DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE Erik Hovenkamp† Jorge Lemus††

ABSTRACT—The Patent Trial and Appeal Board (PTAB) is a recently-formed division of the Patent Office in which patents can be challenged as invalid, and which differs from federal courts in a number of respects. We investigate whether monopolist-patentees and their prospective rivals are using the PTAB—which has not previously received antitrust attention—as a platform for striking settlements that delay the rivals’ entry. Such settlements are common in pharmaceutical markets, and are typically antitrust violations in cases where the patentee pays the challenger (“pay for delay”). However, problematic statutory inducements lead to excessively-delayed competition even in lieu of such payments. Our empirical findings suggest that delayed entry settlements are now commonly executed in the PTAB, and that they comprise a large majority of all PTAB settlements reached between pharmaceutical rivals. Further, nearly half of the delayed entry settlements were reached after the relevant patent claims were deemed “reasonably likely” to be invalid.



A different version of this paper was previously circulated online under the title “Reverse settlement and Holdup at the Patent Office.” We appreciate comments by Shawn Ambwani; Mark Lemley; Scott McKeown; Arti Rai; David Schwartz; Ted Sichelman; John L. Turner; Saurabh Vishnubhakat; and Stephen Yelderman. † Harvard Law School. Erik’s work on this project was supported by the Ewing Marion Kauffman Foundation. Email: [email protected] †† Department of Economics, University of Illinois at Urbana-Champaign. Email: [email protected]

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

I. INTRODUCTION The availability of generic alternatives to brand-name drugs is a major policy concern in economics and law. Our paper investigates the role of the Patent Trial and Appeal Board (PTAB) in shaping generic availability, by discerning what kinds of settlements rivals are striking within it. Since its inauguration in 2012, the Patent Trial and Appeal Board (PTAB) has substantially broadened the scope of post-grant patent reexamination, making it substantially less expensive and time-consuming to challenge patents as invalid.1 By and large, the PTAB has made the patent system more efficient by diminishing the extent to which low quality patents can be shielded from scrutiny by the high costs of district court litigation. However, the PTAB’s impact on the patent landscape is broad and far-reaching, and it raises some policy questions outside the boundaries of pure patent law. An important example is that the PTAB has not yet received significant antitrust attention.2 For instance, the American antitrust agencies (the DOJ and FTC), which commonly undertake litigation or investigations of anticompetitive patent settlements, have not yet directed such efforts at a PTAB settlement, notwithstanding that many PTAB adjudications arise between competitors. This paper addresses the significance of the PTAB to the antitrust-patent interface. The PTAB provides a new platform in which competitors can enter into settlement agreements that restrain competition, but whose judges have no jurisdiction to administer the antitrust laws. Its distinct rules, procedures, and jurisdiction present a number of policy challenges, for some existing antitrust machinery is not well-equipped to police settlements in the forum. These conditions can make the PTAB a strategically advantageous avenue to strike settlements that potentially run afoul of the antitrust laws. We undertake an empirical assessment of competing firms’ propensity to enter into delayed entry settlements in the PTAB. Such settlements arise between a monopolistpatentee and a prospective entrant that had challenged the monopolist’s patent, and involve the challenger’s agreement to terminate the challenge (thereby preserving the patent’s presumptive enforceability) and delay its entry into the market for some number of years.3 Such agreements are best-known for their prevalence in pharmaceutical markets. Until recently, virtually all of these agreements involved a “reverse payment”—a large lump sum transfer from the patentee to the challenger.4 In these cases the agreements are usually called “reverse payment” or “pay for delay” settlements. These arrangements have been addressed extensively in the antitrust literature, albeit in the context of district court litigation.5 In 2013, the Supreme Court’s Actavis decision held that pay for delay agreements may violate

1

PTAB adjudications address only patent validity, not infringement claims. A patent (or more accurately a patent claim) is invalid if it fails to satisfy one or more of the legal requirements for patentability, implying it was granted in error. 2 A recent exception is Sturiale (2016), which provides qualitative discussion of how the PTAB could help to deter reverse payment settlements. 3 To be clear, the challenger gets a license that does not take effect until the end of the contractual delay period. 4 The term “reverse” is used because, in most conventional settlements, money runs in the opposite direction. 5 E.g. Hemphill, 2006; Edlin et al., 2015; Cotter, 2014; Carrier 2009; Dolin, 2011; Shapiro, 2003; Hovenkamp & Lemus, 2017.

2

HOVENKAMP & LEMUS

the antitrust laws.6 However, delayed entry settlements not involving reverse payments—a format we call “pure delay”—are presumptively legal.7 Our analysis allows us to infer delayed entry settlements reached at the PTAB between pharmaceutical rivals. We begin by gathering data on all PTAB adjudications, in particular inter partes review (IPR) proceedings, in which the patentee is a brand-name drug seller and the petitioner is a generic drug maker challenging a patent on the brand-name drug. We then merge this with data from the Food and Drug Administration’s (FDA’s) Orange Book listings, which provides information on approved pharmaceutical drugs; the patents that cover them; and firm identifiers for both the brand-name seller and any approved generics, among other things. If a generic drug maker does not appear in the Orange Book, then we can rule out the possibility that it is selling a generic version of the relevant brand-name drug.8 Using the combined data, we can look for pharmaceutical settlements and discern whether the petitioner (the challenger) began selling a generic version of the patentee’s drug after the settlement. When the generic firm remains inactive well after the settlement, we infer a likelihood of delayed entry. In many instances, several IPRs correspond to the same parties and to a single drug, and are all settled simultaneously. These therefore correspond to a single settlement that happens to span two or more patents, and we therefore lump them together into what we call “consolidated settlement agreements” (CSAs). Over the PTAB’s 4-years of existence, there are 32 applicable CSAs in the data, which subsume a total of 52 IPRs. 24 of the CSAs (75%) satisfy our criteria for inferring delayed entry. Among these 24 CSRs, 10 (42%) occurred after the PTAB judge has “instituted” the IPR, which is a procedural step (discussed further below) reflecting the PTAB judge’s determination that the challenged patent is “reasonably likely” to be invalid.9 Thus, in these post-institution settlements, the parties anticipate a likely victory for the petitioner. Hovenkamp and Lemus (2013) show that, in lieu of any exogenous frictions on challenger entry, pure delay is a socially efficient manner of settlement. The reason is that, without a reverse payment, the firms cannot agree on terms that restrain competition beyond the level necessary for settlement to be mutually-acceptable. But this desirable property does not hold up in the pharmaceutical context, due to some limitations created by the HatchWaxman Act. The relevant statutory provisions, discussed in detail below, were designed to bolster the incentive to challenge drug patents. But in fact they do so only for the first generic to file a drug application with the FDA, while diminishing the incentive to challenge among all later-filing generic firms. Consequently a settlement between the first-filer and patentee acts to forestall entry by most of the generics that would otherwise enter. The result 6

FTC v. Actavis, Inc., 570 U.S. ___, 133 S. Ct. 2233 (2013). The court was persuaded that we can infer a high likelihood of invalidity when there is a large reverse payment—a claim that many scholars had advocated. (Dolin, 2011; Edlin et al., 2015). 7 The Actavis opinion expressly states that firms could settle without a payment as an alternative to reverse payment, which would result in a shorter delay period. 8 However, a generic maker that does show up in the Orange Book is not necessarily actively selling the drug in commerce. We thus use a separate resource, described below, to identify which Orange Book-listed generics are indeed commercially active. 9 The other settlements occur before the PTAB judge makes any institution decision, so in these cases we cannot make any inferences about what the judge thinks about the patent’s validity.

3

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

is ultimately that this settlement will restrain competition excessively, i.e. it will produce significantly less competition over the patent term than litigation would provide in expected value.10 The fact that a PTAB challenger stays out of the market after a settlement agreement is not irrefutable evidence of a reverse payment, although a payment is inferentially more likely in a post-institution settlement in which the generic firm remains out of the market for a significant period of time. But because of the problems created by Hatch-Waxman, even pure delay settlements present serious competition policy concerns in the pharmaceutical context, notwithstanding that they are probably not vulnerable to antitrust intervention. They contribute to a problem that many other papers have identified, which is that contractual restraints on generic entry pose a serious threat to consumer welfare. For example, Helland and Seabury (2016) estimate that restricting the entry of generic drugs reduced consumer surplus by about $800 million over a 5-year period. Although our findings on PTAB settlements illustrate serious competition policy concerns, we think that some relatively simply measures could help to address this issue. To that end, we conclude the paper with a number of policy proposals, some of which make use of the PTAB’s unique institutional rules and procedures.

A. GENERIC COMPETITION WITH PATENTED DRUGS Under the Hatch-Waxman Act, generic drugs can receive expedited FDA approval through an Abbreviated New Drug Applications (ANDA). This requires a demonstration that the generic drug is “bioequivalent” to one that has already gone through the full FDA approval process. This legislation avoids redundant testing of safety and therapeutic efficacy. ANDA approval is predicated on the applicant’s certification that, to the best of its knowledge, its generic drug will not infringe any active patent that is valid and enforceable. If the brand-name drug is indeed patented, the ANDA applicant must certify that the patent is either invalid or would not be infringed by its proposed generic—an option known as “Paragraph IV certification.” If the applicant takes this route, it must immediately provide notice to the patent holder, which will typically then sue the applicant. If the patent holder sues the generic producer within 45 days of receiving notice, ANDA approval is stayed for up to thirty months to allow litigation to proceed. That is, assuming the generic applicant does not obtain a license in a settlement, the ANDA will not be approved until the earlier of the dates on which (a) the ANDA applicant wins in court; or (b) litigation reaches the 30-month mark. Thus, if litigation lasts for more than 30 months, the ANDA will be approved at the 30-month mark, notwithstanding that the patent litigation has not yet concluded. If, by contrast, the generic maker receives a license, the ANDA would then be approved without further delay.11 If the patent is invalidated (by anyone), the first generic party to file for Paragraph IV certification receives a “generic exclusivity” period of 10

We demonstrate explicitly in the online appendix, using a model that builds on Hovenkamp & Lemus (2017). We discuss the results and intuition in Section II(A), below. 11 Another possibility is that the generic applicant could attempt to invalidate the patent through a PTAB challenge, which could expedite Paragraph IV approval.

4

HOVENKAMP & LEMUS

180 days. That is, if a later-filing generic successfully invalidates the patent, the exclusivity reward goes not to this firm, but to the first-filer. B. A BRIEF OVERVIEW OF PTAB The America Invents Act (AIA) in 2011 established the PTAB to conduct three kinds of patentability adjudications: Post Grant Review (PGR), Inter Partes Review (IPR), and Covered Business Method (CBM).12 In all three cases, the petitioner challenges the respondent’s patent as invalid, although the three adjudications vary with respect to availability and scope. Among challenges not directed at business method patents, the IPR format is overwhelmingly more common, and there are no applicable pharmaceutical challenges involving the alternative PGR format. Our focus will thus be on IPRs. When a petition is filed, it is immediately made public. The patentee-respondent then issues a response argument. Absent an early settlement, this initial phase of the adjudication culminates in a (non-appealable) institution decision by the judge. The adjudication will continue if and only if the judge institutes the IPR; the standard for institution is a “reasonable likelihood” that the challenged patent claims are invalid. If an IPR is instituted, the adjudication must be resolved no later than 12 months from the date of institution. If not instituted, the IPR is terminated. The “institution decision” is an example of the ways in which PTAB adjudications differ from full-fledged patent litigation; there is no analogous procedural apparatus in district court litigation. Further, unlike federal court judges, a PTAB judge has the authority to continue an adjudication to final judgment even if the parties settle and the petitioner has withdrawn. Additionally, the PTAB has no standing requirement: any party can challenge any patent. PTAB adjudications have become very popular as a substitute for more costly federal court litigation. Vishnubhakat, Rai, and Kesan (2016) find that about 70% of PTAB trials involve a petitioner who has been sued for infringement by the patentee. Thus in many instances the PTAB is utilized as a substitute for raising an invalidity defense in federal court. In other instances, the PTAB is utilized to forestall or deter potential patent assertions. For example, Unified Patents is an organization that regularly challenges low quality patents (usually those held by patent assertion entities, pejoratively known as “patent trolls”) that might be asserted against member firms, refusing to settle unless all members receive royalty-free licenses. Figure 1 shows the number of PTAB petitions over time. Most petitions challenge the patent of either an operating companies (52%) or a non-practicing entity (46%). Out of the 7,358 petitions in our dataset, about 25% are pending (i.e. not terminated). There are different reasons to terminate a petition: some are not instituted based on merits (22.36%), some are subject to a procedural termination (14%), some reach a final decision (33.25%), and some settle before a final decision (30%). We also collected patent information from the USPTO. The NBER classification numbers in Hall et al. (2001) reveal that PTAB petitions are driven by patents in the NBER category “Computers and Communications,”

12

https://www.uspto.gov/patents-application-process/patent-trial-and-appeal-board-0

5

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

representing about 50% of the patents. About 12% of the patents correspond to the category “Drugs and Medical.”

F IGURE 1: M ONTHLY PTAB PETIT IONS BE TWEEN S E PTEMBER 16, 2012 AND A UGUST 15, 2017.

Pharmaceutical firms will go to great lengths to shield their patents from legal scrutiny, and PTAB thus presents a major threat to pharmaceutical profits—at least in cases where a competition-restraining settlement is not on the table, such as when the petitioner is not a rival drug maker. A recent example demonstrates this acutely. In September of 2017, pharmaceutical firm Allergan reached an agreement to transfer ownership of a high-value drug patent to Saint Regis Mohawk, a native American tribe. The impetus for the transaction is to take advantage of the sovereign immunity maintained by native American tribes like Saint Regis Mohawk, which potentially immunizes the patent from PTAB challenges. 13 In effect, the deal gives Saint Regist Mohawk a share of the drug for as long as the patent remains valid and enforceable.

II. COMPETITION POLICY CONCERNS Patent challenges—and settlements of such challenges—tie into a number of prominent issues at the interface of competition and innovation policy. One well-studied concern involves deficient incentives to challenge patents—and, similarly, to carry such challenges to final judgment.14 (Farrell and Merges 2004; Lemley and Shapiro 2005; Choi 2005). However, this problem does not necessarily raise antitrust concerns, even if the parties are competitors. Rather, in the context of patent challenge settlements, antitrust concerns arise 13

Challenges in federal court are still available, however. A self-interested challenger generally prefers a royalty-free licensing settlement to an outright invalidation, since the latter keeps the patent valid and enforceable against all third parties. The result may be an undersupply of fully-litigated validity challenges. On the other hand, several authors have identified a few reasons why the firms may not settle in all cases (Lanjouw and Schankerman, 2001). First, firms may have different beliefs about the likelihood of invalidation, e.g. under private information, so settlement agreements may be infeasible (Bebchuk, 1984). Second, a patentee may decline to settle for reputational reasons—i.e. to deter future challenges by cultivating a reputation for vigorously defending its patents to the bitter end. 14

6

HOVENKAMP & LEMUS

when competing firms settle on terms that subvert competition materially beyond the level needed to support a mutually-acceptable settlement. This is consistent with antitrust’s broader treatment of horizontal restraints, which is that they are permissible only to the extent reasonably necessary to effect some cognizable efficiency. When the parties are rivals with market power and the patent constitutes a potential entry barrier, different litigation outcomes (e.g. injunction versus invalidation) produce substantially different effects on competition, resulting in very different levels of total profit. Since a settlement must at least match the total profits expected to result from litigation, the firms’ expectations about litigation determine the extent to which their settlement must necessarily restrain competition in order to be mutually-acceptable. If the patentee is relatively less likely to prevail, then they can agree on less restrictive terms, all else being equal. But, problematically, rival firms always have an interest in restraining competition to the level that maximizes their joint profits, just so long as they are able to distribute profits in a way that satisfies all of them. (Hovenkamp & Lemus, 2017; Shapiro, 2003). Pay for delay settlements are a clear illustration of this. Assuming that monopoly maximizes total profits, the firms always have a joint-interest in restraining competition for the full remainder of the patent term; the patentee’s probability of winning in court affects only the size of the reverse payment, not the extent to which competition is restrained. But even if the delay period ends prior to patent expiration, any pay for delay settlement involving a nonnegligible reverse payment must restrain competition more than necessary, because the payment acts to persuade the challenger to delay entry longer than it otherwise would. The firms could instead have entered into a “pure delay” settlement—one that delays the challenger but involves no reverse payment.15 Hovenkamp & Lemus (2017) shows that different kinds of restraints imposed on a prospective rival (e.g. entry limitations, royalties, output caps, etc.) vary systematically in their propensity to elicit “proportional” settlements—those generating total profits equal to what the firms privately expect they would get from litigation (ignoring litigation costs). A proportional settlement has the property that competition is restrained no more than necessary for the parties to mutually-prefer settlement to litigation.16 The paper shows that, in lieu of any exogenous constraints on entry, pure delay is “perfectly proportional” in the sense that, ignoring bargaining over avoided litigation costs, the firms simply cannot agree on anything other than the settlement whose profitability exactly matches the firms’ private expectations about litigation.17 However, the Hatch-Waxman Act’s first-filer exclusivity system undermines this desirable property.

15

The Actavis opinion notes this point explicitly, leaving little question that pure delay settlement are lawful. Shapiro (2003) proposed a very similar antitrust standard, which is that the settlement must provide no less consumer welfare than the expected result of litigation. See Hovenkamp & Lemus (2017) for discussion of the differences. 17 In fact pure delay is a special case, because consumer welfare must also be the same as the expected level that would result from litigation. 16

7

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

A. THE HATCH-WAXMAN PROBLEM Hatch-Waxman’s generic exclusivity system was intended to encourage more generic competition to help lower the cost of pharmaceutical drugs. Because a brand-name drug and its generic equivalents are essentially fungible, competition between them is often very aggressive, and the profits that ensue may be too meager to incentivize patent challenges. Hatch-Waxman’s first-filer rule is an imperfect attempt to resolve this problem, as many authors have recognized. (Hemphill, 2006; Carrier, 2009; Edlin et al., 2015). First, it allocates the exclusivity reward on a basis that lacks any inherent connection to the behavior it seeks to induce; the first-filer always gets the exclusivity reward, even if it was someone else who undertook the costly task of challenging the patent. Second, it creates a regulatory bottleneck by preventing later-filing generics from obtaining FDA approval before the firstfiler’s exclusivity period ends, and the exclusivity period may itself be delayed via settlement. Hence, later-filing generics’ ability to incite competition is largely at the mercy of the first-filer and its dealings with the patentee.18 For these reasons, Hatch-Waxman has the perverse effect of diminishing the incentive to challenge among all later-filers. If there were no statutory exclusivity system at all, a laterfiler would at least be able to enter immediately if it prevailed in challenging the patent. But this is not so under the current statutory regime. One widely-recognized result is that anticompetitive pay for delay settlements become much more stable, because the first-filer and patentee can confidently preserve monopoly through their own bilateral settlement, for they know that third party generics have little motivation to bring their own challenges. But there is a second problem, which is that even pure delay settlements are problematic as a result of Hatch-Waxman, and will generally restrain competition disproportionately, i.e. more than necessary to match the joint profits firms expect to accrue from litigation. We prove this formally in the online appendix using a model that builds on Hovenkamp & Lemus (2017). The intuition for this result is straightforward. Consider the case in which, by default, challenges are unprofitable due to aggressive post-invalidation competition. HatchWaxman further weakens the incentive to challenge among all later-filers, while increasing it for the first filer. Supposing this gives the first-filer a credible threat to challenge, the patentee will offer a settlement only to this firm, since third party generics will not bring separate challenges. As a result, one of the gains from settlement is that it forestalls entry by third party generics, which benefits both the patentee and first-filer. They will thus bargain over the profits created by this entry barrier, and the patentee claims its share by taking a longer delay period. Nash bargaining thus results in competition being delayed beyond the level needed to support mutual agreement.

III. INFERRING DELAYED ENTRY IN PTAB SETTLEMENTS Like virtually all settlements, PTAB settlements terms are almost always kept confidential.19 But in the present context we can take advantage of certain legal and 18

There are statutory provisions stipulating conditions under which a first-filer forfeits its exclusivity, but in practice they are largely impotent, and have never been invoked in practice (as of the time of writing). 19 If the parties wish to settle and terminate the proceeding, the Patent Act permits them to request that the settlement agreement be treated as “business confidential,” in which case it is not publicly available.

8

HOVENKAMP & LEMUS

institutional features to draw strong inferences a settlement’s impact on entry. To do so, we search the data for IPRs satisfying all of the following conditions: (C1) The parties are competing drug manufacturers and the disputed patent covers an FDA-approved brand-name drug; (C2) the dispute was settled, resulting in termination of the proceeding; and (C3) following the settlement, the challenger did not commence sales of a generic version of the brand-name drug. Condition (C3) is particularly important. It ensures that our inference procedure does not flag “ordinary” patent settlements that would not raise antitrust concerns—namely those that give the generic firm a (non-delayed) license in exchange for its promise to pay royalties or other consideration. Such agreements are expressly authorized by the patent act, whereas agreements that forestall patent challenges are not (Hovenkamp, 2016b; Cheng, 2016). Our data includes all PTAB trials from September 19, 2012 to August 15, 2017. We focus on trials occurring between pharmaceutical firms, and in which the relevant patents cover a brand-name drug sold by the patentee. To identify these trials, we make use of the FDA’s Orange Book, which provides information on FDA-approved drugs (both proprietary and generic); applicable patents; and the identities of firms that sell those drugs. 20 In particular, the Orange Book lists generic manufacturers when they obtain approval for a previously-filed Abbreviated New Drug Application (ANDA). We complement our dataset with characteristics of the underlying patents (e.g. NBER classifications, number of claims and citations) and public information from district court filings and decrees. When the IPR begins, the generic firm is always unlisted in the Orange Book, because it has not yet demonstrated that the relevant patents are invalid or uninfringed, and thus has not yet obtained ANDA approval under Paragraph IV. If the generic-petitioner receives a non-delayed license to practice the relevant patents, then it has everything it needs to satisfy its Paragraph IV certification and thereby garner FDA approval. We would thus expect to see these petitioners show up in the Orange Book relatively soon after the settlement, and some of the generic firms do. On the other hand, if the parties were to enter into a delayed entry settlement, this would typically withhold the relevant patent rights from the generic firm for some period of time, and as such it will usually not show up in the Orange Book listings ex post. If a challenger firm is not listed in the Orange Book, then we know it is not presently selling a generic drug.21 However, converse is not necessarily true. That is, a generic firm listed in the Orange Book is not necessarily an active seller of the drug, notwithstanding that it has the approval needed to do so lawfully. To provide a check against this, we utilize the Anticipated Availability of First-Time Generics (AAFTG), which is a subscription-based resource offered to pharmacology practitioners.22 The AAFTG provides information about 20

The Orange Book is publicly accessible at http://www.accessdata.fda.gov/scripts/cder/ob/. By contrast, some generic firms in the Orange Book after settlement may not be actively selling a generic drug, despite having the right to do so. 22 The AAFTG (updated August, 2017) is circulated in Pharmacist’s Letter, available at: http://pharmacistsletter.therapeuticresearch.com/home.aspx?cs=NDPTL~CP&s=PL 21

9

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

generic drugs (whose manufacturers are identified) that are not currently being sold, and whether they are anticipated to enter relatively soon. There are a few settlements where the generic firm shows up in the Orange Book post-settlement, but is also listed in the AAFTG as not being available for some number of years.23 When an Orange Book-listed generic is not listed in the AAFTG, it is actively being sold, which is generally easy to confirm through other pharmaceutical resources.24 Thus, to identify condition (C3) above, we rely principally on the Orange Book, supplemented when necessary by the AAFTG. A. EVALUATING ALTERNATIVE EXPLANATIONS In this section we discuss three alternative hypotheses for why a petitioner that settle a PTAB invalidity challenge has not entered the market. In each case, we conclude the alternative hypothesis is unable to explain our findings. 1. Authorized Generic Agreements An “authorized generic”—the brand-name firm’s authorized generic version, which must be an identical copy of its branded drug—does not require ANDA approval, and its manufacturer is therefore not listed in the Orange Book. In principle, a PTAB settlement could give the generic-petitioner the right to sell an authorized generic. This would be largely equivalent to a licensing settlement. Fortunately, we can use of the FDA’s listings of authorized generics, which specify the particular date on which each authorized generic enters the market.25 For each settlement in our database, we checked whether an authorized generic was launched within a year after the PTAB settlement. There is just one settlement in the data that appears to have resulted in an agreement that permits the petitioner to sell an authorized generic,26 and thus the possibility of authorized generic agreements does not undermine our inference procedure. In fact, for a very large majority of the drugs in these settlements, there is no authorized generic at all.27 In fact, additional details suggest this settlement nevertheless involved delayed entry. The petitioner, Mylan, challenged a patent that covers Loestrin and Moestrin. The only difference between these drugs is that the latter is a chewable version of the former. The patent holder is Warner Chilcott, which is owned by Actavis. An authorized generic of Loestrin was approved for sale on November 5, 2015—just three months after the PTAB 23

The anticipated availability dates listed by the AAFTG are sometimes set to the date of patent expiration by default (presumably because there is no available information signaling earlier entry). In these cases we can infer only that the generic is not currently selling the drug; we cannot rule out the possibility of a private agreement permitting pre-expiration entry at some later date. However, in other cases the AAFTG lists a preexpiration date, reflecting publicly disclosed information that a generic will become available before the relevant patent expires. 24 For example, drugs.com provides information on generic availability of brand-name drugs, including identifiers of the firms that sell such generics. 25 FDA Listing of Authorized Generics (June 30, 2017). Available on the FDA website. 26 The apparent authorized generic settlement occurred between Mylan (petitioner) and Warner Chilcott (patent owner). See IPR2015-00682, paper 10 (PTAB, Aug. 20, 2015). See the notes on this settlement in the appendices for more information. 27 There were only two additional cases such that there is some authorized generic on record for the relevant drug. However, in these cases the generics had been authorized by the patentee long before the settlements in our data, implying the manufacturer of the authorized generic is not the petitioner from the settlement.

10

HOVENKAMP & LEMUS

settlement, suggesting that petitioner Mylan may have obtained the right to sell an authorized generic in the settlement. However, the AAFTG indicates that sales of Mylan’s generic were delayed by the settlement.28 Incidentally, Warner Chilcott has previously been accused of paying generic firms to delay entry of Loestrin, and then lengthening its own patent exclusivity by switching over to the chewable form, Moestrin—a process known as patent “evergreening.” (Upadhye, 2014). Retail pharmacists CVS and Rite Aid filed an antitrust claim against Warner Chilcott and two generic firms (Lupin and Watson) for this alleged agreement to delay generic entry.29 2 Existence of Additional, Uncontested Patents A brand-name drug could in principle be covered by additional patents that were not subject to the settled IPRs. In this case, the petitioner’s challenge may not eradicate the barrier to generic entry, even if it is successful. However, these uncontested patents do not pose a threat to generic entry unless they are “essential” in the sense that one cannot market a viable generic without reading on them. In fact, there is reason to believe that many Orange Book patents are not essential. As Eisenberg and Crane (2015) note, the FDA makes virtually no effort to make sure listed patents actually cover the drugs that supposedly read on them, preferring “to avoid any responsibility for reading patents.” As such, drug companies can list patents that do not actually cover their drugs, but which nevertheless create legal obstacles to generic entry. As such, the fact that a generic-petitioner challenges only some patents may reflect that it is only challenging the ones that actually cover the brand-name drug. In a similar vein, drug companies may strategically list additional patents in the Orange Book over time, with each new addition triggering an additional 30-month stay on generic approval pending litigation—a practice known as “evergreening” (Upadhye, 2014). A brand-name drug maker may use this strategy to protract the generic approval process, even if the newly-listed patents do not actually cover the drug in question. A generic petitioner may recognize that the evergreen patents do not actually cover the drug, but it may nevertheless want them to persist, so that they will continue to exclude third parties postsettlement. Thus the generic firm may challenge only the originally-listed patents, demanding a license that begins when these patents expire, ensuring that it will be the sole generic firm for some material period of time. 3 Involuntary Delays in Commercialization In principle, a settlement might provide an immediate license, but the petitioner may be significantly delayed—for reasons outside its control—in getting its generic product to market. For example, it might suffer some setbacks in production, or there may be a delay

28

In the AAFTG, the listed manufacturer for the delayed generic is Jai Pharma, which is a wholly-owned subsidiary of Mylan. 29 See Kass, Dani, "CVS, Rite Aid Hit Warner Chilcott with Loestrin Antitrust Suit," Law360 (April 6, 2016). Available at http://www.law360.com/articles/781046/cvs-rite-aid-hit-warner-chilcott-with-loestrinantitrust-suit

11

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

in administrative approval of the license at the FDA.30 However, even if we focus on settlements that took place as long as three years ago, we still find that many of these settlements satisfy our conditions for inferring delayed entry settlement.

B. SETTLEMENT EXAMPLE To provide context for subsequent discussion, we first provide an illustrative example of a horizontal PTAB settlement with the markings of a delayed entry agreement.31 Pharmaceutical producer Metrics (and two later-joining co-petitioners) challenged a drug patent owned by Senju. The drug, marketed in the U.S. by co-respondent Bausch & Lomb, is called Prolensa. The joint request to settle was filed after the IPR was instituted. An Orange Book search shows that Metrics did not go on to market a generic version of Prolensa (nor, apparently, did any of the co-petitioners) as of the date of writing. However, the petition filed by Metrics notes that it had previously filed an ANDA for a generic version of Prolensa. The joint request to settle notes that Metrics and Senju had also agreed to settle their ongoing infringement litigation involving the challenged patent, which arose out of Metrics’ ANDA filing. Specifically, the litigation was resolved through a stipulated consent judgment (a judgment that memorializes the parties’ settlement terms), which was attached as an exhibit in the PTAB proceeding.32 The stipulated judgment applies to all of the patents with equal force and effect. The consent decree begins by resolving the infringement and validity issues in patentee Senju’s favor. It stipulates that the patents are "valid, enforceable, and would be infringed by … the Metrics Product.” Here the “Metrics Product” refers to the generic drug embodied in Metrics’ ANDA. The stipulated judgment goes on to issue a permanent injunction against Metrics’ generic. It states that Metrics “will be enjoined until expiration of [the patents] … from making, using, offering to sell, selling, or importing … the Metrics Product” (emphasis added). The consent decree has a preclusive effect on the parties, meaning that it bars the petitioners from ever relitigating the invalidity or infringement questions just resolved. The judgment concludes by acknowledging that Metrics can still obtain Paragraph IV ANDA approval by receiving a license from Senju. But the Orange Book listings imply this has not happened.

The FDA recently announced the “Drug Competition Action Plan” to accelerate the entry of generics. Specifically, complex drugs like metered dose inhalers used to treat asthma, are hard to “genericize.” https://blogs.fda.gov/fdavoice/index.php/2017/10/reducing-the-hurdles-for-complex-generic-drugdevelopment/ 31 The example involves the IPR Metrics, Inc. v. Senju Pharmaceutical Co. Ltd. (settled July 8, 2015). 32 The district court litigation also involved several related patents directed at the same active ingredient found in Prolensa, but Metrics does not offer any approved drug with this active ingredient. 30

12

HOVENKAMP & LEMUS

IV. EMPIRICAL FINDINGS We linked the patents involved in PTAB trials with those in the Orange Book.33 Combining these databases we get a comprehensive look at PTAB settlements that involve Orange Book-listed patents, with the petitioner being a generic drug maker. There are 372 PTAB petitions involving a patent listed in the Orange Book, which corresponds to 204 unique patents. Table 3 shows the current status of PTAB trials involving Orange Book patents. In our data there are 105 petitions pending as of August 2017. Institution Decision Reached? Yes No Total Yes 58 47 105 Pending as of August 31, 2016? No 227 40 267 285 87 372 T AB LE 1: I NS T ITUTION D EC IS IONS AND C URRENT S TATUS FOR P ETIT IONS INVOLV ING O R ANGE B OOK PATENTS

We focus on the 267 finalized IPRs. There are several reasons for cases to be finalized: final written decision; settlement (before or after institution); a decision to deny institution; or a patentee’s request adverse judgment, not instituted by merits, procedural terminations, final written decisions, and settlements. In narrowing our focus to PTAB settlements, we exclude IPRs that reached a final written decision and those terminated following a party’s request for adverse judgment against itself. We also exclude IPR settlements that occurred after the generic firm was successfully enjoined in district court. After doing this, our database includes 52 applicable IPR settlements that involve Orange Book patents, and which arise between pharmaceutical firms, with the patent holder being the seller of an FDAapproved brand-name drug. Figure 2 provides an overview of the timing of settlements relative to institution decisions. The figure shows that 30 IPRs settled before an institution decision was reached. From the cases not settled before an institution decision, 110 of them were not instituted. From the once that were instituted, 22 cases settled and 105 cases reached a final decision. The 22 settlements reached after the PTAB decided to institute the IPR are disconcerting, because an institution decision is (by definition) a signal that the patent claims are reasonably likely to be invalidated. Hence, a post-institution settlement is more likely to involve a relatively low quality patent. If instead, the petition is denied institution, the trial ends automatically, and this decision is non-appealable.

33

For each referenced PTAB trial, the relevant documents (excluding confidential materials) are easily accessed on the PTAB database (PRPS). Orange book data is available at: http://www.fda.gov/Drugs/InformationOnDrugs/default.htm

13

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

F IGURE 2: T ERM INAL E VE NT F REQUE NC IES IN P HAR MACEUTIC AL IPR S

In many instances there are groups of several IPRs that: (a) involve the same parties; (b) surround patents covering a common drug (or a common set of closely-related drugs) sold by the patentee; and (c) settle at the same time. Every such grouping corresponds to what is effectively single challenge, but which happens to subsume two or more patents. Thus, we aggregate these IPRs into groupings we call “consolidated settlement agreements” (CSAs). Table A1 in the appendix provides a summary for each CSAs, including details about whether they satisfy our inference criteria, and whether they occur post-institution. Our Online Appendix provides further details on each one of these CSAs. There are 32 unique CSAs in the PTAB data. In 24 of them (75%), which are highlighted Table A1, our inference criteria for delayed entry settlement are satisfied. That is, in these cases, we found no evidence of the petitioner selling a generic version of the disputed drug after the settlement. Moreover, of the 24 settlements satisfying our inference conditions, 10 (42%) were settled post-institution, implying that the challenged patent claims were reasonably likely to be invalidated.34 In most (but not all) of the settlements we are able to find some public records indicating that the petitioner had indeed filed an ANDA before the PTAB settlement date (usually a court document stating that fact). It should be noted, however, that even if we cannot find evidence of a pre-settlement ANDA-filing,35 we cannot presume that no such filing occurred. The FDA does not itself publish the identities of ANDA-filers unless and until they have been approved (which usually does not happen in a delayed entry settlement). A. PREDICTORS OF SETTLEMENT To further explore PTAB settlements involving Orange Book patents, for each patent in our dataset, we collected the number of claims, the number of words, the number of figures, the number of references, and the number of forward citations, the remaining time of protection, and whether or not there are exclusivity periods. With these information we study whether settlements are explained by patent characteristics. To this end, we estimate the following model:

34

In one of these CSAs, which involved two settled IPRs, just one of the IPRs had been instituted prior to settlement. In all other post-institution CSAs, every IPR had been instituted prior to settlement. 35 See the table in the appendix for the CSAs in which we found no evidence of a pre-settlement ANDAfiling by the petitioner.

14

HOVENKAMP & LEMUS

𝑃𝑟𝑜𝑏(𝑦𝑖 ) = 𝛼 + 𝛽 𝑋𝑖 + 𝛾𝑡 + 𝜀𝑖 . The left-hand side gives the probability of event 𝑦𝑖 in trial i, while the right-hand side includes a vector of patent characteristics 𝑋𝑖 and a time fixed effect 𝛾𝑡 . Following Figure 2, we study three different events, corresponding to three values for 𝑦𝑖 . They are: (a) settlement before institution; (b) a decision to institute, given that the firms did not settle prior to the institution decision; and (c) settlement after a decision to institute.

T AB LE 2: P ROB IT R E GRESSION R E SULTS

Table 3 (Column 1) shows that trials involving patents that are highly cited and have longer life are less likely to settle pre-institution. The number of patent citations is a common proxy for patent strength, and therefore one interpretation of this result is that stronger patents are less likely to settle prior to institution. Also, a patentee holding a (relatively stronger) patent with a longer statutory life has less incentives to settle prior to institution. Finally, patents with more figures are also less likely to settle. Column 2 shows that patents with more claims, fewer words, and fewer non-patent prior art citations are more likely to be instituted. This is intuitive, since broader claims tend to have fewer words and are typically more likely to be invalidated by virtue of overclaiming. And one would expect that the combination of broad claims with fewer nonpatent prior art citations would similarly cut in favor of a higher likelihood of invalidation. Column 3 shows that, conditional on a case being instituted, cases that settle are more likely to involve patents with longer statutory life and patents with no FDA exclusivity designation. This suggests that, when patent holders anticipate their patent-based rents will accrue over many years into the future—rather than arriving mostly in the near-term—then they are more reluctant to settle pre-institution. This may reflect an interest in discouraging challenges by developing a reputation for making it relatively harder and more expensive 15

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

for a challenger to garner a settlement. Note that, while settlement likelihoods correlate significantly with factors affecting profit expectations, such as the remaining term or FDA exclusivity designations, such factors have no significant predictive power on a judge’s institution decision. A final point is that it important to be cautious in applying these results, since the PTAB’s relatively young age leaves us with a relatively small sample size.

V. PROPOSED REFORMS The biggest problem looming over pharmaceutical patent settlements is the HatchWaxman’s first-filer exclusivity system. Many authors have recognized the problematic incentive problems it creates (e.g., Hemphill, 2006; Carrier 2009; Edlin et al., 2015). Ideally, the statute would be amended so that the exclusivity reward is given to the first generic firm to obtain a judgment that the patent is invalid or not infringed, and not necessarily the first firm to file an ANDA. This way, later-filers remain a viable source of competitive pressure, so that the first-filer knows that if it accepts a lengthy delayed entry settlement, it may forego the exclusivity period. What about reforms directed specifically at the PTAB? In fact, PTAB judges have a useful tool at their disposal that federal courts do not possess: they can continue a patent challenge to final judgment even after the parties have reached settlement terms and the petitioner has withdrawn from the adjudication.36 Whether this will occur is largely at the discretion of the judge, and rarely occurs in present practice. However, such continuation could be very helpful in pharmaceutical settlements where the patent seems particularly likely to be invalid, in which cases settlement raises serious concerns about injuries to competition and consumer welfare. For example, one reasonable policy might be to continue to final judgment following post-institution settlements in which the petitioner was the first generic maker to file an ANDA—at least if the settlement does not give the petitioner a right to enter the market immediately. Another valuable improvement would be to establish a more concrete requirement that pharmaceutical PTAB settlements be disclosed to the FTC for antitrust review. At present, the Medicare Modernization Act (MMA) does include a reporting requirement for most settlements between Paragraph IV ANDA-filers and brand-name drug sellers.37 In particular, settlements between such parties must be reported to the FTC if they affect the “commercialization or sale” of either firm’s drug, or if they affect the exclusivity period in some way. The FTC has interpreted this as effectively applying to any settlements that definitively preclude the possibility of pre-expiration entry by the generic firm.38

36

35 U.S.C. 317(a). Medicare Prescription Drug, Improvement, and Modernization Act. Pub. L. 108-173, 117 Stat. 2461. Sec. 1112. 38 Federal Trade Commission, Letter to Helene D. Jaffe, Counsel for Sanofi-Aventis, May 9, 2011 (asserting that a settlement must be reported if it requires the generic firm to switch its ANDA from Paragraph IV to Paragraph III, which would prevent pre-expiration entry). Available at https://www.ftc.gov/tips-advice/competition-guidance/industry-guidance/health-care/pharmaceuticalagreement-filings. 37

16

HOVENKAMP & LEMUS

However, this reporting requirement appears to have been designed in contemplation of district court litigation. By contrast, PTAB adjudications assess validity alone. And a decision that a patent is valid (or a settlement that ends a validity challenge) does not definitively block entry as a matter of law, because a showing of noninfringement would still be sufficient to secure a right to enter. However, in practice the validity issue may be the only one that is really in dispute, since generic drugs will usually infringe the brandname drug’s patents. As such, in practical effect, a validity settlement may make preexpiration entry virtually impossible (at least for the settling petitioner), even though it technically leaves open the unlikely possibility of entry based on noninfringement. This undermines the ability of the MMA’s reporting requirement, as validity settlements do not necessarily fall within its purview.39 As such, another helpful reform would be to expand the reporting requirement to include PTAB settlements in which the generic firm does not obtain the right to enter the market immediately.

VI. CONCLUSION Antitrust maintains very stringent restrictions on market division agreements. But a valid patent entitles its owner to exclude rivals from selling infringing goods, and hence delayed entry patent settlements are an acute example of the tensions that arise between antitrust and IP policy. Pay for delay settlements have been a major antitrust concern throughout the last two decades. The debate came to a head in the Supreme Court’s 2013 Actavis decision, which concluded definitively that pay for delay settlements are anticompetitive when the payment seems to have to reasonable explanation other than to further delay competition. But the Hatch-Waxman Act’s problematic “first-filer exclusivity” system has the effect of inducing excessive delays of competitive entry, even in the absence of reverse payments. Thus, payment-free delayed entry agreements in pharmaceutical markets remain a serious competition policy concern, notwithstanding that such arrangements do not violate the antitrust laws. This article explores whether—and with what frequency—pharmaceutical rivals are shifting their settlement dealings to validity adjudications in the PTAB, a recently-formed division of the patent office that maintains a comparatively lower profile than federal court litigation. Our empirical approach focuses on PTAB settlements between brand-name drug sellers and rival generic drug makers, and looks to various sources of generic entry postsettlement. When entry does not occur post-settlement, this provides a strong inference of delayed entry. About 75% of applicable settlements satisfy this inference condition, suggesting that delayed entry is widely-utilized in the PTAB pharmaceutical settlements. And nearly half of these settlements occurred after institution, which (by definition) suggests that the judge considers it “reasonably likely” that one or more patent claims are invalid. The PTAB has not previously received antitrust attention, but our analysis makes it clear that this is a problematic oversight. The PTAB differs from federal courts in a number of ways that can be strategically advantageous for firms. It has distinct procedural rules, and The FTC may very well maintain the position that such PTAB settlements are indeed covered by the MMA’s reporting requirement. But the fact is this interpretation has not been endorsed by any court, and thus the issue remains unresolved. 39

17

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

its adjudications are much narrower in scope, leaving it with a comparatively low profile in relation to full-fledged litigation. But settlements in the forum present substantially the same anticompetitive potential as those reached in federal court. As such, antitrust cannot hope to maintain robust oversight of horizontal patent agreements without acknowledging the important new role played by the PTAB.

REFERENCES Bebchuk LA. Litigation and settlement under imperfect information. The RAND Journal of Economics. 1984 Oct 1:404-15. Carrier, Michael A., “Solving the Drug Settlement Problem: the Legislative Approach,” Rutgers Law Journal (2009), vol. 31, pp81-101. Carrier, Michael A., “Why the Supreme Court Should Deny Certiorari in King Drug,” Antitrust Chronicle, September 16, 2016, pp.1-8. Cheng, Thomas K., 2016, ANTITRUST TREATMENT OF THE NO CHALLENGE, Journal of Intellectual Property and Entertainment Law, Vol. 5, No. 2, pp 437-512 Choi JP. Live and let live: A tale of weak patents. Journal of the European Economic Association. 2005 Apr 5;3(2‐3):724-33. Cotter TF. FTC v. Actavis, Inc., 2014. When Is the Rule of Reason Not the Rule of Reason. Minn. JL Sci. & Tech.15:41. Dolin, G., 2011. Reverse Settlements as Patent Invalidity Signals. Harvard Journal of Law & Technology, 24(2), pp.281-333. Edlin, A., Hemphill, S., Hovenkamp, H. and Shapiro, C., 2015. Actavis Inference: Theory and Practice, The. Rutgers UL Rev., 67, p.585. Eisenberg, Rebecca S. and Daniel A. Crane. Patent Punting: How FDA and Antitrust Cournts Undermine the Hatch-Waxman Act to Avoid Dealing with Patents. Michigan Telecommunications and Technology Law Review, 21(2):197-262 (2015). Farrell, Joseph and Robert P. Merges. 2004. Incentives to Challenge and Defend Patents: Why Litigation Won’t Reliably Fix Patent Office Errors and Why Administrative Patent Review Might Help. Berkeley Technology Law Journal 19(3):943-70. Hall, Bronwyn H, Adam B Jaffe, and Manuel Trajtenberg. The NBER patent citation data file: Lessons, insights and methodological tools. Technical report, National Bureau of Economic Research (2001).

18

HOVENKAMP & LEMUS

Helland, E. and Seabury, S.A., 2016. Are Settlements in Patent Litigation Collusive? Evidence from Paragraph IV Challenges (No. w22194). National Bureau of Economic Research. Hemphil, Scott. Paying for Delay: Pharmaceutical Patent Settlement as a Regulatory Design Problem, 81 N.Y.U. L.Rev. 1553, 1579 (2006). Hovenkamp, Erik, 2016a. Predatory Patent Litigation: How Patent Assertion Entities Use Reputation to Monetize Bad Patents. Antitrust Chronicle, vol. 1, pp46-55. Hovenkamp, Erik. 2016b. Competition, Inalienability, and the Economic Analysis of Patent Law. Forthcoming, Stanford Technology Law Review. Hovenkamp, Erik and Lemus, Jorge. Proportional Restraints and the Patent System. 2017. Working paper. Hovenkamp, Herbert J. 2015. The Rule of Reason and the Scope of the Patent. San Diego Law Review 52:515-54. Jacobo-Rubio, Ruben and Turner, John L. and Williams, Jonathan W., Generic Entry, Payfor-Delay Settlements, and the Distribution of Surplus in the US Pharmaceutical Industry (October 7, 2014). Available at SSRN: http://ssrn.com/abstract=2481908 Lanjouw JO, Schankerman M. Characteristics of patent litigation: a window on competition. RAND journal of economics. 2001 Apr 1:129-51. Lemley, M.A. and Shapiro, C., 2005. Probabilistic patents. The Journal of Economic Perspectives, 19(2), pp.75-98. Shapiro C. Patent licensing and R & D rivalry. The American Economic Review. 1985 May 1;75(2):25-30. Shapiro, Carl. Antitrust Limits to Patent Settlements. 2003. RAND Journal of Economics 34(2):391-411. Sturiale, Jennifer E., The Hatch-Waxman Act, Post Grant Review, and the PTAB: A New Sort of Competition. Working Paper (2016). Upadhye, Shashank. There's a Hole in My Bucket Dear Liza, Dear Liza: The 30-Year Anniversary of the Hatch-Waxman Act: Resolved and Unresolved Gaps and Court-Driven Policy Gap Filling, William and Mary Law Review, 40(4):1307-1369 (2014). Vishnubhakat, S., Rai, A.K. and Kesan, J.P., 2016. Strategic Decision Making in Dual PTAB and District Court Proceedings. Berkeley Technology Law Journal, Forthcoming (2016).

19

DELAYED ENTRY SETTLEMENTS AT THE PATENT OFFICE

APPENDIX Table A1: Consolidated Settlement Agreements (CSAs in bold are those in which we infer delayed entry settlement) CSA* (Settlement date; proprietary drug) * Bold font reflects inference of delayed entry settlement

Settled PostInstitution? (Institution date) Yes (3/19/2013) Yes (3/5/2013) Yes (1/2/2014) Yes (7/29/2014)

Parallel district court litigation?40

Petitioner filed ANDA?

Petitioner in Orange Book?

Anticipated Entry Date listed in AAFTG?

Yes

Yes

No

March 2020

Yes

Yes

No

June 2018

No

Yes

Yes

Sep. 2017

Yes

Yes

No

PACK Pharmaceuticals v. Alza Corporation (9/8/2014; Glucotrol XL) Apotex Corp. v. ViiV Healthcare Co. (8/3/2015; Trizivir, Triumeq, and Epzicom) Metrics, Inc. v. Senju Pharmaceutical Co. Ltd. (7/8/2015; Prolensa) Antares Pharma, Inc. v. Medac (4/30/2015; Rasuvo)

No

Yes

Yes

Yes (but as Perigo41) No

Yes (12/9/2014) Yes (2/19/2015) Yes (1/6/2015)

No

No

No

No

Yes

Yes

No

No

Yes

Yes

Yes (Otrexup)

Rasuvo and Otrexup are substitutes

Agila Specialities Inc. v. Cubist Pharmaceuticals (4/28/2015; Cubicin) Ranbaxy, Inc. v. Adamas Pharmaceuticals (5/15/2015; Namenda Xr &Namzaric) Agila Specialities Inc. v. Cephalon, Inc. (11/16/2015; Treanda) Mylan Pharmaceuticals v. Warner Chilcott (8/20/2015; Loestrin & Minastrin)

No

Yes, but before PTAB Yes

Yes

No

No

Yes

No

No

Yes (7/20/2015) No

Yes

Yes

No

No

Yes

Yes

No

Dr. Reddy's Laboratories v. Fresenius Kabi USA (4/2/2015; Diprivan) Apotex Corp. v. Allergan, Inc. (12/16/2015; Restasis) Dr. Reddy's Laboratories v. Helsinn Healthcare (10/14/2015; Aloxi) Accord Healthcare Inc. v. Helsinn Healthcare (8/31/2016; Aloxi) Ranbaxy, Inc. v. Jazz Pharmaceutical, Inc. (5/23/2016; Xyrem) Par Pharmaceutical, Inc. v. Novartis AG (4/1/2016; Afinitor)

No

Yes

Yes

No

Feb. 2029 (Loestrin); Mar. 2017 (Minastrin) No

No

Yes

Yes

No

No

No

Yes

Yes

Yes

Sep. 2018

No

Yes

Yes

No

No

Yes

Yes

Yes

No

Dec. 2025

Yes

Yes

No

No

Apotex Corp. v. Alcon Pharmaceuticals Ltd. (11/15/2013; Moxeza & Vigamox) Ranbaxy vs Vertex Pharmaceuticals (11/15/2013; Lexiva) Apotex Corp. v. Alcon Research, Ltd. (7/21/2014; Travatan Z) Impax Laboratories v. Meda Pharmaceuticals (7/29/2014; Astepro)

No

No

(4/12/2016)42

No

40

This column addresses specifically parallel litigation between the same parties, i.e. it does not address for infringement litigation in which the defendant is someone other than the PTAB petitioner. 41 Perigo is listed, and it appears to be in a parent-subsidiary relationship with the petitioner. 42 In one IPR, not the other.

20

HOVENKAMP & LEMUS

Mylan Pharmaceuticals Baxter International (11/19/2015; Esmolol and Brevibloc) Mylan Pharmaceuticals v. Senju Pharmaceutical (8/31/2016; Bepreve) Argentum Pharmaceuticals LLC v. Alcon Research Ltd. (11/30/2016; PAZEO) Frontier Therapeutics, LLC v. Medac Gesellschaft Fur Klinische (12/8/2016; Rasuvo) Fresenius Kabi USA LLC v. Cephalon Inc. (8/25/2016; Treanda, Bendeka) Mylan Pharmaceuticals Inc. v. 3M Company (11/18/2016; Flovent HFA)

No

Yes

Yes

Yes

No

No

No

Unknown

No

No

Yes

No

Unknown

No

No

Yes

No

Unknown

No

No

Yes

Yes

Yes

No

No

Yes

No

Unknown

No

No

Mylan Pharmaceuticals Inc. v. Senju Pharmaceutical Co. Ltd. (11/14/2016; Bepreve) Argentum Pharmaceuticals LLC v. Allergan, Inc. (12/8/2016; Restasis) Mylan Pharmaceuticals Inc. v. AstraZeneca AB (1/10/2017; Faslodex)

No

No

Unknown

No

No

No

No

Unknown

No

No

Yes for

Yes

Yes

No

No

No for the others. Yes

Yes

Yes

No

No

Yes

Yes

Yes

No

No

No

No

Unknown

No

No

No

No

Unknown

No

No

No

No

Unknown

No

No

Amneal Pharmaceuticals LLC v. Hospira, Inc. (5/19/2017; Precedex) Mylan Pharmaceuticals Inc. v. ICOS Corporation (8/16/2017; Cialis) Amerigen Pharmaceuticals Limited v. Shire, LLC (5/2/2017; Mydayis) Mylan Pharmaceuticals Inc v. Fresenius Kabi USA, LLC (4/19/2017; Levothyroxine ) Par Pharmaceutical, Inc. v. Sumitomo Dainippon Pharma Co., Ltd. (8/14/2017)

IPR2016-01325

(12/14/2016)

21