Estrogen receptor expression in breast cancer: we cannot ignore the shades of gray.

In this issue of the Journal, Gomez-Fernandez et al 1 report their experience with estrogen receptor (ER) immu-nohistochemical expression in 278 patients with recurrent and/or metastatic breast cancer. By using ER antibody 1D5, they found that 9 of 159 previously positive cases (5.7%) did not express ER in their metastatic or recurrent sites. In addition , of 119 initially ER– primary breast cancers, no cases changed to a positive ER status in the metastatic or recurrent sample. The conclusion of the study is that ER status, using well-fixed tissues and the 1D5 antibody, remains stable through disease progression. In breast cancers that originally test positive for ER expression (particularly those that are strongly positive), it does not seem to be necessary to routinely repeat ER staining except for diagnostic purposes (ie, to confirm metastatic disease is from a previously ER+ breast primary site). Not only are oncologists loathe to withhold hormonal therapy if a case was considered previously ER+, it is difficult to be confident that the more recent result is not a false-negative. As emphasized by Gomez-Fernandez et al, 1 because of a lack of internal controls in metastatic sites, it is often not possible to exclude poor antigen preservation in these cases. The deleterious effects of decalcification in bone marrow tissue stained for ER is one possible source of poor antigen preservation. Others have reported similar findings when comparing ER in breast needle core biopsy specimens to the final surgical specimen, with the most common change in ER being " false-negative " results in the final surgical specimen, possibly owing to worse fixation of larger specimens. 2,3 But what to do with cases that originally test ER– or only weakly positive? If the initial result is a false-negative, potentially beneficial hormone therapy will be withheld. In the study by Gomez-Fernandez et al, 1 there were no cases that changed from initially negative for ER expression to positive for ER expression at the metastatic or recurrent site. So no new candidates for hormonal therapy were identified. Are we to then assume that repeated testing in previously ER– cases is not warranted? The rate of false-negative ER staining of initial core needle biopsy specimens is estimated to be as low as 2% to 3%. 2 It is not mentioned in the study by Gomez-Fernandez et al 1 if their initial ER determination was performed on needle core biopsy or excision …