Davis v. McKesson Corporation et al

Filing 217

ORDER granting in part and denying in part 153 Omnibus motion; granting 157 Motion to exclude; granting 209 Motion for Leave to File. Signed by Senior Judge David G Campbell on 8/2/2019.(DGC, nvo)

1

WO

2
3
4
5
6

IN THE UNITED STATES DISTRICT COURT

7

FOR THE DISTRICT OF ARIZONA

8

Hilary Davis,

9
Plaintiff,

10
11

v.

12

No. CV-18-1157-PHX-DGC
CV-18-1159-PHX-DGC
CV-18-1778-PHX-DGC

McKesson Corporation, et al.,

13

ORDER
Defendants.

14
15

Plaintiffs Hillary Davis, Srihari Munnuru, and Susan Fischer have sued

16

Defendants Bayer Healthcare Pharmaceuticals, Inc., McKesson Corporation, Guerbet,

17

LLC, and other gadolinium-based contrast agent (“GBCA”) manufacturers and

18

distributors. Plaintiffs were administered GBCAs for the medical procedure known as

19

magnetic resonance imaging (“MRI”), and claim they developed various health problems

20

as a result. After conferring with the parties, the Court decided that the issue of general

21

medical causation – whether Plaintiffs can present admissible evidence that GBCAs

22

cause the health problems they claim – should be decided first. Doc. 110, 111, 115.

23

Accordingly, after focused discovery and Plaintiffs’ production of expert reports on

24

general causation, Defendants filed motions to exclude the experts’ testimony under

25

Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 597 (1993), and Rule 702

26

of the Federal Rules of Evidence. Docs. 153, 157. The motions are fully briefed, and the

27

Court heard oral arguments on July 3, 2019. For the reasons that follow, the Court will

28

1

grant Defendants’ motions with respect to the medical causation experts and deny them

2

with respect to the chemistry tutorial.1

3

I.

Background.

4

“Gadolinium is a lanthanide element (rare earth metal), which exhibits high

5

paramagnetism, a form of magnetism occurring only in the presence of an externally

6

applied magnetic field.” In re Gadolinium-Based Contrast Agents Prods. Liab. Litig.,

7

No. 1:08 GD 50000, 2010 WL 1796334, at * 3 (N.D. Ohio May 4, 2010) [hereinafter In

8

re GBCAs]. When GBCAs are injected intravenously, they enhance MRI scans, resulting

9

in high quality images that aid in identifying serious health conditions such as cancer,

10

infections, and bleeding. Docs. 154-27 at 2, 154-47 at 2. Since their initial approval in

11

1988, GBCAs have been used more than 450 million times worldwide. Doc. 154-24 at 2.

12

Gadolinium in its free state is toxic to humans. In re GBCAs, 2010 WL 1796334,

13

at *3. Because of this toxicity, gadolinium must be chemically bound to a chelate (a

14

compound that can be bound to a metal atom) before it is intravenously injected. This

15

binding prevents the gadolinium from interacting with human tissue. Id. GBCAs can be

16

chelated in a linear model or a macrocyclic model. Doc. 154-24 at 4. While all GBCAs

17

are mostly eliminated from the body through the kidneys, scientific evidence clearly

18

suggests that linear GBCAs are less stable and tend to dechelate in the body (the

19

gadolinium separates from the chelate) more readily than macrocyclic GBCAs.

20

Doc. 154-47 at 2; see also Doc. 154-24 at 5 (92-96% excreted in 1 hour).

21

Defendants’ products in this case are linear GBCAs.

All of

22
23
24
25
26
27
28

1

Three cases have been consolidated in this action: Davis v. McKesson
Corporation, 2:18-cv-00157; Munnuru v. Guerbet, LLC, No. 2:18-cv-01159-DGC; and
Fischer v. Bayer Healthcare Pharmaceuticals, Inc., No 2:18-cv-01778-DGC.
Documents cited from the lead case, Davis, will be referred to simply as “Doc.”
Documents cited from the Fischer and Munnuru cases will be preceded by the Plaintiff’s
last name. All citations are to public versions of the documents unless otherwise noted,
and are to page numbers added at the top of the documents by the Court’s electronic
filing system.
-2-

1

In 2006, several doctors and researchers noted a strong association between

2

GBCA use in patients with advanced kidney disease and the development of a medical

3

condition known as nephrogenic systemic fibrosis (“NSF”).

4

Subsequent studies revealed that the elimination of GBCAs from patients with kidney

5

disease takes significantly longer than from patients with healthy kidneys, and that the

6

GBCAs were never entirely eliminated. In re GBCAs, 2010 WL 1796334, at *6.

Doc. 154-14 at 84.

7

NSF is a disease that primarily involves the skin and subcutaneous tissues, but also

8

may involve other organs such as the lungs, esophagus, heart, and skeletal muscles. Id.

9

Symptoms may develop and progress rapidly or appear months or even years after GBCA

10

exposure. Id. Once the medical community stopped using GBCAs in renally-impaired

11

patients, NSF was “essentially eradicated.” Doc. 154-24 at 2.

12

Many of the persons who developed NSF brought suit against the manufacturers

13

of GBCAs, and the cases were consolidated in a multi-district litigation in 2010 (the

14

“NSF MDL”). See In re GBCAs, 2010 WL 1796334, at *1. Experts in the NSF MDL

15

opined that NSF was caused by dechelation of GBCAs after they were injected into the

16

body, which exposed the body to toxic gadolinium. Id. at *6. Although the experts could

17

not identify the precise mechanism by which gadolinium caused NSF, the district court

18

admitted their causation opinions. Id. The court found sufficient evidence – including

19

the rapid elimination of NSF when GBCAs were no longer used in renally-impaired

20

patients – that GBCAs caused NSF. Id. at *6. One jury trial was held in the NSF MDL,

21

resulting in a verdict for the plaintiff, and the remainder of the MDL cases settled.

22

This case differs from the NSF MDL in two important respects.

Although

23

Plaintiffs in this matter allege that they were exposed to GBCAs when they received

24

MRIs, they do not have impaired kidneys and they have not been diagnosed with NSF.

25

Rather, Plaintiffs and other patients with normal kidney function claim they began to

26

suffer a wide range of symptoms after they received GBCAs. Doc. 191-2 at 2. This

27

range of symptoms has not been recognized as a disease by regulatory authorities or

28
-3-

1

medical associations, but the collection of symptoms has been referred to in some

2

literature as “gadolinium deposition disease” (“GDD”). Id. For the sake of simplicity,

3

the Court will refer to Plaintiffs’ alleged illnesses as either GDD or gadolinium toxicity.

4

In doing so, the Court makes no judgment about whether GDD is or should be recognized

5

as a distinct illness, or whether GBCAs are in fact toxic in patients who have healthy

6

kidneys.

7

Plaintiff Hilary Davis alleges that she contracted GDD after receiving an

8

intravenous GBCA injection. Doc. 1 ¶ 2. In her first and amended complaints, she

9

alleged extreme cognitive impairment; difficult mentation; headaches; swollen, red,

10

thickening, and peeling skin; and pain throughout her body. Id. ¶ 39, Doc. 4 ¶ 43. In her

11

second amended complaint, she alleges that the gadolinium from the GBCA deposited

12

indefinitely in her organs and soft tissues and caused fibrosis in her organs, bone, and

13

skin, and crossed the blood-brain barrier to deposit in the neuronal nuclei of her brain.

14

Doc. 142 ¶ 12.

15

Plaintiff Srihari Munnuru received one or multiple MRIs where he was injected
Munnuru Doc. 1 ¶ 4.

He alleges that he developed GDD “soon

16

with a GBCA.

17

thereafter,” which has caused him severe injury and pain and suffering due to GDD. Id.

18

at 8 ¶ 17. In his deposition, he testified to leg stiffness, weakness, bone pain and joint

19

popping, foot pain, brain fog, and eye and teeth pain. Doc. 154-3 at 14.

20

Plaintiff Susan Fischer alleges that she contracted GDD from intravenous GBCA

21

injections. Fischer Doc. 1 ¶ 2. She alleges the following symptoms: burning sensation;

22

violent

23

hypoglycemia; difficult, painful movement; low body temperature; inflammation,

24

especially throughout her lymphatic system; muscle cramps; numbness; tingling

25

sensation; aching joints; weight loss; hair loss; lumps and rashes on her body; kidney

26

damage; and osteoporosis.

27

gadolinium retention in multiple organs years after the GBCAs were administered.

shaking;

tremors;

clouded

Id. ¶ 16.

mentation;

confusion;

weakness;

fatigue;

In her amended complaint, Fischer alleges

28
-4-

1

Fischer Doc. 10 ¶ 2. She alleges that the gadolinium has caused fibrosis in her organs,

2

bone, and skin, and that it has crossed the blood-brain barrier and deposited in the

3

neuronal nuclei of her brain. Id. ¶ 26; see also Fischer Doc. 60 ¶ 30.

4

Plaintiffs’ cases are among several actions pending in district courts across the

5

United States. Fischer Doc. 39 at 1. All claim adverse health effects related to GBCA

6

administration.

7

considered and denied the motion of 17 plaintiffs to centralize pretrial proceedings in an

8

MDL. Id. The panel determined that even though there are common issues about

9

gadolinium toxicity and retention, the injuries alleged in each case appear highly specific

10

to each individual plaintiff and the actions involve GBCAs manufactured by four

11

different defendant groups, involving different GBCA formulations. Id.

12

II.

13
14
15
16
17
18
19

On October 10, 2018, the Judicial Panel on Multidistrict Litigation

Relevant Legal Standards.
Defendants’ motions are governed by Rule 702 of the Federal Rules of Evidence.

That rule provides:
A witness who is qualified as an expert by knowledge, skill,
experience, training, or education may testify in the form of an opinion or
otherwise if:
(a) the expert’s scientific, technical, or other specialized knowledge
will help the trier of fact to understand the evidence or to determine a fact
in issue;

20

(b) the testimony is based on sufficient facts or data;

21
22

(c) the testimony is the product of reliable principles and methods;
and

23
24

(d) the expert has reliably applied the principles and methods to the
facts of the case.

25
26

Fed. R. Evid. 702.

27
28
-5-

1

Curiously, the parties make few references in their briefs to Rule 702, relying

2

instead almost exclusively on Daubert and related cases. The Court notes, however, that

3

Daubert applied the then-existing version of Rule 702 and held that it superseded the

4

common law rule in Frye v. United States, 293 F. 1013, 1014 (D.C. Cir. 1923). See

5

Daubert, 509 U.S. at 587-89. Rule 702 was amended in 2000, seven years after Daubert

6

was decided. Amendment of the rule required approval by the Supreme Court and

7

acceptance by Congress under the Rules Enabling Act, and the amended rule superseded

8

any other law. See 28 U.S.C. § 2072(b) (“All laws in conflict with such rules shall be of

9

no further force or effect after such rules have taken effect.”). Thus, Rule 702 provides

10

the governing law for this case. Because the 2000 amendment sought to codify and

11

clarify the admissibility of expert testimony in light of Daubert and related cases, those

12

cases clearly remain relevant, but the Court will structure its analysis primarily around

13

the requirements of Rule 702. See Fed. R. Evid. 702 advisory committee’s note to 2000

14

amendment.

15
16

Three legal principles are particularly relevant to the Court’s gatekeeping role in
this case.

17

First, the proponent of expert testimony bears the burden of showing that the

18

proposed testimony is admissible under Rule 702. Cooper v. Brown, 510 F.3d 870, 942

19

(9th Cir. 2007); Lust ex rel. Lust v. Merrell Dow Pharm., Inc., 89 F.3d 594, 598 (9th

20

Cir. 1996). If Plaintiffs fail to carry this burden, the opinions of their experts are not

21

admissible. Cooper, 510 F.3d at 942.

22

Second, the admissibility of expert testimony under Rule 702 is a preliminary

23

question the Court must decide under Federal Rule of Evidence 104(a), and the

24

Rule 104(a) decision must be made by a preponderance of the evidence. See Daubert,

25

509 U.S. at 592 & n.10; Bourjaily v. United States, 483 U.S. 171, 175-76 (1987). Some

26

cases mistakenly suggest that some of the elements in Rule 702 are jury questions – that

27

whether an expert’s opinion is based on reliable principles and methods, and whether

28
-6-

1

those principles and methods have been applied to the facts of the case reliably, go to the

2

weight of the evidence and should be decided by the jury after cross examination and

3

argument at trial. But the requirements of Rule 702 are conditions for admissibility, and

4

the Supreme Court has made clear that “the trial judge must determine at the outset,

5

pursuant to Rule 104(a),” whether the expert’s testimony is admissible under Rule 702.

6

Daubert, 509 U.S. at 592. Thus, the Court’s task in this order is to determine whether

7

Plaintiffs have shown by a preponderance of the evidence that the requirements of Rule

8

702(a)-(d) have been satisfied with respect to each expert’s opinions. Id. at 592 n.10;

9

Bourjaily, 483 U.S. at 175-76. Both sides agreed with this principle during the hearing

10

on the motions.2

11

Third, experts must explain the basis for their conclusions in a manner that allows

12

the Court to determine whether they are using reliable principles and methods and are

13

applying them to the facts of the case in a reliable manner. Fed. R. Evid. 702(c), (d).

14

Although “[t]rained experts commonly extrapolate from existing data . . . nothing in

15

either Daubert or the Federal Rules of Evidence requires a district court to admit opinion

16

evidence that is connected to existing data only by the ipse dixit of the expert. A court

17

may conclude that there is simply too great an analytical gap between the data and the

18

opinion proffered.” Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997).

19

With these legal standards in mind, the Court will discuss the current state of the

20

science with respect to NSF and GDD and the basic approach to causation used by

21
22
23
24
25
26
27
28

2

The Ninth Circuit has emphasized the statement in Daubert that a district court
should conduct the admissibility analysis “with a ‘liberal thrust’ favoring admission.”
Messick v. Novartis Pharm. Corp., 747 F.3d 1193, 1196 (9th Cir. 2014) (quoting
Daubert, 509 U.S. at 588). It has stated that the gatekeeping function is meant to “screen
the jury from unreliable nonsense opinions, but not to exclude opinions merely because
they are impeachable.” Alaska Rent-A-Car, Inc. v. Avis Budget Group, Inc., 738 F.3d
960, 969 (9th Cir. 2013). The Court will follow this guidance, but with the understanding,
established by the Supreme Court cases cited above, that the Court may admit expert
opinions only if it can determine, under Rule 104(a), that Plaintiffs have shown each of
the Rule 702 requirements to be satisfied by a preponderance of the evidence.
-7-

1

Plaintiffs’ medical causation experts.

2

causation experts, two cases cited heavily by Plaintiffs, and Plaintiffs’ chemistry expert.

3

III.

The Court then will address the individual

Background Information on NSF and GDD.

4

A.

NSF.

5

Borrowing liberally from a decision of the district court in the NSF MDL, the

6

Court provides the following description of NSF. See In re GBCAs, 2010 WL 1796334,

7

at *3. NSF was first described in the medical literature in 2000, with the first reported

8

cases going back to 1997. NSF causes fibrosis of the skin, connective tissue, and organs

9

throughout the body, and is a painful, progressive, and debilitating disease. While the

10

precise pathogenesis of NSF is unknown, it has been reported only in patients who have

11

severe kidney disease and, with the exception of a few reported cases with inconclusive

12

medical histories, has been found exclusively in kidney patients who have had one or

13

more exposures to GBCAs.

14

In June 2006, the FDA issued a Public Health Advisory notifying healthcare

15

professionals and the public about the risk of NSF following the administration of

16

GBCAs. In December 2006, the FDA issued an updated Public Health Advisory stating

17

that there was a potential for NSF to occur in at-risk patients following administration of

18

GBCAs. In May 2007, the FDA asked GBCA license holders to issue a warning about

19

the risk of NSF in patients with renal failure. This warning, along with policies and

20

procedures adopted by healthcare facilities and notice to healthcare providers, led to the

21

virtual elimination of new NSF cases. Id.

22

In response to the growing number of NSF cases, medical professionals identified

23

specific diagnostic criteria for identifying the disease. Doc. 154-17. The court in the

24

NSF MDL described NSF as “a singular disease.” In re GBCAs, 2010 WL 1796334 at

25

*6. Although the plaintiffs’ experts in the MDL could not identify with certainty the

26

mechanism by which gadolinium in GBCAs caused NSF, the court found their general

27

causation opinions admissible, providing this explanation:

28
-8-

1

The dominant theory is that dechelation occurs through
transmetalation (simply, a chemical reaction involving the exchange of
ligands between two metal centers), although there are other theories
including that dechelated (or, free) gadolinium has a proliferative effect on
human dermal fibrosis and gadolinium's propensities as a calcium blocker
triggers the fibrotic process. In any event, given the wealth of evidence on
causation – that is, the rapid emergence and decline of NSF associated with
the rise and fall of its use in renally impaired persons, the presence of
gadolinium in the tissue of NSF patients, the known toxicity of gadolinium,
and the majority view in the published and peer reviewed studies and
articles that dechelated gadolinium causes NSF – the Court concludes that it
is not necessary for Plaintiffs’ experts to identify the precise mechanism by
which dechelated gadolinium causes NSF in order to present the theory to a
jury.

2
3
4
5
6
7
8
9
10
11

Id.

12

In short, the general medical causation opinions were admissible in the NSF MDL

13

because (1) the increase and decrease of NSF cases tracked the increase and decrease of

14

GBCA use in renally-impaired patients, (2) a majority of published and peer reviewed

15

studies found a causal link between GBCAs and NSF in such patients, (3) gadolinium

16

was found in the tissue of NSF patients, and (4) free gadolinium was known to be toxic.

17

The evidence in this case is less clear.

18

B.

19

Plaintiffs do not argue that GDD has been recognized as a distinct disease by any

20

regulatory body or medical association, but note that it has been proposed as an illness in

21

some published literature. See Semelka et al. (2016), Doc. 156-16 (proposing GDD as an

22

illness). Other literature has asserted that labelling symptoms as GDD is too strong a

23

suggestion that it is an actual illness, and has proposed instead that the name be changed

24

to “gadolinium-associated symptoms.” See Parillo et al. (2018), Doc. 154-26 at 7.

25

GDD.

1.

MIDAC Review.

26

In 2017, the FDA convened its Medical Imaging Drugs Advisory Committee

27

(“MIDAC”) to consider “the potential risk of gadolinium retention . . . in patients

28
-9-

1

receiving [GBCAs].” Doc. 154-42 at 2. The committee consisted of 15 voting members,

2

all of whom (with the exception of one patient and one consumer representative) were

3

medical doctors or Ph.D.s.

4

laboratory directors, and directors from the National Institutes of Health and the National

5

Cancer Institute. See Transcript of September 8, 2017 MIDAC Conference, Doc. 156-7

6

at 3-9 [hereinafter MIDAC Trans.]. The FDA convened MIDAC because “[t]he evidence

7

indicating retention following the use of GBCAs has led to concerns that gadolinium

8

retention may cause adverse reactions, if not immediately then at some later date.”

9

Doc. 154-43 at 4. The focus of MIDAC’s inquiry was the connection, if any, between the

10

retention of GBCAs or gadolinium in the body and various symptoms reported in patients

11

with healthy kidneys.

They included seven professors at major universities,

12

The FDA prepared a lengthy briefing document for the MIDAC inquiry that

13

reviewed all medical literature since 1988, case reports, regulatory decisions, and reports

14

from the FDA’s adverse event reporting system (“FAERS”). Doc. 154-43. MIDAC then

15

held a conference on September 8, 2017, where it heard presentations from FDA medical

16

officers, invited scholars, industry representatives, and the public. A 401-page transcript

17

records all of the presentations and discussions, as well as MIDAC’s vote on key

18

questions. See MIDAC Trans., Doc. 156-7. After considering all of the evidence and

19

hearing presentations from scholars, experts, and patient advocacy groups, MIDAC

20

unanimously concluded that the medical and scientific evidence does not establish that

21

GBCAs cause GDD. See id. at 338-56. This does not appear to have been a difficult

22

decision for the committee members – there was no equivocation in their views. See id.

23

The chair, Dr. Peter Herscovitch, summarized the committee’s views in these words:

24

“I think there is fair uniformity that there is no evidence of a causal relationship between

25

the symptoms and signs in patients with normal renal function and the retention of

26

gadolinium.” Id. at 356. MIDAC recommended that the FDA revise GBCA warnings to

27

note that retention of gadolinium in some organs, including the brain, is possible

28
- 10 -

1

following GBCA use, but otherwise did not recommend that GBCA use be restricted. Id.

2

at 377; Doc. 154-45 at 8-9.
2.

3
4
5
6
7
8
9
10
11
12

The FDA and Other Regulatory and Medical Organizations.

Following MIDAC’s conference, the FDA approved this revised GBCA product
label in April 2018:
Gadolinium is retained for months and years in several organs. . . .
Consequences of gadolinium retention in the brain have not been
established. . . . There are rare reports of pathologic skin changes in patients
with normal renal function. Adverse events involving multiple organ
systems have been reported in patients with normal renal function without
an established causal link to gadolinium retention. . . . [C]linical
consequences of gadolinium retention have not been established in
plaintiffs with normal renal function[.]
Doc. 154-29 at 5.

13

Due to the retention of gadolinium in the body, other national organizations such

14

as the Australian Department of Health, Health Canada, the European Medicines Agency

15

(“EMA”), the Japanese Pharmaceuticals and Medical Devices Agency, and the New

16

Zealand Ministry of Health have suspended the use of some linear GBCAs. Doc. 153

17

at 12. But these organizations have also concluded that there is insufficient evidence to

18

show that gadolinium deposition in brain tissues causes harm to patients. See Docs. 154-

19

34 at 2, 154-33 at 2, 154-35 at 2, 154-36 at 2, 154-38 at 2. 3

20

Additionally, many scientific and medical societies have agreed that no adverse

21

health effects from gadolinium retention in renally-healthy patients have been proven.

22

See Doc. 153 at 13. These include 2018 findings by the National Institutes of Health, the

23
24
25
26
27
28

At the MIDAC conference, an FDA medical officer described EMA’s reasons
for suspending linear GBCAs: “there’s a theoretical risk associated with gadolinium
retention in the brain; . . . the clinical consequences could take many years to identify; . . .
the concentration of gadolinium in the brain seems to be higher after the linear agents
compared to the macrocyclics; gadolinium clearance from the brain occurs at a much
faster rate after macrocyclics as compared to linears; . . . there is greater disassociation of
gadolinium from the linear GBCAs compared to the macrocyclics; and . . . clinically, the
multipurpose GBCAs are interchangeable.” MIDAC Trans., Doc. 156-7 at 188.
3

- 11 -

1

American College of Radiology, the Radiological Society of North America, and the

2

Canadian Association of Radiologists (Docs. 154-24 at 2, 154-16 at 2), and 2017 findings

3

by the American Society of Neuroradiology and the International Society for Magnetic

4

Resonance in Medicine (Docs. 154-13 at 2, 154-18 at 2, 154-12 at 2).

5

3.

Evidence Supporting GDD and its GBCA Cause.

6

Although regulators and scientific organizations have not found enough evidence

7

to conclude that GBCAs cause an illness like GDD, there is evidence that some level of

8

GBCAs and gadolinium are retrained in the brain, bone, and other organs of patients who

9

have received GBCAs in connection with an MRI, even when they do not have impaired

10

renal function. See Kanda et al. (2014), Doc. 154-24 at 2; 191-5 (hypothesizing the

11

existence of intracranial gadolinium retention based on the observation of unenhanced T1

12

signal intensity increases in the dentate and globus pallidus); see also Semelka et al.

13

(2016), Doc. 187-34 at 2 (describing studies that determined gadolinium was being

14

retained in the brain and bone). No party appears to contest this, but the parties disagree

15

on the form of the retained gadolinium and whether it is toxic in that form and at all

16

doses. And as already noted, research indicates that linear GBCAs deposit in larger

17

amounts or in dechelated forms more often than macrocyclic forms. Doc. 154-24 at 4

18

(study showed after two weeks in human serum that macrocyclic GBCA disassociations

19

were undetectable but linear GBCAs showed variable degrees of dissociation).

20

Plaintiffs also point to publications that have drawn a connection between retained

21

GBCAs and various symptoms in renally-healthy patients. These include Burke (2016),

22

where researchers conducted an anonymous online survey of patients who believed that

23

they suffered from gadolinium toxicity. Doc. 187-33 at 2. The researchers received 50

24

responses, 33 of which described the onset of symptoms immediately following GBCA

25

administration and 16 of which described the onset within six weeks. Id. The most

26

commonly reported symptoms were bone and joint pain and head and neck symptoms,

27

including headaches, vision change, and hearing change. Id. Skin changes were also

28
- 12 -

1

indicated by about 60% of the respondents.

Id.

The authors of the Burke study

2

acknowledged that the “risk of severe adverse reaction to GBCA exposure is extremely

3

small, hence it may be difficult to determine characteristics of individuals that predispose

4

to this type of reaction.” Id. at 3.

5

In Semelka (2016), researchers reviewed the medical history and physical

6

examinations of four individuals with normal renal function who reported development

7

of new disease features within hours to four weeks of receiving GBCA injections.

8

Doc. 187-34 at 2. The authors concluded that gadolinium toxicity may occur in subjects

9

with normal renal function. Id. Central torso and peripheral arm and leg pain were

10

common features. Id. Distal arm and leg pain and rubbery subcutaneous tissue were

11

seen in later stages. Id. Clouded mentation was also common. Id. The authors noted

12

features that resembled NSF. Id. at 4.

13

In another Semelka article (2016), the authors recruited participants with normal

14

renal function and evidence of gadolinium in their system beyond 30 days from two

15

online gadolinium toxicity support groups. Doc. 156-16 at 2. Participants were asked to

16

answer survey questions about their symptoms. Id. The most common symptoms were

17

central pain, peripheral pain, headache, and bone pain. Id. Subjects with distal leg and

18

arm distribution described skin thickening. Id. Clouded mentation and headaches were

19

described as persistent beyond three months in 29 of the 42 subjects. Id. From the

20

results, the authors proposed to call this condition GDD. Id. at 4. Many of the reported

21

symptoms appeared to resemble less severe symptoms of NSF. Id. The authors theorized

22

that the symptoms may come from an immune system response to these agents. Id. They

23

were careful to note the limitations of their study – self reported survey answers,

24

subjectivity, predefined questions, and lack of a control group, noting that “further

25

prospective investigation is needed to verify this condition.” Id.

26
27
28
- 13 -

1

C.

2

The history of NSF and GDD frame the issue in this case. On one hand, we have

3

the recognized disease of NSF in renally-impaired patients and its accepted link to

4

GBCAs. On the other hand, we have the broad-ranging proposed illness of GDD in

5

renally-healthy patients and its alleged link to GBCAs. The distinction between the two

6

was captured in this comment by Dr. Jeffrey Brent, a member of the FDA’s MIDAC

7

committee, during MIDAC’s 2017 conference:

8
9
10
11
12
13
14
15

The Issue.

We know NSF is a disease. It’s a very clear-cut unambiguous
disease caused by gadolinium retention in patients who have renal failure.
It’s interesting to contrast that with the gadolinium retention disease we’re
hearing about today in patients with normal renal function. NSF is an
unmistakable, easy to diagnose, clear cut, limited but devastatingly serious
clinical condition, limited in the sense of clinical manifestations. What we
heard about today is multisystem, multisymptom disease without any
unifying presentations that would suggest a physiology that would seem to
make sense. So there are a lot of questions here about what this actually is,
if it actually is a disease.
MIDAC Trans., Doc. 156-7 at 329-30.

16

The question to be answered in this case is whether Plaintiffs can present

17

admissible expert opinions that GBCAs cause an illness described as GDD, or whether

18

such causation remains in the realm of reasonable suspicion, unsupported by sufficient

19

scientific and medical evidence to satisfy Rule 702. In resolving this question, the Court

20

must remember that reliable scientific principles and methods, applied reliably, are

21

required by Rule 702. Reasonable suspicions and plausible theories are not enough. As

22

other courts have noted, “an opinion that is an insightful, or even inspired, hunch is not

23

admissible if it lacks scientific rigor.” Anderson v. Bristol Myers Squibb, No. Civ.A. H-

24

95-0093, 1998 WL 35178199, at *12 (S.D. Tex. Apr. 20, 1998). “The courtroom is not

25

the place for scientific guess work, even of the inspired sort. Law lags science; it does

26

not lead it.” Rosen v. Ciba-Geigy Corp., 78 F.3d 316, 319 (7th Cir. 1996).

27
28
- 14 -

1

Plaintiffs’ counsel argued at the hearing on these Daubert motions that the

2

relevant question is whether GBCAs should be included in the list of potential causes to

3

be considered in any differential diagnosis of Plaintiffs’ illnesses.

4

vagueness in this argument. On one hand, it could mean that Plaintiffs can clear the

5

Daubert hurdle on general causation with expert opinions that GBCAs could potentially

6

cause GDD, even if GBCAs have not been identified as a known cause. On the other

7

hand, it could mean that Plaintiffs must present expert opinions that GBCAs are in fact a

8

cause of GDD, and that GBCAs are therefore a potential cause of Plaintiffs’ illnesses

9

when considered with all of the other potential causes in a particular Plaintiff such as, for

10

example, fibromyalgia, other autoimmune illnesses, muscle and bone maladies, or other

11

factors from Plaintiffs’ personal medical histories.

There is some

12

The latter formulation is correct. The Court’s Case Management Order identified

13

the subject of this order as “Daubert motions on general causation.” Doc. 115 at 4.

14

Plaintiffs must show that their experts can present admissible opinions that GBCAs cause

15

GDD. Stated differently, the issue to be decided is whether Plaintiffs’ experts can present

16

admissible opinions under Rule 702 that GBCAs are in fact a cause of GDD. Whether

17

GBCAs actually caused any particular Plaintiff’s illness is an issue of specific causation

18

to be decided later in this litigation.4

19
20
21
22
23
24
25
26
27
28

Plaintiffs’ counsel emphasized in their briefing and argument that the Court’s
Case Management Order identified the issue as “‘whether exposure to a substance for
which a defendant is responsible . . . is capable of causing a particular injury or condition
in the general population.’” Doc. 110 at 11 (quoting Jaros v. E.I. DuPont, 292 F.3d
1124, 1133 (9th Cir. 2002)) (emphasis added). Plaintiffs focus on the word “capable” in
this statement, suggesting that it means the mere potential to cause Plaintiffs’ illnesses.
The Court does not agree. Plaintiffs must present admissible expert opinions that GBCAs
cause GDD. The Court has reviewed the record in this case and is confident that
Plaintiffs have understood this to be the relevant inquiry from the outset. See Docs. 110,
111. This is confirmed by Plaintiffs’ expert reports, each of which opines to a reasonable
degree of medical or scientific certainty that GBCAs cause GDD. The Court must now
decide whether those opinions are admissible.
4

- 15 -

1

IV.

Rules 702(a) and (b) – Qualifications and Sufficiency of Facts and Data.

2

Plaintiffs offer four experts to support their theory of general medical causation:

3

Brent Wagner, M.D., Jody Tversky, M.D., Margert Whittaker, Ph.D., and Kenneth

4

Raymond, Ph.D. The three causation experts – Drs. Wagner, Tversky, and Whittaker –

5

opine that GBCAs cause GDD in patients with healthy kidneys. The fourth expert, Dr.

6

Raymond, is a chemist who explains the nature of GBCAs and how they react in the

7

human body. Plaintiffs assert that Dr. Raymond’s chemistry testimony will be essential

8

to the jury’s understanding of GBCAs and their fate in the body, but that they will not use

9

him to state a causation opinion.
Rule 702(a) – Qualifications.

10

A.

11

Dr. Wagner is a board-certified physician in internal medicine and nephrology.

12

Doc. 154-10 at 2. He currently serves as the Director of the Kidney Institute of New

13

Mexico, as the Renal Section Chief of the Medicine Section of the New Mexico Veterans

14

Health Care System, and as an Associate Professor of Medicine at the University of New

15

Mexico Health Science Center. Id. He has done research on patients with impaired

16

kidney function exposed to GBCAs and the subsequent development of NSF. Id. For

17

nearly two decades, Dr. Wagner’s research has focused on renal development and

18

mechanisms implicated in kidney injury. Id. at 4. He was the keynote speaker at

19

MIDAC’s 2017 conference. See MIDAC Trans., Doc. 156-7 at 44-62.

20

Dr. Tversky is a full-time faculty member and former clinical director of the

21

Division of Allergy and Clinical Immunology at The John Hopkins University School of

22

Medicine. Doc. 154-9 at 2. He is board certified in internal medicine and allergy. Id.

23

He is a practicing allergist “with extensive experience managing drug reactions,

24

anaphylaxis, and immunological pathology.” Id. at 3.

25

Dr. Whittaker is the managing director and chief toxicologist of ToxServices LLC,

26

a scientific consulting firm headquartered in Washington, D.C. Doc. 154-11 at 5. She is

27

a Diplomate of the American Board of Toxicology and a board-certified toxicologist. Id.

28
- 16 -

1

She has more than 25 years of experience assessing human health and environmental

2

hazards from substances including inorganic and organic chemicals, microorganisms, and

3

radiological substances. Id.

4

Dr. Raymond is an organic chemist and professor emeritus in chemistry at the

5

University of California, Berkeley. Doc. 154 8 at 2. He has lectured extensively on

6

metal chelate design and has an interest in lanthanide coordination chemistry. Id. For

7

many years, he had National Institutes of Health support for the development of GBCAs,

8

which resulted in patents that have been licensed to GBCA providers. Id. He has served

9

as a consultant on MRI contrast agents and studies the role of metals in biology and

10

medicine. Id.

11

The Court finds by a preponderance of the evidence that these experts are

12

qualified by training and experience to opine on the subjects each will address, and that

13

their testimony, if admitted, would be helpful to the jury. Rule 702(a) is satisfied.
Rule 702(b) – Sufficiency of Facts and Data.

14

B.

15

Plaintiffs’ experts do not opine on specific causation – whether any Plaintiff’s

16

symptoms have in fact been caused by his or her exposure to GBCAs. As a result, they

17

have not focused on Plaintiffs or their medical conditions and histories. The facts and

18

data considered by the experts consist of studies, surveys, and case reports related to the

19

question of gadolinium toxicity in patients with healthy kidneys. The Court finds that

20

their opinions are based on sufficient facts and data. Rule 702(b) is satisfied by a

21

preponderance of the evidence.

22

V.

Rule 702(c) and (d) – General Discussion of the Medical Causation Opinions.

23

In this section of the order, the Court will engage in a general discussion of the

24

causation experts’ opinions, identifying serious flaws in some of the principles and

25

methods they use, as well as a lack of reliability in how they apply some otherwise

26

reliable principles and methods. In the three sections that follow, the Court will address

27

each of the causation experts individually, applying this general discussion.

28
- 17 -

1

A.

Unproven Hypotheses Are Not Admissible.

2

As the Supreme Court explained in Daubert: “Scientific methodology today is

3

based on generating hypotheses and testing them to see if they can be falsified; indeed,

4

this methodology is what distinguishes science from other fields of human inquiry.” 509

5

U.S. at 593 (quotation marks and citations omitted). Merely generating hypotheses –

6

even reasonable ones – is not enough. “[I]n order to qualify as ‘scientific knowledge’ an

7

inference or assertion must be derived by the scientific method. Proposed testimony must

8

be supported by appropriate validation – i.e., ‘good grounds,’ based on what is known.”

9

Id. at 590; Claar v. Burlington N. R.R. Co., 29 F.3d 499, 502-03 (9th Cir. 1994)

10

(“[S]cientists whose conviction about the ultimate conclusion of their research is so firm

11

that they are willing to aver under oath that it is correct prior to performing the necessary

12

validating tests could properly be viewed by the district court as lacking the objectivity

13

that is the hallmark of the scientific method.”); see also Nease v. Ford Motor Co., 848

14

F.3d 219, 232 (4th Cir. 2017); Mitchell v. Gencorp, Inc., 165 F.3d 778, 783 (10th Cir.

15

1999).

16

Daubert identified several means for determining whether a theory has been

17

reliably validated. “[A] key question to be answered in determining whether a theory or

18

technique is scientific knowledge that will assist the trier of fact will be whether it can be

19

(and has been) tested.” 509 U.S. at 593. “Another pertinent consideration is whether the

20

theory or technique has been subjected to peer review and publication.”

21

“Additionally, in the case of a particular scientific technique, the court ordinarily should

22

consider the known or potential rate of error[.]” Id. at 594. Finally, “[w]idespread

23

acceptance can be an important factor in ruling particular evidence admissible, and a

24

known technique which has been able to attract only minimal support within the

25

community may properly be viewed with skepticism.” Id. (quotation marks and citations

26

omitted). Additional considerations are identified in the Advisory Committee Note to the

27

2000 amendment of Rule 702 and will be discussed below.

28
- 18 -

Id.

1
2

These factors underscore the basic Daubert inquiry – whether an expert’s theory
or hypothesis has been tested and shown to be valid. If not, it is inadmissible.

3

B.

The Common Approach of Plaintiffs’ Experts.

4

Plaintiffs’ causation experts offer theories about the existence of GDD and how it

5

is caused, and cite sources which suggest that the theories are worth exploring, but they

6

do not present reliable scientific principles and methods, or apply accepted principles and

7

methods reliably, to show that their theories have been validated. The starting point for

8

the experts’ theories consists of several facts Plaintiffs claim to be undisputed:

9

(a) GBCAs are an accepted cause of NSF; (b) GBCAs dechelate in the human body,

10

releasing free gadolinium; (c) some level of GBCAs and free gadolinium is retained in

11

the bodies of patients with healthy kidneys; and (d) free gadolinium is toxic in the human

12

body. See Doc. 154-10 at 7-16 (Wagner Report); Doc. 154-9 at 4-5 (Tversky Report);

13

Doc. 154-11 at 15-23 (Whittaker Report); see also Doc. 185 at 22.5 But these facts alone

14

are not enough to establish general causation in this case, as Plaintiffs’ experts all seem to

15

concede. These facts suggest that GBCAs can cause adverse symptoms in renally-

16

healthy patients, and they may be enough to raise a reasonable hypothesis that must be

17

tested. But more is needed to conclude reliably that the variety of symptoms associated

18
19
20
21
22
23
24
25
26
27
28

5

Defendants do not dispute these four facts as general propositions but argue that
they are not relevant without information about dose. For example, Defendants concede
that free gadolinium can be toxic to humans, but argue that dose matters and that
Plaintiffs’ experts identify no minimum dose for gadolinium toxicity or for the
development of GDD. Defendants are correct that dose matters. See Federal Judicial
Center Reference Manual on Scientific Evidence (3d ed.), 2011 WL 7724262 at *3 (“‘the
dose makes the poison’; this implies that all chemical agents are intrinsically hazardous –
whether they cause harm is only a question of dose”). But dose primarily is relevant to
the question of specific causation – whether a particular exposure caused the plaintiff’s
injury. Courts have placed less emphasis on dose when addressing general causation.
See Clausen v. M/V New Carissa, 339 F.3d 1049,1059 (9th Cir. 2003). This is especially
true in cases like this that do not involve common background levels of exposure to a
substance, or where there is no identified benign exposure level. See In re: Zicam Cold
Remedy Mktg, Sales Practices, & Prods. Liab. Litig., 797 F. Supp. 2d 940, 944-45 (D.
Ariz. 2011); Henricksen v. ConocoPhillips Co., 605 F. Supp. 2d 1142, (E.D. Wash.
2009). Because dose is not essential to the general causation inquiry, the four facts that
form the basis for Plaintiffs’ expert opinions will be, for purposes of this order,
considered undisputed. The Court notes, however, that dose does matter when the
experts seek to apply animal or in vitro studies to humans, as discussed below.
- 19 -

1

with GDD are in fact an illness, and that the illness is caused by GBCAs. That has been

2

the conclusion of the FDA and every other regulatory and medical body that has

3

considered the question. These bodies are all well aware of Plaintiffs’ four foundational

4

facts, and they have found them to be good reasons for further study, but they

5

unanimously have found that there is not enough scientific evidence to conclude that

6

GBCAs cause GDD. See Doc. 153 at 11-13.

7

The key question in this case, therefore, is whether Plaintiffs’ experts present

8

reliable principles and methods, applied reliably, to bridge the gap between these

9

foundational facts and the conclusion that GBCAs cause GDD. The gap cannot be

10

bridged merely by the mere ipse dixit – the say-so – of the experts. Joiner, 522 U.S.

11

at 146. There must be a reliable scientific basis.

12

Plaintiffs’ experts attempt to span the gap by positing that GBCAs cause a

13

“continuum” of symptoms, the most severe of which constitute NSF and the less severe

14

of which are commonly seen in GDD.

15

evidence: case reports, surveys, animal studies, and in vitro studies. See Doc. 154-10

16

at 17-31 (Wagner Report); Doc. 154-9 at 5-10 (Tversky Report); Doc. 154-11 at 23-32

17

(Whittaker Report).

18

category of evidence, and other concerns about the experts’ principles and methods.6

They rely primarily on four categories of

The Court will address the experts’ proposed continuum, each

19

C.

The Continuum.

20

Plaintiffs’ causation experts posit that there is a continuum of symptoms caused by

21

GBCAs, with the more severe symptoms commonly seen in patients with renal

22

impairment (NSF) and less severe symptoms commonly seen in patients without renal

23

impairment (GDD). Docs. 154-9 at 3 (Tversky Report); 154-10 at 31 (Wagner Report).

24

If this is true, the continuum essentially establishes the general medical causation

25

Plaintiffs are trying to prove because all of the symptoms on the continuum, including

26
27
28

6

Wagner also states that he relies on his treatment of patients suffering from
gadolinium toxicity. Doc. 154-10 at 2, 4. This factor will be discussed below in the
specific review of Dr. Wagner’s opinions.
- 20 -

1

those experienced by Plaintiffs, are caused by GBCAs. The Court will address the

2

continuum evidence cited by Dr. Wagner and Dr. Tversky (Dr. Whittaker provides little

3

discussion of the continuum), and then address a basic question presented by these

4

experts – if GBCAs cause NSF, why doesn’t simple logic lead to the conclusion that they

5

also cause the range of symptoms labeled GDD?
1.

6

Dr. Wagner.

7

After opining that a continuum of symptoms is caused by GBCAs, Dr. Wagner

8

asserts that “[m]any individuals with different backgrounds agree on this point[.]”

9

Doc. 154-10 at 31. He then cites one source: the briefing document from the 2017

10

MIDAC conference. Id. at 38 ¶ 78. But the briefing document, which is part of the

11

record at Doc. 154-43 (and is not available at the website address provided by Dr.

12

Wagner), makes no mention of the three doctors Dr. Wagner names as holding the same

13

opinion – Drs. Pierre Desche, Gene Williams, and Peter Herscovitch. See Doc. 154-10 at

14

31-32.

15

Because Dr. Wagner’s continuum opinion is a very important issue in this case,

16

and the Court could find no mention of these doctors in the FDA briefing document, the

17

Court reviewed the entire transcript of the 2017 MIDAC conference. Sure enough, all

18

three of the named doctors spoke at the conference, and each one mentioned the word

19

“continuum.” But contrary to Dr. Wagner’s claim, none of them supported his opinion of

20

a continuum of symptoms caused by GBCAs.

21

Dr. Desche made this statement: “our conclusion is that brain hyperintensities and

22

NSF are, in fact, part of the same continuum from gadolinium retention to gadolinium

23

toxicity, and renal dysfunction acts as a catalyst.”

24

at 107; see also 113 (same). The continuum Dr. Desche mentions is from gadolinium

25

retention to gadolinium toxicity causing NSF. He does not say there is a continuum that

26

includes a wide range of milder symptoms that encompass GDD, as Dr. Wagner posits.

27
28
- 21 -

See MIDAC Trans., Doc. 156-7

1

Indeed, during his deposition Dr. Desche made clear that he was referring to retention

2

that leads to NSF, not some broader continuum. See Doc. 202 at 10-15 (sealed).

3

Dr. Williams made this statement at the MIDAC conference: “if you had

4

information on patients with varying degrees of renal impairment and you could tease out

5

the effect of the renal impairment itself, you might expect it to be a continuum. The idea

6

forwarded by the advisory committee member from a clinical pharmacology standpoint, I

7

think is a reasonable idea.” MIDAC Trans., Doc. 156-7 at 222-23. This statement is

8

classic hypothesis. Dr. Williams suggests that there “might” be a continuum detected if

9

the varying degrees of renal impairment could be separated. And he is talking about the

10

universe of patients with renal impairment, not individuals like Plaintiffs who have no

11

renal impairment. Even then, the most he says is that there “might” be a continuum and

12

it is “a reasonable idea.” Id. at 223. He does not endorse Dr. Wagner’s continuum

13

conclusion.

14

Dr. Herscovitch’s statement at the MIDAC meeting was this: “renal function is not

15

either normal or abnormal, but as one of our discussants pointed out, it’s a continuum.”

16

MIDAC Trans., Doc. 156-7 at 358. This statement has nothing to do with a continuum of

17

symptoms. It concerns a continuum of renal function. And Dr. Herscovitch, who chaired

18

the MIDAC conference, actually made statements that contradict Dr. Wagner’s medical

19

causation opinion: “I think there is fair uniformity that there is no evidence of a causal

20

relationship between the symptoms and signs in patients with normal renal function and

21

the retention of gadolinium.” Id. at 355. He also presciently addressed the kinds of

22

evidence primarily relied on by Plaintiffs’ experts, saying: “the FAERS data and other

23

anecdotal reports really perhaps raise questions, but in themselves do not have a scientific

24

foundation for reaching any conclusions.” Id. at 356.

25

In short, assuming Dr. Wagner was citing to the MIDAC conference as support for

26

his claim that “[m]any individuals with different backgrounds agree” with his posited

27
28
- 22 -

1

continuum of symptoms caused by GBCAs, he is simply incorrect. That Dr. Wagner

2

would make such a strong but incorrect claim is concerning.
2.

3

Dr. Tversky.

4

Dr. Tversky does not cite other authorities to support his claim of a continuum.

5

He instead provides this rationale: “it was noted that in patients with NSF there were

6

early signs of the disease that progressed over time to these more characteristic later

7

manifestations. This shows that there is a range or continuum of signs and symptoms

8

caused by gadolinium toxicity exposure.” Doc. 154-9 at 3. This assertion – that there is

9

a range of symptoms in NSF patients – is used by Dr. Tversky to conclude that the same

10

range of symptoms must exist in everyone else exposed to GBCAs, including renally-

11

healthy patients. But this leap rests on nothing more than Dr. Tversky’s say-so. He cites

12

no other authority to support the continuum. And the Court notes that at least one

13

speaker at MIDAC’s conference, Dr. Nicholas Dainiak, characterized this assertion –

14

“that the brain deposits may occur prior to eventual evolution to NSF” – as a mere

15

“hypothesis.” MIDAC Trans., Doc. 156-7 at 316.

16

Later in his report Dr. Tversky notes that case reports suggest milder symptoms

17

for GDD patients than NSF patients, and he compares those milder symptoms to the early

18

stages of NSF, again suggesting it’s all one continuum. Doc. 154-9 at 8. This reasoning

19

rests on the reliability of the case reports he cites. As explained in the discussion below,

20

the Court finds the case reports insufficiently reliable to support Dr. Tversky’s proposed

21

continuum.
3.

22

If GBCAs Cause NSF, Then Why Not GDD?

23

Dr. Tversky and Dr. Wagner assert, as do Plaintiffs generally, that NSF is a

24

misnomer – that the use of “nephrogenic” in the name of the disease is incorrect because

25

impaired kidneys don’t cause NSF, they simply allow gadolinium to accumulate in the

26

body to a point where it causes NSF. It is the gadolinium, not the kidneys, that causes

27

NSF.

Docs. 154-9 at 3-4 (Tversky Report), 154-10 at 16 (Wagner Report).

28
- 23 -

This

1

rationale leads logically to the conclusion that the accumulation of gadolinium in patients

2

with healthy kidneys will likewise cause the symptoms seen in GDD. But if it were that

3

simple, MIDAC, the FDA, and other organizations would have concluded that GBCAs

4

cause GDD. It is not that simple. As Dr. Anthony Fotenos, an FDA medical officer,

5

noted at the MIDAC conference, the vast majority of patients who receive GBCAs

6

develop no symptoms.

7

benefited from GBCAs without reported adverse reactions since the first drug in the class

8

was approved in 1988.” MIDAC Trans., Doc. 156-7 at 195; see also Endrikat et al.

9

(2016), Doc. 189-20 at 2 (most patients with severe renal impairment who receive even

10

multiple doses of GBCAs do not develop NSF). The mere receipt and even retention of

11

GBCAs clearly does not cause illness in most patients.

“[I]n the postmarketing setting, millions of patients have

12

And as many comments at the MIDAC conference make clear, the causative

13

mechanism of NSF, although linked to gadolinium, it still not understood. Dr. Wagner

14

himself provided this explanation at the conference:

15
16
17
18
19
20

In 2006, when [NSF] was clearly linked to gadolinium, we know
that gadolinium is the cause, so that the kidney is a risk factor – renal
insufficiency is a risk factor – but the kidney per se is not really causing the
disease.
There is now evidence that gadolinium is deposited in the central
nervous system. I believe that the central nervous deposition warrants
more study, and I am going to show that gadolinium-based contrast agents
are biologically active.

21
22
23
24
25
26
27

Very little is known about the metabolism of these agents, their
biologic effects, and the implications of retaining gadolinium in the tissues.
The toxic effects and the mechanisms of how they impart their
pathophysiology is a major gap in our knowledge.
If we understand how the disease processes occur, we are going to
know quite a bit more for stratifying patients who are at risk, and it will add
to our future knowledge. How these different agents behave once they
enter the body is an active area of investigation.

28
- 24 -

1

MIDAC Trans., Doc. 156-7 at 46-47 (emphasis added). Thus, Plaintiffs’ own causation

2

expert stated less than two years ago that much is not known about the effects of GBCA

3

retention and that further investigation is needed. He also told MIDAC that “dechelation

4

of gadolinium is a hypothetical pathologic mechanism,” and that “[s]tudies concerning

5

the biologic effects of rare earth minerals in general and their retention in human organs

6

are in the nascent stage, and the science on this topic is at ground zero.” Id. at 60.

7

In fact, uncertainty about the precise mechanism by which GBCAs causes NSF is

8

recognized by many sources. Dr. Wagner stated at the MIDAC conference that “there are

9

risk factors out there that have not been defined.” Id. at 135. One of the FDA medical

10

officers stated: “I also want to note – and this is similar to what Dr. Wagner said – while

11

consideration of the comparative exposure to gadolinium from each GBCA is important,

12

patient factors in addition to renal function are likely to play an important role[.]” Id.

13

at 180. Dr. Semelka, the author of several papers cited by Plaintiffs, suggested an

14

entirely different pathology at the MIDAC conference: “My opinion is [that GDD] is a

15

genetic disease of the immune system[.]” Id. at 302. Runge (2018) states that NSF is a

16

manifestation of toxicity from gadolinium released by MRI contrast media, but that the

17

illness probably develops in the presence of “cofactors.” Doc. 189-17 at 5. Endrikat et

18

al. (2016) notes that multiple cofactors for NSF have been proposed, including metabolic

19

acidosis, vascular surgery, or treatment with erythropoietin. Doc. 189-20 at 2; see also

20

Wagner et al. (2016), Doc. 189-21 at 5-6 (noting that many patients with end-stage renal

21

disease on chronic dialysis do not acquire the disease despite several exposures to

22

GBCAs, and some individuals develop NSF without gadolinium exposure). Finally, Dr.

23

Tversky himself asserts that there are other potential contributing factors to NSF,

24

including the activation of free radicals, increased inflammatory chemicals, and the

25

induction of complex immunological cascades. Doc. 154-9 at 7.

26

If the precise causes of NSF are not yet understood, one cannot simply conclude

27

that if GBCAs contribute to NSF in renally-impaired patients they must also cause GDD

28
- 25 -

1

in renally-healthy patients.

2

causation is more complicated than such a simple proposition.
4.

3

As MIDAC and other organizations have recognized,

Continuum Conclusion.

4

Plaintiffs have not shown by a preponderance of the evidence that the continuum

5

theory posited by Drs. Wagner and Tversky is the product of reliable principles and

6

methods applied reliably. The Court will next look to the actual categories of evidence

7

relied on by the causation experts.

8

D.

Individual Case Reports and Surveys.

9

The most common form of evidence cited by the causation experts consists of

10

individual case reports (descriptions of the experiences of single patients) and surveys of

11

individuals who believe they developed adverse health effects after receiving GBCAs.

12

See Docs. 154-10 at 17, 21-26, 28, 30 (Wagner Report); Doc. 154-9 at 8 (Tversky

13

Report); Doc. 154-11 at 25, 30 (Whittaker Report). Some of these reports are from the

14

FDA’s FAERS database and others are from published articles, but all of them constitute

15

the experience of single patients, usually as reported by the patients with no complete

16

study of the patients and no control group.

17

“Such case reports are not reliable scientific evidence of causation, because they

18

simply described reported phenomena without comparison to the rate at which the

19

phenomena occur in the general population or in a defined control group; do not isolate

20

and exclude potentially alternative causes; and do not investigate or explain the

21

mechanism of causation.” Casey v. Ohio Med. Prods., 877 F. Supp. 1380, 1385 (N.D.

22

Cal. 1995). As the Eleventh Circuit has explained:

23
24
25
26
27

[C]ourts must consider that case reports are merely accounts of
medical events.
They reflect only reported data, not scientific
methodology. Some case reports are a very basic form report of symptoms
with little or no patient history, description of course of treatment, or
reasoning to exclude other possible causes. The contents of these case
reports were inadequate, even under the plaintiffs’ expert’s standards, to
demonstrate a relationship between a drug and a potential side effect.

28
- 26 -

1
2
3
4
5
6
7
8
9
10
11
12

Some case reports do contain details of the treatment and differential
diagnosis. Even these more detailed case reports, however, are not reliable
enough, by themselves, to demonstrate the causal link the plaintiffs assert
that they do because they report symptoms observed in a single patient in
an uncontrolled context. They may rule out other potential causes of the
effect, but they do not rule out the possibility that the effect manifested in
the reported patient’s case is simply idiosyncratic or the result of unknown
confounding factors.
Rider v. Sandoz Pharm. Corp., 295 F.3d 1194, 1199 (11th Cir. 2002) (citation omitted).
The Federal Judicial Center Reference Manual on Scientific Evidence includes
this caution about case reports:
Anecdotal reports may be of value, but they are ordinarily more helpful in
generating lines of inquiry than in proving causation. . . . Anecdotal
evidence usually amounts to reports that events of one kind are followed by
events of another kind. Typically, the reports are not even sufficient to
show association, because there is no comparison group.

13
14

2011 WL 7724256 at *4 [hereinafter Reference Manual].7

15

And the FDA itself has identified these limitations for its FAERS database:

16

Yes, FAERS data does have limitations. First, there is no certainty
that the reported event (adverse event or medication error) was due to the
product. FDA does not require that a causal relationship between a product
and event be proven, and reports do not always contain enough detail to
properly evaluate an event. Furthermore, FDA does not receive reports for
every adverse event or medication error that occurs with a product. Many
factors can influence whether an event will be reported, such as the time a
product has been marketed and publicity about an event. There are also
duplicate reports where the same report was submitted by a consumer and
by the sponsor. Therefore, FAERS data cannot be used to calculate the
incidence of an adverse event or medication error in the U.S. population.

17
18
19
20
21
22
23
24
25
26
27
28

7

The Court will refer frequently in this order to the Reference Manual. The FJC
is the research and education arm of the federal judicial system. Established by Congress
in 1967, the FJC’s duties include conducting and promoting education for federal judges
and other court personnel. Federal Judicial Center, https://www.fjc.gov/ (last visited
July 22, 2019). The Reference Manual is now in its third edition, has been found reliable
by many federal courts, and has been cited in hundreds of court decisions. See e.g.,
Comcast Corp. v. Behrend, 569 U.S. 27, 38 (2013); Elliott v. Google, Inc., 860 F.3d
1151, 1160 (9th Cir. 2017).
- 27 -

1

https://www.fda.gov/drugs/surveillance/fda-adverse-event-reporting-system-faers

(last

2

visited July 8, 2019); see also MIDAC Trans., Doc. 156-7 at 148 (FDA official

3

describing limitations of FAERS database).8

4

“Simply stated, case reports raise questions, they do not answer them.” McLain v.

5

Metabolife Int’l, Inc., 401 F.3d 1233, 1245 (11th Cir. 2005); see also Cloud v. Pfizer, 198

6

F. Supp. 2d 1118, 1133-34 (D. Ariz. 2001); Jones v. United States, 933 F. Supp. 894,

7

899-900 (N.D. Cal. 1996), affirmed 127 F.3d 1154 (9th Cir. 1997), cert. denied 524 U.S.

8

946 (1998). Indeed, Dr. Wagner admitted in his deposition that case reports do not

9

provide reliable scientific evidence of causation because they are anecdotal reports of

10

occurrences, without support from the scientific method. Doc. 154-6 at 42.

11

Surveys are even less reliable. They consist of unscreened answers from people

12

who believe they suffer from GBCA-caused illnesses, often provided anonymously.

13

Unlike case reports, the survey responses generally are not made by physicians. They

14

include statements from medically-untrained individuals, with no verification of exposure

15

to GBCAs or physical symptoms, and certainly no control group. And yet all three

16

causation experts rely on surveys without providing any discussion or acknowledgement

17

of their limitations. See Doc. 154-10 at 18, 25, 28-29 (Wagner Report); Doc. 154-9 at 8

18

(Tversky Report); Doc. 154-11 at 24 (Whittaker Report).

19

Dr. Brent, a professor at the University of Colorado School of Medicine and a

20

member of MIDAC during the 2017 conference, provided this relevant comment about

21

relying on anecdotal information:

22
23
24
25
26
27
28

8

The Reference Manual provides similar cautions about the FAERS database:

It is a voluntary system with the following limitations: (1) uncertainty that
the drug caused the reported event, (2) no requirement for proof of a causal
relationship between product and event, (3) insufficient detail to evaluate
events, (4) incomplete reporting of all adverse events, and (5) inability to
determine the incidence of an adverse events because the actual number of
patients receiving a product and the duration of use of those products are
unknown.
2011 WL 7724263 at *28.
- 28 -

1
2
3
4
5
6
7
8
9
10
11

We are in an unfortunate situation here. This is not an easy
scientific question. What we do know without any question is that
gadolinium is retained, and it’s retained in at least some patients
irrespective of how good their renal function is.
What we really don’t know is whether that has any physiological
consequences at all. We have some anecdotal data – some of it is very
powerful, some of it is very emotionally concerning – that it may be
harmful to some people, but that is fundamentally anecdotal data.
Sometimes following our emotions in this kind of data can lead us in
unscientific directions.
What we really need to do is to focus our attention on determining
what, if any, the adverse consequences are of this gadolinium retention
before we decide how we should act on it or to what degree we should be
making restrictions, to what degree we should be changing our clinical
practice.

12
13

MIDAC Trans., Doc. 156-7 at 327-28.

14

As discussed in more detail below, the Court cannot conclude by a preponderance

15

of the evidence that the causation experts’ reliance on individual case reports and

16

surveys, including those in the FAERS database, satisfies the Rule 702 requirements of

17

reliable principles and methods applied reliably.

18

E.

19

Animal studies clearly can be relevant in evaluating the effects of substances on

20

Animal Studies.

humans, but there are limitations here also.

21

First, experts must consider the differences between the animals studied and

22

human anatomy and physiology. As the Reference Manual explains: “The expert should

23

review similarities and differences between the animal species in which the compound

24

has been tested and humans. This analysis should form the basis of the expert’s opinion

25

regarding whether extrapolation from animals to humans is warranted.”

26

7724262 at *16. Thus, while animal studies are not per se inadmissible, there must be a

27

basis for extrapolating them to the human population. See Daubert v. Merrell Dow

28
- 29 -

2011 WL

1

Pharm., Inc. (Daubert II), 43 F.3d 1311, 1319 (9th Cir. 1995); see also Sanderson v. Int’l

2

Flavors & Fragrances, Inc., 950 F. Supp. 981, 997 (C.D. Cal. 1996) (“[I]n order for

3

animal studies to be admissible to prove causation in humans, there must be good

4

grounds to extrapolate from animals to humans[.]”). In this case, none of Plaintiffs’

5

experts explains why the animal studies they cite can reliably be extrapolated to humans

6

with normal kidney function.

7

Second, experts should account for any unique characteristics of the study that

8

might make it less relevant to humans or to the question addressed in the causation

9

analysis. See Joiner, 522 U.S. at 144-45 (expert opinions excluded where animal studies

10

were “so dissimilar to the facts presented in . . . litigation”); Allison v. McGhan Med.

11

Corp., 184 F.3d 1300, 1313-14 (11th Cir. 1999) (expert failed to “explain the correlation

12

of the results . . . of rat studies in which the rats were directly injected with silicone to

13

symptoms in a human patient where [the silicone implants] remained intact”). In this

14

case, Plaintiffs rely on a number of animal studies that included procedures that might

15

make the studies less relevant to whether GBCAs can cause GDD in patients with healthy

16

kidneys.

For example, several of the studies involve rodents whose kidneys were

17

removed.

Rasschaert, et. al. (2018), Doc. 189-14; Do et. al. (2014), Doc. 191-7;

18

Doc. 154-11 at 30-31 (Whittaker Report). In a study of dogs, the blood-brain barrier was

19

deliberately breached. Doc. 154-10 at 18 (Wagner Report); Doc. 154-11 at 28 (Whittaker

20

Report). In another, GBCAs were injected directly into the animal’s brain. Doc. 154-10

21

at 18 (Wagner Report); see also Doc. 154-11 at 28 (Whittaker Report) (discussing

22

negative effects of gadolinium when allowed direct access to the brain). And in another,

23

GBCAs were administered to pregnant mice, and their babies were studied to determine

24

whether gadolinium can affect animals in utero.

25

Doc. 154-11 at 29 (Whittaker Report). Plaintiffs’ experts do not address these unique

26

features or explain why, even with these features, the studies are relevant to GBCA

27

exposure in humans with healthy kidneys.

28
- 30 -

Doc. 154-10 at 22 (Wagner Report);

1

Third, dose matters when extrapolating from animal studies. As the Reference

2

Manual logically explains, “the dose at which mercury causes [an] effect in laboratory

3

animals is modified by many internal factors, and the exact dose-response curve may be

4

different from that for humans.” 2011 WL 7724262 at *8; see also Schudel v. Gen. Elec.

5

Co., 120 F.3d 991, 997 (9th Cir. 1997), (expert inappropriately extrapolated data from

6

different dosages, exposures, and chemical compounds without explaining why the

7

extrapolation was scientifically acceptable), abrogated on other grounds by Weisgram v.

8

Marley Co., 528 U.S. 440, 453 (2000); Turpin v. Merrell Dow Pharm. Inc., 959 F.2d

9

1349, 1360-61 (6th Cir. 1992) (expert failed to clarify why the varying doses of the drug

10

and other chemicals given to rats, rabbits, and in vitro animal cells should be extrapolated

11

to humans). In this case, many of the animal studies cited by Plaintiffs’ experts involved

12

very high and continuous doses of GBCAs, gadolinium, or gadolinium chloride (which

13

Defendants’ assert is never used in GBCAs).

14

Doc. 155-13; Doc. 154-9 at 6 (Tversky Report); Doc. 154-10 at 12, 29 (Wagner Report);

15

Doc. 154-11 at 18-19, 30-31 (Whittaker Report). And yet Plaintiffs’ experts never

16

address the studies’ doses or their use of gadolinium chloride, and never explain why the

17

studies are relevant to the causation question in light of these differences.9

See, e.g., Lohrke et. al. (2017),

18

Fourth, Plaintiffs’ experts cite some studies without acknowledging that the

19

studies’ results or the authors’ conclusions are inconsistent with the experts’ opinions.

20

These will be discussed below in connection with the individual experts.

21

The Court cannot find that the experts are using reliable principles and methods, or

22

using them reliably, when they fail to say anything about the kinds of animals used in the

23

studies, the unique characteristics of the studies, or the substances and doses used.

24

Without such analysis, the relevancy of the animal studies is connected to the experts’

25

opinions merely by their say-so, which is not sufficient. Plaintiffs have provided no basis

26
27
28

An FDA medical officer explained at the MIDAC conference that “biodistribution of gadolinium chloride is very different to the GBCAs.” MIDAC Trans.,
Doc. 156-7 at 132
9

- 31 -

1

for the Court to conclude by a preponderance of the evidence that use of the animal

2

studies in this case constitutes a reliable principle or method applied reliably.

3

The Court also notes that all of the animal studies cited by Plaintiffs’ experts are

4

known to MIDAC, the FDA, and other regulatory and medical bodies. And yet none of

5

these organizations has found the studies sufficient to conclude that GBCAs cause GDD.

6

F.

7

As with animal studies, dose matters when considering in vitro studies. The

8
9
10
11
12

In Vitro Studies.

Reference Manual explains:
Cellular and tissue culture research can be particularly helpful in
identifying mechanisms of toxic action and potential target-organ toxicity.
The major barrier to the use of in vitro results is the frequent inability to
relate doses that cause cellular toxicity to doses that cause whole-animal
toxicity. In many critical areas, knowledge that permits such quantitative
extrapolation is lacking.

13
14

2011 WL 7724262 at *17; see also Turpin, 959 F.2d at 1360-61; In re Rezulin Prods.

15

Liab. Litig., 369 F. Supp. 2d 398, 428-29 (S.D.N.Y 2005) (“[C]aution must always be

16

used in extrapolating results in tissue culture to effects in live humans.”); In re Bausch &

17

Lomb, Inc. Contact Lens Solution Prods. Liab. Litig., No. 2:06-MN-77777-DCN, MDL

18

No. 1785, 2009 WL 2750462, at *12 (D.S.C. Aug. 26, 2009) (“In vitro tests generate

19

hypotheses but lack sufficient reliability, standing alone, to demonstrate causation in

20

humans.”).

21

Plaintiffs’ causation experts cite various in vitro studies.

Doc. 154-9 at 5-6

22

(Tversky Report); Doc. 154-10 at 12, 15, 18 (Wagner Report); Doc. 154-11 at 18, 27,

23

30-31 (Whittaker Report). But none of the experts discusses the doses used in the studies

24

or explains how those doses can reliably be extrapolated to the effects of GBCAs in

25

humans with healthy kidneys.

26

inappropriately extrapolated data from different dosages, exposures, and chemical

27

compounds without explaining why the extrapolation was scientifically acceptable). In

See Schudel, 120 F.3d at 997 (finding that expert

28
- 32 -

1

the absence of such analysis and explanation, the Court cannot find by a preponderance

2

of the evidence that the experts’ use of in vitro studies constitutes a reliable principle or

3

method used reliably.

4

The Court also notes that all of the in vitro studies cited by Plaintiffs’ experts are

5

known to MIDAC, the FDA, and other regulatory and medical bodies. And yet none of

6

them has found the studies sufficient to confirm the hypothesis that GBCAs cause GDD.

7

G.

Incorrect Application of the Bradford Hill Criteria.

8

In order to determine whether a specific compound causes a particular malady, an

9

expert must use valid scientific methodology. Rains v. PPG Indus., Inc., 361 F. Supp. 2d

10

829, 835 (S.D. Ill. 2004).

Many scientists use the Bradford Hill criteria for their

11

causation analysis. Id. These criteria were presented in a 1965 lecture by Sir Austin

12

Bradford Hill. Doc. 154-19. All three of Plaintiffs’ causation experts mention the

13

Bradford Hill criteria, and two of them – Drs. Tversky and Whittaker – purport to apply

14

the criteria. But as will be explained below, these experts do not apply the criteria

15

correctly. Thus, even though the Bradford Hill criteria represent reliable principles and

16

methods, they have not been applied reliably to the facts of this case as required by

17

Rule 702(d).

18

H.

Other Daubert Considerations.

19

Several other considerations identified in Daubert are relevant here. None of the

20

experts’ opinions “has been subjected to peer review and publication.” Daubert, 509

21

U.S. at 593. None has identified a “known or potential rate of error[.]” Id. at 594. And

22

their principles and methods for finding GBCA causation of GDD have not been

23

accepted by other experts, regulatory bodies, or professional associations. As a result,

24

Daubert explained, those principles and methods “may properly be viewed with

25

skepticism.” Id. (quotation marks and citations omitted).

26

The Advisory Committee Note to Rule 702 identifies additional relevant

27

considerations. One is whether the experts are testifying about matters growing naturally

28
- 33 -

1

and directly out of research they have conducted independent of the litigation, or whether

2

they have developed their opinions expressly for purposes of testifying. See Fed. R.

3

Evid. advisory committee’s note to 2000 amendment (citing Daubert II, 43 F.3d at 1317).

4

Dr. Wagner satisfies this criterion, as he has done research on GBCAs and their

5

connection to adverse health effects, but Drs. Tversky and Whittaker have not done such

6

research and their opinions appear to have been prepared solely for this litigation.

7

I.

Conclusion.

8

These general problems with the causation experts’ opinions lead the Court to

9

conclude that they are not admissible under Rule 702. To provide more detail, the Court

10

will address each expert separately.

11

VI.

Rule 702(c) and (d) – Specific Discussion of Dr. Wagner’s Opinions.

12

A.

Opinions.

13

Dr. Wagner begins with the foundational facts described above. He opines first

14

that gadolinium in many free forms is toxic to humans, citing studies of gadolinium dust

15

and gadolinium chloride. Doc. 154-10 at 2. He notes that most GBCAs are eliminated

16

from the body by the kidneys, but that there is always retained gadolinium in the body,

17

even with normal renal function. Id. at 10. He identifies various human and animal

18

studies that show GBCA retention, including in patients with and without renal

19

insufficiency. Id. at 12. He opines that there is no debate that, in the population of

20

patients with renal impairment, GBCAs cause NSF. Id. at 15. He asserts that GBCAs

21

activate fibroblasts, “which are the cells responsible for multiple diseases, including

22

scleroderma and systemic fibrosis.” Id. Although the severity of systemic fibrosis may

23

increase with repeated GBCA exposure, he states that there does not appear to be a

24

minimum threshold dose for the appearance or occurrence of NSF. Id. at 16.

25

Dr. Wagner then opines that GBCAs cause a variety of health problems. Id.

26

Throughout his discussion of these various symptoms, he mixes studies regarding NSF

27

with case reports, surveys, animal data, and in vitro data.

28
- 34 -

1

For neurological effects, Dr. Wagner discusses a study involving the direct

2

injection of GBCAs into the brains of rats, noting that the rats developed a number of

3

behavioral and neurologic problems.

4

injection of GBCAs could increase seizures in dogs when the permeability of the blood-

5

brain barrier was altered. Id. And one rat study showed detectable metabolic changes

6

from GBCA administration. Id. But Dr. Wagner does not explain why the results of

7

these studies can reliably be extrapolated to humans; he does not discuss doses in the

8

studies and how they relate to doses in humans; and he does not explain why direct brain

9

injections in the rats and alteration of the blood-brain barrier in the dogs provide reliable

10

Id.

Another study showed that intravenous

information about the effects of GBCAs in humans with healthy kidneys. Id. at 17-18.

11

Dr. Wagner turns next to “radiographic abnormalities.” Id. at 19. He cites results

12

from the Burke (2016) anonymous survey where 77.6% of participants noted headaches,

13

vision changes, hearing changes, and other head and neck symptoms that they attributed

14

to GBCA exposure. Id. He then cites various human studies showing that gadolinium

15

was retained in the brains of patients who received GBCAs, and that many of the patients

16

had abnormal brain signals. Id. at 19-20. He also cites a case report of a 55-year-old

17

woman who had numerous GBCA exposures and normal renal function, and yet

18

developed symptoms of intense pain in her forearms, hands, feet, and toes two to three

19

weeks after the last GBCA dose. Id. at 20. From this evidence, Dr, Wagner opines that

20

GBCAs can cause neurological effects in patients with normal kidneys. But he does not

21

explain why reliance on an anonymous survey and a single case report constitutes a

22

reliable scientific method, and he never links the abnormal brain signals observed in

23

some GBCA recipients to any illnesses, whether exhibited in the study patients or in

24

Plaintiffs. Id. In fact, the brain signal studies were presented at the MIDAC conference

25

and discussed extensively, and yet MIDAC concluded that they did not establish a causal

26

link between GBCAs and any illnesses. See MIDAC Trans., Doc. 156-7 at 2-402 (scores

27

of references to brain studies).

28
- 35 -

1

Dr. Wagner then discusses behavioral effects. Id. at 21. He notes a study of

2

gadolinium transfer from pregnant mice to developing fetuses which found that the mice

3

pups demonstrated behavioral abnormalities seven to nine weeks after birth. Id. at 22.

4

He discusses three case reports where patients were accidentally injected with gadolinium

5

in their spinal canal and experienced seizures, a vegetative state, confusion, nausea,

6

drowsiness, blurred vision, and delirium. Id. at 22-23. He details another case of a

7

woman who received multiple doses of GBCAs and experienced “gadolinium

8

encephalopathy” characterized by drowsiness, fluctuating levels of consciousness, and

9

mild left-sided paralysis. Id. at 23. In another case report, a 58-year-old woman received

10

one dose of a GBCA and for three months afterward experienced intense, burning pain in

11

her torso, arm, and leg; rashes; nausea; bladder pain; clouded mentation; disorientation;

12

and a metabolic taste. Id. at 23. But Dr. Wagner never explains why the transfer of

13

gadolinium from pregnant mice to their pups, or the accidental injection of GBCAs into

14

the spinal canals of patients, or two case reports with their wide and differing range of

15

symptoms can, through reliable scientific principles and methods, be extrapolated to

16

show that GBCAs cause GDD in patients with healthy kidneys. Nor does he explain why

17

this section of his report on behavioral effects has any relevance to Plaintiffs, none of

18

whom claims adverse behavioral effects from GBCAs.

19

Dr. Wagner addresses “skin diseases” by starting with a discussion of NSF. Id.

20

at 22. Symptoms of NSF may include sclerosis, puckering, pruritis, burning, extremity

21

swelling, Raynaud’s phenomenon (discoloration of fingers and toes), muscle induration,

22

joint flexion contractures, and neuropathy. Id. at 23-24. He then cites two studies where

23

researchers found biological markers of NSF in patients with normal renal function after

24

GBCA exposure and several case reports involving skin issues. Dr. Wagner appears to

25

cite these studies as proof that GBCAs can cause Plaintiffs’ skin issues.

26

Dr. Wagner first notes that gadolinium-associated plaques have been reported in

27

patients without kidney disease, that these plaques are sometimes asymptomatic and

28
- 36 -

1

sometimes pruritic and burning, and that specimens taken from these plaques contained

2

specific CD68- and XIIIa-factor spindle-shaped fibrocytes and CD34-postive cells, all of

3

which are distinctive markers for patients with NSF. Id. at 24. These markers have also

4

been detected in preclinical models showing that bone marrow-derived fibrocytes are

5

involved in mediating fibrosis. Id.

6

The plaque study involved one patient without kidney failure and one patient who

7

had “chronic renal insufficiency,” but not renal disease or NSF. See Gathings et al.

8

(2015), Doc. 191-11. The patient with normal kidney function suffered from a pruritic,

9

burning rash on both hands. Id. at 2. He had a history of cervical spine disease, prostate

10

carcinoma, hypothyroidism, varicella, and gallstones, and had been exposed to multiple

11

GBCA doses. Id. He had six small, red, ring-shaped plaques on the back of his hands.

12

Id. at 2-3. The patient with renal insufficiency had a “slowly enlarging, asymptomatic

13

tan-brown plaque on the right anterior lower leg” and had been exposed to “an unknown

14

dose of gadolinium for several contrast-enhanced radiologic procedures.” Id. at 3. The

15

authors of the study noted that similar plaques had only been described in patients with

16

NSF and that finding them in these patients supported the hypothesis that the plaques

17

were not pathognomonic to NSF. Id. As a result, the authors coined a new clinical

18

entity – “gadolinium-associated plaques” (“GAP”) – and concluded that gadolinium

19

exposure should be considered in any case of multiple or solitary plaques associated with

20

distinctive histopathologic findings of sclerotic bodies regardless of NSF diagnosis or

21

renal function. Id. at 5.

22

Dr. Wagner’s asserts that his second cited case study identified “specific markers

23

for GBCA-induced systemic fibrosis” – NSF – in a patient with normal renal function

24

who underwent 61 contrast-enhanced MRIs over 11 years.

25

According to Dr. Wagner, a skin biopsy demonstrated the presence of gadolinium, and

26

the patient had joint contractures – a clinical criterion for NSF – in his neck and all four

27

extremities to the point that he was immobilized. Id. The study involved a patient with a

28
- 37 -

Doc. 154-10 at 25.

1

history of glioblastoma of the left temporal lobe, seizures, and hypothyroidism. Roberts

2

et. al. (2016), Doc. 187-32 at 4. The authors noted that the patient had “no skin issues,

3

with no complaints of rashes, pruritus, sclerosis of the skin, discoloration, or other

4

changes.” Id. Skin biopsies revealed normal skin and subcutis with “no evidence of

5

increased dermal cellularity, fibrosis, abnormal collagen bundles, osseous metaplasia, or

6

abnormal numbers of fibroblast like cells or macrophages.” Id. at 7. But there was

7

increased CD34 reactivity “indicating inflammation and/or tissue injury.” Id.

8

The authors concluded that while gadolinium was present in the samples, the

9

“form of the gadolinium retained” was unclear, and it was also unclear which specific

10

gadolinium species deposited in the skin. Id. at 10. They noted that the patient’s joint

11

contractures were of unknown cause, and “without a joint biopsy, a definite association

12

of the joint contractures with the high levels of gadolinium could not be confirmed or

13

excluded.” Id. The authors also concluded that, while they found the accumulation or

14

retention of gadolinium in the skin, long-term consequences were unknown. Id.

15

The Court assumes that Dr. Wagner cites these two studies in support of his

16

continuum hypothesis. That is, if distinct biomarkers of NSF are found in patients with

17

normal renal function, it is evidence of a milder GDD-form of NSF due to GBCA

18

retention. The Court finds Dr. Wagner’s use of these studies problematic for several

19

reasons.

20

First, Dr. Wagner does not fully explain what these biomarkers are or how they fit

21

into his opinions. The Court is left to assume their significance. He indicates that CD34

22

cells have been historically associated with NSF lesions, but the Roberts study stated that

23

CD34 is associated with inflammation or tissue injury, which, the Court can only

24

conclude absent further explanation, may or may not be associated with NSF. The

25

authors of the Roberts study do not mention NSF in connection with CD34.

26

although Dr. Wagner states that the study demonstrated “specific markers” of NSF, he

27

does not fully explain what the specific markers are, and the authors do not discuss any

28
- 38 -

And

1

specific NSF markers, instead concluding that connection between their patient case

2

study and NSF is indeterminate.

3

Second, these are single case studies, which, as explained above, are incapable of

4

proving general causation. The plaque study contains only one patient with normal renal

5

function and exposure to cumulative GBCA doses. The Roberts study considered only

6

one patient as well. Both patients had histories of other serious medical issues that were

7

not ruled out as causes for the plaques or joint contractures. The Court finds it troubling

8

that Dr. Wagner represented the Roberts study as concluding that the patient’s joint

9

contractures were necessarily gadolinium related, when that was clearly not the case.

10

Third, assuming the identified markers suggest that gadolinium retention in

11

individuals with normal kidney function mimics a milder form of NSF, these case studies

12

at most support a hypothesis that the myriad skin issues identified by Dr. Wagner – skin

13

changes, burning sensations, thickening, rashes – are caused by gadolinium retention.

14

Doc. 154-10 at 32. As noted above, “case reports raise questions, they do not answer

15

them.” McLain, 401 F.3d at 1245. What is more, the Roberts study clearly noted that the

16

patient did not actually suffer from any skin changes or skin disorders. And the plaque

17

study was specifically limited to the development of sclerotic bodies associated with

18

gadolinium exposure. The authors did not attempt to use the findings to connect GBCA

19

exposure to any other type of skin issues.

20

Dr. Wagner then describes the case of a 58-year-old woman with normal renal

21

function who had one GBCA dose and, seven years later, noted skin discoloration,

22

tightness in the webs between her fingers, discoloration of the legs, and a red and rubbery

23

skin texture in the knee regions. Id. at 24. Another patient with normal renal function

24

and a history of multiple administrations of GBCAs began to experience symptoms two

25

to three weeks after exposure to a linear GBCA. Id. at 25. She experienced erythema

26

and warmth on the bottoms of her feet and the palms of her hands. These symptoms

27
28
- 39 -

1

remained for eight years after the GBCA exposure. Id. And Dr. Wagner cites the Burke

2

(2016) survey where over half of the participants noted “skin changes.” Id.

3

Again, these are case studies and an informal anonymous survey. They raise

4

questions about possible skin effects, but do not answer them. And Dr. Wagner seems to

5

acknowledge this point in his opinion. He concludes that the fact that GBCAs “are not

6

entirely physiologically inert, and even demonstrate the same disease-inducing traits as

7

witnessed in patients with [NSF] raises concern about the potential problems they pose

8

when retained long-term in any individual.” Id. Raising concern, of course, is not the

9

same as proving causation.

10

Dr. Wagner notes additional effects observed in infants exposed to gadolinium in

11

utero. Infants exposed during the first trimester had higher risks for connective tissue

12

diseases, rheumatological, inflammatory, and infiltrative skin conditions. Id. Rates of

13

stillbirth and neonatal deaths increased when examining GBCA exposure during any

14

trimester of pregnancy. Id. at 25. GBCAs also had an effect on cultured skin cells,

15

causing an increase in growth of human skin cells over one to three days. Id. at 26. Dr.

16

Wagner does not explain how the in utero exposure relates to patients with healthy

17

kidneys, nor does he address why the in vitro study of skin growth, and its doses, can be

18

reliably extrapolated to humans with no renal impairment.

19

Dr. Wagner next turns to joint and bone pain associated with GBCAs. He cites a

20

case report of a 23-year old man with a history of end-stage renal disease who suffered

21

from progressive ankle pain, stiffness, and worsening mobility. Id. at 26. Another case

22

report detailed three NSF patients who experienced “diffuse osteopenia, joint

23

contractures, intra-articular and periarticular soft tissue calcification, mature heterotopic

24

bone formation, joint ankylosis, and smooth periosteal reaction along regional long

25

bones.” Id. at 27. He asserts that other studies have indicated that gadolinium is retained

26

and incorporated into the bone following GBCA exposure, which implies that the metal

27

will only be liberated when the bone undergoes remodeling. Id. at 26-27. In the Burke

28
- 40 -

1

(2016) survey, 77.6% of respondents noted bone or joint pain, or both. Id. at 28. Dr.

2

Wagner does not explain why case reports of specific NSF patients can reliably be

3

extrapolated to GDD, how gadolinium retention in bone leads to a reliable conclusion

4

regarding the effects of GBCAs in patients like Plaintiffs, or why reliance on an

5

anonymous survey is appropriate.

6

Dr. Wagner also addresses heart, respiratory, kidney, and liver effects, largely

7

relying on NSF data. He notes that gadolinium has been found in the hearts of several

8

NSF patients. Id. at 28. Gadolinium was also detected in the lungs, kidneys, and liver of

9

a dog two weeks after GBCA administration. Id. at 28-29. Case reports of NSF patients

10

demonstrate experiences with diaphragmatic fibrosis and lung calcifications. Id. at 28.

11

In the Burke (2016) survey, 21% of respondents noted difficulty breathing or other chest

12

symptoms. Id. at 28-29. Gadolinium and markers of fibrosis were also found in the liver

13

of GBCA-treated rats, and gadolinium chloride irreversibly paralyzed a guinea pig

14

intestine. Id. at 30. But Dr. Wagner does not explain why case reports from NSF

15

patients can reliably be applied to reach any conclusions, including conclusions regarding

16

patients who have healthy kidneys and do not have NSF; why the use of a survey or the

17

effects on a dog can be applied reliably to the same patients; or, for that matter, why

18

heart, respiratory, and liver effects have anything to do with Plaintiffs, none of whom

19

claim such ill effects.

20

In a broad category of “other health effects,” Dr. Wagner describes a rat study that

21

showed death of testicular cells and hypogonadism after repeated doses of GBCAs. Id.

22

at 30. In another rat study there was evidence of cell death in salivary glands after eight

23

days of daily doses of GBCAs. Id. But Dr. Wagner does not explain why these rat

24

studies can reliably be applied to humans, why the effects on testicles and salivary gland

25

cells have anything to do with Plaintiffs in this case, or why the doses received by the

26

animals in these studies are relevant to humans receiving GBCAs.

27
28
- 41 -

1

Based on these sources, Dr. Wagner reaches his conclusions: “There is clear

2

evidence that the phenomena noted in patients with normal renal function and [NSF] are

3

part of the same continuum of disease.” Id. “And based on what is currently known in

4

the medical and scientific communities, the consequence of long-term retention of a rare

5

earth metal within the cells of multiple critical organs has demonstrable and harmful

6

long-term health effects.” Id. at 31. The “symptoms known to be caused by these

7

injuries include joint pain, pain with movement, pain in extremities, burning skin

8

sensation, skin changes, skin thickening and rashes, clouded mentation, inflammation,

9

and bone pain.” Id. These symptoms affect patients with both normal renal function and

10

impaired renal function.

11

Although he devoted portions of his report to behavioral, heart, respiratory, liver,

12

testicle, and salivary gland effects, he does not include any of these in his final opinions,

13

perhaps because they are not claimed by Plaintiffs. One wonders, then, why they were

14

included in the report at all.

15

B.

Analysis.

16

As mentioned at various points above, Dr. Wagner never explains how he is using

17

reliable principles and methods, or how he is applying those principles and methods

18

reliably. His connection between GBCAs and the wide range of symptoms characterized

19

as GDD depends on animal studies, but he never explains why they are reliably applied to

20

humans without impaired kidneys, why their dose is analogous, or why they are relevant

21

notwithstanding unique characteristics of the studies such as the removal of kidneys from

22

mice or breaching of the blood-brain barrier in rodents and dogs. He depends on case

23

reports and surveys even though he admitted in his deposition that they do not provide

24

reliable scientific evidence of causation because they are anecdotal, without support from

25

the scientific method. Doc. 154-6 at 42. And he cites some in vitro studies, but never

26

shows why they can be extrapolated to Plaintiffs and the causation issue in this case.

27
28
- 42 -

1
2

These concerns are compounded by other problems with Dr. Wagner’s opinion.
1.

Other Studies.

3

Defendants note that there have been relevant human epidemiologic studies on this

4

issue and that Dr. Wagner never mentions them. Doc. 157 at 12. Defendants identify

5

several such studies.

6

McDonald (2017) was a published abstract for a study that was presented at

7

multiple medical conferences and at the 2017 MIDAC conference. Doc. 153 at 19. It

8

compared 1,315 patients who received GBCAs to 2,946 patients with no history of

9

GBCA exposure. The authors found no association between GBCA use and cognitive

10

decline, poor neuropsychological performance, or diminished motor performance. Id.;

11

see also Doc. 154-23.

12

Welk (2016) compared the risk of new-onset Parkinsonism (stiffness, tremor,

13

imbalance, slowness) between patients who received GBCA-enhanced MRIs and patients

14

who received unenhanced MRIs. The researchers controlled for kidney function and

15

found no association between GBCA exposure and the development of these symptoms.

16

Doc. 153 at 19; see also Doc. 154-28.

17

Parillo (2018) conducted a prospective cohort study of 1,088 MRI patients to test

18

what the authors called “the debated definition of gadolinium deposition disease (GDD).”

19

Doc. 154-26 at 2. Some of the patients received GBCA-enhanced MRIs and others did

20

not. The patients were questioned before and 24 hours after the MRIs. Id. at 3. The

21

questions concerned symptoms in the proposed definition of GDD, many of which are

22

claimed by Plaintiffs – central torso pain, arm or leg pain, bone pain, headache, skin

23

redness (any site of the body), fatigue, and mental confusion. Id. The study found that

24

most symptoms allegedly related to gadolinium retention manifested at similarly low

25

rates in both the GBCA-enhanced and unenhanced MRIs.

26

statistically significant increase of two proposed GDD symptoms – fatigue and mental

27
28
- 43 -

The study did find a

1

confusion – in patients that received GBCAs (about 17%) when compared to those who

2

did not (about 8%). The authors stated this conclusion:

3
4
5
6
7
8
9
10
11

this study shows that, using a robust methodology with a structured
questionnaire administered before and after the MRI scan, and with a
reasonable sample size, only 2 (namely, fatigue and mental confusion) of
the 7 proposed GDD-like symptoms survived after comparison of [MRI
patients that did and did not receive GBCAs]. These results suggest that
the proposed term “gadolinium deposition disease” should be revised.
Taking into consideration that persistent symptoms may be rarely observed
after exposure to GBCAs and that there is not, at least based on currently
reported cases and in our population, a dose-dependent relationship with
previous exposure to multiple doses of GBCAs, the more conservative term
“gadolinium-associated symptoms” could be proposed.
Id. at 7.

12

Mallio (2019) found no association between imaging evidence of gadolinium

13

retention and resting state brain functional imaging in patients with normal kidney

14

functions and Crohn’s disease. Doc. 153 at 20.

15

Cocozza (2019) found no evidence that gadolinium retention in the brains of

16

patients with multiple sclerosis and normal kidney function affects their clinical status

17

over time. Id.

18

Although it is true, as Plaintiffs claim, that some of these studies focused on

19

narrow populations and examined a narrow set of adverse effects, they are human studies

20

performed by medical professionals, as opposed to case reports, anonymous surveys,

21

animal studies, and in vitro studies, and they certainly are relevant to the general

22

causation question in this case. It is concerning that Dr. Wagner never mentions them.

23

2.

Retention Opinions.

24

Dr. Wagner fails to connect retained gadolinium in normal kidney patients to the

25

dechelated gadolinium that causes his noted adverse-health effects in NSF cases. He

26

cites evidence from the FDA that gadolinium is retained in the body, and his own studies

27

suggest that GBCAs are not entirely eliminated. Doc. 154-10 at 12-13. But he never

28
- 44 -

1

specifies – with reliable evidence – whether the gadolinium that is retained in the body is

2

in a chelated or dechelated form. His report refers to GBCAs, gadolinium, and free

3

gadolinium interchangeably, never specifying the form of the substance he claims causes

4

GDD. And when he cites studies, he seldom if ever identifies the type of gadolinium

5

found in tissues.

6

The acting chair at the MIDAC conference, Dr. Herscovitch, provided this

7

summary about the state of knowledge regarding how GBCAs are stored in the body:

8

“There is a lot of unknown, not only about the class of agents but the different chemical

9

forms, linear versus macrocyclic classes, but also whether there is free gadolinium or

10

gadolinium as it’s originally used in agents or perhaps bound to other macromolecules in

11

the body.” MIDAC Trans., Doc. 156-7 at 335. The MIDAC conference also suggested

12

that identifying the form of gadolinium can be very difficult, and that methods for doing

13

so are still being developed. Id. at 132-33, 186. This suggests that such identification

14

rarely is done, which would explain why it does not appear to be addressed in the studies.

15

And yet Dr. Wagner and the other causation experts readily opine that free gadolinium is

16

retained in the bodies of patients with healthy kidneys and causes GDD.

17

Dr. Wagner cites animal studies showing that linear GBCAs are more prone to

18

dechelate than macrocyclic GBCAs and discussing the nanoparticles in the kidney and

19

the skin in his rodent model. Doc. 154-10 at 15. But he does not explain how these

20

experiments relate to the method of gadolinium retention in patients with normal kidneys,

21

or what symptoms, if any, are associated with appearance of the nanoparticles. And an

22

FDA medical officer who reviewed the emerging science on gadolinium retention at the

23

MIDAC conference stated that although linear GBCAs appear to dechelate more readily,

24

“we don’t know the clinical significance of this.” MIDAC Trans., Doc. 156-7 at 187.

25

Further, even if Dr. Wagner could show that retained gadolinium is dechelated, he

26

does not explain how retention of small amounts of gadolinium can cause the adverse-

27

health effects he attributes to the gadolinium retention. He notes that GBCAs are the

28
- 45 -

1

essential cause of NSF, and that NSF symptoms can occur after even small doses of

2

GBCAs.

3

amounts of reported gadolinium retention in patients without renal failure must cause less

4

severe adverse effects on a continuum with NSF. But he never explains how NSF

5

patients retain gadolinium and whether renally-impaired retention is the same as normal-

6

kidney retention. Id. at 10-11.

Doc. 154-10 at 16.

He then extrapolates from these facts that the small

7

He also offers no clear guidance on the quantifiable differences in gadolinium

8

retention. The chart he inserts in his report, but barely mentions, demonstrates that in the

9

same timeframe an individual with stage-IV kidney disease may retain nearly 25% of

10

gadolinium, an individual with stage-III would be closer to 8% and a patient with normal

11

renal function would be closer to 5%. See id. at 10. He does not connect those retention

12

levels to his underlying theory, which is essentially a proportional relationship between

13

retention and adverse-health effects on an NSF continuum.

14

Dr. Wagner simply presents what is known of NSF and what is known about

15

gadolinium toxicity in a free form and assumes that this would be the explanation for any

16

symptoms reported by patients with normal kidney function after GBCA administration.

17

See id. at 19 (assuming, because GBCAs are retained in multiple and vital brain areas,

18

any neurologic symptoms manifested soon after GBCA exposure are caused by

19

gadolinium until proven otherwise), 28 (assuming, based on NSF research, that GBCAs

20

are the cause of bone and joint pain reported in patients). In a complex area of evolving

21

science, where regulatory agencies and medical researchers uniformly acknowledge

22

many unknowns, Dr. Wagner’s assumptions simply do not fill the gaps. Under Rule 702,

23

an expert’s ipse dixit cannot be used to connect existing data to an unproven conclusion.

24

Joiner, 522 U.S. at 146; see also Turpin, 959 F.2d at 1360-61 (“a jury should not be

25

asked to speculate on the issue of causation.”).

26
27
28
- 46 -

1

3.

Dr. Wagner’s MIDAC Testimony.

2

Dr. Wagner was the keynote speaker during the 2017 MIDAC conference. In that

3

setting, he made several statements quite inconsistent with his current opinions. He

4

stated that (1) central nervous system gadolinium toxicity warrants more study; (2) little

5

is known about the metabolism of GBCAs, their biologic effects, and the implications of

6

retained gadolinium; (3) the toxic effects and mechanisms of GBCAs are a major gap in

7

current knowledge; (4) experiments show that renal insufficiency is not required for

8

fibrosis to occur; (5) dechelation of gadolinium is a hypothetical pathologic mechanism;

9

(6) studies concerning the effects of gadolinium retention in human organs are in the

10

“nascent stage”; and (7) the science on this topic is at “ground zero.” Doc. 154-48

11

at 4-21. Referring to brain retention of gadolinium, Dr. Wagner also said: “We don’t

12

know what the toxic effects are of this retention, if there are any.” MIDAC Trans.,

13

Doc. 156-7 at 135.

14

The Supreme Court has instructed that an expert must “employ in the courtroom

15

the same level of intellectual rigor that characterizes the practice of an expert in the

16

relevant field.” Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 152 (1999). During

17

his deposition, Defendants asked Dr. Wagner what changed from his 2017 MIDAC

18

presentation, where dechelation of gadolinium was a hypothetical pathologic mechanism,

19

to his expert report in January 2019, where GBCAs are proven to cause disease through

20

gadolinium dechelation. Doc. 158-3 at 27. Defendants also asked why Dr. Wagner told

21

MIDAC that it was not known what health effects gadolinium retention had on patients,

22

and yet opined in his report that several injuries are caused by gadolinium retention. Id.

23

Dr. Wagner responded that, after the 2017 presentation, he took research results

24

showing that nanoparticles were forming in the skin and kidneys in his animal models to

25

a new colleague at the University of New Mexico. Id. at 27-28. The colleague pointed

26

out that the animals’ skins and kidneys were forming gadolinium-rich structures that

27

appeared similar to what happens when a lanthanide is in a phosphate free solution, and

28
- 47 -

1

as the particle attached to bacteria it would leach phosphate out of the cell membrane. Id.

2

at 28. Dr. Wagner explained that after seeing the nanoparticles in that light, he did not

3

know how the gadolinium could exist in that form and still be chelated. Id. But he did

4

not identify any study he undertook on the basis of his colleague’s apparently casual

5

observation, or how, if at all, he sought to validate the hypothesis.

6

Dr. Wagner also discussed 2018 papers which determined that gadolinium was

7

being liberated from linear chelates, and his own unpublished working papers that

8

showed GBCAs caused fibrocytes to infiltrate the kidneys and that the nanostructures

9

apparent in the cells are damaging the lysosomes, which is causing a cell signal that can

10

cause fibrosis. Id. at 30. But if these studies were pivotal in Dr. Wagner’s formation of

11

the causation opinions expressed in his expert report, it is curious that they are never

12

mentioned in his report. Id.

13

He also noted that he observed symptoms while treating patients. See Docs. 188

14

at 7, 154-6 at 30. But again, if patient observations are important to his current opinion,

15

it is odd that Dr. Wagner does not describe any of those observations in his report.

16

Defendants also note that (1) the images of the nanoparticles showing dechelation were

17

the same images he presented to the MIDAC conference, and (2) the studies he pointed to

18

were animal studies that merely showed that more gadolinium was retained after linear

19

GBCA administration than macrocyclic administration. Doc. 157 at 11-12.

20

The Court finds the stark inconsistencies between Dr. Wagner’s MIDAC

21

statements and his expert report to be very concerning, particularly since the data he cited

22

in his deposition as having changed his opinions are not mentioned in his report on those

23

opinions.

24

inadmissible, they do suggest that he has not employed the same level of intellectual rigor

25

to his opinions as he does in practice. Kumho Tire, 526 U.S. at 152.

While these inconsistencies alone are not sufficient to find his opinions

26
27
28
- 48 -

1
2

4.

General Acceptance in the Scientific Community.

Defendants argue that Dr. Wagner’s opinions are not generally accepted by the

3

regulatory and scientific communities.

4

acceptance is a factor the Court may consider in deciding whether an expert’s opinion is

5

sufficiently reliable to be admitted in evidence. Daubert, 509 U.S. at 594. The general

6

acceptance must be of the expert’s methodologies, not his conclusions. Id. But the

7

Advisory Committee Note to Rule 702 provides this important explanation:

8

Doc. 157 at 8.

As noted above, general

11

Under the amendment [to Rule 702], as under Daubert, when an expert
purports to apply principles and methods in accordance with professional
standards, and yet reaches a conclusion that other experts in the field would
not reach, the trial court may fairly suspect that the principles and methods
have not been faithfully applied.

12

Fed. R. Evid. 702 advisory committee’s note to 2000 amendment; see also Lust, 89 F.3d

13

at 598 (same).

9
10

14

Dr. Wagner’s causation conclusions are not shared by regulatory and medical

15

organizations that have examined the same evidence, including MIDAC, the FDA and

16

government regulators in Canada, Europe, Australia, Japan, and New Zealand, as well as

17

the National Institutes of Health, the American College of Radiology, the Radiological

18

Society of North America, the American Society of Neuroradiology, the Canadian

19

Association of Radiologists, and the International Society for Magnetic Resonance in

20

Medicine. Such a broad consensus permits the Court to suspect that Dr Wagner has not

21

applied reliable principles and methods reliably, a suspicion confirmed by the close

22

examination of his opinions described above.

23

5.

Patient Experience.

24

Finally, Dr. Wagner states that he bases his opinions in part on his experience

25

treating patients suffering from gadolinium toxicity. Doc. 154-10 at 2. As explained in

26

the Advisory Committee Note to Rule 702:

27
28
- 49 -

1
2
3
4

If the witness is relying solely or primarily on experience, then the
witness must explain how that experience leads to the conclusion reached,
why that experience is a sufficient basis for the opinion, and how that
experience is reliably applied to the facts. The trial court’s gatekeeping
function requires more than simply “taking the expert’s word for it.”

5

Fed. R. Evid. 702 advisory committee’s note to 2000 amendment (citing Daubert II, 43

6

F.3d at 1319).

7

As already noted, Dr. Wagner says nothing in his report about his patient-

8

treatment experience or what he has learned from it. He does not describe the conditions

9

of patients he has treated, whether they had impaired or healthy kidneys, the severity or

10

cause of their symptoms, the nature of their exposures to GBCAs, or anything else that

11

might help the Court evaluate whether his patient experience constitutes a reliable basis

12

for rendering opinions in this case.

13

6.

Bradford Hill Criteria.

14

As noted above, the Bradford Hill criteria provide a well-recognized method for

15

evaluating general causation. Dr. Wagner makes only one passing reference to the

16

criteria: “The methodology I employed here is consistent with the sound deductive and

17

inferential reasoning principles that I employ as a physician in both clinical practice and

18

research, including the Bradford Hill Criteria for evaluating causal connections.”

19

Doc. 154-10 at 3. That’s it. No further explanation is provided of how the nine-factor

20

Bradford Hill analysis was performed or supports his causation opinion. Dr. Wagner’s

21

passing reference to the criteria does nothing to show that he has applied them reliably as

22

required by Rule 702(d).

23

7.

Conclusion.

24

Considering all of the factors outlined above, Plaintiffs have not shown by a

25

preponderance of the evidence that Dr. Wagner’s testimony is based on reliable

26

principles and methods that have been applied to the facts of this case reliably. His

27

opinions therefore are not admissible under Rule 702(c) and (d).

28
- 50 -

Plaintiffs’ Motion to File Supplemental References.

1

C.

2

Plaintiffs filed a motion for leave to file supplemental references to support Dr.

3

Wagner’s opinion. Doc. 209. They seek to file two recently published studies. Plaintiffs

4

assert that the articles are directly relevant to the general causation inquiry. Id. Plaintiffs

5

offer only that the studies will help provide the Court a “complete picture of the current

6

state of the scientific evidence.” Id. at 3.

7

The first study – “Gadolinium-Based Contrast Agents: Stimulators of Myeloid-

8

Induced Renal Fibrosis and Major Metabolic Disruptors” – involved a rodent experiment

9

on obesity and GBCA-induced disease. Doc. 209-1 at 6. The test mice underwent 5/6

10

nephrectomies to model chronic kidney disease. Id. at 7. Then they were split into two

11

groups: one received a 60% saturated fat diet for 22 weeks and one received a control diet

12

for 22 weeks. Id. at 8. At 18 weeks, each group was subdivided into untreated and

13

GBCA-treated groups. Id. The GBCA-treated group received 20 doses of GBCAs over

14

the final four weeks. Id.

15

The authors noted that serum chemistries collected at the endpoint revealed

16

increased creatinine and decreased inverse creatinine in the gadolinium-treated group. Id.

17

They concluded that these parameters are highly specific for impaired glomerular

18

filtration rates and these data, along with additional findings, demonstrated that GBCAs

19

have a profound impact on renal function and histology. Id. The authors go on to discuss

20

the effect of GBCA treatment on the metabolic profile of the kidney in contrast-treated

21

mice with “normal renal function.” Id. at 10. They state: “This allowed determination of

22

whether pre-existing chronic kidney disease was requisite for gadolinium-based contrast

23

agent-induced systemic fibrosis or not and provided a baseline to gauge the similarity

24

with high fat diet-induced kidney fibrosis with that induced by gadolinium and the impact

25

of the combination.” Id.

26

It is unclear to the Court how the authors draw conclusions for individuals with

27

normal renal functions when the mice in this experiment were all given 5/6

28
- 51 -

1

nephrectomies to simulate chronic kidney disease. The Court guesses that based on the

2

metrics cited above, the authors distinguished between individuals with chronic kidney

3

disease and renal insufficiency. But distinctions on that basis are not consistent with the

4

discussion of NSF. See Doc. 154-10 at 16 (Wagner Report) (noting that there is no

5

specific degree of kidney impairment that “plays a causal role in the pathogenesis of the

6

disease”). And Dr. Wagner has failed to provide any other sort of explanation for why

7

the Court should consider these results in relation to his opinions about patients with

8

normal renal function. Doc. 209-1 at 2. Moreover, like with the studies discussed above,

9

Dr. Wagner has not explained why the use of the mice, the dose, and other unique aspects

10

of this study can be extrapolated to humans with healthy kidneys who receive GBCAs.

11

In the second in vitro study, “Gadolinium-Based MRI Contrast Agents Induce

12

Mitochondrial Toxicity and Cell Death in Human Neurons, and Toxicity Increases with

13

Reduced Kinetic Stability of the Agent,” the authors considered whether “GBCAs are

14

toxic to the cellular function of differentiated human neurons that are representative of

15

those found in one of the brain regions with high T1-dependent MRI signal intensity

16

changes.” Doc. 209-1 at 21. The authors used dopaminergic neurons and exposed them

17

to increasing concentrations of eight GBCAs for seven days. Id. The concentrations

18

ranged from “clinically relevant concentrations measured in some autopsy patients who

19

had received repeated injections of contrast for MRI to higher concentrations to reveal

20

dose-dependent trends.” Id. For all GBCAs tested, cell death increased with exposure

21

dose and clinically relevant doses of agents with lower kinetic stability exhibited toxicity.

22

Id. “Reduction of mitochondrial membrane potential and oxidative respiratory function

23

also generally mirrored the agents’ structural kinetic stabilities, with greater impairment

24

at lower concentration for the less stable agents.” Id. The authors concluded that

25

GBCAs caused a toxic effect on mitochondrial respiratory function and cell viability, and

26

the toxicity increases as agent concentration increases and as the kinetic stability

27

decreases. Id.

28
- 52 -

1
2
3

The authors noted that the study was conducted in vitro, “which has its inherent
limitations in directly applying results to human beings.” Id. at 29.

6

First, contrast agents in culture medium have easy access to cell
membranes. Second, the medium itself is clearly different from the
interstitium in human brain. Third, cells in culture inherently have a
different environment to those in the brain and do not replicate the tissue
and synaptic complexity in vivo.

7

Id. The authors also concluded, however, that “it is not unreasonable to project that

8

neurons in the human brain would experience some degree of the toxicity observed in

9

cultured human neurons.” Id. at 30. The findings suggested that “a single contrast-

10

enhanced MRI scan would not be expected to achieve this concentration of agent within

11

the brain. However, with numerous repeated administrations, and evidence that the lower

12

stability agents may accumulate and persist longer, our findings suggest that repeated

13

administrations, particularly of the lower kinetic stability of GBCAs, could result in

14

neural toxicity.” Id. Still the authors also noted that clinical effects resulting from the

15

cellular dysfunction are not clear. Id.

4
5

16

Defendants argue this study cannot be relevant because it was a test tube study

17

where the cells were essentially bathed in GBCAs. 7/3/2019, LiveNote Hearing Tr. at 32.

18

Further, the authors drew no connections to clinical effects and instead noted the

19

limitations of applying the study to humans. Id.

20

Plaintiffs provide very little explanation regarding the significance of this study.

21

They do not address the limitations identified by the authors. Nor do they explain why

22

this study may reliably be extrapolated to humans with healthy kidneys considering the

23

methods in which the cells were exposed to the GBCA – a test tube bath in GBCAs. And

24

because the study does not actually identify GBCA-caused clinical adverse effects, it is

25

unclear how it can support Dr. Wagner’s general causation opinion. The authors noted

26

that loss of neurons “was concerning” and could “accelerate cell death due to normal

27

aging,” or cause other problems in patients with neural disease. Doc. 209-1 at 30. The

28
- 53 -

1

actual implications of the study appear to be speculative and, at most, it seems to offer

2

evidence researchers must consider to continue evaluating GBCAs and GBCA retention.
Plaintiffs’ supplemental references do not change the Court’s determination that

3
4

Dr. Wagner’s opinions are inadmissible.

5

VII.

Rule 702(c) and (d) – Specific Discussion of Dr. Tversky’s Opinions.

6

A.

Opinions.

7

Dr. Tversky starts with the same foundational facts as Dr. Wagner. He states that

8

gadolinium is toxic, GBCAs dechelate in the body, gadolinium is retained in the bodies

9

of individuals with normal renal function, and it causes NSF, which is just a form of

10

gadolinium-induced toxicity in humans that represents the severe end of a continuum of

11

adverse effects caused by gadolinium. Doc. 154-9 at 1-8. He opines that impaired renal

12

functioning is not a prerequisite for the development of NSF, nor does every gadolinium-

13

induced injury manifest as NSF. Id. at 4, 14. He opines that NSF is caused by an

14

increased retention of gadolinium, and it is probable that the tissue injury that occurs in

15

NSF patients is similar to the injuries experienced by patients with normal kidney

16

function because both injuries arise from exposure to gadolinium retained in the body in a

17

dechelated state. Id. at 6.

18

Dr. Tversky states that medical case reports describe adverse events attributed to

19

GBCAs in patients with normal renal function. Patients appear to suffer from symptoms

20

that are milder than NSF, but in many cases are consistent with the early symptomatology

21

seen in NSF patients. Id. at 8. He notes that by September 2018, the FDA accumulated

22

several hundred adverse events reports tied to linear GBCAs in patients with normal renal

23

function. Id. Symptoms varied, but tended to cluster around skin, muscle, bone, and the

24

nervous system, as well as chronic pain. Id. Other symptoms included thickening and

25

tightening of the skin, pain in the extremities, persistent headache, and mental fogginess.

26

Id. It is unknown what conditions increase the risk for an individual to experience

27

gadolinium toxicity. Id. at 10. The disease presentations are varied and seem unique to

28
- 54 -

1

individual patients. Id. at 15. Finally, Dr. Tversky applies the Bradford Hill criteria to

2

evaluate the association between gadolinium retention and toxic effects in patients with

3

normal renal function. Id. at 10-14.

4
5

B.

Analysis.
1.

Dr. Tversky’s Assumptions.

6

Throughout his opinion, Dr. Tversky relies on NSF data. He concludes that

7

because dechelated gadolinium is toxic, gadolinium is retained in the tissues and organs

8

of individuals with normal kidney function, and the accumulation of gadolinium appears

9

to be dose-dependent – meaning that more exposure to GBCAs equates to a higher risk of

10

accumulation – then “gadolinium accumulation and toxicity among patients with renal

11

insufficiency . . . share parallel mechanisms compared to those with normal renal

12

function.” Doc. 154-9 at 5. But he cites no evidence that retention in patients with

13

normal renal function is equal in method or quantity to the accumulation in renally-

14

impaired patients, and he merely assumes as probable that “tissue injury that occurs from

15

GBCAs in renally impaired patients shares the same mechanisms of toxic injury with

16

patients who have normal kidney function.” Id. at 6. Confusingly, he also asserts twice

17

that gadolinium tissue retention alone does not lead to disease, when his theory seems to

18

be that gadolinium retention causes these adverse effects. Id. at 7-8. And he offers no

19

citation for one of his most crucial assertions – that patients’ symptoms matched early

20

manifestations of NSF.

21

Dr. Tversky’s vague language describing the disease process makes it near

22

impossible to assess the reliability of his theory.

23

mechanisms of causing injury), 6 (stating that it is probable that there are shared

24

mechanisms for causing toxic injury). It is unclear how Dr. Tversky can make the leap

25

that gadolinium retention in patients with healthy kidneys is the same as accumulation in

26

NSF patients and will cause the same type of symptoms, particularly when there are

27
28
- 55 -

See id. at 5 (describing parallel

1

unknown variables (as he acknowledges) that affect the progression of NSF. Doc. 154-9

2

at 7. And he provides no concrete information demonstrating how the injury occurs.

3

Essentially, Dr. Tversky presents a hypothesis that if gadolinium is not cleared

4

from the body in one type of patient and causes disease, then gadolinium that is retained

5

in another type of patient, possibly in a different manner, causes disease too. The disease

6

caused by the latter situation is both similar to the former and different, and it is unknown

7

how or why it is different or what causes the differences. This broad hypothesis, which

8

Dr. Tversky does not show to be validated by reliable principles and methods, is not

9

admissible under Rule 702(c). See In re Mirena IUD Prods. Liab. Litig., 169 F. Supp. 3d

10

396, 450 (S.D.N.Y. 2016) (“By assuming, without any apparent basis, that secondary

11

perforation is an undisputed phenomenon, and then positing a mechanism by which she

12

believes it may occur, [the doctor] indulges in at most, scientifically-grounded

13

speculation: an untested and potentially untestable hypothesis.” (internal quotation marks

14

and citation omitted)); Dunn v. Sandoz Pharm. Corp., 275 F. Supp. 2d 672, 684 (M.D.

15

N.C. 2003) (“While hypothesis is essential in the scientific community because it leads to

16

advances in science, speculation in the courtroom cannot aid the fact finder in making a

17

determination of whether liability exists. Ultimately, speculation is unreliable evidence

18

and is inadmissible.”).

19

2.

Questionable Evidence.

20

Dr. Tversky relies on a number of studies to assert that gadolinium causes acute

21

and chronic tissue damage in both humans and animals. See Doc. 154-9 at 6. But he

22

offers only one-line conclusions summarizing the results of these studies with no

23

explanation about why these studies can reliably be applied to the causation issue in this

24

case or how they relate to his opinions. For example, without context, he states: “free

25

gadolinium has been shown to damage lung tissue in rats” and “[g]adolinium causes

26

destruction and death of fibroblast cells in rats that may lead to tissue fibrosis and

27

scarring.” Id. at 6. This is particularly concerning given that most of the results of the

28
- 56 -

1

studies seem irrelevant to the question at hand. For example, Dr. Tversky cites a study

2

examining gadolinium chloride – which Plaintiffs never show to be relevant to any issue

3

in this case – and another concerning NSF. Id. He cites two NSF studies to support the

4

statement that gadolinium damages rat and human fibroblasts. Id. The Court cannot

5

determine how reliably these studies can be used to support Dr. Tversky’s conclusions

6

regarding a gadolinium-illness continuum because he provides no analysis or support.

7

Dr. Tversky may rely on other scientists’ work, but he must explain why it supports his

8

opinion and why it is reliable. See Daubert II, 43 F.3d at 1318.

9

Dr. Tversky also relies largely on case reports.

But he mischaracterizes the

10

number of reports that support his theories. He states that “[b]y September 2018, the

11

FDA accumulated over 14,000 adverse events associated with linear GBCAs.”

12

Doc. 154-9 at 8. He states that “at least several hundred” of these individual case reports

13

occurred in patients with normal renal function who developed chronic symptoms after

14

having been exposed to a GBCA. Id. During his deposition, Dr. Tversky testified that he

15

came up with these numbers in his own review of the FDA database. But surprisingly, he

16

admitted that the 14,000 number included NSF cases, and he could not provide the

17

number of non-NSF cases, a significant deficiency for his opinions in this non-NSF case.

18

The best he could say is that “[i]t’s somewhere between 132 and 14,000.” Doc. 155-2

19

at 40. This kind of imprecision compounds the lack of reliability in case reports.10

20

Plaintiffs argue that Dr. Tversky properly relied on extensive literature to support

21

his opinions, including animal, in vitro, in vivo, toxicological, mechanistic, and

22

observational studies.

Plaintiffs assert that the following studies considered by Dr.

23
24
25
26
27
28

10

The 2017 MIDAC report Dr. Tversky cites to support this statement identified
139 patients with normal renal function and adverse events in conjunction with
gadolinium retention – 98 from the medical literature and 41 from FAERS. Doc. 154-43
at 40. The only other source he cites to support the number of identified adverse events is
the FDA public dashboard, which is a link to the FDA’s homepage and does not provide
any specific information related to the FAERS reports. Id. And he cites the Semelka
(2016) study instead of any of the FAERS reports when categorizing the symptoms found
in the FAERS reports. Id.
- 57 -

1

Tversky support his opinions: (1) Agarwal (2014), which is a comprehensive review and

2

meta-analysis of NSF; (2) Do (2014), a study analyzing a type of MRI contrast, tissue

3

gadolinium, and fibrosis; (3) Gibby (2004), a study demonstrating GBCA retention in

4

human bone; (4) Kanda (2014), a study demonstrating retention in the human brain;

5

(5) Knoepp (2017), a study demonstrating the effect of GBCAs on human epithelial

6

channels; and (6) Gathings (2015), a case study about gadolinium-associated plaques.

7

After reviewing these studies, the Court cannot conclude that they reliably move

8

Dr. Tversky’s opinion from hypothesis to conclusion. Agarwal and Do involved NSF

9

and do not look at adverse health effects in patients with normal renal function. See

10

Agarwal et al. (2014), Doc. 191-14 at 2 (finding an association between GBCAs and the

11

development of NSF); Do et al. (2014), Doc. 191-7 (concluding that different types of

12

GBCAs can cause NSF). Gibby and Kanda examined gadolinium retention, but they do

13

not identify any injuries or adverse effects. Gibby et al. (2004), Doc. 191-8; Kanda et al.

14

(2014), Doc. 191-9. Knoepp found that gadolinium ions released from linear GBCAs

15

reduced the activity of human epithelial sodium channels, and that this effect was not

16

observed with macrocyclic GBCAs. Knoepp et al. (2017), Doc. 191-10 at 9. But the

17

authors noted that all remarks were speculative and required further investigation. Id.

18

The authors also concluded that the low incidence of side effects should not affect the use

19

of GBCAs in healthy individuals. Id. The final study, Gathings, is the plaque case study

20

involving two patients who exhibited skin plaques with markers of NSF. See Gathings et

21

al. (2015), Doc. 191-11.

22

hypothesis that GBCAs can cause NSF-like symptoms, but it does not provide a reliable

23

basis for Dr. Tversky’s general causation opinion. Again, “case reports raise questions,

24

they do not answer them.” McLain, 401 F.3d at 1245.

As noted above, this case study certainly could prompt a

25

Finally, Dr. Tversky fails to address the five human studies discussed above with

26

respect to Dr. Wagner’s opinions. These relevant studies suggest that there is no link

27

between GBCAs and many of the symptoms amalgamated in GDD.

28
- 58 -

1

3.

Bradford Hill Criteria.

2

“The Bradford Hill criteria are metrics that epidemiologists use to distinguish a

3

causal connection from a mere association.” In re Zoloft Prods. Liab. Litig., 858 F.3d

4

787, 795 (3rd Cir. 2017). “These criteria ‘start with an association demonstrated by

5

epidemiology and then apply’ eight or nine criteria to determine whether that association

6

is causal.” In re Mirena IUS Levonorgestrel-Related Prods. Liab. Litig. (No. II), 341 F.

7

Supp. 3d 213, 242 (S.D.N.Y. 2018) (quoting In re Breast Implant Litig., 11 F. Supp. 2d

8

1217, 1234 (D. Colo. 1998)). The criteria can be summarized as follows:

9
10
11
12
13
14
15
16
17
18
19
20

Strength of Association. There must be some degree of statistical
association between a cause and its effect. A strong association (large in
magnitude) is more likely to represent causation than a weak association
(small in magnitude).
Temporality. A cause must precede its effect. Strength in temporality,
such as when a cause immediately precedes its effect, supports an inference
of causation.
Biological Plausibility. A cause and effect relationship between exposure
and disease should be biologically plausible with other information about
the disease or harm.
Coherence. A cause and effect relationship between exposure and disease
should be consistent with other information about the disease or harm.
Dose-Response Effect. Causation is more likely if greater amounts of the
putative cause are associated with corresponding increases in the
occurrence of disease or harm.

21
22
23
24
25
26
27

Consistency.
When similar findings are generated by several
epidemiological studies involving various investigators, causation tends to
be supported.
Analogy. Substantiation of relationships similar to the putative causal
relationship increases the likelihood of causation.
Experimental Evidence. Causation is more likely if removing the
exposure in a population results in a decrease in the occurrence of disease
or harm.

28
- 59 -

1
2
3

Specificity. When there is but a single putative cause for the disease or
harm, causation is supported.
See In re Mirena IUS, 341 F. Supp. 3d at 242-43; see also Zoloft, 858 F.3d at 795.

4

Dr. Tversky’s Bradford Hill analysis is the centerpiece of his report. Doc. 154-9

5

at 10-14. He uses this analysis to conclude that GBCAs cause a range of symptoms in

6

patients with normal renal function.

7

recognized method for evaluating causation, the Court finds that Dr. Tversky has not

8

applied it reliably as required by Rule 702(d). The Court will address the criteria in the

9

order they appear in his expert report.
a.

10
11
12
13
14
15
16
17
18

Id. at 16.

Although Bradford Hill is a well-

Strength of Association.

The Reference Manual provides this helpful guidance on the need for an
association:
the first question an epidemiologist addresses is whether an association
exists between exposure to the agent and disease. An association between
exposure to an agent and disease exists when they occur together more
frequently than one would expect by chance. Although a causal
relationship is one possible explanation for an observed association
between an exposure and a disease, an association does not necessarily
mean that there is a cause-effect relationship.
2011 WL 7724261 at *10 (footnote omitted).

19

The Reference Manual makes clear that the observed association is the trigger –

20

the starting point – for the causation analysis. “Once an association has been found

21

between exposure to an agent and development of a disease, researchers consider whether

22

the association reflects a true cause-effect relationship.” Id. at *28 (emphasis added).

23

“We emphasize that [the Bradford Hill criteria] are employed only after a study finds an

24

association to determine whether that association reflects a true causal relationship.” Id.

25

(emphasis added); see also In re Mirena IUS, 341 F. Supp. 3d at 242 (“There must be

26

some degree of statistical association between a cause and its effect.”); Dunn 275 F.

27

Supp. 2d at 678-79 (“The greater weight of authority supports Sandoz’ assertion that [use

28
- 60 -

1

of] the Bradford Hill criteria is a method for determining whether the results of an

2

epidemiologic study can be said to demonstrate causation and not a method for testing an

3

unproven hypothesis.”); Soldo v. Sandoz Pharm. Corp., 244 F. Supp. 2d 434, 514 (W.D.

4

Pa. 2003) (“The Bradford-Hill criteria ‘were developed as a mean[s] of interpreting an

5

established association based on a body of epidemiologic research for the purpose of

6

trying to judge whether the observed association reflects a causal relation between an

7

exposure and disease.’” (quoting report of court-appointed expert)).

8

As these authorities make clear, the association that starts the Bradford Hill

9

analysis normally is established by an epidemiologic study. But in this case, Plaintiffs

10

admit that “there are no epidemiologic studies in patients with normal renal function

11

concerning the long-term effects of gadolinium.” Doc. 192 at 9. And Dr. Tversky does

12

not find such an association elsewhere – he identifies no study-established association

13

between GBCAs and GDD.

14

Not to be deterred from reaching his causation conclusion, Dr. Tversky launches

15

his Bradford Hill analysis on the basis of a different association – the association between

16

GBCAs and NSF. Doc. 154-9 at 10. He begins: “The best-known and most studied

17

population of patients that have suffered injury from gadolinium toxicity are those renally

18

impaired patients diagnosed with NSF.” Id. But the purpose of his Bradford Hill

19

analysis is not to show a causal link between GBCAs and NSF. It is to show a causal link

20

between GBCAs and GDD. Dr. Tversky starts his causation analysis with the wrong

21

illness.

22

After discussing the association between GBCAs and NSF and citing studies that

23

support such an association, Dr. Tversky says this about Plaintiffs’ illnesses: “As to the

24

effect of gadolinium retention in patients with normal renal function, there is no medical

25

or scientific reason to expect any less toxic effect of retained gadolinium in human tissue

26

from linear GBCAs.”

27

between GBCAs and GDD, he assumes an association. Rather than presenting evidence

Id.

Thus, rather than identifying an established association

28
- 61 -

1

of an association, he flips the burden of proof and states that there is no reason to expect

2

it does not exist. As the authorities cited above emphasize, however, this is an incorrect

3

application of the first Bradford Hill criterion. Dr. Tversky never attempts to explain

4

why the first criterion can be applied to an association between GBCAs and NSF. And

5

the lack of a study-established association between GBCAs and GDD appears to be a

6

primary reason that regulatory and medical bodies around the world have been unwilling

7

to find causation.

8

At the end of his discussion of the first Bradford Hill criterion, Dr. Tversky seems

9

to concede that he cannot evaluate strength of association when no association has been

10

shown to exist. He refers to strength of association as “effect size” (Doc. 154-9 at 10),

11

explains his view that doctors may be confused into underdiagnosing GDD because the

12

misnomer in NSF (nephrogenic) causes them to assume that the symptoms can appear

13

only in patients with impaired kidneys, and then states: “For this and other reasons [that

14

he never identifies] it is not possible to state precisely the size of the effect of gadolinium

15

toxicity in renally sufficient patients.” Doc. 154-9 at 10-11. In other words, it is not

16

possible to state the strength of association between GBCAs and GDD in kidney-healthy

17

patients like Plaintiffs. Id. And yet that is the very thing the first Bradford Hill factor

18

requires an expert to address. See In re Mirena IUS, 341 F. Supp. 3d at 242 (“There must

19

be some degree of statistical association between a cause and its effect.

20

association (large in magnitude) is more likely to represent causation than a weak

21

association (small in magnitude).”).

22

b.

A strong

Consistency.

23

This factor considers whether similar findings have been generated by several

24

studies involving various investigators. Is there a consistency among the various studies

25

that have evaluated the link between the substance and the illness it is alleged to cause?

26

Dr. Tversky does not start his analysis of this factor with a consistency among any studies

27

about GBCAs and GDD. Instead, he again he looks at NSF: “This criterion has been

28
- 62 -

1

met, certainly in the context of NSF where the strongest signal of gadolinium toxicity

2

was first recognized.” Id. at 11.

3

Although he later appears to broaden his focus to gadolinium-caused symptoms

4

generally, the only source he cites is the FAERS database. Id. He notes that 80% of

5

symptoms reported in the database associated with GBCAs involved the skin, nervous

6

system, or musculoskeletal system. Id. As noted above, however, the FAERs database

7

consists of a collection of individual reports made to the FDA from a broad range of

8

sources, many of them not scientific or even medical. It certainly does not meet the

9

consistency criterion of Bradford Hill – it does not constitute a series of scientific studies

10

completed under carefully controlled conditions. Even the FDA states that “FAERS data

11

cannot be used to calculate the incidence of an adverse event[.]” Questions and Answers

12

on FAERS, https://www.fda.gov/ drugs/surveillance/fda-adverse-event-reporting-system-

13

faers (last visited July 8, 2019).

14

c.

Specificity.

15

The Reference Manual provides this explanation of specificity:

16

An association exhibits specificity if the exposure is associated only
with a single disease or type of disease. The vast majority of agents do not
cause a wide variety of effects. For example, asbestos causes mesothelioma
and lung cancer and may cause one or two other cancers, but there is no
evidence that it causes any other types of cancers. Thus, a study that finds
that an agent is associated with many different diseases should be examined
skeptically.

17
18
19
20
21

2011 WL 7724261 at *31 (footnotes omitted).

22

Dr. Tversky’s discussion of specificity focuses on NSF. He notes that gadolinium

23

is the primary cause of NSF, there have been no reliable reports of NSF without

24

gadolinium exposure, and cases of NSF dropped when the use of GBCAs in renally-

25

impaired patients dropped. Doc. 154-9 at 11-12. “This fact alone,” he states, “is strong

26

confirmation of a cause and effect relationship.” Id. at 11. But again, the question in this

27

case is not whether GBCAs cause NSF. And the only comment Tversky makes about

28
- 63 -

1

GDD and specificity is this: “the pattern of disease, history of exposure to linear GBCA

2

and cluster of symptoms noted in NSF is very similar to that seen among patients with

3

normal renal function.” Id. Dr. Tversky does not explain how “very similar” is the

4

equivalent of specificity.

5

Nor does he attempt to reconcile the fact that GBCAs are claimed in this case to

6

cause a wide variety of symptoms in patients with healthy kidneys, with the fact that

7

“[a]n association exhibits specificity if the exposure is associated only with a single

8

disease or type of disease.” Reference Manual, 2011 WL 7724261 at *31. To the

9

contrary, Dr. Tversky acknowledges that “a universal set of symptoms associated with

10

exposure to GBCAs has not yet been established among patients with normal renal

11

function[.]” Doc. 154-9 at 12. And even if there was a study linking GBCAs to the wide

12

range of symptoms claimed by Plaintiffs, it would be viewed with doubt: “a study that

13

finds that an agent is associated with many different diseases should be examined

14

skeptically.” Reference Manual, 2011 WL 7724261 at *31.

15

d.

Temporality.

16

This factor considers timing – when the disease appears in relation to exposure.

17

Logically, “[i]f an exposure causes disease, the exposure must occur before the disease

18

develops. Reference Manual, 2011 WL 7724261 at *29. Dr. Tversky again talks mostly

19

about NSF, but states: “all credible reports of GBCA induced disease in patients with

20

normal renal function must occur sometime after exposure.” Doc. 154-9 at 12. He states

21

that credible reports “must occur” after exposure, but he does not state whether they do.

22

And he cites no studies or other authorities on the subject.

23

Temporality also considers whether symptoms arise close in time to exposure.

24

“Strength in temporality, such as when a cause immediately precedes its effect, supports

25

an inference of causation.” In re Mirena IUS, 341 F. Supp. 3d at 242. On this point, Dr.

26

Tversky notes that “[f]rom cases of NSF we learned that there can be a wide range of

27

temporality in terms of the onset of symptoms after exposure to GBCAs.” Doc. 154-9

28
- 64 -

1

at 12. Elsewhere in his report he notes that symptoms can occur days, months, or even

2

years after exposure. Id. at 15. And he acknowledges that “we can expect a range in

3

time from exposure to manifestation of disease.” Id. at 12. He never explains how such

4

wide-ranging timing supports the temporality criterion of causation.
e.

5

Biological Gradient.

6

This factor is referred to in different sources as the “dose-response effect.” It is

7

premised on the notion that “[c]ausation is more likely if greater amounts of the putative

8

cause are associated with corresponding increases in the occurrence of disease or harm.”

9

In re Mirena IUS, 341 F. Supp. 3d at 242. Without citing any specific study, Dr. Tversky

10

asserts that gadolinium toxicity “has been shown to be dose dependent in both humans

11

and animals.” Doc. 154-9 at 12. He does not explain how this assertion squares with Dr.

12

Wagner’s opinion that “there does not appear to be a minimum threshold dose for

13

appearance or occurrence of NSF.” Doc. 154-10 at 16; see also id. at 17 (“[E]ven a

14

miniscule amount of a GBCA is capable of triggering GBCA-induced systemic

15

fibrosis.”).

16

With respect to dose dependency in GDD, Dr. Tversky offers only this: “Selection

17

bias is plausible, however given the lessons learned from renally impaired patients and

18

animal models, in my opinion a true biological gradient does exist.” Doc. 154-9 at 12.

19

This is unvarnished ipse dixit. And he adds: “Individual susceptibility as well as other

20

risk factors play a role in the extent of health effects experienced by any given patient.”

21

Id. He never identifies these “other risk factors,” but this statement seems to be a

22

concession that GDD symptoms are not dose dependent – that other factors influence

23

their appearance and severity.

24

f.

Plausibility.

25

This factor notes that “[a] cause and effect relationship between exposure and

26

disease should be biologically plausible with other information about the disease or

27

harm.” In re Mirena IUS, 341 F. Supp. 3d at 242. As Dr. Tversky notes, this normally

28
- 65 -

1

requires “a plausible mechanism between cause and effect.” Doc. 154-9 at 12. Dr.

2

Tversky admits that the mechanisms for gadolinium toxicity “are not fully elucidated.”

3

Id.

4

plausible, he again resorts to burden-shifting: “The experience of NSF established that

5

de-chelated gadolinium is toxic and causes severe even lethal damage in humans. There

6

is no evidence that renal function alters the fact or nature of that toxicity.” Id. at 13.

7

This, of course, presents no plausible mechanism by which GBCAs cause GDD. Dr.

8

Tversky simply assumes plausibility in the absence of contrary evidence.

9

And in place of describing the mechanism that renders his causation opinion

g.

Coherence.

10

This factor notes that “[a] cause and effect relationship between exposure and

11

disease should be consistent with other information about the disease or harm.” In re

12

Mirena IUS, 341 F. Supp. 3d at 242. After discussing the recognized cause of NSF, Dr.

13

Tversky resorts to conclusions: “symptom[s] among individuals without renal

14

impairment exposed to GBCAs overlap[] with those of patients with NSF,” patients with

15

healthy kidneys retain gadolinium, and “[t]he evidence shows that gadolinium exposure

16

from linear GBCAs will lead to injury in some patients regardless of renal function.”

17

Doc. 154-9 at 13. He summarily compares GDD causation to NSF, but he does not

18

address what this factor considers: whether GBCA causation of GDD is consistent with

19

other information known about GDD. In re Mirena IUS, 341 F. Supp. 3d at 242.

20

h.

Experiment.

21

This factor focuses on what has been learned from experimentation: “Causation is

22

more likely if removing the exposure in a population results in a decrease in the

23

occurrence of disease or harm.” Id. Dr. Tversky does not identify any experimentation

24

regarding GDD, but instead notes the evidence that GBCAs cause NSF. He then states:

25

“As awareness increases, well-vetted case reports will be added to the FDA database and

26

other directories so that more robust calculations can be made.” Doc. 154-9 at 13. His

27
28
- 66 -

1

message seems to be: “we don’t have experimentation data yet, but I am sure it is

2

coming.” Such speculation cannot be used to prove causation.11
i.

3

Analogy.

4

This factor notes that “[s]ubstantiation of relationships similar to the putative

5

causal relationship increases the likelihood of causation.” In re Mirena IUS, 341 F. Supp.

6

3d at 242. The analogy Dr. Tversky points to, of course, is NSF. Doc. 154-9 at 13. And

7

it does appear to be relevant.
j.

8

Bradford Hill Summary.

9

In summary, with the exception of the last of nine factors – analogy – Dr. Tversky

10

does not reliably apply any of the Bradford Hill criteria. His opinion therefore is not

11

admissible under Rule 702(d).

12

4.

General Acceptance in the Scientific Community.

13

As with Dr. Wagner’s opinions, the methods Dr. Tversky uses for reaching his

14

opinions are not shared by the many regulatory and medical organizations that have

15

examined the same evidence.

16

principles and methods have not been faithfully applied.” Fed. R. Evid. 702 advisory

17

committee’s note to 2000 amendment; see also Lust, 89 F.3d at 598.

The Court therefore “may fairly suspect that [his]

18

C.

Conclusion.

19

Plaintiffs have not shown by a preponderance of the evidence that Dr. Tversky’s

20

opinions are based on reliable principles and methods that have been applied to the facts

21

of this case reliably. His opinions therefore are not admissible under Rule 702(c) and (d).

22
23
24
25
26
27
28

11

With respect to experimentation, Plaintiffs assert repeatedly in their briefs that
prospective, blind studies – giving GBCAs to some patients and not to others to see what
differences are produced – are not ethical. See, e.g., Doc. 185 at 17. But they never
dispute that other kinds of valid epidemiological studies are possible. Dr. Tversky notes
that “[e]xperimental evidence must be based upon animal models and retrospective
cohort analysis in humans.” Doc. 154-9 at 13. But he does not discuss the five human
studies discussed above. And the authors of the Parillo (2016) study propose prospective
studies with active surveillance of symptoms. Doc. 154-26 at 7.
- 67 -

1

VIII. Rule 702(c) and (d) – Specific Discussion of Dr. Whittaker’s Opinions.

2

Dr. Whittaker, a Ph.D. toxicologist, adopts an approach similar to Drs. Wagner

3

and Tversky. She notes that linear GBCAs are retained in the bodies of MRI patients

4

with healthy kidneys. Doc. 154-11 at 17. She opines that GBCAs dechelate in the body.

5

Id. at 18. And she asserts that gadolinium is highly reactive and toxic in its free, ionic

6

form. Id. at 19. She opines that gadolinium and linear GBCAs have a proliferative effect

7

on fibroblasts – the type of cell that is responsible for making and maintaining connective

8

tissue. Id. at 20. An overabundance of fibroblasts can result in a condition known as

9

fibrosis. Id. Dechelated or free gadolinium can also become a competitive inhibitor of

10

important physiologic processes that depend on calcium. Id. And she reviews the

11

science showing that GBCAs cause NSF. Id. at 22-23.

12

A.

Description of GDD.

13

To define and describe GDD, Dr. Whittaker turns to other evidence. She cites

14

three articles by Semelka, et al., for the proposition that GDD is a disease in subjects with

15

normal or near normal renal function who develop persistent symptoms after receiving

16

GBCAs. Id. at 23, 43-44.

17

The first article does not purport to show that GBCAs cause GDD. In fact, it is

18

titled “Presumed Gadolinium Toxicity in Subjects With Normal Renal Function.”

19

Doc. 187-34 (emphasis added). The article is a five-page report on four individual cases

20

of persons who developed various symptoms after receiving GBCA-enhanced MRIs. Id.

21

This is the article’s conclusion:

22
23
24
25
26

In summary, we report 4 patients with normal renal function, who
developed clinical symptoms arising shortly after receiving GBCA
administration and who developed an apparent toxicity to it. This disease
may share clinical manifestations with NSF but in less severe form.
Importantly, it is unknown which GBCA agents may cause this collection
of signs and symptoms. Well-designed clinical studies are warranted to
investigate this presumed disease more fully[.]

27
28
- 68 -

1

Id. at 6. The article treats case reports appropriately – it uses them to raise questions, not

2

answer them. See McLain, 401 F.3d at 1245. The article advances a hypothesis – “[t]his

3

disease may share clinical manifestations with NSF” – and recommends that it be

4

investigated through clinical studies.

5

The second Semelka article is not a scientific study. It is a four-page article based

6

on an on-line survey of 42 people from two gadolinium toxicity support groups.

7

Doc. 156-16 at 2-3. The survey was anonymous – meaning that the people responding

8

did not even have to identify themselves – and asked a series of questions about their

9

claimed conditions. Id. The authors acknowledged the limitations of their work:

10
11
12
13
14
15

There are various limitations of our study. Major problems include
that it is a survey and relies on knowledge and integrity of the respondents.
Despite that a survey provides all the participants with a standardized
stimulus, eliminating researcher’s own biases, it is recognized that selfreport studies may have validity problems. Although the study is
subjective, in that we rely on patient descriptions, it should also be
appreciated that there was a generally common pattern of symptomatology.
A control group of subjects who received multiple administrations of
GBCAs, but who do not have symptoms would be helpful.

16
17

Id. at 5 (emphasis added).

18

The third Semelka article does not purport to establish causation. The four-page

19

article notes that some studies have confirmed that GBCAs can deposit in the brain and in

20

bone, but states: “To our knowledge, neither the bone deposition first reported by Gibby

21

et al. nor the brain deposition first reported by Kanda et al. have been associated with

22

recognized disease.” Semelka et al., Gadolinium in Humans: A Family of Disorders,

23

https://www.ajronline.org/doi/pdf/10.2214/AJR.15. 15842 at 229-30 (last visited July 8,

24

2019) (footnotes omitted). The article recounts reports from patient advocacy groups of

25

various symptoms that developed after receipt of GBCAs, and states: “Our preliminary

26

investigation has convinced us that this phenomenon is a true disease process, which we

27

propose naming ‘gadolinium deposition disease.’” Id. at 230 (emphasis added).

28
- 69 -

1

Dr. Whittaker next cites a survey that reports various symptoms in 17 patients.

2

Doc. 154-11 at 24. The survey was copyrighted by www.GadoliniumToxicity.com, the

3

website for a gadolinium-related patient advocacy group and was written by two of the

4

survey participants.

5

Retained Gadolinium from Contrast MRIs, https://gdtoxicity.files.wordpress.com/2014/

6

09/gd-symptom-survey.pdf at 1, 10 (last visited July 9, 2019). The authors stated that

7

“[t]he conclusions are based on a small sample and suggest[ed] that a larger study should

8

be conducted,” and that “[n]o attempt is made to determine what, if any, level of

9

Gadolinium, is safe to remain in the body.” Id. at 1.

See Gadolinium Toxicity: A Survey of the Chronic Effects of

10

Dr. Whittaker cites the similar Burke (2016) survey of 50 anonymous, self-

11

reporting individuals who identified the top three adverse effects of GDD as bone and

12

joint pain, head and neck pain, and skin problems. Docs. 154-11 at 24; 187-33 at 2. The

13

survey authors acknowledged that, as an anonymous survey, the study has validation

14

issues due to patients forgetting important details or exaggerating symptoms.

15

Doc. 187-33 at 3. Another recognized limitation of this survey is selection bias, “as self-

16

selected volunteers completed [the] survey” and the symptoms may have been self-

17

reported by patients who did not have documented gadolinium retention. Id. The authors

18

also stated that the “risk of severe adverse reaction to GBCA exposure is extremely

19

small, hence it may be difficult to determine characteristics of individuals that predispose

20

to this type of reaction,” but it could be caused by an individual genetic abnormality. Id.

21

(footnotes omitted).

22

The authors of the Burke article concluded:

23

This survey represents a first description of patients with normal
renal function who self-described toxicity related to GBCA administration.
Despite its limitations, our opinion is that there most likely is toxicity
associated with GBCA administration in patients with normal renal
function. At the very least, this study highlights the need to further
investigate the subject of patients with normal renal function who complain
of severe long-lasting symptomatology following GBCA administration.

24
25
26
27
28

- 70 -

1

Id. Dr. Whittaker does not address the inherent limitations of this survey, nor those

2

recognized by the authors.

3

Dr. Whittaker next turns to the FDA’s FAERS database. She notes that a 2017

4

FDA analysis identified 139 adverse events related to GBCAs in healthy patients, 41

5

from FAERS and 98 from medical literature. Id. at 25. She asserts that the FDA analysis

6

identified a clustering of symptoms around cutaneous, musculoskeletal, cognitive or

7

neurological, and pain syndrome clinical categories. Id. at 25-26.

8

Dr. Whittaker does not note the limitations in this data. The FDA medical officer

9

who presented this data at the 2017 MIDAC conference, Dr. David Croteau, stated that 7

10

of the 41 cases from the FAERS database reflected gadolinium retention with no adverse

11

health events, only 5 of the 98 literature cases had documented gadolinium retention, and

12

92 of the literature cases “were from two online surveys posted to a private blog on a

13

gadolinium toxicity support group website and a public gadolinium toxicity Facebook

14

page.” MIDAC Trans., Doc. 156-7 at 149-50. The data actually included a wide range

15

of 39 symptoms, with a median of 7 symptoms per patient. Id. at 152. Dr. Croteau

16

further explained: “Even though published in peer-reviewed journals, those two surveys

17

could have been excluded from the review given the nature of the data provided, but it

18

was felt important to include all the data published, given the overall paucity of data on

19

the topic.” Id. at 150. Dr. Croteau added this caveat: “It’s also important to point out that

20

the bulk of the medical literature cases, that is, 96 out of 98 cases, were authored by the

21

same UNC Chapel Hill radiology group, leading to possible duplicate cases and an actual

22

total number of medical literature cases lower than 98.” Id.

23

Additionally, although Dr. Whittaker focuses on clustering around four categories

24

of symptoms, Dr. Croteau stated that “[d]espite clustering around certain clinical

25

categories, the marked heterogeneity of clinical adverse events reported makes

26

interpretation of the data challenging.

27

interpretation of the data and may result in either an over- or underestimation of the

In addition, many factors may influence the

28
- 71 -

1

importance of the problem.” MIDAC Trans., Doc. 156-7 at 158; see also Doc. 154-43

2

at 14 (MIDAC Briefing Document) (“[W]e observed some clustering around cutaneous,

3

musculoskeletal, neurological/cognitive, and pain syndromes clinical categories.

4

However, the clinical category “other” accounted for the highest number of adverse

5

events, emphasizing the heterogeneity of the adverse events reported.”). Dr. Croteau then

6

discussed some of these factors, including “unverified self-reported information” and

7

“alternative etiology” – meaning alternative causes – for the symptoms. Id. at 157-58.

8

He concluded: “at this juncture, considering the totality of the data, including case

9

reports, case series, and published online surveys identifying our review along with the

10

data limitations outlined earlier, we are unable to determine a causal association between

11

reported clinical adverse events and GBCA exposure.” Id. at 161.

12

Dr. Whittaker notes that in 2017 and 2018, the FDA reportedly received another

13

623 claimed adverse events related to linear GBCAs. Id. at 26. These included 25

14

different categories of symptoms, and, she notes, large numbers of cutaneous,

15

musculoskeletal, cognitive or neurological, and pain syndrome clinical categories. Id.

16

But her citation for this data is simply to the FAERS Public Dashboard (id. at 46),

17

suggesting that Dr. Whittaker must have counted these events herself. She says nothing

18

about how she identified or counted these events, and nothing about the reliability of the

19

data – clearly an important issue in light of the FDA’s own recognition of the data’s

20

limitations in its MIDAC presentation noted above.

21

B.

Toxicological Studies.

22

Dr. Whittaker next turns to toxicological investigations of adverse effects of

23

GBCAs.

She notes, but does not discuss, 16 toxicological investigations done in

24

connection with the FDA’s approval of six linear GBCAs. Id. at 28. The investigations

25

presumably contributed to the FDA’s decision to approve the GBCAs for public medical

26

use. Dr. Whittaker then focuses her attention on studies related to three categories of

27

potential harm: neurotoxicity, cutaneous toxicity, and nephrotoxicity.

28
- 72 -

1

1.

Neurotoxicity.

2

Dr. Whittaker discusses three rodent studies: (1) a 1986 paper by Taber and Bryan

3

that concluded there was a “potential danger” if GBCAs “were allowed direct access to

4

the brain”; (2) a 1994 study where rats were injected with GBCAs following disruption of

5

their blood-brain barrier, which “reported no gross neurological toxicity”; and (3) a 2018

6

study, discussed above, where the babies of mice injected with GBCAs displayed

7

abnormal behavior and decreased muscle strength. Id. at 28-29. She also discusses the

8

study in which dogs whose blood-brain barriers had been disrupted were injected with

9

GBCAs. Id. at 28.

10

Apparently because two of the three rodent studies she discusses were less than

11

conclusive – one reported only a “potential danger” and the other found no neurological

12

toxicity – Dr. Whittaker states that “[t]here is disagreement among scientists regarding

13

the utility of using rodents for evaluating the neurotoxicity of GBCAs.” Id. at 29. She

14

then cites a fourth rodent study by Lohrke et al. (2017), which asserted that rats were

15

suitable test subjects but found no “histological [i.e, microscopic] changes in the brain of

16

rodents administered GBCAs.” Id. Dr. Whittaker dismisses this fourth study as lacking a

17

“robust assessment of neurotoxicity,” concludes that it “does not answer the question of

18

whether exposure to any of the GBCAs resulted in neurological impairment,” and

19

alternatively notes that “it may also be the case that the rodent is not a good model to

20

assess impaired neurological function following exposure to linear GBCAs.” Id.

21

All of this leaves Dr. Whittaker’s discussion of neurotoxicity rather muddled.

22

Three of the four studies are inconclusive, two suggest that GBCAs do not cause

23

neurotoxicity, and the fourth concerned only behavior and muscle strength in baby mice.

24

The only other neurotoxicity study mentioned by Dr. Whittaker is the dog study, which

25

she seeks to elevate with this speculation: “the primitive nature of the rat brain . . . may

26

explain why adverse neurological effects were reported in a canine study.” Id. But she

27

never explains why the dogs in question were suitable test subjects for evaluating the

28
- 73 -

1

effects of GBCAs in humans with healthy kidneys, whether the doses given to the dogs

2

were relevant to humans, and why disruption of the dogs’ blood-brain barriers produced

3

results that can be used to draw conclusions for humans with healthy kidneys. In sum,

4

Dr. Whittaker’s discussion of neurotoxicity studies provides no scientific basis for

5

concluding that GBCAs cause GDD.

6

2.

Cutaneous Toxicity.

7

Dr. Whittaker opines that “[s]kin lesions similar to those reported among patients

8

with NSF have been reported in laboratory animals exposed to linear GBCAs, and animal

9

studies also confirm that exposure to linear GBCAs is associated with organ-specific

10
11

increases in fibrotic tissue.” Id. at 30. She further asserts:

14

For assessing cutaneous toxicity of linear GBCAs, rodent studies
reliably mimic effects in the skin that appear to be associated with GBCA
exposure. Available human case report data and studies in animal models
and cell systems indicate that cutaneous toxicity is a risk factor for humans
with normal renal function who are exposed to linear GBCAs.

15

Id. But she provides no citations for this statement, and no explanation of why rodent

16

and humans skin are sufficiently similar to justify the comparison.

12
13

17

She then cites three rodent studies, Plunkett et al. (1992), Lohrke et al. (2017), and

18

Do et al. (2014), to support her opinion that GBCAs can cause skin problems in humans

19

with normal kidney function. Doc. 154-11 at 30. (The Lohrke study, by the way, is the

20

one she criticized on the previous page as lacking a “robust assessment.” Id. at 29.) But

21

she never addresses dose or other distinguishing characteristics of these studies. As noted

22

above, dose matters when extrapolating from animal studies. See Reference Manual,

23

2011 WL 7724262 at *8.

24

The Plunkett study injected rats with GBCA doses of 0.1, 2.0, and 5.0 mmol/kg

25

three times per week for three weeks. Doc. 154-11 at 30. It found skin lesions only in

26

rats that received the highest dose – 50 times higher than the lowest doses. Id. Rats in

27

the Lohrke study were injected with 20 daily intravenous injections of GBCAs at a dose

28
- 74 -

1

of 2.5 mmol/kg. Four of ten in the linear GBCA group were later found to have skin

2

lesions. Rats in the Do study had their kidneys removed and were injected with a daily

3

GBCA dose of 2.5 mmol/kg over a period of four weeks. Some of the animals were later

4

found to have greater epidermal thickness and fibronectin production. Id.; Do et al.

5

(2014), Doc. 191-7.

6

Dr. Whittaker never explains why the high doses in these studies are comparable

7

to doses humans might receive with MRI contrast agents. Nor does she explain why

8

removal of the kidneys in the Do study produced results relevant to GBCA effects in

9

humans with normal kidneys.

10

3.

11

Nephrotoxicity.

Nephrotoxicity means toxicity to the kidneys.

Dr. Whittaker discusses three

12

studies, one on dogs and two on rats, that suggest GBCAs can result in nephrotoxicity.

13

Doc. 154-11 at 31. But she never explains why the rats and dogs in these studies are

14

appropriate subjects for demonstrating human effects, and she never addresses the

15

relevancy of the doses used in the studies. What is more, she does not explain why

16

nephrotoxicity is even relevant to this case. It is not one of the primary symptoms Dr.

17

Whittaker identifies for GDD (cutaneous, musculoskeletal, neurological/cognitive, and

18

pain syndromes (Doc. 154-11 at 25-26)), and although Ms. Fischer alleged kidney

19

damage in her original complaint, she did not include kidney injury in her amended

20

complaints or her deposition testimony. See Doc. 154-2 at 15; Fischer Docs. 1, 10, 60.

21

Additionally, both Ms. Fischer and Ms. Davis allege gadolinium retention in the kidney

22

as an injury, but neither refers to toxicity or injury of the kidney. Doc. 142 ¶ 4; Fischer

23

Doc. 10 ¶ 3.

24

4.

Toxicological Studies Summary.

25

Although Dr. Whittaker is a toxicologist, her discussion of toxicological studies

26

suffers from many shortcomings. The studies she addresses under neurotoxicity are more

27

unfavorable than favorable to her opinion, and she never explains why they reliably

28
- 75 -

1

support her opinions. She fails to show the relevancy of the studies she discusses under

2

cutaneous toxicity. She provides no support for her assertion that rodent studies reliably

3

mimic human skin response, she never addresses whether the high doses in these studies

4

are comparable to doses humans might receive with MRI contrast agents, and she does

5

not explain why removal of the kidneys from rodents in the Do study produced results

6

relevant to this case.

7

relevancy of rodent and canine studies or the doses used in the studies, and she does not

8

explain the relevancy of nephrotoxicity to any issue in this case. The Court cannot

9

conclude by a preponderance of the evidence that her reliance on these various studies

10

Her nephrotoxicity discussion likewise fails to address the

constitutes the use of reliable principles and methods as required by Rule 702(c).

11

C.

12

Dr. Whittaker engages in a brief discussion of the Bradford Hill criteria. This is

13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Bradford Hill Criteria.

her discussion in full:
Case report data in humans with normal to near-normal renal
function identify strength of association, specificity, and temporality
among patients exposed to linear GBCAs followed by development of
consistent symptoms. Secondly, data derived from studies in humans and
test animals exposed to linear GBCAs establish a biological gradient, with
higher doses of linear GBCAs associated with longer residence times. This
biological gradient is also seen among patients with compromised renal
function who developed NSF following exposure to GBCAs, with patients
exposed to linear GBCAs demonstrating higher incidences of NSF
compared to those exposed to macrocyclic GBCAs. Third, injury from
linear GBCAs is plausible and coherent based on data demonstrating that
gadolinium is released from linear GBCAs, as well as data demonstrating
that the gadolinium ion (Gd3+) is toxic.
Fourth, experimental
manipulation of GBCAs administered to patients demonstrate that adverse
symptomology is lower among patients administered macrocyclic GBCAs
compared to linear GBCAs. Finally, our understanding of the etiology of
NSF disease among patients with impaired renal function who were
exposed to linear GBCAs provides analogous information on the causeeffect relationship that help explains the symptomology reported among
patients with adequate kidney function following exposure to linear
GBCAs.

28
- 76 -

1

Doc. 154-11 at 32-33 (emphasis in original). This analysis, like Dr. Tversky’s, is deeply

2

flawed.
1.

3

Strength of Association, Specificity, and Temporality.

4

For strength of association, Dr. Whittaker relies solely on case reports. Id. But as

5

discussed above, case reports are among the weakest of scientific data and far from the

6

epidemiological studies normally relied on to provide the starting point for a Bradford

7

Hill analysis – an association.

8

(unsystematic clinical observations or case reports “are at the bottom of the evidence

9

hierarchy”). Case reports have no control group to answer the fundamental question of

10

an association: whether exposure to GBCAs and GDD “occur together more frequently

11

than one would expect by chance.” Reference Manual, 2011 WL 7724261 at *10. The

12

way to determine whether an association even exists is to compare individuals exposed to

13

a substance with individuals not exposed and see if disease occurs in the exposed group

14

more frequently than would occur by mere chance, as seen in the unexposed group. As

15

the Reference Manual explains: “Typically, [case] reports are not even sufficient to show

16

association, because there is no comparison group.” Id. at *4. Not only can case reports

17

not establish an association, they certainly cannot indicate the strength of the

18

association.12

See Reference Manual, 2011 WL 7724263 at *22

19

The case reports also fail to establish specificity. As noted above with respect to

20

specificity: “The vast majority of agents do not cause a wide variety of effects. . . . Thus,

21

a study that finds that an agent is associated with many different diseases should be

22

examined skeptically.” Reference Manual, 2011 WL 7724261 at *31 (footnotes omitted).

23
24
25
26
27
28

12

Several courts have held that it is not proper methodology for an expert to apply
the Bradford Hill factors without data from a controlled study that shows an association.
In re Fosamax Prods. Liab. Litig., 645 F. Supp. 2d 164, 188 (S.D.N.Y. 2009); In re
Neurontin Mktg., Sales Practices & Prods. Liab. Litig., 612 F. Supp. 2d 116, 126-27 (D.
Mass. 2009); Dunn, 275 F. Supp. 2d at 678 (finding that “pharmacological
properties . . . , statistically insignificant epidemiological studies, clinical studies, case
reports, and animal studies” are insufficient for establishing an association for Bradford
Hill analysis).
- 77 -

1

Dr. Whittaker’s own report notes that case reports in the FAERS database – upon which

2

she relies – purport to associate GBCAs with 25 different categories of symptoms,

3

including disorders of the nervous system, heart, ear, reproductive system, and even

4

“infections and infestations.” See Doc. 154-11 at 27, Figure 4. The case reports lack

5

specificity.

6

And Dr. Whittaker does not explain how the case reports show temporality. She

7

does not describe when symptoms appeared in case reports after exposure, and therefore

8

makes no effort to show that temporality is satisfied.

9

2.

Biological Gradient.

10

As noted above, this factor is also referred to as the “dose-response effect.” It

11

holds that “[c]ausation is more likely if greater amounts of the putative cause are

12

associated with corresponding increases in the occurrence of disease or harm.” In re

13

Mirena IUS, 341 F. Supp. 3d at 242. Asked simply, does increased exposure result in

14

increased disease? As Sir Bradford Hill stated in his seminal lecture: “For instance, the

15

fact that the death rate from cancer of the lung rises linearly with the number of cigarettes

16

smoked daily, adds a very great deal to the simpler evidence that cigarettes smokers have

17

a higher death rate than non-smokers.” Doc. 154-19 at 5.

18

Dr. Whittaker does not address whether increased exposure to GBCAs results in

19

more or worse GDD. She states instead that “higher doses of linear GBCAs [are]

20

associated with longer residence times.” Doc. 154-11 at 32. But the key link in this

21

criterion is not between dose and residence times, but between dose and the existence or

22

seriousness of the illness. She also cites NSF evidence, noting that “patients exposed to

23

linear GBCAs demonstrat[e] higher incidences of NSF compared to those exposed to

24

macrocyclic GBCAs.” Id. This is the same tack Dr. Tversky took, looking to NSF

25

evidence to prove GDD causation. Dr. Whittaker, like Dr. Tversky, does not explain why

26

this is an appropriate application of Bradford Hill.

27
28
- 78 -

3.

1

Experimental Manipulation.

2

This factor focuses on what has been learned from experimentation: “Causation is

3

more likely if removing the exposure in a population results in a decrease in the

4

occurrence of disease or harm.” In re Mirena IUS, 341 F. Supp. 3d at 242. Dr. Whittaker

5

states that “experimental manipulation of GBCAs administered to patients demonstrate

6

that adverse symptomology is lower among patients administered macrocyclic GBCAs

7

compared to linear GBCAs.” Doc. 154-11 at 32. Dr. Whittaker does not say what

8

experiments she is referring to, but her report identifies several studies which have shown

9

that linear GBCAs break down in the body more readily than macrocyclic GBCAs. See,

10

e.g., Doc. 154-11 at 23. She also cites case reports and an anonymous survey that have

11

attributed adverse symptoms more often to linear than macrocyclic GBCAs. Id. at 20

12

(Burke et al. (2016)), 26 (FDA literature review). But this analysis again relies on types

13

of data that cannot prove causation, and Dr. Whittaker does not explain why the least

14

reliable class of data provides a reliable scientific basis for the Bradford Hill evaluation

15

of causation.

16
17
18
19
20
21
22
23
24
25
26
27

4.

Analogy.

On her last criterion, Dr. Whittaker identifies NSF as an analogous disease. Id.
at 32-33. As with Dr. Tversky, this comparison is relevant.
5.

Dr. Whittaker’s Additional Explanation.

As part of her Bradford Hill discussion, Dr. Whittaker seems to acknowledge the
lack of genuine science on GDD causation. She states:
An enormous challenge associated with preparing an unbiased
opinion is posed by the state of peer-reviewed publications as well as the
state of the pre-clinical toxicology dataset. The majority of peer-reviewed
publications on this topic are authored by individuals who work directly
for, or have been compensated by, GBCA manufacturers. For some
publications, this conflict of interest is apparent, as exampled by overly
exuberant conclusions relating to the safety of GBCAs, e.g., “their
accumulated safety record is extraordinarily positive” (e.g., Kanal and
Tweedle 2015); however, for many peer-reviewed publications, the bias of

28
- 79 -

1

3

authors is unclear. Disclosure of past or present affiliations in a
publication’s conflict of interest section merely discloses the affiliation but
does not guarantee that the publication itself is free of bias or selective data
reporting.

4

Doc. 154-11 at 33. This seems to be a lament that there is not enough evidence to prove

5

that GBCAs cause GDD. Dr. Whittaker seeks to excuse the lack of evidence by asserting

6

that it is due to bias, even in authors whose bias is “unclear.”

2

7

She continues:

8

An equal challenge associated with discerning the toxicity of linear
GBCAs using toxicological data is posed by inherent limitations of studies
performed on these compounds as part of the drug approval process. The
majority of preclinical toxicity studies on linear GBCAs were not designed
to evaluate endpoints associated with GDD such as: cutaneous toxicity
(rash, erythema, skin lesions), musculoskeletal toxicity (e.g.,
arthralgia/joint pain, muscle weakness, bone pain, muscle spasms), or
neurological/cognitive effects (e.g., paranesthesia/tingling, tremor,
dizziness).

9
10
11
12
13
14

Id. Here, Dr. Whittaker seems to be explaining why there is not enough evidence to

15

address GDD causation on the basis of toxicological data. But even if her concerns about

16

the lack of usable data are warranted, they do not make her opinion admissible. The

17

Bradford Hill criteria cannot be applied to unreliable data simply because reliable data do

18

not exist. And neither is it the role of the Court to admit unreliable evidence because an

19

expert is at the disadvantage of not having reliable data. Rule 702 requires opinions

20

based on adequate facts and data and reliable principles and methods, applied reliably.

21

Fed. R. Evid. 702(b)-(d). There is no exception for opinions that do not meet these

22

standards.

23

6.

Bradford Hill Conclusion.

24

With the exception of the last factor – analogy – Dr. Whittaker does not reliably

25

apply the Bradford Hill criteria. Thus, even though the criteria represent a well-accepted

26

method for evaluating causation, her application of the criteria is not admissible under

27

Rule 702(d).

28
- 80 -

1

D.

Other Concerns.

2

Defendants argue that Dr. Whittaker’s opinions lack independent research. In the

3

absence of independent research or peer review, an expert must explain the process by

4

which she reaches her conclusions and identify the objective sources she relies on.

5

Daubert II, 43 F.3d at 1318-19. The expert should also demonstrate that the “research

6

and analysis supporting the proffered conclusions have been subjected to normal

7

scientific scrutiny through peer review and publication.” Id. at 1317-18. For all of the

8

reasons discussed above, Dr. Whittaker has not shown that her opinion has been

9

subjected to normal scientific scrutiny.

10

Defendants also argue that Dr. Whittaker’s opinions should be excluded because

11

they were prepared solely for the purpose of litigation. Doc. 156 at 16. Dr. Whittaker

12

has never published her opinions or subjected them to peer review.

13

conducted any clinical studies, animal studies, or other studies on gadolinium. See

14

Doc. 156-1 at 46-48. Indeed, Dr. Whittaker seems to have very little experience with

15

gadolinium prior to writing her expert report. Id. As discussed above, in the absence of

16

independent research or peer review, experts must explain the process by which they

17

reach their conclusions and identify what objective sources they rely on. Daubert II,

18

43 F.3d at 1318-19. As explained above, Dr. Whittaker fails to cite reliable evidence to

19

support her opinion.

Nor has she

20

Finally, like Drs. Wagner and Tversky, Dr. Whittaker never discusses important

21

human epidemiological studies on this issue. Doc. 157 at 12. These include McDonald

22

(2017), which compared 1,315 patients who received GBCAs to 2,946 patients with no

23

history of GBCA exposure and found no association between GBCA use and cognitive

24

decline, poor neuropsychological performance, or diminished motor performance.

25

Doc. 153 at 19; see also Doc. 154-23.

26
27
28
- 81 -

1

E.

Dr. Whittaker Conclusion.

2

Plaintiffs have failed to show by a preponderance of the evidence that Dr.

3

Whittaker has used reliable principles and methods and has applied them to the facts of

4

this case reliably.

5

Rule 702(c) and (d).

6

IX.

As a result, her opinions are not admissible in evidence under

Two Cases Cited by Plaintiffs.

7

A.

Wendell.

8

Plaintiffs cite the Ninth Circuit’s decision in Wendell v. GlaxoSmithKline, LLC,

9

858 F.3d 1227 (9th Cir. 2017), throughout their briefing. In Wendell, the parents of a

10

young man who developed Hepatosplenic T-cell lymphoma (HSTCL), a rare and

11

aggressive form of cancer, sued the manufacturer of drugs their son had taken for other

12

health issues. The district court found two of plaintiffs’ doctors unreliable and excluded

13

their testimony. Id. at 1232-33. The Ninth Circuit found it to be a “close case,” but

14

reversed, finding the district court failed to consider the “broader picture of the experts’

15

overall methodology.” Id. at 1233. The court of appeals found both doctors to be highly

16

qualified and noted that one had conducted clinical research on T-cell leukemia and

17

lymphomas and had experience treating “hundreds of patients with T-cell lymphomas,

18

and thousands of patients with lymphomas, including seven patients with HSTCL.” Id. at

19

1233 (internal quotations omitted). Because HSTCL is rare, the doctor had seen more

20

cases than 99% of the oncologists in the country. Id. at 1234. The second doctor had

21

extensive research and clinical experience with non-Hodgkin’s lymphoma. Id.

22

One doctor conducted a literature search and found articles showing an increased

23

risk of HSTCL in patients prescribed the drug the plaintiffs’ son had taken, and,

24

significantly, performed a differential diagnosis to determine the cause of the son’s

25

HSTCL. Id. The Ninth Circuit stated: “Nothing in Daubert, or its progeny, properly

26

understood, suggests that the most experienced and credentialed doctors in a given field

27

should be barred from testifying based on a differential diagnosis.” Id. at 1235.

28
- 82 -

1

The other doctor “based his opinion on ‘a summary of the medical records of [the

2

son] as well as copies of the pathology reports, and the original slides of the diagnostic

3

bone marrow,’ which he evaluated with over 30 years of experience diagnosing non-

4

Hodgkin lymphoma.” Id. He “also weighed other risk factors, including [the son’s] sex

5

and age, and determined that those were ‘weak risk factors; whereas, the disease he had,

6

particularly in the setting of the drugs he received would be considered very strong risk

7

factors.’” Id.

8

The court concluded that “the experts relied not just on these studies – which not

9

only examined reported cases but also used statistical analysis to come up with risk rates

10

– but also on their own wealth of experience and additional literature.” Id. at 1236. The

11

court found that “when an expert establishes causation based on a differential diagnosis,

12

the expert may rely on his or her extensive clinical experience as a basis for ruling out a

13

potential cause of the disease.” Id. at 1237.

14

This case is different. The HSTCL at issue in Wendell was a recognized illness;

15

the GDD in this case is a wide-ranging and imprecisely defined illness proposed in some

16

medical literature. The experts in Wendell performed a differential diagnosis on the basis

17

of facts specific to the plaintiffs’ son; the experts in this case have not performed a

18

differential diagnosis and have not examined Plaintiffs or their medical histories.

19

Granted, this is because the current issue in this case is general causation, not specific

20

causation, but this fact points up a distinction between this case and Wendell. Wendell

21

addressed a different set of experts, with considerable relevant experience, addressing a

22

known illness, and still found it to be a close case. The Ninth Circuit relied heavily on

23

the doctors’ experience, their differential diagnoses, and their case-specific analysis of

24

the son’s medical history. Similar facts are not present here. Wendell is distinguishable.

25

B.

In Re GBCAs.

26

Plaintiffs cite the decision of Judge Polster in the NSF MDL holding that the

27

plaintiffs’ experts provided admissible opinions on general causation. In re GBCAs, 2010

28
- 83 -

1

WL 1796334. But there are significant differences between that case and this one. In the

2

NSF MDL, the fact that GBCAs caused NSF was strongly supported by the timing of

3

NSF’s rise and decline. As Judge Polster explained: “NSF is a singular disease. NSF

4

rapidly emerged, and just as rapidly declined, with the administration of GBCAs to

5

persons with severe renal impairment.” Id. at *6. That is, “the rapid emergence and

6

decline of NSF associated with the rise and fall of [GBCAs’] use in renally impaired

7

persons” strongly suggested that the GBCAs caused NSF. Id. Additionally, Plaintiff’s

8

theory that GBCAs caused NSF had “been subjected to publication and peer review” and

9

had “been generally accepted in the relevant scientific and medical community.” Id.; see

10

also id. at *7. This case has no comparable increase and decrease of the illness linked to

11

the increase and decrease of GBCA use, nor has Plaintiffs’ causation theory been

12

embraced by the scientific and medical communities. To the contrary, every regulatory

13

and medical organization to consider the question has found insufficient evidence to

14

conclude that GBCAs cause GDD. In re GBCAs is distinguishable.

15

X.

Chemistry Opinion of Dr. Raymond.

16

Plaintiffs’ counsel stated in their briefs and during the hearing that they will not

17

ask Dr. Raymond, a chemist, to provide an opinion on general medical causation.

18

Plaintiffs intend to have him teach the jury about the chemistry of gadolinium and

19

GBCAs, and experts can appropriately be used to provide tutorials to the jury. See Fed.

20

R. Evid. 702 advisory committee’s note to 2000 amendment (“The amendment does not

21

alter the venerable practice of using expert testimony to educate the factfinder on general

22

principles.”).

23

testimony contained in Dr. Raymond’s report, the Court will not attempt to draw

24

admissibility distinctions now.

25

representation that he will not provide general medical causation opinions.

Because the parties have not briefed the admissibility of the tutorial

The Court will, however, hold Plaintiffs to their

26
27
28
- 84 -

1

XI.

Conclusion.

2

For the reasons stated above, the Court finds that (1) the opinions of Drs. Brent

3

Wagner, Jody Tversky, and Margret Whittaker are inadmissible under Rule 702; and

4

(2) Dr. Raymond’s opinions will be limited to the chemistry of GBCAs and gadolinium.

5

IT IS ORDERED:

6

1.

Defendants’ omnibus motion to exclude the testimony of Plaintiffs’ experts

7

(Doc. 153) is granted with respect to Drs. Brent Wagner, Jody Tversky,

8

and Margret Whittaker, and denied with respect to Dr. Kenneth Raymond.

9

Dr. Raymond’s opinions will be limited to the chemistry of GBCAs and
gadolinium and will not include medical causation.

10
11

2.

(Doc. 157) is granted.

12
13

3.

Plaintiffs’ Motion for leave to file supplement references for Brent Wagner,
M.D. (Doc. 209) is granted.

14
15

Defendants motion to exclude the expert testimony of Brent Wagner, M.D.

4.

Within 14 days of this order, the parties shall provide the Court with a joint

16

memorandum setting forth their views of what should happen next in these

17

cases in light of the Court’s ruling. The parties should include a discussion

18

of what should happen with the Drescher case, CV 19-00096.

19

Dated this 2nd day of August, 2019.

20
21
22
23
24
25
26
27
28
- 85 -