西方攝護腺癌風險計算應用程式於台灣單一醫學中心之驗證
陳秉韜、陳逸軒、林仁泰、蔡政諭、余家政、吳東霖
高雄榮民總醫院外科部泌尿外科1
國防醫學中心三軍總醫院外科部泌尿外科2
國立陽明大學醫學院3
External Validation of Western Prostate Cancer Risk Calculator Apps in a Taiwanese Tertiary Medical Center
Bing-Tau Chen1,3,4, I-Hsuan Chen1,2,3, Jen-Tai Lin1,3,
Jeng-Yu Tsai1,3, Chia-Cheng Yu1,2,3, Tony Wu1,3
1 Division of Urology, Department of Surgery, Kaohsiung Veterans General Hospital
2 Division of Urology, Department of Surgery,
Tri-Service General Hospital, National Defense Medical Center
3 School of Medicine, National Yang-Ming University
4 Division of Urology, Department of Surgery, Kaohsiung Armed Forces General Hospital
Objective:
The Rotterdam and Coral apps, prostate cancer (PCa) risk calculators, have both showed better predictive accuracy of PCa and high-grade PCa than traditional PSA and/or DRE. The present study attempts to be the first to validate the PCa risk calculator apps using a Taiwanese cohort of patients. Also, to utilize post-prostatectomy pathology outcomes to assess the accuracy of both apps with regard to high-grade PCa.
Methods:
From 2012 to 2018, a total of 1134 men who undergoing prostate biopsies in a tertiary center were retrospectively included in the current study. Comparing the results of biopsy and prostatectomy with the probabilities of PCa and high-grade PCa (Gleason score ≧7), which were calculated using the Rotterdam and Coral apps. Calibration and discrimination were assessed; decision curve analysis was performed.
Results:
Overall, 21.7% (246/1134) patients were diagnosed with PCa, and 63% (155/246) patients had high-grade PCa according to biopsy results. After confirmation with prostatectomy pathological outcomes, 47.2% (25/53) patients were upgraded to high-grade PCa and 1.2% (1/84) patients were downgraded. The discriminative ability for both PCa (AUC: 0.779 vs 0.687, DeLong's method: P<0.001) and high-grade PCa (AUC: 0.862 vs 0.758, P<0.001) was significantly better for the Rotterdam app. In the prostatectomy cohort, there was no significant difference between both apps (AUC: 0.857 vs 0.777, P=0.128).
Conclusions: The Rotterdam and Coral apps can be applied to the Taiwanese cohort. The Rotterdam app outperformed the Coral app in the prediction of PCa and high-grade PCa.