Another "Bright-Line" Rule from the Federal Circuit
By Kevin E. Noonan --
The Supreme Court has had the occasion to overrule the Federal Circuit with prolixity over the past ten years. And in many of those cases (including KSR Int’l Co. v. Teleflex Inc., eBay Inc. v. MercExchange, LLC, Quanta Computer, Inc. v. LG Electronics, Inc., Microsoft Corp. v. AT&T Corp., Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Dickinson v. Zurko, and Medimmune, Inc. v. Genentech, Inc.), at least part of the Court's reasoning has been an attempt to prevent the Federal Circuit from developing and enforcing "bright line" rules where the Court believes a "totality of the circumstances" approach is more flexible and more consistent with the public interest that affects patent rights. In part, this difference of opinion also stems from the two courts' fundamental disagreement about the Federal Circuit's role: the appellate court sees its Congressional mandate to include not only harmonizing U.S. patent law nationwide but bringing consistency to patent law interpretation. In this way, business people will be able to understand the metes and bounds of the patent grant with sufficient clarity that they can both avoid infringing activity and be prompted to further innovation without undue concern about the scope of third party patent rights. (The Supreme Court, of course, sees the Federal Circuit more as a regional circuit court, subject to the Court's own vision of consistency in how certain legal principles -- such as declaratory judgment actions and injunctions, recently -- are applied.)
The Federal Circuit had the opportunity to make another "bright line" determination in Takeda Pharmaceutical Co. v. Doll, where the issue was whether after-developed technology could be considered when assessing obviousness-type double patenting. The patents at issue involved cepham antibiotics, having an original U.S. filing date of December 19, 1975 (for the product patent, U.S. Patent No. 4,098,888) and a later filing date of January 8, 1990 (for the process patent, U.S. Patent No. 5,583,216). The issue was raised in a re-examination proceeding initiated in 1998 by two anonymous requests. The Examiner asserted a rejection on obviousness-type double patenting grounds, which rejection was affirmed by the Board. Takeda appealed to the District Court of the District of Columbia under 35 U.S.C. § 145, and introduced additional evidence with regard to alternative synthetic methods first disclosed in 2002 and 2005. The parties stipulated that at least one of these methods was in fact a "materially-distinct" method, the District Court granted summary judgment that Takeda was entitled to a reexamination certificate, and the government appealed.
The Federal Circuit reversed and remanded, in part because there were genuine issues of material fact unresolved below. This was due to the parties wanting the Federal Circuit to address the fundamental point of difference: whether it was proper for courts and the Patent Office to consider after-developed technology (i.e., alternative, patentably-distinct processes for producing the cepham compounds claimed in the '888 patent) when deciding whether process claims are unpatentable under the judicially-created doctrine of obviousness-type double patenting. In an opinion by Judge Rader, joined by Judge Moore and joined, in part, by Judge Schall, the Court started the analysis by noting that it had long been the law that the obviousness-type double patenting "doctrine bars an applicant from obtaining separate patents with separate terms for both a product and process for making that product, unless the product and process are 'patentably distinct,'" citing In re Taylor, 360 F.2d 232, 234 (C.C.P.A. 1966) and In re Cady, 77 F.2d 106, 109 (C.C.P.A. 1935). The question of whether a product and a process for making that product were patentably-distinct was whether "the product as claimed can be made by another materially different process," citing M.P.E.P. § 806.05. Returning to the Cady decision, the majority noted that "product and process claims are patentably distinct if multiple processes for creating a product exist at the time of the invention," Cady at 109.
But this begs the question: that is "the time of the invention?" The Patent Office argued that this was the priority date, or in the alternative no later than the date that the first patent issues; the Office provided no support for these propositions, except by analogy to other requirements for patentability (including §§ 102, 103, and 112). Takeda, of course, believed that later-developed technology could be used to show that a process was patentably-distinct from the product produced thereby. Takeda cited several cases in support of this proposition, but according to the Federal Circuit the main case (U.S. Steel Corp. v. Phillips Petroleum Co.) does not address the issue.
The Court decided that the Patent Office rule was too indeterminate, due to difficulties in deciding the "date of invention" (which could include the priority date, the filing date, the date of conception or reduction to practice, etc.); as noted in the dissent, these concerns seem unrealistic, since they are issues the Office and courts address and resolve every day. The Court also rejected Takeda's position, as giving applicants "the best of both worlds: the applicant can use the filing date as a shield, enjoying the earlier priority date in order to avoid prior art, and rely on later-developed alternative processes as a sword to defeat double patenting challenges."
Instead, the Court considered the purpose of the obviousness-type double patenting doctrine: preventing the "unjustified extension of the patent term." From this perspective, the filing date of the second application is what "triggers" obviousness-type double patenting concerns, because, of course, without this filing there would be no "unjustified extension of the patent term." The Court also considered the filing of the second application (without filing a terminal disclaimer) as constituting an assertion by the applicant that the claims are patentably distinct from the earlier-granted claims.
While seeming Solomonic, prudence suggests that the logic and effects of this decision should be assessed before hailing its wisdom. When an application with product claims is filed, the application must disclose a method for making the claimed compounds (to satisfy § 112, first paragraph). These methods can either be patentable (and thus be the only method known for making the product) or known in the art (and thus irrelevant to the double-patenting analysis). If the product claims are patentable on the original filing date, this should also be true on the date the later-filed process patent is filed. In the absence of alternative, patentably-distinct (and presumably non-infringing) methods, granting a separate patent under these circumstances would permit the patentee to effectively extend patent term, since there would be no alternative, non-infringing method for making the product even after the product patent expired. Thus, determining whether there are alternative, patentably-distinct (and presumably non-infringing) methods on the filing date of the second, process claim containing application as decided by the majority opinion is consistent with the policy goals of the obviousness-type double patenting doctrine.
On the other hand, if the patentee was limited to her filing date, then the only way a process would be independently patentable would be if there were known methods for making a patented compound. This would be limited to instances where a known method of making was used to make a novel compound, or where known methods of making were modified to make a novel compound. Otherwise, using the original filing date as espoused by the PTO would prevent independent patenting of compounds and methods per se.
The Court's decision also prevents later "validation" of patentability by alternative methods developed after the filing date of the process patent application. In addition to the uncertainty of when such methods are evaluated (During prosecution? In an infringement action? During a reexamination or interference proceeding?), it would also perhaps encourage development of those methods by patentee, disclosed or not. But in any event, permitting a patentee to use after-developed technology after the process patent filing date would introduce further uncertainty to the analysis. This result is inconsistent with the Court's preoccupation with business certainty.
Judge Schall's dissent focused on the other requirements of the statute, which he says are relevant because obviousness-type double patenting is a patentability doctrine (as opposed to an equitable limit on separate patenting to prevent "unjustified extension of the patent term," which is the majority position). His dissent was hardly scathing: he characterizes the majority opinion as a "not-unreasonable attempt at compromise." However, he dissents because there is, in his view, a fundamental logical inconsistency that after-developed processes can make an unpatentable process patentable. He thinks the majority opinion permits an applicant to "have it both ways," by relying on its priority date to avoid subsequent art while relying on subsequent art to supplement its disclosure and render its process patentable. "[T]here is no other doctrine or rule that allows unpatentable material to spring back into patentability based on later developments in the field," he writes. He also has public policy concerns:
I am also concerned that the majority's approach could, in certain cases, result in the upsetting of reasonable expectations as to what is in the public domain. . . . Takeda is granted a product patent in year one. At that time, only one process exists (Process A). That process is disclosed, but not claimed, in the patent. In year eighteen, the product patent expires, putting Process A fully into the public domain. In addition, during the seventeen-year patent term, no other process has been developed, and thus Process A remains unpatentable separately because it is coextensive with the product patent . . . Then, in year nineteen, a second process, called Process B, is invented for making the patented product. The initial inventor then files a patent application on Process A (assume proper co-pendency with the product patent), triggering double-patenting concerns. According to the majority's approach, this second process would cause Process A to become patentable, despite the fact that Process A and the resulting product were in the public domain for one year. It seems to me such a result runs counter to public expectation and patent law's public notice function.
While Judge Rader characterized these concerns as "a stretch" (footnote 1 in the majority opinion), the real limitation on the effects of this decision is the change in U.S. law, pursuant to adoption of the GATT provisions in 1995, in the length of patent term for a U.S. patent to be 20 years from the earliest priority date instead of 17 years from grant date. The decision thus can be seen as merely once again illustrating how the Federal Circuit continues to approach these policy decisions from its own unique perspective, something only occasionally in evidence recently.
April 27-28, 2009 - 
