In re: Wellbutrin XL Antitrust

Filing

PRECEDENTIAL OPINION Coram: JORDAN, VANASKIE and NYGAARD, Circuit Judges. Total Pages: 78. Authoring Judge: Jordan [15-2875, 15-3559, 15-3681, 15-3591, 15-3682] Opinion Previously filed as a Sealed Opinion on 08/09/17.

Case: 15-2875

Document: 003112704120

Page: 1

Date Filed: 08/17/2017

PRECEDENTIAL
UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
_____________
Nos. 15-2875/3559/3591/3681/3682
_____________
IN RE: WELLBUTRIN XL ANTITRUST LITIGATION
Indirect Purchaser Class,
Appellants in 15-2875
Aetna Health of California Inc.; IBEW-NECA Local 505
Health
and Welfare Plan; Bricklayers and Masons Union Local
Union
No. 5 Ohio Health and Welfare Fund; Mechanical
Contractors-United
Association Local 119 Health and Welfare Plan; Painters
District
Council No. 30 Health and Welfare Fund; Plumbers and
Pipefitters
Local 572 Health and Welfare Fund; Aetna, Inc.,
Appellants in 15-3559
Professional Drug Company, Inc., individually and on behalf
of the Direct Purchaser Class,
Appellant in 15-3591
SmithKline Beecham Corporation d/b/a
GlaxoSmithKline

Case: 15-2875

Document: 003112704120

Page: 2

Date Filed: 08/17/2017

and GlaxoSmithKline plc,
Appellants in 15-3681/3682
_______________
On Appeal from the United States District Court
for the Eastern District of Pennsylvania
(D.C. Nos. 2-08-cv-2431 and 2-08-cv-2433)
District Judge: Hon. Mary A. McLaughlin
_______________
Argued
September 7, 2016
Before: JORDAN, VANASKIE, and NYGAARD, Circuit
Judges.
(Filed: August 9, 2017)
_______________
Kenneth A. Wexler
Justin N. Boley
Wexler Wallace LLP
55 W. Monroe St. – Ste. 3300
Chicago, IL 60603

2

Case: 15-2875

Document: 003112704120

Page: 3

Date Filed: 08/17/2017

Peter D. St. Phillip, Jr. [ARGUED]
Richard W. Cohen
Gerald Lawrence, Jr.
Uriel Rabinovitz
Melissa Cabrera
Lowey Dannenberg Cohen & Hart, P.C.
One North Broadway – Ste. 509
White Plains, NY 10601
Karen Iannance
Kissel Hirsch & Wilmer
370 Lexington Avenue – Ste. 1200
New York, NY 10017
James G. Stranch, III
Joe P. Leniski
Branstetter, Stranch & Jennings, PLLC
227 Second Ave. North – 4th Fl.
Nashville, TN 37201
Counsel for Appellant Indirect Purchaser Class/
End-Payor Class
David F. Sorensen [ARGUED]
Andrew C. Curley
Caitlin G. Coslett
Nicholas Urban
Berger & Montague, P.C.
1622 Locust Street
Philadelphia, PA 19103

3

Case: 15-2875

Document: 003112704120

Page: 4

Date Filed: 08/17/2017

Thomas M. Sobol [ARGUED]
Gregory T. Arnold
David S. Nalven
Kristen A. Johnson
Kristie A. LaSalle
Hagens Berman Sobol Shapiro LLP
55 Cambridge Parkway – Ste. 3-1
Cambridge, MA 021142
John W. Barrett
Barrett Law Group
400 Court Square North
P.O. Box 927
Lexington, MS 39095
Peter Kohn
Faruqi & Faruqi
101 Greenwood Avenue - #600
Jenkintown, PA 19046
Dianne M. Nast
NastLaw
1101 Market Street - #2801
Philadelphia, PA 19107
Counsel for Appellant Direct Purchaser Class

4

Case: 15-2875

Document: 003112704120

Leslie E. John [ARGUED]
Edward D. Rogers
Stephen J. Kastenberg [ARGUED]
Jason A. Leckerman
Jessica M. Anthony
Marcel S. Pratt
Ballard Spahr
1735 Market Street – 51st Fl.
Philadelphia, PA 19103
Taimarie N. Adams
640 N. Broad Street - #529
Philadelphia, PA 19130
Daniel J. Boland
217 Ryers Avenue
Philadelphia, PA 19102
Timothy K. Gilman
Kirkland & Ellis
601 Lexington Avenue
New York, NY 10022
Emily P. Hughes
Jason R. Parish
Kirkland & Ellis
655 15th Street, NW – Ste. 1200
Washington, DC 20005
Simeon G. Papacostas
Kirkland & Ellis

5

Page: 5

Date Filed: 08/17/2017

Case: 15-2875

Document: 003112704120

Page: 6

Date Filed: 08/17/2017

300 N. LaSalle Street - #2400
Chicago, IL 65054
Susanna R. Greenberg
University of Pennsylvania School of Law
3400 Chestnut Street
Philadelphia, PA 19104
Chong S. Park
Ropes & Gray
2009 Pennsylvania Avenue, NW
Washington, DC 20006
Counsel for Appellee SmithKline Beecham Corp,
dba GlaxoSmithKline; GlaxoSmithKline PLC
Linda E. Kelly
Quentin Riegel
Leland P. Frost
Manufacturers’ Center for Legal Action
733 10th Street, NW – Ste. 700
Washington, DC 20001
Brian H. Pandya
Wiley Rein
1776 K Street NW
Washington, DC 20006
Counsel for Amicus, National Association of
Manufacturers

6

Case: 15-2875

Document: 003112704120

Page: 7

Date Filed: 08/17/2017

Deborah L. Feinstein
Markus H. Meier
Bradley S. Alberg
Daniel W. Butrymowicz
Elizabeth R. Hilder
Jamis R. Towey
David C. Shonka
Joel Marcus
Mark S. Hegedus
Federal Trade Commission
MS-582
600 Pennsylvania Avenue, NW
Washington, DC 20580
Counsel for Non Party Amicus, Federal
Trade Commission
Barbara W. Mather
Robin P. Sumner
Lindsay D. Breedlove
Pepper Hamilton
3000 Two Logan Square
Eighteenth & Arch Streets
Philadelphia, PA 19103
Counsel for Amicus, Antitrust Economists
Sarah K. Frederick
Goodwin Procter LLP
Exchange Place
53 State Street
Boston, MA 02109
Counsel for Amicus, Generic Pharmaceutical

7

Case: 15-2875

Document: 003112704120

Page: 8

Date Filed: 08/17/2017

Association
David W. Ogden
Wilmer Cutler Pickering Hale and Dorr LLP
1875 Pennsylvania Avenue, NW
Washington, DC 20006
Mark A. Ford
Peter A. Spaeth
Daniel C. Wewers
Wilmer Cutler Pickering Hale and Dorr LLP
60 State Street
Boston, MA 02109
Counsel for Amicus, Pharmaceutical Research and
Manufacturers of America
Daniel G. Brown
Latham & Watkins
885 Third Avenue – Ste. 1000
New York, NY 10022
Counsel for Par Pharmaceutical Inc., and its
predecessor in interest Anchen
Pharmaceuticals
_______________
OPINION OF THE COURT
_______________

8

Case: 15-2875

Document: 003112704120

Page: 9

Date Filed: 08/17/2017

Table of Contents
Background ................................................................... 14

I.
A.

The Hatch-Waxman Act ..................................... 14

B.

Factual and Procedural Background ................... 18

II.

Discussion ..................................................................... 24
A.

Sham Litigation .................................................. 24
1.

Applicable Law ................................................... 25

2.

The Anchen Lawsuit ........................................... 29

3.

The Abrika Lawsuit ............................................ 36

4.

The Impax and Watson Lawsuits and the
Appellants’ Conspiracy Theory .......................... 37

5.

The FDA Citizen Petition ................................... 40

6.

Serial Petitioning ................................................ 46

B.

Reverse Payment Settlement Agreement ........... 49
1.

Events Leading to the Settlement ....................... 49

2.

The Settlement .................................................... 51

3.

The Appellants Cannot Prevail on Their Antitrust
Claims Pertaining to the Alleged Reverse
Payment .............................................................. 54
a) The Agreements Are Not Immune from Antitrust
Scrutiny; the Rule of Reason Test Applies ......... 55
b) The Appellants Do Not Have Antitrust
Standing .............................................................. 61
(1)

License-Based Scenario ............................. 68

9

Case: 15-2875

Document: 003112704120

(2)
C.
III.

Page: 10

Date Filed: 08/17/2017

Litigation-Based Scenario .......................... 71

Class Certification, Daubert, and Intervention
Issues................................................................... 77
Conclusion .......................................................... 78

10

Case: 15-2875

Document: 003112704120

Page: 11

Date Filed: 08/17/2017

JORDAN, Circuit Judge.
This appeal lies at the confluence of intellectual
property and antitrust law. Following the Supreme Court’s
decision in FTC v. Actavis, Inc., 133 S. Ct. 2223 (2013), we
are tasked with balancing a patent owner’s right to exclude
and the public’s right to benefit from fair and open
competition.
The Appellants in this case are the direct and indirect
purchasers of Wellbutrin XL, a drug designed to treat
depression.
(Consolidated Brief of Appellees/CrossAppellants (“Ans. Br.”) 6, 19.)
The direct-purchaser
Appellants bring claims under federal antitrust law, alleging
that the Appellee, GlaxoSmithKline (“GSK”), 1 violated
Sections One and Two of the Sherman Antitrust Act by
entering into an unlawful conspiracy with a company called
Biovail, 2 GSK’s partner in the development of Wellbutrin
XL, to delay the launch of generic versions of the drug.
(Consolidated Brief of Direct Purchaser and End-Payor Class
1

“GlaxoSmithKline,” or “GSK,” refers collectively to
SmithKline Beecham Corporation and GlaxoSmithKline
PLC, the producers and distributors of Wellbutrin XL.
2

“Biovail” refers collectively to Biovail Corporation
(n/k/a Valeant Pharmaceuticals International, Inc.) and
Biovail Laboratories International SRL (n/k/a Valeant
International Bermuda). Biovail was originally a defendant in
the case but settled with the Appellants prior to the appeal.

11

Case: 15-2875

Document: 003112704120

Page: 12

Date Filed: 08/17/2017

Plaintiffs-Appellants (“Op. Br.”) 2; JA 11465-68.) The
indirect-purchaser Appellants assert similar theories, but
under state, rather than federal law. They also allege that
GSK’s actions violated common law principles and state
statutes mandating fair trade practices.
According to the Appellants, GSK is liable under two
theories. First, the Appellants claim that GSK delayed the
launch of generic versions of Wellbutrin XL by supporting
baseless patent infringement suits and a baseless FDA Citizen
Petition aimed at generic drug companies. Second, they
claim that GSK delayed the launch of those generic drugs by
entering into an unlawful reverse payment settlement
agreement with its potential competitors. 3 The District Court
granted summary judgment on the merits to GSK with respect
to both of those theories. It concluded that there was
insufficient evidence that GSK’s patent litigation was a sham
or that the settlement delayed the launch of generic versions
of Wellbutrin XL. At the same time, the Court granted
GSK’s Daubert motion to exclude the testimony of the
Appellants’ economic expert. The Court also granted a
motion to decertify the indirect-purchaser class for lack of
3

Ordinarily, when a plaintiff sues a defendant, one
expects that, if there is a settlement, it will involve a payment
from the defendant to the plaintiff. A so-called “reverse
payment” settlement takes place when the plaintiff settles the
case by paying the defendant. See FTC v. Actavis, Inc., 133
S. Ct. 2223, 2227 (2013) (“Because the settlement requires
the patentee to pay the alleged infringer, rather than the other
way around, this kind of settlement agreement is often called
a ‘reverse payment’ settlement agreement.”).

12

Case: 15-2875

Document: 003112704120

Page: 13

Date Filed: 08/17/2017

ascertainability and dismissed the indirect-purchaser claims
brought under the laws of any state that was not the home of a
named class representative. 4 Finally, the Court denied a
motion filed by Aetna, Inc. to intervene on the side of the
indirect purchasers. 5
4

The District Court had certified both the directpurchaser and indirect-purchaser classes in August 2011. The
indirect purchasers allege antitrust claims under the laws of
Arizona, California, the District of Columbia, Florida,
Hawaii, Iowa, Kansas, Louisiana, Maine, Michigan,
Minnesota, Mississippi, Nebraska, Nevada, New Mexico,
North Carolina, North Dakota, South Dakota, Tennessee,
Utah, Vermont, West Virginia and Wisconsin. They assert
violations of consumer protection laws in Alaska, Arizona,
Arkansas, California, Colorado, Connecticut, Delaware,
Florida, Georgia, Hawaii, Idaho, Illinois, Kansas, Louisiana,
Maine, Maryland, Massachusetts, Michigan, Minnesota,
Missouri, Montana, Nebraska, Nevada, New Hampshire, New
Mexico, New York, North Carolina, North Dakota, Ohio,
Oklahoma, Oregon, Pennsylvania, Rhode Island, South
Carolina, South Dakota, Texas, Utah, Vermont, Washington
and West Virginia, as well as the District of Columbia. The
District Court concluded that the indirect-purchaser
Appellants only have standing to bring their claims under the
laws of states where their members reside – that is,
California, Florida, Illinois, Nevada, New York, Ohio,
Pennsylvania, Tennessee, Texas, and Wisconsin.
5

Aetna, an insurance provider, had purchased brand
and generic versions of Wellbutrin XL in all 50 states. It
sought to intervene in order to save the indirect purchasers’
claims from dismissal.

13

Case: 15-2875

Document: 003112704120

Page: 14

Date Filed: 08/17/2017

This appeal followed. Both the direct-purchaser and
indirect-purchaser Appellants seek review of the District
Court’s summary judgment and Daubert rulings. The
indirect-purchaser Appellants also contest the order
decertifying their class and the denial of Aetna’s motion to
intervene. GSK filed a conditional cross-appeal challenging
on numerosity grounds the certification of the directpurchaser class. GSK filed a second conditional cross-appeal
with respect to the indirect-purchaser class, asking that, if we
were to disagree with the District Court’s decertification on
ascertainability grounds, we nevertheless affirm on
numerosity grounds. The direct-purchaser and indirectpurchaser Appellants filed a joint brief addressing the
summary judgment orders and the order denying Aetna’s
intervention; the indirect-purchaser Appellants also filed a
separate brief addressing the decertification order.
We agree with the District Court’s conclusions that the
Appellants have failed to establish a genuine dispute of fact
either as to whether GSK engaged in sham litigation or
whether GSK’s actions delayed the launch of any generic
version of Wellbutrin XL. Consequently, we will affirm the
District Court’s grant of summary judgment and do not reach
the remaining issues on appeal.
I.

Background
A.

The Hatch-Waxman Act

14

Case: 15-2875

Document: 003112704120

Page: 15

Date Filed: 08/17/2017

To better explain the antitrust issues in this case, we
first describe the regulatory scheme that governs the testing
and approval of new drugs in the United States. That
framework was established by the Drug Price Competition
and Patent Term Restoration Act of 1984, 98 Stat. 1585, as
amended, which is commonly known as the Hatch-Waxman
Act (“the Act”), or simply Hatch-Waxman. Actavis, 133 S.
Ct. at 2227-28.
A drug manufacturer seeking to market a new drug
“must submit a New Drug Application [(NDA)] to the federal
Food and Drug Administration (FDA) ... and undergo a long,
comprehensive, and costly testing process, after which, if
successful, the manufacturer will receive marketing approval
from the FDA.” Id. at 2228 (citing 21 U.S.C. § 355(b)(1)).
One of the goals of Hatch-Waxman is to increase competition
between generic and brand-name drugs. To that end, the Act
allows the manufacturers of generic drugs to obtain FDA
approval without having to endure the gauntlet of procedures
associated with NDAs.
[O]nce the FDA has approved a brand-name
drug ... a manufacturer of a generic drug can
obtain similar marketing approval through the
use of abbreviated procedures. The [Act]
permits a generic manufacturer to file an
Abbreviated New Drug Application [(ANDA)]
specifying that the generic has the “same active
ingredients as,” and is “biologically equivalent”
to, the already-approved brand-name drug. ...
[B]y allowing the generic to piggy-back on the
pioneer’s approval efforts, [the Act] “speed[s]

15

Case: 15-2875

Document: 003112704120

Page: 16

Date Filed: 08/17/2017

the introduction of low-cost generic drugs to
market,” thereby furthering drug competition.
Id. (last alteration in original) (internal citations omitted)
(quoting Caraco Pharm. Labs., Ltd. v. Novo Nordisk A/S, 566
U.S. 399, 405 (2012)).
In addition to streamlining the drug approval process,
the Hatch-Waxman Act provides specialized procedures for
brand-name and generic drug manufacturers to resolve
intellectual property disputes. The Act “requires the ... brandname manufacturer to list in its [NDA] the number and the
expiration date of any relevant patent. And it requires the
generic manufacturer in its [ANDA] to assure the FDA that
the generic will not infringe the brand-name’s patents.” Id.
(internal quotation marks and citations omitted). One way for
generic manufacturers to make that assurance is to “certify
that any listed, relevant patent ‘is invalid or will not be
infringed by the manufacture, use, or sale’ of the drug
described in the [ANDA].”
Id. (quoting 21 U.S.C.
§ 355(j)(2)(A)(vii)). That assurance is referred to as a
paragraph IV certification. Id. To facilitate the filing of
infringement suits, a paragraph IV certification
“automatically counts as patent infringement.” 6 Id. (citation
6

The “infringement” in those circumstances is a legal
construct that permits a patent holder to initiate suit without
having to wait for the generic manufacturer to actually make,
use, or sell a generic version of the patented drug. Eli Lilly &
Co. v. Medtronic, Inc., 496 U.S. 661, 678 (1990) (explaining
that “the defined act of infringement [is] artificial” and exists
to “enable the judicial adjudication upon which the ANDA ...
scheme[] depend[s]”). Because a paragraph IV certification

16

Case: 15-2875

Document: 003112704120

Page: 17

Date Filed: 08/17/2017

omitted). Hatch-Waxman further states that “[i]t shall be an
act of infringement to submit an [ANDA] for a drug claimed
in a patent[.]” 35 U.S.C. § 271(e)(2)(A).
The Act also encourages brand-name manufacturers to
file patent infringement suits quickly. If a patentee files an
infringement suit against a generic manufacturer within 45
days of receiving notice of the filing of a paragraph IV
certification, the patentee is rewarded with some breathing
space before competition can begin: the FDA is required to
withhold approval of the generic drug for 30 months or until
the infringement case is resolved, whichever comes first. 21
U.S.C. § 355(j)(5)(B)(iii).
Finally, the Act “provides a special incentive for a
generic to be the first to file an [ANDA] taking the paragraph
IV route.” Actavis, 133 S. Ct. at 2228-29. The first applicant
is entitled to an exclusivity period during which no generic
drug other than the first-filer’s can compete with the brandname drug. More specifically, the Act prohibits the FDA
from approving any ANDA other than the one first filed until
180 days after the first-filer starts marketing its drug.
21 U.S.C. § 355(j)(5)(B)(iv). In effect, that allows the firstfiling generic to exclude other generics from the market for
is defined as a technical act of infringement, it allows a patent
owner to sue, but it does not speak to whether the disclosed
generic drug does, in fact, infringe the cited patent. Glaxo,
Inc. v. Novopharm, Ltd., 110 F.3d 1562, 1569 (Fed. Cir.
1997) (“The occurrence of the defined ‘act of infringement’
does not determine the ultimate question whether what will be
sold will infringe any relevant patent.”).

17

Case: 15-2875

Document: 003112704120

Page: 18

Date Filed: 08/17/2017

longer than 180 days because it may delay or decline to
launch its drug. 7
B.
Factual and Procedural Background
In 1985, GSK obtained FDA approval for bupropion
hydrochloride, a drug for the treatment of major depressive
disorders. The drug became branded as “Wellbutrin.” Over
the years, several companies, including GSK, sought to
develop an extended release formulation of bupropion
hydrochloride. While GSK was unsuccessful, at least two
companies – Biovail and Andrx Pharmaceuticals, LLC –
found success and obtained patents covering extended release
7

While the exclusivity period can be forfeited if the
first-filer fails to market its drug, it can take time to trigger
the forfeiture. “Forfeiture applies only upon the satisfaction
of two statutory conditions. The first condition [(i.e., the
failure to market the drug either within 75 days of the date the
ANDA was approved or within 30 months of the date the
ANDA was submitted, whichever comes earlier)] is relatively
easy to satisfy. The second is triggered only if an appeals
court rules that the relevant patents are invalid or not
infringed, or if a settlement reaches a similar result.” C. Scott
Hemphill, An Aggregate Approach to Antitrust: Using New
Data and Rulemaking to Preserve Drug Competition, 109
Colum. L. Rev. 629, 660-61 (2009) (footnotes omitted). That
rule “allows first-filers to retain their exclusivity by settling.”
Chika Seidel, Comment, Settlement Should be the End of
Story: A Proposed Procedure to Settle Hatch-Waxman
Paragraph IV Litigations Modeled After Rule 23 Class Action
Settlement Procedure, 46 Seton Hall L. Rev. 697, 706-07
(2016) (footnote omitted).

18

Case: 15-2875

Document: 003112704120

Page: 19

Date Filed: 08/17/2017

formulations of the drug. 8 To gain access to an extended
release formulation, GSK obtained an exclusive license to
certain of Biovail’s patents. Then, in August 2002, GSK filed
an NDA for that new formulation, which was approved the
following year. The extended release Wellbutrin was named
“Wellbutrin XL.”
Between September 2004 and May 2005, four generic
manufacturers filed ANDAs seeking approval to market
generic versions of Wellbutrin XL. Each of the four
companies – Anchen, Abrika, Impax, and Watson – filed a
paragraph IV certification. 9 Of those companies, Anchen was
the first to file its ANDA, and, as a result, was entitled to the
180-day period of exclusivity.
Biovail filed patent infringement suits against all four
generic companies. With one exception, it filed its several
suits within 45 days of receiving each of the would-be
8

Biovail obtained U.S. Patents 6,096,341 and
6,143,327 (the ’341 and ’327 patents), while Andrx obtained
U.S. Patent 6,905,708 (the ’708 patent). At the time that
Biovail and Andrx were conducting their research, it appears
that most or all of the information relating to bupropion
hydrochloride and its delivery mechanisms was in the public
domain. Neither Biovail nor Andrx needed to obtain a license
from GSK in order to conduct its research.
9

The companies’ full names are Anchen
Pharmaceuticals, Inc., Abrika Pharmaceuticals, LLP, Impax
Laboratories, Inc., and Watson Pharmaceuticals, Inc.,
respectively.

19

Case: 15-2875

Document: 003112704120

Page: 20

Date Filed: 08/17/2017

competitors’ paragraph IV certifications. As explained
above, that triggered a stay that generally prevented the FDA
from approving the ANDAs for 30 months, or until the
resolution of the respective patent suits, whichever came first.
Biovail did not file suit within the required 45-day period
against Impax’s 300 mg dosage of extended release
bupropion hydrochloride. Impax was therefore not subject to
the 30-month stay with respect to that product. GSK joined
Biovail’s suits against Anchen and Abrika but not the suits
against Impax and Watson. 10
In addition to its lawsuits, Biovail filed a “Citizen
Petition” with the FDA on December 20, 2005. 11 Biovail
asked the FDA to impose certain requirements for approval of
any generic version of Wellbutrin XL. The FDA issued a
final response to the Petition in December 2006, granting it in
part and denying it in part.
On December 21, 2005, Andrx filed suit against GSK,
alleging that Wellbutrin XL, in 150 mg dosages, violated
Andrx’s ’708 patent, see supra n.8. Andrx also filed suit
10

GSK’s CEO, Jean-Pierre Garner, explained that
GSK elected not to join the suits against Impax or Watson
because it did not believe it would have sufficient control
over the litigation.
11

Interested citizens may “petition the Commissioner
[of the FDA] to issue, amend, or revoke a regulation or order,
or to take or refrain from taking any other form of
administrative action.” 21 C.F.R. § 10.25. GSK elected not
to join Biovail’s FDA petition.

20

Case: 15-2875

Document: 003112704120

Page: 21

Date Filed: 08/17/2017

against Anchen for infringing the same patent with a generic
version of Wellbutrin XL. In both cases, Andrx sought
damages and an injunction against the sale of infringing
products. In February 2007, all of the parties involved in the
Wellbutrin-related patent litigation, except for Abrika,
entered into a settlement. 12
The next year, in May 2008, this litigation began. Two
putative classes – a class of direct purchasers (e.g., entities
like pharmacies that purchased Wellbutrin XL directly from
GSK) and a class of indirect purchasers (e.g., consumers) –
filed suits against Biovail and GSK. 13 As noted at the outset
12

Abrika settled with Biovail a few months later, after
GSK had withdrawn from the suit.
13

The direct-purchaser class includes “[a]ll persons or
entities in the United States ... who purchased Wellbutrin XL
directly from any of the Defendants at any time during the
period November 14, 2005 through August 31, 2009... .” (JA
3.) In total, the class contains over 30 members. When
certified, the indirect-purchaser class included “[a]ll persons
or entities who purchased an [appropriately FDA- rated]
generic bioequivalent of Wellbutrin XL ... at any time
[between November 14, 2005 and April 29, 2011] in
California, Florida, Nevada, New York, Tennessee and
Wisconsin; and ... [a]ll entities that purchased 150 mg or
300 mg Wellbutrin XL before an [appropriately FDA-rated]
generic bioequivalent was available for such dosages AND
purchased generic XL in the same state after generic XL
became available in California, Florida, Nevada, New York,
Tennessee and Wisconsin.” (Appendix for 15-2875 at 6.)

21

Case: 15-2875

Document: 003112704120

Page: 22

Date Filed: 08/17/2017

of this opinion, both sets of plaintiffs alleged that Biovail and
GSK conspired to prevent generic versions of Wellbutrin XL
from entering the market. The instrumentalities of the alleged
conspiracy were, according to the Plaintiffs, sham lawsuits, a
sham FDA petition, and an unlawful reverse payment
settlement. The direct purchasers brought their claims under
federal law, while the indirect purchasers brought their claims
under various state laws. Biovail settled with both classes in
November 2012, so only GSK has remained as a defendant. 14
The District Court had earlier granted summary
judgment for GSK on the merits on all of the claims. First,
the Court granted summary judgment on the sham petition
claims. Shortly after that, it stayed both the direct-purchaser
and indirect-purchaser cases while the Supreme Court
considered potentially relevant petitions for writs of
certiorari.
The District Court continued the stay in
anticipation of the Supreme Court’s decision in FTC v.
Actavis, Inc., 133 S. Ct. 2223 (2013). After the Actavis
opinion issued, the District Court granted summary judgment
for GSK on the reverse payment claims.
The District Court also rendered decisions regarding
class certification. It at first certified both putative classes.
Later, however, it concluded that the indirect-purchaser class
only had standing “under the laws of those states where the
plaintiffs are located or their members reside.” (JA 243.)
The Court thus dismissed the claims arising under the laws of
states that were not represented by one of the named
plaintiffs. In response, Aetna moved to intervene in the
14

The two cases proceeded independently in the
District Court, but were consolidated on appeal.

22

Case: 15-2875

Document: 003112704120

Page: 23

Date Filed: 08/17/2017

indirect-purchaser suit. It alleged that it had purchased brand
and generic Wellbutrin XL in all fifty states, and that,
consequently, its intervention would alleviate the standing
issues. The Court denied Aetna’s motion. In June 2015, the
Court granted a motion to decertify the indirect-purchaser
class on ascertainability grounds.
Before us on appeal are the following rulings: the grant
of summary judgment to GSK on all claims, the exclusion of
the testimony of the Appellants’ economic expert, the
decertification of the indirect-purchaser class, the dismissal of
certain of the indirect-purchaser Appellants’ claims for lack
of standing, and the denial of Aetna’s motion to intervene.
GSK also conditionally challenges the certification of the
direct-purchaser class. And, should the indirect purchasers
succeed in overcoming the ascertainability objection to
certification of their class, GSK also conditionally appeals
any certification of that class, citing problems with
numerosity.

23

Case: 15-2875

II.

Document: 003112704120

Page: 24

Date Filed: 08/17/2017

Discussion 15
A.

Sham Litigation

The first broad issue on appeal is whether the District
Court erred in granting summary judgment on the sham
litigation claims. The Appellants argue that GSK violated
antitrust laws by conspiring with Biovail to prosecute sham
lawsuits against Anchen, Abrika, Impax, and Watson, and to
file a sham petition with the FDA. According to the
Appellants, GSK and Biovail worked together to press the
infringement lawsuits in order to exploit the mandatory 30month stay created by the Hatch-Waxman Act.
The
Appellants also allege that, but for the lawsuits and the FDA
petition, the FDA would have approved Anchen’s ANDA
immediately and likewise would have approved the other
three ANDAs at the end of Anchen’s 180-day exclusivity
15

The District Court had subject matter jurisdiction
over the direct purchasers’ claims under 15 U.S.C. § 15(a)
and 28 U.S.C. §§ 1331, 1332, and 1337. The District Court
had subject matter jurisdiction over the indirect purchasers’
claims under 28 U.S.C § 1332(d)(2). We have jurisdiction
over the appeal pursuant to 28 U.S.C. § 1291.
“We exercise plenary review over a district court’s
order granting summary judgment, applying the same
standard as the district court. We will affirm only if drawing
all reasonable inferences in favor of the nonmoving party,
there is no genuine issue as to any material fact and the
moving party is entitled to judgment as a matter of law.”
Young v. Martin, 801 F.3d 172, 177 (3d Cir. 2015) (internal
quotation marks, citations, and modifications omitted).

24

Case: 15-2875

Document: 003112704120

Page: 25

Date Filed: 08/17/2017

period. The assertion is that, without the delay in ANDA
approvals, Anchen and the other generics would have
launched their products sooner, resulting in increased
competition and lower drug prices for pharmacies and
consumers.
1.

Applicable Law

A plaintiff claiming that a lawsuit is, by its very
existence, anticompetitive and unlawful faces an uphill battle.
It is well-established that the First Amendment protects the
right to petition the government and to have access to the
courts. Prof’l Real Estate Inv’rs, Inc. v. Columbia Pictures
Indus., Inc., 508 U.S. 49, 56-57 (1993); 16 Cal. Motor Transp.
Co. v. Trucking Unlimited, 404 U.S. 508, 515 (1972); see also
U.S. Const. amend. I (“Congress shall make no law ...
abridging ... the right of the people ... to petition the
Government for a redress of grievances.”). That protection is
the basis of the Noerr-Pennington doctrine, which holds that
“[t]hose who petition [the] government for redress are
generally immune from antitrust liability.” 17 PRE, 508 U.S.
16

We will refer to Professional Real Estate Investors
v. Columbia Pictures Industries, 508 U.S. 49 (1993),
throughout this opinion as “PRE.”
17

The Noerr-Pennington doctrine “takes its name from
Eastern R.R. Presidents Conference v. Noerr Motor Freight,
Inc., 365 U.S. 127 (1961) (holding that railroads’ publicity
campaign to promote support for laws harmful to trucking
interest was immune from antitrust liability), and United Mine
Workers of America v. Pennington, 381 U.S. 657 (1965)
(joint efforts by miners’ union and large coal companies to

25

Case: 15-2875

Document: 003112704120

Page: 26

Date Filed: 08/17/2017

at 56. Noerr-Pennington immunity, however, is not absolute.
“[A]ctivity ‘ostensibly directed toward influencing
governmental action’ does not qualify for [first amendment]
immunity if it ‘is a mere sham to cover ... an attempt to
interfere directly with the business relationships of a
competitor.’” Id. at 51 (third alteration in original) (quoting
E. R.R. Presidents Conference v. Noerr Motor Freight, Inc.,
365 U.S. 127, 144 (1961)).
To determine whether a lawsuit qualifies as a “sham,”
courts apply a two-part test:
First, the lawsuit must be objectively baseless in
the sense that no reasonable litigant could
realistically expect success on the merits. If an
objective litigant could conclude that the suit is
reasonably calculated to elicit a favorable
outcome, the suit is immunized under Noerr,
and an antitrust claim premised on the sham
exception must fail.
Only if challenged
litigation is objectively meritless may a court
examine the litigant’s subjective motivation.
Under this second part ..., the court should
focus on whether the baseless lawsuit conceals
an attempt to interfere directly with the business
relationships of a competitor through the use of
the governmental process—as opposed to the
outcome of that process—as an anticompetitive
have federal agency impose higher minimum wage for coal
suppliers to TVA were immune from antitrust liability).”
Mercatus Grp., LLC v. Lake Forest Hosp., 641 F.3d 834, 841
(7th Cir. 2011).

26

Case: 15-2875

Document: 003112704120

Page: 27

Date Filed: 08/17/2017

weapon. This two-tiered process requires the
plaintiff to disprove the challenged lawsuit’s
legal viability before the court will entertain
evidence of the suit’s economic viability.
PRE, 508 U.S. at 60-61 (internal quotation marks, citations,
alteration, and footnote omitted). 18
The Supreme Court has explained that “[t]he existence
of probable cause to institute legal proceedings precludes a
finding that an antitrust defendant has engaged in sham
litigation.” Id. at 62. In selecting “probable cause” as the
standard by which to judge objective baselessness, the Court
said that it was drawing from “[t]he notion of probable cause,
as understood and applied in the common law tort of
wrongful civil proceedings[.]” Id. A litigant has probable
cause to initiate a suit if the litigant has “a reasonable belief
that there is a chance that a claim may be held valid upon
adjudication.” Id. at 62-63 (internal citations, quotation, and
alterations omitted); see also Restatement (Second) of Torts
§ 675. In other words, the essential question is not whether
18

The Supreme Court in PRE indicated that the
plaintiff in an antitrust suit has the burden of proving that the
defendant is not entitled to immunity under the NoerrPennington doctrine. See PRE, 508 U.S. at 61 (explaining
that a plaintiff must “demonstrat[e] both the objective and the
subjective components of a sham”). The Court was silent,
however, as to the standard of proof (i.e., clear and
convincing evidence, or preponderance of the evidence)
needed to show objective baselessness. Because our decision
in this case does not hinge on the standard of proof, we leave
that question for another day.

27

Case: 15-2875

Document: 003112704120

Page: 28

Date Filed: 08/17/2017

the suit succeeds, but whether the suit was a sham at the time
it was filed. See PRE, 508 U.S. at 60 n.5 (cautioning that
“when the antitrust defendant has lost the underlying
litigation, a court must resist the ... temptation to engage in
post hoc reasoning by concluding that an ultimately
unsuccessful action must have been unreasonable or without
foundation” (internal quotations omitted)).
In addition, it is not enough for a plaintiff to show that
a defendant engaged in sham litigation. “[A] plaintiff who
defeats the defendant’s claim to Noerr immunity ... must still
prove a substantive antitrust violation.” Id. at 61. That
includes proving the challenged lawsuit is “causally linked”
to an antitrust injury. Brunswick Corp. v. Pueblo Bowl-OMat, Inc., 429 U.S. 477, 489 (1977) (describing antitrust
injury as “injury of the type the antitrust laws were intended
to prevent and that flows from that which makes defendants’
acts unlawful”).
As noted earlier, the Appellants argue that each of the
patent infringement suits relating to generic versions of
Wellbutrin XL (that is, each of the suits against Anchen,
Abrika, Watson, and Impax), as well as the Citizen Petition,
was an instance of anticompetitive sham litigation or sham
petitioning that caused antitrust injury by delaying the entry
of generic versions of Wellbutrin XL into the market. The
District Court granted summary judgment to GSK with
respect to each of the five challenged actions. We agree that
the sham litigation claims fail, for reasons we now endeavor
to explain.

28

Case: 15-2875

Document: 003112704120

2.

Page: 29

Date Filed: 08/17/2017

The Anchen Lawsuit

The sham litigation claim relating to the Anchen suit
fails for the simple reason that an act of infringement plainly
occurred. The already high hurdle for stating an antitrust
claim for anticompetitive litigation, PRE, 508 U.S. at 56, is
higher still in the context of an ANDA case because, as
described above, the Hatch-Waxman Act states that “[i]t shall
be an act of infringement to submit” an ANDA for a drug
claimed in a patent, 35 U.S.C. § 271(e)(2). Since the
submission of an ANDA is, by statutory definition, an
infringing act, an infringement suit filed in response to an
ANDA with a paragraph IV certification could only be
objectively baseless if no reasonable person could disagree
with the assertions of noninfringement or invalidity in the
certification. See AstraZeneca AB v. Mylan Labs., Inc., No.
00-cv-6749, 2010 WL 2079722, at *4 (S.D.N.Y. May 19,
2010) (“[A]t the outset of Astra’s case, Mylan gave Astra an
objectively reasonable basis to sue: Mylan provided Astra
notice of its Paragraph IV certification.”), aff’d sub nom. In re
Omeprazole Patent Litig., 412 Fed. App’x 297 (Fed Cir.
2011). It suffices here to say that this case does not present
such a circumstance. Anchen filed an ANDA for a drug that
was claimed in Biovail’s ’341 patent. There is nothing in the
record indicating that Biovail, the patentee, and GSK, the
exclusive licensee, 19 were less than objectively reasonable in

19

“Because the legally protected interests in a patent
are exclusionary rights created by the Patent Act, a party
holding one or more of those exclusionary rights—such as an
exclusive licensee—suffers a legally cognizable injury when
an unauthorized party encroaches upon those rights and

29

Case: 15-2875

Document: 003112704120

Page: 30

Date Filed: 08/17/2017

acting on that technical act of infringement, and that alone
provides a sufficient basis for us to affirm the District Court’s
grant of summary judgment.
The content of the ANDA bolsters that conclusion. As
explained above, Wellbutrin XL used a formulation of
bupropion hydrochloride described in Biovail’s ’341 patent.
That patent discloses, among other things, “a core comprising
bupropion hydrochloride and conventional excipients, free of
stabilizer” (JA 3117, ’341 patent at 9:50-51 (emphasis
added).) Anchen’s paragraph IV certification asserted that
“Anchen’s proposed product cannot be deemed literally to
infringe [the patent] because it includes a stabilizing amount
of hydrochloric acid in the core[,]” and thus does not satisfy
the “free of stabilizer” limitation. (JA 35714); see also
Pfizer, Inc. v. Teva Pharm., USA, Inc., 429 F.3d 1364, 1376
(Fed. Cir. 2005) (explaining that a product must satisfy each
of a claim’s limitations in order to infringe). But Anchen’s
ANDA suggested otherwise. It included several tables listing
the ingredients that would be present in Anchen’s drug, along
with the relative percentages of each. While the tables listed
hydrochloric acid, which can serve as a stabilizer, 20 they
indicated that the acid would compose 0% of the final
product. To the same effect, the percentages associated with
the other listed ingredients summed to 100%. As if to
emphasize that point, the tables explicitly stated that the
therefore has standing to sue.” WiAV Solutions LLC v.
Motorola, Inc., 631 F.3d 1257, 1264-65 (Fed. Cir. 2010).
20

Hydrochloric acid is a stabilizing agent. (See
Anchen’s paragraph IV certification, JA 35714 (referring to
hydrochloric acid as a stabilizing agent).)

30

Case: 15-2875

Document: 003112704120

Page: 31

Date Filed: 08/17/2017

hydrochloric acid had been “removed” or “evaporated” from
the drug. (JA 11748-52.) That language provided GSK and
Biovail with sufficient probable cause to file its infringement
suit, and no reasonable jury – i.e., no jury considering the
sham litigation claim – could conclude otherwise. See Abbott
Labs. v. TorPharm, Inc., 300 F.3d 1367, 1373 (Fed. Cir.
2002) (holding that the ANDA specification governs the
infringement inquiry).
The parties and the District Court invested
considerable effort in addressing two subsidiary questions –
whether FDA regulations required Anchen to quantify the
amount of stabilizer present in its drug, and whether Biovail
asserted a frivolous claim construction position. Those
disputes are ultimately irrelevant. The question here is
whether GSK and Biovail could have perceived “some
likelihood of success” in their case at the time of filing. PRE,
508 U.S. at 65; Rohm & Haas Co. v. Brotech Corp., 127 F.3d
1089, 1093 (Fed. Cir. 1997) (concluding that courts should
evaluate the question of objective baselessness “in light of ...
information [available] at the time of filing”). At that time,
the only information they had access to was an excerpt of
Anchen’s ANDA – an excerpt that, under a plain reading,
suggested the non-infringement theory offered in Anchen’s
paragraph IV certification was, or at least could be, infirm.
Viewed in that light, the FDA’s rules regarding quantification
are insufficient to override the probable cause provided in the
ANDA. Again, the fact that one might conclude, after a
thorough investigation, that Anchen’s ANDA did not
definitively exclude the presence of hydrochloric acid does
not mean it was unreasonable for GSK and Biovail to file
their suit, as it was not unreasonable for them to take the

31

Case: 15-2875

Document: 003112704120

Page: 32

Date Filed: 08/17/2017

ANDA at face value. 21 Similarly, the fact that a court (in the
underlying patent litigation) rejected Biovail’s later proposed
claim construction does not bear on whether the patent
infringement suit was objectively baseless from the outset. 22
21

As explained above, Anchen’s ANDA quantified the
amount of hydrochloric acid in its drug as “--” and indicated
that the other ingredients summed to 100%. The Appellants
argue that, under standard industry convention, “--” denotes a
residual quantity greater than zero. However, they do not
point to any evidence showing that GSK and Biovail’s
interpretation of either “--” or “100%” was unreasonable.
More to the point, they have not demonstrated that it was
unreasonable to view the claim language “free of stabilizer”
as covering a residual amount so small as to not register in the
tables they provided.
22

It is worth noting that GSK withdrew from the case
well before claim construction began. While it is no doubt
important to think about possible constructions for patent
claims before filing a case, it would be unfair to require
parties to divine the outcome of claim construction before
filing. That is especially true in the Hatch-Waxman context,
where many details about the potentially infringing drug
(details that could shape a plaintiff’s claim construction
position) cannot be known at the time a suit is filed and where
there are congressionally designed pressures to file suit
quickly. See 21 U.S.C. § 355(b)(3)(D) (stating that an ANDA
applicant is required to provide notice to patentees
“includ[ing] a detailed statement of the factual and legal basis
of the opinion of the applicant that the patent[s] [are] invalid
or will not be infringed,” but refraining from requiring
ANDA applicants to submit any additional information). The

32

Case: 15-2875

Document: 003112704120

Page: 33

Date Filed: 08/17/2017

See Rohm & Haas Co., 127 F.3d at 1092 (explaining that
“[p]atent litigation is complex, long, and difficult” and that
parties and courts rely on “discovery procedures, partial or
complete summary judgment, and evidentiary rules to narrow
the issues”).
The Appellants also argue that the District Court
“usurp[ed] the role of the jury” by “find[ing] facts.” (Op. Br.
53-54.) We disagree. In PRE, the Supreme Court held that
courts can grant summary judgment on the issue of objective
baselessness if “there is no dispute over the predicate facts of
the underlying legal proceeding.” 508 U.S. at 63. Here, the
predicate facts include the content of Anchen’s ANDA. The
existence of that content – as opposed to its accuracy – is not
in dispute. Instead, the parties disagree about whether that
content was sufficient to establish probable cause for the
time limits imposed by the Hatch-Waxman Act embody a
“file-now, discover-details-later” policy, and while the merit
of that policy may make for an interesting debate, Aaron S.
Kesselheim & Jonathan J. Darrow, Hatch-Waxman Turns 30:
Do We Need a Redesigned Approach for the Modern Era?, 15
Yale J. Health Pol’y, L. & Ethics 293 (2015) (describing the
debate over whether the Act is good policy), it is not our
place – nor was it GSK’s – to take that debate on. We decline
to apply the antitrust laws in a way that would undermine the
operation of Hatch-Waxman.
It is likewise a painful stretch to say that Biovail’s
claim construction arguments show GSK was wrong to join in
the initial decision to file against Anchen. At the time the suit
was filed, GSK was not obligated to know the details of claim
construction arguments that Biovail would later present.

33

Case: 15-2875

Document: 003112704120

Page: 34

Date Filed: 08/17/2017

objective baselessness inquiry. PRE, 508 U.S. at 62. That,
however, is a legal question, not a factual one. Highmark,
Inc. v. Allcare Health Mgmt. Sys., Inc., 701 F.3d 1351, 1353
(Fed. Cir. 2012) (“Under PRE, the reasonableness of a legal
position … is itself a question of law[.]”); Stewart v.
Sonneborn, 98 U.S. 187, 194 (1878) (“[P]robable cause is a
question of law in a very important sense. ... Whether the
circumstances alleged to show it probable are true, and
existed, is a matter of fact; but whether, supposing them to be
true, they amount to a probable cause, is a question of law.”).
In granting summary judgment, the District Court decided
that GSK’s suit “[did] not fit the profile of objectively
baseless sham litigation.” (JA 72, 95.) It was entitled to
reach that legal conclusion. 23
There is an additional problem with the Appellants’
argument that warrants discussion. As we noted earlier, to
establish an antitrust claim for anticompetitive litigation, the
Appellants had to show not only that GSK’s litigation was a
sham, but also that it caused an antitrust injury by delaying
generic competition. Based on the current record, they would
have difficulty making such a showing, for at least two
reasons. First, generic entry would have been blocked by
Biovail’s continuing litigation against Anchen, in which GSK
did not participate. Under Hatch-Waxman, the rule requiring
the FDA to delay approving an ANDA is based not simply on
the filing of a lawsuit but on the ongoing presence of a
lawsuit. See 21 U.S.C. § 355(j)(5)(B)(iii) (stating that the
FDA may approve an ANDA as soon as “the district court
23

That the Court elected at times to use the verb “find”
instead of “conclude” does not change our decision.

34

Case: 15-2875

Document: 003112704120

Page: 35

Date Filed: 08/17/2017

decides that the patent is invalid or not infringed”). GSK
withdrew from the underlying litigation just a few months
after the case was filed. Biovail, however, continued to
pursue the suit. That is significant, as it means that the delay
in competition based on the lawsuit should likely be attributed
to Biovail rather than to GSK. 24
Second, and perhaps more formidably, generic entry
would have been blocked by the ’708 patent owned by Andrx.
We address the Andrx patent in more detail in the context of
the reverse payment settlement.
The implications for
causation, however, apply just as much to the Appellants’
sham litigation claims as they do to their reverse payment
claims.

24

GSK and Biovail filed their suit against Anchen on
December 21, 2004. GSK withdrew from the case exactly
four months later, on April 21, 2005, long before the 30month stay expired. In order to prevail against GSK, the
Appellants must show that at least some delay can be
attributed to GSK’s actions in the case – that is, they must
show that at least some delay can be attributed to the first four
months of the litigation. There is no evidence in the record
indicating that any delay can be linked to that period of time.
In a heading in their statement of undisputed facts, the
Appellants state that “GSK Withdrew From the Anchen and
Abrika Cases But Not the Conspiracy[.]” It takes some
chutzpah to use that language, as GSK plainly disputes that it
was ever in a conspiracy with Biovail. We consider the
conspiracy argument in more detail below.

35

Case: 15-2875

Document: 003112704120

3.

Page: 36

Date Filed: 08/17/2017

The Abrika Lawsuit

The Appellants contend that GSK and Biovail’s suit
against Abrika was another instance of anticompetitive
litigation. As before, they argue that GSK and Biovail are not
entitled to Noerr-Pennington immunity because the lawsuit
was a sham. The District Court granted summary judgment
to GSK based on its conclusion that there was insufficient
evidence to show that the litigation delayed Abrika’s entry
into the market. Again, we agree.
As an initial matter, we note that two of the arguments
that defeated the Appellants’ claim relating to the Anchen
litigation also justify affirming the District Court on this
point. First, Abrika’s ANDA, including the paragraph IV
certification, provided GSK with an objectively reasonable
basis to file its suit. 25 Additionally, as in the Anchen case,
GSK initially joined with Biovail in the infringement suit, but
then withdrew, and Biovail continued to litigate. That means
that any delay attributable to the litigation would have existed
even without GSK’s involvement.
Moreover, there is an independent problem with the
causation theory as it relates to the Abrika litigation. The
Appellants argue that the infringement suit against Abrika
delayed Abrika’s entry into the market because the suit
25

Without parsing the Abrika ANDA in the same
detail as we did Anchen’s, it suffices to say that the
Appellants have not provided evidence to demonstrate that it
was objectively unreasonable for Biovail and GSK to act on
the technical act of infringement that the ANDA and
paragraph IV certification provided.

36

Case: 15-2875

Document: 003112704120

Page: 37

Date Filed: 08/17/2017

imposed a 30-month stay on the FDA’s approval of Abrika’s
ANDA. There is, however, no evidence that Abrika could
have launched even in the absence of the 30-month stay. To
the contrary, it is undisputed that the FDA could not have
approved Abrika’s ANDA until the end of Anchen’s 180-day
first-filer exclusivity period, a period that would not even
start until Anchen launched its drug. Thus, it should surprise
no one to learn that, while the 30-month stay imposed by
GSK’s suit expired on June 21, 2007, Abrika’s ANDA was
not approved until over a year later, after Anchen’s
exclusivity period came to an end. In responding to those
facts, the Appellants in their Reply Brief appear to abandon
their argument that it was the Abrika lawsuit that caused
delay. Instead, they argue that the delay was caused by the
suit against Anchen. That argument, however, is unavailing
for the reasons already stated. Because there is no evidence
showing that GSK’s lawsuit against Abrika actually delayed
Abrika’s entry into the market, the District Court rightly
rejected it as a basis of liability.
4.

The Impax and Watson Lawsuits and the
Appellants’ Conspiracy Theory

In contrast with the Anchen and Abrika lawsuits, GSK
never joined the infringement litigation against Impax and
Watson.
Biovail pursued those suits on its own.
Nevertheless, the Appellants argue that all of Biovail’s
Wellbutrin-related litigation was brought in furtherance of a
conspiracy with GSK. Once again, their arguments are
wanting.
To avoid an adverse summary judgment on an antitrust
conspiracy claim, a plaintiff must “present evidence ‘that

37

Case: 15-2875

Document: 003112704120

Page: 38

Date Filed: 08/17/2017

tends to exclude the possibility’ that the alleged conspirators
acted independently.” Matsushita Elec. Indus. Co., Ltd. v.
Zenith Radio Corp., 475 U.S. 574, 588 (1986) (quoting
Monsanto Co. v. Spray-Rite Serv. Corp., 465 U.S. 752, 764
(1984)). “[A] plaintiff must offer enough evidence that the
inference of conspiracy is reasonable in light of the
competing inferences of independent action ... .” Cosmetic
Gallery, Inc. v. Schoeneman Corp., 495 F.3d 46, 51 (3d Cir.
2007) (internal quotation marks and citation omitted). Mere
communication between alleged co-conspirators, without
more, is not sufficient to defeat the presumption of
independent action. See In re Baby Food Antitrust Litig., 166
F.3d 112, 133 (3d Cir. 1999) (concluding that courts reject
conspiracy claims that “seek to infer agreement from …
communications despite a lack of independent evidence
tending to show an agreement” (citation omitted) (alteration
in original)); Alvord-Polk, Inc. v. F. Schumacher & Co., 37
F.3d 996, 1014 (3d Cir. 1994) (“Plaintiffs … seek to infer an
agreement from those communications despite a lack of
independent evidence tending to show an agreement and in
the face of uncontradicted testimony that only informational
exchanges took place. Without more, they cannot do so.”).
The Appellants have markedly failed to offer
meaningful evidence that excludes the possibility that Biovail
acted independently. With respect to the suits against Impax
and Watson, the evidence that the Appellants do point to is a
“common interest agreement” between Biovail and GSK, 26 a
26

GSK and Biovail did not produce the common
interest agreement. Instead, the parties stipulated that “[u]pon
receiving [Anchen’s] Paragraph [IV] certification ... GSK and
Biovail reached a common interest agreement with respect to

38

Case: 15-2875

Document: 003112704120

Page: 39

Date Filed: 08/17/2017

communication between Biovail and GSK in which GSK
forwarded Impax’s paragraph IV certification, and an email
from Biovail to GSK’s outside counsel stating that Biovail
had not heard from GSK with respect to the Impax litigation
and that Biovail “[did] not want to let the 45-day clock expire
without consciously dealing with the issue.” 27 (JA 2347.)
That evidence is insufficient. Biovail was the patentee and
GSK the exclusive licensee on directly relevant technology,
so communication between them acknowledging a common
interest is hardly surprising and does not come close to
supporting an inference that there was an unlawful conspiracy
to stifle competition. Likewise, it is not surprising to see that
the companies exchanged e-mails relating to their shared
interest. The e-mails containing Impax’s paragraph IV
certification and acknowledging Hatch-Waxman’s 45-day
litigation window are the sort one would expect from two
companies that share an interest in a pharmaceutical patent.
That communication does not amount to a conspiracy to
engage in sham litigation. See In re Nexium (Esomeprazole)
Antitrust Litig., 842 F.3d 34, 56 (1st Cir. 2016) (explaining
that pharmaceutical companies have valid reasons for
their common legal interest in potential infringement of [the
’341 and ’327 patents] by Anchen or the filers of any
additional ANDAs and related Paragraph [IV] notices ... .”
(JA 11513.) GSK and Biovail further stipulated that the
common interest agreement “related to the Paragraph [IV]
certifications of [Anchen, Abrika, and Impax]” as well as the
Anchen and Abrika actions. (Id.)
27

The Appellants do not have any evidence regarding
communication between Biovail and GSK with respect to the
Watson suit.

39

Case: 15-2875

Document: 003112704120

Page: 40

Date Filed: 08/17/2017

communicating with each other, and concluding that evidence
of such communication, without more, is not enough to
establish an unlawful conspiracy).
The Appellants do not fare any better with respect to
their claim that the Impax and Watson suits were brought as
part of a larger conspiracy involving all four infringement
actions, the Citizen Petition filed with the FDA, and the
overall settlement agreement. Bare allegations cannot defeat
summary judgment, and the Appellants have not pointed to
any evidence to support their theory that there was a larger,
overarching conspiracy.
5.

The FDA Citizen Petition

Biovail (but not GSK) filed a Citizen Petition with the
FDA, expressing concern regarding the sufficiency of the
FDA’s bioequivalence criteria for generic versions of
Wellbutrin XL. 28 Biovail requested the FDA to require all
generic manufacturers of the drug to do the following four
things:
28

As already discussed, for a generic manufacturer to
obtain FDA approval of an ANDA, the manufacturer must
demonstrate that its drug is bioequivalent to a drug that went
through the rigorous NDA approval process. Actavis, 133 S.
Ct. at 2228. “Bioequivalence is the absence of a significant
difference in the rate and extent to which the active ingredient
or active moiety in pharmaceutical equivalents or
pharmaceutical alternatives becomes available at the site of
drug action when administered at the same molar dose under
similar conditions in an appropriately designed study.” 21
C.F.R. § 314.3 (2016).

40

Case: 15-2875

Document: 003112704120

Page: 41

Date Filed: 08/17/2017

(1) demonstrate that the generic formulation of
[the drug] is bioequivalent to Wellbutrin ...
Wellbutrin SR ... and Wellbutrin XL;
(2) calculate and evaluate parameters in all of
its bioequivalence trials based on concentrations
of the patent drug and active metabolites; ...
(3) conduct its bioequivalence trials at steadystate... [; and]
[(4)] ... provide in vitro data demonstrating the absence
of dose dumping if generic bupropion HCl extendedrelease tablets are consumed with alcohol.
(JA 37509.) Six months after receiving the Petition, the FDA
issued an interim response stating that the FDA “has been
unable to reach a decision on [the] petition because it raises
complex issues requiring extensive review and analysis by
Agency officials.” (JA 37507.)
Another six months passed and, on December 14,
2006, the FDA issued its final response to the Petition,
granting it in part, and denying it in part. The final response
came on the same day that the FDA approved Anchen’s
ANDA. The Appellants allege that Biovail’s Citizen Petition
was anticompetitive and unlawful because, again, it was filed
in furtherance of a conspiracy with GSK to delay generic
entry. The District Court rejected that contention, concluding
that the Appellants failed to “raise[] a genuine issue of
material fact as to whether the unsuccessful and allegedly
sham requests[(among the four requests Biovail made to the
FDA)] caused any delay beyond the non-sham requests[.]”

41

Case: 15-2875

Document: 003112704120

Page: 42

Date Filed: 08/17/2017

(JA 127.) The Court also concluded that the Appellants failed
to show that Biovail filed the Petition as part of a conspiracy
with GSK.
On appeal, the Appellants argue that each of the four
requests in the Petition was a sham, that there is no
requirement to show the extent to which the delay was caused
by sham requests (as opposed to meritorious requests), and
that there was evidence to show that the Petition was filed as
part of a conspiracy between Biovail and GSK. Those
arguments, though, are no more persuasive now than they
were before the District Court.
The most straightforward basis for affirmance is, once
more, that the Appellants have failed to identify evidence
showing that there was a conspiracy between Biovail and
GSK, in this instance relating to the FDA petition. Just as
with the sham litigation conspiracy claims, the Appellants
must “present evidence ‘that tends to exclude the possibility’
that the alleged conspirators acted independently[,]”
Matsushita, 475 U.S. at 588 (quoting Monsanto Co., 465 U.S.
at 764), and must “offer enough evidence that the inference of
conspiracy is reasonable in light of the competing inferences
of independent action … .” Cosmetic Gallery, 495 F.3d at 51
(internal quotation marks omitted). They have not met that
burden. There is no evidence in the record showing that
Biovail and GSK collaborated, let alone illegally conspired,
on Biovail’s Citizen Petition, and there is no evidence
showing that Biovail filed the Petition at the direction of GSK
or in furtherance of a plan involving GSK. On the contrary,
the record indicates that GSK was not aware of Biovail’s

42

Case: 15-2875

Document: 003112704120

Page: 43

Date Filed: 08/17/2017

specific plans to file a petition, 29 that neither GSK nor Biovail
wanted to collaborate on a petition, 30 that GSK refused to
share its data with Biovail for use in a petition, 31 and that
GSK disagreed with the general premise of Biovail’s
Petition. 32

29

Eight days before Biovail filed its Petition, GSK
employees sent an internal email expressing uncertainty as to
whether Biovail had filed a petition.
30

GSK sent a fax to Biovail indicating that it “[did]
not wish to participate in or be associated with ... Biovail
explorations, deliberations, strategizing, decision-making, or
ultimate advocacy with the FDA.” (JA 12356.) Biovail
responded and confirmed that it “[did] not seek the
participation of GSK[.]” (JA 12358.)
31

In e-mail correspondence, GSK employees
acknowledged that “Biovail is curious if we have any
information on metabolites that might form the basis of a
challenge to the standard bioequivalence testing/standards.”
(JA 13281.) In response, Stan Hull, a GSK employee,
explained that “[the] information is available internally, but
has not been shared with Biovail, and it is our
recommendation not to share metabolite data with Biovail.”
(JA 13282.)
32

Indeed, in a heading in their statement of undisputed
facts, the Appellants state that “GSK concluded that the
bioequivalence ... argument was wrong.”

43

Case: 15-2875

Document: 003112704120

Page: 44

Date Filed: 08/17/2017

In arguing otherwise, the Appellants point to an e-mail
between two Biovail employees noting that “David [S]tout
[who, at the time, was the President of U.S. Pharmaceutical
Operations for GSK ... said that a generic to Wellbutrin XL
would have to prove bioequivalence to [Wellbutrin IR] not to
[Wellbutrin XL] to get approved.” (JA 12621.) Assuming
the relevance of that statement, however, simple
communication does not establish a conspiracy.
The
Appellants claim that the e-mail shows that “GSK developed
the [bioequivalence argument] and gave it to Biovail to put in
the petition.” (JA 2383.) But the e-mail does not support that
claim. Nothing in the e-mail indicates that GSK wanted
Biovail to include the bioequivalence argument in a Citizen
Petition – the e-mail does not mention a petition and, in fact,
there was no petition to which GSK could refer – the e-mail
was written over a year and a half before any FDA petition
was filed.
The Appellants also point to an e-mail from David
Stout to several GSK employees that asked the employees to
“coordinate with Biovail on ... [d]eveloping an agreement for
[the concerns relating to bioequivalence, steady-state testing,
and dose-dumping] and formulat[ing] a plan for the filing of
the petition.” (JA 13261.) At the same time, though, the
Appellants ignore a follow-up e-mail indicating that GSK did
not want to move forward with the Petition. A week after
Stout sent his e-mail, he received a response concluding that
“a Citizen’s Petition requesting that all generic versions of
Wellbutrin extended-release products be required to
demonstrate a lack of food effect appears unnecessary as
these are current confirmed requirements.” (JA 13279.) The
evidence is thus inadequate to support the Appellants’ claim
that GSK and Biovail collaborated on Biovail’s Petition,

44

Case: 15-2875

Document: 003112704120

Page: 45

Date Filed: 08/17/2017

much less that they conspired to use the Petition to suppress
competition. 33 Because we can affirm the District Court on
that basis, we do not need to consider whether GSK is entitled
to Noerr-Pennington immunity with respect to the Petition. 34
33

As was the case with each of the four sham lawsuit
claims, there is another causation problem. As we will
explain in more detail below, Anchen’s launch would have
been delayed by a blocking patent owned by Andrx. That
means that any delay caused by Biovail’s FDA Petition is
irrelevant – the blocking patent would have prevented a
lawful launch even in the absence of Biovail’s Petition.
34

In evaluating Biovail’s Petition, the District Court
considered the Petition as a series of four requests. The Court
concluded that two of the four requests were successful, and
thus not baseless, and that two of the four requests were
potentially baseless. The Court then concluded that GSK was
entitled to summary judgment because the Appellants had
failed to show that their injury was attributable to the
unsuccessful (and potentially sham) requests, rather than to
the successful requests.
We have doubts about that reasoning. The flaw is in
viewing the Petition as four independent requests, rather than
as a single petition. When considering whether a petition is
entitled to immunity, courts should consider whether the
petition as a whole is objectively baseless. See Tyco
Healthcare Grp. LP v. Mut. Pharm. Co., 762 F.3d 1338, 1347
(Fed. Cir. 2014); Cheminor Drugs, Ltd. v. Ethyl Corp., 168
F.3d 119, 123 (3d Cir. 1999) (“[W]e will determine whether
Ethyl’s petition was objectively baseless ... .” (emphasis
added)). While the District Court considered the merit of
each of the Petition’s constituent requests, it did not reach any

45

Case: 15-2875

Document: 003112704120

6.

Page: 46

Date Filed: 08/17/2017

Serial Petitioning

In addition to arguing that GSK engaged in sham
litigation with respect to each of the four suits against generic
manufacturers and the Citizen Petition, the Appellants
contend that GSK engaged in serial petitioning, and thus in an
abuse of the opportunity to litigate. They say that we should
vacate and remand to allow the District Court to evaluate
GSK’s actions in light of Hanover 3201 Realty, LLC v.
Village Supermarkets, Inc., 806 F.3d 162 (3d Cir. 2015), an
opinion we issued after the District Court’s final judgment.
GSK responds that the serial petitioning argument was
waived and that, even if we consider it, Hanover is readily
distinguishable. We decline to vacate and remand because
the Appellants have not demonstrated that GSK engaged in
serial petitioning. 35
conclusions regarding whether the Petition, in toto, was
objectively baseless. As a result, the Court’s consideration of
causation and delay was premature.
35

GSK argues that the Appellants did not properly
preserve the serial petitioning claim below, despite the fact
that the Supreme Court precedent on which Hanover 3201
Realty, LLC v. Village Supermarkets, Inc., 806 F.3d 162 (3d
Cir. 2015), relied was decided “more than 40 years ago,” and
that the Appellants’ trial counsel was familiar with it from a
prior case. (Ans. Br. 64.) The Appellants respond that
asserting a general theory of sham litigation was sufficient
because in doing so, they “recounted a series of meritless
petitions filed to frustrate competition and prolong the
defendants’ monopoly.” (Consolidated Reply Brief of Direct
Purchaser and End-Payor Class Plaintiffs-Appellants 74.) In

46

Case: 15-2875

Document: 003112704120

Page: 47

Date Filed: 08/17/2017

In Hanover, we held that a plaintiff could more easily
overcome Noerr-Pennington immunity when the defendant
had engaged in multiple legal actions against the plaintiff.
806 F.3d at 180. We explained that, “[w]here there is only
one alleged sham petition, [PRE]’s exacting two-step test
properly places a heavy thumb on the scale in favor of the
defendant.” Id. at 180. “In contrast, a more flexible standard
is appropriate when dealing with a pattern of petitioning.” Id.
In the latter context, we ask “whether a series of petitions
were filed with or without regard to merit and for the purpose
of using the governmental process (as opposed to the outcome
of that process) to harm a market rival and restrain trade.” Id.
To determine whether a practice of petitioning the
government without regard to merit was used, “a court should
perform a holistic review that may include looking at the
defendant’s filing success—i.e., win-loss percentage—as
circumstantial evidence of the defendant’s subjective
motivations.” Id.
At the outset, we reject the contention that GSK
engaged in serial petitioning through “four lawsuits and a
petition[.]” (Op. Br. 116.) GSK was only involved in two of
Hanover, we rejected an argument that the plaintiffs had
waived a serial litigation claim where the plaintiffs
“consistently” argued “that the sham exception applie[d]” and
“alleg[ed] an ‘illegal scheme’ through a ‘series of sham
litigations,’ [that] put Defendants on notice of the relevant
facts[.]” 806 F.3d at 179 n.13. Although the Appellants did
not independently articulate the “series” claim in the District
Court, we conclude that, as in Hanover, the recitation of
sequential litigious activity was sufficient to preserve the
claim.

47

Case: 15-2875

Document: 003112704120

Page: 48

Date Filed: 08/17/2017

the suits, and then only briefly. GSK did not file suit against
Impax or Watson, nor did it join the Citizen Petition that was
filed with the FDA. Biovail took those steps alone. And we
have already rejected the Appellants’ arguments that GSK
was engaged in a conspiracy with Biovail. When the
Appellants’ serial petitioning claim is reduced to only the
lawsuits against Anchen and Abrika, both of which GSK
withdrew from, it must fail. The test for serial petitioning
announced in Hanover explicitly applies to “a series of legal
proceedings” or “a pattern of petitioning[,]” 806 F.3d at 180,
and two proceedings – each against an independent defendant
– does not constitute a pattern.
The serial petitioning charge is particularly inapt
because GSK’s actions were consistent with the design and
intent of Hatch-Waxman. The Act incentivizes brand-name
drug manufacturers to promptly file patent infringement suits
by rewarding them with a stay of up to 30 months if they do
so. 21 U.S.C. § 355(j)(5)(B)(iii). We are not inclined to
penalize a brand-name manufacturer whose “litigiousness
was a product of Hatch-Waxman.” Kaiser Found. Health
Plan, Inc. v. Abbott Labs., Inc., 552 F.3d 1033, 1047 (9th Cir.
2009). Doing so would punish behavior that Congress sought
to encourage. See id. (recognizing that the “volume of ...
suits” filed by a brand-name manufacturer is “dependent on
the number of generic companies attempting to enter the …
marketplace, a matter over which the [brand-name
manufacturer] ha[s] no control”). For that reason too, we
agree with the District Court’s rejection of the Appellants’
serial petitioning argument.

48

Case: 15-2875

Document: 003112704120

B.

Page: 49

Date Filed: 08/17/2017

Reverse Payment Settlement Agreement

The second major point of contention on appeal relates
to a set of agreements that resolved Biovail’s infringement
suits against Anchen, Impax, and Watson, and Andrx’s
infringement suits against GSK and Anchen. Each of the
agreements was entered into on February 9, 2007, and
together they settled many of the patent disputes related to
Wellbutrin XL. The Appellants argue that the overall
settlement was unlawful and anticompetitive. Before delving
into the details of the agreements, we give some background
on the events leading to the settlement.
1.

Events Leading to the Settlement

In January 2006, in anticipation of the FDA’s approval
of Impax’s ANDA, Anchen, Impax, and Teva
Pharmaceuticals U.S.A entered into an agreement under
which Anchen would waive its first-filer exclusivity to allow
Teva to market Impax’s 300 mg version of Wellbutrin XL. 36
Pursuant to that agreement, Impax and Teva launched a
generic version of 300 mg Wellbutrin XL in December 2006.
36

This is Teva’s first appearance in this case. It is
another producer and distributor of generic drugs.
As explained above, because Biovail filed its suit
against Anchen promptly, it was able to delay the approval of
Anchen’s ANDA for 30 months. However, Biovail did not
file against Impax within 45 days of receiving Impax’s
paragraph IV certification. As a result, the only barrier to the
approval of Impax’s ANDA was Anchen’s first-filer
exclusivity period.

49

Case: 15-2875

Document: 003112704120

Page: 50

Date Filed: 08/17/2017

That same month, Anchen’s ANDA was approved, and
Anchen and Teva entered into a “Distribution and Supply
Agreement,” under which Teva would launch Anchen’s
150 mg version of Wellbutrin XL. At that point, GSK,
Biovail, and Teva (as the distributor for Anchen and Impax)
entered into a “‘standstill’ agreement under which Teva,
Anchen, and Impax agreed not to launch any more 300 mg
generic product or any 150 mg generic product, and Biovail
... agreed not to launch any authorized generic version of
either dosage strength.” 37 (JA 2435.)
In the midst of that standstill, in February 2007, the
parties entered into the series of agreements constituting the
settlement. By that time, Biovail had lost its infringement
suit against Anchen in district court and had an appeal
pending in the United States Court of Appeals for the Federal
Circuit. 38 Each of the other infringement cases was pending
in district court. Also pending was a federal case filed by
Biovail against the FDA challenging the FDA’s treatment of
Biovail’s Citizen Petition.

37

An “authorized generic” is a non-branded version of
a brand-name drug that is produced by the brand-name
company itself.
38

The appeals process was at an early stage – Biovail
filed its notice of appeal on September 13, 2006, and its
opening brief on December 14, 2006.

50

Case: 15-2875

Document: 003112704120

2.

Page: 51

Date Filed: 08/17/2017

The Settlement

Five agreements constituting the overall settlement are
relevant here. The first was between GSK and Andrx and
provided that GSK would settle with Andrx, “paying $35
million to cover past use of the technology described in
Andrx’s patent, plus an ongoing royalty rate in exchange for a
license to the patent.” 39 (Ans Br. 14; see also JA 34043.)
GSK also obtained the right to sublicense the Andrx patent.
The second agreement was a license between Teva and
Biovail. It contained three relevant provisions. First, it
granted Teva a 180-day exclusive license to certain Biovail
patents, 40 so that Teva could sell a 150 mg version of generic
Wellbutrin XL beginning on May 30, 2008, or earlier if
Biovail lost its appeal in the Anchen case. 41 Second, the
agreement granted Teva an exclusive license to Biovail
patents so that it could sell a 300 mg version of generic
39

In November 2006, Andrx was acquired by Watson.
For simplicity, we will continue to refer to Andrx by that
name.
40

To ensure that the license would, in fact, be
exclusive, Biovail amended its development agreement with
GSK. To facilitate the settlement, GSK agreed to refrain
from launching an authorized generic version of 150 mg
Wellbutrin XL for the duration of Teva’s exclusive license.
41

The agreement contained five other “trigger”
provisions that would allow Teva to market generic versions
of Wellbutrin. None of the other provisions is relevant here.

51

Case: 15-2875

Document: 003112704120

Page: 52

Date Filed: 08/17/2017

Wellbutrin XL. 42 The license ran from December 13, 2006
through June 12, 2007. 43 Finally, the agreement required
Biovail to provide Teva with a supply of 150 mg and 300 mg
generic Wellbutrin XL. 44 Specifically, Biovail agreed to
supply Teva with 75 million tablets of the 150 mg dosage.
The agreement also contained two unlimited supply
provisions (one for the 150 mg dosage and one for the 300
mg dosage), obligating Biovail to provide Teva with an
unlimited supply of Wellbutrin XL in the event that Biovail’s
Citizen Petition ended up interfering with Teva’s launch.
The third agreement was between Anchen and Biovail.
In that agreement, Biovail granted Anchen a sublicense to
Andrx’s ’708 patent – the patent that had been the subject of
litigation involving Anchen’s 150 mg generic version of
Wellbutrin XL. 45 The parties also acknowledged that the
42

Biovail and GSK similarly amended their
development agreement to preclude GSK from launching an
authorized generic version of 300 mg Wellbutrin XL.
43

The license was written to retroactively authorize the
sales that Teva had made prior to entering the license
agreement. The Appellants contend that the license was
meaningless and that “Teva did not need a license to make
and sell the product because Teva was already doing so.” (JA
2635.)
44

The agreement specified that the tablets must be
“Generic NDA Equivalent.” (JA 34081.)
45

The agreement explains that Biovail was an assignee
of the license agreement between Andrx and GSK, and thus

52

Case: 15-2875

Document: 003112704120

Page: 53

Date Filed: 08/17/2017

agreement did not “settle or otherwise end the Biovail
Anchen lawsuit.” (JA 3697.)
The fourth agreement was between Biovail and Impax,
in which Biovail agreed to dismiss its infringement suit
against Impax and agreed not to sue Impax for selling or
manufacturing generic versions of Wellbutrin XL outside of
Anchen’s 180-day exclusivity period.
The fifth agreement was an “omnibus” one in which
the several parties acknowledged that the second through fifth
agreements were related to each other and agreed to submit
those agreements to the FTC for approval. The parties further
agreed to modify the agreements in response to any FTC
concerns. 46
that Biovail had the ability to grant a sublicense to Andrx’s
patent.
46

By statute, the parties were required to submit their
several settlement agreements to the FTC within 10 days.
Medicare
Prescription
Drug,
Improvement,
and
Modernization Act of 2003, Pub. L. No. 108-173, § 1112(a),
117 Stat 2066, 2461-62 (2003). The omnibus agreement
required the parties to submit the agreements to the FTC
within two days and to either revise or terminate the
agreement in response to any FTC concerns. In addition to
the five agreements described above, Andrx sent a letter to
Anchen explaining that, in light of the license agreement
between GSK and Andrx, Andrx would dismiss its
infringement suit against Anchen. That same day, Biovail
and GSK amended their development agreement (the
agreement that granted GSK an exclusive license to Biovail’s

53

Case: 15-2875

Document: 003112704120

Page: 54

Date Filed: 08/17/2017

Pursuant to the terms of the agreements, Anchen
waited until May 2008 to launch its 150 mg generic version
of Wellbutrin XL, and GSK waited 180 days to launch
authorized generic versions of both 150 mg and 300 mg
Wellbutrin XL.
3.

The Appellants Cannot Prevail on Their
Antitrust Claims Pertaining to the
Alleged Reverse Payment

In order to prevail on an antitrust claim, a private
plaintiff must establish antitrust standing, Ethypharm S.A.
France v. Abbott Laboratories, 707 F.3d 223, 232-33 (3d Cir.
2013), and must show that the defendant’s actions violated
antitrust law. Phillip E. Areeda & Herbert Hovenkamp,
Fundamentals of Antitrust Law 3-16 (4th ed. 2015). In this
case, there is an additional threshold question – whether the
challenged agreements are immune from antitrust scrutiny as
the valid exercise of patent rights. See Dawson Chem. Co. v.
Rohm & Haas Co., 448 U.S. 176, 215 (1980) (explaining that
“the essence of a patent grant is the right to exclude others
from profiting by the patented invention”); Actavis, 133 S. Ct.
at 2238 (Roberts, C.J., dissenting) (“A patent grants the right
to exclude others from profiting by the patented invention. In
doing so it provides an exception to antitrust law, and the
scope of the patent ... forms the zone within which the patent

patents) to bring it into compliance with the substance of the
Teva-Biovail agreement. Finally, Biovail agreed to dismiss
its suit against Watson. There was no payment (reverse or
otherwise) associated with the Biovail-Watson settlement.

54

Case: 15-2875

Document: 003112704120

Page: 55

Date Filed: 08/17/2017

holder may operate without facing antitrust liability.”
(internal quotation marks and citation omitted)).
a)

The Agreements Are Not Immune
from Antitrust Scrutiny; the Rule
of Reason Test Applies

The Supreme Court considered the legality of reverse
payment settlements in FTC v. Actavis, 133 S. Ct. 2223. In
that case, a brand-name drug manufacturer sued an ANDA
applicant. Id. at 2229. After litigating the case for several
years, the parties entered into a settlement agreement whereby
the brand manufacturer paid the generic manufacturer tens of
millions of dollars in exchange for the generic’s agreement to
delay its entry into the market for nine years. Id. at 2229-30.
The FTC filed suit challenging the settlement agreement. Id.
at 2227. Although the United States Court of Appeals for the
Eleventh Circuit held that reverse payment settlements should
be immune from antitrust liability, as long as they fall within
the scope of the relevant patents, FTC v. Watson Pharm., Inc.,
677 F.3d 1298, 1312 (11th Cir. 2012), the Supreme Court saw
it differently. It said that “reverse payment settlements ... can
sometimes violate the antitrust laws” and that “courts
reviewing such agreements should ... apply[] [the] ‘rule of
reason’ [test].” Actavis, 133 S. Ct. at 2227, 2237. 47 In
reaching that conclusion, the Court observed that “it would be
incongruous to determine antitrust legality by measuring the
47

The Supreme Court explicitly rejected the claim that
reverse payment settlement agreements are “presumptively
unlawful” and concluded that it would also be improper to
evaluate reverse payment agreements via a “quick-look”
approach. Id. at 2237.

55

Case: 15-2875

Document: 003112704120

Page: 56

Date Filed: 08/17/2017

settlement’s anticompetitive effects solely against patent law
policy, rather than by measuring them against procompetitive
antitrust policies as well.” Id. at 2231. The Court then
explained that reverse payments can generate “genuine
adverse effects on competition” by allowing brand
manufacturers to “avoid the risk of patent invalidation or a
finding of noninfringement.” Id. at 2235-36. Ultimately, the
Court concluded that “a reverse payment, where large and
unjustified, can bring with it the risk of significant
anticompetitive effects[.]” Id. at 2237.
In King Drug Co. of Florence, Inc. v. Smithkline
Beecham Corp., 791 F.3d 388 (3d Cir. 2015), we considered
whether antitrust scrutiny under Actavis was limited to
reverse payments of cash, or whether other “transfers of
value” would also be subject to scrutiny. Like the agreement
at issue in this case, King involved a settlement in which a
brand manufacturer agreed not to produce an “authorized
generic” version of its drug – a so-called “no-AG agreement.”
Id. at 394. The antitrust plaintiffs in that case alleged that the
no-AG agreement qualified as a reverse payment under
Actavis. See id. We agreed and held that “Actavis’s holding
[could not] be limited to reverse payments of cash.” Id. at
403. We explained:
[A] brand’s commitment not to produce an
authorized generic means that it must give up
the valuable right to capture profits ... . The noAG agreement transfers the profits the patentee
would have made from its authorized generic to
the settling generic—plus potentially more ...
because there will now be a generic monopoly
instead of a generic duopoly.

56

Case: 15-2875

Document: 003112704120

Page: 57

Date Filed: 08/17/2017

Id. at 405. As a result, we concluded that “no-AG agreements
are likely to present the same types of problems as reverse
payments of cash” and that “[t]he anticompetitive
consequences of [a no-AG agreement] may be as harmful as
those resulting from reverse payments of cash.” Id. at 40405.
When evaluating the challenged settlements in this
case, the District Court suggested, but did not hold, that they
might be beyond the reach of antitrust law. According to the
Court, “the Wellbutrin Settlement does not present the same
antitrust concerns that motivated the court in Actavis to
subject the settlement to antitrust scrutiny” because “the
Wellbutrin Settlement required the underlying patent
litigation to continue, maintaining the risk of a finding of
patent invalidity or non-infringement[.]”
(JA 182-83.)
Despite that intimation, the Court declined to hold that “any
reverse payment that allows the underlying patent litigation to
continue is automatically exempt from the antitrust laws.”
(JA 184.) Instead, the Court analyzed the settlement using
the rule of reason. On appeal, GSK echoes the initial
intimation of the District Court and maintains that “[t]he
settlement did not pose the anticompetitive harm the Supreme
Court identified in Actavis[.]” (Ans. Br. 67.) We disagree.
In light of Actavis and our decision in King, the
agreements at issue in this case, as they relate to Anchen’s
generic version of 150 mg Wellbutrin XL, must be evaluated
under the rule of reason test. As explained above, the
agreements include an alleged reverse payment and pay-fordelay scheme: in exchange for a 180-day no-AG agreement
from Biovail and GSK (the reverse payment), Anchen agreed
not to launch a generic version of 150 mg Wellbutrin XL until

57

Case: 15-2875

Document: 003112704120

Page: 58

Date Filed: 08/17/2017

the occurrence of a triggering event. Moreover, there is some
support in the record for the assertion that the reverse
payment is large and unjustified, see Actavis, 133 S. Ct. at
2237 (suggesting that reverse payments are especially
problematic if they are “large and unjustified”). First, the
payment can be said to be large. According to the
Appellants’ economic expert, 48 the no-AG agreement was
worth $233 million to Anchen, Teva, and Impax – an amount
that would qualify as large in most any context. See Actavis,
133 S. Ct. at 2237 (explaining that “the likelihood of a reverse
payment bringing about anticompetitive effects depends upon
its size [and] scale in relation to the payor’s anticipated future
litigation costs”). The “payment,” i.e., the no-AG agreement,
could also be said to be unjustified in the sense of being
unexplained. 49 In particular, it was not tied to the merits of
the litigation between Biovail and Anchen. We know that the
48

The District Court granted a Daubert motion to
exclude that expert’s opinion relating to the rule of reason
analysis. However, the Daubert motion did not appear to
challenge the expert’s opinions with respect to the value
associated with the no-AG agreement. And, in Actavis, the
Supreme Court recognized the immense value associated with
market exclusivity. See Actavis, 133 S. Ct. at 2229 (citing C.
Scott Hemphill, Paying for Delay: Pharmaceutical Patent
Settlement as a Regulatory Design Problem, 81 N.Y.U. L.
Rev. 1553, 1579 (2006)).
49

We intend no comment on whether a no-AG
promise could be justified in the sense of being a sound
exercise of business judgment and consonant with good
public policy.

58

Case: 15-2875

Document: 003112704120

Page: 59

Date Filed: 08/17/2017

no-AG agreement was not linked to the merits of the
litigation because its value did not depend on the outcome of
the appeal before the Federal Circuit. The duration of the noAG promise was fixed at 180 days, regardless of who
prevailed in the case, and that duration provided value to
Anchen, as well as to Teva and Impax. 50 Because the
50

It is worth noting that this case differs from Actavis
and King in at least one important respect: in both Actavis and
King, the challenged agreements ended litigation between the
brand-name manufacturer and the generic manufacturer. In
this case, however, the agreements allowed Biovail’s appeal
to continue. (See JA 3697 (explicitly noting that the
agreement did not “settle or otherwise end the Biovail
Anchen lawsuit”).) In acknowledging that difference, the
District Court concluded that the agreements in this case
“do[] not present the same antitrust concerns that motivated
the court in Actavis to subject the settlement to antitrust
scrutiny.” (JA 183.) We question that conclusion. While
there is language in Actavis that describes the premature
termination of litigation as an anticompetitive harm, see 133
S. Ct. at 2236 (explaining that a patentee should not be
allowed to “us[e] its monopoly profits to avoid the risk of
patent invalidation or a finding of noninfringement”), the
Supreme Court’s holding was not so narrow. Instead, Actavis
stands for the broader proposition that both “patent and
antitrust policies are ... relevant in determining the ‘scope of
the patent monopoly’—and consequently antitrust law
immunity—that is conferred by a patent.” Id. at 2231. In
other words, the Court took issue with reverse payments not
simply because they could lead to the premature termination
of litigation, but rather because they eliminate the risk of
competition. Id. at 2236; King, 791 F.3d at 405.

59

Case: 15-2875

Document: 003112704120

Page: 60

Date Filed: 08/17/2017

agreements at issue here are such as to implicate the concerns
identified in Actavis, they are not immune from antitrust
scrutiny and must, to a degree, be evaluated under the rule of
reason test.
That “to a degree” qualifier is added because our
conclusion is limited to the agreements as they relate to
Anchen’s generic version of 150 mg Wellbutrin XL. We
reach a different conclusion with respect to the agreements as
they relate to Anchen’s 300 mg Wellbutrin XL. 51 As
While the agreements at issue here did not end the
litigation between Biovail and Anchen, they nevertheless
implicate the kinds of concerns articulated in Actavis by
delaying the entry of 150 mg generic Wellbutrin XL and by
delaying the entry of an authorized generic version of both
150 and 300 mg Wellbutrin XL. That conclusion follows
directly from Actavis and King and is also supported by the
FTC, which filed an amicus brief in this case. (See FTC Br.
15 (“An agreement that forecloses the possibility of at-risk
entry into the market (in exchange for shared monopoly
profits) can also be anticompetitive under that analysis.”).)
The view of the law espoused by the FTC, adopted by the
majority in Actavis, and followed by our Court in King, has
been subject to cogent criticism, see, e.g., Actavis, 133 S. Ct.
at 2240-47 (Roberts, C.J., dissenting), but the controlling
precedent is what it is.
51

It does not appear that the Appellants have presented
any arguments relating exclusively to Anchen’s generic
version of 300 mg Wellbutrin XL. As a result, any arguments
the Appellants might have regarding the 300 mg product
could be viewed as waived. Nagle v. Alspach, 8 F.3d 141,

60

Case: 15-2875

Document: 003112704120

Page: 61

Date Filed: 08/17/2017

explained above, Anchen, in partnership with Impax and
Teva, launched a 300 mg version of Wellbutrin XL in
December 2006 – as soon as its ANDA was approved. The
agreements reached in February 2007 allowed Teva to
continue marketing that product. As a result, there was no
delay associated with the 300 mg product and the analysis in
Actavis does not apply. As a result, any pay-for-delay claim
unique to Anchen’s 300 mg product must fail. 52
b)

The Appellants Do Not Have
Antitrust Standing

In order to maintain an antitrust suit, a plaintiff must
establish antitrust standing, which is distinct from Article III
standing. While Article III standing is rooted in the
Constitution, antitrust standing is a judge-made doctrine. 53
143 (3d Cir. 1993). However, for the reasons stated above,
the Appellants would not prevail on that issue even if it were
not waived.
52

One could argue that the no-AG agreement relating
to the 300 mg product was part of the payment used to
persuade Anchen to delay its launch of the 150 mg product.
If one adopts that view, then the 300 mg no-AG agreement
would be subject to the same analysis as the 150 mg no-AG
agreement and there would not be any claim unique to the
generic 300 mg product.
53

Though judge-made, federal antitrust standing is
rooted in federal statutory law, and antitrust standing under
state law is likewise rooted in the respective statutes of the
several states represented within the ranks of the indirect-

61

Case: 15-2875

Document: 003112704120

Page: 62

Date Filed: 08/17/2017

Associated Gen. Contractors of Cal., Inc. v. Cal. State
Council of Carpenters, 459 U.S. 519, 534-35 & n.31 (1983);
Ethypharm S.A. France v. Abbott Labs., 707 F.3d 223, 232
n.17 (2013) (“[A]ntitrust standing is based on prudential
principles.”). It is not a jurisdictional requirement. In re
Modafinil Antitrust Litig., 837 F.3d 238, 263 n.30 (3d Cir.
2016). And while “[h]arm to the antitrust plaintiff is
sufficient to satisfy the constitutional standing requirement of
injury in fact,” courts must also consider “whether the
plaintiff is a proper party to bring [the] private antitrust
action.” Associated Gen. Contractors, 459 U.S. at 535 n.31.
In that sense, antitrust standing is more properly viewed as an
element of an antitrust claim that can be resolved at summary
judgment. Ethypharm S.A. France, 707 F.3d at 232 n.15
purchaser Appellants. The state requirements for antitrust
standing are functionally identical to the federal requirements
in this respect – each of the state antitrust laws asserted in this
case requires antitrust plaintiffs to prove that they have
suffered an antitrust injury. Moreover, the standards for
proving antitrust injury under the state laws appear to be
identical to the standards under federal law. See Breakdown
Servs., Ltd. v. Now Casting, Inc., 550 F. Supp. 2d 1123, 1141
(C.D. Cal. 2007); Boulware v. State of Nev., Dept. of Human
Res., 960 F.2d 793, 800-01 (9th Cir. 1992) (Nevada);
Benjamin of Forest Hills Realty, Inc. v. Austin Sheppard
Realty, Inc., 823 N.Y.S.2d 79, 94 (App. Div. 2006); Lerma v.
Univision Commc’ns, Inc., 52 F. Supp. 2d 1011, 1016 (E.D.
Wis. 1999); Rockholt Furniture, Inc. v. Kincaid Furniture
Co., 1998 WL 1661384, at *7 (E.D. Tenn. July 6, 1998). As
a result, our standing and causation analysis on this issue
applies equally to the direct purchasers’ claims and the
indirect purchasers’ claims.

62

Case: 15-2875

Document: 003112704120

Page: 63

Date Filed: 08/17/2017

(indicating that antitrust standing is a “merits issue”); see also
Barton & Pittinos, Inc. v. SmithKline Beecham Corp., 118
F.3d 178, 182 (3d Cir. 1997) (considering the question of
antitrust standing at summary judgment, and determining
whether the plaintiff “adduced sufficient evidence to permit a
reasonable factfinder to conclude that it competed in the
market in which trade was allegedly restrained, such that its
alleged injury would constitute ‘antitrust injury’”); McCarthy
v. Recordex Serv., Inc., 80 F.3d 842, 852-54 (3d Cir. 1996)
(resolving the question of antitrust standing at summary
judgment).
To establish antitrust standing, a plaintiff must show
that it has suffered an antitrust injury 54 – that is, an “injury of
54

“The Supreme Court ... [has] articulated several
factors to be considered when deciding whether a
complainant has antitrust standing.” Ethypharm S.A. France,
707 F.3d at 232 (3d Cir. 2013) (internal quotation marks and
citations omitted). Those factors include:
(1) the causal connection between the antitrust
violation and the harm to the plaintiff and the
intent by the defendant to cause that harm, with
neither factor alone conferring standing; (2)
whether the plaintiff’s alleged injury is of the
type for which the antitrust laws were intended
to provide redress; (3) the directness of the
injury, which addresses the concerns that liberal
application of standing principles might
produce speculative claims; (4) the existence of
more direct victims of the alleged antitrust
violations; and (5) the potential for duplicative

63

Case: 15-2875

Document: 003112704120

Page: 64

Date Filed: 08/17/2017

the type the antitrust laws were intended to prevent and that
flows from that which makes [the] defendants’ acts
unlawful.” 55 Ethypharm S.A. France, 707 F.3d at 233 (3d
recovery or
damages.

complex

apportionment

of

Id. at 232-33 (internal citations omitted). Because “[t]he
second factor, antitrust injury, is a necessary but insufficient
condition of antitrust standing[,] ... if it is lacking, we need
not address the remaining ... factors.” Id. at 233 (internal
quotation marks and citation omitted); see also City of
Pittsburgh v. W. Penn Power Co., 147 F.3d 256, 265
(“[B]ecause there is no causal connection and no antitrust
injury, we need not examine the other ... standing factors.”).
55

In Illinois Brick Co. v. Illinois, the Supreme Court
held that indirect purchasers do not have standing to bring
antitrust suits under federal law. 431 U.S. 720, 730-31
(1977); see also In re Lower Lake Erie Iron Ore Antitrust
Litig., 998 F.2d 1144, 1163 n.10 (3d Cir. 1993) (“[In Illinois
Brick Co.,] [t]he Court held that § 4 [of the Clayton Act] did
not permit ... indirect purchasers ... to recover for the
overcharge passed through the chain of distribution.”).
However, indirect purchasers do have standing to assert
antitrust claims in each of the state causes of action asserted
here. Nev. Rev. Stat. § 598A.210 (Nevada statute granting
standing to indirect purchasers to recover for antitrust
violations); In re Dynamic Random Access Memory (Dram)
Antitrust Litig., 516 F. Supp. 2d 1072, 1094-95 (N.D. Cal.
2007) (recognizing that indirect purchaser suits are permitted
in Arizona, Kansas, Maine, Michigan, Minnesota,
Mississippi, Nebraska, Nevada, New Mexico, North Carolina,

64

Case: 15-2875

Document: 003112704120

Page: 65

Date Filed: 08/17/2017

Cir. 2013) (alteration in original) (quoting Brunswick Corp.,
429 U.S. at 489).
In order to establish antitrust injury here, the
Appellants must show that the harm they say they
experienced – increased drug prices for Wellbutrin XL (and
its generic equivalents) – was caused by the settlement they
are complaining about. See Zenith Radio Corp. v. Hazeltine
Res., Inc., 395 U.S. 100, 114 n.9 (1969) (explaining that,
under the Clayton Act, a plaintiff must prove that it has
suffered at least “some damage flowing from the unlawful
conspiracy”). The Appellants attempt to meet their burden by
pointing to evidence showing that, in the absence of the
agreements, Anchen (partnering with Teva) would have
launched its 150 mg generic no later than the middle of 2007.
At first glance, that argument seems appealing.
Indeed, the District Court found that there was at least a
question of fact as to whether Anchen would have launched
the drug in June 2007. The problem with the argument,
however, is that it does not take into account Andrx’s
blocking patent, the ’708 patent. It is not enough for the
Appellants to show that Anchen wanted to launch its drug;
they must also show that the launch would have been legal.
After all, if the launch were stopped because it was illegal,
then the Appellants’ injury (if it could still be called that)
North Dakota, South Dakota, and Wisconsin); Indirect
Purchaser Lawsuits: A State-by-State Survey 27, 215, 287,
337 (Eric McCarthy et al., eds., 2010) (explaining that
California, New York, Tennessee, and Wisconsin each allow
indirect purchasers to assert antitrust claims).

65

Case: 15-2875

Document: 003112704120

Page: 66

Date Filed: 08/17/2017

would be caused not by the settlement but by the patent laws
prohibiting the launch. See In re Nexium (Esomeprazole)
Antitrust Litig., 842 F.3d 34, 62-63 (1st Cir. 2016) (“[T]he
argument that [the generic manufacturer] would have incurred
the risk of launching at risk or that [it] would have won its ...
suit against [the patent holder] depends on the theory that ...
[the] patents were invalid or not infringed by a generic
version.”); Phillip E. Areeda & Herbert Hovenkamp,
Fundamentals of Antitrust Law § 3.04[B] (rev. 4th ed. Supp.
2015) (“[A] plaintiff cannot be injured in fact by private
conduct excluding it from the market when a statute prevents
the plaintiff from entering that market in any event.”). 56
That a regulatory or legislative bar can break the chain
of causation in an antitrust case is beyond fair dispute. For
example, in RSA Media, Inc. v. AK Media Grp., Inc., 260
F.3d 10, 15 (1st Cir. 2001), the First Circuit decided that the
plaintiff was excluded from the outdoor billboard market not
because of the defendant’s actions but rather “because the
Massachusetts regulatory scheme ... [prevented] new
billboards from being built.” Similarly, in In re Canadian
Import Antitrust Litigation, 470 F.3d 785, 790-91 (8th Cir.
2006), the Eighth Circuit held that the plaintiffs faced higher
drug prices not because drug companies excluded cheaper
56

GSK also argues that Anchen’s launch would have
been blocked by FDA regulations relating to Anchen’s
production facilities. We do not consider that argument
because, even if it were given full weight, it would only show
that Anchen would have had to wait until June 12, 2007 to
launch – a date that was almost a year prior to its actual
launch. In other words, the argument relates to the length of
delay, rather than the existence of a delay.

66

Case: 15-2875

Document: 003112704120

Page: 67

Date Filed: 08/17/2017

Canadian drugs from the market but because federal law
excluded the cheaper Canadian drugs. See City of Pittsburgh
v. W. Penn Power Co., 147 F.3d 256, 265 (3d Cir. 1998)
(applying the same principle and concluding that any injury
suffered by the plaintiff resulted from “the realities of the
regulated environment” rather than from the defendants’
actions). In this case, the launch of Anchen’s 150 mg version
of Wellbutrin XL was effectively blocked by federal patent
law, which, through Andrx’s ’708 patent, would have
prevented market entry.
The Appellants offer two arguments to fend off that
conclusion – one legal and one factual. Their legal argument
is that the reasoning just given was repudiated by our decision
in Consolidated Express, Inc., v. New York Shipping
Association. 602 F.2d 494 (3d Cir. 1979), vacated 448 U.S.
902 (1980), remanded and affirmed, 641 F.2d 90 (3d Cir.
1981). They misread that case. In Consolidated Express, we
held that an antitrust plaintiff’s improper conduct did not
preclude that plaintiff from asserting an antitrust claim
unrelated to the improper conduct. Id. at 508. By contrast,
our holding in this case is that the antitrust claim fails because
the actions of GSK, the defendant, did not actually cause the
Appellants’ claimed injury. But even if the Appellants had a
correct reading of Consolidated Express, their argument
would still fail because that case predates significant
developments
in
antitrust
standing
jurisprudence.
Consolidated Express was decided in 1979, before the
Supreme Court established its antitrust standing “factors” in
Associated General Contractors four years later.
See
Merican, Inc. v. Caterpillar Tractor Co., 713 F.2d 958, 965
(3d Cir. 1983) (applying the Supreme Court’s decision in
Associated General Contractors). We later adopted the very

67

Case: 15-2875

Document: 003112704120

Page: 68

Date Filed: 08/17/2017

argument that the Appellants now claim is not good law. In
City of Pittsburgh we said that no antitrust standing exists
when a plaintiff’s grievance is caused by a regulatory scheme
rather than by the defendant’s actions. 147 F.3d at 266. We
decline to deviate from the well-reasoned path marked in City
of Pittsburgh.
The Appellants’ factual response is that, but for the
challenged agreements, Anchen would have been able to
launch its 150 mg version of Wellbutrin XL without running
afoul of Andrx’s patent. They offer two scenarios. First, they
argue that, in the absence of the challenged agreements,
Anchen would have obtained a license to Andrx’s patent. We
will refer to that as the license-based scenario. Alternatively,
they argue that, in the absence of the challenged agreements,
Anchen would have prevailed against Andrx in litigation. We
will refer to that as the litigation-based scenario. The record
supports neither.
(1)

License-Based Scenario

The Appellants contend that, for at least three reasons,
Anchen would have obtained a license from Andrx. First,
they say that GSK failed to produce evidence “showing [that]
… GSK’s no-AG payment or the generic delay … were
necessary in order [for Anchen] to secure a ... license [to
Andrx’s patent].” (Op. Br. 74.) That argument, however,
flips the burden of proof. As the plaintiffs, the Appellants
have the burden of proving that they have been injured. In
order to withstand summary judgment, they must point to
evidence affirmatively showing that Anchen could have
launched. See W. Penn Allegheny Health Sys., Inc. v. UPMC,
627 F.3d 85, 101 (3d Cir. 2010) (“[T]he plaintiff must

68

Case: 15-2875

Document: 003112704120

Page: 69

Date Filed: 08/17/2017

establish that it suffered an antitrust injury.”). It is no good
saying that the defendants have failed to disprove causation.
See id.
Second, the Appellants say that Andrx had “an
independent economic interest” in providing a license to
Anchen. (Op. Br. 74.) Their reasoning is that Andrx was a
non-practicing entity and thus “could only profit from its ‘708
patent through licenses.” (Id. at 74.) That argument is both
incorrect and insufficient. The argument is incorrect because,
as noted above, supra n.39, Watson acquired Andrx in
November 2006. That means that Andrx was, by that time,
not a non-practicing entity and in fact had a reason to deny
Anchen a license. If Anchen were precluded from launching
its product, then Anchen would waive its exclusivity period,
allowing Watson (a/k/a Andrx) to enter the market earlier.
See 21 U.S.C. § 355(j)(5)(D)(i) (outlining the conditions in
which a first-filer waives its exclusivity). But, even if the
Appellants’ argument were better rooted in reality, it would
be insufficient. In order to withstand summary judgment, the
Appellants must produce evidence from which a reasonable
jury could conclude that it is more likely than not that Anchen
would have obtained a license. Evidence showing that
Anchen may have been able to obtain a license does not meet
that standard. A plaintiff cannot satisfy the summary
judgment burden based on speculation alone. See Halsey v.
Pfeiffer, 750 F.3d 273, 287 (3d Cir. 2014) (“[A]n inference
based upon a speculation or conjecture does not create a
material factual dispute sufficient to defeat [entry of]
summary judgment.” (quoting Robertson v. Allied Signal,
Inc., 914 F.2d 360, 382 n.12 (3d Cir. 1990))); Fedorczyk v.
Caribbean Cruise Lines, Ltd., 82 F.3d 69, 76 (3d Cir. 1996)
(affirming a grant of summary judgment because “[b]ased on

69

Case: 15-2875

Document: 003112704120

Page: 70

Date Filed: 08/17/2017

the evidence presented, a jury could only speculate” as to
whether the defendant’s actions actually caused the claimed
injury). 57
Third, the Appellants argue that Anchen was
negotiating a license agreement with Andrx in the days
preceding the agreements and had agreed on all but one term.
Based on those negotiations, the Appellants argue, a
57

The Appellants point to our recent decision in In re
Fosamax (Alendronate Sodium) Products Liability Litigation,
852 F.3d 268 (3d Cir. 2017), to support their claim that juries
are “‘routinely’ given questions that permissibly require them
to ‘predict the outcome of a hypothetical scenario’ or to
‘speculate’ or ‘guess what could have happened in a
counterfactual setting’ by assessing corporate motives and
thought processes or by ‘evaluat[ing] inference[s] about
human behavior.’” (March 27, 2017 28(j) letter at 2
(alterations in original) (quoting Fosamax, 852 F.3d at 289,
297, 299).) Their argument is correct, but irrelevant. The
fact that juries may predict the outcome of hypothetical
scenarios says nothing about the type or amount of evidence
that is needed for a plaintiff to withstand summary judgment
on a claim involving a counterfactual scenario. As explained
above, the Appellants have not presented sufficient evidence
upon which a reasonable jury could rely to conclude that it is
more likely than not that Anchen and Andrx would have
entered into a license agreement in the counterfactual world.
While it may be better than speculative that Anchen and
Andrx would have had an incentive to talk, it is, on this
record, pure speculation that they would have reached an
agreement.

70

Case: 15-2875

Document: 003112704120

Page: 71

Date Filed: 08/17/2017

reasonable jury could infer that the two companies would
have reached an agreement. But this argument too is
completely speculative. It is certainly possible that Anchen
and Andrx would have reached an agreement, but it is also
certainly possible that the negotiations would have stalled and
failed. Many a contract has foundered on a single dealbreaker point. Without more specific or concrete evidence,
the jury in this case would be left with nothing on which it
could rely to reach a conclusion one way or the other.
Summary judgment was thus appropriate.
(2)

Litigation-Based Scenario

The Appellants’ litigation-based scenario is premised
on the idea that Anchen would have prevailed in Andrx’s
infringement suit. If Andrx’s ’708 patent were invalid, or if
it did not cover Anchen’s product, then patent law would not
have prevented Anchen’s launch. In order to evaluate the
merit of the litigation-based scenario, we must consider the
substance of that underlying litigation. 58
58

In the Actavis decision, there was a debate between
the majority and the Chief Justice on whether lower courts
would be required to resolve substantive patent questions in
order to adequately assess the merits of reverse payment
antitrust claims. Writing for the majority, Justice Breyer
asserted that “it is normally not necessary to litigate patent
validity to answer the antitrust question ... .” Actavis, 133 S.
Ct. at 2236. The Chief Justice disagreed:
[S]ettling a patent claim cannot possibly impose
unlawful anticompetitive harm if the patent
holder is acting within the scope of a valid

71

Case: 15-2875

Document: 003112704120

Page: 72

Date Filed: 08/17/2017

The Appellants make two arguments relative to the
merits of the patent litigation. First, they say that we should
view the size of the reverse payment as “a surrogate for [the]
patent’s weakness” and conclude that GSK “knew [that
Andrx’s patent] could not prevent generic competition.” (Op.
Br. 86.) While the size of a reverse payment may have some
relevance in determining how confident a litigant is in the
strength of its case, Actavis, 133 S. Ct. at 2236-37 (“In a
word, the size of the unexplained reverse payment can
provide a workable surrogate for a patent’s weakness ... .”), it
is far from dispositive. That is especially so when, as in this
case, the settlement is complex and multi-faceted. For
patent and therefore permitted to do precisely
what the antitrust suit claims is unlawful. This
means that in any such antitrust suit, the
defendant (patent holder) will want to use the
validity of his patent as a defense—in other
words, he’ll want to say “I can do this because I
have a valid patent that lets me do this.” I
therefore don’t see how the majority can
conclude that it won’t normally be “necessary
to litigate patent validity to answer the antitrust
question[.]”
Id. at 2244 (Roberts, C.J., dissenting). The present case
appears to vindicate the Chief Justice’s analysis. As he
predicted, GSK argues that the Andrx patent (which was a
central component of the agreements) defeats the Appellants’
suit, and, as he predicted, we cannot resolve this aspect of the
case without considering the merits of the underlying patent
dispute.

72

Case: 15-2875

Document: 003112704120

Page: 73

Date Filed: 08/17/2017

example, GSK and Biovail may have offered the reverse
payment not because they thought Andrx had a weak patent
but rather because they thought Anchen would improperly
evaluate the patent and launch at-risk. 59 In that scenario,
GSK would lose substantial revenue from having a generic
competitor and would not be entitled to damages if the patent
were vindicated because the patent belonged to Andrx, not to
GSK. That there are multiple plausible ways to interpret the
reverse payment in this case means that the payment alone
tells us less about the merits of the underlying case than the
Appellants wish.
We are also persuaded by an argument raised in the
amicus brief filed by a group of antitrust economists (“the
Economists”). That group explains why risk aversion makes
it difficult to use the size of a settlement as a proxy for the
brand-name’s likelihood of success in litigation:
To explore why risk aversion could lead to the
exchange of consideration having nothing to do
with delayed entry, consider a lottery ticket with
a 50% chance of a $0 payoff and a 50% chance
of a $100 million payoff—i.e., the lottery ticket
has an expected payoff of $50 million. Most
people holding such a ticket would be willing to
accept less than the expected payoff amount to
59

In the context of patent litigation, a launch is said to
be “at-risk” if it takes place before the questions of
infringement and validity are resolved, either through
litigation or a license. See In re Modafinil Antitrust Litig.,
837 F.3d at 244.

73

Case: 15-2875

Document: 003112704120

Page: 74

Date Filed: 08/17/2017

achieve certainty. If a person would trade the
aforementioned lottery ticket for a certain
outcome of $20 million, he or she would
essentially be willing to pay $30 million dollars
to eliminate the risk of holding the lottery ticket
that might result in the $0 payoff. Accepting
the certain outcome of $20 million dollars,
however, does not reflect a belief that a $0
payoff is anything more than [a] 50% [risk].
(Antitrust Economists Br. 11 (internal citation omitted).) We
think that reasoning serves as an effective rebuttal to the
Appellants’ claim that the size of the reverse payment is a
“surrogate” for the weakness of the ’708 patent.
The Appellants’ second argument relating to the
litigation-based scenario relies on testimony provided by
Martin Adelman, GSK’s expert. Adelman estimated that
Andrx had an 80% chance of prevailing with respect to
infringement, a 50% chance of prevailing with respect to
validity, and a 90% chance of prevailing with respect to
inequitable conduct. 60 Because Andrx would have to prevail
with respect to all three issues in order to win the case,
Adelman concluded that “Andrx had approximately a one out
of three chance of winning the cases.” 61 (JA 38717.)
60

Adelman’s estimates and analysis are uncontested.

61

The probability that Andrx would prevail on all
three issues was calculated by multiplying the probability of
success on each issue.
0.8*0.9*0.5=0.36, which is
approximately 1/3.

74

Case: 15-2875

Document: 003112704120

Page: 75

Date Filed: 08/17/2017

However, in a bit of historical irony, it was Anchen’s founder
and CEO who made the invention disclosed in the ’708
patent, and he assigned his rights in it to Andrx. Accordingly,
it is highly likely that assignor estoppel would have prevented
Anchen from arguing that the ’708 patent was invalid or that
the patent was unenforceable because of inequitable
conduct. 62 Mentor Graphics Corp. v. Quickturn Design Sys.,
Inc., 150 F.3d 1374, 1378-79 (Fed. Cir. 1998) (describing the
doctrine of assignor estoppel and explaining that the doctrine
“also prevents parties in privity with an estopped assignor
from challenging the validity of the patent”); Shamrock
Techs., Inc. v. Med. Sterilization, Inc., 903 F.2d 789, 793
(Fed. Cir. 1990) (same). As a result, the only topic left for
litigation would be whether Anchen infringed. On that point,
Adelman’s unrebutted analysis was that Andrx would have an
62

“Assignor estoppel prevents a party who assigns a
patent to another from later challenging the validity of the
assigned patent. This doctrine prevents the unfairness and
injustice of permitting a party to sell something and later to
assert that what was sold is worthless. ... [A]n assignment
contains an implicit representation by the assignor that the
patent rights that he is assigning ... are not worthless.”
Mentor Graphics Corp., v. Quickturn Design Sys., Inc., 150
F.3d 1374, 1378 (Fed. Cir. 1998) (internal quotation marks
and citations omitted); see Shamrock Techs., Inc. v. Med.
Sterilization, Inc., 903 F.2d 789, 793 (Fed. Cir. 1990)
(explaining assignor estoppel also applies to those who are in
privity with the assignor). Because the estoppel applies not
only to the individual inventor but also to those in privity,
Anchen itself, and not just its founder and CEO, would likely
have been estopped.

75

Case: 15-2875

Document: 003112704120

Page: 76

Date Filed: 08/17/2017

80% chance of proving infringement – or, in other words, that
Anchen would only have a 20% chance of winning the suit. 63
Neither the Appellants nor GSK identify any other evidence
in the record that speaks to the possible outcomes of the
Anchen-Andrx litigation. On this record, then, no reasonable
jury could conclude that Anchen would have been more likely
than not to prevail.
Because both of the scenarios advanced by the
Appellants fail to show that Anchen would have been able to
launch its 150 mg version of Wellbutrin XL without running
afoul of the Andrx patent, we conclude that the Appellants
have also failed to show that their injuries were caused by the
overall settlement. Because the Appellants thus do not have
antitrust standing, we will affirm the District Court’s grant of
summary judgment. 64
63

The Appellants’ only response to that analysis is that
GSK never moved for summary judgment on inventor
estoppel and that Anchen had not raised the issue of assignor
estoppel before it settled the case. Even if those assertions
are true, they do not show that Andrx would not have asserted
the estoppel argument as the case progressed.
64

Having concluded that the Appellants lack antitrust
standing, we do not need to consider the District Court’s
application of the rule of reason. We note, however, that the
rule of reason inquiry is fact intensive and is not easy to
resolve at the summary judgment stage. See Poller v.
Columbia Broad. Sys., Inc., 368 U.S. 464, 473 (1962)
(“[S]ummary procedures should be used sparingly in complex
antitrust litigation where motive and intent play leading roles,
the proof is largely in the hands of the alleged conspirators,

76

Case: 15-2875

Document: 003112704120

C.

Page: 77

Class
Certification,
Intervention Issues

Date Filed: 08/17/2017

Daubert,

and

Because we affirm the District Court’s grant of
summary judgment on the merits, we need not address those
other issues on appeal. 65 Cf. Bowen v. Owens, 476 U.S. 340,
344 n.4 (1986) (“Because we reject the equal protection
claim, we do not reach the class certification issue.”); Wilson
v. Quadramed Corp., 225 F.3d 350, 353 n.3 (3d Cir. 2000)
(“We do not reach the class certification issue raised by
Wilson since we [will] affirm the District Court’s dismissal of
the complaint ... .”).

and hostile witnesses thicken the plot.”); W. Penn Allegheny
Health Sys., 627 F.3d at 99 (describing the rule-of-reason as
“fact intensive”); cf. King, 791 F.3d at 411 n.36 (describing
the significance of fact finding in the rule of reason analysis).
65

To recap, those issues are the decisions excluding
the testimony of their economic expert and denying Aetna’s
motion to intervene. Additionally, the indirect-purchaser
Appellants challenge the District Court orders decertifying
the indirect-purchaser class and dismissing certain of the
indirect purchasers’ claims for lack of standing. Finally, GSK
conditionally cross-appeals the Court’s certification of the
direct-purchaser class as well as the Court’s conclusion that
the indirect purchasers satisfy the predominance requirement
of Rule 23.

77

Case: 15-2875

III.

Document: 003112704120

Page: 78

Date Filed: 08/17/2017

Conclusion

For the foregoing reasons, we will affirm the District
Court’s grant of summary judgment.

78