基于影像学和血清学特征构建列线图模型对PSA 4~10 ng/mL患者穿刺阳性的预测价值  

Predictive value of a nomogram model constructed based on imaging and serological characteristics for prostate biopsy positivity in patients with PSA levels of 4-10 ng/mL

作　　者：梁建嫦 吴锦锋 张泳欣 沈俊鑫 陈振杰 谭健 钟睿 赵卫[4] 卢扬柏[2] 袁润强[2] LIANG Jianchang;WU Jinfeng;ZHANG Yongxin;SHEN Junxin;CHEN Zhenjie;TAN Jian;ZHONG Rui;ZHAO Wei;LUYangbai;YUAN Runqiang(Department of Clinical Laboratory,Yunfu Hospital of TCM,Yunfu 527300,China;Department of Urology,Zhongshan City People's Hospital,Zhongshan 528403,China;The First Clinical Medical College,Guangdong Medical University,Zhanjiang 524023,China;School of Public Health,Southern Medical University,Guangzhou 510515,China;Imaging Center,Zhongshan City People's Hospital,Zhongshan 528403,China)

机构地区：[1]云浮市中医院检验科,广东云浮527300 [2]中山市人民医院泌尿外科,广东中山528403 [3]广东医科大学第一临床医学院,广东湛江524023 [4]南方医科大学公共卫生学院,广东广州510515 [5]中山市人民医院影像中心,广东中山528403

出　　处：《中国医学物理学杂志》2024年第12期1494-1500,共7页Chinese Journal of Medical Physics

基　　金：国家重点研发计划(2018YFC1602206);广东省基础与应用基础研究基金(2022A1515220032);广东省医学科学技术研究基金(B2023195);中山市科技计划项目(2020B1073);中山市人民医院重大科研基金(BG20228249);中山市人民医院优秀青年项目(SG2023106);中山市第三批社会公益与基础研究项目(2023B3006)。

摘　　要：目的:探讨基于影像学联合前列腺特异性抗原(PSA)及其相关参数构建的列线图模型对PSA 4~10 ng/mL患者穿刺阳性的预测价值。方法:回顾性分析2018年1月至2023年12月191例在中山市人民医院和/或云浮市中医院行血清学PSA及相关指标检测并接受经直肠超声穿刺前列腺首次活检的患者的临床血清学和影像学资料,应用多因素Logistic回归分析前列腺癌相关独立风险因素,构建PSA 4~10 ng/mL患者的列线图模型,使用受试者工作特征曲线、校准曲线和决策曲线对模型进行评估。结果:多因素Logistic回归分析结果显示游离PSA、前列腺体积、移行带体积、PSA密度及前列腺影像报告和数据系统(PI-RADS v2.1)为前列腺癌的独立风险因素。基于这些显著变量构建的融合模型表现最佳,AUC为0.750(95%CI:0.678~0.821),敏感性为72.7%,特异性为77.2%,准确性为74.9%。校准曲线显示该模型预测的前列腺癌概率与病理结果有良好的一致性;决策曲线分析进一步证明该模型具有较高的临床应用价值。结论:构建的列线图及预测模型在术前能较好地预测PSA 4~10 ng/mL患者前列腺癌的风险,为临床医师提供直观的预估工具,有助于根据前列腺癌发生的风险调整治疗计划,从而优化患者的生存结果。Objective To investigate the predictive value of a nomogrammodel constructed based on imaging combined with prostatespecific antigen(PSA)and its related parameters for biopsy in patients with PSAlevels of 4-10 ng/mL.Methods The serological and imaging data of 191 patients who were detected for PSA and related indicators and underwent the first biopsy of prostate by transrectal ultrasound at Zhongshan City People's Hospital and/or Yunfu Hospital of TCM from January 2018 to December 2023 were analyzed retrospectively.Multivariate Logistic regression identified independent risk factors for prostate cancer,and a nomogram model was developed for patients with PSA levels of 4-10 ng/mL.The predictive performance of the model was evaluated using receiver operating characteristic(ROC)curves,calibration curves,and decision curves.Results The multivariate Logistic regression analysis showed that free PSA,prostate volume,transition zone volume,PSAdensity,and the prostate imagingreporting and data system(PI-RADS v2.1)score were independent risk factors for prostate cancer.The model incorporating these significant variables demonstrated the best performance,with an area under the curve(AUC)of 0.750(95%CI:0.678-0.821),sensitivity of 72.7%,specificity of 77.2%,and accuracy of 74.9%.The calibration curve indicated good agreement between the predicted probability and the actual probability of prostate cancer;and the decision curve analysis further confirmed that the model had high clinical utility.Conclusion The constructed nomogram prediction model can effectively estimate the preoperative risk of prostate cancer in patients with PSA levels of 4-10 ng/mL,providing clinicians with an intuitive tool to adjust treatment plans based on the assessed risk,thereby optimizing patient outcomes.

关 键 词：前列腺癌 列线图模型 前列腺特异性抗原 预测模型 影像学 血清学 

分 类 号：R318[医药卫生—生物医学工程] R737.25[医药卫生—基础医学]