A non-invasive test for prenatal diagnosis based on fetal DNA present in maternal blood: a preliminary study

Background: Use of free fetal DNA to diagnose fetal chromosomal abnormalities has been hindered by the inability to distinguish fetal DNA from maternal DNA. Our aim was to establish whether single nucleotide polymorphisms (SNPs) can be used to distinguish fetal DNA from maternal DNA-and to determine the number of fetal chromosomes-in maternal blood samples. Methods: Formaldehyde-treated blood samples from 60 pregnant women and the stated biological fathers were analysed. Maternal plasma fractions were quantified at multiple SNPs, and the ratio of the unique fetal allele signal to the combined maternal and fetal allele signal calculated. The mean ratios of SNPs on chromosomes 13 and 21 were compared to test for potential fetal chromosomal abnormalities. Findings: The mean proportion of free fetal DNA was 34·0% (median 32·5%, range 17·0-93·8). We identified three samples with significant differences in the fetal DNA ratios for chromosome 13 and chromosome 21, indicative of trisomy 21; the remaining 57 samples were deemed to be normal. Amniocentesis or newborn reports from the clinical sites confirmed that the copy number of fetal chromosomes 13 and 21 was established correctly for 58 of the 60 samples, identifying 56 of the 57 normal samples, and two of the three trisomy 21 samples. Of the incorrectly identified samples, one was a false negative and one was a false positive. The sensitivity and positive predictive value were both 66·7% (95% CI 12·5-98·2) and the specificity and negative predictive values were both 98·2% (89·4-99·9). Interpretation: The copy number of chromosomes of interest can be directly established from maternal plasma. Such a non-invasive prenatal test could provide a useful complement to currently used screening tests.