By Kevin E. Noonan —

Amgen Inc., the real party in interest, and the Biotechnology Industry Organization filed briefs recently asking the Federal Circuit to overturn the decision by the U.S. Patent and Trademark Office Board of Patent Appeals and Interferences in Ex parte Kubin (see Patent Docs post).  The Board upheld an examiner’s rejection that the claims were unpatentable under 35 U.S.C. § 103 for being obvious and under 35 U.S.C. § 112, first paragraph, for failing to satisfy the written description requirement.  The Board based its obviousness determination on its interpretation that the Supreme Court’s KSR Int’l Co. v. Teleflex Inc. decision weakened prevailing Federal Circuit law (In re Deuel) that novel nucleic acid molecules were non-obvious.

The Board determined that all the claims would stand or fall together, since they were not argued separately; accordingly, their analysis was confined to claim 73:

The Board based its decision on language in the KSR decision that inventions that are "obvious to try" may also be obvious.  Here, the Board determined that it would have been "obvious to try" to isolate the claimed gene, Natural killer cell Activation Inducing Ligand (NAIL), using well-established molecular biological techniques.  The Board cited three prior art references, two of which were Valiente, which identified the existence of a protein, p38, present on the cell surface of natural killer cells and a monoclonal antibody specific for p38; and Sambrook et al., a standard reference work describing gene cloning techniques (otherwise known as the "Maniatis manual").  The Valiente reference contained a prophetic example to the effect that the gene encoding the p38 protein could be isolated using methods such as those disclosed in Sambrook et al.  The Board considered but did not rely upon a reference to Matthews, which disclosed a mouse protein, 2B4, expressed on the cell surface of natural killer cells (2B4 is the mouse ortholog of human p38 although that was not disclosed in the art).  Although the Board stated that it found Matthews "cumulative" of the teachings of Valiente and Sambrook, it used this reference as an "illustration" of how a gene could be isolated.

In its brief, Amgen argues that the Board both misinterpreted and misapplied the Supreme Court’s KSR decision.  In its argument, Amgen emphasized the Board’s determination that the art was "complex and unpredictable," and asserted that this unpredictability rebutted the Board’s application of the KSR rubrics with regard to when something "obvious to try" was also obvious.  The relevant portion of the KSR decision mandated that the result be predictable, which precluded KSR from being properly applied against Amgen’s claims:

When there is a design need or market pressure to solve a problem and there are a finite number of identified, predictable solutions, a person of ordinary skill has good reason to pursue the known options within his or her technical grasp.  If this leads to the anticipated success, it is likely the product not of innovation but of ordinary skill and common sense. In that instance the fact that a combination was obvious to try might show that it was obvious under §103.

In making this argument, Amgen distinguished the simple mechanical subject matter in KSR with the complexity of Kubin’s invention.

Amgen also cataloged the deficiencies in the cited art.  According to Amgen, the Valiente reference failed to disclose the isolation of p38 protein, or any amino acid sequence of the protein, or any ligand or binding partner of the p38 protein.  Valiente also provided no disclosure of any method for cloning a gene encoding the p38 protein.  The principle deficiency of the Matthews reference is that the relationship between 2B4 and p38 — that 2B4 was the mouse ortholog of human p38 — was not known in the art, and thus the Board’s reliance on Matthews was hindsight reconstruction.  In contrast, Amgen argued that Kubin identified a ligand, CD48, with which the p38 protein specifically binds on the NK cell surface, as well as a unique method for cloning the NAIL gene from pooled mRNA populations from activated and resting human NK cells that was not disclosed in any prior art (but which used Valiente’s p38-specific monoclonal antibodies).

The focus on Amgen’s argument was not the non-obviousness of the way the gene was isolated but rather on the complete absence of any disclosure in the cited art of structure that would render obvious the claimed gene.  Amgen also argued that the Board’s application of KSR was hindsight reconstruction, since once the gene was obtained, its isolation could always be considered "obvious."  Finally, Amgen argued that KSR did not overturn Federal Circuit precedent regarding the standards for finding obvious a chemical compound, specifically as it relates to nucleic acids in In re Deuel, but including In re Bell and also Amgen Inc. v. Chugai Pharmaceutical Co., Noelle v. Lederman, and Takeda Chem. Indust. v. Alphapharma Pty. Inc.

BIO’s brief, which was limited to the obviousness question, argued that the Board was incorrect in ignoring Deuel as binding precedent, and misapplied the KSR decision to have "weakened" the precedential value of Deuel.  The brief also characterized the Board’s "evidence" of factual distinctions between Kubin and Deuel (namely, that advances in technology made isolation of the NAIL gene more predictable than isolation of the subject matter in Deuel) "unfounded," and further challenged the Board’s reliance on In re Wallach to support the idea that improvements in technology could influence the obviousness vel non of nucleic acid claims.  BIO argued that the Board has misinterpreted Wallach as well as KSR; in Wallach the prior art disclosed the complete, full-length amino acid sequence of the protein encoded by the claimed gene, whereas here the art contained not even a partial amino acid sequence of the NAIL-encoded protein (p38).  BIO proffered one unique argument in support of the patentability of Amgen’s claimed gene:  that the Board’s method for determining obviousness was contrary to statute, since it depended on the obviousness of the "way" in which the NAIL gene was isolated, and the statute clearly states that "[p]atentability shall not be negatived by the manner in which the invention was made."  BIO’s brief also emphasized the policy considerations behind overturning the Board and the negative consequences to the biotechnology industry should the Federal Circuit refuse to do so.  BIO argued that, applying the Board’s test for obviousness, almost all biotechnology inventions (and many chemical ones) could be considered obvious and that this threatened innovation.

While most attention has been focused on the obviousness question, the Board’s written description decision may be at least as important, since it presents a question not yet squarely put to the Federal Circuit.  The Board held that Amgen was not entitled to a claim to an isolated nucleic acid encoding a protein having 80% identity to the disclosed amino acid sequence and that bound to CD48.  Amgen’s specification disclosed the amino acid sequence of the protein encoded by the NAIL gene, as well as domains of the protein (signal peptide, extracellular domain, transmembrane domain, and cytoplasmic domain), including the portion of the protein that bound to CD48 (amino acids 22-221).  The specification also disclosed conservative substitutions of these sequences, in addition to deletions, insertions, and fusions.  These teachings were generic, however, and no specific amino acid sequence variants were disclosed.  The Board held that the absence of any specifically-disclosed variants was a failure to satisfy the written description requirement and upheld rejection on these grounds.  As Amgen pointed out, however, this determination was directly contrary to the Written Description Guidelines promulgated by the Office on January 6, 2001 (Example 14), as well as the more recently-revised set of Revised Training Materials (Example 11B) (see "An Analysis of the New Written Description Training Materials – DNA Hybridization & Percent Identity"), and the Board offered no explanation other than that the Guidelines did not constitute a "rigid rule" that the Board was compelled to follow.

Amgen argued that its specification contained sufficient disclosure to satisfy the written description requirement and was in compliance with both sets of examination guidelines.  Amgen argued that it had disclosed a correlation between structure and function — CD48 binding to amino acids 22-221 — as required under Enzo Biochem v. Genprobe, and that the sequence identity limitation of 80% defined the claimed genus of related species sufficiently for the skilled artisan to understand that Kubin was in possession of the invention throughout its entire scope.  The essence of Amgen’s argument is succinctly expressed in two paragraphs:

The Board’s requirement for a further identification of which particular amino acids are required for binding amounts to nothing less than imposing a per se requirement that an adequate description of protein variants must be presented solely in terms of structure – a rigid approach to §112 that this Court has rejected.  Indeed, "there is no per se rule that an adequate written description of an invention that involves a biological macromolecule must contain a recitation of known structure."  Falkner, 448 F.3d at 1366; see also Capon v. Eshhar, 41 8 F.3d 1349, 1357 (Fed. Cir. 2005) (the descriptive text required varies with the nature and scope of the invention and the state of knowledge in the field); Univ. of Rochester v. G.D. Searle & Co., 358 F.3d 916, 927-28 (Fed. Cir. 2004) (applicants have flexibility in how they opt to describe their invention).  Nor is there any requirement that the specification expressly identify which sequences fall outside the scope of the claim.  See Vas-Cath Inc. v. Mahurkar, 935 F.2d 1555, 1566 (Fed. Cir. 1991) (court erred by applying a legal standard that requires the drawings relied upon for §112, ¶1 support to necessarily exclude all diameters other than those within the claimed range).

To require patent applicants to enumerate each and every amino acid that can or cannot be altered within the genus, as the Board would require, ignores the high level of skill that exists in the art of biotechnology, the very same high level of skill that the Board recognized in its obviousness analysis.  Indeed, it would place an unduly onerous burden on the biotechnology industry to require a level of specificity that is unnecessary for skilled artisans to understand whether an applicant actually invented what is claimed.  And it is unwarranted by the Court’s case law, which specifically endorses a flexible approach.  See Rochester, 358 F.3d at 927-28.  Here, the descriptive text conveys to those of ordinary skill that Appellants have in fact invented the NAIL-peptide variants recited in claim 73.  This is all the law requires.

While the Federal Circuit has developed its written description jurisprudence in University of California v. Eli Lilly, Enzo, and University of Rochester v. G.D. Searle generally in the context of patent infringement litigation, the Patent Office has applied its Guidelines to effectively preclude an applicant from claiming "conservative substitutions," on the basis that there was no disclosure of a "representative number of species."  (Of the 8,367 issued U.S. patents reciting "isolated" and "nucleic acid" in the claims, only 20 contain the term "conservative substitution.")  It has become clear that an applicant could never disclose a sufficient number of substitutions that the Office would consider "representative" based on the type of generic disclosure set forth in Kubin’s specification, and that it was impractical if not impossible to comply with the Office’s requirements.  The Federal Circuit has not spoken on this aspect of how the Office has applied its written description jurisprudence (which continues to have its detractors on the Court).  The Office has extended the CAFC’s requirement that structural domains important for biological activity be identified, to require that an applicant identify which amino acid residues can be modified and which cannot.  This is a clear extension of the analysis set forth in Eli Lilly that has no support in any Federal Circuit decision.  Yet, the Office routinely refuses to allow claims containing "conservative substitutions" within their scope.  In re Kubin puts the question squarely before the Federal Circuit for the first time, and perhaps this will be the occasion for the CAFC to finally review en banc the law of written description as it has developed for the past 12 years.

The Patent Office brief is currently due on July 21st.