Torkie-Tork v. Wyeth

Filing 235

MEMORANDUM OPINION re 97 MOTION for Summary Judgment by Wyeth. Signed by District Judge T. S. Ellis, III on 10/4/2010. (tche)

Torkie-Tork v. Wyeth

Doc. 235

IN T H E U N I T E D S T A T E S D I S T R I C T C O U R T
FOR THE EASTERN DISTRICT OF VIRGINL Alexandria Division

OCT " 4 2010
CLERK. U.S. DISTRICT COURT
ALEXANDRIA. VIRGINIA

GEORGIA TORKIE-TORK,

Plaintiff,
No. I:04cv945

WYETH,
Defendant.

MEMORANDUM OPINION

In this removed diversity product liability action, plaintiff Georgia Torkie-Tork sues defendant Wyeth for compensatory and punitive damages, alleging that Prempro, a drug

manufactured and sold by Wyeth, caused her to suffer breast cancer. At issue following
discovery is whether summary judgment is appropriate in favor o f Wyeth on p l a i n t i f f s claims for negligent design defect and fraud. Wyeth argues that no genuine issue of material fact exists

on either claim because, inter alia, (i) plaintifi'has not adduced evidence of an alternative design
for Prempro that would have avoided p l a i n t i f f s breast cancer, and (ii) the summary judgment
record does not support a finding o f fraud in the Prempro label.

For the reasons that follow, summary judgment in Wyeth's favor is appropriate with

respect to part of the fraud claim, but summary judgment must be denied as to the remaining
portion o f the fraud claim and as to the claim for negligent design defect.

Wyeth moved for summary judgment on all claims except the claim for negligent failure to warn about the risks of breast cancer. Def. Br. at 1-2. Plaintiff conceded that summary judgment is appropriate on the claims for (i) strict liability for failure to warn, (ii) strict liability for design defect, (iii) negligent misrepresentation, and (iv) breach o f express warranty. PI. Br. at 16. Plaintiff opposes summary judgment on the two remaining claims, namely negligent
design d e f e c t and fraud.

Dockets.Justia.com

I.2

Plaintiff Georgia Torkie-Tork is a citizen o f Virginia. Defendant Wyeth is a Delaware corporation with its principal place o f business in New Jersey. During times relevant to this

litigation, defendant was one of the world's largest pharmaceutical companies3 and the maker of
Prempro, a hormone therapy drug approved by the Food and Drug Administration ("FDA") that contains a combination o f estrogen and a progestin ("E+P"), and is indicated for treatment o f menopausal symptoms. Beginning in or about 1996, plaintiff began experiencing severe menopausal symptoms. Her then-physician, Dr. Joel Schulman, prescribed Prempro for treatment o f those symptoms.
Between 1996 and 2002, several doctors filled out Prempro prescriptions for plaintiff, including Dr. William Hurwitz. Although he d o e s not specifically r e m e m b e r prescribing Prempro for

plaintiff, Dr. Hurwitz has stated that it is his general practice to read and rely upon the warning

labels on any drug before prescribing it to a patient.
Between 1997 and 2002, the Prempro label contained the following statements: Some studies have reported a moderately increased risk o f breast cancer (relative risk o f 1.3 to 2.0) in those w o m e n on estrogen replacement therapy taking higher doses, or in those taking lower doses for prolonged periods o f time, especially in excess o f 10 years. The majority o f studies, however, have not s h o w n an association in women who have ever used estrogen replacement therapy. The effect o f added progestins on the risk o f breast cancer is unknown, although a moderately increased risk in taking combination estrogen/progestin therapy has been reported. O t h e r studies have not shown this relationship. In a o n e - y e a r
c l i n i c a l trial o f P R E M P R O , P R E M P H A S E a n d P r e m a r i n a l o n e , 5 n e w c a s e s o f breast c a n c e r w e r e d e t e c t e d a m o n g 1377 w o m e n who r e c e i v e d the c o m b i n a t i o n treatments, w h i l e no n e w c a s e s were detected a m o n g 347 w o m e n who received

2The facts recited herein are derived from the pleadings and the record taken as a whole, and are
not m a t e r i a l l y d i s p u t e d e x c e p t w h e r e s p e c i f i c a l l y noted. W h e r e s u c h d i s p u t e s are n o t e d , t h e analysis p r o c e e d s by a s s u m i n g p l a i n t i f f s claim o f fact. See Estate o f Cloaninger v. McDevitt, 555 F.3d 324, 332 (4th Cir. 2 0 0 9 ) (citing Scott v. Harris, 550 U.S. 3 7 2 , 377 (2007)).

3 Pfizer Inc. purchased Wyeth at some point since the filing of this action.

Premarin a l o n e . T h e o v e r a l l i n c i d e n c e o f b r e a s t c a n c e r i n t h i s c l i n i c a l t r i a l d o e s

not exceed that expected in the general population.

In the three y e a r clinical Postmenopausal Estrogen Progestin Intervention (PEPI) trial o f 875 women to assess differences among placebo, unopposed Premarin, and three different c o m b i n a t i o n h o r m o n e therapy regimens, o n e (1) n e w c a s e o f breast caner was detected in the placebo group (n=174), one in the Premarin alone group (n=175), none in the continuous Premarin plus continuous medroxyprogesterone acetate group (n=174) and two (2) in the continuous Premarin plus cyclic medroxyprogesterone acetate group (n=174). On August 7 , 2 0 0 0 , the FDA wrote Wyeth to request that certain changes be made to the

label for E+P hormone therapy drugs. See FDA Letter to Wyeth (Aug. 7, 2000). Specifically,
the F D A ' s proposed changes included the following statements:
While some e p i d e m i o l o g i c s t u d i e s s u g g e s t a very m o d e s t increase in b r e a s t c a n c e r

risk for estrogen alone users versus non-users, other studies have not shown any increased risk. The addition o f progestin to estrogen may increase the risk for breast cancer over that noted in non-hormone users more significantly (by about 24-40%), although this is based solely on epidemiologic studies, and definitive conclusions await prospective controlled clinical trials.

Studies examining the risk o f breast cancer among women using estrogen alone and combined estrogen/progestin therapy have suggested that there may be a mildly increased risk o f breast cancer in women taking the combined therapy. After receiving this letter, W y e t h ' s counsel responded to the FDA, noting that the FDA did not

have the power to "dictate proposed language for an applicant labeling without providing a
meaningful opportunity for dialogue between the applicant and the agency." Arnold & Porter
Letter to FDA (Nov. 7, 2000). Wyeth also proposed alternative label revisions for the F D A ' s

review with explanations for the areas o f disagreement. Wyeth Letter to FDA (Aug. 11,2000).
For example, Wyeth stated:

We strongly disagree with the presentation o f the risk attributable to progestin use as 24-40%. First, we believe it is questionable for increases in risk to be stated only in percentages because this tends to exaggerate risk, particularly when

absolute risk is small. Secondly, w h e n stated o n l y in this manner, the information is easily misinterpreted, i.e., one may interpret that 24-40% o f all HRT users will develop breast cancer, a clearly inappropriate conclusion.

Wyeth then proposed the following alternative language:

Epidemiological studies suggest that the addition o f progestin to estrogen therapy may enhance [the risk o f breast cancer over estrogen-only therapy]. Definitive conclusions await prospective controlled clinical trials.

This dialogue between the FDA and Wyeth concerning possible Prempro label changes
continued until March 2001, at which time the FDA approved final revisions to the Prempro label. PI. Ex. 25. Despite this approval in March 2001, Wyeth did not implement changes to its
label until after a new study was released in July 2002 by the W o m e n ' s Health Initiative

("WHI"). The WHI study showed a statistically significant link between the use o f Prempro and
breast cancer. Following the release o f the WHI study, Wyeth updated the warnings on the

Prempro label. Plaintiff alleges that the label on which her doctor relied when prescribing her Prempro was the pre-2002 version o f the label, and Wyeth has conceded this reliance solely for
the purposes o f resolving the s u m m a r y j u d g m e n t motion.

The use o f Prempro proved effective for the treatment o f plaintiff s symptoms, and she
continued using the drug until June 2002, at which time an abnormality was noted on her annual

mammogram. At the direction o f Dr. Ronald Orleans, she immediately discontinued her use o f
Prempro, and a follow-up sonogram and needle biopsy were performed. Based on the results o f these procedures, plaintiff was diagnosed with breast cancer on June 18, 2002. Thereafter on

June 27,2002, she underwent a partial mastectomy to remove the cancerous tissue. A pathology report signed on July 3, 2002 confirmed that the cancer was hormone receptor positive, meaning
that the cancer was o f a type caused by hormones such as those contained in Prempro. A

surgical procedure on July 24, 2002 confirmed that the June 2 7 , 2 0 0 2 mastectomy had removed
all cancerous tissue. The cancer has not recurred.

Plaintiff filed the instant action in Virginia state court on July 2 , 2 0 0 4 , and it was

removed to this district on August 13, 2004. In her complaint, plaintiff alleges that Wyeth is
liable for the personal injury that she suffered--namely, breast cancer--as a result o f her
prescribed use o f Prempro. Because numerous suits o f this nature had been filed, the Judicial

Panel on Multidistrict Litigation convened multidistrict litigation ("MDL") proceedings in the

Eastern District of Arkansas, and this matter was transferred to that district for participation in
the MDL proceedings. See Torkie-Tork v. Wyeth, No. I:04cv945 (E.D. Va. Nov. 1, 2004)

(Conditional Transfer Order). At the conclusion of the MDL proceedings, by Order dated April
8,2010, the matter was returned to this district for all further proceedings, including casespecific discovery, summary judgment, and if necessary, a trial. See Torkie-Tork v. Wyeth, No.
I:04cv945 (E.D. Va. Apr. 8 , 2 0 1 0 ) (Conditional Remand Order).
II.

The summary judgment standard is too well-settled to require elaboration here. In
essence, summary judgment is appropriate under Rule 56, Fed. R. Civ. P., only where, on the

basis of undisputed material facts, the moving party is entitled to judgment as a matter of law.

Celotex Corp. v. Catrett, All U.S. 317, 322 (1986). Importantly, to defeat summary judgment
the non-moving party may not rest upon a "mere scintilla" o f evidence, but must set forth

specific facts showing a genuine issue for trial. Id. at 324; Anderson v. Liberty Lobby, Inc., A l l
U.S. 242,252 (1986). Thus, the party with the burden of proof on an issue cannot prevail at
summary judgment on that issue unless that party adduces evidence that would be sufficient, i f
believed, to carry the burden o f p r o o f on that issue at trial. See Celotex, A l l U.S. at 322.

III.

Only two o f p l a i n t i f f s claims are contested for the purposes o f this summary judgment

motion: (i) negligent design defect and (ii) fraud in the Prempro label.4 Wyeth argues that
p l a i n t i f f cannot establish certain elements o f these claims and hence neither claim survives
s u m m a r y j u d g m e n t . Additionally, W y e t h also argues t h a t b o t h c l a i m s fail for lack o f e v i d e n c e o f

causation.5 Each of these arguments is addressed separately.
A. Negligent Design Defect

To prevail on a negligent design defect claim under Virginia law,6 "the plaintiff must
prove that the product contained a defect which [sic] rendered it unreasonably dangerous for
ordinary or foreseeable use." Alevromagiros v. Hechinger Co., 993 F.2d 4 1 7 , 4 2 0 (4th Cir. 1993). Although not conclusive in determining what constitutes an unreasonably dangerous defect, "a court will consider safety standards promulgated by the government or the relevant
industry, as well as the reasonable e x p e c t a t i o n s o f c o n s u m e r s . " Id.

Wyeth contends that p l a i n t i f f s claim for negligent design defect fails for two reasons: (i)

Prempro is "safe and effective" in light o f the F D A ' s continued approval o f the drug and doctors'

continued prescription o f the drug in its original form; and (ii) plaintiff has not adduced any
evidence that an alternative design o f Prempro, whether a variance in the dosage or an alternative
formulation, w o u l d h a v e a v o i d e d p l a i n t i f f s b r e a s t cancer.

4See n.l supra. At oral argument, plaintiff conceded that the only viable claim for fraud was
fraud in the Prempro label.

5Wyeth further argues that all of the claims in plaintiffs complaint, including the claim for
n e g l i g e n t f a i l u r e to w a r n , s h o u l d be d i s m i s s e d for l a c k o f c a u s a t i o n .

6It has already been determined that Virginia law applies to this suit. See Torkie-Tork v. Wyeth,
No. I:04cv945 (E.D. Va. Jun. 1 6 , 2 0 1 0 ) (Mem. Op.).

As to W y e t h ' s first a r g u m e n t , p l a i n t i f f correctly notes, and W y e t h concedes, t h a t FDA

approval o f a drug does not preempt an action for defective design. PI. Br. at 17 (citing Wyeth v.
Levine, -- U.S. --, 129 S. Ct. 1187, 1 2 0 4 ( 2 0 0 9 ) ; H i l l v. Searle Labs, 8 8 4 F . 2 d 1064, 1068 ( 8 t h

Cir. 1989)); see also Def. Reply Br. at 19 ("Wyeth has never argued or implied that ' F D A
approval is s o m e h o w s a c r o s a n c t ' as to P l a i n t i f f s design d e f e c t c l a i m " ) . T h i s is so b e c a u s e

"FDA regulations are generally minimal standards o f conduct" absent a clear intent by Congress
to preempt state law, w h i c h has not o c c u r r e d in this area. S e e Hill, 884 F.2d at 1068. O f course,

the F D A ' s continued approval o f Prempro in its current form is strong evidence o f

reasonableness in the Prempro design, and a j u r y may well conclude from this fact that

Prempro's design is not defective. But this fact alone does not resolve the claim for the purposes
o f summary judgment.

Wyeth's second argument focuses on p l a i n t i f f s two alternative designs that would have decreased the risks o f breast cancer for plaintiff and those using Prempro. The first proposed alternative design is simply a change in the dosage o f the drug itself. The second proposed

alternative design would have Wyeth use oral micronized (natural) progesterone instead o f the

synthetic progestin currently in Prempro.7 Analysis of this argument must begin with
consideration o f the question whether the two proposed designs are in fact alternative designs for the purposes o f the negligent design claim. Important in this regard is that an alternative design must not be an altogether essentially different product. As has been often stated, "[a] motorcycle
could be made safer by adding two additional wheels and a cab, but then it is no longer a

7According to the Physician's Desk Reference, Prempro comes in several dosages, each of
which contains a mixture o f conjugated estrogens found in Premarin tablets--Premarin being derived from a mixture o f sodium estrone sulfate and sodium equilin sulfate--and medroxyprogesterone acetate for oral administration. Prempro, P h y s i c i a n ' s Desk Reference (2010). The conjugated estrogens used in Premarin and Prempro are blended to match the average composition o f material derived from pregnant m a r e s ' urine. Id.

motorcycle." Caterpillar, Inc. v. Shears, 911 S . W . 2 d 3 7 9 , 385 (Tex. 1995); s e e also Kimball v.

RJ Reynolds Tobacco Co., 2006 U.S. Dist. LEXIS 27138 (W.D. Wash. Apr. 2 6 , 2 0 0 6 ) (noting
that " ' [ f j w o - w h e e l e d n e s s ' is an essential characteristic o f a motorcycle"). Put another way, an alternative design is not reasonable i f it alters a fundamental and necessary characteristic o f the

product. This is, o f course, typically a question o f fact, not law. Kimball, 2006 U.S. Dist.
L E X I S 2 7 1 3 8 , at *8.

As to the first proposed alternative design, it may well be that the dosage o f a drug is a

fundamental characteristic o f the drug, since a lower dosage may well alter or affect the positive impact the drug is designed to have on the human body. In her brief, p l a i n t i f f offers little explanation for the costs and benefits o f a change in the dosage o f Prempro, i f such an analysis is
even feasible with the current science available. Nevertheless, the decision properly rests with a
j u r y to d e t e r m i n e w h e t h e r an a l t e r n a t i v e d o s a g e o f P r e m p r o w o u l d so fundamentally a l t e r the

drug as to render it an entirely different product. P l a i n t i f f s second proposed alternative design similarly presents an issue o f fact properly submitted to a jury. If Wyeth could have used a natural progesterone instead o f synthetic progestin and accomplished a similar positive therapeutic effect, a j u r y may reasonably decide that the refusal to employ such a design was negligent. On the other hand, Wyeth may martial evidence to show that this proposed alternative design would fundamentally alter Prempro, in which event a j u r y might reasonably conclude that such an alteration would result in a wholly different product--Prempro would no longer be

Prempro, much as a four-wheel vehicle with a cab would cease to be a motorcycle. In short, on
this issue--alternative design--the summary j u d g m e n t record presents a genuine issue o f fact for
trial.

It remains then to resolve Wyeth's argument that p l a i n t i f f s proposed alternative designs would not have prevented p l a i n t i f f s breast cancer. Wyeth argues that plaintiff has only identified "generic experts" to discuss diminished cancer risks from p l a i n t i f f s alternative Prempro designs, rather than "case specific experts" who will show how the alternative designs would have avoided cancer in this plaintiff. Def. Reply. Br. at 17. Wyeth's characterization o f the experts as "generic" is misleading and unhelpful; p l a i n t i f f s expert reports indicate that alternative designs to Prempro would present little or no

risk o f breast cancer to anyone, which, o f course, includes plaintiff. For example, Dr. Don
Austin, one o f p l a i n t i f f s expert witnesses, reviewed studies in this area and concluded "that

[E+P hormone therapy] containing [natural] micronized progesterone or dydrogesterone has no
elevated risk, in contrast to [E+P hormone therapy] containing [medroxyprogesterone acetate],"

the synthetic form o f progesterone. See PI. Ex. 67, at 2 0 , 2 7 (Report o f Dr. Don Austin). Where an alternative drug design would nearly eliminate the overall risk o f cancer, it follows a fortiori

that it would also diminish that risk in the p l a i n t i f f s specific case. Wyeth may dispute Dr.
A u s t i n ' s conclusion, but viewing the record in a light most favorable to the plaintiff, which is
appropriate at this stage, a g e n u i n e i s s u e o f material fact r e m a i n s on the c a u s a t i o n e l e m e n t o f this

claim. Accordingly, summary judgment is not appropriate for the negligent design defect claim.8

o

This result is consistent with the decision in Scroggin v. Wyeth denying Wyeth summary j u d g m e n t on the negligent design d e f e c t claims under Arkansas law. 2008 U.S. Dist. LEXIS

24027 (E.D. Ark. Jan. 14,2008) (finding a genuine dispute o f fact as to whether Prempro, "in the dosage consumed by Plaintiff, was in a defective condition and unreasonably dangerous"). On the other hand, the analysis in Brockert v. Wyeth is distinguishable from the present case. 287
S . W . 3 d 760, 769 (Tex. App. Ct. 2009). There, a state a p p e l l a t e c o u r t in T e x a s upheld the grant

o f summary judgment to Wyeth on a negligent design defect claim. The plaintiff in Brockert contended that the safer alternative design o f Prempro would have been estrogen alone, and the Texas court recognized that such a design would essentially mean that "Prempro should have been a different product[,] [namely] its predecessor[,] Premarin." Id. at 769-71. Here, plaintiff argues that an alternative dosage formulation or a substitution o f progestin with its natural

B.

Fraud in the Label

It is well-settled in Virginia that two elements are essential to a fraud claim: (i) a knowing

misrepresentation or concealment o f material fact, and (ii) reasonable and detrimental reliance on that misrepresentation or concealment. See Allen Realty Corp. v. Holbert, 318 S.E.2d 592, 597
(Va. 1984); Packard Norfolk, Inc. v. Miller, 95 S.E.2d 207, 210 (Va. 1956). P l a i n t i f f s fraud

claim rests exclusively on the Prempro label, which plaintiff argues both misstates and conceals

material facts about the cancer risks associated with Prempro.9 To succeed on her fraudulent
c o n c e a l m e n t allegation, p l a i n t i f f m u s t s h o w that W y e t h " c o n c e a l [ e d ] a fact that is material to the

transaction, knowing that the [plaintiff or her doctors are] acting on the assumption that no such
fact e x i s t s . " Clay v. Butler, 112 S.E. 697, 700 (Va. 1922). U n d e r s u c h c i r c u m s t a n c e s , " t h e

concealment is as much a fraud as i f the existence o f the fact w e r e expressly denied." Id. The label contains four distinct statements, e a c h o f w h i c h must be separately analyzed for any potential fraudulent content.
i. Statement No. 1

S o m e studies h a v e reported a moderately increased risk o f b r e a s t c a n c e r (relative risk o f 1.3 to 2.0) in those w o m e n on estrogen replacement therapy taking higher doses, or in those taking lower doses for prolonged periods o f time, especially in excess o f 10 years.

counterpart would have been safer; whether such changes w o u l d fundamentally transform Prempro into a completely different product is a genuine issue o f fact appropriate for j u r y
resolution.

9At oral argument, plaintiff conceded that her fraud claim rests solely on allegedly fraudulent
statements or concealments in the Prempro label, thus abandoning any other allegations o f fraud contained in her complaint, such as fraud in W y e t h ' s marketing material. Although Wyeth disputes whether p l a i n t i f f s doctors actually relied on the label, Wyeth concedes that a genuine dispute o f fact exists on the r e l i a n c e issue, at least with r e s p e c t to Dr. Hurwitz. As such, the central question in the s u m m a r y j u d g m e n t analysis o f the fraud claim is whether the label
included a n y fraudulent statements.

10

Plaintiff c o n t e n d s t h a t t h i s s t a t e m e n t w a s f r a u d u l e n t b e c a u s e , a t t h e t i m e t h e s t a t e m e n t

was made,10 Wyeth knew that a risk of breast cancer existed for all doses of E+P, not just higher
doses. But as Wyeth correctly points out, Statement No. 1 deals only with "estrogen

replacement therapy," not E+P. The parties do not contest that "estrogen replacement therapy" is
understood to mean estrogen-only therapy, and thus the term does not include estrogen-progestin

(E+P) combination therapy. Furthermore, the parties do not dispute the truth o f the statement with respect to estrogen-only therapy. Accordingly, Statement No. 1, by itself, was not
fraudulent. ii. Statement No. 2

The majority o f studies, however, have not shown an association [with breast cancer] in women who have ever used estrogen replacement therapy.
Plaintiff argues that Statement No. 2 was false at the t i m e it was made because a majority

o f pre-WHI studies--32 out o f 43--in fact showed a risk o f breast cancer. Wyeth protests that
p l a i n t i f f s tally o f the studies is incorrect because it includes studies showing a statistically-

nonsignificant increase in the risk o f cancer from estrogen-related therapies. By defendant's count, only 18 o f the 43 pre-WHI studies showed a statistically significant risk, making
Statement No. 2 an accurate reflection o f then-existing studies.

Defendant's emphasis on statistical significance is entirely appropriate. The concept o f
statistical significance is critical to a p r o p e r u n d e r s t a n d i n g o f statistical analysis. M a n y s t u d i e s

provide their results in the form o f a confidence interval, which is " a range o f values calculated

from the results o f a study, within which the true value is likely to fall." Federal Judiciary

10 For the purposes of the summary judgment analysis, the statements on the Prempro label are
evaluated for falsity d u r i n g the p e r i o d that p l a i n t i f f s p h y s i c i a n a l l e g e d l y relied on the label in prescribing the drug to plaintiff, namely the period from 1996 to mid-2002, j u s t prior to the publication o f the WHI study.

11

Center, Reference Manual on Scientific Evidence 360 (2d. ed. 2000). Confidence intervals are

commonly used in epidemiological studies, and they must be carefully understood before
drawing a conclusion from such studies.
The c a n c e r studies cited by the parties typically report the c o r r e l a t i o n b e t w e e n a d r u g

therapy and the incidence o f breast cancer as a numerical relative risk; a relative risk greater than
1.0 generally indicates a p o s i t i v e correlation between the use o f the d r u g and the incidence o f

cancer. Consider a study stating that the relative risk is 1.27, but where that value was contained in a 95% confidence interval bounded between 0.84 and 1.94. An equivalent way to state these
results would be to say that there is a 95% probability t h a t the relative risk value falls s o m e w h e r e

between 0.84 and 1.94. The range from 0.84 to 1.94 is called the confidence interval, and

because o f the nature o f confidence intervals, the true value is no more likely to be at the low end
o f this range than at the high end. Id. at 361. Since a range o f relative risk from 0.84 to 1.94 encompasses 1.0--the value at which there is no increase in the risk o f cancer associated with

the drug--then one must conclude that the results o f the study are not statistically significant.
N u m b e r s like these are not hypothetical; indeed, these are the n u m b e r s t a k e n from the 2 0 0 2

epidemiological study by Dr. Stephen Hulley and published in the Journal o f the American

Medical Association--a study relied upon by plaintiff. See Hulley et ah, Noncardiovascular
Disease Outcomes During 6.8 Years o f Hormone Therapy, 288 JAMA 58, 60 (2002) (PI. Ex. 71JJ). While plaintiff suggests that this study shows a risk o f breast cancer associated with

hormone therapy, defendant appropriately recognizes that the results o f this study were
inconclusive."

1' Statistical significance is a concept that is poorly understood in many contexts. For example, a
political poll may show that one candidate leads another candidate in the polls 51% to 49%, but the fine print might note that the poll had a statistical error rate o f plus-or-minus 2.5%. While
12

Of the 43 studies that the parties recognize as the universe o f relevant pre-WHI cancer
studies, p l a i n t i f f identified 32 s t u d i e s t h a t she believes s h o w e d a risk o f b r e a s t cancer. D e f e n d a n t

contends that o f these 32 studies, 14 do not show such a risk with statistical significance. Essentially, defendant claims that 14 o f the studies cited by plaintiff either showed no risk o f
breast cancer or showed a relative risk o f cancer within a confidence interval that includes 1.0,

the latter o f which would indicate an association that is not statistically significant. Each o f these
fifteen studies has been reviewed to verify defendant's claim, and the results o f the studies are
summarized in the table below:12
PI. Ex. #
Ex. 71G

Author/Publication
E w e r t z , I n t ' l J. C a n c e r

Year
1988

Statistical Results

RR 1.04 (95% CI 0.74-1.46) for pre menopausal women; RR 1.16 ( 9 5 % CI 0.64-2.11) for menopausal women; RR 1.28 (95% CI 0.96-1.71) for post menopausal women; RR 1.04 (95% CI

0.69-1.57) for artificial menopause13
Ex.9

Bergkvist, N. Engl. J. Med.

1989

RR 1.1 ( 9 5 % CI 1 . 0 - 1 . 3 ) o v e r a l l for
estrogen users

Ex. 7IH

Kaufman, Am. J. Epidem.

1991

RR 1.2 (95% CI 1.0-1.4) for estrogen unopposed by progestogens; RR 1.7 (95% CI 0.9-3.3) for estrogen opposed by
progestogens

Ex.711

Colditz, Cancer Causes & Control

1992

RR 1.08 ( 9 5 % CI 0 . 9 6 - 1 . 2 2 ) for ever-use o f postmenopausal hormones

some would assert that the poll shows one candidate to be leading the other by two percentage
points, the only proper conclusion is that the candidates are in a statistical dead heat, because a 50-50 split is within the range o f error. Arguments based on statistics must be evaluated carefully for such misleading statements.

12 In reciting the results of the studies, the terms "relative risk" and "confidence interval" are
abbreviated " R R " and " C I , " respectively.

13 These results reflect the risk of breast cancer associated with the use of non-contraceptive sex
hormones in the respective groups. As the study summarized, "Exposure to [estrogen] or [progestogen], alone or in combination-type therapy, did not affect the breast cancer risk."
E w e r t z , 4 2 I n t ' l . J. C a n c e r a t 8 3 5 .

13

PI. E x . #
Ex. 71M

Author/Publication

Year
1992

Statistical Results

Yang, C a n c e r C a u s e s & C o n t r o l

RR 1.0 ( 9 5 % CI 0 . 8 - 1 . 3 ) for ever-use o f u n o p p o s e d estrogen; RR 1.0 ( 9 5 % CI 1.02 . 0 ) for c u r r e n t u s e r s
RR 1.09 ( 9 5 % CI 0 . 8 6 - 1 . 3 7 ) for e v e r - u s e

Ex. 7IN

Weinstein, Int'l J. Epidem.
Schairer, Cancer Causes & Control

1993

o f menopausal estrogen pills
Ex. 71P

1994

RR 1.0 ( 9 5 % CI 0.9-1.2) for estrogeno n l y ; RR 1.2 ( 9 5 % CI 1.0-1.6) for estrogen and progestin
RR 1.2 ( 9 5 % CI 0 . 9 - 1 . 5 ) for e v e r - u s e o f

Ex. 7 1 Q
Ex. 71U

La V e c c h i a , B r i t i s h J. C a n c e r

1995

hormone replacement therapy

Levi, European J. o f Cancer Prevention
P e r s s o n , I n t ' l J. C a n c e r

1996

RR 1.3 ( 9 5 % CI 0 . 9 - 1 . 9 ) for ever-use o f

hormone replacement therapy
Ex. 7IV
1997

Results for breast cancer association with

estrogen replacement therapy not clear
Ex. 71X

Brinton, Menopause Hulley, JAMA
Kirsh, C a n c e r C a u s e s & C o n t r o l

1998

R R 0 . 9 ( 9 5 % CI 0 . 7 - 1 . 2 ) for e v e r - u s e o f

h o r m o n e replacement therapy
Ex. 71JJ
2002

RR 1.27 ( 9 5 % CI 0.84-1.94) for estrogen plus progestin therapy
RR 1.14 ( 9 5 % CI 0 . 8 1 - 1 . 5 9 ) for h o r m o n e

Ex. 7IKK

2002

replacement therapy
Ex. 7ILL

Newcomb, Cancer Epidem., Biomarkers
& Prevention

2002

RR 1.14 ( 9 5 % C I 1.04-1.26) f o r e v e r - u s e o f postmenopausal hormones; RR 1.28 ( 9 5 % CI 1 . 1 6 - 1 . 4 3 ) for e v e r - u s e o f

postmenopausal hormones after adjusting
for other observed factors

As the table indicates, one o f the studies, the Persson ( 1 9 9 7 ) study, d i d not p r o v i d e its results in a

form that facilitated straightforward analysis o f the breast cancer risks from estrogen replacement therapy. Because the study lacks clarity as to its conclusions, it is appropriate to read the study in favor o f plaintiff. Additionally, contrary to W y e t h ' s assertion, the Newcomb (2002) study apparently found a statistically significant association between hormone therapy and breast cancer. But even viewing the above 14 studies in a light most favorable to plaintiff, 12 studies
do not s h o w a statistically s i g n i f i c a n t association b e t w e e n e s t r o g e n r e p l a c e m e n t therapy and
breast cancer.

14

Removing these 12 studies from the 32 originally cited by plaintiff leaves just 20 preWHI studies that show a statistically significant risk o f breast cancer. As previously stated, the

parties agree that 43 studies constitute the proper universe o f relevant, pre-WHI studies. Put
another way then, 23 o f 43 pre-WHI studies did not show a statistically significant association
between breast cancer in w o m e n who have e v e r used estrogen replacement therapy. Accordingly, Statement No. 2 was not false.
iii. Statement No. 3

In a one-year clinical trial o f PREMPRO, PREMPHASE and Premarin alone, 5 new cases o f breast c a n c e r w e r e d e t e c t e d a m o n g 1377 w o m e n w h o received the combination treatments, w h i l e no n e w c a s e s were detected a m o n g 347 w o m e n
who received Premarin alone. The overall incidence of breast cancer in this

clinical trial does not exceed that expected in the general population.

Plaintiff argues that Statement No. 3 is false in two respects. First, plaintiff notes that
although Statement No. 3 indicates that the study in question included 1,724 participants in all, almost one-fourth o f the participants "dropped out" o f the study. PI. Br. at 12. Wyeth responds
that dropouts in such s t u d i e s are c o m m o n , and in any event, t h a t S t a t e m e n t No. 3 m a k e s no

representation about the participants who left the study. While the rate o f participation in an

epidemiological study may be appropriate to report along with the study's findings,14 Wyeth did
not act fraudulently in summarizing the study in question in Statement No. 3. An

epidemiological study will include a multitude of factors that may affect the analysis and validity
of, and confidence in, the results. Yet, including all o f this information would have transformed the Prempro label into an extensive rehashing o f the s t u d y ' s findings. The label is intended to

provide a brief summary o f the findings, and in this respect, the statement is reasonably worded

14 For example, the Consolidated Standards of Reporting Trials Group recommends that such
information be included in the "results" section o f any randomized trial study. See CONSORT Statement § 13a (2010), available at http://www.consort-statement.org/consort-statement/13-19--results/iteml3a_participant-flow/ (last visited Sept. 2 2 , 2 0 1 0 ) .

15

and not false. In light o f the elevated standard o f p r o o f in a fraud claim, no reasonable j u r y could
find fraud in this statement.

P l a i n t i f f s s e c o n d a r g u m e n t for fraud based on S t a t e m e n t No. 3 c o n c e r n s the assertion

that the "overall incidence o f breast cancer in this clinical trial does not exceed that expected in
the general population." P l a i n t i f f contends that Wyeth knew this representation was false

because an internal Wyeth document stated that the rate o f breast cancer among study participants actually exceeded the rate o f breast cancer among the population. But the internal document referenced by plaintiff clearly stated that the results o f that study showed a
standardized incidence ratio of 1.47 with a 95% confidence interval bounded between 0.47 and

3.43. Once again, p l a i n t i f f relies on the 1.47 value and i g n o r e s the fact that the c o n f i d e n c e

interval includes 1.0, making the results o f this study statistically inconclusive. Accordingly, the summary judgment record does not support an allegation o f fraud with respect to Statement No.
3. iv. Statement No. 4

In the three year clinical Postmenopausal Estrogen Progestin Intervention (PEPI) trial o f 875 women to assess differences among placebo, unopposed Premarin, and three different combination hormone therapy regimens, one (1) new case o f breast caner was detected in the placebo group (n=174), one in the Premarin alone group (n=175), n o n e in the c o n t i n u o u s Premarin plus c o n t i n u o u s medroxyprogesterone a c e t a t e g r o u p ( n = 1 7 4 ) and two (2) in the c o n t i n u o u s Premarin plus cyclic medroxyprogesterone acetate group (n=174).

Plaintiff argues that Statement No. 4 was "intended to deceive physicians and the public
alike into believing that the PEPI trial was evidence that E+P does not causes [sic] breast

cancer," even though "Wyeth knew all along that the PEPI trial told the world nothing about the
breast cancer risk." PI. Br. at 13. For her assertion about W y e t h ' s knowledge o f the s t u d y ' s purpose, p l a i n t i f f relies on a Wyeth internal email r e c o m m e n d i n g a g a i n s t including Statement
No. 4 on the P r e m p r o label b e c a u s e " P E P I w a s not d e s i g n e d to e v a l u a t e the r i s k or i n c i d e n c e o f

16

breast cancer" in light o f the small s a m p l e size o f the t r e a t m e n t groups. See Email to M a i d a

Burka (Jan. 10, 2001) (PI. Ex. 55) ("January 10 Email").15 This argument fails for two reasons.
First, the suggestion in the January 10 Email was simply that "this information does not add any
information to physicians or patients above our revised wording," not that the statement in the

label was false. PI. Ex. 55. Id. Second, Statement No. 4 specifically included the sample size
(i.e., "n=174") in its description o f the study. Since the sample size was the basis for the January 10 Email's suggestion that the PEPI study was not designed to evaluate cancer risks,

providing this sample size information on the label ensured that the study did not mislead

doctors.16 As such, neither the email nor the statement, standing alone, could lead a reasonable
j u r y to conclude that Statement No. 4 is fraudulent.

Finally, in addition to alleging fraud in the statements on the Prempro label, plaintiff
alleges that Wyeth acted fraudulently by e x c l u d i n g from the label the results o f two additional

studies in 2000. As recounted in Part I supra, in August 2000, the FDA requested that Wyeth change its Prempro label to include stronger statements about the risks o f E+P hormone therapy. Wyeth and the FDA then corresponded for several months about alternative wording for these
revisions before the FDA ultimately approved revisions in March 2001.

15 The email does not state, and the parties do not identify, the sender of the email, nor does the
record explain the role of Maida Burka at Wyeth. Nevertheless, consistent with the obligation to view the evidence in a light most favorable to plaintiff, the sender and recipient are assumed to be experts with influential roles in advising Wyeth as to its drug labeling.

16 One must assume for the purposes of this analysis that the doctor reading and relying upon the
Prempro label had sufficient e x p e r t i s e to understand the s i g n i f i c a n c e o f this information and evaluate the s t u d y ' s merits accordingly. If this were not true--that is, i f a d o c t o r who read this description o f the PEPI study were unable to discern its meaning or significance--then the d o c t o r would be u n r e a s o n a b l e in r e l y i n g on the information, and the fraud c l a i m w o u l d fail for
want of reasonable reliance.

17

At first glance, W y e t h ' s c o r r e s p o n d e n c e w i t h the F D A w o u l d s e e m to b e l i e any

allegation o f fraud. Wyeth carefully explained its concerns with the F D A ' s original draft o f the
revised warnings, and Wyeth ultimately reached an agreement with the FDA as to the appropriate changes to make. Had Wyeth implemented this agreed-upon revision, it would be

very difficult for plaintiff to meet her elevated burden to show fraud by clear and convincing
evidence. Yet, Wyeth did not actually implement changes to the label based on its discussion with the FDA. Indeed, changes were apparently not made to the 1997 Prempro label until after the 2002 WHI study. As such, Wyeth cannot rely on its negotiations with the FDA to justify its
refusal to update its warnings between 2000 and 2002.
Wyeth argues that the f r a u d u l e n t c o n c e a l m e n t c l a i m n e v e r t h e l e s s fails for two reasons: (i)

plaintiff has not provided evidence o f intent to commit fraud, and (ii) the scientific evidence in
2 0 0 0 regarding the risks o f E+P h o r m o n e t h e r a p y was inconclusive. As to W y e t h ' s first

argument, it is well-settled that intent to deceive is most often proven by circumstantial evidence
rather than " s m o k i n g gun" evidence. See French v. Beville, 62 S.E.2d 883, 889 (Va. 1951)

(noting that "[fjraud is seldom, i f ever, provable by direct testimony"). I f plaintiff succeeds in showing that Wyeth knowingly concealed material information about risks o f Prempro, then a
reasonable j u r y might infer that Wyeth acted with the requisite fraudulent intent.

As to Wyeth's second argument, it is true o f course that conflicting or inconclusive scientific studies may be a reasonable basis for Wyeth's refusal to include certain warnings on

the Prempro label. Indeed, that argument lies at the heart o f Wyeth's defense to the negligent
failure to w a r n claim. Yet, as W y e t h c o n c e d e s , a g e n u i n e d i s p u t e o f fact exists on that issue. So,

too, does such a genuine dispute o f material fact exist with respect to this concealment in the

label. While the fraud claim requires a higher showing o f intent and a higher standard o f p r o o f

18

than the negligent failure to w a r n c l a i m , the fraud claim is n e v e r t h e l e s s b o u n d up in a battle o f

the experts o v e r whether the scientific evidence compelled c h a n g e s to the Prempro label before

2002. Put succinctly, the summary j u d g m e n t record, when viewed in a light most favorable to plaintiff, could lead a reasonable j u r y to find by clear and convincing evidence that Wyeth
k n o w i n g l y concealed i n f o r m a t i o n from the 2 0 0 0 s t u d i e s c o n c e r n i n g the increased risk o f breast

cancer attributable to the use o f drugs like Prempro. Accordingly, while summary judgment is appropriate for Wyeth on the allegations o f fraudulent misrepresentation in the Prempro label,

plaintiffs allegation of fraudulent concealment cannot be resolved on summary judgment.17
C. Causation

In addition to its specific attacks on p l a i n t i f f s fraud claims, Wyeth also argues that plaintiff has failed to prove causation in any o f her claims because p l a i n t i f f s causation expert,
Dr. Michael Wertheimer, bases his opinion on unreliable methods in violation o f Rules 702 and
703, Fed. R. Evid., and Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993). W y e t h

has challenged Dr. Wertheimer's testimony in a separate Daubert motion that remains pending.
While success on this motion may entitle Wyeth to summary j u d g m e n t on p l a i n t i f f s remaining

claims, that issue is not yet ripe for adjudication. Accordingly, the denial in part o f summary
j u d g m e n t reflected in this M e m o r a n d u m Opinion is without prejudice to W y e t h ' s right to file

17 Wyeth argued that courts in other Prempro cases have granted Wyeth summary judgment on
the fraud claim. Yet, cases cited by Wyeth in this regard are distinguishable. In Rush v. Wyeth, the plaintiff argued that Wyeth committed fraud in the promotion o f P r e m p r o ' s cardiovascular benefits. See Rush v. Wyeth, No. 4 : 0 3 c v l 5 0 7 (E.D. Ark. Dec. 1 4 , 2 0 0 6 ) (Order). By contrast, the present action alleges fraud in s p e c i f i c statements on the P r e m p r o label, and o m i s s i o n s therefrom, concerning the risks o f breast cancer associated with taking Prempro. And in Bailey v. Wyeth, the New Jersey state court found that the fraud claims were subsumed by New J e r s e y ' s statutory product liability laws. See Bailey v. Wyeth, No. MID-L-0999-06 MT (Sup. Ct. N.J. July 1 1 , 2 0 0 8 ) (Opinion). As such, these t w o c a s e s are u n p e r s u a s i v e here.

19

another motion for summary judgment on the issue of causation should resolution of its Dauhen
motion warrant doing so.
IV.

Accordingly, summary judgment is appropriate in Wyeth's favor on the claim for

fraudulent misrepresentation, but not for the claims of fraudulent concealment and negligent
design defect Additionally, summary judgment is appropriate in Wyeth's favor on the claims

for sirici liability for failure to warn, strict liability for design defect, negligent misrepresentation,
and breach of express warranty, since plaintiff has conceded these claims.
A n appropriate O r d e r will issue.

Alexandria, Virginia O c t o b e r s 2010

T . S . E l l i s , III

United Slates District Judge

20