German Federal Court of Justice (Bundesgerichtshof), judgment of 17 March 2026 – X ZR 165/23 – on appeal from the Federal Patent Court
In a decision that commentators are already calling a milestone for pharmaceutical patents, the X. Civil Senate (Patent Senate) of Germany’s Federal Court of Justice (Bundesgerichtshof, BGH) upheld a patent covering sustained-release fampridine (Fampridine-SR) for improving walking speed in patients with multiple sclerosis (MS). The judgment is notable not only for its outcome—it reversed an invalidation by the Federal Patent Court—but for what it says about two pillars of patent law: the novelty of medical uses and the assessment of inventive step where clinical hurdles stand in the way.
Background: the patent and the dispute
The patent in suit concerned the use of a sustained-release formulation of fampridine (4-aminopyridine) to increase walking speed in MS patients—in essence, a dosing-regimen invention in a disease characterised by demyelination of nerve fibres, which slows or blocks the transmission of impulses along those fibres.
Generic challengers brought nullity actions. At first instance, the 3rd Senate (Nullity Senate) of the Federal Patent Court (Bundespatentgericht) declared the patent null by judgment of 11 July 2023. The patentee appealed (Berufung) to the BGH.
The outcome
The BGH reversed. It amended the Federal Patent Court’s judgment, dismissed the nullity actions, and ordered the claimants to bear the costs of the proceedings. The patent therefore stands.
Key findings of the decision
1. A high bar for novelty of a medical use (the “Memantine” line confirmed)
The Senate reaffirmed its Memantine case law: the suitability of a substance for a particular therapeutic effect is only disclosed in a novelty-destroying way if a prior-art reference makes clear, directly and unambiguously, that the substance actually has that effect. Mere expectations, estimates, or announcements of planned clinical studies—for example in investor materials or conference abstracts—are not sufficient to anticipate the invention. The Court drew a sharp line between disclosed facts and mere hopes.
2. Inventive step and the “roadblock” principle
Even where a prior-art document highlights a particular solution path, inventive step can be denied only if that path could actually have been pursued successfully with the knowledge and means available at the filing or priority date, judged on an ex-ante basis.
It is not enough that a path was theoretically conceivable. If, after setting out along it, obstacles or other circumstances would have arisen that—from a skilled person’s perspective—made continuing inadvisable, inventive step is not negated.
Crucially, the assessment of a reasonable expectation of success must take account of difficulties that were not yet known in the prior art but that would have materialised once the skilled person embarked on the path.
3. Clinical failures can support inventiveness
Where a clinical study (here, the MS-F202 trial) missed its primary endpoint and failed to show statistical significance, that failure can amount to a technical obstacle weighing against obviousness. Far from undermining the patent, a setback in earlier development can serve as evidence that the claimed solution was not obvious—because, without the invention, the skilled person would have faced what looked like a dead end.
4. Post-hoc analyses are a legitimate means of proof
The Court accepted that a post-hoc responder analysis is a valid way to establish a technical property. Although it is a statistical method, it is relevant under patent law because it speaks to a technical matter—the actual effect of the substance—and can support the efficacy of a specific dosing regimen even where the overall study cohort did not reach a significant result.
Practical takeaways
- Document clinical setbacks. Missing a primary endpoint in early trials is not necessarily fatal to a patent; it can demonstrate that the eventual solution was not obvious.
- Scrutinise the prior art closely. In nullity proceedings, distinguish carefully between references that disclose facts and those that convey only expectations or “hope values”—the BGH treats these very differently.
- Use subgroup / responder analyses. Identifying responders through later statistical evaluation is a legitimate way to substantiate the effect of a specific dosing regimen.
- Good news for dosing-regimen patents. The decision reinforces that the path to regulatory approval often runs through unforeseeable technical hurdles that can justify an inventive step.
Why it matters
For research-based pharmaceutical companies, the judgment is a meaningful signal. It confirms a demanding standard for anticipating a medical use, and it recognises that real-world drug development is rarely a straight line: obstacles encountered along the way—including failed trials—can be part of what makes an invention non-obvious. For generic challengers, it underscores that prior-art references must disclose the relevant therapeutic effect clearly and directly, not merely foreshadow it.
The full text of the judgment is available from the Federal Court of Justice (PDF).
