T 2416/18 (Cross-neutralsing antibodies/POMONA RICERCA) 14-09-2021

1. The appeal complies with Articles 106 to 108 and

Rule 99 EPC and is admissible.

Main request - claim 1

Claimed invention - claim construction

2. The claim relates to a human monoclonal antibody (mAb) directed against influenza A virus haemagglutinin antigen capable of binding to and neutralising at least one influenza A virus subtype containing

haemagglutinin H1 and at least one influenza A virus subtype containing haemagglutinin H3. In other words, the claim is directed to a human mAb functionally characterised in that it displays a heterosubtype cross-neutralising activity for influenza A virus subtypes H1 and H3. It is undisputed that the claimed antibody has to recognise an epitope that is shared between haemagglutinins H1 and H3.

Sufficiency of disclosure (Article 83 EPC)

3. Article 83 EPC requires that the application disclose the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art. In the case at hand, to carry out the claimed invention, the skilled person must be able, on the basis of the disclosure in the application and of common general knowledge, to obtain the claimed human mAbs without undue burden (see Case Law of the Boards of Appeal of the European Patent Office,

9th edition 2019, (CLBA), section II.C.4.1).

4. The parties referred in their submissions to the patent and not the application as filed. The board ascertained that there is no difference in disclosure in the relevant passages between the application and the patent and, therefore, saw no need to correct the parties' references.

5. The board agrees with the appellant that the skilled person in the case at hand consists of a team of an immunologist, a virologist and a biotechnologist expert in antibody production. This was not contested by the respondents.

Teaching of the application

6. The present invention lies in the field of human mAbs directed against the haemagglutinin antigen of influenza A virus. Within influenza A viruses, subtypes are distinguished based on the antigenic features of two viral surface proteins, haemagglutinin and neuraminidase. Subtypes that have affected humans in the course of recent history are subtypes H1N1 and H3N2 (see also paragraph [0009] of the patent). Immunity against influenza A virus subtypes is distinguished as follows: (i) homologous immunity relating to the individual isolate, e.g. to isolate A/PR/8/34 of subtype H1N1; (ii) homosubtype immunity relating to isolates belonging to the same subtype, e.g. to isolate A/PR/8/34 and isolate A/SC/1918, both of subtype H1N1; and (iii) heterosubtype immunity relating to isolates belonging to different subtypes, e.g. to isolate A/PR/8/34 of subtype H1N1 and to isolate A/PC/1/73 of subtype H3N2 (see also paragraph [0012] of the patent). After an infection or vaccination, homologous or homosubtype immunity is seen in humans, but heterosubtype immunity to influenza A viruses is "extremely rare both in case of natural infection and in case of vaccination" (see paragraph [0012] of the patent) and achieving mAbs with such properties has so far proved to be "extremely difficult" (see paragraph [0021] of the patent).

7. The patent discloses antigen binding fragment 28 (Fab28), which has a heterosubtype cross-neutralising activity against the reference isolates of virus A/PR/8/34 (H1N1) and virus A/PC/1/73 (H3N2) (see paragraphs [0067] and [0068] and Figures 1 and 2) and thus possesses the functional properties defined in claim 1. The patent thus discloses one way of carrying out the claimed invention. This is not disputed.

8. At issue is whether the patent, when considered alone or in combination with common general knowledge, provides guidance which allows the skilled person to obtain substantially all the embodiments falling within the ambit of the claim without undue burden.

9. The patent proposes obtaining the claimed human mAbs by way of a process comprising the following steps:

(i) selection of individuals for the generation of human mAbs by following the inclusion criteria defined in paragraph [0055] of the patent, (ii) vaccination of the individuals selected accordingly and (iii) cloning of monoclonal cross-reactive anti-influenza antibodies from their EBV-transformed peripheral blood B lymphocytes (see experimental section,

paragraphs [0055] ff).

10. Furthermore, the patent states that "in particular, it is described that some individuals, despite continuous exposure to influenza virus (sometimes for professional reasons, as physicians, pediatricians, people working in kindergartens and schools), do not contract the disease. These rare individuals were thought to be less susceptible to influenza virus infection due to the development, for still unknown reasons, of an effective heterosubtype immunity. For this reason they were thought to be the best candidates for the generation of human mAbs" (see paragraph [0055]).

11. In examining the issue of sufficiency of disclosure in the case at hand, a question of particular relevance is whether or not it can be accepted that the patent, as submitted by the appellant, describes selection criteria that allow for the reliable identification of individuals who are highly likely to produce the antibody of claim 1.

12. The following inclusion criteria are disclosed in the patent for the selection of these individuals:

"(1)- between 25 and 55 years of age;

(2)- recent pathological medical history, for the ten years preceding the study, negative for clinical influenza syndromes;

(3)- antibody titer higher than 1:1000 against virus isolates, subtypes HlNl and H3N2 responsible for the annual epidemics during the five years preceding the study;

(4)- high neutralizing titer (IC50 >= 1:400) against virus isolates, subtypes HlNl and H3N2 responsible for the annual epidemics during the five years preceding the study;

(5)- detectable neutralizing titer (IC50 >=l:20) against two reference subtype A virus isolates

(A/PR/8/34 subtype HlNl; A/PC/1/73 subtype H3N2);

(6)- no prior anti-influenza vaccination;

(7)- compliance to receive anti-influenza vaccination"

(see paragraph [0055] of the patent; numbering (1)

to (7) has been added by the board for ease of reference).

13. Although the patent provides no further information on the purpose of the various inclusion criteria, it can be inferred from the wording of criteria (1), (2) and (6) that these criteria serve to select individuals who, despite not having been vaccinated against influenza virus, had not contracted the disease in the past ten years, referred to in the following as the "healthy donor" concept (see also point 10 above).

14. As evident from point 10 above, the healthy donor concept reflected in selection criteria (1), (2)

and (6) is a mere hypothesis and it is unknown whether individuals fulfilling these criteria do indeed produce an effective heterosubtype immunity.

15. As for the remaining selection criteria, criteria (3) and (4) relate to serological binding and neutralisation tests against virus isolates of subtypes HlNl and H3N2 responsible for the annual epidemics in the recent past, while criterion (5) relates to serological neutralisation tests against older reference virus isolates of subtypes HlNl and H3N2. These tests are performed with sera, i.e. the portion of serum remaining after coagulation of blood. It is undisputed that sera comprise a heterologous mixture of antibodies directed at various different epitopes, also referred to as polyclonal antibodies.

16. While the purpose of the tests of criteria (3) to (5) is not immediately apparent, what is evident is that none of the tests assesses for the presence of cross-neutralising antibodies recognising a shared epitope of haemagglutinins H1 and H3. Indeed, since these tests are performed with a mixture of polyclonal antibodies they cannot distinguish between individuals producing antibodies binding to H1 haemagglutinin or H3 haemagglutinin (homosubtype immunity) and individuals producing antibodies binding to an epitope shared by both haemagglutinins (heterosubtype immunity), as binding of the mixture of polyclonal antibodies will be detected in both cases. Analogous considerations apply to the neutralisation tests. Furthermore, since the tests are performed on intact viruses, the tests are unsuitable for distinguishing between antibodies binding to haemagglutinin and antibodies binding to neuraminidase.

17. The board therefore agrees with respondent II that the tests of selection criteria (3) to (5) provide a read-out for individuals who have been exposed to the tested virus isolates and produce antibodies against these isolates. Being polyclonal, these antibodies are most likely separate antibodies recognising different epitopes. Accordingly, producing a strong polyclonal antibody response cannot be indicative of the presence of cross-reactive mAbs recognising a shared haemagglutinin epitope.

18. A connection between selection criteria (3) to (5) and the presence of cross-neutralising antibodies recognising a shared epitope of haemagglutinins H1 and H3 is thus not apparent. Accordingly, the appellant's assertion that criteria (3) to (5) ensure the identification of individuals who are likely to produce antibodies capable of recognising at least one antigenic site that was not subject to changes, i.e. a conserved part representing a shared epitope of haemagglutinins H1 and H3, is not considered persuasive.

19. It is common ground that criterion (7) does not contribute to the selection of individuals having a higher probability of producing heterosubtype immunity. This criterion need not be considered further.

20. In view of the above consideration of selection criteria (1) to (6), the board concurs with

respondent II that there is no scientifically credible reason to accept that the selection method of paragraph [0055] of the patent ensures the selection of individuals having a higher probability of producing heterosubtype immunity.

21. While the patent states that heterosubtype immunity is "extremely rare", see paragraph [0012], it does not reveal, for example, how many individuals were screened to identify Fab28. As argued by respondent II, the patent provides no comparative data showing that the selection criteria set out in paragraph [0055] increase the chances of finding a cross-neutralising antibody versus mere "brute force" screening of a large number of individuals. In the absence of such comparative data, the provision of Fab28 (see point 7 above) cannot serve as evidence that the criteria set out in paragraph [0055] of the patent allow for the reliable identification of individuals who are highly likely to produce the antibody of claim 1.

22. Moreover, the board concurs with respondent II that document D35, a scientific article relating to the characterisation of PN-SIA28 (= Fab28, the sole antibody isolated in the patent), provides evidence that the serological criteria in paragraph [0055] of the patent are arbitrary. Thus, while criterion (4) in paragraph [0055] of the patent selects for neutralisation of modern H3N2 isolates, H3N2 isolates circulating after 1975 are in fact not neutralised by Fab28 (see page 2, right-hand column, first paragraph, and Table 1), showing a disconnect between the criteria set out in paragraph [0055] of the patent and the properties of the one antibody disclosed in the patent.

23. In view of the above considerations, the board does not accept that the patent describes in paragraph [0055] selection criteria that allow for the reliable identification of individuals who are highly likely to produce the antibody of claim 1.

Post-published evidence

24. The requirements set out in Article 83 EPC must be met on the filing date of the application. This principle applies to any claim request filed in opposition appeal proceedings on the basis of which maintenance of the patent in an amended form is requested. Against this background, post-published documents may be used as evidence that the disclosure is reproducible without undue burden only under certain circumstances (see

also CLBA, section II.C.6.8).

25. In the case at hand, the question of whether or not post-published evidence can be taken into account in assessing the sufficiency of disclosure need not be answered. Indeed, consideration of the documents relied on by the appellant would not lead to a different assessment of the board because none of the documents shows that the selection criteria in paragraph [0055] of the patent ensure the reliable identification of individuals who are highly likely to produce the antibody of claim 1. The board's reasoning in this respect is set out below (see points 26 to 28).

26. Document D40, in particular, concerns the isolation of broadly neutralising antibodies against influenza A viruses from plasma cells isolated from "selected donors who had been previously found to produce a strong heterosubtypic antibody response (9), shortly after natural infection with influenza A or vaccination" (see page 851, left-hand column, end of second paragraph). However, how the donors producing a strong heterosubtypic antibody response were selected is not explained in document D40. This is the subject of a reference document, document (9), which is not

on file.

27. Document D7, in turn, uses memory B cells isolated from recently vaccinated donors and antibody phage display to search for broadly neutralising H5N1 mAbs using combinatorial libraries (see paragraph bridging pages 2 and 3). Finally, document D29 (see page 320, left-hand column) and document D30 (see page 149, right-hand column) are silent about the healthy donor concept, and the antibody titer, neutralising titer and lack of prior anti-influenza vaccination mentioned in criteria (3), (4) and (6) of paragraph [0055] of the patent are not mentioned either.

28. In sum, the selection criteria employed for the identification of the donors of document D40 producing a strong heterosubtypic antibody response are unknown and the selection criteria employed in documents D7, D29 and D30 do not match the patient selection criteria set out in paragraph [0055] of the patent. Therefore, and contrary to the appellant's submissions, none of these documents provides confirmation of the suitability of the screening concept set out in paragraph [0055] of the patent.

Case law

29. As for the case law relied on by the appellant, it is not applicable to the facts of the present case. In contrast to the cases underlying decision T 431/96 (see Reasons, points 6, 7, 10 and 11) and decision T 877/03 (see Reasons, point 23), for example, a suitable antigen, i.e. an antigen representing an epitope shared by haemagglutinins H1 and H3, allowing the skilled person to screen for the claimed antibodies by applying routine methodology is not disclosed in the patent. Screening on the basis of different influenza strains is unsuitable for identifying individuals producing the claimed human mAbs (see points 16 and 17 above). For these reasons, the appellant's line of argument based on case law is not convincing.

Conclusion on sufficiency of disclosure (Article 83 EPC)

30. The board concludes from the above considerations with respect to the selection method provided in paragraph [0055] of the patent that this does not increase the skilled person's chances of finding a cross-neutralising antibody. Furthermore, the board agrees with respondent II that Fab28 represents a "single lucky event" and that finding other cross-neutralising antibodies by following the screening method disclosed in paragraph [0055] of the patent is at best a chance event. In the absence of any evidence that such chance events occur frequently enough and can reliably be identified by the selection criteria disclosed in the patent, and further considering that individuals producing such antibodies are acknowledged in the patent as being "extremely rare", the provision of the claimed antibodies amounts to an undue burden.

31. The opposition division had decided to reject claim 1 as granted for lack of sufficiency of disclosure. Therefore, the onus was on the patent proprietor, as the appellant who pursues a claim identical to claim 1 as granted with its main request, to present a detailed line of argument as to why that decision was not correct on the merits (see also CLBA,

section III.G.5.1.2c). Contrary to the appellant's assertion, in the case at hand the respondents were under no obligation to provide experimental evidence to support the insufficiency attack.

32. The invention claimed in claim 1 of the main request does not meet the requirements of Article 83 EPC.

Auxiliary request 1

Admittance (Article 12(4) RPBA 2007)

33. Amended claim 1 of auxiliary request 1 is based on claim 4 as granted and differs from claim 1 of the main request in that the claimed antibody is further characterised as being capable of binding the haemagglutinin conformational epitope specifically recognised by a reference antibody (see section VIII).

34. The request was filed for the first time with the statement of grounds of appeal (VI and VIII) and its admittance is thus governed by Article 12(4) RPBA 2007. Pursuant to Article 12(4) RPBA 2007, the board has discretion to hold requests filed with the statement of grounds of appeal inadmissible if they "could have been presented or were not admitted in the first instance proceedings". The appellant's line of argument that the "could have been presented" consideration did not apply to the case at hand because this consideration was only relevant under the new RPBA, as in force from 1 January 2020, therefore cannot succeed.

35. Admittance of auxiliary request 1 hinges, inter alia, on the question of whether the appellant was in a position to make its submissions earlier, and whether it could have been expected to do so under the circumstances (see CLBA, section V.A.4.11.1).

36. In response to the notice of opposition, the appellant had filed sets of claims according to a main request and according to auxiliary requests 1 to 3. In all of these claim requests, subject-matter corresponding to that of claim 4 as granted had been deleted in response to an objection under Article 83 EPC raised by the opponents (see appellant's letter dated 13 March 2018, page 1, last paragraph, to page 2, fourth paragraph). Subject-matter corresponding to that of claim 4 as granted had also been deleted in auxiliary requests 4 to 7 filed during the opposition proceedings.

37. Evidently, if the appellant had wanted to pursue the subject-matter of claim 4 as granted it should not have deleted claims directed to that subject-matter from all of its requests pending before the opposition division. The board therefore considers that the appellant could and should have presented auxiliary request 1 in the opposition proceedings.

38. Furthermore, a substantiation requirement applies to claim requests submitted in the appeal proceedings, see Article 12(2) RPBA 2007. In accordance with this requirement, when an auxiliary request is submitted the patent proprietor (or applicant) must also provide reasons as to the extent to which the objections raised in the decision under appeal are overcome by the amendments made, unless this is self-explanatory (see CLBA, V.A.4.12.5). Pursuant to Article 12(4) RPBA 2007,

claim requests filed with the statement of grounds of appeal that do not meet the substantiation requirement are not considered by the board.

39. In the case at hand, the appellant provided no reasoning in the statement of grounds of appeal as to why the amendment in auxiliary request 1 addressed the opposition division's finding of a lack of sufficiency of disclosure. Moreover, no reasoning was provided by the appellant in response to the respondent's reply or the board's communication, both objecting to the lack of substantiation of the amendments made in auxiliary request 1. Furthermore, the board concurs with respondent II that the re-introduction of the subject-matter of dependent claim 4 as granted cannot be considered a response to any aspect of the decision under appeal. Indeed, that in the present case an explanation was needed because it is not self-explanatory as to why the amendment was made has not been contested by the appellant.

40. In view of the above considerations, the board has decided to hold auxiliary request 1 inadmissible (Article 12(4) RPBA 2007).

Conclusion

41. The main request, which is the sole request in the appeal proceedings, does not meet the requirements of Article 83 EPC. Accordingly, the patent cannot be maintained in amended form on the basis of this request. Hence, the decision under appeal cannot be set aside and the appeal is to be dismissed, with the consequence that the opposition division's interlocutory decision becomes final.