Limits on Use of Foreign Prosecution History

Canmar Foods Ltd v TA Foods Ltd 2021 FCA 7 de Montigny JA: Pelletier JA, Rivoalen JJA affg 2019 FC 1233 Manson J

            2,582,376 / method for roasting oil seed

Yesterday’s post the summary judgment aspects of the Canmar appeal; this post turns to the issue of s 53.1, which permits the use of prosecution history to aid in claim construction in certain circumstances. Justice Manson’s decision at first instance was the first to interpret s 53.1, and a difficult issue was raised right off the bat. The defendant, TA Foods, wanted to refer to the US prosecution history, while s 53.1 on its face only permits use of communications between the patentee and “an officer or employee of the Patent Office”—which is to say, the Canadian prosecution history. As discussed here, Manson J held that foreign prosecution history should be admissible to aid in a construction of Canadian claims only in “extraordinary circumstances” [73], [74], [77]. In general, this means that “prosecution of the foreign application is made part of the prosecution history of the Canadian patent” [77, original emphasis]. While Manson J did consider the foreign prosecution history [85]–[90], he expressly stated that he would have come to the same conclusion in any event [79].

Justice de Montigny ultimately held that Manson J should not have referred to the foreign prosecution history on the facts in this case [77]. The FCA’s restrictive approach to the use of foreign prosecution history reflects the text of s 53.1, but at this point we cannot rule out the possibility that foreign prosecution history might be admissible if made part of the Canadian prosecution history with sufficient particularity. That is a question “left for another day” [74], and de Montigny JA emphasized that “I wish to express no firm views on the matter” [77].

More generally, de Montigny JA’s decision for the FCA provides a useful discussion of a number of issues, but ultimately refrained from deciding most of them, on the view that “courts [should] refrain from deciding beyond what is strictly necessary for the resolution of the case of which they are seized” [74]. In my view this is a salutary principle—I often wish the SCC would pay more heed to it. The result is that the decision illuminates the FCA’s thinking on this issue, while leaving the specifics to be developed in future cases that raise them more directly. This is the way the law should develop.

Value of foreign jurisprudence

Justice de Montigny noted that s 53.1 appeared to respond in part to disquiet expressed in some FC decisions about always excluding consideration of prosecution history, especially in some particularly egregious cases [51]–[53], and he also noted that “it can also be perceived as a step to better align Canadian law with its British and American counterparts” [53]. That is true in broad strokes—as de Montigny JA noted, US law of course has a well-developed doctrine of prosecution history estoppel [54]–[56], and even the UK has made some steps in that direction [57]–[62]. With that said, I do not see this as an invitation to draw detailed guidance from US or UK law (which are very different from each other), as de Montigny JA noted that s 53.1 “is carefully tailored, and it would go against statutory interpretation principles to try to go beyond its original intent” [70]. Our law relating to prosecution history will therefore turn on the interpretation s 53.1.

Need there be a specific representation?

Part of that careful tailoring is that prosecution history may be admitted “to rebut any representation” made by the patentee in the action. Justice de Montigny noted that a question has arisen as to whether there must be some specific representation as to the construction of the claim [65], or whether the emphasis “is not so much on the rebuttal of a particular representation, but rather on the interpretive process itself,” as suggested in Bauer Hockey 2020 FC 624 (discussed here): [66]. Justice de Montigny left this question for another day, as it was not necessary to decide the case at hand [67]. My impression is that he was leaning to the latter interpretation, but I can’t point to anything specific in the reasons to support that view, and in any event he expressly left the matter open.

Foreign prosecution history

The main issue raised on the facts was the use of foreign prosecution history. Justice de Montigny emphasized the need to adhere to the text of the provision [70], which suggests a very limited role, if any, for foreign prosecution history. Further, he noted that “There are also public policy reasons for treading carefully in allowing extrinsic evidence” [70]. Allowing foreign patent prosecution history into the analysis “might lead to overly contentious and expensive litigation,” given different law, differences in language potentially requiring translation, and differences in the registration system. He noted that “in this case, the patent application in the United States was eventually dropped, while registration in Canada was completed” [71]. Thus both textual and purposive considerations support a restrictive approach to the use of foreign prosecution history. At the same time he acknowledged the force of the argument for “keeping those who have previously disclaimed elements from their patent from re-claiming them in future infringement cases,” both in general [72] and on the facts of this case [73].

In this case, Manson J held that foreign prosecution history would be admissible in “extraordinary circumstances” [FC 73, 74, 77]. As noted above, Manson J described such circumstances as being where “that prosecution of the foreign application is made part of the prosecution history of the Canadian patent” [FC 77, original emphasis]; Canmar referred to this as an “incorporation by reference” theory [69]. On the facts of the case, “the patentee specifically referred to the corresponding US Application prosecution history and acknowledged that the amendments to the claims in the ‘376 file history were made to overcome novelty and obviousness concerns as raised in the US Application prosecution history” [FC 70], [68].

Justice de Montigny did not dismiss the possibility that some form of incorporation by reference doctrine might permit foreign prosecution history to be used under s 53.1, but he left this question “for another day,” as he was of the view that the facts at hand “do not lend themselves to a proper finding of incorporation by reference” [74]. Justice de Montigny noted that “There is nothing in the prosecution file of the ‘376 Patent that identifies with any detailed particularity what specific ‘written communication’ from the US prosecution history is incorporated and where that written communication can be found” [75]. The only mention is a response to the Examiner stating that the new claims correspond to those submitted during prosecution “of a related United States application” [75], and he remarked that “It is a stretch to say that the prosecution of the US Application is incorporated by reference when the specific application is not even cited within the document” [76]. There is some suggestion in the decision that US doctrine of incorporation by reference might provide some guidance [76], though it remains to be seen how influential this will be, as the matter will ultimately turn on the interpretation of s 53.1.

The FCA decision in Canmar therefore establishes that the foreign prosecution history cannot be used unless the Canadian prosecution history contains some fairly specific reference to the foreign prosecution history. It should be emphasized, however, that the converse is not necessarily true: this does not mean that the foreign prosecution history will be considered if there is a sufficiently specific reference in the Canadian file. Whether foreign prosecution history will be admissible at all, and if so, exactly what kind of reference is needed, is a question left for another day [74[.

In my view the FCA was entirely right to confine itself to the holding necessary to dispose of the appeal, while leaving other questions for another day. Section 53.1 is novel—it does not appear to reflect any particular foreign jurisprudence—and it is detailed. Accordingly it will be best for the law to develop incrementally, at least for a period as we get a better understanding of the issues that will be raised by s 53.1 in different factual contexts.

Ten Year Anniversary

The first post on this blog went up ten years ago today. Blogger tells me I’ve had just over one million page views since then, and lately it’s running around 30k views per month. What’s far more satisfying than the numbers is the engagement. Thank you for making this project worthwhile!

Landmark FCA Decision on Summary Judgment

Canmar Foods Ltd v TA Foods Ltd 2021 FCA 7 de Montigny JA: Pelletier JA, Rivoalen JJA affg 2019 FC 1233 Manson J

            2,582,376 / method for roasting oil seed

This landmark decision from the FCA, affirming Manson J’s use of summary judgment, is a very welcome development that will improve the operation of the patent system by permitting expeditious disposition of patent cases in appropriate circumstances. The decision is also notable for its considered discussion of the interpretation of s 53.1, which permits the use of summary judgment in certain circumstances. I’ll discuss that issue in my next post.

At trial, Manson J granted summary judgment on the basis of non-infringement in a decision which turned on claim construction, and particularly whether two claim elements were essential. This was the first decision in many years to grant summary judgment on substantive grounds in a patent matter. As discussed here, Manson J was critical of the FCA decision in MacNeil Estate 2004 FCA 50, which, in his view, resulted in an approach in which “summary judgment as a just, efficient and expeditious means to resolve disputes on a proportionate basis was lost” [45]. He described the SCC decision in Hryniak v Mauldin 2014 SCC 7 as resulting in “a culture shift” that “opened the door for a more reasoned approach to the use of summary judgment motions” [46]. After Manson J’s decision, three more FC decisions granted summary judgement: see here, here and here.

In his decision for the FCA, affirming Manson J, de Montigny JA did not comment on whether there has been a “culture shift”—indeed, he did not mention MacNeil Estate at all, which may be significant in itself. Nonetheless, the fact that the FCA approved the use of summary judgment on the facts of this case sends an important message to the FC judges, who will evidently be receptive to it judging from the spate of recent decisions.

On the facts, two issues were raised in respect of summary judgment. The first is whether Manson J erred in granting summary judgment on the basis of non-infringement before any discovery had taken place [19]. Justice de Montigny noted that this was clearly permitted by Rule 213: “Whether or not discovery had taken place at this stage is not a factor contemplated by Rule 213, and ought not to be regarded as such” [26, my emphasis]. Moreover, notwithstanding the absence of discovery, the patentee in this case had sufficient information as to the defendant’s method to allow it to marshal the relevant evidence [28].

The second question was whether Manson J erred in making a determination that the two contentious elements were not essential in the absence of any expert evidence on the issue. In effect, the patentee argued that expert evidence is required for claim construction, and that it is necessarily an error of law for the Court to construe a patent in its absence [31], [33]. Justice de Montigny rejected this proposition, both because of a lack of authority to that effect [34], and as a matter of principle: “If the construction of a claim is a matter of law and the judge is entitled to adopt a construction different from that put forward by the parties and their experts, surely the judge can also construe a claim without relying on such evidence in appropriate circumstances” [36].

While de Montigny JA also held that this was an appropriate case to do so, there was an important caveat:

[37] Claims must always be construed in an informed and purposive way, and it is only in the clearest of cases that judges should feel confident enough to construe the claims of a patent as they would be understood by a skilled person, without the help of any expert evidence.

In this case, the invention is a simple mechanical system and the key terms—”stream of air” and “insulated”—and the claim construction exercise was quite straightforward [38]–[47]. This might be taken to suggest that claim construction in the absence of expert evidence is only appropriate in cases in which the claim terms are themselves easy to interpret.

But that suggestion was certainly not explicit and I don’t think it is what was intended. Justice de Montigny continued by saying, in the same paragraph:

[37] In the case at bar, the appellant had the obligation to put its best foot forward. However, it made the strategic decision not to present expert evidence on the summary judgment motion, thereby foregoing the possibility to impress upon the Judge the need to rely on such expertise to construe the patent. Be that as it may, what matters at the end of the day is not so much how and on what basis the judge came to his or her interpretation of a claim, but whether such interpretation is correct or flawed.

This indicates that it is the nature of the evidence, not the nature of the invention, that is important in determining whether a judge should feel confident enough to construe the claims. That seems right to me; even ordinary English words can take on a technical meaning when used in a patent, so it must be the evidence that is key.

I’m sure Canmar will not be the last word on summary judgment, focusing as it did on the facts of the case, but it is the first words in a promising new chapter.

Clarification of Federal Court Jurisdiction over Patent-related Contractual Matters

McCain Foods Limited v JR Simplot Company 2021 FCA 4 Locke JA: Near, Mactavish JJA varg 2019 FC 1635 McVeigh J

2,412,841 / Process for Treating Vegetables and Fruit Before Cooking

The Federal Court’s jurisdiction over patent-related contractual matters was recently expanded in Salt v Baker 2020 FCA 127 (here). This FCA decision on a motion to add a third-party claim clarifies the limits of that jurisdiction. As discussed here, in the action underlying this motion, McCain sued Simplot, one of its competitors, for infringement of McCain’s 841 patent which relates to a system that uses a pulsed electric field (“PEF”) to facilitate cutting of frozen fruits and vegetables [1]. Simplot implemented the system using a PEF system manufactured by Elea, a German company [FC 3]-[5]. Simplot’s contract with Elea specified that Elea would “defend, save harmless, and indemnify Simplot” for patent infringement cases arising from products supplied by Elea to Simplot [FC 74]. Accordingly, when Simplot was sued by McCain, it requested an indemnity from Elea [FC 8, 74]. Elea refused [FC 9].

Rather than bringing an action against Elea on the contract, in the motion under appeal Simplot sought to bring Elea into the infringement action with a third party claim. This was allowed by Aylen J at first instance and by McVeigh J on appeal to the FC. Locke JA’s decision for the FCA reversed on this point and refused to allow the third party claim. The bottom line is that for the Federal Court to have jurisdiction, it is not enough that the claim relates to a patent: the claim must have a basis in federal law, which as a practical matter means a basis in a federal Act.

A third party claim must stand on its own as an independent proceeding [26]. The question therefore is whether the Federal Court would have jurisdiction over Simplot’s claim against Elea if brought as an independent action. Because the Federal Court is a statutory court, its jurisdiction must be founded on some statutory grant of jurisdiction by Parliament [58].

Of course the Federal Court has jurisdiction over patent infringement [62]. Simplot therefore argued that Elea had induced Simplot to infringe the 841 patent. Locke JA disposed of this with a neat logical point:

[69] either Simplot does not infringe McCain’s patent (in which case it will have no liability to McCain that could ground a claim against Elea), or Simplot does infringe (in which case it cannot claim to be a person claiming under the patentee, and hence lacks standing to sue).

So, Elea may indeed be an infringer by inducement; if McCain had brought an action against Elea for inducing Simplot to infringe, this would be a classic inducement scenario and there would be no question that the Federal Court would have jurisdiction. But the question is not whether McCain can bring an infringement action against Elea, but whether Simplot can. Elea can only be liable for inducement (to anyone) if Simplot is an infringer [70]. But if Simplot is an infringer, it can’t be the plaintiff in a patent action; under s 55(1) only a “person claiming under the patentee” has standing to bring an infringement action, and an infringer is not a person claiming under the patentee [73]–[80]. This analysis provides a clear and logical legal foundation to support the intuition that a direct infringer can’t sue the party who induced it to infringe (at least not for infringement). This means this argument was disposed of not as a matter of Federal Court jurisdiction, but as a matter of standing. This seems right to me: the question isn’t whether Elea is inducing infringement, but whether Simplot can bring the action.

Simplot’s next argument was that the Federal Court has jurisdiction pursuant to s 20(2) of the Federal Courts Act:

20(2) The Federal Court has concurrent jurisdiction in all cases, other than those mentioned in subsection (1), in which a remedy is sought under the authority of an Act of Parliament or at law or in equity respecting any patent of invention, certificate of supplementary protection issued under the Patent Act, copyright, trademark, industrial design or topography referred to in paragraph (1)(a).

Under this provision the party has to be (1) seeking a remedy (2) under an Act of Parliament or at law or in equity (3) respecting any patent of invention. It’s not entirely clear from this portion of the decision, but I take it that the remedy sought by Simplot was indemnification under its contract with Elea [90], [91], and that Simplot argued that this was a remedy “respecting any patent of invention.” These points were not elaborated as the matter was disposed of on the basis of element (2), requiring an “Act of Parliament or at law or in equity.”

Locke JA rejected this argument on the basis that “Act of Parliament” in this provision cannot refer to the Federal Courts Act itself, as this would be impermissibly self-referential [86]. In so holding, he followed Windsor Bridge 2016 SCC 54 [52] on the interpretation of s 23, which deals with different subject matter (eg bills of exchange), but is otherwise similarly worded [85]. Section 20(2) gives jurisdiction when the party is seeking a remedy under an Act of Parliament, but the Federal Courts Act does not confer rights for which a remedy might be sought, so it cannot be the requisite “Act of Parliament.” That is, the Federal Courts Act “it is not itself a federal law under which the Company can seek relief, however ‘relief’ is defined”: Windsor Bridge [52].

This provides a good basis for distinguishing Salt v Baker, which expanded the previous understanding of the Federal Court’s jurisdiction. Locke JA noted that “Salt concerned an application under section 52 of the Patent Act to vary an entry in the records of the Patent Office. Section 52 of the Patent Act therefore constituted a specific statutory grant of jurisdiction to the Federal Court” [63]. That strikes me as an entirely sound distinction. It is not merely technical. Section 52 gives the Federal Court jurisdiction to vary records relating to title. Determining who actually has title will normally require interpretation of the relevant assignments, but also application of the Patent Act priority rules which are set out in what is now s 49 of the Act. Thus, varying the title records is almost always closely entwined with application of substantive provisions of the Act.

Turning back to the jurisdiction conferred by s 20(2), what about focusing on a remedy “at law”? If Simplot is seeking to enforce its contractual right to indemnity, that looks like a remedy “at law.” Locke JA rejected this argument, saying “Because both the jurisdiction of the Federal Court and patent law in general must be based on statute, Simplot must nevertheless base its argument on a federal statute” [88], and moreover, a federal statute other than the Federal Courts Act [88]. This remark doesn’t directly address the “at law” phrase, but Windsor Bridge 2016 SCC 54 and especially Quebec North Shore Paper [1977] 2 SCR 1054, 1058, 1065-66, make it clear that “at law” or “or otherwise” must refer to federal law, which is to say either a statute or federal common law (if the latter exists).

So, Locke JA’s decision is solidly based in the SCC jurisprudence he relies on. My difficulty with all this is that the SCC in Kellogg [1941] SCR 242 held that the Exchequer Court had jurisdiction to determine ownership of a patent based on the “in equity” phrase of s 22(c) of the Exchequer Court Act, which is the same as s 20(2) in relevant respects. The appellant in that case sought a declaration that it was the owner of a patent by virtue of the fact that the inventor was an employee and consequently a trustee of the invention for the company [p246]. The SCC noted that the provision gave the Exchequer Court jurisdiction “in all cases where a ‘remedy is sought’ ‘respecting any patent of invention’ ‘under the authority of any Act of the Parliament of Canada or at Common Law or in Equity.’ The remedy sought by the appellant . . . is evidently a remedy in Equity respecting a patent of invention” [p250]. Now, I can readily see how Kellogg might be distinguished on the basis that a claim respecting ownership of a patent is more closely connected to patent law than a claim under the indemnity provision at issue in this case. But that distinction would presumably turn on the phrase ““respecting any patent of invention,” and the question at hand is the interpretation of the “at law or in equity / or otherwise” phrase in s 20 and 23. On that point I’m having a hard time reconciling Kellogg with Quebec North Shore Paper and Windsor Bridge. Though I don’t want to express a firm opinion, so far as I can tell, Kellogg is probably no longer good law on this point. And after all, that provision was only a alternative basis for jurisdiction, which was founded primarily on the statutory conflict provisions of the Patent Act. This aspect of Kellogg does seem to be universally ignored. (I'm not sure I'm persuaded by the SCC interpretation of s 23 in Quebec North Shore Paper and Windsor Bridge, but that's beside the point.)

Finally, “Simplot argues that it would be put in an absurd position of having to prove infringement in a separate provincial court proceeding in order claim its indemnity” [90]. Locke JA was expressly sympathetic to Simplot’s situation “of having to commence separate legal proceedings in a different court to enforce its claims in warranty and indemnity against Elea” [91]. But as he noted, “concerns about a multiplicity of proceedings and/or inconsistent results cannot justify finding that the Federal Court has jurisdiction where no federal statute grants it jurisdiction” [90] I have to quibble a bit; while it is of course true that such concerns alone cannot be the basis for Federal Court jurisdiction, they may be relevant to a purposive interpretation of the relevant provisions. For example, the FCA in Salt v Baker [40]–[44] relied on such considerations in interpreting s 52. With that said, I have to agree that such purposive considerations cannot overcome the authorities relied on by Locke JA. So far as I can see, while there may be a real problem here, it will need some statutory fix to extend the Federal Court’s jurisdiction in order to solve it, assuming that is constitutionality possible.

I must say that it’s not clear to me that there will necessarily be duplication on the particular facts of this case, as we have been given very little of the contractual terms. To the extent that the contract requires Elea to indemnify Simplit for infringement, then once infringement is established in this action, it is not clear to me why non-infringement would need to be re-litigated in an action on the indemnity claim. To the extent that the contract requires Elea to defend the action, then I don’t see how the indemnity claim on that point would turn on whether Simplot actually infringed.

To summarize, while Simplot had a convoluted infringement argument that failed, at the end of the day, Simplot was essentially trying to bring an action against Elea in Federal Court to enforce the contractual indemnification provision. It is perfectly clear that the Federal Court does not have jurisdiction in a case “purely and simply concerning a contract between subject and subject”: Kellogg 249. The issue is whether the fact that the contractual provision at issue is in respect of an indemnity related to patent infringement is sufficient to bring it within the jurisdiction of the Federal Court. The answer is no. It is not enough that the claim relate to “any patent of invention,” the claim must have a basis in federal law.

A final point on this topic is that in Libbey-Owens-Ford Glass Co v Ford Motor Co (1968), 55 CPR165 (Ex Ct), the Exchequer Court held that there was a “good arguable case” that a claim against a third party for indemnity based on contract by a defendant in the context of a patent infringement action fell within the jurisdiction of the Exchequer Court [92]. Locke JA held it cannot be reconciled with the SCC jurisprudence he relied on and it can no longer be considered good law.

To wrap this all up, Simplot also sought to amend its statement of defence to allege a defence based on an implied licence or acquiescence, flowing ultimately from the fact (as alleged) that Elea sold its systems worldwide with McCain’s knowledge. This was permitted by Aylen J and McVeigh J. As discussed here, I found that aspect of McVeigh J’s decision a bit difficult to follow, but Locke JA has clarified that it really just turns on the facts, and while “Simplot’s implicit license defence is light on details, and it may prove difficult for it to adduce evidence sufficient to prove the existence of a license that would benefit it” [32], this is not enough to refuse it to go forward. (And see similarly [53] regarding the acquiescence defence.)

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].

Nova v Dow Flipped the Approach to Accounting of Profits

Rovi Guides, Inc v Videotron Ltd 2021 FC 19 Lafrenière J

In this decision Lafrenière J granted Videotron’s motion to reopen the evidentiary portion of the patent infringement trial to allow additional expert evidence relating to the accounting of profits remedy, given that the FCA in Nova v Dow 2020 FCA 141 “effectively flipped the default approach to accounting of profits” [14]. (Oral argument is scheduled commencing January 20, 2021 [13].) As Lafrenière J noted [26], this is consistent with a similar motion granted in the Bauer Hockey 2020 FC 1123 trademark case, where Grammond J remarked that “I am satisfied that, given the understanding of the law prevailing among intellectual property lawyers prior to Nova v Dow, CCM could not have been expected to file evidence regarding total costs earlier” [7].

In this case, Rovi, opposing the motion, argued that

[21] the Federal Court of Appeal’s decision in Nova Chemicals is either consistent with Schmeiser [2004 SCC 34], which would mean that there has been no fundamental change in the law, or that Nova Chemicals is inconsistent with Schmeiser, and therefore wrongly decided and should not be followed.

Lafrenière J noted that Nova Chemicals is “viewed by some as departing from established law both in statements of principle and in the result,” quoting from my post Nova v Dow: A Radical Departure from Established Law, which argued that Nova is inconsistent with Schmeiser.

Lafrenière J remarked that whatever the merits of those views, he was bound to follow the FCA decision in Nova, given that it “unambiguously provides that the full costs approach should be the default one” [23]. That is surely right. While I remain of the view that the principles expressed in Schmeiser are inconsistent with the decision in Nova, the FCA is evidently of a different opinion, and the SCC did not address the issue of fixed costs specifically. There would be an interesting issue of precedent if the SCC had specifically addressed fixed costs and the FCA had declined to follow it on that point, but on how to understand the implications of the principles set out in Schmeiser, the FC must of course follow the FCA.

Tensar v Enviro-Pro Affirmed on the Facts

Tensar Technologies, Limited v Enviro-Pro Geosynthetics, Ltd 2021 FCA 3 Locke JA: Gauthier, de Montigny JJA aff’g 2019 FC 277 Manson J

2,491,858

At trial, Manson J held that Tensar’s 858 patent was valid but not infringed in a decision that turned primarily on the evidence, as discussed here. The FCA has now affirmed on the basis that all of Manson J’s conclusions were open to him on the record, and Tensar had therefore not established that Manson J had made any reviewable error.

Locke JA concluded by noting that

[48] In the end, it may be that the respondent’s Tri-Grid products use the general idea described in the 858 Patent. This may cause great frustration to the appellants, but it is not enough to establish patent infringement. Infringement focuses on the claims of the patent (not the description therein of the invention), and requires that the allegedly infringing product incorporate all of the essential elements of at least one of the claims. If the claims have a different focus than the disclosure, which is not uncommon, then there is room for the described invention to be used without infringing any of the claims of the patent. In addition, the burden to establish infringement is on the patentee (the appellants here), and the evidence in this case did not satisfy the Trial Judge. That conclusion was open to him on this record.

The Limits of Anticipation by Speculation

Teva Canada Innovation v Pharmascience Inc 2020 FC 1158 Kane J

2,702,437 / 2,760,802 / glatiramer acetate / Copaxone / Glatect / NOC

The patents at issue in this case relate to the use of glatiramer acetate for the treatment of multiple sclerosis [MS]. Glatiramer acetate was known to be effective in treating MS, with a standard dose of 20mg daily, administered by subcutaneous injection. The 437 patent claims the early use of glatiramer acetate for the purpose of delaying the development of MS, while the 802 patent claims a dosing regime of 40mg three times a week. The decision turned largely on the facts (to the tune of 936 paragraphs!), but a few points of interest were raised, particularly in the discussion of the 437 patent, which Kane J held to be novel, but invalid for obviousness. Kane J's holding that the 802 patent was not invalid for obviousness or lack of utility, and that it would be infringed by Pharmascience's 40mg product, turned on the facts.

MS is a disease of the central nervous system in which the patient’s own immune system attacks the myelin of the CNS. The early course of the disease is characterized by intermittent attacks, which include episodes of numbness, muscle weakness and spasticity, etc [73]. The disease also gives rise to lesions on the CNS resulting from destruction of the myelin sheath [71]. Patients who present with a single clinical attack having features typical of MS were considered to have a “clinically isolated syndrome” (CIS) [77]. CIS patients are at elevated risk for MS itself (also referred to as “clinically definite MS” or CDMS), but not all CIS is caused by MS, so CIS patients are not considered to actually have MS.

The inventive concept of the 437 patent is that glatiramer acetate could be used to delay the onset of clinically definite MS in CIS patents at risk of developing MS [569]. Accordingly, Claim 1 was to glatiramer acetate “for use in delaying the onset of clinically definite multiple sclerosis,” in a patient who presents with at least one lesion “suggestive of multiple sclerosis” and who is “at risk of developing” clinically definite MS, and “prior to development of clinically definite multiple sclerosis” [251].

The claim construction debate turned on the fact that around the time of filing, in 2007, the criteria for diagnosing MS were changing. The older Poser criteria relied on clinical signs, ie two clinical attacks. The new McDonald criteria were developed to take advantage of the use of MRI for diagnosis using direct detection of lesions characteristic of MS with less reliance on clinical signs. The result is that some patients who would not have been diagnosed as actually having MS using the Poser criteria, would be diagnosed as having it using the McDonald criteria. The 437 patent was apparently developed with the Poser criteria in mind and there was a long and involved argument as to whether the patients with MS according to the McDonald criteria would be included in the claim; if so, the claim would not be novel, as the use of glatiramer acetate for treating MS was known [305].

I must admit that I was having trouble following the argument and its significance, and I was relieved when Kane J stated that “In my view, far too much time and effort was devoted to debating the meaning of the terms, the evolving diagnostic criteria and the scope of the patients covered by the claims” [306] . All the experts construed the claims in accordance with their plain language, which did not specifically refer to either set of criteria by name, and “the claims mean what they say” [307]. Kane J held that on a plain reading of the claims they were directed to a patient who did not already have a confirmed diagnosis of MS, and so did not include patients who already met the McDonald criteria [329]. This reading seems eminently sensible to me. The result of this claim construction holding was to negate both anticipation based on the Pinchasi 2007 prior art [454] as well as the Gillette defence [396], [403].

In light of this conclusion on claim construction, the more interesting anticipation argument was based on the Karussis 2006 prior art. This was a report of an International Working Group for Treatment Optimization in MS [458]. The report suggested the use of therapies approved for treatment of MS itself, which prominently includes glatiramer acetate, as a possible treatment for patients at risk of developing MS (ie, those with CIS), but acknowledged that this was an inference and was not based on evidence [464], [465]. So, the Working Group concluded that “as [other therapies] and glatiramer acetate have been shown to be effective in [the most common form of diagnosed MS], and a patient with a CIS is highly likely to progress to MS, it is reasonable to propose early treatment with [therapies approved for treatment of MS] at the doses approved for MS, even though there is not yet evidence to support this approach” [465]. The question is whether this and similar statements anticipated the claims of the 437 patent.

This raises the issue of what I have called “anticipation by speculation,” as raised in Hospira 2020 FCA 30 and Biogen v Taro 2020 FC 621, which held that similarly speculative statements were anticipatory. As discussed here and here, my view is that Hospira and Biogen v Taro stretched anticipation to the very limit, and perhaps beyond. In this case, Kane J held that Karussis 2006 was not anticipatory. This is not to say that Kane J’s holding is necessarily inconsistent with Hospira and Biogen v Taro, which Kane J considered and distinguished on the facts: [479]–[482]. I won’t go through all the particular grounds, though I will say that the distinctions she draw struck me as generally reasonable; as always in borderline cases, some of the distinctions to be drawn will necessarily be fine and turn on the particular facts. The broader point is that Kane J’s holding illustrates that anticipation by speculation can only be pushed so far and the line is somewhere in this general vicinity.

One general observation made by Kane J is worth pointing out (my emphasis):

[484] In my view, if Karussis 2006 is anticipatory art, what would prevent a group of experts in any field from having discussions and speculating on the positive outcome of a clinical trial in progress, developed and implemented by others, in order to later argue that their speculation that the trial in progress would be successful anticipated the claims of a patent that are based on such clinical trials. This approach seems to be inconsistent with the objectives of the experimental use exemption established in the common law.

[485] As noted by Teva, the law recognizes that there is no disclosure for the purposes of anticipation where a prior use is experimental. If a patent cannot be anticipated by its own clinical study, in my view, it cannot be anticipated by the speculative recommendations of a group of experts who discuss the state of MS treatment, note the lack of evidence, and note that trials are underway.

Kane J then went on to hold that Karussis 2006 did not in any event satisfy the enablement branch of anticipation: [486]–[493].

All that was for nought, so far as Allergan is concerned, as Kane J then went on to find the 437 patent invalid for obviousness on an obvious-to-try analysis. This is likely to be common in cases where a clear suggestion to try the invention can be found in the prior art, especially since the holding in Hospira 2020 FCA 30 that the state of the art for purposes of an obviousness attack includes all prior art. Exactly where we draw the line on anticipation is now less important, because if an obscure piece of prior art does not quite reach the level of anticipation, it can still be used as the basis for an obviousness attack. With that said, it does not appear that the Hospira rule made any difference in this case; while Karussis 2006 was not common general knowledge, it reflected several studies that were [573].

I’ll finish with two quick points on obviousness.

First, Kane J clearly stated that the mere fact that it is not self-evident that the invention will work, before any experiments are undertaken, does not mean that it is not obvious.

[593] Dr. Selchen’s opinion that the invention was not self-evident appears to conflate the obvious to try test with the factor or consideration that informs the test, i.e., whether it was self-evident that it would work, in the sense of being effective for CIS patients to delay the onset of MS and achieve the other claimed benefits.

I entirely agree with Kane J on this, but it is a point that continues to pop up and cause confusion, as it did for Dr Selchen in this case.

Second, Kane J recognized and applied the “golden bonus” rule, usually traced back to Hallen v Brabantia [1991] RPC 195, 216 (CA), which says that if an invention is obvious for one reason, it does not become non-obvious merely because it has some non-obvious benefits (the "golden bonus"):

[594] Teva’s submission – that the therapeutic benefits or outcomes of the ‘437 Patent were not obvious or expected – does not diminish that the invention was obvious to try. In Janssen [2015 FC 184] at para 100 [see here], the Court explained, that “if a patentee obtains a workable formulation, the later discovery of one of its inherent characteristics does not add anything inventive to what had already been discovered: see Alcon Canada Inc. v Apotex Inc., 2012 FC 410 at para 45, [2012] FCJ No 1707 (QL).” In the present circumstances, once the POSITA pursued the use of glatiramer acetate for CIS patients, the claimed benefits would follow.

As Kane J notes, this is not an entirely new development in Canadian law, but it is nonetheless welcome to see the point expressly recognized and affirmed. I believe that this is the first time that the phrase “golden bonus” has actually been used in a Canadian case [510].

Prosecution History May Not Be Used to Rebut Representations Made by a Licensee

Allergan Inc v Sandoz Canada Inc 2020 FC 1189 Crampton CJ

2,507,002 / silodosin / RAPAFLO / NOC

Section 53.1 of the Patent Act permits the use of prosecution history in claim construction in certain circumstances. In this decision, Crampton CJ held that s 53.1 only permits use of prosecution history to rebut a representation made in litigation by a patentee, not representations made by a licensee. This is a truck-sized loophole in s 53.1, and Crampton CJ noted the prosecution history in question “provides a glaring example of the mischief that is implicitly permitted by the current wording” [135]. Nonetheless, Crampton CJ’s analysis of the provision is entirely compelling.

As noted in Monday’s post, the key issue on infringement was whether “wet granulation” was an essential element of the claims in question. Allergan argued it was not. Sandoz wished to rely on representations made during patent prosecution to rebut this assertion. The difficulty is that s 53.1 permits prosecution history to be introduced to rebut any representation “made by the patentee in the action.” Allergan is not the patentee; it is the exclusive Canadian licensee [1] and is the “first person” under the PM(NOC) Regulations [5]. The patent owner is Kissei Pharmaceutical and the patent was prosecuted on Kissei’s behalf [1], [3]. Kissei was also a defendant to this NOC action, joined pursuant to s 6(2), but crucially, Kissei took no position on the infringement issue, though it adopted Allergan's submissions relating to validity. (Kissei did not appear during the trial [7].)

So, Allergan’s argument was that the prosecution history could not be introduced because it can only be used to rebut a representation by the “patentee” and the patentee, Kissei had not made any representations. The question facing Crampton CJ was therefore whether s 53.1 should be interpreted to allow prosecution history to be used to rebut representations made by a licensee.

As noted, I find Crampton CJ’s reasons on this point to entirely convincing. I don’t have anything to add, so I will simply reproduce the key passage (with some references omitted):

[124] In my view, none of the arguments advanced by Sandoz can overcome the plain wording of subsection 53.1(1), a contextual reading of the Act or the jurisprudence in respect of the definition of the word “patentee” in section 2 of that legislation.

[125] It is trite law that “the words of a statute must be read ‘in their entire context and in their grammatical and ordinary sense harmoniously with the scheme of the Act, the object of the Act, and the intention of Parliament’”.

[126] The “chapeau” in subsection 53.1(1) plainly limits the scope of that provision to permitting certain written communications to be admitted into evidence to rebut any representation made by the patentee in an action or proceeding, in respect of the construction of a claim in a patent that is at issue in the action or proceeding. In the present proceeding, it is admitted that the patentee is the defendant Kissei, which has not made any representation to the Court with respect to the construction of the ‘002 Patent. Accordingly, in the absence of any clear indication elsewhere within the scheme or object of the Act that Parliament intended to word “patentee” to include a licensee of a patent, subsection 53.1(1) cannot be invoked in this proceeding.

[127] I do not agree with Sandoz’s position that a licensee falls within the meaning of the word “patentee”, as defined in section 2 of the Act, namely, “the person for the time being entitled to the benefit of a patent”. This position was specifically considered and rejected in Electric Chain Co of Canadas Ltd v Art Metal Works Inc, [1933] SCR 581 at 586-587 [Electric Chain]. The effect of that decision was that a licensee had no right to be a party to an infringement action in Canada. As a result, what is subsection 55(1) was added to the Act: American Cyanamid Co v Novopharm, [1972] FC 739 at paras 23-24 (FCA) [American Cyanamid]. That provision, which has undergone some minor amendments that are not germane for the present purposes, states:

“A person who infringes a patent is liable to the patentee and to all persons claiming under the patentee for all damage sustained by the patentee or by any such person, after the grant of the patent, by reason of the infringement.”

[128] It has since been confirmed that a person who is a licensee under a patent is a “person claiming under” the patentee within the meaning of subsection 55(1): Armstrong Cork Canada v Domco Industries Ltd, [1982] 1 SCR 907 at 914; American Cyanamid, above, at paras 31-32.

[129] What is instructive for the present purposes is that while the Act was amended to permit a licensee to sue for infringement, the definition of “patentee” was not amended following the interpretation that it was given in Electric Chain, above.

[130] Moreover, given that Parliament included the words “the applicant for a patent” in clause 53.1(1)(b)(i), but not in the “chapeau” of subsection 53.1(1), it can be inferred that (i) Parliament was aware of the distinction between a patentee and a person who is not the patentee, and (ii) Parliament decided to strictly limit the scope of the “chapeau” to a person who is a patentee.

[131] This interpretation of Parliament’s intent finds some support in the legislative history. In particular, in the submission that IPIC made to the [Senate Banking, Trade and Commerce Committee], discussed at paragraph 123 above, IPIC noted that the language of section 53.1 created a “loophole” that would permit “a patentee to circumvent the operation of this section by acting through a licensee”: IPIC Submission, above, at 12. To address this “loophole”, IPIC recommended that subsection 53.1(1) be amended to include “persons claiming under the patentee”. The specific amendment proposed by IPIC is the underlined wording in the chapeau of that provision:

53.1 (1) In any action or proceeding respecting a patent, a written communication or any part of such a communication, may be admitted into evidence to rebut any representation made by the patentee or a person claiming under the patentee in the action or proceeding as to the construction of a claim in the patent if […]

[132] Ultimately, IPIC’s recommendation was not accepted, and Bill C-86 was passed without any change to the language of subsection 53.1(1). This legislative history provides additional support for the view that, at the time Parliament added subsection 53.1(1) to the Act, it was aware of the distinction between a patentee and a person who is not a patentee, yet it chose to limit the scope of section 53.1 to representations made by patentees.

While Crampton CJ did not rely primarily on the last point, I must say that I find it particularly compelling; this very issue was before the responsible Parliamentary committee, along with a very simple fix, and it was not implemented. The inference that it was considered and rejected is strong, which implies that the “loophole” was specifically intended.

[133] In summary, the plain and ordinary meaning of the language in subsection 53.1(1), together with a contextual reading of the Act and the jurisprudence discussed above, support the view that the word “patentee” in subsection 53.1(1) does not include a licensee. The legislative history also provides some additional support for this interpretation. Sandoz has not identified any contextual consideration to support the alternative interpretation that it has advanced.

While the considerations reviewed by Crampton CJ are textual and contextual, rather than purposive, and there is evidently some kind of purposive argument to be made that this loophole frustrates the operation of the provision, Crampton CJ’s textual and contextual analysis is so entirely compelling that I cannot see that any other conclusion would be possible without an impermissible usurpation of the legislative role.

Sandoz also argued that the prosecution history was admissible entirely apart from s 53.1, relying on the FC decision in Distrimedic 2013 FC 1043 at [210]: [116]. Crampton CJ rejected this argument, in part on the basis that Distrimedic is distinguishable [117]. More importantly Crampton CJ also observed that in light of the bar on the use of prosecution history set out in of Free World 2000 SCC 66 [66], Distrimedic has been “overcome” by the enactment of section 53.1 [115], [118]. I understand Crampton CJ to be saying that the use of prosecution history is now confined to the circumstances set out in s 53.1. This strikes me as undoubtedly correct. Any use of prosecution history in Canada now turns entirely on the interpretation of s 53.1 and any subsequent legislative amendments.

Crampton CJ therefore concluded that “given that no representation has been made by the patentee (Kissei) of the ‘002 Patent in the present action, the file prosecution history is not admissible in evidence in this action. It is barred by the prohibition against extrinsic evidence: Free World, above, at para 66" [133].

Crampton CJ’s holding was driven inexorably by the legislation; he did not necessarily welcome it. On the contrary, he observed “in passing,” that the “file prosecution history in question provides a glaring example of the mischief that is implicitly permitted by the current wording of subsection 53.1(1)” [135]. For those interested in looking at the prosecution history themselves, the documents 02507002 - 2011-08-25 - PRO and 02507002 - 2011-01-19 - PRO emphasize wet granulation to overcome various objections.

Reconciling the Two Essentiality Tests in Free World

Allergan Inc v Sandoz Canada Inc 2020 FC 1189 Crampton CJ

2,507,002 / silodosin / RAPAFLO / NOC

In this NOC decision Crampton CJ held that Allergan’s 002 patent was not invalid for obviousness, but it was not infringed by Sandoz’s silodosin formulation. These holdings turned largely on the facts, though Crampton CJ’s discussion raised several interesting legal issues, the most important relating to the interpretation of s 53.1, which permits the use of prosecution history in claim construction in certain circumstances. I’ll discuss this in a subsequent post. This post will provide the background and discuss claim construction, which was determinative of infringement. In particular, I’ll consider the question of how to reconcile the two apparently different tests for essentiality set out by the SCC in paragraph 55 of Free World 2000 SCC 66. Really this is a discussion of a point that was raised more directly by Locke J in Shire v Apotex 2016 FC 382, though I didn’t discuss it when I blogged about that case.

The 002 patent relates to a solid dosage form of silodosin. Silodosin was known to be useful in treating dysuria associated with a kind of benign prostate tumour. However, it was difficult to formulate in a solid dosage form, in part because of its potent adhesive properties. The 002 patent reflects the inventors’ solution to this problem. There were two independent claims at issue. Claim 1 was to a capsule comprising a granule “prepared by wet granulation” of silodosin and various excipients, with a specified dissolution profile [63]. Claim 6 was a method claim comprising granulating silodosin, and other excipients by a “wet granualtion process” [63]. The meaning of the terms was not disputed [64], nor was it disputed that the Sandoz product included all the specified ingredients [21]. The only dispute was as to whether the wet granulation elements were essential [64].

Crampton CJ’s starting point is that all elements are presumed to be essential: [46]. He then adopted at [67] the two-pronged approach to essentiality set out by the SCC in Free World 2000 SCC 66 at the beginning of para 55:

For an element to be considered non-essential and thus substitutable, it must be shown either

(i) that on a purposive construction of the words of the claim it was clearly not intended to be essential, or

(ii) that at the date of publication of the patent, the skilled addressees would have appreciated that a particular element could be substituted without affecting the working of the invention

The difficulty is that in the same paragraph, only two sentences later, the SCC quoted with approval the three part-test set out by Hoffmann J in Improver [1989] RPC 69, which was intended to summarize Catnic:

(i) Does the variant have a material effect upon the way the invention works? If yes, the variant is outside the claim. If no: –

(ii) Would this (i.e.: that the variant had no material effect) have been obvious at the date of publication of the patent to a reader skilled in the art? If no, the variant is outside the claim. If yes: –

(iii) Would the reader skilled in the art nevertheless have understood from the language of the claim that the patentee intended that strict compliance with the primary meaning was an essential requirement of the invention? If yes, the variant is outside the claim.

As Locke J pointed out in Shire v Apotex 2016 FC 382 [135], these two tests are not entirely consistent. The first prong of the two-part test corresponds roughly to the third Improver question, and the second prong corresponds roughly to the first and second Improver questions, but the two-part test is disjunctive while the Improver questions are conjunctive [135]. The problem, then, is to try to understand what the SCC intended. Locke J suggested that the SCC probably intended the two-part test to be conjunctive, so that it would be necessary to establish both elements in order to establish that the element at issue is not essential [137].

Lear v Adkins Should Not Be Followed

Loops LLC v Maxill Inc 2020 ONSC 5438 Lederer J: Swinton, Penny, JJ revg 2020 ONSC 971 Templeton J

            2,577,109 / Toothbrush

In Monday’s post I reviewed Lederer J’s decision in this case, holding that a no-challenge clause in a settlement agreement is enforceable, notwithstanding the decision of the US Supreme Court in Lear Inc v Adkins 395 US 653 (1969) which abolished the doctrine of licensee estoppel in US law. In Tuesday’s post I explained that the doctrine of licensee estoppel is not based on estoppel at all, but is based entirely on contract law. Absent an express agreement to the contrary, the licensee’s obligation to pay royalties will not normally be contingent on the validity of the patent. This is partly because the courts will not normally imply a warranty of validity into the agreement and partly invalidity will not normally constitute complete failure of consideration, but might instead give rise to a counter-claim for damages. In this post I will step back and consider the bigger question: as a matter of policy, should Lear v Adkins be followed in Canada? Recall that while Lederer J was of the view that holding a no-challenge clause in a settlement agreement to be unenforceable was a leap too far [64], but he left open the possibility that it might be adopted in more incremental steps. That is the question I will focus on.

There Is No Such Thing as Licensee Estoppel

Loops LLC v Maxill Inc 2020 ONSC 5438 Lederer J: Swinton, Penny, JJ revg 2020 ONSC 971 Templeton J

            2,577,109 / Toothbrush

In Lear Inc v Adkins 395 US 653 (1969) the US Supreme Court abolished the doctrine of “licensee estoppel,” which (more or less) prevents a licensee from challenging the validity of the licensed patent. Monday’s post summarized the decision of Lederer J for the Ontario Divisional Court holding that a “no-challenge” clause in a settlement agreement is enforceable, Lear v Adkins notwithstanding. In this post I will argue that there is really no such thing as licensee estoppel in Canadian law. More precisely, the rule at issue is simply based on contract law and it is not helpful to characterize it as based on an estoppel.

First, what exactly is the rule? The precise nature of the licensee estoppel rule was not at issue in this case, as the question was only whether the express no-challenge clause was enforceable. Lederer J at [61] did quote from Asturiana (1979) 55 CPR(2d) 131 (Ont HCJ) referring to the consistent Anglo-Canadian authority “that a licensee is not permitted to dispute the validity of the patents.” This is not exactly right because there are many circumstances in which a licensee is permitted to dispute validity. It is more accurate to say that a licensee cannot raise invalidity of the patent as a defence to an action for payment of royalties owing under a licence agreement, except in the case of fraud, unless the agreement contains an express covenant as to validity of the patent. (Asturiana was an action for payment of royalties, so the quoted statement is correct in that context.)

Now, in some respects the rule is very strict. In particular, the licensee cannot raise invalidity as a defence to an action for payment of royalties even when the patent has actually been held invalid in other proceedings: Mills v Carson (1893) 10 RPC 9 (CA); African Gold Recovery Co Ltd (1897) 14 RPC 660 cited with approval in Anderson v EJ Shepard Ltd (1930), 66 OLR 105 (CA) [20]; Duryea v Kaufman (1910), 21 OLR 165 (OHCJ) [23] (in obiter).

On the other hand, it is subject to a major exception. While a licensee cannot raise invalidity as a defence to an action for payment of royalties under the licence, a licensee, even a current licensee, may always raise invalidity as a defence to an action for infringement: this was in fact the holding in Duryea v Kaufman, the leading case that firmly established licensee estoppel in Canadian law with a thorough review of the English cases; see also Bayer AG v Apotex Inc (1995) 60 CPR(3d) 58 (OGD) 73 (“a licensee will not be estopped from alleging that the patent is invalid if the action brought against it is for infringement of the patent”) aff’d (1998) 82 CPR (3d) 526 (Ont CA) 532-33; Fuel Economy v Murray (1930) 47 RPC 346 (Ch) at 353 (“an estoppel cannot arise in an action for infringement”) aff’d 358 (CA). So, for example, a current licensee may argue invalidity if the patentee alleges the licensee has infringed by carrying out activities outside the scope of the licence, whether by falling outside of the permitted uses (Apotex v Tanabe (1994) 59 CPR(3d) 38 (OGD) 49; Bayer AG v Apotex Inc) or outside the geographic scope of the licence (Fuel Economy v Murray). A defendant claiming a right by licence may plead invalidity in the alternative: Duryea [35], [41]-[42]. A former licensee whose licence has come to an end by revocation, lapse of time, or any other method, may raise invalidity as a defence to an action for infringement in the same manner as if it had never had a licence: Duryea [35]; Crossley v Dixon (1863), 10 HLC 293, 308. Conversely, a former licensee may not raise invalidity as a defence in an action for unpaid royalties owing under the agreement that was formerly in effect: Crossley v Dixon (1863), 10 HLC 293, 308; Philco Products Ltd v Thermionics Ltd [1943] SCR 396, 419.

Now, while this rule is commonly referred to as an estoppel, it is not based on any recognized form of estoppel. Indeed, in every case I have found in which a recognized estoppel has been argued, it has been rejected: see Curtiss-Wright Corporation v Canada [1968] 1 ExCR 519 (rejecting estoppel in pais and promissory estoppel); Cropper v Smith (1884), 1 RPC 81 (CA) (rejecting arguments based on estoppel of record, estoppel by deed and estoppel in pais) aff’d on this point (1885) 10 AC 249 (HL) at 259. In Clark v Adie (No 2) (1877), 2 AC 423, 435-36 (HL) Lord Blackburn drew a loose analogy with the position of a tenant of lands who is his estopped from denying his lessor’s title so long as the lease remains in force, and the tenant has not been evicted from the land, he is estopped from denying that his lessor had a title to that land. However, Lord Kenyon in Hayne v Maltby (1789) 100 ER 665 at 666, rejected the analogy.

This should perhaps not be very surprising. Generally speaking, an estoppel is an equitable doctrine that supplements contract law by giving effect to some representation that is not supported by consideration, such as in promissory estoppel when a representation by one party to an agreement leads the other party to suppose that the agreement will not be strictly enforced, even though no consideration may have been given for the forebearance: John Burrows Ltd [1968] SCR 607 at 614-16; Curtiss-Wright Corporation v Canada [1968] 1 ExCR 519 at 530; and see generally John D McCamus, The Law of Contracts, (2005), Ch 8. It is sometimes said that the licensee estoppel arises “from the fact of the relationship itself,” (Fox on Patents 3^rd ed Vol I, 618; Fox 4^th at 320) but this is clearly not true; a licensee may raise invalidity as a defence if there is an express guarantee that the patent is valid, notwithstanding that the nature of the relationship is otherwise the same. In any event, it is very difficult to see how the act of entering into an agreement could in itself give rise to an estoppel making it inequitable to insist on the terms of the agreement.

In Lear the US Supreme Court at 656 explained that

The theory underlying this doctrine is that a licensee should not be permitted to enjoy the benefit afforded by the agreement while simultaneously urging that the patent which forms the basis of the agreement is void.

Perhaps this was the theory in US law, but it cannot be the explanation for the Anglo-Canadian rule; it is perfectly clear, as discussed above, that when sued for infringement a current licensee may indeed enjoy the benefit of the licence agreement while simultaneously urging that the patent is void.

So, we have a bit of a puzzle. Licensee estoppel is an estoppel that is not related to any other estoppel, that does not prevent either party from relying on the strict terms of the contract, and that operates only in respect of a claim brought on the contract. In short, it looks a great deal like a contract doctrine.

And indeed it is. The basis for the “estoppel” doctrine is really very simple. A licensee cannot raise invalidity as a defence to an action for unpaid royalties under a licence because the validity of the patent is normally immaterial to the obligation to pay under the contract. This has nothing to do with estoppel; it is purely a matter of contract. The licensee is contractually obliged to pay and unless otherwise provided in the contract, that obligation is not contingent on the validity of the contract. The “estoppel” such as it is, does not arise from the relationship between the parties, it arises from the terms of the contract. See Aidan Robertson, “Is the Licensee Estoppel Rule Still Good Law? Was It Ever?” 10 EIPR 373 (1991) making much the same point.

This explains what might otherwise be counter-intuitive aspects of the doctrine. It is why the actual invalidity of the patent, established in different proceedings, has not afforded a defence. It is not because the courts have buried their heads in the sand, like ostriches, resolutely ignoring the fact of invalidity; rather, the courts have in effect said, “Yes, we know the patent is invalid, but that doesn’t affect your obligations under the contract.” That is why fraud is an exception; contracts induced by fraud are not enforceable as a matter of general contract law doctrine. That is why an express provision is an exception; it is no exception at all, as the question is simply a matter of interpreting the contract, and an express term will always trump an implied term. That is why a licensee in good standing can always raise invalidity as a defence to an infringement action; invalidity is always material to the licensee’s obligation to pay damages for infringement, even when it is not material to the licensee’s obligation to pay royalties. That is why licensee estoppel is not related to any other established form of estoppel: it is not an estoppel, but purely a matter of contract. When an allegation of invalidity by the licensee is struck, it is simply on the basis of the procedural rule that allegations that can have no effect on the outcome of the case should be struck as a waste of judicial resources.

The more interesting question is why, as a matter of contract law, the obligation to pay royalties does not turn on the validity of the patent (unless of course, the agreement itself so provides). The answer is most clearly articulated in the early cases, decided before the matter was settled simply by reference to authority. It follows from two principles. First, a warranty or guarantee of validity will not normally be implied into a licence agreement; absent express language or special circumstances, either a licence or assignment is presumed to be transferring only such rights as the letters patent conferred upon the patentee: Electric Fireproofing Co of Canada v Electric Fireproofing Co (1910) 43 SCR 182 at 184-85 (Davies J), 189-90 (per Duff J), 195 (Anglin J) (assignment); Vermilyea v Canniff (1886), 12 OR 164 (OntHCJ), 169 (licence); Duryea (1910), 21 OLR 165 [10]-[14] (reviewing the cases); Anderson v EJ Shepard Ltd (1930), 66 OLR 105 (Ont CA) [19] (licence); Rymland v Regal Bedding Co Ltd (1966), 51 CPR 137 (Man CA) 141 (licence). This is because the validity of the patent turns on objective considerations—the early cases dealt primarily with anticipation by a third party publication—and not on information peculiarly within the knowledge of the patentee, and therefore the licensee is generally as well placed as the patentee to assess the validity of the patent: Hall v Conder (1857) 140 ER 318, 327; Taylor v Hare (1805) 127 ER 461, 462; Lawes v Purser (1856) 119 ER 1110, 1112; Vermilyea v Canniff (1886), 12 OR 164 (OntHCJ), 169; Rymland v Regal Bedding Co Ltd (1966), 51 CPR 137 (Man CA) 141. As the exception that proves the rule, if the patentee had actual knowledge of invalidity, this would likely constitute fraud that would enable the licensee to raise invalidity even in an action for payment under the contract: see Lawes v Purser (1856) 119 ER 1110, 1112. Moreover, this is not a strict rule as such; it is simply the usual outcome of interpreting the contract in light of the fact that the licensee is normally in as good a position to assess validity as the licensor. Thus the most accurate statement of the law is that “Courts are not prepared to readily extend the meaning of doubtful words so as to convert them into an express warranty of the validity of a patent”: Rymland 141.

Second, invalidity of the patent does not normally constitute complete failure of consideration, as even an invalid patent may effectively confer the benefit of exclusivity until such time as it is challenged: Vermilyea v Canniff at 169; Lawes v Purser (1856) 119 ER 1110, 1112; Taylor v Hare (1805) 127 ER 461, 462. So, in the early English case of Taylor v Hare, the licensee had practised the invention for 6 years before discovering prior art said to anticipate, and sought on that basis to recover all the royalties that had been paid to date. The court rejected the action, with Heath J stating at 462 “It might as well be said, that if a man lease land, and the lessee pay rent, and afterwards be evicted, that he shall recover back the rent, though he has taken the fruits of the land.” A licence agreement may also involve a transfer of know-how or confidential information, or more than one patent, so that it would be unwarranted to read in a provision absolving the licensee of the obligation to pay on the basis of the invalidity of one patent. Consequently, even when there is language in the agreement relating to the validity of the patent, the courts will not interpret it as absolving the licensee of the obligation to pay royalties (unless it is expressly to that effect); it is more likely to be interpreted as a warranty, breach of which might be the basis for a counter-claim in damages: see eg Hall v Conder (1857) 140 ER 318, 323; Gillard v Watson (1924) 26 OWN 77 (OCA).

In summary, there is no rule of licensee estoppel, because there is no rule and it is not an estoppel. While a licensee cannot normally raise invalidity as a defence to an action for payment of royalties, this is simply because the obligation to pay royalties under the licence is not normally conditional on validity.

No-Challenge Clause in Settlement is Enforceable

Loops LLC v Maxill Inc 2020 ONSC 5438 Lederer J: Swinton, Penny, JJ revg 2020 ONSC 971 Templeton J

            2,577,109 / Toothbrush

This interlocutory motion concerns the enforceability of a “no-challenge” clause in a settlement agreement. It also indirectly raises the question of whether Canadian law should follow the decision of the US Supreme Court in Lear Inc v Adkins 395 US 653 (1969) and abolish the doctrine of “licensee estoppel,” which (more or less) prevents a licensee from challenging the validity of the licensed patent. At first instance Templeton J held that the no-challenge clause at issue was not enforceable (see here). In this appeal, Lederer J, for a unanimous panel of the Ontario Divisional Court, reversed, holding that a no-challenge clause in a settlement agreement is indeed enforceable according to its terms. At the same time, he left the door open to the possibility that Lear might be followed in a different context as part of the incremental evolution of the law; my sense is that he did not particularly favour such a development, but was simply concerned not to decide any more than was necessary to dispose of this motion.

In this post, I will summarize Lederer J’s decision. I agree entirely with his reasons, both in the substantive analysis and in his balanced approach to legal change. In two subsequent posts I will address issues that are raised by the decision, but which go beyond what was required by this appeal. My next post will review the Anglo-Canadian law and argue that there is no such thing as licensee estoppel; more precisely, “estoppel” is a very unhelpful label for a doctrine grounded entirely in contract law and not on any kind of equitable principle. In my third post I will argue that Lear is not sound as a matter of policy and should not be followed in any context, even as a part of a more incremental development of the law.

Facts

Loops and Maxill are in the toothbrush business. In 2012 Loops sued Maxill in Canada for infringement of Loops’ 109 patent. Two years later they reached a global settlement, which contained a “no-challenge” clause stating that Maxill would not “directly or indirectly assist any person attacking the validity” of either the Canadian 109 patent or the corresponding US Patent 8,448,285 [3]–[5], [ONSC 971/15]. Maxill then created another toothbrush to try to design around the 109 patent. In 2015 Loops sued Maxill in Canada over the new toothbrush, this time for breach of the Agreement [7], [ONSC 971/20]. In 2017 Loops also sued Maxill in the US, alleging the same new toothbrush infringed the US 285 patent [9]. Shortly thereafter, a US subsidiary of Maxill, Maxill (Ohio), brought an action seeking a declaratory order that the 285 patent was invalid [10]. The US actions were consolidated in Washington [11]. The result of the US consolidation is that Maxill itself was attacking the validity of the US patent, notwithstanding the no-challenge clause in the settlement agreement; under Lear, as interpreted in the US 9^th Circuit (which includes Washington), the no-challenge clause is unenforceable [ONSC 971/62]. Loops then brought this Canadian motion for an interlocutory injunction prohibiting Maxill from challenging the validity of the US 285 patent in the US litigation, on the basis that this was a breach of the no-challenge clause in the Agreement [ONSC 971/1]. In effect, Maxill is seeking a partial anti-suit injunction to enforce the no-challenge clause in the settlement agreement.

Because the injunction sought involved the enforcement of a restrictive covenant, this required Loops to make out a “strong prima facie case” on the merits [15]. In the decision under appeal, Templeton J held that Loops had failed to make out a strong prima facie case, for three reasons, the first two of which are largely peculiar to the case. The Divisional Court reversed on all points.