Molecular Biology of Preliferative Breast Disease
A number of genes have been identified as putative oncogenes in breast cancer based on expression in breast cancer and correlation with prognosis or as tumor suppressor genes by LOH. A genetic sequence of events has been deduced for colon carcinoma (Fearon and Vogelstein, 1990). Although it is generally believed that breast cancer, too, is the culmination of numerous genetic alterations, a genetic sequence involved in progression for breast has not been deciphered.
The road to cancer in the breast may take many routes. Perhaps complementation groups such as those described for immortahzation (Pereira-Smith and Smith, 1988) exist such that one gene increases the probability of cancer only in the presence of second, specific genetic alteration(s). Undoubtedly genes yet to be identified, as well as the genes to be discussed, are involved.
ErbB-1 encodes for the 170-kDa protein epidermal growth factor receptor (EGFR). More than 5% of cells overexpress EGFR in 56% of early invasive breast cancers, and expression may be associated with early relapse (Gasparini et al., 1992). In particular, cases that are EGFR positive with a growth fraction exceeding 7.5% as determined by antibody against Ki-67 are much more likely to relapse.
EGFR may be overexpressed in about 25% of DCIS cases (Moller et al., 1989), but a possible role in the development of PBD is suggested by the report that EGFR was detected more frequently in breast smears of complex breast cyst fluids (12/17) than in smears of simple cyst fluids (5/23; Athanassiadou et al., 1992). Simple cysts are those with flattened epithelium and complex cysts have apocrine or hyperplastic epithelium. The prevalence of overexpressed EGFR in fine-needle aspirates obtained from high risk patients (35%) was significantly greater than from women of low risk (4%; Fabian et al., 1994). High risk patients were defined as those with a first-degree relative with breast cancer, prior breast cancer, or precancerous mastopathy. There is also evidence that overexpression of EGFR is common in fibroadenomas (Zelada-Hedman et al., 1994).
An association of EGFR expression and p53 expression has been described leading to the suggestion that the two interact in the pathogenesis of cancer (Horak et al., 1991). In that study, p53-positive tumors (staining-detectable with Pab240) were found to have a mean concentration of 31 fmol EGFR per mg of membrane protein compared to 14 fmol/mg for p53-negative tumors.
However, nearly half of the p53-positive cases were EGFR negative. A number of studies have found a negative correlation between expression of EGFR and ER (Sainsbury et al., 1987; Horak et al., 1991; Koenders et al., 1991) but many cases express both receptors (Koenders et al., 1991). The latter study found that, although erbB-1 levels decreased with increasing ER levels, 173/531 cases (33%) expressed both using at least 10 fmol/mg cytosolic protein as the criterion for ER status and 0.5 fmol/mg of membrane protein as the criterion for EGFR expression (median values were 40 fmol/mg).
Using different criteria (5 fmol/mg for ER and 3 fmol/mg for EGFR), 25% of normal breast samples and 33% of benign/fibrocystic samples were found to be positive for both ER and EGFR (Barker et al., 1989). However, individual cells within breast cancers apparently are never positive for both (Sharma et al., 1994). Thus, EGFR and ER expression appear to be mutually exclusive at the cellular level.
Normal breast epithelium, fibrocystic disease, and fibroadenomas all stained with weak to moderate intensity with antibodies to both erbB-2 and EGFR (Tsutsumi et al., 1990). Invasive carcinomas overexpressed one or the other, but not both, in 13 of 36 cases. In a single case, both erbB-2 and EGFR were overexpressed but in a reciprocal manner in which erbB-2 was overexpressed in the DCIS component and the EGFR was overexpressed in the invasive component (Tsutsumi et al., 1990).
Based on this case, the authors suggest that erbB-2 loss at the stage of invasion might explain the decrease in incidence of erbB-2 expression in invasive cancer from DCIS. However, this case seems to be the rare exception (see below). Both erbB-2 and EGFR were detected in only 6/126 breast cancers by immunohistochemistry, but these patients had significantly worse prognoses than did patients with invasive cancers expressing one or the other or neither (Toi et al., 1994). This study did not address whether EGFR and erbB-2 were detected in the same or separate cells within the six EGFR+/erbB-2+ cancers.