First published in EPLAW, July 2019
Takeda UK Limited v F. Hoffman-La Roche AG, High Court of Justice, Patents Court, London, UK, 17 July 2019,  EWHC 1911 (Pat)
This judgment of Birss J concerns a patent relating to “Glycosylated Antibodies” (the “Patent”). The Claimant (“Takeda”) had brought a claim for revocation of the Patent. The Defendant patentee (“Roche”) had counterclaimed that Takeda’s monoclonal antibody vedolizumab (Entyvio) infringed the Patent. Vedolizumab is approved for the treatment of ulcerative colitis and Crohn’s disease.
The claims of the Patent are fairly complex. However, the critical feature of claim 1, which much of the judgment concerned, is a requirement for an antibody to exhibit at least “99% fucosylation”. Fucose is one of the monosaccharides found in glycosylation patterns on proteins. Antibodies with a high degree of fucosylation exhibit reduced “antibody dependent cellular cytotoxicity” (“ADCC”). Whilst high ADCC is generally a useful characteristic for therapeutic antibodies used to treat cancer, low ADCC is in fact advantageous when treating other conditions, for example autoimmune diseases.
First Birss J construed two important aspects of the claims. The first related to the definition of “antibody”. The Patent defined this term in such a way that it excluded certain things that might otherwise have been called an antibody. Importantly, Birss J noted, the definition required the antibody to have a “functionally active Fc region” (also termed “FcR binding”). The second aspect concerned the calculation required to establish >99% fucosylation. Both parties had suggested methods that could be used to calculate this. However, Birss J concluded that Roche’s fall back method (known as the “99%-TRM” method) was the most suitable.
Birss J made a further important preliminary finding which was that the skilled team had within its common general knowledge at the priority date in 2006 the know-how to make an antibody to a given target antigen. While this required a lot of work it was not without undue burden.
There were two key issues to be determined on infringement:
(1) whether vedolizumab had >99% fucosylation as defined in the claim;
and (2) whether vedolizumab was an “antibody” within the meaning of the claim.
Using the 99%-TRM calculation method, vedolizumab exhibited over 99% fucosylation and thus satisfied issue (1).
With respect to issue (2), Takeda argued that since vedolizumab had a mutation in its Fc region and exhibited no ADCC, the antibody did not have a functionally active Fc region and thus fell outside the patent’s definition of antibody and therefore also the claim. Roche submitted experimental evidence that supported its argument that vedolizumab exhibited some (albeit little) FcR binding. After considering the arguments, Birss J stated that the point was “a close call” but ultimately he agreed with Roche – the mutation in vedolizumab strongly reduced, but would not necessarily eliminate, FcR binding. Birss J therefore decided that vedolizumab fell within the claims of the Patent.
Takeda cited a number of prior art documents that disclosed antibodies with >99% fucosylation, and used these as the basis for a lack of novelty attack on the Patent. Roche argued that, whist these documents disclosed antibodies that fell within the claims of the Patent, the disclosures were not enabling. This was on the basis that the skilled person could not have made exactly what was taught in the prior art documents.
In making such arguments, Roche relied on a line of EPO prior use case law, starting with G1/92 (“Availability to the Public”). This case law makes clear that a product freely on sale pre-priority is not novelty destroying unless the skilled person can analyse it sufficiently (without undue burden) in order to be able to reproduce it for themselves. Roche argued that this principle also applies to documentary disclosures.
Birss J did not accept this submission on the basis that even though the skilled team could not make exactly what was taught by the prior art, they could make a variation of the disclosure which itself would fall within the scope of the claim. Notably, Birss J held that if the prior art discloses an amino acid sequence of an antibody, and if the skilled team can make an antibody with that amino acid sequence, then a claim to that amino acid sequence lacks novelty.
The fact that the document states that the amino acid sequence is from an antibody called “Antibody A” (which may contain additional features not relevant to the claims of the patent) and the fact the skilled team would never know how to make Antibody A specifically, is irrelevant to whether there is an enabling disclosure of the amino acid sequence. He held that this was still a lack of novelty attack and did not stray into obviousness.
Takeda’s main obviousness attack was that the patents lacked a technical contribution over the prior art. Roche, in reply, identified three contributions to the art which, it argued, the Patent made. Birss J rejected each alleged contribution in turn. In particular, Birss J held that the idea of increasing fucose for a therapeutically useful purpose formed part of the common general knowledge and therefore did not make the Patent inventive. In addition, whilst the idea of increasing fucose to >99% to reduce ADCC to background would be a technical advance, it was neither plausible from the disclosure in the Patent nor necessarily true.
On insufficiency, Takeda’s primary argument was that the Patent was ambiguous as it was unclear what method to use to measure the amounts of the various sugars in the antibody. Birss J concluded that the skilled team, given the Patent, would consider using one of two measurement systems. Depending on which system they used, the result could either fall inside the claim or outside it. He held that “the ambiguity is not of a kind which reveals a fuzzy boundary at the edge of the claim. The claim is truly ambiguous and therefore invalid.”
Birss J concluded therefore that while vedolizumab was an antibody which fell within the claims of the Patent, all the relevant claims were invalid for lack of novelty, obviousness/lack of technical contribution and insufficiency.
A copy of the decision can be found here.