To Treat or Not to Treat: The Use of Hormone Replacement Therapy in Patients With Ovarian Cancer

The surgical treatment of epithelial ovarian cancer (EOC) results in the abrupt induction of menopause in pre- and perimenopausal women, who account for approximately 20% of all patients with ovarian cancer.¹ The loss of ovarian estrogen results in disruptive vasomotor symptoms, which can negatively affect quality of life, increases the risk of bone loss and fractures, and increases the risk of coronary heart disease (CHD).^2,3 These adverse effects of the loss of ovarian function can be readily treated with hormone replacement therapy (HRT).^2,4,5 However, many physicians treating patients with ovarian cancer have been reluctant to use HRT.

Nearly 80% of serous and endometrioid EOCs express estrogen receptor (ER).^6,7 ER expression is highest in the subtypes of EOC, serous and endometrioid, which account for > 75% of all EOC.⁶ Clinical trials have shown limited activity of agents that disrupt ER signaling (eg, tamoxifen, aromatase inhibitors) in advanced recurrent ovarian cancer, consistent with the preclinical suggestion that these receptors may not be functional. Simpkins et al⁷ reviewed 20 published trials of tamoxifen, a treatment agnostic of ER status, finding an overall response rate of 13% with 35% stable disease. However, epidemiologic studies have demonstrated a correlation between the use of either HRT or oral contraceptives and a reduced incidence of ovarian cancer.⁸ The most compelling concern regarding HRT use in patients with EOC is that HRT may increase the incidence of breast cancer, CHD, and thromboembolic events in these patients. Epidemiologic data from the Women’s Health Initiative show that HRT with estrogen and progestin increases all these risks in healthy postmenopausal women⁹; however, HRT with estrogen alone, as would be used in patients with EOC posthysterectomy, does not increase the risk of breast cancer, CHD, or all-cause mortality.^10–12 Still, in the absence of data that HRT does not worsen the outcomes of patients with EOC, patients continue to experience vasomotor symptoms unabated by HRT.

Few studies have addressed the issue of HRT in women with EOC. Two small retrospective studies (78 and 24 patients) found no difference in survival for patients who had taken HRT after the diagnosis of ovarian cancer.^13,14 Similarly, a randomized trial of 130 patients by Guidozzi and Daponte¹⁵ found that HRT did not significantly affect disease-free interval or survival. These studies, although supportive of the safety of HRT in patients with ovarian cancer, were small, and all were biased toward younger women and well-differentiated tumors.

In the article that accompanies this editorial, Eeles et al¹⁶ present the results of a long-awaited multinational, randomized, nonblinded phase III trial of HRT in pre- and postmenopausal women receiving a recent diagnosis of EOC. Patients were randomly assigned 1:1 to treatment with adjuvant hormonal therapy, HRT in the postdiagnosis period, or no treatment. Patients randomly assigned to treatment with HRT were to continue to a minimum of 5 years if tolerated and were treated for ovarian cancer according to standard guidelines at the time. The study excluded patients in whom ovarian function was preserved or who had a history of a hormone-dependent malignancy. The primary end point for this study was overall survival (OS), and the main secondary end point was relapse-free survival (RFS). Other specified secondary end points were compliance with HRT, myocardial infarction, fracture, transient ischemic attack, cerebrovascular accident, and second cancer. All analyses were based on the intention to treat. The trial was closed early as a result of difficulty in accruing the 570 patients projected for the OS analysis.

The trial randomly assigned 150 patients, and median follow-up at the time of this analysis for patients who had not died was approximately 19 years. The patients in both cohorts were well balanced for clinical characteristics. Twenty-three percent of the patients were premenopausal, and 63% were in stage III of the International Federation of Gynecology and Obstetrics staging system. Seventy-two of the 75 patients randomly assigned to the HRT group received at least 1 day of HRT, and 67 of the 75 patients randomly assigned to the no HRT group received no hormone therapy during the study period. Long-term OS, both nonadjusted and adjusted for stratification factors, was superior in the HRT group (adjusted hazard ratio [HR], 0.45; P < .001). Thus, the primary end point showed that HRT results in significantly improved OS in patients taking HRT. Progression-free survival was better in the HRT group than in the no HRT group (adjusted HR, 0.53; P = .004). In the prior randomized study by Guidozzi and Daponte,¹⁵ the data showed a nonsignificant trend toward improved OS in patients treated with HRT. Eeles et al¹⁶ combined their data with the randomized Guidozzi and Daponte¹⁵ study and found that the HR for OS of the two studies together is 0.68 in favor of HRT. Interestingly, although the Eeles et al¹⁶ study was not powered to address differences in individual causes of mortality, the patients taking HRT had a decrease in death due to ovarian cancer (67% v 75%) as well as other causes (4% v 16%), including CHD and thromboembolic disease.

The surprising finding was that this OS benefit and trend to disease-specific survival occurred with a median HRT exposure on the treatment arm of 1.14 years, with more than one-half of the patients discontinuing therapy prematurely. Only approximately one-third of patients continued HRT until death or last follow-up, and the benefits of HRT persisted throughout the duration of the study. As the authors point out, persistence of treatment effects beyond the duration of therapy has also been seen in breast cancer prevention trials¹⁷ and in the adjuvant setting in early stage breast cancer.¹⁸ Curiously, those studies used tamoxifen, which blocks the effects of estrogen on the initiation or progression of breast cancer, whereas Eeles et al¹⁶ treated patients with estrogen. Mechanisms underlying the benefits of estrogen on OS and RFS require further investigation. Unknown from the data presented in Eeles et al¹⁶ is the OS and RFS in patients who continued the HRT compared with those who stopped.

HRT is not a benign intervention. Importantly, there were no differences in serious adverse events between the groups that received HRT and no HRT (12% v 16%). Development of breast cancer in patients receiving HRT who survived ovarian cancer is a major concern independent of patient BRCA1/2 mutational status. Eeles et al¹⁶ reported two cases of breast cancer in the HRT arm and one in the control arm. The total number of patients enrolled and the low incidence of breast cancer in the study do not allow conclusions to be drawn about a differential risk of breast cancer in the HRT arm. However, the Women’s Health Initiative data indicating that estrogen-only HRT does not increase the risk of breast cancer in healthy women suggest that HRT is unlikely to increase breast cancer in this population.¹¹

Eeles et al¹⁶ are to be congratulated for recognizing the importance of this study and observing the accrued cohort for nearly two decades. These findings, unlikely to be repeated, support the hypothesis that HRT improves outcomes in patients with ovarian cancer by reducing ovarian cancer relapse, ovarian cancer-specific death, and other causes of death. In the Women’s Health Initiative study, estrogen did not significantly affect all-cause mortality, making the decrease in mortality in the patients studied here of interest.¹⁰

Where do we stand in 2015? To treat or not to treat? Although there are many unanswered questions (eg, the mechanisms of how estrogen improves OS, the optimal duration of therapy), the data from Eeles et al¹⁶ combined with that of previous studies allow oncologists to feel comfortable offering patients HRT after the treatment of EOC to reduce vasomotor and other postmenopausal symptoms and should, therefore, improve the quality of life for patients with epithelial ovarian cancer.