Amicus Curiae Brief — International Union, United Automobile, Aerospace & Agricultural Implement v. Johnson Controls, Inc.

Supreme Court, US |

\

: Whe FILED

sy, i} 1590

No. 89-1215 { wei

‘ Beit FINA) jr

CTERA

IN THE .

Supreme Court of the United States

OCTOBER TERM, 1989

INTERNATIONAL UNION, UNITED AUTOMOBILE,

AEROSPACE AND AGRICULTURAL IMPLEMENT

WORKERS OF AMERICA, ef a/.,

Petitioners,

Ve

JOHNSON CONTROLS, INC.,

Respondent.

On Writ of Certiorari to the

United States Court of Appeals

for the Seventh Circuit

BRIEF OF

NATURAL RESOURCES DEFENSE COUNCIL, INC.

AS AMICUS CURIAE IN SUPPORT OF PETITIONERS

THOMAS O. MCGARITY

Counsel of Record

University of Texas

School of Law

727 E. 26th Street

Austin, Texas 78705

(512) 471-5151

ALBERT H. MEYERHOFF

90 New Montgomery

San Francisco, CA 94105

Attorneys for Amicus Curiae

A TT SL DESEO SSL ICES TENET - 5 EY. A TEES S20, BT TILER RIE I ELI OL GTS

WILSON - EPES PRINTING Co., INC. - 789-0096 - WASHINGTON, D.C. 20001

® ot » 60

TABLE OF CONTENTS

Page

rs Cr ee scnonheumnbnahiueseseneess 1

be 8 PY 4 ale ae 8 6 5) | 4. y er 2

ARGUMENT:

THE COURT OF APPEALS ERRED IN CON-

CLUDING THAT SCIENTIFIC DATA FROM

EXPERIMENTS WITH LABORATORY ANI-

MALS ARE TOO SPECULATIVE TO SUPPORT

CONCLUSIONS ABOUT THE REPRODUCTIVE

TOXICITY OF CHEMICALS IN HUMAN

INIT -Uccandpbarentascanpinndtaccumicackcareiiartaestelenactnseses eleemeebiin 3

A. An Established Body of Scienti*c Opinion Sup-

ports the Use of Laboratory Animal Studies in

Assessing the Reproductive Toxicity of Chem-

ee es ee a eeeeeugunesonbenshens dD

B. Courts and Administrative Agencies Have Con-

sistently Relied Upon Laboratory Anima! Strd-

ies in Assessing the Risks of Chemicals to Hu-

rete easegcteunmesinsenounsoniomn SE 12

CERN AIIIU OD vcecctsassecesssersecncese CA ee OLE IS,2- e 20

il

TABLE OF AUTHORITIES

CASES: Page

B. 1! v. Goddard, 366 F.2d 177 (7th Cir. 1966) ...... 16

Environmental De fense Fund, Ine. v. Costle, 578

F.2d 3387 (D.C. Cie. 1970) 00 15

Environmental Defense Fund, Inc. v. EPA, 510

F.2d 1292 (D.C. Cie. 1076) cee 15

Environmental Defense Fund, Inc. v. EPA, 548

F.2d 996 (D.C. Civ. BOGG) ..ccscocnceeee 15

Environmental Defense Fund, Inc. v. EPA, 598

F.2d 62 (D.C. Cit. BOGB) ..cccscee 15

Environmental Defense Fund, Ine. v. Ruckelshaus,

439 F.2d 684 (D.C. Cis. 1971) We 15

Herevles, Inc. v. EPA, 598 F.2d 91 (D.C. Cir.

Industrial Union De pt. AFL-CIO v. American

Petroleum Inst., 448 U.S. 607 (1980) .........0000. 14

International Union v. Johnson Controls, Inc., 886

F.2d 871 (7th Cir. 1968) 0 4, 7,16

Society of Plastics Industry v. OSHA, 509 F.2d

1301 (D.C. Civ, 1GGG) ..2c.ceccccccsceuceuene 15

Synthetic Organie Chem. Mfrs. Assn. v. Brennan,

003 F.2d 1155 (3rd Cir. 1974) ....22.20000 15

STATUTES AND REGULATIONS:

16 C.F.R. § 1500.17 (a) (10) (1989) 12

29 C.F.R. § 1900.111 (a) (1980) ee 14

29 C.F.R. § 1990.1438(g¢) (1969) ................................ 14

40 C.F.R. § 61, Subpart F (1989) nana 13

40 C.F.R. § 154.7(a) (2) (i) (1989) 12

40 C.F.R. § 261.11(a) (1968) 14

40 C.F.R. § 786 (1968) ........... eee 13

45 C.F.R. § 46 (1969) ...............0.. eee 5

$8 Fed. Reg. 108029 (1978) ........- 2... 14

39 Fed. Reg. 35890 (1974) 2.2000. -apsiaeetenaaa 14

40 Fed. Reg. 6 69532 Ri: | RAP 13

41 Fed. Reg. 30476-77 (1976) a 13

42 Fed. Reg. 6532 (1977) ............... 13

43 Fed. Reg. 11514 (1978) 0... ieee 16

43 Fed. Reg. 12308-12310 (1978) .... 12

43 Fed. Reg. 52952 (1978) 17

ili

TABLE OF AUTHORITIES—Continued

i oo. cnsnevcsseuscnccserenscsesssccees

EE ND oo. cereccsesesnsnecsenseceseoreess- wu

i ee ccc cnesccessancseeecaccessesoses

53 Fed. Reg. 24834 (1984) ................................

53 Fed. Reg. 24850 (1988) 0.0 2.

54 Fed. Reg. 9386 (1989)... 7

7 U.S.C. § 136-1387y (1988)

15 U.S.C. $§ 1261, 1262 (1988)

15 U.S.C. § 2079 (a) (1988)

15 U.S.C. $$ 2601-2629 (1988)

21 U.S.C. § 348(c) (3) (A) (1988)

29 U.S.C. § 655 (1988)

33 U.S.C. < 1317(a) (1958)

42 U.S.C. g 300(f) ef seg. (1982)

42 U.S.C. R 6421 (1982)

42 U.S.C. § 6901 ef seq. (1982)

42 U.S.C. § 7412 (1982)

OTHER AUTHORITIES:

Equal Employment Opportunity Commission,

Policy Guidance on lU'nited Auto Workers v.

Johnson Controls, Inc.

Neen TTT ree cesuecucscesrereee

National Academy of Sciences-National Research

Council, Biologic Markers in Reproductive Toxi-

Ce cueenucccssceces

National Academy of Sciences-National Research

Council, Risk Assessment in the Federal Gov-

ernment: Managing the Process (1983)

National Academy of Sciences-National Research

Council, Principles for Evaluating Chemicals

in the Environment (1975) -

National Academy of Sciences-National Research

Council, Evaluating the Safety of Food Chem-

icals (1970) |

I. Nisbet & N. Karch, Chemical Hazards to Hu-

man Reproduction (1983)

Note, E.vrelusionary Employment Practices in Haz-

ardous Industries: Protections or Discrimina-

tions? 5 Colum. J. Envy. L. 97 (1978)

10

IN THE

Siuprenw Court of the Wniied States

OCTOLER TERM, 1989

No. 89-1215

INTERNATIONAL UNION, UNITED AUTOMOBILE,

AEROSPACE AND AGRICULTURAL IMPLEMENT

WORKERS OF AMERICA, ef a/.,

Petitions rs,

.

JOHNSON CONTROLS, INC.,

Respondent.

On Writ of Certiorari to the

United States Court of Appeals

for the Seventh Circuit

BRIEF OF

NATURAL RESOURCES DEFENSE COUNCIL, INC.

AS AMICUS CURIAE IN SUPPORT OF PETITIONERS

INTERESTS OF AMICUS!

The Natural Resources Defense Council (NRDC)}, a

nonprofit corporation whose principal place of business is

at 122 East 42nd Street, New York, N.Y., has offices in

Washington, D.C. and San Francisco, California, and a

1 All parties have consented to the submission of this) brief,

Written jetters of consent will be on file with the Clerk.

2

membership of more than 125,000 persons across the

United States and in several other nations. NRDC is

dedicated to the perservation, enhancement and defense

of the human environment and natural resources of the

United States. In carrying out these objectives, NRDC

has undertaken efforts to obtain clean air, clean water

and toxic substances control, including activities to en-

sure effective implementation of health and _ environ-

mental protection statutes. In connection with these ac-

tivities, NRDC has on numerous occasions relied upon

and supported governmental reliance upon laboratory

animal studies to assess the risks that exposure to chem-

icals poses to humans.

SUMMARY OF ARGUMENT

The Court of Appeals erred in concluding that scientific

data from experiments with laboratory animals are too

speculative to support conclusions about the reproductive

toxicity of chemicals in human males. Scientists rou-

tinely use laboratory animal testing to assess the risks

that chemicals may pose to human beings. The regula-

tory agencies that Congress has charged with protecting

humans from the harmful effects of chemicals in the en-

vironment have consistently adhered to the principle that

laboratory animal studies are predictive of human health

risks, absent clear proof to the contrary. The Court of

Appeals was therefore wrong as a matter of scientific

fact, as a matter of law, and as a matter of sound pub-

lie policy in concluding that laboratory animal evidence

of the reproductive toxicity of lead in human males was

“speculative and unconvincing.”

3

ARGUMENT

THE COURT OF APPEALS ERRED IN CONCLUDING

THAT SCIENTIFIC DATA FROM EXPERIMENTS

WITH LABORATORY ANIMALS ARE TOO SPECULA-

TIVE TO SUPPORT CONCLUSIONS ABOUT THE RE-

PRODUCTIVE TOXICITY OF CHEMICALS IN HU-

MAN MALES.

In upholding Respondent Johnson Controls’ “business

necessity’’ defense, both the District Court and the Court

of Appeals rejected Petitioner International Union’s ar-

gument that the risks to unborn children posed by worker

exposure to lead in the workplace is not limited to fe-

males.” The International Union offered expert testi-

mony to establish that male worker exposure to lead in

the workplace also poses a substantial risk to unborn

children. Although the International Union’s experts did

not attempt to prove conclusively that any particular

human malformity had resulted from any particular

father’s exposure to lead in the workplace, they concluded

inter alia on the basis of scientific studies conducted with

laboratory animals that the exposure of human males to

lead in the workplace posed a risk of genetic damage to

their offspring. The Court of Appeals found, apparently

as a matter of law, that this expert testimony would be

inadequate, even if fully accepted by the factfinder, to

support a conclusion that male worker exposure to lead

* The Court of Appeals addressed ¢! stion of the isefulness

of animal studies in evaluating reproduectivg risxs in the context o!

the “business necessity” defense, whieh that t found applicable,

despite the fact that the defendant's em) ment practices clearly

constituted disparate treatment. If this Court accepts Petitioners

argument (and the position of Judge Easterbrook below) that in

this case of disparate treatment the employer has the burden of

proving that the employment practice constitutes a bona fide oceu-

pational qualification, then the court need not reach the question

of the appropriateness of animal studies for assessing human risks,

4

posed substantial risks to unborn children. In relevant

part, the Court of Appeals reasoned as follows:

... |T}he UAW witnesses posited that animal stud-

ies had demonstrated that there was a possible risk

of genetic damage to human offspring as a result of

male lead exposure. The UAW witnesses attempt to

bridge the wide chasm between the results of animal

studies and a conclusion of genetic harm allegedly

transmitted through the male human being with

human studies merely establishing a correlation be-

tween male lead exposure and changes in sperm

shape. It is interesting to note that the UAW has

not presented any medical evidence in the record of

any human study scientifically documenting genetic

defects in human beings resulting from male lead

exposure. It is this lack of convincing scientific data

that the plaintiffs attempt to gloss over and cast

aside in ignoring the differences between the effect

of lead on the human and animal reproductive

systems.

. . . Unlike the record evidence of a substantial

risk of harm resulting to an unborn child from ex-

posure to lead through the mother’s blood stream

and placenta, the evidence of risk to the unborn

child resulting from exposure of the father to the

lead levels currently present in Johnson Controls’

battery manufacturing factories is, at best, specu-

lative and unconvincing. The UAW’s animal re-

search evidence does not present the type of solid

scientine data necessary for a reasonable factfinder

to reach a nonspeculative conclusion that a father’s

exposure to lead presents the same danger to the un-

born child as that resulting from a female employee's

exposure to lead. . .

International Union v. Johnson Controls, Ine., 886 F.2d

871, 889-90 (7th Cir. 1989). This statement is erroneous

as a matter of law, as a matter of sound publie policy, and

as a niatter of scientific analysis.

5

A. An Established Body of Scientific Opinion Supports

the Use of Laboratory Animal Studies in Assessing

the Reproductive Toxicity of Chemicals.

Scientists attempting to assess the toxicity of chemicals

to human beings have a limited number of testing op-

tions. They can, of course, expose a group of human

beings to various predetermined doses of the chemical

being tested, observe the effects of the chemical on the

exposed humans and their offspring, and compare the re-

sponses of the exposed humans to humans in a “control”

group of unexposed, but otherwise equivalent humans.

Obviously, when a chemical is suspected of producing

toxic effects in humans or their offspring, this kind of

con‘rolled human testing raises grave moral questions.

Past experiences with abuses of human experimentation

have led federal agencies to restrict severely the condi-

tions under which federally sponsored human experi-

ments may take place in this country.‘ -ven if human

testing were ethically allowable, it would take several

exposed groups containing thousands of individual hu-

mans to draw conclusions about the effects of low ex-

posures ‘of the sort ordinarily encountered in the work-

place! with any degree of statistical confidence. Obvi-

ously, such studies would be exceedingly burdensome and

expensive.

In the absence of controlled human experiments, the

next best source of scientific data on the human toxicity

of chemicals is an epidemiological study in which a group

of humans who have historically been exposed to greater

than normal amounts of the chemical is isolated and com-

pared to an equivalent “control” group. A whole host of

methodological problems typically plague epidemiological

studies of workplace chemicals. For example, it is often

very difficult to determine the historical exposure levels

of the workers in the exposed group. Poor medical rec-

‘See 45 CFR. pt. 16 61989) (Department of Health and Human

Services Regulations on the Protection of Human Subjects.

6

ords also make it difficult to ascertain the incidence of

particular diseases and to determine the cause of par-

ticular deaths. Selection and information biases can con-

found attempts to analyze the data objectively. Finally,

because most epidemiological studies ean examine only a

limited number of subjects, only very strong effects can

be confidently isolated.*

Because of the serious limitations that plague human

testing and human epidemiological studies, scientists test

chemicals in laboratory animals to discover their toxic

effects. Although scientists generally agree that labora-

tory rodents are not identical in all regards to human

beings, there is also general agreement that the results

of laboratory »nimal toxicity testing are relevant to as-

sessing the risks of human exposures to the tested sub-

stances. According to the widely-quoted Report of the

National Academy of Sciences’ Committee on the Institu-

tional Means for Assessment of Risks to Public Health:

“The inference that results from animal experiments are

applicable to humans is fundamental to toxicological re-

search; this premise underlies much experimental biology

and medicine. . . .”"° The Carcinogen Guidelines of the

*See United States Environmental Pretection Arency, Proposed

Guidelines for Assessing Male Reproductive Risk and Request for

Comments, 53 Fed. Reg. 24850, 24861 (198%): United States En-

vironmental Protection Agency, Proposed Amendments to the

Guidelines for the Health Assessment of Suspect Developmental

Toxicants; Request for Comments; Notice, 54 Fed. Reg. 9386,

I2RGH-99 (1989). See generally National Academy of Sciences-

National Research Council, Principles for Evaluating Chemicals

in the Environment 126 (1975): National Academy of Sciences-

National Research Counci!, Evaluating the Safety of Food Chemi-

cals 44 (1970,; Office of Science and Technology Policy, Chemical

Carcinogens: A Review of the Science and its Associated Prin-

ciples, 50 Fed. Reg. 10572. 10379-80 (1985).

* National Academy of Sciences-National Research Council, Risk

Assessment in the Federal Government: Managing the Process 22

(1983).

7

Federal Office of Science and Technology Policy, which

reflect the combined efforts of senior scientists from all

of the major federal regulatory and research agencies,

start with the proposition that

it is reasonable to treat an animal carcinogen as if

it were a human carcinogen. ... This principle has

been accepted by all health and regulatory agencies,

and is widely regarded by scientists in industry and

academia as a justifiable and necessary inference.”

These strong statements by panels of experts drawn

from the most prestigious members of the scientific com-

munity contradict the sugestion in the affidavits of John-

son Controls’ paid experts (untested by cross-eXamination

or peer review) that the results of animal studies “were

not scientifically established as being applicable to hu-

mans.” International Union v. Johnson Controls, Ine.,

886 F.2d 871, 889 (7th Cir. 1989). The Court of Ap-

peals erred in relying on the affidavits of the experts

hired by Johnson Controls for the proposition that the

International Union’s “animal research evidence does not

present the type of solid scientific data necessary for a

reasonable factfinder to reach a nonspecuiative conclu-

sion....” Jd. at 889.

In particular, animal studies are relevant (indeed,

often critical) to the assessment of the reproductive risks

that chemicals in the workplace environment pose to hu-

man beings. Nisbet and Karch, for example, devote a

chapter of their took “Chemical Hazards to Reproduc-

tion” to “Concordance Between Reported Effects in

Humans and Measured Effects in Animais.”” The authors

report that ‘“‘{a] basic assumption in toxicology is that

effects obser "ed in experimental animals can be used to

infer their likely effects (or lack of effects) in humans,

with appropriate consideration of the biological! differ-

" Office of Science and Technology Policy, Chemical Carcinogens:

A Review of the Science and its Associated Principles, 50 Fed.

Reg. 10372 (1985).

8

ences between species.” * After examining the human

and animal data for a selected group of chemicals su-

spected of causing reproductive effects, the authors con-

clude:

Teratogenic effects in humans are matched by terato-

genic effects in one or more species of experimental

animal, and are also matched by the effects of sim-

ilar risks in at least one species. ... |T]here may

be a fairly close concordance in the dose levels at

which humans and animals respond to those agents,

humans appearing somewhat more susceptible than

the most sensitive animal species tested. In addi-

tion, .. . there may be similar concordance between

humans and animals for other types of effects on

reproduction, although in at least one case an ani-

mal species appeared to be much more sensitive than

humans.*

A National Academy of Sciences expert panel began its

discussion of the use of biologic markers in reproductive

toxicology with the observation that ‘“|e|xtrapolations to

humans are to be based on the most sensitive animal

species tested, barring clear evidence that the species is

toxicologically distinct from humans.” ”

The Court of Appeals may also have misunderstood

the reason that animal studies typically use one or more

dose levels that are considerably higher than the level to

which humans are exposed in the workplace. Testing

animals at high doses is simply a matter of practical

necessity. To detect effects that are significant in large

exposed populations requires either testing at higher dose

levels or testing an impractical number of animals. One

National Academy of Sciences panel explained:

‘I. Nisbet & N. Karch, Chemical Hazards to Human Reproduc-

tion 93 (1983).

“Jd. at 111.

* National Academy of Sciences-National Research Council, Pio-

logic Markers in Reproductive Toxicology 3 (1989).

9

The study of reproduction and development poses

major resource and logistic problems for those work-

ing with laboratory animals. For instance, manage-

able sample populations do not reveal increases in

toxic events of less than 5 to 10%. For some health

effects associated with reproduction and neurode-

velopment, such as mutagenesis and teratogenesis,

incidences in a human population of 3 per 10,000 are

significant. Obviously, these effects cannot be well

defined in whole-body ctudies of thousands of ex-

perimental animals at a time. Classic toxicology

studies of rodents involve the exposure and _ path-

ological analyses of 200 animals for 2 years. In

such assays, each animal is a surrogate for 1,000,000

people. Birth defects undetectable in this rodent

population could be epidemic within ten generations,

if they occurred in humans."

In this connection, it is important to note that Dr. Silber-

geld (one of the plaintiffs’ experts) testified emphatically

in her deposition that “|t|here are no clear thresholds

for any of these effects of lead.” Silbergeld Deposition at

17. This testimony, which must be taken as true for

purposes of a motion for summary judgment, indicates

that high dose animal studies are relevant for assessing

low dose human risks. It is therefore prudent to assume

that, in the absence of a demonstrated threshold if toad

causes adverse reproductive effects in laboratory animals

at high doses, it will pose some risk of adverse effects in

humans even at substantially lower exposure levels.

The Court of Appeals apparently failed to recognize

that workplace chemicals such as lead can affect fetal

development and the health of the subsequently born

child in several ways."’ After a fetus is in existence,

1” National Academy of Science-National Research Council, Bio-

logic Markers in Reproductive Toxicology 4 (1989).

11This failure on the part of the Court of Appeals is und -r-

standable. Judges have no expertise in toxicology and should there-

10

it can be harmed by exposure to a chemical substance

(transported across the placenta) in much the same way

that an adult can be affected, although the fetus may be

more susceptible io the chemical’s toxic effects than a

child or adult. The unborn child can also be damaged

prior to conception when a chemical causes mutation in

the genetic material of the sperm cell or the egg cell.

Although such muiations may not affect the fetus in any

significant way, they may well cause adverse effects that

ear severely affect the life of the human being post par-

tum.’* The first kind of adverse effect only occurs when

a pregnant female is exposed to a toxic chemical. The

second kind of effect can occur when either a male or a

female is exposed. As a general matter, it cannot be

said that one effect is more significant than the other,

and for that reason it is impossible to conclude that ex-

posure to a chemical that causes only the first kind of

effect is riskier than exposure to a chemical that causes

only the second kind of effect. Exposure to one chemical

may present a high probability of a relatively modest

consequence ‘e.g., slightly reduced I.Q.) while exposure

to another chemicz! may pose a low probability of a dis-

astrous consequence (e.g., hydrocephalus). In the ab-

sence of a great deal of information about the relative

probabilities and without some very difficult judgments

about the relative severity of disease end-points, it is im-

possible to conclude t! at one exposure poses a higher

risk to the unborn human than the other.

fore be cautious about drawing conclusions about the worthiness

scientific data. This caution is especially warranted when the scien-

tific information available to the court comes in the form of un-

tested expert testimony in litigation affidavits.

12 See generally Note, FErclusionary Employment Practices in

Hazardous Industries: Protection or Discrimination? 5 Colum.

J. Envy. L. 97, 99-100 (1978). Dr. Legator’s desposition testimony,

which must be accepted av true for purposes of a motion for sum-

mary judgment, was that “lead is one of the very few chemicals

... that probably causes a genetic lesion during spermatogenesis.”

Legator Deposition at 20.

11

The real world is more complex than the above example

suggests. Exposure to a single substance may pose a risk

of both kinds of adverse effects. Preconception exposure

of either the male or female may pose a small probability

of mutation leading to a grotesque malformity, while

post-conception exposure of the female may lead to more

modest health effects. Thus, the International Union’s

claim with respect to lead exposure in the workplace is

not at all fanciful. The post-conception scenario is rela-

tively well-understood. Scientists know that lead passes

from the mother's blood to the fetus at some time during

pregnancy. Scientists also know that elevated lead levels

in children are associated with various diseases ranging

in severity from relatively minor adverse effects at low

blood levels to more severe effects at high blood levels.

The knowledge-base for these relatively strong scientific

conclusions consists of human epidemiological studies, a

large number of animal studies and occasional clinical

observations.

It is much more difficult to explore preconception tox-

icity with epidemivlogical studies. Proper epidemiology

would require knowledge about both the mother’s and

father’s exposures and about all pregnancy outcomes,

even though some mutations result in spontaneous abor-

tions. This information is much harder to acquire.

Scientists, regulatory agencies and courts are therefore

forced to rely more heavily upon laboratory animal stud-

ies as the knowledge-base from which to draw conclu-

sions about the effects of preconception exposure to

chemical substances. Reliance on animal studies is con-

cededly more “speculative,” but it was wrong for the

court to conclude that animal studies are not “convinc-

ing’ or “solid” scientific data. When existing epidemio-

logical studies do not provide adequate information on

human exposures to the substance at issue, it is not at all

“unscientific”? or ‘unreasonable’ to draw conclusions about

human risks from existing animal studies.

12

8. Courts and Administrative Agencies Have Consist-

ently Relied Upon Laboratory Animal Studies in

Assessing the Risks of Chemicals to Humans.

From the beginning of this century, Congress has

enacted precatory statutes designed to protect the pub-

lie health from the toxic effects of chemicals in food, the

workplace and the general enivornment. The regulatory

agencies that Congress created to implement these statutes

have, with the approval and often active encouragement

of the reviewing courts, consistently relied upon labora-

tory animal studies in evaluating human risks. In the

last two decades, governmental reliance on laboratory

animal studies has become thoroughly interwoven into

the regulatory fabric of our society.

A partial list of federal laws in which animal studies

are routinely used to assess human health risks includes:

the Federal Insecticide, Fungicide and Rodenticide Act; "’

the Federal Food, Drug and Cosmetie Act ;'* the Consumer

Product Safety Act: the Safe Drinking Water Act;"

7 USC. $ 136-137"; EPA routinely relies upon anima! data in

reaching pesticides regulatory decisions, see, e.g. 40 C.F.R. Part

154.7 a1 2)(i) (“validated test data” used for cancellation deci-

sions. due to carcinogenicity, teratogenicity or reproductive toxicity

f<.m “effects demonstrated in humans or experimental animals”).

149] USC. $348(¢c)(3)(A) (no additive shall be deemed to

he safe if it is found to induce cancer when ingested by man or

animal.” )

1515 U.S.C. § 2079(a) and the Federal Hazardous Substances Act

$9(f)(1)(A), 15 U.S.C. § 1261, 1262. See e.g., 16 C.F.R. § 1500.17

(a) (10). and 43 Fed. Reg. 12308-12310 (animal studies are the

basis for classification as banned hazardous substance of self-

pressurized household substances containing vinyl chloride mon-

omer).

16 42 U.S.C. $$ 300(f) et seq. See, eg., 44 Fed. Reg. 68624, Na-

tional Interim Primary Drinking Water Regulations; Control of

Trihalomethanes in Drinking Water:

sufficient scientific evidence had been accumulated to con-

clude that chloroform is an animal carcinogen as shown from

13

the Toxic Substances Control Act;'’ the Federal Water

Pollution Control Act;'* the Federal Clean Air Act;'’ the

a properly conducted bioassay and sheuld be presumed to be a

risk to humans and that, as such, prudent public health pol-

icy warrants reasonable measures to reduce human exposure.

... In the specific case of chloroform and other THMs, EPA

had relied primarily on animal studies demonstrating the

toxicology of chloroform. ... EPA has extrapolated from the

results of animal studies to assess the risk posed by THMs to

humans.

Id. at 68627.

1715 U.S.C. §§ 2601-2629 (1976). § 2603(b)(2)(A) provides:

The health and evironmental effects for which standards for

the development of test data may be prescribed include carcino-

genesis, mutagenesis, teratogenesis, behavioral disorders, cum-

ulative or synergistic effects, and any other effect which may

present an unreasonable risk of injury to health or the environ-

ment. ... The methodologies that may be prescribed in such

standards include epidemiologic studies, serial or hierarchical

test, in vitro test, and whole animal tests,....

(Emphasis supplied); see also 40 C.F.R. Part 798 which specifies

the use of numerous animal test procedures to determine chemical

oncogenicity ($$ 798.3300, 798.320) and _ reproductive toxicity

($$ 798.5100-798.5955) for purposes of implementing $4 of the

Act.

iSFWPCA §$ 307(a), 33 U.S.C. $1317(a) (1976). See, eg., 41

Fed. Reg. 30476-77 (1976) and 42 Fed. Reg. 6532 (1977) (pro-

posed and final discarges standards for PCBs based on evidence of

animal carcinogenicity ).

18 42 U.S.C. § 7412. See, e.g. National Emission Standards for

Hazardous Air Pollutants: Proposed Standard for Vinyl Chloride,

40 Fed. Reg. 69532 (December 24, 1975). (The rationale for the

standard included animal test data demonstrating carcinogenicity

of vinyl chloride as well as epidemiological studies.) See also, Na-

tion Emission Standard for Vinyl Chloride, 40 C.F.R. Part 61,

Subpart F.

14

Occupational Safety and Health Act;*” and the Resource

Conservation and Recovery Act.”

Under these statutes, the federal courts have uniformly

recognized the need to regulate carcinogens and repro-

ductive toxins on the basis of animal test results in the

absence of definitive human epidemiological data. The

plurality opinion of this Court in the Benzene case recog-

nized the practical necessity of basing public policy deci-

sions affecting public health on laboratory animal studies.

The opinion observed that several “rational”’ techniques

existed for determining whether human exposure to a

‘oxic substance in the workplace presented a “significant

risk” of material health impairment and noted that “|iJn

other proceedings, the Agency has had a good deal of

data from animal experiments on which it could base a

conclusion on the significance of the risk.” /ndustrial

Union Dept., AFL-CIO v. American Petroleum Inst., 448

U.S. 607, 657 n.64 (1980). If, as the Court of Appeals

in this ease maintained, evidence from animal studies is

in fact “speculative” and “unconvincing,” it would have

been disingenuous for this Court to have suggested that

2099 U.S.C. $655. See OSHA Cancer Policy which “establishes

the criteria and procedures under which substances will be regu-

lated by OSHA as potential occupational carcinogens.” 29 C.F.R.

$ 1990.111(a). Positive results in high-dosage animal tests “will

he used to establish the qualitative inference of carcinogenic hazard

to workers.” 29 C.F.R. §$ 1990.143(g). See, e.g. 39 Fed. Reg.

95290, 35891 (1974), 38 Fed. Reg. 10929 (1973) (OSHA exposure

standards for vinvl chloride and ethyleneimine, respectively, based

on animal studies indicative of carcinogenicity ).

21 42 U.S.C. § 6901 ef seq. See 42 U.S.C. § 6421 and implementing

regulations at 40 C.F.R. 261.11(a):

The Administrator shall list a solid waste as a hazardous

waste [if]... (3) it contains any of the toxic constituents

listed in Appendix VIII .. . Substances will be listed in Ap-

pendix VIII only if they have been shown in scientific studies

to have toxic, carcinogenic, mutagenic or teratogenic effects on

humans or other life forms.

(Emphasis supplied).

15

OSHA could meet the Court’s “significant risk” test by

relying solely upon such evidence.

The United States Court of Appeals for the District of

Columbia has observed that the government

has a ‘heavy burden’ to ‘explain the basis for .

la] decision to permit the continued use of a chem-

ical known to produce cancer in experimental ani-

mals.’ [citations omitted’. When firm evidence es-

tablishes that a chemical! is a carcinogen, statutes

leave an administratee-no alternative but to step in

and protect the public.

Environmental Defense Fund, lnc. v. EPA, 598 F.2d 62,

88 (D.C. Cir. 1978. See also Environmental Defers

Fund, Inc. v. EPA, 510 F.2d 1292 (D.C. Cir. 1975) Cup-

holding the ban of the pesticides aldrin and dieldrin based

upon carcinogenicity in rats and mice); Environmental

Defense Fund, Inc. v. Ruckelshaus, 439 F.2d 584, 590

(D.C. Cir. 1971) (DDT); EDF v. Costle, 578 F.2d 337%,

345 (D.C. Cir. 1978) (court remanded to EPA regula-

tions implementing the Safe Water Drinking Act that

failed adequately to protect the public upon “ample evi-

dence . . . our drinking water is contaminated with a

large variety of organic substances, of demonstrated car-

cinogenicity in animals”); Environmental Defense Fund,

Inc. v. EPA, 548 F.2d 998, 1005-06 (D.C. Cir. 1976),

cert. denied, 431 U.S. 925 (1977) theptachlor and chlor-

dane) (court approved EPA’s use of “cancer principles”

that “accept the use of animal test data to evaluate bu-

man cancer risks; [and] consider a positive carcinogenic

effect in test animals as sufficient to characterize a pesti-

cide as posing a cancer risk to man... .”); Hercules Ine.

v. EPA, 598 F.2d 91 (D.C. Cir. 1978) (EPA regulation

of toxaphene and endrin under section 307 of the Clean

Water Act, 33 U.S.C. § 1317): Society of Plastics Indus-

try v. OSHA, 509 F.2d 1301 (D.C. Cir. 1975) ‘OSHA

regulation of vinyl! chloride): Synthetic Orgqanie Chem,

Mfrs. Assn. v. Brennan, 503 F.2d 1155 (2rd Cir. 1974)

16

(court upheld OSHA’s workplace exposure standard for

ethyleneimine, upon its demonstrated carcinogenicity in

mice and rats, noting “the extrapolation gathered from

data in these rodent experiments to humans is justified by

the report of the Ad Hoc Committee on the Evaluation of

Low Levels of Environmental Chemical Carcinogens to the

Surgeon General’); Bell v. Goddard, 366 F.2d 177, 180-81

(7th Cir. 1966) (court affirmed an FDA ban on diethelstil-

bestrol (DES) as a poultry feed supplement on evidence of

its carcinogenicity in test animals, noting “there is a

histologically remarkable similarity between cancers of

the endometrium in humans and animals”).

’

Judge Easterbrook, in his dissenting opinion in the

Court of Appeals below, speculated that ‘t]he medical

profession, like the Food and Drug Administration, will

be stunned to discover that animal studies are too ‘specu-

lative.” 886 F.2d at 919. Judge Easterbrook was right.

Not only have the medical p: .ession and the Food and

Drug Administration consistently relied upon animal

studies to assess human health risks, but virtually every

other federal agency charged with protecting the public

health from hazardous chemicals has also extrapolated

from “mouse-to-man” in evaluating human health risks.

'ndeed, it would be irresponsible for an agency charged

with protecting health to do otherwise.**

-- In his deposition testimony Dr. Legator gave an example of an

instance in which the Environmental Protection Agency’s failure

to act on the basis of animal studies on the pesticide DBCP resulted

in sterility in workers exposed to that substance:

With DBCP we had substantial information about the fact that

this chemical at five parts per million caused testicular atrophy

in a variety of animals. The study was generated in 1963.

And basically we ignored the animal data, until the workers

found an effect themselves in 1977.

Leyator Deposition at 52. Upon discovering the human conse-

qucpce, EPA cancelled the registration of DBCP. 50 Fed. Reg.

1122 (1985). OSHA promulated « standard to protect workers

from DBCP in 1978. 13 Fed. Reg. 11514 (1978).

17

In promulgating its standard for workplace exposure

to lead, the Occupational Safety and Health Administra-

tion concluded:

The record in this rulemaking is clear that male

workers may be adversely affected by lead as well as

women. Male workers may be rendered infertile or

impotent, and both men and women are subject to

genetic damage which may affect both the course and

outcome of pregnancy... .”°

In forming this conclusion, the agency relied heavily upon

the very animal studies that the Court of Appeals in this

ase summarily rejected on the basis of a summary judg-

ment record devoid of cross-examination, peer review or

any other validating technique.

The United States Environmental Protection Agency

in 1988 proposed guidelines for assessing male and fe-

male reproductive risks “‘ and for assessing the risks of

suspected developmental toxicants.** Both sets of cuide-

lines rely heavily upon data from animal experiments in

evaluating human risks. Under the “weight of the evi-

dence” approach that the guidelines adopt for character-

izing male reproductive risks due to chemical exposures,

the guidelines posit that a “known positive” reproductive

effect exists when “/a] convincing body of evidence exists

that an agent causes an adverse effect on the male repro-

ductive system in humans.” A “probable positive” effect

exists when “{a] ecnvincing body of evidence exists that

“Occupational Sefety and Health ‘dministration, Final Stand-

ard for Oecupational Exposure to Lead, 18 Fed. Rey. 52952, 52966

(1T97R).

“' Enviconmental Protection Ageney, Proposed Guidelines for

Assessing Female Reproductive Risk, 53 Fed. Reg. 21834 (1988):

Environmental Protection \e@enev, Proposed Guidelines for Assess-

ing Male Reproductive Pisk. 53 Fed. Reg. 24850 (1988).

** Environmental Protection Agency, Proposed Amendments to

the Guidelines for the Health Assessment of Suspect Developmental

Toxicants, 54 Fed. Reg. 9386 (1989),

18

an agent causes an adverse effect on the male reproduc-

tive system in nonhuman mammals.” 53 Fed. Reg. at

24863. While it is clear that human testing and epidemio-

logical studies are the strongest evidence of male repro-

ductive effects, the EPA has concluded that animal stud-

ies alone are not “specuiative” or “unconvincing” evidence

of human reproductive effects.

The Guidelines for Developmental Toxicity address,

inter alia, the risks that pre- and post-conception expo-

sures pose to unborn humans.** One of the fundamental

assumptions underlying the guidelines is that

an agent that produces an adverse developmental

effect in experimental animal studies is assumed

to pose a potential hazard to humans following eX-

posure during development. This assumption 1s based

on the data for known human developmental toxi-

cants. In almost all cases, the experimental animal

data would have predicted a developmental effect in

humans.

54 Fed. Reg. at 9390. The guidelines provide that ‘ade-

quate evidence of potential human developmental toxicity

exists where there is ‘data from an appropriate, well-

executed study in a single experimental animal species

that demonstrates developmental toxicity... .” 54 Fed.

Reg. at 9399. It goes without saying that “adequate”

evidence is not “speculative” or “unconvincing.”

Clearly, the Court of Appeals’ conclusion that labora-

tory animal data do not constitute “convincing scientific

data” does not comport with the scientific judgment of

26 The guidelines define “developrnental toxicology” as “Ttlhe

study of adverse effects on the developing organism that may re-

sult from exposure prior to conception (either parent), during pre-

natal development, or postnatally to the time of sexual maturation.

_.. The major manifestations of developmental toxicity include:

1) death of the developing organism, (2) structural abnormality,

(3) altered growth and (4) functional deficiency.” 51 Fed. Reg. at

9390-91.

19

the agency that Congress has charged with regulating

environmental risks to human reproduction.

The Equal Employment Opportunity Commission is-

sued policy guidance to its investigators in the wake of

the Court of Appeals’ decisions in this case that spoke

directly to the use of laboratory animal data to assess

human risks in both sexes:

The majority in Johnson Controls rejected UAW’s

animal research evidence, stating that it was “specu-

lative,” “unconvincing” and not “solid scientific

data.” . . . Commission investigators should not re-

ject animal studies in such a wholesale manner... .

These studies may in some cases provide the best

available evidence to evaluate fetal or other human

health hazards in the workplace. Certainly, the data

from the animal research in this case, along with the

other data provided, should have been sufficient to

create a genuine issue of material fact warranting

denial of summary judgment.”

The Commission was clearly correct in telling its investi-

gators not to fall into the trap that ensnared a majority

of Court of Appeals in this case.

The Commission correctly pointed out in its policy

guidance that the question of the extent to which human

health risk assessment should rely upon animal toxicity

data was more properly the domain of the public health

agencies like the Environmental Protection Agency and

the Occupational Safety and Heaith Administration.”

27 Equa! Employment Opportunity Commission, Policy Guidance

on United Auto Workers v. Johnson Controls, Inc., reprinted at

p. 127a of the Appendix to the Petition for Certiorari.

-* The Policy Guidance states:

In determining whether the harm is mediated only through one

sex, animal research should be considered along with other

objective scientific evidence. The Commission will defer to

pertinent findings by government agencies such as OSHA and

20

The reviewing courts, inclucing this Court, should heed

this sage advice. Court are not well qualified, either

technically or institutionally, to engage in wholesale

science policy judgments of the sort that the Court of

Appeals in this case mandated. As we have seen, vir-

tually every federal agency that is charged with imple

menting public health laws routinely relies upon animal

testing data in assessing human health risks. This Court

should defer to that expert judgment.

CONCLUSION

The International Union in this ease contends that

Johnson Controls bas the burden of proving that its dis-

parate treatment of women reflects a “bona fide occupa-

‘onal qualification.” If this Court accepts this view of

the Jaw, then there is no need to address the use of animal

studies in evaluating human reproductive risks. To meet

the “business necessity” test adopted by the Court of Ap-

neals én thts ease however, the International Union of-

fered cvidenee from laboratery animal studies tending to

prove thst lead exnosure to male humans at levels com-

sayable te thes found in the Johnson Controls plant

posed a significant risk of adverse effect to unborn chil-

dren through mutations of the genetic material in sperm

cells. The Internations! Union was not allowed to intro-

duce this evidence in its legitimate attempt to persuade

the factfinder that the overnll risk posed by male exposure

i's roughly comparable ‘if not higher) than the risk posed

by female exposure through transplacental passage of

lend into the blood stream of existing fetuses In finding

that the evidence from animal testing was “speculative”

and “uneonvineing,” the Court of Appeals departed from

a long-aecep ted and well-established practice of scientists

NIOSH {the National Institute for Occupational Safety and

W Ith

Lp; pendix te the Petition for Certiorari at (an

21

reru) dtory paencics end yoviewin’ courts of usin anime!

stud es to evaluate human health risks.

Respectfully submitted,

THOMAS O. MCGARITY

Counsel of Record

University of Texas

School of Law

727 E. 26th Street

Austin, Texas 78705

(512) 471-5151

ALBERT H. MEYERHOFF

90 New Montgemery

San Francisco, CA 94105

Attorneys for Amicus Curiae