Optimizing Biopsy Decisions in PI-RADS 3-4 Lesions: Integrating PSA-derived Biomarkers to Reduce Unnecessary Procedures

作者：J.S. Kuo, Yin‐Chien Ou, Che‐Yuan Hu, Kuan‐Yu Wu, Chan-Jung Liu, Ze‐Hong Lu, Hau-Chern Jan, Kun-Che Lin, Yi‐Chia Hsieh, Chun‐Quan Ou, Yuh‐Shyan Tsai · 发表于：Anticancer Research · 年份：2025 · DOI：10.21873/anticanres.17904 · 被引用次数：1 · 研究领域：Prostate Cancer Diagnosis and Treatment、Cardiac Imaging and Diagnostics、MRI in cancer diagnosis

BACKGROUND/AIM: Prostate cancer is a common malignancy among men. However, identifying clinically significant prostate cancer (csPCa) remains challenging, particularly in the Prostate Imaging Reporting and Data System (PI-RADS) 3-4 lesions. This study evaluated whether combining prostate-specific antigen (PSA)-derived biomarkers and clinical parameters with multiparametric magnetic resonance imaging (mpMRI) improves risk stratification and reduces unnecessary biopsies. PATIENTS AND METHODS: Comprehensive clinical data of 165 patients who underwent mpMRI and MRI-ultrasound fusion-guided prostate biopsy between 2021 and 2024 was retrospectively analyzed. Patients were stratified by PI-RADS scores into two groups: 3-4 and 5. csPCa was defined as International Society of Urological Pathology grade group 2 or higher. Logistic regression models were used to identify predictors of csPCa. The diagnostic utility of PI-RADS scores alone and in combination with PSA density and PSA velocity was evaluated through receiver operating characteristic curves and decision curve analysis (DCA). RESULTS: =0.034]. In the PI-RADS 3-4 subgroup, the combination of PSA density and PSA velocity achieved an area under the curve of 0.673. DCA showed that the combined model provided the highest net benefit within the 18%-50% threshold range, reducing unnecessary biopsies by 68.0% while missing 33.3% of csPCa cases. CONCLUSION: Integrating PSA density and PSA velocity improves diagnostic performance in PI-...