NEW YORK — Whole-genome sequencing is a promising alternative to currently used assays to diagnose fetal anomalies, a new analysis has found.
Fetal structural anomalies arise in about 3 percent of pregnancies and can range from mild to severe. While karyotyping, FISH, and array-based approaches have been used to identify genetic causes of fetal anomalies uncovered by ultrasound, sequencing approaches are being adopted or explored to improve diagnostic rates.
In a recent study, researchers from Mount Sinai Hospital in Toronto sequenced the full genomes of a prospective cohort of 37 fetuses with structural anomalies detected by ultrasound, while also performing standard diagnostic testing. Sequencing, they found, had a diagnostic yield of nearly 20 percent and could identify both sequence variants and CNVs as well as variants of uncertain significance, or VUS, with potential clinical impact in 19 percent of cases, suggesting that genome sequencing could replace the range of tests that fetuses with structural anomalies often undergo.
"Genome sequencing is now widely used postnatal and it's getting more and more popular," said Abdul Noor, the director of genetic diagnostics at Mount Sinai. "So, we wanted to see how we can utilize it in the prenatal setting, and particularly we were interested in comparing the diagnostic yield … to the current standard-of-care tests in our own institution."
But whether full genome sequencing is a better approach than exome sequencing is not yet clear.
For their analysis, Noor and his colleagues recruited 42 pregnant individuals with a fetus with one or more structural anomalies to their study, which they reported in the journal Prenatal Diagnosis.
DNA samples from the fetuses underwent standard prenatal diagnostic testing. In Canada, the current prenatal diagnostic protocol includes rapid aneuploidy detection, or RAD, by either QF-PCR or FISH to detect aneuploidies of chromosomes 13, 18, 21, X, and Y, followed by microarray analysis. Fetuses negative on those tests are then offered exome sequencing analysis. The 37 fetuses with negative RAD results then underwent singleton genome sequencing.
For five cases, Noor and his colleagues identified a pathogenic or likely pathogenic sequence variant that could explain the ultrasound findings and present a molecular diagnosis. This included variants in NIPBL, FOXF1, RERE, and FLT4 as well as in the X-linked gene, AMMECR1. The researchers additionally uncovered structural variants in 5 percent of cases.
At the same time, they identified seven VUS, in 19 percent of samples, that could be informative. They noted that one VUS they uncovered was later found through a trio analysis to be a de novo mutation and was reclassified to likely pathogenic.
In all, genome sequencing had a diagnostic yield of 19 percent, but the researchers further argued that with the VUS data, whole-genome sequencing provided clinically useful or potentially useful data for 38 percent of cases. This, they noted, is higher than the 6 percent diagnostic yield for chromosomal microarrays and the 8.5 percent to 10 percent diagnostic yield reported for exome sequencing after negative RAD and CMA findings.
Noor added that while their study was not designed to look at trio analysis, he suspected that if they did trio analyses or follow-up targeted analyses in parents, the diagnostic yield could be closer to 30 percent, which is about what is seen in postnatal testing.
But others argue that full-genome sequencing does not currently offer that much more information over exome sequencing in the prenatal diagnostic space. "I don't see genome sequencing at the moment as a diagnostic tool," said Heinz Gabriel, a clinical laboratory geneticist at Praxis für Humangenetik Tübingen, a German diagnostic lab that offers prenatal trio exome sequencing as a standard test for both post- and prenatal cases.
He and his colleagues recently examined the diagnostic yield of exome trio sequencing in a cohort of 500 pregnancies with fetal anomalies found on ultrasounds. They reported in the journal Prenatal Diagnosis that they could detect pathogenic or likely pathogenic variants in 37.8 percent of cases. Nearly half of those, they noted, were de novo variants.
Gabriel added that in Noor's study, the variants that were uncovered by genome sequencing would have also been found through exome sequencing, which Noor also acknowledged. At the same time, Gabriel said genome sequencing could miss mosaicism that exome sequencing could capture due to its generally higher coverage level. In their analysis, Gabriel and his colleagues noted evidence of mosaicism in seven fetuses suspected to be causative.
However, genome sequencing has the advantage of better CNV calling, Noor said — a benefit Gabriel also noted. Noor added that their analysis uncovered two cases of pathogenic CNVs. Additionally, Noor said a benefit of a full genomic sequencing approach is that data that could be beneficial throughout the individual's life.
Gabriel noted, though, that researchers and clinicians do not know what to make of many intronic variants at the moment, leading him to not yet see the benefit of genome sequencing, except perhaps in the uncommon case of translocations.
But the situation could change, he said. "Let's talk together in two years or three years," Gabriel added. "I will tell you probably a different story, but at the moment I don't see the benefit in this."
Noor said he plans to push toward offering full genome sequencing in the prenatal diagnostic clinic at Mount Sinai. Currently, he is exploring opportunities with partners like the Ministry of Health in Ontario. In particular, he wants to conduct a large-scale pilot study in the same vein as the Genome-Wide Sequencing Ontario project — a two-year pilot project involving the health ministry, the Hospital for Sick Children, and the Children's Hospital of Eastern Ontario that is offering clinical genome sequencing for rare diseases — but for prenatal diagnoses.
"That will help us to establish the infrastructure we need for this and refine our workflow and analysis," Noor said.