Prosecution History and Summary Judgment

Canmar Foods Ltd v TA Foods Ltd 2019 FC 1233 Manson J
            2,582,376 / method for roasting oil seed

Canmar is the first case to invoke the new s 53.1, which provides that the prosecution history may be used in an action to rebut any representation made by the patentee regarding claim construction. As importantly, it is one of the first patent cases in many years to grant summary judgment on non-infringement / invalidity.

My previous post on Canmar discussed Manson J’s holding that foreign prosecution history may, in some conditions, be considered pursuant to s 53.1. I argued that despite this holding, Manson J did not actually consider the US prosecution history. He undoubtedly did consider the Canadian prosecution history, and moreover, he considered it in order to dispose of the action by way of summary judgment. In this post I will suggest that Manson J used summary judgment in order to side-step the restrictive approach to summary judgment that seems to have prevailed in patent cases for some time. The key claim construction question at issue is one that would often call for expert evidence, but invoking prosecution history turned the issue into a matter of law instead. In effect, Manson J solved a problem in the law relating to summary judgment by invoking prosecution history. The difficulty is that if indeed there is a problem with an overly restrictive approach to summary judgment, prosecution history is not a cure; it will work in only cases where the key prior art was the subject of an objection, and not where the applicant had drafted around it from the outset. Moreover, it is always risky to try leveraging one area of the law to fix a defect in another area, as that will often lead to law that is deficient in both areas.

Summary judgment and prosecution history

Summary judgment is potentially an important tool for preventing abuse of the patent system. Patent litigation is notoriously expensive, and the mere threat of patent litigation and the associated costs may have a chilling effect on legitimate activity, even if the defendant could be confident of prevailing if the patent went to trial. The problem is most prominent with patent assertion entities (PAEs) particularly those that the FTC has described as “Litigation PAEs,” whose strategy is to assert dubious claims against end-users and small entities, and settle for amounts below the cost of defending a case through the end of discovery. But the problem is not confined to PAEs. The high cost of defending a patent action makes it a potent weapon for putting pressure on a competitor, particularly when the defendant is a small or medium sized company, regardless of the merits of the claim. The problem is severe enough that some respected scholars argue that the patent system is a net hindrance to innovation in many industries. Regardless of whether the patent system is good or bad overall, there is no question that the cost of patent litigation lends itself to abusive practices that bring the whole patent system into disrepute. Consequently, when an action clearly lacks merit, it is not enough that counsel for the defendant can advise their client that they will surely win; it is important that they will win cheaply.

On its face, Canmar was a case that cried out for summary judgment. Canmar’s 376 patent claim, a method for roasting oil seeds, with flax seed being the product at issue, comprised

heating the oil seed in a stream of air for less than 2 minutes … [and] transferring the heated oil seed into an insulated or partially insulated roasting chamber or tower. [8, emphasis added]

Roasted flax seeds as such were not claimed [22]. The uncontested evidence was that the defendant TA Foods used a “Micronizer” to roast flax seeds:

[18] The Micronizer operates by subjecting oil seed to infrared radiation as the seed travels along a vibratory steel plate under a series of heated ceramic tiles. …

[19] From the vibrating steel plate, the oil seed exits the Micronizer down a slide into a triangular shaped hopper. The hopper is made of steel plating, and is uninsulated. From the bottom of the hopper, the seed travels up an uninsulated auger trough to an uninsulated cooling tower made of steel sheeting.

Patents are property rights and the claims provide public notice of the bounds of the monopoly, by “defining distinctly and in explicit terms the subject-matter of the invention for which an exclusive privilege or property is claimed” s 27(4); Whirlpool 2000 SCC 67 [42]. In some cases, perhaps many, claim construction is complex, requiring expert evidence as to the meaning of technical terms like “pentahydric” or “heterogeneously branched polymer.” But Canmar doesn’t look like such a case. One might have thought that it would not be so difficult to decide that an infrared heater is not a stream of air, and that “insulated” does not mean “uninsulated.” We might even aspire to a patent system in which the defendant, who knows what its own technology is, could read the patent and feel secure that it was not infringing, even without obtaining legal advice.

And indeed, in one sense this really was an easy case. On its face, and as ultimately construed by Manson J, the claimed invention requires that the seed be heated in a stream of air [92] and transferred into an insulated roasting chamber [98]. On its face, and as ultimately found by Manson J, the defendant’s process used infrared radiation, not a stream of air [92], and transferred the product to an uninsulated hopper, not an insulated receptacle [99]. But that hardly matters if the cost of winning is more than the product line is worth. Fee-shifting does not really affect the calculus: in principle Canada follows the English rule that costs follow the cause, but in practice we are much closer to the American rule that each side bears its own costs. That is why it is important not just that the patentee in this case lost, but that it lost on summary judgment.

Despite its simplicity, there was a hurdle to the expeditious disposition of this case. Some claim elements are not “essential,” which means in effect that they can be read out of the claim: Free World 2000 SCC 66 [20]. If the elements “in a stream of air” and “insulated or partially insulated” are not essential, then it does not matter that they were not used in the defendant’s process. One aspect of the inquiry is that an element is non-essential if it obviously does not make any difference to the way the invention works: Free World [31(e)(iii)]. But that inquiry is inherently factual, so the determination of essentiality is a hurdle to summary judgment — unless there is some way of determining essentiality without addressing that aspect. (Further, in general, even essential claim elements typically require expert evidence to interpret purposively: even if it is conceded that a term like “heterogeneously branched polymer” is essential, the court may require expert assistance to understand what that means. This concern did not play a significant role on the facts in Canmar.)

Thus, while it is well-established that claim construction is a matter for the court, this does not mean that expert witnesses are unnecessary. Rather, it means that the role of the expert is only to put the court in the position of being able to construe the claims purposively, and that the court is entitled to adopt a construction of the claims that differs from that put forward by the parties: Whirlpool [57], [61]. To the extent that the meaning of a term, or its essentiality, turns on the facts, experts will be necessary to educate the judge.

This creates a difficulty for summary judgment, as the FCA has held that issues of credibility, including conflicting expert evidence, “ought not to be decided on summary judgment applications”: Trojan Technologies 2004 FCA 140 [20] rev’g 2003 FC 825, citing MacNeil Estate 2004 FCA 50; and see Wenzel Downhole 2010 FC 966 [31] indicating that construction of the claims requires expert evidence and should not be decided on summary judgment; but see Sterling Lumber 2010 FCA 21 granting summary judgment on de novo review, albeit on the basis of admissions and presumptions of law.

Manson J was keenly aware of the need to use summary judgment to promote expeditious disposition of patent cases. He 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]. The reference to a culture shift evidently reflected the remarks of Stratas JA in Amgen 2016 FCA 121 [24], but that comment related to a different issue (questionable tactical conduct), and in Manitoba v Canada 2015 FCA 57 [11]-[17], Stratas JA seemed to take a more limited view of the effect of Hryniak on the summary judgment rules.

In any event, despite his “culture shift” remark, Manson J did not really rely on a changed approach to summary judgment. Instead, he relied on jurisprudence to the effect that “Where the judge can construe the patent as it would be understood by a POSITA, expert evidence is not required” [80]. In particular, expert assistance may not be required “if the meaning of a term is evident from the patent specification,” (Excalibre Oil Tools 2016 FC 1279 Manson J [119]), or if a determination can be made “based on law and the application of uncontroverted relevant evidence or admissions or plain and obvious findings on the evidence” (Pfizer Canada Inc v Pharmascience Inc 2007 FC 188 Hughes J [16]). The principle reflected in these authorities is that the court does not require expert evidence when it can rely instead on matters within its own expertise, namely construction of the words of the specification alone, or matters of law; or when the evidence is uncontroverted.

The evidence as to the defendant’s process, described above, was uncontroverted, so the key question potentially requiring factual evidence was whether the disputed claim elements were essential. Manson J held he was able to purposively construe the disputed claim elements based on “the claim, the disclosure, and the prosecution history” [81]; which is to say, without the need for a trial to resolve factual issues. In particular, the prosecution history enabled Manson J to come to the conclusion that the two contentious elements were essential [87].

Thus, the prosecution history was important in this case, not because it resulted in a construction different from what would have been arrived at without its aid. On the contrary, the prosecution history merely served to confirm a claim construction that was apparent on its face. Prosecution history was important because it turned what would have been a matter of fact, requiring expert witness and a trial, into a matter of law, which could be determined by the court on summary judgment.

While I entirely agree with Manson J’s desire for an expanded use of summary judgment, I have some concerns about using prosecution history for that purpose. Before elaborating, I should emphasize that I do not disagree with the result. On the contrary, based on the description of the prior art in the US prosecution history, I agree entirely with Manson J’s conclusion that the disputed elements were essential to patentability. In particular, the “stream of air” limitation was added to avoid a rejection over GB2182234 (Newton), which disclosed roasting oil seeds using infrared radiation — the same method used by the defendant in this case: Applicant remarks of 10-11-2011, p12-13. (The prior art relating to the “insulated or partially insulated” limitation was even more extensive.) I also agree that the matter should have been disposed of by summary judgment. It would have been an abuse of the patent system to burden the defendant by allowing the matter to go to trial.

Canadian prosecution history

The first issue relates to the use of the Canadian prosecution history. Manson J quoted passages from the plaintiff’s letter of 24 August 2012 to the Canadian Patent Office, responding to the obviousness rejection of 27 Feb 2012, in which the applicant stated that the “newly submitted claims are much narrower in scope than the previously examined claims in view of the introduction of significant limitations derived from the description and previous dependent claims,” and that the new claims were submitted “to encompass both novel and non-obvious subject matter,” [86] as “clear evidence of the inventor’s express intention that these two elements, introduced to advance novel and non-obvious subject matter, are essential” [87].

I have some difficulty with this conclusion. Here is the relevant passage from the Canadian examiner’s obviousness rejection of 27 Feb 2012:

The examiner has considered the documents cited in the ISR available on or prior to 2012-01-30.

The claims on file do not comply with section 28.3 of the Patent Act. The examiner considers that the analysis presented in the international preliminary report on patentability (IPRP) in regard to inventive step is applicable under the Canadian standard for obviousness. In view of the lack of amendment to the claims, the deficiencies mentioned during international examination in regard to inventive step are hereby identified, in respect of the claims identified in the IPRP, as defects for the purposes of national examination.

Here is the entire passage from p2 of the applicant’s letter of 24 August 2012, responding to the obviousness rejection of 27 Feb 2012, with the portions quoted by Manson J italicized, and my underling of other passages:

At the outset, the Examiner is advised that the claims on file have been replaced with a new set of claims fully supported by the description as filed and believed to more accurately and appropriately define the scope of protection sought for the invention. New claims 1-19 correspond substantially to those submitted during prosecution of a related United States application. Former claims 22-27 have been renumbered as new claims 20-25 flowing from new claims 1-19.

Turning to the contents of the Examiner's Report, the Examiner asserted that the claims previously on file failed to comply with Section 28.3 of the Patent Act in view of the analysis presented in the International Preliminary Report on Patentability (IPRP). Attached to the present response is another copy of the IPRP. From page 3 of the IPRP the Examiner will note that all claims previously on file were considered to harbor the attributes of novelty, inventive step and industrial applicability in view of references D1, D2 and D3 cited by the international Examiner. Therefore, the Canadian Examiner appears to be mistaken in asserting that the claims previously on file encompass obvious subject matter in accordance with the IPRP. Withdrawal of the objection is thus appropriate and respectfully requested.

In any event, the newly submitted claims are much narrower in scope that the previously examined claims in view of the introduction of significant limitations derived from the description and dependent claims. The new claims are submitted to encompass both novel and non-obvious subject matter.

The letter as a whole makes it clear that the limitations were not necessary to respond to Canadian examiner’s obviousness objection; on the contrary, the second paragraph asserts that that objection was ill-founded. (It certainly appears that the examiner’s objection was ill-founded, as the IPRP concludes at p3(6) that “For these reasons, claims 1 to 27 are considered to be the result of inventive ingenuity and are considered to comply with Article 33(3) PCT.”) There’s a bare assertion that the new claims are the same as the US claims, and that they are also new and non-obvious, without any link between the two. Now, if we assume the exchange took place with one eye on the US prosecution, we can read a bit more into it. I’d read the applicant’s letter as as saying, in effect, “Your obviousness objection was ill-founded, but the point is moot because I limited the claims in response to US novelty and obviousness objections, and these claims are new and non-obvious in Canadian law.’ This might be taken as an indirect admission that the limitations were necessary to novelty and obviousness in US law, but I have difficulty reading it as an admission that the same was true in Canadian law. Because of differences in the law — what if the US prior art was secret sales?— it is not clear that an admission related to US law is determinative in Canadian law, without proof of US law and consideration of the US objections themselves.

US prosecution history

In contrast, the US prosecution history, as discussed above, does contain what I read as a clear acknowledgment that the “insulated or partially insulated” and “stream of air” limitations were essential to novelty and non-obviousness. Entirely apart from the question of whether the scope of s 53.1 encompasses foreign prosecution history (discussed in my last post), the problem with using US prosecution history is that presumably, in principle, some expert evidence of US law would be required, as an admission that the claims had to be narrowed to overcome novelty or obviousness objections in US law does not necessarily imply that they would have been invalid in Canadian law. That might be reason to refuse summary judgment, at least on a stringent view of the requirements.

My point in all this is not that Manson J was wrong to consider this statement to be an admission. He had a fuller record and the benefit of argument. There are two broader points.

First, if an admission in the prosecution history is required to support summary judgment, summary judgement can be readily avoided in the future by using appropriate language. For example, in Actavis [2017] UKSC 48 [78] the applicant had responded to an objection by stating that "in order to expedite the application proceeding to grant," it had "elected to amend the claims so as to more closely reflect the explicit examples provided." The UKSC [88] considered that this did not constitute a concession that the claims would be invalid without the limitation. Such a response would be particularly apposite in a situation in which the examiner's objection was in fact ill-founded, as appears to have been the case in both Canmar and Actavis, but circumspect language could always legitimately be used. 

Second, to the extent that the admission is in the US prosecution history, prosecution history cannot be used to avoid the need for experts, so allowing summary judgment, because expert witnesses as to US law will be required. This is likely to be a pervasive problem, because, as Manson J pointed out, prosecution of Canadian patents “often follows prosecution of corresponding patent applications in other jurisdictions” [73].

General approach to prosecution history

One response, particularly to the first objection, might be to say that Manson J was applying too strict a standard in looking for an admission by the applicant that the claim limitation was necessary to overcome anticipation and obviousness objections. In US law, as I understand it, the applicant is not entitled to recover claim scope that was given up in response to an objection related to a matter of patentability. (This glosses over a number of details.) There is no requirement that the applicant admit that the objection was sound or that the limitation was necessary. On that view, the Canadian prosecution history might be enough to set up an estoppel. The effect is that the normal process of claim construction can be short-cut, at least for the limitations at issue.

The difficulty is that this goes to a substantive question on the use of prosecution history that was not at issue. In UK and German law, prosecution history is an additional consideration that “cannot be readily used as the sole basis for construction”: Actavis [2017] UKSC 48 [84], quoting with approval the BGH decision in Eli Lilly v Actavis Group PTC No X ZR 29/15. The pattern so far has been that the court will first construe the claims without reference to prosecution history, then turn to the prosecution history only if (i) necessary to clarify any ambiguity, or (ii) if “it would be contrary to the public interest for the contents of the file to be ignored” (Actavis [88]). Branch (i) clearly requires a prior normal process of claim construction. 

While the UK case law to date is limited, it appears that branch (ii) requires either an admission by the applicant (Actavis [88]), or a consideration of the entire context, including whether the examiner’s objection was substantively well-founded: Actavis [89]; Icescape [2018] EWCA Civ 2219 [79]. Like the US approach, branch (ii) of the UK approach is motivated by considerations of fairness, but it reflects a distinct conception of what fairness requires. Broadly speaking, US law implements a strict rule that the patentee cannot reclaim subject-matter given up in prosecution, on the view that is necessarily unfair to give up subject-matter that was surrendered in prosecution (at least in the circumstances where the various requirements of the doctrine of prosecution history estoppel are satisfied). Actavis [88]-[89], in contrast, suggests that it is not necessarily unjust to allow the patentee to reclaim subject matter that was given up in response to an ill-founded objection, so that prior claim construction and consideration of the merits would be required before the prosecution history could be invoked to limit the scope of the claims; though such an inquiry would be unnecessary in the face of a direct admission by the applicant that the limitation was necessary.

Of course, we can’t say at this point whether Canadian law will follow the US or European lead. Section 53.1 provides that the prosecution history “may” be used to rebut a representation made in an action. This certainly leaves the door open to an approach like that in the UK, where prosecution history will only be used in limited circumstances, but the statutory language is also consistent with a more aggressive use of prosecution history, along US lines. The open-ended language strongly suggests that the intent of the drafters was merely to open the door to the use of prosecution history, while leaving it to the courts to develop more detailed principles governing its use. And therein lies the rub. To use the prosecution history alone to determine whether the elements at issue are essential implies a very robust role for prosecution history in claim construction, along the lines of the US approach. Those questions were not addressed in the Canmar decision, and Manson J cannot be taken to have decided them by an indirect implication.

If the law of prosecution history does develop along US lines, so that it might be helpful in summary judgment, all well and good. But the question of when prosecution history might properly be used to limit the scope of claims as understood on a purposive consideration of the patent raises many difficult questions, as is illustrated both by the complex state of US case law, and by the divergent path so far taken in Europe. These issues must be addressed on their own merits; a desire to facilitate summary judgment should not drive the law related to the use of prosecution history.

Summary judgment redux

Recall the problem facing Manson J. The facts of the case demanded summary judgment, yet the case law suggests that summary judgment is inappropriate where expert evidence is normally adduced, such as for claim construction. The use of prosecution history promised a way out of this conundrum by turning the claim construction exercise into a matter of law, turning only on admissions by the patentee which could be discerned in the prosecution history. I have argued that this promise was not fulfilled, because the use of prosecution history in this case would require reference to the US history, and therefore experts on US law; and because the law related to the use of prosecution history is itself unclear.

More importantly, prosecution history on its own will not be able to revitalize summary judgment. Suppose that the applicant had found all the prior art cited by the US patent office at the outset, and had drafted accordingly, so that the original claims included both the controversial limitations and so had never met any objection. There would then be nothing in the prosecution history to suggest whether the limitations were essential, and the prosecution history would be useless in supporting summary judgment. Yet summary judgment on the basis of Newton would, in my view, be every bit as appropriate.

At the end of the day, the reason that summary judgment was appropriate in this case was not because of the prosecution history, but because of the prior art itself. Newton makes the point in respect of the “stream of air” limitation. Newton disclosed an apparatus and method for roasting oil seeds using infrared radiation. The allegedly infringing method roasted oil seeds using infrared radiation. The case can be disposed of without construing the claims on the basis of the Gillette defence; if “stream of air” is not essential, the claim is anticipated, and if it is essential, it is not infringed. Why wouldn’t it be possible to dispose of this case on summary judgment just on the basis of Newton, without worrying about prosecution history at all?

As an academic, I’m sure there are procedural niceties that escape me, and I certainly appreciate the need for the courts to limit themselves to their proper role, but this just doesn’t seem so hard. This post has been more than long enough, and I haven’t examined the summary judgment jurisprudence in detail, but if MacNeil really precludes the court from taking a hard look at Newton and disposing of the case on that basis, then so much the worse for MacNeil, and I hope that Manson J is right about the effect of Hryniak.