A stricter interpretation
Is the grant of a reasonable scope for antibody-related inventions at the European Patent Office a phenomenon of the past? Joachim Wachenfeld and Florian Grasser of Vossius & Partner report.
The invention of the antibody hybridoma technology by Nobel prize winners Georges Köhler and César Milstein in 1975 paved the way to the development of therapeutic antibodies.
The first therapeutic monoclonal antibody obtained market approval in the US in 1986. Since then the market value of therapeutic antibodies has grown to approximately $115.2 billion in 2018 and is expected to reach $300 billion in 2025.
Along with this development, the number of patent applications seeking protection for therapeutic antibodies at the European Patent Office (EPO) has constantly grown and is growing further.
Traditionally, an antibody was held patentable in generic form at the EPO if it was directed to a novel and inventive protein which was sufficiently distinct in its amino acid sequence from prior art proteins.
These cases are, however, rare exceptions nowadays in light of the published sequences of nearly all human proteins. Typically, in order to overcome the inventive step threshold, it is now necessary to show that the antibody was difficult to obtain and/or displays unexpected or superior properties (T 735/00 and T 187/04).
It has been common practice at the EPO for some time to accept that such properties inherent to an exemplified antibody can give rise to a more generic claim if the claim indicates the complete sequences of all six complementarity determining regions (CDRs) of the exemplified antibody. This is because the CDRs predominantly determine the antibody’s epitope specificity.
The EPO typically applies the “plausibility” requirement when assessing the actual achievement of a claimed invention across the claimed breadth (eg, T 1329/04). Currently, we observe a growing number of cases in prosecution, wherein it is deemed necessary to incorporate the complete sequences of the heavy and light chain variable regions, ie, including those of the framework regions (FRs) of the exemplified antibody into the main claim.
According to the Examining Divisions in charge of these cases, this measure is required in order to render it plausible that the property relied upon is achieved across the claimed breadth.
Often, the solution to the technical problem is considered credibly solved only by the specific antibody described in the examples, and challenging this position is particularly difficult if there is only a single antibody described. Alternatively, an argument is presented that the FRs are known to contribute to the binding specificity and thus their sequences need to be incorporated in the main claim as well.
In a further alternative, we have seen the argument that the claims cover future inventions. This principle, however, has been sanctioned by a number of decisions of the Boards of Appeal in the past.
If an antibody has been deposited according to the Budapest Treaty, a requirement to restrict the claims to that deposited antibody may well be expected. This approach taken by the Examining Divisions of the EPO increasingly resembles the manner in which the US Patent and Trademark Office assesses the written description requirement under US law.
”The board held, by relying on published art, that in some cases the FRs were relevant for retaining binding specificity whereas in other cases they were not.”
Since generally a range of mutations can be introduced into the FRs without losing the desired property of the exemplified antibody, the incorporation of the complete sequences of the variable regions of the heavy and light chains into the main claim results in a rather narrow scope of protection. Third parties might be tempted to design around such a narrow claim by generating highly similar antibodies that fall outside the literal scope of such a patent.
In case no agreement on a reasonable scope of protection can be reached with the Examining Division, the question arises as to whether resorting to the Board of Appeal will result in a more favourable outcome. We have therefore reviewed some cases that will hopefully assist applicants in securing a meaningful scope of protection for their antibody inventions.
In T 386/08, the board allowed claims relating to a humanised antibody specifically binding to an epitope of the amyloid beta peptide, characterised by the six CDRs and not further specified by framework region sequences. It was sufficient that the framework regions were characterised as being “from a human immunoglobulin light or heavy chain”.
It certainly helped the patentee’s case that the patent disclosed potential positions of modification in the variable regions. The board held, by relying on published art, that in some cases the FRs were relevant for retaining binding specificity whereas in other cases they were not.
The board emphasised the established concept that, for a disclosure to be considered as sufficient, it is not required to demonstrate that each and every conceivable embodiment falling under the scope of a claim can indeed be obtained. This decision supports the view that an antibody can be sufficiently characterised in a claim by reference to the six CDRs.
In T 617/07, the board dealt with claims covering a monoclonal antibody as well as “synthetic and biotechnological” derivatives thereof, which were defined structurally by at least one CDR sequence of the light chain or heavy chain and functionally by several parameters, inter alia the binding to a nerve growth factor receptor.
The board accepted that the claim covered antibodies that do not have the desired functional properties. However, it also held that the skilled person, having regard to the advanced knowledge of antibody production from 1999, was able to avoid non-functional variants, even if this might require some time-consuming efforts.
In this case, it helped the patentee that it could present an antibody that had five CDRs in common with the antibody exemplified in the patent, but differed in several amino acids in one CDR while retaining the required function. This decision thus allowed an even broader extent of generalisation as regards sequence requirements needed to characterise an antibody.
”The skilled person therefore does not have to perform extensive trial and error experiments in order to find further embodiments falling under the scope of the claims.”
T 418/07 dealt with an antibody defined structurally by the heavy and light chain CDR3 regions, each of which could carry up to five conservative amino acid substitutions and defined functionally by the superior binding characteristics to tumour necrosis factor as determined by Koff rate constant values.
Several antibodies having the required CDRs and also carrying certain conservative substitutions were shown in the specification. The board held that a proper construction of claim 1 rules out any antibody not complying with the structural and functional requirements and also that the skilled person routinely prepares mutations of amino acid sequences and tests for the required function.
The skilled person therefore does not have to perform extensive trial and error experiments in order to find further embodiments falling under the scope of the claims. The requirements for the inclusion of specific sequences contained in the variable portion of an antibody thus come close to those established in T 617/07.
In T 1300/05 the claimed antibody was required to be capable of specifically inhibiting the fusion of a macrophage-tropic primary isolate of HIV-1 to a CD4+ cell susceptible to infection by a macrophage-tropic HIV-1, as determined by a particular method stated in the claims. The antibody was solely characterised by functional features.
The board decided that this antibody was sufficiently disclosed by the application, since it was a matter of common general knowledge to raise antibodies against a given antigen and because various methods referred to or disclosed in the application to quantify HIV-1 envelope protein-mediated membrane fusion between cell lines existed, so that the claimed antibody could have been reproduced by the skilled person without undue burden.
The board rejected the Examining Division’s insistence on having structural features incorporated into the claims and commented: “Yet, the sequence of an antibody is not likely to provide any useful information as to its characteristics, the determination of epitopes is a downstream development of the isolation of the antibody per se and an accession number is not suited to characterise a family of antibodies.”
In T 2101/09, the claims covered an antibody that “specifically” binds to a novel and inventive polypeptide, including polypeptides having a sequence identity of at least 95% to the specifically disclosed polypeptide. No structural information was provided in the claim.
The board held that the claimed antibodies are limited to those that bind only to an epitope of the novel and inventive polypeptide. It was apparently the board‘s understanding that such an epitope does not occur on prior art polypeptides. Methods for identifying appropriate antibodies that do not cross-react with other polypeptides of the same class were available to the skilled person.
In T 2332/10, the main claim was directed to an antibody or fragment thereof that binds to the complement factors C5 and C5a, does not prevent the activation of C5 and does not prevent formation of or inhibit the activity of C5b.
As in T2101/09, the antibody was solely characterised by functional features. The board was satisfied that the claimed antibody solved the technical problem in view of the experimental data of the patent showing that the antibody allows C5 cleavage into C5a and C5b by the convertase of both the classical and the alternative complement activation pathways.
”Methods of how antibodies falling under the scope of the claims can be generated should be identified in the application as filed and by referring to prior art.”
A decision where solely functional features characterising a generic antibody (in this case an antibody to IL-6 or the IL-6 receptor) were not accepted is T 352/07. One of the arguments presented by the board was that the application did not direct the skilled person’s attention to suitable tests for identifying the required antibodies.
The prior art identified in the patent was also held unsuitable for providing the skilled person with an unambiguous choice of suitable tests.
In the more recent decision T 511/14, the board denied patentability of an antibody characterised by reference to the six CDRs and binding to the polypeptide Dkk-1.
Antibodies to Dkk-1 had been described in the prior art and the applicant did not provide any support for superior properties over a specific prior art antibody.
Even if such properties were available, patentability would have been acknowledged only for the corresponding specific antibody (meaning that the complete variable region sequences and potentially also the constant region sequences would have to be included into the claim).
The above summarised decisions of the EPO’s boards of appeal provide various options to argue against objections requiring the incorporation of complete variable region sequences into an antibody claim.
Two notions appear of particular importance. Methods of how antibodies falling under the scope of the claims can be generated should be identified in the application as filed and by referring to prior art. This is particularly advisable if the claims are solely characterised by functional features.
If claims refer to sequences, a broader scope is often easier obtained if sequence variations are indeed tested in the examples. For example, if a claim is characterised solely by a binding specificity and CDRs, it may help to confirm that antibodies having modifications in the FRs display the same specificity.
Alternatively, a promising approach seems to include, besides the sequences of the CDRs, sequence identity language with respect to the FR regions into the claims.
It remains to be seen, however, whether the jurisdiction of the Board of Appeal that so far has allowed broader claims will adapt to the practice of the Examining Divisions over time.
We are aware that the above is not a complete overview of more recent decisions of the boards in the antibody field and that there may be other options to obtain broader claims.
In this regard, we refer to the excellent monograph “Antibody Patenting, A Practitioner’s Guide to Drafting, Prosecution and Enforcement”, Wolters Kluwer 2019, edited by our partners Jürgen Meier and Oswin Ridderbusch.
This article represents the personal opinion of the authors and is not to be construed as an opinion of the law firm Vossius & Partner.
Joachim Wachenfeld is a partner at Vossius & Partner, specialising in biology and biotechnology. He can be contacted at: firstname.lastname@example.org
Florian Grasser is a partner at Vossius & Partner, specialising in biology and biotechnology. He can be contacted at: email@example.com
Image: Shutterstock.com / Christoph Burgstedt