Antibody patents: Danger ahead for biologics

In recent years, the written description requirements at the US Patent and Trademark Office (USPTO) for broadly claiming antibodies have been significantly raised. These heightened requirements follow a series of cases by the Court of Appeals for the Federal Circuit (CAFC) over the last few years.

This article will explore what requirements are “mandated” by CAFC precedents, how USPTO is currently examining claims that recite an antibody using broad terms, and when and how applicants can push back when claims that recite antibodies broadly are necessary to protect an invention.

CAFC precedent

Before 2014, the majority of antibody patents invalidated for lack of written description involved issues of claiming embodiments which were clearly not possessed. For example, in Chiron v Genentech, 363 F.3d 1247 (Fed Cir 2004), a humanised antibody claim was later added to a case that was filed before humanised antibodies were in fact in existence. After 2014, several patents with functional claims against certain antigens have been invalided, eg, “an antibody against protein X”.

The primary case in 2014 invalidating an antibody product claim was AbbVie v Janssen, 759 F.3d 1285 (Fed Cir 2014). AbbVie sued Janssen for infringement of several claims of US patent 6,914,128 and US patent 7,504,485. An exemplary claim of the ‘128 patent read as follows: “29. A neutralising isolated human antibody, or antigen-binding portion thereof that binds to human IL-12 and dissociates from human IL-12 with a koff rate constant of 1 × 10-2s-1 or less, as determined by surface plasmon resonance.”

Janssen challenged the written description of the AbbVie patents. AbbVie developed its IL-12 antibodies using phage display, which involved creating a large library of human DNA fragments and screening for those fragments that encoded an antibody fragment with IL-12 binding affinity. AbbVie isolated only one primary clone, Joe-9, from which several derivatives via amino acid mutations were formed. The court held that AbbVie did not possess the broad claim at issue.

After AbbVie v Janssen, the USPTO did not immediately and significantly alter its standards with respect to obtaining a patent. For example, USPTO maintained its previous guidance found in the March 25, 2008, “Written Description Training Materials”.

Example 13 of the training materials held that a claim such as “An isolated antibody capable of binding to antigen X” was sufficiently described by a specification which described purifying and isolating antigen X only because the level of skill and knowledge in the art of antibodies is such that production of antibodies against a well-characterised antigen is conventional.

However, AbbVie v Janssen was followed by Amgen v Sanofi, 872 F.3d 1367 (Fed Cir 2017). Amgen sued Sanofi for infringement of several claims of US patent 8,829,165 and US patent 8,859,741. An exemplary claim of the ‘165 patent read as follows: “1. An isolated monoclonal antibody, wherein, when bound to PCSK9, the monoclonal antibody binds to at least one of the following residues: S153, I154, P155, R194, D238, A239, I369, S372, D374, C375, T377, C378, F379, V380, or S381 of SEQ ID NO:3, and wherein the monoclonal antibody blocks binding of PCSK9 to LDL[-]R.”

Sanofi challenged the written description of the asserted Amgen patent claims. In the patent disclosure, 3,000 human monoclonal antibodies were rescreened for binding to wild-type PCSK9 to confirm stability, which were eventually narrowed to 85 antibodies that blocked the interaction between PCSK9 and the low density lipoprotein receptor (LDLR) at greater than 90%.

The specification also contained 22 antibodies that compete with Repatha (evolocumab, a monoclonal antibody designed for the treatment of hyperlipidaemia) for binding to PCSK9. The Federal Circuit did not issue a final decision on the merits of written description, but instead, remanded the case for further consideration by the district court. The Federal Circuit noted in the remand that the district court judge instructed the jury that, in the case of an antibody, the correlation between structure and function may be satisfied by the disclosure of a newly characterised antigen, if the level of skill was such that the production of antibodies was conventional or routine.

The exact language of the jury instructions was: “In the case of a claim to antibodies, the correlation between structure and function may also be satisfied by the disclosure of a newly characterised antigen by its structure, formula, chemical name, or physical properties if you find that the level of skill and knowledge in the art of antibodies at the time of filing was such that production of antibodies against such an antigen was conventional or routine” (Amgen v Sanofi, at 1359).

The Federal Circuit took issue with the jury instruction and found that it permitted the jury to dispense with the required finding of a “written description of the invention”. The Federal Circuit stated that it was unwilling to hold that “make and use” (routine or conventional production) of antibodies equates to the required description because such jury instruction would improperly transform a factual issue into a legally required inference, wherein it has been hotly disputed that knowledge of the chemical structure of an antigen gives the required kind of structure-identifying information about the corresponding antibodies.

The Federal Circuit noted that judicial notice is proper only when a fact is either “generally known” or “accurately and readily discernible from sources whose accuracy cannot reasonably be questioned”. The Federal Circuit concluded that the instruction improperly contradicted the statutory quid pro quo by allowing the patentees to claim antibodies by describing something that is not the invention, ie, the antigen.

Current USPTO standards

After the Federal Circuit decision in Amgen v Sanofi the USPTO significantly raised its standard for obtaining a patent in connection to written description. On February 22, 2018, USPTO issued a memorandum titled “Clarification of Written Description Guidance for Claims Drawn to Antibodies and Status of 2008 Training Materials” from Robert W. Bahr, deputy commissioner for patent examination policy.

The USPTO memorandum suspended the guidance found in the March 25, 2008 “Written Description Training Materials”, namely example 13, and directed USPTO personnel to the general written description guidance of MPEP 2161.01 and MPEP 2163.

Furthermore, after Amgen v Sanofi, a number of decisions of written description of antibodies came down at the Patent Trial and Appeal Board (PTAB). These decisions include In re Kyung Jin Kim, Appeal 2017-006385 (US 13/327,624; decided October 30, 2018), In re Peter C Gray, Appeal 2017-001821 (US 12/615,033; decided November 30, 2018), and In re Shiv Srivastava, Appeal 2017-010877 (US 14/287,599; decided December 26, 2018).

In In re Kyung Jin Kim, several claims were appealed. Claim 15 is illustrative of the issues addressed: “15. A method of treating cancer in a patient comprising administering to the patient a pharmaceutical composition comprising a monoclonal antibody that competes for binding to FGF2 with an antibody produced by hybridoma PTA-8864, wherein the monoclonal antibody neutralises FGF2, and the cancer is a type of cancer that responds to FGF2.”

The examiner argued that the knowledge of the amino acid sequence of FGF2 did not put the applicant in possession of all monoclonal antibodies that compete for binding to FGF2 with an antibody produced by hybridoma PTA-8864. The appellant argued that an epitope can be defined by a competition assay in which the genus of antibodies competes with an archetypal antibody for binding to a well-defined antigen.

The PTAB found a lack of written description as the specification contained only a single example within the claim, ie, monoclonal antibody GAL-F2 produced by hybridoma PTA-8864.

In In re Peter C Gray several claims were appealed. Claim 15 is illustrative of the issued addressed: “15. A method of treating a hyperproliferative disease in a human subject comprising administering to the subject an amount of a selective targeting compound that is effective to inhibit Cripto signalling in hyperproliferative cells that express glucose regulated protein 78 (CRP78) on their surfaces and thereby reduce the cells’ proliferation, wherein the selective targeting compound is an antibody that binds within the sequence region defined by amino acids 19-68 of GRP78 and inhibits the formation of complexes between human Cripto and GRP78.”

The examiner argued that the specification as filed disclosed only a single polyclonal antibody, and the polyclonal antibody was not representative of the claimed genus, which includes monoclonal antibodies. The PTAB found a lack of written description citing Sanofi v Amgen because the appellants attempted to describe the invention by describing something that is not the invention, ie, the antigen.

In In re Shiv Srivastava several claims were appealed. Claim 1 is illustrative of the issues addressed: “1. A monoclonal antibody that binds a conformational epitope formed by amino acids 42–66 of Seq ID No: 1.”

The examiner argued that the claims were broadly drawn to any antibody which binds to a conformational epitope formed by amino acids 42–66 of Seq ID No: 1, wherein the specification described only one anti-ERG antibody which binds an epitope within amino acids 42–66 of 42-66 of Seq ID No: 1.

The PTAB found a lack of written description, citing Sanofi v Amgen, because the appellants attempted to describe the invention by something that is not the invention, ie, the antigens to which the antibodies may bind. Appellants pointed to nothing else in the disclosure that describes the antibodies
as required.

New position

As a result of the USPTO memorandum of February 22, 2018, as well as the above-listed CAFC and PTAB decisions, the majority of patent examiners and supervisory patent examiners at USPTO have begun taking the position that any functional antigen-based antibody claim lacks written description, ie, an antibody against protein X.

The examiners contend that the genus of antibodies against any protein is extremely large and thus, any applicant is in “possession” of only a very small number of species within the very large genus. In support of the position the examiners cite Sanofi v Amgen as well as the above-listed PTAB decisions.

A number of examining art units at USPTO have taken the reasoning of Sanofi v Amgen further and applied similar rationales to method of treatment claims which comprise administering an antibody functionally defined by binding to a specific antigen, even if many antibodies with the functional characteristics are previously known in the art.

The USPTO personnel state that applicant has possession only of an animal model showing successful treatment with one or two antibody clones. Applicant is not in possession of the method of treatment for any antibody with the functional characteristics, considering the large size of the genus of antibodies functionally included within the scope of the claim. Examiners are even applying those principles to diagnostic methods using antibodies.

Companies have taken notice of the current environment and have begun using the USPTO’s strict position on written description to attempt to invalidate patents at the USPTO. On April 2, 2019, Eli Lilly filed for a post-grant review (PGR) of Genentech’s US patent 10,011,654. Eli Lilly alleged in its petition that all seven claims of the ‘654 patent are unpatentable for lack of written description.

Just a day earlier UCB v Genentech (PTAB-PGR2019-00044) also challenged the validity of the ‘654 patent with regard to written description by filing a separate PGR petition.

The USPTO’s current position probably places at risk a large portion of current biological pharmaceuticals. Biologics currently account for as much as 40% of prescription drug spending in the US. The biologics market is dominated by monoclonal antibodies, with a share of 43%.

Furthermore many, if not most, of those monoclonal antibody drugs are protected at least in part by patents which functionally claim the antibody used by reference to the antigen, often without severely limiting the claims to the actual amino acid sequence of the antibody. Accordingly, it is likely that this field of patent law will continue to see more decisions further clarifying exactly where is the line with respect to written description.

Where do we go from here?

It is important that patent owners and practitioners take measures to protect themselves, during patent prosecution and during challenges at the USPTO. It is best to claim antibodies with respect to the complementarity-determining region (CDR) sequences, despite the limited protection offered by such claim design. It is also advisable to claim narrow epitopes on a target antigen with several clones specific to the epitope such as to show a possession of a larger portion of the genus of antibodies covered by specific claim language.

If the patent application itself does not clearly demonstrate that the inventors have tested a large number of antibodies of diverse sequences and confirmed that they have the claimed functional characteristics and work in accordance with any other requirements of the claims, there is still hope.

A patent applicant should be able to file a declaration that provides a detailed scientific explanation as to why a person of ordinary skill in the art as of the application filing date, upon reading the specification, would understand that the inventors were in possession of the claimed invention. The declaration can be supplemented with details of further experiments performed after the application was filed. The use of such declarations is often necessary in order to overcome examiner’s rejections and is always required when defending patents in a PGR challenge.

Regarding possible future developments in the case law, it is noted that on remand in Amgen v Sanofi the jury found in February this year, after receiving new and presumably correct jury instructions, that most of the asserted claims of the patents met the written description requirement, despite the fact that none of the claims required specific sequences for the antibody.

This case should be closely watched as it progresses, as it should provide clarity on this important issue.

John H. Heithaus is an associate at Birch, Stewart, Kolasch & Birch. He specialises in drafting and prosecuting patent applications in the areas of biotechnology, oil and gas, biochemistry and pharmaceuticals. He has filed multiple appeal briefs before the PTAB and participated as counsel in litigation matters in the US. Heithaus previously worked at a leading biological and chemistry firm in Osaka, Japan, advising international clients on Japanese prosecution and litigation. He can be contacted at: jhh@bskb.com

Gerald M. Murphy is a partner at Birch, Stewart, Kolasch & Birch. He has prosecuted thousands of patent applications in the areas of biotechnology, biopharmaceuticals, therapeutic antibodies, organic chemistry, biochemistry, and pharmaceuticals. In the biopharmaceutical field (including antibody-related inventions), in addition to patent prosecution and validity and infringement opinions, Murphy has served as an arbitrator in patent disputes, as an expert witness in district court litigations and handled interferences and trials before the PTAB. He can be contacted at: gmm@bskb.com