T 0416/18 (Human colorectal cancers/JOHNS HOPKINS UNIVERSITY) 09-05-2023

Admittance and consideration of new lines of argument based on the full copy of document A6 (Article 13(2) RPBA)

1. As reaction to the board's communication pursuant to Article 15(1) RPBA, the respondent submitted a new version of document A6, namely a version comprising a full copy of Chapter 6 rather than just page 102 thereof as was on file. The respondent argued that this full copy of document A6 should be admitted because, since the board considered for the first time that document A6 was relevant for the patentability of the claims, there had been a change from the previous position taken by the opposition division as regards the relevance of document A6. The document A6 which was already in the proceedings was only an excerpt, whose teaching, if not read in context, could not be properly understood. Moreover, the appellant would not be presented with new information since they were certainly aware of the content of the full chapter from which they had isolated one page.

2. The board cannot accept this view, as the argumentation used in the board's preliminary opinion under Article 15(1) RPBA was essentially based on appellant's arguments on inventive step submitted during appeal proceedings and also in opposition proceedings (statement of grounds of appeal, page 14, last paragraph to page 15 first paragraph and notice of opposition starting on page 12). In its reply to the statement of grounds of appeal, the respondent referred to document A6 (page 4, fourth paragraph) but, while observing that only an extract of document A6 was on file, did not at that point consider it necessary to file a full copy of document A6. As the board's communication contains a preliminary opinion based on issues raised by the parties and their arguments, that communication cannot be taken as a justification for submitting supplementary evidence at this late stage of the proceedings.

3. Even if the opposition division's decision did not indicate why document A6 was not relevant, the fact that the board does not follow this view and instead agrees with the appellant is one of the possible outcomes to be expected and cannot justify the admittance of respondent's new evidence and arguments based thereon at such a late stage of the proceedings. Hence the board considers that there were no exceptional circumstances justified with cogent reasons for the late submission of the new evidence and argumentation. Hence, the board decided not to admit them into the appeal proceedings.

Main request - Inventive step (Article 56 EPC) - Claim 1

4. The present invention relates to a method for diagnosing colorectal cancer in a human, wherein a somatic mutation in the KRAS gene or its encoded mRNA or protein, said mutation being codon K117N, is determined in a test sample (patent, paragraph [0005], claim 2).

5. Document A2 is the closest prior art. This has not been disputed. It discloses a pyrosequencing method for detecting known KRAS mutations, a technique that allows the detection of a minority of mutant KRAS alleles among abundant wild-type alleles (title, abstract). Using said technique, mutations at codons 12 and 13 of KRAS were detected in patients who developed incident colorectal cancer (Figure 2 of document A2).

6. The difference between document A2 and the method of claim 1 is that the KRAS mutation to be detected according to claim 1 is the K117N. A technical effect, let alone an unexpected technical effect, associated with this difference is neither shown in the patent nor made plausible from the prior art. This has not been disputed.

7. Accordingly, the board agrees with the objective technical problem defined in the decision under appeal and formulated by the appellant, which is to provide an alternative method for diagnosing colorectal cancer in humans. The board moreover agrees that the method of claim 1, comprising the detection of a KRAS mutation K117N, solves this problem.

8. Thus, it remains to be assessed whether or not the skilled person starting from the disclosure in document A2 and faced with the problem defined above, would have arrived at the claimed solution in an obvious manner.

9. The board agrees with the respondent and the decision under appeal that neither document A2 nor any other available prior art documents disclose or point at the specific KRAS mutation K117N being associated with human colorectal cancer. However, since the technical problem is just the provision of alternative methods, there is no need for a pointer in the prior art. Moreover, documents A5 and A6 already pointed to mutations at K117 being linked to cancer.

Document A5, which focuses on mutations at codon 63 in a chemically induced rat renal tumour (title, abstract), teaches that the incidence of mutations in genes of the ras family is high in carcinomas of the colon and explicitly mentions mutations in codon 117 (document A5, page 136 left column, first paragraph). On page 137, right column, first paragraph of the "Discussion" section, document A5 again mentions codon 117, among others, as a point mutation, albeit rare, associated with oncogenic ras. Document A6, on the other hand, teaches codon K117, as well as codons D116 and D119, as belonging to a so-called second set of "activating" ras mutations.

10. The skilled person starting from document A2 faced with the technical problem identified above could and would also have turned to documents A5 or A6 and select another/further KRAS somatic mutation for diagnosing human colorectal cancer. Since there was no need to achieve any technical effect beyond what was disclosed in prior art document A2 and there was no technical difficulties to be expected, the skilled person would have been motivated to use the method for diagnosing colorectal cancer in a human disclosed in document A2 but applying another specific somatic KRAS mutation mentioned in the prior art.

11. Thus, the combination of the method disclosed in document A2 with the teaching of documents A5 or A6 would have led the skilled person to consider the KRAS gene previously identified as a colorectal cancer gene, and the second set of "activating" mutations in positions 116, 117 or 119 within this gene (document A6), known to negatively affect guanine nucleotide binding activity thereby resulting in a constitutively activated protein. Without the need of a pointer, the skilled person would have selected one mutation, among all the equally possible mutations in this motif, including mutation K117N, and accordingly would have arrived at the solution of claim 1 in an obvious manner. Consequently, the claimed subject-matter does not involve an inventive step.

12. The respondent argued that document A5 mentioned the mutation at codon 117 in a different context, namely in the context of reference [14], which related to oncogenes activated in a B6C3F1 hybrid mouse causing liver tumours. Hence document A5 referred to mutation at codon position 117 in association with the wrong species and the wrong disease. Moreover it neither specified the member of the ras gene family nor the specific mutation concerned.

The first reference to codon 117 in document A5 is on page 136, first paragraph of the Introduction section. This passage reads:

"Mutational activation of ras in experimental animal tumors has also been extensively documented and reviewed (5]. In both human and experimental tumors, with rare exceptions (codons 117 and 146), activating mutations occur in either of two hotspot regions in any of the ras genes: codons 12 and 13 in exon 1 or codons 59-61 in exon 2."

13. The board notes that no difference is made between human and experimental tumours in the above statement. Secondly, it is explicitly stated that, apart from the two hotspot regions where the activating mutations typically occur, they can also occur at codons 117 and 146. Even if not explicitly referring to K-ras, the reference to "any of the ras genes" embraces all known ras gene family members, such as H-ras, K-ras and N-ras. Thus, from the above passage, the skilled person would have derived that an activating mutation at codon 117 is to be expected in human tumours in any one of the ras gene family member, including KRAS.

14. The second reference to the mutation at codon 117 in document A5 is found in the first paragraph of the Discussion section (page 137, right-hand column) where it is associated with reference [14]. As is apparent from the title of reference 14 in the list of References (page 139, right column), this publication is related to oncogenes activated in a B6C3F1 hybrid mouse causing liver tumours. In view of the whole teaching of document A5 however, the skilled person would have no reason to consider that this teaching in the Discussion section was limited to the specific content of the cited publication it refers to. Clearly, said publication relates to animal models, from which the skilled person extrapolates to humans (in agreement with the first passage above, which explicitly mentions both experimental and human tumours).

15. As to document A6, the respondent argued that although K117 is explicitly mentioned (third paragraph), it was not assigned to the KRAS gene nor limited to K117N, let alone to the type of tumour in which this mutation occurs. Moreover, document A6 stated that, despite many mutations generated in vitro, only a few were found in human malignancies (last paragraph).

16. The third paragraph of document A6 reads as follows:

"A second set of "activating" mutations - although they are very rarely found in human tumors - are those containing substitutions in the binding site of the guanine ring. Thus, mutation of either D116, Kl17, or D119 results in an appreciable reduction in the affinity of the protein for GDP- giving rise to a molecule which will exchange its nucleotide more rapidly, presumably spending more time in the active GTP-bound state.**(38,39)"

17. From this passage, the skilled person would clearly derive that another set of "activating" mutations at position D116, Kl17, or D119, albeit very rarely found in human tumors, results in an appreciable reduction in the affinity of the protein for GDP. As discussed above in relation to document A5, the fact that this statement is supported by references 38 and 39 would not limit its teaching to the experimental setting of the publications cited as references. As to the respondent's argument that the specific mutations referred to in the second sentence of this paragraph were not necessarily the second set of "activating mutations" referred to in the first sentence, the board cannot share this view and notes that, even in the absence of the linking term "thus" at the beginning of the second sentence, this would be still the most logical interpretation.

The last paragraph of document A6 reads:

"Despite the very large number of transforming Ras mutations which can be produced in vitro, relatively few have been identified in human malignancies. It is reasonable to suppose that this is either because certain areas of the gene may be considered "mutational hotspots" or, perhaps more probably, the NIH 3T3 focus assay does not always reflect the ability of the protein to promote oncogenesis in vivo."

18. Contrary to the respondent's arguments, this passage does not raise doubts as to the value of the K117 mutation in a method for diagnosing human cancer; even if this mutation was one of the artificially generated mutations tested for transforming activity in NIH3T3 cells, still the skilled person would be motivated to test it further rather than discard it. Moreover, in the third paragraph of document A6, it is in fact disclosed as a mutation which is found, albeit rarely, in human tumours.

19. The respondent further argued that neither document A3, which related to KRAS mutations in colorectal cancer, nor document A7, which was a review of activating mutations in the three ras genes, H-ras, K-ras, and N-ras, disclosed or suggested that there could be mutations in KRAS other than at codons 12, 13, or 61 in human malignancies, let alone another specific K117N mutation in KRAS associated with colorectal cancer. Although this fact is undisputed, the contrary statement that no other mutations are to be expected in KRAS is incorrect and is inconsistent with the teaching of documents A5, A6, which refer to rare mutations events in human tumours in any of the ras genes at codon 117. The limited number of 30 and 60 samples tested and the conventional test assays and animal models might explain why the rare mutation position 117 was not detected in documents A3 or A7. Hence, the information allegedly missing from these documents cannot call into question the assertions made in two other independent prior art documents A5 and A6.

20. Finally, the respondent also argued that even if the mutation K117 could be detected, it would not be possible to determine whether or not it was a non-functional "passenger" mutation elicited by repeated rounds of cell division in the tumor or in the progenitor stem cells. Reference was made to Example 4 of the patent.

21. The board disagrees. The statistical methods in the patent estimate the probability that the number of mutations observed in a given gene during a discovery and validation screen is greater than the one expected from the background mutation rate. The cancer mutation frequency (CaMP) score is the result of this analysis and reflects a cancer mutation prevalence for each gene. This score intends to distinguish between genes likely to contribute to tumorigenesis from those in which the mutations occur by chance (patent, paragraph [0040]). Since the KRAS gene is known in the art to be mutated in cancer tissues, there is no need for a skilled person to apply the statistical method of the patent to establish whether or not KRAS is involved in tumorigenesis. The CaMP score for KRAS is high and confirms this fact (patent, paragraph [0043] and Table 3 in Figure 7A).

22. Therefore, the subject-matter of claim 1 of the main request lacks an inventive step (Article 56 EPC).

First auxiliary request - Inventive step (Article 56 EPC) - Claim 1

23. Claim 1 of this request further specifies that the test sample is a colorectal tissue sample or a suspected colorectal cancer metastasis and the normal sample is a colorectal tissue sample. The respondent provided no arguments as to why this further limitation contributed for inventive step. The board also fails to see that this feature achieves a technical effect going beyond what is achieved by the claimed method of the main request. Hence, the objective technical problem for the assessment of inventive step starting from the disclosure in document A2 representing the closest prior art is the same as for the main request.

24. Considering that document A2 discloses a method which involves identifying the sample as colorectal cancer when the somatic mutation is determined relative to a normal sample of the human (e.g. Figure 2 and its legend), the claimed method of the first auxiliary request lacks an inventive step for the same reasons as the main request.