Strict US written description requirement applied to CAR-T-cell therapy (Juno v Kite)

In the US, functional antibody claims have increasingly failed to satisfy the strict "written description" sufficiency requirement. The written description requirement stipulates that a patent specification should sufficiently describe the claimed invention such that a skilled person would be convinced that the inventor had possession of the claimed subject matter at the filing date. In a Court Appeal of the Federal Circuit (CAFC) decision last year, the same reasoning was applied to a broadly claimed molecule for CAR-T-cell therapy (Juno v Kite). The decision in Juno v Kite is not a surprise in light of the recent CAFC case law on written description for antibodies, and represents yet another nail in the coffin of functional genus claiming for biomolecules in the US. 

Science background: CAR-T-cell therapy

The application from which the patent claimed priority dated back to 2002, a time dubbed by one of the inventors as "the birth of the CAR-T field". Chimeric antigen receptor (CAR) T-cell therapy is a revolutionary form of immunotherapy in which a patient's own T cells are reprogrammed to target harmful (e.g. cancer) cells (Singh et al.). 

Cancer targeting T-cell

T cells are a type of immune cell that are directed to a particular target by their highly specific surface T-cell receptors (TCRs). TCRs share many structural and functional similarities with antibodies, and resemble antibodies in their diversity and complexity. TCRs are multi-domain proteins, having an intra-cellualar signalling domain, a costimulatory domain, and a target specific binding domain.  When the binding domain of a TCR expressed by a T cell binds its specific target, the intracellular signalling and costimulatory domains work together to activate the T cell. The activated T cell initiates tumour killing and divides to produce more T cells expressing the specific TCR. 

CAR-T cancer therapy involves engineering T cells to express a synthetic equivalent of a TCR, a "CAR". In cancer CAR-T cell therapy, the CARs expressed by the engineered T cells orchestrate tumour specific killing. CARs have the advantage of not requiring secondary signalling in order to activate the T cell. In modern CAR-T, the binding domain of a CAR is generally borrowed from an antibody, and takes the form of a single-chain antibody variable fragment (svFc), i.e. single antibody VH and VL domains connected by a linker.  

Legal background: Written description

Written description is a type of sufficiency requirement, derived from the stipulation in US law that a patent specification "shall contain a written description of the invention" (35 US Code § 112(a)). The written description has been understood by the US courts as requiring the patent specification to describe the invention such that it reasonably conveys to a skilled person that the inventor had possession of the claimed subject matter as of the filing date. As such, "a mere wish or plan" for obtaining the claimed invention is not sufficient to satisfy the written description requirement.   

Over the years, the US courts have grown stricter and stricter in their application of the written description requirement to functional language or genus claims for biological inventions such as antibodies. The written description requirement has been interpreted as requiring demonstration in the specification that the patent applicant "has invented species sufficient to support a claim to the functionally-defined genus" (Ariad v Eli Lilly). In practice, the bar for what constitutes a sufficient number of species has been set very high. In Abbvie v Janssen, for example, a claim directed to a functionally defined anti-Il-12 antibody was found invalid for lack of written description despite disclosure in the specification of 300 example antibodies. In this case, the CAFC found that that 300 examples did not sufficiently represent all antibodies that might fall under the scope of the claim. 

To this Kat's knowledge, the decision in Juno v Kite is the first time the CAFC has applied the written description case law to a CAR-T-cell invention. 

Full-scope written description 

The patent in question (US 7,446,190) was exclusively licensed by Juno Therapeutics (a subsidiary of BMS) and related to CARs for a T cell. The disputed claims specified a nucleic acid encoding a CAR comprising an intracellular domain, a costimulatory signalling domain comprising a defined amino acid sequence, and a scFv binding element that interacts with a selected target. During infringement proceedings relating to the patent, the defendant (Kite Pharma Inc.) counterclaimed that the claims were invalid for not being supported by sufficient written description.

The CAFC noted that the claims included the functional definition that the scFv binding element of the CAR "specifically interacts with a target". The description defined a target as "any target of clinical interest". In other words, the CAFC pointed out, the claimed CAR claimed "any scFv for binding any target" (CAFC's emphasis). The question thus became whether the specification provided a sufficiently representative sample of the species to support such a broadly claimed genus. 

The specification disclosed two working embodiments, binding two different targets of interest. Juno argued that these embodiments were representative of all scFvs, given that scFvs are well known and essentially interchangeable. As their expert argued, "[scFvs ] all do the same thing. They bind to the antigen". As such, Juno argued, a skilled person would reasonably believe from the examples that the inventor possessed the full scope of the claimed invention. 

The CAFC did not agree. First, the CAFC noted that only scant details were provided of the example scFvs (in the form of a reference number). Furthermore, the specification provided no information that would help a skilled person determine how or whether the example scFvs were representative of the entire claimed genus. The CAFC's reasoning will be familiar to those that have followed the case law on written description as applied to monoclonal antibodies. Particularly, the CAFC found that "the disclosure...does not provide information sufficient to establish that a skilled artisan would understand how to identify the species of scFvs capable of binding to the limitless number of targets as the claim require".  The CAFC also dismissed the expert evidence on the general interchangeability of scFvs given that "[w]ithout more in the disclosure, such as the characteristics of the exemplary scFvs that allow them to bind to particular targets [...], the mere fact that scFvs in general bind does not demonstrate that the inventors were in possession of the claimed invention". 

The CAFC also dismissed the argument from Juno that the specification disclosed the structural features common to members of the claimed genus of scFvs. Whilst the CAFC agreed that scFvs have a common general structure, the CAFC pointed out that the structural feature of importance was the target binding region of the molecules, and that these would be unique for each target. The CAFC noted that the patent not only failed "to disclose structural features common to scFvs capable of binding specific targets, it also fails to disclose a way to distinguish those scFvs capable of binding from svFvs incapable of binding those targets". 

The CAFC thus overturned the lower court decision (and the $1.2 billion awarded in damages), and found that the patent did not contain written description support for the claims. 

Final thoughts

The problem with the patent in this case for the CAFC was not just that it claimed a broad genus of binding molecules ("millions of billions") of which only two examples were provided, but that a way of identifying suitable targets "of clinical interest" for the binding molecules was not provided in the specification. This raises the question as to whether a different definition of "target" would have helped the patentee, e.g. if it didn't matter whether the scFv bound a clinically interesting target or not. After all, a particular CAR structure has potential utility beyond its use as a clinical molecule (e.g. as a control vector). Notably, the claimed invention of the patent at stake in Juno v Kite has been broadly recognised for its significance in progressing the CAR-T field.  

The US focus on written description thus contrasts with the European/UK approach, which takes more account of the inventive contribution of the claimed invention, as opposed to whether a skilled person believed the inventor had possession of all embodiments falling within the scope of the claim. By stark contrast, for example, the EPO only requires that the specification sufficiently discloses a single embodiment falling under the scope of the claim (see also, Regeneron v Kymab, (IPKat) and Illumina v MGI (IPKat)).  The EPO is thus prepared to grant patents with functional claims, provided that the inventive contribution is convincing (IPKat). 

The distinction between the EPO and US approach seems partly situated in the different interpretations of what these jurisdictions consider "routine". One can imagine, for example, the EPO finding that a claim directed to a specific CAR to be uninventive and the result of routine experimentation, in view of a general disclosure of the CAR structure. Inventors can get stuck between these opposing positions. A functional genus of a biomolecule will, for most intents and purposes, be considered insufficiently disclosed by the US courts. A specific species of the genus will often-as-not be considered obvious in Europe. When then should applicants file their patent application?

Further reading

Patenting Antibodies: The epitope claim is dead, long live the epitope claim (15 Nov 2021)