This is the final post in a series on decision T 0655/24 and its implications for antibody practice at the EPO.
In the first post, Chris discussed the Board’s strict approach to post-filing evidence to support inventive step. In the second, he looked at how combining structural and functional claim features can help avoid unnecessarily narrow structural definitions in antibody claims. In this article, Chris explains how the decision illustrates two approaches to arguing inventive step for antibody inventions, which can be very effective at the EPO.
Wasn’t the combination of mutations obvious?
At first glance, the outcome in T 0655/24 may seem surprising.
The claimed antibody included the triple Fc mutation L234F/L235E/D265A (“FEA”). The prior art proposed the double mutant L234F/L235E (“FE”) for reducing effector function (albeit without testing it) and separately disclosed the D265A mutation for the same purpose.
The Opponent argued that it would have been obvious to combine these teachings.
The Board disagreed.
1. Mutation combinations are inherently unpredictable
The Board noted that the prior art had not tested the FE double mutant as such, only in a different triple mutant with P331S (“FES”). In that context, they considered that it would have been unclear whether the double mutant would be effective to reduce effector function without P331S, and even more unclear how it would behave in combination with the D265A mutation.
2. Unpredictable effects on other properties
The Proprietor argued that the effect of the mutations on a different property of the antibody was also unpredictable.
The Board noted in the decision that the patent also discloses a different triple mutant (“FEQ”) that combines FE with N297Q instead of D265A. This mutant showed comparable reduction in effector function, but much diminished stability. Although the Board did not discuss FEQ in their reasons on inventive step, it seems very likely that it contributed to the Board’s finding that inventive step for FEA lay in the improved reduction in effector function while maintaining antibody stability. Without evidence that another mutation reduced stability, it seems unlikely that the Board would have consider that the maintained stability of FEA was significant.
Not a “mix-and-match” exercise
Taken together, these arguments reflect a broader theme in antibody practice.
Antibodies are not a simple “mix-and-match” system. Mutations within a given domain interact in complex and often unpredictable ways. It cannot be assumed that combining known features will produce an additive effect, or that modifying one property will leave others untouched.
EPO examiners often recognise this in practice, particularly where the application includes supporting data. It is very encouraging to see the Board following the same approach.
From a drafting point of view, it can be very useful to include details of other, less preferred or even “failed”, antibodies, to support these arguments and illustrate that the path to the claimed antibody(ies) was not straightforward.
-3.png)
