Therapeutic antibodies represent a significant area of drug development with 8 of the top 20 selling drugs globally in 2020 being antibody-based therapies. Humira®, Abbvie’s anti-TNFα antibody drug, attracted global sales of over $20 billion in 2020 alone.
Obtaining strong patent protection for therapeutic antibodies is therefore of the upmost importance and value. Patenting antibodies via the European Patent Office (EPO) has become more difficult in recent years due to what is perceived to be the implementation of a higher hurdle for patentability in areas such as inventive step and enablement.
In a series of two articles dedicated to the topic, we provide an update on the current requirements for patenting antibodies and related biologics at the EPO. In this first article, we look at the different ways an antibody can be claimed at the EPO.
Defining Antibodies by Structure
Immunoglobulin G (IgG) antibodies, the basic functional unit of most therapeutic antibodies, are large Y-shaped molecules comprising two identical heavy chain and light chain subunits. Each heavy chain and light chain are comprised of a constant domain region and variable domain region. The variable domain regions comprise sets of complementarity-determining regions (CDRs) which are typically responsible for an antibody’s binding activity.
Under recent guidelines for examination, the EPO often requires claimed antibodies to be defined by the CDRs necessary for the molecule’s activity. This is typically all three light chain and all three heavy chain CDR sequences. It is possible to define an antibody by less than six CDRs at the EPO, but usually only when it has been shown experimentally that the omitted CDRs are not required for the antibody’s proposed binding activity. Thus, for an antibody of known sequence, all essential CDRs should be disclosed in the patent application as a minimum and are likely to have to be recited in the main claim. A recent EPO Board of Appeal decision, T 1628/16 went even further and ruled that the framework regions of the variable domain are also known to be key to the affinity of an antibody and therefore required the claims to recite the framework regions in addition to the CDRs. Whether or not Examiners at the EPO and other Boards of Appeal follow this decision remains to be seen. Nevertheless, if the purported technical effect underlying the antibody relates primarily to its binding properties, the variable regions of the antibody, if not the whole antibody sequence, should also at least be disclosed within the application as filed, if available.
It is possible to claim an antibody by CDRs or variable domains with sequence identity below 100% relative to a defined sequence, but this typically must be done in combination with a clear functional feature. The permissible deviation from 100% identity will likely depend on the experimental evidence included in the application. Data or disclosures indicating acceptable substitutions, deletions or additions in the CDRs and/or variable regions may be supportive of a broader scope of protection for the antibody.
Defining Antibodies by Functional features
An antibody can also be defined in a claim solely by functional features. Such claim types are valuable because they do not limit the claimed antibody to a particular structure or sequence, and can, in theory, relate to any functional feature identified by the inventor. However, these type of claims are becoming increasingly difficult to prosecute to grant at the EPO.
Commonly, an antibody may be defined by the antigen it binds to. If an antibody is claimed in this way at the EPO, and the antigen in question is defined by a protein sequence, the antigen cannot be defined with open language, such as comprising or including, nor can it be defined by percentage identity or variability relative to a protein sequence disclosed in the application. In other words, the antigen must be defined by a limited sequence. If an antibody is defined by an antigen not specified by a limited sequence, it will be deemed to lack novelty in view of existing antibodies.
Other functional features that may be used to define an antibody product include, for example, binding affinity, neutralisation properties, internalisation of receptors, activation of receptors, or inhibition of receptors. However, it is important to note that if, when claiming antibodies exclusively via functional features, a prior art document discloses an antibody directed to the same antigen using an immunisation and screening protocol that produces antibodies having the claimed properties, the claimed antibody will be assumed to lack novelty.
Further, if an antibody is claimed exclusively via functional features, an EPO examiner will consider whether the patent application provides an enabling disclosure across the whole scope claimed i.e. whether the purported functional feature is achieved across substantially the whole of scope of the claim. If, in view of the patent application as filed, a skilled person is left tasked with conducting research to determine whether the antibody does indeed achieve the supposed effect across the breadth of the claim, the claims will likely be deemed insufficient. Thus, the data in the patent application will be key to obtaining broad protection for an antibody defined exclusively by functional features.
A recent EPO Board of Appeal decision, T 941/16 decided that not only must a patent application provide the information necessary to allow the skilled person to obtain substantially all claimed antibodies fulfilling the functional requirements of the claim, but also that recited functional features must clearly define the claimed subject matter to a person skilled in the art. In this instance, the claim in question related to an antibody defined by the feature “binds strongly to LNCAP cells but not or only minimally to cells which lack expression of PSMA”. Although this language had not been an issue before the Examining Division, the Board of Appeal determined that a skilled person’s assessment of an antibody's binding property is not an absolute criterion and would differ between individuals. As a result, the limits of a strong or minimal binding of the claimed antibody, and hence, the functional limitation of the subject matter, were deemed unclear. The decision highlights the need to define functional features in a way that it provides a clear instruction to a skilled person to reduce it to practice without undue burden. This requirement can often demand functional features be defined by precise parameters, which in turn demands the relevant experimental evidence to be included in the patent application to ensure the skilled person does not have to perform research to determine whether the antibody achieves the supposed effect. Thus, the data included in a patent application is not only of importance with regard to sufficiency of disclosure, but may also be important for ensuring an antibody can be clearly defined to ensure the patent application meets the requirements of Article 84 EPC.
It is important to note that an antibody defined by functional features within the context of a second medical use claim may be patentable with a broader scope than an antibody per se. For instance, a claim having the formula ‘An antibody targeting antigen X for use in the treatment of Y” may not require the structures of the claimed antibody to be disclosed in order to satisfy sufficiency of disclosure requirements, particularly if such antibodies are well known in the art. Nevertheless, an Applicant will be required to show experimental evidence within the application as filed that shows that it is at least plausible the antibody is able to therapeutically treat the claimed condition.
Defining antibodies by Epitope
Antibodies defined by their epitope (the particular region of an antigen that an antibody recognises and binds to) may also be patentable at the EPO.
If the epitope is a linear epitope (i.e., a continuous portion of amino acids of an antigen) and defined by a protein sequence, it cannot be defined with open language and must be defined by a clearly limited fragment with closed wording. If the epitope is a non-linear epitope (i.e., a discontinuous portion of amino acids with multiple distinct segments of an antigen) and defined by a protein sequence, the specific amino acid residues making up the epitope must be clearly identified and the method by which the epitope is determined must also be indicated in the claim. Furthermore, the patent application must provide an enabling disclosure allowing the skilled person to determine whether antibodies not disclosed in the patent application bind this epitope and also enable the production of additional antibodies that bind to the same epitope without undue burden.
A consideration when defining antibodies by their epitope is that it is often difficult to ascertain how an antibody defined in this way is distinguished from other antibodies that bind to the same antigen. Accordingly, comparative data illustrating that known antibodies bind to the same antigen via a differing epitope would advantageous if an antibody is to be defined in this way.
Defining Antibodies by Production Process
It is also possible to define an antibody by the process of their production. At the EPO, claims defining products in terms of a process of production are construed as a claim to the product as such. Thus, the technical substance of an antibody claimed by its production process lies in the properties imparted upon the antibody by its production process and not the process per se. Accordingly, it is the antibody itself that must be new and inventive. Evidence that the process imparts distinguishing properties upon the antibody must therefore be provided in the application as filed.
It is important to note that where an antibody is defined by an immunisation process with a particular antigen, the antigen must be defined by a sequence that is 100% identical to a specified sequenced with no potential variation. To do otherwise would likely incur clarity objections at the EPO because the use of variants renders the scope of the antibodies obtained by the immunisation process unclear.
Of course, a novel and inventive method for producing a particular antibody, for example a novel and inventive production method that improves yield or purity, can be claimed without a need to prove that the antibody derived from the process is distinguished from the prior art. In addition, the provision of Article 64(2) EPC establishes that any patent directed to such a process also confers protection to any antibodies directly obtained by the claimed process. Whether or not antibodies that have been subsequently modified or formulated are also protected under Article 64(2) EPC may be dependent on the national law of the individual EPC contracting states.
Defining antibodies by Hybridoma
Hybridomas are cells resulting from the fusion of B cells and immortal myeloma cells, producing long-lived cells able to constitutively express a single monoclonal antibody. By virtue of Rule 31 EPC, it is possible to define an antibody by reference to a deposited hybridoma cell line. However, it should be noted that a sample of the cell line must be provided to a recognised depositary institution not later than the date of filing of the application and that the deposit information is provided in the patent application as filed in order to fulfil the EPO’s sufficiency of disclosure requirements.
The Board of Appeal in T0032/17 concluded that a product by process claim referring to an antibody produced by a deposited hybridoma did not provide sufficient information for the skilled person to ascertain the chemical structure or amino acid sequence of the antibody. Accordingly, the product by process claim did not limit the claim to the antibodies structure or sequence. It therefore appears applicants should be cautious when relying solely on reference to a hybridoma deposit and that structural information regarding the antibody should also be provided in the application as filed. Otherwise, applicants may ultimately have to settle for protection of the hybridoma cell line as such.
Defining Antibodies with reference to a competitive antibody
A further way of defining an antibody in a claim is by reference to its ability to bind an epitope which is bound by a known antibody e.g. An antibody which binds to Epitope X and competes with Known Antibody Y. Such a claim is potentially valuable in that it potentially covers different antibodies. However, such claims may face clarity and patentability objections.
In summary, there are a number of ways to claim antibodies at the EPO, with each approach requiring the fulfilment of specific criteria to ensure that the claimed antibody is both sufficiently disclosed and clearly defined. In this respect, the scope of protection attainable is often closely tied to the quantity and nature of the data provided in the application as filed. Of course, the inclusion of data is also especially relevant to proving the inventive step of a claimed antibody. In our next article regarding the patentability of antibodies, we will look in greater detail at the requirements for establishing the inventive step of antibodies at the EPO.