Executive Summary: Just over two years of the Unified Patent Court – Legal Trends and Performance Overview

Introduction

The Unified Patent Court (UPC) opened its doors on June 1, 2023, marking a new era for patent litigation in Europe. Two years on, an emerging body of case law sheds light on how the UPC interprets patentability and enforces patent rights across member states. This report analyzes four major trends in the UPC’s early jurisprudence: 

(1) its approach to inventive step (obviousness) and use of the European Patent Office (EPO)’s problem-solution framework, 

(2) the granting of preliminary and permanent injunctions (with comparisons to German and UK practices), 

(3) the success rates of patentees versus defendants in infringement and validity disputes, and 

(4) the UPC’s stance on anti-anti-suit injunctions (AASIs) in global patent battles. Comparisons are drawn with EPO Boards of Appeal jurisprudence and national courts in Germany (DE) and the UK (GB) to highlight consistency, deviations, and procedural outcomes.

1 – Inventive Step at the UPC: Problem-Solution Approach or New Path?

UPC vs. EPO Approach: The UPC is bound by the European Patent Convention’s standards (Art. 56 EPC) for inventive step, but early decisions reveal nuanced departures from the EPO’s classic problem-solution approach. The EPO’s Boards of Appeal require identifying the “closest prior art” and formulating a specific objective technical problem based on the differences between that prior art and the claimed invention. In contrast, some early UPC panels have adopted a more flexible framework, focusing on a “realistic starting point” rather than the single most promising prior art. For example, in a landmark revocation case Sanofi v. Amgen (UPC Central Division Munich, July 16, 2024), the court outlined a four-step obviousness test that deviated from the strict EPO approach:

a) underlying problem to be identified together with

b) realistic starting point – the UPC has been open to multiple starting references, as long as they relate to a similar problem,

c) motivation or incentive: assess whether, starting from a given prior art, the skilled person would have motivation to adopt the claimed solution as the next logical step, and

d) is there a reasonable expectation of success?

Notably, the court explicitly did not require pinpointing one closest prior art reference – multiple “realistic” starting points were acceptable, and the traditional EPO notion of the one most promising prior art was downplayed.

Defining the Technical Problem: Similarly, in NJOY v. Juul Labs (UPC Central Division Paris, Nov. 5, 2024), the court emphasized an “objective” assessment of inventive step. If the patent did not clearly state the problem it solves, the UPC deduced it from the interplay of claim features, effectively defining a single underlying technical problem from the patent itself. This again contrasts with the EPO’s practice of potentially reformulating the problem for each closest prior art. The Paris panel also affirmed that inventive step can be attacked from any element of the prior art – though for efficiency the debate may focus on key prior art chosen by the parties. In other words, the UPC allowed obviousness arguments from several different references as starting points (as presented by the claimant), rather than insisting on one EPO-style starting document.

Consistency with EPO Jurisprudence: Despite these early deviations, there are signs of convergence. In 10x Genomics v. Nanostring (UPC Court of Appeal, Feb. 26, 2024), which involved a preliminary injunction, the appellate judges underscored the importance of aligning with EPO practice. The Court of Appeal stressed that claims must be interpreted through the eyes of the skilled person (citing Art. 69 EPC on claim construction), and ultimately found the patent likely obvious over a known method adapted to the claimed setting. The impact of that decision on the EPO can be seen for example in the Enlarged Board of appeal decision G 1/24 which itself has been widely reported elsewhere, but provide the common sense approach of claim interpretation in light of the full specification rather than merely relying on the language of the claims.

Most significantly, a recent decision of the UPC’s Munich Local Division (Edwards Lifesciences v. Meril, April 4, 2025) explicitly embraced the EPO’s problem-solution approach. In its headnote, the Munich panel announced that the problem-solution approach “shall primarily be applied as a tool wherever feasible, to enhance legal certainty and align UPC jurisprudence with that of the EPO Boards of Appeal”. This suggests an emerging consensus that strict adherence to the well-tested EPO framework is desirable to avoid divergent standards. Indeed, the Edwards decision, which upheld a heart valve patent as inventive, indicates the UPC’s willingness to course-correct toward EPO doctrine for consistency.

Takeaway: The UPC’s inventive step case law is still maturing and not entirely uniform. Some first-instance decisions trialed a more flexible obviousness inquiry (multiple starting points, fixed problem from the patent, etc.), differing from the formalistic EPO method. However, there is a clear trend – championed by the Court of Appeal and at least one major local division – to apply the traditional problem-solution approach as the primary tool, in the interest of harmonization. In practice, both the UPC and EPO require a showing of some teaching or “prompt” in the prior art and a reasonable expectation of success to find a claim obvious.

The nuanced differences (like what counts as a proper starting prior art, or how the problem is formulated) will likely be refined by future Court of Appeal judgments. For now, patent litigants should be prepared for either approach: expect UPC judges to invoke EPO case law heavily in inventive step analysis, but also be ready for arguments on what is a “realistic” prior art starting point if no single closest prior art stands out.

2 – Injunction Trends: Preliminary vs. Permanent Relief

Preliminary Injunctions in the UPC – A More Proactive Stance.

One of the hallmarks of the UPC’s first two years is an apparent willingness to grant preliminary injunctions (PIs) more readily than some national courts. Under Article 62 UPC Agreement (and Rule 211 RoP), the UPC can issue interim injunctions to prevent imminent infringement, provided the claimant shows a “sufficient degree of certainty” of a valid claim on the merits.

In 10x Genomics v. Nanostring, the UPC Court of Appeal clarified this standard: a PI may be granted if it is “more likely than not” that the patent is valid and infringed – in other words, the applicant must establish a >50% likelihood of success. If the court instead finds it more likely than not that the patent will be held invalid (based on the defendant’s invalidity arguments), then a PI will be refused.

This “balance of probabilities” merits test is notably stringent in that UPC judges delve deeply into validity and infringement at the preliminary stage, effectively conducting a mini-trial on these issues. This contrasts with the UK approach (from American Cyanamid v. Ethicon), which only requires a “serious question to be tried” and focuses on balancing convenience and adequacy of damages rather than conclusively assessing validity upfront. The UPC’s approach is closer to the Continental practice (e.g. the Dutch and somewhat the German approach) of requiring a strong prima facie case on the merits, though the UPC explicitly weighs parties’ interests and any delay by the patentee as well.

Success Rate and Statistics: In its first 18 months, the UPC decided roughly two dozen PI applications. About 50–60% of these applications have been granted, indicating an interim relief grant rate slightly above a coin toss.

By comparison, preliminary injunctions in German patent courts have historically been harder to obtain, often requiring the patent to have survived opposition or validity challenges, and UK courts rarely grant PIs in patent cases due to the stringent balance-of-convenience test and the availability of quick trials.

The UPC’s roughly 58% PI grant rate reflect[s] the court’s willingness to intervene quickly, where necessary, making it an attractive forum for patentees seeking fast pan-European relief. Moreover, the UPC has shown it can act with speed: ex parte urgent PIs have been granted in a matter of days in appropriate cases (especially when a trade fair or imminent product launch threatened irreparable harm).

Key Factors – Urgency and Harm: UPC judges have been attentive to urgency. A PI applicant must act promptly once infringement is known, or risk denial for undue delay. The Munich Local Division has informally treated a filing within ~2 months of learning of infringement as a “safe harbour” for urgency in some cases (e.g. Dyson v. SharkNinja), whereas the Düsseldorf division suggested even one month may be the limit in other cases. Exceptions exist – e.g. a 12-month delay was excused in a biotech case due to time needed to gather evidence, but generally patentees are advised to move fast. The Court of Appeal’s Ortovox v. Mammut decision confirmed that the clock starts when the patentee becomes aware (or should have known) of the infringement, and while it declined to set a strict deadline, undue delay can be fatal.

On the harm element, the UPC does consider potential damage on both sides. Interestingly, the Court of Appeal noted that in patent cases, irreparable harm need not be strictly proven – the threat of lost market share or price erosion can be presumed in some scenarios, given the exclusivity conferred by a patent. Nonetheless, patentees should articulate why only an interim injunction (and not waiting for trial) will prevent harm that cannot later be compensated. Defendants, in turn, have invoked the requirement of “necessity”, arguing that if the patentee can be made whole by damages or if the product’s market impact is limited, a PI isn’t warranted.

So far, however, no categorical strictness has emerged: the UPC seems to weigh all circumstances, much like German courts’ proportionality analysis for PIs. If validity is seriously in question (e.g. newly unearthed prior art or pending EPO oppositions suggesting vulnerability), the UPC – like the German courts – tends to err on the side of refusing a PI. Conversely, a patent that has already survived opposition or has been licensed by others might bolster the case for a PI, akin to German practice.

Comparison – Germany and UK: In German courts, preliminary injunctions in patent cases were traditionally rare and subject to a high bar (the patent had to be essentially “validated” and the matter urgent). Recent German case law (e.g. OLG Munich in 2019) softened the criteria slightly, but even so, German judges require near-clear validity and urgency – ongoing opposition or close questions usually preclude a PI.

The UPC’s balance-of-probabilities standard is arguably a bit more lenient to patentees on validity (more akin to “likelihood” than absolute clarity), and the data (about half of PIs granted) suggests a higher success rate than the historically very low German rate.

Meanwhile, UK courts remain far more conservative on interim injunctions: the classic British approach is that if validity is contested and a quick trial can be had, an interim injunction is generally not granted. The UPC explicitly rejected the UK’s American Cyanamid threshold and instead conducts a detailed merits analysis early, making it a more assertive injunctive forum. In sum, the UPC has in its first two years proven to be a relatively PI-friendly forum for patent holders, provided they have a strong case and act swiftly.

Permanent Injunctions and Proportionality.

For permanent injunctions (final remedies after a finding of infringement), the UPC’s practices also show both alignment with traditional European norms and a touch of modern proportionality. Under Article 63 UPC Agreement, the court may grant a permanent injunction “to prevent the continuation of the infringement.” In virtually all cases where a patent was found valid and infringed on the merits, the UPC has granted a permanent injunction as a matter of course.

By the end of 2024, roughly 60% of infringement judgments favored the patentee, all of which resulted in injunctive relief (sometimes with minor qualifications or carve-outs). In other words, if you win on liability in the UPC, you can expect an injunction covering all infringing acts in the UPC member states, much like in national courts. There have been no reported instances in these two years of the UPC outright denying a final injunction to a prevailing patent owner on equity or public interest grounds. This mirrors practice in national courts, such as the historical German rule of automatic injunctive relief and the UK practice that a winning patentee is entitled to an injunction (barring exceptional circumstances such as public health or competition concerns, which have not arisen in early UPC cases).

However, the UPC has begun to articulate a proportionality/continuation test for permanent injunctions, influenced by modern German law and EU principles. In the first final judgment issued by the court – Franz Kaldewei GmbH v. Bette GmbH (UPC Düsseldorf Local Division, July 3, 2024) – the defendant argued that no injunction should issue because it had ceased selling the infringing product and switched to a non-infringing alternative. The UPC panel acknowledged that Article 63(1) UPCA targets injunctions at preventing “the continuation of the infringement.” The judges suggested that if an infringer can prove it has genuinely stopped all infringement and will not resume, then the need for an injunction “may indeed fall away”. In this Kaldewei case, Bette failed to meet the heavy burden of proof – evidence showed lingering sales in at least one country – so the injunction was granted in full.

But the court’s reasoning is telling: the UPC’s threshold for a permanent injunction can be higher than Germany’s traditional approach. Under German practice, any past infringement creates a presumption of future repetition (and thus an injunction) unless the defendant gives a binding cease-and-desist undertaking with a penalty clause. Simply stopping sales unilaterally is not enough in Germany to avoid an injunction. The UPC, by contrast, was open to the idea that an injunction might be unnecessary if infringement has truly been abandoned (though in practice the bar for proving this is high, and defendants would likely need to provide enforceable undertakings or other clear evidence to satisfy the court).

Beyond that scenario, the UPC has not yet had to deeply consider other proportionality arguments (such as an infringer claiming an injunction would be disproportional due to public interest or if the patent covers a minor component – issues sometimes raised under the new German Patent Act §139(1) or in U.S. cases post-eBay). It is expected that UPC courts could entertain such arguments, as the UPCA and Rules give discretion in formulating remedies. Indeed, the Kaldewei decision’s reasoning ties into fundamental concepts of equity and proportional relief, resonating with the EU Intellectual Property Enforcement Directive’s mandate that remedies be fair and equitable. But so far, the default UPC position is to grant permanent injunctions upon a finding of infringement. In the handful of final infringement rulings to date, all injunction requests were granted, sometimes accompanied by additional orders (such as product recalls, information for calculating damages, and provisional damages awards).

Comparative note: In Germany, even after the 2021 introduction of a proportionality clause, there have been virtually no cases denying a patentee an injunction outright – any exceptions are exceedingly rare and require extreme disproportional harm to the infringer or third parties. The UPC appears to be following a similar path: acknowledging proportionality in theory but applying injunctions as the norm.

In UK practice, permanent injunctions are generally granted to patentees as well, except in unique circumstances (for example, in FRAND/SEP cases where courts will grant an injunction only if the implementer refuses to take a license on court-determined FRAND terms, effectively allowing the infringer to avoid the injunction by complying with a license). No such scenario has yet been tested at the UPC; it remains to be seen how the UPC will handle standard-essential patents and FRAND defences in terms of injunctive relief. Early indications (from a 2024 Mannheim decision on SEPs) are that the UPC will issue injunctions against an “unwilling” licensee, similar to the UK Supreme Court’s approach, but might stay enforcement if a FRAND undertaking is being negotiated – certainly it would seem the implementer has to engage reasonably with the SEP holder with reference having been made to the Huawei v ZTE standard.

Generally, the UPC’s approach to final injunctions can be characterized as patentee-friendly but mindful of abuse: a victorious patentee should get an effective stop to infringement across Europe, but defendants have a small window to argue that an injunction is unnecessary (if, for instance, they have completely pulled the product from the market). So far, no defendant has successfully avoided an injunction on proportionality grounds, but the door has been cracked open by Kaldewei for future cases to test that principle.