Artículo en PDF
How to cite
Complete issue
More information about this article
Journal's homepage in redalyc.org
Sistema de Información Científica
Red de Revistas Científicas de América Latina y el Caribe, España y Portugal
Rev. Int. Contam. Ambient. 21 (3) 133­142, 2005
DDT’S, HCH AND HCB LEVELS IN BREAST ADIPOSE TISSUE IN
WOMEN WITH BREAST TUMORS
Stefan M. WALISZEWSKI
1
, Marco T. BERMUDEZ
2
, Carmen S. SILVA
1
, Rosa M. INFANZON
1
,
Octavio CARVAJAL
1
, Sandra GOMEZARROYO
2
, Rafael VILLALOBOS PIETRINI
2
,
Patricia TRUJILLO
1
, Vicente SALDAÑA
3
, Guadalupe MELO
1
, Sergio ESQUIVEL
3
, Felipe CASTRO
3
,
Héctor OCAMPO
3
, J. TORRES
3
and Patricia M. HAYWARD­JONES
4
1
Institute of Forensic Medicine, University of Veracruz, 94290 Boca del Río, Ver. México
2
Center of Atmospheric Sciences, UNAM, México
3
The Mexican Institute of Social Security (IMSS Hospital), Veracruz
4
Chemical­Biological Area University of Veracruz
(Recibidomayo2005,aceptadoagosto2005)
Key words: pesticides, breast tumors
ABSTRACT
Organochlorine pesticides have been used in Mexico in agriculture, sanitation, malaria
programs and livestock to combat ectoparasites. Due to their chemical stability, persis­
tence and lipophility, their residues bioconcentrate in lipid­rich tissues, according to the
equilibrium pattern ofinternal transport and lipid tissuecontent. Thestudies on their role
as estrogenic or antiandrogenic agents have related these pesticides to breast cancer.
Therefore, organochlorine pesticides were determined in breast adipose tissue ofwomen
subjected to surgerydue to the diagnosis of breast abnormalities. The study pool consti­
tuted127 women with malignant breast tumors, 127with benign breast tumors and 127in
a control group without breast abnormalities. The pesticides determined were: HCB,
b
­
HCH, pp’DDE, op’DDT, pp’DDT and
S
­DDT. Comparingthestudied groups, thehighest
residue levels were determined in the benign breast tumor group. The determined levels
from the control group through malignant to benign cases increased stepwise; HCB:
0.045,0.099,0.116mgkg
­1
,
b
­HCH:0.163,0.265,0.319mgkg
­1
,pp’DDE:0.782,0.980,1.761
mgkg
­1
,op’DDT:0.035,0.094, 0.176mgkg
­1
,pp’DDT: 0.296,0.351,0.661mgkg
­1
,
S
­DDT:
1.112, 1.423, 2.601 mg kg
­1
, respectively. Paired analyses of the three groups revealed
divergences among them and correlation of these pesticide levels with breast tumors.
Calculated relative risk (RR) values for the benign to control group and the malignant to
controlcasespresentedthefollowingvalues: HCB:2.11, 2.01,
b
­HCH:1.96, 1.58,pp’DDE:
2.13, 1.17, op’DDT: 4.42, 2.27, pp’DDT: 2.33, 1.33,
S
­DDT: 2.21, 1.19 respectively. The
obtained results indicate higher organochlorine pesticide residue levels in women with
benign breast abnormalities and higher relative risk related to op’DDT presence.
Palabras clave: plaguicidas, tumores de seno
RESUMEN
En México, los plaguicidas organoclorados se ha utilizado en agricultura, salud pública
en programas del combate a la malaria ya ectoparásitos de ganado. Debido a su estabi­
S.M. Waliszewski
etal.
134
lidad química, persistencia y lipofilidad, sus residuos se biomagnifican en tejidos ricos
en grasa de acuerdo con el estado de equilibrio entre el transporte interno y el conteni­
do de lípidos en tejidos. Los estudios sobre su papel como agentes estrogénicos o
antiandrogénicos relacionan a los plaguicidas organoclorados con el cáncer mamario.
Por esta razón se determinaron los niveles de plaguicidas organoclorados en mujeres
sujetas a cirugía para diagnosticar anomalías mamarias. En el estudio participaron 127
pacientes con tumor maligno de seno, 127 pacientes con tumor benigno de seno y 127
pacientes como testigos sin anomalías mamarias. Los plaguicidas estudiados fueron:
HCB,
b
­HCH, pp’DDE, op’DDT, pp’DDT y
S
­DDT. Comparando con el grupotestigo,
el nivel superior se determinó en las pacientes con tumor benigno de mama. Los nive­
les de plaguicidas organoclorados desde el grupo testigo al grupo de casos malignos y
benignos se incrementó paulatinamente: HCB: 0.045, 0.099, 0.116 mg kg
­1
,
b
­HCH:
0.163, 0.265, 0.319 mg kg
­1
, pp’DDE: 0.782, 0.980, 1.761 mg kg
­1
, op’DDT: 0.035,
0.094, 0.176 mg kg
­1
, pp’DDT: 0.296, 0.351, 0.661 mg kg
­1
,
S
­DDT: 1.112, 1.423,
2.601 mg kg
­1
, respectivamente. El análisis pareado de medias reveló divergencia entre
los tres grupos y correlación de los niveles de plaguicidas organoclorados en el tumor
mamario benigno. Al calcular los valores del riesgo relativo (RR) para los grupos con
tumor mamario benigno y maligno, se obtuvieron los valores siguientes: HCB: 2.11,
2.01,
b
­HCH: 1.96, 1.58, pp’DDE: 2.13, 1.17, op’DDT: 4.42, 2.27, pp’DDT: 2.33, 1.33,
S
­DDT: 2.21, 1.19, respectivamente. Los resultados obtenidos indican mayores con­
centraciones de plaguicidas organoclorados en mujeres con anomalías benignas de seno
y un riesgo relativo mayor relacionado con la presencia del op’DDT.
INTRODUCTION
Persistent organochlorine pesticides (HCB­
hexachlorobenzene, HCH’s­ hexachlorocyclohexanes
and DDT­dichlorodiphenilethane) have been used
extensively in plant protection and sanitary actions
throughout the world (Tomlin 2000). Their use per­
mitted the protection of agriculture and the assur­
ance of harvests. In sanitation, they were applied to
combat the vectors that transmit malaria and typhoid,
as well as ectoparasites in humans and livestock. In
the 1970’s, their use in agriculture for plant protec­
tion was discontinued and recently, in Mexico, since
1999, in sanitation to combat malaria. The common
characteristics of these pesticides are persistence in
the environment, resistance to metabolic degrada­
tion, lipophilityand bioconcentrationinthelipidicphase
of organisms. Thus, their residues persist in the envi­
ronment: soil, air, plants, and the human food chain
(Waliszewski
etal.
2003a,b), reaching elevated lev­
els in humans (Waliszewski
etal.
2003c). That is
why the determination of the degree of body con­
tamination serves as an indicator of exposure and
permits to know the correlation between exposure
rate and adverse health effects (Wolff 1995, Wolff
andToniolo 1995, Wolff andWeston 1997, Safe 2000,
Snedeker 2001, Zou and Matsumura 2003).
The studies on the possible role of estrogenic or
antiandrogenic pesticides in promoting breast can­
cer have indicated a positive relationship (Robinson
and Stancel 1982, Bulger and Kupfer 1983, Bustos
etal.
1988, Kelce
etal.
1995, Golden
etal.
1998,
Woolcott
etal.
2001, Calle
etal.
2002, Starek 2003).
In these studies, adipose tissue (Güttes
etal.
1988,
Woolcott
etal.
2001, Waliszewski
etal.
2003d, Mus­
cat
etal.
2003), serum or plasma (Schecter
etal.
1997, Güttes
etal.
1988, Ward
etal.
2000, Wolff
et
al.
2000, López­Carrillo
etal.
2002) that were con­
venient to particular objectives were selected. The
obtained results reached conflicting conclusions in­
dicating positive or negative relationships between
organochlorine pesticide levels and the incidence of
breast cancer.
There are many reports that determine the rate
of exposure to persistent organochlorine pesticides
on breast cancer using blood due to the easiness to
obtain thesesamples. However, blood lipid levels may
fluctuate, biasing the results. The main contributor
of this bias is the transportation mechanism and par­
titioning of each pesticide into serum lipids, lipopro­
teins and various blood components according to their
physical­chemical properties and blood cell compo­
sition (Petreas
etal.
2004). Therefore, adipose tis­
sue not blood, sample of choice for assessing the
steady state of organochlorine pesticides in the body.
However, the disadvantage consists in that invasive
techniques are required to obtain samples.
The use of breast adipose tissue to determine the
level of organochlorine pesticides in breast cancer
studies limits the pool to subjects who were submit­
PESTICIDES IN WOMEN WITH BREAST TUMORS
135
ted to a biopsy after abnormal findings in the breast.
The serum/adipose tissue partition coefficient
(Waliszewski
etal.
2004) is the ratio of the concen­
tration of a pesticide in blood (blood as the transport
compartment and adipose tissue as deposit compart­
ment at equilibrium). This indicates the degree at
which the pesticide accumulates in fatty tissues of
the body. Because of interindividual differences in
the toxicokinetics of organochlorine pesticides, the
serum/adipose tissue partition coefficients have their
own ranges (Waliszewski
etal.
2004). This makes
the comparison of organochlorine pesticide concen­
trations in blood lipids and lipids of adiposetissue dif­
ficult. Minor variations have been observed in the
distribution of organochlorine pesticides in various
types of adipose tissues (Waliszewski
etal.
2003e).
The aim of this study was to determine the orga­
nochlorine pesticides levels in breast adipose tissue
of women subjected to surgical intervention result­
ing from thediagnosis of breast abnormalities. More­
over, the pool was divided according to the histo­
pathological diagnosis of benign and malignant breast
tumors and the results were statistically compared
to determine possible differences between study
groups and the control group.
MATERIALS AND METHODS
Study design. For the study, 254 patients admit­
ted to the Mexican Instituteof Social Security (IMSS)
Hospital for breast adipose tissue biopsies were cho­
sen to participate. All patients were voluntary par­
ticipants, who signed agreements to participate after
a clear explanation of the objectives of the study.
The preliminary diagnosis of participants submitted
to biopsies was breast tumor. The pooled group was
divided into malignant (127 cases) and benign (127
cases) breast tumors which were confirmed by a
histopathological study. Volunteers were selected
from those who have lived for at least one year in
Veracruz or its suburban zone and have not presented
additional disorders. During biopsy surgery, surgeons’
collected small amounts, of approximately 1­3 grams
of breast adipose tissue adhered to the tumor.
No differences in organochlorine pesticide con­
tents between abdominal and breast adipose have
been found (Waliszewski
etal.
2003e), so the con­
trol group was represented by 127 abdominal adi­
pose tissue samples from women subjected to au­
topsies in the Institute of Forensic Medicine of the
University of Veracruz as the consequence of auto­
mobile accidents.
For each participant we asked for additional data:
age, place of residence, demographic characteris­
tics, any lactation and medical history that could in­
fluence organochlorine pesticide content behavior.
Histopathology. The bioptic samples were pro­
cessed and evaluated in the Pathology Laboratory
of the IMSS Hospital to determine all histological
breast lesions. These diagnoses were for invasive
malignant diseases with metastasis and benign histo­
logical changes.
Sample Analysis
Apparatus
a)
Gaschromatograph
­Varian Model 3400CX (Palo
Alto, CA) equipped with
63
Ni electron capture de­
tector. Operating conditions: capillary chromatog­
raphy column J & W Scientific DB­608, 30 m x
0.32 mm id. and 0.83 micron film thickness; tem­
perature program, 193 °C (7 min) rising to 250 °C
at 6 °C/min, hold 20 min; carrier gas, nitrogen at
27 cm/min; 1 µL was injected in splitless mode.
b)
Integrator­
PC computer with Star Chromatog­
raphy Works Station Software Version 4.51
c)
Gaschromatograph­Massspectrometer.
Varian Model 3800 ­ Saturn 2000 GC­MS­MS
was used to confirm the determination of com­
pounds corresponding to organochlorine pesticide
peaks. The confirmation of peaks equivalent to
organochlorine pesticides, eluted from DB­608 30
m x 0.32 mm id. and 0.83 micron film thickness
capillary column, was performed comparing the
obtained mass spectra of substances from adi­
pose tissue extracts to those of standard sub­
stances, selecting the following specific ions ob­
tained from an ion trap detector (m/z values) of
HCB M+282
: 249, 214, 142 of HCH isomers
M+288
: 254, 219, 181 of
pp’
DDT M+352
and
op’
DDT:
M+352
: 235, 199, 165 and of
pp’
DDE
M+318
: 246, 210, 176. In the analyzed samples,
from the HCH isomers, only the presence of
b
­
HCH was confirmed in the adipose tissue
samples.
d)
RotaryevaporatorR114­
Büchi (Flawil, Swit­
zerland)
Reagents
(a)
Solvents­
Hexane Mallinckrodt Nanograde (Ken­
tucky, USA; part No. 4159)
(b)
Sulfuricacid­
95­97%
for analyses (Merck,
Mexico City, Mexico; part No. 1/15851).
(c)
Anhydroussodiumsulfate­
Powder (J.T. Baker,
part No. 3898­32) heated at 650 °C for 16 hours
before use.
S.M. Waliszewski
etal.
136
(d)
Analyticalstandardsofpesticides­
Hexachlo­
robenzene (HCB),
b
­ hexachlorocyclohexane (
b
­
HCH),
pp’
DDT,
op’
DDT,
pp’
DDE were pur­
chased from Supelco Inc. (USA)
Adipose tissue analysis. Breast adipose tissue
samples adhered to the tumor and abdominal adi­
pose tissue were analyzed for organochlorine pesti­
cide residues, according to the method described by
Waliszewski and Szymczynski(1982).Approximately
1­3 grams from a fat sample were ground in a mor­
tar with enough anhydrous sodium sulfate to obtain a
coarse powder. The pulverized sample was trans­
ferred to a 50 x 1 cm i.d. chromatographic column
with a fused­in fritted disk and Teflon stopcock. The
fats were extracted with 150 mL hexane, by passing
them through the column with a 3 mL/min flow rate
and collecting them in a 500 mL round­bottom flask.
Ten milliliters of extract were transferred to a previ­
ously weighed 50 mL round­bottom flask and the sol­
vent was rotary evaporated to dryness for gravimet­
ric fat determination. Because of the clean­up ca­
pacity, the fat content did not exceed 500 mg. An­
other 10 mL of extract were pipetted into a 15 mL
tube with a Teflon stopper and 1 mL concentrated
sulfuric acid was added. The tube was stoppered
tightly and vigorously shaken for half a minute and
then left to stand for 2­3 minutes to produce a good
phase separation. The supernatant was decanted,
dried by passing it through a 3­5 g layer of sodium
sulfate, and washed with hexane. The cleaned­up
extract, with rinses collected in another 50 mL round­
bottom flask, was rotary evaporated to a few drops
and quantitatively transferred to a 1 mL volumetric
tube. The volume was adjusted to 1 mL with hexane
and 1 µL aliquot was injected into a gas chromato­
graph for qualitative and quantitative analyses. The
concentrated sulfuric acid, used to clean­up fat ex­
tracts, allows quantitative fat precipitation and de­
grades the ubiquitous phthalate esters that interfere
in the gas chromatographic identification of orga­
nochlorine pesticides, permitting their accurate de­
termination (Waliszewski and Szymczynski1990).
All breast adipose and abdominal tissue samples
were analyzed for HCB,
b
­HCH isomer,
pp’
DDE,
op’
DDT and
pp’
DDT. The minimum detection lim­
its for the analyzed residues were 0.001 mg kg
­1
for
HCB and 0.002 mg kg
­1
on a fat basis for
b
­HCH,
pp’
DDE,
op’
­DDT and
pp’
­DDT. To determine the
quality of the method, a recovery study was per­
formed on 10 overspiked replicates of a blank cow
fat sample, which presented contamination levels
below detection limits. The results of the fortifica­
tion study, performed at 0.01 – 0.02 mg kg
­1
levels
for HCB and
b
­HCH,
pp’
DDE,
op’
DDT and
pp’
DDT, depending on the pesticide, showed mean
values from 91.3 to 97.2 %. The standard deviation
ranged from 6.1 to 8.2, indicating excellent method
repeatability.
Statistical analysis
Organochlorine pesticide levels from control,
malignant and benign groups were calculated using
basic statistics, such as mean with standard devia­
tion of means (
X
± SD), standard error of means
(SEM), geometric means (GM), medians and 95 %
of confidence intervals (95 % CI). To calculate the
variability between control and malignant breast tu­
mor groups, control and benign tumor population and
benign versus malignant breast tumor groups, the
obtained results were paired applying t­tests to com­
pare differences between mean values. To deter­
mine the magnitude of correlation between the study
groups, the Pearson correlation coefficients (
r
), co­
efficients of determination (r
2
), coefficients of re­
gression (
b
), F­test, that determines differences be­
tween variances and differences between medians
applying the Mann­Whitney confidenceintervals test,
were calculated. In order to determine the associa­
tion between the levels of organochlorine pesticides
in breast adipose tissue and breast tumor diseases,
the relative risk (RR) was calculated for all orga­
nochlorine pesticides.
RESULTS AND DISCUSSION
In the study, adipose tissue samples from 127
women with malignant breast tumors, 127 with be­
nign breast tumors and 127 in the control group with­
out breast lesions were analyzed. The origin of the
participants was the City of Veracruz and its sub­
urban zone. The mean age of the benign breast tu­
mor group was 48 years, for the malignant breast
tumor group, it was 49 years and for the control
group, it was 44 years. The age of one person, when
it varies significantly, can influence organochlorine
pesticide levels because of a more prolonged accu­
mulation and exposition (Sandau
etal.
2000, Deutch
and Hansen 2000, Glavan­Portillo
etal
. 2002,
Voorspoels
etal.
2002, Waliszewski
etal.
2002).
The determined differences between ages did not
vary significantly (p>0.05), thus the difference did
not influence the organochlorine pesticide levels
determined among the studied groups. The number
of child births were 1.9 ± 1.6 for benign, 1.9 ± 1.6
PESTICIDES IN WOMEN WITH BREAST TUMORS
137
for malignant and 2.0 ± 1.9 for the control group,
indicating no significant differences (p>0.05) result­
ing from the number of children born to these groups.
The lack of differences eliminates the principal fac­
tors: age and number of children that can impact the
organochlorine pesticide levels in breast adipose tis­
sue of the participants and bias the comparisons.
The comparison of mean organochlorine pesticide
levels determined in adiposetissueof thecontrol group
and participants with breast illness are presented in
table I and figure 1. Theconcentrations are expressed
on a lipid basis (mg kg
­1
) as mean values and standard
deviation of means (
X
± SD), standard error of means
(SEM), geometric means (GM), median values and
95 % of confidence intervals (95 % CI). Table I di­
vides the study pool into the control group, women
with malignant breast diseases and the group with
benign breast diseases.
The comparison indicated that all mean orga­
nochlorine pesticide (HCB,
b
­HCH,
pp’
DDE,
op’
DDT,
pp’
DDT and
S
­DDT) levels increased
stepwise from the control group to malignant cases,
finally reaching the highest values in benign breast
tumor cases. The HCB levels incremented from
0.045 mgkg
­1
in the control group to 0.099 mgkg
­1
in
malignant breast cases and to 0.116 mg kg
­1
in be­
nign breast cases. The
b
­HCH rose from 0.163 mg
kg
­1
in the control group to 0.265 mg kg
­1
in malig­
nant breast cases and to 0.319 mg kg
­1
in benign
breast cases. The
pp
’DDE increased from 0.782 mg
kg
­1
in the control group, to 0.980 mg kg
­1
in malig­
nant breast cases and to 1.761 mg kg
­1
in benign
TABLE I. MEAN AND STANDARD DEVIATION OF MEANS (
X
± SD), STAN­
DARD ERROR OF MEAN (SEM), GEOMETRIC MEAN (GM), MEDIAN
AND 95 % CI OF ORGANOCHLORINE PESTICIDE LEVELS (mg kg
­1
on a
fat basis) IN ADIPOSE TISSUE OF PATIENTS WITH BREAST TUMOR
Compound
Control (n=127)
Malignant (n=127)
Benign (n=127)
HCB
0.045 ± 0.032
0.099 ± 0.091
0.116 ± 0.158
X
± SD
0.003
0.008
0.011
SEM
0.020
0.054
0.060
G M
0.034
0.070
0.065
Median
0.039 – 0.050
0.083 – 0.115
0.089 – 0.144
95 % CI
b
­HCH
X
± SD
0.163 ± 0.119
0.265 ± 0.210
0.319 ± 0.292
SEM
0.011
0.019
0.026
G M
0.126
0.287
0.417
Median
0.135
0.189
0.258
95 % CI
0.142 – 0.183
0.228 – 0.301
0.269 – 0.371
pp’DDE
X
± SD
0.782 ± 0.282
0.980 ± 0.627
1.761 ± 1.090
SEM
0.025
0.056
0.097
G M
1.027
0.911
1.552
Median
0.740
0.901
1.543
95 % CI
0.733 – 0.731
0.871 – 1.089
1.571 – 1.950
op’DDT
X
± SD
0.035 ± 0.027
0.094 ± 0.098
0.176 ± 0.170
SEM
0.002
0.009
0.015
G M
0.046
0.047
0.096
Median
0.029
0.060
0.101
95 % CI
0.031 – 0.040
0.077 – 0.112
0.147 – 0.206
pp’DDT
X
± SD
0.296 ± 0.230
0.351 ± 0.291
0.661 ± 0.569
SEM
0.020
0.026
0.051
G M
0.306
0.370
1.149
Median
0.248
0.251
0.553
95 % CI
0.256 – 0.336
0.301 – 0.402
0.562 – 0.760
S
­DDT
X
± SD
1.112 ± 0.433
1.423 ± 0.856
2.601 ± 1.461
SEM
0.038
0.077
0.130
G M
1.445
1.340
2.962
Median
1.040
1.219
2.373
95 % CI
1.037 – 1.187
1.274 – 1.572
2.347 – 2.855
S.M. Waliszewski
etal.
138
breast cases. The
op’
DDT increased from 0.035 mg
kg
­1
in the control group, to 0.094 mg kg
­1
in malig­
nant breast tumors and to 0.176 mg kg
­1
in benign
breast cases. The insecticide
pp’
DDT augmented
from 0.296 mg kg
­1
in the control group to 0.351 mg
kg
­1
in malignant breast tumors and to 0.661 mg kg
­1
in benign cases. The sum of DDT (
S
­DDT=
pp’
DDE +
op’
DDT +
pp’
DDT) was 1.112 mg kg
­1
in thecontrol group, 1.423 mg kg
­1
in malignant breast
tumors and 2.601mg kg
­1
in benign breast tumors.
The standard deviation of means (SD) indicates
how spread out the data are. The SD values for all
organochlorine pesticides do not exceed the means,
indicating that the data points are not scattered sig­
nificantly. The standard error of means (SEM) mea­
sures how precise the population means is, indicat­
ing the extent of variability of observations. The cal­
culated results are low and oscillated from 3 to 9 %
of mean value, showing a low scattering of popula­
tions. The geometric mean (GEM) values confirm
the previous conclusion on the incremented tendency
of organochlorine pesticide levels from the control
group to the malignant breast tumor pool pronounc­
ing their significant increase in the benign breast can­
cer group. Calculated medians of organochlorine
pesticide concentrations evidenced lower values than
the means, expressing that the majority of the ob­
tained results are placed within lower levels. The
medians present the same tendency as arithmetic and
geometric means of increasing from control to ma­
lignant and benign breast tumor groups.
To demonstrate the magnitude of differences
among the studied groups (control group vs. benign
breast tumor, control group vs. malignant breast tu­
mor, benign breast tumor vs. malignant breast tumor),
table II presents the results of statistical compari­
son of means and variances. For that reason, paired
t­tests (that calculated the actual differences of
means between two populations studied), Pearson
correlation coefficients (
r
), coefficients of determi­
nation (
r
2
), the regression coefficient (
b
) (to deter­
mine the magnitude of differences) and F­test (to
determine the equality or homogeneity of variances),
were calculated.
The HCB results demonstrated statistically sig­
nificant differences (p<0.05) between mean levels
of control vs. benign and control vs. malignant breast
tumor cases. The comparison between benign and
malignant groups revealed no statistical differences
(p=0.2939). Moreover, the correlation and regres­
sion tests showed lower results, indicating different
origins in the three study groups. The same behav­
ior was present in
b
­HCH levels, evidencing sig­
nificant differences between control ­ benign and
control ­ malignant breast tumor cases (p=0.0001)
but no differences between benign and malignant
cases (p=0.0855). The correlation and regression
analyses demonstrated that the three study groups
have different origins and presented
bad correla­
tions among them.
The calculation made for
pp’
DDE
op’
DDT and
S
­DDT manifested statistically significant differ­
ences among mean residue levels of these compounds
in the three study groups (p=0.0001). The compari­
son of correlation and regression coefficients, to de­
termine the degree of linear relationship, suggested
bad correlations among study groups and different
accumulation rates of organochlorine pesticides in
control, benign and malignant cases. The compari­
son of
pp’
DDT levels between control and benign
breast tumor patients demonstrated significant dif­
ferences between means (p=0.0001) and a bad cor­
relation, whereas the comparison between control
and malignant breast cancer patients indicated no
differences between means (p=0.0949) and a bad
correlation between these groups. The calculations
made for benign and malignant breast tumor patients
revealed significant differences between mean lev­
els (p=0.0001) and a bad correlation (r= ­0.0306).
To specify the behavior of organochlorine pesti­
cide levels between control, benign and malignant
groups, the F­test was applied to determine equality
of variances. When the F­value is higher than 1, it
indicates that variances are unequal. For the study
groups, only
b
­HCH (benign vs. malignant,
F=0.0003) and
pp’
DDT (control vs. malignant,
F=0.0089) variances values were not different, re­
affirming the previous results of mean comparison
(p=0.0855 and p=0.0949) and lack of differences
control
malignant
benign
HCB
b­HCH
pp´DDE
op´DDT
pp´DDT
S­DDT
0
0.5
1
1.5
2
2.5
3
Control
Malignant
Benign
Fig. 1. Comparison of mean (mg kg
­1
onfat basis) organochlorine
pesticide levels among controls, malignant and benign
breast tumor cases
PESTICIDES IN WOMEN WITH BREAST TUMORS
139
between means. For all other pesticides and com­
parison groups, the F­values were higher than one,
indicating that the standard deviations are not equal,
reaffirming the existence of differences among
means.
In general, the organochlorine pesticide levels
determined in adipose tissue of the three patient
groups presented differences. From the control group,
the levels increased significantly in benign breast tu­
mors that constituted non invasive breast tumors. In
malignant breast tumors, the increase of organochlo­
rine pesticide levels was significant (except
pp’
DDT)
but the increase was not so clearly pronounced as in
benign cases. The comparison of obtained results
between benign and malignant cases, presented only
significant differences among the DDT family, indi­
cating higher values in benign than in malignant cases
and the correlation of these diseases to organochlo­
rine pesticide levels determined in breast adipose tis­
sue (Siddiqui
etal.
2005).
For the comparison of two population medians, a
two­sample Mann­Whitney test was applied (Table
III). The test computes the equality of two popula­
tion medians, corresponding point estimate and con­
fidence intervals. The calculation, made for all study
groups, asseverated the previous statements on dif­
ferences (
a
values smaller then 0.05) and diver­
gences between the population medians in the three
TABLE II. STATISTICAL
COMPARISON:
t
­TEST
(
p
),
PEARSON
CORRELATION
COEFFICIENT (
r
), COEFFICIENT OF DETERMINATION (
r
2
), REGRESSION
COEFFICIENT (
b
) AND F­TEST (F) AMONG CONTROLS, MALIGNANT AND
BENIGN BREAST TUMOR GROUPS
Compound
Control vs. benign
Control vs. malignant
Benign vs. malignant
HCB
0.045 vs 0.116
0.045 vs 0.099
0.116 vs 0.099
p
0.0001
0.0001
0.2939*
r
­0.0130
0.0983
0.1172
r
2
0.0002
0.0097
0.0137
b
­0.0026
0.0341
0.2033
F
4.471
5.960
1.705
b
­HCH
0.163 vs 0.319
0.163 vs 0.265
0.319 vs 0.265
p
0.0001
0.0001
0.0855*
r
­0.0623
0.1599
0.1672
r
2
0.0039
0.0256
0.0279
b
­0.0254
0.0882
0.2331
F
2.687
5.676
0.0003
op’DDE
0.782 vs 1.761
0.782 vs 0.980
1.761 vs 0.980
p
0.0001
0.0013
0.0001
r
­0.2953
0.0299
­0.0321
r
2
0.0381
0.0009
0.0011
b
­0.0505
0.0135
­0.0558
F
1.788
1.332
1.426
op’DDT
0.035 vs 0.176
0.035 vs 0.094
0.176 vs 0.094
p
0.0001
0.0001
0.0001
r
­0.1602
0.1513
0.0859
r
2
0.0256
0.0229
0.0074
b
­0.0259
0.0429
0.1503
F
4.050
1.267
3.223
pp’DDT
0.296 vs 0.661
0.296 vs 0.351
0.661 vs 0.351
p
0.0001
0.0949*
0.0001
r
­0.1308
0.0736
­0.0306
r
2
0.0171
0.0054
0.0009
b
­0.0528
0.0582
­0.0598
F
1.097
0.0089
3.556
S
­DDT
1.112 vs 2.601
1.112 vs 1.423
2.601 vs 1.423
p
0.0001
0.0003
0.0001
r
­0.1897
0.1056
­0.1099
r
2
0.0359
0.0111
0.0121
b
­0.0562
0.0534
­0.1877
F
4.976
1.839
4.529
* no significant differences at p<0.05
S.M. Waliszewski
etal.
140
study groups with the exception of
pp’
DDT. The
Mann­Whitney medians comparison test indicated
a
=0.151 between control and malignant breast tu­
mor cases for
pp’
DDT, stating the previous conclu­
sions of no differences between medians, means (p=
0.0949) and standard deviations (F= 0.089).
To determine the relationship between environ­
mental exposure to organochlorine pesticides and its
consequences for human health, the relative risk (RR)
and its 95 % confidence intervals (CI) were calcu­
lated (Table IV). The values obtained for benign
cases indicated higher relative risk than those deter­
mined for malignant cases, suggesting that their de­
velopment is correlated to organochlorine pesticide
exposure. From all compounds studied, the highest
RR values were those pertaining to
op’
DDT in be­
nign breast tumor cases (4.42; 95 % CI 4.22 – 4.62).
The compound is known as a highly estrogenic agent
that can interfere in thehormonal balance in women’s
bodies (Bulger
etal.
1983, Bustos
etal.
1988,
Robinson
etal.
1982). Thesecond place corresponded
to
pp’
DDT (2.33; 95 % CI 2.23 – 2.43),
S
­DDT
(2.21; 95 % CI 2.15 – 2.28),
pp’
DDE (2.13; 95 %
CI 2.06 – 2.21) and HCB (2.11; 95 % CI 1.98 ­
2.25) in benign breast tumor cases. The values of
relative risk (RR) indicated strong relationships be­
tween the organochlorine pesticide presence in hu­
man breast adipose tissue and a higher risk of devel­
oping benign breast tumors and in consequence a
stronger relationship to active hormonal compounds.
The obtained RR results are in concordance with
those calculated by Muscat
etal.
(2003) and Siddiqui
etal.
(2005), expressing the existence of an elevated
risk from these compounds on human health.
In conclusion, our study, carried out in the tropi­
cal region of Mexico, considered as contaminated
with organochlorine pesticide residues applied prin­
cipally to combat malaria, produced results that indi­
cate the correlation of their presence and levels in
breast adipose tissue to breast diseases. This level
correlates principally to benign breast diseases in
women living in Veracruz, Mexico.
REFERENCES
Bulger W.H. and Kupfer D. (1983). Estrogenic action of
DDTanalogs. Am. J. Ind. Med. 4,163­173.
TABLE III. RESULTS OF MANN­WHITNEY CONFIDENCE INTERVAL TESTS TO
DETERMINE DIFFERENCES AMONG STUDY GROUPS (CONTROL,
BENIGN AND MALIGNANT BREAST TUMOR PATIENTS)
Pesticide
Medians
difference
W
95 % CI
a
=0.05
HCB
b/c
0.065/0.034
0.028
20273
0.020,
0.037
0.000
m/c
0.070/0.034
0.031
20173
0.021,
0.042
0.000
HCH
b/c
0.258/0.135
0.106
20158
0.076,
0.141
0.000
m/c
0.189/0.135
0.062
18835
0.036,
0.095
0.000
pp’DDE
b/c
1.543/0.740
0.785
22446
0.632,
0.943
0.000
m/c
0.901/0.740
0.122
17535
0.019,
0.229
0.022
op’DDT
b/c
0.101/0.029
0.074
22754
0.057,
0.102
0.000
m/c
0.060/0.029
0.031
20591
0.023,
0.040
0.000
pp’DDT
b/c
0.553/0.248
0.252
19899
0.161,
0.366
0.000
m/c
0.251/0.248
0.032
17033
­ 0.014,
0.079
0.151*
S
­DDT
b/c
2.373/1.040
1.269
22138
1.019,
1.518
0.000
m/c
1.219/1.040
0.173
17635
0.036,
0.317
0.014
c= controls, b= benign breast tumors,
m= malignant breast tumors
*difference not significant at
a
=0.05
Compound
Relative risk (RR)
95%CI
HCB benign
2.11
1.98– 2.25
HCBmalignant
2.01
1.94– 2.07
b
­HCHbenign
1.96
1.90– 2.01
b
­HCHmalignant
1.58
1.54– 1.62
pp’DDE benign
2.13
2.06– 2.21
pp’DDE malignant
1.17
1.11– 1.23
op’DDT benign
4.42
4.22– 4.62
op’DDT malignant
2.27
2.18– 2.37
pp’DDT benign
2.33
2.23– 2.43
pp’DDT malignant
1.33
1.25– 1.41
S
­DDT benign
2.21
2.15– 2.28
S
­DDT malignant
1.19
1.15– 1.23
TABLE IV. RELATIVE RISK (RR) VALUES CALCULATED
FOR BENIGN AND MALIGNANT BREAST
TUMOR PATIENTS
PESTICIDES IN WOMEN WITH BREAST TUMORS
141
Bustos S., Denegri J.C., Diaz F. and Tchernitchin A.N.
(1988). pp’­DDT is an estrogenic compound. Bull.
Environ. Contam.Toxicol. 41, 496­501.
Calle E.E., Frumkin H., HenleyS.J., Savitz D.A. and Thun
M.J. (2002). Organochlorines and breast cancer risk.
CA.Cancer J. Clin.52,301­309.
Deutch B. and Hansen J.C. (2000). High human plasma
levels of organochlorine compounds in Greenland.
Danish Med. Bull. 47, 132­137.
Galván­PortilloM., Torres­Sánchez L. yLópez­CarrilloL.
(2002). Dietaryand reproductivefactors associatedwith
benign breast disease in mexican women. Nutr. Cancer,
43,133­140.
Golden R.J., Noller K.L., Titus­ErnstoffL., Kaufman R.H.,
Mittendorf R., Stillman R. and Reese E.A. (1998).
Environmental endocrine modulators and human
health: an assessment of the biological evidence. Crit.
Rev.Toxicol. 28, 109­227.
GüttesS., Failing K., Neumann K., Kleinstein J., Georgii S.
and Brunn H. (1998). Chlororganic pesticides and
polychlorinated biphenyls in breast tissue of women
with benign and malign breast disease. Arch. Environ.
Contam.Toxicol. 35, 140­147.
KelceW.R., Stone C.R., Laws S.C., GrayL.E., Kemppainen
J.A. andWilson E.M. (1995). Persistent DDTmetabolite
pp’DDE is a potent androgen receptor antagonist.
Nature15, 581­585.
López­CarrilloL., López­Cervantes M.,Torres­Sánchez L.,
Blair A., Cebrián­García M. and García R.M. (2002).
Serum levels ofbeta­hexachlorocyclohexane, hexachlo­
robenzene and polychlorinated biphenyls and breast
cancer in Mexican women. Eur. J. Cancer Prev. 11, 1­8.
Muscat J.E., Britton J.A., Djordjevic M.V., Citron M.L.,
Kemeny M., Busch­Devereaux E., Pittman B. and
Stellman S.D. (2003). Adipose concentrations of
organochlorine compounds and breast cancer
recurrencein LongIsland,NewYork. Cancer Epidemiol.
BiomarkersPrev. 12,1474­1478.
Petreas M., Smith D., HurleyS., JeffreyS.S., Gilliss D. and
Reynolds P. (2004). Distribution of persistent, lipid­so­
lublechemicals in breast and abdominal adipose tissue:
Lessons learned from a breast cancer study. Cancer
Epidemiol.Biomark. Prev. 13,416­424.
Robinson A.K. and Stancel G.M. (1982). The estrogenic
activity of DDT: correlation of estrogenic effect with
nuclearlevelofestrogen receptor.LifeSci. 31,2479­2484.
Safe S. (2000). Endocrine disruptors and human health – is
there a problem? An update. Environ. Hlth. Perspect.
108,487­493.
Schecter A., Toniolo P., Dai L.C., Thuy L.T.B. and Wolff
M.S. (1997). Blood levels of DDT and breast cancer
riskamongwomen livingintheNorth ofVietnam.Arch.
Environ. Contam.Toxicol. 33, 453­456.
Siddiqui M.K.,Anand M., Mehrotra P.K., Sarangi R. and
Mathur N. (2005). Biomonitoringof organochlorines in
women with benign and malignant breast disease.
Environ. Res. 98, 250­257.
Snedeker S.M. (2001). Pesticides and breast cancer risk:A
review of DDT, DDE and Dieldrin. Environ. Hlth.
Perspect. 109, 35­47.
Starek A. (2003). Estrogens and organochlorine
xenoestrogens and breast cancer risk. Int. J. Occupat.
Med. Environ. Hlth. 16, 113­124.
Tomlin C.D.S. (2000). The pesticide manual. 12
th
edition.
British Crop Protection Council, UK.
Voorspoels S., Covaci A., Maervoet J. and Schepens P.
(2002). Relationship between age and levels of orga­
nochlorine contaminants in human serum of a Belgian
population. Bull. Environ. Contam. Toxicol. 69, 22­29.
Waliszewski S.M. and Infanzon R.M. (2003a). Diferencias
en concentración de plaguicidas organoclorados per­
sistentes en suelo, paja y granos de trigo. Rev. Int.
Contam.Ambient. 19, 5­11.
Waliszewski S.M., Villalobos­Pietrini R., Gómez­ArroyoS.
and Infanzon R.M. (2003b). Persistent organochlorine
pesticide levels in cow’s milk samples from tropical
regions of Mexico. FoodAdd. Contam. 20, 270­275.
Waliszewski S.M., Infanzon R.M. and Hart M.M. (2003c).
Differences in persistent organochlorine pesticides
concentration between breast adipose tissue and blood
serum. Bull. Environ. Contam.Toxicol. 70, 920­926.
Waliszewski S.M., Meza Hernandez M.V., Infanzon R.M.,
TrujilloM.P. and Morales Guzman M.I. (2003d). Nive­
les de plaguicidas organoclorados persistentes en mu­
jeres con carcinoma mamario en Veracruz. Rev. Int.
Contam.Ambient. 19, 59­65.
Waliszewski S.M., Gomez­Arroyo S., Infanzon R.M.,
Villalobos­Pietrini R. and Maxwell Hart M. (2003e).
Comparison oforganochlorine pesticide levels between
abdominal and breast adipose tissue. Bull. Environ.
Contam.Toxicol. 71, 156­162.
Waliszewski S.M., Carvajal O., Infanzon R.M., TrujilloP.
and Maxwell Hart M. (2004a). Copartition rates of
persistent organochlorine pesticides between human
adipose tissue and blood serum lipids. Bull. Environ.
Contam.Toxicol. 73(4),732­738.
WardE.M., Schulte P.,Grajewski B.,AndersenA., Patterson
D.G., Turner W., Jellum E., Deddens J.A., Friedland J.,
Roeleveld N., Waters M., Butler M.A., DiPietro E. and
Needham L.L. (2000b).Serum organochlorinelevelsand
breast cancer: a nested case­control study of
Norwegian women. Cancer Epidem. Biomark. Prev. 9,
1357­1367.
WolffM.S. (1995) Pesticides– howresearch has succeeded
and failed in informing policy: DDT and the link with
breast cancer. Environ. Hlth. Perspect. 103, 87­91.
S.M. Waliszewski
etal.
142
Wolff M.S. and Toniolo P.G. (1995). Environmental
organochlorine exposure as a potential etiologic factor
in breast cancer. Environ. Hlth. Perspect. 103, 141­145.
Wolff M.S. and WestonA. (1997). Breast cancer risk and
environmental exposures. Environ. Hlth. Perspect. 105,
891­896.
WolffM.S., Zeleniuch­JacquotteA., Dubin N. andToniolo
P. (2000). Risk of breast cancer and organochlorine
exposure. Cancer Epidemiol. Biomark. Prev. 9, 271­277.
Woolcott Ch.G.,Aronson K.J., HannaW.M., SenGuotaS.K.,
McCready D.R., Sterns E.E. and Miller A.B. (2001).
Organochlorines and breast cancer risk by receptor
status, tumor size, and grade (Canada). Cancer Causes
andControl 12, 395­404.
Zou E. and Matsumura F. (2003). Long­term exposureto
b
­
hexachlorocyclohexane (
b
­HCH)promotestransforma­
tion and invasiveness of MCF­7 human breast cancer
cells. Biochem. Pharmacol. 66, 831­840.
logo_pie_uaemex.mx