检验医学 ›› 2017, Vol. 32 ›› Issue (5): 353-360.DOI: 10.3969/j.issn.1673-8640.2017.05.002

• 临床应用研究·论著 • 上一篇    下一篇

探索15项生化指标联合检测对肝脏恶性肿瘤首次治疗后的应用价值

刘风华, 王李洁, 周运恒, 李戬, 胡琼   

  1. 武警上海总队医院检验科,上海 201103
  • 收稿日期:2016-06-20 出版日期:2017-05-20 发布日期:2017-06-06
  • 作者简介:null

    作者简介:刘凤华,女,1974年生,学士,副主任技师,主要从事临床生化检验工作。

    通信作者:李 戬,联系电话:021-62429837。

Combined determination of 15 biochemical indicators for the first treatment of malignant liver tumors

LIU Fenghua, WANG Lijie, ZHOU Yunheng, LI Jian, HU Qiong   

  1. Department of Clinical Labortory,the Armed Police Hospital of Shanghai,Shanghai 201103,China
  • Received:2016-06-20 Online:2017-05-20 Published:2017-06-06

摘要:

目的 探索15项生化指标联合检测在肝脏恶性肿瘤首次治疗后的应用价值。方法 测定54例原发性肝癌(PHC)患者、54例继发性肝癌(MHC)患者及120名正常对照(NC)者血清甘胆酸(CG)、前白蛋白(PA)、总胆汁酸(TBA)、天门冬氨酸氨基转移酶线粒体同工酶(m-AST)、丙氨酸氨基转移酶(ALT)、天门冬氨酸氨基转移酶(AST)、γ-谷氨酰基转移酶(GGT)、碱性磷酸酶(ALP)、乳酸脱氢酶(LDH)、胆碱酯酶(CHE)、总胆红素(TB)、直接胆红素(DBil)、总蛋白(TP)、白蛋白(Alb)水平和白蛋白/球蛋白(A/G)比值。采用受试者工作特征(ROC)曲线分析各项指标的诊断价值,并采用探索性因子分析(EFA)和Logistic回归模型逐步探索从单项到多指标联合检测的应用价值。结果 PHC组、MHC组15项生化指标与NC 组比较,差异均有统计学意义(P<0.05、P<0.001)。PHC组TP、AST、m-AST、ALT、GGT、LDH、CG和TBA水平与MHC组比较差异有统计学意义(P<0.05、P<0.001),其他指标2个组之间差异均无统计学意义(P>0.05)。ROC曲线分析显示PHC组除TP[ROC曲线下面积(AUC)=0.626]外,其他指标的AUC均≥0.7;MHC组除ALT(AUC=0.600)和TB(AUC=0.566)外,其他指标的AUC均≥0.7。采用EFA,PHC组和MHC组分别获得6个独立的公共因子,对每个因子内的指标运用综合评价法筛选出P9模型(由PHC组PA、CHE、Alb、CG、DBil、GGT、ALT、AST、LDH联合)和M9模型(由MHC组PA、CHE、Alb、TBA、DBil、GGT、AST、LDH、TP联合)。PHC组P9模型及MHC组M9模型的诊断价值与15项生化指标联合检测的诊断价值相当(Z=0.590,P=0.555 4;Z=0.515,P=0.606 8)。P9和M9模型的AUC均>0.95、敏感性和特异性均>95%、阳性似然比均>10、阴性似然比均>0.1,最佳临界值分别为1.66、0.58,二者的诊断价值均明显优于单项指标。结论 P9和M9模型既有最佳的诊断效能又有最优的经济效益,完全可以替代15项生化指标联合检测,用于肝脏恶性肿瘤治疗后期的疗效评估和病情监控的随访指标。

关键词: 生化指标, 肝脏恶性肿瘤, 探索性因子分析, Logistic回归, 受试者工作特征曲线

Abstract:

Objective To investigate the combined determination of 15 biochemical indicators for the first treatment of malignant liver tumors. Methodse A total of 54 patients with primary hepatic carcinoma(PHC),54 patients with metastatic hepatic carcinoma(MHC) and 120 healthy subjects [normal control(NC)group] were enrolled. Serum cholyglycine(CG),prealbumin(PA),total bile acid(TBA),mitochondrial aspartate aminotransferase(m-AST),alanine aminotransferase(ALT),aspartate aminotransferase(AST),gamma-glutamyltransferase(GGT),alkaline phosphatase(ALP),lactate dehydrogenase(LDH),cholinesterase(CHE),total bilirubin(TB),direct bilirubin(DBil),total protein(TP),albumin(Alb)and Alb/globulin(A/G) ratio were determined. Receiver operating characteristic(ROC)curve was used to evaluate diagnosis performance. Exploratory factor analysis(EFA) and Logistic regression model were used to investigate the significance from single determinations to combined determination. Results The 15 biochemical indicators were statistically significant between PHC and NC groups and MHC and NC groups (P<0.05,P<0.001). There was statistical significance for TP,AST,m-AST,ALT,GGT,LDH,CG and TBA between PHC and MHC groups(P<0.05,P<0.001),and there was no statistical significance for the other indicators (P>0.05). Except for TP [area under ROC curve(AUC)=0.626],the other indicators' AUC in PHC group were ≥0.7. Except for ALT(AUC=0.600)and TB(AUC=0.566),the other indicators' AUC in MHC group were ≥0.7. EFA showed that there were 6 potential factors in PHC and MHC groups. Each potential factor was screened by comprehensive evaluation method. P9 model was made up of PA,CHE,Alb,CG,DBil,GGT,ALT,AST and LDH in PHC group. M9 model was made up of PA,CHE,Alb,TBA,DBil,GGT,AST,LDH and TP in MHC group. The diagnosis performance of P9 and M9 was similar to the combined determination of 15 biochemical indicators (Z=0.590,P=0.555 4;Z=0.515,P=0.606 8). The AUC of P9 and M9 were > 0.95,the sensitivities and specificities were > 95%,the positive likelihood ratios were >10,and the negative likelihood ratios were < 0.1. The optimal cut-off values of P9 and M9 were 1.66 and 0.58. Conclusions P9 and M9 models are effective for diagnosis performance and economical,which can replace the combined determination of 15 biochemical indicators,for the evaluation of treatment efficiency and follow-up monitoring of liver cancer.

Key words: Biochemical indicators, Malignant liver tumor, Exploratory factor analysis, Logistic regression, Receiver operating characteristic curve

中图分类号: