The UPC going into 2026 – three key takeaways

2026 Week 1

2025 was a seriously busy year for the UPC. We came into the year with one case management system and we left it with a shiny new one. We go into 2026 with a new court fee schedule. In the meantime, the workload of the court has cranked up with some divisions, and the Court of Appeal, having lots of cases but keeping up with the flow.

Looking back over 2025, three key topics catch the eye. So we thought we should have a recap before we get distracted by new decisions as we start the new year.

Preliminary injunctions – the UPC approach is now clear

An entire story played out in the course of 2025 in preliminary injunction (PI) proceedings between Boehringer Ingelheim and Zentiva.  

The first instance proceedings were the subject of UPC Weekly 2025 w19, with the Lisbon Local Division refusing to grant a PI to Boehringer Ingelheim against Zentiva. It was clear that the product (nintedanib for use in the treatment of idiopathic pulmonary fibrosis) fell inside the scope of the patent, there were no validity issues raised and the patent proprietor had acted without undue delay in bringing the PI proceedings. The LD decided that there was no imminent infringement.

The first instance decision was appealed and the UPC Court of Appeal (CoA) issued its decision Boehringer Ingelheim v. Zentiva in August 2025. We reported on this in UPC Weekly 2025 w33. Overturning the decision at first instance, the CoA decided that infringement was imminent, and therefore that a PI should be granted. Zentiva had applied for obtained a MA for their generic product. This in itself did not amount to an imminent infringement. However, Zentiva had gone further and had completed national procedures for pricing and reimbursement. This meant that Zentiva had, in the view of the CoA, “set the stage” for infringement. This tipped the balance in favour of awarding the PI to Boehringer Ingelheim.

The UPC’s approach to PIs more generally was nicely summarised in a decision of The Hague Local Division in Cilag v. Rivolution, where the key question for the court was whether there was the required urgency to grant a PI. As we explained in UPC Weekly 2025 w36, the facts of the case showed that Cilag had delayed for too long before requesting a PI, which was therefore refused.

Jurisdiction – how long is the long arm?

Early in 2025 the Düsseldorf Local Division issued its decision in Fujifilm v. Kodak, deciding in principle that the UPC would have jurisdiction over a defendant domiciled in Germany in relation to infringement of the UK part of a European patent (see UPC Weekly 2025 w5). This got a bit ahead of the landmark CJEU decision BSH v. Electrolux, which came out shortly afterwards and explained how long arm jurisdiction for an EU national court should work in patent litigation cases, and how the court should deal with a counterclaim that the relevant patent was invalid.

The key point in Fujifilm and also in BSH was that the defendants were domiciled in the same country as the location of the court. So, with respect to the UPC, this simply confirmed that the court has so-called “long arm” jurisdiction over “its” defendants.

Later in the year came UPC Weekly 2025 w34. This reported on Dyson v. Dreame, where there were four defendants, three based inside the UPC and one outside both the UPC and the EU (Hong Kong). The Hamburg Local Division decided that it had jurisdiction over the Hong Kong defendant in relation to alleged infringements of the Spanish part of the European patent. It was relevant that one of the UPC-domiciled defendants served as an “anchor” defendant, which had a close connection with the Hong Kong defendant, to draw it into the UPC’s jurisdiction. This case extended the boundaries of the jurisdiction of the UPC beyond the Fujifilm case (and therefore the BSH case too).

Looking forward into 2026 we can expect to see several other cases reach decisions on questions of jurisdiction, including some cases with large numbers of allegedly interconnected defendants domiciled in different countries and with activities in many different EPC states. We may also see the CoA weigh in on the first instance cases decided above.

Inventive step – a new framework

This was the main UPC event of 2025. The two different panels of the CoA issued two coordinated decisions that centred on inventive step: Amgen v. Sanofi / Regeneron in the field of antibodies for therapeutic applications, and Edwards v. Meril, relating to prosthetic heart valves. We summarised these decisions in UPC Weekly 2025 w48, comparing the outcome on appeal with the respective first instance decisions. As it happened, both appeal decisions resulted in a positive finding on inventive step.

The following week, we were lucky to see three first instance decisions all tackling inventive step with the same CoA framework, reviewed in UPC Weekly 2025 w49. Given the wordiness of the guidance in the CoA decisions, we thought it might be helpful to try to distil the framework down as far as possible. Nothing should substitute going back to the source text of the decisions to understand the CoA approach fully. But if you strip it back, there are basically three parts to the test, with lots of guidance to provide colour for their application. Our abbreviated version for these steps is set out below, with some additional guidance points set out to assist the final step.

1. Establish the objective problem. Be holistic. What does the invention add to the state of the art? The objective problem should not contain pointers to the claimed solution.

2. Find a realistic starting point in the relevant field of technology. There can be more than one realistic starting point.

3. Starting from the realistic starting point and wishing to solve the objective problem, would the skilled person have arrived at the claimed solution?

a. The skilled person has no inventive skills and no imagination.

b. The skilled person requires a pointer or motivation that directs them to implement a next step in the direction of the claimed invention from the realistic starting point.

c. A claimed solution is obvious when the skilled person would take the next step and arrive at the claimed solution:

i. prompted by the pointer or as a matter of routine and/or

ii. in expectation of finding an envisaged solution to the technical problem, when the results of the next step were clearly predictable or where there was a reasonable expectation of success. 

d. A non-obvious alternative to solutions known in the prior art can be inventive. An improvement is not required.

In Amgen, the claims were inventive because the skilled person did not have a reasonable expectation of success to take the required next step from the realistic starting point, to reach the invention.

In Edwards, inventive step was found to be present because there was no motivation for the skilled person to make the required changes to the different realistic starting points considered, whether or not the claimed invention provided an improvement over the prior art.

The big benefit of the CoA guidance is certainty – the framework for the assessment of inventive step at the UPC is now fixed. These CoA decisions resolve a significant uncertainty for prospective parties to UPC litigation and iron out any perceived differences between different UPC divisions in their assessment of patent validity. This increases the predictability of the likely outcome of UPC disputes, helping parties to assess the validity of European patents for FTO purposes, without resorting to litigation.

Roll on 2026 … !