Molecular evidence has recently suggested a number of different pathways leading to the development of ductal carcinoma of the breast. The links between atypical ductal hyperplasia and low-grade ductal carcinoma in situ and lobular neoplasia and lobular carcinoma are well known pathologically, but high-grade in situ and invasive carcinomas appear to have a different biological oncogenetic pathway. Morphologically there is a similarity between apocrine cells and some cases of high-grade ductal carcinoma. In order to investigate this possibility a number of different biological markers known to occur in high-grade breast carcinomas were assessed in both apocrine metaplasia (APM) and a putative premalignant lesion called apocrine change within sclerosing adenosis (AA). In 64 cases of APM and 18 cases of AA we examined for expression of c-erbB2, p53, Bcl-2, Bax, c-myc and Ki-67 proteins using immunocytochemistry. c-erbB2 expression was seen in 55.6% of AA cases and in 10.9% of APM cases. p53 expression was detected in 27.8% of AA cases but only 1.6% of APM cases. All cases of AA and APM were negative for the anti-apoptotic protein Bcl-2, but all the APM and 33.3% of AA cases showed cytoplasmic positivity for Bax, a pro-apoptotic protein. All the cases of AA and APM were positive for c-myc oncoprotein, however, the mean percentage of nuclear positivity was 50% in AA and 37% in cases of APM cases. The mean percentage positivity for Ki-67, a proliferation associated antigen, was 3.6% in AA and 1.3% in APM. The results indicate that a subset of breast lesions containing APM epithelium show abnormal oncoprotein and apoptosis-related protein expression and have a higher proliferation rate.

Loss of heterozygosity (LOH) and allelic imbalance (AI) at loci reported to show allele loss and/or imbalance in preinvasive and invasive breast cancer were examined in 41 cases of apocrine metaplasia (APM) of the breast using a microdissection technique, polymorphic microsatellite markers, and the polymerase chain reaction (PCR). Occasional examples of LOH and/or AI were identified in 2/28 (7.1%) informative cases at 1p (MYCL1), 2/14 (14.3%) at 11q (INT2), 1/15 (6.7%) at 13q (D13S267), 3/22 (13.6%) at 16q (D16S539), 2/23 (8.7%) at 17p (TP53), and 1/11 (9.1%) at 17p (D17S513) and 3/16 (18.8%) at 17q (D17S250). The finding of LOH/AI in cases of APM indicates that a subset of APM appears clonal, but the significance of allelic loss or imbalance in the pathogenesis of APM or its possible subsequent progression to carcinoma is not yet clear and requires further investigation. Clinical follow-up of these particular cases of APM showing LOH/AI would be of further value.