Accuracy and clinical value of maternal incidental findings during noninvasive prenatal testing for fetal aneuploidies

Genome-wide sequencing of cell-free (cf)DNA of pregnant women aims to detect fetal chromosomal imbalances. Because the largest fraction of cfDNA consists of maternal rather than fetal DNA fragments, maternally derived copy-number variants (CNVs) are also measured. Despite their potential clinical relevance, current analyses do not interpret maternal CNVs. Here, we explore the accuracy and clinical value of maternal CNV analysis. Noninvasive prenatal testing was performed by whole-genome shotgun sequencing on plasma samples. Following mapping of the sequencing reads, the landscape of maternal CNVs was charted for 9,882 women using SeqCBS analysis. Recurrent CNVs were validated retrospectively by comparing their incidence with published reports. Nonrecurrent CNVs were prospectively confirmed by array comparative genomic hybridization or fluorescent in situ hybridization analysis on maternal lymphocytes. Consistent with population estimates, 10% nonrecurrent and 0.4% susceptibility CNVs for low-penetrant genomic disorders were identified. Five clinically actionable variants were reported to the pregnant women, including haploinsufficiency of RUNX1, a mosaicism for segmental chromosome 13 deletion, an unbalanced translocation, and two interstitial chromosome X deletions. Shotgun sequencing of cfDNA not only enables the detection of fetal aneuploidies but also reveals the presence of maternal CNVs. Some of those variants are clinically actionable or could potentially be harmful for the fetus. Interrogating the maternal CNV landscape can improve overall pregnancy management, and we propose reporting those variants if clinically relevant. The identification and reporting of such CNVs pose novel counseling dilemmas that warrant further discussions and development of societal guidelines. Genet Med19 3, 306–313.