More than two years after the start of its operations, the Unified Patent Court (UPC) has provided another clear teaching of substantive patent law. Notably, the UPC approach to inventive step has further crystallised following two landmark Court of Appeal decisions from 25 November 2025 – Amgen v Sanofi/Regeneron (“Amgen”) (UPC_CoA_528/2024 and 529/2024) and Meril v Edwards (“Meril”) (UPC_CoA_464/2024, 530/2024 and related appeals).
In these rulings, the Court explains how it evaluates inventive step under Article 56 EPC. Although the technologies at issue in these cases could hardly be more different (therapeutic antibodies versus prosthetic heart valves), the Court applies a single, coherent analytical framework, while remaining sensitive to the technical and factual context of each case.
This article first explains the UPC’s approach to assessing inventive step. It then takes a closer look at the application of this approach in the two landmark cases from November 2025 to, finally, compare the UPC approach to the long-standing EPO practice (problem-solution approach).
1. Inventive step at the UPC
In its November 2025 decisions, the UPC Court of Appeal emphasises that inventive step is a legal question assessed on the basis of underlying facts. The different ‘approaches’ – such as the EPO’s problem-solution approach, which has found its way into different national practices, or the more ‘holistic’ German or UK approaches – are methodological tools rather than binding formulas. In the Court’s view, these tools, when properly applied, should and generally do lead to the same result. In the November 2025 decisions, the Court clarifies the UPC ‘approach’ (methodology), building on its earlier NanoString v 10x Genomics order (UPC_CoA_335/2023) given in PI proceedings.
A proper inventive step analysis starts with identifying the objective technical problem, i.e. asking what the invention adds to the state of the art. This must be assessed from the perspective of the skilled person at the relevant (priority/filing) date, using common general knowledge. The skilled person is an abstract criterion to be defined by the court. The skilled person is a specialist in the technical field of the invention. In certain technical fields, this ‘person’ may be a team of specialists.
Crucially when determining the objective technical problem, and to avoid hindsight, the problem formulation cannot contain pointers to the claimed solution. The Court insists that the patent claim(s) must be compared, as a whole and considering the description and drawings, to the prior art. The inventive concept – the technical teaching conveyed by the claim(s) – is the focus, rather than the individual features of the claim(s).
A second step is the selection of one or more realistic starting points in the prior art. A disclosure is a realistic starting point if, at the relevant date, it would have been of genuine interest to the skilled person seeking to solve the objective technical problem underlying the claimed invention. There may be more than one such realistic starting point. The claimed invention must be inventive over each of these starting points.
A third requirement in the assessment of inventive step, is the ‘would’ rather than ‘could’ standard: obviousness (i.e. absence of inventive step) exists if the skilled person would have arrived at the claimed solution, and not if they merely could have done so. That, in turn, requires a pointer or motivation (for example, common general knowledge, a document, …) in the prior art that directs the skilled person, as a matter of routine development and without inventive imagination, toward the claimed solution. The claimant must demonstrate the existence of such a motivation or pointer. Indeed, the skilled person from whose viewpoint the inventive step assessment must be made, is not an inventor himself. He has no inventive skills and no imagination that would enable him to take an inventive step on his own.
Closely tied to the ‘would’ standard, set out above, is the concept of reasonable expectation of success, on which the Court of Appeal places particular weight. A solution is obvious if the skilled person, starting from a realistic starting point and seeking to solve the objective technical problem, would have taken the next step in the expectation that it would solve that problem. Predictability of results and scientific certainty therefore carry significant (evidentiary) weight. Conversely, where the patentee shows technical uncertainties or difficulties, the burden shifts to the challenger to show that these would not have deterred the skilled person. The reasonable expectation of success proved decisive in Amgen (see hereafter).
2. Amgen
In the Amgen case, the patent claims monoclonal antibodies (proteins) targeting PCSK9, for the treatment of hypercholesterolaemia and related conditions. By targeting (i.e. binding to) PCSK9, the binding of the PCSK9-molecule to low density lipoprotein receptor (LDLR) is, in turn, inhibited. Thus, a therapeutic effect on hypercholesterolaemia and related conditions is achieved.
The patent had been revoked at first instance for lack of inventive step. On appeal, the UPC Court of Appeal reverses the revocation and upholds the patent, notably by acknowledging an inventive step.
Applying its methodological framework, the Court firstly determines the objective problem that the invention aims to solve in light of the patent description, i.e. the provision of a treatment of hypercholesterolemia and other diseases related to elevated serum cholesterol levels. The Court explicitly rejected problem formulations referring to the provision of a monoclonal antibody or PCSK9 as a binding target to regulate LDLR levels, to prevent hindsight.
Secondly, the Court accepted Lagace as a realistic starting point. This document identified PCSK9 as a regulator of LDLR levels and suggested that binding to PCSK9 to inhibit its working could be therapeutically beneficial for hypercholesterolaemia and related conditions.
Thirdly, the question arose whether, at the relevant date, the skilled person would have had a reasonable expectation that an anti‑PCSK9 antibody would be therapeutically effective in vivo. Such reasonable expectation, the Court held, did not exist, for several technical reasons.[1] The skilled person would, in other words, not have embarked on the pathway of antibody development with a reasonable expectation of achieving a meaningful therapeutic effect. Absent this expectation, the claimed invention was held to be inventive.
In brief, this case signals that, in the life sciences sector, documented scientific uncertainty can be a powerful shield against obviousness. Later technical knowledge (hindsight) is not accepted to argue the skilled person’s reasonable expectation of success at the relevant date.
3. Meril
The Meril case presents a very different technological setting. The patent concerns a balloon-expandable transcatheter prosthetic heart valve whose frame is entirely made of hexagonal cells with side struts. The patent survived the first instance in amended form; on appeal the Court equally upholds it.
The inventive step debate focuses on whether the fully hexagonal frame was obvious over prior art heart valve designs, alone or in combination with general knowledge of ‘honeycomb’ structures from vascular stents (which technically differ from heart valves). The Court, referring to caselaw of the EPO Boards of Appeal, reiterates that inventive step does not require a dramatic performance gain over the prior art: a non‑obvious, alternative solution, as in the present matter, can suffice.
The Court then identifies the objective problem solved by the patent as improving the design of a balloon‑expandable transcatheter prosthetic heart valve, with respect to reducing the crimping profile of the heart valve and providing stability during crimping/expansion.
It accepts the prior art document Levi as a realistic starting point. Levi discloses heart valve frames that combine rhomboid and hexagonal cells, which each fulfil specific technical functions in specific regions of the valve. Levi, however, teaches only such mixed designs (rhomboid-hexagonal): it contains no motivation to create a frame using only hexagonal cells. The skilled person would thus have to step away from Levi to arrive at the claimed solution. This is more than ‘routine optimisation’. Meril then attempts to combine Levi with stent prior art, which, however, has fundamentally different functional requirements. As such, there is no motivation to transplant stent design principles to heart valve designs either. Absent a pointer leading the skilled person, as a matter of routine, to an all‑hexagonal heart valve design, the Court concluded that the patent’s solution involved an inventive step.
4. The UPC v. the EPO
While Article 56 EPC is the common legal basis, the EPO continues to apply its established problem-solution approach, which differs from the UPC framework described above. In its simplest form, the EPO methodology proceeds in three main stages: (i) selecting one closest prior art; (ii) identifying the distinguishing features and their technical effect(s) and, on that basis, formulating the objective technical problem; and (iii) applying the EPO’s could-would reasoning to decide whether the skilled person would have arrived at the claimed solution starting from that closest prior art.
Against that background, the UPC’s Court of Appeal framework differs from the EPO approach in several, practically relevant respects:
- Sequence of steps: the UPC starts by identifying the objective problem (what the invention adds to the state of the art) and only then selects one or more starting points; the EPO typically begins by choosing a closest prior art and builds the analysis from that starting document.
- Starting point(s): the UPC works with one or more “realistic starting points”, i.e. disclosures that would genuinely have interested the skilled person trying to solve the objective problem; by contrast, the EPO requires the selection of a “closest prior art” as “most promising springboard”. In a given case, the EPO can identify multiple “closest prior art” as the basis for an inventive step analysis, e.g. if such prior art documents disclose different aspects of the claimed invention, and one cannot be considered ‘closer’ than the other.
- Formulating the objective technical problem: importantly, at the UPC, the problem is derived primarily from the invention/patent as a whole (claims read in light of the description and drawings) and, more specifically, from its contribution over the state of the art; at the EPO, the problem is formulated starting from the closest prior art by (a) identifying the distinguishing features, (b) determining the technical effect(s) of those differences, and (c) expressing the problem as the task the skilled person would set themselves in view of those effects.
- Motivation / pointer: under the UPC’s “would” standard, obviousness requires a pointer or motivation in the selected realistic starting point (or the prior art/common general knowledge seen from that point) that would direct the skilled person, as a matter of routine, toward the claimed solution; at the EPO, the decisive question is whether the skilled person would (not merely could) have modified the closest prior art in the light of the objective technical problem and the overall prior art/common general knowledge – i.e. whether the solution was obvious over that closest prior art.
Practically, this means that EPO validity debates tend to be organised around a single chosen springboard and a tightly framed problem derived from the differences over that springboard, whereas UPC pleadings should be prepared to argue (and defend) the invention’s contribution more globally and to potentially address multiple realistic starting points – which includes, for challengers, articulating a concrete pointer that would have led the skilled person to the claimed solution. The Belgian caselaw admits the latter approach, with multiple springboards, as well.
5. Conclusion
Taken together, the two November 2025 Court of Appeal decisions send clear messages for practice. Problem formulation matters and must avoid hindsight. Reasonable expectation of success is central, especially in uncertain technical fields. Structurally departing from a specific design logic in the prior art can support inventive step even if individual elements are known.
As UPC caselaw matures, these principles will likely form the backbone of inventive step analysis across technologies, and advocacy will depend less on mechanical application of EPO formulas and more on persuasive, evidence‑based narratives about what the skilled person would realistically have done at the relevant date. This will perhaps prove to be the most important consequence of the UPC-specific approach to inventive step.
[1] First, there was uncertainty about PCSK9’s mechanism of action (intra- or extracellular). Second, the relative contribution of extracellular PCSK9 remained uncertain and experimental data suggested that pronounced effects occurred only in situations of overexpression. Third, the literature repeatedly described antibody approaches as conditional on further research, reflecting the field’s (limited) state of knowledge at the time.
For any questions or assistance, please reach out to our Intellectual Property Team | IP@simontbraun.eu – +32 (0)2 543 70 80
***
This newsletter is not a legal advice or a legal opinion. You should seek advice from a legal counsel of your choice before acting upon any of the information in this newsletter.