Docket: A-119-04
Citation: 2004 FCA 398
CORAM: DÉCARY J.A.
BETWEEN:
PFIZER CANADA INC. AND
PFIZER INC.
Appellants
and
APOTEX INC.
THE MINISTER OF HEALTH
Respondents
Heard at Toronto, Ontario, on November 24th, 2004.
Judgment delivered at Toronto, Ontario, on November 25, 2004.
REASONS FOR JUDGMENT BY: EVANS J.A.
CONCURRED IN BY: DÉCARY J.A.
SEXTON J.A.
Date: 20041125
Docket: A-119-04
Citation: 2004 FCA 398
CORAM: DÉCARY J.A.
SEXTON J.A.
EVANS J.A.
BETWEEN:
PFIZER CANADA INC. AND
PFIZER INC.
Appellants
and
APOTEX INC.
THE MINISTER OF HEALTH
Respondents
REASONS FOR JUDGMENT
EVANS J.A.
[1] This is an appeal from the decision of Snider J. of the Federal Court inPfizer Canada Inc. v. Minister of Health, 2003 FC 1428. In this decision, the Judge dismissed an application by Pfizer Canada Inc. and Pfizer Inc. ("Pfizer") under the Patented Medicines (Notice of Compliance) Regulations, SOR/93-133, for an order prohibiting the Minister of Health from issuing a notice of compliance ("NOC") permitting Apotex Inc. to sell its Apo-azithromycin tablets, until after Pfizer's Canadian Letters Patent No. 1,314,876 (the "876 patent") had expired.
[2] The 876 patent is for crystalline azithromycin dihydrate, which is contained in tablets marketed by Pfizer under the name Zithromax, an antibiotic. In the Notice of Allegation ("NOA") filed when it made its Abbreviated New Drug Submission ("ANDS") for a NOC for its azithromycin tablets, Apotex stated that the tablets would not infringe the 876 patent because:
Our tablets will contain azithromycin itself, but will not contain crystalline azithromycin dihydrate.
In fact, Apotex had also stated in its ANDS, and subsequently told Pfizer, that its tablets contained azithromycin isopropanolate monohydrate ("IPA monohydrate").
[3] Snider J. held that Pfizer had failed to prove on a balance of probabilities that, as a result of either the manufacture of the IPA monohydrate, or its subsequent conversion to the dihydrate, Apotex' NOA was not justified. In other words, Pfizer had not proved that the tablets that Apotex would sell would contain the dihydrate.
[4] Pfizer focussed its attack on two aspects of the decision: the Judge's refusal to draw an adverse inference from Apotex' failure to produce samples of the IPA monohydrate tablets in its possession, and her finding that Apotex' NOA did not contain a full statement of the factual and legal bases of its allegation that the sale of its tablets would not infringe the 876 patent.
1. The adverse inference issue
[5] Pfizer says that, while it had the burden of proving on a balance of probabilities that the sale of Apotex' tablets would infringe the 876 patent, Snider J. erred in law in failing to draw an adverse inference from Apotex' refusal to produce the tablets in its possession. Counsel argued that Apotex withheld them at its peril, even though it was not obliged to produce the tablets under subsection 6(7) of the Regulations because it had not supplied tablets to the Minister when it applied for a NOC.
[6] The argument was that, since the content of the tablets is the central fact in dispute in this case, a fact that Apotex was better placed than Pfizer to prove, Apotex' failure to produce samples of the Apo-azithromycin tablets in its possession warranted the inference that the tablets, when sold, would contain the infringing dihydrate. Pfizer's theory is that, even if Apotex' azithromycin tablets contain the monohydrate when first made up, the monohydrate form is unstable, and is likely to absorb moisture from the air and to convert into the dihydrate when stored, thus infringing the 876 patent.
[7] This issue is best addressed by considering, at the outset, the reasons of Snider J., who derived from the case law the proposition that an adverse inference may be drawn from a party's failure to produce evidence of a fact that it is better placed than the other party to prove, when the evidence is peculiarly within the knowledge of that party and the other party lacks the means of proving it. Counsel for Pfizer does not say that this formulation of the test is erroneous in law.
[8] Accordingly, any error would appear to be in its application by the Judge to the facts. This is a question of mixed law and fact and, in the absence of some question of "pure law", is reviewable only for palpable and overriding error: Housen v. Nikolaisen, [2002] 2 S.C.R. 235. Further, the application of a legal rule relating to the proof of facts and the drawing of factual inferences would seem clearly to lie towards the fact end of the law-fact continuum and, as such, very appropriately the subject of deference by an appellate court.
[9] The Applications Judge concluded that, on the facts before her, the inference was not triggered, because Pfizer had the means of discovering the content of the tablets. She relied on the fact that, pursuant to subsection 6(7) of the Regulations, Apotex had been required to disclose to Pfizer information contained in Apotex' ANDS: the process for making bulk IPA monohydrate, the ingredients of the tablets, and the tableting process.
[10] In addition, an Apotex employee had made a sample of the IPA monohydrate according to the instructions filed with the ANDS. This sample, which was not part of the bulk from which the clinical trial tablets had been made, was tested by experts retained by Apotex, who found no trace of the dihydrate. Some of this sample was also given to Pfizer. There is no evidence whether their experts tested it or made tablets from it, even though one of them stated that tableting would be easy.
[11] When this evidence is considered as a whole, it cannot, in my view, be said that Snider J. made a palpable and overriding error in concluding that Pfizer had the means to discover the content of the tablets. It is true that any tablets made by Pfizer from the bulk IPA monohydrate provided by Apotex would not be the exact tablets that Apotex would sell if it obtained a NOC. However, as I explain below, nor were the tablets that Apotex had in its possession.
[12] Since Snider J. was entitled to find on the evidence before her that Pfizer did not lack the means to discover the content of Apotex' azithromycin tablets, it is not strictly necessary to consider whether the tablets' content was peculiarly within Apotex' knowledge. However, since this issue was debated at the hearing of the appeal, it is appropriate to say something about it.
[13] While Apotex asserts that it had manufactured 20,000 tablets for the purpose of the clinical trials, most of which were still in its possession, it argued that production of a sample of these tablets, even if found to contain the dihydrate, would not necessarily prove that its sale of Apo-azithromycin would infringe the 876 patent.
[14] Despite the apparently large number of tablets produced, Apotex says that, since it had made them for the clinical trials, not for sale, they would not necessarily have been handled and stored with the caution required to prevent conversion from the mono to the dihydrate. In contrast, tablets manufactured for sale would receive the very cautious handling needed to ensure that they did not infringe by converting into the dihydrate.
[15] In addition, it was said, the clinical trial tablets' shelf-life expired in November 2002, prior to the release of Snider J.'s reasons for decision. However, Pfizer had requested the tablets in a letter of December 3, 2001, and, prior to November 2002, had unsuccessfully moved for their production.
[16] I do not know, and need not decide, whether all the above evidence about the tablets would have been sufficient to trigger the adverse inference presumption, if it had not been open to Snider J. to conclude that Pfizer had the means of discovering whether tablets made from the IPA monohydrate bulk contained, or would later convert into, the dihydrate.
[17] Counsel for Pfizer argued that Snider J. had erred in law by mischaracterizing the issue. The argument was that the Judge was misled into thinking that the central issue was whether Pfizer's expert had proved infringement by showing that the sample of bulk IPA monohydrate that he had made by following the instructions provided by Apotex contained dihydrate. Whereas, of course, the real issue was what the tablets contained.
[18] With all respect, I do not read the Judge's reasons in this way. In my opinion, she understood quite clearly the key fact in this case. In my view, Snider J. regarded the question of whether IPA monohydrate could be made by following Apotex' instructions as relevant to determining if the availability of the instructions for making it was an adequate surrogate for testing the tablets themselves in order to discover their content.
[19] Counsel also submitted that, whenever the content of tablets is the issue, the tablets themselves are the only evidence of their content. Thus, the availability of instructions for making bulk IPA monohydrate, and the possession of a sample of bulk IPA monohydrate made according to the ANDS with easy-to-follow instructions for making tablets from it, were not adequate surrogates for the tablets themselves when they existed. In my view, this argument cannot succeed.
[20] First, Snider J. found as a fact (at para. 29) that, whether or not the tablets were the best evidence of the content of the tablets that Apotex would sell if it obtained a NOC, they were not the only evidence. I cannot say that this finding of fact was vitiated by palpable and overriding error.
[21] Second, counsel submitted that, as a matter of law, Apotex' refusal to produce the tablets in its possession was sufficient in itself to give rise to the adverse inference presumption since their content was the central issue in the proceeding. In my view, this proposition is not supported by the case law. The law on adverse inferences is as stated by Snider J. and applies in the context of NOC prohibition proceedings. The triggering of the inference is very dependent on the facts of the particular NOC case, just as it is in other types of litigation.
[22] Indeed, there are good reasons for not accepting counsel's argument that the law should more readily require the drawing of an adverse inference from the non-production of drugs when their content is in issue in NOC cases. They include: the summary nature of the prohibition proceeding and the absence of discovery; the absence of a duty on applicants making an ANDS to supply samples of the drug to the Minister and, hence, to provide them under subsection 6(7) to the applicant in the prohibition proceeding; and the fact that the dismissal of the application for an order of prohibition only denies the applicant the benefit of the statutory stay on the entry onto the market of a possibly infringing product, and does not preclude a subsequent action for infringement, complete with discovery.
[23] Nonetheless, the result in this case is grounded firmly in the particular evidence before the Applications Judge and in the findings of fact that she made on that evidence. It is certainly not my intention to discourage judges, on appropriate facts, from drawing an adverse inference from the non-production of medicines when their content is at issue in NOC proceedings. If non-production, and other forms of gamesmanship, are recurring problems in this kind of proceeding (and this is certainly not the first case in which the issue has arisen and been commented on), they are best remedied through an amendment to the Regulations. Meanwhile it falls to the Courts to do what they can to protect the judicial process from abuse.
2. The sufficiency of the NOA
[24] Counsel argued that the Judge had erred in holding that Apotex' NOA complied with the requirement of paragraph 5(3)(a) of the Regulations that it provide "a detailed statement of the legal and factual basis for the allegation". The basis of her conclusion was that, to use the test in SmithKline Beecham Pharma Inc. v. Apotex Inc. (2001), 10 C.P.R. (4th) 338 (F.C.A.), the NOA was "sufficient to make the appellants fully aware of the grounds on which the respondents claimed that the patent would not be infringed if the notice of compliance was issued."
[25] Counsel for Pfizer conceded that Snider J. correctly formulated the legal test for determining the adequacy of an NOA: the dispute is about its application to the facts. Again, this is a question of mixed fact and law and, in the absence of some extricable general legal proposition, this Court can only interfere with the Judge's conclusion that the NOA was sufficiently detailed to comply with paragraph 5(3)(a) if it was vitiated by some palpable and overriding error. I am not satisfied that it was.
[26] It was open to Snider J. to conclude that Apotex' statement in the NOA that its tablets contained "azithromycin itself", which was not the subject of the patent, but did not contain "crystalline azithromycin dihydrate", was sufficient to alert Pfizer to the basis of Apotex' allegation, namely that it would not infringe because its azithromycin tablets will not contain a dihydrate, even though it did not also state that the tablets will contain the monohydrate.
[27] This conclusion is also supported by SmithKline Beecham Pharma, supra, where, on very similar facts, the Court found the NOA to be sufficient.
[28] For these reasons, I would dismiss the appeal with costs.
"John M. Evans"
J.A.
"I agree
Robert Décary, J.A."
"I agree
A. J. Sexton, J.A."
FEDERAL COURT OF APPEAL
NAMES OF COUNSEL AND SOLICITORS OF RECORD
DOCKET: A-119-04
STYLE OF CAUSE: PFIZER CANADA INC.
and PFIZER INC.
Appellants
and
APOTEX INC.
and THE MINISTER OF HEALTH
Respondents
PLACE OF HEARING: TORONTO, ONTARIO
DATE OF HEARING: NOVEMBER 24, 2004
REASONS FOR JUDGMENT: EVANS J.A.
DATED: NOVEMBER 25, 2004
APPEARANCES:
Sheila R. Block
Andrew Shaughnessy
Andrew Bernstein FOR THE APPELLANTS
Harry Radomski
Andrew Brodkin FOR THE RESPONDENT, APOTEX INC.
No appearance FOR THE RESPONDENT, MINISTER OF HEALTH
SOLICITORS OF RECORD:
TORYS LLP
Barristers & Solicitors
Toronto, ON FOR THE APPELLANTS
GOODMANS LLP
Barristers & Solicitors
Toronto, ON FOR THE RESPONDENT, APOTEX INC.
Morris Rosenberg
Deputy Attorney General of Canada FOR THE RESPONDENT, MINISTER OF HEALTH