T 1011/17 12-03-2020

Reasons for the Decision

1. Admissibility of the appeal

The appeal complies with Articles 106 to 108 EPC and Rule 99 EPC and is therefore admissible.

2. Inventive step

Patent in suit

2.1 The patent in suit explains in its "Background" section (see paragraphs [0002] and [0003]) that imatinib (or "STI-571") was being used in the treatment of chronic myeloid leukemia (also called chronic myelogenous leukemia). The drug blocks the activity of the tyrosine kinase protein BcrAbl, an abnormal protein driving the overproduction of abnormal white blood cells characteristic of leukemia. The aberrantly activated tyrosine kinase BcrAbl is causally associated with chronic myeloid leukemia (CML) and acute lymphocytic leukemia (ALL).

2.2 It was known that many patients eventually developed resistance to imatinib treatment due to mutations in the cancer cells, in particular point mutations in the bcrabl gene (see paragraphs [0002] and [0004]).

2.3 The patent in suit seeks to provide a treatment for imatinib-resistant BcrAbl-positive leukemia by providing a suitable alternative to imatinib (see paragraphs [0004], [0012] and [0014]).

2.4 This is achieved, according to the patent in suit (see paragraph [0001] and independent claims 1 and 7), by using bosutinib or a salt of it to treat patients suffering from BcrAbl-positive leukemia having the 949T>C resistance-associated mutation in the bcrabl gene. According to example 1 of the patent in suit (see paragraph [0046]), the amino acid change which corresponds to the nucleic acid mutation 949T>C is F317L (a phenylalanine [F] at amino acid position 317 mutated to leucine [L]).

Starting point in the prior art

2.5 It is common ground that prior-art document D8 represents a suitable starting point for the assessment of inventive step. The board has no reason to use a different starting point.

2.6 Document D8 relates to a study investigating the in vitro and in vivo activity of SKI-606 (bosutinib) against imatinib-resistant BcrAbl-positive neoplastic cells (see D8: title). In particular, four clinically relevant mutants of BcrAbl were selected to be assayed with bosutinib, namely T315I, Y253F, E255K and D276G (D8: paragraph bridging pages 11320 and 11321). The document discloses in vitro results showing that bosutinib retained activity in three of the four point mutants tested, the exception being T315I, an imatinib-resistant mutation known to be particularly difficult to treat (see D8: abstract and page 11317, table 1). D8 also reports that the activity against Y253F, E255K and D276G was confirmed in vivo in mouse models (see D8: page 11318, Figure 4; page 11321, column 1, lines 13 to 30).

2.7 D8 does not discuss the 949T>C resistance-associated mutation in the bcrabl gene (corresponding to F317L). However, the amino acid F317 is marked as a residue of interest in the schematic diagram of interactions of docked bosutinib with the intermediate conformation of the Abl protein in Figure 6D of D8.

Objective technical problem and solution

2.8 The parties agreed that:

- the technical feature distinguishing the claimed subject-matter from the disclosure of document D8 was the treatment of the patient sub-population carrying the 949T>C/F317L mutation,

- the objective technical problem to be solved was the identification of a further patient subgroup within the imatinib-resistant BcrAbl-positive leukemia patient population that would benefit from treatment with bosutinib.

2.9 The appellant did not dispute that the subject-matter defined in claims 1 and 7 solved this technical problem.

Obviousness of the solution

2.10 Thus, the only issue to be determined is the obviousness of the claimed solution.

2.11 Seeing that the focus of document D8 is on screening for the activity of bosutinib against several CML models of resistance to imatinib, and faced with the technical problem defined above (see point 2.8), the person skilled in the art would have consulted further documents relating to the treatment of imatinib-resistant leukemia with second-generation BcrAbl tyrosine kinase inhibitors, such as the review article D3.

2.12 D3 mentions the problem of imatinib resistance and refers to recent progress made in the development of second-generation drugs designed to combat imatinib-resistant mutant forms of BcrAbl (see D3: abstract). The drugs mentioned in D3 include nilotinib, dasatinib and bosutinib (see page 9, 2.2.3 (SKI-606), where it is also mentioned that bosutinib had been reported to have shown activity against Y253H, E255V, E255K and F359V mutant BcrAbl).

Table 1 on page 6 of D3 presents a comparison of imatinib, nilotinib and dasatinib for their effects on autophosphorylation and proliferation in cells expressing native BcrAbl and "some of the most prevalent imatinib-resistant mutant forms of the enzymes identified in patients", which include F317L BcrAbl.

2.13 Thus, it was known that the F317L mutation belonged to a group of prevalent imatinib-resistant mutations. The activity of other second-generation BcrAbl tyrosine kinase inhibitors had already been screened against this group of mutations, including against the F317L mutation.

2.14 While it was not in dispute among the parties that numerous imatinib-resistant mutations were known from the prior art (D3: page 3, column 2, last full sentence mentions "over 35 [...] mutant forms"), the board agrees with the appellant that the person skilled in the art seeking to solve the technical problem would have taken a "try-and-see" approach for the routine screening for the activity of bosutinib against further prevalent BcrAbl mutations.

2.15 The board is not convinced that the person skilled in the art would have made predictions about the activity of bosutinib against the F317L mutation which would have resulted in an expectation of failure preventing this mutation from even being taken into consideration for preliminary screening.

The skilled person would not have formed an expectation that the F317L mutation was going to be as difficult to treat as the notorious T315I mutation (known to be uniquely difficult) as the F317L mutation differed from the T315I mutation in that it was known to be susceptible to treatment with nilotinib and dasatinib (see D3: table 1) and to confer only moderate resistance to imatinib (see D5: page 121, column 1, first full paragraph).

While, in a study comparing the activity of dasatinib against 15 imatinib-resistant BcrAbl mutants, the F317L mutant was found to be less sensitive to the drug than other mutants, this was attributed in the prior art to structural moieties of dasatinib, namely pyrimidine and thiazole rings, which are not present in bosutinib (see D23: page 5795, column 1, lines 1 to 23 and D3: page 4, Figure 1). Hence, it would not have appeared meaningful or conclusive to extrapolate the properties of dasatinib to bosutinib.

Therefore, the person skilled in the art would have had no reason to adopt a particularly sceptical attitude. There existed no technical prejudice or conclusive theory which would have dissuaded the person skilled in the art from testing bosutinib against the F317L mutation using in vitro screening and appropriate animal models.

2.16 The outcome of such tests was, however, uncertain.

2.17 As pointed out moreover by the respondent, especially in cases where genetic mutations are concerned, neither in vitro nor in vivo laboratory models, albeit informative as a first approach, are reliably predictive of actual clinical efficacy since their relevance to mechanisms of resistance in the leukemic cells of patients being treated is unknown (see document D4, page 7390, column 2, second paragraph). It was not contested by the appellant that clinical relevance is difficult to predict in this field.

2.18 As a consequence, there is no straightforward road which would automatically have led the person skilled in the art from the combined teaching of documents D8 and D3 to the subject-matter of claims 1 and 7.

2.19 Present claims 1 and 7 are, in fact, not concerned with a BcrAbl mutant that passed certain routine screening tests but with a (further) BcrAbl mutant susceptible to the therapeutic use of bosutinib in the treatment of BcrAbl-positive leukemia.

2.20 In view of the considerations set out in point 2.17 above, more extensive studies would have been required to confirm the efficacy of bosutinib in human leukemia patients with imatinib-resistant cancer cells carrying the F317L mutation. In this context, the person skilled in the art would no longer have been in a "try-and-see" situation but would have required a reasonable expectation of success as an incentive for moving to further much more extensive and larger studies in a clinical setting.

2.21 At the priority date of the patent in suit, it may have been obvious to include the F317L mutation in routine screening tests with bosutinib, but the outcome of such tests was as yet unknown. Based on the information on file, it cannot be confirmed that the outcome of such tests would have provided an incentive for the skilled person to proceed with more extensive clinical testing specifically in the case of the F317L mutation.

2.22 Nor is the information provided originally in document D8 sufficient to give rise to a positive expectation of success.

According to D8, bosutinib was found to be active against three of the four point mutations tested. The three "treatable" mutations (Y253F, E255K and D276G) are said to be representative of two different mechanisms by which mutations can cause resistance to imatinib (D8: paragraph bridging pages 11320 and 11321). There is no teaching in D8 which would have enabled the person skilled in the art to identify an evident similarity or analogy between the F317L mutation and any of Y253F, E255K and D276G. Thus, the data presented in D8 cannot be extrapolated to conclusively predict the efficacy of bosutinib with regard to the F317L mutation. In any case, the tests reported in D8 (see point 2.6 above) did not go beyond in vitro screening and in vivo animal models and their results, or the statement in D8 that bosutinib was generally found (in such models) to be more active than imatinib, are not necessarily indicative of clinical efficacy.

This assessment is not changed by the fact that document D3 mentions certain further mutants believed to be affected by bosutinib (D3: page 9, 2.2.3; see point 2.12 above).

According to Figure 6D of D8, the F317 residue of the Abl protein was believed to engage in a van der Waals interaction with bosutinib, while the T315I mutation, known to be the most difficult mutant to tackle, was thought to be in a critical position in the binding pocket of the protein and to be involved in direct electrostatic interaction, by hydrogen bonding, with both imatinib and bosutinib. This difference could not, however, have amounted to providing an expectation of clinical success for bosutinib in the case of the F317L mutation.

2.23 For these reasons, the subject-matter of independent claims 1 and 7 and the dependent claims involves an inventive step within the meaning of Article 56 EPC.