Janssen-Ortho Inc. v. Novopharm Ltd., 2004 FC 1631

Date: 20041119

Docket: T-214-03

Citation: 2004 FC 1631

Ottawa, Ontario, this 19^th day of November, 2004

Present:           The Honourable Mr. Justice Mosley                                   

BETWEEN:

                                                     JANSSEN-ORTHO INC. and

DAIICHI PHARMACEUTICAL CO., LTD.

                                                                                                                                           Applicants

                                                                           and

                                                    NOVOPHARM LIMITED and

THE MINISTER OF HEALTH

                                                                                                                                      Respondents

                                            REASONS FOR ORDER AND ORDER

[1]                Janssen-Ortho Inc., and Daiichi Pharmaceutical Co. Ltd., seek an order to prohibit the Minister of Health from issuing a Notice of Compliance to Novopharm Limited, regarding the marketing of tablets of an antimicrobial drug known generically as "levofloxacin" until after the expiry of Daiichi's levofloxacin patent. Janssen-Ortho is licensed by Daiichi to sell the drug in Canada under the brand name LEVAQUIN. Novopharm alleges that the patent is invalid for lack of novelty, obviousness, ambiguity, overbreadth and insufficiency of specification. Novopharm further asserts that its version of levofloxacin would not in any event infringe any of the patent's claims. The substantive issue in these proceedings is whether Novopharm's allegations of non-infringement and invalidity of the patent are justified.

BACKGROUND

The Parties

[2]         Janssen-Ortho is the "first person" referred to in section 2 and subsection 4(1) of the Patented Medicines (Notice of Compliance) Regulations, SOR/93-133, (the "NOC Regulations"). Pursuant to subsection 6(4) of the NOC Regulations, Daiichi Pharmaceutical, as the owner of the patent, was made a party to this proceeding, however, as per subsection 6(1) of the NOC Regulations, the "first person", that is Janssen-Ortho, initiates and seeks relief in this prohibition proceeding. Janssen-Ortho has listed the patent at issue on the Patent Register in relation to its NOCs for LEVAQUIN. It markets the drug in 250 mg and 500 mg tablet form.

[3]                The respondent, Novopharm, is the "second person" cited in section 2 and subsection 5(1) of the NOC Regulations. By letter dated December 20, 2002, Novopharm sent to Janssen-Ortho a Notice of Allegation ("NOA"), including a detailed statement of the legal and factual basis for its allegation, pursuant to paragraph 5(3)(c) of the NOC Regulations. It seeks a NOC to market 250 mg and 500 mg tablets of levofloxacin hemihydrate.

[4]                The respondent, the Minister of Health (the "Minister"), is responsible for the review of applications by drug companies for the issuance of NOCs under the Food and Drug Regulations, C.R.C. 1985, c. 870. The Minister did not file materials in this application.

The Patent at Issue

[5]                To understand the invention claimed in the patent at issue, certain chemistry terms and relationships must be understood. I have had the benefit of counsel's assistance and the expert affidavits filed by both sides. Janssen-Ortho's primary expert, Dr. Klibanov, is a Professor of Chemistry and Bioengineering at the Massachusetts Institute of Technology. Novopharm's expert, Dr. Caldwell, is the Dean of the Faculty of Medicine at the University of Liverpool. Both experts have extensive credentials and background in pharmacology and organic chemistry. For convenience, definitions of the terms employed in these reasons, based on my understanding of the evidence, are set out in Annex "A" .

[6]                The patent at issue, Canadian Patent 1, 304, 080 (the " '080 patent") was granted to Daiichi Pharmaceutical on June 23, 1992. The filing date of the '080 patent is June 19, 1986 and it will expire on June 23, 2009.

[7]                Daiichi is also the owner of Canadian Patent 1, 157, 840 (the " '840 patent"), issued on May 22, 1984, which disclosed and claimed an antibiotic known as "ofloxacin". Janssen-Ortho was also licensed by Daiichi to market ofloxacin in Canada, which it did under the brand name FLOXIN .    The '840 ofloxacin patent expired on May 22, 2001.

Levofloxacin and Ofloxacin

[8]                Levofloxacin, the subject of the '080 patent and these proceedings, and ofloxacin, disclosed in the expired '840 patent, can both be described as broad-spectrum, synthetic antibacterial agents useful in the treatment of infections caused by certain strains of microorganisms, such as certain types of pneumonia, influenza, skin infections and urinary tract infections. Ofloxacin and levofloxacin formulas have also been marketed for eye infections under different brand names.

[9]                Ofloxacin is a racemic compound, that is a substance containing equal amounts of two optical isomers. An isomer is one of a number of molecules that have the same constituent atoms; however, the bonds between the atoms are oriented differently in space. Their optical quality refers to their ability to rotate the plane of polarized light, that is, light focussed to shine only in one direction. Optical isomers that rotate polarized light to the right, in a clockwise direction, are "dextrorotatory" and use the symbols (+) or (d) for "dextro". Optical isomers that rotate polarized right to the left, in a counter-clockwise direction, are "levorotatory" and use the symbols (-) or (l) for "levo".    The drug at issue here is the S(-) optical isomer of ofloxacin, hence "levofloxacin".

[10]            Another term for describing optical isomers is "enantiomer" . An enantiomer is one of a pair of non-superimposable mirror images; by analogy, one's hands. Compounds that exhibit this quality are called "chiral". They possess a chiral centre and in carbon compounds, such as ofloxacin, the chiral centre is a carbon atom that is bound to four different atoms or groups of atoms, where all four groups are different. Chirality confers a specific three-dimensional shape on a molecule.

[11]            When produced, levofloxacin was discovered to possess beneficial properties over both racemic ofloxacin and the R(+) optical isomer. The specification of the '080 patent, under the heading "Background of the Invention" at pages 1-2, discloses the beneficial properties of the S(-) enantiomer of ofloxacin (levofloxacin) in contrast to the racemic ofloxacin and the R(+) compound, as follows:

... The present inventors obtained optically active compounds of the racemic Ofloxacin and found that the S(-)-compound possesses an antimicrobial activity of about 2 times higher than that of the (")-compound and acute toxicity LD₅₀ weaker than that of the (")-compound as determined in mice by intravenous administration. On the other hand, the present inventors found that the R(+)-compound exhibits an antimicrobial activity of only about 1/10 to 1/100 times that of the (")-compound, whereas it possess an acute toxicity substantially equal to that of the (")-compound. That is, the S(-)-form of Ofloxacin has been found to have very desirable properties, i.e., increased antimicrobial activity and reduced toxicity, and is expected to be a very useful pharmaceutical agents [sic] as compared with the (")-compound. Further, both the R(+)- and S(-)-compounds of Ofloxacin in the free form have markedly high water-solubility as compared with the (")-compound and as compared with free compounds of this type, and can be used as injectable preparations. These advantages will be apparent from the experimental data shown hereinafter.

ISSUES

[12]            1. Burden of proof with respect to patent validity.

2. How should the '080 patent be construed?

3. Is Novopharm's allegation that the '080 patent is invalid justified because of

a) Obviousness;

b) Anticipation; or

c) Other allegations of invalidity?

4. Is Novopharm's allegation that the '080 patent will not be infringed by its product justified?

ANALYSIS

Burden of Proof with Respect to Patent Validity

[13]               It is well-established that the legal burden of proving that the second person's allegations are not justified is on the person seeking the prohibition order, namely Janssen-Ortho. Janssen-Ortho must establish, on a balance of probabilities, that Novopharm's allegations are not justified. While Janssen-Ortho bears the overall legal burden, Novopharm has the evidential burden to put "in play" the allegations set out in its NOA and its detailed statement: see Merck Frosst Canada Inc. v. Canada (Minister of National Health and Welfare) (1994), 55 C.P.R. (3d) 302 (F.C.A.), Pharmacia Inc. v. Canada (Minister of National Health & Welfare) (1995), 60 C.P.R. (3d) 328 (F.C.T.D.), aff'd (1995) 64 C.P.R. (3d) 450 (F.C.A.), and Bayer Inc. et a. v. Minister of National Health and Welfare et al. (2000), 6 C.P.R. (4^th) 285 (F.C.A.). The facts contained in the NOA and detailed statement are presumed to be true, at least with respect to allegations of non-infringement, except to the extent that the contrary has been shown by the applicant: see Merck Frosst, supra, at 319.   

[14]            Janssen-Ortho accepts that with respect to infringement, the legal burden is clearly its own. However, with respect to validity, the applicant submits that the burden of proof in a NOC proceeding shifts to the respondent due to the presumption that the patent is valid pursuant to section 45 of the pre-1989 Patent Act. The parties agree that as the application for the '080 patent was made in 1986, the Patent Act, R.S.C. 1985, c. P-4 applies.

[15]            The applicant relies on Merck Frosst, supra in support of this contention, referring to the following words of Hugessen J.A. at page 319:

As I understand the scheme of the regulations, it is the party moving under s. 6, in this case Merck, which, as the initiator of the proceedings, has the carriage of the litigation and bears the initial burden of proof. That burden, as it seems to me, is a difficult one since it must be to disprove some or all of the allegations in the notice of allegation which, if left unchallenged, would allow the Minister to issue a notice of compliance. There may, of course, be some presumptions (such as for example the statutory presumption of validity of a patent) (Patent Act, s. 43) which may help the moving party and have the effect of displacing the burden of proof. ...

[16]            The applicant also relies on the Supreme Court of Canada decisions of 2004 SCC 34">Monsanto Canada Inc. v. Schmeiser, 2004 SCC 34, (2004) 320 N.R. 201 and Apotex Inc. v. Wellcome Foundation Ltd., [2002] 4 S.C.R. 153, arguing that they indicate that the burden of demonstrating invalidity is on the person attacking the patent, that is the second person, or respondent, in a NOC proceeding. The applicant also says that the language used by the Supreme Court in these decisions indicates that the determination of invalidity can only be made after the Court is satisfied that the Commissioner of Patents was "clearly wrong", or made an unreasonable decision in granting the patent in the first place. The applicant submits that the reasoning of the Supreme Court in these patent actions applies equally to NOC proceedings.

[17]            The respondent Novopharm refers to the 2000 Federal Court of Appeal decision in Bayer, supra, where Sharlow J.A. commented on the statutory presumption of validity and how this affected the burden of proof in a NOC proceeding. Novopharm submits that the presumption of validity is spent when the respondent puts forward evidence to the contrary and if, in the end, the evidence on both sides is a "draw", then the applicant cannot be granted a prohibition order, as the overall legal burden, on a balance of probabilities, was its to carry. At paragraph 6 of Bayer, Sharlow J.A. stated as follows:

In seeking to discharge its burden of proving the allegation to be unjustified, Bayer relied on the statutory presumption of the validity of its patent. Because that presumption exists, it may be said that Apotex, as the party responding to the application for a prohibition order, has a burden of proof in this sense: if Apotex had adduced no evidence that was capable of establishing the invalidity of the patent, Bayer could have succeeded on the basis of the statutory presumption alone. As the Motions Judge correctly said at paragraph 15:

A statutory presumption, for example, may assist Bayer and "have the effect of displacing the burden of proof".

[emphasis added]                         

[18]       Justice Sharlow noted that the second person in that case had adduced evidence in the form of an affidavit. She then observed, at paragraph 9, that rebutting the statutory presumption of validity will depend on the strength of the evidence going to invalidity:

The operation of the statutory presumption in the face of evidence of invalidity depends upon the strength of the evidence. If the evidence proves on a balance of probabilities that the patent is invalid, the presumption is rebutted and is no longer relevant: Diversified Products Corp. v. Tye-Sil Corp. (1991), 35 C.P.R. (3d) 350 (F.C.A.) at p. 359.

[19]            The applicant replies that Bayer is not authority for the proposition that as soon as the respondent files a "speck of evidence", the burden shifts to the applicant on the issue of validity in a NOC proceeding. Janssen contends that the evidence of the second person must demonstrate that the patent is invalid on a balance of probabilities before the statutory presumption is rebutted and no longer relevant.

[20]            I read the Bayer decision as indicating that a second person's evidence must not be clearly incapable of sustaining its allegations, and if it meets that threshold, then the statutory presumption will be spent and cannot aid the applicant for the purpose of a NOC summary proceeding. This interpretation is supported by the language of section 45 of the pre-1989 Patent Act, that the presumption is to apply "in the absence of any evidence to the contrary". I note, however, that simply succeeding in rebutting the presumption does not mean that a second person will be successful in a prohibition application, as it must still put sufficient evidence "in play" to challenge the validity of the patent as well as the applicant's expert evidence, while the overall legal burden remains with the applicant.

[21]            The applicant's reliance on the Supreme Court's reasoning in the 2004 SCC 34">Schmeiser and Wellcome decisions, with regard to the burden of proof, is not helpful in my view, as both were patent actions for declarations of invalidity and infringement. I do not agree that such reasoning is directly applicable in a NOC proceeding, which has been held to be akin to a summary proceeding by way of an application, involving its own unique set of principles and regulations, and where any finding regarding infringement or validity is not determinative of that matter in any subsequent patent action. As held in Merck Frosst, supra, at 319, a NOC proceeding is not an action and its object is "solely to prohibit the issuance of a notice of compliance under the Food and Drug Regulations."

Construction of the '080 patent

[22]            At the heart of the dispute between the parties as to the construction of the '080 patent is whether it claims the levorotatory optical isomer, as found or already disclosed within racemic ofloxacin (Novopharm's position) or claims such enantiomer as a separate, inventive compound (Janssen's position). Related to that question is the issue of whether "S(-)" as used to describe the levo-enantiomer of ofloxacin can be construed to mean the S(-) isomer in a substantially pure form.

[23]            As held by the Supreme Court of Canada in Whirlpool Corp. v. Camco Inc., [2000] 2 S.C.R. 1067 and Free World Trust v. Électro Santé Inc., [2000] 2 S.C.R. 1024, a patent claim is to be construed in a "purposive" way, which involves a realistic construction of the language of the claims through the eyes, and with the learning, of a person skilled in the art (para. 44 Whirlpool, supra). Moreover, the Court may look at the entire specification of the patent in order to understand the words as stated in any disputed claim, provided this does not enlarge or contract the scope of the actual patent claims.

[24]            Claim construction remains, in the end, a question of law for the Court to decide: Whirlpool, supra, and Beecham Canada Ltd. v. Procter & Gamble Co. (1982), 61 C.P.R. (2d) 1 (F.C.A.). The key to purposive construction, as articulated by Justice Binnie in Whirlpool, supra, is for the Court to identify the particular words or phrases in the patent claims that describe what the inventor considered to be the "essential elements of his invention" (para. 45).

[25]            There are 19 claims of the '080 patent. Claims 1, 3, 5, 10 and 11 claim various processes to make levofloxacin and related compounds. The parties agree that the process claims are not at issue in these proceedings. Claims 2, 4, 6, 12-17 and 19 claim, among other substances, levofloxacin and/or its salts and hydrates. Claims 7, 8, 9 and 18 relate to pharmaceutical formulations which contain levofloxacin, including its salts and hydrates.

[26]            Janssen argues that a purposive construction of the relevant claims, including a reading of the specifications, discloses that the invention here is the recognition of the unexpected and beneficial properties that are present in the S (-) enantiomer and where the claim can be properly read as covering that enantiomer, the court should support its validity.

[27]            Novopharm contends that all of the claims of the patent to a S (-) compound - without the limitation that it be pure, partially purified or isolated, or unaccompanied by the R (+) compound - cover the racemate, ofloxacin. The drafter of the claims erred, it argues, in not incorporating any purity limitation. Accordingly, Novopharm contends that the claims as written read on the racemate and there is nothing novel in the compounds claimed and the claims cannot be saved by the patent disclosure.

[28]            In my view, the claims of the '080 patent would be read by a person skilled in the art as referring to the levorotatory isomer of ofloxacin and not to racemic ofloxacin. In claims 1, 2 and 10 of the '080 patent the phrase "S(-)... compound" is employed, as are solid-coloured wedges in the graphical depictions of the formulas for the compounds in the claims. The use of the "S" and "(-)" symbols and the solid-coloured wedges designates, from what the expert affidavit evidence has revealed in this proceeding, an optically active isomer with an "S" configuration rotating the plane of polarized light in a counter-clockwise direction. Racemic ofloxacin would not be described using the S(-) prefix. Moreover, the specification of the '080 patent confirms this construction of the claims. The disclosure refers to "optically active compounds of Ofloxacin" or the "racemic Ofloxacin" in contrast to the description of the S(-) enantiomer.

[29]            Having reviewed the claims with the aid of the expert evidence in this proceeding, I am satisfied that the patent disclosure must form part of the construction to give the Court an understanding of the meaning behind the chemical formulae in the claims. This disclosure indicates that the antimicrobial activity of the optical (-) and (+) isomers of ofloxacin against various microorganisms was compared to that of the ofloxacin racemate. This demonstrates, in my view, that discovery of the beneficial properties in the S(-) optical isomer was the object and usefulness of this patent.

[30]            With respect to the question of whether "S(-)" - as used to describe the levo-enantiomer of ofloxacin in the '080 patent - can be construed to mean S(-) in a substantially pure form, I accept that proposition. A chemist skilled in the art when reading "S(-)-pyridobenzoxazine compound" as described in a particular formula would, in my opinion, accept that such compound is principally made up of the S(-) enantiomer of a racemic compound, wherein the optical activity of the S(-) enantiomer is not only detectable, but dominant. The lack of numeric quantification of the purity level of the S(-) enantiomer is not fatal to distinguishing this compound from either the racemic compound or the R(+) enantiomer.

[31]            This construction of the '080 patent is similar to that employed by Judge Keeley of the U.S. Northern District Court of West Virginia in Ortho-McNeil Pharmaceutical Inc., et. al. v. Mylan Laboratories, Inc. et al., 267 F. Supp. 2d 533 (N.D. W. Virg. 2003), where the U.S. patent^[1] to levofloxacin was challenged. While this Court is in no way bound by the Mylan decision as a precedent and is conscious of the differences between Canadian and U.S. patent law, the reasoning was of general use to this Court in construing the '080 patent.

[32]            The existence of such enantiomers of racemic ofloxacin seems undoubtedly accepted prior to 1984 and it was the development of a process for producing the S(-) enantiomer and the discovery of the beneficial properties of this enantiomer over racemic ofloxacin or the R isomer that are claimed in the '080 patent. Whether the invention claimed was obvious or anticipated, or was improperly set out in the patent claims and is vulnerable on other grounds of invalidity, are the issues dealt with in my analysis below.

Invalidity

[33]            As noted above, the parties agree that the Patent Act, R.S.C. 1985, c. P-4 governs this Court's analysis. This version of the Patent Act is commonly referred to as the "pre-October 1, 1989" Patent Act. Subsections 27(1) and 61(1) of that Act are relevant and provide:

Obviousness

[34]            Turning first to obviousness, the Federal Court of Appeal in Diversified Products Corp. v. Tye-Sil Corp. et al. (1991), 35 C.P.R. (3d) 350 (F.C.A.) described the requirement that an invention not be obvious at page 365:

There is no specific section in the Patent Act relating to the requirement for inventiveness or inventive ingenuity, but it has been held and is no longer questioned that by use of the words "invention" and "inventor" throughout the Act, inventiveness or inventive ingenuity is required to obtain a valid patent. ...

[35]            The commonly accepted test for obviousness was set out in Beloit Canada Ltd. v. Valmet Oy, (1986),8 C.P.R. (3d) 289 (F.C.A.) by Hugessen, J.A. as follows at page 294:

The test for obviousness is not to ask what competent inventors did or would have done to solve the problem. Inventors are by definition inventive. The classical touchstone for obviousness is the technician skilled in the art but having no scintilla of inventiveness or imagination; a paragon of deduction and dexterity, wholly devoid of intuition; a triumph of the left hemisphere over the right. The question to be asked is whether this mythical creature (the man in the Clapham omnibus of patent law) would, in the light of the state of the art and of common general knowledge as at the claimed date of invention, have come directly and without difficulty to the solution taught by the patent. It is a very difficult test to satisfy.

[Emphasis added]

[36]            In determining whether a patent claim is obvious, the Court must avoid using the benefit of hindsight. The question to ask is whether the solution taught by the patent would be "plain as day" to the skilled technician who was searching for something novel, without having to do experimentation or research. Suggestions or signposts in the prior art are not sufficient to make a patent invalid for obviousness: Apotex Inc. v. Wellcome Foundation Ltd. (1998), 79 C.P.R. (3d) 193 (F.C.T.D.), varied but not on the issue of obviousness, [2001] 1 F.C. 495 (C.A.), aff'd [2002] 4 S.C.R. 153, Bayer Aktiengesellschaft v. Apotex Inc. (1995), 60 C.P.R. (3d) 58 (Ont. Gen. Div.), varied on other grounds (1998), 82 C.P.R. (3d) 526 (O.C.A.), leave to appeal to the Supreme Court of Canada denied [1998] S.C.C.A. No. 563(QL) and [1979] 2 S.C.R. 929">Fabwerke Hoechst v. Halocarbon (Ont) Ltd., [1979] 2 S.C.R. 929. To refer to another oft-quoted passage from Beloit, supra, at page 295:

Every invention is obvious after it has been made, and to no one more so than an expert in the field. Where the expert has been hired for the purpose of testifying, his infallible hindsight is even more suspect. It is so easy, once the teaching of a patent is known, to say, "I could have done that"; before the assertion can be given any weight, one must have a satisfactory answer to the question, "Why didn't you?"

[37]            With obviousness, the invention need not be disclosed in one single patent or piece of prior art, as is the case for anticipation. The Court is entitled to look at all the patents and other publications that a skilled technician would discover in a "reasonable and diligent search" to determine whether the resulting "mosaic" leads directly to the invention: Illinois Toolworks Inc. v. Cobra Fixations Cie. (2002), 221 F.T.R. 161, aff'd on this point, varied only with respect to costs: (2003), 312 N.R. 184 (F.C.A.).

[38]            The critical date for assessing obviousness is the claimed date of the invention. The '080 patent claims priority from three Japanese patent applications. The respondent submits that the applicant has not led evidence to establish the date of invention; therefore, the relevant date for evaluating obviousness must be taken as being no earlier than the earliest possible priority date, one year prior to the Canadian patent application, June 20, 1985. The applicant conceded this at the hearing and I accept that this is the correct date in light of sections 28 and 29 of the Patent Act as it read prior to October 1, 1989.

[39]            Further, in evaluating the prior art on the question of obviousness, such art must have been in the public domain. The applicant contends that only art that could be located in a "reasonably diligent search by a skilled workman" should be considered.

[40]            The respondents Notice of Allegation dated December 20, 2002, cites 26 alleged prior art references with respect to obviousness including the '840 ofloxacin patent described above, the corresponding U.S. ofloxacin patent^[2], U.S. and Japanese patents relating to other racemic compounds with antimicrobial properties and several academic articles and presentations, notably:

Repta et al., "Utilization of an Enantiomer as a Solution to a Pharmaceutical Problem:Application to Solobilization of a 1,2-Di(4-piperazine-2.6-dione) propane", J.Pharm.Sciences, Vol. 65(2), 238 (1976).

Gerster et al., "Proceedings of the North American Medicinal Chemistry Symposium", Toronto (Canada), 153 (1982.

Gerster et al., Poster Presentation to the North American Medicinal Chemistry Symposium, Toronto, Canada June, 1982.

[41]            Counsel for Janssen takes issue with the fact that Novopharm has not established how the cited references were located or how any of the references would have been reasonably located without the benefit of hindsight. In particular, Janssen refers to the two 1982 Gerster references which one of its affiants, Dr. Hans Schroeder, could not locate in any Canadian library, despite about two days of searching. Since Novopharm has not put forward any evidence suggesting how these references could be or were obtained, the applicant argues that they are not valid prior art references.

[42]            Novopharm maintains that the Gerster references were publicly available in 1982. Both relate to a paper Gerster presented at a 1982 Toronto four-day symposium sponsored by three major professional societies. One of these references was subsequently cited in the list of references in a 1985 article in the Journal of Antimicrobial Chemotherapy and also in the list of references of U.S. Patent No. 5,053,407.

[43]            With respect to the remainder of the references cited by Novopharm, Janssen maintains that they do not disclose the beneficial properties of the S(-) optical isomer, since these were only ascertainable through separating the isomer from the racemate and performing multiple experiments with the S(-) and R(+) enantiomers and racemic ofloxacin. The applicant submits that it was only after the S(-) optical isomer was "made" and then experimentation performed, that the invention, that is levofloxacin as a separate compound having certain benefits, was discovered. This discovery has enough inventive ingenuity to demonstrate that the patent claims are not obvious.

[44]            The applicant refers to an article written by Novopharm's expert, Dr. Caldwell, entitled "Putting Chirality to Work: The Strategy of Chiral Switches", wherein he recognized that in pairs of enantiomers in racemic mixtures it is uncertain which enantiomer is the "antagonist" or "active" one. Janssen stresses, relying on Bayer Aktiengesellschaft, supra that the test for obviousness does not involve testing or experimentation, but rather the mythical, skilled technician is expected to "instantly and spontaneously" know the answer which the patent teaches from the state of knowledge in the art at the date of invention.   

[45]            The respondent, also relying on Bayer Aktiengesellschaft, supra, submits that the applicants' interpretation of the test for obviousness is incorrect since the case law establishes that the notional person skilled in the art is entitled to carry out routine investigations including chemical analyses directed to merely determining characteristics of a known compound. Novopharm describes the tests done to determine the relative pharmacological characteristics of known enantiomers of racemic compounds as in the nature of "mere verification" of inherent, already known attributes of the compound, rather than in the nature of "searching for something novel". This "searching for something novel" must form part of the inventive process, as per Bayer Aktiengesellschaft, supra, and such inventiveness is lacking in the '080 patent.   

[46]            Novopharm refers to the Federal Court of Appeal decision of SmithKline Beecham Pharma Inc. v. Apotex Inc. (2002), 21 C.P.R. (4^th) 129 (F.C.A.) where the prior art disclosed that the drug could be formulated by conventional methods, and that "mechanical skill" rather than inventive genius was required in order to arrive at the claims in the impugned patent; the "improvements" noted in the patent's disclosure merely verified the properties of a known substance formulated using common techniques.

[47]            The '080 patent can be described as a "selection patent" in that levofloxacin was selected out of its class of substances - the racemate and the two enantiomers - because of its beneficial properties. The following passage from H.G. Fox in The Canadian Law and Practice Relating to Letters Patent for Inventions, 4^th Ed., (Toronto: Carswell Company Ltd., 1969), pp.89-90, describes inventiveness in the context of selection patents:

Invention may be exercised by selecting one out of a number of substances for a particular purpose even though others of that class have been used before for the same purpose, provided there is a special advantage to be derived from the use of the selected substance and its selection constitutes a definite advance upon existing knowledge. While one who merely picks out a number of items from an already disclosed group or series has not invented anything, yet it may be otherwise if his researches have led him to the discovery that certain items in the group or series possess qualities or characteristics peculiar to themselves and hitherto unknown.

[Emphasis added]

[48]            Novopharm relies on the following passage from Fox, supra at page 90:

Selection patents are more usually met with in the chemical, than in the other arts. If for practical purposes it is not obvious to skilled chemists in the state of chemical knowledge existing at the date of a selection patent that the selected components possess a special property, there is then, or at least there may be, a sufficient inventive step to support the patent. But mere verification is not invention. Where the method of manufacture is laid down in the originating patent, the selection patent must not be an exact reproduction of the same process coupled with a statement of the properties possessed by the selected bodies. No man can have a patent merely for ascertaining the properties of a known substance.

[Emphasis added]

[49]            Novopharm submits that the '080 patent does not meet the requirements for a selection patent. Adopting Professor Fox's words, it argues that mere verification is not invention and that a selection patent must not reproduce the same process with a statement of the properties possessed by the selected bodies. Unlike the patent at issue in the case of Re E.I. DuPont Nemours & Co. Application, [1982] F.S.R. 303 (H.L.), the respondent asserts that the '080 patent does not make a discovery of a selected entity that has different valuable properties in a different field.

[50]            Janssen's position is that section 32 of the Patent Act permits a patent based on improvements to any patented invention and there was such an inventive step in ascertaining that levofloxacin possesses special properties in comparison to the racemate, ofloxacin.

[51]            What Fox and the jurisprudence suggest, in my view, when dealing with selection patents, is that in order to demonstrate sufficient inventive ingenuity, there must be discovered some special advantage or quality, previously unknown, or an advantage discovered for a new use, which constitutes a definite advance upon the knowledge already existing with regard to the original group or series. In Pfizer Canada Inc. v. Apotex Inc. (1997), 77 C.P.R. (3d) 547 (F.C.T.D.), Richard J. (as he then was), found a selection patent valid that claimed the pharmaceutical preparation of fluconazole, the originating patent having claimed "an enormous class of compounds".

[52]            In that case, the compound described in the subsequent patent was not described in the original patent, and neither were "its superior and previously unknown efficacy" previously known. The Court there relied upon expert evidence that described the surprising features of fluconazole, and the discovery that it was soluble in water as compared to the earlier disclosed compound that was virtually insoluble. Justice Richard concluded that since fluconazole had "unexpected and valuable properties which are not possessed by the structurally closest compounds disclosed" in the original, prior patent, it was not invalid on the grounds of either obviousness or anticipation.

[53]            In my view, in the case at hand, such reasoning is applicable, but to bring about the opposite result. The beneficial properties discovered and set out in the '080 patent were not unknown. They were the same beneficial properties found in ofloxacin, only to a greater degree of effectiveness. Further, I find that knowledge of the existence of and the possibility of separating the two enantiomers of ofloxacin was common to the ordinary chemist, and the determination of which enantiomer possessed a greater amount of the same benefits of the previously known racemic antibiotic was not surprising or inventive. This determination was made from one of three possibilities, not a great number of possibilities; that is, either the S (-) enantiomer, R (+) enantiomer or the racemate would exhibit better qualities, in comparison to each other, relating to the properties of antimicrobial activity, toxicity or solubility.

[54]            With regard to Janssen's contention that if experimentation or testing is required in order to determine the invention, the resulting invention cannot be said to be obvious, I am persuaded by Novopharm's submission that the test for obviousness does not exclude routine testing to determine characteristics of known compounds, not undertaken for the purpose of "searching for something novel", but rather for the purpose of verifying the actual attributes of already known compounds, where the results indicate no new uses or surprising results, or properties that are clearly superior to the already known parent compound. Concerning the issue of determining whether there has been an inventive step or "undue experimentation" in the obviousness inquiry, see the reasoning of Justice Dawson of this Court in Pfizer Canada Inc. v. Apotex Inc. (2002), 22 C.P.R. (4^th) 466 (F.C.T.D.) at paras 103-114.            

[55]            As noted above, the respondent relied upon several U.S. patents filed well prior to 1985 among the 26 references cited as prior art.

[56]            U.S. Patent 3,917,609 disclosed tricylic fluorquinolone, an antibacterial compound, and indicated that one of the optically active isomers was much more active as an antibacterial agent. U.S. Patent 3,976,651 disclosed another tricylic fluorquinolone that had found one isomer to have more antimicrobial activity than the other. According to Dr. Caldwell, these are racemic compounds similar in structure to ofloxacin.   

[57]            Dr. Caldwell, at paragraphs 72-73 of his affidavit, refers to U.S. Patent 4,314,081, which states that with respect to the racemic compound in question, it "can be resolved into its individual d and l isomers by methods well known in the art, particularly, by forming salts with optically active acids and separating the salts by crystallization." This '081 patent was not one of the references cited by the respondent in its NOA and in keeping with the decision of the Federal Court of Appeal in AB Hassle v. Canada (2000), 7 C.P.R. (4^th) 272, the respondent may not rely upon it as prior art. However, I find it useful as an illustration of Dr. Caldwell's evidence that the knowledge of how to resolve racemic optical isomers was common prior to 1985.

[58]            The U.S. '651 patent discloses that pure optical isomers of this compound could be synthesized, but the process was "tedious and economically impractical using presently available techniques." Dr. Caldwell states that the ability of a chemist skilled in the art to resolve a racemate into its optical isomers is not diminished by the difficulty of designing an economically feasible production process to produce the pure enantiomer on a commercial basis.

[59]            Dr. Klibanov, Janssen's expert, does not address these U.S. patents in his affidavit but instead only refers to another piece of prior art cited by Novopharm in its NOA, the Japanese patent application No. 59-219293. He stated that the optical isomer disclosed in that application is a compound that is very different from ofloxacin and that this reference does not indicate any advantages of one optical isomer over the other isomer or the racemate.

[60]            I accept Dr. Caldwell's evidence that the techniques for isolating and verifying optical isomers were generally known and appear to have been understood and used by chemists in the early 1980s. The prior art Dr. Caldwell describes, including several patents published prior to the date of invention that detail the particular benefits of enantiomers isolated from other racemic antibiotics, also demonstrates that synthesizing individual enantiomers from racemic compounds was not the "extremely difficult" proposition maintained at one point by Dr. Klibanov. In fact, Dr. Klibanov recognized that, while making optical isomers with high optical purity was difficult, there were "various techniques known at that [the early 1980s] time" to undertake the process.

[61]            The 1976 article entitled Utilization of an Enantiomer as a Solution to a Pharmaceutical Problem: Application to Solubilization of 1,2-Di(4-piperazine-2,6-dione) propane by Repta et al., supra, cited by the respondent as prior art, refers to the ease in creating optical isomers, but notes (at p. 241) the difficulty in resolving them for an optical purity of close to 100%:

An inherent drawback in the use of an enantiomer is the greater difficulty and expense of preparation relative to the racemic material. Generally, it is not particularly difficult to attain an optical purity greater than 90%.[emphasis added] However, the preparation of essentially 100% optically pure material is extremely difficult and expensive. Therefore, it must be realized that it is not necessary to strive for 100% optically pure samples to obtain greatly increased solubility.

The Repta article describes (p.239) the generally known feature that typically one enantiomer exhibits greater or different biological activity than the racemate:

Although the employment of optically active compounds in pharmaceutical formulations is not new, their past and present usage has been due largely to one enantiomer exhibiting a quantitatively or qualitatively different biological activity than the corresponding racemic compound (6,7) Examples of such drugs include... [four examples mentioned, not ofloxacin]

[62]            I recognize that while Repta et al. noted that the optically active compounds had been known to have one enantiomer exhibiting a different biological activity than the racemate, the results that were actually found in this study were, at page 241 of the article, that the "antitumor activity for each enantiomer was equivalent to that of the (")-I. Therefore any pharmaceutical advantage to be realized from the use of an enantiomer could only arise from its significantly greater solubility relative to (")-I." The Repta article then described the findings of the increased solubility of the individual enantiomers relative to the racemate.

[63]            Dr. Klibanov's interpretation of the Repta article was that while it may have been known that there could be some solubility benefits with certain isomers, there was no general predictability with respect to the solubility of the racemate as compared to that of each of the optical isomers of ofloxacin. This could have only been obtained through experimentation.

[64]            The Repta article and U.S. patents support Dr. Caldwell's position that there were various known techniques for isolating enantiomers of racemic compounds in the early 1980s. Dr. Klibanov, on cross-examination, admitted that the optical purity that he believed the '080 patent was addressing was at least 90%, and he recognized that 100% optical purity of an enantiomer is virtually impossible to achieve and is more a theoretical concept.

[65]            Dr. Caldwell determined that the testing employed in the '080 patent for solubility, antimicrobial activity and toxicity levels of the enantiomers was done using simple, analytical procedures that did not involve any inventiveness. Dr. Klibanov did not address whether the experiments that were performed to determine the properties of the levorotatory isomer of ofloxacin were relatively simple ones. He only addressed the issue, as I read his affidavit and cross-examination, of the isolation of the distinct S(-) enantiomer of ofloxacin, finding that this was difficult to do and not disclosed by any previous ofloxacin patent.

[66]            Dr. Caldwell attests that the results of the testing on the optical isomers of racemic ofloxacin were not unexpected or surprising, given that the state of the scientific knowledge at the time was such that it was known that racemic compounds with certain attributes usually result from one of the enantiomers having quantitatively or qualitatively different biological activity from the racemic compound or the other enantiomer. This statement is in accordance with an article he co-authored, entitled "Putting Chirality to Work: The Strategy of Chiral Switches" Nature Reviews: Drug Discovery, Vol. 1, October 2002, which was brought to his attention on cross-examination by counsel for Janssen-Ortho.

[67]            While not prior art, this article by Dr. Caldwell illustrates a useful premise involving the requirement of inventiveness with chiral drugs. The following is an excerpt from the cross-examination of Dr. Caldwell at pages 876-77 of the applicant's record:

Q. On the same page in the left-hand column under the heading "Patentability of chiral switches", I want to read a sentence about seven lines from the bottom.

A. Yes.

Q. It reads:

"When the two enantiomers are of a chiral drug are sufficiently different in pharmacological effects from each other and from the racemate, it might be possible to obtain a patent for one or both enantiomers in addition to a valid (or expired) patent of the corresponding racemate."

A. Yes

Q. In principle, I assume you agree with that.

A. Yes. I mean when they are sufficiently different from each other and from the racemate. The whole phrase hinges around what we mean by sufficiently different.

[68]            In my opinion, this is in line with the approach in patent law that a patentable discovery must involve some inventive ingenuity. The improvements contemplated on the already discovered racemic compound must demonstrate "sufficiently different" characteristics from what was already known in order for the patent not to be declared invalid for obviousness. This is just a different way of articulating the "special property" concept described by Fox, above, that a selection patent must possess in order to qualify for a new patent on the selection and description of an item from an already known group or series.

[69]            Verification of certain attributes of an already disclosed compound does not meet the threshold of inventiveness required for patent protection. In SmithKline Beecham, supra, the Federal Court of Appeal found that only "mechanical skill", rather than inventive genius, was needed to apply the prior art to arrive at the invention contained in the challenged patent. As stated by the Court at page 138-39, "The only 'improvement' referenced in the disclosure merely verifies properties of a known substance formulated using common techniques."    In my opinion, the present case is similar, given that the inventors of the '080 patent were merely verifying properties of the known racemic ofloxacin by determining which of its enantiomers displayed better results with respect to antimicrobial activity, solubility and toxicity.     

[70]            In Sharp & Dohme Inc. v. Boots Pure Drug Company Ltd. (1928), 45 R.P.C. 153 (C.A.), Lawrence L.J. of the English Court of Appeal stated as follows at 191-92:

In my opinion, the ascertainment of the valuable properties of a chemical substance (important as it may be from the point of view of utility, where it is desired to obtain a patent for an entirely new substance) is not of itself, especially when such substance is one which anybody is entitled to make, a patentable invention, but at most a discovery. The Plaintiffs, when they obtained their four alkyl bodies by [published art] method, no doubt put them through the usual well-known tests to establish their identity and their respective therapeutic value.

[71]            I prefer Dr. Caldwell's evidence that the purportedly superior qualities of the S(-) enantiomer of ofloxacin with respect to antimicrobial activity set out in the '080 patent are not actually different, or of a special and unexpected nature, from the properties set out in the prior art, including in the '840 patent detailing the nature of ofloxacin. From Repta et al., the above-mentioned U.S. patents, and the two Gerster references, I am satisfied on a balance of probabilities that it would have been obvious to the skilled chemist that one of the enantiomers of racemic ofloxacin could exhibit higher beneficial properties and that all that would be needed to confirm this was to conduct routine, simple testing. Determining and noting which enantiomer possessed higher antimicrobial activity, a quality already existing in the racemate, did not require inventive ingenuity.

[72]            Moreover, I accept the evidence that, not only would it be expected from the prior art that the antimicrobial activity would be attributable to one of the two known enantiomers of ofloxacin, but based on another similar compound, a pharmaceutical chemist of ordinary skill in the art would have expected that the (-) enantiomer would be the one to exhibit higher antimicrobial activity. The Gerster flumequine references addressed this subject.

[73]            Relevant to this issue is the unsatisfactory manner in which Janssen's expert, Dr. Schroeder, performed his search for the Gerster flumequine references. These references, cited by Novopharm as prior art in its NOA, dealt with the likelihood of better properties in racemic flumequine existing on the (-) enantiomer. The claimed inability of Schroeder to find these prior art references is relevant to whether a person skilled in the art could have predicted that it would be the (-) enantiomer of ofloxacin that would be responsible for the antimicrobial activity of the racemate.

[74]            Dr. Klibanov attests that flumequine is a "completely different" compound from ofloxacin and that the differences between the two would have meant that extrapolating the results from Gerster's findings to ofloxacin would not have been a reliable or clear way to predict whether one of the ofloxacin optical isomers would have improved properties over the other isomer or over the racemate.

[75]            Dr. Caldwell, on the other hand, describes flumequine as a "structurally similar" antimicrobial compound to racemic ofloxacin. He claims that Dr. Klibanov has "inverted" the diagram of flumequine contained in his affidavit. The Gerster abstract reference revealed that in isolating both optical isomers of flumequine, the (-) isomer had "potent antibacterial activity" and the (+) isomer only "very weak activity". The Gerster poster presentation disclosed both a method for separating the enantiomers of racemic flumequine and that the differential antimicrobial activity of the (-) enantiomer was about two times greater than that of racemic flumequine and between 10 to 100 times greater than the (+) enantiomer. Dr. Caldwell goes into some detail as to why these results would be applicable to ofloxacin.

[76]            On cross-examination, it was apparent that Dr. Schroeder's search was not as thorough as one would reasonably have expected. He did not keep copies of his search results, or the search terms that he used with the various on-line databases. Moreover, he did not have much first-hand knowledge about certain searches for the Gerster references performed by a librarian who was under his supervision. Accordingly, I find that Dr. Schroeder's search was not as "reasonable and diligent" as one would expect of a skilled workman: General Tire and Rubber Co. v. Firestone Tyre and Rubber Co., [1972] R.P.C. 457 (C.A.);Illinois Tool Works Inc. et al v. Cobra Fixations Cie, supra.

[77]            Even if I were to find that the Gerster references dealing with flumequine are not valid prior art references, this would not affect my conclusion that the general state of knowledge for chemists in the field of stereoisomers in 1984 would have included the common understanding that one of the enantiomers of ofloxacin would likely exhibit higher antimicrobial activity. Therefore, verifying which enantiomer possessed this quality - and to what degree in comparison to the racemate and the other optical isomer - did not involve the exercise of any inventive step.

[78]            Turning to the solubility properties noted in the '080 patent, Dr. Klibanov attests that the 1976 Repta article indicates that it is not possible to predict which isomer of the two would be more efficacious with respect to solubility, or whether either isomer will be more soluble than its racemate. Dr. Caldwell refers to the findings set out in the Repta article which disclose that the solubility of pure enantiomers is several times greater than that of the corresponding racemates.

[79]            Dr. Caldwell also refers to the simplicity of the experiment to test for the solubility characteristics of a racemate. Again, Dr. Klibanov has not addressed the issue of whether the testing to confirm the properties of each enantiomer was routine, but rather focussed on the fact that in order to know with certainty the exact comparative nature of the enantiomers with the racemate, experimentation was required.

[80]            With regard to the toxicity findings, Dr. Caldwell attests that the comparative toxicity findings set out in the '080 patent are not at all significant for pharmaceutical formulation purposes, since therapeutic agents use concentrations of the drug substance at levels substantially below their "acute toxicity levels". This statement was not damaged on cross-examination. Further, Dr. Caldwell noted that based on the prior art it would be expected that a single enantiomer of an optically active compound would exhibit lower toxicity than its racemate.

[81]            Dr. Klibanov's statements in his affidavit support the position that the exact properties of the S(-) isomer, compared to the R(+) enantiomer and racemic ofloxacin, were only discovered through testing. However, he makes no comment as to whether these properties were unexpected, superior or of special advantage in comparison to the benefits already disclosed by the parent compound of ofloxacin. In fact, he repeats over and again the view that the exact nature of the properties of the S(-) enantiomer would not have been known without experimentation. The affidavit of Dr.Wendy Arnott, a Janssen-Ortho Vice-President, addresses toxicity, but only with respect to Novopharm's allegation that the toxicity data set out in the '080 patent are misleading and false.

[82]            Dr. Caldwell's evidence is that the tests to establish the beneficial properties of the S(-) enantiomer were entirely routine, uncomplicated, generally accepted well prior to 1985, and did not amount to any inventive step. While the exact antimicrobial, solubility and toxicity characteristics of levofloxacin could not have been predicted with absolute certainty without verification through testing, such characteristics would have been easily predictable or obvious to the skilled chemist with knowledge of enantiomers and stereochemistry, and could have been verified by using the routine analytical tests available to pharmaceutical chemists in 1984.

[83]            While not prior art, a section of the English text by Philip W. Grubb, Patents in Chemistry and Biotechnology (Clarendon Press: Oxford, U.K.,) published in 1986, illustrates the common knowledge at that time among chemists respecting the nature of optical isomers claimed as new compounds in patents.    As stated at page 134:

Perhaps the closest that a new compound can be to the prior art is the situation when the new compound is an optically active enantiomer of a compound previously known only in racemic form. This may be regarded as an extreme form of a selection invention, and it has been argued that the optically active form cannot be regarded as novel if the racemate is known, since the racemate could be considered as an equimolar mixture of the d- and l- forms. However, in cases both in the UK and the USA it has been decided that optical isomers of known racemates may be considered as novel per se.

Usually, however, the problem is regarded as one of obviousness rather than lack of novelty. It is obvious from the presence of an asymmetric centre in the molecule that optically active forms can exist, and it is usually obvious that they can be isolated by one or other of the standard methods of resolution. The only way in which an optical isomer can be patentable is if it has surprisingly superior properties as compared with the racemate, or if it has a use that the racemate did not have.

[Emphasis added]

[84]            As mentioned, I accept Dr. Caldwell's evidence that the properties described in the '080 patent concerning the levorotatory isomer of ofloxacin were not unexpected or surprisingly superior to what had already been disclosed with respect to ofloxacin. Based on the prior art of Repta et al., the Gerster references, the '840 patent disclosing ofloxacin as racemic with particular beneficial properties, and the U.S. patents cited dealing with other racemic compounds similar in structure to ofloxacin^[3], such "discovery" would have been obvious to the ordinary skilled chemist, knowing the state of the art of stereoisomers in 1985.

[85]            I am satisfied that Novopharm has established on a balance of probabilities that a technician skilled in the art would have come directly and without difficulty to the solution taught by the '080 patent by simply conducting known, routine experiments with racemic ofloxacin. I conclude that the claims set out in the '080 patent are obvious; what is claimed as the "discovery" or "invention" in the '080 patent does not suggest the scintilla of inventive ingenuity to qualify for the patent monopoly, but rather was the result of "mere verification" of the attributes of already known components of a compound.

[86]            Furthermore, based on the state of knowledge in 1985, I am satisfied that an average chemist would have been led directly to the conclusion, prior to such testing, that one enantiomer was likely to have more beneficial characteristics than the racemate, or the other optical isomer, in relation to the same beneficial qualities that had already been discovered, namely antimicrobial activity, solubility and toxicity. Therefore, in undertaking experimentation to verify which enantiomer possessed comparatively better qualities in these areas, no inventive step was involved.

[87]            I find that the '080 patent is invalid for obviousness. Overall, Janssen has not demonstrated on a balance of probabilities that Novopharm's allegation of invalidity on this ground is not justified.

Anticipation/ Lack of Novelty

[88]            While my conclusion above concerning obviousness is enough to determine this application, I will address the remaining issues in this proceeding. Paragraphs 27(1)(b) and 61(1)(a) of the pre-1989 Patent Act are applicable with respect to anticipation. The application for the '080 patent was filed on June 19, 1986, therefore any prior art referenced by Novopharm must be either a patent issued two years before this filing date, that is June 19, 1984, or a published paper or article made available to the public before June 19, 1984. Any document that was not published and available to the public before June 19, 1984 cannot be relied upon for anticipation.

[89]       Janssen argues that the prior art cited by Novopharm in its NOA with respect to lack of novelty, does not anticipate the discovered advantages of the particular optical isomer disclosed in the '080 patent. Counsel refers to In re May, 197 U.S. P.Q. 601 (C.C.P.A. 1978), a U.S. Court of Customs and Patent Appeals decision in which a patent claim to a (-) compound was found not lacking in novelty even though the prior art disclosed its racemate.

[90]            The applicant also relies on the English Patent Appeal Tribunal decision of Imperial Chemical Industries Ltd. (Howe's) Application, November 18, 1971 (Patent Appeal Tribunal). In that decision it was held that prior disclosure of a racemate did not anticipate its isomers, even though a chemist would realize that a racemate could be separated into levorotatory and dextrorotatory isomers. The applicant refers to a decision of the Canadian Patent Appeal Board where claims to an isomer in a substantially pure form were permitted even though the specific isomer at a lower purity level had been previously disclosed: University of Strathclyde Patent Application (Re) (1996), 77 C.P.R. (3d) 328 (Patent Appeal Board).   

[91]            The respondent submits that the S(-) enantiomer of ofloxacin was inherently disclosed in the ofloxacin '840 patent in May 1984 and that before 1984, the breakdown of enantiomers from a racemic compound was "routine" and did not involve the exercise of any inventiveness.

[92]            Novopharm refers to other patents for racemic mixtures similar in structure to ofloxacin that can exist in optically active forms. These patents included the optical isomers within the scope of the invention, and stated as such in the patent disclosure.

[93]            In asserting that a patent claim has been anticipated, the argument is that the invention has already been disclosed to the public and is therefore not novel. The well-established approach to anticipation is found inBeloit, supra at page 297:

It will be recalled that anticipation, or lack of novelty, asserts that the invention has been made known to the public prior to the relevant time. The inquiry is directed to the very invention in suit and not, as in the case of obviousness, to the state of the art and to common general knowledge. Also...anticipation must be found in a specific patent or other published document; it is not enough to pick bits and pieces from a variety of prior publications and to meld them together so as to come up with the claimed invention. One must, in effect, be able to look at a prior, single publication and find in it all the information which, for practical purposes, is needed to produce the claimed invention without the exercise of any inventive skill. The prior publication must contain so clear a direction that a skilled person reading and following it would in every case and without possibility of error be led to the claimed invention. Where, as here, the invention consists of a combination of several known elements, any publication which does not teach the combination of all the elements claimed cannot possibly be anticipatory.

[Emphasis added]

[94]            This approach was more recently cited with approval by the Supreme Court of Canada in Free World Trust, supra at paragraph 26. Moreover, Justice Binnie in that decision cautioned against the use of ex post facto deduction in evaluating whether an invention was anticipated by a particular publication, since assembling a dossier of prior art with the benefit of hindsight is "all too easy after an invention has been disclosed" (Free World Trust at para. 25).   

The '840 patent, its US equivalent - the '892 patent - and the 1983 Osada et al. publication

[95]            These three references are the prior art cited by the respondent in its NOA with respect to anticipation or novelty.

[96]            Novopharm's expert, Dr. Caldwell, attests that levofloxacin, as disclosed in the '080 patent, was anticipated by the '840 ofloxacin patent. Dr. Caldwell's position is that a chemist reading the '840 patent (and by implication, its U.S. counterpart, the '892 patent) would "immediately understand" that ofloxacin possesses a single chiral centre and based solely on its chemical structure, it would be apparent to a chemist that there must exist two enantiomers, or optical isomers. Dr. Caldwell also attests that the enantiomeric nature of ofloxacin was expressly disclosed in a publication by Osada et al. in 1983, "Antimycoplasmal Activity of Ofloxacin", where ofloxacin is identified using the (") symbol. The use of this symbol indicates that the authors of this article recognized the existence of both the (+) and (-) enantiomers of ofloxacin.

[97]            Janssen's expert, Dr. Klibanov, agrees that even though the '840 patent does not expressly indicate that ofloxacin is a chiral compound, "this would be apparent to an ordinary chemist based on the chemical structure of ofloxacin" (para. 64 of his affidavit). Novopharm points to Klibanov's cross-examination where he acknowledges that a chemist having read through the '840 patent would understand that the structure therein shown was a racemate.

[98]            Dr. Caldwell stated as follows at paragraph 41 of his affidavit:

Using the analogy that the disclosure of (+/-) ofloxacin can be likened to disclosure of a pair of gloves (i.e., left and right gloves), the above statement that an "S(-)... compound" would not be used to describe racemic ofloxcacin, is akin to saying that the term "a left glove" would not be used...to describe a pair of gloves. However, this does not address the essential fact that the S(-) compound sought to be claimed in CA '080 is disclosed in CA '840 and Osada et al.

[99]            Novopharm submits that prior disclosure of a pair of gloves inherently discloses its constituent left and right-hand gloves, and claiming a valid patent simply by asserting the term "left glove" cannot be viewed as asserting a novel patent claim that was not anticipated by an earlier disclosure of the "pair of gloves". In determining whether the '080 patent claims are novel, Novopharm also argues that Janssen's expert has improperly "read in" to the claims a purity level or limitation from the disclosure of the patent. Novopharm submits that such reading in from the disclosure should not be permitted to expand the scope of the patent claims and make them novel.

[100]        The applicant argues that the Osada et al. reference does not add anything new beyond the teachings of the prior '840 patent and it does not anticipate the '080 patent since following the teachings of this reference would not make levofloxacin in every event, without possibility of error.

[101]        Is there one piece of prior art which teaches the "discovery" claimed in the '080 patent? To paraphrase the accepted test, does a single publication provide all the information needed to produce the claimed invention without the exercise of any inventive skill? And does this prior publication contain so clear a direction that a person skilled in the art reading and following it, would in every case be led to the claimed invention?

[102]        Dr. Klibanov attested that the teachings of the '840 patent and the '892 patent can only result in racemic ofloxacin. At paragraphs 67 and 68 of his affidavit he continues:

There are special techniques that must be used to make an optical isomer, and these techniques are usually compound-specific. In the early 1980s, making an optical isomer with high optical purity was extremely difficult. While there were various techniques known at that time, their successful application to a specific compound, especially that as complex [as] ofloxacin, was impossible to predict.

Moreover, in many situations a specific methodology was required to be developed to obtain an optical isomer. Indeed, much of the '080 Patent is dedicated to describing processes used to obtain the novel ofloxacin optical isomers. For these reasons, I can unequivocally state that a person skilled in the art following the teachings of the '840 and '8[9]2 Patents would never obtain the optical isomers of ofloxacin because the processes set out in these patents result only in racemic ofloxacin.   

He maintained this position on cross-examination. Dr. Klibanov also attested that any particular properties of either optical isomer of ofloxacin over and above the other isomer or the racemate were not disclosed in any piece of prior art referenced by Novopharm.

[103]        Dr. Caldwell and Dr. Klibanov both agree that the '840 patent would suggest, to an ordinary chemist skilled in the art, that ofloxacin was a chiral compound and possessed optically active isomers. Where they disagree is whether this understanding inevitably includes a recognition of the benefits of one enantiomer over the other enantiomer, and over the racemate, and whether a useful process for isolating and making one enantiomer is "so clearly" disclosed from the teaching of a piece of prior art that it is anticipated.

[104]        Dr. Klibanov maintains that the process for the isolation or resolution of the ofloxacin enantiomers would not inevitably be apparent from the '840 patent, and would in fact be difficult to determine. Further, the beneficial properties of the S(-) optical isomer of ofloxacin would not be known from any of the pieces of prior art cited by Novopharm as anticipatory.

[105]        Dr. Caldwell maintains that Dr. Klibanov's statement that it would be very difficult to design a process of isolating enantiomers is not directed at the mere separation of racemic ofloxacin into its optical isomers but instead is directed to the development of processes to produce the enantiomer of interest on a commercially economical basis, after a compound has been isolated and tested to confirm its pharmacological activity. Dr. Caldwell then asserts that the three processes set out in the disclosure of the '080 patent are "production processes" rather than merely resolution processes.

[106]        In my view, Dr. Klibanov's statement is not so qualified. I do not read his words as indicating that the difficulty in "making" the optical isomer with high optical purity is restricted to the production of a particular enantiomer on a commercial scale.

[107]        Dr. Caldwell's glove analogy only goes so far in dealing with compounds that require a particular isolation process that will allow a chemist to note the properties of each enantiomer. A pair of gloves is visible without requiring a process to produce such visibility. In the case of the '840 patent and '892 patents, there is no teaching of how to differentiate enantiomers of racemic ofloxacin, nor is there any indication from one definitive, prior art source that either enantiomer will have certain improved qualities over the other enantiomer or the racemate. Moreover, the Osada reference does not disclose these teachings.

[108]        In my opinion, neither the '840 patent, the US '892 patent, nor the Osada reference, considered separately, provide all the information that a person of ordinary skill in the art would require in order to be led to the '080 patent. It is therefore novel and the claims set out in the '080 patent were not anticipated by the three pieces of prior art cited by Novopharm. Janssen has demonstrated that the allegation of invalidity on this ground is not justified.

Other Grounds of alleged invalidity

Alleged Ambiguity and allegation that claims broader than disclosed invention

[109]        Novopharm further alleges that the claims in the patent are both ambiguous and broader than the invention since they fail to include any "purity limitation" covering compositions which include the S(-) enantiomer mixed with other components, including the R(+) enantiomer. Novopharm therefore argues that the claims in the '080 patent are broader than the invention disclosed in the patent, since the '080 claims can cover any compound containing levofloxacin molecules, including the prior art of racemic ofloxacin (set out in the prior '840 patent), being an equal mixture of S(-) ofloxacin and R(+) ofloxacin, and the claims are invalid on this ground.

[110]        The applicants submit that certain "limiting" words are not required because the language that is employed in the '080 patent is sufficient to demonstrate that a chemist of ordinary skill in the art would never use the "S" or (-) symbol to designate a racemate. The S(-) designation indicates, instead, to an ordinary person skilled in the art that "practically all" of the molecules in the compound are of this configuration and "substantially free" of the molecules that have a "R" configuration. A chemist would also know that the solid wedge symbol indicates that "practically all" of the molecules have a "S" configuration. The claims of the '080 patent, according to Janssen, would be interpreted as covering an "optically pure form" of levofloxacin, since to achieve the invention's beneficial pharmacological effects one must use a pharmaceutical composition made of principally the S(-) enantiomer.

[111]        The applicants rely on the U.S. decision of Ortho-McNeil Pharmaceutical Inc., et al. v. Mylan Laboratories, Inc. et al., supra where the U.S. patent to levofloxacin was construed as not referring to racemic ofloxacin. A "plain-English purity limitation" was not required in the claim since the inventor was presumed to use terms that a person skilled in the art would understand, and the "S(-)" symbols clearly set out that the racemic compound was not covered by the patent. Also in Mylan, supra, the court found that the specification's repeated use of the words "optically active" and the contrast made between racemic ofloxacin and levofloxacin in the patent disclosure indicated that the patent dealt with the levorotatory enantiomer of ofloxacin. Mylan dealt with the defendant's attempt in a patent infringement action to obtain summary judgment as to the patent's invalidity on anticipation. The summary judgment motion was dismissed by the court.

[112]        The applicants also refer to patents granted to Novopharm's parent company, for isomers that did not include any purity limitations.

[113]        According to the experts, optical purity is the measure of the excess of one enantiomer over another in an optically active sample. A racemic mixture has 0% optical purity, since the equality of the enantiomers does not allow for optical activity.             

[114]        In my opinion, the '080 patent is not invalid on this ground. A person skilled in the art, reading the claims in the '080 patent, would understand that the S(-) enantiomer of ofloxacin in an optically pure form is described, as it is only when a substantial optical purity is present that the S(-) enantiomer becomes distinguishable from the racemate and the R(+) optical isomer. Janssen has satisfied me, on a balance of probabilities, that Novopharm's allegation of invalidity on the basis of ambiguity, because the patent lacks a definition or specification of optical purity, is not justified.

Specification insufficient

[115]        Next, Novopharm alleges that the specification of the '080 patent is insufficient, as it fails to correctly and fully describe the invention as required by paragraph 34(1)(a) of the Patent Act , by citing inconsistent toxicity data. In the NOA, Novopharm alleged that if such omission was willfully made for the purposes of misleading the public, then the patent is void pursuant to subsection 53(1) of the Patent Act.

[116]        Janssen submits that there is no evidence to suggest that the toxicity data reported in the '080 patent is incorrect. Janssen brought forward the affidavit of evidence of Dr. Wendy Arnott on this issue. Novopharm did not file any evidence in response.

[117]        Novopharm submits that the toxicity data set out for the racemic mixture in the '080 patent is inconsistent with the toxicity data previously reported by the inventors in the ofloxacin '840 patent, claiming the racemate. Novopharm points out several deficiencies in Dr. Arnott's knowledge of the toxicity studies performed for the '840 and '080 patents which were brought out on cross-examination. The respondent relies on H. Lundbeck A/S v. Apotex, 2003 FC 1334, (2003), 30 C.P.R. (4th) 198, where it was recently affirmed that the allegations set out in the NOA are presumed to be true except to the extent that they are disproved by the applicant.   

[118]        As explained by Dr. Arnott, the acute intravenous toxicity of a substance is determined by administering a range of doses of the substance to test animals and recording the number of test animals that died at each particular dose. This mortality data is used to calculate the LD₅₀ for a substance - the measure of the dose that was administered where the dose was lethal to 50 per cent of the animals tested. LD₅₀ values are typically expressed in units of milligrams of the drug administered per kilogram of test animal mass (mg/kg). A higher LD₅₀ value indicates that a substance is less toxic whereas a lower value indicates less toxicity.

[119]        Dr. Arnott attested that the '080 patent correctly set out that testing done on mice by intravenous administration indicated that levofloxacin was less acutely toxic than ofloxacin. She referred to Table 3 in the '080 patent at page 21 of the patent. At the 200 mg dose level, two out of five mice died when administered racemic ofloxacin, three out of five died when given the R(+) enantiomer of ofloxacin and no deaths occurred of five tested when the S(-) enantiomer was administered. Dr. Arnott also referred to the product monographs for FLOXIN (the brand name for racemic ofloxacin) and LEVAQUIN (levofloxacin), as demonstrating that the acute intravenous toxicity of levofloxacin was less than that of ofloxacin.

[120]        Having reviewed the cross-examination of Dr. Arnott, where Novopharm's counsel drew her attention to LD₅₀ values for oral administration to male mice that appear to show that levofloxacin has greater toxicity than ofloxacin, I do not accept Novopharm's contention that the data in the '080 patent is either incorrect or misleading. Dr. Arnott explained that doses of levofloxacin would appear more toxic when administered orally due to the fact that it is approximately ten times more soluble in water than ofloxacin and a greater amount is absorbed than when the drug is injected intravenously. The patent disclosure indicates that the acute toxicity results were "determined in mice by intravenous administration" and does not purport to offer toxicity data in relation to oral administration. While the data compiled in the product monographs for FLOXIN and LEVAQUIN appear to offer conflicting conclusions in relation to the oral dosages given to mice, the '080 patent does not purport to relate to such oral administration.    

[121]        Janssen bears the legal burden of demonstrating that the allegations are not justified, and their witness, Dr. Arnott, withstood the cross-examination of Novopharm by adequately explaining the difference in the results between oral and intravenous administration of dosages. Moreover, Novopharm continues to have an evidentiary burden and did not put forth any affidavit evidence of its own to counter Dr. Arnott's conclusions. Therefore, I am satisfied that Janssen has demonstrated that Novopharm's allegation of invalidity on this ground is not justified.

Infringement

[122]        Despite the fact that my determination on obviousness is sufficient to dismiss this application for a prohibition order, I will briefly address the issue of infringement. Both parties agree that the only claims at issue in relation to infringement are claims 2, 4, 7-9 and 17-19. As noted above, the process claims (1, 3, 5, 10 and 11) are not in issue. Claim 6 is also not in issue as it appears to claim a compound that is not levofloxacin due to the fact that "ethyl" is included in the formula rather than "methyl". Claims 12-16 are not at issue as the applicant conceded that they clearly relate to salts of levofloxacin.

[123]        In my view, the most significant claims with respect to the alleged infringement are 2, 4, 17 and 18. These claims read as follows:                                     

2. An S(-)-pyridobenzoxazine compound represented by the general formula (VI)

wherein X1 represents a halogen atom, R1 represents an alkyl group having 1 to 4 carbon atoms, and R3 represents an alkyl group having 1 to 3 carbon atoms.

4. S(-)-9-fluro-3-methyl-10-(4-methyl-1-piperazinyl)-7- oxo-2,3-dihydro-7H-pyrido [ 1,2,3-de] [ 1,4] benzoxazine-6-carboxylicacid.

17. A compound of the general formula(VI)as defined in claim 2 and the pharmaceutically acceptable salts thereof wherein said compound is a hydrate.

18. A pharmaceutical composition comprising a compound of the general formula (VI) as defined in claim 2 or the pharmaceutically acceptable salts thereof and one or more compounds selected from the group consisting of carriers, fillers, extenders, binders, humectants, absorption accelerators, wetting agents, adsorbents, lubricants, solubilizing agents, emulsifiers, stabilizers, preservatives.

[124]        The applicant argues that it is clear that Novopharm's product, described in the NOA as levofloxacin hemihydrate made using "the free acid form" of the compound rather than a salt of the compound, will infringe claims 2 and 17 of the '080 patent. These, it asserts, are interdependent and both include the hemihydrate form of levofloxacin. The formula in claim 2, together with a ½ mole of water constitutes levofloxacin hemihydrate. Claim 17 relies on claim 2 where the compound is a hydrate.

[125]        Dr. Klibanov's affidavit evidence is that claim 17 cannot be limited to hydrates of a salt of levofloxacin and that Novopharm's product, as a hemihydrate of levofloxacin, will literally be within the scope of claims 2 and 17, as well as claims 4, 7-9, 18 and 19. Since Novopharm admits it will make tablets that contain levofloxacin hemihydrate together with a carrier, its product will also infringe claim 18 of the patent.

[126]        Novopharm submits that its levofloxacin hemihydrate product will not infringe claim 17 because it will be made using the free acid form of the compound and not a salt of the compound. No salt is present in the Novopharm formulation. Novopharm stresses that claim 17 must be read with claim 18, in that claim 17 requires, due to the word "and", the presence of "pharmaceutical salts" of the compound defined in claim 2, wherein that compound is a hydrate. Counsel stressed that the language in claim 18 where "or" is used, must be evaluated and contrasted with claim 17, indicating that the preparation claimed in 17 is the compound described in claim 2 only when it is in a pharmaceutically acceptable salt form.

[127]        Novopharm argues that Janssen's contention that claims 2, 4, 6, 8 , 9 and 18 comprise all forms, including various hydrates, solvates or polymorphic forms of levofloxacin, cannot stand. In view of Janssen's expert's opinion that approximately 90 per cent of the molecules are S (-) optical isomers of ofloxacin, the patent claims cannot reach as far as covering the hemihydrate form of levofloxacin, since for the hemihydrate form to exist, the S(-) enantiomer of ofloxacin would be linked with ½ mole of water, meaning that 1/3 of the molecules in the hemihydrate form of levofloxacin would not meet the definition of the S(-) enantiomer claimed in the patent.

[128]        In my opinion, Novopharm's interpretation cannot be accepted. As Janssen has pointed out, Novopharm's expert Dr. Caldwell in fact acknowledged in cross-examination that levofloxacin hemihydrate is within claim 17. Novopharm has described its product in the NOA as levofloxacin hemihydrate. Further, Novopharm's interpretation ignores the final words of claim 17: "wherein said compound is a hydrate." The proper construction of claim 17 covers levofloxacin, in any hydrate form, including pharmaceutically acceptable salts thereof, but is not necessarily limited to the salts. However, I do not accept Janssen's contention that "S(-)" - meaning principally S(-) or the S(-) optical isomer of ofloxacin in a substantially pure form - means water cannot exist in the compound form. The '080 patent expressly states that the compound when it is a hydrate is included within claim 17.

[129]        In my view, therefore, Novopharm's levofloxacin hemihydrate product would infringe claims 2 and 17 of the '080 patent if it were valid.

Costs

[130]        As between Novopharm and the applicants, except where otherwise provided by another Order of this Court, Novopharm is entitled to its costs, to be calculated on the ordinary scale, unless, within fourteen days from the date of the Order herein, Novopharm applies to the Court for an Order that its costs be determined on a different basis. If any such application is made, the applicants will be entitled to serve and file a response thereto, and Novopharm to serve and file a reply to any such response, all in accordance with the Rules of this Court. Any such application may be made under Rule 369 or may be made returnable at a location and on a date and at a time satisfactory to the Court and to counsel.

[131]        There will be no order as to costs either in favour of or against the Respondent, the Minister of Health.

ORDER

THIS COURT ORDERS that the application for an order prohibiting the Minister from issuing a Notice of Compliance to the respondent, Novopharm Limited, is dismissed. Costs to the respondent as provided in paragraphs 130 and 131 of the Reasons for this Order.

"Richard G. Mosley"

F.C.J.

ANNEX "A"

Definitions Employed in These Reasons

Isomers: different compounds that have the same molecular formula.

Field of stereochemistry/ Stereoisomers: concerns the three-dimensional spatial orientation of compounds made up of atoms. Stereoisomers have the same molecular formula, but the bonds joining the atoms are arranged differently in space.

Enantiomers: stereoisomers that are non-superimposable mirror images of one another, by analogy, one's hands. Often called "optical isomers".

Optical Isomers: enantiomers; isomers that are optically active, may have different physical and chemical properties than their "racemates" and are regarded as distinct compounds from their racemate.                        

Chiral compounds/ Chiralty: compounds that exhibit the non-superimposable mirror image quality are called "chiral". These compounds possess a chiral centre and in carbon compounds, such as ofloxacin, the chiral centre is a carbon atom that is bound to four different atoms or groups of atoms, where all four groups are different. Chiralty confers a specific three-dimensional shape on a molecule.

"R" and "S" designations: how chemists distinguish optical isomers based on the spatial arrangement of their atoms attached to the chiral centre. Where the direction from the highest to lowest priority atom or group, determined by atomic number, attached to the chiral carbon is clockwise, "R" is designated and where this is counter-clockwise "S" is designated. To show this arrangement graphically, chemists show a chemical bond that projects up from the plane of the paper toward the viewer with a solid wedge shape and one that projects down away from the viewer by a hatched wedge shape.

"(-), (+) and (")" designations: used by chemists to distinguish optical isomers based on their optical activity, i.e. the direction in which they rotate the plane of polarized light. Optical isomers that rotate polarized light to the right, in a clockwise direction, are "dextrorotatory" and use the symbols (+) or (d) for "dextro". Optical isomers that rotate polarized light to the left, in a counter-clockwise direction, are "levorotatory" and use the symbols (-) or (l) for "levo".

Racemic mixtures, compounds/ Racemates: substances containing equal amounts of two optical isomers, commonly represented by the symbols ("), (+/-), (R/S) or (dl), but does not always contain these symbols to designate the racemate. Does not rotate the plane of polarized light and therefore is not "optically active". The rotation caused by the (-) molecules is cancelled out by the equal and opposite rotation of the (+) molecules. Example: ofloxacin is a racemate composed of equal amounts of the two optical isomers (+) ofloxacin and (-) ofloxacin, and is sometimes referred to as (") ofloxacin.

Hydrate forms: a form of a compound where the solid, crystal structure includes water; a compound that crystalizes in the absence of water is an "anhydrate". Where there are two molecules of water present in the crystalline structure for each molecule of compound, this is a "dihydrate"; where there is one molecule of water for each molecule of compound, this is a "monohydrate"; and where there is one molecule of water for every two molecules of compound, this is a "hemihydrate".

                                     FEDERAL COURT

                              SOLICITORS OF RECORD

DOCKET:                  T-214-03

STYLE OF CAUSE: JANSSEN-ORTHO INC. and

DAIICHI PHARMACEUTICAL CO., LTD.

and

NOVOPHARM LIMITED and

THE MINISTER OF HEALTH

PLACE OF HEARING:                                 Ottawa, Ontario

DATE OF HEARING:                                   May 25, 2004

REASONS FOR ORDER

AND ORDER BY:    The Honourable Mr. Justice Mosley

DATED:                     November 19, 2004

APPEARANCES:

Anthony Creber                                                 FOR THE APPLICANTS

Cristin Wagner                                                  (Janssen-Ortho Inc.)

Michael Charles                                                 (Daiichi Pharmaceutical Co., Ltd)

David Aitken                                                     FOR THE RESPONDENTS

J. Bradley White                                                (Novopharm Limited)

Frederick B. Woyiwada                                                (Minister of Health)

SOLICITORS OF RECORD:

ANTHONY CREBER                                                  FOR THE APPLICANTS

CRISTIN WAGNER                                        (Janssen-Ortho Inc.)

Gowling Henderson Lafleur LLP

Ottawa, Ontario

MICHAEL CHARLES                                                 (Daiichi Pharmaceutical Co.,

Bereskin & Parr                                                Ltd)

Toronto, Ontario

DAVID AITKEN                                             FOR THE RESPONDENTS

J. BRADLEY WHITE                                                   (Novopharm Limited)

Osler, Hoskin & Harcourt LLP

Ottawa, Ontario

MORRIS ROSENBERG                                              FOR THE RESPONDENTS

Deputy Attorney General of Canada

Ottawa, Ontario