Oestrogen receptors and breast cancer

It is time for individualised treatment based on oestrogen receptor status

Breast cancer is marked by great clinical and biological diversity. Some women can be cured, while others die rapidly over just a few months. Systemic treatments can have gratifying and possibly curative effects for some, while for others they add only misery and financial hardship. Given this great heterogeneity, are women with breast cancer best served by therapeutic approaches emphasising homogeneity or ones that reflect the variability of the disease in individual choices of treatment? A case in point is the use of oestrogen receptor status in making decisions about endocrine therapy.

Oestrogen receptors are required for oestrogen stimulated growth and proliferation of breast cancer. They are found to some degree in 50-80% of breast tumours. Endocrine treatments are designed to antagonise the effects of oestrogen. Oophorectomy, luteinising hormone releasing hormone agonists, and aromatase inhibition all reduce the level of oestrogen, while antioestrogens such as tamoxifen competitively block binding of oestrogen to receptors and thus antagonise transcriptional activation of genes required for tumour growth.

Twenty five years ago it was first hypothesised that the response to such endocrine treatments would depend on the presence of oestrogen receptors in the tumours. And indeed, 50-60% of receptor positive metastatic tumours do respond.1 For tumours that test negative for oestrogen and progesterone receptors, the response rate (in terms of tumour size reduction or prolonged stability) is less than 5%.2 However, when less stringent cut off points are selected for defining receptor negativity, response rates increase, since even tumours with very low receptor positivity (between 4 and 10 fmol/mg protein) show a 20-30% response to treatment.3

Similarly, in the context of adjuvant endocrine treatment, oestrogen receptor status predicts the benefit of tamoxifen in preventing or delaying recurrence. In a meta-analysis of trials of adjuvant treatment there seemed to be a small benefit even in tumours with low concentrations of oestrogen receptors, but this was with a relatively high cut off point of 10 fmol/mg.4 This group would include women whose tumours had low concentrations of oestrogen receptors as well as those with receptor negative tumours, so that it is not surprising that there was some benefit. Similarly, our group recently found that if tumours contained even 1% of receptor positive cells by immunohistochemistry, there was evidence of significant benefit from adjuvant endocrine treatment.5 Individual trials using more stringent cut off points, such as a Swedish trial,6 find no benefit from adjuvant tamoxifen treatment in receptor negative tumours.

The apparent benefits of endocrine treatment in some “receptor negative” tumours have contributed to suggestions that all patients should receive tamoxifen independent of their oestrogen receptor status. But these apparent benefits depend to a considerable extent on the cut off point that was used. If cut off points are very low, then this suggestion would mean that 19 patients with receptor negative tumours would have to receive ineffective treatment for each one who benefits.

It could also be argued that tamoxifen should be given to all patients, including those with receptor negative tumours, because it has other benefits besides decreasing recurrence. These benefits might include decreasing the incidence of contralateral breast cancer,4 slowing postmenopausal bone loss,7 and improving the blood lipid profile.8 However, none of these potential benefits has much relevance to a woman with metastatic disease, and even a low level of side effects can erode or obliterate small benefits. Tamoxifen is generally well tolerated but is not completely without problems. Women receiving adjuvant tamoxifen treatment have an apparent increased risk of deep venous thromboses and pulmonary emboli of 1-2 patients/1000/year9 and a probable increased risk of endometrial cancer of about one patient/1000/year.10 Whether this is worth a decrease of two to three breast cancers per 1000 per year is debatable.

It has also been argued that determining oestrogen receptor status is too complicated for routine practice. There are, indeed, many assay methods available, but all clinically practical methods are based either on competitive binding of a labelled ligand or on the recognition of receptor protein by specific antibodies.11 The prototype ligand binding method is the dextran coated charcoal assay, in which radiolabelled oestradiol is allowed to bind to the receptor in a tumour cytosol and the amount bound is expressed in fmol/mg total cytosol protein. Antibody recognition of oestrogen receptors is most often assessed by immunohistochemistry, in which a histological section is stained by an antibody specifically directed against unique receptor epitopes and the percentage of cells with stained nuclei is recorded. The main advantages of the newer immunohistochemical method are that only tumour cells are assessed, so that results are not diluted by non-tumour tissue, and that it can be performed inexpensively on routine permanent sections with no particularly specialised equipment. Furthermore, simple immunohistochemistry seems to be at least as good as the ligand binding assay in predicting a better outcome for receptor positive tumours.5

A final point is that the economics of the indiscriminate use of tamoxifen are not favourable. In the United States a five year course of adjuvant tamoxifen costs about $4000 for the drug alone. To treat 100 patients for five years would therefore cost $400 000. The cost of immunohistochemical testing for oestrogen and progesterone receptors is between $50 and $100, so that testing 100 patients would cost at most $10 000. If 20% of these patients had tumours that were negative for both types of receptor and were not given tamoxifen, $80 000 would be saved, eight times the amount spent on the screening test. Though the cost of tamoxifen in Britain might be substantially less, the difference is unlikely to be enough to alter the balance of this financial equation. Thus, if screening for oestrogen and progesterone receptors could identify patients who have little or no chance of responding to tamoxifen, this would save resources and spare patients the morbidity of unnecessary treatment.

In summary, determining oestrogen and progesterone receptor status is easy and inexpensive, particularly by means of immunohistochemistry. When the cut off point is stringently set low and the assay is of high quality, patients with oestrogen and progesterone receptor negative tumours will experience little, if any, benefit from tamoxifen, especially as an adjuvant treatment. For metastatic disease, treating oestrogen receptor negative tumours with tamoxifen simply delays the institution of more appropriate treatment and increases the probability of patients experiencing tumour related side effects while waiting for a response that seldom occurs. It is time to move towards individualised management for women with breast cancer, making treatment more effective, timely, and cost efficient.