By Kevin E. Noonan —

Biotechnology patent law faces the consequences of two decisions handed down last week by the Federal Circuit:  In re Kubin and Ariad Pharmaceuticals, Inc. v. Eli Lilly and Co.  The first of these illustrates the folly of ignoring the wisdom of the In re Deuel decision, in applying established principles of chemical patent practice to avoid the philosophical error of finding a chemical compound obvious because methods for making an unpredictable compound were well known.  The second decision, on the other hand, reflects Judge Lourie's wisdom in recognizing that biotechnology was sufficiently complex that permitting applicants to obtain patents without requiring sufficient disclosure was equal folly, particularly when the technology was developing so rapidly that what an applicant could enable might outstrip what the applicant could describe.  While the Ariad Court reaffirmed this principle, Judge Linn's concurring opinion — disagreeing that a written description requirement exists independently of the enablement requirement — suggests that the Court may be forgetting (or choosing to ignore) the wisdom of making sure an applicant's reach does not exceed his or her grasp.

The Kubin decision was based on a misreading of both the Deuel decision and the Supreme Court's decision in KSR International Co. v. Teleflex Inc.  The situation in Deuel is well known but bears repeating.  The claims at issue were directed to specific nucleic acids encoding human (and bovine) heparin–binding growth factors.  The prior art consisted of a reference to certain brain proteins, disclosing partial amino acid sequence, and the Sambrook cloning manual.  The U.S. Patent and Trademark Office Board of Patent Appeals and Interferences affirmed rejection on obviousness grounds.  The Federal Circuit reversed, based on the absence of any structure in the prior art that could be used to support a finding that the claimed cDNAs would have been obvious to one having ordinary skill in the art.  As said clearly in Judge Lourie's opinion:

The Bell reference, mentioned in the Deuel opinion, disclosed a cDNA sequence for proteins (insulin-like growth factors) wherein the prior art disclosed the complete amino acid sequence.  Bell's claims were limited to the exact nucleotide sequence of the human cDNA, and the Federal Circuit's basis for reversing an obviousness rejection by the Board was that this sequence was but one of 10³⁶ possible nucleotide sequences that could encode these human proteins.  The cited routine methods for cloning could not have predicted this sequence, out of the 10³⁶ possible nucleotide sequences, as being the one that encoded the human gene.

The Deuel opinion reaffirmed this principle on structural non-obviousness grounds, and extended it to encompass claims reciting the nucleotide sequence in terms of the amino acid sequence it encoded.  Judge Lourie, noting that these "genus" claims had not been argued separately, raised but did not address the sufficiency of disclosure issue these claims posed:

Although focusing on the enablement requirement in this passage, in the event (Regents of California v. Eli Lilly) Judge Lourie (at left) focused the inquiry on the written description requirement (ably assisted by a cogent District Court decision by Judge S. Hugh Dillin).  The beauty of this line of reasoning was that it was completely consistent with prior Federal Circuit case law on the topic of sufficiency of disclosure for nucleic acid claims.  These include Amgen Inc. v. C
hugai Pharmaceutical Co., which first articulated the principle that isolated genes should be considered in patent law like other chemical compounds ("A gene is a chemical compound, albeit a complex one, and it is well established in our law that conception of a chemical compound requires that the inventor be able to define it so as to distinguish it from other materials, and to describe how to obtain it," citing Oka v. Youssefyeh, 849 F.2d 581, 583, 7 U.S.P.Q.2d (BNA) 1169, 1171 (Fed. Cir. 1988)), and Fiers v. Revel ("An adequate written description of a DNA requires more than a mere statement that it is part of the invention and reference to a potential method for isolating it; what is required is a description of the DNA itself.  . . .  A bare reference to a DNA with a statement that it can be obtained by reverse transcription is not a description; it does not indicate that Revel was in possession of the DNA.").

In Eli Lilly, Judge Lourie set forth a bright-line rule of disclosure:

Tempered with the possibility that in the appropriate circumstances something other that a nucleotide sequence would satisfy the requirement:

The benefit of this analysis is that it avoids the question of enablement, since a disclosure of one mammalian species ortholog may enable production of others (indeed, this first isolation may be the Rubicon across which all mammalian orthologs are easily obtained, wherein the art did not describe or enable any species prior to its isolation), but without actually obtaining and disclosing it, an applicant is foreclosed from claiming it.

While occasionally losing its way (as in Enzo v. Genprobe I), the calculus developed by Judge Lourie has produced a fruitful equipoise between obviousness and written description for gene patenting over the past decade and a half.  On the one hand, the Court's application of the written description requirement ensured that applicants claiming a gene or nucleic acid must be in possession of the molecule.  Indeed, the U.S. Patent and Trademark Office has promulgated Written Description Guidelines (on January 6, 2001) and revised Training Materials (on March 25, 2009) specifically addressing the scope of disclosure required to support a nucleic acid claim.  This has restricted the scope of these claims to limit them, for example, to encoding the disclosed amino acid sequence and not to encompass substitutions (even conservative ones), insertions, or deletions while permitting fusion proteins in some instances.  This approach was approved sub silentio in Kubin, where the Court, despite evident interest at oral argument, chose not to rule on how the Office was applying these rules.

On the other hand, Deuel permitted genes that were not disclosed in the art to be patented without obviousness objections.  There is certainly a limit where what is known in the art can, asymptotically, approach obviousness, even in the absence of any known nucleic acid sequence.  This is illustrated by In re Bell, where the art disclosed the complete amino acid sequence and thus the claims were limited to a single disclosed nucleic acid species, based on the lack of disclosure in the art as to which of the 10³⁶ possible molecules actually encoded the human gene species.  And as the art develops, it becomes more and more routine to obtain cDNA species, impacting the scope of the gene claims that are non-obvious.

This disruption of Judge Lourie's careful balance in Kubin (and the threat Judge Linn's Ariad concurrence raises to written description as a separate requirement) raises again issues of disproportionate application of patent law to certain patentable subject matter, i.e., DNA.  It is certainly the case that the progress is not promoted by granting claims of broad scope to disclosures not commensurate therewith.  But biotechnology provides something no other technology can:  the capacity to produce useful quantities of proteins that form, in many instances, biologic drugs or the means to obtain them.  We can, of course, promote a policy where developing these drugs and technologies becomes more difficult and attracts less investment.  The question that immediately comes to mind is, why would we want to do that?

    By Andrew Williams —

On Friday, the Federal Circuit issued its decision in Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co., reversing the U.S. District Court for the District of Massachusetts's denial of Lilly's motion for JMOL in view of a jury verdict of infringement and validity of the asserted claims, and affirming the District Court's ruling that Lilly failed to establish the affirmative defense of inequitable conduct.

The Ariad case involved the technology of gene regulation, and specifically the transcription factor NF-κB.  NF-κB was first identified as playing a role in the expression of genes involved in the immune system, specifically in the regulation of the expression of the gene encoding the kappa immunoglobulin gene in B cells, which are specialized immune cells.  It was eventually discovered in the labs of Dr. David Baltimore and his collaborators that NF-κB did much more than regulate a single immune protein, playing a crucial role in the precise control of the expression of various genes and their protein products that are responsible for the response of cells to various and disparate stimuli, including bacterial lipopolysaccharides, certain cytokines, and even sunlight.  NF-κB is now known to be involved in expression of proteins that are associated with many different processes, including the inflammatory response and regulation of the immune system.  As can be imagined, NF-κB activity is tightly controlled in the cell.  When present in the cytoplasm, NF-κB is in an inactive state, bound to another protein, IκB, the natural inhibitor.  NF-κB is then activated when various stimuli external to the cell cause the NF-κB-IκB complex to dissociate, allowing the transcription factor to travel into the nucleus and bind NF-κB recognition sites found in various promoters.  This binding results in the production of many different proteins.  Then, when the stimulus is removed, the cell returns to a state in which NF-κB is inactive.  When this natural cycle of NF-κB control is disrupted, however, NF-κB can continue producing protein, often to the detriment of the cell, and can result in various diseases and conditions, ranging from sepsis, cancer, and AIDS.

In the mid-1980s, Drs. Baltimore, Philip Sharp, Thomas Maniatis, and ten other scientists at the Massachusetts Institute of Technology, the Whitehead Institute for Biomedical Research, and Harvard University, filed several patent applications related to their work identifying and characterizing NF-κB.  These separate applications, with Dr. Baltimore (at right) as the only common thread, were combined in a single application filed on November 13, 1991.  Eventually, on June 5, 1995, the application that gave rise to the patent-at-issue was filed, claiming priority to this previous application and containing essentially an identical specification.  The applicants, in this June 1995 application, sought claims for artificially reducing NF-κB activity in cells in order to prevent the problems when NF-κB activity runs amok.  In the end, after a lengthy prosecution, U.S. Patent No. 6,410,516 issued with 203 claims — 197 of which contain the same single step of either "reducing" or "altering" NF-κB activity in cells (the remaining 6 claims were not asserted in this litigation).  However, none of these 197 claims, including claims 80, 95, 144, and 145 at issue in this case, indicated how NF-κB was to be reduced or altered.

The history of the present litigation dates back to June 25, 2002, when Ariad filed its complaint on the same day that the '516 patent issued, alleging that certain claims were infringed by two of Lilly's drugs, Evista®, used for the prevention and treatment of post-menopausal osteoporosis, and Xigris®, used for the treatment of adult patients with severe sepsis who are at high risk of death.  After a fourteen-day trial in April 2006, a Massachusetts jury determined that the asserted claims were not anticipated by prior art or public use, that the specification enabled and adequately described the claims, and that the use of Evista® and Xigris® infringed the asserted claims.  The jury also determined that the effective filing date of the '516 patent was April 12, 1989.  Lilly subsequently moved for judgment as a matter of law, which the District Court denied.  After a four-day bench trial in August 2006, the District Court ruled that the asserted claims were directed to patentable subject matter and that the '516 patent was not unenforceable due to inequitable conduct or prosecution latches.  Lilly appealed all of these rulings except the ruling on prosecution latches.  In the Ariad decision, however, the Federal Circuit limited its invalidity holding to the issue of written description, and as a result did not comment on all of the other validity issues.  It can be noted, however, that in a current reexamination proceeding of the '516 patent, the Patent Office has finally rejected certain claims, including the asserted claims, based on inherent anticipation — a decision from which the patentees have filed a Notice of Appeal to the Board.  Also, the Federal Circuit's affirmation of the lack of inequitable conduct will be addressed in a subsequent post.

On appeal, the Federal Circuit had the opportunity to further expand its written description jurisprudence, potentially extending the rules as established in cases such as Univ. of Rocherster v. G.D. Searle & Co., 358 F.3d 916 (Fed. Cir. 2004) and Carnegie Mellon Univ. v. Hoffmann La Roche Inc., 541 F.3d 1115 (Fed. Cir. 2008).  However, the Ariad decision predominantly turned on the lack of evidence to support a finding that the inventors were in possession of the claimed invention on April 21, 1989, the effective filing date of the patent as determined by the jury, mainly because Ariad's evidence was directed to demonstrating possession as of November 13, 1991.  As the Court pointed out, "evidence of what one of ordinary skill in the art knew in 1990 or 1991 cannot provide substantial evidence" that the claims were adequately described in 1989.  Another important factor mentioned in determining possession is the context of claimed invention, but because Ariad had touted the fact that the subject matter of the patent "required years of hard work, great skill, and extraordinary creativity," the Court easily found that this patent was "in a new and unpredictable field where the existing knowledge and prior art was scant."  The Court concluded that because there was insufficient evidence that the '516 patent described any method for reducing NF-κB activity as of April 21, 1989, including a description of the molecules that are necessary to perform these methods, the asserted claims were invalid for failing the written description requirement.

Turning to what the Federal Circuit did determine to be disclosed in the '516 application family as of April 21, 1989, there were three, and only three, hypothetical classes of molecules that were disclosed as potentially being capable of reducing NF-κB activity:

It is worth noting that even if the effective priority date had been November 1991, the outcome would likely have been the same.  The Court noted that the '516 patent contains "no working or even prophetic examples of methods that reduce NF-κB activity, and no completed syntheses of any of the molecules prophesized to be capable of reducing NF-κB activity."  And, because the state of the art was primitive and uncertain, Ariad could not rely on "prior art knowledge to fill the gaping holes in its disclosure."  Even with the finding that decoy molecules were provided for in the specification, there was no accompanying description that they could be used to reduce NF-κB activity.  Of course, had Ariad argued for the later effective priority date, then it is possible that they would have had a different set of problems, such as the anticipation rejection(s) found in the Reexamination proceeding where the Patent Office assigned the November 1991 priority date to the asserted claims.  This fact was not mentioned by the Court.

The Court also quickly dispensed with Ariad's attempt to distinguish this case from the prior cases of Rochester, Fiers, and Eli Lilly by arguing that the asserted claims of the '516 were claiming methods for reducing NF-κB, and not specific molecules.  As such, Ariad posited, they were not required to describe the compositions required to carry out the claimed methods.  Ariad pointed out that each of the previous cases explicitly required the non-described compositions in the claims, whereas the '516 patent claims do not.  The Court pointed out, however, that even if a composition is not part of a claim, the specification must demonstrate that Ariad possessed the claimed method by sufficiently disclosing molecules capable of reducing NF-κB activity to demonstrate the patentee was in possession of the invention that is claimed.

Interestingly, the Federal Circuit commented on how broad the claims were, and appeared to take Ariad to task for maintaining such breath through claim construction and into trial.  However, the Court appeared to ignore the fact that Ariad was forced to seek such a broad claim construction in order to maintain infringement suits against such diverse pharmaceuticals as Evista, a small molecule, and Xigris, a recombinant human protein.

Judge Linn joined the opinion of the Court, but provided a separate concurrence.  He reiterated his belief that he expressed dissenting in the denials of rehearing en banc of the Rochester case and Enzo Biochem, Inc. v. Gen-Probe Inc., 323 F.3d 956 (Fed. Cir. 2002), that the engrafting of a separate written description requirement is misguided.  According to Judge Linn, § 112, first paragraph, should require no more than the specification enable a person skilled in the art to make and use the claimed invention and set forth the best mode for carrying out the invention.  Moreover, Judge Linn noted that because the Court decided the validity issue solely on written description, the majority failed to consider the important enablement issue raised by Lilly.  Thus, the issue of wh
ether claims that are broad enough to cover any method to achieve a particular result can ever be valid, since the specification cannot enable unknown methods, was left unresolved — an outcome that would not have occurred, Judge Linn noted, if there was not a separate written description requirement.