INTRODUCTION To evaluate a risk model of the maternal serum α-fetoprotein variants L2 and L3 (AFP-L2 and AFP-L3) for predicting fetal open neural tube defects (ONTD) and abdominal wall defects (AWD). METHODS Subjects were divided into the ONTD group (21), the AWD group (16) and the control group (38). Comparative analysis was performed using a risk model, receiver operating characteristic (ROC) curves and area under the curve (AUC) were used for screening performance of AFP-L2 and AFP-L3. RESULTS AFP-L2 and AFP-L3 concentrations in the ONTD group and the AWD group were both significantly higher than those in the control group (P < 0.001). The AUC for screening ONTD or AWD using AFP-L2 was 0.830 [95% confidence interval (CI):0.734-0.926, P < 0.001], while using AFP-L3 was 0.886 (95% CI: 0.809-0.963, P < 0.001). ROC curves indicated that the optimal threshold values for ONTD or AWD by AFP-L2 and AFP-L3 were 1.467 multiple of the median (MoM) and 1.944 MoM. The corresponding sensitivity, specificity and Youden index for AFP-L2 and AFP-L3 were 0.703, 0.947, 0.650 and 0.730, 0.974, 0.703, respectively. CONCLUSION AFP-L2 and AFP-L3 are favorable biomarkers for screening ONTD and AWD fetuses, and the risk model using AFP-L2 and AFP-L3 is of superior sensitivity and specificity.

中文翻译：

---

母体血清甲胎蛋白变异体（AFP-L2和AFP-L3）在预测胎儿开放性神经管缺损和腹壁缺损中的诊断准确性。

简介评价母体血清甲胎蛋白变异体L2和L3（AFP-L2和AFP-L3）的风险模型，用于预测胎儿开放性神经管缺损（ONTD）和腹壁缺损（AWD）。方法将受试者分为ONTD组（21），AWD组（16）和对照组（38）。使用风险模型进行比较分析，使用受试者工作特征（ROC）曲线和曲线下面积（AUC）筛选AFP-L2和AFP-L3的性能。结果ONTD组和AWD组的AFP-L2和AFP-L3浓度均显着高于对照组（P <0.001）。使用AFP-L2筛查ONTD或AWD的AUC为0.830 [95％置信区间（CI）：0.734-0.926，P <0.001]，而使用AFP-L3筛查的AUC为0.886（95％CI：0.809-0.963，P <0.001 ）。ROC曲线表明，AFP-L2和AFP-L3对ONTD或AWD的最佳阈值是中位数（MoM）的1.467倍和1.944 MoM。AFP-L2和AFP-L3的相应灵敏度，特异性和尤登指数分别为0.703、0.947、0.650和0.730、0.974、0.703。结论AFP-L2和AFP-L3是筛选ONTD和AWD胎儿的良好生物标志物，使用AFP-L2和AFP-L3的风险模型具有较高的敏感性和特异性。