摘要
目的探讨联合检测血清CA72-4、CA242、CA19-9和癌胚抗原(CEA)对胃癌的诊断价值。方法采用电化学发光法测定193例胃癌患者、106例胃良性疾病患者和86例健康对照者血清CA72-4、CA242、CA19-9和CEA水平,进行单因素方差分析、Logistic回归-ROC曲线和偏最小二乘-判别分析(PLS-DA)。结果Ⅰ~Ⅱ期胃癌组患者血清CA242和CEA水平显著高于良性疾病组和健康对照组(P<0.05)。胃癌组Ⅲ~Ⅳ期患者血清CA72-4、CA242、CA19-9和CEA水平显著高于Ⅰ~Ⅱ期、良性疾病组和健康对照组(P<0.01)。4种标志物联合的Logistic回归-ROC曲线下面积为0.956(95%置信区间:0.938~0.957),临界值(概率)为0.56,敏感度和特异性分别为86.0%和95.3%。PLS-DA模型对胃癌组和非胃癌组(胃部良性疾病+健康对照)的分类正确率分别为83.4%和81.8%。PLS-DA模型对Ⅰ~Ⅱ期和Ⅲ~Ⅳ期胃癌的分期诊断正确率分别为100.0%和82.9%。结论基于血清CA72-4、CA242、CA19-9的PLS-DA模型有助于胃癌的鉴别和分期诊断。
Ojeive o evlue he dignoi vlue of eum 72-4,242,19-9 nd ino-emyoni nigen (E) fo gi ne. Mehod eum level of 72-4, 242,19-9 nd E in 193 pien wih gi ne, 106 pien wih enign gi diee nd 86 helhy onol wee meued y eleohemilumineene y, nd ll d w nlyzed y he one-wy nlyi of vine,Logii egeion, eeive opeing heii (O) uve nd pil le que diini- nn nlyi (PL-D). eul he level of eum 242 nd E in he ge Ⅰ--Ⅱ gi ne goup wee ignifinly highe hn hoe in he enign gi diee nd helhy onol goup(P〈0.05). he level of eum 72-4,242 ,19-9 nd E in he ge Ⅲ -- Ⅳ gi ne goup wee ignifinly highe hn hoe in he ge Ⅰ-- Ⅱ gi ne goup, e- nign gi diee nd helhy onol goup(P〈0.01). he e unde he Logii egeion-O uve of he 4-mke om- inion w 0. 956 [95 % onfidene inevl (95 % I):0. 938 - 0. 957 ], he uoff vlue (poiliy) w 0.56, eniiviy nd peifiiy wee 86.0% nd 95.3% epeively. he uie of PL-D model fo he gi ne goup nd non-gi n- e goup (enign gi diee nd helhy onol) wee 83.4% nd 81.8% epeively. he uie of PL-D model fo ge Ⅰ-- Ⅱ nd 111 -- IV gi ne wee 100.0% nd 82.9% epeively. onluion he PL-D model ed on eum 72-4,242 ,19-9 nd E i helpful fo he diffeeniion nd he ge dignoi of gi ne.
出处
《检验医学与临床》
CAS
2017年第1期58-59,62,共3页
Laboratory Medicine and Clinic