Abstract
Objective: We aimed to build a nomogram allowing to predict the probability of prostate cancer (PC) after an initial 21-core biopsy and with readily available clinical data. Methods: 1,490 screened men who underwent an initial 21-core biopsy protocol were included. A multivariate logistic regression was realized including age, prostate volume, prostate-specific antigen (PSA) level, digital rectal examination (DRE) and transrectal ultrasonography (TRUS). Receiver-operating characteristic estimates were used to quantify accuracy of each model. Results: PC was detected in 41.3% of the patients. Median PSA, age and prostate volume were 6.2 ng/ml (range 0.2-50), 64.6 years (range 33-87) and 40 ml (range 10-270), respectively. Abnormal TRUS findings were detected in 14.7% of patients. Age, PSA level, prostate volume, DRE and TRUS were significantly associated with PC (all p ≤ 0.004) in univariable logistic regression analysis. In multivariate logistic regression analysis, significant associations were found for age, PSA level, prostate volume and DRE. Predictive accuracy estimate of this model was equal to 0.70. TRUS was not an independent predictor of PC. Conclusions: We constructed the first prebiopsy predictive nomogram based on an extended 21-core biopsy procedure with age, PSA level, DRE and prostate volume which are readily available clinical data to urologists.