UPC Weekly - CD Milan revokes remdesivir COVID-19 patent – and says the skilled person has no “fear of failure”

2026 Week 20

Remember the early days of the COVID-19 pandemic and the rapid search for drugs with antiviral activity against SARS-CoV-2? The Chinese Academy of Military Medical Sciences’ (AMMS’s) EP 3854403 B1 covers a drug called remdesivir for use in treating a disease or an infection caused by SARS-CoV-2. It has a 21 January 2020 priority date, shortly after SARS-CoV-2 was sequenced but before the WHO declared a Public Health Emergency of International Concern.

This speedily-filed patent was speedily dealt with at the UPC. The UPC Milan Central Division (CD) revoked it in less than 11 months, outpacing a parallel EPO opposition and illustrating how quickly a UPC revocation action can proceed – particularly when the claimant acts rapidly at grant:

• Gilead filed its UPC revocation action on the date of grant (18 June 2025) and obtained a first-instance decision on 4 May 2026 – well within 11 months.
• Gilead filed its opposition at the end of the nine-month opposition period (18 March 2026). Although the proceedings have been accelerated, the deadline for AMMS filing its first response to the opposition does not expire until 30 June 2026.

In Gilead v. AMMS (4 May 2026), the CD revoked the patent for lack of inventive step, with some interesting comments on the skilled person and how reasonable expectation of success fits within the UPC’s holistic ‘next-step’ approach to assessing inventive step.

Technical background

We’ll never forget that SARS-CoV-2 is a coronavirus. So are SARS-CoV-1 and MERS-CoV, which caused the SARS and MERS outbreaks.

Coronaviruses use an enzyme called RNA-dependent RNA polymerase (RdRP) to copy their genetic material. This means that RdRP is critical for viral replication.

RdRP works by polymerising ‘natural nucleotides’ (the building blocks of genetic material) into a strand of RNA. But RdRP can be tricked into incorporating similarly-shaped ‘nucleotide analogs’ like remdesivir. This halts the RdRp enzyme, blocking viral replication.

A familiar inventive step analysis 

Following the UPC’s holistic next-step approach set out in Amgen v Sanofi / Regeneron, the CD first established the objective problem that the patent aims to solve, from the perspective of the skilled person reading the patent at its relevant date (21 January 2020). The CD settled on “finding a drug with antiviral activity against SARS-CoV-2, to be used for the treatment of a disease caused by the infection thereof”. The CD were careful to not overstate the objective problem as the “finding of a cure for the pandemic or the disease”, seemingly because the patent provides only in vitro data for remdesivir (in cells infected with SARS-CoV-2). This modest formulation of the objective problem becomes important when considering whether the skilled person had a reasonable expectation of successfully solving it.

The reasoning supporting the CD’s decision of lack of inventive step was straightforward. Remdesivir was a broad-spectrum antiviral, which mouse and in vitro models had shown to have antiviral activity against other coronaviruses (e.g. SARS-CoV-1 and MERS-CoV). Documents stating that remdesivir was “likely to” be effective against the newly identified coronavirus in Wuhan were realistic starting points. The CD considered that the skilled person would have turned from these documents to an article disclosing that (i) remdesivir targets RdRP and (ii) RdRP is highly preserved among all coronaviruses. The newly published genome showed that SARS-CoV-2 was highly similar to SARS-CoV-1 and MERS-CoV, including in the region encoding RdRp. The obvious next step was to confirm the effectiveness of remdesivir against SARS-CoV-2 by conducting a routine in vitro experiment.

Who is the skilled person?

An eye-catching part of the decision discusses the skilled person. Much of the discussion was consistent with the approach of the EPO: the skilled person is a notional figure with no inventive skills and no imagination, who cannot be identified with any real-world person. They are cautious and conservative and would never go against an established prejudice, but would readily seek appropriate, obvious changes involving no risks or only calculable risks.

However, this caution only goes so far. The CD stated that the skilled person “does not display a fear of failure”. At first glance, this characterisation of the skilled person might seem bolder than we are accustomed to at the EPO. But it seems to be the CD’s way of rejecting AMMS’s attempt to import into the skilled person the subjective and highly cautious attitude of their expert (that small sequence variations between SARS-CoV-2 and other coronaviruses in the region encoding RdRP led to major uncertainty as to whether remdesivir would work against SARS-CoV-2). In other words, as risk and doubt are part of scientific progress, the skilled person would not reject a scientifically valid suggestion that is expected to work merely due to subjective concerns about possible failure¹. This seems consistent with EPO case law saying that the skilled person would prefer to verify whether a potential solution worked, rather than abandon the project because success was not certain.

Also notable are statements that the skilled person is not required to possess (or disregard) a distinct affiliation to a specific company, that such arguments are not admissible, and that such an affiliation would not actually impact the skilled person’s knowledge². This was prompted by AMMS’s arguments that some documents were “promotional” and “commercial in nature”, originating from circles heavily influenced by Gilead (who, at the time, were developing remdesivir as an antiviral). Pragmatically, the CD also stated that mere speculation about industry-funded research cannot undermine the credibility of researchers (who risk their own credibility if they publish unsubstantiated opinions prone to promotional or commercial interests)³.

The CD considered that the skilled person “prioritizes the most recent and detailed sources, as scientific progress requires ongoing reassessment of previous knowledge”, and was particularly interested in documents that “reflect the culmination of scientific research”. The heavy emphasis on the age of documents is understandable here, as knowledge of the new coronavirus changed by the day, but may not translate well to other settings. As this is a first instance decision, these statements alone do not mean that the UPC will go beyond the context-specific approach applied by the EPO.

Reasonable expectation of success

In the EPO’s problem-solution approach, we sometimes ask whether it was obvious for the skilled person to try a suggested approach with a reasonable expectation of success – ‘success’ being to solve the objective technical problem.

The CD explained that the concept of a reasonable expectation of success was consistent with the UPC’s holistic next-step approach, since it considers “what the invention objectively adds to the state of the art, thus also considering the inventive concept underlying the invention”.

In the present case, the CD considered the skilled person to have a high expectation of success because there was a clear and compelling recommendation for remdesivir, the tests to verify its antiviral activity were routine, and minor changes in the RdRP between coronaviruses were within the realm of foreseeable and acceptable risk. It’s important to remember the modest formulation of the objective problem: it was less complex to find a drug with antiviral activity against SARS-CoV-2 than a ‘cure’ for the pandemic or the disease.

Different approaches to inventive step 

Seemingly prompted by arguments from both Gilead and AMMS, the CD identified a “well known” problem with the EPO’s problem-solution approach: that hindsight might influence the selection of technical features when defining the difference between the prior art and the claimed solution, and possibly distort the inventive step analysis. This re-iterates the UPC’s now-clear preference for a more holistic assessment – although, as usual, the CD note the two approaches are supposed to lead to the same outcome.

It’ll be interesting to see how the EPO opposition proceedings – which should use the problem-solution approach, perhaps with a more ambitious problem reflecting the treatment required by medical use claims – unfold. For now, we just have this quick and logical first instance decision from the UPC, which neatly incorporates key principles from EPO case law into the UPC’s framework for assessing inventive step.

References:

1. Headnote 2.

2. 48 of the decision

3. 133 and 135 of the decision. 

This blog was co-authored by Lucy White and Eliot Ward.