02/(&8/$5(3,'(0,2/2*<2)6325$',&

%5($67&$1&(5
7KHUROHRISRO\PRUSKLFJHQHVLQYROYHGLQRHVWURJHQ
ELRV\QWKHVLVDQGPHWDEROLVP

.DWMD0LWUXQHQ

Division of Biochemistry
Department of Biosciences, Faculty of Science
University of Helsinki, Finland

Finnish Institute of Occupational Health

Department of Industrial Hygiene and Toxicology
Helsinki, Finland

$FDGHPLFGLVVHUWDWLRQ

To be presented with the permission of the Faculty of Science of the

University of Helsinki, for public criticism in Auditorium 1041 at Viikki
Biocentre, on May 4th, 2001, at 12 o´clock noon

Helsinki 2001
 6XSHUYLVRU
Dr. Ari Hirvonen
Department of Industrial Hygiene and Toxicology
Finnish Institute of Occupational Health
Helsinki, Finland

5HIHUHHV
Prof. Kaija Holli
Faculty of Medicine
University of Tampere
Tampere, Finland

Dr. Angela Risch

Department of Toxicology and Cancer Risk Factors
German Cancer Research Center
Heidelberg, Germany

2SSRQHQW
Dr. Elisabete Weiderpass
Unit of Field and Intervention Studies
International Agency for Research on Cancer
Lyon, France

ISBN 952-91-3388-X (nid.)

ISBN 951-45-9952-7 (PDF)

Yliopistopaino
Helsinki 2001

2
 7$%/(2)&217(176

$%675$&7 

$%%5(9,$7,216 

/,672)25,*,1$/38%/,&$7,216 

,1752'8&7,21

5(9,(:2)7+(/,7(5$785(
DESCRIPTIVE EPIDEMIOLOGY OF BREAST CANCER .....................................................10
DEVELOPMENT OF NORMAL AND NEOPLASTIC BREAST TISSUE...................................10
RISK FACTORS FOR BREAST CANCER ..........................................................................11
5HSURGXFWLYHIDFWRUV 
%RG\VL]H 
([RJHQRXVRHVWURJHQV
&LJDUHWWHVPRNLQJ 
$OFRKRO 
'LHWDU\IDFWRUV 
*HQHWLFIDFWRUV 
2WKHUULVNIDFWRUV 
OESTROGEN ...............................................................................................................17
2HVWURJHQDVDFDUFLQRJHQ 
2HVWURJHQELRV\QWKHVLV 
2HVWURJHQPHWDEROLVP 
GENETIC POLYMORPHISMS IN OESTROGEN METABOLISING ENZYMES AND BREAST
CANCER .....................................................................................................................22
(Q]\PHVLQYROYHGLQRHVWURJHQELRV\QWKHVLV
(Q]\PHVLQYROYHGLQRHVWURJHQPHWDEROLVP
$,062)7+(678'<

68%-(&76$1'0(7+2'6 
STUDY SUBJECTS .......................................................................................................38
&DVHV 
&RQWUROV
METHODS ..................................................................................................................40
*HQRW\SLQJDQDO\VHV 
6WDWLVWLFDODQDO\VHV
5(68/76
CHARACTERISTICS OF THE STUDY POPULATIONS .......................................................44
CYP17 AND BREAST CANCER RISK (PAPER I) ............................................................46
COMT AND BREAST CANCER RISK (PAPER II) ...........................................................47
MNSOD AND BREAST CANCER (PAPER III)...............................................................48
GSTS AND BREAST CANCER RISK (PAPER IV)............................................................49
COMBINED EFFECT OF COMT AND GST GENES ON BREAST CANCER RISK (PAPER V)
..................................................................................................................................53

3
',6&866,21 
METABOLIC GENOTYPES AND OVERALL BREAST CANCER RISK ..................................55
&<3 
&207 
0Q62' 
*67V
GENE-GENE INTERACTIONS AND BREAST CANCER RISK .............................................58
GENE-ENVIRONMENT INTERACTIONS AND BREAST CANCER RISK...............................60
(QGRJHQRXVKRUPRQHV 
([RJHQRXVKRUPRQHV 
6PRNLQJ
$OFRKRO 
METHODOLOGICAL CONSIDERATIONS........................................................................63
CONCLUDING REMARKS.............................................................................................65
$&.12:/('*(0(176

5()(5(1&(6 

4
$%675$&7

The major known risk factors for female breast cancer are associated with
prolonged exposure to increased levels of oestrogen. The predominant theory
relates to effects of oestrogen on cell growth. Enhanced cell proliferation, induced
either by endogenous or exogenous oestrogens, increases the number of cell
divisions and thereby the possibility for mutation. However, current evidence also
supports a role for oxidative metabolites, in particular catechol oestrogens, in the
initiation of breast cancer.

As observed in drug and chemical metabolism, there is considerable

interindividual variablity (polymorphism) in the conjugation pathways of both
oestrogen and catechol oestrogens. These person-to-person differences, which
are attributed to polymorphisms in the genes encoding for the respective
enzymes, might define subpopulations of women with higher lifetime exposure to
hormone-dependent growth promotion, or to cellular damage from particular
oestrogens and/or oestrogen metabolites. Such variation could explain a portion
of the cancer susceptibility associated with reproductive events and hormone
exposure.

In this thesis, polymorphic genes encoding for cytochrome P450 17 (CYP17),

catechol-2-methyltransferase (COMT), glutathione 6-transferases (GST) M1, M3,
P1 and T1, and manganese superoxide dismutase (MnSOD) were studied in
relation to breast cancer risk in Finnish women. A moderate, around 1.5-fold
increased risk was seen for postmenopausal women with either *670 null
genotype or 0Q62' $ allele containing genotypes, while combination of the
*670 %, *673 Ile/Ile and *677 null genotypes posed a remarkably
increased risk among premenopausal women (10-fold). Although no overall effect
was found for &<3 and &207genotypes, they appeared to modify the breast
cancer risk associated with endo- and exogenous oestrogens. Of particular
interest was the finding that women with long-term use (>30 months) of
postmenopausal hormone replacement therapy and carrying either the 0Q62'
$$ genotype or the &207/ allele containing genotype had a 3-fold increased
risk of breast cancer. Even higher risks were seen for women concurrently
carrying the putative at-risk genotypes of the *673 or *677 genes together
with the &207/ allele containing genotypes. A trend of increasing risk was seen
with increasing number of high-risk genotypes and time of use of hormones.
Smokers, current users of alcohol, or ever users of oral contraceptives were also
shown to be at a somewhat higher risk when carrying certain at-risk genotypes.

5
$%%5(9,$7,216

BBD benign breast disease

BMI body mass index
bp base pair
CE-Q catechol oestrogen quinone
CE-SQ catechol oestrogen semiquinone
CI confidence interval
CIS carcinoma in situ
COMT catechol-2-methyltransferase
CYP cytochrome P450
DCIS ductal carcinoma in situ
E1 oestrone
E2 17β−oestradiol
ER oestrogen receptor
FFTP first full term pregnancy
GSH glutathione
GST glutathione 6-transferase
HAA heterocyclic aromatic amine
HRT hormone replacement therapy
HSD hydroxysteroid dehydrogenase
IDC invasive ductal carcinoma
ILC invasive lobular carcinoma
LCIS lobular carcinoma in situ
MeOE2 methoxyoestradiol
MnSOD manganese superoxide dismutase
MTS mitochondrial targeting sequence
NAT 1-acetyltransferase
OC oral contraceptives
OHE2 hydroxyoestradiol
OR odds ratio
PAH polycyclic aromatic hydrocarbon
PR progesterone receptor
RFLP restriction fragment length polymorphism
ROS reactive oxygen species
SD standard deviation
SOD superoxide dismutase
WHR waist-to-hip ratio
XME xenobiotic metabolising enzyme
/,672)25,*,1$/38%/,&$7,216

This dissertation is based on the following original publications, referred to in

the text by the Roman numerals I -V, as indicated below.

I. Katja Mitrunen, Nadejda Jourenkova, Vesa Kataja, Matti Eskelinen,

Veli-Matti Kosma, Simone Benhamou, Harri Vainio, Matti Uusitupa, and
Ari Hirvonen. Steroid metabolism gene &<3 polymorphism and the
development of breast cancer. &DQFHU (SLGHPLRO %LRPDUN  3UHY.
200091343-1348.

II. Katja Mitrunen, Nadejda Jourenkova, Vesa Kataja, Matti Eskelinen,

Veli-Matti Kosma, Simone Benhamou, Daehee Kang, Harri Vainio,
Matti Uusitupa, and Ari Hirvonen. Polymorphic catechol 2-
methyltransferase gene and breast cancer risk. &DQFHU (SLGHPLRO
%LRPDUN 3UHY 2001, 10: 000-000 (in press).

III. Katja Mitrunen, Pia Sillanpää, Vesa Kataja, Matti Eskelinen, Veli-Matti
Kosma, Simone Benhamou, Matti Uusitupa, and Ari Hirvonen.
Association between manganese superoxide dismutase (0Q62')
gene and breast cancer risk. &DUFLQRJHQHVLV, 2001, 22: 000-000 (in
press).

IV. Katja Mitrunen, Nadejda Jourenkova, Vesa Kataja, Matti Eskelinen,

Veli-Matti Kosma, Simone Benhamou, Harri Vainio, Matti Uusitupa, and
Ari Hirvonen. Glutathione 6-transferase 0 0 3, and 7, genetic
polymorphisms, and susceptibility to breast cancer. &DQFHU(SLGHPLRO
%LRPDUN 3UHY. 2001, 10: 225-232.

V. Katja Mitrunen, Vesa Kataja, Matti Eskelinen, Veli-Matti Kosma,

Simone Benhamou, Harri Vainio, Matti Uusitupa, and Ari Hirvonen.
Combined &207 and *67 genotypes and breast cancer risk among
users of hormone replacement therapy.3KDUPDFRJHQHWLFVsubmitted.
,1752'8&7,21

Breast cancer is the prevailing cancer among women in industrialised countries

(1). Heritable factors are observed in one fourth of breast cancer cases (2).
However, germline mutations in so called high penetrance cancer susceptibility
genes, such as %5&$ and %5&$, have been shown to account for only up to
5 % of all breast cancer cases (3,4), with even lower frequencies reported in
Finns (5). Therefore relatively common genes, acting together with
endogenous/lifestyle risk factors (low penetrance genes), are likely to account for
a much higher portion of the breast cancer cases (6).

The molecular mechanisms underlying the development of breast cancer are not
completely understood. However, it is generally believed that the initiation of
breast cancer is a consequence of cumulative genetic damages leading to
genetic alterations that result in activation of proto-oncogenes and inactivation of
tumour suppressor genes. These in turn are followed by uncontrolled cellular
proliferation and/or aberrant programmed cell death, or apoptosis. Also the role
of reactive oxygen species (ROS) has been related to the aetiology of cancer (7).

Most of the risk factors for breast cancer relate to the increased or prolonged
exposure to oestrogen. The main effect of oestrogens has been thought to be via
stimulation of breast-cell proliferation, thereby increasing the chances that a cell
bearing a potentially cancer-causing mutation will multiply (8,9). The initial
genetic damage was previously thought to arise solely from spontaneous
mutations or damage triggered by external exposures such as radiation and
cigarette smoke. However, current evidence suggests that the metabolic by-
products of oestrogen in the body may also act as initiators of cellular alterations
(10).

Considerable inter-individual variability has been observed in carcinogen

metabolism as well as in the biosynthetic pathways and metabolism of steroid
hormones (11). These person-to-person differences are largely attributed to
polymorphism in the genes encoding for the xenobiotic metabolising enzymes
(XMEs). The XME gene polymorphisms may therefore define subpopulations of
women with higher lifetime exposure to oestrogens, oestrogen metabolites, and
other carcinogens (12). Such variation could explain a portion of the breast
cancer susceptibility associated with reproductive events and hormone exposure,
as well as other lifestyle/environmental risk factors. They are therefore
considered to account for a high proportion of breast cancer cases.

In this work we examined the potential role of polymorphic genes coding for
enzymes involved in oestrogen biosynthesis and metabolism in modulating
individual susceptibility to breast cancer.

8
 Endogenous hormones/
Environmental exposures

*HQHWLFKRVWIDFWRUV

DFWLYDWLRQ GHWR[LILFDWLRQ

Ultimate carcinogens

Genotoxic damage

Altered expression of cell cycle

related genes, oncogenes,
tumour suppressor genes

Aberrant proliferation

Breast cancer

)LJXUH . The potential role of polymorphic metabolic enzymes in breast

carcinogenesis.

9
5(9,(:2)7+(/,7(5$785(

'HVFULSWLYHHSLGHPLRORJ\RIEUHDVWFDQFHU

With 1 million new cases diagnosed in the world annually, breast cancer is by far
the most common female cancer, comprising 21% of all new cancers in women
(1,13). The highest age-adjusted incidence rate is reported for North America,
being 86.3 per 100 000 women per year, while the lowest rate, reported in China,
is only 11.8 (13). In Finland, 3324 new breast cancer cases were recorded in
1997, corresponding to an age-adjusted incidence rate of 78.6 cases per 100 000
women per year (14). Breast cancer follows a steeply increasing age gradient up
to 50 years of age, after which age the rate of increase slows down.

Even though there are three times as many new cases diagnosed annually as in
the late 1960s, breast cancer mortality has remained largely unchanged. This
may at least partly be explained by earlier detection of the disease, due to
effective screening programs and availability of improved therapies (15). The
highest annual mortality rates for breast cancer are reported for the UK,
Netherlands and Denmark, where they are over 25 per 100 000 (15). In Finland
the cumulative 1, 5, and 10-year survival rates are 96%, 80%, and 68%,
respectively (16). Yet, breast cancer is the leading cause of cancer death among
Finnish women; in 1997 there were 807 deaths due to breast cancer, the age-
adjusted mortality rate being 16.3 per 100 000 women per year (14).

The high rates of breast cancer incidence and mortality in industrialised, Western
nations and lower rates for less industrialised and Asian nations are considered
to mirror differences in factors important in the aetiology of the disease such as
parity, hormonal, environmental and dietary exposures.

'HYHORSPHQWRIQRUPDODQGQHRSODVWLFEUHDVWWLVVXH

Oestrogens are ovarian steroid hormones that act on the mammary gland
through their receptors. They are considered to play a major role in promoting the
proliferation of both normal and the neoplastic breast epithelium (17). The ductal
system of the adult human breast consists of 15-25 lactiferous ducts opening at
the nipple. Oestradiol (E2) acts locally on the mammary gland, stimulating DNA
synthesis and promoting bud formation. Lobule formation in the female breast
occurs within 1-2 years after the onset of the first menstrual period. Full
differentiation of the mammary gland is a gradual process, which takes years to
complete. The breast of nulliparous women contains more undifferentiated
structures while in premenopausal parous women the predominant structure is
the most differentiated form. Full lobular differentiation only occurs in parous
women, especially in those experiencing full term pregnancy (FTP) at a young
age. After menopause the breast undergoes regression in both nulliparous and
parous women manifested as an increase in the undifferentiated structures.

10
The most undifferentiated structure in the breast has been found to be the site of
origin of ductal carcinomas. Also more differentiated lobular structures have been
found to give rise to tumours but they are less aggressive (17). In situ breast
carcinoma (CIS) is a transitional phase in the evolution of invasive malignancy
from normal breast tissue. There are two histological forms of CIS. Ductal
carcinoma in situ (DCIS), also known as intraductal carcinoma or noninvasive
ductal carcinoma, does not invade the surrounding stroma. Lobular carcinoma in
situ (LCIS) is characterized by clusters of analplastic small cells of high nuclear
grade that lie within the lobules (18). The change is usually both multicentric and
bilateral. LCIS is considered as a premalignant stage with a substantially higher
risk of subsequent development of carcinoma.

Infiltrating (invasive) ductal carcinoma (IDC) is the most common type of breast
cancer with no special histologic features, accounting for nearly 75% of the cases
(19). It is characterized by stony hardness at palpation. These tumours
commonly metastasise to bone, lung, liver, and axillary lymph nodes and have a
poor prognosis. Other rare types of ductal carcinoma include medullary, tubular,
mucinous and papillary carcinoma, which have a lower incidence of nodal
involvement and better prognosis. The second common type of breast cancer is
infiltrating (invasive) lobular carcinoma (ILC) accounting for about 10% of breast
tumours. It is characterized by ill-defined thickening in the breast, and is
microscopically composed of small cells in a linear arrangement with a tendecy to
grow around ducts and lobules. ILC are also characterized by a greater
proportion of multicentric tumours, either in the same or the opposite breast, and
has the same probability to metastasise as IDC though invading preferentially to
meninges, serosal surfaces, and retroperitoneum.

5LVNIDFWRUVIRUEUHDVWFDQFHU

5HSURGXFWLYHIDFWRUV
Recent studies have revealed the relationship between endogenous hormone
levels and breast cancer risk (20-22). The serum levels of at least oestrone,
oestradiol, oestriol, androstenedione, testosterone, dehydroepiandrostenone,
progesterone, sex-hormone-binding globulin, and prolactin have been considered
as important in this context (23). Lifetime exposure to endogenous sex hormones
is determined by several variables including timing of menarche, age at first full
term pregnancy, number of pregnancies, and age at menopause. These
variables have therefore all been studied in relation to breast cancer risk (24).

The increased breast cancer risk associated with early age at menarche (< 12
years) (25-29), probably results from prolonged exposure of breast epithelium to
oestrogens, earlier onset of regular menstrual cycles, and higher oestrogen
levels for several years after menarche in women with early menarche (26,29).
Similarly, later age at menopause maximises the number of ovulatory cycles, and
may therefore lead to increased risk. For every one-year increase in the age at
menopause, the risk of breast cancer increases by approximately three percent
(30).

11
Higher parity and early age at first birth have both been associated with
decreased lifetime incidence of breast cancer (28,31,32). Women having their
first child before the age of 20 have about half of the risk compared to those
having their first child after the age of 30. The mechanism of the protective effect
of parity is not completely understood but it probably reflects early full
differentiation of mammary gland cells, rendering them less susceptible to
carcinogenic transformations (17). Also prolonged lactation has been associated
with decreased risk (33).

Many of the above mentioned hormone related factors have recently been shown
to be more strongly associated with oestrogen receptor (ER) and progesterone
receptor (PR) positive than negative breast cancers (34,35).

%RG\VL]H
Obesity has been related to both higher endogenous oestrogen levels and
increased risk of breast cancer in postmenopausal women who have most of
their circulating oestrogen derived from conversion of androgen to oestrogen in
adipose tissue (36). It was recently shown that LQVLWX aromatization catalysed by
cytochrome P450 19 (CYP19) can effectively increase tissue oestradiol levels
more than production by the uptake of oestrogen from circulation (37). In
contrast, an inverse association is suggested for obesity and premenopausal
breast cancer; a greater degree of anovulation could result in lower levels of
oestrogen. However, the prevailing theory suggests a more important role for the
localisation of body fat measured as waist-to-hip ratio (WHR) (38-40). Physical
activity has also been associated with decreased breast cancer risk (27,29,41).
Suggested mechanisms include reducing regulatory ovulatory cycles (29) and
increasing the amount of methylated catechol oestrogens (41).

([RJHQRXVRHVWURJHQV
Since sex hormones have become one of the most widely used drugs among
women, concern has been raised about their safety. It has not usually been
possible to study the effects of single hormones; many are used either in
combination or consecutively in the same patient. Therefore, risks are generally
assessed in relation to the therapeutic goal of the treatment, LH, oral
contraception or hormone replacement therapy (HRT).

The results from studies on the role of oral contraceptives in breast cancer
proneness have been somewhat conflicting. However, a recent meta-analysis
using data from 54 studies concluded that current use of oral contraceptives
poses a slight (24%) increase in the risk, which disappears 10 years after the
cessation of use (30).

The data from 51 studies indicated that risk of having breast cancer is slightly
increased in women using HRT (42). A 35% increase in the risk was seen for
women who have used HRT for 5 years or longer, being comparable with the
effect of delaying menopause, and the increase largely disappeared 5 years after
terminating the use of hormones. The combined estrogen-progestin regimen is
associated with greater increases in the risk than estrogen alone (43). Moreover,

12
in some studies, cancers in women who have ever used HRT tend to be less
advanced at diagnosis and biologically less aggressive than those in never users
(44), but also contrasting results exist (45). The overall mortality among HRT
users has been shown to be lower but the benefit diminishes with longer duration
of use (46).

Xeno-oestrogens include pesticides, dyes, pollutants, plasticizers and food

preservatives that have oestrogen-like effects, and they have been shown to
have a role in the aetiology of breast cancer (47-52). Xeno-oestrogens have also
been called endocrine disrupters because they interfere with the actions of
endogenous oestrogens. For instance, catechol metabolites of polychlorinated
biphenyls (PCBs) have been suggested to alter oestrogen metabolism by
inhibiting the inactivation of carcinogenic oestrogen metabolites (51).

&LJDUHWWHVPRNLQJ
The studies on smoking and breast cancer susceptibility have given rise to
controversial results, and the risk has at its best been only weakly associated
with cigarette smoking (53,54). However, statistically significant effects have
been seen for early age at starting, and for heavy, current, and passive smoking
(54,55). On the other hand, some agents in tobacco smoke may have anti-
oestrogenic effects; for instance nicotine has been shown to inhibit the CYP19
enzyme (56). Accordingly, smokers are known to have an earlier age at
menopause.

Cigarettes contain over 3000 compounds, including 30 known carcinogens. The

most important carcinogens in tobacco smoke are polycyclic aromatic
hydrocarbons (PAHs), aryl amines, heterocyclic aromatic amines (HAAs), and 1-
nitrosamines (54,57). The ingested or inhaled PAHs are converted to water-
soluble derivatives mainly via oxidative activation by cytochrome P450 1A1
(CYP1A1) followed by detoxification by phase II enzymes such as glutathione 6-
transferases (GSTs). PAHs have been shown to be mutagenic to breast cell
lines, and as lipophilic compounds they are stored in adipose tissues, including
breast. Accordingly, PAH-DNA-adducts have been identified in normal breast
tissue from women with and without breast cancer (58-60). Aromatic amines can
be detoxified by 1-acetylation catalysed by 1-acetyltransferases (NATs), while
the NAT mediated 2-acetylation of hydroxylamines results in their activation
(61,62).

Cigarette smoke is a rich source of radicals, and some antioxidant enzymes like
manganese superoxide dismutase (MnSOD) have been shown to be induced by
smoke (63,64). The cigarette tar semiquinone radical can reduce oxygen to
produce superoxides, and hence stimulate production of hydrogen peroxide and
the hydroxyl radical (⋅OH) (64). The latter is extremely reactive and therefore a
prime candidate for producing endogenous oxidation of DNA (65).

13
$OFRKRO
Alcohol has been described as being one of the most constant enhancers of
breast cancer risk. In a recent analysis pooling six cohort studies, alcohol was
shown to increase breast cancer risk linearly with alcohol consumption from 1 to
5 drinks/day (66). Rather surprisingly, however, only a modest 15% increase in
the risk was seen in a study of alcoholic women (67).

Drinking has been suggested to be particularly harmful in the early adult life (68).
The mechanism underlying the carcinogenic effect associated with alcohol is
incompletely understood. However, an increase in the oestrogen level in women
consuming alcohol has been hypothesized (69). Other recently proposed
mechanisms include alcohol induced production of ROS (70), and increased
adduct formation, possibly due to decreased protein expression of detoxification
enzymes (71).

'LHWDU\IDFWRUV
The human diet contains a great variety of natural carcinogens and
anticarcinogens (72). Many of these may act through the generation of oxygen
radicals, which in turn may lead to DNA damage. Accordingly, a high intake of
fat, especially that of polyunsaturated fatty acids, has been shown to increase
breast cancer risk (73,74), while intake of fruits and vegetables, sources of
natural antioxidants, has been shown to decrease the risk (75,76). Also the
consumption of meat has been associated with increased breast cancer risk in
some, but not all, studies (77). Dietary fat has long been suspected to be the
reason for this association but recent studies support the role of HAAs found in
well-done meat. HAAs require metabolic activation by NATs to be able to exert
their harmful effects (61,78). Soy, or more specifically genistein, with a chemical
structure similar to steroidal oestrogens, has been shown to have both anti-
carcinogenic and cancer promoting effects (79). An example of a unexpected
interaction has been found for broccoli and colorectal adenoma. GSTM1 was
found to increase the risk, probably because the cancer chemopreventive
isothiocyanate is conjugated by GSTs rendering it inactive (80).

*HQHWLFIDFWRUV
In a recent study including mono- and dizygotic twins from Sweden, Denmark
and Finland, the heritable factors were shown to account for 27% of all breast
cancer cases (2). As mentioned earlier, two classes of susceptibility genes exist:
1) genes with allelic variants that confer high individual risk of breast cancer,LH,
so called high penetrance genes), 2) genes that confer a small to moderate
cancer risk to individuals who carry these relatively common disease-associated
variant alleles (low penetrance genes) (81). As disease-causing allelic variants of
the high penetrance genes are expected to be relatively rare in the general
population, the population attributable risk (LH the proportion of breast cancer in
the population that may be explained by these genotypes) of the low-penetrance
genes is likely to be much higher than that of the high-penetrance genes (81).

14
Breast cancers in families with several affected members are likely to be
hereditary. %5&$ %5&$ $70 and S are the most common high
penetrance genes exhibiting allelic variants predisposing to hereditary breast
cancer (3). They are tumour suppressor genes that are estimated to be involved
in around half of the familial clustering of early onset breast cancer (3) but only
about 5 % - 10% of all breast cancers (4,82). They tend to predispose to both
earlier onset of the disease and multifocal tumours (81).

BRCA1 and BRCA2 proteins are involved in maintaining genomic stability, the
cellular response to DNA damage, transcriptional regulation, and cellular
proliferation (83). %5&$, located on chromosome 17, shows autosomal
dominant transmission, and seems to be responsible for the majority (42%) of
hereditary female breast cancers (84). Recently, loss of a 13q region, harbouring
also %5&$ gene, was shown to be an early event, possibly initiating genetic
alteration in hereditary cancers (85). Mutations in%5&$gene have been shown
to account for 32% of hereditary female breast cancer cases and most of the
male breast cancers (84). In families with women carrying %5&$ or %5&$
mutations, the risk is estimated to be over 80% by the time the carrier reaches
the age of 70.

In Finland, only 10 % of breast cancer families carry mutations in the %5&$ and
11% in %5&$ genes, frequencies that are much lower than those reported in
most other countries (86). Moreover, the frequency of the %5&$ and %5&$
mutations was recently reported to be only 1.8% in unselected Finnish breast
cancer patients (5). Therefore a significant proportion of breast cancer cases
cannot be explained by these genes.

In contrast to familial cases, low-penetrance genes contribute to sporadic cases

of breast cancer that usually appear unilaterally, and have a relatively late age at
diagnosis (81). These tumours are likely to be caused by the interaction between
many genetic and environmental factors.

In general, the search for low-penetrance susceptibility genes relies on

knowledge about biochemical or physiological pathways that are thought to be
involved in breast carcinogenesis. Candidate polymorphic genes include those
encoding for enzymes involved in oestrogen and carcinogen metabolism, and in
the detoxification of reactive oxygen species emerging in these reactions.
Depending on the substrate, the enzymes can have either inactivating or
activating roles in these pathways. Among the most widely studied low-
penetrance genes in relation to breast cancer are those encoding for CYPs,
GSTs, NATs and catechol-2-methyltransferase (COMT) (11,87,88). Some
studies also exist on MnSOD (89), oestrogen receptor (ER) (90), progesterone
receptor (91), and vitamin D receptor gene polymorphisms (92-94). A new
potential group of polymorphic genes of interest in this context are the repair
genes (95).

15
2WKHUULVNIDFWRUV
Ionising radiation has been shown to increase breast cancer risk among female
flight attendants, nurses, chemists and insulators (96-98). Although
electromagnetic fields have also been hypothesised to affect breast cancer risk
by suppressing melatonin production, the data does not provide strong evidence
to support this theory (99). Other occupational studies have provided some
evidence of an association between breast cancer risk and employment in the
pharmaceutical industry, among cosmetologists, beauticians, chemists, teachers,
social workers, and cashiers (100,101).

A positive association has been found between higher socioeconomic status and
breast cancer (102), which is largely explained by differences in the known risk
factors such as nulliparity, later age at first birth, diet, and greater use of synthetic
hormones (102,103).

A history of benign breast disease (BBD) is also a well-established risk factor for
breast cancer. Women with severe atypical epithelial hyperplasia have been
found to have significantly higher risk compared to women who do not have any
proliferative changes in their breast (1).

16
2HVWURJHQ

All oestrogens have an aromatic A ring, a phenolic hydroxyl group at C-3 and a
methyl group at position C-13 (Figure 2). Oestradiol (E2) with a hydroxyl group at
C-17 and oestrone (E1) with a keto-group at this position, are the major
oestrogens in the blood, oestradiol being biologically the most active in breast
tissue (104).

OH
12 18 17
11 16
1 &
13 '
10
2 9 15
8 14
$ %
7
HO 3 5
6
4

)LJXUH. Structure of 17β-oestradiol (E2).

Oestrogens and progesterone exert their cellular actions by forming complexes

with their respective receptors (105). The extent of binding to ER is determined
by the oestrogen concentration and the oestrogen moiety; oestradiol has the
greatest affinity for the ER. The ER is present in malignant tissue at significantly
higher amounts than in normal cells (17).

2HVWURJHQDVDFDUFLQRJHQ
Oestradiol has been shown to induce aneuploidy and structural chromosomal
changes (106). The biological studies in animals have clearly identified oestradiol
as a carcinogen (8,107,108), and according to IARC there is sufficient evidence
for its carcinogenicity in humans (109).

Stimulation of breast-cell proliferation has been proposed as the main effect of

oestrogens in carcinogenesis (8,9,24). It has been hypothesised that the more
rapidly cells proliferate, the greater the chance of genetic error. Single-stranded
DNA, present during cell division, is more susceptible to damage than double-
stranded DNA. Once mutations are introduced into a target cell, oestrogens
enhance the replication of the clones of cells carrying such genetic errors.

17
Recent studies suggest that oestrogen metabolites also act as initiators of
genetic damage (9,106,108,110). Among the reported free-radical mediated DNA
damage induced by oestrogen and their metabolites are single-stranded breaks,
increased concentrations of 8-hydroxyguanine DNA-bases, oestrogen-DNA
adducts, and protein oxidation and lipid peroxidation (106,111-113). These
theories of enhanced cell proliferation and genotoxic metabolites can act in
additive or synergistic fashions.

2HVWURJHQELRV\QWKHVLV
In premenopausal non-pregnant women, nearly all oestrogen is of ovarian origin,
while after menopause most oestrogen is formed by aromatization of
androstenedione to oestrone in peripheral adipose tissue (114-116).

Biosynthesis of oestrogens involves a lengthy series of enzymatic steps from

cholesterol to C19 androgens and C18 oestrogens. In the biosynthesis of
oestrogen CYP11A, CYP17, and CYP19 are particularly important (Figure 3)
(88,117,118). In addition to CYPs, hydroxysteroid dehydrogenases, 3β-HSD and
17β-HSD, are involved in the biosynthesis (104). The cleavage of the side chain
of cholesterol by CYP11A to form pregnanolone and progesterone (C21 steroids)
is a rate-limiting step in the biosynthesis of all steroids. Hydroxylation and
subsequent cleavage of the C21 steroids by CYP17 yields the C19 steroids,
androstenedione and dehydroepiandrosterone. CYP19 catalyses the final steps
from androgens to oestrogens.

2HVWURJHQPHWDEROLVP
Oestrogens undergo extensive oxidative metabolism via the action of several
CYPs (118). The generation of hydroxyl groups at different sites of the molecule
affects the biological property of the respective oestrogen metabolites yielding
estrogenic, nonestrogenic or even carcinogenic metabolites.

Oestradiol and oestrone are metabolised via two major pathways, LH, by
hydroxylation either at the A-ring or the D-ring (119). Hydroxylation at the A-ring
leads to formation of catechol oestrogens (CE), namely 2- or 4-hydroxyoestrone
or -oestradiols (Figure 4), while hydroxylation at the D-ring yields 16α-
hydroxyoestrone. Catechol oestrogens are the major metabolites of oestrogen
(107).

The major metabolic pathway for oestrogen in liver is 2-hydroxylation

(108,120,121). In humans, hepatic 2-hydroxylation is mainly catalysed by
CYP1A2, 3A3 and 3A4, while CYP1A1 is responsible for extrahepatic 2-
hydroxylation (108,120,121). In contrast, CYP1B1 appears to be the main CYP
responsible for the 4-hydroxylation, which dominates in many extrahepatic
tissues such as human breast and uterus (108,120,121). The 4-hydroxyoestradiol
(4-OHE2) to 2-hydroxyoestradiol (2-OHE2) concentration ratio has been reported
to be 4:1 in a human breast cancer extract (106). The same ratio has been
detected in breast cancer microsomes.

18
 &KROHVWHURO

&<3$ O

HO
3UHJQHQRORQH

O
O β+6' &<3
OH

αK\GUR[\SUHJQHQRORQH
O HO
3URJHVWHURQH &<3 &<3 O

O
OH

HO
O 'HK\GURHSLDQGURVWHURQH'+($
O &<3
αK\GUR[\SURJHVWHURQH

O
$QGURVWHQHGLRQH

OH O
β−+6' &<3

O HO
7HVWRVWHURQH &<3 2HVWURQH(
OH β−+6'

HO
2HVWUDGLRO(

)LJXUH. Biosynthesis of oestradiol from cholesterol

19
A greater role in carcinogenesis has been suggested for 4-OHE2, which was
shown to be carcinogenic to the kidney in the Syrian golden hamster model while
2-OHE2 was not (107). In fact, the 2-hydroxylation pathway is thought to be
anticarcinogenic (122). Women with a larger proportion of oestrogen metabolism
via the 16α−hydroxylation pathway were earlier suggested to be at elevated risk
of breast cancer (123), but subsequent studies have not confirmed these findings
(124-126). Finnish women with high breast cancer risk have been shown to have
markedly higher urinary excretion of catechol oestrogens than 16α-hydroxylated
oestrogens compared to Orientals who have three-fold lower breast cancer risk
(124).

The 2- and 4-hydroxylated oestrogens can be inactivated by 2-methylation

catalysed by COMT (108,110). The observed lack of carcinogenic activity of the
2-hydroxyoestrogens may be due to faster rate of 2-methylation. Quantitatively,
the most active conjugation pathway for catechol oestrogens is methylation, but
they can also be conjugated by glucoronidation or sulphation (127).

Further metabolism of the catechol metabolites of oestrogen leads to formation of

semiquinones (CE-SQ) and quinones (CE-Q). CE-2,3Q can bind stably to DNA,
while CE-3,4Q forms depurinating adducts, which are lost from DNA by cleavage
of the glucosidic bond leaving apurinic sites (106,110). Consequently, the latter
event is hypothesised to cause tumour-initiation similarly to that seen with the
depurinating PAH-DNA adducts. In addition, redox cycling between CE-SQs and
CE-Qs produces ROS, which can cause oxidative damage to lipids and DNA
(10,113,119,128-130).

Quinones may be conjugated with glutathione (GSH) catalysed by GSTs or

reduced to catechol oestrogens by quinone reductase (110,131). The
semiquinones may react with molecular oxygen to form superoxide radicals,
which are reduced to hydrogen peroxide either spontaneously or catalysed by
superoxide dismutases such as MnSOD (132).

H2O2 is neutral and rather nonreactive except in the presence of the reduced
transition metal ions (LH Fe2+), which evoke the formation of the most powerful
oxidant, the hydroxyl radical (⋅OH). H2O2 can also readily cross the cellular and
nuclear membranes and reach the DNA in neighboring cells (133,134)

20
 OH

HO

2HVWUDGLRO(

&<3$ &<3%

OH OH

HO
2+( 2+( OH
HO HO
OH OH &207
&207

OMe
HO
0H2( 0H2(
OH OH OMe
HO

O

HO HO
(64 O (64

O2 Redox cycling O2
62' P450 oxidase/ 62'
H202 O2 
reductase O2 
H202
OH OH

*67 *67
*OXWDWKLRQH *OXWDWKLRQH
FRQMXJDWHV O
FRQMXJDWHV
(4 (4
O O
O

Stable adducts Depurinating adducts

Cancer initiation

)LJXUH Metabolism of oestrogen (Modified from [94]).

21
*HQHWLF SRO\PRUSKLVPV LQ RHVWURJHQ PHWDEROLVLQJ HQ]\PHV
DQGEUHDVWFDQFHU

Person-to-person differences in the conjugation pathways of both oestrogen

and catechol oestrogens may define subpopulations of women with higher
lifetime exposure to hormone-dependent growth promotion or to cellular
damage from particular oestrogens and oestrogen metabolites. Such variation
could explain a portion of the cancer susceptibility associated with
reproductive events and hormone exposure. The most widely studied
polymorphisms in low-penetrance genes encoding for enzymes with a
possible role in oestrogen biosynthesis and metabolism (Figure 4), and their
possible role in breast cancer susceptibility (Table 1) are reviewed below.

Free radical mediated damage to DNA, proteins, lipids

CHOLESTEROL
&<3 0Q62' Fe2+
O2 O2 
H202 OH


&<3 DNA damage

β+6' &<3$
&<3%
OESTRADIOL (E2) OH-E2 SEMIQUINONE QUINONE DNA adducts
*670
&207 *670
*673
*677
Inactivated metabolites Glutathione
conjugates

)LJXUH  A schematic presentation of enzymes with known polymorphisms

involved in oestrogen biosynthesis and metabolism.

22
7DEOH. Genetic polymorphisms in oestrogen synthesising and metabolising
enzymes and their possible role in susceptibility to breast cancer.

*HQH 5ROHLQRHVWURJHQ ([SUHVVLRQ 0XWDWLRQ )XQFWLRQDOHIIHFW 5HI

ELRV\QWKHVLV LQEUHDVW

PHWDEROLVP WLVVXH

&<3 Catalyses the rate- No Base change Creates Sp1-type (135)

limiting step in T->C in the 5’ untranslated promoter site and (136)
ovarian and adrenal region might therefore
biosynthesis increase
pathways for transcription
androstenedione

&<3 Converts Yes Polymorphic (TTTA)n repeat May alter splice site (137)
testosterone to in intron 5 close to (138)
oestradiol, and exon/intron border
androstenedione to
oestrone

β Converts oestrone Yes A–>G in exon 6 => None ? (139)

+6' to oestradiol Ser312Gly (140)
C–>T in intron 5 None

&<3$ 2-hydroxylation of E1 Yes $ : T –> C in 3' non-coding Increased activity ? (141)

and E2 region (0VS, (142)
&: A –> G in exon 7 Increased activity (143)
=>Ile462Val /inducibility (144)
: T–>C in 3' non-coding Altered activity ? (145)
region (Afr.-Am. only)
: C–>A in exon 7 Low activity against
=> Thr461Asp progesterone

&<3% 4-hydroxylation of E1 Yes : C–>G in exon 2 Altered activity ? (121)

and E2 (less 2- =>Ala48Gly (146)
hydroxylation) & G–>T Altered activity ? (147)
=>Ala119Ser (148)
: C–>G in exon 3 Increased activity ?
=>Leu432Val
 A–>G in exon 3 Increased activity?
=>Asn453Ser

&207 Methylation and Yes Val to Met 4-times reduced (149)

inactivation of (G–>A) change at position methylation activity (150)
catechol oestrogens 158/108 (151)

*670 Conjugation of Yes* deletion of the gene Lack of activity (152)

oestrogen quinones (153)
*677 to glutathione ? deletion of the gene Lack of activity (154)
(155)
*673 Yes Ile105Val Altered activity

*670 Yes* 3 bp deletion in intron 6 Generation of a

recognition site for
the YY1
transcription factor
and altered
transcription

0Q62' Converts two Yes Val to Ala amino acid May alter protein (89)
superoxide radicals change in -9 position of the targeting to (132)
to H2O2 and O2, signal sequence mitochondria
thereby reducing
oxidative stress

*Not known which of the mu class enzyme is present (156-159).

23
(Q]\PHVLQYROYHGLQRHVWURJHQELRV\QWKHVLV

&<3
The &<3 gene, located on chromosome 10q, encodes for the cytochrome
P450c17α enzyme, which in women is primarily expressed in ovary and
adrenal cortex (136,160). It mediates both steroid 17α-hydroxylase and 17,20-
lyase activities, and catalyses a rate limiting step in sex steroid synthesis
leading to the precursor, dehydroepiandrosterone (Figure 3).

The 5’ untranslated region of &<3 contains a single T to C base

substitution at position 1931 that creates an additional Sp-1-type (CCACC
box) promoter motif 34 bp upstream from the site of initiation of translation in
the variant allele ($ (161). The additional promoter site is thought to
influence gene expression resulting in increased enzyme activity leading to an
increased amount of bioavailable oestrogen (161). It was originally associated
with polycystic ovary syndrome risk (161), and later with breast cancer risk
(135).

Recent LQ YLWUR data suggest that the 5’ Sp-1-type site in CYP17 does not
actually bind the transcription factor (162). However, there is still some
evidence indicating that this polymorphism may influence endogenous steroid
hormone levels. Serum oestrogen and progesterone levels have been found
to be higher in premenopausal women with the &<3 $ (163), whereas
postmenopausal women with the $$ genotype had elevated levels of
oestrone and dehydroepiandrosterone (164). In addition, women with the
&<3 $$ genotype have been found to be half as likely to be current
users of hormone replacement therapy compared to women with two wild type
($$) alleles (165).

The studies completed so far on &<3 and breast cancer risk have given
controversial results (Table 2). The $ allele has been found to be
significantly associated with advanced breast cancer (135,140), male breast
cancer (166), and early-onset breast cancer (167,168). The latter association
has been found especially in women with a family history of the disease
(167,168). Furthermore, in two of the above mentioned studies, the reduced
risk of breast cancer, usually associated with a later age at menarche, was
observed to be limited to those women with the &<3 $$ genotype
(135,164). The other studies have not shown any significant association
between the &<3 genotypes and breast cancer risk (162,169-173).

24
7DEOH. Studies on &<3promoter polymorphism and their association with
breast cancer.

$XWKRUV\HDU FDVHFRQWURO $ &DVH 25&, $JHLQ\HDUV ,QWHUDFWLRQV

HWKQLFLW\ IUHT GHWDLOV IRU$$ FDVHFRQWURO

$$

Feigelson HW DO 174/285 0.42 All cases 1.32 (0.87-2.00) mean > 13 years
1997 (135)# 46 Asian 0.46 Advanced  63.0/61.4 at menarche
137 Afr.-Am. 0.37 and $$
102 Latino 0.44 25



Helzlsouer HW DO . 115/115 0.42 All 0.89 (0.41-1.95)* mean

1998 (170) US-Caucasian Pre$ 0.70 (0.15-3.19)* 60.4/60.2
Post§ 0.99 (0.39-2.49)*

Weston HW DO 123/240 0.39 All 1.08 (0.69-1.69) ?

1998 (172) 76/148 Cauc. 0.42 Caucasian 0.80 (0.45-1.43)
20/35 Afr.-Am. 0.31 Afr.-Amer. 1.40 (0.44-4.38)
27/57 Hispanic 0.34 Hispanic 1.93 (0.75-5.01)

Dunning HW DO 835/591 0.38 All 1.10 (0.89-1.37) <55/45-74

1998 (169) English Advanced 0.88 (0.38-2.01)

Haiman HW DO 464/619 0.40 All 0.85 (0.65-1.12) 43-69, > 13 years
1999 (164) American Pre 0.51 (0.24-1.06) mean 58.3 at menarche
Post 0.94 (0.70-1.25) and $$
25



Bergman- 109/117 0.31 Pre  22-36/18-39

Jungestrom HW Swedish
DO 1999 (167)

Young HW DO . 64/81 0.33 Male  26-91/?

1998 (166) 39/58 0.29 Female 0.56 (0.25-1.28) <50/?
Scottish

Spurdle HW DO 369/284 0.35 All  <40 with family
2000 (168) Australian history
25 



Nedelcheva- 510/201 0.38 not mentioned 27-91/

Kristensen HW DO Norwegian no significant 20-44
1999 (162) association

Huang HW DO . 150/150 0.53 All 1.28 (0.73-2.27) ?

1999 (171) Taiwanese pre 52/50 1.18 (0.50-2.83)
post 68/75 1.31 (0.59-2.90)

Feigelson HW DO . 615/1508 0.40 All 1.08 (0.81-1.44)* 64.6/62.8 with 17β-
2001 (140)# 536 Afr.-Am. Advanced 1.45 (0.96-2.20)* HSD.
400 Japanese
481 Latinos
402 White
# from the same cohort
* for $$

premenopausal
†
postmenopausal

25
&<3
The &<3 gene located on chromosome 15q encodes a steroid aromatase
that catalyses the conversion of C19 androgens to oestrogens (104). In
postmenopausal women, oestrogen is predominantly circulating as oestrone,
derived from the peripheral aromatization of androstenedione (114,116). The
aromatase activity has been found to be more strongly expressed in adipose
stromal cells surrounding mammary carcinoma cells than in stromal cells
distal to the tumour, indicating that enzymatic aromatization could be an
important source of oestrogen for mammary tumours (9,37).

A tetranucleotide repeat polymorphism (TTTA)n in &<3 gene, which may

alter the mRNA splicing site, has been studied in relation to breast cancer risk
(137,174,175); significantly increased risks have been found for carriers of the
longest 12 repeat variant among Norwegian and Swedish individuals (137),
for carriers of 10 repeat variant in mostly US Caucasian (176), and for carriers
of the shortest 7 repeat allele among US Caucasian (175). Although a recent
meta-analysis supported a role for (TTTA)10 as a breast cancer risk factor
(OR, 2.33; 95% CI, 1.36-4.17) (11), also negative results exist (174,177,178).

Another polymorphism in the &<3 gene giving rise to the amino acid
substitution Arg264Cys has also been studied in relation to breast cancer risk,
but appears to have no effect in this context (177,178).

β−+6'
17β−HSD is mapped on chromosome 17q in the vicinity of %5&$(179). It is
a key enzyme in oestrogen metabolism catalysing a reversible conversion of
oestrone to the more potent oestradiol, and is expressed in both normal and
malignant breast tissues.

There are several polymorphisms described in the β−+6' gene, two of

which have been studied in relation to breast cancer risk (Table 1) (139,140).
No association between breast cancer risk and either the common Ser312Gly
polymorphism (139,140), or the rare intron 5 C to T polymorphism (139) was
observed.

26
(Q]\PHVLQYROYHGLQRHVWURJHQPHWDEROLVP

&207
In humans, the COMT protein exists as two length variants,LH., as a soluble
form (S-COMT) and as a membrane-bound form (MB-COMT), which are both
encoded by a single gene localised in chromosome 22q (180). COMT is a
phase II enzyme that inactivates catechol oestrogens by transfer of a methyl
group. Most studies have shown that methylated catechol oestrogens have
little or no binding affinities for the classical oestrogen receptor, and that they
lack oestrogenic activity in the uterus (181). Moreover, an antitumourigenic
effect has been proposed for 2-methoxyoestradiol (122).

COMT is found in various mammalian tissues including breast (149,150), and

the level of the protein has been shown to be higher in breast tumour tissue
compared to normal mammary tissue of the same individual (150). Recently
the nuclear localisation of S-COMT was proposed to reflect a protective
response to increased local catechol load in women living in the industrialised
countries (182).

A single G to A base pair change in the &207/ allele results in a valine to

methionine amino acid change at codon 108/158 in the cytosolic/membrane-
bound form of the protein. This change has been associated with 3- to 4-fold
decreased activity of the COMT compared with the wild type &207+allele
(151). The &207 allele frequencies have recently been shown to vary
significantly between different ethnic groups (183). Nearly equal frequency of
the two alleles exists in Europe, while the &207+ allele is much more
common than the &207-/ allele in populations in all other parts of the world.

The few studies reported so far on &207 genotypes and breast cancer risk
have given discrepant results (Table 3) (171,184-187). In some studies, the
&207/ allele was associated with increased risk in premenopausal women,
and with decreased risk among postmenopausal women (187), while in other
studies the increased risk was mainly attributable to postmenopausal women
(171). No overall association has been found in other studies (184,185). In a
recent study, the &207/allele was found to be associated with progression
and lymph node metastasis of breast cancer in Japanese women (186).
Moreover, in some of the studies the risk of breast cancer has appeared to be
modified by BMI (184,187).

27
7DEOHStudies conducted on &207allele variants and breast cancer risk.

$XWKRUV FDVHFRQWURO &207/ &DVH 25&, $JHLQ ,QWHUDFWLRQV

\HDU HWKQLFLW\ IUHT GHWDLOV IRU&207// \HDUV

FDVH

FRQWURO

Lavigne HW 115/115 0.52 All 1.42 (0.69-2.94) 60.4/60.2 Pre with

DO 1997 US Caucasian Pre$ 0.24 (0.04-1.51) BMI >24.5
(184) Post§ 2.18 (0.93-5.11) 25



Millikan HW 278/271Afr.-Am. 0.35 Afr.-Am. 0.8 (0.4-1.5) 20-74

DO1998 410/392 White 0.53 White 0.7 (0.5-1.1)
(185) Pre all 0.7 (0.4-1.2)
Post all 0.8 (0.5-1.4)

Thompson 281/289 0.49 All 0.8 (0.5-1.4) Pre with

HWDO 1998 US Caucasian 0.43 Pre  46.2/46.8 BMI>27
(187) 0.54 Post  62.9/ 63.3 25



Post with
BMI < 23
25



Huang HWDO. 150/150 0.25 All  *

1999 (171) Taiwanese 0.23 Pre 2.00 (0.46-10.36) *

0.27 Post  *

Matsui HWDO. 140 cases only 0.34 not mentioned 52.3 / associated

2000 (186) with lymph node

metastasis

* OR shown for combination of &207/ and &207// genotypes.

$
premenopausal
§
postmenopausal

&<3V
&<3$ has been mostly studied as the principal enzyme activating cigarette
smoke constituents and other environmental pollutants such as PAHs leading
to electrophilic, carcinogenic molecules (57). However, it also catalyses the 2-
hydroxylation of oestradiol in several extrahepatic tissues including breast
(Figure 3) (138,147,188).

To date four mutations have been identified in the &<3$ gene (Table 1)
(189,148). The first described mutation, located downstream from the
polyadenylation site (&<3$ $) (141), and the mutation in the heme-
binding region of the gene, leading to an amino acid change from isoleucine
to valine at codon 462 (&<3$ &), are strictly linked in Caucasians (142).
These mutations have been suggested to lead to higher enzyme activity and
therefore higher rates of carcinogen activation (190). Furthermore, an African-
American specific mutation has been found in the 3´noncoding region
upstream from the polyadenylation site (&<3$ ) (143), and a C to A
substitution in H[RQ leading to a threonine to aspartic acid change at codon
461 (&<3$ ) (144). The latter change has been reported to cause lower
catalytic activity against progesterone (145).

28
The frequencies of the &<3$ $and &<3$ &alleles vary considerably
by race, being very rare in the Caucasians but considerably more common in
Asians. While no significant overall association has been found in the
Caucasian population for any of the variant alleles (191-196), significant
associations with breast cancer risk have been found for &<3$ $ and
&<3$ &allelesamong smokers (192,193), and for the &<3$ &allele
among women with above median serum PCB levels (191). Furthermore, the
&<3$ $ allele has been associated with breast cancer among Chinese
(197) and African-American women (196).

CYP1B1, the gene of which is located at 2p, appears to be the main CYP450
enzyme responsible for the 4-hydroxylation of oestradiol (Figure 4) (120,121),
but also activates many PAHs and arylamines (198). As 4-hydroxylated
metabolites represent only a minor portion of total urine oestrogens, it was
previously considered as merely a minor route for metabolism. However,
CYP1B1 is present in virtually all adult and fetal tissues with high levels found
in breast (138,199). Moreover, since it is co-localised with CYP19, the
enzyme producing oestrogen, this could lead to high local production of
potentially carcinogenic oestrogen metabolites (9).

Several polymorphisms have been described in the &<3% gene (146). Four
of these result in amino acid substitutions, as shown in Table1 (121,146). The
Val432Leu polymorphism has recently been found to have the most profound
impact on the catalytic properties of CYP1B1, with the 4-hydroxylase activity
of the Val432 allele displaying four-fold higher activity compared to Leu432 allele
while other mutations did not have any obvious effect on the enzyme activity
(200). However, contrasting results have also been observed (201,202).

The &<3% Val/Val genotype has been associated with oestrogen and
progesterone receptor positive breast cancer in Caucasian women (203),
while Chinese women carrying the Leu/Leu genotype were at increased risk
for postmenopausal breast cancer (204). In contrast, in Japanese women this
polymorphism had no effect, whereas the Ala119Ser change was associated
with increased breast cancer risk (205). No association was found between
the Asn453Ser polymorphism and breast cancer risk (203).

29
*67V
Members of a superfamily of cytosolic GSTs catalyse the conjugation of
glutathione (GSH) to a variety of endogenous and exogenous electrophiles
(206). The incorporation of glutathione increases the molecule’s water
solubility and excretability (207). The most studied substrates for GSTs are
carcinogenic PAH diolepoxides, but they also play a critical protecting role
against oxidative stress (208). In oestrogen metabolism, GSTs play a role in
the catalysis of GSH conjugation of catechol oestrogen quinones, the reactive
intermediates of oestrogen metabolism capable of bind to DNA (127). GSTs
may also be involved in the activation of some carcinogens such as
halogenated hydrocarbons (148,209).

Human tissues show differential expression of the multiple forms of GSTs

(210). The absence of specific isoenzymes affects the tolerance of organisms
to chemical challenges and may result in an increased rate of somatic
mutations and higher susceptibility to disease. The ability of many tumours to
exhibit increased levels of intracellular GST expression has been linked to
mechanisms of chemotherapeutic drug resistance (211).

Cytosolic GSTs are dimeric proteins and they can be differentiated into five
classes in mammalian cells, termed alpha (A), mu (M), pi (P), theta (T), and
zeta (Z). Allelic variation has been found in genes encoding for these GSTs
(209,212,213). The active site of a GST molecule is composed of two
subunits, the glutathione binding site and the hydrophobic substrate binding
site. The residues forming the glutathione binding site are conserved in the
different classes, while those forming the substrate binding site vary
considerably, leading to a wide substrate specificity (210). There are no
reports regarding the specific GST isoforms in the formation of CE-Q-GSH
conjugates or detoxification of reactive oxidant damage.

The two exclusively membrane-bound proteins, microsomal GST and

leukotriene C4 synthase are structurally unrelated to the cytosolic GSTs (214).

GSTM1
GSTM1 is located in the middle of a cluster of five mu class genes on
chromosome 1p (215). The homozygous deletion (null genotype) of the
*670 gene leads to the total absence of the respective enzyme activity
(152). The elimination of the gene has been proposed to have occurred by
unequal crossing over (215). The frequency of the null genotype is around
50% in Caucasians and Asians, but only 27% in Africans (216). Allelic
variants of *670, LH, A and B have also been found to exist, but based on
current knowledge they have no consequences in the catalytic activity of the
enzyme (217). In addition, a gene duplication has been found to exist in
Saudi-Arabians (218).

The *670 genotype has been examined in relation to individual breast

cancer risk in several recent studies (Table 4), some of which pointed to an
association between *670 null genotype and breast cancer risk in

30
postmenopausal women (219,220). Most of the studies have found no
association (192,194,221-224). An increased risk for premenopausal women
has also been shown, but only in one study (225).

7DEOHStudies on *670 null genotype and breast cancer.

$XWKRUV FDVHFRQWURO &DVHGHWDLOV *670 25&, $JHLQ ,QWHUDFWLRQV

\HDU HWKQLFLW\ QXOOIUHT IRU*670QXOO \HDUV

FDVH

FRQWURO

ZhongHWDO 197/225 0.42 0.79 (?) ?

1993 (223) Caucasian

Ambrosone 216/282 Post * 0.50 1.10 (0.73-1.64) ?

HWDO 1995

(192)
Kelsey HWDO 245/245 Prevalent cases 0.51 1.30 (0.91-1.86) Age
1997 (222) 240/240 Incident cases 0.48 1.08 (0.74-1.57) matched
US Caucausian Pre § 0.83 (0.25-2.73)
Post 1.11 (0.71-1.76)

Bailey HWDO 164/164 US All Caucasian 0.62 0.77 (0.5-1.2)

1998 (194) Caucasian All Afr.-Am. 0.41 0.75 (0.35-1.58)
59/59 Afr.-Am.

Helzlsouer 115/115 All 0.46  mean post with

HWDO 1998 Pre 0.64 1.00 (0.29-3.45) 60.4/60.2 BMI >24.47
(220) Post 0.41  25



CharrierHW 361/437 0.51 range

DO1999 French < 50 yrs 0.54 1.03 (0.74-1.45) 26-80/
(219,230) > 50 yrs 0.43  18-101,
mean
51/47

Ambrosone 279/340 Pre 0.52 0.9 (0.4-2.0) ?

HWDO1999 Post 0.50 1.1 (0.7-1.6)
(221)
Garcia- 466/466 All 0.49 1.05 (0.80-1.37) mean
ClosasHWDO US Caucasian Pre 0.52 0.92 (0.42-2.02) 58/58
1999 (231) Post 0.49 1.06 (0.78-1.45)

Curran HWDO 129/129 ? 0.56 not significant, Age

2000 (232) Australian OR not matched
Caucasian mentioned
Park HWDO 189/189 All 0.53 1.3 (0.84-2.06) Age alcohol
2000 (225) Korean Pre 0.50  matched users
Post 0.58 0.9 (0.45-1.93) 29-70+ 25



Millikan HW 278/271 Afr.-Am. All Afr.-Am. 0.28 0.9 (0.6-1.3) 20-74 With family
2000
DO 410/392 US All US whites 0.52 0.9 (0.7-1.2) history
(224) whites Pre all 0.45 0.8 (0.6-1.1) 25

Post all 0.39 1.0 (0.8-1.4) 

Smoking >
20 yrs
25



* postmenopausal
§
premenopausal

31
Despite the above mentioned discrepant findings, a recent meta-analysis
suggested that the *670 null genotype poses a moderately increased risk
for postmenopausal breast cancer (OR, 1.33; 95% CI, 1.01-1.73) (11).
Furthermore, the risk has been shown to be modified by BMI (220), family
history (224), use of alcohol (225), and smoking (224).

The *670 null genotype was not found to be associated with tumour
characteristics or survival in one study (226) but has been suggested to be
associated with both longer (222) and shorter (227) survival times in other
reports. In the study showing a reduced suvival time, the concurrent presence
of both the *670 null and *677 null genotypes was associated with
positive lymph node status. A strong association has also been shown
between *670 deletion and increased PAH-DNA adduct levels in breast
tumour tissue (228).

GSTM3
The *670 is another member of the mu family genes located on
chromosome 1p (229). It shares 70% of amino acid sequence identity but no
similarity outside the protein coding region with the other class mu proteins
(229). The *670 gene has two alleles, *670 $ and *670 %, which
differ from each other by a three base pair deletion in intron 6. This mutation
in the noncoding region of the gene is believed to generate a recognition
sequence for the YY1 transcription factor in the latter allele (155). YY1
regulates gene expression from intragenic sites and influences the
transcription of many genes. This suggests that *670 $ and % alleles may
be expressed at different levels with different efficiencies in the metabolism of
harmful agents (233). No studies have yet been reported on *670 genotype
and breast cancer risk.

The mu class GSTs have been shown to be expressed both in normal and
malignant breast tissue, but it is not known which of the mu class enzymes is
present (156-159). The expression of the GSTM1 and GSTM3 enzymes
appears to be co-ordinated (234), and a linkage disequilibrium has been
noted between the *670 $ and *670 % alleles (155,235). Moreover,
variation in GSTM3 expression has been seen in individuals with different
*670genotypes (234,236).

GSTP1
GSTP1 is the major GST expressed consistently in both normal and tumour
breast tissue (157-159). The *673gene has been mapped to a small region
of chromosome 11q. Two variant alleles, *673 % and &, have been
detected in the *673 gene addition to the wild-type allele *673 $ (154).
In both variant alleles, a point mutation at nucleotide 313 results in a single
amino acid change from isoleucine (Ile) to valine (Val), at codon 105. In
addition to this base substitution, *673 & has another point mutation
yielding an alanine (Ala) to valine change at codon 114. Both of the affected

32
codons lie in close proximity to the hydrophobic binding site. The Val105 variant
has been demonstrated to have either lower or higher specific activity and
affinity depending on the substrate (154,237). In contrast, the Ala114Val
polymorphism has not been described to influence the enzyme activity.

Around 5-10% of Caucasian have been shown to be homozygous for the Val
allele (220,224,238,239). Women with *673 Val/Val genotype have been
shown to have better survival, probably due to lower activity in the metabolism
of chemotherapeutics (240).

Two recent studies revealed no significant association between the *673

Ile105Val polymorphism and breast cancer proneness (232, 238), while one
study suggested a trend for increasing risk with higher number of *673
Val105 alleles (220). In a recent meta-analysis, the Val105 allele carriers were
also shown to be at increased risk (OR, 1.6, 95% CI, 1.08-2.39), while in the
largest study conducted to date on this topic, a significant increase in the risk
was found for former smokers for the homozygous *673 Ile105 allele
containing genotype (224) (Table 5).

In the sole study conducted to date on the Ala114Val polymorphism and breast
cancer, the *673 &allele was reported to have a protective role (241).

7DEOH Studies on *673genotype and breast cancer.

$XWKRUV\HDU FDVHFRQWURO &DVHGHWDLOV *673 25&, $JH ,QWHUDFWLRQV

HWKQLFLW\ 9DOIUHT IRU9DO9DO LQ\HDUV

FDVHFRQWURO

Harries HWDO. 62/155 ? 0.28 1.3 (0.4-4.3)* ?/60.3

1997 (238) Caucasian

HelzlsouerHW 115/115 All 0.29 1.97 (0.77-5.02) 60.4/60.2

1998 (220)
DO US Caucasian Pre § 0.38 0.54 (0.04-6.67)
Post $ 0.27 2.71 (0.91-8.03)

Curran HWDO 129/129 0.29 not mentioned age matched

2000 (232) Australian no association
Caucasian

Millikan HWDO 278/271 All Afr.-Am. 0.51 0.9 (0.5-1.4) 20-74 Former
2000 (224) Afr.-Am. All whites 0.35 0.7 (0.4-1.2) smokers
410/392 Pre all 0.40 0.9 (0.5-1.5) with Ile/Ile
US whites Post all 0.43 0.7 (0.4-1.1) 25



*for Ile/Val and Val/ Val combined

§
premenopausal
$
postmenopausal

33
GSTT1
The human *677 gene is localised on chromosome 22q with the *677
gene and a pseudogene (153,242). Similarly to *670, homozygous deletion
of the *677 gene, leading to total absence of the respective enzyme
activity, has been observed. Large inter-ethnic differences have been reported
in the frequencies of the *677 null genotype being significantly lower among
Caucasians (20%) compared to Asians (60%) (243). The frequency of the
*677 null genotype is only about 13% among Finns (239), similar to findings
from other Scandinavian study populations (216).

GSTT1 is expressed in human erythrocytes, and various tissues including

liver but no expression in breast tissue has been reported (153), (244).
In most of the studies on *677 genotype and breast cancer risk, no
significant association has been found (Table 6) (194,220,224,225,232).
However, in one study the risk was found to be modified by the use of alcohol
(220), and in another study a remarkably lower risk was suggested for
premenopausal women lacking the *677 gene (231).

7DEOHStudies on *677genotype and breast cancer.

$XWKRUV FDVHFRQWURO &DVHGHWDLOV *677 25&, $JHLQ\HDUV ,QWHUDFWLRQV

\HDU HWKQLFLW\ QXOO IRU*677QXOO

IUHT FDVHFRQWURO

BaileyHWDO. 223/223
1998 (194) Caucasian 0.27 1.08 (0.66-1.75)
328 0.29 0.63 (0.27-1.47)
Afr.-Am. 118
Helzlsouer 115/115 All 0.21 1.50 (0.76-2.95) 60.4/60.2 Alcohol
HWDO1998 Pre * 0.20 1.50 (0.42-5.32) 25

(220) Post § 0.22 1.50 (0.67-3.34) 

Garcia- 466/466 All 0.17 0.86 (0.61-1.21) 58/58

Closas HWDO. US Pre 0.21 

1999 (231) Caucasian Post 0.17 0.90 (0.59-1.35)

Curran HWDO 129/129 0.16 not mentioned

2000 (232) Australian not significant
Caucasian

Park HWDO 189/189 All 0.42 1.6 (0.98-2.54) Age matched Together
2000 (225) Korean Pre 0.43 1.7 (0.94-3.24) with the
Post 0.40 1.3 (0.64-2.80) *670 null
25



Millikan HW 278/271 Afr.-Am. 0.17 1.3 (0.8-2.0) 20-74

DO 2000 Afr.-Am. US Caucasian 0.16 0.8 (0.6-1.2)
(224) 410/392 Pre 1.0 (0.7-1.6)
US Post 1.0 (0.7-1.5)
Caucasian
* premenopausal
§
postmenopausal

34
0Q62'
The semiquinones produced in the metabolism of oestrogen may react with
molecular oxygen to form ROS. ROS are also generated in the metabolism of
xenobiotics, polyunsaturated fatty acids, ethanol, and tobacco smoke.
Superoxide dismutases (SODs) comprise a family of metalloenzymes that
catalyse the conversion of superoxide anion (O2-) into hydrogen peroxide
(H2O2) and molecular oxygen (O2), and thus protect cells from damage
induced by free radicals (132). There are three known forms of SOD; the
cytosolic CuZnSOD, the extracellular CuZnSOD, and mitochondrial MnSOD.
MnSOD differs structurally from CuZnSOD in that it forms a homotetramer
instead of a homodimer (245).

The human MnSOD,thegene of which has been localised to 6q25 (246), is

synthesised in the cytosol as a larger precursor and posttranscriptionally
modified for transport into the mitochondria (247). Import of precursor proteins
involves several steps including binding of precursor to mitochodria,
translocation across mitochondrial membranes, and proteolytic cleavage of 1-
terminal leader peptides (248).

MnSOD mRNA expression has been shown to be increased in response to

exposure to hydrogen peroxide (249). A variety of cancer cells are known to
have reduced levels of antioxidant enzymes, especially MnSOD, when
compared to their normal counterparts (250). Furthermore, the increased
expression of MnSOD has been found to suppress the malignant phenotype
of human breast cancer cells, suggesting that itsgene is a tumour suppressor
gene in human breast cancer (251).

A one base pair transition from T to C in 0Q62'gene leads to an amino acid

change from Val to Ala at –9 position of the mitochondrial targeting sequence
(MTS) of the MnSOD protein. This change encoded by the 0Q62' $ allele is
predicted to form a typical amphiphilic helical structure, which is crucial for
effective transport and processing of mitochondrial proteins (252). It has been
suggested that presence of the0Q62'9 allele in this position could result in
mistargeting of the respective protein, thereby leading to a deficiency of this
enzyme within mitochondria (253). There appears to be significant ethnic
differences in the 0Q62'allele frequencies; the variant allele is much rarer in
Asians than in Caucasians. In Finns, the frequency of the 0Q62'$ allele is
about 48% (254).

Although another polymorphism, LH, Ile to Thr amino acid change at codon
58, has been shown to result in lower enzymatic activity due to destabilisation
of the tetrameric structure (245,255), its frequency seems too low to have any
detectable effect on cancer risk (256).

35
In the sole study performed to date on the 0Q62' genotypes and breast
cancer risk, the 0Q62' $$ genotype was found to increase the risk,
especially among those premenopausal women with a low consumption of
dietary sources of antioxidants (89).

36
$,062)7+(678'<
The main objective of this thesis was to examine the role of polymorphic
metabolic genes in individual susceptibility to sporadic breast cancer in the
Finnish population. The genes involved in oestrogen synthesis and
metabolism were of particular interest.

The specific aims were:

• To examine if allelic variants in an oestrogen synthesising gene, the

&<3, and in the oestrogen metabolising gene, &207, could affect the
risk of sporadic breast cancer, especially among individuals exposed to
exo- and endogenous oestrogen.

• To examine if the genetic polymorphism in the mitochondrial targeting

sequence of the endogenous antioxidant coding gene, 0Q62' could
modulate individual susceptibility to sporadic breast cancer, especially
among individuals with high exposure to reactive oxygen species.

• To examine if the genetic polymorphisms of*670*670*673and

*677genes, taking part in detoxification of reactive intermediates, could
influence individual susceptibility to sporadic breast cancer.

• To examine the potential combined effects of the polymorphic Restrogen

metabolising enzyme genotypes in the individual breast cancer risk.

37
68%-(&76$1'0(7+2'6

6WXG\VXEMHFWV

&DVHV
This study was an extension of Kuopio Breast Cancer Study, a case-control
study that follows the protocol of the International Collaborative Study of
Breast and Colorectal Cancer co-ordinated by the European Institute of
Oncology in Milan. All women with a suspect breast lump or breast symptoms
and living in the study catchment area were invited to Kuopio University
Hospital from 1990 to 1995. Altogether 1,919 women visited the hospital
during the recruitment period. Of these 516 were finally diagnosed with
histologically confirmed breast cancer, 717 had benign breast disease, and
686 were diagnosed as being healthy. All women were asked to participate in
the study at the first hospital examination before neither the patient nor the
interviewer knew the diagnosis. Only patients with breast cancer diagnosis
were included in the studies presented in this thesis.

The recruitment protocol missed 51 women within the hospital, all of whom
were private patients who did not enter the hospital through the standard
procedure. According to the comparison with the Finnish Cancer Registry,
only 26 breast cancer cases were treated elsewhere. The contact rate for
cases was therefore 86%. Only twelve women later diagnosed with breast
cancer refused to participate in the study. Since all interviewed women agreed
to donate blood sample, the cooperation rate was 98% and the overall
response rate was therefore 84%. Lymphocyte DNA of sufficient quality was
available for 483 cases (age range from 44.3 to 91.6, mean age 58.9 years).

The malignant breast tumours were categorised on the basis of UICC

Tumour-Node-Metastasis classification (257). Classification of tumours was
possible for 473 case patients for whom full histopathology was available.
Women who had axillary lymph node positive disease or metastatic breast
cancer at diagnosis were considered as advanced cases (Table 7). The cases
diagnosed with a tumour confined to the breast, either in situ or invasive were
designated as local. This categorisation could not be performed for ten
patients because of missing information on lymph node involvement.

38
7DEOH  Breast cancer patients according to tumour stage and receptor
status.

Q
All cases 483 (100)
a
Cases by stage :
In situ
Invasive
40 (8.2)
253 (52.1)
〉 Local cases
Lymph node positive
Metastatic (stage IV)
164 (33.7)
16 (3.3)
〉 Advanced cases
Cases by receptor status:
ER- 89 (24.3)
ER+ 278 (75.7)
PR- 142 (38.9)
PR+ 223 (61.1)
a
Total number does not correspond because the stage of the disease could not be classified for 10
patients due to missing information on lymph node involvement.

&RQWUROV
Healthy population control subjects were drawn from the Finnish National
Population register. They were living in the study area and were originally
individually matched with cases by age (+ 5 years). However, eventually also
some over 77 years old cases without matched controls were included in the
study. In all, 514 controls were interviewed in parallel with the cases. Data on
the exact contact rate of the controls is not available. However, due to the
accurate individual social security number system used in Finland, the
majority of the eligible controls are thought to have been contacted. The
cooperation rate of the controls was 72%; the major reason for non-
participation among controls was refusal. Of these, lymphocyte DNA was still
available for 492 subjects. Four population controls who earlier had had a
breast cancer diagnosis were excluded from the study, as well as four
controls for whom genotype data could not be obtained. Two controls were
also excluded because of their non-Finnish origin Therefore, our final control
population contained 482 healthy population controls (age range from 37.5 to
77.2, mean 53.5 years).

The Joint Ethics Committee of Kuopio University and Kuopio University

Hospital approved the study. Participation was based on written consent.

39
0HWKRGV

*HQRW\SLQJDQDO\VHV
Lymphocyte DNA, extracted by standard techniques, was used as a template
in the PCR-based genotyping analyses (Figure 5) for all studied genes.

*HQRW\SHVGHWHUPLQHG

DNA PCR Gel electrophoresis *670 present/null

*677 present/null

Restriction enzyme digestion: Gel electrophoresis

0QOI *670 $$$%%%

6QD%I  *673 Ile/Ile, Ile/Val,Val/Val

0VS$I  &<3$$$$$$

1ODIII  &207+++///

1JR0,9  0Q62'$$$999

)LJXUH Schematic presentation of genotyping methods used.

All the PCR reactions were carried out in a total volume of 30 µL containing
50-100 ng of lymphocyte DNA, 0.5 µM of each specific primer, 0.2 mM of each
dNTP, 1.5-3.0 mM MgCl2, 50 mM KCl, 10 mM Tris-HCl (pH 9.0), 0.1 % Triton
X-100, and 0.5-0.75 units of Taq polymerase. After initial denaturation at 94°
C, 30-40 cycles of amplification were performed in a Perkin-Elmer 9600
thermal cycler.

All genotyping analyses were performed blinded to the case-control status.

Samples with known genotypes were used as controls in each PCR and
RFLP batch to ensure proper amplification and enzyme digestion,
respectively. All samples with an ambiguous result were reanalysed.
Additionally, a random sample of 10 % was subjected to repeated analysis to
ensure laboratory quality control.

&<3
In the &<3 genotyping analysis, specific primers described in (163) were
used to produce a 431 bp long PCR product. Aliquots of the amplification
product were subsequently digested for 2 h at 37°C using 0VSAI (Promega,
WI, USA), which recognises the base change in $ allele resulting in 309 bp
and 122 bp fragments. The uncut PCR product indicated the presence of wild
type $ alleles.

40
&207
The &207 genotype was determined essentially as described earlier (258).
Briefly, the 217 bp PCR products were digested using the 1ODIII enzyme (New
England BioLabs, MA, USA). The presence of an additional cleavage site
resulted in four digestion products (96, 83, 18, 20 bp) differentiating the
variant &207/ allele from the wild type &207+ allele which produced only
three digestion fragments (114, 83, 18 bp).

*670 *677
A multiplex PCR method was used to detect the presence or absence of the
*670 and *677 genes as described earlier (259,260). Briefly, the *67-
specific primer pairs were used together with a third primer pair for β-globin
(268 bp). The absence of the *670 (136 bp) or *677-specific fragment
(111 bp) indicated the corresponding null genotypes whereas the β-globin
specific fragment (268 bp) confirmed the presence of amplifiable DNA in the
reaction mixture.

*673
The *673 % allele containing a base substitution at the nucleotide 313 was
recognised using primers, which create a restriction site for 6QDBI (New
England BioLabs) restriction enzyme (239); the 173 bp long product is
digested to 142 bp and 31 bp fragments, while the wild-type allele (*673 $)
remains uncut. Since the method used did not differentiate between *673 %
and *673 & alleles, the Val105 alleles are designated as *673 Val
throughout this thesis.

*670
In the *670 analysis the PCR-product was digested with 0QOI (New
England BioLabs) restriction enzyme (155); the wild type *670 $allele was
cut to 125 bp, 86 bp and 51 bp fragments, while the presence of 134 bp and
125 bp fragments indicated the *670 %allele.

0Q62'
In the 0Q62'analysis, the 107 bp PCR product was amplified using specific
primers (89), which create a restriction cut site for 1JR0 IV (New England
Biolabs) in the Ala-allele (0Q62' $). The resulting 89 bp and 18 bp
fragments differentiated the Ala-allele from an undigested 107 bp Val-allele
(0Q62'9).

41
6WDWLVWLFDODQDO\VHV
Associations between polymorphic genes and development of breast cancer
were evaluated by unconditional logistic regression to calculate multivariate
adjusted odds ratios (ORs) and 95% confidence intervals (CIs) (Statistical
Package for Social Sciences, version 9.0). Age (5 year intervals), age at
menarche (<12, 13-14, >15 years), age at first full-term pregnancy
(nulliparous, <25, 25-30, >30 years), number of full term pregnancies
(continuous), first-degree (mother, sister, daughter) family history of breast
cancer (no/yes), and history of benign breast diseases (no/yes) were used as
adjusting variables in all analyses except in Paper III where the adjustment
was done for age only. When indicated, additionally used covariates were: 1)
use of oral contraceptives (never/ever), 2) use of HRT (never/ever), 3) WHR
(<0.91, >0.91), 4) BMI (<25.4 kg/m2, >25.4 kg/m2), 5) education (low,
medium, high), 6) current alcohol intake (never, less than once a month,
monthly-daily), and 7) smoking habits (never/ex-/current smoker). Subjects
with a missing value in any of the variables in the regression model were
excluded from the analysis.

All analyses were performed separately for pre- and postmenopausal women.
Hysterectomy leads to cessation of menstruation without hormonal change
while oophorectomy also terminates the production of oestrogen. Therefore
women who had undergone hysterectomy without ovary removal (40 cases
and 40 controls) were classified as premenopausal unless they were older
than 51 years (median for menopause in Finnish women) and were no longer
menstruating. The same criteria was used for women if the operation
performed was unknown (6 cases and 2 controls). All of the other women
were classified as premenopausal.

Subjects who had ever smoked daily for at least three months were
considered as smokers. Those who reported smoking at the reference date
were considered as current smokers. Ever smokers (ex- or current) were
further categorised to two levels of smoking exposure according to the
approximate median value for smoking years (15 years) and daily
consumption (10 cigarettes) in the control population.

Similar categorisation was performed for the total months of using oral
contraceptives (36 months) or postmenopausal HRT (30 months). We do not
have individual data on the type of hormone products used. However, at the
time of the study the use of combined oestrogen-progestogen preparations
were by far most commonly used for the HRT in Finland (44). Women who
reported ever taking vitamin supplements (C-, E-, or A-vitamin) at least 3
times per week for longer than 6 months were considered as users of vitamin
supplements (Paper III). WHR, BMI, age at first full term pregnancy (FFTP),
and age at menarche, were dichotomised based on the median values for
population controls.

Possible gene-environment interactions were evaluated using stratified

analyses. Tests for interaction were assessed by the likelihood ratio test to
compare goodness of fit of the model with the interaction term (genotype by

42
stratified variable) to that of the model including the main effect variable
(genotype), stratified variable and main adjusting variables in the reduced
model (261).

Based on data from previous reports, individuals having the *670 and
*677 genes, and those carrying the&<3$$&207++*670$$
*673,OH,OHand 0Q62'99genotypes, served as a reference category for
all separate analyses of these genotypes.

All reported S-values were two-sided. 3-value less than 0.05 was considered
as significant.

43
5(68/76

&KDUDFWHULVWLFVRIWKHVWXG\SRSXODWLRQV

The study included older postmenopausal cases (66.5 years) compared to

controls (60.7 years) (Table 8). Both in pre- and postmenopausal women, the
risk of breast cancer was around 2-fold lower for parous women and for
women with earlier age at FFTP, compared to the nulliparous women (Table
9). In contrast, the 1.6-fold increase in the risk among women with WHR over
the median value, and the 3.8-fold increase in women with a family history of
breast cancer, were confined to the premenopausal women. Moreover, when
pre- and postmenopausal women were considered together, a history of
benign breast diseases was associated with increased risk (OR, 1.3; 95% CI,
1.0-1.8), and use of oral contraceptives with decreased risk (OR, 0.7; 95% CI,
0.5-0.9). Age at menarche and age at FFTP were similar among both pre- and
postmenopausal women (Table 8).

As expected, higher number of premenopausal women (~70%) had ever used

oral contraceptives compared to postmenopausal women (~25%) (Table 9).
Furthermore, on average premenopausal cases had used oral contraceptives
somewhat longer (61.5 months) compared to controls (50.2 months), but this
difference was not statistically significant. In contrast, postmenopausal cases
had used hormone replacement therapy significantly longer compared to
controls (61.4 months YHUVXV 45.7 months, respectively, S=0.005) (Table 8).

Smoking (22%) and use of alcohol (26%) were clearly more common among
premenopausal women compared to postmenopausal women (8% and 12%,
respectively) (Table 9), but no significant differences were seen between
cases and controls (Tables 8 and 9).

7DEOH  The mean values (SD) for selected characteristics of the study
population.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
&DVHV &RQWUROV S &DVHV &RQWUROV S

Age, years 44.4 (6.32) 43.6 (5.90) 0.240 66.5 (11.1) 60.7 (7.50) 

Age at menarche, years 13.2 (1.39) 13.4 (1.40) 0.262 14.1 (1.65) 13.8 (1.53) 0.056
Age at FFTP, years 24.9 (4.39) 25.1 (4.26) 0.762 24.9 (4.71) 24.3 (3.98) 0.124
Duration of OC, months 61.5 (60.8) 50.2 (48.5) 0.099 39.6 (44.2) 38.3 (41.1) 0.860
Duration of HRT, months 61.4 (62.8) 45.7 (51.3) 0.005
Daily tobacco consumption 10.5 (6.7) 9.8 (7.1) 0.541 12.0 (11.7) 8.9 (6.8) 0.107
Duration of smoking, years 14.5 (8.54) 12.3 (7.80) 0.126 17.9 (12.7) 17.8 (13.7) 0.963

44
 7DEOHAssociations between selected characteristics of the study subjectsa
and breast cancer risk stratified by menopausal status.

3UHPHQRSDXVDOZRPHQ 3RVWPHQRSDXVDOZRPHQ
&DVHV &RQWUROV &DVHV &RQWUROV

n % n % ORb (95% CI) n % n % ORb (95% CI)

Age at menarche
< 12 49 (29.9) 45 (22.1) 1.0 49 (16.2) 56 (20.4) 1.0
13-14 85 (51.8) 125 (61.3) 0.64 (0.39-1.05) 134 (44.2) 126 (45.8) 1.08 (0.67-1.74)
> 15 30 (18.3) 34 (16.7) 0.84 (0.44-1.60) 120 (39.6) 93 (33.8) 1.21 (0.74-2.00)

Age at FFTP
Nulliparous 30 (18.3) 20 (9.8) 1.0 72 (22.8) 37 (13.3) 1.0
” 83 (50.6) 104 (51.0)  154 (48.7) 159 (57.2) 
26-30 34 (20.7) 59 (28.9)  60 (19.0) 63 (22.7) 
> 31 17 (10.4) 21 (10.3) 0.51 (0.22-1.22) 30 (9.5) 19 (6.8) 0.71 (0.34-1.52)

Number of FTP
Nulliparous 30 (18.3) 20 (9.8) 1.0 72 (22.6) 37 (13.3) 1.0
1 25 (15.2) 33 (16.2) 0.50 (0.23-1.09) 43 (13.5) 31 (11.2) 0.68 (0.35-1.31)
2 59 (36.0) 93 (45.6)  82 (25.8) 88 (31.7) 
3+ 50 (30.5) 58 (28.4) 0.53 (0.26-1.06) 121 (38.1) 122 (43.9) 

Use of OC
Never 56 (34.1) 52 (25.5) 1.0 257 (82.1) 191 (68.7) 1.0
Ever 108 (65.9) 152 (74.5) 0.67 (0.42-1.07) 56 (17.9) 87 (31.3) 0.67 (0.42-1.05)

Postmenopausal HRT use

Never 221 (71.1) 164 (59.4) 1.0
1-30 months 34 (10.9) 57 (20.7) 0.64 (0.39-1.05)
>30 months 56 (18.0) 55 (19.9) 1.10 (0.71-1.73)

WHR
” 71 (43.6) 113 (55.9) 1.0 116 (36.8) 123 (44.4) 1.0
> 0.91 92 (56.4) 89 (44.1)  199 (63.2) 154 (55.6) 1.22 (0.86-1.73)

BMI
” 100 (61.0) 126 (62.4) 1.0 108 (35.5) 114 (41.2) 1.0
> 25.4 64 (39.0) 76 (37.6) 0.99 (0.64-1.55) 196 (64.5) 163 (58.8) 1.40 (0.97-2.01)

First-degree family history of breast cancer

No 143 (87.2) 197 (96.6) 1.0 281 (89.5) 262 (94.6) 1.0
Yes 21 (12.8) 7 (3.4)  33 (10.5) 15 (5.4) 1.82 (0.92-3.59)

History of BBD
No 92 (56.1) 129 (63.2) 1.0 204 (65.4) 184 (66.7) 1.0
Yes 72 (43.9) 75 (36.8) 1.35 (0.89-2.07) 108 (34.6) 92 (33.3) 1.33 (0.93-1.92)

Education
Low 69 (42.1) 84 (41.2) 1.0 223 (70.3) 193 (69.7) 1.0
Medium 55 (33.5) 73 (35.8) 0.97 (0.59-1.58) 72 (22.7) 62 (22.4) 1.21 (0.80-1.84)
High 40 (24.4) 47 (23.0) 1.11 (0.63-1.96) 22 (6.9) 22 (7.9) 1.16 (0.61-2.21)

Current alcohol intake

Never 52 (31.7) 54 (26.5) 1.0 219 (69.3) 152 (54.7) 1.0
<once a month 66 (40.2) 100 (49.0) 0.72 (0.43-1.18) 68 (21.5) 87 (31.3) 0.80 (0.53-1.20)
Daily-weekly 46 (28.0) 50 (24.5) 1.05 (0.60-1.84) 29 (9.2) 39 (14.0) 0.72 (0.41-1.26)

Smoking habits
Never 101 (61.6) 124 (60.8) 1.0 262 (83.2) 227 (81.7) 1.0
Ex 24 (14.6) 39 (19.1) 0.83 (0.46-1.49) 28 (8.9) 28 (10.1) 1.08 (0.61-1.93)
Current 39 (23.8) 41 (20.1) 1.28 (0.75-2.17) 25 (7.9) 23 (8.3) 1.31 (0.70-2.46)
a
Total number of cases and controls does not correspond because of missing values.
b
adjusted for age

45
&<3DQGEUHDVWFDQFHUULVN3DSHU,

The distribution of &<3 alleles and the ORs associated with breast cancer
are shown in Table 10. The $ allele did not significantly affect the overall
breast cancer risk, neither when all women were considered together nor after
stratification by menopausal status.

When cases were considered by tumour stage, premenopausal patients

having advanced disease were less frequently carriers of $ alleles (48%)
compared to controls (56%). There was a clear tendency of lower risk of this
type of breast cancer in carriers of $allele (OR 0.58, 95% CI 0.31-1.07). In
contrast, in postmenopausal cases, no such tendency was observed (OR,
0.99; 95% CI, 0.57-1.69).

In subgroup analyses, the protective effect of later age at menarche (> 13

years) was found to be more pronounced among premenopausal women with
the &<3 $$ genotype (OR, 0.34; 95% CI, 0.15-0.76) compared to
women with the $ allele containing genotypes (OR, 0.82; 95% CI, 0.42-1.63)
(S for interaction = 0.10). An especially low risk (OR, 0.22; 95% CI, 0.07-0.62)
was observed for premenopausal women with at least one child and the
$$ genotype, while the same was not seen among those with the $allele
containing genotypes (OR, 0.79; 95% CI, 0.31-2.02) (S for interaction = 0.11).
In contrast, increased risk of breast cancer with borderline significance (OR,
1.70; 95% CI, 0.98-2.94) was observed for postmenopausal women with the
$$ or $$ genotype and BMI over 25.4 kg/m2.

7DEOH  Association between &<3 genotypes and breast cancer risk D

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
&<3 &DVHV &RQWUROV &DVHV &RQWUROV
E E
JHQRW\SH Q Q 25 &, Q Q 25 &,

$$ 77 (47.2) 88 (43.3) 1.0 22 (38.6) 112 (40.4) 1.0

$$ 71 (43.6) 88 (43.3) 0.87 (0.54-1.39) 56 (49.4) 132 (47.7) 1.00 (0.68-1.48)
$$ 15 (9.2) 27 (13.3) 0.63 (0.30-1.32) 38 (12.0) 33 (11.9) 0.98 (0.53-1.80)
$$ $$ 86 (52.8) 115 (56.7) 0.81 (0.52-1.27) 94 (61.4) 165 (59.6) 1.00 (0.69-1.45)

a
Genotype data missing for 4 cases and 2 controls.
b
Adjusted for age, age at menarche, age at first full term pregnancy, number of full term pregnancies, first-degree
family history of breast cancer, history of BBD, use of oral contraceptives, WHR, and postmenopausal use of
oestrogen.

46
&207DQGEUHDVWFDQFHUULVN3DSHU,,

The &207 genotype was not significantly associated with the overall breast
cancer risk when all cases were considered together or when they were
stratified by menopausal status (Table 11).

An inverse association was found between the &207/ allele containing

genotypes and premenopausal advanced stage breast cancer (OR 0.44, 95%
CI 0.22-0.87), and between &207// genotype and the postmenopausal
local carcinoma and (OR 0.55, 95% CI 0.31-0.98). In contrast, a tendency of
increased risk was seen among &207/ allele carriers among
postmenopausal women with advanced stage of the disease (OR, 1.50; 95%
CI, 0.74-3.05).

When the risk associated with &207 genotypes was studied stratified by
factors known to contribute to endogenous oestrogen load, postmenopausal
women with the &207// genotype and BMI <25.4 kg/m2 were found to have
significantly decreased risk for developing breast cancer (OR, 0.33; 95% CI,
0.13-0.83). In contrast, a substantial high increase in the risk was seen among
postmenopausal women with <12 years at menarche and &207/ allele
containing genotypes (OR 8.59, 95% CI 1.85-39.8, S for interaction = 0.008).
An increased risk was also seen for postmenopausal women who had the
&207// genotype and had undergone HRT for longer than 30 months (OR
4.02, 95% CI 1.13-14.3); a significant trend (S=0.028) of increasing risk was
seen with increasing number of at-risk alleles.

7DEOH  Association between &207 genotypes  and breast cancer risk D

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
&207 &DVHV &RQWUROV &DVHV &RQWUROV

JHQRW\SH Q Q 25&, Q Q 25&,

++ 44 (26.8) 41 (20.3) 1.0 71 (22.4) 59 (21.2) 1.0

+/ 79 (48.2) 112 (55.4) 0.63 (0.37-1.09) 159 (50.2) 125 (45.0) 1.04 (0.65-1.68)
// 41 (25.0) 49 (24.3) 0.70 (0.37-1.31) 87 (27.4) 94 (33.8) 0.81 (0.48-1.35)
+/ // 120 (73.2) 161 (79.7) 0.65 (0.39-1.09) 246 (77.6) 219 (78.8) 0.94 (0.61-1.47)

a
Genotype data missing for 2 controls and 2 cases.
b
ORs and confidence intervals adjusted for age, age at menarche, age at first full term pregnancy, number of full
term pregnancies, first degree family history of breast cancer, history of BBD, use of HRT, and BMI.

47
0Q62'DQGEUHDVWFDQFHU3DSHU,,,

The prevalence of 0Q62' $ allele containing genotypes was slightly higher in

case patients (74%) compared to controls (68%), resulting in a 1.46-fold (95%
CI, 1.07-1.98) overall risk for breast cancer compared to the 0Q62' 99
genotype. This difference was confined to postmenopausal women with at
least one 0Q62' $ allele (OR 1.68, 95% CI 1.13-2.50) (Table 12). After
adjustment for only age (in Paper III) the ORs were 1.45 (95% CI, 1.08-1.95),
1.15 (95%CI, 0.73-1.83), and 1.71 (95% CI, 1.17-2.49), for all, pre- and
postmenopausal women, respectively. No notable difference was seen when
the association between the 0Q62' genotype and breast cancer was
considered according to the stage of the disease.

When the risk associated with the 0Q62' $ allele was studied stratified by
factors suspected to have a role in ROS production/scavenging, none of the
analyses revealed significant associations when all women were considered
together. However, after stratification by menopausal status, postmenopausal
women who reported ever smoking and who carried the 0Q62' $ allele
containing genotypes were found to have a clearly increased risk for breast
cancer (multivariate adjusted OR 4.56, 95% CI 1.46-14.2). The highest risk
was seen for postmenopausal women who had been smoking for longer than
15 years or those who smoked more than 10 cigarettes/day, giving an
unadjusted OR of 5.12 (95% CI, 1.42-18.4), and OR of 5.48 (95% CI, 1.28-
23.4), respectively.

Increased risks associated with the 0Q62' $ allele were also found for
women who reported current use of alcohol (multivariate adjusted OR 2.33,
95% CI 1.20-4.54), use of oral contraceptives at some time (multivariate
adjusted OR 3.23, 95% CI 1.34-7.78), or use of HRT (multivariate adjusted
OR 2.51, 95% CI 1.24-5.11). A significant trend (S=0.01) of increasing risk
with increasing number of high risk alleles could be seen for HRT users.
Furthermore, when the adjustment was done only for age, the interaction
between 0Q62'and HRT use was statistically significant (S=0.02).

7DEOH  Association between 0Q62' genotypesa and development of

breast cancer according to menopausal status at diagnosis of the case patient.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
0Q62' &DVHV &RQWUROV &DVHV &RQWUROV

JHQRW\SH Q Q 25&, Q Q 25&,

99 46 (28.0) 62 (30.4) 1.0 78 (24.8) 91 (32.7) 1.0

9$ 82 (50.0) 106 (52.0) 1.12 (0.68-1.86) 173 (54.9) 125 (45.0) 

$$ 36 (22.0) 36 (17.6) 1.40 (0.74-2.64) 64 (20.3) 62 (22.3) 1.32 (0.79-2.18)

9$ $$ 118 (72.0) 142 (69.6) 1.19 (0.74-1.93) 237 (75.2) 187 (67.3) 

a
Genotype data missing for 4 cases.
b
ORs and 95 % CIs adjusted for age, age at menarche, age at first full term pregnancy, number of full term
pregnancies, first-degree family history of breast cancer, history of BBD, and BMI.

48
*67VDQGEUHDVWFDQFHUULVN3DSHU,9

The *670null genotype was found in 46 % of cases and 42 % of population

controls. It was significantly (S=0.02) more frequent among premenopausal
controls (48 %) compared to postmenopausal controls (38%), giving an OR of
1.49 (95% CI, 1.03-2.15) for breast cancer for postmenopausal women
lacking the *670gene (Table 13). Moreover, the effect of the*670 null
genotype was somewhat more pronounced among postmenopausal current
alcohol users (OR, 1.96; 95% CI, 1.05-3.69) compared to those who reported
not using alcohol (OR, 1.20; 95% CI, 0.74-1.94).

7DEOH  Association between *670 genotypes and breast cancer risk
D

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
*670 &DVHV &RQWUROV &DVHV &RQWUROV
E E

JHQRW\SH Q Q 25 &, Q Q 25 &,

SUHVHQW 89 (54.3) 105 (52.2) 1.0 171 (53.9) 173 (62.5) 1.0
QXOO 75 (45.7) 96 (47.8) 0.95 (0.61-1.46) 146 (46.1) 104 (37.5) 

a
Genotype data missing for 4 controls and 2 cases.
b
Adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first-degree family
history of breast cancer, and history of BBD.

The prevalences of the *670 $ % or % % genotypes were similar in cases

(24% and 1.5%) and controls (24% and 1.7%) when pre- and postmenopausal
women were considered together. There were somewhat fewer individuals
with the *670 % allele containing genotypes among premenopausal (21%)
than postmenopausal controls (28%), but this difference was not statistically
significant (S=0.09) (Table 14). The adjusted OR for women with the
*670 % allele containing genotypes was 1.48 (95% CI, 0.89-2.47) among
premenopausal cases, and 0.79 (95% CI, 0.53-1.20) among postmenopausal
cases. Furthermore, the protective effect of *670 % allele among
postmenopausal women was limited to those who reportedly did not use
alcohol (OR, 0.55; 95 % CI, 0.32-0.94; Sfor interaction 0.04).

7DEOH  Association between *670  genotypes and breast cancer risk
D

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
*670 &DVHV &RQWUROV &DVHV &RQWUROV
EF EF

JHQRW\SH Q Q 25 &, Q Q 25 &,

$ $ 121 (73.8) 160 (78.8) 1.0 240 (75.7) 199 (71.8) 1.0
$ % 42 (25.6) 41 (20.2) 71 (22.4) 72 (26.0)
1.48 (0.89-2.47) 0.79 (0.53-1.20)
% % 1 (0.6) 2 (1.0) 6 (1.9) 6 (2.2)

D
Genotype data missing for 2 controls and 2 case patients.
b
Adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first-degree family
history of breast cancer, and history of BBD.
c
OR calculated for combination of $ % and % % genotypes.

49
The frequency of the *673 Val allele was 26% among controls and 23%
among cases. A tendency of decreased risk for breast cancer was associated
with the *673 Val allele in both pre- and postmenopausal women, but the
associations did not reach statistical significance (Table 15). However, among
women who had ever used HRT the risk paralleled the number of *673 Val
alleles; OR was 0.44 (95% CI, 0.23-0.86) for heterozygotes and 0.17 (95% CI,
0.04-0.77) for homozygotes for the Val allele compared to the women with
*673 Ile/Ile genotype (S for interaction 0.048). No such effect was seen
among women who had never used HRT (OR, 1.22; 95% CI, 0.75-1.98, and
OR, 0.60; 95% CI, 0.18-2.03, respectively for Ile/Val and Val/Val carriers).

7DEOH  Association between *673 genotypes  and breast cancer risk D

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
*673 &DVHV &RQWUROV &DVHV &RQWUROV
E E
JHQRW\SH Q Q 25 &, Q Q 25 &,

,OH,OH 98 (59.8) 113 (55.7) 1.0 185 (58.0) 153 (55.0) 1.0
,OH9DO 57 (34.7) 75 (36.9) 0.89 (0.56-1.41) 121 (37.9) 106 (38.1) 0.88 (0.61-1.29)
9DO9DO 9 (5.5) 15 (7.4) 0.66 (0.27-1.63)c 13 (4.1) 19 (6.8) 0.49 (0.21-1.14)d

a
Genotype data missing for 1 control subject.
b
Adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first-degree family history
of breast cancer, and history of BBD.
c
S for trend 0.353
d
S for trend 0.139

Although the frequency of the *677 null genotype was slightly higher among
premenopausal cases (15%) compared to controls (12%), resulting in a
tendency for increased risk among premenopausal women, no statistically
significant associations were found (Table 16). However, postmenopausal
women who had ever received HRT and lacked the *677 gene had an
almost three-fold risk of breast cancer (OR, 2.91; 95% CI, 1.11-7.64, S for
interaction 0.03) compared with those who had the gene.

7DEOHAssociation between *677 genotypea and breast cancer risk

stratified by menopausal status.

3UHPHQRSDXVDO 3RVWPHQRSDXVDO
*677 &DVHV &RQWUROV &DVHV &RQWUROV
E E
JHQRW\SH Q Q 25 &, Q Q 25 &,

SUHVHQW 139 (84.8) 176 (87.6) 1.0 272 (85.8) 239 (86.3) 1.0
QXOO 25 (15.2) 25 (12.4) 1.50 (0.79-2.84) 45 (14.2) 38 (13.7) 1.13 (0.67-1.91)
a
Genotype data missing for 4 controls and 2 case patients.
b
Adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first-degree
family history of breast cancer, and history of BBD.

50
When combined effects of the *67 genotypes were examined,
premenopausal women having the *670 % allele in combination with the
*673 Ile/Ile genotype had about a 2-fold risk of breast cancer (OR, 2.07;
95% CI, 1.02-4.18) (Table 17). Furthermore, a significant gene-gene
interaction was found between these two genotypes both in pre- (S for
interaction = 0.035) and postmenopausal women (S for interaction = 0.012). 

However, no statistically significant combined effects were seen for

postmenopausal women simultaneously carrying any combination of two of
the at-risk genotypes.

7DEOHAssociation between combinations of two putative *67at-risk

genotypes and breast cancer risk stratified by menopausal status.

D
&DVHV&RQWUROV 25 &,
3UHPHQRSDXVDO
*670 *670

present  $ $ 58/73 1.0

null  $ %RU % % 12/11 1.60 (0.63-4.11)
*670 *673

present Ile/Val or Val/Val 30/44 1.0

null Ile/Ile 39/50 1.17 (0.60-2.26)
*670 *677

present present 77/89 1.0

null null 13/9 2.18 (0.84-5.68)
*673 *670

Ile/Val or Val/Val  $ $ 53/68 1.0

Ile/Ile  $ %RU % % 30/21 

*673 *677

Ile/Val or Val/Val present 57/79 1.0

Ile/Ile null 16/14 1.85 (0.79-4.30)
*677 *670

present  $ $ 105/139 1.0

null  $ %RU % % 9/6 2.58 (0.84-8.01)
3RVWPHQRSDXVDO
*670 *670

present $ %RU % % 54/61 1.0

null  $ $ 123/88 1.66 (0.99-2.77)
*670 *673

present Ile/Val or Val/Val 63/76 1.0

null Ile/Ile 76/56 

*670 *677

present present 142/147 1.0

null null 16/12 1.48 (0.60-3.65)
*673 *670

Ile/Val or Val/Val $ %RU % % 36/25 1.0

Ile/Ile  $ $ 143/99 1.08 (0.58-2.01)
*673 *677

Ile/Val or Val/Val present 111/112 1.0

Ile/Ile null 23/26 1.17 (0.59-2.33)
*677 *670

present $ %RU % % 65/65 1.0

null  $ $ 33/26 1.37 (0.68-2.76)
a
ORs adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first-degree
family history of breast cancer and history of BBD.
Tests for interaction between *67 genotypes for pre- and postmenopausal women, respectively: *670 and
*670, 3=0.841 and 3=0.869; *670 and *673, 3=0.695 and 3=0.249; *670 and *677, 3=0.117 and

3=0.581; *673 and *670, 3=0.035 and 3=0.012; *673 and *677, 3=0.893 and 3=0.518; *677 and

*670, 3=0.633 and 3=0.680.

51
When three of the putative at-risk genotypes were combined no significant
increased risks were seen among the postmenopausal women, but clear
combined effects were observed among the premenopausal women (Table
18). The most remarkable risk was seen for premenopausal women
concurrently carrying the *670 % allele containing genotype, the *677
null genotype, and the *673 Ile/Ile genotype, who had a 10-fold risk (OR,
9.93; 95% CI, 1.10-90.0) of breast cancer. Substantially increased risk of
borderline significance was also seen for the*670 null women carrying the
*673 Ile/Ile genotype together with the *677 null genotype (OR, 3.96;
95% CI, 0.99-15.8).

7DEOHAssociation between combinations of three putative GST at-risk

genotypes and breast cancer risk stratified by menopausal status.

&DVHV&RQWUROV 25D&,

3UHPHQRSDXVDO
*670 *670 *673

present  $ $ Ile/Val or Val/Val 20/29 1.0

null  $ %RU % % Ile/Ile 9/4 3.23 (0.82-12.7)
*670 *670 *677

present  $ $ present 53/60 1.0

null  $ %RU % % null 2/3 0.76 (0.10-5.54)
*670 *673 *677

present Ile/Val or Val/Val present 26/38 1.0

null Ile/Ile null 8/4 

*677 *670 *673

present  $ $ Ile/Val or Val/Val 47/62 1.0

null  $ %RU % % Ile/Ile 6/1 

3RVWPHQRSDXVDO
*670 *670 *673

present  $ %RU % % Ile/Val or Val/Val 22/15 1.0

null  $ $ Ile/Ile 67/49 1.08 (0.47-2.49)
*670 *670 *677

present  $ %RU % % present 45/49 1.0

null  $ $ null 13/12 1.26 (0.45-3.53)
*670 *673 *677

present Ile/Val or Val/Val present 49/69 1.0

null Ile/Ile null 8/7 2.74 (0.81-9.32)
*677 *670 *673

present  $ %RU % % Ile/Val or Val/Val 31/23 1.0

null  $ $ Ile/Ile 16/15 1.08 (0.47-2.49)
a
ORs adjusted for age, age at menarche, age at first full term pregnancy, number of pregnancies, first degree
family history of breast cancer, and history of BBD.

We were restricted from evaluating the potential combined effects of all four
*67 genes among the premenopausal women because none of the controls
and only two of the cases carried all the four at risk genotypes simultaneously.
Among the postmenopausal women, on the other hand, no significant
increased risks were seen (data not shown).

52
&RPELQHGHIIHFWRI&207DQG*67JHQHVRQEUHDVWFDQFHU
ULVN3DSHU9

No combined effect was seen between the &207 genotype and any of the
four polymorphic *67 genes studied in this thesis. Stratification by
menopause did not affect the estimates. However, after stratifying for the use
of HRT, substantially increased risks were seen. Women who reported having
ever received HRT, and who carried the &207/allele containing genotypes
together with either the *673 Ile/Ile genotype or the *677 null genotype
had a 4-fold risk of breast cancer (OR, 4.10; 95% CI, 1.24-13.6; and OR, 4.19;
95% CI, 1.30-13.5, respectively) (Table 19). When data of all the three
genotypes were combined, a trend of increasing risk was seen when there
was a higher number of at-risk genotypes (S for trend 0.004) (Table 20).
Having any two at-risk genotypes compared to having the three most
advantageous genotypes of these genes posed a 9.86-fold (95% CI, 1.55-
62.9) risk of breast cancer. Also the combinations of &207 *670 and
*673 as well as &207 *677 and *673 at-risk genotypes posed
increased risks with ORs of 12.8 (95%CI, 1.33-124.0) and 26.3 (95% CI, 2.17-
318.3), respectively.

7DEOH  Association between combinations of &207 and *67 genotypes

and breast cancer risk stratified by HRT.
D D
1HYHUXVHRI+57 (YHUXVHRI+57
&DVH &DVH
*HQRW\SHV E E
&RQWURO 25 &, &RQWURO 25 &,

&207 *670
HH present 55/37 1.0 8/17 1.0
HL/LL null 133/124 0.80 (0.48-1.32) 32/31 2.43 (0.85-6.93)

&207 *670
HH AA 69/51 1.0 11/19 1.0
HL/LL AB/BB 64/64 0.76 (0.45-1.29) 22/28 1.45 (0.54-3.90)

&207 *673
HH Ile/Val & Val/Val 48/27 1.0 4/17 1.0
F
HL/LL Ile/Ile 163/153 0.65 (0.38-1.13) 52/56 

&207 *677
HH present 81/63 1.0 12/26 1.0
G
HL/LL null 35/44 0.63 (0.35-1.14) 14/9 

a
Information on HRT use missing for 7 subjects.
b
Adjusted for age, age at menarche, number of pregnancies, age at first full term pregnancy, first-
degree family history of breast cancer, history of BBD, and BMI.
S for interaction between combined genotypes and HRT use c0.019, and d0.019.

53
7DEOH   The combined effect of &207 *673 and *677 genotypes
among HRT users.
D
&207 *673 *677JHQRW\SHV &DVH&RQWURO 25 &,

No at risk genotypesb 2/16 1.0

One at risk genotype 40/58 6.27 (1.31-30.0)

Two at risk genotypes 52/50 9.69 (2.02-46.4)

Three at risk genotypes 8/8 9.86 (1.55-62.9)c

a
Adjusted for age, age at menarche, number of pregnancies, age at first full term pregnancy, first-degree family
history of breast cancer, history of BBD, and BMI.
b
Subjects with &207++*673 Ile/Val & Val/Valand *677present genotypes serve as a referent category.
c
S for trend 0.004

54
',6&866,21

0HWDEROLFJHQRW\SHVDQGRYHUDOOEUHDVWFDQFHUULVN

&<3
The &<3$allele is hypothesised to encode a protein with an additional
promoter site, leading to increased enzyme activity and a higher lifetime level
of oestrogen. Accordingly, this allele was shown to increase the risk for
advanced stage of breast cancer (135) motivating a number of groups to
study this issue further in different case-control settings, employing diverse
ethnicities, and varying age ranges (135,166-170,172,173). Most of the
subsequent studies failed to provide support for the original finding. The only
supporting evidence comes from a recent study which did have a larger
number of cases and controls from the same cohort but with a lower risk
estimate with only borderline significance (140). Furthermore, they show a
combined effect with 17β-HSD in women with low oestrogen levels. However,
since the stage of the disease has not been considered in most of these
studies, the possibility of the risk modification by the &<3 genotype could
not be fully excluded.

In agreement with the majority of the previous studies, our study did not
reveal any significant association between &<3$ allele and overall risk of
breast cancer, either when all women were considered together or when the
cases were considered according to the stage of the disease. This contrasts
with the findings in two recent studies showing an increased risk for young
(<40 years) Swedish (167) and Australian (168) women, especially those with
a family history of breast cancer (168). However, since our study includes only
a few women of this age (38 cases and 50 controls) and even fewer
simultaneously having a positive family history (2 controls and 2 cases), this
possible association could not be evaluated thoroughly in our study
population.

&207
COMT participates in the inactivation of reactive catechol oestrogen
metabolites recognized to have a role in breast carcinogenesis. A genetic
polymorphism leads to 4-fold individual differences in the activity of the
respective enzyme and it is therefore considered to be an important
susceptibility factor in the development of breast cancer. However, the studies
conducted on COMT have shown differing results, with both increased and
decreased risks being associated with the hypothesised high risk (low activity)
allele, depending on menopausal status (171,184-187).

In our study no overall effect on breast cancer susceptibility could be seen,

either in pre- or postmenopausal women. When stratified by the stage of the
disease the hypothesised high risk allele was seen to be associated with

55
lower risk of advanced stage disease in premenopausal but with local
carcinoma in postmenopausal women. This further supports the conflicting
role found for the &207 genotypes. Since it has been shown that the ratio of
catechol oestrogens to 16α-hydroxylated oestrogens was four to five times
higher in the Finnish women compared to Orientals (124), an important
modifying role for the &207 could have been suspected among Finns.
However, as discussed in detail later, even if the &207 genotype does not
seem to confer high individual susceptibility to breast cancer at least in
Finnish women, potentially important interactions were found with
endogenous and exogenous oestrogens.

0Q62'
MnSOD is an endogenous antioxidant enzyme that takes part in the
detoxification of reactive oxygen species emerging in cells due to normal
metabolism including metabolism of oestrogen and various xenobiotics (132).
In the presence of excessive production of ROS, or when the cellular defence
mechanisms are insufficient, oxidative stress may lead to DNA damage. The
polymorphism in the mitochondrial targeting sequence of the MnSOD is
hypothesised to affect its transport to mitochondria, where it would be
biologically available, and therefore lead to increased risk of oxidative stress
and cancer.

In the only study reported to date on MnSOD polymorphism and breast

cancer, a positive association was found between the 0Q62' $ allele
containing genotype and breast cancer risk, especially among women who
consumed lower amounts of dietary antioxidants (89). In accordance with
these findings, in our study women with the 0Q62' $ allele containing
genotypes were found to have a 1.5-fold increased risk of developing breast
cancer. However, this was seen to be mainly limited to postmenopausal
women while in the earlier study it was seen for premenopausal women.
Since no dietary information was available for our study subjects, the effect of
antioxidants in the diet could not be evaluated.

It has to be noted, that the amphiphilic helical structure essential for the
correct transport is predicted to be associated with the 0Q62' $ allele. This
allele appeared, however, as the risk allele both in this study and the previous
study. The possible mechanistic explanation for this may be the simultaneous
accumulation of H2O2 at least in the absence of H2O2 scavenging enzymes
such as glutathione peroxidase and catalase (Figure 4). In the presence of
transition metals, this in turn gives rise to the generation of potent free
radicals and hence mitochondrial damage. Furthermore, as earlier
speculated, the prediction of an altered cellular location of the MnSOD is
based on a limited statistical model and should thus be interpreted with
caution (89).

56
*67V
GSTs play an important role in the metabolism of environmental carcinogens
such as PAHs and heterocyclic aromatic amines (207) but they may also take
part in the metabolism of oestrogen and lipid peroxidation (110). The
inactivation of semiquinones, the metabolic products of catechol oestrogens,
which are capable of binding directly to DNA or redox cycling, is achieved by
conjugation with GSH. The GSTs, especially GSTM1 and GSTT1, are
probably the most extensively studied putative cancer susceptibility genes,
when all cancers are considered (262). The *670null genotype has been
shown to increase the risk of postmenopausal breast cancer in two studies
(219,220). These findings were recently supported by the meta-analysis
combining the results of most of the studies conducted to date (OR, 1.33;
95% CI, 1.01-1.76) (11). In contrast, in one study assessing Korean women,
increased risk was seen among premenopausal women (225). However, also
negative results exist (192,194,221-224).

In our study, a 1.5-fold increased risk was found for postmenopausal breast
cancer among women carrying the *670 null genotype, supporting the
outcome of the above mentioned meta-analysis (11). It was particularly
interesting that the association was found to originate from the significantly
lower (38%) frequency of *670 null genotype in older postmenopausal
controls compared to premenopausal controls (48%), while the respective
frequencies in cases remained unchanged (46%). In both of the earlier
studies where a positive association for the *670 null genotype and breast
cancer have been found, a similar decreasing frequency of the *670 null
genotype can be seen among older healthy controls (see Table 4) (219,220).
Moreover, in a recent study including a comparable number of subjects to
ours, the frequency of the *670 null carrying women was somewhat higher
in premenopausal than in postmenopausal controls even though no significant
overall association was found (224). Instead, a positive association was seen
between the *670 null genotype and breast cancer risk for women with a
positive family history. The same could not be seen in our study, possibly
because rather few women with a positive family history were included in our
study (54 cases and 22 controls). It could be speculated that lower frequency
among older cancer free subjects is actually what should be seen if a
genotype is a real risk genotype.

The effect of polymorphism in the *670 gene is not clear but it has been
shown to have overlapping substrate specificities with the *670 gene. In
addition, individuals carrying the *670 $ allele have been shown to have a
reduced expression of GSTM3 (235). Given the putative modifying role of
*670 genotypes in breast cancer susceptibility, the effect of the *670
gene polymorphism should most probably be studied simultaneously for more
reliable results. However, no such studies have yet been conducted.

In our study no significant association was found when the *670genotype

was studied alone. A tendency of increased risk was, however, found for the
premenopausal women carrying the *670 % allele containing genotypes. In
postmenopausal women, an inverse nonsignificant association was seen.

57
Moreover, significant associations were found when the *670 genotype
was studied together with the other *67 genotypes.

Even if GSTP1 is the main GST expressed in breast tissue and the *673
Val105 allele has been shown to confer increased survival (240), its potential
role in breast cancer susceptibility has been studied in only a few studies
(220,224,232,238). None of the individual studies have found a significant
overall effect for the *673 genotypes. However, in a recent meta-analysis
combining the results of the two small studies conducted by the time of the
analysis (220,238), increased risk was seen for the Val allele containing
genotypes (11). Quite unexpectedly, we found a tendency of lower risk
associated with the *673 Val allele containing genotypes in both pre- and
postmenopausal women. On the other hand, *673 Val allele was recently
found to be more active than the Ile allele in conjugation reactions involving
bulky diol epoxides of PAH (237). The activity of the enzyme encoded by the
*673 Val allele could be higher also toward other substrates, such as
oestrogens, thereby conferring protective effect against carcinogenicity of
these factors. A tendency of lower risk can also be seen in a recent study
among white women carrying the Val allele containing genotypes, while the
same is not seen for African-American women (224). Former smokers with
the *673 Ile/Ile were also seen to be at increased risk in the above
mentioned study which is in accordance with the lower activity seen for the Ile
allele against PAHs.

No statistically significant associations were found when the *677 genotype

was studied alone. However, premenopausal women with breast cancer
lacked the *677 gene somewhat more often compared to controls. It should
be noted that due to the low frequency of the *6777 null genotypes, the
power in our study was not adequate to detect any possible effect among
premenopausal women.

*HQHJHQHLQWHUDFWLRQVDQGEUHDVWFDQFHUULVN

Assessment of single polymorphic genotypes is not expected to be sufficient

for evaluating individual susceptibility to the various endogenous or
exogenous harmful agents (263). The study of interactions between different
polymorphic genes should, however, be based on biologically feasible
hypotheses. For instance simultaneous examination of genes for enzymes
with overlapping substrate specificity for a given exposure or separate genes
acting in sequence in the same metabolic pathway are anticipated to be well
justified. Accordingly, some studies have already been carried out to examine
combined effects of separate polymorphic oestrogen metabolising genes in
relation to breast cancer (111,184,194,197,220,224,231,232).

Within the *67supergene family, some isoforms demonstrate an overlap in

substrate specificities and are therefore often studied in combinations
(231,232,239,264). However, these analyses are usually limited by the study

58
size due to the rarity of the at-risk genotypes (such as *673 Val/Val and
*677 null genotypes). One strength of our study was that it had a
reasonably larger number of cases and controls, which provided a good
possibility for studying the putative combined genotype effects.

None of the earlier studies on breast cancer and *67 genotypes have
evaluated the potential role of the *670 polymorphism, possibly because
the biological effect of the polymorphism is not yet completely known. The
GSTM3 enzyme is, however, known to have overlapping substrate specificity
with the GSTM1 enzyme, and is suggested to be in linkage disequilibrium with
the *670 gene (155). We therefore examined also the potential role of
*670 polymorphism in breast cancer proneness. No interaction was found
between the *670 and *670 genotypes in this study. In contrast, an
interaction was found between the *670 and the *673 genotypes.
Accordingly, among premenopausal women the *670 % allele together with
the *673 Ile/Ile genotype posed about a 2-fold risk of breast cancer.
Furthermore, when these women simultaneously carried the *677 null
genotype their risk of developing breast cancer increased to 10-fold. Although
this estimate is based on only 1 control and 6 cases, our results suggest an
important role for the *670genotype, especially in combination with other
*67s among premenopausal women. This emphasises the need to include
*670 genotype analyses in future studies on *67s and breast cancer
susceptibility.

In our study, substantially increased (4-fold) risk of borderline significance was

seen for premenopausal women with the concurrent presence of the*670
null, *677 null and *673 Ile/Ile genotypes. We could not observe the
earlier suggested combined effect for the *670 null *677 null and
*673 Val/Val genotypes (220). The separate analyses of the *673
genotype was indicative at a protective rather than a predisposing role for the
Val/Val genotype, which has been used as an at-risk genotype in the earlier
studies on the combined effects of these genes (220,224,232).

In a recent study of comparable size to ours, the *673Ile/Val and Val/Val

genotypes were also seen to pose a tendency of decreased risk of breast
cancer among white women while the respective risk estimates among
African-American were close to one (224). For some reason, in the combined
analyses, the *673 Val allele containing genotypes were, nevertheless,
used as the at-risk genotypes. Furthermore, African-American and white
women were combined in the analyses, even though the frequencies of the
*673genotypes were significantly different in the two ethnic groups (Table
5). This is anticipated to have masked the potential combined effects. Rather
expectedly no combined effects were seen for the genotypes studied. In fact
the lowest risk was seen for women carrying the combination of the *670
null, *677 null and *673 Val allele containing genotypes. As GSTs act
sequentially with COMT in the metabolism of reactive oestrogen
intermediates, they also studied the combined effect with &207 genotypes.
The lowest risk was seen among carriers of the &207// genotype together
with *673 Ile/Val or Val/Val genotypes (224).

59
Our results on *67s together with the &207 genotypes are in agreement
with the above discussions; lowest risk was seen for postmenopausal women
carrying the &207// genotype together with the *673 Val allele
containing genotypes (OR, 0.52; 95% CI, 0.30-0.89)(Paper V). In contrast, in
another study, a 3-4-fold increased risk of postmenopausal breast cancer was
seen for women carrying the &207// genotype together with either the
*670 null or at least one *673 Val allele containing genotypes (184).
However, only one-fourth of the number of the subjects included in the
present study was evaluated in that study. No increased risks in overall breast
cancer susceptibility were seen for any of the studied *67at risk genotypes
with the &207/ allele containing genotypes in our study population (Paper
V).

*HQHHQYLURQPHQWLQWHUDFWLRQVDQGEUHDVWFDQFHUULVN

(QGRJHQRXVKRUPRQHV
In spite of the lack of overall association between &<3 and breast cancer
risk, we observed an effect modification with age at menarche in agreement
with earlier findings (26). Women with later age at menarche are known to
have somewhat lower risk of breast cancer. In our study, this reduced risk
was seen to be limited to premenopausal women who did not carry the
hypothesised high enzyme activity (A2) allele. This finding is in agreement
with an earlier study (135). Furthermore, a substantial (over 8-fold) risk was
seen among postmenopausal women carrying the putative high-risk &207
// genotype, if they had had an early age at menarche. Although the latter
association has not been shown in earlier studies, it is in accordance with the
above mentioned hypothesis. The low oestrogen synthesis genotype ($$)
together with later age at menarche (associated with both shorter lifetime
exposure and later establishment of regular ovulatory cycles) appears to
protect from breast cancer, while impaired carcinogenic oestrogen
metabolism associated genotype (&207//) with increased exposure, seem
to predispose to breast cancer.

No genotype effect was seen in women with later age at menopause or when
the total years of oestrogen exposure were calculated. This suggests that the
association between early age at menarche and &207genotype appears to
reflect the early establishment of regular ovulatory cycles rather than the total
length of premenopausal years. In general, early exposure (the period before
the first pregnancy) to the hormonal milieu is suggested to be an important
etiologic factor (25,28), probably because the undifferentiated ductal elements
of the breast are more susceptible to the action of genotoxins (17). Moreover
among women with a shorter cycle length, a tendency of increased risk was
seen with &207/ allele containing genotypes, while those with longer cycle
length and &207/ allele had a lower risk (data not shown). It has been
shown that women with shorter cycles spend more of their reproductive years

60
in the luteal phase when oestrogen and progesterone levels are high and
mitotic activity reach its peak (28).

The protective effect of parity was also seen to be confined to premenopausal

women without the &<3 $ allele. In contrast, obese postmenopausal
women with the &<3 $ allele containing genotypes tended to be at
increased risk while lean postmenopausal women with &207// genotype
had decreased risk. All these findings support the idea that increased
exposure to endogenous oestrogens increases the risk of developing breast
cancer, with decreased exposure being beneficial.

([RJHQRXVKRUPRQHV
According to current opinion, postmenopausal HRT is considered to result in
only a slight increase in the risk of breast cancer (2.3% for each year of use)
(42). The excess risk is estimated to be 1.4-fold for women having used HRT
for five years or longer. However, genetic polymorphisms in enzymes involved
in oestrogen metabolism could be hypothesised to affect this risk. Women
with certain at-risk genotypes would have higher concentrations of
carcinogenic catechol oestrogens and their further oxidised metabolites
capable of forming DNA-adducts or reactive oxygen species. They would
therefore be at higher risk than women with the more favourable genotypes.
Accordingly, in our study postmenopausal women who had ever used HRT
had a tendency for increased (1.5-fold) risk of breast cancer if they carried the
&207// genotype compared to those with the &207++ genotype. This
risk was seen to increase to 4-fold among women who had used HRT for
longer than 30 months. Similarly, women with the*677 null or 0Q62' $$
genotype were found to be at almost 3-fold risk, and carriers of the *673
Ile/Ile genotype at 2-fold increased risk of breast cancer when they reported
having ever used HRT.

When the at-risk genotypes of the &207 *673 and *677 genes were
combined, clear additive effects were seen. A 4.5-fold risk was seen for the
carriers of the two defective genotypes, and 7 to 8-fold risks were seen
among long term users (> 30 months) with two at-risk genotypes. The latter
estimates, however, were based on rather low subject numbers and yielded
therefore unstable estimates with wide CIs. Similarly, a trend of increasing risk
with increasing number of high-risk genotypes of all the three genes was
seen.

The use of oral contraceptives was found to be associated only with the
0Q62' genotype. Postmenopausal women who had used oral
contraceptives and carried the 0Q62'$allele containing genotypes were at
3-fold risk compared to 0Q62'99carrying women.

Even if based on low subject numbers in the subgroup analyses, our findings
are considered to offer convincing evidence for the effect of the studied XME
gene polymorphisms in the breast cancer risk associated with HRT. As the
number of postmenopausal women using HRT is growing, our results are
anticipated to be of great importance also from a public health point of view.

61
The possibility to recognise women who are at greater than average risk
would provide valuable additional information for physicians and to the women
themselves as they weight up the benefits and risks of the HRT use.

6PRNLQJ
Cigarette smoke is known to be a very rich source of free radicals and to
induce MnSOD and other antioxidant enzymes (63,64). In this study, women
who had ever been smokers and carried the 0Q62' $ allele were found to
be at over 4-fold risk of breast cancer, with even higher risks seen for women
who smoked more than 10 cigarettes/day or had smoked longer than 15
years. The results from studies on the role of smoking in breast cancer
susceptibility have been inconsistent, and at best only a weak association with
breast cancer risk has been detected (53,54). As the frequency of both the at-
risk genotypes and smokers differs in racially diverse populations, the genetic
difference in individual susceptibility could, at least partly, explain the
conflicting results found for smoking and breast cancer. The antioestrogenic
effect related to tobacco smoke might dominate among women without the at-
risk genotypes, while the women with at-risk genotypes would be more
susceptible to the carcinogenic effects of the tobacco smoke. The prevalence
of smokers in postmenopausal women was 8% in this study. Considering that
smoking is more common in many countries and increasing also among
young women in Finland (22% in the premenopausal women in this study),
the population attributable risk due to smoking could be substantial.

$OFRKRO
In this study women who reported current use of alcohol and carried the
hypothesised impaired activity associated 0Q62' genotypes had a 2-fold
increased risk of breast cancer. This finding is in agreement with the proposed
mechanism for carcinogenicity of alcohol through production of ROS (70).
Another suggested mechanism is by depressing GST activity, especially that
of GSTP1 (71). In our study, a significant interaction was only seen between
the *670 genotypes and alcohol. However, a 2-fold increase in the risk was
also seen among alcohol using women with the *670 null genotype, which
is in accordance with the value found in a Korean study (225). Even though
the number of subjects in these subgroup analyses is limited, these results
support the suggested mechanisms for alcohol carcinogenicity. It should,
however, be noted that we only examined the current users of alcohol, while
in some studies the use before the first birth has been suggested to be more
important (68).

62
0HWKRGRORJLFDOFRQVLGHUDWLRQV

Case-control studies are the most commonly used methods to seek potential
associations between genetic polymorphisms and the risk of common
diseases in the population, and interactions between genetic and
environmental risk factors. To be able to detect moderate to large genetic
effects (OR > 2.0) with an 80% power, at least 200 cases and 200 controls
are usually needed (265). The assessment of interactions requires even
larger studies, especially when the factors under study are very rare or very
common, or when the magnitude of the interaction is modest (266).

Our study is one of the largest studies on breast cancer and polymorphic
genes conducted to date. The power of our study to detect a moderate to
large genetic main effect was over 80% for all the genotypes, except for the
0Q62' and *677 genotypes among premenopausal women. Among
postmenopausal women, the power was actually over 90% for all genotypes
studied. These figures, however, only apply to the total study population and
were not achieved in many of the subgroup analyses.

Concern has been raised that population stratification (LH inclusion of

individuals from a heterogenous genetic background) most likely explains the
observed associations between genotypes and disease, rather than the
physiologic effect of the genetic variant (267). Any genetic (or environmental)
factor, the distribution of which differs between ethnic groups, may falsely
appear to be related to the disease, simply because the ethnic groups are
distributed differently between cases and controls. Inclusion of only Finnish
women who have a genetically homogenous origin, is therefore considered to
be one of the most obvious strengths of the present study.

All women participating in the study provided reliable background information

on reproductive factors, medication, family history of breast cancer, previous
cancers, smoking, and use of alcohol, etc. This enabled us to assess
genotype interactions with these factors. Overall, given the moderate increase
in the risk associated with these low penetrance genes, and the complex
nature of breast cancer, studying the effects in relation to certain exposures is
anticipated to be important. The possible dual role (activating/detoxifying) of
XMEs depending on the substrate, also supports this idea.

The selection of controls should ensure that they are form a representative
sample of the population from which the cases arose. In this study, the
cooperation rate for our controls living in the catchment area of the cases was
72%, while that of included eligible cases was 98%. A possibility of selection
bias cannot be ruled out since highly educated women with a healthier
lifestyle are known to be more likely to volunteer for epidemiological studies.
This might also explain the lack of overall effect for HRT and smoking in
breast cancer risk, and the inverse association between oral contraceptive
use and breast cancer risk. It should also be noted that the ability to recall
early-life events (HJ., age at menarche, use of contraceptives) in
postmenopausal women is anticipated to be poorer than that in

63
premenopausal women. Furthermore, cases aware of an exposure-disease
relationship may under or over-report these factors (268). On the other hand,
although all cases were interviewed before diagnosis, the possibility of
interviewer bias between cases and controls cannot be totally excluded.
However, since neither recall nor the selection bias is considered to be related
to genotype, and as all the analyses in relation to these background variables
were studied within and not between exposure groups (LH genotype effect
among HRT users not between users and non-users) these potential sources
of bias are not considered to have affected the present outcomes.

The goal at this study was to recruit all of the eligible cases in the catchment
area of the hospital. Comparison with Finnish Cancer Registry showed that
this was achieved quite well since only 15% of all eligible cases were missed.
They were all incident cases with histologically confirmed breast cancer.
Incident cases with recent onset of the disease are considered to have a
strong advantage over prevalent cases that may have had the disease for
some years by the time of recruitment. Prevalent cases may be biased toward
those with a longer disease duration, LH., the genotype effect may be more
apparent on survival and not on susceptibility (269).

Many of the earlier studies have not stated the age of the cases and controls,
or the menopausal status of the study subjects. It is nowadays acknowledged
that the age distribution in controls should represent that of the cases; if
controls are significantly younger than cases they may include several future
cases (LH they will develop cancer at a later age) (270). This could mask the
possible associations. This potential source of bias can be overcome by
matching the controls to the cases by age (261). In this study, controls were
originally selected to represent the age of the cases, but the study also
included some cases without matched controls (over 77 years of age).
Therefore, our controls were slightly younger (5 years) than the cases. This
was corrected by adjusting for age in all analyses.

In addition to age, all analyses were adjusted at least for age at menarche,
age at first full term pregnancy, number of pregnancies, first-degree family
history of breast cancer, and history of benign breast diseases. We did not
have data on other susceptibility genes, such as %5&$ and %5&$ The
number of cases with positive family history was, however, small in the
present study. Furthermore, in a recent study these mutations were shown to
only account for a small number of hereditary breast cancer cases in Finnish
women (5). Therefore, the lack of information on these genes is not
considered to have influenced the present results.

A potential weakness of our study is that the information on dietary habits was
unavailable at the time of the analyses. Recent studies have shown that
confounding can be caused by dietary factors, especially isothiocyanates
present in broccoli. They are conjugated by the GSTs and have therefore
been suggested to contribute to the inconsistencies seen in studies on the
role of these enzymes in colorectal adenoma (80) and lung cancer (271).

64
It should also be noted that given the numerous comparisons performed, the
possibility of a chance finding cannot be ruled out. By definition 5% (with 95%
CI) of all results may be due to chance. Corrections for multiple comparisons
have not been used in the present studies. However, analyses were mainly
motivated by D SULRUL hypothesis, and overall more consideration should be
given to causality and biological plausibility of the results than simply to level
of statistical significance. In general, lack of dose-response and weakness of
the association argues against causality (261). Furthermore, no study should
be regarded in isolation but must be viewed in the light of other studies on the
area.

&RQFOXGLQJUHPDUNV

Our results agree with the suggestions that low-penetrance genes may modify
individual susceptibility to breast cancer. Our novel finding that polymorphic
oestrogen metabolising genes could affect breast cancer risk associated with
the use of HRT was especially intriguing. If this should be confirmed in future
studies genotyping for these genes could be used to identify susceptible
women. Moreover, our results support the proposed role of life style factors in
the development of breast cancer. Since the use of alcohol, smoking, and
especially the use of hormones has been increasing among women in
western countries, the increasing incidence of breast cancer could, at least
partly, be due to these factors acting together with specific host factors. These
figures are also favourable in the sense of prevention.

Most of the associations found in this study were restricted to

postmenopausal women. A possible explanation is the lower power to detect
moderate effects due to fewer premenopausal subjects being included in the
study. However, it is probable that the greater proportion of the
premenopausal breast cancer cases is due to some inherited mutation in
high-penetrance genes yet to be identified. Low-penetrance genes are
generally associated with carcinomas detected at later age (81). In contrast,
the combined effects of the GSTs were only seen among premenopausal
women. It is tempting to speculate that in these women carrying
simultaneously multiple low-penetrance at-risk genotypes the genetic burden
resembles that of high-penetrance genes, yielding diagnosis at an earlier age.
However, these results remain to be confirmed in future studies.

It should be noted that often genotype alone usually is an incomplete measure

of the phenotypic effect. Furthermore, levels of the enzyme activity may also
be affected by a number of endogenous and exogenous factors. The effect of
other known genetic polymorphisms in enzymes involved in the steroid
synthesis (LH, CYP19), and metabolism (HJ, CYP1A1 and CYP1B1),
together with the yet unidentified genes, may further complicate the issue. As
described, the strength of the association is an important causal criterion that
may be more difficult to meet with a multifactorial disease such as breast
cancer (87). Therefore, even larger, well-designed studies are needed to

65
prove the present results as well as to study the combined effects of multiple
separate genotypes. The novel high throughput analyses, which will soon be
in routine use, will undoubtedly be of great help in these future studies.

66
$&.12:/('*(0(176
This work was carried out in the Finnish Institute of Occupational Health. I
wish to thank Professor Jorma Rantanen, the Director General of the Institute,
the Head of the Department of Industrial Hygiene and Toxicology, Professor
Kai Savolainen, and the Head of the Laboratory of Molecular and Cellular
Toxicology, Dr. Hannu Norppa for placing the excellent facilities at my
disposal.

I wish to thank Professor Carl G. Gahmberg, the Head of the Division of

Biochemistry, University of Helsinki.

The official referees of my thesis, Prof. Kaija Holli and Dr. Angela Risch are
gratefully acknowledged for their careful review of this thesis.

My deepest thanks goes to my supervisor Dr. Ari Hirvonen for offering me an

opportunity to work on an interesting project in relaxed atmosphere, and for
providing excellent resources for it. I’m grateful for all his help and most of all
for his support and encouragement to keep me going through difficulties.

I also wish to thank:

The co-authors of the original publications; Prof. Matti Eskelinen, Prof.

Daehee Kang, Prof. Veli-Matti Kosma, Prof. Matti Uusitupa, Prof. Harri Vainio,
and especially Dr. Vesa Kataja who has promptly and patiently answered to
my never-ending questions throughout the study. Dr. Simone Benhamou and
Dr. Nadejda Jourenkova for taking me to the world of epidemiology and
statistics during my visit to INSERM (Villejuif, France) in the spring 1999. Pia
Sillanpää for help with genotyping and especially for the refreshing talks late
nights and weekends at work.

All the people in the lab with whom I had an opportunity to work during these
years. Especially Sirkku who originally “fooled” me to join the group, Haiju and
Anu for sharing the office and therefore the numerous personal and work-
related discussions, as well as Maria and Jarno for sharing our tiny old lab in
the good spirits.

ALL the people in the department who have helped me in many ways during
these years.

All my friends outside work for their friendship.

My brother for being the best brother in the world, father, grandmother and
especially my mother for her love and support throughout the years.

The financial support of the Academy of Finland, Finnish Konkordia

Association, and Cancer Society of Finland, is gratefully acknowledged.

67
5()(5(1&(6
1. McPherson, K., Steel C.M. and Dixon J.M. (2000) ABC of breast
diseases. Breast cancer-epidemiology, risk factors, and genetics. %PM,
, 624-8.
2. Lichtenstein, P., Holm N.V., Verkasalo P.K., Iliadou A., Kaprio J.,
Koskenvuo M., Pukkala E., Skytthe A. and Hemminki K. (2000)
Environmental and heritable factors in the causation of cancer--
analyses of cohorts of twins from Sweden, Denmark, and Finland. 1
(QJO-0HG, , 78-85.
3. Easton, D., Ford D. and Peto J. (1993) Inherited susceptibility to breast
cancer. &DQFHU6XUY, , 95-113.
4. Oesterreich, S. and Fuqua S.A. (1999) Tumor suppressor genes in
breast cancer. (QGRFU5HODW&DQFHU, , 405-19.
5. Syrjakoski, K., Vahteristo P., Eerola H., Tamminen A., Kivinummi K.,
Sarantaus L., Holli K., Blomqvist C., Kallioniemi O.P., Kainu T. and
Nevanlinna H. (2000) Population-Based Study of BRCA1 and BRCA2
Mutations in 1035 Unselected Finnish Breast Cancer Patients. - 1DWO
&DQFHU,QVW, , 1529-1531.
6. Johnson-Thompson, M.C. and Guthrie J. (2000) Ongoing research to
identify environmental risk factors in breast carcinoma. &DQFHU, ,
1224-9.
7. Emerit, I. (1994) Reactive oxygen species, chromosome mutation, and
cancer: possible role of clastogenic factors in carcinogenesis. )UHH
5DGLF%LRO0HG, , 99-109.
8. Nandi, S., Guzman R.C. and Yang J. (1995) Hormones and mammary
carcinogenesis in mice, rats, and humans: a unifying hypothesis. 3URF
1DWO$FDG6FL86$, , 3650-7.
9. Jefcoate, C.R., Liehr J.G., Santen R.J., Sutter T.R., Yager J.D., Yue
W., Santner S.J., Tekmal R., Demers L., Pauley R., Naftolin F., Mor G.
and Berstein L. (2000) Tissue-specific synthesis and oxidative
metabolism of estrogens [In Process Citation]. - 1DWO &DQFHU ,QVW
0RQRJU, , 95-112.
10. Yager, J.D. (2000) Endogenous estrogens as carcinogens through
metabolic activation [In Process Citation]. - 1DWO &DQFHU ,QVW 0RQRJU,
, 67-73.
11. Dunning, A.M., Healey C.S., Pharoah P.D., Teare M.D., Ponder B.A.
and Easton D.F. (1999) A systematic review of genetic polymorphisms
and breast cancer risk. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 843-54.
12. Henderson, B.E. and Feigelson H.S. (2000) Hormonal carcinogenesis.
&DUFLQRJHQHVLV, , 427-33.
13. Parkin, D.M. (1998) Epidemiology of cancer: global patterns and
trends. 7R[LFRO/HWW, , 227-34.
14. Finnish Cancer Registry Cancer Incidence in Finland in 1997.
www.cancerregistry.fi.
15. Mettlin, C. (1999) Global breast cancer mortality statistics. &$&DQFHU-
&OLQ, , 138-44.
16. Dickman, P.W., Hakulinen T., Luostarinen T., Pukkala E., Sankila R.,
Soderman B. and Teppo L. (1999) Survival of cancer patients in
Finland 1955-1994. $FWD2QFRO, , 1-103.

68
17. Russo, J., Hu Y.F., Yang X. and Russo I.H. (2000) Developmental,
cellular, and molecular basis of human breast cancer. - 1DWO &DQFHU
,QVW0RQRJU, , 17-37.
18. Frykberg, E.R. (1999) Lobular carcinoma in situ of the breast. 7KH
EUHDVWMRXUQDO, , 296-302.
19. Silva, O.E. and Zurrida S. (1999) Breast cancer: a guide for fellows.
&ULW5HY2QFRO+HPDWRO, , 1-207.
20. Dorgan, J.F., Longcope C., Stephenson H.E., Jr., Falk R.T., Miller R.,
Franz C., Kahle L., Campbell W.S., Tangrea J.A. and Schatzkin A.
(1997) Serum sex hormone levels are related to breast cancer risk in
postmenopausal women. (QYLURQ+HDOWK3HUVSHFW,  6XSSO, 583-
5.
21. Key, T.J. (1999) Serum oestradiol and breast cancer risk. (QGRFU5HODW
&DQFHU, , 175-80.
22. Kabuto, M., Akiba S., Stevens R.G., Neriishi K. and Land C.E. (2000) A
prospective study of estradiol and breast cancer in Japanese women.
&DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 575-9.
23. Bernstein, L. and Ross R.K. (1993) Endogenous hormones and breast
cancer risk. (SLGHPLRO5HY, , 48-65.
24. Feigelson, H.S. and Henderson B.E. (1996) Estrogens and breast
cancer. &DUFLQRJHQHVLV, , 2279-84.
25. Berkey, C.S., Frazier A.L., Gardner J.D. and Colditz G.A. (1999)
Adolescence and breast carcinoma risk. &DQFHU, , 2400-9.
26. Apter, D., Reinila M. and Vihko R. (1989) Some endocrine
characteristics of early menarche, a risk factor for breast cancer, are
preserved into adulthood. ,QW-&DQFHU, , 783-7.
27. Bernstein, L., Henderson B.E., Hanisch R., Sullivan-Halley J. and Ross
R.K. (1994) Physical exercise and reduced risk of breast cancer in
young women. -1DWO&DQFHU,QVW, , 1403-8.
28. Kelsey, J.L., Gammon M.D. and John E.M. (1993) Reproductive factors
and breast cancer. (SLGHPLRO5HY, , 36-47.
29. Henderson, B.E., Ross R.K., Judd H.L., Krailo M.D. and Pike M.C.
(1985) Do regular ovulatory cycles increase breast cancer risk?
&DQFHU, , 1206-8.
30. Collaborative Group on Hormonal Factors in Breast Cancer (1996)
Breast cancer and hormonal contraceptives: collaborative reanalysis of
individual data on 53 297 women with breast cancer and 100 239
women without breast cancer from 54 epidemiological studies. /DQFHW,
, 1713-27.
31. Chie, W.C., Hsieh C., Newcomb P.A., Longnecker M.P., Mittendorf R.,
Greenberg E.R., Clapp R.W., Burke K.P., Titus-Ernstoff L., Trentham-
Dietz A. and MacMahon B. (2000) Age at any full-term pregnancy and
breast cancer risk. $P-(SLGHPLRO, , 715-22.
32. Pathak, D.R., Osuch J.R. and He J. (2000) Breast carcinoma etiology:
current knowledge and new insights into the effects of reproductive and
hormonal risk factors in black and white populations. &DQFHU, , 1230-
8.
33. Lipworth, L., Bailey L.R. and Trichopoulos D. (2000) History of breast-
feeding in relation to breast cancer risk: a review of the epidemiologic
literature. -1DWO&DQFHU,QVW, , 302-12.

69
34. Huang, W.Y., Newman B., Millikan R.C., Schell M.J., Hulka B.S. and
Moorman P.G. (2000) Hormone-related factors and risk of breast
cancer in relation to estrogen receptor and progesterone receptor
status. $P-(SLGHPLRO, , 703-14.
35. Enger, S.M., Ross R.K., Paganini-Hill A., Carpenter C.L. and Bernstein
L. (2000) Body size, physical activity, and breast cancer hormone
receptor status: results from two case-control studies. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 681-7.
36. Hunter, D.J. and Willett W.C. (1993) Diet, body size, and breast
cancer. (SLGHPLRO5HY, , 110-32.
37. Yue, W., Santen R.J., Wang J.P., Hamilton C.J. and Demers L.M.
(1999) Aromatase within the breast. (QGRFU5HODW&DQFHU, , 157-64.
38. Ballard-Barbash, R. (1994) Anthropometry and breast cancer. Body
size--a moving target. &DQFHU, , 1090-100.
39. Mannisto, S., Pietinen P., Pyy M., Palmgren J., Eskelinen M. and
Uusitupa M. (1996) Body-size indicators and risk of breast cancer
according to menopause and estrogen-receptor status. ,QW - &DQFHU,
, 8-13.
40. Hall, I.J., Newman B., Millikan R.C. and Moorman P.G. (2000) Body
size and breast cancer risk in black women and white women: the
Carolina Breast Cancer Study. $P-(SLGHPLRO, , 754-64.
41. De Cree, C., Van Kranenburg G., Geurten P., Fujimori Y. and Keizer
H.A. (1997) 4-Hydroxycatecholestrogen metabolism responses to
exercise and training: possible implications for menstrual cycle
irregularities and breast cancer. )HUWLO6WHULO, , 505-16.
42. Collaborative Group on Hormonal Factors in Breast Cancer (1997)
Breast cancer and hormone replacement therapy: collaborative
reanalysis of data from 51 epidemiological studies of 52,705 women
with breast cancer and 108,411 women without breast cancer. /DQFHW,
, 1047-59.
43. Schairer, C., Lubin J., Troisi R., Sturgeon S., Brinton L. and Hoover R.
(2000) Estrogen-progestin replacement and risk of breast cancer.
-DPD, , 691-4.
44. Holli, K., Isola J. and Cuzick J. (1997) Hormone replacement therapy
and biological aggressiveness of breast cancer. /DQFHW, , 1704-5.
45. Stallard, S., Litherland J.C., Cordiner C.M., Dobson H.M., George
W.D., Mallon E.A. and Hole D. (2000) Effect of hormone replacement
therapy on the pathological stage of breast cancer: population based,
cross sectional study. %PM, , 348-9.
46. Grodstein, F., Stampfer M.J., Colditz G.A., Willett W.C., Manson J.E.,
Joffe M., Rosner B., Fuchs C., Hankinson S.E., Hunter D.J.,
Hennekens C.H. and Speizer F.E. (1997) Postmenopausal hormone
therapy and mortality. 1(QJO-0HG, , 1769-75.
47. Dees, C., Foster J.S., Ahamed S. and Wimalasena J. (1997) Dietary
estrogens stimulate human breast cells to enter the cell cycle. (QYLURQ
+HDOWK3HUVSHFW, 6XSSO, 633-6.
48. Safe, S.H. (1997) Is there an association between exposure to
environmental estrogens and breast cancer? (QYLURQ+HDOWK3HUVSHFW,
6XSSO, 675-8.

70
49. Davis, D.L., Bradlow H.L., Wolff M., Woodruff T., Hoel D.G. and Anton-
Culver H. (1993) Medical hypothesis: xenoestrogens as preventable
causes of breast cancer. (QYLURQ+HDOWK3HUVSHFW, , 372-7.
50. Spink, B.C., Katz B.H., Hussain M.M., Pang S., Connor S.P., Aldous
K.M., Gierthy J.F. and Spink D.C. (2000) SULT1A1 catalyzes 2-
methoxyestradiol sulfonation in MCF-7 breast cancer cells.
&DUFLQRJHQHVLV, , 1947-57.
51. Garner, C.E., Burka L.T., Etheridge A.E. and Matthews H.B. (2000)
Catechol metabolites of polychlorinated biphenyls inhibit the catechol-
O-methyltransferase-mediated metabolism of catechol estrogens.
7R[LFRO$SSO3KDUPDFRO, , 115-23.
52. Garner, C.E., Matthews H.B. and Burka L.T. (2000) Phenolphthalein
metabolite inhibits catechol-O-methyltransferase- mediated metabolism
of catechol estrogens: a possible mechanism for carcinogenicity.
7R[LFRO$SSO3KDUPDFRO, , 124-31.
53. Palmer, J.R. and Rosenberg L. (1993) Cigarette smoking and the risk
of breast cancer. (SLGHPLRO5HY, , 145-56.
54. Ambrosone, C.B. and Shields P.G. (1999) Smoking as a risk factor for
breast cancer. In Bowcock, A.M. (ed.), %UHDVW FDQFHU 0ROHFXODU
JHQHWLFV3DWKRJHQHVLVDQG7KHUDSHXWLFV Human Press Inc., Totowa,
New Jersey, pp. 519-36.
55. Wartenberg, D., Calle E.E., Thun M.J., Heath C.W., Jr., Lally C. and
Woodruff T. (2000) Passive smoking exposure and female breast
cancer mortality. -1DWO&DQFHU,QVW, , 1666-73.
56. Baron, J.A., La Vecchia C. and Levi F. (1990) The antiestrogenic effect
of cigarette smoking in women. $P-2EVWHW*\QHFRO, , 502-14.
57. Bartsch, H., Nair U., Risch A., Rojas M., Wikman H. and Alexandrov K.
(2000) Genetic polymorphism of CYP genes, alone or in combination,
as a risk modifier of tobacco-related cancers. &DQFHU (SLGHPLRO
%LRPDUNHUV3UHY, , 3-28.
58. Li, D., Wang M., Dhingra K. and Hittelman W.N. (1996) Aromatic DNA
adducts in adjacent tissues of breast cancer patients: clues to breast
cancer etiology. &DQFHU5HV, , 287-93.
59. Perera, F.P., Estabrook A., Hewer A., Channing K., Rundle A., Mooney
L.A., Whyatt R. and Phillips D.H. (1995) Carcinogen-DNA adducts in
human breast tissue. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 233-8.
60. Rundle, A., Tang D., Hibshoosh H., Estabrook A., Schnabel F., Cao
W., Grumet S. and Perera F.P. (2000) The relationship between
genetic damage from polycyclic aromatic hydrocarbons in breast tissue
and breast cancer. &DUFLQRJHQHVLV, , 1281-9.
61. Hein, D.W., Doll M.A., Fretland A.J., Leff M.A., Webb S.J., Xiao G.H.,
Devanaboyina U.S., Nangju N.A. and Feng Y. (2000) Molecular
genetics and epidemiology of the NAT1 and NAT2 acetylation
polymorphisms. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 29-42.
62. Pfau, W., Stone E.M., Brockstedt U., Carmichael P.L., Marquardt H.
and Phillips D.H. (1998) DNA adducts in human breast tissue:
association with N- acetyltransferase-2 (NAT2) and NAT1 genotypes.
&DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 1019-25.

71
63. Gilks, C.B., Price K., Wright J.L. and Churg A. (1998) Antioxidant gene
expression in rat lung after exposure to cigarette smoke. $P-3DWKRO,
, 269-78.
64. Pryor, W.A. (1997) Cigarette smoke radicals and the role of free
radicals in chemical carcinogenicity. (QYLURQ +HDOWK 3HUVSHFW, 
6XSSO, 875-82.
65. Marnett, L.J. (2000) Oxyradicals and DNA damage. &DUFLQRJHQHVLV,
, 361-70.
66. Smith-Warner, S.A., Spiegelman D., Yaun S.S., van den Brandt P.A.,
Folsom A.R., Goldbohm R.A., Graham S., Holmberg L., Howe G.R.,
Marshall J.R., Miller A.B., Potter J.D., Speizer F.E., Willett W.C., Wolk
A. and Hunter D.J. (1998) Alcohol and breast cancer in women: a
pooled analysis of cohort studies. -DPD, , 535-40.
67. Kuper, H., Ye W., Weiderpass E., Ekbom A., Trichopoulos D., Nyren O.
and Adami H.O. (2000) Alcohol and breast cancer risk: the alcoholism
paradox. %U-&DQFHU, , 949-51.
68. Colditz, G.A. and Frazier A.L. (1995) Models of breast cancer show
that risk is set by events of early life: prevention efforts must shift focus.
&DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 567-71.
69. Reichman, M.E., Judd J.T., Longcope C., Schatzkin A., Clevidence
B.A., Nair P.P., Campbell W.S. and Taylor P.R. (1993) Effects of
alcohol consumption on plasma and urinary hormone concentrations in
premenopausal women. -1DWO&DQFHU,QVW, , 722-7.
70. Wright, R.M., McManaman J.L. and Repine J.E. (1999) Alcohol-
induced breast cancer: a proposed mechanism. )UHH 5DGLF%LRO 0HG,
, 348-54.
71. Barnes, S.L., Singletary K.W. and Frey R. (2000) Ethanol and
acetaldehyde enhance benzo[a]pyrene-DNA adduct formation in
human mammary epithelial cells. &DUFLQRJHQHVLV, , 2123-8.
72. Sugimura, T. (2000) Nutrition and dietary carcinogens. &DUFLQRJHQHVLV,
, 387-95.
73. Nair, J., Barbin A., Velic I. and Bartsch H. (1999) Etheno DNA-base
adducts from endogenous reactive species. 0XWDW5HV, , 59-69.
74. Bartsch, H., Nair J. and Owen R.W. (1999) Dietary polyunsaturated
fatty acids and cancers of the breast and colorectum: emerging
evidence for their role as risk modifiers. &DUFLQRJHQHVLV, , 2209-18.
75. Lee, I.M. (1999) Antioxidant vitamins in the prevention of cancer. 3URF
$VVRF$P3K\VLFLDQV, , 10-5.
76. McKeown, N. (1999) Antioxidants and breast cancer. 1XWU 5HY, ,
321-4.
77. Zheng, W., Gustafson D.R., Sinha R., Cerhan J.R., Moore D., Hong
C.P., Anderson K.E., Kushi L.H., Sellers T.A. and Folsom A.R. (1998)
Well-done meat intake and the risk of breast cancer. - 1DWO &DQFHU
,QVW, , 1724-9.
78. Hein, D.W. (2000) N-Acetyltransferase genetics and their role in
predisposition to aromatic and heterocyclic amine-induced
carcinogenesis. 7R[LFRO/HWW, , 349-56.
79. Bouker, K.B. and Hilakivi-Clarke L. (2000) Genistein: Does it Prevent or
Promote Breast Cancer? (QYLURQ+HDOWK3HUVSHFW, , 701-708.

72
80. Lin, H.J., Probst-Hensch N.M., Louie A.D., Kau I.H., Witte J.S., Ingles
S.A., Frankl H.D., Lee E.R. and Haile R.W. (1998) Glutathione
transferase null genotype, broccoli, and lower prevalence of colorectal
adenomas. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 647-52.
81. Rebbeck, T.R. (1999) Inherited genetic predisposition in breast cancer.
A population-based perspective. &DQFHU, , 2493-501.
82. Cui, J. and Hopper J.L. (2000) Why are the majority of hereditary cases
of early-onset breast cancer sporadic? A simulation study. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 805-12.
83. Unger, M.A. and Weber B.L. (2000) Recent advances in breast cancer
biology. &XUU2SLQ2QFRO, , 521-5.
84. Ford, D., Easton D.F., Stratton M., Narod S., Goldgar D., Devilee P.,
Bishop D.T., Weber B., Lenoir G., Chang-Claude J., Sobol H., Teare
M.D., Struewing J., Arason A., Scherneck S., Peto J., Rebbeck T.R.,
Tonin P., Neuhausen S., Barkardottir R., Eyfjord J., Lynch H., Ponder
B.A., Gayther S.A., Zelada-Hedman M. et al. (1998) Genetic
heterogeneity and penetrance analysis of the BRCA1 and BRCA2
genes in breast cancer families. The Breast Cancer Linkage
Consortium. $P-+XP*HQHW, , 676-89.
85. Kainu, T., Juo S.H., Desper R., Schaffer A.A., Gillanders E.,
Rozenblum E., Freas-Lutz D., Weaver D., Stephan D., Bailey-Wilson
J., Kallioniemi O.P., Tirkkonen M., Syrjakoski K., Kuukasjarvi T.,
Koivisto P., Karhu R., Holli K., Arason A., Johannesdottir G.,
Bergthorsson J.T., Johannsdottir H., Egilsson V., Barkardottir R.B.,
Johannsson O., Haraldsson K., Sandberg T., Holmberg E., Gronberg
H., Olsson H., Borg A., Vehmanen P., Eerola H., Heikkila P., Pyrhonen
S. and Nevanlinna H. (2000) Somatic deletions in hereditary breast
cancers implicate 13q21 as a putative novel breast cancer
susceptibility locus. 3URF1DWO$FDG6FL86$, , 9603-8.
86. Vehmanen, P., Friedman L.S., Eerola H., McClure M., Ward B.,
Sarantaus L., Kainu T., Syrjakoski K., Pyrhonen S., Kallioniemi O.P.,
Muhonen T., Luce M., Frank T.S. and Nevanlinna H. (1997) Low
proportion of BRCA1 and BRCA2 mutations in Finnish breast cancer
families: evidence for additional susceptibility genes. +XP 0RO *HQHW,
, 2309-15.
87. Coughlin, S.S. and Piper M. (1999) Genetic polymorphisms and risk of
breast cancer. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 1023-32.
88. Kristensen, V.N. and Borresen-Dale A.L. (2000) Molecular
epidemiology of breast cancer: genetic variation in steroid hormone
metabolism. 0XWDW5HV, , 323-33.
89. Ambrosone, C.B., Freudenheim J.L., Thompson P.A., Bowman E.,
Vena J.E., Marshall J.R., Graham S., Laughlin R., Nemoto T. and
Shields P.G. (1999) Manganese superoxide dismutase (MnSOD)
genetic polymorphisms, dietary antioxidants, and risk of breast cancer.
&DQFHU5HV, , 602-6.
90. Schmutzler, R.K., Sanchez M., Lehrer S., Chaparro C.A., Phillips C.,
Rabin J. and Schachter B. (1991) Incidence of an estrogen receptor
polymorphism in breast cancer patients. %UHDVW&DQFHU5HV7UHDW, ,
111-7.

73
91. Wang-Gohrke, S., Chang-Claude J., Becher H., Kieback D.G. and
Runnebaum I.B. (2000) Progesterone receptor gene polymorphism is
associated with decreased risk for breast cancer by age 50. &DQFHU
5HV, , 2348-50.
92. Curran, J.E., Vaughan T., Lea R.A., Weinstein S.R., Morrison N.A. and
Griffiths L.R. (1999) Association of A vitamin D receptor polymorphism
with sporadic breast cancer development. ,QW-&DQFHU, , 723-6.
93. Lundin, A.C., Soderkvist P., Eriksson B., Bergman-Jungestrom M. and
Wingren S. (1999) Association of breast cancer progression with a
vitamin D receptor gene polymorphism. South-East Sweden Breast
Cancer Group. &DQFHU5HV, , 2332-4.
94. Eisman, J.A., Suva L.J. and Martin T.J. (1986) Significance of 1,25-
dihydroxyvitamin D3 receptor in primary breast cancers. &DQFHU 5HV,
, 5406-8.
95. Shen, M.R., Jones I.M. and Mohrenweiser H. (1998) Nonconservative
amino acid substitution variants exist at polymorphic frequency in DNA
repair genes in healthy humans. &DQFHU5HV, , 604-8.
96. John, E.M. and Kelsey J.L. (1993) Radiation and other environmental
exposures and breast cancer. (SLGHPLRO5HY, , 157-62.
97. McCormick, B. (1999) Radiation and local control in early invasive
breast cancer. 7KHEUHDVWMRXUQDO, , 330-4.
98. Weiderpass, E., Pukkala E., Kauppinen T., Mutanen P., Paakkulainen
H., Vasama-Neuvonen K., Boffetta P. and Partanen T. (1999) Breast
cancer and occupational exposures in women in Finland. $P - ,QG
0HG, , 48-53.
99. Kheifets, L.I. and Matkin C.C. (1999) Industrialization, electromagnetic
fields, and breast cancer risk. (QYLURQ +HDOWK 3HUVSHFW,  6XSSO ,
145-54.
100. Goldberg, M.S. and Labreche F. (1996) Occupational risk factors for
female breast cancer: a review. 2FFXS(QYLURQ0HG, , 145-56.
101. Welp, E.A., Weiderpass E., Boffetta P., Vainio H., Vasama-Neuvonen
K., Petralia S. and Partanen T.J. (1998) Environmental risk factors of
breast cancer. 6FDQG-:RUN(QYLURQ+HDOWK, , 3-7.
102. Pukkala, E. and Weiderpass E. (1999) Time trends in socio-economic
differences in incidence rates of cancers of the breast and female
genital organs (Finland, 1971-1995). ,QW-&DQFHU, , 56-61.
103. Heck, K.E. and Pamuk E.R. (1997) Explaining the relation between
education and postmenopausal breast cancer. $P - (SLGHPLRO, ,
366-72.
104. Parl, F., F. (2000) Estrogen, Estrogen Receptor and Breast Cancer,
%LRPHGLFDODQG+HDOWK5HVHDUFK. Vol. 36. IOS Press, Amsterdam, pp.
1-263.
105. Dickson, R.B. and Stancel G.M. (2000) Estrogen receptor-mediated
processes in normal and cancer cells [In Process Citation]. - 1DWO
&DQFHU,QVW0RQRJU, , 135-45.
106. Liehr, J.G. (2000) Is estradiol a genotoxic mutagenic carcinogen?
(QGRFU5HY, , 40-54.
107. Liehr, J.G., Fang W.F., Sirbasku D.A. and Ari-Ulubelen A. (1986)
Carcinogenicity of catechol estrogens in Syrian hamsters. - 6WHURLG
%LRFKHP, , 353-6.

74
108. Zhu, B.T. and Conney A.H. (1998) Functional role of estrogen
metabolism in target cells: review and perspectives. &DUFLQRJHQHVLV,
, 1-27.
109. IARC (1987) Overall Evaluations of Carcinogenicity: An Updating of
IARC Monographs Volumes 1 to 42., 6XSSO.
110. Cavalieri, E.L., Stack D.E., Devanesan P.D., Todorovic R., Dwivedy I.,
Higginbotham S., Johansson S.L., Patil K.D., Gross M.L., Gooden J.K.,
Ramanathan R., Cerny R.L. and Rogan E.G. (1997) Molecular origin of
cancer: catechol estrogen-3,4-quinones as endogenous tumor
initiators. 3URF1DWO$FDG6FL86$, , 10937-42.
111. Matsui, A., Ikeda T., Enomoto K., Hosoda K., Nakashima H., Omae K.,
Watanabe M., Hibi T. and Kitajima M. (2000) Increased formation of
oxidative DNA damage, 8-hydroxy-2’- deoxyguanosine, in human
breast cancer tissue and its relationship to GSTP1 and COMT
genotypes. &DQFHU/HWW, , 87-95.
112. Yoshie, Y. and Ohshima H. (1998) Synergistic induction of DNA strand
breakage by catechol-estrogen and nitric oxide: implications for
hormonal carcinogenesis. )UHH5DGLF%LRO0HG, , 341-8.
113. Mobley, J.A., Bhat A.S. and Brueggemeier R.W. (1999) Measurement
of oxidative DNA damage by catechol estrogens and analogues in
vitro. &KHP5HV7R[LFRO, , 270-7.
114. Grodin, J.M., Siiteri P.K. and MacDonald P.C. (1973) Source of
estrogen production in postmenopausal women. - &OLQ (QGRFULQRO
0HWDE, , 207-14.
115. MacDonald, P.C., Edman C.D., Hemsell D.L., Porter J.C. and Siiteri
P.K. (1978) Effect of obesity on conversion of plasma androstenedione
to estrone in postmenopausal women with and without endometrial
cancer. $P-2EVWHW*\QHFRO, , 448-55.
116. Siiteri, P.K. (1987) Adipose tissue as a source of hormones. $P-&OLQ
1XWU, , 277-82.
117. Omura, T. and Morohashi K. (1995) Gene regulation of
steroidogenesis. -6WHURLG%LRFKHP0RO%LRO, , 19-25.
118. Thompson, P.A. and Ambrosone C. (2000) Molecular epidemiology of
genetic polymorphisms in estrogen metabolizing enzymes in human
breast cancer. -1DWO&DQFHU,QVW0RQRJU, , 125-34.
119. Lippert, T.H., Seeger H. and Mueck A.O. (2000) The impact of
endogenous estradiol metabolites on carcinogenesis. 6WHURLGV, ,
357-69.
120. Hayes, C.L., Spink D.C., Spink B.C., Cao J.Q., Walker N.J. and Sutter
T.R. (1996) 17 beta-estradiol hydroxylation catalyzed by human
cytochrome P450 1B1. 3URF1DWO$FDG6FL86$, , 9776-81.
121. Stoilov, I., Akarsu A.N., Alozie I., Child A., Barsoum-Homsy M., Turacli
M.E., Or M., Lewis R.A., Ozdemir N., Brice G., Aktan S.G., Chevrette
L., Coca-Prados M. and Sarfarazi M. (1998) Sequence analysis and
homology modeling suggest that primary congenital glaucoma on 2p21
results from mutations disrupting either the hinge region or the
conserved core structures of cytochrome P4501B1. $P-+XP*HQHW,
, 573-84.

75
122. Zhu, B.T. and Conney A.H. (1998) Is 2-methoxyestradiol an
endogenous estrogen metabolite that inhibits mammary
carcinogenesis? &DQFHU5HV, , 2269-77.
123. Kabat, G.C., Chang C.J., Sparano J.A., Sepkovie D.W., Hu X.P., Khalil
A., Rosenblatt R. and Bradlow H.L. (1997) Urinary estrogen
metabolites and breast cancer: a case-control study. &DQFHU(SLGHPLRO
%LRPDUNHUV3UHY, , 505-9.
124. Adlercreutz, H., Gorbach S.L., Goldin B.R., Woods M.N., Dwyer J.T.
and Hamalainen E. (1994) Estrogen metabolism and excretion in
Oriental and Caucasian women. -1DWO&DQFHU,QVW, , 1076-82.
125. Ursin, G., London S., Stanczyk F.Z., Gentzschein E., Paganini-Hill A.,
Ross R.K. and Pike M.C. (1999) Urinary 2-hydroxyestrone/16alpha-
hydroxyestrone ratio and risk of breast cancer in postmenopausal
women. -1DWO&DQFHU,QVW, , 1067-72.
126. Ursin, G., London S., Stanczyk F.Z., Gentzschein E., Paganini-Hill A.,
Ross R.K. and Pike M.C. (1997) A pilot study of urinary estrogen
metabolites (16alpha-OHE1 and 2-OHE1) in postmenopausal women
with and without breast cancer. (QYLURQ+HDOWK3HUVSHFW, 6XSSO,
601-5.
127. Raftogianis, R., Creveling C., Weinshilboum R. and Weisz J. (2000)
Estrogen metabolism by conjugation. - 1DWO &DQFHU ,QVW 0RQRJU, ,
113-24.
128. Wang, M.Y. and Liehr J.G. (1995) Induction by estrogens of lipid
peroxidation and lipid peroxide-derived malonaldehyde-DNA adducts in
male Syrian hamsters: role of lipid peroxidation in estrogen-induced
kidney carcinogenesis. &DUFLQRJHQHVLV, , 1941-5.
129. Cavalieri, E., Frenkel K., Liehr J.G., Rogan E. and Roy D. (2000)
Estrogens as endogenous genotoxic agents--DNA adducts and
mutations. -1DWO&DQFHU,QVW0RQRJU, , 75-93.
130. Marnett, L.J. (1999) Lipid peroxidation-DNA damage by
malondialdehyde. 0XWDW5HV, , 83-95.
131. Butterworth, M., Lau S.S. and Monks T.J. (1997) Formation of catechol
estrogen glutathione conjugates and gamma- glutamyl transpeptidase-
dependent nephrotoxicity of 17beta-estradiol in the golden Syrian
hamster. &DUFLQRJHQHVLV, , 561-7.
132. Fridovich, I. (1995) Superoxide radical and superoxide dismutases.
$QQX5HY%LRFKHP, , 97-112.
133. Burdon, R.H. (1995) Superoxide and hydrogen peroxide in relation to
mammalian cell proliferation. )UHH5DGLF%LRO0HG, , 775-94.
134. Cadet, J., Delatour T., Douki T., Gasparutto D., Pouget J.P., Ravanat
J.L. and Sauvaigo S. (1999) Hydroxyl radicals and DNA base damage.
0XWDW5HV, , 9-21.
135. Feigelson, H.S., Coetzee G.A., Kolonel L.N., Ross R.K. and Henderson
B.E. (1997) A polymorphism in the CYP17 gene increases the risk of
breast cancer. &DQFHU5HV, , 1063-5.
136. Picado-Leonard, J. and Miller W.L. (1987) Cloning and sequence of the
human gene for P450c17 (steroid 17 alpha- hydroxylase/17,20 lyase):
similarity with the gene for P450c21. '1$, , 439-48.

76
137. Kristensen, V.N., Andersen T.I., Lindblom A., Erikstein B., Magnus P.
and Borresen-Dale A.L. (1998) A rare CYP19 (aromatase) variant may
increase the risk of breast cancer. 3KDUPDFRJHQHWLFV, , 43-8.
138. Hellmold, H., Rylander T., Magnusson M., Reihner E., Warner M. and
Gustafsson J.A. (1998) Characterization of cytochrome P450 enzymes
in human breast tissue from reduction mammaplasties. - &OLQ
(QGRFULQRO0HWDE, , 886-95.
139. Mannermaa, A., Peltoketo H., Winqvist R., Ponder B.A., Kiviniemi H.,
Easton D.F., Poutanen M., Isomaa V. and Vihko R. (1994) Human
familial and sporadic breast cancer: analysis of the coding regions of
the 17 beta-hydroxysteroid dehydrogenase 2 gene (EDH17B2) using a
single-strand conformation polymorphism assay. +XP*HQHW, , 319-
24.
140. Feigelson, H.S., McKean-Cowdin R., Coetzee G.A., Stram D.O.,
Kolonel. L.N. and Henderson B.E. (2001) Building a multigenic model
of breast cancer susceptibility: CYP17 and HSD17B1 are two important
candidates. &DQFHU5HVHDUFK, , 785-789.
141. Kawajiri, K., Nakachi K., Imai K., Yoshii A., Shinoda N. and Watanabe
J. (1990) Identification of genetically high risk individuals to lung cancer
by DNA polymorphisms of the cytochrome P450IA1 gene. )(%6 /HWW,
, 131-3.
142. Hayashi, S.I., Watanabe J., Nakachi K. and Kawajiri K. (1991) PCR
detection of an A/G polymorphism within exon 7 of the CYP1A1 gene.
1XFOHLF$FLGV5HV, , 4797.
143. Crofts, F., Cosma G.N., Currie D., Taioli E., Toniolo P. and Garte S.J.
(1993) A novel CYP1A1 gene polymorphism in African-Americans.
&DUFLQRJHQHVLV, , 1729-31.
144. Cascorbi I, B.J., Roots I. (1996) A C4887A polymorphism in exon 7 of
human CYP1A1: Population frequency, mutation linkages, and impact
on lung cancer susceptibility. &DQFHU5HV, , 4965-4969.
145. Schwarz, D., Kisselev P., Schunck W.H., Chernogolov A., Boidol W.,
Cascorbi I. and Roots I. (2000) Allelic variants of human cytochrome
P450 1A1 (CYP1A1): effect of T461N and I462V substitutions on
steroid hydroxylase specificity. 3KDUPDFRJHQHWLFV, , 519-30.
146. http://www.imm.ki.se/CYPalleles/cyp1b1.htm
147. Spink, D.C., Spink B.C., Cao J.Q., DePasquale J.A., Pentecost B.T.,
Fasco M.J., Li Y. and Sutter T.R. (1998) Differential expression of
CYP1A1 and CYP1B1 in human breast epithelial cells and breast
tumor cells. &DUFLQRJHQHVLV, , 291-8.
148. Taningher, M., Malacarne D., Izzotti A., Ugolini D. and Parodi S. (1999)
Drug metabolism polymorphisms as modulators of cancer
susceptibility. 0XWDW5HV, , 227-61.
149. Assicot, M., Contesso G. and Bohuon C. (1977) Catechol-O-
methyltransferase in human breast cancers. (XU-&DQFHU, , 961-6.
150. Tenhunen, J., Heikkila P., Alanko A., Heinonen E., Akkila J. and
Ulmanen I. (1999) Soluble and membrane-bound catechol-O-
methyltransferase in normal and malignant mammary gland. &DQFHU
/HWW, , 75-84.
151. Lachman, H.M., Papolos D.F., Saito T., Yu Y.M., Szumlanski C.L. and
Weinshilboum R.M. (1996) Human catechol-O-methyltransferase

77
 pharmacogenetics: description of a functional polymorphism and its
potential application to neuropsychiatric disorders. 3KDUPDFRJHQHWLFV,
, 243-50.
152. Seidegard, J., Vorachek W.R., Pero R.W. and Pearson W.R. (1988)
Hereditary differences in the expression of the human glutathione
transferase active on trans-stilbene oxide are due to a gene deletion.
3URF1DWO$FDG6FL86$, , 7293-7.
153. Pemble, S., Schroeder K.R., Spencer S.R., Meyer D.J., Hallier E., Bolt
H.M., Ketterer B. and Taylor J.B. (1994) Human glutathione S-
transferase theta (GSTT1): cDNA cloning and the characterization of a
genetic polymorphism. %LRFKHP-, , 271-6.
154. Ali-Osman, F., Akande O., Antoun G., Mao J.X. and Buolamwini J.
(1997) Molecular cloning, characterization, and expression in
Escherichia coli of full-length cDNAs of three human glutathione S-
transferase Pi gene variants. Evidence for differential catalytic activity
of the encoded proteins. -%LRO&KHP, , 10004-12.
155. Inskip, A., Elexperu-Camiruaga J., Buxton N., Dias P.S., MacIntosh J.,
Campbell D., Jones P.W., Yengi L., Talbot J.A., Strange R.C. and et al.
(1995) Identification of polymorphism at the glutathione S-transferase,
GSTM3 locus: evidence for linkage with GSTM1*A. %LRFKHP -, ,
713-6.
156. Cairns, J., Wright C., Cattan A.R., Hall A.G., Cantwell B.J., Harris A.L.
and Horne C.H. (1992) Immunohistochemical demonstration of
glutathione S-transferases in primary human breast carcinomas. -
3DWKRO, , 19-25.
157. Shea, T.C., Claflin G., Comstock K.E., Sanderson B.J., Burstein N.A.,
Keenan E.J., Mannervik B. and Henner W.D. (1990) Glutathione
transferase activity and isoenzyme composition in primary human
breast cancers. &DQFHU5HV, , 6848-53.
158. Forrester, L.M., Hayes J.D., Millis R., Barnes D., Harris A.L., Schlager
J.J., Powis G. and Wolf C.R. (1990) Expression of glutathione S-
transferases and cytochrome P450 in normal and tumor breast tissue.
&DUFLQRJHQHVLV, , 2163-70.
159. Albin, N., Massaad L., Toussaint C., Mathieu M.C., Morizet J., Parise
O., Gouyette A. and Chabot G.G. (1993) Main drug-metabolizing
enzyme systems in human breast tumors and peritumoral tissues.
&DQFHU5HV, , 3541-6.
160. Brentano, S.T., Picado-Leonard J., Mellon S.H., Moore C.C. and Miller
W.L. (1990) Tissue-specific, cyclic adenosine 3’,5’-monophosphate-
induced, and phorbol ester-repressed transcription from the human
P450c17 promoter in mouse cells. 0RO(QGRFULQRO, , 1972-9.
161. Carey, A.H., Waterworth D., Patel K., White D., Little J., Novelli P.,
Franks S. and Williamson R. (1994) Polycystic ovaries and premature
male pattern baldness are associated with one allele of the steroid
metabolism gene CYP17. +XP0RO*HQHW, , 1873-6.
162. Nedelcheva Kristensen, V., Haraldsen E.K., Anderson K.B., Lonning
P.E., Erikstein B., Karesen R., Gabrielsen O.S. and Borresen-Dale A.L.
(1999) CYP17 and breast cancer risk: the polymorphism in the 5’
flanking area of the gene does not influence binding to Sp-1. &DQFHU
5HV, , 2825-8.

78
163. Feigelson, H.S., Shames L.S., Pike M.C., Coetzee G.A., Stanczyk F.Z.
and Henderson B.E. (1998) Cytochrome P450c17alpha gene (CYP17)
polymorphism is associated with serum estrogen and progesterone
concentrations. &DQFHU5HV, , 585-7.
164. Haiman, C.A., Hankinson S.E., Spiegelman D., Colditz G.A., Willett
W.C., Speizer F.E., Kelsey K.T. and Hunter D.J. (1999) The
relationship between a polymorphism in CYP17 with plasma hormone
levels and breast cancer. &DQFHU5HV, , 1015-20.
165. Feigelson, H.S., McKean-Cowdin R., Pike M.C., Coetzee G.A., Kolonel
L.N., Nomura A.M., Le Marchand L. and Henderson B.E. (1999)
Cytochrome P450c17alpha gene (CYP17) polymorphism predicts use
of hormone replacement therapy. &DQFHU5HV, , 3908-10.
166. Young, I.E., Kurian K.M., Annink C., Kunkler I.H., Anderson V.A.,
Cohen B.B., Hooper M.L., Wyllie A.H. and Steel C.M. (1999) A
polymorphism in the CYP17 gene is associated with male breast
cancer. %U-&DQFHU, , 141-3.
167. Bergman-Jungestrom, M., Gentile M., Lundin A.C. and Wingren S.
(1999) Association between CYP17 gene polymorphism and risk of
breast cancer in young women. ,QW-&DQFHU, , 350-3.
168. Spurdle, A.B., Hopper J.L., Dite G.S., Chen X., Cui J., McCredie M.R.,
Giles G.G., Southey M.C., Venter D.J., Easton D.F. and Chenevix-
Trench G. (2000) CYP17 Promoter Polymorphism and Breast Cancer
in Australian Women Under Age Forty Years. - 1DWO &DQFHU ,QVW, ,
1674-1681.
169. Dunning, A.M., Healey C.S., Pharoah P.D., Foster N.A., Lipscombe
J.M., Redman K.L., Easton D.F., Day N.E. and Ponder B.A. (1998) No
association between a polymorphism in the steroid metabolism gene
CYP17 and risk of breast cancer. %U-&DQFHU, , 2045-7.
170. Helzlsouer, K.J., Huang H.Y., Strickland P.T., Hoffman S., Alberg A.J.,
Comstock G.W. and Bell D.A. (1998) Association between CYP17
polymorphisms and the development of breast cancer. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 945-9.
171. Huang, C.S., Chern H.D., Chang K.J., Cheng C.W., Hsu S.M. and
Shen C.Y. (1999) Breast cancer risk associated with genotype
polymorphism of the estrogen-metabolizing genes CYP17, CYP1A1,
and COMT: a multigenic study on cancer susceptibility. &DQFHU 5HV,
, 4870-5.
172. Weston, A., Pan C.F., Bleiweiss I.J., Ksieski H.B., Roy N., Maloney N.
and Wolff M.S. (1998) CYP17 genotype and breast cancer risk. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 941-4.
173. Kuligina, E.S., Togo A.V., Suspitsin E.N., Grigoriev M.Y., Pozharisskiy
K.M., Chagunava O.L., Berstein L.M., Theillet C., Hanson K.P. and
Imyanitov E.N. (2000) CYP17 polymorphism in the groups of distinct
breast cancer susceptibility: comparison of patients with the bilateral
disease vs. monolateral breast cancer patients vs. middle-aged female
controls vs. elderly tumor-free women. &DQFHU/HWW, , 45-50.
174. Healey, C.S., Dunning A.M., Durocher F., Teare D., Pharoah P.D.,
Luben R.N., Easton D.F. and Ponder B.A. (2000) Polymorphisms in the
human aromatase cytochrome P450 gene (CYP19) and breast cancer
risk. &DUFLQRJHQHVLV, , 189-93.

79
175. Siegelmann-Danieli, N. and Buetow K.H. (1999) Constitutional genetic
variation at the human aromatase gene (Cyp19) and breast cancer
risk. %U-&DQFHU, , 456-63.
176. Haiman, C.A., Hankinson S.E., Spiegelman D., De Vivo I., Colditz G.A.,
Willett W.C., Speizer F.E. and Hunter D.J. (2000) A tetranucleotide
repeat polymorphism in CYP19 and breast cancer risk. ,QW - &DQFHU,
, 204-10.
177. Probst-Hensch, N.M., Ingles S.A., Diep A.T., Haile R.W., Stanczyk
F.Z., Kolonel L.N. and Henderson B.E. (1999) Aromatase and breast
cancer susceptibility. (QGRFU5HODW&DQFHU, , 165-73.
178. Miyoshi, Y., Iwao K., Ikeda N., Egawa C. and Noguchi S. (2000) Breast
cancer risk associated with polymorphism in CYP19 in Japanese
women. ,QW-&DQFHU, , 325-8.
179. Winqvist, R., Peltoketo H., Isomaa V., Grzeschik K.H., Mannermaa A.
and Vihko R. (1990) The gene for 17 beta-hydroxysteroid
dehydrogenase maps to human chromosome 17, bands q12-q21, and
shows an RFLP with ScaI. +XP*HQHW, , 473-6.
180. Grossman, M.H., Emanuel B.S. and Budarf M.L. (1992) Chromosomal
mapping of the human catechol-O-methyltransferase gene to 22q11.1--
--q11.2. *HQRPLFV, , 822-5.
181. Merriam, G.R., MacLusky N.J., Picard M.K. and Naftolin F. (1980)
Comparative properties of the catechol estrogens, I: methylation by
catechol-O-methyltransferase and binding to cytosol estrogen
receptors. 6WHURLGV, , 1-11.
182. Weisz, J., Fritz-Wolz G., Gestl S., Clawson G.A., Creveling C.R., Liehr
J.G. and Dabbs D. (2000) Nuclear localization of catechol-O-
methyltransferase in neoplastic and nonneoplastic mammary epithelial
cells. $P-3DWKRO, , 1841-8.
183. Palmatier, M.A., Kang A.M. and Kidd K.K. (1999) Global variation in the
frequencies of functionally different catechol- O-methyltransferase
alleles. %LRO3V\FKLDWU\, , 557-67.
184. Lavigne, J.A., Helzlsouer K.J., Huang H.Y., Strickland P.T., Bell D.A.,
Selmin O., Watson M.A., Hoffman S., Comstock G.W. and Yager J.D.
(1997) An association between the allele coding for a low activity
variant of catechol-O-methyltransferase and the risk for breast cancer.
&DQFHU5HV, , 5493-7.
185. Millikan, R.C., Pittman G.S., Tse C.K., Duell E., Newman B., Savitz D.,
Moorman P.G., Boissy R.J. and Bell D.A. (1998) Catechol-O-
methyltransferase and breast cancer risk. &DUFLQRJHQHVLV, , 1943-7.
186. Matsui, A., Ikeda T., Enomoto K., Nakashima H., Omae K., Watanabe
M., Hibi T. and Kitajima M. (2000) Progression of human breast
cancers to the metastatic state is linked to genotypes of catechol-O-
methyltransferase. &DQFHU/HWW, , 23-31.
187. Thompson, P.A., Shields P.G., Freudenheim J.L., Stone A., Vena J.E.,
Marshall J.R., Graham S., Laughlin R., Nemoto T., Kadlubar F.F. and
Ambrosone C.B. (1998) Genetic polymorphisms in catechol-O-
methyltransferase, menopausal status, and breast cancer risk. &DQFHU
5HV, , 2107-10.
188. Spink, D.C., Eugster H.P., Lincoln D.W., 2nd, Schuetz J.D., Schuetz
E.G., Johnson J.A., Kaminsky L.S. and Gierthy J.F. (1992) 17 beta-

80
 estradiol hydroxylation catalyzed by human cytochrome P450 1A1: a
comparison of the activities induced by 2,3,7,8-tetrachlorodibenzo-p-
dioxin in MCF-7 cells with those from heterologous expression of the
cDNA. $UFK%LRFKHP%LRSK\V, , 342-8.
189. http://www.imm.ki.se/CYPalleles/cyp1a1.htm
190. Landi, M.T., Bertazzi P.A., Shields P.G., Clark G., Lucier G.W., Garte
S.J., Cosma G. and Caporaso N.E. (1994) Association between
CYP1A1 genotype, mRNA expression and enzymatic activity in
humans. 3KDUPDFRJHQHWLFV, , 242-6.
191. Moysich, K.B., Shields P.G., Freudenheim J.L., Schisterman E.F.,
Vena J.E., Kostyniak P., Greizerstein H., Marshall J.R., Graham S. and
Ambrosone C.B. (1999) Polychlorinated biphenyls, cytochrome
P4501A1 polymorphism, and postmenopausal breast cancer risk.
&DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 41-4.
192. Ambrosone, C.B., Freudenheim J.L., Graham S., Marshall J.R., Vena
J.E., Brasure J.R., Laughlin R., Nemoto T., Michalek A.M., Harrington
A. and et al. (1995) Cytochrome P4501A1 and glutathione S-
transferase (M1) genetic polymorphisms and postmenopausal breast
cancer risk. &DQFHU5HV, , 3483-5.
193. Ishibe, N., Hankinson S.E., Colditz G.A., Spiegelman D., Willett W.C.,
Speizer F.E., Kelsey K.T. and Hunter D.J. (1998) Cigarette smoking,
cytochrome P450 1A1 polymorphisms, and breast cancer risk in the
Nurses’ Health Study. &DQFHU5HV, , 667-71.
194. Bailey, L.R., Roodi N., Verrier C.S., Yee C.J., Dupont W.D. and Parl
F.F. (1998) Breast cancer and CYPIA1, GSTM1, and GSTT1
polymorphisms: evidence of a lack of association in Caucasians and
African Americans. &DQFHU5HV, , 65-70.
195. Rebbeck, T.R., Rosvold E.A., Duggan D.J., Zhang J. and Buetow K.H.
(1994) Genetics of CYP1A1: coamplification of specific alleles by
polymerase chain reaction and association with breast cancer. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 511-4.
196. Taioli, E., Bradlow H.L., Garbers S.V., Sepkovic D.W., Osborne M.P.,
Trachman J., Ganguly S. and Garte S.J. (1999) Role of estradiol
metabolism and CYP1A1 polymorphisms in breast cancer risk. &DQFHU
'HWHFW3UHY, , 232-7.
197. Huang, C.S., Shen C.Y., Chang K.J., Hsu S.M. and Chern H.D. (1999)
Cytochrome P4501A1 polymorphism as a susceptibility factor for
breast cancer in postmenopausal Chinese women in Taiwan. %U -
&DQFHU, , 1838-43.
198. Guengerich, F.P. and Shimada T. (1998) Activation of procarcinogens
by human cytochrome P450 enzymes. 0XWDW5HV, , 201-13.
199. McKay, J.A., Melvin W.T., Ah-See A.K., Ewen S.W., Greenlee W.F.,
Marcus C.B., Burke M.D. and Murray G.I. (1995) Expression of
cytochrome P450 CYP1B1 in breast cancer. )(%6/HWW, , 270-2.
200. Li, D.N., Seidel A., Pritchard M.P., Wolf C.R. and Friedberg T. (2000)
Polymorphisms in P450 CYP1B1 affect the conversion of estradiol to
the potentially carcinogenic metabolite 4-hydroxyestradiol.
3KDUPDFRJHQHWLFV, , 343-53.

81
201. Shimada, T., Watanabe J., Kawajiri K., Sutter T.R., Guengerich F.P.,
Gillam E.M. and Inoue K. (1999) Catalytic properties of polymorphic
human cytochrome P450 1B1 variants. &DUFLQRJHQHVLV, , 1607-13.
202. Hanna, I.H., Dawling S., Roodi N., Guengerich F.P. and Parl F.F.
(2000) Cytochrome P450 1B1 (CYP1B1) pharmacogenetics:
association of polymorphisms with functional differences in estrogen
hydroxylation activity. &DQFHU5HV, , 3440-4.
203. Bailey, L.R., Roodi N., Dupont W.D. and Parl F.F. (1998) Association of
cytochrome P450 1B1 (CYP1B1) polymorphism with steroid receptor
status in breast cancer. &DQFHU5HV, , 5038-41.
204. Zheng, W., Xie D.W., Jin F., Cheng J.R., Dai Q., Wen W.Q., Shu X.O.
and Gao Y.T. (2000) Genetic polymorphism of cytochrome P450-1B1
and risk of breast cancer. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 147-
50.
205. Watanabe, J., Shimada T., Gillam E.M., Ikuta T., Suemasu K., Higashi
Y., Gotoh O. and Kawajiri K. (2000) Association of CYP1B1 genetic
polymorphism with incidence to breast and lung cancer.
3KDUPDFRJHQHWLFV, , 25-33.
206. Strange, R.C. and Fryer A.A. (1999) Chapter 19. The glutathione S-
transferases: influence of polymorphism on cancer susceptibility. ,$5&
6FL3XEO, , 231-49.
207. Ketterer, B. (1988) Protective role of glutathione and glutathione
transferases in mutagenesis and carcinogenesis. 0XWDW5HV, , 343-
61.
208. Berhane, K., Widersten M., Engstrom A., Kozarich J.W. and Mannervik
B. (1994) Detoxication of base propenals and other alpha, beta-
unsaturated aldehyde products of radical reactions and lipid
peroxidation by human glutathione transferases. 3URF1DWO$FDG6FL8
6$, , 1480-4.
209. Strange, R.C., Jones P.W. and Fryer A.A. (2000) Glutathione S-
transferase: genetics and role in toxicology. 7R[LFRO/HWW, , 357-
63.
210. Johansson, A.-S. and Mannervik B. (2001, in press) Interindividual
variablility of glutathione transferase expression in man. In Pacifici,
G.M. and Pelkonen, O. (eds), ,QWHULQGLYLGXDO YDULDEOLOLW\ LQ GUXJ
PHWDEROLVPLQPDQ. Taylor & Francis, London.
211. Black, S.M., Beggs J.D., Hayes J.D., Bartoszek A., Muramatsu M.,
Sakai M. and Wolf C.R. (1990) Expression of human glutathione S-
transferases in Saccharomyces cerevisiae confers resistance to the
anticancer drugs adriamycin and chlorambucil. %LRFKHP -, , 309-
15.
212. Blackburn, A.C., Tzeng H.F., Anders M.W. and Board P.G. (2000)
Discovery of a functional polymorphism in human glutathione
transferase zeta by expressed sequence tag database analysis.
3KDUPDFRJHQHWLFV, , 49-57.
213. Board, P.G., Baker R.T., Chelvanayagam G. and Jermiin L.S. (1997)
Zeta, a novel class of glutathione transferases in a range of species
from plants to humans. %LRFKHP-, , 929-35.

82
214. Seidegard, J. and Ekstrom G. (1997) The role of human glutathione
transferases and epoxide hydrolases in the metabolism of xenobiotics.
(QYLURQ+HDOWK3HUVSHFW, 6XSSO, 791-9.
215. Pearson, W.R., Vorachek W.R., Xu S.J., Berger R., Hart I., Vannais D.
and Patterson D. (1993) Identification of class-mu glutathione
transferase genes GSTM1-GSTM5 on human chromosome 1p13. $P
-+XP*HQHW, , 220-33.
216. Garte, S., Gaspari L., Alexandrie A.-K., Ambrosone C., Autrup H.,
Autrup J., L., Baranova H. and al. e. (2000) Metabolic gene
polymorphism frequencies in normal control populations. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY.
217. Widersten, M., Pearson W.R., Engstrom A. and Mannervik B. (1991)
Heterologous expression of the allelic variant mu-class glutathione
transferases mu and psi. %LRFKHP-, , 519-24.
218. McLellan, R.A., Oscarson M., Alexandrie A.K., Seidegard J., Evans
D.A., Rannug A. and Ingelman-Sundberg M. (1997) Characterization of
a human glutathione S-transferase mu cluster containing a duplicated
GSTM1 gene that causes ultrarapid enzyme activity. 0RO 3KDUPDFRO,
, 958-65.
219. Charrier, J., Maugard C.M., Le Mevel B. and Bignon Y.J. (1999)
Allelotype influence at glutathione S-transferase M1 locus on breast
cancer susceptibility. %U-&DQFHU, , 346-53.
220. Helzlsouer, K.J., Selmin O., Huang H.Y., Strickland P.T., Hoffman S.,
Alberg A.J., Watson M., Comstock G.W. and Bell D. (1998) Association
between glutathione S-transferase M1, P1, and T1 genetic
polymorphisms and development of breast cancer. -1DWO&DQFHU,QVW,
, 512-8.
221. Ambrosone, C.B., Coles B.F., Freudenheim J.L. and Shields P.G.
(1999) Glutathione-S-transferase (GSTM1) genetic polymorphisms do
not affect human breast cancer risk, regardless of dietary antioxidants.
-1XWU, , 565S-568S.
222. Kelsey, K.T., Hankinson S.E., Colditz G.A., Springer K., Garcia-Closas
M., Spiegelman D., Manson J.E., Garland M., Stampfer M.J., Willett
W.C., Speizer F.E. and Hunter D.J. (1997) Glutathione S-transferase
class mu deletion polymorphism and breast cancer: results from
prevalent versus incident cases. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, ,
511-5.
223. Zhong, S., Wyllie A.H., Barnes D., Wolf C.R. and Spurr N.K. (1993)
Relationship between the GSTM1 genetic polymorphism and
susceptibility to bladder, breast and colon cancer. &DUFLQRJHQHVLV, ,
1821-4.
224. Millikan, R., Pittman G., Tse C.K., Savitz D.A., Newman B. and Bell D.
(2000) Glutathione S-transferases M1, T1, and P1 and breast cancer.
&DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 567-73.
225. Park, S.K., Yoo K.Y., Lee S.J., Kim S.U., Ahn S.H., Noh D.Y., Choe
K.J., Strickland P.T., Hirvonen A. and Kang D. (2000) Alcohol
consumption, glutathione S-transferase M1 and T1 genetic
polymorphisms and breast cancer risk. 3KDUPDFRJHQHWLFV, , 301-9.
226. Lizard-Nacol, S., Coudert B., Colosetti P., Riedinger J.M., Fargeot P.
and Brunet-Lecomte P. (1999) Glutathione S-transferase M1 null

83
 genotype: lack of association with tumour characteristics and survival
in advanced breast cancer. %UHDVW&DQFHU5HV, , 81-87.
227. Nedelcheva Kristensen, V., Andersen T.I., Erikstein B., Geitvik G.,
Skovlund E., Nesland J.M. and Borresen-Dale A.L. (1998) Single tube
multiplex polymerase chain reaction genotype analysis of GSTM1,
GSTT1 and GSTP1: relation of genotypes to TP53 tumor status and
clinicopathological variables in breast cancer patients.
3KDUPDFRJHQHWLFV, , 441-7.
228. Rundle, A., Tang D., Zhou J., Cho S. and Perera F. (2000) The
association between glutathione S-transferase M1 genotype and
polycyclic aromatic hydrocarbon-DNA adducts in breast tissue. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 1079-85.
229. Patskovsky, Y.V., Huang M.Q., Takayama T., Listowsky I. and Pearson
W.R. (1999) Distinctive structure of the human GSTM3 gene-inverted
orientation relative to the mu class glutathione transferase gene
cluster. $UFK%LRFKHP%LRSK\V, , 85-93.
230. Maugard, C.M., Charrier J. and Bignon Y.J. (1998) Allelic deletion at
glutathione S-transferase M1 locus and its association with breast
cancer susceptibility. &KHP%LRO,QWHUDFW, , 365-75.
231. Garcia-Closas, M., Kelsey K.T., Hankinson S.E., Spiegelman D.,
Springer K., Willett W.C., Speizer F.E. and Hunter D.J. (1999)
Glutathione S-transferase mu and theta polymorphisms and breast
cancer susceptibility. -1DWO&DQFHU,QVW, , 1960-4.
232. Curran, J.E., Weinstein S.R. and Griffiths L.R. (2000) Polymorphisms
of glutathione S-transferase genes (GSTM1, GSTP1 and GSTT1) and
breast cancer susceptibility. &DQFHU/HWW, , 113-20.
233. Yengi, L., Inskip A., Gilford J., Alldersea J., Bailey L., Smith A., Lear
J.T., Heagerty A.H., Bowers B., Hand P., Hayes J.D., Jones P.W.,
Strange R.C. and Fryer A.A. (1996) Polymorphism at the glutathione S-
transferase locus GSTM3: interactions with cytochrome P450 and
glutathione S-transferase genotypes as risk factors for multiple
cutaneous basal cell carcinoma. &DQFHU5HV, , 1974-7.
234. Nakajima, T., Elovaara E., Anttila S., Hirvonen A., Camus A.M., Hayes
J.D., Ketterer B. and Vainio H. (1995) Expression and polymorphism of
glutathione S-transferase in human lungs: risk factors in smoking-
related lung cancer. &DUFLQRJHQHVLV, , 707-11.
235. To-Figueras, J., Gene M., Gomez-Catalan J., Piqué E., Borrego N.,
Marfany G., Gonzalez_Duarte R. and Corbella J. (2000) Polymorphism
of glutathione S-transferase M3: interaction with glutathione S-
transferase M1 and lung cancer susceptibility. %LRPDUNHUV, , 73-80.
236. Anttila, S., Luostarinen L., Hirvonen A., Elovaara E., Karjalainen A.,
Nurminen T., Hayes J.D., Vainio H. and Ketterer B. (1995) Pulmonary
expression of glutathione S-transferase M3 in lung cancer patients:
association with GSTM1 polymorphism, smoking, and asbestos
exposure. &DQFHU5HV, , 3305-9.
237. Sundberg, K., Johansson A.S., Stenberg G., Widersten M., Seidel A.,
Mannervik B. and Jernstrom B. (1998) Differences in the catalytic
efficiencies of allelic variants of glutathione transferase P1-1 towards
carcinogenic diol epoxides of polycyclic aromatic hydrocarbons.
&DUFLQRJHQHVLV, , 433-6.

84
238. Harries, L.W., Stubbins M.J., Forman D., Howard G.C. and Wolf C.R.
(1997) Identification of genetic polymorphisms at the glutathione S-
transferase Pi locus and association with susceptibility to bladder,
testicular and prostate cancer. &DUFLQRJHQHVLV, , 641-4.
239. Saarikoski, S.T., Voho A., Reinikainen M., Anttila S., Karjalainen A.,
Malaveille C., Vainio H., Husgafvel-Pursiainen K. and Hirvonen A.
(1998) Combined effect of polymorphic GST genes on individual
susceptibility to lung cancer. ,QW-&DQFHU, , 516-21.
240. Sweeney, C., McClure G.Y., Fares M.Y., Stone A., Coles B.F.,
Thompson P.A., Korourian S., Hutchins L.F., Kadlubar F.F. and
Ambrosone C.B. (2000) Association between survival after treatment
for breast cancer and glutathione S-transferase P1 Ile105Val
polymorphism. &DQFHU5HV, , 5621-4.
241. Maugard, C., M., Charrier J., Pitard A., Campion L., Akande O.,
Pleasants L. and Ali-Osman F. (2001) Genetic polymorphism at the
glutathione S-transferase (GST) P1 locus is a breast cancer modifier.
,QW-&DQFHU, , 334-339.
242. Tan, K.L., Webb G.C., Baker R.T. and Board P.G. (1995) Molecular
cloning of a cDNA and chromosomal localization of a human theta-
class glutathione S-transferase gene (GSTT2) to chromosome 22.
*HQRPLFV, , 381-7.
243. Nelson, H.H., Wiencke J.K., Christiani D.C., Cheng T.J., Zuo Z.F.,
Schwartz B.S., Lee B.K., Spitz M.R., Wang M., Xu X. and et al. (1995)
Ethnic differences in the prevalence of the homozygous deleted
genotype of glutathione S-transferase theta. &DUFLQRJHQHVLV, , 1243-
5.
244. Landi, S. (2000) Mammalian class theta GST and differential
susceptibility to carcinogens: a review. 0XWDW5HV, , 247-283.
245. Borgstahl, G.E., Parge H.E., Hickey M.J., Johnson M.J., Boissinot M.,
Hallewell R.A., Lepock J.R., Cabelli D.E. and Tainer J.A. (1996)
Human mitochondrial manganese superoxide dismutase polymorphic
variant Ile58Thr reduces activity by destabilizing the tetrameric
interface. %LRFKHPLVWU\, , 4287-97.
246. Church, S.L., Grant J.W., Meese E.U. and Trent J.M. (1992)
Sublocalization of the gene encoding manganese superoxide
dismutase (MnSOD/SOD2) to 6q25 by fluorescence in situ
hybridization and somatic cell hybrid mapping. *HQRPLFV, , 823-5.
247. Wispe, J.R., Clark J.C., Burhans M.S., Kropp K.E., Korfhagen T.R. and
Whitsett J.A. (1989) Synthesis and processing of the precursor for
human mangano-superoxide dismutase. %LRFKLP %LRSK\V $FWD, ,
30-6.
248. Hay, R., Bohni P. and Gasser S. (1984) How mitochondria import
proteins. %LRFKLP%LRSK\V$FWD, , 65-87.
249. Rohrdanz, E. and Kahl R. (1998) Alterations of antioxidant enzyme
expression in response to hydrogen peroxide. )UHH5DGLF%LRO0HG, ,
27-38.
250. Oberley, L.W. and Buettner G.R. (1979) Role of superoxide dismutase
in cancer: a review. &DQFHU5HV, , 1141-9.
251. Li, J.J., Oberley L.W., St Clair D.K., Ridnour L.A. and Oberley T.D.
(1995) Phenotypic changes induced in human breast cancer cells by

85
 overexpression of manganese-containing superoxide dismutase.
2QFRJHQH, , 1989-2000.
252. Shimoda-Matsubayashi, S., Matsumine H., Kobayashi T., Nakagawa-
Hattori Y., Shimizu Y. and Mizuno Y. (1996) Structural dimorphism in
the mitochondrial targeting sequence in the human manganese
superoxide dismutase gene. A predictive evidence for conformational
change to influence mitochondrial transport and a study of allelic
association in Parkinson’s disease [published erratum appears in
Biochem Biophys Res Commun 1996 Dec 4;229(1):361]. %LRFKHP
%LRSK\V5HV&RPPXQ, , 561-5.
253. Rosenblum, J.S., Gilula N.B. and Lerner R.A. (1996) On signal
sequence polymorphisms and diseases of distribution. 3URF1DWO$FDG
6FL86$, , 4471-3.
254. Van Landeghem, G.F., Tabatabaie P., Kucinskas V., Saha N. and
Beckman G. (1999) Ethnic variation in the mitochondrial targeting
sequence polymorphism of MnSOD. +XP+HUHG, , 190-3.
255. Zhang, H.J., Yan T., Oberley T.D. and Oberley L.W. (1999)
Comparison of effects of two polymorphic variants of manganese
superoxide dismutase on human breast MCF-7 cancer cell phenotype.
&DQFHU5HV, , 6276-83.
256. Grasbon-Frodl, E.M., Kosel S., Riess O., Muller U., Mehraein P. and
Graeber M.B. (1999) Analysis of mitochondrial targeting sequence and
coding region polymorphisms of the manganese superoxide dismutase
gene in German Parkinson disease patients. %LRFKHP %LRSK\V 5HV
&RPPXQ, , 749-52.
257. Sobin, L.H. and Wittekind C. (1997) %UHDVWWXPRUV (5 edn). Wiley-Liss,
Inc., New York.
258. Li, T., Vallada H., Curtis D., Arranz M., Xu K., Cai G., Deng H., Liu J.,
Murray R., Liu X. and Collier D.A. (1997) Catechol-O-methyltransferase
Val158Met polymorphism: frequency analysis in Han Chinese subjects
and allelic association of the low activity allele with bipolar affective
disorder. 3KDUPDFRJHQHWLFV, , 349-53.
259. Hirvonen, A., Saarikoski S.T., Linnainmaa K., Koskinen K., Husgafvel-
Pursiainen K., Mattson K. and Vainio H. (1996) Glutathione S-
transferase and N-acetyltransferase genotypes and asbestos-
associated pulmonary disorders. -1DWO&DQFHU,QVW, , 1853-6.
260. Chen, H., Sandler D.P., Taylor J.A., Shore D.L., Liu E., Bloomfield C.D.
and Bell D.A. (1996) Increased risk for myelodysplastic syndromes in
individuals with glutathione transferase theta 1 (GSTT1) gene defect.
/DQFHW, , 295-7.
261. Breslow, N.E. and Day N.E. (1980) Statistical methods in cancer
research. Volume I - The analysis of case- control studies. ,$5& 6FL
3XEO, , 5-338.
262. Rebbeck, T.R. (1997) Molecular epidemiology of the human
glutathione S-transferase genotypes GSTM1 and GSTT1 in cancer
susceptibility. &DQFHU(SLGHPLRO%LRPDUNHUV3UHY, , 733-43.
263. Hirvonen, A. (1997) Combinations of susceptible genotypes and
individual responses to toxicants. (QYLURQ+HDOWK3HUVSHFW, 6XSSO
, 755-8.

86
264. Jourenkova-Mironova, N., Voho A., Bouchardy C., Wikman H., Dayer
P., Benhamou S. and Hirvonen A. (1999) Glutathione S-transferase
GSTM3 and GSTP1 genotypes and larynx cancer risk. &DQFHU
(SLGHPLRO%LRPDUNHUV3UHY, , 185-8.
265. Garcia-Closas, M., Hankinson S.E., Ho S., Malins D.C., Polissar N.L.,
Schaefer S.N., Su Y. and Vinson M.A. (2000) Factors critical to the
design and execution of epidemiologic studies and description of an
innovative technology to follow the progression from normal to cancer
tissue. -1DWO&DQFHU,QVW0RQRJU, , 147-56.
266. Hwang, S.J., Beaty T.H., Liang K.Y., Coresh J. and Khoury M.J. (1994)
Minimum sample size estimation to detect gene-environment
interaction in case-control designs. $P-(SLGHPLRO, , 1029-37.
267. Altshuler, D., Kruglyak L. and Lander E. (1998) Genetic polymorphisms
and disease. 1(QJO-0HG, , 1626.
268. Mannisto, S., Pietinen P., Virtanen M., Kataja V. and Uusitupa M.
(1999) Diet and the risk of breast cancer in a case-control study: does
the threat of disease have an influence on recall bias? - &OLQ
(SLGHPLRO, , 429-39.
269. Brennan, P. (1999) Chapter 12. Design and analysis issues in case-
control studies addressing genetic susceptibility. ,$5& 6FL 3XEO, ,
123-32.
270. Fryer, A.A. and Jones P.W. (1999) Chapter 22. Interactions between
detoxifying enzyme polymorphisms and susceptibility to cancer. ,$5&
6FL3XEO, , 303-22.
271. Spitz, M.R., Duphorne C.M., Detry M.A., Pillow P.C., Amos C.I., Lei L.,
de Andrade M., Gu X., Hong W.K. and Wu X. (2000) Dietary intake of
isothiocyanates: evidence of a joint effect with glutathione S-
transferase polymorphisms in lung cancer risk. &DQFHU (SLGHPLRO
%LRPDUNHUV3UHY, , 1017-20.

87