Monday, July 13, 2015

Non-Infringing Alternative "Defence" Rejected Once Again

ADIR v Apotex 2015 FC 721 Gagné J
            1,341,196 / perindopril / COVERSYL

As noted in my last post, this case is the accounting of profits portion of a bifurcated trial. In the liability decision, Laboratoires Servier v Apotex Inc, 2008 FC 825 aff’d 2009 FCA 222, Snider J held that the defendants had infringed the plaintiff’s ‘196 patent and she allowed the plaintiffs to elect an accounting of profits [1]. This post deals with the treatment of non-infringing alternatives in the accounting.

A substantial amount of the infringing material manufactured by Apotex in Canada was destined for export to markets such as the UK and Australia. Apotex argued that it could have manufactured perindopril API and finished dosage in other non-infringing jurisdictions for export to those same markets at exactly the same price, and if it had done so, it would have made exactly the same profits (or even more!) than it made by manufacturing in Canada [79]-[80]. In Schmeiser 2004 SCC 34, the SCC held that the preferred means of calculating an accounting is the differential profit approach, in which “A comparison is to be made between the defendant's profit attributable to the invention and his profit had he used the best non-infringing option” [102]. On this authority, Apotex argued that since it would have made just as much money using the non-infringing offshore manufacturing, the profits attributable to its Canadian infringement were zero. This is referred in the decision to as the “non-infringing alternative” or “NIA” “defence”. (As I have argued, this argument is not a defence, in the sense that it does not normally result in a complete bar to liability.)

Gagné J rejected the NIA argument as a matter of law [106], though she went on to hold that in any event it was not made out on the facts [142]. In my view, Gagné J’s reasons for rejecting the NIA analysis as a matter of law is not persuasive, but her analysis of its application on the facts is.

Gagné J carried out an extensive review of the case law and concluded at [119] that:

courts have developed a variety of different formulas or used different terminologies, depending on the facts of each case, to simply decipher the use of the invention by the infringer and the extent in which this use contributed to the infringer's gross revenues or profits. “Segregation" (as it was employed by the defendants in this case to discuss the indemnity and legal services provided for in the transfer price agreements), "Apportionment" and the “Differential profit approach" are all reformulations of that same notion and they are concepts that attempt to capture causation.

It is true that the courts have used a variety of methods for assessing causation, but that does not mean that all the approaches which have been historically used are equally good. Lubrizol [1997] 2 FC 3, discussed by Gagné J at [92], provides a good example. (Note that Gagné J at [28] cited Lubrizol [1996] 3 FC 40, but she evidently intended the 1997 decision.) The claim in question was for the motor oil containing a dispersant additive—not just for the additive itself—so the whole product infringed the claim. The FCA held that the causation requirement nonetheless required an apportionment, because it was possible that the infringing oil achieved its market share for reasons other than the presence of the additive. Apportionment in this sense was never a very precise matter. It was simply a recognition that in some cases the entire profit from the sale of a patented item could not be attributed to the infringement, so that the overall profits had to be somehow apportioned between the infringing and non-infringing parts. The differential profit analysis gets at exactly the same point, but with an analysis that is more precise, at least conceptually. It says that the profit that is properly attributable to the infringement is the difference between the profit the infringer did make and the profit it would have made had it not infringed. For example, in Lubrizol, if it could be shown that a few customers would not have bought the oil at all but for the additive, then the entire profits on those sales would be attributable to the infringement. Alternatively, if it could be shown that the infringer was able to price its oil slightly higher because of the additive, the increased profit due to the increased price would be attributable to the infringement. Or, I submit, if it could be shown that no customer would purchase motor oil without a dispersant, but there was an alternative unpatented dispersant that the infringer could have used that would have been slightly more expensive, then the increased profit due to the cost saving could be attributed to the infringement.

With that said, the evidence will not always lend itself to a differential profit analysis. When the evidence establishes that the infringer’s profits were not all caused by the infringement, yet the evidence is not adequate for a differential profit analysis, it may nonetheless be appropriate to apportion the profits in a less precise manner. This is why, in my view, the SCC in Schmeiser referred to the differential profit approach as the “preferred” method, rather than saying it was the required method.

The difficulty with Gagné J’s analysis comes in the next sentence [119]:

Tracing causation is a factual endeavour. In some cases, it could almost be as complex as the invention and it will require factual or expert evidence. In other cases, such as the one before me, there is no need for a very sophisticated analysis of the causal relationship between the infringement and the infringer’s profits as the defendants merely sold perindopril, the compound covered by the 196 Patent.

On that basis she held that the non-infringing alternative did not have to be considered, as a matter of law, and moreover (implicitly) that no apportionment was necessary. It is true that tracing causation is a factual matter, but the test to be applied is a matter of law. It is clearly not true that a court has the discretion to assess causation in any way it sees fit: see eg Clements 2012 SCC 32 holding it was an error of law to apply material contribution on the facts [50], and Athey [1996] 3 SCR 458 [42] holding it would be an error of law to apportion causation. Whether the NIA analysis, or some other form of apportionment, is appropriate in this case is a matter of law, just as it was a matter of law in Athey as to whether the material contribution test should be applied. 

Moreover, Gagné J’s statement that “there is no need for a very sophisticated analysis of the causal relationship between the infringement and the infringer’s profits as the defendants merely sold perindopril, the compound covered by the 196 Patent,” is not at all persuasive. Exactly the same could be said on the facts in Schmeiser, in which it was undisputed that the infringer had made an accounting profit directly from the sale of an infringing product. In my view Gagné J's refusal to consider the NIA analysis as a matter of law, simply because the entire product was patented, is directly inconsistent with Schmeiser.

My sense is that Gagné J’s real reason for rejecting the NIA analysis is expressed in this paragraph [121]:

The defendants have asked me to deduct from their profits to be disgorged the profits that they would have made had they manufactured all of the perindopril API and finished tablet dosage that they export during the relevant period from abroad. To accept the defendants' position would be to provide them a perfect shelter against the consequences of any future patent infringement in Canada. Now that the Apotex group of companies own important manufacturing facilities in India and in the Netherlands, along with a [redacted] interest in manufacturing facilities in Mexico, they would never have to disgorge any profits from infringing a Canadian patent, as they would only have to prove that their affiliates had the capacity to manufacture abroad, and that they would have been willing to do so in an effort to advance a NIA defence. Therefore, profits from manufacturing abroad would most likely exceed profits from manufacturing in Canada in any given scenario. These scenarios are exactly the “catch me if you can" situation contemplated by Zinn J in Rivett FC.

The objection is that the NIA “defence” would allow the infringer to get away with infringement scot-free. Again, the NIA analysis is not a defence. Most of the time, it simply recognizes that only a part of the infringer’s entire profits on the sale of the patented invention are caused by the infringement, so that the profits to be disgorged are less than the full profits on the infringer’s sales. So, in Lubrizol, if the profits on a case of oil were $1, it would not seem shocking if the amount attributable to the additive were 10¢, or even less. I suggested above that if in Lubrizol it could be established that the infringer could have used a different unpatented additive that cost slightly more than the patented additive, it would only be required to disgorge the increased profit due to the cost saving. And what if the unpatented additive cost exactly the same as the patented additive? Why should the liability not be reduced to zero? Imagine that a small start-up company is developing the core technology for a new product. Two different in-house teams develop two different technologies, A and B, which are functionally equivalent. The company chooses A, essentially arbitrarily. After the company has built itself into a national force, with $100m in annual profits, a patent troll emerges and sues for infringement based on an obscure patent that the start-up was not aware of when it chose its technology. Why should technology company be forced to disgorge the entire $100m in profits? Why is it not be a good answer to say, “the patented technology was not the cause of our success, as we could equally well have used technology B?” As the FCA said in Lubrizol:

The remedy of an account of profits is an equitable one. Its purpose is not to punish the defendant but simply to have him surrender the actual profits he has made at the plaintiff's expense. But if some part of Imperial's profit on the infringing sales can be shown to have been due not to the appropriation of the Lubrizol invention but to some other factor where is the equity?

In my view, the NIA analysis is sound in equity and in economic logic so long as infringer can prove that it actually had a non-infringing alternative available to it. This is a crucial point. The NIA analysis principle is intuitively sound when it is properly applicable on the facts, but it can easily be made to seem absurd by applying it when the facts do not warrant. So, Gagné J criticized the NIA analysis on the basis that it would provide “a perfect shelter against the consequences of any future patent infringement in Canada” because “they would only have to prove that their affiliates had the capacity to manufacture abroad, and that they would have been willing to do so in an effort to advance a NIA defence.” This assumes that establishing the NIA “defence” is an easy matter – and this is belied by Gagné J’s own analysis of the facts.

Even though she rejected the NIA analysis as a matter of law, Gagné J nonetheless went on to consider it on the facts [127]. She ultimately held that Apotex had not established that an NIA was available to it [142]. Apotex had argued that perindopril was readily available on the international market at the relevant time [135]. Indeed, Apotex argued that it could actually have made more profits by manufacturing abroad [132]. But if this was so, Gagné J asked, “One wonders why then the defendants chose to manufacture in Canada where the plaintiffs had an unexpired patent” [136]. This is an excellent question. In any case in which the defendant infringed the patent, there should be a presumption that the infringer gained some advantage from doing so. Consequently, the burden should be on the defendant to establish that it had a non-infringing alternative actually available to it. And indeed, that is where the burden lies: see Jay-Lor 2007 FC 358, [196]; Grain Processing 185 F.3d 1341, 1349 (Fed. Cir. 1999). As Gagné J noted, the SCC in Schmeiser did not suggest “that in an accounting of profits courts are bound to always consider NIA products, options or scenarios, as fanciful as they may be” [118]. That is quite right. But that is a matter of evidence, not a reason for rejecting the NIA analysis altogether. In this case, on the evidence, Gagné J held that Apotex did not have a non-infringing alternative available to it. Consequently, it is wrong to suppose that an assertion of offshore manufacturing capacity will always allow the infringer to escape liability. 

No comments:

Post a Comment