The prognosis for prostate cancer is largely dependent on the probability o
f metastatic dissemination. Prognostic markers currently in use are very po
or predictors of metastatic potential, and as of yet none of the battery of
new molecular markers has proven greatly superior. This may be due in part
to their inability to assess the degree of interaction of subpopulations o
f prostate cancer cells with each ether and with their microenvironment. A
growing body of evidence indicates that these types of interactions are a m
ajor factor in the eventual genesis of cancer cells capable of metastasis.
Recent research has demonstrated that specialized components of prostate tu
mors may play a critical supporting role for the overall growth of the larg
er tumor. The multifocal nature and apparent polyclonal origins of prostate
tumors suggest that carcinogenesis and tumor progression are promoted by g
lobal influences or "field effects." It appears that these effects extend b
eyond the proliferating epithelial component to the tissue stroma. Prostate
cancer cells and stromal cells seem to act in concert to modify the microe
nvironment, leading to metastasis. An understanding of this synergy may pro
vide a new class of prognostic markers which more accurately measure the co
mplex set of interactions that determine tumor behavior. (C) 2000 Elsevier
Science Inc. All rights reserved.