How Amgen v. Sanofi will affect innovation and the pharmaceutical industry.
By Timothy Bonis
Millions of patients rely on monoclonal antibodies. The global market in 2022 was $210B with a compound annual growth rate of 11 percent. Monoclonal antibody patents are now some of the world’s most valuable intellectual property.
Monoclonal antibody patent law has been in turmoil for the past two decades. The Court of Appeals for the Federal Circuit (the “Federal Circuit”) has repeatedly raised the standards for antibody patents, forcing patentees to rethink how they protect their inventions. In April 2023, the increasingly stringent standards of the Federal Circuit were affirmed by the Supreme Court. In Amgen v. Sanofi, the justices unanimously upheld the invalidation of two antibody patents, potentially transforming patent law across biotech.
This post reviews the emerging post-Amgen literature, distilling the ruling’s likely impact on innovation and the pharmaceutical industry. A companion post addresses the ruling’s legal significance.
This post has three parts. Part I describes monoclonal antibodies and why Amgen is important. Parts II and III address innovation, first by discussing whether Amgen is reflective of modern antibody science and second by exploring how the ruling may affect the pharmaceutical industry.
Part I: Background: Why Amgen v. Sanofi Matters
Monoclonal antibodies refer to antibodies that have been designed to bind to a specific site (referred to as an antigen) and thereby achieve a diagnostic or therapeutic purpose. They are too large to be synthesized through conventional chemistry. There are several approaches to synthesizing monoclonal antibodies. Contemporary approaches involve recombinant DNA technology. Older approaches used mouse hybridoma cells, which offered limited insight into the chemical structure of the antibodies. By contrast, the newer approaches provide a more detailed understanding of their structure.
Historical patent practices reflected scientists’ limited ability to understand the structure of individual antibodies. In the 1980s, 1990s, and early 2000s, antibody patents often consisted of a description of the target antigen and the characteristics of the interaction between the antibody and the antigen, such as where and how strongly the antibody bound. These parameters marked out what antibodies were claimed by the patent, even when the inventor did not know exactly what those antibodies were or even how many antibodies the patent protected. In patent law, these kinds of descriptions are called “functional claims,” meaning they describe what an item does more than what it is. They were also “genus claims,” meaning they were claims to a class of related items, not a single item; it is a scientific reality that a claim for ‘an antibody binding to X protein that targets regions a, b, c, and d with Y strength’ probably encompasses thousands of different antibodies.
Until the early 2000s, the Patent and Trademark Office (PTO) and the courts upheld this method of patenting antibodies. Early antibody cases (like Hybritech v. Monoclonal Antibodies Inc., 1985) protected large genus claims. The PTO had a rule called the “antibody exception”; by statute, patents need a “written description” of what they protect, but the PTO promised to grant antibody patents that described only the antigen, not the antibody itself.
Over the past two decades, however, courts have reined in functional and genus claims for antibodies, emphasizing narrow patents with structural details. Even claims that use a robust mix of structure and function are now vulnerable. For instance, in AbbVie v. Janssen (2014), the Federal Circuit invalidated an antibody genus patent that included over 300 representative amino acid sequences from the genus for failing the written description test. Rulings like AbbVie changed the law, forcing the PTO to eliminate the antibody exception in 2018.
Amgen v. Sanofi dealt with a related patent requirement, enablement. The idea of enablement is that a “person having ordinary skill in the art” (a “PHOSITA”) should be able to make and use the invention from the patent’s disclosures. In the past, inventors usually “enabled” antibody patents by making a cell culture publicly available and detailing how to screen the antibodies in the claim from the output of the culture. That method has long been obsolete, so when Amgen filed its controversial patent in 2014 (a genus claim over some of the antibodies that bind to the protein PCSK9), it included a few structural parameters and a “roadmap” for producing the claimed antibodies. The Federal Circuit and later the Supreme Court felt that this disclosure was insufficient to satisfy the enablement requirement. The courts espoused a doctrine of “full scope” enablement, where the disclosure in a genus claim must explain how to make every embodiment, and argued that a PHOSITA would have to engage in “undue experimentation” to practice the “full scope” of the claim.
(See the companion piece for more information on antibody patent requirements and jurisprudence).
Part II: Does Amgen Align with Modern Antibody Science?
In their article The Antibody Patent Paradox (2023), Mark Lemley and Jake Sherkow argue that the “full scope” requirement is poorly reflective of antibody science and threatens innovation. They think it is unreasonable to hold antibodies to the full scope standard because they are such diverse molecules that no amount of disclosure could ever describe how to make the entirety of any useful patent. Lemley and Sherkow also criticize the Federal Circuit for demanding “representative structures or common structural features” in antibody genus claims (which was not the new holding in Amgen; see AbbVie v. Janssen 2014). They point out that in proteins like antibodies, unlike in smaller molecules, structure and behavior are not closely related; the structure-function relationship is exceedingly complex, so it may be inappropriate to demand structural features that are “common” to a genus — any genus marked out by structural features would be too narrow to offer meaningful protection.
Others have made the opposite argument. Katlin Taylor and Sean Tu applaud the holding in Amgen, emphasizing how, because structure and function are not closely linked in antibodies, broad, functional patents can block inventors from creating new drugs that are structurally unrelated to existing antibody therapies but have similar functions. This could prevent patients from accessing treatments that improve on or offer different benefits from existing therapies, even when the new drug and existing drug are completely unrelated. Taylor explains:
Often when a particular class of antibodies is being developed for therapeutic drug use, each subsequent antibody created within that class has one, if not more, additional, or enhanced feature(s). If the broad functional claims over an antibody class, like those in Amgen, were upheld, this could prevent these incremental improvements within a particular antibody class from being developed.
This is not an abstract risk. In the 1980s, the PTO granted the Chiron Corporation broad genus protection for antibodies that bind to HER2, a protein that helps breast cancer cells grow. In 2004, the Federal Circuit invalidated those patents in a case similar to Amgen (Chiron v. Genentech), giving Chiron and Genentech narrow patents for anti-HER2 antibodies. Under this narrow protection, Genentech introduced the breast cancer drug Herceptin, but because Herceptin’s patent was limited, MacroGenetics was able to introduce a different, more efficacious anti-HER2 antibody in 2020 (Margenza®).
Whereas Lemley and Sherkow criticize the Federal Circuit for demanding ever more structural details, Christopher Holman believes this trend appropriately reflects the progress of antibody science. Biochemists have not only developed technology to understand antibodies’ structure over the past three decades, they have also introduced numerous antibody-based drugs. Structural features are more relevant in pharmaceutical applications than in assays and diagnostics, which were the primary application of antibodies when functional claims were the standard practice.
Part III: How Will Amgen Affect the Pharmaceutical Industry?
Whether Amgen will be good or bad for established drug companies or startups remains unclear. Major industry players filed amici briefs on both sides, AbbVie, Bristol Meyers Squibb, and Merck supporting Amgen, and Pfizer, Eli Lilly, Genentech, and AstraZeneca supporting Sanofi. The Alliance of U.S. Startups supported Amgen, and the decision prompted a range of predictions from experts.
Almost all the pro-Amgen briefs share the argument that only functionally defined genera give inventors in chemical and biological fields broad enough protection to justify investment. AbbVie’s position is representative: “Genus claims with functional limitations promote the progress of science…[they] are especially important in chemistry, pharmaceuticals, and biotechnology, where breakthrough innovations invariably require very significant investments of time and money.” The Alliance of U.S. Startups adds the concern that biotechnology startups need genus protection to attract early-stage funding and acquisitions. The brief attributes a twenty-year decline in the share of venture capital funding received by pharmaceutical startups to the Federal Circuit’s constriction of genus claims in chemical fields.
The pro-Sanofi briefs, in addition to weighing in on the facts of the case (e.g., how many antibodies were in Amgen’s genus, how thorough was Amgen’s disclosure), argue that functional claims like Amgen’s stifle innovation. Many repeat Taylor and Tu’s warning that functional claims to antibodies would block companies from marketing new antibodies with applications that are similar to existing drugs. Genentech and AstraZeneca’s stance is typical: Upholding Amgen’s patent would “foreclose efforts to discover other and better types [of therapies].”
In Paradox, Lemley and Sherkow echo the pro-Amgen briefs. They agree that only genus claims with some functional components provide “effective protection” for antibodies and forecast a fall in investment and innovation, emphasizing “the uncertain and costly road from antigen identification to therapeutic development.” While they do not advocate a return to purely functional claims, they think innovation would be best promoted by a regime that gives patentees some power to block functionally similar but structurally unrelated antibodies.
Sean Tu presents a different outlook on functionally similar antibodies. He argues that narrow protection opens opportunities for the industry, and not necessarily at the expense of existing antibody makers. One of Tu’s examples is Centocor’s Remicade® and Abbott’s Humira®. Both drugs are antibodies that bind to the protein TNF-α, so in 2011, Centocor (which held a broad genus claim for antibodies binding to TNF-α) sued Abbott for infringement. The Federal Circuit invalidated Centocor’s patent, allowing Abbott to sell Humira®. Humira® and Remicade® are not biosimilars; they have varied mechanisms of action in different disorders. The narrower protection allowed Humira® to become the world’s most profitable drug while Remicade® remained a multi-billion dollar product for Centocor (now J&J). It also enabled two other non-biosimilar, anti-TNF-α antibodies to enter the market, Enbrel® and Entyvio®.
Interestingly, Mark Lemley (author of Paradox) implies that in other ways Amgen’s impact on the industry will be limited. In his 2020 article The Death of the Genus Claim (published before Amgen but highly related), Lemley and his co-authors observe that harsh laws against genus claims have not blunted the pharmaceutical industry’s patent activity or profitability. This could reflect inertia among attorneys and the Patent Office, but the article suggests the more significant factor is that the industry relies on protecting species, not genera; drugs that get regulatory approval are single molecules or antibodies, and a generic must be identical or biosimilar. If a drug company can get protection for one species, it will enjoy a monopoly through approval rules and barriers to entry.
