摘要
目的探讨NLR-FIB作为联合预测指标预测结直肠癌预后的准确性。方法选取2009年7月至2012年7月接受治疗的结直肠癌患者330例,收集患者资料。根据随访期间不良预后发生情况,分为不良预后组和预后良好组。通过Kaplan-Meier法计算不良预后率并绘制生存曲线;单因素及多因素Cox对预后影响因素分析;利用ROC曲线分析评价各因素对患者预后的预测能力。结果根据随访分组:预后良好组123例,预后不良组166例。预后不良事件发生的均值时间为43. 95个月(95%CI:41. 83～46. 06),不良预后率为57. 44%。单因素分析显示:性别、吸烟、酗酒史、家族史、病程、肿瘤位置、肿瘤大小、病理类型、CEA对患者预后影响差异无统计学意义(P> 0. 05);年龄(P=0. 003)、Dukes分期(P=0. 005)、侵犯神经(P=0. 028)、NLR(P=0. 012)、FIB(P=0. 017)对疾病预后的影响差异有统计学意义(P <0. 05)。多因素COX回归分析显示:Dukes分期(Rr=5. 213,95%CI:3. 441～7. 899)、FIB(Rr=1. 672,95%CI:1. 427～1. 960)、NLR(Rr=1. 651,95%CI:1. 524～1. 789)、肿瘤侵犯神经(Rr=1. 565,95%CI:1. 204～2. 035)、年龄(Rr=1. 211,95%CI:1. 153～1. 272)。ROC曲线显示:利用FIB预测的AUC为0. 737(95%CI:0. 681～0. 794),最佳诊断点为3. 17 g/L,此时其敏感性和特异性分别为83. 40%、76. 40%;利用NLR预测的AUC为0. 730(95%CI:0. 673～0. 786),最佳诊断点为2. 43,此时其敏感性和特异性分别为79. 10%、74. 60%;利用年龄预测的AUC为0. 657(95%CI:0. 595～0. 720),最佳诊断点为74. 50岁,此时其敏感性和特异性分别为78. 90%、67. 8%。利用NLR、FIB联合指标NLR-FIB预测的AUC为0. 876(95%CI:0. 839～0. 914),最佳诊断点为5. 90,此时其敏感性和特异性分别为89. 60%、85. 50%。结论 NLR和FIB联合的指标NLR-FIB对结直肠癌患者预后有较好的预测能力,有望成为评估结直肠癌患者预后的常规指标。
Objective To investigate the accuracy of neutrophil/lymphocyte ratio(NLR)and detection of plasma fibrinogen(FIB)in predicting the prognosis of colorectal cancer.Methods A total of 330 patients with colorectal cancer who were admitted and treated in our hospital from July 2009 to July 2012 were enrolled in the study.The data of patients were collected.According to the occurrence of poor prognosis during the period of following up,the patients were divided into poor prognosis group and good prognosis group.The Kaplan Meier method was used to calculate the poor prognosis rate and to draw the survival curve.The factors influencing prognosis were analyzed by single factor and multivariate Cox.The predictive ability of different factor on patient’s prognosis was evaluated by ROC curve analysis.Results The patients were divided into good prognosis group(n=123),and poor prognosis group(n=166)according to the results of follow up.The mean time of occurance of poor prognosis event was 43.95 months(95%CI:41.83~46.06),and the ratio of bad prognosis was 57.44%.The univariate analysis showed that there were no significant differences in the effects of gender,smoking,history of intemperance,family history,course of disease,the location of tumour,the size of tumour,pathological pattern,carcino embryonie antigen(CEA)on prognosis of patients(P>0.05).However there were significant differences in the effects of age(P=0.003),Dukes stages(P=0.005),invasion nerves(P=0.005),NLR(P=0.012),FIB(P=0.017)in disease prognostic(P<0.05).Multivariate COX regression analysis showed that Dukes stage(RR=5.213,95%CI:3.441~7.899),FIB(RR=1.672,95%CI:1.427~1.960),NLR(RR=1.651,95%CI:1.524~1.789),tumor invasion of nerves(RR=1.565,95%CI:1.204~2.035),age(RR=1.211,95%CI:1.153 to 1.272).ROC curve showed that the AUC was 0.737(95%CI:0.681~0.794)and the optimal diagnostic point was 3.17g/L,and the sensitivity and specificity were 83.40%and 76.40%respectively when FIB was used as predicted indicator.However when NLR was used as predicted index,the AUC was
作者
张宇澄
盛宇伟
乔治
宋海瑚
沈彪
ZHANG Yucheng;SHENG Yuwei;QIAO Zhi(Department of General Surgery,Baoshan Branch of The First People’s Hospital,Shanghai 200940,China)
出处
《河北医药》
CAS
2018年第23期3535-3539,3544,共6页
Hebei Medical Journal
基金
上海市宝山区科研项目(编号:11-E-22)
