RESULTS
Eighty-six women were admitted with the diagnosis of early-onset severe PE during the study period, 68 of them fulfilled the inclusion criteria and had no maternal complications and no fetal indication for immediate delivery. Five were excluded for not collecting blood samples for angiogenic factors due to a breach of the study protocol, leaving 63 women for analysis.
A total of 18 (28.6%) pregnancies had an adverse neonatal outcome, non-exclusively including 2 (3.2%) stillbirths, 4 (6.4%) neonatal demise, 1 (1.6%) neonatal acidosis, 9 (14.3%) 5-min Apgar score< 7, 5 (7.9%) bronchopulmonary dysplasia, 1 (1.6%) necrotizing enterocolitis, 1 (1.6%) grade III intraventricular hemorrhage, 2 (3.2%) hypoxic-ischemic encephalopathy, 1 (1.6%) acute renal failure and 3 (4.8%) cardiac failures. Table 1 details the characteristics of the study population, pregnancy outcomes and the at-admission parameters by the occurrence of adverse neonatal outcomes. Of note, among the angiogenic factors (PlGF, sFlt-1 and, sFlt-1/PlGF ratio), the PlGF showed the largest difference between affected and unaffected babies, and it was used in the subsequent multivariate models. Table 2 shows the multivariate analysis for the association between at-admission parameters and adverse neonatal outcomes.
Figure 1 and Table 3 show the predictive performance of different combinations of at-admission predictors. Compared with the PREP-L score, both the PREP-L + severe FGR (p=0.041) and PREP-L + PlGF (0.012) significantly added predictive value. The combination of all parameters (PREP-L score, severe FGR and PlGF) did not improve further the prediction capacity.