An End of the Statute of Monopolies x PMNOC Saga

Apotex Inc v Eli Lilly Canada Inc (Olanzapine), 2022 ONCA 587, leave to appeal ref’d 2023 CanLII 33987 (SCC)

In a decision that did not attract much fanfare, the Supreme Court of Canada recently dismissed Apotex’s leave to appeal its long-running Statute of Monopolies claim against Eli Lilly. The Court’s refusal to grant leave means that, for all intents and purposes, the chapter has closed on this intriguing (if somewhat outlandish) legal theory.

It also means that some questions remain unanswered by the Ontario Superior Court of Justice and the Ontario Court of Appeal in their decisions below. In particular, whether a patent is truly void ab initio when it is declared invalid by a court, and to what extent does the Statute of Monopolies continue to apply in non-PMNOC contexts?

The Zyprexa Cases

The product at issue – olanzapine (Zyprexa) – can lay claim to being the poster-child of 2000s Canadian pharmaceutical patent litigation, right next to clopidogrel (Plavix). This drug went through multiple rounds of PMNOC applications (resulting in inconsistent rulings on validity), infringement actions, impeachment proceedings, s. 8 claims, appeals up and down the courts, and even an international arbitration via NAFTA where foreign lawyers opined on the validity of Canadian patent doctrine. Eli Lilly has said (and no doubt would continue to say) that it was among the first and most aggrieved victims of the now-rejected “promise doctrine”.¹ Against this backdrop, one might be tempted to read the Court(s)’ reasons as though they were looking for ways to rid itself of this case and other like cases once and for all.

Eli Lilly’s now-infamous “113 Patent” was both a “selection” patent identifying olanzapine from among a previously-known genus of compounds, and a “use” patent claiming inter alia the use of olanzapine for treatment of schizophrenia.² Eli Lilly had succeeded in its PMNOC prohibition application against Apotex,³ but failed against Novopharm.⁴ This resulted not only in inconsistent findings, but led to a subsequent infringement action against Novopharm (but not Apotex) in which, over the course a trial, two appeals, and a reconsideration, Eli Lilly’s 113 Patent was held invalid only on the basis of inutility.

Statute of Monopolies in lieu of Section 8

Apotex, of course, was not content to sit on the sidelines while Novopharm got the benefit of being on the market. In 2013 Apotex sued Eli Lilly for damages under the 17th-century English Statute of Monopolies, among a plethora of other grounds. This was one among several cases where Apotex pursued this novel legal argument.

In short, the argument goes that the Statute prohibits all “illegal” monopolies; that the Statute was received as law into Upper Canada (Ontario), and then into Canada; that it was never repealed; that it continues to apply notwithstanding any other statutory regime enacted by Parliament; that listing a patent under the PMNOC Regulations where that patent is subsequently invalidated was an exercise of an illegal monopoly; and that exercise of this illegal monopoly results in treble damages according to the Statute.

This intriguing series of cases survived several different motions to strike, and though it seemed they would never be decided, they also seemed “to big to settle” given the damages claim being asserted.

In some ways this case had “bad facts”: since Apotex lost the prohibition proceedings, they had no claim under s. 8 of the PMNOC Regulations.⁵ Thus, there was always the “bad optics” of appearing to circumvent the Regulations to obtain a remedy that was not otherwise available. On the other hand, one could argue (as Apotex did) that there was an insufficient remedy for the harm caused to it in being delayed or denied to market by a patent that was ultimately invalid, though its allegations of invalidity did not succeed at first instance.

Given those facts, it should be somewhat unsurprising that the primary basis on which the motion judge at summary judgment dismissed the case on the ground that the PMNOC Regulations constituted a “complete code”. Since patent law is regarded as “wholly statutory” and the statutory regime created by Parliament did appear to contemplate the situations of improper listing, remedies for failure on a prohibition application, and rights of appeal (at least for the aggrieved generic party), it did seem that any attempt to do an end-run around the Regulations was doomed to fail.

However, the “complete code” argument is in many ways unsatisfying. For one, it didn’t truly address any of the underlying tensions of the pre-2017 Regulations including its likely-unintended consequences of dual-track litigation resulting in inconsistent findings.⁶ Nor, perhaps, would it have been likely that Parliament contemplated the scenario where one generic succeeded on allegations of invalidity while another did not.⁷

Whither the problem of “ab initio” invalidity

More interestingly, or frustratingly, the “complete code” argument entirely ducks the question of whether a patent is void “ab initio” when it is invalidated by the Court. Although on its face it appears to be an academic question, this issue could have had broader impacts on other areas of patent law and not just as the foundation of a Statute of Monopolies claim. ⁸

The distinction between being a patent being void “ab initio” and invalid as of the date of judgment appears to be based in Canada on a passage found in Collette v Lasnier, (1886) 13 SCR 563, where the Supreme Court invalidated a patent as a colorable imitation of an earlier patent (perhaps today it would simply be called anticipation):

The consideration given for a patent is a warranty that all is new which the applicant seeks to protect; otherwise a party by getting a patent would obtain protection at the public expense for an alleged invention which already was in public use. The consideration is entire and covers everything in the patent and specification, and if it fails as to one or more parts the alleged invention, it fails for all, and the patent is therefore void. It is not voidable merely but ab initio void. If void, no action can be maintained for any infringement of it, even if the part of the invention to which the alleged infringement refers was new. (Emphasis added)
Collette v Lasnier, (1886) 13 SCR 563 at 567

However, this passage appears in Henry J.’s judgment where he dissented in part, although it is difficult in these 19th-century judgments to tease out exactly what the “ratio” of a case is since each judge invariably gave their own reasons with differing bases for the outcome.

In Collette there was no practical impact of the patent being void ab initio, rather than merely “voidable”, since the only result of the judgment was that no infringement action could be maintained on the later (invalid) patent. Arguably the Court was simply saying that the patent could, and was, impeached in rem, and nothing turned on the timing of its invalidity. Put another way, the result would be the same if the later patentee sued for infringement and the patent was found to be invalid at that hearing–the action would simply be dismissed.

At least one subsequent case, however, shows why ab initio invalidity matters. In Bergeon v De Kermor Electric Heating Co (1927),⁹ the Exchequer Court refused to grant a reissue patent because the original specification was deemed to be defective by reason of insufficiency. Since the original patent “should never have been granted”, the underlying patent was found “null and void ab initio”. As a result, no underlying patent existed to be reissued. In other words, a patent that was invalid to begin with cannot be cured by post-grant procedure such as reissue, and analogously, re-examination. This is not merely a technical problem since it would arguably remove substantive rights of a patentee under the relevant post-grant portions of the Patent Act. ¹⁰

While I am aware of no cases addressing these points, other plausible scenarios where this distinction might make a difference includes: whether a licensee can claim against a patentee to reclaim past license payments where (perhaps due to third party litigation) the patent is later found invalid; or whether an invalidated patent can be cited as prior art–including whether there is a distinction to be made as between a citable co-pending application under s. 28.2 or merely as a publicly available reference.

Indeed, it is not even fully clear how far ab initio validity goes–does it mean only that the rights granted by the invalid patent are deemed never to have been granted, or does it mean that the entire patent is deemed never to have existed? The earlier cases (cited above) would suggest the former, yet, as matters stand, Olanzapine suggests the latter. In the underlying summary judgment decision affirmed by the Ontario Court of Appeal, Schabas J. wrote:

[132] Additionally, if the patent is void ab initio, as Apotex emphasizes, then it is treated as having never existed. Apotex, however, wants it both ways, saying that Lilly never had a patent, but that Apotex was harmed by the patent. This cannot be right. If Apotex wants to say the patent never existed, then Lilly was never granted a monopoly, let alone an unlawful monopoly, and the prohibition has no application. If one is going to rewrite history, it should at least be done consistently.
Apotex Inc. v. Eli Lilly Canada Inc., 2021 ONSC 1588 at [132] (bold added)

This passage is also likely obiter given that the operative finding was that the act of listing a patent on a Form IV was not an illegal act, but it suggests that if a patent is void from the start, then everything else that follows must also be erased in any “but-for world” construction.

The Ontario Court of Appeal neither endorsed nor rejected this premise. Instead, to my mind, it expressly left open this question when it affirmed the motion judge’s decision. The entirety of ONCA’s reasoning is contained in three short paragraphs:

[46]   Monopolies flowing from patents for new inventions are explicitly excluded from liability under s. 5 of the Statute of Monopolies. There is no question that at the time Eli Lilly’s 113 Patent was granted, it was a patent for a new invention.
[47]   The Statute does not distinguish between valid and subsequently invalidated patents. As the motion judge noted, correctly in my view, this is in keeping with the historical purpose behind the English Statute of Monopolies, first enacted in 1624, upon which the Ontario Statute is based. Parliament passed the Statute of Monopolies in an attempt to limit abuses by the Crown in granting “letters patent”, not “patents of invention”. The Statute was passed in response to the Crown granting letters patent to operate or regulate industries, or to have others act as agents of the Crown in operating monopolies for trade and industry, independent of merit or invention.
[48]   The motion judge’s reliance on Peck v. Hindes, a decision of the Queen’s Bench division of the English High Court of Justice, was not misplaced. As the motion judge noted, the holding in Peck v. Hindes that the Statute of Monopolies “applies in its terms to invalid and improper exercises of the Royal Prerogative, and not to Letters Patent which were perfectly legitimate and protected by law”, is consistent with the motion judge’s reasoning that there was nothing illegitimate or unlawful in the granting of Eli Lilly’s 113 Patent.
Apotex Inc. v. Eli Lilly Canada Inc., 2022 ONCA 587 at [46]-[48] (bold added)

As a result of this brief analysis, it cannot be said that the Court set any definitive precedent as to the question of ab initio validity, leaving all of the above questions unanswered.

Thus the ultimate result is rather anticlimactic given the questions raised. This fact pattern is unlikely to ever be repeated, given that it was uniquely created by the “dual litigation” regime of the pre-2017 PMNOC Regulations. Perhaps that is why the Supreme Court declined to wade into the fray.

1
Interestingly, Eli Lilly’s ultimately unsuccessful application for leave to appeal of the decision invalidating its patent on the basis of the promise doctrine yielded a rare in-person hearing before a 3-justice panel of the Supreme Court of Canada in 2013, including Justice Rothstein. Although no reasons are given for dismissal of a leave application, if one watches the hearing, one gets the distinct feeling that the Court was even then itching to take on the promise doctrine but felt that the decision below at the FCA (comprising a grand total of two sentences) was too thin to appeal from. The Supreme Court ultimately rejected the “promise doctrine” as unsound policy four years later in AstraZeneca, 2017 SCC 36.
2
Technically, it was a “Swiss claim” for the use of olanzapine for the manufacture of a drug for the treatment of schizophrenia, which is a subject for another day.
3
2007 FC 455
4
2007 FC 596.
5
In subsequent cases in respect of other drugs, Apotex sued under both s. 8 and the Statute of Monopolies together.
6
It is often said that Parliament intended the prohibition applications, taking place with a paper-record, no live witnesses, and mostly out-of-court except for a 2-3 day hearing, was supposed to result in speedier dispositions, though it is not at all clear that Parliament ever turned its mind to the consequences of a dual-stream litigation, the asymmetrical rights of appeal, and all of the possible outcomes if a summary disposition upon application was inconsistent with a final disposition upon a full action. The resulting years of litigation put to bed any notion of speedy justice.
7
The Federal Court subsequently developed something of a rule applying the doctrines of comity and abuse of process whereby a subsequent hearing on the same validity allegations would follow the result of the first hearing, unless there were different arguments or “better evidence”. In practice there were very few cases where two parties alleging invalidity obtained different results. This is to be distinguished from two generics getting different results on non-infringement allegations, which should be an expected outcome for non-compound patents.
8
Unrelatedly, the distinction between ab initio invalidity and judgment-date invalidity is of prime importance in family law and constitutional law, as illustrated by the tribulations of Henry VIII and through the entire period of Tudor England.
9
[1927] Ex CR 181
10
This issue is even more interesting in the context of UK or European patent practice, where it is sometimes possible to amend claims during litigation even up to trial to overcome prior art or to avoid ambiguity–arguably, a patent that was originally invalid could not be amended in this manner.

Ordinary Skill