Bristol-Myers Squibb v Isenbruck Bösl Hörschler and others (EPO, T 488/16, 2017)

Dhenne Avocats · 11 October 2026

Our analysis

Summary

Technical Board of Appeal 3.3.01 upheld the revocation of Bristol-Myers Squibb’s European patent EP 1 169 038, limited on appeal to dasatinib and its salts. The application merely asserted that the compounds of the examples had been tested and shown activity as protein tyrosine kinase (PTK) inhibitors, without any data. As the effect was not plausible at the filing date, the post-published documents showing dasatinib’s activity were disregarded, the problem was reduced to providing a further chemical compound and inventive step was denied. The Board also refused a referral to the Enlarged Board of Appeal.

Facts and procedure

The patent, entitled « Cyclic protein tyrosine kinase inhibitors », was opposed by four opponents. By a decision posted on 9 February 2016 the Opposition Division revoked it: the main request was held insufficiently disclosed, the first auxiliary request contrary to Article 123(3) EPC, and the second novel but lacking inventive step (points I, III and IV). On appeal the proprietor defended only that second request, whose sole claim is to dasatinib or salts thereof (point V). At the oral proceedings of 1 February 2017 the proprietor asked for four questions on plausibility and burden of proof to be referred to the Enlarged Board (point X).

The applicable law

Under Article 56 EPC, the Board recalled that inventive step is assessed at the effective date and that post-published evidence may be taken into account only if it is already plausible from the disclosure that the problem is solved (point 4.2, citing T 1329/04 and T 1043/10). Referral is governed by Article 112(1)(a) EPC (point 6.1).

Question

Does a general assertion of activity, without verifiable data, make the technical effect of a compound within a very broad formula plausible, so that post-published evidence can be relied on for inventive step?

Decision

The application discloses 580 compounds, including dasatinib (example 455), and describes assays only generically, without any results (points 4.3 and 4.5). A mere verbal statement that compounds « have been found active », without verifiable technical evidence, does not make it credible that the problem is solved, the formula covering millions of candidates (point 4.5). The proprietor’s expert declarations did not help: the opinion of highly skilled experts does not reflect the view of the notional skilled person (points 4.6.1 to 4.6.5). Experimental data are not always required, but it is a conditio sine qua non that the problem was at least plausibly solved at filing; where the effect is neither self-evident nor predictable nor based on a conclusive theoretical concept, at least some technical evidence is needed (point 4.9). On the burden of proof, the Board did not need to decide whether a presumption exists: the opponents had raised substantiated doubts which the proprietor had to rebut (point 4.10). Post-published documents (9) and (10) were therefore disregarded (point 4.19). The problem was therefore reduced to providing a further compound, which is not inventive, structural originality alone being insufficient (points 5.6, 5.7 and 5.10). The proposed questions, essentially technical, did not warrant a referral (point 6.2). The appeal was dismissed (order).

Key points for practice

• An assertion of activity without results, even accompanied by assay protocols, does not make an effect plausible for a compound drawn from a very broad generic formula (points 4.5 and 4.9).
• Narrowing the claim to a single compound during proceedings does not cure the absence of data in the application as filed (points 4.16.3 and 5.5).
• Practical point: the application must contain verifiable activity data for the compounds that may later be claimed individually; otherwise the problem is reformulated minimally.

Relevance before the UPC

Since the UPC assesses validity against Article 138(1) EPC, which includes Article 56, the analysis of plausibility and of the role of post-published evidence carries over to revocation actions concerning pharmaceutical compound patents.

Provisions applied

European Patent Convention

Art. 56; Art. 112(1)(a); Art. 117(1)

Case law cited

T 22/82; T 939/92; T 1329/04; T 578/06; T 1043/10; T 428/12; T 715/03; T 1642/07; T 1677/11 (among eighteen Board decisions cited); Human Genome Sciences v Eli Lilly [2011] UKSC 51; Generics (UK) v Yeda [2013] EWCA Civ 925; Actavis v Eli Lilly [2015] EWHC 3294

Related decisions

• Johns Hopkins University, GDF-9, EPO, 28 June 2005, applied on post-published evidence and held to be in point (points 4.2 and 4.14).
• Human Genome Sciences v Eli Lilly, UK Supreme Court, 2 November 2011, relied on by the proprietor and distinguished (point 4.16.1).

Prepared by Dhenne Avocats from the text of the decision (epo.org, official PDF of the decision), consulted on 11 October 2026. Only the official text is authoritative.