血清肿瘤标志物对于原发性支气管肺癌病理类型的判别分析
2017-09-22王爱红王明全周小燕
王爱红++++++王明全++++++周小燕
[摘要] 目的 建立血清肿瘤标志物对肺部良性疾病、非小细胞肺癌和小细胞肺癌的贝叶斯判别模型。 方法 收集2010年3月~2015年3月于延安大学附属医院住院和门诊治疗的238例患者的病理诊断、实验室检查资料。采取电化学发光免疫法检测113例非小细胞肺癌、65例小细胞肺癌和60例肺部良性疾病患者的癌胚抗原、细胞角质素片断抗原、组织多肽特异性抗原、神经元特异性烯醇化酶和胃泌素释放肽前体。为了验证函数模型的判别价值,利用相应的指标数据回代预測分组。 结果 肺部良性疾病、非小细胞肺癌和小细胞肺癌患者之间五项血清肿瘤标志物的阳性率比较,差异有统计学意义(P < 0.05)。判别函数对于肺部良性疾病、非小细胞肺癌和小细胞肺癌正确率分别为93.33%、65.49%和64.62%。 结论 贝叶斯判别法对血清肿瘤标志物建立的判别函数鉴别诊断模型可简便、快速、有效地对肺部良性疾病、非小细胞肺癌和小细胞肺癌进行鉴别诊断,值得临床借鉴。
[关键词] 肺癌;肿瘤标志物;判别分析;鉴别诊断
[中图分类号] R734.2 [文献标识码] A [文章编号] 1673-7210(2017)08(b)-0067-04
[Abstract] Objective To establish Bayes discriminant model of serum tumor markers for pulmonary benign disease, non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC). Methods Data of 238 patients with pathological diagnosis, laboratory examination in Affiliated Hospital of Yan′an University inpatient and outpatient from March 2010 to March 2015 were collected. Five serum tumor markers including carcinoma-embryonic antigen, cytokeratin fragment antigen 21-1, tissue polypeptide specific antigen, neuron specific enolase and pro-gastrin-releasing peptide were detected in pulmonary benign disease (60 cases), NSCLS (113 cases) and SCLC (65 cases) by using electro-chemi-luminescence immunoassay. In order to verify the discriminant value of the function model, the corresponding index data was used to return to the forecasting group. Results Positive rate of five serum tumor markers among lung benign disease, NSCLS and SCLC were compared, the differences were statistically significant (P < 0.05). Accuracy rate of discriminant function for lung benign disease, NSCLS and SCLC was 93.33%, 65.49%, 64.62% respectively. Conclusion Bayesian discriminant method for the identification of serum tumor markers to identify the differential diagnosis model can be simple, rapid and effective for lung benign disease, NSCLS and SCLC differential diagnosis. It is worthy of clinical reference.
[Key words] Lung cancer; Tumor marker; Discriminatory analysis; Differential diagnosis
权威数据显示,在全世界范围内的肺癌死亡占癌症死亡总人数的29%(其中男31%,女26%),是重大的公共卫生问题[1]。癌胚抗原(carcinoma-embryonic antigen,CEA)、细胞角蛋白-19片断抗原(cytokeratin fragment antigen 21-1,Cyfra21-1)、组织多肽特异性抗原(tissue polypeptide-specific antigen,TPS)、神经原特异性烯醇化酶(neuron specific enolase,NSE)和胃泌素前体释放肽(pro-gastrin-releasing peptide,ProGRP)是重要的肿瘤标志物,常用于高危人群的筛查、诊断或预后[2-4]。本研究联合检测五种肿瘤标志物,建立判别分析函数模型,探讨其表达水平与肿瘤发生以及病理分型之间的关系。
1 资料与方法
1.1 一般资料
研究对象为2010年3月~2015年3月于延安大学附属医院住院和门诊治疗的病理诊断、实验室检查等资料完整的238例患者,其中男142例,女96例,年龄32~79岁,平均57.5岁。病理确诊的原发性支气管肺癌患者178例,其中非小细胞肺癌113例(Ⅱ期12例、Ⅲ期53例、Ⅳ期48例),小细胞肺癌65例(局限期45例、广泛期16例、未分期4例)。……
