The trusted source for
healthcare information and
Is it Heresy or Good Medicine? HRT in Survivors of Gynecological Malignancies
By Robert L. Coleman, MD, Professor, University of Texas, M.D. Anderson Cancer Center, Houston, is Associate Editor for OB/GYN Clinical Alert.
Dr. Coleman reports no financial relationships to this field of study.
Synopsis: In most scenarios, hormone replacement therapy is safe in survivors of gynecological malignancies as measured by progression-free and overall survival. There are, however, notable exceptions and cautions, and usage should still follow the principles afforded to women without a gynecologic cancer history.
Source: Ibeanu O, et al. Hormone replacement therapy in gynecologic cancer survivors: Why not? Gynecol Oncol 2011;ePub April.
Although the annual incidence of all gynecological malignancies remains between 75,000 and 80,000, a far greater number of women with this history are survivors. Further, a sizable proportion of this cohort is pre- or peri-menopausal and report significant symptoms of estrogen deprivation either as a result of surgery or following prescribed adjuvant chemotherapy and/or radiotherapy. However, there is significant concern regarding the use of hormonal replacement therapy (HRT) in these women, as it may be intimated that therapy could increase the risk of recurrence or secondary malignancy. This concern is particularly evident in women carrying a germline mutation in BRCA1/BRCA2, women who have a personal history of breast cancer, and women with a history of endometrial cancer. In each of these situations, the foundation of apprehension comes from extrapolation of the results from randomized studies of HRT use in healthy women, where associations of increased risk of malignancy have been documented, as well as the noted impact of estrogen/hormones on new cancer development. Fortunately, a review of available (but limited prospective) data in gynecologic cancer cohorts combined with closer scrutiny of these randomized trials of HRT support the short-term safety of HRT in most patient cohorts who exhibit menopausal symptoms. As with women without a personal history of gynecologic cancer, HRT should be administered at the lowest effective dose for the shortest period of time necessary. In addition, symptomatic patients with a personal history of hormone-dependent breast cancer and hormone-expressing ovarian cancers should be managed with non-HRT strategies.
Up to 40% of women diagnosed annually with a gynecologic malignancy in the United States will be pre- or peri-menopausal.1 Since the principle tenet of therapy in most of these situations is to remove or sterilize incumbent disease, a large proportion of these women will be thrust artificially into an estrogen-deprived state, the symptoms of which are exacerbated by the abruptness of hormone fluctuation and the stress in dealing with the new diagnosis. Although HRT is effective in ameliorating symptoms from surgically induced menopause, there has been general concern in using HRT in women with gynecologic malignancy because of the theoretical potential of hastening an unknown natural disease history or stimulating quiescent metastatic disease.2 This concern was further elevated when initial data from randomized controlled trials of HRT use in healthy women documented increased risks of incident breast and ovarian cancer.3 Coupled with the well-described epidemiological and molecular relationship of unopposed estrogen use and the risk of endometrial cancer, most clinicians are reluctant to discuss or offer HRT in women with gynecological cancer. Despite further clarification in the type and duration of HRT exposure and these risks, general concern persists, particularly in light of the paucity of data of HRT use in women with a personal history of gynecologic cancers. However, observational and some prospective studies have been conducted in this setting and provide some guidance for clinicians trying to help women achieve the highest quality of life in dealing with their disease. A summary of the clinical scenario, the recommendation for short-term HRT use, and the level of evidence to support this recommendation is outlined in the Table.
|Table. Recommendations for short-term HRT use in various clinical scenarios|
|Clinical Situation||HRT Acceptable?||Level of Evidence|
|Endometrial cancer (Type I)||Yes – early-stage, low-risk disease after hysterectomy||Level I-III|
|Ovarian cancer||Yes – caveat: caution in women with tumors expressing ER/PR||Level I-III|
|Cervix cancer||Yes||Level II-III|
|BRCA gene mutation carriers||Yes – after RRSO ± hysterectomy AND no personal history of breast cancer||Level II|
|BRCA gene mutation carrier with a personal history of hormone-dependent breast cancer||No||Level II|
|Lynch II syndrome gene mutation carriers||Yes – but no data on safety and best considered if risk-reducing hysterectomy and BSO||None|
HRT: hormone replacement therapy; ER/PR: estrogen receptor/progesterone receptor; RRSO: risk-reducing salpingoophorectomy; BSO: bilateral salpingoophorectomy
Perhaps the most contentious primary site in regard to this debate is one where a strong relationship between estrogen use and development of cancer already exists. Indeed, years ago, incidence rates of epithelial endometrial cancer appeared to parallel the number of prescriptions written for unopposed conjugated estrogen HRT, highlighting the hormone-dependent nature of this cancer. In addition, progesterone and anti-estrogen therapy for advanced stage and metastatic endometrial cancer are still considered valid treatment options for hormone expressing tumors. So, it is not without merit that concerns for HRT exist for women with a history of endometrial cancer. Fortunately, this is also the disease site that the strongest safety data particularly in women with early-stage, low-grade disease. More than 25 years ago, Creasman and colleagues reported that HRT use in Stage I endometrial cancer was associated with a significant decrease in recurrence (2% vs 15%) compared to placebo in a non-randomized comparison.4 However, the strongest data stem from a randomized, placebo-controlled, double-blinded trial conducted by the Gynecologic Oncology Group, in which 1236 women with Stage I-II endometrial cancer were randomized to estrogen or placebo for a planned 3 years duration.5 Unfortunately, publication of the Women's Health Initiative (WHI) greatly influenced the enrollment to this trial, causing it to close prematurely (achieving 60% of its intended accrual). Nevertheless, after a median follow-up of 3 years, recurrence rates (2.3% vs 1.9%), progression-free survival (94.3% vs 95.6%), and overall survival (hazard ratio [HR] 1.27; 80% confidence interval [CI] 0.92-1.77) were all non-significant between the two arms. The ability to infer these results across other stages, histologies, and races is extremely limited and should only be undertaken with caution; however, it would appear that the risk, if present, is low, and short-term use in early-stage, low-risk patients can be considered.
The molecular relationship between estrogen and ovarian cancer is complex and dependent upon not only ligand-receptor interaction and subsequent downstream nuclear processes, but also non-genomic intracellular signaling, which incorporates multiple additional pathways to direct the cellular growth response and differentiation. Therefore, it is not surprising that epidemiological evidence of a strong link between HRT use and lifetime risk of ovarian cancer has been mixed, including a recent meta-analysis concluding the absence of an association between the two.6 However, there are three important considerations in this disease that require further mention: patients with a personal history of ovarian cancer, patients undergoing risk-reducing salpingoophorectomy (RRSO) due to BRCA mutation status, and patients with low-grade serous (or Type I) ovarian cancer.
HRT use in women with a personal history of ovarian cancer has been the subject of several observational and two small randomized controlled trials. In a 1999 report, Guidozzi and colleagues randomized 130 patients with high-grade, advanced stage ovarian cancer to either conjugated estrogen or placebo after surgical debulking and adjuvant chemotherapy.7 Patients who had taken HRT before their diagnosis were excluded. After a median follow-up of 4 years, there was no apparent increase in risk for recurrence or death due to HRT. Several observational trials have concluded the same, including in cohorts of women with low malignant potential tumors. These observations regarding the latter cohort are relevant as the age of diagnosis is substantially lower. The strength of these data would support the short-term use of HRT in symptomatic patients after appropriate counseling.
As has been discussed in OB/GYN Alert previously, women who carry a germline mutation in BRCA1 or BRCA2 are at substantial lifetime risk for the development of ovarian and breast cancer. Many, after appropriate counseling and testing, choose RRSO, as this has been shown to substantially decrease the risk of both cancers. However, most women undergoing the preventive intervention do so before menopause and are likely to exhibit intense menopausal symptoms, which may interfere with their quality of life. In light of the previously established association of HRT use and breast cancer, combined with the heightened lifetime risk of breast cancer due to BRCA mutation, it is not surprising concern exists for HRT use in this population. However, several studies would suggest the practice does not appear to increase this risk. In one prospective cohort study by Rebbeck et al, the reduction in subsequent breast cancer afforded by RRSO (HR 0.40; 95% CI 0.18-0.92) was not negatively influenced by the use of HRT after the procedure.8 The results, while reassuring, also have a molecular basis as it has been well-documented that BRCA mutation carriers (particularly BRCA1) who subsequently develop breast cancer have a "triple-negative" phenotype.9 That is, they don't express the estrogen receptor, progesterone receptor, and are Her-2-neu negative. Since data from the WHI suggested the progesterone component of HRT was associated with the risk for subsequent breast cancer, removal of the uterus at the time of RRSO has been endorsed to help simplify the strategy for HRT in these women. Nevertheless, BRCA mutation carrier patients with a personal history of hormone receptor positive breast cancer should not receive HRT as it appears the use is associated with a significantly increased risk of recurrence, contralateral, and metastatic disease.
Finally, it has become increasingly evident that a proportion of serous (and some endometrioid) ovarian cancers are driven by distinctly different molecular pathways (Type I ovarian cancer) relative to high-grade serous cancer, and may be responsive to progesterone and anti-estrogen based therapy. This is not unlike the situation for endometrial cancer; however, there are no data to support or refute HRT use in this setting and, as such, should be considered cautiously.
Although an uncommon disease in the United States, cervix cancer is frequently diagnosed in premenopausal women. Those with early-stage disease usually are offered surgical extirpation to allow for ovarian preservation and to avoid long-term effects of radiation on the genital track, which include vaginal shortening, loss of plasticity, vaginal dryness, dyspareunia, and menopausal symptoms. The ovaries are exquisitely radiosensitive and are sterilized at doses far lower than that required for microscopic tumor cell kill. As such, a common practice in women with an unknown or suspected risk for postoperative radiation who wish ovarian preservation is adnexal transposition (moving the ovaries out of the pelvic radiation field). However, for those receiving primary chemoradiation or in whom the ovaries were not transposed or failed following the procedure, HRT remains an important intervention to ameliorate menopausal symptoms and to treat the effects of ionizing radiation on the vaginal tissues. Immunohistological assessment for estrogen and progesterone receptors in squamous and adenocarcinomas of the cervix demonstrate their presence, but prognostically, this expression has not been linked to adverse outcomes. One prospective study followed 120 women with invasive cervix cancer treated with HRT (40 treated with estrogen alone, 40 with estrogen/progestin) or placebo (n = 40) after primary surgery or radiation.10 No adverse effect on recurrence or survival was observed with treatment. Although no prospective randomized trials have been conducted in this patient cohort, the risk to benefit ratio appears to favor use and should be considered.
Quality of life considerations are becoming an increasingly important aspect of survivorship in women with gynecologic malignancies, particularly as the cache of patients increases with effective primary treatment. As can be appreciated in this discussion, HRT use, once considered a contraindication to women with a cancer history, may be carefully considered. As is the case for women without this personal history, the lowest effective dose for the shortest period of time should be exercised in those deemed good candidates after careful consideration and counseling.