Obviousness is Objective

E Mishan & Sons, Inc v Supertek Canada Inc 2014 FC 326 Hughes J
            2,779,882

Mishan v Supertek is a straightforward case which illustrates the objective nature of the obviousness inquiry [128]. The inventor believed he had invented an ingenious new type of garden hose, but he was not aware of a key prior art patent which rendered the invention obvious.

The problem addressed by the 882 patent was that of unwieldy garden hoses, which are difficult to store and unravel because of their length. The solution provided was a hose consisting of a elastic inner tube contained within an outer fabric tube, with the tubes coupled to each other at each end. Under pressure, the inner tube would elongate until constrained by the outer tube. When turned off, the inner elastic tube would contract, causing the hose to shrink. The inventor had developed the hose primarily for use as a residential garden hose, but the specification stated that it could be used to convey “fluid” including water, gases and even flowable solids [138]. The prior art was US patent 6,523,539, the McDonald patent, which was titled “Self-elongating oxygen hose for stowable aviation crew oxygen mask.” It operated in essentially the same way as the plaintiff’s hose. The main difference between the McDonald invention and the 882 patent was that the former was intended for air supply in aircraft and for conducting oxygen or air. Given the minor difference, and the fact that the 882 patent stated that it could be used for fluids generally, it was relatively straightforward for Hughes J to conclude that the 882 patent was obvious over McDonald.

 McDonald patent

882 patent

Consequently, there was an important question as to whether the McDonald patent was part of relevant state of the art. On the facts, the answer was straightforward: “McDonald was not only findable but found by those interested in expandable hoses. There is no evidence to the contrary” [91].

Commercial success and motivation were argued as showing invention:

[144] There are a number of secondary factors that have been raised. There was motivation to create a simple, inexpensive garden hose that could be promoted in the direct retail market by television advertising and the like. It was a commercial success. But motivation and success alone do not mean that there was, in the objective sense, an invention. Khubani testified to that when he referred to items such as amber sunglasses and dust mops that had been available for years but were great successes in the direct retail environment.

The inventor of the 882 patent did not begin his research by doing a search of the patent prior art, and apparently the same is true of garden hose designers in general. If they had, the patentee’s solution would have been arrived at sooner. The notional skilled person is probably not as clever as real-world inventors, but she may sometimes be better informed about the prior art.

Note that the costs awarded to the defendant were reduced because it had raised too many issues on which they were unsuccessful. This is a message from Hughes J that the case should be restricted to the real issues [162].

A Predictable and Consistent Framework for the Assessment of S 8 Damages

Teva Canada Ltd v Pfizer Canada Inc 2014 FC 248 Zinn J
            1,248,540 / 2,199,778 / venlafaxine / EFFEXOR XR

While it is still early days for s 8 damages claims, the Teva Venlafaxine s 8 decision turned largely on factual issues, rather than on novel questions of law. This relative clarity in the law seems to be because the courts have adopted a but-for approach to determination of the loss which is consistent with the general damages principles, rather than creating a new jurisprudence specific to s 8. Consistently with this observation, the most contentious legal issues have related to the start and end date for the compensable period, which are governed directly by the statute, and with respect to double ramp-up, which turns on a statutorily based exception to the general principles of causation. Additionally, Zinn J’s holding on the date from which interest runs raised an interesting point of law relating to s 8 damages generally. With respect to the factual issues, Zinn J expressed a general preference for the so-called “analogue approach” to constructing the but-for world, not as a matter of principle, but because he felt that it will often be more closely tailored to the specifics of the product in question. If this view is shared more broadly in the FC, it will bring further predictability to the assessment of s 8 damages.

In Teva s8 venlafaxine, Zinn J summarized the general framework for assessing damages under s 8 of the PM(NOC) Regulations as follows [27]:

1. Determine the period of liability [the Relevant Period];

2. Determine the overall size of the market for the relevant pharmaceutical [the Relevant Pharmaceutical Market] during the Relevant Period;

3. Determine the portion of the Relevant Pharmaceutical Market that would have been held by generic manufacturers during the Relevant Period [The Generic Market];

4. Determine the portion of the Generic Market that would have been held by the plaintiff [the Plaintiff’s Lost Volume]; and

5. Quantify the damages that would have been suffered by the plaintiff in respect of the Plaintiff’s Lost Volume [the Plaintiff’s Net Lost Profit].

Yesterday’s post discussed the first issue. This post discusses the remaining issues. Recall that, as mentioned in yesterday’s post, Pfizer is the corporate successor to Wyeth, and in 2010, Novopharm changed its name to Teva, and Teva and Ratiopharm then amalgamated. Because the events in issue took place in the 2005 to 2007 period, Zinn J referred to the relevant pharmaceutical companies as Ratiopharm, Novopharm and Wyeth [3].

The Relevant Pharmaceutical Market: What is the size of the overall Venlafaxine Market?
The main issue relating to the size of the overall venlafaxine market was the factual question of whether Wyeth’s withdrawal of promotional efforts on generic entry would have resulted in a drop in demand.

Teva’s expert, Dr Hollis, used an “analogue approach” to answering this question. In effect, he started with the presumption that sales in the but-for world would have been the same as sales in the real world when the generic did actually enter, and then he asked whether there was any reason to believe that the but-for world would have been different [67]. Pfizer’s expert, Dr Tepperman, used a regression analysis [71].

One point of general interest is that “Both experts agreed that where there is a good analogue, that approach is preferred over a regression analysis” [80], and Zinn J accepted this, saying:

[80] In my view, Dr. Hollis’ analogue approach is an adequate and appropriate model for three reasons. First, it accounts for any trends that might be unique to Venlafaxine. Second, the difference in the actual time period between the real world and the but-for world is relatively insignificant as generic entry in the real world occurred on December 1, 2006 and on January 10, 2006, in the but-for world. Third, the market dynamics in the real world closely mirror those in the but-for world. Moreover, I agree with the observation of Justice Phelan in Pantoprazole FC 2013 at para 21 that quantification of damages in the but-for world should be grounded in the experience of the real world, and using an analogue approach where an adequate analogue is available is consistent with that approach.

This clearly indicates that an analogue approach is generally preferable. Note, however, that if there were a longer time between entry in the real world and in the but-for world, the analogue approach would be less appealing.

In the result Zinn J preferred Dr Hollis’ approach. While Zinn J did not say that a regression analysis can never be useful, he did reject Dr. Tepperman’s analysis because of problem with the specific analysis [79].

The Generic Market: What is the size of the Generic Venlafaxine Market?
Again, this was largely a factual issue. The experts both used the same “analogue approach” [82-83], which uses the actual sales as the basis for sales in the but-for world. The main adjustment related to the date of formulary listing [85].

Plaintiff’s Lost Volume: What is Ratiopharm’s Market Share?

            (a) Other Generic Entry
The analogue approach was also used in assessing Ratiopharm’s share of the generic market, but here a tricky adjustment had to be made because of the possibility of entry by other generics. In particular, Wyeth argued that Novopharm and Pharmascience would also have entered. Zinn J held that the burden of establishing entry by the other generics rests on Wyeth [91].

As discussed here, in Apotex s8 ramipril FCA 2014 FCA 68, the FCA held that in the hypothetical world, as the Regulatory barrier to entry which face generics in the real world, including the NOC Regulations, also affect all generics in the but-for world, with the proviso that the claimant generic would have been able to enter as of the start date of the compensable period. I called this the “minimalist approach,” as it assumes the smallest possible deviation from the real world. This means that in assessing whether other generics would entered, we must consider whether they would they have legally been able to enter, as well as whether they would have been motivated and had the manufacturing capacity to do so.

The answer to the first question was relatively straightforward with respect to Novopharm, which had entered into an authorized generic agreement with Wyeth which allowed Novopharm to obtain an NOC and seek formulary listing as soon as a second generic receives an NOC [95]. Thus there were no regulatory hurdles to Novopharm entering as soon after Ratiopharm as it was able. Novopharm’s manufacturing capacity was the real issue. There was evidence that Teva Israel, Novopharm’s supplier of venlafaxine, had serious manufacturing difficulties. While there was some suggestion that Novopharm could have entered sooner than it actually did had there been more urgency, in the end Zinn J concluded on the facts that because of the manufacturing problems, it would have entered in the but-for world on the same date as it did in the real world, namely December 1, 2006 [129].

With to respect to Pharmascience, Zinn J held that any generic other than the s 8 claimant must comply with the NOC Regulations even in the but-for world [130]. This is consistent with the FCA’s holding in Apotex s8 ramipril FCA. In view of this, they key question on the facts was whether Pharmascience would even have served an NOA. Zinn J concluded that it would not have done so, essentially because Pharmascience’s strategy was to time its launch to coincide with a positive decision regarding the 778 patent in favour of Ratiopharm, and consequently, it would not have been ready to launch earlier than it actually did [139]. Therefore it would not have served an NOA near the start date of the compensable period in the but-for world, because it would not have been able to launch even if successful.

Consequently Zinn J held that Pharmascience would not have entered the generic market during the Relevant Period [142]. The only two generics would have been Ratiopharm, entering on January 10, 2006, and Novopharm, entering on December 1, 2006 [143].

            (b) Ratiopharm’s Capacity
The next question was whether Ratiopharm would have had the capacity to supply the market. Zinn J held that “the burden of establishing that it could have come to market during the Relevant Period rests with Ratiopharm” [144], and more specifically

[148] To do so, it must both identify an API supplier (since it is the plaintiff generic), and it must show that the API supplier had the capacity to supply the market over the Relevant Period.

On the facts, Zinn J held Ratiopharm had discharged that burden [159].

(c) Formulary Listing

Then the question was when the competing generics would have been listed on the provincial formularies. This was entirely a question of fact [160]-[185]. There are two general points which may be made. First, the formulary analysis illustrates that the s 8 analysis is very fact specific. For example, the precise formulary updating schedule in the different provinces had to be taken into account. Second, in this section (as well as in some other sections), one expert’s evidence was discounted to some extent because of conflicting testimony in a different case [171]. If s 8 cases become common, it will be important for counsel to take the long view of an expert’s evidence, if that expert may be retained in subsequent cases.

Plaintiff’s Net Lost Profit: The Value of Ratiopharm’s Lost Sales

(a) Listing Price

Again, this discussion is very fact intensive. One point of general interest is that in Québec, generic manufacturers had to commit that they would match the best price in any other province in Canada as soon as that price was available [200]. However, in the real world, Novopharm did not drop its price in Québec until four months after the price dropped in Ontario. Zinn J held that:

[201] Although there is no explanation for why in the real world, Novopharm failed to honour its commitment to lower the price in Québec until four months after the price drop in Ontario, I am not prepared to accept that Ratiopharm would have disregarded its obligation, as Novopharm apparently did. It must be assumed in the but-for world that generic manufacturers adhere to their commitments and legal obligations, unless there is convincing evidence that the particular generic does not do so on a regular and consistent basis. There was no such evidence in respect of Ratiopharm’s practices.

This holding is a partial departure from the general rule that the but-for world is constructed purely as a factual inquiry into what actually would have happened, without presumptions or modifications based on rules of law.

(c) Trade-spend

The discussion of trade-spend – the rebates provided by pharmaceutical companies to purchasers – turned entirely on the facts [207] - [232].

Deductions for Regulatory Non-Compliance
Wyeth alleged that “Ratiopharm launched its product without having fully completed its validation and thus Ratiopharm contravened the F&D Regulations. Essentially, Wyeth's argument is grounded in equity: Because Ratiopharm did not comply with all legal requirements at launch, it should be precluded from recovering damages under section 8" [236]. Zinn J held that Wyeth had not established that Ratiopharm breached the F&D Regulations [238], and he therefore never reached the question of whether such a contravention, if established, would be a reason to preclude or reduce recovery of s 8 damages.

Ramp-Up [240]-[254]
In Teva s 8 ramipril FCA, the FCA held that no adjustment should be made to allow for double ramp-up, as discussed here. Zinn J came to the same conclusion, and his decision is therefore consistent with the law as stated by the FCA on this point.

Interest
Prejudgment interest is calculated “from the date the cause of action arose to the date of the order” [255]. The question was whether that date is August 1, 2007, when the Court of Appeal dismissed the Prohibition Application, on the view that Ratiopharm had no right to sue for s 8 damages before that time; or January 10, 2006, the start date of the compensable period, which is when Ratiopharm actually began to suffer a loss. Zinn J held that the latter date was appropriate:

[258] The disposition of a Prohibition Application does not ground liability, it simply confirms that liability exists. The cause of action arises on the date that damages that are the basis for the claim begin to be suffered. Typically, this will coincide with when the Relevant Period begins, as it did in Pantoprazole FC 2012 and as it does in this case. However, because the Relevant Period may begin before damage is actually suffered, this need not always be the case. For that reason, prejudgment interest must be tied to when the loss actually begins to be suffered irrespective of whether that date is the same as the start of the Relevant Period.

This is consistent with Zinn J’s general emphasis on the importance of causation, which was noted in yesterday’s post. This works both ways. On the one hand, the patentee will not be liable for losses not caused by the statutory delay, but on the other hand, it is presumptively liable for all losses that were caused by the statutory delay.

Causation of Loss in Determination of Compensable Period for s 8 Damages

Teva Canada Ltd v Pfizer Canada Inc 2014 FC 248 Zinn J
            1,248,540 / 2,199,778 / venlafaxine / EFFEXOR XR

Under s 8 of the PM(NOC) Regulations, a generic that successfully contests an order for prohibition is entitled to recover its losses due to having been improperly kept out of the market by the statutory stay which is triggered by the patentee’s application for the prohibition order. Section 8 cases now seem to be arriving in a flood. The leading cases to date have been Snider J’s decisions in Teva s8 ramipril FC 2012 FC 552, Apotex s8 ramipril FC 2012 FC 553 (here, here and here) and the very recent FCA decisions on appeal in those cases, Teva s 8 ramipril FCA 2014 FCA 67 and Apotex s 8 ramipril FCA 2014 FCA 68 (here, here and here). Zinn J’s decision in Teva s8 venlafaxine is another important contribution to this case law. It was released to the parties on 14 March 2014, the same date as the FCA decisions were released. This means that Zinn J did not have the benefit of the FCA decision in writing his reasons, and one of the questions is the extent to which his reasons are consistent with those of the FCA.

Zinn J summarized the general framework for assessing s 8 damages as follows [27]:

1. Determine the period of liability [the Relevant Period];

2. Determine the overall size of the market for the relevant pharmaceutical [the Relevant Pharmaceutical Market] during the Relevant Period;

3. Determine the portion of the Relevant Pharmaceutical Market that would have been held by generic manufacturers during the Relevant Period [The Generic Market];

4. Determine the portion of the Generic Market that would have been held by the plaintiff [the Plaintiff’s Lost Volume]; and

5. Quantify the damages that would have been suffered by the plaintiff in respect of the Plaintiff’s Lost Volume [the Plaintiff’s Net Lost Profit].

This post discusses the first issue. Establishing the relevant period turns on two distinct subsidiary questions, namely the start date and the end date. Note that Pfizer is the corporate successor to Wyeth, and in 2010, Novopharm changed its name to Teva, and Teva and Ratiopharm amalgamated. Because the events in issue took place in the 2005 to 2007 period, to avoid confusion, Zinn J referred to the relevant pharmaceutical companies as Ratiopharm, Novopharm and Wyeth [3]

Relevant Period: Start Date
S 8(1) provides that the liability presumptively begins on the patent hold date, “(ii) unless the court concludes that a date other than the certified date is more appropriate.” The patent hold date was December 7, 2005 [15], but in its ANDS and NOA challenging the 778 patent, Ratiopharm had agreed to wait for the expiry of the 540 patent, January 10, 2006, before launching. In Teva s 8 ramipril FCA a similar situation had arisen, and the FCA, affirming Snider J on this point, held that the appropriate start date was the expiry of the later patent, not the patent hold date. Though it did not have the benefit of the FCA decision, in this case, Ratiopharm nonetheless accepted that the appropriate start date should be the expiry date of the 540 patent, rather than the patent hold date. Zinn J accepted Ratiopharm’s submission on this point.

Wyeth argued that the start date should be February 13, 2006, “the date the Minister would have issued a NOC to Ratiopharm if it had served Wyeth with a NOA relating to the 778 Patent and Wyeth had not commenced a Prohibition Application within the 45 day period permitted by the PMNOC Regulations” [48]. That is, if I understand correctly, Wyeth’s position is that we should assume that the Regulations were completely effective, and the start date for the relevant period should be determined on the assumption that rather than applying for an order of prohibition on receiving Ratiopharm’s NOA, Wyeth would have done nothing, so that the NOC would have been issued at the end of the 45 period after service of the NOA.

This position was rejected by Zinn J. Before turning to Zinn J’s reasons, I note that his holding appears to be consistent with that of the FCA in Apotex s8 ramipril FCA. The basic difficulty with Wyeth’s position, in my view, is that it assumes the start date is to be determined as a matter of the construction of the “but for” world used to assess the loss suffered by the generic. If that were the case, then the question of exactly how the "but for" world is to be constructed – whether we should assume that the generic has served an NOA and wait out the reply period, or whether we should assume that the Regulations simply don’t apply at all – would be relevant. However, the “but for” world is relevant to the assessment of the loss, not to the determination of the start date, which is governed directly by s 8(1)(a). As Sharlow J noted in Apotex s8 ramipril FCA says that “the NOC Regulations are to be disregarded in determining the beginning of the section 8 liability period, as long as neither of the stated exceptions applies” [170].

In Teva s8 ramipril FC Snider J had held that as a matter of law, "the liability period cannot predate the statutory stay" [60]. The FCA reversed Snider J on this point [79], though, as noted, it affirmed Snider J in holding that the appropriate date was the later expiry of the later patent. In this case Wyeth also argued that the start date cannot predate the statutory stay. Zinn J rejected this position, consistently with the FCA holding [51].

The most interesting part of Zinn J’s reasons on this point was his very strong emphasis on the need for a causal link between failed NOC proceedings and the claimed loss:

the damages are those that the plaintiff generic suffered “by reason of the delayed market entry of its drug” as stated in the Regulatory Impact Analysis Statement [RIAS] [57]

The question for the Court is whether there is a causal connection between the failed PMNOC proceedings and the loss claimed as damages and if so when did that loss first arise. . . . [T]he damages claimed in an action under section 8 must be causally connected to the Prohibition Application” [61].

See similarly [44] in the context of discussing the end date. In Teva s8 ramipril FC Snider J had made a similar causation argument. As discussed here, the FCA indicated that it was not persuased by the causation point, saying that recovery is governed by s 8, which must prevail over general principles [78]. While that is no doubt true, I suggested that general principles, especially fundamental principles such as causation, should apply unless inconsistent with the Regulations, and I noted that the FCA did not say that causation should never be a consideration under s 8. In my view, Zinn J’s emphasis on causation is entirely sound, and is not inconsistent with the FCA decision in Teva s8 ramipril FCA. It will be interesting to see whether the addition of Zinn J’s voice to that of Snider J will persuade the FCA in future cases. My guess is that even if the FCA never holds that causation of loss is a necessary condition for recovery under s 8 as a matter of law, causation will be such a strong factor that it will be a de facto strict requirement.

Zinn J also noted that in unusual circumstances, the start date might be earlier than the patent hold date, so long as the losses arising prior to the patent hold date were caused by the improper delay [60].

It will be not uncommon that a generic will have agreed to wait for the expiry of a particular patent in its NOA, and will have been put on patent hold long before that expiry date. Between the ramipril decisions and this Venlafaxine decision, it now appears that in such circumstances, the patent expiry date will normally be the appropriate start date for the compensable period.

Relevant Period: End Date
Section 8(1)(b) provides that the relevant period ends “on the date of the withdrawal, the discontinuance, the dismissal or the reversal” of the application for a prohibition order. This contrasts with 8(1)(a)(ii) which explicitly allows the court to choose another more appropriate date. In Apotex s8 ramipril FC Snider J held that the court nonetheless has some discretion to choose a more appropriate end date as well, and this view was affirmed by the FCA. While this is not a live issue in this case, Zinn J disagreed with Justice Snider on this point, and held that the court has no discretion to choose a more appropriate date [44]. However, Zinn J would apparently have come to the same conclusion as Snider J on the facts before her, on the view that the compensable loss ended before the end date of the compensable period [44]. That is, rather than couching the issue as one of when the relevant period ends, he would have framed essentially the same issue as one of when the losses are no longer caused by the delay. This way of arriving at the same conclusion is attractive in that it is formally more consistent with the Regulations. The two approaches are not equivalent, however. Because of the Alendronate 2009 FCA 187 rule that s 8 does not allow for compensation for losses suffered outside the s 8 liability period, the two approaches are equivalent only when the last losses caused by the delay are incurred before the end date. In any event, none of this was a live issue, and on the facts the end date was August 1, 2007, the date on which the FCA dismissed Wyeth's application for an order of prohibition.

Deferential Standard of Review of Re-examination Board

Newco Tank Corp v Canada (Attorney General) 2014 FC 287 Mosley J
            2,421,384

In Newco Tank the patentee unsuccessfully appealed a decision of a Re-examination Board cancelling three of the claims of the ‘384 patent as being invalid for obviousness [1]. The patentee raised a number of points which complained, in effect, that the Board had applied the obviousness test too stringently. The main points illustrated by Newco are that Canadian courts continue to apply “an expansive and flexible approach to an obviousness inquiry rather than an overly rigid rule” [27], citing Sanofi 2008 SCC 61, [61]-[63], and that a deferential “reasonableness” standard of review will be applied to decision by the Board on questions of mixed fact and law, such as the application of the test for obviousness to the facts [22].