In Brief

In a judgment of 18 October 2023 (T-74/21), the EU’s General Court roundly dismissed the appeal by Teva and Cephalon against the European Commission’s 2020 decision imposing fines of approximately €30 million on each of them for concluding an anti-competitive patent settlement agreement back in 2005.  In some ways the case is an historic relic: since the US Supreme Court’s 2013 ruling in Actavis and the Commission’s decision a few days later in Lundbeck, it has been clear that any form of value transfer from the patent holder to a potential generic entrant in settlement of a patent dispute will be subject to close scrutiny and regulatory scepticism to the extent it goes beyond reasonable compensation of legal costs incurred.

Teva (which acquired Cephalon in 2012) reached a landmark settlement with the FTC back in 2015 in which it agreed to make $1.2 billion available to compensate purchasers who overpaid as the result of settlement agreements with four generic manufacturers in the US involving $300 million in payments for API supply and IP licenses that were found to make no economic sense except as payments to forego marketing for up to six years. Any leftover funds were to go to the US Treasury.  As part of that settlement, Teva was barred from entering into similar business deals within 30 days of, or expressly conditioned on, a patent litigation settlement restricting generic entry, without prejudice to Teva’s ability to enter into other types of settlements in which the value transferred “is unlikely to present antitrust concerns such as those providing payment for saved future litigation expenses (up to $7 million)“.

This latest EU judgment breaks no new ground, but is nonetheless a reminder of what to avoid, and the importance of documenting the legitimate reasons for engaging either in patent litigation or in commercial transactions with potential future generic competitors.


Cephalon’s compound patents for modafinil, the API for its sleep disorder drug Provigil, expired in 2003.  Cephalon relied on secondary patents, valid until 2015, to litigate against a number of generic entrants applying for marketing authorisations in Europe.  Teva launched its generic product in the UK in 2005.  During injunctive relief proceedings, it agreed to stop sales in exchange for Cephalon posting a bond to compensate Teva in the event it succeeded in the main proceedings.  Shortly thereafter, the parties reached a settlement including non-compete and non-challenge restraints, whereby Teva agreed not to enter the market before 2012, three years before patent expiry.  In return, Teva: licensed Cephalon its IP rights; agreed to supply Cephalon with its API; was appointed Cephalon’s exclusive distributor in the UK; accepted payments from Cephalon for avoided litigation costs; and took a separate license to certain data co-developed by Cephalon in connection with studies on the treatment of Parkinson’s disease. 

Restriction of Competition “by Object”

The parties argued that each commercial element of the settlement package had a plausible explanation other than market-sharing and that Teva’s early entry date was pro-competitive.  On this basis, they claimed that the Commission had misapplied the two-part test laid down in the 2020 Generics UK ruling (C-307/18, appealing the UK paroxetine decision).  In that case, the European Court of Justice held that a settlement involving transfers of value constitutes a hardcore restriction of competition “by object”, if: (i) the value transfer has no other explanation than the commercial interests of the parties not to compete on the merits, and (ii) the agreement has no proven pro-competitive effects capable of giving rise to a reasonable doubt that the settlement causes a sufficient degree of harm.  That ruling also established that there is no requirement that the net gain transferring should necessarily be greater than the profits which the potential generic entrant would have made had it been successful in the patent litigation proceedings.

(i) The value transfer has no other explanation than an agreement not to compete on the merits

The Commission is required to ascertain whether the elements making up the settlement package would have been concluded on equally favourable terms absent the non-compete and non-challenge clauses.  If the answer is negative, it follows that the only plausible explanation is that the commercial arrangements were simply an inducement for the potential generic entrant to stay off the market, and thereby a restriction of competition by object.

The General Court restated the ruling in Generics that, in making its assessment, the Commission can only rely on legal and economic elements known to the parties at the time of the settlement, to the exclusion of elements arising subsequently. 

The Court considered the Commission’s evaluation of each element of the settlement package:

  • It pointed to contemporaneous evidence that Cephalon had not felt particularly threatened by Teva’s patents and had not done any due diligence on their value before it agreed to pay $125 million for a non-exclusive license to them that was of little or no value to it.
  • It noted that the API supply agreement guaranteed Teva minimum volumes with a 30% margin for five years which the Commission viewed as not being economically rational but for the settlement.  Cephalon faced no long-term supply constraints and had the capacity to meet projected demand.  The prices it paid to Teva were 100-300% higher than the prices paid to another third-party supplier and to Cephalon’s own internal transfer price.  The Commission was therefore entitled to view the supply agreement as an inducement for Teva to stay off the market.
  • The UK distribution appointment netted Teva a one-off €2.5 million payment and a guaranteed 20% distribution margin to the value of €8 million annually which, the Court agreed, it would not have obtained under normal market conditions. 
  • Cephalon paid Teva €3.07 million to end the UK litigation, and €2.5 million to prevent future patent or other litigation in markets outside the UK and US.  The Commission took the view that the total of €5.57 million did not compensate for litigation costs actually incurred by Teva, nor was it based on any estimate of litigation costs avoided by Cephalon.  It corresponded to an amount calculated on the basis of Teva’s anticipated modafinil sales in the UK.  The Court ruled that only compensation paid with respect to actual litigation or other costs incurred may be regarded as justified and therefore not constituting reverse payments.  It did not rule out that payments for future litigation costs could be justified in certain (unspecified) cases, but it is for the parties to provide evidence to substantiate those costs.
  • The grant by Cephalon of a license for clinical and safety data for the treatment of Parkinson’s for $1 million was deemed to be of high value to Teva allowing it to accelerate the commercial launch of its product Azilect.  Contemporaneous Teva documents indicated that delayed access to the data could potentially have cost Teva approximately $200 million in lost revenues.  On this basis, the Court found that the Commission was entitled to conclude that the data license contributed to inducing Teva to agree to the settlement. 

The Court concluded that the parties had sought to find a combination of transactions representing a certain overall value that was sufficient inducement for Teva to agree not to compete, qualifying the settlement package as a restriction of competition “by object”.

(ii) No proven pro-competitive effects

The Court agreed with the Commission that the settlement agreement was not pro-competitive just because it allowed Teva early entry at least three years before Cephalon’s particle size patents expired, compared to the scenario in which Teva did not win the patent infringement proceedings.  Citing Lundbeck (Case C-591/16), the Court ruled that it is not necessary for the Commission to examine scenarios such as where one or other party is successful in a patent dispute; it is sufficient to establish that the conduct reveals a sufficient degree of harm to competition in the specific economic and legal context.  Prior to the settlement, Teva was Cephalon’s most advanced potential competitor with concrete possibilities to enter the market in 2005, a full seven years before the contractually agreed “early entry”.  Even that entry was “delayed, controlled and limited”, being predicated on Teva paying Cephalon royalties of 10-20% of net profits, and given Cephalon’s strategy to switch the market from Provigil to its second-generation product, Nuvigil, to offset the expiry of its patents and counter generic competition.

Restriction of Competition by Effect

The parties argued that the Commission had erred in only assessing the potential effects of the settlement on competition, and not the actual effects given that the agreements had been implemented.  They asserted that the burden was on the Commission to demonstrate a negative impact on prices, output, innovation, variety, or quality that would have ensued compared to the situation where the parties would have continued to litigate.

Despite this plea being moot given the determination that the settlement was a restriction of competition “by object”, the Court nonetheless found it appropriate to address the point.  It reiterated that the Generics (UK) case law means that restrictive effects on competition may be both real and potential but, in both cases, they must be sufficiently appreciable.  It is therefore possible for the Commission to rely solely on the potential competition represented by a new entrant eliminated by the settlement agreement and on the structure of the related market.  The establishment of the counterfactual need not involve any definitive finding in relation to each party’s chances of success or on the probability of them concluding a less restrictive agreement; these are just some factors amongst many to be considered when determining the realistic possibilities of what the generic manufacturer would have done absent the settlement.

Since Teva’s view at the time was that its product was non-infringing and Cephalon’s patents were invalid, the Commission was entitled to conclude that the settlement was not the result of a genuine assessment based on the perceived strength of the disputed patents, but an inducement not to compete. 

Legal Certainty and Legitimate Expectations

The parties’ arguments that there was no precedent back in 2005 suggesting that reverse-payment settlements were unlawful were rejected.  The Court, citing Lundbeck, restated that the principle of legal certainty cannot be interpreted as prohibiting the gradual clarification of the rules on criminal liability by means of evolving case law, provided those interpretations are reasonably foreseeable.  The settlement, which temporarily excluded Teva from the market, constituted an extreme form of market-sharing. Contrary to the parties’ arguments, the prevailing view of US courts at the time was not unanimous and was in any event irrelevant, as was the fact that, at the time, the Commission had not yet imposed fines for similar conduct. 

The principle of legitimate expectations was not available since the parties had not met the legal test which requires them to have received precise, unconditional, and consistent assurances from authorised, reliable institutional sources which was clearly not the case.

Challenges to the fining methodology were also rejected.  The Court upheld the Commission deviating from prior practice in which it took the transfer value as the relevant basis for the fine calculation.  (In reverse-payment settlement cases, it is not possible to apply the usual methodology basing the fine on a percentage of affected revenue since the generic necessarily foregoes revenues in settling.)  In the present case, a concrete value could not be attributed to each element of the settlement package, and the Commission was therefore entitled to fix Teva’s fine by reference to the fine imposed on Cephalon, not least since the gravity and duration of the infringement were the same for both.


Teva has the opportunity to appeal to the EU Court of Justice.  However, the intervening case law from the Court of Justice discussed above, means that the prospect of changing the outcome is remote.


Fiona Carlin is the head of the EU Competition & Regulatory Affairs practice in Brussels.