Date: 20251128

Dockets: A-26-24

A-27-24

Citation: 2025 FCA 212

CORAM:

BOIVIN J.A.* WOODS J.A. LASKIN J.A.

 

Docket: A-26-24

BETWEEN:

SAMSUNG BIOEPIS CO., LTD.

Appellant Respondent to cross-appeal

and

NOVARTIS AG and NOVARTIS PHARMACEUTICALS CANADA INC.

Respondents Appellants by cross-appeal

Docket: A-27-24

AND BETWEEN:

BIOGEN INC., BIOGEN MA INC. and BIOGEN CANADA INC.

Appellants Respondents to cross-appeal

and

NOVARTIS AG and NOVARTIS PHARMACEUTICALS CANADA INC.

Respondents Appellants by cross-appeal

Heard at Toronto, Ontario, on May 28, 2024.

Judgment delivered at Ottawa, Ontario, on November 28, 2025

PUBLIC REASONS FOR JUDGMENT BY:

WOODS J.A. LASKIN J.A.

*Boivin J.A. retired from the Court effective September 1, 2025 and did not participate in the disposition of these appeals. Judgment is being given by the remaining members of the panel in accordance with subsection 45(3) of the Federal Courts Act, R.S.C. 1985, c. F-7.

Date: 20251128

Dockets: A-26-24

A-27-24

Citation: 2025 FCA 212

CORAM:

BOIVIN J.A.* WOODS J.A. LASKIN J.A.

 

Docket: A-26-24

BETWEEN:

SAMSUNG BIOEPIS CO., LTD.

Appellant Respondent to cross-appeal

and

NOVARTIS AG and NOVARTIS PHARMACEUTICALS CANADA INC.

Respondents Appellants by cross-appeal

Docket: A-27-24

AND BETWEEN:

BIOGEN INC., BIOGEN MA INC. and BIOGEN CANADA INC.

Appellants Respondents to cross-appeal

and

NOVARTIS AG and NOVARTIS PHARMACEUTICALS CANADA INC.

Respondents Appellants by cross-appeal

*Boivin J.A. retired from the Court effective September 1, 2025 and did not participate in the disposition of these appeals. Judgment is being given by the remaining members of the panel in accordance with subsection 45(3) of the Federal Courts Act, R.S.C. 1985, c. F-7.

PUBLIC REASONS FOR JUDGMENT

This is a public version of confidential reasons for judgment issued to the parties. The two are identical, there being no confidential information disclosed in the confidential reasons.

WOODS J.A. and LASKIN J.A.

I. Introduction

[1] These reasons address two appeals and two cross-appeals, which were all heard together. In the appeals, the appellants (Samsung and Biogen) ask this Court to set aside a judgment of the Federal Court (2024 FC 52, Pallotta J.). The judgment in question enjoins the appellants and their licensees from using their trademark BYOOVIZ in association with a drug for use in ophthalmology and the prevention and treatment of ocular disorders and diseases, and awards the respondents (collectively, Novartis) $20,000 in nominal damages.

[2] The application judge granted this relief after concluding that the appellants’ use of the BYOOVIZ trademark violated rights of Novartis in its registered trademark BEOVU under s. 20(1) (trademark infringement) and s. 7(b) (passing off) of the Trademarks Act, R.S.C. 1985, c. T-13 (TMA). The appellants seek to have this judgment set aside. The cross-appellants (Novartis) seek to have the remedy varied to order that Novartis be entitled to recover at its election either its damages or the appellants’ profits, to be determined on a reference to the Federal Court under rules 153 and following of the Federal Courts Rules, S.O.R./98-106.

[3] Though the appellants’ arguments in the main appeals differ in some respects, they share the following positions:

(1) the application judge erred by treating patients as relevant consumers in the “likelihood of confusion” analysis;

(2) the application judge erred by failing to take into account the consumer attitudes of ophthalmologists and pharmacists, which are characterized by specialized knowledge and heightened attention, in the “likelihood of confusion” analysis; and

(3) the application judge erred in her consideration of the surrounding circumstances in assessing likelihood of confusion under s.6(5).

[4] The application judge referred to the submissions of the appellants without distinguishing between Samsung and Biogen. We adopt the same approach in these reasons. As explained below, the main appeals and the cross-appeals will be dismissed.

II. Background

A. Statutory framework

[5] Relevant provisions of the TMA are set out in an appendix to these reasons. We also refer to some of them in the discussion that follows.

[6] A trademark is “a sign or combination of signs that is used or proposed to be used by a person for the purpose of distinguishing […] their goods or services from those of others”: TMA, s. 2. Registration of a trademark grants its owner “the exclusive right to the use throughout Canada of the trademark”: TMA, s. 19.

(1) Trademark infringement under subsection 20(1)

[7] A trademark is infringed where any person who is not otherwise entitled to do so under the TMA sells, distributes or advertises any goods or services in association with a confusing trademark or trade name: TMA, s. 20(1)(a).

[8] A trademark is “confusing” with another under s. 20(1)(a) if “the use of both trademarks in the same area would be likely to lead to the inference that the goods or services associated with those trademarks are manufactured, sold, leased, hired or performed by the same person”: TMA, s. 6(2).

[9] The test for confusion is “a matter of first impression in the mind of a casual consumer somewhat in a hurry who sees the [mark], at a time when [they have] no more than an imperfect recollection of the [prior] trade-marks, and [do] not pause to give the matter any detailed consideration or scrutiny, nor […] examine closely the similarities and differences between the marks”: Masterpiece Inc. v. Alavida Lifestyles Inc., 2011 SCC 27 at para. 40[Masterpiece]; Veuve Clicquot Ponsardin v. Boutiques Cliquot Ltée, 2006 SCC 23 at para. 20 [Veuve Clicquot].

[10] In determining whether trademarks are confusing, courts are required to “have regard to all the surrounding circumstances”: TMA, s. 6(5). These surrounding circumstances include, but are not limited to, those set out in that provision:

(a) the inherent distinctiveness of the trademarks or trade names and the extent to which they have become known;

(b) the length of time the trademarks or trade names have been in use;

(c) the nature of the goods, services or business;

(d) the nature of the trade; and

(e) the degree of resemblance between the trademarks or trade names, including in appearance or sound or in the ideas suggested by them.

[11] The assessment is context-specific; different circumstances will be given different weight: Veuve Clicquot at para. 21.

(2) Passing off under para. 7(b)

[12] As set out by the Supreme Court in Kirkbi AG v. Ritvik Holdings Inc., 2005 SCC 65 at para. 66, to succeed in a claim for passing off under s. 7(b), a claimant must establish three elements:

(1) the plaintiff (or applicant) possesses goodwill in the trademark;

(2) the defendant deceived the public by misrepresentation; and

(3) the plaintiff suffered actual or potential damage through the defendant’s actions.

[13] Misrepresentation for the purposes of the second element of passing off is established if the defendant’s trademark is likely to be confused with a trademark owner’s trademark: Sadhu Singh Hamdard Trust v. Navsun Holdings Ltd., 2016 FCA 69 at para. 21.

B. The BEOVU trademark

[14] Novartis owns a Canadian trademark registration for BEOVU. It uses the trademark in association with an anti-vascular endothelial growth factor (anti-VEGF) biologic drug. The drug is approved in Canada for treating neovascular age-related macular degeneration (wet AMD).

[15] Anti-VEGF drugs are prescription drugs, in liquid formulation. They must be refrigerated until they are administered. They must be administered by a qualified ophthalmologist by injection, using a syringe, directly into the eye.

[16] Apart from BYOOVIZ, there were other anti-VEGF biologics available in Canada, including BEOVU. One of these, LUCENTIS, is also manufactured by Novartis.

C. Evidence of clinical practice

[17] The evidence on the applications, tendered through affidavits and cross-examination transcripts, included expert evidence from four ophthalmologists. One gave evidence for Novartis, and the three others, for the appellants. Novartis also submitted, among other evidence, that of the operator of a specialty pharmacy. We will briefly summarize some salient points that these witnesses’ evidence illuminates, drawing substantially on the summaries set out in the reasons of the application judge (at paragraph 13).

[18] Dr. Arif Samad, the ophthalmologist who testified for Novartis, practises in Halifax. He stated that it is the ophthalmologist, who is familiar with the available products and trademarks, who selects the anti-VEGF drug to be administered, and that clinic staff routinely identify the drugs by their brand names. The chosen drug is received from the pharmacy in a box or, if the syringe is filled at a hospital pharmacy, in a plastic bag. Patients may also request a specific drug based on their own research. Dr. Samad did not specify how often this has occurred.

[19] Dr. Robert Devenyi practises in Toronto. He deposed that he does not prescribe BEOVU because it has been shown to cause a condition that can lead to vision loss. Two other drugs that he does regularly prescribe come from the manufacturer in their original packaging. A third drug that he sometimes prescribes comes repackaged by a specialty pharmacy, which will supply syringes bearing a label with the drug’s brand name.

[20] Dr. Devenyi further deposed that patients “almost never” ask for a specific anti-VEGF drug. When a patient attends his clinic for treatment, clinic staff will place a syringe with the proper drug in the treatment room. Before administering it, he checks to ensure that it is the correct drug. He also tells patients what drug he is administering when he first treats them or when he switches to a different medication.

[21] Dr. Farzin Forooghian practises in Vancouver. He stated that his patients almost never ask for a specific anti-VEGF drug. If a patient asked for one that is out of the sequence he follows, he would explain his duty to use drugs that were authorized by the applicable provincial program and most cost-effective.

[22] Dr. Forooghian stated that all anti-VEGF drugs in B.C. are provided by compounding pharmacies. The pharmacy he uses orders the drugs from a manufacturer or distributor, removes them from the original packaging, repackages them in new syringes, places the syringes in colour-coded bags, and delivers them to his clinic where they are refrigerated until used. The bags and syringes bear the name of the drug.

[23] Dr. Forooghian went on to state that he does not see the original manufacturer packaging. When he enters the room to administer the drug, the drug will already be there in its colour-coded bag. He orally confirms the patient’s name and date of birth, cross-references the drug shown on the patient’s chart with the colour of the bag, and verifies visually that it is the correct drug. He stated that he always tells the patient the name of the drug and always tells the patient if he has changed the drug.

[24] Dr. Frank Stockl practises in Manitoba and northwestern Ontario. He stated that in Manitoba, anti-VEGF drugs are delivered to a compounding pharmacy in vials, the original manufacturer packaging. The pharmacy transfers the drug into syringes, which are then labelled with the name of the drug and placed in a colour-coated plastic cassette and sent to the clinic using a refrigerated courier service. In northwestern Ontario, anti-VEGF drugs are obtained from pharmaceutical distributors and may be delivered to the clinic in the manufacturer’s packaging, with a label applied by the pharmacy. The drug name is also etched into the syringe. Drugs supplied in a vial are repackaged into syringes, which bear a label indicating the name of the drug.

[25] In Dr. Stockl’s clinic, a clinical assistant will take the correct drug from the refrigerator, remove it from its package, and place the syringe in the treatment room. A technician verifies name and date of birth, confirms the eye to be treated, and crosschecks the drug. When Dr. Stockl enters the room, he inspects the label to verify the drug, and confirms the drug name and the eye to be treated before the drug is administered. When the drug is to be administered in a hospital, a nurse will confirm, in Dr. Stockl’s presence, the patient’s name and date of birth, the drug to be administered, and which eye or eyes are to be treated.

[26] Dr. Stockl stated that in the vast majority of cases, the patient will not ask for a specific anti-VEGF drug. If a patient asks about a certain drug, he will usually respond that the choice is constrained by rules laid down by the government, but if they still want a specific drug, they can pay for it out-of-pocket or through third party insurance if it is covered.

[27] Joseph Connolly is the co-CEO of a specialty pharmacy in Ontario, which he stated supplies more than 35 percent of ophthalmology clinics across Canada. According to Mr. Connolly, most ophthalmology clinics source their drugs from pharmacies. His pharmacy takes measures aimed at ensuring that patients receive the correct drug. These include affixing a prescription label to the container in which drugs are dispensed that includes their name, strength and manufacturer, unless the prescriber has directed otherwise, and the names of the prescriber and patient. In Mr. Connolly’s experience, the average age of patients who are prescribed anti-VEGF drugs is 81.

[28] While it is possible for patients to pick up their prescriptions from a pharmacy, in almost every case anti-VEGF drugs are sent directly to the clinic where they are to be administered. According to Mr. Connolly, patients typically do not see the writing that may be present on the syringe.

III. Decision of the application judge

[29] Before the application judge, Novartis submitted that the appellants’ unauthorized use of the BYOOVIZ trademark in association with an ophthalmologic drug (1) infringed Novartis’s rights in the BEOVU trademark, with which it was confusingly similar, contrary to ss. 19 and 20 of the TMA, (2) constituted passing off, contrary to s. 7(b) of the TMA, and (3) was likely to have the effect of depreciating the value of the goodwill attaching to Novartis’ mark, contrary to s. 22 of the TMA. It also sought injunctive relief and damages. The s. 22 claim is not pursued in this Court.

[30] In reliance on authority, including this Court’s decision in Sandhu Singh Hamdard Trust v. Navsun Holdings Ltd., 2019 FCA 295 at para. 20, the application judge accepted that s. 19 addresses the use of a trademark that is identical to a registered trademark, while it is s. 20 that applies where a trademark is confusing in light of, but not necessarily identical to, a registered mark. Since the BEOVU and BYOOVIZ marks were not identical, the application judge carried out her infringement analysis by reference to s. 20.

A. The application judge found that the appellants contravened section 20

[31] The application judge first noted (at paragraph 23) the parties’ agreement that the outcome of the proceeding would turn on whether there was a likelihood of confusion between the two trademarks. She then also noted the parties’ disagreement on the identity and attitude of the relevant consumer for the likelihood of confusion analysis, as well as on the relevant point in time for assessing its likelihood.

(1) Relevant consumers

[32] Novartis submitted that physicians, pharmacists, and patients were all relevant consumers for purposes of the likelihood of confusion analysis. The appellants argued that patients are not—that “trademark use in the context of highly specialized anti-VEGF medications does not transcend to the patient level.”

[33] In rejecting this submission, the application judge (at paragraph 33) expressed the view that her conclusion followed from the principles the Supreme Court set out in Ciba-Geigy Canada Ltd. v. Apotex Inc., 1992 CanLII 33 (SCC) [Ciba-Geigy]. There the Court held, in the context of the common law tort of passing off, that patients purchasing prescription tablets at a drug store are relevant consumers in determining whether a competing product is likely to create a risk of confusion.

[34] The application judge rejected (at paragraph 34) the attempt on the part of the appellants to distinguish Ciba-Geigy on the basis that BEOVU and BYOOVIZ, unlike the tablets in Ciba-Geigy, were not interchangeable. Instead, she stated, what the Supreme Court had found significant was that “the patient has a choice, regardless of whether the choice is great or small, easily exercised or not”: Ciba-Geigy at 150.

[35] After quoting this language from Ciba-Geigy, the application judge went on to observe (at paragraphs 34-35) that patients receiving anti-VEGF medications may exercise choice in the following ways.

• Patients are told the brand name of the drug that will be administered to them.

• Patients sign a consent form agreeing to be injected with an anti-VEGF medication.

• Patients, upon hearing the trademark, and possibly also upon seeing the trademark on the syringe, can exercise the choice to refuse administration.

• At least at some clinics, patients can choose between different anti-VEGF medications.

[36] The application judge also expressed the view that in light of general principles of trademark law, patients are relevant consumers of anti-VEGF drugs. This was so, she stated, even when patients do not request a specific anti-VEGF medication, which the evidence suggest occurs the majority of the time. They are informed of the ophthalmologists’ choice of drug and they consent to being injected with the drug that is identified by its trademark.

[37] She found that the facts of NeoRx Corp. v. Cytogen Corp., 1995 CanLII 19221 (CA TMOB) were distinguishable, at least because the decision does not mention any evidence that patients were told the imaging medium that would be injected. In any event, she did not find the Trademarks Opposition Board’s limited discussion of how the circumstances of the case were found distinguishable from Ciba-Geigy to be helpful in deciding whether patients were a relevant consumer in the case before her.

[38] Patients are clearly affected by the ophthalmologist’s choice of medication, she stated, and while the funding models vary from province to province, the drug is purchased for the patient’s benefit and use, injected into the patient’s body at regular intervals, and the patient can be switched to a different drug. Patients know what medication they are on by its brand name. In her view, a patient who is told the name of the anti-VEGF drug they will receive is a relevant consumer, and entitled to protection from confusion.

[39] She concluded that on the facts before her, the patient as the “ultimate end-user of the drug” is a relevant consumer for the likelihood of confusion analysis.

(2) Consumer attitude

[40] The application judge began her discussion of consumer attitude by noting that the casual consumer somewhat in a hurry will differ depending on the context, and an analysis of the likelihood of confusion accounts for the relevant consumer’s “attitude”.

[41] Novartis submitted that consumer attitude or mindset increases the likelihood of confusion for the relevant universe of consumers.

[42] For ophthalmologists, Novartis argued that the high-throughput and other demanding elements of their practices should be accounted for in assessing likelihood of confusion. For patients, Novartis’ position was that the attributes of the average patient consumer and the circumstances in which they would encounter the trademarks at issue would increase the likelihood of confusion.

[43] Novartis submitted that patients are a sensitive population. The average patient receiving one of the drugs is generally over 70 and has a range of anxieties and a diagnosis that could lead to blindness among other health challenges. Novartis stated that it is this sensitive population who encounter a relevant trademark when they hear it spoken by their ophthalmologist that must be taken into account in assessing likelihood of confusion between the trademarks at issue.

[44] The appellants argued that doctors and pharmacists are highly skilled professionals who exercise the utmost care. They are also closer in the chain to manufacturers, and are unlikely to be confused as to the source.

[45] The appellants went on to state that if patients are relevant consumers they are informed consumers, decreasing the likelihood of confusion. They claimed that in the rare instances where the patient is involved in the selection of an anti-VEGF medication the patient is an informed one.

[46] They also went on to submit that Novartis’ arguments about patient anxiety and sensitivity were mischaracterized and taken out of context. They argued that patients are not a particularly sensitive population, and ophthalmologists actively mitigate patient anxiety.

[47] The appellants submitted that given the importance of the decision in selecting anti-VEGF medications patients will be more attentive, alert, and less susceptible to being confused.

[48] The application judge found that the evidence did not establish patients would approach a decision about anti-VEGF drugs with reduced care or attention, or that they were more likely to be confused due to their sensitive nature or their age. However, she disagreed with the appellants that patient confusion is less likely because patients must consult with their doctor before they are given anti-VEGF medication, and are therefore deemed to be informed. The application judge held that the evidence did not establish that the ordinary patient is equipped with knowledge or awareness that would lessen the likelihood of confusion on first impression.

[49] For ophthalmologists and pharmacists, the application judge disagreed with Novartis that the attitude of ophthalmologists (or pharmacists) would effectively be more “hurried” due to the nature of their practice. In her view, because of their training and experience, the average ophthalmologist and pharmacist is more likely to perceive smaller differences in drug name trademarks than the average patient.

(3) Likelihood of confusion

(a) 6(5)(a): inherent distinctiveness and extent known

[50] The application judge found that this factor weighed in favour of Novartis. Distinctiveness is the very essence and is the cardinal requirement of a trademark: Mattel, Inc. v. 3894207 Canada Inc., 2006 SCC 22 at para. 75 [Mattel], citing Western Clock Co. v. Oris Watch Co., [1931] Ex. C.R. 64 at 67. The first factor, 6(5)(a), involves a determination of the strength of a mark in terms of its inherent and/or acquired distinctiveness: United Artists Pictures Inc v. Pink Panther Beauty Corp., [1998] 3 F.C. 534 (FCA) at para. 23 [United Artists].

[51] Novartis submitted that the BEOVU mark is a unique and coined term, and thus should be afforded a greater ambit of protection. Novartis also submitted that the BEOVU mark has acquired distinctiveness through use in Canada. Novartis stated it has promoted the BEOVU drug in Canada, spending over $1.5 million in advertising from 2020 to July 31, 2022, and noted that all the healthcare professional witnesses were aware of the BEOVU mark and drug.

[52] The appellants argued that distinctiveness should be a neutral factor that does not favour either party. BEOVU and BYOOVIZ are both coined words, equally unique and equally inherently distinctive. The appellants contended Novartis presented insufficient evidence to establish that the BEOVU trademark has acquired distinctiveness.

[53] The application judge found that the distinctiveness of the BEOVU mark, which is a coined and inherently distinctive trademark, was a factor that favoured Novartis. She did not accept that distinctiveness should be a neutral factor in this case because BYOOVIZ is also a coined word.

(b) 6(5)(b): length of time

[54] The application judge found that the length of time in use was a neutral factor.

[55] Novartis stated it has been using the BEOVU mark in Canada since May 1, 2020 and the BEOVU drug and trademark have become well known to doctors that treat wet AMD.

[56] The appellants submitted that BEOVU had been in use for less than three years and in that time the drug had already failed. Its reputation suffered due to safety concerns, sales declined significantly and Novartis had dismantled its 20-member field force for BEOVU goods.

[57] While BEOVU had been in use longer than BYOOVIZ, and marginally favoured Novartis’ position, she treated length of time in use as a neutral factor.

(c) 6(5)(c): Nature of the goods, services or business; 6(5)(d): Nature of the trade

[58] The application judge found that the 6(5)(c) and 6(5)(d) factors favoured Novartis.

[59] She held that the BYOOVIZ trademark was used with precisely the same goods that are covered by the BEOVU registration and that the nature of the goods and the nature of the trade for the drugs are essentially identical. She did not accept the appellants’ parsing of the relevant market. Both drugs would likely be encountered in the same manner by each of the relevant consumers groups of ophthalmologists, pharmacists, and patients.

[60] The application judge also accepted that the nature of the trade itself may be considered in assessing the likelihood of confusion. She stated that in this regard, patients, ophthalmologists, and pharmacists are likely to have an elevated level of attention, commensurate with the importance of a medical treatment for preventing vision loss. She noted again that in assessing confusion, the average ophthalmologist and pharmacist is more likely to perceive smaller differences in drug name trademarks due to their training and experience.

[61] However, the application judge did not agree with the appellants that the cost of the injections (or the relative pricing of the respective products), differences in packaging, the fact that BYOOVIZ was expected to replace LUCENTIS in the market, or the safety protocols implemented by pharmacists and ophthalmologists were factors that should be considered to decrease the likelihood of confusion. She stated that while anti-VEGF drugs are not inexpensive, the cost is often covered by provincial plans or insurance. She also noted that consumers (even ophthalmologists) do not always see the packaging, which in any event is not markedly different as between the two drugs, and can change. For the application judge, the safety protocols to prevent medication errors and ensure drugs are not confused prior to administration are the kind of research and deliberation that may dispel trademark confusion. However, for her, this did not mean the trademarks are less likely to be confused on first impression.

(d) 6(5)(e): Degree of resemblance between the trademarks, including in appearance or sound or in the ideas suggested by them

[62] The application judge found that BEOVU and BYOOVIZ resemble each other to at least a moderate degree in appearance and ideas suggested. She stated that both are coined, three syllable words, with a similar sequence of letters. The common B-OV sequence provides a degree of resemblance. Each syllable begins with the same letter, in the same sequence. Also, she noted that both terms have a suffix that conveys a similar idea.

[63] She found a high degree of resemblance in sound between BEOVU and BYOOVIZ. She reasoned that both are three syllable words, and each syllable begins with the same or a very similar sound, in the same sequence. She agreed with Novartis that the terms share a similar sound pattern, rhythm, and emphasis. She noted that as coined words, there is uncertainty regarding how the terms are pronounced, and the evidence in this case demonstrates there is variability and/or subjectivity regarding how the terms would be sounded. There is no consensus on how patients, ophthalmologists, or pharmacists would pronounce the words and the evidence is that the pronunciation can change, depending on what is heard or taught. She found that the variability increases the likelihood of confusion.

[64] She also agreed with Novartis that sound is a particularly important factor in this case in view of the evidence regarding oral communication between health care professionals and staff and the role of this communication in informing patients. She concluded that the spoken trademark is an important way consumers encounter the trademarks in question, and the predominant way patient consumers encounter the trademarks in question.

(e) Other surrounding circumstances

[65] The application judge briefly dealt with the parties’ other submissions concerning surrounding circumstances.

[66] She noted that she considered the names of other anti-VEGF drugs as being relevant. However, she concluded there was no state of the register/state of the marketplace evidence to suggest that the presence of other “close” third party trademarks should reduce the ambit of protection afforded to the BEOVU trademark. Then, she found that any of the other additional surrounding circumstances did not significantly change the confusion analysis.

(f) Conclusion on likelihood of confusion

[67] The application judge found that Novartis had established that the appellants’ use of the BYOOVIZ trademark would likely lead to the inference that the goods associated with it and those associated with the registered trademark BEOVU are manufactured and sold by the same person. Applying the relevant test, she found that the trademarks are likely to be confused.

[68] She also noted that while in her view the likelihood of confusion is higher when considered from the patient’s perspective, Novartis had met its burden of establishing a likelihood of confusion for all relevant consumers.

[69] She concluded that Novartis had established that use of the appellants’ BYOOVIZ trademark infringes Novartis’ rights in its registered BEOVU trademark, contrary to s. 20 of the TMA.

B. The application judge found that the appellants contravened subsection 7(b)

[70] The application judge repeated the test for passing off. She stated that Novartis must establish that it possesses goodwill in the BEOVU mark, that the appellants deceived the public by misrepresentation, and that Novartis suffered actual or potential damage through the appellants’ actions.

[71] As for the first element, she found that Novartis had established that it possesses goodwill in the BEOVU trademark. The BEOVU trademark had been used in association with an anti-VEGF drug for more than three years and it is well known by ophthalmologists. In response to the appellants’ argument that any goodwill associated with BEOVU had been diminished as a result of the safety concerns, she stated that the tort of passing off is not limited to trademarks with a sterling reputation. Negative publicity may have reduced the goodwill in the BEOVU mark, but did not extinguish it.

[72] As for the second element, misrepresentation, she noted that this element will be met if Novartis established that the appellants had used a trademark that was likely to be confused with the BEOVU mark. She stated that for the reasons she had set out, Novartis had established a likelihood of confusion.

[73] As for the third element, she found that Novartis had established actual or potential damage as a result of the appellants’ actions: loss of control of the BEOVU trademark by use of a confusingly similar trademark was sufficient to establish actual or potential damage.

C. Remedy

[74] Novartis sought declaratory and permanent injunctive relief, and $100,000 in damages together with pre- and post-judgment interest. The appellants submitted that Novartis was not entitled to any of the claimed relief.

[75] The application judge denied the requested declaratory relief. In her view, this relief was unnecessary since Novartis had not established that a declaration of infringement or passing off would have any practical benefit or utility in this case.

[76] She granted injunctive relief. The appellants and their licensees were permanently enjoined from using the BYOOVIZ trademark in association with the related drug, or any other trademark that is confusingly similar to the registered BEOVU trademark.

[77] As for damages, Novartis was awarded nominal damages in the amount of $20,000.

IV. Analysis

A. Main appeals

(1) Introduction

[78] The central issue in the main appeals is whether the Federal Court erred in determining that the appellants’ use of the BYOOVIZ trademark violated Novartis’ rights in the BEOVU mark under the infringement and passing off provisions in the TMA. The focus of these arguments is on whether the marks are confusing within the meaning of the statutory test set out in s. 6. Some other issues will be discussed later in these reasons.

[79] While the appellants’ arguments are not as organized as they might be, they essentially argue that the application judge erred in determining that patients are relevant to the confusion analysis and in applying the confusion test to patients as well as to pharmacists and ophthalmologists.

[80] With respect to the standard of review, the decision under review is subject to the appellate standards of review as set out in Housen v. Nikolaisen, 2002 SCC 33 [Housen]. Determinations of fact, inferences of fact and determinations of mixed fact and law are entitled to a very high degree of deference and attract the palpable and overriding error standard of review (see Canada v. Patel, 2025 FCA 145 at paras. 4-13). To find a palpable and overriding error, the error must be obvious and could be determinative of the outcome (Salomon v. Matte‑Thompson, 2019 SCC 14 at para. 33). Determinations of law (including extricable legal questions) are subject to correctness review.

[81] As for the standard of review applicable to confusion determinations, Justice Rothstein of the Supreme Court has stated: “The determination of whether a likelihood of source confusion exists is a fact-finding and inference-drawing exercise, and thus, appellate courts should generally defer to the trial judge’s fact findings and inferences, unless the facts and inferences were based on an error of law or constituted a palpable or overriding error of fact” (Masterpiece at para. 102).

(2) Are patients relevant to the confusion analysis?

(a) Would patients be irrelevant for the confusion analysis if they do not encounter the trademarks as used by the trademark owners?

[82] It is useful to discuss at the outset a key submission the appellants used to support many of their s. 20 arguments. In their view, patients should be disregarded in the confusion analysis because they do not encounter the BYOOVIZ trademark as used by the trademark owners. According to the appellants, the emphasized phrase disqualifies patients from being relevant consumers.

[83] The pertinent statutory provisions in relation to this issue are s. 4 and ss. 6(1), (2) and (5). They are briefly summarized below, in relevant part.

· Subsection 6(1) provides that a trademark is confusing with another if the use of the first mark would cause confusion with the other in the manner and circumstances described in this section.

· Subsection 6(2) provides that confusion is made out “if the use of both trademarks in the same area would likely lead to the inference that [the source of goods is the same]”.

· Subsection 6(5) provides that this determination shall have regard to “all the surrounding circumstances”, including those set out in s. 6(5)(a) – (e).

• Subsection 4(1) (in conjunction with s. 2) provides a meaning of “use” for purposes of the TMA. A trademark is used if, among other things, it is associated with the goods such that notice of the association is given.

[84] It is also useful to mention some interpretative principles. First, the test of confusion to be applied is “a matter of first impression in the mind of a casual consumer somewhat in a hurry who sees the [mark] at a time when he or she has no more than an imperfect recollection of [the other marks] …” (Veuve Clicquot at para. 20).

[85] This test is hypothetical. This is evident from the emphasized words above from ss. 6(1) and 6(2) of the TMA. It is also confirmed by the Supreme Court. For example, in Mattel at para. 56 the appropriate perspective is to consider a mythical casual consumer. Similarly, in Masterpiece (at paragraph 30) Justice Rothstein describes the words “used … in the same area” s. 6(2) as hypothetical so that it is not necessary that the consumer actually encounter the marks in the same area.

[86] Although the test is hypothetical there is flexibility to incorporate surrounding circumstances in delineating the mythical consumer. Justice Binnie in Mattel (at paragraph 58) describes consumers who are functionally bilingual as “those mythical consumers.” Similarly, in Masterpiece at paragraph 69, Justice Rothstein refers to this type of surrounding circumstance as an element of the “broader hypothetical test.”

[87] In addition, at paragraph 51 of Mattel, the Court addresses the interplay between s. 6(2) and s. 6(5) of the TMA. Justice Binnie notes that, whereas the nature of the wares is specifically identified in s. 6(5) as a surrounding circumstance, s. 6(2) provides that the wares need not be of the same general class. He reconciles the two provisions by concluding that the general class of wares “while relevant, is not controlling.”

[88] In these appeals, the appellants acknowledge that the consumer must be a person who is likely to purchase the goods (Baylor University v. Hudson’s Bay Co., 2000 CanLII 15626 (F.C.A.) at para. 27 [Baylor]). However, the appellants suggest there is a further requirement that those prospective purchasers must encounter the trademark as used by the trademark owner. For this principle, they rely on a decision referred to in Baylor (at paragraph 28): Joseph E. Seagram & Sons Ltd. v. Canada (Registrar of Trade Marks), 1990 CarswellNat 826 at para. 48 (F.C.T.D.) [Joseph Seagram]. In Joseph Seagram, Justice MacKay concluded: “I do not believe that the opinions of persons who may never have contemplated using the latter services are directly relevant, for they may not encounter any circumstance likely to give rise to confusion from the trade mark as used by the broker” (emphasis added). The appellants therefore submit that for a person to be considered in the confusion analysis, the person must be likely to purchase the goods and they must also encounter the mark as used by the trademark owner.

[89] If the latter part of this test from Joseph Seagram—i.e., the consumer must encounter the mark as used by the trademark owner—was once good law, it is no longer. It was not applied in Baylor or in any other decision referred to by the appellants. In addition, the requirement is at odds with Mattel and Masterpiece, which both confirm that the test is hypothetical even though it takes the consumer’s surrounding circumstances into account. Accordingly, in applying s. 6(2) the prospective consumers are not actually required to encounter the trademark as used by the owner.

[90] Therefore, we do not accept that patients should be disregarded in the confusion analysis because they do not encounter the BYOOVIZ trademark as used by the appellants. The application judge did not err in this respect.

[91] Similar reasoning applies to other arguments made by the appellants that the patients are not likely to be confused because patients rarely inquire about a particular drug and generally are not aware of the BEOVU brand. The appellants submit that the test for confusion is one of probability, not possibility.

[92] This argument is flawed because the hypothetical test assumes that the relevant consumer is aware of both marks. We disagree with the appellants that this approach goes against the intended purpose of the legislation. To the contrary, this approach enhances the consumer protection aim of the legislation.

(b) Are patients irrelevant for other reasons?

[93] The appellants submit there are other reasons for excluding patients from the confusion analysis: (1) the patients did not purchase the drug; (2) the patients did not choose the drug; and (3) the policy considerations set out in Ciba-Geigy were not relevant.

[94] Regarding the third argument, the Ciba-Geigy decision played an important role in the application judge’s decision. In connection with a common law action for passing off, the Supreme Court considered whether patients who purchase prescription drugs at a pharmacy should be considered in the confusion analysis even though historically they had been excluded. A central issue in the case was whether patients had a choice as to the brand of drug. If there was no choice, arguably they were irrelevant to the confusion analysis.

[95] The Supreme Court concluded that patients did have a choice and should be taken into account in considering confusion. This conclusion was supported by strong policy reasons that patients should have the protection of the TMA: “Whether the choice is great or small, easily exercised or not, does not change anything in the case at bar” (at 150). The Court also cautioned that it was “all the more necessary” for patients to be able to exercise some control if they have no access to a drug (at 147).

[96] Applying the policy considerations in Ciba-Geigy to the facts of this case, the application judge concluded that patients were relevant consumers for purposes of the confusion test. The application judge’s reasons, at paragraphs 33 to 39, are summarized below:

• “[P]atients exercise the requisite control over the anti-VEGF medications they will receive” (at para. 33).

• Patients have a sufficient degree of choice. They are told the brand name of the drug that will be administered; they sign a consent form; the drug name is on the syringe that will be used to inject the drug and the patient may see the name on the syringe before the drug is administered; upon hearing the trademark, and possibly seeing it on the syringe, patients can exercise at least the choice to refuse administration (at para. 34).

• In some circumstances, patients can choose between anti-VEGF medications and would have the option to purchase the drug out-of-pocket or through third-party insurance (at para. 35).

• While funding models vary from province-to-province, the drug is purchased for the patient’s benefit and use. Further, patients know what medication they are on by the brand name (at para. 38).

• Trademarks serve a public interest in assuring customers of the source and quality they associate with a particular mark. Confusion must be avoided in the minds of all customers and is not limited to direct customers. It is ‘all the more necessary’ for patients to have a choice if they have no direct access to the drug (at para. 36, citing Mattel at para. 21; Ciba-Geigy at 140, 147).

• The patient as the ultimate end-user of the drug is a relevant consumer for the likelihood of confusion analysis (at para. 39).

[97] These factually suffused findings are entitled to deference. They reflect no palpable and overriding error. With respect to the policy considerations in Ciba-Geigy, we agree with the application judge that the principles stated in that decision support the view that the patients in this case are relevant consumers (at para. 33). No error was made on that account.

(3) Did the Federal Court wrongly assess s. 6(5) factors and other surrounding circumstances?

[98] Subsection 6(5) of the TMA describes what must, at a minimum, be considered and weighed in the confusion analysis. As mandated by this provision, the application judge had regard to each of the enumerated factors in s. 6(5) and other surrounding circumstances. The appellants submit that the application judge erred with respect to some of these factors. Their submissions are discussed below.

(a) Inherent distinctiveness

[99] The application judge determined that the factor of inherent distinctiveness in s. 6(5)(a) favours Novartis. The judge noted that BEOVU is a coined word which is entitled to greater protection. The judge rejected the appellants’ argument that inherent distinctiveness should be neutral because BYOOVIZ is also a coined word, finding that this would undermine the general principle that coined words are generally provided greater protection.

[100] This finding was open to the Court and it committed no error in making it. As stated by Justice Linden in United Artists: “[W]here the mark is a unique or invented name, such that it could refer to only one thing, it will be extended greater scope of protection” (at para. 23). Contrary to the appellants’ submission, this principle equally applies if there is a similar coined word.

(b) Nature of the goods and trade

[101] The application judge determined that the nature of the goods and trade, which are factors in ss. 6(5)(c) and (5)(d) of the TMA, favour Novartis.

[102] The appellants suggest that the application judge failed to consider the importance of these factors and committed an error of law or palpable and overriding error by ignoring or downplaying relevant facts.

[103] We disagree that the application judge made any error with respect to the nature of the goods or trade. The application judge was meticulous in setting out, and considering, the evidence and the applicable law. In essence, the appellants are requesting a reweighing of the evidence. There is no basis to disturb the findings of the application judge with respect to these s. 6(5) factors.

(c) Degree of resemblance

[104] The application judge considered the degree of resemblance, including in appearance or sound or in the ideas suggested in accordance with s. 6(5)(e). This analysis was highly dependent on the particular facts. The judge concluded (at paragraph 94) that BEOVU and BYOOVIZ resemble each other “to at least a moderate degree” in appearance and ideas suggested. As for sound, the judge concluded that there is a high degree of resemblance.

[105] The application judge also determined that sound is a particularly important factor given the role of oral communication “as between health care professionals and staff,” and “in informing patients” (at paragraph 96).

[106] The appellants submit that the Federal Court made a legal error by finding that sound was a particularly important factor. They submit that “verbal communication, particularly outside the context of the normal course of trade (e.g., a purchase), does not qualify as use of a trademark.” For reasons discussed earlier, the Court did not err in this regard. In essence, the appellants confuse the hypothetical test of encountering a trademark as used by the trademark owner and the surrounding circumstances.

(d) Other surrounding circumstances

[107] The appellants suggest that, while the application judge acknowledges that ophthalmologists and pharmacists are capable of perceiving smaller differences in drug name trademarks, the judge failed to take this fact into account in applying the test for confusion. This clearly misinterprets the application judge’s reasons. She considered the different perspective of health care professionals, as evidenced by her conclusion: “[T]he likelihood of confusion is higher when considered from the patient’s perspective; however, I find Novartis has met its burden of establishing a likelihood of confusion for all relevant consumers” (at paragraph 115). Essentially, the appellants are asking the Court to reweigh this evidence. No palpable and overriding error was made in this regard.

(e) Further comment

[108] It is not necessary to address all of the appellants’ s. 6(5) arguments. In this regard, it bears noting that many of the appellants’ arguments are without merit because they either misinterpret the application judge’s reasons, or improperly seek a reweighing of the evidence. In our view, no errors of law or palpable and overriding errors have been demonstrated.

(4) Did the Federal Court err in concluding that sales had taken place?

[109] The appellants submit that the application judge “was wrong in law” in finding that the appellants’ use of the BYOOVIZ trademark violated Novartis’ rights in the BEOVU trademark. This is so, they submit, because the application was commenced before any infringing activity occurred. Novartis, they say, “did not establish any evidence of use of the BYOOVIZ trademark, and the [a]pplication [j]udge erred in finding that the BYOOVIZ [d]rug was sold and promoted”. In those circumstances, when Novartis had not yet suffered any harm, the only injunctive relief potentially available to Novartis, the appellants argue, was a quia timet injunction; that is, an injunction based on “what one fears” may occur in the future: AstraZeneca Canada Inc. v. Novopharm Limited, 2010 FCA 112 at paras. 6-7, affirming AstraZeneca Canada Inc. v. Novopharm Limited, 2009 FC 1209 at paras. 19-20. But all counsel had agreed, and informed the application judge, that the application had not been brought on a quia timet basis and was not being treated as an application for quia timet relief. Therefore, the appellants submit, the ordinary burden of proof applied, and Novartis had failed to meet it.

[110] However, we agree with Novartis’ submission that the application judge made no reviewable error in finding that the parties had “confirmed that, by the time of the hearing, the [appellants] were promoting and selling the BYOOVIZ drug in Canada.” This finding was an inference of fact, to which the stringent palpable and overriding error standard of review applies: Housen at paras. 5-6; Ontario (Attorney General) v. Restoule, 2024 SCC 27 at paras. 90-91.

[111] The foundation for this finding included exchanges between counsel and the application judge. Counsel for Samsung raised the question of timing in his closing submissions, reminding the application judge that the evidence in the application had all been generated several months earlier, and that “[a]t that point in time, BYOOVIZ was not on the market.” He went on: “This next point that I’m going to make, it’s not in the evidence but we thought we should alert you to it, and that is, that BYOOVIZ is now on the market.” The application judge then observed that “it seems that the parties are treating it as though it’s not like a quia timet type situation.” Counsel for Samsung commented that counsel for Biogen was “pointing out that [Novartis] didn’t ask for that type of injunction, but the practical reality is the product is on the market.” The application judge then made a further comment and asked, “[I]n that situation, what’s the remedy then, like there’s no remedy available? If damages are not available and injunction is not available, what would be the remedy?” Counsel for Samsung responded, “Well, damages are available, they just have to prove them.” (April 18, 2023 recording at 47:58-56:31, Appeal Book at Tab 6).

[112] It was at a minimum reasonable for the application judge to infer that the exchanges with counsel about what was “on the market” encompassed not just promotion but also sales.

[113] As further support for the application judge’s finding regarding the meaning of “on the market,” Novartis points to passages in two of the memoranda filed in the Federal Court, in which the parties plainly used the phrase in referring to actual sales, and not just availability for sale. And even if the phrase did not encompass actual sales, the “promotion” to which it refers would include other activities—distribution and advertising—actionable under s. 20(1)(a) of the TMA.

[114] In addition, the appellants made similar submissions regarding passing off activities. The reasons above are equally applicable to these arguments.

B. Cross-appeals

[115] As set out above, Novartis has brought cross-appeals (identical in each of the two appeals, except for the names of the respondents to the cross-appeal) in which it seeks orders that the application judge’s judgment be “varied” (in the cross-appeal brought against Samsung) to:

(a) order that Novartis AG and Novartis Pharmaceuticals Canada Inc. (“Respondents”) are entitled to recover its damages, or Samsung Bioepis Co. Ltd.’s (“Appellant”) profits, as elected by the Respondents, caused by the infringing sales of pharmaceutical products bearing the Appellant’s BYOOVIZ trademark;

(b) Remit the issue of the quantum of the Respondents’ damages or the Appellant’s profits (as elected by the Respondents) to be determined by way of reference before the Federal Court;

(c) Order the production for inspection and copying by the Respondents of any document or other material relevant to the profits made by the Appellant in view of infringing sales of pharmaceutical products bearing the BYOOVIZ trademark;

(d) Order that the Appellant be examined for discovery on the aforementioned documents;

(e) [Grant t]he Respondents’ costs of its application, the appeal, and th[e] cross-appeal.

[116] In our view, these cross-appeals are misconceived.

[117] The notice of application included in its prayer for relief (Appeal Book at p. 108) a claim for “Damages or an accounting of Biogen/Samsung’s profits, at Novartis’ election before final judgment is rendered, for Biogen/Samsung’s violations of the Trademarks Act as described herein” plus a further claim for punitive damages of $250,000.

[118] Novartis has made its election. It did so in the Federal Court, where it pleaded in its closing submissions (Appeal Book at p. 5415, footnote omitted):

Novartis requests damages in the amount of $100,000. The damages are requested, in addition to the compensatory function, to deter others from carrying out similar activities of selecting drug names similar to the BEOVU trademark.

Novartis repeated its election in oral closing submissions (April 17, 2023 recording at 47:58-1:00:05-1:02:10, Appeal Book at Tab 6). The application judge dealt with Novartis’ request as follows (at para. 136):

The Court may award nominal damages without proof of actual damage; however, I am not satisfied that the award in this case should be $100,000. I would assess a reasonable award of nominal damages to be $20,000.

[119] In its notices of cross-appeal, Novartis seeks to justify its request for a reference on the basis that “there [had] been a fundamental and material change with respect to the damage suffered by [it] and the profits earned by [Samsung and Biogen].” That “fundamental and material change” was that by the time of the hearing, promotion and sales of BYOOVIZ had begun. This, Novartis submits, meant that it was:

not possible for [it] to lead evidence related to its damages (or to compel evidence related to the [other parties’]) profits associated with infringing sales of the … BYOOVIZ drug prior to the hearing of the underlying application.

[120] But Novartis, as it now acknowledges, was made aware of this change in circumstance nearly a month before the hearing of the application and referred to it during the hearing in the Federal Court (April 18, 2023 recording at 1:41.23 to 1:42.03, Appeal Book at Tab 6).

[121] In any event, accepting that there had been a change in circumstances, there were a number of procedural routes open to Novartis, including seeking an adjournment and leave of the Federal Court (if the other parties did not consent) to adduce further evidence. But Novartis did nothing; it lay in the weeds until after the Federal Court rendered its decision and only then raised the change in circumstances in this Court as a ground of appeal.

[122] However, as this Court recognized in AstraZeneca Canada Inc. v. Apotex Inc., 2016 FCA 194 at paras. 17, 19-20, a change in circumstances is not a proper ground of appeal; rather, it can provide a basis for a motion to the first instance decision-maker to set aside or vary under rule 399(2)(a). That rule states that “[o]n motion, the Court may set aside or vary an order […] by reason of a matter that arose or was discovered subsequent to the making of the order.” And as this Court stated in AstraZeneca:

when this Court dismisses an appeal, this Court has concluded that the judgment below is not vitiated by an error of law or a palpable and overriding error of fact or mixed fact and law. This determination is a qualitatively different decision than a decision about whether the judgment ought to be set aside or varied, not because of any error, but because the judgment has been shown to be flawed based on matters discovered after the judgment was made.

The person best placed to decide whether the newly discovered matters would have affected the original judgment is the original decision-maker.

[123] For these reasons, we will dismiss the cross-appeals with costs. It is not necessary for us to consider other grounds for doing so that were argued before us, including issue estoppel and the prerequisites set out in rule 153.

V. Conclusion and disposition

[124] We conclude that the appeals and the cross-appeals should be dismissed with costs.

[125] The request to order elevated costs on the cross-appeals is denied. We also decline to order costs of the applications to Novartis, as the Federal Court is better equipped to consider this.

[126] Finally, the appellants request a 12-month extension of a stay previously ordered by Justice Rennie of this Court. This request should be made by way of a formal motion.

[127] The original of these reasons will be placed in file A-26-24 and a copy will be placed in file A-27-24.

“Judith Woods”

“J.B. Laskin”

APPENDIX

FEDERAL COURT OF APPEAL

NAMES OF COUNSEL AND SOLICITORS OF RECORD

Dockets:		A-26-24 AND A-27-24
DOCKET:	A-26-24
STYLE OF CAUSE:	SAMSUNG BIOEPIS CO., LTD. v. NOVARTIS AG AND NOVARTIS PHARMACEUTICALS CANADA INC.
AND DOCKET:	A-27-24
STYLE OF CAUSE:	BIOGEN INC., BIOGEN MA INC., BIOGEN CANADA INC. v. NOVARTIS AG AND NOVARTIS PHARMACEUTICALS CANADA INC.
PLACE OF HEARING:		Toronto, Ontario
DATE OF HEARING:		May 28, 2024
PUBLIC REASONS FOR JUDGMENT BY:		WOODS J.A. LASKIN J.A.
NOT TAKING PART IN THE JUDGMENT:		BOIVIN J.A.
DATED:		november 28, 2025

APPEARANCES:

J. Bradley White Nathaniel Lipkus Daniel Hnatchuk	For The Appellant SAMSUNG BIOEPIS CO., LTD.
Michael Adams Michael Schwartz	For The Respondents NOVARTIS AG AND NOVARTIS PHARMACEUTICALS CANADA INC.
Scott Miller Deborah Meltzer	For The Appellants BIOGEN INC., BIOGEN MA INC., BIOGEN CANADA INC.

SOLICITORS OF RECORD:

Osler, Hoskin & Harcourt LLP Ottawa, Ontario	For The Appellant SAMSUNG BIOEPIS CO., LTD.
Riches, McKenzie & Herbert LLP Toronto, Ontario	For The Respondents NOVARTIS AG AND NOVARTIS PHARMACEUTICALS CANADA INC.
MBM Intellectual Property Law LLP Ottawa, Ontario	For The Appellants BIOGEN INC., BIOGEN MA INC., BIOGEN CANADA INC.