(…and four other takeaways from the Cephalon decision)
Although the European Commission adopted the latest of its pay-for-delay decisions in November 2020, the Cephalon decision was only published in July of this year. With the Lundbeck agreements first coming under Commission scrutiny in 2003, pay-for-delay enforcement turned 18 this year. Below we set out a short recap of the latest case and our five key takeaways from this case and Europe’s record with respect to these agreements.
Modafinil is intended for patients who suffer from excessive daytime sleepiness (EDS). It was first sold under its brand name Provigil, which was the most important product for Cephalon between 2001 and 2010 and became a blockbuster drug in 2009 when sales surpassed $1 billion.
By 2005, Cephalon held several secondary patents related to modafinil in various countries, but the primary patent had expired. Its internal documents showed that it estimated the chances of its secondary particle size patent being upheld to be around 50% in the US and UK. (Indeed, the patent was found to be invalid in 2011. Although Cephalon appealed, the appeal was settled).
Teva obtained a marketing authorisation (MA) for a generic modafinil product for the UK in June 2005 and launched its product in the same month at 50% off Cephalon’s list price. It expected to win more than 80% of the UK market with its pricing strategy. Teva applied for MAs in 14 other EEA countries, using the mutual recognition procedure.
Cephalon immediately sued Teva for patent infringement in the High Court in July 2005. However, in December 2005 the parties concluded an agreement to settle the litigation. Under the agreement:
- Teva would not enter with its own generic product into the markets in the UK, US, or any other country where Cephalon held modafinil patent rights. Teva was granted a non-exclusive right to manufacture and market generic modafinil from 2012 onwards.
- Cephalon paid sums of £2.1 million and €2.5 million to Teva purportedly in recognition of the litigation costs avoided in the UK and in “European and other markets” than the US and UK.
- The parties concluded several commercial arrangements, all beneficial to Teva. For example, Cephalon purchased a licence to Teva’s modafinil-related IP rights in exchange for royalty payments totalling $125 million and it made Teva its UK distributor for five years with a distribution margin of 20%, and a one-off payment of €2.5 million.
The settlement soon attracted antitrust complaints. A 2006 US private action was settled in 2015, with Teva (which had by then acquired Cephalon) paying the plaintiffs $512 million. The US FTC sued Cephalon in 2008, which Teva settled in 2015 as well. Although it made no admission of wrongdoing, Teva made USD 1.2 billion available to compensate businesses who had overpaid for modafinil.
The Commission started its investigation in December 2009 and adopted its infringement decision in November 2020, imposing a penalty of €30,480,000 on Cephalon and €30,000,000 on Teva.
According to the Commission, Teva was an actual competitor of Cephalon in the UK, and a potential competitor in 27 other EEA countries. Although Teva had only applied for MAs in 14 countries, it could have relied on the mutual recognition procedure to obtain MAs in all other EEA states. Therefore, it was a potential competitor in all 27 countries.
Next, the Commission established that Teva had accepted non-compete and non-challenge obligations that were extensive in scope, geographical reach, and duration (around 7 years in most EEA countries). The Commission then assessed the alleged consideration for these non-challenge and non-compete commitments (the ‘pay’ for the delay). It found that the package of transactions in the agreement resulted in a significant transfer of value to Teva. The parties’ alternative explanations were not plausible, and without the aim of inducing Teva to refrain from competing, Cephalon would not have entered into the transactions.
As a final step, the Commission held that the fact that Teva was allowed to enter the market in 2012 under Cephalon’s licence was not pro-competitive, inter alia because this would only lead to delayed and controlled entry and because Cephalon’s strategy to product-hop to a new form of modafinil (called Nuvigil) would have undermined any pro-competitive effects.
The agreement therefore restricted competition ‘by object’ in the 28 EEA states.
The Commission also assessed whether the agreement restricted competition ‘by effect’ in France, Germany, the Netherlands, Spain, Sweden, and the UK. An effects analysis requires an assessment of the agreement against the ‘counterfactual’, i.e. the world absent the agreement. The Commission concluded that at the time of the agreement, Teva constituted the most significant competitive threat for Cephalon. Without the agreement, this threat would have been maintained; and if entry had occurred, a price decrease would have been likely.
The Commission then analysed the effects of the agreement against this counterfactual. It did so by analysing Cephalon’s market power and ability of extracting rents, the impact of the agreement on other generic manufacturers’ entry plans, and its impact on the likelihood of lower prices. For illustrative purposes, the Commission demonstrated that the price difference between pre- and post-generic entry ranged from 8% in Germany to 88% in the UK. The Commission found that Cephalon could continue to extract significant market power rents above the competitive level after the agreement, since it was not constrained by generic competitors.
The Commission concluded that the agreement eliminated Teva as a potential competitor and thus appreciably restricted competition between the parties and that absent the agreement, Teva would have continued to try to enter and compete with Cephalon. The agreement increased the likelihood that Cephalon’s position on the markets remained uncontested for longer. The agreement replaced the risk of competition and market entry by Teva with the certainty of a significant value transfer to it. Absent the agreement, Teva would have remained a competitive threat. Entry would have had the likely effect to decrease prices. The agreement preserved Cephalon’s market power, allowed it to maintain its significant rents (and the resulting prices) to the detriment of consumers and health systems, and deterred all other challengers from entering the market.
Five key takeaways
- Payment for non-entry can come in many forms
The term ‘pay-for-delay’ captures accurately the alleged mischief in these cases: according to the Commission, the incumbent sole supplier of the drug pays the potential entrant to delay its entry. However, the payment element of the agreement is not always obvious. This is because it appears counter-intuitive for the incumbent (who is alleging a patent infringement) to be paying the potential entrant (or would-be patent infringer), hence the term ‘reverse payment’ which is also used. The payments moreover tend to come in different shapes and forms and tend not to be explicitly referred to as compensation for the non-entry commitment provided by the potential entrant. This is clear from the previous cases:
- In the Commission’s Fentanyl case, the parties had ostensibly agreed that the incumbent would pay the potential entrant for promotional activities. However, the payments considerably exceeded the value of the activities that were actually carried out.
- In the CMA’s Paroxetine case and the Commission’s Lundbeck and Servier cases, there were cash payments but also supplies of fixed quantities of the incumbent’s product to the potential entrant at low prices, for the potential entrant to resell in the market at prevailing and much higher market rates. The transfer of margin in this way can in itself be a form of payment.
The Cephalon case represents a complex set of alleged payments to Teva as the potential entrant, consisting of royal payments to Teva totalling $125 million for a licence that Cephalon does not appear to have needed; the ability for Teva to use clinical and safety data relating to an unrelated drug, which the Commission claims would not have been made available to Teva absent the settlement; the appointment of Teva as Cephalon’s supplier of active pharmaceutical ingredient for modafinil at fixed prices for a period of five years; and its appointment as Cephalon’s distributor for the UK for five years with a distribution margin of 20%, and a one-off payment of €2.5 million.
The Commission found that all individual transactions contemplated under the agreement, although in principle unrelated to each other, were negotiated at the same time and in an interrelated manner with a view of reaching a comprehensive package deal of a certain value satisfactory to Teva. This led to a value transfer that, irrespective of its exact quantification, was sufficiently beneficial to induce Teva to accept the commitment not to independently enter and compete in the markets for modafinil.
If the Commission’s decision is upheld on appeal, the clear lesson is that commercial arrangements entered into in the context of the settlement of patent litigation will be approached with suspicion, in particular where they do not appear to have any other commercial rationale than to induce the acceptance of a commitment to drop efforts to enter the market. It is not necessary for a value transfer to take the form of a cash payment, consideration can be given effect by other means.
There is an interesting parallel with a case in the US involving Humira, AbbVie’s blockbuster biological drug. In the case, which was dismissed at first instance, but is currently under appeal, one of the allegations by the claimants is that the biosimilar entrants were induced not to enter the US market by letting them enter European markets unopposed. In other words, the ‘payment’ to the biosimilar manufacturers for delaying their entry onto the US market was to let them enter the European market. The FTC filed an amicus brief in the case in October 2020 in which it opined that the lower court had incorrectly applied Actavis, the seminal US reverse payments case. In particular, the district court “did not fully consider whether the plaintiffs plausibly alleged that the European settlements served as a payment vehicle to induce the biosimilars to abandon their U.S. patent litigation and accept a deferred entry date”, according to the FTC. It is not clear yet when the Court of Appeals for the Seventh Circuit will hand down its judgment.
2. Belt and braces: Commission continues to find both object and effect agreement
A second point of interest is that even though by the time the Commission adopted its decision the General Court had already upheld its characterisation of pay-for-delay agreements as ‘by object’ infringements, and the Court of Justice had done the same in its preliminary ruling in relation to the CMA’s Paroxetine case, the Commission continued with its practice of finding both by object and by effect infringements in this latest pay-for-delay decision.
Now that the appeal risk against the characterisation of pay-for-delay agreements as ‘by object’ infringements has been settled (it was confirmed again by the Court of Justice in Lundbeck after this decision was adopted), it is questionable what the added value of including an effects analysis is.
The difficulty with respect to effects cases involving potential competition, in particular where the potential competitor must overcome patent litigation of which the outcome is inherently uncertain, is that it is not so easy to determine what the counterfactual looks like. The competition authority has to make a call on whether or not it wants to get into the likelihood of success in that litigation. This ultimately boils down to whether the effects of the agreement are measured on the basis of actual competition (Teva will enter the market in the counterfactual) or potential competition (Teva will continue to have real and concrete possibilities to enter in the counterfactual).
It would go too far to require the authority to hold that absent the agreement the generic would definitely have entered for the agreement to have the effect of restricting competition. Setting such a standard would confuse the concepts of actual and potential competition, both of which merit protection. It should be sufficient that entry remained a real and concrete possibility and that this possibility was removed by the agreement. It is that removal of the possibility of actual entry (and the beneficial impact of that entry on competition and prices) that forms the anti-competitive harm of a pay-for-delay agreement. However, this is sufficiently covered by the object case, so it is not clear why an effects analysis was carried out for the six countries mentioned.
3. Settled law but no settlement
Further, it is notable that the US antitrust claims against the agreement between Teva and Cephalon were all settled by 2015 (and with considerably more significant financial exposure for Teva), whereas it took five more years for the Commission to adopt its decision, which is now being appealed in the General Court (Case T-74/21), with no hearing date planned yet.
Of course, the possibilities for Teva to settle the case were limited. The Commission only has a formal settlement procedure for cartel cases. While it has given discounts in other antitrust cases to companies who cooperate to bring the case to an early resolution (see this factsheet), this policy has not been very popular and was evidently not utilised in this pay-for-delay case.
In addition, as we point out in the next point, the main victims of the agreement in the EEA are national health systems and health insurers. To claim compensation for any harm caused by the agreement, these organisations will now need to wait until the General Court has ruled on Teva’s appeal and possibly even until the higher Court of Justice has ruled on a further appeal before they stand any chance at winning a claim for damages. There is a decent chance that this will only be in 2025 or so, i.e. 20 years after the agreement was concluded.
Although US antitrust law has recently been compared unfavourably to EU competition law, the outcome in the US, with two large settlements involving compensation of harm, do not look like such a bad outcome for the FTC, the relevant claimants, and Teva, which was able to put the proceedings to bed in 2015.
The system should really work better, both for the companies involved in investigations and for victims. It would be interesting to keep an eye on developments in the US, where Congress passed a bill authorising the FTC to seek monetary relief in federal court from businesses that engage inter alia in anticompetitive conduct. The FTC asked for this after the Supreme Court held that existing laws do not authorise the FTC to seek such relief, leading it to withdraw its $500 million recovery case against AbbVie (discussed in our September post on The Thicket). In cases involving direct harm to health authorities or consumers, would a similar system not be beneficial in the EU?
4. MIA: European private actions
Anyone who follows developments in competition law can see private actions for damages popping up left, right, and centre, some of which are pretty ambitious to say the least. However, with the exception of the UK authorities who have so far brought cases in relation to the Lundbeck and Servier pay-for-delay cases (see our blog post from September on the latter), private enforcement cases in relation to pay-for-delay seem to be missing in action in Europe.
There is a huge contrast with the US, where there is a considerable practice of reverse payment cases being brought in the Courts, even though the Supreme Court’s ruling that reverse payments should be reviewed under the ‘rule of reason’ test created complexity for District Courts hearing these cases.
It is possible that in those Member States where prescription drugs are funded by the State, there is still an assumption that competition law is primarily a matter for public enforcement. It may also be that public health authorities simply do not have the resources to take up litigation to recover damages, simply because they are overstretched. In Member States with a private insurance model, like the Netherlands, it may be that the damage per insurer is relatively small, or the insurers may be concerned about the issue of pass-on. It may also be a problem more generally that no public or private entity is able to claim damages across the EU, such that damage suffered in individual Member States may be relatively small. Finally, it may be that health authorities are reluctant to resort to measures like assigning claims to private sector claim vehicles, even though this may be an effective way to outsource litigation for overstretched authorities, providing the relevant national law allows for such assignment.
Ultimately, damages are currently being left unaddressed, and one is left to wonder if the system created by the Damages Directive is also effective for damages caused in European healthcare markets.
This leaves the UK, where on previous form it can be expected that the Secretary of State for Health and Social Care will launch a follow-on claim for damages. The law is clear that follow-on claims can still be brought in relation to European Commission decisions, provided the decision was adopted before the end of the Transition Period, which is the case for this decision.
5. Pay-for-delay hits 18 in the EU. Adulthood, but what comes next?
The Cephalon decision was the last in a series of pay-for-delay cases brought by the Commission. It first became aware of the pay-for-delay agreements that were the subject of the Lundbeck case in October 2003 and since then it has adopted four major decisions: Fentanyl, Lundbeck, Servier (perindopril) and now Cephalon.
So far it has been successful under Article 101, except for the Krka case, which it is appealing. Under Article 102, the Commission suffered a setback when the General Court annulled its findings that Servier’s agreements not only infringed Article 101, but also, combined with Servier’s “patent acquisition practices”, Article 102, on the basis that the Commission’s market definition was flawed. However, from the Court of Justice’s preliminary ruling in Paroxetine it is clear that a dominant undertaking’s strategy to conclude a number of agreements that delay entry by potential competitors can form an abuse of dominance, provided the strategy has the capacity to restrict competition and, in particular, to have exclusionary effects, going beyond the specific anticompetitive effects of each of the settlement agreements.
For cases as complex as the pay-for-delay cases were, the Commission’s record has been good. With the benefit of hindsight and knowing the EU Courts’ judgments, the main criticism one can have is the length of the proceedings. The neutral observer may however be concerned about the lack of cases in the pipeline, both at the EU and – as far as we are aware – at the national level. Experience from the US suggests that, while the increased scrutiny of pay-for-delay agreements has certainly had a big impact, such agreements are still being concluded.
Meanwhile, new ground is being broken in relation to other practices that may be seen around the time of expiry of a product’s main patent. For example, as we discussed last month, the NHS’s case against Servier is not only a follow-on claim for damages based on the Commission’s decision, it also claims that Servier made misleading or dishonest misrepresentations to the European Patent Office (EPO) in obtaining a secondary patent and that this constitutes an abuse of a dominant position. In the EU, the Commission announced that it is investigating whether Teva may have artificially extended the market exclusivity of Copaxone after the initial patent expired by strategically filing and withdrawing divisional patents, repeatedly delaying entry of its generic competitor who was obliged to file a new legal challenge each time. The Commission will also examine whether Teva may have pursued a communication campaign to unduly hinder the use of competing products. The Commission’s investigation is under Article 102.
Pharma remains a sector that is in the spotlight, so it will be important for companies involved in patent litigation to take care when concluding settlement agreements. The appetite for new pay-for-delay cases is unlikely to have diminished.
The author is a partner at Geradin Partners and previously worked at the Competition and Markets Authority. The views in this post are the author’s personal views.