UPC Weekly brings you timely, thoughtful and joined-up insight into the evolving landscape of patent litigation at the Unified Patent Court.
2026 Week 42
This week, we discuss the Milan Central Division’s decision in Biogaran v. Lundbeck (UPC CD Milan, 9 October 2026). Biogaran sought revocation of Lundbeck’s EP 4223296 B1, which covers the medical use of Nalmefene for reducing alcohol consumption in certain patients with alcohol dependence. Before turning to the life sciences aspects of the decision (where novelty and inventiveness were acknowledged based on the identification of a new patient subgroup), it’s worth noting the UPC’s varying reactions when patentees ask to align their claims with those maintained in parallel EPO proceedings.
Parallel EPO proceedings: Different divisions have taken different views on coordinated amendments
The Milan CD admitted Lundbeck’s amendment – the deletion of two dependent claims – into the UPC proceedings to align with the amended patent maintained in the parallel EPO opposition proceedings. This contrasts with the Brussels LD recent order in Barco v. Yealink (UPC LD Brussels, 17 August 2026) – reported in UPC Weekly 2026 w35 – where the court rebuffed Barco’s attempt to file a further auxiliary request that had been admitted very late into the EPO opposition proceedings. As an aside, we note that this first instance order was upheld completely by the UPC Court of Appeal (CoA) in Barco v. Yealink (UPC CoA, 16 October 2026). The CoA refused to interfere with the LD’s wide discretion to manage the case, and also agreed with the LD that the new auxiliary request would change the infringement proceedings but would not in fact affect the outcome of the revocation proceedings (in the very specific circumstances of that case).
In the present case, the Milan CD said the admission was justified because the deletion of two claims simplified the dispute without raising new validity issues. Nevertheless, the Milan CD also took the opportunity to say that limitation to ensure consistency with parallel EPO opposition proceedings was a “fait accompli”:
24. Patent limitation is regarded by this Court as a fait accompli. The need to ensure consistency between the patent texts examined in the parallel proceedings requires that the same amended text submitted before the EPO be admitted in the present action.
In both cases, the amendments in the parallel EPO proceedings were first made at or shortly before the OD hearing. However, a substantial distinction between the amendments waved through by the Milan CD and those refused by the Brussels LD (and now by the CoA) is that the former was the simple deletion of two dependent claims, whereas the latter was a new combination of features from dependent claims being imported into claim 1.
The logic explained by the Milan CD should hold in many cases – it is of course preferable that the claims are aligned in parallel UPC and EPO proceedings where possible and provided that this does not disadvantage the other side. The Barco case is somewhat special, in particular in that the infringement action was based on only a selected set of the granted claims and the counterclaim for revocation was aimed only at those claims. The amendment sought to bring in features from non-asserted and non-attacked claims.
Patient selection as the therapeutic teaching
Lundbeck’s patent did not concern a new compound, disease or dosage regimen. As-needed use of nalmefene, a known opioid-system modulator, in alcohol-dependent patients was already known. Rather, the invention lay in identifying and treating a subgroup that derived greater benefit: patients with a high Drinking Risk Level (DRL) who maintained that level after a one-to-two-week observation period. Claim 1 is as follows:
1. Nalmefene for use in reduction of alcohol consumption in a patient with alcohol dependence, wherein said use comprises the following steps;
a) identifying a patient with alcohol dependence
i) who has a Drinking Risk Level (DRL) corresponding to consumption >60 g/day of pure alcohol for men and >40 g/day for women, and
ii) who maintains the DRL of alcohol consumption after an observation period following initial assessment,
wherein said observation period following initial assessment is 1-2 weeks, and
b) administering nalmefene to the patient identified in step a), wherein said nalmefene is to be administered as-needed, such as on each day the patient perceives a risk of drinking alcohol, preferably 1-2 hours prior to the anticipated time of drinking.
The court construed the claim in a pragmatic and straightforward way before deciding that it was novel, inventive and sufficiently disclosed.
Drawing on EPO case law, for novelty the court required a physiologically or pathologically distinct subgroup whose defining characteristics had a functional relationship with the therapeutic effect:
184 … If the claimed patient group merely constitutes a subset of patients already encompassed by the prior art, the requirement of novelty is, prima facie, not met. Conversely, where the claimed patient population is defined by physiological or pathological characteristics that are functionally linked to the therapeutic effect relied upon, the claimed subject-matter may satisfy the novelty requirement under Article 54(5) EPC.
The 1-2 week observation period excludes “early reducers” from the patient subgroup (i.e., those patients whose alcohol consumption falls substantially before treatment), thereby covering a narrower group of patients who showed a more pronounced response to nalmefene. Persistently high consumption was therefore found to be a “predictive selection criterion”, not merely a narrower description of a known patient population. The court therefore acknowledged novelty.
Sufficiency: the trial protocol need not limit the claim
In contentious proceedings, routine clinical-trial design choices can be used against patentees. The other side often argues that restrictive or unclear eligibility criteria, concomitant therapies and exact dosing regimens must appear in the independent claim for the invention to be sufficiently disclosed.
In this decision, the Milan CD pragmatically distinguished the invention – patient selection – from the conditions that had been used to demonstrate a credible therapeutic effect in the selected patients. A gap in the clinical evidence was not necessarily evidence of a gap in the credibility of the therapeutic effect. For example:
-
Biogaran’s identification of patient populations omitted from the trial was not in itself enough to doubt efficacy in those populations: “Exclusion criteria are common in clinical studies and are normally used to ensure internal validity, patient safety, and methodological reliability. They do not amount, without further evidence, to a technical teaching that nalmefene would not achieve the claimed therapeutic effect in every patient not enrolled in the trials.”
- Although all patients in the patent’s studies also received a motivational and adherence-enhancing intervention (the BRENDA model), Biogaran had not “established that the invention would not work in the absence of such motivational support”.
- The clinical studies in the patent used an 18 mg dose, but the independent claim contained no dosage limitation. This did not create a sufficiency issue: the invention did not lie in a particular dosing regimen and the prior art already indicated efficacy within the patent’s stated dose range. The skilled person could determine a suitable dose without undue burden
The court put it best: “sufficiency does not require every ancillary condition of the clinical protocol to be reproduced in the claim”.
Inventive step: obvious… with hindsight
The court also had little sympathy for an inventive-step attack assembled from more than ten prior-art disclosures and lacking a convincing route through the prior art. The disclosures addressed different technical questions, and none gave the skilled person reason to consult the others.
The key distinction was between (i) evidence that alcohol dependence has subtypes and (ii) evidence that those subtypes predict different treatment responses. Neither realistic starting point suggested selecting patients with a high Drinking Risk Level who maintained that level after a one-to-two-week observation period, or that they would respond particularly well to nalmefene. One was a study containing no subgroup analysis and reporting that diagnosis, family history and treatment-centre type had no significant effect on response.
Mechanistic literature did not fill the gap. It was uncertain and contradictory and, although the opioid system was known to be involved in alcohol dependence, did not make the claimed subgroup’s response predictable.
Biogaran’s mosaic therefore suggested hindsight rather than the course the skilled person would have followed at the effective date. The inventive-step objection was dismissed, and the patent maintained in amended form.
Conclusion
Overall, the decision shows a pragmatic approach from the Milan CD. The court focused on where the inventive contribution actually lay – identifying a technically meaningful patient subgroup – without requiring the claim to recite every feature of the supporting clinical studies.
.png)
.png)