Ramipril Costs

Sanofi-Aventis / ramipril (Costs) 2012 FCA 265 Sharlow JA: Dawson, Trudel JJA

In this brief decision from the bench, the FCA dismissed nine appeals and four cross-appeals, relating primarily to Snider J’s costs awards in ramipril litigation between Sanofi-Aventis and Apotex and Teva, 2009 FC 1138 (Apotex) and 2009 FC 1139 (Teva / Novopharm). The appeals were dismissed on the basis that the costs awards were “well within the discretion of Justice Snider” [4]. Consequently, the neither the FCA decision nor the decision of Snider J establish a strong precedent on costs matters. Nonetheless, a several points from Snider J’s decisions are worth noting.

Scale of Costs

First, with respect to the scale of costs

In my view, the upper end of Column IV is appropriate, and not simply because this award "splits the difference". A review of recent jurisprudence on the issue of awards in intellectual property trials indicates that this scale recognizes the significance and complexity of the various issues in such a trial. [2009 FC 1138 ¶ 14, 2009 FC 1139 ¶ 13]

This is consistent with costs in Eurocopter v Bell, and higher than the scale used by Zinn J in the fenobibrate costs decision.

Offer to Settle

In the Teva / Novopharm case Snider J took into account an offer to settle that did not satisfy Rule 420 (emphasis added):

[22] Sanofi argues that Novopharm has not provided sufficient evidence to support a Rule 420 finding and, in any event, the settlement offers would not meet the criteria of Rule 420. I agree. However, I am also sympathetic to Novopharm's situation. It appears that, even if these offers were not formal or substantive enough to satisfy Rule 420, there is no doubt that they were made in a good faith effort to end the litigation. Further, while Sanofi now explains why the offers could not have been accepted, I see no evidence that Sanofi ever tried to make responding offers. In November 2008, Novopharm went so far as to draw up draft terms of settlement, to which there appears to have been no response. Even on the little evidence before me, I am satisfied that Novopharm made very serious efforts to settle the litigation both prior to and during the trial. Sanofi's attitude throughout, and its defensive, after-the-fact justification for its failure to properly consider the offers are simply not helpful. In the circumstances, while recognizing that the offers do not satisfy Rule 420, I am convinced that the settlement offers by Novopharm should result in an increase in the overall award of 50%.

In the Apotex case Snider J similarly took into account an offer to settle by Apotex , but increased the overall award by only 20% because the offer to settle was vague [22].

Bifurcation

In both cases over half the trial days related to remedies, which, with hindsight, turned out to be wasted as the patent was held to be invalid. Sanofi had brought a motion to bifurcate which had been resisted by both Apotex and Novopharm. Snider J ordered a 10% reduction in the overall award of costs to reflect Apotex and Novopharm’s responsibility for the unnecessary remedies phase. However, Snider J noted that in both cases resisting bifurcation was not initially unreasonable because there was a good faith belief that the remedies phase would be relatively brief.

As the reality became clear, the Plaintiffs could have brought a further motion or a motion for reconsideration. That did not happen. Further, Sanofi contributed to the length of this phase of the trial by not making an election between damages and profits until the commencement of the presentation of oral arguments.

Thus in different circumstances the discount for resisting bifurcation might have been greater.

Mere Verification and Anticipation

AstraZeneca Canada Inc v Pharmascience Inc / esomeprazole (NOC) 2012 FC 1189 O’Keefe J

            2,290,531 NEXIUM

As noted in my first post on this decision, here, the patent at issue claims an enteric coated pharmaceutical formulation of omeprazole using hydroxypropyl methylcellulose (HPMC) with a specified cloud point (CP) as a separating layer between the enteric coating and the active ingredient. The patent asserts that it has surprisingly been found that there is batch variation in the CP of HPMC which affects the rate of release of omeprazole. Pharmascience argued that the patent in question was not an invention on the basis that it merely ascertained the properties of a known substance [173ff]. O’Keefe J dismissed this argument, I believe correctly. However, I do have one concern.

O’Keefe J described the issue as “Is the ‘531 Patent not an invention as defined in section 2 of the Patent Act because it merely ascertains the properties of a known substance?” [92]. I suggest that this is not the best characterization of this issue. In some sense it is true that an invention that does not meet the requirements of patentability as set out in the Act is not an “invention” as that term is used in the Act, but now it is more usual to associate a particular attack with a particular established doctrine and its corresponding section. That is, we no longer say that an obvious invention “lacks subject-matter,” but rather that it is invalid for obviousness under s 28.3. Section 2 is associated with the attack that the invention is not patentable subject matter, as being, for example, a higher life form. I suggest that the argument that the invention merely ascertains the properties of a known substance is more properly considered an attack based on anticipation: see eg see King Pharmaceuticals, Inc v Eon Labs, Inc, 616 F3d 1267 (Fed Cir 2010), cited with approval on this point in Janssen Inc v Mylan Pharmaceuticals ULC / galantamine (NOC), 2010 FC 1123 at [54], treating this issue as being one of anticipation.

Double Ramp-Up in NOC Section 8 Damages

Apotex Inc v Merck Canada Inc / alendronate (NOC) 2012 FC 1235 Hughes J

Under s 8 of the NOC Regulations, if a patentee is unsuccessful in NOC proceedings, a generic which has been kept out of the market by the statutory stay is entitled to damages. While a number of cases have addressed preliminary questions regarding section 8, this is only the third decision quantifying those damages, after the ramipril companion cases decided by Snider J in Teva v Sanofi-Aventis / ramipril 2012 FC 552 and Apotex Inc v Sanofi-Aventis / ramipril 2012 FC 553 (Blogged here, here and here). Hughes J adopted (at [10]), the same approach used by Snider J in the ramipril cases, which compares the “but for” world which would have been if the generic had not been prohibited from entry by the statutory stay, with the real world, with one important caveat. In Merck Frosst Canada Ltd v Apotex Inc / alendronate (NOC) 2009 FCA 187 var’g 2008 FC 1185 the FCA held that the generic is entitled only to losses incurred during the statutory period. Apotex had argued that it was entitled to compensation for loss of permanent market share - springboard damages - but the FCA held that losses incurred after the statutory period are not compensable even if they were caused by the statutory stay.

In many respects the quantification exercise was remarkably straightforward. Hughes J provided a clear and logical structure of issues to be addressed such as the relevant period, the overall market size, generic share and so on [10]. He also identified four key factual determinations relating to when a generic would have entered the marketplace, such as when the generic would have had manufacturing capacity [44]. The parties agreed on many key factual issues, both at a broad level (for example, they agreed that “the [generics'] share in the ‘but for’ world, is the same as the share occupied by the generics in the subsequent ‘real world’ period” [42]) and at a more detailed level (agreement that the generic price would generally be set at 70% of brand price with only one generic in the market, and 63% with multiple generics [75-77]). There were of course a few factual disputes resolved by Hughes J, but on the whole this decision gives some reason for hope that quantification of s 8 damages may not always be a difficult exercise once the basic legal principles are clarified.

There was one contentious point of legal principle, related to the issue of “double ramp-up” [81-87]. “When an organization introduces a new drug product into the marketplace, there is an initial period during which the product has to be made or acquired, orders received from customers, and the product is to be shipped to customers” [81]. This is the ramp-up. The difference between the profits during the ramp-up period and the profits during the same time in the steady state is the “ramp-up loss.” The experts for the parties agreed that to determine Apotex’s profits in the “but for” world, this ramp-up loss must be taken into account, so that Apotex’s profits would be somewhat lower than if it had been able to sell at full capacity for the entire period [82]. The experts for the parties also agreed that as a matter of accounting practice and causation, Apotex should not suffer a double ramp-up loss [82]. Apotex actually suffered a ramp-up loss in the real world, and if a ramp-up loss is also deducted in constructing the “but for” world, then Apotex will face a double ramp-up loss, even though in either the real world or the "but for" world, the loss would have been suffered only once, albeit at different times. 

However, as noted above, the FCA has held that losses incurred outside the statutory period are not compensable, even if they were caused by the stay. Consequently, in Apotex Inc v Sanofi-Aventis / ramipril 2012 FC 553 [265-270] Snider J refused to allow Apotex to deduct the ramp-up loss that occurred in the real world, holding that this was precluded by the FCA decision. Hughes J disagreed with Snider J on this point:

[86] I am not satisfied, particularly given the common view of the accounting experts that, normally, compensation would be made to prevent a double ramp-up loss, that the Court of Appeal had this situation in mind.

Nonetheless, citing “the interests of comity and in the expectation of an inevitable appeal” he followed Snider J and he refused to allow compensation for the double ramp-up.

I expect that Hughes J is right to say that the FCA did not have this situation in mind, although Snider J’s holding that the principle enunciated by the FCA nonetheless covers the ramp-up issue has considerable force. It would have been instructive, and no doubt helpful for the FCA, to have had Hughes J’s analysis of the issue. I am not sure that the principle of comity required him to follow Snider J. The functional justification for following prior decisions of the same level of court is that once enough cases have consistently addressed a point of law it should be considered settled; legal certainty should not require an appellate decision. But a novel point of law should not be considered settled by a single decision. It would be perverse if the first judge to address a legal point decided one way, and five subsequent judges all disagreed, but nonetheless followed the first judge, simply because he was first.

On the whole, this decision suggests that many of the factual points arising from s 8 litigation may be relatively easy to settle, but the legal points will be settled narrowly and so will require many trips to the FCA.

Direct Evidence of Lack of Utility Is Useful

AstraZeneca Canada Inc v Pharmascience Inc / esomeprazole (NOC) 2012 FC 1189 O’Keefe J

            2,290,531 NEXIUM

This decision turned entirely on the evidence regarding utility, and no new points of law appear to be raised. The patent at issue claims an enteric coated pharmaceutical formulation of omeprazole using hydroxypropyl methylcellulose (HPMC) with a specified cloud point (CP) as a separating layer between the enteric coating and the active ingredient. The patent asserts that it has “surprisingly” been found that the CP of HPMC affects the rate of release of omeprazole. Pharmascience’s evidence was that not only was this surprising, it was incredible and false. The expert evidence for Pharmascience was to the effect that the HPMC was merely a separating layer and “could not function to control the rate of drug release” [40].

The case therefore turned primarily on the evidence on the issue of utility. O’Keefe J held that Pharmascience had met its evidentiary burden and put this allegation into play, and the burden therefore lay with AstraZeneca to show on the balance of probabilities that the patented invention was useful [172]. On the facts, O’Keefe J held that the experiments disclosed in the patent were sufficient to carry this burden [191]. Ultimately, Pharmascience’s case failed because AstraZeneca provided direct evidence that the invention worked as claimed, while Pharmascience did not provide any direct evidence to the contrary. There was considerable evidence from Pharmascience’s experts that it was very unlikely that the CP should affect the rate of drug release, but that only goes to show how surprising it is if the patent’s claim is true. Pharmascience did not provide evidence of direct tests of rate of release of formulations with the different CPs, as disclosed in the patent [193]. In the absence of evidence that the invention actually did not work, as opposed to evidence that it was unlikely to work, the patentee’s evidence that it did work carried the day.

Pharmascience also argued that the patent in question was not an invention on the basis that it merely ascertained the properties of a known substance [173ff]. O’Keefe J dismissed this argument. The patent did not merely measure the CP, but also disclosed how to use that knowledge to manufacture a superior product [176]. (Selecting HPMC with the disclosed CP would reduce product discard [97].)

On the fact O’Keefe J held that disclosure was adequate. This was purely a matter of the skilled person’s understanding of the disclosure, and no novel points of law were raised.

Finally, O’Keefe J dismissed Pharmascience’s argument that the invention was obvious. This follows directly from his holding on the evidence that the CP does affect rate of release, as Pharamscience’s own evidence established how surprising this was.

What Is a “Formulation”?

Gilead Sciences Canada, Inc v Canada (Minister of Health) / COMPLERA 2012 FCA 254 Trudel JA: Sharlow, Mainville JJA, aff’g 2012 FC 2 Mosley J (blogged here)

            2,512,475

A question left aside in last week’s week post on this decision was whether the patent in question claimed a medicinal ingredient, and so would be eligible for listing under NOC Reg s 4(2)(a), or a formulation, eligible under 4(2)(b).

The ‘475 patent at issue includes a claim to two medicinal ingredients, tenofovir and emtricitabine, in combination with a third medicinal ingredient from the class of non-nucleoside reverse transcriptase inhibitors [claim 34; FC 24]. A “claim for the formulation” is defined (emphasis added) under the NOC Regs s 2 as

"claim for the formulation" means a claim for a substance that is a mixture of medicinal and non-medicinal ingredients in a drug and that is administered to a patient in a particular dosage form;

As the FCA emphasized, this definition requires that a formulation contain non-medicinal ingredient [27]. The relevant claims do not. Consequently, the FCA held that the eligibility of this claim turned on para 4(2)(a), not (b)[32].

This is a straightforward conclusion on the text of the provision. Why then did the Minister argue, in a point accepted by Mosley J, that the listing should be considered under 4(2)(b) [FC 37]? I think the key to the argument is captured in a statement by Russel J in Bayer Inc v Canada (Minister of Health), 2009 FC 1171 [80], aff’d 2010 FCA 161, quoted by Mosley J at [FC 41]:

The essence of a compound patent is the medicinal ingredient; the essence of a formulation patent is the mixture of ingredients.

In this case, the inventive concept was the discovery that tenofovir and emtricitabine are chemically stable when combined with a third anti-viral of the claimed class [FC 25]. The essence of the invention was the mixture of ingredients, not the discovery of any new medicine. Thus a purposive perspective suggests that the Minister and Mosley J were correct in their interpretation. But text, context and purpose must all be read together, and “[w]hen the words of a provision are precise and unequivocal, the ordinary meaning of the words play a dominant role in the interpretive process” Canada Trustco 2005 SCC 54 [10]. In this case, my view is that the text was sufficiently unequivocal that it should dominate, so that the FCA was correct in its interpretation. However, this case does illustrate a shortcoming in the drafting of 4(2)(b), in that the text is inconsistent with the apparent purpose. In this case no harm was done, as the FCA held that the claim was eligible under 4(2)(a). This implies that a mixture of medicinal ingredients is eligible under 4(2)(a) even when the mixture is desirable for reasons of formulation, and not because of a new medicinal effect. That conclusion is perhaps problematic from a purposive perspective, but it does solve the problem of this type of claim falling between the cracks of 4(2)(a) and (b).

Remand on Infringement Includes Factual Findings

Corlac Inc v Weatherford Canada Ltd 2012 FCA 261 Dawson JA: Nadon, Stratas JJA remanding 2012 FC 76 Phelan J

This decision is a second attempt by the FCA to have Phelan J clarify his reasoning on one issue in the Corlac v Weatherford litigation. The patent at issue relates to a seal assembly for use in oil well pumps. In Weatherford v Corlac 2010 FC 602, Phelan J held claims to the assembly itself to be infringed by Corlac, which sold seal assemblies. Claim 17 was to a method of detecting and restraining oil leaks using that assembly. While Corlac did not practice the infringing method, Corlac sold the assemblies with instruction manuals which, according to Weatherford’s expert, taught the practice described in claim 17 [175]. Phelan J held claim 17 to be infringed, but his reasoning on this point was brief and not entirely clear: see [199]. On appeal, 2011 FCA 228, a decision known primarily for its treatment of fraud on the patent office, the FCA affirmed Phelan J except in relation to claim 17. The FCA held that “the judge's reasons do not provide for meaningful appellate review” and consequently, “I would return the issue of infringement of claim 17, that is inducement, to the judge to be determined in accordance with the established test” [171, emphasis added].

On remand, in 2012 FC 76, (see here) Phelan J addressed each of the legal elements of the three-part test for inducement. However, Phelan J did not revisit a crucial factual issue, holding that it was rea judicata [6]. On appeal of that decision, in 2012 FCA 261, the FCA held that it had remanded the entire "issue of infringement of claim 17" to Phelan J, and "the issue" includes the factual issue. The FCA also pointed out that in his original reasons, Phelan J had never explicitly accepted the evidence of Weatherford’s expert: “At best, such a finding is implicit in the Judge’s finding of infringement” [36]. The FCA also pointed out that in its original decision it had never accepted this implicit factual finding; it merely pointed out how Phelan J had determined the point [33].

The FCA’s reasoning is logical. With that said, I understand why Phelan J did not address the factual issue in his original decision. While the FCA did explicitly remand “the issue of infringement," the FCA focused primarily on the legal test, and, as the FCA pointed out, the factual finding was implicit in Phelan J’s original decision. And since that factual finding was indeed implicit in his original determination, it would be surprising if his determination were to change now that he will be required to make it explicitly.

What Are Normal Costs in Hard Fought Litigation?

Fournier Pharma Inc v Sandoz Canada Inc / fenofibrate (Costs) 2012 FC 1121 Zinn J

A point of general interest in Zinn J’s costs decision from the fenofibrate litigation (blogged here and here) is his remark at [6] (emphasis added) that

Typically costs are assessed at the mid-point of Column III of the Tariff; however, numerous judgments of this Court have assessed costs in pharmaceutical litigation at higher levels. Given the complexity of this litigation, I find that the costs are to be assessed at the higher end of Column III.

Contrast this will Martineau J’s statement in his costs decision in Eurocopter v Bell at [22] that “an assessment of costs in accordance with the higher unit of column IV seem to be commonly maintained in [] lengthy and hard-fought patent litigations” (emphasis added). Eurocopter was an infringement action while the fenofibrate case was an NOC proceeding, but this does not seem to be a distinguishing factor, as Zinn J referred simply to “pharmaceutical litigation.”

Extreme Application of Product Specificity Requirement Affirmed

Gilead Sciences Canada, Inc v Canada (Minister of Health) / COMPLERA 2012 FCA 254 Trudel JA: Sharlow, Mainville JJA, aff’g 2012 FC 2 Mosley J (blogged here)

In this decision the FCA has affirmed a very strict application of the already strict product specificity requirement for listing a patent against a drug on the Patent Register that it set out in Purdue Pharma / TARGIN 2011 FCA 132 (blogged here). In my view, the requirement as set out by the FCA is overly strict and is not consistent with a purposive interpretation of the relevant provisions of the NOC Regulations.

Is the Test for Inducement “Strict”?

Weatherford Canada Ltd. v. Corlac Inc. 2012 FC 76 Phelan J (on remand from 2011 FCA 228)

The patent at issue relates to a seal assembly for use in oil well pumps. In Weatherford Canada Ltd. v. Corlac Inc. 2010 FC 602 varied 2011 FCA 228, Phelan J held a number of claims to be valid and infringed by Corlac. All of the claims in issue except one, claim 17, were to the assembly itself. Claim 17 was a claim to a method of detecting and restraining oil leaks using that assembly. On appeal, the FCA affirmed Phelan J except in relation to claim 17. The FCA at [162] re-affirmed the accepted three-part test for inducement:

            First, the act of infringement must have been completed by the direct infringer.

Second, the completion of the acts of infringement must be influenced by the acts of the alleged inducer to the point that, without the influence, direct infringement would not take place.

Third, the influence must knowingly be exercised by the inducer, that is, the inducer knows that this influence will result in the completion of the act of infringement

Layden-Stevenson JA, for the court, held that Phelan J had not explicitly applied this test, and “[i]n view of the stringent test for inducement, I cannot determine whether the judge, had he explicitly applied the test, would have arrived at a conclusion that the appellants indirectly infringed claim 17" [168]. The FCA therefore remanded the case to Phelan J for re-determination in accordance with the accepted test [171].

Perhaps the most interesting part of Phelan J’s decision concerns the standard of proof. Part of Corlac's complaint on appeal was that there was no direct evidence of direct infringement by end-users. Though the FCA did not accept that such evidence was necessary, it did hold that “[t]here seem to be significant gaps in the evidentiary basis needed to support what appears to be an implicit finding of third party use” [169]. Thus the standard of proof was a significant issue. Further, as Phelan J noted at [10], the FCA variously described the test as a “difficult one to meet”, “a stringent test”, and a “strict” test. But, as Phelan J pointed out at [11], “[t]he Supreme Court has confirmed that there are only two evidentiary standards – ‘balance of probabilities’ and ‘beyond reasonable doubt’ (see F.H. v McDougall, 2008 SCC 53),” and the civil standard is proof on the balance of probabilities, or “more likely than not.”

Having established that the standard of proof is the balance of probabilities, Phelan J then reviewed the evidence, explicitly applied the three-prong test for inducement and held, on that standard, that it had been satisfied, in effect because Corlac has sold the assembly with instructions directing the end-users to use it in an infringing manner. Phelan J was willing to accept, on the evidence that customers were likely to use the assembly according to the instructions and that Corlac would have intended that they do so. Consequently, Corlac was held liable for inducing the infringement of claim 17 [24].

So what does it mean to say that a particular test is “stringent”? As Phelan J noted, it cannot mean any change in the standard of proof. It strikes me that in saying the test was “stringent” or “strict” the FCA may simply have been saying that all three prongs must be met. In saying the test is a difficult one to meet, the FCA may have meant that because the three separate prongs must all be met, it will be generally be difficult for a patentee to establish inducement, even on the balance of probabilities. That is, these remarks were meant as an observation, rather than a proposition of law. But I am not sure it is right to say that the test is particularly difficult to satisfy in practice: as Mandamin J noted in Glaston v Horizon Glass 2010 FC 1191, in a number of cases “[i]nducement has . . . been found where a seller provides a purchaser with instructions or directions for using an infringing method” [89]. This decision adds to that list.

Infringement if Later ≠ Anticipation if Earlier

Bristol-Myers Squibb Co. v. Mylan / efavirenz (NOC) 2012 FC 1142 Barnes J

            2,101,572 – 2,279,198 – efavirenz – SUSTIVA

The traditional rule is that “what would infringe if later, anticipates if earlier”: Abbott Labs v Ratiopharm / clarithromycin (NOC) 2006 FCA 187 [25]; Lightning Fastener Co. v. Colonial Fastener Co., [1933] SCR 377 at 381. This is no longer true, since Sanofi 2008 SCC 61, following Synthon [2005] UKHL 59, held that anticipation requires enabling disclosure. This adage applies only to the disclosure prong of anticipation: Synthon [24]. If the prior use did not disclose the information necessary to enable the invention, it will not anticipate. It will still infringe if practiced later, since infringement does not require knowledge: Merrell Dow v HN Norton [1995] UKHL 14 [47]; Abbott / clarithromycin (NOC) [24]. The consequences is that if the prior use was not enabling, a later patent could confer a right to stop people from doing what they had done before. This is strikingly counter-intuitive: “Ever since the power of the Crown to grant monopolies was curbed by parliament and the courts at the beginning of the seventeenth century, it has been a fundamental principle of United Kingdom patent law that the Crown could not grant a patent which would enable the patentee to stop another trader from doing what he had done before”: Merrell Dow v HN Norton [17]. Nonetheless, that is the direct implication of the rule that anticipation requires enabling disclosure. So far the leading cases Canadian and UK cases have avoided this result on the facts, as in Merrell Dow, but this is not a satisfactory response. (US law has addressed the problem directly with the doctrine of inherent anticipation: “a prior art reference may anticipate without disclosing a feature of the claimed invention if that missing characteristic is necessarily present, or inherent, in the single anticipating reference”Schering Corp v Geneva Pharmaceuticals 339 F.3d 1373, 1377 (Fed Cir 2003).) The problem will arise one day in a context in which it cannot be avoided, and the courts will face a choice between preventing a person from doing what she has been done before, and adhering strictly to the requirement of enabling disclosure.