RISK ASSESSMENT
FOR
4,4'-METHYLENEDIANILINE
CAS NO. 101-77-9
OFFICE OF TOXIC SUBSTANCES
U.S. ENVIRONMENTAL PROTECTION AGENCY
FEBRUARY 2, 1985
J. W. Hirzy, J.A. Wiltse, D. Eberly, B. Cook, G. Grindstaff,
K. Hammerstrom, D. Heggem, J. Helm, R. Kuchkuda, C.R. Mathiessen,
H. Milman, S. Ng, J. Remmers, C. Scott-Siegel, J. Springer,
S. Strassman-Sundy, G. Thies, A. Auletta.
-------
TABLE OP CONTENTS
LIST OF FIGURES AND TABLES iv
ACKNOWLEDGEMENTS vi
EXECUTIVE SUMMARY vii
I. INTRODUCTION 1
II. CHEMICAL IDENTITY 5
III. HEALTH EFFECTS . . 8
A. Absorption and Distribution 9
1. General 9
2. Inhalational and Dermal Absorption 10
a. MBOCA Data 10
i. Percutaneous Penetration 10
ii. Gastro-Intestinal Tract
Penetration 11
iii. Respiratory Tract Penetration .... 12
b. Inferences About 4,4'-MDA 13
c. Limits of this Analysis 14
B. Mutagenicity 15
C. Carcinogenicity 17
1. Human Data 17
2. Animal Data 21
a. NTP Studies 21
b. Hiasa Study 23
c. Deichmann Study 24
d. Schoental Studies 25
e. Griswold Study 26
f. Fukushima Studies 27
g. Steinhoff and Grundmann Study 28
h. Zylberszac Study 29
i. Munn Studies 29
j. Gohlke Studies 30
3. Structural-Activity Relationships 32
4. Summary of Animal Data 35
5. Weight of Evidence 36
a. Animal Studies .37
b. Epidemiological Studies 38
c. Structure-Activity Relationships 39
d. Absorption 40
e. Mutagenicity . 40
D. Other Human Health Effects 41
IV. EXPOSURE ASSESSMENT . 43
A. Workplace Exposure—4,4'-MDA/MDI Manufacturing ... 44
1. Production Processes ...» ..44
a. 4,4'-MDA ' 44
b. MDI 49
2. Production History and Forecast 50
-------
3. Characterization of Non-MDI uses of 4,4'-MDA . . 54
4. Exposure Levels and Duration 54
a. 4,4'-MDA and MDI Production Workers .... 57
b. Non-MDI Uses of 4,4'-MDA 61
i. Epoxy Curing 61
a. Exposures Related to the Curative
Package 61
b. Exposures Related to Handling/Use
of Complete Epoxy Resin
Formulations 66
c. Exposure in an Un-Characterized
Setting 70
ii. Co-reactant in Polyurethanes 71
iii. Wire Coating, Polyimides
and PMR-15 73
iv. Other Uses 74
B. Potential Exposure Related to Consumer Contact With
4,4'-MDA-Containing Articles or Products 76
C. Potential Exposure Related to Drinking Water and
Ambient Air Contamination 80
1. Releases from 4,4'-MDA Manufacturing 83
2. Releases from 4,4'-MDA Use as a Feedstock ... 85
a. Releases from MDI Manufacturing 86
b. Releases from Other Product
Manufacturing 87
3. Releases from Use of MDI in Polyurethane
Manufacture .88
4. Releases from Polyurethane Products 91
5. Releases from Disposal of Wastes 91
6. Potential Releases from Degradation
of Polyurethane 93
7. Environmental Fate and Transport of 4,4'-MDA
Releases 94
a. Environmental Transport of 4,4'-MDA .... 94
b. Photodegradation 95
c. Oxidation 96
d. Hydrolysis 96
e. Volatilization . 96
f. Sorption 97
g. Bioaccumulation 97
h. Biodegradation 97
i. Summary 98
8. Estimated Surface Water Concentrations
Of 4,4'-MDA .98
9. Populations Potentially Exposed to Contaminated
Surface Waters 98
a. Potential Drinking Water Exposures 99
b. Potential Ambient Environmental
Exposures 99
V. QUANTITATIVE RISK ESTIMATION 103
A. Introduction > 103
B. Methods and Results 105
1. High-Dose to Low-Dose Extrapolation Model .... 105
-------
2. Animal to Human Extrapolation 107
a. Dose-Response in the Bioassays 107
b. Extrapolation of Human LADDs to
Animal LADDs 108
3. Tumors Observed in the Bioassay 109
4. Exposure Situations lid
a. Workplace Situations 119
i. Case 1: 4,4'-MDA/MDI
Manufacturing 120
ii-vii. Case 2-7: 4,4'-MDA
Using/Processing 121
viii. Case 8: MDA Using/Processing
Hypothetical Workplace Standard . . 125
b. Drinking Water Case 126
5. Estimation of Risks for Exposed Populations ... 126
a. Workers . .• 126
b. People Drinking Contaminated Water 126
C. Risk Characterization 127
VI. DISCUSSION 135
Appendix A 139
Appendix B 143
Appendix C 157
REFERENCES 167
-------
LIST OP FIGURES AND TABLES
Figures Page
1. 4,4'-MDA/MDI Flow Diagram ,45
Tables
1. Comparison of Results of Chronic NCI/NTP Studies
on 4,4'-MDA and Related Compounds 33
2. Estimated 4f4'-MDA Production 1976-1980 52
3. Producers of 4,4'-MDA and MDI Plant Locations and
Estimated Capacity 53
4. Non-MDI Uses of 4r4'-MDA 55
5. Annual Production of 4,4'-MDA for Non-MDI Uses 56
6. 4,4'-MDA Manufacturing Workplace Airborne
Exposure Levels 59
7. Estimated Dermal Doses in 4,4'-MDA
Manufacturing Plants 59
8. Estimated 4,4'-MDA Manufacturing Workplace LADDs . . . .61
9. Estimates of MDA Available for Absorption by
Non-MDI Use 63
10. 4,4'-MDA Using/Processing Workplace LADDs 77
11. Estimated 4,4'-MDA Levels in Surface Waters 82
12. Populations Potentially Exposed to 4,4'-MDA in
Drinking Water Downstream of 4,4'-MDA
Manufacturing Plants 101
s
13. Populations Swimming in Surface Water Near
4,4'-MDA Manufacturing Plants 102
14. Tumor Incidence by Speciesf Sex and Site of Tumor . . . Ill
2
15. X* Goodness-of-Fit Test, p-Values for Each Tumor
Type in Table 14 115
16. Historical Incidences of Primary Tumors in Untreated
Control B6C3F1/N Mice 116
-------
17. Historical Incidence of Primary Tumors in Untreated
Control F344 Rats 117
18. Incidence of any Malignancy in F344/N Rats and
B6C3fl/N Mice by Sex for Different Dose Levels
from NTP Bioassays on 4,4'-MDA 118
19. Estimated Extra Lifetime Risk of Cancer for Workers . . 132
20. Estimated Extra Lifetime Risk of Cancer
from Drinking Water Exposures . 134
-------
ACKNOWLEDGEMENTS
The authors are grateful for the highly competent and
innovative work that Dr. Mark Boeniger of the National Institute
of Occupational Safety and Health (NIOSH) contributed to this
investigation, for the dedication, competence, and good cheer of
Paulette Grimes who put this document into print, and for the fine
editing work of Karen Flagstaff.
VI
-------
EXECUTIVE SUMMARY
This document characterizes the cancer risks posed by
4,4'-methylenedianiline (4,4'-MDA).
4,4'-MDA is considered to be a probable human carcinogen (B2)
using the criteria of the Revised Interim Guidelines for the
Health Assessment, of suspect Carcinogens. Evidence for this
classification includes observation of elevated tumor incidence at
multiple sites in rats and mice in NTP bioassays, observation of
an elevated incidence of bladder tumors in a group of exposed
workers, observation of the chemical's ability to react with
genetic material in short-terra studies, indications that the
chemical is rapidly absorbed and distributed in mammals, and a
structual relationship with a number of 2-ring aromatic diamines
that are known carcinogens in animals and/or humans.
4,4' -!*iDA is produced at an annual volume of 400-500 million
pounds in the United States. About 97% of production is converted
to methylene aiphenyldiisocyanate, a ployurethane intermediate,
and about 60U workers are exposed through these processes. The
remaining '6% of productioji is used as a component of epoxy resins
or other polymer systems, and about 13,000 workers may be exposed
tnrouyh these uses.
f
There appears to be little or no consumer use of products
containing free, un-reacted 4,4'-MDA, although some adhesives and
sealants used in such trades as foundation wall crack patching are
known to contain the chemical.
Based on the available information, risks of tumor
development in workers exposed to 4,4'-MDA appear to be
VII
-------
RISK ASSESSMENT OP
4,4'-METHYLENEDIANILINE (4,4*-MDA)
I. INTRODUCTION
In 1979, the Interagency Testing Committee (ITC), established
under section 4 of the Toxic Substances Control Act (TSCA),
recommended that 4,4'-MDA (CAS No. 101-77-9) be considered for
testing for carcinogenicity, teratogenicity, mutagenicity, other
chronic effects, environmental effects and epidemiology (44 PR
31885, June 1, 1979) under authority of section 4 of the Toxic
Substances Control Act (TSCA). In making this recommendation, the
ITC cited sketchy toxicity information that included some cancer
studies, along with concern about high production volume and
reports of adverse health effects among exposed humans.
The Office of Toxic Substances (OTS) within the Environmental
Protection Agency (EPA) then undertook a review of available
information on 4,4'-MDA. Contractors were employed to gather this
information, and by 1982 most of it had been summarized in five
reports by Springborn Laboratories (Springborn, 1982), JRB
Associates (JRB, 1980, 1981), MathTech (MathTech, 1982), and
Environmental Science and Engineering, Inc. (ESE, 1981).
In mid-1982, the Chemical Manufacturers Association (CMA)
established a 4,4'-MDA- Project Panel consisting of
representatives of BASF Wyandatte Corp., The Upjohn Co., Mobay
Chemical Corp. Olin Corp., Rubicon Chemicals (ICI Americas
-------
Inc.), and Uniroyal, Inc.* The Panel began to gather production,
use and exposure information for submission to EPA and to develop
a voluntary testing program.
In June 1982, EPA made 4,4'-MDA subject to the reporting
requirements of section 8(a) of TSCA. As a result, manufacturers
and importers were obligated to report on the amount of the
chemical produced or imported and on certain other information
related to uses and releases to the environment (47 PR 26992,
June 22, 1982). In addition, a rule issued under section 8(d) of
TSCA in September 1982 (47 FR 38780, Sept. 2, 1982) required
reporting of unpublished health and safety studies.
As all of this information was being studied by OTS, the
preliminary results of National Toxicology Program (NTP)
carcinogenesis bioassays on the dihydrochloride salt of 4,4'-MDA
became available. This information, coupled with exposure
reports, led the Administrator of EPA in early 1983 to find that
there is a reasonable basis to conclude that 4,4'-MDA may present
a significant risk of serious harm to humans from cancer. This
finding resulted in the invocation of section 4(f), the priority
review provision of TSCA, and a 180-day priority review of the
reasonableness of the cancer risks associated with 4,4'-MDA began
in March (48 FR 19078, April 27, 1983).
Since most of the identified exposure to 4,4'-MDA occurs in
the workplace, EPA contacted the Occupational Safety and Health
Administration (OSHA), and the two agencies agreed that any
* Shell Chemical Co., CIBA-GEIGY Corp., and Pacific Anchor Corp.
subsequently became members of the Panel.
-------
action needed to control the chemical would be undertaken
jointly.
The National Institute of Occupational Safety and Health
(NIOSH) was contacted in May 1983, and plans to increase the
amount of information on workplace exposures were. made.
Subsequently, NIOSH independently decided to determine whether an
epidemiology study of 4,4'-MDA was feasible.
At the end of the 180-day priority review period, EPA and
OSHA published Advance Notices of Proposed Rulemaking (ANPR)
(OSHA: 48 PR 42836; EPA: 48 PR 42898; Sept. 20, 1983),
announcing a joint effort by the agencies to initiate regulatory
action to determine and implement the most effective means of
controlling exposures to the chemical under TSCA and/or the
Occupational Safety and Health Act.
During the 180-day priority review, EPA published its
decision not to pursue further testing of 4,4'-MDA at that time
because of the apparent need to control exposures to the chemical
to lower risks from cancer (48 PR 31806, July 11, 1983).
The present assessment uses the information obtained by EPA
during the period described above, monitoring information •
obtained by NIOSH from two plant visits, information submitted in
response to the April 1983, section 4(f) designation and the
ANPRs of September 20, 1983, and product use information obtained
by a CMA sponsored survey of firms that use and/or process 4,4'-
MDA commercially. Since the information obtained so far is still
incomplete, especially regarding workplace exposure levels,
-------
potential surface water contamination and skin penetration
potential, the risk assessment will be augmented.by further
analysis as those data become available.
-------
II. CHEMICAL IDENTITY (ESE, 1981)
The name approved by the International Union for Pure and
Applied Chemistry for 4,4'-MDA is 4,4l-diaminodiphenylraethane.
Other synonyms and common names are:
4,4'-methylenedianiline
4,4'-methylenebisaniline
4,4'-methylenebisbenzeneamine
p,p'-methylenedianiline
/
methylenedianiline
dianilinomethane
Trade names for the compound and for mixtures containing it
are (CMA, 1983d):
97% Minimum Assay 4,4'-MDA
p,p'-Methylene Dianiline
Phenyl Base G
Laromin B-250
Hardner HT 972
Eposand 112 B
Crude 4,4'-MDA (Mixture with Isomers and Oligomers)
Curithane 103
Curithane 116
Tonox
Tonox R
Tonox 22
Tonox M
-------
Modified Raw Materials Containing Free 4,4'-MDA
Tonox 60/40
Tonox JB
Tonox LC
Araldite Hardener 830
Araldite Hardener 850
HY 932 (XU 205)
HY 2969
Ancamine LO
Ancamine LOS
Ancamine LT
Ancamine TL
Ancamine TLS
Ancamine 1482
Curing Agent Y
Curing Agent Z
The structure of the compound is:
H
C13 H14 N2
The compound in commercially pure form is a light brown to
white crystalline solid with a faint amine-like odor. It is
prepared commercially by the acid-catalyzed condensation of
formaldehyde and aniline.
H+
2 CgH5NH2 + H2CO ___^_>.H2NC6E?4CH2C6H4NH2 * H2°
-------
In addition to the 4,4'-product, some 2,2'- and 2,4'- by-
products are formed, as are isomeric polycondensates.
2X CgH5NH2 + 2X -1 H2CO > H2NCgH4 (CHjCgH^Hj)^^ + 2X -1 H20
The physical properties of 4,4'-MDA follow:
Molecular weight . 198.3
Boiling range (768 mm Hg), °C 398-399
Melting point, °C 91-92
Density at 100°/4°C 1.056
Viscosity at 100°C (cP) 8*04
Flash point, °C 221.1
Fire point, °C 248.9
Heat of vaporization, kJ/mole (kcal/mole) 95.4 (22.8)
Specific heat at 29°C (solid), J/(g°C) [cal/(g °C)] 1.46 (0.35)
Specific heat at 109°C (liquid), J/(g °C)
[cal/(g °O] 2.01 (0.48)
Heat of fusion, kJ/mol (kcal/mole) 19.6 (4.7)
Log P 1.76-2.52
Approximate solubility, g/100 g of solvent at 25° C
Acetone 273.0
Benzene 9.0
Carbon tetrachloride 0.7
Ethyl ether 9.5
Methanol 143.0
Water 0.1
Source: Moore (1978), Windholz (1976), as reported in ESE (1981).
-------
8
III. HEALTH EFFECTS
Information relating to the ability of 4f4'-MDA to be
absorbed by mammals and then to interact with genetic material
will be presented in this chapter along with summaries of studies
on animals aimed at determining the chemical's carcinogenic
potential. Additionally, information from an epidemiology study
and a structure-activity analysis will be given. Then all the
evidence relating to the carcinogenic potential of 4,4'-MDA will
be pulled together in .a weight-of-evi'dence summary.
Since the purpose of this assessment is to investigate the
cancer risks associated with 4,4'-MDA, the principal thrust of
the health effects review given here will be on that topic.
However, it should be noted that 4,4'-MDA has been recognized as
a causative agent in acute liver toxicity in humans and animals
(NIOSH, 1976a, 1976b). Liver toxicity in humans has been
observed following oral exposure in the so-called "Epping
Jaundice" incident in which people in Epping, England consumed
bread made with flour contaminated with 4,4'-MDA (Kopelman et
al. , 1966a). Kopelman (1966b) reported a 2-year follow-up study
on 43 of the 84 persons known to have suffered injury in the
case. While no evidence of progressive liver disease was seen,
some patients reported fat, fruit or alcohol intolerance, weight
loss or other troublesome symtoms.
Dermal exposure to 4,4'-MDA also has been associated with
acute liver toxicity. McGill and Motto (1973) and Williams et
al. (1974) reported on dermal exposures in the workplace which
resulted in liver toxicity. Additional information on these
incidents is given in Section IV.
-------
Retinopathy has also been cited as a toxic effect (Schilling
v. Constatt _et^_al_. , 1966; NIOSH, 1981; Leong _e_t _al_. , 1984), as
has acute myocardiopathy (Brooks et_ a±., 1979) and allergic
derraititis (Emmett, 1976). Retinopathy has been observed in the
cat and the guinea pig and may have occurred in humans, while
myocardiopathy has been seen in humans.
A. Absorption and Distribution (Thies, 1983)
1. General
Tor tore to ot_ jal_. (1961) administered pure (100%) or
technical grade (56%) 4,4'-MDA to male BecSFj^ mice
intraperitonealy in corn oil at the maximum tolerated dose of 250
mg/kg. Animals were examined at intervals between 5 minutes and
12 hours after administration'of 4,4'-MDA for distribution of the
chemical to the blood, liver, kidneys, lungs and spleen.
These workers found that 4,4'-MDA is rapidly absorbed and
reaches peak concentrations in the blood between 10 and 20
minutes after administration. They found the half-life in blood
to be about 6 hours and the rate constant for the beta
(elimination) phase to be about 2 X 10"^ minutes.
They found that 4,4'-MDA is distributed preferentially to
the liver and kidneys arid is eliminated from all examined organs
at similar rates. No notable difference in distribution was seen
between the pure and technical grades.
This study shows that 4,4'-tMDA is rapidly absorbed and well
distributed and that impurities up to 44% have negligible effect
on blood partitioning of the chemical. The liver appears to be
the organ receiving the greatest systemix; dose.
-------
10
2. Inhalational and Dermal Absorption
For the purposes of this analysis, the structural analogue,
methylenebis (ortho-chloroaniline) (MBOCA) , is used to predict
the behavior of 4,4'-MDA while more data on the latter are being
obtained. MBOCA is a good structural analogue for 4,4'-MDA, and
-it has similar physicochemical properties so that it can be used
to make inferences regarding the behavior of 4,4'-MDA.
Some physicochemical properties of 4,4'-MDA and MBOCA
follow:
4,4' -MDA MBOCA
CAS No.: 101-77-9 CAS Ho.: 101-14-4
Molecular Weight: 198.26 259.06
Log P: 1.84* 1.38**
(estimated): 4.89 3.13
It is generally recognized that compounds which exhibit some
degree of lipid and water solubility, which are predominately in
an un-ionized state at physiological pH values, and which have
sufficiently low molecular weight (less than 500) will tend to be
more easily absorbed.
a. MBOCA Data
i. Percutaneous Penetration
Several studies show that MBOCA binds to and penetrates
through human and animal skin.
Chin_e_t^l_. (1983) demonstrated in vitro percutaneous
absorption by, accumulation in, and penetration through neonatal
human foreskin by radiolabelled MBOCA.
* See Section IV.
** Glowinski et al. , 1978.
-------
11
Workers exposed during the manufacture of MBOCA were shown
by Linch jet _al_. (1971) to demonstrate measurable levels of the
chemical in urine.
Radiotracer studies by several groups (Braselton et al.,
1982; Groth as reported by Morton, 1981; Tobes j§_t _al_. , 1983)
using 14C labelled MBOCA demonstrated that MBOCA penetrates human
and animal skin.
Based, on these studies, which, show a range of skin
penetration rates from about 0.13 to about 8% of the applied dose
per hour, a logarithmic mean of 1% absorption per hour can be
used as the dermal penetration rate of MBOCA for humans (Beal,
1982).
ii Gastro—Intestinal Tract Penetration
———————-—————————————— \
MBOCA, its metabolites, or both, penetrate through gastro-
intestinal tract tissue.
Stula ot_ al. (1975 and 1977) and Kommineni _e£ _aJL. (1979)
conducted long-term feeding studies in mice, rats, and dogs that
resulted in production of malignancies in the lungs and urinary
bladder, demonstrating penetration of MBOCA through gastro-
intestinal tract tissue.
Oral LDcQ studies in male rats by Miller and Sherman (1965)
and Reinke (1963) demonstrated that MBOCA, its metabolites, or
both, penetrate gastro-intestinal tract tissue.
Recent single- and multiple-dose pharmacokinetic studies in
rats by Morton _e_t _al_. (1981) and Groth (1981., as cited by Morton
et_ al_., 1981) showed that MBOCA penetrates gastro-intestinal
tract tissue.
-------
12
Barnes (1964) reported profound effects on the blood of dogs
orally exposed to MBOCA, and demonstrated recovery of MBOCA and
its metabolites in urine, thus showing gastro-intestinal tract
absorption.
After oral administration of MBOCA, the major route of
elimination is via the feces, which may indicate that the gastro-
intestinal tract does not completely absorb MBOCA or its
metabolites, or that absorption is followed by some degree of
biliary cycling as shown by Morton'_e_t _al_. (1981).
iii. Respiratory Tract Penetration
No test data are available regarding the penetration of
MBOCA or its metabolites through respiratory tract tissue. But
based on the knowledge that MBOCA penetrates other biological
membranes and its physicochemical properties, it can be assumed
that if MBOCA reaches the alveolar regions of the lung, it will
be absorbed, but to an unknown extent.
A critical factor in estimating the potential for
penetration of substances through lung tissue is particle size.
Particles of MBOCA (or 4,4'-MDA) greater than 5 microns in size
will not be expected to reach the alveolar regions where
extensive absorption could occur; however, such material could
still reach systemic circulation due to the generally capillary-
laden nature of the nasopharyngeal region. Particle sizes
greater than 2, but less than 5 microns will deposit in the upper
respiratory tract, where very little may be absorbed via alveolar
or capillary diffusion; most will be cleared by mucocilliary
movement with subsequent swallowing, and^absorption via gastro-
intestinal tract tissue.
-------
13
Particle size less than 2 microns stand the best chance of
reaching the alveolar regions and, depending on lipid solubility,
crossing the alveolar membrane. MBOCA or 4,4'-MDA in the vapor
phase can be expected to penetrate lung tissue fairly well. This
conclusion is based on knowledge that MBOCA can cross other
biological membranes and on its solubility characteristics. It
is assumed that this penetration may reach at least 50% of the
inhaled material. This assumption is based on gastro-intestinal
absorption of radiolabelled MBOCA in rats of approximately 30%
(Morton _e_t_al_. / 1981).
b. Inferences About 4,4'-MDA
As previously stated, MBOCA is an acceptable analogue to
4,4'-MDA and provides adequate information to conclude that 4,4'-
MDA has a high potential to penetrate biological membranes.
NTP bioassays (NTP, 1983a), on 4,4'-MDA dihydrochloride
indicate the probability that 4,4'-MDA itself will cross gastro-
intestinal tissue. In these studies, rats and mice received the
dihydrochloride salt of 4,4'-MDA in drinking water. The
dihydrochloride moieties would be expected to dissociate rapidly
in the gastro-intestinal tract, leaving 4,4'-MDA as the essential
toxicant. From the NTP studies it is clear that orally
administered dihydrochloride of 4,4'-MDA is a carcinogen in rats
and mice. This provides indirect evidence that 4,4'-MDA, its
metabolites, or both, penetrated gastro-intestinal tract tissue.
while 4,4'-MDA would be almost entirely un-ionized in the
gastro-intestinal tract, it is difficult to support the concept
of 100% absorption, especially when dose levels are not defined
-------
14
by pharmacokinetic studies. Furthermore, Morton et al. (1981)
demonstrated that the major route of MBOA excretion was via the
feces, which may indicate incomplete absorption from the gastro-
intestinal tract. , It can probably be assumed that gastro-
intestinal tract absorption of 4,4'-MDA may be approximately 50%.
The assumption that absorption through lung tissue is
roughly equivalent to gastro-intestinal absorption is plausible,
especially if 4,4'-MDA is in the vapor phase or has a particle
size of less than 2 microns.
In addition to the inferences about 4,4'-MDA's ability to be
absorbed by humans that are based on structural analogy with
MBOCA, further direct evidence for 4,4'-MDA's absorption
potential is given by Vaudaine et al. (1982), Williams et al.
(1974), McGilT and Motto (1974), Dunn and Guirguis (1979) and
Brooks (1979). These workers reported that 4,4'-MDA was found in
the urine of industrial workers who were exposed to the chemical
(Vaudaine et_ _al_. ) or that various toxic manifestations of
exposure occurred in industrial settings. Further information is
given in Section IV. Additional evidence of absorption of 4,4'-
MDA by humans is given by Kopelman et_ _al_. (1966a, 1966b) , who
reported on cases of liver toxicity in people who had eaten bread
contaminated with the chemical.
c. Limits of this Analysis
As with rnost structure activity relationship analyses,
conclusions must be made with the understanding that nothing is a
good substitute for actual testing. while MBOCA appears to be a
good analogue, it is not known what specific effects the two
-------
15
chlorine molecules may have on penetration characteristics. The
strongest conclusion that can be inferred based on the available
information is that 4,4'-i4DA crosses biological membranes.
Quantitative estimates of the"rate of this absorption require
testing, and such testing is being done in an EPA-sponsored
dermal penetration study using radio-labeled 4,4'-MDA.
For the purpose of this assessment, a dermal absorption rate
of 1% per hour of deposited material is assumed, based on MBOCA
data.
B. rtutagenicity
4,4'-MDA induces mutation in the Salmonella typhimurium/
mammalian microsomal assay (Ames assay), induces sister chroraatid
exchanges (SCE) in femoral bone marrow of male mice and binds
covalently to DNA in vivo in the livers of treated mice. It does
not induce chromosomal aberrations in vitro in human peripheral
lymphocytes and is reported to be negative in a Drosophila
melanogaster sex-linked recessive lethal assay, the details of
which were not available for review and whose validity,
therefore, cannot be assessed.
Since the human peripheral leukocyte assay does not measure
the same endpoint as the Ames and SCE assays, a mixture of
positive and negative results is not unusual, and the positive
results seen in the latter are, thus not negated by the negative
results seen in the former. Although not a test for genotoxicity
per se, covalent binding to DNA demonstrates that 4,4'-MDA is
capable of interaction with macromolecules in vivo. Such binding
may be indicative of the ability of this"^ agent to induce
-------
16
mutations and cancer. Taken together the weight of evidence from
all these studies supports the conclusion, reached on the basis
other evidence too, that 4,4'-MDA is an oncogen.
4,4'-MDA has repeatedly been found to be mutagenic in tests
with b. typnimurium strains TA-98 and TA-KJO with metabolic
activation (Godek et_ al_. , 1982; Rao ^t_ _al_. , 1982; Parodi et al. ,
1981; Darby _e_^ a±. , 1978; and Brusick, 1976. While Brusick
(1976) reported positive results with metabolic activation in a
test on strain TA-1538, other workers have found that 4,4'-MDA is
not active in that strain or in strains TA-1535 and TA-1537,
either with or without metabolic activation.
The Ames assay is known to correlate with in vivo
oncogenicity with approximately 81% reliability. In a review
performed for the Gene-Tox Program (EPA, 1933), 122/151 (81%) of
the oncogens which were tested in the Ames assay were correctly
identified.
Parodi _e_t _al_. (1983) reported that 4,4'-MDA induced SCE in
the femoral bone narrow of male mice. Although SCE formation
cannot be used as a quantitative measure of carcinogenic potency,
it can be used as a qualitative indicator of potential in vivo
oncogenicity. In the Gene-Tox Review referred to above, 17/17
(100%) oncogens were correctly identified in this assay. While
adrnitedly a limited data base, it does appear that the in vivo
SCE assay is a sensitive indicator of potential in vivo
oncogenic ity.
Pantarotto (1983) and Parcdi _e_fc _al_. (1981) have both
reported that 4,4'-iv1DA. is capable of covalent binding in vivo to
-------
17
mouse liver DNA. While not a test for genotoxicity per se, these
results do show that 4,4'-MDA is capable of reaching target
tissue in vivo and once there of interacting with cellular
macro-molecules in a manner which may lead to mutations or cancer.
Although Ho et_ jaJ^. (1979) reported that tfDA did not induce
sex-linked recessive lethal mutations in Drosophilay the cited
study is an abstract with no experimental data. In the absence
of such data, no conclusions can be drawn about the validity of
the study.
Nunziata (1983) reported that 4,4'-MDA did not induce
chromosomal aberrations in vitro in cultured human lymphoctes.
The negative results in this study do not lessen the weight of
evidence presented by the studies cited above.
4,4'-MDA induces gene mutation, SCE, and DNA damage. Gene
nutation and chromosomal aberration are separate endpoints. It
is not unusual for a chemical to induce one but not both of these
enopoints. Based upon the results of the studies summarized
above, 4,4'-MDA should be considered a probable oncogen.
C. Carcinogenic ity
1. Human Data (Scott-Siegel, 1984)
The National Institute of Occupational Safety and Health
(NlOSh) conducted a health hazard evaluation of workers employed
by the Boeing Vertol Company, a manufacturer of helicopters and
helicopter parts (NIOSH, 1983). NIOSH evaluated exposures in two
buildings, the blade and pattern shop, where exposure to 4,4'-
MDA, epoxy resins, and solvents (toluene, methyl isobutyl ketone,
and eyelohexanone) were known to occur. "A medical evaluation of
-------
18
workers who had been employed in these areas was also done in two
parts: a medical study of dermatologic conditions and an
epidemiologic evaluation — a proportionate mortality (PMR) and
proportionate cancer mortality (PCMR) study. Among a group of
179 "exposed" worker deaths, NIOSH observed in the PMR study
statistically significant elevations in mortality from cancer of
the bladder, large intestine, and lymphosarcoma/ reticulosarcoma.
The elevation in bladder cancer mortality remained significant in
the PCMR study. This epidemiologic evaluation of the pattern and
blade shops suggests the existence of an increased risk of
mortality from bladder cancer. NIOSH, additionally, conducted
environment monitoring in the pattern and blade shops to
determine current exposure. They found detectable levels of
4,4'-MDA, epoxy resins, and solvents (See section IV).
Analysis of deaths in the "exposed" group show significant
elevations in site-specific mortality from cancer of the bladder
(PMR=374, 3 observed, p<0.05), large intestine (PMR=226, 7
observed, p<0.05), lymphosarcoma/reticulosarcoma (PMR=343, 3
observed, p_<_0.05), and skin (PMR=343, 3 observed, p<_0.05). No
significant excesses in mortality were observed for non-
neoplastic sites. In the PCMR analysis, the excess in bladder
cancer mortality remained significantly elevated (PCMR=341,
p<0.f)5).
Apart from the mortality studies, interviews with living
current and living former employees who worked in either
"exposed" areas or "non-exposed" areas were conducted. Two
additional bladder cancer cases were found in a living current
-------
19
and a living former employee. Both employees had worked in
either the pattern or blade shop, where measurable levels of
4,4'-MDA were detected. A history of cancer was reported by no
workers in a comparison group who worked in packing and storeroom
areas with virtually no exposure to 4,4'-MDA.
PMP and PCMR analyses measure the relative frequencies of
different causes of deaths occuring in a study population.
Limitations of the PMR and PCMR study designs are that they
provide no information on the total force of mortality, the
ratios across diseases are not independent, and the distribution
of deaths not included in the analysis may differ from the
distribution of those included. Proportionate analyses are most
effective when tne disease of interest is relatively rare. Thus,
the above biases may be reduced when a PCMR analysis is employed.
LiKewise, the ability or power of a particular study to
detect an increase in risk depends upon several factors. These
tactors are sample size, magnitude of the increased risk,
background incidence of the disease, desired statistical
significance, and type of analysis. Thus, a study might not
observe significant increases in mortality, when in fact an
increased risk exist, if any one of the above factors is
unfavorable to the study.
Given the discussed design constraints, the epidemiologic
evaluation of the blade and pattern shops suggests the existence
or an increased risK of mortality from bladder cancer. This
finding confirmed the a priori concern for bladder cancer based
on 4,4'-MDA's stuctural analogy to benzidine, a known human
-------
20
bladder carcinogen. However, since two of the three bladder
cancer deatns had latencies of 10 years or less, one cannot
completely discount the possibility of these deaths being a
random cluster, unrelated to 4,4'-MBA exposure.
Exposure monitoring revealed detectable levels of 4,4'-MBA,
methyl isobutyl ketone, toluene, cyclohexanone, and butyl
glycidyl ether. All were found below OSHA and ACGIH standards,
but NIOSH reported that it appeared as though former work
practices may have resulted in exposure levels higher than those.
measured in this study (NIOSH, 1983).
Epoxy resins and 4,4'-MDA have been shown to cause cancer in
laboratory animals. Separating the influences of these two
exposures to identify an etiologic agent of the excess bladder
cancer would be difficult.. However, as NIOSH pointed out in
their conclusion, the following factors lend weight in
implicating 4,4'-MDA as the etiologic agent: 1) detectable
airborne levels of 4,4'-MDA and suggestion of greater exposures
in the past (dermal contact was also a route of exposure); 2)
known toxicological evidence; 3) similarity of 4,4'-MDA to known
human bladder carcinogens such as benzidine; 4) the bladder site
was of concern a priori; and 5) the fact that there were cases of
bladder cancer among living employees with longer exposures and
compatible latency.
In summary, while the confounding exposure to epoxy
compounds and the study, design render this epidemiology work
inadequate for establishing a causal relationship between
exposure to 4,4'-MDA and increased incidence of bladder cancer,
-------
21
tne data are suggestive of such a relationship. This is
especially so because of 1) evidence of higher exposure levels
existing at the Boeing plant in prior years; 2) the close
structural relationship between 4,4'-MDA and benzidine -- a
chemical known to produce bladder cancer in humans; and 3)
4,4'-i*iDA caused urinary bladder tunors in rats in the NTP
bioassay. The evidence presented in this section, along with
that in tne following section which is adequate to establish
4,4'-MDA's carcinogenlcity in animals, leads to classify the
chemical as a probable human carcinogen, (B2).
2. Aninal Data
a. NTP Studies
The Mational Toxicology Program tested 4,4'-iiiDA
a inycirocnioriae ( 4 , 4 '-inDA* 2HC1) for carcinogenicity by
anninistering the chemical in drinking water to both sexes of
Fischer 3*4 rats and B6C3K1 mice (NTP 19fl3a). Fifty rats and
mice of each sex received drinking water containing 150 ppn or
SOU ppm 4,4'-iMDA*2HC1 for 104 weeks, ad libitum. Fifty controls
tor each species and sex received no 4,4'-MDA*2HC1.
These studies were selected as the basis for a quantitative
estimate of possible human risks of contracting cancer from
exposure to ^,4'-MCA because .1) the design is far superior to the
otiiers cited in this section, and 2) the results lend themselves
"/ell to statistical analysis. Statistically significant,
treatment-related increases in malignant and non-malignant
tuiaors, including several rare tumor types, were seen at multiple
sites in both sexes cf both species (Mil&an, 1984).- A discussion
-------
22
of the statistical analysis of the results of these bioassays is
given in Section V of this assessment. A summary of the results
is given here.
Results observed included compound-related, non-neoplastic
lesions of the thyroid in both sexes of rats and mice including
follicular cysts and hyperplasia. An increase in the incidence
of thyroid neoplasms was observed in the high-dose groups
compared to control groups for both sexes of both species. Liver
degeneration was observed in 80% of the low-dose and 60% of the
high-dose male mice, but not in controls.
Thyroid follicular-cell carcinomas were seen in male rats at
rates of: 0/49 control, 0/47 low-dose, and 7/48 high-dose.
Combined thyroid follicular-cell carcinomas and adenomas in male
rats occurred with incidences of: 1/49 control, 4/47 low-dose,
and 10/48 high-dose. Liver neoplastic nodules occurred in male
rats with incidences of: 1/50 control, 12/50 low-dose, and 25/50
high-dose. In female rats, combined thyroid follicular-cell
carcinomas and adenomas showed incidences of: 0/47 control, 4/47
low-dose, and 19/48 high-dose. These animals also showed
incidences of thyroid C-cell carcinomas and adenomas of: 1/47
control, 5/47 low-dose, and 7/48 high-dose. Male mice displayed
incidence of liver -hepatocellular carcinomas of: 10/49 control,
33/50 low-dose, and 29/50 high-dose. Combined incidences of
liver hepatocellular carcinomas .and adenomas in male mice were:
17/49 control, 43/50 low-dose, and 37/50 high-dose. In male mice
adrenal pheochromocytomas were seen at rates of: 2/48 control,
12/49 low-dose, and 14/49 high-dose. Female mice had lung-
-------
23
alveolar/bronchiolar adenomas and carcinomas at rates of: 2/50
control, 3/50 low-dose, and 8/49 high-dose. Malignant lymphomas
occurred in these animals with incidences of: 13/50 control,
28/50 low-dose, and 29/50 high-dose. Liver hepatocellular
carcinomas were seen in female mice at rates of: 1/50 control,
6/50 low-dose, and 11/50 high-dose. Combined liver
hepatocellular carcinomas and adenomas occurred in these mice at
rates of: 4/50 control, 15/50 low-dose, and 23/50 high-dose.
And, finally, thyroid follicular-cell adenomas and carcinomas
occured in female mice at rates of: 0/50 control, 1/47 low-dose,
and 13/50 high-dose.
Several extremely rare tumor types, the probability of whose
spontaneous occurrance is discussed in Section V, were also
observed. These include one bile duct adenoma in a male rat
(none previously diagnosed in 3,633 NTP controls), transitional
cell papillomas of the urinary bladder in three female rats (3
previously diagnosed in 3,644 NTP controls), and granulosa-cell
tumors, including one carcinoma in five female rats (11 such
tumors and 1 such carcinoma previously diagnosed in 3,462 NTP
controls). , . .
b. Hiasa et al. Study
Hiasa ^t^ al_. (1984) treated male inbred Wistar rats i.p.
with a subeffective dose of N,N-bis(2-hydroxypropyl) nitrosamine
(DHPN) namely, 280 mg/100 g body weight followed by a diet
containing 1,000 ppm 4,4'-MDA for 19 weeks. The incidence of
follicular-cell carcinomas of the thyroid was 2/21 (9.5%) and
•.
that of total thyroid tumors was 19/21 (90%). DHPN alone
-------
24
produced only 6/21 (28%) thyroid tumors, and these were all
beniyn neoplasms, while rats receiving 4,4'-KDA alone or saline
controls had no thyroid tumors at the end of the study. The
authors concluded that 4,4'-MDA promoted the action of DHPN, an
initiator of thyroid tumors in this test. Similar results were
seen by these authors with phenobarbital and barbital, known
promoters of hepatocarcinogenesis, and with 3-amino-l,2,4-
triazole a goiter-causing agent (Hiasa _et^ _al_. 1982a, 1982b).
It should be noted that 4,4'-MDA is also a goiter-inducer
(MTP, 1983a). .Thus tne findings of Hiasa et al. in this limited
bioassay of 4,4'-MDA are consisent with the findings of studies
>
on other known tumor promoters and goiter-causing agents.
c. Deichmann Study
Deichmann _e_t _al_. (1978) studied the carcinogenic potential
of purified 4,4'-MDA and crude 4,4'-MOA by oral administration to
dogs, t-iine female beagle dogs, five to six months of age,
received doses of 70 mg by gelatin capsule three days a week
until deatn or termination of the study at 86 months. No control
groups were reported. Study parameters included body weight
gain, cystoscopic examination of the urinary tract (started after
two years of dosing and made at 15-month intervals until
tsrmination), and serum biochemical tests, which included fasting
blood sugar, blood urea nitrogen, creatinine, uric acid, albumin,
total protein, and alkaline phosphatase activity.
Moderate to severe histopathological changes occurred in the
livers of all dogs. These changes included "swollen hepatic
cells, moderate disruption of the lobular patern, passive
-------
25
congestion, fatty infiltration distorting the lohular pattern,
hepatic cell degeneration,' Aortal fibrosis, central zonal fatty
degeneration with hepatic cell necrosis, heraosiderosis, and
lymphoreticular cell infiltration of the portal areas, and bile
ducts distention." Less severe changes were observed in the
kidneys, spleen, and lungs. The authors concluded that purified
and cruae 4,4'-HDA produced no tumors in the urinary bladder or
liver of any dog.
The small number of animals, the lack of controls, and the
less-than-lifetime duration of the study render this study of
limited value in assessing carcinogenic risk.
d. Schoental Studies
Schoental (1968a anc 1968D) prepared a purified 4,4'-MDA
sample from an epox.y resin hardener containing 54% 4,4'-!"iD.A
dissolved in ganma-butyrolactone. The epoxy resin hardener was
known to have contaminated flour used in bread that had oeen
eaten by tue "Lpping jaundice" patients (see above) (Kopelman et
al. , i^bba and ly^fib). Two to five 50 mg doses of 4,4'-HDA in
arachis oil were administered by stomach tube to eight male and
eignt female weanling rats (strain not specific), weighing 45 to
60 g.
Ail rats were maintained until they became ill or died, at
which time they were autopsier!. A liver hepatoiua with many cells
in mitosis and a kidney tumor having a hemangioma-like appearance
were observed in one nale that received five doses. An
adenocarcinoma of the uterus was found in one female 24 months
after the first dor?e. Pituitary tumors were observed in five
-------
26
females and one male. Non-neoplastic lesions of the kidneys,
liver, and lungs were found in most rats.
Twelve white CFW mice, about one month old (sex not
specified), were given a single subcutaneous injection of 5 to 10
my 4,4'-«iDA in arachis oil (50 mg/ml) (hchoental 1968b). Seven
nonths later one mouse "was killed. No tumors were found. The
remaining 11 IP ice were given a second dose of 5 mg 4,4'-MOA in
arachise oil by stomach tube. Nodular hyperplasia of the liv<3r
and a possible nepatoma were observed in one mouse that was
killed when it was 18.5 months old. Wo other lesions were
considered to be compound-related.
Tiiese studies are of extremely limited value for assessing
cancer risks. The number of animals used was small and the
dosing was for a very snort period.
e. Griswold Study
Griswold _et_ _al_. (196H) evaluated the carcinogenic potential
of 3.b aromatic ainin.es including 4,4'-;-1D£. The induction of
mammary gland tumors in female Sprague-Dawley rats was the test
system. Twenty female Sprague-Oawley eats, 40 days of age,
received 10 mg doses of 4,4'-MDA in 1 ml sesame oil by stomach
tube at three-day intervals for 30 days, then were observed for
an additional eight months without treatment. A negative control
:jroup of 140 rats received sosane oil, and a positive control
^roup of *HJ rats received a single dose of 1H mg of 7,12-
diiv,ethylbenz [a] antnracene (Df'i.&A). 4,4'-HDA did not induce tumors
in this test.
-------
The limited number of animals, the short duration of the
study, the low number of doses administered and the lack of a
complete pathology study of the animals riake this study of low
value for assessing carcinogenic risks.
f. Fukushima Studies
Five Croups of taale Wistar rats were fed 1,000 ppm 4r4'-MDA
in the diet for 3, 16, 26, 32, or 40 weeks (r'ukushima et al.,
1979). Within eacn group, three to six rats were killed at
eight-week intervals after cessation.of 4,4'-MDA feeding until
termination of the study at 40 weeks. A control group (eight
rats) received niet alone for 40 weeks. One rat from the control
group and one from each treated group were killed on the first
day of the study. ^io tumors were reported in any organ.
The less-tnan-lifetime duration of this study, along with
tne limited pathology information reported, hinders its use in
. assessing carcinogenic risk.
Fukushima _et_ ^al_. (1977) studied the effect of prior
administration of 4,4'-iviDA on colon cancers produced by 1,2-
aimethylhydrazine (DMH) in rats. Two groups of eight-week-old
male Wistar rats were fed 4,4'-MDA (concentration unspecified) in
their diets for eight weeks. After eight weeks, one group (24
rats) received 4,4'-piDA-containing diet only. The second group
(30 rats) received 12 weekly subcutaneous injections of 10 m.g/kg
U'ir), starting two weeks after d,4'-MDA was removed from the
diet. A third group (24 rats) received untreated diet and 12
weekly suncutaneous injections of 10 rag/kg DMH, starting at 13
weeks of age. £ control group (13 rats)-received only untreated
-------
28
diet. All rats were killed 52 weeks after the first injection of
DMH (70 weeks of age).
Tia 4,4'-MDA-treated groups showed evidence of liver
toxicity but tumorigenic activity was not demonstrated.
The design of this study, especially its limited duration
and enapoint, limits its utility in assessing cancer risks
associated with exposure to 4,4'-MDA alone.
g. Steinhoff and Grundraann Study
Steinhoff and Grundraann (1970) reported that 4,4'-MDA was a
weak carcinogen when injected subcutaneously into rats. A group
ot 25 nale and 25 fenale Wistar rats (age unspecified) received
subcutaneous injections of 30 to 50 mg/kg 4,4'-MDA at intervals
ot one to three weeks, up to a total of 1.41 g/kg. A control
group (specifics not given) received subcutaneous injections of
saline.' The rats were maintained for their lifetime. The mean
lire span of male 4,4 ' -MDA-treated rats was 970 days, for fpir.ales
l,ubO days, ana for controls "(sex not specified) 1,007 days. The
/
4, 4'-HLj.A-treated rsts had 33 malignant and 29 benign tumors and
four hepatomas, while the control had 16, 15, and 0,
respectively, statistical analysis was not provided. The
authors concluded that 4,4' -i>!DA nad weak carcinogenic activity.
wnile carcinogenic activity was observed in this study, the
sraali number of animals useo, the shcrt treatment time, and the
lack of complete pathology information limit its utility in. a
quantitative estimation of rissc.
-------
29
h. Zylberszac Study
Zylberszae (I9bl) reported cirrhosis of the liver after
implantation of crystals of 4,4'-MDA under the skin of 25 male
rats of mixed strain, weighing between 150 and 220 g. Each
.inplanation was 25 rig, and seven implantations were made in five
months. The time of sacrifice was not specified. The surface of
tne livers appeared granular and nodular, but no hepatomas were
found.
The route of p»xposurs and limited information on study
execution make this study of little value.
i. Munn Study
Munn (1967) administered 4,4'-MDA dissolved in arachis oil
by ydvaye in nalo rats five days per week for 121 days,
delivering a total of 330 rig/100 g. Of 24 treated rats, three
•••/ere lost cnrouyh cannibalism. Of the remaining 21 animals, 12
survived wore than two years. All animals had cirrhosis of the
liver, but none developed turaors during the first two years of
che study. Two rats, however, killed at 792 days and 947 days,
respectively, had hepatomas, while a variety of miscellaneous
tumors (not specified) were observed in older animals.
A second experirient was periormsd (neither dosing regimen
nor number of animals specified) during which treatment continued
ror Id nonths. The total cJose ct 4,4'-MPA per animal averaged
600 mg/iOO g.
Two liver tumors were obs^rvec1. on?; tumor (unspecified)
occurred at the end or the tirst year, while the second appeared
more than two years after treatment began. In addition, one
-------
30
intestinal turaor, one pituitary tumor, and two subcutaneous
tibromas were observed more than two'years after teatment began.
The small number of. animals used and incomplete pathology
information limit the utility of this study in quantitative risk
assessment.
j. Gohike Study
Gohlke (1978) conducted a chronic study on the oncogenic
effects of 4,4'-MDA. Male albino rats were divided into four
groups of 120 each and treated as negative controls, positive
controls, or with 4,4'-iwDA (H or 20 mg/ky) by stomach tube five
days/week tor 16 weeks. Ten animals from each group were killed
after ten days and six weeks or treatment, or at termination (16
v/eeKs). In each group, 56-68 additional animals were examined at
their natural death.
In another experiment, 50 animals were divided into two
-jroups ot 30 eacn and used as positive controls, or given 3.2
rrn-j/ky 4,4'-Mf.!A, five days/week for 16 weeks. After six weeks of
treat/merit, at termination, ot treatment (16 weeks), and at 20
weeks, between four and nine animals were killed from both the
control and. 4 , 4 ' -r-iDA-treated groups' for histological examination.
i-io differences were observed between livers of animals given
4,4l-f-iDA at 3.:> mg/kg and controls at up to ten weeks of
exposure. At R p.g/ky, increased hepatocyte mitoses were seen
atter ten clays ot exposure. After six weeks of exposure to 8
mg/kg, 50% of the animals showed hepatocyte swelling with
disseminated isolated fatty degeneration, increased mitoses, and
enlarged nuclei. Increased mitoses also occurred in the bile
ducts. At 16 weeks, no changes were reported.
-------
31
At 20 mg/ky, a decrease in glycoyen in all animals,
proliferation in bile ducts, and large vesicular.'bile duct nuclei
were ooserved after ten days of exposure. After six weeks, 50%
of the animals showed the following effects: increased
nepatccyte and bile duct mitoses; hepatocytes with triple nuclei;
bile duct proliferation with large vesicular nuclei; expanded
fibrous connnective tissue in the portal fields; and septal
interlobular and intralobular connective tissue bridges. After
16 weeks of treatment, bile duct proliferation and the connective
tissue changes were still apparent in 50% of the animals.
Hyperplasias developed after four months and were present
mainly in 4,4'-MDA-treated animals. These were benign, localized
reticular hyperpiasias, benign hepatornas, excessive bile duct
proliferations, and benign vascular tumors of the liver.
ot a total of 437 animals examined, 17 developed 16 tumors
and six systemic diseases, waile seven developed partly tumor-
iike hyperplasias.
The life-span of the aninals (groups unspecified) used in
these experiments was low cue to pulmonary and otogenic
inflammation tnat extended into the meninges. tThe average age at
death was ll.J montns (11.3 months in 4, «i'-MDA-treated, 12.5
montns in controls). All four animal groups had similar growth
curves.
This study clearly demonstrtes- tne hepatotoxicity of 4,4'-
MUA, r>ut the short duration of exposure (16 weeks) puts a severe
limitation on the use of this study for quantitative risk
estimation.
-------
32
3. Structure-Activity Relationships
Evidence from NTP/NCI studies suggests that compounds
belonging to the class of bis-benzenanines separated by -£#2' ~°
or -s are carcinogenic for the tnyro.id gland of rodents. As can
be seen 'in Table 1, the National Cancer Institute tested three
additional analogues of 4,4'-MDA [CI i973b).
Tite carcinogenicity of this class of compounds for the
thyroid gland is of particular interest since these chemicals
b^ar a structural resemblance to the thyroid hormones
triioootnyronine and thyrcxine shown below. Thus 4,4'-lMDA may be
upsetting the hormone balance in tne thyroid or be interfering
with the gland's nornal functioning in some other way.
-------
OSi UP RESULTS OF CHBDN1C NCI/fcflP STUDIES ON 4,4t-HDA
2ND KEIATED CDHFDUNEb
Test substance Structure Species
4,4' -!"!L& Rat
(Current btuay) A^^. ^^ (F344)
H2N-/C jVcH2Af JV-NH2 M^tl
Vs-"»/A Vs — 'V
4,4' -iviethy lenebis Wat
( N , iv-O iiuB thy i ) ( F3 44 )
oenzencffiUPe ivicuse
(NCI, iy?3a) ^"^ /'/^^V (6&C3F1)
(CH3)2N-/f )VcH2-Hnr JN-N(CH3)2
vLer ' s rietone i"NH2 '"C3n)
Sex
M
F •
M
F
M
F
M
F
j
C1
i-! '
F
ivi
r
M
K
M
f
H
H'
Dose
(ppm)
300(a)
300
300
300
750(0
750
2500
2500
50(j(c)
1000
2500
2500
50n(c)
500
800
800
3000 (c)
3000
5000
5000
Site of Neoplast:
Lesion Observed
Liver Thyroi^
M(b> N
M
N N
Cyl
Ni
N
N
N
N
b! N
i\i t\'
1M LV
N M
N
w N
N N
In
water.
N - isiecplastic lesion ocojrrecl at statistically significant incidence (P<0.025 by the Pish
exact test.
in reed.
-------
34
NH-
I
-CH2CHC02H
:H2CHC02H
I " . I
Triiodothyronine
The tnyroia gland is also the target organ for non-
neoplastic effects of bis-benzenanines. Kor example, 4,4'-MDA,
<±, 4 '-oxyaianiline , * , 4 ' -thior. ianiline and 4 ,
-------
35
o With the exception of Michler's ketone, all the
tested analogues of the class of bis-benzenaminesf
including 4,4'-MDAf are carcinogenic for the thyroid
gland of mice.
o All the tested analogues of the class of bis-
benzenamines, including 4,4'-MDA, are carcinogenic
for the liver of rats and mice.
o ,4,4'-MDA is also structurally related to benzidine
and 4,4'-methylene bis(2-methylaniline). All three
compounds are associated with cancer of the urinary
bladder in rats (i.e., 4,4'-MDA) or humans [i.e.,
benzidine, 4,4'-methylene bis-(2-methylaniline) and,
possibly, 4,4'-MDA].
o All the tested analogues of the class of bis-
benzenamines, including 4,4'-MDA, also produce non-
neoplastic effects in the thyroid gland (i.e.,
follicular-cell or papillary hyperplasia).
Additionally, at least two analogues, including
4,4'-MDA, are goiterogenic in mice and rats.
4. Summary of Animal Data
There is sufficient evidence from the NTP bioassays, other
whole animal studies, mutagenicity studies, and structural
analogy with other compounds that have been shown to be
carcinogenic in animals to classify 4,4'-MDA as a carcinogen in
animals.
-------
36
5. Weight of Evidence
Taken together, the strongly positive results in the NTP
cancer bioassays on the dihydrochloride salt of 4,4'-MDA,
evidence of the carcinogenicity in animals and humans of 4,4'-MDA
structural analogues, the demonstrated ability of 4,4'-MDA to
induce bladder tumors in animals and suggestive evidence of MDA-
induced bladder tumors in humans, and data indicating the ability
of 4,4'-MDA to interact with genetic material, lead to the
conclusion that this chemical is carcinogenic in animals and is
probably carcinogenic in humans.
In conducting risk assessments of suspect carcinogens, EPA
generally evaluates the overall weight of evidence including both
primary and secondary evidence of carcinogenicity. As specified
»
in the draft EPA Guidelines for the Health Assessment of Suspect
Carcinogens (EPA, 1984), primary evidence derives from long-term
animal studies and available epidemiological data. Secondary, or
supplemental, evidence includes structure-activity relationships,
the results of short-term tests, pharmacokinetic studies,
comparative metabolism studies, and other toxicological responses
which may be relevant.
Based upon the weight of available evidence, EPA classifies
4,4'-MDA as a probable human carcinogen and places it in category
(B2). The Guidelines cited above give this classification when:
evidence of carcinogenicity from
epidemiological studies ranges from
almost 'sufficient' to 'indequate.' To
-------
37
reflect this range, the category is
divided into higher (Group Bl) and lower
Uirouy 62) degrees of evidence.
Usually, category Bl is reserved for
agents for which there is at least
limited evidence of carcinogenicity to
nunans front epideniological studies. In
the absence of adequate data in numans
it is reasonable, for practical
*
purposes, to regard agents for which
tnere is sufficient evidence of
carcinogenicity in animals as if they
^resented [sic] a carcinogenic ris.K to
hunans. Thererore, agents j:or which
there is inadequate evidence frop hmian
studies and sufficient evioence from
aninai studios, [as with 4 , 4 '-f«iDA ] would
usually result in a classification of
u'2.
a. Animal Studies
There is sufficient evidence of carcinogenicity in aninals
to support the cited classification of 4,4'-MDA. Jn NTP
hioassays, the aihydrochlor.i.de of the chemical was found to be
carcinogenic upon oral administration in hoth sexes of two
species (rats arnJ nice) ana caused turoors at multiple sites in
each species. Significantly increased incidences of tutors were
•.
observed in the thyroid and 'the liver irf both species. The sites
-------
33
of response in the mouse-also included the adrenal glands
(males), and the lung and ly-irpnatic 'system (females). Several
extremely rare tumor types with very low spontaneous incidence
were also observed. These included one bile duct adenoma in a
i-iale rat (spontaneous incidence in historical control rats of
(j/3633), transitional cell papillomas of the urinary bladder in
tnree female rats (spontaneous incidence in historical controls
or 3/3644), and granulosa-cell tumors, including one carcinoma,
in five female rats (il such tumors and one such carcinoma in
3462 historical control rats). observation of these rare tumors
in these studies in test groups consisting of only 50 animals is
a sign of chemical specificity and is highly significant evidence
or tne carcinogenic potential of 4,4'-.M|j£. Thyroid tumors in
rat:i were also observed in a. limited bioassay performed by Hiasa
_^t_ _al_. (Iy>i4). In another study or rats treated with ^, d'-MDA by
suocutaneoiio injection, Steinnoff and (irundnann (1°>70) concluded
that the results suggested carcinogenic activity. This study was
limited in value by cne sw.ail number of test anxrnals (50) and.
incomplete pathology reporting. other studies on the
carcinogenic potential of 4,4'-iMDA nave been conducted, but were
not adequate in design or performance for conclusions to be
reached.
b. Epideniolocjical Studies
only one epidemiology study is available. This proportional
mortality study 'was reported in 1.983 by Niosh (19H3). It was
conducted at a site of manufacture of helicopters where exposure
to 4 , 4 '-rtL'A ard otner cneiaicals was measured. NIOSH studied
-------
39
information on 179 white male deaths that occurred among these
workers from all causes and found a significant excess over the
expected proportion of bladder cancer-related deaths. This
excess remained significant in analysis of only the cancer
deaths. In addition, two cases of bladder cancer were found in
4,4'-MDA - exposed living persons. 4,4'-MDA could not be
definitively concluded to be the causative agent because of
confounding exposure, but the evidence is suggestive given the
corresponding observation of urinary bladder tumors in the NTP
bioassay in rats cited above, and 4,4'-MDA's structural
similarity to benzidine — an agent known to produce this type of
tumor in humans.
c. Structure-Activity Relationships
Structure-activity considerations are strongly supportive
evidence of the human carcinogenic potential of 4,4'-MDA. The
chemical is a member of the structural class of bis 4-
aminobenzenes in which two benzene rings are separated by -Ct^-,
-O-, or -S- groups. Members of this structural class include
4,4'-oxydi'aniline, 4,4'-thiodianiline, and 4,4'-methylenebis
(N,N-dimethylaniline), all of which have been found to cause
neoplasms in the liver and thyroid of rodents, as does 4,4'-
MDA. Other members of this class, 4,4'-methylene bis (2-
methylaniline), and benzidine, the structual analogue of 4,4'-MDA
in which the methylene bridge between the aromatic rings is
absent, have been associated with an increased risk of bladder
tumors in humans.
-------
40
d. Absorption
Further support for EPA's conclusion that the chemical poses
a risk of cancer to humans is the fact that the chemical is
absorbed by the human body. 4,4'-MDA is known to be absorbed by
humans through the skin in workplace settings. Information from
the United States, Canada and France attests to the dermal
.absorption of the chemical, and the scientific literature
documents cases of the liver toxicity of 4,4'-MDA following
dermal exposure along with detection of the chemical in the urine
of workers exposed by this route.
Since the chemical has been shown to penetrate human skin
and to be absorbed through the human gastrointestinal tract (in
the so-called Epping Jaundice Incident), EPA believes it
reasonable to anticipate that the chemical will penetrate lung
tissue as well.
e. Mutagenicity
In short-term tests, 4,4'-MDA has been shown to be a gene
mutagen in prokaryotic systems. The chemical induces sister
chromatid exchanges in femoral bone marrow of male mice; it does
not induce chromosomal aberrations in vitro in human peripheral
lymphocytes. (A mixture of positive and negative results is not
unusual in tests for genotoxicity since the tests for gene
mutations and chromosomal aberrations measure different
endpoints.) In addition, the compound binds covalently to DNA in
vivo in the livers of treated mice, indicating its ability to
interact with macromolecules in vivo.
-------
41
0. Other Hunan Health Effects
This section presents a summary of reports of adverse
effects that 4,4'-r-lOA exposure has caused in hunans.
/•;opeiman _a_t _a_^. (196fia, 196(>b) reported that H4 people who
nad eaten bread contaminated with 4,4'-MD£ developed jaundice,
with nepatoeellular damage evidenced by biochemical tests and
neec.ie biopsy examinations.
•.;il liains et al. (ly?4) reported six cases of hepatitis ariong
workers using 4,4'-ivii.)/» in a surface coating operation at a
construction sice, and McGill and riotto (1974) reported 13 cases
of nepatitis anong workers exposed inhalationally (at 0.1 ppn in
air) and dermally while producing on epcxy resin compound. Of
special significance in the Mcuill and. iviotto report is the fa.ct
that workers who were 3xposed only via the inhalation route (at
u. i ppn in air), in the sane work stations as those wtio also were
exposed oernally ?nd expT)rienct?A to ^enetrata !iunan s
-------
These reports indicate that 4,4'-ML>A. can be absorbed by
humans in ruoloc_,ically significant amounts, anri that apparent
exposure at n.l ^Pri i-n air Goes not result in overt signs of
acute hepatotoxicity, thus there ^ay be no warniny of exposures
tnat could present Giynificant cancer risk.i. (bee Section v).
-------
43
IV. EXPOSURE ASSESSMENT
In this section information on the potential for exposure to
4,4'-MDA by workers, consumers and the general public will be
given. This information will include the production and use
patterns for the chemical and exposure levels, durations and
routes for workers classified as to the various uses of 4,4'-
MDA. Also included will be estimates of releases of 4,4'-MDA
into the air, water and land, the fate of such releases, and the
levels of the chemical that might occur in certain drinking water
supplies.
Humans may be exposed to 4,4'-MDA in the workplace, through
contact with articles or other products containing the chemical,
or through consumption of contaminated water or food. In order
to assess these possible exposures, EPA, in conjunction with
NIOSH, has gathered data on work practices and exposure levels in
v
facilities that manufacture, process or use 4,4-MDA. In
addition, the Agency has gathered information on the types of
articles and products that may contain the chemical, and on the
amount and rate of releases of 4,4'-MDA to the environment
throughout its commercial life cycle.
The assessment indicates 1) that exposures experienced in
some workplace situations are significant and 2) that exposures
through drinking contaminated surface water are probably not
significant.
It is important to note tnat several different analytical
methods for determining levels of 4,4'-MDA in air, water, and
biological samples have been used, and that a systematic attempt
-------
44
at concordance among these methods is underway. Appendix A
summarizes the methods that have been used to measure airborne
concentrations of 4,4'-MDA. In particular, the Marcali
colorimetric method for 4,4'-MDA, used by some 4,4'-MDA
manufacturers who reported levels of the chemical to EPA, suffers
from interference from aromatic amines and isocyanates, both of
which can be present in the air of 4,4'-MDA/methylene diphenyl
diisocyanate (MDI) manufacturing facilities. Such interferences
are less likely to occur at 4,4'-MDA user and processor
facilities. This assessment is based on the assumption that the
exposure levels reported in the literature or to EPA are
accurate.
Information regarding exposure of workers to 4,4'-MDA is
derived from anecdotal reports cited in Versar (1983a), data
voluntarily submitted by the Chemical Manufacturers Association
(CMA, 1983a), individual companies (Docket No. OPTS 64,000a), a
report prepared by PEDCo Environmental, Inc. (PEDCo, 1983) and
the National Institute for Occupational Safety and Health (NIOSH,
1983 and 1984a,b).
Information regarding exposure of the general public is
derived from a draft report prepared by Versar Inc. (Versar
1983b).
A. Workplace Exposure—4,4*-MDA/MDI Manufacturing
1. Production Processes
a. 4,4'-MDA (PEDCO, 1983)
4,4'-MDA is produced commercially by the acid-catalyzed
condensation of aniline and formaldehyde. The initial product is
acidified, crude 4,4'-MDA, almost all of which is purified as
-------
figure |. A,A'-HDA/MIU Flow Diagram. (PBOCo. 1983)
-------
46
outlined below for use as an intermediate for raethylenediphenyl
diisocyanate (MDI) at the same production facilities.* Figure 1
As a flow diagram for this process/ and for conversion of 4,4'-
MDA to MDI.
The production process can be either continuous or batch.
Acid is added to the aniline/ followed by addition of
formaldehyde under agitation. The reaction occurs in a closed
vessel at atmospheric pressure and a temperature of 60° to
100°C. The only product is acidified crude 4,4'-MDA; there are
no significant byproducts, except that up to 40-50% of the crude
4,4'-MDA consists of isomers and.higher homologues.
There are no 4,4'-MDA exposure points other than those that
might occur as a result of fugitive emissions or malfunction of
the vessel or its plumbing. There are no purge or off-gas
streams.
In the neutralization step, the reaction mass from the
reactor tank is neutralized with caustic soda to produce crude
4,4'-MDA. Two layers are formed, a lower layer containing water
and'salt, and an upper layer containing the product and unreacted
aniline. The organic layer is w.ashed with water in a closed,
stainless steel tank at ambient pressure and temperature.
The input materials are the acidified reaction mass, aqueous
sodium hydroxide, and water. The principal output is crude 4,4'-
MDA, which may either be further purified or used for the
production of MDI. A large quantity of sodium chloride is also
* One domestic plant further processes the polymeric 4,4'-MDA to
yield a purified product that finds a number of specialized end
uses.
-------
47
formed and removed in an aqueous purge stream likely contaminated
with 4,4'-MDA and other organics.
The potential for worker exposure in this process is
expected to be limited. The purge stream likely contains some
4,4'-MDA, so exposure could occur in an "open" system around
holding ponds. Neutralizer purge stream levels of 4,4'-MDA are
not well documented in the literature, but monitoring work now in
progress is expected to provide some information on 4,4'-MDA
levels in total process effluent streams'.
The crude 4,4'-MDA from the neutralizer is subjected to
distillation to separate unreacted aniline. The distillation
process is conducted at atmospheric pressure at temperatures
above 185°C in a closed, stainless steel distillation column.
The residual product which solidifies to a hard waxy material at
temperatures below 89°C, may be drummed for distribution and sale
or further purified.
The only input to this step is crude 4,4'-MDA. The product
is a technical grade 4,4'-MDA, which is about 50% 4,4'-MDA plus
oligomers. The process by-products are water and aniline. The
latter is recycled to the reactor vessel.
Liquid and solid wastes containing 4,4'-MDA are produced in
this process. Liquids may be retained in holding ponds and the
solids sent to landfills. No data are currently available
regarding 4,4'-MDA levels in either stream.
Because the distillation system is closed, worker exposure
is not expected except in the case of system disruption.
Filling, sealing, storage, and distribution activities could
-------
48
result in inhalational or dermal exposure to 4,4'-MDA at various
points.
High-purity (97-99% assay) 4,4'-MDA can be isolated from
technical grade 4,4'-MDA. A re-crystallization process (PEDCo,
1983) involves the reformation of the hydrochloride salt of 4,4'-
MDA, followed by filtration and neutralization with sodium
hydroxide. The product 4,4'-MDA is packaged for distribution and
sale. The equipment that would be used is not described in the
literature, but would probably be a closed/ stainless steel
tank. The process would take place at ambient pressure and
temperature. The input materials would be technical grade 4,4'-
MDA, hydrochloric acid/ sodium hydroxide, and water. The product
would be purified 4,4'-MDA in the form of tan flakes. There
would be an aqueous contaminated salt stream that may contain
some 4,4'-MDA.
The system would be closed until output, so worker exposure
•
during purification would be unlikely, although dermal exposures
during manufacture and packaging of the purified form of the
chemical have been reported (NIOSH, 1984a). Subsequent purge
stream holding and disposal operations could also result in
inhalational or dermal exposure.
Technical grade and purified 4,4'-MDA is used in a number of
non-MDI processes. It is packaged and sold as a viscous liquid
or lumps in the technical grade form, or as flakes or granules in
the purified form. Liquid 4,4'-MDA is sold in bulk (tank cars or
tank wagons) or in 55-gallon drums. The lump form of 4,4'-MDA is
sold in bags. Purified 4,4'-MDA is available in bags or kegs.
-------
49
Workers who fill the containers with liquid or solid 4,4'-
MDA are subject to potential exposure, though 4,4'-MDA's vapor
pressure is low even in the liquid state (0.2 mm Hg for Tonox®,
Uniroyal's trade name for 4,4'-MDA, in the melt range).
Inhalation of dust or skin contact from a bagging or drum-filling
operation represent potential risk to the worker. NIOSH (1984a)
reports that such exposure occurs.
Workers may also be exposed to 4,4'-MDA, via inhalational
and dermal routes when handling the filled containers. Drums and
bags must be relocated, stored, and placed on trucks for
shipment. Damage to the container in any of these operations
could expose workers to 4,4'-MDA
b. MDI
Technical grade 4,4'-MDA is reacted with phosgene to produce
the desired product, a crude MDI known commercially as PAPI®.
4,4'-MDA and phosgene are dissolved in a solvent, and the two
solutions are mixed and allowed to react for several hours. The
reaction occurs in a series of closed, agitated, stainless steel
reactor vessels at ambient pressure and a temperature of 200°C.
Input materials are polymeric 4,4'-MDA, phosgene, and a
solvent such as xylene, monochlorobenzene, dichlorobenzene, or
1,2,4-trichlorobenzene. Outputs are the product isocyanate
mixture and a liquid waste stream containing hydrogen chloride,
solvent, and miscellaneous organics. The reaction is reportedly
capable of a 90% yield of isocyanate mixture. Theoretically, the
waste streams generated could contain some unreacted 4,4'-MDA,
although this is unlikely.
-------
50
Workers may be exposed to 4,4'-MDA through dermal contact
with the liquid product or waste streams, or through inhaling MDI
vapors that could be hydrolyzed in the body to 4,4'-MDA.
Crude MDI may be fractionated through several intermediate
grades to a high-purity MDI. Dimers and trimers can also be made
through use of a catalyst, triethyl phosphate. The process takes
place in a closed, stainless steel distillation column, the
operating conditions for which are not reported in the
literature. Temperatures would be elevated.
Input material is crude MDI. The outputs are higher grades
of MDI, as well as recyclable hydrogen chloride and solvent,
which are diverted to a purge stream. This stream also contains
additional organic distillation residue. The product MDI is a
very dark amber, viscous liquid with very low volatility.
The amount of unreacted 4,4'-MDA remaining through this step
depends upon the purity of the MDI produced. There is virtually
no unreacted 4,4'-MDA with high-purity MDI. Waste disposal
consists of extraction of hydrochlogen chloride for process
reuse, disposal t'o holding ponds, and finally municipal
wastewater treatment.
2. Production History and Forecast
U.S. production of 4,4'-MDA has been increasing steadily
since the early 70's. Production is driven primarily by the
demand for MDI, manufacture of which consumes about 98% of 4,4'-
MDA production (JRB, 1981). Since no production data for 4,4'-
MDA have been published in recent years, the estimates presented
•»
in Table 2 below are based on published MDI production data. CMA
-------
51
has estimated that 230 million pounds of 4,4'-MDA were produced
in 1981 by all manufacturers except DuPont (Cox, 1982). Since it
has been reported that DuPont produces approximately 10 million
pounds/year, this would amount to an estimated 240 million
pounds/year 4,4'-MDA production. The difference between CMA's
estimate and that of PEDCo is probably due to the former's
reporting only the production of the 4,4'-isomer, while the
latter includes all isomers and oligomers in technical grade
4,4'-MDA which is converted to commercial grade MDI.
Consolidated data received under the TSCA section 8(a) Level A
reporting rule (47 FR 38780, September 2, 1982) indicate that
between 209 and 286 million pounds were produced in 1981
(Knutson, 1983). Estimates thus range from 209 to 414 million
pounds/year.
Polymeric MDI is used in polyurethane production; demand is
increasing for two major reasons (JRB, 1981):
o Increased substitution of MDI for toluene diisocyanate
(TDI) in the production of polyurethanes due to concern about the
toxicity of MBOCA in MBOCA/TDI systems.
o Increased demand for polyurethane in automotive body
interior applications as a result of Federal regulations for
impact resistance, improved vehicle safety, and improved gas
mileage.
-------
52
TABLE 2
ESTIMATED 4,4'-MD& PRODUCTION 1976-1980 (PEDCo, 1983)
Production (millions of pounds)
Year MDI (USITC) 4,4'-Mm (MATHTEC, 1982)
1976
1977
1978
1979
1980
1981(a)
312.2
352.3
439.5
487.7
511.1
517.9
249
289
352
390
409
414
PEDCo update to referenced data.
MDI production is projected to increase steadily through
1985 at an annual rate of 9% (Mannsville, 1980). Consumption for
the non-MDI uses of 4,4'-MDA are estimated in Table 5.
Imports of 4,4'-MDA have been small, accounting for less
than 0.4% of total supply (1977-1979) (MATHTEC, 1982). Export
data are not published separately for 4,4'-MDA and are assumed to
be small. Exports of MDI, however, accounted for 15 to 18% of
MDI production in 1980 (Mannsville, 1980). Several foreign MDI
plants are expected to begin operation, and this will probably
reduce the demand for MDI exports (MATHTEC, 1982).
4,4'-MDA is currently manufactured by six companies at seven
locations in four states, as shown in Table 3. One additional
company (BASF), which produced 4,4'-MDA in the past, is presently
importing the chemical and plans to open a manufacturing facility
in the future. Three of these companies—Mobay, Rubicon, and
Upjohn—appear to account for over 90% of 4,4'-MDA production
(Springborn, 1982, 1983).
-------
53
TABLE 3
PRODUCERS OP 4,4'-MDA AND MDI
PLANT LOCATIONS AND ESTIMATED CAPACITY (MATHTEC, 1982)
Company
Olin Chemical
BASF Wyandotte
E.I. duPont
itobay Chemical
Rubicon Chemical
(ICI Americas)
Uni royal
Upjohn
location
Moundsville, WV
Geismar, LA
belle, WV
New Martinsville, WV
Bay town, TX
Geismar, LA
Nauytuck, CT
LaPorte, TX
1UTALS
197y-1980
blDPi Capacity
(million
NA
NP
10-50
79
79
79
NA
213
460-500*
1979-1980
MDI Capacity
pounds )
NP
NP
NP
100
100
100
NP
270
570
1985
MDI Capacity
(estimated)
150
100
200
250
NA
270
970+
NA = Not available
NP = Not a producer
-------
54
3. Characterization of Non-MDI Uses of 4 ,,4'-MDA
Information from several sources (CMA 1983a; ICF, 1983;
Springborn, 1982, 1983) was used to characterize the commercial
fate of the 4,4'-MDA that is not converted to MDI. Table 4 (ICF,
1983) summarizes the information from these data sources.
Additional information from a CMA-sponsored survey of users and
processors of 4,4-MDA is in general agreement with the material
presented in Table 4 (CMA, 1984).
For each of the ICF (1983) categories in Table 4, exposure
level information is presented below. The amounts used in each
ICF (1983) category are shown in Table 5.
4. Exposure Levels and Duration
Two populations of potentially exposed workers are of
concern: workers who manufacture 4,4'-MDA and convert it to MDI,
and workers who use or process MDA for other than MDI
applications.
Two major sources of exposure information are used in this
assessment: 1) data submitted voluntarily by manufacturers and
processors, and data compiled by PEDCo (1983), summarized in
Tables 6 and 9; and 3) measurements made by NIOSH representatives
during visits to a 4,4'-MDA manufacturing plant that makes 99%
assay product (NIOSH, 1984a) and to a facility that uses 4,4'-MDA
as a curing agent for epoxy-coated, filament-wound pipe (NIOSH,
1984b), which are summarized in the text of Section V below.
-------
55
TABLE 4
NGN-MDI USES OF 4,4'-MD&
Springborn (1983)
Categories
ICF (1983)
Categories
CMA (1983a)
Categories
Epoxy curing
Urethane curing
MBOCA production
Ketiinine production
Wire coating production
Production of coatings
for circuit boards and
aircraft parts
Dye intermediate
Qiana® intermediate
(not used)(a)
Rubber processing
chemical (not used)
Anti-oxidant in
lubricating oil
(not researched)
(b)
Epoxy uses
Co-reactant in
polyurethane
Wire coating
PMR-15 as a
polyimide
Dye intermediates
Epoxy curing
Producing of TGMDA
Co-reactant in
polyurethane
Polyester- imide
wire coatings
PMR 15 polyimide
MDA as an inter-
mediate for polybis
maleimides
MDA as an inter-
mediate for dyes
and pigments
Qiana® intermediate
(not used)(a)
Rubber Processing
Chemical (not used)
(a)
Corrosion inhibitor
(not researched)^ '
Reporting entites indicate that 4,4'-MDA has been, or could be, used for
this purpose, but is not presently so used.
Springborn (1982) shows zero pounds used.
-------
56
TABLE 5
NQXJKL PRODUCTION OP 4,4'-MD& FOR NON-HDI USES (ICF, 1983)
Use Thousands of Pounds
Epoxy Curing 5,000-7.000(a)
Wire-Coating 200(b)
Coreactant in Production
of Polyurethane 250
Dyes 1,500-2,000
Nuclear Weapons Production^0' 10
PMR-15 N/A
TOTAL 6,950-9,450
High estimate consists of 4 million Ibs. of crude (65% 4,4'-MDA)
and 3 million Ibs. pure 4,4'-MDA (97-100% 4,4'-MDA).
b Springborn (1983).
(c;> DOE (1983b).
The estimates of dermal exposure to 4,4'-MDA used in this
assessment in both MDI- and non-MDI manufacturing facilities are
based on data from the NIObH visits just described. The
conclusion should not be drawn that the exposures — and the
estimated doses and risks derived from those exposures — are
identical to those in all workplaces. There are not enough data
points to permit such a conclusion. Nevertheless, the data used
in this preliminary assessment to estimate dermal exposures were
obtained using reasonable industrial hygiene and analytical
procedures, and this analysis assumes that they are reasonable
estimates of the kind of exposures that may occur in many
workplaces. The information submitted to EPA by users and
processors of 4,4'-MDA (Docket No. 64000a) and that obtained from
the open literature (Dunn and Guirguis, 1979; Vaudaine et al.,
1982; Brooks et al., 1979; McGill and Motto, 1974; Dunn, as cited
-------
57
in NIOSH, 1976a; Emmett, 1976; NIOSH, 1976b) supports this
conclusion.
a. 4y4'-MDA and MDI Production Workers
As described above, 4,4'-MDA is produced in the United
States by six companies at seven locations (PEDCo, 1983). At
these facilities, 284 workers are exposed to 4,4'-MDA for less
than 8 hours per week, 159 are exposed for 9 to 20 hours per
week, and 13'3 are exposed for more than 20 hours per week. This
information, aggregated here, was submitted to EPA by the CMA
project panel on 4,4'-MDA under claims of confidentiality for
data from individual companies (CMA, 1983c) and is in substantial
agreement with information submitted under section 8(a) of TSCA
(Knutson, 1984).
These workers are exposed to 4,4'-MDA chiefly through
inhalation of vapors or particles, or through dermal contact with
the chemical*. Levels of exposure for both routes have been
measured, although there are uncertainties in these measurements.
Chief among the uncertainties in inhalation exposure
measurements is the accuracy of analytical methods used (see
Appendix A). For the dermal route, the rate of absorption is the
least accurately known factor. A dermal absorption study is
planned for completion in 1984, and work is proceeding in several
laboratories toward resolving analytical difficulties.
A reasonable estimate of the exposures that these workers
experience can be made using information supplied by the
* Ingestion, via contaminated smoking mat-erial or food, could
occur, but is not considered here due to lack of data on this
route of exposure.
-------
58
manufacturers of 4,4'-MDA (CMA, 19<83c) and information obtained
during a visit by NIOSH industrial hygienists (NIOSH, 1984a) to a
4,4'-MDA manufacturing plant. The former data include ranges and
average 8-hour Time Weighted Average (TWA) measurements of air
levels of 4,4'-MDA in the workplace, while the latter include the
results of dermal studies.
Table 6 presents estimates of airborne exposure that workers
in the less than 8 hours-, 9 to 20 hours- and more than 20 hours
per week categories may experience.
Table 7 gives the dermal dose estimates for 4,4'-MDA
production workers. These estimates are based on the assumption
that the exposure experienced by the chemical operator studied in
NIOSH (1984a) is a reasonable estimate of exposures that may
occur under similar working conditions. This worker was supplied
with new gloves at the beginning of his snift, and under the
glove were mounted pads on the palm and back of the hand, using a
golf-glove-like device, which collected 4,4'-MDA during the
shift. It should be noted that NIOSH representatives observed
conditions, such as apparently routine re-use of gloves, that
might result in higher doses than those calculated here.
-------
59
TABLE 6
4r4'-MD& MEG. WORKPLACE AIRBORNE EXPOSURE LEVELS (PPM) (CMA, 1983c)
Sample
Conpany(a) Type(b) <8 hrs/wk
A bz
B a
C bz
D a
E bz
Range
<.0002
ND-0.6
<.005
.001-. 89
•-"-"»
Ave.
.
0.07
<.005
.06
— "•-•
DURATION
9-20
Range
___
.001-0.7
<.005
.001-. 89
^•^HM
hrs/wk
Ave.
__
.02(2)
<.005
.06
"««•
>20 hrs/wk
Range
<.001-.009
ND-0.6
.001-. 89
.01-. 10
Ave.
.004
.07
.06
.05
(b)
The sixth domestic 4,4'-MDA manufacturer is not a member of the CMA Panel
and did not .submit monitoring information to EPA.
a = area, bz = breathing zone.
•RBLE 7
ESTIMATED DERMAL DOSES IN 4,4'-MD& MFG PLANTS(a)
8 HRb/WK
0.40 mg/day
20 HRS/WK
2.3 mg/day
40 HRS/WK
4.9 mg/day
Calculations shown in Appendix B. Workers classified in the <8 hr/wk
group are assigned 8 hr/wk exposure; those in the 9-20 hr/wk group are
assigned 20 hr/wk; etc.
-------
60
Table 8 shows lifetime average daily doses (LADDs)* based on
the combined inhalational and dermal exposures presented in
Tables 6 and 7. The inhalational component of these LADDs is
based on the highest reported average 8-hour TWA, not the highest
range, reported in CMA (1983c). An average is considered more
representative of actual exposures than a high range number would
be, and the use of the highest reported average from all
reporting manufacturers is a conservative measure in this risk
assessment. The analytical method used to determine the TWA used
here is subject to interference from aniline also known to be
*LADDs are a tool that makes possible the quantitative estimate
of risk to humans using the results of studies on animals. In
most cancer studies using animals, the animals are exposed to a
known level of the test chemical at a known frequency for their
entire lifetime, and this results in some incidence of cancer.
Some comparable expression of the dose received by exposed humans
is required to translate the dose-response seen in animals to the
human case, and this comparable expression, the LADD, is but an
imperfect estimate of exposure, since humans are almost never
exposed in exactly the same way as the test animals. For
instance, in the 4,4'-MDA case, animals were exposed to the
chemical each time they drank water, over their whole lifetime,
while workers are exposed, at most, for 8 hours per day, 250
workdays per year for as long as they work — perhaps as long as
40 years. Some workers may be exposed for fewer years than
others, some may experience higher' or lower levels of exposure,
and some may be exposed for fewer days or a different number of
hours each workday. These various factors are combined as
illustrated in Appendix B to produce a LADD.
The artificiality of the LADD can be exemplified by looking
at the Epping Jaundice Incident which people ate 4,4'-MDA-
contaminated bread. Surely these people are at higher risk of
cancer from 4,4'-MDA than they would have been had they not been
exposed. In order to quantitatively estimate what their elevated
risk is, one would have to translate the dose they received into
terms that could make use of the dose-response seen in the animal
bioassay. That would be an LADD. We are virtually certain that
they are not being exposed anymore, so to say that they are
receiving X mg/kg/day, as a LADD does, even now, is pure
artifice. One notes, nevertheless, that" those people did suffer
from liver disease, and this illustrates that a short term, high
level exposure to 4,4'-MDA does have a different outcome from „
chronic, low-level exposure (See subsection b, below)..
-------
61
present in the workplace. Because of this possible interference,
the TWA value used here may be high (by as much as a factor of 10
[CMA, 1983c]). This would reduce the LADDs given in Table 8 by
from about 50% (for 8 hours/week duration) to about 25% (for 40
hours/week duration).
8
ESTIMATED 4,4'-MD& MEG. VOFKPLACE LADDs
LADD(a) by Route _ DURATION
Inhalation (mg/kg/day)
Dermal (mg/kg/day)
Total (mg/kg/day)
8 hrs/wk
0.0031
0.0020
0.0051
20 hrs/wk
0.0064
0.014
0.020
40 hrs wk
0.015
0.027
0.042
Lifetime Average Daily Dose. Calculations shown in Appendix B.
b. Non-MDI Uses of 4,4'-MDA
i. Epoxy Curing
a. Exposures Related to the Curative
Package
Epoxy curing uses of 4,4'-MDA can result in exposures during
the' formulation, packaging, and subsequent handling of the
curative "package", which would consist typically of pulverized
4,4'-MDA, fillers and pigments, and during handling and uses of
the blended epoxy resin/curative mixture. These two types of
exposures are discussed below. The total number of workers
exposed is not known, but could range from about 1,500 to about
-------
62
13,000 (NOHS, 1983). Exposure durations are not known, but have
been estimated based on assumptions outlined in Section V below.
Exposure levels for the non-MDI uses that were reported by
manufacturers or processors are summarized in Table 9.
The number of firms engaged in formulating epoxy curative
packages is not known. One such firm responded to EPA's 4(f)
notice on 4,4'-MDA and submitted information on its products and
on worker exposure levels (Ameron, 1983). Among other products,
this firm manufactures curatives for epoxy coatings used to
protect concrete structures in nuclear power plants, steel
members in certain marine structures, and chemical tanks. From 10
to 20 employees are reportedly exposed to airborne concentrations
of 4,4'-MDA ranging from 0.073 to 0.68 mg/m3.
-------
63
TABLE 9
ESTIMATES OF M» AVAILABLE FOR ABSORPTION
BY NUN-MDI USE
(PEDOO, 1983)
Process
Reprocess ing
of MDA
Epoxy uses
Polyurethane
curing
Operation
Dumping
Mixing
Packaging
Unspecified
Mixing of MDA
to make fila-
ment-wound
pipe
Manufacture
of TGMDA
Pulverization
of MDA
Unspecified
Potting roan
near oven
Exposure
route
Inhalation
Inhalation
Inhalation
Dermal
Inhalation
Inhalation
Inhalation
Dermal
Inhalation
dermal
Measured
ppm
0
-
U
(a)
0.0125
0.0075
0.1
Unknown
0.01
(a)
level
mg/m
.12-3.11
0.48
.53-0.68
(a)
0.10
0.06
0.82
-
0.08
(a)
Estimated
exposure time
hrs/day
1
1
6
8
2
8
1
-
8
8
day/yr
250
250
250
250
250
250
250
250
250
250
Amount available
for absorption
Inhalation
mg/kg/yr
7
2
16
N/A
1
2
4
N/A
3
N/A
Dermal
mg-hr/kg/yr
N/A
N/A
N/A
3,600la'
N/A
N/A
N/A
-
N/A
3,600la'
CONTINUED
-------
64
TABLE 9 - CONTINUED
Process
Operation
Exposure
route
Measured level
Estimated
exposure time
Anount available
for absorption
mg/nT
firs/day day/yr
Inhalation
mg/kg/yr
Dermal
mg-hr/kg/yr
Intermediate
for poly-
imides for
wire coating
Intermediate
for poly-
male imides
for aircraft
parts
Intermediate
for dyes
Intermediate
for rubber
additive
Dumping of MDA Inflation
Unknown
100
N/A
Unspecified
Dumping of
MDA into
Dumping of
MDA into
reactor
Inhalation/
dermal
Inhalation/
dermal
Inhalation
dermal
Unknown
Unknown
40
Unknown
250
N/A - Not applicable
2
(a) Based on wipe samples taken of a similar chemical, MBOCA, used to cure polyurethane, averaging 72 ug/cm (17 samples)
Other estimates of the amount of 4,4'-MDA available for absorption in some of these settings are given in Section V.
-------
65
Eleven cases of acute jaundice related to 4,4'-MDA exposure
in a curative formulation plant between 1967 and 1976 were
reported by Dunn and Guirguis (1979). At this plant, located in
Ontario, Canada, previously ground and screened 4,4'-MDA was
received, blended with silica sand, and packaged. Blending was
carried out in two blenders located in a separate room with roof
exhaust. The materials were mixed for 20 minutes, then packaged.
Five to seven workers were employed in the manufacturing area at
any one time. Workers wore respirators when adding 4,4'-MDA and
sand to the mixer. After 1976, the company began providing
workers with coveralls, gauntlets, shoe covers, head and neck
covers, and positive-pressure airline breathing apparatus.
Despite these changes, some workers were still affected,
indicating that special care in the use of protective clothing
and equipment must be taken to ensure against inadvertent
contamination that can lead to exposures via dermal,
inhalational, or ingestion routes (see the report of Vaudaine et
al., 1982, below).
Air sampling data (15-minute samples) taken during charging
of 4,4'-MDA to the blender ranged from 0.2 to 3.11 mg/m ; during
mixing the measurement was 0.48 mg/m ; and during packaging the
measurement ranged from 0.53 to 0.68 mg/m . Improper gasketing
of equipment accounted for a large proportion of dust during the
blending. A flexible exhaust hose (20 cm in diameter) placed
near the operation during charging and bagging was found to have
a large hole and was replaced. During production, the workers
wore coveralls which were changed twice a day, a hat with a wide
-------
66
brim, impervious gloves, and an airline respirator. Coveralls
were found not to be impervious to 4,4'-MDA. ( While Dunn and
Guirguis reported no quantitative information of the penetration
of coveralls by 4,4'-MDA, this permeability is used as one of the
assumptions in applying NIOSH (1984a, 1984b) dermal exposure data
to calculation of dermal LADDs in Appendix 3.)
Additionally/ the armpits of affected workers were stained
with 4,4'-MDA because these individuals had defeated the
protection of the coveralls by making cuts in the material to
relieve the heat stress of summer days. The main route of entry
of 4,4'-MDA to the affected persons was believed to be dermal.
The type and extent of controls used by other curative
formulators is not known, though investigation of both controls
and exposure levels are currently being investigated.
b. Exposures Related to Handling/Use of Complete
Egoxy Resin Formulations
Epoxy resins are used in a variety of applications
including: coatings, laminates and composites, casting and
molding, flooring, and adhesives. 4,4'-MDA has been used in
epoxy molding powders, stick solders, fiberglass cloth-epoxy
laminates, and casting compounds (Springborn, 1982).
Since the applications for epoxies are so varied, there is
no simple process or operation that can be described for curing
of epoxy resins with 4,4'-MDA. In general, either purified or
polymeric 4,4'-MDA is added to the epoxy resin and mixed just
prior to use of the polymer. This can be done at a construction
site, as with use of the epoxy as a concrete coating, or in a
-------
67
closed and automated process, as in the manufacturing of filament
windings. In all cases, the process involves adding the 4,4'-
MDA, mixing with the epoxy resin, and applying the mixture to a
surface or forming the product. All three steps are potential
points of worker exposure, but most information, including NIOSH,
(1984b) indicates that the addition of 4,4'-MDA to the resin is
the operation of greatest concern, because of the propensity for
4,4'-MDA, especially the high assay material, to "dust" and
contaminate nearby surfaces during such handling. The epoxy
cross-linking reaction forming the epoxy is exothermic, and the
mixing and curing processes are usually at elevated temperatures.
4,4'-MDA is used as a curing agent for epoxy resins in a
wide variety of structural laminates including filament winding,
wet lay-up laminates, and potting, casting, and encapsulation
(CMA, 1983a). Filament-wound epoxy pipe cured with 4,4'-MDA has
numerous uses including casings of rockets (Trident, Pershing,
MX, and space shuttle), oil drilling pipe, pipes for chemicals,
and fuel tanks for military aircraft (CMA, 1983a). CMA (1983a)
reports exposure in a European filament winding plant ranged from
0.002 mg/m3 to 0.1 mg/m , with the high value during the
preparation of the mixture of epoxy resin and 4,4'-MDA.
Wet lay-up laminates cured with 4,4'-MDA are used as
structural parts for aircraft. The epoxy is used to impregnate a
reinforcing fiber cloth. The impregnated cloth, called
"pre-preg", is then refrigerated to arrest curing until final
molding occurs under heat and pressure. CMA (1983a) reports
4,4'-MDA concentrations below the detection limit of 0.001 mg/m
-------
68
in the pressroom. No measurements are given for the mixing
operation. NIOSH, however, reports that breathing zone
measurements done on a molder of helicopter blades showed
concentrations of 0.23, less than 0.022, and 0.46 mg/m3 of 4,4'-
MDA for three different 20-minute periods in one shift (NIOSH,
1983).
Liquid epoxy resins cured with 4,4'-MDA are used in the
potting, casting, and encapsulation of electrical components.
4,4'-MDA is preferred because of excellent insulation
characteristics and low shrinkage of the polymer it produces.
Casting and curing occur in closed vacuum chambers (CMA,
1983a). 4,4'-MDA has been specified in numerous nuclear power
plant applications (CMA, 1983a; Brechna, 1965). One manufacturer
of 4,4'-MDA-cured epoxy concrete coatings ships the amine in a
screw-top container into which the liquid epoxy resin is
poured. [Mixing thus occurs in the same vessel in which the 4,4'-
MDA is shipped, reducing dusting and worker exposure. However, in
one report (Williams e_t_ _al_. , 1974), 6 of approximately 300 men
who applied epoxy resins containing 4,4'-MDA to concrete walls at
a nuclear power plant developed clinical hepatitis 2 days to 2
weeks after starting work. 4,4'-MDA had been mixed with liquid
epoxide at the work site and applied to walls with trowels or a
spray gun. Exposure levels were not reported.
Special epoxy resins described as tetraglycidyl-
methylenedianiline (TGMDA)-derived resins, are reinforced with
glass, graphite, boron, or aramide fiber? and used in aerospace
and leisure products, structural adhesives, laminates, tooling
-------
69
and casting applications, and structures such as aircraft wings
and fuselages (Kirk-Othmer, 1980). The only point of exposure is
the introduction of the 4,4'-MDA into the process vessels. The
synthesis reaction is carried out in closed vessels, and the
4,4-MDA is consumed. Controls used in the operation are not
known. The maximum concentration reported in TGMDA manufacture
.was XJ.06 mg/m , but no further data were provided (CMA, 1983a).
A second reported incident of human exposure involved
production of a component of an insulating material (McGill and
Motto, 1974). Between 1966 and 1972, twelve male workers whose
job it was to manually mix 4,4'-MDA into an epoxy resin
contracted hepatitis. A thirteenth individual also contracted
hepatitis; his 4,4'-MDA exposure reportedly occurred during the
pulverization of 4,4'-MDA flakes for the process. Atmospheric
4,4'-MDA levels were measured at 0.1 ppm (analytical method not
reported) during the first survey. Workers who experienced only
inhalational exposure did not contract hepatitis. Each worker
who did contact the disease had at least one hand exposed to the
mixture for several hours per shift. Thus, the probable critical
exposure to 4,4'-MDA by the affected workers was through skin
contact. This is a consistent theme throughout the investigation
of workplace exposure to 4,4'-MDA.
About 100 employees were potentially exposed to 4,4'-MDA at
the facility. Various means were undertaken to reduce 4,4'-MDA
levels, including construction of exhaust ports and respiratory
protection described as breathing helmets. The process has
subsequently been automated to prevent worker skin exposure to
4,4'-MDA (McGill and Motto, 1974).
-------
70
c» Exposure in an Un-Characterized Setting
Vaudaine ^t^ ^1_. (1982) described a toxic agent monitoring
program at a Rhone-Poulenc Industrie facility,* focussing on
4,4'-MDA. Rather than monitor air levels of the chemical, these
researchers chose to monitor workers' urine for 4,4'-MDA on a
"present" or "not-present" basis.
During the period of 1970 to 1978, the chemical was handled
by workers in full protective 'suits ("divers' suits"), yet
workers' urine showed levels of 4,4'-MDA of at least 200 ug/1 in
144 of 965 samples (14.9%) in 1970. Dust from inside the suits
showed "fairly significant" levels of the chemical.
As the decade wore on, the need to avoid dust contamination
was gradually recognized, resulting in lower percentages of 4,4'-
MDA-positive urine samples and reduced levels of the chemical in
those samples that were positive. By 1978, the threshold of
detection of the amine in urine was lowered to 20 ug/1, and 2.7%
of samples contained 20 to 80 ug/1, 2.0% contained 80 to 200
ug/1, and 4.0% contained over 200 ug/1 — for a combined 8.7%
"positivity index."
/From 1978 through 1980, "there was an improvement in working
conditions brought about by cooperation among the company
physician, the manufacturing engineer and the shop personnel."
In 1980, the sampling program found 0.9% of urine samples
contained 20-80 ug/1, with no samples showing higher
* The report did not state the type of use or processing of 4,4'-
MDA carried out at the facility, but did mention empty drums
and 4,4'-MDA powder as being exposure concerns. Contact with
the authors is being sought.
-------
71
concentrations. This case clearly demonstrates that an informed
workforce and committed management can reduce exposure
significantly.
ii. Co-reactant in Polyurethanes
A small quantity of 4,4'-MDA, estimated at 50 thousand
pounds/year, is used as a polyurethane curing agent (Springborn,
1982). The primary use of 4,4'-MDA as a polyurethane curing
agent is with aliphatic isocyanates, in which system pot life is
long enough to permit use in spray applications.
The options for adding 4,4'-MDA to polyurethane batches are
the same as for epoxy curing. However, due to the smaller
quantities used, it is more likely that the operation may be done
by hand. This increases the opportunity for dermal and
inhalation exposure. Unit operations are the same as for epoxy
curing: addition of the 4,4'-MDA, mixing, and application or
formation of the product. There is no estimate of the number of
workers involved nor of the frequency and duration of exposure.
All three operations involve potential worker exposure to
4,4'-MDA, with the potential for skin contact and inhalation
depending on the work practices at the individual facility.
Controls used in these operations are not known. However, it is
likely that cloth gloves, at least, are used for protection from
heat from the 4,4'-MDA melting pot, hot molds, and other
equipment in the work place. As note above, cloth gloves are
permeable to 4,4'-MDA.
-------
72
Three cases of worker exposure to 4,4'-MDA during
polyurethane molding have been reported. In one case reported in
1976, during a 6 month period at least 8 employees in a
polyurethane molding plant developed dermatitis on skin areas
exposed to 4,4'-MDA, usually during the second or third week of
work (Emmett, 1976). Four other employees who worked as molders
during this time did not develop symptoms. No measurements of
exposure levels were.reported.
In another case, also reported in 1976, workers requested
that NIOSH inspect the facility due to the presence of a wide
variety of toxic chemicals, including 4,4'-MDA (NIOSH, 19765).
The 4,4'-MDA was used in the formulation of a polyurethane resin
used to make plastic belts. The operation involved preparation
of special molds, mixing of the polyurethane, and curing of the
polyurethane in molds to form belts. Liquid polyurethane resin
containing free methylene bis(4-cyclohexyl isocyanate) was mixed
with heated liquid 4,4'-MDA. The storage, heating, and mixing of
the chemicals were performed in closed, highly automated
systems. On the initial visit NIOSH industrial hygienists
noticed that conditions and work practices were not consistent
with good industrial hygiene practices. Uncovered containers of
4,4'-MDA were left open in the work area, and open containers of
waste chemicals were discarded in wastepaper baskets, exposing
janitorial workers to skin contact with these materials. During
a second visit, three months later in May of 1975, the work area
and work practices had been improved. Local exhaust ventilation
was being installed at the pouring and curing stations. Twelve
-------
73
air samples were collected on this visit and analyzed for 4,4'-
MDA. All samples showed 4,4'-MDA below the detectable limit of
0.05 mg/sample. Samples ranged from 2.9 to 13.42 liters. Thus,
airborne concentrations were below 3.8 mg/m . Dermal exposure
measurements were not made.
In a third case, NIOSH was asked by the Independent Union of
Rotameter Workers to investigate the health effects of asbestos
fibers and organic vapors upon workers at the Fischer and Porter
Company in Warminster, Pennsylvania (NIOSH, 1980). 4,4'-MDA was
being used to cure polyurethane that covered an epoxy system
employed to hold and encapsulate water flow measuring instruments
in water pipe. The 4,4'-MDA was weighed, heated in an oven,
mixed with the isocyanate, and poured on the previously set
epoxy. The molding was then ground and sprayed with enamel. Of
three ambient measurements taken by NIOSH, the highest 4,4'-MDA
level was 0.08 mg/m in the potting room near the oven.
iii. Wire Coating, Polyimides and PMR-15
Methylenedianiline is used as an ingredient in the
production of polyester-polyimide electrical conductor coatings
and polymeric amide-imide-ester wire enameling compositions.
4,4'-MDA is also used to produce a high-temperature-resistant
polyimide, PMR-15. PMR-15 is beginning to be used commercially
as a replacement for titanium in jet engine components. It is
planned for use in the engines of the F-18 fighter, the B-l
bomber, and new versions of the Boeing 747 and 767 airplanes
(CMA, 1983a).
-------
74
There are no quantitative data on exposure leve^-3 f°r these
uses of 4,4'-MDA. The processes involved indicate that partially
cured, or B-stage, resin systems are handled by wire coaters or
aircraft workers. The B-stage systems are then heated to effect
complete polymerization to cross-linked material. There would be
potential inhalational and dermal exposure to 4,4'-MDA if some of
that monomer were present in the B-stage resin.
Additionally, workers who charge 4,4'-MDA to the reactors in
which the B-stage polyimide is produced may be exposed to the
chemical at levels comparable to those experienced by epoxy
curative workers (see above).
iv. Other Uses
4,4'-MDA is used as a curing agent for polymer systems used
in fabricating nuclear weapons (DOE, 1983a, b, c, d). About
10,000 pounds/year of the amine is used at six sites, with the
greatest (>90%) usage at the Bendix Corp. facility at Kansas
City, Missouri (DOE, 1983b).
Personal air monitoring at the Bendix plant failed to show
detectable levels of 4,4'-MDA, while the highest recorded levet
(DOE, 1983c) at any DOE facility was 0.232 mg/m3. However, DOE
states that respirators are worn for the operation that involved
this exposure level in ambient air, and that dermal exposure is
prevented by the use of protective equipment.
Information on the quantity of 4,4'-MDA used in synthesis of
antioxidants for lubricating oils and greases and general
descriptions of the processes involved can be found in PEDCo
•»
(1983), though CMA (1984) indicates that no 4,4'-MDA is currently
being used for these purposes.
-------
75
in some of the non-MDI uses the purified form of 4,4'-MDA is
required, and the physical form may result in some "dusting"
during use. There are virtually no data on the extent of such
"dusting" in these workplaces. In order to estimate the extent
of possible dermal exposure to workers in these situations,
several different hypothetical cases are considered in
conjunction with dermal monitoring data (NIOSH, 1984a and b). In
one'scenario, worker protection is assumed to be equivalent to
that of workers packaging pure 4,4'-MDA at the manufacturing
site, and 4,4'-MDA would be handled throughout the shift. For
this scenario, dermal exposure calculations are made in the same
way as in the 4,4'-MDA manufacturing section. However, actual
exposures, especially dermal exposures, may be higher if
management and workers are less aware of the hazards of 4,4'-MDA
than their counterparts in the manufacturing setting, and
separate exposure calculations have been made for these
conditions. Another situation covers workers in Department of
Energy contractor facilities where nuclear weapons are
fabricated.
As in the manufacturing setting, workers in these
processing/using cases may be exposed to 4,4'-MDA intermittently
or continuously as they handle the chemical while charging
reactors, mixing batches of ingredients for repackaging, etc.
Recognizing these variables, Table 10 presents lifetime average
daily dose estimates (LADDs) for workers in these cases of varied
exposure durations and settings. Details of each setting
considered here are given in Section V and in Appendix B.
-------
76
The inhalational component of the exposures and LADDs in
Table 10 is based on the mean of airborne exposure level
measurements reported by Ameron (1983) and cited in subsection a.
above. These measurements were made in the vicinity of a work
station where dry ingredients for a protective coating resin
system were packaged. For the DOE workplace no dermal component
is included. Information received from DOE (DOE, 1983 a,b,cfd)
indicates that special care is taken in these workplaces, where
handling of highly toxic or otherwise hazardous substances is
routine, to preclude dermal contact with 4,4'-MDA. Monitoring
data from DOE (1983c) were used to estimate inhalational
exposures and LADDs. The highest recorded exposure level, 0.232
mg/m , was not used in the calculations since DOE indicates (DOE,
1983c) that respirators are worn during this operation. Exposure
to 0.02 mg/m tor 0.2 hours/day was used in the calculations.
B. Potential Exposure Related to Consumer Contact With
4,4*-MDA-Containing Articles or Products
There is no evidence that 4,4'-MDA is used in consumer
products. The Chemical Manufacturers Association recently
sponsored a survey (CMA, 1984) of 312 companies, that buy 4,4'-
MDA from U.S. producers. Sixty-one companies, representing 47%
of the merchant market of 2.64 million pounds in 1982, responded.
The results indicated no consumer products containing the
unreacted chemical. With 53% of the merchant market unaccounted
-------
77
TABLE 10
4r4'-MD& USING/PROCESSING WORKPLACE LADDs
2.5 hrs.
Exposure Duration Per Week
8 hrs.
20 hrs.
40 hrs.
Minimal Dermal Exposure^a'
Duration (Appendix B, Section 2)
LADD by Route in mg/kg/day
Dermal
Inhalational
Total
Continuous Exposure
(Appendix B, Section 3)
LADD by Route in Jig/kg/day
Dermal
Inhalational
Total
Variable Exposure Durations
(Appendix B, Section 4)
LADD by Route in mg/kg/day
Dermal
-Liihalational
Total
bhort-Term Exposures
(Appendix Bf Section 5)
LADD by Route in mg/kg/day
Dermal
Inhalational
Total
DOE contractors
(Appendix B, Section 6)
LADD by Route in mg/kg/day
Dermal
Inhalational
Total
Short-Term Intermittent
Exposure-Best Industrial Hygiene
(Appendix Bf Section 7A)
LADD by Route in mg/kg/day
Dermal
Inhalational
Total
0.00040
0.00066
0.0011
0.000016
0.00066
0.00068
0.0032
0.0021
0.0053
0.0067
0.0052
0.012
0.0010
0.0021
0.0031
0.0011
0.00066
0.0018
nil
0.000013
0.000013
0.0042
0.0052
0.0094
(b)
0.000028
0.0052
0.0052
0.010
0.010
0.020
0.16
0.010
0.17
0.0095
0.010
0.020
0.000028
0.010
0.010
-------
78
TABLE 10 — CONTINUED
2.5 hrs.
Exposure Duration Per week
3 hrs. 20 hrs. 40 hrs.
Appendix B, Section 7B)
IADD by Boute in mg/kg/day
Dermal
Inhalational
Total
Hypothetical Workplace
Standard in Effect
(Appendix B, Section 8)
IADD by Route in mg/kg/day
Dermal
Inhalational
Total
0.00028
0.0013
0.0013
0.00028
0.0026
0.0026
0.00011
0.00022
0.00033
The Exposure Duration Headings for this case correspond to various delay periods between
time of dermal exposure and removal of the 4,4'-MDA by washing. Thus, 2.5-, 8-, 20- and
40- hours per per week headings correspond, for the dermal component of the LADD, to
wash-up 0.25, 2, 4 or 6 hours after exposure.
(b)
Exposure is for 1 hour per week.
-------
79
for, however, it is not possible to state absolutely that no
4,4'-MDA is present as such in consumer products, although use of
4,4'-MDA in production of TGMDA appears to account for most of
the "missing" material. Aggregated data from the TSCA section
8(a) reporting rule, reported by 4,4'-MDA manufacturers, indicate
that in 1982, 660 to 2,600 pounds of the chemical may have been
used by their downstream customers to manufacture consumer
articles from which "limited release" of 4,4'-MDA might occur
(Knutson, 1984). It does not appear that this quantity of 4,4'-
MDA, even were it in fact finding its way into "limited release"
consumer articles, would lead to chronic levels of exposure that
could cause significant cancer risks.
Two patents have been issued for products containing 4,4'-
MDA for the purpose of treating hair. Available information does
not indicate whether 4,4'-MDA is actually being used in this
application, although, as stated above, 4,4'-MDA suppliers claim
that there are no consumer uses for 4,4'-MDA (Springborn, 1983).
The Color Index lists two dyes made using 4,4'-MDA, CI 24750
(Acid Red 9, Milling Red R) and CI 42500 (Basic Red 9,
pararosaniline), but indicates that only CI 42500 is currently
being manufactured. The manufacturer listed is American Cyanamid
Organic Chemicals Division at the Bound Brook Works in New
Jersey. However, contact with the facility indicated the dye is
no longer produced. Contacts at the Dyes Environmental and
Toxicology Organization stated that they knew of no dyes being
produced using 4,4'-MDA (PEDCo, 1983). However, information
•»
submitted by CMA (1983a) indicates that 4,4'-MDA is a non-
isolated intermediate in production.
-------
80
C. Potential Exposure Related to Drinking Water and
Ambient Air Contamination
There is, at this time, no evidence that any drinking water
supply is contaminated with 4,4'-MDA. Surface waters that are
used for drinking water supply, however, do receive effluents,
directly or indirectly, from 4,4'-MDA, MDI and polyurethane
manufacturing operations, which could contain 4,4'-MDA. [MDI can
hydrolyze to 4,4'-MDA when a large excess of water and favorable
mixing conditions are present. It is converted to a stable area
derivative under other conditions (CMA, 1983b)].
Likewise, groundwater used for drinking water supply might
be contaminated through migration of 4,4'-MDA from various
wastes. At present, EPA has no information indicating that such
contamination has occurred, so no risk estimate for this
potential route of exposure can be made beyond that given in the
text discussing Table 11.
EPA is studying the potential for exposure through surface-
supplied drinking water by measuring the amount of 4,4'-MDA
discharged from the treatment works of a 4,4'-MDA/MDI
manufacturing plant. If significant levels are found, steps will
be taken to determine the fate of the discharged 4,4'-MDA. The
Agency is taking this approach, rather than attempting to measure
4,4'-MDA levels at drinking water intakes, because of the current
limitations on analytical sensitivity. A level of about 0.3 ug/1
in drinking water produces an estimated added lifetime risk of
developing cancer by humans of about one in one million (see
Table 11), whereas current analytical methods have a practical
detection limit of about 1-10 ^ug/1.
-------
81
If a significant level of 4,4'-MDA is being discharged, data
on stream dilution and environmental fate will be obtained and
used together with discharge data to calculate possible doses of
4,4'-MDA received by people drinking contaminated water.
In the interim, while this investigation proceeds, estimates
of 4,4'~MDA releases and the associated, consequent doses and
risks have been made. These release estimates are summarized in
Table 11 and explained in further detail below. Because some of
the information used in these estimations has been claimed
confidential, the identities of the companies involved are not
disclosed. Only four of the seven 4,4'-MDA manufacturing plants
discharge waste water into streams that provide drinking water
supplies. These four sites are designated Plants A through D.
-------
82
TABLE 11
ESTIMATED 4,4'-MDA LEVELS IN SURFACE WATERS
Concentration
Plant Estimation mg/1
A Worst Case 5.0 X 10~4
Best Estimate 1.8 X 10~4
a) Worst Case 1.2 X 10~3
Best Estimate 3.0 X 10~4
Worst Case 4.8 X 10~3
Best Estimate 1.2 X 10~3
10"6 Excess Risk 2.8 X 10~4
a' Both plants discharge into the same stream.
The concentrations presented as "Worst Case" and "Best
Estimate" are derived from assumptions about production volume,
process losses, and operating schedules, detailed in Versar,
(1983b), coupled with information on each plant site, such as
total plant effluent flow rate (from waste water discharge
permits) and hydrologic characteristics of the receiving
streams. Instantaneous and complete mixing is assumed to take
place as the effluent enters the river, and no partitioning to
suspended solids or degradative processes are assumed to operate
to remove the 4,4'-MDA from the water column. Since it is
expected that some degradation in the receiving stream,
especially photo - and chemical-oxidation, does occur (EPA,
1982), and since these plants treat their effluent prior to
discharge (CMA, 19835), these estimated concentrations are
believed to be considerably higher than those that are actually
occurring.
-------
83
The concentration estimation titled "10 Excess Risk" is a
calculated value, derived by assuming a 70-year lifetime
consumption of 2 liters/day of contaminated water by a person
weighing 70 kg and using tumor incidence of hepatocellular
carcinomas and adenomas in female mice with the multistage model
(Crump, 1980) upper 95% confidence limit. This tumor type was
selected because it afforded one of the highest estimated risks
of all observed tumor types in the bioassay (NTP, 1983a), and
represents a conservative assumption.
1. Releases from 4,4'-MDA Manufacturing (Versar,
1983b)
During manufacturing, it is expected that nearly all of the
4,4'-MDA releases will occur from neutralization, aniline
separation, and product purification. These production
procedures, described in Section IV A above, are briefly
summarized below.
Neutralization; The crude reaction product is acidic and
must be neutralized. It is piped to a tank where concentrated
aqueous sodium hydroxide is added (Moore, 1978). The water and
salt created in this step are not released at this point, but
piped, with the product, to the next step.
Aniline separation; During aniline separation, the organic
layer is separated and washed with water and then sent to a
distillation facility which removes unreacted aniline (Moore,
1978). This step probably releases the largest quantity of 4,4'-
MDA in the form of contaminated wastewater.
-------
84
Purification; Not all 4,4'-MDA is purified. Little
information is available on this step. Either distillation or
re-crystallization processes could be employed. Some releases,
in the form of contaminated water and organic wastes, are
possible.
Gaseous wastes contaminated with 4,4'-MDA have been reported
to be insignificant (JRB, 1980; ESE, 1981). Furthermore, it is
expected that 4,4'-MDA will not volatilize from the aquatic
environment because of its extremely low Henry's law constant
(ratio of vapor pressure to aqueous solubility). It is therefore
assumed that airborne 4,4'-MDA emissions will be negligible.
No information could be found quantifying 4,4'-MDA water
releases from manufacturing. These releases will depend on
several factors, including: production level, water
requirements, processing conditions, and the level of pollution
control. Since most of these data are unavailable, it was
necessary to use a number of simplifying assumptions in
estimating releases, and these are presented below. Again, the
effluent monitoring work, now underway, is expected to provide
data in this area.
a. The ratio of grams of process and wash
water to grams of product is 4.3 to 1 (Perkins, 1968; Ramney,
1972; Powers, 1970; MATHTECH, 1982). The assumptions used to
derive this value are presented in Appendix A of Versar (1983b).
b. It is assumed that both the process- and
the wash water will be saturated with MDA (1 gm of MDA/liter of
water).
-------
85
Therefore, total MDA water releases from manufacturing
become:
187,988 kkg of MDA x 4.3 gm H20 x 1000 liters .
yr gm MDA kkg of water
x 1 x 10"6 kkg = 808 kkg of MDA released
gm yr
This is a worst-case estimate. In another estimate, it was
assumed that the 4,4'-MDA concentration in the wastewater
(process and wash water) will reach only 10% of the maximum
solubility (0.1 gm MDA/liter of water) due to the salt effect and
the dilution of process water with relatively uncontaminated wash
water. For this latter case, the total MDA releases are 80.8
kkg/yr*. It should be noted, however, that it is possible for
sodium chloride to complex 4,4'-MDA (CMA, 1983a) under certain
conditions, and, thus, sodium chloride might enhance, rather than
diminish, the aqueous solubility of 4,4'-MDA.
2. Releases from 4,4'-MDA Use as a Feedstock (Versar,
1983b)
Most of the 4,4'-MDA produced in the United States (about
90%) is captively converted to MDI at the manufacturing site;
approximately 9% of the total production is converted to MDI at
other locations, and the remaining production is used in other
applications (Springborn, 1982; ESE, 1981).
Background data and some release estimates for the use of
4,4'-MDA as a feedstock are discussed below. Most of the
* Aggregated data reported by manufacturers undei TSCA Section
8(a) indicate that up to 180 kkg of 4,4'-MDA is lost during
^ _ ,-) 1.U-.
-------
36
discussion deals with the production of MDI; however, some
information is also presented on the other uses. It was
estimated that in 1981-/ the 4,4'-MDA releases to air and water
from MDI manufacturing range from 26.9 to 269 kkg.
a. Releases from MDI Manufacturing
In 1982, approximately 361,000 kkg of MDI were produced;
this material was used to manufacture rigid and semiflexible
polyurethane foams -and urethane elastomers.
MDI releases from manufacturing could significantly affect
4,4'-MDA releases, since MDI can hydrolyze to 4,4'-MDA under
certain conditions. However, no quantitative data could be found
concerning MDI manufacturing releases. Therefore, release
estimates were based on a number of simplifying assumptions.
Assumptions for air and water releases are presented below.
Because the vapor pressure of MDI is low (5 x 10~ mm Hg at
25°C) (Woolrich, 1982), it was assumed that air releases of MDI
leaving the manufacturing facility will be very small.
Water releases were estimated based on the data found in
three patents concerning MDI manufacturing (Beck, 1958,- Hidetosh
et_ _ad_. , 1968; Pistor ^t_ _al. , 1977). These patents indicated that
process water is not used in the manufacture of MDI. One patent
(Hidetosh, 1968) called for the use of non-contact cooling water,
which is expected to remain uncontaminated.
Since MDI is known to react with water, no water is used to
wash MDI. However, it is expected that some water will be used
for maintenance, equipment cleaning, and work area washdowns. It
has been assumed that such water requirements for MDI production
-------
87
will be the same as those of 4,4'-MDA production. This is a
worst-case assumption since the organic 4,4'-MDA layer is known
to be washed with water. Wash water constitutes one-third of the
g
water used in 4,4'-MDA manufacturing, i.e. 8.1 x 10 1/yr.
Therefore, the water requirements for MDI manufacturing are
o
assumed to be 2.7 x 10 1/yr.
As a worst case estimate, it is assumed .that any MDI
contacted by wash water would be immediately hydrolyzed to 4,4'-
MDA, and that sufficient 4,4'-MDA would be produced for the water
to become completely saturated. As a more realistic case, it is
assumed that the water would only reach one-tenth of the
saturation point.
According to the worst-case assumptions, 4,4'-MDA releases
from MDI manufacturing become:
2.7 x 108 1/yr x 1 gm 4,4'-MDA x 1 x 10"6 kkg =
liter gm
270 kkg/yr of 4,4'-MDA released
For the more realistic case, the total releases of 4,4'-MDA
are estimated to be 27 kkg/yr*.
b. Releases from Other Product Manufacturing
Releases from the manufacturing of the IGF category products
listed earlier in Table 4 could not be estimated because of the
lack of information on the various manufacturing processes.
* At this time there are no TSCA Section 8(a) data on losses
during manufacture of MDI; however, EPA's monitoring of
effluents from a 4,4'-MDA/MDI manufacturing site is expected to
help determine the accuracy of release estimates made in this
section.
-------
88
3. Releases from Use of MDI in Polyurethane Manufacture
(Versar, 1983b)
No information could be located that quantified the MDI
(4,4'-MDA) releases from polyurethane manufacturing. However,
one report (Smith and LaSalle, 1974) provided total air emissions
data from polyurethane resin manufacturing for a chemical
analogous to MDI, toluene diisocyanate (TDI). Smith and LaSalle
8
(1974) reported emissions of 1.8 x 10 Ibs TDI released/lb of
polyurethane foam produced. Assuming that MDI and TDI emissions
are about the same, airborne MDI releases from a given
manufacturing site would be under 1 kg/yr, since MDI is used to
produce only a portion of polyurethane foams manufactured at
various sites. Thus, the MDI/4,4'-MDA air emissions from a
particular foam manufacturing site would be relatively
insignificant, especially since MDI has a lower vapor pressure
than TDI.
Numerous factors influence water releases of MDI/4,4'-MDA
including the following: production volume, manufacturing
process, amount of excess diisocyanate, water requirements,
processing conditions, ratio of other constituents to MDI, and
level of pollution control. Information on these factors, as
they relate to quantitative releases, could not be found.
Therefore, the following assumptions were necessary to roughly
estimate MDI/4,4'-MDA water releases from polyurethane
manufacturing:
o From the patent literature (Metzler, 1971), it was
estimated that the water requirement for polyurethane
-------
89
manufacturing is between 0 and 10% of all other ingredients. The
exact percentage depends on the type of polyurethane foam being
manufactured and the production process. The worst case scenario
was assumed to require 10% water.
o The amount of process water consumed during the
production of polyurethane foam is not known. As a worst-case
scenario, it was assumed that a negligible amount of water will
actually be consumed in this process. Therefore, process water
releases were still assumed to be 10% of all other- raw materials.
o Also from the patent literature (Metzler, 1971), it was
found that more than the stoichiometric amount of diisocyanate
>
(MDI) was required to complete the reactions and ensure proper
foaming. As a worst case, it was assumed that MDI will rapidly
hydrolyze to 4,4'-MDA and that there will be sufficient excess
MDI present for the resulting 4,4'-MDA to reach maximum
concentration (1 gm of MDA/liter) in the process water.
o According to Youer (1969), in conventional polyurethane
processing, the finished yolyurethane foam is steam cured and
washed. Furthermore, the manufacturing site will use water for
maintenance, equipment cleaning, and work area washdowns. It is
assumed that the miscellaneous water requirements will be equal
to 50% of the process water requirement. This is the same worst
case assumption used for 4,4'-MDA manufacturing. It was assumed
that the 4,4'-MDA concentration in this wash water (from
hydrolysis of unreacted MDI) is 10% of the maximum solubility,
i.e. 0.1 gm MDA/liter of water. This is a worst case assumption
•»
since actual emissions from polymer formation operations are
expected to be insignificant (Hedley _et_ _al_., 1975).
-------
90
o It was assumed that MDI is used to produce approximately
40% of all polyurethane products manufactured in the United
States. This is based on information that indicates TDI is used
in slightly more than 50% of all polyurethane products and that
other non-MDI isocyanates are used in approximately 10% of all
polyurethane products (Suh, 1980). Based on this information,
MDI is used to manufacture approximately 501,600 kkg of
<
polyurethane foam.
o MDI and 4,4'-MDA are not removed from the effluent during
waste treatment.
The calculations for 4,4'-MDA releases from polyurethane
manufacturing are given below:
Releases from process water
Amount of wastewater:
501,600 kkg x 0.10 = 50,160 kkg x 1000 liters/1 kkg
= 50,160,000 liters
Releases:
50,160,000 liters x 1 gm 4,4'-MDA/liter = 50,160,000 gm
= 50.2 kkg of 4,4'-MDA released in process wastewater
Releases from wash water
Amount of wash water: 25,080,000 liters
Releases:
25,080,000 x 0.1 gm 4,4'-MDA/liter = 2,508,000 gm
= 2.5 kkg of 4,4'-MDA released in wash water
Total release (from all polyurethane manufacture)
=50.2 kkg +2.5 kkg =52.7 kkg of 4,4'-MDA
-------
91
As a more realistic case, it was assumed that all the
process water is consumed during the foaming reaction or recycled
during the production of polyurethane foam. Therefore, the only
water releases would be from wash water. The wash water releases
for this case are assumed to be the same as those in the worst-
case scenario. Consequently, for the more realistic case, 2.5
kkg of 4,4'-MDA are assumed to be released to surface waters.
4. Releases from Polyurethane Products (Versar, 1983b)
One report (David, 1969) stated that there may be trace
quantities of isocyanate encapsulated in polyurethane foam.
However, most of this isocyanate, especially MDI which has a low
vapor pressure, is expected to remain in the foam polymer
matrix. Thus, it was assumed that 4,4'-MDA releases from such
polyurethane products will be insignificant.
NIOSH (1981) reported finding 4,4'-MDA in the off-gas from a
spandex fiber sample heated to 200°C to simulate a heat-forming
operation in the apparel industry. This report is being
investigated to determine whether 4,4'-MDA was used, as such, in
the preparation of. the fiber, or if urethane linkages (known to
be labile at high temperatures) might have been broken (after
having been formed from the possible use of MDI in the fiber) to
release 4,4'-MDA. The results of this investigation could have
implications regarding potential exposures to 4,4'-MDA in a
workplace setting not previously identified.
5. Releases from Disposal of Wastes (Versar, 1983b)
No information could be found concerning the disposal
practices for 4,4'-MDA-containing wastes from the manufacture of
the amine or MDI.
-------
92
CMA (1984) reports that of the 312 users and processors it
surveyed, 38 reported some information on environmental releases,
though mostly in a qualitative fashion.
Twenty respondents reported on air emissions. Four reported
using scrubbers or baghouses to limit releases. Six firms
reported air emission values (other than "trace"). One reported
1315 pounds/year released based on mass balance, but there was
some question whether this included material removed by their
scrubber. Another reported 193 pounds/year released. Two
reported about 10 pounds/year. Another firm reported emission
levels at 1.3 parts per billion.
Eight companies (CMA, 1984) reported that their aqueous
effluent contain 4,4'-MDA; seven of these send their effluent to
a publicly owned treatment works. Of these seven, five said the
effluents contained "trace" amounts of the amine and the others
release 50 pounds/year by this route.
Of the 38 respondents (CMA, 1984), 33 reported sending solid
wastes, including spillage, clean-up materials, containers,
reactor rinse solvents and other 4,4'-MDA-containing solutions,
to landfills. Sixteen of these reported that all 4,4'-MDA was
"fully reacted" before being sent to the landfill. Eleven
companies reported disposing of 5 to 100 pounds/year to
landfills. Another firm reported landfilling 270,000 pounds/
year. Three firms reported incineration of waste 4,4'-MDA and
empty bags.
Other disposal methods reported included "RCRA" or "EPA"
disposal and use of waste chemical reclaimer companies.
-------
93
6. Potential Releases from Degradation of Polyurethane
Since about 98 to 99% of 4,4'-MDA production is converted to
MDI and PMDI and thence to various polyurethane products, and
since degradation of such polyurethanes could generate 4,4'-MDA,
there is concern over the fate of polyurethanes disposed of in
landfills.
An example of the behavior of monomers used as starting
materials in the manufacture of such plastics was submitted to
EPA by the International Isocyanate Institute (III, 1979). Ill
submitted the results of a research effort to determine whether
polyether diol-based polyurethane flexible foams made from
toluene diisocyanate, an analogue of MDI, would biodegrade under
the conditions of sanitary landfills, and whether corresponding
amine analogs, 2,4- and 2,6-toluene diamines (TDA), would be
released. Polyurethane foam made with C-labelled toluene
diisocyanates was subjected to three experimental media---sanitary
land fill medium, refuse compost medium, and parabrown earth
medium—of different bacterial activity for three months. The
sanitary fill medium and the refuse compost medium were subjected
to temperatures of 22°C and 50°C. After three months, at 50°C,
0.04% of the C-tagged starting activity in foam extracts was
identified as 2,4- and 2,6-TDA. At 22°C no TDA could be
detected, and no releaso of C02 was identified from any
experiments done with sanitary fill medium. In refuse compost
medium and parabrown earth medium, no detectable TDA was
formed. However, at 22°C and at 50°C, 0.01% and 0.1% of the
starting activity of the labelled foam was detectable as CC.
-------
94
The III paper concludes that polyurethane is very resistant to
microbial degradation. EPA believes that 4,4'-MDA in
polyurethane could be expected to behave in a manner similar to
TDA. Therefore, very little 4,4'-MDA release would be expected
in landfill situations; and the 4,4'-MDA released is expected to
be degraded chemically or microbially fairly rapidly. Since 50°C
is an extreme (though sometimes attainable) environmental
temperature, release of 4,4'-MDA during normal environmental
conditions would not be expected. However, even at the higher
temperatures, any 4,4'-MDA release would probably be very slow
and at very low levels. There are no monitoring data available
for 4,4°-MDA in the terrestrial environment.
7. Environmental Pate and Transport of 4,4'-MDA
Releases
a. Environmental Transport of 4,4'-MDA
Most 4,4'-MDA is converted at the site of its manufacture to
MDI, which is then used to produce polyurethanes. 4,4'-MDA can
be expected to be released as a waste during the conversion to
MDI. It was estimated above that most of the releases will be to
aquatic systems. Typical treatment processes used by 4,4'-MDA
production plants involve the discharging of aqueous waste to a
holding lagoon from which the wastewater is ultimately diverted
to a municipal sewage treatment plant (Young and Parker, 1978) or
surface waters.
In general, no significant quantities of gaseous or solid
wastes contaminated with 4,4'-MDA have been reported (JRB, 1980;
ESE, 1981), although one processor reportes landfilling 270,000
-------
95
Ibs. of waste that contains some 4,4'-MDA (CMA, 1984). It is not
expected that 4,4'-MDA will be transported from aquatic systems
to the atmosphere since its aqueous solubility (1,000 mg/1) and
low vapor pressure (10~7 torr at 25°C) make volatilization from
water unlikely (Callahan et al., 1979). Deposition in sediment
or sorption to soils also is unlikely for 4,4'-MDA dissolved in
water, because of its solubility and its partitioning
preference. Values calculated for the log octanol/water
partition coefficient are 1.76 (Kenaga and Goring, 1980), 1.84
(Banerjee jet_ _al_. , 1980), 1.88 (Leo and Hansch, 1979), and 2.52
(Chiou et_ _al_. , 1977). Values calculated for the log organic
carbon distribution coefficient are 1.79 (Karickhoff et al.,
1979), 1.90 (Briggs, 1973), 1.99 to 2.47 (Kenaga and Goring,
1980), and 2.62 (Karickhoff _et_ _al_. , 1979).
b. Photodegradation
Although no data have been found regarding photolysis of
4,4'-MDA in the aquatic environment, indirect evidence indicates
that photo-oxidation may be the major fate of the compound
released in aqueous waste. 4,4'-MDA crystals darken when they
are exposed to air (Moore, 1978).' Similar behavior of phenols
and other aromatic amines has been attributed to the formation
and photolysis of charge-transfer complexes with oxygen (Joschek
and Miller, 1966). Free radical intermediates and hydroxylated
products are reported.
Landrum and Crosby (1981) report that dilute aqueous
solutions of p-toluidine are oxidized rapidly enough to make its
••
aquatic fate difficult to study. The structure of 4,4'-MDA is
-------
96
sufficiently similar to p-toluidine (both bear benzylic
hydrogens) to expect similar photochemical reactivity. Moreover,
free radical intermediates of 4,4'-MDA, analogous to those which
can be postulated for p-toluidine, should be more readily formed
(Laity ^t_ _al_. , 1973) from structures such as 4,4'-MDA. Zabik and
Kawaguchi (1982) have shown that structurally analogous 4,4'-
methylenebis-2-chloraniline pnotodecomposes in water with a half-
•life of 3.69 hours.
c. Oxidation
Aromatic amines are susceptible to oxidation by a variety of
chemical oxidizing agents (Cason, 1948). The resultant products
are usually quinones. Aromatic amines, such as aniline and p-
chloroaniline, become more easily oxidized after being adsorbed
to the aquatic clay, montmorillonite (Cloos ^t_ ^1_. , 1979).
Similar enhanced reactivity toward oxygen also should be expected
with 4,4'-MDA.
d. Hydrolysis
There are no data to suggest that aromatic amines undergo
hydrolysis under environmentally relevant conditions. The
covalent bond of a substituent attached to an aromatic ring is
resistant to hydrolyis because of the high negative charge-
density of aromatic structures (Morrison and Boyd, 1973).
e. Volatilization
4,4'-MDA's solubility in water (1,000 mg/1) coupled with its
low vapor pressure (10~7 torr at 25°C) diminish the importance of
volatilization as an environmental transport process (Callahan et
al., 1979).
-------
97
f. Sorption
The adsorption by soil of two similar aromatic amines,
aniline and p-chloroaniline, has been found to depend both on
organic matter and clay content of the soil (Moreale and Van
Blade, 1976; Cloos £t_ ajL_. , 1979). The extent of partitioning,
however, between soil and water depends on the solute's relative
affinity for the soil and water phases. The values which have
been calculated for 4,4'-MDA's log organic carbon distribution
coefficient (1.79 to 2.62) and log octanol/water partition
coefficient (1.76 to 2.52) indicate that the compound will not be
strongly sorbed to soils, and that if it were sorbed, it would
probably be subject to leaching (BSE, 1981).
g. Bioaccumulation
Although no direct evidence has been found in the literature
regarding bioconcentration of 4,4'-MDA, bioconcentration factors
have been estimated which range from 1.83 to 46 (ESE, 1981).
Metcalf and Sanborn (1975) point out that compounds with
solubilities of 50 mg/1 or more generally have little potential
for aquatic bioaccumulation. Lu j5t_ ^1_. (1977) reported that
although benzidine was taken up by the organisms of their aquatic
ecosystem, it was not bioaccumulated, and it remained in
equilibrium with the benzidine dissolved in the water. The
structure of benzidine differs from the structure of 4,4'-MDA
only by the absence of the central methylene group.
h. Biodegradation
No information regarding the microbial degradation of 4,4'-
MDA was found. Subba-Rao and Alexander (1977), however, have
-------
98
reported that Pseudomonas putida, isolated from soil, was capable
of degrading structurally similar bis(p-nitrophenyl)-methane
slowly. Bis(p-hydroxyphenyl)methane was not degraded. Inasmuch
as activated sludge can be acclimated to degrade benzidine (Baird
et_ al_. , 1977), 4,4'-MDA could probably be treated in a similar
manner.
i. Summary
Photoxidation may be the major fate of 4,4'-MDA released in
aquatic waste. Volatilization, sorption, and bioaccumulation are
probably not important, leaving dilution as the principal pathway
for dissipating MDA which remains undegraded. Biodegradation
probably would become a viable process only in acclimated sludge.
8. Estimated Surface Water Concentrations of 4,4'-MDA
Concentrations of 4,4'-MDA in surface waters that could be
used as drinking water supplies were estimated using the release
estimates given above, information in EPA's Exposure Analysis
Modeling System (EXAMS), and confidential production data for
4,4'-MDA (and MDI) manufacturing plants derived from the TSCA
section 8(a) Information System. This material is discussed at
the beginning of Section - IV C above and summarized in Table 11.
9. Populations Potentially Exposed to Contaminated
Surface Waters
This section identifies and estimates the sizes of
populations potentially exposed to 4,4'-MDA through ingestion of
drinking water and through dermal contact with contaminated
surface waters. Populations exposed through other pathways, such
as ingestion of potentially contaminated groundwater, could not
-------
99
be identified and were not considered within the scope of this
report.
a. Potential Drinking Water Exposures
Drinking water may be contaminated by 4,4'-MDA released from
4,4'-MDA, MDI and/or polyurethane manufacturing plants. The
Vvater Supply Data Base (WSBD) was checked to determine whether
any drinking water intakes were located downstream of 4,4'-
MDA/MDI manufacturing plants. The WSDB is a computerized data
base, maintained by the Monitoring and Data Support Division
(MDSD) of EPA, that contains information on the location of
surface water utilities; the locations of the utilities'
treatment plants, intakes, and sources of raw water; the
populations served; and the average and maximum daily
production. Table 12 enumerates the populations served by
drinking water facilities at various points downstream from each
plant. 4,4'-MDA manufacturing plants not listed in the table do
not discharge upstream of any drinking water intakes (i.e., no
drinking water intakes are located between the plant discharge
point and the confluence of the receiving water with a salt water
body). As stated above, there is at present no indication that
any drinking water supply is contaminated with 4,4'-MDA. Ongoing
work will resolve the question of whether such contamination
occurs.
b. Potential Ambient Environmental Exposures
Populations may be exposed to 4,4'-MDA in the ambient
environment through inhalation or through dermal contact with
V
surface waters. Only the later will be considered here, since
-------
101
TABLE 12
POPULATIONS POTENTIALLY EXPOSED TO 4,4'-MD&
IN ERINKING fcftTER DCHJSTREAM OP
4,4'-MEft. MANUFACTURING PLANTS
Plant Location
Olin Chemicals Moundsville, WV
Receiving
Stream
Ohio River
*
Location of Drinking
Water Intake (Miles
Downstream frcm Plant) (b'c'd^
0.5
35.3
62.7
Exposed
•Population
900
2,500
25
Mobay(a)
BASF(a) and
ICI Americas
lnc.(a'e)
New Martinsville, Ohio River
W
Geismar, LA
Mississippi
River
Dupont
Belle, wv
Kanawha River
15.4
27.9
38.2
44.9
46.2
71.6
78.9
80.3
83.8
85.4
87.5
88.7
11.3
2,500
4,000
2,000
12,500
5,850
54,800
232,000
550,000
12,000
60,000
25,000
62,700
1,950
(a)
(b)
(c)
(d)
(e)
Also produces MDI.
All drinking water intakes between the Olin and Mobay plants
and the confluence of the Hocking River with the Ohio River
(a distance of about 100 miles from the Olin plant and about
60 miles from the Mobay plant) are listed.
All drinking water intakes located within 90 miles downstream
of the BASF and ICI Americas plants are listed.
All drinking water intakes between the Dupont plant and thex
confluence of the Kanawha River with the Ohio River (a
distance of about 70 miles) are listed.
ICI Americas owns Rubicon, which is the name used by CMA's
4,4-MDA Project Panel member that operates this plant.
Source: EPA's Water Supply Data Base,
-------
102
TABLE 13
POPULATIONS SWIMMING IN SURFACE WXER
NEAR 4,4'-MDA MANUFACTURING PLANTS
Plant
Location
Receiving stream .
Exposed population
Olin Chemical
Mobay
Mobay
BASF
ICI Anericas,
Inc. (Rubicon)
DuPont
Jniroyal
UpJohn
Moundsville,
WV
New Martinsville,
WV
Baytown, TX
Geismar, IA
Geismar, LA
Belle, WV
Deepwater, NJ
Naugatuck, CT
LaPorte, TX
Ohio R. (Marshall Co.) insignificant^3'
Ohio R. (Wetzel Co.)
Cedar Bayou
Mississippi R.
Mississippi R.
Kanawha R. (Kanawha Co.)
Delaware R. (Salem Co.)
Naugatuck R.
San Jacinto Bay
insignificant^3'
0
-------
103
V. QUANTITATIVE RISK ESTIMATION (COOK AND GRINDSTAFF, 1983;
Grindstaff, 1984)
In this seccion the method and results of the quantitative
estimation of cancer risks posed by 4,4'-MDA are presented.
The model used to extrapolate from the dose-response region
of the NTP bioassays down into the estimated human exposure range
was the one-stage version of the Crump (1980) multistage model.
Estimated human response to these exposures (a.k.a. the risk
level) were calculated based on tumor incidences at individual
sites for each sex and species as well as on pooling tumor
incidences in various sexes and species.
The human exposures, or LADDs, that were used in the risk
estimation are explained. They are based on monitoring data and
hypothetical constructions of a variety of workplace and drinking
water exposure situations.
A. Introduction
In National Toxicology Program bioassays (NTP, 1983a) for
the carcinogenicity of 4,4'-MDA, F344/N rats and B6C3F1/N mice of
both sexes were administered the dihydrochloride salt ad libitum
in 'drinking water at concentrations of 150 parts per-million
(ppm) and 300 ppm for 104 weeks. Controls of each species were
given no 4,4'-MDA. Results from these bioassays were used in
*
high-to-low dose extrapolations to derive human carcinogenic risk
following inhalation, dermal, and oral exposure to 4,4'-MDA.
While the bioassays were conducted using the dihydrochloride
of 4,4'-MDA, for the reasons set forth i.n Section III, this
assessment assumes that the results can be used to estimate
cancer risks associated with the parent amine. Likewise, while
-------
104
the bioassays involved drinking water exposures, because of the
demonstrated capability of 4,4'-MDA to penetrate human skin, to
be absorbed through the human gastro-intestinal tract, and to be
rapidly distributed in mammals, this assessment uses the
bioassays to estimate cancer risks for humans exposed dermally,
via inhalation or ingestion.
Certain tumor types displaying a consistent dose-related
increase in incidence in the bioassays were judged appropriate
for use in calculating human carcinogenic risk at that site
(Milman, 1984). These sites are listed in Table 14, along with
incidence rates of tumors in the test animals for different dose
levels. Also listed in Table 14 are data on four tumor types
whose incidence, while statistically signficant, do not permit
their use in quantitative risk estimations; this information is
discussed further in subsection B.3, below.
Also given in Table 14 are the incidences of those
statistically significant tumors useful in assessing risk
aggregated by species/sex, and a similar aggregation of such
tumor incidences in which the malignant tumors alone were
statistically significant. These aggregations have been proposed
4 1
for use in estimating total cancer risk that a substance might
pose (EPA, 1984).
The true mathematical relationship between dose and response
to 4,4'-MDA for animals or humans is not known. To provide an
indicator of risk, data from the NTP bioassays were fitted to the
Global 83 high-to-low dose extrapolation model (Crump, 1980),
modified to reflect the fact that only 2 positive dose-levels
-------
105
were suitable for use in the extrapolation. This method provided
carcinogenic risk estimates for human exposure to low levels of
4,4'-MDA by extrapolating from the higher levels of 4,4'-MDA
given to the test animals.
For the purpose of this assessment, several different sets
of exposure situations representing occupational and
environmental exposures to 4,4'-MDA, described above in Section
IV, were used to estimate carcinogenic risk. These exposures
include environmental exposure through drinking water and
workplace exposure situations. Lifetime average daily doses
(LADDs) for humans were calculated for each situation, and these
LADDs are given in Tables 19 and 20, along with estimated extra
lifetime cancer risks. These situations are described in detail
below, and the calculations are given in Appendix B.
Data were handled using two significant figures in this
section, except where it was obvious that more were appropriate.
Risk estimations are given with one significant figure.
B. Methods and Results
1. High-Dose to Low-Dose Extrapolation Model
. The NTP bioassay data were fitted to the Crump multistage
model (Crump, 1980). The multistage model has been in wide use
in the EPA since the summer of 1980, and is used by EPA's
Carcinogen Assessment Group to set air and water quality criteria
and standards. This model was expected to provide a good fit to
the dose-response data. Other models, such as the logit, probit,
Weibull and the gamma multi-hit were not appropriate for this
analysis, because the number of dose levels in the NTP bioassay
-------
106
was equal to the number of parameters being estimated using these
models. The maximum likelihood estimate of carcinogenic risk and
the upper 95% confidence limit of this risk were calculated for
the human exposure situations (LADDs) at eight sex- and species-
specific tumor sites in the test animals. Risk estimates were
also calculated based on incidences of all the statistically
significant tumors in male and female rats and in female mice,
and another set of risk estimates was derived from incidences of
all the statistically significant tumors in male rats and female
mice for which the incidences of malignant tumors alone were
statistically significant. The "pooling" of tumor incidences was
done only on the sexes/species indicated because it was those
sex/species which responded in the bioassay in a manner amenable
to mathematical analysis. No pooling of male mouse data was done
because only one tumor type showed monotonic dose-response, and
in the female rat there were no statistically significant
malignancies. Liver neoplastic nodules were not taken into
account in any of these calculations, even though the incidence
of this lesion was statistically significant, because the proper
use of such lesions in quantitative risk estimation has not yet
been satisfactorily resolved.
Upon examination of the risk levels presented in Tables 19
and 20, one notes that there is little or no differences between
the levels calculated for the two different kinds of tumor data
pooling. This means that most of the calculated risk is due to
malignancies. If the risk level from pooling tumor data in which
the malignancies alone are statistically significant had been
-------
107
much lower than that derived from pooling all tumors, then one
would conclude that the contribution of malignancies to risk was
relatively low. This was not the case.
Goodness-of-fit tests were performed to evaluate how well
the experimental data fit the model. P-values are presented for
each site, and these are given in Table 15.
2. Animal to Human Extrapolation
In order to extrapolate the expected response in humans to
various lifetime average daily doses of 4,4'-MDA, one must first
determine what the dose-response relationship is in the test
animals from the bioassay data. Then one must take the expected
human LADDs and convert them to animal LADDs using a species
conversion factor, and, using the model described above,
calculate the response that the animals would have shown at those
lower doses. This response is then represented as the "risk to
humans" — in reality, of course, it is the response one would
have expected in the test animals at doses equivalent to those
that humans receive.
a. Dose-Response in the Bioassay
The animal exposures, in ppm in drinking water, were
converted to mg/kg/day LADD by the following relation:
LADD (mg/kg/day) = d (ppm) F (kg/day)
W (kg)
where LADD is lifetime average daily dose, d is the concentration
of the test chemical in the animals' drinking water, F is the
amount of water a test animal consumes per day, and W is the
weight of the test animal. For this analysis it is assumed that
rats weigh 350 grams, mice weigh 20 grams, rats consume 20 grams
-------
108
of water per day, and mice consume 5.5 grams of water per day.
-------
109
Since human weight is assumed to be 70 kg and the rat and
mouse weight, 0.35 and 0.020 kg, respectively, the ratio of
animal dose to human dose is 0.17 for rats and 0.066 for mice.
Hence, for rats,
Human LADD = 0.17 x Animal LADD
and for mice,
Human LADD = 0.066 x Animal LADD
3. Tumors Observed in the Bioassay
Table 14 presents the incidence of tumors from the NTP •
bioassays. Thirteen different species-, sex- and/or site-
specific tumors/lesions were observed in the bioassays at
statistically significant incidences, and 12 were judged
biologically appropriate for use in estimating risks to humans
(Milman, 1984). As stated above, there is at present no
consensus on the proper use of liver neoplastic nodules in this
sort of assessment, so that particular statistically significant
lesion in male rats was not used here. Further, since the
goodness-of-fit to the one-stage model displayed (Table 15) by
the data on male mice liver hepatocellular carcinoma, male mice
liver hepatocellular carcinoma and adenoma, female mice malignant
lymphoma, and female mice thyroid follicular cell carcinoma and
adenoma was either inadequate or marginal, no risk calculations
were made using these individual tumor data.
In order to obtain insight into the total risk for each sex
and species, the incidences of all statistically significant
tumors that each species and sex experienced in the bioassay were
tabulated. That is, for instance, in the case of female mice,
-------
110
all controls and test animals that had at least 1 tumor of the
statistically significant classes: all malignant lymphomas,
liver hepatocellular carcinoma or adenoma, thyroid follicular-
cell carcinoma or adenoma, or lung alveolar/bronchiolar carcinoma
or adenoma; were counted. Each animal bearing such a tumor was
counted only once, even if it had more than one of these tumor
types. Further, in order to obtain insight into the
contributions of benign tumors to the overall risk, a "pooling"
similar to that described above was done, but only for those
tumor types in which the observed malignancies, alone, were
statistically significant. In this assessment, rather than
present risk calculations in the text for all these sets of data,
estimates based on the individual and pooled tumor types showing
highest risk are given in the text, in Tables 19 and 20. Risk
estimates made using the remaining bioassay data are given in
Appendix C in Tables 19A and 20A.
In addition to the tumors in Table 14, other rarer tumors
were observed in the bioassay. These are listed below along with
their incidence in a large group of control animals and the
probability (p-value by Fisher1s Exact Test) of observing this
tumor by chance alone, given the incidence in the historical NTP
program control population. A bile duct adenoma was found in one
(p-value = 0.0136) 150 ppm dose male rat. This tumor had not
been previously diagnosed in 3,633 control male rats in the NTP
bioassay program. Transitional cell papillomas of the urinary
bladder were found in 2/50 (p-value = 0.0017) low-dose and 1/50
•»
(p-value = 0.0531) high-dose female rats compared with 3/3644
-------
Ill
TABIE 14
TUMOR INdDHO B3f SPECIES, SEX AM) SHE OF TUMOR.
r 1983a)
Control
150 ppn
300 ppn
Male Rats
Liver-Neoplastic Nodule (d)
1/50
P_<0.001 *
12/50 25/50
Pf=0.001(c) Pf<0.001
Thyroid-Follicular Cell
Carcinoma
Thyroid-Follicular Cell
Adenoma or Carcinoma
Female Rats
Thyroid-Follicular Cell
Adenoma or Carcinoma
Thyroid-C-cell Adenoma or
Carcinoma
Male Mice
Liver-Hepatocellular
Carcinoma
Liver-Hepatocellular Adenoma
or Carcinoma
Adrenal-Pheochromocytoma
Female Mice
Lung- Aveolar/Bronch iolar-
Adenoma or Carcinoma
0/49
Ps<0.001
PjfO.428
0/47
7/48
P£=0.006
1/49
Ps= 0.001
PL=0.590
0/47
Pg<0.001
PjfO.086
1/47
Pg= 0.055
PL=0.153
10/49
P =0.047
PL<0.001
17/49
Ps<0.001
PL<0.001
2/48
P,,=0.116
4/47
Pf=0.175
4/47
?£=0.058
5/47
Pf=0.102
33/50
P^O.OOl
43/50
Pf<0.001
12/49
p£=0.004
10/48
P£=0.005
19/48
7/48
Pf=0.032
29/50
Pf=0.001
37/50
Pf<0.001
14/49
PL<0.002
2/50
TJ =0.010
3/50
8/49
Pf=0.043
-------
112
TABLE 14 - GQNEINDED
All Malignant Lymphonas
Liver-Hepatocellular
Carcinoma
Liver-Hepatocellular
Adenona or Carcinoma
Thyroid- Poll icular cell
adenoma or carcinoma
Control
13/50
P^O.OOl
PL=O.IOS
1/50
Ps<0.001
Pjf 1.000
4/50
Ps<0.001
PL=0.700
0/50
Pg<0.001
150 ppm
28/50
P£=0.002
6/50
P£=0.056
15/50
Pf=0.005
1/47
Pf=0.485
300 pan
29/50
Pf=0.001
11/50
Pf=0.002
23/50
Pf<0.001
13/50
P<.001
POOLED DATA: ALL STATISTICALLY SIGNIFICANT TOMQRS
(e)
Male rats with thyroid
follicular cell
carcinoma or adenoma
Female rats with thyroid
follicular cell.
carcinoma or adenoma
or thyroid C-cell
carcinoma or adenoma
Female mice with liver
hepatocellular
carcinoma or adenoma,
Control
1/49
Ps=0.001
PjfO.590
1/47
Ps<0l001
P£=0.301
16/50
Ps<0.001
PjjO.052
150 ppn
4/47
P£=0.175
9/47
Pf=0.005
38/50
P£=0.002
300 ppn
10/48
Pf=0.005
25/48
Pf<0.001
44/50
Pf<0.001
carcinoma or adenoma,
lung alveolar/bronchiolar
carcinoma or adenoma or
any malignant lymphoma
-------
113
TABIE 14 - QGNTINDED
POOLED DKEA: ALL SEKT3STICAI12lf SIGNIFICANT TOMQRS
FOR miCH MUJGNMCIES AICNE ARE SIGNIFICANT
Control 150 ppn 300 ppm
Male rats with thyroid 1/49 4/47 10/48
follicular cell Ps=0.001 P £=0.175 p£=0.005
carcinoma or adenoma Pj==0.590
Female mice with 16/50(9> 36/50 43/50
liver hepatocellular Ps<0.001 P£=0.004 Pf<0.001
carcinoma or adenoma PL0.119
or any malignant
lymphoma
(a^ PS= two-sided p-value for positive slope. Small values indicate that the slope
is significantly different from zero, meaning that there is a tendency for
increasing dose to be associated with increasing response.
^ ' r two-sided p-value for departure from linear trend. Small values indicate
rf
th
at the association between dose and response is not linear.
P£= Fisher Exact Test p-value. Small values indicate response in the control
animals is statistically significantly different from the response in the
control animals.
At this time there is no consensus on the proper use of this tumor type in
quantitative risk estimations, so it is not used in this assessment.
Male mouse data were not used because response did not increase monotonically
with dose.
Female rats did not display a response in which malignancies alone were
statistically significant.
NIP report did not identify the animal (could it have been Minnie?) bearing a
liver hepatocellular carcinoma, so this numerator could change by one.
-------
114
control female rats in the bioassay program. Finally, granulosa
cell tumors were found in the ovaries of 2/50 (p-value = 0.0140)
high dose female rats and 3/50 (p-value = 0.0009) low-dose female
rats and 1 (p-value = 0.0283) of the latter tumors was a
granulosa-cell carcinoma. Granulosa-cell tumors were identified
in 11/3462 controls, and granulosa-cell carcinomas have been
observed in only 1/3462 control animals.
• fables 16 and 17 give the historical tumor incidence for a
large group of control animals from the NTP bioassay program
(1983b) for mice and rats, respectively. Included are all tumors
with a spontaneous frequency of at least 0.5%. They are
presented to put the tumor incidence in the controls of this NTP
bioassay into perspective with the incidence of tumors in a large
historical control population. The incidence of specific tumors
in the control group of this bioassay appears to be consistent
with historical groups. For example, the incidence of tumors in
control animals appears to be high in three sites in this
bioassay, yet it is consistent with the historical group. In
control male mice 20% (10/49) developed hepatocellular carcinomas
and 35% (17/49) developed hepatocellular carcinomas or
adenomas. In the historical control population 21.3% (498/2334)
of the male mice developed hepatocellular carcinomas and 31.1%
(725/2334) developed either hepatocellular nodules, adenomas or
carcinomas. In this bioassay, 26% (13/50) of the control female
mice had a malignant lymphoma of the hematopoietic system
compared to 27.2% in the large control population.
-------
115
TRBIE 15
GOODNESS -OP-PIT TEST p-VAIDES FOR EMB
TOMDR TYPE IN TAB££ 14
Species
And Sex
Tumor Type
p-Value*
Male Rats.
Male Rats
Male Rats
Female Rats
Female Rats
Male Mice
Male Mice
Male Mice
Female Mice
Female Mice
Female Mice
Female Mice
Female Mice
Liver Neoplastic Nodules
Thyroid Follicular Cell Carcinomas
Thyoid Follicular Cell Adenomas
or Carcinomas
Thyroid Follicular Cell Adenomas .
or Carcinomas
Thyroid C-Cell Adenomas
or Carcinomas
Liver Hepatocellular Carcinomas
Liver Hepatocellular Adenomas
or Carcinomas
Adrenal Pheochromocytomas
Lung Alveolar/Bronchiolar
Adenomas or Carcinomas
All Malignant Lymphomas
Liver Hepatocellular Carcinomas
Liver Hepatocellular Adenomas
or Carcinomas
Thyrid Follicular Cell
Adenomas or Carcinomas
p> 0.995
0.750 < p < 0.900
0.900 < p < 0.950
0.025 < p < 0.050
p < 0.005
0.500 < p < 0.750
0.950 < p < 0.975
0.250 < p < 0.500
p > 0.995
0.250 < p < 0.500
* This p-value is from the X* goodness-of-fit test. The higher the p-value
the better the fit of the model to the data. Inadequate fits commonly show
p_< 0.05, marginal fits show 0.05 < p < 0.10, and adequate fits show p^.
0.10. A dash indicates that the test was not appropriate, as the number of
experimental dose levels was equal to the number of free parameters in the
model.
-------
116
TABLE 16
HISTORICAL INCIDENCES OF PRIMARY TUMORS(a) IN UNTREATED CONTROL
B6C3F1/N MICE. (NTP, 1983b)
Tumor Site
Lung
Alveolar/Bronchiolar
Adenoma
Alveolar/Bronchiolar
Carcinoma
Liver
Neoplastic Nodule or
Adenoma
Carcinoma
Nodule or Adenoma or
Carcinoma
Adrenal
Pheochromocytoma
Pheochromocytoma, Malignant
Thyroid
C-cell Adenoma
C-cell Carcinoma
Follicular Cell
Adenoma
Carcinoma
Reproductive System
Mammary Gland
Fibroadenoma
Adenocarcinoma
Hematopoietic System
Leukemia
Lymphoma
Le ukem i a/ Lymphoma ^ d '
Male
2343(b)
282(1?. l)(c)
119(5.1)
2334
240(10.3)
498(21.3)
725(31.1)
1903
28(1.2)
2(0.1)
2178
0(0.0)
0(0.0)
22(1.1)
5(0.2)
2343
0(0.0)
0(0.0)
2343
17(0.7)
230(12.0)
297U2.7)
Female
2468
131(5.5)
47(2,0)
2469
98(4.0)
101(4.1)
196(7.9)
2051
16(0.7)
0(0.0)
2203
2(0.0)
0(0.0)
40(1.8)
6(0.3)
2486
8(0.3)
40(1.6)
2468
52(2.1)
625(25.1)
677(27.2)
a' Includes all tumors occurring with a frequency of 0.5% or
greater.
( ' Number of animals examined histopathologically (or, for
certain lesions, the number of animals necropsied).
|<~! Numbers in parentheses are percentages.
(d) This combination is included because certain early studies in
the data base tended to use these terms interchangeable.
-------
117
TABLE 17
HISTORICAL INCIDENCE OP PRIMARY TUMORS IN UNTREATED CONTROL
F344 RATS (NTP, 1983b)
Tumor Site
Lung
Alveolar/Bronchiolar
Adenoma
Alveolar/Bronchiolar
Carcinoma
Liver
Neoplastic Nodule or
Adenoma
Carcinoma
Nodule or Adenoma or
Carcinoma
Adrenal
Pheochromocytoma
Pheochromocytoma, Malignant
Thyroid
C-cell Adenoma
C-cell Carcinoma
Follicular Cell
Adenoma
Carcinoma
Reproductive System
Mammary Gland
Fibroadenoma
Adenocarc inoma
Hematopoietic System
Leukemia
Lymphoma
Le ukemi a/ Lymphoma ' ° '
Male
2305(b)
35(1. 5}
20(0.9)
2306
78(3.4)
18(0.8)
96(4.2)
2280
388(17.0)
23(1.0)
2230
114(5.1)
84(3.8)
22(1.0)
17(0.8)
2320
51(2.2)
6(0.3)
2320
648(27.9)
51(2.2)
699(30.1)
Female
2345
18(0.8)
9(0.4)
2356
71(3.0)
4(0.2)
74(3.1)
2262
81(3.5)
11(0.5)
2265
111(4.9)
81(3.6)
10(0.4)
10(0.4)
2370
527(24.1)
48(2.0)
2370
414(17.5)
36(1.5)
448(18.9)
(d)
Includes all tumors occurring with a frequency of 0.5% or
greater.
Number of animals examined histopathologically (or, for
certain lesions, the number of animals necropsied).
Numbers in parentheses are percentages.
This combination is included because certain early studies in
the data base tended to use these terms interchangeably.
-------
118
TABLE 18
INCIDENCE OF ANY MALIGNANCY IN F344/N RATS
AND B6C3F1/N MICE BY SEX FOR DIFFERENT DOSE
LEVELS FROM NTP BIOASSAYS ON 4,4'-MDA.
F344/N Rats
Male
Total animals with'3'
tumors uncertain
Female
Control
19/50
1/50
16/50
150 ppm
15/50
13/50
20/50
300 ppm
17/50
25/50
14/50
B6C3F1/N Mice(b)
Male
Total animals witlv3'
tumor uncertain
25/49
1/49
38/50
40/50
Tumor diagnosis is uncertain, may be benign or malignant.
May include the same animals as above in each case.
(b)
Data for female B6C3F1/N mice not available.
-------
119
Table 18 shows the incidence of any malignancy observed in
the test animals in the 4,4'-MDA bioassays by dose level.
4. Exposure Situations
Several situations of human exposure were used to estimate
LADDs and human carcinogenic risk. The LADDs were calculated for
each situation1s combination of exposure concentration, exposure
duration, frequency and extent. The LADDs were then applied to
calculate carcinogenic risk using the one-stage model and tumor
incidence data given in Table 14. This section will describe the
various exposure situations which include both workplace and
drinking water situations.
a. Workplace Situations
Mathematical details of the calculation of LADDs from the
situations described here can be found in Appendix B, as can the
assumptions used in the calculations.
There are several important caveats regarding the
assumptions used to calculate workplace LADDs. First, while a
skin absorption rate of 1% per hour of deposited material was
used based on data for MBOCA, the actual penetration rate of
4,4'-MDA remains to measured. Second, while a uniform
permeability of skin to 4,4'-MDA was assumed (to give a uniform
absorption rate of 1% over all surfaces), it is well known that
different skin surfaces display different permeabilities, and
this element of uncertainty does not appear to be amenable to
resolution in this type of assessment. Third, in certain of the
hypothetical exposure cases studied the assumption is made that
-.
all 4,4'-MDA is removed by washing, and-preliminary data from the
-------
120
skin penetration study now underway indicate that soap and water
may not remove all the compound in a simple wash. Fourth, the
data used from NIOSH (1984a) were collected over six hour
periods. At the end of that time the hand pads were removed and
analyzed for 4,4'-MDA. The total amount collected over six hours
was divided by six to derive an hourly deposition rate. It is
not known whether, in fact, deposition occurred linearly with
time.
i. Case 1; 4r4*-MDA/MDI Manufacturing
Information supplied by CMA (1983c) indicates that workers
are potentially exposed to 4,4'-MDA for varying periods during
the work week. LADDs were calculated for workers in the three
exposure-duration classes that were reported by CMA, viz., 8
hours or less-, 9 to 20 hours-, and more than 20 hours-per-
week. For the first class, exposure for 8 hours per week was
assumed; for the second, 20 hours per week; and for the third, 40
hours per week.
Inhalational exposure levels were assumed to be to the
highest average 8-hour Time Weighted Average (TWA) values for the
three exposure duration classes .reported by the five 4,4'-MDA
manufacturers who are members of CMA1s 4,4'-MDA Project Panel.
It is believed that use of the highest average TWA was a
realistically conservative assumption. The highest of the TWA
ranges reported, viz. 7.2 mg/nr, was judged unrealistically high
because it was based on area monitoring in the vicinity of
fugitive emissions to which workers are not exposed throughout
their work time in the unit. Furthermore, the 7.2 mg/m3 value
-------
121
was obtained using the Marcali method, and analytical
interference from aniline is likely to have been significant.
Aniline would be more likely to be a chief constituent of
fugitive emissions from this sort of equipment than would
4,4'-MDA, because of physico-chemical properties and process
conditions.
Dermal exposures were also considered in this situation, and
the dermal information from NIOSH (1984a) was used to calculate
the dermal component of the LADDs. The data were obtained as
described in Section IV above.
ii. Case 2; 4,4'MDA Using/Processing
In this situation, the impact of washing off 4,4'-MDA from
exposed skin, either immediately after exposure or with delays of
2, 4 or 6 hours, is assessed. Workers are considered who may
handle the chemical for varying periods during the work day, viz.
0.5 hours-, 1.6 hours-, 4 hours-, or 8 hours per day as they mix
4,4'-MDA with other materials, charge reactors, or conduct
similar operations, and who then wash-up after some delay.
The inhalational component of the LADDs calculated for these
workers is based on measured air levels (personnel monitoring) of
4,4'-MDA reported by Ameron (1983), a firm that manufactures
epoxy surface coatings for concrete and steel structures and
corrosion resistant piping, among other products. The analytical
method employed by Ameron was to collect 4,4'-MDA on silica gel
tubes and then to measure the amount collected colorimetrically,
a modification of the Marcali method. The employees that were
•.
monitored were engaged in pulverizing, mixing, blending and
-------
122
packaging of dry 4,4'-MDA containing coatings products. Ameron
presented no dermal exposure data.
The dermal component of LADDs derived from this situation is
based on dermal monitoring reported in NIOSH (1984b). This
monitoring was conducted in the epoxy resin mixing room of a
facility than produces filament-wound piping and pipe fittings.
The worker was fitted with a "golf-glove-like" device which
covered glycerine-wetted cotton gauze pads on the palm and back
of the hand. The worker wore no protective gloves. The gauze
pads were collected and analyzed after the worker1s exposure, as
described in NIOSH (1984b). The worker was engaged for less.than
10 minutes in weighing about 150 Ibs. of granular*, 99% assay
4,4'-MDA into a resin mixing vat.
In the situations analyzed in this case, the hypothetical
worker is assumed to handle 4,4'-MDA for only about as long as
the worker studied in NIOSH (1984b) and to receive the same
dermal exposure as that worker, but the worker remains in a work
station where air levels are as indicated for the indicated
durations.
The hypothetical worker wears no gloves, and is exposed via
the palms at the level recorded on the palm of the worker who was
monitored. He or she is exposed through the rest of the hand,
forearms, face and neck at the level recorded on the back of the
hand of the worker who was monitored.
* Use of granular material had recently been instituted in hope
of diminishing the amount of dust formation and distribution
that apparently had been experienced'with the previously used
flake form of 99% assay 4,4'-MDA.
-------
123
LADDs are calculated for situations in which the
hypothetical worker washes thoroughly all exposed skin 0.25, 2, 4
or 6 hours after being exposed, completely removing all deposted
4,4'-MDA. These situations were analyzed in order to assess the
impact of washing after a limited-duration dermal exposure.
iii. Case 3: 4,4*-MDA Using/Processing
In this case the impact of shift-long exposure without the
protection of gloves is assessed.
Inhalational exposures are assumed to be at the level
reported by Ameron (1983), and to be for 8 hours per day.
Assumptions regarding worker weight, breathing rate, etc., used
in calculating doses are the same as in Case 1 and are given in
Appendix B.
Dermal exposure is assumed to occur at the rate experienced
by the worker who was monitored in NIOSH (1984b), namely 43
ug/cm2/10 minutes (250 ug/cm2/hr) for the palms and 4.6 ug/cm2/10
minutes (27 ug/cm2/hr) for the rest of the hands, forearm, face
and neck. Exposure is assumed to occur for four hours on these
body areas twice during the shift, with a mid-shift wash-up for
lun'ch, but presumably the face and neck are not cleansed until an
after-work shower, when all deposited 4,4'-MDA is removed.
Deposition is assumed not to continue during the mid-shift break,
though absorption of already deposited material does continue
during this time.
iv. Case 4; 4,4'-MDA Using/Processing
In this case, the impact of wearing protective gloves is
assessed. •
-------
124
Inhalation exposures and workplace routines considered here
are identical to those described in Case 3, above.
The dermal exposure, however, is continuous for 0.25, 2, 4
or 6 hours per day, and 4,4'-MDA is deposited on the skin at a'
linear rate derived from data in NIOSH (1984a) for the 4,4'-MDA
flakker-bagger. This worker wore mid-forearm length latex
gloves, inside of which was mounted a hand-pad device as
described above.
v. Case 5; 4,4*-MDA Using/Processing
In this case the inhalational exposures and work routines
considered are the same as in Cases 3 and 4 above.
The impact of wearing protective gloves during a single 0.5
hour per day dermal exposure, followed by immediate, thorough
washing to remove deposited 4,4'-MDA from hands and forearms is
assessed. 4,4'-MDA deposited on the rest of the upper body is
assumed to remain there until a shower at shift's end (6 hours
exposure).
vi. Case 6; 4r4'-MDA Using/Processing
This case assesses the impact of the level of worker
protection used by DOE contractors (DOE, 1983c) who use 4,4'-MDA
to fabricate nuclear weapons.
vii. Case 7; 4,4'-MDA Using/Processing
The impact of very protective industrial hygiene practices
is assessed in this case.
Inhalational exposures and work routines are similar to
those described in Cases 3, 4 and 5, except that exposure is for
either 4 or 8 hours per day, which relatively long exposures
-------
125
prompt the use of very protective industrial hygiene practices.
Also, another inhalational exposure level is used in addition to
the one reported by Ameron (1983). The additional level is the
one reported for a resin mixing operation at a filament winding
factory (CMA, 1983a). Thus, this case is divided into two
sections, reflecting the two different inhalational exposure
levels.
The industrial hygiene practices assumed in this case
include wearing gloves, as described above, during two 0.5 hour
exposure periods per shift*, during which dermal exposure could
occur, followed by immediate and thorough washing to remove
deposited material from hands and forearms. Use of a full face
shield and impervious outer garments during periods when dermal
depositon could occur is also assumed, thus restricting dermal
deposition of 4,4'-MDA to that which the worker in NIOSH (1984a)
experienced under the gloves.
viii. Case 8; MDA Using/Processing
Hypothetical Workplace Standard
This case presents a hypothetical situation in which a
workplace standard is in effect that mandates a 0.001 ppm (0.0081
mg/nr) 8-hr TWA airborne exposure level and use of protective
clothing described in Case 7 above. The worker is assumed to
handle 4,4'-MDA for one hour, to wash the exposed areas of the
hands following exposure to completely remove deposited amine, to
handle 4,4',-MDA for and additional hour in the second half of the
shift, and then to wash up, removing any deposited amine.
The 0.5 hour exposures occur at the beginning of each half-
shift.
-------
126
b. Drinking Water Case
In this case it is assumed that an individual consumes 2
liters of water daily over a 70-year lifespan, and that the
person weighs 70 kg during the lifespan and that 50% of the
ingested dose is absorbed.
5. Estimation of Risks for Exposed Populations
a. Workers
Table 19 summarizes the lifetime extra cancer risks for
workers in 4,4'-MDA/MDI manufacturing plants and in 4,4'-MDA
using and processing plants under conditions outlined above and
detailed in Appendix B. These estimates are based on data from
the bioassay (NTP, 1983a) results with female rats that developed
thyroid follicular cell carcinomas and adenomas, female mice,
pooling all statistically significant tumors, and female mice,
pooling all tumors in which malignancies alone were statistically
significant, using the one-stage extrapolation model. Tumors
were observed in both sexes of mice and rats at multiple sites.
Risk estimations based on data for other tumor types and other
tumor poolings are given in Appendix C, Table 19A.
b. People Drinking Contaminated Water
Confidence in the risk estimations for the drinking water
case is low because of the lack of data at this time on actual
releases of 4,4'-MDA into waters that could serve as drinking
water supplies and the parallel lack of information on the fate
of any such releases.
The LADDs shown in Table 20 were derived using the
assumptions stated above and the Best Estimate water
-------
127
concentrations for the locales A through D in Table 11.
Absorption of 50% of the ingested dose is assumed (Thies, 1983).
Risk estimates in Table 20 were based on the same
species/sex tumor data cited in the previous subsection for
worker risks, and additional estimates based on other tumor types
and other tumor poolings are given in Appendix C, Table 20A.
C. RJSK Characterization
Since we do not have sufficient quantitative data detailing
the carcinogenic effects of 4,4'-MDA in humans, this risk
assessment relies on the available data showing animal
carcinogenicity as the source of estimates of human risk from
exposure to the substance. Moreover, when, as here, animal
effects data are available only for exposures at a higher level
than the level of estimated human exposure, we rely on
statistical models to extrapolate the risk to animals from high
to low exposure, then derive human risk estimates from the low
exposure animal risk estimates. The confidence we have in the
estimates of human risk derived in this way is dependant on,
among other things, our degree of confidence that £he animal data
demonstrate a carcinogenic response, and that the character of
the human response will be comparable to the animal response.
Confidence also depends on how accurately we can quantitatively
estimate and compare the doses of the chemical received by the
animals with those received by humans, and on how well the
statistical model portrays the relationship between dose and
response. These matters are discussed in depth separately in
this risk assessment. The purpose of this section is to give a
-------
128
more brief, overall perspective on the results, highlighting some
of the more important observations.
There is evidence demonstrating 4,4-MDA's carcinogenic!ty in
animals. This evidence is chiefly' found in the NTP bioassays on
rats and mice. These studies were well designed and well
conducted, and they showed strong dose-response in both sexes of
both species. In addition to tumors of the liver and thyroid,
which were seen in both species, lymphomas, adrenal, and lung
tumors were observed at statistically significantly elevated
incidences in the mice. Tumors of the urinary bladder occurred
in rats.
Close structural analogues of 4,4-MDA have also been found
to be carcinogenic in the liver and/or thyroid of these two
rodent species.
A number of studies in addition to the NTP bioassays showed
carcinogenic activity in animals, and several others failed to
show such activity. All of these studies, however, were flawed
in terms of duration, number of animals exposed, pathology
reporting or overall design.
Several additional lines of evidence are consistent with a
conclusion that 4,4'-MDA will also be carcinogenic in humans.
The chemical is genotoxic, and it binds to DNA in vivo.
Thus, carcinogenic activity may occur, at least in part, by a
direct genotoxic mechanism involving attack on DNA. If this is
the mechanism, it should operate in humans as well as animals.
Additionally, the chemical is a close structural analogue of
chemicals that are carcinogenic in both a'nimals and humans.
-------
129
Moveover, while the one available epidemiology study is limited
by the presence of confounding exposure to epoxy compounds, the
fact that an elevated incidence of bladder tumors was observed—a
tumor type produced in humans by benzidine, a close structural
analogue of 4,4'-MDA—may be significant, especially in light of
the occurrence of urinary bladder tumors in rats in the NTP
bioassay. Furthermore, 4,4'-MDA is absorbed by humans.
All of these factors, combined with sufficient evidence of
carcinogenicity in animals give a high degree of confidence that
the character of the response in humans will be comparable to
that in animals. An EPA classification of this chemical as a
probable carcinogen in humans [B2] has therefore been assigned.
The chief sources of uncertainty in this assessment relate
to the different exposure routes that humans and the test animals
of the bioassays experience, and to the inherent uncertainties in
extrapolating human risks resulting from low level exposure based
on experiments with animals who experienced relatively high
exposures.
In the workplace, humans are exposed to 4,4'-MDA
intermittently, chiefly via dermal and inhalational routes, while
the test animals were exposed to the dihydrochloride continuously
via drinking water. As discussed earlier, in performing the risk
assessment, the assumption has been made that the dihydrochloride
dissociated to the free amine in the intestine. The differences
between the animals and human routes of exposure, and our lack of
certainty in estimating the absorption rate, and thus exact dose
V
received, affect the quantitative estimation of human risk.
-------
130
These uncertainties are explained throughout this assessment and
are highlighted below.
Confidence is high that humans absorb the chemical.
Evidence from several industrial hygiene reports shows that
dermal exposure to humans results in absorption of significant
amounts of 4,4'-MDA. The structure of the chemical indicates
that inhalational exposure will also result in absorption, and
gastro-intestinal absorption also occurs, as evidence by the
"Epping Jaundice" case described earlier.
The state-of-the-art of quantitatively assessing dermal
exposures under industrial chemical processing conditions is
relatively primitive, but is rapidly developing. Thus, while it
is clear that 4,4'-MDA is absorbed through human skin, and that
this phenomenon occurs in the chemical processing workplace, the
rate of deposition of 4,4'-MDA on the skin and the irate of
absorption of deposited material are not precisely known. The
estimates of each of these rates used in this assessment are
reasonable, based on existing experimental data. These estimates
will be refined using data from studies now ongoing to measure
the in vivo absorption rate of 4,4'-MDA through the skin of
animals, including rhesus monkeys.
There is some question regarding the coupled use of the
estimated deposition and absorption rates in an integral form to
assess dermal exposure in this assessment. It is the author's
view that such coupling is defensible, in that it uses all the
available experimental data, and the uncertainties are duly
noted. It should be noted that an alternate approach that has
-------
131
been used by EPA in the past, namely assuming 100% absorption,
would yield much higher risk estimates.
There are also uncertainties related to high-to-low-dose
extrapolation modeling and animal-to-human extrapolations. In
this case the quantitative risk estimates were made using only
tumor incidence that showed monotonic dose-response behavior
using the one-stage version of the linearized multi-stage model.
The fact that the maximum likelihood estimates of risk differ
from the upper 95% confidence limits generally by less than a
factor of 2 indicates a good fit of these experimental dose-
response data to the model used, and the linear shape of the
dose-response curve at the low doses expected from human
exposures is consistent with the apparent genotoxic mechanism of
4,4'-MDA carcinogenic action.
The reader is cautioned against an assumption that the
quantitative risk estimates made in this assessment represent the
true, known risk to humans. The risk estimates given here are an
upper bound estimate, not a declaration of actual risk levels.
-------
132
TABLE 19
ESTIMATED EXTRA LIFETIME RISK OP CANCER FOR WORKERS
(EPA, 1984)
ADDED RISK BASED ON
Exposure Setting
4,4'-MDA/MDI Mfg.
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4,4'-MDA Use/Proc.
Appendix B, Sec. 2
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
,• t
4,4'-MDA Use/Proc.
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Total LADD
(mg/kg/day)
0.0051
0.020
0.040
o.ooii
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
FRFC/A(a) FMPA^
MLE
2 X 10~3
7 X 10~3
1 X 10~2
4 X 10~4
2 X 10~3
4 X 10~3
7 X 10~J
6 X 10~2
2 X 10~4
1 X 10~3
3 X 10~3
7 X 10~3
6 X 10~4
U95CL(e)
3 X 10"3
1 X 10~2
2 X 10 2
5 X 10~4
3 X 10 3
6 X 10 3
1 X 10~2
8 X 10~2
3 X 10~4
1 X 10~3
4 X 10~3
1 X 10~2
9 X 10~4
MLE
5 X 10~3
2 X 10~2
4 X 10~2
1 X 10~3
5 X 10 *
1 X 10 2
2 X 10
1 X 10"1
6 X l6~4
3 X 10~3
7 X 10"3
2 X 10~2
2 X 10~3
U95CL
7 X 10~3
2 X 10~2
5 X 10~2
1 X 10~3
7 X 10~3
1 X 10 2
2 X 10 2
2 X 10"1
8 X 10~4
4 X 10~3
9 X 10~3
2 X 10~2
2 X 10~3
FMPM
MLE
4 X 10~3
2 X 10~2
3 X 10 2
9 X 10~4
4 X 10~3
1 X 10 2
2 X 10
1 X 10"1
6 X 10~4
3 X 10~3
6 X 10~3
2 X 10~2
2 X 10~3
(c)
U95CL
6 X 10~3
2 X 10~2
4 X 10 2
1 X 10~3
6 X 10~3
1 X 10~2
2 X 10~2
2 X 10"1
7 X 10~4
3 X 10~3
8 X 10~3
2 X 10~2
2 X 10~3
Appendix B, Sec. 6
8 hr/wk
0.000013
5 X 10
,-6
6 X 10
,-6
1 X 10
,-5
2 X 10'
r5
1 X 10
r5 i x io~5
-------
133
TABLE 19
CONTINUED
ADDED RISK BASED ON
Exposure Setting
Total LAPP
(rng/kg/day)
FRFC/A(a)
FMPA
(b)
FMPM
(O
MLE
U95CL
(e)
MLB
U95CL
MLE
Appendix B, Sec. 8
40 hr/wk
0.00033
1 X 10 4 2 X 10 4
3 X 10 4 4 X 10~4
U95CL
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
0.0052
0.010
0.0013
0.0026
2 X 10 3
4 X 10~3
5 X 10~4
9 X 10 4
3 X 10 3
5 X 10~3
6 X 10~4
1 X 10~3
5 X 10 3
9 X 10~3
1 X 10~3
2 X 10~3
6 X 10~3
1 X 10~2
2 X 10~3
3 X 10~3
4 X 10~3
8 X 10~3
1 X 10~3
2 X 10"J
6 X 10 3
1 X 10~2
1 X 10~3
3 X 10"3
3 X 10"4 - 4 X 10 4
Female rat, thyroid follicular cell carcinoma and adenoma. See Appendix C for additional risk data.
'Female mouse, pooled, all significant tumors.
Female mouse, pooled, all tumors in which malignancies alone are significant.
, MLE = maximum likelihood estimate of risk, multi-stage model.
U95CL = upper 95% confidence limit on estimate of risk, multi-stage model.
-------
134
TABLE 20
ESTIMATED EXTRA LIFETIME RISK OF CANCER
FRCM DRINKING WATER EXPOSURES
ADDED RISK BASED ON
4,4'-MDA cone. LADD
Locale . (mg/1 )
A 0.00018
B 0.00015
C 0.00015
B + C 0.00030
D 0.0012
FRFC/A(a) FMPA(b)
(mgAg/day) MLE
-------
135
VI. DISCUSSION
While this assessment focusses on quantitatively estimating
cancer risks, it should be noted that other health effects have
been linked with 4,4'-methylenedianiline (4,4'-MDA). The last
Federal health authority action on 4,4'-MDA, the 1976 Current
Intelligence Bulletin issued by NIOSH, (1976a), addressed the
chemical's acute toxicity to the liver. Retinopathy has also
been cited as a toxic effect (Schilling Von Canstatt et al.,
1966; NIOSH, 1981; Leong et_ al_. , 1984), as has acute
myocardiopathy (Brooks _et^ jal_. , 1979) and allergic dermatitis
(Emmett, 1976). Retinopathy has been observed in two animal
species, the cat and the guinea pig, and may have occurred in
humans, while liver toxicity and myocardiopathy have been
observed in humans.
These adverse health effects, coupled with the results of
the NTP bioassays on the chemical, evidence from mutagenicity
studies, evidence of the carcinogenicity of close structural
analogues in animals and humans, epidemiologic evidence and
evidence of present inhalational and dermal exposures to the
chemical in workplaces that lead to estimated extra lifetime
risks as high as from about one in one hundred to about one in
ten (B2) in certain situations, combine to make 4,4'-MDA a prime
candidate for exposure controls and hazard warnings to those
exposed.
Of special concern is the dermal route of exposure. This
route is insidious. Many workers and managers appear to be
unaware of its significance. For instance, in Vaudaine et al.
-------
136
(1982) it is reported that knowledge of the hazards of 4,4'-MDA
prompted the use of "divers' suits," an extreme measure, to
protect against exposure. Yet 15% of urine samples taken during
that time contained measurable levels of 4,4'-MDA which
subsequent events have shown to have resulted from unsuspected
dermal exposure..
Likewise-in Dunn and Guirguis (1979), even after workers had
been supplied with positive-pressure, supplied air breathing
apparatus to protect against inhalational exposure, cases of
jaundice were still observed among workers who were being exposed
dermally.
The liver toxicity reported in McGill and Motto (1974)
occurred only in workers who were dermally exposed, while co-
workers in the same work station who breathed the same air as
affected workers, but who did not touch the 4,4'-MDA-containing
resin system, were not affected.
While no cases of acute toxicity were reported, the exposure
of the resin mixer reported in NIOSH (1984b) is a concern in the
same regard. Apparently, attempting to lower the potential for
exposure, the company involved had recently switched from the
flake, form of the chemical to the granular form. Even with the
material of "lower dusting potential," (NIOSH, 1984b) the worker
involved received a substantial exposure. This might have
resulted, at least in part, because the worker assumed that
"dusting" had been eliminated, or at least reduced to a
sufficient degree to dispense with the use of protective
clothing. How many other examples of such inadvertant exposure
may there be in such workplaces?
-------
137
Clearly, as shown in Vaudaine et_ al_. (1982), once workers
and managers are made aware of the dermal exposure route and the
hazards it can pose, exposures can be reduced. Such an
educational program for those who manufacture, use or process
4,4'-MDA in the United States is clearly indicated by the
evidence of exposure and the carcinogenicity of the chemical.
While the dermal exposure route is particularly worrisome,
irihalational exposures are also of concern. The LADD
calculations in Appendix B, Cases 4 and 7 and the corresponding
risk estimations in Table 19 illustrate this. When dermal
exposures are sharply limited, inhalational exposures at levels
that have been reported by industry sources still result in
significant risks. It is obvious that attention to both routes
of exposure is called for in order to bring risks down. It is
also obvious from reports of airborne levels of 4,4'^MDA
submitted by CMA (1983a) that control of airborne levels is
feasible. Thus, control of airborne levels by use of engineering
methods and control of dermal exposures by use of protective
clothing and good industrial hygiene practices is feasible, and
these controls result in risk estimations in the range of about
one in ten thousand extra lifetime risk of cancer (Cases 7 and
8).
Aside from the overriding issue of acting expeditiously to
protect workers' health, there are scientific questions the
answers to which could help to reduce the uncertainties in our
understanding of the risks associated with 4,4'-MDA.
-------
138
A metabolism study in Fischer 344 rats coupled with worker
urine monitoring would help to relate the external dose that rats
received in the NTP bioassay to the internal dose that resulted
in the observed tumor incidence in the animals and to the
internal dose that exposed workers are receiving.
Follow-up of workers who were exposed and who were studied
by Vaudaine et_ al_. (1982), Dunn and Guirguis (1979) and McGill
and Motto (1974) could prove useful. Some of these exposures
occurred as long ago as 17 years and may have lasted for as long
as 9 years.
The skin penetration rate of 4,4'-MDA will be a valuable
tool in reducing uncertainty in this case, as will data on
effluent levels of 4,4'-MDA and on the fate of the chemical in
natural surface waters.
Development and use of an analytical method for measuring
airborne levels of 4,4'-MDA that does not suffer from
interference from aniline will be of value in reducing
uncertainty about exposures in the 4,4'-MDA manufacturing
workplace. Likewise, a kinetic study on the rate of deposition
of 4,4'-MDA on various parts of the body would be useful in this
regard. Such studies should not, however, impede exposure
control and hazard warning programs for this probable human
carcinogen.
-------
139
APPENDIX A
ANALYTICAL METHODS
A key issue in assessing workplace exposures to and
associated risks from 4,4'-MDA is the analytical methodology used
for measuring airborne concentrations of the chemical. A related
issue is the-determination of 4,4'-MDA levels in workers' urine
as a possible assessment tool for relating workplace exposures
with actual received doses. The latter issue is under study by
NIOSH, which is developing a protocol for measuring 4,4'-MDA in
urine. Dr. Mark Boeniger, of the Cincinnati office of NIOSH, is
the leader on this project.
Regarding analytical techniques for measuring airborne
levels of 4,4'-MDA, a summary of methods used in the past and
those undergoing development will be given here.
1. Marcali Method
This method, of which there are several permutations,
involves drawing air through a liquid collecting medium of acetic
and hydrochloric acids. The resulting amine salt is then
converted to the diazonium compound and thence to an azo dye by
coupling with 1-N-naphthylethylenediamine. The concentration of
the azo compound is then colorimetrically determined.
The major limitation of this method is its inability to
discriminate among 4,4'-MDA, MDI (which is hydrolyzed to the
amine in the collection medium), aniline (which is often present
at 4,4'-MDA manufacturing plants) or other aromatic amines. This
-------
140
limitation results in reports of 4,4'-MDA levels which may be
erroneously high when interfering compounds are present (CMA,
1983a).
2. Liquid Chromatoqraphic Methods
The most promising methods for measuring 4,4'-MDA levels in
air in the presence of MDI, aniline, and other aromatic amines
involve high-pressure liquid chromatographic (HPLC) separation of
analyte desorbed from a variety of collection media and converted
to elutable derivatives.
a. Acid-Treated Glass Fiber Filter Method
This technique, used by DOE contractors (DOE, 1983c),
involves uniformly coating a 37 mm glass fiber filter with dilute
sulfuric acid, followed by driving off the water in an oven. Air
containing 4,4'-MDA is drawn through the filter, which captures
the amine and stabilizes it toward re-volatilization and against
oxidative loss by converting the amine to the hydrogen sulfate
salt. The analyte is then desorbed using 0.26_N NaOH that is 5%
(v/v)acetonitrile and converted to the diacetyl derivative with
acetic anhydride. The resulting solution is then analyzed on a
high pressure liquid chromatograph using a solvent gradient
(water/ acetonitrile) and a UV detector.
DOE reports (DOE, 1983c) that the method is still under
t
study, but filter collection efficiency and recovery of 4,4'-MDA
from spiked samples are greater than 90%.
EPA and NIOSH are working to further validate this general
method.
-------
141
b. Glass-Fiber Filter/Silica Gel Method
NIOSH (1984a, b) conducted the referenced surveys using an
untreated glass-fiber filter in series with a silica gel tube for
the area monitoring portion of the work. The filter was designed
to trap particulate matter and the silica to collect vapor-phase
material. However, it has been discovered that severe loss of
4,4'-MDA particles from non-acid-treated filters occurs,
rendering the results of those studies suspect insofar as the
reported results of area monitoring are probably lower than
actual levels. Duplicate area and personnel samples were
analyzed by NIOSH and the company using this method (and the
Marcali method for area samples). Agreement among the samples
was not good, and air monitoring data from these NIOSH visits
were not used in this assessment.
c. Silica Gel Collection Method
This method (several permutations) involves collection of
4,4'-MDA from the vapor phase on treated or untreated silica
gel. Silica treated with diethylamine has been used to sample
atmospheres containing 4,4'-MDA, MDI, and dimethyl formamide.
The diethylamine converts MDI to a stable urea, rendering
chromatographically separable the MDI and 4,4'-MDA components
(Lipski, 1982).
Desorption has been accomplished with methanol or diethyl
ether, and the 4,4'-MDA has been analyzed using HPLC, with UV
detection, either per se or as a benzoyl derivative (CMA, 1983a).
-------
142
There is concern that use of a standard, untreated silica
gel collection system might not efficiently remove particles of
4,4'-MDA.
The assumption ha's been made in this assessment that results
obtained with the Marcali method can give an upper bound on
actual exposure levels (and is probably quite reliable in the
absence of interferences), while results obtained with the
untreated glass fiber filter/HPLC method would give a lower bound
on actual exposure levels.
-------
143
APPENDIX B
LIFETIME AVERAGE DAILY DOSE (LAPP) CALCULATIONS
1. 4,4'-MDA/MDI Manufacturing Workplace
Inhalational Component
Assumptions:
o 1.2 m3/hr breathing rate
o 2.50 day/year exposure
o 50% absorption of airborne 4,4'-MDA through the lung
o 40 year working career
o 70 year lifetime
o 70 kg worker weight
o exposure levels for different exposure durations are
highest values from average TWA columns in Table 5.
8 hrs/week = 1.6 hrs/day
AVE. TWA = 0.07 ppm = 0.57 mg/m3
0.57 mg/m3 X 1.2 m3/hr X 1.6 hr/day X 250 X 40 X 0..5 •* 70 kg
365 70
LADD = 0.0031 mg/kg/day
20 hrs/week = 4 hrs/day
AVE. TWA = 0.059 ppm = 0.48 mg/m3
0.48 mg/m3 X 1.2 m3/hr X 4 hr/day X 250 X_40_x °-5 * 70 k<3 =
365 70
LADD = 0.0064
40 hrs/week = 8 hrs/day
AVE. TWA = 0.07 ppm = 0.57 mg/m3
0.57 mg/m3 X 1.2 m3/hr X 8 hr/day X 250 X 40 X 0.5 * 70 kg =
365 "70"
LADD = 0.015 mg/kg/day
-------
144
Dermal Component
Assumptions:
o Deposition rate of 4,4'-MDA is uniform over time.
o Absorption through all skin areas is at a uniform rate
of 1%/hr of deposited material.
o Body surface areas (male and female averages, Snyder et
_al. , 1975) .
Face = 650 cm2
Back of neck = 110 cm2
2
Front of neck and V of chest = 150 cm
Chest and stomach = 3550 cm2
Back = 3500 cm2
Upper arms = 1320 cm2
2
Forearms = 1210 cm
Hands = 820 cm2
o Exposure of the back of the hand under a cotton glove
mimics exposure to the entire upper body (face, neck,
arms, back of hands, chest, stomach and back), which
area totals 10,900 cm2. Area of two palms is 410 cm2.
o Worker showers at the end of the shift, completely
removing all 4,4'-MDA remaining on skin.
f\
o Deposition occurs on both palms at a rate of 9 ug cm""'
hr~| and on the rest of the upper body at 2.5 ug cm
hr"1 (NIOSH, 1984a, Table IV). There is zero deposition
on other body surfaces.
o For 40 hr/wk workers, two 3.5 hour half shifts under the
above conditions is assumed, along with a 1 hour lunch
period during which no deposition occurs. It is assumed
that the hands and forearms are washed free of 4,4'-MDA
at lunch, and that absorption of material already
deposited on the rest of the upper body continues during
lunch.
Dose through the Palms (40 hr/wk)
Daily Dose =
2 X 0.01 hr"1 X 9 ug cm~2 hr'1 X 410 cm2 fjj*5 tdt
= 74 xi/2 X t2 ] ^'5 = 450 ug .
-------
145
Dose through back of hand and forearms (40 hr/wk)
Daily Dose =
2 X .01 hr"1 X 2.5 ug cm"2 hr"1 X 2830 cm2 f3)'5 tdt
= 140 X ^ X t2 ] J'5 = 860 ug
Dose through rest of upper body (40 hr/wk)
Daily Dose =
2 X .01 hr"1 X 2.5 ug cm"2 hr"1 X 9280 cm2 J^*5 tdt
+ 0.01 hr"1 X 2.5 ug cm"2 hr"1 X 3.5 hr X 1 hr X 9280 cm2 =
460 X ^ X t2 ]jj-5 + 810 = 3600 ug
Dose through the palms (8 hr/wk and 20 hr/wk)
Daily Dose =
0.01 hr"1 X 9 ug cm"2 hr"1 X 410 cm2]"? tdt where X = 4.0
1.6
10 X Vfc X t2 ] J = 80 ug
10 X ^ X t2 ] 1<6 = 13 ug
Dose through rest of upper body (8 hr/wk and 20 hr/wk
Daily Dose =
0.01 hr"1 X 2.5 ug cm"2 hr"1 X 10,900 cm2f£ tdt where x = 4.0
J 1.6
270 X ^ X t2 ] £ = 2200 ug
270 X^X t2 ] J'6 = 350 ug
Total daily dermal doses at:
40 hrs/wk = 4900 ug = 4.9 mg
20 hrs/wk = 2300 ug = 2.3 mg
8 hrs/wk = 360 ug = 0.36 mg
LADD = daily dose X 250 X _4£ 4 70 kg
365 70
Dermal LADDs for:
40 hrs/wk = 0.027 mg/kg/day
20 hrs/wk = 0.014 mg/kg/day
8 hrs/wk = 0.0020 mg/kg/day
-------
146
Total LADDs for Durations of
8 hr/wk 20 hr/wk 40 hr/wk
. in mg/kg/day
Dermal 0.0020 0.014 0.027
Inhalational 0.0031 0.0064 0.015
TOTAL 0.0051 0.020 0.042
2. 4,4'-MDA Dsing/Processing Workplace with Minimal Dermal Exposure
Duration and Delayed Wash-up Following Exposure
Inhalational Component
Assumptions:
o Same as above inhalational assumptions, and
o Respirable 4,4'-MDA concentration in air is 0.38 mg/m3
(Ameron, 1983; mean of range limits, Page 4, Section 5.3).
o Worker spends 0.5, 1.6, 4 or 8 hrs/day in a work station with
the above 4,4'-MDA air concentration.
2.5 hrs/wk = 0.5 hr/day
0.38 mg/m3 X 1.2 m3/hr. X 0.5 hr/day X 0.5 X 250 X 40 * 70 kg =
365 70
LADD = 0.00066 mg/kg/day
8 hrs/wk = 1.6 hrs/day
0.38 mg/m3 X 1.2 m3/hr X 1.6 hr/day X 0.5 X 250 X 40 * 70 kg =
365 70
LADD = 0.0021 mg/kg/day
20 hrs/wk = 4 hrs/day
LADD = 0.0052 mg/kg/day
40 hrs/wk = 8 hrs/day
LADD = 0.0iO mg/kg/day
Dermal Component
Assumptions:
o Same as above dermal assumptions, and
-------
147
o Physical form of 4,4'-MDA used is capable of producing
the same "dusting" experienced by the worker in NIOSH
(1984b).
o Worker wears no gloves or other protective gear,
o Worker actually handles 4,4'-MDA only long enough (about
10 minutes) to receive the dermal exposure cited in
NIOSH (1984b), Table IV, and no longer, though he/she
may remain in the same general work area longer.
o Worker thoroughly washes forearms, face, and neck 0.25,
2, 4, or 6 hours after handling 4,4'-MDA.
o Worker is exposed through the palms (410 cm2) at 43 ug
cm"2 and through the rest of the hands, forearm, face
and neck-V (2400 cm2) at 4.6 ug cm"2 (NIOSH, 1984b,
Table IV). . .
Dose through palms
Daily dose =
For 0.25 hr exposure:
.01 hr"1 X 43 ug cm"2 X 0.25 hr X 410 cm2 = 44 ug
For 2 hr exposure:
.01 hr"1 X 43 ug cm"2 X 2 hr X 410 cm2 = 360 ug
For 4 hr exposure:
.01 hr"1 X 43 ug cm"2 X 4 hr X 410 cm2 = 720 ug
For 6 hr exposure:
.01 hr"1 X 43 ug cm"2 X 6 hr X 410 cm2 = 1100 ug
Dose through rest of hands, forearms, face, and neck-V
Daily dose =
.01 hr"1 X 4.6 ug cm"2 X 0.25 hr X 2400 cm2 = 28 ug
For 2 hr exposure:
.01 hr"1 X 4.6 ug cm"2 X 2 hr X 2400 cm2 = 220 ug
For 4 hr exposure:
.01 hr"1 X 4.6 ug cm"2 X 4 hr X 2400 cm2 = 440 ug
-------
148
For 6 hr exposure:
.01 hr'1 X 4.6 ug cm~2 X 6 hr X 2400 cm:
TotaJ daily dose @:
0.25 hrs/day = 72 ug
2 hrs/day - 580 ug
4 hrs/day = 1200 ug
6 hrs/day = 1800 ug
LADD = daily dose X 250 X 40 -f 70 kg
365 70
660 ug
0.072 mg
0.58 rag
1.2 mg
1.8 mg
LADDs @: 0.25 hrs/day =
2 hrs/day =
4 hrs/day =
-6 hrs/day =
Total LADDs for Duration of
0.00040 mg/kg/day
0.0032 mg/kg/day
0.0067 mg/kg/day
0.010 mg/kg/day
2.5 hr/wk 8 hr/wk 20 hr/wk
in mg/kg/day
Dermal
Inhal.
Total
0.00040
0.00066
0.0011
0.0032
0.0021
0.0053
0.0067
0.0052
0.012
40 hr/wk
0.010
0.010
0.020
3. 4r4'-MDA Using/Processing Workplace Without Use of Gloves.
Exposure Duration for Entire 8-Hour Shift
Inhalational Component
The same inhalational component of exposure for the 40-
hours per week situation as given in Case 2, above, is
assumed, namely: LADD = 0.010 mg/kg/day.
Dermal Component
o Same,general assumptipns as in Case 2, except that the
worker is exposed to 4,4'-MDA deposition for the entire
shift (two 4-hour periods, broken by the mid-shift 0.5
hr break) at the rate given in NIOSH, 1984b, Table IV.C,
namely 43 ug cm~2/10 minutes (250 ug cm"2 hr"1), palms
(410 cm2); and 4.6 ug cm~2/10 minutes (27 ug cm"2 hr"1),
rest of hands, forearms, face and entire neck (2500
cm2).
o Worker washes off all deposited 4,4'-MDA from hands and
forearms at mid-shift break, leaving face and neck (910
cm2) un-washed, and removes all 4,4'-MDA with a shift-
end shower.
-------
149
Deposition does not continue during 0.5 hr mid-shift
break, though absorption of already deposited material
continues through this period.
2 X .01 hr"1 X 250 ug cm"2 hr"1 X 410 cm2 j*Q tdt
Dose through palms
*Q
2100 X l/2 x fc2 10= 17'000 U9
Dose through rest of exposed skin
2 X .01 hr"1 X 27- ug cm"2 hr"1 X 2500 cm2 j'o tdt
+ .01 hr"1 X 0.5 hr X 4 hr X 27 ug cm"2 hr"1 X 910 cm2
l 9 4
= 1400 X 1/2 X t2] + 490
0 ug
= 11,000 + 490 ug = 11,000 ug
Total daily dermal dose = 28,000 ug = 28 mg
LADD = daily dose X 250 X 40 * 70 kg
365 70
Dermal LADD = 0.16 mg/kg/day
Inhalational LADD = 0.010 mg/kg day
Total LADD = 0.17 mg/kg/day
4,4'-MDA as ing/Process ing Workplace with Better than Minimal
Industrial Hygiene and Variable Durations of Exposure
Inhalational Component
The same inhalational component of exposure, for the
same time periods, is assumed for this workplace setting as
in Case 2, above.
Dermal Component
Assumptions:
o Same general assumptions as in Case 2, above.
o Worker wears fresh mid- forearm- length neoprene/latex
gloves while handling 4,4'-MDA.
o No further dermal exposure occurs after handling 4,4'-
MDA and removing gloves.
-------
150
o Worker handles 4,4'-MDA for 0.25, 2, 4 or 6 hr/day,
using both hands freely so that there is no difference
between right- and left-hand exposure.
o Deposition of 4,4'-MDA occurs linearly with time, as in
NIOSH, 1984a (Table IV, Samples DM2A, DMB), at 4.2 ug
cm"2 hr"1 on the palms (410 cm2) and 0.7 ug cm""2 hr"1 on
the rest of the upper body (10,900 cm2). No other
dermal exposure.
o 4,4'-MDA penetrates upper body clothing and deposits on
the skin at the same rate as it penetrates the
neoprene/latex gloves, viz. 0.7 ug cm"2 hr"1.
o Mid-shift handwashing is not accounted for in these
calculations.
(O Workers shower at shift's end, completely removing
remaining 4,4'-MDA.
Dose through the palms
Daily Dose =
.01 hr"1 X 4.2 ug cm"2 hr"1 X 4'10 cm2f0tdt x = 0.25 hrs
. . __ 2 hrs
17 X 1/2 X t2]§'25 = 0.53 ug 4 hrs
o •? 6 hrs
17 X 1/2 X t2! g = 34 ug
17 X 1/1 X t2] § = 140 ug
17 X 1/2 X t2] § = 310 ug
Dose through rest of upper body
Daily Dose =
.01 hr"1 X 0.7 ug cm"2 hr"1 X 10,900 cm2 fgtdt x = 0.25 hrs
-) n ->* ' 2 hrs
76 X 1/2 X t2]2-25 = 2.4 ug 4 hrs
6 hrs
76 X 1/2 X t2! 3 = 15° u*
76 X 1/2 X t2] $ = 610 ug
76 X 1/2 X t2] = 1400 ug
-------
151
Total daily dose @: 0.25 hr/day = 2.9 ug = 0.0029 mg
2 hr/day = 180 ug = 0.18 mg
4 hr/day = 750 ug = 0.75 mg
6 hr/day = 1700 ug = 1.7 mg
LADD = daily dose X 250 X 40 * 70 kg
365 70
LADDs @ 0.25 hr/day = 0.000016 mg/kg/day
2 hr/day = 0.0010 mg/kg/day
4 hr/day = 0.0042 mg/kg/day
6 hr/day = 0.0095 mg/kg/day
Total LADDs for Duration of
2.5 hr/wk 8 hr/wk 20 hr/wk 40 hr/wk
in mg/kg/day
Dermal 0.000016 0.0010 0.0042 0.0095
Inhal. 0.00066 0.0021 0.0052 ,0.010
Total 0.00068 0.0031 0.0094 0.020
4,4*-BOA Using/Processing Workplace with Better than Minimal
Industrial Hygiene and Short-Term Exposures
Inhalational Component
Assumptions:
o Same general inhalational assumptions as in Case 1,
above, and
o Respirable 4,4'-MDA air concentration is 0.38 mg/m3
(Ameron, 1983)
o Worker is exposed only in this work station and only for
0.5 hr/day.
0.38 mg/m3 X 1.2 m3/hr X 0.5 hr X 0.5 X 250 X 40 * 70 kg =
365 70
LADD = 0.00066 mg/kg/day
Dermal Component
Assumptions:
o Same general dermal assumptions as in Case 4, above, and
o Worker has but one 0.5 hr exposure/day and thoroughly
washes hands and forearms immediately following the
exposure.
-------
152
o Area of forearms is 1210 cm .
o Remainder of the upper body (9280 cm2) is exposed for 6
hours before shower.
Dose through the palms
Daily dose =
.01 hr'1 X 4.2 ug cm"2 hr"1 X 410 cm2 J"Q tdt
17 X 1/2 X t2]g-5 = 2.1 ug
Dose through the forearms
Daily dose =
.01 hr'1 X 0.7 ug cm'2 hr'1 X 1210 cm2 J*0 tdt
8.5 X 1/2 X t2]0)'5 = 1.1 ug
Dose through rest of upper body
Daily dose =
.01 hr'1 X 0.7 ug cm"2 hr'1 X 0.5 hr X 9280 cm2 X 6 hr = 195 ug
Total daily dermal dose = 200 ug = 0.200 mg
LADD = 0.20 mg/day X 250 X 40 -fr 70 kg
365 70
= 0.0011 mg/kg/day
Total LADD (Inhal. and Dermal) = 0.0018 mg/kg/day
6. DOE Contractor Workplace (DOE, 1983c)
Assumptions:
o As described in the text, only inhalational exposures
occur
o Other inhalational assumptions in Case 1, above, apply
0.02 mg/m3 X 1.2 m3/hr X 0.2 hr/day X 0.5 X 250 X 40 * 70 kg
365 70
LADD = 0.000013 mg/kg/day
-------
153
7. A. 4,4*-MDA Using/Processing Workplace; Intermittent Daily
Exposure; Best Industrial Hygiene Practices
Inhalational Component
Same as Case 2 above, but with 4 or 8 hours total
exposure per day, delivering 0.0052 and 0.010 mg/kg/day
LADD, respectively.
Dermal Component
Assumptions:
o Worker handles 4,4'-MDA for 0.5 hrs in first half and
0.5 hrs in second half of shift.
o Worker wears neoprene/latex, mid-forearm gloves, a
suitable over garment arid full face shield while
handling the chemical, which limit exposure to
—palms (410 cm2) at a deposition rate of 4.2 ug cm~2
hr"1 (NIOSH 1984a, Table IV, Sample DM2A).
—back of hands (410 cm2) at a deposition rate of 0.7 ug
cm~2 hr"1 (NIOSH 1984a, Table IV, Sample DM2B).
o Worker receives no other dermal exposure
o Worker washes hands immediately after exposure,
completely removing any 4,4'-MDA.
Dose to palms
Daily dose =
2 X .01 hr"1 X 4.2 ug cm"2 hr"1 X 410 cm2 f'Q tdt
9 '5
34 X 1/2 X t2] Q = 4.3 ug
Dose to back of hand
Daily dose =
2 X .01 hr"1 X 0.7 ug cm"2 hr"1 X 410 cm2 j""Q tdt
*> .5
5.7 X 1/2 X t2] Q = 0.71 ug
Total daily dermal dose = 5.0 ug = 0.0050 mg
LADD = daily dose X 250 X 40 -» 70 kg
365 70
= 0.000028 mg/kg/day
-------
154
Total LADD = Dermal and Inhalational LADDs
For 4 hr/day inhalational exposure:
(0.000028 + 0.0052) mg/'kg/day
0.0052 mg/kg/day
For 8 hr/day inhalational exposure:
(0.000028 + 0.010) mg/kg/day
0.010 mg/kg/day
B. 4,4'-MDA Using/Processing Workplace; Intermittent Daily
Exposure; Best Industrial Hygiene Practices
Inhalational Component
Same general assumptions as in Case 2 above, except that
the airborne concentration of 4f4'-MDA is the same as that
reported by CMA (1983a) for the resin mixing operation in a
filament winding plant, namely 0.1 mg/nr (See p. 64), and
exposure to this level is for 4 or 8 hours per day.
4 hr/day
0.1 mg/m3 X 1.2 m3/hr X 4 hr/day X 0.5 X 250 X 40 4 70 kg
365 70
LADD = 0.0013 mg/kg/day
8 hr/day
LADD = 0.0026 mg/kg/day
Dermal Component
Same as in Case 7. A above.
Total LADD = Dermal + Inhalational LADDs
4 hr/day inhalational exposure;
(0.000028 + 0.0013) mg/kg/day
0.0013 mg/kg/day
8 hr/day inhalational exposure;
(0.000028 + 0.0026) mg/kg/day
0.0026 mg/kg/day
-------
155
8. 4,4'-MDA Using/Processing Workplace; Workplace Standard in
Effect
Inhalational Component
Same general assumptions, except that exposure is at 0.001 ppm
(0.0081 mg/m3) for 8-hours.
0.0081 mg/m3 X 1.2 m3/hr X 8 hr/day
X 250 X 40 X 0.5 4 70kg = 0.00022 mg/kg/day LADD
365 70
Dermal Component
Assumptions:
o Worker handles 4,4'-MDA for two 1-hour periods at the
beginning of each half-shift, following which deposited
material is immediately washed off.
o Deposition occurs on the palms (410 cm2) at 4.2 ug cm"2
hr"1 (NIOSH, 1984a, Table IV, Sample DM2A); on the back
of the hands (410 cm2) at 0.7 ug cm"2 hr"1 (NIOSH,
1984a, Table IV, Sample DM2B).
o All other dermal exposure is prevented.
Dose to palms
Daily dose =
2 X .01 hr"1 X 4.2 ug cm"2 hr"1 X 410 cm2 fQ tdt
•? 1
34 X 1/2 X t2] 0 = 17 ug
Dose to back of hand
Daily dose =
2 X .01 hr"1 X 0.7 ug cm"2 hr"1 X 410 cm2^ tdt
-> 1
5.7 X 1/2 X t^] Q = 2.8 ug
Total daily dermal dose = 20 ug = 0.020 mg
-------
156
LADD = daily dose X 250 X _40 * 70 kg
365 70"
= 0.00011 mg/kg/day
Total LADD » (0.00022 + 0.00011) mg/kg/day
= 0.00033 mg/kg/day.
-------
157
APPENDIX C
TABLE 19A
ESTIMATED EXTRA
LIFETIME RISK
OP CANCER FOR WXKERS
Extra Risk Based on Tumor Types
Exposure Setting
4,4'-MDA/MDI mfg. '
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4,4'-MDA Use/Proc'e'
Appendix B, Sec. 2
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Appendix B, Sec. 6
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
Appendix B, Sec. 8
Total IADD
nig/kg/day
0.0051
0.020
0.042
0.0011
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
0.000013
0.0052
0.010
0.0013
0.0026
0.00033
MRFC(a)
MLE(C>
5 X 10"4
2 X 10~3
4 X 10"3
1 X 10~4
5 X 10"4
1 X 10"3
2 X 10"3
2 X 10~2
7 X 10~5
3 X 10~4
7 X 10~4
2 X 10"3
2 X 10~4
1 X 10~6
5 X 10~4
1 X 10~3
1 X 10~4
2 X 10~4
3 X 10~5
U95CL(d)
9 X 10~4
3 X 10~3
7 X 10"3
2 X 10"4
9 X 10"4
2 X 10"3
3 X 10~3
3 X 10~2
1 X 10'4
5 X 10~4
1 X 10"3
3 X 10~3
3 X 10~4
2 X 10~6
9 X 10~4
2 X 10~3
2 X 10~4
4 X 10~4
6 X 10"5
MRFC/A(b)
MIE
1 X 10"3
4 X 10"3
7 X 10"3
2 X 10~4
1 X 10~3
2 X 10"3
4 X 10~3
3 X 10~2
1 X 10~4
5 X 10~4
1 X 10"f
4 X 10"3
3 X 10~4
2 X 10~6
9 X 10~4
2 X 10"3
2 X 10~4
5 X 10~4
6 X 10~5
U95CL
1 X 10' 3
6 X 10"3
1 X 10~2
3 X 10"^
1 X 1C"3
3 X 10~^
6 X 10~3
5 X 10~2
2 X 10~4
9 X 10~4
2 X 10"3
6 X 10'3
5 X 10~4
4 X 10~6
1 X 10~3
3 X 10"3
4 X 10"4
7 X 10~4
9 X 10~5
•\fj Male rat, thyroid follicular-cell carcinoma
°- Male rat, thyroid follicular-cell carcinoma and adenoma
c' Maximum likelihood estimate
(d) upper 95% confidence limit
^e* Remaining exposures are all in use/processing category
-------
158
TABLE 19A -CONTINUED
Extra Risk Based on Tumor Types
Exposure Setting
4,4'-MDA/MDI mfg.
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4,4'-MDA Use/Proc
Appendix B, Sec. 2
2.5 hr'/wk
8 hr/wk
20 hr/wk
40 hr/wk '
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Appendix B, Sec. 6
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
' Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
Appendix B, Sec. 8
Total IADD
mgAg/day
0.0051
0.020
0.042
0.0011
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
0.000013
0.0052
0.010
0.0013
0.0026
0.00033
FRCC/A(f) MMAP(9)
MTE
7 X 10"4
3 X 10"3
5 X 10"3
1 X 10~4
7 X 10"4
2 X 10"3
3 X 10~3
2 X 10~2
9 X 10~J
4 X 10~4
1 X 10~3
3 X 10~3
2 X 10"4
2 X 10"6
7 X 10"4
1 X 10~3
2 X 10"4
4 X 10~4
5 X 10~5
U95CL
1 X 10~3
5 X 10"3
9 X 10~3
3 X 10~4
1 X 10~3
3 X 10~3
5 X 10"3
4 X 10"2
2 X 10~4
7 X 10~4
2 X 10"3
5 X 10~3
4 X 10~4
3 X 10~6
1 X 10~3
2 X 10"3
3 X 10"4
6 X 10~4
8 X 10~5 '
MIE
9 X 10~4
3 X 10~3
7 X 10~3
2 X 10~4
9 X 10~4
2 X 10~3
3 X 10~3
3 X 10"2
1 X 10~4
5 X 10~4
1 X 10~3
3 X 10~3
3 X 10~3
2 X 10~6
9 X 10~4
2 X 10"3
2 X 10"3
5 X 10~3
6 X 10~5
U95CL
1 X 10"3
5 X 10" 3
1 X ICT2
3 X 10~4
1 X 10~3
3 X 10~3
5 X 10"3
4 X 10"2
2 X 10"4
8 X 10"4
2 X 10~3
5 X 10"3
5 X 10"3
3 X 10~6
1 X 10~3
3 X 10"3
3 X 10~3
7 X 10~3
8 X 10~5
• | Female rat, thyroid C-cell carcinoma and adenoma
(9) Male mouse, adrenal pheochromocytoma
-------
159
TABIE 19A -OONFINDH)
Extra Risk Based on Tumor Types
Exposure Setting
4,4'-MDA/MDA Mfg.
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4f4'-MDA Use/£roc.
Appendix B, Sec. 2
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Appendix B, Sec. 6
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
Appendix B, Sec. 8
Total IADD
mgAg/day
0.0051
. 0.020
0.042 .
0.0011
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
0.000013
0.0052
0.010
0.0013
0.0026
0.00033
FMABC/A^ FMIHC^
MIE
3 X 10~4
2 X 10~3
2 X 10"3
6 X 10~f
3 X 10~4
7 X 10~4
1 X 10~3
1 X 10~2
4 X 10~5
2 X 10"4
4 X 10~4
1 X 10~3
1 X 10"4
8 X 10"7
3 X 10'4
6 X 10"4
8 X 10~;>
1 X 10~4
2 X 10~5
U95CL
6 X 10'4
2 X 10"3
4 X 10"3
1 X 10"4
6 X 10~4
1 X 10~3
2 X 10~3
2 X 10~2
7 X 10~5
3 X 10~4
8 X 10~4
2 X 10~3
2 X 10"4
1 X 10"6
6 X 10~4
1 X 10~3
1 X 10~4
3 X 10"4
4 X 10~5
MIE
6 X 10"4
2 X 10"3
5 X 10~3
1 X 10~4
6 X 10"4
1 X 10"3
2 X 10"3
2 X 10~2
8 X 10~5
4 X 10"4
8 X 10~4
2 X 10"3
2 X 10~4
1 X 10"6
6 X 10"4
1 X 10~3
1 X 10"4
3 X 10"4
4 X 10"5
U95CL
9 X 1CT4
3 X 10"3
7 X 10~3
2 X 10"4
9 X 10"4
2 X 10"3
3 X 10" 3
3 X 10~2
1 X 10"4
5 X 10"4
1 X 10"3
3 X 10"3
3 X 10"4
2 X 10"6
9 X 10~4
2 X 10~3
2 X 10"4
. 5 X 10~4
6 X 10~5
•.' Female mouse, alveolar-bronchiolar carcinoma and adenoma
'^ Female mouse, liver hepatocellular carcinoma
-------
160
TABIE 19A -OUNL'INUED
Extra Risk Based on Tumor Types
Exposure Setting
4,4'-MDA/MDA Mfg.
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4,4'-MDA Use/E>roc.
Appendix B, Sec. 2
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Appendix B, Sec. 6
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
Appendix 3, Sec. 8
Total IADD
nig/kg/day
0.0051
0.020
0.042
0.0011
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
0.000013
0.0052
0.010
0.0013
0.0026
0.00033
FMIHC/A^ J ^ MRPA^ k)
MI£
1 X 10"3
5 X 10"3
1 X 10"2
3 X 10"4
1 X 10~3
3 X 10"3
5 X 10"3
5 X 10~2
2 X 10"4
8 X 10~*
2 X 10~3
5 X 10~3
5 X 10~4
4 X 10"6
1 X 10"3
3 X 10~3
4 X 10"4
7 X 10~4
9 X 10~5
095CL
2 X 10"3
8 X 10~3
1 X 10~2
4 X 10"4
2 X 10~3
5 X 10~3
8 X 10~3
6 X 10~2
3 X 10"4
1 X 10^
3 X 10 3
8 X 10"3
7 X 10~4
5 X 10~6
2 X 10~3
4 X 10~3
5 X 10~4
1 X 10"3
1 X 10~4
MCE
9 X 10~4
4 X 10~3
7 X 10~3
2 X 10"4
9 X 10~4
2 X 10"3
4 X 10~3
3 X 10~2
1 X 10~4
6 X ID-*
1 X 10 3
4 X 10"3
3 X 10"4
2 X 10~6
9 X 10"4
2 X 10~3
2 X 10"4
5 X 10~4
6 X 10~5
U95CL
1 X 10"3
6 X 10"3
1 X 10~2
3 X 10"4
1 X 10"3
3 X 10~3
6 X 10"J
5 X 10°2
2 X 10~4
9 X 10~4
2 X 10"3
6 X 10"3
5 X 10"4
4 X 10'4
1 X 10"3
3 X 10~3
4 X 10~4
7 X 10"4
9 X 10~5
Female mouse, liver hepatocellular carcinoma and adenoma
Male rat, pooled all statistically significant tumors
-------
161
TABIE ISA -CDNTINDBD
Extra Risk Based on Tumor Types
Exposure Setting
4,4'-MDA/MDA Mfg.
Appendix B, Sec. 1
8 hr/wk
20 hr/wk
40 hr/wk
4,4'-MDA Use/t>roc.
Appendix B, Sec. 2
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 3
Appendix B, Sec. 4
2.5 hr/wk
8 hr/wk
20 hr/wk
40 hr/wk
Appendix B, Sec. 5
Appendix B, Sec. 6
Appendix B, Sec. 7A
20 hr/wk
40 hr/wk
Appendix B, Sec. 7B
20 hr/wk
40 hr/wk
Appendix B, Sec. 8
Total LADD
mg/kg/day
0.0051
0.020
0.042
0.0011
0.0053
0.012
0.020
0.17
0.00068
0.0031
0.0094
0.020
0.0018
0.000013
0.0052
0.010
0.0013
0.0026
0.00033
FRPA(1)
MLE
3 X 10~3
1 X 10~2
2 X 10"2
6 X 10"4
3 X 10~3
7 X 10~3
1 X 10~2
9 X 10"2
4 X 10"4
2 X 10"3
4 X 10~3
1 X 10"2
1 X 10~3
7 X 10~6
3 X 10~3
5 X 10~3
7 X 10"4
1 X 10~3
2 X 10~4
U95CL
4 X 10~3
1 X 10~2
3 X 10~2
8 X 10"4
4 X 10~3
9 X 10~3
1 X 10"2
1 X 10'1
5 X 10"4
2 X 10~3
5 X 10"3
1 X 10~2
2 X 10~3
9 X 10"6
4 X IQ-f
7 X10"3
9 X 10"4
2 X 10"3
2 X 10~4
MRPM(ra)
MLE
9 X 10"4
4 X 10~3
7 X 10~3
2 X 10"4
9 X 10~4
2 X 10"3
4 X 10"3
3 X 10~2
1 X 10~4
5 X 10~4
1 X 10"3
4 X 10"3
3 X 10"4
2 X 10"6
9 X 10~4
2 X 10"3
2 X 10~4
5 X 10~4
6 X 10"5
U95CL
1 X 10"3
6 X 10"3
1 X 10~2
3 X 10"4
1 X 10"3
3 X 10~3
6 X 10"3
5 X 10"2
2 X 10~4
9 X 10~4
2 X 10"3
6 X 10"3
5 X 10"4
4 X 10~6
1 X 10"3
3 X 10"J
4 X 10~4
7 X 10"4
9 X 10"5
(1)
(m)
Female rat, pooled all statistically significant tumors
Male rat, pooled all tumors for which malignancies, alone, are statistically significant
-------
162
TABLE 20A
Estimated Extra Lifetime Risk of Cancer
Fran Drinking Water
Locale
A
B
C
B + C
D
4,4'-MDA Cone. LAED
mg/1 mgAg/day
0.00018
0.00015
O.OOC15
0.00030
0.0012
0.0000026
0.0000021
0.0000021
0.0000042
0.000017
Extra Risk
MRFC(a)
MI£(c)
2 X 10"7
2 X 10~7
2 X 10"7
4 X 10"7
2 X 10"6
U95CL(d)
4 X 10'7
4 X 10~7
4 X 10~7
7 X 10~7
3 X 10~6
Based on Tumors
MRFC/A(b)
MIE
5 X 10"7
4 X 10~7
4 X 10"7
7 X 10~7
3 X 10"6
U95CL
7 X 10"7
6 X 10~7
6 X 10~7
1 X 10~6
5 X 10~6
Male rat, thyroid follicular-cell carcinoma
Male rat, thyroid follicular-cell carcinoma and adenoma
Maximum likelihood estimate
Upper 95% confidence limit
-------
163
TABIE 20A - QONTINDED
Extra Risk Based an Tumors
Locale
A
B
C
B + C
D
4,4'-MDA Cone. IADD
mg/1 mg/kg/day
0.00018
0.00015
0.00015
0.00030
0.0012
0.0000026
0.0000021
0.0000021
0.0000042
0.000017
FRCC/A(e)
MIE
4 X 10~7
3 X 10~7
3 X 10~7
6 X 10~7
2 X 10~6
U95CL
6 X 10"7
5 X 10~7
5 X 10~7
1 X 10"6
4 X 10~6
I1MAP(f)
MLE
5 X 10~7
4 X 10"7
4 X 10~7
7 X 10"7
3 X 10~6
U95CL
7 X 10"7
5 X 10~7
5 X 10~7
1 X 10~6
4 X 10~6
Female rat, thyroid C-cell carcinoma and adenoma
Male mouse, adrenal pheochromocytoma
-------
164
TABIE 20A - ODNl'lNUtl)
Extra Risk Based on Tumors
locale
A
B
C
B + C
D
4,4'-MDA Cone. LACD
mg/1 mgA9/day
0.00018
0.00015
0.00015
0.00030
0.0012
0.0000026
0.0000021
0.0000021
0.0000042
0.000017
FMABC/A(9)
MIE
1 X 10"7
1 X 10~7
1 X 10"7
2 X 10~7
1 X 10~6
U95CL
3 X 10~7
2 X 10~7
2 X 10"7
4 X 10"7
2 X 10~6
FMIflC(h)
MIE
3 X 10~7
2 X 10"7
2 X 10~7
5 X 10~7
2 X 10~6
U95CL
5 X 10"7
4 X 10"7
4 X 10~7
7 X 10"7
3 X 10~6
Female mouse, alveolar-bronchiolar carcinonia and adenoma
Female mouse, liver hepatocellular carcinoma
-------
165
TABIE 20A - ODKTINDED
Extra Risk Based on Tumors
Locale
A
B
C
B + C
D
4,4'-MDA Cone. IADD
mg/1 mgAg/day
0.00018
0.00015
0.00015
0.00030
0.0012
0.0000026
0.0000021
0.0000021 •
0.0000042
0.000017
FMLHC/A(i) MPPA^
MZ£
7 X 10~7
6 X 10~7
6 X 10"7
1 X 10"6
5 X 10~6
U95CL
1 X 10~6
8 X 10"7
8 X 10~7
2 X 10"6
6 X 10"6
MLE
5 X 10~7
4 X 10"7
4 X 10~7
7 X 10'7
3 X 10~6
U95CL
7 X 10"7
6 X 10~7
6 X 10~7
1 X 10~6
5 X 10"6
(i)
.• Female nojjse, liver hepatocellular carcinona and adenoma
Male rat, pooled all statistically significant tumors
-------
166
TABIE 20A - OJNT1NUU)
Extra Risk. Based on Tumors
Locale
A
B
C
B + C
D
4,4'-MDA Cone.
- rag/l
0.00018
0.00015
0.00015
0.00030
0.0012
LADD
mgAg/day
0.0000026
0.0000021
0.0000021 '
0.0000042
0.000017
FEPA
MIE
1 X 10"6
1 X 10"6
1 X 10"6
2 X 10~6
9 X 10~6
(k)
U95CL
2 X 10"6
1 X 10~6
1 X 10~6
3 X 10~6
1 X 10"5
MEE
MLE
5 X 10~7
.4 X 10"7
4 X 10""7
7 X 10~7
3 X 10"6
,M(D
U95CL
7 X 10"7
6 X 10"7
6 X 10"7
1 X 10~6
5 X 10~6
Female rat, pooled all statistically significant tumors
Male rat, pooled all tumors for which malignancies, alone, are
significant
-------
167
REFERENCES
Ameron, 1983. Letter: Reference 4,4'-Methylenedianiline. C.D.
Jones to E. Klein. June 29, 1983.
Andersen, M., Binderup, M-L., Kiel, P., Larsen, H., Mixlid, J,,
1980. Mutagenic Action of Isocyanates Used in the Production of
Urethanes. Scand. J. Work Environ. Health _6, 221 (1980).
Baird, R., Carmona, L. and Jenkins, R.L., 1977. Behavior of
Benzidine and Other Aromatic Amines in Aerobic Wastewater
Treatment. J. Water Pollut. Control Fed. 49, 1609 (1977).
Banerjee, S., Yalkowski, S.H. and Valvani, S.C., 1980. Water-
Solubility and Octanol/Water Partition Coefficients of
Organics. Limitations of the Solubility-Partition Coefficient
Correlation. Environ. Sci. Technol. 14, 1227 (1980).
Barnes, J.R., 1964. Toxicity study on "MOCA" 4,4'-methylene-bis-
(2-chloroaniline). E.I. DuPont de Nemours and Company (Inc.) MR
No. 652 and MR No. 13-D. Wilmington, DE.
Beal, D.D., 1982. HERD Input Into the MOCA PRL-1 Document.
Intra-agency Memorandum. Health and Environmental Review
Division, Office of Toxic Substances, U.S. Environmental
Protection Agency. August 4, 1982.
Beck, T.R. (inventor), E.I. DuPont de Nemours and Company
(assignee), 1958. Process for the manufacture of organic
isocyanates. U.S. patent 2,822,373. February 4, 1958.
Braselton, W.E., Williams, D.E. and McCormack, K.E., 1982.
Pharmacokinetics of 4,4'-Methylenebis (2-chlroaniline).
(Manuscript data April 29, 1981). MBOCA—1982: Research results
and recommendations for environmental and occupational levels,
Lansing, MI, March 10, 1982. Sponsor: Michigan Toxic Substance
Control Commission, Lansing, MI.
Brechna, H., 1965. Effect of Nuclear Radiation on Organic
Materials; Specifically Magnetic Insulation in High-Energy
Accelerators. Technical Report, Contract AT (04-3)-515, U.S
Atomic Energy Commission, San Francisco Operations Office. March
1965.
Briggs, G., 1973. A Simple Relationship Between Soil Adsorption
of Organic Chemicals and Their Octanol/Water Partition
Coefficients. Proceedings of the 7th British Insecticide/
Fungicide Conference, pp. 83-86
-------
168
Brooks, L.J., Neale, J.M., and Pieroni, D.R. , 1979. Acute
Myocardiopathy Following Tripathway Exposure to
Methylenedianiline. JAMA. 242, 1527 (1979).
Brusick, D., 1976. Response of Monsanto Company to TSCA 8(d)
Rule on 4,4'-MDA.
Callahan, M.A., Slimak, M.W. , Gavel, N.W. , May, I.P., Fowler,
C.F. et al. , 1979. Water-related Environmental Fate of 129
Priority Pollutants. Volume I, p. 2-lff. Washington, DC: U.S.
Environmental Protection Agency. EPA-440/4-79-029a.
Cason J., 1948. Synthesis of Benzoquinones by Oxidation.
Organic Reactions. _4, 305 (1948).
Chadwick, D.H., 1981. Organic isocyanates. In: Encyclopedia of
Chemical Technology. Volume 13. 3rd edition. New York: John
Wiley and Sons. pp. 789-818.
Chin, B., Tobes, M.C., and Han, S.S., 1983. Absorption of 4,4-
Methylene bis (2-chloroaniline) by Human Skin. Environ. Res. 32,
167 (1983).
Chiou, C.T., Freed, V.H., Schmedding, D.W. and Kohnert, R.L.,
1977. Partition Coefficient and Bioaccumulation of Selected
Organic Chemicals. Environ. Sci. Technol. 11, 475 (1977).
Cloos, P., Moreale, A., Broers, C. and Badot, C., 1979.
Absorption and Oxidation of Aniline and p-Chloroaniline by
Montmorillpnite. Clay Miner. 14, 307 (1979).
CMA, 1983a. Response of the Chemical Manufacturers Association
Methylenedianiline Panel to EPA1s Section 4(f) Designation for
MDA. June 27, 1983.
CMA, 1983b. Response of the CMA Methylenedianiline Panel to the
EPA and OSHA ANPRs for MDA. Nov. 21, 1983.
CMA, 1983c. Confidential Business Information Version of
Response of the CMA MDA Panel to EPA1s Section 4(f) Designation
for MDA. June 27, 1983.
CMA, 1983d. Draft Survey of Non-MDI Uses of 4,4'-MDA.
CMA, 1984. A Survey of Non-MDI Uses of 4,4'-
Methylenedianiline. A Report Prepared by Risk Science
International, 1101 30th Street, NW., Washington, DC 20007.
April 24, 1984.
Cook, B. and Grindstaff, G., 1983. Memorandum: Carcinogenic
Risk From Exposure to 4,4'-Methylenedianiline. Sept. 19, 1983.
-------
169
Cox, J.V., 1982. Letter to Steven D. Newburg-Rinn, Chief, Test
Rules Development Branch, Office of Toxic Substances, U.S. EPA,
Re: 4,4'-Methylenedianiline. November 4, 1982.
Crump, K.S./ 1980. An Improved Procedure for Low-Dose
Carcinogenic Risk Assessment from Animal Data. J. Environ. Path.
and Toxicol. _5 675 (1980).
Darby, T.D., Johnson, H.J., Northup, S.J., 1978. An Evaluation
of a Polyurethane for Use as a Medical Grade Plastic. Toxicol.
Appl. Pharmacol. _4£, 449 (1978).
David, D.J., 1969. Analytical Chemistry of Polyurethanes.
Wiley-Interscience, New York City, NY.
Deichmann, W.B., MacDonald, W.E., Coplan, M., Woods, F., Blum,
E., 1978. Di-(4-aminophenyl)-methane (HDA): 4-7 Year Dog
Feeding Study. Toxicology_L1, 185 (1978).
DOE, 1983a. Department of Energy. Response of the Albuquerque
Operations Office (AL) and Contractors to the Environmental
Protection Agency (EPA) Request for Information for the Review of
4,4'-Methylenedianiline (MDA). Draft. Received by EPA July 1,
1983.
DOE, 1983b. Department of Energy. Response of the Albuquerque
Operations Office (AL) and Contractors to the Environmental
Protection Agency (EPA) Request for Information for the Review of
4,4'-Methylenedianiline (MDA). Draft. Received by EPA. Aug.
31, 1983.
DOE, 1983c. Information Letter to W. Hirzy from P.M. Ramey,
Department of Energy. Nov. 8, 1983.
DOE, 1983d. DOE Order 5480.1A. Environmental Protection, Safety
and Health Protection Program for DOE Operations. Aug. 13, 1981.
Dunn, G.W. and Guirguis, S.S., 1979. Methylenedianiline (MDA) as
an.Occupational Health Problem: A Suggested Time-Weighted
Average Exposure Level and Medical Program. Ark. Hyg. Rada.
Toksikol. _3_0_' 639 (1979).
Emmett, E.A., 1976. Allergic Contact Dermatitis in Polyurethane
Plastic Molders. J. Occup. Med. 18, 802 (1976).
EPA, 1983. U.S. Environmental Protection Agency. Printout
(Gene-Tox): Carcinogenesis Report IV Retrieved June, 1983.
Washington, D.C.: Office of Pesticides and Toxic Substances.
USEPA.
-------
170
EPA, 1984. Revised Interim Guidelines for the Health Assessment
of Suspect Carcinogens. August, 1984.
ESE, 1981. Environmental Assessment For 4,4'-Methylenedianiline.
Contract No. 68-01-6153. Environmental Science and Engineering,
Inc., P.O. Box ESE, Gainsville, FL, 32602. June 26, 1981.
Fukushima, S., Hibino, T., Shibata, M., Murasaki, G. , Ogiso, T.
and Ito, N., 1977. Effect of Hepatotoxic or Nephrotoxic Agents
on the Induction of Colong Cancers in Rats by 1,2-
Dimethlhydrazine. Toxicol. Appl. Pharmacol. 40, 561 (1977).
Fukushima, S., Shibata, M., Hibino, T., Toshimura, T., Hirose, M.
and Ito, N., 1979. Intrahepatic Bile Duct Proliferation Induced
by 4,4'-Diaminodiphenylmethane in Rats. Toxicol. Appl.
Pharmacol. _48_, 145 (1979).
Glowinski, I.B., Radtke, H.E. and Weber, W.W., 1978. Genetic
Variation in N-Acetylation of Carcinogenic Arylamines by Human
and Rat Liver. Mol. Pharmacol. 14, 940 (1978).
Godeck, E.G., Naismith, R.W., and Matthews, R.J., 1982. Response
of Uniroyal, Inc. To TSCA 8(d) Rule on 4,4'-MDA.
Gohlke, R., 1978. 4,4'-Diaminodiphenylmethane in a Chronic
Experiment. Z. Gesamte Hyg. Ihre Grenzgeb. 24, 159 (1978).
Grindstaff, G., 1984. Memorandum: 4,4'-MDA Risk Calculation
Considerations. Office of Toxic Substances, U.S. Environmental
Protection Agency. November 5, 1984.
Griswold, Jr., D.P., Casey, A.E., Weisburger, E.K. and
Weisburger, J.H., 1968. The Carcinogenicity of Multiple
Intragastic Doses of Aromatic and Heterocyclic Nitro or Amino
Derivatives in Young Female Sprague-Dawley Rats. Cancer Res. 28,
924 (1968).
Hedley, W.H., Mehta, S.M. Moscowitz, C.M., et al., 1975.
Potential Pollutants from Petrochemical Processes. Westport,
CT: Technomic Publication.
Hiasa, T., Itahori, T., Ohshima, M., Fujita, T., Yuasa, T.,
Kinishi, N. and Miyashiro, A., 1982a. Promoting Effect of
Phenobarbital and Barbital on Development of Thyroid Tumors in
Rats Treated with N-Bis(2-Hyroxypropyl) Nitrosamine.
Carcinogenesis 3, 1187 (1982).
Hiasa, Y.> Ohshima, M., Kitahori, Y., Yuasa, T., Fujita, T., and
Iwata, C., 1982b. Promoting Effects of 3-Amino-l,2,4-triazole on
the Development of Thyroid Tumors in Rats Treated with N-Bis(2-
Hydroxypropyl) Nitrosamine. Carcinogenesis_3, 381 (1982).
-------
171
Hiasa, Y., Kitahori, Y., Enoki, N., Konishi, N. and Shimoyama,
T. , 1984. 4,4'-Diaminodiphenylmethane: Promoting Effect on the
Development of Thyrid Tumors in Rats Treated with N-Bis(2-
hydroxypropyl) Nitrosamine. J.N.C.I. _7£, 471 (1984).
Hidetosh, M.f et al. (inventors). Hodogaya Chemical Co.
(assignee), 1968. Process for Preparing Isocyanates. U.S.
Patent 3,381,025. April 30, 1968.
Ho, T. , Hardigree, A.A., Larimer, F.W., Nix, C.E., Rao, T.K.,
Tipton, S.C., Epler, J.L., 1979. Comparative Mutagenicity Study
of Potentially Carcinogenic Industrial Compounds. Environmental
Mutagenesis _1, 167 (1979).
ICF, 1983. MDA Use and Substitutes Analysis in Non-MDI
Applications.
Ill, 1979. Bulletin 1. Study on Microbial Degradation of
Polyurethane Flexible Polyether Foams Under Waste Disposal
Conditions. International Isocyanate Institute, Inc., P.O. Box
1268, 71 Elm Street, New Canaan, CT, 06840. October, 1979.
Joschek, H.I. and Miller, S.I., 1966. Photo-oxidation of Phenol,
Cresols and Dihydroxybenzenes. J. Am. Chem. Soc. 88, 3273
(1966).
JRB, 1980. Support Document, Chapter IV: Health Effects of
4,4'-Methylenedianiline. Contract No. 68-01-4839. JRB
Associates, 8400 West Park Drive, McLean, VA, 22102. December
19, 1980.
JRB, 1981. Human Exposure Assessment: 4,4'-Methylenedianiline.
Contract No. 68-01-4839. JRB Associates, 8400 West Park Drive,
McLean, VA, 22102. October 16, 1981.
Karickhoff, S.W., Brown, D.S., and Scott, T.A., 1979. Sorption
erf Hydrophobia Pollutants on Natural Sediments. Water Res. 13,
241 (1979).
Kenaga, E. and Goring, C., 1980. Relationship Between Water
Solubility, Soil Sorption, Octanol-Water Partitioning, and
Concentration of Chemicals in Biota. Special Technical
Publication No. 707, American Society for Testing Materials, pp.
78-115.
Kirk-Othmer, 1980. Encyclopedia of Chemical Technology, 3rd
Edition, Volume 9, John Wiley and Sons, New York, 1980. p. 272.
Knutson, T., 1983. Memorandum: Aggregated 8(a) Level A Data for
4,4'-MDA. May 10, 1983.
Knutson, T., 1984. Memorandum: Aggregated 8(a) Level A Data for
4,4'-MDA. August 3, 1984. "•
-------
172
Kommineni, C.f Groth, D.H. and Frockt, I.J. , 1979. Determination
of Tumorigenic Potential of Methylene-bis-orthochloraniline. J.
Env. Path. Tox. _2, 149 (1979).
Kopelman, H., Robertson, M.H., Sanders, P.G., et al. , 1966a. The
Epping Jaundice. Br. Med. J. _1, 514 (1966).
Kopelman, J, Scheuer, P.J., and Williams, R., 1966b. The Liver
Lesion of the Epping Jaundice. Q.J. Med. 35, 553 (1966).
Laity, J.L., Burstain, I.G. and Appel, B.R., 1973. Photochemical
Smog and the Atmospheric Reactions of Solvents. In: Tess, R.W.,
et al. Solvents Theory and Practice. Advances in Chemistry
Series 124. Washington, DC: American Chemical Society, pp. 95-
112.
Landrum, P.P. and Crosby, D.G., 1981. Comparison of the
Disposition of Several of Nitrogen-Containing Compounds in the
Sea Urchin and Other Marine Invertebrates. Xenobiotica, 11, 351
(1981) (Abstract only).
Lavoie, E., Tully, L., Fow, E., Hoffmann, D., 1979. Mutagenicity
of Aminophenyl and Nitrophenyl Ethers, Sulfides, and
Disulfides. Mutation Res. JT7, 123 (1979).
Leo, A. and Hansch, C., 1979. Substituent Constants for
Correlation Analysis in Chemistry and Biology. New York: John
Wiley and Sons, Inc.
Leong, B.K.J., Land, J.E., Groehn, J.A., Coombs, -J.K., Sabatis,
C.P. and Griffin, R.L., 1984. Retinopathy from Inhaling 4,4'-
Methylenedianiline Aerosols (MDA) in Guinea Pigs (abstract,
received by EPA March 30, 1984, from the Upjohn Company).
Linch, A.L., O'Connor, G.B., Barnes, J.R., Killian, Jr., A.S. and
Neeld, Jr., W.E., 1971. Methylene-bis-ortho-chloroaniline
(MOCA): Evaluation of Hazards and Exposure Control. Amer.
Industr. Hyg. Ass. J. 32, 802 (1971).
Lipski, K., 1982. Liquid Chromatographic Determination of
Dimethyl Formamide, Methylene Bisphenyl Isocyanate and Methylene
Bisphenyl Amine in Air Samples. Ann. Occup. Hyg. 25, 1 (1982).
Lu, P.Y., Metcalf, R.L., Plummer, N. and Mandel, D. 1977. The
Environmental Fate of Three Carcinogens: Benzo[a] pyrene,
Benzidine, and Vinyl chloride Evaluated in Laboratory Model
Ecosystems. Arch. Environ. Contain. Toxicol. _6_, 129 (1977).
Mannsville Chemical Products. Methylene Diphenyl Isocyanates.
In: Chemical Products Synopsis. Cortland, New York. November
1980.
Mantel, N. and Schneiderman, M.A. , 1975^. Estimating "Safe"
Levels, A Hazardous Undertaking. Cancer Research 35, 1379
(1975).
-------
173
Mathtech, 1982. Draft Report: Level I Economics Evaluation:
4,4'-Methylenedianiline. Contract No. 68-01-5864. MATHTECH,
Inc., 1611 North Kent Street, Arlington, VA 22209. August 1982.
McGill, D.B. and Motto, J.D., 1974. An Industrial Outbreak of
Toxic Hepatitis Due to Methylenedianiline. N. Engl. J. Med. 291,
278 (1974).
Meltzer, Y.L., 1971. Urethane Foams - Technology and
Applications. Park Ridge, NJ: Noyes Data Corporation.
Metcalf, R.L. and Sanborn, J.R., 1975. Pesticides and
Environmental Quality in Illinois. 111. Natl. Hist. Survey Bull.
31, 381 (1975).
Miller, A.L. and Sherman, M., 1965. Test Report on 4,4'-
Methylene-bis-(2-chloroaniline). E.I. DuPont de Nemours and
Company (Inc.) Haskell Laboratory Report No. 167-65, MR No.
815. Wilmington, DE.
Milman, H., 1983. Memorandum: 4,4' -Methylenedianiline.' Office
of Toxic Substances, U.S. Environmental Protection Agency. May
4, 1983.
Milman, H., 1984. Memorandum: 4,4'-MDA. Office of Toxic
Substances, U.S. Environmental Protection Agency. August 30,
1984.
Moore, W.M. 1978. Methylenedianiline. In: Kirk-Othmer
Encyclopedia of Chemical Technology. 3rd edition. New York:
Interscience Publishers. _2, 338.
Moreale, A. and Van Bladel, R., 1976. Influence of Soil
Properties and Absorption of Pesticide Derived Aniline and p-
Chloroaniline. J. Soil. Sci. _2J7, 48 (1976).
a
Morrison, R.T. and Boyd, R.N., 1973. Organic Chemistry. 3rd
ed. Boston: Allyn and Bacon, Inc.
Morton, K.C., 1981. Testimony Submitted to the Occupational
Health Standards Commission of the Michigan Department of Public
Health Concerning 4,4'-Methylene-bis (2-chloroaniline). March
31, 1981.
Munn, A., 1967. Occupational Bladder Tumors and Carcinogens:
Recent Developments in Britain. In: Deichmann, W. and Lampe,
K.F., ed. Bladder Cancer. A Symposium, Birmingham AL:
Aesculapius, pp. 187-193.
NCI, 1978a. Bioassay of 4,4'-Methylene bis (N,N-dimethyl)-
benzeneamine for Possible Carcinogenicity. National Cancer
Institute, Technical Report 186. Department of Health, Education
and Welfare, Bethesda, MD. 1978.
-------
174
NCI, 1978b. Bioassay of Michler's Ketone for Possible
Carcinogenicity. National Cancer Institute Technical Report
181. Department of Health, Welfare and Education, Bethesda,
MD. 1978.
NCI, 1978c. Bioassay of 4,4'-Thiodianiline for Possible
Carcinogenicity, National Cancer Institute Technical Report,
Series No. 47. NCI-CG-TR-47. DHEW Publication No. (NIH) 78-847.
NCI, 1980. Bioassay of 4,4'-Oxydianiline for Possible
Carcinogenicity. National Cancer Institute Technical Report,
Series No. 205. NCI-CG-TR-205. NIH Publication No. 80-1761.
August 1980.
NIOSH, 1976a. Current Intelligence Bulletin 8. 4,4'-
Diaminodiphenylmethane (DDM). January 30, 1976.
NIOSH, 1976b. Background Information on 4,4'-
Diaminodiphenylmethane (DDM). Am. Ind. Hyg. Assoc. J. 37, 26
(1976).
NIOSH, 1980. Health Hazard Evaluation Report HETA 79-141.
Fischer and Porter Co., Warminister, PA.
NIOSH, 1981. Health Hazard Evaluation Report. HETA 81-092-
950. International Playtex Corporation, Newman, GA. September
1981.
NIOSH, 1983. Health Hazard Evaluation Report. HETA 82-146-
1388. Boeing Vertol Co., Philadelphia, PA. November 1983.
NIOSH, 1984a. Industrial Hygiene Survey Report of Olin
Corporation, Moundsville, WV. Dr. Mark Boeniger, Industrial
Hygiene Section, National Institute for Occupational Safety and
Health, Cincinnati, OH. March 1984.
NIOSH, 1984b. Industrial Hygiene Survey Report of A.O.
Smith/Inland Corporation, Little Rock, AR. Dr. Mark Boeniger,
Industrial Hygiene Section, National Institute for Occupational
Safety and Health, Cincinnati, OH. February 1984.
NOHS, 1983. Aniline, 4,4'-Methylenedi- National Occupational
Hazard Survey. Observed Exposures Weighted to National Estimates
of Exposure for 1980. U.S. Department of Health and Human
Services. National Institute for Occupational Safety and
Health. June 1, 1983.
NTP, 1983a. Carcinogenesis Studies of 4,4'-Methylenedianiline
Dihydrochloride in F344/N Rats and B6C3F1 Mice. National
Toxicology Program Technical Report, Series No. 248. NTP-81-
143. NIH Publication No. 83-2504. Juae 1983.
-------
175
NTP, 1983b. National Toxicology Program, Technical Bulletin No.
10 June 1983, pp. 9-10.
Nunizata, A., 1983. Report on Chromosomal Aberrations Assay With
Human Lymphocytes With and Without Metabolic Activation Performed
on 4-4'-Diami'nophenylmethane (DDM). Report from CRF Centro
Ricerca Farmaceutica spa. 15pp. Montedison, Milano, Italy.
Pantarotto, C., 1983. In vivo Studies of Covalent Binding of
4,4' -Diaminophenylmethane to Proteins and DNA of Mouse Liver.
Istituto di Ricerche Farmacologiche "Mario Negri", Milano, Italy.
Parodi, S., Taningher, M., Russo, P., Pala, M., Tamaro, M.,
Monti-Bradigin, C., 1981. DNA-Damaging Activity in vivo and
Bacterial Mutagenicity of Sixteen Aromatic Amines and Azo-
Derivatives, as Related Quantitatively to Their
Carcinogenicity. Carcinogenesis _2 , 1317 (1981)
Parodi, S., Zunino, A., Ottaggio, L., DeFerrari, M., Santi, L. ,
1983. Lack of Correlation Between the Capability of Inducing
Sister-Chromatid Exchanges in vivo and Carcinogenic Potency for
16 Aromatic Amines and Azo-Derivatives. Mutation Res. 108, 225
(1983).
PEDCo, 1983. PEDCo Environmental, Inc. 11499 Chester Road,
Cincinnati, OH., 45246-0100. Exposure to and Control of 4,4'-
Methylenedianiline. Contract No. 68-02-3935. Work Assignment 1-
5. October 1983.
Perkins, G.T. (inventor), E.I. DuPont de Nemours and Co.
(assignee), 1968. Process for the Preparation of 4,4'-
Methylenedianiline. U.S. Patent 3,367,969. February 6, 1968.
Pistor, H.J., Hoffmann, H., Joschek, H., et al. (inventors),
Badiache Anillin and Soda-Fabrik (assignee), 1977. Manufacture
of 4,4'-Diphenylmethane Diisocyanate and a Mixture of
Diisocyanates and Polyisocyanates. U.S. Patent 4,014,914. March
29, 1977.
Powers, E.L. (inventor), Mobay Chemical Co. (assignee), 1979.
Process for the Preparation of Substantially Pure 4,4'-
Diaminodiphenylmethane. U.S. Patent 3,517,062. June 23, 1970.
Ramney, M.W., 1972. Isocyanates Manufacture. Park Ridge, NJ:
Noyes Data Corporation.
Rao, T.K., Dorsey, G.F., Allen, B.E., Epler, J.L., 1982.
Mutagenicity of 4,4'-Methylenedianiline Derivatives in the
Salmonella Histidine Reversion Assay. Arch. Toxicol. 49, 135
(1982).
Reinke, R.E., 1963. Test Report on 4,4'-Methylene-bis-(2-
chloroaniline). E.I. Du Pont Nemours and Company (Inc.) Haskell
Laboratory Report No. 97-63, MR No. 346. Wilmington, DE.
-------
176
Rubino, G.F., Scansetti, G., Piolatto, G.f and Pira, E., 1982.
The Carcinogenic Effect of Aromatic Amines; An Epidemiological
Study on the Role of o-Toludine and 4,4'-Methylene bis(2-
methylaniline) in Inducing Bladder Cancer in Man. Environ. Res.
27, 241 (1982).
Russfield, A.G., Homburger, F. , Boger, E., VanDongen, C.G.,
Weisburger, E.X. and Weisburger, J.H. , 1975. The Carcinogenic
Effect of 4,4'-Methylene-bis-(2-chloroaniline) in Mice and
Rats. Tox. Appl. Pharm. _3J., 47 (1975).
Schilling, B. Von Canstatt, Hofmann, H., Oettel, H., et al. ,
1966. Changes in the Retina of Cats when Poisoned with Chemicals
Applied Perorally and Percutaneously. Trans. Germ. Soc. Path.,
50th Conference, April, 1966.
Schoental, R., 1968a. Carcinogenic and Chronic Effects of 4,4'-
Diaminodiphenylmethane, an Epoxy Hardner. Nature 219, 1162
(1968).
Schoental, R., 1968b. Pathological Lesions, Including Tumors, in
Rats After 4,4' -Diaminodiphenylmethane and gamma-Butyrolacone.
Isr. J. Med. Sci. _4, 1146 (1968).
Scott-Siegel, C. , 1984. Memorandum: NIOSH Health Flazard
Evaluation of Methylenedianiline Exposed Workers. Office of
Toxic Substances, U.S. Environmental Protection Agency. August
2, 1984.
Smith, W.E. and LaSalle, J.R., 1974. Characterization of
Atmospheric Emissions from Polyurethane Resin Manufacture. U.S.
Environmental Protection Agency. EPA-650/2-74-107.
Snyder, W.S., Cook, M.J., Karhausen, L.R., Nasset, E.S., Howells/
G. P. and Tipton, I. H., 1974. Report of the Task Group on
Reference Man.- International Commission on Radiological
Protection. Pergamon Press.
Springborn, 1982. Use Analysis of 4,4'-Methylenedianiline (Non-
Isocyanate Use). Springborn Regulatory Services, Inc. Enfield,
CT., 06082. September 1982.
Springborn, 1983. Non-Isocyanate Uses of 4,4' -Methylenedianiline.
Springborn Regulatory Services, Inc., Enfield, CT, 06-82.
February 1983.
Steinhoff, D. and Grundmann, E., 1970. Carcinogenic Action of
4,4'-Diaminodiphenylmethane and 2,3'-Diaminodiphenylmethane.
Naturwissenschaften 57, 247 (1970).
-------
177
Stula, E.F., Sherman, H., Zapp, Jr., J.A., Clayton, Jr., J.W.,
1975. Experimental Neoplasia in Rats from Oral Administration of
3,3'-Dichlorobenzidine, 4,4'-Methylene-bis-(2-chloroaniline) and
4,4' -Methylene-bis-(2-methylaniline). Tox. Appl. Pharm. 31, 159
(1975).
Stula, E.F., Barnes, J.R., Sherman, H., Reinhardt, C.F. and Zapp,
Jr., J.A., 1977. Urinary Bladder Tumors in Dogs from 4,4'-
Methylene-bis-(2-chloroaniline) (MOCA). J. Env. Path. Tox. _1, 31
(1977).
Subba-Rao, R.V. and Alexander, M., 1977. Products Formed From
Analogues of l,l,l-Trichloro-2,2-bis (p-chlorophenyl) Ethane
(DDT) Metabolites by Pseudomonas putida. Appl. Environ.
Microbiol. 33, 101 (1977).
Suh, K.W., 1980. Foamed Plastics. In: Encyclopedia of Chemical
Technology. Volume 1. 3rd edition. New York: John Wiley and
Sons. pp 82-126.
Thies, G., 1983. Memorandum: Evaluation of Potential Skin and
Lung Penetration of 4,4'-Methylenedianiline. May 24, 1983.
Thies, G., 1984. Personal communication.
Tobes, M.C., Brown, L.E., Chin, B. and Marsh, D.D., 1983.
Kinetics of Tissue Distribution and Elimination of 4,4'-Methylene
bis (2-chloroaniline) in Rats. Toxicol. Lett. 17, 69 (1983).
Tortoreto, M, Urso, R., Bianchi, M., Catalani, P., Blonda, C.,
Paglialunga, S., Pantarotto, C., 1961. Distribution of Pure and
Technical 4,4'-Diaminodiphenylmethane in Blood and Tissues of
B6C3^1 Ma-'-e Mice. Istituto Di Ricerche Farmacologiche Mario
Negri.
Vaudaine, M., Lery, N., Diter, J.N., Droin, M. and Charnaillard,
C. , 1982. Diamino-Diphenyl Methane: An Example of Toxic Agent
Monitoring in the Workplace at Rhone-Poulenc Industrie. J.
Toxicol. Medicale _2, 207 (1982).
Versar, 1983a. Memorandum: Summary of MDA Quick Response
Effort. Doug Dixon to Karen Hammerstrom. March 9, 1983.
Versar. 1983b. Versar Inc., 6850 Versar Center, Springfield, VA,
22151. Exposure Assessment for Methylenedianiline (MDA). Draft
Final Report, Contract No. 68-01-6271. Task 65. September 12,
1983.
Williams, S.V., Bryan, J.A., Burk, J.R. and Wolf, E.S., 1974.
Letter: Toxic Hepatitis and Methylenedianiline. N. Engl. J.
Med. 291, (1974) .
Windholz, M. (Ed.), 1976. Merck Index." Ninth Edition, Merck and
Company, Inc., Rahway, N.J.
-------
178
Woolrich, P.P. , 1982. Toxicology, Industiral Hygiene and Medical
Control of TDI, MDI, and PMPPI. Am. Ind. Hyg. Assoc. J. 43,
(1982).
Young, D.A., and Parker, E.G. (Bendix Corp.), 1978. Removal of
Methylenedianiline from Chemical Plant Wastewater. Washington,
DC: U.S. Department of Energy. (CONF-780455-1). (BDX-613-
1981).
Youer, A.M. (inventor), I.E. DuPont de Nemours and Co.
(assignee), 1969. Process for the Preparation of Polyisocyanate
Reinforced Elastomers. U.S. Patent 3,450,649.
Zabik, M. and Kawaguchi, S., 1982. The Study of the Oxidation,
Sublimation, and Photochemistry of Curene 442 [4,4'-Methylene bis
(2-chlcroaniline)] on Surfaces, as Thin Films, and in Solution.
Preliminary Draft Report. University of Michigan, Pesticide
Research Center.
Zylberszac, S., 1951. Cirrhogenous Action of Insoluable Diamino-
diphenyls on the Rat Liver. .C.R. Soc. Biol. 145, 136 (1951).
------- |