Inventive Step and Sufficiency of Disclosure: the UPC Court of Appeal Establishes Its Own Doctrine

The UPC Method for Inventive Step: A Structured and Holistic Approach

Far from merely importing an existing method, the Court of Appeal has meticulously defined a multi-step approach designed to be a clear guide for all its divisions.

The Steps of the UPC Method

1. The Objective Technical Problem is Defined First: A Shift in Perspective

This is undoubtedly the most fundamental nuance. Contrary to the usual practice of the EPO, which often identifies the closest prior art before formulating the technical problem, the Court of Appeal reverses this sequence. It states that the “objective problem” must be established first. This determination is based on an overall understanding of the invention’s contribution, examining the claim in the context of the description and drawings in order to grasp the technical teaching in its entirety (a clear reference to decision G2/21 of the EPO Enlarged Board of Appeal). This approach prevents the analysis from being reduced to a simple “hunt for differences” with a single document and promotes a broader view of the invention’s merit.

2. The Selection of One or More “Realistic Starting Points”

Once the problem is set, the Court places itself in the shoes of the skilled person and identifies a prior art document that would have genuinely and concretely interested this person in their search for a solution to the objective problem. The Court emphasises that several relevant starting points may exist. In that case—and this is an important warning for patentees—the invention must be considered inventive with respect to each of them. The Meril v. Edwards case provides a perfect illustration of this requirement, as the Court examined inventive step starting from no fewer than six different starting points raised by the parties.

3. The “Could/Would” Test and the Need for a Motivation

The Court formally adopts and endorses the “could/would” distinction, a cornerstone of EPO and several national jurisdictions’ case law. For an invention to be deemed obvious, it is not enough to show that the skilled person could have made it. It must be proven that they would have arrived at it, either through simple routine or by following a clear motivation—a “pointer” (Indice in French or Anlass in German)—present in the prior art. This skilled person, described as being devoid “of inventive skills and imagination,” needs to be guided.

4. The Non-Obvious Alternative: the Invention Does Not Have to Be “Better”

The Court carefully recalls an essential principle: inventive step does not necessarily require a technical improvement or advantage. The mere disclosure of a non-obvious alternative to existing technical solutions is sufficient to meet the threshold of Article 56 EPC.

The Amgen and Meril Cases

1. The Amgen Case: When “Hope of Success” Is Not Enough

The Amgen case provides an ideal testing ground for these principles, in the particularly unpredictable field of biotechnology. In overturning the first-instance decision that had revoked the patent for lack of inventive step, the Court of Appeal delivered a detailed and contextual analysis of the “reasonable expectation of success.”

The Court defines this criterion as “the ability of the skilled person to predict rationally, on the basis of scientific appraisal of the known facts, the successful conclusion of a research project within acceptable time limits.” It established a crucial distinction between:

• the “mere hope to succeed” which motivates exploration of a promising but uncertain field, and
• the “reasonable expectation of success” which requires a much higher degree of predictability.

In Amgen’s case, it was found that, at the priority date, fundamental scientific uncertainties persisted regarding the mechanism of action of the PCSK9 protein. The skilled person could not know whether blocking only the extracellular pathway (the only one accessible to the claimed antibodies) would have a sufficient therapeutic effect to treat or prevent the targeted diseases. They were navigating blindly. The Court therefore concluded that there was a “mere hope to succeed” existed” but not a “reasonable expectation of success.”

Furthermore, the Court clarified the interplay of the burden of proof: it is up to the party alleging invalidity to prove the reasonable expectation of success. If the patentee is able to demonstrate the existence of uncertainties or substantial technical difficulties, the burden of proof shifts to the challenger to prove that these uncertainties or difficulties would not have deterred the skilled person. Finally, the argument that “everyone was working on it” was  swept aside: a patent race in a given field can indicate its attractiveness and exploratory nature just as much as its obviousness.

2. The Meril Case: When the Prior Art Teaches Not to Make the Invention

The Meril v. Edwards case, concerning transcatheter heart valves, provides a perfect counterpoint to the Amgen case. Here, the issue was not biological uncertainty but mechanical engineering and the clear teaching of the prior art. After confirming the invalidation of the main claim for added matter (an unallowed intermediate generalisation), the Court upheld the Edwards patent in an amended version (Auxiliary Request II) claiming a valve frame “entirely made up of hexagonal cells.”

Meril’s argument was that this solution was obvious in light of a key prior‑art document, “Levi.” The Court of Appeal forcefully rejected this argument by applying its method:

• Levi’s Teaching: Levi described a valve with a heterogeneous frame, composed of a combination of rhomboidal and hexagonal cells. More importantly, Levi assigned specific functions to each cell type at precise locations on the frame to optimise its mechanical properties.
• Going Against the Teaching: The Court concluded that, to arrive at Edwards’ invention, the skilled person would have had to ignore Levi’s fundamental teaching. They would have had to abandon Levi’s heterogeneous and sophisticated design in favour of a fully hexagonal homogeneous structure. Levi therefore taught away from the claimed solution. There was no motivation, no pointer, that would have pushed the skilled person to make such a modification.

This analysis perfectly illustrates the “could/would” test: even if the skilled person could have technically built an entirely hexagonal valve, nothing in the prior art encouraged them to do so; on the contrary, the teaching of Levi pushed them in the opposite direction.

Beyond Technology: Proportionality of Injunctions and Public Health

The Meril decision is also remarkable for another aspect: the modulation of the infringement injunction. After confirming infringement, the Court took into account a public interest argument. Meril marketed extra‑large (XL) valves that Edwards did not offer. For some patients, these XL valves were the only viable therapeutic option.

Demonstrating notable judicial pragmatism, the Court created an exception to the injunction. It allowed Meril to continue supplying its XL valves, under the strict condition that a physician notify Edwards confirming that it is the only available treatment option for a given patient. This balance between protecting patent rights and safeguarding public health shows that the UPC positions itself as a court capable of rendering nuanced decisions adapted to human realities.

Sufficiency of Disclosure: One Way is Enough to Obtain Fair Protection

Beyond inventive step, the Court used these decisions to consolidate its position on sufficiency of disclosure (Art. 83 EPC), a particularly sensitive issue for functional claims.

It confirms a pragmatic and balanced approach:

• The “At Least One Way” Principle: to be valid, a patent must teach the skilled person at least one way to carry out the invention across the entire scope of the claim. For a functional feature, it is sufficient to disclose at least one technical concept to achieve it.
• No Requirement for Omniscience: the Court refers to “fair protection.” The patentee is not required to provide an exhaustive instruction manual for every conceivable embodiment. The impossibility of obtaining certain specific variants does not undermine sufficiency, as long as the patent provides the tools to obtain workable embodiments.
• The Acceptance of Trial and Error: a reasonable amount of trial and error is considered normal in implementing an invention and does not in itself constitute an undue burden.

Clear Guidelines and a Strong Signal for Practitioners

These twin decisions are much more than mere rulings. They constitute a genuine treatise on the UPC’s methodology. By creating its own synthesis of the best European practices, the Court of Appeal establishes framework for analysis that aims to be rigorous, flexible, and deeply rooted in the technical reality of each case. For practitioners, the message is crystal clear: arguments before the UPC must be solidly supported by facts, focusing on the real perspective of the skilled person at the relevant date and taking into account all the uncertainties and motivations of the time.

A new era of patent case law in Europe has truly begun, and it will be interesting to see whether the different approaches to assessing inventive step will lead to the same conclusions in parallel proceedings before the UPC and the EPO (in particular, the upcoming decision of the EPO Board of Appeal in Amgen v. Sanofi/Regeneron, T0716/25, which should be handed down in April 2026).

By Anne Rivière, February 10, 1026

Image credit: Markus Winkler (Pexels)