[The preliminary study on the combined application model of serum miRNAs in the diagnosis of pneumoconiosis].

Objective: To explore the role of serum microRNA (miRNAs) levels in the detection of pneumoconiosis, and to establish a combined application model of multiple serums miRNAs for pneumoconiosis diagnosis. Methods: 152 cases were selected in the case group and the control group respectively. The TaqMan Low Density Array method was used to screen out the candidate miRNAs for early screening of pneumoconiosis, and RT-qPCR was used to verify. According to the area under the curve (AUC) , the sensitivity and specificity of the candidate indicators were investigated. The logistic regression model was established by the two-class logistic regression model. Results: The expression of 7 candidate miRNAs in the serum of pneumoconiosis patients was significantly different (P<0.05) . The receiver operating curve (ROC) of the above 7 miRNAs was analyzed, miRNA-21, miRNA-200c, miRNA-16, miRNA-206, miRNA-155, miRNA-29a had statistical significance, and their ROC-AUC is 0.629~0.932. Logistic regression model was: logitP=13.769+0.536×miRNA-21-0.878×miRNA-200C-0.012×miRNA-16-0.111×miRNA-206+0.117×miRNA-155-1.192×miRNA-29a. Conclusion: Multiple serum miRNAs combined application models may be used for the diagnosis of pneumoconiosis patients.目的： 探究血清微RNAs（miRNAs）水平检测在尘肺筛查中的作用，建立用于尘肺诊断的多个血清miRNAs联合应用模型。 方法： 选取天津市工人疗养院就诊的152例男性尘肺患者作为病例组；正常对照组选取同社区健康体检人员152人，体检合格，无接触含游离二氧化硅的矿物粉尘，无心血管、肝肾等器质性病变以及肺部其他疾病。用TaqMan低密度芯片（TaqMan low density array）方法筛选出候选可作为尘肺早期筛查的miRNAs标志物，并采用RT-qPCR方法在152例尘肺较大样本中进行验证。根据曲线下面积（AUC）考察候选指标的灵敏度和特异性，采用二分类logistic回归模型前进法逐步回归建立尘肺和正常人群中诊断尘肺的logistic回归模型。 结果： 与对照组比较，miRNA-204、miRNA-206、miRNA-21、miRNA-16、miRNA-29a、miRNA-155、miRNA-200c 7个候选miRNAs在尘肺患者血清中的表达差异均有统计学意义（P<0.05）；分析上述7个miRNAs的受试者工作曲线（ROC）显示，miRNA-21、miRNA-200c、miRNA-16、miRNA-206、miRNA-155、miRNA-29a有统计学意义，其ROC-AUC为0.629~0.932。建立尘肺和正常人群中诊断尘肺的logistic回归模型：logitP=13.769+0.536×miRNA-21-0.878×miRNA-200c-0.012×miRNA-16-0.111×miRNA-206+0.117×miRNA-155-1.192×miRNA-29a。该回归模型在尘肺和正常人群中诊断尘肺ROC-AUC为0.980（95%CI：0.970~0.993，P<0.01）。 结论： 多个血清miRNAs联合应用模型可能用于尘肺患者的诊断。.