© The Author 2006. Published by Oxford University Press.
ARTICLE |
Endogenous Sex Hormones, Breast Cancer Risk, and Tamoxifen Response: An Ancillary Study in the NSABP Breast Cancer Prevention Trial (P-1)
Affiliations of authors: University of California, San Francisco, Department of Medicine, Division of General Internal Medicine, San Francisco, CA (MSB, SRC); University of Pittsburgh, Department of Medicine, Division of Oncology, Pittsburgh, PA (VGV); National Surgical Adjuvant Breast and Bowel Project, Pittsburgh, PA (JPC, DLW, NW); Academic Department of Biochemistry, Royal Marsden Hospital, London, England (MD, EJF); Harvard School of Public Health, Department of Epidemiology; Channing Laboratory, Department of Medicine, Brigham and Women's Hospital and Harvard Medical School, Boston, MA (WCW, SEH); California Pacific Medical Center, San Francisco, CA (SRC); Harvard School of Public Health, Department of Nutrition, Boston, MA (WCW)
Correspondence to: Mary S. Beattie, MD, MAS, 1635 Divisadero St., Ste. 600, San Francisco, CA 94115 (e-mail: mary.beattie{at}ucsfmedctr.org).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Background: Prospective studies have shown an association between increased serum levels of estradiol and testosterone and breast cancer risk in postmenopausal women. Raloxifene has been shown to reduce breast cancer risk more in women with high estradiol levels than in those with lower levels. The purpose of this study was to determine whether sex hormone levels were associated with breast cancer risk and with response to tamoxifen in a high-risk population. Methods: Using a case–cohort design, we studied 135 women with postmenopausal breast cancer and 275 postmenopausal women without breast cancer who were enrolled in the National Surgical Adjuvant Breast and Bowel Project Cancer Prevention Trial (P-1) and who had been treated with tamoxifen or placebo for 69 months. We measured estradiol, testosterone, and sex hormone–binding globulin by using radioimmunoassay in baseline plasma samples. Relative risks (RRs) and 95% confidence intervals (CIs) for invasive breast cancer were estimated for each quartile of sex hormone level using Cox proportional hazards models. All statistical tests were two-sided. Results: Median plasma levels of estradiol, testosterone, and sex hormone–binding globulin were similar between the case and cohort groups. The relative risk of breast cancer for women in the placebo group was not associated with sex hormone levels (risk of estrogen receptor–positive breast cancer in women by quartile of estradiol: Q1 [lowest], RR = 1.0; Q2, RR = 1.16, 95% CI = 0.49 to 2.7; Q3, RR = 1.08, 95% CI = 0.45 to 2.61; and Q4, RR = 1.29, 95% CI = 0.59 to 2.82). The reduced risk of invasive breast cancer in tamoxifen-treated women compared with placebo-treated women was not associated with sex hormone levels. Conclusions: These data do not support the use of endogenous sex hormone levels to identify women who are at particularly high risk of breast cancer and who are most likely to benefit from chemoprevention with tamoxifen.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Both tamoxifen and raloxifene reduce the risk of estrogen receptor–positive (ER+) breast cancer by competing with endogenous estradiol at estrogen receptors in breast tissue (1,2). These selective estrogen receptor modulators (SERMs) have been studied in several prevention trials (3–5). In the National Surgical Adjuvant Breast and Bowel Project (NSABP) Breast Cancer Prevention Trial (P-1) (3), tamoxifen reduced the overall risk of breast cancer by 49% in high-risk women who were randomly assigned to this SERM over 69 months (relative risk [RR] = 0.51, 95% confidence interval [CI] = 0.39 to 0.66). After 7 years of follow-up in P-1, this risk reduction persisted (RR = 0.57, 95% CI = 0.46 to 0.70) (6). In the NSABP P-1 trial, high risk was defined as having at least a 1.66% 5-year risk of breast cancer as estimated from the modified Gail model (7). In the Continuing Outcomes Relevant to Evista (CORE) trial, raloxifene reduced the risk of invasive breast cancer by 66% in older women with osteoporosis over 8 years (RR = 0.34, 95% CI = 0.22 to 0.50) (5).
Many prospective studies have shown an increased risk of postmenopausal breast cancer with increased endogenous serum levels of estradiol and testosterone and with decreased levels of sex hormone–binding globulin (SHBG) (8–13). Women in the Multiple Outcomes of Raloxifene Evaluation (MORE) trial with serum estradiol levels in the highest tertile (>10 pmol/L) had not only the greatest risk of breast cancer but also the greatest reduction in the relative risk of breast cancer with raloxifene (14). Women with estradiol levels less than the 5-pmol/L limit of detection had very low risk of breast cancer and no further reduction in risk with raloxifene.
We conducted this ancillary study within the NSABP P-1 trial to determine whether plasma levels of estradiol, testosterone, or SHBG were associated with breast cancer risk in a high-risk population. We also tested the hypothesis that women with the highest levels of estradiol and testosterone would achieve the greatest risk reduction with tamoxifen.
|
SUBJECTS AND METHODS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Eligibility
The complete eligibility requirements for participants in NSABP P-1 have been described previously (3). Eligible participants were required be 35 years or older with a 1.66% or greater risk of breast cancer over the next 5 years, as estimated by the model by Gail et al. (7); to have had lobular carcinoma in situ; or to be 60 years or older. Women who had used postmenopausal hormone therapy (PHT) were required to have discontinued hormones at least 6 months before entering the trial.
Study Subjects
The case patients in this ancillary study included all postmenopausal women in the NSABP P-1 trial with invasive breast cancer (n = 135 at the conclusion of the trial in 1998). Women 55 years and older were assumed to be postmenopausal. Women between 50 and 55 years were classified as postmenopausal if their plasma follicle stimulating hormone (FSH) level was greater than 50 IU/L and if their plasma estradiol concentration was less than 100 pmol/L. Also, women who reported having a bilateral oophorectomy and had plasma FSH concentrations greater than 50 IU/L and plasma estradiol concentrations less than 100 pmol/L were also considered to be postmenopausal.
For comparison, we selected a random sample of the whole cohort with oversampling to account for losses anticipated from the exclusion of premenopausal women (using the menopausal criteria above), with a goal of having two cohort subjects per case subject. The random cohort sample was stratified by treatment and age group, with a sampling fraction used to obtain an age distribution similar to that of the case subjects while maintaining equal numbers of tamoxifen- and placebo-treated women. Three placebo- and two tamoxifen-treated case subjects were randomly chosen in the random assignment of the cohort population. All participants in the NSABP P-1 trial provided written informed consent for trial participation. Internal review boards for the University of California–San Francisco, University of Pittsburgh, Allegheny University, and Harvard University approved the protocol for this ancillary study.
Specimen Collection, Storage, and Sex Hormone Assays
Blood samples were obtained from all NSABP P-1 participants at entry to the initial study. At that time, plasma was separated from whole blood and stored at –70 °C. Stored plasma was sent to the Royal Marsden Hospital (London) for sex hormone analyses. Estradiol was measured by radioimmunoassay after extraction with diethyl ether (15). The within- and between-batch coefficients of variation were 8.6% and 10%, respectively, at a concentration of 26 pmol/L. The detection limit of the assay was 3.0 pmol/L. SHBG was measured using a liquid-phase immunoradiometric kit (Orion Diagnostica, Espoo, Finland). Within- and between-batch coefficients of variation were 5.7% and 11%, respectively, at a concentration of 10 nmol/L, and the detection limit was 1.3 nmol/L. Testosterone was measured using a solid-phase radioimmunoassay kit (Diagnostic Products Corporation, Los Angeles, CA). Within- and between-batch coefficients of variation were 3.2% and 9.1%, respectively, at a concentration of 2.9 nmol/L, and the detection limit was 0.14 nmol/L. FSH was measured using an Abbott Axsym autoanalyzer (Abbott Laboratories, Abbott Park, IL). The detection limit for the FSH assay was 0.37 IU/L, and the within- and between-run precision, 3.7% and 2.1%, respectively, at a concentration of 26 IU/L.
Statistical Analyses
The distributions of estradiol, testosterone, and SHBG were examined by box plots, and each distribution was divided into quartiles based on the distribution in the cohort population. The Wilcoxon rank sum test was used to evaluate the statistical significance of differences in sex hormone levels between the case and cohort subjects.
Median values for age, body mass index (BMI, in kg/m2 of body surface area), and Gail risk score were compared between the case and cohort groups using a Wilcoxon rank sum test. Chi-squared analyses were used to evaluate the statistical significance of differences in race, smoking status (never, former, current), physical activity (inactive, light, moderate, vigorous), prior PHT use (ever or never), atypical hyperplasia, and number of first-degree relatives with breast cancer between the case and cohort groups.
To evaluate the relationship of hormone levels with the risk of breast cancer in the normal biologic setting, the placebo group was initially used to determine rates of breast cancer within each quartile. The relative risk across quartiles and associated 95% confidence intervals were determined using a Cox proportional hazards model that incorporated modifications consistent with the nature of the case–cohort design of the study (16,17). Assumptions of proportionality were verified by the method described by Klein and Moeschberger (18), which creates artificial time-dependent covariates. All models used had adequate fit. Univariate and multivariable Cox modeling was used to obtain estimates of relative risk. Similar analyses were used to evaluate the rates of breast cancer by quartile of sex hormone level in the tamoxifen group. The differences in breast cancer risk by treatment group (tamoxifen or placebo) across quartile of sex hormone level were evaluated by including an interaction term in the model.
After viewing the initial results of the unstratified analyses above, we analyzed the association between sex hormones and breast cancer risk following tamoxifen treatment in several subgroups based on prior PHT use, family history of breast cancer, Gail risk score, and age. For these analyses, prior PHT users were classified as never or ever users; family history was categorized as 0, 1, or 2 or more affected first-degree relatives; Gail risk score was divided at 3.00; and age was analyzed as less than 60 years versus 60 years and older. Cutpoints for Gail risk scores and age were chosen as approximate midpoints of the combined case and cohort groups. We also reanalyzed the results after subdividing the first (lowest) quartile of estradiol by its median (10 pmol/L) in the combined case and cohort groups to determine whether there was any association between breast cancer risk and very low estradiol levels. Also, we examined the correlation among plasma levels of estradiol and tamoxifen and BMI using a Spearman correlation coefficient. All statistical tests were two-sided, and P<.05 was considered statistically significant.
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Population Characteristics
The goal of a 2 : 1 ratio of cohort subjects to case subjects was achieved with 280 women in the cohort group and 135 women in the case group (Table 1). Because tamoxifen reduced the risk of breast cancer by approximately 50% in the NSABP P-1 trial, nearly twice as many case subjects randomly assigned to placebo than case subjects randomly assigned to tamoxifen were included in this study. Compared with the cohort group, the case group had a higher median Gail score and more first-degree relatives with breast cancer (Table 2). Baseline variables related to endogenous estradiol levels—including age, BMI, smoking status, previous use of postmenopausal hormones, and physical activity—were not statistically significantly different between the two groups.
|
|
Sex Hormone Levels and Breast Cancer Risk
The median plasma concentrations of estradiol, testosterone, and SHBG were similar between the case and cohort groups. The median plasma estradiol concentration among case subjects was 21 pmol/L (25th–75th percentile = 12 to 33 pmol/L), and that among cohort subjects was 22 pmol/L (25th–75th percentile = 13 to 32 pmol/L) (P = .69). The median testosterone plasma concentration among case subjects was 0.73 nmol/L (25th–75th percentile = 0.42 to 0.98 nmol/L) and among cohort subjects was 0.74 nmol/L (25th–75th percentile = 0.50 to 0.90 nmol/L) (P = .39). The median SHBG plasma concentration among case subjects was 33 nmol/L (25th–75th percentile = 25 to 49 nmol/L) and among cohort subjects was 32 nmol/L (25th–75th percentile = 24 to 46 nmol/L) (P = .86).
Plasma levels of estradiol and testosterone were statistically significantly related (Spearman correlation coefficient, r = .53, P<.001), as were levels of estradiol and SHBG (r = –.27, P<.001). Also, plasma estradiol levels were statistically significantly correlated with BMI (r = .40, P<.001).
The relative risk of invasive breast cancer and ER+ breast cancer were examined by quartile of plasma concentrations of estradiol, testosterone, and SHBG in the placebo group (Fig. 1). When compared with risks for women in the lowest quartile, the relative risks of breast cancer for the other quartiles were not statistically different for any of the hormone measurements. Although confidence intervals crossed 1.0 and were quite wide, the relative risk point estimates for both steroid hormones (estradiol and testosterone) for all breast cancer were actually less than 1.0. Also, although a positive association was suggested between ER+ breast cancer risk and higher estradiol plasma concentration, the confidence intervals were wide and the relative risks were not statistically significantly different from 1.0 (for the lowest to highest quartiles of estradiol, RR = 1.0; RR =1.16, 95% CI = 0.49 to 2.76; RR = 1.08, 95% CI = 0.45 to 2.61; and RR = 1.29, 95% CI = 0.59 to 2.82).
|
Although women treated with tamoxifen had a lower risk of invasive breast cancer than women treated with placebo, no statistically significant difference in relative risk of breast cancer by quartile of sex hormone concentration was observed among tamoxifen-treated women. Adjusting for Gail risk, BMI, smoking status, and physical activity level did not alter these results.
Sex Hormone Levels and Response to Tamoxifen
To assess whether the association between tamoxifen and breast cancer varied by estradiol level, we examined the relative risks of breast cancer according to plasma estradiol level in women who were randomly assigned to tamoxifen. Tamoxifen-treated women had lower risks for both all invasive breast cancer and ER+ breast cancer than placebo-treated women (Fig. 2); however, no statistically significant trend toward increased risk reduction with higher quartiles of estradiol was observed (for all breast cancer, Pinteraction = .89; for ER+ breast cancer, Pinteraction =.66).
|
Subgroup Analyses
Four hypothesis-driven subgroup analyses were performed. The first divided participants into ever versus never users of postmenopausal hormones. Both ever and never users showed no association between breast cancer risk and increasing plasma levels of estradiol, testosterone, and SHBG. Other variables examined in subgroup analyses included age, Gail score at baseline, and the number of affected first-degree relatives. Again, the association between sex hormone level and breast cancer risk did not appear to vary within strata of any of these variables, and confidence intervals were wide. Another analysis was performed using a cross-classification of dichotomized levels of plasma estradiol and testosterone (low estradiol and low testosterone, low estradiol and high testosterone, high estradiol and high testosterone, and high estradiol and low testosterone); none of these groups had a relative risk of breast cancer that differed statistically from 1.0. Finally, the first quartile of estradiol was divided at its median (10 pmol/L) to assess whether women with very low plasma estradiol concentrations had any difference in breast cancer risk compared with that in women with estradiol levels greater than 10 pmol/L. Compared with women with plasma estradiol concentrations of 10 pmol/L or less, the relative risk of breast cancer in the remaining women in the first quartile (11–13 pmol/L) was 2.52 (95% CI = 0.86 to 7.39), and the relative risk of breast cancer among all women with estradiol concentrations greater than 10 pmol/L was 1.64 (95% CI = 0.68 to 4.00). These results remained essentially unchanged after adjustment for Gail risk, BMI, smoking, atypical hyperplasia, physical activity, and prior PHT use.
|
DISCUSSION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
In this population of women at a high risk of breast cancer, levels of sex hormones were not associated with breast cancer risk. Also, women with the highest levels of plasma estradiol or testosterone had a similar reduction in risk from treatment with tamoxifen as did women with the lowest levels of these hormones. Thus, among women who have a high risk of breast cancer, endogenous plasma levels of estradiol or testosterone are not useful for identifying women who will benefit most from treatment with tamoxifen.
Several analyses were performed to examine whether certain subgroups within the NSABP P-1 study might have an increased risk of breast cancer with increasing levels of sex hormones, and these analyses confirmed a lack of association. One subgroup analysis suggested an increased relative risk in women with estradiol levels greater than 10 pmol/L compared with women with 10 pmol/L or less, but the increase was not statistically significant. Only approximately 10% of women in this study had plasma estradiol levels of 10 pmol/L or less. In the MORE study (14), however, the majority of women in this osteoporotic study had very low levels of estradiol. This raises the questions of a possible "threshold effect" (i.e., low risk at very low estradiol levels and equally high risk at estradol levels above a "threshold" value). However, many prior studies have shown increasing risk with increasing estradiol levels (8–13), suggesting no precedent for any "threshold effect".
The main difference in the results of this study, compared with prior studies, is the lack of an association between serum sex hormone levels and breast cancer risk. There are several potential explanations for why the findings in this study differ from those of other prospective studies. First, the NSABP P-1 population had a high baseline risk of breast cancer, as evidenced by high Gail scores, a high prevalence of atypical hyperplasia, and strong family histories of breast cancer. Most previous studies of the association between sex hormone level and breast cancer risk were in more general populations that were not at especially high risk of breast cancer (8–13). Approximately 8% of women in the NSABP P-1 trial had atypical hyperplasia, which is quite rare in the general population, and atypical hyperplasia, combined with other risk factors, may outweigh any effect of endogenous sex hormones.
In this study, tamoxifen-treated women within all quartiles of estradiol had a reduced risk of breast cancer compared with placebo-treated women. Although the differences were not statistically significant, women with higher levels of estradiol had a slightly greater risk reduction in risk of ER+ breast cancer with tamoxifen than women with lower levels of estradiol. These results differ from those of the MORE study (14), which showed a statistically significant interaction between estradiol and response to raloxifene. The MORE population, however, was not "high risk," and the MORE trial was designed to examine osteoporotic fractures primarily. It is not valid to compare estradiol levels determined by different laboratories, but the overall levels of estradiol of women in the NSABP P-1 trial are likely to be higher than those of women in the MORE trial population, who all had osteoporosis. The interaction of SERMs with baseline sex hormone level may be more evident at lower levels of estradiol.
Another possible hypothesis regarding the apparent lack of association between endogenous sex hormones and breast cancer risk in the NSABP P-1 trial relates to the relationship between tissue and serum levels of sex hormones. Estradiol levels in breast tumors can be 10- to 20-fold higher than serum levels due to local production of estrogen (19). Evidence from several (19–23) but not all (24,25) studies suggest that serum estradiol levels are not associated with tumor estradiol levels. Whether peripheral levels of estrogen and estrogen levels in normal breast tissue are associated is unclear. If the association between tissue and circulating hormone levels varies by a woman's underlying risk of breast cancer, with a weaker association among high-risk women, then that might explain why we saw no association in the NSABP P-1 population. The reduction in breast cancer risk seen with tamoxifen in this study in all quartiles of estradiol likely reflects the action of tamoxifen in breast tissue, regardless of circulating estradiol.
The methods used to assay sex hormones differ among laboratories, making between-study comparisons difficult. However, levels of estradiol, testosterone, and SHBG in this study were comparable to those in previous studies, in which similar methodology was used (9–13). In fact, the assay used in this study was the same as that used in the cohort (10) that has shown the strongest association between estradiol and breast cancer risk. Variables related to sex hormone levels, including BMI, were also similar in this study and in prior studies (9–13). Moreover, the association we observed between estradiol and testosterone, between estradiol and BMI, and the inverse association between estradiol and SHBG is similar to that in prior studies (11) with similar correlation coefficients. Thus, the measurement of sex hormone levels appears to be accurate here.
Approximately half of the women in this study had used postmenopausal hormone therapy in the past. Trial entry required at least 6 months to have elapsed between the time from the last use of postmenopausal hormones to enrollment, a period that greatly exceeds the half-life of all postmenopausal hormone preparations. Thus, past PHT use should not have affected current sex hormone measurements here. However, current sex hormone measurements are a better reflection of lifetime exposure to hormones in women who are never users of PHT than in those who have. For these never users, serum estradiol levels are likely to reflect long-term exposure to estradiol. For prior PHT users, however, we hypothesize that serum estradiol levels in this study would not as accurately reflect long-term estradiol exposure, because part of this exposure was pharmacologically determined. This hypothesis led to the subgroup analysis examining ever versus never users of PHT. Although the power for this subgroup analysis was limited, the lack of association of sex hormones with breast cancer risk was confirmed in never users of PHT.
The absence of an association may have been due to chance. However, this study had sufficient power to detect a 2.8-fold increase in breast cancer between the highest and lowest quartiles of estradiol in the placebo group. The magnitude of this increase is similar to that observed in previous studies (9–13). The number of case subjects (n = 89) in the placebo group was similar to, or larger than, that in most of the prior studies that have shown an association (8). In the MORE study, for example, there were only 59 case subjects with invasive breast cancer (14). The lack of a trend toward increasing risk of breast cancer with increasing sex hormone level, and the fact that several of the point estimates of relative risk were actually below 1.0, suggests that there is no association between sex hormone levels and breast cancer risk in this high-risk population. The subgroup analyses further corroborate this lack of association.
This ancillary study is unique because it is the first, to our knowledge, to examine endogenous sex hormone levels in a population at high risk for breast cancer. Because careful risk assessment was required for participants entering the NSABP P-1 trial, these women were assigned accurate Gail scores, had accurate family histories taken, and had careful confirmation of atypical breast tissue. In this high-risk population, endogenous sex hormone levels appear to add no value to risk assessment.
In summary, we found no association between endogenous estradiol, testosterone, or SHBG levels and risk of breast cancer among women who already have a high risk of breast cancer based on family history, reproductive risk factors for breast cancer, number of biopsies, and atypia. These results differ from those of previous prospective studies that observed strong associations between estradiol and testosterone levels and risk of breast cancer in populations that were not selected by their risk for breast cancer. Until our results are tested in other high-risk populations, we believe that, among women who already have a high risk of breast cancer based on the risk factors in the Gail score, sex hormone levels should not be used to assess breast cancer risk or to guide decisions about treatment with tamoxifen. Future studies should examine the association between sex hormones and breast cancer risk in other populations at high risk for breast cancer.
|
NOTES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Supported by grants from the UCSF/Mt. Zion Fund (MSB) and the DaCosta Fund for Breast Cancer Prevention (MSB, SRC, MD, EJF). Dr. Beattie is supported by the American Cancer Society's Cancer Control Career Development Award for Primary Care Physicians. The P-1 study was supported by Public Health Science grants NCI U10-CA-37377 and U10-CA-69974 from the National Cancer Institute, National Institutes of Health, Department of Health and Human Services.
We thank Barbara C. Good, PhD, for editorial assistance.
Dr. Wickerman is on the speaker's bureau of AstraZeneca.
This report is dedicated to the loving memory of Dr. Edwin L. Stanley Jr.
|
REFERENCES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
(1) Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, et al. Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 1997;389:753–8.[CrossRef][Medline]
(2) Grese TA, Sluka JP, Bryant HU, Cullinan GJ, Glasebrook AL, Jones CD, et al. Molecular determinants of tissue selectivity in estrogen receptor modulators. Proc Natl Acad Sci U S A 1997;94:14105–10.
(3) Fisher B, Costantino JP, Wickerham DL, Redmond CK, Kavanah M, Cronin WM, et al. Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 study. J Natl Cancer Inst 1998;90:1371–88.
(4) Cummings SR, Eckert S, Krueger KA, Grady D, Powles TJ, Cauley JA, et al. The effect of raloxifene on risk of breast cancer in postmenopausal women: Results from the MORE randomized trial. Multiple Outcomes of Raloxifene Evaluation. JAMA 1999;281:2189–97. Erratum in JAMA 1999;282:2124.
(5) Martino S, Cauley JA, Barrett-Connor E, Powles TJ, Mershon J, Disch D, et al. Continuing outcomes relevant to Evista: breast cancer incidence in postmenopausal osteoporotic women in a randomized trial of raloxifene. J Natl Cancer Inst 2004;96:1751–61.
(6) Fisher B, Constantino JP, Wickerham DL, Cecchini RS, Cronin WM, Robidoux A, et al. Tamoxifen for the prevention of breast cancer: Current status of the National Surgical Adjuvant Breast and Bowel Project P-1 Study. J Natl Cancer Inst 2005;97:1652–62.
(7) Costantino JP, Gail MH, Pee D, Anderson S, Redmond CK, Benichou J, et al. Validation studies for models projecting the risk of invasive and total breast cancer incidence. J Natl Cancer Inst 1999;91:1541–8.
(8) Key T, Appleby P, Barnes I, Reeves G. Endogenous sex hormones and breast cancer in postmenopausal women: reanalysis of nine prospective studies. J Natl Cancer Inst 2002;94:606–16.
(9) Missmer SA, Eliassen H, Barbier RL, Hankinson SE. Endogenous estrogen, androgen, and progesterone concentrations and breast cancer risk among postmenopausal women. J Natl Cancer Inst 2004;96:1856–65.
(10) Thomas HV, Key TJ, Allen DS, Moore JW, Dowsett M, Fentiman IS, et al. A prospective study of endogenous serum hormone concentrations and breast cancer risk in postmenopausal women on the island of Guernsey. Br J Cancer 1997;75:1075–9.[ISI][Medline]
(11) Cauley JA, Lucas FL, Kuller LH, Stone K, Browner W, Cummings SR. Elevated serum estradiol and testosterone concentrations are associated with a high risk for breast cancer. Study of Osteoporotic Fractures Research Group. Ann Intern Med 1999;130:270–7.
(12) Hankinson SE, Willett WC, Manson JE, Colditz GA, Hunter DJ, Spiegelman D, et al. Plasma sex steroid hormone levels and risk of breast cancer in postmenopausal women. J Natl Cancer Inst 1998;90:1292–9.
(13) Toniolo PG, Levitz M, Zeleniuch-Jacquotte A, Banerjee S, Koenig KL, Shore RE, et al. A prospective study of endogenous estrogens and breast cancer in postmenopausal women. J Natl Cancer Inst 1995;87:190–7.
(14) Cummings SR, Duong T, Kenyon E, Cauley JA, Whitehead M, Krueger KA. Serum estradiol level and risk of breast cancer during treatment with raloxifene. JAMA 2002;287:216–20.
(15) Dowsett M, Goss PE, Powles TJ, Hutchinson G, Brodie AM, Jeffcoate SL, et al. Use of the aromatase inhibitor 4-hydroxyandrostenedione in postmenopausal breast cancer: optimization of therapeutic dose and route. Cancer Res 1987;47:1957–61.
(16) Barlow WE. Robust variance estimation for the case-cohort design. Biometrics 1994;50:1064–72.[CrossRef][ISI][Medline]
(17) Therneau TM, Li H. Computing the Cox model for case cohort designs. Lifetime Data Anal 1999;5:99–112.[CrossRef][ISI][Medline]
(18) Klein JP, Moeschberger ML. Survival analysis: techniques for censored and truncated data. New York (NY): Springer; 1997. p. 276–8.
(19) Geisler J. Breast cancer tissue estrogens and their manipulation with aromatase inhibitors and inactivators. J Steroid Biochem Mol Biol 2003;86:245–53.[CrossRef][ISI][Medline]
(20) Simpson ER. Sources of estrogen and their importance. J Steroid Biochem Mol Biol 2003;86:225–30.[CrossRef][ISI][Medline]
(21) Pasqualini JR, Chetrite G, Blacker C, Feinstein MC, Delalonde L, Talbi M, et al. Concentrations of estrone, estradiol, and estrone suphate and evaluation of sulphatase and aromatase activities in pre-and postmenopausal breast cancer patients. J Clin Endocrinol Metab 1996;81:1460–4.[Abstract]
(22) Miller WR, Stuart M, Sahmoud T, Dixon JM. Anastrozole (‘Arimidex’) blocks oestrogen synthesis both peripherally and within the breast in postmenopausal women with large operable breast cancer. Br J Cancer 2002;87:950–5.[CrossRef][ISI][Medline]
(23) Vermuelen A, Deslypere JP, Paridaens R. Steroid dynamics in the normal and carcinomatous mammary gland. J Steroid Biochem 1986;5B:799–802.
(24) Recchione C, Venturelli E, Manzari A, Cavalleri A, Marinetti A, Secreto G. Testosterone, dihydrotestosterone and oestradiol levels in postmenopausal breast cancer tissues. J Steroid Biochem Mol Biol 1995;6:541–6.
(25) Mady EA, Ramadan EE, Ossman AA. Sex steroid hormones in serum and tissue of benign and malignant breast tumor patients. Dis Markers 2000;16:151–7.[ISI][Medline]
Manuscript received May 18, 2005; revised October 20, 2005; accepted December 7, 2005.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. L Shufelt and G. D Braunstein Testosterone and the breast Menopause Int, September 1, 2008; 14(3): 117 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Rock, S. W. Flatt, G. A. Laughlin, E. B. Gold, C. A. Thomson, L. Natarajan, L. A. Jones, B. J. Caan, M. L. Stefanick, R. A. Hajek, et al. Reproductive Steroid Hormones and Recurrence-Free Survival in Women with a History of Breast Cancer Cancer Epidemiol. Biomarkers Prev., March 1, 2008; 17(3): 614 - 620. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. T. Chlebowski, G. L. Anderson, D. S. Lane, A. K. Aragaki, T. Rohan, S. Yasmeen, G. Sarto, C. A. Rosenberg, F. A. Hubbell, and For the Women's Health Initiative Investigators Predicting Risk of Breast Cancer in Postmenopausal Women by Hormone Receptor Status J Natl Cancer Inst, November 21, 2007; 99(22): 1695 - 1705. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Tamimi, C. Byrne, G. A. Colditz, and S. E. Hankinson Endogenous Hormone Levels, Mammographic Density, and Subsequent Risk of Breast Cancer in Postmenopausal Women J Natl Cancer Inst, August 1, 2007; 99(15): 1178 - 1187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J Santen, N. F Boyd, R. T Chlebowski, S. Cummings, J. Cuzick, M. Dowsett, D. Easton, J. F Forbes, T. Key, S. E Hankinson, et al. Critical assessment of new risk factors for breast cancer: considerations for development of an improved risk prediction model Endocr. Relat. Cancer, June 1, 2007; 14(2): 169 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. G VOGEL Recent Results from Clinical Trials Using SERMs to Reduce the Risk of Breast Cancer Ann. N.Y. Acad. Sci., November 1, 2006; 1089(1): 127 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kendall and M. Dowsett Novel concepts for the chemoprevention of breast cancer through aromatase inhibition. Endocr. Relat. Cancer, September 1, 2006; 13(3): 827 - 837. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. H. Eliassen, S. A. Missmer, S. S. Tworoger, and S. E. Hankinson Endogenous Steroid Hormone Concentrations and Risk of Breast Cancer: Does the Association Vary by a Woman's Predicted Breast Cancer Risk? J. Clin. Oncol., April 20, 2006; 24(12): 1823 - 1830. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. G. Vogel and E. Taioli Have We Found the Ultimate Risk Factor for Breast Cancer? J. Clin. Oncol., April 20, 2006; 24(12): 1791 - 1794. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||