Patentee's own post-published data undermines the credibility of their broad cat antibody patent (T 0709/23)

The decision in T 0709/23, relating to a broadly claimed antibody invention, neatly illustrates the pitfalls of filing for a biotech invention too early before the link between structure, function and therapeutic effect has been sufficiently elucidated. Interestingly, this is also another example of a case in which post-published evidence from the patentee themselves undermined the original application, rendering it insufficiently disclosed over the whole scope of the claim. The case also mentions cats a lot, and so it was obligatory that PatKat take a read. 

EPO standards for sufficiency: Credibility of therapeutic effect (Article 83 EPC & G2/21)

For therapeutic inventions relating to the medical use of a substance or composition, therefore, the case law interprets the disclosure and sufficiency requirement of Article 83 EPC as requiring the claim to satisfy two tests. First, the skilled person must be able to make the substance or composition without undue burden. Second, it must be credible/plausible that the substance or composition actually achieves the claimed therapeutic effect (IPKat). Importantly, in most cases, for the requirement of sufficiency to be met, the patent application must provide suitable evidence for the claimed therapeutic use. A mere assertion of a therapeutic effect is not enough. Instead, the application as filed must provide a credible basis for the claim (IPKat). 

As the EBA confirmed in G2/21 "it is necessary that the patent at the date of its filing renders it credible that the [a] therapeutic agent, i.e. the product, is suitable for the claimed therapeutic application" (G2/21, r.74). According to the Board of Appeal interpretations of G2/21, post-published evidence may also be used to support a technical effect that was made credible by the application as filed (IPKat). 

Credible dog? 

Case T 0709/23: A broad functional claim supported by interspecies data

The patent in question in this case was EP3219729 owned by Zoetis Services and directed to an antibody for treating pruritic (itchy) and allergic conditions in cats. The claims defined the antibody purely by its function in inhibiting the inflammatory cytokine IL-31. The granted claims were impressively broad, with claim 1 reading simply:

An isolated antibody that specifically binds to feline IL-31, wherein said antibody reduces, inhibits or neutralizes feline IL-31-mediated pSTAT signaling in a cell based assay.

The claims of the Main Request on appeal were amended to also specify:

...for use in treating a pruritic condition or an allergic condition in cats.

The patent application provided data  demonstrating that example antibodies were effective at reducing itch in dogs. Based on the genetic similarity between canine and feline IL-31, the Patentee argued that it was credible these antibodies would also work in cats. The Patentee further supported their argument by pointing to documents showing that some medications for atopic dermatitis are used in both dogs and cats. The Opponents, including Boehringer Ingelheim Vetmedica GmbH, argued in response that the extrapolation from dogs to cats was speculative and that the patent failed to provide a sufficient disclosure for the claimed use in cats.

Post-published evidence contradicts the invention

Interestingly, the Board of Appeal decided that it was not necessary to settle the dog versus cat debate. Instead the Board of Appeal based its sufficiency assessment on the broader question of whether the invention could be carried out over its entire scope without undue burden. The Board of Appeal proceeded by simply assuming that the therapeutic effect observed in dogs was also credibly achieved in cats, allowing it to move on to what it considered was the more critical issue of whether the claim as a whole was credible. 

As interpreted by the Board of Appeal, the patent's central teaching was that antibodies binding a specific, conserved region on IL-31 would be effective at treating itching in cats. The application as filed particularly suggested that the antibodies binding this epitope would block the interaction of IL-31 with its receptor. This hypothesis was based on the functional effects observed with respect to an exemplary antibody "11E12" at inhibiting IL-31-mediated pSTAT signalling. 

This is where post-published evidence became relevant. The Patentee submitted post-published evidence describing detailed later experiments. These data showed that whilst the example 11E12 antibody did indeed inhibit pSTAT signalling as required by the claim, it completely failed to reduce itch in cats in a real-world study. However, the post-published document also contained data for a different antibody 15H05, which did reduce itching in cats. Importantly, however, the 15H05 antibody worked by binding to a completely different and previously unknown epitope on IL-31 than the 11E12 antibody. 

Unfortunately for the Patentee, the Board of Appeal considered the post-published evidence as detrimental to the Patentee's case. For the Board of Appeal, the post-published results demonstrated "that there is no clear correlation between the functional features of reducing, inhibiting or neutralising feline IL-31-mediated pSTAT signalling in a cell-based assay" and a therapeutic effect of reducing itching in cats (r.13). The example antibody which was shown to have the cited functional features was subsequently shown to lack any therapeutic effect. 

The Board of Appeal also noted that the patent application as originally filed did not suggest that a skilled person should look for antibodies having a different IL-31 epitope to the antibody used in the examples. The Board of Appeal therefore concluded that the patent's teaching would have not been much help to a skilled person. For the Board of Appeal, to perform the invention, the skilled person would have had to embark on a new research programme to identify an entirely new epitope, which would constitute an undue burden for the skilled person. The appeal was therefore dismissed and the patent was revoked for insufficiency.

Final thoughts

This case is the second recent case to fall down based on harmful post-published data originating from the patentee themselves (IPKat). Shockingly, it seems that post-published evidence is only useful when it supports your original disclosure and not when it contradicts it. 

This decision also demonstrates the limits of functional claiming at the EPO. Whilst the EPO is more permissive than the US when it comes to broad, functionally-defined antibody claims, this case shows us that the disclosure in Europe must still provide a reliable and repeatable path to success across the entire scope claimed. When using functional claim language it is therefore very important to link the specified function, and supporting data, with the actual problem solved by the invention. In this case, the Board of Appeal identified a disconnect between the functional language, the data provided in the application as filed, and the claimed therapeutic effect. 

In the antibody field, the decision contrasts with cases like T 0326/22 and T 0835/21, in which broad functional claims were upheld because the patent provided adequate guidance to obtain antibodies with the claimed properties. Unlike T 0435/20, in which insufficiency arose from the inherent difficulty of making antibodies to a complex conformational epitope without proper guidance, the problem in T 0709/23 was that the guidance on epitope selection provided in the application as filed was actively misleading as the post-published evidence suggested that a different epitope was needed. Indeed, the new epitope identified in the post-published evidence may be considered its own invention. 

On closer investigation, it turns out that the post-published data submitted in the present case was in fact taken from a later patent application filed by the Patentee (WO 2019177697 A2), claiming the new epitope, filed 7 years later. Unsurprisingly, the earlier case is cited as prior art on the later case. This case therefore neatly illustrates the dangers of filing a patent for broad antibody invention too early, before the link between structure, function and invention was fully elucidated (IPKat). This strategy runs the risk of revocation for insufficiency whilst simultaneously scuppering future filings to the clinical lead. 

Further reading

• Patenting Antibodies: The epitope claim is dead, long live the epitope claim (15 Nov 2021) 
• When generating antibodies for a target is more than routine (T 0435/20) (24 Apr 2023) 
• Broad functional claiming at the EPO (T 0835/21) (29 Sep 2023) 
• Epitope claims stand firm: Board of Appeal upholds functional antibody patent despite insufficiency attack (T 0326/22) (19 Jan 2025) 
• Plausibility demystified - a review of EPO case law before G 2/21 (Feb 2023)
• G 2/21 does not permit armchair inventing (T 0258/21) (Oct 2023)
• The plausibility of diagnostic inventions: No data, no patent (T 0589/22) (Apr 2025)