The Top Line
A recent Federal Court decision prevented a generic from pleading a prior art-based defence, based on a fact finding from a case with different parties.
Are prospective generics now collectively limited to one kick at the (in)validity can? And who gets to take the kick? Better hope they're a good striker.
The Federal Court recently rendered an interesting decision on pleading amendments in Janssen v Pharmascience (Macitentan), 2025 FC 718. Although seemingly a simple pleadings motion, there was a notable result that appears to depart from foundational doctrines of res judicata and judicial comity: Pharmascience was held to a factual finding from a case to which it was not party.
The (simple) background of the case: Janssen claims Pharmascience infringes its patent, which claims a combination composition: macitentan and a PDE5 inhibitor. Pharmascience initially pleaded that it does not infringe because its product is a monotherapy, not a combination. But, through its proposed pleading amendments, it sought to plead that it just practices prior art allegedly teaching the combination therapy. In patent parlance, Pharmascience wanted to take advantage of the "Gillette Defence"—if I infringe, your patent is invalid.
The Court refused the proposed amendment. The Court said Pharmascience tried to take inconsistent positions, i.e., that it practices only monotherapy but also practices the combination prior art.
The Court acknowledged that the inconsistency could be addressed by making the allegations in the alternative, as parties regularly do. But the Court denied the amendments anyway.
The reason? A factual finding from an earlier proceeding with different parties concerning the same prior art (Janssen v Sandoz, 2022 FC 715). Based on that decision, the Court concluded "[t]here has been a judicial determination, upheld on appeal" that the prior art does not teach the combination therapy.
The Court therefore denied the proposed amendments. The Court effectively held the prior "judicial determination" made it "plain and obvious that the proposed amendments do not disclose a reasonable cause of action".
The Court's short decision did not provide a clear basis for this finding. In general, the Court has a few doctrines in its toolbox which deal with prior judicial determinations. But, here, none appear to fit.
First, we have res judicata. More specifically, issue estoppel. Issue estoppel has three requirements: (1) the same question was decided, (2) the prior decision was final, and (3) the parties to the prior proceeding were the same. Here, the third factor is missing—Pharmascience was not a party to the prior action that "judicially determined" what the prior art teaches. Issue estoppel cannot support a finding that it is "plain and obvious" that Pharmascience's proposed amendments will fail.
Second, we have judicial comity. Judicial comity seeks to prevent the same legal issue from being decided by different members of the same Court, thereby promoting certainty. But, importantly, judicial comity only applies to determinations of law. It has no application to factual findings1 or findings of mixed fact and law—including in the PMNOC context.2 Here, the issue is whether a piece of prior art discloses something. This is not a question of law and judicial comity should not apply. Like issue estoppel, judicial comity cannot support the Court's "plain and obvious" finding. At best, findings made might be persuasive, but questions of fact or mixed fact and law "must be determined anew based on the evidence adduced" in the second proceeding.3
Here's the issue: the Court found the amendment would fail, because it alleged a fact determined in another proceeding with different parties. The well-established guardrails on res judicata and judicial comity dictate the opposite result—neither captures the prior fact finding.
This decision may have been influenced by the context. Or by judicial realism. First, Pharmascience had not previously alleged invalidity, and the Court did not seem impressed by a brand-new Gillette Defence (the Court called it a "radical departure" from the existing pleadings). Second, this was a PMNOC proceeding where invalidity contentions are typically delivered up-front in a Notice of Allegation, which kick-starts a streamlined process to be concluded in about two years.
However, the decision suggests that a factual finding about prior art is binding on different parties that may happen to later litigate the same patent. In the PMNOC context, this would make a first proceeding (within a set of proceedings relating to the same patent) potentially determinative of all that follow, at least as concerns validity over the prior art. Historically, generic companies in PMNOC proceedings have not been bound by findings of other decisions against other generic companies—they have been free to make their own independent allegations of non-infringement and invalidity. This decision appears to limit these rights.
Footnotes
1. Allergan Inc. v Canada (Minister of Health), 2012 FCA 308 at paras 44-45, 50.
2. Amgen Inc. v Pfizer Canada ULC, 2020 FC 522 at paras 167-169.
3. Bayer Inc. v Cobalt Pharmaceuticals Co., 2016 FC 1013 at paras 52-54.
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
