Optimal fixation conditions for immunocytochemical analysis of estrogen receptor in cytologic specimens of breast carcinoma

Validation of commonly used immunostains on cell-transferred cytologic specimens

BACKGROUND

Immunocytochemical staining (ICC) is often limited by the lack of cell blocks that contain diagnostic cells or by the unavailability of smear materials. The cell-transfer technique, in which original smear material is divided into several pieces and then transferred to multiple slides, can facilitate multiple ICC on limited materials. However the reliability of the staining results has not been systematically evaluated.

METHODS

The authors validated 21 commonly used antibodies (cytokeratin [CK] 7, CK 20, TTF-1, panCK, vimentin, HMB-45, Mart-1, leukocyte common antigen, chromogranin, synaptophysin, estrogen receptor, progesterone receptor, prostate-specific antigen, prostatic acid phosphatase, calretinin, CK5/6, WT-1, BerEP4, MOC-31, p63, and thyroglobulin) on the cell-transferred materials of various tumor samples (n = 20) and reactive effusion samples (n = 2). The staining results were compared with the previously available ICC results (performed either on cell block sections or on smears) of the same specimens.

RESULTS

One hundred pieces of transferred materials were immunostained, of which 5 (5%) were lost during the staining procedure. Of the remaining 95 pieces, 92 (97%) showed staining results that agreed with those of the previous ICC, whereas 3 (3%) showed results that disagreed. The original ICC in the latter three specimens was performed on cell block sections and showed focally positive synaptophysin staining in a sample with a neuroendocrine tumor, focally positive CK5/6 staining in a pleural effusion sample with mesothelioma, and positive TTF-1 staining in a lymph node sample with metastatic lung carcinoma. Their corresponding ICC of cell-transferred materials showed negative results. An increase in background staining was observed in 1 of the 95 tissue pieces, but no false-positive results were observed.

CONCLUSIONS

In general, ICC can be reliably performed on cell-transferred tissue specimens for markers tested. However, caution should be taken when interpreting ICC results in tissue specimens with limited cellularity and heavy background staining.

Oestrogen receptor beta-immunoreactive neurones in the ovine hypothalamus: distribution and colocalisation with gonadotropin-releasing hormone

Oestrogen powerfully affects the secretion of gonadotropin-releasing hormone (GnRH) from the brain in all species investigated, including sheep. Until recently, it was hypothesised that such regulation occurs indirectly because few or no GnRH neurones were found to express oestrogen receptor (ER) alpha. The discovery of a second oestrogen receptor, ERbeta, and its subsequent localisation in numerous GnRH neurones in the rat, led to a reconsideration of this hypothesis. However, colocalisation of immunoreactive ERbeta protein in GnRH neurones has only been demonstrated in the rat, raising the possibility that such putative direct regulation of GnRH neurones by oestrogen may be peculiar to this species. We have previously shown that steroid receptors in the sheep brain are acutely sensitive to fixation and the full complement of immunoreactive cells can only be visualised after antigen retrieval. The aims of this study were therefore to map immunocytochemically the distribution of ERbeta neurones in the ewe brain, and to determine which proportion of GnRH neurones express ERbeta. Brain sections (20 microm) from four ewes killed in anestrus were subjected to high temperature antigen retrieval and immunocytochemistry. Numerous ERbeta-immunoreactive cells were located throughout the hypothalamus and, following dual-label immunocytochemistry, over 50% of the GnRH neurones were found to express immunoreactive ERbeta. The functional significance of these ERbeta-expressing GnRH neurones in the ovine brain remains to be determined.

Estrogen receptor analysis for breast cancer: current issues and keys to increasing testing accuracy

The estrogen receptor (ER) is a regulator of cellular growth, proliferation, and differentiation. In addition to having prognostic value, ER is the most important biologic marker of therapeutic response in breast cancer. Some level of measurable ER protein is expressed in 70-80% of human breast. Immunohistochemistry is the current method of choice for ER assessment, and its predictive value has been shown to be superior to that of biochemically based assays. Although accurate ER protein assessment is critical for optimal treatment of patients with breast cancer, studies have demonstrated inter-laboratory variability in ER detection. False-negative results for tumors with low ER protein levels have been a subject of recent concern. Lack of standardization for immunohistochemistry between laboratories is thought to be the major reason for testing errors, although variability in scoring methods and reporting practices, which can affect results, also plays a role. In this article, we review studies addressing interlaboratory variability and recommend optimal testing techniques and reporting procedures for ER testing, with the goal of increasing interlaboratory standardization for ER analysis by immunohistochemistry. We additionally highlight recent biologic, molecular, and gene expression profiling data related to ER in breast cancer.