Preoperative systemic treatment: Prediction of responsiveness

The use of predictive factors allows a more effective use of available therapies by enabling clinicians to distinguish patients likely to obtain substantial benefit from treatment from those for whom the same therapy is less likely to be effective. A most relevant aspect of clinical research is thus to develop alternative therapeutic approaches which are more efficacious for this latter group, particularly important since treatment effects are likely to be small. In the preoperative setting several predictors of response were identified. They include: diameter of the lesion (larger lesions respond less than smaller lesions), MIB-1 increased expression associated with increased response to chemotherapy, and estrogen receptor (ER) and progesterone receptor (PgR) expression in the tumor typically associated with increased response to endocrine therapies. Other factors include HER-2/neu overexpression, which is a target for treatment with the humanized monoclonal antibody against its extracellular domain, is hypothesized to increase response to anthracycline combination chemotherapy and to lead to an improved response to some endocrine agents (e.g. letrozole) rather than to others. Although primary endocrine therapy demonstrated activity and low profile of side effects in selected populations of older patients, it is infrequently used. On the other hand, chemotherapy remains the mainstay of treatment being considered to be a more active and better documented option. Experience at the European Institute of Oncology on 399 patients with large or locally advanced breast cancer (cT2-T4, N0-2, M0) treated with primary chemotherapy, indicated that a proper selection of primary treatment should be based on tumor characteristics such as ER and PgR status. In particular, patients with tumors with no ER and PgR expression (endocrine-unresponsive disease) at the baseline core-biopsy had a significantly higher response rate to chemotherapy if compared with tumors with some ER/PgR expression. In fact, the absence of ER and PgR expression was the strongest predictors of pCR at the multivariate analysis (P