A new study published in Science Translational Medicine describes a noninvasive technique for sequencing the entire genome of a human fetus. Through analysis of cell-free DNA isolated from maternal blood plasma and containing fetal DNA, the authors were able to predict inherited variation in the fetal genome with 98% accuracy. While new noninvasive prenatal techniques have big implications for the future of prenatal testing and morbidity and mortality rates, they also raise important ethical issues and concerns which will need to be addressed.
Mendelian diseases, genetic disorders caused by mutations within a single gene, are numerous (about 3500 currently) and uncommon in the general population, but collectively make-up ~20% of infant mortality and ~10% of pediatric hospitalizations each year. Current prenatal testing involves either amniocentesis or chorionic villus sampling, invasive procedures that involve a small but potentially deadly risk to the pregnancy. Thus, research has been underway to develop noninvasive techniques for prenatal genetic testing.
13% of cell-free DNA in maternal blood plasma contains fetal DNA. This DNA is increased during pregnancy and then rapidly cleared following birth. Building on techniques recently developed by other research groups, and using a blood sample from the mother and a saliva sample from the father, Kitzman et al. (2012) reported the full genome sequencing of a human fetus at 18.5 weeks of gestation. The authors used the haplotype-resolved genome sequence of a mother, the shotgun genome sequence of a father, and the deep sequencing of maternal cell-free DNA. They went on to replicate the results by sequencing the full genome of a human fetus at 8.2 weeks of gestation, with 95% accuracy. In both cases, cells collected from cord blood following delivery of the child was used to assess accuracy.
Although a major breakthrough in the field of prenatal genetic testing, the authors suggest several areas for improvement. First, there were a large number or sites for which they did not attempt prediction, and improvements in genome-wide haplotyping protocols are required. The authors state: “there remains a critical need for genome-wide haplotyping protocols that are at once robust, scalable, and comprehensive. Significant reductions in cost, along with standardization and automation, will be necessary for compatibility with large-scale clinical application.” Secondly, the authors suggest that improvements in deep sequencing techniques are required to better predict de novo single-nucleotide mutations (new mutations seen only in the fetus and thus representing potential new sources of disease). Finally, while the authors were able to predict single-nucleotide variants (Mendelian diseases), “more robust methods for detecting other forms of variation, for example, insertion-deletions, copy number changes, repeat expansions, and structural rearrangements” will be required.
The noninvasive techniques presented by Kitzman et al. (2012) have huge implications for the future of prenatal testing, and could reach the clinic in the near future. While these new techniques could reduce the incidence of life-threatening genetic disorders and improve prenatal and neonatal treatment for many conditions, the authors warn that more research is required before introduction into clinical practice, in part because “although the noninvasive prediction of a fetal genome may be technically feasible, its interpretation—even for known Mendelian disorders—will remain a major challenge.” To get around some of the ethical issues raised by these new techniques, some researchers recommend only sequencing parts of the genome known to be involved in genetic disorders, but regardless, advances in prenatal testing require “extensive discussions among all stakeholders (physicians, genetic counselors, patient advocacy groups, and the general public), much debate, and great care in implementation.”