Reasonable Expectation of Success: Between Boston College Doug Flutie “Hail Mary” and Wayne Gretsky “Open Net Shot”

Janssen Inc v Apotex Inc 2021 FC 7 Phelan J

            2,661,422 / abiraterone acetate & prednisone / ZYTIGA / NOC

In this decision Phelan J held Janssen’s 422 patent to be invalid for obviousness in a textbook example of the obvious-to-try analysis [198]. He held that it was not invalid for lack of utility, and that it would have been infringed by inducement if it had been valid.

The main argument was focused on obviousness and specifically the obvious-to-try analysis. The 422 patent related to the combination of abiraterone acetate (AA) and prednisone (PN) for the treatment of a prostate cancer. Both of these were known to be useful in treating prostate cancer (with PN being used for palliation rather than treatment as such) [27], [28], and the key question was whether it would have been obvious to try the combination. Phelan J held that it was obvious to try [161] and the effort needed to achieve success did not rise to the level of inventiveness [193]. Consequently, the 422 patent was invalid for obviousness [198].

In a prior decision under the old NOC regulations, 2019 FC 1355 (see here and here) rejected the invalidity attacks and granted an order of prohibition. Phelan J noted that in this action, brought under new Regulations, he had new and better evidence, and consequently his findings and analysis in that decision were irrelevant to this action: [9]–[11] and see eg [123], [161].

Phelan J’s holding turned largely on the facts, with a particularly nice application of the obvious-to-try analysis, so I’ll just note a few points that caught my eye.

Obvious to try is a factor, not the test

In the obvious-to-try analysis [134]:

whereas being “more or less self-evident to try to obtain the invention” is a requirement for obviousness to try, being “more or less self-evident that what is being tried ought to work” is not a requirement but a factor to be considered.

That is, the claims are not non-obvious merely because success was not self-evident. That is because the obvious-to-try analysis is used precisely in areas where advances are won by experiment, so that success cannot be guaranteed before trying. This point is now well-established, but it is good to see it reaffirmed.

Reasonable expectation of success is somewhere between Doug Flutie and Wayne Gretsky

In particular, the fact that clinical trials are necessary to establish that the invention will work is not determinative.

[191] It would be a concern to courts if the normal work of a clinical trial was held out to be the effort that would bar an “obvious to try” analysis. It would mean that no pharmaceutical treatment which would otherwise be obvious to try would ever meet the Supreme Court of Canada’s fourth step because regulatory approval would trump patent law.

On the expectation of success, Phelan J remarked:

[135] As to “ought to work”, it is clear that certainty of success is not required otherwise there would be no point in describing it as something “to try”. “Trying” implies the possibility of failure but with the expectation of success. While never easy to define on a spectrum of likely success, it is neither a Boston College Doug Flutie “Hail Mary” pass nor a Wayne Gretsky “open net shot”. Some limited experimentation is permitted in the context of the second factor. It is not to be arduous, inventive or unusual.

Not new law, but a fun way to state the test.

Golden bonus

The “golden bonus” principle was also raised, albeit indirectly. The ultimate goal of any cancer treatment is to improve survival, but that is difficult to measure directly in research because the only way to determine survival benefit is for the patient to die [23]. Presumably, gathering statistics on that basis would take too long. So a surrogate in the form of some anti-tumour or anti-cancer activity is used. Janssen argued that the combination of AA and PN was not obvious because there was no evidence of survival benefit for either individually. Phelan J rejected this in part because (my emphasis):

[179] If the search for survival benefits was as important as suggested, the positive anti-cancer effects of each drug outlined above in conjunction with O’Donnell 2004 provided good reason for a POS to combine AA and PN. The inventors do not have to be seeking the same solution as the eventual patent discloses (Hospira FCA [2020 FCA 30], para 94).

I take it that this is related to the “golden bonus” point noted in last Friday’s post: if an invention is obvious to try for one reason, it does not become less obvious because it turns out to have some unexpected advantage.

Apotex got caught in a squeeze in its utility attack, as “much of the evidence on which it relies for Obvious to Try supports the utility of the Patent” [203]. That’s not a bad squeeze to be in, given that they succeeded on obviousness. In holding the invention to be useful Phelan J noted that “ For purposes of utility, it is not necessary to meet Guidelines or FDA approvals. Utility is met if some patients, even if only those in dire circumstances, respond” [209].

Courts should be cautious on patentability of methods of medical treatment

Apotex also raised an attack based on patentable subject matter. Phelan J noted that “The arguments of both parties on this issue of unpatentable subject matters are confusing and difficult to consider” [216]. Perhaps this is because the law itself is confusing, as the FCA recently recognized in Hospira. Accordingly, Phelan J noted that “The issue of method of medical treatment is not a settled one and therefore a court should be cautious in striking down claims on this basis” [223]. This strikes me as salutary advice.

Product Monograph important in establishing inducement

Given his holding on validity, it was unnecessary for Phelan J to consider infringement, but he did so nonetheless, in order to facilitate the inevitable appeal [19]. He found that if the patent were valid, Apotex would infringe by inducement by the sale of their AA product “indicated in combination with prednisone for the treatment of metastatic prostate cancer” [246]. This finding was based in large part on Apotex’s draft Product Monograph, which directed use in a manner consistent with the asserted claims [245]–[260]. This is consistent with many prior cases, but it is nonetheless worth reiterating the importance of the PM in respect of inducement.

Do physicians read the PM?

Apotex apparently tried to get around this by arguing that their product would not be used for anti-cancer treatment, as specified in some of the asserted claims, but for other purposes, such as controlling side effects. Phelan J rejected this, in part because (my emphasis):

[250] In their draft Product Monographs the Defendants cite a number of References (publications such as de Bono 2011, the Attard Reports) which address PN contributing to anti-cancer effects of the combination with AA. PN’s role in mitigating side effects is also contained in the References. It is no answer to say that although the Defendants put the References in the Product Monograph, they do not expect the doctors or pharmacists to consider them.

Despite Phelan J’s apparently rhetorical point, I’ll note that there are a number of cases calling into doubt whether doctors or pharmacists consider the PM itself, particularly the generic PM, much less the references found in it: see Solvay 2008 FC 308 [192]; Aventis 2006 FC 861 [44]; Abbot 2006 FC 1411 [40] affd 2007 FCA 251 [26]; Allergan 2011 FC 1316 [161]. Of course, these cases can be distinguished because there was evidence on this issue, and there does not appear to have been similar evidence in this case. Nonetheless, Phelan J’s remark does make the point that it is very counter-intuitive to argue that while the PM directs infringement, that’s permissible, because no one reads it anyway. On that note, I’ll plug my draft article “Is 'But For' Causation Necessary to Establish Inducement?” which argues that the way to resolve this difficulty is to recognize that “but for” causation is not required in the second stage of the inducement test. This article has been on the back burner for too long, and I hope to return to it soon. Here’s the abstract:

The established Weatherford test for inducing patent infringement requires, as its second branch, that inducement by the indirect party be the “but for” cause of the direct infringement. This article shows that such a requirement of “but for” causation leads to problematic results, particularly when the evidence indicates that the direct infringers do not heed the encouragement provided by the indirect party. The article argues that while the “but for” causation requirement is often stated, it has rarely actually been applied, and there are several cases which suggest a less stringent requirement. This article argues that the best way to reconcile the cases is to recognize that causation in the second branch of the inducement test may be established on the basis of material contribution, rather than “but for” causation, while also recognizing that “but for” causation remains necessary for an award of any monetary remedy. This approach, combining material contribution as the substantive causation requirement at the second stage of the inducement test, with “but for” causation in awarding monetary remedies, also appropriately balances the policy concerns of ensuring effective enforcement of the patentee’s rights while avoiding the problem of overbroad enforcement.

Open question whether claim of invalidity can be consolidated with a NOC action

Finally, Phelan J noted an unsettled procedural point. Apotex has alleged in a counterclaim that claims that were not asserted by Janssen were also invalid. Section 6(3)(a) of the Regulations allows the second person to bring a counterclaim for a declaration of invalidity “in respect of any patent claim asserted in the action.” A counterclaim is a separate action, and while s 6(3) “allows a party the convenience of bringing a challenge to a patent in the context of an NOC action in respect of asserted claims,” it is restricted to the claims asserted in the action [234]. “Whether a separate claim of invalidity could be consolidated with a NOC action remains an open question” [234].