[Low serum testosterone level does not predict bone metastasis of prostate cancer]

To evaluate the role of the serum testosterone level as an independent predictor of bone metastasis of prostate cancer.

This study included 165 male patients with prostate cancer confirmed by biopsy. The patients were aged 58－78 (66.6±5.3) years and none had received androgen-deprivation therapy, chemotherapy or radiotherapy previously. We obtained the baseline clinical data from the patients, including prostate biopsy Gleason scores and the levels of serum prostate-specific antigen (PSA), total testosterone (TT), luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol (E2), alkaline phosphatase (ALP) and prolactin. According to the results of bone scanning, we divided the patients into a bone metastasis and a non-bone metastasis group and screened out the differential factors by univariate analysis and the independent predictor of bone metastasis using the multivariate non-conditional logistic regression model.

Univariate analysis showed no statistically significant differences between the bone metastasis and non-bone metastasis groups in age (P = 0.126) or the levels of serum LH (P = 0.930), FSH (P = 0.763) and E2 (P = 0.256), but that the former had remarkably higher Gleason scores (P < 0.01), total PSA (P <0.01) and ALP (P <0.01) but a lower TT level than the latter (P = 0.013). According to the results of multivariate logistic regression analysis, serum ALP (P <0.01, OR = 1.018 ［1.011－1.026］) and total PSA (P <0.01, OR = 1.029 ［1.015－1.044］) could be regarded as independent predictors of bone metastasis of prostate cancer but not low serum TT (P = 0.531, OR = 0.999 ［0.996－1.002］) or biopsy Gleason score (P = 0.898, OR = 0.787 ［0.412－1.9559］).

The low level of serum testosterone is closely associated with but not an independent predictor of bone metastasis of prostate cancer.