Competition between generics and originators – what’s the relevance of a patent?


The Opinion of Advocate General Kokott in the Paroxetine “Pay for delay” litigation (Case C-307/18, Opinion delivered on 22 January 2020).

The Paroxetine case arises from familiar circumstances and has a familiar cast of characters: a blockbuster drug; the expiry of a compound patent (and of data exclusivity); a variety of other relevant patents; existing or contemplated patent litigation by the patentee/originator against various generic manufacturers;  and a series of agreements relating to those disputes under which value was transferred by the originator to the various generics in various ways.

Unlike earlier cases in which the EU Commission and General Court have considered “pay-for-delay” settlements (such as Servier and Others v Commission (Case T‑691/14) and Lundbeck v Commission (Case T-472/13)) the Opinion responds to a request from the UK’s Competition Appeal Tribunal (‘CAT’) for guidance from the CJEU on the legal approach to “pay-for-delay” agreements under EU competition law. This procedural context has various effects:

  • Although it is a chunky 276 paragraphs, it is still much shorter than the GC’s judgments in Lundbeck and Servier.
  • It deals only with issues of law raised by the CAT – the facts are as found by the CAT.
  • AG Kokott is an experienced AG, with a reputation for a conservative approach (particularly in competition matters) and for thorough reasoning – the procedural context allows her to draw out a number of core principles in her opinion.

As would be expected from an Opinion responding to 10 questions (which have a number of sub-questions) AG Kokott deals with a wide range of legal issues. We discuss some of the most interesting parts of the Opinion for those involved in disputes between generics and originators. To keep it snappy, we mention the facts only where needed to understand the conclusion.

What’s the relevance of patents (and a patent dispute) to deciding whether an originator and a generic are competitors?

This is important because:

  • the competition rules apply only to sectors that are open to competition; and
  • where the competition rules do apply, agreements between competitors are much more likely to have an effect on competition than those between non-competitors.

It has been suggested that patents affect both of those considerations because, it could be argued:

  • that entry before a patent has expired, when neither invalidity nor non-infringement has been established, is illegal and the market is not ‘open to competition’;  and that
  • if the market is not open to competition and entry is presumptively illegal, a generic cannot be a competitor (actual or potential) on that market.

The AG’s  view is that whether an undertaking is a potential competitor on a market which is open to competition depends on two things:

  • are there insurmountable barriers to entry to the market in question; and
  • given the market context and the facts, is entry a concrete possibility?

The critical factors will include whether a generic has the objective ability to enter. The conclusion of an agreement which would prevent entry, and the originator’s subjective view of the possibility of entry, would be indicators of a competitive relationship between the generic and the originator.

The AG held that a dispute (whether or not it has yet resulted in litigation) is an indication that the parties are potential competitors while the commencement of proceedings may be an indicator that the generic is actually preparing to enter. However, the objective likelihood of success for either party in litigation is not relevant to the competition analysis. Patent law assessment is exclusively for national patent courts or authorities. The competition authorities must take account of the uncertainty inherent in that situation. The market reality is that the mere existence of patents does not prevent a competitive dynamic between a patentee and alleged infringer because:

  • it is impossible to know whether entry will infringe until a patent court has definitively ruled;
  • uncertainty is ‘a fundamental characteristic of competitive relationships in the pharmaceutical sector’;
  • a presumption that patents are valid does not mean that the entry of generic products could be assumed to be illegal;
  • the purpose of a patent is to ensure exclusivity but the specific subject matter of a patent does not extend to taking steps to prevent validity challenges; and therefore
  • the existence of patents over a product (or, to an even greater degree, a method of manufacturing  a product) does not create a legal barrier excluding all competition.

The AG lists several specific factors which are much more relevant to assessing whether the parties are competitors than any mini-trial of patent rights, which would be speculative:

  • validity challenges; ‘at risk launch’; and related legal proceedings often happen around the time of possible generic market entry;
  • it is not necessary to show non-infringement of patent rights to obtain an MA;
  • potential competition may start well before expiry of the compound patent;
  • process patents are not insurmountable barriers to entry as products produced using non-infringing processes may enter, irrespective of validity;
  • the patentee’s view of competitive threats from the generics is relevant, perhaps illustrated by willingness to transfer value in return (solely) for an agreement to delay entry;
  • the generic’s view of the likelihood of successful entry, its intentions and progress towards entry are also relevant.

Unsurprisingly, given her overall approach, the AG held that the fact that an interim injunction has been granted or an interim undertaking given does not preclude potential competition. As long as generic entry may still take place within a reasonable period, that is sufficient to amount to potential competition even if there is a temporary legal or contractual block, as was the case in the Alpharma and GUK cases. The AG noted that as interim measures are purely preliminary, they do not remove  uncertainty around the patent position.

In summary, the AG concluded that, in assessing whether parties are competitors and whether an agreement can affect competition, the patent context is largely irrelevant other than as part of the overall market dynamic in the pharmaceutical sector. The relevance of the patent litigation position consists largely in providing evidence that the originator regarded the generic as a competitive threat and was taking steps to remove that threat. Competition law does not require any assessment of the likelihood of success at any patent trial.

Pay-for-delay – a ‘by object’ restriction?

The Opinion contains a lengthy discussion of one of the current controversies in competition law – how to distinguish a ‘by object’ restriction of competition from a ‘by effect’ restriction. This is  important because, among other things, it affects the level of proof required to establish an infringement. We will come back to that general issue in later blogs or articles – it was also central in the Court of Appeal’s recent Judgment in PING.

From an IP/pharma perspective, the AG held that the agreements were capable of amounting to ‘by object’ restrictions. Key factors were:

  • a patent dispute with an uncertain outcome;
  • settled by agreement;
  • patentee transfers sufficient value to induce the generic to abandon entry efforts; and
  •  ‘sole’ consideration is abandonment of entry and undertaking not to challenge the patent.

The AG took a standard approach to identifying ‘by object’ infringements and noted that as the agreements were, effectively, arrangements under which one company paid another to exit the market, in principle they were agreements with an anticompetitive objective. However, various patent related issues had to be resolved before the AG could reach her conclusion on ‘by object’:

  • Can an agreement not to enter a market restrict competition when the obligations do not exceed the scope and duration of patent coverage?
  • Does the fact of patent uncertainty mean that it is impossible to determine the ‘counterfactual scenario’ that would have arisen in the absence of the agreement?
  • Can litigation settlements, which are encouraged by public policy constitute ‘by object’ infringements in situations where there is a genuine dispute between the parties?

Relevance of patent scope and duration

The AG concluded that even if restrictions in a settlement agreement did not exceed the temporal and technical scope of the patent, the agreement could still affect competition. There is no special exclusion for IP agreements from the competition rules and the exercise of IP rights (including through licensing or settling litigation) can be an abuse of those rights and affect competition. The AG held that this is entirely in line with the objectives of IP lawas reflected, for example, in recital 12 of the 2004 Enforcement Directive.

AG Kokott reiterated that while a patentee has the right to oppose infringements, a patentee’s exclusive rights do not protect it against validity challenges, therefore agreements which have that object fall outside the scope of protection of the patent. By entering into such an agreement, in return for a value transfer, a generic is not simply acknowledging the validity of the patent or reflecting its perception of the strength of the patent, but is agreeing to refrain from challenge and from entering the market because of that payment.

Given the nature of the pharmaceutical market, where significant price reductions follow generic entry, the fact that a value transfer to a generic is less that the profits that the generic might have expected from market entry does not affect the anti-competitive nature of the agreement. In essence, the generic replaces the uncertainty and effort of market entry with ‘the certainty of obtaining, without further effort, a significant part of the patent holder’s monopoly revenues’.

Relevance of patent uncertainty to the ‘counterfactual’

The AG disagreed with the suggestion that it is impossible to say that settlement agreements are capable of restricting competition simply because the impact of patents is uncertain. The correct question is whether the agreements substituted cooperation between the parties for the risks of competition rather than whether the generics would probably, very probably or definitely have entered in the absence of the agreements. The relevant counterfactual is a situation in which the parties continued to act independently, pursuing their own strategies, including in the context of patent litigation and market entry. The objective validity or infringement of the patent, or uncertainty about those matters, are not relevant to that consideration.

Relevance of nature of the agreement – settlements of litigation

The parties argued that because settlement agreements pursue legitimate objectives, and are encouraged by public policy, they cannot be restrictive of competition by object, save potentially in egregious cases such as where the settlement is a fraud or a sham.

The AG gave this argument short shrift, referring to previous case law of the CJEU holding that EU competition law does not distinguish between settlement agreements and other agreements. She noted that a policy to encourage settlement could not shield settlement agreements from competition law, which is also an aspect of public policy.

AG Kokott also dealt summarily with arguments that settlements would no longer be possible, noting that ‘settlements continue to be possible if they are genuinely intended to settle the disputes at issue and reflect a compromise between the parties reached following an independent assessment of the competitive situation by them’. The AG also pointed out that the particular agreements under review did not, technically, settle the litigation but deferred the dispute in time.

Other Article 101 issues

The AG’s opinion deals with a number of less IP focussed issues, including how benefits to consumers may be relevant to a competition analysis and how to analyse agreements which are argued to restrict competition by effect rather than (or as well as) by object. An analysis of those aspects of the opinion would take significant time, but headline points are:

  • applying Article 101(1) TFEU does not involve a ‘rule of reason’ assessment, weighing up pro- and anti-competitive aspects of an agreement. That balancing exercise happens only in the context of the exemption analysis under Article 101(3);
  • however, where an arrangement has pro- competitive objectives which can only be achieved through restrictions which are necessary to achieve the legitimate objectives; directly related to those objectives; and proportionate to that aim, those restrictions may be held to be ancillary to the pro-competitive purpose and not prohibited by Article 101(1). This does not involve a balancing exercise (see above);
  • the fact that an agreement may give rise to benefits is an aspect of its context which must be assessed when considering whether it restricts competition by object. The mere fact that an agreement may have benefits does not mean that it cannot be restrictive by object;
  • EU competition law protects the structure of competition as well as consumer interests. It is not necessary to identify a direct link between a restriction and consumer harm to find an anti-competitive object;.
  • When assessing the effects of an agreement, a similar approach to issues of patent law and uncertainty will be taken to that summarised above when assessing by object infringement.
Article 102

The Opinion deals with various questions under Article 102.

Market definition

The AG initially sets out the classic approach to market definition under EU competition law. She then expands that definition, explaining that market definition is a dynamic exercise, intended to help identify the competitive constraints faced by companies. Competitive pressure from those who have not yet entered the market may be relevant to market definition if those companies are able to enter the market sufficiently quickly and effectively and the prospect that they will do so exerts significant competitive pressure on incumbents. The AG emphasises the structure of supply and demand and actual competitive conditions as being relevant to the overall context for market definition. In the pharmaceutical sector considerations include:

  • once compound patents have expired, concerns about generic entry may exert significant competitive pressure, notwithstanding the existence of other patents;
  • the actual progress made by generics in preparing for market entry; and
  • evidence of the perception of the patentee about the extent of the competitive threat posed by the generics – and this could include the fact of willingness to transfer substantial value to the generics to stay off the market.


The CAT asked a number of specific questions on abuse. AG Kokott’s most interesting observations are summarised below:

  • Article 101 and 102 can be applied concurrently;
  • Article 102 prohibits a dominant company from eliminating a competitor and from strengthening its position otherwise than through ‘competition on the merits’;
  • An agreement which infringes Article 101 may also infringe Article 102 if it is capable of influencing the structure of the relevant market by hindering or eliminating competition;
  • A dominant company in the pharma sector which concludes a number of agreements settling patent disputes with generics will infringe Article 102 if:
    • the agreements seek to induce the generics to abandon efforts at independent market entry by a value transfer solely in return for market withdrawal; and
    • those agreements are capable of altering market structure so as to hinder or eliminate competition;
  • If benefits arising from the agreements can be identified, they must be taken into account. This is an objective exercise, irrespective of the subjective intent of the parties;
  • A dominant company may objectively justify its conduct by showing either that it is objectively necessary or that any exclusionary effects are counterbalanced by benefits. The test is similar to that under Article 101(3) and is unlikely to be satisfied where the agreements have the effect of eliminating effective competition by removing all or most sources of potential competition.

It remains to be seen to what extent the CJEU will endorse the views of Advocate General Kokott. In any event, given the careful reasoning it contains, the Opinion is likely to be of significant interest to both originator and generic companies and to competition lawyers more generally.  It provides much food for thought for those presently involved in or contemplating settlement. Key takeaways are:

  • The mere fact of a dispute between an originator and a generic may be sufficient to demonstrate that they are potential competitors so care should be taken when drafting licences or other agreements.
  • Conversely, the likelihood of success in litigation (for example, in proceedings about validity or whether a patent has been infringed) is irrelevant to whether the parties are potential competitors and whether an agreement can affect competition.
  • Settlements relating to the validity or infringement of a patent which provide for a value transfer from the originator to the generic manufacturer may infringe competition law.
  • Where agreements involve significant value transfers and the sole consideration given for that value transfer is the generic refraining from market entry and from challenging the patent during an agreed period a finding of by object infringement is very  likely.

Despite its detail, the Opinion leaves some open issues, in part because it is responding to the factual scenario in the national proceedings. For example, how might the analysis change if a value transfer is not made solely in return for market exit and no-challenge commitments, but reflects other relationships between the parties? The answer to that, and other outstanding issues on pay-for-delay which were not raised in the Paroxetine case, may need to await the CJEU Judgments in Servier and Lundbeck.

Those who are interested in the final outcome in the CJEU on Paroxetine need only wait until 30 January 2020, as it is understood that the CJEU will be giving judgment on that day.