A screening test proposal for congenital defects based on maternal serum metabolomics profile

ABSTRACT

Background

Historically, non-invasive techniques are only able to identify chromosomal anomalies that accounted for less than 50% of all congenital defects, while the others are diagnosed via ultrasound evaluations in later stages of pregnancy. Metabolomic analysis may provide a crucial improvement, potentially addressing the need for novel non-invasive and multi-comprehensive early prenatal screening tools. Indeed, a growing body of evidence outlines notable metabolic alterations in different biofluids derived from pregnant women carrying malformed fetuses, suggesting that such an approach may allow the discovery of biomarkers common to most fetal malformations. In addition, metabolomic investigations are inexpensive, fast, and risk-free, and often provide high diagnostic performance also allowing an early diagnosis.

Objective

The purpose of this study was to evaluate the diagnostic accuracy of an ensemble machine learning model based on maternal serum metabolomic signatures for detecting fetal malformations, including both chromosomal anomalies and structural defects.

Study Design

We describe a multi-center observational retrospective study which includes two different arms. In the first, a total of 654 Italian pregnant women (334 cases, carrying malformed fetuses and 320 controls, with normal developing fetuses) were enrolled and used to train an ensemble machine learning classification model based on serum metabolomics profiles. In the second arm, serum samples obtained from 1935 participants of the New Zealand SCOPE study were blindly analyzed and used as a validation cohort. Untargeted metabolomics analysis was performed via gas chromatography-mass spectrometry. Nine individual machine learning classification models were built and optimized via cross-validation (Partial Least Square Discriminant Analysis, Linear Discriminant Analysis, Naïve Bayes, Decision Tree, Random Forest, k-nearest neighbor, Artificial Neural Network, Support Vector Machine, and Logistic regression). Then, an ensemble of them was developed according to a voting scheme statistically weighted by the cross-validation accuracy and classification confidence of the individual models. This ensemble machine learning system was used to screen the validation cohort.

Results

Significant metabolic differences were detected in women carrying malformed fetuses, who exhibited lower amounts of palmitic, myristic and stearic acids, N-α-acetyllysine, glucose, L-acetylcarnitine, fructose, p-cresol and xylose and higher levels of serine, alanine, urea, progesterone and valine (p<0.05) when compared to controls. When applied to the validation cohort, the screening test showed a 99.4%±0.6% accuracy (specificity=99.9±0.1% [1892/1894 controls correctly identified] with a sensitivity=78±6% [32/41 fetal malformations correctly identified]).

Conclusion

In conclusion, the present study provides a clinical validation of a metabolomics-based prenatal screening test to detect the presence of congenital defects. Further investigations are needed in order to enable the identification of the type of malformation as well as to confirm these findings on even larger study populations.