External validation of prostate health index-based nomogram for predicting prostate cancer at extended biopsy

Abstract

Introduction/Objective. Prostate Health Index (PHI)-based nomograms were created by Lughezzani et al. (2012) and Zhu et al. (2015) for predicting prostate cancer (PCa) at extended biopsy. The aim of the study was to externally validate two nomograms in the Serbian population. Methods. This retrospective study comprised 71 patients irrespective of digital rectal examination (DRE) findings, with prostate-specific antigen level < 10 ng/ml, who had undergone prostate biopsies, and PHI testing. Data were collected in accordance with previous nomograms predictors. Independent predictors were identified by using logistic regression. The predictive accuracy was measured by the area under the receiver operating characteristic curve (AUC). The calibration belt was used to assess model calibration. The clinical utility was measured by using decision curve analysis (DCA). Results. There were numerous differences in underlying risk factors between validation dataset and previously available data. Analysis demonstrated that the DRE and PHI were independent predictors. AUCs for both nomograms, in patients with normal DRE had shown to have a good discriminatory ability (77.2?86.2%). In the entire population AUC of nomogram had exceptional discrimination (92.9%). Zhu et al. nomogram is associated with lower false positive predictions. The calibration belt for Zhu et al. nomogram was acceptable. Our DCA suggested that both nomograms are likely to be clinically useful. Conclusion. We performed external validation of two PHI-based nomograms predicting the presence of PCa in both the initial and the repeat biopsy setting. The PHI-based nomograms displayed adequate accuracy and justifies its use in Serbian patients.