摘要
目的探讨核仁组成区相关嗜银蛋白(AgNOR)、癌胚抗原(CEA)、铁蛋白(Ft)定量分析在良、恶性浆膜腔积液中的鉴别诊断价值。方法采用HPIAS-1000高清晰度图像分析系统定量测定57例良、恶性浆膜腔积液中脱落细胞的AgNOR颗粒12项参数,并对AgNOR形态进行观察分型;同时采用化学发光法测定CEA和Ft水平。结果AgNOR定量分析12项参数有9项参数良、恶性组比较,差异有统计学意义(P<0.01),形态分型有助于诊断,特别是聚集型的出现更具诊断价值,发现了汉字样AgNOR异型颗粒。AgNOR、CEA、Ft的敏感性分别为100.0%、57.1%、52.4%,特异性分别为96.2%、88.5%、69.2%,阳性预测值分别为95.5%、80.0%、57.9%,阴性预测值分别为100.0%、71.9%、64.3%,尤登指数(J)分别为0.96、0.46、0.12。结论浆膜腔积液核仁组成区相关嗜银蛋白(AgNOR)、CEA的联合测定对于良、恶性积液有重要的鉴别诊断价值。
Objective To investigate the value of quantitative on analysis of argyrophil nucleolar organizer regions(AgNOR), carcinoembryonic antigen(CEA) and ferritin(Ft) for differentiation diagnosis of benign and malignant serous cavities effusion. Methods HPIAS-2000, a highly-clear image analysis system, was used to quantify 12 parameters of AgNOR particles of falling-off cells from 57 cases of benign and malignant serous cavities effusion. The shapes of AgNOR were observed and sorted. Meanwhile, CEA and Ft were examined with chemiluminescent assay. Results Among 12 parameters of AgNOR quantification, 9 parameters were significantly distinct between benign and malignant serous cavity effusion (P 〈 0.01). The morphological study of AgNOR particles was useful for diagnosis, especially the observation of aggregated particles. What was more interesting was that we discovered a kind of character-like atypical AgNOR particles. The sensitivities of AgNOR, CEA and Ft for diagnosis were respectively 100%, 57.1% and 52.4%. The specificities of them were 96.2%, 88.5% and 62.9%. The positive predictive values of them were 95.5%, 80% and 57.9% and the negative predictive values were 100%, 71.9% and 64.3%. The Youden's indexes were 0.96, 0.46 and 0.12. Conclusion The collaborated assay of AgNOR and CEA has great value for the differentiation of benign and malignant serous cavities effusion.
出处
《国际检验医学杂志》
CAS
2006年第6期502-504,共3页
International Journal of Laboratory Medicine
关键词
核蛋白类
核仁组成区
癌胚抗原
铁蛋白
胸腔积液
诊断
Nucleoproleins
Nucleolus organizer region
Carcinoembryonic antigen
Ferritin
Pleural effusion
Diagnosis