|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2005.03.7432 on March 27 2006 © 2006 American Society of Clinical Oncology. Endogenous Steroid Hormone Concentrations and Risk of Breast Cancer: Does the Association Vary by a Woman's Predicted Breast Cancer Risk?
From the Channing Laboratory, Department of Medicine and the Department of Obstetrics, Gynecology and Reproductive Biology, Brigham and Women's Hospital and Harvard Medical School; Department of Epidemiology, Harvard School of Public Health, Boston, MA Address reprint requests to Heather Eliassen, ScD, Channing Laboratory, Brigham and Women's Hospital, 181 Longwood Avenue, Boston, MA 02115; e-mail: heather.eliassen{at}channing.harvard.edu
PURPOSE: To examine whether the associations of endogenous estrogens and testosterone with breast cancer risk differ between high- and low-risk women, as determined by the Gail model and the Rosner and Colditz model, and by family history of breast cancer. METHODS: We conducted a prospective nested case-control study within the Nurses' Health Study. From 1989 or 1990 until June 2000, blood samples were collected, 418 breast cancer patient cases were identified, and two controls (total n = 817) were matched to each case. We classified women as high or low risk based on their family history of breast cancer, their 5-year Gail risk score, and their 5-year Rosner and Colditz risk score. Multivariate relative risks (RR) and 95% CI were calculated by unconditional logistic regression, adjusting for matching and breast cancer risk factors.
RESULTS: Estrone sulfate was statistically significantly associated with breast cancer risk among women with low (< 1.66%) and high ( CONCLUSION: These data suggest that higher levels of endogenous estrogens and testosterone are associated with increased breast cancer risk, regardless of predicted risk or family history of breast cancer.
Sex steroid hormones have long been hypothesized to increase breast cancer risk, and epidemiologic evidence provides strong support for the etiologic role of endogenous hormones.1-3 Women whose estrogen or androgen concentrations are in the highest quintile have a two- to threefold increased breast cancer risk as compared with those in the lowest quintile.1 However, most epidemiologic studies of these associations have been conducted in populations of women with average breast cancer risk.2-16 Although it is plausible that hormones are associated with breast cancer in high-risk individuals, no association was observed between breast cancer and estradiol or testosterone among high-risk women in the National Surgical Adjuvant Breast and Bowel Project (NSABP) P-1 tamoxifen prevention trial.17 To our knowledge, this is the only study to have evaluated this issue specifically. Models used to predict risk, such as the Gail18 and the Rosner and Colditz19,20 models, do not currently incorporate information on hormone levels. Confirmation of an association between hormone levels and breast cancer in these higher-risk subgroups would further support the inclusion of hormone concentrations into prediction models used to identify high-risk women who could benefit most from increased screening or chemoprevention. Within the prospective Nurses' Health Study (NHS), we previously investigated the relationship of circulating sex steroids and breast cancer risk among postmenopausal women (322 patient cases with follow-up from 1990 through 1998).3 We have now extended this study (418 cases with follow-up through 2000) to examine these associations in accordance with both predicted risk and family history of breast cancer.
Study Population In 1976, 121,700 female, married, registered nurses, ages 30 to 55 years, were enrolled in the NHS. Biennially, participants completed mailed questionnaires on various exposures, including many breast cancer risk factors, and new disease diagnoses. From 1989 to 1990, blood samples were collected from 32,826 cohort members, aged 43 to 69 years. Details regarding the blood collection methods have been published previously.7,21 Briefly, each woman arranged to have her blood drawn and shipped, via overnight courier with an ice pack, to our laboratory, where it was processed; 97% of samples arrived at our laboratory within 26 hours of being drawn. The stability of estrogens and androgens in whole blood for 24 to 48 hours has been documented previously.22 Samples have been stored in continuously monitored liquid nitrogen freezers since collection. As of 2000, the follow-up rate among those who gave blood was 99%. The study was approved by the Committee on the Use of Human Subjects in Research at the Brigham and Women's Hospital (Boston, MA); completion of the self-administered questionnaire was considered to imply informed consent. Case subjects and controls were selected from women who met the following critera at blood collection. The women were postmenopausal, which was defined as having a natural menopause with no periods in the last 12 months; or having bilateral oophorectomy; or having a hysterectomy with at least one ovary remaining in nonsmoking women who were at least 56 years old, or in current-smoker women who were at least 54 years old, which were the ages at which natural menopause had occurred in 90% of these groups in the overall cohort. Participants could not have used postmenopausal hormones (PMH) for at least 3 months prior to study commencement. Eleven thousand one hundred sixty-nine women met these criteria.
Case subjects had no reported cancer diagnosis before blood collection (other than nonmelanoma skin cancer) and were diagnosed with breast cancer between blood collection and June 1, 2000. Overall, 418 cases of breast cancer were reported (336 cases were invasive) and confirmed by medical record review (n = 412) or by verbal confirmation of the diagnosis by the nurse (n = 6). Time from blood collection to diagnosis ranged from less than 1 month to 129 months (median, 66 months). Two control subjects (total, n = 817) were matched per case by age (± 2 years), and month (± 1 month), time of day (± 2 hours), and fasting status at blood collection (
Laboratory Analyses All case-control-control sets were assayed together in the same batch. Samples were ordered by random assignment and labeled so that case-control status was masked to the laboratories. The inter-assay coefficient of variation (CV) from masked replicate plasma samples in each batch ranged from 6% to 13%. When plasma hormone values were reported as less than the detection limit (estradiol, 2 pg/mL; estrone sulfate, 40 pg/mL; testosterone, 1 ng/dL), we set the value to half this limit (estradiol, n = 2; estrone sulfate, n = 7; testosterone, n = 4).
Covariate Data
Risk Prediction We also calculated predicted risk using the Rosner and Colditz model.20,32 This model is restricted to women with natural menopause or surgical menopause with bilateral oophorectomy, and includes age, age at menarche, age at and type of menopause, parity, age at first childbirth, history of benign breast disease, family history of breast cancer, current or past duration and type of PMH use, body mass index (BMI), alcohol intake, and height. Five-year risk from 1990 was calculated using beta coefficients from a cumulative incidence model within the NHS with 3,221 cases from 1980 to 2000.
Statistical Analysis
We removed two matched sets from the analysis because the case subjects' estrogen values were in the premenopausal range. We studentized deviate many-outlier procedure33 to identify and exclude statistical outliers (three estradiol and three testosterone values). In addition, several women did not have an adequate volume of plasma for all assays; the final sample sizes vary by hormone.
Hormone analyses were stratified by 5-year predicted risk (Gail model and Rosner and Colditz model scores) and by the presence or absence of family history of breast cancer. Gail scores were dichotomized at 1.66% risk over 5 years, which is the average 5-year risk of a 60-year old women and the cut point used for inclusion in the NSABP P-1 trial.29 Because there is no established cut point for defining high-risk women using the Rosner and Colditz model, scores were dichotomized using the median score of cases and controls combined. To evaluate women at even higher predicted risk, we stratified by To estimate relative risks (RRs) and 95% CIs, we used unconditional logistic regression, controlling for matching and breast cancer risk factors. Estimates from age- and matching factor–adjusted regression models were similar to those from multivariate models, therefore, only multivariate results are presented. We calculated tests for trend by modeling the continuous natural log hormone level and calculating a Wald statistic. Interactions on the multiplicative scale between hormone levels and predicted risk or family history were evaluated by adding an interaction term (log hormone quartile medians x continuous risk score [or presence of family history]) to the logistic models and calculating a Wald statistic. All analyses were conducted using SAS software, version 8 (SAS Institute, Cary, NC).
Five-year Gail predicted risk ranged from 0.67% to 10.11% across all case subjects and controls (median = 1.80%; 75th percentile = 2.52%). Five-year predicted risks from the Rosner and Colditz model were lower overall than the Gail scores (range, 0.21% to 6.87%; median = 1.57%; 75th percentile = 2.04%). Means and proportions of risk factors used in both models increased across categories of risk, as expected (Table 1). For example, mean age and family history of breast cancer increased with higher predicted risk. Also as expected, variables included in the Rosner and Colditz model, but not the Gail model, such as BMI and parity, varied more by Rosner and Colditz risk score than by Gail score.
The prevalence of past PMH use differed between the two scores, with a lower overall prevalence among those participants with a Rosner and Colditz risk score. This is probably because 89% of these women had a natural menopause, compared with 72% of participants with a Gail score, and women with natural menopause were less likely to have used PMH (27% v 56% for other types of menopause). In addition, the Rosner and Colditz model assigns a negative beta-coefficient to past users20,32; thus, the prevalence of past use is lower with higher risk scores. Several risk factors in the Rosner and Colditz model, such as BMI and alcohol consumption, are associated with endogenous postmenopausal hormone levels.21,34,35 Thus, the risk predicted by this model had a weak, but statistically significant, correlation with the log hormone values (Pearson correlations: estradiol r = 0.17; estrone sulfate r = 0.12; testosterone r = 0.10). Gail predicted risk, however, was not correlated with any of the hormones (Pearson r = –0.08 to 0.03).
Estradiol was associated with similar increased breast cancer risk when the Gail risk score was dichotomized (for Gail risk scores < 1.66%: top v bottom quartile RR = 2.1; 95% CI, 1.2 to 3.6; for scores
Results stratified by the Rosner and Colditz risk score were similar to the Gail risk stratification (Table 3). Estradiol was associated with an increased risk in women with low predicted risk (< 1.57%; RR = 1.9; 95% CI, 1.0 to 3.4) and among women with predicted risk
Higher concentrations of estradiol were associated with an increased breast cancer risk, whether or not the participant had a family history of breast cancer (Table 4). Although the association seemed stronger in those participants with a family history (RR = 3.6; 95% CI, 1.4 to 9.6), the interaction was not statistically significant (P = .36). Whereas estrone sulfate and testosterone were associated with a significant increased risk only among women without family history of breast cancer, the interactions were not significant (P = .29; .48, respectively). Results using updated family history information (updated through 2000, adding 40 cases and 50 controls with a family history of breast cancer) were similar for estradiol and testosterone, whereas estrone sulfate became more strongly associated with risk among those participants with a family history of the disease (RR = 1.9; 95% CI, 0.9 to 3.9; Ptrend = .073).
When cases were restricted to estrogen and progesterone receptor (ER/PR) –positive tumors, results were similar to the overall results (data not shown). Given that age is one of the strongest predictors in both risk prediction models, we repeated the analyses, stratifying by risk within age groups, and results were unchanged. To ensure that the larger population in the Gail analysis did not induce any important differences between the Gail analyses and the Rosner and Colditz analyses, we compared Gail scores between those participants with scores from both models and those with scores from the Gail model only and observed similar distributions. In addition, when we repeated the Gail analysis and restricted it to women for whom we could calculate both scores (281 cases, 606 controls), results were unchanged. Finally, we repeated all analyses after excluding cases diagnosed within 2 years of blood collection, and the results were unchanged.
In this large nested case-control analysis, we observed statistically significant associations of postmenopausal plasma estrogens and testosterone with breast cancer risk, regardless of risks predicted by the Gail model or the Rosner and Colditz model, or by family history of breast cancer. Although estradiol appeared more strongly associated with breast cancer risk among higher-risk women, none of the differences between risk groups were statistically significant and we did not observe a similar pattern with estrone sulfate. Thus, the differing estradiol risk estimates across predicted risk score are more likely due to chance than indicative of biologic differences. Similarly, the lack of association observed with estrone sulfate and testosterone in women with a positive family history of breast cancer could be a chance finding, and indeed, when family history information was updated (resulting in a larger number of women with family history), the point estimates for estrone sulfate increased and were nearly statistically significant. High estradiol and testosterone levels have been consistently associated with an increased breast cancer risk in postmenopausal women.1-3 However, most of the studies to date have been conducted in cohorts that were not selected in accordance with breast cancer risk, and hence, the women represent a more general population sample.2-16 The association among high-risk women has been examined in only one previous study, a case-cohort analysis within the P-1 trial,17 and no statistically significant association was observed between estradiol or testosterone and breast cancer risk in the placebo group. However, this group included only 89 case subjects, and therefore statistical power was limited to detect modest associations. Although the average predicted Gail risk in our study was lower than in the P-1 trial, the distribution of risk among women above the 75th percentile of risk in our study was comparable with the risk in P-1 trial’s population (P-1 trial median scores = 3.48% in cases, 2.90% in controls).17 Our results suggest that estrogen and testosterone are important risk factors, regardless of predicted breast cancer risk. Estrogen promotes cellular growth and contributes to tumor growth by promoting the proliferation of cells with existing mutations or perhaps by increasing the opportunity for mutations.36 Although many established risk factors might act via a hormonal pathway, it is biologically plausible, and likely, that estrogen can still act to promote growth among women who are considered at high risk because of parity, family history, or a combination of risk factors. This is supported by evidence that exogenous hormones increase breast cancer risk in postmenopausal women, regardless of family history.37-39 In addition, selective estrogen receptor modulators (SERMs), such as tamoxifen and raloxifene, prevent breast cancer in women at high predicted risk,29 and among women with BRCA-1 and BRCA-2 mutations.40 Thus, the overall body of evidence suggests that endogenous estrogen is an important risk factor, even among high-risk women. This study has several strengths, including the large number of cases enrolled. Blood samples and risk factor information were collected before diagnosis, minimizing the possibility of reverse causality or recall bias. In addition, the women included in this study, while at average risk overall, represent a broad range of predicted risk. The use of a modified Gail model to calculate risk is a potential limitation. Because we did not have information on the number of biopsies or the presence of atypical hyperplasia, the modified Gail model may have underestimated risk in the small subgroup of women with atypical hyperplasia. Nevertheless, we had information on all the other necessary risk factors and this modified model is still able to discriminate women by estimated risk.31 In addition, we used two different models to define predicted risk and observed similar results, suggesting that the associations are robust, regardless of how the risk was calculated. Overall, these data support the importance of endogenous steroid hormones in the etiology of breast cancer across all levels of predicted risk. Circulating levels of hormones may be important to consider when evaluating a woman's risk of breast cancer, even among women who are at higher predicted risk, based on predictive models or family history.
The authors indicated no potential conflicts of interest.
We thank Bernard Rosner, PhD, and Graham Colditz, MD, DrPH, for their insightful comments. We gratefully acknowledge the Nurses' Health Study participants for their continuing cooperation.
Supported by Research Grants No. CA49449 and CA87969 from the National Cancer Institute, Cancer Education and Career Development Grant No. R25 CA098566-02 from the National Cancer Institute (A.H.E.), and in part by Training Grant in Cancer Epidemiology No. T32 CA090001-281 from the National Cancer Institute (S.A.M. and S.S.T.). Presented in part at the American Association for Cancer Research 96th Annual Meeting, Anaheim, CA, April 16-20, 2005 (poster session). Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
1. Endogenous Hormones and Breast Cancer Collaborative Group: Endogenous sex hormones and breast cancer in postmenopausal women: Reanalysis of nine prospective studies. J Natl Cancer Inst 94:606-616, 2002 2. Zeleniuch-Jacquotte A, Shore RE, Koenig KL, et al: Postmenopausal levels of oestrogen, androgen, and SHBG and breast cancer: Long-term results of a prospective study. Br J Cancer 90:153-159, 2004[CrossRef][Medline] 3. Missmer SA, Eliassen AH, Barbieri RL, et al: Endogenous estrogen, androgen, and progesterone concentrations and breast cancer risk among postmenopausal women. J Natl Cancer Inst 96:1856-1865, 2004 4. Dorgan JF, Longcope C, Stephenson HE Jr, et al: Relation of prediagnostic serum estrogen and androgen levels to breast cancer risk. Cancer Epidemiol Biomarkers Prev 5:533-539, 1996[Abstract] 5. Dorgan JF, Stanczyk FZ, Longcope C, et al: Relationship of serum dehydroepiandrosterone (DHEA), DHEA sulfate, and 5-androstene-3 beta, 17 beta-diol to risk of breast cancer in postmenopausal women. Cancer Epidemiol Biomarkers Prev 6:177-181, 1997[Abstract] 6. Thomas HV, Key TJ, Allen DS, et al: A prospective study of endogenous serum hormone concentrations and breast cancer risk in post-menopausal women on the island of Guernsey. Br J Cancer 76:401-405, 1997[Medline] 7. Hankinson SE, Willett WC, Manson JE, et al: Plasma sex steroid hormone levels and risk of breast cancer in postmenopausal women. J Natl Cancer Inst 90:1292-1299, 1998 8. Toniolo PG, Levitz M, Zeleniuch-Jacquotte A, et al: A prospective study of endogenous estrogens and breast cancer in postmenopausal women. J Natl Cancer Inst 87:190-197, 1995 9. Zeleniuch-Jacquotte A, Bruning PF, Bonfrer JM, et al: Relation of serum levels of testosterone and dehydroepiandrosterone sulfate to risk of breast cancer in postmenopausal women. Am J Epidemiol 145:1030-1038, 1997 10. Berrino F, Muti P, Micheli A, et al: Serum sex hormone levels after menopause and subsequent breast cancer. J Natl Cancer Inst 88:291-296, 1996 11. Barrett-Connor E, Friedlander NJ, Khaw KT: Dehydroepiandrosterone sulfate and breast cancer risk. Cancer Res 50:6571-6574, 1990 12. Garland CF, Friedlander NJ, Barrett-Connor E, et al: Sex hormones and postmenopausal breast cancer: A prospective study in an adult community. Am J Epidemiol 135:1220-1230, 1992 13. Kabuto M, Akiba S, Stevens RG, et al: A prospective study of estradiol and breast cancer in Japanese women. Cancer Epidemiol Biomarkers Prev 9:575-579, 2000 14. Cauley JA, Lucas FL, Kuller LH, et al: Elevated serum estradiol and testosterone concentrations are associated with a high risk for breast cancer. Ann Intern Med 130:270-277, 1999 15. Gordon GB, Bush TL, Helzlsouer KJ, et al: Relationship of serum levels of dehydroepiandrosterone and dehydroepiandrosterone sulfate to the risk of developing postmenopausal breast cancer. Cancer Res 50:3859-3862, 1990 16. Helzlsouer KJ, Alberg AJ, Bush TL, et al: A prospective study of endogenous hormones and breast cancer. Cancer Detect Prev 18:79-85, 1994[Medline] 17. Beattie MS, Costantino JP, Cummings SR, et al: Endogenous sex hormones, breast cancer risk, and tamoxifen response: An ancillary study in the NSABP breast cancer prevention trial (P1). J Natl Cancer Inst 98:110-115, 2006 18. Gail MH, Brinton LA, Byar DP, et al: Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst 81:1879-1886, 1989 19. Rosner B, Colditz GA: Nurses' health study: Log-incidence mathematical model of breast cancer incidence. J Natl Cancer Inst 88:359-364, 1996 20. Colditz GA, Rosner B: Cumulative risk of breast cancer to age 70 years according to risk factor status: Data from the Nurses' Health Study. Am J Epidemiol 152:950-964, 2000 21. Hankinson SE, Willett WC, Manson JE, et al: Alcohol, height, and adiposity in relation to estrogen and prolactin levels in postmenopausal women. J Natl Cancer Inst 87:1297-1302, 1995 22. Hankinson SE, London SJ, Chute CG, et al: Effect of transport conditions on the stability of biochemical markers in blood. Clin Chem 35:2313-2316, 1989 23. Abraham GE, Tulchinsky D, Korenman SG: Chromatographic purification of estradiol-17 for use in radio-ligand assay. Biochem Med 3:365-368, 1970[CrossRef][Medline] 24. Mikhail G, Chung HW: Radioimmunoassay of plasma estrogens: Use of polymerized antibodies, in Peron FG, Caldwells BV (eds): Immunological methods in steroid determination. New York, NY, Appleton-Century-Croft, 1970, pp 113 25. Mikhail G, Wu CH, Ferin M, et al: Radioimmunoassay of plasma estrone and estradiol. Steroids 15:333-352, 1970[CrossRef][Medline] 26. Kinouchi T, Pages L, Horton R: A specific radioimmunossay for testosterone in peripheral plasma. J Lab Clin Med 82:309-316, 1973[Medline] 27. McNatty KP, Smith DM, Makris A, et al: The microenvironment of the human antral follicle: Interrelationships among the steroid levels in antral fluid, the population of granulosa cells, and the status of the oocyte in vivo and in vitro. J Clin Endocrinol Metab 49:851-860, 1979[Medline] 28. Franz C, Watson D, Longcope C: Estrone sulfate and dehydroepiandrosterone sulfate concentrations in normal subjects and men with cirrhosis. Steroids 34:563-573, 1979[CrossRef][Medline] 29. Fisher B, Costantino JP, Wickerham DL, 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 90:1371-1388, 1998 30. Reis LAG, Eisner MP, Kosary CL, et al: SEER Cancer Statistics Review, 1975-2002. Bethesda, MD, National Cancer Institute, 2005. http://seer.cancer.gov/csr/1975_2002/ 31. Rockhill B, Spiegelman D, Byrne C, et al: Validation of the Gail, et al: Model of breast cancer risk prediction and implications for chemoprevention. J Natl Cancer Inst 93:358-366, 2001 32. Rockhill B, Byrne C, Rosner B, et al: Breast cancer risk prediction with a log-incidence model: Evaluation of accuracy. J Clin Epidemiol 56:856-861, 2003[CrossRef][Medline] 33. Rosner B: Percentage points for a generalized ESD many-outlier procedure. Technometrics 25:165-172, 1983[CrossRef] 34. Lukanova A, Lundin E, Zeleniuch-Jacquotte A, et al: Body mass index, circulating levels of sex-steroid hormones, IGF-I and IGF-binding protein-3: A cross-sectional study in healthy women. Eur J Endocrinol 150:161-171, 2004[Abstract] 35. Onland-Moret NC, Peeters PH, van der Schouw YT, et al: Alcohol and endogenous sex steroid levels in postmenopausal women: A cross-sectional study. J Clin Endocrinol Metab 90:1414-1419, 2005 36. Henderson BE, Feigelson HS: Hormonal carcinogenesis. Carcinogenesis 21:427-433, 2000 37. Colditz GA, Rosner BA, Speizer FE: Risk factors for breast cancer according to family history of breast cancer: For the Nurses' Health Study Research Group. J Natl Cancer Inst 88:365-371, 1996 38. Magnusson C, Colditz G, Rosner B, et al: Association of family history and other risk factors with breast cancer risk (Sweden). Cancer Causes Control 9:259-267, 1998[CrossRef][Medline] 39. Ursin G, Tseng CC, Paganini-Hill A, et al: Does menopausal hormone replacement therapy interact with known factors to increase risk of breast cancer? J Clin Oncol 20:699-706, 2002 40. King MC, Wieand S, Hale K, et al: Tamoxifen and breast cancer incidence among women with inherited mutations in BRCA1 and BRCA2: National Surgical Adjuvant Breast and Bowel Project (NSABP-P1) Breast Cancer Prevention Trial. JAMA 286:2251-2256, 2001 Submitted August 5, 2005; accepted December 15, 2005. Related Editorial
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2.52%; 75th percentile) Gail predicted risk (fourth v first quartile RR = 3.6; 95% CI, 1.9 to 7.0; RR = 2.5; 95% CI, 1.2 to 5.1, respectively). Testosterone results were similar across strata of predicted risk, with two times the risk in the fourth (v first) quartile. Estradiol appeared more strongly associated with breast cancer in women with higher predicted risk (RR = 4.5; 95% CI, 2.1 to 9.5) compared with women with lower risk (RR = 2.1; 95% CI, 1.2 to 3.6), but differences were not statistically significant. Results were similar across predicted Rosner and Colditz risk scores. 
