Function Over Form: Amgen Successfully Defends Functionally-Defined Antibody Claims Against Sanofi

Sanofi v Amgen Inc. (No 3) [2025] FCA 387

Introduction

In a lengthy judgment, Justice Nicholas of the Federal Court of Australia has dismissed Sanofi’s appeal to overturn the decision of the Delegate of the Commissioner of Patents dismissing Sanofi’s opposition to the grant of five Amgen patent applications – AU2013203677, AU2013203748, AU2013203685, AU2013203689 and AU2013203751 (the Applications). The Applications cover Amgen’s cholesterol-lowering monoclonal antibody, evolocumab (Repatha®), and potentially cover Sanofi’s competitor antibody, alirocumab (Praluent®).

The Applications were all divisional applications derived from AU2008288791 (the Parent Application) filed on 22 August 2008 as PCT/US2008/074097 and first published on 26 February 2009. Each of the Applications was entitled “Antigen Binding Proteins to Proprotein Convertase Subtilisin Kexin Type 9 (PCSK9)”, with a filing date of 11 April 2013 and an earliest asserted priority date for each of the relevant claims of 23 August 2007 (the Priority Date).

Critically, Sanofi brought its appeal under section 60 of the Patents Act 1990 (the Act) based on the provisions of the Act as they existed prior to the enactment of the Intellectual Property Laws Amendment (Raising the Bar) Act 2012. This meant that the Applications only had to meet the lower validity standards which existed prior to the Raising the Bar amendments, a key factor favouring Amgen and assisting it to achieve a successful outcome.

Background

The Applications had a common specification, which differed as between the Applications only in respect of the consistory-like clauses and some figures. The appeal was argued and determined based on agreed exemplary claims.

Each of the independent claims of the Applications was to an isolated monoclonal antibody (MAb) that binds to human PCSK9. PCSK9 is a protein involved in the degradation of low-density lipoprotein receptor (LDLR). LDLR on the surface of the liver takes up low density lipoprotein (LDL) which is then transported to lysosomes for degradation. The LDLR is then returned to the liver surface to take up more LDL. High levels of LDL cholesterol are associated with increased risk of cardiovascular and related diseases. Reduction in LDL cholesterol concentrations may therefore lead to improved health outcomes.

The exemplary claims did not describe the claimed MAbs by amino acid or nucleic acid coding sequence, or by any other physical attribute. Rather, the claimed MAbs were defined by their physico-chemical or biological results or effects in relation to PCSK9. The claims either:

• identified PCSK9 epitopes or amino acid residues to which the MAb would bind and block or reduce PCSK9’s binding to LDLR; or
• identified the claimed MAb by describing “reference” MAbs with which it would compete to bind PCSK9.

Key Issues and Consideration

Section 40: Failure to define the invention

Sanofi submitted that the invention was not properly defined as, by failing to identify the structure of the claimed MAbs, the claims did not permit others to know in advance whether a MAb proposed to be made was within the claims. Amgen argued that the definition requirement was met if a person skilled in the art produced a MAb and could then tell whether it fell within a claim or not.

Justice Nicholas held that the law did not require an invention to be defined in the claims in terms which make it possible to know whether an article would infringe in advance of it being made. The Court accepted that it is permissible to define the physical characteristics of an article by reference to the results it is to achieve, and that some experimentation may be required to determine whether the article has such characteristics.

In respect of the claims relating to PCSK9 epitope or residue binding, the specification described identifying such results by arginine scanning and X-ray crystallography. The Court held, based on expert evidence, that such techniques would allow a person skilled in the art to carry out experiments to determine without undue difficulty or invention whether an antibody made fell within the claims. The claims were held to sufficiently define the invention, despite lack of sequence information.

Section 18(1)(a): Manner of manufacture

Sanofi failed to prove that the claims did not describe a patentable manner of manufacture. Sanofi argued that the lack of a structural description of any claimed putative MAb meant that the invention claimed was no more than an abstract idea or a “mere desideratum” of any antibody that had the desired functional outcome. Sanofi argued that the claims were at the boundaries of the concept of manner of manufacture, and that the principles in Myriad should be applied to determine if the invention was a patentable manner of manufacture.¹ In particular, Sanofi said that allowing claims of this type would have an impermissible “chilling effect” on researchers, contrary to those principles.

The Court distinguished Myriad on the basis that the claims in that case involved abstract genetic information, while Amgen’s claims in suit were to physical MAbs with effects on the binding of PCSK9 to LDLR. The Myriad principles therefore were not applicable. In any event, there was no evidence of the claimed chilling effect.

Section 40(3): Fair basis

Fair basis was assessed under the pre-Raising the Bar law – i.e. that the body of the specification must contain “a real and reasonably clear disclosure” of what is claimed.

Sanofi put forward the following four arguments in relation to the claims that defined the invention by reference to the PCSK9 epitope or residues to which a claimed MAb would bind:

• First, Sanofi argued that there was no fair basis for claims that included reference to the MAb binding certain PCSK9 amino acid residues, as these claims omitted several residues identified in the specification as being part of the interface between PCSK9 and the example MAbs described. Sanofi also argued that, for some claims, no specific antibody was identified in the specification as having been generated and tested as binding to residues of those claims.

  Justice Nicholas noted that the omission of some elements described in the specification would only result in a lack of fair basis if that omission meant that the claim was to a fundamentally different invention to that described.² That was plainly not the case here. Justice Nicholas considered that the specification described a class of antibodies which block or interact with the residues identified in certain examples. His Honour held that the specification did not need to exemplify every antibody the subject of the claims in order for the claims to be fairly based, and that the specification did not require the person skilled in the art to engage in impermissible extrapolation.

• Sanofi’s second argument was that Amgen should not be able to rely on analysis by its experts to establish that an exemplified MAb formed a claimed bond with a specific PCSK9 amino acid common to a number of the exemplary claims. Even if it could, Sanofi argued that the evidence was insufficient. Justice Nicholas dismissed this argument on the basis that Amgen did not need to prove that the claimed bond formed in order for the claim to be fairly based.
• Sanofi’s third argument was that the designation of certain PCSK9 amino acids as “core” and “boundary” residues would not be understood by a person skilled in the art as defining a structural epitope, or as identifying residues that would form bonds with MAbs or LDLR. It said that the specification did not identify the epitopes for the antibodies claimed. Justice Nicholas disagreed, finding a fair basis for the relevant claims in the examples disclosed and the consistory clauses.
• Similarly to the argument on lack of definition, Sanofi’s fourth argument of lack of fair basis was that the claims did not characterise, by reference to amino acid sequences, all the antibodies that fell within the scope of certain claims across the breadth of those claims. This argument was dismissed on the same grounds as the lack of definition argument.

Sanofi put forward a more limited case of lack of fair basis in respect of the claims that defined the invention by reference to competition between the claimed MAb and reference MAbs in binding to PCSK9.  Sanofi submitted that the “suggestion that an applicant invented both the invention described in the specification, together with anything that may be later demonstrated to compete with that thing at a later point in time, must be wrong as a matter of principle”.  It further submitted that the competing antibodies must be limited to “antibodies defined by way of their unique, specific amino acid sequence in specific figures set out in specification”.  His Honour, having reviewed the specification, found clear and fairly-based disclosure of the relevant elements of these claims in the specification, such that they were not clearly invalid.

Priority Date

The patent claimed a priority date from the first of four relevant US provisional patent applications. The subsequent provisional applications added matter, including examples, to this first application. Justice Nicholas noted that, under the relevant law, a provisional specification merely had to “describe the invention” and did not need to provide an enabling disclosure as has been necessary since the Raising the Bar amendments to the Act.

Justice Nicholas accepted that the earliest priority document did not describe the class of antibodies falling within the claims made by reference to a set of residues to which the claimed MAbs would bind directly, or to residues within the epitopes to which the antibodies would bind. Instead, his Honour found that the language of that document was much more general. Nevertheless, Justice Nicholas held that the later provisional filings did not introduce any feature involving a new inventive step or otherwise depart from the more general description of the invention in the first application. Accordingly, the specification in suit followed “the same line of thought” and the first provisional application did confer priority on the claims.

Inventive Step

Justice Nicholas carefully considered the evidence as to the common general knowledge at the relevant time. His Honour found that, although the notional skilled team would have considered the possibility of generating PCSK9 inhibiting monoclonal antibodies as a potential treatment for hypercholesterolemia and related cardiovascular diseases, there were several other possibilities for inhibiting PCSK9 activity. The choice of which possibility to pursue depended to a significant degree on whether PCSK9 functioned primarily outside or inside of cells – a fact which was not then part of the common general knowledge. It was also uncertain whether it would even be possible to generate an antibody to PCSK9. Accordingly, his Honour concluded that the pathway to the generation of anti-PCSK9 antibodies was not clear or straightforward and was the subject of considerable uncertainty relating to PCSK9’s mechanism of action. Justice Nicholas was not satisfied that the notional skilled team would have been directly led to try an anti-PCSK9 antibody in the expectation that it may well succeed. The evidence therefore fell short of establishing that any of the exemplary claims would, if the applications proceeded to grant, be clearly invalid as not involving an inventive step.

Lack of Clarity

Sanofi argued that the claims to MAbs that competed with stated reference MAbs in binding to PCSK9 lacked clarity because no numerical threshold was given, or workable standard identified, that was said to define “competing.”

Justice Nicholas noted that no experts had suggested in evidence that they could not determine whether two antibodies competed with each other to bind an antigen. The evidence did show that there could be some discrepancy between results of different assays or of choices made in the assay procedure. However, the person skilled in the art would be aware of this and still be able to determine if there was competition. Accordingly, the relevant claims were found to be appropriately clear.

Sufficiency

Sanofi argued that the claims in question were each in effect claims to multiple inventions, being a series of individual MAbs, each of which would bind its own epitope on PCSK9. If that was so, then the sufficiency of the specification ought to be assessed as if each separate invention was the subject of its own claim,³ meaning that the specification ought to enable the person skilled in the art to generate at least one antibody within each such claim.

Justice Nicholas disagreed with this categorisation of the claims, saying that they were each to a class of MAbs. Sanofi conceded that if the claims were properly categorised as each relating to a single invention, then the specification was sufficient.

Outcome and Implications

Sanofi’s appeal was dismissed, the Delegate’s decision was affirmed, and the Applications were directed to proceed to grant.

This case is of considerable interest as it confirms that it is permissible to define a MAb by reference to its functional results rather than by its amino acid sequence or other physical attributes. Such a MAb will be adequately described if a person skilled in the art is able, without undue difficulty or invention, to carry out experiments to determine whether it falls within the claims of the patent. A MAb described in this way is also a patentable “manner of manufacture.”

This decision also confirms that potential means of solving a known problem may be known but not render the use of that means obvious, if there are other known possible solutions and knowledge in the field has not progressed sufficiently to determine whether the means is in fact workable or preferable.

¹ D’Arcy v Myriad Genetics Inc [2015] HCA 35; (2015) 258 CLR 334.

² Relying on AstraZeneca AB v Apotex Pty Ltd [2014] FCAFC 99; (2014) 226 FCR 324 at [254]-[255].

³ Relying on Tramanco Pty Ltd v BPW Transpec Pty Ltd (2014) 105 IPR 18 and subsequent cases.

About Pearce IP

Pearce IP is a specialist firm offering intellectual property specialist lawyers and attorneys with a focus on the life sciences industries.  Pearce IP and its leaders are ranked in every notable legal directory for legal, patent and trade mark excellence, including: Chambers & Partners, Legal 500, IAM Patent 1000, IAM Strategy 300, MIP IP Stars, Doyles Guide, WTR 1000, Best Lawyers, WIPR Leaders, 5 Star IP Lawyers, among others.

In 2025, Pearce IP was recognised by Australasian Lawyer and New Zealand Lawyer’s 5 Star Employer of Choice, and is the “Standout Winner” for inclusion and culture for firms with less than 100 employees. Pearce IP was awarded “IP Team of the Year” by Lawyers Weekly at the 2021 Australian Law Awards. Pearce IP is recognised by Managing IP as the only leading ANZ IP firm with a female founder, and is certified by WEConnect International as women owned.