Predicting fetoplacental chromosomal mosaicism during non-invasive prenatal testing
Maria Neofytou1, Nathalie Brison2, Luc Dehaspe3, Baran Bayindir4, Kris Van Den Bogaert5, Hilde Peeters, Van Esch Hilde6, Griet Van Buggenhout1, Annick Vogels1, Jeroen Breckpot, Thomy de Ravel5, Eric Legius6, Koenraad Devriendt5, Joris Vermeesch5, Leila Dardour7
1Centre for Human Genetics - KU Leuven, Leuven, Belgium, Leuven, Belgium
2Center for Human Genetics, University Hospital Leuven, Leuven, Belgium
3Centre for Human Genetics - KU Leuven, Leuven, Belgium., LEUVEN, Belgium
4Dr, LEUVEN, Belgium
5Center for Medical genetics, Katholieke Universiteit Leuven, Leuven, Belgium
6Center for Human Genetics, KU Leuven, Leuven, Belgium
72Department of Human Genetics, Faculty of Medicine , Sousse, Tunisia
Objectives
Non-invasive prenatal detection of trisomies 21, 18 and 13 can be achieved with high accuracy through sequencing of maternal plasma-derived cell-free DNA (cfDNA). However, fetoplacental mosaicism is a main cause for false positive/negative NIPT results. We further improved the analytical power of genome-wide cfDNA screening by predicting the occurrence of fetoplacental mosaicism.
Methods
Trisomy Z-scores increase linearly with increasing fetal fractions. Z-scores that do not correlate with the actual detected fetal fraction are indicative of the presence of placental chromosomal mosaicism. This analysis pipeline was applied to whole genome sequencing data derived from ~20.000 maternal plasma samples. Following an abnormal NIPT, test results were validated by conventional invasive prenatal or postnatal genetic testing.
Results
The new analysis pipeline identified 134, 24 and 7 non-mosaic trisomies 21, 18 and 13 respectively. All for whom follow-up information was available were confirmed upon invasive testing. The incidence of other, rare autosomal trisomies (RATs) was ~0.3%, with trisomy 7 and 16 being the most prevalent. Three of these RATs, predicted as full trisomies in the placenta, were found to be mosaic in the fetus; 25 other RATs were predicted to be mosaic, 8 of which have been confirmed in placental tissue. The new pipeline also correctly predicted twin pregnancies with discordant fetal sex.
Conclusions
This improved analysis pipeline permits the detection of autosomal aneuploidies and pinpoints pregnancies at risk of fetoplacental mosaicism. This knowledge can influence the estimation of the risk for miscarriage, aid in genetic counseling and improve prenatal management.
