UK – Samsung Bioepis v. Janssen Biotech / ustekinumab

This article was first published in EPLAW blog [August, 2024].

Patents Court finds Janssen Biotech’s patent, for the antibody ustekinumab for treating ulcerative colitis, invalid for obviousness over prior art

On 30 July 2024, Mr Justice Meade handed down a decision in a revocation action brought by Samsung Bioepis against Janssen Biotech’s patent, EP 3 883 606, which claims the antibody ustekinumab for treating ulcerative colitis (UC). The key issues at hand were (i) the common general knowledge of the skilled person at the priority date, (ii) the construction of the claims, and (iii) whether the patent was anticipated or made obvious by the prior art, which included an abstract and poster by Prof Ochsenkühn and a set of slides reporting a phase III study of ustekinumab induction therapy in UC (the Sands Slides).

Meade J conducted a detailed assessment of the common general knowledge. The dispute fundamentally concerned the knowledge and understanding that the skilled person would have in relation to the detailed immunology and mechanisms of inflammatory bowel disease (IBD) and the ways in which drugs affected them. Janssen submitted that the skilled person would have detailed knowledge in this respect in relation to both UC and Crohn’s disease. Applying the principles from Alcon and Illumina, Meade J sided with Janssen on this point, considering that the skilled person’s level of knowledge would be greater than that submitted by Samsung. This was supported by the detailed information in the patent, which the skilled person would be expected to understand, and the level of the literature in the field. However, in this case, the identity of the skilled person and their common general knowledge did not significantly impact the overall conclusion on obviousness.

The anticipation attack was based on the Ochsenkühn poster. The Ochsenkühn abstract had been published on 16 January 2018, however there was a dispute between the parties about the disclosure of the poster, which presented more data than the abstract. Samsung alleged that the poster was presented at a conference between 2 and 5 June 2018. Janssen submitted that Samsung had failed to prove that the poster presented at this time was the same version as the poster Samsung relied on in the proceedings. Prof. Ochsenkühn gave evidence on behalf of Samsung, proving his attendance at the conference with a selfie containing a partial view of the poster, and explaining that although there were some formatting differences, these were immaterial.

Meade J found for Samsung on this factual issue, concluding that the pleaded poster was part of the state of the art in June 2018, prior to the relevant priority date. However, Meade J held that the Ochsenkühn poster did not anticipate the patent; the study it disclosed was a retrospective analysis of a small number of patients with UC and was from an unblinded single centre study with no control group or placebo. On the proper construction of the claim, therapeutic efficacy must actually be achieved by ustekinumab and the data in the Ochsenkühn poster did not demonstrate to the necessary standard that ustekinumab was in fact effective for UC. The patent was therefore not anticipated.

On claim construction, Meade J also rejected an argument by Samsung that the claim wording “wherein the subject is in corticosteroid-free clinical remission at least 44 weeks after week 0” meant that a patient must be remission at some point prior to 44 weeks after “week 0” (which Samsung contended meant the start of the induction phase) but need not still be in remission at 44 weeks. Meade J noted that where a claim is poorly drafted (as here), the skilled person would refer to the specification to obtain important context to understand the meaning of the claim. Although the language of the claims is a powerful consideration and there is some limit to the degree which the specification can extend or cut it down, it is not appropriate to look at the claims alone and use the specification only if the claims appear unambiguous. Taking the information in the specification into account when reading the claim, it would be clear to the skilled person that “week 0” in this context meant the start of the maintenance phase rather than the induction phase. It was also clear that “at least 44 weeks” means remission must be achieved at 44 weeks and it would not be sufficient if remission was only achieved before 44 weeks and not after. This dealt a further blow to Samsung’s anticipation attack: even if the data in the Ochsenkühn poster had been sufficient to demonstrate efficacy, remission was only reported 39 weeks after initiation of the induction phase.

Samsung also alleged obviousness over the three pieces of prior art. Meade J rejected the obviousness case over the Ochsenkühn abstract and poster. Although the work reported in the Ochsenkühn abstract and poster looked superficially impressive, the problems in the methodology and results meant it would not provide the skilled person with confidence about the treatment of UC so as to render the patent obvious: the documents would provide the skilled person with only a hope of success, not a reasonable expectation. Moreover, the attack required illegitimate mosaicking with the Sands Slides. However, the Sands Slides contained positive results for the induction phase of a phase III clinical trial for the use of ustekinumab in UC and proved that ustekinumab had an effect in treating UC. This would give the skilled person a high degree of confidence, albeit not a certainty, that ustekinumab would also work in the maintenance phase. Meade J therefore found the patent invalid for obviousness over the Sands Slides.

The judgment can be read here.