Eli Lilly & Co. v. Novopharm Ltd., docket A-391-97

Date: 20001219

Docket: A-391-97

(T-2432-95)

CORAM:        DESJARDINS J.A.

SEXTON J.A.

SHARLOW J.A.

BETWEEN:

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                        NOVOPHARM LIMITED

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

                                                                                                                              Docket: A-392-97

                                                                                                                                         (T-2433-95)

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                               NU-PHARM INC.

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

                                                                                                                              Docket: A-393-97

                                                                                                                                         (T-2434-95)

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                                  APOTEX INC.

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

Heard at Toronto, Ontario, on Monday, Tuesday, Wednesday, Thursday, Friday, October 16, 17, 18, 19, 20, 2000.

Judgment rendered at Ottawa, Ontario, on Tuesday, December 19, 2000.

REASONS FOR JUDGMENT BY:                                                                 DESJARDINS J.A.

CONCURRED IN BY:                                                                                               SEXTON J.A.

                                                                                                                                SHARLOW J.A.

Date: 20001219

Docket: A-391-97

(T-2432-95)

CORAM:        DESJARDINS J.A.

SEXTON J.A.

SHARLOW J.A.

BETWEEN:

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                        NOVOPHARM LIMITED

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

                                                                                                                              Docket: A-392-97

                                                                                                                                         (T-2433-95)

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                               NU-PHARM INC.

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

                                                                                                                              Docket: A-393-97

                                                                                                                                         (T-2434-95)

                                                    ELI LILLY & COMPANY and

                                                       ELI LILLY CANADA, INC.

                                                                                                                                          Appellants

                                                                                                                                           (Plaintiffs)

                                                                         - and -

                                                                  APOTEX INC.

                                                                                                                                        Respondent

                                                                                                                                        (Defendant)

                                                    REASONS FOR JUDGMENT

DESJARDINS J.A.

[1]                The appellants seek to reverse a judgment of the Trial Division^[1] dismissing their claims that the respondents had passed off their generic fluoxetine hydrochloride capsules as the appellants' Prozac brand fluoxetine hydrochloride capsules, contrary to paragraph 7(b) of the Trade-marks Act (the "Act").^[2] The trial judge found that the appellants had not proven that the size, shape and colour of their Prozac brand fluoxetine hydrochloride capsules had acquired a secondary meaning - that consumers did not associate the appearance of the Prozac capsules with either Lilly or any other particular source - nor that the respondents' sale of their generic fluoxetine hydrochloride capsules was likely to cause confusion in Canada. The trial judge also found that there were no valid licence agreements between Eli Lilly & Company and Eli Lilly Canada Inc., and between Eli Lilly Canada Inc. and two other corporations, Pharmascience and 115013 Canada Inc., to use the size, shape and colour of the Prozac capsules as a trade-mark.

[2]                The issue before us is whether the trial judge erred in so doing.

I - Facts

[3]                The appellant Eli Lilly Canada Inc. ("Lilly Canada") is a wholly owned subsidiary of the appellant Eli Lilly and Company ("Lilly U.S."). Both sell fluoxetine hydrochloride under the brand name Prozac. Lilly U.S. invented fluoxetine hydrochloride ("fluoxetine"), which was the first of a new generation of antidepressants.

[4]                Lilly Canada marketed Prozac in Canada under a 1976 licensing arrangement with Lilly U.S. On January 1, 1991, that agreement was superceded by a new one. This new agreement granted Lilly Canada the right to make, use, distribute and sell certain scheduled Lilly U.S. products, including Prozac. It also granted Lilly Canada the right to use certain scheduled trade-marks, brand names, house marks and other designations belonging to Lilly U.S. Although the relevant schedule (Schedule B) included the brand name "Prozac", as well as designs and packaging relating to various other Lilly products, the size, shape and colour of the Prozac capsules were not mentioned. The 1991 Agreement did not grant Lilly Canada a right to sub-licence any of its rights under the agreement. It also specified that the entire agreement between the parties was contained in the document - there were no oral agreements, representations or undertakings - and that any modifications to the agreement were to be made in writing.^[3] The Canadian trade-mark registration of the word mark Prozac is owned by Lilly U.S. It is not contested that Prozac is a registered trade-mark of Lilly U.S.

[5]                The products at issue in these appeals are capsules containing 10 and 20 mg dosages of Prozac brand fluoxetine. The 20-mg Prozac capsules, introduced to the Canadian market in 1989, are half green and half cream. The 10-mg Prozac capsules, first sold in Canada in 1993, are half green and half pale grey. There are a limited number of standard capsule sizes for prescription medicine; the larger the number, the smaller the size of the capsule. The size and shape of capsules must be such as to be easily consumable. Lilly chose a size 3 for both the 10-mg and 20-mg dosage forms. Lilly's capsule differs from the standard shape in one respect: while the cap end of the Lilly capsule is spherical, the body end is bullet shaped. The capsules are marked "Prozac" and "Lilly", along with the dosage size and an identification number. "Lilly" is written in a stylized script.^[4]

[6]                Lilly Canada had a monopoly on the supply of fluoxetine in Canada between 1989 and the expiry of Lilly U.S.'s patent on March 20, 1996. Between 1990 and 1995, Lilly Canada sold more than four hundred million dollars worth of Prozac brand fluoxetine in Canada. The appellants spent more than twenty-two million dollars promoting it. The drug became the focus of considerable media attention and its significance in the treatment of depression was widely publicized. Prozac's appearance (also known as "trade dress" or "get-up") was depicted on the covers of national magazines.

[7]                In anticipation of the March 20, 1996 expiry of the patent, Lilly decided to market its fluoxetine as a generic product as well as under the trade name Prozac. Arrangements for this purpose commenced in March 1995. On June 30, 1995, Lilly Canada entered into an agreement with Pharmascience and 115013 Canada Inc. (which the trial judge referred to collectively as "PMS") to distribute and sell generic versions of Lilly products including fluoxetine.

[8]                In late October 1995, the January 1991 Agreement between Lilly U.S. and Lilly Canada was amended to "confirm" that Lilly Canada had always been licensed to use the product appearance of the Lilly U.S. products that it was selling or had sold on the Canadian market. Schedule B was amended to include "the trade dress, i.e., the colour, size, shape and other indicia..."^[5] of the appearance of Prozac. The agreement was also amended (retroactive to June 30, 1995) to allow Lilly Canada to sub-licence its rights under the agreement to PMS. Finally, on November 10, 1995, Lilly Canada granted PMS the right to use the product appearance of the Lilly fluoxetine.

[9]                Under the agreement with PMS, Lilly Canada controls all aspects of the manufacture, distribution, promotion and sale of PMS-fluoxetine. The product is virtually identical to Prozac brand fluoxetine. It is produced by Lilly on the same production lines as Prozac. The only difference between the two products is in the markings which consist of "TRADE DRESS LICD", the dosage and an identification number.

[10]            PMS-fluoxetine became available in Canada on December 1, 1995. Lilly Canada sold the product in the hospital market. PMS sold the product to pharmacists with the assistance of marketing efforts by Lilly Canada directed at doctors and pharmacists.

[11]            The respondents, Novopharm, Apotex and Nu-Pharm, are all companies that produce generic versions of fluoxetine.

[12]            The day after the expiry of the patent, on March 21, 1996, an interim injunction was issued in the Trial Division preventing the respondents from adopting trade dresses for their fluoxetine hydrochloride products which were the same as or similar to the Prozac trade dress. The interim injunction was replaced on April 26, 1996, by an interlocutory injunction. When the injunctions were in place, each of the respondents marketed its fluoxetine products in non-look-alike capsules (all cream for the 20-mg fluoxetine dosage form and all grey for the 10-mg dosage form). The interlocutory injunction was overturned by this Court on September 25, 1996, following which the respondents changed the trade dress of their capsules and began marketing their products in a trade dress similar to the Prozac trade dress. There were some differences in appearance. To begin with, all of the respondents' products were sold in standard shaped capsules with two spherical ends. While the 20-mg dosages were all marketed in size 3 capsules, the 10-mg dosages marketed by Apotex and Nu-Pharm were contained in the smaller, size 4, capsules. Finally, the markings on the generic fluoxetine products were different. All generic products were marked in block capitals with only a company designator and the dosage (i.e. NOVO 20, APO 10, etc.). The trial judge found that the differences "only become obvious upon a close viewing of the capsules".^[6]

[13]            The trial of the action began on November 4, 1996, just a few short weeks after the respondents entered the market with their similar appearance products.

II - The decision appealed from

[14]            The trial consisted of forty-two hearing days over a period of five months. The trial judge dismissed the action with costs. Her reasons were extensive and included detailed reasons for her findings of fact and with respect to the credibility of the witnesses who gave evidence before her.

[15]            The trial judge summed up the issue before her in the following manner:^[7]

The issue is whether or not the capsule appearance of the Lilly fluoxetine is distinctive as an indication of trade source or provenance, and whether the use of a similar capsule appearance by the defendants would result in a likelihood of confusion.

(a) The relevant market

[16]            The trial judge considered the relevant market to be potential as well as actual customers, including doctors, pharmacists and patients. She accepted that medical professionals do not prescribe or dispense drugs on the basis of appearance. Her focus was on evidence dealing with actual and potential patients. She listed a number of factors that made consideration of this market (and consideration of it at any particular point in time) difficult. These included the percentage of the population likely to suffer from depression at some point in their lives (about 15%), the unpredictable onset and recurring nature of the condition, the various treatment methods available and the period of time for which an anti-depressant medication might be prescribed. She held that these factors "make the potential customer market difficult if not impossible, to identify",^[8] and that the "actual customer base for fluoxetine will change from day to day and will be significantly different from one six to nine month period to the next".^[9] The trial judge finally declined to reject any evidence on the basis that it surveyed the wrong market. Rather, she assessed the weight of the evidence by reference to the questions asked and the group to whom the questions were put.^[10]

(b) The evidence

[17]            Evidence was received in the form of fifty-one expert reports and in the testimony of sixty-two witnesses, including doctors, pharmacists and Prozac users. Central to the evidence were four separate surveys. One survey, that designed by Dr. Heeler on behalf of Lilly, questioned only respondents who claimed to have been prescribed or to have used an anti-depressant in the preceding five years. Based on expert evidence and her own assessment of the survey design, the trial judge found significant defects in the survey, particularly the biassed nature of the order in which the questions were posed. She recognized that the survey had the advantage of asking for visual recognition on the basis of the actual physical presentation of a green and cream capsule. She concluded, however, that:^[11]

The defects in the survey, however, are so serious that they undercut its usefulness. In this regard, for example, Dr. Heeler's conclusion that the fame of Prozac relates not solely to the brand name but also to the capsule appearance is a leap of logic not supported by the data.

[18]            The other surveys contained some inherent disadvantages, in particular their sampling of broader populations and their testing of recall rather than recognition of appearance. However, the trial judge found them to be professionally conceived and presented. While the populations sampled were broad, the surveys were designed to allow analysis of sub-populations known as the "Prozac Aware" (people who indicated that either they, or a member of their family, had taken Prozac) and the "Prozac Knowledgeable" (people that had experience with Prozac). Another study surveyed persons known to live in households where one of the household members had consulted a doctor regarding depression in the preceding year (the so-called "Anxiety Sample"). Within this group, a smaller population of persons exposed to Prozac through its use by household members or close friends or family members was identified and analysed (the "Claimed Prozac Exposed").^[12] In considering these surveys, the trial judge extrapolated their conclusions to make findings of fact with respect to the market of actual and potential customers that she found relevant.

(c) Distinctive appearance as indicative of source

[19]            The trial judge heard evidence from the respondents that they chose to copy the innovator's capsule appearance for marketing reasons. They asserted that in Canada the appearance of prescription medicine is associated with the type and dosage form of the drug.^[13] She noted the practice of marketing generic drugs of the same colour, shape and size as the innovator's brand to have been standard in the Canadian pharmaceutical industry for a period of some decades.^[14]

[20]            The parties do not contest the trial judge's conclusion that from a pharmaceutical or "bio-availability" point of view, the products of all of the parties were interchangeable. The non-active fillers and the composition of the gelatin cap might vary between manufacturers and the percentages of the active ingredient might vary from capsule to capsule or batch to batch depending upon each manufacturer's self-imposed tolerances. However, the generic fluoxetine products have been found by the Health Protection Branch of the federal Department of Health and Welfare to be equivalent to Prozac brand fluoxetine. Various provincial health agencies have also found the products to be interchangeable. The trial judge found any differences that did exist to be so insignificant that it was accurate to say that the medications were the same.^[15]

[21]            She found that a two-colour capsule for prescription medication was not in itself inherently distinctive. She referred to evidence before her that Lithane, a medicine sometimes prescribed as an adjunct to the treatment of depression, was available in green and yellow capsules, and that Librium, an anti-anxiety medication, was sold in green and yellow capsules in the 20-mg dosage form. While recognizing that the shades were not identical to those used by Lilly, she held that there was nothing inherently distinctive about those specific colours.^[16]

[22]            With respect to the issue of whether consumers associated the appearance of the capsules with a single trade source or provenance, the trial judge considered the evidence that she found to be reliable and held:^[17]

I cannot conclude on the facts set out above that the plaintiffs have proven that the capsule appearance has acquired the requisite reputation in the market place as a distinguishing feature of the plaintiffs' product.                                 [Emphasis in text]

[23]            Rather, she found that if the market place attaches any meaning to the appearance of the capsules, it is with respect to the medicine contained in them and to its therapeutic effect, and not its trade source or provenance.^[18]

(d) Confusion

[24]            The trial judge held that there was no evidence of actual confusion before her. She went on to review the various groups of patients who might potentially be confused by similar appearance. She wrote:^[19]

I cannot conclude that the plaintiffs have proven, on the balance of probabilities, that the defendants' sale of fluoxetine in capsules having a similar appearance to those of the plaintiff would result in any significant likelihood of confusion.

[25]            She also held that:^[20]

The appearance of the defendants' capsules does not operate to lead a customer to request the products of the defendants rather than those of the plaintiffs.

                                                                                                                   [Emphasis in text]

(e) The licensing arrangements

[26]            The trial judge devoted a portion of her reasons to an analysis of the licensing arrangements between the appellants and, particularly, whether the use of the Prozac get-up had been licensed by Lilly U.S. to Lilly Canada and, in turn, by Lilly Canada to PMS. She held, based on a strict reading of the 1991 agreement between Lilly U.S. and Lilly Canada, that the use of the get-up by Lilly Canada was not licensed. In the absence of sufficient evidence regarding the structure of the relationship between Lilly U.S. and its Canadian subsidiary, she refused to hold that the use of the get-up by Lilly Canada would enure to the benefit of its parent. She found that the evidence did not support the conclusion that any licence that she might infer was for use of the get-up as a trade-mark. She rejected the appellants' interpretation of subsection 50(1) of the Trade-marks Act and held that the 1995 amendment did not cure the defect in the 1991 agreement. Finally, she held that the use of the same get-up by PMS, even if it were licensed, would destroy the distinctiveness of Lilly Canada's use.

II - Issues

[27]            The key issue in this appeal is whether the trial judge applied and, if she did, whether she correctly applied the decision of the Supreme Court of Canada in Ciba-Geigy Canada Ltd. v. Apotex Inc.^[21] and the rules of passing-off. A subsidiary issue is whether the trial judge erred in her appreciation of the licensing arrangement between the two appellants and between one appellant and PMS.

III - The appellants' position

[28]            The appellants claim that this case provided the Court with its first opportunity to apply the decision of the Supreme Court of Canada in Ciba-Geigy relating to the use by a generic company of a trade dress for its pharmaceutical product that is similar to that introduced to the market by the inventor of the drug. They argued that the trial judge not only failed to consider or apply the decision of the Supreme Court, but that she disregarded it.

[29]            The appellants claim that the Ciba-Geigy case is authority for more than merely the proposition that patients must be included as part of the relevant market of consumers in a passing-off action with respect to prescription drugs. The trial judge failed, they say, to recognize the direction the Supreme Court of Canada gave in the application of the law of passing-off to the pharmaceutical field; namely that the patient has a choice and that a choice requires protection and a mechanism of control. They claim that the Supreme Court of Canada held that since the patient has no direct access to the product, it is all the more necessary that he/she be able to exercise some kind of control over what is being given, and that appearance is the first information he/she gets about the prescription drug. The get-up must therefore be protected in order to give effect to the patient's right to choose. As a result, competing manufacturers have a duty not to use similar get-ups.

IV - Analysis

[30]            My analysis must, therefore, begin with an examination of the decision of the Supreme Court of Canada in the Ciba-Geigy case.

(a) The Ciba-Geigy decision

[31]            Ciba-Geigy involved passing-off actions brought by an innovator company, Ciba-Geigy, which produced a brand name prescription drug "Lopresor", for low or moderate hypertension. The actions were against Apotex Inc. and Novopharm Ltd. which produced generic versions of the same pharmaceutical formulation that were identical in appearance to the brand name product. The tablets were of the same get-up -- shape, size and colour -- as those of Ciba-Geigy. Apotex Inc. and Novopharm Ltd. filed motions for summary judgment in the Supreme Court of Ontario in response to the passing-off actions. In the alternative, Apotex Inc. and Novopharm Ltd. asked the Court to rule on the following question of law.^[22]

... with respect to the marketing of prescription drugs, a plaintiff in an action for the alleged passing-off of a prescription drug must establish that the conduct complained of is likely to result in the confusion of physicians or pharmacists in choosing whether to prescribe or dispense either the plaintiff's or the defendant's product.

[32]            The motions judge denied the motions, but answered the question of law in the affirmative. Ciba-Geigy appealed to the Ontario Court of Appeal. It argued the trial judge erred in excluding patients who were consumers of the drugs in question. The Ontario Court of Appeal dismissed the appeal. Leave was granted by the Supreme Court of Canada.

[33]            Gonthier J., for the Supreme Court of Canada, defined the issue in the following manner:^[23]

   This court must determine whether, in a passing-off action dealing with prescription drugs, a plaintiff may argue that the public affected by the risk of confusion includes, in addition to physicians, dentists and pharmacists, the patients who consume the drugs, or is instead limited exclusively to the health care professionals in question.

[34]            In the course of his analysis, Gonthier J. reviewed the general principles of passing-off developed by the courts as early as 1842 in Perry v. Truefitt^[24] where it was stated:

[...] no man is entitled to represent his goods as being the goods of another man.

[35]            He confirmed the three necessary components of a passing-off action as being: (1) the existence of goodwill or reputation attached, in the mind of the purchasing public, to goods or services which, by association with the get-up, have become distinctive specifically of the plaintiff's goods and services, (2) deception of the public due to a misrepresentation leading or likely to lead the public to believe that the goods and services of one person are those of the plaintiff, and (3) actual or potential damage to the plaintiff.^[25]

[36]            As to the first component,^[26] Gonthier J. referred to the case of Oxford Pendaflex Canada v. Korr Marketing Ltd. et al.^[27], where it was said that in a passing-off action, the plaintiff, in order to succeed, must establish that its product has acquired a secondary meaning. He noted that the requirements of the passing-off action had evolved somewhat and said:^[28]

... A manufacturer must therefore avoid creating confusion in the public mind, whether deliberately or not, by a get-up identical to that of a product which has acquired a secondary meaning by reasons of its get-up.

[37]            He reminded that although in the Anglo-Saxon legal systems the person chiefly concerned was the competitor affected by the unfair act of passing-off, unfair competitions cases were affected with a public interest, namely the protection of customers.^[29] He considered the market outside the field of pharmaceutical products^[30] and stated, after citing Nadeau and Nadeau,^[31] that it was necessary to avoid confusing "anyone who has an actual or potential, immediate or remote, connection with the product".^[32] He pointed out that the expression "ultimate consumers" was fading out of use in the jurisprudence and that the word "public" was increasingly preferred.

[38]            He then applied the principles of passing-off to the pharmaceutical field and stated as a preliminary observation:^[33]

     There is no need to dwell at length on the purposes of the passing-off action in this field as they are essentially the same as those I have just considered. Adapting the principles set out to cases such as that now before this court leads to the conclusion that competing laboratories must avoid manufacturing and marketing drugs with such a similar get-up that it sows confusion in the customer's mind.

                                                                                                                   [Emphasis added]

[39]            He observed that pharmaceutical companies were limited in the choice of ways to distinguish the get-up of their products. Since pharmacists buy such products in bulk and dispense them to the public in standard containers, which are transparent and anonymous, the only way of drawing the attention of patients to the origin of the products was the capsule or tablet itself. Since the tablets are small, not much that is legible could be written on them. That left only the shape, size and colour of the products as a means of distinguishing them. The size and shape of drugs could not depend solely on imagination since they must meet certain functional requirements resulting from physiological necessities such as ingestion and digestion. As to colour, owing partly to the small size of the products, combinations which might be original in characteristic were also relatively limited.

[40]            Gonthier J. then determined the target clientele of pharmaceutical products.^[34] The manufacturers were obviously meant to be protected by a passing-off action. But, also, he said, those who "buy" or "consume" the products. "As the patient has no direct access to the product" he said, "it is all the more necessary for him to be able to exercise some kind of control over what he is being given".^[35] He noted that in the field of prescription drugs, the first information the patient received when the product was given to him came from its appearance.

[41]            He considered the relevant legislation, namely the Prescription Drug Cost Regulation Act, 1986,^[36] and the Prescription Drug Cost Regulations^[37] of Ontario, and noted that if the physician, for instance, simply puts the trade name of the medication "Lopresor" on the prescription, it was up to the pharmacist to give out an interchangeable product. The patient may, however, he said, exercise the choice himself - as subsection 4(2) on the Ontario Act stated:

4(2) If a prescription directs the dispensing of a specific interchangeable product, the dispenser, on the request of the person for whom the product was prescribed or the person presenting the prescription, shall dispense in its place another product that is designated as interchangeable with it.

                                                                                                                   [Emphasis in text]

[42]            The Ontario Act required the pharmacist to inform the consumer -- the patient -- of his right to indicate which product he would like. The pharmacist's duty to provide this information was fulfilled by the posting of a visible notice to that effect. From this, Gonthier J. concluded that patients had a choice. They needed information and protection just as other consumers do. He then stated:^[38]

[...] Not including him in the clientele covered by the passing-off action, in my opinion, divests him of part of his rights as an individual. He is deprived of the means of protecting himself as an informed person. In Travaux de l'association Henri Capitant on consumer protection, Prof. Lilkoff says (translation):

     Canadian food and drug legislation dates from January 1, 1875 but it is only since 1968 that the Department of Health has sought to take on a more active role in consumer protection. The Act is designed to protect the consumer against products dangerous to health and against misleading advertising and fraud in the use and sale of food, drugs, cosmetics and therapeutic devices.

Emphasis added. "Rapport sur la protection du consommateur en droit pénal canadien", in Travaux de l'association Henri Capitant des amis de la culture juridique française, vol. 24, La protection des consommateurs (Paris: Dalloz, 1975), 331, at pp. 338-9.)

     When the present Act (Food and Drugs Act, supra) states in s. 9(1):

9(1) No person shall label, package, treat, process, sell or advertise any drug in a manner that is false, misleading or deceptive or is likely to create an erroneous impression regarding its character, value, quantity, composition, merit or safety.

there is no basis for saying that it applies only to health care professionals (emphasis added).

[43]            He concluded ultimately that patients/customers were to be included together with the physicians, dentists and pharmacists. He said in the order issued by the Court:^[39]

    [...] with respect to the marketing of prescription drugs, a plaintiff in an action for the alleged passing off of a prescription drug must establish that the conduct complained of is likely to result in the confusion of physicians, pharmacists or patients/customers in choosing whether to prescribe, dispense or request either the plaintiff's or the defendant's product.

                                                                                                                   [Emphasis added]

[44]            The Supreme Court of Canada, in effect, reversed an earlier decision of the Ontario Court of Appeal in Ayerst, McKenna & Harrison Inc. v. Apotex Inc.^[40] which had limited the consumer market for prescription drugs to the professionals.

[45]            Under the caption "(iv) The prescription drug market is like any other", Gonthier J. stated:^[41]

As I mentioned earlier, a manufacturer that wishes to succeed in a passing-off action must usually show that its product has acquired a secondary meaning with its customers, the public, and that the competing product is likely to create a risk of confusion in the public mind. There is no reason for the approach to be any different when the producer or manufacturer is a pharmaceutical laboratory. The prescription pharmaceutical products business is not so fundamentally different from other areas of commercial activity that special rules should apply to it. The courts have no reason in law to deprive pharmaceutical laboratories of means of proof available to other industries.

                                                                                                                   [Emphasis added]

[46]            Gonthier J. gave citations from case law and scholarly comment. The case law he cited is important in the case at bar because the final result in each decision turned on the evidence that had been adduced. He concluded on this point:^[42]

     Whether the get-up of drugs is associated with their effect rather than their commercial origin is a question of fact the consideration of which must be left to the trier of fact, as happens in other commercial fields. That definitely does not mean in law that evidence should be limited to physicians, pharmacists and dentists.

                                                                                                                   [Emphasis added]

[47]            The Ciba-Geigy decision stands, in my view, for two propositions. The first is that the target clientele or the relevant universe in the pharmaceutical field encompasses the patient. Gonthier J. is, however, unclear as to whether "the patient" is limited to that person who was given a prescription and who buys and is expected to swallow the capsule and who, therefore, is in a position to make a choice under the relevant legislation. The order issued at the end of his reasons seems to indicate that Gonthier J. was dealing with the first meaning since he used such words as "patients/customers" and "request". But, according to Gonthier J. the word "patient" might also mean "anyone who has an actual or potential, immediate or remote, connection with the product,"^[43] and anyone whom the Food and Drugs Act is meant to protect.^[44]    He invoked this wider category when he cited "Nadeau and Nadeau" in the general context of passing-off and when, while dealing with the pharmaceutical field, he cited Professor Lilkoff's study on consumer protection.

[48]            The second proposition which Ciba-Geigy stands for is that the general rules in a passing-off action apply to the prescription drug market without any difference or exception. This means that similarities in shape, size and colour of a capsule may found a passing-off action if the three necessary components are met. Ciba-Geigy does not stand, however, for the proposition that pharmaceutical manufacturers cannot, as a start, adopt the trade dress or get-up of capsules found already in the market.^[45] What Ciba-Geigy stands for is that they cannot do so if the prescription drug has acquired distinctiveness and the copying is likely to lead to confusion.

[49]            While Gonthier J. gave a long description of the state of affairs in the pharmaceutical field, I cannot conclude, as the appellants suggest, that the Supreme Court of Canada is of the view that the appearance of a prescription drug constitutes in every case a trade-mark right. Each case must be demonstrated by relevant evidence.

(b) Application of the Ciba-Geigy decision to the judgment appealed from

[50]            The trial judge characterized the Ciba-Geigy case as "of vital importance to the present litigation".^[46] She did not ignore it. The real question before us is whether she applied it correctly. I will first consider whether she correctly defined the relevant universe. I will then deal with her application of the passing-off action.

(i) Relevant universe

[51]            The dispute between the parties concerns the extent to which potential patients and members of the general public are to be included as the target clientele in a passing-off action in the field of prescription drugs. The appellants claim the trial judge erred in defining that relevant universe. That universe, they claim, includes patients who are actual users of fluoxetine. It includes, in the category of potential customers, those members of the public who have suffered depression and have seen their doctors. But, they say, it does not include the rest of the public, contrary to the approach taken by the trial judge.

[52]            If the appellants were right in so defining the potential market, the result would be to exclude those members of the public, who would have been made aware of the enormous publicity the marketing of Prozac has attracted, who might themselves or a member of their family or their friends be inclined to depression, and who might think they should seek the advice of their doctors about their state of health. This group of persons can still reasonably be considered among potential customers of fluoxetine generally, and of Prozac as a possibility. It is true that these persons have not requested a "choice". But they are persons who are potentially within that class. They are certainly also those meant to be protected under the Food and Drugs Act to which Gonthier J. referred.

[53]            The trial judge cited statistics which indicated "that fifteen per cent of all Canadians will, at some point in their lives, suffer from depression. Of these, many may never seek medical assistance. Of those who do, not all will be treated with a prescription drug. Some of those for whom drug treatment is prescribed will be given an anti-depressant, others fluoxetine."^[47] Bearing this in mind and considering the amount of publicity generated when Prozac reached the market, it was certainly open to her, on the particular facts of this case, to conclude that the conditions made identification of potential customers difficult, if not impossible.^[48] It became difficult to determine who, in the general public, were the potential customers. She was careful, however, to consider survey results derived from both the general public and sub-groups such as "Prozac Knowledgeable", "Prozac Aware", "Anxiety Sample" and "Claimed Prozac Exposed". After a review of all four studies, she concluded:^[49]

     Overall, the survey results show very little evidence to support a claim of association between capsule colours being the same and capsule source being the same, very little evidence of an association between the Prozac name and the actual colours of the Prozac capsule, and very little evidence of an association between the capsule colours green and pale yellow and the Prozac name.

[54]            Therefore, whether the relevant universe was a narrow one, as submitted by the appellants, or a larger one, including the potential patients, makes no difference in the case at bar. The conclusion the trial judge arrived at is the same. Her finding of fact does not reveal any "palpable and overriding error".^[50]

(ii) Paragraph 7(b) of the Trade-marks Act

[55]            With respect to appearance or get-up, the appellants claim they have given to their capsules a specific trade dress which is a distinguishing feature of their goods, and that their goods are known on the market and have acquired a reputation by reason of that distinguishing feature.^[51] They assert that the colours of their capsules, green and cream, green and pale grey, were unique in the anti-depressant market in 1989 (with the exception of Lithane which is sometimes described as an adjunct to the treatment of depression, and Librium which is an anti-anxiety medication).^[52] By adopting those same colours, they say, the respondents have created a situation which is likely to cause confusion.

[56]            The trial judge found, as a fact, that the appellants had not proven that the capsule appearance had acquired the requisite reputation in the market place as a distinguishing feature of their product.^[53] She concluded that many of the consumers who have taken the appellants' product and are filling a new prescription, or are obtaining a refill of an existing prescription, will associate the capsule's appearance with the character of the medicine and not its trade source or provenance.^[54] She added:^[55]

[...] Those consumers will not be confused by the defendants' products being of similar appearance to that of the plaintiff.

[57]            Her fact finding is not one that warrants our intervention.

(iii) "Significant" likelihood of confusion

[58]            The appellants also take issue with trial judge's finding of no "significant likelihood of confusion" which, they say, is not the proper legal test. They correctly point out that the word "significant" is not found in paragraph 7(b) of the Act. The case law surrounding the passing-off requires, however, evidence of confusion that exceeds the de minimis threshold. The trial judge based her statements that there must be "some evidence adequate to support a finding of likelihood of confusion" and that "What degree of likelihood of confusion will support a passing-off claim varies with the particular facts of each case" upon accurate interpretations of the authorities which she cited.

[59]            There are no magic words to be used to determine the level to be reached. One can find in the case law phrases such as "significant likelihood of confusion",^[56] "no reasonable likelihood of confusion",^[57] "real likelihood of confusion",^[58] "substantial number"^[59] (of prospective consumers would be deceived) and "significant percentage".^[60]

[60]            Her approach is also supported by the passage from Roche Products Ltd. v. Berk Pharmaceuticals Ltd. that was quoted by the Supreme Court in Ciba-Geigy to the effect that:^[61]

... What must first be shown is that a substantial body of patients would have come to attribute the plaintiffs' product, by reason of colour, shape, size, number, groove on one side, and some word or the other, to some one trade source or provenance..."

                                                                                                                   [Emphasis in text]

[61]            It is clear that in making her determination, the trial judge correctly directed herself to the question of whether the evidence established that there was a likelihood of confusion that surpassed the de minimis standard.

[62]            I conclude on the whole that the trial judge did not misdirect herself in the law and that it was open to her, as a trier of fact, to conclude that the appellants had not met the requirements of the passing-off action.

(iv) Section 50 of the Trade-marks Act

[63]            The trial judge found that no licence agreement existed between Lilly U.S. and its Canadian subsidiary granting to Lilly Canada the right to use the capsule appearance as a trade-mark.

[64]            Since I find that the appellants cannot claim any trade-mark with respect to the size, shape and colour of their Prozac capsules, I need not pronounce on whether the trial judge correctly interpreted the law and particularly subsection 50(1) of the Act in so doing. But, considering the importance of the question and the fact that the trial judge did pronounce on that question, I plan to deal with the issue.

[65]            I do not share the trial judge's conclusion. I find that the 1991 Agreement signed between Lilly USA and Lilly Canada and, in case of doubt, the 1995 Agreement and subsection 50(1) of the Act, dispose completely of the respondents' argument that the use by Lilly Canada of the alleged trade-mark rights of Lilly US was unlicensed and that the appellants' claim should have failed for that reason.

[66]            The statement of claim asserted that Lilly U.S. was the owner of the trade-mark rights in the capsule appearance. Lilly U.S. was not, however, a registered owner of the capsule appearance in Canada. Only Lilly Canada had been selling Prozac in Canada.

[67]            Counsel for the appellants claim that the use by Lilly Canada of the alleged trade-mark rights owned by Lilly U.S. was licensed by the 1976 Agreement and then by the 1991 Agreement between Lilly U.S. and Lilly Canada. In any event, they say, any defect would have been cured by an amendment of November 1995 to the 1991 Agreement read in conjunction with subsection 50(1) of the Act. The 1995 Agreement was meant to "confirm" that Lilly Canada was and had always been licensed by Lilly to use the trade dress, i.e. the colour, shape and size of the Lilly products that had been sold and were sold by Lilly Canada in Canada.

[68]            It can reasonably be said, in my view, that the 1991 Agreement, which precedes the adoption, on June 9, 1993,^[62] of subsection 50(1) of the Act, licensed the use by Lilly Canada of the alleged trade-mark rights in the Prozac capsules and that there was no need for the "confirmation" Agreement of 1995.

[69]            The first recital of that Agreement indicates the meaning to be given to "Lilly Products".

1. Lilly represents and warrants that for Canada, it has the exclusive right to grant licenses to enable the licensee to make, have made, use and sell certain products, including the right to use within Canada, certain patents, trademarks, formulas, discoveries, processes, designs, and other scientific and technical data (whether patentable or not relating to such products and to their preparation, manufacture, processing, and packaging, the products being hereinafter referred to as "Lilly Products", by which is meant products to be sold in Canada under trademarks, brand names, or other designations owned by or employed by Lilly.

[70]            Article 1 of the Agreement then states:

Section 1.1Subject to the terms and conditions of this Agreement, Lilly hereby appoints Lilly Canada the authorized distributor in Canada of Lilly Products.

Section 1.2Lilly further grants to Lilly Canada a non-exclusive sublicense (without right of further sublicense except as further granted in writing by Lilly) under the Canadian patent applications and patents listed in Schedule "A" attached hereto (as amended from time to time by mutual consent), to make, have made, use or sell, and/or import Lilly Products whose preparation is covered by the patent applications and patents.

Section 1.3Lilly further grants to Lilly Canada the right and licence to apply to Lilly Products the house marks, trademarks, brand names, and/or other designations for Lilly Products as may have been established by Lilly, as listed in Schedule "B" attached hereto (as amended from time to time by mutual consent). Those house marks, trademarks, brand names, and/or other designations shall be applied only to Lilly Products sold under the label of Lilly or Lilly Canada. The Lilly Products sold under those house marks, trademarks, brand names, or other designations shall conform to those standards Lilly may set from time to time as to specifications and quality.

Section 1.4Lilly further shall disclose to Lilly Canada and grant a non-exclusive licence to Lilly Canada to enable it to make, have made, use and sell Lilly Products for distribution in Canada complete information on Lilly Products. Such information shall include formulas, processes, designs, and other scientific and technical data, including the manner for preparation, manufacture and compounding the Products into pharmaceutical forms, the labelling and packaging of Lilly Products, engineering and manufacturing drawings, specifications, diagrams, instruction for the construction and operation of factories, and all other scientific and technical data that Lilly is free to disclose in connection therewith, all of which are referred [...]

                                                                                                                   [Emphasis added]

[71]            Under section 1.4 of Article 1, Lilly U.S. undertook to disclose to Lilly Canada complete information on Lilly products and to grant to it a non-exclusive licence to sell its Lilly products in Canada. The information to be disclosed comprised not only the manner for preparation, manufacture and compounding the products into pharmaceutical forms, but also the labelling and packaging of such products. The labelling and packaging, it seems to me, should not be limited to the labelling and packaging of the bottles themselves, but could extend to the capsule with its specific size, shape and colour.

[72]            If this is too generous an interpretation, subsection 50(1) of the Act cures the situation.

[73]            Section 50 of the Act reads thus:

[74]            The trial judge said:^[63]

     There is no doubt that Lilly U.S. controlled Lilly Canada's use of the capsule appearance and directed it to use the green and cream and green and grey capsules. It controlled as well the fluoxetine product that would be marked in those capsules. This does not, however, mean that direction was given to use the capsule appearance as a trade-mark. For example, I draw an analogy to a situation in which a parent company might direct a subsidiary to use a certain type of cardboard container, having a certain strength and configuration, for the shipment of product. Even if the parent directed such use and provided information concerning how to construct the containers that would not mean that the use by the subsidiary was as a trade-mark.

[75]            Her finding that "Lilly U.S. controlled Lilly Canada's use of the capsule appearance and directed it to use the green and cream and green and grey capsules" is key to the application of section 50 of the Act.

[76]            Subsection 50(1) of the Act must be read in the light of the pre-existing registered user provisions it replaced. Those provisions required that a licensee be a registered user if the trade-mark was registered. They came into force on July 1, 1954^[64] following a Report of the Trade Mark Law Revision Committee to the Secretary of State of Canada.^[65] The Committee was appointed on October 28, 1947 to review the Unfair Competition Act, 1932. The Committee submitted a draft statute which was subsequently enacted by Parliament subject to some insignificant amendments.^[66]

[77]            In its Report, the Committee indicated what the previous law was in relation to the transfer and licencing of trade-marks, and why it felt a change was necessary. The following extract is indicative of the mischief the draft statute was meant to cover:^[67]

28. Transfer and Licensing of Trade Mark Rights                                                             There was almost complete unanimity in the opinions received by us condemning the restrictive character of the existing law relating to the transfer and licensing of trade marks. According to the classical concept of trade mark law, a trade mark was to give to the prospective purchaser an indication of the trade source of goods. So long as this view was maintained, any interruption of the line indicative of origin of the goods was regarded as vitiating the trade mark. Thus, any assignment of a trade mark not accompanied by the goodwill of the whole of the business concerned in the goods in association with which the trade mark was used, or any licensing of its use by others than the trade mark owner, rendered the mark, in the eyes of the law, non-distinctive and hence invalid and open to use by anyone. (See, for example, Bowden Wire Ltd. v. Bowden Brake Co. (1914), 31 R.P.C. 385; Pinto v. Badman (1891), 8 R.P.C. 181; Re Vulcan Trade Mark (1914), 22 D.L.R. 214, 15 Ex. C.R. 265; 24 D.L.R. 621, 51 S.C.R. 411; Robert Crean & Co. v. Dobbs & Co., [1930], 3 D.L.R. 22, S.C.R. 307).

     Judicial efforts have been made to ameliorate the harshness occasioned by the strict opinion of this rule (Siegel Kahn Co. v. Peggy Sage Inc., [1935], 2 D.L.R. 225, Ex. C.R. 1; Good Humor Corp. of America v. Good Humor Food Products Ltd. et al., [1937], 4 D.L.R. 145, Ex. C.R. 61; Magazine Repeating Razor Co. of Canada et al. v. Schick Shaver Ltd., [1939], 2 D.L.R. 17, Ex. C.R. 108; [1940], 3 D.L.R. 129, S.C.R. 465). The necessity of transferring rights in a trade mark only if accompanied by the goodwill of the business in association with the wares in respect of which the mark has been registered was given statutory effect in s. 44(2) of The Unfair Competition Act, 1932. But the trend of thought has been increasingly in favour of discarding the indication of origin theory of trade marks. It has gradually come to be recognized that this theory is not based so much on common sense as had been supposed. This is obvious when it is borne in mind that a trade mark may point to the manufacturer as a distinct person, to the place of manufacture, to a certain natural source or to a certain quality or secret formula. Origin and ownership are by no means the sole indication of a trade mark. We feel that modern commercial development has demonstrated that the public is not so much interested in source as in the maintenance of quality, kind and type. Rules which were valid in the days of simple and closely connected trade relationships show many discrepancies and absurdities when forced to accommodate themselves to the established circumstances of modern commercial practice and usage. The indication of origin theory breaks down when a business is sold to a newcomer or changes hands by the modern process of stock transfer, and it has seemed to us that there is no logical reason for fettering the disposition or use of trade marks with artificial and antiquated rules and restrictions. We are all the more confirmed in this view by a recognition of the fact that considerable numbers of trade marks now registered in Canada would necessarily be held invalid if examined in the light of the manner in which they have been used commercially. Many Canadian subsidiaries of foreign corporations have so used their trade marks as to invalidate them even though the purchasing public has in no way been deceived or harmed. These facts let us to make a searching examination into this branch of the subject. The theory of the free assignability of trade marks on the basis of their pure function as offering a guarantee of quality rather than indication of origin has been gaining wide acceptance in the legal thought not only of this country, but particularly in continental Europe (see, for example, Martin-Achard, La Cession Libre de la Marque, Geneva, 1946). We, therefore, felt it proper to endeavor to probe public opinion on this point and have found the overwhelming weight of such opinion in favour of abolishing these artificial restrictions and of permitting limited licensing of trade marks.        [Emphasis added]

[78]            The registered users provisions themselves gave rise to difficulties.^[68] The concept of registered user was revoked on June 9, 1993 and replaced by subsection 50(1).

[79]            The appellants rest their case on the 1991 Agreement and on the retroactive character of subsection 50(1) of the Act, as evidenced by the words "and is deemed always to have had, the same effect as such a use... of the trade-mark in that country by the owner". They claim that section 50 of the Act is expressly stated to be retroactive so that any licensed use in accordance with the terms of section 50, even licensed use predating its enactment, enures to the benefit of the trade-mark owner. In addition, they say that the trial judge failed to recognize that even if the 1991 Agreement did not specifically grant a trade-mark licence, the direction to use the trade dress was an oral licence. They find support in the case of TGI Friday's of Minnesota Inc. v. Registrar of Trademarks,^[69] where this Court held that section 50 is not only retroactive, but includes an oral licence to use the mark.

[80]            I do not accept the concluding part of the following paragraph of the trial judge's decision:^[70]

     Originally, trade-mark rights could not be conveyed to another in the absence of the goodwill of the business also being conveyed. To do so would deceive the public as to the source of the goods and destroy the trade-mark. This was modified in 1954 by the inclusion of registered user provisions in the Trade-marks Act. The right to use a trade-mark by another would not destroy the owner's rights if the use was licensed by the owner and that license was registered. Difficulties still existed under this system. Licences could be granted but there might be neglect or delay in registering them. In that context section 50 was enacted. The deeming provisions thereof cure the situation in which a licence existed but had not been registered. The provision does not deem use prior to the granting of a licence to be use that ensures to the benefit of the owner.

                                                                                                                   [Emphasis added]

[81]            The trial judge ignored completely the direction to use the trade dress given by Lilly U.S. to Lilly Canada, and the element of control retained by Lilly U.S. which was explicit in the 1991 Agreement as key to the retroactive effect of subsection 50(1) of the Act. Moreover, subsection 50(2) deals precisely with PMS's licensed use of the same get-up. As was stated earlier, PMS sold the appellants' capsules assisted by marketing efforts by Lilly Canada directed at doctors and pharmacists.^[71]

[82]            The respondents' position rests on two grounds which I cannot accept. The respondents plead that nowhere in the Agreement was there any mention of a grant of right on the shape, size and colour of the capsule fluoxetine. Section 1.3 of Article 1 of the Agreement makes it clear, they say, that only the items listed in Schedule "A" were transmitted. I find this interpretation of the 1991 Agreement too rigid. An agreement, being a contract, is meant to reflect the intent of the parties. The possibility that the law might recognize a trade-mark in the shape, size and colour of a pharmaceutical capsule was not known to them at the time the 1991 Agreement was signed. This concept came into the open with the decision of the Supreme Court of Canada in Ciba-Geigy, on October 29, 1992. However, the words used by them can reasonably bear the interpretation I have given them.

[83]            The second ground invoked by the respondents is the decision of Crean & Co. Ltd. v. Dobbs & Co.^[72] This very decision is among those cited in the Report of the Trade Mark Law Revision Committee which the 1954 legislation was meant to overturn.

Conclusion

[84]            Having concluded that it was open to the trial judge to dismiss the action in passing-off, I would dismiss these appeals with costs.

                                                                                "Alice Desjardins"                

                                                                                                      J.A.

"I agree

J. Edgar Sexton J.A."

"I agree

Karen R. Sharlow J.A.