Divergent Approaches, Consistent Outcomes: UPC and EPO Tackle Inventive Step in Amgen’s Patent Case

The patent dispute between Amgen, Sanofi, and Regeneron over cholesterol-lowering drugs marks the first case before the UPC. Initially, Amgen’s European patent EP 3 666 797 (the “Patent”) was revoked by the Central Division in Munich (CDM) in the first ever central revocation decision, but the Court of Appeal (COA) has now reversed it (UPC_CoA_528/2024 and UPC_CoA_529/2024, 25 November 2025), offering important guidance on inventive step and claim interpretation, particularly for medical use claims.

1. Inventive step before the COA of the UPC

The UPC had already hinted at its approach to inventive step in the landmark decision UPC_CoA_335/2023 of 26 June 2024. This time, the COA ruled in favour of the validity of the Patent and made explicit its approach to inventive step, which must be presented and put in perspective with that of the EPO.

After recalling that both the EPO’s “problem-solution approach” used in some jurisdictions and the “holistic” approach of certain other jurisdictions, such as Germany or the UK, are merely guidelines for implementing Article 56 EPC, the COA proceeded to set out its principles on inventive step assessment. First, a realistic starting point is selected that would have been of interest to a skilled person who, at the relevant date, wishes to solve the objective problem. This may, for instance, be the case if the relevant piece of prior art already discloses several features similar to those relevant to the invention as claimed and/or addresses the same or a similar underlying problem as that of the claimed invention. According to the COA’s own observation, there can be more than one realistic starting point, and the claimed invention must be inventive starting from each of them. This approach should not result in a different conclusion from any other method.

Then, an “objective problem” must be identified. This is done from the perspective of the “skilled person” by establishing what the invention adds to the state of the art, comparing the claims as a whole to identify the technical teaching of the invention, which must be based on the technical effect that the skilled person understands is achieved by the invention. It adds that this objective problem should not contain any pointer or motivation to the “claimed solution”. This claimed solution is considered non-inventive when the skilled person, starting from the “realistic starting point in the state of the art in the relevant field of technology”, wishing to solve the problem, would (and not only could) have arrived at the claimed solution. If the skilled person would have taken the next step with the expectation of solving the objective problem, or if this next step would have predictable results or a reasonable expectation of success, then the claimed solution will be regarded as obvious and the patent as lacking inventive step.

Finally, the COA recalls that the burden of proof lies with the party claiming a lack of inventive step, both for demonstrating that the results were clearly predictable or that success was reasonably expected, as well as for addressing any difficulties or substantiated uncertainties that would impact the reasonable expectation of success put forward by the patentee.

With this approach, the COA confirms the view that the CDM adopted: a pragmatic view under which, if a skilled team is motivated to take the next logical step in developing the prior art, and does not foresee any particular difficulties in doing so, the claim lacks inventive step. However, contrary to the CDM, the COA found that there were no obvious steps in considering that the development of antibody therapies would succeed in solving the objective problem based on the realistic starting point. The various prior art documents did not present sufficient prospects of success in this regard, as they all indicated that, although this route could be promising, there was no evidence that the therapeutic aim of the invention would be achieved. Each prior art document considered that further experiments or studies were necessary to confirm that this route was viable. This lack of certainty regarding the effectiveness of the antibody in an in vivo environment led the COA to conclude that the skilled person would still have regarded the antibody route as uncertain, meaning that it was not the obvious next step and that the Patent was therefore valid.

2. Different approach from the EPO, same conclusion

This confirms what was hinted at in previous decisions from the UPC, that it would adopt its own approach to inventive step, not necessarily strictly following that of the EPO. The COA sees no risk of divergent rulings as the different approaches are just guidelines which, “when properly applied, should and generally do lead to the same conclusion”. In the present case, despite the different approaches, both the COA and the Opposition Division of the EPO (which upheld the Patent in parallel opposition proceedings) ended up ruling that the Patent was valid. It remains to be seen what the Board of Appeal of the EPO will decide and whether consistency will be maintained.

When comparing the COA’s approach with the “problem-solution” approach, the main differences are as follows: First, the COA focuses on the “objective problem” as a starting point, rather than the closest prior art. While the “objective technical problem” of the EPO is defined by reference to the closest prior art, as “the aim and task of modifying or adapting the closest prior art to achieve the technical effects that the invention offers over the closest prior art”, this is not the case in the COA’s decision. The COA’s objective problem is assessed globally and not with a single piece of prior art. This does not prevent combining the most realistic starting point with other pieces from the state of the art, which was precisely done in this decision by the COA.

This results in differences between the “closest prior art” and the “realistic starting point”. Whereas the closest prior art discloses, “in one single reference, the combination of features which constitutes the most promising starting point for a development leading to the invention”, the COA states that a realistic starting point can “disclose several features similar to those relevant to the invention as claimed and/or address the same or a similar underlying problem”, but also that there can “be more than one realistic starting point and the claimed invention must be inventive starting from each of them”.

At the final step of the problem-solution approach, regarding whether or not the invention would have been obvious to the skilled person, the COA adopts a similar approach to the EPO, with the same requirement that it is not sufficient that the teaching in the prior art could have prompted the skilled person to act as per the claimed invention; rather, it must be shown that this would have prompted the skilled person, as the prior art provided a pointer or motivation to do so with a reasonable expectation of improvement or advantage.

It appears that the COA wanted to keep flexibility in the assessment of inventive step and not make the “problem-solution approach” the sole guideline. Nonetheless, the decision of the COA aligns with that of the EPO Opposition Division and the COA made it clear that different approaches should not, in general, lead to different outcomes.

3. Claim interpretation for claim drafted in “medical use-format”

Another main clarification by the COA is how medical use claims should be interpreted and the consequence on assessment of inventive step. The COA clarified that the interpretation of the independent claim of the Patent must be based solely on its own features, without reference to the additional elements found in dependent claims. The claim covers treatment with the claimed antibodies as the only medicament, and thus these antibodies must be therapeutically effective on their own. The COA agreed with Amgen that the claim implicitly requires the antibodies to be effective as a medicament for treating or preventing the specified diseases. This means that not just any reduction in cholesterol levels is sufficient; the treatment must result in a meaningful, noticeable improvement in the patient’s medical condition. While the claim does not specify a minimum threshold for therapeutic effect, the effect must nonetheless be therapeutically relevant.

As a result, where a claim requires that the antibody must provide a meaningful therapeutic effect, this requirement is assessed based solely on the features of the main claim and is not broadened by the additional elements of dependent claims involving combination therapies. Furthermore, while the technical effect of binding to the catalytic domain is relevant to the assessment of inventive step, the interpretation of the claim itself remains limited to the features expressly stated in the main claim. The fact that the skilled person does not derive any minimum required effect from the claim or the description does not alter this conclusion, since the requirement for a therapeutic effect arises not from the claim language itself, but from the use of the medical use claim format.

The EPO had also taken this view during the examination of the opposition and highlighted that the lack of direct evidence in the prior art for therapeutic efficacy meant that a skilled team could not have reasonably expected success, rendering the invention non-obvious. It referenced the EPO Enlarged Board of Appeal decision G2/21, which requires “plausibility” of therapeutic effect in patent applications for medical use claims. It concluded that if the prior art would not meet that standard, it cannot support a reasonable expectation of success either.

In summary, the COA’s decision in the dispute over the Patent clarified its stance on inventive step and claim interpretation for medical use claims. The COA also made it clear that even using a different approach, the UPC and the EPO should not diverge in the result of their assessment. Both the COA and EPO’s Opposition Division ultimately upheld the Patent with nuanced differences in approach. The EPO appeal proceedings will be followed with great interest to see whether consistency will be maintained to the end. Oral proceedings are already scheduled for 13–15 April 2026.