2,129,287 / travoprost / TRAVATAN Z / NOC
Patent law is a technical area of law, and no doubt always will be, but at times it seems that the conceptual apparatus we deploy make the field even more complex than necessary. In this case, the concepts of selection patents and the inventive concept have generated some problematic analysis.
Travoprost is the isopropyl ester of (+)-fluprostenol [8]. It is a synthetic analogue of PGF2α, which is a naturally occurring prostaglandin. The claims of the ‘287 Patent at issue (12, 27, 35 and 46) relate to the use of a therapeutically effective amount of travoprost for the treatment of glaucoma and ocular hypertension [123]. (The claims are all to a relatively small class of compounds encompassing travoprost – none relates solely to travoprost – but nothing turns on this.) The claims were held invalid as being anticipated and obvious, but an attack based on lack of utility failed.
There was considerable dispute as to whether the ‘287 patent was a selection patent over European Patent application EP 0 364 417. Kane J held that it was not a selection patent [152], but that even if it was the result would have been the same for all three grounds of attack [153]. I have suggested previously that the concept of selection patents generates more heat than light, and should probably be abandoned. Kane J’s Travoprost decision seems to me to reinforce this view, but given that the selection patent issue did not make any difference, I will refrain from further comment on this aspect of the decision. It is in any event clear that the ‘417 application, which discloses the use of a large class of prostaglandin derivatives, encompassing travoprost, for the treatment of glaucoma and ocular hypertension, was highly relevant prior art.
Anticipation
Kane J held the ‘287 patent to be anticipated in light of the ‘417 application, despite the fact that the travoprost was nowhere specifically disclosed in the ‘417 patent, on the basis that the compounds disclosed by the ‘417 patent encompassed travoprost [371]. In my view, Kane J’s holding with respect to anticipation is clearly wrong. It is the same error as was evident in her Valganciclovir decision, 2013 FC 718, as discussed here. I expect that in due course Kane J will be corrected by the FCA on this point, though not soon, as both Valganciclovir and Travoprost are NOC applications so cannot be appealed by the patentee. Kane J did state that “[a] person carrying out the prior disclosure in the ‘417 would, on a balance of probabilities, infringe the claims at issue of the ‘287" [371]. However, I can see no support in her reasons for this proposition. The ‘417 patent encompassed billions of compounds [269], and, as just noted, travoprost was not specifically described or made. Nothing in Kane J's decision explains why a person carrying out the disclosure of the ‘417 patent would make travoprost rather than one of the other billions of compounds encompassed by the disclosure.
Utility
With respect to utility, Apotex argued that the patent promised both greater inter-ocular pressure reduction and reduced side effects as compared with the compounds of the ‘417 application [202], [205]. Alcon argued that the patent promised only that travoprost would be useful or therapeutically effective for the treatment in humans, without any promise that it would be better than the compounds of the ‘417 patent [188]. Kane J held in Alcon’s favour [241]. She emphasized in particular Zinn J’s statement in Fournier / fenofibrate (NOC) 2012 FC 741 [126] that a utility not expressed in the claim portion of the specification “should be presumed to be a mere statement of advantage unless the inventor clearly and unequivocally states that it is part of the promised utility" [181-82], [244]. The claims in question were to the use of travoprost in the treatment of glaucoma and ocular hypertension, without any mention of reduced side effects [185], and this was construed as the promise. Given this modest promise, utility was held to be soundly predicted.
Obviousness
Perhaps the most interesting part of the decision relates to obviousness. The ‘287 patent asserted [70] that
It has now been unexpectedly found that [the claimed compounds] show significantly
greater IOP reduction than the compounds of [the ‘417 application] while having a
similar or lower side effect profile.
Kane J held that the patent was obvious whether or not it was in fact true that the properties of the claimed compounds were unexpected. This is because she found that the inventive concept was simply that travoprost will be therapeutically effective in the treatment of glaucoma and ocular hypertension [461, 167], and it is the inventive concept, and not the disclosed properties of the invention, which are the focus of the obviousness analysis [463]. (See similarly [468-69].) Because the ‘417 application disclosed that all the disclosed compounds were effective in the treatment of glaucoma and ocular hypertension, and the compounds of the ‘287 patent were encompassed by the ‘417 application, it follows straightforwardly that it was obvious that the compounds of the ‘287 patent would be effective in the treatment of glaucoma and ocular hypertension.
This conclusion is striking. On Kane J’s holding, even if the claimed compounds were in fact substantially more effective than the prior art, and had fewer side effects, and no skilled person could have predicted this, the claims would still be obvious. In my view, this conclusion is wrong because Kane J’s identification of the inventive concept is wrong. In assessing the inventive concept Kane J focused on the fact that the claims did not refer to the purported advantages such as increased efficacy and fewer side effects [167], but, as the SCC has pointed out “The inventive concept of the claims is not readily discernable from the claims themselves. A bare chemical formula in a patent claim may not be sufficient to determine its inventiveness” Sanofi 2008 SCC 61 [77]. As the Court’s example indicates, it is the properties of a new chemical compound that will determine whether it is obvious, but these properties will not be set out in the claims, because if the compound is new, the patentee is entitled to claim the compound itself. For that reason it is necessary to look to the disclosure to identify the inventive concept. (See also COMBIGAN 2012 FCA 308 , blogged here.) It is true that in this case, the claims are to the use of travoprost, not to the compound itself, but the principle is the same: "[p]atents are essentially about information as to what to make or do" Aerotel [2006] EWCA Civ 1371, Jacob LJ [32]), and as Lord Hoffmann said in Biogen [1996] UKHL 18, [14], "[w]henever anything inventive is done for the first time it is the result of the addition of a new idea to the existing stock of knowledge." The quid pro quo for a patent is not the thing claimed, but the information disclosed in the specification: Consolboard [1981] 1 SCR 504, 517. Consequently, it is the information in the patent that makes the invention obvious or not, and information is found in the disclosure. The claims merely define the scope of the patent exclusivity.
I cannot criticize Kane J too much for this, as the concept of the “inventive concept” has been cloudy from the outset: the SCC in Sanofi could not even decide whether it is necessary to identify the inventive concept in order to assess obviousness. As I have suggested before, I am increasingly inclined to think that the European problem-solution approach would provide a more structured way of identifying the inventive concept for the purposes of the obviousness analysis. The problem in this case was clearly to find a prostaglandin analogue with improved properties, and the question should be whether travoprost solved that problem, ie whether it in fact had improved properties.
With that said, Kane J’s decision suggests that travoprost might well have been held to be obvious even if the inventive concept had included its properties. In particular, the ‘417 application disclosed utility for treating glaucoma, the “Woodward” prior art suggested improved properties [418], [435], [472], and the combination might have made the improved properties of the ‘287 patent obvious. That is of course just speculation without an express finding to that effect; I only want to be clear that I am not saying that the ‘287 patent was necessarily non-obvious, but only that the analysis was not properly focused.
