Abstract
Aggressive prostate cancer (PC) represents a significant health concern. Conventional screening methods, primarily based on prostate-specific antigen (PSA) levels, lack specificity, leading to an urgent need for more accurate diagnostic tools. This study investigates whether integrating clinical and routine blood laboratory parameters can improve the early non-invasive prediction of aggressive PC. In a pilot study of 578 patients with suspicion of PC, 28 laboratory values alongside data on family history, diet, and lifestyle were analyzed. A logistic regression classifier was developed, with model performance evaluated using repeated k-fold cross-validation on the complete dataset (n = 282). Participants were categorized into healthy, moderate PC (ISUP 1-2), and aggressive PC (ISUP 3-5). Significant associations were found between PC aggressiveness and lower levels of androstenedione, dehydroepiandrosterone-sulfate (DHEA-S) and free PSA%, as well as higher levels of sex hormone binding globulin (SHBG). The integration of these serum markers with clinical parameters into a new multi-stage risk classifier significantly improved the predictive accuracy for aggressive PC, outperforming PSA-only methods. The integration of DHEA-S, androstenedione, and SHBG as widely available and cost-effective novel blood biomarkers offers a more targeted, non-invasive prediction of aggressive PC. This approach could reduce reliance on invasive prostate biopsies and expensive magnetic resonance imaging.