The British Menopause Society & Women’s Health Concern 2020 recommendations on hormone replacement therapy in menopausal women

Breast cancer

Breast cancer is the most commonly occurring female cancer in the UK and it is estimated that approximately 11,400 women die from breast cancer in the UK each year. Early diagnosis of breast cancer through screening and significant improvements in treatment have now resulted in the majority of women with breast cancer surviving their diagnosis.

Observational data from The Million Women Study (MWS) raised concerns over the long-term safety of HRT from the perspective of breast cancer.

Critique of the MWS has illustrated a number of key flaws in the study methodology and findings which limit the ability of the trial to establish a causal association between HRT and breast cancer.

The WHI estrogen and progestogen RCT reported a small increase in risk of breast cancer during the intervention phase after five years of usage of HRT of approximately one extra case per 1000 women per annum (HR 1.24; 95% CI 1.01–1.53), although this increase was no longer statistically significant when appropriate adjustment for confounding variables was made. In the early post-intervention phase (within 2.75 years from intervention), there was a sharp decrease in breast cancer risk in the combined arm and the risk became statistically insignificant (HR 1.23; 95% CI 0.90–1.70). However, during the late post-intervention phase (median 5.5 years post-intervention) a small increase in breast cancer risk was noted (HR 1.37; 95% CI 1.06–1.77).

In the WHI estrogen-alone trial, a small decrease in breast cancer risk was detected. The reduction in risk was not statistically significant during the intervention phase (HR 0.79; 95% CI 061–1.02). However, during the early post-intervention phase (within 2.75 years from intervention) the reduction in breast cancer risk in the estrogen alone arm became statistically significant (HR 0.55; 95% CI 0.34–0.89). The risk reduction subsequently became neutral in the late (median 5.5 years post intervention) post-intervention phase (HR 1.17; 95% CI 0.73–1.87).

The literature review from the NICE 2015 guideline on the diagnosis and management of the menopause concluded that HRT with estrogen alone is associated with little or no change in the risk of breast cancer, while HRT with estrogen and progestogen can be associated with an increase in the risk of breast cancer. It also concluded that any increase in the risk of breast cancer is related to treatment duration and reduces after stopping HRT.

A meta-analysis by the Collaborative Group on Hormonal Factors in Breast Cancer published in the Lancet in August 2019 reported on the risk of breast cancer with HRT in relation to the type and timing of hormonal intake. The review covered the period January 1992 to January 2018 and included information from 58 studies of which 24 were prospective. Prospective follow-up identified 108,647 postmenopausal women who developed breast cancer of which 55,575 (51%) had used HRT. The meta-analysis did not collect information on breast cancer mortality.

The meta-analysis noted a duration-dependent risk of increased breast cancer diagnosis with both unopposed estrogen and combined HRT, risk with the latter being greater. In addition, the risk of breast cancer associated with continuous combined preparations was higher than sequential preparations, although the difference in absolute risk was small.

In absolute numbers, for women taking HRT for five years the increased risk reported in the meta-analysis was as follows:

For women taking continuous combined HRT for five years from the age of 50, the risk of developing breast cancer between the age of 50 and 69 goes up by 1 extra case in 50 over 20 years from a background risk of 3 out of 50 women to 4 out of 50 women.

In past users of HRT, the risk of breast cancer was lower than in current users but the risk remained elevated more than 10 years after stopping. The increase, however, in absolute numbers was small.

The meta-analysis showed no dosage effect for estrogen on the risk of breast cancer with HRT and that vaginal estrogen exposure had no adverse impact on breast cancer risk with HRT.

There are a number of limitations that need to be taken into consideration when interpreting the findings from this meta-analysis. Over 40% of cases included in the meta-analysis were from the MWS, which had significant methodological limitations. In addition, data from the placebo-controlled WHI study were not included for comparison in the meta-analysis and only very small numbers of women on micronised progesterone were included. Data from the French E3N cohort study, which had demonstrated lower breast cancer risk in users of micronised progesterone compared to users of more androgenic progestogens, were excluded from the meta-analysis.

The British Menopause Society is of the view that the meta-analysis provides important additional information on the risk of breast cancer with HRT. The findings, however, are consistent with NICE recommendations and should not be interpreted in isolation. These need to be discussed in the context of the overall benefits and risks associated with HRT to help women make an informed choice.

The WHI long-term randomised clinical trials, published in JAMA 2020, reported on the association of HRT with breast cancer incidence and mortality and involved over 27,000 women, who were enrolled between 1993 and 1998 and followed-up through 2017. The report showed a significant decrease in the risk of breast cancer diagnosis (HR 0.78; 95% CI 0.65–0.93; p = 0.005) and a significant reduction in breast cancer mortality (HR 0.60; 95% CI 0.37–0.97; p = 0.04) when estrogen-only HRT is taken. Women who took combined estrogen and progestogen HRT had an increased risk of breast cancer compared to placebo (HR 1.28; 95% CI 1.13–1.45; p < 0.001), in keeping with the NICE guideline conclusions, but had no significant difference in breast cancer mortality compared with placebo (HR 1.35; 95% CI 0.94–1.95; p = 0.11). These important findings on breast cancer mortality contrast with the most recent follow-up data from the MWS.

Analysis of the WHI data assessed the effect of being overweight or obese on the risk of breast cancer. Women who had a body mass index (BMI) of over 35 had a significantly increased risk of invasive breast cancer compared with women of normal weight (HR 1.58; 95% CI 1.40–1.79). In addition, obesity was associated with an increase in estrogen receptor-positive and progesterone receptor-positive breast cancers (HR 1.86; 95% CI 1.60–2.17), an increase in advanced diseased (HR 2.12; 95% CI 1.67–2.69) and breast cancer mortality (HR 2.11; 95% CI 1.57–2.84) compared with women of normal body weight.

The 2019 meta-analysis by the Collaborative Group on Hormonal Factors in Breast Cancer reported a modifying effect for obesity. The analysis noted an increase in the incidence of estrogen receptor positive breast cancer with increasing BMI in postmenopausal who had never used HRT. This effect was not noted in women who had been using HRT for some years. The review noted that in obese women, use of estrogen only HRT added little to their already elevated breast cancer risk related to the adiposity-associated estrogenic stimulation of their breast tissue.

However, lean and obese women had a similar increase in breast cancer risk by addition of a progestogen.

The authors concluded that obesity attenuated the absolute and the relative excess breast cancer risk associated with both unopposed and combined HRT.

Fournier et al. (2014) reported updated figures from the E3N Cohort, a large observational French study that included 3678 invasive breast cancers between 1992 and 2008 among 78,353 women. HRT regimens that included estrogen and micronised progesterone or dydrogesterone were not associated with an increased risk of invasive breast cancer with short-term use up to five years (HR 1.11; 95% CI 0.89–1.38). Long-term use (more than 5 years) was associated with a slight increase in the risk of breast cancer (HR 1.31; 95% CI 1.15–1.48), but this risk was no longer statistically significant following discontinuation of HRT (HR 1.15; 95% CI 0.93–1.42).

HRT regimens that included estrogen and a progestogen other than micronised progesterone or dydrogesterone had a slightly elevated breast cancer risk with short-term use up to five years (HR 1.70; 95% CI 1.50–1.91) and with long-term use for more than five years (HR 2.02; 95% CI 1.81–2.26). A slight ongoing increase (HR 1.36; 95% CI 1.13–1.64) was also noted following discontinuation of HRT in this group.

In summary, results from the E3N French cohort study suggest:

Estrogens combined with micronised progesterone do not increase breast cancer risk for up to five years of use.

Large observational trial data from the E3N Cohort and the Finnish Cancer Registry have reported no difference in the risk of invasive breast cancer with oral versus transdermal administration of estradiol. In addition, data from the Finnish Cancer Registry have suggested a similar risk of breast cancer with HRT regimens using the levonorgestrel intrauterine system to that noted with regimens using oral progestogens. The latter findings, as well as the effect of dose, duration of exposure and type of regimen require further evaluation in adequately powered prospective studies.

Women who are carriers of the BRCA1 and BRCA2 gene mutation who have undergone risk-reducing BSO, can receive add-back HRT and this has not been shown to diminish the risk-reducing benefit of BSO on subsequent risk of breast cancer diagnosis. A meta-analysis and recent systematic review on this showed no increase in the risk of breast cancer in women with BRCA1 and BRCA2 mutations using HRT after risk-reducing BSO.

In summary, current evidence suggests that estrogen alone HRT is associated with little or no change in the risk of breast cancer while combined HRT can be associated with an increased risk which appears duration dependent and may vary with the type of progestogen used. However, this risk is low in both medical and statistical terms, particularly compared to other modifiable risk factors (e.g. obesity), and should be taken in the context of the overall benefits obtained from using HRT. Large observational trial data suggest that micronised progesterone and dydrogesterone are likely to be associated with a lower risk of invasive breast cancer compared to that noted with other progestogens.

Endometrial cancer

Unopposed estrogen therapy increases the incidence of endometrial cancer and this risk is largely avoided by the use of combined estrogen and progestogen therapy.

Long-term use of sequential combined HRT for more than five years may be associated with a small increase in risk of endometrial cancer, with risk being inversely proportional to the number of days progestogen is given.

The WHI estrogen and progestogen study reported a neutral effect on the risk of endometrial cancer with HRT compared to placebo during the intervention phase after five years of usage of HRT (HR 0.83; 95% CI 0.49–1.40). However, a significant reduction was noted with combined estrogen and progestogen intake compared to placebo in the post-intervention phase (HR 0.58; 95% CI 0.40–0.86) and with long-term cumulative follow-up (HR 0.67; 95% CI 0.49–0.91).

Ovarian cancer

Observational data have suggested an increased risk of ovarian cancer with HRT use.

The WHI estrogen and progestogen trial was the only randomised placebo-controlled trial which studied the incidence of ovarian cancer in women taking combined equine estrogens plus medroxyprogesterone acetate and concluded that over long-term follow-up there was no evidence of increased risk (HR 1.24; 95% CI 0.83–1.87).

Data analysis from the Danish National Cancer Registry revealed a significant increase in the incidence of serous invasive ovarian tumours following eight years use of estrogen alone (IRR 1.7; 95% CI 1.4–2.1) and combined estrogen and progestogen therapy (IRR 1.6; 95% CI 1.4–1.9).

A meta-analysis by the Collaborative Group on Epidemiological Studies of Ovarian Cancer included individual data from 52 epidemiological studies, in which approximately half the postmenopausal women with ovarian cancer had used HRT. Ovarian cancer risk was significantly increased in current users receiving up to five years of HRT (RR 1.43; 95% CI 1.31–1.56). In past users, the risk decreased the longer the duration of time after discontinuation of HRT. However, the ongoing risk remained slightly elevated (HR 1.37; 95% CI 1.29–1.46). The risk did not differ significantly between users of estrogen alone and combined estrogen and progestogen preparations. In addition, the increased risk was only noted for serous and endometrioid cancers. The meta-analysis concluded that women who used HRT for five years starting approximately at the age of 50 years had an additional risk of developing ovarian cancer of approximately one extra case per 1000 users (which equates to one extra case per 5000 women per year) and a risk of having one extra death related to ovarian cancer per 1700 users.

There are a number of limitations that need to be taken into consideration when interpreting the findings, including heterogeneity of the data, differences in study protocols and proportions of women lost to follow up in these studies.

More recently, Liu et al. (2019) reported a meta-analysis on the association of HRT with the risk of ovarian cancer by histological subtype. A total of 36 observational studies involving 4,229,061 participants were included in the meta-analysis and the review suggested a small increase in the risk of ovarian cancer with HRT (RR 1.29; 95% CI 1.19–1.40). Association differed between histological subtypes with an increased risk noted for serous (RR 1.50; 95% CI 1.35–1.68) and endometrioid (RR 1.48; 95% CI 1.13–1.94) cancers.

In summary, there may be a slight increase in the risk of developing serous and endometrioid ovarian cancer associated with HRT use. However, this risk is small in both medical and statistical terms and should be taken in the context of the overall benefits and risks balance for the individual woman.

Cervical cancer

While there is a known association between the combined oral contraceptive pill use and cervical cancer, there is no association between cervical cancer and HRT. The WHI study found no evidence for an increase in the risk of cervical cancer with HRT (HR 1.44; 95% CI 0.47–4.42).

The European Prospective Investigation into Cancer and Nutrition (EPIC) Study (Roura et al., 2016) included a cohort of 308,036 women of which 261 women had invasive cervical cancer with a median follow-up of nine years and 804 women had CIN 3/carcinoma in situ. Ever use of HRT was associated with a statistically significant reduction in the risk of invasive cervical cancer (HR 0.50; 95% CI 0.40–0.80).

Colorectal cancer

Published data suggest a reduced risk of colorectal cancer with the use of oral combined HRT.

The WHI trial showed that the risk of colorectal cancer was reduced in the combined estrogen and progestogen arm (HR 0.62; 95% CI 0.43–0.89), though no effect was shown in the estrogen alone group (HR 1.15; 95% CI 0.81–1.64). The risk reduction of combined estrogen and progestogen reduced after stopping HRT over 13 years cumulative follow-up (HR 0.80; 95% CI 0.63–1.01).

There are no data on transdermal HRT and the risk of colorectal cancer.

Breast cancer

The evidence on the risk of recurrence of breast cancer with the use of HRT is inconclusive as the number of breast cancer events in published studies is too small for definitive conclusions to be made. In addition, some of the analyses in the published literature were not based on a priori hypothesis.

A randomised, non-placebo-controlled Scandinavian RCT HABITS – Holmberg et al. (2004, 2008) was terminated early after two years of follow-up as a significantly increased number of new breast cancer cases was noted in the HRT arm of the trial. The HABITS trial was initiated in 1997 and a total of 447 women were randomly assigned. Most women in the HRT arm received continuous combined or sequential estradiol and norethisterone.

The HABITS steering committee terminated the study in December 2003, when preliminary results based on a median follow-up of 2.1 years showed a significantly increased risk of breast cancer recurrence in the HRT arm of the trial (HR 3.5; 95% CI 1.5–7.4). A total of 442 women were followed up for a median of four years. Thirty-nine of 221 women in the HRT arm and 17 of 221 women in the control arm experienced a new breast cancer event (HR 2.4; 95% CI 1.3–4.2).

The authors concluded that after extended follow-up, there was a clinically and statistically significant increased risk of a new breast cancer event in survivors who took HRT.

The Stockholm trial was an open randomised trial that was initiated in 1997. A total of 188 women with a history of breast cancer were randomised to HRT, while 190 women were randomised to no HRT. The trial was prematurely stopped in 2003 when the HABITS trial findings, described above, were reported.

The Stockholm trial found no evidence for excess risk of breast cancer recurrence with HRT after a median follow-up of 4.1 years at the end date in January 2004 (HR 0.82, 95% CI 0.35–1.9). Long-term follow-up with a median of 10.8 years did not find evidence for any difference in new breast cancer events (60 in the HRT group vs. 48 in the control group (HR 1.3; 95% CI 0.9–1.9). However, there was a significantly higher number of contralateral breast cancers (14 cases) in the HRT group compared to the control group (4 cases; HR 3.6; 95% CI 1.2–10.9). The authors concluded that it was uncertain whether these contralateral tumours should be regarded as a recurrence of the primary cancer or as a new primary malignancy. This finding is based on a very small number of events and there is no biological explanation for this discrepancy. It raises the issue to whether this outcome is due to the small number of breast cancer events.

There were a number of variations in the design of the HABITS and Stockholm trials that may account for the different outcomes noted. It has been suggested that the increased risk of recurrence in the HABITS trial might be attributed to greater progestogen exposure. However, the numbers for the different subgroups are too small to draw meaningful conclusions regarding the risk of recurrence associated with different progestogen regimens. In addition, the proportion of lymph node-positive patients was higher in the HABITS trial (26%) compared to the Stockholm trial (16%) and a greater percentage of women in the Stockholm trial were treated with adjuvant tamoxifen (52%) than in the HABITS trial (21%). However, subgroup analyses by the HABITS study group for use of tamoxifen and nodal status did not show a significant association, although the authors acknowledged that their subgroup analysis lacked sufficient power due to the small numbers included to confirm this conclusion.

The Livial Intervention following Breast cancer: Efficacy, Recurrence, And Tolerability Endpoints (LIBERATE) was placebo-controlled double-blind randomised trial that assessed the safety and efficacy of tibolone in breast cancer patients. A total of 3098 women were included in the intention to treat analysis (1556 in the tibolone group and 1542 in the placebo group). The trial was ended prematurely in 2007, as interim analysis showed an overall increased risk of breast cancer recurrence in the tibolone group. After a median follow-up of 3.1 years, 237/1556 (15.2%) women in the tibolone arm had cancer recurrence, compared with 165/1542 (10.7%) in the placebo group (HR 1.40; 95% CI 1.14–1.70).

Based on current evidence, a history of breast cancer should be considered a contraindication to systemic HRT. Non-hormonal options should be considered for the management of menopausal symptoms in women with breast cancer.

Women with ongoing symptoms who fail to respond to non-hormonal management should be referred to discuss their options with their oncology team and menopause specialist to allow an individualised plan based on the woman’s own circumstances.

Non-hormonal options should be the first-line treatment in women with a history of breast cancer, particularly those receiving tamoxifen or aromatase inhibitors.

Women with breast cancer who do not respond to non-hormonal treatment may consider vaginal estrogens. Off-label use of vaginal estrogen therapy can be considered in women with a history of hormone sensitive malignancies but the advantages and disadvantages of each case should be weighed up carefully and discussed with the woman’s oncology team and menopause specialist.

Women on aromatase inhibitors who wish to use vaginal estrogen treatment should consider switching their adjuvant therapy to tamoxifen given that the mode of action of tamoxifen is through estrogen receptor antagonism while aromatase inhibitors exert their effect by lowering total estrogen levels.

Ospemifene is contraindicated in women undergoing active treatment for breast cancer. However, it may be used once treatment, including adjuvant therapy is complete although there is limited clinical evidence assessing this.

Additional research is needed to assess vaginal DHEA use in cancer survivors.

Endometrial cancer

Studies assessing the use of HRT following treatment for endometrial cancer have either shown no increased risk of recurrence or a reduced recurrence rate with an increased disease-free interval.

A meta-analysis by Shim et al. (2014) included one randomised trial and five observational studies. A total of 896 women with a history of endometrial cancer who used HRT were included as well as 1079 non-users. The meta-analysis showed a decrease in recurrence in women taking HRT compared to the control group (OR 0.53; 95% CI 0.30–0.96).

HRT was initiated 1 to 60 months (on average 3–12 months) after surgery. Most of these studies have been on early stage disease and the findings may be different in advanced cancer where there may be microscopic metastatic deposits.

A Korean population study including 5667 women was conducted in 2019. This found no evidence for risk of endometrial cancer in women taking either estrogen alone (HR 0.78; 95% CI 0.31–1.96) or estrogen plus progestogen (HR 0.57; 95% CI 0.21–1.57). However, although the study accounted for several important variables a key confounding variable, the stage of disease, could not be included in the analysis. This limits the reliability of the study’s findings.

Endometrial sarcomas are estrogen sensitive and should be considered a contraindication to HRT.

Ovarian cancer

There is no evidence that estrogen therapy following treatment for ovarian cancer will adversely affect the prognosis.

Studies have either shown no difference in survival rates or an improvement in survival rates with the use of HRT in women with epithelial ovarian cancer.

Li et al. (2015) reported a systematic review and meta-analysis on postoperative HRT following epithelial ovarian cancer. The review included two RCTs and four cohort studies and included 419 women with epithelial ovarian cancer who used HRT and 1029 women with epithelial ovarian cancer who did not use HRT. The review concluded that postoperative HRT does not have a negative effect on overall survival (HR 0.68; 95% CI 0.54–0.86) and tumour recurrence (RR 0.83; 95% CI 0.64–1.07). The authors, however, acknowledged that there were methodological limitations in the included studies and recommended the need for well-designed and large-scale RCTs to further assess this.

Saeaib et al. (2020) reported a Cochrane review (three RCTs, involving 350 women) on HRT after surgery for epithelial ovarian cancer. The review concluded that HRT may slightly improve overall survival in women who have undergone surgical treatment for epithelial ovarian cancer, but the quality of the evidence was low (HR 0.71; 95% CI 0.54–0.93).

There is no evidence assessing the effect of HRT on women with germ cell tumours. However, a small subset of germ cell tumours may be hormone sensitive and in this subset HRT should be avoided on theoretical grounds.

There are no data on the use of HRT following granulosa cell tumours though HRT should be avoided in this situation largely on theoretical grounds.

Ongoing hormone receptor studies on ovarian cancers may help predict the risk of recurrence.

Cervical cancer

HRT is not contraindicated after treatment for squamous cell carcinoma of the cervix or adenocarcinoma of the cervix.

Ploch (1987) reported on the use of HRT in cervical cancer survivors. Eighty women with cervical cancer received HRT, while 40 women with cervical cancer were used as controls. Overall, no significant difference in recurrence rate or survival was noted between the two groups.

Vaginal and vulval cancer

Most vaginal and vulval cancers are squamous cell carcinomas and behave similarly to squamous cell carcinoma of the cervix and there is no evidence of an adverse effect with regard to recurrence of vulval disease. Thus, HRT would not be contraindicated and systemic and topical HRT can be used following vaginal and vulval carcinoma.

Thromboprophylaxis in women on HRT who are admitted to hospital and women undergoing elective surgery

Hospitalisation for surgery or with medical illness must also prompt risk assessment for VTE. If women using HRT are admitted to hospital they should receive thromboprophylaxis as appropriate and do not need to discontinue their HRT.

If elective surgery is planned, women require risk assessment. HRT can be continued, whether transdermal or oral, as long as suitable thromboprophylaxis is given. Women on HRT containing oral estradiol who have additional risk factors for VTE can switch to transdermal estradiol intake.

Premature ovarian insufficiency

POI has been estimated to affect about 1% of women under the age of 40, 0.1% under 30 and 0.01% of women under the age of 20. However, as cure rates of cancers in young women continue to improve, it is likely that the incidence of iatrogenic prematurely menopausal women will rise.

HRT is strongly recommended in these young women to control menopausal symptoms, maintain sexual function as well as to minimise the risk of cardiovascular disease, osteoporosis, and possibly reduce the risk of cognitive impairment associated with POI.

The majority of women with POI (84–86%) will experience menopausal symptoms while approximately 40–50% will experience symptoms related to urogenital atrophy. Menopausal symptoms experienced by women with POI may vary in intensity and can be intermittent due to the fluctuation in ovarian activity.

Women with POI who do not experience menopausal symptoms would still be advised to consider hormone replacement for the prevention of the long-term sequelae of the condition.

Hormone replacement in POI simply replaces ovarian hormones that would normally be produced at this age.

The aim is to replace hormones as close to physiological levels as possible.

Hormone therapy should generally be continued at least until the estimated age of natural menopause (on average 51 years).

HRT is also important to preserve uterine function in women planning ovum donation.

Five to ten per cent of women with POI achieve a spontaneous pregnancy.

HRT and the combined contraceptive pill containing ethinyl estradiol would both be suitable options for hormone replacement. However, HRT may be more beneficial in improving bone health and cardiovascular risk compared to the combined oral contraceptive pill. Data from two small randomised trials have shown significantly greater improvement in bone density with HRT compared to that noted with the combined contraceptive pill as well as significantly lower mean systolic and diastolic blood pressure. Plasma angiotensin II and serum creatinine were reduced without alteration of plasma aldosterone concentrations with HRT compared with the combined contraceptive pill.

It is well recognised that young women with premature menopause will potentially be at an increased risk of osteoporosis, cardiovascular disease and cognitive impairment if adequate hormonal support is not used.

The WHI study findings do not apply to this young group as this study did not include women under the age of 50.

It is of paramount importance that the patients understand this in view of recent media reports on HRT.

There is an ongoing need to standardise terminology and to determine the causes and scale of the problem through a global registry (e.g. https://poiregistry.net). Good quality observational and RCT data will facilitate the refinement of evidence-based guidelines (e.g. ESHRE Guideline, 2015) which will optimise the management of POI.

POI and its health and fertility implications are discussed in further detail in separate consensus statement produced by the British Menopause Society and Women’s Health Concern.

Early menopause

‘Early menopause’ refers to onset of the menopause between the age of 40 and up to 45 years of age. This group of women has similar long-term risks related to estrogen deficiency as those experienced by women with POI. Rocca et al. (2006) showed that mortality was significantly increased in women who had bilateral oophorectomy before the age of 45 years compared to control women (HR 1·67; 95% CI 1.16–2.40). Increased mortality was noted in women who had not received estrogen replacement up to the age of 45 years. In clinical practice, therefore, both groups (<40 and 40–45) are advised similarly regarding the bone and cardiovascular protective effects of sex steroid hormone replacement and should consider hormone replacement at least until the natural age of the menopause of 51 in the absence of a contraindication.

The meta-analysis by the Collaborative Group on Hormonal Factors in Breast Cancer (2019) reported that the use of HRT in postmenopausal women younger than 50 increases the risk of breast cancer diagnosis which contradicts previous evidence and advice to date. However, the control group in the meta-analysis was age-matched postmenopausal women rather than age-matched normally cycling women which would have provided a more accurate comparison (given that early menopause reduces breast cancer risk). In addition, the number of women in this subgroup was relatively small. These findings need to be taken in the context of the significant bone protective effects and cardiovascular benefits that HRT offers to younger postmenopausal women. Given the limitations above, the current recommendation is that if years of HRT exposure in women with POI/early menopause are to be counted, it should still be from the age of natural menopause of 51 years.

Routes and regimens

Transdermal (gels, patches or spray) and subcutaneous (implants) administration of estradiol avoid the first pass effect through the liver and do not alter the coagulation cascade in the same way that oral estrogens do. Laboratory data have shown a neutral impact on thrombin generation, the coagulation cascade and pro-inflammatory markers with transdermal administration of estradiol. In addition, data from large observational studies have shown that transdermal administration of estradiol is unlikely to increase the risk of VTE or stroke above that of controls and has a lower risk than that which occurs with oral estradiol.

Non-hysterectomised women require progestogen (administered for 12–14 days in a sequential regimen and daily in a continuous combined regimen) to minimise the risk of endometrial hyperplasia and endometrial cancer associated with unopposed estrogen exposure.

Intrauterine and vaginal routes of progestogen administration, such as the levonorgestrel releasing intrauterine system and progesterone gel and pessaries, provide adequate endometrial protection with reduced systemic side-effects.

The 52 mg levonorgestrel releasing intrauterine system provides adequate endometrial protection in women receiving estrogen therapy. Systemic side-effects are reduced though not completely eliminated. The impact on breast cancer risk remains unclear with preliminary data from the Finnish cancer registry showing no significant difference when compared to oral progestogens.

Continuous combined regimens avoid the need for regular withdrawal bleeds but may be associated with continuous low-grade progestogenic side effects. Unscheduled bleeding is higher with continuous combined HRT regimens compared to that with sequential regimens.

Ultra-low dose continuous combined estradiol and progestogen regimens (e.g. 0.5 mg estradiol in combination with dydrogesterone 2.5 mg) appear to maintain the benefits of higher dose regimens whilst allowing minimal use of progestogen to reduce side-effects.

Low-dose vaginal estrogenic creams, rings, tablets and pessaries should be considered for women with symptoms of urogenital atrophy and can be used in conjunction with oral/transdermal HRT. It is estimated that up to 20% of women do not obtain satisfactory control of their VVA symptoms when systemic HRT is used without topical vaginal estrogen treatment.

Long-term usage of topical vaginal estrogen treatment is commonly required as symptoms often return when treatment is discontinued and progestogenic opposition is not required as systemic absorption is minimal with low-dose estradiol and estriol vaginal preparations.

Conventionally regulated (non-compounded) bioidentical (rBHRT) estradiol, progesterone and testosterone are produced from plant extracts and are similar to their biological equivalents in the body. They may have some advantages over non-identical varieties of HRT (e.g. ethinyl estradiol, synthetic progestogens) in terms of VTE and breast cancer risks. Recent case-controlled data indicate that the use of micronised progesterone is associated with a lower risk of VTE and breast cancer compared to that noted with other synthetic progestogens.

Compounded bioidentical hormone replacement therapies (cBHRT) are manufactured as creams, lozenges and vaginal preparations by ‘Specialist Pharmacies’ which have proliferated both physically and online in the UK and abroad. These cBHRT products:

Do not follow the same regulatory pathway of evaluation by the MHRA as conventional pharmaceutical products;

The regulatory view is that compounding can only be justified when a medicine has to be created because the strength, concentration, or dosage form that is required for a specific patient is not commercially available.

Unscheduled bleeding on HRT

Persistent unscheduled bleeding beyond 4–6 months from commencing HRT warrants investigation with ultrasound scan and/or endometrial biopsy.

For the majority of women with unscheduled bleeding on HRT, modifying progestogen intake would often control the bleeding especially in women who experience unscheduled bleeding in the first few months after commencing HRT.

Progestogen intake could be modified as follows:

For continuous combined HRT regimens, the dose of progestogen could be increased (e.g. increase micronised progesterone daily dose from 100 mg to 200 mg daily on continuous basis), particularly when combined with higher dose estrogenic regimens. Those on continuous combined HRT regimens that contain a progestogen in a combined preparation or have the levonorgestrel intrauterine system, could have micronised progesterone/medroxyprogesterone acetate or norethisterone added to their HRT regimen. If they continue to experience ongoing unscheduled bleeding, the HRT regimen could be changed to a cyclical intake of progestogen.

For cyclical HRT regimens, the dose of progestogen could be increased (e.g. micronised progesterone 300 mg for 12 days a month instead of 200 mg) or increase duration of progestogen intake (can take progestogen for 14 days a month or for 21 days out of a 28-day HRT intake cycle).

Women who continue to have unscheduled bleeding beyond six months despite modifying their progestogen intake or where there is a concern about the clinical presentation or bleeding amount/pattern should have pelvic ultrasound scan assessment of the endometrial cavity and an endometrial biopsy where appropriate or assessment of the endometrial cavity by hysteroscopy where clinically indicated and an endometrial biopsy.

If breakthrough bleeding occurs following the switch to continuous combined HRT and does not settle after three to six months, then the woman can be switched back to a sequential regimen for at least another year.

The risk of endometrial cancer in women with unscheduled bleeding on HRT is significantly lower than that with postmenopausal bleeding in women not on HRT especially in women who had not been experiencing bleeding before commencing HRT and who are taking progestogen.

Sexual function/androgens

While there is an age related decline in sexual function including libido, arousal, orgasm and satisfaction, there is a significant decline around the time of the menopause.

Women with distressing low sexual desire and tiredness should be counselled that androgen supplementation is an option particularly if HRT in the form of adequate levels of estrogen with or without progesterone has not been effective.

Assessment of serum androgen levels is unlikely to be beneficial in making the diagnosis of hormone dependent low sexual desire, as there is poor correlation between circulating androgen levels and clinical symptoms.

However, best practice as recommended by the Global Consensus Position Statement on the Use of Testosterone Therapy for Women is that testosterone levels should be checked to exclude high baseline levels and to prevent subsequent supraphysiological replacement.

Assessment of total testosterone levels (ideally by mass spectrometry techniques) should take into account sex hormone binding globulin (SHBG) levels. Free Androgen Index = total testosterone × 100/SHBG. Levels should be kept within the female physiological threshold (1.5–10%).

There are few licensed female androgenic options available globally even though there are accumulating data for efficacy and safety.

Testosterone implants and patches were withdrawn by pharmaceutical companies for commercial, not safety reasons.

Tibolone has a weak androgenic effect which can have a beneficial effect on mood and libido.

One per cent testosterone gel is available in 50 mg, 5 mL sachets and 2% testosterone gel in 60 mL canisters for male use. Off-license prescribing by specialists is an option for female androgen replacement, at a reduced dosage of 0.5 mL/day, or one pump on alternate days, both equating to a daily dose of 5 mg/day. Alternate day administration, however, is likely to result in fluctuations in serum levels.

A 1% testosterone cream (0.5 mL/day) with an indication for female use is manufactured in Australia and is only available privately at present in the UK.

Recent systematic reviews and meta-analyses by Achilli et al. (2017) and Islam et al. (2019) have shown that androgenic side-effects and risks are minimal and reversible if testosterone levels are maintained within the female physiological range.

Some studies have shown benefits on the skeleton, cognition, well-being and the vagina, although these findings require further assessment.

Other options such as oral DHEA require further research to confirm their efficacy and safety. However, vaginal DHEA pessaries have recently been licensed for the treatment of VVA and may have some benefits for low libido. The latter, however, requires further evaluation in adequately powered randomised studies.

Lifestyle/Complementary and alternative therapies (CAMs)/Alternatives to HRT

Optimisation of diet and lifestyle advice should be incorporated into the routine management of all women in the menopause transition and beyond.

This should include advice on bone and cardiovascular health and information on adequate calcium and vitamin D intake, exercise, smoking cessation as well as avoidance of excessive alcohol intake.

Vaginal bioadhesive moisturisers are a more physiological way of replacing vaginal secretions than vaginal gels such as KY. They are hydrophilic and rehydrate vaginal tissues, providing a reasonable alternative to vaginal estrogen. Lubricants should have similar osmolality and pH to that of physiological vaginal secretions.

Pharmacological alternatives

A meta-analysis of 10 RCTs has shown a marginal benefit of clonidine over placebo in the control of menopausal VMS.

A significant amount of evidence exists for the efficacy of selective serotonin reuptake inhibitors such as fluoxetine and paroxetine in treating VMS.

The most convincing data are for the serotonin and norepinephrine reuptake inhibitors (venlafaxine) in a dose of 37.5 mg bd in cancer survivors. However, the most common side-effect, nausea, limits the usefulness of this agent. Fluoxetine and paroxetine should be avoided in tamoxifen users as they interfere with its metabolism and reduce its efficacy.

Low dose paroxetine was licensed in the USA by the FDA in 2013 for the treatment of VMS; systematic review and meta-analysis of moderate evidence support its efficacy.

Small studies with the anti-epileptic drug gabapentin have shown effectiveness for hot flush reduction. A recent meta-analysis has confirmed superior efficacy compared to placebo for treating hot flushes. Its use is limited by side-effects such as drowsiness and somnolence, particularly at high doses. A stepwise increase in dosage by 300 mg per week up to a maximum of 1.2 g is advised to minimise the side-effects.

An RCT, yet to be repeated, has demonstrated good efficacy for extended release oxybutyinin in the treatment of VMS. Side-effects such as dry mouth were problematic in some patients as expected.

Preliminary studies of a novel treatment using neurokinin 1 and 3 receptor (NK1/3R) antagonism have been reported. These treatments appear to be effective in the treatment of hot flushes and may be particularly helpful to women such as those with breast cancer in whom estrogen replacement is contraindicated.

Recent phase 2 clinical trials have demonstrated that neurokinin receptor antagonists are highly effective and well tolerated in alleviating VMS. A randomised, double-blind, placebo-controlled, phase 2a clinical trial in 87 postmenopausal women showed that NK3R 90 mg taken twice a day for 12 weeks of treatment significantly reduced moderate/severe VMS compared to placebo. A phase 2b, randomised, placebo-controlled, double-blind, dose-finding study, reported that all NK3R doses except the lowest dose (15 mg BD) resulted in significant reduction of moderate/severe VMS compared to placebo.

At the time of writing there are no licensed options available. Further assessment in larger phase 3 trials and for longer durations is needed before these agents become available as an option in standard clinical practice.

In summary, published literature shows a marginal benefit for non-hormonal interventions over placebo but they are likely to be less effective than HRT in controlling menopausal symptoms. However, non-hormonal interventions may be of help in women who have a contraindication to receiving HRT or who do not wish to take HRT.

Phytoestrogens and herbal remedies

A recent large survey in the UK has shown that the prevalence of CAM usage including phytoestrogens in the UK is around 25% (1 in 4 women).

Data from some of the better researched phytoestrogen containing preparations appear to demonstrate some benefits, not only for symptom relief, but also on the skeleton and cardiovascular system.

A network meta-analysis undertaken by the NICE menopause guideline group showed that St John’s wort and some isoflavone preparations may be effective for VMS but more research is required to confirm efficacy.

A model based meta-analysis of soy isoflavones showed a 25% reduction in hot flushes after elimination of the placebo effect.

There are as yet no hard data on major outcome measures such as coronary heart disease and fractures or long-term endometrial safety.

If a herbal treatment is chosen patients should look for the traditional herbal remedy marking, validating strength and quality.

Acupuncture

A meta-analysis of the effects of acupuncture on frequency and severity of hot flushes, menopause-related symptoms, and quality of life in women in natural menopause showed a significant reduction in hot flushes and sweats. However, none of the studies included a sham control.

A double-blind, placebo (sham)-controlled randomised clinical trial determined the effects of acupuncture on menopausal hot flushes compared with sham therapy. The study found no evidence of a beneficial effect of acupuncture on VMS compared with sham therapy.

Further research is needed into the effects of acupuncture on VMS before it can be considered a more effective therapy than placebo.

Cognitive behavioural therapy

CBT impacts on both VMS perception and reduction in stress and wellbeing, sleep problems and vasomotor symptomatology.

The efficacy and feasibility of a guided, internet-based CBT programme in alleviating or reducing the severity of menopausal symptoms were evaluated in breast cancer survivors. Eighty-three per cent reported feeling more control over their symptoms after completing the programme. Larger randomised trials are underway to assess the efficacy of the programme.

A fact sheet on the Women’s Health Concern website provides guidance on CBT in a self-help format for women.

Laser for VVA

CO₂ and erbium lasers have been used for the treatment of VVA. This is a novel treatment, but emerging evidence suggests that this may have a place in the treatment of symptoms of VVA.

A double-blind placebo RCT by Cruz et al. (2018) included 45 participants and compared the efficacy of fractional CO₂ vaginal laser treatment to local estriol therapy and the combination of both in the treatment of VVA in postmenopausal women over a 20-week-period. All treatment arms resulted in improvement in VVA symptoms, vaginal health index scores and vaginal cytology analyses. The laser group showed significant worsening in the pain domain of the Female Sexual Function Index. These findings require further evaluation in larger randomised studies and over a longer duration of time.

A systematic review and meta-analysis by Pitsouni et al. (2017) assessed the efficacy of laser therapy for treating VVA. The meta-analysis included 14 studies and 542 participants with a follow-up period between 1 and 18 months. All VVA symptoms including dryness/dyspareunia/itching/burning/dysuria/urgency/frequency and urinary frequency decreased significantly and consistently in all the included studies. The authors concluded that laser therapy appears to be a promising and safe non-pharmaceutical therapeutic option for managing VVA symptoms although the quality of the evidence assessed was low. These findings require further assessment in adequately powered RCTs.