Res Judicata in a Bifurcated Action

Merck & Co Inc v Apotex Inc 2014 FC 883 Lafrenière J
            1,275,350 – lisinopril – PRINIVIL

This motion to amend a Responding Statement of Issues raises a very interesting and difficult issue respecting the application of res judicata to a bifurcated action when the law has changed between the liability and damages portions, though at the end of the day I doubt the resolution of that question would make any difference to the outcome.

In the underlying Lisinopril Liability decision, 2006 FC 524 aff’d 2006 FCA 323, Apotex argued that the ‘350 patent was invalid because it was improperly divided from the parent ‘340 application, which disclosed only one invention. Hughes J rejected this on the basis that he was bound by prior case law, including Boehringer (1962), 39 CPR 201, to hold that each claim constitutes a different invention [116]. The FCA affirmed on this basis [39]. Subsequently, in its Viagra decision 2012 SCC 60 [57], the SCC held that Boehringer “does not stand for the proposition that every claim in a patent application is a separate invention.” In this motion Apotex argued that the law having changed – or at least, the Liability decision having been based on a misunderstanding of the law – it should be able to amend its defence in the damages portion of the action to argue that the Merck did not suffer any damages as a result of Apotex’ activities [28.12].

Essentially, the argument is that even though Apotex has been found liable for infringement, it should not be made to pay damages for infringement of an invalid patent. In support, Apotex relied on the UKSC decision in Virgin Atlantic [2013] UKSC 46. In Virgin Atlantic, the EWCA had held the patent at issue to be valid and infringed, and ordered an enquiry as to damages [11]. Subsequently, the EPO held the patent to be invalid [12]. The effect of the EPO decision is that the patent on which the liability decision was based is deemed never to have existed [27]. However, the EPO decision could not be raised by way of an appeal of the liability decision, as the order of the EWCA was final and not subject to appeal [16]. The question before the UKSC was whether the defendant should be able to raise the invalidity in the enquiry as to damage [16]. The UKSC held that the invalidity could be considered in the damages enquiry [37].

While Virgin Atlantic is very similar to the Lisinorpil situation, Lafrenière J distinguished it on the basis that it turned on the peculiarities of the European patent system in which both the EPO and the English courts have jurisdiction over European patents [52]. Lafrenière J held that as a matter of Canadian law “whether the Supreme Court subsequently changed the law upon which the Liability Judgment was based is irrelevant. A change in the law is not a sufficient basis to re-open previously decided cases” [49].

To my mind, this is a difficult point. It is true that Virgin Atlantic itself can be distinguished on its facts, as noted by Lafrenière J. However, the UKSC expressly approved the decision of the House of Lords in Arnold v National Westminster Bank [1991] 2 AC 93. The dispute in Arnold turned on which of two possible interpretations of a rent review clause in a lease was correct. The issue first arose on the first rent review date. After arbitration, the matter went to court, where Walton J interpreted the clause one way, and leave to appeal to the EWCA was refused, so that action was final. In subsequent cases between different parties, similar clauses were consistently construed in the opposite way, so that by the time the second rent review date arose, there were “powerful grounds for the view that Walton J. wrongly construed the rent review clause in the parties' lease” (p103). The plaintiffs therefore sought to re-open the question of the construction of the same clause which had been decided as between the same parties in the earlier action before Walton J. The defendants argued they were estopped from doing so. The House of Lords held for the plaintiffs, on the basis that the change in the law was a material circumstance allowing the same point to be re-argued in a subsequent case. As the UKSC explained in Virgin Atlantic, this was different from the classic case of res judicata, as illustrated by the old leading case of Henderson v Henderson (1843) 3 Hare 100, in which the parties were estopped from raising issues which could have been raised in a previous proceeding between them. The plaintiff in Arnold “had not failed to bring his whole case forward before Walton J. On the contrary, he had argued the very point which he now wished to reopen and had lost” [20]. The plaintiff was permitted to re-argue the very same point because of “materially altered circumstances” which could not have been raised in the first instance. So, according to the UKSC in Virgin Atlantic both cause of action estoppel and issue estoppel bars the raising in a subsequent proceeding issues which were not raised in the earlier proceeding “if they/it could with reasonable diligence and should in all the circumstances have been raised” [22]. As illustrated by Arnold, a point which could not have been raised with reasonable diligence includes a subsequent change in the law.

Turning to Canadian law, I am not convinced that the cases cited by Lafrenière J for the proposition that “[a] change in the law is not a sufficient basis to re-open previously decided cases,” are really determinative. Both AB Hassle 2008 FC 184 [39-40], aff’d 2008 FCA 416, and Metro Can Construction 2001 FCA 227 [4-5], concerned motions for reconsideration of a prior decision under Rule 399(2)(a). Reconsideration of a previous final decision is different from re-arguing an issue in on-going litigation. In Regie des rentes du Quebec 2013 SCC 46, McLachlin J at [55] cited Roberge v Bolduc, [1991] 1 SCR 374 for the proposition that “the authority of res judicata exists even when there is an error in the judgment.” But in Roberge again, the question was whether a final judgment could be challenged in light of the need for finality: “It is a social necessity of the first order that legal proceedings should not be started over and over again on the same matter,” and “The purpose of the legislature has been to ensure that proceedings being perpetually started over again would not compromise the security and stability of social relationships.” In the case of a bifurcated proceeding, the problem of starting proceedings “over and over again” is not as directly in issue, since the damages proceeding will go forward in any event.

With all that said, the SCC has approached res judicata, including both cause of action estoppel and issue estoppel, as balancing the need for finality and avoidance of duplicative litigation, against the interests of justice between the parties: Danyluk 2001 SCC 44 [18-20]. When the law has changed between the liability phase and the remedies phase of a bifurcated action, it is a very close call as to which way the balance tips, as is illustrated by the fact that Virgin Atlantic reversed settled UK law. In Virgin Atlantic the UKSC indicated that the balance would only tip in favour of finality at the point where the equiry as to damages was concluded, so that there was “a final judgment for a liquidated sum” [36]. But there is also a good functional argument for drawing the line earlier, after the liability decision in a bifurcated proceeding. Strictly, as I have noted above, the damages reference will proceed in any event, so we don’t have exactly the same problem of starting a new legal proceeding, but as Lafrenière J noted, to allow Apotex to raise the validity issue in the context of the reference on damages would transform the nature of that proceeding [56]. Moreover, even if the UKSC in Virgin Atlantic drew the line in the right place on its particular facts, there is a real functional difference between allowing a party to raise the established legal fact of the invalidity of a patent, as in Virgin Atlantic, and allowing a party to re-argue the invalidity of a patent based on a change in the law, as would be required in this case.

In summary, while I certainly see the force of Lafrenière J’s concerns, I am not sure that on the current state of Canadian law, it can be said that it is clear that Apotex’s proposed pleading discloses no reasonable cause of action.

But even if the courts were to hold that changed law related to validity of a patent could be raised in the remedial portion of a bifurcated action, I’m not sure that would help Apotex in this particular case. While both the FC and the FCA in Lisinopril Liability held that the divisional was not improper, both also held that even if the divisional was improper, invalidity is not a consequence. As Hughes J put it:

Division of a patent application is essentially a procedural matter. If several patents claiming the same invention have been granted, a sufficient remedy as to validity exists in the application of the principles of double patenting.

While Hughes J’s reasoning on this point was brief, he was specifically affirmed by the FCA after an extensive analysis of the relevant provisions [40]-[50], which reasons, I might say, strike me as being entirely persuasive. All that can be said against this is that the FCA holding on that point was strictly obiter in light of the holding that the divisional was not improper, but clear and considered obiter carries a great deal of force. My inclination is to think that while Lafrenière J did not emphasize this point, he did address it, and it seems to me to provide a sound basis for his holding that the amendment should not be allowed.