Genome-wide non-invasive prenatal testing in single- and multiple-pregnancies at any risk: Identification of maternal polymorphisms to reduce the number of unnecessary invasive confirmation testing

Abstract

Objective: Non-invasive prenatal testing by targeted or genome-wide copy number profiling (cnNIPT) has the potential to outperform standard NIPT targeting the common trisomies 13, 18, and 21, only. Nevertheless, prospective results and outcome data on cnNIPT are still scarce and there is increasing evidence for maternal copy number variants (CNVs) interfering with results of both, standard and cnNIPT.

Study design: We assessed the performance of cnNIPT in 3053 prospective and 116 retrospective cases with special consideration of maternal CNVs in singleton and multiple gestational pregnancies at any risk, as well as comprehensive follow-up.

Results: A result was achieved in 2998 (98.2%) of total prospective cases (89.2% analyzed genome-wide). Confirmed fetal chromosomal abnormalities were detected in 45 (1.5%) cases, of which five (11%) would have remained undetected in standard NIPTs. Additionally, we observed 4 likely fetal trisomies without follow-up and a likely phenotype associated placental partial trisomy 16. Moreover, we observed clinically relevant confirmed maternal CNVs in 9 (0.3%) cases and likely maternal clonal hematopoiesis in 3 (0.1%). For common fetal trisomies we prospectively observed a very high sensitivity (100% [95% CI: 91.96-100%]) and specificity (>99.9% [95% CI: 99.8-100%]), and positive predictive value (PPV) (97.8% [95% CI: 86.1-99.7%]), but our retrospective control cases demonstrated that due to cases of fetal restricted mosaicism the true sensitivity of NIPT is lower. After showing that 97.3% of small CNVs prospectively observed in 8.3% of genome-wide tests were mostly benign maternal variants, sensitivity (75.0% [95% CI: 19.4%-99.4%]), specificity (99.7% [99.5%-99.9%]) and PPV (30.0% [14.5%-52.1%]) for relevant fetal CNVs were relatively high, too. Maternal autoimmune disorders and medication, such as dalteparin, seem to impair assay quality.

Conclusion: When maternal CNVs are recognized as such, cnNIPT showed a very high sensitivity, specificity and PPV for common trisomies in single and multiple pregnancies at any risk and very good values genome-wide. We found that the resolution for segmental aberrations is generally comparable to standard karyotyping, and exceeds the latter if the fetal fraction is above 10%, which allows detection of the 2.5 Mb 22q11.2 microdeletion associated with the velocardiofacial syndrome, even if the mother is not a carrier.

Keywords: Cell free DNA; Copy number variants; Genome-wide NIPT; NIPT; Non-Invasive prenatal testing.