Candidate Circulating Biomarkers of Spontaneous Miscarriage After IVF-ET Identified via Coupling Machine Learning and Serum Lipidomics Profiling

Reprod Sci. 2022 Mar;29(3):750-760. doi: 10.1007/s43032-021-00830-w. Epub 2022 Jan 24.

Abstract

Spontaneous miscarriage is a common pregnancy complication. Multiple etiologies have been proposed such as genetic aberrations, endocrinology disorder, and immunologic derangement; however, the relevance of circulating lipidomes to the specific condition remains unclear. In the present study, lipidomics profiling was examined on serum of women with spontaneous miscarriage after in vitro fertilization and embryo transfer (IVF-ET). Screening and analysis of differential lipid levels were conducted using a machine learning approach to verify the stability and validity of potential serum biomarkers. Seven lipid species presented significant differences between the abortion and term birth patients, including three types of sphingomyelins (SMs), two types of diglycerides (DGs), one phosphatidylcholine (PC), and one lysophosphatidylethanolamine (LPE). All the SMs presented with a fold change of > 1, while both the PC and LPE had a fold change of < 1. The DG containing two saturated fatty acyl chains was decreased, but that containing two unsaturated fatty acyl chains was increased in the miscarriage group compared to the control group. This study reveals the relevance of lipid profiles to spontaneous abortion after IVF-ET, providing potential biomarkers and therapeutic targets for the specific clinical scenario.

Keywords: Biomarker; In vitro fertilization and embryo transfer; Lipidomics; Machine learning; Spontaneous miscarriage.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Abortion, Spontaneous*
  • Adult
  • Biomarkers / blood*
  • Embryo Transfer*
  • Female
  • Fertilization in Vitro*
  • Humans
  • Lipidomics / methods*
  • Machine Learning*
  • Pregnancy

Substances

  • Biomarkers