Bd. of Trs. of the Leland Stanford Junior Univ. v. Chinese Univ. of H.K., No. 2015-2011, 2017 U.S. App. LEXIS 11382 (Fed. Cir. June 27, 2017) (Before O’Malley, Reyna, and Chen, J.) (Opinion for the court, O’Malley, J.).
The Federal Circuit vacated three interference decisions, in which the Board found that Stanford’s claims were unpatentable for lack of written description, and remanded for further proceedings.
Two competing inventors, Stanford’s Professor Quake and Professor Lo from the Chinese University of Hong Kong (CUHK), each developed methods for diagnosing aneuploidies, using cell-free fetal DNA from maternal blood samples. Aneuploidies is a genetic condition in which a fetus has an abnormally high or low number of chromosomes, e.g., Down’s syndrome or Turner’s syndrome.
Quake’s specification discloses that second-generation massively parallel sequencing (“MPS”) platforms, specifically the Illumina sequencing platform, are useful for performing his method, as MPS can process large numbers of DNA samples simultaneously. MPS can be performed by random methods, in which all DNA in a sample is linked to a leader sequence and amplified using a primer complementary to the leader, or by targeted methods, in which a target sequence is specifically amplified and then sequenced. Quake first filed an application covering his method on February 2, 2007, and Lo first filed an application covering a “random sequencing” method on July 23, 2007. Quake’s originally filed claims were directed to the method of detecting “target sequences,” but during prosecution, Quake canceled all pending claims and added those that explicitly covered random MPS methods.
Both Quake and Lo requested interferences to determine inventorship of the random sequencing method. In each of three related interferences, Lo attacked Quake’s interfering claims as unpatentable for lack of written description. The Board agreed with Lo in all three interferences, finding that (1) Quake’s specification disclosed targeted, rather than random, sequencing and (2) the specification would not have indicated to a person of ordinary skill in the art that Quake was in possession of the claimed random MPS method. The Board credited the testimony of Lo’s expert, Dr. Gabriel, finding that although the Illumina platform referenced in Quake’s specification could be used for both random and targeted sequencing, in the context of the specification as a whole, the Illumina platform referred to targeted sequencing.
Stanford appealed the Board’s interference rulings to a district court, where the parties engaged in discovery, including full expert review. CUHK notified the district court of a potential jurisdictional issue arising out of Biogen Idec MA, Inc. v. Japanese Foundation for Cancer Research, 38 F. Supp. 3d 162 (D. Mass. May 22, 2014). There, a district court held that the AIA permits appeal of a Board interference decision only to the Federal Circuit for interferences declared after September 15, 2012. The district court stayed Stanford’s appeal, pending the Federal Circuit’s review in Biogen. After the Federal Circuit affirmed Biogen and denied rehearing, Stanford’s appeal was transferred to the Federal Circuit.
Stanford asked the Federal Circuit to (1) revisit its jurisdictional holding in Biogen, (2) consider the evidentiary record developed at the district court or instruct the Board to take that record into consideration, and (3) reverse the Board’s lack of written description ruling, even if limited to the record developed before the Board.
The Court declined to reconsider Biogen and refused to consider the district court record, explaining that because the district court did not have subject-matter jurisdiction to review the Board’s interference decisions, “the activities in the district court are a nullity.”
The Court then turned to the written description issue. The Board relied on Dr. Gabriel’s testimony, at least in part, because of “published references” she cited. However, “the Illumina references post-date the 2007 priority date, and the other references discuss a platform not referenced in the … patent.” The Board erroneously relied on references concerning the Roche 454 platform, not the Illumina platform described in Dr. Quake’s specification. The Board did not connect targeted sequencing on the Roche 454 platform to targeted sequencing on the Illumina system and did not explain what it found persuasive about the Roche 454 references as applicable to the Illumina platform.
Stanford offered evidence showing that the Illumina platform, a second-generation MPS platform, came after the Roche 454 platform, a first-generation MPS platform. The two systems operate differently, and targeted sequencing is more difficult using the Illumina platform. According to the Court, neither Dr. Gabriel nor the Board explained why conclusions about the Roche 454 platform should be used to conclude that the Illumina platform essentially teaches only targeted sequencing.
The Board also found that the artisan would have understood the discussion of MPS in Quake’s specification to refer to targeted sequencing and not random sequencing because he or she “‘could have considered the references in the … specification to Illumina products to indicate targeted sequencing,’ in part because ‘the language of the [specification] does not preclude targeted massively parallel sequencing.’” (emphasis added by the Court). The Court explained that this finding “ignores the fact that the same description might be able to disclose both random and targeted sequencing, and the Court thus found that “[t]he Board frames its finding in terms of an erroneous premise: the Board’s task was to determine whether the … written description discloses random MPS sequencing, as recited by the later-added claims, not whether the description does not preclude targeted MPS sequencing.” (emphasis in original).
The Court thus vacated the interference decisions and remanded for the Board to (1) “reconsider whether Quake’s relevant patents and applications satisfy the written description requirement” and (2) examine whether the artisan “would have known, as of the priority date, that the … specification references to Illumina products meant random MPS sequencing as recited in the claims” and whether he or she “would have understood that the … specification disclosed random MPS sequencing, as opposed to whether the specification did not preclude targeted MPS sequencing.”
Prior art describing specific embodiments of a technology should not be generalized or applied to different specific embodiments, absent evidence that the different embodiments should be similarly treated. An adequate written description must describe to the artisan what is claimed; it need not preclude alternatives that are not claimed.