Screening for trisomies in circulating DNA.

In the broadest sense, noninvasive prenatal testing began in the 1970s, when the first articulatedarm diagnostic ultrasound machines produced two-dimensional images of fetuses in utero. Initially, the crude new imaging technology was eagerly adopted for the simple measurement of embryos, fetuses, and fetal parts to develop normative data for gestational-age determination. But as technology and image quality rapidly improved, it quickly became obvious that fetal anatomy could be defined with increasing precision. At approximately the same time, Brock and Sutcliffe1 recognized that alpha-fetoprotein, which is found in very high concentrations in fetal blood and cerebrospinal f luid, could be measured in maternal serum at much lower concentrations. Elevated levels of this protein were associated with, but not perfectly predictive of, open neural-tube defects and other fetal abnormalities. Merkatz et al.2 recognized that unusually low serum levels of maternal alphafetoprotein were associated with an increased risk of fetal Down’s syndrome, and the rush was on to develop other maternal serum and sonographic markers to improve the positive and negative predictive values for the prenatal screening process. What is fundamentally different about the new era in noninvasive prenatal diagnosis of aneuploidy is that for the first time, the analyte of interest in maternal blood is DNA and not a proxy analyte associated with, but not directly responsible for, the fetal phenotype of interest. Lo et al.3,4 opened the door to the new era when they described procedures for identifying qualitatively different DNA sequences in the maternal circulation that could originate only in a fetus: Y chromosome sequences from a male fetus and DNA encoding the rhesus D antigen (so-called Rh factor) in women who were Rh-negative. Initially, these sequences were obtained by disrupting rare fetal cells in the maternal circulation, but later they were obtained from the small fragments of cell-free DNA (cfDNA) that are shed by the placenta into the maternal circulation. The proportion of fetal cfDNA circulating in maternal blood increases during gestation and represents about 10% of free DNA in circulating maternal blood during the first and second trimesters. The most recent advances have been made possible by the development of very rapid DNA sequencing and computer hardware and software capable of matching hundreds of millions of DNA fragments and sequences in hours. Sequencing of the circulating DNA obtained from a woman carrying a euploid fetus provides random sequence reads (the nucleotide sequence of stretches of DNA) of each chromosome. Sequence reads that are derived from a specific autosomal chromosome have a 1:1 ratio with the reads of any other autosomal chromosome. Fetuses with aneuploidy have a chromosomal dosage imbalance, which can be detected through deviations from a 1:1 ratio, in the number of chromosomespecific sequence reads. However, the relatively high proportion of maternal DNA in maternal blood has posed a challenge to the sensitive detection of fetal aneuploidy. Massive parallel sequencing, also called nextgeneration sequencing, generates millions of sequence reads along the length of each chromosome, thereby overcoming the challenge of excess maternal DNA and providing the sensitivity