Development of improved nomogram for prediction of outcome of initial prostate biopsy using readily available clinical information.

OBJECTIVES: To construct a nomogram that can be used to estimate the risk of prostate cancer (PCa) and high-grade PCa using readily available clinical information for men undergoing initial extended prostate biopsy (PBx). Many nomograms have been developed to predict the outcome of initial PBx. However, most require information not available at the decision to biopsy.
METHODS: From March 2000 to April 2010, 1551 men with a prostate-specific antigen (PSA) of ≤10 ng/mL who underwent initial extended PBx were included in the present study. The nomogram predictor variables were patient age, race, prostate-specific antigen (PSA) level, percent free PSA, family history of PCa, and the digital rectal examination findings. The area under the receiver operating characteristic curve was calculated as a measure of discrimination. The calibration was assessed graphically.
RESULTS: Of the 1551 men, 606 (39.1%) had PCa on biopsy. The mean value for age, PSA, and percent free PSA was 63.4 years, 5.1 ng/mL, and 21.4%, respectively. Also, 25.1% and 7.8% of patients with positive PBx findings had digital rectal examination abnormalities and a positive family history, respectively. The univariate and multivariate analyses suggested that all 6 risk factors were predictors of PCa in the study cohort (P < .05). The area under the curve for all factors in a model predicting PCa was 0.73 (95% confidence interval 0.71-0.76). The area under the curve for predicting high-grade PCa was 0.71 (95% confidence interval 0.69-0.74).
CONCLUSIONS: The present predictive model allows an assessment of the risk of PCa and high-grade PCa for men undergoing initial extended PBx using readily available, noninvasively obtained clinical data.