American Journal of Law & Medicine

What potent blood: non-invasive prenatal genetic diagnosis and the transformation of modern prenatal care.

What potent blood hath modest May, What fiery force the earth renews, The wealth of forms, the flush of hues....

--Ralph Waldo Emerson (1)


Someday soon, virtually any pregnant woman (2) will be able to learn--with 98-99% accuracy--whether her fetus has contracted a serious genetic disorder by undergoing nothing more than an inexpensive, non-invasive blood test. For years, scientists have sought a method of prenatal testing that could boast both high levels of accuracy and low levels of risk. The most promising solution lies in an exciting recent discovery: tiny quantities of fetal cells and DNA cross over into the mother's bloodstream during pregnancy. (3) If the fetal genetic material can be successfully isolated from the maternal blood, it can be used to diagnose a number of genetic disorders, such as Down Syndrome, cystic fibrosis, Tay-Sachs disease, and sickle cell anemia. Indeed, researchers have spent the last decade developing ways to accomplish this.

These new blood tests promise significant advantages over present methods of prenatal testing. Unlike current prenatal screening tests, like ultrasound and chemical assays, this new technology could attain extremely high levels of accuracy and be performed as early as 6-10 weeks' gestation. (4) Unlike current prenatal diagnostic tests, like amniocentesis and chorionic villus sampling ("CVS"), the new genetic tests would be non-invasive; as such, they would pose no risk of miscarriage and could be offered to women of all ages and risk levels. (5)

This article introduces the emerging technology of non-invasive prenatal genetic diagnosis ("NPGD') and argues for its impending potential to revolutionize modern prenatal care. In particular, clinical implementation of NPGD--which is non-invasive, accurate, and inexpensive--could dramatically increase the availability of prenatal genetic testing to all pregnant women, change the standard of care, reduce the incidence of serious genetic disorders, and raise (with even greater force and urgency than past advancements in genetics) numerous ethical, legal, and social questions.

Part II offers a brief overview of the two modern methods of prenatal genetic diagnosis: (6) amniocentesis and CVS, both of which are considered "invasive" procedures and pose some risk to both the mother and the developing fetus. Part III explains the science behind two potential noninvasive alternatives for prenatal genetic testing, which I call "maternal serum fetal cell sorting" ("MSFCS") and "maternal plasma DNA recovery" ("MPFDR"). Although clinical implementation of these tests is still years away, scientists expect them to offer high levels of accuracy, early intervention options, and significantly lower prices and costs. Part IV proceeds from the assumption that researchers will successfully develop a highly accurate, clinical version of NPGD and attempts to explain some of the initial legal and social implications, including: NPGD's likely effect of dramatically increasing the number of pregnant women who utilize prenatal genetic testing, its capability of becoming the new standard of care, and its potential to garner both public and private funding through insurance. Finally, Part V discusses several long-term consequences of the likely widespread use of NPGD.


Today, what is termed "prenatal testing" can involve both screening and diagnostic tests. Screening tests impose a lower threshold of accuracy and merely help identify an at-risk population for additional testing, while diagnostic tests are held to stringent accuracy standards (on the order of 98-99% accuracy) and result in a conclusion regarding the fetus's disease status. (7)

Many women who undergo prenatal genetic testing begin with a screening

test. The most common screening tests include: (1) the maternal serum a-fetoprotein ("MSAFP") assay, a maternal blood test measuring levels of [alpha]-fetoprotein, which are considerably higher in cases of Down Syndrome; (8) (2) the "multiple-" or "triple-marker" screen, which measures maternal blood concentrations of a-fetoprotein, as well as two other chemicals associated with chromosomal abnormalities; (9) and (3) ultrasonography, which is used to screen for physical abnormalities (e.g. shortened or missing fetal limbs) and to assess the likelihood of chromosomal defects like Down Syndrome (e.g. by examining "nuchal lucency," the thickness of the skin at the back of the fetal neck). (10) It is important to note that such screening tests are non-invasive (i.e. do not involve entry into the mother's uterus) and do not actually sample fetal genetic material.

If a screening test suggests elevated risks--or for other women who are known to be at elevated risks because of their age or genetic history--a diagnostic test is used to determine with greater certainty whether the fetus has a genetic or chromosomal condition. The two principal diagnostic tests used are: (1) amniocentesis and (2) CVS. (11) Although these tests achieve a very high level of precision (about 99% diagnostic accuracy), (12) both have notable drawbacks. Amniocentesis and CVS are invasive procedures, entailing the use of a needle or catheter to enter the mother's uterus and obtain a sample of fetal tissue. (13) As a result, both tests also involve a small, but significant, risk to the mother and the fetus.

Amniocentesis, which is typically conducted at 15 to 17 weeks' gestation, recovers a sample of fluid from the amniotic sac by means of a needle inserted through the abdomen and guided by real-time ultrasound. (14) The resulting twenty to thirty milliliter sample of amniotic fluid will contain a notable concentration of living cells that have been sloughed off by the fetus. (15) Clinicians then analyze DNA recovered from these cells to diagnose particular genetic conditions. Test results may take up to three weeks to receive, depending upon the particular method used for DNA analysis. (16) The three primary methods are: (17) (1) karyotyping, which reveals chromosomal abnormalities; (18) (2) fluorescent in-situ hybridization ("FISH"), which is employed to diagnose both chromosomal and single-gene abnormalities; (19) and (3) polymerase chain reaction ("PCR'), which is usually used to diagnose single-gene disorders. (20) Because these methods are common to all types of fetal genetic diagnosis--whether the sample is obtained by amniocentesis, by CVS, or by non-invasive means--they are worth describing briefly here.

Karyotyping--a physical, as opposed to molecular, type of analysis--is the most commonly used method today. (21) It involves: (1) culturing ("feeding" and growing) the sample cells, (2) "arresting" cell division at the stage where chromosomes are most clearly visible, (3) staining and photographing the chromosomes, and (4) arranging the chromosomes in corresponding pairs according to function and size. (22) With a full "map" of genetic material, clinicians can determine whether the fetus's chromosomes exist in the correct number, shape, and size. (23) An abnormal number of certain chromosomes indicates a genetic disorder, such as Down Syndrome (3 copies of chromosome 21). (24) Karyotyping typically takes 10-12 days, but may take up to 3 weeks because of the extra time required to culture the cells from the sample. (25)

In contrast, FISH and PCR, the two "molecular" methods of analysis, take considerably less time (only about 2-3 days) to complete. (26) In FISH, a segment of DNA coding for a specific chromosome (usually X, Y, 13, 18, or 21) is labeled with fluorescent "probe" molecules, (27) so that it may then be used to detect the presence of the complementary DNA sequence in target cells to which it will pair itself. (28) If too many (or too few) copies of a given chromosome (or piece of a chromosome) appear colored within the same fetal cell, this indicates an abnormality. (29)

PCR, often used on a single fetal cell, (30) can replicate all or part of a cell's DNA in large quantities. Where a sample is small, PCR can increase the amount of genetic material to be analyzed. (31) PCR may also be used to amplify (i.e. make multiple copies of) a particular portion of DNA known to code for a certain genetic condition, in order to search for abnormalities in that locus. (32) This is typically done in bulk, on a sample of many cells, but may also be conducted on a single fetal cell, which increases reliability by reducing interference by non-fetal cells or DNA. (33)

Although "[t]he safety ... of amniocentesis ha[s] been established by several large collaborative studies," (34) the test is not without some risk to both the mother and the fetus. The procedure causes membrane rupture and fluid leakage in about 1-2% of pregnant women. (35) Mothers may also experience bleeding, infections, or early initiation of labor. (36) With respect to the fetus, amniocentesis increases the risk of miscarriage from about 1% (37) to about 1.5%--a 50% increase over the background rate of miscarriage during the second trimester. (38) In rare cases, it may also cause injury to the umbilical cord, (39) or fetal brain damage. (40)

CVS may be performed earlier than amniocentesis, at approximately 10 to 11 weeks' gestation. (41) In this test, a physician inserts a catheter into the uterus (either transcervically or through the abdominal wall) and takes a tissue sample from the chorionic villi, the finger-like projections that make up part of the placenta. (42) As with amniocentesis, fetal cells are isolated, and DNA analysis is conducted on the genetic material obtained from the cells' nuclei by means of karyotyping (most common), FISH, or PCR.(43)

The primary risk associated with CVS is an increased likelihood of miscarriage. CVS increases this risk from about 3%, (44) to about 4-4.5% (or 133-150% above the background rate at the end of the first trimester). (45) Thus, the miscarriage rate of CVS is higher than that of amniocentesis. Additionally, research suggests--and biology does not refute--that CVS may lead to an increased occurrence of birth defects, such as shortening of the limbs, fingers, or toes. (46)

Relatively few pregnant women undergo amniocentesis or CVS, in part because of the augmented risks and in part because of the costs. Because of the fetal and maternal hazards highlighted above, medical professionals recommend the procedures almost exclusively for women aged 35 and older, (47) unless the fetus is at particular risk for genetic disease or abnormality (e.g. because of a positive family history). (48) Another limiting factor is cost. Amniocentesis and CVS typically cost between $1100-$1200, but may range as high as $2000. (49) Insurance coverage of these procedures varies, but coverage is more likely for women over 35 whose screening tests returned positive results, and for others (like those with a family history for abnormalities) for whom the tests may be deemed a "medical necessity." (50) As a result of the foregoing factors, only about 88,000 women underwent amniocentesis in 2001; (51) 79,000 in 2002; (52) and 67,000 in 2003 (53) (out of over 4,000,000 live births and about 5,000,000 pregnancies in each year). (54) The number of women undergoing CVS testing was likely even lower. In 2002, a study of genetic testing at the University of Connecticut Health Center recorded only 55 CVS referrals, as compared with 878 amniocentesis referrals--a rate of under 6% (of all prenatal genetic testing), which had remained relatively consistent in the years between 1991 and 2002. (55) The numbers were higher at the Stanford University Medical Center, but CVS nonetheless made up only 11% of the prenatal diagnostic procedures performed in 2004, and 12% of the procedures performed in 2005. (56) Thus, if an approximate rate of 10% is applied comparably across the nation, it would indicate a total of only about 8,800 CVS procedures in 2001; 7,900 in 2002; and 6,700 in 2003 (out of 5,000,000 pregnancies). In contrast, ultrasound, which is less precise than amniocentesis and CVS, was utilized in about 2,700,000 pregnancies in 2001-2003, most likely because it is a relatively inexpensive and generally noninvasive procedure. (57)


Two alternative diagnostic tests are currently being developed that would enable doctors to obtain genetic material from the fetus without intruding into the uterine area. Thus, in contrast to the current invasive procedures, these methods pose virtually no risk of injury to the mother and eliminate the procedure-related risks to the fetus. Both of the new tests rely on the presence of minute quantities of fetal genetic material that cross the placenta and circulate in the mother's bloodstream during the pregnancy. (58) Each method merely requires a blood sample from the mother. (59) Section A describes the first method, maternal serum fetal cell sorting ("MSFCS"), which entails isolating intact fetal cells from maternal blood and analyzing the DNA within them. (60) Section B details the second method, maternal plasma fetal DNA recovery ("MPFDR"), which involves analyzing segments of cell-free fetal DNA found circulating in the mother's blood. (61) Neither procedure is ready for clinical use at present, yet each offers distinct possibilities and presents particular challenges.


The presence of fetal cells in maternal circulation was first discovered in the late nineteenth century by German pathologist Christian Georg Schmorl. (62) Twentieth-century research has confirmed the existence of such fetal cells (as well as other types) in maternal blood, (63) but has also underscored their rarity. Although their relative concentration increases as the pregnancy progresses, fetal cells typically make up only about 1 in 1,000,000 to 1 in 10,000,000 cells in maternal blood serum. (64) One group estimated that there were only 19 fetal cells in a 16-milliliter blood sample. (65) As a result, isolating such cells for use in genetic testing has posed significant challenges. Since the 1980s, much of the research in MSFCS has focused on "enrichment" techniques, which aim to identify various types of fetal cells and recover them in increased quantities from maternal blood. (66)

The initial question is what type of fetal cell is easiest and most cost-effective to isolate. Researchers have experimented with three principal types: (1) trophoblasts (embryonic cells responsible for forming the placenta), (67) (2) leukocytes (white blood cells), (68) and (3) erythroblasts (nucleated red blood cells, or NRBCs). (69) Although trophoblasts were the first type to be detected and are easy to identify microscopically because of their unique shape, they also have a high rate of mosaicism ("mixing" with maternal tissue) that complicates analysis, and they are rapidly depleted from the mother's bloodstream, making them harder to locate. (70) Fetal leukocytes have been successfully isolated by several research teams, (71) but pose difficulties because of the lack of specific antibody "markers" to separate them from maternal cells, (72) and because leukocytes from prior pregnancies may persist in maternal blood for many years following birth or spontaneous abortion. (73) Erythroblasts are the most attractive candidate for isolation because NRBCs are rare in adult blood, appear early in the pregnancy, are less likely to persist in maternal blood following pregnancy, and possess several potential antibodies for use in enrichment. (74) Even after enrichment, however, maternal cells still outnumber fetal erythroblasts, raising the possibility of accidental analysis of maternal (rather than fetal) genetic material. (75)

Several potential enrichment procedures have been tested in recent years. In fluorescence-activated cell sorting ("FACS"), target cells, such as NRBCs, (76) are "tagged" with a fluorescent antibody specific to such cells. (77) The tagged cells, with their bright labels, are identified using a "computer-assisted laser-guided device." (78) Magnetic-activated cell sorting ("MACS"), on the other hand, uses antibodies affixed to magnetic beads that "tag" the target cells. (79) When the combined cell mixture (tagged and untagged) is passed through a magnetic field, the desired fetal cells are retained, and the undesired cells flow through and are discarded. (80) MACS is often favored because it is relatively less expensive and does not require a high level of technical expertise. (81) Alternatively, some researchers have experimented with density gradients, which separate cell types based on differentials in relative mass. (82) Alone, or in conjunction with the above techniques, scientists may also employ "micromanipulation," in which a single fetal cell is identified by microscope and recovered for analysis. (83)

After enriching for fetal cells (or isolating a single fetal cell), scientists can then conduct genetic analysis on the fetal genetic material via karyotyping, FISH, or PCR. (84) These methods operate no differently for MSFCS than for the analysis of an amniocentesis or CVS sample.

Therefore, an ideal MSFCS procedure would proceed as follows: at approximately 8-12 weeks' gestation, a pregnant woman arrives at her doctor's office, where she gives a standard 10 milliliter (about 1/3 of an ounce) blood sample. The sample is then enriched for fetal erythroblasts by means of an efficient, automated process of cell sorting, with reliable, fetus-specific antibody markers. The enriched sample is then subjected to multicolored FISH (to search for more than one chromosomal condition at once) or PCR analysis of the DNA, where target abnormalities may be detected. (85) Results are obtained in less than a week.

The benefits of MSFCS are manifold. First and foremost, its non-invasive nature eliminates the risk of miscarriage and other dangers to the mother and the fetus. Moreover, because MSFCS could be used earlier in the pregnancy than amniocentesis (and about the same time as CVS), it would provide families with more time to make difficult decisions concerning the termination of an affected fetus, and allow for a safer and easier termination, if that is the ultimate choice. Moreover, following successful isolation of the fetal cell(s), medical professionals could diagnose virtually any genetic condition for which there are reliable genetic markers. MSFCS preserves the complete fetal genome and may be used to analyze any and all of it in searching for genetic abnormalities. So far, MSFCS has been successfully used to detect fetal gender, (86) muscular dystrophy, (87) Rh incompatibility, (88) and common chromosomal abnormalities like Down Syndrome. (89)

Despite its advantages, the MSFCS is not without challenges, which currently prevent the technology from being put to clinical use. First, there is still a large amount of "background noise," (90) primarily from the vast quantity of maternal cells that remain in the blood sample following enrichment. "Even after enrichment, most cells are likely to be maternal," with "perhaps one [fetal cell] per 100 to 1000." (91) Part of the problem is the lack of sufficiently specific antibody markers. To increase yield, scientists continue to search for more reliable and fetus-specific antibodies, but such markers are still elusive. (92) The more reliable the markers, the more conducive the cell sorting process will be to automation, an ultimate goal of researchers. (93) Indeed, because the future utility of this procedure depends upon achieving a level of efficiency and cost-effectiveness that surpasses current invasive procedures like amniocentesis, automation of cell sorting is critical. Moreover, although it is currently easiest to test for only one disorder at a time, increased accuracy and reliability of multicolored FISH, (94) as well as better techniques for micromanipulation of a single fetal cell for PCR analysis, could allow scientists to efficiently screen for multiple abnormalities at once.

The most troubling problem, however, is the persistence of fetal cells from prior pregnancies (successful or unsuccessful) in maternal blood, which may confound enrichment and analysis of the current fetus's cells. Research studies have demonstrated the continued proliferation of fetal leukocytes in maternal blood 1 year post partum, (95) and 5 years postpartum. (96) One laboratory discovered male genetic material in one mother's blood a full 27 years following the birth of her last son. (97) There are few solutions to this dilemma, except perhaps to compare isolated fetal cells to the cells of a living child to determine whether the cells originated with the current fetus or the child (i.e. previous fetus). This method would not apply, however, where the fetal cells came from earlier miscarriages or abortions, where no basis for comparison exists. This issue would become especially problematic if chromosomally abnormal fetal cells remained in the mother's blood after the birth of an affected child or, more likely, after spontaneous abortion of an affected fetus. (98) The persistence of abnormal cells could lead to a false positive and possibly result in an unnecessary abortion.

All of the studies above, however, were performed on white blood cells and their precursors. In contrast, fetal erythrocytes ("NRBCs') have a lifespan of only about 90 days, (99) making them less likely to confuse prenatal genetic diagnosis via MSFCS. In fact, the authors of one study suggested, as a potential solution to the problem of fetal cell persistence, the use of a "highly differentiated fetal cell type that is unlikely to proliferate, like the fetal nucleated erythrocyte." (100) Indeed, targeting fetal NRBCs may largely solve the dilemma (except in the case of very recent pregnancies),if the NRBCs can be reliably isolated.

Nevertheless, the present-day accuracy and reliability of MSFCS reaches, at best, the level of a screening test. (101) With further research and refinement of procedures, it is possible that MSCFS may reach diagnostic-level precision.


Experimentation with cell-free fetal DNA in maternal blood is a much more recent phenomenon, but one that promises a number of advantages over intact cell isolation (as well as its own set of limitations). First discovered by Y. M. Dennis Lo et al. in 1997, (102) cell-free fetal DNA has been "reliably detected" in maternal serum as early as 5 weeks' gestation. (103) Such DNA exists in small fragments, and while its origin is presently unknown, researchers suspect that it results from the disintegration or death of intact fetal cells and subsequent passage through the placenta. (104) Significantly, cell-free fetal DNA has been shown to appear in maternal blood in greater quantities than intact fetal cells. According to one study, cell-free fetal DNA made up an average of 8.4% of all cell-free DNA in maternal plasma (0.13% in maternal serum) in early pregnancy, and an average of 6.2% of all cell-free DNA in maternal plasma (1.0% in maternal serum) in late pregnancy. (105) Thus, cell-free fetal DNA in maternal plasma may prove easier to enrich, if enrichment is necessary at all, (106) because of its already high relative presence.

Two principal enrichment techniques have been attempted. First, some researchers have found that treating maternal plasma with formaldehyde reduces breakage of maternal cells during analysis, thus minimizing "background" cell-free maternal DNA. (107) This, in turn, makes fetal DNA easier to recover and results in higher yields. Subsequent researchers, however, have failed to replicate these results, instead noting a "lack of dramatic enrichment of fetal DNA" after using formaldehyde. (108) The second technique is "size-fractionalization," which exploits the recently discovered discrepancy between the approximate size of cell-free fetal DNA and that of cell-free maternal DNA. (109) Because fetal DNA fragments are typically smaller (fewer than 300 base pairs [bp]) than maternal DNA fragments (greater than 500 bp), fetal DNA will travel farther when passed through a dense gel and may be extracted with little contamination by maternal genetic material. (110) This technique has thus far met with moderate success.

Following enrichment (or even without substantial enrichment), MPFDR may employ either FISH or PCR to analyze the fetal genetic material. (111) Karyotyping would not be possible for MPFDR, because cell-free fetal DNA exists in fragments, rather than in full chromosomes. (112) This fragmentary nature of fetal DNA should not pose any procedural problems, however, because FISH and PCR typically require cell and DNA breakage already. …

Log in to your account to read this article – and millions more.