Hormone replacement therapy and the endometrium

Endometrial effects of hormone replacement therapy

Common hormone replacement therapy (HRT) regimens are unopposed estrogen for women who have had a hysterectomy, and estrogen and progesterone in a sequential or continuous combined preparation for women with a uterus. The intention of a sequential progesterone regimen is to cause a regular and predictable withdrawal bleed that mimics normal menstruation. Continuous combined regimens of estrogen and progesterone are intended to keep the endometrium suppressed and to avoid withdrawal bleeding. Nonetheless, unscheduled vaginal bleeding is a common problem with HRT use and this causes dissatisfaction and concern for a woman and her physician alike. Irregular bleeding is a common reason for discontinuation of HRT and led to un-blinding of 40% of the study sample in a recent large study of menopausal women on HRT. ¹ Irregular bleeding is more likely to happen in the first year of use with continuous combined therapy than with sequential therapy. In the second year of use and beyond, however, irregular bleeding is more likely with sequential therapy. ²

Tibolone is a unique type of HRT that was initially introduced with the aim of leaving the endometrium unaffected while providing therapeutic benefit. ³ Tibolone is a steroidal prodrug that is metabolized into estrogenic, progestogenic and androgenic molecules. In the first six months of use, tibolone appears to have a more favourable bleeding profile than combined HRT, but this effect is lost with long-term use. ⁴

Irregular HRT-related bleeding is attributed to a number of different mechanisms that affect the endometrial microvasculature. ⁵ Disruption of microvessels is caused by altered production of angiogenic growth factors. This is compounded by progestogen-related changes in tissue factor secretion, which lead to altered vascular permeability. ⁶ Increased production of endometrial metalloproteinases also affects the integrity of capillary endothelial cells, ⁷ alongwith leukocyte infiltration of the endometrium under the chemotactic effect of progesterone. ^8–10

Abnormal bleeding can be difficult to evaluate based on a woman's symptoms alone, as it is subjective. Standardized methods for reporting bleeding have been suggested, which could make comparison of bleeding associated with different preparations across different studies more meaningful. ¹¹ More important clinical considerations are when to investigate abnormal bleeding and how to manage persistent bleeding. Common approaches are to investigate the bleeding using ultrasound scan (USS) and ‘blind’ endometrial biopsy or alternatively to perform hysteroscopy with directed biopsy. Although these investigations are standard they are invasive and expensive. Approximately half of all HRT users undergo at least one endometrial biopsy, USS or hysteroscopy, although usually no pathology is found. ¹²

Postmenopausal endometrial thickness, and its relation to endometrial pathology, has been extensively studied. Endometrial thickness can be variable in women on HRT, depending upon the type of preparation used. ¹³ In women using sequential HRT, the endometrial thickness varies depending on the phase of therapy whereas continuous combined HRT and tibolone cause endometrial atrophy. In women who are not using HRT, postmenopausal bleeding is rarely due to significant endometrial pathology when the endometrial thickness is 4–5 mm or less. ^11,14 HRT does not appear to reduce the sensitivity of USS for detecting endometrial carcinoma, although it does reduce specificity. ¹⁴ For a woman with unscheduled bleeding on sequential therapy, it has been recommended that ultrasonography be performed shortly after subsequent progesterone withdrawal bleeding, when the endometrium is expected to be at its thinnest. ¹⁵ For women taking continuous combined or tibolone therapy, the timing of the ultrasound is unimportant. Based on pre-test probability of a diagnosis of endometrial cancer in particular groups of women, the recommended threshold value for normal endometrial thickness in women taking sequential combined HRT is 5 mm. Below this level, the risk of endometrial cancer underlying any unscheduled bleeding is only 0.2% and endometrial biopsy is therefore unnecessary. ¹⁵ Women using continuous combined HRT have a lower pre-test probability of underlying endometrial cancer and therefore it has been suggested that the threshold endometrial thickness for triggering biopsy should be less (3 mm). ¹⁵ If unscheduled bleeding continues despite a reassuring USS, then further investigation, e.g. hysteroscopy with endometrial biopsy, should be considered. In the absence of pathology, common approaches to stop the problematic bleeding are to change the preparation or route of delivery of HRT. The progestogen-containing intrauterine system may be used to provide continuous endometrial suppression in an attempt to reduce the problem. Lower-dose preparations also have a more favourable side-effect profile, and these should be considered (see below). Ultimately a balance needs to be struck between the beneficial effects of HRT for the individual woman and the degree of inconvenience and concern regarding the bleeding.

HRT and endometrial cancer risk

The use of estrogen-only HRT is clearly associated with an increased risk of developing endometrioid-type endometrial adenocarcinoma. The risk is increased with both the dose of estrogen and the duration of use. ¹⁶ Unopposed estrogen stimulates the endometrium, causing increased proliferation. This increases the opportunity for genetic abnormalities to occur during cell division, leading in turn to an increased risk of endometrial hyperplasia and, subsequently, carcinoma. A Cochrane review of 30 randomized controlled trials concluded that, in contrast to estrogen-only HRT, women who take progestogen with estrogen are likely to reduce their risk of endometrial cancer. ¹⁷ Addition of progesterone acts by inducing secretory change, which prevents continued endometrial proliferation. Although progestogen use reduces the risk of developing significant endometrial pathology, use of a sequential estrogen–progesterone regimen does not completely eliminate the risk. The duration of progestogen use in each cycle is related to the degree of risk reduction with respect to development of endometrial cancer. Women who take progesterone for less than 10 days every month have a slightly reduced risk compared with women using estrogen-only preparations. ² In contrast, those regimens that include progesterone for more than 10 days every month confer no increased risk of endometrial cancer. ² Regimens that use progesterone in a long-cycle regimen, such as once every three months, are associated with a higher risk of endometrial hyperplasia. ¹⁷ The route of administration of estrogen also appears to be a factor in terms of endometrial safety. For example, studies show that vaginally administered estrogen does not increase the risk of endometrial carcinoma when compared with oral preparations. ¹⁸

More recent data from two large population studies (the Women's Health Initiative [WHI] and Million Women studies [MWS]) confirm earlier observations regarding HRT and risk of endometrial disease. ^19,20 The MWS conducted from 1996 to 2001 involved approximately 717,000 postmenopausal women without a previous history of cancer or hysterectomy and analysed a number of outcomes in relation to HRT use. These women were recruited and followed up for an average of 3.4 years. Compared with never users of HRT, the risk of endometrial cancer was reduced with the use of continuous combined HRT preparations, increased with the use of estrogen-only HRT and not significantly altered with the use of sequential combined preparations. Interestingly, body mass index significantly affected these associations, such that adverse events with estrogen-only HRT were greatest in non-obese women and the beneficial effects of continuous combined HRT were greatest in obese women. Overall, this study concluded that continuous combined therapy was associated with a small reduction in the risk of endometrial cancer compared with risk for never users of HRT, whereas in those women using sequential therapy, there was no significant difference. ²⁰

The WHI was launched in 1991 and in total involved approximately 162,000 healthy postmenopausal women in a set of clinical trials and an observational study. Two trials of hormone use were included: a study of combined estrogen–progesterone use in women with a uterus and an estrogen-alone study of women without a uterus. In both studies women were randomly assigned to receive either the hormone preparation or placebo. Women in these studies are now participating in a follow-up phase, which will last until 2010. The estrogen-alone study stopped in 2004. WHI investigators report that women randomized to combined estrogen and progesterone experienced a 19% decrease in endometrial cancer rates compared with women in the placebo arm.

The risks associated with the use of tibolone are less clear than those for other types of HRT. In the MWS, tibolone use was associated with an increased risk of endometrial cancer and this was related to duration of use. The risk was three-fold higher in women with normal weight, but was unchanged in obese women. Although a 79% increased risk of endometrial cancer was reported, when converted into absolute numbers the excess endometrial cancer risk in 1000 women over a five-year period is only three cases. ²⁰ Other studies have reported on the endometrial risks associated with tibolone use, although none have convincingly demonstrated a harmful effect of tibolone on the endometrium (Table 1). ^21–23 The validity of the data regarding tibolone in the MWS has been questioned. A UK General Practice database study indicates that women using tibolone were more likely to have used unopposed estrogen prior to starting tibolone, and it is possible therefore that the increased risk is due to prior use of unopposed estrogens. ²⁴

Newer low-dose preparations of HRT have been developed in order to alleviate menopausal symptoms while reducing unwanted adverse side-effects and endometrial cancer risk. Oral low-dose preparations (0.5 mg oral estradiol and 0.3 mg oral conjugated equine estrogens) and low dose transdermal (14 µg estradiol) preparations are available. The advent of the progestogen-containing intrauterine system also allows direct delivery of progestogen to the endometrium with good endometrial suppression and lower circulating levels than other routes, which may provide a consistent level of protection against the proliferative effects of estrogens. Low-dose HRT appears to be effective with an acceptable side-effect profile. A large randomized controlled trial of low-dose estradiol (0.5 mg) together with either norethisterone acetate 0.25 or 0.1 mg daily compared with placebo over six months showed that the rate of bleeding was similar to that experienced by the placebo group. ²⁵ In a study of approximately 400 women aged 60–80 years, unopposed low-dose transdermal estradiol (14 µg daily) over two years resulted in a small increase in bone density, with no difference in bleeding or endometrial stimulation compared with placebo. ²⁶ Other studies also demonstrated that symptoms can be controlled satisfactorily by the low-dose preparations ¹⁶ and show that low-dose preparations have less effect on breast and endometrial tissue. ²⁷ It has therefore been suggested that low-dose HRT preparations may not require progestogen for endometrial protection, but long-term safety data to support this approach are awaited.

Use of HRT following a diagnosis of endometrial cancer

Endometrial cancer is the fourth most common malignancy in women after breast, colorectal and lung cancers. Moreover, because it is associated with obesity, an increased incidence is expected in future, related to increasing rates of obesity. While endometrial cancer is most frequently diagnosed in postmenopausal women, 20–25% of cases occur premenopausally. Many women will therefore have an abrupt onset of menopausal symptoms following hysterectomy and bilateral oophorectomy and this can have a significant adverse effect on the quality of life following their cancer treatment. Consequently, it is necessary to consider whether these women may safely use HRT in order to treat symptoms. Because the most common form of endometrial cancer is estrogen related (endometrioid type), data sheets accompanying HRT preparations necessarily include endometrial cancer as a contraindication. Despite this, data from five studies have failed to demonstrate an increased risk of recurrence with the use of HRT after treatment for endometrial cancer. ^28–31 In contrast, a reduction in the rate of relapse, longer disease-free intervals and longer survival have been reported. It is likely that selection bias accounts for some of these associations. Where HRT is used for symptom relief after a diagnosis of endometrial cancer, this is most commonly in cases of early-stage, low-risk disease with complete surgical removal of the tumour, i.e. stage 1A and 1B, FIGO grades 1 and 2. The prognosis for most women with these types of disease is extremely good and therefore it is perhaps not surprising that these women have a low relapse rate and better survival. The only randomized controlled trial addressing the issue of use of HRT after treatment for endometrial cancer was inconclusive. ³² Where HRT is considered for use in women with a history of endometrial carcinoma, debate exists concerning the safest type of HRT preparation. In low-risk stage 1 disease, it seems unlikely that there is potential for inadvertent stimulation of residual occult tumour cells. Nonetheless, case reports of malignant transformation of endometriosis in women who had used unopposed estrogen HRT after treatment for endometrial cancer suggest that continuous combined HRT may be the safest choice in this context. ³³ The antidepressant venlafaxine is reasonably effective in treating ‘hot flushes’ and can be used as an alternative to HRT in women with severe symptoms, particularly where there may be related depressive symptoms.

Despite the limitations of available data, it appears safe to use short-term HRT for women with completely excised stages 1 and 2 endometrial carcinoma where quality of life is very adversely affected by menopausal symptoms. This requires a full discussion between the treating gynaecological oncologist and the woman regarding the theoretical risks and the aims and duration of treatment. It seems logical to use the lowest dose that successfully treats the symptoms and to use a continuous combined preparation for systemic symptoms. Given the good safety profile of vaginal preparations, these are probably a reasonable choice for women with problematic vaginal symptoms. ¹⁸ The safety of HRT in other women with a diagnosis of endometrial cancer is unknown and merits future research in this area.

HRT and hereditary endometrial cancer

Hereditary non-polyposis colorectal cancer syndrome (Lynch syndrome) is an autosomal dominant inherited cancer susceptibility syndrome, which is associated with a 60% lifetime risk of developing endometrial cancer as well as the well-documented association with development of colorectal cancer at a young age. Currently, prophylactic hysterectomy is the only proven effective preventive measure against the development of endometrial cancer in this group of women. Unfortunately a prospective multicentre trial of the use of the progestogen-containing intrauterine system was recently abandoned due to lack of recruitment. ³⁴ Therefore no data are available to support the effective use of intrauterine progestogens, either alone or in combination with estrogen-containing HRT in the prevention of hereditary non-polyposis colorectal cancer (HNPCC)-related endometrial cancers. About one in six endometrial cancers in women with HNPCC are diagnosed before the age of 40, and after careful discussion women known to have HNPCC may undergo hysterectomy premenopausally when their family is complete. Considerations regarding HRT use for symptom control, following treatment for endometrial cancer in women with HNPCC, are similar to those for other premenopausal women with endometrial cancer. There is unlikely to be any significant risk with respect to endometrial cancer following prophylactic hysterectomy and HRT is not contraindicated in this situation.

Conclusion

HRT is an effective treatment for menopause-related symptoms and osteoporosis. Unscheduled vaginal bleeding can be problematic for women and causes concern regarding possible endometrial pathology. There is an increased risk of endometrial cancer with estrogen-only preparations and these should not be prescribed for women who retain their uterus. Low-dose vaginal estrogens, however, have a good safety profile and can be prescribed for relief of urogenital symptoms without the need for a break in treatment or progestogen supplementation. Sequential combined HRT with at least 10 days of progesterone is not associated with an increased risk of endometrial cancer and continuous combined HRT is associated with a lower risk of endometrial cancer than non-users. Newer low-dose oral and transdermal preparations appear to offer effective symptom relief with a more acceptable side-effect and risk profile, although there is insufficient evidence at present to use these unopposed in women who retain a uterus. The progestogen-releasing intrauterine system may be used to alleviate problem bleeding on HRT, but at present it is not known whether this strategy is effective at preventing endometrial cancer in women on HRT. It seems reasonable to provide hormone replacement for very symptomatic premenopausal women who have had complete surgery for low-risk, early endometrial carcinoma. The theoretical risks should be weighed against the potential benefits for each individual woman. Women should be offered alternatives, and where HRT is prescribed, the lowest dose required to alleviate symptoms should be used. It may be beneficial to use continuous combined HRT for systemic symptoms in this group despite the absence of a uterus. Vaginal estrogens appear safe for urogenital symptoms. An evidence-based approach to the use of HRT in women with a history of treatment for endometrial cancer is required. For example, it is not known whether there is any risk associated with the use of HRT in women with cancers that are hormone receptor negative. More research is also needed to determine the impact of HRT use on quality of life and risk of disease recurrence for premenopausal women following surgery for early endometrial cancer.

Competing interests

None declared.