Prevention of prostate cancer by androgens:: Experimental paradox or clinical reality

Androgen replacement therapy in the aging male with partial androgen deficiency improved quality of life. However, such treatment is prohibited for men with a preexisting prostate cancer. The possibility of an increased risk of prostate cancer for healthy men has also been suggested on theoretical basis but recent experimental data showed that androgens may act in prevention of prostate cancer. In this review, we try to evaluate benefits and risks associated to a hormonal replacement therapy in regard to recent data. Several studies analyzing the role of testosterone for prostatic epithelial cells evidenced that testosterone acts in prostatic cell differentiation but does not have a direct role for induction of cell proliferation. Moreover, clinical studies have shown that low free testosterone levels in serum is associated with aggressive prostate cancer, like that has been observed in men with prostate cancer under prostate cancer chemo-prevention by finasteride. These data suggest that an androgen pathway disruption in prostate is responsible of cell deregulations that may be associated not only with apoptosis of differentiated prostatic cells but also with potential cell transformation. The effects of androgens withdrawal for prostate cancer therapy induced in a short time the tumor arrest growth. However with time, cells adapt to low levels of androgens leading to the evolution of an androgen-independent tumor, which is more aggressive and most often fatal. The molecular mechanisms of this evolution begin to merge. A hypothesis is that such mechanisms could be initiated in elderly men with an androgen deficiency. The question is raised of whether hormonal replacement therapy could prevent prostate cancer. An encouraging recent study performed on rats demonstrated a protective effect of DHEA for prostate cancer. However, the putative role of the normalization of DHEA or other androgen levels in prevention of prostate cancer should be evaluated in clinical trials. (C) 2004 Elsevier B.V. All rights reserved.