The EU Court of Justice holds that colluding to disseminate information constitute a restriction of competition “by object” (Roche / Novartis)
This article was first published on Concurrences – Antitrust Publications & Events.
On January 23rd, 2018, the ECJ delivered its preliminary ruling on the Roche/Novartis v. Commission case (C-179/16), concerning the compatibility of an agreement entered into between Roche and Novartis with Article 101 TFEU.
The controversy originated before of the ICA, which on 27 February 2014 issued two fines (approximately 92 million euros each) to Roche/Roche Italia and Novartis/Novartis Italia. In particular, the ICA found the agreement between the two companies incompatible with Article 101 TFEU given that, following the authority’s reasoning, the agreement was “designed to achieve an artificial differentiation between the medicinal products Avastin and Lucentis by manipulating the perception of the risks of using Avastin in the field of ophthalmology”.
Both Avastin and Lucentis were developed by Genentech, a company established and operating exclusively in the United States. Genentech allowed Roche, its parent company, to distribute Avastin outside the United States. At the same time, by way of a licensing agreement, Genentech allowed Novartis to do the same with Lucentis, a pharmaceutical product produced to cure eye diseases. In 2005, Avastin received a Marketing Authorization (MA) for the treatment of certain tumorous diseases. However, before Lucentis was placed on the market, several doctors worldwide began prescribing Avastin “off label” for the cure of eye diseases, i.e. in respect of indications not corresponding to those mentioned in the MA for the product.
The ICA held that Avastin, given its widespread off-label use in Italy, positioned itself as the main competitor for Lucentis in the ophthalmology field and this lead Roche and Novartis to enter into an arrangement aimed at disseminating information affecting the public perception of Avastin as a safe product to the end of decreasing its demand for the cure of eye diseases.
Following dismissal of the two undertakings’ appeals by the Regional Administrative Court of Lazio (or “TAR Lazio”), the case was appealed further before Council of State (which is the Italian supreme administrative Court), which stayed the proceeding and refer a set of questions to the ECJ for a preliminary ruling.
The first question referred to the ECJ related to whether the scope of a marketing authorization should be considered as binding for the definition of the relevant market, which is preliminary to properly assess whether or not a contested agreement is capable of affecting competition. In practice, the ECJ was requested whether Avastin, a pharmaceutical product whose MA was limited to the treatment of tumorous diseases, could be comprehended within the same relevant market of Lucentis, a medicinal product specifically envisaged to cure eye diseases.
To do so, the Court firstly clarified that the concept of relevant market entails a sufficient degree of interchangeability between the products concerned. Said degree of interchangeability should be assessed relying on both the objective characteristics of the products and the structure of supply and demand of the market. Secondly, the Court clarified how the legal or illegal nature of the products can affect the definition of the relevant market. In particular, the Court noted that “the fact that pharmaceutical products are manufactured or sold illegally prevents them, in principle, from being regarded as substitutable or interchangeable products, both on the supply side […] and on the demand side” (para. 52). However, the Court found that in the case at hand the marketing of the drug off-label was not clearly illegal as there was no EU legislation that prohibited the off-label prescription of Avastin and therefore there was not preclusion for the AGCM to include Avastin in the same relevant market as Lucentis.
More specifically, the ECJ’s stated that, “for the purpose of [application of Article 101 TFEU] a national competition authority may include in the relevant market, in addition to the medicinal products authorized for the treatment of the diseases concerned, another medicinal product whose MA does not cover that treatment but which is used for that purpose and is thus actually substitutable with the former”(para. 67).
As to second question, the ECJ was requested to give guidance on whether a restriction not expressly envisaged in the licensing agreement between Roche and Novartis (namely, that by which the parties agreed to disseminate information on the Avastin’s risks in treating eye diseases) could be considered “ancillary” (i.e. strictly related and necessary) to that agreement and thus fall outside the scope of the prohibition of Article 101(1) TFEU. In addressing this question the Court reaffirmed a cornerstone judicial elaboration on the concept of ancillarity in relation to Article 101 TFEU: if an operation/activity falls outside the prohibition envisaged in Article 101(1) TFEU because of its neutrality or positive effects on competition, then a restriction of the commercial autonomy of one of the participants to that operation/activity also falls outside the scope of the prohibition if that restriction is “objectively necessary” to the implementation of that operation “and is proportionate to the objectives” of the operation/activity (para. 69).
The ECJ further elaborated on how authorities and courts should address the relationship between a main operation that is not anti-competitive in nature and a restriction connected to that operation to assess if it is ancillary: “it is necessary to inquire whether that operation would be impossible to carry out in the absence of the restriction in question. The fact that that operation is simply more difficult to implement or even less profitable without the restriction concerned cannot be deemed to give that restriction the objective necessity required in order for it to be classified as ancillary. Such an interpretation would effectively extend that concept to restrictions which are not strictly indispensable to the implementation of the main operation. Such an outcome would undermine the effectiveness of the prohibition laid down in Article 101(1) TFEU”.
On a side note, one can argue that the problem with this test is that, as a matter of fact, it requires the judge or authority to substitute the undertaking in assessing how much reduction of expected profits or what degree of difficulties make an operation economically unsustainable and thus “impossible”. Indeed, the dividing line between objective and subjective impossibility may be extremely thin in this space.
With this said, when called upon to assess how these concepts were to be applied in relation to the restriction set forth by Roche and Novartis, the ECJ found that such criteria were not met. In particular, the ECJ held that the conduct consisting in the dissemination of “misleading” information on Avastin’s effects could not be qualified as a restriction of the commercial autonomy of one of the participants of the licensing agreement but was rather set forth to restrict the conduct of third parties. Furthermore, the ECJ also held that the such an agreement could not be deemed “objectively necessary” to pursue the objectives set out in the main operation.
In light of the above the ECJ concluded that “Article 101 (1) TFEU must be interpreted as meaning that an arrangement put in place between the parties to a licensing agreement regarding the exploitation of a medicinal product which, in order to reduce competitive pressure on the use of that product for the treatment of given diseases, is designed to restrict the conduct of third parties promoting the use of another medicinal product for the treatment of those diseases, does not fall outside the application of that provision on the ground that the arrangement is ancillary to that agreement” (par. 75).
With the third question, the ECJ was called to evaluate whether Article 101(1) should be interpreted as meaning that the arrangement devised by Roche and Novartis constituted a restriction of competition “by object” (i.e. which is prima facie presumed incompatible with Article 101 and thus prohibited). In answering this question, the ECJ preliminarily points out that, according to its case-law, the concept of restriction of competition by object should be interpreted strictly given that only specific types of agreements between parties can entail such a degree of harm to competition that it would exclude the necessity to evaluate their effects on the market. In particular, as also repeatedly stated in case-law, to determine whether a restriction of competition by object has been agreed, referring courts should take into consideration: i) the content of its provisions; ii) its objectives and iii) the legal and economic context of which it forms a part.
Paraphrasing a consolidated statement in the ECJ’s case-law and literature on by-object restrictions: where from the above three elements emerges that there is no plausible alternative explanation and economic rationale to the agreement in question other than to restrict competition, the agreement is set to fall within the by-object box as the degree of possible anti-competitive effects is presumed sufficiently high.
The ECJ then considered the EU regulatory framework on pharmacovigilance and information obligations burdening MA holders. In fact, the Court found that Avastin is a product which falls within the pharmacovigilance system provided for by Article 101(1) of the amended Directive 2001/83/EC. The pharmacovigilance obligations ensure that information on the risks associated with the “on-label” or “off-label” use of those products is systematically collected and properly disclosed to the scientific community and to the public. The ECJ also outlined the communication and information obligations to which MA holders are subjected. According to EU law, MA holders are required to inform the European Medicines Agency (“EMA”) and the Commission of any information which may influence the evaluation of the benefits and risks arising in relation to the marketed products. Also, MA holders should always ensure that the information on its products is presented objectively to the public and is not misleading.
In addressing the compatibility of the contested conduct with this regulatory framework, the Court noted that “the requirements for pharmacovigilance that might call for steps to be taken such as the dissemination to healthcare professionals and the general public of information relating to a procedure before the EMA […] rest […] solely with the holder of the MA for that medicinal product” (par. 91). The fact that two undertakings collude to disseminate information related to the product marketed by only one of them, exemplifies that “the dissemination of information pursues objectives unrelated to pharmacovigilance”. Further, the ECJ did not raise doubts that the information disseminated by the parties lacked compliance with the requirements of completeness and accuracy laid down in Article 1(1) of Regulation no. 658/2007.
In light of the above, the ECJ determined that “Article 101 (1) TFEU must be interpreted as meaning that an arrangement put in place between two undertakings marketing two competing products, which concerns the dissemination, in a context of scientific uncertainty, to the EMA, healthcare professionals and the general public of misleading information relating to adverse reactions resulting from the use of one of those products for the treatment of diseases not covered by the MA for that product, with a view to reducing the competitive pressure resulting from such sue on the use of the other medicinal product, constitutes a restriction of competition ‘by object’ for the purposes of that provision”.
It is worth noting that, in the ECJ’s reasoning, the mere fact that two undertakings colluded with respect to the dissemination of the information on the risks of the off-label use of Avastin (where such a task should have been pursued autonomously by the MA holder) is at least a strong indication that the purpose and rationale of the conduct is not pharmacovigilance but rather to alter the demand of the cheaper drug in favor of the more profitable one in the field of ophthalmology. Nonetheless, the misleading nature of the information seems in fact to have played a decisive role in the overall assessment of the conduct and in the outcome of the judgement, i.e. that the agreement constituted a restriction “by-object”.
To wrap up, the case outlined above provides practitioners and firms with useful insights circa the ECJ’s approach when dealing with anticompetitive conducts in the pharmaceutical sector. In particular, the Court established that the scope of the MA should not be considered as binding in the definition of the relevant market. Secondly, the Court further expanded on the notion of “ancillary restraints” and how to assess them. Lastly, the Court shed light on the regulatory framework concerning the pharmacovigilance obligations for pharmaceutical companies: it held that a concerted practice between two competing manufacturers regarding the dissemination of information emphasizing the threat for health of a drug marketed by only one of them can be presumed prima facie contrary to Article 101, particularly if the information is inaccurate and incomplete or not objective and thus misleading.