基于多因素logistic回归模型预测前列腺特异性抗原4～20ng/mL时移行带前列腺癌
Predicting transitional zone prostate cancer with prostate specific antigen of 4-20 ng/m L based on multivariate logistic regression model
聂洋平 1李云1
作者信息
- 1. 郑州市第七人民医院MR科,河南 郑州 450000
- 折叠
摘要
目的 基于前列腺影像报告和数据系统(PI-RADS)v2.1与前列腺特异性抗原(PSA)及其衍生指标建立多因素logistic回归模型,探讨PSA 4～20 ng/m L时对移行带前列腺癌的预测价值.方法 回顾性选取PSA 在4～20 ng/m L的前列腺疾病患者104例,均进行穿刺活检术或根治性前列腺切除术.依据病理结果将患者分为前列腺癌组(30例)和非前列腺癌(前列腺增生或前列腺炎)组(74例).比较2组患者总前列腺特异性抗原(t PSA)、游离前列腺特异性抗原(f PSA)、f PSA与t PSA比值(f PSA/t PSA)、前列腺特异性抗原密度(PSAD)、PI-RADS v2.1评分等指标的差别.建立多因素logistic回归模型,分析患癌的危险因素,使用受试者工作特征(ROC)曲线评价诊断效能.结果 2组患者的t PSA、f PSA/t PSA、PSAD及PI-RADS v2.1评分间差异均有统计学意义(P<0.05).logistic回归分析显示,PSAD、PI-RADS v2.1评分是预测移行带前列腺癌的独立危险因素.模型的曲线下面积(AUC)为0.938,明显高于任一指标,其敏感度为90.00%,特异度为86.49%,准确性为86.54%.结论 基于PI-RADS v2.1与PSAD的logistic回归模型可以提升移行带前列腺癌的诊断效能,为可疑前列腺癌患者的穿刺活检提供更有力的证据,避免过度诊断以及不必要的活检.
Abstract
Objective To establish a multivariate logistic regression model based on prostate imaging reporting and data system(PI-RADS)v2.1 and prostate specific antigen(PSA)and its derived indicators,and to investigate the predictive value of PSA 4-20 ng/m L for transitional zone prostate cancer.Methods A total of 104 patients with prostate disease and PSA of 4-20 ng/m L were retospec-tively selected.All patients underwent puncture biopsy or radical prostatectomy.In accordance with the pathological findings,patients were divided into prostate cancer group(30 cases)and non-prostate cancer(prostatic hyperplasia or prostatitis)group(74 cases).The differences of total prostate specific antigen(t PSA),free prostate specific antigen(f PSA),f PSA/t PSA,prostate specific antigen density(PSAD)and PI-RADS v2.1 scores were compared between the two groups.A multivariate logistic regression model was con-structed to analyze the risk factors of cancer.The receiver operating characteristic(ROC)curve was used to evaluate the diagnostic efficiency.Results There were statistically differences in t PSA,f PSA/t PSA,PSAD,and PI-RADS v2.1 scores between the two groups(P<0.05).Logistic regression analysis indicated that PSAD and PI-RADS v2.1 scores were independent risk factors for pre-dicting transitional zone prostate cancer.The area under the curve(AUC)of the model was 0.938,which was significantly higher than any individual index,with sensitivity of 90.00%,specificity of 86.49%,and accuracy of 86.54%.Conclusion The logistic regression model based on PI-RADS v2.1 and PSAD can enhance the diagnostic efficiency for transitional zone prostate cancer.It provides stron-ger evidence for the needle biopsy of patients with suspicious prostate cancer and helps to avoid overdiagnosis and unnecessary biopsy.
关键词
前列腺癌/前列腺影像报告和数据系统/前列腺特异性抗原/logistic回归Key words
prostate cancer/prostate imaging reporting and data system/prostate specific antigen/logistic regression引用本文复制引用
出版年
2025
NETL
NSTL
万方数据