Prognostic and predictive factors in breast cancer by immunohistochemical analysis

Most recent decisions for breast cancer patients are made on the basis of prognostic and predictive factors. In addition to the traditional tumor/nodal/metastasis staging variables, estrogen and progesterone receptor status as assessed by biochemical ligand-binding assays are the only other factors that have been adequately validated and recommended for routine clinical use. Pathologists today, however, are evaluating estrogen and progesterone receptors almost exclusively by immunohistochemical means. Although many studies suggest that these tests might have equivalent or even superior abilities to predict patient outcome, there are important methodologic shortcomings to resolve before this technology achieves the clinical and technical validation necessary to justify its routine use. Many laboratories are also evaluating other factors for clinical use by immunohistochemical techniques, including, in particular, c-erbB-2, p53, and Ki-67 proliferation indices. Although available studies suggest that these factors might indeed be helpful in making treatment decisions, their clinical usefulness is still controversial, and, like the assessment of hormone receptors, there are important unresolved technical issues, such as how to prepare the tissue, which reagents to use and, most importantly, how to interpret the results. A few laboratories have gone to considerable effort to develop reproducible methods for evaluating these factors, and they have performed comprehensive studies demonstrating the prognostic and predictive significance of their results. Nonetheless, most laboratories offering these tests have not adequately validated them and might not even be aware of the issues needing attention. Unless laboratories validate their tests or follow the procedures of others who have, they run the risk of reporting meaningless and potentially harmful results. In the future, these and other factors will be incorporated into a prognostic index that will better reflect the biologic diversity of breast cancer and that will more accurately predict clinical outcome.