represents something of a landmark in the assessment of cathepsin D as a prognostic marker in breast cancer. It demonstrates unequivocally that cathepsin D is an independent marker of poor prognosis in all groups of breast cancer patients.Cathepsin D is an aspartyl protease that is normally localized within lysosomes and is involved in protein catabolism and tissue remodelling (Westley and May, 1996). For reasons that are not clear, but probably as a result of intracellular misrouting, this protease can be secreted. The secretion of cathepsin D is elevated in some cancer cells, and it has been suggested that the illicit secretion of a protease whose activity is normally restrained to the intracellular degradation of proteins may facilitate the invasion and spread of cancer cells.Cathepsin D synthesis and secretion is regulated by oestrogen in oestrogen-responsive breast cancer cells (Westley and Rochefort, 1980); however, the reason why cathepsin D is regulated by oestrogen remains obscure. Given its role in protein catabolism, it may be involved in tissue involution in oestrogen-responsive tissues, and examples that immediately spring to mind are the involution of ductal and alveolar structures in the breast following the cessation of lactation and the post-partum involution of the uterus. The regulation of cathepsin D expression by oestrogen results from the interaction of the oestrogen receptor with oestrogen response elements in the cathepsin D promoter. The classical model of oestrogen action that involves the interaction of receptor homodimers with a palindromic oestrogen response element of the type found in the Xenopus laevis vitellogenin gene promoter does not seem to apply to cathepsin D.Cathepsin D expression originally became of clinical interest in breast cancer as a potential marker of oestrogen responsiveness. It was hoped that it would prove more reliable than existing markers such as the oestrogen receptor; however, the promise of cathepsin D as a marker of oestrogen responsiveness has not been fulfilled. Interest in this protease has continued as a result of reports that it is a marker of poor prognosis in node-negative breast cancer patients (Spyratos et al, 1989;Foekens et al, 1996). The value of identifying markers of poor prognosis in node-negative breast cancer is that we need to be able to identify those women in this good-prognosis group who are at significant risk of relapse and who would be likely to benefit from a more aggressive treatment regimen.Research into the prognostic value of cathepsin D has been facilitated by the availability of a radiometric immunoassay that has been validated by the EORTC Breast Cancer Cooperative Group (Benraad et al, 1992). Although this assay is expensive to perform, it measures the amount of cathepsin D in tumour cytosols prepared for oestrogen receptor analysis with impressively low inter-assay variation. It should be noted that at least three other methods have been used to measure cathepsin D expression in prognostic studies, namely Western transfe...