EPO decision T 0655/24: A Strict Stance on Post-Filing Evidence

Recently issued EPO decision T 0655/24 will be of great interest to those following antibody patenting at the EPO. It deals with several recurring themes, particularly around inventive step and post-filing evidence.

In this short series of blogs, Chris highlights three aspects of the decision that stand out:

• The Board¹ took a notably strict approach to the use of post-filing evidence to support inventive step, when the patent discloses an effect for the claimed antibody and the Proprietor seeks to use post-filing evidence to show an improvement in this effect compared to the closest prior art.
• The case illustrates an effective approach to claim drafting - combining limited sequence features (here, Fc point mutations) with functional language, to avoid an objection that the claim encompasses embodiments that don’t share the inventive properties of the exemplified antibodies.
• It also reinforces two commonly used arguments for inventive step: that the effects of combining known mutations cannot simply be assumed; and that obtaining one effect without adversely affecting other properties of the antibody is also not predictable.

Background

The claims were limited to antibodies having the triple Fc mutation L234F/L235E/D265A (“FEA”), which potently reduces effector function.

The closest prior art was the double mutant L234F/L235E (“FE”). While FE itself had not been tested for its effect on effector function, it was known that combining L234F/L235E with a different mutation (P331S, “FES”) led to reduced effector function. Separately, the D265A mutation was also reported to reduce effector function.

The patent also disclosed another triple mutant, L234F/L235E/N297Q (“FEQ”), which showed comparable reduction of effector function to FEA, but poorer stability, being cleared from plasma more rapidly. By contrast, FEA retained plasma clearance comparable to the wild type.

Post-filing evidence of improvement: when is it allowed?

To support inventive step, the Proprietor provided post-filing data showing that FEA provided a greater reduction in effector function than FE.

This raised the now-familiar question under G 2/21: is the relied-upon effect “encompassed by the technical teaching [of the patent] and embodied by the same originally disclosed invention”?

The Proprietor referred to a line of recent case law², all citing G 2/21, that if the patent discloses a technical effect, then an improvement in that technical effect is also implicitly derivable from the original teaching and can be supported by post-filing evidence. That is, the fact that the patent disclosed reduced effector function for FEA would be enough to allow post-filing evidence that the reduction is improved compared to FE.

The Board disagreed, citing other cases³ that pre-date G 2/21 but are discussed in that decision:

Applying the principles of G 2/21, the board does not consider that an improvement of an effect … is encompassed by the technical teaching and embodied by the same originally disclosed invention merely because the effect itself (but not the improvement), was shown to be achieved in the application as filed (see G 2/21, Headnote, Reasons 67 and 72).

Interestingly, having set out this strict standard, the Board ultimately found that the improvement over FE was already sufficiently disclosed in the patent. The post-filing evidence was therefore admissible, but unnecessary.

A deliberate message from the Board?

The decision is notable for the space the Board devotes to this point. The “catchword” focuses entirely on the treatment of improvements in light of G 2/21.

This is striking, as the Board could have resolved the case simply by finding that the improvement was disclosed in the patent, without considering the admissibility of the post-filing evidence at all. The extended reasoning on this point suggests a deliberate attempt to influence how future cases approach post-filing comparative data of improvement.

Proprietors advocating a less stringent test for admissibility of comparative data – at least before other boards – might find it helpful to argue that the Board’s reasoning on this point therefore seems to be obiter.

Practical takeaway

From a drafting perspective, this decision highlights the importance of trying to identify the most relevant prior art before filing, and providing evidence in the application as filed of improvement over that prior art.

We appreciate the challenge that this represents in the crowded antibody field!

Is the Board’s approach too strict?

Historically, the EPO has recognised the realities of a “first to file” system: applicants cannot reasonably be expected to identify and provide comparative data for all relevant prior art at the filing stage. On that basis, post-filing comparative data has been accepted, provided the effect shown by the data could be seen as related to the original technical problem addressed by the invention.

Against that background, it is striking that the Board invoked the “first to file” principle to justify a stricter approach, particularly in a case where the closest prior art had not itself demonstrated the relevant effect (Reasons, s. 72):

Furthermore, the approach taken in decisions T 1989/19, T 2716/19 and T 840/22 would allow an applicant to solely rely on post-published evidence to establish inventive step by filing an application disclosing an effect that was already known from the prior art or formed part of common general knowledge and subsequently invoking an improvement of said effect as "implicitly derivable" from the application as filed. This scenario goes against the principle that patents should not be granted for subject-matter whose inventive contribution is identified only after their effective date. Such an outcome directly contradicts the "first to file" principle enshrined in the EPC and confirmed in decisions G 1/03 and G 2/10. It is also in evident tension with the framework established by decision G 2/21 (see also decision T 314/20, points 6.13.3 and 6.14 of the Reasons).

There is also a tension with established practice allowing the reformulation of the objective technical problem in light of prior art identified during examination – an approach that remains reflected in the Guidelines even after G 2/21.⁴

Next in the series: Chris will consider the use of a combination of structural (here, point mutation) and functional definitions.

[1] Board 3.3.04 (Chakravarty, Rutz, Romandini)

[2] T 1989/19, T 2716/19 and T 0840/22

[3] T 0787/14, T 1791/11 and T 1322/17

[4] G-VII, 5.2