In September 2016, the EU General Court (GC) issued its long-awaited decision in the Lundbeck case1 – the first ever European judgment on the legality of reverse payment patent settlement agreements.
Reverse payment patent settlement agreements are a feature of the pharmaceutical sector, where – in lieu of the risks of trial – a patent holder agrees to settle a patent dispute with a patent challenger with a settlement payment that provides the patent holder’s business with ongoing patent protection. Given the uncertainties of patent litigation, such settlements appear ostensibly sensible commercial arrangements, but have been characterised in a series of cases on both sides of the Atlantic as anticompetitive agreements which foreclose market entry (thereby assuming the settlement payment equates to a recognition of the patent’s invalidity).
The Lundbeck decision follows two years after the landmark US Supreme Court’s ruling on reverse payment patent settlement agreements in Actavis2, and less than a year after the UK CMA imposed significant fines on GSK and certain generic drug manufacturers3 in the context of similar settlement agreements.
It was hoped that the GC’s Lundbeck judgment would bring some clarity to the legal standard that applies to reverse payment patent settlement agreements in Europe, but unfortunately the GC has merely upheld the European Commission’s decision in its entirety – a decision which contains conflicting ideas about the interaction of competition law and intellectual property rights in the pharmaceutical sector.
The GC’s judgment now faces a further appeal to the EU’s Court of Justice (CJEU), but for companies involved in patent litigation who may be considering settling a dispute, the GC’s judgment is an important development and continues the trend of antitrust authorities applying competition law to contexts where IP rights would normally dictate that the right holder is legally protected from competition.
Summary of the facts
Lundbeck's compound patent for the citalopram molecule (used as an antidepressant medicine) had expired by the time the settlement agreements were entered into, but patents covering processes for the manufacture and crystallisation of citalopram remained in place as generic manufacturers started to enter the market. Lundbeck started patent infringement proceedings against certain generic companies (Generics UK, Alpharma, Arrow and Ranbaxy) which eventually led to a variety of settlement agreements. As part of these settlement agreements, the generic companies agreed: (i) not to market citalopram which infringed Lundbeck’s patents; and (ii) to sell existing citalopram stocks to Lundbeck, resell Lundbeck citalopram, or receive payments instead of damages or litigation costs.
The European Commission investigated these agreements and found them to be ‘market sharing’ agreements which restricted competition law ‘by object’, fining Lundbeck and the generic companies €146 million. It was important to the European Commission’s assessment that the value of payments made to the generic manufacturers corresponded to the expected profits of the generic companies following market entry: the Commission found that these payments ‘induced’ the generic company to abandon independent efforts to enter.
It is worth noting that the agreements in Lundbeck were not traditional settlement agreements in that they did not finally settle the litigation between the parties. They instead suspended both the litigation and the generic company’s entry for a limited period of time.
GC’s decision – points of significance
There were several points in the GC’s decision which were of particular interest, including:
- The GC’s decision to classify the infringing conduct as ‘by object’ (as opposed to under the more burdensome legal threshold of ‘by effect’);
- How the GC evaluated the existence of ‘potential competition’;
- How the GC considered the ‘transfer of value’ under the settlement agreement from Lundbeck to the generic companies.
We consider each of these points in further detail below.
‘By object’ restriction of competition
According to the GC, patent settlement agreements are a restriction of competition ‘by object’ i.e. they are by their very nature harmful to competition, irrespective of their effects on the market, where:
- they are made between actual or potential competitors;
- they contain a ‘value transfer’ from the patent holder to the patent challenger; and
- this ‘value transfer’ is in return for restrictions on the challenger company’s entry on the market (usually in the form of non-compete and / or no-challenge clauses).
The ‘by object’ classification is important in competition law cases as it removes the burden on the competition authority to establish in any precise detail the anticompetitive effects of the conduct in question. It has been a recurring theme in recent years for the Commission to favour the pursuit of cases under the ‘by object’ test, rather than apply the more onerous ‘by effect’ threshold, where the authority is required to spell out the anticompetitive effects of the arrangements in question. However, it is recognised that cases should not be taken forward under the ‘by object’ classification where they involve ‘novel’ infringements. The Commission’s trend of expanding the range of behaviours which were caught within the ‘by object’ box was recently noted by the CJEU in Cartes Bancaires4 where it set out the key elements of a restriction of competition ‘by object’:
- the concept of a restriction of competition ‘by object’ must be interpreted restrictively;
- it can only apply to conduct that, in itself, reveals a sufficient degree of harm to competition;
- in order to establish whether there is a sufficient degree of harm it is necessary to consider the relevant restrictions in their context, taking into account their content, objectives and the economic and legal context in which they occur.
The CJEU’s judgment in Cartes Bancaires has been characterised as a welcome check on a tendency, for procedural efficiency, for the Commission to push more challenging cases through the ‘by object’ channel. However, the GC’s Lundbeck decision does not appear aligned with the Cartes Bancaires trend, in that there are a number of complex issues – not least the validity of Lundbeck’s IP rights – which are skated over in this decision.
In view of the recent Cartes Bancaires judgment, an appeal to the CJEU in Lundbeck will inevitably focus on the correct legal standard to be applied to competition law consideration of reverse payment patent settlement agreements. The Lundbeck approach is certainly ‘novel’ when it is considered that restrictions of competition ‘by object’ are intended to be reserved for obviously anticompetitive arrangements, such as cartel-type conduct (e.g. price-fixing and market sharing), whereas patent settlements are generally regarded as mutually beneficial by preventing unnecessary litigation and freeing up court resources. Cases should only be classified as ‘by object’ where there is ‘existing experience’ of such agreements being harmful to consumers and society – this again is not obviously the case for patent settlement agreements. This was a factor which led the US Supreme Court in Actavis to decide that patent settlement agreements should be assessed using a ‘rule of reason’ approach (akin to a ‘by effect’ analysis in the EU) rather than a ‘per se’ approach (akin to a ‘by object’ analysis in the EU). However, in Lundbeck, the GC expressly rejected the US Supreme Court’s analysis – noting that US law is not binding in Europe.5
A particularly significant aspect of the Lundbeck case is the idea that the generic companies were potential competitors of Lundbeck regardless of the existence of a ‘blocking position’ (i.e. patent protection held by Lundbeck). However, the GC found the fact that the generic companies had possibilities for entering the market, including by launching ‘at risk’ of infringing Lundbeck’s patent, and that this was sufficient for them to be regarded as potential competitors.
In reaching the conclusion that launching ‘at risk’ was an expression of potential competition, the GC relied on three factors:
- Lundbeck’s compound patent had expired;
- there were other processes available to produce citalopram that were non-infringing; and
- the generics had taken steps and made investments to enter the market in competition with Lundbeck, including obtaining the active pharmaceutical ingredient (API), applying for a Marketing Authorisation (MA), and actively seeking customers for their generic products.
There is an uncomfortable contradiction in this analysis which necessarily assumes that Lundbeck’s patent would have been invalidated upon challenge, despite the GC also acknowledging that patents should be presumed valid once they have been granted. This goes to the heart of the concern patent holders will have that competition authorities might be willing to assume their patents are invalid on the basis of commercial decisions taken in the face of litigation, rather than through any objective analysis of the patent validity itself.
Indeed, in addition to the ‘blocking position’ created by Lundbeck’s patent, the generic companies in Lundbeck explained that there were a number of other reasons why they had not entered the market to compete with Lundbeck, including other regulatory and commercial barriers. For example, some of the generics had not obtained a MA. The GC did not give much weight to these arguments, noting that the very fact that Lundbeck had entered into settlement agreements with the generic companies indicated that it perceived those undertakings as a potential threat.
Another criticism of the GC’s approach in Lundbeck is with regard to the time period required for entry by a potential competitor. The traditional test for a potential competitor requires entry within a short period of time – however, the GC in Lundbeck indicated that competition could occur several years before expiry of the compound patent when generic producers that want to launch a generic product begin developments leading to a product that meets regulatory requirements. Effectively, this means a generic which is up to eight years from market entry could be considered a potential competitor – far longer than is typically considered for a company to be considered a ‘potential competitor’, and also at a point when there is little certainty as to the likelihood of successful entry.
There is a strong argument that potential competition cannot exist where market entry depends on infringement of an IP right – in these circumstances the generic company does not have the ability to enter the market. The fact that the right in question may be declared invalid at a later stage should not be a relevant consideration, because this is true for all patents.
It is no coincidence that the patents in question in Lundbeck, and also in Servier (the second reverse payment patent settlement case brought by the Commission) are process patents. The Commission has been clear in its view that once the compound patent for a molecule has expired, the market is in principle open to generic companies. It therefore appears relevant to the Commission’s analysis of patent settlement agreements whether the patent protects a new molecule rather than a new process. However, the implication of a hierarchy of patents is artificial and sets a potentially dangerous precedent for competition authorities.
The Commission’s view appears to be that medicines should be entitled to the standard period of patent protection, but that this should not be extended by means of process, secondary or formulation patents. However, this ignores the fact that medicines are often legally entitled to more than the standard period of patent protection. Patents can be extended by supplementary protection certificates (SPCs) (to compensate for the lengthy period of time which it takes to obtain a MA) and also paediatric extensions. Exclusivity can also be extended by various other means including regulatory and marketing protection, and market exclusivity for orphan medicinal products. These rights have been granted as an acknowledgement of the particular characteristics of the pharmaceutical sector, but the Commission’s approach to patents cannot be reconciled with this.
Transfer of value
A further aspect of the Lundbeck decision which is of interest is the treatment of the ‘value transfer’ – which was a critical element of the infringement finding. The generic companies had entered into a variety of agreements with Lundbeck. Some involved cash payments by Lundbeck, some involved the purchase and / or destruction of generic stocks, and others involved a distribution element (i.e. rights for the generics to distribute Lundbeck products). All of these were considered ‘value transfers’ by the GC.
In particular, the GC took issue with the size of the payments made by Lundbeck to the generics, concluding that the size of the payments were disproportionate and that this ‘induced’ the generic companies to enter into the settlement agreements, rather than pressing ahead with the patent litigation which could have led to their competitive entry.
However, this assessment discounts the commercial logic that can justify a ‘value transfer’ in a patent settlement agreement. In a patent dispute, views on validity and infringement may differ within organisations and over time. Patent cases are complex, even for patent judges, and there is an asymmetry of risk between the originator and generic given the huge investment associated with a patent protected medicine that has to be recouped over the patent period, as against the position of the generic which will have made minimal investment in reaching the stage of patent litigation. In other words, where the outcome of the litigation is uncertain (which will often be the case given the complexities involved), the originator has more at stake than the generic, and so has an incentive to settle at some cost, rather than risk the loss of the ability to recoup its investment in the patented product.
The consequence is that it can sometimes make sense to settle (in fact English procedural rules encourage parties to settle), and payments often bridge the gap due to the asymmetry of risk between the parties. Generic entry will lead to an almost immediate and irreversible downward spiral in the reimbursement price of the originator medicine, whereas the loss to the generic if the patent is upheld is relatively minor. In such circumstances it is not uncommon for the transfer of value to flow in reverse – i.e. from originator to generic.
While the GC appeared to accept that an ‘asymmetry of risk’ exists between the parties and that this can partly explain why the originator may make a payment to a generic in the context of a patent settlement, the GC ultimately rejects the argument that the value transfer might be justified on such grounds. The GC merely confirms the Commission’s reasoning that, there was significant uncertainty at the time the agreements were concluded, and that that uncertainty was eliminated and replaced by the certainty that the generic undertakings would not enter the market during the term of the settlement agreements, and that this had the effect of excluding generic competition. It does not acknowledge that this would be exactly the same position as if the patent were upheld in litigation.
The GC’s judgment in Lundbeck has created more questions than answers and it is inevitable that the parties in Lundbeck will appeal. In the meantime, the GC is likely to rule on the Commission’s second reverse patent settlement decision in the Servier case. It will be interesting to see whether the facts in Servier make any difference to the GC’s approach to reverse payment patent settlement agreements. For example, one of the key differences between the cases is that in Lundbeck, unlike in Servier, the settlement agreements did not finally resolve the litigation (Lundbeck’s agreements merely suspending the litigation).
The Lundbeck decision will, however, be welcomed by the UK’s Competition and Markets Authority (CMA) who in February of this year adopted a decision against GSK and a number of generic companies for entering into reverse payment patent settlement agreements. No doubt the CMA will rely on the GC’s decision in the pending appeal before the UK Competition Appeal Tribunal (CAT) which is due to be heard in February 2017. The CMA’s decision is evidence of a continued focus by competition authorities of these types of agreements.
Due to the inevitable appeals in both Lundbeck and Servier, it will be several years until pharmaceutical companies and their advisors can obtain any certainty on the legality of reverse payment patent settlement agreements. Indeed, the reality is that their assessment will be fact specific. However, in the meantime, there are some conclusions that can be drawn from the GC’s judgment.
- There are three basic criteria for the finding of a ‘by-object’ infringement of competition law: (1) significant ‘value transfer’ from originator to generic company; (2) restrictions on the generic company’s entry on the market; (3) the generic and originator company being actual or potential competitors.
- The size of the ‘value transfer’ matters. While a payment can be linked to the costs of litigation, if it is linked to the generic company’s anticipated profits post-entry, it is likely to be considered anti-competitive.
- The very attempt to conclude an agreement or to engage in discussions with a generic undertaking not yet present on the market provides a strong indication that that generic undertaking is a potential competitor (irrespective of any commercial, legal, and regulatory barriers to entry).
- There is a likelihood that once the compound patent has expired, the competition authorities may treat the market as in principle open to generic companies. Therefore a generic is more likely to be a potential competitor if the relevant patent is a process patent. The implication is that settling a patent dispute on a compound patent is much less likely to cause concern, but that care should be taken when settling a dispute on a process patent.
- The generic’s internal documents can be used to support the notion that the particular generic was a potential competitor. This would be the case, for example, where the generic company’s internal documents indicate that the generic would be prepared to enter ‘at risk’ and that it is taking steps to enter the market. On the other hand, any documents which indicate that the generic company considers the patent valid could be useful in proving the absence of potential competition.6
- Settlement agreements that are only temporary in nature and do not finally resolve the litigation will be looked at suspiciously.
Ultimately, the legality of any patent settlement agreement under competition law will need to be judged on its own facts. However, the points above can help provide a framework for the approach of both originators and generics who wish to best protect their position in the context of what is likely to be ongoing scrutiny of the sector by competition regulators in Europe, the US and beyond.