Open Access Highly Accessed Research article

Pregnancy-related factors and the risk of breast carcinoma in situ and invasive breast cancer among postmenopausal women in the California Teachers Study cohort

Huiyan Ma1*, Katherine D Henderson1, Jane Sullivan-Halley1, Lei Duan1, Sarah F Marshall1, Giske Ursin34, Pamela L Horn-Ross5, Joan Largent2, Dennis M Deapen3, James V Lacey1 and Leslie Bernstein13

Author Affiliations

1 Department of Population Sciences, Beckman Research Institute, City of Hope, 1500 East Duarte Road, Duarte, CA 91010, USA

2 Department of Epidemiology, School of Medicine, University of California, Irvine, 224 Irvine Hall, Irvine, CA 92697, USA

3 Department of Preventive Medicine, Keck School of Medicine/Norris Cancer Center, University of Southern California, 1441 Eastlake Avenue, Los Angeles, CA 90033, USA

4 Department of Nutrition, University of Oslo, PO Box 1046, Blindern, 0316 Oslo, Norway

5 Cancer Prevention Institute of California, 2201 Walnut Avenue, Suite 300, Fremont, CA 94538, USA

For all author emails, please log on.

Breast Cancer Research 2010, 12:R35  doi:10.1186/bcr2589

The electronic version of this article is the complete one and can be found online at: http://breast-cancer-research.com/content/12/3/R35


Received:1 April 2010
Revisions received:2 June 2010
Accepted:18 June 2010
Published:18 June 2010

© 2010 Ma et al.; licensee BioMed Central Ltd.

This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Introduction

Although pregnancy-related factors such as nulliparity and late age at first full-term pregnancy are well-established risk factors for invasive breast cancer, the roles of these factors in the natural history of breast cancer development remain unclear.

Methods

Among 52,464 postmenopausal women participating in the California Teachers Study (CTS), 624 were diagnosed with breast carcinoma in situ (CIS) and 2,828 with invasive breast cancer between 1995 and 2007. Multivariable Cox proportional hazards regression methods were used to estimate relative risks associated with parity, age at first full-term pregnancy, breastfeeding, nausea or vomiting during pregnancy, and preeclampsia.

Results

Compared with never-pregnant women, an increasing number of full-term pregnancies was associated with greater risk reduction for both breast CIS and invasive breast cancer (both P trend < 0.01). Women having four or more full-term pregnancies had a 31% lower breast CIS risk (RR = 0.69, 95% CI = 0.51 to 0.93) and 18% lower invasive breast cancer risk (RR = 0.82, 95% CI = 0.72 to 0.94). Parous women whose first full-term pregnancy occurred at age 35 years or later had a 118% greater risk for breast CIS (RR = 2.18, 95% CI = 1.36 to 3.49) and 27% greater risk for invasive breast cancer (RR = 1.27, 95% CI = 0.99 to 1.65) than those whose first full-term pregnancy occurred before age 21 years. Furthermore, parity was negatively associated with the risk of estrogen receptor-positive (ER+) or ER+/progesterone receptor-positive (PR+) while age at first full-term pregnancy was positively associated with the risk of ER+ or ER+/PR+ invasive breast cancer. Neither of these factors was statistically significantly associated with the risk of ER-negative (ER-) or ER-/PR- invasive breast cancer, tests for heterogeneity between subtypes did not reach statistical significance. No clear associations were detected for other pregnancy-related factors.

Conclusions

These results provide some epidemiologic evidence that parity and age at first full-term pregnancy are involved in the development of breast cancer among postmenopausal women. The role of these factors in risk of in situ versus invasive, and hormone receptor-positive versus -negative breast cancer merits further exploration.

Introduction

Although pregnancy-related factors such as nulliparity and late age at first full-term pregnancy are well-established risk factors for invasive breast cancer, it remains unclear whether such factors have similar effects on breast carcinoma in situ (CIS) or whether their effects vary across the subtypes of invasive breast cancer defined by the estrogen receptor (ER) or the joint ER and progesterone receptor (PR) status of the tumor. The clarification of these issues may shed light on a deeper understanding of the roles of pregnancy-related factors in the natural history of breast cancer development.

Breast CIS includes ductal carcinoma in situ (DCIS), lobular carcinoma in situ (LCIS), and other relatively rare forms of breast CIS, with DCIS being the predominant subtype. The term 'in situ' indicates that neoplastic cells are present but have not spread past the boundaries of ducts or lobules where the tumor initially developed. Estimates of the percentage of patients with DCIS that will progress to invasive disease over a period of 10 or more years vary widely, ranging from 14 to 53% [1] and likely depend on a variety of factors, including type and extent of treatment and medical surveillance. Unlike DCIS, LCIS indicates neoplastic changes in the breast lobules, a precursor lesion that carries an elevated risk of invasive lobular carcinoma in either breast; thus, it is not included in clinical definitions of in situ breast cancer [2,3]. In contrast, DCIS and other forms of breast CIS are clinically considered pre-invasive lesions that can be associated with the development of invasive breast cancer at the same location in the breast where the CIS is located. Nevertheless, epidemiologic studies have shown that, compared with the general population, the risk of invasive breast cancer increases over four-fold following a diagnosis of either DCIS or LCIS [4,5].

Given that breast CIS is a potential precursor to invasive disease, one would expect that breast CIS and invasive breast cancer would share risk factors involved in the early stages of breast cancer development while factors affecting disease progression would be associated only with invasive breast cancer risk. Few studies have made direct comparisons of the consistency of risk estimates for pregnancy-related risk factors between breast CIS and invasive breast cancer within a study population, particularly among women whose reproductive years have ended. In studies that provided results for postmenopausal women, parity was associated with a decreased risk for both postmenopausal breast CIS and postmenopausal invasive breast cancer; results for age at first full-term pregnancy were inconsistent [6-8]. None of these studies provided results on the effects of breastfeeding, nausea or vomiting during pregnancy, or preeclampsia on the risk of breast CIS or invasive breast cancer.

Parity, early age at first full-term pregnancy, and breastfeeding influence the risk of breast cancer predominantly through hormonal mechanisms that involve estrogen and progesterone [9,10]. Previous studies have also linked severe nausea/persistent vomiting during pregnancy to elevated serum estradiol levels [11], and have shown that preeclampsia is associated with lower maternal serum levels of estriol [12] and insulin-like growth factors [13]. Therefore, severe nausea/persistent vomiting during pregnancy might be associated with an increased risk while preeclampsia might be associated with a decreased risk of breast cancer. Although a number of epidemiologic studies have examined severe pregnancy nausea/vomiting [14-16] or history of preeclampsia [17-23] in relation to invasive breast cancer risk, the data have been inconsistent for both factors. Further, no data have been published specifically for breast CIS.

Both a systematic review [24] and a meta-analysis [25] show that nulliparity and late age at first full-term pregnancy are associated with increased risk of ER-positive (ER+) or ER+/PR+, but not with ER-negative (ER-) or ER-/PR- invasive breast cancer. In contrast, the protection from breastfeeding does not differ by ER/PR status. However, these results are based largely on results from case-control studies and only a small number of these studies provide data for postmenopausal women separately from premenopausal women. Three prospective cohort studies that have assessed reproductive factors in relation to hormone receptor subtypes of invasive breast cancer provide consistent results for parity, but inconsistent results for late age at first full-term pregnancy. Further, no data are given for the association of breastfeeding with the hormone receptor subtypes of invasive breast cancer [26-28].

Thus, important questions remain regarding the roles of pregnancy-related factors in the development of breast CIS and invasive breast cancer among women who have completed their reproductive years. The authors examined the associations of parity, age at first full-term pregnancy, duration of breastfeeding, nausea or vomiting during pregnancy, and preeclampsia with breast CIS and invasive breast cancer (overall and by hormonal receptor subtype) among postmenopausal women participating in the prospective California Teachers Study (CTS) cohort.

Materials and methods

Study population and data collection

Details of the CTS have been described previously [29]. Briefly, the CTS is a prospective study of 133,479 current and retired female California public school teachers and administrators who were recruited from active members of the California State Teachers Retirement System. New diagnoses of first primary breast CIS or invasive breast cancer among cohort members were identified through annual linkages from 1995 through 2007 with the California Cancer Registry (CCR), a legally mandated statewide population-based cancer reporting system. Modeled after the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program, the CCR is composed of three SEER Program registries and maintains the highest standards for data quality and completeness. CCR ascertainment of newly diagnosed cancers is estimated to be 99% complete [30].

All information on pregnancy-related factors except nausea or vomiting during pregnancy and preeclampsia was collected in the questionnaire completed by participants when they joined the cohort. Participants reported information on race/ethnicity, family history of breast cancer, age at menarche, detailed pregnancy (including ages at and outcomes for each pregnancy) and breastfeeding histories, menopausal status, use of menopausal hormone therapy, height, and weight. Participants reported information regarding nausea or vomiting during pregnancy and preeclampsia in the second questionnaire completed in 1997-1998. Preeclampsia was defined as a condition that can occur during the second half of a pregnancy and is marked by elevated blood pressure, protein in the urine and fluid retention.

For the current analysis, using information provided when the cohort was formed, 8,867 women who were not California residents were excluded, followed by 6,350 women who had a prior history of breast CIS or invasive breast cancer or an unknown cancer history and by 5,107 women who were 80 years or older. This yielded a preliminary analytic cohort of 113,155 women. In order to ensure that each woman's complete pregnancy history was covered, eligibility for this analysis was restricted to women who were postmenopausal at cohort entry (n = 53,440). Menopausal status was determined after reviewing a woman's age at last menstrual period, reason for cessation of menstrual periods, hysterectomy status, oophorectomy status and current use of hormonal therapy. Women were considered to be postmenopausal if they reported a natural menopause more than six months before completing the baseline questionnaire (n = 31,207), if they had a bilateral oophorectomy (n = 10,868), if they were 56 years or older and not menstruating regularly (n = 6,675), or if their periods had stopped due to other reasons including pituitary adenoma, medication, chemotherapy, radiation treatment, or another reason (n = 4,690). The age criterion was based on previous work indicating that among those with natural menopause 97% were postmenopausal by age 56 years [31]. Further exclusions were women who were missing information on age at menarche (n = 217), pregnancy history (n = 75), or both factors (n = 684). Therefore, 52,464 postmenopausal women remained in the analytic cohort.

Person-time of follow-up for each woman began with the date she completed the baseline questionnaire and ended with the first of the following events: a breast cancer diagnosis (n = 624 breast CIS including 51 women diagnosed with LCIS, ICD-O morphology code = 8520; n = 2,828 invasive), a permanent move outside of California (n = 4,481), death (n = 5,852), or December 31, 2007 (n = 38,679).

The analytic cohort included 8,937 never-pregnant and 43,527 ever-pregnant women. Of the 43,527 ever-pregnant women, 371 experienced pregnancy but did not provide outcome, and 1,957 had only incomplete pregnancies. Women who experienced pregnancy and did not provide an outcome and those who had only incomplete pregnancies were included when testing the effect of any pregnancy (ever vs. never), but were excluded from analyses of parous women. The effect of incomplete pregnancy on breast cancer risk has previously been studied in this cohort [32].

The 41,199 women who had at least one stillbirth or live childbirth were referred to as parous throughout this article. Still and live births were combined to obtain the total number of full-term pregnancies. Parous women who were missing information on breastfeeding or age at first full-term pregnancy (n = 117) were additionally excluded from analyses restricted to parous women. Analyses of parous women were based on 41,082 women of whom 479 were diagnosed with breast CIS and 2,193 were diagnosed with invasive breast cancer during follow-up.

Of 43,527 ever-pregnant postmenopausal women at cohort entry, 32,084 (73.7%; including 394 with breast CIS and 1,794 with invasive breast cancer) provided information on nausea or vomiting during any pregnancy and 31,459 (72.3%; including 391 with breast CIS and 1,768 with invasive breast cancer) provided information on preeclampsia in the CTS' second questionnaire, which was mailed in 1997, and were included in the analysis of these factors.

Since 1990, National Cancer Institute designated SEER Program registries, including those comprising the CCR, routinely abstract ER and PR status from the medical records following a breast cancer diagnosis. Among 2,828 invasive breast cancers diagnosed in the current analytic cohort, ER status was available for 2,443 (86.4%; 2,104 ER+, 339 ER-). Both ER status and PR status were available for 2,326 (82.2%; 1,651 ER+/PR+, 317 ER-/PR-, 340 ER+/PR-, 18 ER-/PR+).

The use of CTS participants' data for this analysis was approved by the Institutional Review Boards at the City of Hope, the University of Southern California, the Cancer Prevention Institute of California (formerly the Northern California Cancer Center), and the University of California at Irvine, and by the Committee for the Protection of Human Subjects, California Health and Human Services Agency.

Data analyses

The relative hazard (represented as the relative risk (RR) and its 95% confidence interval (CI)) for the association between pregnancy-related factors and breast cancer risk was estimated using multivariable Cox proportional hazards regression methods [33]. Separate analyses were conducted for breast CIS and invasive breast cancer (the latter, overall and by hormone receptor subtypes: ER+ and ER-, ER+/PR+ and ER-/PR-).

For all analyses of breast CIS, women diagnosed with invasive breast cancer during follow-up were excluded as the diagnosis of invasive disease presumes they have passed through the in situ disease stage undetected. For analyses of invasive breast cancer, women who developed breast CIS were censored on the date of this diagnosis. For analyses of each receptor subtype of invasive breast cancer, women who developed all other receptor subtypes or had an unknown subtype were censored on the date of this diagnosis.

In the Cox regression models, the time scale was defined by age (in days) at cohort entry and age (in days) at event or censoring (exit). All multivariable models were adjusted for the following factors, selected a priori, as potential confounders: race (white, African-American, others), family history of breast cancer in a first degree relative, that is, mother, father, sister or brother (yes, no, unknown/adopted), age at menarche (< 13, ≥ 13 years), menopausal hormone therapy (HT) use (never use, ever use: only estrogen therapy (ET), only estrogen in combination with progesterone therapy (E+P), both types of HT, unknown type of HT) and body mass index (< 25, 25 to 29.9, ≥ 30 kg/m2, unknown).

Multivariable models for parous women were additionally mutually adjusted for number of full-term pregnancies (1, 2, 3, ≥ 4), age at first full-term pregnancy (< 21, 22 to 24, 25 to 29, 30 to34, ≥ 35 years), and duration of breastfeeding (never, < 6, 6 to 11, 12 to 23, ≥ 24 months).

The associations of breast CIS or invasive breast cancer with nausea or vomiting during pregnancy or preeclampsia were explored among postmenopausal women with at least one pregnancy as reported at cohort entry who had complete information on one or both of these variables in the second mailed questionnaire. Nausea or vomiting during any pregnancy was examined as was the number of pregnancies during which the ever-pregnant woman experienced these symptoms (1, 2, 3, ≥ 4) and whether treatment was received for these symptoms during the most recent pregnancy or during other pregnancies. Preeclampsia was evaluated as never occurring during a pregnancy, occurring during the most recent pregnancy, or not occurring during the most recent pregnancy but having occurred during at least one of other pregnancies. Multivariable models for nausea or vomiting during pregnancy or preeclampsia were additionally adjusted for pregnancy history (number of full-term pregnancies: 1, 2, 3, ≥ 4; ever pregnant but outcomes unknown; ever pregnant but no full-term pregnancy).

Tests for trend were conducted to examine the dose-response relationship of breast CIS or invasive breast cancer with number full-term pregnancies, age at first full-term pregnancy, duration of breastfeeding, and number of pregnancies during which participant experienced nausea or vomiting by fitting ordinal values corresponding to each exposure category in the statistical models and determining whether the slope parameter differed from zero using the Wald test [34].

Tests for homogeneity (evaluating the null hypothesis of homogeneity) of RR estimates were performed (for dichotomous variables or ordinal variables) by constructing a Z test of the differences in log RR divided by the square root of the sum of the variances of the two log RR estimates for the following comparisons: breast CIS vs. invasive breast cancer, ER+ vs. ER-, and ER+/PR+ vs. ER-/PR- breast cancer.

The analyses for breast CIS were repeated among 44,671 (90%) postmenopausal women who reported having their most recent mammogram less than three years before the CTS baseline survey. The results remained the same. The analysis for breast CIS was also repeated after excluding 51 women diagnosed with LCIS. Again, the results remained essentially identical. The results for breast CIS including LCIS from all postmenopausal women without invasive breast cancer (n = 49,636) are presented.

Two-sided P-values are reported for tests for trend and for homogeneity of RR estimates. All statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, North Carolina, USA).

Results

Characteristics

Women in the analytic cohort were followed for an average of 10.5 years. Table 1 provides the age-adjusted distribution of several participant characteristics according to pregnancy history (never-pregnant, ever-pregnant), age at first full-term pregnancy, and number of full-term pregnancies. Compared with never-pregnant women, those who had ever been pregnant were more likely to be African-American women. Parous African-American women were less likely than other women to have had three or more full-term pregnancies with the first full-term pregnancy at age 25 years or older. Parous women who had more full-term pregnancies (≥ 3) were more likely to have a longer duration of breastfeeding than those who had fewer full-term pregnancies (< 3).

Table 1. Age-adjusteda percent distribution of baseline characteristics

Parity, age at first full-term pregnancy and duration of breastfeeding

Compared with never-pregnant women, women who experienced any pregnancy had reduced risks of breast CIS (RR = 0.85, 95% CI = 0.69 to 1.04) and invasive breast cancer (RR = 0.91, 95% CI = 0.83 to 1.00) (Table 2). Compared with never-pregnant women, the magnitude of risk reductions increased with increasing number of full-term pregnancies for both breast CIS (P trend = 0.008) and invasive breast cancer (P trend = 0.003).

Table 2. Adjusted RRs for the association between pregnancy-related factors and breast CIS and invasive breast cancer

In analyses restricted to parous women, increasing age at first full-term pregnancy was associated with an increased risk for both breast CIS and invasive breast cancer (breast CIS, P trend = 0.09; invasive breast cancer, P trend = 0.002) (Table 2). Women who had a first full-term pregnancy at age 35 years or older had more than a two-fold greater risk of breast CIS (RR = 2.18, 95% CI = 1.36 to 3.49), and 27% (RR = 1.27, CI = 0.99 to 1.65) greater risk of invasive breast cancer compared with women whose first full-term pregnancy was before age 21 years. In a model adjusted for age at first full-term pregnancy and number of full-term pregnancies, duration of breastfeeding was not statistically significantly associated with the risk of either breast CIS or invasive breast cancer among parous women.

Further, the risk estimates for parity, age at first full-term pregnancy, and breastfeeding did not differ statistically between breast CIS and invasive breast cancer (all P for homogeneity > 0.25). In addition, these risk estimates for both breast CIS and invasive breast cancer did not vary within subgroups defined by body mass index (< 25, ≥ 25 kg/m2), HT use (never, ever use), and race (white, others).

In the analyses for invasive breast cancer by ER status and joint ER and PR status, ever being pregnant, having multiple full-term pregnancies, and having an early age at first full-term pregnancy were each statistically significantly associated with reduced risk of ER + and of ER +/PR + invasive breast cancer, and were not statistically significantly associated with ER- or ER-/PR- invasive breast cancer, but none of the tests for homogeneity of risk estimates comparing results for ER+ to those for ER- or comparing those for ER +/PR + to those for ER -/PR - reached statistical significance (Table 3). Duration of breastfeeding was not statistically significantly associated with anyone of ER +, ER -, ER+/PR +, and ER -/PR - invasive breast cancer.

Table 3. Adjusted RRs for the association between pregnancy-related factors and the subtypes of invasive breast cancer

Nausea or vomiting during pregnancy

Among women who experienced any pregnancy, nausea or vomiting during pregnancy was not associated with risk of breast CIS or invasive breast cancer regardless of the number of pregnancies in which it occurred (one or more) or if the woman received treatment for nausea or vomiting of pregnancy (a measure of severity of the condition) or the timing of that treatment (for the most recent or only for earlier pregnancies) (all 95% CIs include 1, all P trend > 0.05; Table 4).

Table 4. Adjusted RRs for the association between pregnancy-related conditions and breast CIS and invasive breast cancer

Preeclampsia

Among women who experienced any pregnancy, preeclampsia diagnosed during the most recent pregnancy was associated with an increased risk of breast CIS (RR = 1.72, 95% CI = 1.05 to 2.81), but was not associated with invasive breast cancer (Homogeneity test P = 0.06, Table 4). Preeclampsia diagnosed during any pregnancies prior to the most recent was not statistically significantly associated with either breast CIS risk (RR = 0.93, 95% CI = 0.53 to 1.62) or invasive breast cancer (RR = 0.77, 95% RR = 0.58 to 1.02).

Discussion

In the CTS, nulliparity and late age at first full-term pregnancy were associated with higher risk for both breast CIS and invasive breast cancer, although not all of the associations with breast CIS reached statistical significance, in part due to the lower incidence of breast CIS in the cohort. Three previous studies have reported data on both breast CIS and invasive breast cancer for women beyond reproductive age [6-8]. In a cohort study of 32,607 postmenopausal women followed for approximately five years, nulliparous women had a 70% higher risk of breast CIS and a 40% higher risk of invasive breast cancer than women who had a full-term pregnancy before age 21 years; the same study showed that later age at first full-term pregnancy was associated with an increased risk of invasive breast cancer, but was not statistically significantly associated with breast CIS risk [8]. The Iowa Women's Health Study whose 37,105 postmenopausal participants were followed for 11 years found that women with late age at first birth (≥ 30 years) had a 92% and a 29% greater risk for DCIS and invasive breast cancer, respectively, than women whose first birth occurred at or before age 20 years [6]. One previous population-based case-control study also reported that parous postmenopausal women had decreased risk of both breast CIS and invasive breast cancer relative to nulliparous women while late age at first full-term pregnancy was not associated with either breast CIS or invasive breast cancer [7]. Although the associations of parity and age at first full-term pregnancy with risk of breast CIS and invasive breast cancer have not been entirely consistent across epidemiologic studies, as a whole, the data provide some evidence that these two factors are involved in the early stages of breast cancer development.

The mechanisms underlying the protective effect of early parity on breast cancer are not yet clear. Mouse models have demonstrated that breast carcinogenesis is significantly inhibited when rats have completed one pregnancy prior to exposure to the carcinogen compared with age-matched virgin rats [35,36]. This protective effect may be attributed to permanent structural and functional changes induced in the mammary parenchyma by the reproductive process, including exposure to pregnancy hormones, resulting in a lower susceptibility of epithelial cells to future carcinogenic stimuli [37,38]. Moreover, rat gene expression data suggest that persistent pregnancy-induced changes in mammary gene expression may account for the protection of parity [39]. Human data suggest that hormonal mechanisms are involved in pregnancy-related protection against breast cancer including lowered circulating estrogen and progesterone [40-42], higher levels of sex hormone-binding globulin [40] and possibly, human chorionic gonadotropin [43], which has also been demonstrated in a rat carcinogenesis model [44].

In the CTS analyses for invasive breast cancer by hormone receptor subtype, nulliparity and late age at first full-term pregnancy were associated with an increased risk of ER + or ER +/PR + invasive breast cancer but not with ER - or ER -/PR - invasive cancer. These are essentially consistent with the conclusions of a systematic review [24] and a meta-analysis [25], both consisting largely of case-control studies. Compared with data from previous cohort studies, the CTS results for both parity and age at first full-term pregnancy are basically consistent with those reported for the Iowa Women's Health Study [28] and Women's Health Initiative Cohort [26], but the CTS results for age at first full-term pregnancy are inconsistent with those reported for the Nurses' Health Study, in which late age at first full-term pregnancy was associated with an increased risk of ER -/PR - but not ER +/PR +, tumors [27]. Therefore, similar to the majority of previous data, these results support the hypothesis that some pregnancy-related risk factors may differentially influence risk for breast cancer subtypes classified by ER and PR status.

Breastfeeding has been proposed to protect against breast cancer through hormonal mechanisms that include postponing the resumption of ovulatory menstrual cycles after a pregnancy [45], reducing estrogen levels in the breast [46], and having fully differentiated breast tissue which is less susceptible to the hormone milieu [47]. In addition, it has been proposed that breastfeeding also has a direct mechanical effect by which carcinogenic agents are excreted from the breast ductal tissue [48]. However, among parous postmenopausal CTS participants, breastfeeding was not associated with either breast CIS or invasive breast cancer (overall or by receptor subtypes). The observed reductions in breast cancer risk in many previous studies have been stronger for or restricted to younger or premenopausal women [10,15,49-52]. However, in some studies, this reduction in risk was observed in the postmenopausal years [53-55] or was negligible in both age groups [56,57]. Two studies found that the protective effect of breastfeeding decreased with the increasing time after a pregnancy [15,58]. Among the parous women ages 35 to 49 year who participated in Women's Contraceptive and Reproductive Experiences study, the odds ratio (95% CI) of invasive breast cancer associating with at least 24-month breastfeeding was 0.38 (0.19 to 0.77) among women who had given birth within five years and 0.69 (0.51 to 0.93) among those beyond five years [58]. Among postmenopausal CTS participants, the average interval between the last pregnancy and CTS baseline survey was 32 years, and only 1% of the intervals were within 15 years. Fine stratification of the interval between the last pregnancy and CTS baseline survey still did not show any association between breastfeeding and breast cancer risk (results not shown). Therefore, the lack of an association between breastfeeding and breast cancer in the CTS analyses may be due to the restriction to postmenopausal women.

Since hyperemesis gravidarum, that is, severe nausea combined with persistent vomiting during pregnancy, has been linked to an elevated serum estradiol level [11], one might expect a positive association between these events and the risk of breast CIS or invasive breast cancer. Being treated served as a proxy for severe nausea or vomiting of pregnancy, with not treated serving as a proxy for mild/moderate symptoms. However, no association was observed by severity of nausea or vomiting. Consistent with these results, a population-based case-control study has also reported no association of breast cancer risk overall or by age (< 50, ≥ 50) with hyperemesis gravidarum [14]. In contrast with the results presented here, one case-control study reported that having ever been treated for nausea or vomiting during pregnancy was associated with an increased risk of breast cancer [15] while another case-control study reported that nausea or vomiting during first pregnancy was associated with a slightly lower risk of breast cancer [16]. Both of these studies were conducted among young women (< 45 years of age). It is possible that the null results from the CTS study are due to an older population of women, or that the impact of this condition or the hormonal changes related to it, dissipates with time since the pregnancy.

Preeclampsia has been associated with lower maternal serum levels of estriol [12] and insulin-like growth factors [13]. Therefore, preeclampsia might be associated with a lower breast cancer risk. However, among postmenopausal women in the CTS, preeclampsia diagnosed during a woman's most recent pregnancy was associated with an increased risk of breast CIS, while preeclampsia diagnosed during any pregnancies prior to the most recent was not associated with breast CIS risk. Preeclampsia was not associated with invasive breast cancer. Although a number of epidemiologic studies have examined history of preeclampsia in relation to invasive breast cancer risk, none have reported data specifically for breast CIS. Among cohort studies reporting results for preeclampsia and invasive breast cancer, one found a 38% (95% CI = 1.00 to 1.89) greater risk of invasive breast cancer among women who ever experience preeclampsia than among women with no such history [17], while another reported that women with preeclampsia and/or hypertension diagnosed during their first pregnancies had 19% (95% CI = 0.71 to 0.91) lower risk for breast cancer than women who had not experienced preeclampsia [18]. A third cohort study reported no association [19]. In contrast, several case-control studies have reported that preeclampsia was associated with a decreased risk of breast cancer [20-23]. Based on the CTS results and those of earlier studies, no conclusion can be drawn as to whether preeclampsia is a risk factor for breast cancer.

Strengths of the current study include its size, the large number of women who have been diagnosed with an incident breast cancer, the ability to identify and confirm cancer diagnoses through the CCR, California's high-quality statewide cancer registry, and collection of pregnancy-related information prior to the diagnosis of breast CIS or invasive breast cancer.

Several limitations of the current study must be considered. ER and PR status results were collected by the regional registries in California from pathology laboratories located throughout the state; these laboratories may vary in their application of immunohistochemical methods and the cutpoints used to assign a positive status. However, it is unlikely that any methodological differences would influence the observed associations, as a large validation study comparing registry reports of receptor status to those of a single expert laboratory found only small differences in risk estimates for ER+/PR+ and ER-/PR- breast cancer between the two sources [59]. Although not all women with invasive breast cancer had ER or PR status available, the numbers missing this information were relatively small (14% of cases missing ER status; 18% missing joint ER and PR status), and the percentage of women with missing information gradually decreased during follow-up, both of which are similar to the rates presented in previous studies conducted within SEER registries [60,61]. Women with invasive breast cancer who had ER status available in the CCR had pregnancy-related factors (ever pregnant, number of full-term pregnancies, age at first full-term pregnancy, and duration of breastfeeding) that were similar to those of women who were missing ER status information (all Pearson's chi-square P > 0.07). These pregnancy-related factors were also similar between women having invasive breast cancer who had information on both receptor status markers and those without this information, except that those with the information for both ER and PR status were slightly more likely to have never been pregnant (18.1% vs. 14.1%, Pearson's chi-square P = 0.03). This difference could cause an overestimate of any protective effect of pregnancies on all subtypes, but it is unlikely that this bias would be restricted to ER+/PR+ invasive breast cancer. In addition, the relative risk estimates for ER+/PR+ subtype differed minimally from those for ER+ subtype and the estimates for ER-/PR- subtype were similar to those for ER- subtype. Therefore, the missing information on ER and/or PR status is unlikely to have introduced measurable bias in this study. Since 75% of women with breast CIS had no data for ER/PR status, we were unable to examine the association between pregnancy-related factors and the risk of breast CIS by ER/PR status.

Another limitation is that the analyses for nausea or vomiting during pregnancy and preeclampsia were based on approximately 73% of ever-pregnant postmenopausal women. These women were similar to those without information on these two factors in terms of race, family history of breast cancer in a first degree relative, age at menarche, HT use, body mass index, number of full-term pregnancy, age at first full-term pregnancy, and duration of breastfeeding. Therefore, selection bias is unlikely to be an important explanation for the results for these two factors.

Conclusions

These results provide some epidemiologic evidence that parity and age at first full-term pregnancy are involved in the development of breast cancer among postmenopausal women. This study did not provide evidence of any association of breastfeeding with postmenopausal diagnosis of breast CIS or invasive breast cancer overall or by hormone receptor subtypes. Further, nausea or vomiting during pregnancy was not associated with stage of breast cancer and no evidence was observed that preeclampsia is clearly associated with lower breast cancer risk.

Abbreviations

BMI: body mass index; CCR: California Cancer Registry; CI: confidence interval; CIS: carcinoma in situ; CTS: California Teachers Study; DCIS: ductal carcinoma in situ; E: estrogen; ER: estrogen receptor; HT: hormonal therapy; LCIS: lobular carcinoma in situ; P: progestin; PR: progesterone receptor; RR: relative risk; SEER: Surveillance, Epidemiology, and End Results.

Competing interests

The authors declare that they have no competing interests.

Authors' contributions

LB and PLHR participated in the study design and supervised the data collection. Data management was conducted by HM, JSH, and SM. HM and SM conducted data analyses under the supervision of LB. HM drafted the manuscript. All authors participated in the revision of the manuscript and have read and approved the final version.

Acknowledgements

This work was supported by the California Breast Cancer Act of 1993; National Institute of Health (grant numbers R01 CA77398 and K05 CA136967 (LB)); and the California Breast Cancer Research Fund (contract 97-10500). The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute's Surveillance, Epidemiology and End Results Program under contract N01-PC-35136 awarded to the Cancer Prevention Institute of California (formerly the Northern California Cancer Center), contract N01-PC-35139 awarded to the University of Southern California, and contract N02-PC-15105 awarded to the Public Health Institute; and the Centers for Disease Control and Prevention's National Program of Cancer Registries, under agreement #U55/CCR921930-02 awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the authors and endorsement by the State of California, Department of Public Health, the National Cancer Institute, the Centers for Disease Control and Prevention, or their contractors and subcontractors is not intended nor should be inferred.

References

  1. Erbas B, Provenzano E, Armes J, Gertig D: The natural history of ductal carcinoma in situ of the breast: a review.

    Breast Cancer Res Treat 2006, 97:135-144. PubMed Abstract | Publisher Full Text OpenURL

  2. Hanby AM, Hughes TA: In situ and invasive lobular neoplasia of the breast.

    Histopathology 2008, 52:58-66. PubMed Abstract | Publisher Full Text OpenURL

  3. Tavassoli FA: Lobular and ductal intraepithelial neoplasia.

    Pathologe 2008, 29:107-111. PubMed Abstract | Publisher Full Text OpenURL

  4. Levi F, Randimbison L, Te VC, La Vecchia C: Invasive breast cancer following ductal and lobular carcinoma in situ of the breast.

    Int J Cancer 2005, 116:820-823. PubMed Abstract | Publisher Full Text OpenURL

  5. Warnberg F, Yuen J, Holmberg L: Risk of subsequent invasive breast cancer after breast carcinoma in situ.

    Lancet 2000, 355:724-725. PubMed Abstract | Publisher Full Text OpenURL

  6. Gapstur SM, Morrow M, Sellers TA: Hormone replacement therapy and risk of breast cancer with a favorable histology: results of the Iowa Women's Health Study.

    JAMA 1999, 281:2091-2097. PubMed Abstract | Publisher Full Text OpenURL

  7. Longnecker MP, Bernstein L, Paganini-Hill A, Enger SM, Ross RK: Risk factors for in situ breast cancer.

    Cancer Epidemiol Biomarkers Prev 1996, 5:961-965. PubMed Abstract | Publisher Full Text OpenURL

  8. Reinier KS, Vacek PM, Geller BM: Risk factors for breast carcinoma in situ versus invasive breast cancer in a prospective study of pre- and post-menopausal women.

    Breast Cancer Res Treat 2007, 103:343-348. PubMed Abstract | Publisher Full Text OpenURL

  9. Henderson BE, Ross RK, Pike MC, Casagrande JT: Endogenous hormones as a major factor in human cancer.

    Cancer Res 1982, 42:3232-3239. PubMed Abstract | Publisher Full Text OpenURL

  10. Kelsey JL, Gammon MD, John EM: Reproductive factors and breast cancer.

    Epidemiol Rev 1993, 15:36-47. PubMed Abstract | Publisher Full Text OpenURL

  11. Depue RH, Bernstein L, Ross RK, Judd HL, Henderson BE: Hyperemesis gravidarum in relation to estradiol levels pregnancy outcome and other maternal factors: a seroepidemiologic study.

    Am J Obstet Gynecol 1987, 156:1137-1141. PubMed Abstract OpenURL

  12. Isouard G: Measurement of serum levels of oestriol and human placental lactogen in the management of pre-eclamptic pregnancies.

    Med J Aust 1979, 2:401-404. PubMed Abstract OpenURL

  13. Altinkaynak K, Aksoy HH, Bakan E, Kumtepe Y: Serum IGF-I and IGFBP-3 in healthy pregnancies and patients with preeclampsia.

    Clin Biochem 2003, 36:221-223. PubMed Abstract | Publisher Full Text OpenURL

  14. Erlandsson G, Lambe M, Cnattingius S, Ekbom A: Hyperemesis gravidarum and subsequent breast cancer risk.

    Br J Cancer 2002, 87:974-976. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  15. Enger SM, Ross RK, Henderson B, Bernstein L: Breastfeeding history pregnancy experience and risk of breast cancer.

    Br J Cancer 1997, 76:118-123. PubMed Abstract | PubMed Central Full Text OpenURL

  16. Troisi R, Weiss HA, Hoover RN, Potischman N, Swanson CA, Brogan DR, Coates RJ, Gammon MD, Malone KE, Daling JR, Brinton LA: Pregnancy characteristics and maternal risk of breast cancer.

    Epidemiology 1998, 9:641-647. PubMed Abstract | Publisher Full Text OpenURL

  17. Paltiel O, Friedlander Y, Tiram E, Barchana M, Xue X, Harlap S: Cancer after pre-eclampsia: follow up of the Jerusalem perinatal study cohort.

    BMJ 2004, 328:919. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  18. Vatten LJ, Romundstad PR, Trichopoulos D, Skjaerven R: Pre-eclampsia in pregnancy and subsequent risk for breast cancer.

    Br J Cancer 2002, 87:971-973. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  19. Mogren I, Stenlund H, Hogberg U: Long-term impact of reproductive factors on the risk of cervical endometrial, ovarian and breast cancer.

    Acta Oncol 2001, 40:849-854. PubMed Abstract | Publisher Full Text OpenURL

  20. Innes KE, Byers TE: First pregnancy characteristics and subsequent breast cancer risk among young women.

    Int J Cancer 2004, 112:306-311. PubMed Abstract | Publisher Full Text OpenURL

  21. Polednak AP, Janerich DT: Characteristics of first pregnancy in relation to early breast cancer. A case-control study.

    J Reprod Med 1983, 28:314-318. PubMed Abstract OpenURL

  22. Terry MB, Perrin M, Salafia CM, Zhang FF, Neugut AI, Teitelbaum SL, Britton J, Gammon MD: Preeclampsia, pregnancy-related hypertension and breast cancer risk.

    Am J Epidemiol 2007, 165:1007-1014. PubMed Abstract | Publisher Full Text OpenURL

  23. Thompson WD, Jacobson HI, Negrini B, Janerich DT: Hypertension, pregnancy, and risk of breast cancer.

    J Natl Cancer Inst 1989, 81:1571-1574. PubMed Abstract | Publisher Full Text OpenURL

  24. Althuis MD, Fergenbaum JH, Garcia-Closas M, Brinton LA, Madigan MP, Sherman ME: Etiology of hormone receptor-defined breast cancer: a systematic review of the literature.

    Cancer Epidemiol Biomarkers Prev 2004, 13:1558-1568. PubMed Abstract | Publisher Full Text OpenURL

  25. Ma H, Bernstein L, Pike MC, Ursin G: Reproductive factors and breast cancer risk according to joint estrogen and progesterone receptor status: a meta-analysis of epidemiological studies.

    Breast Cancer Res 2006, 8:R43. PubMed Abstract | BioMed Central Full Text | PubMed Central Full Text OpenURL

  26. Chlebowski RT, Anderson GL, Lane DS, Aragaki AK, Rohan T, Yasmeen S, Sarto G, Rosenberg CA, Hubbell FA, the Women's Health Initiative I: Predicting Risk of Breast Cancer in Postmenopausal Women by Hormone Receptor Status.

    J Natl Cancer Inst 2007, 99:1695-1705. PubMed Abstract | Publisher Full Text OpenURL

  27. Colditz GA, Rosner BA, Chen WY, Holmes MD, Hankinson SE: Risk factors for breast cancer according to estrogen and progesterone receptor status.

    J Natl Cancer Inst 2004, 96:218-228. PubMed Abstract | Publisher Full Text OpenURL

  28. Potter JD, Cerhan JR, Sellers TA, McGovern PG, Drinkard C, Kushi LR, Folsom AR: Progesterone and estrogen receptors and mammary neoplasia in the Iowa Women's Health Study: how many kinds of breast cancer are there?

    Cancer Epidemiol Biomarkers Prev 1995, 4:319-326. PubMed Abstract | Publisher Full Text OpenURL

  29. Bernstein L, Allen M, Anton-Culver H, Deapen D, Horn-Ross PL, Peel D, Pinder R, Reynolds P, Sullivan-Halley J, West D, Wright W, Ziogas A, Ross RK: High breast cancer incidence rates among California teachers: results from the California Teachers Study (United States).

    Cancer Causes Control 2002, 13:625-635. PubMed Abstract | Publisher Full Text OpenURL

  30. Kwong SL, Perkins CI, Morris CR, Cohen R, Allen M, Wright WE, (Eds): Cancer in California: 1988-1999. Department of Health Services, Cancer Surveillance Section. Sacramento CA; 2001.

  31. Weiss LK, Burkman RT, Cushing-Haugen KL, Voigt LF, Simon MS, Daling JR, Norman SA, Bernstein L, Ursin G, Marchbanks PA, Strom BL, Berlin JA, Weber AL, Doody DR, Wingo PA, McDonald JA, Malone KE, Folger SG, Spirtas R: Hormone replacement therapy regimens and breast cancer risk.

    Obstet Gynecol 2002, 100:1148-1158. PubMed Abstract | Publisher Full Text OpenURL

  32. Henderson KD, Sullivan-Halley J, Reynolds P, Horn-Ross PL, Clarke CA, Chang ET, Neuhausen S, Ursin G, Bernstein L: Incomplete pregnancy is not associated with breast cancer risk: the California Teachers Study.

    Contraception 2008, 77:391-396. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL

  33. Cox DR, Oakes D: Analysis of survival data. London, England: Chapman & Hall; 1984.

  34. Rothman KJ, Greenland S: Modern epidemiology. Philadelphia: Lippincott-Raven; 1998.

  35. Russo J, Russo IH: Susceptibility of the mammary gland to carcinogenesis. II. Pregnancy interruption as a risk factor in tumor incidence.

    Am J Pathol 1980, 100:497-512. PubMed Abstract | PubMed Central Full Text OpenURL

  36. Thordarson G, Jin E, Guzman RC, Swanson SM, Nandi S, Talamantes F: Refractoriness to mammary tumorigenesis in parous rats: is it caused by persistent changes in the hormonal environment or permanent biochemical alterations in the mammary epithelia?

    Carcinogenesis 1995, 16:2847-2853. PubMed Abstract | Publisher Full Text OpenURL

  37. Russo IH, Koszalka M, Russo J: Comparative study of the influence of pregnancy and hormonal treatment on mammary carcinogenesis.

    Br J Cancer 1991, 64:481-484. PubMed Abstract | PubMed Central Full Text OpenURL

  38. Russo J, Russo IH: Biological and molecular bases of mammary carcinogenesis.

    Lab Invest 1987, 57:112-137. PubMed Abstract OpenURL

  39. Blakely CM, Stoddard AJ, Belka GK, Dugan KD, Notarfrancesco KL, Moody SE, D'Cruz CM, Chodosh LA: Hormone-induced protection against mammary tumorigenesis is conserved in multiple rat strains and identifies a core gene expression signature induced by pregnancy.

    Cancer Res 2006, 66:6421-6431. PubMed Abstract | Publisher Full Text OpenURL

  40. Bernstein L, Pike MC, Ross RK, Judd HL, Brown JB, Henderson BE: Estrogen and sex hormone-binding globulin levels in nulliparous and parous women.

    J Natl Cancer Inst 1985, 74:741-745. PubMed Abstract OpenURL

  41. Dorgan JF, Reichman ME, Judd JT, Brown C, Longcope C, Schatzkin A, Campbell WS, Franz C, Kahle L, Taylor PR: Relationships of age and reproductive characteristics with plasma estrogens and androgens in premenopausal women.

    Cancer Epidemiol Biomarkers Prev 1995, 4:381-386. PubMed Abstract | Publisher Full Text OpenURL

  42. Garcia-Closas M, Herbstman J, Schiffman M, Glass A, Dorgan JF: Relationship between serum hormone concentrations reproductive history alcohol consumption and genetic polymorphisms in pre-menopausal women.

    Int J Cancer 2002, 102:172-178. PubMed Abstract | Publisher Full Text OpenURL

  43. Bernstein L, Hanisch R, Sullivan-Halley J, Ross RK: Treatment with human chorionic gonadotropin and risk of breast cancer.

    Cancer Epidemiol Biomarkers Prev 1995, 4:437-440. PubMed Abstract | Publisher Full Text OpenURL

  44. Russo IH, Koszalka M, Russo J: Human chorionic gonadotropin and rat mammary cancer prevention.

    J Natl Cancer Inst 1990, 82:1286-1289. PubMed Abstract | Publisher Full Text OpenURL

  45. Russo J, Russo IH: Toward a physiological approach to breast cancer prevention.

    1994, 3:353-364.

    1055-9965

    PubMed Abstract | Publisher Full Text OpenURL

  46. Petrakis NL, Wrensch MR, Ernster VL, Miike R, Murai J, Simberg N, Siiteri PK: Influence of pregnancy and lactation on serum and breast fluid estrogen levels: implications for breast cancer risk.

    Int J Cancer 1987, 40:587-591. PubMed Abstract | Publisher Full Text OpenURL

  47. Russo J, Hu YF, Yang X, Russo IH: Developmental, cellular and molecular basis of human breast cancer.

    J Natl Cancer Inst Monogr 2000, 27:17-37. PubMed Abstract | Publisher Full Text OpenURL

  48. Murrell TG: Epidemiological and biochemical support for a theory on the cause and prevention of breast cancer.

    Med Hypotheses 1991, 36:389-396. PubMed Abstract | Publisher Full Text OpenURL

  49. McCredie M, Paul C, Skegg DC, Williams S: Reproductive factors and breast cancer in New Zealand.

    Int J Cancer 1998, 76:182-188. PubMed Abstract | Publisher Full Text OpenURL

  50. Newcomb PA, Storer BE, Longnecker MP, Mittendorf R, Greenberg ER, Clapp RW, Burke KP, Willett WC, MacMahon B: Lactation and a reduced risk of premenopausal breast cancer.

    N Engl J Med 1994, 330:81-87. PubMed Abstract | Publisher Full Text OpenURL

  51. Tryggvadottir L, Tulinius H, Eyfjord JE, Sigurvinsson T: Breastfeeding and reduced risk of breast cancer in an Icelandic cohort study.

    Am J Epidemiol 2001, 154:37-42. PubMed Abstract | Publisher Full Text OpenURL

  52. Zheng T, Duan L, Liu Y, Zhang B, Wang Y, Chen Y, Zhang Y, Owens PH: Lactation reduces breast cancer risk in Shandong Province China.

    Am J Epidemiol 2000, 152:1129-1135. PubMed Abstract | Publisher Full Text OpenURL

  53. Enger SM, Ross RK, Paganini-Hill A, Bernstein L: Breastfeeding experience and breast cancer risk among postmenopausal women.

    Cancer Epidemiol Biomarkers Prev 1998, 7:365-369. PubMed Abstract | Publisher Full Text OpenURL

  54. Newcomb PA, Egan KM, Titus-Ernstoff L, Trentham-Dietz A, Greenberg ER, Baron JA, Willett WC, Stampfer MJ: Lactation in relation to postmenopausal breast cancer.

    Am J Epidemiol 1999, 150:174-182. PubMed Abstract | Publisher Full Text OpenURL

  55. Romieu I, Hernandez-Avila M, Lazcano E, Lopez L, Romero-Jaime R: Breast cancer and lactation history in Mexican women.

    Am J Epidemiol 1996, 143:543-552. PubMed Abstract | Publisher Full Text OpenURL

  56. Michels KB, Trichopoulos D, Rosner BA, Hunter DJ, Colditz GA, Hankinson SE, Speizer FE, Willett WC: Being breastfed in infancy and breast cancer incidence in adult life: results from the two nurses' health studies.

    Am J Epidemiol 2001, 153:275-283. PubMed Abstract | Publisher Full Text OpenURL

  57. Negri E, Braga C, La Vecchia C, Levi F, Talamini R, Franceschi S: Lactation and the risk of breast cancer in an Italian population.

    Int J Cancer 1996, 67:161-164. PubMed Abstract | Publisher Full Text OpenURL

  58. Ursin G, Bernstein L, Wang Y, Lord SJ, Deapen D, Liff JM, Norman SA, Weiss LK, Daling JR, Marchbanks PA, Malone KE, Folger SG, McDonald JA, Burkman RT, Simon MS, Strom BL, Spirtas R: Reproductive factors and risk of breast carcinoma in a study of white and African-American women.

    Cancer 2004, 101:353-362. PubMed Abstract | Publisher Full Text OpenURL

  59. Ma H, Wang Y, Sullivan-Halley J, Weiss L, Burkman RT, Simon MS, Malone KE, Strom BL, Ursin G, Marchbanks PA, McDonald JA, Spirtas R, Press MF, Bernstein L: Breast cancer receptor status: do results from a centralized pathology laboratory agree with SEER registry reports?

    Cancer Epidemiol Biomarkers Prev 2009, 18:2214-2220. PubMed Abstract | Publisher Full Text OpenURL

  60. Chu KC, Anderson WF, Fritz A, Ries LA, Brawley OW: Frequency distributions of breast cancer characteristics classified by estrogen receptor and progesterone receptor status for eight racial/ethnic groups.

    Cancer 2001, 92:37-45. PubMed Abstract | Publisher Full Text OpenURL

  61. Ursin G, Bernstein L, Lord SJ, Karim R, Deapen D, Press MF, Daling JR, Norman SA, Liff JM, Marchbanks PA, Folger SG, Simon MS, Strom BL, Burkman RT, Weiss LK, Spirtas R: Reproductive factors and subtypes of breast cancer defined by hormone receptor and histology.

    Br J Cancer 2005, 93:364-371. PubMed Abstract | Publisher Full Text | PubMed Central Full Text OpenURL