Triple Re-Litigation of Factual Issues

Apotex Inc vPfizer Canada Inc / azithromycin (NOC) 2013 FC 493 O'Reilly J
             ZITHROMAX / azithromycin / 1,314,876

O’Reilly J’s decision, which I will call Azithromycin AD (No 2), relates to a claim by Apotex for damages under s 8 of the PM(NOC) Regulations, consequent on Snider J’s dismissal of Pfizer’s application for an order of prohibition in the s 6 proceeding, Pfizer Canada Inc v Apotex Inc, 2003 FC 1428 (Azithromycin AD (No 1)). The key issue is whether a prior factual finding of non-infringement in the s 6 proceeding can be re-litigated in the s 8 proceeding. The answer is yes (although on the facts O’Reilly J arrived at the same conclusion as had Snider J, namely that the patent was not infringed). This means that the same factual issues may be litigated three times: once in the s 6 application, again in the s 8 action, and again in the infringement action. "Seemingly redundant litigation" is already said to be a problem with our version of the patent linkage regime: 2012 FC 767 [69]-[70]. This decision, if it stands, will add to the problem.

The ‘876 patent claims a particular form of azithromycin, namely crystalline azithromycin dihydrate (AD). In 2001 Apotex sought an NOC permitting it to market tablets of azithromycin isopropanolate monohydrate (AIM). In the s 6 proceeding the sole issue was non-infringement. It was not contested that AIM would not infringe the ‘876 patent, and the question was whether Apotex’s product would contain AD, either as a byproduct of initial production, or by conversion of AIM to AD on storage: Azithromycin AD (No 1) [46]. On the evidence before her, Snider J held that Pfizer had not carried its burden of proving infringement by either of these routes [81].

Apotex consequently brought a claim under s 8. This decision by O’Reilly J in Azithromycin AD (No 2), does not concern quantum, but only Apotex’s entitlement to s 8 damages. Pfizer argued that Apotex was not entitled to damages because the evidence showed that the product actually marketed by Apotex did indeed infringe [3]. That is, Pfizer sought to re-litigate, with new evidence, the issue that had been decided against it in Azithromycin AD (No 1). For its part, Apotex argued that s 8 created an entirely free-standing cause of action, separate from the s 6 proceeding, so that Apotex could respond by arguing that the patent was invalid, even though invalidity was not alleged in the s 6 application: [26].

O’Reilly J held that based on the concept of stare decisis and judicial comity, he was bound by Snider J’s construction of the ‘876 patent (unless there were compelling reason to depart from it, which there was not). Since claim construction is considered to be a matter of law, this is consistent with the FCA decision in Apotex Inc v Allergan Inc / COMBIGAN (NOC) 2012 FCA 308 [48] that a Federal Court judge is generally bound by conclusions of law of another Federal Court judge. The point is particularly strong in this case where the same parties are involved, so that each had the chance to makes its arguments in the original proceeding.

O’Reilly J also held s 8 does not create a free-standing right of action, as that “would unfairly expose first persons to allegations that could not have been anticipated when making the decision whether to apply for an order of prohibition under s 6" [21]. This seems to be saying that a patentee who contested an allegation of non-infringement might have chosen not to apply for an order of prohibition if the validity of the patent would be put in question. Consequently, the s 8 action is circumscribed by the issues put in play in the s 6 application. Entirely new allegations of non-infringement or invalidity cannot be raised [22].

However, O’Reilly J held that the factual question of infringement was different, so that “Pfizer is entitled, now that Apotex’s product is actually on the market and can be tested, to introduce evidence showing that Apotex may have actually infringed the ‘876 patent” [23]. This is generally consistent with the FCA’s statement in COMBIGAN that “the principle of judicial comity . . . has no application to factual findings” [44]. O’Reilly consequently considered the evidence that Apotex’s product did infringe, though in the end he held that it did not [55-78]. Consequently, Apotex is entitled to s 8 damages, with quantum to be determined separately [79].

While O’Reilly J’s distinction between legal and factual matters is consistent with the FCA’s COMBIGAN decision, I do not think it is compelled by that decision. In COMBIGAN the FCA addressed the principle of comity in general terms, where it may be invoked in distinct proceedings between different parties. The close relationship between s 6 and s 8, rightly emphasized by O’Reilly J, might imply a more stringent form of comity. In particular, it seems to me that the question may not turn on general principles of comity, but rather on the interpretation of those statutory provisions. Section 8 provides for damages “If an application made under subsection 6(1) is . . . dismissed by the court hearing the application.” It may be that this means that s 8 provides for damages conditioned on the s 6 dismissal alone. If that is so, the statute would of course displace general principles of comity.

From a policy perspective, O’Reilly J’s holding that factual issues may be re-litigated in the s 8 proceedings is driven by the observation that it would be wrong to allow the generic to recover damages for being kept out of the market with a product that in fact infringes. This was express in Pfizer’s argument that recovery in such a case would be barred by ex turpi causa: [3], [24]. This is certainly an understandable and legitimate concern, and it is now clear that a finding in an infringement action that the patent in question is valid and infringed can be raised in a s 8 action: see AstraZeneca v Apotex / omeprazole (NOC) 2013 FCA 77 aff'g 2012 FC 559 (blogged here and here) and Apotex v Merck / lovastatin (NOC) 2012 FC 620 (blogged here). However, it does not necessarily follow that procedurally, these questions can be re-litigated in the s 8 action itself, as opposed to a separate infringement action. Wasteful re-litigation between the NOC proceedings and the infringement action is already a problem that is inherent in our particular system of patent linkage. The approach adopted by O’Reilly J means that exactly the same factual issues may end up being re-litigated three times, once under s 6, a second time under s 8, and a third time in an infringement action.

The counter-argument is that in principle there may be no infringement action, in which case the generic might recover s 8 damages even if the product were in fact infringing. But I suspect that situations in which a patentee that has been unsuccessful in the NOC proceeding, then chooses not to litigate a patent which it believes to be valid and infringed, will be relatively rare in practice. What is normal is that a generic will seek s 8 damages against that unsuccessful patentee, who in turn will bring an infringement action, thus giving rise to the triple re-litigation of the factual issues. It may be that a patentee who has been unsuccessful in the NOC proceedings will choose not to bring an infringement action if it believes that it is unlikely to be successful, but in that case the lack of an opportunity to re-litigate the factual findings is not a cause for concern.