Dynamic change in placental growth factor in patients with preeclampsia and its predictive efficay for pregnancy outcomes

Abstract Objective To analyze the dynamic change in placental growth factor (PLGF) level in patients with preeclampsia (PE) and its role in predicting pregnancy outcomes. Methods A retrospective cohort study was conducted on 198 PE patients admitted to Department of Obstetrics of Tianjin Beichen Hospital from January 2021 to January 2024. Among them, 151 cases were identified as non-severe PE and 47 cases as severe PE. During the same period, another 100 healthy pregnant women were included in the department and served as the control group. According to the final pregnancy status of the PE patients, they were divided into an adverse outcome group (88 cases) and a good outcome group (110 cases). Locally Weighted Scatterplot Smoothing (LOWESS) was used to analyze the relationship between vascular endothelial function indicators and PLGF level. Logistic regression was employed to identify the influencing factors for adverse pregnancy outcomes. Receiver operating characteristic (ROC) curve was plotted to evaluate the value of PLGF level for predicting adverse pregnancy outcomes in patients with non-severe PE and severe PE. Results The PLGF level was in a peak shape at different gestational weeks, relatively low at 28~29 +6 weeks of gestation, gradually increasing with the increment of gestational weeks, reaching a peak at 30~31 +6 weeks of gestation, and then gradually decreased (P<0.05). The PLGF level was gradually reduced in the control group, non-severe PE group, and severe PE group in turn (P<0.05). LOWESS analysis showed that the PLGF level exhibited certain nonlinear relationship with soluble fms-like tyrosine kinase receptor 1 (sFlt-1), vascular endothelial growth factor (VEGF), nitric oxide (NO), thrombomodulin (TM), thromboxane A2 (TXA2), and Von Willebrand factor (VwF). The results of logistic regression analysis indicated that after gradually eliminating collinearity confounding factors and adjusting for each covariate, there was still an independent correlation between the PLGF level at 28~29 +6 gestational weeks and pregnancy outcome (OR=0.593, 95%CI: 0.583~0.778, P<0.001). After converting the PLGF level into a binary variable, there was an independent correlation between high PLGF level and pregnancy outcome (OR=0.773, 95%CI: 0.671~0.885, P<0.001). Compared with the quintile with the lowest PLGF (Q1), as the PLGF level gradually increased (Q2 to Q5), the correlation effect values were (OR=0.765, 95%CI: 0.639~0.837, P=0.092), (OR=0.752, 95%CI: 0.624~0.818, P=0.056), (OR=0.696, 95%CI: 0.615~0.813, P=0.027), and (OR=0.642, 95%CI: 0.591~0.733, P<0.001), and the trend test was statistically significant (Ptrend<0.001). ROC curve analysis revealed that the area under the ROC curve (AUC) of serum PLGF level at 28-29+6 gestational weeks in predicting adverse pregnancy outcomes in non-severe PE patients was 0.830 (95%CI: 0.635~0.917), the optimal critical threshold was 75.47 ng/L, the sensitivity was 88.37%, the specificity was 71.30%, the positive predictive value was 55.07%, and the negative predictive value was 93.90%. Its AUC value of the level for the prediction in patients with severe PE was 0.874 (95%CI: 0.677~0.923), the optimal critical threshold was 38.53 ng/L, the sensitivity was 88.89%, the specificity was 50.00%, the positive predictive value was 97.56%, and the negative predictive value was 16.67%. Conclusion At 28~29+6 gestational weeks, PLGF has an independent correlation with pregnancy outcomes and shows a good predictive performance for adverse pregnancy outcomes in patients with non-severe PE and severe PE.