In re: Avandia Marketing v.

PRECEDENTIAL

UNITED STATES COURT OF APPEALS
FOR THE THIRD CIRCUIT
_____________

No. 18-1010
_____________

IN RE: AVANDIA MARKETING, SALES
AND PRODUCTS LIABILITY LITIGATION

UFCW Local 1776 and Participating Employers Health
and Welfare Fund; J.B. Hunt Transport Services, Inc.,
Appellants
______________

On Appeal from the United States District Court
for the Eastern District of Pennsylvania
(District Court Nos. 2-07-md-01871, 2-10-cv-02475, 2-11-cv-
04013)
District Judge: Hon. Cynthia M. Rufe
______________

Argued: March 6, 2019
______________

Before: SMITH, Chief Judge, AMBRO and RESTREPO,
Circuit Judges.

(Filed: December 17, 2019)
Thomas M. Sobol [ARGUED]
Hannah W. Brennan
Edward Notargiacomo
Kiersten A. Taylor
Hagens Berman Sobol Shapiro
55 Cambridge Parkway
Suite 301
Cambridge, MA 02142

James R. Dugan, II
Douglas R. Plymale
The Dugan Law Firm
365 Canal Street
Suite 1000
New Orleans, LA 70130

Eric L. Young
McEldrew Young
123 South Broad Street
Suite 2250
Philadelphia, PA 19109

Eric H. Gibbs
Anne-Marie J. de Bartolomeo
Gibbs Law Group
505 14th Street
Suite 1110
Oakland, CA 94612
Counsel for Appellants

Jay P. Lefkowitz [ARGUED]
Thomas S. Burnett
Kirkland & Ellis LLP

2
601 Lexington Avenue
New York, NY 10022

Nina M. Gussack
Anthony C. Vale
Sean P. Fahey
Kyle A. Dolinsky
Hedya Aryani
Jeremy D. Heep
Yvonne M. McKenzie
Pepper Hamilton LLP
3000 Two Logan Square
18th and Arch Streets
Philadelphia, PA 19103
Counsel for Appellee

3
 ______________

OPINION OF THE COURT
______________

RESTREPO, Circuit Judge.

Plaintiffs, two health benefit plans (“Plans”), appeal the
District Court’s grant of summary judgment in favor of
Defendant, GlaxoSmithKline LLC (“GSK”), the manufacturer
of the prescription drug Avandia. The Plans brought suit
against GSK under various state consumer-protection laws and
the Racketeer Influenced and Corrupt Organizations Act, 18
U.S.C. ch. 96 (“RICO”), based on, among other things, GSK’s
marketing of Avandia. The District Court granted summary
judgment in favor of GSK on the Plans’ claims, finding, in
relevant part, that (i) the Plans’ state-law consumer-protection
claims were preempted by the Federal Food, Drug, and
Cosmetic Act, 21 U.S.C. ch. 9 (“FDCA”); (ii) the Plans had
failed to identify a sufficient “enterprise” for purposes of
RICO; and (iii) the Plans’ arguments related to GSK’s alleged
attempts to market Avandia as providing cardiovascular
“benefits” were “belated.” The Plans assert that the District
Court erred in granting summary judgment, and we agree.

Applying the guidance recently provided by the
Supreme Court in Merck Sharp & Dohme Corp. v. Albrecht,
139 S. Ct. 1668 (2019), we hold that the Plans’ state-law
consumer-protection claims are not preempted by the FDCA.
With respect to their RICO claims, the Plans should have been
given the opportunity to seek discovery prior to the District
Court’s granting summary judgment on such claims. Further,
from the inception of this litigation, the Plans’ claims have
centered on GSK’s marketing of Avandia as providing superior

4
cardiovascular outcomes—in other words, cardiovascular
benefits—as compared to other forms of treatment, and
therefore, the District Court’s refusal to consider the Plans’
“benefits” arguments was in error because those arguments
were timely raised.

Therefore, for the reasons that follow, we will reverse
in part and vacate in part the order of the District Court granting
summary judgment in favor of GSK, and we will remand to the
District Court for further proceedings consistent with this
opinion.

I.

In May 1999, the Food and Drug Administration
(“FDA”) approved Avandia (Rosiglitazone), a drug developed
by GSK, for the treatment of type-2 diabetes. Prior to the
development of Avandia and similar drugs, physicians
primarily treated type-2 diabetes by prescribing metformin
and/or sulfonylureas. GSK, however, marketed Avandia at a
much higher price point than metformin and sulfonylureas: a
one-month supply of Avandia cost approximately $220,
approximately $140 of which typically was covered by
patients’ health benefit plans, whereas a one-month supply of
metformin or sulfonylureas cost approximately $50, about $45
of which typically was covered by patients’ health benefit
plans.

Despite this cost differential, health benefit plans—
including the Plans—placed Avandia on their formularies as a
“covered” drug. The Plans, for example, determined that it was
advantageous to cover the cost of Avandia because GSK
allegedly marketed Avandia as being capable of both
controlling a patient’s blood sugar levels and reducing

5
cardiovascular risk, the latter of which is particularly pertinent
to type-2 diabetes patients, 65% of whom suffer fatal
cardiovascular-related illnesses or complications. Metformin
and sulfonylureas—the drugs that constituted the “standard of
care” for type-2 diabetes prior to Avandia’s development—did
not decrease cardiovascular risk, and therefore, according to
the Plans, GSK presented Avandia as a cost-effective
alternative to those drugs. As a result, health benefit plans
covered a large portion of the expenses related to patients’
prescriptions for Avandia, resulting in approximately $2.2
billion in U.S. sales in 2006 alone.

In 2006, however, concerns arose that Avandia may in
fact increase certain cardiac risks. In August of that year, GSK
submitted a Prior Approval Supplement to the FDA, in which
GSK sought approval to add information to Avandia’s label
regarding the results of a recent meta-analysis of various
clinical trials. The meta-analysis, “ICT-42,” demonstrated that
use of Avandia was associated with a statistically significant
increase in myocardial ischemic events—events during which
the heart does not receive adequate oxygen because blood flow
to it is reduced. In May 2007, GSK submitted an update to its
Prior Approval Supplement, offering a new formulation of its
proposed warning with respect to myocardial ischemic events
that would, among other things, make the warning more
prominent and clear.

Three days after GSK submitted the update to its Prior
Approval Supplement, the New England Journal of Medicine
published a study authored by Dr. Steve Nissen regarding
Avandia (“Nissen Study”), in which Dr. Nissen concluded that
Avandia “was associated with a significant increase in the risk
of myocardial infarction and with an increase in the risk of
death from cardiovascular causes that had borderline

6
significance.” J. App. 1064. Following the release of the
Nissen Study, a representative of GSK held a telephone
conversation with an official at the FDA regarding progress on
the FDA’s review of the Prior Approval Supplement.
According to GSK’s representative, who wrote a memo
memorializing the details of the conversation, the FDA official
advised that another official within the FDA was “calling for
withdrawal of [the] approval” of Avandia, and thus, it was
difficult for FDA officials to agree on labeling language for
Avandia. Sealed App. 655–56. GSK’s representative then
proposed implementing the labelling changes with respect to
myocardial ischemic events through the Changes Being
Effected (“CBE”) process, which permits a drug manufacturer
to implement a change to its label prior to approval of such
label by the FDA. The FDA official “strongly advised against
proceeding” through the CBE process, stating that doing so
“may give legitimacy to Dr. Nissen’s data” and “will make
people think that GSK must have other information.” Id. at
656. The FDA official concluded the conversation by
reminding the GSK representative that he “knew the
regulations,” which state that the drug manufacturer is
ultimately responsible for making the decision to pursue a
labelling change through the CBE process. Id.

On June 8, 2007, the FDA sent a letter (“Letter”) to
GSK regarding the Prior Approval Supplement. In the Letter,
the FDA stated that it had “reviewed the data provided [by
GSK in its Prior Approval Supplement] and f[ou]nd [that] the
information presented [was] inadequate” and that, therefore,
the Prior Approval Supplement was “not approvable.” Id. at
660. The FDA stated that it had “concluded that the pooled
data require[d] further analysis to adequately convey the
potential risk for increased cardiac ischemia associated” with

7
use of Avandia. In particular, the FDA stated that it had
“identified certain subgroups of patients . . . that may be
particularly vulnerable to experiencing an ischemic event”
while using Avandia. Id. The FDA then directed GSK to
provide additional information “to address the deficiency” in
the Prior Approval Supplement, including “[d]ata from studies
included in a meta-analysis performed by Dr. Steven Nissen
published in the New England Journal of Medicine that were
not included in [GSK’s] pooled analysis,” as well as data from
various other clinical trials. Id. at 661.

The FDA expressed its view that the “potential risk of
increased cardiac ischemia [was] a significant finding that may
impact a large proportion of patients with type[-]2 diabetes,”
and as a result, the FDA scheduled a joint meeting of two FDA
advisory committees (“Joint Meeting”) “to discuss the findings
from th[e Prior Approval Supplement] submission, additional
data recently requested, and accruing information from
ongoing clinical trials” of Avandia. Id. The FDA stated that
the “outcome of th[e Joint M]eeting w[ould] be particularly
germane to any labeling or other regulatory action needed for
[Avandia] and should be factored into any resubmission to
address the above deficiencies.” Id.

Later in 2007, the FDA required GSK to implement
various changes to Avandia’s label. Subsequent to issuing the
Letter, the FDA directed GSK to add a black-box warning to
Avandia’s label with respect to the risk of congestive heart
failure that (i) advised physicians and patients that Avandia
“cause[s] or exacerbate[s] congestive heart failure in some
patients,” (ii) instructed physicians to “observe patients [taking
Avandia] carefully for signs and symptoms of heart failure,”
and (iii) warned patients with certain heart conditions not to
take Avandia. J. App. 708. Following the Joint Meeting, the

8
FDA additionally directed GSK to add a black-box warning to
Avandia’s label with respect to the risk of myocardial ischemic
events, advising physicians and patients that a “meta-analysis
of 42 clinical studies . . . , most of which compared Avandia to
placebo, showed Avandia to be associated with an increased
risk of myocardial ischemic events such as angina or
myocardial infarction” and that “[t]hree other studies . . . ,
comparing Avandia to some other approved oral antidiabetic
agents or placebo, have not confirmed or excluded this risk.”
Id. at 743. The FDA also required GSK to include a longer
explanation of the data with respect to the risk of myocardial
ischemic events elsewhere on Avandia’s label.

Approximately three years later, in 2011, the FDA again
directed GSK to revise the warning on Avandia’s label,
including the black-box warning, with respect to the risk of
myocardial ischemic events. By that time, GSK had completed
fifty-two (52) clinical trials. The FDA’s required warning
advised physicians and patients that “[a] meta-analysis of 52
clinical trials . . . , most of which compared Avandia to placebo,
showed Avandia to be associated with a statistically significant
increased risk of myocardial infraction” and that “[b]ecause of
the potential increased risk of myocardial infarction, Avandia
[was] available only through a restricted distribution program.”
Id. at 786. In a memorandum accompanying its direction to
implement the labelling changes, the FDA noted that the
“evidence pointing to a cardiovascular ischemic risk with
[Avandia] is not robust or consistent,” but that “[n]evertheless,
there are multiple signals of concern, from varied sources of
data, without reliable evidence that refutes them.” Id. at 1397.

In November 2013, however, following the
readjudication of a particular clinical trial (“RECORD Trial”),
the FDA concluded that while “[o]ne cannot entirely discount

9
the results of the meta-analysis” that associated Avandia with
a statistically significant increased risk of myocardial ischemic
events, “the totality of the available evidence does not support
a marked signal of cardiovascular harm.” Id. at 1656. The
FDA determined that, following the readjudication of the
RECORD Trial, Avandia “does not appear to be associated
with an increased risk of major adverse cardiovascular events
or death, although a small amount of residual uncertainty
remains.” Id. at 1657. The FDA directed GSK to revise
Avandia’s label “to reflect the current level of knowledge
regarding [its] cardiovascular risk.” Id.

In 2014, GSK revised Avandia’s label pursuant to the
FDA’s direction. GSK removed information regarding the
restricted-distribution program from the label and information
regarding the risk of myocardial ischemic events from the
black-box warning only. The revised label, however,
continued to warn physicians and patients elsewhere on the
label that “[i]n a meta-analysis of 52 double-blind,
randomized, controlled clinical trials . . . , a statistically
significant increased risk of myocardial infarction with
Avandia versus pooled comparators was observed”—this
information simply was no longer included in the black-box
warning, but this warning nonetheless appeared elsewhere on
the label. Id. at 829. Avandia’s label continued to include a
black-box warning that (i) advised physicians and patients that
Avandia “cause[s] or exacerbate[s] congestive heart failure in
some patients,” (ii) instructed physicians to “observe patients
[taking Avandia] carefully for signs and symptoms of heart
failure,” and (iii) warned patients with certain heart conditions
not to take Avandia. Id. at 825. These warnings remain on
Avandia’s label to this day.

10
 II.

The Plans brought suit alleging that GSK falsely
marketed Avandia and concealed data with respect to its
potential cardiovascular risks and side effects, thereby
violating RICO and various state consumer-protection laws.
The Plans assert that they would not have placed Avandia on
their formularies if GSK had disclosed the cardiovascular risks
that are in fact associated with Avandia. In other words, the
Plans would not have covered the cost of Avandia, which was
considerably more expensive than alternatives, if they had
known that Avandia not only did not reduce cardiovascular risk
in type-2 diabetes patients but also increased cardiovascular
risk as compared to those alternatives.

The Plans first filed suit in May 2010, and their cases
subsequently were consolidated in a multi-district litigation
case, which also included consumer and personal-injury suits
filed by other plaintiffs. In November 2010, GSK filed a
motion to dismiss the Plans’ complaints, arguing that the Plans
lacked standing to bring claims under RICO. In October 2013,
the District Court denied GSK’s motion, and, in October 2015,
we affirmed the decision of the District Court on an
interlocutory appeal. See In re Avandia Mktg., Sales Practices
& Prod. Liab. Litig. (Avandia I), 804 F.3d 633, 646 (3d Cir.
2015).

In May 2016, GSK filed a motion for summary
judgment. It argued that it was entitled to summary judgment
because, among other things, the Plans’ state-law consumer-
protection claims were preempted by the FDCA and the Plans
had failed to identify a distinct “enterprise” for purposes of
RICO. The Plans opposed the motion.

11
 In December 2017, the District Court granted summary
judgment in favor of GSK. First, the District Court refused to
consider the Plans’ arguments that GSK falsely marketed
Avandia as providing cardiovascular benefits in comparison to
alternatives because such arguments were “belated.” Unsealed
App. 4. The District Court noted that the Plans “seemed to
[have] shift[ed] their allegations to focus on Avandia’s
benefits, rather than the risks,” and stated that it only would
“address GSK’s motion for summary judgment as to [the
Plans’] state law claims on cardiovascular risk.” Id. at 3–4. It
stated that it would not “entertain” any of the Plans’ “benefits”
arguments “at th[at] juncture” due to their “belated” nature. Id.
at 4.

Second, the District Court found that the Plans’ state-
law consumer-protection claims were preempted by the FDCA
under the doctrine of “impossibility” preemption. It found that
three separate facts established “clear evidence” that the FDA
would not have approved a change to Avandia’s label with
respect to cardiovascular risks: (a) “the FDA rejected GSK’s
[Prior Approval Supplement],” (b) “the FDA advised against
using the CBE process to unilaterally change the label,” and (c)
“the FDA ultimately concluded that there was no increased
cardiovascular risk with Avandia use in relation to
comparators.” Id. at 24. With respect to the Prior Approval
Supplement, the District Court found that the “rejection of
GSK’s proposed label on the basis of inconclusive data,
considered with other evidence, constitutes clear evidence that
the FDA would not have approved the label change . . . ,
particularly where . . . the FDA wanted to conduct further
review of the data.” Id. at 24–25. Regarding the FDA’s
advising against using the CBE process, the District Court
found that an FDA representative’s statements—that she

12
“strongly advised” against using the CBE process and that
initializing that process would be “looked on with suspicion”
and would “pull the rug out” from the FDA’s then-current
plans for reviewing Avandia’s label—“shows that the FDA
advised against using [the] CBE [process] to make the
proposed label change prior to November 2007.” Id. at 25.
Finally, the District Court placed an emphasis on the FDA’s
“remov[al of] the black[-]box warning and restricted[-]access
information from Avandia’s label,” as well as the FDA’s
“current conclusion that a link between Avandia use and
increased cardiovascular risk does not exist.” Id. at 26. In
summary, the District Court found that the “FDA would not
have approved of a warning for increased cardiovascular risk
in Avandia versus competitors earlier than 2007 . . . and would
not approve one now.” Id.

Third, the District Court concluded that the Plans failed
to identify an “enterprise” that satisfies the “distinctiveness”
requirement of RICO. Specifically, it determined that “GSK
was conducting its own business in selling Avandia, and thus .
. . GSK is both the person and the enterprise.” Id. at 16.
Because “RICO liability ‘depends on showing that the
defendants conducted or participated in the conduct of the
enterprise’s affairs, not just their own affairs,’” the District
Court found that the Plans had not adequately alleged that an
“enterprise” existed because they merely alleged that the
“enterprise” in this case consisted of “GSK and its agents.” Id.
(emphasis in original) (quoting Reeves v. Ernst & Young, 507
U.S. 170, 185 (1993)).

The Plans timely appealed. They also appealed two
orders of the District Court that maintained the vast majority
of the summary-judgement record under seal. We considered
that appeal in In re Avandia Mktg., Sales Practices and Prods.

13
Liab. Litig. (Avandia II), 924 F.3d 662, 680 (3d Cir. 2019), in
which we vacated the District Court’s sealing orders.

III.

The District Court had jurisdiction pursuant to 28
U.S.C. §§ 1331 and 1332(d), and we have jurisdiction under
28 U.S.C. § 1291. We exercise plenary review over a district
court’s grant of summary judgment. Reedy v. Evanson, 615
F.3d 197, 210 (3d Cir. 2010). When a district court grants
summary judgment without considering a declaration filed by
the nonmoving party under Federal Rule of Civil Procedure
56(d), however, we review for abuse of discretion the district
court’s decision to disregard the Rule 56(d) declaration.
Shelton v. Bledsoe, 775 F.3d 554, 568 (3d Cir. 2015).

A.

With the benefit of the Supreme Court’s recent
guidance in Merck, which was decided following oral
argument in this case and well after the District Court’s
issuance of its memorandum opinion, 1 we hold that the Plans’
state-law consumer-protection claims are not preempted by the
FCDA, and we therefore will reverse the District Court’s order
granting summary judgment in favor of GSK on such claims.

In Wyeth v. Levine, 555 U.S. 555, 570–71 (2009), the
Supreme Court recognized that “it has remained a central
premise of federal drug regulation that the manufacturer [of a
pharmaceutical] bears responsibility for the content of its label

1
We subsequently ordered the parties to submit supplemental
letter briefs discussing Merck’s effect, if any, on the disposition
of this case.

14
at all times” and that the manufacturer “is charged both with
crafting an adequate label and with ensuring that its warnings
remain adequate as long as the drug is on the market.” Thus,
when it “bec[o]me[s] apparent” that a drug poses a certain risk
to the health and safety of persons taking it, the manufacturer
of the drug “ha[s] a duty to provide a warning that adequately
describe[s] that risk.” Id. at 571. The manufacturer may warn
persons of that risk by altering the drug’s label through the
CBE process, which “permit[s] it to provide such a warning
before receiving the FDA’s approval.” Id.

Under the FDCA, however, the FDA “retains authority
to reject labeling changes made pursuant to the CBE regulation
in its review of the manufacturer’s supplemental application,
just as it retains such authority in reviewing all supplemental
applications.” Id. Therein lies the conflict that may give rise
to impossibility preemption: even though a drug manufacturer
has the responsibility under state consumer-protection laws to
accurately label a drug and may change the label pursuant to
the CBE process prior to receiving approval from the FDA, it
may reject a label change at any time if it considers the drug to
be “mislabeled” under the FDCA. Thus, a situation may occur
in which a drug company seeks to change its label to add a
warning that it believes is required by state consumer-
protection laws, but the FDA considers the drug “mislabeled”
under the FDCA in light of the new warning that was added to
the label. In that situation, it would be impossible to comply
with both state and federal law. In resolving this conflict, the
Supreme Court struck a balance in Wyeth, holding that the
FDCA does not preempt state-law consumer-protection claims
regarding the labeling of a drug “absent clear evidence that the
FDA would not have approved a change to [the drug]’s label.”
Id.

15
 After we indicated in In re Fosamax (Alendronate
Sodium) Prods. Liab. Litig., 852 F.3d 268, 284 (3d Cir. 2017),
vacated, Merck, 139 S. Ct. 1668, that it would be helpful for
the Supreme Court to “clarif[y] or buil[d] out the doctrine”
espoused in Wyeth, the Supreme Court provided such
interpretive guidance in Merck. “[C]lear evidence,” as used in
Wyeth’s core holding, means “evidence that shows the court
that the drug manufacturer fully informed the FDA of the
justifications for the warning required by state law and that the
FDA, in turn, informed the drug manufacturer that the FDA
would not approve a change to the drug’s label to include that
warning.” 139 S. Ct. at 1672. Thus, to “show[] that federal
law prohibited [a] drug manufacturer from adding a warning
that would satisfy state law,” the drug manufacturer must
demonstrate that (1) “it fully informed the FDA of the
justifications for the warning required by state law” and (2)
“the FDA, in turn, informed the drug manufacturer that the
FDA would not approve changing the drug’s label to include
that warning.” Id. at 1678.

GSK has failed to satisfy either prong of Merck’s two-
prong test, and it therefore is not “entitled to judgment as a
matter of law.” Fed. R. Civ. P. 56(a). First, GSK has not
shown that “it fully informed the FDA of the justifications for
the warning required by state law.” Merck, 139 S. Ct. at 1678.
In the Letter, the FDA itself stated that it had “reviewed the
data provided [by GSK] and f[ou]nd [that] the information
presented is inadequate.” Sealed App. 660 (emphasis added).
Further, the FDA indicated that GSK needed to submit various
data and information “in order to address the deficiency of this
application.” Id. at 661. Thus, GSK cannot demonstrate that
the FDA was “fully informed . . . of the justifications for the
warning,” Merck, 139 S. Ct. at 1678, because the FDA itself

16
stated that it was “inadequate[ly]” informed of the
justifications for the warning, Sealed App. 660.

GSK argues that it “fully informed” the FDA because
GSK (1) provided all “material” information to the FDA and
(2) did not have access to the information that the FDA
requested until after the latter issued the Letter, but these
arguments are unavailing. GSK concedes that the FDA
requested additional data and information in the Letter, yet
GSK argues that none of the data and information that the FDA
actually requested in the Letter was “material” to its proposed
warning on cardiac risk, and that therefore, the FDA was “fully
informed” for purposes of Merck. This argument turns the
regulatory regime on its head. The FDA, not GSK, is the entity
with power to approve or refuse a change to a drug’s label, and
in making such a decision, it has the statutory authority to
conclude that the data and tests submitted by a manufacturer
were not “adequate” or that there is “insufficient information
about the drug to determine whether the product is safe for use
under the conditions prescribed, recommended, or suggested in
its proposed labeling.” 21 C.F.R. §§ 314.125(b)(2), (4). GSK
is not the arbiter of which data and information is or is not
“material” to the FDA’s decision to approve or reject a change
to a drug’s label—the FDA, and only the FDA, can determine
what information is “material” to its own decision to approve
or reject a labelling change.

Additionally, by arguing that it did not have access to
the FDA’s requested data and information until after the FDA’s
issuance of the Letter, GSK undermines its own argument that
the FDA was “fully informed.” Merck noted that “a drug
manufacturer will not ordinarily be able to show that there is
an actual conflict between state and federal law such that it was
impossible to comply with both.” 139 S. Ct. at 1679. Thus we

17
read Merck as holding that, in order to prove impossibility
preemption, the drug manufacturer must show that the “FDA
would not approve changing the drug’s label” and that the FDA
was “fully informed . . . of the justifications for the [proposed]
warning” at the time that the FDA rejected the proposed
warning. Id. at 1678. In other words, the upshot of Merck is
that a drug manufacturer must show that the FDA made a fully
informed decision to reject a change to a drug’s label in order
to establish the “demanding defense” of impossibility
preemption. Id. at 1678. If the question of whether the FDA
was “fully informed” was not tethered in time to the question
of whether the FDA indeed rejected the proposed warning, the
“fully informed” prong of the test espoused in Merck would be
rendered superfluous.

Thus, if GSK wishes to rely on the Letter as proof that
the FDA rejected its proposed label change, it must also
demonstrate that the FDA possessed all the information it
deemed necessary to decide whether to approve or reject the
proposed warning at the time it issued the Letter. By arguing
that it did not have the FDA’s requested data and information
until after the FDA issued its letter, however, GSK is, in effect,
conceding that the FDA was not “fully informed” at the time
of the Letter’s issuance. For that reason, among the others
outlined above, GSK cannot satisfy the first prong of the test
espoused in Merck.

Second, GSK cannot show that the “FDA . . . informed
[it] that the FDA would not approve changing the drug’s label
to include [the relevant] warning.” Id. at 1678. GSK directs
the Court’s attention to the Letter as proof that the FDA
rejected the proposed warning. The Letter indeed stated that
GSK’s Prior Approval Supplement for a label change was “not
approvable,” but the FDA indicated that this was so because

18
the “information presented [by GSK wa]s inadequate.” Sealed
App. 660. The FDA then required GSK to “amend the
supplemental application,” stating that “[a]ny amendment
should respond to all the deficiencies listed” in the Letter. Id.
at 661 (emphasis added). Thus, it is clear from the very text of
the Letter that the FDA did not consider GSK’s Prior Approval
Supplement “not approvable” because it was unconvinced of
the need for a strong warning on myocardial ischemic events;
rather, the FDA considered the Prior Approval Supplement
“not approvable” because it contained various “deficiencies”
that the FDA required GSK to ameliorate prior to the FDA’s
making a final determination. At most, the Letter indicates that
it is possible that the FDA could have rejected the label change
after receiving the various data and information it requested
from GSK, but as the Supreme Court has reiterated, the
“possibility of impossibility [is] not enough.” Merck, 139 S.
Ct. at 1678 (alteration in original) (quoting PLIVA, Inc. v.
Mensing, 564 U.S. 604, 625 n.8 (2011)). We nevertheless need
not speculate regarding the possibility that the FDA would
have rejected the proposed warning upon the receipt of the
requested data and information because it indeed ordered GSK
to include various warnings regarding cardiac risks on
Avandia’s label shorty after issuing the Letter, which alone
undermines GSK’s position that the Letter represents a
rejection of its proposed warning.

Finally, we are not persuaded by any of GSK’s
arguments that the Plans’ claims are preempted because GSK
allegedly was unable to avail itself of the CBE process for
various reasons. GSK primarily argues that it could not use the
CBE process to introduce a warning on ischemic risks prior to
mid-2006, when it submitted its Prior Approval Supplement.
GSK reasons that ICT-42 served as the basis for its belief that

19
an ischemic-risk warning should be included on the label, and
because that study was completed in mid-2006, it did not have
the “newly acquired information” necessary to make a labeling
change prior to that time. This argument, however, is
undermined by GSK’s own admissions. For example, GSK
itself described the results of “ICT-37,” a meta-analysis
completed a year earlier in August 2005, as “generally similar”
to ICT-42, and GSK stated that “[a]ny numerical differences
[between the meta-analyses] were not clinically significant.”
Sealed App. 861. Thus, at the very least, it appears that GSK
could have used the CBE process to add an ischemic-risk
warning as early as August 2005 because, by GSK’s own
admission, ICT-37 and ICT-42 indicated similar results and
had clinically insignificant numerical differences. 2 Further,
GSK cannot rely on its informal phone conversations with an
FDA official to claim that it could not pursue a label change
through the CBE process, nor can GSK rely on the stock
language at the end of the Letter, which advised GSK that
Avandia “may be considered to be misbranded under the
[FDCA] if it is marketed with the[ proposed] changes before
approval of this supplemental application.” Id. at 661. An
informal phone conversation with an FDA official is not an
“agency action taken pursuant to the FDA’s congressionally
delegated authority,” Merck, 139 S. Ct. at 1679, and the stock
language at the end of the Letter is a simple statement of the
law: if a manufacturer makes a label change pursuant to the
CBE process (i.e., without seeking the prior approval of the
FDA), the manufacturer always runs the risk that the FDA will

2
We take no position with respect to whether GSK could have
used the CBE process, or otherwise sought to change
Avandia’s label, to add an ischemic-risk warning prior to
August 2005.

20
later reject the label change and consider the drug as
“mislabeled,” see 21 C.F.R. § 314.70(c)(7). Finally, GSK’s
argument that it could not implement a black-box warning
through the CBE process is a red herring—the Plans are not
arguing that GSK should have added the black box itself
through the CBE process, but rather that GSK should have
added the content of the black-box warning anywhere on the
label.

GSK thus has failed to demonstrate that the Plans’ state-
law consumer-protection claims are preempted by the FDCA,
and GSK therefore is not entitled to summary judgment on
those grounds. Therefore, we will reverse the order of the
District Court granting summary judgment in favor of GSK on
the Plans’ state-law consumer-protection claims.

B.

The District Court erred in granting summary judgment
on the Plans’ RICO claims without giving the Plans the benefit
of discovery on those claims.

“[A] Court ‘is obligated to give a party opposing
summary judgment an adequate opportunity to obtain
discovery.’” Doe v. Abington Friends Sch., 480 F.3d 252, 257
(3d Cir. 2007) (quoting Dowling v. City of Philadelphia, 855
F.2d 136, 139 (3d Cir. 1988)). “If discovery is incomplete, a
district court is rarely justified in granting summary judgment,
unless the discovery request pertains to facts that are not
material to the moving party’s entitlement to judgment as a
matter of law.” Shelton, 775 F.3d at 568.

Rule 56(d) provides that “[i]f a nonmovant shows by
affidavit or declaration that, for specified reasons, it cannot

21
present facts essential to justify its opposition, the court may:
(1) defer considering the motion or deny it; (2) allow time to
obtain affidavits or declarations or to take discovery; or (3)
issue any other appropriate order.” Fed. R. Civ. P. 56(d).
“[D]istrict courts usually grant properly filed requests for
discovery under Rule 56(d) ‘as a matter of course’ . . . .”
Shelton, 775 F.3d at 568 (quoting Murphy v. Millennium Radio
Grp. LLC, 650 F.3d 295, 309–10 (3d Cir. 2011)). “This is
particularly true when there are discovery requests outstanding
or where relevant facts are under control of the party moving
for summary judgment.” Id. A district court abuses its
discretion when it grants summary judgment in favor of the
moving party “without even considering” a Rule 56(d)
declaration filed by the nonmoving party. See id.

The Plans never received discovery related to their
RICO claims, including with respect to whether an “enterprise”
existed for purposes of RICO, and thus when GSK moved for
summary judgment on the Plans’ RICO claims, the Plans
submitted a detailed Rule 56(d) declaration regarding the lack
of discovery on the issues related to RICO. See J. App. 2195–
2198. They subsequently filed a supplemental Rule 56(d)
declaration, further elaborating on their need for discovery on
RICO-related issues. See id. at 2272–76.

The District Court granted summary judgment in favor
of GSK on the Plans’ RICO claims without considering their
Rule 56(d) declaration and their supplemental Rule 56(d)
declaration. This was an abuse of discretion, especially as the
District Court granted summary judgment on the ground that
the Plans could not prove the existence of an “enterprise,”
information related to which is “under control of the party

22
moving for summary judgment”—in this case, GSK. 3 Shelton,
775 F.3d at 568. We therefore vacate the District Court’s order
granting summary judgment in favor of GSK on the Plans’
RICO claims, and we remand to the District Court to give
proper consideration to the Plans’ Rule 56(d) declarations.

IV.

Finally, we note that, on remand, the District Court must
consider the Plans’ arguments that GSK marketed Avandia as
providing cardiovascular benefits. These arguments and
claims are not “belated”; the Plans have pursued this line of
argument since the outset of this litigation. In the Plans’
complaint itself, the Plans alleged that they “rel[ied] upon
[GSK]’s promises of superior treatment and better
cardiovascular outcomes compared with the older diabetes
drugs” in determining that it was worth the increased cost to
cover Avandia. J. App. 1273. They alleged that “better
cardiovascular outcomes” were a crucial part of GSK’s alleged
fraudulent marketing: “[t]he notion that Avandia would
actually lower diabetics’ cardiovascular risk was critical to
Avandia’s marketing” because GSK “needed justification for
the steep price difference between Avandia and the older
established diabetes drugs.” Id. at 1291. While a portion of
the Plans’ claims center on the assertion that GSK should have
disclosed on its label the true nature of the increased
cardiovascular risk that was presented by Avandia as compared
to cheaper alternatives, the increased risk is only relevant to the

3
We refuse to construe the District Court’s grant of summary
judgment in favor of GSK on the Plans’ RICO claims as a
dismissal on the pleadings pursuant to Rule 12(c), particularly
because the Plans’ RICO claims previously survived a Rule
12(b)(6) motion to dismiss. See Avandia I, 804 F.3d at 646.

23
Plans’ claims insofar as the Plans make the following
argument: GSK failed to warn of Avandia’s true
cardiovascular risk, and thus, GSK was continuing—by
omission—to promote Avandia as capable of lowering
patients’ cardiovascular risk, and GSK thereby continued to
induce the Plans to cover the cost of Avandia based on this
perceived “benefit” of lowering cardiovascular risk. Id. at
1316. In short, the Plans have never argued that GSK
promoted Avandia as capable of actually improving patients’
cardiovascular health, but rather as capable of lowering
cardiovascular risk when compared to cheaper alternatives,
which indeed is a “benefit.”

Because the Plans have raised, throughout these
proceedings, arguments that GSK marketed Avandia as
providing cardiovascular benefits, it was error for the District
Court to refuse to consider those arguments. See, e.g., Hillman
v. Resolution Tr. Corp., 66 F.3d 141, 144 (7th Cir. 1995).
Therefore, on remand, the District Court needs to give proper
consideration to these arguments.

V.

For the reasons stated above, we will reverse the order
of the District Court granting summary judgment in favor of
GSK on the Plans’ state-law consumer-protection claims,
vacate the order of the District Court granting summary
judgment in favor of GSK on the Plans’ RICO claims, and
remand to it for proceedings consistent with this opinion. On
remand, the District Court shall give proper consideration to
the Plans’ Rule 56(d) declarations, as well as their arguments
that GSK marketed Avandia as providing cardiovascular
benefits.

24