News

Currently, prenatal diagnosis of recessive autosomal diseases is only possible through invasive testing, such as amniocentesis and chorionic villus sampling, which carries around a 1% risk of miscarriage. However, a new technique developed by Dennis Lo’s laboratory in Hong Kong – based on the presence of cell-free fragments of fetal DNA present in the maternal bloodstream during pregnancy – has raised the possibility that such diseases could potentially be prenatally diagnosed non-invasively [Lun  et al (2008) PNAS 105: 19920-19925].

 

To date, research into the non-invasive prenatal diagnosis of single gene disorders using cell-free fetal DNA has been limited to detection of paternally inherited mutations, which are not otherwise present in the maternal genome. This limits the clinical applications to diagnosis of the inheritance of paternal autosomal dominant diseases, and exclusion of the inheritance of certain autosomal recessive diseases where the maternal and paternal alleles differ. However, in cases where the woman is a heterozygous carrier of a disease – with one normal copy of a gene, and one diseased copy – prenatal diagnosis requires determining whether the maternal disease mutation has been inherited by the fetus.

 

The new technique measures the ‘relative mutation dosage’ of a particular gene in all cell-free DNA in the maternal bloodstream, i.e. the ratio of the diseased to the normal alleles. During pregnancy, the presence of even a small amount of fetal DNA in the maternal bloodstream will affect this ratio, and both an underrepresentation or overrepresentation of the disease allele suggest that the fetus is homozygous (and normal or affected respectively); if the alleles are balanced, then the fetus is a carrier. This allele ratio can be measured using a variety of techniques, the most obvious of which is digital PCR, which allows the number of copies of a particular sequence to be easily counted. However, although this ‘molecular counting’ technique is conceptually simple, it is practically challenging due to the relatively low proportion of fetal DNA versus maternal DNA (less than 10% in early pregnancy). The researchers tackled this problem by exploiting the fact that fetal cell-free DNA fragments are generally shorter than the maternal ones. Using an innovative extension to standard digital PCR, the target gene derived specifically from the fetus was selectively amplified and detected, thus effectively enriching the sample with fetal DNA and increasing the potential affect on the allele ratio.

 

However, despite this technical wizardry, the technique is far from being clinically validated. Although the technique proved to be highly accurate on artificially created test mixtures of male and female blood samples, only 5 out of 10 fetuses at risk of either b-thalassaemia or haemoglobin E were correctly diagnosed at 18-20 weeks by measuring the relative mutation dosage (in the absence of fetal enrichment). When fetal DNA enrichment was combined with this method, 4 out of 5 fetuses were correctly genotyped at 12-14 weeks, using a locus of interest for Down syndrome testing on chromosome 21.

 

Comment: This exciting proof-of-concept study has brought non-invasive prenatal diagnosis of single gene diseases a step closer to clinical application, and most importantly, has offered the first practical method for diagnosing recessive diseases. However, before it could be realistically offered by antenatal services, further laboratory development will be needed to make the technology practicable for use by clinical service laboratories, and extensive clinical evaluation in large cohorts will be required to establish the accuracy of the technique for numerous single gene diseases.