血清HE4和CA125联合预测卵巢肿瘤的风险

目的通过联合检测患者血清中的人附睾蛋白4(HE4)和癌抗原125(CA125),建立一种简单、有效的卵巢肿瘤患者风险预测模型。方法采用化学发光微粒子法(CMIA)测定257例女性血清的CA125、HE4水平,其中盆腔肿瘤患者按术后病理结果分为良性病变组111例和卵巢癌组53例;体检健康女性93例作为正常对照。对所有结果进行统计分析,并通过预测公式分别计算绝经前、后妇女患卵巢癌的风险。结果卵巢癌组CA125和HE4水平明显高于良性病变组和正常对照组(P<0.01);而相比正常对照组,良性病变组CA125水平稍有升高(48.2±127.0 vs 15.9±22.0IU/ml,P>0.05),HE4水平几乎没有变化(47.5±11.0 vs 42.6±8.0pmol/L,P>0.05)。绝经前组77例良性患者中67例被预测为低风险人群,预测特异性为87%;而35例卵巢癌病人有29例被预测为高风险患者,预测敏感性为82.9%;同理,绝经后组34例良性患者中27例被预测为低风险人群,预测特异性为79.4%,而18例卵巢癌病人有15例被预测为高风险患者,预测敏感性为83.3%。结论该模型综合利用血清CA125和HE4水平,成功将83%的卵巢癌患者归类为高风险人群,是临床上辅助卵巢肿瘤诊断、监测和术后风险预知的较好模型。

Prediction Risk of Ovarian Malignancy with Serum HE4 and CA125

Objective To establish a simple and effective model for prediction risk of patients with ovarian neoplasm with human epididymis protein 4(HE4) and cancer antigen 125(CA125).Methods Serum CA125 and HE4 levels in 257 cases were measured by chemiluminescent microparticle immunoassay(CMIA).Among them,111 cases were divided into benign lesions and 53 patients were classified as ovarian cancer according to pathological diagnosis,while 93 cases healthy women were chosen as normal controls.Data were processed by statistical analysis,and premenopausal/postmenopausal women were categorized into low or high risk candidates using corresponding prediction equations.Results CA125 and HE4 levels in ovarian cancer group were significantly higher than benign lesions or normal group(P<0.05).Compared to normal ones,there showed a slightly elevation for CA125 in benign lesions(48.2±127.0 vs 15.9±22.0IU/ml,P>0.05),while HE4 remained nearly unchanged(47.5±11.0 vs 42.6±8.0pmol/L,P>0.05).In premenopausal group,67 of 77 benign cases were predicted as the low-risk populations,with a predictive specificity of 87%;among 35 ovarian cancer cases,29 were forecasted to be high-risk patients,giving a sensitivity of 82.9%.Similarly,27 of 34 postmenopausal benign patients were predicted as the low-risk populations,with a forecast specificity of 79.4%,and 15 of 18 ovarian cancer cases were classified as high-risk patients,providing a sensitivity of 83.3%.Conclusion These equations with serum CA125 and HE4 levels successfully categorize 83% ovarian cancer patients into high-risk group,thus is believed to be a promising model for clinical ovarian cancer auxiliary diagnosis,postoperative monitoring as well as risk prediction.