Is the European Patent Office’s approach to assessing obviousness of antibody inventions consistent?

Abstract

Antibodies can be highly effective in the diagnosis and/or treatment of a wide range of diseases, including certain cancers, viral infections and inflammatory diseases. The development of any new therapeutic antibody is very time consuming, requires significant resources and only a small percentage of initial candidates ever make it onto the market. Patent protection therefore plays a key role in allowing companies to obtain a good return on this investment. However, in Europe, the bar for assessing the inventiveness (obviousness) of antibodies is currently set particularly high, so there is a significant risk that new antibodies may be denied patent protection despite having beneficial properties. This article looks at the rationale behind the European Patent Office’s (EPO) current approach to assessing the patentability of antibodies, contrasting it with US practice as well as the EPO’s assessment of obviousness in the chemical field. It also explores whether a recent decision by an EPO Appeal Board may help to bring about a change in this practice.

Keywords::

• antibodies
• European Patent Office
• innovative
• inventive step
• obviousness

1. Introduction to antibodies

Antibodies are one of the tools used by the immune system to fight infection. Generally speaking, each antibody can specifically recognise and bind to a specific part (epitope) of a foreign target (antigen) in a manner analogus to a lock and key.

Antibodies are large, Y-shaped proteins that share the same general structure including a constant region, but possess so-called variable regions at the tip of each arm of the ‘Y’ that distinguish one antibody from another. Each variable region is composed of three ‘hyper-variable’ complementarity determining regions (CDRs) that are surrounded by framework regions. Broadly speaking, the amino acid sequence of the CDRs determines inter alia the binding specificity and affinity of an antibody.

It is widely accepted in the scientific community and in patent circles that significantly changing the amino acid sequence of the variable region, particularly the CDRs, of an antibody is likely to affect its properties. However, the way in which the variable regions determine the characteristics of antibodies is still largely unknown, so it is impossible to predict or model with any confidence what effects a radical change to the sequence of the variable region would have, or what CDR sequences should be combined to arrive at an antibody having a desired set of properties.

Antibodies have significant potential in a variety of applications, particularly therapeutic and diagnostic ones, so there is a drive to generate new antibodies. The present article is concerned with the patentability of antibodies that differ from existing antibodies in the sequence of the variable region(s). Such antibodies are typically referred to as ‘innovative’ or, in some instances, ‘me-better’ antibodies. Biosimilars are antibodies having the same variable and constant sequences as a known reference antibody and are not considered herein [1].

2. European Patent Office’s current approach to assessing inventiveness

A key requirement for obtaining a patent is that the claimed invention must be non-obvious. This ensures inter alia that patents may not be obtained for arbitrary modifications of known products. The European Patent Office’s (EPO)’s guidelines for examination define the term ‘obvious’ as that which does not go beyond the normal progress of technology but merely follows plainly or logically from the prior art, that is, something that does not involve the exercise of any skill or ability beyond that to be expected of the person skilled in the art [2].

It is also established EPO practice that non-obviousness may be achieved by an invention having an unexpected property or advantage, or by the provision of a non-obvious alternative. However, in the field of antibodies, EPO’s examination practice has developed to require applicants to show an advantage of a new antibody over any known antibodies to the same antigen. In other words, the provision of a new antibody that ‘merely’ represents an alternative to a known antibody to the same antigen is considered to be obvious.

The leading decision commonly quoted by EPO Examiners in this regard is Appeal decision T735/00, which asserts that as early as 1989 the preparation of monoclonal antibodies was ‘a matter of routine experiment’ and alleges that once an antibody to a specific target is known, there is an incentive to look for further antibodies [3]. In T735/00 the Appeal Board indicated that the provision of a further antibody specific for the target should only be considered to be inventive if the antibody has unexpected properties, was particularly difficult to make or if it was unexpected that such an antibody could be produced at all.

EPO examiners therefore place a burden on the applicant to show that his antibody is superior to any known antibodies to the same target; for example, by showing a higher affinity or by having a property that the prior art antibody does not possess, such as the ability to induce target cell death.

The EPO’s position is in stark contrast to the ‘structural non-obviousness’ approach taken, for example, by the US Patent and Trademark Office. The structural non-obviousness approach acknowledges that it could not have been foreseen that an antibody having a different structure (particularly different CDR sequences) from any known antibodies to the same target may nevertheless be specific for that target. In the US, non-obviousness can therefore be acknowledged for a new antibody having a significantly different amino acid sequence from prior art antibodies, regardless of whether or not the new antibody has superior properties compared to the prior art antibodies. Interestingly, the EPO typically does apply the structural non-obviousness approach to other types of inventions; for example, small molecules in the chemical field.

3. Decision T0067/11 of the EPO’s Board of Appeal

In May 2014, one of the EPO’s Appeal Boards issued decision T0067/11 in an appeal against the decision of the Examining Division to refuse patent application No. 04728755.2 [4]. This case concerned the provision of humanised versions of a known murine antibody to a ganglioside (a sugar-containing cell surface molecule).

The applicant initially tried to claim almost any humanised versions of said antibody: This claim required the constant regions, which were of murine origin, to be replaced with constant regions of human origin. This is standard practice when humanising murine antibodies. The only further requirement recited in the claims was the presence of point mutations in the framework regions to reduce the immunogenicity of the antibody. Neither the number, nor the positions or identities of the point mutations were recited in claim 1.

The prior art showed that it was well known that framework regions of murine antibodies can be immunogenic in humans, so the Examining Division considered that it would have been obvious to introduce point mutations to make the framework regions more similar to human framework regions and hence be less immunogenic.

During the appeal, the applicant limited claim 1 to recite the specific sequences of the mutated framework regions (in addition to the CDR sequences) and argued that it could not have been foreseen that an antibody having these specific sequences would have reduced immunogenicity while retaining the binding affinity for its target.

The Appeal Board accepted this argument and acknowledged an inventive step on the basis that the specific claimed antibody has an advantageous property because it is less immunogenic than the murine prior art antibody.

4. Discussion and expert opinion

The EPO has been widely criticised by industry for setting a higher bar when assessing the obviousness of antibody-related inventions compared to small chemical molecule inventions, where structural non-obviousness is typically acknowledged. It is possible, but by no means certain, that T0067/11 may signal a slight lowering of this bar.

In some respects, decision T0067/11 may be argued to be in line with current EPO Examination practice, because non-obviousness was acknowledged only for an antibody having an advantageous property.

However, it can also be argued that it is not surprising that the replacement of murine residues with those commonly found in human antibodies results in a reduction in immunogenicity. The true unexpected element in this case was that the specific antibody sequence would result in reduced immunogenicity. Indeed, the board indicated that it could not be predicted that the claimed specific mutations would result in reduced immunogenicity. Thus, the board seems to consider that the advantageous property of the antibody is intrinsically linked to its sequence.

Once it is accepted that the characteristics of an antibody are determined by its sequence, and that it cannot be predicted how a change in sequence may affect an antibody’s properties, the assessment of obviousness does not seem to be too distant from the structural non-obviousness approach. It is therefore possible that T0067/11 represents a slight shift in the EPO’s treatment of antibody inventions.

Interestingly, the Appeal Board deciding T0067/11 made no reference to T735/00, or indeed any previous Appeal Board decisions; thus, no guidance is provided as to whether this board felt that it was following, or distancing itself from, any of the leading decisions. It is worth noting that in T735/00, the applicant did not try to argue that the provision of an antibody having a specific new sequence was non-obvious. Thus, while apparently more lenient in its assessment of obviousness, T0067/11 is arguably not inconsistent with T735/00.

Further clarification may well be provided when a decision issues in pending case T617/13 [5]. In this case, a new antibody specific for receptor CXCR4 was provided and the Examining Division refused the application, alleging obviousness in light of a prior art antibody having a different sequence but the same functional properties (inhibition of ligand binding and causing apoptosis of cancer cells) as the claimed antibody.

The Examining Division’s rationale was that it is not surprising that further antibodies having this combination of properties can be generated. However, it could be argued, applying the reasoning of T0067/11, that, although it may have been obvious to try to provide a further antibody having this combination of properties, it could not have been foreseen that an antibody having the specific sequence of the new antibody would have this combination of properties. It remains to be seen how the EPO will decide this case, and whether T0067/11 will indeed be considered by the board.

Meanwhile, applicants should continue to try to identify advantageous features of their antibodies, but during the prosecution of applications relating to antibodies for which no advantage over prior art antibodies can be identified, it may well be worth trying to argue for a structural non-obviousness assessment based on T0067/11.

Declaration of interest

The author is an employee at Dehns and is a Patent and Trade Mark Attorney. The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.