No abstract

In 1974 a project was initiated in which patients with apparently operable breast carcinoma underwent special Investigations in an attempt to identify occult metastases in bone and liver. One hundred and seventy‐two patients have been followed for two years or more and their axillary node histological findings reviewed. From analysis of these patients the following conclusions may be drawn. Careful clinical assessment, especially measurement of the primary tumour and palpation of the axillary nodes, remains a fundamental guide to the likelihood of early recurrence. Histological confirmation of node metastases is as valuable as clinical assessment, but is not superior unless other features of node histology are considered. Micrometastases do not increase the probability of early recurrence except in comparison with a group whose nodes lack both metastases and other unfavourable features. The majority of special Investigations do not have additional predictive value, though a bone scan is useful in the small number (less than 10%) in whom it is positive.

Evidence is presented to support the view that biopsy of the pectoral and axillary tail nodes obtained at simple (total) mastectomy gives a reliable estimate of axillary node status. The technique is described.

Summary.-When large human breast cancers were assayed for oestrogen receptors at multiple sites, 5-fold differences were found in the numbers of oestrogen receptors, between the site within a tumour. This may result from variations in the cell: stroma ratio from site to site. Such differences could be significant when receptor levels in the tumour are low (<50 fmol oestradiol bound mg cytosol protein) since the classification distinction between hormone-sensitive and hormone-insensitive breast cancers is based upon numbers of oestrogen receptors detected by the assay. This problem might be remedied by assessment of the cell:stroma ratio in all assayed tumours, and by the combination of the cytoplasmic oestrogen-receptor assay with other hormone-receptor assays.

Summary.-A study instigated by the British Breast Group and involving 3 centres (Edinburgh, Glasgow and Liverpool) was carried out to compare 3 methods for the estimation of urinary hydroxyproline. No significant difference between the first and the second 24 h urine collection was found for each measure of urinary hydroxyproline, within laboratories and within patient groups. Reliable hydroxyproline studies can, therefore, be performed on one 24 h urine collection.The Grant and Ellis/Goldberg methods gave comparable results and the excretion of hydroxyproline in the urine measured by either of these 2 methods could be used to distinguish cases of breast cancer with osseous involvement (as demonstrated by X-rays) from those without. The Hypronosticon Kit method was found to be unreliable as it had 29.4% false negatives in breast-cancer patients with X-ray demonstrable metastases.The incidence of elevated urinary hydroxyproline excretion in breast-cancer patients with negative X-rays was 11/14 (25%), 5/34 (15%) and 8/43 (19%) for the Ellis/Goldberg, Hypronosticon and Grant methods respectively. No conclusion can be drawn regarding the outcome of this group of patients because of the short period of follow-up.