What is "the" Inventive Concept?

Alcon Canada Inc v Apotex Inc 2014 FC 791 Kane J

            2,606,370 / travoprost formulation / TRAVATAN Z / NOC

The ‘370 patent relates to formulations of travoprost, an anti-glaucoma drug, with a non-conventional preservative system, based on zinc ions. The motivation for the invention was to avoid the potentially harmful effects on the cornea of conventional preservatives such as benzalkonium chloride (BAK). In this NOC proceding Apotex successfully attacked the ‘370 patent as being obvious, though a inutility attack failed.

The conclusion that the claims at issue were obvious ultimately turned on the facts. One of Alcon’s own prior art products, the tear replacement product Systane Free, also used a non-conventional preservative system which very similar, even in the specific components, to the claimed formulation [236], [342]. The evidence accepted by Kane J was to the effect that it would have been obvious to extend the same preservative system to travoprost, given the acknowledged motivation to avoid the use of BAK [348].

Inventive Concept

There is, however, one aspect of Kane J’s opinion which I find disquieting. A substantial part of the obviousness analysis concerned the identification of the inventive concept [163]-[197], including extensive testimony from the experts, and a thorough parsing of the specification. (Kane J noted that because the claims were to a bare chemical formula, the inventive concept cannot be identified without looking at the specification [180], which, in my view, is clearly correct.) Kane J’s ultimate identification of the inventive concept drew heavily from elements identified in the specification itself. 

None of this analysis is objectionable in principle; certainly, expert evidence and the specification should be considered in assessing the putative inventive contribution. But the overall tenor of the analysis suggests a focus on identifying “the” inventive concept from the specification, which then becomes the focus of the obviousness analysis. This would be problematic in two respects. First, by focusing on the specification, this analysis could turn the obviousness inquiry into a subjective analysis, in which the question is whether the inventive contribution identified by the applicant is obvious. This is contrary to the principle that the obviousness analysis is objective. It would be evidently be wrong to hold an invention to be non-obvious because the applicant asserts in the specification that it is a work of genius, but it is also wrong to hold a patent obvious because the contribution identified by the applicant turns out to be obvious. For example, the inventor may believe that she has discovered a whole new class of compounds with unexpected properties, and she has claimed both the genus and some particular species. It may turn out that the inventor had overlooked some relevant prior art, and in fact it was obvious or anticipated that some members of the class had the properties in question, but it was nonetheless surprising that the particular claimed species had those properties. The question is not whether what the inventor thought she had contributed is truly inventive; it is whether what she actually contributed is inventive: see Nichia Corporation v Argos Ltd, [2007] EWCA Civ 741 at [14]-[19].

Second, as a corollary, it would be wrong to focus on identifying “the” inventive concept, which is thereafter the sole focus of the obviousness analysis. If a putative inventive concept is examined and found to be wanting, the objective nature of the obviousness analysis makes it necessary to ask whether some other aspect of the invention might be inventive. So, if an invention claims a solution to a problem, it may be non-obvious either because the solution was non-obvious, or because the problem itself was non-obvious; it is necessary to ask both questions, not just one. As the EPO Guidelines explain, the question is the formulation of the objective technical problem:

The objective technical problem derived in this way may not be what the applicant presented as "the problem" in his application. The latter may require reformulation, since the objective technical problem is based on objectively established facts, in particular appearing in the prior art revealed in the course of the proceedings, which may be different from the prior art of which the applicant was actually aware at the time the application was filed. In particular, the prior art cited in the search report may put the invention in an entirely different perspective from that apparent from reading the application only. Reformulation might lead to the objective technical problem being less ambitious than originally envisaged by the application.

It is true, of course, that the Windsurfing/ Pozzoli approach as approved by the SCC in Sanofi 2008 SCC 61 [67] refers to identification of the inventive concept as one of the steps, but the inventive concept is by no means the focus of the analysis. Indeed, the test is very casual about the inventive concept: “Identify the inventive concept of the claim in question or if that cannot readily be done, construe it.” That is, don’t worry too much about the inventive concept – the crucial question [67(4)] is whether the differences between the state of the art and the invention as claimed were obvious.

This is a long-winded discussion for a point that probably had no real effect on the result, but there was a similar problem with Kane J’s analysis in the earlier Travoprost case, 2014 FC 699 discussed here. I am concerned that this may signal a new trend that has the effect of improperly turning the obviousness analysis into a subjective inquiry, in the same way that the promise doctrine has made utility a subjective inquiry.

Utility

Turning then to utility, the issue turned entirely on the construction of the promise. Apotex argued that the patent promised minimization or reduction of side effects [352], while Alcon argued that the utility is simply to be useful and to meet the USP preservative efficacy testing standards [351]. Like O’Keefe J in the recent decision Dow v NOVA decision, 2014 FC 844 (blogged here), Kane J relied both on the leading FCA decision, Plavix 2013 FCA 186, for the proposition that the promised utility must be explicit [408], and that “not every statement of advantage in the specification rises to the level of promised utility” [356], and also on Zinn J’s statement in Fournier / Fenofibrate 2012 FC 741 (discussed here), for the proposition that the focus should be on the claims [360]. In light of this jurisprudence, and noting in particular the absence of any promise of reduced side effects in the claims [416], Kane J held in favour of Alcon’s interpretation of the promise [417]. This construction of the promise effectively decided the issue of utility in favour of Alcon, because Apotex’s argument was focused entirely on the promised utility as Apotex had construed it [427], and did not address the issue of demonstrated or soundly predicted utility from the perspective of “simple utility” [440]. The unchallenged evidence led by Alcon was therefore sufficient to establish utility on the basis of sound prediction [446].

Note: see also 2014 FC 699 in which the claims of a different patent, 2,129,287, were to the use of travoprost for the treatment of glaucoma, as opposed to the travoprost formulation. Kane J held the claims of the '287 patent were obvious and anticipated, though the utility attack also failed: see here.