Quality control for estrogen receptor quantification by dextran-coated charcoal assay: a single laboratory's experience

Quality Assurance for Steroid Receptor Assay in Human Breast Cancer: Six Years Experience of the Italian Committee

Since 1979 the quality control design proposed by the Italian ad hoc Committee has evaluated several lyophilized preparations with scalar receptor content; this permits the identification by linear regression analysis of systematic and non systematic errors. At present 41 laboratories from most of the national regions have joined the Italian Committee. The overall results of five years application of quality assurance in Italy show that there was a different pattern of imprecision with satisfactory indexes for intralaboratory performances but major variations in interlaboratory controls. There was also a remarkable difference of variability indices between the so-called "expert" and "new" laboratories; this problem can be reduced with practical seminars for new centers. On the basis of the results and experience achieved the Committee is starting another program of quality assurance for different new methodologies to provide guidelines for international working reference standards.

Expression of beta 1 integrins in non-neoplastic mammary epithelium, fibroadenoma and carcinoma of the breast

beta 1 Integrins were examined immunohistochemically in normal and mastopathic mammary glands, 12 benign tumours and 90 carcinomas of the breast using monoclonal antibodies against beta 1 and alpha 1 to alpha 6 subunits. When compared with epithelial cells of non-neoplastic mammary glands and of benign tumours, carcinoma cells showed considerable quantitative changes in the pattern of alpha 2, alpha 3 and alpha 6 subunit expression. In contrast, the distribution pattern of beta 1, alpha 1, alpha 4 and alpha 5 antigens corresponded to the situation observed in non-neoplastic mammary gland epithelium in most instances. An abnormal expression of alpha 2 was found in 71.0% of the carcinomas ranging from a remarkably low number of alpha 2-positive tumour cells in 27.5% of the cases to a complete absence of the alpha 2 molecule in 43.5% of the carcinomas. Of the carcinomas 39.9% exhibited quantitative changes in alpha 3 expression with an abnormally low content of alpha 3-positive neoplastic cells in 15.4% and a complete absence of this molecule in 24.5% of the cases. Expression of alpha 6 was abnormal in 73.2% of the carcinomas, consisting in a greater number of alpha 6-negative tumour cells in 31.9% and in a complete absence of alpha 6 in 41.3% of the tumours. The abnormally low expression/absence of alpha 2 and alpha 3 subunits correlated with oestrogen receptor negativity (P < 0.033 and P < 0.04, respectively). In addition, abnormally low expression/absence of alpha 2 correlated with poor differentiation of the tumours (P < 0.014). The quantitative changes in the expression pattern of beta 1-associated alpha subunits in breast carcinomas may cause a disturbed cell-cell and/or cell-matrix interaction that increases the invasive and migratory property of the tumour cells.

Expression of ecto-5'-nucleotidase (CD73) in normal mammary gland and in breast carcinoma

Ecto-5'-nucleotidase (ecto-5'-NT) is a phosphatidylinositol anchored membrane structure recently defined as the lymphocyte differentiation antigen CD73. Using CD73 (1E9.28.1) monoclonal antibody, normal mammary gland and breast carcinoma were immunohistochemically investigated for ecto-5'-NT expression. In normal breast epithelium, CD73 was differentially expressed in lobular, ductal and myoepithelial cells and was most frequently detected in the myoepithelial compartment. The glandular stroma contained fibrocytes, a subset of which was also CD73-positive. Among 102 unselected breast carcinoma primary lesions, only 9 contained CD73-positive tumour cells, whereas in 95 cases, stromal fibroblasts and fibrocytes showed variable degrees of CD73 expression. The extent of stromal CD73 expression correlated positively with the estrogen receptor (ER) status of the tumour (P less than 0.038). We conclude that ecto-5'-NT-expression reflects a still unknown state of activity of normal breast epithelium which is lost in the majority of carcinomas derived therefrom. It may also be indicative of some functional activity of stromal fibroblasts which is significantly enhanced in ER-positive carcinomas.

Identification of the major sources of error in estrogen receptor measurements for individual laboratories using both tissue and cytosol samples

The major sources of error between laboratories performing estrogen receptor measurements in tissue samples were identified for 17 participating laboratories in a trial conducted by the Australasian Quality Assurance programme. Both tissue and cytosol samples were provided, and the In-House assays were compared with the ER-EIA kit (Abbott Laboratories, U.S.A.) as a reference assay. For both the In-House and Abbott assays, tissue samples resulted in a between laboratory CV of about 55% and a within laboratory CV of about 30%. In contrast to tissue samples, the between laboratory CV for cytosol samples was reduced to 41% for the In-House assays and to 33% for the Abbott assay, whereas the within laboratory CV was reduced to 10% for both types of assay. The different methods of tissue homogenization by themselves were not found to be sources of error, and protein extraction efficiency from tissue was strongly correlated with protein measurement (P less than 0.0005). The major sources of error due to protein measurement, cytosol preparation, In-House and Abbott assays were evaluated for individual laboratories. The results indicated absence of any major sources of error for four laboratories, while one, two and three or more sources were indicated for seven, three and three laboratories respectively. The conclusion that about half the participants need to improve their ER assays was confirmed by three independent reviews. Furthermore, the trial demonstrated that tissue samples are essential as quality assurance material for a realistic assessment of ER assays in biopsy specimens.

Sources of improvement in between laboratory variation in estrogen and progestin receptor measurement using tissue samples during the Australian Quality Assurance Programme

Tissue and cytosol samples were compared as quality control material for assessment of between and within laboratory error in measurement of estrogen and progestin receptors (ER and PR) in a series of four trials for a total of 17 participating laboratories during the Australasian Quality Assurance Programme. For tissue samples, a substantial reduction in between laboratory CVs for both ER and PR from about 90 to 50% was achieved during the programme. In contrast, for cytosol samples a substantially lower between laboratory CV of about 30% was obtained. Tissue sample heterogeneity could be excluded as a major source of variation between laboratories. The likely source of the observed improvement in CV for tissue samples during the trials was due to a reduction of the initial under-estimation of receptor concentration in tissue samples by some of the participants. Although cytosol preparation from tissue samples was shown to be one major source of error, other sources of error such as the receptor assay itself and the associated protein measurements were identified. It is concluded that fragmented tissue samples are essential for a realistic assessment of between laboratory error in receptor measurements in biopsy material such as obtained from clinical breast cancer samples.

Nuclear location of hormone-free estrogen receptors by monoclonal antibodies could be a tissue-fixation dependent artifact

The 'two-step' model proposed by Jensen and his collaborators for explaining estrogen action conceptualized hormone-free estrogen receptors (ER) to be cytoplasmic, and hormone-filled, transformed ER to be nuclear. Applying monoclonal antibodies which recognized epitopes in ER and formaldehyde-fixed tissues, King et al demonstrated exclusively nuclear staining in target tissues utilizing immunoperoxidase technique. Recently these antibodies have become commercially available enabling other investigators to conduct studies. In this report, using these monoclonal antibodies we have demonstrated that a change in the concentration of formaldehyde alters the staining pattern yielding cytoplasmic instead of nuclear staining in calf uterus, MCF-7 cells, and ER(+) human breast cancer. In addition, neutralization of the antibody activity was not achieved with freshly prepared ER(+) cytosols. Formaldehyde-treated cytosols were essential. These results ought to caution investigators in determining in vivo location of antigens based on the staining pattern obtained in fixed tissues. Furthermore, this effect of formaldehyde on estrogen receptors may be applicable to other steroid hormone receptors.

Evaluation of a new oestrogen receptor assay

Replicate assays of oestrogen receptor (ER) concentration on 81 specimens of human breast carcinoma were performed to compare the ligand binding Sephadex separation method with a polystyrene bead linked anti-oestrogen receptor monoclonal antibody technique (enzyme immunoassay [EIA], Abbott Laboratories). Each specimen was homogenized and the cytosol divided into four fractions. Replicate assays by each method gave an estimate of reproducibility. The mean CV (coefficient of variation = standard deviation/mean) for the ligand binding method was 9.73% compared with a mean CV of 17.9% for the EIA method. In the clinically significant range of ER values, around the cut-off point between negative and positive assays, there was no difference in the precision of the two assays (Mann Whitney U-test). The correlation between methods gave a Pearson's Product Moment Correlation Coefficient (r value) of 0.822, significant at the P less than 0.001 level, indicating a close correlation between the two methods. The highest CVs for both methods were in the range of oestrogen receptor concentrations, below 10 fmol/mg cytosol protein. Using 10 fmol/mg cytosol protein as a cut-off between negative and positive assays, 9.87% of the specimens would be reclassified as either positive or negative by the new assay method. The new assay method is, therefore, acceptable for clinical use.

Standardization of steroid receptor assays in human breast cancer--III. Selection of reference material for intra- and inter-laboratory quality control

A comparison of estrogen and progesterone receptor (ER and PgR, respectively) analyses using minced frozen tissue and lyophilized cytosols of the same samples demonstrated that intra-laboratory variations in assays performed by 13 members of the EORTC Receptor Group are lowest using the lyophilized samples. Inter-laboratory variation in receptor values was on the same order of magnitude for both types of samples (ca 22% for ER and 30% for PgR). There was no correlation between receptor values measured within each laboratory for either ER or PgR in minced tissues compared to lyophilized cytosols, which illustrates that methods of tissue disruption, extraction of receptors, and preparation of cytosol are sources of intra- and inter-laboratory variation. In some laboratories the handling of the tissue was apparently sub-optimal since a slight but significant difference was found in the overall mean concentration of ER in minced tissue compared to lyophilized cytosol samples. It was concluded that lyophilized tissue samples are the material of choice for routine intra- and inter-laboratory quality controls. However, differences in methods of handling tissue to obtain cytosol should not be disregarded since they lead to increased intra-laboratory variation. A difference was demonstrated between use of a common batch of isotope and the different batches concurrently employed in the laboratories, but the differences were not large enough to warrant use of a common batch for routine inter-laboratory comparisons. Differences in methods used to convert cpm to dpm did not appreciably affect the results when counting samples of tritium containing from 30,000 to 105,000 dpm.

Effect of sodium molybdate on protein measurements: quality control aspects of steroid hormone receptor assays

The quantity of steroid hormone receptors in human mammary carcinomas is of importance for predicting clinical response to hormone modes of therapy. There have been several reports describing an increase in the yield of cytosolic receptors when sodium molybdate is included in the tumor cytosol. Since the quantitative units for receptors are fmol hormone bound per mg cytosolic protein, we examined whether this oxyanion had an interfering or enhancing effect on two methods of protein measurements: those of Bio-Rad and of Lowry. Our results showed that sodium molybdate, even when present in a 5 mM concentration in the protein solution, interferes with the color development by both reagents. This effect of molybdate is consistent, reproducible and statistically significant. We present the data to caution the investigators to be aware of this phenomenon while interpreting their own data on the molybdate effect on steroid receptors.