T 2423/18 (MICA binding agents/INNATE PHARMA) 22-07-2021
Download and more information:
MICA binding agents
I. The appeal is from the applicant (appellant) against the decision of the examining division to refuse European patent application No. 13 704 064.8 ("the application"), published as international patent application WO 2013/117647 with the title "MICA binding agents".
II. In the decision under appeal, the examining division held the subject-matter of the set of claims of a main request and auxiliary requests 1 to 6 to lack an inventive step (Article 56 EPC). It refused the application for this reason.
III. With their statement of grounds of appeal, the appellant re-submitted the sets of claims of the main request and of auxiliary requests 1 to 6. They submitted arguments to the effect that the claimed subject-matter of each request involved an inventive step (Article 56 EPC).
Claim 1 of the main request read:
"1. A monoclonal antibody that binds to a MICA polypeptide comprising an amino acid sequence of
SEQ ID NO: 1, to a MICA polypeptide comprising an amino acid sequence of SEQ ID NO: 2, to a MICA polypeptide comprising an amino acid sequence of SEQ ID NO: 3, and to a MICA polypeptide comprising an amino acid sequence of SEQ ID NO: 4 on the surface of cells for use in the treatment of cancer."
Claim 1 of auxiliary request 1 was identical to claim 1 of the main request but for the insertion of the additional wording ",wherein said antibody comprises a human heavy chain constant region that binds a human FyIIIA receptor or wherein said antibody is conjugated or covalently bound to a toxic agent" at the end of the wording of the claim.
IV. The appellant was summoned to oral proceedings as requested. Subsequently, in a communication pursuant to Article 15(1) RPBA 2020, the board expressed, inter alia, the preliminary view that the claimed subject-matter of the main request lacked an inventive step.
V. In response to the board's communication, the appellant withdrew former auxiliary requests 2 to 6 and filed new auxiliary requests 2 to 4.
VI. During the oral proceedings the appellant withdrew auxiliary requests 2 to 4. At the end of the oral proceedings the chair announced the decision of the board.
VII. The following documents are referred to in this decision:
D1: WO 2008/036981
D3: Yizhou et al. (2002), Human Immunology, vol.63,
No.1, pages 30-39.
D5: Duquesnoy et al. (2008), Human Immunology,
vol.69, No.12, pages 826-832.
D6: Suarez-Alverez et al. (2009), Transplantation,
vol.88, No.3, Suppl., pages S68-S77.
D7: Yizhou et al. (2008), Immunogenetics, vol.61,
No.2, pages 91-10.
D8: Spreu et al. (2006), J. Immunol., vol.177, No.5,
pages 3143-3149.
D9: "Anti-Human MICA/B Monoclonal Antibody BAMO1",
dated 1 January 2004.
Exhibits 1 and 2: filed by the applicant with the
letter dated 24 June 2016.
Exhibit 3: filed by the applicant with the
letter dated 4 January 2017.
VIII. The appellant's arguments, as far as they are relevant for the decision, are summarised as follows:
Main request
Inventive step (Article 56 EPC) - claim 1
Document D1, which represented the closest prior art, admittedly disclosed (e.g. on page 26, lines 8 to 11), inter alia, anti-MICA antibodies for use in the treatment of MICA-expressing cancer by mechanisms such as complement-dependent cellular cytotoxicity (CDC) and antibody-dependent cellular toxicity (ADCC).
However, document D1 did not disclose a concrete monoclonal anti-MICA antibody. It did not consider the polymorphism of MICA either - there are over 50 MICA alleles existent in the human population. Document D1 thus concerned the concept of a patient-specific antibody rather than antibodies which bind to multiple key MICA alleles.
The claimed antibody differed from those disclosed in document D1 in that a monoclonal anti-MICA antibody which is capable of binding to MICA alleles *001, *004, *007 and *008 was used (these alleles are referred to in the claim by polypeptides comprising the amino acid sequences of SEQ ID NO: 1 to SEQ ID NO: 4). The application disclosed nine antibodies with the binding pattern referred to in the claim (page 116, lines 14 to 16).
The effect of this technical difference was that the antibody defined in the claim was pan-allelic by virtue of sequence similarity to other alleles and permitted coverage of substantially all known MICA alleles. This permitted treatment of numerous types of tumours in a large patient population. The antibody thus obviated the need to analyse which MICA allele(s) (homozygous or heterozygous) is/are expressed by a patient in need of treatment. It also obviated the need for the development of a particular antibody for each allele. This reduced the cost of the marketed drug.
The technical problem to be solved was thus to provide an improved treatment for cancer.
The skilled person could have selected, among the antibodies disclosed in document D1, any of those suggested as suitable for therapeutic use. The claimed antibody, however, was not suggested.
Equally, none of the other cited documents suggested using a monoclonal antibody capable of binding to multiple alleles of MICA, especially to MICA alleles *001, *004, *007 and *008, for the treatment of cancer. Each of documents D2, D8 and D9 disclosed anti-MICA monoclonal antibodies for use as research/detection agents. Each of documents D3 and documents D5 to D7 related to allograft rejection and not to cancer therapy.
Document D3 disclosed two antibodies that only bound to MICA *008, two antibodies that bound only to group 2 alleles and two antibodies that bound to both group 1 and group 2 alleles (i.e. *001, *002, *004, *008 and *009.).
Document D5 (pages 830 and 831), document D6 (Table 2 on page S71) and document D7 (Table 3) each disclosed observations in which organ-transplanted patients became immunised against MICA and presented anti-MICA auto-antibodies. They illustrated that sera isolated from human donors never contained antibodies for each of MICA *001, *004, *007 and *008 alleles. Immuno-dominant epitopes on MICA were thus not present on each of these alleles.
The skilled person was thus dissuaded from developing a single antibody directed against several alleles of MICA for a therapeutic purpose as claimed. If anything, the skilled person would deduce from the aforementioned disclosures that an antibody directed against several alleles of MICA could not be used for a therapeutic purpose.
Document D9 was a product sheet disclosing information on the mouse monoclonal antibody BAMO1, a monoclonal antibody binding to the MICA alleles *001, *004, *007 and *008 and to the MICB allele *02. The product sheet, however, only indicated applications for in vitro use, i.e. flow cytometry, ELISA and immunoblot. It further stated the conditions that the mAB was for research use only and not for use in diagnostic or therapeutic procedures.
The skilled person had no motivation to choose the reagent of document D9 as a solution to the technical problem, because the antibody of document D9 was not disclosed for the same purpose as the antibodies that were disclosed in document D1 or claimed. In fact, the combining of the disclosures of document Dl with the disclosure of document D9 was the result of an ex post facto analysis. Once the solution as claimed was known, it could be perceived as obvious to look for a document disclosing an antibody capable of binding to multiple alleles of MICA. However, in the absence of that knowledge, the skilled person had no reason or motivation to combine the disclosures of documents D1 and D9.
The BAMO1 antibody itself was not suitable for human cancer therapy as it was a mouse antibody and its Fc part was not bound by the human Fc receptor.
Exhibits 1 to 3 provided data obtained with BAMO1 and a chimeric derivative thereof. The data illustrated that, even if the skilled person had combined the disclosures of documents D1 and D9, the results with the chimeric derivative would have discouraged or even dissuaded the skilled person from using such an antibody for the treatment of cancer.
Exhibit 1 demonstrated that BAM01 bound to MICA*007 allele with a very high EC50. Thus, while BAMO1 might bind to MICA *007 and other MICA alleles sufficiently for detection in research uses, it was not effective for inducing ADCC as an in vivo therapeutic agent.
Exhibit 2 demonstrated that a chimeric derivative of BAMO1, which was indeed usable in humans, did not induce ADCC at concentrations consistent with its use as a therapeutic agent with MICA *007 target cells.
Exhibit 3 further demonstrated that, in order to achieve receptor saturation in a MICA *007 individual, BAM01 would need to be dosed, for intravenous administration, in the range of 100 mg/kg (i.e. around 6 000 mg for an average human) at least every week. Such a dose and regimen did not provide a treatment of cancer compatible with a therapeutic use.
The results presented in Exhibits 1 to 3 represent experiments that the skilled person would conduct when undertaking to implement a BAM01 derivative for human cancer therapy. In view of those results, the skilled person would not expect any BAM01-derived antibody to ever bind with an affinity and specificity to each of cell-surface MICA *001, *004, *007 and *008 alleles for it to be useful in therapy.
It was implicit in the wording of the claim that the claimed antibody was also useful in the treatment of cancer expressing the MICA alleles mentioned in the claim. The skilled person would thus be dissuaded from continuing to develop the antibody as referred to in the claim, because it was not useful for treating cancer expressing the MICA *007 allele.
Figure 11 of the application, showing results of testing in a mouse long-term RAJI-MICA*01 tumour model in example 11, demonstrated that the claimed antibodies were superior to antibodies only recognising two MICA alleles.
The claimed subject-matter thus involved an inventive step.
Auxiliary request 1 - claim 1
Inventive step (Article 56 EPC)
Essentially the same arguments as submitted for claim 1 of the main request applied to this claim.
IX. The appellant requested that the decision under appeal be set aside and that a patent be granted on the basis of the set of claims of the main request, or, alternatively, of auxiliary request 1, both filed with the statement of grounds of appeal.
1. The appeal complies with Articles 106 to 108 and
Rule 99 EPC and is admissible.
Main request and auxiliary request 1 - claim 1
Claim construction
2. What is claimed is a monoclonal antibody that binds to four MICA polypeptides comprising sequences of the *001, *004, *007 and *008 alleles (these alleles are referred to in the claim by polypeptides comprising the amino acid sequences of SEQ ID NO: 1 to SEQ ID NO: 4) on the surface of cells for use in the treatment of cancer. In auxiliary request 1, the antibody is further specified to comprise a human heavy chain constant region that binds a human FyIIIA receptor (see section III).
3. xcdIt is specified in the claim that the antibody binds to the four alleles comprising the indicated sequences on the surface of cells. The claim does not require, however, that this antibody should also be suitable for the treatment of all cancers expressing each particular one of these alleles.
4. The board cannot, therefore, concur with the appellant that it is implicit in the wording of the claim that the defined antibodies are useful also in the treatment of each and every cancer expressing one of the particular MICA alleles mentioned in the claim.
Inventive step (Article 56 EPC)
5. The board agrees with the appellant that the anti-MICA antibodies for use in the treatment of MICA-expressing cancer by complement-dependent cellular cytotoxicity (CDC) or antibody-dependent cellular toxicity (ADCC), as disclosed in document D1 (see e.g. page 25, line 29, to page 26, line 11), represent the closest prior art for the assessment of inventive step of the claimed subject-matter.
6. The board also agrees with the appellant that the difference between the antibodies and use disclosed in document D1 and the claimed antibody and use is that the latter antibody is capable of binding to MICA alleles *001, *004, *007 and *008 (but see point 4).
7. However, the board has not seen evidence that the antibody in the claim, as alleged by the appellant, is pan-allelic in MICA binding to such an extent that it permits coverage of substantially all known alleles, thus permits treatment of numerous types of tumours in a large patient population and thus obviates the need to analyse which MICA allele is expressed by a given patient in need of treatment. Equally, in fact, the board has not seen evidence that the claimed antibody is capable of treating tumours expressing each of the particular MICA alleles mentioned in the claim.
8. Nevertheless, the board can accept that, as was argued by the appellant, the "multi-allelic" binding character of the claimed antibody (here "tetra-allelic") allows the coverage of more than one MICA allele and thus treatment of different MICA allele-expressing types of cancer.
9. Accordingly, the board accepts this technical effect to justify that the objective technical problem, as submitted by the appellant, can be formulated as the provision of an improved treatment for cancer.
10. Accordingly, when assessing obviousness, it needs to be established whether or not the skilled person, aiming to improve cancer treatment in the sense that cancers expressing different MICA alleles can be treated using a single antibody, would have provided the antibody referred to in the claim.
11. Document D1 itself, as submitted by the appellant, does not consider or refer to the polymorphism of MICA - there are over 50 MICA alleles present in the human population (see application, page 3, lines 19 to 30). However, this polymorphism undoubtedly formed part of the common general knowledge of the person skilled in the art and the appellant has not argued differently.
12. The board notes further that a number of antibodies were known to the person skilled in the art which recognise more than one allele of MICA (see, inter alia, document D3, "DISCUSSION", first paragraph, and document D9, disclosure of BAMO1). This fact was not contested by the appellant either. In fact, the appellant argued that such antibodies had merely not been disclosed for the treatment of cancer.
13. The board considers, therefore, that the skilled person, having knowledge of the existence in the art of anti-MICA antibodies which are reactive to different allelic forms of MICA, would also consider using anti-MICA antibodies with a broad binding spectrum to allelic forms of MICA for treating cancers. In fact, it is the binding character of the antibody to the very allele expressed by a given cancer in need of treatment that determines the therapeutic success of the cancer treatment and not the antibody's multi-allelic binding character. This conclusion is plainly based on the knowledge of the skilled person and, therefore, does not constitute the result of an ex post facto analysis. The board, accordingly, judges that the multi-allelic binding characteristic of the antibody referred to in the claim, as such, does not attribute inventive character to the claimed subject-matter. Therefore, the provision of a MICA multi-allele-recognising antibody for improving cancer treatment was obvious to the skilled person.
14. As to the four particular alleles to which the claimed antibody is said to bind, document D9 discloses the mouse IgG1 monoclonal antibody BAMO1 having a binding specificity for the very same four MICA alleles *001, *004, *007 and *008 as referred to in the claim. The board is satisfied that, accordingly, document D9 teaches that modelling suitable antibodies for cancer therapy on the tetra-allelic binding characteristics of BAMO1 provides an antibody that can reasonably be expected to allow treatment of cancers expressing one or more of the alleles *001, *004, *007 and *008. The skilled person, when formulating a solution to the objective technical problem, would thus in an obvious manner consider using an antibody suitable for the envisaged therapeutic use which is capable of binding to these four MICA alleles.
15. The appellant argued that the skilled person had no motivation to choose the reagent of document D9 as a solution to the technical problem. It argued that the document D9 - a product sheet disclosing information on the mouse monoclonal antibody BAMO1 - did not disclose BAMO1 for the same purpose as the antibodies disclosed in document D1 or referred to in the claim. Instead, it merely indicated applications for in vitro use, i.e. flow cytometry, ELISA and immunoblot. Document D1 further stated that the mAB was for research use only and not for use in diagnostic or therapeutic procedures.
16. However, the fact that document D9 does not disclose the antibody for cancer therapy but merely for in vitro applications is fully comprehensible for the skilled person. In fact, it is part of common general knowledge that the administration of a murine antibody to humans carries huge immunological health risks. In fact, the conclusion reached by the board in point 12, above, takes into account such known risks of antibodies which are not suitable for administration to human patients.
17. The appellant's further argument was that the binding characteristics of BAMO1 are such that the antibody as well as a chimeric derivative of BAMO1, which was usable in humans, were not effective for inducing ADCC as an in vivo therapeutic agent for the treatment of MICA*007-expressing cancer (which was argued with reference to Exhibits 1 to 3). Reference is made in this connection to point 3, above, in which the board concluded that the claim does not require the claimed antibody to be suitable for the treatment of all cancers expressing a particular one of the alleles indicated in the claim. Accordingly, the board is not persuaded by the argument that these binding characteristics would have discouraged or dissuaded the skilled person from formulating antibodies capable of binding to the four MICA alleles indicated in the claim for use in the treatment of cancer, at least such cancer expressing the other indicated MICA alleles.
18. The same applies to the appellant's argument, based on the disclosures in document D3, document D5 (pages 830 and 831), document D6 (Table 2 on page 71) and document D7 (Table 3), that the skilled person was aware that immuno-dominant epitopes were not present on each of the MICA *001, *004, *007 and *008 alleles and that this dissuaded the skilled person from developing a single antibody directed against several alleles of MICA for a therapeutic purpose as claimed. As noted above in points 3 and 16, the claim in fact does not require the suitability of the antibody for the treatment of cancer expressing at least one of the indicated MICA alleles. Furthermore, document D9 discloses a murine antibody which the board considers to be appropriate for modelling suitable antibodies for cancer therapy as claimed (see point 13). The argument accordingly fails to support the inventive character of the claimed subject-matter.
19. Finally, the appellant also submitted that figure 11 of the application, showing results of testing in a mouse long-term RAJI-MICA*01 tumour model in example 11, demonstrated that the antibodies referred to in the claim were superior to antibodies only recognising two MICA alleles. The board refers anew to the conclusion in point 3, above: the claim does not require that the claimed antibody, although binding to the MICA *001 allele, is necessarily required to be effective in treating cancers expressing this particular allele. In fact, in this context therefore, the board is not persuaded that the appellant's argument, namely that particular antibodies recognising the four MICA alleles indicated in the claim are superior to other antibodies only recognising two MICA alleles in the context of the treatment of MICA*001-expressing cancer, is relevant for supporting the inventive character of the claimed subject-matter.
20. In view of the above considerations, the board considers that the subject-matter of claim 1 of both the main request and auxiliary request 1 does not involve an inventive step.
For these reasons it is decided that:
The appeal is dismissed.