Merck Sharp & Dohme Limited v Ono Pharmaceutical Co. Limited & another, and Bristol Myers Squibb Company, Ono Pharmaceutical Co. Limited & another v Merck & Co. Inc. & another, High Court (Patents Court), Mr Justice Birss, 22 October 2015
The UK Patents Court has upheld the validity of Ono’s patent EP (UK) 1 537 878 (under which BMS is an exclusive licensee), which concerns antibodies to the PD-1 receptor for the treatment of all cancers.
Ono and BMS alleged that Merck’s anti-PD-1 antibody pembrolizumab (Keytruda), which it proposes to market as a treatment for unresectable/metastatic melanoma, infringed the patent. Merck accepted that if the patent was valid, its product fell within the scope of the claims. Merck raised a wide variety of attacks on the patent, including added matter, insufficiency, loss of priority, anticipation and obviousness. Ono admitted that if priority was lost, the claims were invalid. Ono further stated that it would not seek an injunction in the UK if its claim to infringement was successful.
Interestingly, the judge held that plausibility forms part of the law of novelty as part of the enablement assessment. In this case, the question was whether the prior art plausibly enabled the claimed therapeutic effect. This issue is frequently raised in cases concerning second medical use claims where there is an issue of whether the disclosure of the therapeutic effect in the prior art is sufficient to be novelty destroying. Birss J’s approach can be contrasted with the approach of Arnold J in Hospira v Genentech  EWHC 1796 (Pat) in which this issue was instead dealt with by asking whether the claimed therapeutic effect was properly disclosed in the prior art (i.e. you must know from the prior art that the claimed therapeutic effect will be obtained).
Other interesting aspects of the judgment include how the judge dealt with Merck’s plausibility attack (which was raised under the heads of priority, sufficiency and AgrEvo obviousness) and his analysis of whether the invention was ‘obvious to try’.
Plausibility – sufficiency/priority/AgrEvo obviousness
The claims in issue related to use of an anti-PD-1 antibody to inhibit the immunosuppressive signal of PD-1 for the treatment of cancer.
Ono relied on two in vivo tumour studies in mice contained in both the priority document and the patent (which demonstrated that inhibiting the PD-1 pathway reduces tumour growth in two specific cancer cell lines) in support of its claim to priority and its argument that the patent was sufficient.
Merck argued that it was not plausible that anti-PD-1 antibodies could be used to treat all cancers. In particular, they said that if a particular tumour did not express a PD-1 ligand (PD-L1 or PD-L2), inhibiting the PD-1 receptor would not be effective to reduce growth of the tumour cells. On the evidence, Birss J said this was not shown to be the case and, as part of their CGK, the skilled team would not expect effective treatment to be limited to PD-1 ligand expressing cells as PD-L1 was also known to be expressed in the lymphatic system.
To address this issue, Merck also argued that the claims were only plausible to the extent that cancer cells either expressed a PD-1 ligand or were otherwise immunogenic (capable of being recognised by the immune system). Merck argued that as not all cancers were known to be immunogenic (indeed the judge held earlier in his judgment that immunotherapy, i.e. treating cancers with the immune system, was viewed pessimistically with a history of failure at the priority date), the claims were unduly broad and therefore insufficient. The judge rejected this argument. In his view, immunogenicity of cancers was not a binary question and could change over time for a variety of reasons. He therefore held that the claims were plausible at the priority date.
Merck’s final plausibility attack was that, as a matter of fact, post-priority evidence showed that anti-PD-1 antibodies are not promising treatments for colorectal and prostate cancers. Merck argued that the claims do not therefore work with substantially everything falling within their scope (see Kitchin LJ at  –  in Regeneron v Genentech  EWCA Civ 93). The judge also rejected this argument, as he thought the claims expressed a principle of general application that was described at a fair level of generality and therefore, on his view, the claims worked for substantially everything within the scope of the claims.
Interestingly, the judge held that plausibility forms part of the law of novelty under the enablement assessment. So with medical use claims, the prior art must enable the therapeutic effect by making it plausible. In this case, the prior art referred to treating cancer but without any real experimental data in support. The mere assertion was not enough to make the disclosed therapeutic effect plausible and the prior art was therefore not novelty destroying.
The judge considered that the strongest piece of prior art was a publication called ‘Latchman’, which suggested that “the PD-L-PD-1 pathway may be an attractive therapeutic target” and “blocking the PD-1 pathway may enhance anti-tumour immunity.”
A critical issue in the case was whether it would have been part of the skilled team’s CGK that the PD-1 pathway was generally regarded as inhibitory (i.e. the binding of PD-1 and its ligands inhibits T-cell activation and downregulates immune response). Birss J held it was. However, he also held that there was a known discrepancy that there were co-stimulatory effects associated with the ligands PD-L1 and PD-L2.
Birss J also held that it was known that PD-1 knockout mice developed autoimmune diseases (i.e. immune response is upregulated in vivo by removing the PD-1 gene), and earlier research showed that inhibiting a homologous pathway (the CTLA-4 receptor), which was also known to be inhibitory, prevented growth of cancer in mice and this was CGK. He held that it was also CGK that research on this pathway would have informed the direction of research on the PD-1 pathway.
The key question the judge had to address for obviousness was whether it would have been obvious to carry out an in vivo mouse tumour model in which PD-1 was blocked using an anti-PD-1 antibody. The judge’s analysis of this question focussed on the degree of expectation of success that the skilled team would have had and focussed on his findings of what was CGK. Birss J held that, on balance, he considered that the skilled addressee would only have a hope that an in vivo mouse tumour model might work not a reasonable expectation that it would work. Key factors in weighing in Ono/BMS’ favour were the uncertainty surrounding the co-stimulatory effects of the PD-1 ligands and Merck’s expert expressed the view in cross-examination that the results of the in vivo mouse tumour experiments in the priority document would have been seen as very exciting.
Interestingly, the judge’s analysis centred on the expectation of success with little consideration of other factors relevant to obviousness, e.g. motive to find a solution, number and extent of possible avenues of research and the effort involved in pursuing them (see Lord Hoffman at  in Conor v Angiotech  UKHL 49).
Read the full decision here.
Head note: Chris Pratt, Marks & Clerk Solicitors