[Predictive factors for bone metastases of prostate cancer].

Objective: To investigate the correlation between a diverse of clinical factors and bone metastases of prostate cancer.

Methods: The clinical data of 80 patients with prostate cancer were collected and analyzed. The correlations of age, alkaline phosphotase (ALP), prostate specific antigen (PSA), erythrocyte sedimentation rate (ESR), Gleason score, and expressions of androgen receptor (AR) and Ki-67 with bone metastases were analyzed by one-way ANOVA and Logistic regression analysis. The cutoff value, sensitivity and specificity of the independent correlation factors were calculated.

Results: Forty-five of the 80 patients (56%) were found to have bone metastasis, who had significantly older age and higher levels of ALP, PSA, ESR, Gleason score, and expressions of AR and Ki-67 than those without bone metastasis (P<0.05). Logistic regression analysis identified PSA, Gleason score and AR expression as independent factors correlated with bone metastasis with OR (95% CI) of 1.005 (1.001, 1.009) (P=0.008), 5.356 (1.431, 20.039) (P=0.013), and 18.594 (2.460, 140.524) (P=0.005), respectively. The cutoff values of PSA, Gleason Score and AR were 67.1 ng/ml, 7.5, and 2.5, respectively; their sensitivities were 55.6%, 75.6%, and 84.0% for predicting bone metastasis with specificities of 97.1%, 82.9%, and 91.4%, respectively.

Conclusion: Of the factors analyzed, PSA, Gleason score and AR expression, but not age, ALP, PSA, ESR, or Ki-67 expression, are the predictive factors of bone metastasis of prostate cancer.