论文部分内容阅读
目的探讨血液分析仪高荧光强度有核细胞(HF)对体液中肿瘤细胞筛查价值。方法收集本院胸腹腔积液82例,分析记录其WBC、RBC、MN、PMN和HF细胞分类及GLU、TP、LDH和CHO生化指标水平,运用非条件二元Logistic回归模型结合受试者工作特征曲线(ROC)分析评价多变量指标在鉴别诊断积液性质中的价值。结果良、恶性胸腹腔积液组间WBC、RBC、单个核细胞数量、单个核细胞所占比例、多个核细胞数量、高荧光有核细胞数量、LDH、多个核细胞所占比例、TP、GLU水平差异均无统计学意义(P>0.05),HF%、CHO水平差异有统计学意义(P<0.05)。ROC曲线分析显示,HF%区分良、恶性积液中的肿瘤细胞具有良好的诊断效能。Logistic回归分析显示,仅CHO、HF%因素进入模型,方程为:y=0.153(HF%)+2.402(CHO)-4.236;对预测概率做ROC曲线分析,AUC为0.906(0.794,1.000),P<0.001,敏感度为0.964,特异性为0.727,cut-off值为0.49,该模型诊断效能较好,优于CHO、HF%单个的诊断效能。结论Sysmex XN-3000体液模式HF%在筛查积液肿瘤细胞,鉴别积液性质中初步显示良好价值。
Objective To investigate the value of hematology analyzer high fluorescence intensity nucleated cells (HF) in the detection of tumor cells in body fluids. Methods Totally 82 cases of pleural fluid and ascitic fluid in our hospital were collected. The classification of WBC, RBC, MN, PMN and HF cells and the biochemical indexes of GLU, TP, LDH and CHO were recorded. The Binary Logistic Regression Model ROC analysis to evaluate the value of multivariate markers in differential diagnosis of effusion. Results WBC, RBC, the number of mononuclear cells, the proportion of mononuclear cells, the number of multiple nucleated cells, the number of highly fluorescent nucleated cells, LDH, the proportion of multiple nucleated cells in benign and malignant pleural effusion group, TP (P> 0.05). The differences of HF% and CHO levels were statistically significant (P <0.05). ROC curve analysis showed that HF% differentiated benign and malignant effusion in the tumor cells have a good diagnostic efficacy. Logistic regression analysis showed that only the factors of CHO and HF entered the model with the equation: y = 0.153 (HF%) + 2.402 (CHO) -4.236. The ROC curve analysis of prediction probability showed that AUC was 0.906 (0.794,1.000), P <0.001, with a sensitivity of 0.964, a specificity of 0.727 and a cut-off value of 0.49. The diagnostic performance of this model is better than that of CHO and HF%. Conclusion Sysmex XN-3000 humoral model of HF% showed good value in screening effusion cells and identifying effusion.