Before prescribing hormone-replacement therapy, healthcare professionals should consider carefully the potential benefits and risks for every woman
Article date: September 2007
This webpage was reviewed in November 2015 and remains an accurate reflection of the current regulatory position.
Since the publication of information about hormone-replacement therapy (HRT) in Current Problems in Pharmacovigilance in October, 2004,1 important new evidence has become available that affects prescribing advice.
HRT effectively relieves vasomotor symptoms. In most cases, 2–3 years’ therapy is sufficient, but some women may need longer—this judgement should be made on a case-by-case basis with regular attempts to discontinue. Symptoms may recur for a short time after stopping HRT.
For all women, the lowest effective dose should be used for the shortest time.
Coronary heart disease (CHD)
Randomised controlled trials have found an increased risk of CHD in women who started combined (oestrogen-progestogen) therapy more than 10 years after menopause.2, 3 Very few randomised controlled trials have assessed younger, newly menopausal women, and some have suggested a lower relative risk in these women compared with older women. The low baseline risk of CHD in most younger women, and the very low attributable risk due to HRT, means that their overall CHD risk is likely to be low. No increased risk of CHD with use of oestrogen-only HRT has been identified to date. 2 Importantly, there are no data from randomised controlled trials to suggest a cardiovascular benefit with oestrogen-only or combined HRT.
Healthcare professionals should assess carefully every woman’s risk of CHD before prescribing HRT, irrespective of her age or time since menopause.
In randomised controlled trials, oestrogen-only and combined HRT increased the risk of stroke (mostly ischaemic) compared with placebo.4, 5 Although the increase in relative risk seems to be similar irrespective of age, 2 baseline risk of stroke increases with age and therefore older women have a greater absolute risk. Limited observational data suggest that this risk may depend on oestrogen dose.6, 7
Venous thromboembolism (VTE)
Oral HRT has been associated with an increased risk of VTE (ie, deep vein thrombosis or pulmonary embolism) in randomised controlled trials and observational studies. Evidence suggests that risk is higher with combined HRT than with oestrogen-only HRT, and that these events are more likely in the first year of use.8
The level of risk associated with other routes of administration has not been clearly established, although it may be lower with transdermal HRT.9
In women with a uterus, use of oestrogen-only HRT substantially increases the risk of endometrial hyperplasia and carcinoma in a way that depends on dose and duration.10 Addition of progestogen cyclically for at least 10 days per 28-day cycle greatly reduces the risk, and addition of progestogen every day eliminates the risk. 11
The risk of breast cancer is increased in women who take HRT for several years:
- Combined HRT has been associated with the highest risk
- For oestrogen-only HRT, risk is lower than with combined HRT.12 Some studies have not shown an increased risk for oestrogen-only HRT 13
- Risk increases with duration of use and returns to baseline within a few years of stopping treatment
HRT, especially combined therapy, may increase mammographic density, which may adversely affect radiological detection of breast cancer. In the Women’s Health Initiative trial,13 14 conjugated equine oestrogens (CEE) and CEE plus medroxyprogesterone increased the likelihood of having an abnormal mammogram that needed further evaluation.
Observational studies suggest that long-term use of oestrogen-only or combined HRT may be associated with a small increased risk of ovarian cancer, which returns to baseline a few years after stopping treatment.15 16
HRT is effective for prevention of osteoporosis, but its beneficial effect on bone diminishes soon after stopping treatment.
Because of the risks associated with long-term use, HRT should be used for prevention of osteoporosis only in women who are unable to use other medicines that are authorised for this purpose.
Advice for healthcare professionals to consider before prescribing HRT:
- The decision to prescribe HRT should be based on a thorough evaluation of the potential benefits and potential risks of treatment
- Healthcare professionals should assess every woman’s overall risk, including cardiovascular risk, particularly in those older than 60 years who have increased baseline risk of serious adverse events
- Evidence for the risks of HRT in women who had premature menopause is limited. However, the baseline risk of adverse events in these younger women is low, and the balance of benefits and risks may be more favourable than in older women
|Age range (years)||Time (years)||Background incidence per 1000 women in Europe*||Oestrogen-only HRT||Oestrogenprogestogen HRT|
|Additional cases per 1000 HRT users†||Risk ratio (95% CI)‡||Additional cases per 1000 HRT users†||Risk ratio (95% CI)‡|
|Cancer risk Breast§|
|50–59||5||10||2 (1-4)||1·2 (1·1–1·4)||6 (5–7)||1·6 (1·5–1·7)|
|60–69||5||15||3 (2–6)||9 (8–11)|
|50–59||10||20||6 (4–10)||1·3 (1·2–1·5)||24 (20–28)||2·2 (2·0–2·4)|
|60–69||10||30||9 (6–15)||36 (30–42)|
|50–59||5||2||4 (3–5)||3·0 (2·5–3·6)||NS||1·0 (0·8–1·2)|
|50–59||10||4||32 (21–48)||9·0 (6·3–12·9)||NS||1·1 (0·9–1·2)|
|50–59||5||2||<1||1·1 (1·0–1·3)||<1||1·1 (1·0–1·3)|
|50–59||10||4||1 (1–2)||1·3 (1·1–1·5)||1 (1–2)||1·3 (1·1–1·5)|
|60–69||10||6||2 (1–3)||2 (1–3)|
|Cardiovascular risk Venous thromboembolism (VTE)|
|50–59||5||5||2 (0–4)||1·3 (1·0–1·7)||7 (5–10)||2·3 (1·8–3·0)|
|60–69||5||8||2 (0–6)||10 (7–16)|
|50–59||5||4||1 (1–2)||1·3 (1·1–1·4)||1 (1–2)||1·3 (1·1–1·4)|
|60–69||10||9||3 (1–4)||3 (1–4)|
|Coronary heart disease (CHD)|
|50–59||5||14||9||NS||0·6 (0·4–1·1)||NS||1·3 (0·8–2·1)|
|60–69||5||31||18||NS||0·9 (0·7–1·2)||NS||1·0 (0·7–1·4)|
|70–79||5||44||29||NS||1·1 (0·8–1·5)||15 (1–32)||1·5 (1·0–2·1)|
|Benefits** Colorectal cancer|
|50–59||5||6||3||NS||0·9 (0·7–1·1)||NS||0·9 (0·7–1·1)|
|Fracture of femur|
|50–59||5||0·5||1·5||0||0·6 (0·4–0·9)||NS||0·7 (0·5–1·0)|
|60–69||5||5·5||5·5||–2 (–3 to –1)||NS|
Table: Risks and benefits of HRT
- Background incidence from: Hospital Admissions in England (HES) for stroke and VTE; WHI trial for CHD; the International Agency for Research on Cancer (IARC) for ovarian cancer and endometrial cancer; and from never-users in the Million Women Study for breast cancer.
† Best estimate and range based on relative risk and 95% CI.
‡ Risk ratios and 95% CI from: meta-analyses of randomised controlled trials (RCTs) for stroke; meta-analyses of RCTs and observational studies for VTE, endometrial cancer, and ovarian cancer; meta-analysis of RCTs and observational studies in Europe only for breast cancer; and from Women’s Health Initiative (WHI) trial for CHD.
§ European studies have generally identified higher breast-cancer risk than North American studies and may be due to differences in prevalence of obesity.
|Progestogen added for 10 days or more per 28-day cycle.|
¶ Estimates from placebo groups from the conjugated equine oestrogens (CEE) and CEE plus medroxyprogesterone placebo arms of WHI trial.2
** Menopausal symptom relief is not included in this table, but is a key benefit of HRT and will play a major part in the decision to prescribe HRT.
NS non-significant difference.
Article citation: Drug Safety Update September 2007, vol 1 issue 2: 2.
Hormone replacement therapy (HRT)
Rossouw JE. JAMA 2007; 297: 1465–77 ↩
Grady D, et al. JAMA 2002; 288: 49–57 ↩
Hendrix SL, et al. Circulation 2006; 113: 2425–34 ↩
Wassertheil-Smoller S, et al. JAMA 2003; 289: 2673–84 ↩
Grodstein F, et al. Ann Intern Med 2000; 133: 933–41 ↩
Lemaitre RN. Arch Intern Med 2006; 166: 399–404 ↩
Peverill RE. Best Pract Res Clin Endocrinol Metab 2003; 17: 149–64 ↩
Scarabin P-Y. Lancet 2003; 362: 428–32 ↩
Grady D, et al. Obs Gynae 1995; 85: 304–13 ↩
Million Women Study Collaborators. Lancet 2005; 365: 1543–51 ↩
Shah NR, et al. Menopause 2005; 12: 668–78 ↩
Stefanick ML. JAMA 2006; 295: 1647–57 ↩
Anderson GL. Maturitas 2006; 55: 103–15 ↩
Beral V. Lancet 2007; 369: 1703–10 ↩
Danforth KN. Br J Cancer 2007; 96: 151–56 ↩