Validation study of a prostate health index-based nomogram model for predicting prostate cancer

Abstract

Objective To validate a nomogram model based on prostate health index (PHI) for predicting prostate cancer (PCa). Methods The pre-operation serum and clinical data were collected for suspected PCa patients (aged 34 to 90 years), who visited Peking University First Hospital from August 2015 to May 2017 and received transrectal ultrasound-guided prostate biopsy. A total of 391 suspected PCa with total prostate-specific antigen (tPSA)>4 ng/ml were selected into this study, including 235 cases with tPSA level of 4-10 ng/ml and 156 cases with tPSA>10 ng/ml. The p2PSA was tested in all cases and then PHI was calculated. The biopsy results were considered as the gold standard to diagnose PCa. The nomogram model established in Shanghai based on PHI, age and prostate volume was validated in all cases enrolled in this study. Receiver operator curves (ROC) were used to assess the ability of nomogram model to predict PCa. Results Of 391 male patients included in this study, 175(44.8%)were finally diagnosed as PCa. ROC curves indicated that, the area under the curve (AUC) of the nomogram model for predicting PCa among 391 cases was higher than that of the traditional indicator tPSA (AUC: 0.786 vs 0.578, P<0.000 1). And f/t(AUC: 0.786 vs 0.672, P=0.000 2). For those people with tPSA level of 4-10 ng/ml, the AUC of the nomogram model was also higher than that of tPSA (AUC: 0.720 vs 0.513, P=0.000 3) and f/t(AUC: 0.720 vs 0.626, P=0.042 5). Conclusion The nomogram model based prostate health index (PHI) was validated to have a good auxiliary diagnostic value for PCa in our center.(Chin J Lab Med, 2018, 41: 536-540) Key words: Prostatic neoplasms; Prostate-specific antigen; Protein Precursors; ROC curve; Health status