NNAS
RNAE
C.IOM
Plans for Clinical and Epidemiologic
Follow-up After Area-Wide
Chemical Contamination
Proceedings of an International Workshop
Committee on Response Strategies to Unusual Chemical Hazards
Board on Toxicology and Environmental Health Hazards
Assembly of Life Sciences
-------
Plans for Clinical and Epidemiologic
Follow-up After Area-Wide
Chemical Contamination
Proceedings of an International Workshop
Washington, D.C.
March 17-19,1980
Luigi Dardanoni, Co-Chairman
Robert W. Miller, Co-Chairman
Committee on Response Strategies to Unusual Chemical Hazards
Board on Toxicology and Environmental Health Hazards
Assembly of Life Sciences
National Research Council
NATIONAL ACADEMY PRESS
Washington, D.C. 1982
-------
NOTICE: The project that is the subject of this report was approved by
the Governing Board of the National Research Council, whose members are
drawn from the councils of the National Academy of Sciences, the
National Academy of Engineering, and the Institute of Medicine. The
members of the committee responsible for the report were chosen for
their special competences and with regard for appropriate balance.
This report has been reviewed by a group other than the authors
according to procedures approved by a Report Review Committee
consisting of members of the National Academy of Sciences, the National
Academy of Engineering, and the Institute of Medicine.
The National Research Council was established by the National
Academy of Sciences in 1916 to associate the broad community of science
and technology with the Academy's purposes of furthering knowledge and
of advising the federal government. The Council operates in accordance
with general policies determined by the Academy under the authority of
its Congressional charter of 1863, which establishes the Academy as a
private, nonprofit, self-governing membership corporation. The Council
has become the principal operating agency of both the National Academy
of Sciences and the National Academy of Engineering in the conduct of
their services to the government, the public, and the scientific and
engineering communities. It is administered jointly by both Academies
and the Institute of Medicine. The National Academy of Engineering and
the Institute of Medicine were established in 1964 and 1970,
respectively, under the charter of the National Academy of Sciences.
The work on which this publication is based was performed pursuant
to Contract 68-02-3211 with the Environmental Protection Agency and
Centers for Disease Control.
ii
-------
COMMITTEE ON RESPONSE STRATEGIES TO UNUSUAL CHEMICAL HAZARDS
Robert Miller, Chairman
Clinical Epidemiologic Branch
National Cancer Institute
Bethesda, Maryland
A.L. Burlingame
Mass Spectrometry Research Resource
University of California
Berkeley, California
Aaron B. Lerner
Department of Dermatology
Yale University
New Haven, Connecticut
John A. Moore
Research Resources Program
National Institute of Environmental
Health Sciences
Research Triangle Park, North Carolina
Sheldon D. Murphy
Department of Pharmacology
University of Texas
Houston, Texas
Robert A. Neal
Department of Biochemistry
Vanderbilt University
Nashville, Tennessee
Milos Novotny
Department of Chemistry
Indiana University
Bloomington, Indiana
Patrick O'Keefe
Division of Laboratories and Research
New York Sate Department of Health
Albany, New York
Alan Poland
Department of Oncology
McArdle Lab for Cancer Research
University of Wisconsin,
Madison, Wisconsin
Staff
Robert G. Tardiff Frances Peter Jacqueline Prince
Project Director Editor Staff Assistant
iii
-------
BOARD ON TOXICOLOGY AND ENVIRONMENTAL HEALTH HAZARDS
Ronald W. Estabrook, Chairman
Department of Biochemistry
University of Texas Medical School
Dallas, Texas
Philip Landrigan, Vice-Chairman
National Institute for
Safety and Health
Cincinnati, Ohio
Edward Bresnick
Department of Biochemistry
University of Vermont
Burlington, Vermont
Theodore Cairns
Greenville, Delaware
Victor Cohn
Department of Pharmacology
George Washington University
Medical Center
Washington, D.C.
A. Myrick Freeman
Department of Economics
Bowdoin College
Brunswick, Maine
Ronald W. Hart
National Center for Toxicological
Research
Jefferson, Arkansas
Michael Lieberman
Department of Pathology
Washington University
St. Louis, Missouri
Richard Merrill
School of Law
University of Virginia
Charlottesville, Virginia
Robert A. Neal
Chemial Industry Institute
of Toxicology
Research Triangle Park, North Carolina
Ian Nisbet
Clement Associates
Washington, D.C.
John Peters
Department of Family and
Preventive Medicine
University of Southern Cal
Los Angeles, California
Liane Russell
Biology Division
Oak Ridge National Laboratory
Oak Ridge, Tennessee
Charles R. Schuster,Jr.
Department of Psychiatry
University of Chicago
Chicago, Illinois
Ex Officio Members
James F. Crow
Genetics Department
University of Wisconsin
Madison, Wisconsin
Roger McClellan
Lovelace Biomedical and
Environmental Research Inst
Albuquerque, New Mexico
Daniel Menzel
Department of Pharmacology
Duke University
College Park, Maryland
Robert Menzer
Department of Entomology
University of Maryland
College Park, Maryland
Robert Miller
National Cancer Institute
Bethesda, Maryland
iv
-------
Sheldon Murphy
Department of Pharmacology
University of Texas
Houston, Texas
Norton Nelson
Institute of Environmental Medicine
New York University Medical Center
New York, New York
James Whittenberger
School of Public Health
Harvard University
Boston, Massachusetts
Staff
Robert G. Tardiff
Executive Director
Board on Toxicology and Environmental
Health Hazards
National Research Council
Gordon Newell
Associate Executive Director
Board on Toxicology and Environmental
Health Hazards
National Research Council
Jacqueline Prince
Staff Assistant
Board on Toxicology and Environmental
Heath Hazards
National Research Council
-------
TABLE OF CONTENTS
Page
PREFACE xi
INTRODUCTION
Francesco Pocchiari 1
CASE STUDIES OF SELECTED AREA-WIDE ENVIRONMENTAL
EXPOSURES
TCDD (Italy)
Luigi Dardanoni and Gaetano Fara 3
Yusho (Japan)
Robert W. Miller 34
Treatment of Chlordecone (Kepone) Poisoning
with Cholestyramine
Philip S . Guzelian 40
DBCP
Donald Whorton 60
Increased Lead Absorption with Anemia and and
Slowed Nerve Conduction in Children Near a Lead
Smelter
Philip J. Landrigan 75
Methyl Mercury (Japan)
Robert W. Miller 86
Chlorinated Hydrocarbons
David Axelrod 90
Cohort Study of Michigan Residents Exposed to
Polybrominated Biphenyls: Epidemiologic and
Immunologic Findings
Philip J. Landrigan 100
Atomic Bomb Casualty Commission
Gilbert W. Beebe 114
vii
-------
ADVERSE EFFECTS ON TARGET SITES AND STUDIES OF TOXICITY
Reproductive Injury: Love Canal
David Axelrod 126
Reproductive Injury: General Considerations
Robert W. Miller 145
Reproductive Injury: General Considerations
Helga Rehder 154
Birth Defects Register in Seveso: A TTCDD-Polluted Area
E. Marni, et. al. 174
Carcinogenic Effects of Chemical and Physical
Agents: Human Observations
Clark W. Heath, Jr 195
Experimental Studies on Carcinogenic Effects of TCDD
Giuseppe Delia Porta, Maria I. Culnaghi,
Tommaso A. Dragani 206
Adverse Neurologic Effects
Alan M. Goldberg 217
Exposure to TCDD: Immunologic Effects
Girolamo G. Sirchia, et. a_l 234
Somatic Cell Mutations
Arthur D. Bloom 267
Cytogenetic Investigations of the Seveso Population
Exposed to TCDD
L. De Carli, ejtjil 292
Cytochrome P-450 Induction by 2,3,7,8-Tetra-
chlorodibenzo-p_-dioxin, Polychlorinated Biphenyls,
and Polybrominated Biphenyls
Robert Neal 320
Hepatic Toxicity of TCDD
Silvio Garattini 335
viii
-------
In Vivo DNA Damaging Activity, In Vivo Covalent DNA
Binding and Bacterial Mutagenicity as Related Quan-
itatively to Carcinogenic Potency
S. Parodi, M. Taningher, L. Santi 364
Liver Injury
Sheldon D. Murphy 366
PANEL ON EPIDEMIOLOGIC APPROACHES TO MEASUREMENT AND
ASSESSMENT OF EXPOSURES 385
SUMMARY AND CONCLUSIONS 406
APPENDIX 1: Epidemiologic Monitoring in an Episode of
Environmental Chemical Pollution: Problems and Programs
in the Seveso Experience
Leonardo Santi 409
APPENDIX 2: Program 413
ix
-------
PREFACE
In 1977, the National Academy of Sciences-National Research Council
(NAS-NRC) was invited by the Italian government to join in a
collaborative effort to investigate the effects of area-wide chemical
contamination at Seveso, Italy. The contamination was the result of an
explosion of a reaction vessel containing highly toxic 2,3,7,8-tetra-
chloro-dibenzo-p-dioxin (TCDD and commonly known as "dioxin"), which
produced a cloud of chemical that was carried southward by the wind,
exposing humans, animals, and plant life for several kilometers. The
NAS-NRC sent a team of American scientists to visit Italy and to
determine with Italian officials the needs and opportunities for
cooperative study. The NAS-NRC team recommended the development of a
continuing relationship of U.S. and Italian scientists for the purposes
of exchanging scientific and technical information, fostering the
conduct of complementary research, organizing workshops and conferences
to examine the impacts on health and the environment, and assisting in
the coordination of exchange of scientists engaged in the analysis of
this accident.
The NAS-NRC formally structured its involvement in this
collaborative venture by establishing the Committee for Binationad
Cooperative Study of Exposure to TCDD which was later renamed the
Committee on Response Strategies to Unusual Chemical Hazards. The
Committee's terms of reference were two-fold: first, the development
of guidelines that might be used to implement a worldwide mechanism for
guiding biomedical researchers at the scene of accidents similar to
xi
-------
that at Seveso (thereby ensuring the most comprehensive collection of
scientific information in a timely manner); and second, the
evaluation—in cooperation with Italian counterpart scientists—of
newer health data from the Seveso accident and the design of future
studies.
One of the Committee's first undertakings was a workshop, held in
March 1980, on Plans for Clinical and Epidemiological Follow-up after
Area-wide Chemical Contamination. These Proceedings are the product of
that workshop. The topic of the workshop was approached from two
points of view: first, by exploring a number of cases in which such
widespread contamination occurred and which served as the basis for
field studies; and second, by evaluating diseases and target organs
that were identified as likely outcomes of chemical exposures. A
synthesis of experiences and guiding principles for future investi-
gations of similar exposures was provided by a panel of experts from
the U.S. and Italy.
Between the time that the workshop was held and the proceedings
were completed, the Committee has been aware of several publications on
this topic and wishes to bring to the attention of the reader the
following: Guidelines for Studies of Human Populations Exposed to
Mutagenic and Reproductive Hazards (Proceedings of a conference held
January 26-27, 1981, edited by Arthur D. Bloom, published by March of
Dimes Birth Defects Foundation) and Chemical Radiation Hazards to
Children (Proceedings of the 84th Ross Conference on Pediatric
Research, Columbus, OH, in press). It is the Committee's hope that all
of these resources will provide useful guidance to the difficult
XII
-------
investigations of adverse effects to human health that may be
associated with chemical exposures of individuals of broad geographic
distribution.
xiii
-------
INTERNATIONAL WORKSHOP ON PLANS FOR CLINICAL AND EPIDEMIOLOGIC FOLLOW-UP
AFTER AREA-WIDE CHEMICAL CONTAMINATION
Introduction
Francesco Pocchiari
This International Workshop on "Plans for Clinical and Epidemiological
Follow-up after Area-wide Chemical Contamination" is a collaborative effort of
the Lombardy Region of Italy, the Istituto Superiore di Sanita in Rome, and
the U.S. National Academy of Sciences, to deal with the long-term medical
followup of the population exposed to dioxin at Seveso, and to explore the
implications of this event toward similar, areawide chemical incidents. This
meeting should help synthesize comprehensive epidemiologic approaches and
develop plans for investigating widespread exposures and their impacts on
human health for use in other present and future chemical accidents.
A number of problems have to be solved in order to define a medium-term
strategy to both protect the population and to decontaminate the environment.
In many respects, the Seveso incident is representative of the difficulties
inherent in dealing with any accidental chemical contamination.
A major goal, in any such event, is the clinical and epidemiologic followup
of the exposed population. But at Seveso, this undertaking was confounded by
several factors:
1. Difficulty in identifying the actual "hit" area. TCDD distribution in
soil was very uneven, with wide differences even between levels at nearby sites.
2. Constraints relative to the sampling and extraction techniques prevented
extensive and rapid monitoring of contamination.
3. Lack of fully adequate and rapid analytic methods to assay TCDD in some
materials, such as atmospheric particles, vegetation, and animal tissues resulted
in delayed and uncertain estimates of the area actually exposed.
-------
4. Overlapping responsibilities of national, regional, and municipal
authorities required coordination.
An impressive amount of work was required to overcome all of the difficulties
and, thus, to derive an exact estimate of the TCDD levels in the affected territory,
to assess the persistence of TCDD in soil, and to predict with reliability the
accident's effects on the health of the exposed population.
-------
TCDD (Italy)
by Luigi Dardanoni and Gaetano Fara
On July 10, 1976, a reactor exploded in Seveso, north of Milan,
sending a cloud of 2,3,7,8-tetrachlorodibenzo-p_-dioxin (TCDD) over a
large area north of Milan. Plants and animals in the vicinity of the
plant were affected, as were humans exposed to the cloud. Levels of
exposure were classified according to three geographic regions near
the plant, and attempts were made to assess the risk to the residents
of these areas, some of whom were evacuated. Techniques included
assessment of soil samples from the various areas for TCDD contamination
and, later, for TCDD persistence; a questionnaire survey of residents
to determine their exposure to the cloud and to contaminated animals;
pathologic studies of animals that spontaneously died in the area
following the accident; autopsy reports of syndromes exhibited by those
animals; and an examination of the general distribution of mortality among
various species of animals. These data, along with the incidence of chloracne
in the exposed population, were correlated with the territorial distribution
of the TCDD derived from the soil samples.
This report summarizes the data on the sources of contamination by
and the degree of population exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin
(TCDD) contamination in Seveso, Italy. It focuses on TCDD distribution
in different environmental areas within the Seveso region and on the degree
of human exposure, based on an evaluation of anamnestic data provided
by area residents, deaths of animals, and human morbidity. Most of the
data presented here have already been published elsewhere.
Industria Chimica Meda Societa Anonoma (ICMESA), a chemical and
pharmaceutical company, was established in Seveso in 1945. It began
producing 2,4,5-trichlorophenol (2,4,5-T) in 1969. On July 10, 1976,
a fast-spreading reaction raised the internal temperature of the reactor
to far above 200°C (perhaps to 400°C), and the pressure increased up
to the critical point of the valve. As a result, TCDD concentration
^Institute of Hygiene, University of Palmero Medical School (L. Dardanoni) and
Institute of Hygiene, University of Milan Medical School (G. Fara).
-------
in the final product was also raised beyond measure. When the protection
disk blew up, the fluid mixture burst into the open air, propelled by
the built-up pressure. The visible part of the cloud rose some 50
meters and subsequently fell back to the ground. The wind caused the
TCDD to spread over a wide area.
Leaves of plants, courtyard animals, and birds near the ICMESA
factory were severely affected. Many animals died within a few days of
the accident. At the same time, humans exposed to the toxic alkaline
cloud began to develop dermal lesions (Pocchiari et^ £!L. , 1979).
A map approximating the contaminated area (Figure 1) was drawn
based on airstream pattern at blow-out time, preliminary analytic
findings, and information on the sites of toxic and pathologic events,
which was obtained through work performed by university and county
laboratories coordinated by the Istituto Superiore di Sanita. As a
first step, on July 26, 1976, Italian authorities evacuated 179 people
from a 15-hectare area immediately southeast of the plant. A few days
later, further analysis of the TCDD content of soil and vegetation
samples prompted authorities to evacuate all the inhabitants (733
people) from a wider area (Zone A, approximately 110 hectares), extending
about 2 km southeast of the ICMESA plant. Moreover, inhabitants of the
surrounding area (Zones B and R) were subjected to a number of public
health regulations. Residents were prohibited from farming, consuming
local agricultural products, and keeping poultry and other animals.
-------
N
t
0 4OO
FIGURE 1. Schematic map of Seveso area by pollution zones.
Regione Lombardia, Ufficio Speciale, 1976.
From
-------
Zone B (270 hectares) was the natural extension of Zone A along
the main pathway of TCDD diffusion. Zone B exhibited a lower dioxin
content than did Zone A. Both zones were enclosed by a larger territory,
Zone R (1,430 hectares) exhibiting undetectable, or near detection-
threshold, contamination levels. At the Zone B-R borderline, the
average TCDD concentration in soil was found to be 5 yg/m2; Zone R
contained nondetectable TCDD levels (more exactly, less than 0.75 yg/m2.
The Zone A-B borderline ran along the line at which the mean TCDD
9
concentration was 50 yg/m . The actual borders were eventually
established where natural or artificial geographic divisions already
existed (DiDomenico e_t al^. , 1980a).
A more accurate map was drawn in September 1976, when the
investigators had the results of sampling along five straight reference
lines, fanning southward from the ICMESA plant (Figure 2). The figure
shows the findings from 108 sampling sites.
A limited area within Zone A, close to the ICMESA plant, exhibited
the highest TCDD content. The main diffusion pathway fell between
reference lines II and IV. This distribution suggests that most of the
chemical cloud was wind-driven over a narrow corridor and that, moving
away from the main diffusion pathway or from the ICMESA plant, a gradual
dilution process occurred within the cloud. The dispersion was due
both to three dimensional airborne motion and to gravitational effects.
-------
TCDD levels detected in Zone A ranged from analytic detection
threshold (0.75 uS/m2) to - 20 mg/m2. Thus, Zone A was divided into
eight subzones, characterized by more homogeneous TCDD levels (Figure
2).
Based on early analytic results and various calculations, the
total amount of TCDD in the soil of Zone A was estimated to amount to a
few hundred grams. By far, most of TCDD was deposited within an 0.6-km
area south of the plant (Bisanti et_ al^. , 1980).
By late September, a new sampling campaign was started, continuing
through December 1976. This resulted in the production of the Zone A
map dated January 1977 (Figure 3).
The data obtained from soil tests on different occasions reveal
that the TCDD fallout resulted in a highly irregular distribution
pattern, which has been only slightly modified since the accident.
Concentrations in any two nearby sites (less than 100 m apart) may
differ by as much as a factor of 100, thereby conferring negligible
significance to individual site data. Table 1 summarizes the analytic
results of soil samples from the various zones.
TCDD levels determined at 44 locations in Zone A during three
surveys carried out at different times (1, 5, and 17 months after the
ICMESA accident), although not specifically intended for such a purpose,
were used to derive a rough estimate of the environmental persistence of
TCDD in soil. The data provide statistically significant (p<0.01)
-------
I—I—I
TCDD/jg/m 2
. 0.75-5
• 5-10
• 10-50
• 50-250
250-TOOO
Zone A
September 1976 \
100 20 m
FIGURE 2. September 1976 map of Zone A. Original polar coordinates are
labeled with Roman numerals I-V. From DiDomenico et al. , 1980a,
-------
Zone A
January 1977
100 a m
=^m=z
FIGURE 3. January 1977 map of Zone A. From Di Domenico et al., 1980a.
-------
TABLE 1
Distribution of TCDD Contamination in the A,B, and R Areas on
the Basis of Gas Chromatographic-Mass Spectrometric Analysis of Soil Samples'1
Zone
A total
Al
A2
A3
A4
A5
A6
A7
B
R
Size
(hectares)
80.3
10.7
5.1
9.2
7.2
16.3
14.0
17.8
269.4
1,430.0C
TCDD values
Average
580.4
421.1
350.5
134.9
62.8
29.9
15.5
3.0
0.9
(Pg/m2)
Top
5447
1700
2015
902
427
270
91.7
43.8
9.7
No. of
Samples
306
51
19
34
26
50
61
65
106
449
Negative
No.
12
1
0
3
3
2
2
1
26
308
Samples
%
3.9
1.9
0
8.8
11.5
4.0
3.2
1.5
24.5
68.6
Estimated Amount
of TCDD in the
Area (g)
147.5
62.1
26.5
32.2
9.7
10.02
4.1
2.7
8.0
8.5
a Data provided by Regione Lombardia, Piano Operativo no. 1. Adapted from Bisanti et al., 1980.
b Less than 0.75 yg/m2
c only 950 hectares mapped.
-------
evidence that the geometric mean of TCDD levels dropped to about one-
half in the unworked soil of Zone A during the first 5 months after the
accident. Following this period, no further decreases in TCDD levels
were detected. One month after the accident, TCDD half-life was
approximately 1 year. Seventeen months after the accident, it was
estimated to be greater than 10 years (Di Domenico et al., 1980b).
Vertical distribution of TCDD dropped sharply in the top 8 cm of
soil, regardless of sampling site location or TCDD concentration. TCDD
levels varied much less in the 8-24 cm range. This trend is illustrated
in Figure 4, where the vertical distribution (expressed as a percentage
of total TCDD recovered) observed at site A,, (a spot just south of the
ICMESA plant in Zone A,) is shown as a typical example. TCDD concentrations
at depths greater than 8 cm were generally less than those detected in
the upper layer at the site by at least one order of magnitude.
TCDD vertical distribution was also investigated in the top 2-cm
soil layer. As Figure 5 shows, the highest TCDD levels were not found
in the topmost soil layer (0.5 cm), but very often in the second (0.5-
1.0 cm) or third (1.0-1.5 cm) strata (Di Domenico e£ al^. , 1980c).
This, and the findings of other investigators of the vertical
distribution of TCDD in soil at various intervals following the accident,
are consistent with the hypothesis that the TCDD abatement observed in
1976 at least partially resulted from photodegradation in the topmost
layer, where most of the TCDD was exposed to direct sunlight.
11
-------
100-
Q 90H
Q
o
I—
o 30-
c
I 701
ftJ
60-
50-
40-
30-
20-
10-
0
FIGURE 4.
0 4 8 12 16 20 24 28 32 36 40 44
depth(cm)
Site A .. : Vertical distribution of TCDD in soil, expressed
as a percentage of total TCDD detected. Sampling was per-
formed on September 27, 1976. From DiDomenico et al., 1980c.
12
-------
40-
30-
20-
10-
Q
Q
U
§ 50-
E
40-
30-
20-
10-
B
0.0 0.5 1.0 1.5 2.0
depth (cm)
FIGURE 5. Vertical distribution of TCDD in the very top soil layer of
Zone A (above) and Zone B (below). Each column represents
the average of the five assays available for Zone A and of
the 12 assays available for Zone B. From DiDomenico et al., 1980c,
13
-------
Air particles were sampled with dustfall jars at different locations
between July 1977 and February 1978. Two cosamples (liquid phase and
sediment) were obtained from each jar. TCDD was identified in 15 of 19
sediment specimens from Subzone A., site 2. Seasonal variations in
dust deposits were observed at this sampling site.
TCDD levels in the deposited dust and TCDD fallout at the site
were calculated from the previous two values. Apparently, maximum
dust and TCDD fallout occurs at the beginning of the summer. TCDD
levels in settling dust also change with the season, but follow a
different pattern. TCDD fallout reached a peak value of approximately
0.2 ng/m^/day in the most contaminated area (Subzone A,). TCDD levels
were occasionally recorded at other sites in 1978 and 1979.
Suspended particles were sampled with high-volume devices at five
locations for 24 hours in June 1977. No traces of TCDD were detected;
however, some TCDD was observed when extracts from several 24-hour
samples were pooled and analyzed. The sensitivity of the method used
was equal to 0.2 ng of TCDD/g of dust. Assuming that dust in the air
is 0.14 mg/m and the TCDD level in the dust is 1 ppb, a constantly
exposed individual thus could have inhaled as much as 1.4 pg of TCDD
daily (Di Domenico et_ al. , 1980d).
Table 2 shows the distribution of the population involved in the
accident, by zone. Zone A includes segments of the population of two
cities, Seveso and Meda; Zone B includes segments of Seveso, Cesano,
and Desio; Zone R, with a population of 31,800 people, includes parts
of six cities (Seveso, Cesano, Meda, Desio, Bovisio, and Barlassina).
14
-------
TABLE 2
Distribution of Inhabitants (No. and %) by City3 and Pollution Zoneb
Cities
Seveso
Cesano M.
Total 1
Meda
Desio
Total 2
Bovisio M.
Barlassina
Seregno
Lent ate
No. and (%)
Total
16,975
33,799
50,774
19,571
33,011
52,582
11,225
5,656
36,838
13,037
Inhabitants
Zone A Zone B
668 628
(3.93) (3.69)
2,736
(8.09)
668 3,364
(1.31) (6.62)
62
(0.31)
1,373
(4.15)
62 1,373
(0.11) (2.61)
— —
— —
— —
—
Zone R
7,945
(46.79)
14,991
(44.35)
22,936
(45.17)
4,017
(20.52)
4,608
(13.95)
8,625
(16.40)
167
(1.48)
72
(1.28)
—
—
Zone
A+B+R
9,241
(54.45)
17,727
(52.44)
26,958
(53.11)
4,079
(20.83)
5,981
(18.10)
10,060
(19.14)
167
(1.48)
72
(1.28)
—
—
Zone
non A+B+R
7,734
(45.55)
16,072
(47.56)
23,806
(46.89)
15,492
(79.17)
27,030
(81.90)
42,522
(80.86)
11,058
(98.52)
5,584
(98.72)
36,838
(100)
13,037
aoo
-------
Table 2 (continued)
Cities
Varedo
Nova M.
X
Muggio
Total 3
Total 1+2+3
No. and % Inhabitants
Total Zone A Zone B Zone R
11,841
19,467
18,690
116,754 — — 239
(0.20)
220,110 730 4,737 31,800
(0.33) (2.15) (14.47)
Zone Zone
A+B+R non A+B+R
11,841
Q-00)
19,467
0.00)
18,690
(LOO)
239 116,515
(0.10) (99.80)
37,257 182,843
17.8 (82.06)
aThe six cities with territory partially classified in one or more of the pollution zones (A,B,R) are members,
with other cities, of three health administrative departments (Consorzi Sanitari di Zona or C.S.Z). Zone
"non A+B+R", in the last column, included five cities that belong to the three C.S.Z. departments, but are
completely outside the pollution zones (e.g., Seregno, Lentate, etc.).
^Adapted from Bisanti et al., 1980.
-------
As Illustrated in Table 2, the 11 cities can be divided into three
groups. In the first group, Seveso and Cesano, more than 50% of the
population was included in the contaminated area (A + B + R); in the
second (Meda and Desio), 20% of the population was involved; and in the
third group (seven cities) only 0.2% of the people were involved.
The health surveillance plan, as described by Fara (1976a), included
the following:
1. clinical followup of each individual living in the polluted
area, primarily for humanitarian reasons and at the request of
the population, rather than for epidemiologic purposes;
2. longitudinal control of groups at higher risk (residents of
Zone A, children of Zone B, ICMESA workers, residents of all
zones who had suffered acute skin lesions or chloracne);
3. special projects, including searching for damages to reproductive,
neurological, and immunologic systems, and studies of neoplastic
and chromosomal effects; and
4. surveillance of general health indicators. Exposure data have
been gathered through individual questionnaires to ascertain the
following:
o exposure (E), i.e., presence in the polluted area on July 10, 1976;
o permanence (P), I.e., presence in the area from July 10 to August 10,
1976;
17
-------
o death of poultry and domestic animals (A) due to suspected intoxication;
o eating of their own or neighbors' vegetables and farm products (C); and
o working in the open air (handling of agricultural products, working in
the streets) or playing outdoors (for children) (G).
Information on exposure risk of any one person is translated into
a configuration of the indicators P, E, A, C, and G, classified as
follows and as shown in Table 3;
1. Absent or improbable risk; All risk indicators (replies) on
the questionnaire are "most favorable" to the subject.
2. Uncertain risk: Information for one or more risk indicators
is lacking. What data are available from the questionnaire
can be considered "favorable" to the subject.
3. Definite risk; Information on the questionnaire for at least
one of the risk indicators is "unfavorable" to the subject.
An attempt was made to correlate the exposure indicators to the
incidence of chloracne by comparing the frequency of maximal risk
configurations of 163 children with chloracne to 51 matched youngsters
without the disease. As Table 4 shows, both diseased and nondiseased
children showed similar exposure indicators.
Acute cutaneous lesions in humans, due to the toxic alkaline cloud,
might also be considered a possible indicator of pollution. As Table 5
shows, the acute skin morbidity rate was higher in residents of Zone
A than in residents of Zone B, but Zone R residents were not always less
affected than were those in Zone B (Fara, 1976b). A similar incidence
18
-------
TABLE 3
Classification of Risk for Individuals Exposed to TCDDa
Type of risk
Absent or improbable
Uncertain
Risk Associated with Factors for Each Individual
J> E A "C "G
(P) E A C G
7 ¥ A "c (?)
(P) E) A "C "G
"P I A (C)(G)
(P)(E)(A) "C G
7 I (A)(C)(G)
(P)(E)(A)(C) G
P (E)(A)(C)(G)
Certain
P IE A "C G
7 E A "C ^G
P E A C (G)
P E A C G
aFrom Caramaschi et al., 1980.
b( ) = Uncertain risk.
X; = Risk not present.
19
-------
TABLE 4
Frequency Distributions5 for Subjects Under 15 Years of Age
With and Without Chloracne11
Subjects with Subjects without
Reported Selected Chloracne Chloracne, Zone A
Risk Patterns0 No. % No. %
None P" ¥ A "C G~
Uncertain ¥1: (A) "C G~
(P) (E) (A) (C) (G) 9 16
Definite 11 ^ "C G
P IS (A) £ £
i> _E (A) c £
_P E (A) £ £
P E_ (A) £ G
£ E (A) £ G
P E (A)_ £ G - -
P E A C £ _
P E (A)_ C G 8 15
P E A C £ - - 5 10
JP E (A) C G 16 10 10 20
PE(A)_CG - - 3 6
PEACG 9 6
P E A C G 32 20 9 18
P E (A) C G 23 14 8 16
PEACG 23 14 6 12
frequencies are given only for those risk patterns presented by more than
5% of the total number of subjects in a group.
bFrom Del Corno e_t al. , 1980.
c( ) = Uncertain risk.
X = Risk not present.
20
-------
TABLE 5
Dermatologic Cases from July 10 to August 18a
(Crude incidence rates710,000 inhabitants)
City
Zone A
Zone B
Zone R
Surrounding
Areas
Meda
Seveso
586 (Meda and
Seveso)
305
120
131
20
46
Cesano
78
50
14
Desio
Other
aFrom Fara, 1976b.
21
-------
pattern was observed in the territorial distribution of the chloracne
cases, as shown in Table 6. Mapping the chloracne cases (Figure 6)
indicates a wider geographic distribution of chloracne than expected
from the known soil pollution pattern.
Additional data on possible human exposure were obtained through
the study of animal mortality (Veterinary Report, 1980). At the time
of the accident, 81,131 domestic animals lived in Zones A, B, and R:
24,885 rabbits; 55,545 poultry and other small animals; 349 cattle; 233
pigs; 49 horses; 21 sheep; and 49 goats. Poultry and rabbits were
raised mainly for household consumption in Zones A, B, and R. Cattle
(5 to 10 head per farm) were raised mainly for milk production.
Mortality started, mostly among rabbits and poultry, some few days
after the accident, and rose markedly within the first 2 weeks. By the
end of August 1976, 3,281 animals were recorded as dead (Table 7).
Rabbit mortality, by zone, is shown in Table 8. Rabbit deaths occurred
at ^ 75% of the farms in Zone A, at ^ 22% of the farms in Zone B, and
at ^ 15% of the farms in Zone R. The number of dead rabbits was 31.9%
of the total in Zone A, 8.8% in Zone B, and 6.8% in Zone R. On farms
where rabbits died, mortality was 42% in Zone A, 23% in Zone B, and 16%
in Zone R. Mortality was higher on farms where rabbits were fed grass
gathered in contaminated areas at various distances from the plant.
There is strong evidence that mortality was directly related to the
distance from ICMESA where the grass was gathered. Farms where rabbits
were fed commercial feed or fodder collected before the accident or
far from ICMESA had a lower mortality rate.
22
-------
TABLE 6
Chloracne Cases
Pollution Zone
Zone A
Zone B
Seveso, Zone R
Meda, Zone R
Cesano, Zone R
Desio, Zone R
Seveso, Zone
non A,B,R
Meda , Zone
non A,B,R
Cesano M. , Zone
non A,B,R
Desio, Zone
non A,B,R
Other cities
Total
(No. and x 1,000
(1) Chloracne
Sept. -Dec. 1976
"Early"
No. xlOOO
46 63.01
0
1 0.13
0
0
0
0
0
0
0
3
50
inhabitants)
(2)
Feb.
No.
15
9
28
20
13
2
13
14
8
5
10
137
by Pollution
Chloracne
-Apr. 1977
"Late"
xlOOO
20.55
1.90
3.52
4.98
0.87
0.43
1.68
0.90
0.50
0.18
Zones3
No
61
9
29
20
13
2
13
14
8
5
13
187
Chloracne
(1) + (2)
xlOOO
83.56
1.90
3.65
4.98
0.87
0.43
1.68
0.90
0.50
0.18
aFrom Bisanti et al., 1980.
23
-------
• •OIOM LOMSAMOI*
A r 4 case
A s 5o case*
FIGURE 6.
Distribution of chloracne cases observed in Seveso area from
July 10, 1976 to July 1977. From Del Corno et_ al., 1980.
24
-------
TABLE 7
Animal
Overall Animal Mortality (July 10 - August 31, 1976)'
No. of dead animals/total
Rabbits
Other small farmyard animals
Cattle
Horses
Pigs
Sheep
Goats
2,062/24,885
1,219/55,545
0/349
0/49
0/233
0/21
0/49
aFrom Veterinary Report, 1980.
25
-------
TABLE 8
Rabbit Mortality on Farms in the Contaminated Zones
Zone
A
B
R
Total
July
No. of Rabbits No.
Total Dead % Total
1,089 348 31.9 45
4,814 426 8.8 303
18,982 1,288 6.8 1,398
24,885 2,062 8.3 1,746
10 - August 31, 1976"
of Farms in Area
Farms with %
Deaths
34 75.5
67 22.1
208 14.9
309 17.7
No. of
Where
Total
825
1,801
7,783
10,409
Rabbits on Farms
Deaths Occurred
Dead %
348 42
426 23
1,288 16
2,062 19
.2
.6
.5
.8
From Veterinary Report, 1980.
-------
During the first period, 926 autopsies were performed on 45% of
the dead rabbits. Some 20% of these animals showed a syndrome
characterized by substernal and retrosternal edema, hemmorrhagic
tracheitis, pleural serous hemorrhage, and dystrophic lesions of hepatic
tissue. The percentage of rabbits with pathologic syndrome was 44.4%,
21.4%, and 18.4% in Zones A, B, and R, as shown in Table 9. Some dead
rabbits were analyzed for the presence of TCDD; 97%, 92%, and 75% of
the samples were positive in Zones A, B, and R, respectively. Although
fewer samples were analyzed for TCDD than were rabbits autopsied, it is
clear that the presence of TCDD did not always correlate with the
described pathologic syndrome.
Figure 7 shows the geographic distribution of animal deaths.
These occurred on farms in all three zones and also in the surrounding
areas. Of course, contaminated fodder may have been transported from
Zone A to nearby areas, thereby reducing the significance of territorial
distribution of animal deaths as an indicator of local TCDD pollution.
In summary, TCDD concentrations varied widely, even from site to
site, within the regions studied. Animal mortality was correlated with
a farm's distance from the ICMESA plant, or with the distance of the
source of the animal's food (grass or fodder) from the plant. Incidence
of chloracne in children was not related to exposure indicators. The
disease occurred over a wider geographic area than was expected from the
known TCDD soil contamination. Acute cutaneous lesions were more
frequent in residents of Zone A than in residents of Zones B and R,
but residents of Zone R did not exhibit fewer such lesions than did those of
Zone B, as was expected.
27
-------
TABLE 9
Autopsies
No. of Dead
Zones Rabbits
A 348
B 426
R 1,288
Total 2,062
on Rabbits
No. of
Autopsies
89
196
641
926
(July - August, 1976)a
% No. of Rabbits
with Pathological
Syndrome"
26 36
46 42
50 118
45 199
% of Examined
Rabbits
40.4
21.4
18.4
21.5
aFrom Veterinary Report, 1980.
^Substernal and retrosternal edema, hemorrhagic tracheitis, pleural serous
hemorrhage, dystrophic lesions of hepatic tissue.
28
-------
N
OESIO
FIGURE 7. Mapping of animal deaths. From Veterinary Report, 1980.
29
-------
REFERENCES
Bisanti, L., F. Bonettl, F. Caramaschi, G. Del Corno, C. Favaretti,
S.E. Giambelluca, E. Marni, E. Montesarchio, V. Puccinelli, G. Remotti,
C. Volpato, E. Zambrelli, G.M. Fara. 1980. Experiences from the
accident of Seveso. Acta Morphol. Acad. Sci. Hung. 28(1-2):139-157.
Caramaschi, F., G. Del Corno, C. Favaretti, S.E. Giambelluca, and E. Montesarchio.
1981. Analysis of exposure to environmental contamination by TCDD in individuals
affected by dermatological lesions. Igiene Moderna, in press.
Del Corno, G., F. Caramaschi, C. Favaretti, S.E. Giambelluca, and E. Montesarchio.
1981. Distribution of chloracne cases in the Seveso area following contamination
by TCDD. Igiene Moderna, in press.
DiDomenico, A. , V. Silano, G. Viviano, and G. Zapponi. 1980a. Accidental release
of TCDD at Seveso (Italy): II. TCDD distribution in the soil surface layer.
Ecotoxicol. Environ. Safety 4(3):238-320.
DiDomenico, A., V. Silano, G. Vivano, and G. Zapponi. 1980b. Accidental
release of TCDD at Seveso (Italy): V. Environmental persistence (half-life)
of TCDD in soil. Ecotoxicol. Environ. Safety 4(3):339-345.
DiDomenico, A. , V. Silano, G. Vivano, and G. Zapponi. 1980c. Accidental
release of TCDD at Seveso (Italy): IV. Vertical distribution of TCDD
in soil. Ecotoxicol. Environ. Safety 4(3):327-338.
DiDomenico, A., V. Silano, G. Vivano, and G. Zapponi. 1980d. Accidental
release of TCDD at Seveso (Italy): VI. TCDD levels in atmospheric
particles. Ecotoxicol. Environ. Safety 4(3):346-356.
Fara, G. M. 1976a. The history of the Seveso accident. Pp. 171-79 in E. B.
van Julsingha, J. M. Tesh, and G. M. Fara, eds. Advances in the Detection
of Congenital Malformations: Proceedings of the 5th Conference of the
European Teratology Society, Gargnano, Sept. 20-23, 1976.
Fara, G. M. 1976b. Introductory report on the epidemiological aspects.
Pp. 39-46 in A. Berlin, A. Buratta, and M. Th.-Van der Venne, eds. Pro-
ceedings of the Experts Meetings on the Problems Raised by TCDD Pollution,
Milano, Sept. 30-Oct. 1976. Regione Lombardia.
Pocchiari, F., V. Silano, and A. Zampieri. 1979. Human health effects from
accidental release of TCDD at Seveso, Italy. Pp. 311-21 in W. J. Nicholson
and J. A. Moore, eds. Volume 320, Human Effects of Halogenated Aromatic
Hydrocarbons. New York Academy of Science, New York.
Regione Lombardia, Ufficio Speciale. 1980. Veterinary Report, offset print.
30
-------
DISCUSSION
DR. MILLER: Dr. Pocchiari has summarized four major problems
related to chemical accidents at the administrative and scientific
levels: uneven soil distribution, difficulties in monitoring, lack of
rapid assay methods, and problems in coordinating the jurisdictions.
Your study, specific to Seveso, has shown that TCDD has a long half-
life (more than 10 years), that you could make some clinical observations
on the basis of experience with TCDD (there are other chemicals for
which there is no past experience), that "heavily exposed" must be
defined specifically, and that you had an advantage in that you could
observe effects in domestic animals — rabbits — acting as biologic
indicators. Autopsies of the rabbits provided you with chemical
analyses.
DR. SUSKIND: Did the group in Seveso regard chloracne as a clinical
marker? Did the appearance of acne signal exposure to a degree that
adverse effects occurred, or were there other adverse effects and other
incidents? How severe were the acne cases? In the United States, we
have heard all kinds of stories about the mildness of acne and the late
occurrence of acne in children.
PROF. DARDANONI: Chloracne is presently the only indicator we
have of human morbidity. Therefore, we rely on it to make a correlation
between soil distribution and human pathology. Severity was rather strong
in the first group of people. Late chloracne cases were much less severe.
There was a spectrum partly correlated with distance of the child's
residence from the ICMESA plant.
31
-------
DR. MOORE: You suggested that the initial half-life of TCDD was
about 3 or 4 months, and now you are talking about a half-life extending
into years. Is time a function of the chemical's location in the soil?
PROF. SILANO: These data applied only to Zone A and only to the
worked soil from this zone. Zone A was evacuated just 2 weeks after
the accident. Since then, it has not been worked, and no people are in
the zone.
This study was not a special activity to evaluate TCDD's half-life
in soil. It was an attempt to record different monitoring at different
times and to get some idea of the persistence of TCDD in soil.
We had three monitorings; the first about 1 month after the
accident, the second after approximately 5 months, and the last one
after 14 or 15 months. We were able to compare TCDD levels at 44 sites,
of which 32 were identical.
We made two estimates. The first was based on 44 sites, and the
second on 32. At the initial monitoring, there were no siginificant or
relevant differences among these estimates. TCDD levels decreased
about 50% on the average (basically a geometric mean) between the first
and the second monitorings, but we found no statistically significant
difference between the second and third monitorings.
The mathematic function that describes the overall pattern of
results indicates that the TCDD half-life in soil changes with time.
In other words, half-life is not a constant value. In fact, from the
32
-------
mathematic function, we calculated that, In measurements made at the
times I described, the half-life in this unworked soil is more than 10
years.
The literature has reported substantial differences in the half-
life of TCDD in different experiments, ranging from a few months up to
3 or more years. The data at Seveso may explain the variability.
Certainly they suggest several hypotheses.
TCDD in the very top soil or even on plants and leaves and so on
may be degraded by sunlight. After the chemical moves down into the
soil, it is not degraded by sunlight, and this fact is certainly
important.
Italy received exceptionally heavy rainfalls between August and
December; so some TCDD certainly penetrated the soil. TCDD really did
move downward. This may be not only because TCDD on the very surface
was destroyed, but also because at the moment we can say that the upper
layer of 20 cm contains most of the TCDD. Soon after the contamination
occurred, the soil layer containing 80% TCDD was much less; it was only
8 cm deep. So, there has since been some deeper penetration into the
soil.
A second hypothesis concerns volatilization of TCDD. The soil is
a very heterogeneous matrix, and one would expect TCDD, with time, to
be redistributed among the soil's colloid matrix and different components.
But stronger bonds seem to develop and prevent volatilization. Both
sunlight and increased resistance to disease may explain our results.
33
-------
Yusho (Japan)
Robert W. Miller1
Yusho (rice-oil) disease was first recognized because of an epidemic
of chloracne that began in February 1968 in Kyushu, Japan. About 1,200
people were affected; their illnesses were quickly traced to polychlorinated
biphenyls (PCB's) used as a heat transfer agent during the manufacture
of cooking oil. Nine women, pregnant at the time of their exposure,
subsequently gave birth to infants who were small-for-date and had
transiently cola-colored skin. Followup of the affected population,
thus far, has not revealed carcinogenesis or other adverse health
effects related to the PCB's. A similar leak occurred soon afterwards
in the United States; chickenfeed was polluted with PCB's, but no human
exposure was reported. Damage was revealed in reduced hatchability of
the eggs. A second chloracne epidemic developed recently in Taiwan,
again traced to cooking oil contaminated with PCB's during manufacture.
The lactational transmission of the substance observed in this instance
has raised questions about the advisability of breastfeeding by women
occupationally or otherwise exposed to PCB's.
Approximately 1,200 people around the city of Fukuoka, on the
Japanese island of Kyushu, developed chloracne from exposure to an
environmental pollutant (Taki et_ al., 1969). The cause was quickly
traced to cooking oil, which was contaminated with polychlorinated
biphenyls (PCB's) used as a heat transfer agent during manufacture of
the oil. In addition to chloracne developed by the general population,
infants born to women who ingested the oil during pregnancy were also
affected, indicating placenta! transfer of the agent. Doses of PCB's
received by the women were estimated through chemical analyses of the
cooking oil and from the history of of its ingestion by the family.
The illness was termed Yusho (rice oil) poisoning.
Clinical Epidemiology Branch, National Cancer Institute, Bethesda, Md.
34
-------
Nine cola-colored infants (as the Japanese described them) were
born with skin transiently stained a deep brown. The discoloration
cleared with time. All of the infants were exposed in utero, and were
small-for-date. In addition, two were born with teeth, a finding that
warrants additional thought.
According to Harada (1976), Children developed chloracne after
ingesting PCB's in breast milk from women who were exposed to the
chemical only after their pregnancies had ended. Breast milk samples
were not studied during the Kyushu epidemic. Generally speaking, the
levels of PCB's in breast milk in Japan are no higher than they are in
the United States (Tatsukawa, 1976).
The experience in Japan heralded PCB problems elsewhere in the
world. In the United States, Holly Farms, Inc., which is a supplier of
chickens and eggs, discovered reduced hatchability of its eggs, and
traced the problem to PCB contamination of chickenfeed (Pichirallo,
1971). The contamination occurred in a manner similar to that of the
Japanese cooking oil. PCB's, used as a heat-transfer agent, leaked
into the feed through pin-hole erosions in the pipes.
Also in the United States, the Hudson River and some of the Great
Lakes were polluted with PCB's leaking from capacitors discarded by
General Electric Co. into the waterways (Dennis, 1976). At the same
time, those areas were being stocked with salmon for sport-fishing.
Another epidemic (1,000 cases) of chloracne has recently been
reported by Chen ^t^ al_. (1980). Again, the source was contaminated
cooking oil.
35
-------
Thus, PCB's have caused areawlde food contamination, which may
have a transplacental effect and produce chloracne. They may also be
transmitted in breast milk. To date, effects in the general population
have been limited to the skin and the fetus.
36
-------
REFERENCES
Chen, P.H., J.M. Gaw, C.K. Wong, and C.J. Chem. 1980. Levels and gas
chromatographic patterns of polychlorinated biphenyls in the blood
of patients after PCB poisoning in Taiwan. Bull. Environ. Contain.
25:325-329.
Dennis, D.S. 1976. Polychlorinated biphenyls in the surface waters
and bottom sediments of the major drainage basins of the United
States. Pp. 183-194 in National Conference on Polychlorinated
Biphenyls (November 1975, Chicago, Illinois), EPA-560/6-75-004,
Environmental Protection Agency, Washington, D.C.
Harada, M. 1976. Intrauterine poisoning: Clinical and epidemiological
studies and significance of the problem. Bull. Inst. Constitutional
Med. (Kumamoto Univ.) Suppl. 25: 1-60.
Pichirallo, J. 1971. PCBs: Leaks of toxic substances raises issue of
effects, regulations. Science 173:899-902.
Taki, I., S. Hisanaga, and Y. Amagase. 1969. Report on Yusho
(chlorobiphenyls poisonings): Pregnant women and their fetuses.
Fukuoka Igaku Zasshi 60:471-474.
Tatsukawa, R. 1976. PCB pollution of the Japanese environment. Pp.
173, 177-78 in K. Higuchi, ed. PCB Poisoning and Pollution.
Academic Press, New York.
37
-------
DISCUSSION
DR. SUSKIND: As a dermatologist, I have seen some of the chloracne
cases at Kyushu University, where there has been an attempt to follow
the mothers and children. There are some reports that second pregnancies
1 year or more after the mother's exposure to PCB's seem to result in
darker than usual—hypermelanotic-infants. In addition, the chloracne
of patients first seen in 1968 persisted for 2 or 3 years.
DR. MILLER: On the Goto Islands, two or three children have been
affected in some families in which several children had been born
subsequent to the PCB exposure. The effect has, however, diminished in
successive pregnancies. The Kyushu mothers who had an affected child
have not had additional children. (Because of Catholicism on the Goto
Islands, birth control is not practiced there as it is on Kyushu.)
One more comment: Some international organization may want to
watch for similar occurrences with PCB's or other substances, so that
they can be brought promptly to international attention and their
implications can be recognized. The transmission of the first information
on PCB's was by coincidence.
DR. NEWELL: The exposure in Kyushu is very interesting, in that
it was prolonged—as opposed to exposures to other agents, which are
usually single, relatively short-term events. Is there any information
on the concentrations of PCB's or on the duration of exposure?
DR. MILLER: I do not know, although a report by Kuratsune (1976)
states that two mothers were exposed throughout their pregnancies and
that the others were exposed for only a few weeks.
38
-------
REFERENCE FOR DISCUSSION
Kuratsune, M. 1976. Epidemiologic studies on Yusho. P. 20 in K.
Higushi, ed. PCS Poisoning and Pollution. Academic Press.
New York.
39
-------
Treatment of Chlordecone (Kepone) Poisoning with Cholestyramine
Philip S. Guzelian X
Due to the lack of adequate industrial hygienic protection,
workers in a factory that manufactured chlordecone as its only
product were exposed to large quantities of the organochlorine
pesticide for many months. The salient clinical features in
32 of the most severely affected workers, all males, involved
the nervous system, liver, and testes. High levels of chlordecone
were found in samples of blood, adipose tissue, and liver. However,
the pesticide was absent from excretory fluids (sweat, urine, and
sebum), but present in gallbladder bile. Only 5-10% of the biliary
chlordecone appeared in the stool, which suggested that chlordecone
may be reabsorbed in the intestine. Oral administration of
cholestyramine, a nonabsorbable anion exchange resin that binds
chlordecone iri vitro, increased the fecal elimination of chlordecone
from the body and accelerated its disappearance. Recent evidence
indicates that chlordecone enters the intestine by an alternate,
nonbiliary mechanism (probably via the gut itself), and that net
excretion of chlordecone from this source can be augmented by
cholestyramine administration. Coincident with elimination of
chlordecone from the tissues was an amelioration of the signs and
symptoms of chlordecone toxicity. Since other lipophilic toxins
also appear to be excreted by this mechanism, oral administration
of selected binding agents may provide a practical, therapeutic
approach to treating poisonings from many types of environmental
chemicals.
In July 1975, a chemical worker for Life Science Products
Corporation, Hopewell, Va., which manufactured chlordecone (Kepone)
as its only product, was discovered to have high concentrations of
that chemical in his blood and severe toxic manifestations involving
his neurologic system, liver, and testes (Cohn et_ al. , 1978).
Epidemiologic followup of this index case by Federal and State health
officials revealed that the small factory consisted of a renovated
gasoline service station, which lacked adequate industrial hygienic
LMedical College of Virginia, Richmond, Va.
40
-------
protective measures and employee safety provisions (Cannon et al.,
1978). These conditions led to frequent exposure of the workers to
large quantities of the organochlorine pesticide. Indeed, two-thirds
of the production workers employed for all or a portion of the 16
months the factory was in operation exhibited clinical evidence of
chlordecone toxicity, and almost all had detectable concentrations
of chlordecone in blood samples (Table 1). The production workers
wore no special protective clothing and took their work clothes
home, which probably explains the contamination of their family
members who also had detectable chlordecone in blood samples (Table
1). Because the factory discharged chlordecone into the air and
water, the city's sewage system became contaminated and sewage workers
and residents of the area were also exposed to chlordecone (Table 1).
It was learned subsequently that Allied Chemical Corporation,
which had been the exclusive manufacturer of chlordecone prior to
its contractual arrangement with Life Science Products Corporation,
had discharged the pesticide into the James River for more than 10
years. Evidence of contamination of the animal and marine life of
the entire tidewater region of Virginia has been gathered, but the
total number of individuals at risk of low-dose exposure to chlordecone
in the environment is presently unknown.
A team of clinical investigators studied a group of 32 of the
most severely affected men, aged 18 to 47. Each had signs or symptoms
(or both) consistent with chlordecone toxicity and a concentration
41
-------
TABLE 1
Epidemiology of Chlordecone Poisoning at
Life Science Products Corporation, Hopewell, Va.
a
Percent
Finding Subjects (No. Affected/No. Studied) Affected
Attack Rate: Production workers (73/114) 67
Chlordecone All workers (105/106) 99
detected in blood:
Family members (30/32) 94
Residents of Hopewell (40/214) 19
Data from Cannon et al., 1978.
42
-------
of chlordecone in the blood greater than 600 ng/ml. Evidence of
neurologic impairment included a motor disorder manifested by
appendicular tremor, stuttering speech, and opsoclonia (arrhythmic
eye movements). Also frequent were findings of anxiety, short-term
memory loss, change in personality, and headaches. In the majority
of these patients, the liver was enlarged, although chemical tests
of liver function showed normal results. Morphologic examination
of liver biopsies revealed proliferation of the smooth endoplasmic
reticulum as the most consistent finding. Evidence for impairment
of testicular function included oligospermia and a decreased per-
centage of motile sperm. The salient features of the clinical syndrome
exhibited by these men have been described by Cohn et_ al. (1978) and
Taylor et_ al. (1978).
No cases of human poisoning with chlordecone had been reported
previously, and no specific treatment for this condition was availa-
ble. Indeed, although toxicity and even death have occurred in
humans exposed to chlorophenothane (DDT) (Case, 1945; Hill and Robinson,
1945; Jenkins and Toole, 1964; Mackerras and West, 1946; Wigglesworth,
1945), dieldrin (Garrettson and Curley, 1969; Hoogendam et al., 1965;
Jindal, 1968; Patel and Rao, 1957), aldrin (Hoogendam et^ al. , 1965;
Jenkins and Toole, 1964; Kazantizis £t aL., 1964; Spiotta, 1951),
endrin (Coble et^ al., 1967; Davies and Lewis, 1956), chlordane (Curley
and Garrettson, 1969; Derbes £t_ al., 1955), polychlorinated biphenyls
(Fishbein, 1974; Hirayama, 1976), and polybrominated biphenyls (Kay,
1977), no antidotes are capable of reversing the toxic lesions created
by these organochlorine chemicals.
43
-------
One approach to therapy was to devise a means to accelerate
chlordecone excretion from the body, on the assumption that the
substance's presence in the tissues is responsible for toxic mani-
festations, constitutes a risk of carcinogenesis, or both. The
issue of carcinogenesis was raised by studies demonstrating that
chlordecone is a hepatic carcinogen in rats and mice (Reuber, 1978).
Accordingly, a study of the distribution and excretion of chlordecone
in these patients revealed that the pesticide is excreted into the
intestine by the liver bile (Cohn e^ al., 1978) and also by a novel,
nonbiliary mechanism (BoyIan et al., 1979). Based on this information
and current concepts of gastrointestinal physiology, Guzelian et al.
(1980) established that oral administration of cholestyramine, a
nonabsorbable anion exchange resin that binds chlordecone in vitro,
increases the excretion of the pesticide in the stool and accelerates
its elimination from tissues, thereby ameliorating the clinical
manifestations of toxicity.
METHODS
Details of the patient population, and the methods used to obtain
and prepare samples of blood, fat, bile, and stool have been published
elsewhere (Blanke et_ a±., 1977; Boylan _e_t al_., 1979; Cohn et al. ,
1978; Guzelian e_t^ al. , 1980). Chlordecone was measured by gas-liquid
chromatography using an electron-capture detector.
44
-------
RESULTS
Distribution and Excretion of Chlordecone and Effects of Cholestyramine
Treatment
The highest concentrations of chlordecone were found in the blood,
liver, and adipose tissue (Table 2). The ratio of the concentration
of chlordecone in the blood compared to its concentration in fat
(approximately 1:7) is extremely high when compared to the ratio
reported for other organochlorine pesticides. One explanation for this
phenomenon is the specific binding of chlordecone by plasma proteins,
particularly by high-density lipoproteins (Skalsky et_ ad., 1979). A
practical consequence of this unusual distribution is that the measurement
of chlordecone in blood is a far more sensitive indicator of exposure
than would be expected for other organochlorine pesticides. Indeed,
across a 5,000-fold range of chlordecone concentrations in adipose
tissue, a strict ratio of 7:1 was maintained with the respective blood
concentration (Cohn et_ al. , 1978).
Of excretory fluids examined (sweat, urine, saliva, gastric juice),
only gallbladder bile contained significant quantities of chlordecone
(Table 2). The rates of excretion of chlordeconed in hepatic bile,
measured by aspiration of duodenal contents, represented a relatively
constant proportion (0.5-1.0%) of the estimated total body content
of chlordecone (Cohn et al., 1978). An important finding, however,
was that only 5-10% of the biliary chlordecone entering the lumen of
the duodenum appeared in the feces each day. This discovery suggested
that 90-95% of the biliary chlordecone was reabsorbed in the intestine
and recirculated to the liver, thus undergoing enterohepatic recirculation.
45
-------
TABLE 2
Distribution of Chlordecone in Workers3
Tissue
Whole blood
Liver
Subcutaneous
fat
Muscle
Gallbladder
bile
No. of
Patients
32
10
29
5
6
Range of
Chlordecone
Concentration (yg/g)
0.6-32.0
13.3-173.0
1.7-62.1
1.2-11.3
2.5-30.0
Partition
Blood
Tissue Range
1.0
15.0 4.6-31.0
6.7 3.8-12.2
2.9 1.8-4.5
2.5 1.4-4.1
From Cohn et_ jil_., 1978; reprinted by permission of The New England Journal
of Medicine 298:243-248, 1978.
46
-------
To test the hypothesis that cholestyramlne, which binds
chlordecone in vitro, might also bind this substance in the intestine,
the investigators administered this resin orally (24 g/day) to nine
patients for 5 days. There were marked increases in the fecal excre-
tion of chlordecone, ranging from 3.3-fold to 17.8-fold (Figure 1).
By comparison, oral administration of activated charcoal to three
patients produced only a minimal increase in fecal excretion of
chlordecone. Similarly, a 2-week cholestyramine-feeding study in
chlordecone-treated rats established that the resin not only
increased fecal excretion of the pesticide, but also decreased
total body content of chlordecone (Boylan et^ al_., 1978).
Encouraged by these preliminary studies, the researchers
carried out a double-blind, randomized clinical trial of
cholestyramine therapy (Cohn et_ al_., 1978). Patients were classed
into three groups, according to the concentration of chlordecone in
their blood when they entered the study. Within each group, the men
were randomly allocated to treatment with either cholestyramine or
placebo, and were observed twice monthly for 5 months. Cholestyramine
significantly accelerated the rate of disappearance of chlordecone
from the blood (Figure 2) and also from adipose tissue (Cohn et al.,
1978).
47
-------
CD 6.0
O 5.0
LaJ
X
LU
LU 3.0
2
O
UJ 2.0
Q
o:
I
o
1.0
o
14.3
j=l
CHOLESTYRAMINE
(24g/d)
CONTROL
r
r
r
8
9 10
PATIENT
II
12
13
14
FIGURE 1. Stimulatory effect of cholestyramine on the rate of excretion
of chlordecone in stool. Control collections of stool were
obtained from each patient for at least 72 hours. Treatment
with cholestyramine was started and, 48 hours later, stool was
collected for a second 72-hour period. From Cohn et al., 1978,
with permission from The New England Journal of MedTcine
298:243-248, 1978. ~~~~
48
-------
Q 250
O
3
CD
— 200
UJ
§
Q
a:
3
o
u.
o
UJ
150
100
50
--I39
I
I
I
I
UNTREATED CHOLESTYRAMINE UNTREATED
PLACEBO
FIGURE 2. Effect of cholestyramlne or placebo on the half-life of
chlordecone in the blood of exposed workers. Open circles
denote patients with statistically significant differences
between the half-lives of chlordecone in the control period
(untreated) and the treatment period (p < 0.05). Average
half-lives are given as mean + S.E.M. (brackets). From
Cohn et_ al^., 1978, reprinted with permission from
The New England Journal of Medicine 298:243-248, 1978.
49
-------
Amelioration of Toxic Manifestations with Cholestyramine Treatment
After the clinical trial, all patients were placed on cholestyramine.
Within 1 year, their blood showed low or undetectable concentrations
of chlordecone. Coinciding with a decrease in chlordecone levels in
blood (and presumably in total body content) was the disappearance
of objective neurologic findings. All but one of the patients was
able to return to work. In addition, in patients with hepatomegaly,
the liver returned to normal size, repeat liver biopsies showed that
the previously observed nonspecific abnormalities had disappeared,
and urinary excretion of glucaric acid (a measure of induction of
smooth endoplasmic reticulum in the liver) had declined from markedly
elevated values into the normal range (Guzelian et_ _al. , 1980).
Finally, the numbers of motile sperm increased in 12 of 13 patients,
and blood chlordecone concentration declined (Cohn et al., 1978).
Evidence for a Nonbiliary Mechanism for Excretion of Chlordecone by the
Gastrointestinal Tract
Discovery of a nonbiliary mechanism for entry of chlordecone into
the intestine was facilitated by a unique opportunity to study biliary
excretion of chlordecone directly in one patient who required
cholecystectomy for gallstones and agreed to have a T-tube implanted
in his common bile duct. Details of the experiments have been published
by Boylan et_ £.L. (1979). Pertinent data are summarized in Table 3.
Hepatic bile was diverted through the T-tube and small samples were
taken for analysis of chlordecone and its metabolite, chlordecone
alcohol. The rest of the bile was reinfused into the duodenum through
50
-------
TABLE 3
Excretion of Chlordecone in T-Tube Bile and Stool in a Single Patient3
Bile Excretion (ug/24 hr) Stool Excretion (yg/24 hr)
Chlordecone Chlordecone
Condition Chlordecone alcohol Chlordecone alcohol
Intact circuit
(bile reinfused) 593 486 88 195
Interrupted circuit
(bile diverted) 258 250 240 <5
aFor each condition, bile was collected for 24 hours during the second day of a
72-hour stool collection; bile was either saved or infused into the duodenum
through a surgically implanted tube (Boylan et^ al. , 1979). Chlordecone and
Chlordecone alcohol were analyzed by modifying a previously described method
(Boylan et al., 1979; Blanke £t al. , 1977) in that samples were digested with
3-glucuronidase and then autoclaved (115°C at 20 psi for 15 min) in the presence
of sulfuric acid (10% final concentration). This treatment appeared to release
Chlordecone, producing higher values than those obtained using the previous
method (Boylan et al., 1979). Data are the average of two experiments.
51
-------
a surgically implanted tube. Under these conditions (enterohepatic
recirculation intact), only 15% of the biliary chlordecone appeared
in the stool each day, confirming earlier studies using duodenal
aspiration techniques (Cohn &t_ al_. , 1978). Similarly, chlordecone
alcohol seemed to undergo enterohepatic recirculation, although to
a lesser extent (60% reabsorbed). When bile diverted from the
T-tube was not reinfused (thus, interrupting enterohepatic recirculation),
chlordecone alcohol failed to appear in the stool (Table 3). This
finding is consistent with the interpretation that chlordecone alcohol
is formed in the liver and that bile is the sole source of the chlordecone
alcohol in the feces. In contrast, such diversion of the bile stream
did not reduce or terminate fecal excretion of chlordecone. Indeed,
the excretion of chlordecone in stool increased following biliary
diversion (Table 3). This observation, confirmed in rats (Boylan
et_ al. , 1979), establishes the existence of a nonbiliary mechanism
(probably the gut itself) for entry of chlordecone into the intestine.
Moreover, it may be inferred that constituents in bile inhibit the
net transportation of chlordecone into the lumen by this nonbiliary
pathway (Table 3).
OVERVIEW
Polyhalogenated hydrocarbon chemicals have been produced and used
in steadily increasing amounts during the last 25 years. As a conse-
quence, they have become ubiquitous contaminants of our environment.
There is little specific information on the pathobiologic effects of
environmental chemicals in humans, perhaps because potential adverse
52
-------
health effects of chronic exposure to low doses of these substances
are difficult to study. The unfortunate industrial exposure of
chemical workers to chlordecone provides several unique advantages
for investigating these effects. The workers were all young, adult
males and were exposed to a single chemical, at high doses, over a
well-defined period of time.
Blanke et^ al. (1977) first developed and validated a method to
measure chlordecone in biologic samples. Methods for analysis of
environmental chemicals often are validated by samples "spiked" with
standards. This technique may be inappropriate for measurements of
chlordecone and probably for many other lipophilic chemicals in
biologic samples. Extraction of chlordecone from excreta or tissues
of animals treated with C-chlordecone differs significantly from
samples of the same tissues spiked with exogenously added standards.
Indeed, amounts of chlordecone extracted from biologic samples varies
widely among different tissues and may even differ among replicate
samples within a given tissue. Whereas appropriate external standards
(monohydrochlordecone) can correct for the latter variation, careful
analysis of endogenously labeled samples is necessary to correct .for
the former.
The researchers determined that chlordecone was stored in the
highest concentrations in adipose tissue and in the liver and was
excreted in bile into the gastrointestinal tract. However, there is
a substantial enterohepatic recirculation of chlordecone, which
curtails the overall elimination of the pesticide from the body.
Accordingly, the hypothesis that cholestyramine might block the
53
-------
reabsorption of chlordecone in the intestine and increase its excretion
in the stool was tested. Both short-term administration of cholestyramine
to humans and animals, as well as a long-term controlled clinical trial
of cholestyramine verified this hypothesis. Coincident with chemical
detoxification of the patients was an amelioration of the signs and
symptoms of chlordecone toxicity. This finding implies that the
clinical manifestations of toxicity sustained in some patients for
the 16 months they worked at Life Science Products Corporation and
for many months thereafter were not irreversible.
The discovery of the nonbiliary mechanism for entry of
chlordecone into the intestine was made possible by an opportunity
to measure directly the excretion of an organochlorine chemical in
human bile. One important feature of the nonbiliary mechanism is
that more chlordecone undergoes net translocation into the intestinal
lumen via this pathway than by biliary excretion. Thus, the intestine
should be considered as a primary excretory organ as well as a conduit
for products of hepatic metabolism. A second important feature of
the nonbiliary pathway is that it may transport chemicals other than
chlordecone. In monkeys and rats, halogenated chemicals such as mirex
(Pittman _e_t al. , 1976), dieldrin (Heath and Vadecar, 1964), and poly-
chlorinated biphenyls (Yoshimura and Yoshihara, 1975), as well as
endogenous lipophilic substances such as unconjugated bilirubin (Lester
_et_ al. , 1962), are excreted in the stool in larger quantities than in
bile. Thus, a variety of hydrophobic chemicals, including those that
are poorly translocated into bile can be eliminated via the nonbiliary
route. If the precise mechanisms of nonbiliary transport can be
54
-------
elucidated and suitable binding agents developed for these chemicals,
then promising means may be at hand to enhance the removal of slowly
excreted lipophilic toxins from the body.
55
-------
REFERENCES
Blanke, R.V., M.W. Fariss, F.D. Griffith, Jr., and P.S. Guzelian.
1977- Analysis of chlordecone (Kepone) in biological specimens.
J. Analyt. Toxicol. 1:57-62.
Boylan, J.J., W.J. Cohn, J.L. Egle, Jr., R.V. Blanke, and P.S. Guzelian.
1979. Excretion of chlordecone by the gastrointestinal tract:
Evidence for a nonbiliary mechanism. Clin. Pharmacol. Ther.
25:579-585.
Boylan, J.J., J.L. Egle, Jr., and P.S. Guzelian. 1978. Use of
cholestyramine as a new therapeutic approach for chlordecone
(Kepone) poisoning. Science 199:893-895.
Cannon, S.B., J.M. Veazey, Jr., R.S. Jackson, V.W. Burse, C. Hayes,
W.E. Straub, P.J. Landrigan, and J.A. Liddle. 1978. Epidemic
Kepone poisoning in chemical workers. Am. J. Epidemiol. 107:529-537.
Case, R.A.M. 1945. Toxic effects of 2,2-bis(p-chloro-phenyl)l,l,l-
trichlorethane (DDT) in man. Br. Med. J. 2:842-845.
Coble, Y., P. Hildebrandt, J. Davis, F. Raasch, and A. Curley. 1967.
Acute endrin poisoning. J. Am. Med. Assoc. 202:489-493.
Cohn, W.J., J.J. Boylan, R.V. Blanke, M.W. Fariss, J.R. Howell, and
P.S. Guzelian. 1978. Treatment of chlordecone (Kepone) toxicity
with cholestyramine: Results of a controlled clinical trial.
N. Engl. J. Med. 298:243-248.
Curley, A., and L.K. Garrettson. 1969. Acute chlordane poisoning.
Arch. Environ. Health 18:211-215.
Davies, G.M., and I. Lewis. 1956. Outbreak of food poisoning from
bread made of chemically contaminated flour. Br. Med. J.
2:393-398.
Derbes, V.J., J.H. Dent, W.W. Forrest, and M.F. Johnson. 1955.
Fatal chlordane poisoning. J. Am. Med. Assoc. 158:1367-1369.
Fishbein, L. 1974. Toxicity of chlorinated biphenyls. Ann. Rev.
Pharmacol. 14:139-156.
Garrettson, L.K., and A. Curley. 1969. Dieldrin: Studies in a
poisoned child. Arch. Environ. Health 19:814-822.
Guzelian, P.S., G. Vranian, J.J. Boylan, W.J. Cohn, and R.V. Blanke.
1980. Liver structure and function in patients poisoned with
chlordane (Kepone). Gastroenterology 78:206-213.
Heath, D. , and M. Vadecar. 1964. Toxicity and metabolism of dieldrin
in rats. Br. J. Ind. Med. 21:269-279.
56
-------
Hill, K.R., and G. Robinson. 1945. A fatal case of DDT poisoning in
a child. Br. Med. J. 2:845.
Hirayama, C. 1976. Clinical aspects of PCB poisoning. Pp. 87-102
in K. Higuchi, ed. PCB Poisoning and Pollution. Academic
Press, New York.
Hoogendam, I., J.P.J. Versteeg, and M. de Vlieger. 1965. Nine years
toxicity control in insecticide plants. Arch. Environ. Health
10:441-448.
Jenkins, R.B., and J.F. Toole. 1964. Polyneuropathy following exposure
to insecticides. Arch. Intern. Med. 113:691-695.
Jindal, H.R. 1968. Bilateral retrobulbar neuritis due to insecticides.
Postgrad Med. J. 44:341-342.
Kay, K. 1977. Polybrominated biphenyls (PBB) environmental contamination
in Michigan, 1973-1976. Environ. Res. 13:74-93.
Kazantizis, G., A.I.G. McLaughlin, and P.P. Prior. 1964. Poisoning
in industrial workers by the insecticide, aldrin. Br. J. Ind.
Med. 21:46-48.
Lester, R. , L. Rainmaker, and R. Schmid. 1962. A new therapeutic approach
to unconjugated hyperbilirubinaemia. Lancet 2:1257.
Mackerras, I.M., and R.F.K. West. 1946. DDT poisoning in man.
Med. J. Aust. 1:400-401.
Patel, T.B., and V.N. Rao. 1957. Dieldrin poisoning in man. Br. Med. J.
1:919-921.
Pittman, K., M. Weiner, and D.H. Treble. 1976. Mirex kinetics in the
Rhesus monkey, Drug Metab. Dispos. 4:288-295.
Reuber, M.D. 1978. Carcinogenicity of Kepone. J. Toxicol. Environ.
Health. Perspect. 4:895-911.
Skalsky, H.L. , M.W. Fariss, R.V. Blanke, and P.S. Guzelian. 1979.
The role of plasma proteins in the transport and distribution
of chlordecone (Kepone) and other polyhalogenated hydrocarbons.
Ann. N.Y. Acad. Sci. 320:231-237.
Spiotta, E.J. 1951. Aldrin poisoning in man. Arch. Ind. Hyg. Occup.
Med. 4:560-566.
Taylor, J.R., J.B. Selhorst, S.A. Houff, and J. Martinez. 1978.
Chlordecone intoxication in man. I. Clinical observations.
Neurology 28:626-630.
Wigglesworth, V.B. 1945. A case of DDT poisoning in man. Br. Med. J.
1:517-
Yoshimura, H., and H. Yoshihara. 1975. A novel route of excretion of
2,4,3',4'tetrachlorobiphenyl in rats. Bull. Environ. Contam.
Toxicol. 13: 681-688.
57
-------
DISCUSSION
DR. MILLER: There were two lessons I learned from your
presentation. First, by understanding pathogenesis, one may hit on
a novel idea for treatment. Second, pollutants can be carried on
the clothes of the worker into their homes, where they can affect
or at least get into the blood of the family. But you did not
mention how Kepone was first recognized as harmful to workers.
DR. GUZELIAN: The workers in this factory believed that there
was some adverse health effect related to Kepone. They ate in the
room where the dust was packaged and developed some clinical mani-
festations — generally tremors, called the "Kepone shakes." The
workers repeatedly asked the company medical directors whether or
not the chemical was harmful. They were told it was not.
Several of the workers sought medical advice in the community.
One in particular did not receive a careful review of his work
history, and his complaints were dismissed as effects of anxiety.
He was treated with tranquilizers. The cause was not identified
until a general practitioner took a careful work history and
forwarded blood samples to the Centers for Disease Control (CDC).
These samples contained high amounts of chlordecone, which led to
the eventual descent on the town by Federal and State health officials.
58
-------
DR. MILLER: The point I wanted to bring out was that an alert
clinician referred the specimens to the CDC.
PROF. GARATTINI: In the preliminary screening, did you find
any other agent that was able to bind chlordecone? According to
some studies we conducted in mice, cholestyramine is not effective
in ridding the body of TCDD (2, 3, 7, 8-tetrachlorodibenzodioxin).
DR. GUZELIAN: We have some other agents, including XAD-2,
which seemed almost as effective as cholestyramine. We have not
studied mice, but there is an important species difference between
humans and rats. I think it would be important to test cholestyramine
in higher animals rather than in rodents because the latter do not
have very prominent inhibition of their nonbiliary mechanism for
handling these substances.
Cholestyramine probably does not act solely by binding Kepone,
as I originally thought, also indirectly by binding bile salts and
releasing inhibition of the nonbiliary mechanism. Bile salts may be
important in altering the disposition of chemicals across the mucosal
barrier of the gut wall. This area needs to be investigated thoroughly.
Despite negative results in rodents, I would not dismiss this approach
to therapy for any of these compounds.
59
-------
DBCP
Donald Whorton-*-
Dibromochloropropane (DBCP) was a widely used nematocide until
it was discovered to cause adverse effects on testicular function
among formulation and production workers. When formulation workers
at a California pesticide plant noticed infertility problems, they
joined together to investigate. Studies at this plant showed a strong
association between altered testicular function (azoospermia and
oligospermia) and workplace exposure to DBCP. Subsequently, studies
at other U.S. and Israeli DBCP production facilities have shown
similar results. Followup studies have shown the effects to be
reversible when testicular function is not severely affected.
Dibromochloropropane (DBCP) is a liquid nematocide used widely
in the United States since the mid-1950's on such crops as citrus
fruits, grapes, peaches, pineapples, tomatoes, and soybeans. In
Central America and Israel, it is used on bananas. Thus, it is
applied both to annual and perennial plants, and harms neither.
The toxicity of DBCP has been recorded in the open scientific
literature. In 1961, it was shown to be a mild irritant and an
hepatic and renal toxin and to produce testicular atrophy in laboratory
animals. (Torkelson et_ ad., 1961). In 1973, it was reported to be
carcinogenic in animals (Olson et_ a.L. , 1973) and in 1975, it was also
determined to be a mutagen in. vitro (Rosenkranz, 1975). Finally, in
1977, it was shown to cause infertility and sterility in humans
(Whorton et al., 1977).
Department of Biomedical and Environmental Health Sciences, School
of Public Health, University of Califorina, Berkeley.
60
-------
The occupational population at risk includes both formulators
and applicators of the compound. The fertility problem was first
noticed at a manufacturing plant in central California, which produced
DBCP, ammonia, and fertilizers (Whorton et_ al. , 1977). The pesticide
formulation area of the plant was small, employing less than 15% of
the company's work force. Men in that part of the plant were young,
and many wanted to have children. However, the investigators noted
that men working in that section of the plant ceased to have children.
After considerable discussion among themselves, five of the
employees decided to conduct their own investigation and took semen
samples to a local laboratory for analysis. The laboratory would not
release the results directly to the men, and I became involved,
receiving and subsequently verifying the abnormal findings.
The remainder of the workers in that area of the plant, 36 men,
were then tested. Of these, 11 had had vasectomies; the 25 remaining
men provided semen samples for analysis. These 25 were divided into
three groups, according to length of exposure, which was determined
by length of employment in that section of the plant. Eleven men in
Group A had the longest exposure: an average of 8 years; range, 4-15
years. The 11 men in Group B were either exposed very briefly ( 0.08
years) or exposed to low concentrations. Group C consisted of three
men whose exposure fell between Groups A and B.
61
-------
The mean sperm counts of these groups were significantly different:
Group A was 200,000, as compared to 93,000,000 in Group B (Table 1).
In addition, 9 of the 11 men in Group A were azoospermic and had
significantly increased levels of follicle-stimulating and luteinizing
hormones. There were no other group findings such as physical symptoms
or laboratory results, other than abnormalities of testicular function.
The three men in Group C had been exposed for approximately 1 year
and had sperm counts ranging between 10-30 million.
Because of the range of findings in the exposed workers, the
rest of the employees at the plant were studied under contract to the
National Institute of Occupation Safety and Health. A control group
of 35 people with no exposure was developed. The duration of exposure
to DBCP was estimated for each of 90 men, but could not be estimated
for 17 others known to have been exposed. The following ratios were
derived. For controls, the ratio of those with a sperm count higher
than 20 million/ml to those with a sperm count less than 20 million/ml
was 34 to 1. For the exposed groups, the ratio decreased with length
of exposure to DBCP — from 11 to 1 for those exposurd 1-6 months to
0.31 to 1 for those exposed more than 42 months (Table 2).
There were six similar studies in the United States and one
Israeli report during 1977-78 (Table 3). All were cross-sectional
epidemiologic clinical evaluations, and have been reported in literature
or at some scientific symposium or forum. All study subjects had
some exposure to DBCP, although exposure was not uniform in duration
or concentration. Of 485 men examined, 14.6% were azoospermic and
62
-------
TABLE 1
Comparison of Nonvasectomized DBCP Workers With Very Low (Group A)
and Normal (Group B) Sperm Counts3
No. of Age, Exposure, Sperm Count Follicle Stimulating Luteinizing Hormone, Testosterone
Group Subjects Years Years x 106/ml Hormone, mlubml mlu/ml ng/dl
A llc 32.7 + 1.6d 8.0 + 1.2e 0.2 + 0.106 11.3 + 1.86 28.4 + 3.3d 459 + 35
B 11 26.7 + 1.2d 0.08 + 0.02e 93 + 18e 2.6+0.46 14.0+2.8d 463 + 31
aFrom Whorton et al., 1977. All results given as mean + standard error of the mean.
mlu = milli International unit
GNine were azoospermic.
dp < 0.01
ep < 0.001
-------
TABLE 2
Ratio of Oligospermia to Normospermia for 126 Nonvasectomized Men
Exposed to DBCP for Different Lengths of Timea
Exposure
No. of
Subjects
Ratio of Subjects with Sperm
Counts > 20 million/ml to those
with Sperm Counts < 20 million/ml
None (Control
group)
35
34
1-6 Months
48
11
7-24 Months
14
2.5
25-42 Months
12
0.5
> 42 Months
17
0.31
aFrom Whorton et al., 1979.
64
-------
TABLE 3
Summary of DBCP Studies 1977-78
Stud-s
Oxy Chem
(California)13
Shell
(Colorado)0
Shell
(Alabama)0
Dow
(Arkansas)
California
Applicators
Israel^
EPA"
(North and South
Carolina and Texas)
Totals
No. of
Subjects
Examined3
114
64
71
86
74
23
53
485
No. of Azoospermic and
Oligospermic Men by Sperm
Count, Millions/ml
Azoospermic
(0 count)
15
5
1
30
6
12
2
Oligospermic
0.1-9 10-19
8
2
6
(17)e
8
6
12
12
7
5
(3)
(7)
0
8
71
59
42
Percent
14.6
12.2
8.7
^Variable exposures.
bWhorton et al., 1977, 1979.
cLipschultz e£ al., 1980.
dScharnweber, 1982; Glass £lt al^. , 1979.
eNumbers in parentheses extrapolated by me from data presented by original author
fGlass j;t al. , 1979.
gPotasnik at al., 1979.
hSandifer et al., 1979.
65
-------
20.9% were oligospermic. Thus, approximately 35% of the men studied
exhibited a significant biologic effect resulting from their DBCP
exposure.
At the central California plant, 12 men were azoospermic in 1977
and remained so in 1978. Of the men who were oligospermic in 1977,
six of nine were normal (> 20 million sperm per milliliter) in 1978,
(Table 4) (Unpublished follow-up data collected by me in 1979 and
1980 show the 12 azoospermic men in 1977 and 1978 to have remained
azoospermic.).
Biopsies from testicles from these exposed California men were
also studied. In an individual exposed less than 3 months, the
seminiferous tubules showed marked cellularity and active spermatogenesis.
However, a biopsy from an azoospermic individual with 10 years'
exposure showed seminiferous tubules that were somewhat collapsed and
devoid of cellularity. The cells present were Sertoli's cells.
There was no evidence of spermatogenesis. The insterstitial stroma
had increased minimally, probably due to the collapse of the seminiferous
tubules rather than to fibres is. There was no evidence of inflammation;
the spermatogenic cells seemed simply to have disappeared. Thus, the
pathological examination showed no inflammation, no fibrosis, a
Sertoli-cell only appearance, and normal interstitial cells, both
under light and electron microscopy.
Several questions remain unanswered regarding DBCP exposure.
Those men who were less exposed and less affected by the DBCP resumed
spermatogenesis. However, no study has been performed of their
66
-------
TABLE 4
1977 vs. 1978 Sperm Count of 21 Men Exposed to DBCP Longer Than 30 Days'
Sperm Count,
million/ml
No. of Men, by year
1977 1978
0
0-9
10-19
> 20
12
4
5
0
12
1
2
6
Total
aFrom Whorton and Mllby, 1980.
21
21
67
-------
Also unaddressed is the possible carcinogenic effect of DBCP in
humans. DBCP is known to be carcinogenic in animals and mutagenic in
in vitro systems. Thus, those men with families and no longer
concerned with future parenthood are especially concerned with the
carcinogenic aspects of their exposure.
The third problem is the unresolved question of the chemical's
effect on women; none of the studies to date has included enough
women to evaluate.
The last concern is the effect of low concentrations of DBCP in
drinking water. The presence of the chemical is currently an important
issue in California, because approximately one-third of the water
supply in the central valley comes from deep wells contaminated with
varying amounts, ranging from parts per trillion to 5-10 parts per
billion.
The last point is a statement rather than a question. Observant
and thoughtful workers are responsible for the discovery of the
fertility problems caused by DBCP exposure.
-------
REFERENCES
Glass, R.L., R.N. Lyness, D.C. Mengle, e_t al. 1979. Sperm count
depression in pesticide applicators exposed to dibromochloro-
propane. Am. J. Epidemiol. 3:346-351.
Lipschultz, L.I., C.E. Ross, M.D. Whorton, et al_. 1980. Dibromo-
chloropropane (DBCP) and its effects on testicular function in
man. J. Urol. 124:464-468.
Olson, W.A., R.T. Habermann, E.K. Weisburger, J.M. Ward, and
J.H. Weisburger. 1973. Induction of stomach cancer in rats
and mice by halogenated aliphatic fumigants. J. Natl. Cancer
Inst. 51:1993-1997.
Potasnik, G., N. Ben-Aderet, R. Israeli, £t al. 1979. Effect of
1,2-dibromo-3-chloropropane on human testicular function.
Israeli J. Med. Sci. 15:438-442.
Rosenkranz, H.S. 1975. Genetic activity in 1,2-dibromo-3-chloropropane,
a widely-used fumigant. Bull. Environ. Contain. Toxicol.
14:8-12.
Sandifer, S.H., R.I. Wilkins, C.B. Loadholt, L.G. Lane, and J.C. Eldridge,
1979. Spermatogenesis in agricultural workers exposed to
dibromochloropropane (DBCP). Bull. Environ. Contain. Toxicol.
23:703-710.
Scharnweber, H.C. 1982. The Dow experience. Pp. 30-42 in C. Meyer,
M. Curry, and J. Lybarger, eds. Proceedings of NIOSH Conference on
Dibromochloropane, Cincinnati, Ohio, October 19-20, 1977. National
Institute for Occupational Safety and Health, Washington, D.C.
Torkelson, T.R., S.E. Sadek, V.K. Rowe, J.K. Kodama, H.H. Anderson,
G.S. Loquvan, and C.H. Hine. 1961. Toxicological investigations
of l,2-dibromo-3-chloropropane. Toxicol. Appl. Pharmacol.
3:545-559.
Whorton, M.D., R.M. Krauss, S. Marshall, and T.H. Milby. 1977.
Chemical induced fertility among employees in a pesticide
formulation facility. Lancet 2:1259-1261.
Whorton, M.D., T.H. Milby, R.M. Krauss, and H.A. Stubbs. 1979.
Testicular function in DBCP-exposed pesticide workers. J. Occup.
Med. 21:161-166.
Whorton, M.D., and T. H. Milby. 1980. Recovery of testicular
function among DBCP workers. J. Occup. Med. 22:177-179.
69
-------
DISCUSSION
DR. REHDER: With experimental mutagenic substances, we know
that, in many cases, after exposure there can be recovery if some
spermatogonia are left, but you showed tubules without spermatogonia.
Have you any experience with these workers after the exposure has
ceased? Is there new spermatogenesis? What do testicular histologies
look like in between those normal and those very excessive alterations?
Where does the damage start — in the spermatogonia — or is there
a spermatocytic arrest before?
DR. WHORTON: We did study men who had a range of exposures and
a range of sperm counts. It appears that the primary spermatogonia
are the target cells for direct toxic effect. Testes of lesser injury
had foci of spermatogenesis that looked normal, intermixed with foci
of large areas containing only Sertoli cells. So, there did not
appear to be an interference with the spermatogenic process once that
process started. The primary spermatogonia appeared to be the target
cell.
As far as recovery goes, several men who have not had any DBCP
exposure for more than 10 years are still azoospermic. Apparently,
the condition is permanent. Probably most of the men, at least in
this group that I have been following who are azoospermic, are probably
going to remain azoospermic because of their long exposure.
70
-------
Some individuals at Dow, in Arkansas, were azoospermic, but have
shown some evidence of recovery. However, other men at the same
place have not recovered, at this point.
Now, if you use animal or human radiation or chemotherapeutic
data as a guide, you may have to wait 3 to 6 years before deciding
that an azoospermic individual will remain so permanently. There
definitely is a dose-response relationship; what it is — how much, I
don't know.
PROF. DARDANONI: Is there any suggestion of nonoccupational
exposure of general population to this compound?
DR. WHORTON: Yes, in California, nonoccupational exposure comes
from drinking water; the route of entry is ingestion. Usual workplace
route of entry is either by inhalation or skin absorption. How much
difference this type of exposure makes is not known.
DBCP was a restricted chemical, used primarily in agriculture,
not in the home garden.
DR. DEROSIER: In what concentration does the compound appear in
drinking water?
DR. WHORTON: The highest I have seen has been about 5 to 10
ppb. Mostly it has been around 1 to 2 ppb.
DR. MILLER: Is there any evidence that the workers brought the
compound home on their clothing and affected the family?
71
-------
DR. WHORTON: At the plant with which I am familiar, they had
special workclothes, took showers, and had to change clothes to come
home. Such precautions do not occur elsewhere.
DR. MILLER: On looking at the pathology in the testes, what
kind of testicular cancer might develop? Can you predict the cell
type?
DR. WHORTON: No, I cannot. For testicular cancer, the annual
incidence rate is something like three in 100,000. Half of the
cancers are seminomas, and the other half belong to six or seven
different groups. Cancer of the testes histologically is defined
according to the predominant cell type, because most of them tend to
be teratomas.
DR. NELSON: Is there any evidence in animal studies, dominant
lethal, or multigenerational studies, of mammalian mutagenicity?
When conception can occur, is there any evidence that the father
transfers the compound to offspring?
DR. WHORTON: At this time, I am not familiar with that research.
Incidentally, there was almost no research on this compound until
1977.
DR. SUSKIND: First, would you comment on the recovery rate of
the oligospermic person? Second, is it true that one case of testicular
cancer has been reported? Third, what evidence is there that that
cancer may have resulted from heavy exposure to DBCP? Fourth, what
is the accuracy of the new film, "The Song of the Canary?"
72
-------
DR. WHORTON: For the most part, oligospermic men recover within
the first year after exposure. I am following some severely
oligospermic individuals who were not exposed for a long period but
they may not recover. Some oligospermic persons have had an
increase in fertility in the first year and a half after cessation
of exposure.
One individual in Arkansas developed testicular cancer, an
embryonal cell carcinoma. He had worked with DBCP for 2 years and
with ethylene dibromide for about 8 years. Apparently, he also had
had a history of testicular trauma as a youth. So the cause of the
cancer is unknown. As far as I know, that is the only case.
The problem has been the lack of a large enough population to
study DBCP's possible other effects. One lone factory population is
not large enough for study. And the National Institute for
Occupational Safety and Health is still in the process of setting up
a registry. As to "The Song of the Canary," a completely,
scientifically accurate film puts everybody to sleep. Overall, I
think it is a fair film.
DR. MOORE: Are researchers also studying applicators or
fieldworkers?
DR. WHORTON: OxyChem has conducted the only study on
formulators. Shell, Denver; Shell, Alabama; and Dow, Arkansas were
done'on primary manufacturers. The California study on applicators
was done by the Centers for Disease Control and the California
Department of Health. The Environmental Protection Agency study was
done by investigators from the University of South Carolina, also on
applicators in the south. The Israeli study was done on
manufacturers.
73
-------
DR. MURPHY: Can you calculate a dosage level that affects
workers?
DR. WHORTON: We have attempted to relate dosage. The problem
in the California study, the one at OxyChem, is that the company has
no worker records, so, we really have had to rely on people's
memories for determining how long they worked.
During the 1960s, DBCP was put in fertilizer, and workers in
that period say exposure levels were very high compared to what they
are now.
DR. HOOPER: Has anyone studied sperm morphology in these
workers? Is there any age distribution in these workers? Are older
men, for instance, more susceptible to azoospermia?
DR. WHORTON: Only in the fact that older men worked around the
chemical longer, but we saw no real age effect.
As far as sperm abnormalities, there is one paper in the
literature by Kapp, who used double Y bodies to look at the Arkansas
workers. There was a statistically significant increase in double Y
bodies among individuals who had been exposed to DBCP.
74
-------
Increased Lead Absorption with Anemia and Slowed
Nerve Conduction in Children Near a Lead Smelter
Philip J. Landrigan
Studies to evaluate the prevalence, sources, and health consequences
of lead absorption were conducted among children living near a primary
lead smelter. Lead levels in air, soil, and dust were highest at the
smelter and decreased with distance. Nearly 100% of 1- to 9-year-old
children living within 1.6 km had lead levels in blood ^>40 y.g/dl,
indicating increased absorption, and 22% had levels 2.80 pg/dl. The
prevalence of lead levels >40yg/dl decreased with distance; 72 km from
the smelter, the concentration was 1%. Erythrocyte protoporphyrin levels
increased with lead levels in blood: 17% of children with lead levels of
^80yg/dl were anemic. There was no overt neurologic toxicity. In 202
5- to 9-year-old children, a significant negative correlation was found
between blood lead levels and motor nerve conduction velocity (r = -0.38,
p <0.02).
Lead smelters and refineries in the United States discharge more
than 900 metric tons of particulate lead into the air each year
(National Academy of Sciences, 1972). Even though this is only a
tiny fraction of the 91,000 to 136,000 metric tons emitted annually
into the atmosphere, most of which comes from automobile emissions,
smelters are fixed sources of emission and thus serve as useful
models for studying human exposure to lead in the environment.
In April 1974, an alert pediatrician in northern Idaho saw two
siblings, 2 and 3 years old, who complained of abdominal pain, otitis
media, and diarrhea. The physician realized that the children lived
within 3 km of a large lead smelting plant, and he therefore took
x-rays of their knees. The x-rays showed characteristic lead lines
in the distal femurs and proximal tibias. The doctor then measured
lead levels in the blood of the children and found concentrations of
^Division of Surveillance, Hazard Evaluations and Field Studies, National
Institute for Occupational Safety and Health, Cincinnati, Ohio.
75
-------
68 and 89 yg/dl. At this point, he contacted public health authorities,
who launched a major epidemiologic investigation.
The study showed that lead concentrations in the atmosphere near
the smelter had increased steadily through the early 1970's. Mean
monthly lead emissions from 1955 to 1964 were 7.3 metric tons/month.
Then, from 1965 through 1973, they increased to 10 metric tons/month,
and, during 1974, rose abruptly to 32 metric tons/month following a
fire, which caused unrepaired damage to the plant's main filtration
facility. The plant ran for a year, without filtration, discharging
at least 1,800 metric tons of finely particulate lead into the
atmosphere.
Concentrations of lead in the air were highest immediately adjacent
to the smelter, and decreased with distance from the plant (Yankel
et al. , 1977). Contamination was virtually nonexistent 55 km away.
For the epidemiologic study (Landrigan e_t^ al., 1976), the valley
surrounding the smelter was divided into a series of concentric
circles. The first had a radius of 1.6 km; the second, a radius of
4 km. Additional circles were drawn with increasing distances from
the plant. Lead levels were measured in the blood of several hundred
children living within each circle. Numerous environmental samples
(air, dust, soil, food, water, and paint) were also evaluated to
measure children's exposure from these sources.
A striking pattern was found in the lead levels in the children's
blood. The closer children lived to the plant, the more lead in their
blood. Within 1.6 km, 98% of almost 200 children had lead concentrations
76
-------
higher than 40 yg/dl. In fact, 20% of the children had concentrations
higher than 80 yg/dl, the level the Centers for Disease Control (1978)
consider to signal a medical emergency. Those 30 children were, for
the most part, referred for intravenous chelation therapy with edetic
acid (EDTA).
At 4 km from the plant, 75% of the children had evidence of
increased lead absorption. The lead concentration decreased to 30%
farther from the plant, then to 10% in more distant circles. At 72
km from the smelter, in a rural district completely removed from
plant emissions, only 1% (one child) had evidence of increased lead
absorption. Interestingly, that child's father was a painter, and
the child may have been exposed to the parent's paint rather than to
the smelter.
Other analyses revealed that lead concentrations in children's
blood also correlated rather closely with the children's exposure to
lead in air; the correlation coefficient between lead in air and
blood was 0.72 in a multiple regression analysis. There was also a
fairly close correlation coefficient (0.59) between lead levels in
soil and blood (Yankel e£ al^. , 1977).
An examination of results from epidemiologic studies in relation
to various demographic factors showed a correlation between lead
concentrations in blood and socioeconomic indicators. Children of
higher socioeconomic status had lower lead concentrations in blood
than did children of lower socioeconomic status living in the area.
The children of smelter workers, as a group, generally had higher
blood concentrations of lead than did other children in the same
77
-------
geographic area. This is another indication that workers carry
noxious substances home, thereby exposing their families to secondary
contamination.
After the epidemiologic studies, pathophysiologic surveys were
conducted to determine whether or not the increased absorption of
lead had produced any harmful effects in the exposed children.
Studies using hematocrit levels as an index of anemia showed that
children in the towns near the smelter were well nourished and that
anemia was not a common problem. Only 2% of the 1,000 children
screened had hematocrit values below 33%. However, there was a
striking correlation between the distribution of anemia in relation
to lead levels in blood. Nearly 20% of the children with blood
levels greater than 80 yg/dl were anemic, and only 1% of the children
with lead levels below 80 yg/dl had evidence of anemia.
There is a fairly strong correlation between erythrocyte
protoporphyrin (EP) concentrations and levels of lead in the blood
(Granick et_ al_. , 1972). In Figure 1, the correlation coefficient
with blood lead on the horizontal axis and the EP logarithm on the
vertical is 0.79, which agrees with findings of other studies of
adults and children exposed to lead (Granick et al., 1972; Piomelli,
1980). The EP elevation is, in all likelihood, a precursor of the
anemia observed in the children with the highest lead concentrations
in blood.
Finally, peripheral nerve motor conduction velocity in relation
to lead exposure was studied. The 5- to 9-year-old children with
lead levels greater than 40 y g of lead were compared to a similar
78
-------
I I I I I I
FEP = 0.043 x (blood lead) -1-0.45 (blood Iead)-2.l4
r= 0.79
n= 1056
15
30 45 60 75
BLOOD LEAD
90 105 120
/100ml)
135 150
FIGURE 1. Free erythrocyte protoporphyrin (FEP) versus blood lead levels in children 1 to 9
years old in Shoshone County, Idaho, in 1974.
-------
group of children with lower concentrations of lead in blood. Each
child first received a very careful neurologic examination: those
found to have obvious, preexisting neurologic lesions were excluded
from further study. Some children, for example, had microcephaly;
others, anoxic birth injuries; and one had been treated for a brain
tumor.
In 202 apparently healthy children there was an interesting
correlation between the level of lead in blood and motor nerve
conduction velocity. As the lead level increased, the conduction
velocity decreased. The correlation coefficient there, in the
negative direction, was 0.39. Making the same analysis with EP
as the independent variable provided essentially the same result
with the same strength of association.
In summary, Increased absorption of lead, defined as a lead
concentration in blood of 40 yg/dl or higher, was highly prevalent
among children who lived near the Idaho smelter. The finding that
98% of the children within 1.6 km of the smelter had increased lead
concentrations in blood is the highest prevalence that has been
recorded. Finally, there were close correlations between blood and
environmental lead levels and between lead absorption and anemia, EP
elevation, and the decrease in nerve conduction velocity.
80
-------
REFERENCES
Center for Disease Control. 1978. Preventing Lead Poisoning in
Young Children. Center for Disease Control, Atlanta.
Granick, S. , S. Sassa, and J. L. Granick. 1972. Assays for
porphyrins, delta aminolevulinic acid dehydratase, and
prophyrinogen synthetase in microliter samples of whole blood:
Applications to metabolic defects involving the heme pathway.
Proc. Natl. Acad. Sci. USA 69:2381.
Landrigan, P. J. , E. L. Baker, R. G. Feldman, ejt a^. 1976.
Increased lead absorption with anemia and slowed nerve conduction
in children near a lead smelter. J. Pediatr. 89:904-910.
National Academy of Sciences. 1972. Lead: Airborne Lead in
Perspective. National Academy of Sciences, Washington, D.C.
330 pp.
Piomelli, S. 1980. The effects of low-level lead exposure on
heme metabolism. Pp. 67-74 in H. Needleman, ed. Low Level
Lead Exposure: The Clinical Implications of Current Research.
Raven, New York.
Yankel, A. J., I. H. von Lindern, and S. D. Walter. 1977.
The Silver Valley lead study: The relationship between
childhood blood lead levels and environmental exposure.
J. Air Pollut. Control Assoc. 27:763-767.
81
-------
DISCUSSION
DR. MILLER: How do the health effects you observed compare
with the dogma that encephalopathy is the only effect of lead in
children, and that lesser effects are not seen?
DR. LANDRIGAN: That point of view prevailed during the 1940's
and perhaps into the 1950's. However, as more sensitive hematologic
and diagnostic techniques have been developed, it has become evident
that there is a wide spectrum of damage induced by lead in children,
ranging from minimal slowing of nerve conduction velocity and minimal
learning deficits to encephalopathy (Needleman and Landrigan, 1981).
This continuum of toxicity relates to increasing exposure to lead.
DR. MILLER: Do you wish to comment about using children's
teeth to measure lead?
DR. LANDRIGAN: Studying teeth has turned out to be a very
exciting way to measure a child's exposure to lead. Whereas the
lead level in blood reflects only very recent lead exposure, perhaps
only that during the previous 3 days or in the past week, and the EP
level reflects exposure over the past 2 or 3 months, the lead level
in a deciduous tooth presents a cumulative exposure record of 7 or
8 years' duration. This technique enables us to correlate a child's
longitudinal neurologic and psychomotor development with past lead
exposure, perhaps even to exposure before birth. The recent studies
of Needleman et al. in Boston showed some very good correlations
between children's lead levels in teeth and the occurrence in those
82
-------
children of various subtle learning disabilities and psychomotor
problems (Needleman e± al., 1979).
DR. MILLER: Can you comment on the famous Glasgow lead study
based on blood specimens collected for PKU (phenylketonurla) screening?
The samples were filed away for years, then retrieved and analyzed
for lead to show the blood concentrations of mentally retarded
children as compared with those of normal children at birth (Seattle
et al., 1975).
DR. LANDRIGAN: The retarded children showed higher blood lead
levels than did the normal children. The presumed source of their
exposure was the very heavily contaminated acidic water of Glasgow,
which was stored in lead cisterns above the children's houses.
DR. GOLDBERG: Were there any followup studies on the children
in Idaho to look at the long-term effects of their lead exposure?
DR. LANDRIGAN: No followup studies have been conducted in
Idaho. However, followup study of children with lead exposure
near a smelter in El Paso, Tex., showed that the level of lead
in children's blood decreased considerably over a period of 5 years
(Landrigan et^ al., 1975; Morse et al. , 1979).
PROF. DARDANONI: Since the Idaho smelter lies in a valley,
the prevailing wind may diffuse the lead in one direction. Is
there any possibility of determining whether "hot spots" — lead
concentrations in the soil — exist and correlating these findings
to different patterns of human exposure?
83
-------
DR. LANDRIGAN: Yes. We probably did not do that as thoroughly
as we would like. The smelter could not be located in a more
unfavorable spot. The valley floor is approximately 1,000 meters
above the sea and the mountains on either side rise to nearly 2,000
meters, so the smelter sits in the base of a very narrow "V."
Furthermore, the valley runs from west to east; thus, the prevailing
wind from the Pacific Ocean carries the lead predominantly to the
east. As we looked at lead concentrations in the environment and
children, we could clearly tell that the pattern of excess contamination
extended much farther to the east of the smelter than to the west.
SPEAKER (UNIDENTIFIED): Have you estimated the heavy metal exposures
of populations living within several kilometers of the 40 or so lead
and copper smelters in the United States.
DR. LANDRIGAN: After we studies lead absorption near the plants
in El Paso, Texas (Landrigan e£ aJ_. , 1975), and Kellogg, Idaho
(Landrigan et al., 1976), we conducted a nationwide survey of heavy
metal absorption in children living near 21 other ore smelters in the
United States (Baker _e£ al., 1977). We measured concentrations of
lead, cadmium, arsenic, and copper in children living with a 3.2 km
radius of each smelter. We also took samples of their blood and urine.
We found that the Kellogg smelter was the most heavily contaminated
with lead. The El Paso smelter was second worst. In terms of lead
absorption, no other smelters, except for one in Herculaneum, Missouri,
caused even a slight problem. We found a problem of arsenic absorption
around copper smelters in Tacoma, Washington (Milham and Strong, 1974),
and Anacona, Montana. Cadmium absorption was a problem in children
living around a zinc smelter in Bartlesville, Oklahoma.
84
-------
REFERENCES FOR DISCUSSION
Baker, E. L., C. G. Hayes, P. J. Landrigan, J. L. Handke, R. T.
Leger, W. J. Housworth, and J. M. Harrington. 1977. A nation-
wide survey of heavy metal absorption of children living near
primary copper, lead, and zinc smelters. Amer. J. Epidemiol.
106:261-273.
Beattie, A. D., M. R. Moore, A. Goldberg, M. J. W. Finlayson,
J. R. Graham, E. M. Mackie, J. C. Main, D. A. McLaren, R. W.
Murdoch, and G. T. Stewart. 1975. Role of low level lead
exposure in the aetiology of mental retardation. Lancet 1:7907-7910.
Landrigan, P. J., S. H. Gehlbach, B. F. Rosenblum, J. M. Shoults,
R. M. Candelaria, W. F. Barthel, J. A. Liddle, A. L. Smrek,
N. W. Staehling, and J. F. Sanders. 1975. Epidemic lead
absorption near an ore smelter: The role of particulate lead.
N. Engl. J. Med. 292:123.
Landrigan, P. J. , E. L. Baker, R. G. Feldman et_ al_. 1976.
Increased lead absorption with anemia and slowed nerve conduction
in children near a lead smelter. J. Pediatr. 89:904-910.
Milham, S., and T. Strong. 1974. Human arsenic exposure in
relation to a copper smelter. Environ. Res. 7:176-182.
Morse, D. L., P. J. Landrigan, B. F. Rosenblum, J. S. Hubert, and
J. Housworth. 1979. El Paso revisited: Epidemiologic follow-up
of an environmental lead problem. J. Amer. Assoc. 242:739-741.
Needleman, H. L., and P- J. Landrigan. 1981. The health effects
of low level exposure to lead. Am. Rev. Public Health 2:277-298.
Needeleman, H. L., C. G. Gunnoe, A. Leviton, R. R. Reed, H. Peresie,
C. Maher, and P. Barrett. 1979. Deficits in psychologic and
classroom performance of children with elevated dentine lead
levels. N. Engl. Md. 300:584-695.
85
-------
Methyl Mercury (Japan)
Robert W. Miller1
A severe, sometimes fatal, neurologic epidemic occurred in the
early 1950's around Minamata Bay in Kyushu, Japan. Humans, fish, birds,
and cats were affected, but the cause was not determined for 5 years,
when it was traced to industrial contamination of the bay with methyl
mercury. Congenital Minamata disease (cerebral palsy) was recognized
even later. A smaller, but similar epidemic occurred a few years later
in Niigata, Japan. Methyl mercury poisoning from food has come mainly
from using the chemical as a fungicide for grain meant to be planted
and not eaten. The later effects of methyl mercury exposure after
areawide contamination have not yet been well described. M. Harada of
Kumamoto University found a unique way to identify previous methyl
mercury exposure by chemical analysis of dried umbilical cords kept for
decades by Japanese families. He was able to identify high levels of
methyl mercury, not only at the time of the epidemic known in the 1950's,
but also of one in the late 1930's, which had been unexplained. This
experience illustrates the importance of preserving specimens now for
future studies of human diseases thought to be environmentally induced.
Mercury, dumped by a plastics factory into Minamata Bay, Japan,
in the 1950's, was converted into organic mercury that was ingested by
fish, thus entering the food chain. Animals, both domestic and wild,
were affected by the mercury contamination, as were the residents of
the area. A privately printed book, Minamata Disease, written in English
(Study Group of Minamata Disease, 1968), said that because of neurologic
disability, fish swam upside down, birds could not fly, and cats went
mad. The fishermen sold the best of their catch; they and their families
ate the worst. Thus, their own families were particularly affected by
the contaminated fish.
Clinical Epidemiology Branch, National Cancer Institute, Bethesda, Md.
86
-------
The severe neurological effects were Immediately recognizable in
children and adults, but the transplacental effects were not described
until 7 years after the epidemic (Matsumoto e* al., 1965). Harada (1976)
included a 1962 photograph (probably a composite) of 13 children with severe
cerebral palsy, attributed to their intrauterine exposure to methyl mercury.
Twenty-three cases were described in the first comprehensive reports; the
expected number was 1 or 2. The most recent publication (Harada, 1978)
reports 40 cases.
Japanese families, by custom, save dried umbilical stumps, and Harada
was able to obtain these from a family with six children born between 1953
and 1966, the time in which the bay was contaminated with methyl mercury.
He also obtained dried umbilical stumps from a family with six children
born between 1927 and 1939. Unexpectedly, he found a higher methyl mercury
content in the cords of two children in this family born during the late
1930's. Apparently, an earlier epidemic had occurred, which Nishigaki and
Harada (1975) related to the production of vinyl plastics at the same factory
that was implicated in the 1950's outbreak. People living near the factory
recalled an unexplained outbreak of psychiatric illness at the time. The
factory did not produce plastic during the war, but resumed production
afterward, at which time the Minamata disease epidemic occurred.
Minamata disease occurred not only in Japan, but also in Guatemala;
Alamogordo, N. Mex.; and Iraq, among other places. In these instances, methyl
mercury was used as a fungicide on grain meant to be planted, but instead
was fed to pigs or used for baking. In Iraq, some 6,000 people (including
six transplacental cases) were reported to be poisoned (Bakir et al. , 1973).
87
-------
Thus, methyl mercury traceable to water, grain, fish, and animals has been
implicated in neurologic disorders observed in adults and newborns.
What we have seen here is a neurologic disorder traceable to contami-
nation of water, grain, fish, and animals. It produces a transplacental
effect. We have learned that preserved specimens of human tissue are
uniquely available in Japan, but could be obtained in this country or
elsewhere. This experience raises the question of whether to save such
specimens as teeth, a spot of dried blood, or part of the placenta as a
means for retracing our past for environmental contamination.
88
-------
REFERENCES
Bakir, F., S.F. Damluji, L. Amin-Zakl, M. Murtadha, A. Khalidi, N.Y. Al-Rawi,
S. Tikriti, H.I. Dhahir, T.W. Clarkson, J.C. Smith, and R.A.
Doherty. 1973. Methyl mercury poisoning in Iraq. Science 181:230-
241.
Harada, M. 1976. Intrauterine poisoning: Clinical and epidemiological
studies and significance of the problem. Bull. Inst. Constitutional
Med. (Suppl.), Kumamoto University 25:1-60.
Harada, M. 1978. Congenital Minamata disease: Intrauterine methyl
mercury poisoning. Teratology 18:285-288.
Matsumoto, H., G. Koya, and T. Takeuchi. 1965. Fetal Minamata disease:
A neuropathological study of two cases of intrauterine intoxication
by a methyl mercury compound. J. Neuropath. Exp. Neurol. 24:563-
574.
Nishigaki, S., and M. Harada. 1975. Methyl mercury and selenium in
umbilical cord of inhabitants of the Minamata area. Nature 258:324-
325.
Study Group of Minamata Disease. 1968. P. 338 in Minamata Disease.
Kumamoto University, Japan.
89
-------
Chlorinated Hydrocarbons
David Axelrod1
Love Canal is a 6.5-hectare landfill located in the southeast corner
of the city of Niagara Falls, N.Y. The Canal takes its name from a water-
way that developer William T. Love began to dig there at the end of the last
century. From 1933 to 1953, the Hooker Chemical Company used the Canal to
dispose of wastes from its Niagara Falls manufacturing complex. After Hooker
sold the landfill in 1950 to the local board of education, an elementary
school was built and the surrounding area developed into a neighborhood of
working class, one-family homes. In the 1970's it was suspected that pesti-
cide residues had leached from the landfill into the nearby Niagara River
and thence into Lake Ontario. State and Federal environmental agencies took
air and soil samples at the landfill in 1976-77. On the basis of their
findings and homeowner complaints about chemical odors and illnesses, the
State Health Department investigated and declared a public health emergency
in 1978. The Governor decided that 239 families should be permanently
relocated away from the landfill site. Hooker has acknowledged depositing
almost 20,000 metric tons of chemical wastes in the Canal from 1942 to 1952.
More than 200 organic chemical compounds have been identified in soil, sedi-
ment, and sludge samples taken from the landfill. Many of these compounds
are toxic to humans. Epidemiologic and environmental testing continues in a
widening circle around the landfill.
The disposal of hazardous industrial wastes is a problem of immense
complexity with both immediate and long-term health implications for society.
One prime example of the problem exists at the Love Canal landfill in Niagara
Falls, N.Y. (Figure 1).
When I first visited the site in 1978, it was barren and foreboding.
Only a single scraggly tree stood on the site. The ground underneath was
gummy, and chemical ooze clung to my shoes. The sickening stench of chemicals
and oddly colored pools of water were everywhere. When I poked a stick into
the soil, a viscous, oily fluid came to the surface. Pools of water stood
on the playing fields of the elementary school, which bordered the landfill.
Near the site's northern edge, mounds of dry, white powder dotted the surface.
LNew York State Department of Health, Albany, N.Y.
90
-------
LOVE CANAL
SITE
FIGURE 1. The Love Canal site in Niagara Falls, N.Y.
91
-------
Someone said this was pure lindane. Fresh motorbike tracks
crisscrossed the area and I had visions of toxic dust clouds rising
as cyclists raced. Most people know Niagara Falls as a honeymoon
resort city of great natural beauty. The landfill is a far cry from
the scenic beauty that appears on postcards.
The principal business of the city is the manufacture of chemicals
and allied products. Niagara County, in which Niagara Falls lies,
harbors an estimated 100 chemical dump sites. According to a survey
made by the State Department of Commerce, nine major chemical companies,
employing 5,200 people, are located in Niagara County.
The largest of the enterprises, with some 2,000 employees, is the
Hooker Chemical Company, which used Love Canal. Love Canal is a 6.5
hectare site in the midst of a residential neighborhood in the southeast
sector of Niagara Falls. The site was originally excavated at the
end of the 19th century as part of a proposed navigable waterway to
link the upper and lower sections of the Niagara River. The river
flows into Lake Ontario, the easternmost of the Great Lakes.
Beginning about 1930, Hooker Chemical Company used the abandoned
canal site for disposal of chemical wastes, including chlorinated
hydrocarbons, processed sludge, and fly ash. The company's records
indicate that during a 10-year period ending in 1952, it deposited
almost 20,000 metric tons of chemical wastes in the Love Canal area.
Whether the canal flowed to the Niagara River is not known. A 1938
aerial view of the area shows no obvious link with the river, although
there is a scar in the earth. The landfill was closed in 1953 and
Hooker sold the property to the city board of education. Soon thereafter
92
-------
residential development of the surrounding area began, and continued
until 1974. In the mid-19501s, a public elementary school was built
on land adjacent to the site.
By 1966, the entire landfill had been covered over, and homes
were built along the canal. In 1978, 99 houses abutted directly on
the waterway, and the first two rings of housing developments around
them contained about 230 houses. Approximately, 280 more homes lay
beyond these.
The hazard posed by the Love Canal landfill to the residents was
noticed by the State Department of Health in 1978. The U.S. Environ-
mental Protection Agency (EPA) and the State Department of Environmental
Conservation, had been concerned about potential contamination of the
Niagara River and, ultimately, of Lake Ontario. During the course of
their investigation, the agencies made a number of soil analyses from
properties near the landfill. Afterwards, the Department of Environ-
mental Conservation asked the health department for additional laboratory
analyses to determine whether chemicals in the landfill had reached
the Niagara River and perhaps accounted for the presence of pesticides
in a Lake Ontario fishery.
At that point it was very clear to the Department of Health that
the key question was not whether Lake Ontario was contaminated, but
whether or not there was a public health problem from the landfill.
When the department realized that there was, in fact, a considerable
hazard, an order was issued to fence the site and to cover over
exposed chemical deposits as a temporary measure until additional
93
-------
studies could be conducted. Infrared photographs showed areas of
"coolness," which meant chemical spoliation, and a complete lack of
vegetation, even up to the school playground itself. The pesticide
lindane, one of the chemicals present, had surfaced and then crumbled
off onto the playground and into the street. Water seeped into some
of the homes in this area, and also onto the land surface.
Children played on the landfill site and used it as a shortcut
to school, even though their shoes, according to anecdote, later
fell apart. One youngster reported that the children also used to
build treehouses at the site (while the trees were still there),
carrying lumber from adjacent areas in the community.
In the spring of 1978, following a period of heavier than normal
rain- and snowfalls in the Niagara area, there were many complaints
about chemical odors from Love Canal. People living there had, in
fact, complained of odors previously, but the odors became much more
severe that spring.
EPA studies had identified some 26 chemicals in the area.
Subsequently, Department of Health studies showed that about 200
chemicals had been deposited in the landfill. Among these compounds
were benzene, toluene, benzoic acid, isomers of dioxin, lindane,
trichlorethylene, dibromoethane, benzaldehydes--in effect, most of
the chlorinated benzenes and toluene present around a chemical
manufacturing plant.
The department is continuing environmental sampling at the site
to determine if the levels and types of chemical exposure can be
correlated to epidemiologic findings. At present, the department
94
-------
has taken some 3,000 soil samples, from both sides of the canal, for
analysis for a variety of different chemicals. The objectives of
the environmental and toxicologic studies have been fourfold: (1) to
identify the chemical agent or agents responsible for adverse health
effects reported by Love Canal residents; (2) to ascertain, if possible,
the route of entry of such agents into the human body; (3) to determine
the extent and means of chemical migration outward from the landfill
proper; and (4) to evaluate the efficacy of the remedial construction
taken at the landfill.
From early 1978 and well into 1979, the health department dispatched
teams of epidemiologic investigators to the site to determine the
extent of adverse health effects. In addition, a nationwide, toll-free
telephone hotline was established and publicized through major news
media to contact persons who had moved away from the area. About
700 former residents have been reached through this hotline. Within
a few weeks in mid-1978, the department had also taken blood samples
from some 2,800 people who resided, or had previously resided, in the
Love Canal area.
The teams administered a 29-page questionnaire to all residents
within a 4-block area of the landfill to obtain detailed information
on present and past health status, as well as on social, occupational,
and residential histories. The questionnaire covered more than 150
physical complaints or symptoms, ranging from excessive weight loss
to headaches and dryness of skin. The Department of Health thus
collected a massive amount of health and morbidity data about the
people in the area.
95
-------
Initially, the department selected four health indicators as
measures of toxic exposure: miscarriages, birth defects, low birth
weight, and liver dysfunction. Adverse pregnancy outcome was selected
because the prepartum period is generally considered to be especially
susceptible to chemical insult; liver function was studied because
many of the chemical compounds identified are known to have hepatic
toxicity.
Complete blood counts were performed on 4,386 samples taken from
an ultimate study population of 3,919 Love Canal residents. These
samples were analyzed for 26 different hematologic and enzymatic
parameters.
The department also made an effort to discover the hydrogeology
of the area surrounding the landfill. It determined that several
"swales" or ditches may have been present in the area. Aerial maps
indicate that if water migrated through these ditches over time, the
migration pattern apparently has since been greatly disturbed by
construction, by the development of storm sewers, and by a variety
of other factors. By 1966, aerial overflights indicated only one
remaining swale.
Throughout the past 2 years, the department has been trying to
discern the normality or abnormality of the epidemiologic evidence.
Constantly, data on canal residents have had to be compared with
information on population groups that are jarringly inconsistent
socially, educationally, economically, and vocationally.
This lack of a control population is the single most important
lesson learned from Love Canal. What is desperately needed is a
96
-------
reservoir of scientific information concerning the risks of long-term,
low-dose exposure to man-made chemicals. Causal relationships must
be established with precision in the face of such confounding health
influences as tobacco, occupational exposure, and even genetic factors,
It is not inconceivable that significant ill health effects from
exposure to the chemicals at Love Canal may not manifest themselves
for a generation or longer.
Today, Love Canal has a clay cap, and drainage ditches have been
placed to carry the canal's liquids to a facility where they are
treated and discharged into the Niagara River.
In summary, much more information is needed to evaluate the
toxicity of low-level exposure to chemicals over long periods of
time. Furthermore, control populations must be established. Finally,
an adequate method is needed to enable a comparison of the kinds of
experiences identified at the canal with those in other parts of the
country.
97
-------
DISCUSSION
DR. MILLER: About how many chemicals are really involved at
Love Canal?
DR. AXELROD: Hundreds, literally hundreds. A whole variety of
chlorinated benzenes, toluenes, phenols, and trichlorophenol in
large concentrations suggested that dioxin was also present. It was
eventually identified, leaching out from the canal through storm
sewers and into the stream-bed in the vicinity.
DR. MILLER: Is chemical dumping limited to Niagara Falls?
DR. AXELROD: No. The Love Canal situation has led to an aware-
ness of a large number of hazardous waste sites throughout the United
States. The problem of chemical dumps is a major one for the Nation,
and the potential danger needs to be examined further.
SPEAKER (UNIDENTIFIED): How many dumps are there in the United
States?
DR. AXELROD: New York State has about 550. A recent EPA report
stated that there were thousands.
SPEAKER (UNIDENTIFIED): At Love Canal, was there an odor of organic
solvents on wet or rainy days prior to the capping?
DR. AXELROD: On the occasions I visited, there was clearly an
odor. The question is whether or not the odor at the site is signifi-
cantly different from that which exists in adjacent areas. The
residents are attuned to the odor and may be able to make that kind
of distinction; I could not. Only when I was walking in the area
and putting my feet directly into the oozing chemicals could I clearly
detect an odor emanating directly from the surface.
98
-------
DR. LESTER: What comparison group have you been able to find to
evaluate some of the health problems?
DR. AXELROD: Generally, the comparison group has come from the
area north of Colvin, north of Love Canal.
DR. LESTER: Thus far, how much do you know about chemical migration
from the canal area?
DR. AXELROD: We are in the process of analyzing some 3,000 soil
samples for evidence of chemical migration. Each sample has to be
analyzed for the presence of individual chemicals.
DR. SPENCER: Have you looked for or identified any specific
neurologic effects in the population?
DR. AXELROD: We have not examined the population for specific
neurologic defects, but the original questionnaire did include a
number of items relating to certain identifiable neurologic diseases.
DR. MILLER: Would any purpose be served by studying urine samples
to see if they were positive in the Ames test?
DR. AXELROD: We discussed doing Ames tests, but they didn't seem
to be the most efficacious method of determining the kinds of exposures.
We will continue to look at urine sampling as a way to determine other
mechanisms for adverse health effects.
DR. MILLER: Couldn't there be an aggregate effect from the chemicals
that might not be determined by identifying Individual chemicals?
DR. AXELROD: We have discussed this possibility with various
individuals involved with the Ames test, and there are conflicting
points of view.
99
-------
Cohort Study of Michigan Residents Exposed to Polybrominated
Biphenyls; Epidemiologic and Immunologic Findings
Philip J. Landrigan*
Polybrominated biphenyls (PBB's) were dispersed widely in
Michigan by a 1973 accident in which PBB's were introduced into
cattle feed. Thousands of people were exposed principally from
ingestion of contaminated dairy food products. To determine whether
PBB exposure has or will cause acute or chronic illness, a prospective
cohort study of 4,545 persons has been undertaken. Three exposure
groups were analyzed: all persons living on PBB-quarantined farms;
persons who had received food directly from such farms; and workers
(and their families) engaged in PBB manufacture. Enrollment rates
were 95.6%, 95.1%, and 78.0% for each group. Another 725 persons
with low-level PBB exposure were also enrolled. All persons were
queried about 17 symptons and conditions possibly related to PBB
exposure. Venous blood was drawn from 3,639 persons and analyzed
for PBB by gas chromatography. Mean serum PBB levels were 26.9 ppb
in quarantined farm families, 17.1 ppb in recipients, 43.0 ppb in
workers, and 3.4 ppb in the low exposure groups. No associations
were found between serum PBB levels and symptom-prevalence rates.
To evaluate peripheral lymphocyte function, T and B cell quantitation
and in vitro responses to three nonspecific mitogens were studied in
the 34 persons with the highest PBB levels (mean, 787 ppb), and in
56 with low values (mean, 2.8 ppb). No statistically significant
differences were noted in lymphocyte number or function.
In a polybrominated biphenyl (PBB) molecule, there are two
6-carbon rings connected by a carbon-carbon bond:
X
Bromine is substituted on the rings at any one or more of the 10
positions, thus allowing for an enormous number of possible isomers.
Division of Surveillance Hazard Evaluations and Field Studies,
National Institute for Occupational Safety and Health, Cincinnati,
Ohio.
100
-------
In the PBB that was the culprit in Michigan, about 80% of the
material was hexabromobiphenyl, but other brominated biphenyls,
carrying anywhere from one to seven or eight Bromine atoms per
molecule, were also detected in the commercial mixture (Matthews
et al., 1977).
In the summer of 1973, a small chemical company in St. Louis,
®
Mich, produced two commercial mixes. Nutrimaster contained magnesium
®
oxide and was used to stimulate lactation in dairy cattle. Firemaster
was used as a flame retardant, principally in the plastics industry.
On one occasion, the company ran out of Nutrimaster bags and packaged
several hundred pounds of Firemaster into Nutrimaster bags. The
company claims to have written the word Firemaster across the top of
the bags, but the written label appeared precisely where a farmer
would rip the bag to pour it into cattle feed.
It is estimated that probably several hundred pounds of
polybrominated biphenyl were introduced into cattle feed in Michigan
that summer. The chemical spread rapidly through the distribution
chain and affected thousands and thousands of cattle, which were
contaminated and had to be quarantined and destroyed. Other cattle
died. The PBB caused devastating illness. Sick cows and calves
became emaciated and had trouble walking. A variety of dermatologic
disorders appeared, including a bizarre overgrowth of the hoof, which
may be the closest manifestation cattle have to chloracne (Jackson
and Halbert, 1974).
101
-------
Many thousands of people were also exposed to PBB by several
routes. Workers in the chemical factory that produced the material
were exposed directly, probably both by inhalation and inadvertently
by ingestion of particulate material. People who lived on contaminated
farms were also heavily exposed. The farmers very likely handled feed
directly, and the entire farm family ate and drank contaminated meat
and milk. Finally, the consumers, the people who bought contaminated
dairy products, were also exposed.
To evaluate PBB exposure and health effects in the State of
Michigan, the Centers for Disease Control (CDC), in collaboration
with the Michigan Department of Public Health, the National
Institute for Environmental Health Sciences, the National Cancer
Institute, the Food and Drug Administration, and the Environmental
Protection Agency (EPA), mounted an enormous cohort evaluation of
some 4,000 Michigan residents beginning in 1976 (Landrigan et al.,
1979). Basically, four categories of Michigan residents were
enrolled in the study. First, there were the farm families who
had been contaminated. Of 2,200 people contacted, 95% agreed to
participate in the study. Second, a number of people were farm
product recipients. These people lived next door to or down the
road from a contaminated farm. They received their milk or meat
directly from a contaminated farm, and there was no dilution of the
contaminated product through normal commercial channels.
102
-------
Third, there was a small group of workers at the chemical plant.
The participation of this group and their families was not quite so
good. Finally, there was a small group of persons who had participated
in a pilot study conducted the previous year by the Michigan Department
of Public Health (Humphrey and Hayner, 1975). There were also approxi-
mately 600 people, who essentially had no documented exposure to PBB,
but who were particularly concerned and wanted to be included in the
study.
Everybody enrolled was visited between 1976 and 1977 by trained
field investigators, who asked in detail about 17 symptoms that might
conceivably be related to PBB exposure (Kuratsune et_ al_. , 1972).
The list of symptoms included those reported by the subjects and a
few related to effects observed in cattle. Fatigue and pains in the
joints were the two most prevalent symptoms.
The distribution of serum concentrations of PBB in the various
groups was an interesting finding. The range of concentration was
very broad in almost every group, yet the medians were very low.
Hence, it is evident that there is a log normal distribution of PBB
in the serum of these population subgroups.
The highest PBB concentrations in serum were observed in the
chemical workers and their families. In fact, looking at just the
workers apart from their families, the serum level was even higher.
Nonetheless, the family members also had levels much increased above
those found in the general population, suggesting a pattern of secondary
contamination as workers transport the substance from the workplace
home to their families.
103
-------
After the workers, the next highest serum PBB levels were found
in persons living on contaminated farms. Next to this group were
the consumers, the recipients of contaminated products. The volunteer
participants had almost uniformly low levels.
The study related the serum PBB concentrations to the prevalence
of symptoms in the 3,300 persons for whom complete data sets were
obtained. The cohort was divided into seven groups, depending on
the range of serum PBB concentrations; for example, no concentration,
1 ppb, 2 to 3 ppb, and so on. Of the persons who had no PBB concen-
tration, 46% reported fatigue; of those with 1 ppb, 40% had fatigue,
and of those with more than 1 ppb, 27% had fatigue. This finding is
essentially an inverse dose-response relationship. A more or less
similar pattern was observed for each of the other symptoms.
There are some rather striking differences among groups. For
example, the chemical workers had a much lower symptom-prevalence
rate than did farmworkers. Prevalence rates within each study group
were looked at separately, but the same pattern remained; that is,
the highest prevalence of symptoms was observed in the persons with
the lowest serum PBB levels.
Thus, there is unequivocal evidence of statewide PBB contamina-
tion. That finding is corroborated by the results of other studies
(Brilliant _et_ _al. , 1978; Selikoff and Anderson, 1979). The evidence
of the exposure was observed first in chemical workers, then in
farmers, then in consumers, as the contamination spread broadly
through the food chain. However, nothing was observed to correlate
104
-------
the dose-response relationship to serum PBB levels or to the occurrence
of symptoms. There are probably people in Michigan who have symptoms
caused by PBB exposure, but those symptoms may never become evident
through the epidemiologic technique because the real organic symptoms
are masked by nonorganlc symptoms.
Analyses of PBB in serum samples taken from several hundred
people in 1974 were compared with results from second samples taken
in 1977. There was virtually no change in the serum PBB concentrations
during that 3-year period. The compound is extremely stable in the
body. It is very fat soluble and highly concentrated in fat. A
person heavily contaminated with PBB carries a substantial portion
of the ingested dose for a lifetime.
In this study, there were also 221 simultaneous paired specimens
of blood and adipose tissue. The concentration gradient from adipose
tissue to blood was 360 to 1. Dr. Irving Selikoff and his colleagues
at Mt. Sinai have also looked at paired adipose and serum tissues of
several hundred people. They, too reported an adipose-to-serum PBB
gradient of about 350 or 360 to 1 (Anderson et_ al_. , 1978). The
two sets of data are thus in close agreement and confirm the hypothesis
that PBB is highly partitioned in fat.
PBB is also concentrated in the fat portion of breast milk. A survey by
the Michigan Department of Public Health showed that 96% of breast milk samples
obtained in Michigan's lower peninsula were contaminated with PBB. Analyses of
breast milk samples from women in the upper peninsula showed that 43% were
contaminated, providing evidence of the geographic transfer of the substance
(Brilliant et al., 1978).
105
-------
In summary, although no definitive evidence of acute or subacute
disease has come to light so far in the Michigan population as the
result of PBB exposure, considerable concern remains about long-term,
possible delayed consequences. Dr. Renate Kimbrough, a toxicologist
at the Centers for Disease Control, has developed evidence indicating
that PBB causes adenocarcinoma of the liver in rats (Kimbrough et
al. , 1978). Because Michigan residents will continue to suffer
internal biological exposure to PBB for the next several decades,
they may have a greater risk of developing cancer. The health agencies
that established the Michigan study cohort plan to follow it for the
next several decades.
106
-------
REFERENCES
Anderson, H. A., R. Lilis, I. J. Selikoff, K. D. Rosenman, J. A.
Valclukas, and S. Freedman. 1978. Unanticipated prevalence of
symptoms among dairy farmers in Michigan and Wisconsin. Environ.
Health Perspect. 23:217-226.
Brilliant, L. B., G. Van Amburg, J. Isbister, H. Humphrey,
K. Wilcox, J. Eyster, A. W. Bloomer, and H. Price. 1978.
Breast milk monitoring to measure Michigan's contamination
with polybrominated biphenyls. Lancet 2:643-646.
Humphrey, H. E. B., and N. S. Hayner. 1975. Polybrominated
biphenyls: An agricultural incident and its consequences,
an epidemiological investigation of human exposure. Presented
at the Ninth Annual Conference on Trace Substances in Environ-
mental Health, Columbia, Mo., June 1975.
Jackson, T. F., and F. L. Halbert. 1974. A toxic syndrome
associated with the feeding of polybrominated biphenyl-
contaminated protein concentrate to dairy cattle.
J. Am. Vet. Med. Assoc. 165:427-439.
Kimbrough, R. D., V. W. Burse, and J. A. Liddle. 1978. Persistent
liver lesions in rats after a single oral dose of polybrominated
biphenyls (Fire Master FF-1) and concomitant PBB tissue levels.
Environ. Health Perspect. 23:265-273.
Kuratsune, M., T. Yoshimura, J. Matsuzaka, and A. Yamaguchi. 1972.
Epidemiologic study on Yusho, a poisoning caused by ingestion of
rice oil contaminated with a commercial brand of polychlorinated
biphenyls. Environ. Health Perspect. Experimental Issue No. 1,
April 1972:119-128.
Landrigan, P- J., D. R. Wilcox, Jr., J. Silva, Jr., H. E. B.
Humphrey, C. Kauffman, and C. W. Heath, Jr. 1979. Cohort
study of Michigan residents exposed to polybrominated biphenyls:
Epidemiologic and immunologic findings. Ann. NY Acad. Sci.
320:284-294.
Matthews, H. B., S. Kato, N. M. Morales, and D. B. Tuey. 1977-
Distribution and excretion of 2,4,5,2',4',5',-hexabromobiphenyl,
the major component of Firemaster BP-6. J. Toxlcol. Environ.
Health 3:599-605.
Selikoff, I. J., and H. A. Anderson. 1979. A survey of the
general population of Michigan for health effects of PBB exposure.
Contract Final Report. Michigan Department of Public Health,
Lansing, Mich.
107
-------
DISCUSSION
DR. MILLER: One of the novel aspects of this experience was
using breast milk to analyze the body burden of a chemical that is
deposited in fat and not easily removed from it.
DR. LANDRIGAN: Yes, a very strong case can be made for using
human breast milk as a device for screening a population's exposure
to fat-soluble compounds, whether they are halogenated organics, such
as PBB's or polychlorinated biphenyls (Landrigan, 1980), or fat-soluble
pesticides (Savage, 1977). Milk is obviously an easy tissue to obtain.
It does not require venipuncture, and it is easily transported. Also,
because milk fat contains highly concentrated fat-soluble compounds,
relatively simple analyses will often suffice for observing trends in
population exposure.
DR. MILLER: Are PBB's metabolic activators?
DR. LANDRIGAN: Yes, very much so. They activate microsomal
enzymes in the liver (Matthews et al. , 1978).
DR. MILLER: Would that change a patient's blood level of a
chemotherapeutic agent for cancer or of some other drug?
DR. LANDRIGAN: Barbiturate metabolism might, for example, be
altered.
DR. NELSON: Can you say anything about the immunologic studies
done in relationship to the PBB studies by the Mt. Sinai group? Are
there special problems in maintaining surveillance of identified
exposed groups for 30 to 35 years?
108
-------
DR. LANDRIGAN: Two immunologic studies have been conducted in
Michigan on persons exposed to PBB's. The first was made by Bekesi
et^ a!L. (1978) from Mt. Sinai Hospital in New York, who examined immune
functions in Michigan Residents and compared them with those from people
in Wisconsin with no exposure to PBB. The data are curious. Every one
of the Michigan residents showed some form, although not the same form,
of lymphocytic malfunction. In contrast, virtually none of the Wisconsin
residents showed any lymphocytic malfunction. Among the Michigan
residents, there was no evidence of a dose-response relationship to
serum PBB concentration. Many instances of dysfunction occurred in
persons with virtually nondetectable serum PBB concentrations.
The Centers for Disease Control undertook a second study with
the University of Michigan and examined 36 persons with very high
exposure to PBB (Silva et^ al_. , 1979). Their mean serum PBB concentration
was 787 ppb, a concentration much higher than that in the group studied
by Bekesi et al. (1978). CDC also examined a control group of 57
persons with mean serum PBB concentrations of 2.8 ppb. No differences
between groups in leukocyte count, in the number of T cells or B cells,
or in the in vitro responses of lymphocytes to any of three exogenous
mitogens were recorded.
DR. MILLER: As to continuing surveillance of exposed groups,
there are three keys to success. One, know who was exposed; two,
know the body burden or the dose; and three, know an easy, almost
effortless way to follow people inexpensively.
109
-------
Our National Death Index, which has recently been established
by the National Center for Health Statistics, will enable easy followup
of any person listed on an exposure registry. When a death is reported,
we can determine the cause from the death certificate. We will
still need to take a census following an acute environmental episode
to determine who was exposed, the dose involved, and the followup
system needed.
DR. NELSON: How were PBB's finally identified in the feed?
DR. LANDRIGAN: It is a long and complex tale best told in the
book Bitter Harvest by Fred Halbert, a chemical engineer turned
farmer in Michigan (Halbert and Halbert, 1978). He lost a prize
herd of some 3,000 cattle to PBB. The discovery that PBB was the
offending agent was due probably more to his efforts than to anything
else.
In the beginning, Halbert went to many different laboratories
around the country with feed samples. The feed was analyzed by
conventional gas chromatographic techniques, and the results were
uniformly negative. Fortunately, he persisted and finally went to
the U.S. Department of Agriculture field station in Ames, Iowa.
Laboratory workers there injected an extract of feed onto the gas
chromatograph and left it running over a long lunch break. PBB,
which has an unusually long retention time because of its high
molecular weight, finally settled during this longer period. When
the workers returned, they found a late peak, subsequently identified
by mass spectrescopy as PBB.
DR. HUNT: Could you elaborate on the maternal cord-blood
relationships?
110
-------
DR. LANDRIGAN: We looked at the PBB concentrations In the
paired maternal and cord blood of 58 maternal-infant pairs. In
general, there was a slight concentration gradient favoring the
mother, and the average ratio of maternal-to-cord PBB concentration
was 7:1. We also had some 32 maternal blood-breast milk pairs, and
the ratio of PBB in whole milk to serum was 122:1.
Thus, it is clear that PBB does cross the placenta almost
unimpeded. Furthermore, it is concentrated in milk fat and therefore
is a continuing hazard to the nursing infant.
DR. HUNT: In terms of quality assurance of analyses of serum
and adipose tissue, have you examined the effect of freezing on the
stability of the chlorinated hydrocarbons in body tissues over a
period of time? This question relates, of course, to whether you can
go back to other samples that have been frozen for a long time.
DR. LANDRIGAN: To maintain quality control, an enormous effort
has been conducted by the laboratories of the Michigan Department of
Public Health and the CDC. Approximately 40% of the total laboratory
work in this project has been quality control, not only because PBB
is a terribly difficult chemical to analyze, but also because there
is a great desire to put these analyses on as close to an absolute
standard as possible. The objective is that results obtained 20 years
hence will be directly comparable with those obtained today. A freeze-
thaw experiment showed no degradation in PBB concentration.
SPEAKER (UNIDENTIFIED): When you discussed the negative dose-
response correlation between the prevalence of symptoms and the body
111
-------
burdens of PBB, they were all based on symptoms such as fatigue and
joint pain. Were any hard signs used to attempt to generate a dose-
response curve?
DR. LANDRIGAN: We did not do physical exams during this study.
We were trying to interview, within 1 year, some 4,000 persons widely
scattered across the state, and we did not have the personnel to
conduct the exams. However, if a person reported seeing a physician
during the preceding 2 or 3 years—for any condition—we got
permission to contact the physician and collect the medical records
on every such visit.
DR. GUZELIAN: People commonly use serum as a way of analyzing
exposure, and changes of exposure, in relation to symptoms. Serum
may be an inadequate indicator, so I am not surprised that you did
not find a dose-response relationship between symptoms frequency and
PBB concentration in serum given the incredibly disproportionate
distribution of PBB between fat and blood. It is possible in field
studies to obtain samples of adipose tissue by using the Garrottson
technique, which involves subcutaneous needle aspiration of adipose
tissue without anesthesia.
DR. LANDRIGAN: We decided not to use wide-scale fat biopsy
because we wanted to have a cohort to look at 30 years from now.
We felt fat biopsies might deter people from continuing to participate.
Surprisingly we found a very high correlation coefficient with
very little scatter between the PBB concentrations in serum and adipose
tissue. It was 0.960. That finding seems to vindicate the use of PBB
concentrations in serum as an indicator and gives us assurance that we
need not do 4,000 fat biopsies.
112
-------
REFERENCES FOR DISCUSSION
Bekesi, J. G., J. F. Holland, H. A. Anderson, A. S. Fischbein,
W. Rom, M. S. Wolff, and I. J. Selikoff. 1978. Lymphocyte
function of Michigan dairy farmers exposed to polybrominated
biphenyls. Science 199:1207-1209.
Halbert, F., and S. Halbert. 1978. Bitter Harvest: The
Investigation of the PBB Contamination — A Personal Story.
William B. Eerdmans Publishing Co., Grand Rapids, Mich.
Landrigan, P. J. 1980. General population exposure to
environmental concentrations of halogenated biphenyls.
Chapter 9A, pp. 267-286 in R. D. Kimbrough, ed. Halogenated
Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins, and Related
Products. Elsevier, Amsterdam.
Matthews, H., G. Fries, A. Gardner, L. Garthoff, J. Goldstein,
Y. Ku, and J. Moore. 1978. Metabolism and biochemical toxicity
of PCBs and PBBs. Environ. Health Perspect. 24:147-155.
Savage, E. P. 1977- National study to determine levels of
chlorinated hydrocarbon insecticides in human milk, 1975-1976,
and supplementary report to the National Milk Study, 1975-1976.
National Technical Information Services, Accession No. PB284393/As,
Springfield, Va.
Silva, J., C. A. Kauffman, D. G. Simon, P. J. Landrigan, H. E. B.
Humphrey, C. W. Heath, Jr., K. R. Wilcox, Jr., G. Van Amburg, R. A.
Kaslow, A. Ringel, and K. Hoff. 1979. Lymphocyte function in
humans exposed to polybrominated biphenyls. J. Reticuloendothelial
Soc. 26:341-347.
113
-------
Case Studies: The Atomic Bomb Casualty Commission
Gilbert W. Beebe1
Investigation of the acute effects of the atomic bombs dropped
on Hiroshima and Nagasaki in 1945 led to a continuing study of their
late effects by the National Academy of Sciences. Under its Committee
on Atomic Casualties, the Atomic Bomb Casualty Commission (ABCC) was
created to conduct the field investigations in Japan. Although the
initial search for genetic effects was well conceived and effectively
carried out, the search for somatic effects, initially characterized
by ad hoc surveys aimed at particular effects, enjoyed only temporary
success. Investigator interest waned, subject participation declined,
and serious suggestions were made for closing the operation. The
venture was rescued by a reconceptualization of the task along epidemi-
ologic lines. Systematic, multiphasic screening strategies then
began to bear fruit in the 1960's as one form of cancer after another
was shown to be occurring in excess among heavily exposed survivors
and, eventually, to be dose-dependent in quantifiable fashion. The
history of the ABCC contains important lessons for those who may
plan long-term studies of the health effects of large-scale disasters.
In September 1945, American medical teams joined Japanese already
at work in Hiroshima and Nagasaki investigating the acute effects of
the disasters (Oughterson and Warren, 1956). On their return to the
United States, they proposed a continuing investigation of late effects,
and President Truman was persuaded to ask the National Academy of
Sciences, as a nonmilitary, nongovernmental organization, to assume
responsibility for directing a study of the late effects of the
ionizing radiation released by the bombs.
The Academy formed the Advisory Committee on Atomic Casualties
(CAC) and a field organization in Japan, the Atomic Bomb Casualty
Clinical Epidemiology Branch, National Cancer Institute,
Bethesda, Md.
114
-------
Commission (ABCC). It is noteworthy that there was no epidemiologist
among the members of the CAC. In the early years, the ABCC emphasized
the building of an organization along conventional, multidepartmental
lines and attacked specific hypotheses concerning the nature of the
late effects. Investigators were invited to survey survivors for
particular end results, but there was no overall research strategy
other than the search for particular effects. This worked well for
the study of genetic effects, which was well conceived and effectively
carried out, but proved to be only temporarily successful with respect
to somatic effects. There were some very important findings in the
early period (1947-1955)—leukemia (Folley e£ al., 1952), cataracts
(Cogan et al., 1949), and, after in utero exposure, small heads
(Plummer, 1952) and mental retardation (Plummer, 1952)—and an
ABCC Radiation Census in Hiroshima and Nagasaki in 1949 paved the
way for a nationwide census of atomic-bomb survivors in 1950 at the
time of the official Japanese National Census, a census that was to
provide a sound sampling base for the program in later years.
By 1952 or 1953, despite the initial accomplishments, the effort
was losing momentum and suggestions were made that it be terminated.
One consideration was the rapidly mounting cost, which the sponsoring
federal agency, the Atomic Energy Commission (AEC), found increasingly
burdensome. There also was a marked decline in participation in the
clinical examination programs on the part of the Japanese subjects.
Once the early effects had been detected, systematic clinical surveys,
115
-------
in which the greatest investment was being made, yielded essentially
negative results. The genetic study, although massive in scale,
also produced essentially negative results (Neel and Schull, 1956).
What had happened, we now know, was that the immediate effects, and
leukemia, which has a short minimum latent period, were rather
easily and quickly detected, but the solid tumors, which were to
become the major findings in later years, were not yet in evidence
because of their longer latent periods (Beebe et^ al^. , 1978).
The situation became critical in 1955, and Dr. R. Keith Cannan,
as Chairman of the Division of Medical Sciences, National Research
Council, appointed an ad hoc committee, the so-called "Francis
Committee", under the chairmanship of Thomas Francis, Jr., the
virologist and epidemiologist. The other members of this committee were
statisticians: Seymour Jablon, of the Academy staff, and Felix Moore,
then at the National Heart Institute. Dr. Cannan asked that this
group visit the operation in Japan and recommend action that might
be taken in order to fulfill the original mission of ABCC. The
Francis Committee prepared a most remarkable document that drew upon
the considerable strengths and experiences of the past program, made
recommendations for eliminating essential weaknesses, suggested
specific methodologic innovations, and visualized an integrated
strategy for the conduct of a continuing program (Francis et_ al. ,
1959). It provided a conceptual overview within which a variety of
research approaches were integrated in a "Unified Study Program."
A considerable emphasis was placed on systematic screening of fixed
116
-------
cohorts in different ways and on continuity of effort within
permanent research designs.
The committee discovered that much of the work started during
the early years remained unfinished. New members of the staff had
arrived, started new projects, and completed their fixed tours of
duty before their work had been completed. There was a morgue of
unfinished projects, tabulations, and manuscripts. To ensure
continuity of effort and program in the face of short-term staff
assignments was a major objective of the Francis recommendations.
Fortunately, Japan has a nearly unique family registration
system that provides for the posting of vital events on a family
register in the "home" city. This record makes it possible to
trace people for mortality on the basis of their permanent family
"home", even if they reside elsewhere. The Francis Committee
recommended that this system be used to trace mortality for fixed
cohorts defined on the basis of exposure to the bombs. Major
samples include survivors' children conceived after the bombs,
those exposed postnatally, and those exposed In utero. All are
critically important groups in the program today (Beebe and Usagawa,
1968).
Almost concurrently with the work of the Francis Committee
plans were begun to develop a physical dosimetry program recommended
by the AEC. ABCC technicians obtained shielding histories from
atomic-bomb survivors. An Oak Ridge National Laboratory group then
created dose-distance curves and arranged for experimental work at the
117
-------
Nevada Test Site in order to develop transmission or attenuation
factors associated with environmental shielding (Auxier, 1975).
Recently, this group has provided the basis for adjusting essentially
external doses to tissue doses for critical organs (Kerr, 1979).
These are individual doses with separate components for gamma and
neutron radiation.
Initially, the research goal was mainly to identify affects.
Now, in contrast, the main emphasis is placed on the measurement
of effects in relation to dose, host factors, and other environmental
factors. This interest is not merely descriptive; it extends to
mechanisms as well (Beebe, 1981).
Since 1955 there have, of course, been changes in the Unified
Study Program, changes that have been to some extent responsive to
technological change and to advances in radiation biology. Although
the Francis Committee proposal for an "epidemiologic detection network,"
a block-based morbidity reporting system, was abandoned at the outset,
all other features were retained: a clinical detection program in
which survivors were seen in outpatient clinics, a postmortem detection
program, a laboratory detection program, and a death certificate
study. These components were unified through common samples or sub-
samples. Additions to the program include cancer registries, newly
designed genetic studies, and, of the utmost importance, a physical
dosimetry program. In 1975 an ambitious biochemical genetics study
was launched (Neel et_ al. , 1980). This study extended earlier
genetic studies of pregnancy termination and birth defects (Neel and
118
-------
Schull, 1956), mortality patterns of the FI generation (Neel et al.,
1974), and chromosomal aberrations in the F, generation that might
be indicative of hereditary changes (Awa, 1975).
The atomic-bomb victims themselves have been very cooperative,
despite the policy of ABCC not to provide direct medical care
except in a few, now-closed diagnostic beds in Hiroshima. There
have been no attitude surveys or other social surveys that might
give the American investigators a good reading on the feelings of
the Japanese victims of the bombings, but the often unfriendly tone
of the local Japanese press contrasts sharply with the continued
cooperation of the subjects themselves. There is at least one fairly
substantial observation: Dr. Robert Lifton (1967), the Yale
psychiatrist, suggested that even those who experienced very low
doses of ionizing radiation are quite fearful of the future,
especially from the standpoint of cancer. This fear exists despite
two facts: (1) it has never been possible to demonstrate excess
cancer at the low doses experienced by the great majority of the
survivors, and (2) the best current estimate is that, from 1950 to
1974, the 285,000 survivors enumerated at the time of the 1950 census
may have suffered about 415 deaths as the late effects of exposure
to ionizing radiation, an increment of about 1 in 690 survivors, or
0.6 percent of the naturally occurring deaths among them in this
period (Beebe et_ al., 1978). The lack of a program to inform the
survivors of their real risks has seemed to me most unfortunate.
119
-------
Another notable fact has been an almost unrelenting hostility
on the part of the local press. Allegations that have been made
repeatedly include: the survivors are merely guinea pigs; the
aggressors should not be entrusted with the studies; secret data are
sent to the United States for use in its own defense; and deaths of
atomic-bomb survivors are mainly "A-bomb deaths", attributable to
radiation. Although much of the research directly benefited the
survivors by providing a diagnostic and referral service, an aggressive
program to procure permission to autopsy survivors living in the
community during the 1960's, although quite successful and apparently
acceptable at first, led eventually to very negative reactions in
the press and, in the end, in the community as well.
One should bear in mind that the U.S. staff of the ABCC was
for many years under the U.S. Embassy and enjoyed some of the
perquisites of diplomatic status. In 1974, when discussions were
under way to replace ABCC with the present organization, the
Radiation Effects Research Foundation, it became clear that diplomatic
privilege was symbolic of foreign control and that the local Japanese,
both physicians and community leaders, were anxious to have Japanese
control. The Government of Japan either did not want this or did
not want to insist on it, and the final agreement called for equality
of both funding and top-level management control. The local people
also wanted a medical care emphasis that had been lacking throughout
the program.
What lessons concerning research strategy can be drawn from
120
-------
the ABCC experience? First, such work must be directed by a compre-
hensive research strategy devised early. This is not to say that
there should be no response to new developments in science, but,
rather, that without a firm navigational plan it is very easy for
an effort of this kind to wander off course. We saw that ABCC
was threatened with closure early in its history. Had it stopped
operation in 1955, we would never have learned about the solid tumor
effects that became evident only in the 1960's, the first being
thyroid cancer in 1962. A research strategy also needs to be
enlivened by imagination about techniques that may only later
become available, laboratory techniques that might require stored
samples of blood or tissue, for example. (These observations and
those described below are discussed in greater detail in Beebe, 1979.)
The ABCC experience clearly indicates the tremendous importance
of registering the exposed population early, before it scatters
widely and becomes infiltrated by people who are looking for com-
pensation or other benefits. In the enumeration process, appropriate
identifiers for long-term follow-up must be obtained.
A third point is that the parameters of exposure should be
ascertained in fine detail, with emphasis on objective physical measure-
ments and on determinations applicable to the individual if at all
possible. This just has to be done early. At ABCC it was possible
to obtain exposure histories long after the bombings because the
bombing was what Lifton has called a supernatural experience, one
that was burned indelibly in the minds of the victims. Even 10 to
15 years after the bombings, it was possible for a technician with
121
-------
a prestrike map to sit down with a survivor and get a pretty accurate
indication of where that person was, what he was doing, and what his
shielding situation was. But such recall must be highly unusual.
The need for fixed cohorts is also very important, not only to
avoid retrospective bias but also to permit absolute risk estimates
to be made. And if mortality can serve as the end point in a long-
term study, much of the cost of a long-term follow-up study can be
(
saved. I think at ABCC we continued a strong clinical emphasis long
after it was most cost-effective.
Staffing patterns are needed that ensure not only competence,
but also continuity. ABCC was fortunate in having one director
for 15 years, and also benefited from arrangements with key U.S.
institutions to assume some responsibility for staffing the major
research departments.
An overall prescription for follow-up studies of the long-term
effects of disasters would include: (1) Devise an adequate overall
strategy early. (2) Register the population of interest as soon as
possible, taking care to include all key identifiers on which follow-up
may depend. (3) Determine the parameters of exposure in detail, with
emphasis on objective physical measurements. (4) Use a cohort approach
and plan statistically powerful comparisons to identify and measure
effects, either in dose-specific fashion or in exposed vs. unexposed
contrasts. (5) Evaluate carefully the potential yield and cost-
effectiveness of alternative end points, especially mortality vs.
122
-------
morbidity and physical defects. (6) Plan staffing patterns to
provide not only excellent leadership and scientific performance,
but also continuity. (7) In an operation of long duration, endeavor
to develop deep local roots.
123
-------
REFERENCES
Auxier, J. A. 1975. Physical dose estimates for A-bomb survivors—
Studies at Oak Ridge, U.S.A. J. Radiat. Res. (Tokyo) (Suppl)
16:1-11.
Awa, A. A. 1975. Chromosome aberrations in somatic cells
(in atomic bomb survivors). J. Radiat. Res. (Tokyo) (Suppl)
16:122-131.
Beebe, G. W. 1979. Reflections on the work of the Atomic Bomb
Casualty Commission in Japan. Epidemiol. Rev. 1:184-210.
Beebe, G. W. 1981. Overall Risks of Cancer in A-Bomb Survivors
and Patients Irradiated for Ankylosing Spondylitis. In
H. Burchenal aoi H. F. Oettgen, eds. Cancer, Achievements,
Challenges, and Prospects for the 1980's. Grune & Stratton, Inc.,
New York.
Beebe, G. W., and M. Usagawa. 1968. The Major ABCC Samples.
Atomic Bomb Casualty Commission TR 12-68, Hiroshima.
Beebe, G. W., H. Kato, and C. E. Land. 1978. Life span study
report 8: Mortality experience of atomic bomb survivors,
1950-74. Hiroshima. Radiat. Res. 75:138-201.
Cogan, D. G., S. F. Martin, and S. J. Kimura. 1949. Atomic
bomb cataracts. Science 110:654-655.
Folley, D. G., W. Borges, and T. Yamawaki. 1952. Incidence
of leukemia in survivors of the atomic bomb in Hiroshima and
Nagasaki, Japan. Am. J. Med. 13:311-321.
Francis, T. Jr., S. Jablon, and F. E. Moore. 1959. Report
of Ad Hoc Committee for Appraisal of ABCC Program, 1955.
Atomic Bomb Casualty Commission TR 33-59, Hiroshima.
Kerr, G. D. 1979. Organ dose estimates for the Japanese atomic-
bomb survivors. Health Physics 37:487-508.
Lifton, R. J. 1967. Death in Life: Survivors of Hiroshima.
Simon and Schuster, New York.
Neel, J. V., and W. J. Schull. 1956. The Effect of Exposure
to the Atomic Bombs on Pregnancy Termination in Hiroshima and
Nagasaki. National Academy of Sciences, Washington, D.C.
124
-------
Neel, J. V., H. Kato, and W. J. Schull. 1974. Mortality in
children of atomic bomb survivors and controls. Genetics
76:311-326.
Neel, J. V., C. Satoh, H. B. Hamilton, M. Otake, K. Goriki,
T. Kageoka, M. Fujita, S. Neriishi, and J. A. Asakawa. 1980.
A search for mutations affecting protein structure in children
of atomic-bomb survivors: A preliminary report. Proc. Natl.
Acad. Sci. USA 77:4221-4225.
Oughterson, A. W., and S. Warren. 1956. Medical Effects of
the Atomic Bomb in Japan. McGraw-Hill, New York.
Plummer, G. 1952. Anomalies occurring in children exposed
in utero to the atomic bomb in Hiroshima. Pediatrics 10:687-693.
DISCUSSION
PROF. DARDANONI: Is there a common registration of patients
in Japanese hospitals? If so, it would be rather easy (and inexpensive)
to monitor morbidity.
DR. BEEBE: I know of no way to do this in Japan. There are
insurance programs in Japan which, under other circumstances, might
suit the need for morbidity data. As a substitute, we developed
tumor registries and tissue registries, which are essentially tumor
registries with histologic slides.
125
-------
Reproductive Injury: Love Canal
David Axelrodl
The most likely source of human exposure to the chemicals
deposited in the Love Canal landfill is through leaching of the
compounds in soil, leading to high concentrations of chlorotoluene
and chlorobenzene in the ambient air of nearby homes. Because fetuses
are especially vulnerable to chemical insult, congenital defects,
spontaneous abortions, and below-normal birth weight were selected as
indicators of toxic human exposure. Adverse pregnancy outcomes were
compared among matching groups in Canada, a neighborhood near but not
contiguous to the landfill, and four sectors of the Love Canal area,
two of them abutting the landfill. Findings suggest that an increased
number of spontaneous abortions and/or low birth weights may have
occurred in certain sections of the canal area. However, the results
with respect to congenital defects were not statistically significant.
The relationship of adverse pregnancy outcomes to evidence of chemical
exposure has not been established.
The original geologic surveys of Love Canal are important to
understanding the possible modes of human exposure and the ultimate
effects of the chemicals on human reproduction. Soil strata at the
canal site generally consisted of a thin mantle of silts and fine
sands on top of low-permeability clay. The clay strata apparently
acted as a barrier to water movement of chemicals below the surface.
Rain and groundwater, accumulated either from natural or human
activity, probably carried the chemical waste to layers of soil that
had higher permeability, thereby facilitating the lateral migration
of the pollutants.
York State Department of Health, Albany, N.Y.
126
-------
The chemicals thus leached into the top soil layers and into
the basements of nearby homes, resulting in human exposure to the
contaminants. From the basements and top soil strata, the chemicals
probably volatilized, producing high air concentrations of various
agents, such as chlorinated organic compounds with low molecular
weight.
Initial air samples, taken by the Environmental Protection Agency
from the basements of 14 houses adjacent to the canal, showed the
presence of 26 organic compounds. These samples showed clear evidence
of chlorotoluene in the basement air of 33 houses (of 99 sampled)
abutting the canal. Only four houses (of 256 sampled) in the outer
areas of the landfill site showed trace evidence of contamination
with chlorotoluene. There was similar evidence of the presence of
chlorobenzene. Neither of these chemicals is commonly found in
household products. Benzene, toluene, chloroform, and many other
compounds were identified in a large number of homes, but the presence
of these chemicals did not necessarily correlate with the presence of
chlorotoluene or chlorobenzene.
Other migration mechanisms may also have been in effect.
Chemicals may have traveled along surface paths, such as along the
lowlands that were present before housing was built peripheral to the
landfill. The topography of the region is generally flat, except for
three creeks north of the canal. These creeks contain water throughout
the year. Shallow depressions (swales) traversed the area; some
127
-------
transected the canal itself before the housing was built. The location
of these swales was determined from aerial photographs, which were
independently interpreted by personnel at Cornell University's School
of Civil and Environmental Engineering. The university researchers
had no prior knowledge of the swale hypothesis. The depressions
could have served as drainage ways and even have produced ponds in
some sections during times of high water. Additional verification of
these historically "wet" areas was obtained by interviewing residents
and reviewing their photographs and motion pictures.
Building construction has since modified the contour and extent
of the swales. Housing built between 1951 and 1956 eliminated the
major one, which had intersected the canal and carried surface water
to peripheral areas. One 1966 aerial photograph shows only one small
swale, northeast of the canal. At the time of this study, no visible,
above-ground evidence remained of these natural swales.
Filling in the swales eliminated the potential chemical transport
by actual surface flow, but not by migration. Permeable fill used in
the swales may still have carried leachate from the canal. Furthermore,
chemically contaminated soil may have been used to fill in low-level
areas during development. For instance, major portions of the pond
sections were filled during 1958. Minutes of a school board meeting
in the 1950's record the transport of earth from the 99th Street
schoolyard to the 93rd Street school property, beyond the area under
study. Aerial photographs show that by May 1958, the vast majority
of the historical depressions were leveled. A map of this area is
shown in Figure 1 of my paper entitled "Chlorinated Hydrocarbons,"
which appears earlier in this volume.
128
-------
Yet another migration mechanism was recently uncovered.
Construction at 99th Street at the edge of the canal unearthed three
pipes (10.0 cm in diameter) approximately 1 meter below the ground
and extending laterally toward a former pond on the periphery of the
study area. Farmers apparently had used these pipes to draw water
from the canal before the chemical wastes were placed there.
Collectively, this information yields a relatively unambiguous
picture of the original land surface, as modified by construction.
And the hypothetical patterns of chemical migration along swales,
through pipes, in earth transported elsewhere as fill could result
in "preferential" contamination of the wet areas. Thus, for this
study, housing was divided into two groups: all houses (116) built
on wet (or water) areas (other than those adjacent to the canal
itself), and those houses (268) built on historically "dry" (nonwater)
areas. Soil samples were taken at every home to verify the former
locations of ponds and swales and to determine the extent of chemical
contamination of the soil. The samples were then divided into two
groups — undisturbed soil and disturbed soil, the latter samples
containing at least some fill dirt. Houses with at least 1 meter of
fill and houses adjacent to these were labeled "fill" houses. A high
coefficient of agreement was found between fill houses and water
houses. Comparison of a map of fill houses to a map of water houses
substantiated the positive correlation, but other, more general modes
of exposure and possible conduits of chemical migration were also
considered. Residents of the area might have been exposed to toxic
129
-------
vapors emanating from the canal and transported by air or to contami-
nation via the public water supply, in existence since 1930.
The study of the temporal pattern of housing development, the
topography of the canal area, and the initial chemical evaluations
suggested that many houses directly adjacent to the canal were
contaminated. Pollution of the more peripheral areas would most
likely have occurred through some general mechanism affecting broad
areas or by a more selective route, such as the historic water area.
Each street was examined separately in all statistical analyses
because the temporal factors might be important and the nouses on
99th Street were considerably older than those on 97th Street. Also,
studying the area street by street might reveal a gradient effect as
distances from the canal increased, which would support the hypothesis
of some general mechanism of contamination. Historic water areas,
where chemicals might have migrated from the canal or been brought in
with contaminated fill, might show more evidence of contamination
than the nonwater areas, which also generally had newer houses.
Specific biologic events—congenital defects, spontaneous
abortions, and low birth-weight infants—were enumerated as possible
end points of exposure, although the correlation of their incidence
in humans to chemical exposure is not well established. However,
these indicators have a shorter induction period than do most adult
chronic diseases, and the prepartum period is generally accepted to
be especially susceptible to chemical insult. It is agreed that
130
-------
many chemicals are hazardous to the conceptus of lower animals,
depending on dosage, route of administration, and stage of gestation
at exposure. Three end points were examined because of the wide
variety of chemicals involved and their differing toxic manifestations.
The monitoring of spontaneous abortion and the identification of
environmental teratogens are important because the number and variety
of congenital anomalies during gestation is far greater than can be
detected from an analysis of full-term births.
Each of these end points has certain limitations. The frequency
of spontaneous abortions, especially of those occurring very early in
pregnancy, is exceedingly difficult to measure. Some congenital
malformations, especially cardiac abnormalities and mental defects,
are difficult to diagnose during the immediate postpartum period.
Birth weight alone may not be a sufficiently sensitive or specific
indicator of toxic effect.
Five epidemiologic hypotheses were advanced concerning possible
distribution of spontaneous abortions, congenital defects, and low
birth weights in specific sections of the study area. First, an
increased incidence of some or all of the indicators might be
demonstrated only among pregnant females residing in houses on 97th
and 99th Streets, the two streets where backyards extend to the dump
site and where the likelihood of chemical contamination was greatest.
Second, an excess might also be observed in more peripheral areas.
Examination of each indicator by location might point to a gradient
effect as the distance from the canal increased. Third, the entire
study area, including 97th and 99th Streets, might have an increased
131
-------
rate of all or some of the biologic indicators, suggesting a general
mechanism of exposure, e.g., air, water. Fourth, increased incidence
might be observed in the historic water areas of the neighborhood,
and a lesser frequency in historic nonwater areas. Finally; the
incidence rates might be related to temporal factors and to the
nature and concentration of the chemicals to which pregnant females
were exposed. For example, 99th Street might be characterized by an
increased incidence of the most severe indicator, spontaneous abortion,
and streets farther from the landfill might have an increased incidence
of low birth weight infants.
Initially, expected numbers of spontaneous abortions were obtained
from a report by Warburton and Fraser (1964). They studied a relatively
large sample of the spontaneous abortion incidence in more than 6,000
Canadian pregnancies, tabulated both by birth order and maternal age
at conception. However, possible differences between demography,
economics, and other characteristics of this population and those of
the Love Canal group dictated the need for additional control groups.
Thus, a similar distribution table was constructed for residents of
houses north of Colvin Avenue. This latter area, directly north of
the canal, was selected because it has a large population residing in
single-family houses near the canal. Most important, residents were
concerned that chemicals might have migrated into their area. More
than 98 percent of the adult residents from this area, as well as
from the Love Canal site, participated in the investigation.
The validity of this comparison group is also limited. A greater
proportion of the Colvin Avenue residents have college educations
132
-------
than do those in the entire study area, although this factor was
controlled for in the statistical analysis. In addition, the area north
of Colvin Avenue may also have been subject to chemical contamination.
Were this so, it might lessen the extent to which the rate of abnormal
reproductive events among Love Canal women appeared to deviate from
the expected.
Finally, a maternal age parity table was constructed based on
the spontaneous abortion experience of females residing in nonwater
areas. This internal comparison group was used to evaluate further
the hypothesis that an increased number of reproductive indicators
might be present in historic water sections. But, again, the
possibility of prior chemical contamination could not be excluded.
The same survey instrument was used throughout the investigation,
with questions relating to medical, therapeutic, social, occupational,
and pregnancy histories answered by all adult residents in a door-to-
door survey. The same field teams were used in all areas studied.
Investigators had no prior knowledge of the specific hypotheses under
evaluation or of the areas selected as comparison groups. During the
field work, supervisors, with at least 4 years of experience, reviewed
completed questionnaires to assure standardized data collection. Two
other supervisors and members of the statistical unit subsequently
reviewed all questionnaires for completeness and possible inconsistencies.
A number of measures were followed to check validity of responses.
The interview procedure had a built-in recall mechanism, and certain
questions were asked repetitively in the questionnaire. Statistical
analyses of the low birth weight indicator were performed separately
133
-------
on Interview data and birth certificates obtained from the Office of
Vital Records, and the results were compared. An effort was made to
confirm all spontaneous abortions through vital records, physician
interviews, and medical records. Questionnaires were keypunched in a
key-to-disc unit with verification. Programs were checked by hand
calculation to guarantee accurate implementation of computational
algorithms. Statistical routines of sufficient complexity to defy
hand computation were run against test data for verification.
Indicators were included only for those females who resided in
the study area as of June 1978, and who lived there during the entire
period of pregnancy. Spontaneous abortions were considered, confirmed,
and included in analyses if they were verified by personal physicians
or hospital records or if there was evidence of pregnancy with a
subsequent history compatible with this outcome. Women who were ever
pregnant at their present address in the study area were categorized
according to their age at delivery and parity for each pregnancy.
For each of the areas studied, the number of miscarriages observed
among the pregnancies was determined. The number of miscarriages expected
among these women was calculated by applying the percentage of miscarriages
for each parity combination of the other control groups to the observed
number of pregnancies in each cell.
The expected numbers for all study areas were derived from the
Warburton and Fraser report and from the controls in the Colvin area.
The nonwater area was used as a control for the water area. The
Mantel-Haensel chi square was used to test the differences between
the observed and expected numbers of miscarriages.
-------
Findings showed that, of the residents on the streets directly
adjacent to the canal, only those on 99th Street may have had an
excess of spontaneous abortions. The number of births in the water
area was 80; in the nonwater area, 149. There was a much higher than
expected incidence of untoward pregnancy outcomes in the wet households:
11 among 80.
With respect to low birth weights (defined as 2.49 kg or less),
the relative risk of a small or a low birth weight infant from
contaminated areas of the Love Canal was twice what was expected from
a noncontaminated area. Comparisons were based on a 1977 New York
State annual report. The New York State number was 3.16 kg for all
white births; in the contaminated areas it was 6.31 kg, and in the
noncontaminated areas it was 3.29 kg (New York State Department of
Health, 1977).
The apparent increased rates of spontaneous abortions and low
birth weights in the 99th Street and water areas may be due to some
confounding variables or combination of factors such as maternal age and
parity.
There was also the possibility that women who had one or
more adverse pregnancy outcomes while residing in the Love Canal
area might have had similar experiences before moving into the
area. No significant number of prior spontaneous abortions was found
using numbers from either the Warburton and Fraser report or those
from the Colvin group. Prior low birth weights were not significantly
increased when compared to rates from New York State. In addition,
135
-------
there was no documented evidence of congenital defects among the 57
live births to female residents of 99th Street and the water areas
before they moved to the canal area. This finding is in juxtaposition
to the 14 such episodes among 122 live births while these females
resided in these two areas.
Workers in certain occupations, such as operating room attendants
and chemical and laboratory operators, might also be subject to
increased rates of spontaneous abortion and congenital malformation.
Controlling for such exposure revealed no significant differences
among households with either spontaneous abortions or congenital
defects. There were too few households with a low birth weight child
for the five-area chi square analysis.
Still other factors must be considered. Certain infections,
such as rubella, mycoplasmas, and toxoplasmosis; prior, induced
spontaneous abortions; alcoholism; and clinical conditions, including
diabetes mellitus, epilepsy, malnutrition; and maternal exposure to
lead, mercury, arsenic, or ionizing radiation, all might adversely
influence pregnancy outcome. The medical histories of females having
had spontaneous abortions and of children with either congenital
defects or low birth weight were reviewed and, with one exception,
none of these factors could be documented as pertinent.
Variability in interviewer technique and/or respondent's recall
could also account for some of the differences. Analysis of the data
by individual interviewer did not suggest that an increased incidence
was attributable to any one interviewer. The time interval between
136
-------
the date of interview and the date of spontaneous abortion reported
by women in the water and nonwater areas was examined. And, finally,
a validity check suggested that the accuracy of respondent recall was
comparable in the Love Canal and Colvin areas.
A comparison of the low birth weight information, as determined
by interview data from each area, to the vital statistics data, revealed
similar results. Only the water areas had a significant number of
low birth weights when compared with the ratio of low birth weights
recorded in New York State, excluding New York City.
Of all the factors, the most important were probably mother's
age, parity, smoking, drinking, and social class, as indicated by
educational status. House age and duration of residence were also
important considerations.
In summary, these findings are consistent with the possibility
that a slight to moderate increase in spontaneous abortions and/or
low birth weight infants may have occurred on 99th Street and in
historic water sections of the Love Canal area. Similar pregnancy
outcome patterns have been observed among women living near a Swedish
smelter that produces metallurgic and chemical products. Birth
weight, obtained from vital records, is clearly the more objective
and reliable indicator.
The results regarding congenital defects were not statistically
significant, but the highest percentages were observed on 99th Street
adjacent to the canal and in historic water sections.
137
-------
Finally, and most important, geographic distributions of adverse
pregnancy outcomes were not correlated with chemical evidence of
exposure.
There is no direct evidence of a cause-effect relationship
between chemicals from the canal and adverse pregnancy outcome.
138
-------
REFERENCES
New York State Department of Health. 1977. Vital Statistics
Report. Albany, N.Y.
Warburton, D., and F. C. Fraser. 1964. Spontaneous abortion risk
in man: Data from reproductive histories collected in a
medical genetics unit. Am. J. Human. Genet. 16:1-25.
139
-------
DISCUSSION
DR. SUSKIND: What about possible chemical contamination from
the swale area into the household? Very early in the study, the
basements of some of the houses were examined for contaminants. Were
basements of houses where families had reported birth defects examined
for contaminants?
DR. AXELROD: We examined virtually all houses to determine the
concentrations of chemicals. At this point, we cannot draw any
conclusions about concentrations of various chemicals in the air with
respect to the presence or absence of a congenital defect. Any air
measurement is taken at one instant. Several times when we went back
and repeated observations, samples produced different results.
DR. MOORE: Given the variety of chemicals that were found in
Love Canal and given that your findings depended on numerous factors,
do you ever expect to find a direct correlation between Love Canal
exposures and a given birth defect or a low birth-weight child?
DR. AXELROD: It would be virtually impossible to exclude or
include any individual case, but the accumulated epidemiologic evidence
strongly suggests a correlation between an adverse response and
exposure to the gamut of chemicals present in the area. We were
unable to correlate a specific adverse response in our studies with
respect to actual concentrations at a given location. A correlation
is not excluded in the generic sense.
DR. MILLER: If you look over the list of situations that have
been discussed involving PCB's, there was a transplacental effect.
140
-------
With kepone, there was effect on fertility; with DBCP, an effect on
sterility; with lead, an effect on the sperm and on the fetus; with
methyl mercury, on the fetus. An effect on reproductive performance
was seen — except at Love Canal where there were 82 or more chemicals.
DR. AXELROD: There is, of course, a problem with small numbers
in the study, and the exposures are very different in some respects
from those at Yusho, Minamata, and the others. Those episodes involved
relatively high dose levels. At Love Canal, we are considering a
chronic lower level of exposure. In fact, the dose cannot even be
measured. In the studies of atomic bomb effects, long latent periods
have now elapsed, so effects can be found where information was
previously lacking. There may or may not be a threshold at Love
Canal. The sensitivity of the Love Canal studies to date is probably
too weak to detect a low-dose effect on fertility rates.
DR. REGGIANI: Was TCDD among those chemicals found at Love
Canal?
DR. AXELROD: Yes. We expected to find dioxin because of
the kind of manufacturing process Hooker Chemical Company used at
Niagara Falls. We also expected it when we learned of the large
potential concentration of trichlorophenol. Subsequently, we found
TCDD in several samples from areas adjacent to the canal and, also,
on the bank of a stream north of the canal where there would have
been sewer discharge.
Based on some grids we laid out, we believe that some construction
work has prevented the TCDD from migrating as rapidly or as far as
141
-------
would normally be anticipated. For example, our samples turn from
positive to negative at the end of the canal area, as if the chemicals
did not cross the street. Construction of the street apparently
served as a barrier, certainly to our analytic capability. The
chemicals may be there but we cannot detect them. We could measure
parts per trillion in the area adjacent to the road.
DR. REGGIANI: Was that the highest level that you measured?
DR. AXELROD: No, there were levels in the parts per billion
range, as well. In the soil samples, as we moved toward the roadway,
there were levels in the parts per trillion range. The difficulty
is that we do not know exactly where the trichlorophenol was placed
in the canal. If we could precisely isolate the area where the
trichlorophenol was, we could probably get higher concentrations.
Basically, we are sampling at random because the chemical distribution
is unknown to us, and, I believe, unknown to anybody.
DR. REGGIANI: You expect a higher concentration?
DR. AXELROD: It is possible. Based on the levels of contamination
known to be present in trichlorophenol, we would not be surprised to
find considerably higher levels, at least in some of the chemical
waste material that was removed from the canal proper. How much
would have migrated, how much would have moved into the soil area,
is anybody's guess.
DR. REGGIANI: I would expect concentrations at the parts per
million level.
142
-------
DR. AXELROD: I would not be surprised to find that. It depends
on where you sample. We could be sampling in one area and find
nothing, and we would sample 2 meters away and find significantly
higher concentrations.
We don't know if disposal of chemicals in the canal was orderly
at all times. According to some stories, the canal was partially
filled in certain areas, with specific chemicals being placed in one
area and not in another. We cannot confirm these.
DR. MOORE: Please elaborate on your concluding remark that
there was no direct correlation or direct evidence of cause and
effect.
DR. AXELROD: I did not mean it in the generic sense. In terms
of soil sampling, and with respect to actual chemical concentrations
in a given location, we were unable to correlate contamination with a
specific adverse response. That is not to say, in the generic sense,
that we do not believe that there is a correlation.
PROF. DARDANONI: What about fertility rates in the different
groups?
DR. AXELROD: They are comparable, both with respect to the
control groups and with respect to the various study populations we
evaluated.
DR. MILLER: Would you recommend that studies be performed at
other hazardous waste disposal areas?
DR. AXELROD: Something at the intensity of the study at Love
Canal is probably not feasible on a grand scale. Where a reasonably
143
-------
direct population exposure can be assumed, then I think this sort of
study has to be made. These sites have to be very carefully identified
and selected because there simply are not enough resources to examine
every one of the thousands of sites within the United States or even
the hundreds of sites in New York State.
There is an absolute lack of data with which to compare the kind
of information we accumulated at Love Canal. The first question to
ask when you start comparing the outcomes of a study is what are you
comparing them with? Certainly, the Warburton-Fraser report was not
on an industrial community. The area north of Colvin was appropriate
because it was in the same community. But there are legitimate
questions that need to be raised regarding the kinds of outcomes for
pregnancy in other industrialized areas where there is exposure to
benzene, chlorotoluene, or any of a large variety of chemicals as a
result of manufacturing processes or leaching.
144
-------
Reproductive Injury; General Considerations
Robert W. Miller1
Reproductive injury may occur either before conception by affecting
the germ cells, or after, by affecting the embryo or fetus. At the Atomic
Bomb Casualty Commission in Japan, seven measures of germ-cell mutation
have been studied: stillbirths and neonatal deaths, congenital mal-
formations, birth weight, sex ratio, anthropometries at about 9 months of
age, F, mortality, and biochemical genetics. Sterility or infertility are
other measures of either preconception or postconception effects. In
addition, injury to the embryo may be measured by the frequency of
miscarriages, malformations due to exposures while in utero, and
transplacental carcinogeneiss. Observations in domestic or wild animals
may provide clues to environmental agents that also cause harm to the
human conceptus.
Reproductive injury occurs either before conception that is, by injury
to germ cells, or after conception by injury to the embryo. The two
possibilities are very different. The first involves a germ-cell mutation;
the second, an embryologic catastrophe affecting the organism in utero.
STERILITY, INFERTILITY, AND MEASURES OF GERMINAL MUTATIONS
One measure of reproductive injury is sterility: no sperm yields no baby.
Another possible effect is infertility: a diminished number of sperm diminishes
the chance of having children.
The most comprehensive measurements of human genetic damage have been
made by the Atomic Bomb Casualty Commission (now called the Radiation
^Clinical Epidemiology Branch, National Cancer Institute, Bethesda, Md,
145
-------
Effects Research Foundation) in Japan. The study by Neel and Schull
(1956) involved six indicators of genetic damage observed in 70,000 children
conceived after the bombs were dropped. Only a small percentage of the parents
had been heavily exposed.
No excess in malformations was observed, nor was there an increased
incidence in the frequency of stillbirths and neonatal deaths combined.
However, any increase less than 1.8-fold would have gone undetected. There
were also no effects attributable to radiation in the third and fourth
measures of genetic effects—birth weight and measurements at 8 to 10
months of age. The fifth—disturbance in the sex ratio, depending on
whether the mother or father was exposed—was observed in the early
years; but this finding, of uncertain significance, disappeared
10 years later. The sixth—death rate in the generation of children
conceived after exposure—also showed no effect.
Recently, a seventh measure involving biochemical genetics has been developed
(Neel et_ al., 1980). An electrophoretic study is made of proteins in the child's
blood. An abnormality found in red cells or plasma that is not present
in the parents' blood is presumed to be a mutation attributable to the exposure.
EMBRYOTOXICITY AND TERATOGENESIS
Measuring the effects of postzygotic exposures includes recording the frequency
of miscarriages, which are very difficult to evaluate because women often do not
know they are miscarrying, especially during early pregnancy. Also, a woman may
provide very different responses at different times. Hospital records, of course,
do not routinely provide information about miscarriages that occur at home.
146
-------
Environmental effects on the embryo during the first weeks of pregnancy may
be lethal (embryotoxic). Later in the first trimester, during the formation
of body organs, such effects can cause congenital anomalies (teratogenesis).
Small head circumference and mental retardation following intrauterine
radiation exposure were known before the atomic bomb exposures in
Japan. Fourteen case reports related to pelvic radiotherapy had been
collected by 1928 (Murphy, 1929). Subsequently; an inexpensive mail
survey of obstetric services in the United States located 16 more children
with small head circumference and mental retardation whose mothers had
received pelvic radiotherapy (Goldstein and Murphy, 1929).
Thus, pediatricians with the Atomic Bomb Casualty Commission knew to look
for these effects, and found them in children exposured to radiation
before they had reached 18 weeks of gestational age (reviewed by Miller
and Mulvihill, 1976). The effect was related to dose.
In Hiroshima, exposures before the fourth week of gestational age apparently
had a lethal effect. Exposure between the 4th and the 17th weeks induced small
head circumference in a substantial proportion of infants; the frequency and
severity of this effect increased as the dosage increased. The effect in
Hiroshima was detectable even among children whose mothers had received
only 10-19 rads in air; the embryo had received less because the mothers'
bodies had attenuated the dose. Higher doses, beginning at about 50 rads, were
associated with mental retardation. In Nagasaki, fewer pregnant women
were exposed; small head size and mental retardation of their infants
147
-------
were not associated with a dosage of less than 150 rads (Blot and Miller,
1973; Miller and Blot, 1972).
Polychlorinated biphenyls (PCB's) cross the placenta and can produce
a teratogenic effect. Affected infants are small for date, have transient
hyperplgmentation, and eyes swollen by Meibomian cysts. In two of nine
cases reported by Taki ejt^ al. (1969), teeth were present at birth.
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is teratogenic in laboratory
animals and was reported by Kimbrough e£ al. (1977) to cause unspecified
malformations in horses. In Moscow Mills, Mo., a waste-oil dealer mixed
TCDD with oil and disposed of it illegally by spreading it on the earth
in horse arenas. As a consequence, 43 horses died and 26 others aborted.
Many foals were stillborn or died soon after birth (Kimbrough, 1977).
TRANSPLACENTAL CARCINOGENESIS
Chemicals that cross the placenta can induce cancer in offspring. The
first transplacental carcinogen recognized in humans was diethylstilbestrol
(DES). In Boston, seven cases of a rare tumor (adenocarcinoma of the vagina)
were observed in young women within a 4-year period. Retrospective study
identified DES, given therapeutically to their mothers during pregnancy,
as the cause (Herbst ££ al_. , 1971). In fact, a mothers whose daughter
was affected suggested DES to the medical investigators who were on the
watch for such an explanation (Ulfelder, 1980). The relationship was
quickly confirmed by the New York State Tumor Registry (Greenwald et
al. , 1971). The risk eventually was shown to be about 1 per 1,000 DES
daughters (Herbst et_ £l. , 1977).
148
-------
After the transplacental carcinogenic effect of DES was discovered,
a high proportion of the women who had been exposed in utero were found to have
congenital anomalies of the lower genital tract (Herbst et al., 1975). An
anomalous contour of the uterine cavity, recorded on hysterosalpingograms, has
now been observed and is associated with abnormal pregnancy outcomes (Kaufman
et al_., 1977). A recent followup of daughters whose mothers participated in
a clinical trial of DES given randomly during pregnancy several decades ago
revealed that the unfavorable pregnancy outcome was related to DES (Herbst,
1980). These results from a clinical trial are not in accord with the sugges-
tion by Barnes et al. (1980) that the effect on outcome of pregnancy may be
heritable.
Among other possible transplacental carcinogens is diphenylhydantoin (DPH),
prescribed for epilepsy. Taken during pregnancy, DPH occasionally
produces a syndrome of birth defects, including hypoplasia of the midface
and hypoplasia or aplasia of the fingernails and toenails (Hanson and
Smith, 1975). Occasionally, it induces lymphoma in adults (Hoover and
Fraumeni, 1975). This observation led to the expectation that lymphoma
might occur at an increased rate in children with the fetal hydantoin syndrome.
Four children have now been reported with the syndrome, but neuroblastoma,
not lymphoma, was observed (Allen ^aJ., 1980). In the fourth
case, three older siblings were unaffected although the mother received
DPH throughout all pregnancies. There may be an intrauterine interaction
involving other environmental factors or genetic susceptibility of the
fourth child. Letters to Lancet detailed the first two cases. Repeated
appeals for more information about other cases led to the report of the
third. Publication of this case drew a report of a fourth case (summarized
149
-------
by Allen et_ al., 1980). Thus, what is probably the second known transplacental
carcinogen in humans was identified by an approach that one might call
"microepidemiology."
VETERINARY OBSERVATIONS
An article published in Modern Veterinary Practice (Crowe, 1969) described
five epidemics of congenital arthrogryposis in pigs on Kentucky farms, perhaps
due to exposures to pesticides or growth stimulants used on tobacco (Crowe and
Swerezek, 1974). Similar episodes occurred in Missouri, where ingestion of
wild black cherries was implicated (Selby et al., 1971). No comparable human
effects are known, but such observations in domestic (or wild) animals may
serve as indicators of potential embryologic hazards in humans.
Chemicals can cross the placenta and produce death, deformity, or cancer
in the embryo or fetus. They can also be transmitted by lactation, through
the fat of breastmilk in which PCB's, TCDD, polybrominated biphenyls (PBB's)
and pesticides are concentrated. Finally, in theory at least, the embryo
or fetus can be affected by dust brought home on a parent's workclothes.
No examples of embryologic effects are known, but asbestos has been
transmitted to children this way in several households, and decades
later those persons developed mesothelioma (Anderson et^ al. , 1976; Li jilt
al., 1978). Chisolm (1978) has referred to such contamination of the home
as fouling one's own nest.
150
-------
REFERENCES
Allen, R.W., Jr., B. Ogden, F.L. Bentley, and A.L. Jung. 1980. Fetal
hydantoin syndrome, neuroblastoma, and hemorrhaglc disease in a neonate.
J. Am. Med. Assoc. 244:1464-1465.
Anderson, H.A., R. Lilis, S.M. Daum, A. S. Fischbein, and I.J. Selikoff. 1976.
Household-contact asbestos: Neoplastic risk. Ann. N.Y. Acad. Sci. 271:311-323.
Barnes, A.B., T. Colton, J. Gundersen, K.L. Noller, B.C. Tilley, T. Strama,
D.E. Townsend, P. Hatab, and P.C. O'Brien. 1980. Fertility and outcome of
pregnancy in women exposed in utero to diethylstilbestrol. N. Engl. J. Med.
302:609-613.
Blot, W.J. , and R.W. Miller. 1973. Mental retardation following in utero exposure
to the atomic bombs of Hiroshima and Nagasaki. Radiology 106:617-619.
Chisolm, J.J., Jr. 1978. Fouling one's own nest. Pediatrics 62:614-617.
Crowe, M.W. 1969. Skeletal anomalies in pigs associated with tobacco. Mod. Vet.
Pract. 50:54-55.
Crowe, M.W., and T.W. Swerczek. 1974. Congenital arthrogryposis in offspring of
sows fed tobacco (Nicotiana tabacum). Am. J. Vet. Res. 35:1071-1073.
Goldstein, L. , and D.P. Murphy. 1929. Etiology of the ill health in children born
after maternal pelvic irradiation. Part II: Defective children born after post-
conception pelvic irradiation. Am. J. Roentgenol. 22:322-331.
Greenwald, P., J.J. Barlow, P.C. Nasca, and W.S. Burnett. 1971. Vaginal cancer
after maternal treatment with synthetic estrogens. N. Engl. J. Med. 285:390-392.
Hanson, J.W., and D.W. Smith. 1975. The fetal hydantoin syndrome. J. Pediatr.
87:285-290.
Herbst, A. L. 1980. A comparison of pregnancy experience in DES-exposed and
DES-unexposed daughters. J. Reprod. Med. 24:62-69.
Herbst, A.L., H. Ulfelder, and B.C. Poskanzer. 1971. Adenocarcinoma of the vagina.
N. Engl. J. Med. 284:878-881.
Herbst, A.L., D.C. Poskanzer, S.J. Robboy, L. Friedlander, and R.E. Scully. 1975.
Prenatal exposure to stilbestrol: A prospective comparison of exposed female
offspring with unexposed controls. N. Engl. J. Med. 292:334-339.
Herbst, A.L., P. Cole, T. Colton, S.J. Robboy, and R.E. Scully. 1977. Age-incidence
and risk of diethylstilbestrol-related clear cell adenocarcinoma of the vagina
and cervix. Am. J. Obstet. Gynecol. 128:43-50.
151
-------
Hoover, R. , and J.F. Fraumeni, Jr. 1975. Drugs. Pp. 185-198 in J.F. Fraumeni, Jr.,
ed. Persons at High Risk of Cancer: An Approach to Cancer Etiology and Control.
Academic Press, Inc., New York.
Kaufman, R.H., G.L. Binder, P.M. Gray, Jr., and E. Adam. 1977. Upper genital tract
changes associated with exposure in utero to diethylstilbestrol. Am. J. Obstet.
Gynecol. 128:51-59.
Kimbrough, R.D., C.D. Carter, J.A. Liddle, R.E. Cline, and P.E. Phillips. 1977.
Epidemiology and pathology of tetrachlorodibenzodioxin poisoning episode. Arch.
Environ. Health 32:77-86.
Li, F.P., J. Lokich, J. Lapey, W.B. Neptune, and E.W. Wilkins, Jr. 1978. Familial
mesothelioma after intense asbestos exposure at home. J. Am. Med. Assoc.
240:467.
Miller, R.W., and W.J. Blot. 1972. Small head size after in-utero exposure to
atomic radiation. Lancet 2:784-787.
Miller, R.W., and J.J. Mulvihill. 1976. Small head size after atomic irradiation.
Teratology 14:355-357.
Murphy, D.P. 1928. Ovarian irradiation, its effects on the health of subsequent
children: Review of the literature, experimental and clinical, with a report of
320 human pregnancies. Surg. Gynecol. Obstet. 47:201-215.
Neel, J.V., and W.J. Schull. 1956. The Effect of Exposure to the Atomic Bombs on
Pregnancy Termination in Hiroshima and Nagasaki. National Academy of Sciences,
Washington, D.C. 241 pp.
Neel, J.V., C. Satoh, H.B. Hamilton, M. Otake, K. Goriki, T. Kageoka, M. Fujita,
S. Neriishi, and J. Asakawa. 1980. Search for mutations affecting protein
structure in children of atomic bomb survivors: Preliminary report. Proc.
Natl. Acad. Sci. USA 77:4221-4225.
Selby, L.A., R.W. Menges, E.G. Houser, R.E. Flatt, and A.A. Case. 1971. Outbreak
of swine malformations associated with the wild black cherry, Prunus serotina.
Arch. Environ. Health 22:198-201.
Taki, I., S. Hisanaga, and Y. Amagase. 1969. Report of Yusho (chlorobiphenyls
poisoning): Pregnant women and their fetuses. Fukuoka Igaku Zasshi 60:471-474.
Ulfelder, H. 1980. The stilbestrol disorders in historical perspective.
Cancer 45:3008-3011.
152
-------
DISCUSSION
DR. REHDER: With respect to leukemia after radiation of newborn children,
do you think there are two different pathomechanisms of inducing tumors by
exposing fetuses to certain substances? In some cases, is there a direct
oncogenic effect as, perhaps, in the leukemias? In other cases, is there
dysgenesis that increases the risk of tumor?
DR. MILLER: Yes, leukemia may have, as its basis, a chromosomal abnormality
induced by an agent, such as radiation, and then a clonal evolution of the
leukemic cell takes place. The other mechanism (dysgenesis) seems
clear from studies of human embryos. Areas of malformed tissue, for
example, have been seen under the capsule of the kidney. In patients
withWilms' tumor of both kidneys (multifocal Wilms' tumor), dysplasia
regularly found under the capsule of the kidney is believed to be due to
a germinal mutation, and a second postgygotic event causes malignant
transformation. But that second event does not have to happen.
Yes, there appear to be different mechanisms. Immunosuppression may be one;
another may relate to enzymes in the embryo or the fetus. As I understand it,
enzymes develop late in fetal life. If enzyme activation is required for
carcinogenesis, exposure before then might not have an effect.
Cancers induced by environmental exposure of fetuses may not appear for
decades, as has happened in experimental studies with nitrosamines, for example.
In DBS-induced cancer, the cancers appear mostly in women from 14 to 29 years
of age. So, one may have to wait a long time to see what, if any, effect there
is.
153
-------
Reproductive Injury; General Considerations
Helga Render1
Reproductive injuries are frequent events. Approximately 20% of
all conceptions are spontaneously aborted, and 30% of all stillborn and
2% of all liveborn infants display major congenital malformations. It
is important to delineate disorders caused by exogenous agents from
those that are genetically determined or due to recognizable chromosomal
aberrations. Proving the association of exogenous disturbances with
known environmental contamination is difficult. Experimental background
information on the biologic effects of a compound is helpful, but the
results may not be absolutely valid when related to other species.
Mutations may not appear when exposed cells are susceptible to lethal
cytotoxic side effects. A teratogenic potency, which is claimed to be
nonspecific for any given substance, depending only on the sensitive
periods of organ development at the time of exposure, may seem to gain
specificity because fetal uptake of different substances occurs at
different developmental periods and there may be selective accumulation
in a single fetal tissue. The possibility of indirect fetal damage
also has to be considered for those compounds that cause lesions of the
yolk sac endoderm or of the placenta, rather than direct injury to the
embryo. Time and grade of exposure as well as the type and pattern of
developmental disturbance must be analyzed for each case in order to
evaluate environmental reproductive hazard.
The modes of detecting reproductive injuries caused by environmental
contamination are diverse. Attention may first be attracted by a sudden
and striking increase in the number of certain congenital anomalies
such as occurred near Minamata Bay in Japan from 1953 to 1960.
Methylmercury was later identified as the pollutant causing the birth
defects (Matsumoto et_ al. , 1965; Tatetsu and Harada, 1968). A similar
pattern of malformations led to the discovery of the reproductive
effects of thalidomide (Lenz and Knapp, 1962). In both of these
instances, reproductive injuries occurred, and the causes were traced.
Institute fur Pathologie der Medizinischen Hochschule Lu'beck, Lubeck,
Federal Republic of Germany.
154
-------
More difficult is the search for after-effects of a known accident,
such as the tetrachlorodibenzodioxin contamination at Seveso (Hay,
1976, 1977a,b; Rehder et^ al_. , 1978), or a correlation of an increased
number of animal deaths in areas of concentrated lead industry to some
effect on humans (Koch and Vahrenholt, 1978).
In the first type of study, a pattern of congenital defects seems
to follow a homogeneous pattern, and the effect of a pollutant on
reproduction seems obvious. For the second type, effects seem
heterogeneous, and any congenital defect is automatically ascribed
specifically to the polluting event. Assessment of abnormalities is
based mainly on epidemiologic data such as fertility pattern, spontaneous
abortion rate, frequency of malformations, neonatal and childhood
mortality, and even childhood malignancies within the exposed group.
Exact data, however, may be unavailable; after a polluting event, the
exposed population and physicians tend to overreport reproductive
failures. In normal times, underreporting is more common. According
to a study in Pennsylvania reported by Babbot and Ingalls (1962), only
60% of malformed liveborn babies were reported as malformed on their
birth certificates. Furthermore, exact control data are often missing.
Centralized monitoring of birth defects has only just started in a few
countries (Ericson et_ al_. , 1977; Klingberg and Papier, 1979).
In addition, too little consideration is given to the fact that
prenatal survival is itself the outcome of natural selection.
Developmental failures are frequent events, occurring spontaneously in
any population and area and without recognizable environmental hazards.
155
-------
In humans, at least 20% of all conceptions are spontaneously aborted
(Tuchmann-Duplessis, 1975) and, although only few systematic
embryopathological investigations have been performed, it is apparent
that the number and variety of anomalies during gestation is far greater
than can be recognized from an analysis of term births (Stein et al.,
1975). Even animals have a constant spontaneous abortion rate. In the
mouse, it is 5% to 8%, the variation dependent on whether the animal is
born in the wild or in the laboratory (Berry and Peters, 1976).
Anomalies are probably caused by a number of factors such as
hormonal changes, intrafollicular aging of the ovum, delayed fertilization,
unfavorable uterine environments, infections, other maternal disorders,
immunologic incompatibilities, spontaneous genetic mutations, and—last
but not least—chromosomal abnormalities. Extensive cytogenetic studies
of almost 1,500 clinical cases of spontaneous abortions between the 3rd
and 12th week showed chromosomal abnormalities in 60% of the specimens
younger than 7 weeks. Whereas, in abortions between the 8th and 12th
gestational week, the frequency was 23% (Boue and Boue, 1976). My own
experiences indicate that the incidence of chromosomal aberrations is
also very high in abortions occurring in late pregnancy. Among newborns,
the rate is only 0.5% (Hamerton et al. , 1975).
-------
Minor anomalies are found in 10% to 15% of newborns (Harden et al.,
1964). These numbers may vary among different obstetric and neonatal
units and among different areas, as was demonstrated most impressively
by the 40-fold difference found in the frequency of anencephaly between
the highest in Belfast and the lowest in Bogota and Ljubljana (Stevenson
et_ al., 1966).
To evaluate environmental reproductive hazards, it is important to
differentiate the cases with reproductive injury caused only by exogenous
influences from those caused by endogenous disorders. This requires
the ability and experience to recognize, for example, genetic disorders
in the fetus. However, since a chemical agent may become effective via
endogenous disturbances, experimental background information on a
compound's biologic effects is necessary.
A chemical agent may act as mutagen, teratogen, and fetotoxin.
A mutagen influences reproduction by affecting germ cells; it may
express itself as a point or gene mutation, as a chromosome mutation,
and as a genome mutation. Single-point or gene mutations generally
have no effect on the individual and are seen as normal polymorphisms.
They will not be recognized as long as they represent recessive gene
mutants. When these genes accumulate within a population, they may
eventually appear in the homozygous state in later generations. Some
of the dominant gene mutants, however, may result in congenital diseases
of fetuses. These are known as genetic diseases and are recognizable
as such in many cases. More than 1,200 autosomal dominant and 900 autosomal
recessive, and approximately 170 X-linked genetic disorders are recognized
(McKusick, 1975). When point mutations have accumulated within a fertilizing
157
-------
germ cell, they exhibit a lethal effect on the heterozygous zygote. It
has to be mentioned in this context that the "spontaneous mutation rate"
is very high, varying between 2 to 3 per million gametes for hemophilia
B and 100 per million gametes for neurofibromatosis (Vogel, 1970).
Chromosome mutations affect the actual structure in the form of
deletions, duplications, inversions, and translocations following
chromosome breaks. When genetically imbalanced, they may kill the
gamete or the early embryo. Cytogenetic studies show chromosome
mutations in 10-15% of second trimester abortions (Ruzicska e£ al.,
1970); among newborns, only 2% carry a structural, mainly balanced
aberration (Hamerton ^t al., 1975).
Genome mutations change the chromosome number. They result from
extrachromosomal disturbances, either from meiotic nondisjunction when
single chromosomes are lost or triplicated or from interference with
cell cleavage in cases of polyploidy. Most genome mutations are lethal.
Very few affected fetuses (such as those exhibiting trisomy 21, 18, or
13) survive until birth.
There are a number of in vivo and in vitro test systems for
mutagenicity using microorganisms or somatic or germ cells of plants,
insects, and mammals. The most relevant one for humans is the dominant
lethal test. Thus, it was shown that metals such as mercury, lead,
and cadmium do not induce point or chromosome mutations in the mammalian
germ cell test (Gebhart, 1977). These metals interfere with chromosomal
segregation during mitosis in human leukocytes (Fiskesjo, 1970) and in
Drosophila (Ramel e£ _al_. , 1969). However, these results cannot very
well be extrapolated to meiosis since none of these substances has, up
158
-------
to now, shown a lethal effect in the dominant lethal test—an effect
that would be expected from the genome mutation in germ cells.
Chlorinated carbohydrates may be another example of compounds that
do not show mutagenic effects in the dominant lethal test system.
Nonetheless, wives whose husbands were exposed to vinyl chloride monomers
at their workplace have an increased abortion rate (Infante et al.,
1976). Abortion rates in wives of unexposed workers at the same plant
were within normal limits, so the increase in the fetal wastage is
attributed to mutation of male germ cells.
Although one may assume that many chemical agents have a mutagenic
effect on human reproduction and that at least some of the so-called
spontaneous mutations are actually exogenously induced, possibly by
chemical compounds. There has been no evidence of an increase of
genetic or chromosomal diseases in areas of detectable chemical
contamination. Such findings may be obscured by a high dominant lethal
effect or to a high spontaneous abortion rate. Unfortunately, there
have been no thorough investigations of abortion material for confirmation.
Furthermore, it has to be considered that mutagenic effects may
not be expressed when exposed cells undergo regression. This finding
became evident among pesticide workers who developed oligospermia or
aspermia when they were exposed to dibromochloropropane (DBCP) (Biava
ei£ a!L., 1978; Whorton and Milby, 1980). Investigators used x-rays and
gamma rays to show that spermatogonia, representing reproductive stem
cells, as well as dictyotene oocytes, are far more susceptible to
159
-------
cytotoxic rather than to mutagenic effects. They also showed that it
is easier to induce mutations in more differentiated germ cells (such
as spermatocytes, spermatids, and in fertilized ova during the pronucleus
phase, that is, before fusion of the two nuclei) (Vogel et_ al^. , 1969).
The conclusion is that, if there is a mutagenic effect on germ cells,
it is temporary, resulting mainly in a reduction of fertility.
Mutagenic effects on the embryo itself (on embryonic somatic cells)
should result in genetic mosaicism, relevant only when occurring in
very early embryonal stages. Again, it was shown that in these, the
cytotoxic effect predominates (Brent, 1979; Russel, 1965) and that
preimplantation embryos respond to cytotoxicity with an "all or nothing
effect," either being killed or progressing into an apparently normal
individual. Possible tumor induction by mutation of somatic cells in
older fetuses will not be discussed in this paper. The lack of more
differentiated sensible germ cell stages in the fetus limits the possible
mutagenic effect on fetal germ cells.
Teratogenic effects of a chemical compound on human reproduction are
relatively easy to detect. "Teratogenesis" refers to impairment of
embryonal organogenesis within a certain early developmental period;
impairment of growth and differentiation or maturation in later develop-
mental periods is a fetotoxic effect. This distinction is important
with respect to the higher sensitivity and regenerative potency of the
undifferentiated embryonal cell and thus to the type of the resulting
anomaly.
160
-------
Teratogenic compounds, if exerting direct action on the embryonal
tissue, interfere with embryonal metabolism. They may cause nonspecific
alterations, such as retarded development of the primordia by arresting
or retarding mitotic activity, by killing cells, by inhibiting cell
migration or cell interaction, or by disturbing inductional processes
(Nishimura and Tanimura, 1976). Resulting organ anomalies are generally
present as "defect malformations", but they also appear as "augmentation
malformations" through overcompensation of early regenerative processes
(Merker, 1977). The type of developmental defect does not reflect the
exact temporal onset of the fetal injury, which generally occurs earlier
than the developmental process, which is disturbed. This delayed
effect is due to the attempted regenerative proliferation of the sur-
rounding primordial cells that precedes the development breakdown and
that may even fully compensate for the defect.
Despite their different biochemical pathways, different chemical
agents can cause the same types of malformations if exposure occurs
with the same sensitive period of organogenesis. However, some types
of defects occur more frequently than others. Neural tube defects or
cleft palate can be induced experimentally in almost any animal species
by many different substances. These defects are the first to attract
attention in areas of suspected chemical contamination such as occurred
in Ohio in 1975, where the occurrence of neural tube defects was attri-
buted to the presence of the polyvinylchloride industry (Edmonds et
al., 1975). In New Zealand and Sweden, the same defects were ascribed
to the spraying of forests with 2,4,5-trichlorophenol (2,4,5-T) (Divi-
sion of Public Health, New Zealand, 1977; Swedish National Board of
161
-------
Health and Welfare, 1977). A rise in the incidence of cleft palate
was also correlated to the use of 2,4,5-T (Nelson jit _al. , 1979).
Since palatine shelf and neural tube closure obviously represent
extremely sensitive developmental processes, neural tube defects and
cleft palate may also result from unspecific disturbances. This hypo-
thesis is supported by the high incidence of these malformations in
discordant homozygous twins (Beliefeuille, 1969) and by their associa-
tion with many chromosomal syndromes, where the malformations are not
directly genetically determined but result from a general retardation
of cell proliferation.
A chemical agent may gain some teratogenic specificity by strongly
and selectively accumulating in single fetal tissues. Thus, experimentally
verified localization of tetracycline in fetal bone structures (Andre,
1956; Blomquist and Hangren, 1966), of thiouracil in the thyroid (Slanina
et^ al. , 1973), and of some polycyclic compounds (chloroquine and
chlorpromazine) in the melanin of the eye and in the inner ear (Dencker
and Lindquist, 1975; Denecker &t al., 1975; Lindquist and Ullberg,
1972) can be correlated to reported organ damage in human fetuses after
application of these substances to pregnant mothers. However, the main
sites of accumulation are the placenta and the yolk sac. Fetal uptake
of the different compounds occurs from these sources at different
developmental stages, thus limiting and specifying their teratogenic
effect.
An earlier placenta! barrier and an accelerated fetal uptake with
advancing gestation has been shown in mouse and hamster for 2 4 5-T
(Dencker, 1976). This finding implies an effect on the fetus only in
162
-------
the late organogenic phase and explains the experimental induction of
cleft palate or hydronephrosis. Neural tube defects would not, however,
fall into that category because they represent disturbances during
early organogenesis.
In contrast to 2,4,5-T, cadmium acts only in the early organogenic
period. It does not cross the placenta at all. Apparently, it enters
the fetal gut via the yolk sac just before closure of the vitelline
duct. Its teratogenic effect then decreases rapidly (Denecker, 1975).
The experimental finding of cleft palate as a late organogenic disturbance
resulting from cadmium application is explained by the intimate contact
of the stomodeum and foregut before the formation of the oral and
pharyngeal cavities.
Salicylic acid has free passage to the fetus throughout gestation,
and it is taken up to a similar degree at all stages. Thus, it may
cause a wide range of malformations and fetotoxic effects. The structure
most severely damaged will be the one undergoing most rapid development
at the time of administration (Kimmel £t al., 1971). Benzole acid,
widely used as a food preservative, increases the embryonic concentration
of salicylic acid.
Embryonic uptake of inorganic mercury varies in relation to
placental development. In early gestation, the substance may enter the
fetal gut via the yolk sac (like cadmium), but it exhibits less
teratogenic activity. Closure of the vitelline duct temporarily arrests
the passage of mercury, but increasing fetal concentrations are again
observed in late gestation (Dencker, 1976). The more lipid-soluble
163
-------
organic mercury compounds, such as methylmercury, are much more readily
transferred to the fetus and are teratogenic during the entire organogenic
period.
A chemical compound can also act on organogenesis if the primary
target is not the embryonic cell but, rather, certain maternal tissues
or the placenta. Inhibited embryotrophic nutrition after chemical
compounds have accumulated within the placenta may, secondarily, lead
to fetal growth retardation or even to fetal malformations. This
pathomechanism has been proposed for the action of trypan blue on the
fetus (Beck e£ a\_. , 1967). Furthermore, researchers have shown that,
after application of methylmercury, a distinct decrease of fetal amino
acids can be related, at least in part, to impaired placental transfer
(Olson and Massaro, 1978).
Although the teratogenic effects of a compound are confined to the
period of organogenesis (leading to a reduction of still undifferentiated
primordial cells and, thus, to a true malformation), fetotoxic damage
in later gestation is characterized by a disturbance in body and organ
growth and by destruction of more or less differentiated tissues.
These alterations are followed by hypotrophy and by a change in the
distribution of parenchymal versus connective (mesenchymal or glial)
tissue, which results in a reduction of organ function and—in more
extended tissue defects—in either secondary cyst formation or
scarring. The fetal brain seems to be the most susceptible to toxic
damage and may display a wide range of congenital alterations.
Microencephaly and microgyria, due to a narrowing of the cerebral and
cerebellar cortex and to nerve cell necrosis, have been described in
164
-------
congenital Minamata disease (Matsumoto et^ _al_., 1965). Experimental
lead intoxication of pregnant rats and chick embryos resulted in brain
hemorrhage and in hydrocephaly due to obstruction of the aqueduct
(Ridgway and Karnofsky, 1952). In some instances, cystic porencephaly,
following more extended destructive brain lesions in midterm fetuses
(when the reactivity of the glial tissue is still limited), may be
related to exogenous disturbances. Hydranencephaly or even ulegyria,
occurring in older fetuses when reactive gliosis responds to the tissue
damage, may also be related to exogenous disturbances. Extracerebral
fetotoxic lesions can be observed in the lens of the eye (Brent, 1979),
in the lymphoid and hematopoietic tissue (Driscoll et al., 1963), and—
mainly by the way of circulatory disturbances—in the liver, heart,
and lungs (own observations), where parenchymal defects are replaced by
mesenchymal proliferations leading to tissue fibrosis and hamartomatosis.
Again, one has to consider placental accumulation of a chemical
compound to be responsible, at least partly, for the fetotoxic effects
such as hypotrophy and fetal death, even in those cases in which the
fetotoxic agent can be traced in placental as well as in fetal tissues.
This rough diagram of possible pathomechamisms of mutagenic,
teratogenic, and fetotoxic compounds is not complete. Rather, it is
meant to focus on difficulties of their recognition and to emphasize
the many different aspects that must be considered before reproductive
injuries can be related to a specific environmental contamination.
Epidemiologic data and experimental background information on the
165
-------
biologic effects of chemical compounds is helpful, but are often
insufficient and not of absolute validity when extrapolated to humans
or to conditions outside the laboratory.
In fact, there has been little evidence of reproductive injury
resulting from chemical contaminants mainly because of the lack of
systematic cytogenetic, toxicologic, and pathoanatomical investigations
of fetal material in humans. It should be emphasized that morphologic
and cytogenetic analyses, even of early fetuses, are possible and are
important. Only when pathomechanisms and effects of certain compounds
on human reproduction have been analyzed can one develop preventive
efforts, which may even include the use of selective antimutagenic and
antiteratogenic substances.
This work was supported by the Deutsche Forschungsgemeinschaft Re 429/3.
166
-------
REFERENCES
Andre, T. 1956. Studies on the distribution of tritium-labelled
dihydrostreptomycin and tetracycline in the body. Acta Radiol.
(Diagn.) Suppl. 142:1-89.
Babbot, J.G., and T.H. Ingalls. 1962. Field studies of selected congenital
malformations occurring in Pennsylvania. Am. J. Public Health 52:2009-
2017.
Beck, F., J.B. Lloyd, and A. Griffiths. 1967. Lysosomal enzyme inhibition
by trypan blue: A theory of teratogenesis. Science 157:1180-1182.
Berry, R.J., and J. Peters. 1976. Genes, survival and adjustment in an
island population of the housemouse. Pp. 23-48 in S. Karlin and E.
Nevo, eds. Population Genetics and Ecology. Academic Press, New
York.
Biava, C.G., E.A. Smuckler, and D. Whorton. 1978. The testicular
morphology of individuals exposed to dibromochloropropane. Exp. Mol.
Pathol. 29:448-458.
Blomquist, L., and A. Hanngren. 1966. Flourescence technique applied to
whole body section for distribution studies of tetracyclines.
Pharmacology 15:215-219.
Boue, J.G., and A. Boue. 1976. Chromosomal anomalies in early spontaneous
abortion. Pp. 193-208 in A. Gropp and K. Benirschke, eds. Current
Topics in Pathology. Vol. 62, Springer, New York.
Brent, R.L. 1979. Effects of ionizing radiation on growth and development.
Contrib. Epidem. Biostatist. 1:147-183.
deBellefeuille, P. 1969. Contribution a 1'etiologie de 1'anencephalie
par l'e"tude des gumeaux. Union Med. Can. 98:437-443.
Dencker, L. 1975. Possible mechanisms of cadmium fetotoxicity in golden
hamsters and mice: Uptake by the embryo, placenta and ovary. J.
Reprod. Fertil. 44:461-471.
Dencker, L. 1976. Tissue localization of some teratogens at early and
late gestation related to fetal effects. Acta Pharmacol. Toxicol. 39
(Suppl. I):5-131.
Dencker, L., and N.G. Lindquist. 1975. Distribution of labelled
chloroquine in the inner ear. Arch. Otolaryngol. 101:185-188.
167
-------
Dencker, L. , N.G. Lindquist, and S. Ullberg. 1975. Distribution of an
^^I-labelled chloroqulne analogue in a pregnant Macaca monkey.
Toxicology 5:255-264.
Division of Public Health, Department of Health, New Zealand. June
1977. 2,4,5-T and human birth defects: A report. Division of Public
Health, New Zealand, pp. 2.
Driscoll, S.G., S.P. Hicks, E.H. Copenhaver, and C.L. Easterday. 1963.
Acute radiation injury in two human fetuses. Arch. Path. (Chicago)
76:113-119.
Edmonds, L.D., H. Falk, and J.E. Nissim. 1975. Congenital malformations
and vinyl chloride. Lancet ii:1098.
Ericson, A., B. Kallen, and J. Winberg. 1977. Surveillance of malforma-
tions at birth: A comparison of two record systems run in parallel.
Int. J. Epidemiol. 6:35-41.
Fiskesjo, G. 1970. The effect of two organic mercury compounds on human
leukocytes in vitro. Hereditas 64:142-146.
Gebhart, E. 1977. Pp. 1-192 in Chemische Mutagenese. G. Fischer,
Stuttgart-New York.
Hamerton, J.L., N. Canning, M. Ray, and S. Smith. 1975. A cytogenetic
survey of 14,069 newborn infants. I, Incidence of chromosome abnormalities.
Clin. Genet. 8:223-243.
Hay; A. 1976. Toxic cloud over Seveso. Nature 262:636-638.
Hay, A. 1977a. Seveso: The aftermath. Nature 263:538-540.
Hay, A. 1977b. Seveso solicitude. Nature 267:384-385.
Infante, P.F. , J.K. Wagoner, A.J. McMichael, R.J. Waxweiler, and H.
Falk. 1976. Genetic risks of vinyl chloride. Lancet 1:734-735.
Kimmel, C.A., J.G. Wilson, and H.J. Schumacher. 1971. Studies on metabolism
and identification of the causative agent in aspirin teratogenesis
in rats. Teratology 4:15-24.
Klingberg, M.A., and C.M. Papier. 1979. Teratoepidemiology. J. Biosoc.
Sci. 11:233-258.
Koch, E.R., and F. Varenholt. 1978. Teuflisches Blei. Pp. 132-138 in
E.R. Koch andj(F. Vahrenholt, eds. Seveso 1st Uberall. Kiepenheuer
und Wit sen, Ko'ln.
Lenz, W. , and K. Knapp. 1962. Thalidomide embryopathy. Arch. Environ.
Health 5:100-105.
168
-------
Lindquist, N.G., and S. Ullberg. 1972. The melanin affinity of
chloroquine and chlorpromazine studied by whole body autoradiography.
Acta Pharmacol. 31 (Suppl. 2): 1-32.
Harden, P.M., D.W. Smith, and M.J. McDonald. 1964. Congenital anomalies
in the newborn infant including minor variations. A study of 4,412
babies by surface examination for anomalies and buccal smear for sex
chromatin. J. Pediatr. 64:357.
Matsumoto, H., G. Koya, and T. Takeuchi. 1965. Fetal Minamata disease.
A neuropathological study of two cases of intrauterine intoxication
by a methyl mercury compound. J. Neuropathol. Exp. Neurol. 24:563-573.
McKusick, V.A. 1975. Pp. 1-837 in Mendelian Inheritance in Man.
Johns Hopkins University Press, Baltimore-London.
Merker, H.J. 1977. Considerations on the problem of critical period
during the development of limb skeleton. Pp. 179-202 in D. Bergsma
and W. Lenz, eds. Morphogenesis and Malformation of the Limb. No. 1
in Birth Defects: Original Article Series, Vol. XIII. Alan R.
Liss, Inc., New York.
Nelson, C.J., J.F. Holson, H.G. Green, and D.W. Gaylor. 1979. Retrospective
study of the relationship between agricultural use of 2,4,5-T and
cleft palate occurrence in Arkansas. Teratology 19:377-383.
Nishimura, H., and T. Tanimura. 1976. General knowledge on teratogenicity
of drugs. Pp. 73-92 in H. Nishimura and T. Tanimura, eds. Clinical
Aspects of the Teratogenicity of Drugs. Excerpta Medica and American
Elsevier Publishing Company, Inc., Amsterdam-Oxford-New York.
Olson, F.C. , and E.J. Massaro. 1978. Effects of methyl mercury on murine
fetal amino acid uptake, protein synthesis and palate closure.
Teratology 16:187-194.
Ramel, C., and J. Magnusson. 1969. Genetic effects of organic mercury
compounds. II, Chromosome segregation in Drosophila melanogaster.
Hereditas 61:231-254.
Rehder, H., L. Sanchioni, F. Cefis, and A. Gropp. 1978. Pathologisch-
embryologische Untersuchungen an Abortusfallen im Zusammenhang mit
dem Seveso-Ungluck. Schweiz. Med. Wochensch. 108:1617-1625.
Ridgway, L.P-, and D.A. Karnofsky. 1952. The effect of metal on the
chick embryo: Toxicity and production of abnormalities in development.
Ann. N.Y. Acad. Sci. 55: 203-215.
Russell, L.B. 1965. Death and chromosome damage from irradiation of
preimplantation stages. Pp. 217ff in G.E.W. Wolstenholme and M.
O'Connor, eds. CIBA Foundation Symposium on Preimplanted Stages of
Pregnancy.
Ruziczka, P., and Z. Czeizel. 1970. Cytogenetic studies on mid-trimester
abortuses. Hum. Genet. 10:273-297.
169
-------
Slanina, P-, S. Ullberg, and L. Hammarstrom. 1973. Distribution and
placental transfer of C-thiourea and C-thiouracil in mice studied
by whole-body autoradiography. Acta Pharmacol. 32:358-368.
Stein, Z., M. Susser, D. Warburton, J. Wittes, and J. Kline. 1975.
Spontaneous abortion as a screening device. The effect of fetal
survival on the incidence of birth defects. Am. J. Epidemiol. 102:275-
290.
Stevenson, A.C., H.A. Johnston, M.I.P. Stewart, and D.R. Golding. 1966.
Congenital malformations: A report of a study of series of consecutive
births in 24 centres. Bull. WHO 34 (Suppl.)l.
Swedish National Board of Health and Welfare. 1977. Pp. 1-11 in Congeni-
tal Malformations and Perinatal Death in a Swedish County: A Study
of the Relation to Exposure to 2,4,5-T. Socialstyrelsen, Byra.
Tatetsu, S., and M. Harada. 1968. Mental deficiency resulting from
intoxication in the prenatal period. Adv. Neurol. Sci. (Tokyo) 12:181-
190.
Tuchmann-Duplessis, H. 1975. Pp. 1-267 in Drug Effects in the Fetus.
Adis Press, New York.
Vogel, F. 1970. Spontaneous mutation in man. Pp. 16-68 in F. Vogel and
G. Rohrborn, eds. Chemical Mutagenesis in Mammals and Man. Springer,
Berlin-Heidelberg-New York.
Vogel, F., G. Rohrborn, E. Schleiermacher, and T.M. Schroder, eds.
1969. Pp. 1-80 in Strahlengenetik der Sa'uger. Thieme, Stuttgart.
Whorton, M.D. , and T. H. Milby. 1980. Recovery of testicular function
among DBCP workers. J. Occup. Med. 22:177-179.
170
-------
DISCUSSION
DR. MILLER: In Minamata disease, is the effect on the brain
fetotoxic or teratogenic?
DR. REHDER: It is mainly fetotoxic. However, it is difficult to
say because brain development takes a long time, and there may be a
disturbance in nerve cell migration even in later gestation. Nerve
cell death occurs in the brain with Minamata disease. Disturbances of
nerve cell migration that would result in nerve cell heterotopias are
not actually seen.
DR. MILLER: In radiation injury we know that the susceptible time
during gestation is during the period of organogenesis, but the effect
is said to be due to cell depletion in the brain. Is that fetotoxic or
teratogenic?
DR. REHDER: The early cell depletion is teratogenic, but cell
depletion in the more differentiated brain has to be called fetotoxic,
even if the result seems to be the same. In one it is hypoplasia, and
in the other one it is hypotrophy.
DR. MURPHY: Please elaborate on reports from New Zealand and
Arkansas of cleft palate associated with 2,4,5-T exposure.
DR. REHDER: At a conference in Washington last year on 2,4,5-T
and dioxin, it was said that the findings are not sufficient to implicate
a correlation. Palatine shelf closure disturbances and neural tube
closure disturbances are easily diagnosed, and people jump to the
conclusion that these are effects of pollution.
171
-------
DR. MILLER: A cluster of women in a certain industry have had
miscarriages. What do you do with a miscarried fetus to make a proper
cytogenetic examination? Do you put it directly into formaldehyde?
DR. REHDER: No. Laboratory specialists and gynecologists should
be informed and should have all the media they need to preserve tissues
and to guarantee that these investigations are done. The studies are
very important. People always say the abortion rate has risen due to
one or another pollutant, but nobody investigates.
DR. MILLER: What is the simplest procedure to ensure that a fetus
is in a state to have its chromosomes examined?
DR. REHDER: Fetuses in spontaneous abortions are often macerated,
and it makes no sense to take fetal tissue if you have autolytic changes
already. The placenta, however, is always preserved — if it is still
attached to the uterine wall. Amniotic membranes (which you can easily
identify) do grow very well, even if the fetus is macerated. You need
just a small amount of medium; insert the specimen, and send it to the
cytogeneticist. You should also preserve the fetus, of course, for
pathoanatomic study.
DR. SUSKIND: Would you comment on the need to be rather careful
about how one interprets both fetotoxicity or teratogenicity in relation
to the nutritional environment of experimental animals? For example,
there is a problem of maintaining an adequate zinc and copper intake,
especially with respect to cadmium teratogenicity or fetotoxicity, and
even with some other teratogenic agents, such as salicylic acid or the
salicylates.
172
-------
DR. REHDER: Effects other than direct toxlcity on the embryos are
very important and play a great role. The potential interaction of
elements is very important and should be evaluated, perhaps by studying
the placenta. Morphologic study of the placenta is a problem, but you
will gain experience if you just start to investigate.
DR. MURPHY: Presently there is a great deal of concern among
Vietnam veterans as to possible malformations in their children. What
is the likelihood of such occurrences long after the exposure period?
Might continued testicular change persist and appear in terata some
years later?
DR. REHDER: The effect of the alterations results more in the
reduction of the fertility rate than in an increase of genetic disease.
This finding can be explained by the high cytotoxic effect versus the
low susceptibility of germ stem cells to mutagenic effects. The mutation
in the offspring occurs in the more differentiated germ cells, which
are eliminated after some years. So, the risk decreases.
DR. MURPHY: Is the germinal stem cell generally more susceptible
to death than to mutation by chemical agents?
DR. REHDER: The cytotoxic effect predominates; the cell degenerates
and does exhibit a mutagenic effect. It does not affect the offspring.
173
-------
Birth Defects Register in Seveso: A TCDD-Polluted Area
by E. Marni, L. Blsanti, L. Abate, C. Borgna-Pignatti,
G. Maggiore, P- Bruzzi, and E. Montesarchio
A Birth Defects Registry was established in 1978 to screen all infants
born to women living near the ICMESA (Industrie Chimiche Meda Societa
Anonima) plant where a reaction vessel containing tetrachlorodibenzodioxin
(TCDD) exploded in July 1976. All live and stillbirths are reported to
the registry by various private and public health sources. A three-level
screening program determines the presence of birth defects and their
severity. Children are followed through age 6. Although case numbers
are small and control groups are difficult to establish, birth defects
are being observed more frequently in children born south of the ICMESA
plant than elsewhere in Italy. This finding, however, may be due to the
more stringent reporting procedures in this region. In addition, many
children born in 1972 and used as controls are still classified as
"unknown"; followup of this group has proved difficult.
Numerous experimental works on animals (Courtney and Moore, 1971;
Greig e_t al. , 1973; Murray et_ a^. , 1977; Neubert e£ al., 1973; Sparschu
et al., 1971) have demonstrated a cause-effect relationship between
2,3,7,8-tetrachlorodibenzodioxin (TCDD) intake by pregnant females and
birth defects in their litters (Table 1). Human exposure, mostly of
males, during chemical manufacture or as a consequence of industrial
accidents, has been described (Bleiberg et al., 1964; Carter et al.,
1975; Goldman, 1972; International Agency for Research on Cancer, 1978;
Jirasek _et_ al. , 1973, 1974; Klmbrough _et_ al. , 1977; Poland £t al. , 1971;
Sparschu e_£ al. , 1971; Thiess and Frentzel-Beyne, 1978) on at least 20
occasions (besides the Vietnam war).
Drs. Marni, Abate, Borgna-Pignatti, and Maggiore are from the Pediatric
Department of the University of Pavia. Drs. Bisanti and Montesarchio
are from the Special Bureau for Seveso of the Regione Lombardia.
Dr. Bruzzi is from the Institute for the Study and the Therapy of Tumors
in Genoa.
174
-------
TABLE 1
Main Toxic Effects Reported in Litters of Female Animals
Treated with TCDD During Pregnancy
Animal
Malformation
Other Problems
Mouse
(Courtney and Moore,
1971; Neubert et_ al.,
1973)
Rat
(Murray et^ al. , 1977;
Sparschu et^ al. , 1971)
Hamster
(Neubert et al. , 1973)
Cleft palate
Kidney abnormalities
Kidney abnormalities
Eye Abnormalities
Hemorrhages
Edema
Interference with
development of
lymphatic system
Fatty infiltration
of the liver
Intestinal hemorrhages
Decreased weight
Reduced survival
Edema
Reduction of fetal
weight
Hemorrhages
Prenatal mortality
175
-------
The information gathered on malformed infants born in areas contaminated
with TCDD is very scanty (Table 2) and poorly documented (Advisory
Committee on 2,4,5-T, 1971; Cutting et_ ail. , 1970; McQueen et_ aJ. , 1977;
Meselson et al., 1971; Rose and Rose, 1972; Tung e£ al. , 1971).
On July 10, 1976, a reactor at the ICMESA plant in the town of
o
Seveso, Italy, exploded and an 18 km area was exposed to the chemical
contaminant dioxin. The monitored area includes 11 municipalities and
has a population of 219,358 inhabitants (Table 3).
Italian law requires that public health officials be informed of the
birth of any "deformed infant" and of the presence of congenital anomalies
of the locomotor system in individuals of any age. These laws, which date
back to 1941 and are concerned only with gross deformities, particularly
of limbs, clearly appear insufficient for detection of birth defects in
normal conditions and are even more inadequate for the present situation.
To overcome these limitations, a "Birth Defects Registry" (BDR) has been
devised for the area involved in the chemical contamination.
The registry was started in the fall of 1978, with retrospective and
prospective studies. It is based on the screening of infants born to
women who were living in the area around the ICMESA plant during July
1976.
Figure 1 shows the organization of the BDR. The registry derives
information from area hospitals, local health authorities, and pediatric
health services. The registry program has two levels of screening; a
third level is added when necessary.
176
-------
TABLE 2
Human Malformations Reported in Populations Exposed to TCDD
Country
Malformation
Vietnam
(Rose and Rose. 1972;
Tung et al., 1971)
Trimosy 21
Spina bifida
Cleft palate
New Zealand
(McQueen et al., 1972)
Anaencephalia
Spina bifida
177
-------
TABLE 3
Population and Area Surface Involved in the Polluting Event
00
Municipality
Meda
Seveso
Cesano M.
Desio
Barlassina
Bovisio
Seregno
Lentate
Varedo
Muggio
Nova M.
Total
population
19,668
16,958
33,132
33,021
5,635
11,170
36,894
13,140
12,014
18,827
18,899
Surface Area Zone R Zone B Zone A
X 10,000 sqm pop.% sur.% pop.% sur.% pop.% sur.%
834 20.4 18.3 a a 0.3 2.1
734 46.8 37.5 3.7 5.7 4.0 9.5
1146 45.1 41.6 8.2 10.2
1479 13.9 29.0 4.1 7.4
287 1.3 3.5
493 1.5 17.4
1301
1399
484
547
581
Total
219,358
aZone B is not present in Meda.
9285
-------
PEDIATRIC HEALTH SERVICES
F
I
R
S
T
L
E
V
E
L
SECOND
LEVEL
THIRD
LEVEL
REGISTER
REGISTRY
OFFICES
CONSULTANT
ORGANIZATION OF
BIRTH DEFECTS
R EG IS TE R
FIGURE 1. Organizational model of the birth defects register.
179
-------
First-level screening includes the following procedures:
- Monthly, registry offices in the 11 towns send a list of new births to
the BDR.
- Local health authorities notify the BDR of any reported malformations.
- Area hospitals continually report all new births, any infants admitted
with malformations, and any suspected to be carriers of birth
defects.
- Pediatric health community services offices send the BDR a list of children
checked by a pediatrician and identify those children suspected of being
malformed.
- Public health service agencies report any malformation observed while
performing physical examinations on children receiving their antipolio
vaccination.
- Starting with birthyear 1972, children born before the ICMESA accident
are examined when they reach age 6 in the first grade of the elementary
school. These data will be compared with information collected on
children born after 1976.
To minimize the number of false negatives, largely inclusive criteria are
used at this first stage.
The BDR staff collects all the information and unifies multiple reports on
the same child. From the list of newborns, it identifies any children who have
not been examined. Furthermore, it invites families of unexamined children to
a public health clinic near their homes (home visits arranged when indicated)
and visits all the children reported to be malformed.
At the end of this first screening, the BDR guarantees complete examinations
of all children on the basic list, the preparation of an inquiry list of any children
suspected to be malformed, and the recall of children for second-level examinations.
During second-level screening, the BDR's pediatric staff examines the
children reported as malformed in the pediatric health clinics or at their homes.
180
-------
Medical data are recorded as is any possible parental exposure to
teratogenic agents. These second-level examinations provide the
following conclusions:
Positive: Child diagnosed as being affected by simple or multiple
malformations.
Dubious; Additional diagnostic tests or reexaminatlon required,
or the clinical picture is difficult to interpret.
Negative; Child free of malformation.
The coordinator of the pediatric staff then lists the positive
cases in the registry.
Medical and town records are reviewed and, if necessary, parents
are interviewed to establish the cause of death of infants stillborn or
dead shortly after birth and to ascertain whether malformations were
present.
At the third level, the coordinator of the pediatric staff refers
cases with dubious diagnoses to a consultant, whose diagnosis is accepted
as definitive.
A cohort study is performed on all newborns and is continued up to
age 6 to identify those malformations that only become apparent later.
All children are contacted by the registry within the first year of
life, between the ages of 1 and 3, and at age 6 when they begin school.
The data presented, updated through September 20, 1979, represent
the results of the first year of BDR activity and should be considered
preliminary. The children studied were born in 1972, 1976, 1977, 1978,
181
-------
and part of 1979. Table 4 shows the number of live births and malformed
infants reported. Table 5 shows the results of the second-level check
for the years 1976, 1977, and 1978. The numbers of live births and
positive and dubious cases are indicated.
Cases are classified as follows:
Unknown: Subjects for whom no medical information is available.
Their percentage varies in different areas and at different times.
Furthermore, these subjects may even have a different probability of
being malformed.
Known: Children for whom medical information has been supplied by
at least one of the sources considered by the registry.
Reported: Subjects suspected of carrying a birth defect.
Reviewed: Children who underwent second-level examination by BDR
pediatric staff.
Known, reported, and reviewed subjects include infants stillborn
or dead shortly after birth, if the existence of malformations was
definitely reported (or excluded) and reviewed.
The secular trend has not yet been analyzed because of the small
number of years so far.
Trend analysis of data from elementary school and neonatal screenings
(birthyears 1972 versus 1978) will be kept distinct because screening
design, observers, and age of population studied are different in the two
groups. Unified analysis will be possible for those malformations
that can be diagnosed independently of age and observer if the unknown
subjects comprise less than 10 percent of the groups.
182
-------
TABLE 4
Malformed Infants Reported to the Public Health Authority
in the Seveso Area Between 1973 and 1978
Malformed
Year Live Births Infants Reported
1973
1974
1975
1976
1977
1978
3,783
3,656
3,516
3,210
2,756
2,747
2
5
3
4
37
53
183
-------
TABLE 5
Results of the Second-Level Check;
Distribution by Zone and Year of Birth
Year of birth
Zone11'1 1976 (II sem.) 1977 1978
Zone A
Live births 243
Possible malformations 000
Dubious diagnoses 000
Zone B
Live births 28 67 75
Possible malformations 026
Dubious diagnoses 010
Zone R
Live births 221 375 428
Possible malformations 7 9 18
Dubious diagnoses 2 10 4
Zone A+B+R
Live births 251 446 506
Possible malformations 7 11 24
Dubious diagnoses 2 11 4
[Zone A+B+R]b)
Live births 1138 2201 2291
Possible malformations 7 45 91
Dubious diagnoses 1 8 15
b)
'Zones A, B, and R represent areas of high, low, and no soil contamination.
A+B+R represents the area of the 11 municipalities, excluding Zones A, B,
and R.
184
-------
Comparisons will be possible only among populations living within
the monitored area due to the difficulties of including in the study
a control population outside of the 11 municipalities.
A map of the area at risk from TCDD is being prepared. It will
permit identification of different degrees of pollution and epidemiologic
comparisons.
The distribution of the positive cases on the area is detailed in
the maps in Figure 2 for the years 1972, 1976, 1977, 1978, and 1979.
An uneven distribution of cases among municipalities has been observed.
This might be due, at least in part, to the small number of birth defect
carriers in each municipality. Birth defects are somewhat more frequent
south of the ICMESA plant. BDR activity and its results are shown for
each year studied in Figure 3.
The malformations identified are indicated in Table 6. Microcephaly
cases are still being investigated to establish whether they are primary
or secondary effects. Four groups of malformations (spina bifida,
microcephaly, Down's syndrome, and hemangioma) appear with particularly
high frequency, but their significance has not yet been established.
The data do suggest definitive reasons for the increase in congenital
malformations observed over time. The increase in absolute frequency
of congenital malformations observed from 1976 to 1978 (Table 4) is
probably due to the recordkeeping of the BDR.
185
-------
1172
1977
1978
FIGURE 2. Maps of the monitored area showing distribution of malformed infants
in the 11 municipalities. Zones A,B, and R represent areas of high,
low, and no soil contamination as defined on the basis of the chemical
analysis performed in the first months after the accident. Birth
defect rates are presented for 1978 only, as in that year the monitoring
was satisfactory. (• = positive cases carrying one or more malforma-
tion; o = malformed infants born before July 10, 1976.
186
-------
o
(D
CO
c
o
Hi
P3
n
it
oo
ilrll.1 In 1*76 III*.,
Tottl blrthi
T»l«l birth*
"rill. In 1BTI
if
o
TO
D"
ro
T)
U»l
rll
-------
TABLE 6
Positive Cases, Updated to September 1979a
Diagnosis
(single malformation) 1979
1° sem.
Central nervous system
01 Anencephaly
02 Encephaloceleb
03 Meningomyelocele, spina
bifida
04 Microcephaly
05 Other
Subtotal
Eye
08 Cataract
09 Colobomata
10 Glaucoma
11 Other
Ear
12 Anotia
13 Other
Subtotal
Cardiovascular system
15/16 Ventricular septal defect 2
17/18 Atrial septal defect
20 Patent ductus arteriosus
21 Aortic coarctation or stenosis
22/23 Other
25/26 Pallet's tetralogy 1
Subtotal 3
1978
1
5
4
10
1
3
1
1
6
6
3
9
Year of birth
1977 1976
2° sem.
1
1
2
1
2
1
4
3
1
4
1972
1
1
1
1
3
1
6
1
1
1
1
3
7
Respiratory and digestive system
31 Harelip 12 1
32 Cleft palate 1
33 Harelip and cleft palate 2
38 Intestinal atresia 1
39 Absent, atresic, or imperforate
anus 1
40/44 Other 2 5
41 Diaphragmatic hernia 2
43 Inguinal hernia 53 4 2 10
Subtotal 6 12 5 3 16
aPart of the list used by the Italian Research Council for Perinatal Preventive
Medicine, Subproject MPP4.
"See multiple malformations
188
-------
TABLE 6 (continued)
Diagnosis
(single malformation)
Year of birth
1979
1°
1978
1977
s em.
1976
2° sem.
1972
Genitourinary system
48 Extrophia of the bladder
51 Hypospadias
52 Bilateral cryptorchidism
53 Other anomalies/male genitalia
54 Other anomalies/female genitalia
Subtotal
Bones and joints
55 Syndactyly
56 Polydactyly
57 Clubfoot
58 Dysplasla of the hip
60 Other limb defects
61 Osteomuscular abnormalities
of head and trunk
63 Chondrodystrophies
Subtotal
Skin
65 Hemangioma > 0.5 cm diameter (raised
or flat in atypical locations) 17
64 Other 5
Subtotal 22
Chromosomal abnormalities
66/67 Trisomy 21
76 Other
Subtotal
Multiple malformations
One case for each group of codes
01+02
05+11
+31
11+25
43+58
51+65
55+60
58+64
+65
64+65
5
1
6
21
26
11
37
02+03
+04+60
04+26
+51+53
18+20
+24
23+54
+56
24+43
43+57
51+55
+60
56+67
64+65
1
6
1
1
2
1
4
5
8
3
2
1
1
8
1
13
8
2
10
2
1
3
05+08
+23
43+51
43+65
55+60
58+65
64+65
1
7
4
2
6
05+13
+52
11+13
64+65
10
1
11
1
7
1
1
2
15
12
14
26
4
4
43+51
43+65
60+64
189
-------
A final diagnosis of "birth defect" has been made in 29% of the
"known" infants born in 1976, in 26% of those born in 1977, and in 46%
of those born in 1978. However, it is not yet possible to establish
whether this increase is due to the ICMESA accident or to the improved
reporting system. The rates obtained for 1978 are similar to those
reported in the literature (Klingberg and Weatherall, 1979; Takemichi,
personal communication, 1979).
There are still many "unknown" children born in 1972; this group
includes stillbirths, deaths between the ages of 0 and 6 years, and
those who have moved away. The search for these children is incomplete
and difficult. Therefore, the real rates for year 1972 could be very
close to those observed for 1978.
None of the nine infants born between 1976 and 1978 to women from
Zone A (the area with the highest TCDD soil contamination) was malformed.
Seven more babies were born in 1979; one of these has a flat angioma on
the thigh.
In 1978, 24 infants (of 506 live births, 47%) had birth defects;
in the surrounding areas, the percentage of birth defects was 40%.
These numbers are small and the detection procedures may have had
varying degrees of accuracy.
The data do not show a large increase of congenital malformations
attributable to the chemical pollution accident from the ICMESA plant.
However, a comparison of these data with those derived from the study
of other pathological consequences of TCDD pollution, and a more careful
analysis of those malformations frequently observed in the area, are
necessary.
190
-------
REFERENCES
Advisory Committee on 2,4,5-T. 1971. Report of the Advisory Committee
on 2,4,5-T to the Administrator of the Environmental Protection Agency,
mimeographed. U.S. Environmental Protection Agency, Washington, DC.
76 pp.
Bleiberg, J., M. Wallen, R. Brodking, and I.L. Applebaum. 1964.
Industrially acquired prophyria. Arch. Dermatol. 89:793.
Carter, C.D., R.D. Kimbrough, J.A. Liddle, R.E. Cline, M.M. Zack, Jr.,
W.F. Barthel, R.E. Koehler, and P.E. Phillips. 1975. Tetrachlorodibenzo:
An accidental poisoning episode in horse arenas. Science 188:738.
Courtney, K.D., and J.A. Moore. 1971. Teratology studies with 2,4,5-
trichlorophenoxyacetic acid and 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Toxicol. Appl. Pharmacol. 20:396.
Cutting, R.T., T.H. Phuoc, J.M. Ballo, M.W. Benenson, and C.H. Evans.
1970. Congenital malformations, hydatidiform males and stillbirths in
the Republic of Vietnam, 1960-1969. Document No. 903.233. Government
Printing Office, Washington, D.C.
Goldman, P.J. 1972. Schwerste akute Chlorakne durch
Trichlorophenol-Zeresetzungsprodukte. Arbeitsmed. Sozialmed. Arbeitshyg.
7:12.
Greig, J.B., G. Jones, W.H. Butler, and J.M. Barnes. 1973. Toxic effects
of 2,3,7,8-tetrachlorodibenzo-p-dioxin. Food Cosmet. Toxicol. 11:585.
International Agency for Research on Cancer. 1978. Coordination of
Epidemiological Studies on the Long-Term Hazards of Chlorinated Dibenzo-
dioxins/Chlorinated Diobenzofurans. Internal Technical Report No. 78/001,
draft. World Health Organization, Lyon, France. 48 pp.
Jirasek, L., J. Kalensky, and K. Kubec. 1973. Acne chlorina and porphyria
cutanea tarda during the manufacture of herbicides. Cesk. Dermatol.
48(5):306.
Jirasek, L., J. Kalensky, K. Kubec, J. Pazderova, and E. Lukas. 1974.
Acne chlorina, prophyria cutanea tarda and other manifestations of
general intoxication during the manufacture of herbicides, Part II.
Cesk. Dermatol. 49(3):145.
191
-------
Kimbrough, R.D., C.D. Carter, J.A. Liddle, R.E. Cline, and P.E. Phillips.
1977. Epidemiology and pathology of a tetrachlorodibenzodioxin poisoning
episode. Arch. Environ. Health. 28:77-
Klingberg, M.A., and J.A.C. Weatherall. 1979. Epidemiologic methods for
detection of teratogens. Vol. 1: Contributions to Epidemiology and
Biostatistics. Karger, Basel, Switzerland. 196 pp.
McQueen, E.G., A.M.O. Veale, W.S. Alexander, and M.N. Bates. 1977-
2,4,5-T and human birth defects, mimeographed. New Zealand Department
of Health, Division of Public Health. 41 pp.
Meselson, M.S., A.H. Westing, and J.D. Constable. 1971. Background
material relevant to presentations at the 1970 annual meeting of the
AAAS concerning the Herbicide Assessment Commission for the American
Association for the Advancment of Science, mimeographed. Washington,
D.C. 47 pp.
Murray, F.J., F.A. Smith, K.D. Nitschke, G.C. Humiston, R.J. Kociba,
and B.A. Schwertz. 1977. Three-generation reproduction study of rats
ingesting 2,3,7,8-tetrachlorodibenzo-p-dioxin. P. 24 in Proceedings of
the 16th Annual Meeting of the American Society for Toxicology, Toronto.
Neubert, D., P. Zens, A. Rothernwallner, and H.J. Merker. 1973. A survey
of the embryotoxic effects of TCDD in mammalian species. Environ. Health
Perspect. N5:67.
Poland, A.P., D. Smith, G. Metter, and P. Fossick. 1971. A health survey
of workers in a 2,4-D and 2,4,5-T plant. Arch. Environ. Health 22:316.
Rose, H.A., and S.P.R. Rose. 1972. Chemical spraying as reported by
refugees from South Vietnam. Science 177:710.
Sparschu, G.L., F.L. Dunn, and V.K. Rowe. 1971. Study of the teratogenicity
of 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Food Cosmet. Toxicol.
9:405.
Thiess, A.M., and R. Frentzel-Beyne. 1978. Mortality study of persons
exposed to dioxin after an accident which occurred in the BASF on 13th
November 1953. Paper presented at the Fifth International Medichem
Congress, 5th-9th September, San Francisco.
Tung, T.T., T.K. Anh, B.Q. Tuyen, D.X. Tra, and N.X. Huyen. 1971.
Clinical effects of massive and continuous utilization of defoliants on
civilians. Vietnamese Studies 29:57.
192
-------
DISCUSSION
DR. MILLER: On the basis of animal studies, do you expect a
teratogenic effect in humans from exposure to TCDD?
DR. MARNI: A cleft lip, cleft palate, kidney abnormalities, edema,
and other anomalies have appeared in mice, rats, hamsters, and some
primates. Only a few defects have been noted in humans, and they are
not definitely related to TCDD pollution. In Vietnam, two cases of
trisomy, a case of palatoschisis, and spina bifida have appeared.
Similar observations were recorded in New Zealand.
DR. MILLER: Animal studies for teratogenesis may not relate very
well to human experience. Aspirin is a good teratogen in rodent, in
which it produces cleft palate and cleft lip, but produces no such
effect in humans. Cortisone does the same thing in rodents, but not in
humans; and thalidomide does nothing to rodents. At least for those
compounds, it is almost as if the effect in humans is opposite to that
in animals.
PROF. DARDANONI: On a list of the most frequent malformations, is
there any tendency toward one particular type, although the total number
of malformations might not clearly be increased? Are certain types of
malformation more frequent in certain years?
193
-------
DR. MARNI: We have few data. We have seen an increase in
malformations only in 1978. There may have been an increase in
malformations in unknown subjects born in 1976 and 1977. We need to
know if three or four types of malformation are at a high incidence
with respect to normal data. Also, we need data for past years. We
hope to see all or a high percentage of children born between 1972 and
1975. It is a very difficult, though possible, job to find these
children who live now not just in our area but, throughout all Italy.
PROF. DARDANONI: An effort is being made to locate these
nonrespondents, who sometimes comprise anywhere from 10% to 50% of the
sample.
194
-------
Carcinogenic Effects of Chemical and Physical Agents;
~~~Human Observations
Clark W. Heath, Jr.1
Study of carcinogenesis in humans is limited by the nature
of cancer itself — its low frequency, its nonspecific clinical
nature (lack of specific etiology), the multiple factors involved
in its etiology, and long latency periods between exposure and
disease appearance. Several specific procedures may help identify
relationships between cancer incidence and exposure: a centralized
registry would pool cohorts with similar exposures; "marker tumors"
could be identified and studied more intensively; and preclinical
markers such as chromosomal or mutagenic damage could be established.
The question of carcinogenic effects is ever present in environ-
mental exposure situations. Will a particular exposure increase
cancer risk in a specific population? Unfortunately, the question is
easy to ask but extraordinarily difficult to answer. There are
three approaches to developing an answer:
o extrapolation from results of animal experiments;
o inferences from chemical structure or from the known actions
of related compounds or similar exposures; and
o direct observations of humans.
Only the last, observations of humans, can provide final answers.
However, the ability to observe is hampered not only by constraints
on human experimentation, which dictate reliance on epidemiologic
studies, but also by the nature of carcinogenesis itself. The limitations
^Chronic Diseases Division, Center for Environmental Health,
Centers for Disease Control, Atlanta, Ga.
195
-------
imposed by carcinogenesis are (1) the low frequency with which specific
cancers occur, especially in low exposure settings; (2) the nonspecific
clinical nature of cancer (that is, specific etiologies cannot be assigned
to specific cases); (3) the multifactorial nature of cancer etiology;
and (4) the long and variable latency periods that occur between
exposures to carcinogens and diagnosis of cancer.
LOW FREQUENCY
When considered collectively over a lifetime, cancer is a common
disease. But particular types of cancer occurring over limited time
spans are rare. Epidemiologic studies must therefore encompass
large populations if they are to measure rates of disease directly.
Since risk of cancer generally increases with dose of the carcinogen,
it is not surprising that most evidence of human oncogenicity has
come from studies of populations with relatively high exposures. In
practical terms, this usually means studying exposures in occupational
settings (Cole and Goldman, 1975). Prominent examples of occupational
carcinogenesis include lung cancer in workers exposed to asbestos;
bladder cancer in workers exposed to dyes; leukemia in workers exposed
to benzene; and lung, skin, and liver cancers in workers exposed to
arsenic. In such instances, high occupational exposures in the past
have led to recent increases in cancer incidence. These increases
have been of sufficient magnitude to be detected epidemiologically
in relatively small cohorts of workers. (An increased incidence of
cancer is not, of course, limited to occupational groups, as evidenced
by the lung cancer of cigarette smokers and cancers of many sorts in
Japanese survivors of the atom bomb.)
196
-------
Some knowledge of actual dose levels is necessary for any observa-
tion of excess cancer in humans to have full impact. Dose-response
patterns, if present, provide strong support for concluding that a
specific exposure actually caused a particular increase in cancer
incidence.
These two elements, therefore—adequate sample size and adequate
dose information—are important in any epidemiologic study. The
interaction of these two factors is clearly illustrated in the epidemio-
logic approach currently being pursued to assess the potential carcino-
genicity of polybrominated biphenyls (PBB's) in human populations in
Michigan (Landrigan et_ al., 1979). A cohort of some 4,000 persons
with relatively high exposure to PBB's is being followed prospectively.
Followup will be maintained for at least 15 to 20 years. The group
has been observed for 6 years thus far. A major intent of the study
is to detect increases in cancer incidence that may occur. Exposure
information is incorporated into the study through the subjects'
responses to questions concerning PBB exposure and through laboratory
measurements of PBB levels in serum. The overall sample size is
considered sufficient to detect a significant rise in liver cancer
over 5 years. Individual exposure data should eventually make it
possible to correlate cancer incidence to PBB levels. For the most
part, however, dose levels are low even in this cohort of relatively
highly exposed persons. It may be unlikely, therefore, that any
increased cancer incidence will be found, given limits in cohort size.
197
-------
CLINICAL NONSPECIFICITY
Observational tools for cancer epidemiology are at a somewhat
primitive stage of development. On the whole, most cancers can be
discriminated only by tissue site and cell type, not by particular
etiology. By analogy, the main advances in acute infectious disease
epidemiology came only after clinical discriminants were supplemented
with etiologic tests that isolated viruses and bacteria and measured
antibody titers.
This observational handicap is evident everywhere in cancer
epidemiology. Assessment of particular etiologies is achieved, not
by pinpointing the cause of specific cases, but by adjusting for the
relative contributions of competing etiologies, and then measuring
any differences that remain.
A simple example is the ongoing study of leukemia incidence among
military personnel present at a 1957 U.S. atmospheric nuclear test
called "Smoky" (Caldwell _et_ al_. , 1980). Followup of this cohort of
approximately 3,000 men has identified nine cases of leukemia, whereas
3.5 cases might be expected from normal incidence patterns. The
study is limited, of course, both by its relatively small size in
terms of low expected cancer incidence and by incomplete exposure data.
Film badge readings provide some measure of penetrating radiation
dose, but no reliable estimates have yet been developed for possible
internal exposures from inhalation or ingestion of radioactive materials.
Under these circumstances, it would be quite useful if means were
at hand for discriminating between leukemia cases caused by radiation
198
-------
and leukemia cases caused by other exposures. While such discrimination
might explain one of the nine cases (a possible variant of chronic
lymphocytic leukemia, a form of leukemia not associated with radiation
in other studies), the remaining eight cases may or may not be caused
by radiation, with five or six related to the nuclear incident.
MULTIFACTORIAL ETIOLOGY
In the absence of clinical discriminants, the multiple causes of
cancer make it difficult for cancer epidemiologists to assess relation-
ships between specific exposures and specific patterns of cancer
incidence. Analyses must account for competing risk factors while
assessing specific etiologic assocations. This process is often not
a major difficulty in occupational settings where relatively high
exposures can occur. It may also pose no problem for particular
types of cancer where the contribution of competing causes may be
small (lung cancer and cigarette smoking, for example) or may even
be nonexistent (mesothelioma and exposure to asbestos).
The relationship between workplace exposure to vinyl chloride
monomer (VCM) and different types of cancer in humans illustrates
this point. The very strong association of VCM exposure with hepatic
angiosarcoma, a rare tumor with no common competing causes (other
causes being arsenic and thorium dioxide), was easily identified and
confirmed (Falk et^ al., 1974). In contrast, however, the association
of occupational VCM exposure with increase in lung cancer incidence
has proved much more difficult to study (Falk e_t al., 1976), despite
the much greater frequency of lung cancer compared to hepatic angiosarcoma.
199
-------
The reason for this difficulty, of course, lies largely in the need
to adjust for cigarette smoking as a strong competing risk factor in
lung cancer etiology.
Unfortunately, the problems produced by competing risk factors
can be expected to be all the greater in situations involving low
levels of toxic exposure. Hence, in the Michigan PBB study, it may
prove extremely difficult to link observed cancer incidence patterns
to actual PBB exposure, given the combination of multiple risk factors
and low tissue PBB levels. Conceivably, of course, a rare form of
cancer involving relatively few competing variables might easily
identify the relationship. For more common tumors, however, such an
outcome should not be expected, given the study's limitations in
size and the likelihood of strong competing variables.
LONG LATENCY
The problem of long latency imposes perhaps the greatest restriction
on human cancer epidemiology, particularly where exposures are recent.
In situations where 10 or more years have passed since exposure began,
and where sufficient numbers of persons were exposed, retrospective
cohort studies can be performed, or case-control studies, if exposures
are relatively widespread. Such retrospective approaches, of course,
are difficult for the investigator who must reconstruct exposure
histories or retrieve complete sets of past disease incidence data.
Prospective studies such as the Michigan PBB project allow one
to develop case incidence data as time passes, and they avoid the
biases that arise from retrospective risk assessment. But these
200
-------
studies demand patience; they cannot be used to derive quick answers
to questions of cancer risk.
CENTRALIZED REGISTRY
Some centralized mechanism is needed whereby persons exposed in
specific environmental situations might be registered for long-term
followup in a collective or pooled manner. Most individual exposure
situations involve too few subjects to justify long-term epidemiologic
followup by themselves. However, several similar exposed populations
could well provide a sufficient sample size if combined for joint
followup. A central registry would maintain records on specific persons
exposed in particular situations and permit access at appropriate times
for coordinated followup with respect to cancer incidence. To ensure
comparability among individual population sets, a standardized data
core would have to be established for recording individual features
such as age, race, sex, smoking history, and dose information.
MARKER TUMORS
In considering cancer incidence among populations, particular
attention might be given to unusual or rare tumor types or to cancers
occurring at young ages. Close observation of such unusual "marker"
cases might well provide clues to etiology that are obscured by
competing risk factors in older subjects or by more common forms of
cancer. Epidemiologic studies of tumors such as hepatic angiosarcoma,
mesothelioma, and vaginal adenocarcinoma in young women have successfully
provided the impetus for this suggestion.
201
-------
PRECLINICAL MARKERS
New observational tools, which go beyond merely recognizing the
clinical occurrence of human cancer and relating such events to
environmental exposures are needed. The effectiveness of present-
day epidemiologic techniques would be enhanced if data concerning
preclinical markers or subclinical biologic damage could be assembled
in such a way that eventual cancer occurrence or "cancerproneness"
might be predicted accurately. Current interest in markers of
chromosomal and mutagenic damage points in this direction and
suggests approaches for epidemiologists. The development of such
markers, of course, is basically a matter of continued research
regarding mechanisms of carcinogenesis.
The effectiveness with which such new knowledge is developed and
then applied to epidemiologic studies will ultimately determine how
useful observations in humans are in assessing cancer risks arising
from potentially toxic environmental exposures.
202
-------
REFERENCES
.Caldwell, G. C., D. B. Kelley, and C. W. Heath, Jr. 1980. Leukemia among
participants in military maneuvers at a nuclear bomb test (Smoky): A
preliminary report. J. Am. Med. Assoc. 244:1575-1578.
Cole, P., and M. B. Goldman. 1975. Occupation. Pp. 167-183 in J. F.
Fraumeni, Jr., ed. Persons at High Risk to Cancer: An Approach to
Cancer Etiology and Control. Academic Press, New York.
Falk, H., and R. J. Waxweiler. 1976. Epidemiological studies of vinyl
chloride health effects in the United States. Proc. R. Soc. Med.
69:303-306.
Falk, H., J. J. Creech, C. W. Heath, Jr., e£ al. 1974. Hepatic disease
among workers at a vinyl chloride polymerization plant. J. Am. Med.
Assoc. 230:59-63.
Landrigan, P. J., K. R. Wilcox, Jr., J. Silva, Jr., et al. , 1979. Cohort
study of Michigan residents exposed to polybrominated biphenyls:
Epidemiologic and immunologic findings. Ann. N.Y. Acad. Sci. 320:284-294.
203
-------
DISCUSSION
DR. SUSKIND: Some of the major factors that need to be considered are
difficult to deal with in constructing an epidemiologic study. Workplace
populations in particular present immense problems. What have they been
exposed to in addition to the compound under study and how long? Many of
these populations have been exposed to several compounds concurrently.
What other types of exposure do they have (e.g., drugs, tobacco, and
alcohol)? One of the most difficult problems is to derive accurate data
about their other exposures prior to the one being studied. And in a
retrospective study, subsequent exposures must also be noted and accounted
for.
DR. MILLER: You suggested the formation of a central registry for
long-term followup of people exposed to certain chemicals, and then said
that your own group is studying veterans exposed to low-dose radiation
during the Smoky exercise, people exposed to PBB's in Michigan, and vinyl
chloride workers. Is the work of the Centers for Disease Control (CDC)
being diluted by having to delegate resources to these huge epidemiologic
studies that should be spent on the registry for following tumor markers?
DR. HEATH: There is really no question that resources are tied up
inappropriately. In the Michigan study, however, the resources are not
ours. The major funds are from other parts of the Federal Government.
CDC is managing the project, and the work is being carried out by the
state health department.
It could be argued, especially in the face of limited resources, that
several large followup studies would be enough in terms of long-term followup
204
-------
and that the maintenance of a central registry of smaller events might
not be necessary. Such a central registry might largely be a rather
dormant project that awakens after 10 years for some linkage.
DR. DARDANONI: Could routinely collected data, such as morbidity
data from hospitalized people, be used to monitor certain diseases
in the exposed population?
DR. HEATH: That kind of surveillance program for environmentally
caused diseases, which are nonspecific, becomes a much larger task
than initially expected. There is more reason to set up wide networks
of disease surveillance to serve a particular exposure problem.
205
-------
Experimental Studies on Carcinogenic Effects of TCDD
Giuseppe Delia Porta, Maria I. Culnaghi, and Tommaso A. Dragani1
The possible carcinogenic activity of 2,3,7,8-tetrachlorodibenzodioxin
(TCDD) has been studied in long term experiments in which the compound
was administered orally to Sprague-Dawley rats and Swiss mice. Both
studies indicated that there were carcinogenic effects. In the rat, an
increased incidence of some tumor types and a decreased incidence of
others were observed only at the highest dosage level (0.1 yg/kg/day in
the diet for 2 years). In the mouse, an increased incidence of liver
tumors was noted after weekly doses of 0.7 y g/kg were administered by
gavage for 1 year. Three different experiments in (C57BL/6JDp x
C3Hf/Dp)F1 and (C57BL/6JDp x BALB/cLacDp)F1 mice are in initial stages.
In the first experiment, 10-day old mice are receiving 5 intraperitoneal
doses of 1, 30, and 60 g/kg once weekly and then observed for 15
months. In the second experiment, female mice are receiving 10 yg/kg
by gavage on days 16-18 of pregnancy, and their progeny are observed
for 15 months. In the third experiment, 6-week-old mice are receiving
weekly doses of 2.5 and 5 yg/kg by gavage for 1 year and then observed
for an additional year.
Carcinogenesis was certainly not among the high priority health
problems Italian authorities and the scientific community faced soon
after the Seveso accident. Certainly, acute and subacute toxicity had
to be investigated first; a late effect such as cancer induction had to
be studied later.
At the time of the Seveso accident, several years after somewhat
similar accidents around the world, practically nothing was known of
the carcinogenic potential of 2,3,7,8-tetrachlorodibenzodioxin (TCDD).
In the only published experiment, Innes £t_ al_« (1969) used 2,4,5-
trichlorophenoxyacetic acid contaminated with a level of TCDD sufficient
to supply 0.27 y g of TCDD/kg/day to mice and obtained negative results.
-'•Division of Experimental Oncology A, Istituto Nazionale per lo Studio
e la Cura dei Tumori, Milan, Italy.
206
-------
A few experiments were underway In the United States and in Hungary,
but no data were yet available.
The Van Miller report (1977) of a 1.5 to 2-year feeding experiment
in SpragueDawley rats showed some indication of carcinogenicity.
Unfortunately, the experiment was conducted on only 10 males per dosage
level, and no formal evaluation of the significance of its results was
possible, particularly because in most treated groups a few single
tumors of different types were observed. However, the four lung tumors,
the four neoplastic nodules of the liver, and the two cholangiocarcinomas
observed in animals receiving the highest dosage level could not be
dismissed easily. The highest level was roughly equivalent to 0.3
1jg/kg body weight per day. Interim reports of other experiments were
also indicating carcinogenicity.
Meanwhile, it had become clear that TCDD was going to remain in
the Seveso area and that not only the initial high-dose exposure but
also a long-term low-level exposure might occur and involve a large
population. Carcinogenicity was therefore considered more seriously
and epidemiologic and additional experimental studies were believed
necessary.
Toward the end of 1977, our institute was asked to prepare a study
of carcinogenicity. One major obstacle was the location of a suitable
facility for conducting the experiment. Facilities for long-term
studies are limited in Italy, especially sites for conducting a dangerous
experiment under some sort of GLP (good laboratory practice) conditions.
207
-------
In addition, the Seveso accident had created a psychologic antipathy in
the general population toward experiments with TCDD.
Since other pharmacologic studies were foreseen, it was decided
that three small laboratory facilities would be prepared and equipped
for TCDD experiments within the boundaries of the contaminating factory.
These rather small laboratories are now located in plastic wagons with
air-conditioning and absolute filters. Strict operational rules were
established to protect the few people allowed to enter the restricted
area. Moreover, it was decided to limit the TCDD load at any given
time. The laboratories were ready by June 1978, and went into full
operation a few months later.
Preliminary reports of Kociba et^ al. (1978) and Toth _et al. (1979)
were then available. Both the Van Miller et_ al. (1977) and Kociba e_t
al. (1978) experiments were performed on Sprague-Dawley rats by feeding
them TCDD in the diet continuously; the Toth ejt al. (1979) experiments
in Hungary were performed on outbred Swiss mice given TCDD by stomach
intubation.
In our study only mice were used, partly because of space limitations.
Two F, hybrids of inbred strains were selected. The hybrid B6C3F,
(C57BL/6JDp x C3Hf/Dp) has been widely used in the United States and
carries from its parental C3H strain the propensity toward a fairly
high spontaneous incidence of liver cell tumors. The B6CF. hybrid
(C57BL/6JDp x BALB/cLacDp) carries a susceptibility to leukemogenesis
from the C57BL strain and to lung carcinogenesis from the BALB/c strain.
The hybrids were also selected because of their strength and longevity
relative to those of their parental strains.
208
-------
For better control of contamination in the laboratory, it was
decided not to mix TCDD with the diet. Thus, in the long-term assay,
TCDD was given by gavage, dissolved first in a small amount of acetone
and then in olive oil to a volume of 0.1 ml/10 g body weight per
administration. Treatment was administered once a week to assure a
sufficiently continuous exposure relative to the retention of TCDD.
Only B6C3F^ were used for the long-term assay. One untreated
control group of 50 males and 50 females is kept in the animal quarters
of the Cancer Institute in Milan. A vehicle control of similar size is
in the Seveso laboratory. There are two TCDD-treated groups of 50
males and 50 females each.
The treatment was started when the animals were 6 weeks old and
will continue for a year. The animals will be observed until the end
of their second year, at which time the experiment will be terminated
and all survivors will be examined pathologically.
Selection of dosage level was not easy since there was no way to
perform a proper 90-day study. In addition, this particular study had
to be restricted to two levels of exposure because of the limitation of
the facilities. Thus, it was decided to keep both levels in the high-
dose zone. Toth et al. (1979) had encountered excess mortality at
weekly oral doses of 7 yg/kg body weight. As a consequence, the dosage
had been lowered to 0.7 yg/kg body weight. On the other hand, Kociba
et al. (1978), working with rats that had an almost 10-fold lower LD..Q
than the mouse, had observed a somewhat high mortality at a dosage
level of 0.1 yg/kg body weight per day. Thus, for this study it was
209
-------
decided to use 5 vg/kg/week as the high dose and 2.5y g/kg/week for the
low dose.
The experiment was started at two successive times, and groups of
15 mice of each sex are now approaching the end of treatment. The test
animals have suffered low mortality. A growth curve of these initial
groups shows a toxic effect in the high-dose males. At 20 weeks of age
and after 14 weeks of treatment, these animals already exhibited a
significant weight loss compared to the control animals. The low-dose
group started to lose weight after 10 weeks. Later on, animals in both
groups weighed considerably less than did the controls. The difference
was much less evident in the females and involved only those receiving
the highest dosage.
Two more studies were initiated. In these, each test group consists
of 50 males and 50 females of the two hybrids, B6C3F and B6CF . In
the first study, TCDD was given intraperitoneally, starting at 10 days
of age, once a week for 5 weeks. These animals will be observed for
1.5 years. This schedule should cover the period of maximal susceptibility
to carcinogenesis previously demonstrated with various carcinogens in
similar experiments.
This experiment includes one vehicle control group and one group
treated with TCDD at 1y g/kg body weight, five times once weekly, a
treatment that did not exert any apparent toxic effect. Mortality
among the suckling mice and at the end of the treatment period was
within control levels for both hybrids. The growth curves showed no
difference between the treated groups and the controls.
210
-------
To select higher doses for the same schedule of treatment, a small
preliminary experiment has been conducted. Results have shown a steep
curve of mortality in animals receiving TCDD from 100 yg to 50 yg/kg of
body weight five times. At 100 yg, all mice died before the end of
treatment, whereas at 50 yg, there was no mortality, but there was a
decrease in the mean body weight gain relative to controls. Therefore,
we have selected 30 and 60 yg/kg body weight dose levels for two
additional experimental groups.
In the second experiment, a transplacenta! study, (one dose of 10
yg/kg/body weight was given by gavage near the end of gestation at days
16-18. No other treatment will be administered. The litters will be
observed for 1.5 years.
Currently, almost 4 years after the Seveso explosion, the implications
of exposure to TCDD for carcinogenicity are still unclear, although
three studies have already been published. The mouse experiment by
Toth et al. (1979) was performed in outbred Swiss male mice, which had
been given 7 yg of TCDD in sunflower oil by gavage once a week for 1
year. A high mortality rate resulted, precluding a proper evaluation
of tumor incidence. The 0.7 yg level produced a significant increase
in the incidence of liver tumors in animals. These growths, both benign
and malignant, developed in 21 of the 44 treated animals (48%), compared
to the matched vehicle controls, of which 7 out of 38 (18%) animals had
liver tumors. However, Toth et al. also presented the data from three
other control groups. One was given carboxymethyl cellulose instead
211
-------
of sunflower oil as the vehicle; the other two were untreated. All
three groups had a much higher incidence of liver cell tumors (33%,
33%, and 26%) than was observed in the matched control group. Thus,
it is difficult to ascribe the increased incidence of liver tumors to
the TCDD. The sunflower oil by itself may have caused a decrease in
the spontaneous incidence of liver cell tumors or, more likely in this
outbred strain, there is a variation in spontaneous incidence. This
experiment seems to indicate hepatocarcinogenicity in the mouse at a
relatively low dose, but this hypothesis is not fully substantiated.
The Van Miller et^ al. (1977) experiment provided the first strong
indication of carcinogenic activity in Sprague-Dawley rats given TCDD
in the diet for 78 weeks, at a dosage level equivalent to 0.3 g/kg
body weight per day. The small size of the groups (10 male rats)
precludes the evaluation of the results obtained at the smaller dosage
levels. It is difficult to ascribe significance to single observations
of various different types of tumor with groups of this size.
Kociba et al. (1978) also studied Sprague-Dawley rats. In their
experiment, groups of 50 males and 50 females received one of three
dosage levels 0.001, 0.01, or 0.1 yg/kg body weight daily in the
diet for 2 years. The untreated control group included 85 males and 86
females. Both males and females receiving the highest dosage exhibited
a significant increase (compared to controls) in the incidence of
squamous cell carcinoma of the hard palate and turbinates. Males
212
-------
receiving the highest dosage had an increased incidence of squamous
cell carcinoma of the tongue (although two cases were observed among
females as well). The incidence of squamous cell carcinoma of the lung
increased in the high-dose female group.
The development of squamous carcinomas is an important observation
since chloracne is based on a squamous metaplasia. Some of the squamous
carcinomas may have developed over squamous metaplasia, involving well-
differentiated (perhaps mucin-producing) epithelial cells. In various
situations, squamous metaplasia is replacing more sophisticated epithelial
elements.
The other group of tumors observed in increased incidence by Kociba
et al. (1978) were liver cell tumors — liver nodules and liver
carcinomas—but only in the female group. Carcinomas occurred only
in the high-dose animals; liver nodules were also observed in those
receiving 0.01 yg TCDD. It is difficult to compare these results with
the Van Miller et^ a^. (1977) data, which are derived only from males.
In the same experiment, Kociba and colleagues showed a decrease of
tumors as well; again, the finding occurred in animals receiving the
highest dose. There was a significant decrease in the occurrence of
endocrine-type tumors, mammary benign tumors, mammary carcinomas,
uterine benign tumors, adrenal pheochromocytomas, and pancreatic adenomas
in males. This decreased incidence should be further investigated in
connection with a possible imbalance produced by a toxic effect of TCDD
on endocrine glands directly or indirectly through liver injury.
213
-------
In an experiment conducted under a National Cancer Institute
contract mice and rats were given TCDD by gavage and dermal application,
but the final report of this study is not yet available. If results of
this experiment confirm the results of previous experiments, particularly
regarding target organs, a mechanism of carcinogenesis might be suggested.
Results of other studies may also have implications for TCDD
carcinogenicity. First, as demonstrated by Poland and Glover (1979),
covalent binding to DNA is extremely low, and mutagenicity testing has
not provided convincing results. On the other hand, TCDD is a potent
enzyme inducer that may augment the activation of other carcinogens
from endogenous or exogenous sources, thereby modifying tumor incidence.
The few experiments using TCDD and known carcinogens to study carcinogenesis
point to the fact that TCDD may modify carcinogenicity in both directions.
In conclusion, public health officials need to know whether there
is sufficient evidence that TCDD is producing tumors in experimental
animals; whether the mechanism, direct or indirect, is known; and
whether the results can be extrapolated to humans. In addition, the
work of epidemiologists will be facilitated or at least the results
will provide stronger evidence if target organs can be identified. Not
all answers are in as yet. Unfortunately- one of the major lessons is
that these experiments should have been performed before the emergency
occurred.
214
-------
REFERENCES
Innes, J.R.M., B.M. Ulland, M.G. Valeric, L. Petrucelli, L. Fishbein,
E.R. Hart, A.J. Pallotta, R.R. Bates, H.L. Fack, J.J. Gart, M. Klein,
I. Mitchell, and J. Peters. 1969. Bioassay of pesticides and industrial
chemicals for tumorigenicity in mice: A preliminary note. J. Natl.
Cancer Inst. 42:1101-1114
Kociba, R.J., D.G. Keyes, J.E. Beyer, R.M. Carreon, C.E. Wade, D.A.
Dittenber, R.P. Kalnins, L.E. Frauson, C.N. Park, S.D. Barnard, R.A.
Hummel, and C.G. Humiston. 1978. Results of a two-year chronic
toxicity and oncogenicity study of 2,3,7,8tetrachlorodibenzo-p-dioxin
in rats. Toxicol. Appl. Pharmacol. 46:279-303.
Poland, A., and E. Glover. 1979. An estimate of the maximum in vivo
covalent binding of 2,3,7,8-tetrachlorodibenzo-p-dioxin to rat liver
protein, ribosomal RNA, and DNA. Cancer Res. 39:3341-3344.
Toth, K. , S. Somfai-Relle, J. Sugar, and J. Bence. 1979. Carcinogenicity
testing of herbicide 2,4,5-trichlorophenoxyethanol-containing dioxin
and pure dioxin in Swiss mice. Nature 278:548-549.
Van Miller, J.P., J.J.. Lalich, and J.R. Allen. 1977. Increased incidence
of neoplasms in rats exposed to low levels of 2,3,7,8-tetrachlorodibenzo-
p-dioxin. Chemosphere 10:625-632.
215
-------
DISCUSSION
DR. MILLER: Regarding the exposure of mice in utero, was the TCDD
administered once between the 16th and 18th days of gestation?
DR. DELIA PORTA: Yes.
DR. MILLER: Suppose the critical day for developing tumors was the
14th or 15th days? Why not administer the TCDD in some experiment throughout
the pregnancy?
DR. DELIA PORTA: That creates a teratogenesis problem.
DR. MILLER: What about in the last half of pregnancy?
DR. DELIA PORTA: Perhaps TCDD could be administered earlier than the
16th day.
DR. MILLER: Is the TCDD you are administering pure?
DR. DELLA PORTA: Yes.
DR. MILLER: In case there are impurities, should another experiment
be conducted using the contamination experienced to see what would happen
in humans?
DR. DELIA PORTA: Yes, but this type of experiment is very dangerous
for the workers exposed. So, there are severe limitations.
DR. MILLER: Were there any kidney tumors?
DR. DELLA PORTA: No, perhaps one or two observations, but none were
statistically significant.
DR. REHDER: Are you going to look for changes such as metaplasia in
the respiratory tract or leukoplakic changes within the mouth area?
DR. DELIA PORTA: Yes, we are, including those cases where no tumors
occur. The Kociba study does not mention the histopathology in the same
areas of animals that had not developed tumors. Perhaps even at the
lower dosage levels there may be this kind of pathology.
216
-------
Adverse Neurologic Effects
Alan M. Goldberg1
A study of neurologic effects of chemical contamination poses
three major problems: Definition of the questions to be addressed,
level and duration of exposure, type of population; current scien-
tific knowledge regarding the chemical's mechanism of action; and
an understanding of the effects of a specific chemical on the human
nervous system. Exposure to lead and Kepone typify the problems
researchers must solve in order to deal with specific polluting
events, in terms of current knowledge regarding a chemical itself
and the biologic effects of human exposure to it, specifically on
nervous tissue.
Response to areawide chemical contamination requires an
understanding of the immediate and longer term biologic effects
resulting from exposures to both high and low doses of a compound.
It is also crucial to understand the interaction between the xeno-
biotics and the multitude of biological systems potentially at risk.
An investigation of adverse neurologic effects has three components:
1. definition of the questions to be addressed;
2. development of scientific understanding of a chemical's
mechanism of action; and
3. understanding the uniqueness of the nervous system as a
target for toxic agents.
In addition to biologic considerations, investigators must take
into account many variables, such as the characteristics of the
exposed population, the ecology, and later consequences of the exposure.
Department of Environmental Health Sciences, Johns Hopkins University,
Baltimore, Md.
217
-------
The questions studied and the data collected will be considerably
different, depending on whether the population is young or old, male
or female, pregnant or nonpregnant. The biological questions must
address the toxicity of the compound and its metabolism in humans.
These points will be illustrated based on the current literature
of exposure to lead and Kepone.
LEAD
Data obtained from the criteria document on lead, published
by the World Health Organization (1977), are plotted in Figure 1
and provide a misleading interpretation to relate disease with
the levels of lead in blood in children and adults. These data
correlate the clinical symptomatology or disease state with blood
lead levels and suggest that lead exposure produces similar, if
not identical, effects in adults and children. Furthermore, they
suggest that children are only slightly more susceptible to the
consequences of lead exposure. However, the consequences of lead
exposure are very different in the two populations; not only are
children more susceptible, the constellation of symptoms is also
different. Moreover, lead concentrations in blood only indicate
recent exposure and are inadequate and misleading as an indicator
of body burden of the metal.
Understanding the consequences of exposure to a chemical agent
and its relationship to health effects must be examined as a function
of both level and duration of exposure. Figure 2 shows a comparison
of a continuous and stable exposure to lead (dotted line) over a
218
-------
80-
70-
60-
E 50-
o
o
o> 40-
_ct 30-
Q.
20-
10-
adults
children
ALAD FEP ANAEMIA MBD ENCEP.
INH.
Fig 1. No detectable effect levels in terms of Blood lead
FIGURE 1. No-detectable-effect levels of lead in blood (Pb.B) of
children and adults.
219
-------
EXPOSURE
(Relative
Units)
TIME
FIGURE 2. Two types of exposure to lead: continuous-equal exposure
(dotted line) and an acute high dose exposure (solid line),
220
-------
prolonged period of time and contrasts it with an acute high dose
exposure (solid line). This point is raised to make one aware that
the health effects from a high dose acute exposure of lead are
different than the health consequences of long term low level
exposure to lead.
Persons exposed to various concentrations of lead for varying
lengths of time exhibit numerous signs of lead poisoning including
drowsiness and fatigue, intellectual deterioration, gastrointestinal
distress, anemia, basophilic stippling, kidney damage, and central
nervous system effects. The appearance of some of these effects
depends on the amount and duration of exposure. Central nervous
system effects recorded after low-dose exposures are considerably
different from those observed after high-dose exposure (National
Academy of Sciences, 1972). Unfortunately, the mechanisms by which
the lead produces its effects are known for only two of the symptoms:
in anemia, heme synthesis is inhibited, whereas basophilic stippling
is the result of a nucleotidase inhibition (Pagalia et^ al., 1975;
Valentine et_ al. 1976).
Lead poisoning is an environmental problem of magnitude. In
Baltimore, more than 70% of some 30 million private homes and
dwellings built before 1940 have surfaces coated with lead-based
paint. Public housing in the city is lead-free. Test children
(447) from public dwellings showed a mean blood lead level of 16.5
yg/dl. Less than 1% had a lead concentration in blood greater than
30 yg/dl. In 155 children living in private dwellings, the mean
221
-------
lead level was 38 jag/dl. Approximately 78% of these children had
a level higher than the 30 ]jg/dl. Almost 20% had a concentration
higher than 50 yg/dl (Chisholm, 1979, personal communication).
Clearly, children living in older private dwellings had elevated
lead concentrations in blood. These findings are summarized
in Table 1.
Settle and Patterson (1980) examined the problem of defining
"low-lead." They compared the amount of lead in tuna caught in
the deep sea and processed lead-free to that of processed tuna
packaged in lead-soldered cans. The deep-sea tuna, which was
processed at sea and remained packed in lead-free containers
throughout the assay procedure, contained approximately 0.3 ng
of lead per gram of fish. The tuna from the lead-soldered cans,
however, contained approximately 1,400 ng/g. These results are
an indication that the human population is continuously exposed
to lead and that the lowest lead exposure level attainable now for
the general population is really quite high. Settle and Patterson
(1980) also studied bones of Peruvian Indians to calculate their
level of "environmental" lead for comparison to current levels.
Exposure of the Indians was less than 200 ng/day; humans now are
exposed to approximately 29,000 ng/day. These data show the
magnitude of the lead-poisoning problem today, both in the United
States and throughout the world.
The measures used to quantify exposure to lead have only
recently been refined sufficiently to identify body burden. Thus,
222
-------
TABLE 1
Concentrations of Lead in the Blood of Children
Residing in Baltimore, Maryland"
Type of
Housing
Public (lead-free)
Older private
dwellings
(>70% with lead-
painted surfaces)
Mean Blood Percentage of Children
No. of Children Lead
Tested (yg/dl)
447
155
with Blood Lead >30 yg/dl
and >50 yg/dl
>30 yg/dl >50 yg/dl
16.5
38
0.7
78
0
19
From J. Chisholm, personal communication.
223
-------
the consequences of low-dose, long-term lead poisoning are just
being recognized. Recent studies by Needleman et_ a^. (1979) have
led to a more refined source from which to measure body burden of
lead—the shed tooth. Using lead levels in dentine, these
researchers concluded that lead exposure of a magnitude too low to
produce clinical symptomatology is associated with neuropsychological
deficits that interfere with normal development of children. With
various test batteries and scales, these researchers demonstrated a
dose-response relationship between dentine lead levels in children
and their overall academic performance. Children with high lead
levels performed significantly less well (than children with lower
lead concentrations) on the Weschler Intelligence Scale and on
attentional performance tests as measured by reaction time studies.
Table 2 shows teachers' ratings of children with high and low
levels of lead in dentine. "Low" was defined as less than 10 ppm;
the term "high" was applied to levels higher than 20 ppm. The
table lists the percentage of teachers who indicated negative
performances in these children for various characteristics.
Children with lead concentrations less than 10 ppm in their teeth
received about 10% overall negative responses; children with more
than 20 ppm in their teeth received, overall, some 25-30% negative
response—almost a threefold increase.
The IQ's (verbal and performance) of the children were matched
against the lead concentrations in their dentine (Table 3). Children
with low lead levels had IQ's three or four points higher than those
224
-------
TABLE 2
Teachers' Ratings of Children with High and Low Lead Levels in Dentine2
% negative responses
Characteristic <10 ppm >20 ppm
Distractibility 1436
Not persistent 9 21
Impulsive 9 25
Easily frustrated 11 25
Daydreamer 15 24
Low performance 8 26
aFrom Needleman et_ al., 1979; reprinted by permission of The New England
Journal of Medicine~300:689-732, 1979.
225
-------
TABLE 3
IQ's of Children with High and Low Lead Levels in Dentine3
Test <10 ppm >20 ppm
Full scale 106.6 102.1
Verbal 103.9 99.3
Performance 108.7 104.9
Needleman et al. , 1979; reprinted by permission of The New England
Journal of MedicineTsOO:689-732,1979.
226
-------
of children with elevated lead concentrations. Lead exposure thus
produced a significant decrease in IQ's in this population. In
these same children, investigators developed a method to measure
reaction time, which also measured the distractability of the
subjects. The high lead group had a significant increase in
reaction time and, thus, a significant decrease in performance.
This compilation of data strongly suggests that there is
considerable exposure to low but toxic levels of lead in our
environment. The magnitude of the problem is first being defined.
Appropriate intervention should be established.
KEPONE
A second environmental contaminant that typifies the problems
caused by area-wide chemical pollution is Kepone (chlordecone), a
unique chlorinated hydrocarbon, previously used as an insecticide.
It was patented in 1952 and registered as a pesticide in 1955.
Kepone was produced at two plants in Hopewell, Va. from 1966 to 1974
and at a third plant in the same area during 1974. There is consider-
able evidence that the chemical found its way into Bailey's Creek,
the James River, and finally into aquatic life as early as 1967.
The consequences to the health of humans who consumed the contaminated
aquatic species, if any, are unknown.
A number of people were exposed to high doses of Kepone during
its production in 1974. Exposure to the substance and the effects of
that exposure were preventable. When the high-dose exposure to Kepone
was first observed in humans, there were no published data pertaining
227
-------
to the consequences of such contamination. The mechanisms by which
Kepone exerts biologic effects are just being published now, some
5 years after the exposure, and almost 30 years after the introduction
and manufacture of the compound. Such data are basic to devising
rational approaches to deal with specific accidents and exposures.
Health effects were not observed in the community neighboring
the plant, although contamination was measurable. The acute,
high-dose exposure produced major neurologic pathology. One of the
most intriguing findings was opsoclonus, a true side-to-side fluttering
of the eyes.
Diagnosis of exposure to high levels of Kepone is possible
because its cluster of effects is unique. Symptoms include memory
difficulties, nervousness, weight loss, joint and chest pain,
tremors, and visual difficulties; physical signs involve tremors
(rest, postural, and intention), opsoclonus, ataxia, and memory
loss. In addition, laboratory tests reveal decreased sperm counts.
Although nerve conduction velocities are normal, myelin changes
have been observed in nerve biopsies. The consequences of exposure
to continuous but low levels of Kepone have not yet been studied.
UNIQUE ASPECTS OF NERVOUS TISSUE
In addressing the problem of neurotoxicity, or the neurological
consequences of a chemical exposure, we must understand the unique
properties of nervous tissue in order to understand which system or
subcomponent of the system is likely to be affected. One of the
most unique features of the nervous system is that the nerve cell
228
-------
has its cell body (the energy generating system, the protein synthesis
systems, etc.) at one location and an axon, which can be very long,
that does not have the capability, e.g., to synthesize protein. Thus,
protein must be transported from the cell body to the nerve terminal.
A second unique feature of the nervous system is the myelin sheath
and it is possible for the myelin to be affected by toxic chemicals.
Clearly, if we know that the myelin sheath around the nerve is
affected by a toxic chemical we might not yet know how to treat
but we will know how to diagnose and evaluate the problem. A third
unique feature of the nervous system are the neurotransmitters and
related enzymes. The possibility is that chemicals can interact
with either the synthesizing enzymes, the integrity of the nerve
terminal, the neurotransmitter, or the processes associated with
neurotransmission. Clearly, these are specific points at which
the nervous system might be susceptible to attack. It is these
unique characteristics which, if understood, may reveal how
chemicals interact with nervous tissue, thereby allowing us
to diagnose, treat, and (hopefully) prevent the damage produced
by these chemicals.
In conclusion, appropriate species must be studied to generate
the data needed to evaluate the toxicity of chemicals in humans. If
an experimental species does not metabolize a compound in the same
manner as humans, the information collected is inappropriate for
extrapolation to humans. Such species may well incur different
biologic consequences of exposure to the parent compound. Second,
229
-------
the basic mechanisms of the interaction of various toxic agents
must be known to evaluate the consequences of exposure. For
instance, if agents interfere directly with neurotransmission
without affecting transmission along the nerve, measurement of
nerve conduction velocity would be inappropriate. Knowing which
measures are appropriate is dependent on understanding the
mechanisms of action of the compound. Last, appropriate tests
must be developed to fully assess the damage within the nervous
system. The central nervous system is both plastic and redundant.
It is possible that chemical damage can result, but it is not
easily measured since other parts of the system compensate. We
must, therefore, develop techniques to truly assess the
consequences to health of such alterations.
230
-------
REFERENCES
National Academy of Sciences. 1972. Airborne Lead in Perspective.
National Academy of Sciences, Washington, D.C. 330 pp.
Needleman, H. L. C. Gunnoe, A. Leviton, R. Reed, H. Peresie,
H. Maher, and P. Barrett. 1979. DEficlts in psychologic and
classroom performance of children with elevated dentine lead
levels. N. Engl. J. Med. 300:689-732.
Paglia, D. E., W. N. Valentine, and J. G. Dahlgren. 1975.
Effects of low-level lead exposure on pyrimidine 5'-nucleotidase
and other erythrocyte enzymes: Possible role of pyrimidine
5'-nucleotidase in the pathogenesis of lead-induced anemia.
J. Clin. Invest. 56:1164-1169.
Settle, D. M., and C. C. Patterson. 1980. Lead in albacore:
Guide to lead pollution in Americans. Science 207:1167-1176.
Valentine, W. N., D. E. Paglia, K. Fink, and G. Madokoro. 1976.
Lead Poisoning. J. Clin. Invest. 58:926-932.
World Health Organization. 1977. Lead, Environmental Health
Criteria. World Health Organization, Geneva.
231
-------
DISCUSSION
DR. MURPHY: What are the significant differences in metabolism
of Kepone in rodent and humans? Is the deficiency of the animal
model one of metabolism or of target site? The estrogenic activity
of Kepone was demonstrated many years ago, as were other standard,
toxicologic effects. I do not know if opsoclonia was observed in
rodents.
DR. GOLDBERG: The strogenic activity of Kepone was
demonstrated in both rats and humans, but the metabolic pathway
present in humans is lacking in the rodent. Thus, some of the
problems caused by Kepone contamination are demonstrable in
rodents, and others may not be. Some of the consequences are
directly due to the Kepone itself, e.g., the estrogenic activity,
while other actions may be due to active metabolites.
DR. MILLER: Where did the Kepone come from in the 1960's
when the Life Science Product Corp. was not yet open?
DR. GOLDBERG: Before 1966, Kepone was manufactured by Allied
Chemical Corporation in Hopewell, Virginia.
DR. MILLER: The fish with differing Kepone levels at different
seasons—did they swim upside down or sideways? Were they obviously
ill?
DR. GOLDBERG: No. Nor do the fish kills during 1976 seem
related to the presence of Kepone, at least not superficially.
The fish in the early spring come from the Chesapeake Bay and are
232
-------
not contaminated. Like the bluefish, they ingest Kepone while
feeding as they enter the river, where they remain during the summer
months.
Some species, such as oysters and female crabs depurate their
Kepone. If they move to clean water, they rid themselves of the
compound very quickly. Other species tend to store it—in both
edible and nonedible parts.
DR. MILLER: Does Kepone affect a fish's reproductive
capacity?
DR. GOLDBERG: That has not been studied.
233
-------
Exposure to TCDD: Immunologic Effects
Girolamo G. Sirchia and the Group for Immunological Monitoring
Three groups of subjects exposed to 2,3,7,8-tetrachlorodibenzo-p_-dioxin
(TCDD) in Seveso were monitored: children 3 to 8 years of age; workers
at the ICMESA (Industrie Chimice Meda Societa Anonyma) plant where the
accident occurred; and soldiers who cordoned off the plant. A battery
of immunologic tests were repeated on the same subjects at different
times. Results were evaluated by comparing results from (1) exposed and
control groups, and (2) exposed chloracne and nonchloracne populations.
The children's group was more homogeneous with respect to TCDD exposure.
Thus, it was screened more frequently. Findings in that cohort are
reported here in more detail. Total serum complement hemolytic activity
(CH50) values were significantly higher in exposed children than in
controls, for all screenings. Children with chloracne had higher CH50
values than did children without chloracne. Exposed children had higher
lymphocytic responses to phytohemagglutinin and pokeweed mitogen than
did controls in the first three screenings; later screenings revealed a
trend toward higher values, but the increases were not significant.
Chloracne incidence did not seem related to these findings. The percentage
of children with increased CH50 values, lymphocytes of peripheral blood,
and lymphocytic response to lectins was repeatedly higher than expected,
especially among children with chloracne. Consistently increased (at
two or more screenings) values for tests were shown only for CH50, PEL,
and lymphocytic response to lectin, more evident among children with
chloracne. A few subjects consistently (at two or more screenings) had
values below the lower tolerance limits. Additional investigation is
needed, especially with tests more sensitive to slight modifications in
values.
1-The following investigators comprise the Group for Immunological
Monitoring: Drs. G. Sirchia (Director), V.E.M. Rosso di San Secondo,
F. Tedesco, C. Fortis, D. Alcini, A.M. Giovanetti, L. Guazzotti (Centre
Trasfusionale e di Immunologia dei Trapianti, Ospedale Policlinico,
Milano) (CT); Drs. C. Zanussi (Director), R. Scorza, G. Fabio, P. Bonara,
P.L. Meroni, M.G. Sabbadini, M. Vanoli, C. Pettenati (Istituto di Clinica
Medica IV, Universita di Milano, Ospedale Policlinico, Milano) (CM4);
Drs. P. Mocarelli (Director), A. Pessina, P. Brambilla, A. Marocchi
(Servizio di Patologia Clinica, Ospedale di Desio, Milano) (OD); Drs.
P. Careddu (Director), M. Bardare, F. Corona (Clinica Pediatrica,
Universita di Milano) (CP); Drs. G. Chiappino (Director), C. Nava,
G. Meregalli (Clinica del Lavoro, Universita di Milano) (CL); Drs.
S. Garattini (Director), F. Spreafico, A. Vecchi, A. Mantovani (Istituto
Mario Negri, Milano) (IMN) - only for the first year; and Drs. E. Marubini
(Director), S. Milan! (Istituto di Biometria e Statistica Medica,
Universitk di Milano).
234
-------
2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) has been reported to cause
abnormalities in the immune system in animals. These effects include
atrophy of thymus and peripheral nodes (Gupta et al., 1973; Vos et
al. , 1973) and decreased resistance bacterial infection (Thigpen et
al. , 1975). Other effects are depression of cell-mediated immune
functions, such as lymphocyte response to mitogens (Vos and Moore,
1974; Vos et^ al^. , 1973), skin graft rejection (Vos and Moore, 1974),
and delayed hypersensitivity responses (Moore and Faith, 1976).
More recently Faith et^ al. (1978) have reported that exposure to TCDD
causes suppression of some T-cell functions (such as delayed hypersensi-
tivity reactions and lymphocyte response to mitogens) but not of
others (such as the helper function). It is more difficult to extrapolate
this experimental information to humans.
Immunologic data on people exposed to dioxin are scanty, even
though dioxin contamination has occurred many times all over the
world. One of the major accidental releases of TCDD occurred in
Seveso, Italy, on July 10, 1976, when a cloud of vapor containing
hundreds of grams of TCDD polluted a densely populated area. Imniuno-
logical evaluation of individuals believed to have received the
greatest exposure has been performed. This paper reports the results
of the first 3 years of this monitoring.
Three cohorts were formed for this study. Some 48 children from
3 to 8 years of age, who lived in the most contaminated area, were
selected first because they had a (1) possibly higher TCDD:body
weight ratio; (2) possibly higher exposure to TCDD because they
played In the contaminated soil; and (3) high incidence of chloracne
(approximately 50%).
235
-------
For the second study group, 103 workers at ICMESA, the company
where the accident occurred, were selected. Although the accident
happened on a Saturday when the plant was closed, most workers were
probably highly contaminated because they lived in the polluted
area. Moreover, some of them worked at the plant for a few days after
the accident. A third group consisted of 75 soldiers who cordoned off
the contaminated area. The TCDD exposures in these two groups of adult
subjects were less homogeneous than those in the group of children.
Thus, this report focuses on the results obtained in children.
Control subjects were not the same persons at each screening,
except for approximately 20% of the children, because of compliance
problems encountered with this population. Control subjects were
apparently healthy people who lived in uncontaminated areas of Greater
Milan. Control children were selected from schools in Lissone.
Control adults were chemical workers who attended the Clinica del
Lavoro in Milan for medical examinations.
The average age of the controls was close to that of exposed
subjects within each sex (Table 1). The same procedures were used
to collect and analyze blood samples of control and exposed subjects.
Thus, any systematic technical errors should affect results from
both groups in the same way.
The monitoring protocol included the periodical performance of
the following screening tests:
• absolute number of lymphocytes of peripheral blood (PEL);
• titer of antisheep red blood cell antibodies;
• titre of A and B isohemagglutinins;
• serum immunoglobulin concentration (Ig);
• total serum complement hemolytic activity (CH50);
• percentage of erythrocytes (E) rosette-forming cells (E-RFC);
-------
Table 1
Ages, Sex, and Number of Individuals in Exposed and Control Groups'
Median Age (and Range of Ages) and Number (N) of Subjects
Children
Dates of
Screenings
November 1976
February 1977
May 1977
December 1977
May 1978
October 1978
May 1979
Sex
F
M
F
M
F
M
F
M
F
M
F
M
F
M
Exposed
4
N
5
N
5
N
5
N
5
N
5
N
5
N
6
N
6
N
7
N
7
N
7
.5
.5
(
.5
=:
.5
.5
(2.5-8)
17
(2.5-11)
27
2.5-8)
15
(3-13)
31
(3-8.5)
16
(3-13.5)
29
.5 (3.5-9)
= 17
.5(3.5-13.5)
= 28
.5
(
(4.5-9.5)
15
4-14)
28
-
.5(4.5-10.5)
= 15
.5 (5-15)
ICMESA Workers Soldiers
Controls Exposed Controls Exposed
5
N
5.
N
5
N
6
N
6
N
6
N
5.
N
6.
N
6.
N
7
N
8
N
7.
(3-8.5) -
= 15
5 (3-8.5) -
=27
(3.5-8) - -
= 17
(3-12.5) -
= 29
(2.5-9) - -
= 19
(3.5-11) - -
= 28
5 (4-9.5) - -
= 17
5 (4-9.5) -
= 29
5 (4.5-9.5) - -
= 15
(4.5-13) -
= 26
35 (19-56) 33.5 (18-50)
N = 4 N = 10
39 (19-65) 35.5 (23-62)
N = 91 N = 76
(6-11) - -
= 10
5 (5-14.5) - - - -
Controls
—
-
-
-
-
-
-
-------
Table 1 - continued
February 1979 F
M
22 (20-47) 22 (19-37)
N = 75 N = 64
May 1979
M
34 (19-56) 29.5 (20-41)
N = 3 N = 6
39 (18-65) 39 (19-67)
N = 76 N = 80
to
00
-------
• percentage of zymosan-complement rosette-forming cells (ZyC-RFC);
• lymphocyte response to phytohemagglutinin (PHA);
• lymphocyte response to pokeweed mitogen (PWM); and
• lymphocyte response to alloantigens in the mixed lymphocyte
culture (MLC).
Each screening lasted approximately 1 month; approximately five
exposed and five control subjects were examined each day. The tests
were blind, and technicians did not know whether samples had been
obtained from exposed or control subjects. Rosette determination
and lymphocyte transformation tests were conducted in two different
laboratories to allow for evaluation of laboratory performances.
Tests were performed as follows:
The absolute number of PBL's was obtained from routine hematologic
data (Dacie and Lewis, 1975). To establish the titer of antibodies
against sheep erythrocytes and A and B isohemagglutinins, 100 yl of
a twofold serial dilution of the test sera, in phosphate-buffered
saline, pH 7.4, and containing 0.01 M EDTA, were mixed with 50 yl of
a 1% suspension of sheep erythrocytes. Incubation was carried out
strictly at 20°C for 1 hour. Agglutination was read microscopically-
and the titer of the antibodies was defined as the reciprocal of the
highest dilution of the serum that was still able to induce formation
of numerous clumps of three or more cells. The same procedure was
followed for the titration A and B isohemagglutinins, except that 1%
suspensions of Al, B, and 0 human erythrocytes were used.
Serum immunoglobulins IgG, IgA, and IgM were quantified by the
single radial diffusion procedure (Mancini £££!_•> 1965), using agar
gel plates (Quantiplates Kallestad Laboratories, Chaska, Minn.).
239
-------
The diameters of the precipitin rings were compared with those of known
standard reference preparations. Serum IgE levels were measured by
radioimmunoassay (Nye e_t^ a^., 1975).
The Lachmann and Hobart (1978) technique was used to measure serum
complement hemolytic activity. Veronal buffered saline, pH 7.4, mixed
with an equal volume of 5% glucose and containing 0.15 mM calcium and
Q
0.5 mM magnesium, was used as diluent. Sheep red blood cells (3 x 10
ml) were sensitized by eight hemolytic doses of rabbit IgM antisheep
erythrocytes. An equal number of sera of exposed and control subjects,
as well as a pool of sera from 20 healthy blood donors (stored frozen in
liquid nitrogen), was tested each day under the same experimental condi-
tions. The results were expressed as CH50 units per milliliter of
serum, 1 unit representing the amount of serum required to lyse 50% of
sensitized erythrocytes. To compute CH50 units, the observed percentages
of lysis at diffeent concentrations of test serum were fitted by a
logistic curve according to maximum likelihood principle (following
Bliss, 1970).
The test for E-RFC was performed according to Jondal et_ £l. (1972),
with slight modifications. Sheep red blood cells (SRBC) were collected
in Alsever's solution and stored at 4°C for no longer than 1 week.
They were washed five times in buffered salt solution (BSS) and adjusted
to a concentration of 10% in BSS containing 20% heat-inactivated fetal
bovine serum (FBS) absorbed with SRBC. The lymphocyte suspension was
adjusted to a concentration of 2 x 10 cells per milliliter, in BSS
containing 20% heat-inactivated FBS absorbed with SRBC. Equal volumes
(0.1 ml) of SRBC and lymphocytic suspensions were mixed in round-bottom
tubes. These were incubated for 37°C for 15 minutes before centrifugation
240
-------
at 200 g for 10 minutes at room temperature. Tubes were incubated
overnight at 4°C. The pellet was gently resuspended with a Pasteur
pipette and mounted on glass slides. Some 300 viable cells were counted.
Those that bound at least three SRBC's were considered to be E-RFC's.
The test for ZyC-RFC was performed according to Huber and Wigzell
(1975), with slight modifications. Plastic tubes (Falcon Plastic, Los
Angeles, Calif.) containing a mixture of 25 yl of lymphocyte suspension
(4 x 106 cells per milliliter) and 25 yl of baker's yeast (0.5 rag/ml),
prepared according to Hadding et al. (1967) and coated with mouse
complement (1 mg of yeast for 0.05 ml of serum), were centrifuged at
90 g for 5 minutes at room temperature and kept in an ice bath for 30
minutes. The pellet was gently resuspended with a Pasteur pipette and
mounted with 0.2% trypan blue on glass slides. Some 300 viable cells
were counted. Those binding at least three yeast beads were considered
to be ZyC-RFC's.
Tests for PHA and PWM stimulations were performed according to
Greaves et al. (1974) in round-bottom plastic plates (Falcon Plastic)
containing 0.5 x 105 cells in 0.4 ml of HEPES [4-(2-hydroxyethyl-l-
piperazineethanesulfonic acid] vated FBS. All tests were performed in
triplicate. PHA (Wellcome purified PHA, Wellcome Italia) and PWM
(GIBCO, Bio-Cult Ltd., Scotland) were used at concentrations of 0.1%,
1%, and 5%. Blood lymphocyte preparations were incubated for 3 days.
Approximately 12 hours before the end of the culture period, 0.5 Ci
of ^H-thymidine (Amersham, England) were added to each well. Cells
were harvested on glass fiber filters, and radioactivity was measured
in a liquid scintillation counter (Packard Instrument Co., 111.).
Tests for MLC stimulation were performed according to Helgelsen
e£ al. (1973), with slight modifications. Round-bottom plastic plates
(Falcon Plastic) contained 0.5 x 10$ responding and 0.5 x 105 stimulating
241
-------
cells per well in 0.4 ml of HEPES (40 mM) buffered RPMI 1640, supplemented
with 10% heat-inactivated human AB serum. All tests were performed in
triplicate. Stimulating cells (pool of frozen and thawed cells from five
unrelated donors) were inactivated by irradiation (4,000 rads). Cultures
were incubated for 6 days. Approximately 12 hours before the end of the
3
culture period. 0.5 yCi of H-thymidine were added to each well. Cells
were harvested on glass fiber filters, and radioactivity was measured in a
liquid scintillation counter.
For the chilren, two approaches were devised to obtain surveywide
information and to evaluate each subject:
1. Findings from the following groups were compared at each screening
of exposed subjects vs controls, and exposed subjects with chloracne vs.
exposed subjects without chloracne (Scheffe, 1961).
2. A flowchart of results obtained at different times was set up for
each exposed subject and for each immunologic test. Data from each test
obtained at each screening, for every single subject, were plotted,
together with their reference values. Only data from laboratories that
performed at least four screenings were included on the charts.
Since the adults had received only one or two screenings, the approach
to their analysis was limited to comparisons of results from exposed and
control subjects.
Results of all the tests performed at each screening were evaluated,
using analysis of variance techniques. For most of the immunologic tests,
the experimental error is positively skewed and its dispersion is related
to the mean. Thus, the assumptions underlying the analysis of variance are
not fulfilled (Snedecor and Cochran, 1972). Hence, original values were
first transformed, as indicated in Table 2, to distribute errors symmetrically
242
-------
and to make them independent of the mean. A 10% level of significance
(two tails) was adopted as a threshold to improve the power of the test.
To construct the individual flowcharts, reference values were given
in terms of tolerance limits (Guenther, 1977). The limits were computed
with a confidence level of y = 0.95. Thus, at least a prefixed percentage
(_P) of the control population would respond within the limits if their
tests were performed with the same precision and accuracy as those of
the study subjects.
Tolerance limits at each screening (TL ) were computed as follows,
8
TL (y,P) = x~+ r (n,P) u(f,j).s
o — S
where: x is the average response of the control group; r is a function of
the number (n) of control subjects and of the percentage (P) of the
population included between the limits: P values were 80%, 90%, and 98%;
u is a function of the degrees of freedom (f) of s and of confidence level
S
Y (values of r and u are reported in Owen, 1962);
s is the total standard deviation of the control group and is an estimate
S
of biologic variability "between control subjects" plus technical
errors "within day" and "between days."
As n and f increase, r tends toward the percentile of gaussian distri-
bution corresponding to P, and u tends toward 1. However, s is particularly
S
high in some tests, such as in the lymphocyte transformation analyses
(response to lectins and alloantigens). Therefore, the results obtained
for the same subject under the same experimental conditions, but in
different days, are highly discordant. The discrepancies are lower
among data obtained from subsamples of blood taken from the same subject
at the same time and examined separately (Rosso di San Secondo et^ al. ,
1979). Table 3 shows details of these variabilities.
The major component of technical variability is "between days." The
reference values obtained on different days were widely spread. Thus,
243
-------
TABLE 2
Transformation of the Data for Symmetry of
Distribution and Homogeneity of Variance
Immunologic Test Expression of Results
Lymphocytes
IgG
IgA
IgM
IgE
CH50
Anti-A, RBC
Anti-B RBC
Antisheep RBC
E-RFC
ZyC-RFC
PHA
PWM
MLC
no./yl
mg/dl
mg/dl
mg/dl
I.U./ml
U/ml
titer
titer
titer
%
%
c . p. m.
c. p.m.
c . p .m.
Metanieter
Square root
Logarithmic
Logarithmic
Logarithmic
Logarithmic
Logarithmic
Logarithmic
Logarithmic
Logarithmic
Angul ar
Angular
Logarithmic
Logarithmic
Logarithmic
244
-------
only gross abnormalities could be detected. To counteract this limita-
tion, the results of lymphocyte transformation tests of exposed
subjects were also evaluated in relation to only the values obtained
in control subjects tested on the same day in the same experiment.
This approach introduced a new source of variability. Different
controls were used for each day of screening; but the size of this
variability is still lower than that of the variability "between
days." Its effects can also be reduced by selecting the control
subjects and increasing their number. Table 4 shows how this approach
allows the reduction of variabilities and how that reduction is a function
of the number of controls (Rosso di San Secondo and Milani, unpublished
data).
Tolerance limits for the lymphocyte transformation were obtained
for each day of analysis, as follows:
TLd(Y,P) = xd + r(ndP) u(f p,y ) Sp
where x is the average response of the n, control subjects tested on the
d-th day;
s is the standard deviation "within day" with f degrees of freedom
that estimates biologic variability "between control subjects"
plus technical errors "within day."
This approach was used only for lymphocyte transformation tests. No
information is| available on the relative importance of the different
components of variability for the other immunologic tests.
Thus, for all the immunologic tests performed, reference values were
obtained at each screening from the values of all the control subjects
examined at that screening (within-screening tolerance limit = TL ).
S
For lymphocyte transformation tests, reference values were also
calculated daily from the values of the control subjects examined
that day (within-day tolerance limit = TLd).
245
-------
TABLE 3
Contribution of Each Indicated Source of Variability to Whole Error3
Source of variability
Replicates
Subsamples
Days
Whole error
MLC
1.6 (5)
2.7 (45)
2.8 (50)
6.2 (100)
PWM
1.8 (1)
2.8 (8)
9.5 (91)
10.4 (100)
PHA
1.1 CD
2.7 (19)
5.5 (80)
7.5 (100)
Results are expressed as coefficients of variation; figures in parenthesis
indicate the percentage of whole error.
246
-------
For the comparison between exposed and control subjects, 140 F tests
of significance were performed at six screenings of 10 tests each. (Lympho-
cyte response to lectins was tested at three lectin concentrations, and
some tests were carried out by two laboratories concurrently.) Another
140 significance tests were performed to compare findings from exposed
subjects with and without chloracne. The results can be summarized as
follows:
Under the hypothesis that no real difference exists between the two
groups of children, it is expected that, by chance, out of 140 tests of
significance, 1 or 2 are significant at the 1% level, 7 at the 5% level,
and 14 at the 10% level. The observed significant differences are 7,
16, and 25 respectively. Moreover, all the significant differences
(except one) are in the same direction (values of exposed children are
higher than those of controls).
The highest number of significant differences was observed in the
first three screenings (performed between November 1976 and May 1977).
In fact, 19 of the 80 tests performed in the first three screenings were
significant at the 10% level. Only eight had been expected.
CH50 activity was unique in that exposed subjects had values
significantly higher than did controls in all the screenings. CH50
was significantly higher (at the 1% level) only at the first screening
in children with chloracne (E+) than in those without chloracne (E~).
However, if the differences obtained in the remaining five screenings
are included, children with chloracne apparently had mean values
higher than did children without chloracne. The probability of
this occurring by chance is low (P = 0.031), indicating that
children in screenings 2 to 6 had a E > E~ tendency (Figure 1)
247
-------
Exposed children showed PHA values significantly higher than did con-
trols at the first screening (performed at the Istituto Mario Negri) and
at the second and third screenings at Ospedale di Desio (Figures 2, 3,
and 4). If, however, results obtained at the three PHA concentrations
at each of the six screenings at Ospedale di Desio are considered together,
the averages of the values for exposed subjects are higher than those of
controls 16 out of 18 times (P =0.001). A similar tendency is not evident
for studies at the Istituto Mario Negri, but only the first three screenings
were performed there. Only at the second screening at Ospedale di Desio
did children with chloracne differ from children without chloracne signifi-
cantly. There was no tendency for this difference to occur at any of
the other screenings.
Exposed children showed PWM values significantly higher than those of
controls at the first screening (at the Istituto Mario Negri) and at the
second and third screening at the Ospedale di Desio (Figures 5, 6, and
7). Again, there seems to be a tendency for exposed subjects to have
values higher than controls (16 out of 18 times; P =0.001). In this
test also, children with chloracne did not seem to differ from children
without chloracne.
Exposed subjects showed PEL values higher than did controls at each
screening, but the difference was significant only at the second and
sixth screenings (Figure 8).
Except for CH50, number of PBL's, and lymphocyte response to lectins,
test results in exposed subjects were not clearly different from control
values, even though sporadic significant differences have been observed.
248
-------
The individual flow charts indicated which subjects and which immuno-
logic tests had values outside the tolerance limits. They also suggested
a pattern of differences (from the mean of control values) in the 3-year
period. Tables were prepared from the charts to highlight single data
elements.
The percentage of exposed children with values below the 10th and
above the 90th percentiles of control value distribution was calculated
for each test at each screening. Analysis of the data indicated that
only values for CH50, PBL, and lymphocyte response to PHA and PMW are
of interest, of these two percentile groups. In fact, the percentage
of subjects with values that exceed the 90th percentile is repeatedly
higher than expected for these tests, and is more evident in children
with chloracne.
It also became apparent that the calculation of tolerance limits
from controls examined within-day (TL,) affords a more powerful means
to detect abnormalities, provided that at least five control subjects
are tested at the same time as the exposed subjects (Table 4). In
fact, on average, the percentage of subjects with values outside the
percentiles (using TL,) is approximately 30% higher than that
obtained using TL .
s
The number of exposed children who, at two or more screenings, had
values lower than the 10th, 5th, and 1st percentiles or greater than
the 90th, 95th, and 99th percentiles in each of the immunologic
tests performed was evaluated.
249
-------
TABLE 4
Error of Difference in Absolute Value of the Test Subjects, Minus Average
Value of Controls Tested Contemporaneously, as a Function of Number of Controlsa
No. of Controls
1
2
3
4
5
6
7
MLC
7.1
6.4
6.1
5.9
5.8
5.7
5.6
PWM
8.4
7.4
7.0
6.7
6.6
6.4
6.3
PHA
8.1
7.0
6.5
6.3
6.1
5.9
5.8
Results expressed as coefficients of variation.
250
-------
Increased values as reported by the same test, in the same subject,
at two or more screenings, were consistently found only for CH50,
number of PBL's, and lymphocyte response to lectins, and was more
evident in children with chloracne. In fact, seven subjects (four
with chloracne) showed values of CH50 higher than the 90th percentile
at two or more screenings (expected number = 5.8; 2.6 with chloracne).
Of these seven subjects, two (both with chloracne) showed values higher
than the 99th percentile (expected numbers, 0.1; 0.03 with chloracne).
No subjects had values lower than the 10th percentile.
In the PBL analyses, 11 subjects (8 with chloracne) had values higher
than the 90th percentile at two or more screenings expected number
same as above. Of these, seven (five with chloracne) showed values
higher than the 95th percentile (expected number = 1.5; 0.7 with
chloracne).
Considering tolerance limits obtained from TL, controls, seven
children (all with chlorance) had values of lymphocyte response to
PHA 5% higher than the 90th percentile at two or more screenings
(expected number, 5.8; 2.6 with chloracne). Similar data were recorded
when lymphocyte responses to PHA, 0.1%, and PMW, 1% and 5%, were
considered. Finally, a few subjects, at two or more screenings and
in some tests, had values below the lower tolerance limits. Such
results occurred especially for the levels of IgA and IgM (for which
one and two subjects, respectively, repeatedly fell below the 1st
percentile) and, to a lesser extent, for the titer of antisheep RBC
antibodies and for the lymphocyte response to PWM, 0.1% and 1%.
251
-------
For the groups of ICMESA workers and soldiers, the comparison
between exposed vs control subjects at each of the screenings and
for the 10 immunological tests gave a number of significant differences
compared to those expected by chance.
Has the TCDD released at Seveso caused some transient or permanent
alteration to the immune status of exposed people? The study group
tried to answer the question by concentrating on results obtained from
monitoring children, probably highly exposed to TCDD, for a 3-year
period.
Data indicate that exposed subjects showed a CH50 significantly higher
than that of controls at each of the six screenings performed. They
also had values of lymphocyte response to lectins that were higher
than those of the controls at each screening, although significant
differences were observed only in the first three screenings. Exposed
subjects also tend to have higher PEL values than do controls.
These findings are somewhat surprising because experiments in animals
indicate that TCDD has an immunosuppressive effect (Faith et al.,
1978; Moore and Faith, 1976; Vos et al., 1973; Vos and Moore, 1974).
However, those experiments were performed under different conditions
(dose, timing, route of exposure) than the ones discussed in this
paper. These variables can influence whether physical and chemical
agents, induce either stimulation or suppression, thereby affecting
test results. Furthermore, animal data on CH50 are lacking.
252
-------
Nonetheless, the data from this study must be interpreted with care;
every effort was made to obtain the best possible selection of control
subjects, but they cannot be considered ideal from an epidemiologic
point of view. Secondly, data were not obtained from a specifically
designed experiment, but from an observational study, which is exposed
to a high risk of erroneous conclusions.
If systematic differences between exposed and control subjects
not due to TCDD exposure can be ruled out (as may be suggested by
the fact that increased values seem to appear more frequently in
children with chloracne than in those without), the mechanism of the
increases becomes an interesting subject for speculation. Increased
responses of lymphocytes to lectins could result either from a direct
stimulating effect of TCDD on lymphocytes or from an indirect action.
Vos and Moore (1974) reported that lymphocyte response to lectins in
vitro is not modified by the presence of TCDD, but there is some
evidence (Pandian and Talwar, 1971) that TCDD in vivo stimulates the
lymphocytic incorporation of thymidine through the release of somatotrophin
by the pituitary gland. More recently, however, Vecchi et al. (1980)
have reported that the in vitro response of mouse lymphocytes to
Concanavalin A increases significantly in the presence of TCDD, if
appropriate amounts of the chemical are used. Alternatively, increased
lymphocytic responses to lectins could be the result of the depressive
action of TCDD on some suppressor mechanisms, which has been shown
to occur for small-dose radiation (Anderson and Lefkovits, 1979).
253
-------
The increased CH50 hemolytic activity could be explained by the
increased production of complement components as well as other "acute
phase proteins" by the liver in response to tissue damage by TCDD. If
TCDD does cause the modifications observed in vitro, one must then
determine their clinical importance. Immune reactivity may have increased;
however, elevated rates in some tests do not necessarily imply increased
immune reactivity. An increase of some indices, such as the level of
immunoglobulins, is even found in some congenital immunodeficiencies
(World Health Organization Scientific Group on Immunodeficiency, 1979).
Moreover, even if the increases observed in this study are the result of
increased immune reactivity, immunodepression may follow the stimulation.
Thus, the few subjects with decreased IgA, IgM, antisheep RBC antibodies,
and PMW values should not be disregarded.
During the 3-year observation period, no particular clinical event
besides chloracne seems to have occurred in the exposed children.
Thus, it is difficult to correlate in vitro and in vivo events and to
draw conclusions about the clinical value of the data from in vitro
studies. This information may be obtained only by continuing appropriately
planned immunologic and clinical investigations of exposed children
for many years.
Present studies indicate several areas requiring modification in
future investigations. Some tests used in this investigation show a
variability that could obscure slight modifications occurring over a
period of time. An jid hoc experiment to quantify the principal sources
of variability of the most variable tests (such as the lymphocyte
254
-------
transformation analyses) was performed (Owen, 1962; Rosso di San Secondo
and Milani, 1979), and the information collected was used to improve
the design of the protocol. It is clear, however, that tests for
immunologic monitoring must be made more sensitive to slight varia-
tions. It is difficult to monitor the immune status of a large
number of subjects effectively, i.e., to reconcile effectiveness and
feasibility. The protocol in this study seemed the only one possible
3 years ago. A different set of immunologic tests might have provided
a deeper insight into the problem. Other tests should be added,
including some sophisticated ones, to evaluate T and B lymphocytes and
phagocyte function. In addition, this study examined only a selected
group of exposed people. It was assumed that, if no abnormality
became apparent in the children, then abnormalities would probably
not be detected in cohorts that were exposed to lower concentrations.
Regardless of the limitations, the Seveso study has taught at least
three major lessons:
1. A permanent investigative group, able to cope with future areawide
chemical contamination, should be established.
2. Tests used must detect slight modifications of the immune status
of exposed subjects.
3. The clinical significance of the in vitro findings must be
investigated further.
4. Accidental environmental contaminations are occurring with
increasing frequency. Thus, national systems to evaluate the immune
status of exposed populations and international cooperation are
needed.
255
-------
CH50
36,
34.
o>
£ 33
O)
o
3.1.
30
B
1976
1977
1978
1979
14,
1.2
10
I
FIGURE 1 - Group of children. Complement activity.
Upper portion of figure: Comparison between control (left-hand box) and
exposed children (right-hand box) at each screening. Boxes indicate the
mean and the confidence limits for exposed subjects with (interrupted
line) and without (continuous line) chloracne. • significant at the
10% level; •• significant at the 5% level; ooo significant at the 1%
level.
Lower portion of figure: As above, but control values are indicated as
a baseline so that differences from controls occurring in exposed children
(vertical lines) are more evident. Ordinates are the ratio of mean value
of exposed subjects to mean value of controls with the 95th confidence
limits.
256
-------
E
OL
o
O)
O
46.
44
4.2
40
38
36
34
PHA 0.1 /
'
I
[
i
•
1976
1977
1978
1979
30
20
1.5-
ra
05
FIGURE 2
Group of children. Lymphocyte response to PHA 0.1%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
257
-------
PHA 1 /
48-
46.
£ 4.4.
a
0
Q) A o
o 42
1
_J
40.
38-
T
l
4-T
L
T
1
4
«
»•
T
1
1
•
r-
l
».
t
ff-
i n !
i *
i r
4-, L
Bl ~ "
T
t
1 •— '
1 '
—
• •
f
1976
1977
1978
1979
3.0,
25
20
15
1.0.
0.5J
FIGURE 3
Group of children. Lymphocyte response to PHA 1%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
258
-------
4.2
1976
PHA 5X
40.
TT
1
nlil
e as.
Q.
6 1
0) 3.6.
34.
32.
_f
i
U.
•
•
W
T
1 T
i T
i rr
i_ T !
r*~
i
IT
1
i.
!. n HI
t
i
i
L
4-
— l
l__
1 '
1977
L
1978
t
i
1979
35
30-
25,
20,
05,
FIGURE 4
Group of children. Lymphocyte response to PHA 5%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
259
-------
PWM 0.1 X
40
38
36
E
CL
0
34
0)
0
_j
32
30
28
1T
ffl
n i
T
~~ 1 T
TT r*-
! ! i
1 ' i
r~i i~ i i
I «
• r~i ,- ~ »-
1 T 1
— UJ i i
i tit U9 n
U } M- !
A i •
i i ~~
T
1-
"1 *
1
t
— \i.
i
T
I
1
1
i
i
i
1976
1977
1978
1979
25,
20
1.5-1
10
05J
FIGURE 5
Group of children. Lymphocyte response to PWM 0.1%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
260
-------
4.4-
PWM 1X
4.2-
4.0-
O)
O
3.8-
3.6.
34-
3.2
•
I
1976
i
1977
1978
1979
2°1
1.5-
10
0.5.
FIGURE 6
Group of children. Lymphocyte response to PWM 1%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
261
-------
PWM 5X
o
44
42
40
38
36
34
32
30
••
T
1976
1977
1978
1979
2
FIGURE 7 - Group of children. Lymphocyte response to PWM 5%
(performed at Ospedale di Desio).
For the legend, see Figure 1.
262
-------
80,
PBL
o
g 70
0)
CO
6O
CD
CL
50-
\.
0
e
40
1976
1977
1978
1979
1.5,
1.4
1.3.
1.2.
1.1.
ra -
0.9
0.8
FIGURE 8 - Group of children. Number of peripheral blood lymphocytes.
For the legend, see Figure 1.
263
-------
REFERENCES
Anderson, R.E. , and I. Lefkovits. 1979. In vitro evaluation of radiation-
induced augmentation of the immune response. Am. J. Pathol. 97:456-472.
Bliss, C.I. 1970. Pp. 172-178 in Statistics in Biology. Vol. II. McGraw
Hill, New York.
Dacie, J.V., and S.M. Lewis. 1975. Pp. 46-51, in Practical Haematology,
5th edition. Churchill Livingston, Edinburgh.
Faith, R.E., M.I. Luster, and J.A. Moore, 1978. Chemical separation of
helper cell function and delayed hypersensitivity responses. Cell
Immunol. 40:275-284.
Greaves, M., G. Janossy, and M. Doenhoff. 1974. Selective triggering of
human T and B lymphocytes in vitro by polyclonal mitogens. J. Exp.
Med. 140:1-18.
Guenther, W.C. 1977. Sampling inspection in statistical quality control.
P. 155 in Griffin's Statistical Monographs and Courses, No. 37. Griffin,
London, England.
Gupta, B.N., J.G. Vos, J.A. Moore, J.G. Zinkl, and B.C. Bullock. 1973.
Pathologic effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin in laboratory
animals. Environ. Health Perspect. 5:125-140.
Hadding, U., D. Bitter-Suermann, and F. Melchert. 1967. A tool for the
detection of C'6 deficiencies. Pp.319-321 in H. Peeters ed. Vol. 17.
Protides of the Biological Fluids. Elsevier, Amsterdam.
Helgesen, A., H. Hirschberg, and E. Thorsby. 1973. Modified micro-mixed
lymphocyte culture technique. P. 75 in J.G. Ray, D.B. Hare, and D.E.
Kayhoe, eds. Manual of Tissue Typing Techniques. DHEW (NIH) publication
74-545.
Huber, C., and H. Wigzell. 1975. A simple rosette assay for demonstration
of complement receptor sites using complement-coated zymosan beads.
Eur. J. Immunol. 5:432-435.
Jondal, M., G. Holm, and H. Wigzell. 1972. Surface markers on human T and
B lymphocytes. Part I, A large population of lymphocytes forming
non-immune rosettes with sheep red blood cells. J. Exp. Med. 136:207-215.
Lachmann, P.J. and M.J. Hobart. 1978. Complement technology. Chapter 5A
in D.M. Weir, ed. Handbook of Experimental Immunology, 3d ed. Blackwell
Scientific Publications, Oxford, England.
264
-------
Manclni, G., A.O. Carbonara, and J.F. Heremans. 1965. Immunechemical
quantitation of antigens by single radial immunodiffusion. Immunochemistry
2:235-254.
Moore, J.A., and R.E. Faith. 1976. Immunologic response and factors affecting
its assessment. Environ. Health Perspect. 18:125-131.
Nye, L., T.G. Merret, J. Landon, and R.J. White. 1975. A detailed investigation
of circulating IgE levels in a normal population. Clin. Allergy 5:13-24.
Owen, D.B. 1962. Handbook of Statistical Tables. Addison-Wesley Publishing Co.,
Reading, Mas s.
Pandian, M.R., and G.P. Talwar. 1971. Effect of growth hormone on the metabolism
of thymus and on the immune response against sheep erythrocytes. J. Exp. Med.
134:1095-1113.
Rosso di San Secondo, V.E.M., S. Milani, C. Fortis, E. Marubini, and G. Sirchia.
1979. The variability of lymphocyte transformation tests. Transplant Proc.
11:1379-1380.
Scheffe, J. 1961. Pp. 66-83 in The Analysis of Variance. John Wiley & Sons,
New York.
Snedecor, G.W., and W.G. Cochran. 1972. Statistical Methods. The Iowa State
University Press, Ames, Iowa.
Thigpen, J.E. , R.E. Faith, E.E. McConnell, and J.A. Moore. 1975. Increased
susceptibility to bacterial infection as a sequela of exposure to 2,3,7,8-
tetrachlorodibenzo-p-dioxin. Infect. Immun. 12:1319-1324.
Vecchi, A. A. Mantovani, M. Sizoni, W. Luini, M. Cairo, and S. Garattini.
1980. Effect of acute exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
on humoral antibody production in mice.
Vos, J.G., and J.A. Moore. 1974. Suppression of cellular immunity in rats and
mice by maternal treatment with 2,3,7,8-tetrachlorodibenzo-p-dioxin.
Int. Arch. Allergy Appl. Immunol. 47:777-794.
Vos, J.G., J.A. Moore, and J.G. Zinkl. 1973. Effect of 2,3,7,8-tetrachlorodibenzo-
p-dioxin on the immune system of laboratory animals. Environ. Health. Perspect.
5:149-162.
World Health Organization Scientific Group on Immunodeficiency. 1979. Immuno-
deficiency Clin. Immunol. Immunopath. 13:296-359.
265
-------
DISCUSSION
DR. MILLER: What about the ascertainment of mortality in the
population? Is mortality being studied?
DR. SIRCHIA: No, this cohort is very small. I believe that mortality
changes will become apparent over a very long period. This is also true
for morbidity.
DR. DARDANONI: Dr. Sirchia is speaking about mortality within the
sample studied, a rather small sample of the exposed population. I
think you are referring to the entire population. Mortality and hospital
morbidity are going to be monitored for the entire population through an
assessment of general health indicators, such as mortality by age and
sex, mortality from specific causes, and morbidity severe enough to
require hospitalization. Data have already been collected, but not yet
analyzed. There is circumstantial evidence that general morbidity has
not changed.
DR. MILLER: What of individual causes of morbidity? When people
develop infections, are the infections more severe if the people were
exposed or if they had chloracne or not?
DR. DARDANONI: I do not think there is any approach to evaluate
that.
DR. MOORE: Is any consideration being given to testing some of the
population with recall antigen response or something like it?
DR. SIRCHIA: It was scheduled, but it was impossible because people
did not comply. They did not want skin tines.
DR. MOORE: Does Italy vaccinate for TB? Could you use that as a
recall type?
DR. SIRCHIA: Yes.
DR. DARDANONI: Unfortunately, that has not been done.
266
-------
Somatic Cell Mutations
Arthur D. Bloom1
Environmental chemicals may induce both chromosomal and specific
locus mutations. Chromosomal mutations in mammalian somatic cells
include the classic forms of chromatid and chromosome-type lesions,
as well as the more recently described sister chromatid exchanges,
shown by differential staining techniques. Since specific locus
mutation rates cannot presently be studied in vivo, several loci (in
either cultured fibroblasts or cultured lymphocytes) must be analyzed
to identify in vitro mutagenesis with known or putative agents. The
two-step cancer hypothesis suggests that some individuals are geneti-
cally at increased risk of cancer. Induced somatic mutation in
these individuals is superimposed on an inherited genie or chromosomal
abnormality. Among the most promising direct mutation assays is one
that determines the frequency of cells deficient in HGPRT (hypoxanthine-
guanine phosphoribosyltransferase) among a wild-type population. At
this time, the determination of somatic mutation in populations
exposed to chemical mutagens is based on studies of clastogenesis
and sister chromatid exchange, rather than on specific locus assays.
Genetic alterations of somatic cells exposed to environmental
mutagens in vivo are of two basic kinds: chromosomal and genie.
Chromosomal mutations induced in mammalian somatic cells after such
exposures include the classic forms of chromatid- and chromosome-type
lesions, involving, respectively, one or both arms of the chromosomes,
which may undergo simple breakage or more complex breakage and re-
arrangement (Bloom, 1972).
In recent years, through the use of differential staining tech-
niques in which only one arm of the chromosome is stained, the propor-
tion of sister chromatid exchanges (SCE's) representing crossovers
Departments of Pediatrics and Human Genetics, College of Physicians
and Surgeons, Columbia University, New York.
267
-------
between the chromatids of the same chromosome has been determined.
Numerous reports have shown these SCE's to correlate well with the
induction of single gene mutations, particularly when exposure is to
chemical (rather than to physical) mutagens (Perry and Wolff, 1974).
A last chromosomal effect involves alteration of the mitotic apparatus
and resulting segregational errors. This leads to somatic nondisjunc-
tion and secondary trisomies and monosomies.
At the single gene level, a small number of loci can be used
to study forward and reverse mutation in cultured mammalian cells.
The principle, of course, is that the cells must be cultured; in
practical terms, the effects of environmental contaminants can be
studied in vitro, but as yet there is no technique to sample somatic
cells directly and to determine the mutation rate for a specific
locus in vivo.
The most promising direct assays are those involving a search
for individual cells with abnormal, induced, mutant hemoglobins
among a normal or wild-type population of red blood cells. Other
promising assays analyze the presence of hypoxanthine-guanine
phosphoribosyltransferase (HGPRT), the enzyme that is deficient
in the X-linked Lesch-Nyhan syndrome. In this latter system,
Albertini at the University of Vermont is attempting to isolate
HGPRT-deficient blood cells from HGPRT-positive cells. Both the
hemoglobin and HGPRT assays for direct mutation in somatic cells are
in early developmental stages, and patients exposed to radiation or
chemotherapy (i.e. , exposed to known mutagens) are being sought for
the study.
268
-------
In the meantime, the effects of chemical mutagens at specific
loci are being studied in cultured cells. The loci most commonly
used are (1) the hpt locus, which specifies HGPRT synthesis;
(2) the tk locus, which specifies the synthesis of thymidine
kinase; and (3) the oua locus, which specifies resistance or
sensitivity to the cardiomimetic compound ouabain, a membrane
function determined by the enzyme Na K ATPase. A few other loci
are sometimes used in these in vitro studies, but these three have
been examined most thoroughly.
Many biologic and clinical effects of compounds such as DBCP
(1,2-dibromo-3-chloropropane, an agricultural nematocide and known
carcinogen) and TCDD (2,3,7,8-tetrachlorodibenzo-p-dioxin; a highly
toxic contaminant formed during 2,4,5-trichlorophenol production)
have been reported. However, the somatic cell mutagenicity of these
and related compounds causes the most concern. As described in their
paper in this volume, Pocchiari ^t jil. have recently reviewed the
health effects of the Seveso accident involving TCDD, in terms of
chloracne, peripheral nervous system effects, the possible increased
incidence of abortions, and the possible increase in chromosomal
aberrations in persons with chloracne. The long range concern,
however, is cancer, because of the relationship between somatic cell
mutations and neoplastic disease.
The suggestion by Knudson that some cancer, specifically retino-
blastoma, develops through a two-step mutational process has led to
the more general notion that a significant proportion of human cancers
may result from a person's genetic predisposition, with or without
superimposed environmental insult (Knudson, 1979). What began as a
269
-------
mutational hypothesis for a specific tumor type can now, appropriately
modified, be accepted as a theoretical framework for numerous types
of neoplasms. Thus, the mutational origins of human cancers are
becoming clarified.
Some individuals are born with a genie or chromosomal mutation
that puts them, in utero and thereafter, at increased risk of neo-
plastic disease. The underlying mutation can be vertically trans-
mitted by parents who themselves may be at risk. Or, the inherited
genie or chromosomal mutation may have arisen de novo during gameto-
genesis. The genes involved may be autosomal recessive or dominant.
The recessive disorders studied most extensively to date are Fanconi's
anemia (FA), Bloom's syndrome (BS), ataxia telangiectasia (AT), and
xeroderma pigmentosum (XP). Deficiencies in DNA repair have been
demonstrated for all but BS. The most fully studied autosomal-dominant
disorders are retinoblastoma and Wilms's tumor. In each of these, one
(in the dominant disorders) or two (for homozygotes in the recessive
disorders and one in the heterozygous state) mutant genes are present
from conception.
Similarly, karyotypic abnormalities can exert a similar effect.
A deletion of chromosome 13 can be associated with retinoblastoma,
and a deletion involving chromosome 11 is found in some Wilms's tumor
cases. In addition, a chromosomal translocation (t3:8), seemingly
balanced, was observed in a family with numerous cases of renal
carcinoma. Translocation carriers have an 87% risk of renal carcinoma
by age 57 (Cohen et a^., 1979). Thus, some persons begin life with
a genetic predisposition to cancer; perhaps 1-2% of the people in
the U.S. population carry these genes.
270
-------
Superimposed on this background is somatic mutation. Some somatic
mutations are spontaneous—of unknown cause; others are induced by
identified environmental mutagens, either physical (such as ionizing
radiation) or chemical. Although it is likely that massive exposure
of a population to a strong mutagen will lead to carcinogenicity in
many persons with or without underlying genetic predispositions,
lower level exposures (particularly to mutagens of intermediate
strength) may well act selectively on genetically at-risk persons.
It is somewhat ironic that, despite an enhanced understanding
of the relationship between genie mutation and cancer, little more
can be done now to detect exposure to environmental mutagens (in
terms of somatic mutation) than was possible 15 years ago when the
peripheral blood lymphocytes of A-bomb-exposed persons were being
studied in Japan. Then, as now, cytogenetic monitoring was perhaps
one of the most sensitive measures of mutagenic exposure.
-In suspected incidents of areawide contamination, however, the
sampling of peripheral lymphoctyes for chromosomal mutation studies
is indicated. The proportion of aneuploid cells can be determined
from such studies as well as the proportion of cells with chromatid
and chromosome breaks and rearrangements. Clearly, SCE frequencies
after in vivo exposures must also be determined on these same blood
cells. Early studies on SCE's were conducted in cultured cells
exposed to varying doses of known mutagens. At the 1980 meeting of
the Environmental Mutagen Society, Carrano et^ ajL. (1980) described
data on SCE frequencies in petroleum refinery workers. Of 22 workers
271
-------
studied, 11 had SCE frequencies more than two standard errors above
the mean for the "nonexposed" group. The investigators reviewed the
personal health and work histories of the workers and concluded that
occupational or other environmental exposure (not smoking, medication,
and so on) was probably responsible for the increased SCE.
Although the issue of genetic screening in the workplace is
still controversial, identifying genetically at-risk persons may be
lifesaving. Judiciously done, such examinations are a serious
exercise in preventive medicine. For example, AT patients are known
to be peculiarly susceptible to the effects of X-rays and gamma
rays. Cells from FA patients are particularly susceptible to the
effects of DNA crossliriking compounds. And yet, of the chromosome
breakage syndromes, only XP cells have thus far been found found to
be hypermutable in vitro. BS patients, however, have the highest
spontaneous SCE frequencies and a marked response in terms of SCE
increases when their cells are exposed to mutagens.
Thus, when dealing with exposures of a large population, it
must be recognized that a subset (1-2%) of the population is more
at risk than the rest. Specific individuals may not be identifiable
by chromosome breakage and SCE studies after exposure. However, it
may soon be possible to identify these persons generally with ir\_
vitro tests that exert stress on cells by exposing them to chemical
mutagens (as illustrated by the effect of diepoxybutane on FA cells)
(Auerbach and Wolman, 1976).
In the next year or two, the Albertini HGPRT test should become
helpful in estimating in vivo somatic mutation rates. This test
272
-------
will use lymphocytes, incubated briefly in the selective medium
6-thioguanine (6TG), to isolate 6TG-resistant cells, which are hpt~,
forward mutants. The frequency of these cells in normal subjects is
proving to be 1/10-10 cells, a somewhat higher background rate
than desired, perhaps, but not unmanageable.
Although this method will be valuable, it measures effects only
at one locus, and might not provide reliable estimates of mutability
at other loci. Ideally, multiple loci should be studied, and different
kinds of genie products specified. The specific genes involved
should be analyzed by restriction enzyme analysis. These techniques
are not likely to be feasible in the near future because restriction
enzyme technology currently requires both a large amount of DNA and
specific mRNA probes (as is required for hemoglobin).
In summary, cytogenetic technology is still the best available
for assessing genetic changes in somatic cells of exposed persons.
In terms of specific locus mutation in somatic cells, known or putative
mutagens can only be studied in vitro.
273
-------
REFERENCES
Auerbach, A. D., and S. R. Wolman. 1976. Susceptibility of Fanconi's
Anemia fibroblasts to chromosome damage by carcinogens. Nature
261:494-496.
Bloom, A. D. 1972. Induced chromosomal aberrations in man. Pp. 99-172
in H. Harris and H. Hirschhorn eds. Advances in Human Genetics, Vol. 3.
Plenum Press, New York.
Carrano, A. V-, J. L. Minkler, D. G. Stetka, and D. H. Moore, II.
1980. Variation in the baseline sister chromatid exchange frequency
in human lymphocytes. Environ. Mutagen 2:325-337.
Cohen, A. J., F. P. Li, S. Berg, D. J. Marchetto, S. Tsai, S. C. Jacobs
and R. S. Brown. 1979. Hereditary renal-cell carcinoma associated
with a chromosomal translocation. N. Engl. J. Med. 301:592-595.
Knudson, A. G. 1979. Persons at high risk of cancer. (Editorial)
N. Engl. J. Med. 301:606-607.
Perry, P., and S. Wolff. 1974. New Giemsa method for the differential
staining of sister chromatids. Nature 251:156-158.
274
-------
DISCUSSION
DR. MILLER: What studies would you recommend to examine body
fluids or chromosomes of persons exposed to hazardous wastes to
determine the biologic effects of a chemical mixture?
DR. BLOOM: We have to distinguish between biologic and clini-
cal consequences. One major approach is to test for mutagenicity,
because the correlation between rautagenicity and carcinogenicity is
very high. So the approach is to examine body fluids, mixtures of
chemicals (as from Love Canal), or individual chemicals in multiple
tests across the prokaryotic and eukaryotic systems. For example,
urine concentrates can be studied in the Ames assay; suspected carcin-
ogens in certain fungi or yeast systems; and mammalian cell assays—
mouse cell and human cell—can be helpful.
Estimates of mutation frequencies ^.n vitro, how they apply ^n
vivo, of course, remain to be determined. Theoretically, it is
possible to determine whether or not individuals are exposed through
the use of chromosomal assays and SCE frequencies in short-term
cultured cells. Such evidence is, in a sense, all indirect. Most
of the testing has to be performed in vitro. But some of it can be
done directly, using these approaches. Somatic cell mutation studies
are very limited.
DR. MILLER: Suppose you test 50 people and find that the Ames
test is positive in 25 of the exposed group and in none of the control
population. What does this mean?
275
-------
DR. BLOOM: Well, the Ames test simply measures mutagenicity
in an isolated strain of Salmonella. It won't be a direct statement
of mutagenicity in vivo. A positive Ames assay says only that the
compounds in the body fluids are, in fact, mutagenic.
Are the concentrations in body fluids in vivo the same as the
concentrations producing mutation in vitro? There are usually
enormous differences that make it difficult to extrapolate results
from the in vitro test to the _in vivo situation. Mutagenicity in
the Ames test or in some other test systems usually occurs at dosage
levels considerably higher than those found in vivo.
DR. MILLER: Can you identify the agent in urine that causes
the Ames test to yield positive results?
DR. BLOOM: Yes, many chemical separation techniques can be
used.
DR. DELIA PORTA: Do you have any good examples of the correla-
tion of positive results for the Ames assay to an in vivo study of
the same individual?
DR. BLOOM: Not for _in vivo studies. Few studies have been
systematic enough to definitely show that. Studies only show that
a specimen is mutagenic in vitro.
DR. DELIA PORTA: Is there no example of the correlation based
on results from the same individual?
DR. BLOOM: No.
DR. DARDANONI: Every person has some chromosomal aberrations
of lymphocytes. Is it possible to choose a value that can discriminate
between most of the pathologic and normal aberrations?
276
-------
DR. BLOOM: Each laboratory has to standardize its own methods
and determine the normal frequency of aberrations for individuals
in that laboratory, particularly for persons of different ages.
There is a lot of interlaboratory variation; no absolute value can
be given for the frequency of chromosomal aberrations. Values should
be described as either abnormal or normal for a specific laboratory.
DR. DARDANONI: If a test is designed to identify susceptible
people, nonexposed people must be used as controls.
DR. BLOOM: Yes, that's true. The ideal is for an individual
to be his or her own control (before or after exposure), but that
doesn't usually happen.
DR. DELIA PORTA: Is the background noise higher when you
look for chromosomal aberrations or for increased SCE's?
DR. BLOOM: Usually there is more variation in chromosomal
aberration frequencies than there is in SCE frequencies. Again,
each laboratory has its own baseline, and the variation around it
tends to be much less than the discrepancy would be for a group
of laboratories. The variation is much less around SCE's, perhaps
because SCE's occur less frequently.
DR. NELSON: Tests on urine by an array of mutants, forward
or reverse, are essentially surrogates for chemical analysis, which
have a much greater sensitivity. Combining these tests with chemical
separations and, perhaps, more selective tests might be of great
value in analyzing exposure more sensitively. Screening body fluids
for mutagenicity only examines exposure.
277
-------
In some instances, the substances sought are transient, furtive,
or labile. The exposure to be measured can have occurred yesterday,
a few weeks ago, or even longer ago than that. Suppose that the
exposed population at Love Canal had been removed from the area
and kept in pristine, clean conditions. What approaches would you
recommend for an historic review of genetic injury?
DR. BLOOM: If you are looking for genetic injury, the only
persisting effects you are likely to see will be cytogenetic damage.
A great deal of the damage induced by low-level exposure is ultimately
repaired. The problem is to find a stable end point if you want to
go back in time. That end point clearly has to be in the lymphocytes,
which by chance are long-lived cells. Lymphocytes can be examined for
chromosome breakage for a long time after exposure. Among specific
genetic tests available now, an analysis of chromosome breakage is
probably the only one that can be used.
DR. NELSON: What specifically needs to be done now to detect
persistent genetic injury in somatic cells, that is, to identify
chromosomal abnormalities, not SCE's?
DR. BLOOM: If the cell has undergone more than two replications
after exposure, SCE's will not be detectable. It is now becoming
more clearly established that short-term, high-dose in vivo exposure
to mutagens does induce detectable SCE's. Demonstration of an increase
in SCE with longer term, low-dose, recurrent exposure is very likely,
but has not yet been shown. SCE will probably not be visible or
detectable many generations after high-dose exposure.
278
-------
Cell-survival curves are determined in an interesting way. If
the different types of chromosomal aberrations are classified accord-
ing to involvement of only one arm of a chromosome, or of both arms
(which produces rearrangements such as dicentric chromosomes) then
different kinetics are involved. The frequency of chromosomal
aberrations differs by type. The simpler kinds of chromosomal damage
are more easily repairable. More complex damage, for example, dicentric
chromosomes, tricentric chromosomes, rings, and translocations, tends
to persist, particularly if the cell is not undergoing many cell
divisions. Most lymphocytes in the peripheral blood are in G zero—
they are quiescent.
The phenomenon of amplification can be added to the baseline
level of induced aberrations. Under certain conditions, subpopula-
tions of the lymphocytes will be triggered to divide, for example,
by specific antigenic stimulation. In this way, chromosomally
abnormal cells can be induced to form clones. A single cell that
has a stable translocation, for example, can be antigenically
stimulated, by a variety of means, to proliferate in vivo. Suddenly
there may be not just a single copy of that cell in a million lympho-
cytes but several thousand copies. Is that a mutant, malignant
clone? Some evidence indicates that such clone formation may well
be related to the development of specific kinds of cancers that arise
from single cells.
The biology, then, involves not only repair of damage to DNA
with a resultant decrease in the frequency of chromosomal aberrations,
but also an increase in the frequency of mutant clones within an
exposed population of cells.
279
-------
In vivo clone formation has not really been demonstrated
with chemical carcinogens or mutagens. The formation has been
shown with ionizing radiation. Whether it occurs after exposure
to chemical carcinogens or not is unclear. Chemicals tend to be
less efficient than ionizing radiation in inducing chromosomal
abnormalities. On the other hand, SCE frequencies tend to be
much higher with chemical carcinogens than with physical mutagens.
There are differences and to some extent they are complementary.
DR. NELSON: Is antigenic stimulation convertible into a
provocative test that would be safe and useful?
DR. BLOOM: That is the purpose of looking at the cells JLn
vitro but it may not be a reasonable thing to do in vivo.
DR. DELLA PORTA: Is there any example of an exposure followed
by chromosomal aberrations and cancer?
DR. BLOOM: There are many types of compounds, benzopyrene
or benzene, for example, that can produce chromosomal damage in
cell systems and in vivo; these compounds are associated with
specific kinds of neoplasia. The Hiroshima-Nagasaki radiation
studies are perhaps the classic example of that.
DR. DELLA PORTA: Vigliani's group described a relationship
between chromosomal aberrations and leukemia in workers exposed
to benzene. But has the exposure of a given population to a specific
chemical been studied enough to define the cytogenetic abnormalities?
DR. MILLER: Two studies, of benzene and radiation, have been
performed but we don't know that the chromosomal aberration leads
to the leukemia.
280
-------
DR. DE CARLI: What is the possibility of using unscheduled
DNA synthesis to evaluate genetic injury in epidemiologic studies?
DR. BLOOM: It is probably a reasonable approach to try to
identify individuals at risk in terms of their having deficiencies
in DNA repair. This may not work in a population already exposed.
The technique really requires rather short-term followup of
cellular response to high-dose exposure. The problem is that
screening for unscheduled DNA synthesis is a very nonspecific kind
of test; it does not define the nature of the repair problem. An
abnormal result relates to a general phenomenon (deficient repair);
it reveals nothing about the mechanism. Probably many mechanisms
are involved, particularly in mammalian cells.
DR. DELLA PORTA: What do you think is the mechanism underlying
SCE's? Repair?
DR. BLOOM: The test uses bromodeoxyuridine to substitute uridine
for thymidine. When up to two cell divisions occur, thymidine is re-
placed by uridine in one chromatid of a chromosome in a fair proportion
of cells. The presence of uridine quenches the stain in the two
chromatids. The mechanism is not clear. Presumably we are simply
looking at a manifestation of somatic crossover. The results suggest
that there are break points in the nucleotide sequence. There are at
least two in most regional exchanges. Undoubtedly, specific target
sites exist on the nucleotide sequence where many of these chemicals
will act. Exactly where the breakpoints are has not been established.
The exchange that occurs was not detectable until these tech-
niques became available. Compounds that are effective with regard to
281
-------
specific locus mutation are probably the same. These are the
compounds that can induce increases in SCE frequencies. So SCE
works in much the same way as point mutation does, and, therefore,
is a very reasonable indicator of point mutagenicity.
DR. WEINBERG: Given that these test systems can be used
for exposed populations at chemical dumpsites, and given that there
is a large amount of variability from laboratory to laboratory,
do you see the need for more careful planning of studies of several
dumpsites? More specifically, should a central laboratory be
designated to perform all the testing so that the results can be
better compared?
DR. BLOOM: Yes. The Environmental Mutagen Society is consider-
ing the possibility of licensing laboratories for specific kinds of
studies with chemical mutagens. These laboratories, each with ex-
pertise in a small number of bioassay systems, could probably be
scattered around the world. Otherwise, many laboratories, particu-
larly commerical ones, try to perform too many of these tests. Many
firms will perform multiple kinds of assays, at times with minimal
and superficial knowledge, even within the genetic test system,
not to mention other kinds of test systems. They may be competent
in cytogenetics, but not necessarily in specific locus mutation
tests or in the Ames assay. Or they may do the Ames well, and know
nothing about cytogenetics.
This problem surfaced in Harrisburg, for example, when the
Three Mile Island accident occurred. Some of the tests obviously
282
-------
should have been performed right away. There was great discussion
about which laboratory to use. Was the closest laboratory the best
one? Was it even a reasonable laboratory for determining SCE's or
chromosomal aberrations?
DR. MILLER: As I understand what you have said, if you had
been called in August 1978 for the Love Canal incident, you would
have recommended (1) SCE studies, (2) other cytogenetic studies,
and (3) possible examination of urine in short-term assays. Is
there anything else?
DR. BLOOM: The systematic examination of certain individuals
such as those who were pregnant when exposed. A major somatic mani-
festation of many chemical carcinogens is teratogenicity. Thus,
consideration should be given to examination of pregnant women and
of their fetuses, perhaps using amniocentesis, and certainly to follow
up of the newborn infants.
Various observations should be made, but the general medical
approach really involves setting up cohorts and following the
individuals for morbidity and mortality to watch for disorders that
are likely to develop. The kinds of studies made in Seveso were
appropriate.
DR. MILLER: In an exposed area, if a person has surgery,
should a sample of skin fibroblasts be obtained for storage?
DR. BLOOM: Not really.
DR. MILLER: Suppose a woman miscarries; should a special
study be made of the fetus?
DR. BLOOM: Absolutely.
283
-------
DR. MILLER: Did Dr. Rehder suggest using the amnion if the
fetus is macerated?
DR. REHDER: The fetus is mostly autolytic. Placental tissue
will yield mostly maternal cells; this is a risk. You can exclude
this possibility by studying the amniotic membrane.
DR. BLOOM: Certainly it is possible to culture the fetal cells.
DR. REHDER: Yes. It is certainly better to study fetal cells,
if the fetus is not macerated.
DR. MILLER: What tissues or body fluids could be saved and by
what methods? For instance, if the fetus is put in formalin, cyto-
genetic studies are not possible. And how should urine samples be
handled? Should they be lyophilized or somehow reduced in volume?
How can lymphocytes be preserved for future cytogenetic study?
DR. BLOOM: It is difficult. If you can separate the lympho-
cytes and put them in medium, they can be studied for days after an
exposure. But after several weeks, nothing much can be done.
DR. MILLER: Could lymphocytes be processed to a point for
cytogenetic analysis and stopped short?
DR. BLOOM: Absolutely. For example, white cells can be
separated from blood. The buffy coat could be put into tissue
culture medium—that would be the ideal if it were available—
and left standing for days. That would rescue the lymphocytes.
Freeze them for later use.
284
-------
DR. HEATH: The sequence of events of what you would do in
the case of another Love Canal is logical, but you pointed out
that in the case of Three Mile Island, the choices are contingent
on dosage considerations.
DR. BLOOM: Yes.
DR. HEATH: At Three Mile Island there was a lot of talk
about cytogenetic studies, but the decision was made not to
proceed because the dosage was only a fraction of the background
levels. The exposure certainly was there, but the dosage to more
than a few people around the canal was, perhaps, not very large.
Would you still base that judgment on size of dose? I don't know
what the judgment should have been at Love Canal, or should be if a
similar situation were to occur. But size of dose is important to
the decision to perform an expensive battery of tests.
The question has to do with your reference to an important
aspect of preventive medicine—screening the workplace (or other)
population particularly after an exposure. You said 1-2% would
be predisposed to adverse effects. Where did that figure come from?
DR. BLOOM: Michael Swift at the University of North Carolina
studied ataxia telangiectasia (AT). He has begun to estimate the
frequency within the population of heterozygotes for AT and Fanconi's
anemia. His estimate in the AT study was that about 1% of the general
population may be heterozygous for the AT gene.
DR. HEATH: Would you add on the frequencies of other high-risk
heterozygotes?
285
-------
DR. BLOOM: Yes.
DR. HEATH: Has any such screening of the population in Japan
taken place, to follow at least that cohort in terms of increased
risk?
DR. BLOOM: No. It has not been established that these indi-
viduals can be detected by screening the general population.
DR. IREY: The chromosomal aberrations you are picking up
are in the cell line of the lymphocytes. What is going on in other
cell lines of the same organism?
DR. BLOOM: The presumption is that the lymphocytic response
is simply a manifestation of what is going on throughout the body.
On the other hand, there are undoubtedly tissue differences in
susceptibility. Because we can only study certain tissues, we
really do not know what these differences are. The lymphocyte-
chromosome response is very likely the same as it is in other
systems. When we see chromosomal damage in the lymphocytes and
thyroid cancer, it makes sense to say that they are related.
DR. GOLDBERG: Have studies been conducted in an industrial
setting of SCE's, somatic mutations, chromosomal aberrations, and
positive urine results in the Ames test to give an idea of the
influence of smoking on an industrial population?
DR. BLOOM: The SCE studies in the industrial setting are
just beginning. Carrano's is perhaps one of the first, and that
has only been described in abstract form. Studies of chromosome
286
-------
breakage in workers have been most heavily applied to the radiation
industry, in Dow's plutonium workers and a few selected groups of
that sort. Genetic screening in the workplace has not been widely
performed.
DR. GOLDBERG: What about the Ames test?
DR. BLOOM: The Ames assay has been applied to urine from
people working in a variety of industrial settings and from a
variety of people exposed to putative mutagens. Results of the
assay in general correlate well when one has a known mutagen,
or metabolite of a known mutagen, in the urine. But there have
been only a few such studies of workers.
DR. MILLER: In the past, policymakers in oncology said, "Why
study genetics? You can't fix genetics." It now seems to me that
genetics is being used to study the environment.
DR. BLOOM: That's right. Basically we are talking about the
response of an organism to an environmental agent. We cannot alter
the genetic response, but when genetic measures are known to be
very sensitive, they can be used to help identify deleterious
agents. They are of value in protecting the population. We are
not trying to alter the genotype; we are trying to use the responses
of the chromosomes, specific loci, and so on, as measures of environ-
mental risk.
If somatic cell damage leads to clinical disease such as cancer
or birth defects, it is assumed then that the mutational effects are
likely to be reproduced in the germ cells because the meiotic chromo-
somal material there is essentially the same as it is in somatic
287
-------
cells. The end points are difficult to define; ways of estimating
vertically transmitted damage are not always available.
Somatic cell effects are reasonable measures to use to identify
those compounds in the environment that are capable of producing
both somatic and potentially genie effects, which will be reflected
in later generations.
My concern is primarily with somatic effects. Unless a really
massive exposure to a potent mutagen occurs, it is unlikely that
environmentally induced, inherited abnormalities can be detected.
If we could not detect them after the atomic bomb exposures in
Japan, it is very unlikely that we will identify them anywhere else.
DR. MILLER: What about the people exposed to Kepone or DBCP?
Should their children be studied by biochemical genetics?
DR. BLOOM: I think not. Unless the numbers are sufficiently
large, it would not be worthwhile.
DR. TARDIFF: With respect to several classes of chemical mutagens
and chemical carcinogens, advances have been made in the last several
years in measuring the formation of adducts to nuclear material.
Presumably an initiating step leads to an improper replication leading
to other events. Please comment on the possibility of using that
kind of analysis as an adjunct to measuring mutagenic activity of
SCE.
DR. BLOOM: Such studies are really in an early stage.
Undoubtedly some mutagens are going to work on cell membranes;
others are going to work directly on the DMA. Some mutagenic
288
-------
compounds are going to affect cell replication. We are calling all
of these substances mutagens, even though they may not all be muta-
gens. Defining the end point is very different from defining the
mechanism by which a mutagen works.
When we talk about chromosomal aberrations or SCE, we are
not really discussing the mechanism by which an event takes place.
For instance, the range of mechanisms is enormous for nitrosamines
versus alkylating agents. To generalize is not necessarily reasonable.
DR. TARDIFF: For those agents known to produce electrophiles
and for which there is a substantial likelihood that there would be
adduct formation with genetic material, wouldn't it be reasonable to
study adduct formation under those circumstances? Of course, while
adducts do not necessarily lead to an immediate expression of a
genetic alteration, genetic alterations could occur at a substantially
later time.
Can this mechanism be used to identify people whose risk may
not be immediately observable, but who may themselves be at risk in
the future, or, perhaps, whose progeny may be at risk from parental
exposure?
DR. BLOOM: How would you identify them?
DR. TARDIFF: If one could identify people with adduct formations
(as a result of exposures to certain materials that are metabolized
to electrophiles), one might then be able to follow those people as
unique study populations to find out if, in fact, their risk materializes
and to what extent. Then perhaps one could use the information to
predict effects for other, similarly exposed populations.
289
-------
DR. BLOOM: I am not aware of any applications of that tech-
nology in large-scale human studies.
DR. TARDIFF: Might it be feasible to apply the technique
to populations that have known exposures to agents that have this
kind of mechanism?
DR. BLOOM: Yes, it's possible.
DR. MILLER: Dr. Rehder and Dr. Gropp have studied embryos
from Seveso. What are your findings?
DR. REHDER: We studied 32 fetuses from pregnancies interrupted
at the mothers' requests, because the women said they had been exposed.
Analysis of where these women lived revealed that none of them had
been heavily exposed. Those who were exposed were exposed for a very
short time. Most of the women said they had been exposed while riding
through the area in cars.
Abortions were performed by curettage, rather early in gestation,
and most of the material was lost. Our publication lists the study
samples, but mostly we had only one leg and a bit of placenta. We
did not have whole fetues, so we could not say whether or not they
were deformed. In the few cases where we had complete fetuses, we
found no malformations except for lesions caused by the abortions.
We also studied specimens from three spontaneous abortions; these
did have anomalies. In one, there was a total infarction of the
placenta, which would explain the abortion; another had atypical
facies, a heart malformation, and some minor internal stigmata.
This fetus probably had chromosomal aberrations and also Down's
syndrome.
290
-------
The third case had a molar degeneration of the placenta, which Is
not a true mole. It Is a form of degeneration found frequently in
chromosomal aberrations, with a closed amniotic sac, but without an
embryo—like a blighted ovum. Chromosomal aberrations have been
found in samples such as this, for example, by the Boves in Paris.
Of course, some alterations would be expected in the case of these
spontaneous abortions—something to explain the abortion.
We did not see anything abnormal in the fetuses from the
induced abortions. Our conclusion, thus, was that we did not find
anything beyond the normal incidence of malformations.
One year later, we analyzed another 50 cases of spontaneous
abortions and neonatal deaths located by Dr. Sancioni, and we
analyzed all of the malformations she had found. Of these, 24%
had major and minor malformations and cases of trlsomy 13 and 18.
Again, we compared our findings to those of another sample from
outside the area and to published data on malformations,
aberrations, and minor anomalies among newborns. The results from
the 50 cases were below the ranges established by the comparison
groups. Although the 24% rate had seemed high at the beginning of
the study, it really was not above the norm. The malformations and
anomalies were very heterogeneous; there was no specificity at all.
291
-------
Cytogenetic Investigations of the Seveso Population Exposed to TCDD
L. De Carli, A. Mottura, F. Nuzzo, G. Zei, M.L. Tenchini, M. Fraccaro,
B. Nicoletti, G. Simoni, and P- Mocarelli1
The scoring of chromosomal aberrations in fibroblasts and
lymphocytes from subjects exposed to toxic compounds is of limited
value as an indicator of somatic genetic damage. Nonetheless, it is
the only method available to evaluate immediate effects on genetic
material. After the Seveso accident, the Institutes of General
Biology and Genetics of the universities of Milan, Pavia, and Rome
began a cooperative program of cytogenetic investigations of indi-
viduals from the TCDD-contaminated area. Chromosomal analysis was
performed on 28 cases of induced abortions for a comparison with
the frequencies of chromosomal aberrations in cultured cells from
maternal and fetal tissues. A sample of 301 individuals from the
Seveso population was also investigated. This study included 145
residents of Zone A near the ICMESA (Industrie Chimiche Meda Societa
Anoniraa) farm; 69 nonresident ICMESA workers; and 87 control subjects.
The differences in the frequency of aberrant cells between
exposed and control subjects were not generally significantly greater
than those among individuals and among observations from different
laboratories. However, in some cases, the frequency of mitotic
samples with at least one aberrant cell was higher in the exposed
groups than in controls. The results so far do not suggest a
straightforward conclusion, but indicate the need to continue
the cytogenetic monitoring of the Seveso population.
The detection of chromosomal aberrations in lymphocyte or
fibroblast cultures from exposed individuals is virtually the only
immediate and practical tool for evaluating genetic damage induced
by a toxic substance accidentally released into the environment
(Evans and O'Riordan, 1975). Genie mutations cannot be detected so
easily. The possibility of recognizing them in cells from tissues
of exposed individuals is confined to a few loci in the hemizygous
state (Albertini and De Mars, 1973; Strauss and Albertini, 1977).
Since lymphocytes and fibroblasts are almost ubiquitous cells,
Cytogenetics Collaborative Group, Regione Lombardia, Italy
Project coordinator: G. Morganti
292
-------
they may be regarded as targets that are representative of the organism,
although the sensitivity of different tissues can vary greatly.
The significance of chromosomal lesions in cells cultured In vitro
for the pathology of an individual and progeny varies. An abnormally
high frequency of chromosome breaks, when the capacity to repair DNA
damage is normal, indicates that a chemical or physical agent has
directly or indirectly interacted with the genetic material to bring
about a change in its structure. However, different types of chromo-
somal aberrations have different genetic consequences, either at the
somatic or germinal level. Completely or partially unstable aberrations
(chromatid and chromosome breaks, rings, and dicentrics) are relevant
mainly for cell survival and for the control of cell proliferation insofar
as they can determine cellular lethality or can represent a critical
step in the succession of events ending in neoplastic transformation
(Bloom, 1972). Stable chromosome aberrations (translocations and
inversions) may represent transmissible genetic damage, and therefore
can indicate a risk to progeny. However, a sizable fraction of
chromosome structural variation detectable in vitro may not have
any pathologic significance because it does not reflect the condi-
tion in_ vivo. A variety of factors inherent to cell culture,
including pH, quality of medium, and contamination by mycoplasma,
induce chromosomal alterations (Keck and Emerit, 1979; Schneider
et al., 1974). The presence of either chromatid or isochromatid
and chromosomal aberrations is generally considered a criterion
for discriminating preexisting chromosome damage from that produced
in vitro, but it is not absolutely valid. Due to the interference
293
-------
of technical factors, the cell cultures from exposed persons must
always be established in tandem with those of controls. Moreover,
data on the incidence of chromosome lesions in cell samples cultured
in different laboratories or at different times are hardly comparable.
Finally, scoring for chromosomal aberrations in vitro is
informative only if a positive response, providing evidence of a
deleterious change in the genetic material, is observed. A negative
response is not sufficient to exclude the occurrence of such a change,
even at a chromosomal level (Schinzel and Schmid, 1976).
Despite all these limitations, cytogenetic analysis of exposed
subjects appears to be unavoidable, at least for those cases in which
the mutagenicity of a chemical contaminant can be predicted on the
basis of results of laboratory tests or from inferences drawn from
the molecular structure of the compound. Such is the case with tetra-
chlorodibenzodioxin (TCDD), a weak mutagen both in prokaryotes and
lower eukaryotes (Wasson _et_ a\_. , 1978).
After the Seveso accident of July 10, 1976, the first chromosome
analyses were performed at the request of hospitals. Analyses were
performed on blood from eight children aged 2 to 10 years and from
four pregnant women living in the contaminated area and admitted to
the hospitals because of dermatologic lesions, presumably caused by
TCDD exposure.
A second sample group included 28 women who chose abortion
because of a presumed risk of teratogenesis. Of these, 21 were in
the first trimester of pregnancy at the time of the accident; 7 had
become pregnant immediately afterward. Preliminary data on these
294
-------
cases have already been reported (Tenchini et^ al., 1979). At that
time, Italian law did not permit interruption of pregnancy. Therefore,
a parallel sample could not be obtained for nonexposed, interrupted
pregnancies. Such samples were studied later, after abortion was
legalized in Italy for clinically normal cases, which could serve as
suitable controls. Thus, samples from 16 pregnancies became available
for cytogenetic study some 2 years later.
In a collaborative program sponsored by the administration of
the Regione Lombardia, five Italian laboratories have undertaken a
more extensive cytogenetic analysis of the Seveso population. During
a 5-month period starting in October 1976, investigators sampled 330
persons, who were distributed into three exposure categories: acute—162
persons who lived near the ICMESA plant in Zone A; chronic—73 persons
who worked at the ICMESA plant but did not live in Zone A; and
controls—95 persons who lived in the area surrounding the contaminated
zone and matched exposed persons by age and sex.
The types of chromosomal aberrations studied were (1) chromatid
and isochromatid gaps and chromatid and chromosomal breaks and (2)
chromatid and chromosomal rearrangements (chromatid interchanges,
rings, dicentrics, and morphologically atypical chromosomes).
Examples of these aberrations are illustrated in Figure 1. To
evaluate the total amount of chromosome damage, the overall frequencies
of chromosome lesions were established, including and excluding gaps,
because the latter type of alterations is a poor indicator of cyto-
genetic risk. In the analysis of the large sample from the Seveso
population, care was taken to ensure that the scorer did not know
the classifications of the subjects under examination.
295
-------
CHROMATID
ISOCHROMflTID OR CHROMOSOME
GflPS
BREflKS
RERRRRN6EMENTS
"VJ
\
\B '
i**™-*
1
FIGURE 1. Types of chromosome aberrations investigated at Seveso.
296
-------
SAMPLE OF HOSPITAL PATIENTS
The frequencies of chromosomal aberrations in lymphocytes from
72-hour cultures of peripheral blood are shown in Table 1. Standard
laboratory values, calculated on 10 adults examined during a comparable
period are also provided for reference. The difference in the frequencies
of chromosomal aberrations between the patients and the unmatched
controls is not significant, except for the category of "aberrations
including gaps" in pregnant women, which is at the limit of significance.
SAMPLE OF INTERRUPTED PREGNANCIES
For each pregnancy, fragment cultures were made from fetal tissues,
placenta, and umbilical cord. Lymphocyte cultures were established
from maternal blood drawn immediately before the abortion. Not all
types of tissue were available for all cases. For the controls,
particular care was taken to ensure that culture conditions, cytologic
procedures, and scoring criteria were as identical as possible to
those used for analyses of samples from exposed pregnancies.
The frequencies of chromosomal aberrations observed in cells from
maternal and fetal tissues in exposed and nonexposed pregnancies are
reported in Table 2. Data are recorded only for those specimens that
showed a high growth rate within the first month of culture. The
frequencies of chromosomal aberrations are consistently different
between tissues within categories. Blood cultures have the lowest
values and the placenta and umbilical cord cultures have values about
twice as high. In the controls, fetal cells have aberrational
frequencies comparable to those of the other solid tissues. In the
exposed pregnancies, fetal cells have values almost twice as high.
297
-------
TABLE 1
Frequencies of Chromosomal Aberrations in Blood Cultures of Hospital Patients
No. of No. of % Aberrant Cells
Subject
Children
Pregnant women
Controls
Individuals Mitoses Including Gaps
8 439 5.92
ta= 1.4057 n.s.b
4 185 7.57
t = 2.4367
P = < 0.05
10 553 3.98
Excluding Gaps
1.59
t = 0.4727 n.s,
1.62
t = 0.4353 n.s,
1.99
a t = test of significance for comparison between two percentages
(transformed values).
b n.s. = not significant (P > 0.05).
298
-------
TABLE 2
Frequencies of Chromosomal Aberrations in Maternal and Fetal Tissues;
Type of
Tissue
Maternal
blood
Sample
of
Interrupted
No. of No. of
Pregnancies
% aberrant
Categories Pregnancies Mitoses Including
Control
Exposed
11
13
1172
1403
4.10 + 1.
3.35 + 0.
cells
Gaps
13
94
+s
.d,
a
i
Excluding
2.
1.
22
78
+ 0.
+ 0.
Gaps
84
69
t = 0.93 n.s. t = 0.73 n.s.
Placenta Control 11 1547 7.50+1.31 4.91+1.08
and
umbilical Exposed 12 2001 8.40+1.22 4.50+0.91
cord
t = 0.99 n.s. t = 0.56 n.s.
Fetal Control 12 1600 7.56+1.30 4.44+1.01
tissue
Exposed 12 2050 15.17+1.55 9.12+1.25
t = 7.24 t = 5.68
P < 0.001 P < 0.001
as.d. = standard deviation
299
-------
The differences among categories within tissues are not statistically
significant for maternal blood, placenta, and umbilical cord. They
are highly significant for fetal tissue.
SAMPLE OF SEVESO POPULATION AND ICMESA WORKERS
Four blood cultures were established for each individual, and
preparations were made after 48 and 72 hours of incubation. To ensure
homogeneity of cell culture conditions and cytologic procedures, this
part of the work was concentrated in a single laboratory, the Institute
of General Biology at the University of Milan. The coded slides were
randomly distributed for a preliminary examination among the five
laboratories participating in the program. The frequencies of the
various types of aberrations have been determined on all the scorable
preparations obtained from the 72-hour cultures. Although these
preparations were less informative for determining preexistent
chromosomal damage in vivo, they did permit analysis of a larger number
of individuals than did the 48-hour preparations, which did not always
show a sufficient yield of mitoses.
The number of cells analyzed per person in the first scoring
ranged from 30 to 80. The cytogenetic analysis was performed on
blood cultures and preparations were made at 48 and 72 hours. Com-
plete data are available for 72-hour cultures. The pooled data from
the different laboratories are shown in Table 3. Variance analysis
was performed on the individual values, according to the criteria of
exposure categories and scorers in different laboratories. The
results are shown in Table 4. The differences in the frequencies of
chromosomal aberrations, including gaps between categories, are at
the limit of significance when compared with the differences among
300
-------
TABLE 3
Frequency of Chromosomal Aberrations in Lymphocytes; Preliminary
Cytogenic Analysis of Samples from the Seveso Population
No. of No. of % Aberrant Cells (Variance Interval)
Category Individuals Mitoses Including Gaps Excluding Gaps
Chronic
exposure 69 3040 2.53 0.92
(0.00 - 12.00) (0.00 - 12.00)
Acute
exposure 145 6470 2.49 0.99
(0.00 - 18.00) (0.00 - 8.00)
Controls 87 3958 1.64 0.48
(0.00 - 8.90) (0.00 - 4.00)
301
-------
TABLE 4
Variance Analysis on Transformed Values (arc sinj"p) of Frequencies
of Chromosomal Aberrations in Lymphocytes; Preliminary Cytogenetic
Analysis; Samples from the Seveso population
Sources of Variation S.S.;
Including gaps
Total
Exposures
Laboratories
within exposure
Individuals
within laboratory,
within exposure
Excluding gaps
Total
Exposures
Laboratories
within exposure
Individuals
within laboratory,
within exposure
d.f. Variance Fc
Statistical
Significance
9880.3552 247
252.7708 2 126.3854 3.4499 < 0.05
981.8648 9 109.0961 2.9780 < 0.005
8645.7196 236 36.6344
4935.8563 247
102.2117 2 51.1059 2.5987 < 0.05
192.4735 9
4641.1711 236
21.3859 1.0875 < 0.05
19.6660
a S.S. = sum of squares.
d.f. = degrees of freedom.
c F = variance ratio.
302
-------
individuals within categories. However, the differences among scorers
are highly significant. When gaps are excluded, these differences are
no longer significant.
The results of the statistical analysis of these preliminary
data prompted the extension of the analysis to a larger number of
mitoses on selected samples from the three exposure classes. The
samples had to be comparable in size and characterized by a high
mitotic yield on both the 48- and 72-hour cultures. Each group
contained samples from 45 persons and was subdivided into subgroups
of 15 each. The subgroup samples were distributed among three
laboratories according to the scheme shown in Table 5. Thus,
preparations for each person were scored by two different laboratories.
The frequencies of chromosomal aberrations, both including and
excluding gaps, and the number of aberrations per cell for the control,
for both the acute and the chronic exposure categories, are reported in
Tables 6, 7, and 8. Data from 48- and 72-hour cultures are recorded
for each subgroup and for each of the two scorers. The variability
in observations among the different laboratories is notable. Table
9 shows the average percentage of aberrant cells for the sets of 30
individuals scored in each category by each laboratory. Whereas two
of the laboratories show consistently higher scorers for both chronic
and acute exposure categories (as compared to the controls), the third
shows almost equal values for the acute exposure category and the
controls, but lower values for the chronic category.
303
-------
TABLE 5
Experimental Design; Balanced Incomplete Blocks
Exposures
Laboratory3 Acute Chronic Controls
PVG B! B2 B4 B5 By Bg
RD 121 "DD TJT3 H
oi o o "A D£ By JJQ
T>TTlJ TJ T) T> p TJ -n
TV Oo DO DC D/- Do JJQ
/ j jo o y
aPVG = Istituto di Genetica, Universit
-------
TABLE 6
Frequencies of Chromosomal Aberrations for the Control Category,
Ul
Analyzed in a Block Design;
No. of
Blocks Individuals Laboratories
48 hours
B? 15 RB
PVG
Bg 15 PVG
PVB
Bg 15 PVB
RB
72 hours
B? 15 RB
PVG
Bg 15 PVG
PVB
B9 15 PVB
RB
No. Scored
per Block
15
15
15
15
9
10
11
14
15
15
10
6
No. of
Mitoses
962
1144
1482
1197
483
659
1027
1276
1536
1475
783
558
Sample of Seveso Population
% Aberrant
Including
Gaps
1.247
0.786
0.877
0.334
2.070
1.669
1.168
0.628
0.520
1.423
1.149
0.716
Cells
Excluding
Gaps
0.415
0.174
0.269
0.167
1.035
1.062
0.681
0.078
0.000
0.813
0.255
0.179
No. Aberrations/Cell
Including
Gaps
1.083
1.000
1.000
1.000
1.200
1.090
1.250
1.000
1.000
1.142
1.000
1.000
Excluding
Gaps
1.000
1.000
1.000
1.000
1.200
1.000
1.000
1.000
0.000
1.000
1.000
1.000
-------
TABLE 7
Frequencies of Chromosomal Aberrations for the Acute Exposure Category
UJ
o
Analyzed in a Block Design; Sample of Seveso
No. of
Blocks Individuals Laboratory
48 hours
Bj 15 PVG
RB
B2 15 PVG
PVB
B3 15 RB
PVB
72 hours
Bx 15 PVG
RB
B2 15 PVG
PVB
B3 15 RB
PVB
No. Scored
per Block
12
14
15
15
15
15
12
14
15
15
14
15
No. of
Mitoses
1173
1250
1540
1059
1415
1400
1193
1176
1510
1650
1254
1490
% Aberrant
Including
Gaps
0.767
1.200
1.363
3.871
1.625
3.000
1.424
1.275
2.317
3.575
0.956
2.281
Population
Cells
Excluding
Gaps
0.255
0.320
0.324
1.605
0.565
0.785
0.335
0.595
0.993
1.939
0.558
0.939
No. Aberrations/Cell
Including
Gaps
1.000
1.133
1.000
1.170
1.000
1.166
1.058
1.200
1.085
1.050
1.250
1.147
Excluding
Gaps
1.000
1.000
1.000
1.235
1.000
1.272
1.000
1.142
1.000
1.062
1.000
1.071
-------
TABLE 8
Frequencies of Chromosomal Aberrations for the Chronic Exposure Category, Analyzed in a Block Design:
Sample of Seveso Population
No. of
Blocks Individuals
48 hours
B4 15
B5 15
B6 15
72 hours
B4 15
B5 15
B6 15
Laboratories
PVG
RB
PVG
PVB
RB
PVB
PVG
RB
PVG
PVB
RB
PVB
No. Scored
per Block
10
14
11
11
13
13
9
14
9
13
13
13
No. of
Mitoses
868
1385
592
489
1250
1142
801
1354
844
1083
1300
1170
% Aberrant
Including
Gaps
1.036
0.433
3.209
2.658
1.120
1.488
1.872
0.516
1.895
2.123
0.307
1.965
Cells
Excluding
Gaps
0.345
0.072
1.520
0.817
0.480
0.437
0.624
0.221
0.947
1.015
0.076
0.598
No. Aberrations/Cell
Including
Gaps
1.000
1.000
1.000
1.000
1.000
1.176
1.000
1.000
1.125
1.130
1.250
1.217
Excluding
Gaps
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.000
1.125
1.000
2.000
1.428
-------
TABLE 9
Frequencies of Chromosomal Aberrations for Each Exposure Category
and Each Laboratory
Acute Exposure, Chronic Exposure, Controls,
% Aberrant Cells (No. of Mitoses) % Aberrant Cells (No. of Mitoses) % Aberrant Cells (No. of Mitoses)
Laboratories Including Gaps Excluding Gaps Including Gaps Excluding Gaps Including Gaps Excluding gaps
48 h
PVG 1.105 0.294 1.917 0.821 0.837 0.228
o, (2713) (1460) (2626)
o
RB 1.425 0.450 0.759 0.265 1.418 0.678
(2665) (2635) (1621)
PVB 3.375 1.138 1.839 0.551 0.833 0.416
(2459) (1631) (1680)
72 h
PVG 1.923 0.702 1.884 0.790 0.568 0.035
(2703) (1645) (2812)
RB 1.111 0.576 0.414 0.150 1.009 0.504
(2430) (2654) (1585)
PVB 2.961 1.464 2.041 0.798 1.328 0.620
(3140) (2253) (2258)
-------
For the variance analysis, individual values were transformed
by the sin ~* /^transformation. Only mitotic samples for which
exact 100 counts were available were included in the analysis. The
analysis was based on the following model.
Yijq = ^*i + %J + -q + *tjq
where are represented the effects of
p = mean
IT = exposure
g = incomplete block
T = laboratory
e = intrablock residual or error
Analysis of Variance
Sources of Variation: d.f.
Exposures: (c-1)
Laboratories (unadjusted): (t-1)
Block within exposures (adjusted): (b-c)
Intrablock error: (tr-t-b+1)
where t = 3 = no. of laboratories, r = 6 = no. of replications, and
b = 9 = no. of blocks. This is repeated for c = 3 = no. of exposures.
309
-------
The results are shown in Tables 10 and 11. No significant
differences were observed between exposure categories for any of
the four combinations (48 and 72 hours, with aberrations inclusive
or exclusive of gaps). Significant differences were observed be-
tween laboratories in the frequencies of aberrations at 72 hours
of culture. The differences at 48 hours are insignificant.
An alternative criterion for analyzing differences in the in-
duced chromosomal damage among exposure categories would be to use
the proportion of individuals with at least one aberrant cell in
either one or the other of the two scoring replicates. Such an
analysis seemed particularly relevant because of the pronounced
variability among subjects in the frequencies of chromosomal aber-
rations and also because the exposure conditions were most likely
not homogenous. The values calculated in this way are shown in
Table 12. A significant chi-square value was recorded for aber-
rations including gaps at 48 hours of culture because of the lower
proportion of individuals with aberrations in the chronic exposure
category.
Conclusions
From the analysis of the cytogenetic findings in the samples
from TCDD-exposed persons, the following conclusions can be drawn.
Owing to the limited size of the hospital patient sample and to the
lack of controls from appropriately matched individuals, the results
of the cytogenetic analysis are not very informative. Nevertheless,
the average frequencies of chromosomal aberrations for the two groups
310
-------
TABLE 10
F-Ratios in Analysis of Variance for the Comparison Between
Exposures, Following the Balanced Incomplete Block Design
Aberrations
Including gaps
Culture
Period, hours
48
72
r\
s Between Exposures/ Statistical
s2 Error = F (2,13 d.f.)a Significance
4.3742
5^036 - °'8095
4.2779
1T9854 - 2'1547
n.s.
n.s.
Excluding gaps
48
72
2.4329
5^974 - °'4593
5.7274
2.5826
= 2.2176
n.s.
n.s.
F (2, 13 d.f.) = variance ratio for 2 and 13 degrees of freedom.
311
-------
TABLE 11
F-Ratios in Analysis of Variance for the Comparison Between
Laboratories, Following the Balanced Incomplete Block Design
Aberrations
Including gaps
Excluding gaps
Culture
Period, hours
48
72
48
72
s^ Between Laboratories/ Statistical
s2 Error = F (2,7 d.f.)a Significance
7.7967
2.3325 ~ 3-3426
17.1429
1.1107
15.4343
1.2686
371146 = °'3946
9.4400
1.7693
= 5.3355
n.s,
P<0.01
n.s.
P<0.05
F (2, 7 d.f.) = variance ratio for 2 and 7 degrees of freedom
312
-------
TABLE 12
Overall Frequencies of Individuals with Chromosomal Aberrations
% of Individuals with
Exposure
Categories
48 hours
Acute
Chronic
Controls
No. of
Individuals
29
19
13
Chromosomal
Including Gaps
96.55
73.68
92.31
X2 = 6.1418
Aberrations
Excluding Gaps
62.07
42.11
53.85
X2 = 1.8428
72 hours
Acute
Chronic
Controls
31
23
27
P <0.05
93.55
86.96
85.19
X2 =
1.1434
n.s.
80.65
56.52
55.56
5.1437
n.s.
313
-------
of patients are not significantly different from the standard labora-
tory values for the relevant period.
The data regarding the differences between blood and other tissues
obtained from control pregnancies are consistent with those obtained
from exposed pregnancies 2 years earlier. Placental and umbilical cord
tissues showed frequencies of chromosomal aberrations two times higher
than those of maternal blood, a finding that correlated well with the
ratio usually observed between values in cells from solid tissues and
in lymphocytes (Simoni et al., 1979). A significant deviation from this
trend was found only for fetal cells from exposed pregnancies. These
showed a fourfold increase in chromosomal aberrations with respect to
maternal blood. It is difficult to ascribe such a difference to factors
inherent in the cell culture even if established at different periods,
because the values are comparable in the two samples for all other
tissues. Therefore, the possible effect of TCDD exposure on the mothers
has to be considered. Under this assumption, a transplacental action
of TCDD and a higher sensitivity of the fetal cells to the chemical's
clastogenic effect can be postulated.
Preliminary analysis of the Seveso population data revealed a
tendency toward an increase in the frequency of chromosomal aberrations
in both acutely and chronically exposed persons. However, the differ-
ences (when compared with results from controls) were not statistically
significant. The trend was only partially confirmed by the later
analysis. As shown by the results of the variance analysis, there
was a considerable subjective component in the scoring of the various
314
-------
types of aberrations; observer differences were often significant.
However, the differences in the frequencies of chromosomal aberrations
(both including and excluding gaps) between exposure categories did
not become significant even after correction for observer variance.
This conclusion applies to results obtained both from 48- and
72-hour cultures. If the incidence of chromosomal aberrations in
the exposed groups is significant after increasing the number of
mitoses analyzed, the cytogenetic risk following the Seveso accident
will not be higher for persons living in the contaminated area than
for the ICMESA workers, who were not exposed to the toxic cloud.
Czeizel and Kiraly (1976) conducted a similar study. In comparison
to controls, there was a higher frequency of chromosomal aberrations
in blood of workers employed at a herbicide-producing factory in
Budapest, regardless of whether or not they had been directly involved
in the chemical production.
Indication of the lack of an obvious effect of acute exposure
to TCDD on the incidence of chromosome lesions in blood is provided
by analyzing the overall frequencies of persons with aberrant cells
in the various categories.
In conclusion, the combined cytogenetic findings in the three
samples examined—hospital patients, interrupted pregnancies, and
individuals from Seveso and ICMESA populations— may justify con-
tinuing cytogenetic monitoring of selected samples of the Seveso
population, even though they have not provided consistent evidence
of chromosomal effects associated with TCDD exposure.
315
-------
REFERENCES
Albertini, R. J., and R. De Mars. 1973. Detection and quantification
of X-ray induced mutation in cultured diploid human cells. Mutat.
Res. 18:199-224.
Bloom, A. D. 1972. Induced chromosomal aberrations in man. Pp. 99-
112 in H. Harris and K. Hirschhorn, eds. Vol. 3, Advances in
Human Genetics. Plenum Press, New York.
Czeizel, E., and J. Kiraly. 1976. Chromosome examinations in workers
producing Klorinol and Buvinol. Pp. 239-256 in L. Bank, ed.
The Development of a Pesticide as a Complex Scientific Task.
Medicina, Budapest, Hungary.
Evans, H. J., and M. L. O'Riordan. 1975. Human peripheral blood
lymphocytes for the analysis of chromosome aberrations in muta-
gen test. Mutat. Res. 31:135-148.
Keck, M,, and I. Emerit. 1979. The influence of culture medium
composition on the incidence of chromosomal breakage. Hum.
Genet. 50:277-283.
Schinzel, A., and W. Schmid. 1976. Lymphocyte chromosome studies
in humans exposed to chemical mutagens. The validity of the
method in 67 patients under cytostatic therapy. Mutat. Res.
40:139-166.
Schneider, E. L., E. J. Standbridge, C. J. Epstein, M. Golbus,
G. Abbo, and G. Rogers. 1974. Mycoplasma contamination of
cultured amniotic fluid cells: Potential hazard to prenatal
chromosomal diagnosis. Science 184:477-480.
Simoni, G., L. Larizza, N. Sacchi, G. Delia Valle, F. Dambrosio,
and L. De Carli. 1979. Chromosome lesions in amniotic fluid
cell cultures. Hum. Genet. 49:327-332.
Strauss, G. H., and R. J. Albertini. 1977. 6-Thioguanin resistant
lymphocytes in human peripheral blood. Pp. 327-334 in D. Scott,
B. A. Bridges, and F. H. Sobels, eds. Progress in Genetic Toxi-
cology. Elsevier North-Holland Biomedical Press.
Tenchini, M. L., R. Giorgi, C. Crimaudo, G. Simoni, F. Nuzzo, and
L. De Carli. 1979. Approaches to examination of genetic damage
after a major hazard in chemical industry: Preliminary citogenetic
findings on TCDD exposed subjects after Seveso accident. Pp. 301-
317 in K. Berg, ed. Genetic Damage in Man Caused by Environmental
Agents. Academic Press, New York.
316
-------
Wasson, J. S., J. E. Huff, and N. Loprieno. 1978. A review of the
genetic toxicology of chlorinated dibenzo-p-dioxins. Mutat. Res.
47 (2/4):141-160.
Zavola, C. , P. Arroyo, R. Lisker, A. Carnevale, F. Salamanca, J. I.
Navarrete, F. M. Jimenez, B. Blanco, V. Vasquez, J. Sanchez, and
S. Canun. 1979. Variability between and within laboratories
in the analysis of structural chromosomal abnormalities. Clin.
Genet. 15:377-381.
317
-------
DISCUSSION
DR. MILLER: Is there a plan to study sister-chromatid exchange
(SCE) in the chronically exposed people?
DR. DE CARLI: Yes. We are studying SCE's in individuals from
the three categories, including the ICMESA workers who showed exceed-
ingly high frequencies of chromosomal aberrations in the first analysis.
We selected some 30 subjects from the three categories and matched
controls. We examined SCE's and unscheduled DNA synthesis, and again,
chromosomal aberrations. The results are being analyzed now.
DR. DARDANONI: The Epidemiological Commission of Seveso con-
cluded that the study of frequency in a population of mixed exposure
indicators may underestimate the real presence of damage because
(1) what might be considered a homogeneous group of 163 people may
be a mixture with different exposure conditions; (2) high intralab-
oratory variation may obscure the existence of aberration; and
(3) inter-group variation may exist because not everyone had the
same exposure. Could the data be reexamined along two lines? One,
to confirm in the same individuals the stability of the gaps; and
two, to determine if the data can be petitioned according to expo-
sure indicates?
DR. DE CARLI: Yes, but we also are monitoring with different
parameters the individuals from the three categories that showed
the highest values. We can, though, both reexamine the old samples
and examine the selected sample in successive analyses.
318
-------
DR. MOORE: In Zone A, 163 people were heavily exposed and some
developed frank chloracne. Have they been separated in the study of
chromosome gaps or other aberrations from those who still might be in
an exposed area but failed to show chloracne or enlarged liver?
DR. DE CARLI: Yes, we selected a sample of subjects showing
chloracne, but they were not from Zone A; they were from Zones B and R.
The differences were not significant in this sample when compared to
nonchloracne cases.
DR. MILLER: Can you predict the likelihood of chromosomal effects
on the basis of the chemical structure or biologic tests that have been
made previously with TCDD? Would you expect TCDD to damage chromosomes?
DR. DE CARLI: The available data from chromosome studies are
inconsistent. Results differ, even within the same group of investi-
gators. Negative results and occasionally positive results have been
reported using bone marrow of rats continuously treated with TCDD.
DR. MALTONI: Do you explain the difference between laboratories
as resulting from the different samples or from the criteria used to
characterize the lesions?
DR. DE CARLI: The differences can be ascribed both to differ-
ences in criteria for classifying aberrations and to sampling effects.
DR. MALTONI: Did you have any type of collegial discussion to
clear up the difference?
DR. DE CARLI: Yes, with the conclusion that probably there is
a sampling fluctuation. There is no doubt, however, that one of the
laboratories tends to give higher estimates than the others. But
this effect is systematic, and we can correct for it.
319
-------
Cytochrome P-450 Induction By 2,3,7,8-Tetrachlorodibenzo-p-dioxin,
Polychlorinated Biphenyls, and Polybrominated Biphenyls
Robert Neal1
The major enzyme systems involved In metabolizing xenobiotic
compounds are the monooxygenase systems containing cytochrome P-450.
Recent studies have indicated the presence of numerous species of
cytochrome P-450 in the organs of a single animal; for example,
the liver of the Sprague-Dawley rat contains at least six. These
various species of cytochrome P-450 are inducible, to varying degrees,
by a large number of hydrophobic compounds, which can be divided
into two classes. One class contains the compounds that induce
various species of cytochrome P-450 in a manner similar to pheno-
barbital; the second contains the compounds that induce various
species of cytochrome P-450 in a manner similar to 3-methylcholanthrene.
The chemical 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) is also
a potent inducer of species of cytochrome P-450 that are similar to
those induced by 3-methylcholanthrene.
Various polychlorinated biphenyl (PCB) and polybrominated bi-
phenyl (PBB) isomers are metabolized by the cytochrome-P-450-containing
monooxygenase systems. The biologic half-life of various isomers of
PCB's and PBB's appears to be related to the ability of the cytochrome
P-450 monooxygenase system to metabolize them. The results of vari-
ous studies indicate that PCB and PBB isomers with two or more adja-
cent, unsubstituted carbons are readily metabolized by the cytochrome
P-450 monooxygenase systems.
Recent work in various laboratories has indicated that TCDD is
metabolized in the rodent to more polar metabolites as shown by radio-
activity excreted in urine and bile of hamsters that were given t^C] -
or [3H] - TCDD.
Xenobiotic metabolism is altered after exposure to environ-
mental chemicals. Polycyclic aromatic hydrocarbons are particularly
potent inducers of the enzymes involved in metabolizing xenobiotics,
and cytochrome P-450 is the most important of these enzymes. Cyto-
chrome P-450 is involved in the metabolism of numerous xenobiotic
compounds and is particularly important in terms of biotransformation
of potentially toxic compounds.
Chemical Industry Institute of Toxicology, Research Triangle Park, N.C.
320
-------
The term "cytochrome P-450" describes a component of the so-
called drug-metabolizing enzyme system that is predominantly found
in the liver; however, it is also present in other tissues in smaller
amounts. It gets its name from the absorption maximum of the carbon
monoxide complex with the reduced form of this heme enzyme. Thus,
the carbon monoxide complex with the reduced form of this enzyme
has a peak absorption at 450 nm.
The mechanism of the metabolism of substrates by the cytochrome-
P-450-containing enzyme system is as follows: First, the substrate
is bound to the enzyme. Next, two electrons, supplied by the reduced
pyridine nucleotide NADPH, are transferred by way of a flavoprotein
to cytochrome P-450. Molecular oxygen then binds to the heme iron
of the reduced cytochrome P-450. In a final series of reactions,
one of the atoms of the molecular oxygen is introduced into the sub-
strate to produce an oxygenated substrate. The other oxygen atom
receives the two electrons originally donated by NADPH and, in the
presence of two protons, is converted to water. This enzyme system
is variously called the drug-metabolizing enzyme system, the P-450
monooxygenase enzyme system, and the mixed-function oxidase enzyme
system.
One of the interesting developments in recent years is the dis-
cussion of the heterogeneity of cytochrome P-450 in animal organisms.
Initially- it was believed that cytochrome P-450 was a single enzyme.
Then, emerging data began to suggest that there were probably two
species of this enzyme. Studies had shown that the spectral properties
of the enzyme induced by phenobarbital were different from those in-
duced by 3-methylcholanthrene. The species induced by phenobarbital
321
-------
had ay at ^^0 nm and was called cytochrome P-450. The species
rnrtX
induced by 3-methylcholanthrene had a Ymax of its carbon monoxide
complex with the reduced enzyme at 448 nm, and thus was called cyto-
chrome P-448.
More recent data suggest that there are multiple cytochrome
P-450 enzymes present in biologic tissues. As many as 8 to 12
different species of cytochrome P-450 can be identified in rat
liver by using various inducers of the cytochrome P-4501s such as
phenobarbital or 3-methylcholanthrene. These various species of
cytochrome P-450 differ both in molecular weight and spectral
properties. The biologic reason for the existence of so many
cytochrome P-450 enzymes is unclear. Many may be present because
of evolutionary processes. For example, new forms may have developed
with the capability of metabolizing the different nonpolar exogenous
substrates to which our ancestors were exposed in the changing environment,
The various species of cytochrome P-450 do have somewhat different
substrate specificity and therefore can metabolize some compounds more
readily than others. However, in almost all cases there is a great
deal of overlap in the ability of the various species of cytochrome
P-450 to metabolize various nonpolar compounds.
Polychlorinated biphenyls (PCB's) and polybrominated biphenyls
(PBB's) are potent inducers of the activity of the cytochrome P-450
monooxygenase systems. It recently became clear that if one exposes
animals to Arochlor 1254 (which is in fact a mixture of compounds),
the activity of a number of species of cytochrome P-450 is induced.
322
-------
As noted previously, there are two types of inducers of cytochrome
P-450: the phenobarbltal-like inducers and compounds that induce
species of P-450 with a spectral maximum of the reduced carbon monoxide
complex of approximately 450 nm. The 3-methylcholanthrene-like
compounds induce cytochrome P-450's with a reduced carbon monoxide
spectral maximum that averages about 448 nm. Experiments with PCB's
and PBB's have shown a mixed-type induction, that is, both types of
cytochrome P-450 were induced.
Investigations of individual isomers of PCB's have provided an
understanding of why PCB's and PBB's are mixed inducers. Upon examination
of the inducing properties of 2,4,5,2',4',5'-hexachlorobiphenyl,
one observes a species of cytochrome P-450 with a spectral maximum
of 450 nm. However, with 3,4,5,3',4',5'-hexachlorobiphenyl, the
spectral maximum averages approximately 448 nm. Thus, the mixed-type
induction observed for commercial PCB's and PBB's undoubtedly results
from the fact that the mixture of polychlorinated biphenyls contains
different structural isomers of the polyhalogenated biphenyls that
can induce different spectral types of P-450.
Dr. Joyce Goldstein, at the National Institute of Environmental
Health Sciences (NIEHS), has identified various individual PCB iso-
mers that induce cytochrome P-450 and cytochrome P-448 (Goldstein et
al., 1977). In addition, she has shown that individual PCB isomers
can be strong, moderate, weak, or inactive inducers. The isomers
that are more highly chlorinated and that are chlorinated at positions
ortho to the ring juncture are strong P-450 inducers. The more
highly chlorinated these compounds are, the less readily they are
323
-------
metabolized and the more persistent they are in experimental animals.
This is probably an important factor in their ability to be either
strong or weak inducers of cytochrome P-450.
3,4,3',4'-Tetrachlorobiphenyl and 3,4,5,3',4',5'-hexachlorobiphenyl
are among the few PCB isomers that induce P-450 enzymes with a reduced
carbon monoxide at 448 nm, rather than 450 nm. They are also
H13.X
the most toxic of the PBC isomers so far examined (McKinney et_ al. ,
1976). These two PCB isomers can assume a planar structure since
they are not chlorinated in the position ortho to the ring juncture.
Isomers that are chlorinated ortho to the ring juncture cannot be
planar because the rings have to rotate to accommodate the steric
properties of the halogen atom. Both 2,3,7,8-tetrachlorodibenzo-
j)-dioxin (TCDD) and the corresponding tetrachlorodibenzofuran are
planar molecules and strong P-448 inducers.
PCB's are metabolized by the cytochrome-P-450-containing enzyme
systems. The PCB isomers that are the best substrates for such
systems have two adjacent carbons that are not substituted with
chlorine atoms. Fully chlorinated biphenyl, as well as those biphenyls
containing 9, 8, and 7 chlorines and some compounds containing 6
chlorines, are poor substrates for the P-450-containing monooxygenases.
TCDD does not have two adjacent unchlorinated carbons. Therefore,
it is not a good substrate for the cytochrome P-450 enzyme systems.
Until quite recently, it was believed that TCDD was not metabolized
by mammalian enzymes. However, recent data indicate that TCDD is
metabolized in certain rodent species (Olson et al., 1980; Poiger
et al., 1979).
324
-------
The single dose LD5Q of TCDD is quite variable among species.
For example, in the guinea pig the LD™ is approximately 1 yg/kg; in
the rat, it ranges from 20 yg/kg to 40 yg/kg; in the chicken, from
25 yg/kg to 50 yg/kg; in the mouse, from approximately 100 yg/kg to
300 yg/kg; and in the monkey, it is approximately 70 yg/kg. The
LD5Q of TCDD in the Golden Syrian hamster has been found to be
greater than 3,000 yg/kg when administered by intraperitoneal injec-
tion, and approximately 1,000 yg/kg when administered orally. Thus,
of the species examined so far, the hamster is the most resistant to
the toxicity of TCDD.
The half-life of excretion of TCDD has been examined in various
rodent species. In the guinea pig and rat, the half-life is approxi-
mately 30 days. In the hamster, it is about 12 to 15 days.
Thus, hamsters excrete TCDD much more rapidly than does the guinea
pig or the rat.
Hamsters given 600 yg/kg [^C] TCDD by intraperitoneal injec-
tion excrete approximately 30% to 35% of the radioactivity in urine.
In contrast, the rat excretes only about 2% to 5% by this route.
These and other data suggest that radioactivity in the urine of
hamsters and, perhaps, rats is in the form of metabolites. TCDD
metabolism has been examined in the hamster in greater detail.
Examination for urinary metabolites using high-pressure liquid
chromatography (HPLC) has revealed that all the radioactivity in
urine elutes from the column as metabolites of TCDD. No TCDD itself
could be detected in the urine.
325
-------
The radioactivity excreted in bile following administration
of [^C] TCDD to hamsters was also examined. Again, none of the
radioactivity excreted in hamster bile appeared as TCDD, but, in
fact, occurred as a major metabolite and as a number of minor
metabolites of TCDD.
The nature of the radioactivity remaining in liver following
administration of [ C] TCDD to hamsters was also examined. The
radioactivity in liver was unchanged TCDD, suggesting that TCDD
is very rapidly excreted after metabolism.
If the major peak of radioactivity excreted in bile is treated
with g-glucuronidase and rechromatographed, all the radioactivity is
eluted from the HPLC column at a later time, suggesting that the
major product excreted in bile is a glucuronide derivative of TCDD.
This implies that TCDD is being hydroxylated and a glucuronide
formed before excretion in bile.
An examination of TCDD metabolism ±n_ vitro (using hamster micro-
somes) shows evidence of covalent binding of radioactivity from
[ C] TCDD to macromolecules of the microsomes. Little or no
binding is observed when TCDD is incubated with boiled microsomes
or with microsomes in the absence of an NADPH-generating system.
When isolated primary hepatocytes from hamsters are used, TCDD is
metabolized to products that accumulate in the incubation medium.
The major metabolite formed using this hepatocyte system appears
to be the same glucuronide derivative detected in bile.
326
-------
In summary, the P-450-containing enzyme systems are important
in the metabolism of foreign compounds. They are strongly induced
by chlorinated aromatic compounds. Particularly potent inducers of
these enzymes are TCDD, the dibenzofurans, and 3,4,5,3*,4',5'-hex-
achloro- or hexabromobiphenyl.
327
-------
REFERENCES
Goldstein, J. A., P. Hickman, H. Bergman, J. D. McKlnney, and
M. P. Walker. 1977. Separation of pure polychlorinated biphenyl
isomers into two types of inducers on the basis of induction of
cytochrome P-450 or P-448. Chem. Biol. Interact. 17:69-87.
McKinney, J. D., K. Chae, B. N. Gupta, J. A. Moore, and J. A.
Goldstein. 1976. Toxicological assessment of hexachlorobiphenyl
isomers and 2,3,78-tetrachlorodibenzofuran in chicks.
I. Relationship of chemical parameters. Toxicol. Appl.
Pharmacol. 36:65-80.
Olson, J. R. , T. A. Gasiewicz, and R. A. Neal. 1980. Tissue
distribution, excretion, and metabolism of 2,3,7-tetrachlorodibenzo-p-
dioxin (TCDD) in the golden Syrian hamster. Toxicol. Appl. Pharmacol.
56:78-85.
Poiger, H., and Ch. Schlatter. 1979. Biological degradation of TCDD
in rats. Nature 281:706-707.
328
-------
DISCUSSION
DR. MILLER: You showed the hamster to be relatively resistant
to TCDD. Does this have foreseeable implications for the treatment
of people—to make them react more like hamsters?
DR. NEAL: The interesting thing to determine is whether humans
are more like hamsters or more like guinea pigs. We could, for
example, use an in vitro metabolizing system to compare the abilities
of rat, hamster, and human tissue to metabolize TCDD. But the key
question is whether the hamster is more resistant to TCDD because
it metabolizes it more rapidly or because of other factors. The
most popular theory concerning TCDD's acute toxicity is that TCDD
binds to a receptor in cytosol. This receptor-TCDD complex is then
translocated into the nucleus, followed by the coordinate derepression
of the synthesis of a number of enzymes. The increase in enzyme
synthesis caused by this derepression is postulated to cause TCDD's
acute toxicity.
An examination of hamster liver reveals the presence of about
the same level of receptor as seen in rat; however, as noted previously,
the hamster is orders of magnitude less susceptible to TCDD toxicity
than is the rat. Thus, toxicity does not appear to be related to the
level of receptor in livers of the two species. I don't know whether
the increased rate of metabolism or the increased rate of TCDD excretion
explains why the hamster is so resistant to TCDD's acute toxicity. But
it is clearly an interesting avenue to pursue.
329
-------
DR. MURPHY: Is there any relationship between Goldstein's
observation of apparent correspondence between inducibility of
cytochrome P-450 monooxygenase activity and toxicity of these
compounds to various species, especially the hamster?
DR. NEAL: TCDD does induce enzymes in the hamster. I don't
know that we can say "easily," because we haven't compared the rat
and hamster at the same dosage level to see if we get the same degree
of induction of cytochrome P-450 or other enzymes in both species.
However, 3-methylcholanthrene can induce the same enzymes as TCDD,
but the toxicity seen with TCDD is not evident in the case of
3-methylcholanthrene. Therefore, there is no reason to believe that
the induction of the enzymes by TCDD is responsible for the acute
toxicity. There may be an induction of the activity of some unknown
enzymes or an increase in synthesis of some unknown proteins, perhaps
a membrane protein, which could lead to toxicity.
DR. GARATTINI: If you give the same dose, for example, a rela-
tively low dose of 10 yg/kg to rats and hamsters, do you get the same
kind of induction for P-448?
DR. NEAL: We need to look at that—that is, we need to compare
the ability of TCDD to induce P-448 in the hamster and other species.
DR. GARATTINI: That would be extremely important in deter-
mining whether the sensitivity of the system or the metabolism is
involved. Do you have any suggestion about the preferential site
for metabolism of TCDD? Which chlorine is going out?
330
-------
DR. MEAL: We don't know. We now have about 3 g of the major
TCDD metabolite isolated in a reasonably pure form, and we are
going to perform mass spectral analysis on it. It should be easy
to tell whether we have lost a chlorine.
DR. GARATTINI: It would be difficult to find the position.
DR. NEAL: Yes. However, if all four chlorines remain in the
metabolite, then the four unsubstituted positions are equivalent.
In other words, if all four chlorines are still there, then all
four unsubstituted positions would be equivalent relative to
hydroxylation. Thus, the key questions are whether the tricyclic
structure is still intact and whether all four chlorines remain in
the metabolite. If that is the case, then the identification of
the metabolite should be fairly straightforward.
DR. GARATTINI: We tried to determine if there was covalent
binding to proteins with liver microsomes from the rat and mouse.
<
We could find no evidence that agrees with what you found.
As to the half-life of TCDD in hamsters, you tried to explain
the reduced toxicity by the relatively shorter half-life; however,
in mice, the half-life is still shorter. Yet, the toxicity is much
higher, on the order of 50 Vg/kg.
DR. NEAL: I wasn't aware that the half-life studies in the
mouse had been done. You carried out these studies?
DR. GARATTINI: Yes. Apparently TCDD is excreted, but we
couldn't find any metabolite excreted as such.
DR. NEAL: About covalent binding, we do see it with the hamster
microsome system, but not with the rat.
331
-------
DR. MILLER: In Seveso, where there was exposure to TCDD, which
induces P-450, would you expect newborn infants to have a shorter
duration of neonatal jaundice because of the induction of the enzyme?
DR. NEAL: It would probably depend on whether the species of
P-450 induced will in fact work on bilirubin. Different species
of P-450 display different substrate specificity. The species
induced by TCDD may not, in fact, work on bilirubin. I don't know
the answer to your question.
DR. MILLER: If the people still had TCDD in their bodies, is
there a drug test to measure the influence of the TCDD on the
metabolism of the drug?
DR. NEAL: No. There are so many compounds—literally hundreds
of them—known to induce cytochrome P-450 or P-448 or this whole
spectrum of P-450 enzymes; a number of these compounds are drugs.
We are exposed to many inducers on a daily basis because they are
present in our environment, including our food. Therefore, I think
a drug metabolism test would be useless in trying to assess a body
burden of TCDD. The only way to identify TCDD in tissue is to detect
it analytically.
DR. GARATTINI: A metabolism test in children might not
definitively measure aryl hydrocarbon hydroxylase (AHH), but it
certainly could indicate a strong increase.
DR. NEAL: Except that TCDD is not a particularly good inducer
of P-450 in a quantitative sense. The activity of P-450-like reac-
tions is actually lower in TCDD-treated animals than in untreated
ones. Thus, in TCDD-treated rats, benzamphetamine-demethylase activity
is actually decreased whereas benzo(a_)pyrene activity is increased.
332
-------
If people with TCDD ate cabbage or brussels sprouts the day
before the enzyme activity was measured, you would get a false
result. Certain foods contain some very potent inducers of those
AHH enzymes.
DR. GARATTINI: In given control conditions, perhaps in
children, because they are not smokers, AHH might serve as an indi-
cator if you find a substantial increase.
DR. MEAL: You would probably have to have a great amount of
data on the level in normal individuals before you could attribute
an increase in AHH to the presence of TCDD.
SPEAKER (UNIDENTIFIED): What fraction of the material that was
radiolabeled was excreted as glucuronide versus the fraction bound
covalently to cellular macromolecules?
DR. NEAL: If we do the experiments with either [14C]- or [3H]-
labeled TCDD (1,6-[3H] TCDD), we get covalent binding of radio-
activity. I have shown data using [C]-labeled TCDD. If we use
1,6-[3H]-TCDD, we obtain an extra peak off the HPLC columns. This
peak elutes very early and appears to be tritiated water. These
data also suggest that P-450 is hydroxylating TCDD in the 1, 4, 6, or
o
9 position and, as a result, tritium is lost as f^O.
Thus, we see the binding both of [14C]- and tritium-labeled TCDD.
Both C and tritium are observed in the metabolites isolated from
urine and bile, although an extra peak is observed in the HPLC columns
o
from urine and bile of animals administered [JH]-TCDD. The compound
333
-------
appears to be tritiated water. You can evaporate off the radioactivity
in that HPLC peak.
DR. MURPHY: Couldn't you use a microsomal enzyme assay as
suggestive evidence of exposure to TCDD or related compounds?
DR. NEAL: No, because so many compounds induce P-450. Also,
I don't think you can use an in vivo test of drug metabolism. In
some methods, you give a drug you know is metabolized by an enzyme
system and look at excretion products in urine. In this case, the
test would be hard to control. It would be difficult to determine
a normal level in a human population. Fred Guengerich of our center
has been studying human cytochrome P-450's, isolating and purifying
them from liver. He has looked at about 20 human liver samples and
purified enzymes from some of them. The amounts of enzyme he sees
are quite variable. The P-450 monooxygenase activities appear to be
quite variable, as is the substrate specificity. Therefore, I
don't believe you can use an enzymatic assay to analyze body burden
of TCDD.
334
-------
Hepatic Toxicity of TCDD
Silvio Garattini1
2,3,7,8-Tetrachlorodibenzodioxin (TCDD) is a contaminant that
shows an exceptional tropism for liver in several animal species.
The chemical persists in liver for a long time, inducing marked
morphologic and biochemical changes. The liver subcellular distri-
bution of TCDD changes with time after administration. Hypertrophy
of the reticuloendothelial system and induction of various microsomal
enzymes have been reported in several animal species. Differences
among mouse strains have been observed; this difference correlates
with the immunosuppresive effect of TCDD. TCDD affects porphyrin
metabolism in rats treated with low doses for long periods. Changes
in porphyrin excretion may represent a sensitive indicator of TCDD
exposure. Studies have also attempted to affect the persistence
of TCDD in liver.
Animals living in the Seveso area were useful in defining the
contamination and for starting a number of studies. At the time of
the accident, some 81,000 domestic animals lived in Zones A, B, and
R. Mortality, mostly among rabbits and poultry, started some days
after the accident, rising markedly within the first 2 weeks. By
the end of August 1976, 3,281 dead animals had been recorded. Figure
1 shows overall animal mortality figures and the location of farms
where deaths occurred. Rabbit mortality is given in Table 1. Deaths
occurred on 75% of the farms in Zone A, on 22% of the farms in Zone
B, and on 14% of those in Zone R. In Zone A, 31.9% of the rabbits
died, 8.8% in Zone B, and 6.8% in Zone R. On farms where rabbits died,
the percentage of mortality was 42% in Zone A, 23% in Zone B, and 16%
Istituto di Ricerche Farmacologiche "Mario Negri," 62, Via Eritrea,
20157 Milan, Italy.
335
-------
Overall animal mortality
(July 10 - August 31, 1976
Animal
No. of dead animals
over total
Rabbits
Other small
farmyard
animals
Cattle
Horses
Pigs
Sheep
Goats
2,062/24,885
1,219/55,545
0/349
O/ 49
0/233
O/ 21
O/ 49
FIGURE 1. Distribution of animal mortality in the Seveso area;
• = Farms with mortality. On the basis of soil
analysis for TCDD three zones were designated:
A = TCDD levels in soil from <0.75 to 5,477 vg/m2
B = TCDD levels in soil from <0.75 to 43.8
R = TCDD levels in soil from <0.74 to 5 vg/m
Vgm
2
Data from Servizio Veterinazio della Lombardia
336
-------
TABLE 1
Rabbit Mortality on Farms in the Contaminated Zones,
July 10 - August 31, 1976a
No. of Rabbits
No. of Farms in Area
No. of Rabbits on Farms
Where Deaths Occurred
Farms with
Zoneb
A
B
R
Total
Total
1,089
4,814
18,982
24,885
Dead
348
426
1,288
2,062
%
31.9
8.8
6.8
8.3
Total
45
303
1,398
1,746
deaths
34
67
208
309
%
75.5
22.1
14.9
17.7
Total
825
1,801
7,783
10,409
Dead
348
426
1,288
2,062
%
42
23
16
19.8
aData from Servizio Veterinario della Lombardia.
See Figure 1 for soil contamination levels in Zones A, B, and R.
337
-------
ICMESA (Industrie Chimiche Meda Societa Azionaria) factory, and there
is strong evidence that mortality was directly related to how close
to ICMESA fresh fodder was gathered. Farms where rabbits were fed
commercial feed and/or fodder collected before the accident or far
from ICMESA had lower mortality rates. TCDD was analyzed in samples
from domestic and wild animals that had died or been slaughtered for
safety reasons to monitor contamination and to measure the accumula-
tion of TCDD in living organisms (Table 2). Milk collected 2 weeks
after the accident (from cows living close to the chemical plant and
fed contaminated fodder) contained up to 8 g of TCDD per liter.
Rabbits that died (or that were slaughtered within 1 month of the
accident) had levels of TCDD in liver ranging from 0.25 ng/g to 633
ng/g. Twenty percent of 900 autopsied rabbits had a syndrome char-
acterized by substernal and retrosternal edema, hemorrhagic tracheitis,
pleural serous hemorrhage, and dystrophic lesions of hepatic tissue.
TCDD played an important part in this toxic syndrome, as indicated
by the presence of large amounts of the chemical in the liver of
many of the animals. Similar liver alterations and TCDD concentra-
tions (100 ng/g) were found in rabbits that died or were killed
within 1 month of receiving a single oral dose of 60-120 yg/kg of
TCDD.
Rabbits were believed to be a reliable model of the risk of TCDD
exposure because the chemical accumulates in their tissues and becomes
detectable in their liver after ingestion of the compound or the environ-
mental exposure. At the end of 1976, all rabbits in zones A, B,
338
-------
TABLE 2
Overall TCDD Analyses of Animals from Contaminated Zones
and Surrounding Areas3
Animals
Farm animals
Rabbitb
Poultry
Cattle
Horses
Pigs
Goats
Guinea pigs
Cats
Wildlife
Hares
Field mice
Rats
Earthworms
Frogs
Snakes
Cow's milk
No. of Samples
698 (all
83 liver)
43
12
13
25
4
1
% TCDD
Positive
62
42
5
17
0
68
0
0
6 (liver)
14 (whole body)
1 (pool of four
livers)
2 (pool)
1 (liver)
1 (liver)
103
67
100
Maximum Detected
Level (ppb)
633
24
10
88
1,253
13
49
28
12
0.2
26
Total
1,007
aGas chromatography-mass fragmentography was used for detection,
identification, and measurement of TCDD in tissues (limit of
detection 0.25 ng/g). A detailed description of the method is
reported in Fanelli et_ al. , 1980.
^Figures include rabbits kept in the special test plots of con-
taminated ground for experimental purposes.
339
-------
and R were killed, and pilot breedings were established in Zone R in
r\
fields with known soil contamination (approximately 1.5 yg/nr). TCDD
was analyzed in liver from these rabbits after periods ranging from
60 to 300 days. Samples of rabbits from farms outside Zone R were
also analyzed. Table 3 shows the percentages of positive samples
from 1976 to 1979. In pilot breedings and surrounding areas, during
the years 1978 and 1979, only few animals of those examined were found
contaminated and the concentration of TCDD in their liver never
exceded 750 pg/g.
The study was helpful for several reasons: The animal data
confirmed the division into three geographic areas and uncovered
some contamination outside Zone R. Pilot breedings showed that
traces of TCDD could still be found in liver of rabbits kept in
f\
areas with a soil contamination of the order of 1.5 yg/m .
Studies are still in progress on the relationship between
liver pathology and liver concentrations of TCDD. In the
rabbit, as in the rat and mouse, the liver is not only the site
of accumulation of TCDD, but also is more severely altered than
other organs (Gupta e£ al. , 1973; McConnell et^ al_. , 1977; Rose et^
al. , 1976). When rabbits were treated orally with [3H]-TCDD (2.5,
25, and 500 ng/kg every other day) for 30 days, the levels of [3H]-TCDD
almost reached a steady state (Table 4) as predicted by the rate of
elimination of TCDD from liver (half-life about 7 days). Microscopic
340
-------
TABLE 3
Presence of TCDD as Percentage of Positive Samples in Liver
of Rabbits from Contaminated Zones, 1976-1979a
Surrounding
Time of Sampling Zone A Zone B Zone R Areas
July-August 1976 97 92 75 27
(29) (37) (122) (22)
September-December 98 91 77 43
1976 (42) (11) (127) (7)
1977 all animals killed 93b 12
(31)b (41)
1978 - - 5b 15.9
(59)b (44)
1979 - - 2.4b 2.8
(42)b (70)
aFigures in parentheses are numbers of animals analyzed for TCDD.
For the determination of TCDD in biologic samples, see Fanelli
et_al., 1980. The sensitivity of the method was 0.25 ng of TCDD
per gram.
bln 1977, except for a small number of pilot breedings set up in
Zone R, there were no more animals in Zones A, B, or R. These
figures refer to the pilot breedings. Samples were also taken
from farms located outside Zone R (surrounding areas).
341
-------
TABLE 4
Concentration of [3H]-TCDD in Rabbit Liver3
Oral Dose
Every Other 3H-TCDD in the Liver, ng/g Theoretical
Day, ng/kg 8 Days 22 Days 30 Days Steady State
2.5 - b 0.15 + 0.03 0.12 + .008 0.16
25 0.8+0.09 0.59+0.14 0.92+0.12 1.4
500 7.9 + 1.7 27.7 + 1.7 25.3 + 2 23.4
aRabbits were given [3H]-TCDD orally at a dosage of 2.5 ng (0.4 Ci),
25 ng (4 Ci) and 500 ng (5.9 Ci) per kilogram every other day for
30 days. Three or four rabbits were killed after 8, 22, and 30
days. A portion of the livers was homogenized and its radioactivity
extracted, identified by thin-layer chromatography, and determined
by liquid scintillation counting.
Not measured.
342
-------
examination of the liver showed alterations related to the liver
concentrations of [3H]-TCDD (swelling of cells, fatty degenerations,
cell necrosis), and also showed that the time of exposure to TCDD
was very important. For example, liver with 7.9 ng of TCDD per gram,
dissected from animals exposed for 8 days to 500 ng/kg every other
day, had alterations similar to those of liver with 0.6 ng of TCDD
per gram, dissected from animals exposed for 30 days to 25 ng/kg
every other day.
Over time, TCDD can change location within the cell as seen by
administering TCDD to mice and then looking at different subcellu-
lar fractions of the liver (nuclear, mitochondrial, microsomal,
soluble). Figure 2 shows that the distribution profile of TCDD in
subcellular fractions 7 days after receipt of a single dose of
TCDD is very similar to that observed when TCDD is added in_ vitro.
The largest percentage is in the nuclear fraction, and smaller
amounts are in the mitochondrial, microsomal, and soluble fractions.
However, 14 days after TCDD injection, the subcellular distribution
profile is completely different. There is markedly less TCDD in
the nuclear fraction, but more in the mitochondrial and microsomal
fractions. This shift of TCDD in the cells occurs without substan-
tial changes in total levels of TCDD in the liver. Thus, even when
the total TCDD level is the same, its subcellular distribution can
differ.
343
-------
% OCCURRENCE
IN FRACTION
AND S.D. n = 3
B
50-
40-
30-
^k ^k
20-
10-
p
1
I
1
p2
T
1
^3
•L
S
p
1
T
1
PZ
1
P3
I_
s
FIGURE 2. TCDD distribution in mouse liver subcellular fractions.
Male C57BL/6J mice were injected intraperitoneally with
4 mCI, 25 pg/kg[3H]-TCDD and were killed either 7 (b)
or 14 (c) days later. [3H]-TCDD (88 ng/g) was added
to livers from nonintoxicated mice (A). Tissues were homo-
genized in five volumes of potassium chloride, 1.15% buffer
plus 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid
(HEPES), 20 mM, pH 7.5, and subcellular fractions were pre-
pared according to Lucier et^ a.L. (1973). P, (nuclear), P~
(mitochondrial), and P« (microsomal) fractions were the
sediments of centrifugations 670, 10,000, and 105,000 xg,
respectively, and S (soluble fraction) was the high speed
supernatant. Radioactivity in fractions was assayed
(see Table 9). Occasional chromatography did not reveal
evidence of biotransformation of TCDD in any fraction.
Total liver TCDD concentrations (ng/g + S.D.) were
103.9 + 14.0 and 95.0 + 25.4 for B and C, respectively.
Three livers were processed individually for each group.
From unpublished data of P. Coccia, T. Croci, and L. Manara,
1980.
344
-------
Continuous treatment with TCDD causes porphyria in both labor-
atory animals and humans (Goldstein et_ al_. , 1973; Poland et al. ,
1971). In mice, porphyria is inherited, as is the induction of aryl
hydrocarbon hydroxylase (AHH) (Jones and Sweeney, 1977, 1980). How-
ever, it was not known what dose of TCDD, given to animals continuously,
had no porphyrogenic effect and caused no microsomal enzyme induction.
Nor was it clear whether evaluation of the pattern of excretion of
porphyrins throughout TCDD administration was a sensitive parameter
for monitoring TCDD exposure. TCDD was given to female rats at
three dose levels—10, 100, and 1,000 ng/kg, once a week for 45
weeks. At the end of treatment, body weights were unchanged in
animals receiving the two lower doses, but an 8% loss was observed
in animals receiving the highest dose. Liver weight increased pro-
portionally to the dose administered (19% increase with the lowest
dose, corresponding to 1.4 ng of TCDD per kilogram per day).
The P-450 and mixed-function oxidase (MFO) activities are shown
in Table 5. P-450 was raised only in animals receiving 1,000 ng/kg/wk.
AHH was markedly increased at all doses. There was also a marked dose-
dependent increase in the levels of 7-ethoxy-coumarin-O-deethylase (7-EC),
which is known to be induced by pretreatment with both phenobarbital
and 3-methylcholanthrene. Increases were noted at the lowest (4-fold)
and highest (30-fold) doses. The same trend of induction was observed
when enzymatic activities (per nanomole of cytochrome P-450) were
calculated. The porphyrogenic effect of TCDD was marked in animals
345
-------
TABLE 5
Liver Microsomal Cytochrome and Mixed-Function Oxidase
Activity After Chronic TCDD Treatment*1
TCDD Dose
(ng/g/week)
None
10
100
1,000
P-450
(nmol/mg protein
0.56 + 0.05
0.60 + 0.06
0.68 + 0.03
0.91 + 0.13C
AHH 7 -EC
(pmol/min/mg) (pmol/min/mg protein)
19.54+ 8.80
143.25 + 17.51b
341.67 + 12. 4b
283.17 + 50.25°
131.27 +
583.43 +
1,565.33 +
4,513.31 +
31.68
103.40°
25.14b
l,105.99b
aData from Cantoni e_t al. , 1981a. CD-COBS rats were killed after
45 weeks of treatment. Values represent mean + S.E. of three or
four animals. Cytochrome P-450, AHH, and 7-EC were determined as
described by Omura and Sata (1964), Nebert and Gelboin (1968),
Greenlee and Poland (1978), respectively.
bp < 0.01, student's t-test.
cp < 0.05, student's t-test.
346
-------
treated with the highest concentration (1,000 ng/kg/wk) for 10 months.
The onset and development of porphyria was monitored by measuring
the levels of the various porphyrins in urine throughout the experi-
ment. In all three treated groups, there was an increase in the
absolute amount of coproporphyrin excreted during the 24-hour period.
After 10 months, this increase was significant, even in animals
receiving the lowest dose of TCDD (Figure 3).
The levels of other toxicologically important porphyrins ex-
creted in urine were also measured. The amount of uroporphyrin was
increased in animals receiving 100 and 1,000 ng of TCDD per kilogram
per week. The heptacarboxylic porphyrin level was elevated only in
animals receiving the highest dosage. Hexacarboxylic porphyrin was
never detected in control animals, but was measurable in urine of
animals receiving 1,000 ng/kg/wk.
In animals given the highest doses of TCDD, increases in the
excretion of porphyrins with a high number of carboxyl groups were
accompanied by changes in the relative percentage distribution of
each porphyrin. The principal change is the inversion of the
copro:uro ratio (from 6 to 8 months after treatment), as com-
pared to that in control animals. Total urinary prophyrin levels
were also elevated 10 months after treatment. The difference at
the lowest dose of TCDD was significant (Table 6). This increase was
not accompanied by any change in porphyrins in liver. The marked rise
347
-------
et
CC.
^:
C\J
cr
>-
x
o_
cc
o
QL
LJ
CC
100-
50-
15-
9-
6-
3-
2-
1-
D CONTROL
Q TOngTCDD/kg
m 100
E]iooo
*
/
/
rn / ^
4678 4678
**
4678
7 8
FIGURE 3. Urinary porphyrin excretion after TCDD treatment for
45 weeks. Values represent the mean of four animals.
4, coproporphyrin; 6, hexacarboxylic porphyrin;
7, heptacarboxylic porphyrin; 8, uroporphyrin.
* = p < 0.05, student's t-test; ** = p < 0.01, student's
t-test. Redrawn from Canton! et al., 1981b.
348
-------
TABLE 6
Hepatic Accumulation of TCDD and Total Porphyrln Content of Liver
~~and Urine After TCDD Treatment for 45 Weeks3
TCDD Dose
(ng/kg/week)
Liver TCDD
(ng/g wet tissue)
Total Liver
Porphyrins
(ng/g wet tissue)
Total Urinary
Porphyrins
(nmol/24 hr)
None
10
100
1,000
0.05
1.55 + 0.02
4.74 + 0.54
30.70 + 3.34
1.59 + 0.25
0.92 + 0.20
4.11 + 2.96
724.67 + 2.67C
2.27 + 0.49
5.55+ 0.85b
7.62 + 1.79b
196.89 + 63.14°
aData from Cantoni et al., 1981b. CD-COBS rats were killed
after 45 weeks of treatment; values represent mean + S.E. of three
or four animals. TCDD content of liver and total liver porphyrins
were determined as described by Fanelli et^ al. (1980) and Abbritti
and De Matteis (1971, 1972). Total urinary porphyrins were measured
from the sum of each individual porphyrin, as described by Doss
(1970).
bp <0.05, student's t-test.
cp <0.01, student's t-test.
349
-------
in porphyrins in the liver and urine of animals receiving the largest
dose of TCDD reflects established prophyria.
In conclusion, a clear porphyrogenic effect was present only
when the animals were treated with 1,000 ng of TCDD per kilogram
weekly for 45 weeks. At the 10 ng dose, coproporphyrin excretion
was altered, possibly an early sign of intoxication. When porphyria
is established, the observed pattern of porphyrin excretion supports
the hypothesis that activity of the liver enzyme uroporphyrinogen
decarboxylase diminishes, as has been shown for hexachlorobenzene
(Elder, 1978).
It is well known that TCDD causes atrophy of the thymus and of
the thymus-dependent areas of lymphoid organs in the mouse, rat,
guinea pig, and monkey (McConnell et al., 1978; Vos et al., 1973).
When TCDD is administered during the perinatal period through ma-
ternal treatment (before or after birth), severe consequences are
observed in the immune system and cell-mediated responses are
inhibited (Vos and Moore, 1974). More recent studies (Mantovani et
£l_. , 1980; McConnell et_ £l_. , 1978; Vecchi et_ al. , 1980 indicate that
a single dose of TCDD (l-30yg/kg) administered orally or intraperi-
toneally to 6- to 8-week-old C57BL/6 mice does not reduce cell-mediated
activities (splenocyte blastogenic response to mitogens, graft versus
host reaction, macrophage, and natural killer-cell cytotoxic activities)
on a per cell basis, but does reduce the number of splenocytes and
macrophages, indicating general impairment of the immune response
(Mantovani et al., 1980).
350
-------
In contrast, the same dose levels produce a functional, dose-
dependent inhibition of primary and secondary antibody production
to T-dependent and T-independent antigens (sheep erythrocytes and
pneumococcal polysaccharide-type III), as indicated by the smaller
number of antibody-producing cells per 10 splenocytes (Vecchi et al.,
1980). Immunosuppression is marked and long-lasting. Thirty micro-
grams per kilogram inhibits humoral responses by 80% up to 6 weeks
after a single dose.
TCDD is also a potent inducer of AHH in several animal species
(Kouri et_ al., 1974; Poland e£ a^. , 1974). It induces hepatic AHH
in mouse strains such as DBA/2, AKR, and SJL/J, which are not suscepti-
ble to 3-methylcholanthrene and have been described as nonresponsive
strains (Poland et al., 1974). Nonresponsive strains require dose
levels of TCDD approximately 10 times higher than those needed by
responsive strains (e.g., C57BL/6, C3H/He, BALB/c, and A/J) to induce
the AHH system (Robinson et_ al. , 1974). This enzyme complex is
believed to play an important role in the biotransformation of some
aromatic compounds, producing and/or destroying toxic metabolites.
Thus, the immunosuppressive activity of TCDD was investigated in
both responsive and nonresponsive mice. In addition to the humoral
antibody production, thymus weight was used as parameter of TCDD
exposure.
As shown in Table 7, 6 yg of TCDD per kilogram significantly
reduced thymus weight in responsive mice (C57BL/6 and C3H/He) only,
not in the nonresponsive strains DBA/2 and AKR. Moreover, the same dose
level caused a marked inhibition of antibody humoral response (80-90%)
351
-------
TABLE 7
Effect of TCDD on Primary Humoral Response in
Different Mouse Strains'1
Mouse Strain
C57BL/6
C3H/He
TCDD
(yig/kg)
0
1.2
6
30
0
1.2
6
PFC/Spleen X + _!_ S.E.
27334 (24106 - 30993)
10514 ( 8698 - 12708)b
3723 ( 2666 - 5199)°
915 ( 765 - 1095)°
108239 (105177 - 111390)
37378 ( 30028 - 46402)°
5383 ( 3562 - 8137)°
% Control
100
38
14
3
100
35
5
DBA/2
AKR
0
1.2
6
30
0
1.2
6
16413 (13973 - 19279)
13583 (11522 - 16014)
8362 ( 7630 - 9160)
5852 ( 4168 - 8215)1
51185 (50176 - 52213)
31682 (26040 - 38546)
100
83
50
36
100
61.9
Unpublished data by A. Vecchi, A. Mantovani, M. Sironi, W. Luini, 1980.
TCDD was injected intraperitoneally in acetone roil solution (1:6),
7 days before antigen challenge (4.108 SRBC); the test was performed
5 days later.
p <0.05, student's t-test.
cp <0.01, student's t-test.
352
-------
in C57BL/6 and C3H/He mice. In DBA/2 and AKR mice, the response was
reduced by only 40-50% (Table 8). These data support the suggestion
that the different sensitivities of various mouse strains to AHH
induction reflect a more general difference in susceptibility to
TCDD toxicity (Poland and Glover, 1980).
In laboratories at the Istituto Ricerche Farmacologiche "Mario
Negri", investigators are searching for a way to shorten the persistence
of TCDD in the body and to reduce its toxic action (Coccia et_ al_. , 1981).
In these studies a single dose of [3H]-TCDD was administered orally
by stomach intubation or intraperitoneally to male C57BL/6J mice,
weighing 18 to 21 g, which were maintained in controlled environmental
conditions. The animals were fed diets containing different additives
o
or standard powdered chow. Controls received standard chow throughout
the experiment. The amount of TCDD in the livers of the treated
mice was measured days later by radioassay and liquid scintillation.
Thin-layer chromatography (n-hexane : ethyl ether, 16:1; R = 0.80) of
selected samples gave no evidence of TCDD biotransformation in any
group of animals. Representative results are summarized in Table 9.
After 14 days approximately 10% to 17% of the administered dose is
recovered in the livers of animals fed the standard diet. The addition
of 4% cholestyramine does not change the results. However, when 5%
charcoal is added to the diet, the amount of TCDD present in the
liver is substantially reduced. The concentrations of TCDD in the livers
Isotopics, USA, S.A. 51C:/mM, 90% radiochemically pure
dissolved in acetone: corn oil, 1:6, 5 ml/kg.
2"Altromin R," A. Rieper, Italy.
353
-------
of animals receiving charcoal were as much as 60% lower than those
in controls receiving only TCDD. The addition of cholic acid, which
stimulates bile flow, also lowered the amount of TCDD in the liver.
In another experiment, mice were given charcoal 3 days after
they received TCDD; thus, the mice absorbed most of the TCDD
through the intestinal tract before the charcoal could act (Unpublished
data by P. Coccia, T. Croci, and L. Manara, 1981.) Even so, there was
a much lower percentage of TCDD in their liver than in a control group.
Finally, two different preparations of charcoal reduced the amount
of TCDD in liver after intraperitoneal administration of TCDD.
Interestingly enough, the effect of these two preparations was further
increased by the addition of 1% cholic acid to the diet. Additional
experiments indicated that charcoal and cholic acid are effective in
reducing the mortality of mice given lethal doses of TCDD.
354
-------
TABLE 8
Effect of TCDD on Thymus Weight in Different Mouse Strains£
TCDD Thymus Weight
Mouse Strain ^yg/kg) (mg) % of Control
C57BL/6 0 64.3+2.9 100
1.2 57.0 + 4.0 89
6 45.0 + 4.3b 70
30 24.8 + 3.0b 38
C3H/He 0 36.8+2.5 100
1.2 29.3 + 2.9 80
6 22.3 + 2.1b 60
30 14.0 + 2.2b 38
DBA/2 0 27.7+2.2 100
1.2 22.4 + 2.9 81
6 23.4 + 3.5 84
30 20.3 + 4.5 73
AKR 0 54.7 + 3.5 100
6 40.0 + 9.3 73
aUnpublished data by A. Vecchi, A. Mantovani, M. Sironi, and W. Luini,
1980. TCDD was injected intraperitoneally in acetone:oil solution
(1:6) 12 days before killing.
bp <0.05, student's t-test.
355
-------
TABLE 9
Radioactivity in Liver of Mice Fed Chow with Different Additives
Measured 14 Days After a Single Dose of [3H]-TCDD was Administered3
Dose Recovered in Liver, mean percentage + S.D.
(Number of animals)
Diet
7.6 yg/kg
Oral Dose
11.0 yg/kg
Intraperi-
toneal Dosec
5% Animal charcoal
13.8 + 1.3d
12.5 yg/kg
Oral Dosec
Standard chow
5% Vegetable charcoal
0.5% Cholic acid
17.3 + 3.2
(6)
6.3 + l.ld
("6)
j
13.1 + 2.8d
(6)
19.2 + 1.6
(6)
15.3 + 2.6e
(7)
10.3 + 0.8
(5)
5.6 + 0.8e
(5)
5% Vegetable charcoal
plus 1% Cholic
acid
10.9 + 2.4C
5% Animal charcoal
plus 1% Cholic
acid
4% Cholestyramine
14.5 + 1.7
(6)
9.1 + l.lc
f
dDATA from Coccia et al., 1981.
Diets with added test substances were given to mice immediately after
TCDD dose.
cDiets with added test substances were given to mice 3 days after
TCDD dose.
p <0.05, Duncan's new multiple test.
ep <0.01, Duncan's new multiple test.
Not investigated.
356
-------
REFERENCES
Abbritti, G., and F. De Matteis. 1971-1972. Decreased levels of
cytochrome P-450 and catalase in hepatic porphyria caused by
substituted acetamldes and barbiturates. Importance of the
allyl group in the molecule of the active drugs. Chem. Biol.
Interact. 4:281-286.
Cantoni, L., M. Salmona and M. Rizzardini. 1981b. Porphyrogenic
effect of chronic treatment with 2,3,7,8-tetrachlorodibenzo-p-
dioxin in female rats. Dose-effect relationship following
urinary excretion of porphyrins. Toxicol. Appl. Pharmacol.
57:156-163.
Cantoni, L., M. Rizzardini, G. Belvedere, R. Cantoni, R. Fanelli,
and M. Salmona. 1981a. Induction of mixed-function oxidase by
chronic treatment with 2,3,7,8-tetrachloro-dibenzo-£-dioxin
in female rats. Toxicology 21:159-167.
Coccia, P., T. Croci, and L. Manara. 1981. Less TCDD persists in
liver 2 weeks after single dose to mice fed chow with added
charcoal or cholic acid. Br. J. Pharmacol. 72: 181-182.
Doss, M. 1970. Analytical and preparative thin-layer chomatography
of porphyrin methyl esters. J. Clin. Chem., Clin., Biochem. 8:
197-207.
Elder, G. H. 1978. Porphyria caused by hexachlorobenzene and other
polyhalogenated aromatic hydrocarbons. Pp. 157-200 in Handbook
of Experimental Pharmacology, Vol. 44, Heme and Hemoproteins.
Springer Verlag, Berlin.
Fanelli, R., M. P. Bertoni, M. Bonfanti, M. G. Castelli, C. Chiabrando,
G. P. Martelli, M. A. Noe, A. Noseda, and C. Sbarra. 1980. Rou-
tine analysis of 2,3,7,8-tetrachlorodibenzo-p-dioxin by gas chrom-
atography-mass fragmentography in biological samples from the
contaminated area of Seveso. Bull. Environ. Contain. Toxicol.
24: 818-823.
Goldstein, J. A., P. Hickman, H. Bergman, and J. G. Vos. 1973.
Hepatic porphyria induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin
in the mouse. Res. Commun. Chem. Pathol. Pharmacol. 6:919-928.
Greenlee, W. F., and A. Poland. 1978. An improved assay of 7-ethoxy-
coumarin 0-deethylase activity: Induction of hepatic enzyme
activity in C57BL/6J and DBA/2J mice by phenobarbital, 3-methyl-
cholanthrene and 2,3,7,8-tetrachlorodibenzo-p-dioxin. J. Pharm-
acol. Exp. Ther. 205:596-605.
357
-------
Gupta, B. N. , J. G. Vos, J. A. Moore, J. G. Zinkl, and B. C. Bullock.
1973. Pathologic effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin
in laboratory animals. Environ. Health Perspect. 5:125-140.
Jones, K. G., and G. D. Sweeney. 1977. Association between induction
of aryl hydrocarbon hydroxylase and depression of uroporphyrinogen
decarboxylase activity. Res. Commun. Chem. Pathol. Pharmacol. 17:
631-637.
Jones, K. G., and G. D. Sweeney. 1980. Dependence of the por-
phyrogenic effect of 2,3,7,8-tetrachlorodibenzo(p)dioxin upon
inheritance of aryl hydrocarbon hydroxylase responsiveness.
Toxicol. Appl. Pharmacol. 53:42-49.
Kouri, R. E. , H. Ratrie, S. A. Atlas, A. Niwa, and D. W. Nebert,
1974. Aryl hydrocarbon hydroxylase induction in human lymph-
ocyte cultures by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Life
Sci. 15:1585-1595.
Lucier, G. W., 0. S. McDaniel, G. E. R. Hook, B. A. Fowler, B. R.
Sonawane, and E. Faeder. 1973. TCDD-induced changes in rat
liver microsomal enzymes. Environ. Health Perspect. 5:199-
209.
Mantovani, A., A. Vecchi, W. Luini, M. Sironi, G. P. Candiani, F.
Spreafico, and S. Garattini. 1980. Effect of 2,3,7,8-tetra-
chlorodibenzo-p-dioxin on macrophage and natural killer cell-
mediated cytotoxicity in mice. Biomedicine. 32: 200-204.
McConnell, E. E., J. A. Moore, J. K. Haseman, and M. W. Harris.
1977. The comparative toxicity of chlorinated dibenzo-p-dioxin
isomers in mice and guinea pigs. Toxicol. Appl. Pharmacol. 44:
335-356.
McConnell, E. E., J. A. Moore, and D. W. Dalgard. 1978. Toxicity
of 2,3 ,7 ,8-tetrachlorodibenzo-p-dioxin (TCDD) in rhesus monkeys
(Macaca mulatta) following a single oral dose. Toxicol. Appl.
Pharmacol. 43:175-187.
Nebert, D. W., and H. V. Gelboin. 1968. Substrate-inducible micro-
somal aryl hydroxylase in mammalian cell culture. I: Assay and
properties of induced enzyme. J. Biol. Chem. 243:6242-6249.
Omura, T., and R. Sato. 1964. The carbon monoxide-binding pigment
of liver microsomes. I: Evidence for its hemoprotein nature.
J. Biol. Chem. 239:2370-2378.
Poland, A., and E. Glover. 1980. 2,3,7,8-Tetrachlorodibenzo-p-dioxin:
Segregation of toxicity with the Ah locus. Mol. Pharmacol. 17:
86-94.
358
-------
Poland, A. P., D. Smith, G. Metter, and P- Fossick. 1971. A health
survey of workers in a 2,4-D and 2,4,5-T plant. With special
attention to chloracne, porphyria cutanea tarda, and psychologic
parameters. Arch. Environ. Health 22:316-327.
Poland, A. P., E. Glover, J. R. Robinson, and D. W. Nebert. 1974.
Genetic expression of aryl hydrocarbon hydroxylase activity.
Induction of mono-oxygenase activities and cytochrome P-450
formation by 2,3,7,8-tetrachlorodibenzo-p-dioxin in mice
genetically "nonresponsive" to other aromatic hydrocarbons.
J. Biol. Chem. 249:5599-5606.
Robinson, J. R., N. Considine, and D. W. Nebert. 1974. Genetic
expression of aryl hydrocarbon hydroxylase induction: evi-
dence for the involvement of other genetic loci. J. Biol.
Chem. 249:5851-5859.
Rose, J. Q., J. C. Ramsey, T. H. Wentzler, R. A. Hummel, and P- J.
Gehring. 1976. The fate of 2 ,3 ,7 ,8-tetrachlorodibenzo-p-
dioxin following single and repeated oral doses to the rat.
Toxicol. Appl. Pharmacol. 36:209-226.
Vecchi, A., A. Mantovani, M. Sironi, W. Luini, M. Cairo, and S.
Garattini. 1980. Effect of acute exposure to 2,3,7,8-tetra-
chlorodibenzo-p-dioxin on humoral antibody production in mice.
Chem. Biol. Interact. 30:337-342.
Vos, J. G., and J. A. Moore. 1974. Suppression of cellular immunity
in rats and mice by maternal treatment with 2,3,7,8-tetrachlor-
odibenzo-p-dioxin. Int. Arch. Allergy Appl. Immunol. 47:777-794.
Vos, J. G., J. A. Moore, and J. G. Zinkl. 1973. Effect of 2,3,7,8-
tetrachlorodibenzo-p-dioxin on the immune system of laboratory
animals. Environ. Health Perspect. 5:149-162.
359
-------
DISCUSSION
DR. MOORE: Over time, there was a marked difference in the
compartmentalization of TCDD within the cell. Have you tried to
study the charcoal depletion phenomenon, perhaps 14 days after you
administered TCDD, when it has left the cytosol and pretty much gone
to the microsomal or the mitochondrial fractions? Did you see if,
indeed, you could still move it out then?
DR. GARATTINI: Yes. I don't have any results yet, but we are
trying to understand if there is a redistribution because of this
treatment.
The mechanism is probably dual: one, to drain TCDD from the
bile, and two (perhaps more important), to bind the TCDD that is
excreted.
DR. WEINBERG: Did you determine exactly what the immediate
cause of death was for these animals? Obviously, there is a large
toxic effect on the liver and other organs. But were animals dying
from toxicity alone? Or was death due to respiratory paralysis or
neurologic destruction? Was autopsy analysis detailed enough to
determine if other kinds of tissue destruction had occurred?
DR. GARATTINI: I don't know.
DR. MOORE: Some pathologists have necropsied the animals and
studied the histopathology. They described a lesion in rodent liver
and the absence of a lesion in guinea pig. Then, when asked the
360
-------
cause of death, they say, "I don't know." The simplest answer is
that animals look as if they had wasted away.
DR. GARATTINI: Some pathologists believe that the kind of
liver degeneration observed, particularly after high doses, is
not compatible with life; but it is difficult to answer the question.
DR. NEAL: Have you looked at LD.-Q determinations in the
charcoal-treated C57BL/6 mice to see if toxicity is decreased, even
though TCDD is cleared from the liver?
DR. GARATTINI: Yes. There is a decrease in mortality.
DR. NEAL: The hamster, for example, may have 100 times more
TCDD than the amount found in the liver of the rat, but still sustain
no liver damage. Your observation of variable effects on liver with
varying doses suggests something really much more complicated.
DR. GARATTINI: Certainly.
DR. NEAL: The wasting syndrome was mentioned, but death can
occur from TCDD without this effect. We fed rats parenterally
(involuntarily). They gained weight equally with control animals,
but they died "right on schedule," without having lost any weight.
Weight loss is an indirect effect of the compound.
DR. REHDER: Pathologists are not always able to say how many
alterations have been the cause of death, but they can look at a
certain histologic picture of the liver and determine whether the
361
-------
alterations were sufficient to cause the death of an animal or a human
being. Your histologic pictures showed fatty metamorphosis, swelling
of liver cells, and single cell necrosis. These changes are more or
less reversible. Could irreversible changes, like fibrosis, bile-
duct proliferation, or even cirrhosis be expected? In your experi-
ments, did you wait long enough to see if irreversible changes occurred
in the liver?
DR. GARATTINI: Not for liver, but we did for other types of
toxicity, such as immunodepression. We found that if the dose is
low enough mice can recover completely in about 60 days.
DR. MOORE: TCDD-induced thymic atrophy, after a long enough
interval, will show repopulation of the cortical area. The severity
of liver pathology in some of the rodent species could be interpreted
as a contributing cause of death. However, for the Rhesus monkey or
the guinea pig, there is no pathology in the liver.
DR. MURPHY: In your three strains of mice, the C57BL/6 strain
was the one susceptible both to toxicity amd thymus atrophy.
The thymus was about twice the weight of that gland in the
DBA2 strain, which was resistant. Is there any connection?
DR. GARATTINI: There are marked differences in the thymus
weight, depending on strain. I don't think it has any significance;
percentage-wise, there was no decrease in the thymus weight in the
DBA2, even with doses that were several times the amount effective in the
sensitive strain.
362
-------
These strains were actually selected on the basis of the effect
of TCDD on P-448 induction. The strain that is not sensitive to
TCDD, or that is less sensitive to TCDD in terms of liver induction,
is also less sensitive in terms of thymus atrophy and also from the
point of view of antibody production.
I don't know if the three things are linked. What may be
important for this kind of toxicity is a different sensitivity of
whatever is the receptor.
363
-------
In Vivo DNA Damaging Activity, In Vivo Covalent DNA Binding and
Bacterial Mutagenicity as Related Quantitatively to Carcinogenic
Potency
S. Parodi, M. Taningher, and L. Santi1
The following is a brief report on two different studies: one
on 16 hydrazine derivatives, and another on 21 compounds in different
chemical classes.
For these compounds, the following parameters were examined:
o DNA-damaging potency (DPI). This was evaluated according
to the in vivo alkaline elution assay, using the following formula:
DFI = (K - K )/(dosage administered in mMol/kg), where K is the
elution rate constant of eluted DNA (as a first approximation a
parameter directly proportional to the amount of damage).
o Covalent binding index (CBI) according to the formula:
CBI = (ymol chemical bound per Mol nucleotides)/(mMol chemical
administered/kg).
o Mutagenic potency (MPI) in the Ames test, evaluated according
to the following formula: MPI = (hystidine revertants over controls
per plate/(nMol of chemical per plate).
o Carcinogenic potency (OPI) evaluated as follows: OPI =-ln
(fraction of tumor-free animals)/[(cumulative dose in mMol/kg) x
(time of exposure)]. For the set of hydrazine derivatives, OPI was
calculated using a slightly different formula.
Istituto Scientifico per lo Studio e la Cura dei Tumori and
University Department of Oncology, Viale Benedetto XV, 10
16132 Genoa.
364
-------
The results of these two experiments were the following:
(1) The Ames1 test was absolutely not predictive for the 16 hydrazine
derivatives. (2) The ^.n vivo DNA damage was predictive for the 16
hydrazine derivatives (r = 0.6). (3) For the 21 compounds of
different chemical classes, all three tests were predictive in the
following order: CBI: r = 0.60; DFI: r = 0.55; MPI: r = 0.42.
The correlation between MPI and DFI was very modest (r =
0.15). The correlation between CBI and DFI was very strong (r = 0.80).
A multiple correlation of OPI = f (DFI, MPI) showed improvement in
respect to the simple correlations (^,03 = 0«64 against rj~ = 0.55
and r,~ = 0.42). This seems to suggest that a battery of two short-
term tests can indeed be more predictive than each one of them
individually.
365
-------
Liver Injury
Sheldon D. Murphy1
The liver is a common site of injury resulting from exposure to
a wide variety of chemical contaminants. Altered hepatic function
and structure often serve as early indicators of excessive exposure.
However, because of the liver's large reserve and capacity for repair,
chemically induced effects in the liver often do not represent the
most serious health effects produced by a particular chemical. It
is almost axiomatic that mammals exposed to halogenated organics
will have functional and morphologic changes in liver. The nature
of these changes and their importance in relation to other effects
can differ, depending on the specific compound, tjhe species exposed,
the duration of exposure, and the dosage. So many substances induce
hepatotoxicity that liver function tests are not very specific early
indices of health effects associated with environmental exposure.
More insight into the mechanisms of hepatoxicity (in relation to
other toxic manifestations) will permit a better assessment of the
utility of these tests in epidemiologic followups of chemical contami-
nation.
For the king of Babylon stood at the parting of the way,,
at the head of the two ways, to use divination: he made
his arrows bright, he consulted with images, he looked
in the liver.—Ezekiel 21:22
The basis of the following discussion of hepatotoxicity was
drawn largely from Zimmerman (1978). The three hepatotoxic candidates
for area-wide contamination selected for discussion in this presentation
are polychlorinated biphenyls (PCB's), tetrachlorodibenzodioxin (TCDD),
and hexachlorobenzene (HCB).
The response of the liver to exposure to hepatotoxic chemicals can
take several forms. If the exposure is brief and at a high dosage,
and if the chemical is capable of injuring parenchymal cells, acute
Division of Toxicology, University of Texas Medical School at
Houston, Houston, Texas.
366
-------
necrosis can be expected. If injury is sufficiently severe, it can
lead to death from hepatic failure in a few days. However, complete
recovery can occur, because the liver generally is capable of complete
functional and morphologic repair if injury is not too massive or too
frequent. A single exposure to chemicals such as white phosphorous
can result in massive liver injury, which can lead to macronodular
cirrhosis. On the other hand, cirrhosis is more likely to be the
outcome of prolonged or frequently repeated exposures that produce
repeated subtle injuries. These subtle injuries may never be mani-
fested by characteristic findings in tests of acute liver injury;
but in the long term, fibrosis occurs and the chronic clinical condi-
tion of cirrhosis becomes manifest.
At the cellular level, acute exposure to hepatotoxins generally
results first in injury to intracellular and cell-surface membranes.
With this injury, cellular constituents commonly leak out of the cell
and into the extracellular fluids. A measurement of these constituents
in plasma is the basis for many of the clinical tests for liver
injury. This increased plasma activity of enzymes such as serum
glutamic oxaloacetic transaminase (SCOT), and pyruvic transamlnase,
or isocitric dehydrogenase) has been used as an index of acute
hepatocellular injury.
Carbon tetrachloride is often used to illustrate chemical-induced
hepatotoxicity. The endoplasmic reticulum of the liver converts
carbon tetrachloride to a reactive free radical form. This reactive
intermediate actually injures the endoplasmic reticulum and causes
decreased activities of enzymes (e.g., glucose-6-phosphatase)
367
-------
associated with that intracellular structure. But this reactive
molecule is also capable of reacting with other intracellular and
cell-surface membranes. These injurious interactions cause altera-
tion of the critical semipermeable characteristics of the cell, with
resultant leakage of constituents into the extracellular fluid.
This injury is also manifest by certain structural changes that can
be observed by light and electron microscopy. Biochemical manifesta-
tions of structural injury can be also detected by assaying for
hepatocellular constituents in plasma.
One common outcome of exposure to hepatotoxic chemicals is the
accumulation of lipid in the hepatocytes. Free fatty acids are
incorporated into triglycerides in the hepatocytes, and the tri-
glycerides are then transported out of the cell. This transport
mechanism depends on the presence of very low density lipoproteins.
Thus, anything that interferes with the synthesis of such lipoproteins
can lead to an impairment of the egress of lipids from the cell and
accumulation of triglycerides in the liver (steatosis).
Various compounds that typically cause the accumulation of
lipid as one of their manifestations may do so by quite different
mechanisms. Dialkyl nitrosamines react with the DNA or RNA that
controls the formation of the template for synthesis of the apoprotein
that becomes a part of the low density lipoprotein. On the other hand,
carbon tetrachloride attacks the membranes of the endoplasmic reticulum
and the ribosomes, causing the same net effect—reducing the synthesis
of the transport protein. To varying degrees, both aliphatic and
aromatic halogenated compounds can produce both necrosis and steatosis.
368
-------
Table 1 illustrates that a wide variety of aliphatic halogenated
hydrocarbons produce both fatty liver and centrizonal necrosis. A
few of the compounds seem to be fairly specific in producing only
steatosis, and some closely related compounds (some merely stereoisomers
of each other) apparently have very little potential for producing
liver injury. Although this discussion does not focus on these
compounds, consideration of liver injury as a consequence of area-wide
chemical contamination probably cannot ignore these relatively low
molecular weight, halogenated compounds. Even though these chemicals
may not be persistent, many of them are volatile, and they may be ones
to which people are exposed through slow vaporization from dumps, leaking
storage containers, and so forth.
Table 2 lists several aliphatic and aromatic halogenated compounds
that produce hepatic necrosis in laboratory animals. Several of the
more persistent of these, such as dichlorodiphenyltrichloroethane
(DDT), chloronaphthalenes, and chlorobenzenes, have contributed to
area-wide contamination. Some of these chemicals, such as chloroform,
are volatile, low molecular weight compounds. Nevertheless, some
are fairly ubiquitous at low concentrations in urban potable water
supplies because they are formed in the process of water chlorination.
Almost all of these compounds, in sufficient doses, can produce
centrizonal necrosis of the liver. They are fairly specific in
their site of injury within the liver. Most of them will also produce
steatosis. As indicated in Table 2, the halogenated compounds are
not the only environmental contaminants that produce liver injury.
369
-------
TABLE 1
Liver Injury by Halogenated Aliphatic Hydrocarbons5
Steatosis
and
Centrizonal
Necrosis
CC1,
4
CI4
CBr,
CCl3Br
CHCK
CHI3
CHBr,
CHC12CHC12
CH2BrCH2Br
CHC12CC13
CHC1=CC10
CH3CHC1CH3
CH3CHC1CH2C1
Steatosis
Only
CH2ClBr
CHC1=CHC1 (cis)
cci2=cci2
CH3CH2CHC1CH3
Slight
Steatosis
or
Injury
CH3C1
CH3Br
CH.,1
CHC1=CHC1 (trans)
CH3CH2CH2CH2C1
From Zimmerman, 1978, with permission.
370
-------
TABLE 2
Some Halogenated and Nonhalogenated Compounds that Produce Hepatic
Necrosis in Experimental Animalsa
Compounds
Site of Accompanied
Halogenated Necrosis by Steatosis
Bromobenzene CZ +
Bromotrichloromethane CZ +
Carbon tetrachloride CZ +
Chlorobenzenes CZ
Chloroform CZ +
Chloroprene CZ +
Chlorinated biphenyls CZ +
Chloronaphthalenes CZ -
Chloropropane CZ +
Dichloropropane CZ +
Dichlorodiphenyltri-
chloroethane CZ +
Ethylene dibromide CA
Ethylene dichloride CZ +
lodoform CZ
Methylchloroform CZ +
Methylene chloride CZ +
Nonhalogenated
Aflatoxins CZ,PZ +
Allyl compounds PZ -
-Amanitin CZ +
Anthrapyrimidine CZ,MZ
Antimony - +
Arsenic (inorg.) CZ,PZ
Beryllium MZ
Botulinus toxin CZ
Dimethylnitrosamine CZ +
Dinitrobenzene CZ +
Ferrous sulfate PZ
Manganese compounds PZ -
Naphthalene CZ +
Paraquat CZ,MZ
Pyridine CZ
Xylidine CZ +
^Adapted from Zimmerman, 1978.
bCZ=centrizonal; MZ=midzonal; PZ=peripheral zonal.
371
-------
The liver is also a target organ for a variety of other environmental
contaminants, not all of which produce centrizonal necrosis. The
allyl compounds, for example, act primarily in the portal region.
Also, several of the nonhalogenated compounds produce necrosis without
characteristically producing fatty livers.
Kuratsune et al. (1972) reviewed the signs and symptoms of
patients with Yusho disease caused by rice oil contaminated with
PCB's. Jaundice was the effect most likely to be associated with liver
injury. However, this symptom was reported only at an 11% frequency.
Many of the other symptoms occurred much more often. For example,
brown pigmentation of the nails was reported at a frequency of
75%-83%. One might suspect that this symptom had some relationship
to porphyria, which is often associated with liver injury. Nevertheless,
in summarizing the clinical and pathologic findings in Yusho patients,
Kuratsune (1972) indicated that there is little objective evidence
of liver injury based on standard clinical tests such as plasma
enzyme assays. On the other hand, some liver biopsies showed
increased endoplasmic reticulum, which is somewhat characteristic of
the response observed in laboratory animals exposed to PCB's.
The effects of PCB's in humans, monkeys, and rats are summarized
in Table 3. Hepatic hypertrophy is listed as being a response to
PCB's exhibited by human and subhuman primates as well as by the rat.
But the health significance of hepatic hypertrophy depends on what is
causing it, e.g., fatty infiltration, neoplasia, or induced endoplasmic
reticulum. Characteristically, PCB's tend to cause increases in liver
372
-------
TABLE 3
Responses of Primates and Rats to PCB'sa
Response Humans Monkey Rat
Susceptibility to toxicity High High Moderate
Acne Yes Yes No
Hyperpigmentation of skin Yes Infants only No
Alopecia NAb Yes No
Hyperactive meibomian glands Yes Yes No
Conjunctivitis Yes Yes No
Edema of eyelids Yes Yes No
Subcutaneous edema Yes Yes No
Keratin cysts in hair follicles Yes Yes No
Hyperplasia of hair follicle
epithelium Yes Yes No
Gastric hyperplasia NA Yes No
Thymic atrophy NA NAb Yes
Hepatic hypertrophy Yes Yes Yes
Liver enzyme change NA Yes Yes
Decreased number of red blood
cells Yes Yes No
Decreased hemoglobin Yes Yes No
Serum hyperlipidemia Yes Hypolipidemia Yes
Leukocytosis Yes Yes No
^International Agency for Research on Cancer, 1978, with permission.
NA=not available.
373
-------
weight and the membranous smooth endoplasmic reticulum. For most of
the other toxic symptoms and signs caused by PCB exposure, the rat
does not appear to provide a very good model for humans.
Bruckner (1973, 1974) studied rats exposed to Aroclor 1242 and
described some of the effects of that compound on the liver. Ad-
ministering 100 mg/kg of Aroclor 1242 to rats on alternate days for
3 weeks resulted in an approximately two-fold increase in relative
liver weight and in SCOT. The increase in SCOT is a minimal effect,
since a much greater increase would be expected with serious liver
injury. Routinely, some reduction in hematocrit level was also
observed. In the rat, the most striking effect of exposure to Aroclor
was an approximately tenfold increase in the liver microsomal hydroxylation
of acetanilide. However, there was not a uniformly large increase
in all microsomal enzyme activities as oxidative demethylase activity
increased only a small amount. This liver enzyme-inducing effect
persisted in rats exposed to PCB's. After a single dose, the peak
induction of hydroxylation activity lasted approximately 5 days.
Furthermore, the level of activity remained significantly above
that in controls for a longer time, possibly reflecting a flux of
PCB's from lipid storage depots into the bloodstream and the liver
where the chemical could continue to stimulate the microsomal system.
At exposures to concentrations as little as 5 ppm, there was, within
a few days, a twofold increase in the activity of acetanilide hydroxylase
activity, which persisted for several weeks after the animals were
returned to control diets. Additional studies showed that feeding
rats Araclor 1242 for 6 months at 5 and 25 ppm in the diet caused a
374
-------
significant increase in liver weight and in liver lipids. There was
also a significant twofold increase in the urinary coproporphyrin
levels in animals fed as little as 5 ppm. Histologically, the liver
had an increased fat content as shown by a Sudan-4 stain. Otherwise,
it appeared normal.
Liver damage has been reported to occur in humans exposed to
TCDD (National Institute of Environmental Health Sciences, 1978).
A review (Kociba et al., 1978) of the results of numerous toxicity
studies in laboratory animals indicates that the changes in clinical
chemistry associated with liver injury occur at chronic dosage levels
of less than 1 mg/kg/day. Thus, it appears that lesions in the liver
may be a fairly sensitive index of effect, but apparently no more
sensitive than a reduction in fertility in rats (Murray et_ suU, 1979),
Both the thymus and liver are target organs of TCDD. An early
study of reported effects of TCDD on relative weights of liver and
thymus in guinea pig and mouse indicated that significant changes in
the relative weight of liver require a higher dosage than does a
change in thymus weight in both species (Harris et_ al. , 1973). This
finding could be taken to suggest that thymus is a more sensitive
tissue than liver in terms of response to TCDD exposure. However, a
later report (McConnell et_ al., 1978) indicates that changes in
relative thymus and liver weights were about equally sensitive for
detecting an effect, and neither was more sensitive than reduced
body weight gain.
375
-------
The discovery of a cytoplasmic TCDD receptor, which appears to
have binding affinity for chlorinated dioxins that correlates with their
toxicity (Poland et al., 1979), suggested that the presence of these
receptors in various tissues may correlate with specific organ
susceptibility. Studies by Carlstedt-Duke (1979) in the rat showed that
the ratios of TCDD receptor concentrations in extrahepatic tissue
to the liver concentrations were 1.81, 1.50, 0.90, 0.25, and 0.20
for thymus, lung, kidney, testis, and brain, respectively. Thus,
although liver contained an appreciable quantity of receptor protein, the
thymus and lung contained more. The high concentration of TCDD-binding
protein in thymus is consistent with the sensitivity of this tissue
to injury from TCDD. The lung, however, has not generally been
considered a specific target tissue. These molecular studies tend
to support the position that the liver is not the most TCDD-sensitive
tissue. At the same time, the high concentration of cytoplasmic
TCDD-binding protein in the lung fails to make a clear case for the
causal association of this protein with specific TCDD organ toxicity.
An early report by Goldstein _et^ al. (1973) demonstrated that
TCDD increased liver porphyrin levels in mice given four weekly doses
of 25 mg/kg. Under those subacute exposure conditions, lower doses of
1 or 5 mg/kg/week increased liver weight significantly. At the lower
doses of TCDD, these increases occurred in the absence of statistically
significant changes in porphyrins or of aminolevulinic acid (ALA)
synthetase activity. The 5 mg/kg dose group had slight pathological
changes in liver. On the other hand, in addition to the increase in
376
-------
porphyrins accompanying the higher dose, there was an increase in
ALA synthetase activity, increased iron content, marked liver changes,
and 50% mortality of the treated group. These observations suggest
that liver porphyrin content was not a good early indicator of toxicity
in the mouse. However, in chronic exposures of the rat to TCDD,
Kociba et^ _al_. (1978) reported increased urinary porphyrin excretion
to be one of the most sensitive assays of an effect. The porphyrogenic
action of TCDD may thus be more prominent in chronic studies than in
the rather high-dose, subacute studies. The specific porphyrin
derivatives measured and possible species or strain differences also
need to be considered when making such comparisons.
Jones and Butler (1974) made some interesting observations on
liver histopathology induced by TCDD. Rats given single TCDD doses
of 200 ug/kg were sequentially killed and changes in their livers
were observed. Necrotic foci in the centrolobular zone were reported
7 days after the animals received the TCDD. At 14 days, clearly
dilated sinusoids were surrounded by enlarged parenchymal cells.
Multinucleated cells lined up along the trabeculae in the liver and
sometimes surrounded the sinusoids. These multinucleated cells,
which the authors considered as somewhat characteristic of response
to TCDD exposure, sometimes appeared to have as many as 20 nuclei.
Jones and Butler concluded that the TCDD affected the plasma membrane,
and that the multinucleated giant cells arose from a coalescing of
parenchymal cells—secondary to damage to the cell membranes. Other
histologic changes observed at longer times included some mild fibrotic
377
-------
changes and a disordered endoplasmic reticulum. Electronmicrographic
observations of liver from TCDD-treated rats were described by Norback
and Allen (1972) as including a tortuous disarray of the endoplasmic
reticulum, sometimes engulfing other organelles and lipid inclusions.
Hexachlorobenzene (HCB) also produces histopathologic changes which,
when viewed by electronmicroscopy, include a disordering of the
endoplasmic reticulum that is not too different from that resulting
from TCDD. Kuiper-Goodman and coworkers (1977) reported the morphologic
and biochemical effects of HCB in relation to tissue residues in
rats fed various doses in the diet for 15 weeks. The test dosages
ranged from 0.5 to 32 mg/kg/day. Even at the lowest dose, HCB accumulated
in liver, but approximately 10 times as much accumulated in adipose
tissue. In addition, relative liver weights were increased above
those of controls at the two highest dosages of 8 and 32 mg/kg/day.
These increased liver weights were associated with increased microsomal
oxidase activities. There was consistent histologic evidence of
mild liver injury at a dosage of 2 mg/kg/day, and some suggestion of
an effect even at the lowest dosage (0.5 mg/kg/day). Liver porphyrin
concentrations increased in female rats fed 8 and 32 mg/kg/day during
the 15-week exposure period. Lower dosages did not cause liver
porphyrin levels to increase during feeding, but an occasional animal
at the 2- or 0.5 mg/kg/day dose had an increased liver porphyrin
level during the 16-week recovery period. These studies indicated
that histopathologic examinations were at least as sensitive in
detecting effects of HCB on liver as were measurements of the level
378
-------
of liver porphyrin. The results of plasma enzyme assays of sorbitol
dehydrogenase were compared with porphyrin concentrations in liver
from rats given 8 and 32 mg/kg/day. The extent of leakage of the
liver enzyme into plasma and the increase in porphyrins in liver
showed good correlation.
Den Tonkelaar et_ al. (1978) observed pigs fed RGB's in dosages
of 0.05, 0.5, 5.0, and 50 mg/kg/day for 90 days. There were signifi-
cant increases in liver, kidney, and thyroid weights for the 5.0
mg/kg group. Histopathologic changes were observed only in liver of
animals receiving less than 50 mg/kg/day. Whorls of smooth endo-
plasmic reticulum were occasionally seen in pigs fed as little as
0.5 mg/kg/day. No changes of any kind were observed in animals that
received 0.05 mg/kg/day. At the 0.5 mg/kg dosage, microsomal oxidase
activity in liver was increased significantly as was urinary copro-
porphyrin excretion. The blood level of HCB corresponding to these
minimal biochemical and histopathologic changes in pigs given 0.5
mg/kg/day ranged from 0.235 to 0.285 ppm. This closely corresponded
with a 0.338 ppm level of HCB in blood reported (Mazzei and Mazzei,
1973) in a case of chronic HCB intoxication in a human with clinical
signs of porphyria and liver damage. Furthermore, the daily dosages
that produced evidence of liver injury in both the rat and the pig
correspond to the estimated dosage (1 to 4 mg/kg/day) associated
with symptoms of porphyria that resulted from an incident of mass
chronic poisoning of humans with HCB (Cam and Nigogosyan, 1963).
In conclusion, although tests for liver injury may prove especially
sensitive for use in epidemiologic followups after area-wide chemical
379
-------
contamination involving compounds such as HCB, compounds such as
PCB's and TCDD may injure the liver only at dosages or durations of
exposure that are more severe than those causing serious effects in
other organ systems. Clinical tests for liver function or structural
integrity are not likely to be very discriminating in an epidemiologic
study because so many chemicals that can contaminate the environment
have some capacity for liver injury. Thus, although, "looking in the
liver'1 has been one way to search for knowledge and guidance since
ancient times, this technique cannot be considered the panacea for
today's environmental health problems.
380
-------
REFERENCES
Bruckner, J. V., K. L. Khanna, and H. H. Cornish. 1973. Biological
responses of the rat to polychlorinated biphenyls. Toxicol. Appl.
Pharmacol. 24:434-448.
Bruckner, J. V., K. L. Khanna, and H. H. Cornish. 1974. Polychlorinated
biphenyl-induced alteration of biologic parameters in the rat. Toxicol.
Appl. Pharmacol. 28:189-199.
Cam, C., and G. Nigogosyan. 1963. Acquired toxic porphyria cutanea
tarda due to hexachlorobenzene: Report of 438 cases caused by this
fungicide. J.A.M.A. 183:88-91.
Carlstedt-Duke, J. M. B. 1979. Tissue distribution of the receptor
for 2,3,7,8-tetrachlorodibenzo-p-dioxin in the rat. Cancer Res.
39:3172-3176.
Den Tonkelaar, E. M., H. G. Verschuuren, J. Bankovska, T. De Vries,
R. Krves, and G. J. Van Esch. 1978. Hexachlorobenzene toxicity in
pigs. Toxicol. Appl. Pharmacol. 43:137-145.
Goldstein, J. A., P. Hickman, H. Bergman, and J. G. Vos. 1973.
Hepatic prophyria induced by 2,3,7 ,8-tetrachlorodibenzo-p-dioxin
in the mouse. Res. Commun. Chem. Pathol. Pharmacol. 6:919.
Harris, M. W., J. A. Moore, J. G. Vos, and B. M. Gupta. 1973.
General biological effects of TCDD in laboratory animals.
Environ. Health Perspect. Exp. Issue No. 5:101-109.
International Agency for Research on Cancer. 1978. Evaluations
of the Carcinogenic Risk of Chemicals to Humans: Polychlorinated
Biphenyls and Polybrominated Biphenyls. Volume 18. World Health
Organization, Lyon, France. 140 pp.
Jones, G., and W. A. Butler. 1974. A morphological study of the
liver lesion induced by 2,3,7,8-tetrachlorodibenzo-p-dioxin in
rats. J. Pathol. 112:93-97-
Kociba, R. J., D. G. Keyes, J. E. Beyer, R. M. Carreon, C. E. Wade,
R. P. Dittenber, R.P. Kalnins, L. E. Frauson, C. N. Park, S. A. Barnaed,
R. A. Hunmel, and C. G. Humiston. 1978. Results of a two year
chronic toxicity and oncogeniclty study of 2,3,7-tetrachloro-p-dioxin
in rats. Toxicol. Appl. Pharmacol. 46:279-303.
Kuiper-Goodman, T., D. L. Grant, C. A. Moodie, G. 0. Korsrud, and
I. C. Munro. 1977. Subacute toxicity of hexachlorobenzene in
the rat. Toxicol. Appl. Pharmacol. 40:529-549.
381
-------
Kuratsune, M., T. Yoshimura, J. Matsunaka, and A. Yamaguchi. 1972.
Epidemiologic study on Yusho, a poisoning caused by ingestion of
rice oil contaminated with a commercial brand of polychlorinated
biphenyls. Environ. Health Perspect. Exp. Issue No. 1:119-128.
Mazzei, E. S. , and C. M. Mazzei. 1973. Une intoxication par un
fungicide, 1'hexachlorobenzene, souillant les grains de ble.
Sem. Hop. Paris. 49:63-69.
McConnell, E. E., J. A. Moore, J. K. Haseman, and M. W. Harris.
1978. The comparative toxicity of chlorinated dibenzo-p-dioxins
in mice and guinea pigs. Toxicol. Appl. Pharmacol. 44:335-356.
Murray, F. J., F. A. Smith, K. D. Nitschke, C. G. Humiston, R. J.
Kociba, and B. A. Schwetz. 1979. Three-generation reproduction
study of rats given 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)
in the diet. Toxicol. Appl. Pharmacol. 50:241-252.
National Institute of Environmental Health Sciences/International
Agency for Research on Cancer. June 1978. Long Term Hazards
of Polychlorinated Dibenzodioxins and Polychlorinated Dibenzo-
furans. Joint working group report, Internal Technical Report
78/001. International Agency for Research on Cancer, Lyon, France.
Norback, D. H., and J. R. Allen. 1972. Chlorinated aromatic hydro-
carbon induced modifications of the hepatic endoplasmic reticulum:
Concentric membrane arrays. Environ. Health Perspect. Exp. Issue
No. 1:137-143.
Poland, A., W. F. Greenlee, and A. S. Kende. 1979. Studies on the
mechanism of action of the chlorinated dibenzo-p-dioxins and related
compounds. Ann. N.Y. Acad. Sci. 320:214-230.
Zimmerman, H. J. 1978. Hepatotoxicity, The Adverse Effects of
Drugs and Other Chemicals on the Liver. 1978. Appleton-Century-
Crofts, New York. 597 pp.
382
-------
DISCUSSION
DR. GOLDBERG: Dr. Murphy mentioned other effects of TCDD
that have not been discussed, and I wonder particularly if sensory
effects or peripheral neuropathies have been studied in Seveso.
DR. DARDANONI: Most of the data have been collected, but they
have not yet been analyzed. It is difficult to correlate exposure
to symptoms and objective signs. Among a number of central nervous
symptoms, such as fatigue, sleepiness, and headache, only headache
has been found to correlate statistically with chloracne in comparison
with the nonchloracne group. This finding holds true both in adults
and children. As for effects on the peripheral nervous system, recent
studies have demonstrated some sort of damage, measured by nerve
conduction and myographic aspects, in addition to direct symptoms such
as fatigue, paresthesia, and other subjective observations.
Some of these studies have used control groups rather far from
the polluted area and statistically significant differences were
found. The method of data collection on subjective symptoms is not
entirely free of suspicion of bias. Objective data for the myographic
and conduction studies are also under review by several experts.
DR. LINGAMEN: Could some of the neurologic and perhaps some of
the psychiatric symptoms be due to porphyria? Acute intermittent
porphyria, which is probably a different disease, is associated with
psychiatric and neurologic symptoms. Dioxin patients develop
porphyria cutanea tarda, and there may be some overlapping of these
porphyric syndromes.
383
-------
DR. NEAL: The porphyria cutanea tarda observed periodically
has not been seen consistently. Thus, it is difficult to know whether
porphyria cutanea tarda is a symptom of TCDD contamination in humans.
DR. MURPHY: In Turkey, perhaps 15 years ago, a lot of porphyria
cutanea tarda resulted. I don't recall any outstanding psychiatric
problems associated with that incident.
DR. GOLDBERG: I don't specifically know that incident, but in
ongoing studies with lead poisoning, investigators are asking that
same question: Are the CNS symptoms associated with lead poisoning
due to porphyria? The answers are still very early in development.
384
-------
Panel on Epidemiologic Approaches to Measurement and
Assessment of Exposures
The following Panel was convened to consider the information
presented in the Workshop and to extract, if possible, some general
principles of investigation of human health impact from extensive
environmental contamination:
Prof. P- Bruzzi Prof. Cesare Maltoni
Prof. Luigi Dardanoni Dr. Robert Miller, Moderator
Dr. Clark Heath Prof. Bruno Paccagnella
Dr. George Hutchison
DR. MILLER: In the last minutes of the workshop, we need to
develop some ideas about how to organize future studies of area-
wide contamination. What action should be taken first after a
catastrophe, whether it is radiation contamination, a runaway
reaction in a factory, or poisoning of a river?
DR. HUTCHISON: Your question is not how to handle the emergency
involving sick people—whether to take them to a hospital or how to
treat them—but how to approach the situation. It is generally very
difficult to get any cool and objective investigation underway if
the population has already realized that something "bad" has happened,
and that some "bad" person is probably responsible for it. An early
step is to develop some rapport with the people in the area to be
studied so that they accept the investigational team. The issues
of dealing with the population, with government representatives at
all levels, and with commercial interests are tremendously complicated,
385
-------
After rapport with the public has been established, the investi-
gation should follow the usual epidemiologic steps. Characterize the
population, determine the kind of contaminant and estimate exposure
to it by area and by individual. Then correlate the data on the
population and the outcome with the exposure, over the short and
long term.
DR. MILLER: At a series of meetings about 10 years ago, the
Defense Atomic Support Agency wanted to know what actions to take
if there were a nuclear disaster. A psychiatrist who specialized
in disaster studies suggested that the best public relations activity,
which was also useful scientifically, was to take a census. Not only
is a census valuable for knowing who was exposed, it also gives the
people a sense of belonging to society again. It is important to
register the population and to get identifying information of a unique
sort for each person. In the United States, most adults have a social
security number that could be used as an identifier. This would
simplify followup studies on exposed population, no matter where an
individual goes.
DR. BRUZZI: People can be traced almost everywhere in Italy.
The demographic register in municipalities gives forwarding addresses
within the area so you can follow people even when they move. The place
of residence on the register sometimes does not correspond to the real
residence, but on the average it is a good estimator of the population
living in each municipality.
386
-------
An actual census, however, is Important too. In Seveso, in
August 1979, there was a census of Zone A and of some part of Zone B,
but none for Zone R. Nothing was known about the population of the
surrounding zones, in which some exposure also occurred. So in
August 1979 a census was taken—of the population present now and of
a cohort present at the time of exposure.
DR. MILLER: Then a census must count the current population and
also try to estimate who was there at the time of the accident and has
since left. What percentage of the people had left Seveso?
DR. BRUZZI: We still don't know because we did not match the
two populations. We estimate that about 5% of the population is
leaving each year. However, the percentage that emigrates is about
equal to that which comes in.
DR. HEATH: A census could not have been taken at Three Mile
Island right away. We had to wait for the people to return. In the
5-mile radius of the census, maybe 40% of the people had left at
some time during the 2 weeks after the accident.
We had to undertake the census quickly, as soon as people came
back because, at least in United States, there is a great deal of
mobility. In the 2 or 3 months that intervened between the accident
and the census at Three Mile Island, many people moved away, not just
because of the accident, but because of normal migration.
DR. MILLER: Did the actual mechanism, the procedure of taking a
census, have any effect on the population? Did it make the people
feel better?
387
-------
DR. HEATH: I suppose it did, particularly because we recruited
census takers from the population itself. The U.S. Census Bureau
helped us take the survey, but people from the local community, under
the supervision of state and federal personnel, were recruited to
interview residents.
DR. MILLER: Identifying the chemical involved in the exposure
can take time.
DR. PACCAGNELLA: Certainly, in any area-wide contamination, a
first step is to measure the level of contamination. Sometimes it
takes a long time, not only because the contaminant is not known, but
because there are methodological and analytic aspects to develop and
people expert in both analysis and methodology must be gathered.
This process takes a long time, sometimes weeks.
Defining the limits of a polluted area and determining the
levels of pollution, can take several weeks, as it did in one case
when well water was polluted by organic chemicals. In such situations,
there is a need to estimate exposure as closely as possible at the
beginning, while simultaneously dealing with emotional problems
within the population.
DR. MILLER: You have to know what chemical polluted the area
and also its source—such as the well water or the cooking oil. For
instance, in Japan, is it correct to say that cooking oil was polluted
with polychlorinated biphenyls (PCB's)? Wasn't it more than just
PCB's? You can be looking for effects from PCB's, but the effects
observed may not be due to PCB's if impurities are present. The
388
-------
suspect material must be fully analyzed to know what chemicals are
involved.
You need specimens of the chemical to analyze in order to learn
what people were actually exposed to. It may not be what you think.
Impurities may be more responsible than the compound for whatever
effects are observed. So to conduct a proper study, you need to
obtain samples of what actually contaminated the area.
DR. HEATH: Sometimes you have to do the best you can, as in
the Michigan polybrominated biphenyl (PBB) problem. The contamination
happened months before; to retrieve what actually went into the mills
and got mixed up in the feed was always a bit uncertain. Even today
there is concern that the batch of PBB available for analysis may not
have been representative of contamination. Contamination by furans
may have occurred and never fully analyzed. Nor can the analysis be
made at this date.
Perhaps the most difficult situations are those where people are
exposed to multiple chemicals, as at Love Canal or Pittston, Pennsylvania.
DR. PACCAGNELLA: Exposures can be divided into two kinds:
those caused by unusual contaminants and those by known contaminants.
In the latter situation, criteria and knowledge can be developed in
a short time, sometimes even before a disaster. When there is unusual
contamination, the situation is quite different and more difficult.
Uncertainties and the lack of knowledge impede planning for epidemio-
logic evaluation and protection.
389
-------
This lack of certainty and decision making heightens people's
emotions. In turn, the ensuing turmoil makes many aspects of
organization more difficult.
DR. MALTONI: The source of a calamity may not be known, or if
it is known, there may be no data on possible effects to monitor.
In Seveso, it was soon known that dioxin was the contaminant, and it
was not the first time human beings were exposed to it. Occupational
exposures to dioxin had occurred since 1948—in Germany in 1953 at BSF
and later at another factory, and in the United Kingdom. These exposures
were limited, but strong; yet, in 1976 when the Seveso accident occurred,
none of us knew what to do—how to sample, how to monitor, how to be-
have with respect to a compound with which there had already been
experience.
The need emerged, therefore, to have an archive of responses
to accidents with known toxic compounds. More than three decades
after the first accident in Germany, we do not have epidemiologic
data on chronic effects of dioxin.
DR. MILLER: One problem is that analytic methods are still
evolving, just as laboratory analyses are. General epidemiologic
principles, however, can be applied, whether the exposure is to
dioxin, to methylmercury, or to PBB.
There was literature on the subject of accidental exposures,
but knowledge of that literature was not instantly available in Milan.
You need to know where to go in a particular city to get information
quickly on a chemical that you may not even have known was being manu-
factured in the neighborhood. How can investigators know what they
390
-------
are dealing with? They have to begin by going to the library, unless
there is an international or national agency.
DR. MALTONI: In 1976, was public information on the human effects
of dioxin sufficiently available on an international basis to develop
a behavioral protocol for Seveso?
DR. DARDANONI: There was a delay of 8 days before investigators
identified the chemical. It was long known by the factory. Dr. Cavalarro
suggested dioxin, because he had studied the literature. The monitoring
plan, prepared in a very few weeks, was established on the basis of all
data available from the literature. The commission dealing with the
Seveso affair stated behavioral norms immediately after the accident.
The first people were evacuated very rapidly and the ban on consuming
food and animals was established immediately.
DR. MALTONI: I was thinking about the lack of information
generally, which could have been provided from knowledge of the
previous accident and reports in the international scientific
literature. These are two entirely different things. Because so
many accidents had happened before, much more should have been known
in 1976 at the international level. The scientific community had
not collected enough information.
DR. NELSON: Many of these accidents occurred in the privacy
of industry and were learned about only later. It was not until
1969 that pieces began to fit together during the Science Advisory
Board study of the Agent Orange used in Vietnam. That brought dioxin
into consideration of 2,4,5-trichlorophenol. After that, laboratory
391
-------
research accelerated markedly. The earlier episodes were then connected
retrospectively, but they were never synthesized. Certainly, by the
time of the Seveso accident, there was a vast literature, but it was
mostly constructed after the intense interest in Agent Orange. It
is a mistake to assume that in 1946 we knew about dioxin. We didn't.
It was only in 1969 that the pieces began to fit together.
DR. MOORE: In 1973, a bibliography of all the published scientific
data dealing with 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) was put together,
However, although those who worked on it knew of the book's existence,
it was not well known to the rest of the world.
Perhaps in this country we profited from an accident like Seveso.
We now have available a crisis response team to search literature.
Whatever the next area-wide pollutant may be, after it has been
identified a complete literature search and printout will be avail-
able within 24 hours from the Toxicology Information Response Center
(TIRC) at Oak Ridge. The existence of the TIRC, I suspect, as a
revelation to the majority of the people in this room. It is under
the general auspices of the National Library of Medicine, Department
of Health and Human Services.
DR. PACCAGNELLA: There is certainly a need for the literature
to be available to everybody. The International Register for Poten-
tial Toxic Compounds includes only 10,000-25,000 compounds, in contrast
to the 600,000 or 700,000 chemical compounds estimated to be in the
environment.
DR. MILLER: Is the Oak Ridge facility open to anybody in the
world?
392
-------
DR. MOORE: Yes, I assume so.
DR. MILLER: After you have taken the census, have identified
the exposed people, have identified the chemical, and have the
makings of a study, you then have to establish the body burden
or the dose. How do you do that?
DR. PACCAGNELLA: It depends on the nature of the chemical.
Methylmercury can be measured in blood and in hair. Other organic
compounds, such as pesticides, are concentrated in fat.
DR. MILLER: You have to be thinking about what samples to
collect, perhaps blood, urine, or fat.
DR. PACCAGNELLA: There is a need for criteria, and sometimes
in events of unusual contamination, they are lacking. It is im-
possible to know in advance what kinds of samples—urine, blood, or
hair—to collect. What about biopsies?
DR. MILLER: —And autopsies?
DR. PACCAGNELLA: It is difficult to ask people for a biopsy
although it is the correct procedure to follow when you know the
kinetics of the compounds. For unusual contaminants, how do you
collect biopsies?
DR. BRUZZI: It also depends on the size of the population to
be monitored. Blood samples or biopsy specimens are not likely to
be obtained from some hundred thousand or ten thousand people.
DR. MILLER: Get material from surgery performed for other
reasons and from autopsies. Get specimens from placental or fetal
tissue. People may not realize that the placenta is a source of
specimens for chemical analysis. Baby teeth and breast milk are
393
-------
other sources. It is important, at the outset of area-wide con-
tamination, to gather specimens appropriate to the chemical for
analysis.
DR. PACCAGNEL1A: First, you must distinguish between acute
or nonacute and usual or unusual contaminants. It is rather
difficult to get specimens after acute episodes or unusual contami-
nations. There may be some other criteria for usual and acute
contamination, but the samples mentioned here are certainly important.
DR. HEATH: Among the things to be considered are sampling of the
environment and sampling of chemicals both transiently and persistently
in the body. Environmental sampling is the other half of the dosimetry
question. If an exposure involves large numbers of people, there is
no way to collect specimens from all of them. Also, when radiation
exposure occurs, dosimetry has to be used to extrapolate the dose
from environmental samples. People can't wear film badges all the
time in case an accident occurs.
Often sampling is practical only with chemicals that can be
measured easily or without too much expense. For chemicals that
come one minute and disappear the next, like trichloroethylene,
sampling has to take place right at the time of exposure. It is
crucial then to know what specimens to get at the time of the
accident or soon afterwards. If the contaminants are chemicals
that persist for a person*s lifetime, specimens can be traced and
obtained later. To get specimens from people who die or from
aborted tissue at the time of the accident, investigators have to
be on top of the situation.
394
-------
DR. MILLER: How many thousands of years old were the exhumed
Peruvian Indians whose bodies provided specimens for lead content?
DR. GOLDBERG: Prehistoric.
DR. MILLER: Samples kept for other purposes can also provide
useful information; for instance, in Japan, umbilical cords are
traditionally preserved. At Yale University, blood specimens were
taken from all entering students. When some students developed
infectious mononucleosis, it was easy decades later to pair blood
serum to show that Epstein-Barr virus was involved in infectious
mononucleosis.
DR. HEATH: To evaluate radiation effects, it has been suggested
that old thyroid biopsies or thyroid surgery specimens, saved since
days of fallout over southern Utah, be studied.
DR. MILLER: In St. Louis in the 1950's, naturally shed children's
teeth were collected to measure fallout. Those teeth are probably
retrievable now.
So far now in our hypothetical contamination, the census has
been taken, the chemical identified, specimens obtained to analyze
the dose, environmental samples taken, the literature reviewed—
what next?
DR. HUTCHISON: We haven't identified the effect that we are
going to study. Ideally, if the literature is complete, we will
know the effects to look for.
Monitoring for acute early effects and long-term effects is
totally different. In all likelihood, the long-term effects will be
395
-------
the less well known, and the investigation may turn to identifying
etiology, that is, cause-effect relationships. Short-term relation-
ships, in many instances, are already known. But the magnitude of
the episode and the persons affected by it must be documented.
DR. MILLER: Let's say we know that people are suffering from
severe neurologic disability. How do you find the cases as they
develop in a community? Hospitals, practitioners, insurance systems?
Death certificates to retrieve information on patients who may have
escaped your net?
Look at the system used by the Atomic Bomb Casualty Commission,
for example—tissue and tumor registries, hospitals, and patient
examinations.
How do you identify patients with specific effects from an
exposure? What is being done in Seveso?
DR. BRUZZI: Epidemiologic studies have to be carefully designed.
Fishing expeditions have not proved very useful. In Seveso we are
establishing two approaches for cancer: first, a mortality study
(it is still too soon to get results) and second, an incidence study
using a register based on mortality data and hospital records.
Cancer registries, of course, are expensive and difficult to
establish. Fortunately, the region is computerizing all admission-
discharge information from hospital charts for administrative reasons.
The quality of these data is not very good, but it may be useful
for getting an idea about the incidence of cancer in the area.
396
-------
DR. MOORE: It is one thing to collect data to amass a greater
base of knowledge for science and mankind, but how do we handle the
concerns of the exposed people? A tumor registry isn't really going
to help me if I am one of the affected people. People have symptoms
that they experience themselves, that they have read about, or somebody
has told them might occur. Such symptoms tend to be subjective
complaints: "I'm, tired," "My memory isn't as good as it used to
be," "My muscles hurt." Or the population might experience a slight
increase in fetal resorption, an effect difficult to measure. We
aren't doing a very good job of answering the immediate complaints.
DR. HEATH: This is a key problem. Such illnesses—and they
are generally the acute ones, the ones that people are worried about
immediately—are not going to be found in hospitals. The only approach
to take is to examine people and question them about symptoms. It is
a very difficult process, and as an epidemiologist I would resist that
approach. The medical investigator who has to respond immediately
to an exposure cannot say publicly he is not going to give examinations
but he might say privately that is his last resort. The trouble
with questionnaire surveys and asking people about symptoms is that
people perceive their health in relation to what they think their
exposure might have been. Distortion occurs. Too often, the findings
cannot be documented.
I can't say that in either the Love Canal or Michigan situations the
massive physical examinations to find illness possibly related to those
exposures have been successful. They have put to rest some of the questions
that were raised, but have let others linger. They have opened some
397
-------
difficult situations in Michigan. Perhaps the most difficult
concern immunologic problems.
To date, questionnaires have not proved the presence of any
abnormality in Michigan or at Love Canal. The one abnormality
found at the canal was revealed by using objective records. The
investigators had to look at the birth weight on birth certificates;
the records showed an apparent increase in the number of children
who weighed less than 2,500 g at birth.
DR. MALTONI: The method of monitoring effects is also
influenced by the types of diseases sought. One of the most direct
and correct methods of cancer followup is epidemiologic evaluation
of death certificates. But this takes time, and does not answer
people's immediate concerns. For some types of cancer, monitoring
the risk is possible; but is not possible for liver cancer. For
exposure to vinyl chloride, the use of all possible tests of liver
function did not effectively monitor the risk to the population.
Certain procedures, such as sputum cytology, allow for correct
monitoring. An example of this is provided by studies of uranium
miners in the United States. We have learned to monitor, when
possible, for precancerous lesions from the few experiences through-
out the world of carcinogenic exposure to chemicals.
Of course, it does not help a person very much to say, "You have
such a precursor," because we may not know what to do about it. But
we may help science by comparing two populations, exposed and not
exposed. Several years later, a mass survey by cytology of sputum
398
-------
or urine will show if an excess of dysplasia exists in the exposed
group. This test provides some information about the risk of the
exposed group. After experimental data showed that dioxin produced
lung tumors in highly resistant animals, such as the rat, the time had
come for cytologic monitoring of lung and urine.
DR. MOORE: We need to admit that we do not always do epidemio-
logic studies well. These studies raise the population's expectations.
If we can't deliver, the outome is sometimes worse than doing nothing
at all. We should identify those methods that are imprecise and work
to improve them. For example, we need better methods to assess a
slight decrement in reproductive performance in females in Wyoming.
The most helpful records may be those we collect historically, such
as birth certificates.
DR. MILLER: Who should respond in this country to area-wide
contamination?
DR. HEATH: Area-wide problems are essentially local problems,
and the immediate and long-term responsibility rests with local
health departments. If the situation is serious enough, it draws in
larger bodies such as public health groups and private groups. If
your question is who in the Federal group should have that responsi-
bility for the government, the answer is debatable.
DR. MILLER: What happened in the investigation of asbestos
exposure in Globe, Arizona.
DR. HEATH: The local people quickly went from step A to step
B to step C, and then the Federal people came in.
399
-------
DR. PACCAGNELLA: In Italy, it is more or less the same.
The local authority first has full responsibility. But the public
health officers in the villages are really responsible for general
public health matters, not for special episodes. They have no
technical equipment. They have no means to control environment,
to control pollution, to do the analysis. From the legal point of
view, however, they are responsible first.
DR. HEATH: Local authorities need to know at what point they
should call for assistance from the next higher echelon, and, in
the United States, the next step up is the state. Local groups should
not hesitate to call for help. A cooperative effort should develop
among all levels, but the local one has the continuing responsibility.
It is there after the State and Federal Governments go home.
DR. MALTONI: The greatest responsibility lies not with the
official there when an accident happens, but with the person (or
group) who originally allowed the danger, especially if that official
did not assess the danger. In Seveso, the public health officer
there when the factory was established was not the one there when
the accident occurred. It was known, by the way, that Seveso was
the third choice for the factory. Two other places refused to have it.
Under a new law, the mayor of a town is responsible for everything.
Under the previous law, both the mayor and the public health officer
were equally responsible.
DR. PACCAGNELLA: Yes, but the final decisions are not in the
hands of the health authorities. Even before, the final decisions
400
-------
were in the hands of the community through their representatives.
If the health authority could limit or stop development, the official
would be too powerful. This is true everywhere.
DR. BRUZZI: Cooperation with local services is very important.
One of the main failures in Seveso was a break between local health
services and the special bureau. The population lost faith in the
special bureau because there was no collaboration, physicians were
ignored, and there was a gap between these two organizations.
DR. MILLER: Should the local health officer do very sophisti-
cated studies?
DR. HEATH: No study of that sort will be made at the local
level. There are exposure incidents where sophisticated studies
are not needed, and some practical public health control measures
are needed.
If you wanted to conduct sophisticated studies of all accidents,
you would have to reorder your priorities. Trichloroethylene
exposures in this country, for instance, are a dime a dozen; you
would have to tell the local EPA people at the State and county
levels not to become involved. You would call on a central Federal
person or agency to plug the local study into a large protocol.
That would be quite a switch in priorities. Problems that need
sophisticated study go through a substantial filtering system in
the United States until they reach the top. Now, some that are
important get suppressed, not intentionally, but often they are not
detected early. We should lubricate the system so it works better.
401
-------
DR. MILLER: How is TIRC at Oak Ridge being advertised? Do
local health officers know it exists?
DR. HEATH: It was mostly set up for agencies at the federal
level.
DR. MOORE: Since it was established it has never been called
on, perhaps because nobody knows it is there or because no one
thinks a specific crisis is big enough. It is a chemical crisis
response group; it is not concerned with radiation or infectious
diseases.
DR. MILLER: The Department of Transportation listed 603 toxic
chemical spills reported to it in 1978. What was the response to
those spills?
DR. HEATH: A consortium of federal agencies, called the National
Response Team, was set up mainly to handle oil spills. Many of
these spills were probably oil spills, and were handled through that
team in the initial emergency phase. When the Pittston Mine was
found draining chemicals into the Susquehanna River, the National
Response Team was called to implement emergency measures. The team
works with its state counterparts and deals with the immediate
environmental protection aspects of a problem, not necessarily the
health aspects. It works on cleaning up the oil, and sequestering
the chemicals. It has a fairly practical system. The team currently
is under EPA, but often works with the Coast Guard because the team
focuses on spills in water.
402
-------
The National Response Team is concerned not only with water. The
Crestview train derailment in northern Florida, which caused
considerably more toxic exposure of people than did Three Mile Island,
was handled by the southeast regional National Response Team. The team
also acted at a fire in a warehouse where toxic chemicals had been
stored.
MS. CONWAY: The people within EPA who handle oil spills also
handle hazardous waste problems. The Resource Recovery and
Conservation Act covers hazardous materials in transportation and
disposal. There is a so-called "cradle-to-grave" manifest system for
hazardous materials.
DR. IREY: When does the TIRC in Oak Ridge act?
DR. MOORE: It answeres queries, such as, "We had a spill of hexa-
clorobenzene. We know the name of the chemical; we aren't familiar
with what it can do or what previous accidents have shown us, or what
is known about it." Within a short period, Oak Ridge will produce a
computer search printout of all of the pertinent information in its
files. This group does not send anyone out to the field.
DR. PACCAGNELLA: During the last few decades when several
epidemiologic studies were conducted on usual environmental pollution
(air pollution and respirable disease, pesticides in rural areas, human
reactions to traffic or community noise) residents and local health
services generally cooperated.
But implementing epidemiologic programs in Italy has been difficult
in acute and unusual episodes. In emotional situations, the people in
charge of the task forces dealing with the problem lack credibility.
403
-------
Frequently, sometimes successful attempts have been made during
the last 10 years to establish credibility. The people's cooperation
has been achieved by asking for international cooperation. It is
sort of an international validation of our programs. How do you
gain the population's credibility in the United States? Do the people
believe in your service? Do they oppose it? Do they believe that
you are not objective?
DR. HEATH: When officials are perceived as not reacting promptly,
they lose credibility. If they are on the scene doing something,
even if it is not much, their activity helps maintain credibility.
If they can bring the people to understand that not much can be done,
that is ideal. In so many situations, there isn't much that can be
done, but you must respond to the public's concern.
DR. PACCAGNELLA: Do the people believe in your service?
DR. HEATH: Well, I hope they do, but we don't always succeed.
If people realize the lack of activity is for technical reasons,
scientific gaps, or whatever, they will understand the delay. But
often a delay is perceived as bureaucratic obfuscation. People think
officials don't know what they are doing, or are not doing their jobs.
DR. PACCAGNELLA: Does the public think a task force tries
to avoid responsibility?
DR. HEATH: It is a matter of communication and rapid response.
Effective rapid response can sometimes be achieved by communicating
and by staying at your desk. We often do that. If we can deal with
a problem on the phone, that is an adequate response; we don't have
to do more. But if a letter sits on our desk for months and gets
passed back and forth to different people, we lose credibility.
404
-------
Drs. Miller and Dardanoni thanked the Panel for their perceptive
comments and lucid exchange. Both agreed that this novel analysis
of past experiences was a substantial contribution to help guide future
investigations of area-wide chemical contamination.
405
-------
Summary and Conclusions
A workshop was held to explore the steps that should be undertaken to
investigate the impact on human health of area-wide chemical contamination.
Two approaches were used to explore the topic. On the one hand, case
studies of known exposures to toxicants were developed in relation to
chemical manufacture. Correspondingly, case studies of types of adverse
effects (e.g., reproductive injury, cancer, neurologic effects,
genotoxicity. biochemical toxicity, and liver injury) were studied to
indicate the complexity of the disease characteristics in relation to
various causative substances.
Several studies of acute exposures of humans to substances widely
distributed in the environment were identified as likely candidates to
provide some insight into investigative approaches and principles
demonstrated to be particularly beneficial in establishing cause-effect
relationships. The case studies included TCDD ("dioxin"), Yusho (PCBs),
Kepone, dibromochloropropane, lead, methyl mercury, chlorinated hydrocarbon
mixtures at the Love Canal, polybrominated biphenyls, and acute radiation
exposure from the atomic bomb. The case studies exemplified a wide
diversity of exposures and of adverse effects, emphasized the complexities
of obtaining reliable exposure data, and demonstrated a substantial range of
success in delineating cause-effect relationships. A strong basic science
data base coupled with creative approaches and carefully controlled study
conditions were elements essential to reaching definitive conclusions about
impacts on human health.
The second approach focused on various adverse effects, some of which
were associated with specific organs (e.g., reproductive toxicity) while
406
-------
others were not (e.g., cancer). This approach elucidated some of the more
serious difficulties of finding causes for diseases which are relatively
common in an environment where exposures to chemical and physical agents are
highly diverse and complex. A variety of study characteristics were
identified to enhance the value of such experimental undertakings.
Through the use of the case studies, several general principles emerged:
1. An effective and credible rapport must be established between
scientists and the public before any epidemiologic investigation
can proceed with its conventional approaches.
2. A necessary first step to the study of an exposed population is a
comprehensive census of individuals in the geographic area.
3. To the extent feasible, the contaminant(s) should be identified and
its (their) concentraton in the media of human exposure should be
*
documented.
4. If the contaminant(s) is known, literature describing its (their)
toxic properties should be reviewed to obtain possible pathologic
and biochemical indexes that could be useful leads in epidemiologic
studies.
5. In the early phases of investigation, the study should include the
analysis of body burdens in order to more accurately assess
exposure.
6. Throughout the investigation, humanitarian concern for the exposed
population should be preeminent and should dictate a sensitivity in
dealing with these individuals such that their anxiety about
possible adverse health consequences is kept at a minimum.
7. There was agreement that there should be at least one organization
given the responsibility worldwide to investigate the health impact
407
-------
of large scale chemical contamination; however, no existing
agencies were identified as the likely candidates for such a
function. The complementary roles of local and national
authorities were identified as a subject that needed sharper
definition in order to enhance the value of such studies.
An important element in the success of such epidemiologic studies
is credibility of the investigators and the authorities they
represent. Factors influencing credibility include rapidity of
response, level of activity, and communications about progress and
results. A strong sense of public concern was recommended to
tailor interactions in the community and to establish the
credibility required to conduct a study whose results have a
greater probability of being accepted by the public.
408
-------
Appendix 1*
Leonardo Santi: Epidemiclogic Monitoring in an Episode of
Environmental Chemical Pollution: Problems
and Programs in the Seveso Experience
When area-wide chemical contamination occurs, particular care
must be given to the conciliation of emergency activities with long-
term epidemiologic monitoring programs. Immediately after an accident,
when surveillance and health care programs are to be set in motion
as quickly as possible, programs intended to provide useful epidemio-
logic data over a longer period may be overlooked. Therefore, emergency
activities must be fitted into a general epidemiologic plan with
mid-term and long-term monitoring activities.
1. Ad hoc intervention programs must be coordinated by a
team of epidemiologists.
2. Ad hoc activities must be performed, in tandem with health
operators and services already present in the area, in order to ensure
an early involvement of those who will progressively have to take on
the management of the long-term program.
An intervention program that does not account for services
already operating in the area is likely to face serious problems
later when trying to followup monitored populations.
Furthermore, the basic activities essential to many subsequent
epidemiologic studies must be started as soon as possible in order
to avoid losing information that is difficult to gather later. Of
*This paper was not presented at the workshop but was submitted
later.
409
-------
these activities, a demographic register is a necessary tool for
operative purposes, as well as for identifying and following up
subgroups of the whole population at different levels of exposure.
Another problem that has to be taken care of soon after area-
wide contamination is the evaluation, on an individual basis, if
possible, of the extent of exposure to the chemicals. At Seveso,
emergency problems are over, and long-term monitoring programs
have been established. The Epidemiology Department of the Istituto
Scientifico per lo Studio e la Cura dei Tumori di Geneva was asked
to prepare a general plan of epidemiologic research for Seveso.
Personnel decided to involve the public health services of the area
in the monitoring program and established cooperative ties with the
Regione Lombardia's Epidemiological Service and with local health
agencies. Now, more than 3 years after the accident, it does not
appear fruitful to allocate efforts and resources in ambitious but
demanding clinical research protocols involving large population
groups. Current research is based on three main criteria: feasibility,
clinical and/or social relevance, and association with the polluting
agent 2, 3, 7,8-tetrachlorodibenzo-_p-dioxin (TCDD).
The first problem was to define the evaluation of exposure and,
at the same time, to select suitable control groups, which are essential for
epidemiologic studies. Because reliable indicators of dioxin
exposure are not available (even chloracne, which is a quite specific
indicator, does not seem to be very sensitive), and information about
individual exposure was scanty and of poor quality, it was decided to
410
-------
use place of residence at the moment of the accident and during the
following periods for the stratification of the population into
subgroups having received different levels of TCDD exposure. Due to
the organizational and administrative impossibility of selecting and
following suitable control groups outside the area, it became necessary
to identify areas inside the monitored zone that can be regarded as
free of TCDD contamination. This judgment was based on data concern-
ing the amount of TCDD in soil, animal casualties, and acute dermal
lesions or chloracne in individuals in every area. The population
of the TCDD-free areas serves as control groups for the large epidemic-
logic studies.
The following groups of individuals, however, are classified as
exposed, independent of their place of residence:
1. individuals with acute or chronic skin lesions attributable
to the toxic cloud or to TCDD;
2. occupationally exposed workers; and
3. individuals in highly polluted zones at the time of the
accident.
These groups are to be followed with specific protocols, along
with suitable control groups.
Some basic studies were given top priority because they were
expected to provide general and essential information and to serve
as preliminary steps for more specific research. These included:
1. a mortality study, both on the current population of the
area and on the cohort present in July 1976;
411
-------
2. a cancer incidence study by means of a cancer registry; and
3. a study of the frequency of birth defects in newborns from
1976 on and of abortions performed in the area.
These programs are being performed now, along with detailed followup
protocols for different subgroups at defined levels of exposure.
This program, admittedly smaller than that planned earlier, is
still comprehensive and seems feasible in all its parts. It will
not provide final answers to all questions raised by the ICMESA
accident, but will generate reliable data regarding most of the
major problems.
412
-------
Appendix 2
' ' •"^•''••t'-W1;"i~3?**1i^''^'
•* *&* A* Sli?"?*--!
t'fi'f i-^siKHiaBtsHBSif:-.«
;ii ••^
and Epide
*i^'li*^"-'»o;>;
Follo>v-up
m^!,:^xSSSf,^^m^ j< Ii
Prof. Gaetanb Maria Fara^
f NATIONAL ACADEMY OF SCIENCES
jfe^BdjJMuin'^
Washington; D.C;^:-.^.- '?li^^^
413
-------
Sponsored by
COMMITTEE ON RESPONSE STRATEGIES
TO UNUSUAL CHEMICAL HAZARDS
BOARD ON TOXICOLOGY AND
ENVIRONMENTAL HEALTH HAZARDS
Assembly of Dfe Sciences
National Research Council
National Academy of Sciences
PURPOSE
Through the review of various episodes of area-wide
contamination by chemical and physical agents, this
workshop will seek general principles and formulate
plans for the investigation of wide-spread exposures
and their impacts on human health.
COMMITTEE ON RESPONSE STRATEGIES
TO UNUSUAL CHEMICAL HAZARDS
Dr. Norton Nelson, Chairman, Institute of Environ-
mental Medicine, New York University MedicaJ
Center, New York, N.Y. 10016
Dr. A. L. Burlingame, Mass Spectrometry Research
Resource, University of California, Berkeley,
Calif. 94720
Dr. Aaron B. Lerner, Department of Dermatology,
Yale University, New Haven, Conn. 06510
Dr. Robert Miller, Vice-Chairman, Epidemiologic
Branch, National Cancer Institute, Bethesda,
Md. 20205
Dr. John A. Moore, Research Resources Program, Na-
tional Institute of Environmental Health Sciences,
Research Triangle Park, N.C. 27709
Dr. Sheldon D. Murphy, Department of Pharma-
cology, University of Texas, Houston, Tx. 77025
Dr. Robert A. Neal, Department of Biochemistry,
Vanderbih University, Nashville, Tenn. 37203
Dr. Milos Novotny, Department of Chemistry, Indiana
University, Bloomington, Ind. 47401
Dr. Patrick O'Keefe, New York State Department of
Health, Albany, N.Y. 12201
Dr. Alan Poland, Department of Oncology, University
of Wisconsin, Madison, Wis. 53706
PROGRAM
MONDAY, MARCH 17,1980
8:00 REGISTRATION AND COFFEE
9:00 WELCOME
Dr. Philip Handler,President
National Academy of Sciences
INTRODUCTION
Dr. Norton Nelson
Prof. Francesco Pocchiari
I. CASE STUDIES OF SELECTED AREA-
WIDE ENVIRONMENTAL EXPOSURES
A brief background description of each occurrence
will be presented with emphasis on the source(s)
of contamination, nature and degree of exposure,
population exposed, and other pertinent exposure
assessment parameters.
9:30 TCDD (Italy)
Prof. Gaetano Fara
9:45 TCDD (U.S.)
Dr. Raymond Suskind
10:00 Yusho Disease (Japan)
Dr. Robert Miller
10:15 COFFEE
10:30 Kepone(U.S.)
Dr. Philip Guzelian
10:45 DBCP(U.S.)
Dr. Donald Whorton
11:00 Lead (U.S.)
Dr. Philip Landrigan
11:15 Methyl Mercury (Japan)
Dr. Robert Miller
11:30 Chlorinated Hydrocarbons (U.S. Lo\c Canal)
Dr. David Axelrod
12:00 LUNCH
1:30
1:45
PBB's (U.S.)
Dr. Philip Landrigan
ABCC* (Japan)
Dr. Gilbert Beebe
•Atomic Bomb Casualty Commission
414
-------
II. ADVERSE EFFECTS ON TARGET SITES
REPRODUCTIVE INJURY
2:00 Dr. David Axelrod
2:40 Prof. Gianni Remotti
3:20 COFFEE
3:30 Dr. Robert Miller
4:10 Dr. Helga Rehder
TUESDAY, MARCH 18, 1980
DERMATOLOG1C EFFECTS
9:00 Dr. Aaron Lerner
9:40 Prof. Vittorio Puccinelli
10:20 COFFEE
CLINICAL EFFECTS OF EXPOSURES IN U.S. TO
2,4,5-T AND ITS CONTAMINANTS
10:30 Dr. Raymond Suskind
CARCINOGENIC EFFECTS
11:10 Human Observations
Dr. Clark Heath
11:50 Experimental Studies
Prof. Guiseppe Delia Porta
12:30 LUNCH
1:30 Experimental Studies
Dr. John Moore
NEUROLOGICAL AND BEHAVIORAL EFFECTS
2:10 Dr. Alan Goldberg
2:50 Prof. Guglielmo Scarlato
3:30 COFFEE
IMMUNOLOGICAL EFFECTS
3:45 Dr. John Moore
4:25 Prof. Girolamo Sirchia
WEDNESDAY, MARCH 19, 1980
SOMATIC CELL MUTATIONS
9:00 Dr. Arthur Bloom
9:40 Prof. Luigi De Carli
10:20 COFFEE
LIVER INJURY AND OTHER TOXIC EFFECTS
10:30 Dr. David Axelrod
11:10 Prof. Carlo Zanussi
12:00 LUNCH
1:30 Dr. Robert Neal
2:10 Prof. Silvio Garattini
2:50 Dr. Sheldon Murphy
3:30 COFFEE
0
r
ADVANCE REGISTRATION
There is no charge for registration. Registering in
advance is required.
The registration form and any inquiries concerning
the program should be addressed to:
Dr. Robert G. Tardiff
Executive Director
Board on Toxicology and Environmental
Health Hazards
National Academy of Sciences
2101 Constitution Avenue
Washington, D.C. 20418
(202) 389-6914
A'jwr
Address
Cf/v
Stair-
Zip.
Telephone.
415
-------
3:45
III.
PANEL ON EPIDEMIOLOGIC
APPROACHES TO MEASUREMENT
AND ASSESSMENT OF EXPOSURES
Prof. Luigi Dardanoni
Dr. Clark Heath
Dr. George Hutchison
Dr. Robert Miller, Moderator
Prof. Bruno Paccagnella
Prof. Alfredo Zampieri
4:45
IV.
SUMMARY AND PROJECTIONS
FOR THE FUTURE
Dr. Robert Miller
Prof. Gaetano Fara
GENERAL INFORMATION
Registration and Luncheons
Badges will be issued to all registrants at registration
desks at the C Street entrance of the National Acad-
emy of Sciences between 8:00 and 9:00 a.m.
Tickets for luncheons will be provided for participants
of the workshop on March 17-19 and sold to all others
at a cost of $4.50 per day.
Message Center
Incoming calls to meeting attendees should be di-
rected to (202) 389-6821. Messages received will be
placed on the bulletin board at the entrance to the
NAS Auditorium.
Parking
The Academy does not have parking facilities for
participants or attendees. However, if you plan to
drive, some side street parking is usually available.
PARTICIPANTS OF THE INTERNATIONAL
WORKSHOP ON PLANS FOR CLINICAL
AND EPIDEMIOLOGIC FOLLOW-UP AFTER
AREA-WIDE CHEMICAL CONTAMINATION
DR. DAVID AXELROD, Commissioner of Health,
New York State Department of Health, Nelson
A. Rockefeller Empire State Plaza, Albany,
N.Y.12237
DR. GILBERT BEEBE, Clinical Epidemiology
Branch, National Cancer Institute, Bethesda,
Md. 20205
DR. ARTHUR BLOOM, Department of Pediatrics,
Columbia University, New York, N.Y. 10032
PROF LUIGI DARDANONI, Istituto di Igiene, Uni-
versita di Palermo, 90100 Palermo, Italy
PROF. LUIGI DE CARLI, Istituto Genetica, Facolla
Scienza, Via S. Epifanio 14, 27100 Pavia, Italy
PROF. GIUSEPPE DELLA PORTA, Istituto per lo
Studio dei Tumori, Via Venziani, 20100 Milano,
Italy
PROF GAETANO MARIA FARA, Istituto di Igiene,
Universita di Milano, 20100 Milano, Italy
PROF. SILVIO GARATTINI, Istituto Rjcerche
Farmacologiche "Mario Negri," Via Eritre A 62,
20157 Milano, Italy
DR. ALAN GOLDBERG, Department of Environ-
mental Health, Johns Hopkins University,
Baltimore, Md. 21025
DR. PHILIP GUZELIAJM, Medical College of Virginia,
Box 265, Richmond, Va. 23298
DR. CLARK HEATH, Chronic Diseases Division,
Bureau of Epidemiology, Center for Disease
Control, Atlanta, Ga. 30333
DR. GEORGE HUTCHISON, Department of Epi-
demiology, Harvard School of Public Health,
Boston, Mass. 02115
DR. PHILIP LANDRIGAN, NIOSH/DSHEFS/MSF-],
Robert Taft Laboratories, Cincinnati, Ohio 45226
DR. AARON LERNER, Department of Dermatology,
Yale University, New Haven, Conn. 06510
DR. ROBERT MILLER, Epidemiologic Branch, Na-
tional Cancer Institute, Bethesda, Md. 20205
DR. JOHN MOORE, Research Resources Program,
National Institute of Environmental Health Sci-
ences, Research Triangle Park, N.C. 27709
DR. SHELDON D. MURPHY, Department of Pharma-
cology, University of Texas, Houston, Tx. 77025
DR. ROBERT NEAL, Department of Biochemistry.
Vanderbilt University, Nashville, Tenn. 37203
416
-------
DR. NORTON NELSON, Institute of Environmental
Medicine, New York University Medical Center,
550 First Street, New York, N.Y. 10016
PROF BRUNO PACCAGNELLA, Istituto di Igiene,
Universita di Padova, 35100 Padova, Italy
PROF. FRANCESCO POCCHIARI, Direttore, Istituto
Superiore di Sanita, Viale Regina Elena 299, 00161
Roma, Italy
PROF. V1TTORIO PUCC1NELLI, Clinica Dermato-
logica, Universita di Milano, 20100 Milano, Italy
DR. HELGA REHDER, Institut fur Pathologie Medi-
zinischen Hochschule Lubeck, Ratzeburger Alice
160, D-24, Lubeck, Lubeck, West Germany
PROF. GIANNI REMOTTI.Prima Clinica Osterica,
Universita di Milano, Via della Commenda 12,
20110 Milano, Italy
PROF. GUGLIELMO SCARLATO, Istituto Neuro-
logia, Universita di Milano, Via. A. Sforza 35,
20100 Milano, Italy
PROF. V1TTORIO SILANO, Laboratorio di Tossi-
cologia, Istituto Superiore di Sanita, Viale Regina
Elena 299,00161 Roma, Italy
PROF. GIROLAMO SIRCHIA, Ospedale Maggiore
Policlinco di Milano, Via F. Sforza 35, 20122
Milano, Italy
DR. RAYMOND SUSKIND, Kettering Laboratories,
University of Cincinnati, Cincinnati, Ohio 45219
DR. DONALD WHORTON, Labor Occupational
Health Program, University of California, Berkeley,
Calif. 94720
PROF. ALFREDO ZAMPIERI, Laboratorio di Epi-
dcmiologia et Biostatislica, Istituto Superiore di
Sanita, Viale Regina Elena 299,00161 Roma,
Italy
PROF. CARLO ZANUSSI, Clinica Medica. Quarta,
Padiglione Litta, Via Sforza 35, 20122 Milano,
Italy
417
------- |