Skip to content

Patent settlements and competition law – guidance from Europe’s top court

On 25 March 2021, the European Court of Justice (ECJ) ruled for the second time on a “pay-for-delay” settlement. These are settlements of a patent dispute that involve payments or other value transfers from the originator manufacturer to the generic manufacturer in return for a delayed market entry.

The ECJ dismissed appeals by the originator manufacturer Lundbeck and five generic manufacturers against the ruling of the General Court (GC) of September 2016, which in turn had upheld the European Commission’s (Commission) infringement decision against them of June 2013.

The ECJ’s judgment refines the framework established in its first pay-for-delay ruling (GSK (Paroxetine), judgment of 30 January 2020) and introduces a novel duty of care for undertakings to preserve evidence.

This article summarises the judgment, puts it in the context of other pay-for-delay decisions by the Commission and European courts and assesses the judgment’s practical implications.

Facts of the case

Lundbeck, a Danish pharmaceutical company, had developed the antidepressant medicine “citalopram”. After the compound patent had expired, Lundbeck continued to hold a number of “secondary patents” protecting certain manufacturing methods. Meanwhile, manufacturers of generic versions of citalopram were preparing their market entry.

After Lundbeck had brought a number of patent-infringement cases, the company concluded settlement agreements with four generic manufacturers. The generic companies committed not to enter the citalopram market for a certain period and received cash payments from Lundbeck. In addition, Lundbeck purchased their stock of generic products and concluded – limited – supply agreements with the generic companies relating to the original drug.

Shortly after the agreements expired, the Commission launched an antitrust investigation against Lundbeck and concluded in 2013 that the settlement agreements infringed Art. 101 TFEU. It imposed severe fines on Lundbeck (EUR 94 million) and the generic manufacturers (EUR 52 million).


The Commission’s decision in 2013 was the first infringement decision by the Commission on a pay-for-delay case, followed by a number of similar investigations. In November 2020, the Commission concluded the most recent investigation, imposing a EUR 60.5 million fine on Teva and Cephalon (see here), and has now turned its attention to other types of potentially anti-competitive practices in the context of the market entry of generic drugs (see here).

Nonetheless, many questions with regard to pay-for-delay cases are still unsettled. A year ago, in GSK (Paroxetine), the ECJ established the framework for assessing patent settlements under Art. 101 and 102 TFEU and answered some open questions. Further clarification is now provided in the Lundbeck judgment.

Cartel prohibition applies despite patent protection – parties are potential competitors

First, the ECJ dealt with the question of whether Lundbeck and the generic companies were competitors. They had strongly contested that they were even potential competitors. In their views, the agreements were therefore out of scope of Art. 101 TFEU.

The ECJ did not follow their arguments and considered Lundbeck and the generic companies as potential competitors:

  • The ECJ followed previous case law confirming that the legal test to establish whether companies are potential competitors is as follows: In the absence of the settlement agreements, would there have been “real and concrete possibilities” for the generic companies to enter that market and compete with the originator (para. 55)? This test has two limbs:
  • Firstly, one has to determine whether the generic company has “a firm intention and an inherent ability” to enter the market (para. 56). In this context, the ECJ assessed the generic companies’ preparatory steps taken to enter the market and whether these steps would enable it to enter the market within such a period of time as would impose competitive pressure on the originator (para. 57). Preparatory steps include obtaining a marketing authorisation, building up stock or challenging patents held by the originator.
  • Secondly, it must be determined that the market entry of such a manufacturer of generic medicines does not meet barriers to entry that are “insurmountable” (paras. 56 and 57). One fundamental aspect in this context is the existence of patents held by the originator.
  • The application of these criteria to the Lundbeck case was similar to the GSK (Paroxetine) case, with some additions. The ECJ did not find any errors in law in the GC’s assessment that the generic companies had undertaken sufficient preparatory steps to demonstrate “a firm intention and an inherent ability” to enter the market (paras. 59 et seqq.). The ECJ clarified – and this is a new element compared to GSK (Paroxetine) – that it is irrelevant whether generics already hold marketing authorisations (MA) or not (para. 83). The absence of MAs cannot preclude potential competition. Even without an MA, a generic company can have taken sufficient steps to prepare its market entry and thus put competitive pressure on the originator (para. 84).
  • The existence of a patent protecting the manufacturing process of an active ingredient that is already in the public domain cannot, as such, be regarded as an “insurmountable barrier” (para. 58). The ECJ distinguished between process and compound patents and concluded that the existence of a process patent does not preclude potential competition (para. 59), in particular if there are other processes for manufacturing the drug that might not infringe the patent (para. 62). It only matters whether the generic manufacturer has shown readiness to challenge the validity of that process patent and has taken the risk of being subject to infringement proceedings brought by the patent holder (para. 59).
  • Moreover, the ECJ made clear that it does not matter whether the patent is ultimately found valid or not (para. 71). In the Lundbeck case, the EPO Board of Appeal and the Netherlands Patent Office had later confirmed that the patents in question were valid. However, the ECJ held that this was not relevant for the assessment of whether the parties to the agreements were potential competitors because this fact was unknown to the parties at the time they concluded the agreements – only events known at the time of the agreements should be taken into account (para. 69).
  • Finally, the ECJ referred to the perception by the originator of the degree of risk that the generic medicines present (para. 76). The ECJ did not object to the GC’s statement that the strongest evidence for potential competition is the fact that Lundbeck entered into agreements with the generic companies and paid them for them delaying their market entry.

Restriction of competition “by object”

The second question that was decisive in the Lundbeck case was whether the settlement agreements constituted restrictions of competition “by object”, i.e. whether they were anti-competitive by their very nature, comparable to a market sharing agreement.

This question was fundamental for the Commission since it meant that the Commission did not have to prove actual anti-competitive effects.

The ECJ confirmed the Commission’s and GC’s assessments that the agreements were “by object”-restrictions:

  • The ECJ started off, as in the GSK (Paroxetine) judgment, by stating that pay-for-delay agreements are not generally anti-competitive – they cannot be considered restrictions “by object” in all cases (para. 113). Value transfers do not necessarily indicate a clear anti-competitive objective. The ECJ sets a high hurdle for the “by object”-category: It must be apparent that the transfer of value cannot have any explanation other than it being in the interest of both the holder of the patent at issue and the party allegedly infringing the patent not to engage in competition on the merits (para. 114). In other words, with the agreement, the originator must buy off competition, to classify it as anti-competitive “by object”.
  • Whether this is the case must be determined on a case-by-case basis and depends mainly on an assessment of the transferred value – was the net gain of the transfers of value from the originator to the generic manufacturer sufficiently significant to act as an incentive to the generic manufacturer to refrain from entering the market (para. 115)?
  • In Lundbeck, this was the case, according to the GC, and the ECJ did not object to this (para. 117). The ECJ found that it was primarily the cash payments that induced the generic manufacturers to accept the agreements. The amounts paid to the generic manufacturers by Lundbeck took into consideration the profit that the generic companies expected to make if they had entered the market. In contrast, the existing patents themselves appeared not to be sufficient for the generic manufacturers to accept the settlements agreements – they had already undertaken significant investments to prepare for their market entry and were challenging Lundbeck’s patents.

Duties of care triggered by sector inquiry

A novel element of the Lundbeck decision is the duty of care that the ECJ imposed on undertakings active in a sector that underwent a sector inquiry by a competition authority.

Two generic manufacturers argued that the Commission had infringed their rights of defence by waiting too long to initiate proceedings against them. While the Commission was investigating Lundbeck’s conduct from December 2003, the proceedings against the two generic companies were started only in 2010 and 2011, respectively. By the time the Commission initiated proceedings against them, they had already deleted documents that were relevant for their defence.

The ECJ rejected this argument. The ECJ noted that the sector inquiry in the pharmaceutical sector that the Commission started in January 2008 had triggered a specific duty of care. This duty concerned undertakings belonging to the sector and in particular undertakings that had participated in practices explicitly mentioned in the decision initiating the sector inquiry (such as pay-for-delay agreements). The duty obliged the two generic manufacturers to retain all relevant documents and evidence, given that sector inquiries often lead to subsequent investigations (para. 154).

The ECJ found that the duty of care applied from January 2008 on, which was four and a half years after the agreements had expired (para. 155). This was sufficient in the ECJ’s view so that the court could leave open whether the two generic manufacturers had become aware of the investigation against Lundbeck before it was extended to them and whether the Commission had unreasonably delayed this step.

Practical implications and outlook

Overall, the ECJ’s judgment contains few surprises, but provides further clarity on the framework for assessing patent settlement agreements and introduces a duty of care following sector inquiries.

The ECJ’s findings on potential competition make clear that pay-for-delay agreements will rarely be exempt from scrutiny under Art. 101 TFEU. This is understandable from an enforcement perspective. However, the consideration that the very fact that the originator felt a need to conclude an agreement indicates potential competition appears circular. A broad definition of potential competition also has implications at other levels of the legal assessment. For instance, fines are calculated based on the value of the sales affected by the infringement. The ECJ deemed sales in all countries affected where Lundbeck was threatened by potential competition (para. 190). The broader the notion of potential competition, the more sales are included in the calculation of the fine.

What the Lundbeck judgment leaves open, is in which scenario existing patents suffice to preclude potential competition because they constitute an “insurmountable barrier”. Both the GSK (Paroxetine) and Lundbeck cases revolved around process patents and the ECJ made continuous references to this fact in its reasoning. If, in contrast, a patent protects the compound, the answer might be different, at least where market entry is sufficiently far away. Another potential “insurmountable barrier” could be an orphan drug status, i.e. market exclusivity that is granted for drugs developed to treat rare (“orphan”) diseases. In any case, the strength of the patent does not play a crucial role. The ECJ wants to avoid requiring competition authorities to ascertain the strength of a patent or anticipate the outcome of challenges and infringement proceedings (para. 60).

The ECJ puts the type of agreements concluded by Lundbeck in the “by object”-category. At the other end of the spectrum lie settlement agreements that involve payments that only compensate for the disruption and costs caused by the dispute – such agreements are clearly admissible. Many settlements involving value transfers will however fall in the grey area between these two extremes. In this regard, the Servier case that is already pending at the ECJ should provide further clarity.

The duty of care on undertakings active in a sector that is or was the subject of a sector inquiry could have broad practical implications. The Lundbeck judgment mentions one possible consequence if companies fail to comply with this duty – they cannot blame the Commission if documents relevant to their defence are deleted. However, it leaves open what other consequences violations might have. In any case, the duty of care is relevant beyond the pharmaceutical sector, particularly in the tech sector where competition authorities around the world have launched sector inquiries.