Yueyue Zhang1,2, Guiqi Zhu3, Wenlu Zhao1, Chaogang Wei1, Tong Chen1, Qi Ma4, Yongsheng Zhang5, Boxin Xue6, Junkang Shen1,2. 1. Department of Radiology, The Second Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215004, People's Republic of China. 2. Department of Radiotherapy Institute, The Second Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215004, People's Republic of China. 3. Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai 200032, People's Republic of China. 4. Department of Ultrasound, The Second Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215004, People's Republic of China. 5. Department of Pathology, The Second Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215004, People's Republic of China. 6. Department of Urology, The Second Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215004, People's Republic of China.
Abstract
PURPOSE: To develop and validate a PI-RADS-based nomogram for predicting the probability of clinically significant prostate cancer (csPCa) at initial prostate biopsy. PATIENTS AND METHODS: From February 2015 to October 2018, 573 consecutive patients made up the development cohort (DC), and another 253 patients were included as an independent validation cohort (VC). Univariate and multivariate analysis were used for determining the dependent clinical risk factors for csPCa. Prediction model1 was constructed by integrating independent clinical risk factors. Then added the PI-RADS score to model1 to develop the prediction model2 and present it in the form of a nomogram. The performance of the nomogram was assessed by receiver operating characteristic curve, net reclassification improvement analysis, calibration curve, and decision curve. RESULTS: All clinical candidate factors were significantly different between csPCa and non-csPCa in both the DC and VC. Age, PSA density (PSAD), and free-to-total PSA ratio (f/t) were ultimately determined as dependent clinical risk factors for csPCa and integrated into prediction model1. Then, prediction model2 was developed and presented in a nomogram. In the DC, the nomogram (AUC=0.894) was superior to model1, PI-RADS score, or other clinical factors alone in detecting csPCa. Similar result (AUC=0.891) was obtained in the VC. NRI analysis showed that the nomogram improved the classification of patients significantly compared with model1. Furthermore, the nomogram showed favorable calibration and great clinical usefulness. CONCLUSION: This study developed and validated a nomogram that integrates PI-RADS score with other independent clinical risk factors to facilitate prebiopsy individualized prediction in high-risk patients with csPCa.
PURPOSE: To develop and validate a PI-RADS-based nomogram for predicting the probability of clinically significant prostate cancer (csPCa) at initial prostate biopsy. PATIENTS AND METHODS: From February 2015 to October 2018, 573 consecutive patients made up the development cohort (DC), and another 253 patients were included as an independent validation cohort (VC). Univariate and multivariate analysis were used for determining the dependent clinical risk factors for csPCa. Prediction model1 was constructed by integrating independent clinical risk factors. Then added the PI-RADS score to model1 to develop the prediction model2 and present it in the form of a nomogram. The performance of the nomogram was assessed by receiver operating characteristic curve, net reclassification improvement analysis, calibration curve, and decision curve. RESULTS: All clinical candidate factors were significantly different between csPCa and non-csPCa in both the DC and VC. Age, PSA density (PSAD), and free-to-total PSA ratio (f/t) were ultimately determined as dependent clinical risk factors for csPCa and integrated into prediction model1. Then, prediction model2 was developed and presented in a nomogram. In the DC, the nomogram (AUC=0.894) was superior to model1, PI-RADS score, or other clinical factors alone in detecting csPCa. Similar result (AUC=0.891) was obtained in the VC. NRI analysis showed that the nomogram improved the classification of patients significantly compared with model1. Furthermore, the nomogram showed favorable calibration and great clinical usefulness. CONCLUSION: This study developed and validated a nomogram that integrates PI-RADS score with other independent clinical risk factors to facilitate prebiopsy individualized prediction in high-risk patients with csPCa.
Authors: Leander Van Neste; Rianne J Hendriks; Siebren Dijkstra; Geert Trooskens; Erik B Cornel; Sander A Jannink; Hans de Jong; Daphne Hessels; Frank P Smit; Willem J G Melchers; Gisèle H J M Leyten; Theo M de Reijke; Henk Vergunst; Paul Kil; Ben C Knipscheer; Christina A Hulsbergen-van de Kaa; Peter F A Mulders; Inge M van Oort; Wim Van Criekinge; Jack A Schalken Journal: Eur Urol Date: 2016-04-20 Impact factor: 20.096