state law, but instead whether the FDA considered the safety and efficacy information giving rise to the need for that warning. id. Stated differently, when the FDA has considered the safety and efficacy information required by state law in the course of approving a prescription drug’s label, there exists clear evidence that the FDA would not have approved the warning required by state law. Cf. Albrecht , 587 U.S. at 310 (requiring, for post-approval claims, “‘clear evidence’ that the FDA would not have approved the warning that state law requires” (quoting Wyeth , 555 U.S. at 571)).
Applying Celexa ’s standard, Warner’s claim based on her pre-approval theory is preempted because the FDA was fully aware of the exclusion criteria used in Aimovig’s clinical trials before it approved the Aimovig label in May 2018. In the Clinical Review, the FDA acknowledged that the exclusion of patients who had ever been diagnosed with, among other things, seizure disorders and cerebrovascular disease, could “limit the generalizability of the safety data to the larger population when considering that postmarketing use [of Aimovig] will be much less restrictive.” ECF 43-1, at 131; see also id. at 213 (Clinical Review noting that a review of potential safety concerns regarding Aimovig toxicity was “extremely limited as the [clinical study] population was primarily young, and healthy”). The Clinical Review also “examined in detail” the “theoretical” cardiovascular and cerebrovascular safety risks that Aimovig might pose, including the risk that Aimovig might induce “a potential lack of compensatory vasodil[ation] in the context of ischemia.” Id. at 195-96, 213. On the other hand, the limited data before the FDA “revealed no potential safety concerns in regards to toxicity associated with [Aimovig] use” for individuals with “cardiovascular and cerebrovascular disorders.” at 213. Having considered the range of safety information and the labeling proposed by Amgen and recommended by the reviewer, the FDA did not require Amgen to provide the complete list of exclusion criteria used in the clinical trials on the Aimovig label. Cf. Zofran , 57 F.4th at 330 (noting that one of the FDCA’s objectives is “to prevent overwarning, which may deter appropriate use of medical products, or overshadow more important warnings” (quotation marks and citation omitted)).
The record thus shows not only that the FDA was aware that individuals who had ever been diagnosed with a seizure disorder or cerebrovascular disease were excluded from Aimovig’s clinical trials, but also that it considered the risks and safety concerns associated with these exclusions. Put differently, the alleged need to strengthen the label’s warnings is based on information “that was plainly known to the FDA prior to approving the label.” Celexa , 779 F.3d at 43. Since, under First Circuit precedent, the FDA is the “exclusive judge of safety and efficacy based on information available at the commencement of marketing,” Warner’s pre-approval failure-to-warn claim is necessarily preempted by federal law. Id. at 41; cf. id. at 44 (holding that state-law claim that a manufacturer misleadingly omitted material efficacy information, based on results of clinical trials, was preempted because the clinical trial results were “plainly known to the FDA prior to approving the label”); Zofran , 57 F.4th at 337-39 (finding failure-to-warn claim preempted because the FDA considered the clinical studies cited by the plaintiffs before approving the relevant product label).
III. Preemption of Warner’s Post-Approval Claim.
Warner separately contends that, after Aimovig’s approval, Amgen had a duty under Massachusetts law to strengthen the drug’s label with a warning that people with a history of seizure disorders or cerebrovascular disease were excluded from the clinical trials. She also asks to amend her complaint to make further allegations that Amgen is liable for failing to warn, post- approval, that Aimovig poses “cerebral risks.” ECF 42, at 1-2. She argues that newly acquired information would have permitted Amgen to use the CBE regulation, in the month between Aimovig’s approval by the FDA and Lucas’s injection of the drug, to unilaterally add this warning to the Aimovig label.
An impossibility preemption defense for this type of post-approval claim generally hinges on whether the manufacturer could have strengthened the label warnings through the FDA’s CBE regulation. See, e.g. , Zofran , 57 F.4th at 336-37; Celexa , 779 F.3d at 41. The CBE regulation authorizes manufacturers to unilaterally change FDA-approved drug labels “to reflect newly acquired information” when, as relevant here, the revision would “add or strengthen a contraindication, warning, precaution, or adverse reaction for which the evidence of a causal association satisfies the standard for inclusion in the labeling under [21 C.F.R.] § 201.57(c).” 21 C.F.R. § 314.70(c)(6)(iii)(A). When the CBE procedure is available to a manufacturer, state law claims are not preempted because it is not impossible for a manufacturer to comply with both state and federal requirements. See Wyeth , 555 U.S. at 573. But when the CBE procedure is not available—when, for example, a manufacturer does not have newly acquired information showing a causal association between the drug and a contraindication or adverse event—the state law claim is preempted because the manufacturer cannot give the warnings that state law requires. See Gibbons v. Bristol-Myers Squibb Co. , 919 F.3d 699, 708 (2d Cir. 2019); Celexa , 779 F.3d at 41. And even when the CBE process is available, a manufacturer can prevail on an impossibility preemption defense if it can show “clear evidence” that the FDA would have ultimately rejected the label change. Albrecht , 587 U.S. at 302-03; Wyeth , 555 U.S. at 571.
Warner’s complaint does not allege that Amgen had newly acquired information that would have permitted it to unilaterally alter Aimovig’s label in the month between the drug’s approval and Lucas’s injection. But she has requested leave to amend the complaint, seeking to add factual allegations supporting her contention that Amgen could have modified the Aimovig label via the CBE regulation. ECF 42, at 2-3; see also ECF 33, at 20. Federal Rule of Civil Procedure 15(a)(2) provides that a court “should freely give leave [to amend] when justice so requires.” Fed. R. Civ. P. 15(a)(2). “In the absence of any apparent or declared reason—such as undue delay, bad faith or dilatory motive on the part of the movant, repeated failure to cure deficiencies by amendments previously allowed, undue prejudice to the opposing party by virtue of allowance of the amendment, futility of amendment, etc.—the leave sought should, as the rules require, be ‘freely given.’” Foman v. Davis , 371 U.S. 178, 182 (1962).
Amgen contends that amendment would be futile because, even if permitted, Warner cannot plausibly allege that Amgen could have used the CBE process to unilaterally modify the Aimovig label. Amendment is futile if “‘the complaint, as amended, would fail to state a claim upon which relief could be granted.’” Amyndas Pharms., S.A. v. Zealand Pharma A/S , 48 F.4th 18, 40 (1st Cir. 2022) (quoting Glassman v. Computervision Corp. , 90 F.3d 617, 623 (1st Cir. 1996)). The Court will thus evaluate the proposed amendment under the same pleading standard as the operative complaint.
In support of her request to file an amended complaint, Warner has submitted nine articles that, she avers, constitute newly acquired information of the “cerebral risks presented by CGRP- blocking drugs, like Aimovig.” ECF 42, at 2. In Warner’s view, the articles establish “a reasonable basis for believing that a CGRP blockade posed a type, severity, or frequency of risk that was not submitted to or analyzed by the FDA prior to Aimovig’s approval.” at 4. Amgen responds (1) that the articles do not constitute newly acquired information; (2) that even if they did, it could not have utilized the CBE regulation; and (3) that it could not have plausibly updated the label in the time between when Aimovig was approved and when Lucas received his injection. ECF 43, at 1-2. To prevail on its preemption defense and defeat Warner’s request for leave to amend, Amgen bears the burden of establishing that, even with all reasonable inferences drawn in Warner’s favor, it is not plausible that the articles constitute newly acquired information demonstrating a causal association between Aimovig and a risk or adverse event that would require a stronger label warning. See Twombly , 550 U.S. at 559.
For multiple reasons, Amgen has carried its burden. First, none of the articles link Aimovig or any other CGRP-blocking drug to an adverse event or risk. The articles do not, therefore, plausibly show any “causal association” between Aimovig and a “contraindication, warning, precaution, or adverse reaction” that “satisfies the standard for inclusion in [Aimovig’s] labeling.” 21 C.F.R. § 314.70(c)(6)(iii)(A); see Pietrantoni v. Corcept Therapeutics Inc. , 640 F. Supp. 3d 197, 214 (D. Mass. 2022) (“[T]he CBE regulation requires reasonable evidence of a causal association between the drug and clinically significant adverse reactions.” (quotation marks and emphasis omitted)). Rather, the articles discuss CGRP generally or the process of “ischemic conditioning,” which can be facilitated by CGRP. Ischemic conditioning involves intentionally exposing tissue or organs to brief periods of ischemia (i.e., reduced blood flow and oxygenation), followed by reperfusion (i.e., restoration of blood flow to the tissue or organ), to protect against organ damage following a major ischemic event, such as a heart attack or stroke. See, e.g. , ECF 42-2, at 1-2.
Second, the articles do not plausibly constitute newly acquired information, because they do not contain information that was unknown to the FDA when it approved Aimovig. Under FDA regulations, information is “newly acquired” if it consists of “data, analyses, or other information not previously submitted to the Agency , which may include (but is 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 .” 21 C.F.R. § 314.3(b) (emphases added). Thus, newly acquired information “must be information the FDA lacked when it approved the drug.” Pietrantoni , 640 F. Supp. 3d at 214 (citing cases). This may include, for example, “an increasing body of data of an inherent risk with the drug.” Celexa , 779 F.3d at 42.
Warner contends, at a high level of generality, that the articles reveal new “cerebral risks” posed by Aimovig that were not known to the FDA in May 2018. ECF 42, at 1-2. But the FDA’s pre-approval Clinical Review, as discussed, assessed safety concerns for individuals with cerebrovascular disorders and certain theoretical cerebrovascular risks posed by Aimovig, including the specific risk of a “potential lack of compensatory vasodi[lation] in the setting of ischemia.” See id. at 195-96, 213-14. The DCRP Report concluded that “if CGRP plays a role in coronary flow regulation, there are redundant systems to override inhibition of [CGRP]—both in healthy and diseased animals.” ECF 43-3, DCRP Report, at 2. It could “locate no information to suggest that the same is not true in the cerebral . . . circulations.” Thus, to the extent that Warner asserts that Aimovig poses “cerebral risks” in general, this category of risk was “plainly known” to the FDA before it approved Aimovig. Celexa , 779 F.3d at 43.
Six of the nine articles were published after the submission of the Clinical Review, certain Non-Clinical Reviews, and the DCRP report to the FDA. Warner contends that the three older articles—Kokkoris et al. (2012), [7] Liu et al. (2011), [8] and Rehni et al. (2007) [9] —demonstrate that CGRP reduces the occurrence or severity of cerebral injuries. From there, she argues by inference that a drug like Aimovig, which blocks CGRP, risks impairing CGRP’s protective effects. But again, the FDA considered this information before approving Aimovig. See, e.g. , ECF 43-1, at 129, 195-96, 213-16. The DCRP reviewed “the world’s literature” before preparing its report on the theoretical risks associated with CGRP antagonism. ECF 43-3, DCRP Report, at 1. It is not plausible that articles bearing on that subject, published at least four years before the DCRP began preparing its report, “reveal risks of a different type or greater severity or frequency than previously included in submissions to FDA.” 21 C.F.R. § 314.3(b).
The remaining six articles provide only tenuous support, at best, for Warner’s assertion that CGRP-blocking drugs pose “cerebral risks” that were unknown to the FDA when it approved Aimovig. Notably, none of the articles discuss the cerebral injury suffered by Lucas as a result of Aimovig—namely, a seizure. The only article that provides any direct support for Warner’s contention that CGRP antagonism creates more general “cerebral risks”—Zhai et al. (2018) [10] — demonstrates that genetically modified mice with a CGRP deficiency experience worse outcomes following cerebral ischemia than those without a CGRP deficiency. See ECF 42-3, at 2. This study’s findings are, for several reasons, too attenuated from Warner’s claim to plausibly constitute newly acquired information. First, the article examined how CGRP mediates the effects of cerebral ischemia and strokes, not seizures, and it does not link cerebral ischemia to seizures. See generally ECF 42-3. Second, the study does not address whether its findings about mice with a CGRP deficiency generalize to drugs, like Aimovig, that are CGRP antagonists. See id. ; cf. ECF 43-3, DCRP Report, at 8-9, 11-12 (discussing animal studies finding that CGRP antagonism did not worsen outcomes following myocardial ischemia in pigs or dogs). Third, while animal studies involving a manufacturer’s drug product are relevant to the FDA’s risk assessments and can constitute newly acquired information, see, e.g. , Zofran , 57 F.4th at 337-39, Warner does not cite any precedent for treating a single animal study not involving the manufacturer’s drug as newly acquired information, nor has the Court identified any such precedent. In view of these considerations, this study does not plausibly demonstrate that Aimovig posed a different type or severity of risk than was previously known to the FDA.
The other five articles are even more attenuated. Two of the articles—Zhao et al. (2018), [11] a meta-analysis, and Szeto et al. (2018), [12] a literature review—do not even mention CGRP. Zhao et al. (2018) finds that a process called “remote ischemic postconditioning” “might offer cerebral protection for stroke patients suffering from or . . . at risk of [ischemia-reperfusion injury].” ECF 42-4, at 2. Szeto et al. (2018) addresses how diabetes drugs that block “K ATP channels” might increase a diabetes patient’s risk of stroke. See ECF 42-5, at 2, 7-8. In a conclusory manner, Warner asserts that because CGRP plays a role in ischemic conditioning and the activation of K ATP channels, a drug like Aimovig, which blocks CGRP, creates cerebral risks. See ECF 42, at 4-5. But it is implausible that these articles—which fail to mention CGRP or CGRP-blocking drugs— “reveal risks of a different type or greater severity or frequency” involving Aimovig “than previously included in submissions to [the] FDA.” 21 C.F.R. § 314.3(b); cf. Celexa , 779 F.3d at 42 (even if the court assumed that an article evaluating the general efficacy of antidepressant medications could constitute newly acquired information, the article did not support the plaintiffs’ contention that the defendants’ antidepressant label misled consumers about its efficacy for adolescents).
The two articles that postdate Aimovig’s approval and predate Lucas’s injection—Guo et al. (2018), [13] a rat study, and Basalay et al. (2018), [14] a literature review—also discuss ischemic conditioning. But the record shows that the FDA considered CGRP’s role in ischemic conditioning before it approved Aimovig. The DCRP Report evaluated three articles addressing the cardioprotective role that CGRP plays in ischemic pre- and post-conditioning. See ECF 43-3, DCRP Report, at 13-16. Reviewers within the Division of Neurology Products (“DNP”) discussed this report and additionally noted “numerous publications on [ischemic] preconditioning and its potential beneficial effects, not only on the cardiovascular system but also on other (e.g., neurological, and renal) systems.” ECF 43-2, DNP Memo, at 4; see ECF 43, at 2 & n.3 (Amgen citing same). The DCRP Report and DNP reviewers each acknowledged the potential risk that CGRP antagonism could inhibit “compensatory vasodilation in the setting of ischemic vascular events.” ECF 43-3, DCRP Report, at 1; see ECF 43-2, DNP Memo, at 2. And the DCRP Report concluded that there are “redundant systems to override inhibition of this modulator” in the cardio- and cerebrovascular contexts. ECF 43-3, DCRP Report, at 1; see also ECF 43-1, at 215-16 (Clinical Review noting that the Medical Policy and Program Review Council unanimously agreed that the animal data related to the theoretical risks associated with CGRP antagonism in patients with major cardiovascular disease “was not compelling enough to include a warning in . . . the label”). The DNP reviewers did “not reach a conclusion regarding the potential for [Aimovig] to antagonize the potential beneficial effects of preconditioning,” but they concluded that the limited extant data “suggest[ed] that the contribution of CGRP to compensatory vasodilation may be relatively small.” ECF 43-2, DNP Memo, at 5, 6. Thus, Guo et al. (2018) and Basalay et al. (2018) do not reveal different or more severe risks than those already considered by the FDA.
The final article—Sorby-Adams et al. (2017), a literature review—cannot plausibly constitute newly acquired information for similar reasons. Warner claims that this literature review “establishes that CGRP improves the outcome of stroke, helps stabilize the blood-brain barrier, and reduces cerebral injury.” ECF 42, at 5. The article says that “CGRP administered at the onset of reperfusion produced a significant reduction in infarct volume, [blood-brain barrier] permeability and cerebral oedema following rodent stroke.” ECF 42-6, at 17. This is not, however, new information: Warner submitted Liu et al. (2011), the sole study cited in support of this assertion. As discussed, it is not plausible that the FDA overlooked a 2011 study analyzing the cerebroprotective effects of CGRP when it evaluated the theoretical risks associated with CGRP antagonism “in the setting of ischemic vascular events.” ECF 43-3, DCRP Report, at 1. A one- sentence reference to this study in a literature review published six years later does not alter this conclusion.
In sum, Warner does not plausibly allege, through the articles she seeks to submit as part of an amended complaint, the existence of newly acquired information or a causal association between Aimovig and an adverse event or risk. The proposed amendment would thus be futile because Amgen could not have used the CBE regulation to unilaterally modify the Aimovig label between May 17, 2018, when it received FDA approval, and June 17, 2018, when Lucas received his injection. Accordingly, Warner’s request to amend will be denied, and the Court concludes that Warner’s post-approval claim against Amgen, like her pre-approval claim, is preempted.
CONCLUSION AND ORDER
For the foregoing reasons, Amgen’s motion to dismiss, ECF 12, is GRANTED. The complaint is DISMISSED with prejudice and without leave to amend.
SO ORDERED.
/s/ Julia E. Kobick JULIA E. KOBICK Dated: February 13, 2025 UNITED STATES DISTRICT JUDGE
[1] The Court follows the parties’ lead in referring to Lucas Warner as “Lucas,” so as to avoid confusing him with his mother, plaintiff Elissa Warner, referred to as “Warner” in this decision.
[2] In this context, the term “label” refers “to the written material that is sent to the physician who prescribes the drug and the written material that comes with the [drug’s packaging].” Merck Sharp & Dohme Corp. v. Albrecht , 587 U.S. 299, 303-04 (2019).
[3] Warner and Amgen each ask the Court to consider documents prepared by the FDA in connection with its review of Aimovig’s NDA. When deciding a motion to dismiss, a court may “consider not only the complaint but also matters fairly incorporated within it and matters susceptible to judicial notice.” In re Colonial Mortg. Bankers Corp. , 324 F.3d 12, 15 (1st Cir. 2003); see also Boateng v. InterAmerican Univ. , Inc., 210 F.3d 56, 60 (1st Cir. 2000) (noting that a court “may look to matters of public record in deciding a Rule 12(b)(6) motion”). The documents submitted by the parties are publicly accessible on the FDA’s website, their authenticity is not disputed, and they are fairly incorporated into the complaint. The Court will therefore consider the following documents when deciding Amgen’s motion to dismiss: the FDA’s May 2018 Aimovig Approval Package, ECF 1-3; the FDA-Approved Aimovig Label, ECF 13-1; and the FDA’s Clinical, Non-Clinical, and “Other” Reviews prepared in connection with its review of Aimovig’s NDA, ECF 43-1, 43-2, 43-3. See Pietrantoni v. Corcept Therapeutics Inc. , 640 F. Supp. 3d 197, 204-05 (D. Mass. 2022) (taking judicial notice of FDA documents when deciding motion to dismiss drug product liability claims on preemption grounds and collecting cases doing the same); Gustavsen v. Alcon Lab’ys, Inc. , 272 F. Supp. 3d 241, 252 (D. Mass. 2017) (same), aff’d , 903 F.3d 1 (1st Cir. 2018).
[4] Amgen disputes that Massachusetts law imposes such a duty to warn, contending that Warner’s wrongful death claim is barred by the learned intermediary doctrine. ECF 13, at 10-12. Amgen also argues that the complaint fails to plead causation or adequately allege a request for punitive damages. at 12-14. Because Warner’s claim is preempted, the Court does not reach these arguments.
[5] Warner does not assert that any of her articles disclose a specific risk posed by Aimovig to individuals, like Lucas, with an AVM or a history of seizures. In similar cases, the newly acquired information at issue revealed a more specific risk, and bore a closer nexus to the manufacturer’s drug product, than the information in Warner’s articles. See, e.g. , Wyeth , 555 U.S. at 569-70 (discussing post-approval adverse events showing that the manufacturer’s drug created risk of gangrene when injected directly into a patient’s vein); Albrecht , 587 U.S. at 305-07 (describing evidence that long-term use of the manufacturer’s drug increased risk of atypical femoral fractures); Zofran , 57 F.4th at 337-38 (plaintiffs alleging that animal studies demonstrated that the manufacturer’s drug could cause fetal defects); Holley , 379 F.Supp.3d at 828-29 (plaintiffs alleging that pre- and post-approval evidence demonstrated that the manufacturer’s drug posed safety risks to patients’ kidneys and bones).
[6] Four of the articles submitted by Warner were published between August 2017 and April 2018. See ECF 42-3 (Exhibit C), ECF 42-4 (Exhibit D), ECF 42-5 (Exhibit E), ECF 42-6 (Exhibit F). Amgen contends that these articles cannot constitute newly acquired information because they were published before the FDA approved Aimovig in 2018. ECF 43, at 4. The Court does not agree. While the FDA’s Aimovig-related reviews were not marked as “complete” until 2018, several were submitted in May 2017. See, e.g. , ECF 43-1, at 1 (Clinical Review listing submission date of May 17, 2017); ECF 43-2, Pharmacology/Toxicology NDA/BLA Review and Evaluation, at 1 (“Applicant’s letter date” of May 17, 2017); ECF 43-3, DCRP Report (listing “Date of Document” as May 17, 2017). Because the Court must draw all reasonable inferences in Warner’s favor, it is plausible that the FDA did not review articles published between the date when the various reviews and reports were initially submitted in 2017 and when they were finalized in 2018.
[7] See ECF 42-7 (Exhibit G), Kokkoris, et al., Role of calcitonin gene-related peptide in cerebral vasospasm, and as a therapeutic approach to subarachnoid hemorrhage , 3 F RONTIERS IN E NDOCRINOLOGY 135 (Nov. 15, 2012).
[8] See ECF 42-8 (Exhibit H), Liu, et al., Calcitonin gene-related peptide prevents blood-brain barrier injury and brain edema induced by focal cerebral ischemia reperfusion , 171 R EGUL . P EPTIDES 19 (June 28, 2011).
[9] See ECF 42-9 (Exhibit I), Rehni, et al., Possible involvement of insulin, endogenous opioids and calcitonin gene related peptide in remote ischaemic preconditioning of brain , 127 P HARM . S OC ’ Y OF J APAN 1013 (June 1, 2007).
[10] ECF 42-3 (Exhibit C), Zhai, et al., Endogenous calcitonin gene-related peptide suppresses ischemic brain injuries and progression of cognitive decline , 36 J. H YPERTENSION 876 (Apr. 2018).
[11] See ECF 42-4 (Exhibit D), Zhao, et al., Remote ischemic postconditioning for ischemic stroke: A systematic review and meta-analysis of randomized controlled trials , 131 C HINESE M ED . J. 956 (Apr. 20, 2018).
[12] ECF 42-5 (Exhibit E), Szeto, et al., The role of K ATP channels in cerebral ischemic stroke and diabetes , 39 A CTA P HARMACOLOGICA S INICA 683 (Apr. 19, 2018).
[13] See ECF 42-1 (Exhibit A), Guo, et al., Independent roles of CGRP in cardioprotection and hemodynamic regulation in ischemic postconditioning , 828 E UR . J. P HARMACOLOGY 18 (June 5, 2018). This article found that CGRP “may” mediate the “cardioprotective” effects of a certain type of ischemic conditioning. ECF 42-1, at 2. A discussion of the cardiovascular benefits of a process mediated by CGRP provides no direct support for Warner’s contention that these articles reveal different or more severe cerebral risks associated with CGRP antagonism.
[14] ECF 42-2 (Exhibit B), Basalay, et al., Neural mechanisms in remote ischaemic conditioning in the heart and brain: mechanistic and translational aspects , 113 B ASIC R SCH . C ARDIOLOGY 25 (June 1, 2018).
[15] ECF 42-6 (Exhibit F), Sorby-Adams, et al., The role of neurogenic inflammation in blood-brain barrier disruption and development of cerebral oedema following acute central nervous system (CNS) injury , 18 I NT ’ L J. M OLECULAR S CI . 1788 (Aug. 17, 2017).
[16] Warner also argues that since Amgen did not propose a label listing all exclusion criteria before Aimovig received FDA approval, it could have treated the exclusion criteria as newly acquired information and modified the Aimovig label via the CBE pathway. See ECF 33, at 13- 14. This argument is contradicted and foreclosed by Warner’s admission that the FDA was fully informed of the exclusion criteria before it approved Aimovig. ECF 33, at 6. Information plainly known to the FDA cannot constitute newly acquired information. See 21 C.F.R. § 314.3(b); Celexa , 779 F.3d at 43.