T 0351/01 (Tissue Factor Protein/GENENTECH) 02-07-2003

I. European patent No 0 278 776 with the title "Methods and deoxyribonucleic acid for the preparation of tissue factor protein" was granted with 26 claims based on European patent application No.88 301 190.0, claiming priority from US 13743 of 12 February 1987, from US 35409 of 7 April 1987 and from US 152698 of 5 February 1988.

Granted claims 1, 11 and 20 read as follows:

"1. A polynucleotide as defined in Figure 2, or a variant thereof, encoding a biologically active tissue factor protein or a biologically active variant or fragment of the said tissue factor protein wherein at least one amino acid has been selectively inserted, deleted or substituted."

"11. A biologically active tissue factor protein capable of being encoded by a polynucleotide as defined in claim 1 or a biologically active variant or fragment of said tissue factor protein which either lacks glycosylation, or has non-mammalian glycosylation."

"20. A nucleotide sequence encoding a biologically active tissue factor protein or variant or fragment as defined in any one of Claims 11 to 17."

Claims 2 and 3 defined further features of the polynucleotide of claim 1. Claims 4 to 8 were directed to methods for producing the tissue factor protein (TFP) encoded by the polynucleotide of claims 1 or 2. Claims 9 and 10, 12. to 19 were directed to various biologically active TFPs encoded by a polynucleotide as defined in claim 1. Claim 21 related to an expression vector comprising a nucleotide sequence according to any one of claims 1 to 3. or 20; claims 22 to 25 respectively related to cells comprising this vector and to a method for producing TFP from these cells. Claim 26 was directed to a method for obtaining a polynucleotide encoding a biologically active TFP.

II.Two notices of opposition were filed. Opponents 1 withdrew their opposition when the case was pending before the Opposition Division. By a decision within the meaning of Article 102(3) EPC dated 29 January 2001, the Opposition Division maintained the patent on the basis of the auxiliary request then on file.

Claims 1, 9, 11, 12 and 20 of the said auxiliary request read as follows:

"1. A polynucleotide as defined in Figure 2, or a variant thereof, encoding a biologically active tissue factor protein or a biologically active variant or fragment of the said tissue factor protein which induces coagulation wherein at least one amino acid has been selectively inserted, deleted or substituted." (emphasis added by the Board)

"9. A biologically active tissue factor protein or a biologically active variant or fragment of said tissue factor protein which induces coagulation obtainable by the method of claim 4."

"11. A biologically active tissue factor protein capable of being encoded by a polynucleotide as defined in claim 1 or a biologically active variant or fragment of said tissue factor protein which induces coagulation which either lacks glycosylation, or has non-mammalian glycosylation."

"12. A biologically active tissue factor protein variant which induces coagulation capable of being encoded by a polynucleotide as defined in claim 1 wherein the transmembrane domain of native tissue factor protein is not present."

"20. A nucleotide sequence encoding a biologically active tissue factor protein or variant or fragment as defined in any one of Claims 11 to 17."

The expression "which induces coagulation" was also inserted in claims 2, 4, 6, 10, 13, 16 to 18 as a feature of the biologically active TFP, TFP variant or fragment.

III. The Appellants (Opponents) lodged an appeal against the decision of the Opposition Division, paid the appeal fee and filed a statement of grounds of appeal.

IV. Submissions in response to the appeal were filed by the Respondents (Patentees).

V. The Board sent a communication pursuant to Article 11(2) of the Rules of Procedure of the Boards of Appeal, conveying its preliminary non-binding opinion on some of the issues to be decided.

VI. Both parties answered the communication. With their letter dated 2 June 2003, the Respondents filed auxiliary requests 1 to 5 for consideration by the Board.

VII. At oral proceedings which took place on 2 July 2003, the Respondents made the auxiliary request 5 their auxiliary request 1, the other requests being renumbered accordingly. Claim 1 of said auxiliary request 5 read as follows:

"1. A polynucleotide encoding a biologically active tissue factor protein whose sequence is given in Figure 2 or a biologically active variant or fragment of the said tissue factor protein which induces coagulation wherein at least one amino acid has been selectively inserted, deleted or substituted." (emphasis added by the Board)

The other claims remained identical to the claims of the request accepted by the Opposition Division (section II, supra).

VIII. The following documents are referred to in this decision:

(N1): Bach, R., documents in relation to the oral presentation at the 7th National Conference on Thrombosis and Hemostasis, The American Heart Association's 59th scientific sessions, November 19, 1986;

(N2): Morrissey, J. et al., abstract Nr. 1632 from the American Heart Association's 59th Scientific sessions. Circulation Supplement, Part 2, Vol. 74, No. 4, October 1986;

(N4): Fisher, K.L. et al., Thrombosis Research, Vol. 48, pages 89 to 99, 1987;

(N15): Broze, G.J. et al., The Journal of Biological Chemistry, Vol. 260, No. 20, pages 10917 to 10920, 1985;

(N16): Guha, A. et al., Proc.Natl.Acad.Sci.USA, Vol. 83, pages 299 to 302, January 1986;

(N29): Giercksky, K-E., Scand.J.Haematol., Vol.19, pages 385 to 395, 1977;

(N32): Paborsky, L.R. et al., The Journal of Biological Chemistry, Vol. 266, No. 2, pages 21911 to 21916, 1991;

(N38): EP-A-0 266 993;

(N41): Old, R.W. et al., "Principles of Gene Manipulation", N.G. Carr, J.L. Ingraham and S.C. Rittenberg Eds, Blackwell Scientific Publications, Third Edition, pages 38 to 44, 1985;

(N46): Neuenschwander, P.F. et al., The Journal of Biological Chemistry, Vol. 267, No. 20, pages 14477 to 14482, 1992;

(N47): Wildgoose et al., Blood, Vol. 80, No. 1, pages 25 to 28, 1992;

(N49): Declaration by Dr J.H. Morrissey dated 30 May 2003.

IX. The Appellants' arguments in writing and during oral proceedings can be summarized as follows:

Main request; claims accepted by the Opposition Division

Article 123(2)(3) EPC in relation to claim 12

In the application as filed, the general part of the description taught that the term "tissue factor protein" covered molecules which had coagulation activity and others which had not (pages 12 to 18). Nothing was said about the activity of a TFP deleted for the transmembrane domain (page 16, lines 11 to 18). Thus, the combination in claim 12 of the two features: "wherein the transmembrane domain...is not present" and "which induces coagulation" constituted hitherto undisclosed subject-matter. Moreover, this combination conferred to the claimed subject-matter a scope of protection different from that conferred by granted claim 12. The requirements of Article 123(2)(3) EPC were, thus, not fulfilled.

Article 84 EPC; clarity of claim 1

Claim 1 was unclear because it did not specify which assay was to be used to measure coagulating activity whereas the two assays proposed in the description, namely, the chromogenic assay and the one stage clotting assay, could give different results. For example, the TFP variant lacking the transmembrane domain would not exhibit coagulating activity in the first assay whereas some activity would be seen in the second assay. This could be explained by the fact that Factor VIIa which was needed for coagulation to occur was not present in the chromogenic assay whereas it was present to some extent in the plasma used to measure coagulating activity by the one stage clotting assay. The relevant documents to illustrate this point were documents (N32) and (N49).

Article 83 EPC in relation to Article 84 EPC

The skilled person would be unable to carry out the assays measuring coagulating activity on the basis of the information given in the patent in suit, which was unclear and, therefore, he/she could not obtain the claimed TFPs without undue burden: for the chromogenic assay to be repeatable, it would be necessary that the molecular amounts of each of the reagents be disclosed; one could not repeat the one stage clotting assay as described since the origin of the plasma (eg from patients suffering from hemophilia A or B) was not adequately identified. In both cases, the incubation time was not mentioned, which was an essential feature of the assay (document (N46)).

Article 83 EPC; sufficiency of disclosure

- Document (N32) provided evidence that RFP deleted in the transmembrane domain had no coagulating activity when tested by the same chromogenic assay as described in the patent in suit. This demonstrated that the subject-matter of claim 12 could not be reproduced.

- The claims encompassed a myriad of compounds, yet no teachings were made available in the specification as to which TFP regions should be kept in order to retain coagulating activity. Very few examples were shown which were not even qualitatively satisfactory since they concerned, in particular, the whole TFP or a variant deleted in the cytoplasmic domain, a region which was of no interest. The scope of the claim was not commensurate with the contribution to the art.

- The present case was alike to that dealt with in decision T 923/92 (OJ EPO 1996, 564) where a claim considered to relate to a vast catalog of tPA derivatives of unspecified structure and vaguely defined function was refused. The then competent Board decided that the examples and information given in the patent were not sufficient to allow the skilled person to perform the invention without undue burden over the whole area claimed.

Articles 87 and 88 EPC: priority rights, Article 54 EPC: novelty of claim 1

The sequence shown in Figure 2 of the priority documents I and II differed from the sequence shown in Figure 2 of the patent in suit by five nucleotides. To assess priority, it was irrelevant that these nucleotides were not in the TFP coding region because, on the one hand, the claimed polynucleotide was larger than the coding region and, on the other, in accordance with the Enlarged Board of Appeal's opinion G 2/98 (point 9, OJ EPO 2001, 413), it was inappropriate and prejudicial to a proper exercise of priority rights to make a distinction between technical features related to the function and effect of the invention and technical features which were not.

The claimed polynucleotide did not enjoy priority rights from the filing dates of priority documents I or II. Document (N4), the scientific article disclosing the claimed invention was, thus, part of the prior art and destroyed the novelty of, in particular, the subject- matter of claim 1.

Auxiliary request 1

Article 123(3) EPC; claim 1

The scope of claim 1 was larger than that of granted claim 20 when dependent on claim 11. Indeed, in contrast to claim 1, granted claim 20 did not comprise, for example, the polynucleotide encoding full-length natural TFP since it related to a nucleotide sequence encoding unglycosylated TFP, ie to a nucleotide sequence which did not comprise the codons encoding the amino acids which were the sites where glycosylation occurred.

Articles 87 and 88 EPC; priority rights, Article 83 EPC; sufficiency of disclosure

The subject-matter of claim 1 did not enjoy priority rights from the dates of filing of priority documents I or II nor was it sufficiently disclosed for the same reasons as given for claim 1 of the main request because the added feature " whose sequence is given in Figure 2" related to the polynucleotide and not to the tissue factor protein.

Article 54 EPC; novelty: claims 9 and 11

The subject-matter of claims 9 and 11 was not novel over the teachings of document (N2), which enabled a 100.000 fold purification of TFP and its subsequent deglycosylation. Document (N2) was enabling as it mentioned Factor VII-affinity chromatography as the key step in the purification method, the other steps being directly derivable from the teachings of documents (N15) or (N16).

It also lacked novelty under Article 54(3)(4) EPC over the teachings of document (N38). Moreover, the human TFP described in documents (N15) and (N16) fell within the scope of claim 11.

Article 56 EPC; inventive step: claim 1

There were three prior art documents which described a protein identified as TFP or portions thereof, characterized by partial amino acid sequences: (N15), (N16) and (N1). It was not disputed that the sequences shown in document (N1) had been presented by Dr Bach at the 59th scientific sessions of the 7th National Conference on Thrombosis and Hemostasis before the first priority date. Since the oral disclosure of Dr Bach represented the latest development, it would be considered by the skilled person as the most reliable teaching: it was the closest prior art.

Starting from document (N1) which disclosed 55% of the TFP protein sequence, the problem to be solved could be defined as obtaining the full length protein.

Based on the partial amino acid sequences and using common general knowledge such as that found in document (N41) or mentioned in document (N49), it would have been a matter of routine work in 1987 to clone the TFP encoding cDNA. In fact, no special difficulties were mentioned either in Example I of the patent in suit or in the corresponding scientific publication, document (N4):

- It could be seen in Example 1 that the cDNA library was prepared according to a known technique.

- It was known that tissue factor could be found in all tissues and, besides, the Respondents' choice of adipose tissue was obvious in view of the teachings of document (N29).

- The scarcity of the mRNA was no problem as millions of clones could be screened with the probe known from document (N1). Its large size was also of no consequence since libraries were commercially available at the priority date which contained cDNAs of an average length of 2000 to 3000 bases. If only part of the cDNA was obtained, one could as a matter of routine, use this part as a primer to isolate the rest.

- The situation was quite different from that encountered in case T 223/96 of 29 January 1999 relied upon by the Respondents. There, inventive step was acknowledged since there existed real cloning difficulties, in particular, only one tissue could be the source of the relevant mRNA. In the present case, no such difficulties existed, thus inventive step had to be denied.

X. The Respondents' arguments in writing and during oral proceedings can be summarized as follows:

Main request; claims accepted by the Opposition Division

Article 123(2)(3) EPC in relation to claim 12

A basis for the subject-matter of this claim could be found in the application as filed on page 12, lines 5 to 7 where it was stated that variants which demonstrated tissue factor protein activity were comprised within the scope of the claim. Amongst these variants were those lacking the transmembrane domain (page 16). In contrast, derivatives which were not coagulating were only mentioned for the first time on page 18. The requirements of Article 123(2) EPC were, thus, fulfilled.

Article 84 EPC; clarity of claim 1

Claim 1 was clear although not mentioning the assay used to measure coagulation. It was mere speculation in the Appellants' statement that the two assays described in the patent in suit would give different results. They both relied on the same biological effect and both depended on Factor VIIa being present. In the chromogenic assay, the presence of Factor VIIa was insured by the presence of Factor VII and Factor IXa, Factor IXa being able to convert Factor VII into Factor VIIa (document (N47)). The plasma used in the one stage clotting assay could contain different amounts of Factor VIIa, yet, it did not mean that coagulation would not be seen but only that, as shown in document (N46),(Figure 2, page 14479), a little more incubation time was necessary in order to see it when the levels of Factor VIIa were low.

Article 83 EPC in relation to Article 84 EPC

In order to establish which amounts of Factor IXa and X were needed to carry out the chromogenic assays, it would suffice to establish a standard curve. Document (N38) showed that it was perfectly possible to perform the assay starting from a commercially available mixture of the two factors. The one stage clotting assay was described in document (N47) as exquisitely sensitive to trace amounts of Factor VIIa, which meant that it could be made to work irrespective of the origin of the plasma chosen to carry it out with. The average incubation time necessary for coagulation to take place was 120 seconds ie. much less than the incubation times contemplated in the patent in suit.

Article 83 EPC; sufficiency of disclosure

- Document (N32) did not provide evidence that TFP deleted in the transmembrane domain did not show coagulating activity when tested by the chromogenic assay as described in the patent in suit because it did not make use of this assay but of an alternative assay where Factor IXa was not present, ie where Factor VIIa could not have been produced from Factor VII.

- Claim 1 was not an overly broad claim. It would be within the skilled person's ability to obtain the correct variants with a reasonable amount of trial and errors.

- The present case was alike to that dealt with in decision T 923/92 (see supra) where sufficiency of disclosure was recognized to derivatives of tPA as long as adequate information was given on how to produce human tPA and the functions to be tested were clearly indicated. In fact, the situation was even better than in this earlier case since several examples of how to produce TFP variants were provided and at least two such variants were shown to be functional in the chromogenic assay.

Articles 87 and 88 EPC; priority rights; claim 1

The formulation of claim 1 as being directed to "a polynucleotide as defined in Figure 2" implied that said polynucleotide needed not have the whole sequence shown in claim 2. On the contrary, the polynucleotide was that which encoded TFP ie, the coding portion of the sequence of Figure 2. This interpretation of the claim was consistent with the thrust of the disclosure. On page 2, the invention was defined as "the coding portion". In the legend to Figure 2, no emphasis was put on the non-coding part of the sequence. In claim 12 as originally filed, the invention was defined as the DNA sequence which encoded tissue factor protein.

The skilled person wanting to reproduce the invention would never attempt to obtain the DNA of Figure 2 as a whole but only the coding region. In addition, it had to be taken into account that the five differences between the sequences shown in Figure 2 of priority documents I and II and Figure 2 of the patent were irrelevant for the subject-matter of the invention. It would be going too far to apply the opinion G 2/98 (see supra) so strictly that irrelevant differences would jeopardize the right to priority.

Auxiliary request 1

Article 123(3) EPC; claim 1

The subject-matter of claim 1 did not go beyond that of granted claim 20 when dependent on granted claim 11. Granted claim 11 comprised, in particular, full length unglycosylated TFP (or full length TFP carrying non- mammalian glycosylation) as being produced in hosts incapable of glycosylation. Granted claim 20 related to a nucleotide sequence encoding the TFP of granted claim 11, which sequence was the same as that of the polynucleotide which encoded full length natural TFP as comprised in claim 1. Thus, granted claim 20 had the same scope as claim 1. The requirements of Article 123(3) EPC were satisfied.

Articles 87 and 88 EPC; priority rights, Article 83 EPC; sufficiency of disclosure

There was no doubt that the expression "whose sequence is given in Figure 2" in claim 1 was meant to apply to the TFP protein and not to the polynucleotide encoding it. The TFP amino acid sequence disclosed in the priority documents I and II was the same as that disclosed in Figure 2 of the patent in suit. Said documents also disclosed polynucleotides encoding said amino acid sequence. Priority rights could be acknowledged from the filing date of the first priority document.

Sufficiency of disclosure could be acknowledged for the same reasons as given in relation to the main request.

Article 54 EPC; novelty: claims 9 and 11

While reporting the purification and subsequent deglycosylation of natural TFP, document (N2) was not enabling insofar as the purification was concerned, as it only briefly mentioned one of the steps in the process. Document (N38) was also not enabling with regard to the production of human TFP by recombinant means as it merely mentioned this possibility on page 5, lines 31 to 35 without providing any further technical details. Documents (N15) and (N16) disclosed purified preparations of proteins which had partial amino acid sequences different from that of TFP and which were glycosylated. None of these documents destroyed the novelty of the subject-matter of claims 9 or 11.

Article 56 EPC; claim 1

Document (N1) was the report of an oral disclosure of partial TFP amino acid sequences which took place at the 59th Scientific Sessions of the Conference on Thrombosis and Hemostasis. It did not make the disclosed sequences available to the public in the way required by the case law. In particular, evidence from some members of the audience was missing, which would have been necessary to ascertain what had really been said.

Even if the disclosure in document (N1) was taken at its face value, there was no reasons why the person skilled in the art wanting to clone the TFP cDNA would have isolated a probe from the amino acid sequences given therein. Seeing the discrepancies between said sequences and those disclosed in documents (N15) or (N16), the skilled person would rather have chosen to re-isolate the natural protein and partially sequence it, which was a very difficult task.

Starting with the amino acid sequences shown in document (N1), the cloning of TFP cDNA involved many difficulties which could not be solved without inventive skills.

- Brain tissue and placenta were the tissues from which TFP was commonly isolated . The TFPs from both these sources were known to have different N-terminal sequences (document (N2)) and this would leave the skilled person uncertain as to which tissue was to be chosen as an mRNA source. The Respondents had shown that attempts at cloning the cDNA from placenta mRNA would not succeed. To obtain it, they had to start from adipose tissue which, although having been identified as a source of TFP in a document dating from 1977 (document (N29)), was not an obvious source of the factor.

- The mRNA was in very low abundancy and of a high size (2350bp). Reverse transcription of the full length molecule would, thus, not be expected to occur, the average length of cDNA obtainable by the then available methods being of about 600 bp.

- This cloning was at least as difficult as the one performed in the patent on appeal under the number T 223/96 (see supra) where at least a partial cDNA sequence had already been disclosed.

XI. The Appellants requested that the decision under appeal be set aside and that the European patent No. 0 278 776 be revoked.

As main request, the Respondents requested that the appeal be dismissed. As auxiliary request 1, the Respondents requested that the decision under appeal be set aside and that the patent be maintained with the claims of auxiliary request 5 filed on 3 June 2003. As auxiliary requests 2 to 5, the Respondents requested that the patent be maintained with the claims of auxiliary requests 1 to 4 filed on 3 June 2003, taken in their numerical order.