Assessing potential competition and by object and effect infringements in the pharmaceutical market

AG Kokott’s opinion in Case C-307/18 Generics (UK) Ltd and Ors v Competition and Markets Authority (Paroxetine), Court of Justice of the European Union.

On 22 January 2020, AG Kokott delivered her opinion following the CAT’s request for a preliminary ruling in the appeal by GSK, Generics (UK) Ltd, Xellia Pharmaceuticals, Alpharma, Actavis and Merck of the CMA’s infringement decision adopted on 12 February 2016.[1] This is due to be followed, in very short (pre-Brexit) order, by the Court of Justice’s judgment, expected on 30th January, 2020.

The CAT sought guidance on the application and interpretation of both Articles 101 and 102 where the regulatory findings of anti-competitive conduct concern patent litigation and settlements of the type commonplace in the pharmaceutical industry. While the CAT referred a number of questions concerning the interpretation of both Article 101 and Article 102, this article will focus only on the questions concerning the interpretation of Article 101. Specifically, it considers only the concepts of: (i) potential competition and (ii) restriction of competition by object in the pharmaceutical market – the same issues as are currently under consideration by the Court of Justice (CJEU) in the Lundbeck and Servier appeals.

Potential competition

The starting point is that the type of agreements entered into between GSK and the generic manufacturers could not have been found to have infringed competition law, unless – as the European General Court (GC) recently reiterated in Lundbeck[2] – it was first established that the generic manufacturers were, in fact, competitors of GSK. The assessment of patent dispute settlements from a competition perspective poses a conceptual difficulty in this regard; as patents are, by their very nature, intended to ensure exclusivity. The question must therefore be whether there is potential competition which could have been restricted by the patent dispute settlement. 

The CAT, live to the appeals in both Lundbeck and Servier addressing the same legal issues, was concerned that different considerations (and potentially defences) might apply to pharmaceutical patent dispute settlements due to the impossibility of knowing whether the underlying patent was valid and therefore whether it was infringed, which, in turn, made it impossible to determine whether there was potential competition which could be restricted. The European courts have previously held that the presumption of validity from which patents[3] benefit – at least where they are the subject of scrutiny from a competition authority rather than an authority or court with patent jurisdiction – does not mean that generic products should be presumed to be infringing the patent.[4]

The GC explained in Lundbeck that the test for potential competition necessarily differs from that of actual competition because it must take account of the lack of certainty that necessarily exists in predicting what might have happened absent the relevant agreement. The court confirmed that the test for potential competition remains as set out by the CJEU in Delimitis nearly thirty years ago and requires an assessment of the barriers to entry to the relevant market; and whether, absent the agreement, there would have been real concrete possibilities for the potential competitor undertaking to enter the market and compete.[5] 

In her Paroxetine Opinion, AG Kokott states that the validity of the underlying patent and the likelihood of a generic manufacturer being successful in its patent dispute should be put to one side when assessing whether potential competition exists. Kokott’s interpretation is that the “essential factor” in the assessment of potential competition is “whether an undertaking has the ability to enter a particular market” and that “its intention to enter it may also be of relevance”.[6]  She then goes on to give guidance on conducting this assessment specific to pharmaceutical patent disputes and settlements. She warns that the uncertainty around the validity of the underlying patent should not be interpreted as an insurmountable barrier to entry but is, instead, a “fundamental characteristic” of competitive relationships in the pharmaceutical industry. Indeed, reiterating the GC’s findings in Lundbeck, she notes that the conclusion of an agreement whose purpose is to keep one party out of a specific market is a strong indication that a competitive relationship exists.[7] According to Kokott, the question for the competition authority is whether, despite the existence of the patent rights at issue, the generic manufacturer has real concrete possibilities to enter the market at the relevant time. In this assessment, the perception of the patent holder as to the competitive pressure which may be exerted by the generic manufacturer is highly relevant. Potential competition can exist even before the expiry of a patent as generics will want to be ready to enter the market the moment the patent expires. Kokott also notes that while the speed of the potential entry by a competitor is important, the test is not whether there can be immediate entry; rather, that entry could take place within a reasonable period such that it is capable of forming a competitive constraint on current market participants. Again, legal proceedings are indicative of the existence of potential competition.

By object restrictions

The CAT’s central question for the CJEU was, in light of the requirement that an agreement must “clearly reveal a sufficient degree of harm to competition”[8] to constitute a restriction ‘by object’, whether a patent settlement agreement could constitute such a restriction where it was not possible to determine which party was likely to succeed. 

Kokott’s view was that this approach conflates object and effect restrictions, which are alternative conditions for assessing whether an agreement infringes Article 101. When conducting a ‘by object’ restriction assessment, effect is irrelevant; the appropriate question is whether the agreement pursued an anti-competitive purpose.[9] In so far as whether patent dispute settlements necessitate any special considerations, Kokott noted that context was relevant in any ‘by object’ restriction assessment and an agreement must be considered in the context of the specific market where it is implemented, which involves considering the nature of the goods or services affected and the functioning and structure of the market in question.[10] 

By way of guidance specific to patent settlement agreements, she states that the entering into of a patent settlement agreement pursuant to which the sole consideration from a potential competitor is an agreement not to not to enter the market and to cease its challenge of the patent in return for a substantial payment is strongly indicative of an anti-competitive motive where there is no other plausible explanation for the settlement. Such an agreement does not indicate a belief in the strength of the patent, instead it indicates a conscious decision to substitute the uncertainty of the possibility of the profits that would accrue to the generic manufacturer on entry into the market (the risks of competition) for an assured return achieved through co-operation and this constitutes a by object restriction. [11]  

The key factor in this assessment is one of general application, if the undertakings no longer determine independently their conduct but agree a concerted position, thereby substituting their co-operation for the risks of competition, then the agreement constitutes a ‘by object’ restriction. Whether the substitution of co-operation for competition was the sole motivation for the agreement is a matter for the national court to review.


[1] Opinion of AG Kokott of 22 January 2020, Generics (UK) Ltd and others v Competition and Markets Authority, C-307/18, EU:C:2020:28 (the Paroxetine Opinion).

[2] Judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraphs 98-99.

[3] Judgment of 1 July 2010, AstraZeneca v Commission, T-321/05, EU:T:2010:266, paragraph 362.

[4] Paroxetine Opinion, paragraph 84 and Judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraph 121 and Judgment of 12 December 2018, Servier and Others v Commission, T-691/14, EU:T:2018:922, paragraph 359.

[5] Judgment of 28 February 1991, Delimitis v Henninger Bräu, C-234/89, EU:C:1991:91, paragraph 21.

[6] Paroxetine Opinion, paragraph 59 and Judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraph 101.

[7] Judgment of 8 September 2016, Lundbeck v Commission, T-472/13, EU:T:2016:449, paragraph 144.

[8] Judgment of 26 November 2015, Maxima Latvija, C-345/14, EU:C:2015:784, paragraphs 18-23.

[9] Judgment of 4 June 2009, T-Mobile Netherlands and Others, C-8/08, EU:C:2009:343, and AG Kokott’s opinion in the same case of 19 February 2009, T-Mobile Netherlands and Others, C-8/08, EU:C:2009:110.

[10] Paroxetine Opinion, paragraph101.

[11] Paroxetine Opinion, paragraphs 117 and 118.