The Duty to Disclose “The Invention”: The Wrong Tool for the Job

Teva Canada Ltd v Pfizer Canada Inc 2012 SCC 60

2,163,446 – Sildendafil – VIAGRA

My initial response to the SCC’s Sildenafil decision was quite positive about the general reasoning, though I was somewhat more critical of the doctrinal details of the decision. I have since completed a thorough analysis of the decision, which has caused me to substantially revise my opinion. In law, the doctrinal details matter, and in my article I argue that by using the wrong tool for the job, the SCC in Sildenafil did a poor job on the facts and damaged the tool of disclosure. My article, entitled “The Duty to Disclose ‘The Invention’: The Wrong Tool for the Job,” is forthcoming in the IPJ and is available in draft from SSRN. Here is the abstract:

In its Sildenafil decision, the Supreme Court of Canada held Pfizer’s Viagra patent to be invalid for failure to disclose “the invention,” though it provided no explicit definition of that term. I argue that on the best reading of the decision, “the invention” means the new, useful, non-obvious contribution to knowledge which is disclosed in the patent. While this is initially attractive, I argue that on this definition, a duty to disclose the invention is unsound as a matter of law and policy, because it implies that if any important claim in a patent is invalid, all the claims will be invalid. This eviscerates section 58 of the Patent Act, which provides that claims stand or fall independently. I argue that the Court’s reasoning was motivated by a mistaken belief that at the time of the application Pfizer had tested the other claimed compounds and knew them to be ineffective in treating ED, and that it claimed those compounds primarily to conceal the true invention. I argue that even if the Court’s understanding of the facts was correct, a duty to disclose the invention is not the appropriate tool to address the issue. The case properly raises questions relating to the duty of good faith disclosure, not a duty to disclose the invention.

My full analysis of the Sildenafil decision also confirms my initial view that Zinn J was wrong to grant summary judgment to Apotex in the subsequent impeachment action, Apotex v Pfizer / sildenafil 2012 FC 1339, in which he held the ‘446 patent to be of no force or effect, in part on the basis that “the determination by the Supreme Court in its Reasons that the '446 Patent fails to meet the requirement of sufficient disclosure pursuant to subsection 27(3) of the Patent Act is a legal determination binding on this Court and is dispositive of Apotex's claim in this action” [30]. While it is remarkably difficult to pin down precisely the core of the SCC’s legal reasoning in Sildenafil, the decision was clearly motivated by the belief that Pfizer had tried to “game” the system [Sildenafil 80; and see Imatinib 2013 FC 141, blogged here, in which Snider J at [387] emphasized this aspect of Sildenafil]. The SCC held this view because it apparently believed that Pfizer had tested the other compounds and knew them to lack utility. This is a factual question, and the SCC’s view appears to be wrong, in large part because the nature of the NOC proceedings did not allow the question to be fully explored. This illustrates the wisdom of the current rule that the results of NOC proceedings are not binding in a subsequent infringement or impeachment action. As the SCC said in Eli Lilly & Co v Novopharm Ltd [1998] 2 SCR 129, in affirming the rule developed in the Federal Courts (SCC’s emphasis):

[95] [T]he summary judicial review that is to be conducted on an application for a prohibition order under the Regulations is highly fact-specific and is generally considered to be binding only on the parties in the specific litigation. This is only appropriate, given the limited nature of the proceedings, the question that is to be answered, and the record generated for this limited purpose. [my emphasis]

[96] This point was reinforced more recently by Strayer J.A. in David Bull Laboratories, supra, at p. 600:

This Court is not prepared to accept that patentees and generic companies alike have been forced to make their sole assertion of their private rights through the summary procedure of a judicial review application. [SCC emphasis]

Note that in Eli Lilly v Novopharm the question on which the SCC refused to provide a binding declaration was whether the supply agreement constituted a sublicence. As a question of contractual interpretation, this would be a question of law, and despite having answered the question, the SCC refused to make a binding declaration, saying “For this Court to make a binding declaration concerning the private rights and obligations of the parties to the agreement would go well beyond the limited scope of the proceeding. Accordingly, I would deny the declarator relief requested by Novopharm” [97]. The same reasoning must surely hold with respect to the question of invalidity in Sildenafil, and for the same reason; my article on Sildenafil illustrates the perils of making a binding declaration based on the basis of the summary NOC proceeding.

A Principled Approach to Awarding an Accounting of Profits

Varco Canada Ltd v Pason Systems Corp 2013 FC 750 Phelan J
            2,094,313

The remedy of an accounting of profits has undergone a resurgence in Canada since the 1980s, and it is now regularly sought by a successful patentee. There are two major questions that have emerged: (1) what is the basis for the calculation of the quantum; (2) what is the basis for an entitlement to elect an accounting? The first question has been greatly clarified by the landmark SCC decision in Monsanto v Schmeiser, 2004 SCC 34 and the subsequent FCA decision in Monsanto v Rivett FCA 2010 FCA 207 which largely affirmed the approach taken by Zinn J at first instance, 2009 FC 317. Phelan J’s decision in Varco suggests that the Federal Court may be approaching a principled and theoretically sound consensus on the second question.

Canadian Duty of Candor in Response to a Requisition is Not the Same as the US Duty to Disclose Material Prior Art

Varco Canada Ltd v Pason Systems Corp 2013 FC 750 Phelan J
            2,094,313

The main invalidity argument made by the defendants in Varco v Pason was that the patent had been abandoned pursuant to s 73(1)(a) of the Act by reason failure to reply in good faith to the Examiner’s requisition [341]. The requisition asked for “an identification of any prior art cited in respect of the corresponding United States and European patent applications.” This reply to requisition duly provided the references cited to the USPTO [342]. The defendant argued that the disclosure to the USPTO did not comply with the US disclosure requirements, because it failed to disclose material prior art, and this failure to comply with the US disclosure requirements implied lack of good faith in responding to the Canadian requisition, even though even though the response did respond strictly to the requisition, in that all references cited to the USPTO had been identified [345].

The first difficulty with this argument is that in Corlac v Weatherford 2011 FCA 228 [150] the FCA held that s 73 cannot be used to attack the validity of a granted patent. Curiously, Phelan J did not rely on this holding, even though he had made the same point at trial in Weatherford 2010 FC 602, [344].

Instead, Phelan J emphasized that the US disclosure requirements are different from the Canadian disclosure requirements [348-49]:

Section 73 of the Canadian Patent Act requires a good faith response to a requisition from one examiner, rather than the free standing duty as described in US law.

It is not accurate to say or suggest that Canada does not have a duty of candor – it is just not of the same nature as the US principle. The question is “good faith in respect of what?” In Canada the good faith and duty of candor is in response to a question or inquiry. The response has to be fair and responsive, but it need not respond beyond that which is requested.

Whether or not the US disclosure requirement had been complied with, the relevant question was whether the Canadian requirement had been complied with [352]. In this case, it had been: “the term 'prior art cited', as used in the examiner’s requisition, meant the prior art reviewed by an examiner and applied against a foreign application. It does not mean any prior art listed in a search report; nor does it mean all the prior art of which an applicant is aware,” [357] and “There is no requirement to go beyond answering that which is asked or requested” [360].

Phelan J made it clear that he did not regret the narrower scope of the Canadian requirement:

[364] Despite the not so veiled invitation of the Defendants to have this Court go down the US route of “inequitable conduct”, I will not do so. The Canadian legislation is clear, the obligation is to respond directly and honestly. The US principle is highly unsatisfactory, not supported by Canadian principles or practice, and has been described by US witnesses as a “mess”.

This embraces the message of the FCA in Weatherford: the US experience with the inequitable conduct doctrine shows that an expansive duty of good faith disclosure is a cure worse than the disease, and it will not be judicially read into the Canadian Act.

Anticipation and Obviousness in Varco v Pason

Varco Canada Ltd v Pason Systems Corp 2013 FC 750 Phelan J
            2,094,313

Yesterday’s post looked at claim construction and infringement. The anticipation, obviousness and utility issues in Varco v Pason were also fairly straightforward, though the facts provide an interesting illustration of the problems of avoiding novelty-destroying disclosure of the invention during pre-filing field testing. The most interesting legal point is Phelan J’s comments regarding the relationship between the US and Canadian duties in disclosure to the patent office, made in relation to the abandonment argument, which I will discuss in a separate post.

Anticipation
The defendant’s anticipation attack based on the prior art wholly failed, as the closest prior art cited by the defendant did not come close to anticipating the essential elements of the invention [273-288]. The argument based on the inventor’s field testing [289]-[308] was more interesting, particularly when compared with Wenzel Downhole Tools 2012 FCA 333. New designs in downhole rig equipment have to be tested on operating wells, often wells operated by a third party, and the question is what steps have to be taken in order to avoid disclosing the invention during the testing process.

In Wenzel the question was whether a third party’s downhole bearing assembly had been disclosed during its testing. The assembly embodied all the essential elements of the invention and so would be anticipatory if it was made available to the public. Since the drilling company that had tested the assembly was not bound by any obligation of confidence, the question was only whether the assembly had been “made available.” The bearing assembly was enclosed in a metal casing giving no hint of what was inside. Had the user opened the casing and inspected the assembly, it would have been able to discern the essential elements of the invention, but it would not have been usual or easy for the user to open the casing and on the facts in Wenzel the user made no effort to do so. As discussed here, the FCA, in a split decision, held that this was anticipatory: “being available does not require that access to the information be easy, simply that it be possible using known methods and instruments” [69], and the test is “whether, in this case, there was an opportunity to access the relevant information during the rental” [74, original emphasis].

In considering whether the inventor’s field testing in Varco was anticipatory, Phelan J distinguished Wenzel on the facts. As in Wenzel, inspection of the exterior would not reveal the working of the invention, but in contrast to Wenzel, the inventor in this case took steps to ensure that the device could not be inspected, by locking the device when he was absent [297], [301]. And while there was no express confidentiality agreement with the driller, that was not fatal: “Although there was no evidence as to either industry practice or corporate understanding, Bowden’s actions to keep the invention from being observed or disclosed is consistent with his intent to keep it confidential. There is no suggestion that Union Pacific thought that the testing was not confidential nor is there evidence of enabling disclosure during the test period” [304].

While the facts do legitimately distinguish Wenzel, this case illustrates the potentially difficult nature of the “opportunity to access” test. Here, the driller had opportunity to access the invention in the sense of having physical access to the device in the inventor’s absence. It would no doubt have taken lock picking tools to access the device, which the driller presumably would not have had on hand, but in Wenzel it would also have taken special tools, available only off-site, to inspect the interior of the device. It seems to me that the real distinction between the cases is with respect to intent. In Wenzel it did not appear that the third party developer of the prior art device was attempting to keep the device secret, whereas the inventor in Varco clearly was, in anticipation of obtaining a patent. As I argued in my post on Wenzel, this is consistent with a purposive analysis of the novelty requirement. If the third party in Wenzel did not care whether the device was made public, it evidently did not think the lure of a patent was necessary to provide an incentive to develop the device. In this case, that the inventor took steps to the keep the device secret, shows that he was relying on the patent incentive. The difficulty with this analysis is that anticipation should be an objective test. I suggest that these cases show that while it is an objective test, in borderline cases subjective considerations may become relevant to tipping the balance one way or the other.

Obviousness
Phelan J relied heavily on long-felt need and commercial success in holding that the invention was not obvious: see [316], [319], [320], and especially [321]. He started his analysis with the question: “if it was so obvious, why was it not done before?” (See also Nichia v Argos [2007] EWCA Civ 741 [21], Beloit v Valmet 8 CPR (3d) 289, 295.) While these considerations are sometimes disparaged as being “secondary factors” which must be kept firmly in its place,” in my article "Secondary” Evidence of Obviousness is Not Secondary, (2012) 28 CIPR 279, I have argued in that when this question is not rebutted, as in this case, it speaks far more powerfully to non-obviousness than the evidence of expert witnesses, which is inevitably tainted by hindsight, as Phelan J pointed out [316].

Correction of Error in the Claim
The only utility argument of significance turned entirely on a drafting error: the words “increased” and “decreased” had been switched in Claim 9. Interpreted literally, the invention as claimed would not work. It was uncontroversial that a skilled reader would have readily recognized that “increased” should be read to mean “decreased”, and vice versa [336], and Phelan J accordingly interpreted the claim that way, with the result that the utility challenge failed. This is consistent with a long line of authority that the entire specification must be read with “a mind willing to understand,” and consequently readily discernable errors, even in the claims, will be read as corrected: see eg Procter & Gamble Co. v. Bristol-Myers Canada Ltd. (1979), 42 CPR (2d) 33, 36-37 (FCA); 2009 FC 991, [159]; 2010 FC 1265, [99].

(Thanks to Alan Macek's IPPractice blog for posting this decision before it has appeared on the FC website. )

Background and Infringement in Varco v Pason

Varco Canada Ltd v Pason Systems Corp 2013 FC 750 Phelan J
            2,094,313

The ‘313 patent at issue in this litigation is for an automatic drilling system for use in the petroleum industry. The most interesting part of the decision concerns remedies, and particularly the question of entitlement to an accounting. This post provides an overview of the facts and infringement. Subsequent posts will deal with validity and remedies.

Objective ("Secondary") Obviousness Factors in the Federal Circuit

In my article “Secondary” Evidence of Obviousness is Not Secondary, (2012) 28 CIPR 279 (draft version here), I argued that the dictum that the “primary” evidence of obviousness is that expert witnesses, and that “secondary” evidence “must be kept firmly in its place,” is unsound in law and policy. Both types of evidence have strengths and weaknesses and both must be assessed with due regard to their relative advantages. Expert witness evidence is susceptible to the hindsight bias, while it may be difficult to establish a causal link between objective (“secondary”) evidence and obviousness. In its recent decision in Plantronics, Inc v Aiph Inc (Fed Cir 2013), the US Federal Circuit has reaffirmed that failure to consider objective evidence is an error of law (slip op 18-19):

As a safeguard against ‘slipping into use of hindsight and to resist the temptation to read into the prior art the teachings of the invention in issue,’ we have required courts to consider evidence of the objective indicia of nonobviousness prior to making the ultimate determination of whether an invention is obvious. to . . . [A]ll evidence pertaining to the objective indicia of nonobviousness must be considered before reaching an obviousness conclusion. The significance of this fourth Graham factor cannot be overlooked or be relegated to ‘secondary status.’

Update: See also Leo Pharmaceutical Products, Ltd v Rae, 2012-1520 (Fed Cir 2013) in which the Fed Cir emphasized that “Objective indicia of nonobviousness play a critical role in the obviousness analysis. . . . This case illustrates a good reason for considering objective indicia as a critical piece of the obviousness analysis: Objective indicia “can be the most probative evidence of nonobviousness in the record, and enables the court to avert the trap of hindsight.” Crocs, Inc. v. Int’l Trade Comm’n, 598 F.3d 1294, 1310 (Fed. Cir. 2010)”

No New Cases for the Week of 4 August

No new patent / NOC / data protection cases were posted on the Federal Courts' websites for the week of 4 August 2013.

Gauthier J on Disclosure of Factual Basis for Sound Prediction

Sanofi-Aventis v Apotex Inc / clopidogrel 2013 FCA 186 Pelletier JA, concurred in by Noël, Gauthier JJA; separate concurring reasons by Gauthier JA, rev’g 2011 FC 1486 Boivin J
            1,336,777 – PLAVIX

In Lilly v Apotex / raloxifene 2009 FCA 97, [15] aff'g 2008 FC 142 [164], and subsequently in Novopharm v Lilly / atomoxetine 2011 FCA 220, [46-47] aff’g 2010 FC 915, [116-117], the FCA held that when utility is based on sound prediction, the factual basis for the prediction must be disclosed in the patent itself. The ultimate authority for this was a passage in the SCC decision in Wellcome / AZT 2002 SCC 77, [70]. In my article, “Must the Factual Basis for Sound Prediction be Disclosed in the Patent?” (2012) 28 CIPR 39 (draft version here), I argued that Wellcome / AZT did not mandate such a rule. I also argued that the requirement is unsound, in part because it creates a sharp distinction between demonstrated utility and sound prediction. Such a distinction cannot be supported on the Act, which contains only a single utility requirement.

The requirement to disclose the factual basis for sound prediction is conceptually distinct from the promise of the patent doctrine, but the two are linked in practice; the promise doctrine raises the bar for demonstrated utility, making it more likely that sound prediction will be invoked, and the factual basis doctrine makes it more difficult to establish sound prediction. This is illustrated by Boivin J’s decision at first instance in Clopidogrel. As explained here, because he held that the patent promised a relatively high level of utility, he found that the promise was not met. Boivin J held that there was, however, a good factual basis and sound line of reasoning supporting a sound prediction of utility, but the test for sound prediction was not satisfied because the factual basis was not disclosed in the patent itself. But for the promise doctrine, sound prediction would not have been raised, and but for the factual basis doctrine, the test for sound prediction would have been satisfied.

The Promise of Selection Patents

Sanofi-Aventis v Apotex Inc / clopidogrel 2013 FCA 186 Pelletier JA, concurred in by Noël, Gauthier JJA; separate concurring reasons by Gauthier JA, rev’g 2011 FC 1486 Boivin J
            1,336,777 – PLAVIX

The ‘777 patent at issue in Clopidogrel is a selection patent over Canadian patent 1,194,875. This decision illustrates how the promise of the patent doctrine has distorted the analysis of selection patents in Canadian law, but it also illustrates that ultimately nothing turns on this. This reinforces the view that it may be time to abandon the doctrine entirely.

In Clopidogrel the FCA re-affirmed the point made by the SCC in Sanofi 2008 SCC 61 [9] and subsequently by the FCA in Olanzapine (No 1) 2010 FCA 197 [33] that “a selection patent is like any other patent” [44], in that it must satisfy the requirements that the invention be new, useful and non-obvious. It is also clear that in order to be valid, a selection patent must possess some unexpected property as compared with the genus, be it the presence of an advantage or the avoidance of a disadvantage. The question then, is whether the unexpected property is required as a matter of utility, or as a matter of non-obviousness?