c-erbB-2 oncogene expression in prostatic intraepithelial neoplasia: mounting evidence for a precursor role.

The search for the precursor of invasive prostatic adenocarcinoma has intensified in recent years, focusing on the spectrum of histopathologic changes referred to as high-grade prostatic intraepithelial neoplasia. This microscopic finding, considered to be the most likely precursor of invasive carcinoma (7), is characterized by cellular proliferations within pre-existing ducts and glands, with cytologic changes mimicking cancer, including nuclear and nucleolar enlargement. Prostatic intraepithelial neoplasia coexists with cancer in more than 85% of cases but retains an intact or fragmented basal cell layer, unlike cancer that lacks a basal cell layer (2). The clinical importance of recognizing prostatic intraepithelial neoplasia is based on its strong association with prostatic carcinoma. Prostatic intraepithelial neoplasia has a high predictive value as a marker for adenocarcinoma, and its identification in biopsy specimens of the prostate warrants further search for concurrent invasive carcinoma. Studies, to date, have not determined whether prostatic intraepithelial neoplasia remains stable, regresses, or progresses, although the implication is that it can progress (2). The identification of specific and consistent molecular changes in prostatic intraepithelial neoplasia, such as pl60 rbB3 and pl85~ overexpression, adds further support to the expanding body of evidence indicating the premalignant nature of this lesion. In this issue of the Journal, Myers et al. (3) describe immunoreactivity in paraffin-embedded archival tissue sections for p,60erbB-3 a n d p J 85erbB-2 j n m e b a s a , c e , , s o f t h e b e n j g n p r o S ta t ic epithelium, in basal cells of prostatic intraepithelial neoplasia, in secretory luminal cells of prostatic intraepithelial neoplasia, and in carcinoma. This pattern of immunoreactivity for oncoproteins in prostatic intraepithelial neoplasia and cancer is similar to that of other biomarkers in the prostate, including epidermal growth factor, epidermal growth factor receptor, type IV collagenase, Lewis Y antigen, transforming growth factor-cx (4), apoptotic bodies, and proliferating cell nuclear antigen (5). Increased cell proliferation is the most likely explanation for this phenotypic similarity of normal basal cells, prostatic intraepithelial neoplasia, and cancer, suggesting that the basal cells are the regenerative or stem cells of the prostate. Other biomarkers show progressive loss with increasing grades of prostatic intraepithelial neoplasia and cancer and include markers of secretory differentiation such as prostatespecific antigen, secretory proteins, cytoskeletal proteins, glycoproteins, and neuroendocrine cells (6). These results indicate that there is progressive impairment of cell differentiation and regulatory control with advancing stages of prostatic carcinogenesis. Changes in cytoskeletal proteins in prostatic intraepithelial neoplasia may affect transport of cell products, accounting for the differences in secretory protein distribution. Virtually all studies to date of biomarkers have indicated that high-grade prostatic intraepithelial neoplasia is more closely related to carcinoma than to benign epithelium. As members of the epidermal growth factor family and its associated oncoproteins, pl60" and pl85" reflect an increase in cellular proliferative activity when they are overexpressed. The c-erbB-2 (HER-2/neu) oncogene encodes pl85', a 185-kd transmembrane phosphoglycoprotein with 43% homology with epidermal growth factor receptor in extracellular sequences. The findings of Myers et al. (J) confirm and extend the results of other studies that indicate the presence of this oncoprotein in basal cells of the prostate and in the secretory luminal cells of high-grade prostatic intraepithelial neoplasia. Using TA1, a monoclonal antibody that recognizes the extracellular domain of the transmembrane oncoprotein, Ibrahim et al. (7) found a similar pattern of immunoreactivity for c-erbB-2 in frozen sections of prostatic intraepithelial neoplasia and prostate cancer. Using AB-3, a monoclonal antibody directed against the C-terminal end of the c-erbB-2 oncoprotein, Giri et al. (8) identified basal cell staining in normal and hyperplastic prostatic epithelia and in adenocarcinoma. Conversely, other studies of c-erbB-2 oncoprotein expression have found inconsistent patterns of immunoreactivity, probably resulting from differences in tissue preparation and fixation, immunohistochemical techniques, antibody specificity, and inter-