ABSTRACT
PURPOSE: Copy-number variant (CNV) assessment is recommended for patients undergoing prenatal diagnostic testing. Noninvasive screening tests have not been extensively validated for CNV detection. The objective of this study was to compare the ability of genome-wide noninvasive prenatal screening (NIPS) to chromosomal microarray to detect clinically significant findings. METHODS: We prospectively enrolled 198 subjects at the time of consent for diagnostic prenatal testing. Genome-wide NIPS results were compared with diagnostic testing results to assess NIPS test performance (n = 160, 38 subjects without microarray results excluded). Cohen's kappa statistic was used to assess test agreement. RESULTS: Genome-wide NIPS did not detect clinically significant chromosomal abnormalities at the same rate as diagnostic testing, κ = 0.75 (95% confidence interval [CI], 0.62-0.87). When excluding CNVs <7 Mb and findings outside the limits of genome-wide NIPS, test agreement improved, κ = 0.88 (0.79-0.97) driven by agreement for common aneuploidies (κ = 1.0). However, among patients with an abnormal fetal survey, agreement was only fair, κ = 0.38 (0.08-0.67). CONCLUSION: While NIPS is an excellent screening test for common aneuploidies, genome-wide NIPS misses clinically significant findings detected on routine diagnostic testing. False positive and false negative cases highlight the importance of pretest counseling regarding NIPS limitations, especially in the setting of fetal anomalies.
