The Relationship Between Utility and Obviousness

Hoffmann-La Roche Limited v Apotex Inc / MMF (NOC) 2011 FC 875 O'Reilly J

Regular readers of this blog will recall that I have been critical of the “false promise” doctrine, which holds that utility is measured by the promise of the patent. In the MMF case, the patentee, Roche, directly challenged this doctrine, arguing that an invention need only satisfy the minimum general requirement for utility, and not whatever greater utility might have been promised in the description. O'Reilly J rejected Roche’s submissions on this point. Unfortunately, the nature of the invention in this case complicated the argument by implicating difficult questions regarding the treatment of utility in selection patents, and I will argue in this post that O'Reilly J's reasons for affirming the false promise doctrine are closely related to issues raised by selection patents. This means that in order to revise the false promise doctrine, it will first be necessary to clarify the law relating to selection patents.

The claim at issue was to the compound MMF. MMF is an ester of MPA, a known compound used as an immunosuppressive drug. MMF acts as a prodrug; that is, it is metabolized to MPA once ingested. It was developed in an attempt to improve on the limited bioavailability of MPA. The interesting question of law was whether utility was established only if MMF was superior to MPA as an immunosuppressant by virtue of its better bioavailability, or whether it was sufficient to establish that MMF was useful as an immunosuppressant (which it clearly was). O’Reilly J adopted the standard view that utility is measured by the promise of the patent. He therefore construed the patent to determine the promised utility, and concluded that it promised an improvement over MPA. On the facts, he upheld the patent nonetheless, as being useful even as measured by the promise of the patent.

Roche had argued that “a patent holder can rely on the advantages described in a patent to show that the invention was not obvious, even though those advantages are not relevant to utility” [20] and that “these two inquiries are completely separate and unrelated” [21]. O’Reilly J rejected this argument [21], primarily on the basis that “where advantages form part of the stated invention, it would be unfair to allow the patent holder to rely on those advantages to show that the invention was unobvious and, at the same time, dismiss those advantages as being irrelevant to utility. A patent holder cannot read up the invention for obviousness and read it down for utility” [22].

As an initial point, it is clear that obviousness and utility are often “completely separate and unrelated.” On the facts of this case there was no particular difficulty in creating MMF, and the inventive ingenuity lay in the discovery that of all the hundreds of candidate compounds related to MPA, MMF was one which had improved bioavailability. O’Reilly J construed the patent to promise improved utility, with the consequence that the utility and obviousness inquiries were related. But suppose that the creation of MMF had required great ingenuity, perhaps because there was no known way to create an ester, and the patentees had developed a new reaction. Whether the MMF had better, worse, or the same properties as MPA would be entirely unrelated to whether it was a difficult compound to make. In such a case MMF could be the subject of a valid patent, so long as it had some utility as an immunosuppressant, even if it was worse than MPA itself, according to the well-known maxim that “it is sufficient utility to support a patent that the invention. . . affords the public a useful choice” (Consolboard [1981] 1 SCR 504 at 525).

With that said, O’Reilly J did not say that obviousness and utility are necessarily related. Rather he said that “where advantages form part of the stated invention,” it would be unfair to separate the two. I suggest that this point is fundamentally related to the debate over selection patents. On the facts, the MMF patent is very similar to a selection patent. A selection patent consists of a selection of a species from a genus where the genus is typically a well-defined class of compounds with related properties described in a prior patent. It will necessarily be obvious how to make all of the compounds in the genus, or the prior patent would be invalid for insufficiency. All of the compounds in the genus will also share a common utility, and the inventive ingenuity in the selection patent is identifying, from among all of the compounds in the genus, a species with improved properties. The MMF patent is essentially of this nature, except that the prior “genus” was not a well-defined class, but rather the open ended class of all easy-to-make MPA derivatives that might function as prodrugs. Nonetheless, the fundamental question was the same as in a selection patent: which of this class would have advantageous properties. Roche conceded that a higher utility requirement may arise in selection patents, and the “substantial advantage” branch of Maugham J’s I. G. Farbenindustrie test for selection patents has been related to the utility requirement by the FCA in Eli Lilly Canada Inc v Novopharm Ltd / olanzapine 2010 FCA 197 at [90], [99]. With this in mind, O’Reilly J’s holding amounts to the observation that in the case of selection patents, utility and obviousness are related. It does not necessarily imply that utility must always be measured by the promise of the patent.

This brings us to the question of why selection patents are different. I would argue that they are not. This is not because utility and obviousness are always related, but because they are always separate; the "substantial advantage" branch of the selection patent test is not related to utility, as the FCA has indicated, but rather to obviousness. So, in this case O’Reilly J stated that “[w]ithout those qualities, MMF might technically still be a prodrug, but it would not be a useful prodrug” [29]. I would suggest that without those qualities, MMF would be a useful prodrug, but it would not be an inventive prodrug. I acknowledge that this is an uphill argument. The view that advantages of a selection patent go solely to utility was argued by Lilly before the Court of Appeal in the olanzapine decision ([26]) and was rejected by the Court at [90]. Accordingly, it is understandable that Roche would have conceded that a higher standard of utility required in that context. But one this concession was made, O'Reilly J's conclusions follow.

This is not the place for a full analysis of the olanzapine decision. For now I will conclude with the observation that the MMF decision shows that the debate over selection patents has direct ramifications for the false promise doctrine.