Language of document : ECLI:EU:T:2017:637

Provisional text

JUDGMENT OF THE GENERAL COURT (Fifth Chamber)

21 September 2017 (*)

(EU trade mark — Opposition proceedings — Application for the EU word mark Zymara — Earlier EU word mark FEMARA — Relative ground for refusal — Article 8(1)(b) of Regulation (EC) No 207/2009 — Article 75 of Regulation No 207/2009 — Article 76(1) of Regulation No 207/2009)

In Case T‑214/15,

Novartis AG, established in Basle (Switzerland), represented by M. Douglas, lawyer,

applicant,

v

European Union Intellectual Property Office (EUIPO), represented by H. O’Neill, acting as Agent,

defendant,

the other party to the proceedings before the Board of Appeal of EUIPO being:

Meda AB, established in Solna (Sweden),

ACTION brought against the decision of the Fifth Board of Appeal of EUIPO of 6 February 2015 (Case R 550/2014-5) relating to opposition proceedings between Novartis and Meda,

THE GENERAL COURT (Fifth Chamber),

composed of D. Gratsias, President, A. Dittrich (Rapporteur) and P.G. Xuereb, Judges,

Registrar: I. Dragan, Administrator,

having regard to the application lodged at the Court Registry on 23 April 2015,

having regard to the response lodged at the Court Registry on 27 August 2015,

having regard to the reply lodged at the Court Registry on 15 December 2015,

having regard to the rejoinder lodged at the Court Registry on 4 March 2016,

further to the hearing on 12 January 2017,

gives the following

Judgment

 Background to the dispute

1        On 20 May 2011, Meda AB filed an application for registration of an EU trade mark with the European Union Intellectual Property Office (EUIPO) pursuant to Council Regulation (EC) No 207/2009 of 26 February 2009 on the European Union trade mark (OJ 2009 L 78, p. 1).

2        Registration as a mark was sought for the word sign Zymara.

3        The goods in respect of which registration was sought are in Classes 3 and 5 of the Nice Agreement concerning the International Classification of Goods and Services for the Purposes of the Registration of Marks of 15 June 1957, as revised and amended, and correspond, for each of those classes, following the restrictions made during the opposition proceedings, to the following description:

–        Class 3: ‘Soaps; Perfumery, essential oils, cosmetics (not for dental use), hair lotions’;

–        Class 5: ‘Pharmaceutical preparations; Sanitary preparations for medical purposes; Dietetic substances adapted for medical use, food for babies; Disinfectants; except for drugs for use in the cardio-vascular system (all of the aforementioned products not for use in the dental field and excluding ophthalmic preparations)’.

4        The EU trade mark application was published in Community Trade Marks Bulletin No 116/2011 of 22 June 2011.

5        On 19 September 2011, the applicant, Novartis AG, filed a notice of opposition, pursuant to Article 41 of Regulation No 207/2009, to registration of the mark applied for in respect of the following goods in Class 5: ‘Pharmaceutical preparations; sanitary preparations for medical purposes; dietetic substances adapted for medical use’.

6        The opposition was based on the EU word mark FEMARA, registered on 28 May 1999 for the goods in Class 5 corresponding to the following description: ‘Pharmaceutical preparations’.

7        The ground relied on in support of the opposition was that set out in Article 8(1)(b) of Regulation No 207/2009.

8        By letter of 27 December 2012, Meda requested proof of genuine use of the earlier mark on which the opposition was based. The Opposition Division therefore invited the applicant, pursuant to Article 42(2) of Regulation No 207/2009, to furnish that proof.

9        On 17 December 2013, the Opposition Division rejected the opposition. In particular, it found that the applicant had demonstrated genuine use of the earlier mark, during the relevant period, only for pharmaceutical preparations for the treatment of breast cancer. Moreover, it found that the goods at issue were partly identical, partly similar to a high degree, and partly similar, and that the marks at issue were visually and phonetically similar to a certain degree. It found that there was no likelihood of confusion, taking account in particular of the fact that, in its view, the beginning of the mark applied for was very different from that of the earlier mark.

10      On 17 February 2014, the applicant filed an appeal with EUIPO, pursuant to Articles 58 to 64 of Regulation No 207/2009, against the decision of the Opposition Division.

11      By decision of 6 February 2015 (‘the contested decision’), the Fifth Board of Appeal of EUIPO dismissed the appeal. In particular, it found that the Opposition Division had correctly defined the sub-category for which the applicant had demonstrated genuine use of the earlier mark, namely that of pharmaceutical preparations for the treatment of breast cancer. Moreover, it noted that, even if EUIPO had considered that genuine use had been demonstrated for the category of ‘pharmaceutical preparations for the treatment of cancer’, in accordance with the applicant’s line of argument, that would not have changed anything, with regard to either the result of the comparison of the goods or the outcome of the case.

12      The Board of Appeal stated that the relevant territory was that of the European Union and that the goods at issue were aimed at professionals in the medical sector and average consumers as end consumers. It also found, in paragraph 33 of the contested decision, that the relevant public displayed a high level of attention in relation to the goods at issue.

13      The Board of Appeal found that the ‘pharmaceutical preparations’ covered by the mark applied for were identical to ‘pharmaceutical preparations for the treatment of breast cancer’, for which genuine use of the earlier mark had been demonstrated. Furthermore, it found that there was a low degree of similarity between the ‘sanitary preparations for medical purposes’ and the ‘dietetic substances adapted for medical use’, covered by the mark applied for, on the one hand, and the ‘pharmaceutical preparations for the treatment of breast cancer’, on the other.

14      According to the Board of Appeal, there is a very low degree of visual similarity and a low degree of phonetic similarity between the mark applied for and the earlier mark. As regards the phonetic similarity, it found that the marks were pronounced ‘zy-ma-ra’ and ‘fe-ma-ra’ in accordance with the pronunciation rules of the respective languages. As regards the English-speaking part of the relevant public, the Board of Appeal found that it would pronounce the letter ‘y’ of the Zymara mark like an ‘i’ and the letter ‘e’ in the FEMARA mark like the ‘e’ in the English word ‘lemon’.

15      The Board of Appeal concluded that there was no likelihood of confusion, in particular in the light of the low degree of similarity between the mark applied for and the earlier mark, the fact that the difference between those marks was at the beginning of them, and the high level of attention of the relevant public concerning the goods at issue.

 Forms of order sought

16      The applicant claims that the Court should:

–        annul the contested decision;

–        order EUIPO to pay the costs.

17      EUIPO contends that the Court should:

–        dismiss the application;

–        order the applicant to pay the costs.

 Law

18      The applicant essentially raises a single plea in law, alleging infringement of Article 8(1)(b) of Regulation No 207/2009. The applicant claims that there is a likelihood of confusion between the marks at issue for the goods in question. In the context of that plea in law, the applicant raises objections alleging infringement of Articles 42(2), 76(1) and 75 of Regulation No 207/2009.

19      It should be noted that Article 8(1)(b) of Regulation No 207/2009 provides that, upon opposition by the proprietor of an earlier trade mark, the trade mark applied for must not be registered if, because of its identity with, or similarity to, an earlier trade mark and the identity or similarity of the goods or services covered by the trade marks, there exists a likelihood of confusion on the part of the public in the territory in which the earlier trade mark is protected. The likelihood of confusion includes the likelihood of association with the earlier trade mark.

20      According to settled case-law, the risk that the public might believe that the goods or services in question come from the same undertaking or from economically-linked undertakings constitutes a likelihood of confusion. According to the same case-law, the likelihood of confusion must be assessed globally, according to the relevant public’s perception of the signs and goods or services in question and taking into account all factors relevant to the circumstances of the case, in particular the interdependence between the similarity of the signs and that of the goods or services covered (see judgment of 9 July 2003, Laboratorios RTB v OHIM — Giorgio Beverly Hills (GIORGIO BEVERLY HILLS), T‑162/01, EU:T:2003:199, paragraphs 30 to 33 and the case-law cited).

21      Where the protection of the earlier trade mark extends to the entirety of the European Union, it is necessary to take into account the perception of the marks at issue by the consumer of the goods in question in that territory. However, it must be borne in mind that, for an EU trade mark to be refused registration, it is sufficient that a relative ground for refusal for the purposes of Article 8(1)(b) of Regulation No 207/2009 exists in part of the European Union (see, to that effect, judgment of 14 December 2006, Mast-Jägermeister v OHIM — Licorera Zacapaneca (VENADO with frame and others), T‑81/03, T‑82/03 and T‑103/03, EU:T:2006:397, paragraph 76 and the case-law cited).

 The goods for which the earlier mark is deemed to be registered

22      The applicant claims that the category of goods for which it demonstrated genuine use of the earlier mark, by virtue of Article 42(2) of Regulation No 207/2009, must be defined as ‘pharmaceutical preparations for the treatment of cancer’ and that the Board of Appeal was wrong to restrict that category to pharmaceutical preparations for the treatment of breast cancer alone.

23      EUIPO disputes that argument and furthermore claims that it cannot have any bearing on the outcome of the case.

24      In that regard, it should be noted that the applicant has not raised any argument seeking to dispute the finding of the Board of Appeal that, if genuine use had been demonstrated for the category of ‘pharmaceutical preparations for the treatment of cancer’, that would not have changed anything, either with regard to the outcome of the comparison of the goods or the outcome of the case.

25      It must be noted that, in fact, for the purposes of comparing the goods at issue, it is irrelevant whether the earlier mark is deemed to have been registered for ‘pharmaceutical preparations for the treatment of breast cancer’, in accordance with the view of the Board of Appeal, or for ‘pharmaceutical preparations for the treatment of cancer’, in accordance with the applicant’s view (see paragraph 71 below). The applicant’s argument is therefore ineffective.

26      As regards the applicant’s argument that it is important to obtain a decision on that question, which would also give guidance for the future, it must be noted that the Court is not obliged to rule on questions which are irrelevant to the outcome of the dispute on the ground that one party considers that it would be useful, for other cases, to receive clarification on those questions.

 The relevant public and its level of attention

27      According to the case-law, in the global assessment of the likelihood of confusion, account should be taken of the average consumer of the category of goods concerned, who is reasonably well informed and reasonably observant and circumspect. It should also be borne in mind that the average consumer’s level of attention is likely to vary according to the category of goods or services in question (see judgment of 13 February 2007, Mundipharma v OHIM — Altana Pharma (RESPICUR), T‑256/04, EU:T:2007:46, paragraph 42 and the case-law cited).

28      In the present case, it should be noted that, having regard to the fact that the earlier mark is an EU trade mark, the relevant territory for the assessment of the likelihood of confusion is the territory of the European Union, as the Board of Appeal correctly stated, which, moreover, is not disputed by the applicant.

29      The Board of Appeal was furthermore right to make the finding, which the applicant does not dispute, that the goods at issue were intended for professionals in the medical sector and for average consumers as end consumers.

30      The applicant, however, disputes the Board of Appeal’s assessment of the level of attention of the relevant public.

31      In the first place, the applicant submits that the contested decision is not consistent, in so far as it states, on the one hand, that the level of attention displayed by the relevant public is heightened (paragraph 34 of the contested decision) and, on the other, that that level is high. In the second place, it maintains, in the application, that the level of attention of the end-user when purchasing the goods in question is average. In the reply, the applicant concedes that it may be the case that the level of attention of the relevant public is above average. It maintains, however, that this does not mean that the consumer always displays a high level of attention.

32      EUIPO submits that the Board of Appeal used the expression ‘heightened degree of attentiveness’ to mean ‘a high degree of attention’. Furthermore, it submits that the Board of Appeal was correct to find that there exists a high level of attention.

33      First, as regards the alleged lack of consistency of the contested decision, it should be noted that the Board of Appeal observed, in paragraphs 33 and 45 of that decision, that the level of attention of the relevant public was ‘high’and, in paragraph 34 of that decision, that that level was ‘heightened’.

34      It should be pointed out that those words were used as synonyms by the Board of Appeal. It should be noted that, in paragraph 34 of the contested decision, the Board of Appeal referred to judgments in which the Court had found that there was a ‘heightened’ level of attention, which explains the use of that word by the Board of Appeal in that paragraph of the contested decision.

35      Moreover, the use of those words as synonyms is consistent with the case-law. Thus, in the judgment of 15 December 2010, Novartis v OHIM — Sanochemia Pharmazeutika (TOLPOSAN) (T‑331/09, EU:T:2010:520, paragraph 26), those words were used as synonyms by the Court, which noted, first, that consumers were likely to have a ‘high’ degree of attentiveness upon prescription of medicines and secondly that ‘medicines, whether or not issued on prescription, [could] be regarded as receiving a heightened degree of attentiveness by consumers who are reasonably well informed and reasonably observant and circumspect’. This was also the case in the judgment of 26 November 2015, Bionecs v OHIM — Fidia farmaceutici (BIONECS) (T‑262/14, not published, EU:T:2015:888, paragraph 17).

36      In any event, it is necessary to reject the applicant’s argument that, from a grammatical point of view, a heightened level of attention is only a comparison with a different level of attention, and that it is difficult to know whether the result of that heightened level of attention will be a high level of attention or merely a normal level of attention, compared with a low level of attention. That reasoning is based solely on a grammatical analysis of the word ‘heightened’. However, it is clear from the judgments cited in paragraph 35 above that the Court referred to a level of attention which was ‘heightened’ in comparison with an average level of attention. It is also clear from the contested decision that the Board of Appeal referred to a level of attention which was ‘heightened’ in comparison with an average level of attention, since it also noted, on two occasions (paragraphs 33 and 45 of the contested decision), the existence of a ‘high’ level of attention.

37      It follows from the foregoing that the applicant’s argument alleging that the contested decision is inconsistent must be rejected.

38      Next, as regards the level of attention of the relevant public, the following must be noted.

39      As regards pharmaceutical goods, it is settled case-law that professionals display a high level of attention (judgments of 21 October 2008, Aventis Pharma v OHIM — Nycomed (PRAZOL), T‑95/07, not published, EU:T:2008:455, paragraph 29, and of 15 December 2010, TOLPOSAN, T‑331/09, EU:T:2010:520, paragraph 26). The applicant does not dispute that professionals display such a level of attention.

40      It is also necessary to consider that end-users display a high level of attention with regard to pharmaceutical goods (judgments of 11 June 2014, Golam v OHIM — meta Fackler Arzneimittel (METABIOMAX), T‑281/13, not published, EU:T:2014:440, paragraph 30, and of 26 November 2015, BIONECS, T‑262/14, not published, EU:T:2015:888, paragraphs 17 and 20). That high level of attention is due to the implications of pharmaceutical goods for human health.

41      The applicant claims that end-users do not display a high level of attention and submits that pharmaceutical goods may be bought via mail order, like other consumer goods, without the assistance of a healthcare professional.

42      However, as the Court has already had occasion to hold, the fact that an end-user could possibly obtain on the internet a medicinal product sold without prescription, without the advice of a pharmacist or physician, is not such as to lower that consumer’s level of attention when purchasing such goods (judgment of 15 December 2010, TOLPOSAN, T‑331/09, EU:T:2010:520, paragraph 28). As EUIPO maintains, the high level of attention of non-professional users of medicines is associated with the health implications of the medicines, and not with the intervention of a specialist.

43      The applicant also claims that, if the consumer’s level of attention when purchasing medicines cheaply on the internet had to be considered to be heightened or high, it would also be necessary to note that the level of attention concerning foodstuffs was high, since health problems could also be caused by eating the wrong product, on account of allergies caused by specific ingredients.

44      That argument cannot be upheld. It is based on the very specific situation of consumers suffering from an allergy. While it is true that consumers suffering from an allergy may display a heightened level of attention when purchasing foodstuffs, that is not the case for average consumers. The health implications relating to the choice of foodstuffs are not comparable to those relating to the choice of pharmaceutical goods.

45      As regards ‘dietetic substances adapted for medical use’, it must also be held that the level of attention displayed by the public with regard to those goods is high (see, to that effect, judgment of 12 July 2012, Pharmazeutische Fabrik Evers v OHIM — Ozone Laboratories Pharma (HYPOCHOL), T‑517/10, not published, EU:T:2012:372, paragraph 21). It should be noted that goods ‘for medical use’ have health implications for consumers, which justifies the view that their level of attention is high, in the same way as for pharmaceutical goods.

46      Finally, it must also be concluded that the relevant public has a high level of attention in respect of ‘sanitary preparations for medical purposes’ (judgment of 11 June 2014, METABIOMAX, T‑281/13, not published, EU:T:2014:440, paragraph 30).

47      That result is not called into question by the judgment of 15 March 2012, Cadila Healthcare v OHIM — Novartis (ZYDUS) (T‑288/08, not published, EU:T:2012:124, paragraph 36). In that judgment, the Court held that ‘the relevant public’s level of attention in respect of sanitary preparations in Class 5, without being high, [was] above average, since those preparations include[d] both goods intended for everyday use and goods intended for medical use’. By contrast, in the present case, the sanitary preparations covered by the mark applied for are expressly limited to those ‘for medical purposes’.

48      It follows from all of the foregoing that the Board of Appeal was right to find that the relevant public has a high level of attention in respect of the goods in question.

 The comparison of the goods

49      According to settled case-law, in assessing the similarity of the goods or services at issue, all the relevant factors relating to those goods or services should be taken into account. Those factors include, inter alia, their nature, their intended purpose and their method of use and whether they are in competition with each other or are complementary. Other factors may also be taken into account, such as the distribution channels of the goods concerned (see judgment of 11 July 2007, El Corte Inglés v OHIM — Bolaños Sabri (PiraÑAM diseño original Juan Bolaños), T‑443/05, EU:T:2007:219, paragraph 37 and the case-law cited).

50      In the present case, it is necessary to compare ‘pharmaceutical preparations for the treatment of breast cancer’, in respect of which the earlier mark is deemed to be registered, with the following goods, covered by the mark applied for following the restrictions of the list of goods made by Meda during the administrative procedure: ‘Pharmaceutical preparations; sanitary preparations for medical purposes; dietetic substances adapted for medical use; except for drugs for use in the cardiovascular system (all of the aforementioned products not for use in the dental field and excluding ophthalmic preparations)’.

51      In the remainder of the present judgment, the restrictions to the list of goods covered by the mark applied for, namely the goods expressly excluded from the list, are no longer expressly referred to, as they have no relevance to the comparison of the goods in question.

 ‘Pharmaceutical preparations’

52      The Board of Appeal was correct in stating, and is not contradicted on this point by the applicant, that the ‘pharmaceutical preparations’ covered by the mark applied for were identical to the ‘pharmaceutical preparations for the treatment of breast cancer’, for which the earlier mark was deemed to be registered. The latter goods are included in the broader category of ‘pharmaceutical preparations’ covered by the mark applied for.

 ‘Sanitary preparations for medical purposes’

53      The Board of Appeal found that the ‘sanitary preparations for medical purposes’ covered by the mark applied for and ‘pharmaceutical preparations for the treatment of breast cancer’, for which the earlier mark was deemed to be registered, were similar to a low degree. It noted that ‘sanitary preparations for medical purposes’ included various goods which are used for sterilising and disinfecting in hospitals and clinics and that they were, inter alia, intended for use in medical treatments or in surgical operations. According to the Board of Appeal, they, together with pharmaceutical preparations, belonged to the same category of goods for healthcare, they could be manufactured by the same economic operators, namely pharmaceutical companies, and they were sold in the same establishments. In support of its analysis, the Board of Appeal cited the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124, paragraph 44).

54      The applicant claims that it is apparent from the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124, paragraph 44) that ‘pharmaceutical preparations’ and ‘sanitary preparations for medical purposes’ are very similar, and not similar to a low degree, as the Board of Appeal considered.

55      In this regard, it should be noted that, in the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124), the Court compared ‘sanitary preparations’ in Class 5 and ‘pharmaceutical preparations’. After noting that Class 5 within the meaning of the Nice Agreement included, inter alia, sanitary preparations for medical purposes and for personal hygiene, but did not include sanitary preparations for personal hygiene which are toiletries, which are covered instead by Class 3, the Court held that there was a ‘relatively high’ degree of similarity between ‘sanitary preparations’, as thus defined, and ‘pharmaceutical preparations’ (paragraph 44 of that judgment). In support of that finding, the Court based its view on the following elements characterising those goods (paragraph 44 of that judgment):

–        they are aimed at the same public, namely healthcare professionals and end consumers;

–        they have the same intended purpose, being intended, inter alia, for use in medical treatment or in a surgical operation;

–        they are also of the same nature, given that they are perceived by consumers as goods belonging to the same general category of healthcare products;

–        they are closely complementary, in so far as certain sanitary preparations for medical purposes, such as antiseptics or antibacterial lotions, may prove to be indispensable when certain pharmaceutical preparations are administered;

–        they may be manufactured by the same economic operators (pharmaceutical companies) and are sold in the same establishments, that is to say, mainly pharmacies but also, sometimes, supermarkets.

56      EUIPO maintains that the finding of the Board of Appeal that the degree of similarity is low is not contrary to the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124). According to EUIPO, the Court’s finding in that judgment cannot be interpreted as meaning that all ‘pharmaceutical preparations’ are highly similar to ‘sanitary preparations for medical purposes’, but merely that some of them are. Other goods, such as ‘pharmaceutical preparations for the treatment of breast cancer’, are not highly similar to ‘sanitary preparations for medical purposes’. In that regard, EUIPO maintains that the finding of complementarity in the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124) is not valid with regard to ‘pharmaceutical preparations for the treatment of breast cancer’, for which the earlier mark is deemed to be registered, which are 2.5 mg film-coated tablets to be taken orally.

57      However, as the applicant maintains, the goods for which the earlier mark is deemed to be registered are not restricted to goods taken orally in the form of film-coated tablets. The applicant is correct in claiming that ‘pharmaceutical preparations for the treatment of (breast) cancer’ can also be administered intravenously. The Court’s finding in the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124, paragraph 44), that certain sanitary preparations for medical use, such as antiseptics or antibacterial lotions, may prove to be indispensable when certain pharmaceutical preparations are administered is entirely valid concerning ‘pharmaceutical preparations for the treatment of breast cancer’ administered intravenously. Those goods are therefore complementary.

58      The findings in paragraph 44 of the judgment of 15 March 2012, ZYDUS (T‑288/08, not published, EU:T:2012:124), as summarised in paragraph 55 above, are valid in so far as concerns the comparison of ‘sanitary preparations for medical use’ with ‘pharmaceutical preparations for the treatment of breast cancer’.

59      Moreover, in the judgment of 11 June 2014, METABIOMAX (T‑281/13, not published, EU:T:2014:440, paragraph 37), the Court also held that there was a ‘relatively high’ degree of similarity between ‘sanitary preparations for medical purposes’ and ‘medicines’.

60      Finally, it should be noted that the Board of Appeal listed several factors which indicate similarity between ‘sanitary preparations for medical purposes’ and ‘pharmaceutical preparations for the treatment of breast cancer’, and that it did not in any way support its finding that the degree of similarity had to be classified as ‘low’.

61      It follows from the foregoing that the Board of Appeal erred in classifying the degree of similarity between ‘sanitary preparations for medical purposes’ and ‘pharmaceutical preparations for the treatment of breast cancer’ as low. The degree of similarity between those goods must be regarded as ‘relatively high’.

 ‘Dietetic substances adapted for medical use’

62      The Board of Appeal found that the ‘dietetic substances adapted for medical use’ covered by the mark applied for and ‘pharmaceutical preparations for the treatment of breast cancer’, for which the earlier mark was deemed to be registered, were similar to a low degree. It noted that ‘dietetic substances adapted for medical use’ included various substances prepared for special dietary requirements for the purpose of treating or preventing disease and that they were often used to support and supplement a medical treatment. Furthermore, it considered that those goods served a related purpose, shared the same distribution channels and sometimes had the same commercial origin. In support of its analysis, the Board of Appeal referred to the judgment of 15 December 2009, TRUBION (T‑412/08, not published, EU:T:2009:507, paragraphs 30 and 33).

63      The applicant claims that the degree of similarity between the ‘dietetic substances adapted for medical use’ covered by the mark applied for and ‘pharmaceutical preparations for the treatment of breast cancer’ is high.

64      It should be noted that, in the judgment of 15 December 2009, TRUBION (T‑412/08, not published, EU:T:2009:507, paragraph 33), the Court noted that, on the one hand, ‘pharmaceutical preparations’ and ‘pharmaceutical preparations for use in the fields of immunology and oncology’ and, on the other hand, ‘dietetic preparations for medical use’ were ‘similar’, without explicitly specifying the degree of similarity. In paragraph 32 of that judgment, the Court held that those goods fulfilled, in general terms, the same function, namely to improve the medical condition of patients. Furthermore, it held that they displayed a certain complementary connection, given that a doctor may prescribe dietetic preparations for medical use in the course of medical treatment and that that finding was all the more relevant for serious illnesses such as cancer or immunological disorders. Finally, the Court observed that those goods generally shared the same distribution channels.

65      It must be held that those findings support the conclusion that there is at least an average degree of similarity.

66      It is, admittedly, true that, as EUIPO maintains, the Court, in the judgment of 15 December 2009, TRUBION (T‑412/08, not published, EU:T:2009:507), did not expressly specify the degree of similarity, with the result that the finding of a low degree of similarity by the Board of Appeal does not directly contradict that judgment. Nevertheless, given that the Court listed, in paragraph 32 of that judgment, several factors of similarity, including ‘a certain complementary connection’, but did not point out elements capable of reducing the degree of similarity, it cannot be held that the degree of similarity, as found by the Court, was low.

67      That finding is not called into question by EUIPO’s argument that the Court found, in the judgment of 15 December 2009, TRUBION (T‑412/08, not published, EU:T:2009:507, paragraph 30), that ‘pharmaceutical preparations’ were of a ‘somewhat different’ nature to ‘dietetic preparations for medical use’, which, according to EUIPO, supports the finding of a low degree of similarity. In paragraph 30 of that judgment, the Court summarised the content of the contested decision in that case. The Court did not repeat that statement in the context of its own analysis, in paragraph 32 of that judgment

68      EUIPO maintains that, according to case-law, ‘health-related goods having different therapeutic indications [are] similar to a certain degree’. It is, admittedly, true that, as EUIPO notes, in the judgment of 17 October 2006, Armour Pharmaceutical v OHIM-Teva Pharmaceutical Industries (GALZIN) (T‑483/04, EU:T:2006:323, paragraphs 70 and 71), the Court held that there was ‘some degree of’ similarity between pharmaceutical preparations with different therapeutic indications. However, in that judgment, the Court did not specify the exact degree of similarity between those goods. Furthermore, in that judgment, the Court did not find that there was a complementary connection between the goods in question.

69      By contrast, the Court has already found that there was a certain complementary connection between ‘dietetic preparations for medical use’ and ‘pharmaceutical preparations for use in the [field] of … oncology’ (see paragraph 64 above).

70      It follows from the foregoing that the Board of Appeal erred in classifying the degree of similarity between ‘dietetic preparations adapted for medical use’, covered by the mark applied for, and ‘pharmaceutical preparations for the treatment of breast cancer’, for which the earlier mark is deemed to be registered, as low. The degree of similarity between those goods must be classified as at least average.

71      It should be noted that the findings relating to the comparison of the goods, in paragraphs 52 to 70 above, would be the same if it were held that the earlier mark was deemed to be registered for ‘pharmaceutical preparations for the treatment of cancer’ in general.

 The comparison of the signs

72      The global assessment of the likelihood of confusion must, so far as concerns the visual, phonetic or conceptual similarity of the signs at issue, be based on the overall impression given by the signs, bearing in mind, in particular, their distinctive and dominant elements. The perception of marks by the average consumer of the goods or services in question plays a decisive role in the global assessment of that likelihood. In this regard, the average consumer normally perceives a mark as a whole and does not engage in an analysis of its various details (see judgment of 12 June 2007, OHIM v Shaker, C‑334/05 P, EU:C:2007:333, paragraph 35 and the case-law cited).

 Visual similarity

73      The Board of Appeal noted that, visually, the marks differed in terms of their first two letters (‘z’ and ‘y’, on the one hand, and ‘f’ and ‘e’, on the other) and coincided in their ending (‘mara’). It noted that, since consumers tend to pay greater attention to the beginning of a mark than to the end, and since the element ‘mara’ was unlikely to be singled out from the marks as a whole, the coincidence in those letters is not immediately perceivable. It concluded that the visual similarity between the marks at issue was very low.

74      The applicant submits that the Board of Appeal put too much weight on the beginning of the words in the context of the visual comparison. According to the applicant, the marks at issue have an average degree of visual similarity or are visually similar ‘at least to an average degree’.

75      According to the case-law, the initial part of a word mark is likely to retain the consumer’s attention more than the parts that follow (see, to that effect, judgments of 17 March 2004, El Corte Inglés v OHIM — González Cabello and Iberia Líneas Aéreas de España (MUNDICOR), T‑183/02 and T‑184/02, EU:T:2004:79, paragraph 81, and of 16 March 2005, L’Oréal v OHIM — Revlon (FLEXI AIR), T‑112/03, EU:T:2005:102, paragraphs 64 and 65). Admittedly, that consideration cannot apply in all cases and it cannot, in any event, undermine the principle that the examination of the similarity between the marks must take account of the overall impression given by them, since the average consumer normally perceives a mark as a whole and does not examine its individual details (judgment of 9 April 2014, Farmaceutisk Laboratorium Ferring v OHIM — Tillotts Pharma (OCTASA), T‑501/12, not published, EU:T:2014:194, paragraph 58).

76      In the present case, the applicant has not put forward any argument to support the view that, as regards the marks at issue, the public does not attach more importance to the initial part of those marks than to their ends.

77      It is true that, as the applicant maintains, in the judgment of 9 April 2014, OCTASA (T‑501/12, not published, EU:T:2014:194, paragraphs 59 and 60), the Court held that the differences between the first parts of the signs PENTASA and OCTASA, despite their position at the beginning of the marks, were incapable of negating a certain degree of visual and phonetic similarity in the overall impression given by those signs and that the Board of Appeal was therefore wrong to state that there was no visual or phonetic similarity between those signs.

78      However, in the present case, the Board of Appeal did not consider that there was no similarity between the marks at issue, but on the contrary held that there was visual and phonetic similarity, albeit a very low degree of visual similarity and a low degree of phonetic similarity.

79      It should be noted that, in the present case, the first syllables of the marks at issue, namely ‘zy’ and ‘fe’, are visually very different and have no letters in common. The second and third syllables, namely ‘ma’ and ‘ra’, are identical.

80      In view of the fact that the public generally pays greater attention to the beginning of a mark than to its end, it must be noted that the degree of visual similarity between the signs at issue must be considered low. However, since the second and third syllables of the marks at issue are the same, that visual similarity cannot be considered, as it was by the Board of Appeal, to be ‘very low’.

81      The finding of a low degree of visual similarity between the marks at issue is not called into question by the applicant’s arguments.

82      The applicant maintains that, in the judgments of 1 July 2014, Jyoti Ceramic Industries v OHIM — DeguDent (ZIECON) (T‑239/12, not published, EU:T:2014:592); of 24 September 2014, Sanofi v OHIM — GP Pharm (GEPRAL) (T‑493/12, not published, EU:T:2014:807); and of 2 December 2014, Boehringer Ingelheim Pharma v OHIM — Nepentes Pharma (Momarid) (T‑75/13, not published, EU:T:2014:1017), the Court held that there was visual similarity between the signs CERCON and ZIERCON (figurative), DELPRAL and GEPRAL and Momarid and LONARID, even though the differences between those signs were to be found at the beginning of them.

83      In that regard, it should be noted that the comparison between two signs must be carried out according to the particular circumstances of each case, with the result that a comparison with the specific degree of similarity found to exist between other signs in other cases is of only limited value (judgments of 3 June 2015, Giovanni Cosmetics v OHIM — Vasconcelos & Gonçalves (GIOVANNI GALLI), T‑559/13, not published, EU:T:2015:353, paragraph 83, and of 5 February 2016, Kicktipp v OHIM — Italiana Calzature (kicktipp), T‑135/14, not published, EU:T:2016:69, paragraph 158). Moreover, in all the cases referred to in paragraph 82 above, the first syllables of the marks at issue had one letter in common. In the judgment of 24 September 2014, GEPRAL (T‑493/12, not published, EU:T:2014:807, paragraph 30), the Court also held that ‘the difference between the respective upper case first letters, “D” and “G”, the contour and shape of which [were] visually similar, [was] not significant’. In the present case, the contour and shape of the first letters of the marks at issue are very different.

84      As regards the Board of Appeal’s finding that, in so far as the element ‘mara’ is unlikely to be singled out from the marks as a whole, the coincidence in those letters is not immediately perceivable, the applicant claims that the Board of Appeal misapplied the principle that, when a mark consists of several elements, an element that is singled out from the whole has an independent distinctive position within the composite sign. The applicant submits that, for the assessment of the visual similarity, the identical part need not be singled out from the whole.

85      However, this argument of the applicant is based on a misreading of the second sentence of paragraph 43 of the contested decision, which is worded as follows: ‘As consumers tend to pay greater attention to the beginning of a mark than to the end ... and “-MARA” is unlikely to be singled out from the marks as a whole, the coincidence in these letters is not immediately perceivable. ’

86      Contrary to what the applicant claims, the Board of Appeal did not take the view that it was necessary for the identical part to be singled out from the marks at issue as a whole, in order for those marks to be considered similar. The Board of Appeal concluded that there was visual similarity, albeit to a very low degree, between the marks at issue, although it noted that the common element ‘mara’ was not singled out from the marks as a whole. The Board of Appeal essentially took the view that partial identity between two marks was less visible where it concerned a part of the marks which was not singled out from those marks as a whole.

87      In any event, the existence of a merely low degree of similarity between the marks at issue may be established even disregarding that additional point raised by the Board of Appeal.

 Phonetic similarity

88      The Board of Appeal noted, in paragraph 44 of the contested decision, that the marks were pronounced ‘zy-ma-ra’ and ‘fe-ma-ra’ in accordance with the pronunciation rules of the respective languages. It noted that, while it is true that, in English, the letter ‘y’ and the letter ‘e’ were sometimes pronounced in a similar way, that was not, however, so in the present case. The Board of Appeal noted that the letter ‘Y’ in ‘ZYMARA’ was pronounced like an ‘i’ and the ‘E’ in ‘FEMARA’ like the ‘e’ in ‘lemon’. The Board of Appeal concluded that the phonetic similarity was low.

89      The applicant submits that, in the present case, the initial vowels of the marks at issue will be pronounced by an English-speaking public in the same way, as an ‘i’.

90      EUIPO maintains that the Board of Appeal correctly identified the pronunciation of the marks at issue.

91      It should be noted, first, that the public to be taken into consideration is that of the entire European Union, since the earlier mark is an EU trade mark.

92      The applicant maintains that all languages of the European Union must be taken into account when assessing the phonetic similarity of the signs. The specific arguments raised by the applicant in order to demonstrate a high degree of phonetic similarity concern, however, the English-speaking part of the relevant public. As regards the non-English-speaking part of the relevant public, the applicant did not put forward any argument capable of calling into question the Board of Appeal’s assessment that the degree of phonetic similarity was low (see, for more details, paragraph 145 below).

93      It is therefore particularly in relation to the perception of the English-speaking public that the phonetic similarity of the marks at issue should be examined.

–       The pronunciation of the mark applied for

94      As regards the pronunciation of the mark applied for, it should be noted that the Board of Appeal found that ‘the “y” of “ZYMARA” [was] pronounced as an “I”’. The applicant maintains that the initial vowel in that mark is pronounced as an ‘i’, something which EUIPO does not dispute.

95      The interpretation of the parties, which take the view, essentially, that the initial vowel of the mark applied for will be pronounced like the sound [ɪ] in the international phonetic alphabet published by the International Phonetic Association, must be upheld.

96      As regards the pronunciation of the first letter of the mark Zymara, the applicant claims that it will be pronounced like the letter ‘s’.

97      In that regard, it should be noted that, in English, the letters ‘z’ and ‘s’ are not pronounced in the same way. The letter ‘z’ is a voiced consonant, while the letter ‘s’ is an unvoiced consonant. As EUIPO maintains, in English, the words ‘zip’ and ‘sip’ are not pronounced in the same way.

–       The pronunciation of the earlier mark

98      The applicant puts forward several arguments in order to challenge the Board of Appeal’s assessment that the English-speaking public would pronounce the letter ‘e’ of the mark FEMARA like the letter ‘e’ in the English word ‘lemon’ (that is to say like the [e] sound in the international phonetic alphabet).

99      In the first place, the applicant claims that the Board of Appeal infringed Article 76(1) of Regulation No 207/2009, which provides that, in proceedings relating to relative grounds for refusal of registration, EUIPO is to be restricted in its examination to the facts, evidence and arguments provided by the parties and the relief sought. In that regard, it maintains that Meda did not, during the administrative procedure, challenge the applicant’s argument relating to the pronunciation of the letter ‘e’ in the mark FEMARA. The applicant submits that, when the pronunciation of a particular vowel is not disputed, EUIPO may not adopt a different pronunciation rule out of the blue and that the Board of Appeal ‘cannot interfere for the benefit of one of the parties by adapting their own facts’.

100    This argument must be rejected. The question of the pronunciation of a word in an EU language is a question of well-known facts. Restricting the factual basis of the examination by EUIPO does not preclude it from taking into consideration, in addition to the facts which have been expressly put forward by the parties, facts which are well known, that is, facts which are likely to be known by anyone or which may be learnt from generally accessible sources (see judgments of 9 December 2010, Tresplain Investments v OHIM — Hoo Hing (Golden Elephant Brand), T‑303/08, EU:T:2010:505, paragraph 67 and the case-law cited, and of 19 June 2014, Kampol v OHIM — Colmol (Nobel), T‑382/12, not published, EU:T:2014:563, paragraph 57 and the case-law cited).

101    Furthermore, when a party to opposition proceedings relies on an allegedly well-known fact, such as in the present case the pronunciation of the letter ‘e’ in the earlier mark as an ‘i’, and the other party does not challenge it, that does not mean that the Board of Appeal must regard this as an established fact. It should be noted that the rules of pronunciation of EU languages are not subject to the discretion of the parties and that it is for the Board of Appeal to make its own assessment of the pronunciation of the marks at issue. In that regard, it must be stated that Article 76(1), in fine, of Regulation No 207/2009 cannot have the purpose of compelling the Opposition Division or Board of Appeal consciously to adopt a decision on the basis of factual hypotheses which are manifestly incomplete or contrary to reality (judgment of 22 June 2004, Ruiz-Picasso and Others v OHIM — DaimlerChrysler (PICARO), T‑185/02, EU:T:2004:189, paragraph 32).

102    Moreover, in accordance with Rule 20(3) of Commission Regulation (EC) No 2868/95 of 13 December 1995 implementing Council Regulation (EC) No 40/94 on the Community trade mark (OJ 1995 L 303, p. 1), as amended, if the trade mark applicant submits no observations on the opposition, EUIPO is to base its ruling on that opposition on the evidence before it. A trade mark applicant therefore has a right, but not an obligation, to submit observations and, if he does not submit any, the consequence is not that EUIPO must base its ruling exclusively on the facts as alleged by the opponent, but that it is to base its ruling on the evidence before it. EUIPO cannot therefore be required to base its rulings on facts alleged by the opponent, where they are contrary to well-known facts, even if they have not been disputed by the other party.

103    As regards well-known facts, the Board of Appeal is not therefore bound by the arguments of the parties, contrary to what the applicant claims.

104    The applicant also submits that Meda has not adduced any evidence concerning the way in which the earlier mark is pronounced. In that regard, it should be noted that facts which are well known may be taken into consideration by the Board of Appeal even if they have not been explicitly put forward by the parties to the opposition proceedings. A fortiori, parties to opposition proceedings are not required to prove facts which are well known (judgment of 19 June 2014, Nobel, T‑382/12, not published, EU:T:2014:563, paragraph 57).

105    As regards the applicant’s claim that, when the pronunciation of a particular vowel is not disputed, the Board of Appeal cannot adopt a different pronunciation rule ‘out of the blue’, it should be noted that, if the applicant’s assertion were to be interpreted as a claim of infringement of the second sentence of Article 75 of Regulation No 207/2009, that argument would, in any event, have to be rejected.

106    Under the second sentence of Article 75 of Regulation No 207/2009, EUIPO’s decisions are to be based only on reasons or evidence on which the parties concerned have had an opportunity to present their comments. In the present case, there is no doubt that the applicant did indeed have an opportunity to present its comments on the pronunciation of the earlier mark, which, moreover, is something that it did. The right to be heard extends to all the matters of fact or of law which form the basis for the decision-making act, but not to the final position which the administration intends to adopt (see judgment of 13 April 2011, Safariland v OHIM — DEF-TEC Defense Technology (FIRST DEFENSE AEROSOL PEPPER PROJECTOR), T‑262/09, EU:T:2011:171, paragraph 80 and the case-law cited). An applicant may therefore not validly claim a right to be heard in respect of each well-known fact on which the Board of Appeal relies in its statement of reasons (judgment of 1 June 2016, Wolf Oil v EUIPO — SCT Lubricants (CHEMPIOIL), T‑34/15, not published, EU:T:2016:330, paragraph 83).

107    In the second place, the applicant maintains that EUIPO did not satisfy the requirements to give reasons. It claims that, even if EUIPO was not bound by the uncontested pronunciation of the mark FEMARA, the Board of Appeal ought to have set out the reasons why the letter ‘e’ in that mark should be pronounced like the ‘e’ in the word ‘lemon’. It maintains that the Board of Appeal did not give any reasons why, in its view, the first vowel of the word ‘femara’ and the first vowel of the word ‘lemon’ would be pronounced in the same way.

108    In this regard, it must be recalled that, in accordance with Article 75 of Regulation No 207/2009, decisions of EUIPO must state the reasons on which they are based. That duty to state reasons has the same scope as that under Article 296 TFEU, pursuant to which the reasoning of the author of the act must be shown clearly and unequivocally. The purpose of that duty is twofold: first, to enable the persons concerned to ascertain the reasons for the measure in order to defend their rights and, secondly, to enable the Courts of the European Union to exercise their power to review the legality of the decision (judgment of 19 May 2010, Zeta Europe v OHIM (Superleggera), T‑464/08, not published, EU:T:2010:212, paragraph 47; see also, to that effect, judgment of 21 October 2004, KWS Saat v OHIM, C‑447/02 P, EU:C:2004:649, paragraphs 63 to 65). However, the Boards of Appeal cannot be required to provide an account that follows exhaustively and one by one all the lines of reasoning articulated by the parties before them (see judgment of 28 November 2013, Herbacin cosmetic v OHIM — Laboratoire Garnier (HERBA SHINE), T‑34/12, not published, EU:T:2013:618, paragraph 42 and the case-law cited).

109    As regards well-known facts, it is settled case-law that the bodies of EUIPO are not obliged to establish in their decisions the accuracy of such facts (see judgment of 1 March 2016, Peri v OHIM (Multiprop), T‑538/14, not published, EU:T:2016:117, paragraph 14 and the case-law cited). The Board of Appeal was accordingly not obliged to establish the accuracy of its finding concerning the pronunciation of the letter ‘e’ in the earlier mark, which means that it was not under an obligation to explain how it arrived at that finding.

110    The applicant’s argument alleging that the reasoning given in the contested decision was inadequate must therefore be rejected.

111    In the third place, the applicant maintains that where EUIPO has accepted that two letters are pronounced in a similar or identical way, then this is a matter of fact which cannot be called into question in the subsequent decision.

112    In making this claim, the applicant, in essence, maintains that, where the Opposition Division has found that a letter was pronounced in a certain way, the Board of Appeal may not depart from that assessment.

113    In the present case, the Opposition Division had taken the view that, in certain languages, such as, for example, English, the letters ‘e’ and ‘y’ were pronounced almost identically, like an ‘i’. The applicant, in essence, claims that the Board of Appeal was not entitled to depart from that assessment in the absence of evidence from Meda that the rules of pronunciation had changed since that time.

114    The applicant’s argument must be rejected. It should be noted that, under Article 62(1) of Regulation No 207/2009, by virtue of the appeal brought before it, the Board of Appeal is called upon to carry out a new, full examination of the merits of the opposition, in terms of both law and fact (judgment of 13 March 2007, OHIM v Kaul, C‑29/05 P, EU:C:2007:162, paragraph 57).

115    Moreover, it is apparent from settled case-law that the decisions concerning the registration of a sign as an EU trade mark which the Boards of Appeal of EUIPO take under Regulation No 207/2009 are adopted in the exercise of circumscribed powers and are not a matter of discretion. Accordingly, the legality of those decisions must be assessed solely on the basis of that regulation and not on the basis of any previous decision-making practice (judgment of 26 April 2007, Alcon v OHIM, C‑412/05 P, EU:C:2007:252, paragraph 65, and judgment of 24 November 2005, Sadas v OHIM — LTJ Diffusion (ARTHUR ET FELICIE), T‑346/04, EU:T:2005:420, paragraph 71).

116    The Board of Appeal was therefore not bound by the findings of the Opposition Division with regard to the pronunciation of the initial vowels of the marks at issue.

117    In the fourth place, the applicant submits that the Board of Appeal’s assessment in respect of the pronunciation, by an English-speaking public, of the first vowel in the mark FEMARA is incorrect. It maintains that it submitted several examples of English words in which the letter ‘e’ is pronounced as an ‘i’. The applicant claims that it is very likely that the earlier mark will be pronounced ‘fi-ma-ra’.

118    In that regard, EUIPO maintains that it would be counter-intuitive for a native English speaker to pronounce the first syllable of the mark FEMARA like the first two letters of the English words ‘feline’ or ‘female’, that is to say [fiː] in accordance with the international phonetic alphabet. An EU English speaker would pronounce the first syllable of the mark FEMARA like the first two letters of the English word ‘feminine’, that is to say [fe] in accordance with the international phonetic alphabet.

119    EUIPO has produced, in Annex B1 to the response, a list of words starting with the letter combination ‘fe’, taken from the Pocket Oxford Dictionary, 1996 edition. It maintains that it is clear that the ‘e’ in the letter combination ‘fe’, when placed before a consonant, is pronounced in the great majority of cases like the ‘e’ in the English word ‘lemon’. Furthermore, it submits that, out of approximately 100 entries where the letter combination ‘fe’ is followed by a consonant, it is pronounced [fiː] only in six cases, namely ‘feline’, ‘female’, ‘femur’, ‘fever’, ‘feverfew’ and ‘feverish’.

120    In that regard, the applicant claims that Annex B1 is illegible and accordingly inadmissible.

121    That argument must be rejected. Annex B1 is indeed legible. As EUIPO points out, if the applicant had received an illegible document, it should have contacted the Court in order to resolve that problem.

122    When questioned on that issue at the hearing, the applicant maintained that it had received Annex B1 by fax and that the copy which it received was illegible. Furthermore, it conceded that it had not explicitly requested that Annex B1 be resent. It must be pointed out that the fact that the copy received by the applicant by fax may have been illegible cannot call into question the admissibility of Annex B1.

123    It must also be noted that it follows from the case-law that EUIPO may submit documents to the Court in order to demonstrate the accuracy of a well-known fact, even if it was not established in the EUIPO decision contested before the Court (see judgment of 1 March 2016, Multiprop, T‑538/14, not published, EU:T:2016:117, paragraph 14 and the case-law cited).

124    It should be noted that the pronunciation [fe] of the first syllable of the mark FEMARA (in accordance with the international phonetic alphabet), that is to say, like the first two letters of the English word ‘feminine’, is the most natural for an English speaker, as confirmed by Annex B1. As EUIPO maintains, it is apparent from that annex that, out of 100 entries in which the letter combination ‘fe’ is followed by a consonant, it is pronounced [fiː] in six cases only.

125    The applicant’s argument that the vowel ‘e’ in the letter combination ‘fe’ is pronounced as an ‘e’ if the letter combination is followed by at least two consonants and that the vowel ‘e’ in that letter combination is pronounced, by contrast, as an ‘i’ if it is followed by a single consonant, as in the words ‘female’, ‘fever’ or ‘femur’, must be rejected. EUIPO rightly referred to the following examples, in which the letter ‘e’ is not pronounced as an ‘i’: ‘fecund’, ‘federal’, ‘felicity’, ‘felon’, ‘feral’, ‘fetid’, ‘fetish’. The word ‘feminine’ is a further example. The pronunciation rule invoked by the applicant therefore does not exist.

126    The applicant claims furthermore that, for the purposes of assessing the similarity of signs, it is sufficient to note that it cannot be ruled out that the pronunciation ‘fi-ma-ra’ is the correct pronunciation. It maintains that it is not unlikely that the mark FEMARA will be pronounced ‘fi-ma-ra’ in so far as a consumer might consider that there is a connection with the word ‘female’, which is pronounced ‘fi-male’; As long as there is a ‘reasonable risk’ that the mark FEMARA is pronounced ‘fi-ma-ra’, it submits, EUIPO must take into consideration that pronunciation regardless of whether different pronunciations exist.

127    In that regard, it should be noted that, where there are several significant groups of consumers who pronounce a mark differently, it is sufficient that there is a likelihood of confusion for one of those significant groups (see, to that effect, concerning the interpretation of a figurative mark as representing a specific letter, judgment of 10 November 2011, Esprit International v OHIM — Marc O’Polo International (Representation of a letter on a pocket), T‑22/10, not published, EU:T:2011:651, paragraphs 118 and 119).

128    It is therefore necessary to examine whether a significant part of the English-speaking public will pronounce the first vowel of the mark FEMARA in the same way as the first vowel of the English word ‘female’.

129    In that regard, it must be recalled that the most natural pronunciation of the first syllable of the mark FEMARA corresponds to the sound [fe] in the international phonetic alphabet.

130    The applicant claims, however, that is not unlikely that the mark FEMARA will be pronounced ‘fi-ma-ra’ in so far as an English-speaking consumer might consider there to be a link with the English word ‘female’ and that the element ‘fem’ refers to the fact that the product is aimed at women.

131    It must, however, be noted that, even if part of the English-speaking public would understand the beginning of the earlier mark as being a reference to the fact that the product sold under that mark is aimed at women, there is no reason to consider that that part of the relevant public will pronounce the first syllable of the mark FEMARA in the same way as the first two letters of the English word ‘female’. Contrary to what the applicant claims – in maintaining that a consumer who understands the beginning of the earlier mark as a reference to the fact that the product sold under that mark is aimed at women will also pronounce it in the same way as ‘the’ word referring to women, namely the word ‘female’ – there is no single word which refers to women in English.

132    In English, the first two letters of the word ‘feminine’ are pronounced [fe], in accordance with the international phonetic alphabet. The first two letters of the English words ‘femininity’, ‘feminism, ‘feminize’ and ‘Fem Lib’, which all refer to women, are also pronounced [fe], in accordance with the international phonetic alphabet.

133    The mere fact that part of the relevant public might understand the beginning of the earlier mark as being a reference to the fact that the product sold under that mark is aimed at women would therefore, if proved, not be sufficient for the view to be taken that a significant part of the English-speaking public would pronounce the first syllable of the earlier mark as [fiː]. Given that, among English words referring to women which contain the element ‘fe’, that element is, in the vast majority of cases, pronounced [fe], in accordance with the international phonetic alphabet, there is no reason for an English speaker to pronounce the first two letters of the earlier mark differently.

134    It follows from the foregoing that the Board of Appeal was right to find that the first vowel of the earlier mark will be pronounced like the letter ‘e’ in the English word ‘lemon’.

–       The phonetic comparison of the marks at issue

135    It should be noted that the second and third syllables of the marks at issue, namely ‘ma’ and ‘ra’, coincide. The English-speaking part of the relevant public will pronounce those two syllables in the same way in both marks.

136    The applicant claims that the pronunciation of the first letter of the mark applied for and that of the first letter of the earlier mark are similar and that the pronunciation of the first vowel of the mark applied for and that of the first vowel of the earlier mark are identical, a claim which EUIPO disputes.

137    It should be noted that the applicant submits that the first letter of the mark applied for is pronounced as an ‘s’ (see paragraph 96 above).

138    The applicant maintains that the letter ‘s’ — as in the English word ‘sip’ — is a voiceless sibilant, sibilants being a subset of fricatives. The letter ‘f’, as in the English word ‘fine’, is a voiceless labiodental fricative. The applicant submits that, since both letters are fricatives, they are pronounced in a very similar way.

139    That argument cannot be upheld. It should be noted that the first letter of the earlier mark will be pronounced, not like the letter ‘s’ in the English word ‘sip’, but like the letter ‘z’ in the English word ‘zip’ (see paragraph 97 above). It is a voiced consonant, whereas the letter ‘f’ is an unvoiced consonant.

140    It is true that the letters ‘z’ and ‘f’ belong to the group of fricatives. That fact alone is not, however, sufficient for the view to be taken that they are pronounced in a similar way. On the contrary, the pronunciation of the letter ‘z’ is very different from the pronunciation of the letter ‘f’, since ‘z’ is a sibilant.

141    It follows from the foregoing that the pronunciation of the initial letter of the mark applied for is very different from that of the initial letter of the earlier mark.

142    As regards the pronunciation of the initial vowels, that is to say, the second letters, of the marks at issue, it should be noted that the sound [ɪ] of the international phonetic alphabet, which corresponds to the pronunciation of the first vowel of the mark applied for, is clearly distinct from the sound [e], which corresponds to the pronunciation of the first vowel of the earlier mark.

143    It must therefore be held that the pronunciation of the first syllable of the mark applied for and that of the first syllable of the earlier mark, which are likely to attract the relevant consumer’s attention to a greater extent, are very different. In that context, it must be stressed that a sibilant is very clearly distinguished from other consonants. As regards hushing sibilants, the Court noted, in the judgment of 25 November 2010, Vidieffe v OHIM — Ellis International Group (GOTHA) (T‑169/09, not published, EU:T:2010:484, paragraph 42), concerning the phonetic comparison of the marks GOTHA and gotcha, that the presence of the letter ‘c’ in the mark gotcha had the effect of transforming the ‘th’ sound into a hushing sound and that, accordingly, the words ‘gotha’ and ‘gotcha’ were pronounced differently.

144    Under those circumstances, it should be considered that the degree of phonetic similarity between the marks at issue is low for the English-speaking section of the relevant public.

145    In so far as concerns the non-English-speaking part of the relevant public, it is also necessary to take the view that the degree of phonetic similarity between the marks at issue is low. For that part of the relevant public, the initial syllables of the marks at issue are also pronounced very differently. The applicant has not put forward any argument capable of allowing the view that, in a language other than English, the pronunciation of the syllable ‘fe’ and that of the syllable ‘zy’ are similar. While it is true that the applicant has claimed that, in most languages, the letters ‘y’ and ‘i’ are pronounced identically, it has not identified any language other than English in which, in its view, the letter ‘e’ of the mark FEMARA would be pronounced as an ‘i’.

146    The Board of Appeal was therefore right to conclude that there was a low degree of phonetic similarity between the marks at issue.

 Conceptual similarity

147    The Board of Appeal found that neither of the marks at issue had any meaning in the relevant territory.

148    The parties expressly agree with that finding.

149    It is necessary to uphold the Board of Appeal’s finding that neither of the marks at issue has any meaning in the relevant territory. This means that a conceptual comparison between the marks at issue is not possible.

150    However, it should be pointed out that the applicant’s reasoning is not consistent in so far as it claims, first, that part of the relevant public will recognise the first part of the earlier mark as a reference to the fact that the product sold under that mark is aimed at women, and, secondly, that neither of the marks at issue has any meaning in the relevant territory.

 The likelihood of confusion

151    A global assessment of the likelihood of confusion implies some interdependence between the factors taken into account and, in particular, between the similarity of the trade marks and that of the goods or services covered. Accordingly, a low degree of similarity between those goods or services may be offset by a high degree of similarity between the marks, and vice versa (judgments of 29 September 1998, Canon, C‑39/97, EU:C:1998:442, paragraph 17, and of 14 December 2006, VENADO with frame and others, T‑81/03, T‑82/03 and T‑103/03, EU:T:2006:397, paragraph 74).

152    In the context of a global assessment of the likelihood of confusion, the Board of Appeal noted that the goods in question were identical or similar and that the marks were similar to a low degree. It considered that, given that the difference lay at the beginning of the marks at issue and that the relevant public was likely to display a high level of attention with regard to the goods in question, there was no likelihood of confusion in the present case.

153    The applicant submits that there is a likelihood of confusion in the present case. It claims in particular that the phonetic comparison is more important than the visual comparison.

154    EUIPO contests the applicant’s arguments.

155    It should be recalled that, in the present case, the goods in question are partly identical, that they have in part a relatively high degree of similarity and that they have in part at least an average degree of similarity (see paragraphs 52, 61 and 70 above). Furthermore, it should be noted that the degree of visual similarity and the degree of phonetic similarity of the marks at issue are low (see paragraphs 80 and 146 above) and that a conceptual comparison is not possible (see paragraph 149 above).

156    In that context, it should be noted that the applicant’s argument that the phonetic comparison is more important than the visual comparison is ineffective, since the marks at issue are visually and phonetically similar to the same degree.

157    It should also be recalled that, for all the goods at issue, the relevant public has a high level of attention (see paragraph 48 above).

158    In the light of the fact that the level of attention of the relevant public is high and that the degree of visual similarity and phonetic similarity is merely weak, it must be concluded that there is no likelihood of confusion between the marks at issue, even in respect of identical goods.

159    The Board of Appeal was therefore right to conclude that there was no likelihood of confusion between the marks at issue for the goods in question.

160    The errors committed by the Board of Appeal, first, in classifying the degree of visual similarity between the signs at issue as very low rather than as low (see paragraph 80 above) and, secondly, in classifying as low the degree of similarity between ‘sanitary preparations for medical purposes’ and ‘dietetic substances adapted for medical use’, on the one hand, and ‘pharmaceutical preparations for the treatment of breast cancer’, on the other hand (see paragraphs 61 and 70 above), cannot result in the annulment of the contested decision, since they had no influence on the outcome (see, to that effect, judgment of 15 January 2003, Mystery Drinks v OHIM — Karlsberg Brauerei (MYSTERY), T‑99/01, EU:T:2003:7, paragraph 36).

161    The applicant’s single plea in law must therefore be rejected and, consequently, the action must be dismissed in its entirety.

 Costs

162    Under Article 134(1) of the Rules of Procedure of the General Court, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.

163    Since the applicant has been unsuccessful, it must be ordered to pay the costs, in accordance with the form of order sought by EUIPO.

On those grounds,

THE GENERAL COURT (Fifth Chamber)

hereby:

1.      Dismisses the action;


2.      Orders Novartis AG to pay the costs.


Gratsias

Dittrich

Xuereb


Delivered in open court in Luxembourg on 21 September 2017.


E. Coulon

 

D. Gratsias

Registrar

 

President


*      Language of the case: English.