With typical rapidity (ahem), the full text of the Commission merger decision in Teva/Cephalon has been published two years and four months after it was notified to the parties.
While the main decision – authorisation of the merger subject to divestment of Cephalon’s generic version of its Provigil (modafinil) drug, treating narcolepsy – is old news, the Commission’s reasoning contains some interesting indications of its policy in the pharma field, not least in relation to patent settlement agreements.
Looking first at the market definition analysis, the decision is typical in that it does not reach a firm conclusion on any of the markets where potential overlaps were identified and even for the divestment product the Commission concludes that competition concerns arose even on the widest market identified by the parties. There are, however, a number of indications about the Commission’s thinking, and those indications exhibit a preference for narrow markets over a more standard (in the pharma context) ATC3 analysis. Given that this tie-up was between a (mostly) generic company and an originator, this is not especially surprising – the Commission’s practice in such cases is to look for all possible markets where the parties may have substitutable products. Whereas in a merger between innovators, ATC3 or 4 would be most likely to produce overlaps, here, narrower markets and higher market shares are more likely to indicate potential competition concerns. In this case, as a few years ago in Sanofi-Aventis/Zentiva, the Commission notes that generic products are usually the closest substitute for those of originators as they are specifically designed to compete with the originator’s medicine. Even if a relevant market may be wider than just the molecule, “the competitive constraints stemming from the generic version [of the relevant molecule] would […] be significantly stronger than any […] competitive pressure from other molecules under any market definition”.
More novel is the suggestion that the narrowest possible relevant market may be even narrower than the pharmaceutical molecule. This could be the case where a particular dosage, pharmaceutical form or route of administration is used – the Commission refers to “New Form Codes” used by IMS and EphMRA which distinguish between e.g. topical/systemic administration and between long-acting and ordinary forms. Anyone want to bet against such narrow market definitions being adopted in an abuse of dominance case?
Regardless of the introductory paragraphs, the only overlap which is analysed in depth in the decision is in relation to modafinil, where Teva had a pipeline generic product. Indeed, the role played by potential competition (as is appropriate for pipeline products) is possibly the most interesting part of the decision. The outcome of the Commission’s analysis is that Teva, a company which did not have its own product on the market, was to be regarded as a potential competitor, whereas other generic companies, which actually marketed generic products at the time, were not to be counted as actual or potential competitors. The reason for this counterintuitive finding, at least so far as the third party generics are concerned, is the existence of a Cephalon patent against which the generic companies had entered at risk. Due to the continued patent protection, and the fact that Cephalon had asserted the patent against all of the existing entrants, the decision notes that the generics’ access to the market was not guaranteed to be “sustainable” and proceeds on the basis that these market players should be disregarded.
While this is expressed as a “conservative” basis on which to proceed, it had the significant consequence of requiring a divestment remedy on the part of Cephalon (although it could be assumed that withdrawing the patent litigation would have had the same outcome…*). It also assumes the very opposite to what appears to have been concluded in the recent Lundbeck patent settlement decision (so far as can be judged from the details of the appeals and from the Commission press release – someone please assure me that it isn’t going to be over two years until this is made public?!), where the generic companies are treated as potential competitors of Lundbeck even in cases where they had not yet entered the market. Here, by contrast, it is assumed that Cephalon would have gone on to prove that its patent (protecting the particle size of modafinil) was valid and infringed by the generics. This is a significant assumption, particularly given that the UK designation of the patent had, at the time of the decision, been revoked by the English High Court, and Cephalon had failed to obtain interim injunctive relief in cases in other jurisdictions.
Leaving this discrepancy aside, the approach to potential competition in the merger decision appears quite reasonable, with three key issues identified as determinative of the existence of potential competition at the level of generics:
1. Has a generic version of the originator drug been developed?
2. Has a marketing authorisation (MA) been granted?
3. Does the generic have the right continuously to market its product in the period up to patent expiry – i.e., is the product blocked by an existing patent?
As regards the conclusion that Teva was a potential competitor despite not marketing the product, there is some interesting discussion of the difference in competitive terms between having an application for a first MA and having a lapsed MA. Where there is actual competition from other generics the Commission’s conclusion seems to be that there is not much difference between the two situations (as the existing competitors can quickly mop up market share previously held by the exiting company), but in circumstances where there is no other competition, a lapsed MA is likely to allow a company to come (back) to market more rapidly than where a new MA has to be applied for.
Attentive readers will wonder why Cephalon’s patent was not regarded as liable to exclude Teva also. The answer is in the link with the Commission’s ongoing antitrust investigation. The merger decision describes what is assumed to be the factual basis for this investigation (at paragraph 95). It appears that in 2005, Teva entered at risk with a generic of Cephalon’s Provigil, in response to which Cephalon applied for a PI under its “particle size” patent. Following a settlement agreement, Teva agreed to postpone entry until 3 years prior to patent expiry (or earlier in the event of the launch of another generic), and was also appointed as the exclusive distributor of Cephalon’s modafinil products. In line with the concerns expressed about early entry agreements in the 4th patent settlement monitoring report (see discussion here), it appears that the Commission is indeed investigating such an agreement. How it proposes to find an infringement (which presumably relies on Teva being a potential competitor at the time when the agreement was entered into) without contradicting its conclusions in this merger decision (which excludes a finding of potential competition if the market entry could prove to be unsustainable) remains to be seen.
* This might of course have resulted in an “invalid” patent remaining on the register, another of the Commission’s pet hates. Expect more from us on this topic soon…