Detection of fetal copy number variants by non-invasive prenatal testing for common aneuploidies.

Abstract

To evaluate the clinical performance of non-invasive prenatal testing (NIPT) in detecting fetal copy number variants (CNVs). NIPT of cell-free fetal DNA was performed retrospectively, using stored maternal plasma, at an average gestational age of 21.1 weeks in 117 pregnant women who had previously undergone invasive prenatal testing for chromosome microarray analysis (CMA). Among the fetal samples tested by CMA, 18 had positive results (CNVs > 1 megabase (Mb)) and 99 had negative results (CNVs < 1 Mb or without CNVs detected). The results of NIPT and CMA were then compared. In the 11 CMA-positive samples with CNVs > 5 Mb, the detection rate of CNVs was 90.9%; the one case missed by NIPT had a fetal fraction of 4.7%. For the other seven CMA-positive samples with CNVs < 5 Mb, the detection rate was 14.3%; only one case with a 2.82-Mb duplication was detected, with a fetal fraction of 26.7%. For 35.7% (5/14) of CNVs detected by both NIPT and CMA, the differences in fragment length were within ± 1 Mb. In this study, the overall sensitivity and specificity of NIPT for detecting CNVs > 1 Mb were 61.1% and 95.0%, respectively, with a false-positive rate of 5.0%. Our results demonstrate that NIPT for common aneuploidies can detect fetal CNVs > 5 Mb with high sensitivity, provided that fetal fraction is high enough, without increasing sequencing depth. The detection power of NIPT is determined mostly by fetal fraction and CNV size. A positive NIPT screening result for CNVs must be interpreted with caution and validated by additional diagnostic study.