THE STAY DILEMMA: EXAMINING BRAND AND GENERIC
INCENTIVES FOR DELAYING THE RESOLUTION OF
The Hatch-Waxman Act encourages generic drug companies to chal-lenge the patents on brand name drugs by awarding the first generic chal-lenger 180 days of generic marketing exclusivity, a bounty often worth mil-lions of dollars. Challenging a patent usually triggers patent infringementlitigation from the brand name firm. In response to the increased numbers ofpatent challenges, brand name firms have adopted an “evergreening” strat-egy—filing for multiple patents for each drug hoping that the generic firmwill not be able to successfully challenge all of the patents and that the con-tinued validity of just one of them will prevent generic entry. Evergreeninginevitably results in patents of different strengths. Generics have responded tothis strategy by challenging only the weaker patents on a drug and thenfiling for a stay of the subsequent patent infringement lawsuit until thestrong patents are about to expire—which is often many years in the future. Generics seek these stays because an early litigation victory would grant aperiod of exclusivity that they could not use because the strong patents con-tinue to block generic entry. Brand name firms should also favor stays be-cause by delaying the generic’s exclusivity period, the stay also ultimatelydelays full competition. Yet in two out of three cases, the brand name firmopposed the stay motion. Courts are split on whether to grant the motions buthave not addressed the anticompetitive consequences of stays. This Note ex-plores the cases, regulatory background, and economic incentives surround-ing the stays and concludes that stay motions should be denied.
The Hatch-Waxman Act1 encourages generic challenges to brand
name pharmaceutical patents. It does so by giving the first generic drugcompany to file for Food and Drug Administration (FDA) approval 180days of generic exclusivity when the company introduces its generic drugon the market in competition with the brand’s drug. Filing with the FDAfor a drug under patent protection, however, usually triggers patent in-fringement litigation by the brand name firm holding the applicable pat-ents. Exclusivity is awarded as long as the generic firm does not lose that
* J.D. Candidate 2012, Columbia Law School. 1. The Hatch-Waxman Act is officially called the Drug Price Competition and Patent
Term Restoration Act of 1984, Pub. L. No. 98-417, 98 Stat. 1585 (codified as amended inscattered sections of 15, 21, 35, and 42 U.S.C.).
patent infringement suit.2 The stakes are high: 180 days of generic exclu-sivity is worth millions of dollars for major drugs.3
One strategic response by brand name firms who seek to prevent
generic competition has been to file an increasingly large number of pat-ents for each of their drugs.4 The theory is that generics will not be ableto invalidate all of the patents and the continued validity of at least one ofthem prevents generic entry. This process—known as “evergreening”—usually leads to patents of varying strengths or “quality.”5
For generics that want to receive the 180 days of exclusivity, this situ-
ation often leads to a dilemma. Suppose there are two patents on a drug. One of them is very strong and therefore not worth challenging.6 Theother is weak. Ordinarily, several generic firms would race to be the firstto challenge the weak one so as to gain the 180-day exclusivity period.7Even if the generic firm were to prevail in the resulting patent litigation,however, the strong patent would continue to block FDA approval of thegeneric drug product for the duration of the strong patent’s term. Thus,there is a risk that the generic firm might win the patent litigation tooearly, resulting in a premature period of exclusivity during which the ge-neric firm still cannot feasibly enter the market. In such a case, the ge-neric firm would ultimately forfeit exclusivity under the Hatch-WaxmanAct’s “failure to market” forfeiture provision.8
Generic firms that want to ensure themselves the 180-day exclusivity
period after the strong patent expires recently have adopted a new strate-gic response. The strategy is filing with the FDA challenging only theweak patent and then asking the court for a stay of the subsequent patentlitigation for many years, until the strong patent is about to expire. Thestay would allow enough time to pass so that a generic victory as to theweak patent would permit the generic firm to launch its drug with exclu-sivity upon the expiration of the strong patent.
2. The simplest such outcome is where the generic firm wins the lawsuit. A generic
can win the suit by successfully claiming that the generic product does not infringe thebrand name patent or the brand name patent is invalid. See infra text accompanying notes45–50 (discussing claims in greater detail). However, the generic firm does not have to
defeat the brand name firm to win exclusivity—it just cannot lose. Generic firms canreceive exclusivity after settlements with brand name firms or where they are not sued. Seeinfra text accompanying notes 246–253 (discussing ways generics have received
3. See infra text accompanying note 56 (discussing value of exclusivity).
4. See infra text accompanying note 72 (discussing this trend).
5. In other words, some patents may be very strong, or of high quality, while others
are weak. See infra Part I.B (discussing patent quality).
6. It might even be sanctionable to challenge an obviously ironclad patent. See, e.g.,
Takeda Chem. Indus., Ltd. v. Mylan Labs., Inc., 549 F.3d 1381, 1391 (Fed. Cir. 2008)(affirming award of $16,800,000 in attorneys’ fees, expenses, and expert fees to brandname firm as sanction after finding generic firm challenged patents on pioglitazone in badfaith).
7. See infra Part I.A.2 (discussing procedure for challenging patents). 8. See infra text accompanying notes 95–102 (discussing forfeiture provision).
This new strategy has anticompetitive consequences. Such a stay
harms consumers because it manipulates the 180-day exclusivity period inorder to exclude competition: During the stay, the untriggered exclusivityperiod serves as a “bottleneck” that prevents subsequent generic chal-lenges to the brand’s patents.9 A stay followed by a generic victory wouldresult in consumers having to pay heightened duopoly prices during theexclusivity period after the strong patent expires without any offsettingsocietal benefit.10 And the stay delays for years the public interest in re-moving weak patents from the market.11 Two courts have split onwhether to grant these stay motions;12 however, they have not grappledwith the anticompetitive consequences of a stay.
District courts confronting these motions lack adequate briefing and
argument of these anticompetitive effects because, while a stay obviouslyserves the generic firm’s interests, it also serves the brand name firm’sinterests. Should the generic firm succeed in obtaining exclusivity follow-ing the expiration of the strong patent, it would mean less competitionfor the brand name firm during the period of exclusivity. The brandwould make much more money in 180 days of duopoly than if full compe-tition commenced immediately after the strong patent expired.13 Thus it
9. In other contexts courts have condemned blatant attempts to manipulate the 180-
day exclusivity period to exclude competitors as antitrust violations. See Ark. CarpentersHealth & Welfare Fund v. Bayer AG, 604 F.3d 98, 106 (2d Cir. 2010) (“[A] plaintiff canhave antitrust claims where a Hatch-Waxman settlement allows the generic manufacturerto manipulate the 180-day exclusivity period in a manner that bars subsequent challengesto the patent or precludes the generic manufacturer from marketing non-infringingproducts unrelated to the patent.”), cert. denied, 131 S. Ct. 1606 (2011); see also In reTamoxifen Citrate Antitrust Litig., 466 F.3d 187, 216–19 (2d Cir. 2006) (agreeingagreement between brand name and generic drug firms to “manipulat[e] [the] exclusivityperiod in order to protect [brand] from competition from other generic manufacturers”would be antitrust violation); In re Cardizem CD Antitrust Litig., 332 F.3d 896, 907–09 (6thCir. 2003) (finding per se antitrust violation where generic firm agreed to delay entry inmarket in exchange for payment from brand name firm and, by not settling underlyingpatent litigation, companies manipulated 180-day exclusivity to prevent other genericsfrom entering market).
10. See infra note 25 and accompanying text (describing dramatic impact on price
resulting from full generic competition, as compared to duopoly pricing); cf. infra textaccompanying notes 250–253 (describing troubling trend of awards of exclusivity without
any corresponding societal benefit).
11. See infra text accompanying notes 75–76 (discussing deleterious effect of weak
patents on society); see also Cardinal Chem. Co. v. Morton Int’l, Inc., 508 U.S. 83, 100–01(1993) (noting “importance to the public at large of resolving questions of patentvalidity”); Edward Katzinger Co. v. Chi. Metallic Mfg. Co., 329 U.S. 394, 400 (1947) (noting“necessity of protecting our competitive economy by keeping open the way for interestedpersons to challenge the validity of patents which might be shown to be invalid”).
12. In Abbott Laboratories v. Matrix Laboratories, Inc., No. 09-cv-1586, 2009 WL
3719214, at *5 (N.D. Ill. Nov. 5, 2009), the stay was granted. In MillenniumPharmaceuticals, Inc. v. Teva Parenteral Medicines, Inc., Nos. 09-cv-105, 09-cv-204, 10-cv-137, 2010 WL 1507655, at *4 (D. Del. Apr. 14, 2010), it was denied.
13. The FDA has found that in a duopoly generated by the 180-day exclusivity period,
the generic drug’s price is on average 94% of the brand name drug’s price. Ctr. for Drug
was not surprising that in one case the drug companies agreed to stay thepatent infringement litigation for four years.14 In another case, the brandname firm waited six years after the generic challenged its patent to initi-ate litigation, thereby achieving the effects of a litigation stay.15
Contrary to their economic incentives, brand name firms opposed
these stay motions in two other cases. Their arguments in opposition,however, were largely boilerplate and failed to reflect the brands’ and thegenerics’ aligned economic incentives. In fact, brand opposition may behalf-hearted because brands do not want to reveal to the court their eco-nomic incentives favoring delay or invite antitrust or regulatory scru-tiny.16 Thus, sole reliance on the brand name firms’ submissions is inade-quate for district courts to evaluate the merits of the stay motions.
This Note addresses the cases, regulatory background, and incentives
surrounding the stay decisions. Part I explains the regulatory backgroundsurrounding pharmaceutical competition and the ways in which drugcompanies have used it to delay generic entry. Part II explores the casesaddressing the stay motions and evaluates the incentives of both brandname and generic firms. Part III offers district courts guidance on how torule on these stay motions and addresses promising legislative proposalsfor removing the incentives of drug companies to collude. This Note con-cludes that these stay motions should be denied because of the stay’santicompetitive consequences and the strong public interest in ensuringprompt adjudication of patent disputes so consumers can enjoy lowerprices from generic competition.
I. THE STATUTORY AND REGULATORY LANDSCAPE SURROUNDING
This Note requires an understanding of the incentive scheme cre-
ated by the Drug Price Competition and Patent Term Restoration Act of1984.17 This Act, more commonly known as the Hatch-Waxman Act, cre-
Evaluation & Research, FDA, Generic Competition and Drug Prices [hereinafter FDA,Generic Competition], http://www.fda.gov/AboutFDA/CentersOffices/CDER/ucm129385.htm (on file with the Columbia Law Review) (last updated Mar. 1, 2010). Whentwo generic drug companies’ products are on the market, the generic drugs are priced atabout half the price of the brand name drug, and that price continues to fall as moregeneric firms enter the market. Id.
14. Joint Agreed Order Staying the Litigation and Tolling Automatic FDA Stay at 2,
Sepracor Inc. v. Teva Parenteral Meds., Inc., No. 10-cv-00746 (S.D.N.Y. July 8, 2010)[hereinafter Brovana Stay Order].
15. See Complaint at 4, Pfizer Inc. v. Teva Pharm. USA Inc., No. 2:10-cv-00128 (E.D.
Va. Mar. 24, 2010) [hereinafter Viagra Complaint] (noting Pfizer received notice that Tevachallenged its patent on Viagra in 2004, but Pfizer did not sue until 2010).
16. See infra Part II.D.2 (analyzing possibility brand opposition is half-hearted). 17. Pub. L. No. 98-417, 98 Stat. 1585 (codified as amended in scattered sections of 15,
21, 35, and 42 U.S.C.). Congress amended this Act with the passage of the MedicarePrescription Drug, Improvement, and Modernization Act of 2003, Pub. L. No. 108-173, tit. XI, subtits. A–B, 117 Stat. 2066, 2448–64 (codified at 21 U.S.C. § 355 (2006)).
ated the regulatory structure that governs competition in the pharmaceu-tical industry. Part I.A evaluates the purposes and history of the Act andthe regulatory framework that it establishes. Part I.B introduces the topicof “patent quality”—the notion that approved patents are of differentstrengths and that the Patent and Trademark Office (PTO) routinelyapproves even weak patents. Part I.C addresses the so-called “pay for de-lay” settlement problem, where brand name firms pay generic firms todelay or forgo introducing a competing generic drug on the market. Finally, Part I.D discusses a recent statutory change to the 180-day exclu-sivity period, which has led to the stay motions later discussed in Part II.
A. The Hatch-Waxman Act’s Regulatory and Competition Framework
1. The Act’s Purposes and Effects. — The Hatch-Waxman Act “was an
unprecedented attempt to achieve two seemingly contradictory objec-tives, namely, 1) to make lower-costing generic copies of approved drugsmore widely available and 2) to assure that there were adequate incen-tives to invest in the development of new drugs.”18 It did so by establish-ing a regime to streamline generic challenges to brand name pharmaceu-tical patents while also extending the life of those patents to compensatefor the time necessary to satisfy FDA drug safety and efficacy require-ments.19 One of the Act’s main purposes was to encourage generic com-petition to bring down the price of prescription drugs.20 The Hatch-Waxman Act sought to foster generic competition for drugs that were offpatent,21 as well as drugs that were protected by one or more patents.22
Over the next fifteen years, the Hatch-Waxman framework was suc-
cessful in encouraging the development and introduction of genericdrugs. For example, in 1998 the Congressional Budget Office (CBO) re-leased a study showing that by 1996, generic drugs accounted for 42.6%of all drugs sold, compared to only 18.6% at the time the Act waspassed.23 The CBO also reported that whereas before the Act was passedonly about one-third of off-patent drugs had generic versions available, by
18. Alfred B. Engelberg, Special Patent Provisions for Pharmaceuticals: Have They
Outlived Their Usefulness?, 39 IDEA 389, 389 (1999).
19. See id. at 390–91 (explaining general framework of Hatch-Waxman Act). 20. See H.R. Rep. No. 98-857, pt. 1, at 14 (1984), reprinted in 1984 U.S.C.C.A.N.
2647, 2647 (noting Act sought “to make available more low cost generic drugs byestablishing a generic drug approval procedure for pioneer drugs”).
21. Cf. id. at 17, 1984 U.S.C.C.A.N. at 2650 (lamenting there were “approximately 150
drugs” off patent but for which there was no generic equivalent).
22. In addition to streamlining generic challenges to brand name patents, the Act
also provided a safe harbor provision which allowed generic companies to experiment withand develop generic alternatives to drugs under patent protection without the fear ofpatent infringement litigation, unless the generic sought to sell the drug to the public. 35U.S.C. § 271(e)(1) (2006).
23. Cong. Budget Office, How Increased Competition from Generic Drugs Has
Affected Prices and Returns in the Pharmaceutical Industry 27 (1998) [hereinafter CBO,Increased Competition], available at http://www.cbo.gov/ftpdocs/6xx/doc655/pharm.pdf (on file with the Columbia Law Review).
1998, virtually all of them did.24 The introduction of generic drugs incompetition with brand name drugs has a profound and lasting impacton price. The Federal Trade Commission (FTC) has estimated thatwithin one year after the introduction of generic competition, the aver-age price of the drug falls 77%.25 Thus, by 1994, generic substitutionsaved consumers roughly $8 to $10 billion a year26—far more than theHouse Report estimated when the Hatch-Waxman Act was introduced.27
In addition, the Hatch-Waxman framework was successful in facilitat-
ing challenges to weak patents. Between 1984 and 2003, generic drugfirms challenged patents on over 200 drugs,28 and did so at an increasingrate.29 These challenges were successful in securing generic competitionprior to patent expiration on many drugs. An FTC study analyzing alleighty-two resolved challenges to drug patents between 1992 and 2000found that the challenges produced pre-expiration competition for morethan half of the drugs.30 That included at least four of the top ten best-selling drugs in 2000.31
The Hatch-Waxman Act was also successful in preserving the incen-
tives of brand name firms to continue investing in the development ofnew drugs. Throughout the 1990s, brand name firms’ profits dramatically
24. Id. at 37. 25. Fed. Trade Comm’n, Pay-for-Delay: How Drug Company Pay-Offs Cost Consumers
Billions 8 (2010) [hereinafter FTC, Pay-for-Delay Study]. The FTC determined that it takesaround one year after the introduction of generic competition for the lower-priced genericdrugs to achieve market-wide penetration. See id. (noting within one year “genericpenetration rate” reaches 90%, meaning “pharmacists fill 90 [out] of every 100prescriptions for the molecule with . . . [a bioequivalent] generic”). The generic drugproduct is usually around 85% cheaper than the brand name drug, but for whateverreason, 10% of consumers stay with the more expensive brand name drug. Id.; see alsoFDA, Generic Competition, supra note 13 (showing the more different generic versions of
a drug are on the market, the lower the price).
26. CBO, Increased Competition, supra note 23, at 31.
27. See H.R. Rep. No. 98-857, pt. 1, at 17 (1984), reprinted in 1984 U.S.C.C.A.N.
2647, 2650 (estimating Hatch-Waxman framework would save consumers only $920 millionover twelve years following Act’s passage).
28. C. Scott Hemphill, Paying for Delay: Pharmaceutical Patent Settlement as a
Regulatory Design Problem, 81 N.Y.U. L. Rev. 1553, 1567 (2006) [hereinafter Hemphill,Paying for Delay].
29. See Fed. Trade Comm’n, Generic Drug Entry Prior to Patent Expiration 10
(2002) [hereinafter FTC, Generic Drug Study], available at http://www.ftc.gov/os/2002/07/genericdrugstudy.pdf (on file with the Columbia Law Review) (finding 104 of 130 drugschallenged between 1984 and 2001 were challenged after 1992).
30. The FTC found that of the 104 total challenges, the brand name firm declined to
sue the generic in twenty-nine cases, allowing immediate generic entry. Id. at 15 fig.2-1. Ofthe remaining seventy-five challenges, fifty-three had been resolved by the end of the study. Id. Of those, the generic firm won twenty-two. Id. The brand name firm only won eightcases. Id.; see also Hemphill, Paying for Delay, supra note 28, at 1567 (describing these
31. See Hemphill, Paying for Delay, supra note 28, at 1567 n.57 (reporting that Paxil,
Prilosec, Prozac, and Zocor have had pre-expiration competition).
increased.32 Moreover, “the re-investment of those profits in research,both in total dollars and as a percentage of sales” was the highest inhistory.33
By the early 2000s, however, a new and troubling trend emerged that
continues to threaten the initial success of the Hatch-Waxman Act. Brandname and generic firms have increasingly settled patent infringement liti-gation with settlement provisions that include a large payment from thebrand name firm to the generic in exchange for the generic agreeing todelay or forgo introducing its competing generic drug on the market.34 A2010 FTC study found that these settlements delay the introduction ofgeneric competition by a median of around one and a half years and costAmerican consumers at least $3.5 billion per year.35 Another estimatefound that a one-year delay in generic entry across twenty-one drugs costsconsumers about $14 billion.36 These kinds of settlements are unheard ofoutside of the Hatch-Waxman context.
2. The Hatch-Waxman Act’s Regulatory Framework. — Under the Hatch-
Waxman Act, a pharmaceutical manufacturer must receive approval tomarket a new drug by filing what is known as a “New Drug Application”(NDA) with the FDA.37 The NDA is essentially a demonstration that thedrug is safe and effective.38 Making this demonstration requires lengthyclinical trials, which cost many millions of dollars.39 The innovator firmmust list in the NDA any patents that it holds that it believes could beasserted against the unauthorized manufacture, sale, or use of the drug.40These patents are listed in an FDA compendium called Approved DrugProducts with Therapeutic Equivalence Evaluations, more commonlyknown as the Orange Book.
32. From 1991 to 1999, the seven largest brand name pharmaceutical companies saw
their market capitalization increase by $655 billion, a gain of 536%. Engelberg, supra note18, at 390 n.3.
33. Id. at 390. 34. See infra Part I.C (discussing problem in greater detail). Congressman Henry A.
Waxman, who coauthored the Hatch-Waxman Act, lamented that these settlement dealshave “turned [the Act] on its head. We were trying to encourage more generics andthrough different business arrangements, the reverse has happened.” Sheryl Gay Stolberg& Jeff Gerth, How Companies Stall Generics and Keep Themselves Healthy, N.Y. Times,July 23, 2000, at A1.
35. FTC, Pay-for-Delay Study, supra note 25, at 10.
36. C. Scott Hemphill, An Aggregate Approach to Antitrust: Using New Data and
Rulemaking to Preserve Drug Competition, 109 Colum. L. Rev. 629, 650 (2009)[hereinafter Hemphill, Aggregate Approach].
37. 21 U.S.C. § 355 (2006). 38. The safety and efficacy requirement is mandated by the Federal Food, Drug, and
39. 21 U.S.C. § 355(b). For costs of this process, see Joseph A. DiMasi, Ronald W.
Hansen & Henry G. Grabowski, The Price of Innovation: New Estimates of DrugDevelopment Costs, 22 J. Health Econ. 151, 162–65 & tbl.1 (2003) (estimating that clinicaltests cost on average $130 million).
The Hatch-Waxman Act attempted to streamline the introduction of
generic drugs by providing that once the NDA is approved, a genericdrug manufacturer would not have to submit an NDA before marketing ageneric version of the drug. Instead, it can file a considerably cheaperand less time-consuming application called an “Abbreviated New DrugApplication” or ANDA.41 An ANDA requires the generic firm to demon-strate primarily that its product is a bioequivalent to the brand name drugit seeks to compete against without having to conduct an independentdetermination as to the safety and efficacy of the drug.42
An ANDA contains a variety of certifications with respect to the pat-
ents on the brand name drug. A “Paragraph III” certification means thatthe generic firm seeks approval to market its generic drug only after therelevant patents have expired.43 Meanwhile, a “Paragraph IV” certifica-tion expresses the generic firm’s desire to market its drug before the rele-vant patents have expired because the generic believes “the patent or pat-ents in question are invalid or not infringed by the generic product.”44
A patent could be invalid because it was inequitably procured,45 was
inherently anticipated by the prior art already on the market,46 or be-
41. See Requirements for Submission of In Vivo Bioequivalence Data, 68 Fed. Reg.
61,640, 61,645 (proposed Oct. 29, 2003) (to be codified at 21 C.F.R. pts. 314, 320)(estimating costs of filing ANDA between $300,000 and $1 million). However, litigationexpenses can raise the expense of an ANDA to around $10 million. Marc Goodman, GaryNachman & Louise Chen, Morgan Stanley Equity Research, Quantifying the Impact fromAuthorized Generics 9 (2004).
42. 21 U.S.C. § 355(j)(2)(A), (8)(B). In addition to demonstrating bioequivalence as
defined in subsection (8)(B), the ANDA also must demonstrate that its proposed genericdrug has the same indications, active ingredient, route of administration, dosage form,strength, and labeling. Id. § 355(j)(2)(A)(i)–(v); see also Hemphill, Paying for Delay,supra note 28, at 1565 (explaining ANDA requirements).
43. 21 U.S.C. § 355(j)(2)(A)(vii)(III); see also Michael R. Herman, Consulting
Legislative History to Condemn Pay-for-Delay Settlements, Bright Ideas, Spring/Summer2011, at 19, 22 (explaining Paragraph III certification is “acknowledge[ment] that one ormore patents exist to block generic entry until patent expiration”).
44. Herman, supra note 43, at 22; see 21 U.S.C. § 355(j)(2)(A)(vii)(IV). Paragraphs I
and II permit immediate approval on the grounds that the brand name firm has not filedrequired information with the FDA or the patents in question have already expired. See id. § 355(j)(2)(A)(vii)(I)–(II).
45. See, e.g., Aventis Pharma S.A. v. Hospira, Inc., 743 F. Supp. 2d 305, 350–51 (D.
Del. 2010) (discussing this type of claim). Under Federal Circuit precedent, a partyalleging inequitable conduct before the PTO must establish by clear and convincingevidence that the patent applicant “(1) made an affirmative misrepresentation of materialfact, failed to disclose material information, or submitted false material information to thePTO; and (2) intended to deceive the PTO.” Id. at 350 (citing Star Scientific, Inc. v. R.J. Reynolds Tobacco Co., 537 F.3d 1357, 1365 (Fed. Cir. 2008)).
46. Under the Patent Act, a patent cannot be obtained “if the differences between the
subject matter sought to be patented and the prior art are such that the subject matter as awhole would have been obvious at the time the invention was made to a person havingordinary skill in the art.” 35 U.S.C. § 103(a) (2006) (emphasis added). Patents arepresumed to be nonobvious. Structural Rubber Prods. Co. v. Park Rubber Co., 749 F.2d707, 714 (Fed. Cir. 1984). A defendant can rebut that presumption by demonstrating “byclear and convincing evidence that a skilled artisan would have been motivated to combine
cause the drug testing violated the public use bar.47 Alternatively, the ge-neric drug product might not violate the patent because the generic man-ufacturer has devised a way to create a bioequivalent drug by a differentprocess,48 or has created a different structure of the same active ingredi-ent,49 or a different delivery mechanism.50
Filing an ANDA with a Paragraph IV certification is considered an
act of patent infringement, and the brand name firm may sue the genericfirm, even though no generic drug has appeared on the market.51 TheHatch-Waxman Act encourages the brand name firm to sue the genericfirm promptly by providing that if the brand name firm sues within forty-five days of notification of a Paragraph IV-certified ANDA, the suit auto-matically stays the date on which the FDA may approve the ANDA forthirty months unless the patent expires or is found to be invalid beforethen.52 If the brand name firm declines to sue, the FDA may approve thegeneric firm’s ANDA prior to patent expiration and the generic firm maymarket its drug shortly thereafter.
the teachings of the prior art references to achieve the claimed invention, and that theskilled artisan would have had a reasonable expectation of success in doing so.” Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1361 (Fed. Cir. 2007). This is generally a hard claim to winwhen the PTO was aware of the prior art because “[w]hen the prior art was before theexaminer during prosecution of the application, there is a particularly heavy burden inestablishing invalidity.” Impax Labs., Inc. v. Aventis Pharm., Inc., 468 F.3d 1366, 1378 (Fed. Cir. 2006). “The rationale underlying the presumption of validity is, however, ‘muchdiminished’ when the prior art relied upon was not considered by the PTO duringprosecution.” Eli Lilly & Co. v. Sicor Pharm., Inc., 705 F. Supp. 2d 971, 989 (S.D. Ind. 2010) (quoting KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 426 (2007)); see also id. at 997(holding defendants failed to establish plaintiff’s patents for the cancer drug Gemzar wereanticipated by prior art).
47. A patent claim is not valid if “the invention was . . . in public use . . . in this
country, more than one year prior to the date of the application for patent in the UnitedStates.” 35 U.S.C. § 102(b); see also, e.g., SmithKline Beecham Corp. v. Apotex Corp., 365F.3d 1306, 1308 (Fed. Cir. 2004) (invalidating patent for violating public use bar of 35U.S.C. § 102(b) during clinical trials), reh’g granted in part and vacated, 403 F.3d 1328(Fed. Cir.) (en banc), remanded to 403 F.3d 1331 (Fed. Cir. 2005).
48. See, e.g., In re Omeprazole Patent Litig., 281 F. App’x 974, 977 (Fed. Cir. 2008)
(affirming district court’s finding of noninfringement where generic firm was able to getaround patent of alkaline reacting compound that protected active ingredient fromdegradation by creating noninfringing “talc”).
49. See, e.g., SmithKline Beecham Corp. v. Apotex Corp., 247 F. Supp. 2d 1011, 1023
(N.D. Ill. 2003) (describing such a noninfringement claim), aff’d, 365 F.3d 1306 (Fed. Cir. 2004), reh’g granted in part and vacated, 403 F.3d 1328 (Fed. Cir.) (en banc), aff’d onother grounds on remand, 403 F.3d 1331 (Fed. Cir. 2005).
50. See, e.g., Alza Corp. v. Andrx Pharm., LLC, 607 F. Supp. 2d 614, 629 (D. Del.
2009) (finding generic version of drug to treat Attention Deficit Hyperactivity Disorder didnot infringe patent because patent only pertained to time-release tablet that released drugwithin one hour whereas generic version took more than ninety minutes), aff’d on othergrounds, 603 F.3d 935 (Fed. Cir. 2010).
51. 35 U.S.C. § 271(e)(1)–(2). 52. 21 U.S.C. § 355(j)(5)(B)(iii) (2006). This thirty-month stay, however, can actually
last more than three years. Hemphill, Paying for Delay, supra note 28, at 1566 & n.50
(discussing ways in which thirty-month stay can last for longer than three years).
The Hatch-Waxman Act also encourages the filing of Paragraph IV-
certified ANDAs by granting the first such filer the opportunity to marketa generic version of the drug for 180 days without competition from anyother generic manufacturer.53 The exclusivity period begins on the datethat the FDA approves the generic firm’s ANDA, which usually occursafter a final court decision finding the relevant patents invalid or not in-fringed.54 FDA approval also occurs if the brand name firm declines tosue the first filer or after a litigation settlement.55
This 180-day exclusivity period—effectively a mini-patent—is a
“bounty worth hundreds of millions of dollars” for major drugs.56 In fact,“most generic drug companies estimate that 60% to 80% of their poten-tial profit for any one product is made during this exclusivity period.”57Perhaps because of the large amount of money involved, “[i]t is widelyunderstood that the 180-day exclusivity period offers the potential for col-lusive settlement arrangements between pioneers and generics.”58 To un-derstand these abuses, it is first important to understand the concept of“patent quality.”
A main motivator behind the Hatch-Waxman Act was the realization
that many patents preventing or delaying generic entry were of poor qual-ity.59 A patent is considered to be of poor quality if it was improperlyissued by the Patent and Trademark Office.60 Thus, it is important tounderstand the process by which patents are granted, and their relativestrengths.
53. 21 U.S.C. § 355(j)(5)(B)(iv). 54. Hemphill, Paying for Delay, supra note 28, at 1578.
55. Id. at 1579. 56. Id. at 1560 (internal quotation marks omitted). 57. Daniel F. Coughlin & Rochelle A. Dede, Hatch-Waxman Game-Playing from a
Generic Manufacturer Perspective: From Ticlid to Pravachol, Apotex Has DifficultyTelling Who’s on First, 25 Biotechnology L. Rep. 525, 525–26 (2006); see also, e.g.,Gardiner Harris & Joanna Slater, Bitter Pills: Drug Makers See “Branded Generics” Eatinginto Profits, Wall St. J., Apr. 17, 2003, at A1 (reporting Barr’s generic version of Prozacearned $366 million in revenue during its 180-day exclusivity period, and only $4 million insubsequent 180 days).
58. Herbert Hovenkamp, Mark Janis & Mark A. Lemley, Anticompetitive Settlement
of Intellectual Property Disputes, 87 Minn. L. Rev. 1719, 1755 (2003).
59. See S. Rep. No. 107-167, at 4 (2002) (“Under Hatch-Waxman, manufacturers of
generic drugs are encouraged to challenge weak or invalid patents on brand name drugsso consumers can enjoy lower drug prices.”).
60. See Mark A. Lemley & Carl Shapiro, Probabilistic Patents, J. Econ. Persp., Spring
2005, at 75, 77 (defining low-quality patents as those “patents improperly issued”). Thereare many reasons why a patent might be improperly issued. See supra notes 45–47 and
accompanying text (describing ways in which patent may be invalid). A valid patent conferson the innovator “the right to exclude others from making, using, offering for sale, orselling the invention throughout the United States or importing the invention into theUnited States” for a term of twenty years. 35 U.S.C. § 154(a)(1)–(2) (2006).
As Mark Lemley and Carl Shapiro explain, “[n]early 200,000 patents
are issued every year after a very limited examination process.”61 Thestructure of the patent review process ensures that almost every patentapplication is approved.62 The review process is fairly cursory: “[A] patentexaminer spends only 18 hours per application on average.”63 Moreover,a patent will be approved unless the PTO can provide a reason not toissue the patent.64 Additionally, “structural problems [exist] that en-courage the PTO to grant patents of doubtful quality, including high ex-aminer turnover and an incentive system that rewards examiners for al-lowing but not for rejecting applications.”65 Finally, the PTO has tocontend with the “unprecedented explosion” of patent applications inrecent years.66 Thus, there remains a great risk that the PTO issues “pat-ents of questionable validity.”67
Perhaps as a result, “[r]oughly half of all litigated patents are found
to be invalid, including some of great commercial significance.”68 Thus,economists refer to patents as “probabilistic” since they do not confer anabsolute privilege to exclude alleged infringers but only a right to try toexclude through uncertain litigation.69 The FTC has found that in patentlitigation between brand name firms and generics under the Hatch-Waxman framework, generics prevailed in 73% of cases.70 This is notable
61. Lemley & Shapiro, supra note 60, at 75.
62. The estimated “Grant Rate”—“the number of applications that were granted
during the reporting period, divided by the number of disposals in the reporting period(applications granted plus those abandoned)”—was 98% in 2000. See Cecil D. Quillen, Jr.,Ogden H. Webster & Richard Eichmann, Continuing Patent Applications andPerformance of the U.S. Patent and Trademark Office—Extended, 12 Fed. Cir. B.J. 35, 46,47 (2003) (internal quotations omitted). This is considerably higher than in Europe (67%)and Japan (64%). Fed. Trade Comm’n, To Promote Innovation: A Proper Balance ofCompetition and Patent Law and Policy, ch. 5, at 6 (2003) [hereinafter, FTC, To PromoteInnovation], available at http://www.ftc.gov/os/2003/10/innovationrpt.pdf (on file withthe Columbia Law Review).
63. Lemley & Shapiro, supra note 60, at 79.
64. Id. at 78. 65. Id. at 79. The compensation regime for patent examiners is composed of a base
salary and a bonus; bonus points are accumulated only for “dispositions”—i.e., finalallowances or rejections of patents—but since final rejections do not usually result in theend of the examination due to postfinal actions and amendments, “the only way to earnbonus points with confidence is to allow a patent application.” Robert P. Merges, As Manyas Six Impossible Patents Before Breakfast: Property Rights for Business Concepts andPatent System Reform, 14 Berkeley Tech. L.J. 577, 607 (1999).
66. FTC, To Promote Innovation, supra note 62, ch. 5, at 4. The FTC reported that, in
the twelve years leading up to their study, applications had doubled and were increasing byabout 10% per year. Id.
67. Id. at 31. 68. Lemley & Shapiro, supra note 60, at 76.
69. C. Scott Hemphill & Bhaven N. Sampat, When Do Generics Challenge Drug
Patents? 6 (Columbia Law & Econ. Working Paper No. 379, 2011), available at http://ssrn.com/abstract=1640512 (on file with the Columbia Law Review).
70. FTC, Generic Drug Study, supra note 29, at 13. Of those decisions, “there were
slightly more non-infringement decisions (14) than invalidity ones (11).” Id.
given the high burden of proving patent invalidity: Patents are afforded a“presumption of validity” and require “clear and convincing evidence” todeclare them invalid.71
Brand name firms have responded to this trend by applying for more
patents per drug for high-earning drugs and listing additional patents inthe Orange Book even after a generic firm files its Paragraph IV-certifiedANDA.72 This strategy is referred to as “evergreening.”73 Many of thesepatents do not cover the compound at issue. Rather, they are “ancillarypatents on chemical variants, alternative formulations, methods of use,and relatively minor aspects of the drug.”74 Many such patents tend to beweak. Even though they are weak, they “have the effect of extending thetotal duration of protection for a brand-name drug, compared to the pro-tection offered by the compound patent alone.”75 Accordingly, theseweak patents have a deleterious effect on social welfare.76 As discussed inthe next section, they also provide an incentive for collusive settlementsof patent litigation.77
C. The Pay for Delay Settlement Problem
The Hatch-Waxman Act has been remarkably effective in encourag-
ing generic challenges to brand name firms’ patents prior to expiration. Since the generic firm has usually not launched its drug before the con-clusion of patent infringement litigation, it would not have to pay dam-ages should it lose. As a result, a generic firm can take a flier on a brandname firm’s drug with little downside other than litigation costs. And theupside is huge: If the generic firm was the first to file the Paragraph IV-certified ANDA, it would receive the 180-day generic exclusivity period,
71. See, e.g., Iovate Health Scis., Inc. v. Bio-Engineered Supplements & Nutrition,
Inc., 586 F.3d 1376, 1380 (Fed. Cir. 2009) (“Patents enjoy a presumption of validity and aparty seeking to invalidate a patent must overcome this presumption by facts supported byclear and convincing evidence.” (citation omitted)).
72. FTC, Generic Drug Study, supra note 29, at 39–40.
73. See Rebecca S. Eisenberg, The Role of the FDA in Innovation Policy, 13 Mich.
Telecomm. & Tech. L. Rev. 345, 354 (2007).
74. Hemphill & Sampat, supra note 69, at 4.
75. Id. 76. See FTC, To Promote Innovation, supra note 62, ch. 5, at 31 (“Issuance of patents
of questionable validity may affect competition and innovation by discouraging entry andresearch efforts, inducing unnecessary licenses and royalty payments, and imposinglitigation costs.”); Joseph Farrell & Carl Shapiro, How Strong Are Weak Patents?, 98 Am. Econ. Rev. 1347, 1362 (2008) (“With unrestricted two-part tariffs, weak patents can be usedto impose high per-unit royalties (along with negative fixed fees), raising downstreammarginal costs and thus moving the downstream price closer to the monopoly price.”).
77. See In re Tamoxifen Citrate Antitrust Litig., 466 F.3d 187, 211 (2d Cir. 2006)
(“[W]eak patent cases will likely be settled even though such settlements will inevitablyprotect patent monopolies that are, perhaps, undeserved.”); id. at 212 (stating reversepayment settlements can lead to “the survival of monopolies created by . . . fatally weakpatents”). For a discussion of these settlements, see infra Part I.C.
worth millions of dollars for major drugs.78 These challenges are now thenorm for major drugs.79
Once the brand name firm sues the generic firm, it has little to gain
and much to lose from litigating through to judgment. As explainedabove, even a litigation victory would not likely yield damages because thegeneric firm has yet to market its generic product. Also, a litigation vic-tory would not stop a different generic manufacturer from subsequentlyfiling its own ANDA challenging the same patents.80 Meanwhile, a courtruling of noninfringement or invalidity in favor of the generic firm wouldpermit all generic firms to enter the market, causing the brand namefirm to lose its ability to charge monopoly prices for the drug.
For the generic firm, even a litigation victory is not the most
favorable outcome. Due to the brand name firm’s ability to charge mo-nopoly prices for its drug, if the patent is found invalid “the total profits ofthe patent holder and the generic manufacturer on the drug in the com-petitive market will be lower than the total profits of the patent holderalone under a patent-conferred monopoly.”81 Accordingly, it is often inthe brand name firm’s economic interest “to pay some portion of thatdifference to the generic manufacturer to maintain the patent-monopoly
78. Hemphill, Paying for Delay, supra note 28, at 1560.
79. Id. at 1567. In fact, since 1984, generic firms have filed ANDAs with Paragraph IV
certifications over 200 times. Id. at 1567; see also FTC, Generic Drug Study, supra note 29,
at 10 (finding between 1984 and 2000 generic firms filed Paragraph IV-certified ANDAs for130 drugs and filing of Paragraph IV-certified ANDAs is increasing at accelerated rate). That includes challenges to nine of the top ten best-selling drugs in 2000. Hemphill,Paying for Delay, supra note 28, at 1567.
80. That is because a court ruling only applies to the specific ANDA at issue—in many
cases, a brand name firm has to fend off challenges from multiple generic manufacturersin succession. For example, Bayer successfully defended its patent on the antibiotic Ciproagainst four different generic manufacturers. See In re Ciprofloxacin HydrochlorideAntitrust Litig., 261 F. Supp. 2d 188, 197 (E.D.N.Y. 2003) (“[F]our generic companies havefiled [Paragraph IV-certified ANDAs] for Cipro . . . ; one challenge was dismissed, andthree challenges were unsuccessful.” (citing Bayer AG v. Schein Pharm., Inc., 301 F.3d1306 (Fed. Cir. 2002) (finding for Bayer in consolidated action); Bayer AG v. CarlsbadTech., Inc., No. Civ. 01-867-B, 2001 WL 34125673 (S.D. Cal. Oct. 24, 2001) (denyingCarlsbad’s motion for summary judgment); Bayer AG v. Ranbaxy Pharm., Inc., No. 98 Civ. 4464 (D.N.J. Oct. 29, 1999) (dismissing case per stipulation after Ranbaxy withdrew itsParagraph IV certification)). Collateral estoppel does not apply since the defendants areall different parties.
81. Tamoxifen, 466 F.3d at 209; see also In re Schering-Plough Corp., 136 F.T.C. 956,
989 (2003) (“The anticipated profits of the patent holder in the absence of genericcompetition are greater than the sum of its profits and the profits of the generic entrantwhen the two compete.”), vacated, 402 F.3d 1056 (11th Cir. 2005). A brand name firmoperating under a patent-conferred monopoly can set prices significantly abovecompetitive levels by restricting its output, and that output restriction cannot be offset byother firms who are blocked by the patent. This leads to greater market-wide profits for themonopolized good than would exist under full competition. For the general effects ofmonopoly pricing, see 2B Phillip E. Areeda et al., Antitrust Law ¶ 403, at 7–9 (3d ed. 2007).
market for itself.”82 And because the payment is often larger than thegeneric firm’s expected gain from the litigation, “it [makes] obvious eco-nomic sense for the generic manufacturer to accept such a payment if itis offered” and agree to end its patent challenge and not compete withthe brand name firm for some or all of the remaining term of thepatent.83
These payments, often called “reverse payment” or “pay for delay”
settlements, are “typically in the tens or hundreds of millions of dol-lars.”84 The result of these arrangements is that brand name and genericdrug companies split the monopoly profits of the drugs at the expense ofconsumers, who are denied the benefits of generic entry.85 The FTC esti-mated that these settlements “cost American consumers $3.5 billion peryear.”86
These settlements have attracted significant antitrust scrutiny from
both private plaintiffs and the government. A large reverse payment set-tlement tends to indicate that the patent is weak and will likely be invali-dated—especially since generic firms win close to three-quarters of thecases that proceed to judgment.87 As a result, plaintiffs have argued thatthese settlements are properly condemned under the antitrust laws asways for the brand name firm to pay the generic firm to stay out of themarket, resulting in significant harm to consumer welfare. The FTC andthe DOJ believe that these settlements are presumptively unlawful undersection 1 of the Sherman Act.88 Antitrust litigation over these settlements,however, “has resulted in a general, though not across-the-board, rejec-tion of antitrust liability.”89 Perhaps owing to the courts’ largely permis-
82. Tamoxifen, 466 F.3d at 209. 83. Id. 84. Hemphill, Paying for Delay, supra note 28, at 1568; see, e.g., Ark. Carpenters
Health & Welfare Fund v. Bayer AG, 604 F.3d 98, 102 n.8 (2d Cir. 2010) (payment of $398million), cert. denied, 131 S. Ct. 1606 (2011); Schering-Plough, 402 F.3d at 1068 (payment of$60 million).
86. FTC, Pay-for-Delay Study, supra note 25, at 2. The FTC calculates this number by
multiplying the 77% savings obtained by consumers after generic entry by the $3.2 billionin drug sales that will be affected by reverse payment settlements per year, and in turnmultiplying that product by 1.42 years, the median delay achieved by such settlements. That yields $3.5 billion per year. See id. at 8–10 (detailing this calculation).
87. See FTC, Generic Drug Study, supra note 29, at vi (“Generic applicants have
prevailed in 73 percent of the cases in which a court has resolved the patent dispute.”).
88. Brief for the United States in Response to the Court’s Invitation at 21–27, Ark.Carpenters, 604 F.3d 98 (No. 05-2851-cv(L)) [hereinafter DOJ Cipro Brief]; Plaintiff FederalTrade Commission’s Memorandum in Opposition to Defendant Cephalon’s Motion toDismiss at 12–13, FTC v. Cephalon, Inc., 702 F. Supp. 2d 514 (E.D. Pa. 2010) (No. 08-cv-2141-MSG).
89. Herman, supra note 43, at 24; compare Ark. Carpenters, 604 F.3d at 110 (finding
no liability under Tamoxifen standard), In re Ciprofloxacin Hydrochloride Antitrust Litig.,544 F.3d 1323, 1336 (Fed. Cir. 2008) (holding agreements presumptively lawful providedlitigation was not a sham, there was no fraud on Patent and Trademark Office, andsettlement was within exclusionary zone of patent), and In re Tamoxifen Citrate Antitrust
sive view of these settlements, the number of pay for delay settlements hasmore than doubled from 2007 to 2010.90
D. The Medicare Modernization Act and Its Effect on Exclusivity
In 2003, the Hatch-Waxman framework changed following the pas-
sage of the Medicare Prescription Drug, Improvement, and Moderniza-tion Act, more commonly known as the Medicare Modernization Act(MMA).91 The most significant changes were to the 180-day exclusivityperiod.
First, the MMA changed the way that the FDA awarded generic ex-
clusivity, providing that only one generic firm could receive the 180-dayexclusivity period for a given drug.92 Prior to the passage of the MMA,multiple generic firms could receive exclusivity based on different patentson the same drug. In other words, the FDA could award 180-day exclusiv-ity periods to each ANDA applicant who was the first to submit aParagraph IV certification for a different patent on the same drug.93 Thiscreated overlapping exclusivity periods—called “mutual blocking exclu-sivities”—which meant that no generic firm could launch its drug until allof the overlapping exclusivity periods expired, further delaying genericentry.94 The change from a patent-by-patent to a drug-product-based ex-clusivity award was thought to prevent mutual blocking exclusivities byproviding that only one generic drug firm could receive exclusivity perdrug.
Litig., 466 F.3d 187, 208 (2d Cir. 2006) (stating agreements presumptively lawful providedinfringement litigation is not sham and scope of settlement does not exceed exclusionaryzone of patent), with Valley Drug Co. v. Geneva Pharm., Inc., 344 F.3d 1294, 1311–12(11th Cir. 2003) (applying rule of reason analysis “requir[ing] consideration of the scopeof the exclusionary potential of the patent, the extent to which these provisions of the[a]greements exceed that scope, and the anticompetitive effects thereof”), and In reCardizem CD Antitrust Litig., 332 F.3d 896, 900 (6th Cir. 2003) (finding agreement per seunlawful, at least in case where agreement did not settle patent litigation). The SupremeCourt has never opined on the appropriate standard to apply and has denied certioraripetitions in all of the above cases where a petition has been filed.
90. See Fed. Trade Comm’n, Agreements Filed with the Federal Trade Commission
Under the Medicare Prescription Drug, Improvement, and Modernization Act of 2003:Overview of Agreements Filed in FY 2010, at 2 (2011) (finding thirty-one pay-for-delaysettlements in 2010, compared with only fourteen in 2007).
91. Pub. L. No. 108-173, 117 Stat. 2066 (codified as amended in scattered titles of the
92. See 21 U.S.C. § 355(j)(5)(B)(iv)(I) (2006) (awarding exclusivity only to “first
applicant”); id. § 355(j)(5)(B)(iv)(II)(bb) (defining “first applicant” as first firm to submitParagraph IV-certified ANDA for a“drug” rather than a patent).
93. See Apotex Inc. v. FDA, 414 F. Supp. 2d 61, 72–74 (D.D.C. 2006) (granting
Chevron deference to FDA’s interpretation of 21 U.S.C. § 355(j)(5)(B)(iv) as providingseparate exclusivity for separate patents), aff’d, 226 F. App’x 4 (D.C. Cir. 2007) (percuriam).
94. See Shashank Upadhye, Generic Pharmaceutical Patent and FDA Law § 13:4
(2011) (illustrating mutual blocking exclusivity phenomenon).
Second, the MMA introduced a new forfeiture procedure whereby a
generic firm would forfeit its exclusivity period under certain circum-stances.95 The forfeiture procedure was added in response to provisionsin reverse payment settlements between brand name firms and first ge-neric challengers that blocked the FDA from approving subsequent ge-neric challengers’ ANDAs. As noted, the FDA cannot approve a subse-quently filed Paragraph IV-certified ANDA until the first filer’s 180-dayexclusivity period expires.96 The exclusivity period is not triggered untilthe first filer either introduces its generic drug product or there is a judi-cial determination of invalidity or noninfringement.97 Drug companiesstructured their reverse payment settlements so that the first-filing ge-neric firm would not market its generic drug and the patent infringe-ment litigation would end so that there would not be a judicial determi-nation of invalidity or noninfringement.98 Accordingly, the exclusivityperiod would not begin and no subsequent generic filer could gain FDAapproval of its ANDA. Generic entry would be “blocked entirely.”99 Thisis often referred to as an “approval” or “statutory” bottleneck.100
The forfeiture procedure was intended to remedy this bottleneck. It
provided that a generic firm would forfeit exclusivity if it failed to marketits generic product within a specified time frame.101 To simplify an overlycomplicated process, the generic firm would forfeit exclusivity if it failedto market its generic drug by the later of two dates: (1) seventy-five daysafter the date on which the ANDA is approved or thirty months after theANDA is submitted to the FDA, whichever is earlier; and (2) essentiallyseventy-five days after the patent in question is adjudicated or agreed tobe invalid or not infringed.102
Most observers concluded, however, that the forfeiture provision
failed to fully prevent the statutory bottleneck.103 The FDA, for example,still believes the bottleneck exists in situations where the brand namefirm settles with the first generic applicant, preventing the 180-day exclu-sivity period from beginning.104 A subsequent generic applicant may be
95. See 21 U.S.C. § 355(j)(5)(D) (describing forfeiture provision). 96. 21 U.S.C. § 355(j)(5)(B)(iv). 97. Id. 98. Herman, supra note 43, at 23.
99. See Hemphill, Paying for Delay, supra note 28, at 1586.
100. See id. at 1586–88 (discussing “approval” or “statutory bottleneck”). 101. See 21 U.S.C. § 355(j)(5)(D) (describing forfeiture provision). The MMA also
provided for five other ways that a generic firm could forfeit exclusivity, but thoseforfeiture events are not relevant for the purposes of this Note. Id. § 355(j)(5)(D)(i)(II)–(VI).
102. Id. § 355 (j)(5)(D)(i)(I)(aa)–(bb); see infra Part II.A (discussing failure to
market forfeiture event in greater detail).
103. See, e.g., Hemphill, Aggregate Approach, supra note 36, at 660–61 (“[T]he new
rules still contain a bottleneck.”).
104. The FDA’s Director of the Office of Generic Drugs wrote:Inherent in the structure of the “failure to market” forfeiture provisions is thepossibility that a first applicant would be able to enter into a settlement
able to bring about forfeiture in certain circumstances by obtaining a de-claratory judgment against the brand name firm that the applicable pat-ents are invalid or noninfringed.105 It has little incentive to do so, how-ever, because winning such a challenge would give the 180-day exclusivityto the first generic filer.106 Thus, there remains an entry-delaying statu-tory bottleneck without a sufficient remedy.
The forfeiture provisions coupled with the drug-product-based exclu-
sivity rule have created a unique strategic dilemma for generic firmswhere there are multiple patents of different strengths on a single drug. This is a recurring problem because, as discussed above, increasing num-bers of drugs now have multiple patents on them.107 Generic firms wantto secure exclusivity because it is worth millions of dollars for majordrugs.108 However, it may not be feasible to challenge all of the patentson a single drug because some of the patents are too strong. In thesecircumstances generics have started to race to be the first ANDA filer andto file Paragraph IV certifications for the weak patents and Paragraph IIIcertifications for the strong patents.109 In the first such case, the brand
agreement with the [brand name firm] in which a court does not enter a finaljudgment of invalidity or non-infringement (i.e., without a forfeiture event undersubpart (bb) occurring), and that subsequent applicants would be unable toinitiate a forfeiture with a declaratory judgment action. This inability to force aforfeiture of 180-day exclusivity could result in delays in the approval of otherwiseapprovable ANDAs owned by applicants that would market their generic drugs ifthey could but obtain approval. This potential scenario is not one for which thestatute currently provides a remedy.
Letter from Gary J. Buehler, Dir., Office of Generic Drugs, Food & Drug Admin., to MarcA. Goshko, Exec. Dir., Teva N. Am. 5 n.6 (Jan. 17, 2008) [hereinafter FDA GranisetronLetter], available at http://www.fda.gov/ohrms/DOCKETS/dockets/07n0389/07n-0389-let0003.pdf (on file with the Columbia Law Review).
105. See Teva Pharm. USA, Inc. v. Eisai Co., 620 F.3d 1341, 1347 (Fed. Cir. 2010)
(allowing generic firm Teva to bring declaratory judgment action in these circumstances).
106. Cf. Hemphill, Paying for Delay, supra note 28, at 1586 (noting settling with first
generic filer “removes from consideration the most motivated challenger, and the oneclosest to introducing competition”).
107. See supra text accompanying notes 72–73 (explaining brand name firms’
practice of filing increasing number of patents for each drug through process known as“evergreening”).
108. Hemphill, Paying for Delay, supra note 28, at 1560.
109. The practice is becoming a “concern” for the FDA. In a recent letter it wrote:[A] point that has been a source of some concern to the [FDA] irrespective of theforfeiture provisions . . . is that the “race” to earn 180-day exclusivity by submittingthe first ANDA to challenge a patent may result in the submission of ANDAs thatmay also contain one or more paragraph III certifications to patents that do notexpire until well into the future and that will preclude approval of the applicationfor many years. We have received ANDAs for which, based on the patentexpiration dates and corresponding certifications, the sponsor has no intentionto obtain approval and market the generic drug for 12 years or more.
FDA Granisetron Letter, supra note 104, at 6 n.7 (emphasis added).
name firm declined to sue the generic on its Paragraph IV certifications,but in three subsequent cases, the brands did initiate litigation. Genericshave responded by seeking to stay the patent litigation. This Part exam-ines the generics’ strategy and the brands’ recent decision to initiate pat-ent infringement litigation, discusses brands’ mysterious legal positionopposing stays in two out of the three cases, and offers possible explana-tions for brands’ behavior.
The FDA first encountered the generic strategy of filing Paragraph
III and Paragraph IV certifications for different patents on a single drugin 2004—for a drug called Granisetron, which is used to treat nausea andvomiting following chemotherapy.110 Hoffmann-La Roche Inc. (Roche),the brand name firm, listed three patents in the Orange Book on thisformulation of Granisetron111 and marketed it under the name Kytril.112A generic firm, Teva Parenteral Medicines (Teva), filed its ANDA in June2004, containing a Paragraph III certification for the strong patent set toexpire in December 2007, a so-called “section viii carve out”113 for onepatent set to expire in September 2016, and a Paragraph IV certificationfor the weak patent set to expire in May 2019.114 Significantly, the corepatent on the drug was set to expire in 2007 but, through evergreening,Roche had attempted to extend its monopoly by twelve more years.
110. See Granisetron, PubMed Health, http://www.ncbi.nlm.nih.gov/pubmed
health/PMH0000159 (on file with the Columbia Law Review) (last updated Sept. 1, 2008).
111. Upadhye, supra note 94, § 14:10.
112. FDA Granisetron Letter, supra note 104, at 1. 113. A generic applicant may elect not to certify that its generic product should be
used for each and every use of the brand name drug. See 21 U.S.C. § 355 (j)(2)(a)(viii)(2006). If so, it may file a “section viii” statement which states that the “the applicant is notseeking approval to market the product for one or more patented methods of use coveringthe listed drug.” FDA Confirms Its View on Generic Drug Label Carve-Outs, McDermott,Will & Emery (Aug. 21, 2008), http://www.mwe.com/index.cfm/fuseaction/publications.nldetail/object_id/158ffb98-bd2e-4aed-a7a3-da9be792e3ce.cfm (on file with the ColumbiaLaw Review). In this case, the patent at issue only pertained to a specific use of Granisetron,and this carve out allowed Teva to market its drug only to those uses not covered by thepatent.
114. Upadhye, supra note 94, § 14:10. The strong patent discloses the Granisetron
molecule itself and is therefore likely unchallengeable. The patent expiring in 2016 is amethod of use patent for Granisetron for the treatment of Post-Operative Nausea andVomiting, a condition that results in nausea after the patient undergoes generalanesthesia. The patent expiring in 2019 is a dosage patent, disclosing multiple dosages forcertain patients. For detailed information on these patents, see Aseptically Filled MultidoseInjectable Dosage Forms of Granisetron, Int’l Patent Pub. No. WO 2007/069070 A2, WorldIntellectual Property Organization (June 21, 2007), available at http://www.wipo.int/patentscope/search/en/detail.jsf?docId=WO2007069070&recNum=1&maxRec=1&office=&prevFilter=&sortOption=&queryString=FP%3A%28Aseptically+Filled+Multidose+Injectable+Dosage+Forms+of+Granisetron%29&tab=PCTDescription (on file with the ColumbiaLaw Review).
Roche opted not to sue Teva on its Paragraph IV certification.115 Teva’sANDA was therefore “tentatively” approved—the FDA could not awardTeva final approval because the outstanding Paragraph III certificationprevented generic entry until that patent expired in December 2007.116That date was forty-three months after Teva filed its ANDA, and so itcould have been deemed to trigger the thirty-months-to-market forfeitureevent in the MMA.117
Contrary to some expectations,118 however, the FDA held that Teva
was entitled to the 180-day exclusivity period and that it was not for-feited.119 The failure to market forfeiture event occurs at the “later” oftwo specific dates, as defined by subsections (aa) and (bb) of the failureto market provision in the MMA. The (aa) date occurs seventy-five daysafter the date on which the ANDA is approved or thirty months followingthe submission of the ANDA, whichever is earlier.120 The (bb) date essen-tially occurs seventy-five days after the patent is adjudicated or agreed tobe invalid or not infringed.121 In this case, in a novel interpretation of theforfeiture doctrine, the FDA determined that since only one of those twodates had occurred—the (aa) date but not the (bb) date—it could notdetermine which date was “later” than the other, and therefore a forfei-ture did not occur.122
115. Upadhye, supra note 94, § 14:10.
116. Letter from Marc Goshko, Exec. Dir., Teva N. Am. to Gary J. Buehler, Dir., Office
of Generic Drugs, Food & Drug Admin. 2–3 (Sept. 28, 2007) [hereinafter TevaGranisetron Letter], available at http://www.regulations.gov/#!documentDetail;D=FDA-2007-N-0269-0004 (on file with the Columbia Law Review).
117. See 21 U.S.C. § 355 (j)(5)(D)(i)(I)(aa)(BB) (providing that forfeiture could
occur if generic does not bring its drug to market by “30 months after the date ofsubmission of” its ANDA).
118. See, e.g., Upadhye, supra note 94, § 14:10 n.1 (noting author incorrectly
119. See FDA Granisetron Letter, supra note 104, at 1 (“[I]t is the Agency’s position
that Teva is entitled to 180-day exclusivity with respect to [its Granisetron] ANDA.”).
120. 21 U.S.C. § 355 (j)(5)(D)(i)(I)(aa). 121. Specifically, the (bb) date occurs seventy-five days after at least one of the
following events occurs: A court has issued a final judgment holding the patent eitherinvalid or not infringed, a settlement between the parties that includes a finding that thepatent is invalid or not infringed, or the patent holder delists the patent from the Orangebook. Id. § 355(j)(5)(D)(i)(I)(bb).
122. FDA Granisetron Letter, supra note 104, at 5. For its forfeiture analysis, first the
FDA determined the (aa) date, which requires comparing the dates computed undersubsections (aa)(AA) and (aa)(BB). In order to determine the date under subsection(aa)(AA), the FDA determined that Teva received final launch approval on December 31,2007 (at the expiration of the strong patent), which meant that seventy-five days later wasMarch 15, 2008. Id. at 4. It then determined that the date under subsection (aa)(BB) isthirty months after the submission of Teva’s ANDA, which would be December 1, 2006. Id. Since that date was before March 15, 2008, it would be the applicable date under thesubpart (aa) analysis. Id. The MMA directs the FDA to look to the “later of” the December1, 2006 (aa) date and the date computed under subpart (bb). Id.; see 21 U.S.C. § 355(j)(5)(D)(i)(I) (describing mechanism for computing this calculation). The trickier partwas figuring out what the (bb) date should be in order to make an appropriate forfeiture
This determination constituted a huge win for Teva and other poten-
tial generic manufacturers.123 It meant that a generic that was the first tofile an ANDA containing both Paragraph III and IV certifications couldretain exclusivity when the strong patent expired. That would occur evenif the Paragraph III patent expired much later than the (aa) event so longas no (bb) event took place. Essentially, so long as the brand failed to suethe generic or delist its patent, no (bb) event would take place and exclu-sivity would be assured.124 The result is that consumers have to paysupracompetitive duopoly prices for 180 days after the strong patent ex-pires. This leads to not only a decline in consumer welfare, but also anallocative inefficiency due to the duopoly price. Moreover, the result issocially wasteful because the generic adds nothing to society or the con-sumer: It, like everyone else, could not start marketing a genericGranisetron until the strong patent expired. The generics likely believedthat this strategy was a good one. However, things did not go as the gener-ics planned.
B. Cases Addressing the Stay Dilemma
In three subsequent cases, brand name firms have sued generics who
have adopted this strategy, based on their Paragraph IV certifications. Inall three cases, the generic defendant has sought a stay of the litigation. As explained more fully below, these stays are obviously in the interest ofthe generic defendant who does not want to forfeit its exclusivity. Underthe FDA’s Granisetron analysis, a generic victory in the litigation wouldconstitute a (bb) event. Since the strong patents prevent generic entry formany years, the generic firm would forfeit exclusivity.
One would expect brand name firms not to oppose these stay mo-
tions because even if they were to lose the litigation, they would preferthat the generic firms retain exclusivity since the brands would earn sub-stantially more money in 180 days of duopoly than in full competition.
comparison. Here, the FDA determined that no (bb) event had occurred: There was nolitigation (under (AA) or (BB)), and the patentee did not delist its patents (under (CC)). FDA Granisetron Letter, supra note 104, at 5. The FDA determined that since there was no“later” event with which to compare the (aa) event, forfeiture could not occur, especiallysince it was at least theoretically possible that a (bb) event could take place in the future. Id. Although such an action might be unlikely, a subsequent generic filer could still initiatea declaratory judgment action that the challenged patent is invalid, which would trigger a(bb) event. Id; see also Minn. Mining & Mfg. Co. v. Barr Labs., Inc., 289 F.3d 775, 779–81(Fed. Cir. 2002) (holding 180-day exclusivity period can be “triggered” by termination ofaction by second ANDA filer).
123. While it was a huge win for generics, it was a mixed bag for consumers. On the
one hand, approval of Teva’s ANDA ensured the entry of a lower-priced Granisetron inDecember 2007, more than a decade before the final patent expired, and the exclusivitypotential might have been what motivated Teva to file its ANDA. On the other hand,consumers would have benefited more if Teva had not gained exclusivity, since it wouldhave resulted in greater competition starting in December 2007.
124. It is unlikely that another generic firm would initiate a declaratory judgment
action to trigger a forfeiture under (bb)(BB) since there would be no benefit to it.
Accordingly, in one case, the drug companies stipulated to a four-yearstay, which was approved by the court.125 However, in two other cases, thebrand name firms have opposed these stay requests. Facing virtually iden-tical stay motions in these two cases, one court granted a stay and onecourt denied a stay. Neither those two courts nor the court that approvedthe stipulated stay, however, addressed the competitive concerns sur-rounding a stay motion under these circumstances, or the effect of a stayon consumer welfare.
1. Abbott v. Matrix. — Plaintiff Abbott Laboratories (Abbott) mar-
kets Kaletra, a drug cocktail containing lopinavir and ritonavir that isused to control the human immunodeficiency virus (HIV).126 Abbottlisted eleven patents in the Orange Book on this drug. These patents canbe divided into two groups. The first group includes the first nine pat-ents,127 and it likely makes up the core compound patents for the drug. The latest expiring patent in that group is set to expire on December 26,2016.128 The second group of patents contains a patent to a polymorph—i.e., a different crystalline structure of the drug—and one to a water-soluble polymer of the drug.129 Those patents are set to expire January19, 2020 and November 10, 2021, respectively.130 On December 23, 2008,defendant Matrix Laboratories, Inc. (Matrix) filed an ANDA withParagraph III certifications for all of the patents in the first group andParagraph IV certifications to both patents in the second group.131 As aresult of the Paragraph III certifications, Matrix cannot market its genericlopinavir/ritonavir tablets until the first group of patents expires in 2016. Nevertheless, Abbott decided to sue Matrix over its paragraph-IV certifi-cations. Prior to discovery, Matrix filed a motion to stay the litigation forfive years, from 2009 to July 2014.132 It proposed that during this stay thethirty-month statutory stay period would be tolled and that either partycould move to vacate the stay for “good cause.”133 In its motion, Matrixwas upfront about its reasons for a stay:
125. Brovana Stay Order, supra note 14.
126. Abbott is no stranger to litigation over this drug. See, e.g., Doe v. Abbott Labs.,
571 F.3d 930, 932, 935 (9th Cir. 2009) (alleging that Abbott is attempting to obtain amonopoly in all protease inhibitors by charging too much for ritonavir and too little forKaletra, creating a form of “price squeeze”); Schor v. Abbott Labs., 457 F.3d 608, 610 (7thCir. 2006) (same).
127. Upadhye, supra note 94, § 11:14.
128. Id. 129. Id. 130. Electronic Orange Book Patent and Exclusivity Information, Lopinavir/
Ritonavir 200mg/50mg, http://www.accessdata.fda.gov/scripts/cder/ob/docs/patexclnew.cfm?Appl_No=021906&Product_No=001&table1=OB_Rx (on file with the ColumbiaLaw Review) (last visited Sept. 25, 2011).
131. Memorandum in Support of Defendants’ Motion to Stay at 4, Abbott Labs. v.
Matrix Labs., Inc., No. 09-cv-1586, 2009 WL 3719214 (N.D. Ill. Nov. 5, 2009) [hereinafterMatrix Stay Motion].
In the present action, absent a stay, Defendants are in a Catch-22 situation in view of the forfeiture provision. If Defendantslitigate now and win, their exclusivity will almost inevitably beforfeited because a final court decision would almost certainlyoccur more than 75 days prior to December 26, 2016 [when thefirst group of patents expires]. In that event, Defendants wouldhave undertaken the expense of a successful challenge to thepatent obstacle to generic entry, but would then compete withgeneric competitors who . . . would not have borne the competi-tive burden of the litigation expense. . . . If Defendants lose,then they would presumably be enjoined . . . and unable to com-pete with later generic applicants until expiration of the pat-ent(s) they would be found to infringe. In sum, if Defendantswin soon, they lose in the marketplace, and if Defendants losesoon they lose in the marketplace. The only way to avoid theCatch-22 is to stay the action.134Abbott recognized that “staying this litigation may provide some ben-
efit to both parties.”135 In fact, it maintained that it “has no objection, intheory, to a five year hiatus from litigating the two patents in suit.”136Nonetheless, Abbott did not consent to Matrix’s stay motion. It arguedthat the stay would be contrary to the Hatch-Waxman Act’s purposes ofexpediting patent litigation and that the “stay is not sufficiently tailoredto protect against potential prejudice and harm to Abbott.”137 Thisclaimed prejudice included uncertainty about what would constitute“good cause” to lift the stay and potential evidentiary problems, sincemany of defendant’s employees reside in India.138 Abbott also placedMatrix’s “Catch-22” situation squarely on Matrix because Matrix “know-ingly controlled the timing of this lawsuit when they submitted and noti-fied Abbott of their Paragraph IV Certifications.”139
Significantly, Abbott did not claim that it was opposing the stay mo-
tion, only that it “cannot consent” to the stay.140 As a result, it did notspecifically request that the court deny the motion.141
The district court granted the stay motion based on the court’s “in-
herent power to manage [its] docket[ ], including the authority to stayproceedings.”142 It assumed that the traditional factors governing staymotions applied: “(i) whether a stay will unduly prejudice or tactically
134. Id. at 9–10. 135. Plaintiff’s Response to Defendants’ Motion to Stay at 1, Abbott, No. 09-cv-1586,
2009 WL 3719214 [hereinafter Abbott Stay Response].
136. Id. at 4. 137. Id. at 1. 138. Id. at 7–8. 139. Id. at 4. 140. Id. at 8. 141. The “Conclusion” section of Abbott’s motion stated only that it “cannot consent”
to the stay and that “oral argument” was requested. Id. at 8–9. It did not explicitly ask thatthe stay motion be denied.
142. Abbott, 2009 WL 3719214, at *2.
disadvantage the non-moving party, (ii) whether a stay will simplify theissues in question and streamline the trial, and (iii) whether a stay willreduce the burden of litigation on the parties and on the court.”143 Thecourt held “there is no evidence that delaying the litigation will undulyprejudice or tactically disadvantage Abbott” since the stay would also tollthe thirty-month statutory stay period.144 The court brushed asideAbbott’s argument that a stay would be contrary to the Hatch-WaxmanAct’s duty to “reasonably cooperate in expediting” patent litigation.145 Itnoted that there is a statutory remedy—tolling the thirty-month period—which is exactly what the stay would do.146 Moreover, the court creditedMatrix’s assurance that it would not market its generic version of Kaletrauntil 2016, further showing that Abbott would not suffer any prejudice.147Finally, the court found that Abbott’s general concern about evidentiaryproblems was insufficiently specific and in any event Abbott would be freeto seek a modification of the stay motion should it come up with morespecific evidentiary worries.148
As for the second and third factors, the court first held that there
would be a negligible simplification of the issues, but it decided not togive that factor much weight.149 It attached greater weight to the relativeburdens on the parties.150 Finding that no party would be burdened, thecourt granted the stay.151
2. Millennium v. Teva. — A very similar situation occurred soon af-
ter Abbott, but this time the district court denied the generic defendant’sstay motion. Millennium Pharmaceuticals, Inc. and Schering Corporation(Millennium) market the drug Integrilin, which is an injection used totreat individuals suffering heart attacks.152 Millennium listed five patentson the drug in the Orange Book. The first group of patents—which arelikely the core patents—consists of two patents set to expire in 2014.153 A
143. Id. 144. Id. 145. 21 U.S.C. § 355(j)(5)(B)(iii) (2006). 146. Abbott, 2009 WL 3719214, at *3. 147. Id. 148. Id. at *4. 149. Id. 150. Id. Interestingly, the court did not find that defendant’s “Catch-22” situation was
that relevant because it found that such a situation was implicitly authorized by theregulatory regime set up by Hatch-Waxman. See id. at *4 n.1 (noting it is not “self-evidentthat such an outcome would frustrate Congress’s intent”).
151. Id. at *4–*5. 152. Millennium Pharm., Inc. v. Teva Parenteral Meds., Inc., Nos. 09-cv-105, 09-cv-204,
10-cv-137, 2010 WL 1507655, at *1 (D. Del. Apr. 14, 2010).
153. Electronic Orange Book Patent and Exclusivity Information, Eptifibatide
2mg/mL, http://www.accessdata.fda.gov/scripts/cder/ob/docs/patexclnew.cfm?Appl_No=020718&Product_No=001&table1=OB_Rx (on file with the Columbia Law Review) (lastvisited Aug. 12, 2011).
second group—consisting of three patents—is set to expire in 2015.154Teva, the same generic manufacturer that was involved in theGranisetron matter, filed an ANDA with Paragraph III certifications forthe 2014 patents and Paragraph IV certifications for the 2015 patents.155Soon after, Millennium sued Teva over the Paragraph IV certifications.
Teva responded by filing a motion to stay the litigation for around
two years, from March 2010 until May 11, 2012, and its motion reliedheavily on Abbott.156 It repeated many of the arguments made by Matrixin that case, including the argument that forfeiting its 180-day exclusivityperiod should be considered ample prejudice supporting a stay.157 It alsoargued that Millennium would not be prejudiced since even if Teva wonthe case, Teva could not market a generic version of Integrilin for manyyears.158 It also requested a stay because it would conserve judicial re-sources and because discovery had not yet begun.159
In this case, Millennium opposed the stay far more vociferously than
Abbott did.160 Millennium characterized Teva’s position as “asking to berescued from a dilemma entirely of their own making” since Teva con-trolled “the timing and scope of these actions.”161 Millennium arguedthat a stay was not authorized by the Hatch-Waxman Act because districtcourts do not have the ability to delay the administrative thirty-monthstay.162 It further argued that a delay would cause it significant prejudicefor two reasons: (1) It would prevent prompt resolution of the case toward off not only Teva but also other potential generic challengers,163and (2) the stay may lead to evidentiary problems such as “faded memo-
154. Id. The 2015 patents are polymers and use patents for the core molecule; some
clearly have nothing to do with the intended use of Integrelin to treat patients who arehaving heart attacks. In fact, two of the three 2015 patents concern snake venom. See U.S. Patent No. 5,807,825 (filed June 7, 1995) (“An assay for screening snake venom for thepresence or absence of platelet aggregation inhibitors (PAIs) based on specific receptorbinding is described.”); U.S. Patent No. 5,968,902 (filed June 7, 1995) (“The isolated andpurified PAI from several active snake venoms is described and characterized.”).
155. Only a public version of the briefing in this case is available, and the parties have
redacted which patents Teva is challenging with Paragraph IV certifications as well as forwhich patents it has filed Paragraph III certifications. However, that information isavailable online. See Aaron F. Barkoff, District Court Declines to Stay Integrilin Case,Putting Teva at Risk of Forfeiting 180-Day Exclusivity, Orange Book Blog (Apr. 19, 2010),http://www.orangebookblog.com/2010/04/district-court-declines-to-stay-integrelin-case-putting-teva-at-risk-of-forfeiting-180-day-exclusivity.html (on file with the Columbia LawReview).
156. Teva’s Motion to Stay at 1, 7–9, Millennium, Nos. 09-cv-105, 09-cv-204, 10-cv-137,
157. Id. at 6. 158. Id. at 7. 159. Id. at 7–8. 160. Plaintiffs’ Opposition to Teva’s Motion to Stay, Millennium, Nos. 09-cv-105, 09-cv-
204, 10-cv-137, 2010 WL 1507655 [hereinafter Millennium Stay Response].
161. Id. at 1. 162. Id. 163. Id. at 10.
ries and the potential loss of useful testimonial evidence.”164 Finally,Millennium argued that the proposed stay would give Teva an unfair tac-tical advantage, in that Teva would have no incentive to settle while thestay was pending.165
Two days after the parties finished submitting their briefs, the district
court denied the stay motion.166 It held that Millennium would beprejudiced by “not being able to timely ‘clear the cloud’ that has beencast over the validity of their  patents.”167 Further, the court heldthat a multi-year stay would prejudice Millennium by making it harder forit to prove the validity of its patents, believing that evidentiary problemswould exist.168 Finally, the court held that Millennium would beprejudiced because a stay would contravene the “statutorily provided paththat was meant to expedite the consideration of patent validity.”169 Thecourt did not give much weight to Teva’s arguments that absent a stay itwould be prejudiced by the forfeiture of its 180-day exclusivity period. Itreasoned that Teva “ha[s] only themselves to blame” for that result be-cause Teva “control[led] the timing of the present litigation.”170 Thecourt characterized Teva’s actions as follows:
Rather than wait until they could fully take advantage of theirposition as first filer, however, Defendants sought to prema-turely reserve their place at the front of the line, and now seekan order from this Court that allows them to preserve that posi-tion. . . . Although Defendants may suffer a hardship, it is one oftheir own creation . . . .171Finally, the court accepted Millennium’s contention that the stay
would remove the incentive for Teva to settle.172 Accordingly, the courtdenied the stay motion.
3. Sepracor v. Teva. — While the brand name firm opposed the ge-
neric firm’s stay motion in the two prior cases, in a third case, the compa-nies agreed on a four-year stay.173 Sepracor, the brand name firm, mar-kets Brovana, an inhalation used to control the symptoms of chronicpulmonary diseases, such as bronchitis and emphysema.174 Sepracorlisted eight patents in the Orange Book on Brovana.175 In November
164. Id. at 10–11. 165. Id. at 11–12. 166. Teva filed its reply brief on April 12, 2010, and the district court denied the stay
167. Millennium, 2010 WL 1507655, at *3. 168. Id. 169. Id. 170. Id. 171. Id. 172. Id. at *4. 173. Brovana Stay Order, supra note 14, at 2.
174. Arformoterol Inhalation, PubMed Health, http://www.ncbi.nlm.nih.gov/
pubmedhealth/PMH0000410/ (on file with the Columbia Law Review) (last updated May 1,2010).
175. Brovana Stay Order, supra note 14, at 1–2.
2009, Teva filed an ANDA with Paragraph IV certifications for three pat-ents and Paragraph III certifications for five patents. The result would begeneric entry on November 12, 2016, when the last unchallenged patentexpires.176 Sepracor sued Teva on its Paragraph IV certifications. Lessthan six months after Pfizer filed its complaint and before Teva even filedits answer, the two companies agreed on a four-year stay, from April 22,2010 through February 1, 2014.177 As a result, if Teva wins the suit, it willnot forfeit exclusivity.
While the courts in Abbott and Millennium have split over whether to
grant the stay motions, neither court considered adequately the parties’incentives and likely benefits from granting or denying the stays. The ge-neric firms have an obvious interest in seeking a stay, since a denial fol-lowed by a litigation victory would cause them to lose their 180-day exclu-sivity period. But the brand name plaintiffs also have a strong incentive tofavor a stay. Thus, one would expect the brand name firms to agree to astay, as was the case in Sepracor. However, brand name firms have opposedthe stay motions in the two other cases, and their arguments in opposi-tion do not line up with their incentives. This section examines thoseincentives.
1. Brand Name Firm Incentives. — Despite their stated opposition to
generics’ stay motions in Abbott and Millennium, brand name firms wouldappear to benefit just as much as the generic firms from a stay. Absent astay, a generic firm’s victory would cause the generic firm to forfeit exclu-sivity. As a result, all generic drug companies could immediately enter themarket once the strong patents expire.178 Many generic firms would belikely to do so because the time between the litigation victory and theexpiration of the strong patents would give them ample time to preparetheir generic products.179 Thus, instead of a 180-day duopoly after patentexpiration, the brand name firm would be looking at a completely com-petitive landscape, which would force prices down considerably more
176. Id. 177. Id. at 2. Judge William H. Pauley III approved the stay on July 8, 2010, but the
stay was made effective as of April 22, 2010. Id. at 2–3.
178. See 21 U.S.C. § 355 (j)(5)(D)(iii)(I)–(II) (2006) (“If all first applicants forfeit
the 180-day exclusivity period . . . approval of any application containing a [Paragraph IVcertification] shall be made effective . . . and no applicant shall be eligible for a 180-dayexclusivity period.”).
179. See FTC, Pay-for-Delay Study, supra note 25, at 8 (“Publicly available information
about recent generic launches suggests that a generic market typically matures about oneyear after the first entrant comes on the market.”). In all cases where the stay dilemma ispresent, a generic victory absent a stay would almost certainly occur more than one yearbefore the strong patents expire. Moreover, the FTC study assumed that the first genericfirm would have exclusivity, meaning that its one year period included six months ofexclusivity. Thus, without exclusivity, a generic market would likely take considerably lessthan one year to mature.
than in a duopoly situation.180 Moreover, duopoly—or at least less thanfull competition—usually lasts for many months beyond the expiration ofthe 180-day exclusivity period. In fact, it usually takes another 180 days fora complete generic market to mature because the first generic filer wouldhave a 180-day marketing and distribution head start.181 The head startmight even prevent some generic firms from entering at all.182 Thus, ageneric litigation victory followed by a forfeiture of exclusivity would costthe brand name manufacturer at least 180 days, and most likely up to ayear, of reduced competition for its drug. That reduced competitioncould amount to hundreds of millions of dollars for the brand namefirm.183
These incentives are not merely theoretical: In similar contexts,
brand name firms have adopted strategies meant to ensure that the ge-neric firm does not forfeit its 180-day exclusivity period should it win theresulting litigation. In addition to the stipulated stay in Sepracor,184
Pfizer’s strategy concerning the drug Viagra provides another apt exam-
ple. There are two relevant patents on Viagra: one patent set to expire onMarch 27, 2012, which claims the invention of sildenafil, the active ingre-dient in Viagra;185 and another patent set to expire on October 22, 2019,which essentially covers using Viagra to treat erectile dysfunction (ED).186On December 17, 2004, the generic manufacturer Teva—the same ge-neric manufacturer in Sepracor, Millennium, and the Granisetron matter—filed an ANDA with a Paragraph IV certification for only the later-expir-ing patent.187 Initially Pfizer did not sue and the FDA “tentatively” ap-proved Teva’s ANDA,188 meaning that Teva would be able to launch ageneric version of Viagra after the strong patent expires in March 2012. However, in March 2010—close to six years after Teva notified Pfizer of
180. See supra note 13 (detailing drug prices in duopoly and in full competition).
181. See FTC, Pay-for-Delay Study, supra note 25, at 8 (finding generic market
“matures” in about one year, or six months after expiration of 180-day exclusivity period).
182. Cf. Martin A. Voet, The Generic Challenge: Understanding Patents, FDA and
Pharmaceutical Life-Cycle Management 61 (2005) (noting generic firm that receivesexclusivity usually maintains majority of sales even after its exclusivity period expires).
183. See supra note 56 and accompanying text (explaining that exclusivity period is
potentially worth hundreds of millions of dollars).
184. See supra Part II.B.3 (discussing Sepracor). 185. Pfizer Inc. v. Teva Pharm. USA, Inc., No. 2:10-cv-00128, 2011 WL 3563112, at *1
186. See id. at *1 (describing the later-expiring patent as “the oral administration of
sildenafil and other compounds for the treatment of ED”). Sildenafil was originallydeveloped to treat high blood pressure and Pfizer only later discovered that it could beused to treat ED. Pfizer then filed for a patent covering that use. See id. at *3–*4 (tracinghistory of development of sildenafil). Teva claimed that this use patent was anticipated byprior art and was therefore invalid. See id. at *20–*21 (reviewing this claim).
187. See Viagra Complaint, supra note 15, at 4 (noting that Teva’s ANDA challenged
188. Id.; see also Pfizer, 2011 WL 3563112, at *1.
its ANDA—Pfizer sued Teva on its Paragraph IV certification.189 Sincepatent litigation typically takes around two years to complete,190 if Tevawins, the victory would occur around the time that the 2012 patent wouldexpire. As a result, Teva would be able to launch its generic Viagra withexclusivity. If Pfizer had sued when it was first notified of Teva’s ANDAback in 2004, a Teva victory would have caused it to forfeit exclusivity. Accordingly, the six-year delay appeared to be timed entirely to ensurethat if Teva wins the lawsuit, it could enter with exclusivity.191
In other situations, once generic entry becomes inevitable, brand
name firms have even declined to sue generics on challenged patents—effectively sacrificing one of their patents so that a generic could retainexclusivity. This strategy is illustrated in the case surrounding the populardrug Lipitor. In that case—which was governed by the pre-MMA rulesand thus lacked a forfeiture possibility—the generic firm, Ranbaxy, wonits challenge to one patent associated with the drug in the FederalCircuit.192 This triggered exclusivity, but prematurely, since the othervalid and infringed patents prevented FDA approval of Ranbaxy’s ANDA. The result would have been nonexclusive generic entry in March 2010.193Ranbaxy likely was the first ANDA filer for the other patents but Pfizerdid not sue.194 As Scott Hemphill explains, “[b]y declining to sueRanbaxy on these patents, Pfizer preserved Ranbaxy’s exclusivity despitethe initial trigger, a preferable result for both parties.”195
A similar result occurred for the popular drug Zoloft, where the
brand name and generic firms preserved the generic’s exclusivity periodby explicit agreement. Pfizer had two patents on Zoloft, a strong patent
189. See Viagra Complaint, supra note 15, at 4, 9 (admitting Pfizer knew of Teva’s
ANDA on December 17, 2004, but did not sue until March 24, 2010).
190. See FTC, Generic Drug Study, supra note 29, at 47 tbl.4-1 (finding average time
between complaint and district court decision in Hatch-Waxman litigation for first genericapplicant is “25 months, 21 days”).
191. However, this strategy came at a cost: Since Pfizer did not sue within forty-five
days of receiving notification of Teva’s ANDA, it was not entitled to the thirty-monthadministrative stay. See 21 U.S.C. § 355(j)(5)(B)(iii) (2006) (awarding thirty-month stayonly if brand name firm sues within forty-five days of receiving notification of genericfirm’s ANDA).
192. See Pfizer, Inc. v. Ranbaxy Labs. Ltd., 457 F.3d 1284, 1291–92 (Fed. Cir. 2006)
(invalidating one patent because of drafting error in patent application).
193. Hemphill, Aggregate Approach, supra note 36, at 653 n.99.
194. As explained above, under pre-MMA law, each patent provided a new exclusivity
period. The MMA changed that “patent-by-patent” approach with a single opportunity forexclusivity for each product. 21 U.S.C. § 355(j)(5)(B)(iv)(I) (making exclusivity availableonly to “first applicant”); id. § 355(j)(5)(B)(iv)(II)(bb) (defining “first applicant” byreference to drug, not patent).
195. Hemphill, Aggregate Approach, supra note 36, at 653 n.99. Two years later,
Pfizer and Ranbaxy settled, allowing Ranbaxy to market its generic version of Lipitor inNovember 2011, with retained exclusivity. See Pfizer Inc. v. Apotex Inc., 726 F. Supp. 2d921, 926 (N.D. Ill. 2010) (“Pfizer and Ranbaxy entered a settlement agreement wherebyRanbaxy agreed not to market its product until November 30, 2011. Therefore, Ranbaxy’sexclusivity period will not begin to run until that date at the earliest . . . .”).
expiring in 2006 and a weak patent expiring in 2013.196 In 1999, Zenith, apotential generic competitor, challenged the weak patent but not thestrong one.197 If Zenith won its lawsuit as to the weak patent, it wouldalmost certainly have wasted its exclusivity because the strong patent pre-vented generic entry. Instead, Pfizer and Zenith entered into a settlementallowing Zenith to enter with exclusivity at the expiration of the strongpatent.198 No money changed hands.199 Presumably Pfizer felt that itcould not win the infringement action as to its weak patent, and thereforeensured that Zenith could enter with exclusivity.
While the brand name firm is incentivized to prefer that the generic
firm retain exclusivity should it lose the patent infringement litigation,there is also little incentive for the brand to oppose the stay, even if it isfairly sure it will win. That is because of the unique statutory scheme setup by the Hatch-Waxman Act, which ensures that a brand victory simplypreserves the status quo.200 While the lawsuit accuses the generic defen-dant of patent infringement, the generic has yet to market the competingdrug, so the brand name manufacturer has not suffered any damages. The absence of any damages to the brand name firm is why courts, in-cluding the court in Millennium, refer to the filing of a Paragraph IV-certified ANDA as an “‘artificial act of infringement.’”201 That is particu-larly true in situations where not only has the generic firm not produceda competing product, but it is in fact prevented from doing so for manyyears by the continued validity of the strong patents.
Thus, Abbott’s understated concession that “staying this litigation
may provide some benefit to both parties” is not surprising.202 In fact, inone case under slightly different circumstances, it was the plaintiff brandname firm that requested the stay, not the defendant generic firm.203 Inthat case, the generic firm, Exela, filed a Paragraph IV certification forthe patents on the brand name firm’s glaucoma drug Alphagan. Allergan,the brand name firm, filed a patent infringement suit against Exela andthen subsequently filed a motion to stay the litigation. Just like in Abbott
and Millennium, the generic firm, Exela, was the first generic filer and so
196. Hemphill, Aggregate Approach, supra note 36, at 653 n.99.
197. Id. 198. Id.; see also Stipulation of Filing of Redacted Settlement Agreement, Exhibit B.
at 5, 7, Pfizer, Inc. v. Zenith Goldline Pharm., Inc., Nos. 00-0408, 01-6007 (D.N.J. June 14,2002) [hereinafter Zoloft Agreement].
199. See Zoloft Agreement, supra note 198. The only form of compensation was
200. See supra Part I.C (explaining that brand name firms have little to gain from
litigation victory other than preserving status quo).
201. Millennium Pharm., Inc. v. Teva Parenteral Meds., Inc., Nos. 09-cv-105, 09-cv-204,
10-cv-137, 2010 WL 1507655, at *3 (D. Del. Apr. 14, 2010) (emphasis added) (quoting EliLilly & Co. v. Medtronic, Inc., 496 U.S. 661, 678 (1990)).
202. Abbott Stay Response, supra note 135, at 1.
203. See In re Brimonidine Patent Litig., No. 07-md-1866, 2008 WL 4809037, at *1 (D.
was eligible for the 180-day exclusivity period.204 Allergan claimed thatExela could not market its generic version of Alphagan because Exela’sjoint venture partner terminated its relationship with Exela.205 Allergancontended that a stay was necessary until Exela could find a joint venturepartner to ensure that the trial would address the correct generic prod-uct.206 The court denied the stay for the obvious reason that it was Exela’sANDA that was at issue, not how Exela went about manufacturing itsproduct.207
It seems that Allergan’s main motivation was to delay resolution of
the case as a means to delay the introduction of generic competition. That was certainly Exela’s belief—it argued that a stay “will only serve todelay market entry of a generic alternative to Allergan’s brand-name drugfor the sole benefit of Allergan, but to the detriment of both Exela andconsumers.”208 The evidence suggests that Exela was correct. Prior to thecase, Allergan entered into an agreement with another generic drug firm,Alcon, in which Alcon would sell a version of Alphagan, but not untilSeptember 30, 2009.209 In return, Allergan would drop its patent in-fringement suit against Alcon and would receive royalties from thosesales.210 Those royalties would be significantly lower if another genericcompetitor could enter before September 30, 2009, as Exela was threat-ening—especially since Exela would enter with exclusivity.211 Accord-ingly, it made economic sense for Allergan to delay the resolution of itslitigation with Exela until sometime after Alcon entered the market. Thus, the combination of brand name firms’ incentives favoring a stay
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . This essay responds to some of the recent publications in the philos-ophy of mathematics which either are or might be connected withthe semantics of set theory. It is concerned particularly with therelevance of forcing to questions in the semantics of set theor
COMMON SKIN PROBLEMS— Steven E. Prawer, MD, Clinical Professor of Dermatology, University of Minnesota Medical School, Minneapolis Acne vulgaris: grade 1— comedones ( ie, whiteheads, blackheads); most common form; best treated with comedolytic agents, eg, Retin-A (tretinoin), Differin (adapalene), and Tazorac (tazarotene); avoid these agents in adults because of drying effects; patie