Abstract
Abstract Various findings implicate sex hormones in prostate growth and development and also in prostate carcinogenesis. We investigated if addition of sex steroid hormone and sex hormone binding globulin (SHBG) serum levels to standard risk assessment parameters [prostate-specific antigen (PSA), free PSA percentage (fPSA%), and age] improves prostate cancer prediction in a PSA screening setting. Steroid hormones testosterone (T), free testosterone (fT), and estradiol (E2), and binding protein SHBG levels were measured in 762 men undergoing prostate biopsy due to suspect PSA serum levels. Prostate cancer was diagnosed in 286 (37.5%) of these men. Our data confirmed that PSA (mean BE=5.09; mean CA=6.05; p=1.24×10-5), fPSA% (mean BE=22.08; mean CA=18.67; p=1.97×10-7), and age (mean BE=60.64; mean CA=64.5; p=7.05×10-10) differentiate men with cancer (CA) and men with benign disease (BE), such as benign prostate hyperplasia. In addition, SHBG (mean BE=50.3; mean CA=54.9; p=0.008) also differed statistically significantly between these two groups. All hormones except E2 and tumor markers correlated significantly with age (T: ρ=-0.09; fT: ρ=-0.27; SHBG: ρ=0.21; PSA: ρ=0.32; and fPSA%: ρ=0.22). Furthermore, we found that PSA correlates with E2 (ρ=0.08), and fPSA% with SHBG (ρ=0.1) and fT (ρ=-0.09). Addition of hormones and SHBG to a baseline marker model including PSA, fPSA%, and age improved cancer prediction in three multivariate classification methods; however, the improvement was minimal. The best improvement by 0.8% was obtained in the logistic regression model with the addition of T and SHBG or of E2 and SHBG, or in the support vector machine model with the addition of SHBG and all steroid hormones to the combination of standard markers PSA, fPSA%, and age; however, this additional gain of accuracy is too small to justify the additional efforts and costs.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call