NEW YORK (GenomeWeb) – In a correspondence article published last night in the New England Journal of Medicine, researchers from Illumina and the Tufts Medical Center presented evidence suggesting false-positive results related to maternal copy number can be better winnowed out of non-invasive prenatal test results using an updated analytical method.
For the Illumina-funded work, the team re-analyzed sequence data generated for 1,914 pregnant women for the Comparison of Aneuploidy Risk Evaluations (CARE) study, which focused on women believed to be at elevated fetal aneuploidy risk.
Using an analytic algorithm employed in Illumina's clinical laboratory, which was recently updated to bump up normalization steps related to sub-chromosomal coverage and to better filter out sub-chromosomal copy number variants, the investigators found that they were able to correctly reclassify six of the 11 false-positive cases in the pregnancies considered.
Of those, three of the corrected cases had previously been considered aneuploid based on false-positive detection of maternal CNVs, while the other three re-classifications hinged on enhanced normalization steps, authors of the correspondence article noted.
A 2015 study in NEJM by researchers at the University of Washington and the Fred Hutchinson Cancer Research Center suggested that maternal copy number variants may impact the interpretation of NIPT, potentially leading to false-positive results.
In the Illumina-Tufts re-analysis, the five false-positive cases that were not re-reclassified in the new analysis, appeared to stem from autosomal copy number gains and/or fetal mosaicism.
Overall, the researchers' updated analytical algorithm increased the positive predictive value for trisomy 21 from 46 percent to 63 percent, while the PPV for trisomy 18 edged up from 40 percent to 67 percent. For trisomy 13, meanwhile, PPV jumped from 24 percent to 50 percent.
"Re-analysis of the CARE data suggests that there are both technical and biologic reasons for the original false-positive results," they wrote, "including fetoplacental mosaicism and variants in maternal copy number."
Along with efforts to curb false positives from CNVs in NIPT technologies, there is ongoing debate over the feasibility and utility of attempting to detect CNVs and other sub-chromosomal abnormalities using fetal DNA in maternal circulation.
In an article published in the American Journal of Human Genetics late last year, for example, a team from the UK reported it had limited success picking up microdeletions and microduplications smaller than 6 million bases, with mothers carrying three of the five duplications detected in that size range.
Just last week, Swiss researchers reporting in the journal Genetics in Medicine presented evidence in favor of expanding NIPT to include CNVs and other sub-chromosomal abnormalities, as well as autosomal trisomies that are rarer than trisomy 21, trisomy 18, or trisomy 13 (the three most common trisomies).
In data from nearly 6,400 consecutive singleton pregnancies, that team saw three-dozen suspicious CNVs, 50 rare autosomal trisomies, and 53 sex-chromosome abnormalities, along with 119 common trisomies. Even so, authors of that study called for a focus on disease-associated, recurrent CNVs, since the phenotypic effects of most CNVs are not yet known.