Combined conjugated esterified estrogen plus methyltestosterone supplementation and risk of breast cancer in postmenopausal women

Maturitas 79 (2014) 70–76

Contents lists available at ScienceDirect

Maturitas journal homepage: www.elsevier.com/locate/maturitas

Combined conjugated esterified estrogen plus methyltestosterone supplementation and risk of breast cancer in postmenopausal women Geoffrey C. Kabat a,∗ , Victor Kamensky a , Moonseong Heo a , Jennifer W. Bea b , Lifang Hou c , Dorothy S. Lane d , Simin Liu e , LiHong Qi f , Michael S. Simon g , Jean Wactawski-Wende h , Thomas E. Rohan a a

Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY, United States Department of Medicine, Arizona Cancer Center, Tucson, AZ, United States c Department of Preventive Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL, United States d Department of Preventive Medicine, School of Medicine, State University of New York at Stony Brook, Stony Brook, NY, United States e Departments of Epidemiology and Medicine, Brown University, Providence, RI, United States f Department of Public Health Sciences, School of Medicine, University of California at Davis, Davis, CA, United States g Karmanos Cancer Institute, Detroit, MI, United States h Department of Social and Preventive Medicine, School of Public Health and Health Professions, University at Buffalo, Buffalo, NY, United States b

a r t i c l e

i n f o

Article history: Received 18 March 2014 Received in revised form 5 June 2014 Accepted 6 June 2014 Keywords: Androgens Conjugated esterified estrogen plus methyltestosterone supplementation Breast cancer Postmenopausal women

a b s t r a c t Objectives: Testosterone supplementation is being prescribed increasingly to treat symptoms of hormone deficiency in pre- and postmenopausal women; however, studies of the association of testosterone therapy, alone or in combination with estrogen, with risk of breast cancer are limited. The current study assessed the association of combination conjugated esterified estrogen and methyltestosterone (CEE + MT) use and breast cancer risk in postmenopausal women in the Women’s Health Initiative (WHI). Study design: At Year 3 of follow-up, women in the WHI observational study (N = 71,964) provided information on CEE + MT use in the past two years, duration of use, and the brand name of the product. In addition, in each of years 4–8, women were asked whether they had used CEE + MT in the previous year. After 10 years of follow-up, 2832 incident breast cancer cases were identified. Cox proportional hazards models were used to estimate hazard ratios (HR) and 95% confidence intervals (95% CI) for the association of CEE + MT use (irrespective of use of other hormones) and of exclusive CEE + MT use in relation to breast cancer risk. Results: Neither CEE + MT use nor exclusive use of CEE + MT was associated with risk: multivariableadjusted HR 1.06, 95% CI 0.82–1.36 and HR 1.22, 95% CI 0.78–1.92, respectively. Among women with a natural menopause, the HR for exclusive use was 1.32 (95% CI 0.68–2.55). There was no indication of an association when repeated measures of CEE + MT use were included in a time-dependent covariates analysis. Conclusion: The present study, the largest prospective study to date, did not show a significant association of CEE + MT supplementation and risk of breast cancer. © 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Female sexual dysfunction and hypoactive sexual disorder are prevalent in postmenopausal women and are associated with reduced levels of endogenous testosterone [1]. A survey of

∗ Corresponding author at: Department of Epidemiology and Population Health, Albert Einstein College of Medicine, Bronx, NY 10461, United States. Tel.: +1 7184303038; fax: +1 7184308653. E-mail address: [email protected] (G.C. Kabat). http://dx.doi.org/10.1016/j.maturitas.2014.06.006 0378-5122/© 2014 Elsevier Ireland Ltd. All rights reserved.

prescriptions written by U.S. physicians revealed that 2 million prescriptions for testosterone for women were written in 2006–2007 [1]. This figure represents an increase since 2004. Furthermore, 21% of prescriptions for branded male testosterone products were written for women [1]. Combination oral esterified estrogen plus methyltestosterone (CEE + MT) has been widely used in the United States since the 1970s, although it has not been approved by the Food and Drug Administration (FDA) [2–4]. It is marketed for treatment of hot flashes [3]. The most commonly used products contain 1.25 mg of esterified estrogens and 2.5 mg of methyltestosterone

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

(introduced in 1964) and 0.625 mg esterified estrogens and 1.25 mg methyltestosterone (introduced in 1974) [2]. Because these products contained relatively high dosages of estrogen and testosterone, they have largely been replaced by non-oral preparations, particularly the transdermal patch [5]. Whether oral CEE + MT therapy influences a woman’s risk of breast cancer is unknown [6–14]. There is suggestive evidence that the risk of breast cancer may vary according to the balance between estrogen and testosterone (i.e., higher E-to-T ratio) [8]. Testosterone supplementation may have indirect effects on breast cancer risk by modifying the bioavailability of estrogen [9]. An increase in serum testosterone levels could lead to a decrease in the percent of estradiol bound to sex hormone-binding globulin (SHBG), thereby increasing risk [10,15]. On the other hand, experimental evidence indicates that testosterone has an inhibitory influence on the mitogenic and cancer-promoting effects of estrogen in breast cells and enhances apoptosis via the androgen receptor [11,12]. These findings are further supported by experiments in animals, including primates, that demonstrate that testosterone down-regulates cell proliferation at the molecular level [16,17]. Two recent prospective epidemiologic studies [6,7] have reported positive associations of combined estrogen plus testosterone supplementation (E + T) with risk of breast cancer. Tamimi et al. [7] found that the risk of breast cancer was 1.8-fold greater among current users of estrogen plus testosterone therapies than among never users of postmenopausal hormones, and a previous analysis of the Women’s Health Initiative (WHI) [6] reported a non-significant positive association based on 35 exposed cases. In contrast, two prospective studies involving other testosterone preparations (implants and patch) showed no increase in risk [8,14]. Given the dearth of prospective studies, we reexamined the association of CEE + MT supplementation and breast cancer risk in the WHI observational study with an additional 5.3 years of followup and roughly five times as many incident breast cancer cases and a total of 75 exposed cases.

2. Methods The Women’s Health Initiative (WHI) is a large, multi-center, multi-pronged prospective study designed to advance our understanding of the determinants of major chronic diseases in older women [18]. It is composed of a clinical trial component (CT) and an observational study (OS). Women between the ages of 50 and 79 and representing major racial/ethnic groups were recruited from the general population at 40 clinical centers throughout the US between 1993 and 1998. Details of the design and reliability of the baseline measures have been published [18,19]. Self-administered questionnaires, completed at study entry, were used to collect information on demographics, medical, reproductive and family history, and on dietary and lifestyle factors, including smoking history, alcohol consumption, and recreational physical activity. Three questions regarding the use of CEE + MT were asked of OS participants in the Year 3 follow-up: (1) “In the past 2 years, did you use female hormone PILLS prescribed by a doctor which contained both ESTROGEN and TESTOSTERONE COMBINED in the same pill?”; (2) “In the past 2 years, how many months did you use COMBINED female hormone pills which contained both ESTROGEN and TESTOSTERONE?”; (3) “In the past 2 years, what type of COMBINED ESTROGEN and TESTOSTERONE pill did you use the longest?” In succeeding annual questionnaires (years 4–8), the same questions were asked with reference to the previous year. Questions about use of other hormones (estrogen and progesterone) distinguished between products containing estrogen alone (E), progesterone alone (P), or both (E + P).

71

At baseline (Year 0), current use of medication was assessed through direct review of participants’ medication bottles. At Year 3, women in the WHI OS were asked to record all current medications and were asked to bring in any new medications at the Year 3 clinical visit. Information recorded included the product name, mode of use (transdermal patch, injection, pill), and the dose. We examined all responses containing the words “androgens,” “testosterone,” or “estrogen plus testosterone” in the current medications questionnaire to determine to what extent other testosterone preparations were used. 2.1. Ascertainment of breast cancer New breast cancer diagnoses were updated annually in the OS using in-person, mailed, or telephone questionnaires. Self-reports of a breast cancer diagnosis were verified by centralized review of medical records and pathology reports by trained physician adjudicators [20]. 2.2. Analytic sample Of the 93,676 women enrolled in the OS at baseline, 82,560 completed the Year 3 questionnaire (Fig. 1). We excluded respondents with a history of breast cancer prior to Year 0 (N = 5303), those who reported a breast cancer diagnosis between Year 0 and Year 3 (N = 1297), and those with missing information on hormone use at Year 3 (N = 3996), leaving 71,964 women in the analysis. We proceeded to create two datasets. Dataset #1 (“CEE + MT use” – whether or not other hormones were used) included all women with information on use (yes, no) of hormones at Year 3 (N = 71,964). In this dataset we compared users of CEE + MT, regardless of whether or not they had used other hormones (N = 1714: 75 cases and 1639 non-cases) to women with no hormone use or who only used hormones other than CEE + MT (i.e., E or E + P) (N = 70,250: 2757 cases and 67,493 non-cases). A second dataset (dataset #2 – “exclusive CEE + MT use”) was created by further restricting dataset #1 to women who used CEE + MT only or who reported no hormone use at Year 3 (N = 30,889). In this dataset we compared exclusive users of CEE + MT reported at Year 3 (N = 497: 22 cases and 475 non-cases) to women with no reported hormone use at Year 3 (N = 30,392: 1041 cases and 29,351 non-cases). Over a mean followup of 10 years, as of December 11, 2012, 2832 incident cases of invasive breast cancer were identified in dataset #1 and 1063 cases were identified in dataset #2. 2.3. Statistical analysis We used Cox proportional hazards models to assess the association of use of CEE + MT supplementation at the Year 3 visit and subsequent risk of breast cancer in both datasets. Women who did not develop breast cancer during follow-up were censored at death, cancer diagnosis (other than the index cancer), or end-of-followup, whichever occurred earliest. Measures of exposure included: CEE + MT supplementation in the past 2 years (yes, no); duration of use (no. of months); and type of CEE + MT preparation (Estratest, Estratest HS, other). We computed age-adjusted hazard ratios and multivariable-adjusted HRs. Owing to the large number of covariates and the relatively small number of exposed cases, we dichotomized categorical covariates and used continuous variables wherever possible to maximize statistical power. We used a number of alternative models, adjusting for a range of breast cancer risk factors. These gave similar results, and we present the results of the model including the following covariates: age (continuous), family history of breast cancer in a first degree relative (yes, no), previous breast biopsy (yes, no), number of mammograms in past 5 years (≥4, <4), pack-years of smoking (continuous), and body mass index

72

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

WHI Observaonal Study N = 93,676

Completed Year 3 quesonnaire on hormone use N = 82,560

10,596 excluded 5,303 hx breast cancer prior to Yr 0 1,297 breast cancer before Yr 3 3,996 missing informaon about hormone use at Yr 3 Dataset #1* N = 71,964 (2,832 cases; 69,132 non-cases)

Dataset #2** N = 30,889 (1,063cases; 29,826 non-cases)

*Those who responded to CEE + MT quesons, whether or not they reported other hormone use at Yr 3. **CEE + MT only or no hormone use reported at Yr 3. Fig. 1. Flowchart showing WHI observational study participants used in analysis of CEE + MT and breast cancer.

(kg/m2 – continuous). Since dataset #1 included users of other hormone therapy in addition to CEE + MT, use of E (yes, no) and use of E + P (yes, no) at Year 3 were additionally included in the model. The same model was used in the subset of women with a natural menopause. Additional adjustment for prior hormone use at Year 0 (none, E alone, E + P, both), educational level (less than high school grad, high school grad/some college, college grad, post college), physical activity (hours per week), alcohol intake (servings per day), and ethnicity (white, black, other) did not materially affect the results. With the exception of BMI, for which we used repeated measures (starting with BMI at Year 3), covariates were obtained from the baseline questionnaire (Year 0). In addition to using only the Year 3 exposure data, in a second stage of the analysis we analyzed the repeated CEE + MT use measurements as time-dependent covariates in Cox proportional hazards models to account for fluctuations in the measurements over time [21]. In this analysis in dataset #1 there were 3360 women who reported using CEE + MT on at least one questionnaire between Year 3 and Year 8, in contrast to 1714 users at Year 3. In dataset #2 the corresponding numbers were 632 and 497. The relevant time-dependent covariate for subjects at risk at time t was a function of responses obtained only until the time of diagnosis in the index case or censoring of the non-cases. Responses obtained within 1 year prior to diagnosis or censoring were excluded from all analyses. The same covariates as used in the Year 3 analyses were included in the repeated measures analysis. Time-varying covariates included CEE + MT use and BMI, and, in dataset #1, other hormone use (estrogen alone; estrogen plus progestin). Other covariates were those reported at Year 0. 3. Results Based on reporting of current medications at baseline and at Year 3, use of testosterone preparations other than CEE + MT in

this population was virtually nil. Based on the questions regarding hormone use at years 3–8, the number of women reporting use of CEE + MT in dataset #1 was 1714 (Year 3), 1469 (Year 4), 1312 (Year 5), 1032 (Year 6), 747 (Year 7), and 511 (Year 8). The corresponding numbers in dataset #2 were: 497, 431, 389, 289, 199, and 130. Based on dataset #1, most women who reported use of CEE + MT did so in Year 3 only (N = 688), in Year 4 only (N = 356), in Years 3 and 4 (N = 236), in Year 5 only (N = 307), or in Year 6 only (N = 209). These women accounted for 53% of the 3360 women who used CEE + MT at some time between Year 3 and Year 8. Characteristics of CEE + MT users and non-users (irrespective of other hormone use) and “exclusive users” (datasets 1 and 2, respectively) are compared in Table 1. In both datasets, compared to non-users of CEE + MT, users of CEE + MT were younger, leaner, had lower parity, higher mean alcohol intake, fewer pack-years of smoking, and a higher level of physical activity. In addition, CEE + MT users tended to have more years of education, higher income, earlier age at menarche, earlier age at first birth, and greater proportions of ever use of oral contraceptives and hormone therapy (any type). The largest differences between CEE + MT users and non-users were seen for past hormone use, mammography screening, prevalence of hysterectomy, and prevalence of natural menopause. These differences were more extreme among exclusive CEE + MT users vs. non-hormone users at Year 3. Among this group, 91.4% (454/497) of CEE + MT exclusive users had used hormone therapy in the past (reported at Year 0) vs. 27.2% (8277/30,392) of non-hormone users at Year 3. Among the non-exclusive users of CEE + MT, 23.3% (399/1714) of CEE + MT users did not use any other type of hormone at Year 3 vs. 49.5% (34,774/70,250) of nonCEE + MT users. The prevalence of CEE + MT use among women who completed the Year 3 questionnaire was 2.4% (1714/70,250) in dataset #1 and 1.6% (497/30,392) in dataset #2.

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

73

Table 1 Baseline characteristics by CEE + MT use in the Women’s Health Initiative observational study. Characteristic

Mean age as of Year 3 Mean BMI (kg/m2 ) Parity Mean alcohol servings per week Mean pack-years of smoking Mean MET-h/wka Education, % >16y Ethnicity White, % Black, % Other, % Income ≥$75,000, % Age at menarche, % <12y Age at first birth, %> 30y Oral contraceptive use, % ever Hormone therapy (Year 0), % ever Type of past hormone therapy, (Year 0) % None E alone E+P Both Other hormone use at yr 3 None (other than CEE + MT) E P E+P Other combinations Current smokers, % Mammography screeningb Hysterectomy (% yes) Bilateral oophorectomy (% yes) Natural menopausec

Dataset #2 Exclusive CEE + MT use at Year 3

Dataset #1 CEE + MT use at Year 3 Yes (N = 1714)

No (N = 70,250)

P-value

Yes (N = 497)

No hormone use (N = 30,392)

P-value

62.2 26.8 2.3 2.9 8.7 16.1 34.4

66.2 27.3 2.5 2.6 9.3 13.9 31.6

<0.0001 0.0004 0.01 0.01 0.14 <0.0001 0.006 0.03

61.2 26.4 2.4 3.1 8.9 16.9 35.4

67.9 28.1 2.5 2.3 9.5 13.0 27.7

<0.0001 <0.0001 0.04 0.001 0.43 <0.0001 <0.0001 <0.0001

84.4 6.1 9.6 34.5 23.0 8.3 56.0 88.6

85.5 6.9 7.6 23.4 22.0 9.5 40.8 60.4

91.4 4.3 4.3 36.9 22.3 7.1 59.1 90.8

82.4 9.8 7.9 17.9 21.7 10.5 32.3 27.2

11.4 39.6 37.4 11.7

39.3 30.5 23.5 6.3

9.2 44.4 30.8 15.6

72.8 16.1 9.3 1.8

23.3 22.6 9.8 13.6 30.7 6.1 73.5 47.8 26.3 51.5

49.5 25.3 1.0 9.4 15.3 5.7 63.2 40.2 19.7 59.6

–

–

– – 6.9 73.4 58.5 34.3 40.9

– – 6.7 51.2 28.3 12.4 71.4

<0.0001 0.01 <0.0001 <0.0001 <0.0001 <0.0001

<0.0001 0.04 0.0005 <0.0001 <0.0001 <0.0001

<0.0001

0.08 <0.0001 <0.0001 <0.0001 <0.0001

0.02 <0.0001 <0.0001 <0.0001 <0.0001

a MET – metabolic equivalent tasks – defined as caloric need per kilogram of body weight per hour of activity divided by the caloric need per kilogram of body weight per hour of rest, per hour per week. b Had 4 or more mammograms vs. <4 within past 5 years. c Did not have a hysterectomy or bilateral oophorectomy.

3.1. Dataset #1

4. Discussion

In dataset #1, ever use of CEE + MT at Year 3 was not associated with breast cancer risk (adjusted HR 1.06, 95% CI 0.82–1.36) (Table 2). Duration of CEE + MT use and type also were not associated with risk. Among women who had a natural menopause, the HR for CEE + MT use was 1.15, 95% CI 0.84–1.59. This analysis included 883 breast cancer cases (46 exposed, 837 not exposed) and 41,839 non-cases (1753 exposed, 40,086 not exposed).

The present study is the largest cohort study to date to examine the use of CEE + MT supplementation in relation to breast cancer risk. In the “CEE + MT use” analysis (i.e., whether or not other hormones were used) (dataset #1), compared to women who had not used CEE + MT over the previous 2 years, CEE + MT use was not associated with increased risk, either before or after adjustment for other hormone use at Year 3 as well as for previous hormone use. In the “exclusive CEE + MT” analysis, CEE + MT use at Year 3 also was not associated with risk. Results of the repeated measures analysis were consistent with those of the baseline analysis. Two recent cohort studies have reported a positive association between estrogen and testosterone (E + T) use and breast cancer. In an analysis of E + T use in the Nurses’ Health Study from 1978 to 2002, Tamimi et al. [7] identified 4610 cases of invasive breast cancer among postmenopausal women. Compared to never users of hormone therapy, the relative risk (RR) of breast cancer among women who were current users of E + T (based on 29 exposed cases) was 1.77, 95% CI 1.22–2.56. Among women with a natural menopause, the RR for breast cancer among E + T users was 2.48, 95% CI 1.53–4.04 relative to never users of hormone therapy. This estimate was based on 17 exposed cases. Ness et al. [6] analyzed CEE + MT supplementation (ascertained at the Year 3 visit) in relation to breast cancer risk in the observational study of the Women’s Health Initiative. A total of 31,842 women in the OS were followed for a mean of 4.7 years. Thirty-five women using CEE + MT developed invasive breast cancer: adjusted

3.2. Dataset #2 In the analysis of exclusive CEE + MT use at Year 3, the HR associated with ever use was not statistically significant: 1.22, 95% CI 0.78–1.92. Duration of use and type also showed no association. The HR for exclusive CEE + MT use among women who had a natural menopause was 1.32, 95% CI 0.68–2.55. This analysis was based on 202 breast cancer cases (11 exposed, 191 not exposed) and 21,692 non-cases (773 exposed, 20,919 not exposed). 3.3. Time-dependent covariate analysis In the time-dependent covariate analysis using repeated measures of CEE + MT use between Year 3 and Year 8, in dataset #1, the adjusted HR for CEE + MT use from Year 3 to Year 8 was 1.06, 95% CI 0.74–1.54. In dataset #2, the adjusted HR for exclusive CEE + MT use from Year 3 to Year 8 was 1.10, 95% CI 0.57–2.12.

74

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

Table 2 Hazard ratios (HR) and 95% confidence intervals (95% CI) for the association of CEE + MT use with risk of breast cancer in the Women’s Health Initiative observational study. CEE + MT exposure

Dataset #1 CEE + MT usea

Dataset #2 Exclusive CEE + MT use

Cases

Non-cases

Age-adjusted

MV-adjustedb

Cases

Non-cases

Age-adjusted

MV-adjustedb

29,826

HR

95% CI

HR

95% CI

2832

69,132

HR

95% CI

HR

95% CI

1063

Used CEE + MT No Yes Missing

2757 75 0

67,493 1639 35

1.00 1.11

Ref. (0.88–1.39)

1.00 1.06

Ref. (0.82–1.36)

1041 22

29,351 475

1.00 1.23

Ref. (0.80–1.89)

1.00 1.22

Ref. (0.78–1.92)

Duration of CEE + MT Never <7 mo 7–12 mo >12 mo Missing

2757 19 16 40 0

67,493 440 276 888 35

1.00 1.04 1.40 1.09

Ref. (0.66–1.64) (0.85–2.28) (0.80–1.49)

1.00 0.92 1.51 1.04

Ref. (0.56–1.51) (0.92–2.47) (0.74–1.46)

1041 2 1 19 0

29,351 41 48 382 4

1.00 1.31 0.64 1.30

Ref. (0.33–5.27) (0.09–4.54) (0.82–2.05)

1.00 1.29 0.73 1.33

Ref. (0.32–5.16) (0.10–5.18) (0.83–2.13)

Type of CEE + MT None Estratest Estratest HS Other Missing

2757 35 19 16 5

67,493 699 425 421 94

1.00 1.19 1.03 0.97

Ref. (0.85–1.67) (0.66–1.63) (0.60–1.59)

1.00 1.14 1.10 0.90

Ref. (0.79–1.64) (0.69–1.75) (0.54–1.51)

1041 11 8 2 1

29,351 244 203 24 4

1.00 1.20 1.03 2.59

Ref. (0.66–2.17) (0.51–2.08) (0.65–10.34)

1.00 1.13 1.07 0.99

Ref. (0.80–1.61) (0.68–1.68) (0.59–1.64)

1799

40,923

784

21,110

1753 46 0

40,086 837 0

773 11 0

20,919 191 0

1.00 1.46

Ref. 0.80–2.65

1.00 1.32

Ref. 0.68–2.55

Women with a natural menopause Used CEE + MT No Yes Missing

1.00 1.25

Ref. 0.93–1.67

1.00 1.15

Ref. 0.84–1.59

a

Whether or not other hormones were used. Multivariable-adjusted HR: Adjusted for age (continuous), family history of breast cancer in a first degree relative (yes, no), previous breast biopsy (yes, no), mammograms in past 5 years (≥4 vs. <4), pack-years of smoking (continuous), body mass index (kg/m2 – continuous). Estrogen use and estrogen plus progestin use at Year 3 were additionally included in the model for dataset #1. b

HR 1.42, 95% CI 0.95–2.11. The most commonly used CEE + MT preparation, Estratest, was associated with a statistically significant elevation in risk of invasive breast cancer: adjusted HR 1.78, 95% CI 1.05–3.01. However, only shorter duration (<1–12 mo) of CEE + MT use showed a statistically significant association with risk: HR 1.90, 95% CI 1.05–3.01, whereas longer-term use (13–24 mo) showed no association: HR 1.09, 95% CI 0.61–1.93. It appears that in both cohort studies described above [6,7] a large proportion of current E + T users had used other hormone preparations before switching to E + T, whereas the comparison group was comprised of those who had never used any type of hormone preparations. Tamimi et al. [7] state that, “Women reporting current use of E&T therapy in 1998 were mostly past users of other types of hormones including testosterone only (54.6%), estrogen only (28.8%), E&P (19.1%), progesterone only (2.1%) and other types of hormone (7.5%). Only 2.4% of current E&T users in 1998 were never users prior to initiating E&T use” [7]. This is also true of the Ness et al. [7] study in which other hormone users were excluded from the “no hormone use” group but not from the CEE + MT users. Thus, it appears that an effect of CEE + MT in these studies may be confounded by the use of other hormone preparations. In addition to the cohort studies mentioned above, two casecontrol studies [13,22] and 3 other cohort studies [8,14,23] (Ref. [23] is an early analysis of the Nurses’ Health Study) have examined the association of testosterone supplementation and breast cancer risk, with conflicting results. Studies to date have had a number of limitations, as pointed out by Bitzer [24], including small numbers of exposed cases, particularly in the cohort studies (4 [23], 12 [14], 7 [8], 29 [7], and 35 [6]) and limited information on dose and type of hormone preparation. Additionally, two studies lacked a nonexposed group and compared breast cancer incidence in cohort participants to incidence reported in the literature or in the general

population [8,14]. Because some of these studies focused on use of the testosterone patch [14] or other hormone preparations [8,13], their results may not be relevant to the effects of combined oral esterified estrogen plus methyltestosterone studied here. In spite of the large sample size in the present study, only 2.4% of respondents to the Year 3 questionnaire had used CEE + MT in the previous 2 years. This is similar to the prevalence of use in the Nurses’ Health Study [7] and in the Ness et al. analysis [6]. With 10 years of follow-up of the WHI OS cohort starting in Year 3, we had nearly 5 times the number of invasive breast cancers included in the Ness et al. analysis (2832 vs. 593). However, as pointed out above, over 75% of CEE + MT users at Year 3 had used other hormone preparations in the past two years. When we restricted our analysis to exclusive users of CEE + MT compared to non-hormone users, only 22 out of 1063 breast cancer cases were exclusive users. Therefore, we had limited power to assess the association of exclusive CEE + MT exposure with breast cancer. In addition, the problem of potential confounding by past hormone use is particularly challenging in our study and in similar studies [6,7]. Although the evidence from observational studies regarding the association of CEE + MT supplementation and breast cancer is rudimentary at present, experimental studies do not provide a strong rationale for testosterone acting as a breast carcinogen. Indeed, there is some evidence to suggest that testosterone might have protective effects. Specifically, in in vitro studies, testosterone administration exhibits growth-inhibitory and apoptotic effects in some breast cancer cell lines [11]. In primate studies, testosterone reduces estrogen-induced breast epithelial proliferation [16,17] and abolishes estrogen-induced gene expression [17]. A recent review concluded that the majority of in vitro and primate studies support testosterone’s inhibitory effects on mammary epithelial proliferation [12]. Additionally, clinical studies

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

suggest that testosterone administration may counteract mammary epithelial cell proliferation induced by estrogen/progestogen therapy in postmenopausal women [25–27]. Nevertheless, there is some evidence to suggest that androgens may have both pro- and anti-carcinogenic effects. Specifically, androgens may have different effects in ER+ versus ER− tumors, inhibiting cell proliferation in the former [28], but promoting cell proliferation in the latter [29,30]. Strengths of the present study include its prospective nature, detailed information on past use of exogenous hormones and use during follow-up, use of repeated measurement data on CEE + MT, and the use of two different analytic approaches, which yielded consistent results. In addition, data from a separate questionnaire regarding all current medications (administered at Year 0 and Year 3) indicated that use of other, off-label testosterone preparations was insignificant in this population. Our study has several limitations. Although the number of exposed cases was larger compared to previous cohort studies, the statistical power to adjust for many potential confounding variables (and, in particular, concurrent and prior use of other hormone preparations) in the Cox models was still limited. The number of breast cancer cases ascertained among exclusive CEE + MT users was small (N = 22). However, in analyses of both datasets, all covariates included in the model were highly statistically significant, indicating that we had the power to detect moderate-to-strong associations of risk factors that are prevalent in the study population. The small number of cases precluded examination of the association by estrogen receptor status. Finally, CEE + MT use was based on self-report, rather than on direct examination by study personnel of the label on the pill bottles that was used to validate exposure to E and E + P in the Year 0 assessment of hormone therapy use in WHI [31]. Given the increasing use of low-dose E + T to treat reduced libido in postmenopausal women, and the possibility that the balance between androgens and estrogens, and their metabolites, may play a role in breast carcinogenesis, study of the association of CEE + MT with cancer risk is an important area of research. However, due to the low prevalence of CEE + MT supplementation in published reports from cohort studies and the problem of concurrent and prior use of other hormone preparations, all studies to date have had severely limited power to address this question. Adequate assessment of the long-term association of CEE + MT therapy with risk of breast cancer will require pooling of data from large prospective studies and continued follow-up in those studies. This is particularly the case with respect to ensuring adequate statistical power to examine the association of CEE + MT with different types of breast cancer. 5. Conclusion Our results, from the largest study carried out to date, add to the sparse available evidence regarding the long-term effects of CEE + MT supplementation on breast cancer risk in postmenopausal women. In two separate analyses we were unable to demonstrate a statistically significant association of CEE + MT use with breast cancer. Authors contributors Conception and design: G.C. Kabat, T.E. Rohan; development and methodology: G.C. Kabat, V. Kamensky, T.E. Rohan; acquisition of data: G.C. Kabat, D.S. Lane, J. Wactawski-Wende, T.E. Rohan; analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): G.C. Kabat, V. Kamensky, M. Heo, J.W. Bea, L. Hou, D.S. Lane, S. Liu, L.-H. Qi, M.S. Simon, J. Wactawski-Wende,

75

T.E. Rohan; writing, review, and/or revision of the manuscript: G.C. Kabat, V. Kamensky, M. Heo, J.W. Bea, L. Hou, D.S. Lane, S. Liu, L.-H. Qi, M.S. Simon, J. Wactawski-Wende, T.E. Rohan; administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): G.C. Kabat, V. Kamensky; study supervision: G.C. Kabat, T.E. Rohan. Competing interests No potential conflicts of interest were disclosed. Funding This work was supported with institutional funds from the Albert Einstein College of Medicine. T.E. Rohan is supported in part by a grant from the Breast Cancer Research Foundation. Ethical approval IRB approval was obtained in the original WHI study. Acknowledgments The authors wish to thank the following key investigators from the WHI: Program Office: (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Jacques Rossouw, Shari Ludlam, Dale Burwen, Joan McGowan, Leslie Ford, Nancy Geller. Clinical Coordinating Center: (Fred Hutchinson Cancer Research Center, Seattle, WA) Garnet Anderson, Ross Prentice, Andrea LaCroix, Charles Kooperberg, Barbara Cochrane, Julie Hunt, Marian Neuhouser, Lesley Tinker, Susan Heckbert, Alex Reiner. Regional Centers: (Brigham and Women’s Hospital, Harvard Medical School, Boston, MA) JoAnn E. Manson, Kathryn M. Rexrode, Brian Walsh, J. Michael Gaziano, Maria Bueche; (MedStar Health Research Institute/Howard University, Washington, DC) Barbara V. Howard, Lucile Adams-Campbell, Lawrence Lessin, Cheryl Iglesia, Brian Walitt, Amy Park; (The Ohio State University, Columbus, OH) Rebecca Jackson, Randall Harris, Electra Paskett, W. Jerry Mysiw, Michael Blumenfeld; (Stanford Prevention Research Center, Stanford, CA) Marcia L. Stefanick, Mark A. Hlatky, Manisha Desai, Jean Tang, Stacy T. Sims; (University of Arizona, Tucson/Phoenix, AZ) Cynthia A. Thomson, Tamsen Bassford, Cheryl Ritenbaugh, Zhao Chen, Marcia Ko; (University at Buffalo, Buffalo, NY) Jean Wactawski-Wende, Maurizio Trevisan, Ellen Smit, Amy Millen, Michael LaMonte; (University of Florida, Gainesville/Jacksonville, FL) Marian Limacher, Michael Perri, Andrew Kaunitz, R. Stan Williams, Yvonne Brinson; (University of Iowa, Iowa City/Davenport, IA) Robert Wallace, James Torner, Susan Johnson, Linda Snetselaar, Jennifer Robinson; (University of Pittsburgh, Pittsburgh, PA) Lewis Kuller, Jane Cauley, N. Carole Milas; (University of Tennessee Health Science Center, Memphis, TN) Karen C. Johnson, Suzanne Satterfield, Rongling Li, Stephanie Connelly, Fran Tylavsky; (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker, Stephen Rapp, Claudine Legault, Mark Espeland, Laura Coker, Michelle Naughton. Women’s Health Initiative Memory Study: (Wake Forest University School of Medicine, Winston-Salem, NC) Sally Shumaker, Stephen Rapp, Claudine Legault, Mark Espeland, Laura Coker, Michelle Naughton. Former Principal Investigators and Project Officers: (Albert Einstein College of Medicine, Bronx, NY) Sylvia Wassertheil-Smoller (Baylor College of Medicine, Houston, TX) Haleh Sangi-Haghpeykar,

76

G.C. Kabat et al. / Maturitas 79 (2014) 70–76

Aleksandar Rajkovic, Jennifer Hays, John Foreyt; (Brown University, Providence, RI) Charles B. Eaton, Annlouise R. Assaf; (Emory University, Atlanta, GA) Lawrence S. Phillips, Nelson Watts, Sally McNagny, Dallas Hall; (Fred Hutchinson Cancer Research Center, Seattle, WA) Shirley A.A. Beresford, Maureen Henderson; (George Washington University, Washington, DC) Lisa Martin, Judith Hsia, Valery Miller; (Harbor-UCLA Research and Education Institute, Torrance, CA) Rowan Chlebowski (Kaiser Permanente Center for Health Research, Portland, OR) Erin LeBlanc, Yvonne Michael, Evelyn Whitlock, Cheryl Ritenbaugh, Barbara Valanis; (Kaiser Permanente Division of Research, Oakland, CA) Bette Caan, Robert Hiatt; (National Cancer Institute, Bethesda, MD) Carolyn Clifford1 ; (Medical College of Wisconsin, Milwaukee, WI) Jane Morley Kotchen; (National Heart, Lung, and Blood Institute, Bethesda, Maryland) Linda Pottern; (Northwestern University, Chicago/Evanston, IL) Linda Van Horn, Philip Greenland; (Rush University Medical Center, Chicago, IL) Lynda Powell, William Elliott, Henry Black; (State University of New York at Stony Brook, Stony Brook, NY) Dorothy Lane, Iris Granek; (University at Buffalo, Buffalo, NY) Maurizio Trevisan; (University of Alabama at Birmingham, Birmingham, AL) Cora E. Lewis, Albert Oberman; (University of Arizona, Tucson/Phoenix, AZ) Tamsen Bassford, Cheryl Ritenbaugh, Tom Moon; (University of California at Davis, Sacramento, CA) John Robbins; (University of California at Irvine, CA) F. Allan Hubbell, Frank Meyskens, Jr.; (University of California at Los Angeles, CA) Lauren Nathan, Howard Judd1 ; (University of California at San Diego, LaJolla/Chula Vista, CA) Robert D. Langer; (University of Cincinnati, Cincinnati, OH) Michael Thomas, Margery Gass, James Liu; (University of Hawaii, Honolulu, HI) J. David Curb1 ; (University of Massachusetts/Fallon Clinic, Worcester, MA) Judith Ockene; (University of Medicine and Dentistry of New Jersey, Newark, NJ) Norman Lasser; (University of Miami, Miami, FL) Mary Jo O’Sullivan, Marianna Baum; (University of Minnesota, Minneapolis, MN) Karen L. Margolis, Richard Grimm; (University of Nevada, Reno, NV) Robert Brunner, Sandra Daugherty1 ; (University of North Carolina, Chapel Hill, NC) Gerardo Heiss, Barbara Hulka, David Sheps; (University of Tennessee Health Science Center, Memphis, TN) Karen Johnson, William Applegate; (University of Texas Health Science Center, San Antonio, TX) Robert Brzyski, Robert Schenken; (University of Wisconsin, Madison, WI) Gloria E. Sarto, Catherine Allen1 ; (Wake Forest University School of Medicine, Winston-Salem, NC) Mara Vitolins, Denise Bonds, Electra Paskett, Greg Burke; (Wayne State University School of Medicine/Karmanos Cancer Institute, Detroit, MI) Michael S. Simon, Susan Hendrix. References [1] Snabes MC, Simes SM. Approved hormonal treatments for HSDD: an unmet medical need (Commentary). J Sex Med 2009;6:1846–9. [2] Phillips EH, Ryan S, Ferrari R, Green C. Estratest and Estratest HS (esterified estrogens and methyltestosterone) therapy: a summary of safety surveillance data January 1989 to August 2002. Clin Ther 2003;25: 3027–43. [3] Margo K, Winn R. Testosterone treatments: why, when, and how? Am Fam Physician 2006;73:1591–8. [4] Cameron Braunstein GD. Androgen replacement therapy in women. Fertil Steril 2004;82:273–89.

1

Deceased.

[5] Basaria S, Dobs A. Controversies regarding transdermal androgen therapy in postmenopausal women. J Clin Endocrinol Metab 2006;91:4743–52. [6] Ness RB, Albano JD, McTiernan A, Cauley JA. Influence of estrogen plus testosterone supplementation on breast cancer. Arch Intern Med 2009;169:41–6. [7] Tamimi RM, Hankinson SE, Chen WY, Rosner B, Colditz G. Combined estrogen and testosterone use and risk of breast cancer in postmenopausal women. Arch Intern Med 2006;166:1483–9. [8] Dimitrakakis C, Jones RA, Liu A, Bondy CA. Breast cancer incidence in postmenopausal women using testosterone in addition to the usual hormone therapy. Menopause 2004;11:531–5. [9] Dimitrakakis C. Androgens and breast cancer in men and women. Endocrinol Metab Clin North Am 2011;40:533–47. [10] Simon JA. Safety of estrogen/androgen regimens. J Reprod Med 2001;46:281–90. [11] Somboonporn W, Davis SR. Testosterone effects on the breast: implications for testosterone therapy for women. Endocr Rev 2004;25:374–88. [12] Shufelt CL, Braunstein GD. Testosterone and the breast. Menopause Int 2008;14:117–22. [13] Ewertz M. Influence of non-contraceptive exogenous and endogenous sex hormones on breast cancer risk in Denmark. Int J Cancer 1988;42:832–8. [14] Davis SR, Moreau M, Kroll R, et al. Testosterone for low libido in postmenopausal women not taking estrogen. New Engl J Med 2008;359:2005–17. [15] Lobo RA, Rosen RC, Yang H-M, Block B, van der Hoop RG. Comparative effects of oral esterified estrogens with and without methyltestosterone on endocrine profiles and dimensions of sexual function in postmenopausal women with hypoactive sexual desire. Fertil Steril 2003;79:1341–52. [16] Dimitrakakis C, Zhou J, Wang J, et al. A physiologic role for testosterone in limiting estrogenic stimulation of the breast. Menopause 2003;10:292–8. [17] Zhou J, Ng S, Adesanya-Famuiya O, Anderson K, Bondy CA. Testosterone inhibits estrogen-induced mammary epithelial proliferation and suppresses estrogen receptor expression. FASEB J 2000;14:1725–30. [18] The Women’s Health Initiative Study Group. Design of the Women’s Health Initiative clinical trial and observational study. Control Clin Trials 1998;19:61–109. [19] Langer RD, White E, Lewis CE, et al. The Women’s Health Initiative observational study: baseline characteristics of participants and reliability of baseline measures. Ann Epidemiol 2003;13(9 Suppl.):S107–21. [20] Curb JD, McTiernan A, Heckbert SR, et al. Outcomes ascertainment and adjudication methods in the Women’s Health Initiative. Ann Epidemiol 2003;13(9 Suppl.):S122–8. [21] Gail MH. Evaluating serial cancer marker studies in patients at risk of recurrent disease. Biometrics 1981;37:67–78. [22] Brinton LA, Hoover R, Fraumeni Jr JF. Menopausal oestrogens and breast cancer risk: an expanded case–control study. Br J Cancer 1986;54:825–32. [23] Colditz GA, Hankinson SE, Hunter DJ, et al. The use of estrogens and progestin and the risk of breast cancer in postmenopausal women. N Engl J Med 1995;332:1589–93. [24] Bitzer J, Kenemans P, Mueck AO. Breast cancer risk in postmenopausal women using testosterone in combination with hormone replacement therapy. Maturitas 2008;59:209–18. [25] Somboonporn W, Davis SR. Postmenopausal testosterone therapy and breast cancer risk. Maturitas 2004;49:267–75. [26] Hofling M, Hirschberg AL, Skoog L, Tani E, Hägerström T, von Schoultz B. Testosterone inhibits estrogen/progestogen-induced cell proliferation in postmenopausal women. Menopause 2007;14:183–90. [27] Hofling M, Löfgren L, von Schoultz B, Carlström K, Söderqvist G. Associations between serum testosterone levels, cell proliferation and progesterone receptor content in normal and malignant breast tissue in postmenopausal women. Gynecol Endocrinol 2008;24:405–10. [28] Peters AA, Buchanan G, Ricciardelli C, et al. Androgen receptor inhibits estrogen receptor-alpha activity and is prognostic of breast cancer. Cancer Res 2009;69:6131–40. [29] Hickey TE, Robinson JL, Carroll JS, Tilley WD. Minireview: the androgen receptor in breast tissues: growth inhibitor, tumor suppressor, oncogene? Mol Endocrinol 2012;26:1252–67. [30] Doane AS, Danso M, Lal P, et al. An estrogen receptor-negative breast cancer subset characterized by hormonally regulated transcriptional program and response to androgen. Oncogene 2006;25:3994–4008. [31] Kabat GC, Kim MY, Luo J, et al. Menstrual and reproductive factors and endogenous hormone use and risk of transitional cell bladder cancer in postmenopausal women. Eur J Cancer Prev 2013;22:409–16.