Tumor marker expression in breast carcinomas and relationship to prognosis. An immunohistochemical study

Abstract

A collaborative immunohistochemical study was carried out to examine the expression and prognostic significance of tumor markers in a retrospective series of 233 invasive breast carcinomas. The patterns of tumor marker expression in 94 patients with short remission duration (recurrence within five years) were compared with 50 patients with intermediate (at five to ten years) and 89 patients with long (no recurrence at ten years or longer) remission durations. The antigens examined were carcinoembryonic antigen (CEA), human chorionic gonadotropin (HCG), placental lactogen, alpha-lactalbumin, and pregnancy-specific beta-1 glycoprotein. Carcino-embryonic antigen was the most frequently expressed antigen, whereas HCG was demonstrated least frequently. Also, the ABH isoantigen status was examined using monoclonal antibodies; isoantigen expression was observed in a subset of breast carcinomas, contrary to previous reports of total deletion in breast cancer. Two of the markers, CEA and HCG, were examined by both laboratories, each with two different antisera and also with both PAP and ABC immunohistochemical technics. Meticulous efforts were taken to provide quality control and ensure reproducibility of results. These included the use of serial sections of duplicate pathologic material by both institutions, standardization of experimental conditions and interpretation criteria, double-blind evaluation of exchanged slides, and use of standardized data sheets to record staining extent and intensity. No significant disagreements were observed between data obtained through the different approaches. The steps that were taken to minimize interobserver and interinstitutional differences in this study are presented as a model for collaborative immunohistochemical studies. The expression of tumor markers, alone or in combination, was not found to bear any significant relationship to prognostic indicators, such as the likelihood of recurrence, interval before recurrence, or presence of metastasis.