Position Statements

Hormone Replacement Therapy


Current clinical practice must continually be reevaluated in view of new evidence. The current clinical practice of oestrogen replacement therapy (ERT) and hormone replacement therapy (HRT) for the short-term alleviation of menopausal symptoms is unchallenged. Numerous randomized controlled trials (RCTs) agree on benefit, without any reason to expect unacceptable risk, if therapy is instituted shortly after the start of menopause and discontinued after 5 years.

The rationale of initiation of ERT or HRT after age 55, or continuation after 5 years, relies heavily on observational data indicating a risk reduction of 50% in coronary heart disease (CHD). The observational data has been consistent and backed by plausible biological explanations. As CHD is the main cause of death in postmenopausal women, the potential benefit of ERT/HRT outweighed the possible small increase in breast cancer as seen in observational studies. It has previously been proposed that the observational data may be flawed by a healthy users effect. This suspicion was supported by results of the HERS study, a RCT which failed to show secondary protection of HRT in elderly cardio-symptomatic patients after 4 years' use.

In order to assess the long-term benefits and risks of ERT/ HRT, the National Institute of Health (USA) launched the Women's Health Initiative (WHI). In a large RCT the effect of daily oral conjugated equine estrogen (CEE) 0.625 mg alone (ERT study arm), or daily oral continuous combined CEE and medroxy-progesterone acetate 2.5 mg (HRT study arm) were evaluated against placebo for the outcome of coronary heart disease (CHD), breast cancer, stroke, pulmonary embolism, hip fracture, colorectal cancer and death due to other causes.

On the 31 May 2002, the HRT arm was prematurely discontinued after 5.2 years, as the incidence of breast cancer in the HRT arm exceeded a predetermined "hazard" level. Furthermore, risk exceeded benefits. The ERT arm was not stopped. Publication of the results on 17 July caused great concern amongst patients and health care providers.(1)

It is recommended that all members of SASOG study the original paper in depth. The following discussion is aimed at providing some help in the interpretation. Members are encouraged to attend the 3 rd annual congress of the South African Menopause Society in Durban on 30/8-1/9/2002, where international experts will address the issue.

Outcome Number Relative risk(RR) Increased absolute risk per
10 000 w/y
Increased absolute benefit per
10 000 w/y
Myocardial infarction 286 1.29 7
Stroke 212 1.41 8
Breast cancer 290 1.26 8
Thrombo embolism 101 2.11 18
Hip fracture 106 0.66 5
Colon ca 112 0.63 6
Risk / benefit 41 11

W/y=women per year

Interpretation of results:


Conclusions are only valid for continuous combined HRT, and not for unopposed ERT.
Conclusions are not valid for other hormone regimes or different drugs.
Conclusions may not be valid for low dose HRT.

Breast cancer:

Although the trial was stopped primarily because the risk of incident breast cancer crossed a predetermined value, this is surprising as the increased incidence was consistent with available epidemiological studies (2). Breast cancer in patients on HRT is generally accepted as having a better prognosis than cases diagnosed in non-users. This may happen as a result of oestrogen promoting growth in pre-existing cancers and thus allowing earlier diagnosis and treatment. Tumours in the WHI study have not yet been analyzed in this regard. The results are consistent with two recent trials indicating risk only when progestin is added. The WHI results do not significantly change present thinking on breast cancer risk.

Cardiovascular Disease:

The lack of cardiovascular protection in a primary prevention study is the most significant finding of this study. As in HERS, the increased risk of cardiac events was mostly confined to the first year of treatment. Increases in thromboembolism though occurred during all 5 years of the study. As myocardial infarction is the main cause of death in menopausal patients, the WHI results are very significant for long-term use of HRT.

Fracture prevention

The WHI paper is the first trial with definitive data to support the ability of HRT to significantly prevent osteoporotic fractures. The magnitude of this effect will obviously be much greater in a population at high risk for fracture.

Bowel cancer

WHI is the first large study to support the beneficial effect as seen in observational studies


When assessing the risk/benefit ratio of HRT use, it is important to note that the WHI study did not include the following proven HRT benefits:
Alleviation of menopausal vasomotor symptoms.
Overall quality of life assessment.
Effects on cognitive function and brain dementia

Conclusion reached by the authors

In healthy asymptomatic postmenopausal women with an intact uterus, this particular regime of HRT should not be initiated or continued for the sole purpose of primary prevention of CHD.

Suggested clinical guidelines

It is important to understand that there is no need for panic. Patients must be assured that they can continue with their present medication, but be encouraged to consult their health care provider, when convenient, to assess their individual indication for HRT. This should be reassessed annually.

Although the risk and benefits of HRT given for the treatment of acute menopausal symptoms were not addressed in this study, there is still good reason to suspect that risk in the first 5 years after menopause is small. Undisputed scientific evidence supports the fact that only ERT/HRT is successful in the alleviation of vasomotor symptoms. There is no evidence to support the use of alternate medication such as phyto-oestrogens. It is concluded that present practice as regards the management of acute symptoms of menopause be unchanged.

A decision to continue with HRT after 5 years will have to be based on an individual assessment of risk and benefit. The WHI results make this task easier than before. The difficult part remains the value placed on the improvement of quality of life. This has not been quantified and will differ between individuals. For many patients at present, this constitutes the only reason why they persist with the use of HRT. The patient must understand that the absolute risks as found in the WHI study. It is expected that patients will be most concerned by the breast cancer results. It is important to be prepared to spend some time to explain that the increased diagnosis does not necessarily imply causality. Furthermore, neither in this nor in other studies has mortality been shown to be increased.

The suggestion not to use HRT for the sole purpose of prevention of CHD should be strictly observed. The beneficial effect of lifestyle changes on the prevention of CHD must be stressed. Optimal medical treatment of hypertension, diabetes and hypercholesterolemia must be assured.

There is evidence from observational data to suspect that the results of the ERT arm of WHI study may differ significantly. This issue will be resolved in 2005. In the meantime, it is recommended that the results of this WHI study and suggested guidelines be applied to ERT and HRT users. The medical profession must put pressure on companies providing other drugs and routes of administration of HRT to undertake appropriate clinical trials. Without such evidence, we simply have no other choice but to apply the results of the WHI study to all methods of prescribing ERT/HRT when assessing the need of individual patients. This is recommended not only on a scientific basis, but also from a medico-legal point of view.


  1. Writing group for the Women's Health Initiative Investigators. Risk and benefit of estrogen plus progestin in healthy postmenopausal women JAMA 2002; 288: 321-333.

  2. Collaborative group on hormonal factors in breast cancer. Brest cancer and hormone replacement therapy: collaborative re-analysis of data from 51 epidemiological studies of 52705 women with breast cancer and 108411 women without breast cancer. Lancet 1997; 350: 1047-1059.

Compiled By Dr Tobie de Villiers
Edited by Prof Gerhard Lindeque
August 2002