Apotex Inc. v. Takeda Canada Inc.

Apotex Inc. v. Takeda Canada Inc.
Court (s) Database
Federal Court Decisions
Date
2013-12-11
Neutral citation
2013 FC 1237
File numbers
T-1786-08
Decision Content
Date: 20131211
Docket:
T-1786-08
Citation: 2013 FC 1237
BETWEEN:
APOTEX INC.
Plaintiff
and
TAKEDA CANADA INC.
Defendant
REASONS FOR JUDGMENT
PHELAN J.
I. INTRODUCTION
[1] This is a case under section 8 of the Patented Medicines (Notice of Compliance) Regulations, SOR/93-133 [Regulations], concerning the calculation of loss for which Apotex is entitled to compensation under the Regulations.
[2] The liability for damages arises from a decision of Justice Gauthier, then of this Court, in Solvay Pharma Inc v Apotex Inc, 2008 FC 308, 323 FTR 1, which found that Apotex successfully addressed each of the patents in issue. Takeda’s Notices of Application were dismissed by Justice Gauthier on March 3, 2008 with public reasons on March 6, 2008. Apotex is entitled to compensation for the period it was prevented from entering the market with its drug, Apo-pantoprazole.
[3] The Court is not asked, at this time, to calculate the amount of the loss but to determine certain issues and from which determination the parties believe that they can arrive at the amount of compensation. Failing agreement, the Court may be required to settle the specific amount of compensation.
[4] The relevant provisions of the Regulations are:
8. (1) If an application made under subsection 6(1) is withdrawn or discontinued by the first person or is dismissed by the court hearing the application or if an order preventing the Minister from issuing a notice of compliance, made pursuant to that subsection, is reversed on appeal, the first person is liable to the second person for any loss suffered during the period
(a) beginning on the date, as certified by the Minister, on which a notice of compliance would have been issued in the absence of these Regulations, unless the court concludes that
(i) the certified date was, by the operation of An Act to amend the Patent Act and the Food and Drugs Act (The Jean Chrétien Pledge to Africa), chapter 23 of the Statutes of Canada, 2004, earlier than it would otherwise have been and therefore a date later than the certified date is more appropriate, or
(ii) a date other than the certified date is more appropriate; and
(b) ending on the date of the withdrawal, the discontinuance, the dismissal or the reversal.
(2) A second person may, by action against a first person, apply to the court for an order requiring the first person to compensate the second person for the loss referred to in subsection (1).
(3) The court may make an order under this section without regard to whether the first person has commenced an action for the infringement of a patent that is the subject matter of the application.
(4) If a court orders a first person to compensate a second person under subsection (1), the court may, in respect of any loss referred to in that subsection, make any order for relief by way of damages that the circumstances require.
(5) In assessing the amount of compensation the court shall take into account all matters that it considers relevant to the assessment of the amount, including any conduct of the first or second person which contributed to delay the disposition of the application under subsection 6(1).
(6) The Minister is not liable for damages under this section.
8. (1) Si la demande présentée aux termes du paragraphe 6(1) est retirée ou fait l’objet d’un désistement par la première personne ou est rejetée par le tribunal qui en est saisi, ou si l’ordonnance interdisant au ministre de délivrer un avis de conformité, rendue aux termes de ce paragraphe, est annulée lors d’un appel, la première personne est responsable envers la seconde personne de toute perte subie au cours de la période :
a) débutant à la date, attestée par le ministre, à laquelle un avis de conformité aurait été délivré en l’absence du présent règlement, sauf si le tribunal conclut :
(i) soit que la date attestée est devancée en raison de l’application de la Loi modifiant la Loi sur les brevets et la Loi sur les aliments et drogues (engagement de Jean Chrétien envers l’Afrique), chapitre 23 des Lois du Canada (2004), et qu’en conséquence une date postérieure à celle-ci est plus appropriée,
(ii) soit qu’une date autre que la date attestée est plus appropriée;
b) se terminant à la date du retrait, du désistement ou du rejet de la demande ou de l’annulation de l’ordonnance.
(2) La seconde personne peut, par voie d’action contre la première personne, demander au tribunal de rendre une ordonnance enjoignant à cette dernière de lui verser une indemnité pour la perte visée au paragraphe (1).
(3) Le tribunal peut rendre une ordonnance aux termes du présent article sans tenir compte du fait que la première personne a institué ou non une action en contrefaçon du brevet visé par la demande.
(4) Lorsque le tribunal enjoint à la première personne de verser à la seconde personne une indemnité pour la perte visée au paragraphe (1), il peut rendre l’ordonnance qu’il juge indiquée pour accorder réparation par recouvrement de dommages-intérêts à l’égard de cette perte.
(5) Pour déterminer le montant de l’indemnité à accorder, le tribunal tient compte des facteurs qu’il juge pertinents à cette fin, y compris, le cas échéant, la conduite de la première personne ou de la seconde personne qui a contribué à retarder le règlement de la demande visée au paragraphe 6(1).
(6) Le ministre ne peut être tenu pour responsable des dommages-intérêts au titre du présent article.
[5] The analytical framework for s 8 damages has been set out in Sanofi-Aventis Canada Inc v Teva Canada Ltd, 2012 FC 552, 410 FTR 1 [Teva-Ramipril] and Apotex Inc v Merck Canada Inc, 2012 FC 1235, 105 CPR (4th) 399 [Alendronate 2012].
I adopt that framework and the steps therein. Laid out below are the steps and the unsettled issues for resolution are emboldened.
[6] The steps and issues to be resolved are:
1. The burden of proof (Issue 1);
2. The relevant period for determination of loss: the parties agree that the period is March 9, 2007 to March 5, 2008 [Relevant Period];
3. The size of the total pantoprazole market during the Relevant Period: the parties agree that the total pantoprazole market is the quantity of brand pantoprazole sold during the Relevant Period which is 184,329,100 units of 40 mg tablets (including hospital sales) and 896,000 units of 20 mg tablets;
4. Size of the total generic segment of the pantoprazole market during the Relevant Period: the parties accept the real world experience of the generics entry; therefore, March 3, 2008 for the year to 2009 applied back one year to the Relevant Period is the basis for the size of the generic market within the whole of the pantoprazole market;
5. Apotex’s share of the generic segment of pantoprazole market during the Relevant Period:
a) Number and identity of generic market entrants (Issue 2);
b) If Apotex is alone in market: parties agree Apotex would have 100% of generic market;
c) If Apotex is in competition – Apotex’s percentage of market share (Issue 3);
d) Apotex’s production capacity: parties agree;
e) Timing of formulary listing: parties agree;
6. Apotex’s lost revenue:
a)Price (Issue 4);
b) Inventory adjustment (Issue 5);
c)Double ramp-up (Issue 6).
7. Apotex’s deduction from lost sales:
a)Rebates (Issue 7);
b) Prompt payment discount: parties agree at 1.755% of Apotex’s lost sales revenue;
c)Cost of sales: parties agree on $0.0745 per tablet;
d) Sales commission: parties agree on .22% of Apotex’s lost sales revenues;
e)Freight and distribution: parties agree on $0.001 per tablet;
8. Interest (Issue 8);
9. Discretion to reduce award based on Apotex’s misconduct (Issue 9).
II. GENERAL FACTS
[7] The brand drug at issue in this case sold by Takeda (formerly Nycomed Canada Inc.) is known as Pantoloc. Pantoloc is a protein pump inhibitor [PPI] that contains pantoprazole as its active ingredient and is available in 20 mg and 40 mg tablets.
[8] Pantoprazole was covered by five patents listed on the Patent Register maintained by the Minister of Health; Patent Nos. 2,109,697 [the 697 Patent], 2,310,585 [the 585 Patent], 2,092,694 [the 694 Patent], and 2,089,748 [the 748 Patent]. The fifth patent, No. 1,254,215 [the 215 Patent] was the substance patent for pantoprazole and expired May 16, 2006. Apotex did not challenge the 215 Patent and therefore the Notice of Compliance [NOC] to allow Apotex’s version of pantoprazole on the market could not issue until after May 16, 2006.
[9] Apotex served three Notices of Allegation [NOAs] on October 26, 2005 for the 585 Patent; on January 18, 2006 for the 694 and 748 Patents; and on January 30, 2006 for the 697 Patent.
Apotex served two earlier NOAs on Takeda for the 694 and 748 Patents which were withdrawn.
[10] Takeda filed a Notice of Application against the January 18, 2006 NOA in regard to the 694 and 748 Patents (Court File T-427-06). Takeda did not commence Notices of Application against the other NOAs.
[11] The Minister of Health certified that the examination of Apo-pantoprazole was complete on March 9, 2007. By virtue of the Regulations, Apo-pantoprazole went on “patent hold” and would not receive its NOC until the conditions in the Regulations had been met.
[12] The parties have agreed that Apotex would have received its NOC on March 9, 2007 “in the absence of these Regulations” (paragraph 8(1)(a)). The Court accepts the parties’ agreement on this matter as it does on all other matters where there is agreement.
[13] Justice Gauthier dismissed Takeda’s Notice of Application on March 3, 2008, with public reasons on March 6, 2008. For purposes of calculating Apotex’s loss, the parties have agreed that the end of the Relevant Period in this case is March 5, 2008.
[14] Reflecting Justice Snider’s description of the task in assessing compensation to be awarded (see Apotex Inc v Sanofi-Aventis, 2012 FC 553, 410 FTR 78 [Apo-Ramipril] at paragraph 6 thereof), the question before this Court in this case is:
“What would have happened if Takeda had not brought an application for prohibition?”
The task is complex, theoretic and akin to revising history by changing some facts and predicting the outcome – a “but for” form of analysis.
III. ANALYSIS
A. Issue 1 – Burden
[15] The parties have raised two aspects with respect to onus which are addressed differently. The first aspect is to determine which party bears the onus for which issue. For ease of reference, the Court has generally dealt with the onus matter at the beginning of each issue, on a case-by-case basis.
[16] The second aspect is to deal with the weight of the burden. In general, Apotex’s argument is that a percentage weight may be appropriate – that the Court should consider a percentage probability and then apportion the deduction from damages according to that percentage.
[17] The percentage approach was referred to by Justice Hughes in Alendronate 2012 at paragraphs 37 and 38 referring to the Supreme Court of Canada decision in Athey v Leonati, [1996] 3 SCR 458:
37 Also, both parties have cited and relied upon the decision of the Supreme Court of Canada in Athey v Leonati, [1996] 3 SCR 458, at paragraphs 26 and 27, which I repeat:
26 The respondents argued that the trial judge's assessment of probabilities in causation was similar to the assessment of probabilities routinely undertaken by courts in adjusting damages to reflect contingencies. This argument overlooks the fundamental distinction between the way in which courts deal with alleged past events and the way in which courts deal with potential future or hypothetical events.
27 Hypothetical events (such as how the plaintiff's life would have proceeded without the tortious injury) or future events need not be proven on a balance of probabilities. Instead, they are simply given weight according to their relative likelihood: Mallett v. McMonagle, [1970] A.C. 166 (H.L.); Malec v. J. C. Hutton Proprietary Ltd. (1990), 169 C.L.R. 638 (Aust. H.C.); Janiak v. Ippolito, [1985] 1 S.C.R. 146. For example, if there is a 30 percent chance that the plaintiff's injuries will worsen, then the damage award may be increased by 30 percent of the anticipated extra damages to reflect that risk. A future or hypothetical possibility will be taken into consideration as long as it is a real and substantial possibility and not mere speculation: Schrump v. Koot (1977), 18 O.R. (2d) 337 (C.A.); Graham v. Rourke (1990), 74 D.L.R. (4th) 1 (Ont. C.A.).
38 I emphasize the last sentence of that decision; the hypothetical possibility must be real and substantial not mere speculation. Put another way, the possibility must be realistic and not simply hopeful.
[18] Justice Hughes’ comments arise in the context of pricing only. He did not apply the percentage approach in that case.
[19] Justice Gauthier, then of the Federal Court, rejected that percentage approach in Eli Lilly & Co v Apotex Inc, 2009 FC 991 at paras 760-761, aff’d 2010 FCA 240, because it was only useful in the evaluation of potential future or hypothetical events for the purposes of assessing quantum of damages in a personal injury case.
[20] Like Justices Hughes and Gauthier, I will not adopt that approach. There were no arguments as to the scope of this approach; Apotex did not press this argument strongly. It adds a further and unnecessary level of speculative complexity to a process which is already speculative – where the question is “what would have happened if Apotex has not been prevented from entering the market?” This is also referred to as the “hypothetical world” from time to time here and in other cases.
[21] The better approach is to mirror as much as possible real world circumstances – to use history as the basis of the calculation of the hypothetical world. In this case the parties start from the premise that real world events post Apotex’s NOC give the basis upon which to then work out what likely would have happened if Apotex had not been held back approximately one year.
[22] The evidentiary standard for any proposition asserted must be the civil standard of balance of probabilities that the proposition would be realized. Therefore, Apotex’s approach on this aspect of the burden (to the extent that it continues to assert it) is rejected.
B. Issue 2 - Number and identity of generic market entrants
[23] This issue is directed at when and from whom Apotex would face competition from generic companies. It is Takeda who asserts that such generic competition would occur. Therefore, it is Takeda who bears the burden of establishing who the generic competition market entrants would have been and when they would have entered the market.
Justice Hughes in Apotex Inc v AstraZeneca Canada, 2012 FC 559, 410 FTR 559, at paragraph 35, neatly summarized the governing principle:
35 In brief, it may be said that the party who has led sufficient evidence to put an issue "in play", must, to succeed on that issue, put in sufficient evidence so that on the balance of probabilities, the relevant facts are accepted by the Court as having been proved. Thus Apotex must put in play and subsequently prove on the balance of probabilities the facts that it needs to establish its case for compensation. AstraZeneca must put in play and subsequently prove those facts that it asserts disqualifies Apotex or reduces or negates Apotex's claim for compensation.
[24] On this issue, The Court heard from a number of industry and party witnesses:
• Dr. Bernard Sherman, Chair, Apotex
• Gordon Fahner, Vice President, Business Operations and Finance, Apotex
• Peter Hardwick, Senior Vice President, Sales and Marketing, Apotex
• Frank Murphy, Vice-President, Finance IT Operations, Takeda
• Michel Robidoux, President and General Manager of Sandoz Canada Inc.
• William David Boughner, Director of Strategic Initiatives, Teva
• C. Benjamin Gray, Vice President, Legal and General Counsel, Mylan Pharmaceuticals
• James Erb, Entrepreneur, GRX Healthcare
• Ajay Vashisht, Vice-President of Commercial Operations, Ranbaxy
[25] Both parties advanced different scenarios for the Relevant Period in respect of the degree of competition:
• Apotex’s Scenario 1 assumed Apotex was the sole generic for the Relevant Period.
• Apotex’s Scenario 2 (also Takeda’s Scenario 3), which was the “Common Scenario”, had Apotex as the sole generic for three months at which point Ranbaxy entered. This was described as “Common” not because it was an agreed scenario but it is the only one both parties costed out fully.
• Takeda’s Scenario 1 assumed Teva and Ranbaxy were the first entrants followed by the other generics six months later.
• Takeda’s Scenario 2 had Apotex as the sole generic for one month after which Ranbaxy entered, followed by Teva two months later and a series of other generics three months later.
[26] As will be seen from these Reasons, the Court had to make its own determination rather than accepting one specific scenario. The use of multiple scenarios underscores the speculative nature of the parties’ positions and generally are unhelpful.
[27] Takeda accepts that “but for” its application for prohibition, Apotex would have received an NOC for the Apo-pantoprazole 20 and 40 mg tablets on March 9, 2007 and that Apotex had and has available sufficient manufacturing capacity and motivation to enter the market.
[28] Takeda also takes the position that there would have been three non-related party generics in the market upon filing for their respective NOCs and the entry dates in the hypothetical world correspond to entry dates in the “real world”. Therefore, Takeda says that the NOC dates for these generics would have been: Teva on September 1, 2006; Mylan on September 21, 2007 and Sandoz on October 19, 2007.
[29] With respect to Takeda’s authorized generic [AG] Ranbaxy, Takeda contends that Takeda would have consented to Ranbaxy being in the market and that Ranbaxy would have entered the market at the same time as Teva.
[30] Takeda arrives at this position, at least in part, by concluding that the Regulations are removed from consideration in the hypothetical world in respect of all potential entrants. Takeda interprets the words “… would have been issued in the absence of these Regulations” in paragraph 8(1)(a) as applying to any generic who had applied for an NOC rather than applying only to the successful generic (in this case, Apotex).
[31] Takeda’s position on the number and time of generics’ entry into the market for pantoprazole is based on the following propositions:
• Ranbaxy would have entered the generic market and launched at the same time as the first entrant generic pantoprazole.
• Takeda would have received early warning of Apotex’s imminent entry into the pantoprazole market.
• Given Takeda’s interpretation of the Regulations, other generics are to be injected into the market in the hypothetical world on their patent hold dates.
• No permission from Takeda is required to enter the generic market in the hypothetical world.
• If this Court finds that permission is required, Takeda would have provided it on or after Apotex’s launch.
[32] Apotex’s position in summary is:
• Apotex would have entered the pantoprazole market on March 9, 2007 and Takeda would not have launched its AG (Ranbaxy in this instance) in the hypothetical world. This is based on the assumption that Apotex’s 2007 launch would have been a surprise; Takeda would not have been in a position to launch an AG immediately and absent simultaneous launch of the AG with Apotex, Ranbaxy would not have entered the market.
• There are no other generics in the hypothetical world because Takeda would not have authorized other generics into the market. This requires a consideration of four scenarios developed by Takeda as to what would or could have happened.
[33] A critical consideration for this Issue 2 is how the Regulations are applied in the hypothetical world as it has impact on the entry of other generics. Either these other generics would be allowed to enter the market as of their patent hold date (a result favourable to Takeda) or they would be forced to litigate or be allowed into the market by Takeda (a result favourable to Apotex).
[34] In Apo-Ramipril, the Court held that in the hypothetical world, notice would not have been given. Underlying that conclusion is the premise that notice is part of a scheme of inter-related benefits that accrue to the first person under the Regulations; the other benefits being the right to file a prohibition application and the 24-month stay.
[35] The Apo-Ramipril decision is under appeal but until the Court of Appeal concludes otherwise, in constructing the hypothetical world, I conclude that Apotex would not have given notice of its NOA. The Court of Appeal’s decision in Apotex Inc v Merck & Co, 2011 FCA 329, does not assist Takeda as that decision is distinguishable in that the case dealt with a prohibition application and preliminary application; not section 8 damages (see paragraphs 74-75 thereof).
[36] Although Apotex would not have given notice of its NOA on January 18, 2006 (which would suggest that Takeda would have been surprised by Apotex’s market entry and would take some time to react to the competitive change), the situation is complicated by other NOAs in play.
[37] Firstly, Apotex submitted NOAs on August 15, 2005 relevant to the patents in issue as well as the 585 Patent and 697 Patent but withdrew them. It then filed the relevant NOAs again, which are the subject of Justice Gauthier’s decision. As a consequence of Apo-Ramipril, those notices are therefore not considered as part of the factual matrix of the hypothetical world. Takeda would not have had warning of Apotex’s intent to enter the market on the basis of these NOAs.
[38] Furthermore, in the real world in the last quarter of 2006, Takeda’s head office decided not to proceed with its AG despite both Apotex and Teva providing NOAs. Takeda was not prepared to launch its own AG in the face of warning that other generics were prepared to launch.
[39] On the facts surrounding these NOAs, I have concluded that while Takeda would not necessarily have been caught by complete surprise that Apotex entered the market, the real issue is whether Takeda was prepared or at least sufficiently prepared to compete with Apotex’s generic drug and what strategy it would have employed.
[40] I conclude that in the hypothetical world, Apotex acts without the obligations and limitations of the Regulations. I also find that even with the warning of impending entry, Takeda did not decide to pursue an AG strategy prior to Apotex’s launch and therefore would not have done so in a hypothetical world.
[41] The next matter to consider is whether Takeda would have launched an AG to meet Apotex after it had launched and when it would have done so.
[42] Takeda has summarized five factors to be addressed in this regard. While it is not an exhaustive or a minimal list, it is sufficient for purposes of this case:
1. the importance of the drug to the brand company – an important factor;
2. whether in fact an AG was authorized;
3. the brand company’s financial consideration to introduce an AG;
4. the business arrangements under which the drug is sold and the financial viability of the AG; and
5. whether the brand company had contemplated the possibility of generic entry and possibility of a launch of an AG.
[43] I conclude that Pantoloc was very important to Takeda as it was virtually its only drug in the Canadian market representing 96-98% of its sales. As Frank Murphy, Vice-President, Finance IT Operations of Takeda explained, in around 2002 the brand company began to consider how to extend and defend this brand.
While the drug was important to Takeda in Canada, it appeared to be less so to the head office in Europe. There was not the same level of support from head office as the Canadian subsidiary seemed to require.
[44] With regard to the second factor, an AG – Ranbaxy – was ultimately authorized by Takeda. Apotex contends that it was a self-serving authorization to minimize section 8 damages. Even if that were so, the AG was authorized and was a real factor in the market. It was more than a “shell” or “stalking horse”.
[45] It is with regard to the third factor that Takeda’s case becomes troublesome (at least for Takeda). The financial considerations of Takeda to introduce an AG show problems in determining what strategy to apply and in obtaining corporate support from Takeda’s head office.
[46] At a meeting in November 2005, Takeda executives presented to Takeda’s head office a document entitled “Priorities”. The document outlined three strategies: (1) defending the intellectual property; (2) launching Pantoloc M made with a new salt to extend the patent; and (3) launching an AG.
[47] Murphy testified that there was no hierarchy amongst these priorities but he did admit that they were listed in the order of greatest revenues.
[48] The document is critical evidence as it sets out the strategies considered to deal with Apotex’s market entry. It shows that Apotex’s entry was not a surprise but it also shows limited preparation to deal with generic competition. It also addresses the fifth factor that Takeda had contemplated an AG launch.
[49] The Court must determine whether Takeda would have launched an AG at all (and when) or would have pursued one of the other two strategies. Takeda’s position that it would have launched an AG and done so immediately is weakened by head office’s decision to shelve the AG concept pending resolution of the difficult transfer pricing problem. The position is further weakened by the absence of substantial explanation for abandoning the “priority one” strategy of defending the intellectual property and moving to “priority three”.
[50] Takeda has established that the Canadian office had proceeded quite far down the road of creating an AG. Even in the face of head office opposition, the Canadian operation kept the AG strategy as a fall-back position to be able to launch when and if it could.
[51] The Canadian operation had taken several steps – a generic partner had been selected, heads of agreement and confidentiality agreements were in place. However, both Takeda and Ranbaxy needed their respective head office’s approval.
[52] In the real world, Takeda received approval for an AG about one year later in the fall of 2007. In the hypothetical world, the issue is how quickly Takeda could move. Ranbaxy Canada required approval from both its US office and its India office; something that could proceed concurrently. There are no timelines advanced for these approvals. A cross-reference would have to be sought from Health Canada; a two to four week process.
[53] Even with a motive to move quickly on the part of Takeda, the evidence establishes that Takeda’s head office did not fully endorse its Canadian strategy of an AG, that its Canadian executives were to some extent swimming against the head office current and that the resolution of transfer pricing issues was important and complex. However, the AG strategy seems to be the best option available.
[54] Takeda alleges it would have launched an AG to respond to Apotex’s market entry on March 9, 2007, but Ranbaxy would not be up and running as quickly as Takeda alleges due to the inherent corporate approvals, less than fulsome enthusiasm from Takeda’s head office and related internal issues.
[55] As a consequence, I find that the most reasonable date for Ranbaxy’s launch in the hypothetical world is three months (June 9, 2007) after Apotex’s entry. This is the date estimated in the Third (or Common) Scenario the parties examined.
[56] With respect to the entry of other generics, the same type of factors are taken into consideration. As indicated earlier, the issue of whether in Apotex’s hypothetical world the other generics are subject to the Regulations has been resolved in Apo-Ramipril. In considering Teva’s and other generics’ entry in the hypothetical world, they are subject to the Regulations and Takeda does not enjoy the benefit of claiming that because Apotex was successful, all other generics are, for purposes of hypothetical world analysis, free to enter the market.
[57] Takeda has also raised the issue that since other generics could make section 8 claims, there could be multiple recoveries against Takeda which exceed the total of real losses.
[58] There are several methods by which this problem may be alleviated but ultimately it is the loser’s risk. Justice Snider in Teva Ramapril suggested that multiple recoveries could be accounted for by judicial discretion under subsection 8(5). Further, this Court attempts to have the same judge hear related section 8 damages claims; as was planned in respect of this drug (however, the other section 8 damages case was settled under a confidential agreement). The best resolution is for a defendant to put forward a compelling hypothetical world which accurately reflects the results of other claims.
In the end, this Court need not resolve the problem of multiple recoveries as that is not the matter before it.
[59] In considering whether other generics would enter the market, Justice Hughes in Alendronate 2012 outlined three factors to determine if a generic would enter the market:
• when the generic would have received its NOC;
• whether the generic had the capacity to manufacture or acquire the product in the relevant time; and
• whether the generic was motivated or dissuaded from entering the market during the relevant period.
[60] Takeda’s principal position is that it would have given permission to the other generics Teva (then known as Novopharm), Sandoz, Mylan (then known as Genpharm) and Cobalt to enter the market on their patent hold date in the hypothetical world where Apotex entered as of March 9, 2007. Takeda goes so far as to say that it would have allowed Teva into the market in either September 1, 2006 or October 20, 2006 – six months ahead of Apotex.
[61] The date on which there would be generic competition to Apotex in the hypothetical world is important for a number of reasons. Not only does such competition influence the size and share of the generic market calculations but it impacts the price at which tablets are sold, particularly where Apotex would be the sole generic.
[62] Takeda relies on Alendronate 2012 in support of its position that Teva should be included in the market from its patent hold date six months before Apotex’s deemed entry. In that decision Justice Hughes injected Teva into the market on the date it would have been on patent hold.
[63] However, the facts in Alendronate 2012 are quite different from those in this case. In Alendronate 2012, the parties let the generics on the market in the real world and therefore in the hypothetical world. In the present case, Takeda resisted the various generics at every turn, most particularly Teva.
[64] Takeda’s actions in the real world belie its position that it would have consented to Teva’s entry in a hypothetical world. In summary, when Takeda lost its monopoly, it continued to stifle other generics’ entry into the market including prosecution of ongoing applications against Teva, Cobalt, Sandoz and Mylan, denying generic requests to be let into the market, fighting the generics’ paragraph 6(5)(a) motions to have the prohibition applications dismissed and bringing a motion for reconsideration when Teva succeeded in defeating Takeda’s application.
[65] There is no credible evidence that in a hypothetical world, Takeda would have allowed Teva on the market ahead of Apotex. In fact, the best evidence is that even after Apotex would be in the hypothetical market, Takeda would have fought Teva to the bitter end.
Therefore, in a hypothetical world, I find that Teva would have entered as of its NOC hearing date of September 2008.
[66] With respect to the other generics, Takeda negotiated to allow Cobalt, Mylan and Sandoz on the market in exchange for waiving section 8 damages. It says that it would have done the same in the hypothetical world because Apotex had unlocked the patent.
[67] While there would not, in a hypothetical world, have been any unlocking by Apotex because there would have been no prohibition proceeding, it is more likely than not that in respect of these smaller generics, Takeda would have permitted their entry as of their respective patent hold dates.
[68] In conclusion on this Issue 2, I have concluded that Ranbaxy would only have come into the market three months after Apotex. A third generic (Teva) would have entered on its NOC hearing date and other generics would have come in on their respective patent hold dates.
C. Issue 3 - Apotex’s Share of the Generic Market
[69] This issue is an affirmative claim made by Apotex and therefore it has the onus to establish what its share of the generic market would have been in the hypothetical world.
[70] The issue is determined largely through the expert evidence commenting on various scenarios (earlier described) presented by the parties which assumed certain competition circumstances.
[71] The experts who testified on this issue were:
For Apotex:
Andrew Harington who was qualified as a chartered accountant, chartered business valuator and chartered financial analyst with expertise in investigative and forensic accounting, business valuation and loss quantification in commercial and intellectual property disputes.
Dr. Andrew Tepperman, an expert in industrial organization and economics with particular expertise in pharmaceutical markets and in competition in pharmaceutical markets.
For Takeda
Dr. Paul Grootendorst, an expert in economics with particular expertise in pharmaceutical markets.
[72] Both Grottendorst and Tepperman used econometric models in their calculations. Harington’s evidence was based on data to which business judgment was applied to determine market shares.
[73] While there are differences in approach, methodology and results between the experts, each was well qualified, forthright, candid and credible. They fulfilled their obligations as experts. In preferring one expert over another, the Court does so because it finds their approach and results more compelling in the context of all the evidence, not because of any issues of integrity or candor.
[74] There are significant differences in the market shares calculated by Harington and Grootendorst:
• Harington concluded that Ranbaxy would gain market share in the first three months in the market before reaching “steady state” (conceptually a form of mature market position);
• Grootendorst suggested Ranbaxy would reach this steady state upon entry into the market;
• After three months, both experts reached similar conclusions with respect to Apotex’s and Ranbaxy’s market shares.
[75] Harington used sales data for five molecules including two PPIs, calculated average market share for all five molecules on a province-by-province and month-by-month basis. After performing these calculations, Harington applied business judgment to the calculations in order to determine market share values, discounting anomalous market share figures in certain provinces.
[76] Grootendorst engaged in an econometric analysis using data from 23 comparator molecules. His model then used the number of competitors, the number of months that Apotex preceded the second generic market entrant and the number of months the Apotex product was on the market as variables. Several other variables were used but the equation was never set out in the experts’ report.
[77] I find that the first results of the three months of the hypothetical world scenario is the critical difference between these two experts and the remaining eight months’ results which were modelled are essentially the same.
[78] The third expert, Tepperman, was called largely to rebut and criticize Grootendorst’s econometric model. However, both parties were substantially comfortable with his approach which was used for the scenarios Takeda 1 and 2. Tepperman never analyzed the Common Scenario, the scenario largely accepted by the Court.
[79] Grootendorst’s modelling suffered from using molecules with small sales which distorted their utility as comparators. The model used comparator molecules where the second generic entered the market more than ten months after the first generic and molecules that have simultaneous entry, thus distorting the timing element of comparison. It also included scenarios with more than three market participants.
[80] I find that Grootendorst’s analysis is not a reliable basis for predicting Apotex’s market share because it is unnecessarily complicated and opaque compared to Harington or Tepperman; it reached an unreasonable conclusion that steady state would occur instantly, its comparators were flawed; no regard was given for the “first moves” advantage; the results contained numerous inconsistencies and prediction errors.
[81] The Harington analysis was not free from problems. Harington took a more conventional market analysis approach, relying on real numbers and applying business judgment – the very common features of opinion evidence. However, many of the calculations were subjective and not particularly well explained. As Tepperman noted, Harington did not say how he selected various percentages specifically.
[82] However, opinion evidence always contains subjective elements, educated choices even in a purely mathematical model. As between Harington and Grootendorst, the Court favours Harington because the approach is more realistic and accords more closely in the result to the other facts surrounding this case. The fact that Tepperman supported Harington’s approach generally (despite some reservations) is significant. There is no magic nor necessarily any compelling reason to use an econometric model in the Common Scenario.
[83] Although Tepperman did not assess the Common Scenario, the Court notes that Tepperman was particularly persuasive in his evidence. His model used a large number of observations, it was objective, his approach was balanced, he was the least criticized witness and even Grootendorst acknowledged that it was a good model.
[84] In a “perfect world”, Tepperman would have analyzed the Common Scenario. However, even though Apotex suggested that the Court order Tepperman to determine Apotex’s lost volumes and Tepperman acknowledged that mechanically speaking, it would be reasonably straightforward to do, he put in an important caveat:
Q: Okay. And having modelled it for scenario 1, which involved Teva, Apotex, Ranbaxy and three other generics, so six, and scenario 2 and seeing your results, you used the term “economist inference”. Is there any inference that you have on its applicability to scenario 3?
A: Well, I wasn’t asked to analyze scenario 3, so I haven’t used it to analyze scenario 3. If I were asked to consider scenario 3 using an econometric model, I think there are a number of things that I would want to think about before I set out and mechanically apply this model to scenario 3. It’s quite a different scenario. In scenario 3, Apotex is one of two, where the only other market participant is a much smaller authorized generic. And in scenarios 1 and 2, Apotex is one of six. It’s a much different circumstance.
(Trial Transcript at 855-856).
[85] The effect of sending the matter back to Tepperman is not simply some mathematical exercise but the creation of an entirely new and different expert’s report with no advance knowledge or acceptance of the assumptions or other circumstances leading to the calculations.
[86] Absent the parties’ consent to creating a new and independent expert’s report, the Court must decide the matter on the basis of the best available evidence presented. In that regard, the Court must accept the Harington Report as the basis for calculating Apotex’s share of the gene