Increasing cancer survival rates have led to the need to manage menopausal symptoms in survivors of epithelial ovarian cancer.1 Menopause resulting from cancer treatment has a more abrupt onset and severe symptoms, including high rates of sexual dysfunction and decreased quality of life.1–3 In addition, there are long-term health consequences of premature menopause, including bone loss and cardiovascular disease.4 Hormone therapy (HT) is effective in treating these symptoms and improving long-term health.5
Care providers are hesitant to prescribe HT after the results of the Women's Health Initiative and Nurses' Health Study.6–9 The use of HT in survivors of ovarian cancer presents the added theoretical concern of increasing the risk of recurrence, especially in survivors of nonserous epithelial ovarian cancer. These cancers may be associated with endometriosis, which is known to be hormonally responsive.10–15 Recent studies have also shown that nonserous subtypes are more likely to be associated with long-term survival and improved prognosis.16,17 Theoretically, the women who could benefit the most from HT are also the women who have the highest risk of recurrence.
We hypothesized that HT use after treatment of nonserous epithelial ovarian cancer could negatively affect survival as a result of the association of this subtype with endometriosis and the potential for hormonal responsiveness. Our study looks to determine whether hormone use and duration of hormone use, specifically after treatment of nonserous epithelial ovarian cancer, are associated with a decrease in overall survival and disease-free survival.
MATERIALS AND METHODS
This retrospective cohort study was approved by Manitoba Health (HIPC 2012/2013–17), The University of Manitoba Research Ethics Board (REB #HS15610[H2012:258]), and CancerCare Manitoba (RRIC #29-2012). The Winnipeg HSC Medical Staff Council Award and the University of Manitoba Department of Obstetrics, Gynecology and Reproductive Sciences Resident Research Fund provided funding.
Data held by the Manitoba Cancer Registry were used to extract the cancer cohort as well as the demographic, diagnostic, treatment, and death information. The HT data were extracted from the Drug Programs Information Network at Manitoba Health, which captures prescriptions filled in the community. Clinical information and personal histories were extracted by chart review. All women of any age with known nonserous epithelial ovarian, fallopian tube, or primary peritoneal cancer between 1995 and 2010 were included. Only patients receiving treatment for their ovarian cancer (ie, surgery, chemotherapy, or radiation) and who had follow-up information were included in the analysis. All patients who underwent surgical treatment had unilateral or bilateral oophorectomy plus any additional surgery (ie, hysterectomy, omentectomy, full debulking).
Outcomes included overall survival and disease-free survival. December 31, 2011, was the end of study date. The date of first treatment (either surgery, chemotherapy, or radiation) was used as the start time. Overall survival is defined as time from first treatment to death, last contact, or end-of-study date. Disease-free survival is defined as time from first treatment to the first of recurrence, death, last contact, or end-of-study date. Hormone therapy use was defined with the Anatomic Therapeutic Classification codes of G03A, G03C, G03D, and G03F. Hormone therapy use was analyzed using two approaches: a landmark analysis, in which follow-up started 6 months after the date of first treatment, and HT was defined as presence or absence during those first 6 months and a time-varying approach that included duration of use. With the supply days and dates provided for each prescription, a time-varying variable of HT cumulative use was defined as nonuser, initiation to 12 months of use, and 12 or more months of use. The initiation of HT use had to begin within 6 months of first treatment or the patient was censored at the date of HT initiation. Twelve months of HT use had to begin within 21 months of first treatment or the patient was censored at the date of 12 months' use. Censoring was used to prevent delayed HT exposure from influencing the results. Other predictors included treatment (surgery, chemotherapy, or radiation), body mass index (calculated as weight (kg)/[height (m)]2) at diagnosis, period of diagnosis, age at diagnosis, grade, smoking status, neighborhood income, residence (Winnipeg, non-Winnipeg), International Federation of Gynecology and Obstetrics (FIGO) stage, imaging (lymph node-positive, omentum-positive, bowel-positive, carcinomatosis, ascites), and family history (breast, ovarian, pancreatic, prostate, and colon cancer).
Separate analyses were conducted for younger (younger than 55 years of age) and older women (55 years of age and older). Additional analyses were conducted for the subcohort that received surgery as first treatment. Landmark and time-varying Cox regression models were used to analyze overall survival and disease-free survival. Time-varying Cox regression models were used to analyze overall survival and disease-free survival. Time-varying models were used to allow the inclusion of the time-varying predictors of HT use and treatment. Proportional hazard assumptions were tested using Schoenfeld residuals. The change-in-estimate approach was used for model building: a change of 10% in either of the coefficients of the HT variable was required for inclusion as a control variable. The survival package was used for analyzing the outcomes of overall survival and disease-free survival using R 3.1.0. Missing FIGO stage, grade, body mass index, and smoking status were assumed to be missing at random. The mice package was used to produce 30 imputations. Imputations were verified by comparing the distributions of observed and imputed data conditional on propensity score.
A total of 391 patients met the criteria for nonserous epithelial ovarian cancer between 1995 and 2010. Seventeen patients were excluded because the patients did not receive treatment for their diagnosis of cancer, and 17 were excluded for lack of follow-up, leaving 357 patients who met our inclusion criteria. Ninety-four women received HT after treatment and 263 women did not (Fig. 1). The median follow-up of the cohort was 3.18 years (25th percentile: 1.39; 75th percentile: 5.08).
The average age of the entire cohort was 57.8 years old. We split the cohort into two groups: women older than 55 years of age and women younger than 55 years of age. Baseline characteristics of the two cohorts are presented in Table 1. Women younger than 55 years of age who received HT were more likely to be younger, be prescribed HT before 2002, more likely to have negative omentum on imaging, more likely to have surgery as their sole treatment, and had longer follow-up than women who did not receive HT. Women older than 55 years of age who received HT were more likely to be younger, more likely to be prescribed HT before 2002, less likely to have been on hormonal medications any time before diagnosis (ie, HT or hormonal contraception), and had longer follow-up than women who did not receive HT. There were no differences between tumor histology between the two groups (Table 1).
Of the women who received hormones, 76% received oral conjugated equine estrogen; only 7% of women received vaginal conjugated equine estrogen either alone or in combination with oral conjugated equine estrogen. Fourteen percent of women received some type of estrogen combined with a progestin, and 2% of women received a progestin alone. Dosages were variable, but the most common dose of oral conjugated equine estrogen was 0.625 mg. Women were more likely to be prescribed HT after treatment for ovarian cancer before 2002. The median HT duration (for those with more than 1 year of use) for those younger than 55 years of age was 4.76 years. The median HT duration (for those with more than 1 year of use) for those 55 years of age and older was 4.04 years.
We used a landmark analysis to assess overall and disease-free survival for women who used HT at the 6-month mark. This landmark was chosen because we presumed that women who survived their treatment for ovarian cancer would have become symptomatic from menopause by that time. Among patients younger than 55 years of age who were alive at 6 months after treatment, overall survival was higher in the HT group compared with the non-HT group according to the univariable analysis (hazard ratio [HR] 0.410, 95% confidence interval [CI] 0.19–0.89, P=.023), but this result was not significant in the multivariable model correcting for FIGO stage and chemotherapy use (Table 2). Disease-free survival in this group was also improved in both the univariable model (HR 0.339, 95% CI 0.17–0.69, P=.003) and the multivariable model (adjusted HR 0.354, 95% CI 0.17–0.74, P=.006) (Table 3). Similarly, in our landmark analysis for women younger than 55 years of age who underwent surgery as their first treatment, overall survival was higher in the univariable analysis (HR 0.387, 95% CI 0.17–0.86, P=.021) but was not statistically significant in the multivariable model. Disease-free survival in this group was improved in both the univariable model (HR 0.319, 95% CI 0.15–0.67, P=.003) and the multivariable model (adjusted HR 0.342, 95% CI 0.16–0.73, P=.006). International Federation of Gynecology and Obstetrics stage III and IV and the need for chemotherapy decreased overall and disease-free survival in both patient groups in the multivariable analysis; however, the results for chemotherapy were not statistically significant (Tables 2 and 3). Among patients older than 55 years of age who were alive at 6 months after treatment, HT use did not affect overall or disease-free survival in both the univariable and multivariable analysis. Similarly, in our landmark analysis for women older than 55 years of age who underwent surgery as their first treatment, HT use did not affect overall or disease-free survival in both the univariable and multivariable analysis. Age at diagnosis and FIGO stage III and IV negatively affected overall and disease-free survival in both patient groups. The use of chemotherapy only as treatment negatively affected overall survival but not disease-free survival in women older than 55 years of age (Tables 4 and 5).
We used a time-varying Cox regression model to assess the effect that duration of hormone use would have on overall and disease-free survival. According to the time-varying Cox regression model, HT use among women younger than 55 years of age was related to significantly greater overall survival (HR 0.291, 95% CI 0.11–0.79, P=.016) and disease-free survival (HR 0.233, 95% CI 0.09–0.63, P=.004) when used for longer than 12 months (Tables 6 and 7). When adjusting for FIGO stage and chemotherapy use, disease-free survival was also improved (adjusted HR 0.212, 95% CI 0.07–0.60, P=.004), but overall survival was no longer statistically significant. Hormone use for less than 12 months was also associated with increased overall and disease-free survival, but the results were not statistically significant. Similarly, among women younger than 55 years of age who underwent surgery as their first treatment, HT use for longer than 12 months was associated with improved overall survival (HR 0.305, 95% CI 0.11–0.83, P=.021) and disease-free survival (HR 0.242, 95% CI 0.09–0.65, P=.005). When adjusting for FIGO stage and chemotherapy use, disease-free survival was also improved (adjusted HR 2.17, 95% CI 0.08–0.62, P=.004), but overall survival was no longer statistically significant. Hormone use for less than 12 months in this cohort was also associated with improved overall and disease-free survival, but the results were not statistically significant. International Federation of Gynecology and Obstetrics stage III and IV disease was associated with decreased overall and disease-free survival in women younger than 55 years of age and in women in this age group who underwent surgery as their first treatment. The need for chemotherapy was also associated with decreased overall and disease-free survival in both groups, but this was not statistically significant (Tables 6 and 7). According to the time-varying Cox regression model, HT use for any duration among women older than 55 years of age was not associated with significant differences with respect to overall and disease-free survival. Women in this age group who underwent surgery as their first treatment also did not have significant differences in overall and disease-free survival with respect to hormone use for any duration. Age at diagnosis and FIGO stage III and IV disease were associated with decreased overall and disease-free survival in women younger than 55 years of age and in women younger than 55 years of age who underwent surgery as their first treatment. The need for chemotherapy was also associated with decreased overall and disease-free survival in both groups, but this was not statistically significant (Tables 8 and 9).
In our study, the use of HT after treatment of nonserous epithelial ovarian cancer was not associated with harm. We did not see decreased overall or disease-free survival. In fact, in our landmark analysis, HT use in women younger than 55 years of age was associated with increased disease-free survival. In addition, extended use of HT in women younger than 55 years of age was associated with an increase in overall survival and disease-free survival even when correcting for better prognostic factors such as FIGO stage and the need for adjuvant treatment.
There was a proportion of women who were not prescribed HT but who theoretically would have been good candidates for HT given their age. In our cohort, 42.1% of women younger than 50 years of age did not receive HT. There were 23 women (21.5%) who were treated for endometrioid epithelial ovarian cancer, 5 women (4.7%) who were treated for clear cell ovarian cancer, and 12 women (11.2%) who were treated for mucinous epithelial ovarian cancer age younger than 50 years old who did not receive HT. These women were likely premenopausal before treatment. In addition to the alleviation of vasomotor symptoms, HT would have provided bone and cardiovascular protection to these young newly postmenopausal women.
Limitations of our study include the fact that it is a retrospective cohort study with a small sample size, which limits our ability to discern differences in HR. There was a selection bias favoring HT use in women who were younger with better prognostic factors. This makes sense because these women are more likely to live longer and experience menopausal symptoms. We attempted to correct for this in our analysis by conducting a separate analysis for younger women. In addition, most women were on oral conjugated equine estrogen rather than a transdermal preparation. As a result of the results of recent studies, including the Women's Health Initiative publications, prescribing practices favor transdermal estradiol preparations as a result of potential decreased risk of venous thromboembolism.18 In our study, more women were prescribed HT before 2002. In addition, there was a small proportion of women who received vaginal estrogen (7%) and we were unable to determine what percentage of that population also received oral treatment.
The strength of our study is that we focused solely on nonserous epithelial ovarian cancer, which is crucial because these histologies are linked to hormone stimulation and are associated with longer survival.16 These are the women who might benefit most from HT but who are also at highest theoretical risk of recurrence as a result of exposure to hormones. We found that these women may instead actually benefit from HT use. This is similar to other studies that include both serous and nonserous epithelial ovarian cancer.19–25 Another strength of our study is that we studied the effect of duration of HT use on our primary outcomes. In addition, we used drug programs' information network information to capture HT use, which is more accurate than gathering this information from chart review or telephone interviews.
Published evidence supports the notion that HT improves quality of life and reduces risk of coronary heart disease, osteoporosis, colon cancer, and diabetes in healthy women with postmenopausal vasomotor symptoms.5 The studies on female survivors of cancer show that women rate menopausal symptoms, sexual dysfunction, and infertility as the most bothersome issues associated with survivorship.1,2,26 The use of HT to treat these symptoms can significantly improve the quality of life of cancer survivors. According to the results of our study, combined with the evidence in the literature, HT may be considered in survivors of nonserous epithelial ovarian cancer who experience menopausal symptoms shortly after treatment.
1. Absolom K, Eiser C, Turner L, Ledger W, Ross R, Davies H, et al.. Ovarian failure following cancer treatment: current management and quality of life. Hum Reprod 2008;23:2506–12.
2. King J, Wynne CH, Assersohn L, Jones A. Hormone replacement therapy and women with premature menopause—a cancer survivorship issue. Eur J Cancer 2011;47:1623–32.
3. Wenzel LB, Donnelly JP, Fowler JM, Habbal R, Taylor TH, Aziz N, et al.. Resilience, reflection, and residual stress in ovarian cancer survivorship: a gynecologic oncology group study. Psychooncology 2002;11:142–53.
4. Maclaran K, Horner E, Panay N. Premature ovarian failure: long-term sequelae. Menopause Int 2015;16:38–41.
5. North American Menopause Society. The 2012 hormone therapy position statement of: The North American Menopause Society. Menopause 2012;19:257–71.
6. Stampfer MJ, Colditz GA, Willett WC, Manson JE, Rosner B, Speizer FE, et al.. Postmenopausal estrogen therapy and cardiovascular disease. Ten-year follow-up from the nurses' health study. N Engl J Med 1991;325:756–62.
7. Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, et al.. Risks and benefits of estrogen plus progestin in healthy postmenopausal women. JAMA 2002;288:321–33.
8. Cauley JA, Robbins J, Chen Z, Cummings SR, Jackson RD, Lacroix AZ, et al.. Effects of estrogen plus progestin on risk of fracture and bone mineral density: the Women's Health Initiative randomized trial. JAMA 2003;290:1729–38.
9. Rossouw JE, Prentice RL, Manson JE, Wu L, Barad D, Barnabei VM, et al.. Postmenopausal hormone therapy and risk of cardiovascular disease by age and years since menopause. JAMA 2007;297:1465–77.
10. Brinton LA, Gridley G, Persson I, Baron J, Bergqvist A. Cancer risk after a hospital discharge diagnosis of endometriosis. Am J Obstet Gynecol 1997;176:572–9.
11. Yang HP, Trabert B, Murphy MA, Sherman ME, Sampson JN, Brinton LA, et al.. Ovarian cancer risk factors by histologic subtypes in the NIH-AARP Diet and Health Study. Int J Cancer 2012;131:938–48.
12. Sayasneh A, Tsivos D, Crawford R. Endometriosis and ovarian cancer: a systematic review. ISRN Obstet Gynecol 2011;2011:140310.
13. Morch LS, Løkkegaard E, Andreasen AH, Kjaer SK, Lidegaard O. Hormone therapy and different ovarian cancers: a national cohort study. Am J Epidemiol 2012;175:1234–42.
14. Koskela-Niska V, Pukkala E, Lyytinen H, Ylikorkala O, Dyba T. Effect of various forms of postmenopausal hormone therapy on the risk of ovarian cancer—a population-based case control study from Finland. Int J Cancer 2013;133:1680–8.
15. Koskela-Niska V, Lyytinen H, Riska a, Pukkala E, Ylikorkala O. Ovarian cancer risk in postmenopausal women using estradiol–progestin therapy—a nationwide study. Climacteric 2013;16:48–53.
16. Cress RD, Chen YS, Morris CR, Petersen M, Leiserowitz GS. Characteristics of long-term survivors of epithelial ovarian cancer. Obstet Gynecol 2015;126:491–7.
17. Heintz AP, Odicino F, Maisonneuve P, Beller U, Benedet JL, Creasman WT, et al.. Carcinoma of the ovary. FIGO 26th annual report on the results of treatment in gynecological cancer. Int J Gynaecol Obstet 2006;95(suppl 1):S161–92.
18. Scarabin PY, Oger E, Plu-Bureau G; EStrogen and THromboEmbolism Risk Study Group. Differential association of oral and transdermal oestrogen-replacement therapy with venous thromboembolism risk. Lancet 2003;362:428–32.
19. Bebar S, Ursic-Vrscaj M. Hormone replacement therapy after epithelial ovarian cancer treatment. Eur J Gynaecol Oncol 2000;21:192–6.
20. Eeles RA, Tan S, Wiltshaw E, Fryatt I, A'Hern RP, Shepherd JH, et al.. Hormone replacement therapy and survival after surgery for ovarian cancer. BMJ 1991;302:259–62.
21. Mascarenhas C, Lambe M, Bellocco R, Bergfeldt K, Riman T, Persson I, et al.. Use of hormone replacement therapy before and after ovarian cancer diagnosis and ovarian cancer survival. Int J Cancer 2006;119:2907–15.
22. Wen Y, Huang H, Huang H, Wu M, Shen K, Pan L. The safety of postoperative hormone replacement therapy in epithelial ovarian cancer patients in China. Climacteric 2013;16:673–81.
23. Guidozzi F, Daponte A. Estrogen replacement therapy for ovarian carcinoma survivors: A randomized controlled trial. Cancer 1999;86:1013–8.
24. Eeles RA, Morden JP, Gore M, Mansi J, Glees J, Wenczl M. Adjuvant Hormone Therapy May Improve Survival in Epithelial Ovarian Cancer: Results of the AHT Randomized Trial. J Clin Oncol 2015;33:4138–44.
25. Ursic-Vrscaj M, Bebar S, Zakelj MP. Hormone replacement therapy after invasive ovarian serous cystadenocarcinoma treatment: the effect on survival. Menopause 2001;8:70–5.
© 2016 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
26. Ganz PA, Rowland JH, Desmond K, Meyerowitz BE, Wyatt G. Life after breast cancer: understanding women's health-related quality of life and sexual functioning. J Clin Oncol 1998;16:501–14.