Résumé
Purpose A growing body of evidence suggests that low testosterone can be an independent predictor of adverse clinicopathological features and worse prognosis in prostate cancer (PCa) patients. However, this association is still incompletely understood and the results are divisive. The aim of this study was to analyze testosterone as a predictor of aggressive disease in subjects with clinically localized PCa. Materials and Methods A cohort was conducted including the patients submitted to radical prostatectomy in our institution during a period of four years. The patients had clinically localized disease and their total testosterone (TT) was routinely measured preoperatively in the morning before surgery. They were stratified in groups with low (< 300 ng/dL) and normal TT (≥ 300 ng/dL). Tumor aggressiveness was inferred based on preoperative PSA levels, pathological Gleason score (lower, equal or greater than 7), TNM stage and surgical margins status. Results After analyzing 164 patients we found a significant association between mean preoperative TT and extraprostatic disease (379 for pT3 vs. 421 ng/for pT2 - p < 0.001, AUC > 0.99). Conversely, men with high Gleason score had similar mean TT compared to those with lower scores. Preoperative low TT (defined as TT < 300 ng/dL) could not be statistically correlated with either preoperative PSA levels, pathological Gleason score, extraprostatic extension, positive surgical margins or seminal vesicles involvement. Conclusions This study indicates that testosterone may be a useful predictive tool once pathological extraprostatic extension was somewhat signaled by lower TT levels preoperatively. However, it does not consolidate a clear association between aggressive tumor biology and hypogonadism. .