血清sFlt-1、PLGF水平及比值变化对伴发子痫前期的风险评估及模型构建

Risk assessment and model construction of serum sFlt-1,PLGF levels and ratio changes associated with preeclampsi

目的 探究血清可溶性血管内皮生长因子受体1(sFlt-1)、胎盘生长因子(PLGF)水平及比值变化对伴发子痫前期(PE)的风险评估及模型构建。方法 选取2021年8月至2022年8月山东济宁医学院附属医院产科收治137例PE患者设为PE组,根据病情程度分为轻度PE(MPE)组和重度PE(SPE)组,另选取同期规律体检的137例健康孕妇设为对照组,比较3组患者血清sFlt-1、PLGF水平、比值及临床资料,多因素Logistic回归分析PE发生的危险因素,构建风险预测模型并验证区分度、拟合度及预测效能评估。结果 MPE组与SPE组血清sFlt-1、sFlt-1/PLGF比值均高于对照组,且SPE组高于MPE组(P<0.05);MPE组与SPE组血清PLGF低于对照组,且SPE组低于MPE组(P<0.05)。多因素Logistic回归分析显示,年龄>35岁、孕前体质指数(BMI)≥28 kg/m^(2)、有高血压家族史、妊娠期糖尿病或肾病、血小板计数减少、sFlt-1/PLGF升高是PE发生的独立危险因素(P<0.05),构建风险预测模型方程:Logit(P)=年龄>35岁×0.415+孕前BMI≥28 kg/m^(2)×1.320+有高血压家族史×1.552+妊娠期糖尿病或肾病×0.944+血小板计数减少×0.552+sFlt-1/PLGF升高×1.011~2.501,该风险预测方程的区分度好(C-index∶95%CI=0.862∶0.756~0.905),拟合度优(Hosmer-Lemeshow χ2=0.585,P=0.312)。应用PE预测模型方程回顾性验证,ROC曲线下面积(AUC)为0.876(95%CI:0.823~0.917),优于sFlt-1/PLGF比值的AUC(0.776)(95%CI:0.715~0.830,Z=2.105,P=0.034)。结论 年龄>35岁、BMI≥28 kg/m^(2)、有高血压家族史、妊娠期糖尿病或肾病、血小板计数减少、sFlt-1/PLGF升高是PE发生的独立危险因素,建立的风险预测模型对PE发生有较好的预测效能。

Objective To evaluate the risk assessment and model construction of serum soluble vascular endothelial growth factor receptor 1(sFlt-1)and placental growth factor(PLGF)levels and their ratios in patients with preeclampsia(PE).Methods From August 2021 to August 2022,137 patients with PE admitted to Affiliated Hospital of Jining Medical University were selected as PE group,and divided into mild PE(MPE)group and severe PE(SPE)group according to the degree of disease.Another 137 healthy pregnant women who had regular physical examination during the same period were selected as control group.The serum sFlt-1,PLGF levels and ratios,and the clinical data of the three groups of the patients were compared.Multivariate Logistic regression analysis of risk factors for PE.Build a risk prediction model,and verify the evaluation of discrimination,fit and prediction efficiency.Results The ratios of serum sFlt-1 and sFlt-1/PLGF in MPE group and SPE group were higher than those in control group,and SPE group was higher than MPE group(P<0.05).Serum PLGF in MPE group and SPE group was lower than that in control group,and SPE group was lower than MPE group(P<0.05).Multivariate Logistic regression analysis showed that age>35 years old,pre-pregnancy body mass index(BMI)≥28 kg/m^(2),family history of hypertension,gestational diabetes or nephropathy,decreased platelet countand elevated sFlt-1/PLGF were independent risk factors for PE(P<0.05).Construct the risk prediction model equation:Logit(P)=age>35 years×0.415+pre-pregnancy BMI≥28 kg/m^(2)×1.320+family history of hypertension×1.552+gestational diabetes or kidney disease×0.944+decreased platelet count×0.552+sFlt-1/PLGF increased×1.011~2.501.The risk prediction equation had good discrimination(Cindex:95%CI=0.862∶0.756~0.905)and good fit(Hosmer-Lemeshowχ2=0.585,P=0.312).The PE prediction model equation was used for retrospective verification,and the area under the ROC curve(AUC)was 0.876(95%CI:0.823~0.917),which was better than the AUC of the sFlt-1/PLGF ratio was 0.776(95%CI:0.715~0.830,Z=2.105,P=0.034).Conclusion Age>35 years old,BMI≥28 kg/m^(2),family history of hypertension,gestational diabetes mellitus or nephropathy,decreased platelet count,and elevated sFlt-1/PLGF are independent risk factors for PE,and the establishedrisk prediction model has good predictive performance for PEoccurrence.