MRI阴性患者前列腺穿刺活检阳性的相关危险因素分析及预测模型的建立

Analysis of risk factors related to positive prostate biopsy in MRI-negative patients and establishment of a prediction model

目的探究磁共振成像(MRI)影像表现为阴性的患者在超声引导下行经直肠前列腺穿刺活检为阳性的危险因素,建立Logistic回归预测模型。方法回顾性分析2011年1月至2018年6月在西安交通大学第一附属医院检查MRI为阴性且行前列腺穿刺活检患者的临床资料,采用单因素分析和Logistic回归分析法筛选相关危险因素,建立Logistic回归模型,通过受试者工作特征(ROC)曲线下面积(AUC)验证效能。结果本研究共纳入227例患者,单因素分析显示年龄、总前列腺特异性抗原(tPSA)、游离/总前列腺特异性抗原(f/t PSA)、前列腺体积(PV)和tPSA密度(PSAD)与MRI阴性且活检阳性的发生显著相关(P<0.05)。多因素分析表明年龄(OR=1.058,95%CI:1.010~1.108,P=0.017)、tPSA(OR=1.039,95%CI:1.014~1.065,P=0.002)、PV(OR=0.978,95%CI:0.962~0.994,P=0.008)和PSAD(OR=0.489,95%CI:0.313~0.764,P=0.002)是MRI阴性但穿刺活检阳性的独立危险因素。根据多因素分析结果建立预测模型:Logit P=IN{P/(1-P)}=-3.559+0.056×年龄+0.038×tPSA-0.023×PV-0.716×PSAD。当预测概率P>0.20时,患者穿刺结果为阳性的可能性较大,模型预测概率的AUC为0.774。结论年龄、tPSA、PV和PSAD是MRI阴性患者穿刺活检呈阳性的预测因素,运用预测模型可为临床医生决策提供更好的理论依据。

Objective To investigate the risk factors related to positive ultrasound-guided transrectal prostate biopsy in patients with MRI-negative images and to develop a Logistic regression model.Methods Clinical data of patients with MRI-negative images who underwent ultrasound-guided transrectal prostate biopsy during Jan.2011 and Jun.2018 at our hospital were retrospectively analyzed.The relevant risk factors were screened with univariate analysis and Logistic regression analysis,and a Logistic regression prediction model was established.The effectiveness of the prediction model was verified with the area under the ROC curve(AUC).Results A total of 227 patients were involved.Univariate analysis showed a significant correlation between age,tPSA,f/t PSA,PV,PSAD,and negative MRI but positive biopsy(P<0.05).Multivariable analysis showed that age(OR=1.058,95%CI:1.010-1.108,P=0.017),tPSA(OR=1.039,95%CI:1.014-1.065,P=0.002),PV(OR=0.978,95%CI:0.962-0.994,P=0.008),and PSAD(OR=0.489,95%CI:0.313-0.764,P=0.002)were independent risk factors of negative MRI but positive biopsy(P<0.05).A prediction model was established based on multivariable analysis:Logit P=IN{P/(1-P)}=-3.559+0.056×Age+0.038×tPSA-0.023×PV-0.716×PSAD.When P>0.20,the biopsy was more likely to be positive,and the AUC was 0.774.Conclusion Age,tPSA,PV and PSAD are predictors of positive biopsy in MRI-negative patients,and the prediction model may provide a good theoretical basis for clinicians' decision-making.