By Kevin E. Noonan --
Oral argument in Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co. was heard on Monday before an en banc panel of the Federal Circuit. Counsel for Ariad (John M. Whealan), Lilly (Charles E. Lipsey) and the U.S. Patent and Trademark Office (Mark R. Freeman) argued before an active Court, with most questioning coming from Judges Rader, Dyk, Lourie, Linn, Moore, and Newman. While active, the Court and counsel avoided the principal question lurking under the question of whether there is a separate written description requirement (one of the questions posed to the parties and amici by the Court). That question is what is the standard for assessing patentability under circumstances when the mere recitation of an invention -- i.e., simply naming it -- is insufficient to inform the skilled worker of what the invention is, even if the disclosure as a whole enables the skilled worker to make and use the invention.
Mr. Lipsey, Lilly's counsel, began the argument, and was almost immediately sidetracked by two questions from the Court. First, Judge Lourie asked whether written description should be considered a question of law or fact. Mr. Lipsey answered that it always was described as a question of fact, but had "come to be resolved like claim construction," being determined from "the four corners of the document." He said written description was akin to the question of conception, citing Professor Liivak's amicus brief to the effect that the specification is the "corroboration of conception." This colloquy continued, with Judge Gajarsa suggesting that the written description issue could be treated like obviousness, being a question of law with underlying facts, to which Mr. Lipsey noted that this is how conception is assessed, so it would provide a logical way to do it.
At this point Chief Judge Michel (at right) asked whether there is any evidence that there is a harm in considering the written description requirement as a question of fact? And more importantly, whether there was any evidence that any cases would have a decision changed if written description was considered a question of law versus a question of fact? And finally, he pointedly asked Mr. Lipsey whether, if it's a question of fact, "do you lose?" When Mr Lipsey answered "no," Judge Rader quickly asked, "Then why do you care?" Lilly's counsel quickly responded, "I don't, I was asked the question," prompting the first of several outbursts of laughter from the gallery.
Chief Judge Michel also delayed counsel arriving at the substantive portion of his argument by asking whether the District Court had ruled on alternative jury instructions regarding the written description issue. This portion of the argument devolved quickly into specific questions of whether a timely objection was made and alternative jury instruction proffered, and whether there was a charge conference where the District Court considered the "proper" instruction. Mr. Lipsey replied that the jury instruction permitted Ariad's claim to satisfy the written description requirement if the claims recited NF-kB inhibitors functionally, while Lilly insisted that the Federal Circuit's jurisprudence required there be some disclosed structural correlation with function.
Judge Linn (at left) then asked, "if the original claims are published, how can they fail the public notice test?" Mr. Lipsey responded that claims satisfy the notice function as to what the boundary of the invention is but do not describe the invention. The Court countered, asking "don't the claims describe what the inventor considers to be his invention?" "They sometimes do, but they often do not," according to Mr. Lipsey, reciting Judge Rich in a footnote in the In re Vanco case. There, Judge Rich said, patent claims describe the boundaries of the invention, like lot lines do with real property. But those boundaries do not describe the land itself, or in this context, the invention according to Mr. Lipsey's argument. Judge Lourie then applied the question to genus/species claims, asking: "Isn't the question of support that you possess a sufficient portion of the genus to be entitled to the boundary?" Mr. Lipsey's answer was that this is always the inquiry if you claim a genus, whether you have sufficient support for a generic invention in the description, which can be satisfied "in any one of a number of ways," reciting a sufficient number of species or a generic description of a chemical compound as examples. Judge Linn then asked: "Doesn't the statute specify the way to do it, sufficiency is measured by enablement?" Mr. Lipsey responded that this is one thing the statute requires. The second thing the statute requires, he argued, is that the thing claimed be something that the applicant invented -- as shown by an objective analysis of the four corners of the patent document. He maintained that this description of what the inventor had invented had "always been required," citing O'Reilly, Schriber-Schroth, In re Moore, and In re Ruschig.
Judge Dyk (at right) asked: "In light of what Mr. Whealan has conceded in his brief, what is the practical difference in Lilly's position and Ariad's position?" (It should be noted here that Ariad's reply brief seems to have hurt its position somewhat, because some of the judges seem to believe that Ariad had "retreated" from the position it espoused in its opening brief.) Mr. Lipsey took this occasion to make the strongest statement of Lilly's position, saying that "the question of whether the specification describes what an inventor has invented is fundamentally different than the question of what a hypothetical person of ordinary skill (in possession of all that is known in the art) might without undue experimentation be able to do." Judge Dyk pressed its question, asking: "But don't you take Mr. Whealan to have conceded that the specification must provide a written description as part of the enablement requriement?" Mr. Lipsey answered by both taking advantage of this characterization by the Court yet distinguishing Lilly's position from Ariad's. He said that while Mr. Whealen "does concede that, but then he circularly says the sufficiency of it is to be determined by enablement." This was the error in Ariad's position, according to Mr. Lipsey, since "the case law is clear that if an inventor has not originated the subject matter, has not invented it, as reflected in the corroborated conception, the patent document, he is not entitled to a patent on it, that is how our system is set up."
Mr. Lipsey continued in this vein, that the patent must go to the "originator" and not to a third party who could build upon what is contained in the patent specification. He cited Schriber-Schroth as making this "abundantly clear," because the Supreme Court "accepted that people of ordinary skill in the art would recognize that a flexible web could be used" with the claimed invention, but that wasn't described in the patent. Judge Rader asked whether the case said that the flexible-web embodiment of the invention in Schriber-Schroth had not been "described or identified" (foreshadowing a distinction made later by Mr. Whealan). The reason for this distinction, according to Judge Rader, was that applying the written description requirement as advocated by Lilly would "turn claim construction into a validity doctrine." (Mr. Lipsey later expressly answered this question, saying that that was claimed in the Schriber–Schroth case "was not the invention described, so it was a matter of description there.") The Judge went on to say that "[c]laim construction is the most difficult area, given the philosophical imprecision of language, we find that varies even from panel to panel, yet now you are making that a validity doctrine, does that trouble you at all?" Mr. Lipsey responded that he "wasn't making anything anything," and that the statute and a long history of court decisions supported his position. Judge Rader disagreed, asking (repeatedly) for one example of a case where the written description requirement was used outside the priority context. Mr. Lipsey responded that if there was a written description requirement in the statute (and both the Court and Mr. Lipsey agreed that Ariad had conceded that there was), "it has to apply to all claims in all applications in all arts." In response to repeated demands from Judge Rader for a case law example supporting Lilly's position, Mr. Lipsey proferred In re Sus and In re Moore without further argument.
Mr. Lipsey was permitted to make the following cogent statement of his argument:
The cases are clear, from O'Reilly to the present day, that there is a requirement separate and apart from what anyone else looking might have been able to do it, that will limit what a patent applicant can claim in his application, and that is that he must have demonstrated within the four corners of the application as it is read by one of skill in the art that this is something he invented, i.e., something he conceived, and because not all ideas are legally sufficient conceptions, and indeed many of them are not, research plans, the germ of an idea as in Genentech v, Novo Nordisk, not all ideas are conceptions, and it's up to the court to look at what this inventor actually invented what he is now claiming, and normally, I agree, in 99 out of a 100 cases, written description and enablement stand and fall together, but they need not as Schriber–Schroth indicates and it is now and has always been a validity defense and a validity requirement.
He then turned to a further response to Judge Rader's
question of whether the written description requirement applies outside the
priority context, making reference to the language of § 120 of the
Section 120 of the statute tells you when you are entitled to an earlier application filing date and it doesn't say that the claim has to be presented in the earlier case. There is only one requirement, it says it has to comply with Section 112, and if it does, you are entitled to that date. And indeed, I believe it has been conceded by Ariad that in that environment, the written description analysis is performed all the time, and if that's the case it can only go on under that statutory framework if it is a statutory requirement of Section 112, and if you accept that, there is no principled reason why the written description requirement does not apply to all claims in all applications at all times in all arts, which is what we contend.
(Judge Lourie helpfully asked, "isn't it true that there is no limitation of 112 to the priority question?" to which Mr. Lipsey answered: "Indeed, it does not.")
Judge Rader then asked: "Why did Judge Rich create the written description requirement in Ruschig? Why did he say in Rasmussen that he was applying it as it had been applied in priority questions? It was created for that purpose and he moved it out of the 130 Section because he was uncomfortable with the narrow interpretation of new matter? So if it was created to be so narrow, how did it suddently get to be so broad?"
Mr. Lipsey replied that he begged to differ: "It [the written description requirement] was not created in In re Ruschig, it had been in existence a long time by the time of Ruschig and was not an unusual rule." Responding to Judge Rader's challenge for "evidence that it had been in existence a long time," Mr. Lipsey replied that "O'Reilly v. Morse said, you are trying to claim something you didn't describe and haven't invented, you're out of here, Schriber–Schroth said you are trying to claim something you didn't describe." When Judge Rader characterized these cases as "enablement cases," Mr. Lipsey replied that "the word they used was described." Judge Rader then made his most direct statement of his previously-asserted position on the question, saying that the Court in these earlier cases used the word "described" "consistent with the statutory requirement, because the description was so as to enable, so you can use either term and be discussing the same thing." Mr. Lipsey asserted in response that "the Schriber–Schroth case described the thing changed that everyone would have known, and the only reason not to allow the patentee to claim it is that it was not what he had described." Judge Rader countered that "Schriber–Schroth said that the patentee must describe his invention so that others may construct and us it. They used description in its proper enablement context, identification and enablement, the claim scope identifying and the description enabling."
In his response, and final argument before his time was up, Mr. Lipsey made Lilly's argument for the difference between Ariad (and Judge Rader's) view, along with a suggestion that "overturning settled expectations" was not the Court's proper role:
What gets left out is the fundamental requirement of a U.S. patent that the person who applied for it invented what is being claimed, and that is judged by reference to his own description of his invention, the corroborated evidence of what he conceived, and if he didn't invent it, he isn't entitled to a patent, whether anyone else could have done it or not. And that's the structure of our system, and if this is to be changed, this interpretation of the statute has been in effect from 30, 40, 50, 100 years, and the Act has been amended many times, and not once has it been suggested that it be changed, and we suggest that the prudent thing, rather than upset settled expectations, it is for Congress.
The Court then heard from the Patent Office representative, Mr. Freeman, who directly addressed Judge Rader's argument that the early Supreme Court cases cited in support of a written description requirement were enablement cases, saying: "We don't think there is any way to read Schriber–Schroth, Permutit, or Moore as merely reflecting an enablement question," and then proceeded to recite the Question Presented in Schriber–Schroth, asking whether the defendant could be held liable wherein elements of the claimed combination were not described.
Judge Moore (at right) asked whether any of Ariad's claims at issue were originally-filed claims, and whether originally-filed claims could fail the written description requirement. She also asked specifically whether there were cases outside the genus/species situation in chemical cases where an originally-filed claim could fail the WD requirement? Mr. Freeman answered that, yes, claiming in purely functional terms, even as an original claim, could fail the requirement, citing the Office's Written Description Guidelines "where the specification describes how to make a compound but it is claimed by function, having a scope to any compound having that function." He went on to say that fulfillment of the written description requirement by originally-filed claims is presumed, and cases where these claims failed to satisfy the written description requirement should be "rare." He said that the "fight over original claims is a fight over whether it is always true that original claims describe themselves or if in the rare instance they may not."
Judge Dyk asked whether there was "really a difference" between what the Patent Office was saying and Ariad was saying (again, seeming to adhere to the position espoused in Ariad's reply brief rather than its principal brief) -- "what are we arguing about," he asked. Mr. Freeman responded that this would be "a great question for Mr. Whealan," provoking another round of laughter. Responding to the question, Mr. Freemand characterized Ariad's reply brief as being "baffling," and said that if all agree this means that when we judge "whether a patent does A and B, we look only at B -- we don't think it consistent with the text [of the statute]. The statute does not say the description is required only insofar as it enables."
Returning to the fray, Chief Judge Michel asked: "Why does the PTO care? How many applications are there that can't be rejected on other grounds?" and "I'm not asking about history, I'm asking about impact?" Mr. Freeman responded that he didn't know the precise numbers, "but number must be high -- there are 400,000 applications examined each year, and it has become increasing popular to claim inventions by their functions or effect or to describe species and claim the genus." The written description requirement was "one of the basic tools" the Office uses "to constrain these claims and to make the patentee make clear to the public what the scope of the invention is."
Judge Rader countered that he didn't see any evidence of the written description requirement being used in Office rejections in the 300-400 cases that come to the Court each year -- not a "a single case." Although Mr. Freeman began a substantive response -- "You cannot assume away 400,000 cases, the vast majority of the PTO's decisions are not litigated," Judge Newman (at left) chimed in, quietly saying: "Let me help you" (and provoking another round of laughter). Judge Newman said:
When I started, chemistry was the hot area, and the question of how many species you needed was assessed by the written description requirement, and while many things have happened [since] . . . we have come full circle, where there are generic concepts, this case is a good example, and how one describes the genus depends on how much detail is required. In some areas, description and enablement are not co-extensive. And generally when it comes up in the Office it is resolved, so I agree completely that it is quite rare to see those cases on appeal; they are not insoluble, it is much easier to run some experiments than to appeal and come back a decade later. The principle is importantly illustrated in this case, where the outcome may depend on how we answer the questions.
Judge Lourie then threw Mr. Freeman a softball, asking: "Doesn't it occur that someone files a broad claim and exemplifies it with very little, because it is all they have, and the PTO rejects on written description?"
In answering "Yes" to this question, Mr. Freeman used the Tronzo case as an example. Judge Rader remained unconvinced, asking: "Why isn't that covered in the vast majority of cases by the doctrine of enabling throughout the scope of the claim?" While agreeing that this could be the case, Mr. Freeman maintained that while it could be true in some cases, it is "not always true that what isn't described isn't enabled," using In re DiLeone as an example.
To this, Judge Lourie said: "It's because they are related, you have enabled other species but have not described it," and Mr. Freeman agreed as Judge Rader informed him that his time had expired.
Mr. Whealan (after exchanging pleasantries with Chief Judge Michel) was immediately directed to Judge Lourie's perception that there were significant differences in the positions taken in Ariad's opening brief and its reply brief. He was pointedly asked if Ariad had "retreated," and parried this question with a clear statement of Ariad's position:
What this Court has done is delinked the written description of the invention from the rest of the words of the statute, and what we are suggesting is that the description of the invention, and of how to make and use it is measured by the following terms. Therefore it has to be clear, precise and exact, and it has to enable. If this were not the case, then Congress would have said that the written description of the invention does not have to be clear, precise and exact terms, but the written description of the manner of making and using it does. That makes no sense.
He then also announced that he was not going to use "labels" (presumably, "written description" and "enablement"), but was stopped from developing this theme by a question from Judge Prost (at left), citing to that inconvenient reply brief -- with reference to the statements in the reply brief about the written description matching the inventor's contribution: "Are you saying that there is no way that you can enforce the written description requirement except by the enablement portion?" "Absolutely not," Mr. Whealan replied. Ariad's position was that the Court had delinked the written description and enablement requirements, leaving the Court with no standard for assessing compliance with the written description requirement. This led to the development of the "possession" standard, which led to the requirement to supply sequences and structure. "Possession does not exist in the statute," he said, "and it doesn't exist in the case law."
The main portion of his argument was linked to the Telephone cases and (to a lesser extent) to other early cases, where he maintained that the evidence was that the inventor had not made the invention and had not actually made all embodiments within the scope of the patent claims, was permitted to claim broadly because he had enabled the skilled worker to practice the invention. He said that the standard under the Court's jurisprudence had shifted from whether the inventor put the public in possession of the invention to whether the inventor was in possession of the invention, and ultimately to what the inventor had actually done. He said that Bell would have failed under the possession test used by the Court, because Bell didn't get the invention to work, others did. Judge Dyk noted that Lilly was not espousing that there has to be an actual reduction to practice, and Judge Linn asked what was the difference between possession and identification? Mr. Whealan used a perpetual motion machine as an example, where you could identify it but not enable it.
Judge Lourie (at right) then mentioned that almost all the amici, across all technologies and sizes, and most of the major patent bar associations, all disagreed with Ariad's position (at least the one taken in its principal brief). "What do they know that you don't?" he asked. When his initial response ("people like invalidating patents for any reason") was shot down by the bench, noting that the opinion was held "on both sides" (presumably, companies like Microsoft that frequently criticize patents and companies like Amgen that depend on patents), Mr. Whealan charged that Lilly and amici had set up "a strawman," that Ariad was not suggesting that the written description requirement be "read out" of the statute (something that might come as a surprise to anyone who read Ariad's principal brief and the grammatical parsing of the statutory language), but that the Court needed to enunciate the right standard. He also said that, in view of the vast sentiment against Ariad's position (characterizing the amici as "noise"), Ariad could only prevail if he were to convince the Court that ruling in Ariad's favor would not "harm the system."
When he began his argument with regard to the Supreme Court never having held, as the Federal Circuit has done, that there are separate written description and enablement requirements, he was interrupted by Judge Newman, who noted that while there are not cases that have come before the Court, that is because of the nature of the inventions in which the problem arises that they can be enabled -- "getting a gene out of a chromosome," which is "well known" how to do -- without being described. Mr. Whealan asserted in response that an applicant would not get a patent under these circumstances, without supporting that statement with any citation or other authority (perhaps erroneously believing that In re Kubin has foreclosed patentability to gene claims).
He also maintained that "an original claim defines itself," and that Lilly had changed this requirement because "under Lilly you would have to show possession." In response to Judge Lourie's question concerning whether the factfinder needs to believe a genus has been described, Mr. Whealan said the Federal Circuit should "follow the Supreme Court" with regard to "whether the description matched the inventor's contribution." He then contrasted the Telephone cases -- where Bell was granted broad claims to his manner of transmitting voice over a distance -- to the Incandescent Light cases, where the broad claims were not valid since most of the claimed "fibrous material" did not work (prompting one of the judges to ask whether Mr. Whealan was arguing a utility aspect to the question). Judge Dyk asked Mr. Whealan to describe the difference between "identifying" and "describing" in view of his position that he would use the former but not the latter term. Without directly responding, Mr. Whealan argued that "putting the statute back together" would lead the Court to applying the Telephone cases and that precedent, which he characterized as Supreme Court teaching that a patentee would not be limited to "how you did it" (i.e., practiced the invention). If you had "changed the way to do something, then [you are] entitled [to a patent,]" he said.
Judge Lourie then asked whether there were "a lot of broad claims with little support that [have] to be rejected?" Mr. Whealan turned again to the Incandescent Light cases, to which Judge Lourie said: "That was 150 years ago, we're dealing with today," and asked about the Gentry Gallery case. The question wasn't that the claim wasn't enabled, Mr. Whealan said, but that if you failed to identify, you failed enablement. Judge Moore asked in this regard whether requiring that a claim be enabled throughout its full scope would "satisfy this problem." Again pointedly reminding the Court that he was "avoiding labels," Mr. Whealan agreed. Judge Moore then turned to the facts of this case, and asked whether the other examples disclosed in Ariad's specification would "work like decoy molecules." Mr. Whealan said the answer is no, but that the patentee had identified the concept of preventing NF-kB from "hitting" (actually, entering) the cell nucleus, and identified three ways to achieve that goal. He then asserted that "others" did just what the patentees had told them to do and "got it to work." He cited the "Michigan paper" that the patentee had "told them what to do" and that they "just did it with the exact sequences disclosed by the patentee." He characterized the inventors -- "research university people" -- as "figuring out the key thing to do," and that "the Michigan people wouldn't have known to look at NF-kB."
Judge Moore then asked whether the Court should use the principle of stare decisis to "leave it alone?" Mr. Whealan responded that since the Supreme Court had not "done the analysis," there was no stare decisis (nicely disparaging the Federal Circuit's own precedent). He also noted the disagreements between the judges of the Federal Circuit, and said that this case was a "progression of the Court's jurisprudence" on the issue.
Judge Newman acknowledged that this case presented questions that concerned "very fundamental aspects of policy based on areas of technology -- how far an applicant must go with a scientific concept before the fundamental requirements of patentability will be met." She said it was "easy to understand why this issue has not been before the Supreme Court," and that in the earlier cases "the issues were much easier," mentioning that in O'Reilly, "the Supreme Court said of course you couldn't patent electromagnetism." Mr. Whealan countered that to the people in the late 19th Century, the telephone was "revolutionary" technology.
In his final remarks, Mr. Whealan restated Ariad's position that "if you teach enough that someone could do it, that satisfies it [the disclosure requirements of Section 112]." He said that "Ariad taught people what to do and they did it."
In rebuttal, Mr. Lipsey dismissed Ariad's reliance on the Telephone cases, saying that these cases "have nothing whatsoever to do with written description," involving a "unique" claim set reciting the apparatus "as herein described." He said that there was "no holding that identification of a common principle entitles you to claim every way of attaining it. The cases are legion that what you are trying to do is claim the desired result, as in Morse and the ice cutting case, those claims are improper."
He also dismissed the argument about the word "possession," saying that "Carnegie Mellon and LizardTech makes clear that it is a synonym for 'did he invent it?'"
Judge Dyk asked whether the issue was "the difference between a research plan and an invention, and that the parties would draw the line in different places?" Mr. Lipsey agreed, but said that "the only way to get there is to do away with the statutory requirement." The statute defines what is required -- did the inventor conceive the invention and reflect that conception? He also agreed with the final question posed by Judge Lourie, to the effect that "possession" is a synonym for what an inventor invented and disclosed.
While there is little to be gained by speculating on how the Court will rule based on the questioning from the bench, a remarkable aspect of the argument is that neither party addressed four-squarely the issue underlying the argument: how much disclosure is required for inventions and in arts where merely naming something is insufficient to describe it? There are consequences (positive and negative) for patentees and the public under the Court's current written description jurisprudence, as evidenced by the arguments from the amici involved, as targets or enforcers of patents under the current interpretation of § 112. What the Court should consider is whether or not these consequences are in sufficient balance, and what the consequences would be if the Court significantly modified its application of the written description requirement, thereby unsettling the expectations that have arisen in the patenting community during the development of the modern form of the written description requirement.
Patent Docs thanks Professor Oskar Liivak of Cornell Law School, who attended the oral argument and kindly provided identifications of a number of judges who asked questions during the hearing.