The Servier Case: The Game isn’t Over Yet

On 27 June 2024, French pharmaceutical company Servier – which between 2005 and 2007 concluded several patent settlement agreements involving reverse payments with five generic companies (Niche/Unichem, Matrix, Teva, Lupin and Krka) – largely lost its battle against the European Commission (the “Commission”). Following appeals lodged by almost all parties involved, including Servier and the Commission, the European Court of Justice (the “ECJ”) issued seven judgments siding with the view expressed in prior cases that certain so-called “pay-for-delay” agreements are inherently anticompetitive. But more is yet to come in this over 10-year saga, as the ECJ referred part of the case back to the General Court on (i) the compatibility with Article 101 of the Treaty on the Functioning of the European Union (“TFEU”) of the agreement assigning to Servier two patent applications filed by generic company Krka and (ii) the abuse of dominance allegedly committed by Servier on the perindopril market.

Background of the case

Servier developed perindopril, a drug primarily intended for the treatment of hypertension and heart failure, by inhibiting the angiotensin converting enzyme (“ACE”). Until the early 2000’s, Servier benefitted from patent protection and, as such, enjoyed a monopoly in the market. Alongside obtaining supplementary protection certificates covering perindopril’s active ingredient (which took the form of a salt, erbumine), allowing Servier to prolong its exclusivity beyond 2001 in the UK and various Member States (including France and Italy), Servier filed a number of secondary patent applications relating notably to (i) the manufacturing process of perindopril’s active ingredient and (ii) the alpha crystalline form of perindopril erbumine and the process for its preparation (known as the “947 patent”, granted in 2004).

Several manufacturers preparing their market entry with a generic version of perindopril filed opposition proceedings against the 947 patent before the European Patent Office, seeking its revocation on grounds of lack of novelty, lack of inventive step, and insufficient disclosure of the invention. Patent litigation disputes over the validity of the 947 patent were also brought before national courts, mainly in the UK.

These proceedings eventually led to the conclusion, between 2005 and 2007, of several patent settlement agreements between Servier and four generic companies (Niche/Unichem, Matrix, Teva and Lupin).  The generic companies in essence agreed to stop their efforts to market a generic version of perindopril until the expiry of Servier’s patents (“non-marketing” clauses) and to withdraw and refrain from bringing claims challenging the validity of Servier’s patents (“non-challenge” clauses), in exchange for Servier’s commitment not to bring infringement proceedings against them. These agreements also provided for significant payments to be made by Servier to the generic companies.

Servier also reached a settlement with a fifth generic company, Krka, which had begun to market a generic version of perindopril in several Member States in Central and Eastern Europe. This settlement included three distinct agreements:

• a classic patent settlement agreement, whereby Servier withdrew its actions for infringement against Krka while Krka agreed to non-challenge and non-marketing clauses; this agreement did not involve any value transfer;
• a license agreement pursuant to which Servier granted Krka, for the duration of the 947 patent, an exclusive, irrevocable license on such patent, covering Krka’s main markets in Eastern and Central Europe, against the payment of 3% royalties on net sales in these territories; and
• an assignment and license agreement, whereby Krka assigned two patent applications relating to perindopril to Servier for 30 million euros (together the “Krka agreements”).

In July 2014, the Commission fined Servier and the above-mentioned generic manufacturers a total of 427.7 million euros for infringing Article 101 TFEU by agreeing to substitute practical collaboration for the risks of competition, thereby significantly delaying generic entry, to the detriment of patients and health systems. The Commission also found that Servier abused its dominant position in the markets for perindopril and the corresponding technology, in breach of Article 102 TFEU, by implementing a global anticompetitive strategy consisting in acquiring any technology that was a possible source of competition.

In a judgment of 12 December 2018, the General Court partially annulled the Commission’s decision. While upholding the Commission’s findings insofar as they related to the Niche/Unichem, Matrix, Teva and Lupin agreements, the General Court considered that the Krka agreements did not amount to a restriction of competition. In addition, the General Court, following a detailed analysis of the competitive constraints exerted on Servier’s perindopril by other ACE drugs, found that the Commission had incorrectly delineated the relevant market by limiting it to perindopril alone, in its brand-name and generic versions.

Through seven rulings of 27 June 2024, the ECJ largely confirmed the General Court’s judgment in respect of the Niche/Unichem, Matrix, Teva and Lupin agreements and the corresponding fines which are now definitive. However, the ECJ disavowed the General Court with regard to the Krka patent settlement and license agreements, which were found anticompetitive. The ECJ decided to remit the case to the General Court regarding the compatibility with Article 101 TFEU of the assignment and license agreement between Servier and Krka. Finally, the ECJ also criticized the General Court’s findings relating to market definition, thereby reopening the debate on the abuse of dominance allegation, which will now have to be examined by the General Court.

The view that certain so-called “pay-for-delay” agreements are anticompetitive is largely confirmed by the ECJ

In line with the strict approach already taken in the Generics UK and Lundbeck precedents, the ECJ applied its now well-defined legal test for the assessment of reverse payment patent settlement agreements under Article 101 TFEU. In this respect, it should be recalled that not all patent settlement agreements involving a value transfer are anticompetitive. To the contrary, there may be circumstances where it is perfectly legitimate for a generic manufacturer, after assessing the chances of success of patent litigation in court proceedings, to settle the dispute with the originator company and eventually to receive sums from that originator company in compensation for the costs of or disruption caused by the litigation.

The legal test applied by the ECJ requires to determine, at an initial stage, whether the two pharmaceutical companies that are parties to the agreement are in a situation of competition, if not actually, at least potentially, i.e. whether at the time the agreement was concluded, the generic manufacturer (i) had already taken significant preparatory steps to enter the market within a timeframe sufficient to impose competitive pressure on the originator company (which can be of several years) and (ii) did not face insurmountable barriers to do so (the existence of secondary patents such as process patents not being considered insurmountable barriers).

If the conclusion is that the parties to the agreement were indeed potential competitors, the second stage of the analysis focuses on the content of the agreement itself, its objectives and overall context. According to the ECJ, in order to determine whether a patent dispute settlement agreement involving value transfers from the originator company to the generic company (whichever the form), constitutes a restriction of competition by object, it is necessary to ascertain, first whether the net gain from these transfers is justified and, if not, it must be verified, second, if these transfers find their justification only in the commercial interest of these drug manufacturers not to engage in competition on the merits.

In the present case, the ECJ first confirmed the Commission’s and the General Court’s view that the agreements between Servier, on the one hand, and Niche/Unichem, Matrix, Teva and Lupin, on the other hand, were anticompetitive. In particular, rejecting Servier’s argument that some of these agreements brought forward the date on which generic companies could legally enter the market, the ECJ recalled that agreements with potential competitors including non-marketing and non-challenge clauses not based on the recognition of the validity of Servier’s patents but on significant value transfers inducing the generic companies to stay out of the market, amount to a restriction of competition by-object, with the consequence that the examination of their effects on competition is unnecessary to characterize an infringement of Article 101 TFEU.

Second, substituting its analysis for that of the General Court, the ECJ found that two of the Krka agreements, i.e. the settlement agreement and the license agreement, also amounted to a by-object restriction of competition. According to the ECJ, although these two agreements – unlike the other agreements concluded by Servier and unlike the agreements which the ECJ examined in the Generics UK and Lundbeck precedents –  did not involve any reverse payments from Servier to Krka, nonetheless resulted in a geographical market sharing between Servier and Krka of their respective core national markets for perindopril within the European Union, in violation of Article 101 TFEU. Specifically, the ECJ criticized the General Court for having looked at these two agreements separately whereas they were closely connected, as supported by evidence in the case file showing that Krka would not have entered into the settlement agreement but for the license agreement. According to the ECJ, the license agreement in effect constituted the incentive for Krka to stay out of Servier’s core markets in Western Europe and was therefore akin to a value transfer. In other words, the ECJ concluded that both agreements were complementary: on the one hand, through the settlement agreement, Servier was guaranteed that Krka would not launch a generic version of perindopril in Servier’s main markets, including France and the UK, while, on the other hand, through the license agreement, Krka was given the possibility to continue operating in Eastern and Central Europe without running the risk of being sued by Servier for patent infringement.The ECJ also rejected Servier’s argument that the license agreement was pro-competitive insofar as it allowed Krka to legally commercialize a generic version of perindopril in certain markets, which it could not have done otherwise. According to the ECJ, the potential pro-competitive effects of the license agreement in eastern European markets where Krka was active were irrelevant: indeed, while the grant of a license in exchange for an undertaking not to challenge the validity of a patent may, in certain circumstances, be legitimate, the ECJ stated that the evidence of the case showed that Krka voluntarily sacrificed certain markets to keep or gain immediate access to its core markets. The purpose of Servier and Krka’s arrangement was therefore clearly to organize a market sharing, to the benefit of both parties, with the effect that competition may have been enhanced in Eastern Europe but was, through the elimination of a potential competitor, restricted on Servier’s core markets, which were precisely those concerned by the Commission’s finding of infringement. In other words, what the ECJ deems anticompetitive is not the grant of the license itself to Krka, but the combination of the settlement agreement and the license agreement, which together amounted to a geographical market sharing between Servier and Krka, with the result that Servier was protected from the competition of Krka in its core markets. Such result is directly in breach of the fundamental principles of the EU, which has established an internal market with the purpose of removing barriers to free movement of goods between Member States, and therefore prohibits any conduct, such as market partitioning, that artificially recreates such barriers.

Finally, although an effect-based analysis is not necessary when a conduct is deemed a restriction of competition by object, the ECJ upheld the Commission’s finding that, irrespective of Krka’s actual possibility to enter Servier’s core markets with its generic version of perindopril, Krka forwent this possibility by entering into the settlement and license agreements with Servier, which resulted in the elimination of an immediate source of potential competition for Servier and, as such, had a concrete – and not only hypothetical – impact on competition.

The return to a price-centered approach to market definition in the pharma industry?

The other important drawback for Servier stemming from the ECJ’s judgment is the latter’s very strict application to prescription drugs of the classic economic principles guiding the analysis of demand-side substitutability for the purposes of market definition.

The Commission, consistently with its previous practice when competition from generics is at stake, had in its decision narrowly defined the market at the level of perindopril’s active ingredient, which allowed to establish a dominant position of Servier on that market. On appeal, however, the General Court concluded that the market was in fact wider and encompassed not only perindopril and its generics but also all the other drugs classified as ACE inhibitors. To reach this conclusion, the General Court went beyond price competition, which is the most immediate factor in originator-generic competition. Underlying the unique characteristics of the pharmaceutical industry, which is heavily regulated and where demand for prescription drugs is largely influenced by doctors’ choice and less sensitive to price variations, the General Court emphasized the importance of qualitative factors in the assessment of demand-side substitutability, such as therapeutic indications, therapeutic recommendations, doctors’ experience and perceptions as to a possible superiority of perindopril over other ACE inhibitors.

In a rather unexpected move, the ECJ completely overturned the General Court’s judgment, considering that the latter contradicted itself and disregarded the principles that undermine market definition by including other ACE inhibitors in the relevant market based on qualitative criteria, even though the demand for perindopril was not impacted by the important price difference between perindopril and other ACE inhibitors.

According to the ECJ, in the pharmaceutical sector as in any other industry, interchangeability or substitutability between two products is not assessed solely by reference to the objective characteristics of the products and whether, from a functional point of view, they are capable of satisfying the same need, but also requires a determination of economic substitutability, i.e. whether a relative change in the price of a given product shifts demand towards another, cheaper product. Therefore, the General Court’s finding that demand for perindopril remained stable despite an important decrease in the price of other ACE inhibitors should have been, according to the Court, a strong indicator of the absence of substitutability between perindopril and other ACE inhibitors, regardless of whether this stability of demand may have been due to other non-price related factors, such as perindopril’s intrinsic qualities, the role of prescribing doctors or Servier’s promotional efforts.

This “U-turn”, which puts back price-based factors at the heart of the assessment of demand-side substitutability and relegates other – more qualitative – factors to a second position, contrasts with the position of the ECJ in the Hoffmann-La Roche 2018 preliminary ruling where it underlined the specific features of competition in the pharmaceutical sector, notably the fact that prescribing doctors are primarily guided by considerations of therapeutic appropriateness and the efficacy of medicines.¹   

It therefore remains to be seen whether the ECJ’s position in Servier is case-specific and how it will translate into the case law of the Member States, where the bulk of pharmaceutical cases lie. In particular, it will be interesting to see whether the price-based approach deriving from the Servier judgment will remain confined to cases involving originator-generic competition or whether it will extend to other situations.

For now, two things remain certain for sure:

• first, despite the ECJ’s strong stance and dissent with the General Court, there is still room for debate on market definition in the pharmaceutical sector, particularly in cases where a relative price elasticity of demand for different drugs can be established; and
• second, this calls for great caution from pharmaceutical companies, which should be conscious that their commercial vision of their competitive environment may not always align with the legal and economic approach of competition authorities; this misalignment may have important consequences in terms of their actual degree of market power and the responsibility deriving from the finding of a potential dominant position.