Background: Androgen-deprivation remains the standard of care for metastatic prostate cancer, yet its total impact on the course of disease is incompletely established.
Methods: We have examined the long-term effects of castration upon the progression of established cancer in the TRAMP transgenic mouse model of prostate cancer. Mice castrated at 15-weeks of age, as well as intact littermates, were followed until spontaneous death from cancer.
Results: Statistical analyses of age-at-death versus primary tumor mass revealed that mice segregate into two categories of response to androgen-deprivation. In Category One, the act of castration paradoxically triggers the growth of microscopic androgen-independent lesions, as evidenced by a statistical synchronization of primary tumor growth. Delaying castration until 20-weeks of age delays the synchronized growth of these tumors, as well as the deaths of the host mice. In Category Two, castration eliminates or substantially delays primary tumor growth, but fails to eliminate metastases. so that castration is found to impart no long-term survival advantage.
Conclusions: We propose a two-step model for the alteration of androgen signaling in prostate cancer capable of explaining the above paradoxical results, which is based upon the dualistic role of androgens as both survival and differentiation factors. This model makes specific predictions for clinical intervention that are discussed in light of published studies.
(c) 2004 Wiley-Liss, Inc.