Teva Pharmaceuticals, USA, Inc v. Kathleen Sebelius

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued December 7, 2009 Decided March 2, 2010
No. 09-5281
TEVA PHARMACEUTICALS USA, INC.,
APPELLANT
v.
KATHLEEN SEBELIUS, IN HER OFFICIAL CAPACITY AS
SECRETARY OF HEALTH AND HUMAN SERVICES, ET AL.,
APPELLEES
Consolidated with 09-5308
Appeals from the United States District Court
for the District of Columbia
(No. 1:09-cv-01111-RMC)
Michael D. Shumsky argued the cause for appellant. With
him on the briefs were Jay P. Lefkowitz and Gregory L.
Skidmore.
Carmen M. Shepard and Kate C. Beardsley were on the
briefs for cross-appellant Apotex, Inc. in No. 09-5308.
Drake Cutini, Attorney, U.S. Department of Justice,
argued the cause for appellees. With him on the brief were
2
Eugene M. Thirolf Jr., Director, David S. Cade, Acting
General Counsel, United States Food and Drug
Administration, Michael M. Landa, Acting Associate General
Counsel, and Eric M. Blumberg, Deputy Chief Counsel.
Carmen M. Shepard and Kate C. Beardsley were on the
brief for amicus curiae Apotex, Inc. in support of appellees.
Before: HENDERSON and GRIFFITH, Circuit Judges, and
WILLIAMS, Senior Circuit Judge.
Opinion for the Court filed by Senior Circuit Judge
WILLIAMS.
Dissenting opinion filed by Circuit Judge HENDERSON.
WILLIAMS, Senior Circuit Judge: This is the latest
installment in a long-running series of cases concerning an
incentive that Congress established for companies to bring
“generic” versions of branded drugs to market faster than they
otherwise might. Teva Pharmaceuticals USA, Inc., a
manufacturer of generics, has received tentative approval
from the U.S. Food and Drug Administration to sell losartan
potassium products—used primarily to treat hypertension.
The approval will become final once the “pediatric exclusivity
period”1 ends, following the expiration of the last remaining
patent on Merck’s pioneered versions of the same drugs, sold
under the names Cozaar and Hyzaar. When that date arrives
(April 6, 2010), Teva believes that it should be entitled to the
six-month period of marketing exclusivity that generic drug
makers earn, in some circumstances, for successfully taking
1
This is a six-month extension of the time during which all
generic competition against a branded drug is prohibited, see 21
U.S.C. § 355a; it is not a subject of dispute here.
3
the risks and bearing the costs of showing the invalidity or
inefficacy of a patent that a brand-name drug maker has said
blocks competing products. See Mova Pharmaceutical Corp.
v. Shalala, 140 F.3d 1060, 1063-65 (D.C. Cir. 1998)
(describing the incentive regime established by the Hatch-
Waxman Act of 1984); Ranbaxy Laboratories Ltd. v. Leavitt,
469 F.3d 120, 121-22 (D.C. Cir. 2006).
Thwarting its receipt of that entitlement, however, is an
FDA interpretation of the operative statutory regime (the
Food, Drug, and Cosmetic Act, as amended by various other
laws, codified in relevant part at 21 U.S.C. § 355) that will
allow not only Teva but all generic manufacturers to sell their
approved losartan potassium products right out of the gate. In
short, Teva says that, effective April 6, 2010, the agency’s
interpretation will deprive the company of the competitive
advantage Congress has said it should enjoy.
To ward off this danger, Teva filed suit in the federal
district court for the District of Columbia in June 2009,
seeking a declaration that the relevant FDA policy is unlawful
and an injunction compelling the agency to act in accordance
with Teva’s reading of the statute. Despite protestations by
the government that the matter was not ripe for review and
that Teva lacked standing, the district court reached the merits
of the claim—but ruled in the FDA’s favor. Teva
Pharmaceuticals U.S.A, Inc. v. Sebelius, 638 F.Supp.2d 42(D.D.C. 2009). Teva now appeals that decision. We agree
that the suit is justiciable, and hold that the FDA’s
interpretation is inconsistent with, and thus foreclosed by, the
statutory scheme.
4
* * *
In the process of obtaining FDA approval to sell a
pioneering new drug, an applicant lists publicly all of the
patents that, it believes, would be infringed by
“bioequivalent” versions of the product sold by other
companies. Ranbaxy, 469 F.3d at 121-22 (discussing 21
U.S.C. § 355(a)-(b)(1)). Prospective generic competitors need
not, however, take these lists as gospel. After a new drug hits
the market, they can effectively challenge the brand maker’s
pronouncement by filing a certification that a proposed
generic version of the brand drug would not run afoul of one
(or more) of the putatively blocking patents, either because
the patent is invalid or because the generic maker has found a
way to design around it. See id. at 122 (discussing 21 U.S.C.
§ 355(j)(2)(A)(vii)(IV)). The generic producer’s filing, called
a “paragraph IV certification” in our past cases, comes in the
course of the generic’s own application for FDA approval,
known as an Abbreviated New Drug Application, or ANDA.
See id. (discussing 21 U.S.C. § 355(j)(2)).
Filing a paragraph IV certification comes with a risk,
though: it constitutes an act of patent infringement, 35 U.S.C.
§ 271(e)(2)(A), with the hazard of sparking costly litigation.
In order, then, to “compensate [generic] manufacturers for
research and development costs as well as the risk of litigation
from patent holders,” Teva Pharmaceuticals USA, Inc. v.
Leavitt, 548 F.3d 103, 104 (D.C. Cir. 2008), the statute
provides that the first company to file an ANDA containing a
paragraph IV certification earns an “exclusivity” period of
180 days, during which the FDA may not approve for sale any
competing generic version of the drug at issue, id. (discussing
21 U.S.C. § 355(j)(5)(B)(iv)). This promise of initial
marketing exclusivity is thus intended to increase competition
by expediting the availability of generic equivalents. See id.;
5
Serono Laboratories, Inc. v. Shalala, 158 F.3d 1313, 1326
(D.C. Cir. 1998).
A potential bug in the system is the ability of the brand
manufacturer, after a generic has filed a paragraph IV
certification, to announce that in fact the challenged patent is
not one that protects the drug at issue and to ask the FDA to
“delist” the patent, thus purporting to pull the rug from under
the paragraph IV certification. In Ranbaxy we considered
“whether the FDA may delist a patent upon the request of the
[brand manufacturer] after a generic manufacturer has filed an
ANDA containing a paragraph IV certification so that the
effect of delisting is to deprive the applicant of a period of
marketing exclusivity.” 469 F.3d at 125. The answer, we
said, was no; an FDA policy that allowed brand manufacturers
to strategically delist challenged patents, thereby unilaterally
stripping generic manufacturers of marketing exclusivity, was
“inconsistent with the structure of the statute.” Id.Ranbaxy, however, interpreted the law as it stood before
Congress amended it in 2003 via the Medicare Prescription
Drug, Improvement, and Modernization Act, Pub. L. No. 108-
173, 117 Stat. 2066. Id. at 122 n.*. Three times since the
effective date of the amendments, the same series of events at
issue in Ranbaxy has arisen—once involving the generic
manufacturer Cobalt Pharmaceuticals and the brand drug
Precose, made by Bayer; once involving the generic
manufacturer Hi-Tech Pharmacal Co. and the brand drug
COSOPT, made by Merck; and now involving Teva, the drugs
Cozaar and Hyzaar, and Merck. In the first two instances, the
generic makers presented arguments to the FDA why they
should still, in the modified statutory regime, be entitled to
exclusivity notwithstanding the brand companies’ delisting a
challenged patent. Teva itself responded to the FDA’s
solicitation of comments in the Cobalt matter, advocating the
same pro-exclusivity reading of the amended statute’s
6
treatment of post-paragraph-IV-filing delisting requests. See
Letter from Marc Goshko, Executive Director, Teva North
America, In Response to FDA Request for Comments re
Generic Drug Applications for Acarbose Tablets (Oct. 16,
2007), in Joint Appendix (“J.A.”) 78 et seq. In both cases, the
FDA ruled that the 2003 amendments required a different
outcome from the one Ranbaxy ordered under the old version
of the law.
The agency pointed to the 2003 amendments’ addition of
a critical new term to the statute: the “forfeiture event.” See
21 U.S.C. § 355(j)(5)(D)(ii). On the occurrence of any one of
six defined scenarios, the law now says, the entitlement to a
180-day exclusivity period “shall be forfeited by a first
applicant.” See id. In both the Cobalt and Hi-Tech disputes,
the FDA decided that the facts at issue, paralleling those in
Ranbaxy and our case, had satisfied the terms of the first listed
forfeiture event, “failure to market,” and in each case denied
the generic manufacturer exclusivity.
The statutory definition of the first listed forfeiture event
is as follows:
(I) FAILURE TO MARKET. — The first applicant fails
to market the drug by the later of —
(aa) the earlier of the date that is —
(AA) 75 days after the date on which the
approval of the application of the first applicant
is made effective under subparagraph (B)(iii); or
(BB) 30 months after the date of submission of
the application of the first applicant; or
(bb) with respect to the first applicant or any other
applicant (which other applicant has received
7
tentative approval), the date that is 75 days after the
date as of which, as to each of the patents with
respect to which the first applicant submitted and
lawfully maintained a certification qualifying the first
applicant for the 180-day exclusivity period under
subparagraph (B)(iv), at least 1 of the following has
occurred:
(AA) In an infringement action brought against
that applicant with respect to the patent or in a
declaratory judgment action brought by that
applicant with respect to the patent, a court
enters a final decision from which no appeal
(other than a petition to the Supreme Court for a
writ of certiorari) has been or can be taken that
the patent is invalid or not infringed.
(BB) In an infringement action or a declaratory
judgment action described in subitem (AA), a
court signs a settlement order or consent decree
that enters a final judgment that includes a
finding that the patent is invalid or not infringed.
(CC) The patent information submitted under
subsection (b) or (c) of this section is withdrawn
by the holder of the application approved under
subsection (b) of this section.
21 U.S.C. § 355(j)(5)(D)(i)(I) (emphasis added).
The FDA stated its view of the matter in terms echoing
the so-called “first prong” of Chevron, U.S.A. Inc. v. NRDC,
467 U.S. 837 (1984), see, e.g., Mova, 140 F.3d at 1067,
explaining: “The effect of patent delisting on eligibility for
180-day exclusivity is expressly addressed by the [preceding]
plain language.” Dorzolamide Hydrochloride-Timolo
8
Maleate Ophthalmic Solution — 180-day generic drug
exclusivity, Dear ANDA Applicant Letter (Oct. 28, 2008)
(“Hi-Tech Letter”) at 14 n.15., J.A. 121 n.15. A company
otherwise entitled to exclusivity always forfeits it, said the
agency, if events occur satisfying both paragraphs (aa) and
(bb). Id. Paragraph (aa) gets checked off, thanks to its
subsection (BB), as soon as 30 months have passed since the
generic maker filed its ANDA—which had long since
happened in both Cobalt’s and Hi-Tech’s cases. And
paragraph (bb) is taken care of 75 days after the brand
manufacturer delists the challenged patent (under subsection
(CC)), regardless of the purpose or circumstance of the
delisting request. Id. In the later of the two letter rulings, the
FDA wrote that it had “considered and rejected in both this
case and in the matter described in the [Cobalt] Decision, the
argument that eligibility for 180-day exclusivity following the
[brand maker’s] voluntary withdrawal of its patent should be
governed not by the [new] forfeiture provisions, but by the
rule established in Ranbaxy.” Hi-Tech Letter at 14, J.A. 121.
Even though neither Cobalt nor Hi-Tech could have sold its
generic drug before the date that the FDA said amounted to a
“failure to market” event (since unchallenged patents
protected the relevant brand drugs until a good deal later), the
agency announced that both companies had forfeited
exclusivity. Both Cobalt and Hi-Tech sought judicial review,
were denied relief in district court, and didn’t appeal.
Teva filed the ANDAs at issue in this case on December
18, 2003, for Cozaar, and May 24, 2004, for Hyzaar. Both
contained a paragraph IV certification targeting Merck’s U.S.
patent No. 5,608,075, which does not expire until 2014, and
left unchallenged Merck’s other, earlier-expiring patents on
the drugs. In response to Teva’s filing, Merck chose not to
sue for infringement, as it might have. Instead, on March 18,
2005, Merck asked the FDA to delist the 075 patent, which
the agency did, though without making the action public until
9
April 18, 2008. Appellees’ Br. at 17. As of the present date,
the FDA has awarded tentative approval to Teva’s ANDAs,
see Teva, 638 F.Supp.2d at 58 n.12, and also to an ANDA
filed by a competitor of Teva’s, Apotex Inc., to sell generic
Hyzaar, see Reply Br. at 11 n.9. Though the FDA does not
formally announce which ANDA filer was the first to submit
a paragraph IV certification with respect to a brand drug (or
whether any generic manufacturer is officially entitled to
exclusivity) until the date on which generic sales can actually
begin, see 21 C.F.R. § 314.430(b), Teva has every reason to
believe that it was the first filer for both drugs at issue here: it
points to the fact that the FDA’s own website lists the first
paragraph IV certification against Hyzaar (i.e., “Losartan
Potassium and Hydrochlorothiazide”) as having been filed on
the very day that Teva filed its own Hyzaar ANDA. See
http://www.fda.gov (enter “Hyzaar ANDA” in search box;
select sole result, “[PDF] Paragraph IV Patent Certifications”;
scroll to page 16) (last visited December 21, 2009).
But in light of the Hi-Tech Letter, Teva saw the writing
on the wall: under the interpretation of the “plain language” of
the amended statute that the FDA had twice adopted, Teva
had by the fall of 2008 already forfeited the exclusivity it
believed it had earned—on August 12, 2006 for the generic
Cozaar ANDA, and on January 16, 2007 for the generic
Hyzaar ANDA.2 Moreover, the agency had twice rejected the
2
The calculation under the FDA’s understanding of the statute
looks like this: With respect to Cozaar, the date satisfying paragraph
(aa) of the “Failure to Market” forfeiture event is August 12, 2006
(30 months since the filing of the ANDA, see subsection (BB))—
and the date satisfying paragraph (bb) is 75 days after March 18,
2005 (when Merck asked that the drug be delisted, see subsection
(CC)); of the two dates, August 12, 2006 is the later one, hence
(under the opening clause of § 355(j)(5)(D)(i)(I)) the forfeiture
event. With respect to Hyzaar, the analysis is the same, except that
10
contention, made once by Teva itself as a commenter, that its
chosen interpretation of the statute was untenable for a
number of reasons, among them that it was inconsistent with
Ranbaxy. Eschewing presentation of the same argument to
the agency for yet a third time, though the first time with its
own ANDA directly on the line, Teva went straight to the
district court, hoping for a declaratory judgment rejecting the
FDA’s interpretation and an order that the FDA grant it
exclusivity on the date that generic losartan potassium
competition would begin, April 6, 2010.
* * *
The posture of this case raises several significant
questions about its justiciability. One concerns conventional
ripeness. A second, an issue of standing, implicates a
potential—though ultimately illusory—conflict between, on
one hand, decisions of this court regarding a plaintiff’s ability
to obtain pre-enforcement review of a policy adopted by an
agency in an adjudication and, on the other hand, the well-
established teaching of Lujan v. Defenders of Wildlife, 504
U.S. 555, 560-61 (1992), that the imminent threat of injury
inflicted by the defendant and redressable by the court suffices
for constitutional standing.
Ripeness
Pre-enforcement judicial review of an agency’s policy is
available only if the dispute is ripe. Nat’l Park Hospitality
30 months from the date of the ANDA’s filing fell on January 16,
2007. See also Hi-Tech Letter at 15 (saying that the subsection
(CC) event is calculated from the date of the brand maker’s
delisting request, not the date that FDA makes public the delisting).
11
Ass’n v. Dep’t of Interior, 538 U.S. 803 (2003). The ripeness
inquiry probes the fitness for review of the legal issue
presented, along with (in at least some cases) “the hardship to
the parties of withholding court consideration.” Id. at 808.
The “fitness” prong of the analysis generally addresses
“whether the issue is purely legal, whether consideration of
the issue would benefit from a more concrete setting, and
whether the agency’s action is sufficiently final.” National
Ass’n of Home Builders v. U.S. Army Corps of Engineers, 440
F.3d 459, 463 (D.C. Cir. 2006).
In this case, the substantive issues Teva raises are
undoubtedly “purely legal” in the relevant sense. They turn
on questions of statutory construction, see Shays v. FEC, 414
F.3d 76, 95 (D.C. Cir. 2005), and the interpretations chosen
by the FDA and proposed by Teva both constitute bright-line
rules, impervious, so far as appears, to factual variation. This
in itself largely answers the question whether delay might
afford additional “concrete[ness]”; it would not. As to
finality, that largely resolves into the questions whether the
FDA actually has a policy, whether it’s clear what will happen
when the FDA applies the policy to Teva, and whether in any
event it’s sufficiently likely that the policy will matter at all,
given possible uncertainty whether Teva would be entitled to
exclusivity even if the agency’s take on 21 U.S.C.
§ 355(j)(5)(D)(i)(I)(bb)(CC) matched Teva’s.
While the FDA could in principle change its position as
to the effect on generics’ exclusivity of brand makers’
requests to delist, an about-face seems extraordinarily
unlikely. In its brief, the agency maintains, as it did in the
Cobalt matter and the Hi-Tech matter, that the interpretation it
adopted in those instances is compelled by the statute and that
arguments to the contrary are plainly futile. Appellees’ Br. at
42-43 (“[T]he plain language of subsection (CC) makes clear
that the provision applies whenever a patent is withdrawn by
12
the [patent holder.]” (emphasis in original)). The mere
theoretical possibility that an agency could alter its views on a
legal issue before enforcing them against a party has not, in
the past, precluded pre-enforcement review of those views.
The same possibility exists for rulemakings, as we observed in
Association of Bituminous Contractors, Inc. v. Andrus, 581
F.2d 853, 859 (D.C. Cir. 1978), and for less finely chiseled
agency decisions, see Appalachian Power Co. v. EPA, 208
F.3d 1015, 1022 (D.C. Cir. 2000). As in Appalachian Power,
there is here virtually no doubt, as a practical matter, what
approach the agency will apply to Teva. And the implication
of the FDA’s position for any exclusivity that Teva would
otherwise merit is equally clear: as discussed above, the
unambiguous result of the agency’s interpretation is that any
such entitlement is already forfeited.
The government argues, however—relying chiefly on
Pfizer Inc. v. Shalala, 182 F.3d 975 (D.C. Cir. 1999)—that the
issue nevertheless remains unfit for review because the
agency’s challenged interpretation may not be dispositive of
the question whether Teva ultimately deserves exclusivity. In
Pfizer a brand manufacturer (Pfizer) filed suit alleging that the
FDA’s mere acceptance of an ANDA for processing was
unlawful because the proposed generic drug differed in a
crucial respect from the product it sought to replicate. 182
F.3d at 978. We found the suit unripe, suggesting that despite
the FDA’s tentative approval of the generic’s ANDA, grounds
for uncertainty over whether the generic drug would ever be
approved for sale persisted, posing concerns for “piecemeal
litigation”: we instanced a possible FDA finding of a lack of
bioequivalence, a matter that we obviously assumed the
tentative approval left open. Id. at 980.
The absence of any colorable factual dispute in Teva’s
case compels a different outcome from Pfizer. The FDA
makes no suggestion that any possible deficiency or
13
uncertainty in Teva’s ANDA could thwart final approval. It
offers no reason to doubt the conclusion that the first
paragraph IV certification against Hyzaar, filed on May 24,
2004, was the paragraph IV certification against Hyzaar that
Teva filed on May 24, 2004—which in turn dictates that Teva
has satisfied the threshold requirement for exclusivity. The
agency does caution that one or more of the statutory
“forfeiture events” other than a “failure to market” might in
any case deprive Teva of exclusivity before final approval—
but as Teva’s counsel ably demonstrated at oral argument, any
such outcome is virtually inconceivable: Teva will not
withdraw its ANDA, see 21 U.S.C. § 355(j)(5)(D)(i)(II); it
will not amend its paragraph IV certification, see § (D)(i)(III);
it has already obtained tentative approval, see § (D)(i)(IV);
there is no indication that it will enter a collusive agreement
with Merck, see § (D)(i)(V); and the now-delisted patent will
not expire, see § (D)(i)(VI). See Oral Argument Tr. at 29-30
(Dec. 7, 2009). In short, the question before us is one of pure
statutory interpretation; we know precisely what the FDA
thinks the answer is; and its resolution will almost certainly
determine whether Teva is entitled to the exclusivity it claims.
The second prong of the ripeness analysis addresses
“whether postponing judicial review would impose an undue
burden on” the parties. National Ass’n of Home Builders, 440
F.3d at 464 (emphasis in original). This court has frequently
suggested that hardship is not a sine qua non of ripeness. See
id. at 465 (“[W]here . . . there are no significant agency or
judicial interests militating in favor of delay, [lack of]
hardship cannot tip the balance against judicial review.”
(second bracketed alteration in original, internal quotation
marks omitted)); Electric Power Supply Ass’n v. FERC, 391
F.3d 1255, 1263 (D.C. Cir. 2004) (“The hardship prong under
the ripeness doctrine is largely irrelevant in cases . . . in which
neither the agency nor the court have [sic] a significant
interest in postponing review.”); AT&T Corp. v. FCC, 349
14
F.3d 692, 700 (D.C. Cir. 2003) (“The ‘hardship’ prong of the
Abbott Laboratories [v. Gardner, 387 U.S. 136 (1967)] test is
not an independent requirement divorced from the
consideration of the institutional interests of the court and
agency. Thus, where there are no institutional interests
favoring postponement of review, a petitioner need not satisfy
the hardship prong.” (internal citation omitted)); Village of
Bensenville v. FAA, 376 F.3d 1114, 1120 (D.C. Cir. 2004)
(“[A]lthough the FAA reasonably asserts that the
municipalities will not ‘suffer [any] immediate hardship from
an EIS,’ Respondent’s Br. at 23, we see no benefit to us in
postponing review[.]” (emphasis and second bracketed
alteration in original)). In this case we need not consider the
effect of a failure to show hardship, as Teva faces at least one
harm from delayed judicial review cognizable in the ripeness
analysis: a near-certain loss of the first-mover advantage to
which the company claims entitlement.3
3
Teva also alleges hardship resulting from the severe impact
of uncertainty on investment decisions that it must make well
before the first legal opportunity to sell its generic, whether as an
exclusive (as it claims) or not (under the FDA’s view). Delayed
resolution of the issues in this case will, depending on the
assumptions under which it operates, either (1) cost the company
much of a valuable (and lawful) commercial opportunity, if it
mistakenly assumes that the FDA view will prevail and therefore
refrains from investing sufficient resources to prepare for the
increased demand that would accompany an exclusive as opposed
to a non-exclusive product launch, or (2) waste hundreds of millions
in company resources invested in anticipation of fully exploiting its
exclusivity, if it mistakenly assumes that its view will prevail. See
Declaration of David Marshall, Vice President of New Products
Portfolio Strategy for Teva Pharmaceuticals USA, Inc., at 4-8, J.A.
128-32. (Of course a straddling investment decision would entail
some of each cost.) We express no view as to whether such harm
counts in the ripeness analysis. Cf. Exxon Mobil Corp. v. FERC,
15
If Teva is right on the merits (as we must assume it is for
purposes of the ripeness inquiry, see U.S. Air Tour Ass’n v.
FAA, 298 F.3d 997, 1014 (D.C. Cir. 2002)), then as of April 6,
2010, it will be entitled to start enjoying its exclusivity period
and to continue doing so for 180 days before additional firms
lawfully enter the market. This “first-mover advantage” is a
valuable asset. In Mova we observed “the loss of [a generic’s]
officially sanctioned head start” can, at least in some
circumstances, yield a “severe economic impact.” 140 F.3d at
1066 n.6 (internal quotation marks omitted). If we refrained
from adjudicating this dispute now, Teva would almost
certainly face competition from Apotex on April 6, see 21
C.F.R. § 314.105(d) (explaining that a “tentative” approval is
the same as a final approval with a delayed effective date)—
an injury that would not be remedied by Teva’s securing 180
days of exclusivity later on.
District courts in this circuit routinely reach the merits of
generic manufacturers’ claims to exclusivity before the FDA
has granted final approval to any ANDA concerning the drug
at issue. See, e.g., Teva, 548 F.3d 103 (earliest possible date
of generic competition June 29, 2008, see Appellee’s Br. at 5;
district court decision April 11, 2008, id. at 4); Ranbaxy, 469
F.3d 120 (earliest possible date of generic competition June
23, 2006, see Appellants’ Br. at 11; district court decision
April 30, 2006, id. at 1). This makes good sense; the
exclusivity reward that Congress made available as an
incentive for patent challenges is time-sensitive, and where
there is no material ambiguity about essential facts a court can
501 F.3d 204, 208 (D.C. Cir. 2007) (hardship ample where
postponing review would cause uncertainty and cost to prospective
applicant for approval to build pipeline and would “tend to inhibit
or delay investment” in a project Congress had deemed important).
16
readily decide whether it has been earned in advance of
generic competition’s onset. The alternative approach—
delaying review until the agency has made its technically
tentative decisions final—puts a court in an awkward bind,
unless it miraculously manages to resolve the merits issue
more or less instantaneously. Apart from that risky and
improbable course, there would be two possible stopgaps
available to preserve the first-mover advantage. The court
could delay all generic competition, thereby thwarting the
statutory purpose of achieving swift competition by generics
(a factor that would in turn weigh against preliminary
injunctive relief under the “public interest” component of the
standard test). Or it could delay the entrance of the
exclusivity claimant’s generic rivals into the market, thereby
giving the claimant precisely the relief it seeks, simply in
order to allow the court time to decide whether such relief was
warranted. The technical possibility that a judge might
embrace one of these highly imperfect alternatives can hardly
be thought to protect Teva from the hardship made likely by
delayed review.
When the question at issue is well-defined, and when
withholding judicial consideration would cause undeniable
harm, as here, ripeness concerns pose no obstacle to pre-
enforcement review.
Standing
The FDA embraced the statutory interpretation that Teva
now seeks to challenge not in a rulemaking but in two
adjudications to which Teva was not a party (though actively
commenting in one). Our past cases suggest some uncertainty
whether a dispute in that posture can ever be justiciable. See,
e.g., Radiofone, Inc. v. FCC, 759 F.2d 936, 938 (D.C. Cir.
1985) (opinion of then-Judge Scalia) (“All persons adversely
affected by [a] rule [“addressed, so to speak, to the world at
17
large”] would have standing to challenge its compliance with
legal prescriptions designed for their protection. . . . The
situation is different, however, when an interpretation of a
statute, or some other legal principle, is set forth as the
rationale of an adjudication.” (emphasis in original)).
But straightforward application of hornbook doctrine
yields the conclusion that Teva has standing. Article III of the
Constitution requires that a federal court plaintiff allege an
actual or imminent injury that is fairly traceable to the
defendant’s challenged conduct and redressable in the judicial
proceeding. Lujan, 504 U.S. at 560-61. In this instance, the
latter two elements are clearly satisfied. Any imminent
deprivation of Teva’s allegedly deserved exclusivity would be
directly attributable to the FDA’s statutory interpretation.
And if we agreed with Teva on the merits, we (or the district
court) could issue precisely the declaration it has sought,
announcing that requests to delist challenged patents should
have no more legal significance in the amended statutory
regime than they did in the old one, as per Ranbaxy, 469 F.3d
at 126.
The “injury” prong of the standard standing inquiry is a
bit thornier—but only to the extent of the trivial uncertainty
whether the FDA will on April 6, 2010 stick to the
interpretation that Teva attacks here. As discussed in the
ripeness analysis above, however, we find no uncertainty to
speak of on the matter. It is clear what the FDA will do
absent judicial intervention and what the effect of the
agency’s action will be. The inescapable implication is that
Teva faces an imminent threat of the same harm that has
sufficed for Article-III injury purposes in all of our past drug-
approval cases: the impending prospect of allegedly unlawful
competition in the relevant market. See, e.g., Bristol-Myers
Squibb Co. v. Shalala, 91 F.3d 1493, 1497 (D.C. Cir. 1996)
(“[W]here . . . a statutory provision reflects a legislative
18
purpose to protect a competitive interest, the protected
competitor has standing to require compliance with that
provision.”); Ranbaxy, 469 F.3d 120 (adjudicating a dispute in
which the only injury at issue was the prospective loss of a
generic manufacturer’s 180-day period of marketing
exclusivity). For the purpose of the classic constitutional
standing analysis, it makes no difference to the “injury”
inquiry whether the agency adopted the policy at issue in an
adjudication, a rulemaking, a guidance document, or indeed
by ouija board; provided the projected sequence of events is
sufficiently certain, the prospective injury flows from what the
agency is going to do, not how it decided to do it. Cf. City of
Los Angeles v. Lyons, 461 U.S. 95, 106 n.7 (1983) (“[T]o have
a case or controversy . . . [plaintiff] would have to credibly
allege that he faced a realistic threat from the future
application of the City’s policy.”).
The question, then, is whether the normal application of
the constitutional standing doctrine is suspended when the
court’s knowledge that an agency is about to inflict injury on a
party derives from an agency policy that originated in an
adjudication (or several). The strongest support for such a
principle would be Sea-Land Service, Inc. v. Department of
Transportation, 137 F.3d 640, 648 (D.C. Cir. 1998), in which
we rejected a pre-enforcement challenge to an agency
interpretation born of an adjudication, noting that a policy’s
“mere precedential effect within an agency is not, alone,
enough to create Article III standing, no matter how
foreseeable the future litigation” involving the plaintiff. We
have articulated a similar idea, albeit in weaker form, on
numerous other occasions. See, e.g., Shipbuilders Council of
America v. United States, 868 F.2d 452, 456 (D.C. Cir. 1989)
(“[W]e know of no authority recognizing that the mere
potential precedential effect of an agency action affords a
bystander to that action a basis for complaint.”); American
Family Life Assurance Co. v. FCC, 129 F.3d 625, 629 (D.C.
19
Cir. 1997) (“AFLAC”) (“[W]e have said before, and we say
again, that the ‘mere precedential effect of [an] agency’s
rationale in later adjudications’ is not an injury sufficient to
confer standing on someone seeking judicial review of the
agency’s ruling.” (quoting Radiofone, 759 F.2d at 939)).
In all of these cases, we rebuffed efforts to obtain pre-
enforcement review of policies embraced by agencies in
adjudications. In each instance, however, the failure to
demonstrate standing is more naturally understood as arising
from the lack of a sufficiently imminent and concrete injury
than from some sort of ad hoc exception to otherwise-
universally applicable constitutional doctrine. Radiofone, for
example, addressed whether parties allegedly aggrieved by
reasoning employed by the FCC in an adjudication could
appeal the agency’s order even though the recipient of the
order had since ceased doing business. 759 F.2d at 937-38.
There was no suggestion in any of the panel’s three
opinions—including then-Judge Scalia’s, which didn’t in any
case garner a majority for its standing passage—that the
parties seeking review were at risk of injury from imminent
application of the principle the agency had articulated.
Shipbuilders similarly concerned no identifiable prospective
application of the allegedly offending policy. We explicitly
noted, in fact, that plaintiffs had failed to present “specific,
concrete facts demonstrating that the challenged [ruling
would] harm” them, adding that their “hypothesizing . . .
never descends from a highly general plane; it remains at a
considerable distance from the more concrete pleas” needed to
establish standing. Shipbuilders, 868 F.2d at 457. While the
opinion also framed the complaint as an impermissible
“request for judicial advice—a declaration that a line of
agency rulings should henceforth have no precedential effect,”
id. at 456, we simply did not address the scenario in which a
line of agency rulings threatened a party with an imminent
injury otherwise ample for Article III purposes.
20
Sea-Land, too, did not involve a party pointing to a
particular imminent application of the disputed agency policy.
The justiciability problem in that case arose from the
“principle that prevailing parties lack standing to appeal,” 137
F.3d at 647—which is undoubtedly correct as a general
matter, but which does not foreclose review of a case in which
a party is aggrieved not by the “mere potential precedential
effect of an agency action,” Shipbuilders, 868 F.2d at 456, but
instead by the impending application of an agency’s statutory
interpretation, the firmness of which is not in dispute, on a
fast-arriving date certain. The Sea-Land opinion, to be sure,
phrased the proscription against challenges to agency
precedent qua precedent as one applying “no matter how
foreseeable the future litigation.” 137 F.3d at 648. But we
could not possibly have purported to overturn well-established
Supreme Court precedent holding that an imminent threat of
injury suffices for standing, see Lujan, 504 U.S. at 560—
particularly in a case not involving the slightest allegation of
such a threat. A more sensible reading of Sea-Land is one that
leaves it consistent with Lujan and its equally binding
progeny: merely foreseeable future litigation resulting from a
statutory interpretation that an agency has adopted in an
adjudication is, “alone,” 137 F.3d at 648—i.e., without
more—too speculative to satisfy Article III’s injury-in-fact
requirement. An agency’s imminent application of its
established interpretation of a statute, at the potential cost of
hundreds of millions of dollars to the regulated firm, remains,
by contrast, as sufficient for standing purposes today as it was
before Sea-Land. See Marshall Declaration at 4-5, J.A. 128-
29 (explaining why Teva “stands to lose hundreds of millions
of dollars in net revenues during its first year of generic
losartan potassium products sales as a direct result of the
[FDA’s policy]”).
No other case we’ve decided concerning a pre-
enforcement challenge to an agency interpretation adopted via
21
adjudication counsels a contrary result. See AFLAC, 129 F.3d
at 628 (“Petitioner reports no litigation on the horizon . . . no
simmering disputes about to erupt into a lawsuit[.]”); Shell Oil
Co. v. FERC, 47 F.3d 1186, 1202 (D.C. Cir. 1995) (“Shell’s
allegations of injury rest on a hypothetical scenario . . . .
Although such injury is not inconceivable, we are
unpersuaded that it is imminent.” (emphasis in original));
Crowley Caribbean Transp., Inc. v. Pena, 37 F.3d 671, 674
(D.C. Cir. 1994) (finding impact of agency’s challenged
position on party seeking review “nebulous and remote”);
Aeronautical Radio, Inc. v. FCC, 983 F.2d 275, 284 (D.C. Cir.
1993) (“There is no indication in the record . . . that the
Commission is likely to attempt to [enforce the challenged
interpretation against TRW, the party seeking review].
TRW’s alleged injury is therefore merely conjectural.”
(internal quotation marks omitted)).
We have, on the other hand, allowed a party to challenge
in advance an agency policy adopted via adjudication when
the prospect of impending harm was effectively certain. In
International Brotherhood of Electrical Workers v. ICC, 862
F.2d 330 (D.C. Cir. 1988), a union sought judicial review of
the Interstate Commerce Commission’s exercise of
jurisdiction to review an arbitration award—even though the
ICC, having accepted jurisdiction, had ruled in favor of the
union. 862 F.2d at 334. We found ripeness and standing
requirements satisfied, noting that “[b]ecause of the ICC’s
decision to review arbitration awards, the union will be
subject to agency review in future cases involving disputes” of
the same type. Id. As we later explained, International
Brotherhood stands for the proposition that the “concrete cost
of an additional proceeding is a cognizable Article III injury,”
Sea-Land, 137 F.3d at 648—notwithstanding that the source
of the harm was an agency position adopted in an adjudication
whose outcome was no longer at issue. Teva’s alleged injury
22
threatens to impose no less of a “concrete cost” and with no
less certainty.
We have, moreover, explicitly sanctioned review of a case
in the present posture—albeit while framing the justiciability
question as one of ripeness rather than standing. Association
of Bituminous Contractors v. Andrus, 581 F.2d 853, 858-59
(D.C. Cir. 1978), was precisely a pre-enforcement challenge
to a policy adopted in a previous adjudication by the Interior
Board of Mine Operations Appeals. The doctrine of standing
has undoubtedly evolved significantly since the time of that
decision (though not generating any new limits on imminent
injuries that happen to be traceable to adjudicative rules)—but
the case does demonstrate that we have previously considered
the lawfulness of an agency policy with precisely the kind of
provenance as the policy Teva challenges, where imminent
application of the policy was about to inflict injury. See also
Independent Insurance Agents of America, Inc. v. Hawke, 211
F.3d 638 (D.C. Cir. 2000) (adjudicating dispute over agency
interpretation adopted in letter ruling to which district court
plaintiff was not party); Air Transport Ass’n of America, Inc.
v. FAA, 291 F.3d 49 (D.C. Cir. 2002) (same, where dispute
concerned letter ruling to which circuit court petitioner was
not party).
We see no basis for concluding that this court has created
an exception to the Supreme Court’s constitutional standing
doctrine excising cases like Teva’s from the class of otherwise
justiciable matters. Teva presents a valid Article III case or
controversy.
* * *
On the merits, we review de novo the district court’s
grant of summary judgment to the FDA. See Kersey v.
23
Washington Metropolitan Area Transit Authority, 586 F.3d
13, 16 (D.C. Cir. 2009). We evaluate the FDA’s
interpretations of the Food, Drug, and Cosmetic Act adopted
in letter rulings under the familiar two-part Chevron
framework. Mylan Labs., Inc. v. Thompson, 389 F.3d 1272,
1280 (D.C. Cir. 2004). But see Matthew C. Stephenson and
Adrian Vermeule, Chevron Has Only One Step, 95 Va. L.
Rev. 597 (2009).
Teva offers two principal reasons to conclude that the
FDA may not allow a brand manufacturer’s request to delist a
challenged patent to trigger a statutory “forfeiture event”
resulting in the loss of a generic’s exclusivity. One reason
takes the form of linguistic analysis focused almost entirely on
the text of the “failure to market” forfeiture event and a
related provision. The 2003 amendments, Teva explains,
introduced a new procedure, a counterclaim in the brand
manufacturer’s patent infringement suit, through which
generic companies can force brand companies to delist an
improperly asserted patent. See 21 U.S.C.
§ 355(j)(5)(C)(ii)(I).4 This counterclaim provision is the only
portion of the statute that explicitly provides for the delisting
of a patent after it has been challenged in an ANDA. In the
company’s view, that singular reference requires the
4
The purpose of this procedure, says Teva, is to offer generics
a means of combating brand companies’ practice of delaying
generic competition by listing “sham patents,” baiting a generic into
filing a paragraph IV certification, and then filing an infringement
suit—which typically brings a 30-month stay of generic
competition. Appellant’s Br. at 42; see 21 U.S.C. § 355(j)(5)(B)(iii)
(creating the stay and subjecting it to various limits such as the
generic manufacturer’s earlier success in the suit); aaiPharma Inc.
v. Thompson, 296 F.3d 227, 236 (4th Cir. 2002) (describing
precisely this delay tactic).
24
conclusion that the counterclaim provision describes the only
scenario in which the FDA may delist a challenged patent.
Obviously, then, no other kind of delisting could ever serve as
an occurrence satisfying the terms of the “failure to market”
forfeiture trigger listed at 21 U.S.C.
§ 355(j)(5)(D)(i)(I)(bb)(CC).
The FDA, for its part, responds that “the plain language
of the statute contains no limitation on when delisting can
occur.” Appellees’ Br. at 44. Brand manufacturers are thus
free to delist challenged patents whenever they please—and
any such delisting satisfies subsection (CC) of the “failure to
market” forfeiture section. Id. at 45-46. In effect, the agency
says, the counterclaim provision says nothing about its being
an exclusive route to delisting, and if Congress meant to
confine subsection (CC) delistings to those arising from the
counterclaim procedure, it would have been natural for it to
place that limitation in (CC).
While Teva’s purely linguistic argument shows its
understanding of the relevant language to be perfectly
plausible, it hardly rules out alternative readings that, absent
consideration of statutory structure, also appear plausible. See
Chevron, 467 U.S. at 844-45; INS v. Cardoza Fonseca, 480
U.S. 421, 443 (1987) (considering “the structure of the Act” at
Chevron step one). As the FDA notes, there is simply no
express preclusion of non-counterclaim delistings, or of such
delistings’ triggering forfeiture, in either of the places one
might expect to find one, the counterclaim section or (CC).
This brings us to Teva’s structural argument. Ranbaxy,
Teva notes, concerned an FDA policy with a virtually
identical effect. See 469 F.3d at 125. This court condemned
that rule, partly because it allowed a brand manufacturer,
25
by delisting its patent, to deprive the generic applicant of
a period of marketing exclusivity. By thus reducing the
certainty of receiving a period of marketing exclusivity,
the FDA’s delisting policy diminishe[d] the incentive for
a manufacturer of generic drugs to challenge a patent . . .
in the hope of bringing to market a generic competitor for
an approved drug without waiting for the patent to expire.
The FDA may not, however, change the incentive
structure adopted by the Congress, for the agency is
bound “not only by the ultimate purposes Congress has
selected, but by the means it has deemed appropriate, and
prescribed, for the pursuit of those purposes.”
Id. at 126 (emphasis added, citation omitted). Nothing in the
2003 amendments to the Food, Drug, and Cosmetic Act
altered that essential incentive structure, says Teva, so the
preceding portion of Ranbaxy remains applicable even under
the new regime. Indeed, it is true that the 2003 amendments
say nothing specific to undermine our prior understanding of
the statute’s intended incentive structure.
But the FDA sees a way in which its interpretation of
subsection (CC) accomplishes at least some congressional
purpose. Without the possibility of a forfeiture of exclusivity
resulting from the delisting of a challenged patent, a generic
manufacturer that had been awarded exclusivity could delay
all generic competition more or less indefinitely, since by
statute the agency can’t approve competing generics until 180
days after the first paragraph-IV filer has begun commercial
marketing of its newly approved product. See 21 U.S.C.
§ 355(j)(5)(B)(iv)(I). Congress enacted the “failure to
market” provision, in the agency’s view, precisely to avoid
such “parking” of exclusivity; allowing a brand maker to
trigger forfeiture by delisting a challenged patent positively
furthers that legislative aim. Appellees’ Br. at 45. Besides,
the agency says, “Consumers benefit from lower drug prices
26
immediately without having to wait for one generic company
to enjoy 180 days of exclusivity when the patent owner itself
takes the position that a patent should not hinder FDA
approval of ANDAs.” Id.The real issue, then, is whether the FDA is right that the
2003 addition of the “failure to market” forfeiture provision,
21 U.S.C. § 355(j)(5)(D)(i)(I), altered the statute’s incentive
structure to the point that Ranbaxy’s reasoning no longer
controls the agency’s treatment of a delisting request in the
wake of a paragraph-IV filing.
The terms of § 355(j)(5)(D)(i)(I), quoted in full in the
opening of this opinion, create five possible dates on which a
generic manufacturer otherwise entitled to exclusivity can
forfeit it: (1) 75 days after the agency finally approves the
relevant ANDA; (2) 30 months after the generic submits the
relevant ANDA; (3) 75 days after a court judgment that the
challenged patent is invalid or not infringed; (4) 75 days after
a suit over the challenged patent is settled favorably to the
ANDA filer; and (5) 75 days after the challenged patent is
delisted. No forfeiture occurs, however, unless one of dates
(1)-(2) and one of dates (3)-(5) have come to pass. See id.;
FDA Letter re 180-day exclusivity, Docket No. 2007N-0389,
ANDA 77-165: Granisetron Hydrochloride Injection, 1
mg/mL, at 5, J.A. 68 (“We find that under the plain language
of the statute, 180-day exclusivity is not forfeited for failure to
market when an event under subpart (aa) has occurred, but . . .
none of the events in subpart (bb) has occurred.”). Setting
aside the subsection at issue in this case—listed as (5) above,
and codified as (bb)(CC)—the “failure to market” forfeiture
provision does not permit a brand manufacturer to vitiate a
generic’s exclusivity without the generic manufacturer’s
having had some say in the matter. No forfeiture can take
place unless the brand manufacturer brings an infringement
suit against the generic and either loses on the merits or enters
27
an unfavorable settlement agreement. The latter necessarily
entails some participation by the generic; the former
invariably involves significant expense for the brand
manufacturer, and affords the victorious generic the
opportunity to ask the court to delay entering final judgment
until a date that would not trigger forfeiture prematurely—
before the agency grants final approval to the relevant ANDA.
The FDA’s view turns the last alternative among events
(3)-(5) into a fundamentally different forfeiture trigger: it is
satisfied when the patent targeted in a paragraph-IV filing “is
withdrawn by the” brand manufacturer, full stop—meaning
that Congress has now explicitly provided for a scenario in
which the brand maker can unilaterally deprive the generic of
its exclusivity. The agency, however, offers not a single
cogent reason why Congress might have permitted brand
manufacturers to trigger subsection (CC) by withdrawing a
challenged patent, outside the counterclaim scenario identified
by Teva.
The argument that the plain language of the statute
imposes no limit on the circumstances in which the agency
may effectuate delisting requests fails. Precisely the same
could have been said of the version of the statute that Ranbaxy
addressed, and we nevertheless concluded that its structure
precluded an FDA rule allowing the agency “to delist a patent
upon the request of the [brand manufacturer]” when the
delisting would rob the generic maker of earned exclusivity.
469 F.3d at 125.
The agency fares no better in suggesting that allowing the
delisting of challenged patents prevents the ANDA filer from
“creat[ing] a bottleneck” blocking generic competition by
“parking” its exclusivity. Appellees’ Br. at 45. As a parking-
prevention device, letting brand makers delist challenged
patents in order to trigger a forfeiture of exclusivity would be
28
completely ineffective; given the incentives for the brand
manufacturer, it will be used only where its impact on
Congress’s scheme is most destructive. If the generic appears
likely to park its exclusivity, the brand maker will simply
refrain from delisting altogether, thus enjoying an extended
period during which it faces no generic competition while the
exclusivity-holder bides its time.5 If the generic appears
unlikely to park its exclusivity, the brand maker can delist
well before the generic can go to market, thus eviscerating the
exclusivity incentive altogether. In other words, the only case
in which a unilateral right for brand makers to delist
challenged patents actually results in the outcome the FDA
touts is when the brand maker deliberately accelerates the
onset of generic competition—an utterly implausible scenario.
In other cases, the brand maker either does nothing to prevent
parking, or prevents parking that was unlikely to have
occurred in any event, but with precisely the effect that
Ranbaxy proscribed. Thus the “parking” concern offers no
reason to conclude that the 2003 addition of forfeiture
provisions meant to give the brand manufacturer a right to
unilaterally vitiate a generic’s exclusivity.
Finally, the FDA’s sole effort to root its interpretation in
the policy underlying Hatch-Waxman—the thought that the
interpretation benefits consumers by allowing full generic
competition without a 180-day delay—betrays a
misunderstanding of the exclusivity incentive. The statute’s
grant of a 180-day delay in multiple generic competition for
the first successful paragraph IV filer is a pro-consumer
5
We note, in fact, that many instances of generics’ parking
their exclusivity have evidently arisen thanks to agreements with the
brand maker itself to delay generic competition. See Federal Trade
Commission, Authorized Generics: An Interim Report ch. 2, at 1
(2009).
29
device. And it happens to be precisely the device Congress
has chosen to induce challenges to patents claimed to support
brand drugs. The statute thus deliberately sacrifices the
benefits of full generic competition at the first chance allowed
by the brand manufacturer’s patents, in favor of the benefits of
earlier generic competition, brought about by the promise of a
reward for generics that stick out their necks (at the potential
cost of a patent infringement suit) by claiming that patent law
does not extend the brand maker’s monopoly as long as the
brand maker has asserted. As Congress deliberately created
the 180-day exclusivity bonus, the FDA cannot justify its
interpretation by proudly proclaiming that it has eviscerated
that bonus.
We see nothing in the 2003 amendments to the Food,
Drug, and Cosmetic Act that changes the structure of the
statute such that brand companies should be newly able to
delist challenged patents, thereby triggering a forfeiture event
that deprives generic companies of the period of marketing
exclusivity they otherwise deserve. For that reason, the
interpretation of the statute that the FDA has adopted in two
recent adjudications, and that it regards itself as bound by law
to apply to Teva’s ANDAs for losartan products, fails at
Chevron step one. Cf. Ranbaxy, 469 F.3d at 126; Cardoza
Fonseca, 480 U.S. at 443.
* * *
One matter remains. Teva’s prospective generic losartan
competitor, Apotex, sought to intervene as a defendant in
Teva’s suit before the district court. The court denied the
intervention on the ground that Apotex lacked standing. Teva,
638 F.Supp.2d at 59. Apotex has appealed that ruling, but has
also, with the consent of both parties, expressed its substantive
views of this case in an amicus brief, which we have
30
considered no less than if Apotex had formally intervened. As
Apotex and the FDA are as a practical matter identically
positioned on the issues (though from radically different
perspectives), we think it prudent to follow the line of
precedent in this circuit declining to assess a would-be
intervenor’s standing when answering the question wouldn’t
affect the outcome of the case. See Comcast Corp. v. FCC,
579 F.3d 1, 6 (D.C. Cir. 2009) (“We need not decide whether
[the harm alleged by a prospective intervenor] is too
‘conjectural or hypothetical’ to support standing . . . because
‘if one party has standing in an action, a court need not reach
the issue of the standing of other parties when it makes no
difference to the merits of the case.’” (quoting Railway Labor
Executives Ass’n v. United States, 987 F.3d 806, 810 (D.C.
Cir. 1993))); see also McConnell v. FEC, 540 U.S. 93, 233
(2003) (“It is clear . . . that the Federal Election Commission
(FEC) has standing, and therefore we need not address the
standing of the intervenor-defendants, whose position here is
identical to the FEC’s.”). We note that courts appear not to
have considered whether a party whose attempt to intervene
has been pretermitted in this fashion (or a party whose
standing has otherwise been left unresolved) can seek review
of the court’s decision on the merits, as a successful
intervenor could. Perhaps courts have assumed that that issue
could reasonably be kicked up the road to the possible
appellate body. Finally, we also note that Apotex might move
again for intervention in future proceedings before the district
court in this case in light of changed circumstances—
specifically that Apotex’s ANDA has now earned tentative
approval from the FDA, effectively removing the obstacle to
standing on which the district court relied.
31
* * *
We therefore reverse the judgment of the district court,
but, as the court has yet to address the appropriateness of each
form of relief that Teva has sought, we remand for further
proceedings not inconsistent with this opinion.
So ordered.
KAREN LECRAFT HENDERSON, Circuit Judge, dissenting:
I dissent from the majority opinion because the issue Teva
seeks to litigate—its statutory eligibility vel non to exclusively
market generic versions of Cozaar and Hyzaar, brand name
drugs manufactured by Merck & Co., Inc. (Merck)—will not be
ripe unless and until the United States Food and Drug
Administration (FDA) issues its final decision either granting or
denying Teva’s Abbreviated New Drug Application (ANDA).
The United States Supreme Court has established a two-pronged
test for determining ripeness, requiring that the court analyze:
“(1) the fitness of the issues for judicial decision and (2) the
hardship to the parties of withholding court consideration.”
Nat’l Park Hospitality Ass’n v. Dep’t of Interior, 538 U.S. 803,
808 (2003) (citing Abbott Labs. v. Gardner, 387 U.S. 136, 149
(1967)). This action satisfies neither prong of the ripeness test
as is clear from our decision in Pfizer Inc. v. Shalala, 182 F.3d
975 (D.C. Cir. 1999).
In that case, Pfizer filed a “citizen petition” with the FDA
asking that the agency recognize as “a distinct dosage form” a
patented “osmotic pump” used as an extended release
mechanism for Pfizer’s brand drug Procardia XL. Pfizer, 182
F.3d at 977. Almost four years later—with the petition still
pending—Mylan Pharmaceuticals, Inc. (Mylan) filed an ANDA
to market a generic version of Procardia XL, claiming
pharmaceutical equivalence notwithstanding Mylan’s product
used a different release mechanism. After the FDA accepted
Mylan’s ANDA for processing but before it decided whether to
approve it, Pfizer filed a suit in district court challenging the
FDA’s acceptance of the ANDA on the ground that the two
products were not equivalent because Pfizer’s osmotic pump
was a unique dosage form and thus distinct from Mylan’s
mechanism. The district court held that Pfizer’s challenge to
Mylan’s application was not ripe for judicial review but that its
unresolved citizen petition was. Id. at 978. On appeal, we
found neither challenge ripe.
2
We first rejected Pfizer’s argument that “once having
decided, based upon the information contained in Mylan’s
application, that Mylan’s drug uses the same dosage form as
Procardia XL®, the FDA will not ‘alter its views with respect to
the necessity of Mylan filing a suitability petition.’ ” Id. at 978.
We explained:
The decision to accept Mylan’s ANDA for processing
as a pharmaceutical equivalent to Procardia XL® is . . .
merely the first step in the agency’s approval process.
The critical fact remains that the FDA may never
approve Mylan’s application—whether because it
decides in the end that the dosage form of Mylan’s
drug is different from that of Procardia XL® or for
some entirely different reason, such as a lack of
bioequivalence. Therefore, “depending upon the
agency’s future actions . . . review now may turn out to
have been unnecessary” and could deprive the agency
of the opportunity to apply its expertise and to correct
any mistakes it may have made.
Id. (quoting Ohio Forestry Ass’n v. Sierra Club, 523 U.S. 726,
736 (1998)) (first ellipsis added). Teva faces the same hurdle
here. We do not know whether the FDA’s final decision will
approve Teva’s ANDA or what the FDA’s reasoning will be if,
as the majority forecasts, maj. op. at 11-13, it does not. The
FDA may conclude Teva forfeited its eligibility upon Merck’s
delisting of its patents, as Teva and the majority insist it will, or
it may reject Teva’s application based on one of the other
forfeiture provisions “or for some entirely different reason, such
as a lack of bioequivalence.” Pfizer, 182 F.3d at 978.1 Because
1
The FDA’s “tentative approval” of Teva’s ANDA is not, as Teva
suggests, Reply Br. at 10-11, the final word on its generic drug’s
equivalence. See Pfizer, 182 F.3d at 980 (although FDA’s post-oral
argument tentative approval of Mylan’s generic made it “more likely
3
the FDA has not yet issued its decision we are unable to divine
its substance. Given this uncertainty and the consequent
possibility the court may not need to resolve the
delisting/forfeiture issue after the FDA’s final decision, Teva’s
challenge to the FDA’s previous decisions in other proceedings
is not now fit for review under the first prong of the ripeness
test. In short, “[i]t makes no sense for us to anticipate a wrong
when none may ever arise.” Cronin v. FAA, 73 F.3d 1126, 1132
(D.C. Cir. 1996).
Nor does Teva fare better under the test’s hardship prong as
we applied it in Pfizer. There we explained that Pfizer was not
able to “point to any imminent hardship arising from the FDA’s
acceptance of Mylan’s ANDA”:
Before Pfizer could suffer its claimed “economic injury
from unlawful competition,” FDA approval for a
pharmaceutical equivalent to Procardia XL® would
have to be not only sought but granted. That has not
happened. Therefore “no irremediable adverse
consequences flow from requiring a later challenge.”
Pfizer, 182 F.3d 979 (quoting Toilet Goods Ass’n v. Gardner,
387 U.S. 158, 164 (1967)). For the same reason, Teva too will
suffer no imminent hardship if review is postponed. See Fed.
Express Corp. v. Mineta, 373 F.3d 112, 119 (D.C. Cir. 2004)
(hardship prong not satisfied because “postponing review . . .
w[ould] not be a hardship to [petitioners], let alone a hardship
that is ‘immediate, direct, and significant.’ ” (quoting State Farm
Mut. Auto. Ins. Co. v. Dole, 802 F.2d 474, 480 (D.C. Cir.
1986))) (emphasis added). As in Pfizer, the delay will not
“foreclose[ the appellant’s] right ever to get meaningful judicial
that the FDA w[ould] eventually approve Mylan’s drug, the agency’s
tentative approval cause[d] Pfizer no hardship at present or in the near
future, nor d[id] it render Pfizer’s challenge fit for review”).
4
review,” 182 F.3d at 979; upon the FDA’s issuance of an
adverse final order, Teva is free to seek judicial
review—forestalling generic competition and the loss of the
“first-mover advantage,” maj. op. at 15, through appropriate and
immediate injunctive relief.2
For the foregoing reasons, I would find the appeal is unripe
and dismiss it for lack of jurisdiction.3
2
And contrary to my colleagues’ lack of confidence in judicial
alacrity, maj. op. at 15-16, courts make speedy decisions on injunction
applications in ANDA cases all the time. See, e.g., Apotex, Inc. v.
FDA, C.A. No. 06-627 (D.D.C. Apr. 19 2006); Biovail Corp. v. FDA,
C.A. No. 06-1487 (D.D.C. Aug. 25, 2006); Merck & Co. v. FDA, C.A.
No. 01-1343 (D.D.C. June 20, 2001).
3
In support of ripeness, the majority asserts: “District courts
routinely reach the merits of generic manufacturers’ claims to
exclusivity before the FDA has granted final approval to any ANDA
concerning the drug at issue.” Maj. op. at 15 (citing Teva Pharms.,
USA, Inc. v. Leavitt, 548 F.3d 103 (D.C. Cir. 2008); Ranbaxy Labs.,
Ltd. v. Leavitt, 469 F.3d 120 (D.C. Cir. 2006)). Leaving aside what
effect a court’s routine practice may have on an issue’s ripeness vel
non, I know of no instance where the district court reached the merits
of an ANDA before the FDA has issued any decision regarding the
plaintiff and the issue raised. In the two cases the majority cites, the
district court directly reviewed FDA decisions denying relief to the
plaintiffs. See Teva Pharms., 548 F.3d at 105 (reviewing denial of
citizen petition contesting FDA’s delisting of patent certified in its
ANDA); Ranbaxy Labs., Ltd., 469 F.3d at 121 (reviewing denial of
citizen petition denial). Here, by contrast, the FDA has taken no
adverse action whatsoever regarding the effect of delisting on Teva’s
ANDA—and apparently will not do so unless and until it denies final
approval. Had Teva raised the delisting issue before the FDA in the
first instance, its status here might be different.