EPO Guidelines on Antibodies: an assessment of adherence

In March 2021, the European Patent Office (EPO) Guidelines for Examination were updated to include a section devoted to the examination of antibody-related patent applications (Guidelines G-II, 5.6). This new section provided a much anticipated outline of the EPO requirements for different approaches to defining antibodies in patent claims (Guidelines G-II, 5.6.1) and helpful guidance in relation to the inventive step of antibodies in Europe (Guidelines G-II, 5.6.2). The new antibodies section of the EPO Guidelines was discussed in detail in our article “An antidote to antibody patent confusion in Europe”. Here we look at the effect the new Guidelines have had on the prosecution of antibody patent applications over the last year.

Antibodies are typically in the form of immunoglobulin G and comprise a Y-shaped protein consisting of two identical light chains and two identical heavy chains, each folded into constant and variable domains. Each of the heavy and light chain variable domains (referred to as VH and VL) contains three hypervariable regions, referred to as “complementarity-determining regions” (CDRs), which together provide the antigen-specific binding interface.

In the March 2022 revision of the EPO Guidelines, amendments to the antibody sections have been made. However, these changes are relatively minor, and appear to be generally intended only to more explicitly acknowledge the patentability of unconventional immunoglobulin structures, such as heavy-chain-only antibodies and antibody derivatives, such as antibody fragments, bi-specific or multi-specific antibodies, and antibody fusion products.

Since the introduction of the new antibody Guidelines in March 2021, the EPO has granted in the region of 300 patents relating to new antibodies and antibody derivatives. We have analysed the claims of these granted patents to investigate how EPO Examiners have started implementing the new antibody Guidelines, and in the following brief review, a number of emerging trends and perspectives are discussed.

As an initial observation, very few antibody patents granted since the introduction of the new Guidelines (less than 5%), define the antibodies in terms of their method of production, such as reciting an immunisation strategy, or on the basis of a deposited hybridoma.

The vast majority (about 85%) of all antibody patents granted in this period define the antibody, at least in part, on the basis of an amino acid sequence of the antibody. Of these patents, about 55% recite in the granted claims the sequences of all six CDRs. A further 40% recite the full heavy and light chain variable domain sequences, and about 2% recite the full antibody sequence in the claims. In view of these figures, it seems that it is becoming increasingly challenging to obtain a broad scope of patent protection, without reciting at least the sequences of all six CDRs.

The Guidelines state that it is possible to define an antibody on the basis of less than 100% sequence identity to the variable domains or CDRs when this feature is combined with a functional feature. This appears to be the most common approach taken by antibody patentees, and about 55% of granted patents define the antibody using a combination of both a specific sequence and a functional property. Functional limitations typically include minimum antigen binding affinity, binding specificity, or blocking capability, for example.

Using this approach, a number of patents have been granted using a broader sequence definition. In some examples, the antibody sequence is defined on the basis of percentage sequence identity (such as at least 90% sequence identity), or a maximum number of residue substitutions, relative to specific VH and VL sequences. Another approach that has been used successfully is to define a number of possible alternative sequences for each CDR, or the presence of conservative amino acid substitutions in specific positions of the CDRs.

The Guidelines also provide scope for defining an antibody using less than all of the CDRs, if experimental support is provided to evidence that the omitted CDR(s) are not required for epitope binding. However, there do not appear to have been any patents granted that take advantage of this approach. This may be a reflection of the involvement of the CDRs in antibody binding, or it may be that applicants are unable to meet the high threshold for the level of evidence required. We may see more patents defining antibodies in this way in the future as this type of data become easier to obtain, and as applications are drafted and filed with the knowledge and certainty provided by the new antibody Guidelines in relation to the potential importance of this type of data.

The Guidelines also provide for antibodies to be defined on the basis of the target antigen. Perhaps surprisingly, however, only a small proportion of the antibody patents granted since the introduction of the new Guidelines define the antibody on this basis. These patents represent only about 5% of the patents granted since the introduction of the antibody Guidelines.

The Guidelines indicate that, for antibodies defined in this way, no sequence variability and no open language may be used in the definition of the antigen. Interestingly, however, Examiners do appear to have been willing to relax these requirements for the antigen definition slightly. For example, patents have been granted in which the antibodies are defined on the basis of binding to an antigenic peptide that includes a specific sequence, or on the basis of binding to the same epitope as a comparable antibody.

Thus, it appears that defining an antibody on the basis of the target antigen may be a promising approach to obtain a potentially useful scope of protection. Again, this is an approach that we may see used more frequently in the future, as applications are drafted and filed in light of the new Guidelines.

One of the clear messages from our analysis of patents granted since the introduction of the antibody Guidelines is that if at all possible, it would be advisable to obtain 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.

If you would like more information about patenting antibodies, please check out our antibodies web page which also has our key team contacts should you wish to reach out.

Matthew Handley, Partner - Venner Shipley LLP
mhandley@vennershipley.co.uk

Kyle Burgess, Associate - Venner Shipley LLP
kburgess@vennershipley.co.uk