Janssen Pharmaceutica Inc. c. Apotex Inc. (C.A.) [2002] 1 C.F. 393
Date: 20010802
Docket: A-85-00
Neutral citation: 2001 FCA 247
CORAM: LINDEN J.A.
BETWEEN:
APOTEX INC.
Appellant
(Respondent)
- and -
JANSSEN PHARMACEUTICA INC. and
JANSSEN PHARMACEUTICA naamloze vennootschap
Respondents
(Applicants)
- and -
THE MINISTER OF NATIONAL HEALTH AND WELFARE
Respondent
(Respondent)
LINDEN J.A.
Introduction
[1] This is an appeal from an order of the Federal Court, Trial Division dated February 7, 2000 prohibiting the Minister of National Health and Welfare (the "Minister") from issuing a Notice of Compliance ("NOC") to the appellant, Apotex Inc. ("Apotex"), in respect of the drug cisapride monohydrate ("cisapride") until the expiration of Canadian Letters Patent No. 1,183,847 (the "Janssen Patent") (See Janssen Pharmaceutica Inc. v. Apotex Inc. (2000), 5 C.P.R. (4th) 53, [2000] F.C.J. No. 164 (QL)).
[2] The order of the Motions Judge arose from facts which may be summarised briefly. The respondent Janssen Pharmaceutica naamloze vennootschap ("Janssen Belgium") owns the Janssen Patent which relates to a number of novel compounds having utility as new pharmaceuticals. Of interest in this appeal is the Janssen Patent's process and product by process claims to cisapride, a peristaltic stimulant used to treat gastrointestinal disorders. The second respondent, Janssen Pharmaceutica Inc. ("Janssen Canada"), is the exclusive licensee under the Janssen Patent, which was included in the Patent Lists submitted to the Minister by Janssen Canada pursuant to subsection 4(1) of the Patented Medicines (Notice of Compliance) Regulations, S.O.R. 93-133 (the "Regulations") in connection with NOCs issued to it in respect of its 5 mg and 10 mg tablets, and 1 mg/ml oral suspension of cisapride.
[3] By letter dated July 4, 1994, Apotex made a Notice of Allegation pursuant to subsection 5(3) of the Regulations alleging that the making, constructing, using or selling by Apotex of tablets containing cisapride would not infringe the Janssen Patent. Subject to a protective order dated November 1, 1994, Apotex served and filed a Detailed Statement of Fact and Law on November 3, 1994. In that Statement, Apotex described the process by which its supplier, Torcan Chemical Ltd. ("Torcan"), manufactures cisapride (the "Torcan Process"). Also included in the Statement was an affidavit of Dr. McClelland sworn on September 2, 1994 in which Dr. McClelland explains why the Torcan Process is not within the scope of the Janssen Patent.
[4] On August 22, 1994, the respondents Janssen Belgium and Janssen Canada (hereinafter "Janssen") filed an Originating Notice of Motion seeking an order pursuant to subsection 6(1) of the Regulations prohibiting the Minister from issuing a NOC to Apotex in connection with the drug cisapride until the expiration of the Janssen Patent on March 12, 2002. Janssen took the view that the Notice of Allegation made by Apotex was not justified because the Torcan Process described in it infringes the process claimed in the Janssen Patent or is an obvious chemical equivalent thereof.
[5] In December of 1998 and January of 1999, the motion was heard in the Trial Division. As previously mentioned, the Motions Judge decided to grant the application and issued a prohibition order on February 7, 2000. He did so on the basis that, in his view, the Torcan Process is either within the scope of the process claimed in the Janssen Patent or involves an obvious chemical equivalent of the essential reaction described therein.
The Issue
[6] There are several issues to be decided on appeal. However, the essential issue for determination is whether, with respect to the critical acylation reaction necessary for making cisapride, the intramolecular reaction employed in the Torcan Process is an "obvious chemical equivalent" of the intermolecular condensation reaction described and claimed in the Janssen Patent.
The Competing Processes
[7] In order to appreciate the nature of the issue to be determined, it is necessary to describe the relevant claims in the Janssen Patent as well as the method for producing cisapride set out in the Torcan Process.
(a) The Janssen Patent
[8] Of concern in this appeal are claims 1 and 5 of the Janssen Patent. Claim 1 sets out three alternative processes for preparing certain compounds, including cisapride, by forming a carbon-nitrogen-carbonyl linkage (or "amide bond") between the piperidine ring and a substituted benzoyl group, followed by a number of optional subsequent steps. Claim 1 is best understood by careful reference to the written and diagrammatic description of it at pages 125 to 130 of the Janssen Patent (Appeal Book, Vol. I, Tab 7 at 203-208).
[9] The desired end-product of the processes described in claim 1 has a formula referred to in the Janssen Patent as "formula (I)" which is described and illustrated at pages 125-127 of the Patent. The first of the three alternative processes for producing the compounds of formula (I) is the most relevant to this appeal. In the words of claim 1, that process is characterized by
...reacting a piperidine of formula (II) with a carboxylic acid of formula (III) or an appropriate functional derivative thereof, in a suitable medium; ... and, if desired, where R1 is hydrogen, converting a compound of formula (I-a-1) into a compound of formula (I-a-2) by reacting (I-a-1) with an appropriate alkylating agent of formula R1-a in a suitable medium, said R1-a -W (VI) having the meaning of R1 provided that hydrogen is excluded; ... and/or preparing stereochemically isomeric forms thereof. [Diagrams omitted]
[10] In the simpler language of Janssen's factum, the above process relates to making the compounds of interest through the formation of an amide bond by way of a nucleophilic substitution or an acylation reaction between the nitrogen of the piperidine of formula (II) and the carbonyl group on the carboxylic acid, or a functional derivative of carboxylic acid, of formula (III). The first relevant optional step thereafter is the replacement of the hydrogen at R1 with a methyl group (CH3), otherwise known as O-methylation. The second claimed optional step relevant to the appeal is the preparation of stereochemical isomers of the compounds of interest; in other words, making the cis or trans isomers of the compounds. Accordingly, claim 1 of the Janssen Patent, insofar as it is relevant to the appeal, describes a three-step process for making cisapride comprised of
(1) the formation of the amide bond by an acylation reaction;
(2) O-methylation; and/or
(3) preparing stereoisomers of the compounds.
[11] Claim 5 of the Janssen Patent claims the chemical compound expressed in formula (I) above "whenever prepared or produced by the process of claim 1 or by any obvious chemical equivalent thereof". Therefore, the Janssen Patent claims the three-step process enumerated in the last paragraph and "any obvious chemical equivalent thereof".
(b) The Torcan Process
[12] Given the absence of a patent for the Torcan Process, it is difficult to find an objective description of the process in the record. However, reference can be made to the "General Description of Manufacturing Procedure" included in the New Drug Submission filed by Apotex with respect to its cisapride tablets (See Appeal Book, Vol. II, Tab 24 at 314). That description reads as follows:
3,4,4-Trimethoxypiperidine 1 is alkylated with 1-(4- fluorophenoxy) propyl chloride 2 in refluxing methyl isobutyl ketone in the presence of anhydrous potassium carbonate to yield N-1-[3-(4-fluorophenoxy)propyl]-3,4,4-trimethoxy-4-piperidine 3. The intermediate 3 without isolation is subjected for acid catalyzed hydrolysis. After basification and extraction with an organic solvent the crystalline N-1-[3-(4-fluorophenoxy) propyl]-3-methoxy-4-piperidone 4 is isolated by filtration from a suitable solvent or a solvent system such as ethyl acetate: hexane. The intermediate 4 is reacted under reductive amination conditions with benzylamine 5 to yield the benzylamino derivative 6. Hydrogenation of the last intermediate results in the cis-N-1-[3-(4-fluorophenoxy)propyl]-3-methoxy-4-piperidamine 7. The mixed anhydride method using carbobenzoxy chloride and triethylamine is used to couple 4-amino-5-chloro-2-methoxybenzoic acid 8 and the intermediate 7. After work up of the crude cisapride 9 is isolated by filtration and is purified by recrystallization from a suitable solvent or solvent system such as ethanol: water.
[13] In the factum filed by Apotex, The Torcan Process is described in more digestible language (See Appellant's Memorandum of Fact and Law at paragraphs 11-21). The process is described by Apotex as involving the reaction of piperidine and a benzoyl unit to form a larger molecule in which the linkage between the piperidine and benzoyl unit occurs at position 3 of the piperidine ring. Specifically, the carbon at position 3 is linked to an oxygen attached to the benzoyl group. In order to bring about the requisite amide linkage at position 4 of the piperidine ring, the Torcan process then uses an intramolecular reaction which disconnects the piperidine ring and benzoyl group at position 3 and then reconnects them through the central nitrogen at position 4. The altered molecule is subsequently reacted to eliminate the methoxy group at position 4 of the piperidine ring using triethylsilane and a strong acid to produce a 3-keto derivative. With the "bulky" amide (NH) at the 4 position, reduction of the C=0 results in the C-H bond at the 3 position of the piperidine ring opposite the amide. As a result, the OH and amide groups (NH) in the final compound are in the required cis arrangement.
Decision of the Motions Judge
[14] The Motions Judge began by construing claims 1 and 5 of the Janssen Patent. In doing so, he noted that a purposive construction should be given to the claims in light of the decision by the House of Lords in Catnic Components Limited and Another v. Hill and Smith Limited, [1982] R.P.C. 183 (H.L.). The Motions Judge reproduced a passage found at page 243 of the Catnic decision which reads, in part, as follows:
The question in each case is: whether persons with practical knowledge and experience of the kind of work in which the invention was intended to be used, would understand that strict compliance with a particular descriptive word or phrase appearing in a claim was intended by the patentee to be an essential requirement of the invention so that any variant would fall outside the monopoly claimed, even though it could have no material effect upon the way the invention worked.
[15] The Motions Judge proceeded to restate some of the well-established principles of patent construction. Specifically, he noted that patent construction must be neither benevolent nor harsh, that it must be fair and reasonable to the patentee and the public, that it must be done with a mind willing to understand the invention and that it should endeavour to give the inventor protection for that which he has in good faith invented.
[16] The Motions Judge also took note of the fact that at the time the Janssen Patent was filed, cisapride was a novel compound. In this connection, he observed that if it were not for the limitations contained in subsection 41(1) of the Patent Act, R.S.C. 1970, c. P-4 at the time of filing, cisapride could have been claimed in a product per se format, regardless of how the compound was made. Drawing support from the decision of Richard J. (as he then was) in Pfizer Canada Inc. v. Apotex Inc. (1997), 77 C.P.R. (3d) 547 (F.C.T.D.), the Motions Judge noted that because the types of reactions included in the Janssen Patent were general reactions known at the time of filing and because there was nothing inventive in the idea of using these reactions, the patentee could not have intended to be limited to the specific processes set forth in the claims. Rather, the process limitations were only included to comply with subsection 41(1) of the Patent Act. Given claim 5's coverage of the processes in claim 1 and all other obvious chemical equivalents, the Motions Judge determined that the specific process limitations in claim 1 could not be read as essential features.
[17] Having considered the expert evidence adduced by the parties, the Motions Judge construed the Janssen Patent. At paragraph 58 of his reasons, he came to the following conclusion:
In my view, on a purposive construction of the '847 [Janssen] Patent "and construing the claims with a mind willing to understand the true essentials of the invention", the true purpose of the first process of claim 1 is to make cisapride by the formation of the amide bond, and optionally, 0-methylation and/or preparing the cis isomer. With respect to claim 5 of the '847 [Janssen] Patent, it is my view that the claim claims cisapride when made or prepared by the first process of claim 1, i.e. formation of the amide bond, and optionally, 0-methylation and/or preparing the cis isomer, and by all other obvious chemical equivalents of the claim 1 process. [footnotes omitted]
[18] The Motions Judge then turned to the issue of infringement. In doing so, he set out (at paragraph 59) the test established by Lord Diplock in Catnic, supra as restated by Hoffmann J. (as he then was) in Improver Corporation v. Remington Consumer Products Ltd. (1989), 17 F.S.R. 181 (Pat.Ct.) at 189:
1. Does the variant have a material effect upon the way the invention works? If yes, the variant is outside the claim. If no, -
2. Would this (i.e. that the variant had no material effect) have been obvious at the date of publication of the patent to a reader skilled in the art? If no, the variant is outside the claim. If yes, -
3. Would the reader skilled in the art nevertheless have understood from the language of the claim that the patentee intended that strict compliance with the primary meaning was an essential requirement of the invention? If yes, the variant is outside the claim.
The Motions Judge added his view, based on Novocol Chemical Manufacturing Company of Canada Limited v. W.R. MacFarlane et al., [1939] Ex.C.R. 151, that where a variant performs the same function in substantially the same manner as the claimed process, it will constitute an equivalent.
[19] In applying the Catnic test, the Motions Judge centred his analysis on the formation of the amide bond in the competing processes for making cisapride. In this connection, he framed the infringement issue (at paragraph 62) in this way:
It is undeniable, in my view, that the Torcan amide bond is made in a manner analogous to the '847 [Janssen] Patent amide bond, i.e. by way of a nucleophilic acyl substitution reaction involving the non-bonded electrons of the nitrogen attacking the carbonyl carbon, which results in the displacement of an appropriate leaving group Y, to form the nitrogen-carbonyl carbon bond. There are a number of variants between the '847 [Janssen] Patent and the Torcan Process, of which the principal is the use of an intramolecular acylation in lieu of an intermolecular acylation. The issue, therefore, is whether, by reason of these variants, the Torcan Process falls outside the scope of the '847 [Janssen] Patent.
[20] The Motions Judge assessed the expert evidence in order to determine whether the variants - especially the intramolecular method of acylation reaction - in the Torcan Process fell outside the scope of the Janssen Patent. He accepted the credibility of the evidence adduced by Janssen's experts, particularly the evidence given by Dr. Victor Snieckus. By contrast, the Motions Judge found that Apotex's chief expert, Dr. McClelland, failed to construe the Janssen Patent with a mind willing to understand the true essentials of the invention. Moreover, he was of the opinion that Dr. McClelland had no interest in addressing the functional similarities between the processes but instead examined the Torcan Process and Janssen Patent for the specific purpose of emphasizing their differences.
[21] Adopting the analysis of Dr. Snieckus, the Motions Judge found that the only difference between the intramolecular reaction in the Torcan Process and the intermolecular reaction described in the Janssen Patent was that in the former reaction, the two reactive groups (an amine and an ester) are part of the same molecule whereas in the latter reaction, they are part of different molecules. This difference had no effect on the nature of the reaction involved, which is determined by the nature of the reactive groups. Those groups were the same in each process.
[22] The Motions Judge further accepted evidence that synthetic chemists have known for quite some time that they should consider intramolecular versions of an intermolecular reaction when planning a synthetic pathway. He also accepted evidence that the intramolecular reaction used in the Torcan Process is well known and has been performed since 1937. Repeating his view that there is nothing in the Janssen Patent limiting the main reaction to an intermolecular one, the Motions Judge concluded that the intramolecular reaction in the Torcan Process either falls within the scope of claim 1 of the Janssen Patent or constitutes an obvious chemical equivalent of the acylation reaction described in claim 1.
[23] Addressing the other variants in the Torcan Process, the Motions Judge agreed with Dr. Snieckus that those variants did not affect the substance of the main reaction. He pointed out that all nine steps of claim 1 as identified by Dr. Snieckus (see schematic at paragraph 54 of Dr. Snieckus' Affidavit, Appeal Book, Vol. II, Tab 33 at 371) were used in the Torcan Process. He agreed with Dr. Snieckus that the differences essentially consisted in the interposition of the additional steps necessary to perform the acylation reaction as an intramolecular reaction instead of an intermolecular one. The Torcan Process, therefore, attempted to add complexity to the process claimed in the Janssen Patent for the purpose of differentiating itself from the patented process.
[24] In the result, the Motions Judge concluded that the sale of cisapride manufactured by the Torcan Process would infringe the Janssen Patent. Accordingly, an order was issued pursuant to subsection 6(2) of the Regulations prohibiting the Minister from issuing a NOC to Apotex until the expiry of the Janssen Patent.
Arguments on Appeal
[25] Mr. Radomski, counsel for Apotex, argues, in his typically thorough style, that the Motions Judge committed a number of reversible errors in arriving at his conclusion that the Torcan Process infringes the Janssen Patent. The first grouping of alleged errors relates to the appreciation of the expert evidence by the Motions Judge. In this regard, Apotex contends that there was no valid basis upon which the Motions Judge rejected the evidence of Dr. McClelland in favour of the evidence submitted by Janssen's chief expert, Dr. Snieckus. Specifically, it is argued that Dr. McClelland did not disregard the similarities between the Torcan Process and the process claimed in the Janssen Patent but rather took as an accepted first premise that both processes employed the essential nitrogen-carbonyl linkage to produce the same end product. Contrary to the statement by the Motions Judge, Apotex asserts that determining equivalence requires a comparison of both the similarities and the differences between the competing processes.
[26] Apotex also challenges the view of the Motions Judge that Dr. McClelland failed to address the issue of whether the processes were "obvious chemical equivalents" within the meaning of that phrase in claim 5 of the Janssen Patent. It does so on the basis that Dr. McClelland's affidavit clearly refers to the issue of obvious chemical equivalence and specifically concludes that the two processes are not equivalents. Apotex additionally takes issue with the criticism by the Motions Judge regarding Dr. McClelland's failure to address a schematic prepared by Dr. Snieckus representing the transition from the formation of an imine and an amine to the end of the acylation reaction in both processes. Apotex argues that such criticism is unfair given the fact that Dr. Snieckus conceived of this transition by the use of intermediates 7a and 7b, neither of which are claimed in the Janssen Patent. Furthermore, these intermediates are merely Dr. Snieckus' conception of what is occurring in the Torcan Process and constitute only one possible means by which the essential reaction may be effected.
[27] It is also submitted by Apotex that the Motions Judge erred in accepting the evidence of Janssen's expert, Dr. Snieckus. Apotex highlights two principal ways in which the evidence of Dr. Snieckus is tainted. First, it is argued that Dr. Snieckus was of the opinion that the key step in both processes was a "condensation" reaction. However, once Dr. Snieckus realised that the Torcan Process did not employ such a reaction due to its intramolecular nature, he changed his definition of "condensation" in order to include an intramolecular reaction such as that in the Torcan Process that does not produce water. Counsel for Apotex calls this change in definition a "transparent attempt to ‘cover up'" on the part of Dr. Snieckus (Appellant's Memorandum of Fact and Law at paragraph 63). Given this inconsistency in the evidence of Dr. Snieckus, the acceptance of that evidence by the Motions Judge constitutes, in the opinion of Apotex, a palpable and overriding error.
[28] The second alleged source of taint in Dr. Snieckus' evidence is the fact that, according to Apotex, Dr. Snieckus admitted on cross-examination that virtually the entire contents of his affidavit were prepared by Counsel for Janssen. That Dr. Snieckus' evidence was prepared by another person should have led the Motions Judge to accord less weight to it. This fact similarly undermines the evidence of another Janssen witness, Dr. Van Lommen, who gave his evidence following a review of Dr. Snieckus' affidavit. In addition, Dr. Van Lommen is said to lack the requisite independence to be considered an expert. He is a long-time employee of Janssen who had an obvious interest in the outcome of the proceeding. His evidence should therefore have been given little or no weight by the Motions Judge.
[29] The other grouping of errors alleged by Apotex to have been committed by the Motions Judge includes a number of legal errors. These errors relate to the construction of the Janssen Patent and the application of the doctrine of equivalents in order to determine the issue of infringement. Apotex submits that the Motions Judge erred in concluding that, on a purposive construction, the Janssen Patent was not limited to the processes claimed therein and in particular was not limited to an intermolecular reaction. However, Apotex contends that what claim 1 does not include is not instructive as to the breadth of the claimed process. Rather, the applicable jurisprudence requires that claim 1 be construed on the basis of what is explicitly included in it. Contrary to the approach adopted by the Motions Judge, the language of claim 1 cannot be broadened by reference to the words "all obvious chemical equivalents" in claim 5. Any obvious equivalence must be based on the explicit terms of claim 1.
[30] On a proper construction of the Janssen Patent, therefore, Apotex contends that claim 1 only covers an intermolecular reaction. There is no evidence that the Janssen Patent contemplates an intramolecular reaction such as the one employed in the Torcan Process. Indeed, in Apotex's submission, claim 1's silence on the possibility of an intramolecular reaction demonstrates that an intramolecular alternative is not at all obvious.
[31] Another alleged error in the construction of the Janssen Patent is said to be the determination by the Motions Judge that the true purpose of claim 1 is to make cisapride by the formation of an amide bond. Apotex argues that this construction of the patented process essentially converts the Janssen Patent into a product per se claim to cisapride which is prohibited by the terms of subsection 41(1) of the Patent Act pursuant to which the Janssen Patent was filed. This is because the manufacture of cisapride necessarily requires the formation of an amide bond. Consequently, the construction given to claim 1 by the Motions Judge amounts to accepting that a legitimate purpose of a process claim is to protect the end product itself. To avoid this absurdity, Apotex submits that the only reasonable construction of claim 1 must be that it cover the manufacture of cisapride by the formation of an amide bond through an intermolecular condensation reaction. On these terms, Torcan's intramolecular reaction is not infringing.
[32] It is further argued that the Motions Judge erred by misapplying the doctrine of equivalents. Specifically, Apotex takes exception to the proposition allegedly espoused by the Motions Judge that a process will be equivalent where it functions to effect the same result as a patented process. If this were true, it would be impossible not to infringe a product by process patent because any process developed to produce the same compound would be taken as an equivalent. That the end result is the same in two processes does not assist the analysis. In Apotex's view, it is the differences in approach and chemical strategy which are relevant to determining chemical equivalence. However, the Motions Judge is said to have incorrectly limited his analysis to the functional similarities between the processes. By adopting this approach, it is submitted that the Motions Judge proceeded at a level of generality and superficiality which neglected to address the key differences in chemical strategy which distinguish the Torcan Process from the process described in claim 1 of the Janssen Patent.
[33] Finally, Apotex contends that the Motions Judge erred in his understanding of what constitutes an "obvious" chemical equivalent. It argues that the Motions Judge was wrong to find that Torcan's intramolecular acylation reaction was an obvious chemical equivalent purely on the basis that such a reaction was known to chemists at the time. Apotex urges a higher threshold of obviousness. In its view, Torcan's reaction would only be obvious if the unimaginative but skilled chemist knew for certain that an intramolecular acylation reaction would work, not simply that such reaction was worth a try. In the circumstances, Apotex argues that there was no indication that an intramolecular acylation reaction was certain to work. Indeed, the failure of claim 1 to contemplate such a chemical strategy certainly indicates that it could not have been an obvious alternative to Janssen's intermolecular reaction.
Analysis
(A) The Law of Patent Construction
[34] In order to verify the general allegation by Apotex that the Motions Judge erred in construing the Janssen Patent, it is necessary to review, if only briefly, the law applicable to patent construction. The recent decisions of the Supreme Court in Free World Trust v. Électro Santé Inc., [2000] 2 S.C.R. 1024, 2000 SCC 66 and Whirlpool Corp. v. Camco Inc., [2000] 2 S.C.R. 1067, 2000 SCC 67, which were released after the decision being appealed, explain in comprehensive fashion the principles of patent construction which must be followed in cases like the present appeal. In Free World Trust, Binnie J. explained the Court's view of the nature of patent construction in Canadian law (at paragraphs 14-15):
Patent claims are frequently analogized to "fences" and "boundaries", giving the "fields" of the monopoly a comfortable pretense of bright line demarcation ...
In reality, the "fences" often consist of complex layers of definitions of different elements (or "components" or "features" or "integers") of differing complexity, substitutability and ingenuity. A matrix of descriptive words and phrases defines the monopoly, warns the public and ensnares the infringer. In some instances, the precise elements of the "fence" may be crucial or "essential" to the working of the invention as claimed; in others the inventor may contemplate, and the reader skilled in the art appreciate, that variants could easily be used or substituted without making any material difference to the working of the invention. The interpretative task of the court in claims construction is to separate the one from the other, to distinguish the essential from the inessential, and to give to the "field" framed by the former the legal protection to which the holder of a valid patent is entitled. [emphasis added]
The significance of distinguishing essential from non-essential elements is that the substitution or omission of an essential element of a patented invention will defeat an allegation of infringement whereas the substitution or omission of a non-essential element will not necessarily foreclose a patentee's claim of infringement.
[35] In Camco, Binnie J. provided another useful summary of the law of patent construction. He noted that the phrase "purposive construction" coined by Lord Diplock in Catnic did not represent a particular innovation in the law of patent interpretation. Rather, he viewed the approach explained in Catnic as a restatement of the existing jurisprudence on the subject. At paragraph 48 of the Camco decision, Binnie J. expressed the Court's opinion as follows:
In Catnic, as in the earlier case law, the scope of the monopoly remains a function of the written claims but, as before, flexibility and fairness is achieved by differentiating the essential features ("the pith and marrow") from the unessential, based on a knowledgeable reading of the whole specification through the eyes of the skilled addressee rather than on the basis of "the kind of meticulous verbal analysis in which lawyers are too often tempted by their training to indulge" (Catnic, supra, p. 243).
Though patent construction must be tied to the language of the patent, a simple "dictionary" or "grammatical" approach to patent construction is to be avoided. The terms of the specification, including the claims, must be given meaning and purpose by the skilled addressee applying his or her knowledge in the field to which the patent relates (Camco at paragraphs 52-53).
[36] One important caveat bears mentioning. While it is in practice difficult to divorce the initial task of claims construction from the specific issues of infringement raised in any particular case, it is critical that the essential features, or "pith and marrow", of the claim be determined without reference to the specific variant in the allegedly infringing process (Camco at paragraph 49; Dableh v. Ontario Hydro, [1996] 3 F.C. 751 (C.A.) at pages 773-774). Consideration of the allegedly infringing variant will only take place once the essential scope of the patent has been determined.
[37] Construing a patent for the purposes of determining infringement is the next step in the analysis. With the Supreme Court's decision in Free World Trust, courts now have the benefit of a thorough road map to patent construction for these purposes. In that decision, Binnie J. set out a series of propositions (at paragraph 31) which must guide courts in determining infringement issues. These propositions are the following:
(a) The Patent Act promotes adherence to the language of the claims.
(b) Adherence to the claims in turn promotes both fairness and predictability.
(c) The claim language must, however, be read in an informed and purposive way.
(d) The language of the claims thus construed defines the monopoly. There is no recourse to such vague notions as the "spirit of the invention" to expand it further.
(e) The claims language will, on a purposive construction, show that some elements of the claimed invention are essential while others are non-essential. The identification of elements as essential or non-essential is made:
(i) on the basis of the common knowledge of the worker skilled in the art to which the patent relates;
(ii) as of the date the patent is published;
(iii) having regard to whether or not it was obvious to the skilled reader at the time the patent was published that a variant of a particular element would not make a difference to the way in which the invention works; or
(iv) according to the intent of the inventor, expressed or inferred from the claims, that a particular element is essential irrespective of its practical effect;
(v) without, however, resort to extrinsic evidence of the inventor's intention.
(f) There is no infringement if an essential element is different or omitted. There may still be infringement, however, if non-essential elements are substituted or omitted. [emphasis added]
[38] Proposition (e) underlined in the above excerpt from Free World Trust lies at the heart of the infringement analysis. That proposition recognizes that "[i]t would be unfair to allow a patent monopoly to be breached with impunity by a copycat device that simply switched the bells and whistles, to escape the literal claims of the patent" (Free World Trust at paragraph 55). Binnie J. succinctly explained the analysis necessary with respect to determining equivalence (at paragraph 55):
For an element to be considered non-essential and thus substitutable, it must be shown either (i) that on a purposive construction of the words of the claim it was clearly not intended to be essential, or (ii) that at the date of publication of the patent, the skilled addressees would have appreciated that a particular element could be substituted without affecting the working of the invention, i.e., had the skilled worker at that time been told of both the element specified in the claim and the variant and "asked whether the variant would obviously work in the same way", the answer would be yes: Improver Corp. v. Remington, supra, at p. 192. In this context, I think "work in the same way" should be taken for our purposes as meaning that the variant (or component) would perform substantially the same function in substantially the same way to obtain substantially the same result.
[39] Having reviewed the law of patent construction as comprehensively restated in the recent jurisprudence of the Supreme Court, it remains to determine whether the Motions Judge erred in applying that law as Apotex claims.
(B) The Application of the Law by the Motions Judge
[40] Although the Motions Judge did not have the benefit of the Supreme Court's recent guidance when he construed the Janssen Patent, I am of the opinion, and both counsel in argument agreed on this, that he anticipated and applied the general approach expressed in the propositions laid down by Binnie J. in Free World Trust. Having employed the correct general principles, the decision of the Motions Judge to accept the evidence of Janssen's experts as an aid to his construction of the claims is entitled to some deference. In the circumstances, I am not persuaded that any palpable and overriding error was made by the Motions Judge in this respect and I therefore see no reason to interfere with his conclusion as to infringement.
[41] Applying the purposive approach to patent construction, the Motions Judge construed the essentials of the Janssen Patent as follows (at paragraph 58):
In my view, on a purposive construction of the '847 [Janssen] Patent "and construing the claims with a mind willing to understand the true essentials of the invention", the true purpose of the first process of claim 1 is to make cisapride by the formation of the amide bond, and optionally, 0-methylation and/or preparing the cis isomer. With respect to claim 5 of the '847 [Janssen] Patent, it is my view that the claim claims cisapride when made or prepared by the first process of claim 1, i.e. formation of the amide bond, and optionally, 0-methylation and/or preparing the cis isomer, and by all other obvious chemical equivalents of the claim 1 process. [footnotes omitted]
Apotex criticises the Motions Judge for not viewing the intermolecular "condensation" reaction as an essential part of the claim. While the view of the Motions Judge clearly narrows the scope for Apotex to prove that its Torcan Process is materially different from that claimed in the Janssen Patent, it was a view that, in my opinion, was open to the Motions Judge on the evidence. Indeed, the Motions Judge noted at paragraph 50 of his reasons that Dr. McClelland agreed in cross-examination that the essence of the first process in claim 1 was the acylation of an amine to form a nitrogen carbonyl bond and that "the most important thing" of the condensation reaction was the formation of the bond between the nitrogen and the carbon of the carbonyl. This admission is consistent with the view propounded by Dr. Snieckus which the Motions Judge found to be credible and persuasive.
[42] I should note, however, that the construction of the claims adopted by the Motions Judge did not foreclose argument on the issue of infringement. The Motions Judge was careful to construe the essentials of the claims without reference to the particular variants in the Torcan Process. Only after he had completed his initial construction did he proceed to consider the issue of equivalence vis-à-vis the Torcan Process. His position that the process limitations in the claims were not intended to be essential was, in my view, nothing more than a recognition of the clear protection in subsection 41(1) of the Patent Act which extends to cover all obvious chemical equivalents of a process described and claimed in a patent. At the relevant time, subsection 41(1) read as follows:
|
41 (1) In the case of inventions relating to substances prepared or produced by chemical processes and intended for food or medicine, the specification shall not include claims for the substance itself, except when prepared or produced by the methods or processes of manufacture particularly described and claimed or by their obvious chemical equivalents. |
41(1) Lorsqu'il s'agit d'inventions couvrant des substances préparées ou produites par des procédés chimiques et destinées à l'alimentation ou à la médication, le mémoire descriptif ne doit pas comprendre les revendications pour la substance même, excepté lorsque la substance est préparée ou produite par les modes ou procédés de fabrication décrits en détail et revendiqués, ou par leurs équivalents chimiques manifestes. |
To this extent, claim 5 of the Janssen Patent claiming all obvious chemical equivalents to the processes claimed in claim 1 was superfluous. The allegation by Apotex that the Motions Judge improperly used claim 5 to broaden the scope of claim 1 misses the point that claim 1, by virtue of subsection 41(1) of the Patent Act, already contemplates obvious chemical equivalents to the processes it specifically describes and claims.
[43] In any event, the manner in which the Motions Judge framed the infringement issue suggests that his initial construction of the Janssen Patent may not have been as broad as Apotex alleges or as may appear from the language adopted by the Motions Judge himself in the excerpt quoted above. In this regard, it is perhaps worth repeating what he stated at paragraph 62 of his reasons:
It is undeniable, in my view, that the Torcan amide bond is made in a manner analogous to the '847 [Janssen] Patent amide bond, i.e. by way of a nucleophilic acyl substitution reaction involving the non-bonded electrons of the nitrogen attacking the carbonyl carbon, which results in the displacement of an appropriate leaving group Y, to form the nitrogen-carbonyl carbon bond. There are a number of variants between the '847 [Janssen] Patent and the Torcan Process, of which the principal is the use of an intramolecular acylation in lieu of an intermolecular acylation. The issue, therefore, is whether, by reason of these variants, the Torcan Process falls outside the scope of the '847 [Janssen] Patent.
[44] On the basis of this passage, the allegation by Apotex that the Motions Judge converted the Janssen Patent into a product per se claim to cisapride is not accurate. The Motions Judge went to the heart of the issue, as recognised by both parties in their submissions to and before this Court. He was clearly aware of the main issue being whether Torcan's intramolecular acylation reaction was materially different from Janssen's intermolecular acylation reaction. In this way, the Motions Judge was open to the possibility that an intramolecular acylation reaction could bring the Torcan Process outside the scope of the Janssen Patent.
[45] Furthermore, contrary to the position advocated by Apotex, the Motions Judge clearly recognized that there were other differences between the competing processes in addition to the form of acylation reaction involved. To that extent, it cannot be said that the Motions Judge ultimately insisted upon an erroneous rule that only similarities between the processes must be considered for infringement purposes. Although at one point in his reasons, he appears to have relied on such a rule, it is clear to me that he did not in fact apply that rule at the end of the day in deciding the application. Rather, the language used initially by the Motions Judge, in my opinion, flowed from his attempt to articulate his criticism of Dr. McClelland's evidence in failing to address, in a forthright manner, what he considered to be the key issue, namely the similarities and differences between the acylation reactions in each process. In the end, he examined both the similarities and the differences.
[46] Nor am I persuaded that the Motions Judge applied the wrong test for equivalence. The allegation by Apotex that the Motions Judge reduced equivalence to the notion of "effecting the same result" is nowhere substantiated on a reading of the decision under appeal. The Motions Judge applied the Catnic analysis as restated in Hoffmann J.'s three part test from Improver, supra. That test was endorsed by the Supreme Court in Free World Trust. The Motions Judge spoke of an equivalent as a process which "performs the same function in substantially the same manner" as the claimed process (at paragraph 60). He also spoke of equivalence in terms of working in a similar manner to achieve the same result (at paragraph 70). This hardly amounts to defining an equivalent as a process which merely effects the same result as the claimed process. Indeed, the definition stated by the Motions Judge captures the essence of the Supreme Court's notion of equivalence as arising from the performance of substantially the same function in substantially the same way to obtain substantially the same result (Free World Trust at paragraph 55).
[47] Applying the correct test for equivalence, the Motions Judge determined that Torcan's intramolecular acylation reaction was an obvious chemical equivalent of the intermolecular acylation reaction described in claim 1 of the Janssen Patent. I am not persuaded by the argument raised by Apotex that the Motions Judge misunderstood the concept of "obviousness" in arriving at his conclusion. In my view, he did understand the test for obvious chemical equivalence as stated by Hoffmann J. in Improver and repeated by Binnie J. for the Supreme Court in Free World Trust, and he applied it properly.
[48] Specifically, the question of equivalence supposes that the person skilled in the art is told of both the invention and the variant and asked whether the variant would obviously work in the same way (Improver at page 192, Free World Trust at paragraph 55). Apotex has ignored the particulars of this test in order to argue that the skilled person must view the variant as being obvious without being told of the variant's existence. This is not the correct approach but approximates the analysis for determining the obviousness of an invention in order to decide the issue of patent validity. However, the Motions Judge was not concerned with the inventiveness or validity of the Janssen Patent. He was concerned with deciding the issue of infringement, and he applied the correct notion of obviousness for the purpose of determining equivalence.
[49] In doing so, the Motions Judge concluded on the evidence that the intramolecular acylation reaction in the Torcan Process obviously works in the same way as the intermolecular acylation reaction described in the Janssen Patent. In other words, the introduction of an intramolecular context for the very same reaction, involving the very same reactive groups, was not a sufficient difference to take the Torcan Process outside of the intended scope of claim 1. This conclusion was, in my opinion, open to the Motions Judge on the evidence. He relied on evidence to the effect that intramolecular versions of intermolecular reactions should be considered by synthetic organic chemists when planning a synthetic pathway and to the effect that Torcan's intramolecular reaction is well known and has been performed since 1937.
[50] In my view, the Motions Judge was justified, given all the evidence presented to him, in preferring the position of Dr. Snieckus to that of Dr. McClelland. In view of the construction given to claim 1 by the Motions Judge, he was clearly interested at the infringement stage of the analysis to compare the acylation reactions in each of the processes. For whatever reason, Dr. McClelland failed to address, to the satisfaction of the Motions Judge, what was occurring in the Torcan Process during the critical acylation reaction. It is not enough to say, as Apotex has done on appeal, that no weight should have been given to the description Dr. Snieckus gave of Torcan's acylation reaction because intermediates 7a and 7b as identified by Dr. Snieckus are merely hypothetical intermediates which may or may not exist in the Torcan Process. If those intermediates did not exist in the Torcan Process, it was up to Apotex, through its experts, to offer something in reply, rather than to rely on the uncertainty which flowed from a lack of analysis. In the circumstances, the Motions Judge found the analysis of Dr. Snieckus to be persuasive. It was within his discretion to do so.
[51] Apotex makes much of the decision by Dr. Snieckus to change his definition of the condensation reaction employed by Janssen in its claimed process. There is little doubt that Dr. Snieckus regretted using the term "condensation" in his affidavit, and the pointed cross-examination of Mr. Radomski successfully dramatised this point. However, this fact does not materially affect the outcome of the case. It was recognised both by Dr. Snieckus and the Motions Judge that the acylation reaction specifically described in claim 1 of the Janssen Patent occurred as an intermolecular condensation reaction in the narrow sense of the term "condensation". It was similarly recognised that the Torcan Process involved an intramolecular acylation that did not result in the loss of a smaller water or alcohol molecule. The fact that the Janssen acylation reaction occurs by condensation reaction does not demonstrate that the Torcan Process is not infringing. It merely requires that there be consideration of the possible equivalence between the intra- and intermolecular forms of amide bond formation. In truth, this was the key battleground between the parties.
[52] The attempt by Dr. Snieckus to correct the imprecision in his earlier use of the term "condensation" was made in order to demonstrate his opinion that the essence of an acylation reaction, whether by intra- or intermolecular means, is the fact of a nucleophilic acyl substitution also known as a nitrogen carbonyl bond-forming reaction or an amide bond formation. This opinion was clearly explained in cross-examination (see Cross-examination of Dr. Snieckus, Appeal Book, Vol. III, Tab 63 at 662-702). While the use of the term "condensation" in the affidavits sworn by Dr. Snieckus made for some awkward moments for him in cross-examination, it is, in my view, arguable, from the explanations provided by Dr. Snieckus, that he did not make a fundamental error which impugns his evidence. Accordingly, the Motions Judge was entitled to accord weight to that evidence and to prefer it over that of Dr. McClelland.
[53] As to the independence of Dr. Snieckus, the only evidence in the record regarding the preparation of his affidavits indicates that Counsel for Janssen prepared the first drafts of the affidavits after consultation with Dr. Snieckus. There is nothing unusual in this practice, as long as the evidence is that of the affiant, not the lawyer. I have not been convinced that the credibility of the evidence adduced by Dr. Snieckus was fatally undermined in the circumstances, though it may have been weakened.
[54] This was a complex case, which was well-argued by Counsel for both sides. In the end, the Motions Judge, in my view, applied the law correctly to very difficult facts as he found them. He is entitled, in my opinion, to a degree of latitude in determining the factual issue of obvious chemical equivalence. Having reviewed the record, I am not prepared to find any palpable and overriding error on the part of the Motions Judge in assessing the evidence and arriving at the conclusion that the Torcan Process falls within the scope of the Janssen Patent.
Disposition
[55] For the reasons given above, I would dismiss the appeal. The respondents Janssen Belgium and Janssen Canada are entitled to one set of costs between them.
"A.M. Linden"
J.A.
"I agree
Julius A. Isaac J.A."
"I agree
B. Malone J.A."