In Re: Taxotere (Docetaxel) Eye Injury Products Liability Litigation

1
UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF LOUISIANA
IN RE: TAXOTERE (DOCETAXEL) 	) 	MDL NO. 3023
EYE INJURY 	)
PRODUCTS LIABILITY LITIGATION 	)
)
This document relates to: 	) 	SECTION: “H” (5)
All cases 	)
ORDER AND REASONS
Before the Court are four Motions filed by Defendant Sanofi: (1) Motion
to Exclude Testimony of Dr. Gerald McGwin (Doc. 462); (2) Motion to Exclude
General Causation Opinions of Dr. Josef Thundiyil (Doc. 446); (3) Motion to
Exclude Certain Testimony of Dr. Vikram Durairaj (Doc. 455); and (4) Motion
for Summary Judgment on General Causation (Doc. 450). For the reasons set
forth herein, Sanofi’s Motion to Exclude Dr. Gerald McGwin is DENIED.
Sanofi’s Motion to Exclude Dr. Thundiyil is DENIED. Sanofi’s Motion to
Exclude Dr. Durairaj is GRANTED IN PART and DENIED IN PART.
Sanofi’s Motion for Summary Judgment is DENIED.
BACKGROUND
Plaintiffs 	in 	this 	multidistrict 	litigation 	(“MDL”) 	are 	suing 	several
pharmaceutical 	companies 	that 	manufactured 	and/or 	distributed 	a
chemotherapy drug, Taxotere or docetaxel, that Plaintiffs were administered
for the treatment of cancer, primarily breast cancer.1 Among these companies
are 	Defendants 	sanofi-aventis 	U.S. 	LLC 	and 	Sanofi 	U.S. 	Services 	Inc.
1 Taxotere is the brand name for the drug docetaxel.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 1 of 67

2
(collectively, “Sanofi”), Accord Healthcare, Inc. (“Accord”), and Sandoz Inc.
(“Sandoz”). Plaintiffs allege that Taxotere or docetaxel caused them to sustain
injuries 	to 	their 	lacrimal 	systems, 	including 	punctal, 	canalicular, 	and/or
nasolacrimal duct stenosis (narrowing), which resulted in excessive tearing,
otherwise known as epiphora.2
On May 2, 2023, this Court issued Case Management Order No. 16
(“CMO 16”).3 Among other things, CMO 16 set deadlines for discovery and
motion practice, and bifurcated general dispositive issues, including but not
limited to preemption, adequacy of the label, and general causation. Pursuant
to CMO 16, Plaintiffs provided the expert reports of (1) Dr. Gerald McGwin,
expert epidemiologist; (2) Dr. Josef Thundiyil, expert toxicologist; and (3) Dr.
Vikram 	Durairaj, 	a 	board-certified 	oculoplastic 	surgeon 	whom 	Plaintiffs
contend specializes in treating the injury at issue. On February 28, 2025,
Defendants Accord, Sandoz, and Sanofi jointly filed three Motions seeking to
exclude certain opinions, including those related to general causation, of Drs.
Thundiyil,4 Durairaj,5 and McGwin.6 The same day, Defendants filed a Motion
for Summary Judgment on General Causation as to the claims of Plaintiffs
naming them as Defendants.7 Plaintiffs oppose all four Motions.8
2 For reference, this Court provided extensive background on the injury at issue and
the facts giving rise to this case in its prior Orders addressing preemption. Doc. 583; Doc.
591. 3 Doc. 115.
4 Doc. 446.
5 Doc. 445.
6 Doc. 462.
7 Doc. 450.
8 Doc. 488; Doc. 482; Doc. 487; Doc. 492.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 2 of 67

3
On December 1, 2025, this Court denied Sanofi’s Motion for Judgment
based on Preemption.9 On January 16, 2026, this Court granted Accord’s and
Sandoz’s Motions for Summary Judgment based on preemption. This Court
entered Judgment in favor of Defendants Accord and Sandoz on March 6, 2026,
dismissing Plaintiffs’ claims against them with prejudice.10 On February 27,
2026, this Court entered an Order and Reasons granting Sanofi’s Motion to
Certify this Court’s December 1, 2025 Order and Reasons for Interlocutory
Appeal.11 As such, as to the four Motions now before the Court, Sanofi is the
only remaining Defendant.
LEGAL STANDARD
I. 	Summary Judgment
Summary 	judgment 	is 	appropriate 	“if 	the 	pleadings, 	depositions,
answers to interrogatories, and admissions on file, together with affidavits, if
any, show that there is no genuine issue as to any material fact and that the
moving party is entitled to a judgment as a matter of law.”12
In determining whether the movant is entitled to summary judgment,
the Court views facts in the light most favorable to the non-movant and draws
all reasonable inferences in his favor.13 “If the moving party meets the initial
burden of showing that there is no genuine issue of material fact, the burden
shifts to the non-moving party to produce evidence or designate specific facts
9 Doc. 583.
10 Doc. 606.
11 Id. The Order certifying this Court’s December 1, 2025 Order and Reasons was
entered on March 4, 2026. Doc. 604.
12 Sherman v. Hallbauer, 455 F.2d 1236, 1241 (5th Cir. 1972).
13 Coleman v. Houston Indep. Sch. Dist., 113 F.3d 528, 532 (5th Cir. 1997).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 3 of 67

4
showing the existence of a genuine issue for trial.”14 Summary judgment is
appropriate if the non-movant “fails to make a showing sufficient to establish
the existence of an element essential to that party’s case.”15 “In response to a
properly supported motion for summary judgment, the non-movant must
identify specific evidence in the record and articulate the manner in which that
evidence supports that party’s claim, and such evidence must be sufficient to
sustain a finding in favor of the non-movant on all issues as to which the non-
movant would bear the burden of proof at trial.”16 “We do not . . . in the absence
of any proof, assume that the nonmoving party could or would prove the
necessary facts.”17 Additionally, “[t]he mere argued existence of a factual
dispute will not defeat an otherwise properly supported motion.”18
II. 	Federal Rule of Evidence 702
The admissibility of expert testimony is governed by Federal Rule of
Evidence 702, which provides as follows:
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; (b) the
testimony is based on sufficient facts or data; (c) the testimony is
the product of reliable principles and methods; and (d) the expert
has reliably applied the principles and methods to the facts of the
case.
14 Engstrom v. First Nat’l Bank of Eagle Lake, 47 F.3d 1459, 1462 (5th Cir. 1995).
15 Celotex Corp. v. Catrett, 477 U.S. 317, 324 (1986).
16 Johnson v. Deep E. Tex. Reg. Narcotics Trafficking Task Force, 379 F.3d 293, 301
(5th Cir. 2004) (internal citations omitted).
17 Badon v. R J R Nabisco, Inc., 224 F.3d 382, 394 (5th Cir. 2000) (quoting Little v.
Liquid Air Corp., 37 F.3d 1069, 1075 (5th Cir. 1994)).
18 Boudreaux v. Banctec, Inc., 366 F. Supp. 2d 425, 430 (E.D. La. 2005).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 4 of 67

5
The current version of Rule 702 reflects the Supreme Court’s decisions
in 	Daubert 	v. 	Merrell 	Dow 	Pharms., 	Inc.,19 	and 	Kumho 	Tire 	Co. 	v.
Carmichael.20 	The threshold inquiry is whether the expert possesses the
requisite qualifications to render opinion on a particular subject matter.21
Having defined the permissible scope of the expert’s testimony, a court next
inquires whether the opinions are reliable and relevant.22 “Under Federal Rule
of Evidence 702 and Daubert,” expert testimony is reliable if “the reasoning or
methodology underlying the [expert’s] testimony is scientifically valid.”23 In
assessing reliability, the Court may consider several nonexclusive factors,
such as “whether the expert’s theory or technique: (1) can be or has been tested;
(2) has been subjected to peer review and publication; (3) has a known or
potential 	rate 	of 	error 	or 	standards 	controlling 	its 	operation; 	and 	(4) 	is
generally accepted in the relevant scientific community.”24 “The reliability
analysis applies to all aspects of an expert’s testimony: the methodology, the
facts underlying the expert’s opinion, the link between the facts and the
conclusion, et alia.”25
19 509 U.S. 579 (1993).
20 526 U.S. 137 (1999).
21 Wagoner v. Exxon Mobil Corp., 813 F. Supp. 2d 771, 799 (E.D. La. 2011); see also
Wilson v. Woods, 163 F.3d 935, 937 (5th Cir. 1999) (“A district court should refuse to allow
an expert witness to testify if it finds that the witness is not qualified to testify in a particular
field or on a given subject.”).
22 See United States v. Valencia, 600 F.3d 389, 424 (5th Cir. 2010).
23 Ruffin v. BP Expl. & Prod., Inc., 137 F.4th 276, 280 (5th Cir. 2025) (alterations in
original) (citation modified).
24 Burleson v. Tex. Dept. of Crim. Just., 393 F.3d 577, 584 (5th Cir. 2004) (citing
Daubert, 509 U.S. at 593–94).
25 Nestle v. BP Expl. & Prod., Inc., 627 F. Supp. 3d 577, 583 (E.D. La. 2022)
(alterations in original) (quoting Knight v. Kirby Inland Marine Inc., 482 F.3d 347, 355 (5th
Cir. 2007)).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 5 of 67

6
In turn, “[i]n determining admissibility of expert testimony, the court
must determine whether expert’s reasoning or methodology fits facts of case,
and whether it will thereby assist trier of fact to understand evidence; in other
words, it must determine whether it is relevant.”26
In 	undertaking 	this 	tripartite 	analysis, 	courts 	must 	give 	proper
deference to the traditional adversary system and the role of the jury within
that system.27 “Vigorous cross-examination, presentation of contrary evidence,
and 	careful 	instruction 	on 	the 	burden 	of 	proof 	are 	the 	traditional 	and
appropriate means of attacking shaky but admissible evidence.”28 As the
“gatekeeper” of expert testimony, the trial court enjoys broad discretion in
determining admissibility.29
III. 	Applicable Law
In this case, Plaintiffs assert claims under the products liability laws of
their respective states. Because this Court exercises diversity jurisdiction, it
applies 	state 	substantive 	law.30 	Where 	a 	transferee 	court 	presides 	over
multiple diversity actions consolidated for multidistrict litigation, it must
apply the substantive law of each transferor forum.31 Here, however, Plaintiffs
26 Id. at 583.
27 See Daubert, 509 U.S. at 596.
28 Id.
29 Wellogix, Inc. v. Accenture, L.L.P., 716 F.3d 867, 881 (5th Cir. 2013).
30 See Newsome v. Int’l Paper Co., 123 F.4th 754, 761 (5th Cir. 2024) (citing Smith v.
Christus Saint Michaels Health Sys., 496 F. App’x 468, 470 (5th Cir. 2012)); see also Ruffin,
137 F.4th at 282.
31 In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 16-2740, 2022 WL 110494,
at *2 (E.D. La. Jan. 12, 2022) (quoting In re Air Disaster at Ramstein Air Base, Germany on
8/29/90, 81 F.3d 570, 576 (5th Cir. 1996)); In re: Zicam Cold Remedy Mktg., Sales Pracs., &
Prods. Liab. Litig., 797 F. Supp. 2d 940, 942 (D. Ariz. 2011) (“Where a transferee court
presides over several diversity actions consolidated under the multidistrict rules, the choice
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 6 of 67

7
do not dispute that they must provide expert evidence on general causation to
survive summary judgment.32 While state law supplies the elements of a
plaintiff’s claims and the relevant burden of proof, the admissibility of expert
testimony 	is 	generally 	governed 	by 	the 	Federal 	Rules 	of 	Evidence 	and
Daubert.33 Sanofi moves for summary judgment solely on the basis that
Plaintiffs’ experts’ general causation testimony is inadmissible, and the parties
do not dispute that Fifth Circuit law applies.34 The Court therefore assesses
Sanofi’s Motion for Summary Judgment through this lens, and will address the
of law rules of each jurisdiction in which the transferred actions were originally filed must
be applied.”).
32 See Doc. 450-1 at 3. Several courts—including those presiding over pharmaceutical
MDLs—have concluded that most, if not all, jurisdictions require expert testimony to show
general causation in complex medical cases. See In re Onglyza (Saxagliptin) & Kombiglyze
(Saxagliptin & Metformin) Prods. Liab. Litig., 93 F.4th 339, 349 (6th Cir. 2024) (“So we
conclude that the district court did not err in finding that all jurisdictions require expert
testimony to show general causation in complex medical cases such as this MDL.”); see also
In re Lipitor (Atorvastatin Calcium) Mktg., Sales Pracs. & Prods. Liab. Litig., 227 F. Supp.
3d 452, 469 (D.S.C. 2017), aff’d, 892 F.3d 624 (4th Cir. 2018) (“While the specific language
used by courts vary to some degree, all jurisdictions require expert testimony at least where
the issues are medically complex and outside common knowledge and lay experience.”); In re
Mirena IUD Prods. Liab. Litig., 202 F. Supp. 3d 304, 311 (S.D. N.Y. 2016) (“The parties agree,
and the substantive law across all relevant jurisdictions holds, that where a causal link is
beyond the knowledge or expertise of a lay jury, expert testimony is required to establish
causation.” (citation modified)).
33 See Coleman v. United States, 912 F.3d 824, 832 (5th Cir. 2019) (citing Huss v.
Gayden, 571 F.3d 442, 452 & n.7 (5th Cir. 2009)). State rules, however, “become pertinent to
the extent that the federal rules make them so.” See id.; see also Ruffin, 137 F.4th at 280 n.1,
282. 34 See In re Chinese-Manufactured Drywall Prods. Liab. Litig., No. 09-02047, 2017 WL
1476595, at *18 (E.D. La. April 12, 2017) (“noting that “[a]s the MDL transferee court, this
Court is required to apply the substantive state law of the transferor court, and the federal
law of its own circuit”); see also In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 16-
2740, 2018 WL 5016219, at *3 (E.D. La. Oct. 16, 2018) (agreeing that “when interpreting and
applying the federal Constitution, any federal statute, or the Federal Rules of Civil
Procedure,” an MDL court should apply the law of the circuit in which it sits).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 7 of 67

8
admissibility of each expert’s general causation testimony separately. The
Court will then turn to the remaining arguments in the Motions to Exclude.
IV. 	Law on General Causation
General 	causation 	is 	generally 	defined 	as 	“whether 	a 	substance 	is
capable of causing a particular injury or condition in the general population,
while specific causation is whether a substance caused a particular individual’s
injury.”35 This terminology and the distinction between general and specific
causation are “widely recognized.”36 	General causation “has typically been
understood to mean the capacity of a toxic agent . . . to cause the illnesses
complained 	of 	by 	plaintiffs. 	If 	such 	capacity 	is 	established, 	‘individual
causation’ answers whether that toxic agent actually caused a particular
plaintiff’s illness.”37
Before proceeding with the analysis, the Court provides background on
the methodologies at issue. “Epidemiology is the scientific field that ‘attempts
to define a relationship between a disease [in humans] and a factor suspected
of 	causing 	it.’”38 Toxicology 	is 	primarily 	concerned 	with 	“identifying 	and
understanding the adverse effects of external chemical and physical agents on
biological systems.”39 Whereas epidemiology focuses on “what causes disease
35 Wells v. SmithKline Beecham Corp., 601 F.3d 375, 377–378 (5th Cir. 2010); see also
Knight, 482 F.3d at 351.
36 Steve C. Gold et al., REFERENCE G UIDE ON EPIDEMIOLOGY , in FED. JUD. CTR.,
REFERENCE M ANUAL ON S CIENTIFIC EVIDENCE 901 n.8 (4th ed. 2025) (“RMSE”) (collecting
cases). 37 In re Hanford Nuclear Rsrv. Litig., 292 F.3d 1124, 1129 (citing Sterling v. Velsicol
Chem. Corp., 855 F.2d 1188, 1200 (6th Cir. 1988)).
38 Wagoner, 813 F. Supp. 2d at 799 (quoting Brock v. Merrell Dow Pharm., Inc., 874
F.2d 307, 311 (5th Cir.1989), modified by 884 F.2d 166 (5th Cir. 1989)).
39 David L. Eaton et al., REFERENCE G UIDE ON TOXICOLOGY , in RMSE, supra note 36,
at 1030.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 8 of 67

9
in humans,” toxicology focuses on “what effects substances have on humans
and animals.”40 Toxicology provides scientific information relevant to the
following questions: “(1) What hazards does a chemical or physical agent
present to human populations or the environment? (2) What degree of risk is
associated with chemical exposure at any given dose?”41
A. 	Epidemiologic Methodology
The Fifth Circuit has recognized that epidemiological evidence is “the
most useful and conclusive type of [general causation] evidence”—specifically
“in toxic tort cases.”42 In this context, “[t]o assess whether general causation
exists between an agent and a disease, the case law recognizes a two-prong
test.”43 First, there must be an association between exposure to a drug and a
disease.44 “An association exists between exposure to a drug and a disease
when the two ‘occur together more frequently than one would expect by
chance.’”45 In other words, an association exists when people exposed to the
drug have a higher incidence of the disease and the difference is not simply
due to chance.46
40 Perrotti v. Lockheed Martin Corp., No. 6:22-cv-1338, 2025 WL 2554425, at *11 (M.D.
Fla. Sept. 2, 2025).
41 RMSE, supra note 36, at 1031.
42 Ruffin, 137 F.4th at 282 (emphasis added) (citation modified); see also Burst v. Shell
Oil Co., No. 14-109, 2015 WL 3755953, at *4 (E.D. La. June 16, 2011), aff’d, 650 F. App’x 170
(5th Cir. 2016).
43 In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 2740, 2021 WL 311865, at
*3 (E.D. La. Jan. 19, 2021) (citing Burst, 2015 WL 3755953, at *5).
44 Wagoner, 813 F. Supp. 2d at 803–04.
45 In re Lipitor (Atorvastatin Calcium) Mktg., Sales Pracs. & Prods. Liab. Litig., 174
F. Supp. 3d 911 (D.S.C. 2016) (quoting Michael D. Green et. al, REFERENCE G UIDE O N
EPIDEMIOLOGY , I N FED. JUD. CTR., REFERENCE M ANUAL O N S CIENTIFIC EVIDENCE 566 (3d
ed. 2011) (“2011 RMSE”) (collecting cases).
46 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 9 of 67

10
An 	association, 	however, 	“is 	not 	equivalent 	to 	causation.”47 	“An
association identified in an epidemiologic study may or may not be causal.”48
Thus, as to the second prong, once an association is found, the expert assesses
whether a true causal relationship underlies the association.49 Typically, an
expert applies the Bradford Hill criteria to evaluate this second prong.50 The
Bradford 	Hill 	criteria 	are: 	(1) 	temporal 	relationship; 	(2) 	strength 	of 	the
association; 	(3) 	dose-response 	relationship; 	(4) 	replication 	of 	findings; 	(5)
biological 	plausibility; 	(6) 	consideration 	of 	alternative 	explanations; 	(7)
cessation of exposure; (8) specificity of the association; and (9) consistency with
other knowledge.51 Determining whether an association is causal is a matter
of scientific judgment, and scientists “reliably applying the Bradford Hill
factors may reasonably come to different conclusions about whether a causal
inference may be drawn.”52
The Reference Manual on Scientific Evidence (hereinafter “RMSE”)
describes the common ways of expressing the existence of an association
between exposure and disease, including in terms of a relative risk (RR), an
odds ratio, (OR) or an attributable risk (AR). 	The most commonly used
approach is “relative risk,” defined as the “ratio of the risk of disease or death
among people exposed to an agent to the risk among the unexposed.”53 “For
47 RMSE, supra note 36, at 902.
48 Id.
49 See id.
50 See id. at 973.
51 Burst, 2015 WL 3755953, at *5 (citing 2011 RMSE, supra note 45, at 600).
52 In re Abilify (Aripiprazole) Prods. Liab. Litig., 299 F. Supp. 3d 1291, 1307 (N.D. Fla.
2018) (citing Milward v. Acuity Spec. Prods. Grp., Inc., 639 F.3d 11, 18 (1st Cir. 2011)).
53 RMSE, supra note 36, at 1019.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 10 of 67

11
instance, if 10% of all people exposed to a chemical develop a disease, compared
with 5% of people who are not exposed, the disease occurs twice as frequently
among the exposed people,” and the relative risk is 2.54 A relative risk of 1
indicates no association between exposure and disease.55 “A relative risk ratio
above 1 indicates an increased risk in the exposed group, and a relative risk
ratio less than one indicates a decreased risk in the exposed group.”56
Next, epidemiologists consider whether an association may be due to
chance, bias, or confounding. Two common ways for evaluating whether a
difference between those exposed to a drug and those not exposed could have
occurred simply by chance are by calculating a p-value or a confidence interval
for the relative risk.57
A 	p-value 	“represents 	the 	probability 	that 	an 	observed 	positive
association could result from random error even if no association were in fact
present.”58 “A p-value of .1 means that if in fact there is no association, 10% of
all similar studies would be expected to yield an association the same as, or
greater than, the one found in the study due to random error.”59 Stated another
way, there is a 10% chance that values at least as large as the observed relative
risk could have occurred by random error, with no association actually present
in the population.60
54 Id.
55 Id.
56 In re Lipitor, 174 F. Supp. 3d at 915.
57 Id. at 914 (citing 2011 RMSE, supra note 45, at 576, 580).
58 RMSE, supra note 36, at 933.
59 Id. at 934 n.107.
60 See id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 11 of 67

12
“To minimize false positives, epidemiologists use a convention that the
p-value must fall below some selected level known as alpha or significance level
for the results of the study to be statistically significant.”61 “The lower the p-
value, the less likely that random error would have produced 	the observed
relative risk if the true relative risk is 1.”62 “The most common significance
level in science is .05.” 63 Thus, “the statement ‘p < .05’ means that p is less
than 5%, and, by convention, the result is deemed statistically significant.”64
A confidence interval is “a range of values that reflects random error.”65
“A confidence interval is essentially a ‘margin of error’ for the estimated
relative risk ratio.”66 If the confidence interval for relative risk includes results
which do not show any increased risk, that is, includes values less than or
equal to 1, “we would say the study does not demonstrate a ‘statistically
significant’ increased risk of an adverse outcome.”67
B. 	Types of Evidence
In assessing causation, epidemiologists evaluate several types of studies,
including 	randomized 	and/or 	clinical 	trials, 	observational 	studies, 	and
toxicological studies. Randomized or clinical trials, which are considered the
“gold standard,” evaluate the development of a disease in two groups of
patients: one group that is exposed to the agent and one group that is not.68
61 In re Lipitor, 174 F. Supp. 3d at 915 (citing 2011 RMSE, supra note 45, at 580).
62 RMSE, supra note 36, at 1018.
63 In re Lipitor, 174 F. Supp. 3d at 915 (citing 2011 RMSE, supra note 45, at 577).
64 RMSE, supra note 36, at 1018.
65 Id. at 1011.
66 In re Lipitor, 174 F. Supp. 3d at 915 (quoting In re Bextra & Celebrex Mktg. Sales
Pracs. & Prods. Liab. Litig., 524 F. Supp. 2d 1166, 1174 (N.D. Cal. 2007)).
67 Id. (citation modified).
68 See RMSE, supra note 36, at 905.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 12 of 67

13
Observational studies observe a group of individuals who are exposed and
compare them with individuals who are not exposed.69 The categories of
observational studies include: (1) cohort studies; (2) case control studies; (3)
cross-sectional studies; and (4) ecological studies.70 “Cohort studies measure
and compare health outcomes . . . in the exposed and unexposed groups . . .,
while case-control studies measure and compare the frequency (or level) of
exposure in the group with the disease (the cases) and the group without the
disease 	(the 	controls).”71 	In 	cohort 	studies, 	“the 	subjects’ 	exposure 	is
determined before their health status,” 	and the “risk of disease among the
exposed can then be compared with the risk of disease among the unexposed.”72
By contrast, in a case-control study, “the health status is determined first,” and
“[t]he odds that someone with the disease was exposed to a suspected agent
can then be compared with the odds that someone without the disease was
similarly exposed.”73 In cross-sectional studies, individuals are examined for
the presence of both the exposure of interest and the disease of interest at a
single point in time.74 Ecological studies, or demographic studies, focus on the
occurrence of disease based on data 	from populations, rather than from
individuals.75 Finally, toxicological studies may be based on animals or human
or animal tissue.76
69 Id. at 906.
70 Id. at 906–11.
71 Id. at 907.
72 Id.
73 Id.
74 Id. at 911.
75 Id. at 1013.
76 See id. at 1037.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 13 of 67

14
Notably, all studies have limitations that add uncertainty about the
proper interpretation of the results.77 “Assessing whether an association is
causal requires an understanding of the strengths and weaknesses of the
study’s design and implementation, as well as a judgment about how the study
findings fit with other scientific knowledge.”78
C. 	Background on Dose-Related Issues
Because Sanofi raises several issues related to dose, a brief explanation
is 	warranted. 	As 	to 	the 	third 	Bradford 	Hill 	criterion, 	“dose-response
relationship,” the RMSE explains that “[i]f exposure to an agent causes a
disease, 	higher 	exposures 	would 	generally 	be 	expected 	to 	increase 	the
incidence or severity of that disease.”79 The RMSE defines the term as a
“relationship in which a change in amount, intensity, or duration of exposure
to an agent is associated with a change—either an increase or decrease—in
risk of disease.”80 	In turn, “dose” “refers to the intensity or magnitude of
exposure to an agent, taking into account the amount or concentration of the
agent and the duration or frequency of exposure.”81 “A linear dose response
exists when the response increases proportionally to increasing dose,” though
“[s]ome causal agents do not exhibit a linear dose–response relationship. For
example, some agents may not cause disease until the exposure exceeds a
77 See id. at 903.
78 Id. at 902–03.
79 Id. at 977.
80 Id. at 1013; see also In re Deepwater Horizon Belo Cases, 119 F.4th 937, 941 (11th
Cir. 2024) (alterations in original) (quoting McClain v. Metabolife Int’l, Inc., 401 F.3d 1233,
1241–42 (11th Cir. 2005)).
81 RMSE, supra note 36, at 1012.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 14 of 67

15
certain dose.”82 “[A] dose-response relationship is strong, but not essential,
evidence that the relationship between an agent and disease is causal.”83
A 	dose-response 	relationship 	is 	also 	an 	important 	component 	of
toxicological research. “Thus, most toxicological studies generally test a range
of doses of the chemical.”84 In this context, “dose-response relationships” are
defined as “[t]he extent to which a living organism responds to specific doses
of a toxic substance. The more time spent in contact with a toxic substance, or
the higher the dose, the greater the organism’s response.”85 For example, “a
small dose of carbon monoxide will cause drowsiness; a large dose can be
fatal.”86
Finally, 	“[e]vidence 	of 	a 	dose-response 	relationship 	as 	bearing 	on
whether an inference of general causation is justified is analytically distinct
from determining whether evidence of the dose to which a plaintiff was exposed
is required in order to establish specific causation.”87
LAW AND ANALYSIS
Sanofi seeks to exclude the general causation testimony of Drs. McGwin,
Thundiyil, and Durairaj as inadmissible pursuant to Federal Rule of Evidence
702. Sanofi also requests that the Court exclude certain testimony of Drs.
McGwin and Durairaj related to label adequacy and regulatory obligations.
Relying primarily on the arguments advanced in its Motions to Exclude, Sanofi
82 Id. at 980.
83 Id. at 981.
84 Id. at 1051.
85 Id. at 1098.
86 Id.
87 Id. at 981 n.244.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 15 of 67

16
moves for summary judgment solely on the basis that Plaintiffs lack general
causation evidence. The Court will address each expert in turn.
I. 	MOTION TO EXCLUDE DR. MCGWIN
Plaintiffs proffered Dr. McGwin as an expert in epidemiology and in the
design and analysis of observational and non-observational (i.e., clinical trial)
studies. At the outset, Sanofi does not challenge Dr. McGwin’s qualifications.
Indeed, he is a tenured professor and Vice Chair in the Department of
Epidemiology at the University of Alabama at Birmingham (UAB). He holds
an M.S. degree from Harvard University and a Ph.D. from UAB. He is a former
associate editor for the American Journal of Epidemiology and has authored or
co-authored more than 850 peer-reviewed manuscripts, with an emphasis on
the epidemiology of eye disease. He has specifically published work related to
the 	association 	between 	chemotherapeutic 	drugs, 	including
Taxotere/docetaxel, and epiphora.
Sanofi moves to exclude Dr. McGwin’s “regulatory opinions,” as well as
his general causation opinions. 	As to the former, Sanofi requests that the
Court preclude Dr. McGwin from testifying that “[s]ince Sanofi’s 2002 label
change, new and accumulating evidence contradicts the characterization that
docetaxel-related excessive tearing and cases of ‘lacrimal duct obstruction’ that
required procedural intervention were rare adverse reactions.”88 Sanofi argues
that 	in 	opining 	on 	“new 	and 	accumulating 	information,” 	Dr. 	McGwin 	is
attempting to opine that there existed “newly acquired information” such that
Sanofi could have changed its label pursuant to the changes-being-effected
88 Doc. 462-4 at 10.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 16 of 67

17
(“CBE”) regulation. In its December 1, 2025 Order and Reasons denying
Sanofi’s Motion for Summary Judgment, this Court held that Sanofi had
“newly acquired information” such that Sanofi could have changed its label
pursuant to the CBE regulation. Sanofi does not argue that Dr. McGwin is
attempting to offer any other type of “regulatory” opinion. As such, Sanofi’s
request 	to 	exclude 	Dr. 	McGwin’s 	purported 	regulatory 	opinions 	must 	be
DENIED AS MOOT.
As to general causation, in his report, Dr. McGwin applies the two-prong
methodology described above. He opines that “the epidemiological research is
sufficient to support a causal relationship between docetaxel and stenosis of
the nasolacrimal drainage system of the eye.”89 In reaching this conclusion, he
first 	notes 	that 	between 	2001 	and 	2014, 	several 	case 	series, 	as 	well 	as
descriptive and prospective studies, were published describing “docetaxel-
associated 	stenosis 	of 	the 	nasolacrimal 	drainage 	system 	of 	the 	eye 	and
epiphora.”90 He also cites five “etiological studies” published between 2020 and
2023 regarding the association between docetaxel and eye disorders, as well as
Sanofi’s clinical trial data. Next, he considers whether the association is causal
by applying the Bradford Hill criteria.
Sanofi raises four arguments in support of its position that Dr. McGwin’s
opinion is unreliable. First, Sanofi argues that in identifying an association,
Dr. McGwin impermissibly relied on a “disproportionality analysis.” Second, it
argues that Dr. McGwin must identify a statistically significant association at
each relevant “dose” (every-three-week and weekly docetaxel administration).
89 Id. at 17.
90 Id. at 12.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 17 of 67

18
Third, Sanofi points out that Dr. McGwin did not identify a “threshold dose” at
which Taxotere causes stenosis. Fourth, according to Sanofi, Dr. McGwin did
not reliably apply the Bradford Hill Factors because (1) he relied on studies
that do not focus on stenosis and/or docetaxel specifically and/or studies that
had overwhelming confounders and (2) failed to identify a “background rate”
of stenosis.
A. 	Statistically Significant Association
“An 	epidemiological 	study 	identifying 	a 	statistically 	significant
association 	between 	the 	use 	of 	a 	drug 	and 	a 	particular 	adverse 	effect,
accompanied by a reliable expert opinion that the association is causal, is
‘powerful’ 	evidence 	of 	general 	causation.”91 	“The 	purpose 	of 	statistical
significance . . . is to indicate a certain level of confidence in the results of an
analysis.”92 Here, Sanofi appears to argue that general causation testimony is
per 	se 	unreliable 	absent 	epidemiologic 	studies 	showing 	a 	statistically
significant 	association. 	To 	the 	extent 	they 	do 	so, 	the 	Court 	rejects 	that
argument.
While 	an 	expert 	who 	invokes 	the 	epidemiologic 	methodology 	must
reliably apply it, there is no requirement that an expert use it at all. In
displacing the Frye “general acceptance” test, Daubert established that the
inquiry into the reliability of expert testimony under Rule 702 is a “flexible
one”; nothing in Rule 702 renders any particular type of proof an absolute
91 In re Abilify, 299 F. Supp. 3d at 1306.
92 In re Lipitor, 892 F.3d at 641. “That a study’s results are statistically significant
says nothing about the importance of the magnitude of any association (i.e., the relative risk
or odds ratio) found in a study or about the biological or clinical importance of the finding.”
RMSE supra note 36, at 930.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 18 of 67

19
prerequisite 	to 	admissibility.93 	After 	Daubert, 	per 	se 	evidentiary 	rules
requiring experts to employ specific methodologies or to produce specific forms
of proof as a condition of admissibility are “not viable.”94 In the Fifth Circuit,
epidemiologic evidence is not necessarily required for a plaintiff to prove
general causation.95 That is not to say that Plaintiffs’ experts need not choose
a reliable methodology. And naturally, an expert must also “reliably apply the
method he chooses—but that does not mean only that method is reliable.”96
Here, the Court will follow other courts and “decline to establish a bright-line
rule 	requiring 	experts 	to 	rely 	only 	on 	[epidemiologic] 	evidence 	that 	is
statistically significant or else have their opinions excluded.”97
Regardless, 	Dr. 	McGwin 	does 	purport 	to 	identify 	a 	statistically
significant association, primarily relying on the disproportionality analysis in
his own study published in 2023 (“2023 McGwin Study”). The 2023 McGwin
Study investigated the association between lacrimal and retinal disorders and
the use of docetaxel and paclitaxel using the FDA Adverse Event Reporting
93 See U.S. v. Posado, 57 F.3d 428, 435 (5th Cir. 1995).
94 See id. at 432.
95 See Ruffin, 137 F.4th at 283.
96 Perrotti, 2025 WL 2554425, at *11.
97 In re Lipitor, 892 F.3d at 642. In Lipitor, the Fourth Circuit noted that “[a]
significant p-value is not, however, some all-purpose salve, nor is it a get-out-of-Daubert-free
card. Just as statistically significant evidence won’t result in automatic admission, the
absence of a p-value that is smaller than .05 (or some other threshold) isn’t necessarily fatal
to a case.” Id. See also RMSE, supra note 36, at 936 n.114 (collecting cases declining to require
epidemiological studies showing statistically significant results); Matrixx Initiatives, Inc. v.
Siracusano, 563 U.S. 27, 40 (2011) (noting that, in the context of materiality, “[a] lack of
statistically significant data does not mean that medical experts have no reliable basis for
inferring a causal link between a drug and adverse events,” and that “courts frequently
permit expert testimony on causation based on evidence other than statistical significance”).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 19 of 67

20
Systems (“FAERS”).98 The FDA describes “disproportionality methods” as
“largely used to identify statistical associations between products and events
in their respective databases of safety reports.”99 These methods “compare the
observed count for a product-event combination with an ‘expected’ count.”100
“Unexpectedly high reporting associations ‘signal’ that there may be a causal
association between the particular adverse event and the product.”101 The
“Proportional Reporting Ratio (PRR),” which is the “foundational concept for
many disproportionality methods,” is “the degree of disproportionate reporting
of an adverse event for a product of interest compared to this same event for
all other products in the database,” so the “entire database is used as a
background ‘expected.’”102 Stated another way, a signal of disproportionality
such as a PRR measures a statistical association within a collection of adverse
event reports.
In the 2023 McGwin Study, all adverse event reports including the terms
“docetaxel” or “paclitaxel” were selected, and lacrimal adverse events were
identified 	using 	the 	lacrimal 	disorders 	Standardization 	MedDRA 	Query
(“SMQ”). The PRR for the association between docetaxel and any lacrimal
disorder, as well as “dacryostenosis” specifically, was calculated. As to the
association between docetaxel and lacrimal events, the PRR is 2.47, and the
range is reported at 2.03-3.02 with a 95% confidence interval (“CI”). The p-
value 	is 	calculated 	at 	<.001.103 	With 	regard 	to 	the 	association 	between
98 Doc. 487-2 at 3.
99 Doc. 521-1 at 3 (FDA WHITE PAPER, D ATA M INING AT FDA (Aug. 20, 2018)).
100 Id.
101 Id.
102 Id.
103 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 20 of 67

21
docetaxel and dacryostenosis, the PRR is 19.54, with a range of 7.19-53.13,
(95% CI); the p-value is the same (<.001). Dr. McGwin concludes that the
“specific 	events 	of 	dacryostenosis, 	lacrimation 	disorder 	and 	lacrimation
increased” demonstrated “significant increased risks.”104 Therefore, the 2023
McGwin Study provides a statistically significant disproportionality signal for
lacrimal adverse events, including dacryostenosis, indicating that these events
are reported more frequently with docetaxel than comparator drugs. The 2023
McGwin 	Study, 	however, 	did 	not 	specify 	its 	findings 	as 	to 	frequency 	of
administration and/or dosages.
Dr. McGwin also provides context by citing studies “wherein it is possible
to 	calculate 	the 	proportion 	of 	docetaxel-treated 	patients 	that 	are 	either
recommended 	for 	or 	receive 	surgical 	intervention,” 	though 	these 	studies
provide 	no 	comparators.105 	Next, 	he 	references 	“case 	series 	of
docetaxel-associated stenosis of the nasolacrimal drainage system of the eye
and epiphora reported in the scientific literature.”106 Much of this literature
was authored by Sanofi’s expert, Dr. Bita Esmaeli.
He then cites, in addition to the 2023 McGwin Study, four “observational
epidemiological studies” regarding the association between docetaxel and eye
disorders. The first is a retrospective study published by Blake H. Fortes and
others in 2022 (“Fortes Study”), wherein the authors compared docetaxel to
paclitaxel in a defined clinical population (371 vs 1,533 patients) and found
“nasolacrimal duct adverse effects” occurred only in the docetaxel group (4/371
104 Doc. 487-1 at 14.
105 Doc. 462-4 at 9.
106 Id. at 12.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 21 of 67

22
vs 	0/1533, 	p 	< 	.01).107 	The 	Fortes 	Study 	therefore 	supplies 	a 	direct
exposed-versus-comparator incidence comparison, albeit with small numbers.
Second, he cites the findings of a large claims-based, retrospective cohort
study published by Mohit Sodhi and others in 2022 (“Sodhi Study”). The
authors reported an approximately five-fold increased risk of epiphora in
taxane users compared to tamoxifen users. The adjusted hazard ratio (“HR”)
for epiphora was 5.15 (95% CI, 2.79-9.54).108 The Sodhi Study provides a true
epidemiologic risk estimate, although it is not stenosis-specific and does not
isolate docetaxel alone.109
Third, he cites to a case-control study published by Yusuke Noguchi and
others published in 2020, which focused on patients administered docetaxel
every three weeks (“Noguchi Study”).110 The Noguchi Study revealed that a
cumulative docetaxel dose of ≥300 mg/m² (odds ratio: 15.50, 95% CI: 1.37-
175.00, p=.0027) was a significant risk factor associated with “eye disorders,”
which included watery eyes.
Fourth, he cites to another study using the FAERS database published
by Wang and others in 2022 (“Wang Study”). Dr. McGwin explains that in
Wang Study, the authors compared positive signals of adverse drug events
among paclitaxel and docetaxel to evaluate the accuracy of current drug
package information. For docetaxel, “periorbital and eyelid disorders” had the
strongest signal.
107 Doc. 493-135.
108 Doc. 487-12.
109 Id. at 3, 5.
110 Doc. 462-5 at 219.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 22 of 67

23
1. 	Whether the PRR analysis sufficiently identifies an
association
Sanofi does not argue that the PRR methodology is itself unreliable.
Rather, Sanofi contends that the 2023 McGwin Study does not present data
capable of establishing a reliable, statistically significant association under the
Bradford Hill methodology. They point out that because Bradford Hill assumes
the existence of an association, courts have excluded expert testimony “in the
absence of any epidemiologic studies finding an association.”111 Sanofi avers
that the 2023 McGwin Study is not an “epidemiologic study” because it is not
a randomized trial, cohort study, case-control study, or meta-analysis.
The Court disagrees. As the FDA notes, the PRR method “relies on an
‘independence assumption,’ i.e., that there is no association between products
and events mentioned in reports.”112 So, if “there is disproportionate reporting
of an event for a particular product, then this independence assumption is
questionable”—that is, “there may be an association between a particular
adverse event and the product.”113 In this sense, a PRR calculation may provide
evidence of an association, and nothing in the RMSE precludes Dr. McGwin
from relying on a disproportionality analysis. Indeed, courts frequently allow
general causation experts to identify an association via this methodology.114
111 Doc. 521 (citing Frischhertz v. SmithKline Beecham Corp., No. 10-2125, 2012 WL
6697124, at *3 (E.D. La. Dec. 21, 2012)).
112 Doc. 521-1 at 3–4.
113 Id.
114 See, e.g., In re Abilify, 2021 WL 4951944, at *5 (holding that statistical methods
employed by expert—including PRR—were reliable and allowing expert to rely on same in
identifying 	an 	association 	for 	general 	causation 	purposes); 	see 	also 	In 	re 	Fosamax
(Alendronate Sodium) Prods. Liab. Litig., Nos. 11-5304, 08–08, 2013 WL 1558690, at *8
(D.N.J. Apr. 10, 2013); In re Yasmin & YAZ (Drospirenone) Mktg., Sales Pracs. & Prods. Liab.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 23 of 67

24
To be sure, a FAERS analyses measure reporting disproportionality, not
population-level 	disease 	risk. 	Without 	incidence 	data 	(i.e., 	a 	defined
denominator), 	a 	PRR 	cannot 	establish 	“relative 	risk” 	in 	the 	traditional
epidemiologic sense. The 2023 McGwin Study, however, acknowledges this
limitation, noting that “[u]nlike a cohort study, the lack of a true denominator
(i.e. the number of persons prescribed or who use a product) precludes the
calculation of incidence rates,” but explains that the “calculation of reporting
ratios . . . from pharmacovigilance data as valid estimates of relative risks has
been described and their use is widely accepted.”115 Even so, the FDA describes
an association expressed by a PRR as “statistical; thus, it cannot be interpreted
as causal or related to risk.”116
Courts have recognized these limitations. In In re Abilify, for example,
the 	district 	court 	observed 	that 	“the 	safety 	signals 	identified 	through
disproportionality analyses in this case do not, by themselves, demonstrate a
causal 	association,” 	but 	nonetheless 	permitted 	their 	use 	as 	supporting
evidence.117 In a subsequent decision, the court again found such analyses
reliable 	to 	“investigate 	the 	possible 	existence 	of 	a 	statistical 	association
between 	Abilify 	and 	reports 	of 	certain 	adverse 	events 	in 	the 	FAERS
database.”118 The court noted that “[i]nvestigation and contextualization of a
Litig., No. 3:09–md–02100, MDL No. 2100, 2011 WL 6302573, at *17 (S.D. Ill. Dec. 16, 2011).
Indeed, this Court previously allowed an expert to identify an association, for general
causation purposes, via a PRR methodology. In re Taxotere (Docetaxel) Prods. Liab. Litig.,
No. 16-2740, 16-17144, 2019 WL 3997122, at *4–5 (E.D. La. Aug. 23, 2019).
115 Doc. 487-2 at 6.
116 Doc. 521-1 at 3 (emphasis added).
117 In re Abilify, 299 F. Supp. 3d at 1346 (citation modified).
118 In re Abilify (Aripiprazole) Prods. Liab. Litig., No. 3:16-md-2734, 2021 WL 4951944,
at *6 (N.D. Fla. July 15, 2021).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 24 of 67

25
statistical signal are critical to determining whether the signal represents a
causal relationship between a drug and a reported adverse event.”119 That
process “involves examining supporting information such as published case
reports, 	biological 	and 	clinical 	plausibility, 	clinical 	trials 	data, 	temporal
relationship, similarity to other drugs, and may also require [conducting]
epidemiological studies in healthcare databases.”120
That is how Dr. McGwin uses the database here. After identifying a
potential association—relying primarily on the 2023 McGwin Study, but also
citing additional studies in support—he applies the Bradford Hill criteria as to
numerous studies. As such, Dr. McGwin does not treat the 2023 McGwin Study
as definitive proof of causation. Notably, in many of Dr. McGwin’s cited
publications, there appears to be a consensus that docetaxel is, at the very
least, associated with an increased risk of stenosis. Dr. McGwin’s approach is
not unlike a “weight of the evidence” approach that courts have held to be
permissible.121
119 Id.
120 Id. (alterations in original) (citation modified) (first citing Harpaz et al., Novel
Data-Mining Methodologies for Adverse Drug Event Discovery and Analysis, 91 CLINICAL
PHARMACOLOGY & THERAPEUTICS 1010, 1013 (2012); then citing Harpaz et al., Performance
of Pharmacovigilance Signal Detection Algorithms for the FDA Adverse Event Reporting
System, 93 CLINICAL PHARMACOLOGY & THERAPEUTICS 539 (2013)); see also Doc. 521-1 at 12
(noting that “[h]ands-on case reviews,” as well as analyses of other 	“public scientific
literature,” and “further epidemiologic assessments are necessary to characterize the clinical
and public health significance of safety ‘signals’ generated by data mining analyses”).
121 See David L. Faigman et al., 3 M ODERN SCIENTIFIC EVIDENCE § 23:15 (2025-2026
ed.) (noting that “[t]he weight-of-the-evidence approach is frequently discussed in terms of
the Bradford Hill criteria” and that “[a]lthough Sir Bradford Hill imagined the criteria as a
way to assess causation in epidemiologic research, many of the factors, e.g., specificity,
temporality, dose-response, and plausibility, are clearly relevant to the assessment of other
types of evidence and courts frequently invoke the Bradford Hill criteria as a way to assess
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 25 of 67

26
Sanofi further asserts that the 2023 McGwin Study is unreliable because
FAERS is composed of unverified, uncontrolled, and anecdotal case reports
that are not subject to medical corroboration. They also contend that when
asked about the fact that half of all the docetaxel case reports in the 2023
McGwin Study were reported in a 24-month period, Dr. McGwin acknowledged
that these were presumably related to this litigation. Although there are
limitations with FAERS data, this does not warrant exclusion. Sanofi may
explore 	these 	issues 	through 	cross-examination 	and 	the 	presentation 	of
contrary evidence.122
2. 	Whether 	Dr. 	McGwin 	must 	show 	an 	association 	at
weekly and every-three-week administration
Next, Sanofi argues that Dr. McGwin’s opinion is unreliable because he
failed to show a statistically significant association at each relevant schedule
of administration (that is, weekly and every-three-weeks). It asserts that for
purposes of identifying a statistically significant association, “dose matters,”
and that dose is “a product of both the concentration of a chemical or physical
agent and the duration or frequency of exposure.”123 Plaintiffs respond that Dr.
McGwin sufficiently identified an association via his PRR analysis and that he
need not do so for each relevant “dose.”
The Court finds that neither the caselaw nor the RMSE require Dr.
McGwin to identify an independent association at each administration of
an expert’s weight-of-the evidence analysis”); see also In re Abilify, 299 F. Supp. 3d at 1311;
In re Abilify, 2021 WL 4951944, at *6; In re Fosamax, 2013 WL 1558690, at *4.
122 See In re Abilify, 2021 WL 4951944, at *5.
123 Doc. 450-1 at 5 (citing Simon v. Grand Isle Shipyard, Inc., No. CV 11-1432, 2023
WL 2430048, at *5 (E.D. La. Mar. 9, 2023)).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 26 of 67

27
docetaxel. In support of its argument, Sanofi cites to two district court cases
outside of this jurisdiction, In re Lipitor and In re Bextra.124 In those cases, the
district courts assessed the experts’ causation analyses separately for each
dose at which the drugs at issue were administered. Other district courts,
however, 	have 	rejected 	any 	such 	requirement.125 	This 	makes 	sense: 	a
“dose-response” relationship is assessed through the Bradford Hill criteria, the
second step of the analysis.
But even if Bextra and Lipitor were binding on this Court, both are
distinguishable. In those cases, the drug at issue was prescribed at one of
several commercially available doses.126 By contrast, Plaintiffs in this case
were administered docetaxel at various dosing schedules (such as once a week
or every three weeks), as well as in different treatment settings (metastatic,
adjuvant, and neoadjuvant) and for varying durations, all of which affect
dose.127 Further, while Sanofi contends that Dr. McGwin agrees that the
124 Doc. 462-1 at 5–6 (citing In re Bextra, 524 F. Supp. 2d 1166); Doc. 526 at 3 n.2
(citing In re Lipitor, 174 F. Supp. 3d 911).
125 See In re: Zicam, 797 F. Supp. 2d at 945.
126 See In re Bextra, 524 F. Supp. 2d 1166; In re Lipitor, 174 F. Supp. 3d 911.
127 In an editorial responding to a 2013 study, Sanofi’s expert Dr. Esmaeli and her co-
authors stated that while docetaxel administered every three weeks for a shorter duration
are not “the subject of greatest concern,” those who receive docetaxel weekly, or every three
weeks for extended periods (such as in the metastatic setting), are among those patients who
are “at the greatest risk of canalicular stenosis.” Doc. 493-131 at 2. Dr. Esmaeli also opines
in her Report that when considering the association between stenosis and weekly docetaxel,
“the duration of treatment, frequency of intravenous administration of the drug, and
cumulative dose are all factors to consider. Doc. 462-3 at 22. While she appears to
acknowledge that there is an association between docetaxel administered in the weekly
setting and stenosis, she also opines that there is insufficient evidence with respect to the
association between docetaxel administered every three weeks in shorter durations (such as
in the adjuvant setting). Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 27 of 67

28
evidence differs with respect to dosing and that there is a “dose-response”
relationship, 	it 	does 	not 	specifically 	identify 	any 	studies 	that 	directly
undermine his conclusions, as was the case in Bextra and Lipitor. The Court
will not, therefore, require Dr. McGwin to provide a separate analysis for
weekly and every-three-week administration or evidence of an association at
every conceivable dose. That kind of inquiry is best addressed through a
dose-response analysis and/or specific causation.128
B. 	Whether 	a 	Showing 	of 	“Threshold” 	or 	“Toxic” 	Dose 	Is
Required
Next, Sanofi contends that Dr. McGwin’s opinion is unreliable because
he failed to show a “toxic” or “threshold” dose at which Taxotere causes
stenosis—which Sanofi contends is a requirement for a general causation
expert’s opinion to be admissible under Fifth Circuit caselaw. This argument
is without merit. In the recent decision Ruffin v. BP, the Fifth Circuit explicitly
rejected the defendant’s argument that “to be admissible, an expert must
‘identify the minimum amount of a particular chemical necessary to cause a
plaintiff’s 	alleged 	condition 	in 	the 	general 	population.”’129 	Framing 	the
defendant’s contention as a relevance objection, the court noted that Daubert
128 Other courts have explained that “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.” Davis v. McKesson Corp.,
Nos. CV-18-1157, CV-18-1159, CV-18-1778, 2019 WL 3532179, at *11 n.5 (D. Ariz. Aug. 2,
2019) (citing 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.” Id. (first citing In re:
Zicam, 797 F. Supp. 2d at 944–45; then citing Henricksen v. ConocoPhillips Co., 605 F. Supp.
2d 1142, (E.D. Wash. 2009)).
129 Ruffin, 137 F.4th at 281.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 28 of 67

29
and Federal Rule of Evidence 702 “control whether evidence is admissible, and
they each provide that evidence is admissible if, among other things, it is
‘relevant’ to the applicable legal element,” that is, general causation.130 Under
the governing law, this required only a showing that the substance was capable
of causing the condition at issue in the general population.131 The court further
explained that a quantitative estimate was not required and that, “even when
an expert does provide a quantitative-dosage estimate, we have allowed a
numerical ‘range’ supported by other, non-dosage associations, such as the
chemical’s 	‘toxicological 	profile’ 	and 	the 	‘temporal 	connection’ 	between
workplace exposure to a chemical and the incidence of a condition.”132
The court also stressed that the Bradford Hill criteria 	“consider a
dose-response relationship” as “but one factor among others,”133 and cited
authority contrasting general causation, in which dose is not a central issue,
with specific causation, in which the primary issue is whether there has been
exposure to a sufficient dose to be a likely cause of the condition.134 It
ultimately concluded that the “general-causation caselaw is inconsistent with
130 Id. (quoting Daubert, 509 U.S. at 591).
131 Id. (quoting Knight, 482 F.3d at 351). The Fifth Circuit acknowledged that this
standard—articulated in Knight v. Kirby—derived from Texas state law; nevertheless, the
court applied it in the maritime context. See id. at 280 n.1 (noting that “we have still
previously required general and specific causation for toxic-tort claims arising under general
maritime law,” and that “even though the Knight rule derives from state law, we nevertheless
applied it there to a claim that arose under the Jones Act (citing Knight, 482 F.3d at 350)).
132 Id. at 283 (citing Curtis v. M&S Petroleum, Inc., 174 F.3d 661, 669–70, 670 n.8 (5th
Cir. 1999)).
133 Id. (citation modified).
134 See id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 29 of 67

30
requiring a quantitative dosage for expert testimony to be relevant under Rule
702 or Daubert.”135
As one district court has explained, Ruffin “teaches us not to confuse
what is sufficient to prove causation—a precise quantitative dose—with what
is required—some level of exposure that shows the substance is capable of
causing 	the 	relevant 	harm.”136 	This 	makes 	sense 	because 	“[d]ose 	is 	not
synonymous 	with 	specific 	concentration 	amounts 	anyway—part 	of 	dose
includes duration of exposure.”137 Other courts have followed suit and have
declined to extend the “toxic dose” requirement “from environmental exposure
litigation to drug products liability actions.”138 “That the toxic dose issue does
not arise more frequently in the drug products liability context makes sense,”
because “[i]n environmental exposure litigation, plaintiffs may allege injury
caused by a substance with which many people interact harmlessly at lesser
degrees 	of 	exposure.”139 Thus, 	“to 	explain 	how 	a 	pervasive 	substance 	is
harmful, one must show that at a particular level of exposure, the substance
becomes toxic.”140 Absent such a requirement, there is a risk of plaintiffs filing
meritless claims based on harmless exposures. “But that risk is less severe in
drug product liability actions.”141
135 Id. at 283.
136 Perrotti, 2025 WL 2554425, at *15 (citing Ruffin, 137 F.4th at 282).
137 Id.
138 In re: Zicam, 797 F. Supp. 2d at 945.
139 Id.
140 Id.
141 See id. at 946.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 30 of 67

31
As such, the Court finds that Dr. McGwin need not provide a “toxic” or
“threshold dose” to be admissible. Rather, the pertinent question is whether he
adequately considered the “dose-response” factor.142
C. 	Dose-Response Analysis
Sanofi argues that Dr. McGwin did not reliably apply the dose-response
factor. The Court disagrees. In the epidemiological context, a dose-response
relationship 	means 	that 	“[a] 	relationship 	in 	which 	a 	change 	in 	amount,
intensity, or duration of exposure to an agent is associated with a change—
either an increase or a decrease—in risk of disease.”143 In considering this
factor (which Dr. McGwin refers to as “biological gradient”), Dr. McGwin
opines that a number of studies have demonstrated that patients treated with
docetaxel who develop stenosis have a higher cumulative dose than those who
do not.
In assessing this criterion, Dr. McGwin cites a study published by Dr.
Esmaeli and others in 2002 in Ophthalmology titled “Canalicular Stenosis
Secondary 	to 	Weekly 	Versus 	Every-3-Weeks 	Docetaxel 	in 	Patients 	with
Metastatic Breast Cancer” (“2002 Esmaeli Ophthalmology Study”), wherein
the authors compared canalicular stenosis secondary to weekly versus every-
three-weeks 	docetaxel 	in 	patients 	with 	metastatic 	breast 	cancer.144 	The
authors reported on the pattern of docetaxel administration, as well as the
“dose intensity” and “mean cumulative dose” that each patient received.145
142 Notably, as to weekly administration, Dr. Esmaeli admits that there is no precise
threshold dose.
143 See RMSE, supra note 36, at 1013.
144 Doc. 462-5 at 44.
145 Id. at 44–46.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 31 of 67

32
“The 	mean 	cumulative 	dose 	of 	docetaxel 	was 	higher 	in 	patients 	with
canalicular stenosis than in patients without canalicular stenosis,” with 712
mg/m² (±201) for stenosis versus 361 mg/m² (±172) without. The authors
concluded that “canalicular stenosis and epiphora were more common in
patients receiving docetaxel weekly.”146
Dr. McGwin also cites additional studies that indicate that a higher
cumulative dose of docetaxel may increase the incidence of stenosis. For
example, in 2002, Dr. Esmaeli and others published a study in the Annals of
Oncology 	titled 	“Canalicular 	stenosis 	secondary 	to 	weekly 	docetaxel: 	a
potentially preventable side effect” (“2002 Esmaeli Annals of Oncology Study”).
There, the authors tied severity to cumulative dose.147 The authors concluded
that “[t]he mean cumulative dose of docetaxel at the time of diagnosis of
epiphora was higher in patients with severe canalicular stenosis than in
patients with moderate canalicular stenosis,” (399 mg/m² vs. 501 mg/m²,
P=0.09).
He also cites a 2006 study authored by Medy Taslic and others titled
“Epiphora (Excessive Tearing) and Other Ocular Manifestations Related to
Weekly Docetaxel,” which was published in Medical Oncology (“2006 Tsalic
Study”). 	The 	authors 	reported 	epiphora 	related 	to 	canalicular 	stenosis
occurring in seven of twenty-one (33%) patients,” with two patients developing
“complete canalicular stenosis requiring surgery” who received docetaxel at
208 and 645 mg/m².148
146 Id. at 46.
147 Doc. 487-7 at 2.
148 Doc. 462-5 at 162.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 32 of 67

33
Sanofi 	also 	argues 	that 	Dr. 	McGwin 	failed 	to 	reliably 	apply 	the
dose-response factor because he relies on the Noguchi Study, which does not
focus on stenosis specifically. The Noguchi Study reported on patients taking
docetaxel every three weeks who developed “eye disorders,” defined as “an
event in which the pharmacist confirmed the symptoms in a patient interview
and the ophthalmologist diagnosed the disorder.”149 The authors reported that
7 out of 89 total patients (7.9%) had eye disorders and symptoms of “watering
eyes.”150 Of these, 2.2% had a stye and eye discharge. The reported cases also
included 	corneal 	and 	conjunctival 	disorder, 	visual 	acuity 	reduction, 	and
blepharedema (1.1% each). Four patients who presented with watering eyes,
eye discharge, or corneal and conjunctival disorder showed improvement with
the use of eye drops such as artificial tears. The authors concluded that
docetaxel-related eye disorders might be influenced by the cumulative dose of
docetaxel, 	as 	well 	as 	cumulative 	docetaxel 	dose 	and 	concomitant
cyclophosphamide use.
The Noguchi Study, standing alone, would be insufficient. Exclusion is
not warranted, however, because Noguchi supports Dr. McGwin’s conclusions
as to the 2002 Esmaeli Study; the authors specifically cite it as reporting that
the cumulative dose of docetaxel was “significantly higher in patients with
canalicular stenosis than in those without canalicular stenosis.”151
Here, 	the 	Court 	finds 	that 	Dr. 	McGwin 	adequately 	addressed 	the
dose-response criterion, and he need not identify a threshold dose to be
149 Doc. 462-5 at 221.
150 Id. at 219.
151 Id. at 220.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 33 of 67

34
admissible.152 Any differences with respect to weekly v. every-three-week
administration go to the weight of Dr. McGwin’s opinion, not its reliability.
D. 	Strength
Next, 	Sanofi 	argues 	that 	Dr. 	McGwin 	failed 	to 	reliably 	apply 	the
“strength” factor. “The relative risk is a commonly used measure of association
that represents the strength of an association.”153 “Larger relative risks (or
stronger associations using other statistical measures) are often believed to be
more likely to be causal than smaller ones.”154 While “lower relative risks can
reflect causality, the epidemiologist will scrutinize such associations more
closely 	because 	there 	is 	a 	greater 	chance 	that 	they 	are 	the 	result 	of
uncontrolled confounding or biases.”155 Dr. McGwin opines that the results of
(1) the 2023 McGwin Study; (2) the Fortes Study; (3) the Wang Study; (4) the
Sodhi Study; and (5) Sanofi’s clinical trial data are consistent with a strong
association 	between 	docetaxel 	exposure 	and 	stenosis 	of 	the 	nasolacrimal
drainage system of the eye resulting in epiphora. Sanofi argues that some of
the studies upon which Dr. McGwin relies do not specifically identify stenosis.
Further, it points out that while Dr. McGwin acknowledged that not every
report of tearing in a study is a report of stenosis of the tear ducts, he does not
152 See In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 16-2740, 16-17144, 2019
WL 3997122, at *8 (E.D. La. Aug. 23, 2019) (finding expert (1) adequately considered “dose
response factor” and (2) was not required to give a definitive dose, where she testified that
“not every study is identical with when the observations are seen what dose,” and that she
thought that “the pertinent point is that there was a dose response”).
153 RMSE, supra note 36, at 977.
154 Id.
155 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 34 of 67

35
attempt to differentiate between reports of stenosis and those of tearing
without underlying stenosis.
The Fifth Circuit does not “require an expert to back his or her opinion
with published studies that unequivocally support his or her conclusions.”156
“Nonetheless, the expert’s testimony must be reliable at each and every step
or else it is inadmissible.”157 “District courts must carefully analyze the studies
on which experts rely for their opinions before admitting their testimony.”158
Further, “while ‘[t]rained experts commonly extrapolate from existing data,’
the Supreme Court has held that ‘nothing in either Daubert or the Federal
Rules of Evidence requires a district court to admit opinion evidence that is
connected to existing data only by the ipse dixit of the expert.’”159 “In such a
situation, ‘a court may conclude that there is simply too great an analytical gap
between the data and the opinion offered.’”160 Additionally, studies that do not
examine the precise disease at issue may not be relevant, that is, “fit” the facts
of the case and thereby assist the trier of fact to understand the evidence.161
In support of its argument, Sanofi cites a decision from another section
of this Court, Burst v. Shell Oil.162 There, the plaintiff’s expert asserted that
exposure to low levels of benzene present in gasoline can cause acute myeloid
leukemia (“AML”) and that the decedent’s exposure to gasoline containing
156 See Burst, 650 F. App’x at 174.
157 Id.
158 Id.
159 Id. at 173 (quoting Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146 (1997)).
160 Id. (quoting Joiner, 522 U.S. at 146).
161 Burst, 2015 WL 3755953, at *3.
162 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 35 of 67

36
benzene caused his disease.163 The expert relied primarily on studies that
focused on benzene, rather than gasoline.164 The parties did not dispute that
gasoline contains benzene or that benzene, at sufficiently high exposure levels,
is capable of causing AML.165 But “no scientific authority” had “classified
gasoline as a human carcinogen.”166 The defendant argued, among other
things, that the expert’s reliance on benzene studies was improper.167
The court agreed, 	emphasizing that the scientific literature addressing
gasoline 	exposure 	did 	not 	mirror 	the 	literature 	concerning 	benzene
exposure.168 The defendant also presented evidence that benzene is present in
gasoline only in small concentrations and that gasoline constituents can
interfere with benzene’s toxic effects through competitive inhibition.169 The
court ultimately determined that although the evaluation of the benzene
literature was generally relevant to the expert’s ultimate opinion, 	it alone
could not provide a reliable basis for that opinion.170 The court also found
numerous 	methodological 	flaws 	in 	the 	expert’s 	analysis 	of 	the 	gasoline
literature.
Applying Burst to the present facts, Dr. McGwin’s reliance on studies
involving 	taxanes 	generally, 	other 	comparator 	drugs, 	or 	broader 	ocular
conditions does not automatically render his opinion unreliable. Rather, Burst
163 Id.
164 See id. at *4.
165 Id. at *9.
166 Id.
167 Id. at *4.
168 Id. at *9.
169 Id. at *10.
170 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 36 of 67

37
suggests that such studies may be considered relevant supporting evidence.171
Here, Dr. McGwin adequately explains how these studies provide support for
his opinions. Dr. McGwin first notes that Fortes and the 2023 McGwin study
“report statistically significant higher risks of nasolacrimal duct adverse effect
and 	lacrimal 	adverse 	events 	(specifically 	dacryostenosis 	and 	increased
lacrimation), 	respectively, 	for 	those 	receiving 	docetaxel 	compared 	to
paclitaxel.”172 He then cites the findings in Sodhi and Wang as additional
support. He explains that “the literature inconsistently refers to stenosis,
epiphora, and lacrimal duct obstruction,” but concludes that “there is sufficient
overlap to support the strength of association.”173
Unlike in Burst, this conclusion is not necessarily contradicted by the
literature. For example, though Sodhi reported on the association between
“epiphora” 	and 	taxanes, 	“epiphora” 	at 	least 	included 	reported 	cases 	of
“stenosis.”174 The Sodhi authors also recognized that the “proposed mechanism
for taxane-induced epiphora is stenosis.”175 Thus, while Sodhi does not allow
for the extraction of 	the exact number of reported cases of stenosis, it is
supportive of Dr. McGwin’s opinion, particularly when combined with the 2023
McGwin Study’s findings.
171 See id. at *9–10.
172 As to Fortes, of the 371 docetaxel patients, there were 11 reported cases of an
“ophthalmic effect.” Of these, there were 4 reported cases of “canalicular obstruction,” 1 case
of “punctal stenosis,” 1 case of “binocular diplopia,” 1 case of “cranial nerve six palsy,” 2 cases
of “meibomian gland dysfunction” (MGD”), 1 “chalazion,” and 1 “blepharitis with MGD.” Doc.
493-135.
173 Doc. 487-12 at 3.
174 Id.
175 Id. at 5.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 37 of 67

38
Further, Dr. McGwin identifies additional limitations in the studies. For
example, as to Sodhi, Dr. McGwin notes that although the authors do not
report taxane-specific associations, “Fortes and McGwin suggest that docetaxel
is driving the association.”176 As to Fortes, although it was not possible to
“calculate a quantitative measure of association,” this was so because there
were “zero events in the paclitaxel group,” the “comparator” group.177 As to
Wang, it “reported a remarkably strong association between docetaxel and
periorbital and eyelid disorders, though associations for specific conditions
within this class of adverse effects are not reported.”178 As to Sanofi’s clinical
trial data, he acknowledges its limitations by noting that the adverse events
reported in these trials, which are summarized in the docetaxel label, include
many events such as “lacrimation disorder or related conditions.”179
Here, there is not “simply too great an analytical gap between the
underlying data and the opinion offered.”180 The Court recognizes that it “may
exclude expert testimony based on epidemiological studies where the studies
are insufficient, whether considered individually or collectively, to support the
expert’s causation opinion.”181 Nevertheless, the Court finds that Dr. McGwin’s
opinions 	are 	sufficiently 	supported 	by 	the 	literature. 	Certainty 	is 	not
required.182
176 Doc. 462-4 at 15.
177 Id.
178 Id.
179 Id.
180 See Burst, 650 F. App’x at 173.
181 Burst, 2015 WL 3755953, at *6.
182 See Burst, 650 F. App’x at 174.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 38 of 67

39
E. 	Failure to Identify “Background Risk” of Stenosis
“A temporal, or chronological, relationship must exist for causation to
exist.”183 “If an exposure causes disease, the exposure must occur before the
disease develops.”184 In assessing temporality, Dr. McGwin opines that “[t]he
majority of etiologic epidemiologic studies are randomized controlled trials or
prospective 	studies, 	thus 	the 	issue 	of 	the 	temporality 	of 	the 	observed
association is clear.”185 Sanofi contends that Dr. McGwin did not reliably apply
the “temporality” criterion because he failed to consider the “background rate”
of stenosis. As Plaintiffs point out, however, temporality refers to the temporal
relationship between a drug and disease, not necessarily the background rate.
Plaintiffs further argue that Dr. McGwin adequately considered the
background rate because in comparative studies, the background rate of
disease applies to persons exposed and unexposed to docetaxel, so any observed
increased risk is over and above the background rate of those symptoms and
conditions. They contend that a reliable analysis need not take this into
account other than with the choice of a suitable comparison group. In his
deposition, Dr. McGwin explained:
Q: 	You 	would 	agree 	that 	there 	is 	a 	certain 	segment 	of 	the
population that has stenosis of the nasolacrimal drainage system
before they are ever exposed to Taxotere; correct?
A: I would agree with you that that is true.
183 RMSE, supra note 36, at 975.
184 Id.
185 Doc. 462-4 at 16.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 39 of 67

40
Q: And in forming your opinions in this case, what determination
did you make as to what that background rate was for individuals
before they are ever exposed to Taxotere?
A: In this case, I didn’t need to make that determination. The
individual 	studies 	that 	I 	included 	would 	have 	made 	that
determination. If I were doing my own study, it would have been a
different issue.186
Indeed, in his report, Dr. McGwin cites studies that compare patients exposed
to docetaxel and paclitaxel, such as the 2022 Sodhi Study and 2023 McGwin
Study. The data from both the exposed and unexposed groups would account
for any alleged background rate. Dr. McGwin also cites studies that do not
necessarily have a “comparison” group. In some of those studies, however, a
baseline evaluation was performed on patients at the start of the study.187
Thus, since one can assume that those studies did not include pre-existing
cases of stenosis, Dr. McGwin would not need to consider “background risk.”
Sanofi points out that in several studies cited by Plaintiffs in their
Opposition, the authors only evaluated patients after they received docetaxel.
Specifically, Sanofi cites to the 2002 Esmaeli Annals of Oncology Study, the
2002 Esmaeli Ophthalmology Study, and the 2006 Tsalic Study. Although in
these studies patients were not excluded at baseline based on asymptomatic
186 Doc. 462-2 at 6.
187 For example, in 2006, Dr. Esmaeli and others published “Prospective Study of
Incidence and Severity of Epiphora and Canalicular Stenosis in Patients With Metastatic
Breast Cancer Receiving Docetaxel” in the Journal of Clinical Oncology (“2006 Esmaeli
Study”), which reports the findings of a prospective study of patients receiving docetaxel
weekly and every three weeks. Doc. 462-5 at 167. The authors noted that before treatment
with docetaxel, patients were evaluated by the same ophthalmologist and underwent a
comprehensive ophthalmologic examination. Id. at 168.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 40 of 67

41
stenosis, each study enrolled patients who were symptom-free at entry and
identified 	cases 	only 	when 	clinically 	apparent 	epiphora 	developed 	after
docetaxel exposure, at which point stenosis was confirmed through objective
ophthalmologic evaluation. Thus, at the very least, the relevant outcome
(stenosis manifesting as epiphora) was not present at baseline.
Even so, Sanofi points out that courts have recognized that “[i]gnoring
available evidence about background risks may be fatal to an expert’s general
causation opinion.”188 In support, it cites one publication that it contends
addresses background risk of stenosis (the “2013 Chan Study”), which is a
prospective study that aimed to assess tearing in early breast cancer patients
receiving docetaxel every three weeks. The authors reported that, at baseline
evaluation, asymptomatic “lacrimal duct obstruction” (“LDO”) was present in
17–18% of patients.
At the outset, Sanofi argues that Dr. McGwin does not identify or
describe this (or any) reported background rate in his report. But in arguing
that Dr. McGwin impermissibly relies “heavily on studies that report on
conditions other than stenosis of the nasolacrimal drainage system,” Sanofi
specifically cites, among other publications, the 2013 Chan Study as one that
Dr. McGwin “concedes” does not specifically address stenosis.189 That Sanofi
faults Dr. McGwin for failing to rely on Chan’s findings on background risk
188 In re Deepwater Horizon Belo Cases, No. 3:19CV963, 2022 WL 17721595, at *13
(N.D. Fla. Dec. 15, 2022), report & recommendations adopted by No. 3:19CV963, 2023 WL
2711573 (N.D. Fla. Mar. 30, 2023), aff’d, 119 F.4th 937 (11th Cir. 2024).
189 Doc. 462-1 at 12 n.18.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 41 of 67

42
while objecting to his reliance on it because it reports on a different injury is
puzzling.190
Regardless, Dr. McGwin does not ignore Chan’s findings. Rather, he
explains that although the authors concluded that docetaxel-related tearing
was 	likely 	“not 	caused 	by 	lacrimal 	duct 	obstruction,” 	the 	study’s 	design
undermines that conclusion for at least three reasons. First, according to the
study’s methodology, patients with asymptomatic LDO should have been
excluded, but the results suggest otherwise.191 Second, patients were evaluated
using 	two 	diagnostic 	approaches 	(CT-DCG 	and 	ophthalmic 	examination),
which 	“demonstrated 	poor 	agreement, 	thus 	raising 	the 	issue 	of
misclassification.”192 	Dr. 	McGwin 	further 	notes 	that 	his 	critiques 	are
consistent with the opinions of Dr. Durairaj, who explains that “CT-DCG is not
the standard of care.”193 Third, the authors themselves acknowledge that the
2013 Chan Study “was not designed to evaluate specific causes of tearing,”
meaning “their conclusion that docetaxel-related tearing is not caused by
lacrimal duct obstruction cannot be supported by the study design.”194 These
diagnostic limitations call into question not only the study’s overall conclusions
but also the reliability of the reported “background rate.”
190 Like other studies upon which Plaintiffs’ experts do rely, the 2013 Chan Study
identifies cases of “lacrimal duct obstruction.” Elsewhere in this litigation, Defendants have
taken the position, and Plaintiffs disagree, that the term “lacrimal duct obstruction” on the
Taxotere label sufficiently encompasses “stenosis.” The Court clarifies that it makes no
determination on that issue, or any issue related to label adequacy, here.
191 Doc. 462-4 at 14.
192 Id. at 14.
193 Doc. 455-2 at 27.
194 Doc. 462-4 at 14. The authors in the 2013 Esmaeli Editorial echo these concerns.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 42 of 67

43
Although Sanofi may disagree with Dr. McGwin’s conclusions, “where an
expert’s testimony is within ‘the range where experts may reasonably differ,’
the jury, not the trial court, should ‘decide among the conflicting views of
different experts.’”195 Here, because Dr. McGwin identifies and accounts for the
“limited body of scientific literature currently available on background risk,”
the fact that he does not “offer a more expansive analysis of background risk .
. 	. 	does 	not 	present 	a 	‘serious 	methodological 	deficiency’ 	or 	‘substantial
weakness’” in his general causation opinions.196
F. 	Remaining Bradford Hill Criteria
Finally, although Sanofi does not specifically address the remaining
Bradford Hill criteria, the Court finds that Dr. McGwin adequately addresses
them. For example, as to the fifth factor, where the biological mechanism of an
injury is not fully established, the plausibility inquiry asks whether the
proposed causal explanation is consistent with existing medical and scientific
knowledge, recognizing that well-supported hypotheses may suffice even in the
absence of definitive mechanistic proof.197 In support of this factor, which Dr.
McGwin refers to as “plausibility and coherence,” Dr. McGwin relies on two
studies published by Dr. Esmaeli and others examining how docetaxel may
affect the eye’s tear drainage system.
First, he cites a study published by Dr. Esmaeli and others in titled
“Docetaxel secretion in tears: association with lacrimal drainage obstruction”
in the Archives of Ophthalmology (“2002 Esmaeli Arch. Ophthal. Study”). The
195 See In re Abilify, 2021 WL 4951944, at *3 (quoting Kumho Tire, 526 U.S. at 137).
196 See In re Abilify, 299 F. Supp. 3d at 1308 (quoting Chapman v. Procter & Gamble
Distrib., LLC, 766 F.3d 1296, 1308 (11th Cir. 2014)).
197 See Milward, 639 F.3d at 25; see also RMSE, supra note 36, at 605.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 43 of 67

44
2002 Esmaeli Arch. Ophthal. Study was a prospective trial of six patients who
received docetaxel on a weekly or every-3-week interval that measured the
subjects’ tears for the concentration of the drug. All six patients had docetaxel
measured in their tears, and the two patients with the two highest doses of
docetaxel had the highest concentration in their tears. The authors stated that
“the detection of docetaxel in tears collected” from all six patients after “its
injection confirms our hypothesis that the ocular surface irritation and fibrosis
of the tear-drainage ducts seen following treatment with this drug may be
partly due to direct contact between the drug and both the conjunctival surface
and mucosal lining of the tear drainage apparatus.”198 Dr. McGwin explains
that the 2002 Esmaeli Arch. Ophthal. Study “provided evidence that the
secretion 	of 	docetaxel 	in 	tears 	may 	be 	a 	mechanism 	for 	canalicular
inflammation.”199
Then, 	in 	2003, 	Dr. 	Esmaeli 	and 	others 	published 	a 	study 	titled
“Docetaxel Induced Histological Changes in the Lacrimal Sac and Nasal
Mucosa” in Ophthalmic Plastic and Reconstructive Surgery (2003 Esmaeli
Ophthalmic 	Study”). 	Dr. 	McGwin 	explains 	that 	in 	the 	2003 	Esmaeli
Ophthalmic Study, “biopsies of patients revealed histopathological changes in
the lacrimal sac and the nasal mucosa caused by docetaxel.”200 The authors
ultimately 	concluded 	that 	“that 	nasolacrimal 	duct 	obstruction 	following
docetaxel treatment may be due to stromal fibrosis in the mucosal lining of the
198 Doc. 446-6 at 3.
199 Doc. 462-4 at 16.
200 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 44 of 67

45
lacrimal drainage apparatus.”201 Thus, Dr. McGwin identifies a plausible
mechanism for stenosis.
In sum, the Court finds that Dr. McGwin’s general causation testimony
is admissible. Sanofi’s Motion to Exclude Dr. McGwin is therefore DENIED.
II. 	MOTION TO EXCLUDE DR. THUNDIYIL
Plaintiffs offer Dr. Thundyil as an expert in toxicology. Dr. Thundiyil
performed a review of the literature with respect to the causal association
between docetaxel and ocular pathology. Like Dr. McGwin, he applies the
Bradford Hill factors. Although Sanofi does not challenge his qualifications,
Dr. Thundiyil is a licensed medical doctor practicing in the state of Florida with
Board 	Certifications 	in 	Emergency 	Medicine, 	Medical 	Toxicology, 	and
Occupational and Environmental Medicine. He notes that he has more than 20
years of clinical experience with independent practice.
Sanofi 	asserts 	three 	arguments 	that 	Dr. 	Thundiyil’s 	opinion 	is
unreliable: (1) he does not identify a statistically significant association; (2) he
did not reliably apply the dose-response criterion; and (3) he failed to reliably
apply the remainder of the Bradford Hill criterion because his analysis is “rife
with errors, cherry-picking, untested hypotheses, fit problems, and missing
analyses.”202 The Court will consider each in turn.
A. 	Statistically Significant Association
Sanofi points out that Dr. Thundiyil did not specifically address the first
prong of the analysis, that is, identify a statistically significant association
201 Id.
202 Doc. 446-1 at 17.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 45 of 67

46
between docetaxel and stenosis. As explained above, ordinarily, where an
expert 	invokes 	the 	two-prong 	epidemiologic 	methodology, 	the 	expert 	is
required to first identify an association that is statistically significant. Here,
however, the Court finds that Dr. McGwin has adequately identified an
association 	between 	docetaxel 	and 	stenosis 	via 	studies 	that 	include
statistically significant results. As such, Dr. Thundiyil need not do so. “From a
scientific perspective and for legal causation purposes, there is no requirement
that the same expert who identifies a statistical signal conduct the entire
causation inquiry.”203 The Court recognizes that while Plaintiffs argue that
they provided “complimentary” expert opinion testimony from Dr. Thundiyil
and Dr. McGwin, Dr. Thundiyil provides a standalone general causation
analysis. Nevertheless, Dr. Thundiyil cites several studies also relied upon by
Dr. McGwin, including the 2023 McGwin Study. Accordingly, Dr. Thundiyil
need 	not 	explicitly 	identify 	the 	existence 	of 	a 	statistically 	significant
association to be admissible.
B. 	Dose-Response Analysis
Next, Sanofi contends that Dr. Thundiyil’s dose-response analysis is
inadmissible because he did not provide a threshold dose at which Taxotere
causes stenosis. As explained above, however, neither the RMSE nor the
caselaw require this. Accordingly, this argument must be denied.
Sanofi further argues that Dr. Thundiyil’s dose-response analysis is not
reliable 	because he 	failed to identify 	reliable 	evidence in support of his
203 In re Abilify, 2021 WL 4951944, at *6; see also In re Taxotere, 2019 WL 3997122, at
*3; In re Abilify, 299 F. Supp. 3d at 1312-30, 1361-68 (finding Dr. Madigan “amply qualified”
to offer a biostatistical analysis of scientific evidence but unqualified to offer expert opinions
on medical causation).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 46 of 67

47
conclusions. Sanofi points out that Dr. Thundiyil only specifically cites two
sources in support of his dose-response inquiry. First, he cites the 2002 Esmaeli
Arch. Ophthal. Study. Second, he cites deposition testimony of Sanofi’s Chief
Medical Officer Dr. Barry Childs.
1. 	Reliance on 2002 Esmaeli Arch. Ophthal. Study
Dr. Thundiyil concludes that the 2002 Arch. Ophthal. Study “provides
not 	only 	proof 	of 	exposure 	through 	tears, 	but 	also 	a 	dose 	dependent
relationship.”204 Sanofi argues that the 2002 Arch. Ophthal. Study did not
actually determine that patients had stenosis but rather only measured the
presence 	of 	docetaxel 	in 	tears. 	It 	further 	contends 	that 	Dr. 	Thundiyil
overlooked conflicting data in the study—i.e., that it did not find consistently
higher concentrations of docetaxel in both the plasma and tears of specific
patient groups (weekly administration of smaller doses v. every-three-week
administration of higher doses). Sanofi further argues that when pressed, Dr.
Thundiyil acknowledged that he could not identify a dose-response correlation
in the 2002 Esmaeli Arch. Ophthal. Study, pointing to the following testimony:
Q: But the takeaway here is that the higher dose results in a
higher presence in the tears; that’s the takeaway from this study?
A. Yes -- well, that’s one the of takeaways. I mean, it is important
to note, though, that, you know, even in patient 1, who had the
lower dose, the 35 with the total of 58, her concentration in tears
at that time was still not that much different than someone who
had the higher 	dose. So 	there’s 	not 	a -- necessarily 	a 	direct
correlation. There’s a trend for a dose response, but it – there’s
probably some other patient factor that we don’t yet understand
204 Doc. 446-3 at 9–10.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 47 of 67

48
that might cause some people to secrete it more in tears than
others.205
The Court finds that Dr. Thundiyil’s reliance on the 2002 Esmaeli Arch.
Ophthal. Study does not render his opinion unreliable. First, Dr. Thundiyil did
not “acknowledge” that there was no dose-response relationship. Rather, he
indicated that there was at least a trend. Second, although the authors did not
examine patients for the presence of stenosis, the authors indicated that their
findings 	of 	docetaxel 	in 	tears 	confirmed 	their 	hypothesis 	regarding 	the
mechanism of injury for docetaxel-induced stenosis. Indeed, the authors stated
that “[w]e have previously reported that stenosis of the canaliculi and blockage
of the nasolacrimal ducts are common adverse effects of weekly administration
of docetaxel.”206 They further explained that “in this study, we sought to test
the hypothesis that docetaxel may be secreted in tears. The secretion of
docetaxel 	in 	tears 	would 	suggest 	that 	closure 	of 	the 	lacrimal 	drainage
apparatus may be due in part to its direct contact with the drug.”207 Thus, the
2002 Arch. Ophthal. Study is sufficiently related to stenosis. Third, when
asked 	if 	there 	was 	“any 	way 	to 	explain 	[the 	plasma] 	discrepancy,” 	Dr.
Thundiyil stated that “[i]n something of this small sample size, I don’t know
why that is, but it illustrates a couple of things: That not -- the plasma
concentration may not directly be related to the dose or the tear concentration.
Meaning, that how much you have in your plasma may not in everybody be
directly related to how much is secreted in the tears.”208
205 Doc. 446-4 at 44.
206 Doc. 446-2.
207 Id.
208 Doc. 446-4 at 44.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 48 of 67

49
Fourth, like Dr. McGwin, Dr. Thundiyil relies on additional studies that
support the dose-response criterion. Although he does not directly cite them in
the “dose-response” section of his report, Dr. Thundiyil relies on several studies
that 	indicate 	that 	higher 	doses 	or 	cumulative 	exposures 	correlate 	with
increased incidence or severity. For example, like Dr. McGwin, Dr. Thundiyil
cites the 2002 Esmaeli Annals of Oncology Study and the 2002 Esmaeli
Ophthalmology Study.209
2. 	Reliance on deposition testimony
Finally, 	Sanofi 	argues 	that 	Dr. 	Thundiyil 	impermissibly 	relied 	on
deposition testimony of Dr. Barry Childs, Sanofi’s Chief Medical Officer. In his
report, Dr. Thundiyil cites to Dr. Childs deposition testimony that “it was
known that docetaxel was secreted in tears.”210 Dr. Thundiyil also cites Dr.
Childs’s testimony describing an exhibit he prepared, which states that (1)
“Taxotere secretion in tears may contribute to the canalicular stenosis”; (2) in
patients with hyperlacrimation, this injury “occurs most often with weekly
dosing”; 	and (3) “incidence appears to increase with cumulative dose.”211
Sanofi points out that Dr. Thundiyil admitted that he does not usually rely
209 He also relies on a study published in 2006 by Dr. Esmaeli and others in the Journal
of Clinical Oncology (“2006 Esmaeli Study”). There, the authors reported that of the twenty-
eight patients taking weekly docetaxel, “[n]ine patients (32%) had worsening of canalicular
stenosis despite tobramycin and dexamethasone (to grade 2 in eight patients and to grade 3
in one patient), so surgery was recommended.” Doc. 462-5 at 168. Of the twenty-eight patients
taking docetaxel every two or three weeks, “[t]wo patients (7.1%) had persistent epiphora and
grade 1 canalicular stenosis and underwent bilateral DCR with temporary silicone tube
placement.” Id. at 169. Plaintiffs point out that patients requiring surgery for epiphora had
a median cumulative docetaxel dose of 889.5 mg, whereas those who developed epiphora (but
did not necessarily require surgery in this study) had median doses of 496.5 mg (weekly
group) and 420 mg (every-three-week group).
210 Doc. 446-3 at 9.
211 Id. at 9–10.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 49 of 67

50
upon company witness testimony in considering the Bradford Hill factors. It
further avers that Dr. Thundiyil overlooked the dose limitations of the cited
statements. Sanofi points out that Dr. Childs states that “[t]he three weeks on
and one week off is consistent with conventional dosing and helps avoid some
of the chronic toxicities like . . . hyperlacrimation . . . .”212
Here, Dr. Thundiyil does not treat Dr. Child’s statements as conclusive
proof 	of 	a 	dose-response 	relationship. 	Rather, 	cites 	them 	as 	additional,
supportive evidence. Dr. Thundiyil also includes the statements regarding
every-three-week administration in his report. Although Dr. Thundiyil does
not separate out his analysis by weekly and every-three-week administration,
again, this is not required to satisfy the dose-response factor. While Sanofi
takes issue with Dr. Thundiyil’s “failure to consider differences in dosing,” as
well as alleged discrepancies in the studies upon which he relies, here, those
issues go to the weight of his testimony, and Sanofi may address them on
cross-examination or by the presentation of contradictory evidence.
C. 	Plausibility
Sanofi 	also 	argues 	that 	Dr. 	Thundiyil 	failed 	to 	reliably 	apply 	the
plausibility factor because he failed to cite reliable studies in support of his
opinions. In his report, Dr. Thundiyil begins by explaining, as background, how
docetaxel works at the cellular level. He explains that docetaxel works by
inhibiting the cell division of cancer cells. When cells cannot replicate, this
leads to cell death, and he notes that docetaxel may also influence proteins
involved in apoptosis (programmed cell death). He then states that these same
212 Doc. 446-4 at 10.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 50 of 67

51
“cytotoxic” (cell-killing) properties that allow docetaxel to destroy cancer cells
can also affect other cells in the body, particularly rapidly dividing cells, and
can result in a range of adverse effects.
He then turns, under the plausibility factor, to the mechanisms of injury.
He begins by describing the body’s general response to tissue damage: injury
leads to the release of cytokines and immune cells, causing inflammation,
along with angiogenesis, extracellular matrix deposition, and the formation of
new tissue. He explains that fibroblasts produce extracellular matrix to repair
damage, but that excess matrix formation can result in stiffening and scarring,
particularly in tissues with limited regenerative capacity.
After describing this injury-response process, he opines that “[a]lthough
the exact mechanism of docetaxel induced canalicular stenosis is not known,”
there are several plausible mechanisms which likely play a factor in the
development of this pathology.”213 He then identifies two specific mechanisms
related to docetaxel. First, he opines that one mechanism is edema, explaining
that 	docetaxel 	increases 	vascular 	permeability 	and 	causes 	fluid 	leakage.
Second, like Dr. McGwin, Dr. Thundiyil notes that in the 2003 Esmaeli
Ophthalmic Study, biopsies revealed extensive fibrotic changes in the nasal
mucosa and lacrimal sac. He also explains that lymphocytic infiltrates and
chronic fibrosis of the stroma were found in the patients in this study. He
ultimately concludes that “[i]t is plausible and very clear that exposure to
docetaxel in the eye can cause inflammation and trauma to the eye. Due to the
unique embryologic origin of the canaliculi, the damaged cells that line it
213 Doc. 446-3 at 8.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 51 of 67

52
undergo 	scarring 	as 	confirmed 	by 	biopsy. 	This 	scarring 	is 	the 	plausible
mechanism for stenosis.”214
Sanofi argues that Dr. Thundiyil’s biological plausibility opinions are
speculative 	because 	he 	concedes 	that 	the 	precise 	mechanism 	by 	which
docetaxel purportedly causes tear-duct stenosis is unknown and admits there
is no evidence linking apoptosis to lacrimal duct injury. Plaintiffs respond that
Dr. Thundiyil’s opinions are reliable and that he reviewed docetaxel’s known
cytotoxic effects, relevant anatomy, and study findings to propose plausible
mechanisms like increased edema and scarring. They further argue that a
precise mechanism is not required under Daubert.
As to his apoptosis theory, Dr. Thundiyil testified as follows:
Q. Okay. And is that meant to say that the cell killing properties
of docetaxel, that allow it to destroy cancer cells, are the same
properties that are leading to stenosis of the lacrimal duct system?
A: All I’m really trying to say here is there’s a plausible mechanism
for cell damage, because we know it kills cancer cells, we know the
mechanism is to inhibit mitosis, which causes apoptosis. What my
point 	is 	here 	is 	that, 	in 	addition 	to 	the 	desired 	effects 	that
chemotherapeutic drugs give, which is the killing of cancer cells,
there are also some of the undesired effects, and I list some of the
symptoms here. And I am explaining why it’s plausible that this
can also cause damage to certain parts of the eye.
Q. Okay. And, precisely, how does apoptosis lead to damage to the
lacrimal duct system?
A. In this case, I’m not aware of apoptosis and lacrimal duct
damage, or at least it’s not been specifically shown.215
214 Id. at 9.
215 Doc. 446-4 at 34.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 52 of 67

53
As explained above, the RMSE and Hill recognize that plausibility depends on
the “biological knowledge of the day,” so courts cannot demand mechanistic
certainty.216 Courts also recognize that “biological plausibility” is but one
criterion in the Bradford Hill analysis; thus, to the extent an expert is unable
to articulate a biologically plausible mechanism at all, “that is an issue that
goes to the weight of their opinion, and not the issue of admissibility.”217
Here, the Court finds that the absence of a fully elucidated mechanism
does not render Dr. Thundiyil’s opinion inadmissible. While the lack of support
for the connection between docetaxel, apoptosis, and tear duct damage may
create fodder for cross-examination, it does not render it unreliable. Dr.
Thundiyil mentions apoptosis as representative of one mechanism through
which a cytotoxic agent such as docetaxel may injure cells.
Next, Sanofi contends that the 2003 Esmaeli Ophthalmic Study does not
support Dr. Thundiyil’s opinion because in the biopsies taken, the authors
found notable scarring in the nasal mucosa and lacrimal sac, which are
different body tissues from the “punctal, canalicular, and/or nasolacrimal duct”
targeted by Dr. Thundiyil’s opinions. In his deposition, Dr. Thundiyil did agree
that the study was a three-case report, that only one patient had a lacrimal-
sac specimen, that two patients had only nasal-mucosa specimens, and that
none of the patients had biopsy tissue from the canaliculi or punctum. He also
216 See Doc. 488 at 25 (citing 2011 RMSE, supra note 45, at 605); see also Austin
Bradford Hill, The Environment and Disease: Association or Causation, 58 PROC. ROYAL
S OC’ Y M ED. 295, 295 (1965)).
217 Wagoner, 813 F. Supp. 2d at 804 (citing In re Viagra Prods. Liab. Litig., 658 F.
Supp. 2d 936, 946 (D. Minn. 2009)).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 53 of 67

54
agreed that the lacrimal sac is a different structure from the canaliculi and
punctum. Even so, his reliance on the 2003 Esmaeli Ophthalmic Study does
not render his opinions unreliable.
Indeed, 	the 	authors 	were 	examining 	patients 	who 	had 	canalicular
narrowing and nasolacrimal duct obstruction, and they interpreted the fibrosis
they observed in the lacrimal sac and nasal mucosa as evidence that docetaxel
causes fibrotic changes in the lacrimal drainage apparatus more broadly.
Indeed, 	they 	explicitly 	stated 	that 	the 	drug 	probably 	affects 	the 	“entire
lacrimal drainage apparatus,” while acknowledging the narrower scope of the
tissue sampled.218 Accordingly, the 2003 Esmaeli Ophthalmic Study provides
support for Dr. Thundiyil’s conclusions. In his report, Dr. Thundiyil uses it as
support for his 	broader biological-plausibility 	chain: docetaxel can injure
ocular tissues, injury can trigger inflammation and fibroblast-driven scarring,
and scarring can obstruct the lacrimal drainage system.
As additional support, Dr. Thundiyil also cites a study published by
Abdul Hassan and others in 1998 titled “Epiphora in 	patients receiving
systemic 5-flourouracil therapy (“5-FU”) (“1998 Hassan Study”). He contends
that the 1998 Hassan Study “revealed a dose dependent relationship between
5-Fluorouracil and epiphora,” that “[t]his study demonstrates a precedent in
an analogous drug which have a similar effect to docetaxel.”219 Sanofi contends
that there is a “fit” problem because the 1998 Hassan Study focused on a
different drug. They point out that when pressed about the mechanisms of
action resulting in scarring and inflammation for both drugs, Dr. Thundiyil
218 Id. at 4.
219 Doc. 446-3 at 8.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 54 of 67

55
admitted that “their mechanisms [of action] are different,” with one drug
affecting 	mRNA 	and 	DNA 	replicability 	and 	the 	other 	damaging 	cells’
microtubules.220 	Sanofi 	also 	argues 	that 	Dr. 	Thundiyil 	“ignored 	contrary
evidence” and cherry-picked favorable data because he summarily rejected the
1998 Hassan article’s conclusions that fibrosis might be related to excessive
eye-rubbing.
The Court disagrees. In his deposition, Dr. Thundiyil explained why he
cited the 1998 Hassan Study: “With the limited understanding that exists for
the mechanism, the important thing that my opinion carries is the cause. The
proposed mechanism, that’s--as I mentioned, is analogous in other drugs, or
other settings, is this common theme of scarring that causes stenosis.”221 As
for the authors’ conclusions regarding “punctal excoriation,” or eye rubbing,
Dr. Thundiyil testified that he did not believe that the study proved that
conclusion.
To 	be 	sure, 	the 	1998 	Hassan 	Study 	does 	not 	establish 	that
chemotherapy-induced 	fibrosis 	is 	the 	definitive 	mechanism 	of 	tearing.
Nevertheless, like the 2003 Esmaeli Ophthalmic Study, it provides some
support for an association between treatment, canalicular fibrosis, and tear-
duct dysfunction, rendering Dr. Thundiyil’s reliance on it a matter of weight
rather than admissibility. Here again, Sanofi may explore these issues on
cross-examination.
D. 	Failure to Identify Background Risk of Stenosis
220 Doc. 446-1 at 18.
221 Doc. 446-4 at 62.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 55 of 67

56
Finally, as with Dr. McGwin, Sanofi argues that Dr. Thundiyil failed to
consider the “background rate” for lacrimal duct stenosis. This too must be
rejected. Again, Sanofi does not identify any evidence of a “background rate”
that Dr. Thunidyil ignores. Rather, when pressed, Dr. Thundiyil testified as
follows:
Q: Okay. And the evaluation of your assignment in this case, did
you make any effort to identify the background rate for lacrimal
duct stenosis in the general population?
A. I tried to do that. It's not--there is not--I couldn’t find a good
answer to that, because there’s wide varieties, or wide intervals
like this, and there are a variety of biases, which go into that--the
biggest thing is either a referral bias, or lot of people -- many don’t
go seek care for minor causes of tearing. And then, in some cases,
not all ophthalmologists do some of the gold standard procedures,
which would be probing and irrigation, to identify it. So there
didn’t 	seem 	to 	be 	a 	uniform 	method 	for 	evaluating 	an[sic]
diagnosing it, so it was hard to get a background rate for this.222
Thus, in this case, Dr. Thundiyil contends that while he attempted to do such
an analysis, given the limited body of evidence, he could not do so. Sanofi
nevertheless contends that Dr. Thundiyil “admitted there were international
studies and comparator populations, but he did not consider them.”223 When
asked about these studies, however, Dr. Thundiyil gave a reasoned explanation
as to why each one was not sufficient to give the prevalence or incidence of
stenosis in the general population, such as biases.224 He ultimately testified
that “I have not seen an accurate study that reflects what the baseline
222 Id. at 52 (emphasis added).
223 Doc. 508 at 10.
224 Doc. 446-4 at 52–54.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 56 of 67

57
prevalence, or incidence, of canalicular stenosis is.”225 As such, Dr. Thundiyil
did not simply “disregard” these studies, as Sanofi contends. Rather, as was
the case with Dr. McGwin, the record evidence indicates that Dr. Thundiyil
“identified and accounted for the limited body of scientific literature currently
available on background risk for” stenosis.226 In so doing, he “satisfied Rule
702 and Daubert.”227
In sum, after reviewing Dr. Thundiyil’s report, deposition testimony, and
the arguments presented by the parties, the Court concludes that Sanofi’s
objections go to the weight of Dr. Thundiyil’s opinions, not their admissibility.
Accordingly, Sanofi’s Motion to Exclude Dr. Thundiyil is DENIED.
III. 	MOTION TO EXCLUDE DR. DURAIRAJ
Plaintiffs proffered Dr. Durairaj as a medical expert who specializes in
treating the injury at issue. Dr. Durairaj is a board-certified ophthalmologist
who is fellowship trained in oculoplastic and orbital surgery. He completed his
fellowship at the Mayo Clinic, served as a professor at the University of
Colorado and the University of Texas at Austin’s Dell Medical School, and is
currently the managing partner of an oculoplastic clinic. Dr. Durairaj opines
that, based on his review of the medical literature and his experience with the
patients he has treated, Taxotere/docetaxel can cause permanent stenosis to
the nasolacrimal drainage system. Sanofi moves to exclude Dr. Durairaj’s
general causation opinions, as well as his opinions related to label adequacy
225 Id. at 54.
226 In re Abilify, 299 F. Supp. 3d at 1332.
227 See id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 57 of 67

58
and warnings, arguing that he is not qualified and that he did not apply a
reliable methodology. The Court will consider each argument in turn.
A. 	General Causation Opinions
Sanofi argues that Dr. Durairaj is not qualified to offer general causation
opinions because he (1) is not epidemiologist, toxicologist, or pharmacologist;
(2) has never published anything about Taxotere and stenosis; and (3) has
never conducted a study to determine whether Taxotere is capable of causing
stenosis nor otherwise assessed whether any drug is capable of causing a side
effect. Sanofi further contends that Dr. Durairaj’s methodology is not reliable
because he did not invoke the epidemiological or toxicological methodologies
described above, and that his opinions, which are based only on his review of
the medical literature as well as his own personal experience, are not a reliable
substitute. Plaintiffs counter that Dr. Durairaj is qualified based on his clinical
experience. They point out that in 2002, Dr. Durairaj gave a presentation
regarding canalicular stenosis as a complication of Taxotere to the Department
of Oncology at the University of Colorado. According to Plaintiffs, this makes
Dr. Durairaj “one of the earliest surgeons in the world familiar with this
specific injury.”228 They further contend that Dr. Durairaj’s opinions are
reliable because they are based on his clinical experience and a general
consensus within his field.
Although usually in the context of specific causation, the Fifth Circuit
“has upheld the admission of expert testimony where it was based on the
expert’s 	specialized 	knowledge, 	training, 	experience, 	and 	first-hand
228 Doc. 482 at 10.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 58 of 67

59
observation while supported by solid evidence in the scientific community.”229
Here, however, Dr. Durairaj admitted his limitations:
Q: Absolutely. So as far as any sort of causation opinion regarding
Taxotere 	and 	stenosis 	or 	dosing 	issues 	or 	infusion 	schedules
involving 	Taxotere 	and 	stenosis, 	would 	you 	defer 	to 	the
epidemiologist and toxicologist listed in this litigation?
A: I’m not sure how to answer that question. I would just say that’s
outside of my area of expertise.
Q: Okay. So any causation opinion would be outside your area of
expertise?
A: In saying “causation,” I don’t – I mean I think that’s a difficult
question to ask, and the reason why is – I’ll give you an example,
like if somebody got hit by a baseball and it caused an orbital
fracture, part of my area of expertise is I would say getting hit by
a baseball caused an orbital fracture. So yeah.
Q: Well, maybe I can ask you this. Would you agree that you're not
an expert in determining causation?
A: I mean, I think it depends on the situation.
Dr. Durairaj also testified that he does not know whether an association differs
from causation and acknowledged that he has never assessed whether any
drug can cause a particular side effect. He testified that once he has diagnosed
stenosis, he “assumes it is probably Taxotere,” based upon a consensus within
this field.230
This Court need not decide whether Dr. Durairaj is qualified, however,
because he failed to apply a reliable methodology. Although epidemiological
229 Pipitone v. Biomatrix, Inc., 288 F.3d 239, 247 (5th Cir. 2002) (citing Skidmore v.
Precision Printing and Packaging, Inc., 188 F.3d 606, 617 (5th Cir. 1999)).
230 Doc. 455-3 at 76.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 59 of 67

60
evidence is not strictly required, he must still employ a reliable method of
analyzing the available evidence. The literature relied upon by Dr. Durairaj
admittedly 	describes 	epiphora 	associated 	with 	docetaxel 	and 	identifies
canalicular stenosis as a proposed mechanism. But Dr. Durairaj does not
analyze or synthesize that literature to establish general causation. He does
not evaluate the underlying studies, assess their strengths or limitations, or
explain 	how 	they 	support 	a 	causal 	conclusion. 	Instead, 	he 	adopts 	the
conclusions of the literature without demonstrating that they are sufficient to
establish causation. His opinion therefore amounts to an unsupported ipse
dixit.
Accordingly, the Court finds that Dr. Durairaj’s ultimate opinion on
general causation is inadmissible. Dr. Durairaj, however, may offer other
relevant supporting opinions. For example, he may testify about the nature
and treatment of stenosis, as well as describe and/or critique the findings
reported in the literature and his field. He may also testify as to the mechanism
of injury. Based on his clinical experience, he opines that the “unique cellular
structure of the drainage system is susceptible to damage” and that “scarring
and/or fibrosis can progress until pathway[s] closes and predominantly affects
the punctal and canalicular systems, and sometimes the nasolacrimal duct.”231
He supports this opinion by citing literature, which he contends indicates that
“the mechanism of injury occurs through secretion of the drug in tears causing
inflammation, 	scarring, 	and/or 	fibrosis 	in 	the 	nasolacrimal 	drainage
231 Id.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 60 of 67

61
system.”232 In sum, while he may testify about these issues, Dr. Durairaj may
not offer an opinion on general causation.
B. 	Regulatory and Labeling Opinions
Sanofi moved to exclude Dr. Durairaj “from offering any opinions on the
adequacy of Taxotere’s label or how prescribing oncologists would interpret it.”
Specifically, it seeks to preclude the following opinions:
1. 	The 	use 	of 	the 	term 	“lacrimal 	duct 	obstruction” 	in 	the
Taxotere 	label 	is 	“imprecis[e],” 	“unhelpful 	to 	treating
physicians,” and has “no place in a prescription drug label.”
2. 	The Taxotere label should direct oncologists to refer patients
with epiphora for a prompt ophthalmologic examination.
3. 	The Taxotere label should warn that stenosis is potentially
permanent, and “many oncologists do not fully appreciate
that Taxotere / Docetaxel-related epiphora is not a minor
transitory 	inconvenience 	that 	will 	dissipate 	at 	the
completion of chemotherapy[.]”
4. 	The “need for an improved warning in the [Taxotere] label
was heightened by the publication” of the 2013 Chan Study.
5. 	Based on the foregoing, the Taxotere label and the 2013
Chan 	Study 	created 	confusion 	within 	the 	oncology
community about the nature and severity of the injury.
Sanofi seeks to exclude these opinions on the basis that (1) Dr. Durairaj is not
qualified; (2) he has no reliable basis for these opinions and/or that they are
mere speculation; and (3) his opinions do not fit the facts of the bellwether
cases. As a threshold matter, Sanofi points out that Dr. Durairaj conceded that
232 Doc. 455-2 at 9.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 61 of 67

62
he does not hold himself out to be an expert in labeling, appropriate labeling,
or regulatory matters. Plaintiffs respond that Dr. Durairaj is not offering
opinions on whether the label meets regulatory requirements but rather offers
fact-based observations about the label’s content and associated side effects he
has personally treated.
An expert may be qualified to testify based on the expert’s “knowledge,
skill, experience, training, or education.”233 Generally, “qualification is viewed
liberally and is determined by comparing the area in which the witness has
superior knowledge, skill, experience, or education with the subject matter of
the witness’s testimony.”234 An expert who is qualified may “draw a conclusion
from a set of observations based on extensive and specialized experience.”235 If,
however, “the witness is relying solely or primarily on experience,” he or she
“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.”236 The Court now considers each of Dr. Durairaj’s
opinions in turn.
1. 	Use of the term “lacrimal duct obstruction”
Dr. 	Durairaj 	opines 	that 	the 	term 	“lacrimal 	duct 	obstruction” 	is
“imprecise,” “unhelpful,” and has “no place in a prescription drug label.”237
233 FED. R. EVID. 702.
234 In re Fosamax, 645 F. Supp. 2d 164, 190 (S.D. N.Y. 2009) (citing In re Methyl
Tertiary Butyl Ether (“MTBE”) Prods. Liab. Litig., Nos. 1:00-1898, M21-88, MDL No. 1358,
2008 WL 1971538, at *6 (S.D.N.Y. May 7, 2008)).
235 Id. (citing In re MTBE, 2008 WL 1971538, at *6).
236 Andrews v. Rosewood Hotels & Resorts, LLC, 575 F. Supp. 3d 728, 734 (N.D. Tex.
2021) (citing FED. R. EVID. 702 advisory committee’s note to 2000 amendment).
237 Doc. 455-2 at 6.
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 62 of 67

63
Courts in this jurisdiction have held that “doctors are fully qualified to opine
on the medical facts and science regarding the risks and benefits” of drugs, and
to “compare that knowledge with what was provided in the text of labeling and
warnings for FDA approved drugs.”238 However, courts draw an important
distinction 	between 	permissible 	medical 	testimony 	and 	impermissible
regulatory opinions. While a physician may testify about medical risks and
whether those risks are reflected in the label, he “must possess additional
expertise” to opine on what information should or should not be included in a
drug label.239 In another MDL, this Court allowed an oncologist and expert
treating physician to testify about Taxotere’s label but precluded testimony
“about whether the label complied with the FDA regulations,” which fell
outside of his expertise.240
Here, to the extent Dr. Durairaj explains the medical meaning of the
term “lacrimal duct obstruction” and whether it accurately describes the
condition at issue, such testimony is admissible. As a clinician, he may also
testify how, from a medical perspective, the terminology would be understood
by physicians. Dr. Durairaj’s opinion, however, that this language has “no
place in a prescription drug label” constitutes an impermissible opinion about
what language should or should not be included in the label, which falls outside
his expertise.
238 Conn v. C.R. Bard, Inc., No. 4:14-CV-298, 2021 WL 2346036, at *3 (S.D. Tex. June
8, 2021) (citation modified).
239 Id. (citation modified).
240 In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 16-2740, 16-17039, 2021
WL 111778, at *3 (E.D. La. Jan. 12, 2021).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 63 of 67

64
2. 	Label should direct referral to an ophthalmologist
and 	warn 	of 	permanence; 	oncologists 	do 	not
appreciate severity
Dr. Durairaj states in his report that prompt referral is necessary to
prevent the progression of stenosis and that it is rare that patients who he
diagnoses 	with 	epiphora 	secondary 	to 	docetaxel 	are 	referred 	by 	their
oncologists. He opines that the fact that patients with this side effect have
continued to consistently arrive in oculoplastic surgeons’ offices with advanced
cases of stenosis is evidence itself of the fact that many oncologists are not
aware of this side effect. Finally, he contends that it is crucial that the Taxotere
label include a clear warning to oncologists alerting them to the fact that this
potentially permanent side effect can be prevented.
Dr. Durairaj may testify, based on his clinical experience, about the
general need for prompt referral and the risk of permanent stenosis. He may
also testify about patients who frequently present with advanced stenosis.
However, 	his 	opinion 	that 	these 	observations 	demonstrate 	that 	“many
oncologists are not aware” of the side effect is inadmissible. This conclusion
constitutes an unsupported and speculative inference regarding the knowledge
of other physicians and is not grounded in a reliable methodology.241 Finally,
again, he may not opine on what the label should have warned about, i.e., the
risk of permanence or the need for prompt referral, as this falls outside of this
expertise.
241 See In re Fosamax, 645 F. Supp. 2d at 189 (holding inadmissible the expert’s
testimony about the manufacturer’s “state of mind, knowledge, or intent”).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 64 of 67

65
3. 	The 	2013 	Chan 	Study 	heightened 	the 	need 	for 	an
improved warning
Dr. Durairaj opines that publication of the 2013 Chan Study heightened
the need for an improved warning. This opinion constitutes a regulatory
judgment regarding when a label should be modified. Again, such testimony
requires 	expertise 	in 	drug 	labeling 	and 	regulatory 	standards, 	which 	Dr.
Durairaj does not possess.242 He also opines that the Taxotere label and the
Chan Study created confusion within the oncology community. This opinion is
an unsupported assertion about the collective understanding of oncologists and
is not grounded in any reliable methodology. Courts routinely exclude such
broad, speculative claims.243 He may, however, testify regarding the medical
findings of the 2013 Chan Study, their significance to clinicians, and offer
critiques. He just may not testify that the 2013 Chan Study created a duty or
“need” to change the label.
4. 	Sanofi’s argument regarding treating physicians
Finally, Sanofi argues that under this Court’s prior rulings, testimony
regarding how a warning would have affected a physician’s decision must come
from the treating physician, not a retained expert. In failure-to-warn cases,
causation turns on 	whether 	a different warning would 	have altered 	the
prescribing physician’s decision.244 That argument is correct as to case-specific
causation. This Court has previously held that when determining whether a
242 See C.R. Bard., 2021 WL 2346036, at *3.
243 See In re Fosamax, 645 F. Supp. 2d at 197–98 (finding that medical expert’s
testimony that “osteoporosis is ‘controversial’ or that it is not really a disease was “mere
personal opinion going beyond his expertise and, as such, would not be helpful to the jury”).
244 See Willett v. Baxter Int’l, Inc., 929 F.2d 1094, 1098–99 (5th Cir. 1991).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 65 of 67

66
different warning would have altered a prescribing decision, the appropriate
witness 	is 	the 	treating 	physician, 	particularly 	where 	that 	physician 	is
available to testify.245
Plaintiffs, however, represent that Dr. Durairaj does not intend to offer
opinions 	regarding 	how 	any 	bellwether 	plaintiff’s 	oncologist 	would 	have
interpreted the label or altered treatment decisions. To the extent Dr. Durairaj
limits his testimony to general medical principles—such as the nature of the
risk, its severity, and the importance of timely treatment—such testimony is
permissible. 	Dr. 	Durairaj 	may 	not, 	however, 	offer 	case-specific 	opinions
regarding (1) what a prescribing oncologist would have done; (2) how any
oncologist would have interpreted the label; or (3) whether a different warning
would have changed treatment decisions. Those opinions are reserved for the
treating physicians. Accordingly, Sanofi’s Motion to Exclude Dr. Durairaj is
GRANTED IN PART and DENIED IN PART, as outlined herein.
IV. 	MOTION 	FOR 	SUMMARY 	JUDGMENT 	ON 	GENERAL
CAUSATION
As explained above, Sanofi moves for summary judgment solely on the
basis that Plaintiffs lack admissible expert testimony. Having found that
Plaintiffs have presented admissible expert testimony on general causation,
Sanofi’s Motion for Summary Judgment must be DENIED.
245 In re Taxotere (Docetaxel) Prods. Liab. Litig., MDL No. 16-2740, No. 16-17039,
2021 WL 321661, at *3 (E.D. La. Feb. 1, 2021).
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 66 of 67

67
CONCLUSION
For the foregoing reasons,
IT IS ORDERED that Sanofi’s Motion to Exclude Opinions of Dr.
McGwin (Doc. 462) is DENIED.
IT IS FURTHER ORDERED that Sanofi’s Motion to Exclude General-
Causation Opinions of Dr. Thundiyil (Doc. 446) is DENIED.
IT IS FURTHER ORDERED that Sanofi’s Motion to Exclude Certain
Testimony of Dr. Vikram Durairaj (Doc. 455) is GRANTED IN PART and
DENIED IN PART. Dr. Durairaj may offer medical testimony regarding the
risks, severity, and treatment of stenosis and may compare those risks to what
is disclosed in the Taxotere label. He may not offer opinions on the ultimate
issues of general causation, label adequacy, regulatory requirements, or how
prescribing physicians would have interpreted or acted upon the label.
IT 	IS 	FINALLY 	ORDERED 	that 	Sanofi’s 	Motion 	for 	Summary
Judgment on General Causation (Doc. 450) is DENIED.
New Orleans, Louisiana this 2nd day of April, 2026.
____________________________________
JANE TRICHE MILAZZO
UNITED STATES DISTRICT JUDGE
Case 2:22-md-03023-JTM-MBN Document 619 Filed 04/02/26 Page 67 of 67