摘要
目的探讨并建立乳腺癌患者组织多项组化标记与血清CA153,CA125,CEA,25-OH-D3水平关联性分析模型,评估模型在预后判断中的临床价值。方法选取2016年1月~4月在中国石油天然气集团公司中心医院就诊的由组织病理学确诊的新发乳腺癌患者60例作为乳腺癌组,选取乳腺良性疾病患者60例作为乳腺良性疾病组,检测患者肿瘤标记物(CA153,CA125,CEA)和25-OH-D3;对所有检测数据建立Apriori模型,对各检验指标与不同组化组合阳性乳腺癌间的关联性分析,并进行特征规则集分析,搜集30例首次就诊患者作为验证组,获取模型性能评估信息。结果①乳腺癌组的CEA和CA125水平均高于乳腺良性疾病组,而25-OH-D3水平则低于乳腺良性疾病组,差异均具有统计学意义(t=3.347,2.790,6.727,均P<0.05);CA153水平在乳腺癌组和乳腺良性疾病组间进行比较,差异无统计学意义(t=-0.054,P>0.05);②乳腺癌组多项组化指标间关联性分析结果,预后较好指标组合分别为ER,PR和KiF67;PR和ER;ER,EFcadherin和KiF67,实验室指标预测变量分别为是否绝经,CA125水平,25-OH-D3水平和CEA水平;预后较差指标组合为CFerbBF2和KiF67;Calponin和KiF67;PF63和KiF67,实验室指标预测变量分别为CA153水平,25-OH-D3水平,年龄和是否绝经;③不同组合之间灵敏度分别为85.00%,77.78%,94.74%,88.89%,100.00%和65.00%;特异度分别为54.54%,76.92%,58.33%,50.00%,50.00%和72.73%;阳性预测值分别为77.27%,82.35%,78.26%,92.31%,89.29%和81.25%;阴性预测值分别为66.67%,71.43%,87.50%,40.00%,100.00%和53.33%。结论乳腺癌确诊患者其病理组化结果与血清CA153,CA125,CEA,25-OH-D3水平关联度较高,规则集在预测组化结果方面灵敏度高、特异度好,在患者预后中意义重大,具有临床应用价值。
Objective To explore and establish a correlation analysis model between multi-grouped markers of breast cancer patients and serum CA153,CA125,CEA,25-OH-D3 levels,and evaluate the clinical value of the model in prognosis.Methods Sixty patients with newly diagnosed breast cancer diagnosed by histopathology at the China National Petroleum Corporation Central Hospital from January to April 2016 were selected,and the tumor markers(CA153,CA125,CEA) and 25-OH-D3 were also tested.Apriori model was established for all test data,and the correlation analysis between each test index and different grouped positive breast cancer was performed,and characteristic rule set analysis was performed.30 cases of first-time patients were collected as verification group to obtain model performance.Results ①The CEA and CA125 levels in the breast cancer group were higher than those in the benign breast disease group,while the 25-OH-D3 level was lower than that in the benign breast disease group(t=3.347,2.790,6.727,all P<0.05).The CA153 level was in breast cancer patients.Compared with the benign breast disease group,the difference was not statistically significant(t=-0.054,P>0.05).②The correlation analysis between the multi-grouped indicators of breast cancer patients was ranked according to the confidence level from high to low,and the prognosis was better.The index combinations were ER,PR and KiF67,PR and ER,ER,EFcadherin and KiF67,with confidence levels of 100%,97.509% and 97.509%,respectively.The combination of poor prognosis indicators were CFerbBF2 and KiF67,Calponin and KiF67,PF63 and KiF67,with confidence levels of 96.000%,95.833% and 95.455%,respectively.③The sensitivity between different combinations were 85.00%,77.78%,94.74%,88.89%,100.00% and 65.00% respectively.The specificity were 54.54%,76.92%,58.33%,50.00%,50.00% and 72.73% respectively.The positive predictive values were 77.27%,82.35%,78.26%,92.31%,89.29% and 81.25% respectively.The negative predictive values were 66.67%,71.43%,87.50%,40.00%,100.00% and 53.33%,respe
作者
尹志辉
周志伟
刘继勇
杨保昌
王凌凌
YIN Zhi-hui;ZHOU Zhi-wei;LIU Ji-yong;YANG Bao-chang;WANG Ling-ling(Department of Clinical Laboratory,Xingtai Third Hospital,Hebei Xingtai 054000,China;Department of Clinical Laboratory,the Central Hospital of China National Petroleum Corporation,Hebei Langfang 065000,China;Department of Oncology,the Central Hospital of China National Petroleum Corporation,Hebei Langfang 065000,China)
出处
《现代检验医学杂志》
CAS
2019年第3期47-50,54,共5页
Journal of Modern Laboratory Medicine
基金
廊坊市科技支撑计划项目,编号:2016013025
