CURIA - Documents

Home > Search form > List of results > Documents

Start printing

Language of document : ECLI:EU:C:2012:293

OPINION OF ADVOCATE GENERAL

MAZÁK

delivered on 15 May 2012 (1)

Case C‑457/10 P

AstraZeneca AB and AstraZeneca plc

European Commission

(Appeals – Competition – Abuse of a dominant position – Market in anti-ulcer medicines – Abuse of procedures relating to supplementary protection certificates for medicinal products and of marketing authorisation procedures for medicinal products – Misleading representations – Deregistration of marketing authorisations – Obstacles to the marketing of generic medicinal products and to parallel imports)

I – Introduction

1. By their appeal, AstraZeneca AB and AstraZeneca plc (‘the appellants’) seek to have set aside the judgment of the General Court of the European Union of 1 July 2010 in Case T‑321/05 AstraZeneca v Commission, (2) whereby the General Court largely dismissed their action for annulment of Commission Decision C(2005) 1757. (3) In accordance with the contested decision, the Commission imposed a fine of EUR 60 million on those companies for having abused the patents system and the procedures for marketing pharmaceutical products in order to prevent or delay the arrival of competing generic medicinal products on the market and to impede parallel trade.

2. The European Federation of Pharmaceutical Industries and Associations (EFPIA), which intervened in the case at first instance in support of the form of order sought by the appellants, has lodged a cross-appeal seeking to have set aside the judgment under appeal and the annulment of the contested decision. A cross-appeal has also been lodged by the Commission seeking to have set aside the judgment under appeal in so far as it annulled in part and varied the contested decision.

II – Facts at the origin of the dispute

3. The AstraZeneca plc group form a pharmaceutical group (‘AZ’) which is active worldwide in the sector of the invention, development and marketing of innovative products. Its business is focused on a number of pharmaceutical domains including, in particular, the domain of gastrointestinal conditions. In that regard, one of the main products marketed by AZ is known as Losec, a brand name used in most European markets. That omeprazole-based medicinal product, used in the treatment of gastrointestinal conditions linked with hyperacidity and, in particular, to inhibit proactively acid secretion into the stomach, was the first on the market to act directly on the proton pump, that is to say, the specific enzyme inside the parietal cells along the stomach wall, which pumps acid into the stomach.

4. On 12 May 1999, Generics (UK) Ltd and Scandinavian Pharmaceuticals Generics AB complained to the Commission of AZ’s conduct aimed at preventing them from introducing generic versions of omeprazole on a number of EEA markets. By decision of 9 February 2000 the Commission ordered AZ to submit to investigations at its premises in London and Södertälje. On 25 July 2003, the Commission adopted a decision to initiate the procedure and on 29 July 2003 it sent a statement of objections to AZ. Following a number of oral and written exchanges between 2003 and 2005, the Commission on 15 June 2005 adopted the contested decision, in which it found that AstraZeneca AB and AstraZeneca plc had committed two abuses of a dominant position, thereby infringing Article 82 EC (now Article 102 TFEU) and Article 54 of the EEA Agreement.

5. According to Article 1(1) of the contested decision, the first abuse consisted in misleading representations to patent offices in Germany, Belgium, Denmark, Norway, the Netherlands and the United Kingdom and also before the national courts in Germany and Norway. The Commission considered in that regard that those representations formed part of an overall strategy designed to keep manufacturers of generic products away from the market by obtaining or maintaining supplementary protection certificates (‘SPCs’) (4) for omeprazole to which AZ was not entitled or to which it was entitled for a shorter duration.

6. Under Article 1(2) of the contested decision, the second abuse consisted in the submission of requests for deregistration of the marketing authorisations for Losec capsules in Denmark, Norway and Sweden, combined with the withdrawal of Losec capsules from the market and the launch of Losec MUPS (‘Multiple Unit Pellet System’) tablets in those three countries. In the Commission’s submission, those steps were taken in order to ensure that the abridged registration route provided for in point 8(a)(iii), paragraph 3, of Article 4 of Council Directive 65/65/EEC of 26 January 1965 on the approximation of provisions laid down by law, regulation or administrative action relating to proprietary medicinal products (5) would not be available to producers of generic omeprazole and they also had the consequence that parallel importers were likely to lose their parallel import licences. The Commission took issue, in particular, with the appellants’ strategic implementation of the regulatory framework in order to artificially protect from competition products that were no longer protected by a patent and for which the period of data exclusivity had expired.

7. For those two abuses, the Commission imposed on the appellants jointly and severally a fine of EUR 46 million and on AstraZeneca AB a fine of EUR 14 million.

8. By application lodged at the Registry of the General Court on 25 August 2005, the appellants brought an action for annulment of the decision in issue. That action called into question the lawfulness of that decision with respect to the definition of the relevant market, the assessment of dominance, the first and second abuses of a dominant position and the amount of the fines.

9. By the judgment under appeal, the General Court upheld the contested decision in large part. The General Court however annulled Article 1(2) of the contested decision relating to the second abuse in so far as it found that the appellants had infringed Article 82 EC and Article 54 of the EEA Agreement by requesting the deregistration of the Losec capsule marketing authorisations in Denmark and Norway in combination with the withdrawal from the market of Losec capsules and the launch of Losec MUPS tablets in those two countries, inasmuch as it was found that those actions were capable of restricting parallel imports of Losec capsules in those countries. The General Court therefore reduced the amount of the fine imposed jointly and severally on the appellants to EUR 40 250 000 and the fine imposed on AstraZeneca AB to EUR 12 250 000 and dismissed the action for the remainder.

III – Forms of order sought by the parties before the Court

10. By their appeal, the appellants claim that the Court should set aside the judgment under appeal and annul the contested decision; in the alternative, reduce, at the Court’s discretion, the fine imposed on the appellants by Article 2 of the contested decision; and order the Commission to pay the costs at first instance and on appeal.

11. EFPIA claims that the Court should set aside the judgment under appeal and annul the contested decision and order the Commission to pay the costs at first instance and on appeal, including those relating to EFPIA’s intervention.

12. The Commission contends that the Court should dismiss the appeal and EFPIA’s cross-appeal; allow the Commission’s cross-appeal; order the appellants to pay the costs of the appeal and order EFPIA to pay the costs of its cross-appeal.

IV – The appeal

13. The appellants’ grounds of appeal may be classified under four headings.

A – First heading: definition of the relevant product market

14. The appellants raise two grounds of appeal with respect to market definition.

1. First ground: failure to properly consider the gradual nature of the increase in sales of PPIs at the expense of H2 blockers

a) Argument

15. The appellants claim that the General Court made an error of law in failing to examine properly the relevance of the gradual nature of the increase in the use of proton pump inhibitors (PPIs) at the expense of H2 blockers (antihistamines). This plea is divided into two parts.

16. Firstly, the appellants claim that the General Court failed to conduct a temporal analysis. Thus, the judgment under appeal, and in particular paragraphs 66 to 82 thereof, does not recognise the need to examine the development of the competitive relationship between PPIs and H2 blockers during the relevant infringement periods and does not take account of the changes which occurred on the relevant geographic markets. It is wrong as a matter of law to adjudicate on the relevant product market in a particular country in 1993 on the basis of the state of competition in 2000. Furthermore, the fact that the relationship between PPIs and H2 blockers changed over time is clear from the statements of the medical experts on which the General Court relied.

17. Secondly, the appellants claim that the General Court failed to recognise the relevance of the inertia that characterised the dissemination of knowledge relating to PPIs within the medical community and prescribing practices, which was the reason for the gradual replacement over time of H2 blockers by PPIs. The General Court was wrong to reject, at paragraphs 83 to 107 of the judgment under appeal, the appellants’ argument that H2 blockers necessarily exercised considerable competitive constraint on PPIs, since sales of PPIs increased only gradually at the expense of H2 blockers and therefore less rapidly than would have been expected given the therapeutic superiority of PPIs. The appellants submit, in particular, that the General Court artificially compartmentalised the different advantages and disadvantages of H2 blockers and PPIs, which were necessarily interlinked. In effect, if a doctor decides to prescribe an H2 blocker because he has concerns about the side effects of PPIs, that decision is not a function solely of concerns about PPIs, but necessarily concerns an evaluation of the quality and therapeutic profile of H2 blockers, including the fact that they present fewer risks.

18. EFPIA, which supports this first plea, claims that the General Court reversed the burden of proof by requiring that the appellants show that the gradual replacement of H2 blockers by PPIs is relevant to market definition.

19. The Commission contends that the first ground of appeal is ineffective, because it challenges only one of the elements of the General Court’s reasoning. The gradual nature of the substitution trends is only one aspect of the overall assessment of the relevant market and any error of law in relation to that aspect would not undermine that assessment. The Commission further claims that a large part of this plea is inadmissible in that it requests the Court to reappraise findings of fact.

20. The Commission submits that in any event this plea is unfounded. As regards the first part, the Commission maintains that the General Court did not restrict its examination to evidence from the end of the reference period but, on the contrary, directed its attention to the need to establish the existence of the market from the start of the reference period. Furthermore, the General Court was correct to hold that the gradual nature of the growth of a new product is not inconsistent with the existence of a separate product market for that product alone. In addition, the fact, not disputed by the appellants, that the relationship between PPIs and H2 blockers was characterised by ‘asymmetrical’ substitution at the expense of H2 blockers and a repositioning of H2 blockers towards milder gastrointestinal conditions is relevant for the purpose of demonstrating that H2 blockers did not constitute a significant competitive constraint on PPIs. Last, the emergence of a ‘new’ market does not necessarily mean that the ‘old’ market has disappeared or that the new market already records more sales than the old market.

21. With regard to the second part, the Commission contends that it is based on a misreading of the judgment under appeal. Thus, the General Court recognised the importance of inertia but held that that did not mean that PPIs had been subjected to significant pressure on the part of H2 blockers during the reference period, since in the present case the inertia resulted primarily from the lack of information on PPIs and not from the qualities of H2 blockers.

b) Assessment

22. In my view, the appellants’ first ground of appeal concerning failure to consider the gradual nature of the increase in the sales of PPIs at the expense of H2 blockers is not, as submitted by the Commission, ineffective. It is true, as claimed by the Commission, that the assessment of the relevant market is based on a number of factors which take into account the entire relevant period between 1993 and 2000, and not just the end of that period. (6) However, I consider that the extent to which products are interchangeable or substitutes is a key element of any assessment of a relevant product market for the purposes of Article 102 TFEU. (7) Given the fact that the sales of PPIs and H2 blockers evolved over time, (8) in the light of the finding of the General Court that the first abuse started in Germany, Belgium, Denmark, the Netherlands and the United Kingdom on 30 June 1993 at the latest (9) and ended in Denmark on 30 November 1994 and in the United Kingdom on 16 June 1994, (10) it is of material importance to the assessment of the behaviour in question pursuant to Article 102 TFEU that the relevant product market was correctly established with respect to the entire relevant period and in particular with reference to 1993 and 1994 by taking account of that evolution.

23. As regards the plea of inadmissibility raised by the Commission, I consider that the appellants, by reciting in their pleadings before the Court of Justice evidence inter alia from a number of medical experts and the IMS report (11) which was also before the General Court, are in large part seeking a reassessment of that evidence by the Court of Justice. Given that appeals are confined to questions of law, the Court of Justice may not carry out that reassessment, in the absence of a claim that the General Court has distorted the clear sense of the evidence. (12) The appellants have not however claimed that the evidence in question was distorted. In my view, the present ground of appeal, to the extent that it seeks a reassessment of the facts in question, is therefore inadmissible.

24. I consider that the present ground of appeal raises nonetheless questions of law which I shall now examine.

25. As regards the first part of the first ground of appeal, the appellants consider that the General Court’s reliance on its findings at paragraphs 68 to 72 of the judgment under appeal to uphold the Commission’s decision concerning the relevant product markets in the different countries between 1993 and 2000 (1999 Denmark) is materially flawed because it fails to take account of the changes in those markets over the relevant period and bases market definition for a given period on the position some years later. Contrary to the appellants’ claim, I consider that the General Court recognised the legal relevance of gradual developments in the relevant markets. It is clear from the judgment under appeal that the General Court examined in detail the patterns of substitution of H2 blockers by PPIs (13) between 1991 and 2000 in the context of the plea before that court on gradual substitution in order to assess whether during the relevant period H2 blockers exercised significant competitive constraint over PPIs. At paragraph 84 of the judgment under appeal, the General Court acknowledged that both the number and value of PPI treatments prescribed increased gradually and it is evident that that court was aware that treatments of H2 blockers were greater than PPIs during part of the relevant period. (14) The General Court however considered that the gradual developments did not support a finding that H2 blockers exercised any significant competitive constraint over PPIs during the relevant period. Those considerations are based on two issues.

26. Firstly, the General Court found at paragraph 91 of the judgment under appeal that, in principle and even in the case of pharmaceutical product markets, the gradual nature of the increase in sales of a new product substituting for an existing product is not sufficient to conclude that the existing product necessarily exercises a significant competitive constraint over the new one. I would note that the appellants however have not challenged that finding in their appeal or the theoretical framework for that finding laid down by that court in paragraphs 86 to 90 of the judgment under appeal. Nor have the appellants challenged the General Court’s finding at paragraph 92 of the judgment under appeal that they had failed to adduce evidence permitting the inference that the gradual increase in sales of PPIs was caused by a significant competitive constraint exercised by H2 blockers. I therefore consider that the General Court correctly found that the appellants merely postulated a presumption of a causal link between the gradual nature of the increase in sales of PPIs and a competitive constraint exercised by H2 blockers over PPIs. The General Court thus rightly found that there was no such presumption in principle and that there were no specific elements in the case to find such a causal link. In so doing, I consider that the General Court did not reverse the burden of proof which lies on the Commission to establish the relevant product markets. The General Court merely stated that the plea in law, raised before it, was unsupported by evidence.

27. Secondly, the General Court found at paragraph 96 of the judgment under appeal that despite the fact that sales of PPIs were much lower than those of H2 blockers in 1993, this did not permit the conclusion that the latter exercised a significant competitive constraint over PPIs during that year as the trend of asymmetrical substitution characterised by the growth in sales of PPIs and the decrease or stagnation in sales of H2 blockers, in conjunction with the finding of a repositioning in the use of H2 blockers towards the treatment of the milder forms of the conditions supported the view that H2 blockers did not exercise any significant competitive constraint over PPIs. These findings by the General Court also have not been challenged by the appellants.

28. In my view, an analysis of which product sells more at a particular point in time may be insufficient for the purposes of defining a relevant market pursuant to competition law. Thus in the case of evolving markets, sales and substitution trends must be examined over time. The mere fact that there were significant sales of H2 blockers at the end of the relevant period does not mean, as suggested by the appellants, that H2 blockers and PPIs were part of the same relevant product market. It is possible for a ‘new’ and ‘old’ product to coexist in two separate markets.

29. I therefore consider that the Court should dismiss the first part of the first ground of appeal as in part inadmissible and in part unfounded.

30. As regards the question of inertia, in my view the appellants’ claim that the relative advantages and disadvantages of PPIs and H2 blockers are necessarily interlinked should be rejected as unfounded as it attempts, in my view, to raise a quasi-presumption which is unsupported by the clear findings of fact of the General Court on the specific circumstances of the case. (15)

31. The General Court acknowledged that the degree of ‘inertia’ of prescribing doctors slowed down sales of PPIs and, accordingly, the process of substitution of PPIs for H2 blockers. (16) However, the General Court found that that does not show that H2 blockers exercised a significant competitive constraint over PPIs. (17) While the General Court expressly accepted that the quality of a pre-existing product may influence the degree of inertia of prescribing doctors if its therapeutic efficacy is deemed sufficient, (18) it found on the basis of the evidence in the file before it that the ‘inertia’ stemmed primarily from caution towards a new product and, more specifically, from concerns as to the possible carcinogenic side effects of PPIs. In addition, the General Court noted, inter alia, that the fact that PPIs were deemed to be the only effective treatment for the severe forms of gastrointestinal conditions, that PPIs and H2 blockers therefore had different therapeutic uses and that the growth in PPIs was in many cases very largely not at the expense of H2 blockers supported the argument that the ‘inertia’ of doctors depended more on the accumulation and dissemination of information on the properties of PPIs than on the quality of H2 blockers. (19) In my view, those findings of fact cannot be challenged on appeal, in the absence of distortion, which is not claimed by the appellants.

32. I also consider that the General Court’s approach to inertia in the context of market definition and dominance is not, as claimed by the appellants, inconsistent. Inertia in doctor’s prescribing practices was examined both in the context of market definition and the assessment of dominance, with rather different conclusions being drawn. However, in my view, such differences can be reconciled with the fact that the definition of a market and the assessment of dominance are quite distinct exercises from a competition law perspective. In addition and more importantly, the different treatment of inertia in the definition of a market and the assessment of dominance is wholly consistent and comprehensible in the light of the General Court’s specific findings of fact. In that regard, the General Court considered that while inertia slowed the process of substitution of PPIs to H2 blockers, this did not show that H2 blockers exercised a competitive constraint over PPIs as the inertia did not stem from the therapeutic qualities of H2 blockers but rather from a lack of knowledge of PPIs, which were in fact therapeutically superior. However, on the question of dominance, the General Court found that in the PPI market, and thus in relation to products which were therapeutically similar, the inertia on the part of prescribing doctors coupled with AZ’s first-mover status and the strong brand image of Losec gave AZ an appreciable competitive advantage. (20)

33. I therefore consider that the Court should dismiss the second part of the first ground of appeal as in part inadmissible and in part unfounded.

2. Second ground: failure to consider the overall cost of treatment for PPIs and H2 blockers when assessing the Commission’s reliance on price factors

a) Argument

34. The appellants submit that the General Court failed to examine the general cost of treatment with PPIs by comparison with the cost of treatment with H2 blockers when it evaluated the price indicators on which the Commission relied. They maintain in that regard that although the cost of a daily dose of PPIs is higher than the cost of a daily dose of H2 blockers, the general cost of treatment is virtually identical because PPIs treat patients more rapidly. Although the General Court recognised that fact at paragraphs 188 and 193 of the judgment under appeal, it held at paragraphs 189 and 190 of that judgment that, since quantification of cost-effectiveness is likely to be particularly complex and uncertain, the Commission did not make a manifest error of assessment in taking into account the price of the medicines for an identical period of treatment. In fact, that approach by the General Court is legally incorrect, in that it reverses the burden of proof. Thus, when the Commission seeks to rely on complex and uncertain factors, such as price indicators, it should either analyse those factors in a satisfactory manner or refrain from relying on them if it is unable to prove them because of their complexity.

35. EFPIA supports this plea and takes issue with the General Court for not having correctly applied the test of substitutability in finding that the Commission had not made a manifest error of assessment in comparing prices for the same treatment period.

36. The Commission contends that this plea is ineffective, as it does not challenge the finding made at paragraph 191 of the judgment under appeal that H2 blockers were not capable of exercising a significant competitive constraint on PPIs by means of lower prices, in view, first, of the limited sensitivity of doctors and patients to price differences and, second, of the regulatory systems in force. This plea is also unfounded. The fact that the decision in issue is based on treatment over 28 days cannot be considered a manifest error of assessment, as it would be impossible to determine the precise duration of each treatment. The Commission maintains in this context that the appellants’ view of the assessment of cost effectiveness is over-simplistic and does not take account of the multitude of conditions and individual treatments possible. Furthermore, the fact that the General Court considered that the data relating to price differentials was relevant indicates that, in spite of the lack of certainty, it considered those data sufficiently reliable to form part of the overall assessment. That assessment cannot be challenged on appeal.

b) Assessment

37. In my view, the present ground of appeal is ineffective. The General Court found at paragraph 196 of the judgment under appeal that price-based indicators constitute an important element of the Commission’s definition of the relevant market in the present case. However, even if the General Court erred by finding at paragraph 190 of the judgment under appeal that the Commission did not commit a manifest error of assessment in taking into account the price of the medicines for an identical period of treatment (28 days), (21) this does not call into question the uncontested findings of the General Court at paragraphs 171 to 175 and 177 of the judgment under appeal that H2 blockers were not capable of exercising a significant competitive constraint over PPIs by means of lower prices. (22)

38. In addition, I consider that the present ground of appeal is unfounded. While the overall price difference between H2 blockers and PPIs may be lower due to the cost-effectiveness of PPIs, as claimed by the appellants and indeed as expressly accepted by the General Court, in my view, the appellants have merely asserted that the General Court erred by accepting the Commission’s reliance on the difference in cost of PPIs and H2 blockers based on a 28-day treatment period. The appellants have not challenged, however, the General Court’s finding that the quantification of cost-effectiveness was likely to be particularly complex and uncertain. I therefore consider that while the 28-day treatment period is not a perfectly reliable price-based indicator, the General Court did not err in finding that the Commission could take it into account in the contested decision when defining the relevant product market along with other more cogent price-based indicators which were outlined in the judgment under appeal.

39. I therefore consider that the second ground of appeal is ineffective and unfounded and should be dismissed by the Court.

B – Second heading: first abuse of a dominant position

40. The appellants raise two grounds of appeal with respect to the first abuse.

1. First ground: no failure to compete on the merits and AZ relying on a bona fides interpretation of the law

a) Argument

41. The appellants consider that the General Court’s approach to competition on the merits is legally flawed. The General Court was wrong, in assessing whether the appellants’ representations to the patent offices were objectively misleading, to have dismissed as irrelevant the reasonableness and bona fides of AZ’s understanding of its legal rights to a SPC pursuant to Article 19 of Regulation No 1768/92.

42. The appellants claim that the General Court misinterpreted the concept of ‘competition on the merits’ by characterising as a breach of such competition the fact that they did not disclose their interpretation of Article 19 of Regulation No 1768/92 to the national patent offices and therefore in particular the circumstance that the reference to the first authorisation on which they relied in support of their applications for SPCs was not the authorisation under Directive 65/65 but the reference to the subsequent authorisation linked with the publication of prices. A ‘lack of transparency’ cannot suffice for an abuse and the General Court ought to have required at least knowledge on the appellants’ part that they were not entitled to the SPC. After dismissing as irrelevant the fact that, at the time of submission of the applications, it was reasonable, given the ambiguity of Article 19 of Regulation No 1768/92, to consider that the appellants were entitled to the SPCs, the General Court set the threshold too low, promoting to the rank of an abuse the mere fact that an undertaking in a dominant position seeks a right from which it thinks it can benefit without disclosing the elements on which it bases its opinion. The General Court’s reasoning is based on the premiss that the appellants were not entitled to the SPC and is therefore made with the benefit of hindsight, taking account of the clarification provided by the judgment in Hässle. (23)

43. The appellants maintain that there are compelling political and legal reasons why deliberate fraud or deceit should be a requirement for a finding of abuse in circumstances such as those of the present case. Thus, a concept of abuse as severe as that applied by the General Court will be likely to impede and delay applications for intellectual property rights in Europe, particularly if it is combined with the Commission’s strict approach to market definition. By way of comparison, in United States law only patents obtained fraudulently can be challenged under competition law, in order not to chill patent applications. A parallel should also be drawn between the case‑law on litigation abuses and the two conditions, objective and subjective, established by the General Court in ITT Promedia v Commission, (24) should be applied, neither of those conditions being satisfied in the present case.

44. EFPIA supports this plea and further submits that, according to the General Court’s interpretation, an ‘objectively misleading’ representation in reality means an ‘objectively wrong’ representation. If that standard were to be applied, dominant undertakings would have to be infallible in their dealings with regulatory authorities. Thus, even an error that was made unintentionally and immediately rectified could give rise to liability under Article 102 TFEU. EFPIA maintains, in particular, that it would be legally indefensible to apply that concept to patent applications, since a number of such applications would be rejected each year on the ground that they were not objectively correct, as their objective did not satisfy the patentability criteria. EFPIA emphasises that patent law is particularly complex and that the search and examination procedures take years to process.

45. The Commission contends that the appellants are attempting, by this plea, to play down the abuse by presenting it as a mere lack of transparency, whereas the General Court found that their conduct was deliberate and highly misleading. Thus, the appellants merely describe the facts in a way that differs from the findings made by the General Court, which, in particular, observed that they could not be unaware that both patent agents and patent offices had understood the concept of ‘marketing authorisation’ as referring to authorisation under Directive 65/65. This plea is therefore inadmissible, since it seeks in reality to obtain a reappraisal of the facts underlying the first abuse.

46. The Commission emphasises that the first abuse consisted not solely in the non-disclosure of a legal interpretation of the SPC Regulation, but also in the fact that the appellants knowingly misled the competent authorities by not disclosing very specific factual information which was necessary for the purpose of determining whether SPCs should be granted and also, where appropriate, their duration. Nor is it necessary to prove bad faith in the context of an abuse of a dominant position, as that abuse is an objective concept. The misleading nature of a representation does not depend on whether or not the person making it perceives it as misleading. The decisive question is whether the conduct was objectively of such a kind as to restrict competition, which the General Court carefully examined. Furthermore, the appellants’ argument is tantamount to saying that if a company believes it can benefit from an exclusive right there is nothing to prevent it from making false, deceptive or misleading representations to the public authorities, which is inconceivable. Last, the judgment in ITT Promedia v Commission (25) is not relevant in the present case.

b) Assessment

47. It is clear from paragraph 496 of the judgment under appeal that the General Court considered that AZ’s purported good faith in the interpretation of Regulation No 1768/92 and the reasonableness of that interpretation was not at all an issue in the first abuse. Indeed, the General Court previously stated that it follows from the objective nature of the concept of abuse that the misleading nature of representations made to public authorities must be assessed on the basis of objective factors and that proof of the deliberate nature of the conduct and of the bad faith of the undertaking in a dominant position is not required for the purposes of identifying an abuse of a dominant position. (26) The appellants submit that the General Court was wrong to dismiss as irrelevant the bona fides of AZ’s interpretation of the law and that to do so in effect made it a per se abuse for a dominant company to apply for a right it thinks it is entitled to without disclosing the basis of its belief.

48. In my view, the appellants’ submissions are wholly unsupported by the very detailed and clear findings of fact of the General Court based on the concrete actions of AZ. In that regard, it must be noted that the General Court found that the representations made by AZ to the patent offices for the purposes of SPC applications were ‘characterised by a manifest lack of transparency’ (27) and were ‘highly misleading’. (28) In accordance with that court’s findings of fact, the SPC applications were presented in such a manner as to lead the patent offices in question to consider that the dates submitted in relation to France and Luxembourg corresponded to the issue of the technical marketing authorisation rather than the date of publication of the price of the medicinal product. (29)

49. I consider therefore that the General Court found that the misrepresentations in question did not rest purely on AZ’s lack of disclosure in the SPC applications as regards its interpretation of Article 19 of Regulation No 1768/92, but rather on highly misleading representations made by AZ during the application procedure. The General Court’s reference in paragraph 494 of the judgment under appeal to the absence of proactive disclosure of the nature of the dates mentioned in relation to the Luxembourg and French marketing authorisations, on the one hand, and of the interpretation of Regulation No 1768/92 which led to the choice of those dates, on the other, cannot be viewed in isolation but rather in the context of the General Court’s detailed factual findings on the highly misleading representations made by AZ during the application procedure. Indeed, the General Court found that AZ had, on a number of occasions, deliberately (30) tried to mislead the relevant authorities by not disclosing factual information which was relevant for the granting of SPCs.

50. It is settled case-law that the concept of abuse of dominance is an objective concept. (31) I consider therefore that, in the context of an abuse of dominance, in assessing whether a particular course of behaviour is misleading, the General Court was not obliged, as claimed by the appellants, to assess AZ’s alleged subjective beliefs on an interpretation of law, bona fides or otherwise, but rather to examine their actual conduct. (32) Moreover, the appellants’ submission regarding a requirement of proof that AZ knew that it was not entitled to an SPC and was thus acting fraudulently, in my view, radically departs from the principle that abuse of dominance is an objective concept. It also constitutes an attempt to apply criminal evidential standards to a procedure which the Court of Justice has stated is administrative rather than criminal in nature (33) and is somewhat incoherent with Article 23(5) of Council Regulation (EC) No 1/2003, (34) which provides that fines imposed pursuant to that provision shall not be of a criminal law nature.

51. The fact that a preliminary reference was made to the Court of Justice in the Hässle (35) case for a clarification of Article 19 of Regulation No 1768/92 or that in 1994, one year after the commencement of the first abuse, two law firms hired by AZ wrote legal opinions supporting the ‘effective marketing theory’ is not relevant and cannot detract from AZ’s objectively misleading representations, which I would stress, in the light of the General Court’s findings, clearly exceeded any bona fide interpretation of the applicable law. In my opinion, the General Court has not, as claimed by the appellants, made it a per se abuse for a dominant company to apply for a right it thinks it is entitled to without disclosing the basis for its belief. Rather, the General Court found that an undertaking in a dominant position may not make objectively misleading representations to public authorities to obtain a right, irrespective of whether that undertaking believes it is entitled to that right. Such an approach does not set a low threshold for abuse and will not in my view have a chilling effect on or delay applications for intellectual property rights in Europe by increasing the regulatory, legal and bureaucratic burden on companies, as claimed by the appellants and also EFPIA, but rather will curtail abuse of dominance resulting from highly misleading representations made to patent, or other intellectual property, authorities.

52. I also consider that the General Court correctly found that the judgment of the General Court in ITT Promedia v Commission (36) was not relevant to the present proceedings. The General Court did not actually rule on the criteria necessary in order to establish whether legal proceedings constitute an abuse of a dominant position in the ITT Promedia v Commission case. Thus the appellants’ reference to those ‘criteria’ in its pleadings is somewhat speculative. (37) In addition, I consider that no meaningful parallel can in any event be drawn between what the appellants’ refer to as litigation and regulatory abuse cases. The extreme restraint that must be exercised, in order to preserve the fundamental right of access to justice, before deeming that litigation is abusive in nature is not warranted in the present case in the absence of any necessity to preserve that fundamental right and also given the fact that the abuse in question was characterised by highly misleading representations to patent authorities.

53. I therefore consider that the Court should dismiss the present ground of appeal as unfounded.

2. Second ground: failure to find an effect on competition or a tendency to restrict competition

a) Argument

54. The appellants maintain that the General Court made an error of law in not correctly identifying the time when the first abuse of a dominant position started. Thus, the General Court wrongly held that the mere fact of applying for an SPC already constituted an abuse, without considering whether competition was affected or whether the impugned conduct had a tendency to restrict competition. If the General Court had carried out such an examination, it ought to have found that an abuse started not with the application for an SPC but only with the grant of that certificate. The appellants further observe that the SPC applications were filed between five and six years before they entered into force and that up to that point their rights were protected by patents.

55. The appellants claim, in particular, that conduct cannot be impugned under Article 102 TFEU on the sole ground that with the benefit of hindsight it is found to be misleading. In order for there to be an exclusionary abuse, the misleading conduct must have either an actual effect on competition or a tendency to have such an effect. Competition could not be affected while the exclusive right sought had not been granted, when the appellants’ competitors were not aware of the exclusive right and when the existence of that right was not liable to affect those competitors’ conduct. In support of their analysis, the appellants rely, in particular, on the Opinion of Advocate General Ruiz-Jarabo Colomer in Sot. Lélos kai Sia and Others, (38) a number of judgments of this Court and of the General Court and also United States competition law, under which there is no abuse unless the patent is implemented.

56. EFPIA also takes issue with the General Court for having held that a misleading representation may constitute an abuse even if it had no external effect because the error was corrected by a patent office or by third parties using correction mechanisms such as opposition procedures or invalidity litigation.

57. The Commission contends that, contrary to the appellants’ assertions, the General Court did not rely on an analysis demonstrating that misleading representations are abusive ‘per se’, but carried out a very thorough examination of the potential effects of the impugned conduct, explaining in detail its reasons for taking the view that such conduct was likely to restrict competition and finding that the conduct in question had produced effects on the market. The Commission refers in that regard to paragraphs 357, 361, 377, 380, 493, 591, 593, 598, 602 to 608 and 903 of the judgment under appeal, which contain findings of fact which are not amenable to review on appeal.

58. In so far as the appellants require that it be established that the abuse has by itself a direct effect on competition, such a claim is contrary to the case-law and it was correctly rejected at paragraphs 376 and 377 of the judgment under appeal. It follows from the case-law, moreover, that the criterion of potential competition could be appropriate for the purpose of defining anti-competitive conduct. In addition, the fact that the effects on the market may depend on further action on the part of the public authorities does not preclude the existence of an abuse. If misleading representations distort those authorities’ decision-making process, the resulting anti-competitive effect is not imputable to the action of the State, but to those representations.

59. As regards the argument relating to the fact that the SPC was not granted in certain countries, the Commission submits that, in so far as the impugned conduct forms part of an overall strategy, the existence of an abuse is not affected by the fact that that strategy did not succeed in certain countries. The decisive criterion is whether the chain of events can be established with sufficient likelihood. Last, the Commission maintains that the solution adopted in United States law cannot be transposed to the European context and that the judgment under appeal, and in particular paragraphs 362 and 368, is sufficiently reasoned in that regard.

b) Assessment

60. It is settled case-law that Article 102 TFEU refers to the conduct of a dominant undertaking which, on a market where the degree of competition is already weakened precisely because of the presence of the undertaking concerned, through recourse to methods different from those governing normal competition in products or services on the basis of the transactions of commercial operators, has the effect of hindering the maintenance of the degree of competition still existing in the market or the growth of that competition. (39)

61. It is therefore necessary that anti-competitive effect be demonstrated. (40)

62. However, the extent to which an anti-competitive effect must be demonstrated in order for an abuse of dominance to be found is the subject of much debate and is of central importance to the proper and timely enforcement of Article 102 TFEU. If the requirement of demonstrating the anti-competitive effect of a practice is set too high, thereby requiring proof of actual effect or a high probability or a likelihood (41) that such an effect will arise, there is a risk that anti-competitive behaviour which is detrimental inter alia to consumers will go unchallenged by the relevant competition authorities because the evidentiary burden placed on them is too high. On the other hand, if the requirement of demonstrating the anti-competitive effects of certain practices is set too low by assuming that they are per se abusive or by requiring little more than a vague or theoretical assertion that they produce anti-competitive effects, this risks stifling the legitimate efforts of dominant companies which compete, perhaps ‘aggressively’, but nonetheless on the merits. There is thus a need to carve out a via media between these two extremes.

63. I therefore consider that competition authorities must demonstrate, in a manner tailored to the specificities and facts of each case, that a particular practice ‘tends’ to restrict competition in the sense that it has the potential to hinder competition. It must thus be demonstrated that it is plausible that the practice harms or will harm competition. Abstract, purely hypothetical or remote assertions or theories of harm, which are not linked to the specificities of the case at hand, will thus not suffice.

64. In order to establish whether a practice has the requisite anti-competitive (potential/plausible) effects, I consider that those effects must be assessed at the time when the practice was actually carried out or implemented. (42) A ‘wait and see’ approach, thereby assessing anti-competitive effects at some later point in time, may be tantamount to introducing a standard approaching a requirement of actual, concrete anti-competitive effects and may raise the evidentiary bar too high. It follows therefore in my view that the actual, subsequent knowledge or reactions of third parties to a particular practice which has already been implemented are also, in principle, not relevant in assessing whether that practice tends to have anti-competitive effects. I fully concur with the General Court’s finding at paragraph 377 of the judgment under appeal that ‘representations designed to obtain exclusive rights unlawfully constitute an abuse only if it is established that, in view of the objective context in which they are made, those representations are actually liable to lead the public authorities to grant the exclusive right applied for’.

65. I consider as a preliminary matter that the appellants’ claim at point 55 above that their behaviour was found to be abusive solely with the benefit of hindsight should not be accepted. As stated above at point 48 et seq., the General Court found as a matter of fact that the SPC applications in question were ‘characterised by a manifest lack of transparency’ and were ‘highly misleading’ and exceeded any bona fide interpretation of the applicable law.

66. I also consider that the SPC applications in question had, at the time they were made, the potential to hinder competition. In that regard, the fact that the SPCs actually granted could only come into force after a number of years when the basic patents expired, or were never actually granted (43) in certain countries, does not detract from the fact that the applications themselves had the potential to detrimentally affect or hinder competition due to the exclusionary effect of SPCs.

67. A finding of anti-competitive effect does not require that the abusive behaviour is successful (44) or, I would submit, is successful within a particular time frame, provided that the anti-competitive effect is not too remote as to be implausible.

68. I consider that the General Court correctly found at paragraph 360 of the judgment under appeal that the fact that certain public authorities did not allow themselves to be misled or that competitors obtained the revocation of the SPCs does not mean that the misleading representations were not capable of having anti-competitive effect at the time they were made. I thus consider that EFPIA’s claim at point 56 above should be rejected. In the case at hand, were it not for the intervention of third parties, it is plausible that the SPC applications would have resulted in the grant of SPCs and given rise to regulatory obstacles to competition. Contrary to the appellants’ submissions before this court, this is not a situation where conduct ‘would only restrict competition if a series of further contingencies were to occur’. Rather, this is clearly more akin to a situation where conduct would restrict competition unless further contingencies (such as the intervention of third parties) occurred to prevent that happening.

69. In my view, the Commission is correct in stating that the additional criterion of ‘knowledge by competitors’ advanced by the appellants would introduce a subjective element into the concept of abuse of dominance which is inconsistent with its objective nature. Moreover, and as indicated by the Commission, since the undertaking in a dominant position may not be able to know whether its competitors are aware of its conduct, that requirement would also run counter to legal certainty.

70. As regards the references by the appellants to United States law, suffice is to state that United States law is not relevant in the context of the present proceedings which concern the application of Article 102 TFEU. The General Court therefore correctly held at paragraph 368 of the judgment under appeal that United States law cannot take precedence over that adopted by the European Union. In any event, I consider that the standard of evidence of anti-competitive effects advocated by the appellants by analogy with United States law should not be accepted. In that regard, the appellants, quoting from a United States District Court (Federal) judgment, (45) note in their pleadings that as ‘a general proposition, merely obtaining a patent by fraud, with no subsequent enforcement attempt, is not an antitrust violation’. Firstly, as I stated at point 50 above, a requirement of fraud constitutes an untoward attempt to apply criminal evidential standards in a field which is not criminal in nature. Secondly, a requirement of potential/plausible anti-competitive effects ensures that Article 102 TFEU creates sufficient deterrents in order to prevent abuse of dominance, while avoiding a formulaic or per se application of that provision which would risk stifling competition on the merits. I therefore consider that a requirement of a subsequent enforcement attempt is clearly approaching a requirement to demonstrate actual anti-competitive effects. Such a requirement thus raises the evidentiary bar in respect of anti-competitive effects too high and would risk greatly diminishing the deterrent effect of Article 102 TFEU. I consider that the General Court correctly found at paragraph 362 of the judgment under appeal that there was no necessity that the SPCs actually be enforced as ‘[t]he mere possession by an undertaking of an exclusive right normally results in keeping competitors away, since public regulations require them to respect that exclusive right’.

71. I therefore consider that the Court should dismiss the present ground of appeal as unfounded.

C – Third heading: second abuse of a dominant position

72. The appellants raise two grounds of appeal with respect to the second abuse of a dominant position.

1. First ground: competition on the merits

a) Argument

73. The appellants claim that the General Court misinterpreted the concept of ‘competition on the merits’ in considering that the mere exercise of a right conferred by EU law is incompatible with such competition. The right to withdraw a marketing authorisation cannot logically be granted by the European Union and at the same time prohibited by the European Union. The appellants maintain in that context that the EU regulation of pharmaceutical matters confers on the holder of a marketing authorisation the right to request the withdrawal of that authorisation, just like the right not to renew it upon its expiry. The Commission itself, and also Advocates General La Pergola and Geelhoed, expressly recognised in Rhône-Poulenc Rorer and May & Baker (46) and Ferring (47) that the owner may exercise that right at any time without having to provide any reasons and without having to take account of the interests of manufacturers of generic products and parallel importers. Those principles also follow from the judgment in the latter case.

74. The appellants emphasise in that regard that the existence of a marketing authorisation imposes stringent pharmacovigilance obligations and the permanent costs which it is lawful to dispose of if the authorised product is no longer marketed. That a company in a dominant position should be deprived of a right of withdrawal and be required to maintain in force an authorisation which it no longer needs and should thus be forced to incur effort and costs and to assume public health liability for the accuracy of the information which it supplies, without any recompense on the part of their competitors, stretches too far the special responsibility of companies in a dominant position. Nor would that withdrawal prevent either parallel imports or the marketing of generic products already on the market.

75. The appellants further take issue with the General Court for having provided insufficient reasons, at paragraph 677 of the judgment under appeal, for its conclusion that the illegality of abusive conduct under Article 102 TFEU is unrelated to its compliance with other legal rules. Thus, the General Court ought to have explained how the exercise by AZ of a legitimate right constituted an abuse in this case. In addition, the EU regulations governing pharmaceutical matters themselves seek to reconcile the encouragement of innovation with the protection of competition. The appellants further contend that the General Court identified as constituting abuse a different set of conduct from that identified by the Commission and in doing so exceeded its jurisdiction.

76. The Commission observes, first of all, that the General Court found that AZ’s intention in withdrawing the marketing authorisations was to hinder the introduction of generic products and parallel imports and that there was no objective justification for its conduct. Next, it observes that the appellants have distorted both the Commission’s position and that of the General Court. The Commission submits that the mere fact that Directive 65/65 imposes no condition as to whether the holder of the marketing authorisation may request the deregistration of a product does not mean that there is a right in favour of that holder which deserves protection. There is a significant difference, moreover, between allowing the authorisation to lapse without requesting its renewal and requesting its withdrawal before expiry of its term of validity in such a way as to raise barriers to the entry on the market of generic products and parallel imports. The contested decision did not establish positive obligations, but found that a series of actions were abusive. The Commission maintains that the illegality of abusive conduct under Article 102 TFEU results from the consequences which it is capable of having on competition and is unrelated to its compliance with other legal regimes. Furthermore, as Directive 65/65 was not adopted on the basis of the provisions of primary law on competition, it does not pursue the same objective as Article 102 TFEU.

b) Assessment

77. As regards the alleged disagreement between the Commission and the General Court on the relevant conduct which constituted the second abuse, (48) I consider that it is clear from paragraph 789 of the contested decision that the Commission considered that the abuse concerned AZ’s selective requests for deregistration of Losec capsules in Denmark, Norway and Sweden combined with the Losec MUPS tablet/Losec capsule switch. At paragraph 792 of the contested decision, the Commission states that single acts involving the launch, withdrawal or requests for deregistration of a pharmaceutical product would not normally be regarded as an abuse. The Commission however clearly underscored at paragraph 793 of the contested decision that it is not its case that the launch of a new formulation of Losec (Losec MUPS) and/or the withdrawal of Losec capsules would as such constitute an abuse. The General Court thus correctly stated, in my view, at paragraph 807 of the judgment under appeal that the central feature of the second abuse consists in the deregistration of the Losec capsule marketing authorisations, the conversion of sales of Losec capsules to Losec MUPS being the context in which the deregistrations of the marketing authorisations were carried out. Thus both the Commission and General Court agree that, while the abuse of dominance consists of the deregistration of the marketing authorisations, the context in which that abuse arose is not irrelevant. Such an approach is, in my view, wholly consistent with a case-by-case assessment of abuse of dominance which takes account of the factual and regulatory framework in which a particular practice takes place and avoids any formulaic methodology.

78. The appellants claim that they had an unfettered right to withdraw their own marketing authorisation and they rely to a great extent on Rhône-Poulenc Rorer and May & Baker (49) and Ferring (50) and in particular the Opinions of the Advocates General and the arguments of the Commission in those cases. It must be stressed that the present proceedings concern the application of Article 102 TFEU and that no reference whatsoever to that provision or indeed any of the rules on competition laid down by the Treaty is contained in the aforementioned judgments or the Advocates General’s Opinions in those cases which concerned, respectively, the application of Directive 65/65, as amended, and the rules on the free movement of goods. Thus any statements in those judgments, opinions of Advocates General or indeed arguments of the Commission cannot be read out of context and transformed into general statements which are necessarily applicable, inter alia, to cases concerning Article 102 TFEU. While a pharmaceutical company may be free in accordance with Directive 65/65 to surrender a marketing authorisation, this does not mean that such behaviour is free from scrutiny pursuant to other rules of EU law, including Article 102 TFEU. Moreover, the fact that Directive 65/65 establishes an EU regulatory scheme rather than a national scheme or that the provisions of that directive may indirectly promote, inter alia, competition in the EU, does not alter that analysis and sanction what would in effect be the non-application of Article 102 TFEU. I would add that given that the legal basis of Directive 65/65 is Article 100 EEC (now Article 114(1) TFEU), the rules of that harmonising directive cannot pre-empt the application of Article 102 TFEU. Moreover, it is clear from the recitals in the preamble to that directive that its primary purpose is to safeguard public health while eliminating disparities between certain national provisions which hinder trade in medicinal products within the Union. Directive 65/65 does not therefore, as claimed by the appellants, pursue much the same objectives as Article 102 TFEU.

79. I therefore fully concur not only with the finding of the General Court at paragraph 677 of the judgment under appeal but also with the sufficiency of its reasoning. The fact that AZ was entitled to request the withdrawal of its marketing authorisations for Losec capsules pursuant to Directive 65/65 in no way causes that conduct to escape the prohibition laid down in Article 102 TFEU. As the Commission pointed out in its pleadings, the illegality of abusive conduct under Article 102 TFEU is unrelated to the compliance or non-compliance of that conduct with other legal regimes.

80. It must be noted that the contested decision and the judgment under appeal concern active steps taken by AZ to deregister marketing authorisations. No parallel may thus be drawn, as claimed by the appellants, between the specific facts of the case at hand and the natural lapse of a marketing authorisation after a five-year period. The contested decision and the judgment under appeal do not concern a positive obligation on AZ to renew a lapsed or lapsing marketing authorisation. As regards the appellants’ claims concerning pharmacovigilance obligations, these should in my view be dismissed in the light of the clear findings of fact in paragraphs 688 to 694 of the judgment under appeal in which the General Court found that the pharmacovigilance obligations to which AZ was subject in Denmark, Norway and Sweden were not particularly burdensome and thus did not constitute an objective justification for the requests for deregistration of the marketing authorisations for Losec in those countries.

81. I therefore consider that the Court should dismiss the present ground of appeal as unfounded.

2. Second ground: conduct tending to restrict competition

a) Argument

82. The appellants maintain that the General Court misunderstood the requirements for a distortion of competition by considering that the mere exercise of a right lawfully afforded by EU law tends to restrict competition. The exercise of such a right could in principle amount to an abuse only in exceptional circumstances, namely where there is an elimination of effective competition. An analogy should be drawn with compulsory licensing cases, such as that dealt with in IMS Health. (51) That analogy is justified not only by virtue of the effective expropriation of the right to withdraw the marketing authorisation but also by virtue of the fact that the prohibition on withdrawal is a form of compulsory licensing.

83. Furthermore, contrary to the General Court’s assertion at paragraph 830 of the judgment under appeal, AZ still held exclusive rights in the clinical data, which remained confidential, even after the expiry of the exclusivity period conferred by Directive 65/65. That directive contains no obligation for companies supplying confidential information to share that information with their competitors, which is confirmed by the opinion delivered by the European Parliament in the preparatory stages of Council Directive 87/21/EEC of 22 December 1986 amending Directive 65/65. (52)

84. It follows that, contrary to the General Court’s assertions at paragraphs 817 and 829 of the judgment under appeal, it is insufficient in the present case to demonstrate merely that the withdrawal of marketing authorisation rendered competition ‘more difficult’, but it must also be demonstrated that the withdrawal has a disproportionate effect on competition.

85. According to the appellants, competition by generic companies was not eliminated. Indeed it was not substantially affected. The surrender of a marketing authorisation did not remove the right of generic companies already on the market from continuing to market their products. For generics that had not yet come to market, there were several routes to market other than the abridged procedure provided for in point 8(a)(iii) of paragraph 3 of Article 4 of Directive 65/65. There were realistic ‘alternative solutions’ even if they were ‘less advantageous’. (53)

86. The appellants also claim that the part of the contested decision relating to the second abuse and parallel imports, ought to have been annulled in so far as it applied to Sweden as well. Any impediment to competition in Sweden was caused by the incorrect application of Union law by the Swedish authority, as the Court of Justice has held that Articles 28 EC and 30 EC preclude the withdrawal of marketing authorisation for a pharmaceutical product from entailing in itself the withdrawal of authorisation of parallel imports in the absence of a risk to health. (54)

87. The Commission contends that, by their ‘compulsory licensing’ arguments, the appellants merely reiterate the arguments which they have already put forward at first instance, without giving reasons why the General Court’s examination of those arguments is flawed. This line of argument is therefore inadmissible.

88. The Commission also observes in this context that the existence of an original marketing authorisation merely allows the pharmaceutical authorities to refer – for the purposes of authorising another medicinal product under the abridged procedure – to a dossier already in their possession. Since the appellants have lost the exclusive right to use the information in the dossier on the original medicinal product, there is no question of granting a ‘compulsory licence’ to producers of generic medicinal products. Even on the assumption that the dossier contained ‘confidential commercial information’, the application of the abridged procedure would in no way interfere with that confidentiality, since the pharmaceutical authority would never make that information public or disclose it to the second applicant. The finding of the second abuse therefore does not have the consequence that competitors are given access to AZ’s data. It is clear that, in those circumstances, the ‘essential facilities’ case‑law is irrelevant.

b) Assessment

89. In the light of my findings at points 79 and 80 above, I do not consider that the fact that the deregistration of a marketing authorisation may be permissible pursuant to Directive 65/65 removes that conduct from scrutiny pursuant to Article 102 TFEU. Moreover, the right to withdraw a marketing authorisation is not in any manner akin to a property right, but merely constitutes a course of action which is available to undertakings pursuant to the terms of Directive 65/65. The application of Article 102 TFEU does not, in my view, constitute an effective expropriation of the right to withdraw a marketing authorisation, as claimed by the appellants. Thus a requirement of the elimination of effective competition, as in compulsory licence cases, should not be applied in this case.

90. The appellants have also based their claims on the requirement of an elimination of effective competition on the premiss that AZ enjoyed proprietary rights in its clinical data. The appellants also refer to the confidential nature of the information. In my view, that premiss is unfounded.

91. It is clear from paragraphs 668 and 680 of the judgment under appeal, which have not been challenged by the appellants, that after the expiry of a period of 6 or 10 years which starts to run from the grant of the first marketing authorisation, Directive 65/65 no longer confers on the owner of an original proprietary medicinal product the exclusive right to make use of the results of the pharmacological and toxicological tests and clinical trials placed in the file. On the contrary, that information may be taken into account by the national authorities for the purpose of granting marketing authorisations for essentially similar products under the abridged procedure provided for in point 8(a)(iii) of paragraph 3 of Article 4. In my view, the General Court correctly found at paragraph 681 of the judgment under appeal that any right enjoyed by AZ in the information in question was limited by the aforementioned provision at the relevant point in time.

92. Thus, despite the fact that the confidential information in question was not made available directly to other companies, Directive 65/65, as the appellants themselves indicated in their application before the General Court, (55) ‘created an exception to AZ’s confidentiality in so far as it excused a subsequent applicant, in specified conditions, from being required to provide its own data package’.

93. In the light of the foregoing, I consider that the General Court did not, as claimed by the appellants, err in law at paragraph 830 of the judgment under appeal by stating that ‘AZ no longer had the exclusive right to use the results of the pharmacological and toxicological tests and clinical trials’ as AZ could not prevent the national authorities from relying on the data in question in the abridged procedure. (56) I therefore consider that the appellants have not demonstrated that any of AZ’s proprietary rights were expropriated or that a compulsory licence has been granted to AZ’s competitors (57) due to the application of Article 102 TFEU in the contested decision.

94. In addition, I consider that the IMS Health case-law (58) is wholly inapplicable as the present case does not concern, inter alia, a refusal by a dominant undertaking to provide access to or licence information which is indispensable in order to allow a potential competitor to have access to the market in which the undertaking which owns the right occupies a dominant position. It is clear that the contested decision did not oblige AZ to transfer an asset or enter into a contract with persons with whom it has not chosen to contract. (59)

95. The extremely rigorous standards imposed in essential facility cases, which are exceptional in nature and which thus require, inter alia, a demonstration of the elimination of competition, (60) cannot be extrapolated to the wholly unrelated circumstances and facts of the present case.

96. The appellants have also in their pleadings recited evidence which is aimed at demonstrating that in January to February 2003 four generic companies launched their generic omeprazole capsules in Sweden. In addition, AZ submits evidence that generic companies could easily have obtained authorisation for a generic capsule version using the published literature procedure. Given that appeals are confined to questions of law, the Court of Justice may not carry out a reassessment of the facts, in the absence of a claim that the General Court has distorted the clear sense of the evidence. The appellants have not however claimed that the evidence in question was distorted. In my view, the present ground of appeal, to the extent that it seeks a reassessment of the facts in question, is therefore inadmissible.

97. In my view, the General Court did not err in law in finding that the conduct (the deregistration of the marketing authorisations) had the requisite anti-competitive effect pursuant to Article 102 TFEU in relation to the marketing of generic products in Denmark, Norway and Sweden. The General Court found, at paragraph 833 of the judgment under appeal, that the published literature procedure or the hybrid procedure requires conditions to be satisfied, such as the submission of additional data, which go beyond those required by the abridged procedure referred to in point 8(a)(iii) of paragraph 3 of Article 4 of Directive 65/65. Those other procedures were found, as a matter of fact by the General Court, to be more burdensome for manufacturers of generic products and necessarily take more time than the abridged procedure. The deregistration of the marketing authorisations thus enabled AZ to delay, at least temporarily, the significant competitive pressure that generic products exerted on it. The General Court found, in view of the sales volumes at stake, that any delay in the entry of generic products onto the market was worthwhile for AZ. (61) Contrary to the claims of the appellants, I consider that the delay in question is material and is sufficient for the withdrawal of the marketing authorisation to hinder the maintenance of the degree of competition still existing in the market or the growth of that competition.

98. As regards the correct test to be applied to parallel imports in so far as Sweden is concerned, it is clear from paragraph 862 of the judgment under appeal that the General Court found as a matter of fact that the Swedish Medical Products Agency (SMPA) considered that parallel import licences could be granted only if valid marketing authorisations were in place (62) and that that agency withdrew the parallel import licences as a result of the deregistration of the Losec capsule marketing authorisation. The General Court therefore found that the deregistration of the marketing authorisations was such as to impede parallel imports in Sweden.

99. The fact that the practice of the Swedish authorities was contrary to EU law, as claimed by the appellants and indeed as clarified by the Court in later judgments, (63) is not in my view such as to negate the fact that at the time of the deregistration of the relevant marketing authorisations by AZ it was plausible, in the light of documentary evidence of the practice of those authorities, that the deregistration would have the effect of impeding parallel trade in Sweden.

100. I therefore consider that the Court should dismiss the present ground of appeal as in part inadmissible and in part unfounded.

D – Fourth heading: fine

1. Argument

101. In this plea, which is divided into two parts, the appellants claim that the amount of the fine imposed on them is excessive.

102. The appellants claim firstly that the General Court ought to have reduced the amount of the fine on the ground that the abuses were novel. In the present case, the competition rules relating to the abuses had never been established before, which, in accordance with the judgment in AKZO v Commission, (64) justifies the imposition of a symbolic fine. For the reasons set out in the context of their plea concerning the first abuse and the absence of failure to compete on the merits, (65) the appellants dispute the General Court’s analysis, according to which the practices constituting the first abuse were manifestly contrary to competition on the merits, so that a reduction of the fine to take account of their novelty was excluded. The appellants consider that Nederlandsche Banden-Industrie-Michelin v Commission (66) on which the General Court based its analysis is inapplicable, as it relates to a completely different scenario. As regards the second abuse, the appellants claim that it is novel to describe the exercise of an EU right as abusive and moreover the fact that AZ’s request to withdraw its marketing authorisation was permitted under EU pharmaceutical law ought to be regarded as a mitigating circumstance that would justify a reduction of the fine.

103. In the context of the second part of this plea, the appellants maintain that the absence of anti-competitive effects is a factor that the General Court ought to take into account when reviewing the amount of a fine. They rely in that regard on T-Mobile Netherlands and Others (67) and ARBED v Commission. (68) Thus, as regards the first abuse, there were no anti-competitive effects in Denmark and the United Kingdom because SPCs were not granted. In Germany, while an SPC was granted, it was revoked so long before it entered into force that it could not therefore have affected competition. Moreover, there is no evidence that competition was actually restricted in Norway, the Netherlands and Belgium. As regards the second abuse, there is little concrete evidence that it produced any restrictive effects.

104. The Commission contends that this plea is inadmissible, since its purpose is to secure a general re-examination of the fines. Thus, it is not for this Court, in appeal proceedings, to substitute, on grounds of fairness, its own assessment for that of the General Court exercising its unlimited jurisdiction to rule on the amount of fines imposed on undertakings for infringements of competition law. In addition, the General Court correctly examined all the elements relevant to the calculation of the fine, including the alleged novelty of the abuses and the alleged absence of effects.

2. Assessment

105. As regards the question of inadmissibility raised by the Commission, it is settled case‑law that it is not for the Court of Justice, when ruling on questions of law in the context of an appeal, to substitute, on grounds of fairness, its own assessment for that of the General Court exercising its unlimited jurisdiction to rule on the amount of fines imposed on undertakings for infringements of EU law. (69) I consider that the present ground of appeal is not inadmissible as the appellants are not, as claimed by the Commission, merely seeking a general re-examination of the fines imposed. Rather the appellants claim that the General Court failed to assess in a legally correct manner the novelty of the infringements in question and the effects of those infringements for the purpose of calculating the fines. The present ground of appeal is thus, in my view, admissible.

106. On the question of novelty, it follows from paragraph 901 of the judgment under appeal and the reference made by the General Court at paragraph 903 of the judgment under appeal to paragraph 908 of the contested decision that the General Court and indeed the Commission considered that the abuses in question were novel.

107. It is nonetheless clear from those provisions that the General Court considered that the abuses were novel as regards the means used (70) and in that specific and limited regard were not clear cut.

108. The appellants’ claim that the novelty of the abuses warrants the imposition of a symbolic fine should, in my view, be rejected. Such a claim wholly ignores the fact that while the means used were novel, as there was no Commission decision or judgment of the Court on conduct using those same methods, the actual substance of the abuses in question was not novel and clearly departed from competition on the merits. (71) I consider that the General Court correctly found, on examination of the actual substance of the abuses in question, (72) that those abuses were serious infringements. In Deutsche Telekom v Commission, the Court held that, in relation to the question whether the infringements were committed intentionally or negligently and are, therefore, liable to be punished by a fine, it follows from the case-law of the Court that that condition is satisfied where the undertaking concerned cannot be unaware of the anti-competitive nature of its conduct, whether or not it is aware that it is infringing the competition rules of the Treaty. (73) I consider that the General Court correctly referred at paragraph 901 of the judgment under appeal to paragraph 107 of the Michelin I judgment (74) and found that AZ could not be exonerated from the fines. AZ ought to have expected that the abuses in question would fall within the sphere of application of Article 102 TFEU, even though conduct which used the same means or methods (75) had not been examined by the Commission or the Court. Moreover, the appellants’ claim should be rejected for policy reasons. Such an approach, which would privilege form rather than substance, would, in my view, undermine the deterrent role of fines for infringements of competition law.

109. As regards the appellants’ claim with respect to mitigating circumstances and the fact that AZ’s deregistration of marketing authorisation was permitted under Directive 65/65, I consider that the General Court correctly found at paragraph 914 of the judgment under appeal that the appellants reiterate once more the arguments taken into consideration at the stage of examining the abuse of a dominant position or assessing the gravity of the infringement. Moreover, no parallel can be drawn between the circumstances in Case T‑271/03 Deutsche Telekom v Commission (76) which led to a 10% reduction in the fine and the fact that Directive 65/65 does not prevent the deregistration of marketing authorisations. In the Deutsche Telekom v Commission case, the General Court held that the Commission had correctly used its margin of appreciation in fixing fines by considering that the repeated, active and specific intervention of a national regulator in the fixing of Deutsche Telekom’s prices in the telecommunications sector and the examination by that regulator of whether Deutsche Telekom’s prices led to margin squeeze warranted a 10% reduction in the fine. (77)

110. On the claim that the General Court failed to reduce the fine on account of minimal effects, I consider that the General Court found at paragraph 902 of the judgment under appeal that the practices relating to the first and second abuses were highly anti-competitive in that they were capable of having a significant effect on competition. I consider therefore that the General Court rightly found at paragraphs 902 and 911 of the judgment under appeal that factors relating to the object of a course of conduct may be more significant for the purposes of setting the amount of the fine than those relating to its effects. (78) In addition, it is clear from the file before the Court that the fact that the actual effects of the first abuse were limited for example in Denmark and the United Kingdom was due to the intervention of third parties. I consider that it would be inordinate if the appellants were to derive a benefit from that intervention. Moreover, the deterrent role of Article 102 TFEU would be greatly undermined if such an approach were adopted. (79)

111. I therefore consider that the Court should dismiss the present ground of appeal as unfounded.

V – EFPIA’s cross-appeal

112. EFPIA put forward two grounds in support of its cross-appeal relating to the existence of a dominant position. EFPIA claims that the General Court erred in law, firstly, in failing to properly consider the role of the State as a monopsonist buyer and, secondly, in finding that AZ’s intellectual property rights, first-mover status and its financial strength constituted evidence of AZ’s dominance.

113. Prior to examining in detail and on an individual basis these two grounds of appeal, I would note as a preliminary matter that, in accordance with the case‑law of the Court, although the importance of market shares may vary from one market to another, the possession over time of a very large market share is in itself, save in exceptional circumstances, evidence of the existence of a dominant position. (80) In addition, a market share of between 70% and 80% is in itself a clear indication of the existence of a dominant position. (81)

114. It is evident from paragraphs 245 to 254 of the judgment under appeal that the General Court found that the Commission’s finding of dominance rested largely on AZ’s generally very large market share, which was out of all comparison to those of its competitors, throughout the entire relevant period in all the countries concerned and which thus ensured that AZ was always the leading player on the PPI market. (82) The General Court also stated at paragraph 244 of the judgment under appeal that the Commission rightly did not base its finding of AZ’s dominance solely on market shares but also examined various other factors. The other factors taken into account in the contested decision and upheld by the General Court in the judgment under appeal included inter alia price levels charged for Losec, the existence and use of intellectual property rights, AZ’s first-mover status and AZ’s financial strength.

115. Given the case-law on the probative value of high market shares indicated at point 113 above, I consider that EFPIA’s grounds of appeal concerning the role of the State as a monopsonist buyer and AZ’s intellectual property rights, first-mover status and its financial strength, even if upheld, will be ineffective unless they call into question the soundness of the overall finding of dominance, by the Commission as confirmed by the General Court, which is largely based on market share.

116. Given that I consider that EFPIA’s two grounds of appeal should be dismissed, it is not necessary in this instance to examine the effectiveness of those two grounds with regard to the overall findings of dominance.

A – Error of law in respect of the role of the State – monopsony power

1. Argument

117. EFPIA considers that the General Court erred in law by failing to consider whether AZ’s high market share allowed it to act independently of its competitors and customers or, rather, whether the role of the State as a monopsonist buyer of prescription medicines and simultaneously as price regulator excluded or at least mitigated AZ’s alleged market power.

118. The General Court merely confirmed, at paragraph 257 of the judgment under appeal, the Commission’s findings that, first, pharmaceutical undertakings which offer for the first time on the market products with a high added therapeutic value as a result of their innovativeness are able to extract from public authorities higher prices or reimbursement levels than those of existing products and, second, pharmaceutical undertakings have bargaining power because price and reimbursement levels are set by public authorities in dialogue with those undertakings. In fact, neither of those findings is sufficient to support the allegation that AZ was able to act independently in circumstances in which the market was heavily regulated in terms of pricing and there was fierce competition in terms of innovation. Nor did the General Court consider the extent to which the pharmaceutical undertakings’ bargaining power gives them leverage over the State’s bargaining power.

119. It follows, moreover, from the General Court’s finding at paragraphs 191 and 262 of the judgment under appeal that, first, the sensitivity of doctors and patients to price differences is limited owing to the importance of the role played by therapeutic efficacy and, second, the costs of medicines are fully or largely covered by social security systems, that price will have a limited impact on the number of Losec prescriptions and hence on AZ’s market share. Contrary to the General Court’s finding at paragraph 261 of that judgment, therefore, no meaningful conclusion with respect to market power can be derived from the fact that AZ was able to maintain higher shares than its competitors while charging higher prices.

120. The Commission contends that this plea is inadmissible, since EFPIA merely requests the Court to reassess the findings of fact made by the General Court. In addition, the arguments put forward in the context of this plea, which have already been correctly examined by the General Court at paragraphs 258 to 268 of the judgment under appeal, are unfounded and constitute an attempt to deny even the possibility of the existence of a dominant position on the prescription medicines markets.

2. Assessment

121. As regards the question of inadmissibility raised by the Commission, I consider that EFPIA is not contesting the facts found by the General Court but rather the legal inferences to be drawn from those facts and specifically whether certain facts support or not a finding of dominance on the part of AZ. The present ground of appeal is thus, in my view, admissible.

122. On the substance of the present ground of appeal, I would note that it is settled case‑law that the dominant position referred to in Article 102 TFEU relates to a position of economic strength enjoyed by an undertaking which enables it to prevent effective competition being maintained on the relevant market by affording it the power to behave to an appreciable extent independently of its competitors, its customers and ultimately of consumers. (83)

123. It is uncontested by EFPIA that AZ was able to maintain a much higher market share than those of its competitors while charging prices higher than those charged for other PPIs. (84) EFPIA however asserts that due to inelastic demand, price will have a limited impact on demand and thus market share. In my view, that assertion is wholly vague and abstract and fails to demonstrate that the General Court erred at paragraph 262 of the judgment under appeal in finding that the health systems tend to reinforce the market power of pharmaceutical companies due to inelastic demand. In contrast, the General Court tailored its analysis and findings of inelasticity to the specificities of the particular situation of omeprazole and stated that where a pharmaceutical company is first to market an innovative product it is able to obtain a higher price from public authorities than similar products with only limited added therapeutic value. (85) In addition, EFPIA has not challenged the General Court’s finding that public authorities were making efforts to reduce health expenditure in order to compensate for the limited sensitivity to price of prescribing doctors and patients. (86) It follows that as those authorities were sensitive to price, the General Court did not err in stating that price could be a relevant criterion in the assessment of market power in certain circumstances. (87)

124. Moreover, contrary to EFPIA’s claims, the General Court examined in great detail the role of the State as a monopsony buyer in the specific context of the PPI market and in particular AZ’s product omeprazole. (88) The General Court correctly found, in my view, that the bargaining power of pharmaceutical undertakings varies according to the added therapeutic value of their products in comparison with pre-existing products. In that regard, national authorities which set reimbursement levels or prices of medicines on account of their public interest mission have more limited bargaining power in relation to products which contribute significantly to the improvement of public health. The General Court on the specificities of the case at hand found that given that AZ was the first undertaking to offer a PPI (89) whose therapeutic value was incontestably much higher than that of the existing products on the market, AZ was able to obtain a higher price from public authorities, and this despite the latter’s sensitivity to price. (90) By contrast, pharmaceutical undertakings marketing other PPIs could not obtain such prices since such products offered only limited added therapeutic value. (91) In my view, the fact that pharmaceutical companies have an interest in obtaining price and reimbursement approval as quickly as possible does not negate the fact that in certain specific circumstances, such as the case of omeprazole as outlined above, a pharmaceutical company may enjoy bargaining power in price negotiations with the State. I therefore consider, contrary to EFPIA’s claims, that the General Court did consider the extent to which pharmaceutical undertakings’ bargaining power gives them leverage over the State’s bargaining power.

125. EFPIA’s claim that the General Court failed to take into account the fact that AZ faced fierce competition in terms of innovation is merely asserted and is wholly unsupported by the file before the Court of Justice. In addition, EFPIA’s claim that the market is heavily regulated in terms of supply is again merely asserted. In any event, the fact that Losec was a prescription drug and that its supply was regulated was taken into account by the General Court in the context of price levels. (92)

126. The General Court thus rightly considered that AZ’s higher prices was a relevant factor showing that AZ’s behaviour was not subject to an appreciable extent to competitive constraints.

127. I therefore consider that the Court should dismiss the present ground of appeal as unfounded.

B – Error of law in respect of AZ’s intellectual property rights, its first-mover status and its financial strength

1. Argument

128. EFPIA maintains that the General Court erred in law in considering that AZ’s intellectual property rights, its first-mover status and its financial strength constituted evidence of its dominant position. Those three characteristics are typically shared by many innovative companies that successfully engage in research for new products and do not allow a meaningful distinction to be drawn between dominant and non-dominant undertakings. The General Court thus misapplied the case-law of this Court, and in particular the judgments in RTE and ITP v Commission (‘Magill’) (93) and in IMS Health, (94) which confirmed that the mere possession of intellectual property rights is not sufficient to establish the existence of a dominant position. What led the Court of Justice to conclude that there was dominance in Magill was the existence of elements on the basis of which the Court regarded Magill’s programme listings as effectively amounting to an essential facility. (95) The judgment under appeal has significant implications in that it finds, in reality, that a company which is the first to enter the market with an innovative product must refrain from acquiring a comprehensive portfolio of intellectual property rights or from enforcing those rights if it is not to risk being regarded as dominant. EFPIA further criticises the General Court for having failed to confirm that AZ’s intellectual property rights enabled it to act independently on the market.

129. The Commission contends that this plea is based on a recurring confusion between the assessment of dominance and the qualification of certain behaviour as abusive. The recognition of the importance of patents as a factor to be taken into account for the purpose of determining whether an undertaking has a dominant position is as old as EU competition law itself and was already recognised in the judgment in Istituto Chemioterapico Italiano and Commercial Solvents v Commission. (96) In addition, the existence of a dominant position on the part of the holder of a patent can be established only after a specific analysis of the situation of the market, which in the present case is explained in teens of recitals to the decision in issue and was confirmed by the General Court. Furthermore, the fact that a patent is not automatically synonymous with a dominant position does not alter the fact that it may constitute a serious obstacle to the entry of competitors on the market or to their expansion.

2. Assessment

130. In my view, EFPIA has merely asserted but failed to indicate how the General Court erred in law by taking the issues of first-mover strength and financial status into account in its overall assessment of AZ’s dominance. I therefore consider EFPIA’s claims on those matters to be inadmissible.

131. As regards intellectual property rights, in my view the possession of such exclusive rights does not necessarily imply that an undertaking holds a dominant position in a relevant market as there may be substitutes for the products or services in question. Thus, as indicated by the Commission in its pleadings, there is no presumption that the possession of such rights gives rise to market power. Indeed, many products which are subject to patent, copyright, trade mark and design protection are commercially unsuccessful. However, in certain concrete cases the possession of such rights may be sufficient in itself to confer a dominant position on an undertaking. Alternatively, the possession of such rights may, in conjunction with other factors, lead to a finding of dominance. Any assessment of dominance must therefore be done on a case-by-case basis and intellectual property rights should be largely treated as similar to other property rights, with due account being taken of the specificities of intellectual property rights.

132. EFPIA’s submission that intellectual property rights may only confer dominance where such rights constitute an essential facility is entirely unsupported by the case‑law invoked by that party (97) which concern the possible abuse of a dominant position by refusing to license such rights. Moreover, while holding an intellectual property right which is indispensable to compete in a relevant market will undoubtedly confer dominance on an undertaking in respect of that market due to barriers to entry, indispensability is not the sine qua non of a finding of dominance in such situations. (98)

133. The finding that an undertaking has a dominant position is not in itself a ground of criticism of the undertaking concerned. (99) It is only the abuse of that position which is subject to sanctions pursuant to Article 102 TFEU. Consequently, the fact that the General Court confirmed that the Commission may take into account, inter alia, AZ’s intellectual property rights, its first-mover status and its financial strength as indicia of a dominant position, by no means chills legitimate competition on the merits either by AZ itself or indeed any pharmaceutical company.

134. In the light of the uncontested finding of the General Court at paragraph 271 of the judgment under appeal that as the first PPI to be introduced on the market, Losec enjoyed particularly strong patent protection, on the basis of which AZ brought a series of legal actions which enabled it to impose significant constraints on its competitors (100) and to dictate to a large extent market-entry terms to them, I consider that the General Court did not err in law in finding at paragraph 272 of the judgment under appeal that the patent protection enjoyed by Losec enabled AZ to exert significant pressure on its competitors and was therefore a relevant indicator, in itself, (101) of its dominant position. Thus the terms ‘in itself’, objected to by EFPIA, must be read in context and in the light of the specific and clear reasoning of the General Court. In any event, given the fact that the General Court examined other factors, not least the extremely high market shares held by AZ in the relevant markets, the judgment under appeal clearly requires more than ‘mere possession’ of intellectual rights for a finding of dominance as claimed by EFPIA.

135. I therefore consider that the Court should dismiss the present ground of appeal as partly inadmissible and partly unfounded.

VI – The Commission’s cross-appeal

A – Argument

136. The Commission’s cross-appeal is directed against the General Court’s assessment, at paragraphs 840 to 861 of the judgment under appeal, on the basis of which that court held that the Commission demonstrated for Sweden, but not for Denmark and Norway, that the deregistration of the marketing authorisation of Losec capsules was capable of excluding parallel imports of those products and therefore likely to restrict competition.

137. The Commission submits that the General Court misapplied the rules on the burden and standard of proof by requiring that the Commission demonstrate that the national authorities were inclined to withdraw, or indeed did habitually withdraw, parallel import licences following deregistration. In reality, the General Court focused on the actual effects of the practice or, rather, on a particular concept of the ‘effects’, instead of applying the legal test which it had set for itself. The General Court’s reasoning is contradictory and has paradoxical consequences. Thus, Denmark was specifically the only country in which AZ’s deregistration strategy proved to be wholly effective, and yet the General Court found that there was no abuse in that country, which illustrates that the test of causality applied was too narrow. Thus, the mere fact that other factors might have contributed to the exclusion of all parallel trade is no justification for the conclusion that deregistration was not also apt to have that effect. Furthermore, in so far as the legal context in the three countries was exactly the same, it is contradictory to arrive at different results. In addition, the General Court failed, at paragraph 850 of the judgment under appeal, to assess crucial evidence and at paragraphs 839 and 846 of that judgment made a manifestly flawed application of the presumption of innocence.

138. In addition, the General Court’s finding, at paragraphs 848 and 849 of the judgment under appeal, that the AZ documents referred to by the Commission reflected only the personal opinion, or the expectations, of AZ employees and could at the very most show that AZ had the intention of excluding parallel imports by deregistering the Losec capsules authorisation, constitutes a manifest distortion of the clear sense of the evidence. Those documents show that AZ had carried out its own research into the practices of the national authorities and had concluded that its strategy was likely to succeed in the three countries concerned. In those circumstances the General Court was wrong to require that the Commission investigate, ex post facto, years after the events, what an authority’s attitude might have been, when AZ’s research into the authorities’ attitude was particularly reliable. Nor is the Commission to be criticised for not having ascertained a practice that did not exist, owing to the fact that the ‘switch and deregistration’ operation was unprecedented. The General Court was wrong, moreover, to reject at paragraph 849 the relevance of evidence of intention, contrary to the test which it had set for itself and to the case-law of this Court.

B – Assessment

139. The appellants claimed at first instance that the decline in parallel imports of Losec capsules in Sweden, Denmark and Norway was due to the success of Losec MUPS rather than the deregistration of marketing authorisations. The Commission considered, however, that there was a causal link between the elimination of parallel trade and deregistration. (102)

140. The General Court correctly found that the burden of proof lay on the Commission to establish the necessary anti-competitive effects of the practice of deregistration on parallel trade. It is thus clear that, contrary to its assertions, the Commission did not have to establish an actual causal link between the deregistration of Losec capsules marketing authorisation and an impediment to parallel trade, but merely that ‘national authorities were liable to withdraw or did usually withdraw parallel import licences following the deregistration ...’. (103)

141. The General Court found that the Commission had not adduced tangible evidence that, in the wake of deregistration of the marketing authorisations for Losec capsules in Denmark and Norway, the national authorities were liable to withdraw or did usually withdraw parallel import licences. That court found that the Commission, in the case of Denmark and Norway, had failed to establish the anti-competitive effect of deregistration, as it had relied on evidence which merely reflected AZ’s expectations of the likely reactions of the relevant authorities in those countries to deregistration. However, in the case of Sweden, the contested decision was upheld on the matter as that decision referred to documentary evidence from the SMPA, which had been obtained from the latter by AZ, and which indicated that that agency considered that parallel import licences could be granted only if valid marketing authorisations were in place. (104) The Commission itself admits that no such tangible evidence was available in respect of Norway and Denmark.

142. It is clear from the judgment under appeal that AZ had carried out its own research into the practices of the national authorities and had concluded that its strategy was likely to succeed in the three countries concerned. (105) In my view, contrary to the claims of the Commission, the General Court did not misapply the rules on the burden and standard of proof and correctly dismissed evidence reflecting AZ’s own assessment of whether the Danish and Norwegian authorities were inclined to withdraw parallel import licences following deregistration of marketing authorisations. I consider that the General Court correctly found that AZ’s informed, but nonetheless subjective, expectations of the reaction of the Danish and Norwegian authorities to deregistration, based on the advice of in-house counsel, (106) constituted evidence of AZ’s anti-competitive intent, but were insufficient in themselves to satisfy the requirement of establishing anti-competitive effect in the absence of any tangible or objective evidence which would corroborate those personal opinions or expectations.

143. In my view the fact that AZ believed, on the basis of considerable research and expert advice, that its actions would have the desired anti-competitive effect is insufficient in itself, as it follows from the objective nature of the concept of abuse that the anti-competitive effects of a practice must be assessed on the basis of objective factors. Tangible evidence, over and above evidence of anti-competitive intent, is necessary in order to establish that the conduct objectively tends to restrict competition. As regards the Commission’s claim that such corroborative evidence is difficult to obtain after the events, that claim should be dismissed, in the light to the burden of proof borne by the Commission. I would also note that, in any event, the Commission did not submit any evidence or even assert in its pleadings that it had unsuccessfully attempted to investigate what the attitude of the relevant authorities in Denmark and Norway was to deregistration of marketing authorisations and licences for parallel imports.

144. In my view, the General Court’s assessment at paragraph 850 of the judgment under appeal is not erroneous. While that paragraph does not specifically mention paragraph 302 of the contested decision, but rather paragraph 311 which refers to paragraph 302 of that decision, it is clear that the latter paragraph merely establishes AZ’s personal expectations of a practice and thus its anti-competitive intent. In that regard, paragraph 302 of the contested decision refers to the Norwegian LPPS Strategy (107) document which outlines that it was expected ‘that parallel trade of Losec capsules will gradually cease …’ and will mimic the situation in Denmark following the introduction of Losec MUPS. I consider that proof of anti-competitive intent does not establish the necessary anti-competitive causal link between the deregistration of Losec capsule marketing authorisation and the exclusion of parallel imports. Contrary to the Commission’s assertions, the General Court did not require that the disappearance of parallel trade in Denmark was exclusively caused by deregistration as the General Court found at paragraph 850 of the judgment under appeal that ‘no link is established between the deregistration of the Losec capsule marketing authorisation and the exclusion of parallel imports’.

145. In addition, the fact that it subsequently transpires that parallel trade in Losec capsules was affected in Denmark and was not affected in Sweden, as claimed by the Commission, is not paradoxical. In the first case, proof of the necessary causal link was absent from the contested decision, a flaw that cannot be remedied by subsequent evidence produced after the adoption of that decision. The contested decision must be assessed on its contents. In the second case, the fact that a particular anti-competitive practice was unsuccessful does not negate its potential/plausible effects at the time of implementation of that practice.

146. I therefore consider that the Court should dismiss the Commission’s cross-appeal as unfounded.

VII – Costs

147. Under Article 69(2) of the Rules of Procedure, which, under Article 118 of those rules, applies to appeals, the unsuccessful party is to be ordered to pay the costs if they have been applied for in the successful party’s pleadings.

148. Since the appellants have been unsuccessful with their appeal, they should be ordered to pay the costs of that appeal, in accordance with the form of order sought by the Commission.

149. Since EFPIA has been unsuccessful with its cross-appeal, it should be ordered to pay the costs of that appeal, in accordance with the form of order sought by the Commission. EFPIA should bear its own costs in connection with its intervention in support of the appeal brought by the appellants. As the Commission did not request that EFPIA be ordered to pays the costs of the Commission incurred in connection with EFPIA’s intervention, EFPIA shall not bear those costs.

150. Since the Commission has been unsuccessful with its cross-appeal, in view of the particular circumstances of the case where the appellants did not lodge written pleadings in respect of that cross-appeal, the Commission must be ordered to bear its own costs.

VIII – Conclusion

151. For the foregoing reasons, I suggest to the Court that it should decide as follows:

(i) dismiss the appeal brought by AstraZeneca AB and AstraZeneca plc;

(ii) dismiss the cross-appeal brought by the European Federation of Pharmaceutical Industries and Associations (EFPIA);

(iii) dismiss the cross-appeal brought by the Commission;

(iv) order AstraZeneca AB and AstraZeneca plc to pay in respect of their appeal their own costs and the costs of the Commission;

(v) order EFPIA to bear in respect of its cross-appeal its own costs and the costs of the Commission;

(vi) order EFPIA to bear in respect of the appeal brought by AstraZeneca AB and AstraZeneca plc its own costs;

(vii) order the Commission to bear in respect of its cross-appeal its own costs.

1 – Original language: English.

2 – [2010] ECR II‑2805 (‘the judgment under appeal’).

3 – Decision of 15 June 2005 relating to a proceeding under Article 82 [EC] and Article 54 of the European Economic Area (EEA) Agreement (Case COMP/A 37.507/F3 – AstraZeneca) (‘the contested decision’).

4 – Council Regulation (EEC) No 1768/92 of 18 June 1992 concerning the creation of a supplementary protection certificate for medicinal products (OJ 1992 L 182, p. 1) provides for the creation of a supplementary protection certificate, the purpose of which is to extend the duration of the exclusive right guaranteed by a patent and, therefore, to confer an additional protection period. The SPC is designed to compensate for the reduction in the period of effective protection conferred by the patent, corresponding to the period between the filing of a patent application in respect of a medicinal product and the granting of authorisation to place that product on the market. The aforementioned regulation will hereinafter be referred to as the ‘SPC Regulation’.

5 – OJ, English Special Edition 1965‑66(I), p. 24.

6 – See for example paragraphs 68 and 69 of the judgment under appeal.

7 – It is settled case-law that, for the purposes of applying Article 102 TFEU, the market for the product or service in question comprises all the products or services which in view of their characteristics are particularly suited to satisfy constant needs and are only to a limited extent interchangeable with other products or services; see Case C‑7/97 Bronner [1998] ECR I‑7791, paragraph 33 and the case‑law cited therein.

8 – See paragraph 84 of the judgment under appeal.

9 – See paragraphs 381 and 612 of the judgment under appeal.

10 – See paragraph 613 of the judgment under appeal.

11 – A report prepared by IMS Health; see paragraph 37 of the judgment under appeal.

12 – See Case C‑185/95 P Baustahlgewebe v Commission [1998] ECR I‑8417, paragraph 23, and Case C‑551/03 P General Motors v Commission [2006] ECR I‑3173, paragraph 51.

13 – See paragraphs 83 to 107 of the judgment under appeal, in particular paragraphs 84 and 101.

14 – See also paragraphs 95 and 96 of the judgment under appeal.

15 – I consider that much of the evidence presented by the appellants in the context of this part of the first ground of appeal is inadmissible as it merely seeks a reassessment of the findings of fact of the General Court. See point 23 above.

16 – See paragraph 94 of the judgment under appeal.

17 – See paragraph 94 of the judgment under appeal.

18 – See paragraph 98 of the judgment under appeal.

19 – See paragraph 102 of the judgment under appeal.

20 – See paragraph 278 of the judgment under appeal.

21 – As the quantification of cost-effectiveness was likely to be particularly complex and uncertain.

22 – Due to ‘(i) the limited sensitivity of doctors and patients to price differences on account of the importance of the role played by therapeutic efficacy in the choice of what to prescribe, and (ii) of the regulatory systems in force in the relevant States, which were not designed in such a way as to enable the prices of H2 blockers to exert downward pressure on sales or prices of PPIs’. See the summary at paragraph 191 of the judgment under appeal.

23 – Case C‑127/00 [2003] ECR I‑14781.

24 – Case T‑111/96 [1998] ECR II‑2937, paragraphs 54 to 60.

25 – Cited in footnote 24.

26 – See paragraph 356 of the judgment under appeal.

27 – See paragraph 493 of the judgment under appeal.

28 – See paragraph 495 of the judgment under appeal.

29 – See, for example, paragraphs 491, 495 and 497 of the judgment under appeal.

30 – See paragraphs 573, 588 and 599 of the judgment under appeal.

31 – Case C‑280/08 P Deutsche Telekom v Commission [2010] ECR I‑9555, paragraph 174 and the case-law cited.

32 – I consider that the General Court correctly stated that proof of intention to resort to practices falling outside the scope of competition on the merits may, however, be relevant where it supports the conclusion based on objective factors that an undertaking abused its dominant position. See, to that effect, paragraph 359 of the judgment under appeal.

33 – See Joined Cases C‑204/00 P, C‑205/00 P, C‑211/00 P, C‑213/00 P, C‑217/00 P and C‑219/00 P Aalborg Portland and Others v Commission [2004] ECR I‑123, paragraph 200.

34 – Regulation of 16 December 2002 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty, OJ 2003 L 1, p. 1. See also Article 15(4) of Council Regulation No 17 of 6 February 1962: First Regulation implementing Articles 85 and 86 of the Treaty, OJ, English Special Edition 1959-1962(I), p. 87.

35 – Cited in footnote 23.

36 – Case cited in footnote 24.

37 – In that case the Commission stated that, in order to be able to determine the cases in which legal proceedings are an abuse, two cumulative criteria must be met. It is necessary that the action firstly, cannot reasonably be considered as an attempt to establish the rights of the undertaking concerned and can therefore only serve to harass the opposite party and secondly, it is conceived in the framework of a plan whose goal is to eliminate competition. It must, however, be noted that the General Court examined whether the Commission correctly applied the two cumulative criteria and did not rule on the correctness of the criteria chosen by the Commission. See paragraph 58 of that judgment.

38 – Joined Cases C‑468/06 to C‑478/06 [2008] ECR I‑7139.

39 – See, to that effect, Case 85/76 Hoffmann-La Roche v Commission [1979] ECR 461, paragraph 91, and Case C‑62/86 AKZO v Commission [1991] ECR I‑3359, paragraph 69.

40 – Deutsche Telekom v Commission, cited in footnote 31, paragraph 250. In that case the Court found that the anti-competitive effect which the Commission is required to demonstrate, as regards pricing practices of a dominant undertaking resulting in a margin squeeze of its equally efficient competitors, relates to the possible barriers which the appellant’s pricing practices could have created for the growth of products on the retail market in end-user access services and, therefore, on the degree of competition in that market (emphasis added) (see paragraph 252). It was also found in that case that the particular pricing practices gave rise to actual exclusionary effects (see paragraph 259).

41 – I dislike the use of the term ‘likely effects’ in the case-law in this context. It brings to mind the tort law standard of ‘on the balance of probabilities’ and thus sets the evidentiary bar too high. On the other side of the spectrum, the term ‘capable’ may set the evidentiary bar too low, whereby any remote possibility of anti-competitive effects is sufficient to establish abuse.

42 – If a practice at the time of implementation is incapable of hindering competition, then that practice does not infringe Article 102 TFEU. See, to that effect, Deutsche Telekom v Commission, cited in footnote 31, paragraph 254.

43 – I consider that the General Court correctly found at paragraph 548 of the judgment under appeal in relation to the initial SPC application to the United Kingdom office (a country in which no SPC was granted to AZ) that ‘it is absolutely clear from all the documentary evidence submitted for the Court’s attention, … that the initial SPC application filed with the United Kingdom patent office was part of an overall strategy on SPC applications, designed to base those applications on the date of 21 March 1988 instead of on the date of 15 April 1987, which corresponded to the first marketing authorisation granted in the Community’.

44 – See, by analogy, Deutsche Telekom v Commission, cited in footnote 31, paragraph 254. In that regard, I also fully concur with the General Court’s finding at paragraph 379 of the judgment under appeal that ‘[t]he fact that AZ was no longer in a dominant position at the time when its abusive behaviour was able to produce its effects does not alter the legal classification to be attached to its acts, since those acts were committed at a time when AZ was under a special responsibility not to allow its behaviour to impair genuine undistorted competition on the common market’. I consider that the effects referred to by the General Court are actual effects which are clearly not required pursuant to the Court’s case-law. I consider that the Commission has correctly argued in its pleadings that the legality of an act must be assessed at the time it is implemented rather than at the time actual concrete effects arise.

45 – K-Lath Division of Tree Island Wire (USA) Iinc v Davis Wire Corporation and Others, 15 F.Supp. 2d 952 (C:D: Cal. 1998).

46 – Case C‑94/98 [1999] ECR I‑8789.

47 – Case C‑172/00 [2002] ECR I‑6891.

48 – See point 75 above.

49 – Cited in footnote 46.

50 – Cited in footnote 47.

51 – Case C‑418/01 [2004] ECR I‑5039.

52 – OJ 1987 L 15, p. 36.

53 – See IMS Health, cited in footnote 51, paragraph 22.

54 – Case C‑15/01 Paranova Läkemedel and Others [2003] ECR I‑4175, paragraphs 25 to 28 and 33, and Case C‑113/01 Paranova [2003] ECR I‑4243, paragraphs 26 to 29 and 34.

55 – See paragraph 492(b).

56 – The abridged procedure was however unavailable due to the positive actions of AZ in requesting deregistration of the marketing authorisations for Losec capsules in the relevant countries.

57 – Who are not granted direct access to the data in question pursuant to Directive 65/65.

58 – Or duty to deal or essential facility case-law.

59 – See Case T‑65/98 Van den Bergh Foods v Commission [2003] ECR II‑4653, paragraph 161, upheld by the Court of Justice in Case C‑552/03 P Unilever Bestfoods v Commission [2006] ECR I‑9091, paragraph 137.

60 – See IMS Health, cited in footnote 51, paragraph 52.

61 – See paragraph 834 of the judgment under appeal.

62 – It is clear from paragraph 315 of the contested decision that, as a result of replies to a questionnaire returned to AZ by the SMPA in 1997, there was documentary evidence about the possible/plausible effects of the deregistration of Losec capsules on parallel imports in Sweden.

63 – Paranova Läkemedel and Others, cited in footnote 54, paragraphs 25 to 28 and 33, and Paranova, cited in footnote 54, paragraphs 26 to 29 and 34.

64 – Cited in footnote 39.

65 – See points 41 to 43 above.

66 – Case 322/81 [1983] ECR 3461 (‘Michelin I’).

67 – Case C‑8/08 [2009] ECR I‑4529.

68 – Case T‑137/94 [1999] ECR II‑303.

69 – Case C‑219/95 P Ferriere Nord v Commission [1997] ECR I‑4411, paragraph 31, and Baustahlgewebe v Commission, cited in footnote 12, paragraph 129.

70 – See paragraph 908 of the contested decision which refers to the use of public procedures and regulation with an exclusionary intent.

71 – See points 47 et seq. and 77 et seq. above.

72 – Which consisted firstly, in misleading representations made deliberately in order to obtain exclusive rights to which AZ was not entitled or to which it was entitled for a shorter period and secondly, the deregistration of the marketing authorisations in order to create obstacles to the market entry of generic products in Denmark, Norway and Sweden and to parallel imports in Sweden, thus resulting in partitioning of the common market.

73 – Case cited in footnote 31, paragraph 124.

74 – Case cited in footnote 66.

75 – In the Michelin I case the Court referred to discount systems having the same features (cited in footnote 66).

76 – [2008] ECR II‑477, paragraphs 312 and 313.

77 – Upheld on appeal. See Case C‑280/08 P Deutsche Telekom v Commission, cited in footnote 31, paragraphs 279 and 286.

78 – It is clear from the context that the effects in question are actual effects.

79 – See, by analogy, Joined Cases C‑101/07 P and C‑110/07 P Coop de France Bétail et Viande v Commission [2008] ECR I‑10193, paragraphs 96 to 98.

80 – See Hoffmann-La Roche v Commission, cited in footnote 39, paragraph 41.

81 – Case T‑30/89 Hilti v Commission [1991] ECR II‑1439, paragraph 92.

82 – See paragraph 245 of the judgment under appeal. At paragraph 294 of the judgment under appeal the General Court concluded that the Commission did not commit a manifest error of assessment in finding that AZ held a dominant position within the meaning of Article 82 EC and Article 54 of the EEA Agreement on the PPI market in Germany from 1993 until the end of 1997, in Belgium from 1993 until the end of 2000, in Denmark from 1993 until the end of 1999, in Norway from 1994 until the end of 2000, in the Netherlands from 1993 until the end of 2000, in the United Kingdom from 1993 until the end of 1999, and in Sweden from 1993 until the end of 2000.

83 – Case 27/76 United Brands and United Brands Continentaal v Commission [1978] ECR 207, paragraph 65. Unlike a monopoly or quasi-monopoly situation, such a position does not preclude some competition but enables the undertaking which profits by it, if not to determine, at least to have an appreciable influence on the conditions under which that competition will develop, and in any case to act largely in disregard of it and without suffering any adverse effects as a result of its attitude. See Hoffmann-La Roche v Commission, cited in footnote 39, paragraph 39.

84 – See paragraph 261 of the judgment under appeal.

85 – See paragraphs 259 to 262 of the judgment under appeal.

86 – See paragraph 264 of the judgment under appeal.

87 – See paragraph 269 of the judgment under appeal.

88 – See paragraph 256 of the judgment under appeal.

89 – AZ was the first-mover on a market which it pioneered. See paragraph 260 of the judgment under appeal.

90 – See paragraphs 259 and 264 of the judgment under appeal.

91 – See paragraph 259 of the judgment under appeal.

92 – See, inter alia, paragraph 264 of the judgment under appeal.

93 – Joined Cases C‑241/91 P and C‑242/91 P [1995] ECR I‑743.

94 – Case cited in footnote 51.

95 – See paragraph 47 of Magill (cited in footnote 93).

96 – Joined Cases 6/73 and 7/73 [1974] ECR 223.

97 – Magill, cited in footnote 93; IMS Health, cited in footnote 51; and Case 238/87 Volvo [1988] ECR 6211.

98 – Indispensability, however, is of paramount importance in order to establish an abuse in such essential facility cases.

99 – See Michelin I, cited in footnote 66, paragraph 57, and Joined Cases C‑395/96 P and C‑396/96 P Compagnie maritime belge transports and Others v Commission [2000] ECR I‑1365, paragraph 37.

100 – Takeda, Byk Gulden and Eisai.

101 – Including, inter alia, extremely high market shares.

102 – See, paragraph 753 of the judgment under appeal.

103 – See, paragraph 846 of the judgment under appeal, see also paragraph 839.

104 – See paragraph 862 of the judgment under appeal which refers to paragraph 315 of the contested decision.

105 – See, for example, paragraphs 780 and 848.

106 – Which according to the Commission was based on thorough, exhaustive research.

107 – Losec Post Patent Strategy.