Analyzing the metastatic phenotype.

The dissemination of cells from a primary tumor, resulting in the progressive growth of metastatic carcinoma in distant sites, is the most common cause of death of cancer patients. The observations from clinical studies and the results of experimental studies using rodent tumors and human cancer cells implanted into immunodeficient host animals suggest that metastasis is not a random event, but rather the result of a sequence of selective events, many of which involve interactions with elements of the microenvironment of the primary and metastatic tumors. Analysis of the metastatic potential of a human tumor cell population has been greatly improved by the introduction of orthotopic models of tumor growth and metastasis, which have demonstrated that implanting human tumor cells into the appropriate tissue in an immunodeficient rodent can increase both tumor take and incidence of metastasis. These will be the models that should be used to validate the identity of candidate metastasis-associated genes, and to determine the value of new forms of therapy, either genetic or pharmacological, for controlling metastatic cancer growth.