This post is based on a previous article on IPKat.
The US Supreme Court recently ruled in the high profile Amgen versus Sanofi patent dispute. Agreeing with the Federal Circuit (No. 20-1074, Fed. Cir. 2021), the Supreme Court found Amgen’s function and epitope defined PCSK9 antibody patents to lack enablement (Amgen Inc v Sanofi, No. 21-757). The US Supreme Court was keen to stress that its reasoning did not alter the law on enablement. The decision in Amgen v Sanofi thus does not change the status quo whereby achieving functional claim language for antibody inventions represents a significant challenge in the US. The US decision in Amgen v Sanofialso highlights fundamental differences in the underlying assumptions between Europe and the US with respect to antibody inventions and the capabilities of the skilled person.
Enablement and Written Description
In US patent law, there are two distinct sufficiency requirements: enablement and written description. US patent law specifically requires that:
“The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same” (35 U.S.C. § 112(a))
Functional claim languagewill often fall foul of either or both of the enablement and written description requirements. Amgen v Sanofi related specifically to enablement. The question before the Supreme Court was whether the specification of Amgen’s anti-PCSK9 antibody patents contained enough information to enable a skilled person to make and use the broadly claimed invention.
Amgen v Sanofi – Defining an antibody by epitope and function
Antibodies may be defined in a patent claim by their function, their amino acid sequence and/or by the sequence of the target (epitope) to which the antibody binds. Functional and epitope claims for antibodies will provide the broadest scope of protection. Functional claims cover any antibodies with the same function. Epitope claims include in their scope any antibody that binds to the specified epitope. Epitope and functional claims can therefore protect against not only biosimilars of an antibody therapeutic (i.e. antibodies with an identical sequence) but also competitor antibodies binding the same target epitope.
The patents in dispute in Amgen v Sanofi defined a class of antibodies by their function and epitope. The granted claims were directed to antibodies capable of 1) binding to a specific epitope on PCSK9, and 2) blocking the activity of PCSK9. Blocking PCSK9 reduces low-density lipoprotein (LDL) cholesterol in a patient in order to treat cardiovascular disease (US 8829165 and US 8859741). The broad epitope and functional language of the claims covered Amgen’s own monoclonal antibody therapeutic (Repatha) as well as Sanofi’s independently developed competitor anti-PCSK9 antibody (Praluent). Amgen sued Sanofi for infringement and Sanofi counterclaimed that Amgen’s patents were invalid for lack of enablement.
US Supreme Court decision: “the more a party claims, the more it must enable“
The question before the US Supreme court was whether the methods provided in the disputed patents were sufficiently clear so as to enable the claimed invention to a skilled person (35 U.S.C. § 112(a)). In its judgment, the US Supreme Court agreed with the Federal Circuit that the broad claim language in the Amgen patents placed too high a burden on the skilled person attempting to perform the invention. As such, the Supreme Court upheld the Federal Circuit decision that the patents were not enabled. By using the broad claim language Amgen was attempting to claim “an entire universe of antibodies” containing “at least millions of candidates“. However, the Supreme Court reasoned, “the more a party claims, the broader the monopoly it demands, the more it must enable”.
So how much guidance did the disputed patents provide? Amgen’s patents described the sequences of 26 example antibodies falling under the scope of the claims. The specification also suggested two different methods whereby a skilled person might make antibodies binding the claimed epitope and possessing the claimed function. Particularly, the skilled person could either perform a de novo antibody screen for antibodies with the same target epitope and function, or they could mutate the sequence of the example antibodies described in the specification and check for changes in function or binding behaviour.
The Supreme Court accepted that the enablement requirement permits the need for a “reasonable amount of experimentation” on behalf of the skilled person attempting to perform the invention. However, in the case in question, the Supreme Court found that the level of additional work required went beyond what would be considered reasonable. The methods suggested by the specifications, for the Supreme Court, amounted to “little more than two research assignments” involving “painstaking experimentation“. The Supreme Court referenced an existing analogy to highlight the challenge faced by the skilled person trying to perform the invention. The claimed invention was equated to successful combinations of a combination lock, where the inventor had disclosed 26 successful combinations, whilst claiming all possible combinations.
The Supreme Court also emphasised that the law on enablement applied in the case was not to be considered as specific to antibody inventions. Instead, the enablement test applied was to be understood as merely the application of the long-established law on enablement from the mechanical field.
European sufficiency and functional claiming
The sufficiency bar for epitope and functional claims in Europe is far lower than in the US. The point of departure between the two jurisdictions rests in a different understanding of what would or would not be routine for the skilled person.
The Europe Patent Office maintains as a general principle that generating antibodies against a known target epitope is a matter of routine. In a world where many targets are known, convincing the EPO that a new antibody against a known target is inventive is often a significant challenge. On the flip side, the sufficiency bar for antibody inventions is far lower in Europe than it is in the US. In Europe, provided that the patentee can demonstrate that the finding and testing of candidate antibodies requires only standard techniques and tools available to a skilled person, even very broad antibody claims are permissible. It is thus possible to achieve grant in Europe of claims defining an antibody only by its epitope and/or function.
The decision of the US Supreme Court in Amgen v Sanofi confirmed that US patent law makes the opposite assumption to the EPO with respect to antibody development. The Supreme Court in Amgen v Sanofi re-enforced that, in the US, finding and testing candidate antibodies is not considered routine. The US Supreme Court also emphasised that it is not just that it would take a long time for a skilled person to perform the large amount of trial and error required to find more antibodies falling under the claim. For the Supreme Court, the challenge faced by the skilled person in view of the broad claim language somehow represented a qualitatively different challenge that went beyond mere trial and error (page 18, first paragraph).
Plausibility and enablement
There are clear conceptual differences between the US and the EPO on the question of how routine or not it is to make new antibodies to a known target. Nonetheless, there are many parallels to be drawn between the US enablement requirement considered in Amgen v Sanofi and the concept of plausibility employed by the EPO and recently discussed in G 2/21. In Europe, an invention defined by broad functional features must be disclosed sufficiently clear and complete enough so as to enable a skilled person to perform the invention across the whole scope of the claim (Article 83 EPC, see also EPO Guidelines for Examination, F-III-1). There are many similarities between the case law and reasoning underlying the EPO sufficiency requirement and enablement in the US. All that really differs, it appears, is how much the EPO is prepared to permit the need for onerous trial-and-error on behalf of the skilled person attempting to perform the claimed invention.
However, whist the US Supreme Court reasoning in Amgen v Sanofi highlights a clear distinction between the US and Europe on functional and epitope language in general, it is not thereby clear how equivalent claims would have fared in Europe. The European patents corresponding to the US patents invalidated by the Supreme Court were amended in proceedings before the Boards of Appeal to include structural features (i.e. so as to be no-longer reliant on epitope or functional claim language). The validity of claims equivalent to those litigated in the US has thus not been tested in Europe.
Furthermore, the claim language used in the Amgen patents is exceedingly broad, even for an epitope claim. Particularly, the claims of US 8829165 for example do not specify a linear epitope, but instead define a set of only partly contiguous amino acids, any of which the claimed antibody may bind. As recently considered by the EPO Boards of Appeal in T 0435/20, generating new antibodies to a known structural epitope represents a greater challenge than if the epitope is defined as a linear sequence. However, unlike the patent in T 0435/20, the Amgen patents do suggest standard methods for use in functional screening of likely antibodies. It is thus difficult to say whether the Amgen patents would have been upheld in Europe, even taking into account the EPO’s clearly more lenient approach to functional and epitope claim language.
Final thoughts
The US Supreme Court decision in Amgen v Sanofi confirms existing US case law on enablement. Particularly, in view of Amgen v Sanofi it is confirmed that, whilst epitope claims are not ruled out in principle, the enablement bar is so high as to make epitope claims exceedingly difficult (if not quite impossible) to enforce. The prosecution and enforcement challenges for functional and epitope claims in the US thus remain. However, the reasoning in the Supreme Court decision nonetheless leaves open the theoretical possibility of functional and epitope claims provided that “painstaking experimentation” on behalf of the skilled person is not required. I wonder whether the rise of artificial intelligence aided antibody design will eventually (or soon?) erode the barriers to enablement of broad antibody claims that the Supreme Court currently presumes to exist.
Author: Rose Hughes
