The Influence of Endogenous Testosterone Density on Unfavorable Disease and Tumor Load at Final Pathology in Intermediate-Risk Prostate Cancer: Results in 338 Patients Treated with Radical Prostatectomy and Extended Pelvic Lymph Node Dissection

Abstract

<b><i>Objective:</i></b> The aim of this study is to evaluate the influence of endogenous testosterone density (ETD) on features of aggressive prostate cancer (PCa) in intermediate-risk disease treated with radical prostatectomy and extended pelvic lymph node dissection. <b><i>Materials and Methods:</i></b> Density measurements included the ratio of endogenous testosterone (ET), prostate-specific antigen (PSA), and percentage of biopsy positive cores (BPC) on prostate volume (ETD, PSAD, and BPCD, respectively). The ratio of percentage of cancer invading the gland (tumor load, TL) on prostate weight (TLD) was also calculated. Unfavorable disease (UD) was defined as tumor upgrading (ISUP &#x3e;3) and/or upstaging (pT &#x3e;2) and/or lymph node invasion (LNI). Associations of ETD with features of aggressive PCa, including UD and TLD, were evaluated by logistic and linear regression models. <b><i>Results:</i></b> Evaluated cases were 338. Subjects with upgrading, upstaging, and LNI were 61/338 (18%), 73/338 (21%), and 25/338 (7.4%), respectively. TLD correlated with UD (Pearson’s correlation coefficient, <i>r</i> = 0.204; <i>p</i> &#x3c; 0.0001), PSAD (<i>r</i> = 0.342; <i>p</i> &#x3c; 0.0001), BPCD (<i>r</i> = 0.364; <i>p</i> &#x3c; 0.0001), and ETD (<i>r</i> = 0.214; <i>p</i> &#x3c; 0.0001), which also correlated with BMI (<i>r</i> = −0.223; <i>p</i> &#x3c; 0.0001), PSAD (<i>r</i> = 0.391; <i>p</i> &#x3c; 0.0001), and BPCD (<i>r</i> = 0.407; <i>p</i> &#x3c; 0.0001). TLD was the strongest independent predictor of UD (OR = 2.244; 95% CI = 1.146–4.395; <i>p</i> = 0.018). In the multivariate linear regression model predicting BPCD, ETD was an independent predictor (linear regression coefficient, <i>b</i> = 0.026; 95% CI: 0.016–0.036; <i>p</i> &#x3c; 0.0001) together with PSAD (<i>b</i> = 1.599; 95% CI: 0.863–2.334; <i>p</i> &#x3c; 0.0001) and TLD (<i>b</i> = 0.489; 95% CI: 0.274–0.706; <i>p</i> &#x3c; 0.0001). According to models, TLD increased as ETD increased accordingly, but mean ET levels were significantly lower for patients with UD. <b><i>Conclusions:</i></b> As ETD measurements incremented, the risk of large tumors extending beyond the prostate increased accordingly, and patients with lower ET levels were more likely to occult UD. The influence of ETD on PCa biology should be addressed by prospective studies.