Antibodies – EPO user consultation on new EPO guidelines

AA Thornton’s Pharma & Biotech experts contribute to the European Patent Office’s peer consultation on the EPOs Guidelines for Examination; Part G chapter II,  section 5.6 concerning Antibodies.

On 1st March 2021 the EPO published new guidelines for the examination of claims to antibodies which has raised some some specific concerns regarding the proposed wording in Part G chapter II, section 5.6.  After discussion on the topic with industry clients, Partners Craig Turner, Geoff Hussey, Mike Jennings and Dan Byrne have provided the following submission to the EPO with recommendations.



AA Thornton is a firm of patent attorneys, trade mark attorneys, solicitors and barristers providing intellectual property services to clients in a great many countries and industries. We were one of the UK’s first specialist IP firms – supporting industry since 1911. We are fully supportive of the EPO’s efforts to increase and improve consultation with all users of the European patent system. We applaud the progress the EPO has already made in this important area.


1. We are very grateful to the EPO for the opportunity afforded to users to comment on the proposed new Guidelines for Examination in the EPO. The EPO’s Guidelines have become a very useful tool for examiners and the EPO’s worldwide user community, thanks to frequent updates to reflect Board of Appeal decisions and thanks to a great deal of effort to ensure that the Guidelines are clear and consistent and useful. The Guidelines have an important role to help Examining Divisions and Opposition Divisions achieve consistency and predictability of outcomes. We have some recommendations that we hope will improve clarity and consistency.

1.1 We have some specific concerns regarding the proposed wording for the Guidelines for Examination in the EPO, Part G chapter II,  section 5.6  concerning Antibodies. We are aware that there may be a number of submissions concerning this section, and in particular we are aware of proposals being discussed by the European Federation of Pharmaceutical Industries and Associations (EFPIA).   AA Thornton represents a number of clients in the antibody field, and are generally supportive of the approach we understand EFPIA to be taking on section 5.6.  Nevertheless, the comments which follow reflect our own views on this section.


2. 5.6.1 General Remarks

2.1  We are concerned that the general remarks section oversimplifies the antibody field, and that some of the proposed terms are potentially misleading or ill-defined (e.g. what is a “conventional antibody”?) and could result in misrepresentation of this technical field.   Terms such as “conventional antibodies”, “designed by nature”, “Y–shaped proteins” and “naturally produced by plasma B cells” are some specific examples of wording that we consider to be unhelpful.  Many useful antibodies, including therapeutic antibodies, are not produced by nature, nor are they necessarily Y-shaped.    Yet such antibodies are rightly capable of patent protection, and this is of course recognised by the EPO.    We propose that some updating of the currently proposed wording would be useful, to better reflect the current state-of-the-art.

2.2  We would support amendment of the wording of this section along the following lines:

Proposal for amendments of 5.6.1 General remarks section

Antibodies may be large, Y-shaped proteins composed of two identical light chains and two identical heavy chains, both containing variable and constant domains. This description corresponds to a format of antibodies produced in the human body. Antibodies bind specifically to antigen targets via the antigen binding region which contains complementarity-determining regions (CDRs).

Formats based on other immunoglobulin structures are also known, such as heavy-chain-only antibodies that consist of only two identical heavy chains (with variable and constant domains) and the antigen-binding region consists of a single variable domain with three CDRs, a format derived from camelid antibodies.

Knowledge of the structure-function relationships of parts of the antibody may allow for the provision of derivatives for a multitude of applications. Variants of antibodies, antibody fragments, bispecific or multispecific antibodies, and antibody fusion products are commonly designed and produced.

In general, antibodies can be defined by (but are not limited to):

(a) their own structure (amino acid sequences);

(b) nucleic acid sequences encoding the antibody;

(c) reference to the target antigen;

(d) target antigen and further functional features;

(e) functional and structural features;

(f) the production process

(g) the epitope

(h) the hybridoma producing the antibody.


3. Definition by structure of the antibody

The proposed wording for this section calls for the antibody to be defined by at least six CDRs to fulfil the requirements of Article 84 EPC. We think this restriction is unnecessary, and again is an oversimplification.  Many antibodies, including those based on a single light or heavy chain, do not possess six CDRs.  We would suggest a very modest amendment whereby the proposed wording of “six CDRs” is replaced by “required CDRs”.  The final paragraph of section clearly recognises that specific antibody formats exist which allow for epitope recognition by fewer than six CDRs. Accordingly, we think it would be clearer for the rest of the section to be made consistent with this.

3.1 We would support amendment of the wording of this section along the following lines:

Proposal for amendments of Definition by structure of the antibody

For an antibody to be uniquely defined by its structure only and its binding specificity, all the CDR sequences required for such binding need to be defined to fulfil the requirements of Art. 84.

CDRs when not defined by their specific sequence must be defined according to a numbering scheme, for example, chosen from that of Kabat, Chothia or IMGT.

If an antibody is defined by fewer than the required CDRs sequences, the claim will be objected to under Art. 84 because it lacks an essential technical feature.

A claim to an antibody defined by its structure by fewer than six CDRs may be considered to fulfil the requirements of Art. 84 if it is experimentally shown that one or more of the six CDRs do not interact with the target epitope or if it concerns a specific antibody format allowing for epitope recognition by fewer CDRs.


4  5.6.2    Inventive step of antibodies

4.1  Our primary concern with this section is the addition of a requirement to show a “surprising technical effect”   in order for an inventive step to be acknowledged for a novel, further antibody binding to a known antigen.  We believe the effect of this would be to add a further requirement to qualify for an inventive step in the antibody field, which requirement is not applied in other fields, thus leading to inconsistency of approach.  We believe the EPO have worked admirably over many years to improve the consistency of the Guidelines, but we think there is a real danger of a backward step being taken here, should the currently proposed wording be approved.  The statutory test provided by the EPC is that an inventive step is to be recognized where the claimed subject matter is not obvious to a person skilled in the art, bearing in mind the state-of-the-art.  A surprising technical effect is an example of one way in which non-obviousness may be shown, but it is not the only way.  For example, in the chemical field, a novel pharmaceutical drug which has the same technical effect as the known drugs (for example, the treatment of a particular disease), but not having a “surprising” technical effect such as an unexpected improvement or advantage over the known drugs,  may  nevertheless be considered inventive  by virtue of its non-obvious structure, for example. In other words, it may be a non-obvious alternative solution to a known problem. Many other similar examples of this could be cited from other fields.  We believe that structurally novel antibodies should be capable of patent protection in the same way, and on the same basis, provided it is not obvious from the state-of-the-art that the alleged technical effect for that novel antibody would be achieved.    In line with the case law of the Boards of Appeal of the EPO, we propose it would be more appropriate and more consistent to make reference to “no reasonable expectation of success”.  This would also be consistent with the existing specific guidance for Biotechnological inventions provided in the Guidelines at G.VII.13.  We think that reference to a “surprising technical effect” could still be retained within the revised guidelines, but with suitable modification  of the language to reflect that it may be one way (out of many) by which support for an inventive step may be provided.


4.2  One further point to note is that we do not believe the Guidelines should  imply  or suggest that antibodies are  necessarily obtained by applying ”routine methods”. It is our experience that examiners often start with this presupposition, and it is frequently not correct. Particularly in the growing field of therapeutic antibodies, a good deal of effort and innovation may have gone into the successful manufacture and production of the claimed antibodies. For example, specific technical difficulties may have been overcome in the engineering and production of the antibodies. We believe this ought to be specifically recognized in the revised Guidelines.


4.3  For the above reasons, we would support amendment of the wording of this section along the following lines:

Proposal for amendments of 5.6.2 Inventive step of antibodies

If inventive step relies on an improved property versus the enabled antibodies of the prior art, the main characteristics of the method for determining the property must also be indicated in the claim or indicated by reference to the description (F-IV, 4.11.1).

If the technical effect involves the binding affinity, and the affinity feature is included in the claims, the method of measuring the affinity must be included. If the antibody is defined by affinity and by sequence by the CDRs, the structural requirements for antibodies inherently reflecting this affinity must comprise the required CDRs (i.e. the CDRs which interact with the target epitope see G-II, paragraph 4). If the technical effect involves binding affinity, and the antibody is defined by sequence, without the affinity feature, then the required CDRs and the framework regions (with some degree of variability in the framework regions) must be included in the claims because the framework regions also can influence the affinity.

If a novel antibody binds to the same antigen as known antibodies, inventive step is not acknowledged solely on the basis that the novel antibody is structurally different from the known antibodies. Arriving at alternative antibodies by applying techniques known in the art is considered to be obvious to the skilled person. The fact that the structure of the thus obtained alternative antibodies, i.e. their amino acid sequences, is not predictable is not a reason for considering these antibodies as non-obvious (see T 605/14, section 24; T 187/04, section 11). On the other hand, inventive step may be acknowledged where changes into the structure bring about differences in the final technical effect or if no antibody to the same antigen has been structurally characterized in the prior art and if there is no reasonable expectation of success from the prior art concerning the impact of such a novel structure (see T0067/11 section 24 and T 2127/16 sections 27 to 35).


If you would like to discuss the guidelines, or our recommendations please contact a member of our expert Pharma & Biotech team members.

Category: News | Author: Daniel Byrne, Geoff Hussey, Mike Jennings, Craig Turner | Published: | Read more