As I mentioned last week, the European Commission has now published a "provisional non-confidential version" of its 565-page decision in Case AT.40588—Teva Copaxone. The decision finds that Teva abused its dominant position in the market for glatiramer acetate (GA) within several E.U. countries. GA is the active ingredient in Copaxone, an immunomodulator drug used for treating relapsing remitting multiple sclerosis. According to the E.C., the abuse consisted of the combination of two strategies: the misuse of divisional applications and the disparagement of a rival generic drug. The portion of the decision addressing the misuse of divisionals is particularly relevant to my forthcoming book, Wrongful Patent Assertion: A Comparative Law and Economics Analysis, and so I will provide a few (very preliminary) thoughts about it here.
By way
of background, under the European Patent Convention, a patent applicant can
file a divisional application either “to remedy a lack of unity of invention
disclosed in an earlier application,” para. 47, or (more commonly) on a voluntary
basis to cover “different aspects of the same invention, such as distinct
therapeutic uses, different production processes, or dosage regimes” (para. 48). An applicant may file a divisional “if it
relates to an earlier pending European patent application, be it the original
parent patent application or an earlier divisional application,” and only “up
to the date on which its (direct) parent patent is granted by the EPO. However, it is possible to file multiple
generations of divisionals on the basis of an original parent patent”
(para.49). In other words, while a
parent application is pending, you can file a divisional, and regardless of
what happens to the parent the divisional is entitled to its own independent
examination; and while any given divisional is pending, you can file another divisional
based on that application (regardless of whether the initial parent
application remains pending), which will be entitled to its own independent examination,
and so on. If granted and not revoked, however,
a divisional will expire the same day as its parent. Thus, “by continuously filing divisional
applications, it is possible to preserve the patent family pending throughout
the initial exclusivity period of the original parent patent (even if the
original parent patent has already been granted, revoked or withdrawn) and
thereby extend the uncertainty as to the validity of relevant patents for
potential or actual competitors” (para. 52).
Further, a divisional “may not claim any additional subject-matter to
that disclosed in that parent application,” but “it is possible to file a
divisional application, which has a scope that is very similar to the parent
application (that is, with nearly identical claims),” as long as the subject
matter claimed is not identical (paras. 53-55).
In the present case, Teva obtained its “main patent protecting the composition of GA within specific molecular weight range” in 2002, and that patent expired in 2015 (para. 101). The divisionals at issue, however, were derived from two subsequent families of secondary patents. In 2005, Teva filed a parent patent claiming a production process for GA, from which it later split off two divisional patents; and in 2010, a parent “dosage” patent, relating to a 40 mg formulation of the drug, from which it split off four divisional patents (paras. 102-03; see also tables 13 and 14, pp. 250-52). (The dosage patent applications also included one, the so-called “ghost divisional,” that was withdrawn before grant, see para. 1137. An opposition to the last of the granted dosage divisionals was still pending as of the date of the decision, see p.252.) Altogether, Teva filed the divisionals over a period of eight years, from 2010 to 2018, but “[t]he intervals between the filings were not driven by any ongoing research” (para. 1018). So why did it file them when it did? According to the Commission, Teva staggered the patent filings, sought preliminary injunctions on the basis of patents that the EPO granted, and then (sometimes) “strategically withdrew the challenged patents before the adoption of a reasoned decision of those challenged patents by the TBA [EPO Technical Board of Appeals],” because “a reasoned revocation would have had a strong precedent effect on the remaining divisional patents of the same patent family” (paras. 1019-20). In other words:
Teva carefully timed the filing and withdrawal of these divisionals, using these patents to delay or prevent market entry of competing GA products and in particular [generic competitor] Synthon GA. Correspondingly, Synthon and its licensees had to challenge the validity of each of those patents successively, as Teva repeatedly and in sequence decided to file and withdraw one patent only to rely on the next patent. Thus, instead of being able to obtain a final decision resolving the legal uncertainty concerning patent claims relevant for the entire divisional patent family, Synthon had to repeatedly re-start litigation from the very beginning before it could finally remove the patent barrier. This way, Teva managed to obstruct effective legal review of its patents and to prolong legal uncertainty surrounding their validity, thereby hindering Synthon GA’s entry on the markets. None of the patents relied on by Teva was confirmed as valid and enforceable in a final decision by the EPO (para. 140).
The
EC concludes that this scheme constituted an abuse of dominant position in view
of the following principles as applied to the facts. First, the general legal framework with regard
to exclusionary practices is that “a dominant undertaking’s conduct can only be
characterised as abusive if two cumulative conditions are met: (i) the practice
must be based on the use of means other than those which come within the scope
of competition on the merits; and (ii) the practice must be capable of
producing an exclusionary effect. Where those two conditions are fulfilled, the
undertaking in a dominant position can nevertheless escape the prohibition laid
down in Article 102 TFEU if it shows that the practice at issue was either
objectively justified and proportionate to that justification, or
counterbalanced or even outweighed by advantages in terms of efficiency that
also benefit consumers” (para. 1035). Second,
on the basis of AstraZeneca AB v. European Commission, Case C-457/10 P
(CJEU 2012), an abuse of dominant position may “consist of behaviour which is
otherwise lawful under branches of law other than competition law” (para.
1061). In my view, this is correct both
as a matter of law and policy. Much the
same reasoning underlies the decisions in the E.U. and the U.S. holding that
pay-for-delay settlements can violate competition law, even though the scope of
the agreed-upon term of exclusion falls within the potential exclusionary scope
of the patent. (Later in the decision,
the Commission also cites AstraZeneca for the proposition that “[t]he
mere possession by an originator company of an exclusive right can deter
generic competitors from entering the market” (para. 1346). By way of contrast, in U.S. cases in which, as
in AstraZeneca, the plaintiff alleges that the defendant’s submission of
misleading information to the Patent Office caused the issuance of an
undeserved right, there must be proof of
enforcement or threatened enforcement, not just possession, for there to be a potential
competition law offense. I take this
issue up in chapter 3 of the forthcoming book.)
Third, citing ITT Promedia NV v. Commission of the European Communities, Case T-111/96 (Ct. First Inst. 1998), which addressed sham litigation antitrust claims, the Commission states that it “does not in this case . . . take issue with Teva’s recourse to courts,” but it “does take issue with Teva’s filing of multiple, staggered divisional patents with overlapping content and strategic withdrawal of challenged patents before a reasoned decision on their validity could be issued” (para. 1062). Further, “[a]lthough, in isolation, a patent holder may exercise its rights by filing and withdrawing its patents, such conduct falls outside the scope of competition on the merits in cases where the regulatory procedures are used in such a way as to prevent or to hinder the entry of competitors on the market and where the conduct serves no legitimate interests. The combination of staggered divisional filings and strategic withdrawals enabled Teva to game the system and shield its patents from damaging precedents for as long as possible” (para. 1316; see also para. 1323). The stated rationale makes me think of the common-law tort of abuse of process, under which liability hinges on the use of judicial process “to achieve an objective that is outside the purpose for which the process is intended,” Restatement (Third) of Torts: Liability for Economic Harm § 26 (Am. L. Inst. 2020), as well as some descriptions of the civil-law doctrine of abuse of right.
Fourth,
it’s not too difficult to obtain a preliminary injunction in some E.U. member
states, because “courts do not make a substantial assessment of the validity of
the patent invoked in the request for preliminary injunctions and grant these
on a prima facie basis” (para. 1136; see also paras. 1371-72). This is, in my view, something that
individual nations should consider altering, to the extent permissible under Phoenix
Contact GmbH & Co. KG v. HARTING Deutschland GmbH & Co. KG, Case
C-44/21 (CJEU 2022).
Fifth,
the Commission repeatedly notes that the patents at issue were weak; that they
were likely to be invalidated despite having been initially granted and, in
some instances, having survived an initial opposition; and that Teva personnel
were aware of these infirmities (see, e.g., paras. 1136, 1141, 1160, 1178,
1181, 1182-1213, 1255, 1258, 1290, 1294, 1296, 1315, 1357). What I’m still trying to figure out is whether
the Commission views such evidence as necessary for the staggered filing and subsequent
withdrawal of divisional applications to constitute an abuse of dominant
position. Maybe it’s enough for a firm simply
to employ the strategy of strategic filing and withdrawal, just to avoid the
possibility of a reasoned decision on invalidity that could, potentially,
topple all the dominoes (although the perceived weakness of the patents surely makes
it more likely that the firm’s conduct is abusive). But then, if you thought the patent probably would
survive a decision by the TBA, and thus remain in force for several years to
come, why would you withdraw it at the last minute? (In a similar vein, see also para. 1247,
noting that Teva’s conduct of staggering the filing of the divisionals “shortened
and thereby reduced the value of each of its divisional patents”.) On the other hand, as noted above, there was
one divisional patent that was still pending before the TBA as of the date of
the E.C.’s decision—though it wasn’t filed until 2018, and so it may be that absent
the divisionals “game” Teva wouldn’t have been able to file or maintain that application for as long as it did;
and in any event, the decision states that that particular patent “was only maintained
in Turkey after being granted” and is apparently not all that relevant here other
than for context (para. 1121). Moreover,
in response to Teva’s argument that the Commission failed to prove
anticompetitive effects—because its “theory is grounded on the premise that
absent the withdrawals, the patents in question would have been invalidated,” when
in fact according to Teva other factors need to be considered as well (para.
1326)—the Commission states, among other things, that “[c]ontrary to Teva’s
claims . . . the Commission does not, and does not need to, independently
conclude on the validity of Teva’s patents. . . . It does however take into account Teva’s
contemporaneous documents, including those in which Teva assesses the strength
and validity of its patents as well as the documents originating from the EPO
(including ODs and TBAs decisions on Teva’s patents) to show that, from an ex
ante perspective, Teva’s comprehensive patenting conduct was capable of having
exclusionary effects” (para. 1331). The
Commission’s principal rejoinder, though, is that needs to consider the
defendant’s conduct as a whole, not just matters viewed in isolation.
Sixth,
based on the Commission’s findings, I got the sense that perhaps there would
have been a basis for alleging abuse by means of product hopping or sham litigation.
But the Commission states early on that its decision focuses on the divisionals
game and on the disparagement campaign (not addressed in this blog post), and
that the Commission was taking no “position on the legality of Teva’s other
practices, which are part of Teva’s overall strategy to hinder the entry and
uptake of competing GA” (para. 180).
Seventh and last, it seems to me that a potentially more effective way of preventing firms from gaming the system is to change the rules that enable the gaming. As the decision notes, the EPO changed the rules in 2014 to permit the voluntary filing of divisional applications as long as the parent remained pending, rather than as before limiting them to 24 months after the receipt of the examination report. Proposals to tighten the rules on divisionals, however, might meet the same fate as analogous proposals in the U.S. in recent years regarding matters such as continuations and double patenting. Entrenched interests typically do their utmost to retain their ability to engage in rent-seeking, which then leaves it to courts and agencies to clear up some of the mess ex post. Not an ideal situation, to be sure, but perhaps the best that can be hoped for in the near term.
