T 1616/22 (FVIII production/OCTAPHARMA) 08-07-2024

Admission of documents D31 to D40

1. As documents D31 to D40, filed by the appellant, were not needed to reach the present decision, it was not necessary to decide on their admission.

Main request - claim 1

Claim interpretation

2. Claim 1 is directed to a method of increasing the productivity of recombinant factor VIII (rFVIII) characterised by:

(a) the composition of the cell culture:

- suspension of human cells

- in a culturing medium

- containing not more than 500 myM CaCl2

- at least a non-ionic detergent

- other nutrient components needed for the cells to grow and produce rFVIII

and by:

(b) the cell culture conditions:

- inducing a shear stress by mechanical means directly to the suspension of human cells by adding a power density of 50 to 1000 W/m**(3)

3. Claim 1 further states that the purpose of the claimed method is to increase the productivity of "recombinant factor VIII (rFVIII) produced in a eukaryotic cell suspension". Increasing the productivity of a protein implies that the protein is produced.

4. The boards have consistently found that the criteria set out by the Enlarged Board of Appeal in G 2/88 and G 6/88 (OJ EPO 1990, 93 and 114) can be applied only to claims directed to the use of a substance for achieving an effect and cannot be extended to claims directed to a process for producing a product characterised by process steps where the purpose of carrying out the process steps is indicated in the claim (see Case Law of the Boards of Appeal of the EPO, 10th edn. 2022, I.C.8.1.3.a)).

5. "[I]ncreasing the productivity, in particular cell-specific productivity, of recombinant factor VIII (rFVIII)" therefore imposes no limitation on the method other than it being suitable to achieve an increase in productivity. Since there is no reference defined in the claim, any minimal increase in productivity compared to any reference method fulfils this criterion.

6. It has not been contested that "the expression 'other nutrient components needed for the cells to grow and produce rFVIII' must be construed broadly, including even trivial ingredients such as 'water'" (see decision under appeal, point 2.4.3.2).

7. It is also undisputed that applying a power density of 50 to 1000 W/m**(3) (energy per volume) by mechanical means to a cell suspension in a culturing medium induces shear stress (see decision under appeal, point 2.4.6.2) as, in general, any mechanical movement in a cell culture does.

Inventive step (Article 100(a) and Article 56 EPC)

Document D1 as the starting point

8. Document D1 discloses expression of a B-domain deleted form of rFVIII (see page 358, left-hand column, last full paragraph) in human HEK 293 cells grown in suspension in Freestyle 293 Expression Medium**(TM) (Invitrogen). Shear stress is induced by mechanical means directly to the suspension of the cells by shakers (see point 2.5 on page 359). The composition of the medium is not provided in document D1, but given the successful production of FVIII, it is undisputed that the medium contains "nutrient components needed for the cells to grow and produce rFVIII" as required by the claim. It is also undisputed that Freestyle 293 Expression Medium**(TM) comprises the non-ionic detergent Pluronic (see document D12, page 9, answer to question 7).

9. For the purpose of assessing inventive step, the board assumes, in favour of the respondent, that the feature "not more than 500 myM CaCl2" is not disclosed in document D1 and thus represents a difference.

10. The parties agreed on the further difference to the disclosure of document D1 of "adding a power density of 50 W/m**(3) to 1000 W/m**(3)".

Effects and objective technical problem

11. With regard to the first difference, "not more than 500 myM CaCl2", the respondent argued that a culture medium with CaCl2 in this concentration range prevented aggregation of the cells (see paragraph [0026] of the patent). Based on the state of the art as summarised, for example in document D6, it is credible that a reduced CaCl2 concentration indeed prevents cell aggregation (see page 194, left-hand column, last paragraph: "Reduction in calcium level was designed to minimize the HEK293 clumping which was shown to be calcium-dependent (Peshwa et al. 1993; Nadeau et al. 1996)").

12. The respondent further argued that the reduced CaCl2 concentration ("not more than 500 myM") in combination with the specified power density range resulted in increased productivity of rFVIII.

13. The board does not agree because there is no evidence to support an effect of the CaCl2 concentration on this productivity. Although productivity of rFVIII is measured in the examples of the patent (see Examples 2 and 3), the CaCl2 concentration in the cell culture is not provided. The only CaCl2 concentration indicated is 30 mM, which is applied for release of rFVIII from the cells and lies almost two orders of magnitude above the upper limit of the claimed range (see Examples).

14. The respondent argued that the skilled person would have recognised that the examples were carried out using a CaCl2 concentration within the claimed range as this was also the only range mentioned in the description (see also decision under appeal, points 2.7.2.2 and 2.7.5.5).

15. Even assuming, in favour of the respondent, that this was the case, no comparison of different CaCl2 concentrations, i.e. within and above the claimed concentration range, is provided in the patent. No effect on rFVIII productivity can therefore be derived for the CaCl2 concentration range specified in the claim.

16. With regard to the second difference, "adding a power density of 50 W/m**(3) to 1000 W/m**(3)", the respondent argued that applying this measure to a culture of rFVIII expressing human cells resulted in increased rFVIII productivity, as evidenced by the examples of the patent. Hence, it was not necessary to calculate or measure shear stress, nor to show a correlation between this shear stress and the power density added to the system.

17. The board disagrees. The only mechanistic explanation provided in the patent for an effect of increased power density on rFVIII productivity is based on increased shear stress exerted on the cells by mechanical means. In paragraphs [0008] and [0009], prior-art studies which report a positive effect of shear stress on protein production of human cells are summarised. This sets the background for the claimed invention. Paragraph [0019] then states that: "The shear stress is achieved by adding a power density of 50 W/m**(3) to 1000 W/m**(3) to the cell suspension. The conditions inducing a shear stress are events, which induce mechanical movements of the cell suspension or the cells in the suspension". Paragraph [0038] points out that "increased productivity can be achieved also in large-scale cultures by increasing the shear forces and energy input by increased stirring" (see also paragraph [0039]). The relevance of "shear" or "shear forces" is further emphasised in the examples (see column 10, lines 4 to 7; column 10, lines 32 to 36) and by the wording of claim 1 (emphasis added by the board): "inducing a shear stress by mechanical means directly to the suspension of human cells by adding a power density of 50 W/m**(3) to 1000 W/m**(3)".

18. The skilled person would therefore have taken from the patent that higher rFVIII productivity results from the increased shear stress experienced by the cells in the culture.

19. It is also undisputed that power input or density (equivalent to "energy dissipation (rate) epsilon", see paragraphs [0021] and [0030] of the patent) is not directly correlated with the amount of shear stress (see also point 2.4.6.1 of the decision under appeal).

20. It was furthermore common general knowledge at the relevant date that parameters such as the volume and form of the vessel, the type of mechanical means to apply a power input (e.g. stirrer, shaker, propeller, impeller, centrifuge, sparger or hollow fibre unit) and the viscosity of the medium have an effect on the shear stress generated (see document D8, paragraph bridging pages 101 and 102).

21. As evidence for the influence of other parameters on shear stress, the appellant referred to document D8. Figure 4.13 on page 102 shows the correlation between shear stress (y-axis) and culture volume (x-axis) for different mechanical means at a constant power density. To allow comparison, the shear stress ratio and the volume ratio are plotted, i.e. both normalised to 1 (Tau/Tau0 and V/V0). This graph shows that a volume increase by a factor of 1000 for a stirred culture results in an increase of the shear stress by a factor of about 4 to 5 fold.

22. It was also common general knowledge that high shear stress results in cell damage and thus reduced productivity (see document D8, page 90, Table 4.1: "Threshold shear stress"). This is confirmed in the patent in paragraph [0021]: "The power added directly to the cell suspension to introduce shear stress should not exceed a value where the cells are destroyed". The examples of the patent also show that at higher stirring rates, the rFVIII productivity decreases (see Table 1 in Example 2).

23. In conclusion, it is not credible that at the higher end of the power density range defined in the claim an increase of the shear stress by 5 fold or more depending on the scale-up factor would be tolerated by the cells and result in a rFVIII productivity increase. It is therefore not credible that selecting a power density in the upper end of the range gives rise to the alleged increase of rFVIII productivity.

24. On the other hand, in a larger vessel of 100 litres, a power density below the lower range defined in the claim (29 W/m**(3)) is shown to be sufficient to increase rFVIII production compared to an even lower power density (6 W/m**(3)) (see Table 2 in the patent). The lower end of the range, 50 W/m**(3), is therefore not associated with an effect which would occur over the whole scope claimed, e.g. for any culture volume.

25. The respondent confirms that in the prior art "the individual cellular response to power density addition must still be assessed on a case-by-case basis" and that this depends on various parameters (see reply to the appeal, page 27). The patent, however, contains no data on the influence of any of these parameters, e.g. cell type, culture volume or mechanical means on the (cell-specific) productivity of rFVIII. Even the examples which the respondent considers to fall under the scope of the claim contain additional sources of power density, namely a sparger stone in Example 2 and a hollow fibre filter in Example 3 for which the contribution in terms of power density was not assessed.

26. During the oral proceedings, the respondent argued that it was up to the appellant to show that the effect was not achieved over the whole scope. The board agrees that it is usually up to the party that alleges a fact in opposition to provide evidence for it. However, if the patent proprietor alleges the fact that the claimed invention improves a technical effect, the burden of proof for that fact rests upon it (see also Case Law of the Boards of Appeal of the EPO, 10th edn. 2022, I.D.4.3.1). In the case at hand, moreover, the appellant has provided, inter alia, document D8 representing common general knowledge to raise serious doubts that the effect can be achieved over the whole scope claimed. It is therefore up to the respondent to dispel these doubts.

27. In the decision under appeal, the opposition division considered that the claimed method was improved because document D1 required additional factors to reduce the membrane attachment of FVIII, i.e. von Willebrand Factor, Annexin V and ortho-phospho-L-serine (see sheet 16 of the decision under appeal, first and second full paragraph).

28. The board does not agree because the claim does not exclude the addition of supplementary factors to the medium ("culturing medium containing [...] at least [...]"). The claim is also not concerned with the distribution of FVIII between the soluble and the membrane fraction but only mentions "increasing the productivity". The absence of additional factors is therefore not a difference to the disclosure of document D1 and cannot result in an additional technical effect.

29. Summing up the above considerations, the board finds that the technical effect achieved by the claimed CaCl2 concentration is the prevention of cell aggregation. However, an increase of rFVIII productivity by means of the claimed power density range in combination with the claimed CaCl2 concentration cannot be accepted as being achievable over the whole scope of the claim. As a consequence, only the technical effect of preventing cell aggregation can be taken into account in the formulation of the objective technical problem posed.

30. The objective technical problem is thus the provision of a production method for rFVIII which prevents aggregation of human cells in suspension.

Obviousness

31. Adding a power density within the range of 50 to 1000 W/m**(3) is disclosed in the state of the art for human cell suspension cultures (see e.g. documents D3, D4 or D8).

31.1 Document D3 discloses a power density of 56 W m**(-3) for the production of rFVIII in HKB-11 (hybrid of kidney and B cell) cell line (page 17, line 4). The board does not agree with the respondent's argument that the cell line in document D3 was "sticky" (see page 15, lines 15 to 16: "klebrige humane Hybrid-Zelllinie HKB-11") and thus not representative for a suspension cell culture. The claim is not limited to any particular cell line, so any disclosure in the prior art for suspension cell culture would have been considered relevant by the skilled person.

31.2 Document D4 discloses power densities of "about 50 W/m**(3), about 60 W/m**(3), about 70 W/m**(3), about 80 W/m**(3) or more" (see paragraph [0031]) for the agitation of cultures of non-anchorage dependent mammalian cells, including 293 cells (see paragraph [0048]). Document D4 is not limited to the ADAMTS13 protein used in the examples, also mentioning other proteins, e.g. FVIII (see paragraph [0050]).

31.3 Document D8 discloses in a sub-chapter on "Non-Anchorage Dependent Cells Grown in Suspension" that "cell damage on suspended hybridom cells grown in a small stirred vessel was first observed at a stirrer speed of 300 rpm [corresponding to a power density of 88 W/m**(3)]; at 580 rpm [above 400 W/m**(3)] cell growth stopped completely" (see Table 4.2 on page 97 and accompanying text) and concludes in general terms that "[i]n stirred tanks the mean energy dissipation rate should be below ~ 1,000 W m**(-3)" (see page 101, fifth bullet point).

32. Moreover, document D7, which relates to large-scale animal cell culture, discloses typical values of power densities for animal cell cultures of 0.01 to 0.15 W kg**(-1) which corresponds to 10 to 150 W/m**(3)(see page 12, right-hand column, first paragraph). It also states that with mean specific energy dissipation values of from 0.01 up to 0.24 W kg**(-1) in baffled bioreactors with different impellers, a wide range of cells could be grown equally well (see page 13, paragraph bridging columns). The respondent argued that document D7 did not relate to human cell cultures. The board disagrees because the skilled person would consider human cells to be encompassed in the term "animal cells". Even though in document D7 only non-human mammalian cells are mentioned (e.g. CHO, BHK and mouse hybridoma), the skilled person would have considered the teaching relevant for human cells.

33. The skilled person looking for alternative production methods of rFVIII would thus have found indications of suitable power densities in the disclosure of documents D3, D4, D7 and D8. There was no reason for the skilled person to assume that these power densities which are within the claimed range could not be used in the method disclosed in document D1.

34. The effect of the indicated CaCl2 concentration on cell aggregation (see point 11. above) was known from the prior art, in particular for HEK 293 cells, i.e. the cells used in the patent. Document D6 states in this regard:

"The basic formula of the medium used in this report (IS293-V from Irvine Scientific) contains a relatively low concentration of calcium ion (~108 myM) as opposed to ~2 mM in a typical DMEM medium. Reduction in calcium level was designed to minimize the HEK293 clumping which was shown to be calcium-dependent (Peshwa et al. 1993; Nadeau et al. 1996). In our experience, the calcium concentration can be as low as 10 myM without significant effect on cell growth or viability (data not shown)." (page 194, left-hand column, last paragraph)

And:

"Since higher concentrations of calcium cause increased levels of cell aggregation, which may at least in theory interfere with the virus infection, the time of calcium addition during subsequent development work was set at 2h post infection." (page 195, paragraph bridging columns)

35. The respondent argued that document D6 concerned the production of adenovirus in HEK 293 cells and would therefore not have been consulted by the skilled person concerned with rFVIII expression in human cells. The board does not agree because the state of the art summarised in document D6 provides general recommendations for the culture of HEK 293 cells and is not limited to adenovirus production (see titles of references: Peshwa et al., 1993: "Cultivation of mammalian cells as aggregates in bioreactors: effect of calcium concentration on spatial distribution of viability" and Nadeau et al., 1996: "Improvement of recombinant protein production with the human adenovirus/293S expression system using fed-batch strategies").

36. In light of the preceding considerations, the board concludes that the subject-matter of claim 1 lacks an inventive step. The ground for opposition under Article 100(a) in conjunction with Article 56 EPC thus prejudices the maintenance of the patent as granted.

Auxiliary requests 1 to 10

Admittance of auxiliary requests 1 to 10 (Article 12(4) RPBA)

37. The sets of claims of auxiliary requests 1 to 10 are identical to those of auxiliary requests 1 to 10, respectively, submitted during the opposition proceedings but not dealt with in the decision under appeal. The board considered them to have been admissibly raised and maintained.

Auxiliary request 5 - claim 1

Inventive step (Article 56 EPC)

38. The initial argument by the respondent in oral proceedings that the features rBDD-FVIII and HEK 293 cells represented differences over the disclosure of document D1 was not maintained. Indeed, these features are disclosed in document D1 (see page 358, right-hand column, section 2.2 and page 359, left-hand column, section 2.5). The subject-matter of claim 1 thus differs from the disclosure of document D1 in the limitation of the CaCl2 concentration, the use of a rotating impeller and the specified range of power densities. The board has not been presented with any evidence that a rotating impeller achieved any technical effect compared to the shakers used in document D1. The obviousness of the limited CaCl2 concentration has been addressed for the main request (see points 34. and 35. above).

39. The more restricted range of power densities defined in the claim suffers from the same problem as the broader range defined in claim 1 of the main request, namely that further parameters such as culture volume, vessel form, etc. are lacking meaning that achieving an increase in productivity over the whole claimed range is not credible (see points 16. to 25. above).

40. The objective technical problem is therefore the same as for the main request, and an obvious solution to this problem is provided in documents D8 and D7 (see points 31.3 and 32. above).

41. The subject-matter of claim 1 of auxiliary request 5 lacks an inventive step within the meaning of Article 56 EPC.

Auxiliary requests 1 to 4 and 6 to 10 - claim 1

Inventive step (Article 56 EPC)

42. In writing, the respondent referred to its submissions on inventive step of the main request (see reply, point III, pages 40 to 44). During the oral proceedings, the respondent referred to its written submissions.

43. Accordingly, the considerations for the main request or for auxiliary request 5 on inventive step equally apply to these auxiliary requests.

44. The subject-matter of claim 1 of each of auxiliary requests 1 to 4 and 6 to 10 therefore lacks an inventive step.