Second Medical Use Claims Before the UPC: Proving the Therapeutic Purpose

In Sanofi and Regeneron v. Amgen, the Düsseldorf Local Division upheld a second medical use patent but dismissed the infringement action. The difficulty was proving that the medicine would be prescribed for the claimed therapeutic purpose. A pharmacological effect mentioned in the product information did not establish that link.

Second medical use claims present a familiar difficulty: the same medicine can be prescribed for a patented purpose or for an unpatented one. The decision of 13 May 2025 shows how that distinction affects both claim construction and proof of infringement. It also raises a question for patentees: how can they prove that a therapeutic benefit influenced prescribing without being required to show that it was the principal reason for treatment?1

What the claim required

The dispute concerned Repatha, Amgen’s medicine containing evolocumab, an antibody that inhibits PCSK9. Claim 1 of Regeneron’s patent concerned the use of a PCSK9-inhibiting antibody or antigen-binding fragment to reduce lipoprotein(a), or Lp(a), in a defined patient group. The patients had to have serum Lp(a) above 30 mg/dL, be diagnosed with or identified as being at risk of cardiovascular or thrombotic occlusive disease or disorder, and not be receiving a therapeutic statin regimen at the time of administration. Sanofi was the exclusive licensee.2

The Division rejected a construction under which reducing Lp(a) had to be the dominant or sole reason for administering the medicine. Treatment could also aim to reduce LDL cholesterol. But the distinction between a therapeutic purpose and an incidental effect remained essential: administration solely to reduce LDL cholesterol fell outside the claim, even if Lp(a) happened to fall as well.3

The Court allowed concurrent therapeutic purposes while retaining the limitations that made the claimed use distinct. It also held that the patient’s Lp(a) level had to have been measured before treatment, although the claim did not specify when, how or by whom. Evidence concerning patients treated with Repatha in general would therefore not necessarily establish use within the claim.4

Actual or expected use, and knowledge

The Division approached direct infringement through Article 25 UPCA and Article 69 EPC. It did not confine protection to cases where the medicine had already been administered for the patented purpose. The objective element could also be established by circumstances showing that such use was to be expected. The supplier had, in addition, to know or reasonably to have known that offering or marketing the product would or might lead to that use. The claimants bore the burden of establishing these elements.5

The summary of product characteristics (SmPC) and the package insert were relevant, but the assessment extended to prescribing practices, the extent of the claimed use in the relevant market, and measures taken by the supplier to encourage or prevent it. The judgment therefore calls for a factual inquiry beyond the authorised indications. It does not, however, make a merely conceivable off-label use sufficient.6

Nor should the knowledge requirement be recast as a requirement to prove a deliberate plan to infringe. The test expressly includes what the supplier reasonably should have known. Conversely, knowledge that a medicine has a particular biological effect does not by itself establish knowledge that it will be prescribed for the claimed purpose. Those are different propositions, and the evidence must address both.

Why the evidence did not suffice

The claimants relied heavily on section 5.1 of Repatha’s SmPC, which reported a reduction in Lp(a) in clinical trials. The Division distinguished this pharmacodynamic information from the authorised therapeutic indications in section 4.1. In the circumstances of the case, reporting that effect did not show that Amgen was presenting Repatha for the claimed use or that physicians would prescribe it for that purpose.7

The clinical evidence did not resolve the difficulty. The claimants’ expert explained that an elevated Lp(a) level could influence the choice of a PCSK9 inhibitor. Amgen’s experts disputed the significance of that effect in prescribing. The Court also distinguished treating an elevated Lp(a) level from responding to the associated cardiovascular risk by lowering LDL cholesterol more aggressively. Only the former could establish the therapeutic purpose at issue. The difficulties of justifying off-label prescribing, including the risk of refusal of reimbursement or recourse by health insurers, further weakened the claimants’ case.8

There were additional gaps. The claimants had not established the claimed use specifically in patients with Lp(a) above 30 mg/dL, or that Amgen knew or should have known of that use. The action therefore failed on more than the interpretation of the SmPC. The counterclaim for revocation was also dismissed.9

Under Rule 9.2 RoP, the Division rejected additional briefs filed shortly before the hearing because the matters they addressed could have been raised earlier. It then explained that the four patient letters relied on would not have changed the outcome in any event: they did not establish the relevant prescribing practice, and one concerned a patient apparently receiving a statin.10

Proving a concurrent therapeutic purpose

For a patentee, the starting point should be the claimed patient group and the reason for prescribing. Aggregate sales or dispensing figures may establish the scale of use, but usually say little about its therapeutic purpose. Properly substantiated prescribing evidence, linked to the claim limitations, would address that issue more directly. Evidence of the supplier’s communications with healthcare professionals could then help establish what it knew or should have known. Depending on the market, treatment guidelines, reimbursement rules or tender documents may provide further context. These are possible sources of proof, not a checklist imposed by the judgment.

For a supplier, the absence of the patented use from the authorised indications is relevant but not conclusive. Any measures relied on to discourage that use should be assessed alongside actual prescribing practices and the supplier’s knowledge of them. The judgment does not transfer the general burden of proving infringement to the defendant or require it to account for every off-label prescription. Nor does it establish that a disclaimer will necessarily prevent liability.

A difficulty remains in the Division’s reasoning. At the claim-construction stage, reducing Lp(a) need not be the principal reason for treatment. Yet in assessing infringement, the Court required evidence that lowering Lp(a) was itself a therapeutic aim, and treated the prospect of an incidental reduction as insufficient. The distinction is defensible, but it can be difficult to prove where a physician chooses one treatment for several reasons. In my view, the evidence should be capable of establishing a concurrent therapeutic purpose without having to isolate it as the decisive reason for prescribing.

If the evidential standard in practice demands a predominant purpose, it would narrow the claim after the Court has declined to do so through interpretation. The task is to prove that reducing Lp(a) formed part of the therapeutic decision for patients within the claim, together with the supplier’s actual or constructive knowledge of that use. Merely repeating the pharmacodynamic information cannot do that work.

• 1UPC Local Division Düsseldorf, 13 May 2025, Sanofi Biotechnology SAS and Regeneron Pharmaceuticals Inc. v. Amgen Inc. and others, EP 3 536 712 B1, UPC_CFI_505/2023 (identified as UPC_CFI_505/2024 on the cover of the decision), ACT_597355/2023.
• 2Ibid., paras 1–5.
• 3Ibid., paras 107–109 and 191.
• 4Ibid., para 109.
• 5Ibid., paras 180–185.
• 6Ibid., paras 182–184 and 192–193.
• 7Ibid., paras 186–191.
• 8Ibid., paras 192–198.
• 9Ibid., para 199 and operative part, points I–II.
• 10Ibid., paras 200–201.