2,177,576 / celecoxib / CELEBREX
The decisions under appeal, Mylan / celecoxib (NOC) 2014 FC 38 and Apotex / celecoxib (NOC) 2014 FC 314 (blogged here and here, respectively), turned primarily on the construction of the promised utility of the ‘576 patent. In both cases Harrington J interpreted the promise of the patent modestly and granted an order of prohibition to Pfizer. In a decision which is primarily an application of Plavix FCA 2013 FCA 186 (blogged here), the FCA has now affirmed both of Harrington J’s decisions. (While Harrington J’s decisions were separate, the reasoning was common, and the appeals were heard together .) While the FCA decision elaborates on some points relating to the promise of the patent, it is perhaps more noteworthy in its treatment of abuse of process, particularly the holding that it is not necessarily an abuse of process for an innovator to switch positions across NOC proceedings.
With respect to the promise doctrine, the FCA reaffirmed the law as set out in Plavix FCA, to the effect that “the threshold that must be proven to establish utility is generally quite low, described as being no more than a ‘scintilla of utility’,” but “[t]he promise doctrine represents an exception to the above minimum statutory requirements,” so that if the inventor explicitly promises a specific result, that patent will be invalid if the promise is not met, even if the invention satisfies the “scintilla” threshold -.
The most significant development in the law is the holding that the promise doctrine applies on a claim-by-claim basis. This arose while considering a putative promise that the invention would be useful in treating colorectal cancer, which had been claimed in Claim 16.
[85 The central point at issue is whether, as a matter of law, any given promise made in a patent must be construed as overarching to the invention and all of the patent’s claims. As I understand the argument, a patent containing a failed promise cannot be saved by severance pursuant to section 58 of the Patent Act.
 In effect, Apotex argues that any given promise must be construed as overarching, and that, because the respondent has failed to rebut Apotex’ allegation that the promise of preventing colorectal cancer was neither demonstrated nor soundly predicted as of the filing date, the validity of the patent as a whole is undermined.
The FCA rejected this proposition, accepting the alternative view that “that some promises are properly construed so as to touch only a subset of claims” :
 Though I accept that some promises may impose utility requirements across each of a patent’s claim, Apotex has offered no reason to depart from the Federal Court judge’s determination that the promise of colorectal cancer prevention, if any, can be severed [by application of s 58].
Within this legal framework, the FCA reviewed the construction of promise de novo, as a matter of law , and upheld Harrington J’s conclusion that the patent promised neither utility in humans nor reduced side effects.
Abuse of Process, Stare Decisis, and Comity
Turning to stare decisis, the issue arose because in Novopharm 2007 FC 81 rev’d 2007 FCA 173, an NOC proceeding concerning the same ‘576 patent, Hughes J had accepted a concession by counsel for the patentee that the promise of the patent included reduced side effects , and had construed the promise of the patent accordingly.
Apotex argued that it was abuse of process to allow the patentee to take a position contrary to its concession before Hughes J, on the basis that an innovator cannot switch positions across NOC proceedings on the same patent . The FCA rejected this, saying that while “there may be instances where a concession made in one proceeding under the Regulations may be construed as binding upon the conceding party in a later proceeding involving a different party” , there is no support for the notion that shifting positions across proceedings is in itself an abuse .
Apotex also appealed to Ramipril FCA 2007 FCA 163, which is the leading case holding that when an a patentee has failed in an NOC proceeding against one generic, it is an abuse of process to relitigate the same allegation of invalidity when made by second generic. The FCA distinguished Ramipril FCA on two grounds. First, “the respondent in the current proceeding was successful in the dispute disposed of in the earlier proceeding” . That is, it may be an abuse for a patentee which was unsuccessful in obtaining a prohibition order against a first generic from trying again against a second, but it is not necessarily an abuse for a patentee that was successful in the first application from shifting its position in a second application.
The second ground was that in Ramipril FCA the patentee had tried to rely on new evidence that it could have adduced in the prior proceeding, while in this case the patentee did not want to be bound by a prior concession. While it is true that in Ramipril FCA the Court emphasized the evidentiary issue, saying eg “The doctrine of abuse of process calls for the innovator to bring forth all its evidence on each ground of invalidity raised” , I must say it is not clear to me what principle underlies this distinction. If the general rule is, as was said in Ramipril FCA  that the patentee is “required to put their best foot forward in the earlier proceedings,” this seems broad enough to encompass concessions as well as evidence. But it also seems to me that such a broad principle would be unfair to the patentee. In any litigation each party necessarily responds to the specific arguments made by the other party. A new generic bringing a new challenge to a patent will likely use different arguments from those that were unsuccessful in the first application, and to require the patentee to put its best foot forward in all respects in the first application would require it to anticipate all the arguments that might be made at any time in the future. A concession might be made for tactical reasons, or for even for cost reasons, that are unrelated to the merits. But perhaps the same might be said of the Ramipril FCA holding that it is an abuse to relitigate against a second generic. In any event, whether the distinction is perfectly principled or not, it is now apparent that a successful patentee has at least some room to change its position in a second proceeding.
With respect to stare decisis, the principle that a lower court is bound by holdings of a higher court, the FCA pointed out that the Court in Novopharm FCA never turned its mind to the question of the promise of the patent, and consequently stare decision did not apply . Comity on the other hand, is not binding: “There is no legal sanction for a judge’s failure to abide by comity” . Consequently, Harrington J could not be reversed simply for construing the promise different from Hughes J .