Progress toward noninvasive prenatal diagnosis.

For decades, noninvasive analysis of the fetal genotype has been the holy grail of the field of prenatal diagnosis. Noninvasive prenatal diagnosis (NIPD) would use a sample source other than amniocentesis or chorionic villus sampling. The focus has been primarily on the use of maternal blood samples, with less attention given to the possibility of recovering fetal cells from maternal cervical samples (as would be collected for cervical cancer screening). The initial efforts focused on the presence of fetal cells in the maternal circulation, and evidence indicated as early as 1979 that fetal cells could be recovered via fluorescence-activated cell sorting. The elegant strategy of using antibodies to paternal HLA antigens to identify fetal cells was demonstrated. Attempts to recover fetal cells focused for a time on fetal nucleated red blood cells via analysis with fluorescence in situ hybridization and the PCR, but the disappointing results led to skepticism that NIPD would ever become a reality. By 1996 there were reports of tumor DNA detected in the plasma of cancer patients, and Lo et al. first reported in 1997 the presence of fetal DNA in maternal plasma. The use of fetal DNA in the plasma has important advantages over the use of fetal cells in that fetal DNA can be obtained consistently in virtually 100% of pregnancies, whereas fetal cells are recovered much less consistently. In recent years, the use of fetal DNA in maternal plasma to detect fetal sex has become a well-established clinical test [see references in Lo et al. (1 ) for background on this test and other details]. In fact, determining whether any genotype present in the father but not the mother has been transmitted to the fetus is relatively straightforward. This approach has been used to determine whether a fetus is Rh negative or positive in the context of the risk for Rh incompatibility. Similarly, a fetus at risk of inheriting a dominant mutation (e.g., polyposis coli) from the father can be tested for the presence of the mutation. In the case of recessive disorders (e.g., cystic fibrosis) in which the parents have different mutations, one can determine whether the fetus has inherited the paternal mutation. Two newer technologies have greatly altered the prospects for NIPD. One technology is chromosomal microarray analysis, which has opened the possibility of detecting a wide range of disease-causing copy number variants (CNVs), such as those causing DiGeorge, Williams, and Smith–Magenis syndromes. This technology expands the emphasis on NIPD from detection of trisomy 21 alone to broader detection of tens and even hundreds of serious disabilities caused by microdeletions or microduplications. Array comparative genomic hybridization (CGH) enables the detection of CNVs and aneuploidy in single blastomeres. Thus, if single fetal cells could be recovered reliably from maternal blood or the cervix, NIPD could potentially be used to detect all forms of aneuploidy and most disease-causing CNVs. If such cells could be obtained, testing could be implemented with current technology and at costs similar to those currently for invasive prenatal diagnosis. The second new technology is next-generation (Next-Gen) sequencing. This term refers to the many instruments and platforms that permit the collection of enormous amounts of DNA sequence data at lower costs. A recent report by Lo et al. (1 ) applied Next-Gen sequencing to the analysis of fetal DNA in maternal plasma and provided evidence that NIPD is closer at hand than ever. Although the report successfully analyzed a single pregnancy at risk for -thalassemia, the findings have much broader implications. The investigators sequenced plasma DNA to provide 65-fold coverage, meaning that the amount of sequencing was equivalent to that of 65 haploid genomes. They used single-nucleotide polymorphism data from both parents in an elegant computational way to distinguish fetal sequence reads from maternal reads to the extent possible. Importantly, they demonstrated for the first time that the entire fetal genome was represented in a uniform manner, with close to 11% fetal sequence and 89% maternal sequence for each chromosome. This result is equivalent to about 7-fold coverage of the fetal haploid genome or about 3.5-fold coverage of each diploid allele. The authors demonstrated the feasibility of deciphering the entire genome of the fetus from a 1 Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, TX. * Address correspondence to the author at: Department of Molecular and Human Genetics, Baylor College of Medicine, One Baylor Plaza, BCM225, Houston, TX 77030. Fax 713-798-7773; e-mail [email protected]. Received March 30, 2011; accepted March 31, 2011. Previously published online at DOI: 10.1373/clinchem.2011.165563 2 Nonstandard abbreviations: NIPD, noninvasive prenatal diagnosis; CNV, copy number variant; CGH, comparative genomic hybridization; Next-Gen, nextgeneration (sequencing). Clinical Chemistry 57:6 000 – 000 (2011) Perspectives