External validation of a PCA-3-based nomogram for predicting prostate cancer and high-grade cancer on initial prostate biopsy

Abstract

INTRODUCTIONThe aim of this study was to externally validate a previously developed PCA3-based nomogram for the prediction of prostate cancer (PCa) and high-grade (intermediate and/or high-grade) prostate cancer (HGPCa) at the time of initial prostate biopsy. METHODSA retrospective review was performed on a cohort of 336 men from a large urban academic medical center. All men had serum PSA <20ng/ml and underwent initial transrectal ultrasound-guided prostate biopsy with at least 10 cores sampling for suspicious exam and/or elevated PSA. Covariates were collected for the nomogram and included age, ethnicity, family history (FH) of PCa, PSA at diagnosis, PCA3, total prostate volume (TPV), and abnormal finding on digital rectal exam (DRE). These variables were used to test the accuracy (concordance index) and calibration of a previously published PCA3 nomogram. RESULTSBiopsy confirms PCa and HGPCa in 51.0% and 30.4% of validation patients, respectively. This differed from the original cohort in that it had significantly more PCa and HGPCA (51% vs. 44%, P=0.019; and 30.4% vs. 19.1%, P<0.001). Despite the differences in PCa detection the concordance index was 75% and 77% for overall PCa and HGPCa, respectively. Calibration for overall PCa was good. CONCLUSIONSThis represents the first external validation of a PCA3-based prostate cancer predictive nomogram in a North American population. Prostate 76:1019-1023, 2016. (c) 2016 Wiley Periodicals, Inc.