PI-RADS≤3分患者预测前列腺穿刺阳性列线图模型的构建  

Construction of a nomogram for predicting the risk of positiver prostate biospy with PI-RADS≤3

作　　者：张泳欣 王忠泉 张水兴 肖学红[1] 杨昂[1] 唐秉航[1] 黄红星 袁润强 卢扬柏 Zhang Yongxin;Wang Zhongquan;Zhang Shuixing;Xiao Xuehong;Yang Ang;Tang Binghang;Huang Hongxing;Yuan Runqiang;Lu Yangbai(Department of Radiology,Zhongshan City People's Hospital Afiliated to Sun Yat sen University,Guangdong 528403,China;Department of Kangyi VIP Outpatient Clinic,Zhongshan City People's Hospital Afiliated to Sun Yat sen University,Guangdong 528403,China;Department of Urology,Zhongshan City People's Hospital Afiliated to Sun Yat sen University,Guangdong 528403,China;Department of Radiology,the First Affiliated Hospital of Jinan University,Guangzhou 510630,China)

机构地区：[1]中山市人民医院影像中心,广东528403 [2]中山市人民医院特诊中心门诊部,广东528403 [3]中山市人民医院泌尿外科,广东528403 [4]暨南大学附属第一医院影像中心,广州510630

出　　处：《中华腔镜泌尿外科杂志（电子版）》2022年第6期501-507,共7页Chinese Journal of Endourology(Electronic Edition)

基　　金：中山市人民医院放射影像中心重点专科科研项目(T2020016);中山市科技计划项目(2019B1063,2020B1070,2020B1073);中山市人民医院2022年度医院科研基金重大项目(BG20228249)中山市人民医院高水平医院建设项目-泌尿外科(G330102097008)。

摘　　要：目的探讨前列腺影像报告及数据系统(PI-RADS.v2.1)联合前列腺特异性抗原及其他参数构建的列线图模型对PI-RADS≤3分患者活检阳性的预测价值。方法回顾性分析2018年1月至2021年12月198例在中山市人民医院接受经直肠超声穿刺前列腺首次活检患者的临床血清学和影像学资料,应用Logistic多因素回归分析前列腺癌相关独立风险因素,并构建对前列腺PI-RADS≤3分病变的列线图模型,利用受试者工作曲线、校准曲线和决策曲线对模型进行评估。结果多因素Logistic回归分析显示年龄(P<0.001)PI-RADS(P=0.017)游离PSA/总PSA(FPSA/TPSA)(P=0.049)及移行带体积(TZV)(P<0.001)是前列腺癌的独立危险因素。基于多变量构建的融合模型效能最优(AUC=0.823,95%CI=0.762~0.885),敏感性81.3%,特异性78.8%,准确性79.8%。校准曲线显示其预测概率与病理结果有良好的一致性。决策曲线显示模型具有良好的临床应用价值。结论基于多变量构建的列线图及预测模型能较好地术前预测患前列腺癌的风险。Objective To investigate the predictive value of the nomogram model based on prostate imaging reporting and data system(PI-RADS v2.1)combined with prostate-specific antigen and other parameters for puncture in patients with PI-RADS≤3.MethodsThe clinical serological and imaging data of 198 patients who underwent transrectal ultrasound for the first prostate biopsy in Zhongshan People's Hospital from January 2018 to December 2021 were retrospectively analyzed,and the risk scores were analyzed by Logistic multifactor regression.The independent risk factors related to prostate cancer were analyzed,and the rosette model of prostate PI-RADS≤3 was constructed,and the model was evaluated by the subject operating curve,calibration curve and decision curve.ResultsMultivariate logistic regression analysis showed that age(P<0.001),PI-RADS(P=0.017),FPSA/TPSA(P=0.049)and TZV(P<0.001)were statistically significant independent risk factors for prostate cancer.The fusion model based on multivariable construction had the best performance,with(AUC-0.823,95%CI-0.762-0.885),sensitivity 81.3%,specificity 78.8%,accuracy 79.8%.The calibration curve showed a good agreement between the predicted probabilities of fusion model and pathologic findings.The decision curve model had good clinical application value.ConclusionnThe nomogram and prediction model can better predict the riskof prostate cancerbefore surgery.

关 键 词：育前列腺癌 PI-RADS 列线图 预测模型 

分 类 号：R737.25[医药卫生—肿瘤]