【作者】 吴迪；

【导师】 李红美；

【作者基本信息】 苏州大学 ， 公共卫生（专业学位）， 2023， 硕士

【摘要】 研究背景妊娠期高血压疾病(hypertensive disorders of pregnancy,HDP)是女性妊娠期一种常见的疾病。HDP不仅严重威胁孕产妇的健康,而且对胎儿、新生儿甚至以后的生长发育也有严重影响,是孕产妇和胎儿发病和死亡的主要原因之一。然而,并非所有HDP患者都有较高的传统危险因素。临床上常见的问题是未能早期发现HDP,一旦患者被确诊,一般都符合临床诊断标准或有严重的靶器官并发症,需要转到三级医院治疗。因此,及时预测和准确诊断是预防HDP导致严重后果的关键。血管生成标志物,如胎盘生长因子(placental growth factor,PlGF)和血管生成相关生物标志物,如可溶性 fms 样酪氨酸激酶-1(Soluble fms-like tyrosine kinase-1,sFlt-1)与 PlGF 的比值,即sFlt-1/PlGF,已被发现可以预测先兆子痫。但是,血管生成标志物对中国女性HDP及其亚型的预测价值仍需进一步探索。研究目的1.探讨PlGF及sFlt-1/PlGF与HDP及其亚型发病的关联关系,并分析它们与早发型和晚发型HDP及其亚型发病风险之间的关系。2.评价PlGF及sFlt-1/PlGF对HDP及其亚型、以及早发型和晚发型HDP及其亚型的预测价值。研究对象与方法自2019年8月到2021年6月,在江苏省昆山市的10家医院和15家社区卫生服务中心建立前瞻性队列。收集该队列人群的建卡信息、每次产检信息及分娩信息,采集孕中期(16-20周)的血液样本,离心后分离冻存。最后,共有11731例孕妇完成了信息采集及血液标本采集,但仅有9447名孕产妇符合本研究的纳入与排除标准。队列中确诊为HDP的孕妇共有415名,按照既定的纳入和排除标准,共有403例患者被纳入病例组。按照年龄±1岁、产次相同及1:1的原则从未发生HDP的孕妇中匹配对照,构建巢式病例对照。根据发病孕周将HDP及其亚型分为早发型(发病孕周<34)和晚发型(发病孕周≥34)。采用酶联免疫吸附法检测巢式病例对照人群孕16-20周血液中PlGF和sFlt-1浓度,计算sFlt-1与PlGF的比值。实验室检测后,16对研究对象因PlGF检测失败而被排除。最终387对(n=774)信息完整者被纳入本次分析。描述和比较病例组和对照组的一般特征时,因本研究中数值变量资料均不满足正态分布,故采用中位数和四分位数间距进行描述,并使用Wilcoxon符号秩检验比较组间差异。分类变量采用率或构成比描述,使用McNemarχ2检验或其精确检验比较组间差异。采用条件logistic回归分析PlGF、sFlt-1/PlGF与HDP及其亚型之间的关系。进行多因素分析时,调整变量包括BMI,首次产检收缩压,首次产检空腹血糖,教育程度、血小板计数和白细胞计数。计算优势比(odds ratio,OR)和95%可信区间(95%confidenceinterval,95%CI)来描述 PlGF 和 sFlt-1/PlGF 与 HDP 发生风险的关系。采用受试者工作特征(receiver operating characteristic curve,ROC)和曲线下面积(areaunder curve,AUC)评价预测模型的有效性。计算综合判别改进指数(integrated discrimination improvement,IDI)和净重分类指数(net reclassification index,NRI)用于描述传统因素(traditional factors,TFs)+P1GF 和 TFs+sFlt-1/PlGF 模型相对于TFs模型的改善程度。使用R软件(4.1.2版)进行统计学分析,本研究所有统计学检验均为双侧,检验水准均设为α=0.05。结果1.HDP及其对照组人群的基本特征387 例 HDP 患者中,妊娠高血压(gestational hypertension,GH)291 例、子痫前期(preeclampsia,PE)82 例、慢性高血压合并先兆子痫(chronic hypertension with superimposedpreeclampsia,CH-PE)14 例。HDP 组PlGF 水平低于对照组(74.46pg/ml vs.92.93 pg/ml,P<0.0001),而 sFlt-1/PlGF 比值水平则高于对照组(35.44 vs.31.07,P=0.0058),并且病例组文化程度较低者所占的比例、有高血压家族史者、首次体检时的血压值、空腹血糖、血小板、BMI≥24kg/m2者的比例,以及孕中期血糖水平均高于对照组。2.血管生成标志物与HDP及其亚型的关系低水平的PlGF(<100 pg/ml)及高水平的sFlt-1/PlGF(≥85)比值均与HDP发病风险增加有关,OR 值分别为 1.85(95%CI,1.36-2.53)、2.35(95%CI,1.52-3.63)。且经过多因素调整后,OR值仍具有统计学意义,分别为2.05(95%CI:1.28-3.26)、3.10(95%CI:1.66-5.79)。低水平PlGF(<100 pg/ml)和高水平sFlt-1/PlGF(≥85)比值与GH发生风险增加相关,单因素及多因素分析均有统计学意义,单因素分析中OR值分别为1.66(95%CI:1.17-2.36)、1.74(95%CI:1.06-2.84),多因素分析中 OR 值分别为 2.01(95%CI:1.14-3.53)、2.15(95%CI:1.02-4.52);而对于 PE 来说,PlGF<100pg/ml 是PE发生的危险因素,OR=2.78(95%CI:1.30-5.95)。但经多因素校正后,校正OR值无统计学意义。sFlt-1/PlGF比值≥85与PE发生风险增加相关,单因素和多因素分析中 OR 值均有意义,分别为 6.12(95%CI:2.06-18.17)、5.86(95%CI:1.66-20.65)。3.血管生成标志物对HDP及其亚型的预测价值预测HDP的传统因素(TFs)模型中分别加入PlGF和sFlt-1/PlGF后对于模型的预测能力均有所改善,NRI分别为0.806(P<0.0001)、0.651(P<0.0001)。在预测GH发生的传统模型中加入PlGF提升了对GH发生的预测能力(IDI=0.017,P=0.0026);在PE方面,传统模型中加入PlGF及sFlt-1/PlGF对PE的预测能力均有所改善,而sFlt-1/PlGF+TFs 模型具有更高的预测能力,NRI 为 0.415(95%CI:0.241-0.589)。4.血管生成标志物在早发型、晚发型HDP及亚型与对照组人群中的水平比较早发型 HDP(early-onset HDP,EO-HDP)组(75.66 pg/ml vs.94.10 pg/ml,P<0.0001)及早发型 GH(early-onset GH,EO-GH)组(77.99pg/ml vs.92.56pg/ml,P=0.0185)孕妇的PlGF水平均低于其匹配的对照组,且具有统计学意义;早发型PE(early-onset PE,EO-PE)组孕妇与对照组相比,有较低的PlGF水平(55.89 pg/ml vs.102.04 pg/ml,P=0.0001)和较高的 sFlt-1/PlGF 比值(43.50 vs.23.56,P=0.0460)。晚发型HDP(late-onset HDP,LO-HDP)组孕妇的PlGF水平低于对照组(70.42 pg/ml vs.80.90pg/ml,P=0.0009),sFlt-1/PlGF 值高于正常组(37.73 vs.32.71,P=0.0032),且均有统计学意义;晚发型GH(late-onset GH,LO-GH)组孕妇的PlGF水平低于对照组(75.65 pg/ml vs.94.65 pg/ml,P=0.0195),尽管 sFlt-1/PlGF 值高于正常组(34.77 vs.31.23,P=0.189),但没有显著意义;晚发型 PE(late-onset PE,LO-PE)组孕妇的PlGF水平低于对照组(46.14 pg/ml vs.78.67 pg/ml,P=0.0039),且具有统计学意义。5.血管生成标志物与早发型和晚发型HDP及亚型的关系低水平PlGF(<100pg/ml)和高水平sFlt-1/PlGF(≥85)比值与EO-HDP风险增加相关,OR值分别为 1.85(95%CI:1.26-2.72)、3.11(95%CI:1.53-6.33),多变量校正后,他们之间的关联性仍有统计学意义,OR值分别为1.89(95%CI:1.10-3.27)和2.99(95%CI:1.16-7.71)。然而,仅有PlGF与EO-GH发生风险增加有关联性,OR=1.54(95%CI:1.01-2.36),并且多因素后仍有统计学意义,OR为3.29(95%CI:1.37-7.95)。低PlGF与高sFlt-1/PlGF比值均与EO-PE发生风险增加有关联,OR值分别为 5.33(95%CI:1.55-18.30)、6.19(95%CI:1.32-29.10),经多因素调整后,仅有PlGF与EO-PE的关联性具有统计学意义,OR=5.52(95%CI:1.03-29.43)。PlGF 水平低(<100pg/ml)可增加 LO-HDP 的风险,OR=1.86(95%CI:1.09-3.16),但经多变量调整后,OR值无统计学意义;高的sFlt-1/PlGF(≥85)比值与LO-HDP的风险增加相关(OR=3.26,95%CI:1.56-6.82),并且在多变量校正后仍具有统计学意义,OR 为 3.20(95%CI:1.32-7.79)。低的 PlGF(<100pg/ml)和高的 sFlt-1/PlGF(≥85)比值均与LO-GH的风险增加相关,其OR值分别为1.93(95%CI:1.04-3.61)、2.48(95%CI:1.05-5.88),而多因素调整后无统计学意义。仅有高水平的sFlt-1/PlGF比值与LO-PE发病风险增加有关(OR=5.72,95%CI:1.24-26.45),且多因素调整后仍具有统计学意义,OR值为5.83(95%CI:1.22-27.92)。6.血管生成标志物对早发型和晚发型HDP及其亚型的预测价值TFs模型中加入PlGF和sFlt-1/PlGF对EO-HDP的预测价值均有明显改善,NRI分别为 1.008(95%CI:0.856-1.161)、0.878(95%CI:0.719-1.037);TFs 模型中加入 PlGF对EO-GH和EO-PE的预测价值均有所提高,NRI分别为1.242(1.085-1.400)、0.209(0.005-0.319),IDI 分别为 0.036(0.016-0.056)、0.107(0.039-0.175)。在传统模型中加入PlGF可以提高对LO-HDP的预测能力,NRI和IDI分别为0.454(95%CI:0.227-0.681)、0.031(95%CI:0.009-0.052);在 TFs 模型中加入 sFlt-1/PlGF可以改善对LO-PE的预测能力,与TFs模型相比,NRI(0.385,P=0.0078)和IDI(0.103,P=0.0036)均有统计学意义。结论1.孕16-20周血浆低水平的PlGF(<100pg/ml)和高水平的sFlt-1/PlGF比值(≥85)增加HDP及其亚型GH和PE的发病风险。2.传统模型中加入PlGF、sFlt-1/PlGF对于HDP及其亚型PE的预测力均有所提高,而对于预测GH的发生,只有在模型中加入PlGF其预测力才有所提高,并且PlGF联合传统因素对早发型HDP及其亚型的预测能力更好。3.建议监测孕产妇血液中PlGF及sFlt-1/PlGF水平,以便早期发现异常,采取应对措施,帮助临床及妇幼保健工作者更好地预防和管理HDP,防止其不良结局的发生。更多还原

【Abstract】 BackgroundHypertensive disorders of pregnancy(HDP)are a common disorder in women during pregnancy,which not only endangers maternal health,but also endangers fetal,neonatal health and even later growth and development.Furtherly,HDP is one of the main causes of maternal and fetal morbidity and mortality.However,not all patients with HDP have high levels of traditional risk factors.A common clinical problem is the failure to detect HDP early.Once patients with HDP are diagnosed,they generally meet clinical diagnostic criteria or have severe target organ complications that require referral to tertiary care.Therefore,timely prediction and accurate diagnosis are keys to prevent morbidity and mortality due to HDP.Angiogenic markers,such as placental growth factor(PlGF)and soluble fms-like tyrosine kinase-1(sFlt-1)/PlGF,have been shown to predict pre-eclampsia in pregnant women.However,the predictive value of angiogenic markers for HDP and its subtypes in Chinese women still needs to be further explored.Purposes1.To evaluate the association of PlGF and sFlt-1/PlGF with the development of HDP and its subtypes,and to analyze their relationship with early-onset and late-onset HDP and its subtypes.2.To investigate the predictive value of PlGF and sFlt-1/PIGF on HDP and its subtypes,as well as the early-onset and late-onset HDP and its subtypes.Participants and MethodsA prospective cohort was established in 10 hospitals and 15 community health service centers in Kunshan,Jiangsu Province,from August 2019 to June 2021.All participants’information at the first antenatal care appointment,each prenatal care visit and hospital delivery was collected,and blood samples were collected in the early second trimester(16-20 gestationgal weeks,gw),centrifuged and then separated for freezing.A total of 11,731 pregnant women completed information and blood specimen collection,but only 9,447 met the inclusion and exclusion criteria of this study.415 pregnant women were diagnosed with HDP in the cohort.According to established inclusion and exclusion criteria,403 patients were included in the case group.Matched controls were selected from pregnant women who had never experienced HDP.Based on the principle of age ± 1 year old,same parity,and 1:lratio,a nested case-control study was constructed.According to the gestational week of onset,HDP and its subtypes were classified into early onset(gestational week of onset<34gw)and late onset(gestational week of onset>34gw)types.The concentrations of PlGF and sFlt-1 in the plasma at 16-20gw of the nested case-control participants were measured by enzyme-linked immunosorbent assay,and the ratio of sFlt-1 to PlGF was calculated.16 pairs were excluded due to failure of PlGF detection after laboratory testing.A total of 387(n=774)study pairs ended up with complete information.The general characteristics between the case and control groups were described and compared.Because all numerical variables didn’t follow a normal distribution,median and interquartile range were calculated to describe the description,and Wilcoxon signed rank test was used to compare the differences between groups.The distribution of categorical data was described by rates or proportions,and differences were compared using McNemar’s χ2 test or exact test.Conditional logistic regression was used to analyze the relationship between PlGF,sFlt-1/PlGF and HDP and its subtypes.In the multivariate analysis,the adjusted variables included education level,body mass index,systolic blood pressure,fasting blood glucose blood platelet and white blood cell at the first antenatal care appointment.The odds ratio(OR)and 95%confidence interval(95%CI)were calculated to describe the relationship between PlGF and sFlt-1/PlGF and the risk of HDP.The receiver operating characteristic curve(ROC)and area under the curve(AUC)were used to evaluate the validity of the prediction models.Integrated discriminant improvement(IDI)and net reclassification improvement(NRI)were used to describe the overall improvement of the traditional factors(TFs)+PlGF and TFs+sFlt-1/PlGF models relative to the TFs model.Statistical analysis was performed using R software(version 4.1.2),and all statistical tests in this study were two-sided,and the significant level was set at a=0.05.Results1.Basic characteristics of HDP and its control groupsAmong the 387 HDP patients,291 had gestational hypertension(GH),82 had preeclampsia(PE),and 14 had chronic hypertension with superimposed preeclampsia(CH-PE).Median level of PlGF was lower in HDP group than that in the control group(74.46 pg/ml vs.92.93 pg/ml,P<0.0001),whereas sFlt-1/PIGF ratio was higher than that in the control group(35.44 vs.31.07,P=0.0058).The proportions of pregnant women with lower education,and with a family history of hypertension were higher in HDP than that in the control group.Moreover,average levels of blood pressures,fasting glucose,platelets,and proportion of women with BMI>24kg/m2 at the first antenatal care,and mid-pregnancy glucose level were higher in the case group than that in the control group.2.Relationship between angiogenic markers and HDP and its subtypesBoth lower levels of PlGF(<100 pg/ml)and higher levels of sFlt-1/PlGF(≥85)were associated with a higher risk of HDP occurrence,with ORs of 1.85(95%CI:1.36-2.53)and 2.35(95%CI:1.52-3.63),respectively.And the association remained statistically significant after adjustment for covariates,with ORs of 2.05(95%CI:1.28-3.26),3.10(95%CI:1.66-5.79),respectively.Lower PlGF(<100 pg/ml)and higher sFlt-1/PlGF(≥85)were associated with increased risk of GH occurrence.The OR values in the univariate analysis were 1.66(95%CI:1.17-2.36)and 1.74(95%CI:1.06-2.84),respectively,and which was 2.01(95%CI:1.14-3.53)and 2.15(95%CI:1.02-4.52),respectively in the multivariate analysis.Lower PlGF(<100pg/ml)was found to increase the risk of PE,with OR of 2.78(95%CI:1.30-5.95).However,after multivariate adjustment,the adjusted OR value was not statistically significant.Furthermore,higher sFlt-1/PlGF(≥85)ratio was associated with an increased risk of PE,with OR of 6.12(95%CI:2.06-18.17),and the association remained statistically significant after multivariate adjustment,with adjusted OR of 5.86(95%CI:1.66-20.65).3.Predictive value of angiogenic markers for HDP and its subtypesPlGF and sFlt-1/PlGF were added to the traditional factors(TFs)model for predicting HDP,which improved the predictive ability of the TFs model with NRIs of 0.806(P<0.0001)and 0.651(P<0.0001),respectively.The addition of PlGF to the TFs model for predicting GH occurrence improved the predictive ability(IDI=0.017,P=0.0026).For PE,the addition of PlGF and sFlt-1/PlGF to the TFs model improved the predictive ability,while the sFlt-1/PlGF+TFs model had a higher predictive ability,with NRI of 0.415(95%Cl:0.241-0.589).4.Angiogenic markers levels of early-onset and late-onset HDP and its subtypes and control groupsPlGF levels were lower in pregnant women in the early-onset HDP(EO-HDP)group(75.66 pg/ml vs.94.10 pg/ml,P<0.0001)and early-onset GH(EO-GH)group(77.99 pg/ml vs.92.56 pg/ml,P=0.0185)compared to their controls.Compared to the control group,the early-onset PE(EO-PE)group of pregnant women had a lower PlGF level(55.89 pg/ml vs.102.04 pg/ml,P=0.0001)and a higher sFlt-1/PlGF ratio(43.50 vs.23.56,P=0.0460).Pregnant women in the late-onset HDP(LO-HDP)group had a lower PlGF level(70.42 pg/ml vs.80.90 pg/ml,P=0.0009)and a higher sFlt-1/PlGF value(37.73 vs.32.71,P=0.0032)than the control group.Pregnant women in the late-onset GH(LO-GH)group had a lower PlGF level than the control group(75.65 pg/ml vs.94.65 pg/ml,P=0.0195)and a higher sFlt-1/PlGF value than the control group(34.77 vs.31.23,P=0.1189),but the difference was not significant.The PlGF level in the late-onset PE(LO-PE)group was significantly lower than that in the control group(46.14 pg/ml vs.78.67 pg/ml,P=0.0039).5.Association of angiogenic markers with early-onset and late-onset HDP and subtypesLower PlGF(<100pg/ml)and higher sFlt-1/PlGF(≥85)ratio were associated with an increased risk of EO-HDP,with ORs of 1.85(95%CI,1.26-2.72)and 3.11(95%CI,1.53-6.33),respectively.The associations remained significant after multivariable adjustment,with ORs of 1.89(95%CI:1.10-3.27)and 2.99(95%CI:1.16-7.71),respectively.PlGF was associated with an increased risk of EO-GH occurrence,OR=1.54(95%CI:1.01-2.36),which remained statistically significant after adjustment(adjusted OR=3.29,95%CI:1.37-7.95).In the univariate analysis,both lower PlGF and higher sFlt-1/PIGF ratio were associated with increased risk of EO-PE,with ORs of 5.33(95%CI:1.55-18.30)and 6.19(95%CI:1.32-29.10),respectively.However,after adjustment for covariates,only PIGF was associated with the risk of PE statistically,with OR of 5.52(95%CI:1.03-29.43).Low PlGF(<100pg/ml)was associated with an increased risk of LO-HDP(OR=1.86,95%CI:1.09-3.16),but the adjusted OR value was not statistically significant after multivariate adjustment.Higher sFlt-1/PlGF(≥85)ratio was associated with an increased risk of LO-HDP(OR=3.26,95%CI:1.56-6.82),and the association remained statistically significant after multivariate adjustment,with OR of 3.20(95%CI:1.32-7.79),respectively.Lower PlGF(<100pg/ml)and higher sFlt-1/PlGF(≥85)ratio were both associated with an increased risk of LO-GH,with ORs of 1.93(95%CI:1.04-3.61)and 2.48(95%CI:1.05-5.88),respectively.But after multivariate adjustment,it is not statistically significant anymore.Only higher sFlt-1/PlGF ratio was associated with an increased risk of LO-PE(OR=5.72 95%CI:1.24-26.45),and the association was still statistically significant after multivariate adjustment,the OR was 5.83(95%CI:1.22-27.92).6.Predictive ability of angiogenic markers for early-onset and late-onset HDP and its subtypesThe addition of PlGF and sFlt-1/PlGF in the TFs model significantly improved the predictive value of EO-HDP,with NRI of 1.008(95%CI:0.856-1.161)and 0.878(95%CI:0.719-1.037),respectively.The TFs model with PlGF had a higher predictive value for EO-GH and EO-PE,with NRI of 1.242(1.085-1.400),0.209(0.005-0.319)and IDI of 0.036(0.016-0.056),0.107(0.039-0.175),respectively.The addition of PlGF to the TFs model improved the predictive ability for LO-HDP,with NRI and IDI of 0.454(95%CI:0.227-0.681)and 0.031(95%CI:0.009-0.052),respectively.The addition of sFlt-1/PlGF to the TFs model improved the predictive ability for LO-PE.Compared with TFs model,NRI(0.385,P=0.0078)and IDI(0.103,P=0.0036)were statistically significant.Conclusions1.Low level of PlGF(<100pg/ml)and high level of sFlt-1/PlGF(>85)ratio in plasma during 16-20 weeks of pregnancy increase the risks of HDP and its subtypes GH and PE.2.The addition of PlGF and sFlt-1/PlGF to the TFs model can significantly improve the predictive ability for HDP and its subtype PE,while for predicting GH,only the addition of PlGF to the TFs model can improve the predictive ability.And the combination of PlGF and TFs has better predictive ability for early-onset HDP and its subtypes.3.It is recommended to monitor the levels of PlGF and sFlt-1/PlGF in the blood,in order to find abnormalities and take response measures in the early stage,to help clinical and maternal and child health workers to better prevent and manage HDP,and to prevent its adverse outcomes更多还原