Therapeutic implications of a stem cell model for human breast cancer: a hypothesis.

A model for human breast cancer is proposed in which a subpopulation of relatively undifferentiated tumor stem cells gives rise to clones of cells by a process of exponential division. The progeny progressively become more differentiated but lose their proliferative potential. The model implies that breast cancers are heterogeneous with respect to differentiation features such as hormone receptors, both between different clones and within individual clones. Therefore, endocrine therapies are preferentially directed towards differentiated cells and are unlikely to be curative either in advanced disease or in the adjuvant setting. The reduction of differentiated cells by hormonal maneuvers may leave a population of more actively dividing tumor stem cells which are more sensitive to chemotherapy. Hormonal therapy followed by cytotoxic chemotherapy may be more likely to kill tumor stem cells than either form of therapy given alone.