On 12 December 2018, the EU General Court (GC) delivered its judgment in the Servier reverse payment patent settlement case, the second GC judgment to date on reverse payment patent settlements (after the 2016 Lundbeck judgment).

The GC confirmed that such agreements fall within the scope of Article 101 of the Treaty on the Functioning of the European Union (TFEU) and may constitute a restriction of competition by object. The GC also confirmed that, to the extent that an infringement is a restriction by object, it is not necessary to analyse its effects. Finally, the GC annulled the fine imposed by the European Commission (Commission) because the Commission failed to establish that the market was limited to the perindopril molecule.


After the molecule patent to its blockbuster medicine, perindopril, expired in 2003, Servier sought to rely on secondary, process and form, patents to protect perindopril. However, these patents only offered limited protection, and generic companies prepared to enter the market, either by avoiding the protected processes and form, or by challenging Servier’s remaining patents.

In July 2014, the Commission fined Servier and the generics a total of €427.7 million for having entered into agreements that delayed market entry of generic versions of perindopril and protected Servier from price competition in the EU (see, our analysis here and here). In addition to the settlement agreements, the Commission found that Servier engaged in conduct that was part of an overall strategy to delay or prevent generic entry, namely:

  • Between 2005 and 2007, Servier entered into settlement agreements with generics that were challenging its patents. The Commission found that these were not legitimate litigation settlements because their purpose was to keep generic competitors out of the market by offering them a portion of perindopril profits. All of these agreements entailed Servier making significant payments (or providing other inducements) to the generics.
  • Servier entered into ten distribution agreements with “friendly generics”, granting them the right to distribute a so-called “authorised generic”. This effectively gave Servier control over generic entry. In return, the generics committed to not sell other generic versions of perindopril.
  • In 2004, Servier acquired alternative technologies for the production of perindopril, forcing the generics to halt their projects and market entry. The Commission found that Servier did not exploit this technology and quoted Servier as admitting that it was using the acquisition to “strengthen the defence mechanism”.

The General Court’s Judgment

In its Servier decision, the Commission went beyond Lundbeck and in its Fentanyl decision regarding an agreement between Sandoz and Janssen Cilag, in that it not only looked at whether the settlement agreements between Servier and generics restricted competition by object under Article 101 TFEU, it also analysed the effects of the agreements and whether Servier’s conduct amounted to an abuse of dominance under Article 102 TFEU.

The “By Object” Analysis

The GC confirmed the analytical framework used by the Commission in both Servier and Lundbeck to assess whether the patent settlement agreements had the potential to restrict competition “by their very nature”.

While a patent granted by a public authority is presumed valid and its ownership is presumed lawful, the GC stressed the importance and legitimacy of settlement agreements, which provide a route to bring litigation to an end. However, the GC upheld the Commission’s conclusion that, when an originator induces a generic to refrain from entering the market and/or challenging its patent, the agreement amounts to market exclusion. The GC concluded that it is this inducement – rather than the recognition of the validity of the patent – that results in settlement agreements restricting competition.

As in Lundbeck, the GC confirmed that Servier and the generics were at least potential competitors; the generics had a real, concrete possibility of market entry, notwithstanding the existence of barriers linked to Servier’s patents, the difficulties in obtaining market authorisations, etc. The GC went on to find that the commitment by the generics to limit their efforts to enter the market and the value transfer from Servier to the generics meant that all but one of the agreements amounted to restrictions of competition by object.

However, the GC found that Servier’s agreement with Krka did not amount to a restriction by object because the Commission had not established the existence of an inducement from Servier for Krka to withdraw from the market. The GC rejected the Commission’s position that a royalty paid by Krka to Servier under a license was not concluded at arm’s length.

Finally, the GC found that, because of overlaps between the infringements, the Commission should have further reduced the fine imposed on Servier in respect of its agreement with Matrix. The GC reduced this amount by 30% (from €330.99 million to €228.32 million).

The “By Effect” Analysis

In its Servier decision, the Commission concluded that each agreement appreciably restricted potential competition among Servier and the relevant generic. Further, the settlement agreements appreciably increased the likelihood that Servier’s significant market power would remain uncontested for longer, and that consumers would be deprived of the significant reduction of prices that would result from timely and effective generic entry.

The GC noted that taking into account the concrete effects of an agreement is “superfluous” where the agreement restrains competition by object. As a result, where the Commission finds infringement on both a by-object and by-effect basis, “an error rendering unlawful the ground based on the existence of a restriction by effect does not, in any event, have a decisive effect on the operative part adopted by the Commission in that decision, inasmuch as the ground based on the existence of a restriction by object […] is not vitiated by an illegality”.

However, the GC considered the Commission’s effects analysis in connection with Servier’s agreements with Servier and Krka (since it found that those agreements did not restrict competition by object). The GC found that the Commission had not established that:

  • Without the settlement agreement, Krka would probably have entered the market;
  • Continuing the patent litigation against Servier would probably have led to invalidation of the patent quickly; or
  • The agreement whereby Krka sold its technology to Servier had anti-competitive effects. The Commission merely indicated that Krka would otherwise have been free to license its technology.

The GC concluded that it is insufficient for the Commission to merely show that the agreements are “likely” to give rise to “potential” anti-competitive effects (e.g., exclusion of a potential competitor). Rather, the Commission should have examined the real events that took place between the signature of the agreement and the publication of its decision. Finally, the GC noted that it is not its role to examine, on the basis of the file, whether the agreements would have had restrictive effects, and annulled the part of the decision relating to the agreement between Servier and Krka.

The Article 102 TFEU Analysis

The Commission found that Servier abused its dominant position by (i) acquiring API technology (which it never used), preventing generics from developing non-infringing perindopril formulations, and (ii) inducing generics to accept restrictions of competition in the settlement agreements, thus strengthening or maintaining Servier’s market power and foreclosing its competitors.

In a perhaps surprising turn of events, the GC annulled the entire fine imposed on Servier for the infringement of Article 102 TFEU. Importantly, the GC overturned the Commission’s abuse of dominance finding on the basis that the Commission erroneously defined the relevant market. The issue of market definition is relatively infrequently raised in annulment proceedings before the EU Courts and even more rarely accepted.

In its Decision, the Commission defined the market as the single molecule of perindopril, in its originator and generic versions (the Commission also defined an upstream market for perindopril API technology). This narrow definition was mainly based on the fact that other angiotensin converting enzyme (ACE) inhibitors were not sufficiently close competitors to perindopril. The GC rejected this analysis. First, it stressed the specificities of competitive relationships in the pharmaceutical sector, resulting mostly from the fact that demand for prescription medicines is typically determined not by consumers, but by prescribing doctors, who are less guided by the cost of the treatment than by its therapeutic use.

The GC went on to conclude that the Commission committed a series of errors in defining the relevant market, namely, the Commission (i) wrongly found that perindopril differed from other ACE inhibitors in terms of therapeutic use, (ii) erroneously found that doctors’ inertia significantly restricted competitive pressure on Servier’s perindopril, (iii) under-estimated the likelihood of switching between different ACE inhibitors, (iv) did not sufficiently take into account companies’ promotional efforts, and (v) attributed too much importance to price in its analysis. In relation to price, the GC held that the Commission did not fully pay sufficient attention to qualitative, non-price factors evidencing competitive pressure on Servier.

As a result, the GC concluded that the Commission’s failure to show that the product market was limited to the perindopril molecule alone vitiated its Article 102 TFEU assessment.

Looking Forward

This judgment will likely be important in future judgments by the EU Courts. The Court of Justice of the EU (CJEU) is expected to provide further guidance on “by object” infringement through reverse payment settlements in the appeal against the GC’s Lundbeck judgment. It will also address these issues in the context of the preliminary questions referred by the UK Competition Appeal Tribunal in the appeal by GlaxoSmithKline and generics against the UK Competition and Markets Authority’s decision fining them approximately £45 million for allegedly delaying entry of generic versions of GlaxoSmithKline’s anti-depressant Seroxat (paroxetine) in the UK.