The ECJ’s ruling in Servier: Never Settle For Less

More is yet to come: the ECJ sent the case back to the GC (i) to reassess the European Commission’s allegation that Servier had abused a dominant position, and (ii) to rule on a third settlement agreement between Servier and generics company Krka. 

For now, pharma companies should use this ruling, together with earlier case law, as guidance for the antitrust assessment of their patent settlement agreements and keep in mind that relevant pharma markets may be defined narrowly with a resulting risk of dominance looming.

Background

Servier is the originator manufacturer of perindopril, a cardiovascular medicine. After Servier’s patent over the active substance of perindopril expired, Servier relied on a “secondary” patent covering the processes for the manufacture of the active ingredient of perindopril to apply for interim injunctions against companies attempting to sell a generic version of perindopril. The validity of the secondary patent was challenged in court by five generics producers, Niche/Unichem, Matrix, Teva, Lupin and Krka. All companies eventually settled their litigation with Servier, agreeing to refrain from entering the market or challenging Servier’s patent. In return, the generic companies received payments from Servier. Apart from these settlement agreements, Servier also acquired certain technology from the generic companies, which led to the termination of a number of generic projects. Alongside the settlement agreement, Servier licenced perindopril to Krka in seven markets. In addition, an assignment and licence agreement was concluded, pursuant to which Krka assigned two patent applications to Servier.

According to the Commission, the technology acquisitions and series of patent settlements breached the EU antitrust rules. The Commission imposed fines on Servier and the generic companies totalling EUR 427.7 million.

On appeal, the GC partly annulled the Commission's decision and lowered the total fines to EUR 315 million (see our January 2019 newsletter). It upheld the Commission's finding that the agreements with four generics producers constituted a restriction of competition 'by object'. However, it deemed the agreements with Krka not to be anticompetitive but based on Krka’s genuine expectation that Servier’s patent was valid. Servier had, as part of the settlement, provided a license to Krka to commercialise perindopril in seven markets. The GC found that, in that context, a settlement agreement linked to a licence agreement can constitute a ‘by object’ restriction only if the Commission demonstrates that the licence agreement was not granted under normal market conditions and thus masks a reverse payment (i.e., a payment by the originator company to the generic company to stay off the market). The Commission had failed to do so.

Furthermore, the GC annulled the Commission’s finding that Servier had – in addition to an infringement of the cartel prohibition in Article 101 TFEU – abused its dominant position by drawing up and implementing, through the technology acquisitions and the settlement agreements, an exclusionary strategy aimed at preventing generic entry. According to the GC, the Commission had attached too much weight to price differences in its market definition on which it based its finding of Servier’s dominance. 

Both the European Commission, Servier and the generic companies appealed the GC’s ruling to the ECJ. The ECJ’s judgment essentially covers two issues: (i) the legal framework for the assessment of pharma settlement agreements under Article 101 TFEU and (ii) the definition of the relevant market for assessing dominance under Article 102 TFEU in the pharmaceutical sector.

(i) The ECJ ruling on the patent settlement agreements

The ECJ confirmed the GC’s ruling that the patent settlement agreements with Niche/Unichem, Matrix, Teva, Lupin constituted restrictions of competition ‘by object’. It did, however, reduce Servier’s fine for its agreement with Lupin from EUR 37 million to EUR 34 million because of a shorter infringement period than the Commission considered in its fine calculations. 

In its judgment, the ECJ followed the rules of thumb for the antitrust assessment of patent settlement agreements set out earlier case law (see our January 2021  newsletter):

• to assess whether a company is a potential competitor, the test is whether there are "real and concrete possibilities" of it joining the market and competing with the companies present in it;
• settlement agreements offering significant value transfers from the originator to the generic manufacturer that act as an inducement for the latter to refrain from entering the market are likely to be considered as ‘by object’ restrictions, and
• any pro-competitive effects associated with the settlement agreements cannot rebut their characterisation as a ‘by object’ restriction. Interestingly, the ECJ, therefore, seems to have explicitly moved away from its earlier case law in which pro-competitive effects could play a role in the examination of an agreement’s anticompetitive object. 

Following these rules of thumb, the ECJ found that the GC had erroneously rejected the Commission’s reasoning that the settlement and licence agreements between Krka and Servier were an infringement by object. The GC had disregarded that the infringement attributed to Servier and Krka was not limited to a patent dispute settlement agreement. Instead, it pursued the broader objective of sharing markets between the two companies, with the Krka licence agreement being the inducement for Krka’s commitment not to compete with Servier on its core markets. 

The ECJ therefore found that the Commission’s appeal on this point was successful and referred the matter back to the GC to rule on the assignment and licence agreement between Krka and Servier.

(ii) The ECJ ruling on the relevant market for abuse of dominance

Following the Commission’s ground of appeal, the ECJ disagreed with the GC’s finding that the Commission relied too excessively on price to conclude that the relevant market for the dominance assessment consisted of perindopril only. The ECJ ruled that, irrespective of the specific characteristics of the pharmaceutical sector, a distinct market exists if a price change does not result in a shift in sales between medicinal products intended for the same therapeutic indication. The GC therefore erred in law in holding that the Commission had wrongly based its dominance assessment on a too narrow market definition.

The ECJ set aside the GC’s decision and has sent the case back to the GC to rule on Servier’s alleged abuse of dominance.

Conclusion

The perindopril-saga continues with the modafinil pay-for-delay tale on its tail. However, pharma companies can already note that markets may be defined quite narrowly, thereby increasing the risk of a finding of a dominant position. In addition, playing by the above-mentioned rules of thumb when concluding patent settlement agreements will safeguard these agreements from raising antitrust flags.

Pharma companies should also keep in mind that the Commission has moved on to other pastures: disparagement is the current talk of the town (see our May 2024 newsletter), together with an ongoing investigation into a pharmaceutical cartel. More developments are therefore yet to come.