RESUMO
BACKGROUND: Accurate prediction of the individual risk of venous thromboembolism (VTE) remains suboptimal in children, and biomarkers are currently not used to stratify the risk of VTE in children. OBJECTIVES: This study aimed to assess which biological or radiological biomarkers may predict VTE or VTE complications in children. PATIENTS/METHODS: A literature search was performed for peer-reviewed publications (1990-2022). We included studies addressing the use of biomarkers for patients aged 29 days to 18 years to predict VTE or its complications, including but not limited to TE-related death, VTE recurrence, or postthrombotic syndrome. Given the heterogeneity of the study designs, populations, and outcomes, no quantitative data synthesis was performed. RESULTS: Forty studies were included, totaling 10,987 participants (median age: 4.7 years). Reports were often lacking critical methodological data, including blood collection method (68% of studies) and timepoints, laboratory testing technique (41%), or primary outcome definition (20%). Forty-six individual biomarkers were assessed for VTE prediction (32 studies, 9525 participants), including d-dimers, fibrinogen, platelet count, white blood cell count, and factor VIII. Albumin, C-reactive protein, d-dimers, factor VIII, and thrombin-antithrombin levels showed promising results for VTE prediction. In 9 studies (1606 participants), no biomarker was consistently predictive of postthrombotic syndrome, VTE persistence, or VTE recurrence in children. CONCLUSIONS: Several candidate biomarkers were promising in the prediction of VTE in children. Still, discrepancies between different studies and the high risk of bias from the current literature prevent their widespread use in the clinical setting. Further prospective research in various pediatric subpopulations is required.
