© The Author 2006. Published by Oxford University Press.
ARTICLE |
Menopausal Hormone Therapy and Ovarian Cancer Risk in the National Institutes of Health–AARP Diet and Health Study Cohort
Affiliations of authors: Hormonal and Reproductive Epidemiology Branch (JVL, LAB), Nutritional Epidemiology Branch (MFL, TM, AS), and Epidemiology and Biostatistics Program (PH), Division of Cancer Epidemiology and Genetics, National Cancer Institute, National Institutes of Health, Rockville, MD; AARP, Washington DC (AH)
Correspondence to: James V. Lacey Jr., PhD, 6120 Executive Blvd., Room 5030, Rockville, MD 20852-7234 (e-mail: jimlacey{at}nih.gov).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Background: Recent studies offer conflicting data on risks of ovarian cancer in users of menopausal hormone therapy. Some findings of increased risks associated with unopposed estrogen use are based on older studies of women with intact uteri, and small sample size and incomplete exposure information have limited the data on estrogen plus progestin associations. Methods: The National Institutes of Health–AARP Diet and Health Study Cohort included 97 638 women aged 50–71 years at baseline who completed two questionnaires (1995–1996 and 1996–1997). We identified 214 incident ovarian cancers among these women through the year 2000 using data from state cancer registries and mortality indexes. We estimated relative risks (RRs) of ovarian cancer for detailed hormone therapy exposures using multivariable proportional hazards regression models. All statistical tests were two-sided. Results: Use of unopposed estrogen for fewer than 10 years was not associated with ovarian cancer. Compared with use of no hormone therapy, use of unopposed estrogen for 10 or more years was statistically significantly associated with ovarian cancer among all women (RR = 1.89, 95% confidence interval [CI] = 1.22 to 2.95; P = .004; 56 versus 72 ovarian cancers per 100 000 person-years, respectively) and, albeit not statistically significantly, among women with hysterectomy (n = 19 359, RR = 1.70, 95% CI = 0.87 to 3.31; P = .06). Among the 73 483 women with intact uteri, 51 698 had used no hormone therapy or only estrogen plus progestin. Compared with no hormone therapy use, 5 or more years of use of sequential (progestin for <15 days per cycle; RR = 3.09, 95% CI = 1.68 to 5.68; P<.001; 49 versus 108 per 100 000 person-years) or continuous (progestin for
15 days per cycle; RR = 1.82, 95% CI = 1.03 to 3.23; P = .02; 49 versus 66 per 100 000 person-years) estrogen plus progestin regimens were statistically significantly associated with ovarian cancer. Conclusions: Long durations of use of unopposed estrogen and of estrogen plus progestin, especially sequential regimens, are associated with increased ovarian cancer risk. These data expand the range of possible risks associated with menopausal hormone therapy.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Data from early studies (1–4) show no association between menopausal hormone therapy and ovarian cancer, but recent studies (5–11) suggest that long-duration use of unopposed estrogen is associated with increased ovarian cancer risk. Methodologic issues, such as small sample size or residual confounding by oral contraceptives or hysterectomy (12), might contribute to the discrepancy. However, other recent large studies without obvious limitations (13,14) found no associations between ovarian cancer and hormone therapy.
Many of the reports of higher ovarian cancer risks in women who used hormone therapy arose from study populations in which women with intact uteri had used unopposed estrogen (5–7,9–11). This exposure combination, although etiologically and historically intriguing, is of less relevance today because, since the early 1990s, clinical guidelines (15–17) have recommended use of estrogen plus progestin formulations for women with intact uteri and use of unopposed estrogen formulations for women with hysterectomy. Data on ovarian cancer risk associated with use according to those guidelines are sparse. Long-duration unopposed estrogen use among women with hysterectomy was associated with statistically significantly increased ovarian cancer risk in two studies (5,7), statistically nonsignificantly increased risk in another (14), and no increased risk in a fourth (13).
Much of the limited data on exposure to estrogen plus progestin and risk of ovarian cancer came from studies of women who previously used unopposed estrogen (11,14) or from studies that did not evaluate individual estrogen plus progestin regimens (10,13,14). In one Swedish case–control study (7), use of sequential estrogen plus progestin regimens (i.e., daily estrogen and progestin taken for 
15 days per cycle) was associated with increased ovarian cancer risk but use of continuous regimens (i.e., daily estrogen and progestin taken for 19 days per cycle) was not. Menopausal hormone therapies used in Sweden contain different estrogens and progestins than those used in the United States (7). The difference in formulations may be particularly important for ovarian cancer risk (18). The most detailed US data on continuous combined estrogen plus progestin come from the Women's Health Initiative (WHI), a randomized clinical trial in which this regimen was associated with a statistically nonsignificantly increased ovarian cancer risk, based on 32 participants who developed ovarian cancer (19).
The studies to date do not provide clear evidence on ovarian cancer risk associated with the common patterns of menopausal hormone use in the United States since the early 1990s. Studies with sufficient sample size and exposure details to evaluate specific formulations, regimens, and durations of use are needed to address these questions. Toward that end, we analyzed data from the National Institutes of Health (NIH)–AARP Diet and Health Study, a large prospective study of US women.
|
SUBJECTS AND METHODS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Study Population
As previously described (20), the NIH– AARP Diet and Health Study was established in 1995–1996 when a baseline questionnaire (21) eliciting information on demographic characteristics, dietary intake, and numerous health-related behaviors was mailed to 3.5 million AARP members. Recipients included members between 50 and 71 years of age who resided in one of six US states (California, Florida, Louisiana, New Jersey, North Carolina, and Pennsylvania) or two metropolitan areas (Atlanta, GA, and Detroit, MI). A total of 617 119 persons (17.6%) returned the questionnaire, among whom 566 407 persons (16.2%) satisfactorily completed it. In 1996–1997, we sent a second questionnaire to collect additional information on diet, family history of cancer, anthropometry, physical activity, and use of menopausal hormone therapy. A total of 337 076 respondents (59.5%) completed this questionnaire. After excluding participants who died (n = 1619) or moved out of the study area (n = 547) before their completed second questionnaires were received and scanned, proxy respondents to the baseline questionnaire (n = 6959) or second questionnaire (n = 3424), and all 188 118 men, the study population included 136 409 potentially eligible women.
Exposure Ascertainment
The baseline questionnaire asked whether women were currently taking "replacement hormones," and, if so, for how many years. Participants also reported whether they had had a hysterectomy or oophorectomy. Other characteristics, such as demographics, reproductive history, oral contraceptive use, menopausal status, family history of cancer, and smoking, were included in the baseline questionnaire.
The second questionnaire collected detailed data on use of hormone therapy, including ever use of different preparations and formulations. Estrogen or progestin pill users were asked to report dates of first and last use, total duration of use, regimen, usual dose, and name of the pill that they took for the longest period of time. The questionnaire did not ask about the continuous combined estrogen plus progestin pill, which was first marketed in 1995 (22), but instead asked separately about estrogen and progestin use. The second questionnaire did not update hysterectomy or oophorectomy status.
We considered women who reported taking both estrogen and progestin pills to have used only estrogen plus progestin if the reported dates of first use were within 90 days of each other or if the reported durations of use were identical. We created separate exposure categories for estrogen plus progestin use after use of unopposed estrogen (n = 3964; 4.1%) or unopposed progestin (n = 306; 0.3%), estrogen plus progestin use followed by use of unopposed estrogen (n = 1083; 1.1%) or unopposed progestin (n = 191; 0.2%), use of unopposed estrogen followed by use of unopposed progestin (n = 260; 0.3%), and use of unopposed progestin followed by use of unopposed estrogen (n = 347; 0.4%).
Sequential regimens included estrogen plus progestin use for fewer than 15 days per cycle. Continuous estrogen plus progestin regimens included estrogen plus progestin use every day of the cycle. The 2446 women (3.3% of women without hysterectomy) who reported taking progestin for 15–19 or 20–25 days per cycle were categorized as having used the continuous regimen.
Cohort Follow-up
Study follow-up occurred via two annual linkages. For address changes, the cohort was matched to the National Change of Address database (maintained by the US Postal Service) and also updated based on undeliverable mail, other address change update services, and participants' notifications. For vital status, the Social Security Administration Death Master File identified cohort members who are presumed deceased. A follow-up search of the National Death Index Plus provided verification.
Incident Cancers
A probabilistic linkage to eight state cancer registries, using names, address history, sex, date of birth, and Social Security Number provided on the baseline questionnaire, identified incident cancers. All suspected matches underwent review to reject the potential matches that were unlikely to be true (an estimated 4%), and uncertain matches underwent final manual review. An earlier validation study that compared registry findings with self-reports and medical records estimated that linkage validly identified approximately 90% of all incident cancers (23). The North American Association of Central Cancer Registries certifies all eight registries, which are estimated to be 95% complete for cancers reported up to 24 months prior (23). Date and cause of death from the National Death Index linkage also identified fatal cancers. The Special Studies Institutional Review Board of the US National Cancer Institute approved this study. All participants provided written informed consent.
Analytic Population
To use the detailed hormone therapy data, we restricted analyses to the 136 409 women who completed the second questionnaire. We excluded 9039 women who reported a personal history of cancer other than nonmelanoma skin cancer on either questionnaire (including 398 ovarian cancers), 27 602 who reported a bilateral oophorectomy before baseline, and 2118 whose oophorectomy status was unknown at baseline. We also excluded 12 women who developed nonepithelial ovarian cancer during follow-up. Analysis therefore included 97 638 women.
Through December 31, 2000, 214 women developed epithelial ovarian cancer: 155 in women with intact uteri and 59 in women with hysterectomy. Registry data provided diagnosis dates. Ninety-six cancers were serous, 9 were mucinous, 18 were endometrioid, 16 were papillary, 9 were clear cell, 40 were other adenocarcinomas, and 26 were unclassified histologic types.
Statistical Analysis
We used Cox proportional hazards regression (using SAS 8.2 software, SAS Institute, Inc, Cary, NC), with age as the time scale and ties handled by complete enumeration (24), to calculate hazard ratios and estimate the relative risk (RR) of developing ovarian cancer. Tests of the proportional hazards assumptions for exposures and other variables included in statistical models revealed no departures. Follow-up began at the age at which the second questionnaire was received and scanned. Follow-up ended on December 31, 2000 (n = 91 958; 94.2%), or the earliest of the following dates: participant was diagnosed with ovarian cancer (n = 214; 0.2%), moved out of her registry catchment area (n = 3218; 2.2%), or died from any cause (n = 2248; 2.3%).
Most women who use unopposed estrogens today have had hysterectomy. However, older women with intact uteri likely had opportunities to take unopposed estrogen before adding progestins became routine for such women (22). We therefore analyzed unopposed estrogen therapy associations both in the entire cohort and in the 19 359 women with hysterectomy at baseline who never used hormones (n = 6335) or only used unopposed estrogen (n = 13 024). The other 4363 women with hysterectomy at baseline reported use of other or unknown hormone therapy formulations. None of the 433 women with unknown hysterectomy status developed ovarian cancer.
We limited the assessment of estrogen plus progestin associations to the 73 483 women with intact uteri at baseline and further restricted analyses to women who never used hormones (n = 38 005) or only used estrogen plus progestin (n = 20 850). In a priori subgroup analyses, we further excluded estrogen plus progestin–only users whose regimen and dose combinations were unknown (n = 767) or differed from historical prescribing patterns in the United States (16): 1) sequential estrogen plus progestin with a progestin dose of less than 1.0 mg/day or 2.5 mg/day (n = 1881), unknown (n = 1351), or "other" (n = 64); 2) continuous estrogen plus progestin with a progestin dose of less than 1.0 mg/day (n = 431), 10.0 mg/day (n = 227), unknown (n = 1931), or "other" (n = 146); or 3) other regimens (n = 359).
We based our detailed hormone therapy variables on the close-ended response categories from the second questionnaire. For both estrogen and progestin, this questionnaire asked women to report total duration of use in 1-year increments up to 10 years and grouped all use greater than 10 years into a single exposure category. All regression models used women who reported no hormone therapy use as the referent group. The models that evaluated more than one hormone therapy formulation or regimen relied on mutually exclusive exposure categories. To assess combinations of exposures (e.g., recency and duration, regimen and recency, or regimen and duration), we created single variables based on cross-tabulations of the two original variables, collapsing cells with small sample sizes as necessary. The statistical models evaluated each exposure class separately, except for estrogen plus progestin regimen; all regimen-specific relative risks were obtained from models that included terms for both regimens.
We initially evaluated potential confounding by all available factors but ultimately chose a parsimonious combination of variables that were associated with both exposure and outcome and changed the hormone therapy parameter estimates compared with estimates from models adjusted for only age at entry. Our statistical models adjusted for continuous age at entry (years) and race/ethnicity (white versus other/unknown), duration of oral contraceptive use (none, <10 years, 10 years), menopausal status (premenopausal, postmenopausal, or unknown), and body mass index (BMI; kg/m2; <25, 25–29, 30, or unknown). Analyses of the entire cohort also adjusted for hysterectomy and menopausal status (premenopausal, natural menopause, surgical menopause, or unknown). Analyses of unopposed estrogen in women with hysterectomy were not adjusted for menopausal status because only six women with hysterectomy (none of whom developed ovarian cancer) were premenopausal. Analyses of estrogen plus progestin only in women with intact uteri were adjusted for menopausal status (premenopausal, postmenopausal, or unknown). Additional adjustment for calendar time or other factors (e.g., parity or family history of cancer) did not change the results. Two-sided P values were calculated using Wald chi-square tests of categorical (ever use) or ordinal (recency and duration) variables, in which no hormone therapy use was the referent; P<.05 was considered statistically significant.
A woman's current age can influence whether she is prescribed sequential or continuous estrogen plus progestin regimens (16). We therefore conducted sensitivity analyses of associations between ovarian cancer and estrogen plus progestin regimens by stratifying regression models by age group at entry.
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
Characteristics of the Cohort
The 97 638 women accrued 386 468 total person-years. The mean durations of follow-up in women who developed ovarian cancer and women who did not develop ovarian cancer during follow-up were 2.0 years (range = 1 day–4.1 years) and 4.0 years (range = 1 day–4.2 years), respectively. Two women were diagnosed with ovarian cancer and 61 women were censored within 1 month of entry. The mean ± standard deviation ages at entry and exit were 62.5 ± 5.4 years and 66.4 ± 5.4 years, respectively. The standardized incidence ratio for ovarian cancer in the full cohort compared with National Cancer Institute's Surveillance, Epidemiology, and End Results rates (ages 50–79 years) was 0.94 (95% confidence interval [CI] = 0.83 to 1.06).
Most women in the cohort were white, postmenopausal, and in their 60s when they completed the second questionnaire. Women who were overweight (BMI = 25–29 kg/m2) or obese (BMI 30 kg/m2) at baseline contributed just more than half of the total person-time. Overall, ovarian cancer was positively associated with family history of breast cancer; inversely associated with oral contraceptive use, parity, and non-Caucasian race/ethnicity; and not associated with BMI, smoking, age at menarche, or age at natural menopause (Table 1).
|
Compared with women who never used hormone therapy, women who had used hormone therapy were more likely to be white, to be married, to have used oral contraceptives, to have formerly smoked, and to have a BMI below 25 kg/m2; women who had used unopposed estrogen were more likely to be postmenopausal, to have had a hysterectomy, and to have given birth at younger ages; and women who had used estrogen plus progestin were more likely to be younger, to have graduated from college, and to have reported excellent or very good overall health (versus good, fair, or poor health) at baseline. Other ovarian cancer risk factors did not differ by duration of unopposed estrogen use or by estrogen plus progestin regimen (data not shown).
Unopposed Estrogen Therapy
In analyses that included all 97 638 participants, women who ever or currently used unopposed estrogen had statistically nonsignificantly increased ovarian cancer risks compared with women who never used hormone therapy (Table 2). Compared with women who never used hormone therapy, women who had used unopposed estrogen for long durations (10 or more years)—most of whom were also current users—had a statistically significantly increased risk of ovarian cancer (RR = 1.89, 95% CI = 1.22 to 2.95; P = .004; 56 versus 72 ovarian cancers per 100 000 person-years, respectively).
|
Among women with hysterectomy, the associations between ovarian cancer and unopposed estrogen use were slightly attenuated. All 26 long-duration users who developed ovarian cancer had had a hysterectomy. Compared with never use of hormone therapy, the relative risks associated with long-duration and current long-duration unopposed estrogen use were 1.70 (95% CI = 0.87 to 3.31) and 1.71 (95% CI = 0.87 to 3.35), respectively.
Most of the reported unopposed estrogen use was 0.625 mg of Premarin (conjugated equine estrogens) each day of the cycle. Among women who did not report daily estrogen use, 93% reported taking estrogen at least 20 days per cycle. Approximately 4% and 7% of users of unopposed estrogens reported usually taking lower (0.3 mg/day) and higher (1.25 mg/day) doses, respectively.
Estrogen Plus Progestin in Women With Intact Uteri
Compared with women with intact uteri who never used hormone therapy, women with intact uteri who used only estrogen plus progestin had a statistically significantly increased risk of ovarian cancer (RR = 1.50, 95% CI = 1.03 to 2.19; P = .04; Table 3). Risks in women who used unopposed estrogen or unopposed estrogen followed by estrogen plus progestin were elevated, but the increases were not statistically significant. Too few women in these exposure categories developed ovarian cancer to further explore these combinations.
|
Compared with women with intact uteri who never used hormone therapy, women who used estrogen plus progestin for fewer than 10 years did not have an increased risk of ovarian cancer. Women who used estrogen plus progestin for 10 or more years had statistically significantly increased risks. For users of estrogen plus progestin, as for users of unopposed estrogen, former users and short-duration current users did not have an increased risk of ovarian cancer compared with never users.
Compared with women with intact uteri who never used hormone therapy, risks of ovarian cancer were higher for women taking sequential (RR = 1.94, 95% CI = 1.17 to 3.22; P = .01) than continuous (RR = 1.41, 95% CI = 0.90 to 2.22; P = .14) regimens (Table 4). Most of the women who used estrogen plus progestin and developed ovarian cancer were current users at the time of the second questionnaire. The association of ovarian cancer risk with current sequential use was stronger than that with current continuous use. For both regimens, the inconsistently elevated risks for short-duration (<5 years) use were based on few ovarian cancers, and there was no consistent duration response. Compared with no use, long-duration use (5 years) of sequential regimens was associated with statistically significantly increased risk (RR = 1.92, 95% CI = 1.07 to 3.46; P = .02; 49 versus 108 per 100 000 person-years), but long-duration use of continuous regimens was associated with statistically nonsignificantly increased risk (RR = 1.55, 95% CI = 0.97 to 2.87; P = .04; 49 versus 66 per 100 000 person-years; Table 4). The association with use of continuous regimens was nearly identical after excluding the women who reported 15–25 days of progestin use per cycle (data not shown).
|
Analyses restricted to sequential regimens containing 5.0 mg/day or 10.0 mg/day of medroxyprogesterone acetate or continuous regimens containing 2.5 mg/day or 5.0 mg/day of medroxyprogesterone acetate produced similar associations to those based on analyses of all reported sequential or continuous regimen use (RR = 3.09, 95% CI = 1.68 to 5.68; P<.001 and RR = 1.82, 95% CI = 1.03 to 3.23; P = .02, respectively) (Table 4). For both regimens, all associations restricted to these usual regimen–dose combinations were slightly stronger than the associations based on all reported estrogen plus progestin–only use, but again, no consistent duration response was observed for either regimen (Table 4).
Provera (medroxyprogesterone acetate) was the most commonly reported type of progestin. Strong overlap between usual dose and regimen limited within-regimen analyses. None of the women who reported taking continuous regimens containing 5.0 mg/day of progestin developed ovarian cancer. The relative risks associated with sequential regimens did not differ markedly by progestin dose (5.0 mg/day versus 10.0 mg/day) (data not shown).
Women who reported taking the equivalent of the common single-pill continuous combined estrogen plus progestin regimen—daily conjugated equine estrogens at 0.625 mg/day plus daily medroxyprogesterone acetate at 2.5 mg/day—accrued 16 232 person-years. Eight of these women developed ovarian cancer. The relative risk for ever use compared with never use was 1.28 (95% CI = 0.61 to 2.72).
The average reported age at first use of estrogen plus progestin was lower among sequential regimen users who developed cancer (52.1 years) and who did not (51.7 years) than among continuous regimen users who developed ovarian cancer (55.7 years) and who did not (53.9 years). Almost all the ovarian cancers among estrogen plus progestin users occurred in women who were postmenopausal and between ages 57 and 68 at baseline. Stratification by age within that range (57–60, 61–64, and 65–68 years) revealed that the associations of cancer risk with sequential estrogen plus progestin use, although statistically significant in all three groups, were highest among women aged 57–60 years and declined as age increased (data not shown). Conversely, the associations with continuous estrogen plus progestin use increased as age group increased and were statistically significant only among women aged 65–68 at baseline (data not shown). Results were similar after excluding the 8054 women who reported a previous other ovarian surgery at baseline (data not shown).
Cumulative Incidence
The cumulative incidence of developing ovarian cancer in this population was 5.5 per 10 000 person-years. Among women with a hysterectomy, unopposed estrogen users had a 28% higher cumulative incidence compared with never users (7.2 versus 5.6 per 10 000 person-years). Among women with intact uteri, the cumulative incidence for women who never used hormone therapy was 4.9 per 10 000 person-years, whereas cumulative incidence in estrogen plus progestin users (6.0 per 10 000 person-years), sequential estrogen plus progestin users (10.2 per 10 000 person-years), and continuous estrogen plus progestin users (6.6 per 10 000 person-years) were 22%, 108%, and 35% higher, respectively.
|
DISCUSSION |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
In this large cohort, women who used menopausal hormone therapy had elevated risks of developing ovarian cancer compared with women who used no therapy. The increased risks differed by hormone therapy formulation and regimen and varied according to hysterectomy status. The changing formulations, regimens, and patterns of use since the 1970s (22,25) pose challenges for elucidating ovarian cancer risks, which might not emerge until well after use begins. Our relatively detailed data on substantial numbers of hormone therapy users extend the understanding of ovarian cancer by revealing increased risks for women with intact uteri who used estrogen plus progestin formulations.
The cohort's large size and recent data collection allowed us to investigate contemporary hormone therapy use, including unopposed estrogen among women with hysterectomy and estrogen plus progestin among women with intact uteri who only used this formulation. Emphasizing these patterns of use among women who reported specific regimens and doses further increased both the internal and external validity of the findings. We considered most known ovarian cancer risk factors, which reduced the potential for confounding.
The increased risks among long-duration unopposed estrogen users match what other recent US studies observed in analyses that adjusted for hysterectomy status. Three cohort studies (5,6,8) and one case–control study (10) found increased risk among 10-year users. Another cohort study (Danforth KN, Tworoger SS, Hecht J, Rosner BA, Colditz FA, Hankinson SE: personal communication) reported increased risks among women with 5 or more years of estrogen use. Risk estimates in women with hysterectomy are less consistent, but our data indicate that risk of ovarian cancer might be elevated yet slightly attenuated in women with hysterectomy who take unopposed estrogen for long durations. One cohort study found stronger associations with ovarian cancer risk among women with hysterectomy who took unopposed estrogen for a long duration than women with intact uteri (5), and a case–control study reported a statistically nonsignificant positive association with more than 5 years of estrogen use (14). Two other case–control studies (9,13) reported weak and statistically nonsignificant inverse associations with conjugated equine estrogens.
Duration of unopposed estrogen use appeared to be more important than recency of use in our study. Consistent with other studies (5,6,10), we observed that most long-duration unopposed estrogen users were also current users. Millions of US women have used unopposed estrogens, but only a minority have used them for long durations (26). If ovarian cancer risks decline after cessation of long-duration use but remain elevated, as seems to occur with hormone therapy–associated breast cancer risks (27), then ovarian cancer risk could be a concern for long-duration users for some time after use ceases. Adherence to current recommendations, which emphasize short-duration low-dose use for menopausal symptom management only (28,29), would decrease the future number of long-duration unopposed estrogen users. The populations at risk might then only include women who undergo premature surgical menopause and women whose persistent menopausal symptoms long after menopause necessitate use of unopposed estrogen for many years (30,31). At present, there remain insufficient data to estimate risk after stopping unopposed estrogen therapy.
Existing US studies included primarily women who likely used at least 0.625 mg/day of conjugated equine estrogens. Even this large study of more than 97 000 women included few who used other estrogen preparations or doses. Whether estrogens taken at those doses are associated with ovarian cancer is unknown.
Women who used estrogen plus progestin for long durations also had statistically significantly increased risks for ovarian cancer compared with women who used no hormone therapy. This finding contradicts the null associations from our earlier study (5), which included 18 ovarian cancers in women who used only estrogen plus progestin, and one case–control study (13), which included 57 exposed case patients. The three other US studies that showed positive associations between ovarian cancer and menopausal hormone therapy use (6,8,9) collected exposure information when unopposed estrogen would have been used by most, if not all, women.
Other null associations of ovarian cancer risk with estrogen plus progestin were reported in studies that statistically adjusted for hysterectomy (10,14) or use of multiple formulations, including previous unopposed estrogen use (7,11,13). The latter approach might produce statistically misleading results (32), especially if long-duration unopposed estrogen use preceded estrogen plus progestin use. In the current and our previous (5) analyses, the increased risks of ovarian cancer among women who switched from unopposed estrogen to estrogen plus progestin were elevated and similar to the increased risks among unopposed estrogen users.
Our findings of higher risks of ovarian cancer in association with sequential than with continuous regimens are generally similar to results from a large Swedish case–control study (7). The main analyses of that study included previous users of unopposed estrogens, but sensitivity analyses that were restricted to estrogen plus progestin–only users produced stronger associations with ovarian cancer risk for women taking sequential regimens (odds ratio = 1.98, 95% CI = 1.40 to 2.78) than women taking continuous regimens (odds ratio = 1.11, 95% CI = 0.71 to 1.74). The estrogens (estradiol and estriol) and progestins (19-nortestosterone derivatives, such as norethisterone) that are predominantly used in Europe differ from those used in the United States. Only 28 case patients and 138 control subjects in the Swedish study had used the conjugated equine estrogens and 17-hydroxyprogesterone derivatives, such as medroxyprogesterone acetate, that dominate the US market. These differences do not necessarily translate into different risk profiles for each preparation, dose, and regimen (28,31), but they do hinder a direct comparison across studies (11,33).
Sequential regimens and continuous regimens have been used by different groups of women. Monthly withdrawal bleeding accompanies sequential regimens, which have been reported to be used most commonly by perimenopausal and early postmenopausal women (15). The convenience of continuous regimens (especially continuous combined regimens) and absence of breakthrough bleeding in most women after the first few months of use (15) are thought to contribute to preferential use among women who are years past menopause (16). Because users of sequential regimens were slightly younger than users of continuous regimens in our study, a lower absolute risk of ovarian cancer among these younger users of sequential regimens might have generated higher relative risks for sequential regimens. However, age alone seems unlikely to account for the different relative risks because the absolute difference in age was minimal between groups, both groups of women began use in their mid-50s, we adjusted for continuous age, and other risk factor differences between regimens were negligible. Future studies of ovarian cancer risk in association with hormone therapy use should consider both true differences between regimens and confounding by indication (e.g., by menopausal symptoms), which we could not address because we did not ascertain reasons for hormone therapy use.
Like the WHI estrogen plus progestin trial, our study population included more than 8000 estrogen plus progestin users whose mean age was 63 years at study entry (19). With our slightly shorter follow-up period, too few women reported the equivalent dose–regimen combination to accurately estimate risk associated with the single-pill continuous combined estrogen plus progestin regimen. However, risks associated with 5 or more years of use of continuous regimens in our study (RR = 1.55, 95% CI = 0.97 to 2.87 and RR = 1.82, 95% CI = 1.03 to 3.23) were similar to the WHI hazard ratio after an average of 5.6 years of continuous use of combined estrogen plus progestin (hazard ratio = 1.58, 95% CI = 0.77 to 3.24).
The role of steroid hormones in ovarian carcinogenesis is unclear (18). Unopposed estrogen and regimens that contain estrogen plus progestin both produce equivalent increases in circulating serum estrone (34), but epidemiologic studies have not firmly linked higher circulating hormone levels with increased ovarian cancer risk (35,36). Experimental studies in which exogenous hormones stimulated ovarian surface epithelium (37) and altered ovarian expression of estrogen and progesterone receptor subtypes (38–40) supply speculative mechanistic support for the associations we observed.
Study limitations affect our findings. Some analyses relied on small numbers of ovarian cancers. We lacked information on hormone therapy use after the second questionnaire, but the short follow-up period minimized potential exposure misclassification after study entry. Overall hormone therapy use increased in the United States between 1996 and 2000, and therefore, we expect that most participants who reported current use at baseline continued their hormone use during the study period. Consequently, reported duration of use at baseline would have systematically underestimated the true total duration of hormone therapy use in the population during the study period. We could not evaluate whether any cessation of or changes in use after baseline differed by exposure or ovarian cancer status. Even larger studies will be needed to address potential histology-specific associations with hormone therapy. Despite its usual detection at advanced stages, ovarian cancer has potentially recognizable symptoms (41,42). However, the absence of specific symptoms decreases the chance that undetected ovarian cancers in hormone therapy users would have biased their self-reported exposures. We had no data on gynecologic surgery after baseline. Overall hysterectomy prevalence in the United States increased during the study period, with bilateral oophorectomy accompanying approximately half of those hysterectomy procedures. However, a decline in hysterectomy after age 55 years (43) means that only a small number of NIH–AARP study participants would be expected to have had gynecologic surgery after baseline. The baseline questionnaire, which was sent to a large and representative group of women over age 50, generated a low response, but 62% of its respondents completed the second questionnaire. Response was not associated with hysterectomy status or menopause. However, compared with nonrespondents, respondents to the second questionnaire were more likely to be older and non-Hispanic white, to report overall excellent or very good health, to report a BMI below 25 kg/m2, to have been current and longer duration hormone therapy users, and to have attended college. We therefore expect that our results are generalizable to a similar population of women over age 50 years in the United States.
Since the late 1960s, menopausal hormone therapy use has grown, declined, expanded, narrowed, and shifted. Especially for rare outcomes with long latency periods, such as ovarian cancer, these dynamic exposures pose major challenges to accurately assessing risk. The increased ovarian cancer risks we observed among long-duration users of unopposed estrogens will likely diminish in importance if recent trends in use continue. Our study provides evidence that links use of estrogen plus progestin, especially sequential regimens, with increased ovarian cancer risk. The increased absolute risks appear to be small, and other risk–benefit considerations may dominate patients' and clinicians' decision making regarding hormone therapy. Nonetheless, these associations, if real, represent potentially avoidable risk factors for a highly fatal cancer and therefore warrant continued investigation.
|
NOTES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
This research was supported in part by the Intramural Research Program of the NIH, National Cancer Institute. The funding agency had no role in the data collection and analysis, the interpretation of the findings, or the preparation of the manuscript. Cancer incidence data from the Atlanta metropolitan area were collected by the Georgia Center for Cancer Statistics, Department of Epidemiology, Rollins School of Public Health, Emory University. Cancer incidence data from California were collected by the California Department of Health Services, Cancer Surveillance Section. Cancer incidence data from the Detroit metropolitan area were collected by the Michigan Cancer Surveillance Program, Community Health Administration, State of Michigan. The Florida cancer incidence data used in this report were collected by the Florida Cancer Data System under contract to the Department of Health (DOH). The views expressed herein are solely those of the authors and do not necessarily reflect those of the contractor or DOH. Cancer incidence data from Louisiana were collected by the Louisiana Tumor Registry, Louisiana State University Medical Center in New Orleans. Cancer incidence data from New Jersey were collected by the New Jersey State Cancer Registry, Cancer Epidemiology Services, New Jersey State Department of Health and Senior Services. Cancer incidence data from North Carolina were collected by the North Carolina Central Cancer Registry. Cancer incidence data from Pennsylvania were supplied by the Division of Health Statistics and Research, Pennsylvania Department of Health, Harrisburg, PA. The Pennsylvania Department of Health specifically disclaims responsibility for any analyses, interpretations, or conclusions.
We thank Leslie Carroll, Matthew Butcher, and Dave Campbell at Information Management Services, Inc, Silver Spring, MD, for data support.
|
REFERENCES |
|---|
|
TopNotesAbstractIntroductionSubjects and methodsResultsDiscussionReferences |
|---|
(1) Whittemore AS, Harris R, Itnyre J. Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. II. Invasive epithelial ovarian cancers in white women. Collaborative Ovarian Cancer Group. Am J Epidemiol 1992;136:1184–203.
(2) Kaufman DW, Kelly JP, Welch WR, Rosenberg L, Stolley PD, Warshauer ME, et al. Noncontraceptive estrogen use and epithelial ovarian cancer. Am J Epidemiol 1989;130:1142–51.
(3) Hildreth NG, Kelsey JL, LiVolsi VA, Fischer DB, Holford TR, Mostow ED, et al. An epidemiologic study of epithelial carcinoma of the ovary. Am J Epidemiol 1981;114:398–405.
(4) Purdie DM, Bain CJ, Siskind V, Russell P, Hacker NF, Ward BG, et al. Hormone replacement therapy and risk of epithelial ovarian cancer. Br J Cancer 1999;81:559–63.[CrossRef][ISI][Medline]
(5) Lacey JV Jr, Mink PJ, Lubin JH, Sherman ME, Troisi R, Hartge P, et al. Menopausal hormone replacement therapy and risk of ovarian cancer. JAMA 2002;288:334–41.
(6) Folsom AR, Anderson JP, Ross JA. Estrogen replacement therapy and ovarian cancer. Epidemiology 2004;15:100–4.[CrossRef][ISI][Medline]
(7) Riman T, Dickman PW, Nilsson S, Correia N, Nordliner H, Magnusson CM, et al. Hormone replacement therapy and the risk of invasive epithelial ovarian cancer in Swedish women. J Natl Cancer Inst 2002;94:497–504.
(8) Rodriguez C, Patel AV, Calle EE, Jacob EJ, Thun MJ. Estrogen replacement therapy and ovarian cancer mortality in a large prospective study of US women. JAMA 2001;285:1460–5.
(9) Mills PK, Riordan DG, Cress RD. Epithelial ovarian cancer risk by invasiveness and cell type in the Central Valley of California. Gynecol Oncol 2004;95:215–25.[CrossRef][ISI][Medline]
(10) Moorman PG, Schildkraut JM, Calingaert B, Halabi S, Berchuck A. Menopausal hormones and risk of ovarian cancer. Am J Obstet Gynecol 2005;193:76–82.[CrossRef][ISI][Medline]
(11) Glud E, Kjaer SK, Thomsen BL, Hogdall C, Christensen L, Hogdall E, et al. Hormone therapy and the impact of estrogen intake on the risk of ovarian cancer. Arch Intern Med 2004;164:2253–9.
(12) Beral V, Banks E, Reeves G, Appleby P. Use of HRT and the subsequent risk of cancer. J Epidemiol Biostat 1999;4:191–210.[Medline]
(13) Sit AS, Modugno F, Weissfeld JL, Berga SL, Ness RB. Hormone replacement therapy formulations and risk of epithelial ovarian carcinoma. Gynecol Oncol 2002;86:118–23.[CrossRef][ISI][Medline]
(14) Pike MC, Pearce CL, Peters R, Cozen W, Wan P, Wu AH. Hormonal factors and the risk of invasive ovarian cancer: a population-based case-control study. Fertil Steril 2004;82:186–95.[ISI][Medline]
(15) Guidelines for counseling postmenopausal women about preventive hormone therapy. American College of Physicians. Ann Intern Med 1992;117:1038–41.[ISI][Medline]
(16) McNagny SE. Prescribing hormone replacement therapy for menopausal symptoms. Ann Intern Med 1999;131:605–16.
(17) Nelson HD. Postmenopausal estrogen for treatment of hot flashes: clinical applications. JAMA 2004;291:1621–5.
(18) Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis concerning the role of androgens and progesterone. J Natl Cancer Inst 1998;90:1774–86.
(19) Anderson GL, Judd HL, Kaunitz AM, Barad DH, Beresford SA, Pettinger M, et al. Effects of estrogen plus progestin on gynecologic cancers and associated diagnostic procedures: the Women's Health Initiative randomized trial. JAMA 2003;290:1739–48.
(20) Schatzkin A, Subar AF, Thompson FE, Harlan LC, Tangrea J, Hollenbeck AR, et al. Design and serendipity in establishing a large cohort with wide dietary intake distributions: the National Institutes of Health-American Association of Retired Persons Diet and Health Study. Am J Epidemiol 2001;154:1119–25.
(21) Available at: http://dietandhealth.cancer.gov/questionnaires.html. [Last accessed: August 7, 2006.]
(22) Stefanick ML. Estrogens and progestins: background and history, trends in use, and guidelines and regimens approved by the US Food and Drug Administration. Am J Med 2005;118:64–73.[CrossRef][Medline]
(23) Michaud DS, Midthune D, Hermansen S, Leitzmann M, Harlan LC, Kipnis V, et al. Comparison of cancer registry case ascertainment with SEER estimates and self-reporting in a subset of the NIH-AARP Diet and Health Study. J Regist Manag 2005;32:70–5.
(24) Gail MH, Lubin JH, Rubinstein LV. Likelihood calculations for matched case-control studies and survival studies with tied death times. Biometrika 1981;68:703–7.
(25) Hersh AL, Stefanick ML, Stafford RS. National use of postmenopausal hormone therapy: annual trends and response to recent evidence. JAMA 2004;291:47–53.
(26) Brett KM, Reuben CA. Prevalence of estrogen or estrogen-progestin hormone therapy use. Obstet Gynecol 2003;102:1240–9.
(27) Pike MC, Spicer DV. Hormonal contraception and chemoprevention of female cancers. Endocr Relat Cancer 2000;7:73–83.[Abstract]
(28) Recommendations for estrogen and progestogen use in peri-and postmenopausal women: October 2004 position statement of The North American Menopause Society. Menopause 2004;11:589–600.[CrossRef][Medline]
(29) Postmenopausal hormone replacement therapy for primary prevention of chronic conditions: recommendations and rationale. Ann Intern Med 2002;137:834–9.
(30) Grady D. A 60-year-old woman trying to discontinue hormone replacement therapy. JAMA 2002;287:2130–7.
(31) American College of Obstetricians and Gynecologists Women's Health Care Physicians. Executive summary. Hormone therapy. Obstet Gynecol 2004;104:1S–4S.
(32) McKnight B, Cook LS, Weiss NS. Logistic regression analysis for more than one characteristic of exposure. Am J Epidemiol 1999;149:984–92.
(33) Bakken K, Alsaker E, Eggen AE, Lund E. Hormone replacement therapy and incidence of hormone-dependent cancers in the Norwegian Women and Cancer study. Int J Cancer 2004;112:130–4.[CrossRef][ISI][Medline]
(34) Ursin G, Palla SL, Reboussin BA, Slone S, Wasilauskas C, Pike MC, et al. Post-treatment change in serum estrone predicts mammographic percent density changes in women who received combination estrogen and progestin in the Postmenopausal Estrogen/Progestin Interventions (PEPI) trial. J Clin Oncol 2004;22:2842–8.
(35) Helzlsouer KJ, Alberg AJ, Gordon GB, Longcope C, Bush TL, Hoffman SC, et al. Serum gonadotropins and steroid hormones and the development of ovarian cancer. JAMA 1995;274:1926–30.[Abstract]
(36) Lukanova A, Lundin E, Akhmedkhanov A, Micheli A, Rinaldi S, Zeleniuch-Jacquotte A, et al. Circulating levels of sex steroid hormones and risk of ovarian cancer. Int J Cancer 2003;104:636–42.[CrossRef][ISI][Medline]
(37) Syed V, Ulinski G, Mok SC, Yiu GK, Ho SM. Expression of gonadotropin receptor and growth responses to key reproductive hormones in normal and malignant human ovarian surface epithelial cells. Cancer Res 2001;61:6768–76.
(38) Lindgren PR, Cajander S, Backstrom T, Gustafsson JA, Makela S, Olofsson JI. Estrogen and progesterone receptors in ovarian epithelial tumors. Mol Cell Endocrinol 2004;221:97–104.[CrossRef][ISI][Medline]
(39) Li AJ, Baldwin RL, Karlan BY. Estrogen and progesterone receptor subtype expression in normal and malignant ovarian epithelial cell cultures. Am J Obstet Gynecol 2003;189:22–7.[CrossRef][ISI][Medline]
(40) Mukherjee K, Syed V, Ho SM. Estrogen-induced loss of progesterone receptor expression in normal and malignant ovarian surface epithelial cells. Oncogene 2005;24:4388–400.[CrossRef][ISI][Medline]
(41) Smith LH, Morris CR, Yasmeen S, Parikh-Patel A, Cress RD, Romano PS. Ovarian cancer: can we make the clinical diagnosis earlier? Cancer 2005;104:1398–407.[CrossRef][ISI][Medline]
(42) Goff BA, Mandel LS, Melancon CH, Muntz HG. Frequency of symptoms of ovarian cancer in women presenting to primary care clinics. JAMA 2004;291:2705–12.
(43) Keshavarz H, Hillis SD, Marchbanks PA. Hysterectomy surveillance—United States, 1994–1999. Atlanta (GA): Centers for Disease Control and Prevention; 2002 Jul 12 Report No.: SS-5.
Manuscript received March 28, 2006; revised July 17, 2006; accepted August 17, 2006.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  J. V. Lacey Jr. Ovarian Cancer and Menopausal Hormone Therapy: More Data and New Questions Cancer Epidemiol. Biomarkers Prev., December 1, 2007; 16(12): 2524 - 2525. [Full Text] [PDF]
|
|
|
|
|
|
M. A. Rossing, K. L. Cushing-Haugen, K. G. Wicklund, J. A. Doherty, and N. S. Weiss Menopausal Hormone Therapy and Risk of Epithelial Ovarian Cancer Cancer Epidemiol. Biomarkers Prev., December 1, 2007; 16(12): 2548 - 2556. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Beesley, S. J. Jordan, A. B. Spurdle, H. Song, S. J. Ramus, S. K. Kjaer, E. Hogdall, R. A. DiCioccio, V. McGuire, A. S. Whittemore, et al. Association Between Single-Nucleotide Polymorphisms in Hormone Metabolism and DNA Repair Genes and Epithelial Ovarian Cancer: Results from Two Australian Studies and an Additional Validation Set Cancer Epidemiol. Biomarkers Prev., December 1, 2007; 16(12): 2557 - 2565. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Baik, M. Lambe, Q. Liu, S. Cnattingius, L. A. Mucci, T. Riman, A. Ekbom, H.-O. Adami, and C.-C. Hsieh Gender of Offspring and Maternal Risk of Invasive Epithelial Ovarian Cancer Cancer Epidemiol. Biomarkers Prev., November 1, 2007; 16(11): 2314 - 2320. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Greiser, E. M. Greiser, and M. Doren Menopausal hormone therapy and risk of ovarian cancer: systematic review and meta-analysis Hum. Reprod. Update, September 1, 2007; 13(5): 453 - 463. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Does Menopausal Hormone Therapy Increase Risk for Ovarian Cancer? Journal Watch (General), November 2, 2006; 2006(1102): 3 - 3. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||