Eli Lilly & Co. v. Kennedy,
2025 U.S. Dist. LEXIS 196211 (S.D. Ind. 2025).

Andrew Wasson*

Why It Made the List

The so-called definitional controversies of food and drug law have exercised jurists, litigants, and FDA since the earliest days of food and drug regulation. Along one axis, definitional legal controversies in food and drug law pose difficulties because of the lack of scientific and cultural consensus about the conceptual categorical boundaries for the objects of regulation, like food, drugs, and biological products. In more recent days, these difficult definitional questions are compounded along an additional axis—questions of deference in the context of administrative law. Eli Lilly & Co. v. Kennedy is found at the intersection of these axes.

Discussion

Legal Background

FDA’s authority to regulate biological products stretches back to 1906, when Congress passed the Biologics Control Act of 1902 in response to the 1901 St. Louis diphtheria tragedy, where diphtheria antitoxin was produced using the blood of a tetanus-infected horse. The Biologics Control Act was eventually superseded by the 1944 Public Health Services Act (PHSA), which attempted to consolidate all of the various laws relating to public health. In its original formulation, the PHSA prohibited introduction of a “virus, therapeutic serum, toxin, antitoxin, or analogous product, or arsphenamine or its derivatives” applicable to the prevention, treatment of disease or injury, unless it was produced in an establishment with the appropriate license into interstate commerce.[1] Being many years before the significant advances of the biotechnology revolution, the 1944 PHSA did not contemplate proteins or antibodies.

Incredibly, it was not until the Biologics Price Competition and Innovation Act of 2009 that Congress amended the PHSA to first include “protein (except any chemically synthesized polypeptide)” before the phrase “or analogous product” to bring proteins within the definition of biological product. In October 2010, FDA asked the public to comment on the scientific and technical factors it should consider if it were to develop a regulatory definition of “protein” or “any chemically synthesized polypeptide.”[2] In asking for comment, FDA acknowledged the “absence of scientific consensus on the distinction between the categories of ‘protein’ and ‘polypeptide’ or ‘peptide’ . . . .”[3]

In a proposed rulemaking, FDA reviewed comments from the public in several dockets following the call for comment and early draft guidance, and also independently reviewed scientific literature and dictionaries.[4] Canvassing the comments, dictionaries, and textbook definitions for protein and polypeptide, FDA observed that there was not a “precise, agreed-upon” definition for protein, polypeptide, and peptide, but that there was agreement on “certain aspects” of the meanings.[5] Notwithstanding—or perhaps because of—the disagreement, FDA explained that it sought to “establish a scientifically reasonable, bright-line rule that provides regulatory clarity and facilitates the implementation of the BPCI Act.”[6]

Summarizing its technical analysis, FDA determined that the terms protein, polypeptide, and peptide all referred to amino acid polymers comprising alpha amino acids linked by peptide bonds.[7] FDA observed that peptides were distinguished from proteins as referring to “smaller, simpler” amino acid chains while proteins were “long, complex” polymers.[8] FDA further observed that proteins were characterized by chains with a “specific, defined sequence” of amino acids.[9] FDA proposed defining “protein (except any chemically synthesized polypeptide)” to mean “any alpha amino acid polymer with a specific, defined, sequence that is greater than 40 amino acids in size.”[10]

Following public comment, FDA finalized the definition of protein as proposed. Altogether, the final rule defined “protein” as:

any alpha amino acid polymer with a specific, defined sequence that is greater than 40 amino acids in size. When two or more amino acid chains in an amino acid polymer are associated with each other in a manner that occurs in nature, the size of the amino acid polymer . . . will be based on the total number of amino acids in those chains, and will not be limited to the number of amino acids in a contiguous sequence.[11]

In finalizing the definition, FDA rejected concerns about lack of consensus or scientific sufficiency, prioritizing setting a “bright-line rule that provides regulatory certainty.”[12]

Notwithstanding FDA’s hope of providing regulatory certainty, the line between biological products and drugs has been repeatedly challenged over the last several years. For instance, the D.C. District Court upheld the agency’s position that Teva’s Copaxone (glatiramer acetate) was not a biological product because it failed to satisfy the definitional requirement of a “specific, defined sequence.”[13] More recently, the D.C. Circuit upheld FDA’s decision that a depot form of octapeptide lanreotide acetate is a drug (and not a biological product).[14]

Factual Background

Eli Lilly is the sponsor of the active retatrutide, which is being studied for the treatment of obesity, obstructive sleep apnea, knee osteoarthritis, and type 2 diabetes.[15] Lilly submitted a Request for Designation in November 2023 asking FDA to classify retatrutide as a biological product.[16] In its Request for Designation, Lilly described retatrutide as:

an alpha amino acid polymer that contains a specific, defined sequence of 41 amino acids cumulatively, comprising a backbone of 39 alpha amino acids and a second (associated) chain of one residue each of gamma-glutamate and 8-amino-3,6-dioxaoctanoic acid (ADO). The associated chain of two amino acids is covalently bound to the backbone of 39 amino acids through an amide bond between the epsilon-amino group of a lysine residue and the carboxyl group of ADO.[17]

Lilly argued that retatrutide met the definition of protein because it had a “specific, defined sequence,” the amino acid chains “are associated with each other in a manner that occurs in nature,” and it is an “alpha amino polymer that is greater than 40 amino acids . . . .”[18] Regarding this last point, Lilly argued that FDA should count all of the amino acids in retatrutide—even an amino acid like ADO, which is not an alpha amino acid.[19]

Despite Lilly’s arguments, FDA declined to classify retatrutide as a biological product, instead determining that it was a drug.[20] According to FDA, retatrutide was “not an alpha amino acid polymer with a specific defined sequence greater than 40 amino acids in size.”[21] In particular, FDA interpreted the definition of protein to require more than 40 alpha amino acids—not just more than 40 amino acids, with at least one being an alpha amino acid. By contrast, FDA found that retatrutide contained 40 alpha amino acids and a single non-alpha amino acid (i.e., ADO).[22] In its letter of classification, FDA stated that it was “common scientific knowledge” that proteins alpha amino acids formed the building blocks of proteins in humans.[23]

FDA also determined that retatrutide did not fall within the definition of biological product by virtue of being “analogous” to a protein.[24] Specifically, FDA found that it would be inappropriate to classify a molecule as a biological product if it was otherwise excluded from the express definition because it failed to meet a fundamental attribute of protein.[25] Here, FDA found that failure to contain more than 40 alpha amino acids was the fundamental defining property.[26] FDA observed that including molecules otherwise excluded in this fashion would “defeat the purpose of the bright line rule . . . .”[27]

Following FDA’s letter of designation, Eli Lilly filed a complaint for a declaratory judgment and injunctive relief on September 3, 2024, in United States District Court for the Southern District of Indiana.[28] Lilly asserted that FDA’s refusal to designate retatrutide as a biological product exceeded the agency’s statutory authority, violated the agency’s regulations, and was arbitrary and capricious.[29] Lilly moved for summary judgment on January 28, 2025 and the Defendants cross-moved for summary judgment on March 18, 2025.[30] The court heard oral argument in August 2025.

Decision

The court granted-in-part Lilly’s motion for summary judgment with respect to the agency’s determination that retatrutide is not “analogous” to a protein, but denied Lilly’s motion regarding FDA’s decision that retatrutide was not a protein. Correlatively, the district court granted the FDA’s motion for summary judgment with respect to FDA’s decision that retatrutide was not a protein but denied the motion with respect to FDA’s decision that retatrutide was not “analogous” to a protein. The district court vacated and remanded to the agency for further consideration as to whether retatrutide was “analogous” to a protein.

The district court found that FDA did not act arbitrarily and capriciously by finding that retatrutide failed to meet FDA’s interpretation of its regulatory definition of “protein.” Because the question was about FDA’s interpretation of its own regulatory definition, the district court applied the Kisor/Auer framework of deference.[31] Kisor/Auer deference is a three-part test.[32] In the first step, a court must decide whether the regulation is “genuinely ambiguous” by exhausting the “traditional tools of construction.”[33] Second, a court must determine whether the agency’s interpretation was reasonable in the face of genuine ambiguity.[34] And third, the “‘the character and context of the agency interpretation’ must entitle it to ‘controlling weight.’”[35] Here, the district court determined that “the regulation unambiguously forecloses Lilly’s reading” and thus, did not reach the second and third Kisor steps.[36]

The district court found that Lilly’s interpretation of the regulatory definition of “protein” was unambiguously foreclosed by the plain language of the regulation. As described above, “protein” is defined as “any alpha amino acid polymer with a specific, defined sequence that is greater than 40 amino acids in size.”[37] The district court reasoned that: “the most natural reading of ‘alpha amino acid polymer . . . that is more than 40 amino acids’ is a polymer with more than forty alpha amino acids. Otherwise, the ‘alpha’ modifier does no work, and the term becomes superfluous.”[38] In addition, the district court credited FDA’s contention that it was common knowledge that amino acid “almost always” means an alpha amino acid.[39] The district court also credited scientific textbooks which, in the district court’s view, confirmed this position.[40]

While the district court acknowledged that courts often presume that drafters intend meaning by including language in one place and omitting the same language elsewhere, the district court found that such a presumption typically occurs when at least one of the phrases has a consistent meaning throughout the statute or regulation.[41] Here, however, the district court found that “alpha amino acid” did not have a consistent meaning throughout the statute. Because the agency made the “deliberate choice” to limit the definition to alpha amino acids, the district court reasoned that “the most natural reading” of alpha amino acid polymer requires all references to amino acids in the definition to mean alpha amino acids.[42]

Quite differently, the district court rejected FDA’s interpretation of the statutory term “analogous” as applied to protein.[43] Unlike for “protein,” where FDA applied the Auer/Kisor framework for agency interpretation of its own regulations, to interpret the statutory term “analogous,” the district court applied the Supreme Court’s newly landmarked Loper Bright decision.[44] As the district court noted, the Supreme Court held in Loper Bright that “courts must exercise independent judgment in determining the meaning of statutory provisions.”[45] Still, the district court also allowed for some measure of deference to an agency—citing even Loper Bright’s acknowledgement that “an agency’s interpretation of a statute ‘may be especially informative to the extent it rests on factual premises within the agency’s expertise.’”[46] Still further, “[e]ven in the wake of Loper Bright, courts are permitted to leave those niche factual questions for the agency to decide.”[47]

Having said that, the district court called into question FDA’s logic and whether FDA’s standard was clearly defined and consistently applied. For one, the court questioned FDA’s logic that the “fundamental defining property” of a protein was the length of amino acids, in view of FDA’s previous position in the Copaxone case that the specific, defined sequence was the fundamental defining property of proteins.[48] Moreover, the district court criticized FDA’s “bright line” approach to “analogous” products—the district court pointed out “[b]y definition, then, ‘analogous products’ need not fit into a ‘bright line’ rule for any category of biological products.”[49] As a result, the district court found that, “[b]y requiring ‘analogous’ products to meet each and every requirement for a ‘protein,’ FDA effectively reads ‘analogous product’ out of the statute.”[50]

On February 12, 2026, Eli Lilly appealed the district court’s decision that retatrutide did not meet the definition of protein.

Impact of the Decision

The Eli Lilly decision deserves attention for a number of reasons—some quite practical and others increasingly esoteric.

Most practically, the decision provides some clarity, at least provisionally, about whether the definition of protein requires more than 40 alpha amino acids—or, as Lilly contends, whether it requires a total of 40 amino acids, some of which are alpha amino acids. The district court’s decision also raises questions about the agency’s interpretation of “analogous” here and elsewhere. And if the agency’s approach to interpreting “analogous” is untenable, the decision here raises the question of how FDA will structure its analysis regarding this statutory term in the future.

The district court’s decision in Eli Lilly also provides a data point in the ever-expanding wake of Loper Bright. In the Eli Lilly decision, we saw an example of a court applying Loper Bright (at least with regard to FDA’s interpretation of “analogous”) to a decision with deep scientific underpinnings. Not only that, we saw a willingness of the district court to still extend some deference to the agency using dicta in Loper Bright itself as a basis. Notwithstanding this potential deference, the district court still broadly struck down the agency’s interpretation. Thus, while courts may find ways to acknowledge agency expertise, district courts still have substantial leeway to vacate agency decisions.

The district court’s opinion into the agency’s interpretation of its own regulation defining protein user Kisor/Auer deference is another takeaway. While Loper Bright did not overrule Kisor/Auer deference, after Loper Bright, some scholars wondered whether Kisor/Auer remained a stable doctrine.[51] Notwithstanding these concerns, we see in Eli Lilly an example of a court upholding an agency interpretation using the Kisor/Auer framework (although, admittedly not based on any deference due the agency).

And finally, on the spectrum esoteric, the facts of Eli Lilly highlight the tensions and difficulties inherent in attempts to provide statutory and regulatory definitions of FDA-regulated products that straddle the line between nature and artifact. Products that are peptides, polypeptides, or proteins carry with them the ontological baggage from long unsettled philosophical disputes that go beyond the typical (and also difficult) vagaries of molding language to fit policy goals. In Eli Lilly, these difficulties played out in real time, as we see the parties weighing the benefits of a “bright line rule” against a lack of precise scientific consensus on terminology.

* Andrew Wasson is a partner and Chair of Haug Partners’ FDA practice in the New York office. Andrew’s practice relates to patent law and FDA regulatory law in the life sciences, including Hatch–Waxman litigation at the district court and appellate levels, Inter Partes Review proceedings, due diligence, and strategic counseling.

[1] Pub. L. 78-410.

[2] 75 Fed. Reg. 61497 at 61499 (Oct. 5, 2010).

[3] Id.

[4] 83 Fed. Reg. 63817 at 63819 (Dec. 12, 2018).

[5] Id. at 63820.

[6] Id.

[7] Id.

[8] Id.

[9] Id.

[10] Id.

[11] See also 21 C.F.R. § 600.3(h)(6).

[12] 85 Fed. Reg. 10057 at 10060 (Feb. 21, 2020).

[13] Teva Pharms. USA, Inc. v. U.S. Food & Drug Admin., 514 F. Supp. 3d 66 (D.D.C. 2020).

[14] Ipsen Biopharmaceuticals, Inc. v. Becerra, 108 F.4th 836 (D.C. Cir. 2024).

[15] Eli Lilly at *4.

[16] Id. at *4–5.

[17] Id. at 5.

[18] Id. at *5–6.

[19] Id. at *6.

[20] Id. at *6.

[21] Id.

[22] Id.

[23] Id. at *6–7.

[24] Id. at *7.

[25] Id.

[26] Id.

[27] Id. at *7–8.

[28] Id. at *8.

[29] Id.

[30] Id.

[31] Id. at *11.

[32] Auer v. Robbins, 519 U.S. 452 (1997); see also Kisor v. Wilkie, 588 U.S. 558 (2019).

[33] E.g., Kisor, 519 U.S. at 574.

[34] Id.

[35] Eli Lilly at *12, quoting Kisor at 577–79.

[36] Eli Lilly at *12.

[37] 21 C.F.R. § 600.3(h)(6).

[38] Eli Lilly at *14.

[39] Id. at 15.

[40] Id. at *15–16.

[41] Id. at 14–15.

[42] Id. at 15.

[43] Eli Lilly at *18–*23.

[44] See id. at *18; see also Loper Bright v. Raimondo, 603 U.S. 369 (2024).

[45] Eli Lilly at *18, quoting Loper Bright, 603 U.S. at 394.

[46] Id. quoting Loper Bright, 603 U.S. at 402.

[47] Id. at *20 (citing Seven Cnty. Infrastructure Coal. v. Eagle Cnty., 605 U.S. 168 (2025)).

[48] But see Teva Pharms. USA, Inc. v. U.S. Food & Drug Admin., 514 F. Supp. 3d 66, 116 (D.D.C. 2020).

[49] Eli Lilly at *22.

[50] Id.

[51] See, e.g., Notice and Comment Blog, Chad Squitieri, Auer after Loper Bright, Yale Journal on Regulation (Oct. 15, 2024).