U.S. Supreme Court Referral: Diverging Approaches To Antibody Epitope Claims Between The EPO and USPTO?

Antibody patents are one of the most valuable of all intellectual property rights, with the annual global therapeutic monoclonal antibody market expected to be worth $300 billion (USD) by 2025 according to a recent report . Obtaining patent protection for antibodies involves different challenges in different jurisdictions, but the good news for antibody developers is that, in Europe, at least, it is possible to obtain broad claims based on the target epitope. In the U.S., on 3 November 2022, the Supreme Court granted a petition by Amgen for a review that could influence the breadth of allowable epitope patent claims, as we discuss below.

Antibodies are Y-shaped tetrameric protein molecules consisting of two heavy (H) and two light (L) di-sulphide-linked polypeptide chains. Each of the H- and L-chains has a variable domain (termed VH and VL), and each of the VH and VL domains include complementarity determining regions (CDRs), that determine the antibody’s binding specificity to the target part of the antigen molecule (i.e. the epitope).

Many antibody patent applications are directed not to conventional antibodies per se, but to antibody-related technologies, including, for example, truncated, bi-specific, and drug-conjugated antibodies, antibody fusion proteins, and cell surface receptors, such as chimeric antigen receptors (known as CARs) comprising antibody protein domains.

Patent protection for antibodies per se, on the other hand, may be obtained using patent claims that define the antibodies on the basis of their structural and/or functional properties, their use, for example, in specific medical treatments, and/or their method of production. In practice, to provide robust protection for a valuable antibody, all of these approaches are likely to be used in separate patent applications, with different filing dates where appropriate, in order to maximise the scope and duration of the patent rights.

Patent Claims Based on Structural Properties of Antibodies
Both in Europe and the U.S., antibody patent claims that refer directly and exclusively to the amino acid sequences of the antibodies provide potentially the most robust patent protection for specific antibodies, albeit at the expense of claim breadth. Such claims may, for example take the form of: “An isolated antibody comprising a VL domain comprising the amino acid sequence of SEQ. ID. No. 1 and a VH domain comprising the amino acid sequence of SEQ ID. No. 2.”

The European Patent Office (EPO) takes the view that if the inventive step of the claimed antibodies is based on an improved binding affinity, then the full VH and VL domain sequences of the antibody must be recited in the claim. Broader antibody structure-based patent claims may be possible at the EPO, however, if the inventive feature of the antibodies relates to a different surprising advantage, such as, for example, an improved therapeutic property, a reduced toxicity or immunogenicity, or an unexpected species cross-reactivity. In such cases, it may be possible to obtain patent protection for an antibody that is structurally defined on the basis of the CDR sequences only. In most cases, all six CDRs must be recited in the claims, although it is theoretically possible to claim fewer than six CDRs, for example, if appropriate supporting data are available, or if the claim additionally includes a functional limitation.

Broader antibody structure-based patent claims are also potentially allowable in both Europe and the U.S., as long as the specification includes an appropriate level of supporting data. For example, if the data are available, then it may be possible to obtain patent claims for sequences based on a level of sequence identity, variability in specific regions, or alterations in selected amino acids of the antibody. Such data could include, for example, evidence that certain amino acid residues in the antibody sequence are not involved in binding to the specific antigen, and/or that particular residues may be substituted without affecting the binding properties of the antibody or the claimed technical effect.

Patent Claims Based on the Target Antigen (Epitope Claims)
In general, the broadest antibody patent claims are those based on the function of the antibody, such as the target epitope to which the antibody specifically binds. For example, such claims may have the form “An isolated antibody binding to epitope X”.

This type of patent claim, which is sometimes referred to as an “epitope claim”, has historically been widely used because it provides the possibility of obtaining broad protection for any antibody binding to the same epitope, regardless of the antibody sequence.

Claims in this format, however, based on the specific binding properties of the antibody, have become progressively more difficult to obtain in the U.S. In the U.S., 35 U.S.C. § 112 sets forth (i) a written description requirement, and (ii) an enablement requirement, for all patent specifications. To meet the written description requirement, a patent specification must demonstrate that the inventors were in possession of the claimed invention at the filing date of the patent application. To meet the enablement requirement, a patent specification must provide sufficient information for a person skilled in the art to practice the claimed invention without having to engage in undue experimentation.

The USPTO previously used the “antigen test” in which a functional antibody would satisfy the written description requirement if the patent specification adequately described the target antigen, usually on the basis of the amino acid sequence, and if the production of the claimed antibodies was conventional or routine; even if the specification did not include a structural definition of the claimed antibodies.

In a number of recent decisions, however, the Federal Circuit has undermined and subsequently disposed of the antigen test. The recent decision of the Federal Circuit in Amgen v. Sanofi (No. 20-1074, Fed. Cir. 2021) confirmed the threshold for the written description requirement for functional antibody claims. The Federal Circuit found that claims directed to anti-PCSK9 antibodies defined only by their target epitope placed an excessive burden on the skilled person, on the basis that they would have to screen millions of antibody candidates to determine which antibodies were encompassed by the functional definition of the claim. This is despite the patents-in-suit providing the sequences of at least 20 example antibodies, and the 3D structures of two of the antibodies, including structures showing how these antibodies bind to epitopes on the PCSK9 receptor to thereby prevent binding of the ligand to the receptor.

Amgen subsequently filed a petition for the U.S. Supreme Court to review the decision, arguing that that the Federal Circuit’s standard for enablement is “impossible to satisfy”, and that it requires the claims to be enabled throughout their full scope, even if there is no evidence that there are any particular embodiments that would require undue experimentation. According to Amgen, the appropriate standard is the provision of sufficient disclosure to be able to make and use the invention, which does not require disclosure throughout the entire scope of the claims.

The fact that the Supreme Court granted Amgen’s petition for a review is interesting given that the United States Solicitor General had recommended that the Court deny the petition. The Supreme Court granted a review with respect to the following question presented in Amgen’s petition:

2. Whether enablement is governed by the statutory requirement that the specification teach those skilled in the art to “make and use” the claimed invention, 35 U.S.C. §112, or whether it must instead enable those skilled in the art “to reach the full scope of claimed embodiments” without undue experimentation—i.e., to cumulatively identify and make all or nearly all embodiments of the invention without substantial “‘time and effort,’” Pet.App. 14a (emphasis added).

In essence, the Court will consider the balance between the claim scope that a patent applicant may be entitled to expect in view of the contribution that their invention makes to the art, versus the need to avoid the inhibitory effect of overly broad claims on future development of related, but undisclosed, embodiments.

For the time being, a high bar for the written description and enablement requirements exists, and it is extremely challenging to achieve grant of a functional antibody (epitope) claim in the U.S.

However, antibody developers will be pleased to hear that this is not necessarily true for Europe, where it is still possible to obtain broad epitope claims, at least in principle. Inventive step, rather than sufficient disclosure, is likely to present more of a challenge in Europe in relation to epitope claims, however. The EPO takes the view that a claim to a new antibody binding to a known target does not involve an inventive step unless the antibody provides a surprising technical effect (EPO Guidelines II-5.6.2).

That said, the EPO’s Guidelines for Examination explicitly permit the definition of an antibody in a patent claim using various different functional parameters, including the target antigen or epitope (G.II, 5.6.1). Moreover, such claims are not only theoretically allowable, but have been tested and survived scrutiny by the EPO Boards of Appeal.

For example, in T 1964/18, the claimed antibody was defined on the basis of specifically blocking the NKG2A receptor which inhibits natural killer (NK) cells but not the related, activatory, receptors NKG2C and E, and binding the same epitope on NKG2A as a deposited antibody, such that binding of the antibody to its target on a NK cell activates NK-cell killing. The prior art cited in relation to inventive step differed from the claimed antibody in that it additionally bound the NKG2C and NKG2E receptors, as well as NKG2A. The Board of Appeal thus considered the question of whether it would have been obvious to modify the prior art antibodies to block only the activity of NKG2A, as claimed, and not NKG2C and NKG2E.

Based on a teaching in the prior art that blocking all three receptors with an antibody would lead to NK cell activation, the Board of Appeal found that a skilled person would not have been motivated to produce antibodies that specifically bind NKG2A, but not NKG2C or HKG2E. Accordingly, the broad functional epitope claim was found to be inventive. The allowed claim read:
1.           A monoclonal antibody or a fragment thereof characterized by:
a. specifically binding to NKG2A, wherein the antibody or fragment is characterized by binding the same epitope on NKG2A as the antibody produced by the cell deposited at the CNCM under accession number I-3549;
b. not specifically binding to human NKG2C;
c. not specifically binding to human NKG2E;
d. not binding, via its Fc region, to a human Fc gamma receptor; and
e. when bound to NKG2A on a human NK cell, causing said NK cell to lyse a target human cell bearing HLA-E or Qalb on the target cell surface, when said target cell comes into contact with said NK cell.

Currently, in the U.S., in the corresponding patent application, similar functionally defined epitope claims were not considered allowable, for failing to satisfy the written description requirement. Related U.S. patent applications are still in prosecution, but the only patent claims that have been granted in the U.S. relate to a method of treatment comprising the use of the antibody defined on the basis of all six CDR sequences (US 10,160,810), and a cell line capable of specifically producing the antibody (US 8,993,319).

Clearly, Applicants face increasingly distinct and sometimes conflicting challenges when attempting to protect antibody inventions across different jurisdictions, presenting particular challenges when drafting patent specifications. One clear message, however, is that the specification should include as many different examples of the claimed antibodies as possible, and evidence that the antibodies meet the functional requirements. Ideally, the example antibodies should be as diverse as possible, highlighting the importance of consistent features, such as specific amino acid residues, that are conserved across the various examples.

If at all possible, it would be advisable to generate and include experimental data relating to the interaction of the antibody with the target epitope, or residues of the antibody involved in binding the antigen, as this type of information is likely be very helpful to support the validity and inventive step of the claimed antibodies, in all jurisdictions. In Europe specifically, these data would make it possible to claim not only the amino acids of the antibody’s CDRs that bind to the target antigen, but also those residues constituting the epitope itself. Moreover, information on whether the epitope is a linear or conformational epitope can also be valuable, both as a way of potentially distinguishing over prior art antibodies, and, if associated with a surprising technical effect, also to support an inventive step.

Antibody patent applications should, therefore be drafted in such a way that they provide the possibility of achieving the broadest available scope of protection in different territories, including broad epitope claims in Europe.

If you would like more information about patenting antibodies, please contact our specialist Antibody Team, who would be happy to assist. We have considerable experience in drafting and prosecuting antibody patent applications, and also in proceedings before the European Patent Office Opposition Divisions and Boards of Appeal concerning opposed antibody patents. Please also let us know if you would like to receive our regular antibody newsletter.