In relation to antibodies, the EPO does not generally apply the principle of “structural non-obviousness.” The approach of the EPO is that once an antibody having a particular function is known, then other antibodies having the same function can generally be obtained without difficulty. Prima facie, the specific sequence of a new antibody embodying a known function therefore cannot confer inventive step: it is simply an “arbitrary selection” from among multiple, equally obvious, alternatives to the prior art. (For instance as established in the Board of Appeal decision T0735/00). Historically, this approach derived from a technical background in which the antibodies in question were generally obtained as a direct result of a screening. The rationale was that once you knew that an antibody having certain properties could be obtained from a screen, it would be generally possible to apply the same screen to obtain a second antibody also having these properties, without undue difficulty. As engineering approaches have become more common, the EPO has tended to maintain this rationale as the starting point, and to “bolt on” objections that certain engineering methods represent routine steps, not capable of conferring inventive step. Thus, an EPO Examiner may assert that a humanized, affinity matured antibody to a particular target lacks inventive step over the parent mouse antibody, or even over a different mouse antibody with similar properties, alleging it could be obtained by routine methods. However, particularly as such objections extend to cover more of the tools of antibody engineering, this approach risks losing touch with the technical reality of the engineering process. It also risks shortchanging innovator pharma applicants on the level of the reward provided in return for disclosing complex inventions.

Underlying each of the techniques of antibody engineering are a multitude of interrelated considerations, with a sometimes unpredictable relationship with drug efficacy. For instance, affinity may affect tumor penetration and diffusion rates or may modify internalization, with a critical effect on therapeutic efficacy; changes made to improve manufacturing stability may adversely affect affinity or specificity; and changes to the overall charge of the antibody can affect pharmacokinetics. In some cases, kd (off-rate) may affect cellular response to antibodies, independently of overall KD (affinity). Similarly, optimization of T-cell engaging bispecific antibodies may require balancing the affinities of the two specificities for maximizing therapeutic efficacy while minimizing off-target effects. We are seeing many of our clients paying close attention to the specific biology of a pathway, with a great deal of care being taken at the design and engineering stage. The good news is that a detailed understanding of the antibody’s properties will often support a strong argument for inventive step during the patenting process.

The sophistication of these engineering techniques may make it easier to rely on specific sequence features to support inventive step, even in the presence of close prior art. For example, it will often be possible to point to unexpected results or effectiveness; to the antibody being subject to multiple constraints; and/or to simultaneous improvement of multiple properties. Some exemplary scenarios are below. More complex ones could certainly be envisaged.

• The prior art antibodies bind human target strongly and cynomolgus target weakly; improving affinity for cynomolgus without reducing affinity for human might confer inventive step.
• The antibody has a CDR residue that causes production problems. Several mutants are tried. As expected, most result in reduced affinity but one surprisingly also increases affinity as well as avoiding the production problem.
• The antibody has multiple residues that cause production problems. Several mutants are tried for each residue. Certain combinations have much less deleterious effect on binding than the majority.

T 0067/11 gives a hint of such an approach being endorsed by the Boards, in a finding that the precise modifications giving a combination of low immunogenicity and high affinity to a humanized antibody could not have been predicted. We should caution that due to the raising of the bar as technology develops, a different decision may be reached on the basis of these same facts today. However, the general principle should still apply, namely that inventive step can arise from a need to balance different constraints and from a lack predictability as to how this can be achieved. From the patent attorney’s perspective, the more understanding there is of these kinds of issues when the patent is filed, the better the patent application can be. It can be extremely useful to tell a good technical story about the multiple problems solved in developing the antibody, along with data to support that those problems have been solved, and that other approaches did not solve the problem. Comparative data can be invaluable when there is close prior art, especially the applicant’s own earlier antibodies. Where highly engineered antibodies are to be defined by structure, then engineered residues will probably need to be precisely defined, even if some variation is permitted elsewhere. And if the inventive step story depends on interactions between different parts of the antibody, then it can be expected that these different parts will need to be defined in the claims. By having a good understanding of the likely arguments at the drafting stage, appropriate basis for claims and amendments can be built into the application.

Another opportunity lies in a renewed role for functional definitions. Over recent years, we have seen increased resistance at the EPO to claims which define antibodies in purely functional terms. The EPO often objects that the claim merely defines the underlying problem to be solved, especially where the feature appears to be something that would be desirable, e.g., high affinity. There is now a growing realization in the field that subtle differences in properties can give large advantages, e.g., in terms of reduction of side effects. For instance, in many cases there is a move from a general view that “high affinity is good” towards tuning the precise affinity that is optimal, which may be lower or intermediate, depending on the situation. This may open the door to new possibilities for claiming antibodies by function, in which one or more parameters such as target affinity are defined by a (relatively) narrow range. Claims which relate to the use of the antibody (e.g., medical use claims) may also be important, by providing the context in which the parameter range is beneficial. The EPO’s purpose-limited-product approach to medical use claims offers considerable flexibility for claim drafting. Medical use claims may also reduce the burden on the applicant in regard to the demonstration of novelty. Generally, the burden of proof for showing novelty for functionally-defined antibodies falls on the applicant (e.g., T753/00 and the Guidelines for Examination G-VI-6). For a product claim, and where there are many existing antibodies against a target, this burden can be onerous: by using a medical use claim (where appropriate), the extent of the art over which it is necessary to show novelty can be greatly reduced. As always, it will be important to carefully define the functional parameters and to ensure that the description provides a method by which they can be reliably and reproducibly measured.

The increasing prevalence of highly engineered and fine-tuned antibodies being developed by applicants appears to present the opportunity for some shifts in patenting strategy at the EPO. In the right circumstances, the need for complex engineering of antibodies appears to open the door for antibody technology to be treated more like other technical areas, and for the non-obviousness of the specific structures which achieve the desired effect to be given more weight. This perhaps also brings some elements of US-style structural non-obviousness into EPO practice. The downside of this approach is the likelihood of relatively narrow claims. Of course, even narrow claims may be highly valuable for protecting biopharmaceuticals against generic competition. Alternatively, with the right scenario and data, functional definitions may provide an alternative route offering broader protection. This is supported at a technical level by a move away from properties which may be considered generally or obviously desirable by an Examiner, and towards properties which are tuned to specific contexts. The EPO applies a high bar to the protection of antibody inventions. As patent practitioners, ensuring that we fully understand the complexity of the engineering work at the drafting stage offers the best possible chance of clearing the bar easily and with a useful scope of claim. It is also encouraging to remember that positioning the application with the EPO’s requirements in mind should give the best chance of the application succeeding in all major patent jurisdictions.