The Federal Circuit’s decision in Merck Serono S.A. v. Hopewell Pharma Ventures, Inc., Nos. 2025-1210, -1211 (Fed. Cir. Oct. 30, 2025), delivers a clarifying—and unforgiving—interpretation of what it means for prior art to be “by another” under pre-AIA 35 U.S.C. § 102. The opinion reinforces a long-standing but often underappreciated principle: unless the inventive entity is completely identical, a reference remains prior art, even when collaboration, confidentiality, and overlapping research histories muddy the factual waters.
The case arose from inter partes reviews challenging Merck’s patents covering oral cladribine dosing regimens for treating multiple sclerosis. The patents claimed priority to 2004 filings and named four Merck inventors. Hopewell relied primarily on a published international patent application (“Bodor”) filed by Ivax researchers during the period when Ivax and Merck were collaborating on oral cladribine development. Merck argued that Bodor should not qualify as prior art because the disclosure reflected Merck’s own inventive work shared during the collaboration. The Board—and ultimately the Federal Circuit—was not persuaded.
The Core Dispute: Collaboration Versus Inventive Identity
At the heart of the appeal was whether Bodor qualified as prior art “by another” under pre-AIA §§ 102(a) and (e). Merck’s theory was straightforward: the relevant six-line dosing disclosure in Bodor originated from Merck’s confidential research and therefore could not constitute prior art against Merck’s later-filed patents. In other words, Merck contended that the disclosure was not truly “by another,” even though the named inventors differed.
The Federal Circuit rejected that framing and reaffirmed a doctrinal line stretching back nearly sixty years. As the court explained, the statutory inquiry does not turn on collaboration, information flow, or even confidentiality obligations. It turns on inventive identity. Where the inventive entities are not completely identical, the reference is “by another,” unless the patentee proves that the specific portions relied upon reflect the collective work of the same inventive entity named on the challenged patent.
This point bears emphasis. The court did not adopt a bright-line rule that authorship controls. Instead, it required proof that the disclosure itself embodies the joint invention of the same group of inventors. Any incongruity—whether inventors are added or subtracted—renders the reference prior art unless the patentee can bridge that gap with evidence.
Reaffirming Land and Its Progeny
Much of the opinion is devoted to situating the dispute within the Federal Circuit’s—and the CCPA’s—prior precedent. The court leaned heavily on In re Land, 368 F.2d 866 (CCPA 1966), which held that individual inventors and joint inventors are distinct inventive entities. A disclosure by one inventor is not automatically the work of a joint inventive entity, even if the subject matter overlaps.
Merck attempted to limit Land to its unusual facts and instead relied on later cases such as Applied Materials v. Gemini Research and Allergan v. Apotex. The court was unpersuaded, explaining that those cases do not undermine Land’s core principle. Rather, they confirm that the key inquiry is whether the disclosure evidences knowledge by the same inventive entity—not whether there is some overlap in inventorship or a shared research lineage.
As the opinion makes clear, later cases including Riverwood, EmeraChem, Google v. IPA Technologies, and Duncan Parking consistently require complete identity of inventive entities to exclude a reference from the prior art. Where a reference includes contributions from even one inventor not named on the challenged patent, it remains “by another.”
Evidentiary Burdens and the Inventorship Trap
Merck also argued that the Board improperly shifted the burden of persuasion by requiring proof that one of Merck’s inventors, Dr. De Luca, made a specific inventive contribution to Bodor’s six-line disclosure. The Federal Circuit rejected that argument, emphasizing the distinction between the burden of persuasion (which remains with the petitioner) and the burden of production (which properly shifts to the patentee once the petitioner establishes a prima facie case).
Here, Hopewell showed that Bodor was filed and published before Merck’s priority date and named different inventors. That showing shifted the burden to Merck to produce evidence that the disclosure reflected the work of the same inventive entity. The court agreed with the Board that Merck failed to meet that burden. Testimony from Merck witnesses could not identify any concrete inventive contribution by Dr. De Luca to the Bodor regimen, and documentary evidence such as meeting minutes and briefing documents fell short of establishing joint inventorship of the disclosure itself.
The court also rejected Merck’s reliance on the “rule of reason” for corroboration, explaining that corroboration alone is not enough. The alleged contribution must be significant when measured against the full anticipating disclosure. Vague involvement or project-level participation does not suffice.
Obviousness: Retreatment, Result-Effective Variables, and Expectation of Success
After resolving the prior art issue, the court turned to obviousness and largely deferred to the Board’s fact-finding. Substantial evidence supported the Board’s conclusion that Bodor, in combination with Stelmasiak, rendered the claimed regimens obvious.
The Board credited expert testimony that multiple sclerosis is a chronic disease requiring retreatment, that Bodor’s defined cladribine-free periods logically imply retreatment, and that Stelmasiak expressly taught cyclic cladribine administration. The court rejected Merck’s attempt to characterize retreatment as speculative or conditional, noting that the claims did not require retreatment in every case—only that retreatment occur.
The Federal Circuit also upheld the Board’s determination that dosing optimization was a result-effective variable. The prior art taught monitoring lymphocyte counts to balance efficacy and safety, and the claims imposed no requirement that dosing be determined by a particular calculation method. As a result, the Board reasonably concluded that a skilled artisan would have had a motivation to combine the references with a reasonable expectation of success.
MPEP Reliance and Procedural Fairness
Merck further argued that it was unfairly surprised by the Board’s application of the complete-identity rule, pointing to language in the MPEP suggesting that a disclosure by “at least one joint inventor” cannot be used as prior art. The court dismissed that argument, noting that the MPEP itself incorporates Land and expressly states that inventive entities differ when not all inventors are the same.
More importantly, the court reiterated that the MPEP does not override binding precedent. To the extent the MPEP could be read to suggest otherwise, it does not control substantive law. There was no APA violation, and no remand was warranted.
Why This Decision Matters
This decision is a reminder that collaborative innovation carries structural patent risks—particularly under pre-AIA law. Joint research, shared data, and overlapping development efforts do not collapse inventive entities into one. Unless inventorship aligns precisely, earlier disclosures can and will be used as prior art, even where they arise from close collaboration and shared objectives.
Charles Gideon Korrell has long observed that companies often underestimate the inventorship consequences of collaboration agreements, especially when R&D partners file independently. This case underscores the importance of proactive inventorship analysis, coordinated filing strategies, and—where possible—joint research agreements that anticipate how disclosures will be treated in later validity challenges.
Charles Gideon Korrell also notes that the Federal Circuit’s reaffirmation of Land leaves little room for equitable arguments based on fairness or confidentiality. The statute asks who invented the subject matter, not who shared it first or under what expectations. For companies operating in highly collaborative technical environments, that distinction can be outcome-determinative.
Finally, Charles Gideon Korrell believes the opinion highlights the sharp contrast between pre-AIA and post-AIA regimes. While the AIA introduced exceptions designed to soften the impact of secret prior art in collaborative settings, those provisions were unavailable to Merck here. As older patents continue to be litigated, similar disputes are likely to arise with equally unforgiving results.
Takeaways
The Federal Circuit’s decision in Merck Serono v. Hopewell is doctrinally orthodox but practically sobering. A reference is “by another” unless the patentee can prove complete inventive identity for the relied-upon disclosure. Collaboration does not equal co-inventorship. Confidential sharing does not negate prior art status. And once a disclosure qualifies as prior art, well-supported obviousness challenges will be difficult to overcome.
For companies and counsel alike, the lesson is clear: inventorship alignment matters just as much as novelty and nonobviousness. Ignore it at your peril.