Federal Court of Australia

Sanofi v Amgen Inc. [2023] FCA 264

ORDERS

THE COURT ORDERS THAT:

1.    The appellant be granted leave pursuant to r 34.50(2)(b) of the Federal Court Rules 2011 (Cth) to rely on the experiments referred to in para 9(1) of Mr Muratore’s affidavit dated 1 March 2023 conducted by Professor van der Merwe, as reported in his declaration, on the condition that:

(a)    the appellant provide to the respondent copies of all documents constituting or recording any instructions provided to Professor van der Merwe in relation to the experiments; and

(b)    any evidence in chief to be given by Professor van der Merwe in relation to his experiments be by affidavit in admissible form and which complies with any relevant Rule of Court or Practice Note concerning the giving of expert evidence.

2.    The appellant be granted leave pursuant to r 34.50(2)(b) of the Federal Court Rules 2011 (Cth) to rely on the experiments referred to in para 14 of Mr Muratore’s affidavit dated 1 March 2023 conducted by Dr Frenzel, as reported in his declaration dated 6 July 2019, on the condition that:

(a)    the appellant provide to the respondent copies of the raw data from the experiments;

(b)    the appellant provide to the respondent copies of any experimental protocol prepared by Dr Frenzel and used by him for the purpose of conducting the experiments;

(c)    the appellant provide to the respondent copies of all documents constituting or recording any instructions provided to Dr Frenzel in relation to the experiments; and

(d)    any evidence in chief to be given by Dr Frenzel in relation to his experiments be by affidavit in admissible form and which complies with any relevant Rule of Court or Practice Note concerning the giving of expert evidence.

3.    The interlocutory application be otherwise dismissed.

4.    The proceeding be listed for hearing:

(a)    for the taking of evidence commencing at 10.15am on 6 November 2023 for 10 days; and

(b)    for submissions commencing at 10.15am on 11 December 2023 for 3 days.

5.    The respondent file and serve any evidence in reply by 4.00pm, 29 September 2023.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

NICHOLAS J:

Background

1    Before me is an interlocutory application filed by the appellant (“Sanofi”) in an appeal brought by Sanofi pursuant to s 60(4) of the Patents Act 1990 (Cth) (“the Act”) from a decision of a delegate of the Commissioner of Patents (“the Delegate”) directing that five applications for standard patents filed by the respondent (“Amgen”) proceed to grant. By its interlocutory application, Sanofi seeks orders for discovery and leave to rely on experimental proof.

2    The five oppositions were heard by the Delegate in June 2019. Senior Counsel for Sanofi, Mr Shavin KC, also appeared before the Delegate as did Senior Counsel for Amgen, Mr Dimitriadis SC. The Delegate’s decision is dated 26 September 2022.

3    Under the relevant transitional provisions, the oppositions to the applications are to be determined by reference to the provisions of the Act before amendment by the Intellectual Property Laws Amendment (Raising the Bar) Act 2012 (Cth) (“RTB Act”). It is common ground that the test to be applied for the purpose of determining whether or not the applications should proceed to grant is whether Sanofi has established that any claim, if it were the subject of a grant, would be clearly invalid or whether, as the test is sometimes expressed, it would be practically certain that any such claim would be invalid.

4    An appeal under s 60(4) is a hearing de novo in the original jurisdiction of the Court. The onus of proving facts which are relied on to make out a ground of opposition is on the opponent. The rules of evidence, including, in particular, relevant provisions of the Evidence Act 1995 (Cth), apply at the hearing of the appeal. Sections 37M-37P of the Federal Court of Australia Act 1976 (Cth) also apply to the proceeding. Section 37M(1) provides:

The overarching purpose of the civil practice and procedure provisions is to facilitate the just resolution of disputes:

(a)    according to law; and

(b)    as quickly, inexpensively and efficiently as possible.

5    There is a significant public interest in ensuring that patents that are invalid are not granted. But there is also a significant public interest in ensuring that opposition proceedings are not used improperly to delay the grant of a patent. These considerations were referred to by the Full Court in Genetics Institute Inc v Kirin-Amgen Inc (1999) 92 FCR 106 (Black CJ, Merkel and Goldberg JJ). The Court also stated at [19] “… the purpose of pre-grant opposition proceedings is to provide a swift and economical means of settling disputes that would otherwise need to be dealt with by the courts in more expensive and time-consuming post-grant litigation; that is, to decrease the occasion for costly revocation proceedings by ensuring that bad patents do not proceed to grant …”.

6    Each of the patent applications is a divisional application of a parent application filed in 2008. Each is entitled “Antigen Binding Proteins to Proprotein Convertase Subtilisin Kexin Type 9” (“PCSK9”). For convenience, I will refer to the five patent applications as the 677, the 685, the 689, the 748 and the 751 applications.

7    The 677 application has 42 claims including two independent claims. The 185, 689 and 748 application have 30 claims, with only one independent claim. The 751 application has 42 claims including four independent claims.

8    The Delegate noted that the specification for the five applications were essentially the same save for consistory statements and some minor differences in text and figures. She used the 677 application as the exemplar specification.

9    Claim 1 of the 677 application (an example of a “binding claim”) is to:

An isolated monoclonal antibody that binds an epitope on hPCSK9 comprising one or more of amino acid residues 207, 208, 162, 164, 167, or 123 of SEQ ID NO: 1, and wherein the monoclonal antibody blocks binding of PCSK9 to LDLR.

10    Claim 1 of the 748 application (an example of a “competition claim”) is to:

An isolated monoclonal antibody that binds to human PCSK9 and is neutralizing in that an excess of said antibody reduces the quantity of human PCSK9 bound to LDLR in an in vitro competitive binding assay, wherein said monoclonal antibody competes for binding to PCSK9 with an antibody that comprises: a heavy chain variable region of the amino acid sequence in SEQ ID NO: 49; and a light chain variable region of the amino acid sequence in SEQ ID NO: 23.

11    The evidence relied on by Sanofi at the hearing of the interlocutory application included affidavit evidence from its solicitors, Mr Anthony Muratore and Ms Lisa Taliadoros, both of whom are experienced practitioners well versed in the conduct of patent proceedings in this Court involving pharmaceutical patents. Affidavit evidence for Amgen was given by its solicitor, Mr Gary Cox, who is also an experienced practitioner who has acted in patent proceedings in this Court including those involving pharmaceutical patents.

12    In her first affidavit, Ms Taliadoros identifies the issues in the proceeding and the matters relevant to Sanofi’s oppositions. According to para 11 of her first affidavit, the various claims fall into two broad categories:

(a)    Epitope/Residue Claims (677, 689 and 751 Applications), being claims to an isolated monoclonal antibody that:

(i)    either binds to at least one of the nominated residues of PCSK9, or binds (or “recognizes”) an epitope comprising one or more nominated residues of PCSK9; and

(ii)    either “blocks” binding of PCSK9 to LDLR, or “reduces” binding between PCSK9 and (an) EGFa domain of LDLR.

(b)    Competition Claims (685 and 748 Applications), being claims to an isolated monoclonal antibody that:

(i)    binds to human PCSK9;

(ii)    is “neutralizing” in that an excess of said antibody “reduces” the quantity of human PCSK9 bound to LDLR in an in vitro competitive binding assay, or “reduces” binding between human PCSK9 and an EGFa domain of LDLR; and

(iii)    “competes” for binding to PCSK9 with a nominated ‘reference antibody’.

13    Sanofi alleges that each of the claims of Amgen’s applications is invalid on the basis that the invention as claimed is not a manner of manufacture, lacks fair basis, is not fully described (including best method), is not sufficiently defined, lacks clarity, lacks utility, lacks novelty, and does not involve an inventive step.

14    Sanofi contends that even if the patent applications disclose two monoclonal antibodies (21B12 and 31H4) that binds to an epitope of PCSK9 and blocks or reduces binding between the antigen PCSK9 and LDLR, the claims of each of the applications has been drafted far more widely than is legally permissible. Sanofi contends that the claims are of such breadth that they claim, in effect, a monopoly over any antibody that happens to interact with PCSK9, regardless of the antibody’s amino acid sequence or 3D structure, and irrespective of whether the claimed antibody has the desired biological outcome of lowering plasma LDL levels (ie. “bad cholesterol”). Sanofi also contends that, through the use of very broad definitions for terms including “neutralize” and “compete” in the 685 and 748 applications, Amgen has sought to extend the scope of its claims to antibodies that “modulate” the activity of PCSK9 in any way whatsoever (in the case of “neutralize”) and to antibodies that occupy any physical space that is proximal or even remote to the spaces to be occupied by either of the two antibodies described in the patent applications (in the case of “competes”).

15    In para 13 of her first affidavit, Ms Taliadoros identifies 14 matters relevant to Sanofi’s opposition all of which appear to have been drawn from written submissions relied on by Sanofi before the Delegate. I will refer to four of those matters (a, d, f and k) by way of example:

a.    The claims are of such a breadth that they claim in effect a monopoly over any antibody that happens to interact with PCSK9, regardless of the antibody's amino acid sequence or 3D structure and irrespective of whether the claimed antibody has the desired biological outcome of lowering plasma LDL levels …

d.    Having an amino acid sequence of an antibody in hand does not of itself enable one to know (or predict) the 3D structure of the antibody, let alone the antigen to which it will bind (and where on the antigen it will bind and with what affinity). Equally, the identification of a useful function for a (hypothetical) antibody does not enable one to know (or predict) the amino acid sequence of such an antibody …

f.    Knowledge of the structure the 3D structure of the antigen (even knowledge of the epitopes or residues on PCSK9 to which the antibody binds) does not assist the skilled reader in identifying the amino acid sequence of that antibody or in obtaining an antibody that will have the necessary concomitant structure. Nor is there any correlation between where an antibody binds and its amino acid sequence. In addition, a change made to a single amino acid, for example in a CDR, may so severely impact binding that it may be lost altogether because of the impact on 3D configuration …

k.    There is no reasonable scientific basis to extrapolate from 21B12 and 31H4 to the wider classes of antibodies being the subject matter claimed …

16    Each of the matters referred to in para 13 of Ms Taliadoros’ first affidavit concerns the type of issue or question that is almost invariably the subject of independent expert evidence. This was the approach taken by Sanofi in the proceeding before the Delegate whose reasons for decision identify the declarations relied on by Sanofi’s own experts in the fields of monoclonal antibodies and structural biology.

DISCOVERY

17    By its interlocutory application, Sanofi seeks an order requiring Amgen to provide it with four categories of documents as set out in what is referred to as the Further Revised Sanofi Notice (“the Sanofi Notice”). The categories of documents are identified in Pt A of the Sanofi Notice.

18    The first of the orders sought by Sanofi is not expressed in terms of an order for discovery, but there can be no doubt that it is discovery that Sanofi seeks. It is an application in relation to which Pt 20 of the Federal Court Rules 2011 (Cth) (“FCR”) applies. Rule 20.11 of the FCR provides that:

A party must not apply for an order for discovery unless the making of the order sought will facilitate the just resolution of the proceeding as quickly, inexpensively and efficiently as possible.

19    Rule 20.14 of the FCR is concerned with standard discovery. Rule 20.15 of the FCR is concerned with “non-standard and more extensive discovery”. Rule 20.15(3) of the FCR requires a party seeking more extensive discovery than is required under r 20.14 of the FCR to file a supporting affidavit stating why the order should be made. Here there is a supporting affidavit made by Ms Taliadoros. According to her first affidavit, the documents sought by Sanofi will assist the Court in the resolution of the various grounds of opposition to which she refers.

20    In Rinehart v Rinehart (No 2) [2015] FCA 339 Gleeson J referred to the principles relevant to discovery. Her Honour said at [36]-[39]:

[36]    The Court will not order discovery as a matter of course unless discovery is necessary for the determination of issues in the proceeding: Alliance Craton Explorer Pty Ltd v Quasar Resources Pty Ltd (No 4) [2013] FCA 1044 at [33]; Construction, Forestry, Mining & Energy Union v Rio Tinto Coal Australia Pty Ltd [2014] FCA 462 at [92].

[37]    In Trade Practices Commission v CC (New South Wales) Pty Ltd (No 4) (1995) 58 FCR 426 at 436 to 437, Lindgren J considered the meaning of the word “necessary” in the context of Order 15 r 15 of the former Federal Court Rules. Although the requirement of “necessity” is not explicitly retained in the current rules, the citations above make plain that the concept has continued relevance in justifying an application for discovery. In particular, it directs attention to the potential evidentiary difficulties of a party taking into account the interests of the party to whom the order is to be directed. Lindgren J concluded that, where one party and not the other was likely to have documents relating to a matter in question, it was prima facie “necessary” that discovery be ordered.

[38]    This conclusion was subject to the qualification that the Court will not grant an application for discovery where to do so would permit “fishing”, that is, to seek evidence to support a claim which is essentially speculative in nature.

[39]    Even if the Court is persuaded to make an order for discovery, the Court will fashion the order to suit the particular circumstances of the case: Taylor v Saloniklis [2013] FCA 679 at [7].

21    I also refer to the observations of Burchett J in F Hoffman-La Roche v Chiron Corporation (2000) 47 IPR 516 when considering an application for discovery in a patent revocation proceeding. His Honour referred to the policy of the rules that “discovery should only be ordered to the extent that it is necessary for the attainment of the ends of justice”.

Category 1

22    Ms Taliadoros has annexed to her first affidavit some extracts from the transcript of examinations in the US proceeding of three of the 11 named inventors. The questions and answers appear to be focused on the role of binning experiments and binning data in determining whether antibodies will bind to residues on PCSK9. Binning experiments, also referred to as competition assays, are used to determine whether antibodies compete with each other and, if competition assays show that antibodies are competitive with each other, may indicate that they are binding to the same area on the antigen.

23    It is not suggested by Ms Taliadoros that production of any of the documents is necessary to enable Sanofi to establish any of its grounds of opposition. Rather, she states in para 20 of her first affidavit, that based on the extracts from the depositions annexed to her affidavit, it is her opinion that “… the inventors have provided evidence that will assist the Court to make an assessment as to the matters identified in paragraph 13 [of her affidavit] in so far as they concern the grounds of lack of manner of manufacture, lack of fair basis, and lack of definition and are therefore relevant to the determination of these grounds”. In his oral submissions, Mr Shavin KC expressed the point slightly differently. He submitted that there was a “good possibility” that the documents would assist Sanofi’s case.

24    Based on the limited information contained within the extracts of the depositions exhibited to Ms Taliadoros’ first affidavit, it is simply not possible to say whether any statements made by any of the inventors either in depositions, declarations or oral evidence in other proceedings is likely to assist the Court in resolving any of the issues to which they are said to be relevant.

25    Questions concerning the scope of the relevant claims, whether they are for a manner of manufacture, whether they lack fair basis, or whether they suffer from a lack of definition or clarity, are all matters that will be determined based on a reading of the relevant patent applications through the eyes of the notional skilled addressee equipped with the common general knowledge.

26    Mr Shavin KC submitted that the documents would be relevant to manner of manufacture, fair basis, and what is said to be a failure of the claims to adequately define the invention.

27    As to manner of manufacture, the question is whether the requirements of s 18(1)(a) of the Act are satisfied. This depends on the character of the invention disclosed and claimed in the complete specification: D’Arcy v Myriad Genetics Inc (2015) 258 CLR 334. The same is also true of fair basis. Whether or not the claims are fair based ultimately depends on the application of the principles considered by the High Court in Lockwood Security Products Pty Ltd v Doric Products Pty Ltd (2014) 217 CLR 274.

28    Sanofi’s contention that the claims do not sufficiently define the scope of the invention focuses on what is said to be the functional nature of the claims which do not identify the physical structure or composition of the relevant antibody and which extend to any antibody that (using claim 1 of the 677 application as an example) binds to the PCSK9 antigen to block binding of PCSK9 to LDLR. Sanofi’s argument is that claims of this nature, at least in the case of monoclonal antibodies, do not sufficiently define the invention.

29    I think it unlikely that the inventors’ evidence would be of any assistance in determining whether the monoclonal antibodies claimed are a manner of manufacture, fairly based or whether they fail to sufficiently define the invention described.

Category 2

30    The category 2 documents are said by Sanofi to be directed to manner of manufacture, lack of fair basis, lack of definition and lack of utility. It is not suggested that they are relevant to any other ground of opposition.

31    The fact that the documents referred to in category 2 may have been produced in the US proceeding is not of itself a basis for requiring them to be produced in this proceeding given significant differences in adjectival and substantive law. In particular, the law in the United States relating to the written description and enablement requirements, which appears to have been the main focus of the US proceedings, is different from Australian law at least as the Act stood prior to the RTB Act amendments taking effect: Kimberly Clark Australia Pty Ltd v Arico Trading International Pty Ltd (2001) 207 CLR 1 at [25]; cf. Amgen Inc v Sanofi, United States Court of Appeals for the Federal Circuit, February 11, 2021 at pp 10-11.

32    Category 2 seeks the results of testing or analysis by Amgen on various antibodies identified in Table 2 of the complete specification which are said to bind with or neutralise PCSK9 so as to prevent it binding with LDLR. Although Ms Taliadoros’ first affidavit identifies the grounds of opposition to which Sanofi says the category 2 documents are relevant, it does not identify with any particularity what it is they are likely to prove.

33    With regard to utility, Sanofi’s case is that the claims are not limited by the promised result (ie. a reduction in plasma LDL levels), and will necessarily include monoclonal antibodies that, even though binding to a specified epitope, and blocking binding of PCKS9 to LDLR (the LDL receptor), do not achieve that result. However, there is nothing in the material before me to indicate that the category 2 documents are likely to include evidence of embodiments of the invention that do not achieve a promised result.

Category 3

34    As to category 3, the documents within this category are quite limited in their number. They relate to what was referred to as Amgen’s “catabolic research program”.

35    Amgen contends that the category 3 documents relate to a post-priority date research program that is not relevant to any issue in the opposition. In addressing Ms Taliadoros’ evidence on this topic, Mr Cox said in his affidavit at para 44:

Ms. Taliadoros asserts in paragraph 34 of her affidavit that the category 3 documents are relevant because they “relate to Amgen’s attempts to generate antibodies binding to a specific region of PCSK9, the so called ‘sweet spot’.” She bases this on her review of Sanofi’s US appellate brief, not on a review of the underlying documents or facts. However, I understand that the US Court, which had the benefit of the underlying documents and facts, considered Sanofi’s argument and came to a different conclusion. In particular, the US Court acknowledged that certain catabolic antibody program documents, including category 3 documents, could have been relevant if in fact Amgen was actually “trying to make other antibodies within the scope of the patent” (e.g., antibodies that bind the sweet spot). Sanofi was given the opportunity at trial, through its cross examination of one of the inventors, Dr Jackson, to show that Amgen was in fact “trying to make other antibodies within the scope of the patent.” But Sanofi “did not make this showing,” so the US Court “continued to exclude these documents.” (see Annexure GBC-2 at page 26).

Ms Taliadoros’ response to this evidence in her affidavit in reply does not dispute the correctness of what is said by Mr Cox.

36    Sanofi sought to rely on the category 3 documents (or some of them) in the US proceedings to show that Amgen continued to look for antibodies within the claims of the 677 application and its equivalent for more than four years after the priority date but never found them. However, the record of the US proceedings suggests that the documents in question did not show that this is what occurred. It is apparent that Judge Andrews, the US District Judge who heard the case, who was presumably familiar with the documents in question, doubted that they had any relevance to any issue in the case but that, even if they had some marginal relevance, they should be excluded from evidence.

Category 4

37    Pursuant to category 4, Sanofi seeks all documents created by or on behalf of Amgen that refer to the category 3 documents. What I have said concerning category 3 applies also to the category 4 documents.

38    Having regard to the nature of this proceeding, the issues to be determined, and s 37M of the Federal Court of Australia Act 1976 (Cth), I would not be disposed to order the discovery sought by Sanofi unless I was persuaded that the documents were directly relevant to an issue to be determined and likely to assist in the resolution of that issue. In a pre-RTB Act appeal under s 60(4) bought by an opponent who contends that a grant would clearly be invalid, it will be a rare case in which discovery will be necessary for the just resolution of the proceeding. In this case I am not persuaded that the documents in issue are either directly relevant to any matter in issue or that their production is necessary for the just resolution of the proceeding.

Experiments

39    By its interlocutory application, Sanofi also seeks leave to rely on experiments. The experiments in respect of which leave is sought are identified in Pt B, paras 1-3 of the Sanofi Notice.

40    Rule 34.50 of the FCR provides:

34.50    Experimental proof as evidence

(1)    If a party (the proponent) proposes to tender, as evidence in a proceeding, experimental proof of a fact, the proponent must apply for orders in relation to the experimental proof, including orders about any of the following:

(a)    the service on other parties of particulars of the experiment and of each fact that the proponent asserts is, will or may be proved by the experiment;

(b)    any persons who must be permitted to attend the conduct of the experiment;

(c)    the time when, and the place where, the experiment must be conducted;

(d)    the means by which the conduct and results of the experiment must be recorded;

(e)    the time by which any other party (the opponent) must notify the proponent of any grounds on which the opponent will contend that the experiment does not prove a fact that the proponent asserts is, will or may be proved by the experiment.

(2)    Evidence of the conduct and results of the experiment is admissible in the proceeding, only:

(a)    if the proponent has complied with subrule (1) and any orders given under that subrule; or

(b)    with the leave of the Court.

(3)    If an order mentioned in paragraph (1)(e) has been made, and the opponent has not complied with the order in relation to a ground, the opponent may rely on the ground only with the leave of the Court.

Professor Anton van der Merwe’s experiments

41    Sanofi seeks leave under r 34.50(2)(b) of the FCR to rely on experiments conducted by Professor Anton van der Merwe which are relied on by Sanofi in proceedings in Israel and experiments conducted by Dr Andrew Frenzel which are relied on by Sanofi in EPO proceedings. It is common ground that r 34.50(1) of the FCR was not complied with. The relevant experiments were conducted by experts retained by Sanofi or Regeneron for the purpose of proceedings in other jurisdictions and prior to the commencement of this proceeding.

42    Experimental evidence can give rise to expensive and time consuming debates between expert witnesses and has the potential to create a “trial within a trial” as to the appropriateness of the methodology used, the conditions of the material and the facility used to conduct the experiment, the accuracy of any observations or measurements made, and the interpretation of the results. It is incumbent on Sanofi to show that the experiments in respect of which it seeks leave are relevant to an issue in the proceeding in the sense that they could rationally affect (directly or indirectly) the assessment of the probability of the existence of a fact in issue in the proceeding: see s 55 of the Evidence Act 1995 (Cth). It would not be consistent with the overarching purpose to grant Sanofi leave to rely on experiments in the circumstances unless Sanofi can establish the relevance of those experiments to those matters which it says they will be relied on to prove.

43    Sanofi seeks leave to rely on Professor van der Merwe’s experiments and declaration as proof that Eli Lilly’s antibody LY 3015014 produced by Absolute Antibody and another antibody known as REGN2625 produced by Regeneron (together, “the test antibodies”) were tested for their ability to compete with Amgen’s reference antibody 21B12 and that the order in which the antibodies were added to the assay can impact the outcome of the competition assay. Further, Sanofi also seek to rely on Professor van der Merwe’s experiments as proof that competition observed was due to steric hindrance rather than the test antibodies binding to the same residues on PCSK9 to which the reference antibody 21B12 binds.

44    Professor Yoram Reiter, the author of a declaration dated 21 August 2021 filed by Amgen in the Israeli proceedings, has commented on Professor van der Merwe’s experiments and declaration in some detail. Professor Reiter’s declaration does not criticise Professor van der Merwe’s methodology. Indeed, Professor Reiter appears to accept Professor van der Merwe’s methods and conclusions, except in so far as Professor van der Merwe suggests that Eli Lilly’s antibody LY 3015014 inhibits PCKS9 binding to LDLR. Professor Reiter states that Professor van der Merwe did not perform experiments to ascertain whether the test antibodies neutralise the binding of PCSK9 to LDLR and that neither the Eli Lilly patent nor the article by Schroeder et al relied on by Professor van der Merwe provides the data that would permit that conclusion to be drawn. On that point, Professor Reiter is contradicted in a later declaration prepared by another of Sanofi’s experts, Professor Michael Eck, who refers to data in both the Eli Lilly patent and Schroeder article which he says supports Professor van der Merwe’s conclusion that the Eli Lilly antibody is neutralising.

45    One difficulty is that each of the Eli Lilly patent and Schroeder article reports on experimental work in respect of which Sanofi would need to seek and obtain leave if it were to rely on it to support Professor van der Merwe’s opinion that the Eli Lilly antibody inhibits PCKS9 binding to LDLR. No such leave has been sought. To be clear, I would not allow Sanofi to rely on Professor van der Merwe’s experiments as evidence that the Eli Lilly antibody LY 3015014 was within any of the relevant claims including, in particular, claim 1 of the 748 application.

46    However, I will grant leave to Sanofi to rely on Professor van der Merwe’s experiments to show that Eli Lilly’s antibody LY 3015014 produced by Absolute Antibody, when tested for its ability to compete with antibody 21B12, was seen to be dependant on the order in which the test and reference antibodies were added to the assay, and that competition observed was due to steric hindrance rather than the test antibodies binding to the same residues on PCSK9 to which the reference antibody 21B12 binds. Those are the only matters Mr Muratore says Professor van der Merwe’s experiments will be relied on to prove. In my view they are directly relevant to Sanofi’s case based on lack of clarity and what it calls “lack of definition”.

47    Mr Dimitriadis SC submitted that there is no admissible evidence to show how the two test antibodies were produced. This did not appear to raise any concern for Professor Reiter who responded to Professor van der Merwe’s declaration in some detail and who did not consider it surprising that the order of addition of the test and reference antibodies can influence the outcome of the competition experiments.

48    I would need to be persuaded that there was a genuine dispute as to the composition or provenance of the test antibodies if I was to refuse Sanofi leave to rely on Professor van der Merwe’s experiments for the reason advanced by Mr Dimitriadis SC. Based on the material before me, which includes the appendices to the declaration, and Mr Cox’s evidence at para 69 and 70 of his affidavit, there does not appear to be a genuine dispute in relation to those matters. Of course, that is not to say that Sanofi is not required to adduce admissible evidence to prove those matters at the hearing.

49    Leave will be granted to Sanofi to rely on Professor van der Merwe’s experiments referred to in para 9(1) of Mr Muratore’s first affidavit subject to any proper objection as to the admissibility of any evidence relating to such experiments. Any evidence in chief to be given by Professor van der Merwe in relation to his experiments must be by affidavit in admissible form and comply with any relevant Rule of Court or Practice Note concerning the giving of expert evidence.

Dr Andre Frenzel’s experiments

50    Sanofi also seeks leave to rely on two sets of experiments conducted by Dr Andre Frenzel, the first of which is the subject of a declaration dated 6 July 2019 (“Frenzel 1”), and the second of which is the subject of a declaration dated 22 October 2019 (“Frenzel 2”). The evidence before me also includes a declaration of Dr Lutz Riechmann dated 7 July 2019 (“Riechmann 1”) and another declaration by him dated 22 October 2019 (“Riechmann 2”) (both filed and relied on by Sanofi in the EPO proceedings). There was another declaration of Dr Riechmann that was not included in the evidence in this application. The evidence also includes a declaration made by Dr Anthony Rees dated 21 February 2020 (filed and relied on by Amgen in the EPO proceedings) responding to Riechmann 2 and Frenzel 2.

51    Frenzel 1 indicates that Dr Frenzel was requested by Regeneron’s counsel to test various antibodies binding to hPCSK9 for their ability to compete with two reference antibodies 21B12 and 31H4. He was also asked by Regeneron’s counsel to test the ability of those antibodies to decrease the binding of PCSK9 to LDLR. Regeneron is a co-defendant in the US proceedings and also an opponent in the EPO proceedings.

52    Dr Frenzel does not express any conclusion in Frenzel 1. In fact, Frenzel 1 does not appear to contain anything in the nature of a conclusion. An exhibit to his declaration (B1) records tests results, all of which appear to be directed to determining whether or not any of the antibodies tested was within claim 1 of EP2215124. However, Dr Riechmann in Riechmann 2 points to some of those results as suggesting that “competition can occur through a number of mechanisms” and “that competition with [the reference antibodies] 21B12 and/or 31H4 provides no prediction that an antibody will neutralize the hPCSK9-LDLR interaction”. Dr Riechmann goes on to say that:

… the property of competition with 21B12/31H4 is simply not predictive as to whether any given antibody would have useful therapeutic properties, since it in no way predicts any mechanism or level of neutralization, and certainly not the mechanism and level of neutralization of 21B12 and 31H4.

53    In para 4 of Mr Muratore’s first affidavit, he expresses the opinion that Dr Frenzel’s experiments will assist the Court in making an assessment as to the following matters:

(a)    whether competition assays, in which an antibody (the competing antibody) is tested for its ability to bind to an antigen when a reference antibody is bound informs as to the binding site of the competing antibody; and

(b)    having regard to the fact that there are different known competition assays, whether the term “competes” as used in the claims of the 748 and 685 applications, lacks clarity.

54    I accept that Riechmann 1 suggests that Dr Frenzel’s first set of experiments reported in Frenzel 1 may provide evidence to support the contention that competition between antibodies may occur through a number of mechanisms and that competition with reference antibodies 21B12 or 31H4 provides no prediction that an antibody will neutralise the hPCSK9-LDLR interaction.

55    In para 61 of Mr Cox’s affidavit, he explains why he does not consider that any of Dr Frenzel’s experiment results reported in Frenzel 1 will assist the Court in addressing the two issues identified by Mr Muratore. Mr Cox says that Dr Frenzel’s first set of experiments reported in Frenzel 1 was not properly controlled and the data was not properly analysed, and that Amgen presented expert evidence to this effect in proceedings in Japan. Various other criticisms are made including that Sanofi has not provided to Amgen Dr Frenzel’s raw data. However, what is apparent is that Amgen has previously addressed Dr Frenzel’s experimental evidence in at least one other jurisdiction. Amgen’s own evidence on the interlocutory application would suggest that Amgen is already in a position to address Dr Frenzel’s first set of experiments referred to in Frenzel 1 in expert evidence of its own.

56    Leave will be granted to Sanofi to rely on Professor Frenzel’s set of experiments referred to in para 14 of Mr Muratore’s first affidavit on the condition that:

(a)    Sanofi provide to Amgen copies of the raw data from the experiments;

(b)    Sanofi provide to Amgen copies of any experimental protocol prepared by Dr Frenzel and used by him for the purpose of conducting the experiments;

(c)    Sanofi provide to Amgen copies of all documents constituting or recording any instructions provided to Dr Frenzel in relation to the experiments; and

(d)    any evidence in chief to be given by Dr Frenzel in relation to his experiments be by affidavit in admissible form and which complies with any relevant Rule of Court or Practice Note concerning the giving of expert evidence.

57    With regard to Dr Frenzel’s second set of experiments referred to in Frenzel 2, Mr Muratore states in his first affidavit that:

22.    Based on my reading of [Frenzel 2], Dr Frenzel discusses experiments conducted on antibodies which were obtained by screening of an antibody-phage library comprising antibody sequences disclosed in libraries designated HAL4/HAL7, which based on the evidence of Dr Frenzel were available prior to the claimed priority date of the Applications (23 August 2007)) (the Phage Display Antibodies).

23.    Based on my reading of [Riechmann 2], Dr Riechmann discusses the data obtained from experiments discussed in [Frenzel 2], in support of the scientific proposition that at least 2 of the antibodies tested by Dr Frenzel, were found to compete with both 21B12 and 31H4, in an ELISA competition assay conducted in accordance with the protocol set out in Dr Frenzel's declaration and reduced the binding of LDLR to PCSK9 by at least 40% at a concentration of 10 ug/ml.

24.    In this proceeding, Sanofi seeks to rely on the experiments reported in [Frenzel 2] as proving the propositions set out in paragraphs 22 and 23 above.

58    Mr Muratore’s evidence provides very little information as to the ground of opposition to which Frenzel 2 is directed. At para 27 of his first affidavit, Mr Muratore states that Dr Frenzel’s experiments will assist the Court in determining the grounds of lack of manner of manufacture, lack of definition and lack of inventive step. I am not persuaded that Dr Frenzel’s second set of experiments could be of any assistance in the determination of the first and second of those grounds. In any event, Dr Frenzel’s first set of experiments and Professor van der Merwe’s experiments will presumably be relied on by Sanofi in support of them.

59    As to inventive step, Mr Muratore refers to Riechmann 2 which appears to indicate that in the EPO proceeding Sanofi was relying on Frenzel 2 and Dr Riechmann’s evidence to show that a skilled person in 2007, following the teaching of a Novartis patent (WO 2008/125623), and using standard techniques, protocols and widely available reagents, would have been able to produce human antibodies that neutralise hPCSK9-LDLR binding and compete with 21B12 or 31H4 for hPCSK9 binding.

60    No case of lack of inventive step was relied on before the Delegate. I understand this ground of opposition was abandoned by Sanofi either during or shortly prior to the hearing before the Delegate. The notice of appeal raises allegations of lack of inventive step but does not refer to the availability of either of the relevant antibodies or the antibody library from which they are said to have been obtained. Nor does the notice of appeal refer to the Novartis patent in support of any allegation that the claims do not involve an inventive step. The Novartis patent is relied on solely in support of a lack of novelty contention.

61    Even if Dr Frenzel’s second set of experiments is relevant to Sanofi’s lack of novelty contention, there are in my view good reasons why Sanofi should not be granted the leave it seeks.

62    It is apparent from the declaration of Dr Rees filed by Amgen in the EPO proceedings that there is a significant debate as to what conclusions may be drawn from Dr Frenzel’s second set of experiments. Dr Rees is critical of both Dr Frenzel’s experimental protocol and data analysis. He says that it is not possible to conclude from Dr Frenzel’s results that the relevant antibodies compete with antibodies 21B12 or 31H4. He also draws attention to the instructions that were provided to Dr Frenzel by Sanofi’s counsel who, according to Dr Frenzel’s declaration, “asked him to use HAL4 and HAL7 antibody gene libraries for the isolation of antibodies specifically binding to a peptide derived from hPCSK9 … to test these antibodies for their ability to compete with two reference antibodies [ie. 21B12 and 31H4]”. Dr Rees says that these antibody libraries do not appear to have been publicly available as of 22 August 2008. Further, Dr Rees says that these antibody libraries would not have come to mind when the ordinary scientist was working in the field. It is not apparent from anything in Frenzel 2 why Dr Frenzel was instructed to use those particular antibody libraries. While it is not my function to rule on any of those criticisms for the purpose of determining the present application, I do infer that there is a significant debate as to the appropriateness of Dr Frenzel’s starting point and the suitability of his experimental protocol and data analysis.

63    It was suggested on behalf of Sanofi that it was open to Amgen to repeat Dr Frenzel’s second set of experiments and, presumably, make any modifications to the experimental protocol and data analysis that Dr Rees considered appropriate. According to Sanofi, Amgen decided not to take up the opportunity to conduct its own set of experiments in response to Frenzel 2.

64    If there were to be another set of experiments conducted aimed at demonstrating that the Novartis patent provides clear directions to the notional skilled addressee to make one or more of the relevant antibodies (ie. antibodies within the claims), then that is a matter that should in my view be the subject of an experiment conducted in accordance with a protocol either agreed between the parties or, if not agreed, then as ordered by the Court. Frenzel 2 suggests that experiments of the type Dr Frenzel conducted would take a scientist working in the area of antibody analytics about three weeks to perform.

65    In my view, permitting Sanofi to rely on Dr Frenzel’s second set of experiments is likely to give rise to a substantial and undue waste of time and costs that is better avoided by requiring the relevant experiments to be carried out in accordance with an agreed or court ordered protocol that is sensitive to the criticisms made by Dr Rees and the novelty test applicable under Australian law: see Bristol Myers Squibb Co v F H Faulding & Co Ltd (2000) 97 FCR 527 at [67]; AstraZeneca AB v Apotex Pty Ltd (2014) 226 FCR 324 at [293]. In the circumstances, and having regard to the nature of this proceeding, and the overarching purpose, I am not persuaded it is appropriate to grant Sanofi the leave it seeks in relation to Dr Frenzel’s second set of experiments referred to in para 21 of Mr Muratore’s first affidavit.

66    Orders accordingly.

Associate:

Dated:    24 March 2023