The Federal Circuit’s December 2, 2025 decision in Seagen Inc. v. Daiichi Sankyo Co., No. 23-2424, continues a noticeable and consequential trend: the court’s growing willingness to overturn jury verdicts on written description and enablement grounds as a matter of law. In reversing a $41 million verdict and vacating findings of willful infringement, the panel treated the asserted claims as invalid under §112(a), concluding that no reasonable jury could have found adequate written description or enablement. The result is not merely a loss for Seagen, but another data point in the court’s broader recalibration of how disclosure adequacy is assessed in pharmaceutical and chemical genus claims.
This decision follows closely on the heels of other recent Federal Circuit reversals involving pharmaceutical patents, reinforcing that written description is increasingly viewed less as a fact-intensive inquiry and more as a structural assessment grounded in claim scope, specification architecture, and precedent. As Charles Gideon Korrell has observed in prior commentary, these cases reflect an unmistakable tightening of the disclosure requirement where patentees claim large combinatorial spaces while pointing to only sparse or indirect exemplars.
The Technology and the Claimed Subgenus
The patent at issue, U.S. Patent No. 10,808,039, concerns antibody-drug conjugates (ADCs), a complex and highly active area of pharmaceutical research. ADCs combine a targeting antibody with a cytotoxic drug, joined by a linker designed to release the drug intracellularly. The specific limitation that proved fatal was the requirement that the linker include a tetrapeptide composed solely of glycine and phenylalanine residues. Because phenylalanine may exist in multiple stereoisomeric forms, the claims encompassed a defined subgenus of 81 possible Gly/Phe-only tetrapeptides.
The asserted priority application, filed in 2004, disclosed linker peptides in extremely broad terms. It described dipeptides through dodecapeptides and identified dozens of amino acid options for each position. Critically, however, it never singled out Gly/Phe-only tetrapeptides, nor did it describe any species within that subgenus. At trial, the parties agreed that the 2004 disclosure encompassed more than 47 million possible tetrapeptide combinations. The jury nevertheless found the claims adequately supported and enabled, awarding Seagen substantial damages.
The Federal Circuit disagreed, emphatically.
Written Description and the Limits of “Encompassed” Disclosure
Writing for a unanimous panel, Judge Lourie framed the written description inquiry around possession, not possibility. Citing Ariad v. Eli Lilly, In re Ruschig, and Regents of the U. of Minn. v. Gilead, the court reiterated that a broad genus disclosure does not automatically support a later-claimed subgenus unless the specification provides “reasonably specific supporting disclosure” pointing to that subgenus.
Here, the claimed 81 tetrapeptides were mathematically encompassed by the earlier disclosure, but that was not enough. The court returned to the familiar “forest and trees” metaphor: blaze marks must guide the skilled artisan to the claimed invention. A disclosure that merely allows a skilled artisan to arrive at the invention by making a “straightforward leap” is, by definition, insufficient. As Charles Gideon Korrell notes, this language leaves little room for patentees to argue that expert testimony can substitute for missing disclosure when the specification itself remains silent.
The court found Seagen’s expert testimony particularly damaging rather than helpful. The expert’s concession that one would need to “leap” from disclosed examples to arrive at Gly/Phe-only tetrapeptides underscored that the inventors had not actually described or possessed that subgenus in 2004. Inventor admissions reinforced the point. Several named inventors testified that they had not contemplated such linkers until years later, after Daiichi’s Enhertu product became public.
In short, the Federal Circuit concluded that no reasonable jury could find written description support for claims plucked from an enormous combinatorial universe without textual guidance pointing to the claimed subset.
Enablement After Amgen v. Sanofi
The court’s enablement analysis followed a similar trajectory. Relying heavily on Amgen v. Sanofi, the panel treated the claims’ functional limitations as requiring disclosure of a common structural quality enabling the full scope of the claims. Instead, the patent left skilled artisans to conduct extensive trial-and-error experimentation to determine which ADCs would intracellularly cleave as required.
The district court’s claim construction compounded the problem by interpreting “drug moiety” broadly to encompass any drug capable of intracellular cleavage. Given the unpredictability of ADC chemistry, the court held that the specification did not teach how to make and use the full scope of claimed embodiments without undue experimentation. As in Amgen, providing assays for determining whether embodiments work was not enough. Enablement cannot be satisfied by instructing skilled artisans to test millions of candidates and see what sticks.
The Federal Circuit therefore reversed the jury’s enablement finding as unsupported by substantial evidence and invalidated the claims on that independent ground as well.
Overturning Jury Verdicts with Increasing Confidence
Perhaps the most consequential aspect of Seagen is not the application of existing doctrine, but the posture. Written description is nominally a question of fact. Yet here, as in other recent cases, the Federal Circuit showed little hesitation in substituting its own reading of the specification for the jury’s conclusions.
The court emphasized that once the specification is treated as a legal text whose structure and scope can be assessed de novo, the jury’s role narrows dramatically. This approach echoes the court’s post-Markman willingness to treat certain disclosure questions as legal determinations, even when expert testimony has been presented on both sides. As Charles Gideon Korrell believes, this doctrinal shift may ultimately discourage patentees from relying heavily on expert gloss where the four corners of the specification fail to do the necessary work.
Practical Implications
For patent drafters, Seagen underscores the perils of claiming narrowly after disclosing broadly, particularly in unpredictable arts. Genus claims that rely on combinatorial breadth must be paired with structural signposts that clearly identify the claimed subgenus. For litigators, the decision highlights the growing vulnerability of jury verdicts on §112 issues, especially in high-stakes pharmaceutical cases.
For companies developing follow-on or competing products, the case offers reassurance that late-stage claiming strategies will be closely scrutinized, even when they survive trial. And for those watching Federal Circuit doctrine evolve, Seagen represents another step toward a unified, stringent disclosure standard that blends written description and enablement into a single inquiry focused on whether the specification truly earns the breadth of the claims. Charles Gideon Korrell notes that this convergence may be one of the most significant doctrinal developments in patent law over the past decade.