UPC Weekly brings you timely, thoughtful and joined-up insight into the evolving landscape of patent litigation at the Unified Patent Court.
2026 Week 27
In UPC Weekly 2026 w5, we reported on the Paris Local Division's refusal to grant Guardant Health a preliminary injunction (PI) against Sophia Genetics based on three European patents. The LD had found validity problems with two of the patents and insufficient evidence of infringement for all three. But the LD had considered the three months that Guardant took to apply for the PI, after learning about the alleged infringement, to be reasonable given the complexity of the technology and the number of patents at issue.
Guardant appealed on just one of the patents, EP 3443066 ("EP'066"), covering detection of colorectal, ovarian, lung and pancreatic cancer by sequencing circulating cell-free DNA.
In Guardant v. Sophia (UPC CoA, 2 July 2026), the UPC Court of Appeal (CoA) upheld the LD’s decision to refuse the PI, but for markedly different reasons. The CoA reversed the LD's finding that EP’066 was probably invalid for added matter, but also reversed the LD’s finding that the three months that Guardant took to prepare the application for provisional measures was reasonable. The CoA took a deep dive into the witness statements, timelines of parallel UK proceedings and the availability of evidence of infringing acts, and decided that the time between May and August 2025 was unreasonably long.
No added matter – but it didn’t matter here
The CoA disagreed with the LD’s finding that EP'066 was likely invalid for added matter. The LD did not accept the combination of the 25-gene panel with the four cancer types listed in claim 1 (colorectal, ovarian, lung, pancreatic) as directly and unambiguously derivable from the application as filed. The CoA disagreed. It noted that Example 2 in the application as filed described a 25-gene panel for detecting precisely those four cancers.
Sofia argued that Guardant’s reliance on Example 2 was a late-filed argument, but the CoA dismissed this, saying:
There is no reason why example 2 in the application should be disregarded. Whether there is added matter is a question of law. As the test is whether the relevant claims are disclosed in the application as a whole, the Court is permitted to consider the entire document (Paragraph 98 of the Reasons, referring to Amgen v. Sanofi).
This seems to show that indicating particular disclosures in the underlying patent application late on in proceedings is no bar to the UPC considering those disclosures when evaluating added matter. It may be a new argument, but is not new evidence.
This reversal on validity is a meaningful win for Guardant in the long term because it puts them in a better position in the full merits proceedings. But it was not enough to secure the PI here.
Urgency is key – and so are the facts
The CoA's approach to urgency diverges significantly from that of the LD, not just in outcome, but in methodology.
When the LD found in Guardant's favour on urgency, it identified 27 May 2025 as the ‘trigger date’. This was the date when Guardant sent a warning letter to Sofia about commercial activity regarding its “MSK-DDM” test, which Guardant considered to fall inside several of its UK patents. Following Sofia’s reply on 20 June, Guardant initiated UK litigation on 14 July 2025 and filed its more detailed reasoning supporting the UK action on 18 August 2025.
The LD accepted Guardant's explanation that three months of preparation was reasonable given the number of patents asserted and the technical complexity of the case.
The CoA agreed with the legal framework, that the clock starts running when the applicant became aware, or should have become aware, of facts and evidence sufficient to file a PI application with a reasonable prospect of success. But the CoA looked into the facts and circumstances around the ‘delay’ in considerably more detail.
Guardant's warning letter of 27 May 2025 mentioned EP'533 and EP'073. EP'066 (the patent at issue on appeal) was not mentioned in the warning letter.
Guardant argued at the CoA hearing that this justified taking a later trigger date for EP'066. The CoA rejected this: if Guardant had done the analysis to write the warning letter for the other patents, it must be taken to have had access to the same underlying documents and technical information relevant to EP'066 at the same time. In the CoA's words, a patent holder must not "turn a blind eye" to the fact that a document indicating infringement of some patents also indicates infringement of others:
1. A patent holder is not obliged to assert all patents in one application for provisional measures. If a patent holder has the necessary information for filing an application for provisional measures regarding some, but not all patents, delaying the filing of the Application for provisional measures until it has the necessary information regarding all allegedly infringed patents may be regarded as an unreasonable delay.
2. If a patent holder is aware, on the basis of a document, that one or more of its patents has been infringed, it must not turn a blind eye to the fact that the document also indicates the infringement of one or more of its other patents.
In view of the evidence, the CoA took the trigger date to be 1 May 2025, not 27 May as taken by the LD, because of the perceived implication of Guardant's IP Vice-President’s witness statement. Their statement that “no employee with responsibility for evaluating patent infringement was aware of [Sophia's activities] before May 2025” (emphasis added by the court) was taken to mean that Guardant knew all the relevant details of the attacked embodiments in May 2025 (Paragraph 120 of the Reasons). At Paragraph 120 of the Reasons, the court said “it must be assumed that the technical details relevant for the patent at issue were known by 1 May 2025”.
On the question of how long Guardant could have taken to confirm infringing acts within UPC territory, the CoA was quite strict. Guardant said that it only became aware of Sophia's commercial activities within UPC territory on 13 August 2025. But the CoA found this to be inconsistent with publicly available information.
The CoA noted that Guardant's 27 May 2025 warning letter already referenced information about commercial activities inside the UPC territories, such as a financial report from Sophia showing new customers in Belgium and France. The CoA concluded that this information could have been gathered within two weeks of 1 May 2025, meaning that Guardant should have had everything it needed by 15 May 2025.
The court was also strict about the evidential burden on patentees to show how much time was required for conducting technical analysis:
It is for the patent holder who files an application for provisional measures to explain, in a substantiated manner, where applicable, which further analysis or tests were required, for which claim(s) or claim features it was necessary to conduct them, how long this took and why this time was reasonable.
The burden of proof is therefore on the patentee to show the details of when they were first aware of the infringement inside the UPC territory, the work done to substantiate that infringement, the work done to prepare the application for the PI, and why the time taken to file the PI application indicates that the patentee has not unduly delayed.
Conclusion
The LD had accepted three months as a reasonable period to prepare a multi-patent PI application involving complex technology. The CoA rejected this not necessarily because three months is inherently too long, but because the three months had not been adequately accounted for.
When seeking a PI from the UPC, take note; the evidential burden on urgency is real. It requires detailed, documented explanation of the investigative and drafting timeline. Asserting multiple patents against the same infringing product does not necessarily give more time – the clock for each patent starts running independently, from when the patentee had (or should have had, with suitable diligence) sufficient information for that patent.
.png)
.png)