Noninvasive screening for congenital heart defects using a serum metabolomics approach

Objective Heart anomalies represent nearly one-third of all congenital anomalies. They are currently diagnosed using ultrasound. However, there is a strong need for a more accurate and less operator-dependent screening method. Here we report a metabolomics characterization of maternal serum in order to describe a metabolomic fingerprint representative of heart congenital anomalies. Methods Metabolomic profiles were obtained from serum of 350 mothers (280 controls and 70 cases). Nine classification models were built and optimized. An ensemble model was built based on the results from the individual models. Results The ensemble machine learning model correctly classified all cases and controls. Malonic, 3-hydroxybutyric and methyl glutaric acid, urea, androstenedione, fructose, tocopherol, leucine, and putrescine were determined as the most relevant metabolites in class separation. Conclusion The metabolomic signature of second trimester maternal serum from pregnancies affected by a fetal heart anomaly is quantifiably different from that of a normal pregnancy. Maternal serum metabolomics is a promising tool for the accurate and sensitive screening of such congenital defects. Moreover, the revelation of the associated metabolites and their respective biochemical pathways allows a better understanding of the overall pathophysiology of affected pregnancies.

Automated summary

Metabolomic profiling of maternal serum was used to identify a unique fingerprint for fetal heart anomalies, resulting in accurate classification models and the discovery of key metabolites. This approach shows promise for sensitive and effective screening of congenital defects, providing insights into the underlying pathophysiology of affected pregnancies.