A low maternal soluble fms-like tyrosine kinase-1 before 11 weeks is associated with a greater risk of early preeclampsia

OBJECTIVE: Soluble fms-like tyrosine kinase-1 (sFlt-1) is a vascular endothelial growth factor and placental growth factor (PlGF) antagonist that is produced in several tissues, including the placenta.1 Although maternal blood sFlt-1 is increased in preeclampsia (PE), it could be lowered early in pregnancy and raised at the end among women who developed preterm PE.2 We aimed to test this hypothesis, particularly before 11 weeks’ gestation, a time when biomarkers of fetal aneuploidy are frequently analyzed. STUDY DESIGN: We carried out a post hoc analysis of a casecontrol study that included pregnant women who had PE with delivery before 34 weeks’ gestation (early-PE) and compared them with women without a pregnancy complication, matched by date at blood sampling between 8þ0 and 13þ6 weeks’ gestation.3 Participants’ serum sFlt-1 and PlGF concentrations were measured and converted to multiples of the median (MoM) after adjustment for gestational age and body mass index (BMI). The median concentrations with interquartile ranges (IQRs) from cases and controls were compared using nonparametric tests and using the area under the curve (AUC) of receiver operating characteristics (ROC) curves. The analyses were stratified according to the gestational age at blood sampling (<11 and ≥11 weeks of pregnancy). RESULTS: Atotal of 37 patients with early PE were compared with 99 controls, including 21 cases and 43 controls with blood drawn before 11 weeks. Before 11 weeks, the maternal serum sFlt-1 concentrations were lower in cases (median, 496; IQR, 472e676 pg/mL) than in controls (1001; 743e1277 pg/mL; P1/4.03), but this was not the case after 11 weeks (745; IQR, 608e1425 pg/mL vs 1003; IQR, 747e1327 pg/mL; P1/4.33). A predictive model that combined maternal BMI with maternal serum sFlt-1 and PlGF concentrations before 11 weeks could detect (sensitivity) 52% of early PE at a false positive rate of 10% (1-specificity) and could detect 71% of early PE at a false positive rate of 20% (Figure). CONCLUSION: Low sFlt-1 concentrations before 11 weeks’ gestation are associated with early PE. Our observation agrees with that of Fillion et al2 who observed that sFlt-1 seemed to go from a low value in the first trimester to an increased value in the second and third trimesters in patients with preterm PE. Given that the same trend occurs with other maternal biochemical marker, this improves the support for sFlt-1 concentration before 11 weeks as a predictive biomarker of early PE. Our finding is important because it could promote access to PE screening as proposed by the Fetal Medicine Foundation. The screening is usually performed between 11 and 13 weeks’ gestation at the time of the first-trimester ultrasound, and if a patient was identified as being at high risk, aspirin was started to prevent PE.4 However, in the absence of an onsite laboratory and in many remote areas, blood test results can be delayed, thereby reducing the possibility of timely risk calculations. In many of these settings, for fetal aneuploidy screening, blood samples (pregnancy associated plasma protein-A and free Beta-human chorionic gonadotropin) are taken between 9 and 11 weeks and the results are available at the time of the ultrasound examination that includes nuchal translucency screening. Although the current PE screening with PlGF before 11 weeks is limited,5 our results suggest that its combination with sFlt-1 using the same procedure is associated with good sensitivity and specificity. Nevertheless, the case-control design gives only an estimate of the specificity, and we cannot speculate on specific predictive values. Therefore, large prospective cohorts are required. Our results support the hypothesis that maternal sFlt-1 is low before 11 weeks among women who will develop early PE, and therefore it could be used, in combination with PlGF, as a predictive measure.