Serum biomarkers for the prediction and diagnosis of preeclampsia: a systematic meta-analysis of diagnostic accuracy.

BACKGROUND: Preeclampsia, a life-threatening gestational complication for maternal-fetal health, necessitates prediction (across all trimesters) to improve clinical outcomes. Serum biomarkers, by revealing core pathological mechanisms such as angiogenic imbalance, have emerged as potential screening tools. However, the heterogeneity in their predictive efficacy and clinical utility requires evidence-based synthesis. OBJECTIVE: To comprehensively evaluate the predictive potential of seven categories of serum biomarkers and clarify their clinical application value. METHODS: Databases including PubMed, Embase, Cochrane Library, and Web of Science were searched from inception to January 20, 2026, to identify studies evaluating the diagnostic performance of serum biomarkers for preeclampsia. The methodological quality of included studies was assessed using the QUADAS-2 tool. Pooled sensitivity, specificity, positive likelihood ratio (PLR), negative likelihood ratio (NLR), diagnostic odds ratio (DOR), and area under the curve (AUC) were calculated using bivariate random-effects models and the hierarchical summary receiver operating characteristic (HSROC) model. RESULTS: A total of 22 studies (7,772 participants) were included. The sFlt-1/PIGF ratio demonstrated the highest diagnostic accuracy, with a pooled sensitivity of 85%, specificity of 83%, and AUC of 0.923. For individual biomarkers, sFlt-1 showed a sensitivity of 78% (95% CI: 74-82%) and specificity of 79% (95% CI: 75-83%) with an AUC of 0.881; PIGF showed a sensitivity of 74% (95% CI: 70-78%) and specificity of 76% (95% CI: 72-80%) with an AUC of 0.859. Endoglin and leptin exhibited moderate diagnostic performance (AUCs of 0.814 and 0.768, respectively) but were limited by high heterogeneity (I 2 > 75%). Activin A and inhibin A showed preliminary diagnostic potential (AUCs of 0.785 and 0.796, respectively), but these findings are based on only three studies each and require validation in larger prospective cohorts. Subgroup analysis revealed that diagnostic accuracy varied by gestational trimester, with the sFlt-1/PIGF ratio achieving the highest AUC (0.94) in the second trimester. Sensitivity analyses confirmed the robustness of the pooled estimates, and publication bias was assessed using Deeks' funnel plot. CONCLUSION: This meta-analysis demonstrates that the sFlt-1/PIGF ratio exhibits superior diagnostic accuracy for preeclampsia, particularly during the second trimester.