T 0400/99 (Gamma Interferons/GENENTECH) 12-10-2000

1. The appeal is admissible.

Article 123(2) EPC

2. At the oral proceedings, the appellant sought to object that the claims extended beyond the content of the application as filed. The board, however, refused to allow this issue into the proceedings pursuant to Article 114 (2) EPC, as it had not been submitted in due time, the respondent had not had the opportunity to assess it or to provide its own counterarguments, and its relevance was not immediately apparent.

Novelty

3. The board agrees that the specific embodiment Met-1, des-[Cys1-Tyr2-Cys3]-IFN- is not mentioned expressis verbis in document (A1). The question to be answered, however, is whether or not this compound is nevertheless directly and unambiguously derivable from said document.

4. Turning to the passages of document (A1) relied upon by the appellant for questioning the novelty, the board views them as comparable to generic chemical formulae containing multiple substituents R1, R2, R3, etc., contrary to the appellant's contention. This is because the wording:

"Other classes of IFN analogs consists of polypeptides wherein the [Trp39] residue is replaced by [Phe39], and/or wherein one or more of the methionine residues at amino acid positions 48, 80, 120 and 137 are replaced by, e.g., leucine, and/or wherein cysteines at amino acid positions 1 and 3 are replaced by, e.g., serine or are completely eliminated." (paragraph bridging pages 48 and 49 of document (A1)),

interpreted in the light of the paragraph on page 49, line 33 to page 50, line 13, relating to three examples of replacement/deletion of the cysteines at positions 1 and 3, is susceptible of being expressed in an equivalent but more concise manner under the form of a generic chemical formula, for example:

Met-1-R1-R2-R3-U1-R39-U2-R48-U3-R80-R81-U4-R120-U5-R137-U6

wherein R1 to R137 are the amino acid substituents at the positions indicated in the cited passages or the "native" residue (cf. the "and/or"), wherein R1 and/or R3 or R1 to R3 may additionally represent a covalent bond (ie, residues R1 and/or R3 or R1 to R3 are absent) and wherein U1, U2, U3, U4, U5 and U6 designate, for the sake of simplicity, the unaltered stretches of IFN- . Both the "phrased" version, ie, the cited passages of document (A1) and the above more concise "formula" version, having identical content as regards the technical information which can be derived therefrom, conceptually cover several hundreds of possible IFN- variants and are interchangeable for the purpose of the present decision.

5. It is jurisprudence of the Boards of Appeal that a class of chemical compounds, defined only by a general structural formula having one or more variable group(s) does not specifically disclose each and any of the individual compounds which would result from the mental combination of all possible variants within such group(s). Rather, the novelty of an individual chemical compound can only be denied if there is a direct and unambiguous disclosure in the prior art of this very same compound in the form of a technical teaching (see T 181/82, OJ EPO 1984, 401, No. 8 of the reasons, and T 296/87, OJ EPO 1990, 195, Nos. 6 and 7 of the reasons).

6. Turning to the present case, it is therefore not sufficient for denying novelty of the claims at issue covering Met-1, des-[Cys1-Tyr2-Cys3]-IFN- , that the latter compound belongs conceptually to the group of possible variants embraced by the cited passages of document (A1) and its equivalent, more concise formula set out above, unless there is a direct and unambiguous pointer in the prior art document to this individual member. When applying this principle to document (A1), it turns out that Met-1, des-[Cys1-Tyr2-Cys3]-IFN- is neither exemplified, nor specifically mentioned, nor the inevitable result of carrying out the instructions disclosed. Under these circumstances, in accordance with the rationale emerging from decision T 181/82 (loc. cit.) and T 296/87 (loc. cit.), it must be concluded that there is no direct and unambiguous disclosure in the prior art of this compound in the form of a technical teaching. Consequently, the subject-matter of claim 1 and dependent claims 2-26 meets the requirements of Article 54 EPC. This conclusion also applies to claims 1-21 for the Contracting State AT.

7. The appellant argues that Met-1, des-[Cys1-Tyr2-Cys3]-IFN- is the inevitable result of carrying out the instructions given on page 49, lines 33-34 of document (A1) to "only alter subunit IF-4" and to assemble a DNA encoding for an IFN- analog using the IF-4 along with subunits IF-1, IF-2 and IF-3 of Table IV, which DNA yields upon expression this protein. In the board's judgement, however, this passage cannot be considered as an instruction to alter only subunit IF-4 since it does not per se exclude the possibility of modifying one or more of subunits IF-1, IF-2 and IF-3 of Table IV. There is also no teaching in document (A1) that combining modified subunit IF-4 with unmodified subunits IF-1, IF-2 and IF-3 of Table IV is a privileged direction to be followed, among the remaining hundreds of possible mental combinations, when making IFN- variants. Under these circumstances, the conclusion cannot be drawn that Met-1, des-[Cys1-Tyr2-Cys3]-IFN- is the inevitable result of carrying out the instructions given in document (A1).

8. The appellant maintains that document (A1) discloses Met-1, des-[Cys1-Tyr2-Cys3]-IFN- variants which are not excluded by the disclaimer in claim 1 of the patent in suit. However, in the board's view, claim 1 also requires that any variant should "exhibit biological activity equivalent to or greater than that of gamma interferon of the amino acid sequence 1-143 depicted in Fig. 1 hereof". Since the experimental results referred to in Table VI on page 52 of document (A1) show that the alterations proposed in document (A1) are all deleterious to the biological activity, the above expression in claim 1 is a distinguishing feature vis-à-vis any interferons- variant disclosed by document (A1).

Inventive step (Article 56 EPC)

9. The closest prior art is represented by document (A1) relating to various Cys1-Tyr2-Cys3-IFN- variants. This document comprises on page 49, lines 1 to 6, a suggestion to replace or eliminate Cys1 and/or Cys3 for avoiding intermolecular bonds, and hence for facilitating isolation of the variants. This cysteine replacement/deletion class is illustrated by three examples on page 49, line 33 to page 50, line 13, one of which is Met-1, des-[Cys1-Tyr2-Cys3]-Lys81-IFN- . The problem to be solved by the patent in suit is to provide IFN- variants that are more active and more stable than the prior art ones (see eg the patent in suit on page 5, lines 33-35 and page 12, lines 40-43). In view of the experimental results of Table 2 and Fig. 3 showing respectively the superior antiproliferative activity and stability of des-[Cys1-Tyr2-Cys3]-IFN- compared with those of Cys1-Tyr2-Cys3-IFN- , the board is satisfied that the above problem is solved by the claimed subject-matter.

10. Departing from the disclosure of document (A1), the proper question to be addressed is whether the skilled person is motivated to make an IFN- analog containing the Met-1, des-[Cys1-Tyr2-Cys3] modification without the [Lys81] modification or in general to go into the direction of removing Cys1-Tyr2-Cys3 alone, optionally in the presence of a C-terminal truncation, in order to solve the problem of providing IFN- variants that are more active and more stable than the prior art ones.

11. The board observes that document (A1) opens the way to a great number (several hundreds) of possible variants. There is no pointer for the skilled person to go into the direction of removing Cys1-Tyr2-Cys3 alone in order to obtain IFN- variants that are more active and more stable than the prior art ones. Rather, the experimental results referred to in Table VI on page 52 of document (A1) show that the alterations, including removing Cys1-Tyr2-Cys3 proposed in this Table are deleterious to the biological activity: if anything, this fact does not appear to be encouraging for somebody looking for more active compounds.

12. It is true that the document comprises on page 49, lines 1 to 6 a suggestion to replace or eliminate Cys1 and/or Cys3 for avoiding intermolecular bonds, and hence for facilitating isolation of the variants. However, nothing is said about the destination of Tyr2 or about the necessity of effecting this change alone. Therefore, a suggestion cannot be derived from this passage to eliminate Cys1-Tyr2-Cys3 alone, optionally in the presence of a C-terminal truncation, in order to obtain IFN- variants that are more active and more stable than the prior art ones.

13. The appellant argues that C-terminal truncation is either already contemplated in the prior art or reduces activity and, hence, these C-terminal truncated variants do not involve an inventive step.

14. In the board's judgement, however, even assuming that the C-terminal truncation is an obvious measure already contemplated in the prior art, the des-[Cys1-Tyr2-Cys3] feature shared by all the claimed compounds, as seen above, is not.

15. As for the reduction of biological activity achieved by C-terminal truncation, it has to be noted that the distinguishing feature vis-à-vis the prior art is the des-[Cys1-Tyr2-Cys3] feature exhibited by all the claimed compounds rather than the C-terminal truncation, which is an optional feature. It is this distinguishing feature which has to be responsible for the improved effect invoked in the patent in suit, following decision T 197/86 (OJ EPO 1989, 371). Therefore, the proper question to be answered in the context of the inventive step of C-terminal truncated variants is whether or not the des-Cys1-Tyr2-Cys3 feature achieves an improved effect also for these C-terminal truncated variants. Reduction of the biological activity of Met-1, des-[Cys1-Tyr2-Cys3]-IFN- (see Table 1 of the patent in suit) upon C-terminal truncation, as invoked by the appellant, fails to provide an answer to this question and is therefore irrelevant. Rather, in order to deny the presence of an inventive step, the board must have been persuaded that the des-[Cys1-Tyr2-Cys3] feature fails to achieve an improved effect in the case of C-terminal truncated variants, ie that a des-[Cys1-Tyr2-Cys3], C-truncated IFN- variant is less active than the corresponding Cys1-Tyr2-Cys3, C-truncated IFN- variant having the same C-terminal. However, since there is no evidence before the board that such is the case, the patentee should be given the benefit of doubt that the des-[Cys1-Tyr2-Cys3] feature achieves an improved effect irrespective of C-terminal truncation. Hence, the presence of an inventive step has to be acknowledged also for the claimed C-terminal truncated IFN- variants.

16. Therefore, it must be concluded that the claimed Cys1-Tyr2-Cys3 deleted IFN- variants do not follow in an obvious fashion from the prior art. Consequently, the subject-matter of claim 1 and dependent claims 2-26 meets the requirements of Article 56 EPC. This conclusion also applies to claims 1-21 for the Contracting State AT.