Abstract
BACKGROUND: Genetic abnormalities in embryos are responsible for most miscarriages and repeated embryo implantation failures, so a reliable preimplantation genetic screening method is urgently needed. Non-invasive preimplantation genetic testing (niPGT) is a potential method for embryo genetic diagnosis. However, the value of its application is controversial. This meta-analysis aimed to investigate and validate the diagnostic value of niPGT in patients undergoing in vitro fertilization (IVF). METHODS: This review used the "Preferred Reporting Items" as a systematic review and meta-analysis of the diagnostic test accuracy (PRISMA-DTA) statement. We searched PubMed, Embase, Web of Science Core Collection, and Cochrane Library up to May 2022 to retrieve non-invasive preimplantation gene detection studies. The eligible research quality was evaluated following the quality assessment study-2 system for diagnostic accuracy. The pooled receiver operator characteristic curve (SROC) and the area under SROC (AUC) were used to evaluate diagnostic performance quantitatively. Threshold effect, subgroup analysis, and meta-regression analysis were used to explore the source of heterogeneity. Deeks' funnel plots and sensitivity analyses were used to test the publication bias and stability of the meta-analysis, respectively. FINDINGS: Twenty studies met the inclusion criteria. The pooled sensitivity, specificity, and AUC were 0.84 (95% CI 0.72-0.91), 0.85 (95% CI 0.74-0.92), and 0.91 (95% CI 0.88-0.93), respectively. Subgroup analysis showed that the spent culture medium (SCM) subgroup had higher sensitivity and lower specificity than the SCM combined with the blastocoel fluid (BF) subgroup. Subgroup analysis showed that the study sensitivity and specificity of < 100 cases were higher than those of ≥ 100. Heterogeneity (chi-square) analysis revealed that sample size might be a potential source of heterogeneity. Sensitivity analysis and Deeks' funnel plots indicated that our results were relatively robust and free from publication bias. INTERPRETATION: The present meta-analysis indicated that the pooled sensitivity, specificity, and AUC of niPGT in preimplantation genetic testing were 0.84, 0.85, and 0.91, respectively. niPGT may have high detection accuracy and may serve as an alternative model for embryonic analysis. Additionally, by subgroup analysis, we found that BF did not improve the accuracy of niPGT in embryos. In the future, large-scale studies are needed to determine the detection value of niPGT.