Comparative analysis of vascular endothelial growth factor levels in placental, fetal-membrane, and umbilical-cord tissues: associations with maternal obstetric complications

Abstract

Introduction Vascular endothelial growth factor (VEGF) is a central regulator of placental angiogenesis, but the extent to which tissue-level VEGF expression differs across obstetric complications and among gestational compartments remains unclear. This study evaluated VEGF protein concentrations in placental villous tissue, fetal membranes, and umbilical cord across diverse pregnancy phenotypes. Methods A prospective observational cohort study with subsequent group classification was conducted at King Abdullah University Hospital, Irbid, Jordan. Eighty-six pregnancies were classified into five groups: uncomplicated controls, small-for-gestational-age (SGA), pre-eclampsia/eclampsia, gestational diabetes mellitus (GDM), and thrombophilia. VEGF concentrations were quantified by enzyme-linked immunosorbent assay (ELISA) in placental, membrane, and umbilical cord tissues obtained at delivery. Between-group differences were assessed using non-parametric analyses, while multinomial logistic regression, Spearman correlation, principal component analysis, clustering, and cross-validated receiver operating characteristic (ROC) analysis were performed. Results VEGF concentrations showed substantial overlap among clinical groups. No significant differences were observed in placental VEGF ( P = 0.802), membrane VEGF ( P = 0.139), or umbilical cord VEGF ( P = 0.525), with negligible effect sizes ( ε 2 ≤ 0.04). Multivariable modelling demonstrated no significant association between any VEGF compartment and pathological pregnancy groups. Correlations between compartments were weak, with modest associations between membrane and umbilical cord VEGF ( ρ = 0.29, P = 0.006) and placental and membrane VEGF ( ρ = 0.22, P = 0.046). Multivariate analyses failed to identify distinct VEGF expression patterns, and a three-marker classifier showed poor discrimination after cross-validation (AUC = 0.52). Conclusion In this cohort, total tissue VEGF measured at delivery did not clearly distinguish pregnancies complicated by SGA, pre-eclampsia/eclampsia, GDM, or thrombophilia from uncomplicated pregnancies. The limited sample size, however, means that smaller effects cannot be excluded. The weak inter-compartment correlations may be more consistent with locally regulated angiogenic processes than with a unified tissue signature, although sampling and methodological factors may also contribute. These findings highlight the limitations of total VEGF tissue measurement as a biomarker of placental disease.