Prediction model for extraprostatic extension of prostate based on MRI and clinical indicators
10.3969/j.issn.1006-7795.2025.02.010
- VernacularTitle:基于MRI及临床指标的前列腺被膜外侵犯预测模型
- Author:
Yunpeng FAN
1
;
Tianyu XIONG
;
Kun YANG
;
Zhanliang LIU
;
Song JIN
;
Ping XIE
;
Yinong NIU
Author Information
1. 首都医科大学附属北京友谊医院泌尿外科,北京 100050;北京市卫生健康委员会泌尿外科研究所,北京 100050
- Publication Type:Journal Article
- Keywords:
prostate cancer;
magnetic resonance imaging;
prediction model;
Nomogram;
neoplasm grading;
extraprostatic extension
- From:
Journal of Capital Medical University
2025;46(2):243-251
- CountryChina
- Language:Chinese
-
Abstract:
Objective To develop a Nomogram clinical prediction model for the pathological occurrence of extraprostatic extension(EPE)after radical prostatectomy in prostate cancer patients,using simplified site-specific magnetic resonance imaging(MRI)indicators and other clinical parameters.Methods A total of 181 prostate cancer patients[mean age(69.0±7.3)years]who underwent radical prostatectomy were included.These patients had received 3-Tesla multi-parametric magnetic resonance imaging(3-T mpMRI)within 6 months prior to surgery.Based on mpMRI measurements[capsular contact length(CCL)>15 mm,capsular bulging/irregularities,diameter of index lesion(dIL),and evident extraprostatic extension(eEPE)],the dIL?sEPE grading system was derived.The optimal cut-off value of dIL(denoted as dIL)was determined using the Youden J index,and categorized it into a binary variable.A Logistic regression model was established based on the dIL?sEPE grading and clinical scores.The predictive performance of clinical indicators,MRI indicators,and combined clinical and MRI indicators were compared.Finally,a clinical prediction model integrating both clinical and MRI data was developed.Results Pathological EPE was found in 46 out of 181 cases(25.4% ).A Nomogram prediction model for EPE was established with a combination of the dIL?sEPE grading and clinical indicators.Conclusion The combination of dIL?sEPE grading with clinical indicators accurately predicts extracapsular extension in prostate cancer.The Nomogram model that established,based on MRI imaging characteristics and clinical indicators has good performance and is easy to use.It is beneficial to stratifying management for prostate cancer patients,and it provides valuable guidance for patients suitable for nerve-sparing surgery.
