Loss of heterozygosity at 17p13.3-ter, distal to TP53, correlates with negative hormonal phenotype in sporadic breast cancer

Genetic alterations on chromosome 17p are frequent in a variety of human malignancies such as sporadic breast carcinomas. The clinico-pathological significance of it remains to be elucidated. The purpose of this study was to explore whether the allelic loss (LOH) in 17p13.3, which is suggested by the presence of tumour suppressor genes, independent of TP53, are related to current clinico-pathological criteria for characterising breast cancer. A group of sporadic breast carcinomas, with no alteration in TP53 locus, were analysed for the presence of LOH in D17S34 and D17S30/5 loci, mapped to the 17p13.3 region, distinct from and telomeric to TP53. LOH by at least one marker was observed in 13 of 47 informative cases (27.6%). Clinicopathological parameters such as age, menopausal status, histological type, tumour size, nodal status, grading, ploidy, labelling index, S-phase fraction and hormonal phenotype, were evaluated. LOH at distal 17p13.3 significantly correlated with the absence of oestrogen receptors (ER) (P=0.018), progesterone receptors (PgR) (P=0.009) and concordant absence of either ER or PgR (P=0.006). This may provide a basis for speculation as to the function of the putative tumour suppressor genes in 17p13-ter in sporadic breast cancer.     

Relationship of PS2 with response to tamoxifen therapy in patients with recurrent breast cancer.

PS2, an oestrogen-inducible protein, was measured in the cytosol of 230 primary tumours from patients who were subjected to first-line tamoxifen therapy for advanced disease without prior adjuvant therapy with tamoxifen. PS2 correlated positively with oestrogen receptor (ER, P < 0.01) and progesterone receptor content (PgR, P < 0.001), and with the length of progression-free survival (PFS, P = 0.05). Although not statistically significant, higher levels of PS2 (> or = 10 ng mg-1 protein) were also associated with increased probability of response to tamoxifen treatment and a longer total post-relapse survival (PRS). ER, PgR, menopausal status, site of disease and prior adjuvant chemotherapy were all associated with response to tamoxifen therapy and with PFS. In multivariate analysis for PFS, low levels of ER and PgR, visceral metastasis, a disease-free interval of less than 1 year and prior adjuvant chemotherapy were all significantly associated with an increased probability of a rapid disease progression after start of tamoxifen therapy. In the subset of 83 tumours with intermediate levels of ER and PgR (both > or = 10, but not both > or = 75 fmol mg-1 protein), PS2 was positively related with the length of PFS (P < 0.01) and PRS (P < 0.05). PS2 remained the strongest factor in multivariate analysis for PFS (P < 0.01) in this ER/PgR intermediate subgroup, but was not of predictive value in univariate or multivariate analysis for both PFS and PRS in tumours classified as ER/PgR low or high (> or = 75 fmol mg-1 protein). It is concluded that PS2 status may be used as a parameter, additional to ER and PgR, for better refinement of prediction of response to tamoxifen treatment in advanced breast cancer patients especially with intermediate ER/PgR levels in their primary tumour.     

Correlation of allelic losses and clinicopathological factors in primary breast cancers

Human breast cancers frequently show allelic loss or loss of heterozygosity (LOH) at specific chromosomal regions. To understand the possible role of these genetic alterations in tumor development and progression, we examined LOH at loci on chromosomal arms lp, 3p, 11p, 13q, 16q, 18q, and 22q in 140 to 246 cases of primary breast cancers and compared it with lymph node metastasis, histological type, tumor stage, estrogen receptor (ER) and progesterone receptor (PgR) status. LOH at 1_p22-31 correlated with lymph node metastasis and a tumor size of greater than 2 cm. LOH at 13ql2-14 and 18q21 were most frequently observed in tumors of the solid-tubular type. LOH at 1p34-36 was more frequent in tumors of the scirrhous and solidtubular types than in other less aggressive histological types. Furthermore, a significant association was observed between LOH at 3pl4-21, 11p11-15 and 13ql2-14 and the absence of progesterone receptors. These results suggest that some clinical characteristics of breast cancers are determined by loss of tumor suppressor genes present at specific chromosomal regions.     

Clinical impact of detection of loss of heterozygosity of BRCA1 and BRCA2 markers in sporadic breast cancer.

The development of familial and sporadic breast cancer is based on genetic alterations of tumour-suppressor genes, for which loss of heterozygosity (LOH) is one mechanism of gene inactivation. To investigate LOH of BRCA1 (17q21) and BRCA2 (13-q12-13) in sporadic breast cancer, polymerase chain reaction (PCR)-based fluorescent DNA technology for detection of microsatellite polymorphisms was applied. A total of 137 breast cancer and 15 benign breast specimens with matched normal tissue were examined. Fluorescent-labelled PCR products were analysed in an automated DNA sequencer (ALFTM Pharmacia). Losses at both loci were correlated with different histological types, age, tumour size, lymph node status, grading and steroid hormone receptor expression, [SHR: oestrogen receptor (ER), progesterone receptor (PgR)]. For BRCA1 (D17S855, THRA1, D17S579) losses could be detected in invasive ductal carcinoma (IDC; n = 108) in 32-38%, invasive lobular carcinoma (ILC; n = 19) in 21-42% depending on the marker applied, but not in benign breast tumours (n = 15). Losses of BRCA1 markers correlated with larger tumour size, higher grade, and PgR expression. For BRCA2 (D13S260, D13S267, D13S171) losses could be detected in 108 IDCs in 30-38%, in 19 ILCs in 17-39% depending on the marker applied, but not in benign breast tumours. Losses of BRCA2 markers correlated only with higher grade. Microsatellite analyses combined with detection of fluorescent-labelled PCR products by an automated laser DNA sequencer can be used for routine determination of LOH. In sporadic breast cancer, LOH of BRCA1 of BRCA2 does not add decisive prognostic value as stated for familial breast cancer.     

Detection of loss of heterozygosity at RAD51, RAD52, RAD54 and BRCA1 and BRCA2 loci in breast cancer: pathological correlations.

Loss of heterozygosity (LOH) in loci of the 15q15.1, 12p13, 1p32, 17q21 and 13q12-13 regions may collaborate in the inactivation of RAD51, RAD52, RAD54, BRCA1, BRCA2 and possibly other genes implicated in the repair of double-stranded DNA and in DNA recombination. We investigate allelic losses in microsatellites of the RAD51, RAD52, RAD54, BRCA1 and BRCA2 regions, and their correlations with nine pathologic parameters in 127 breast carcinomas. The LOH analysis was performed by amplifying DNA by PCR, using 15 markers of the 15q15.1, 12p13.3, 1p32, 17q21 and 13q12-13 regions. LOH was found in the RAD51 region in 32% of tumours, in the RAD52 region in 16%, in RAD54 in 20% and in the BRCA1 and BRCA2 regions in 49% and 44% respectively. Significant correlations between one or more regions with concomitant LOH and pathologic parameters were observed with respect to age (P = 0.008), oestrogen receptor content (P = 0.03), progesterone receptors (P = 0.003), higher grade (P = 0.001), more advanced stage (P = 0.004) and peritumoural vessel involvement (P < 0.0001). The number of cases in which LOH was observed simultaneously in two or more regions was always higher than expected on the basis of their statistical probability, and curiously, the three patients with LOH at five regions concomitantly were under the age of 30 years. These results suggest that LOH at these regions could be related to breast cancer, and probably to a poor tumour prognosis.     

Allelotype of squamous cell carcinoma of the head and neck: fractional allele loss correlates with survival.

Allelic imbalance or loss of heterozygosity (LOH) studies have been used extensively to identify regions on chromosomes that may contain putative tumour-suppressor genes. We have undertaken an extensive allelotype of 80 specimens of squamous cell carcinoma of the head and neck (SCCHN) using 145 polymorphic microsatellite markers on 39 chromosome arms. Allelic imbalances were found most frequently on chromosome arms 3p, 9p, 17p and 18q with over 45% LOH and imbalances on 1p, 1q, 2p, 5q, 6p, 6q, 8p, 8q, 9q, 11q, 13q, 17q and 19q were found in more than 20% of SCCHN. These LOH data were analysed against a range of clinicopathological parameters which included previously untreated and previously treated tumours; correlations were found between LOH on 9q and nodes at pathology (P = 0.02) and between histopathological grade and LOH on 12q (P = 0.02) and 13q (P = 0.01). In the group of previously untreated tumours, a correlation was found between site of tumour and LOH on 3p (P = 0.019), and 8p (P = 0.029), while TNM staging correlated with LOH on 3p (P = 0.019) and 17p (P = 0.016). Fractional allele loss (FAL) was calculated for 52 tumours with LOH data on nine or more chromosomal arms and found to have a median value of 0.22 (range 0.0-0.80). Correlations were found between FAL > median value and nodes at pathology (P = 0.01) and tumour grade (P = 0.06), demonstrating that advanced tumours with lymph node metastasis often had LOH at multiple sites. FAL > median value was found to correlate with a poor survival (P < 0.03) and, furthermore, FAL > median value correlated with poor survival in the previously untreated patients (P < 0.019). These results indicate that assessment of the accumulation of genetic damage, as provided by allelotype data, provides a useful molecular indicator of the tumour behaviour and clinical outcome.     

Close correlation between progesterone receptor concentration and hormonal sensitivity in DMBA-induced mammary tumours of the rat

Cyclic administration of oestrogen alone and with progesterone to ovariectomized rats bearing DMBA-induced tumours resulted in more uniform cytosol oestrogen and progesterone receptor (ER and PgR) concentrations in the uteri and a stronger correlation between ER and PgR concentrations in the tumours than in a group of naturally cycling animals. Using this model, the concentration of PgR was a better indicator of tumour response to ovariectomy and hormonal restimulation than the concentration of ER. In ten unequivocally responsive adenocarcinomas receptor values, expressed as fmol/mg cytosol protein, were: ER 47 ± S.D. 19, PgR 246 ± 134, PgR:ER 5.9 ± 3.0, compared with ER 18 ± 2, PgR 33 ± 8 and PgR:ER 1.8 ± 0.3 in three autonomous adenocarcinomas. All 3 parameters were significantly different between the two groups (0.02 < P < 0.05), but there was overlap in the ER values, whereas no overlap occurred in the PgR values or the PgR:ER ratios. Ten other adenocarcinomas showed a different sensitivity to restimulation than to ovariectomy and six tumours were fibroadenomas or of mixed histopathology. For the whole tumour population the response to hormonal restimulation at the time of excision suggested that a PgR concentration >60 fm/mg protein combined with an ER concentration >20 fm/mg protein is necessary for the maintenance of hormonal sensitivity. Using these criteria, response was related to receptor concentrations in 93% of the tumours.     

Oestrogen-regulated genes in breast cancer: association of pLIV1 with response to endocrine therapy.

Northern hybridization analyses of the oestrogen-inducible mRNAs pLIV1 and pS2 were compared with oestrogen receptor (ER) immunocytochemistry assessments in 40 untreated primary or early recurrent breast tumours. Significant associations were observed between pLIV1/ER (P < 0.03), pS2/ER (P < 0.001) and pLIV1/pS2 (P < 0.04) status. After disease recurrence, patients were treated with assessable courses of endocrine therapies. Positive pLIV1, pS2 and ER statuses in primary disease were consequently found to be predictive of endocrine responsiveness in the secondary lesions (P < 0.03, P < 0.02, P < 0.005 respectively). However, despite these associations, a number of pLIV1- and/or pS2-positive tumours failed to respond to therapy.     

Oestrogen receptor negative-progesterone receptor positive phenotype in 1,211 breast tumours.

From 1,211 breast cancers, 15 oestrogen receptor (ER) negative-progesterone receptor (PgR) positive breast cancers by conventional dextran coated charcoal steroid binding assays in cytosol were reassessed using Elisa techniques with monoclonal antireceptors antibodies in the cytosolic and nuclear fractions, and immunocytochemistry on cryostat sections. Three categories of results were found in this series. Two tumours were false negative ER due to receptor sites occupancy by hormonal contraceptive treatment. A second group of ten tumours, with high PgR concentrations and immunoreactive ER, corresponds to non ER-binding forms of receptors. One PgR positive tumour was found to be devoid of PgR by using monoclonal antiPgR antibodies might contain a progesterone binding cyst protein. Only two tumours were found to be true ER negative-PgR positive by all methods. This rare phenotype deserves further study of the regulation of the PgR gene.     

Analysis of Genetic Alterations Related to the Development and Progression of Breast Carcinoma

To study genetic alterations related to the development and/or progression of breast carcinoma, we examined amplification of the ERBB2, INT2, and MYC genes, as well as loss of heterozygosity (LOH) at loci on 11p, 16q, 17p ( D17S5 and TP53 ), 17q ( D17S74 and NME1 ), and 18q by restriction fragment length polymorphism analysis. The subjects were 26 patients with small breast carcinomas (≦2 cm) and 88 patients with larger breast carcinomas (2 to < 5 cm). All patients were free of distant metastasis. As tumor diameter increased, the frequency of oncogene amplification and LOH at all loci except D17S5 increased. However, there was no relationship between tumor diameter and amplification of specific oncogenes or allelic loss at specific loci. LOH at D17S5 was detected in 40% of small breast carcinomas (≦2 cm) and 43% of larger breast carcinomas (2 to < 5 cm). There was a significant correlation of LOH at D17S5 with INT2 amplification or with LOH on 11p, 16q, and 18q. These findings suggest that LOH at D17S5 may be involved in the early stage of breast carcinoma development, while INT2 amplification and LOH at 11p, 16q, and 18q appear to be genetic alterations that occur with tumor progression. In addition, as lymph node metastases were significantly related to amplification of the ERBB2 and MYC genes, and LOH of the NME1 gene, these genetic alterations may play a role in the mechanism of lymph node metastases.     

Immunocytochemical localization of estrogen and progesterone receptor and prognosis in human primary breast cancer

An immunocytochemical assay (ICA) for the measurement of estrogen receptor (ER) and progesterone receptor (PgR) has been evaluated in 426 human primary breast carcinomas. For estrogen receptor determination ER ICA was used. PgR ICA was performed using the monoclonal antibody KD 68. Assay results for progesterone receptor immunocytochemistry were in agreement (P less than 0.0001) with those of biochemical determination in 74%. Progesterone receptor positivity determined with a semiquantified approach based on intensity and heterogeneity of immunocytochemical staining correlated significantly with biochemically determined progesterone receptor levels (P = 0.0001). Survival data showed a significantly better overall survival for patients with either ER ICA- or PgR ICA-positive carcinomas (ER ICA, P less than 0.00001; PgR ICA, P = 0.004). Patients with both negative ER ICA and PgR ICA showed a poorer prognosis than patients with only one negative receptor. In ER ICA- and PgR ICA-positive carcinomas a trend could be found that patients whose carcinomas contained high numbers of receptor-positive tumor cells had a better survival. This study demonstrates that ER ICA and PgR ICA are strong prognostic indicators and that the proportion of steroid hormone receptor-positive tumor cells seems to be of clinical importance.     

Prognostic value of Ki-67 immunolabelling in primary operable breast cancer.

The prognostic value of Ki-67 immunohistochemical labelling was evaluated in 327 operable primary carcinomas of the breast. The follow-up time was up to 4 years (mean 2.7 years). The disease-free survival in Ki-67 positive patients was shorter than in Ki-67 negative patients (P < 0.005). By combining the Ki-67 expression with ER receptors and stage, subgroups with a different disease-free survival were identified. In stage II patients there was a significant difference (P < 0.005) in disease-free survival between Ki-67 positive/ER negative and Ki-67 negative/ER positive patients. In node negative patients there was no such difference. The disease-free survival according to different prognostic factors, stage, ER and node status, were separately examined using a Cox''s proportional hazards model. ER (P < 0.0001), the Ki-67 (P < 0.02), tumour size (P < 0.0001) and nodal status (P < 0.006) were independent prognostic factors. We conclude that the potential value of Ki-67 labelling for prognostic evaluation of patients with breast carcinoma is good.     

Prostate-specific antigen and gross cystic disease fluid protein-15 are co-expressed in androgen receptor-positive breast tumours.

Androgens regulate breast cancer cell proliferation via androgen receptor (AR)-mediated mechanisms. To investigate further the androgen-responsiveness of human breast tumours, we examined the immunohistochemical expression of the AR and two androgen-regulated proteins, prostate-specific antigen (PSA) and gross cystic disease fluid protein-15 (GCDFP-15), in 72 primary breast tumours. AR immunoreactivity was present in the nuclei of breast tumour cells and was correlated with oestrogen receptor (ER; P < 0.05) and progesterone receptor (PR; P < 0.01) status. PSA and GCDFP-15 immunoreactivity was present in the cytoplasm of tumour cells but not the adjacent stromal cells. AR-positive cells were present in 85% (61/72) of breast tumours, and 98% (43/44) of PSA-positive and 92% (44/48) of GCDFP-15-positive tumours were also positive for AR. Positive immunoreactivity for both PSA and GCDFP-15 in breast tumours was highly dependent on AR status (odds ratios of 24.0 and 4.5 respectively), but unrelated to age, ER and PR status and axillary lymph node involvement. PSA immunoreactivity was more frequently observed in moderate and well-differentiated tumours and was significantly (P < 0.001) associated with GCDFP-15 immunoreactivity. In conclusion, PSA and GCDFP-15 immunoreactivity was dependent on the presence of AR, but not ER or PR in primary breast tumours.     

Phosphorylation of tyrosine 1248-ERBB2 measured by chemiluminescence-linked immunoassay is an independent predictor of poor prognosis in primary breast cancer patients

ERBB2 (HER2/Neu) gene amplification and overexpression is associated with increased risk of metastases and shorter survival in breast cancer. Tyrosine 1248 is a major phosphorylation site of ERBB2 and reflects the activation status of the receptor. The aim of this study was to investigate the relationships between quantitative levels of pY1248-ERBB2 (p-ERBB2) and the expression of epidermal growth factor receptor (EGFR)-family members, and whether p-ERBB2 could provide additional prognostic value compared with established prognostic markers. For this purpose we developed a highly sensitive chemiluminescence-linked immunoassay (CLISA) and detected p-ERBB2 levels in 70 primary breast cancer biopsies. Phosphorylated ERBB2 correlated with EGFR and ERBB2, and inversely with oestrogen receptor (ER), progesterone receptor (PgR) and ERBB4 expression levels. Additionally, p-ERBB2 was associated with poor clinical outcome in univariate and multivariate Cox regression analysis. Further studies are needed to evaluate the predictive value of p-ERBB2.     

Biomarker profile and genetic abnormalities in lobular carcinoma <Emphasis Type="Italic">in situ</Emphasis>

Summary The predisposition of patients with lobular carcinoma in situ (LCIS) to develop invasive breast cancer (IBC) is well known. However, relatively little is known about the biologic characteristics, which may be involved in the development and progression of LCIS. This study evaluated 59 cases of LCIS (29 pure, 30 with synchronous IBC) for five biomarkers known to be important in IBC (ER, PgR, c-erbB-2, p53 and Ki-67 proliferation rate) by immunohistochemistry. A comprehensive analysis of loss-of-heterozygosity (LOH) was performed in 12 cases (10 pure, 2 with synchronous IBC) at 15 genetic loci on 9 chromosomes. LCIS demonstrated a low grade/favorable biophenotype that was not significantly different in cases with and without synchronous IBC (ER 98%, PgR 84%, c-erbB-24%, p53 19% and proliferation rate 2%). LOH was present in 80% of pure LCIS and the highest rates of LOH were at loci on 9p (30%), 16q (63%), 17p (33%) and 17q (50%). The clustering of LOH at these four foci suggests that inactivated tumor suppressor genes in these regions may be particularly important. LOH was present in both cases of LCIS with synchronous IBC and the LOH phenotype was shared by LCIS and IBC. Our findings suggest that five known prognostic factors in IBC do not have prognostic utility in LCIS. Multiple genetic mechanisms may be involved in the development of LCIS.     

Does vimentin help to delineate the so-called 'basal type breast cancer'?

Background

Vimentin is one of the cytoplasmic intermediate filament proteins which are the major component of the cytoskeleton. In our study we checked the usefulness of vimentin expression in identifying cases of breast cancer with poorer prognosis, by adding vimentin to the immunopanel consisting of basal type cytokeratins, estrogen, progesterone, and HER2 receptors.

Methods

179 tissue specimens of invasive operable ductal breast cancer were assessed by the use of immunohistochemistry. The median follow-up period for censored cases was 90 months.

Results

38 cases (21.2%) were identified as being vimentin-positive. Vimentin-positive tumours affected younger women (p = 0.024), usually lacked estrogen and progesterone receptor (p < 0.001), more often expressed basal cytokeratins (<0.001), and were high-grade cancers (p < 0.001). Survival analysis showed that vimentin did not help to delineate basal type phenotype in a triple negative (ER, PgR, HER2-negative) group. For patients with ''vimentin or CK5/6, 14, 17-positive'' tumours, 5-year estimated survival rate was 78.6%, whereas for patients with ''vimentin, or CK5/6, 14, 17-negative'' tumours it was 58.3% (log-rank p = 0.227).

Conclusion

We were not able to better delineate an immunohistochemical definition of basal type of breast cancer by adding vimentin to the immunopanel consisted of ER, PgR, HER2, CK5/6, 14 and 17 markers, when overall survival was a primary end-point.     

Variation of hormonal receptor, pS2, c-erbB-2 and GSTpi contents in breast carcinomas under tamoxifen: a study of 74 cases.

Seventy-four post menopausal patients with primary non-metastatic invasive ductal carcinomas of the breast were first treated with tamoxifen alone (30 mg p.o. daily) for 5 months. To study changes induced by tamoxifen, core biopsies before treatment and surgical specimens after hormonal therapy were assayed by immunohistochemistry for oestrogen (ER) and progesterone receptors (PR), pS2, GSTpi and c-erbB2. After tamoxifen, ER and PR significantly decreased in 60 and 44 cases respectively, whereas 11 and 19 cases showed no variation and 2 and 11 cases showed an increase (P<10(-4)). GSTpi and pS2 showed a significant increase in 43 and 41 cases, a decrease in 2 and 21 cases and no variation in 29 and 12 cases (P<10(-4) and P=0.04 respectively). c-erbB-2 showed no significant variation under tamoxifen, increased in only three cases and decreased in 13 cases. No relation was found between these variations and efficiency of hormone therapy. Our results allow a better knowledge of protein expression modifications occurring in breast cancer cells under tamoxifen therapy. They are also more consistent with clone selection rather than with phenotype modification.     

Prognostic value of bcl-2 expression in invasive breast cancer.

Expression of the bcl-2 proto-oncogene was studied immunohistochemically in 251 invasive ductal breast carcinomas (median follow-up time 91 months, range 24-186 months) and the results were correlated with clinicopathological data and prognostic variables. Sixty-three (25%) tumours were scored bcl-2 negative and 188 (75%) tumours were bcl-2 positive. No relationship could be observed between bcl-2 status and tumour grade, pTNM staging or menopausal status. A strong positive relationship was demonstrated between bcl-2 immunoreactivity and oestrogen receptor status (P < 0.001) and progesterone receptor status (P < 0.001). No prognostic value was demonstrated for bcl-2 expression on disease-free survival and overall survival in axillary node-negative breast cancer patients. However, in axillary node-positive breast cancer patients multivariate analysis demonstrated absence of bcl-2 expression to be independently related to shortened disease-free survival (P = 0.003) and shortened overall survival (P < 0.001). Our results suggest a potential important role for bcl-2 expression as a modulator of response to adjuvant therapy in breast cancer.     

Immunoradiometric detection of pS2 and total cathepsin D in primary breast cancer biopsies: their correlation with steroid receptors.

Commercially available immunoradiometric assays were used for pS2 and total cathepsin D determination in the cytosol fraction obtained from 266 primary breast cancers. We show that pS2 and cathepsin D values were significantly associated (Spearman''s rank correlation: P < 0.0001) in tumours from lymph node-positive patients (N+), while such association did not reach significance in tumours taken from patients with negative lymph nodes (N-). Moreover, cathepsin D concentrations in pS2-rich tumours (pS2 above the median value, 5 ng mg-1 protein) were significantly higher (Mann-Whitney-Wilcoxon''s rank-sum test: P = 0.00001) than those obtained in the samples expressing less than 5 ng of pS2 per mg of protein. pS2 was also correlated to both the oestrogen receptor (ER) (Spearman''s rank correlation: P < 0.0001) and the progesterone receptor (PR) (Spearman''s rank correlation: P = 0.022). No significant differences in the expression of pS2 and cathepsin D taken from N+ and N- patients were found. Furthermore, no significant differences in pS2 and cathepsin D expression were obtained by stratifying tumours on the basis of their size (T). pS2 and cathepsin D values obtained in ER-positive/PR-positive tumours did not significantly differ from the values obtained in ER-positive/PR-negative and in ER-negative/PR-positive tumours. We conclude that pS2 could have a role in cathepsin D expression, and that it can be used in the assessment of a functioning oestrogen response machinery in those tumours that express only ER.