Novozymes Loses “Possession” Of Its Enzyme
I am not sure how I missed this decision, which came down on July 22nd, but it offers a rather scary hi-def picture as to where the written description requirement of s. 112 has been and where it is headed. This decision, Novozymes A/S v. Dupont Nutrition Biosciences, Appeal No. 2012-1433 (Fed. Cir., July 22, 2013) (copy available at end of this post) rendered by a split panel (Judges Schall and Bryson, with Rader dissenting), affirmed a district court’s JMOL ruling nullifying a jury verdict that the claims of Novozymes’ U.S. Patent No. 7,717,723 (a copy found at the end of this post) met the written description requirement.
Novozyme had identified “promising mutation sites” among the approximately 500 amino acid residues that are found in alpha amylase, a useful starch-degrading enzyme. Using random protein design and random mutagenesis, Novozymes identified 33 sites at which a switch in amino acid residues might alter and hopefully improve the stability of the enzyme. The provisional and the regular filing disclosed seven “parent”alpha-amylases. “Position 29” was identified as a useful location and one analog, S239W, was prepared – although it was not found to have improved properties, as were 13/19 substitutions at this position. While Novozymes was prosecuting a series of applications claiming alpha-amylase variants, DuPont obtained a patent covering, inter alia, a S239Q variant that was thermostable. Since Novozyme had disclosed variations at this position were useful to obtain thermostable enzymes, Novozyme filed and obtained claims in the application that yielded the ‘723 patent, including:
1. An isolated variant of a parent alpha-amylase wherein:
(a) The variant has at least 90% sequence identity to SEQ ID NO: 6,
(b) The variant comprises a substitution of serine at position 239 relative to the parent alpha-amylase…,
(c) The variant has increased thermostability relative to the parent…at pH 4.5, 90 degrees C. and 5 ppm calcium and has alpha amylase activity.
Read in the context of prior case law, this claim certainly seems to meet the UC v. Lilly/University of Rochester requirements that
“[f]actors to be considered in determining whether or not there is sufficient evidence of [legal] possession include the level of skill and knowledge in the art, partial structure, physical and/or chemical properties, functional characteristics alone…and the method of making the claimed invention” [MPEP 2163(II)(A)(1)(a)].
This claim seems to meet almost all of these factors – no one disputed that the level of skill in the art was not high, or that the art was not well-developed. Elements (a) and (b) are certainly “partial structure” and (c) recited both a physical property and a functional characteristic.
However, the district court did not agree, and the Fed. Cir. affirmed. Where did Novozymes go wrong? Novozymes argued that decisions like Boston Scientific [supra] could be distinguished because “the 2000 application expressly discloses each limitation of the asserted claims [as in claim 1, above]” and that the art worker would have found that the patent and parent application contain a clear disclosure of the invention. DuPont, on the other hand, argued that the WDR requires “the disclosure of the actual invention, and that the Novozymes application was no more than a “laundry list of potential solutions” of the stability problem. Dupont also accused Novozyme as using “hindsight to work backward …to show that, given knowledge of the claimed invention, each limitation could be retroactively derived….[Slip op. at 17-18].
The Fed. Cir. bought Dupont’s arguments, reasoning that “the supporting disclosure [for claim 1] provides only generalized guidance listing several variables that might, in some combination lead to a useful result.” In other words, the court relied on the needle-in-haystack approach of In re Ruschig where the CCPA in 1967 found that the specification did not provide sufficient guidance to lead the art worker from the disclosed genus to the claimed species. The Fed. Cir. then took language from Boston Scientific, see supra, that “’nothing in the [disclosure] indicate[d] that the claimed triene analogs [of rapamycin] might be of special interest’” and “because the disclosure did not identify any such analogs or any reliable means for divining one,” the written description failed to demonstrate that the inventors were in possession of the claimed invention.” Slip op. at 19-20.
Where the Fed. Cir. opinion doctrinally goes off the rail is at page 27 of the opinion, where the majority says that “this case is very analogous to University of Rochester, where the patent specification failed to disclose any compounds that could be used in the claimed methods [to inhibit COX-2].” This factual situation is nothing like the situation in University of Rochester, in which the specification did not disclose anything meaningful about the identity of an agent that might inhibit COX-2, except that it must be “non-steroidal”. The key phrase is “any compounds,” but it is as clear as it can be in view of UC v. Lilly, as discussed above, this level of disclosure may be optimal, but it is certainly not required.
Good lawyering led the majority of the Fed. Cir. panel that there were a myriad of possible variants disclosed in the Novozyme specification and that Novozymes did not provide sufficient “blaze marks’ to demonstrate possession of the DuPont variant. However, apart from the fact that claim 1 is directed to a genus of variants, not to a single compound, the district court and the majority of the panel essentially required actual reduction to practice of the DuPont variants. I won’t quote and quote, but language in the opinion clearly demonstrates that the absence of a working example or express disclosure of the DuPont variant was a major, if not the only basis, for the majority’s holding: “The [Novozymes application], however, contains no disclosure of any variant that actually satisfies the claims, nor is there anything to suggest that Novozymes actually possessed such a variant at the time of filing”… “ [The application] nowhere describes the actual functioning, thermostable alpha-amylase variants that those [claim 1] limitations together define.” Finally, the panel gave weight to expert testimony that improved properties could not be predicted “without actually making and testing the variants.”
Judge Rader authored a brief concurring opinion that begins: “Although a separate [WDR] and the vague notion of ‘possession’ that it embodies, still troubles me [citing Ariad], I write today to ask the court instead to give full attention to the rules that it has created.” Judge Rader goes on to argue that deference should have been given to the original jury verdict, and reviews some the factual findings that supported it. In Judge Rader’s view, in view of the high level of skill in the art, it would be routine, and not require undue effort, to make and to screen all of the possible mutations at the relatively small number of positions on the enzyme identified as responsible for the recited improved thermostability: “The court might also have credited the patentee with reducing the original 500 total amino acid positions down to a mere thirty-three.”
Since the factors supporting “possession” a la Lilly are so well developed here, this might be a good case for the largely new Fed. Cir. to consider en banc. In view of the increasing role of the WDR in halting patents for early-stage technologies, practitioners, litigators and inventors deserve no less.