糖链抗原125在预测卵巢癌复发中的价值

Predictive Value of Carbohydrate Antigen 125 for Recurrence of Ovarian Cancer

目的评价初诊糖链抗原125(carbohydrate antigen 125,CA125)对卵巢癌复发的预测价值。方法回顾性地从作者医院住院病例库中提取136例初诊卵巢癌患者的国际妇产科协会(Federation International Of GynecologyAnd Obstetrics,FIGO)分期、年龄、病理类型、CA125水平等资料,对患者进行电话随访,获取本研究的观察终点无瘤生存期。采用受试者工作特征曲线(receiver operating characteristic curve,ROC)法、Kaplan-Meier生存曲线法和Cox风险比例模型,分析初诊CA125水平与卵巢癌患者无瘤生存期的关系。结果随访期间复发的83例卵巢癌患者,血清CA125水平明显高于未复发的53例患者。CA125预测卵巢癌复发的曲线下面积(area under the curve,AUC)为0.76(0.68～0.84),最佳预测界值132kU/L,此时的预测敏感度0.71(95%CI:0.60～0.81),特异度0.66(95%CI:0.52～0.78)。CA125≥132kU/L和<132kU/L的卵巢癌的患者无瘤生存期分别为3.1年和4.9年(P<0.01)。在校正了年龄、FIGO分期之后,CA125≥132kU/L仍然与无瘤生存期有关(HR=3.20,P<0.01)。结论CA125可以作为初诊卵巢癌患者术后复发的独立预测因子。

Objective To evaluate the predictive value of carbohydrate antigen 125（CA125） for the recurrence of ovarian cancer.Methods The medical records,including age,Federation International Of Gynecology And Obstetrics（FIGO） stages and CA125 levels were retrospectively extracted from 136 patients with newly diagnosed ovarian cancer.All these patients were followed up by telephone interview to obtain the data of ovarian cancer free survival,the endpoint of this study.Receiver operating characteristic curve analysis,Kaplan-Meier approach and Cox model were used to analyze the relationship between CA125 level and the ovarian cancer free survival.Results Increased serum CA125 was observed in 83 patients with ovarian cancer recurrence during the follow up.The area under curve（AUC） of serum CA125 was 0.76（95% CI：0.68-0.84）,with the predictive sensitivity of 0.71（95% CI：0.60-0.81） and specificity of 0.66（95% CI：0.52-0.78） at the optimal threshold of 132 kU/L.The median ovarian cancer free survival for patients with increased（CA125≥132 kU/L） and decreased（CA125132 kU/L） CA125 were 3.1 and 4.9 years respectively（P0.01）.After correction for age and FIGO stages,CA125≥132 kU/L was independently associated with the recurrence of ovarian cancer（HR=3.20,P0.01）.Conclusion In patients with newly diagnosed ovarian cancer,CA125 is an independent predictor for recurrence.