It is likely that, soon after amniocentesis became accepted widely for the prenatal diagnosis of fetal chromosome and certain structural abnormalities, clinicians and researchers took interest in developing technologies that would permit prenatal detection of fetal abnormalities without placing pregnancies at increased risk for fetal loss as the result of invasive procedures. The development of such technologies would require considerable work, first to determine whether evaluation of the fetus was even possible by direct analysis of fetal tissues or cells that were obtained directly from the mother without interrupting the fetoplacental unit and then whether such analysis could be developed that would differentiate abnormal from normal pregnancies accurately. Accordingly, the initial, and presumably more achievable, first step would be the development of more effective approaches to screening the population that would reduce the use of invasive procedures by the identification of those pregnancies that were truly at increased risk for detectable abnormalities and the reduction of the number of procedures that are offered to women who carry unaffected fetuses. Although the personal and family history was and is considered to be a vital component of the counseling process, it is clear that it is a poor “stand-alone” screening tool for the identification of pregnancies that are at an increased risk for detectable fetal abnormalities, most of which (eg, fetal chromosome abnormalities) are not associated with Mendelian disorders.

The development of diagnostic ultrasound scanning clearly provided for a safe approach to the identification of certain fetal abnormalities, especially structural anomalies. However, despite continuing advancements in technology and expertise, it continues to fall short of the sensitivity and specificity that is needed to be considered a diagnostic test for many of the conditions that are evaluated by prenatal cytogenetic and molecular testing. In the 1970s, the identification of the association of amniotic fluid and the maternal serum alpha-fetoprotein with open neural tube defects and open structural abnormalities provided an early approach to the diagnosis (amniotic fluid) and screening (maternal serum) for these fetal abnormalities. Indeed, an early step in the improvement of screening for fetal chromosome abnormalities was the recognition that low levels of maternal serum alpha-fetoprotein in the second trimester were associated with an increased risk for fetal Down syndrome and trisomy 18, which provided for the initial nonhistoric and biochemical approach to population-based screening for fetal chromosome abnormalities. Soon thereafter, the incorporation of other maternal serum analytes and ultrasound scanning for the measurement of the fetal nuchal translucency in the late first trimester were found to improve the detection rate, extend screening to the first trimester, and expand the number of fetal abnormalities that were amenable to screening.

Such screening algorithms improved the detection rates for fetal Down syndrome and certain other fetal chromosome abnormalities (eg, trisomy 18), reduced the overall number of invasive procedures that were being performed by providing a more accurate screening of the general and high-risk population and were clearly superior to detection rates that had been achieved by maternal age and family history estimations alone. However, such screening protocols are viewed by some obstetricians as cumbersome and somewhat confusing to incorporate into the obstetrics practice. If a patient is found to be at an increased risk for fetal Down syndrome and the diagnostic test result is normal, patients and clinicians frequently are concerned that the screening protocol is flawed. All too often, patients report that their friend was told that her baby had Down syndrome but that the test turned out to be wrong. In fact, nothing was wrong; the actual story was almost always that her friend was told that her screening test was positive for Down syndrome but that the diagnostic test (ie, chorionic villus sampling, amniocentesis) was normal. Educating clinicians and their patients about the differences between screening and diagnostic tests eventually facilitated the incorporation of such screening tests (for Mendelian disorders such as Tay-Sachs, sickle cell disease, and cystic fibrosis) into routine preconception and prenatal care. Nonetheless, all such screening tests eventually led to an invasive diagnostic test to provide for a definitive fetal clinical outcome, tests that were and continue to be associated with an increased risk for fetal loss. With the increasing number of fetal conditions that are amenable to prenatal screening, it was and is the increased risk for fetal loss, albeit now shown to be negligible in some recent studies, that has kept many women from considering such testing and has encouraged the development of even more effective screening algorithms and noninvasive prenatal diagnostic testing.

The aforementioned prenatal screening modalities all incorporate surrogate fetal markers; none of the protocols use fetal-derived cells, analytes, or tissues that could be collected easily from the mother and potentially provide a more accurate evaluation of a particular fetal abnormality. Similarly, sonographic approaches to screening use a set of nonspecific “soft markers” to adjust a woman’s risk. The modern attempt to incorporate such maternally obtained fetal tissues began with the work by Walknowska et al in 1969, who treated maternal blood from 22 women who each were carrying a male fetus with a lymphocyte mitogen and then recovered 46,XY cells from 19 of the samples. In 1979, Herzenberg reported the use of fluorescence-activated cell sorting that could be used to enrich the apparently minuscule population of fetal cells in maternal blood and then be used to identify the rare cells. These technologic advancements, especially with regard to cell sorting, led to the work by Bianchi et al, who were the first to enrich the fetal nucleated erythrocytes in maternal blood by sorting with a monoclonal antibody to CD71, which is an antigen that is found in nucleated red blood cells. At the same time, Lo et al in Hong Kong reported on the detection of Y-specific DNA sequences in maternal blood samples by polymerase chain reaction; soon thereafter Price et al and Elias et al reported on the first-trimester prenatal diagnosis of fetal trisomy 21 in fetal cells that had been obtained from maternal blood that was enriched by flow cytometry and detected by fluorescent in situ hybridization. Others reported on a variety of other enrichment and cell separation techniques and other candidate fetal cells that included trophoblasts and lymphocytes. However, for a variety of reasons that ranged from cost to reproducibility, no single technology or cell type emerged as a likely candidate for the development of an effective prenatal screening or diagnostic modality.

That is, until now. Ehrich et al, using a multiplexed massively parallel shotgun sequencing assay for fetal trisomy 21 and based on the fundamental work by Lo et al that demonstrated the presence of circulating cell-free fetal DNA in maternal plasma, obtained 480 prospectively collected samples and eventually evaluated 449 samples for fetal trisomy 21. The assay correctly identified all 39 samples that had been obtained from women who carried a fetus with trisomy 21 and misclassified only a single sample as being trisomy 21 in a woman who carried a fetus with 2 copies of chromosome 21. The overall classification showed 100% sensitivity (95% confidence interval, 89–100%) and 99.7% specificity (95% confidence interval, 98.5–99.9%). These findings are comparable with those reported by Chiu et al who, using similar technology in a separate and independent study, found a slightly lower sensitivity and slightly higher specificity with increased (8-plex vs 2-plex) multiplexing.

Although this preliminary study presents promising results, current screening and diagnostic modalities are not yet ready to be pushed aside. The results of this study point more to a continuation in the improvement in the various noninvasive aneuploidy detection technologies that have been studied for the past 2 decades; hence, the results of this study can be considered a small step. Further studies that would include economic feasibility studies, portability studies, and larger studies to better assess the detection rate of this assay for fetal trisomy 21, especially in lower-risk populations, are required before such technology is incorporated into prenatal screening algorithms, let alone used to replace current screening protocols.

However, one can also look at this study and its results and consider it to be a giant leap. Granted, numerous studies of other noninvasive prenatal fetal aneuploidy detection technologies have shown promising initial results. Why is this technology potentially the giant leap? First, the study is not the initial evaluation of this technology, but rather the study was undertaken after several pilot studies demonstrated important technologic issues that were used to improve the overall approach to the evaluation of maternal plasma samples for fetal trisomy 21. Although further and more extensive studies of this technology are needed to best assess its ability to detect fetal trisomy 21 in maternal blood samples from a more comprehensive risk cohort, the technology presented here potentially is amenable for the study and screening of other aneuploidies and even gene mutations and rearrangements.

Whenever a new technology demonstrates outstanding clinical results, the best approach for clinicians is caution. This technology is clearly not ready to replace invasive prenatal testing, or even prenatal screening, because it evaluates only a single fetal chromosome abnormality. With regard to screening, it is hoped that this technology eventually may be used to further improve current screening algorithms, with thoughts to replace our current evidence-based approaches to prenatal screening being best entrusted to researchers and prognosticators. Clinicians must fight the urge, which will likely be promoted and encouraged by the lay press, to consider the current study to be the representation of a new and improved approach to prenatal diagnosis that will replace amniocentesis and chorionic villus sampling. Although health and science reporters are required to simplify new technological advancements considerably to best communicate them to the public, we must not fall into a “new clinical development” stupor that could arise from this study. Does the current study present promising results? Yes. Is there a new modality for screening pregnant women for fetal Down syndrome? Not yet. Is it time to throw away the needles and catheters that are needed for invasive prenatal diagnosis? Not for a while.

The first step on the moon by Neil Armstrong was the result of numerous studies, test flights, missions, and even failures, many of which occurred concurrently or after successes and failures by the Soviet space program. This current study by Ehrich et al is an impressive and important first step, although other technologies and other fetal tissues and cells may yet be shown to be as or more effective than massively parallel shotgun sequencing in the development of more effective prenatal screening protocols or even the development of a reliable noninvasive prenatal diagnosis. Many more steps and studies will be needed to arrive at the goal that most of us who provide and study prenatal screening and diagnosis have long sought after: a facile, economical, and more accurate approach to prenatal screening and, eventually, an effective noninvasive alternative to invasive prenatal testing. We may not be there yet, but it seems that we have come a bit closer to our ultimate objective.