This browser is not actively supported anymore. For the best passle experience, we strongly recommend you upgrade your browser.
| 14 minutes read

A detailed analysis of the first substantive UPC Court of Appeal decision – lifting a preliminary injunction against NanoString

NanoString Technologies Inc.; NanoString Technologies Germany GmbH; NanoString Technologies Netherlands B.V.  v President and Fellows of Harvard College; 10x Genomics, Inc. (UPC_CoA_335/2023)


The first ever substantive UPC Court of Appeal decision was issued on 26 February 2024, relating to the preliminary injunction (PI) granted to 10x Genomics (Genomics) against NanoString Technologies (NanoString) concerning Genomics’ patent (EP 4108782) for a method of detecting analytes such as RNA or proteins in cell or tissue samples. The Munich Local Division issued the PI in September 2023 and the appeal was heard in Luxembourg in December 2023. The relatively swift timeline of five months from the initial PI decision to appeal is promising for the UPC’s overarching aim of expediency, but we will have to wait and see if the Court of Appeal can maintain this level of efficiency as its workload increases, particularly with more substantive appeals (including main action appeals). 

The expected timeframe for first instance main action decisions in the UPC still stands at 12 to 14 months, so the ability to obtain a PI remains important to patent proprietors looking for more rapid relief pending those decisions. 

The Court of Appeal’s PI decision in the Genomics case assesses specific technical knowledge and practices within the field of analyte detection and offers a contrasting legal analysis to that given by the Munich Local Division at first instance, particularly in relation to the assessment of validity. This decision also marks the first occasion on which the UPC Court of Appeal has issued a detailed judgment on a PI application, and therefore gives some welcome clarity on the correct approach to be taken in the assessment of such applications. Below we provide a detailed summary of the Court of Appeal’s reasoning. However, one of the key takeaways relates to the respective thresholds for infringement and validity that must be met for a PI to be granted. In short, the Court must consider it at least more likely than not that the applicant is entitled to initiate proceedings (which includes that it is more likely than not that the patent is valid) and that the patent is infringed.

The Court of Appeal’s findings will also be relevant to the drafting of protective letters, as these should seek to address the thresholds of infringement and/or validity necessary to prevent a PI being granted (i.e. show it is more likely than not that the patent is invalid and/or not infringed).

Also of note in this case is that the Court of Appeal issued a second order on the same day, rejecting Genomics’ request to stay the proceedings due to NanoString’s chapter 11 (insolvency) filing in the US. This is briefly summarised at the end of this article. 

More generally with regard to the role of the Court of Appeal, and given the divergence between Local Divisions in some of the early cases (for example the conflicting decisions relating to file wrapper estoppel by the Munich Local Division in SES-imagotag v Hanshow Technology and the Düsseldorf Local Division in Ortovox v Mammut), it is hoped that through further appeals and well-reasoned judgments, the Court of Appeal will iron out such inconsistencies with a view to more uniform (and predictable) decision-making across the UPC. 

The patent

Genomics’ patent, EP 4 108 782 (EP 782), relates to the detection of analytes in a cell or tissue sample mounted on a solid support (i.e. in situ). In particular, to a multiplexed biological assay and readout, in which a multitude of detection reagents comprising one or more probes are applied to the sample, allowing the detection reagents to bind target molecules or analytes in the sample, which can then be identified in a temporally-sequential manner. The patented method thereby enables the simultaneous in situ analysis of a multitude of analytes (e.g., RNA or proteins) in a cell or tissue sample. This is desirable as the test samples are often precious and those analysing them either do not know in advance precisely what to look for, or they must extract the most information they can from a single sample. 

The first instance PI decision

At first instance, the Munich Local Division issued a PI on 19 September 2023 in relation to alleged infringement by NanoString of EP 782. Separately, in October 2023, the same panel denied a PI against NanoString in relation to Genomics’ patent EP 2 794 928, the parent application of EP 782.  

In the EP 782 case subject to this appeal, the Local Division began by finding that it had jurisdiction and that the PI application was admissible. It then considered in some detail questions of validity and infringement, finding with sufficient certainty for the purposes of preliminary relief that EP 782 was valid and infringed. Of note in the assessment of validity was that the Local Division seemingly adopted an EPO ‘problem-and-solution’ approach to assessing inventive step, starting from the document the Local Division had identified as the closest prior art. As discussed below, the Court of Appeal appears to have taken a different approach. 

Following its assessment of validity and infringement, the Local Division went on to consider the necessity of ordering provisional measures. In terms of urgency, and noting that EP 782 is a unitary patent, it concluded that enforcement measures before national courts are not equivalent to those before the UPC. Therefore, whether there had been unreasonable delay in this case should be evaluated from 1 June 2023, the day on which claims before the UPC could first be filed (and on which date the PI request was made). Regarding harm, it concluded that NanoString’s market entry would cause Genomics significant, long-lasting harm. 

Finally, the Local Division balanced the parties’ interests in granting/rejecting the PI and found that Genomics’ interests were more important than NanoString’s interests. A crucial factor in this assessment was the Local Division’s finding that there was a high likelihood of the patent being found valid and infringed at trial. A PI was therefore granted to Genomics. 

Notably, the PI was ordered without any requirement for Genomics to provide security. The Local Division was satisfied based on the economic condition of Genomics and US enforcement law that there were no indications that there would be difficulties enforcing a claim for compensation of NanoString if the PI was later held to be wrongly granted. The Local Division also ordered that NanoString make a penalty payment of up to €250,000 for each individual breach of the PI order. On 5 December 2023 the Local Division ordered NanoString to pay a penalty of €100,000 for its failure to comply with the PI order in a timely manner as it failed to insert notices on all relevant webpages stating that the alleged infringing CosMx Spatial Molecular Imager and corresponding decoder probes must not be used for RNA detection in any UPC member states.

The Appeal

On appeal to the Court of Appeal in Luxembourg, NanoString raised a number of grounds of appeal. These were that the Local Division: 

a) had incorrectly assumed jurisdiction; 

b) had erred in its narrow interpretation of the claims; 

c) had erred in assuming that EP 782 was most likely valid; 

d) had failed to consider that the PI application was inadmissible as it did not fulfil the requirements of R. 206.2(c), (d) and (e) of the Rules of Procedure;

e) had misapplied both the facts and antitrust/competition law, which in turn meant that the patentee had an obligation to grant NanoString a licence; and

f) did not take proper account of facts on irreparable harm, proportionality and urgency that counted against the grant of a PI.

In the alternative, NanoString argued that they should have been allowed to continue the acts complained of on provision of security, and in the further alternative, that the PI applicants (Genomics) should have to provide security (in the amount of €20 million) to secure the PI if maintained.

Jurisdiction (ground (a) of the appeal)

On jurisdiction, the Court of Appeal held that the Munich Local Division did have jurisdiction on the basis that the allegedly infringing embodiments had been offered in Germany. However, the Court of Appeal did not expressly deal with the argument raised by NanoString and rejected by the Local Division that the UPC should decline jurisdiction (at least in respect of the German territory) given that Genomics had already sought a national PI under the parent patent in Germany which was issued by the District Court of Munich. It is not clear from the judgment whether this argument was maintained in the appeal. 

Construction (ground (b) of the appeal)

On claim construction, the Court of Appeal departed from the conclusions of the Local Division on a number of specific issues. However, given its decision on validity (see below), it did not deal with the overall question of whether infringement was likely as a result.  

Of interest are the conclusions on claim interpretation. The Court confirmed that “the patent claim is not only the starting point, but the decisive basis for determining the scope of protection”, however this does not limit the claim construction to the strict literal meaning of the wording and so “the description and the drawings must always be used as explanatory aids for the interpretation of the patent claim and not only to resolve any ambiguities in the patent claim”. In its decision, the Court of Appeal made references to many parts of the description of the patent when justifying its interpretation of claim features. Such an approach is in accordance with Art. 69 EPC (in conjunction with the Protocol on the Interpretation of Art. 69 EPC), with wording from the Court of Appeal on the balance to be struck mirroring that of the Protocol on Interpretation: “In applying these principles, the aim is to combine adequate protection for the patent proprietor with sufficient legal certainty for third parties.” This approach is also similar to the concept of ‘purposive construction’ followed by the English Courts.

Validity (ground (c) of the appeal)

The most interesting part of the judgment is the approach to validity. In relation to the standard of proof, the Court of Appeal agreed with the Local Division that as an order for provisional measures is by way of summary proceedings, in which the opportunities for the parties to present facts and evidence are limited, the standard of proof for a PI must not be set too high. This is particularly important if delays associated with proceedings on the merits (i.e. to main action proceedings without a PI) would cause irreparable harm to the patentee. In this finding the Court referred in part to the CJEU decision in Phoenix Contact (C-44/21). The Court also confirmed that the standard of proof should not be set too low in order to prevent the defendant from being harmed by an order that is subsequently revoked. 

Rule 211.2 of the Rules of Procedure in conjunction with Art.62(4) UPCA provide that the Court may invite the PI applicant to provide reasonable evidence to satisfy the Court to a sufficient degree of certainty that the applicant is entitled to institute proceedings, that the patent is valid and is being infringed (or infringement is imminent). According to the Court of Appeal, such a degree of certainty requires the Court to consider it at least more likely than not that the applicant is entitled to initiate proceedings (which includes that it is more likely than not that the patent is valid) and that the patent is infringed

As to which party bears the burden of proof, the Court of Appeal held that this lies with the applicant in relation to entitlement to initiate proceedings and infringement, as well as other circumstances supporting the applicant’s requests. Except in the situation of an ex parte PI request, the burden regarding (lack of) validity and other circumstances supporting the respondent’s request lie with the respondent. The Court of Appeal noted that this is aligned with the position in main action proceedings.

Applying this standard of proof, the Court of Appeal disagreed with the Munich Local Division on validity, holding that on the balance of probability, it was more likely than not that the subject matter of claim 1 would prove to be invalid for obviousness. In this finding, the Court of Appeal conducted a detailed analysis of its own, sitting with two Technically Qualified Judges on the panel.

As stated above, at first instance the Munich Local Division appeared to adopt the ‘problem-and-solution’ approach of the EPO to assessing inventive step, starting from an identification of the closest prior art, usually being a prior art document that discloses subject matter directed to the same purpose or with the same aim as the claimed invention and sharing the main technical features (and/or requiring the fewest structural changes). Applying this approach, the Local Division had taken the view that one of the prior art citations advanced by the respondents (D6) was not a realistic starting point, let alone the closest prior art. In contrast, the Court of Appeal did not apply a formalistic ‘closest prior art’ analysis and held that D6 would have been of interest to a person skilled in the art seeking to develop certain methods. 

The Court of Appeal’s decision placed emphasis on the idea that a skilled person, motivated by the need for in situ analysis, would find it obvious to adapt the method in D6. Contrary to the patent claims, D6 was primarily directed to an in vitro application. However, the Court of Appeal found that it also hinted at in situ applications, particularly in one footnote to another publication. Similar publications cited by the respondent also demonstrated the feasibility of this concept. The Court considered that in situ analysis issues like molecular crowding and autofluorescence were hurdles a skilled person would expect and be equipped to tackle. These problems did not make an in situ adaptation an unpredictable leap. The patent itself offered no unique solutions to in situ problems, which the Court of Appeal found would strengthen the argument that the adaptation would be within the routine skill of someone in the art. 

The Court of Appeal’s approach was therefore in contrast to that of the Local Division at first instance, as there was no apparent application of the EPO’s problem-and solution approach. However, the Court of Appeal did not explicitly comment on the approach taken to validity at first instance and this is, of course, only a decision on a PI application. No final decision on validity has been issued by the UPC at either first instance or on appeal as the first trials will not take place until later this year. As such, we will need to wait for further decisions and appeals on inventive step for clarity on what the correct approach to assessing inventive step will be for the UPC. 

Auxiliary Requests

An auxiliary request to amend the claims was submitted by Genomics at the first instance hearing and they sought to rely on this at appeal. 

The Court of Appeal found that the amendments proposed to the patent in an auxiliary request would not solve the problem of obviousness. In short, limiting the patent to immunohistochemistry and fluorescence in situ hybridization would not address the core problem – the court considered that these were obvious and established techniques for a skilled person seeking to do in situ analysis. On this basis, the Court found it more likely than not that the patent would be revoked in the future, making a PI inappropriate. 

An interesting point was raised by NanoString in relation to the admissibility of the auxiliary request. As it was only filed at the first instance hearing (and not before) they argued that it was inadmissible under R. 221.2 of the Rules of Procedure because they had no opportunity in the proceedings before the Court of First Instance to comment on the validity of the request. The Court of Appeal found that the fact the Respondents did not present their arguments for obviousness relating to the auxiliary request until the oral hearing in the appeal proceedings did not violate R. 221. In view of the fact this was the first appeal case in which auxiliary requests and R. 221 were relevant, the Court had instructed the parties at the hearing to comment on the validity of the auxiliary request. These arguments largely matched those of the main request and so the Court found that NanoStrings’s defence was not duly disadvantaged by this.

Admissibility of the PI application (ground (d) of the appeal)

NanoString had argued that the application for the PI was deficient in that it did not fulfil the requirements of R. 206.2(c), (d) and (e) of the Rules of Procedure. The Court of Appeal dismissed this, drawing a distinction between R. 206.2(a) and R. 206(2)(b) to (e). The Court pointed out that the requirements for the application for provisional measures set out in R. 206.2(a) in conjunction with Rules 13.1(a) to (i) are of a formal nature, whereas the requirements set out in R. 206.2(b) to (e) relate to the content of the application for provisional measures. Accordingly, compliance with the requirements of R. 206.2(b) to (e) is not checked by the Registry and no time limit is set in the event of non-compliance, which can lead to a decision by default under R. 355.1(a). Rather, compliance with the requirements set out in R. 206.2(b) to (e) concerns the merits of the application for provisional measures, the examination of which is the responsibility of the judge and must be considered by the judge when making orders under Rules 209, 211 and 212. The Court noted that, in the context of orders to be made by a judge in the exercise of their discretion, non-compliance with the requirements set out in R. 206.2(b) to (e) may be to the detriment of the Applicant, however the alleged breach of R. 206.2(c), (d) and (e) would not render the application inadmissible.

Other Matters

Following its finding on patent invalidity, and thus rejection of the PI, the Court of Appeal did not deal with all the other issues raised in the appeal, including grounds (e) and (f) above: the EU/antitrust issue and consideration of irreparable harm, or the provision of security. Some questions therefore remain unanswered. 

The Second Order – Chapter 11 (insolvency) filing

Also of interest is the second decision by the Court of Appeal, the order for which was issued on the same day. This related to NanoString’s filing for Chapter 11 bankruptcy protection in the US and request for a stay of the proceedings. Crucially this request was made after the oral hearing for the main PI appeal took place in December 2023. Genomics agreed to the stay, however the Court of Appeal held that a stay would not be granted. The Court concluded that the Rules of Procedure on stays were designed for situations where insolvency was declared before crucial legal proceedings were completed. In this case, all major procedural steps (including the oral hearing) had been completed before the bankruptcy filing. The Court considered that by suspending the case now, this would cause unnecessary delay and expense without impacting the bankruptcy proceedings themselves and therefore in the interest of efficiency and fairness, the Court of Appeal should deliver its PI decision as planned. The Court also commented that several other countries' legal systems have similar provisions, allowing ongoing court cases to reach a conclusion even if a party declares bankruptcy near their conclusion.


biotech, life sciences, patent litigation, upc