Prenatal Diagnosis of Congenital Disorders

Likelihood Ratios

The impact of a positive screening test depends on the pretest (a priori) risk of an affected pregnancy. Likelihood ratios are the likelihood that a specific test result would be expected in a patient with the target disorder compared with the likelihood that the same result would occur in a patient without the target disorder; that is, likelihood ratios are basically a ratio of the probability that a test result is correct to the probability that the test result is incorrect.

For binary risk factors that are either present or absent, likelihood ratios are determined by comparing the frequency of positive tests in a population of affected pregnancies with the frequency of a positive test in unaffected pregnancies. This is calculated as the sensitivity of the test divided by its false-positive rate. For tests that use continuous variables (e.g., serum marker measurements), likelihood ratios are calculated by comparing gaussian distributions of the results from normal and affected pregnancies. Once a likelihood ratio is determined, it can be used to modify the a priori risk ( Table 30.2 ). If more than one likelihood ratio is available and if these are independent of each other, they can each be used separately to modify risk. In this way, multiple factors (e.g., individual serum analyte results and ultrasound findings) can be simultaneously used to modify the a priori risk (e.g., maternal age).

Maternal Age as a Screening Test

The association of Down syndrome with advancing maternal age was first reported in 1909. It was not until 50 years later that karyotype analysis was developed and correlated the Down syndrome phenotype with an extra G chromosome. Based on recommendations from a National Institute of Child Health and Human Development (NICHD) consensus conference in 1979, invasive testing for Down syndrome began being offered routinely to women age 35 years or older at delivery, consistent with a second-trimester risk of 1:270 or higher (liveborn risk of 1:380). The age of 35 seemed a natural cutoff because women age 30 to 34 had a risk of 1:880, and the risk for women age 35 to 40 was almost fourfold higher. Other factors initially considered in determining this threshold included the relative risk of invasive testing compared with the risk of trisomy 21, the availability of resources capable of performing amniocentesis and karyotype, and the cost-to-benefit analysis.

The risk for Down syndrome is now recognized to be continuous, which emphasizes the arbitrary nature of an absolute age threshold of 35. In addition to maternal age, the risk for trisomy 21 depends on the gestational age at which testing is performed, as only 69% of first-trimester and 76% of second-trimester Down syndrome pregnancies are viable ( Table 30.3 ). In addition, although the risk for Down syndrome increases with age, 70% of affected pregnancies occur in women younger than 35. For these reasons as well as because more refined risk analysis has been developed, maternal age alone is no longer used as an independent indication for invasive testing.

In 2007 (reaffirmed in 2016 and 2020), the American College of Obstetricians and Gynecologists (ACOG) recommended that both screening and diagnostic testing should be available to all women who present for prenatal care before 20 weeks of gestation regardless of maternal age. The decision for a woman older than 35 to have invasive testing is based on many factors, including the risk that the fetus will have an abnormality, the risk for pregnancy loss from an invasive procedure, and the patient’s perceived consequences of having an affected child. Studies of women’s preferences have shown that they weigh these potential outcomes differently. ^,

First-Trimester Screening for Aneuploidy: First-Trimester Biochemical Screening

Two maternal serum analytes are useful for first-trimester aneuploidy risk modification. Pregnancy-associated plasma protein A (PAPP-A) has been demonstrated to have a mean value of 0.4 multiples of the median (MoM) in trisomy 21 pregnancies. The free β subunit of human chorionic gonadotropin (hCG) is elevated in Down syndrome pregnancies, with a mean value of 1.8 MoM. Screening using PAPP-A alone identifies about 40% to 45% of trisomy 21 pregnancies, and free β-hCG identifies about 23%, both with a screen-positive rate of 5%. Combining both free β-hCG and PAPP-A can identify 60% to 65% of trisomy 21 pregnancies, for a similar 5% screen-positive rate.

The total hCG molecule can also be used for first-trimester screening but has slightly less discrimination than the free β subunit, especially at less than 11 weeks’ gestation. Free β-hCG begins to perform as a Down syndrome marker by 9 weeks’ gestation, reaching values almost twice those in unaffected pregnancies by 13 weeks. Levels of total hCG begin to increase above levels in unaffected gestations at 11 weeks. ^, The impact of substituting total hCG for the free β subunit on overall Down syndrome screening remains uncertain. A meta-analysis showed that in younger patients (<35 years old), detection of Down syndrome increased by 4, 5, 6, and 7 percentage points at 9, 10, 11, and 12 weeks when free β subunit was added to PAPP-A and nuchal translucency (NT), compared with 0, 0, 2, and 4 percentage points when intact hCG was added. In patients with advanced maternal age (>35 years old), inclusion of free β-hCG reduced the false-positive rate by 2.5, 3.1, 3.8, and 4.4 percentage points compared with 0.1, 0.3, 1.0, and 2.2 percentage points for intact hCG at 9, 10, 11, and 12 weeks. Other authors have found less impact. Using samples from the First and Second Trimester Evaluation of Risk (FASTER) study, Canick and coworkers showed that at 12 weeks’ gestation, the addition of free β-hCG to NT and PAPP-A increased detection only 0.9% (−3.3 to 6.3). However, at earlier gestational ages, the impact of free β-hCG would be greater.

First-Trimester Ultrasound Screening for Aneuploidy

In his initial description of the syndrome that bears his name, Langdon Down described skin so deficient in elasticity that it appeared to be too large for the body. This was particularly noticeable in the neck area. The skin and underlying lymphatic fluid in the fetal neck can now be seen with ultrasound at 10 to 12 weeks’ gestation. NT is defined as the collection of fluid under the skin behind the neck in fetuses between 11 and 14 weeks’ gestation. It can be successfully measured by transabdominal ultrasound examination in approximately 95% of cases, and quantification of NT is used for first-trimester Down syndrome screening.

An association between increased NT and fetal chromosomal defects is now well known. This association allows the NT measurement to be converted into a likelihood ratio. Because the median NT increases with gestational age and thus crown-rump length, the actual measurement must first be converted into gestational age–specific MoM or a measurement of the deviation from the expected mean (delta NT) before conversion into a likelihood ratio. Fig. 30.2 illustrates the NT measurement between 11 and 14 weeks’ gestation.

The performance of the NT test combined with maternal and gestational age to assess the risk for Down syndrome was studied in a series of more than 100,000 pregnancies. The risk for Down syndrome was calculated by the maternal age–specific and gestational age–specific prevalence multiplied by the NT likelihood ratio, with a cutoff of greater than 1:300 used as screen positive. The sample included 326 fetuses with trisomy 21; 82% of trisomy 21 fetuses were identified, for a screen-positive rate of 8.3%. When a screen-positive rate of 5% was selected, the sensitivity was 77% (95% confidence interval [CI], 72%–82%). Subsequent studies have demonstrated similar Down syndrome detection rates, between 70% and 75% ( Table 30.4 ).

A screening paradigm using an ultrasound measurement to determine a likelihood ratio is reliable only if NT is measured in a standard fashion. Standards for NT measurements include the following:

• 1.

  The minimal crown-rump length should be 45 mm, and the maximal length should be 84 mm. The success rate for accomplishing a measurement for these gestational ages is between 98% and 100%. The success rate decreases to 90% at 14 weeks and onward.

• 2.

  Either transabdominal or transvaginal scanning can be used; about 95% of cases can be imaged by the transabdominal route.

• 3.

  A true midline sagittal view of the fetal spine must be seen in the cervical and thoracic region with the tip of the nose seen in the face area and the third and fourth ventricle seen in the fetal central nervous system.

• 4.

  The magnification must be such that the fetal head, neck, and upper thorax occupy greater than 50% of the image.

• 5.

  Care must be taken to clearly distinguish between the fetal skin and the amnion. At this gestational age, both structures appear as thin membranes. The structures can be distinguished by either waiting for spontaneous fetal movement away from the amniotic membrane or by bouncing the fetus off the amnion by asking the mother to cough or by tapping on her abdomen ( Fig. 30.3 ).

  

• 6.

  The maximal thickness of the subcutaneous translucency between the skin and the soft tissue overlying the cervical spine should be measured by placing the “+” calipers on the inner borders of the nuchal space with none of the horizontal crossbar protruding into the space, as illustrated in Fig. 30.4 .

  

• 7.

  The maximum measurement is recorded and used for Down syndrome risk calculation.

• 8.

  The NT should be measured with the fetal head in the neutral position. When the fetal neck is hyperextended, the measurement can be increased by 0.6 mm, and when the neck is flexed, the measurement can be decreased by 0.4 mm.

• 9.

  The umbilical cord is found around the fetal neck in approximately 5% to 10% of cases, which may produce a falsely increased NT, adding approximately 0.8 mm to the measurement. In such cases, the measurements of NT above and below the cord differ, and the smaller measurement is the more appropriate one.

Even with these criteria, standardization of NT measurements is difficult. The ability to achieve a reliable measurement has been linked to the motivation of the sonographer. One study compared results obtained from hospitals where NT measurement was used for clinical purposes with results from hospitals where NT was merely measured but not acted on; in the interventional groups, successful measurement was achieved in 100% of cases, whereas the noninterventional groups achieved successful measurement in only 85%. In a prospective study, NT was measured by two to four operators in 200 pregnant women; after an initial measurement, a second one made by the same operator or another operator varied from the initial measurement by less than 0.5 or 0.6 mm, respectively, in 95% of cases. It is suggested that placement of the calipers rather than generation of the appropriate image accounts for a large part of the variation between operators. Subsequent studies continue to report small interoperator differences.

Because NT values are incorporated into a standardized algorithm along with biochemical analyte values, it is critical that these ultrasound measurements be performed and monitored appropriately. To accomplish this, certification and quality review programs have been developed to ensure that accurate and precise NT measurements are obtained. The Fetal Medicine Foundation of London was the first to offer formalized NT training and quality review. In the United States, the Nuchal Translucency Quality Review program was initiated in 2005. Both programs teach the mechanics of obtaining an NT measurement, have an image review process to ensure that the standard technique is used correctly, and perform ongoing epidemiologic monitoring of sonographer and sonologist performance. Two studies have evaluated the techniques used to ensure consistent NT results. Both confirmed that ongoing expert review of images is an inefficient and impractical approach. Epidemiologic monitoring in which an individual operator’s performance is compared with expected standards is preferable.

Combined First-Trimester Nuchal Translucency and Biochemistry Screening

Combining NT with serum analytes improves Down syndrome detection rates in the first trimester. Table 30.5 summarizes the large international experience with first-trimester Down syndrome screening using free β-hCG, PAPP-A, and NT measurements. Overall, for a 5% screen-positive rate, combined first-trimester risk assessment provides a Down syndrome detection rate of approximately 88% (95% CI, 84.0%–89.4%). In women older than age 35, 90% to 92% of trisomy 21 pregnancies can be identified with a 16% to 22% false-positive rate. ^, First-trimester screening can also identify trisomy 18 pregnancies. Greater than 90% of such pregnancies are screen positive when combined biochemical and NT screening is used.

Gestational age–specific variation in the performance of individual analytes can affect screening performance. ^, At all gestational ages between 9 and 12 weeks, NT and PAPP-A are the most efficient markers. In combination, they are most efficient at 11 weeks—a gestational age when free and total hCG are less efficient. In practice, screening is performed between 11 and 13 weeks of gestation.

Ultrasound Markers

Nasal Bone

Assessment of the fetal nasal bone (NB) can be used in the first trimester to predict trisomy 21. This ultrasound marker is based on the flat nasal bridge area, which is a well-described component of the Down syndrome phenotype, and is supported by histopathologic and radiographic studies demonstrating differences in NBs of fetuses with Down syndrome. Stempfle and colleagues found that NB ossification was absent in one-quarter of fetuses with Down syndrome investigated between 15 and 40 weeks’ gestation compared with none of the control fetuses. Similarly, Tuxen and colleagues evaluated fetuses with Down syndrome between 14 and 25 weeks of gestational age by radiographic and pathologic studies and found that the NB was absent in one-third.

Sonek and coworkers published the first large prospective trial of aneuploid risk evaluation using first-trimester ultrasound assessment of the fetal NB. They determined that the fetal NB could be imaged routinely and that its absence was associated with trisomy 21 ( Fig. 30.5 ). The NB was absent in 73% of trisomy 21 fetuses compared with only 0.5% of euploid fetuses. They estimated that if NB assessment were combined with maternal age and NT measurement, 93% of Down syndrome pregnancies would be detected with a false-positive rate of 5%, and 85% would be detected with a false-positive rate of 1%.

A review of the literature by Rosen and D’Alton evaluated 35,312 women having first-trimester ultrasound assessment for NB. In 33,314 cases (94.3%), the NB was successfully imaged. The sensitivity of NB alone for detecting trisomy 21 was 65%, with a false-positive rate of 0.8%. The positive predictive value (PPV) of the screen was 54%, meaning that approximately one in two fetuses with an absent NB had trisomy 21. If the NB was absent, the likelihood that a fetus had trisomy 21 was increased 87-fold. The negative likelihood ratio with a normal NB was 0.35 (95% CI, 0.32–0.39).

As experience with NB has increased, relationships between absent NB, fetal crown-rump length (i.e., gestational age), NT, and ethnicity have been established. Current data demonstrate that in euploid pregnancies, NB absence occurs more frequently with increasing NT. In a series of 5,851 high-risk patients with 333 trisomy 21 fetuses, absence of the NB had a likelihood ratio of 37.1 when the NT was less than the 95th percentile, and this was reduced to 13.4 when the NT was 4 mm or greater. The same study showed that the NB was more likely to be absent at earlier gestational ages. For example, in euploid fetuses with a crown-rump length between 45 and 54 mm, the NB was absent in 4.7% of cases. At a crown-rump length between 75 and 84 mm, the NB was absent in only 1.0% of cases. Prefumo and colleagues found that NB hypoplasia was more common in the euploid fetuses of women of African descent than in either Asian or White populations (odds ratio = 2.3). Cicero and colleagues also found an increased incidence of absent fetal NB in the first trimester in women of Afro-Caribbean and southern Asian descent. The NB was absent in 2.5%, 9.0%, and 5.0% of White, Afro-Caribbean, and southern Asian populations. Likelihood ratios for trisomy 21 with absent NB were 31.3, 8.8, and 14.2 in these three populations.

NB status is independent of serum biochemistry, allowing NB assessment to be combined with measurements of NT and maternal serum markers to increase first-trimester screening performance. In a retrospective case-control study of a high-risk population with a median maternal age of more than 38 years assessed by NT, NB, and biochemistry, it was estimated that 97% of Down syndrome cases would be detected, with a false-positive rate of 5%. For a false-positive rate of 0.5%, the detection rate would be 90.5%. Although these data are promising, performance using this combined screen would be expected to be significantly lower in an unselected population.

NB assessment is technically more difficult to perform than NT, which may limit its usefulness as a component of first-tier screening. To address this, Nicolaides and associates proposed a two-stage screen, reserving NB assessment for patients at intermediate risk after the combined first-trimester screen is complete. In this model, patients evaluated by NT and serum markers with a risk of 1:100 or greater would be offered chorionic villus sampling (CVS), and patients with a risk of less than 1:1000 would be deemed to have such a low risk that no further testing would be offered. Patients with a risk between 1:101 and 1:1000 would have NB evaluation. In initial studies, performance of this two-stage approach was similar to using NB assessment as part of the initial screen. The two-stage approach would have a significant advantage, as only about 15% of pregnancies would require NB evaluation, which could be performed in centers that have developed special expertise in this technique.

Impact of Spontaneous Miscarriages on First-Trimester Screening

A potential disadvantage of earlier screening is that chromosomally abnormal pregnancies that are destined to miscarry will be identified. Of trisomy 21 fetuses living in the first trimester, 69% will be born alive, and 76% of trisomy 21 fetuses identified in the second trimester will be born alive. Using this information, Dunstan and Nix calculated that a detection rate of 80% in the first trimester is approximately equivalent to a second-trimester sensitivity of 75%, suggesting that even when early spontaneous losses of trisomy 21 pregnancies are considered, first-trimester screening is superior to screening presently available in the second trimester.

First-trimester screening would be less desirable if screen-positive pregnancies or pregnancies with enlarged NTs were naturally aborted. In a study of 108 fetuses with trisomy 21 diagnosed in the first trimester because of increased NT, Hyett and colleagues found that six patients elected to continue the pregnancy. In five of the six fetuses, the translucency resolved, and at the second trimester scan the nuchal fold thickness was normal. All six of these trisomy 21 fetuses were born alive. Wapner and colleagues calculated that greater than 80% of screen-positive trisomy 21 pregnancies will be born alive.

Second-Trimester Maternal Serum Screening for Aneuploidy

Serum screening performed in the second trimester (approximately 16 to 18 weeks’ gestation) was the primary tool for Down syndrome risk assessment before the development of first-trimester screening. This approach may be still used for patients who present for prenatal care after the first trimester given that maternal serum alpha fetoprotein (MSAFP) levels reported in women carrying a Down syndrome fetus have median MSAFP values of 0.75 multiples of the unaffected median. Using this deviation to calculate a likelihood ratio, the age-related risk for Down syndrome could be modified. When the standard 1:270 cutoff was used, approximately 25% of Down syndrome pregnancies among women younger than 35 years of age were screen positive.

Elevated hCG (mean, 2.3 MoM) and reduced levels of unconjugated estriol (mean, 0.7 MoM) were subsequently linked to increased trisomy 21 risk and the term “triple screen” entered common usage in the ^, risk calculation with MSAFP. The sensitivity of the triple screen for Down syndrome detection in women younger than 35 years is between 57% and 67% if the false-positive rate is held constant at 5%. Overall, the odds of having an affected pregnancy with a positive screen are approximately 1:33 to 1:62, depending on the age range of the population studied, an improvement over the 1:100 odds when maternal age was the sole screening parameter. Because of the impact of maternal age on the risk analysis, screening women who are 35 years of age or more increases the sensitivity, using the same cutoffs, to approximately 87% but with a false-positive rate of nearly 25%. ^,

Inhibin-A, a protein produced initially by the corpus luteum and later by the placenta, is routinely included in second-trimester Down syndrome screening, resulting in a quad screen. Inhibin-A levels are elevated in Down syndrome pregnancies (1.3 to 2.5 MoM) and do not vary with gestational age in the second trimester. There is, however, a small correlation with hCG levels, making the added sensitivity for Down syndrome detection less robust. The detection rate for Down syndrome using a quad screen consisting of AFP, hCG, unconjugated estriol, and inhibin-A is approximately 75% (screen-positive rate of 5%) in women younger than 35 years. For women older than 35 years, the detection rate is approximately 92%, with a screen-positive rate of 13%.

Other analytes or combinations of analytes have been tested to further increase sensitivity. Hyperglycosylated hCG excreted in maternal urine has been tested as a marker for Down syndrome. One study of nearly 1500 women (1448 control subjects and 39 Down syndrome pregnancies) reported a sensitivity of 96% of affected pregnancies with a 5% false-positive rate and 71% detection with a 1% false-positive rate when a combination of hyperglycosylated hCG, urine β-core hCG fragment, MSAFP, and maternal age was used. However, this detection rate has not been duplicated by others. Hyperglycosylated hCG can also be used as a serum marker for Down syndrome screening and is offered by some laboratories as part of a “penta-test.”

Combining First- and Second-Trimester Screening Tests

Screening performance may be improved by combining analytes assessed in the first and second trimesters. ^, These approaches include the following.

Serum Markers

Numerous significant associations between abnormal maternal serum analytes and adverse obstetric outcomes have been reported and are reviewed in this section and summarized in Table 30.6 . The associations noted in Table 30.6 were derived from studies that included data on adjusted odds ratios. The likelihood of an adverse obstetric outcome is increased as the values of the markers become more extreme. In addition, the association with adverse outcomes strengthens as the number of abnormal markers increases. Although there are many significant associations between abnormal first-trimester and second-trimester maternal serum markers and adverse obstetric outcomes, the sensitivity and PPVs for the individual outcomes are too low for them to be clinically useful as screening tests.

First Trimester

Second Trimester

Second-Trimester Ultrasound Markers of Aneuploidy

As noninvasive screening for trisomy 21, 18, and 13 has a sensitivity of over 90%, initial identification of cases by second-trimester ultrasound has become infrequent. For patients presenting late to prenatal care and patients not desiring earlier testing, ultrasound may still have value in Down syndrome risk assessment.

Second-Trimester Ultrasound Markers of Trisomy 21

The diagnosis of Down syndrome is suspected when certain associated anomalies or physical features are noted on an ultrasound examination. Certain structural anomalies, such as atrioventricular canal defect or duodenal atresia, have a strong association with Down syndrome. Although these anomalies are highly specific for trisomy 21, they have low sensitivity and are not useful for population screening. For example, 40% of pregnancies with duodenal atresia have trisomy 21, but it is seen in only 8% of affected fetuses.

Physical characteristics that are not structural anomalies but that occur more commonly in fetuses with Down syndrome are called soft markers. The ratio of the prevalence of these markers in fetuses with Down syndrome to their prevalence in the normal population will result in a likelihood ratio that can be used to modify the a priori age or screening risk. Available markers and their likelihood ratios are listed in Box 30.2 and Table 30.8 , respectively.

Ultrasound markers modifying the risk for Down syndrome include the following:

• 1.

  Increased nuchal fold (>6 mm) in the second trimester is the most distinctive marker. The fetal head is imaged in the transverse plane used to measure the biparietal diameter. The thalami and the upper portion of the cerebellum should be in the image. The distance between the external surface of the occipital bone and the external surface of the skin is then measured. About 35% of fetuses with Down syndrome, but only 0.7% of normal fetuses, have a nuchal skin fold measurement greater than 5 mm. This ratio yields a likelihood ratio of 50 but includes fetuses with more than one marker. When an increased nuchal fold is an isolated finding, the likelihood ratio is still strong at 20-fold. This high likelihood ratio is obtained because of the rarity of an increased nuchal fold in an unaffected population (i.e., high specificity).

• 2.

  The fetal NB is hypoplastic or absent in up to 60% of Down syndrome pregnancies imaged in the second trimester but in only about 1% to 2% of unaffected pregnancies. Complete absence occurs in about 37% of affected cases, and hypoplasia occurs in about 50%. In normal pregnancies, absence of NB occurs in 0.9% of cases and hypoplasia occurs in 2.4%. NB length can be converted to a likelihood ratio and used for Down syndrome risk assessment. When performed by experienced operators, NB evaluation may be the best single ultrasound marker for second-trimester risk assessment; however, as discussed in the earlier section on first-trimester NB screening, ethnic variation can occur.

• 3.

  Fetuses with Down syndrome may have short proximal extremities (humerus and femur) relative to the expected length for their biparietal diameter in the second trimester. This can be used to identify at-risk pregnancies by calculating a ratio of observed-to-expected femur length based on the biparietal diameter of the fetus. An observed-to-expected ratio of less than 0.91 or a biparietal diameter-to-femur ratio of more than 1.5 has a reported likelihood ratio of 1.5 to 2.7 when present as an isolated finding. A short humerus is more strongly related to Down syndrome, with reported likelihood ratios of 2.5 to 7.5. Bahado-Singh and coworkers combined humerus length with nuchal skin fold to estimate Down syndrome risk and calculated the likelihood ratios for various measurements to adjust estimated Down syndrome risk for each patient.

• 4.

  Echogenic intracardiac foci is observed in up to 5% of normal pregnancies and in approximately 13% to 18% of Down syndrome pregnancies. The likelihood ratio for Down syndrome when an echogenic focus is present as an isolated marker is 1.8 to 2.8 but may be lower in an Asian population, where the frequency in unaffected pregnancies may be higher. The risk does not seem to vary if the focus is in the right or left ventricle or if it is unilateral or bilateral.

• 5.

  Increased echogenicity of the fetal bowel, when brighter than the surrounding bone, has a likelihood ratio for Down syndrome of 5.5 to 6.7. This finding can also be seen with fetal cystic fibrosis (CF), congenital cytomegalovirus infection, swallowed bloody amniotic fluid, and severe FGR or placental insufficiency. Therefore, if amniocentesis is performed for this finding, testing for the other potential etiologies should be considered.

• 6.

  Mild fetal pyelectasis (renal pelvis anterior-posterior diameter >4 mm) has been suggested as a marker for Down syndrome. As an isolated marker, the likelihood ratio is 1.5 to 1.9 (see Table 30.8 ). However, Snijders and coworkers found that mild renal pyelectasis is not significantly more frequent in Down syndrome pregnancies than in normal pregnancies.

• 7.

  Other markers described include a hypoplastic fifth middle phalanx of the hand, short ears, a sandal gap between the first and second toes, an abnormal iliac wing angle, an altered foot-to-femur ratio, an altered frontomaxillary angle, ^, and increased prenasal thickness. ^, These markers are inconsistently used because of the time and expertise required to obtain them.

Use of Second-Trimester Ultrasound in Down Syndrome Screening

As biochemical and DNA screening tests have improved, the value of ultrasound soft markers in Down syndrome screening has become less significant. Soft markers should not independently be used to calculate Down syndrome risk but should be part of a total risk analysis. Many centers do not inform patients of a single soft marker unless the overall risk for trisomy 21 exceeds that at which they routinely offer invasive testing. This cutoff is frequently between 1: 150 and 1:300. With the use of cell-free DNA screening, which has a positive likelihood ratio of 1720 and a negative likelihood ratio of 0.006, it is recommended that ultrasound should not be used to modify a positive cell-free DNA result.

Second-Trimester Ultrasound Screening for Other Chromosomal Abnormalities

Fetal aneuploidy other than Down syndrome can be suspected on the basis of ultrasound findings ( Table 30.9 ). Choroid plexus cysts occur in 1% of fetuses between 16 and 24 weeks’ gestation and have been associated with trisomy 18. Among fetuses with trisomy 18, 30% to 35% have choroid plexus cysts. Among fetuses with a choroid plexus cyst, about 3% have trisomy 18, and most (65% to 90%) of these have other ultrasound findings ( Table 30.10 ). Although an isolated choroid plexus cyst was estimated to yield a probability of trisomy 18 of 1:150, many series contain a high proportion of older women, which would overstate the risk. Snijders and coworkers calculated that an isolated choroid plexus cyst has a likelihood ratio of 1.5 for trisomy 18, and thus it can be used to calculate an individual’s risk for trisomy 18 when combined with other screening modalities. The size, location, or persistence of the cyst does not alter this risk. In women who screen negative for trisomy 18 (either first-trimester or second-trimester screening or cell-free DNA) and in whom no other fetal structural abnormalities are visualized, the finding of an isolated choroid plexus cyst does not require additional genetic testing.

Table 30.9 displays the magnitude of the associations between various ultrasound findings and aneuploid conditions as estimated from a referral population. The rates noted may overestimate the strength of the association when such findings are noted on a screening examination.

Aneuploidy Screening Using Cell-Free Fetal DNA in the Maternal Circulation

Cell-Free DNA Methods Used to Screen for Fetal Aneuploidy

Three approaches to cell-free DNA screening have been validated by different laboratories. The first technique reported was massively parallel sequencing, in which all of the DNA that is extracted from a maternal sample is sequenced. In 2008 two groups identified Down syndrome pregnancies ^, by analyzing the cell-free fetal DNA in the maternal plasma using massively parallel shotgun sequencing. The first 36 bases of all circulating DNA fragments (both maternal and fetal) were sequenced and identified to determine their specific chromosomal origin ( Fig. 30.6A ). If the fetus has a third chromosome 21, the percentage of chromosome 21 fragments compared with disomic chromosomes will be slightly higher than expected. In a pregnancy in which 10% of the cell-free DNA is fetal, a woman carrying a fetus with trisomy 21 should have approximately 1.05 times more DNA from chromosome 21 fragments than a woman carrying a fetus with disomy 21. Prediction of trisomy 21 is based on the ability to distinguish this difference by sequencing millions of fragments, identifying their chromosome origin, and then quantifying their relative proportion ( Fig. 30.6B ).

Targeted massive parallel sequencing, also known as chromosome-specific sequencing, selects the genomic regions of interest, such as nonpolymorphic loci on chromosomes 21, 18, and 13, and then reads and counts only those. The ability to sequence specific regions of the genome in cell-free DNA allows for a focused analysis of clinically important chromosomes such as 13, 18, 21, X, and Y. This strategy significantly reduces the total number of reads analyzed with a concomitant improvement in efficiency and a 10-fold reduction in overall costs. Also, the reduced read requirement enables the use of small bench-top sequencers that affords both financial and practical benefits to the laboratory.

Microarray-based technology is used by some labs to quantify the relative contributions from specific chromosomes and does not require next-generation sequencing. In this approach DNA fragments from the chromosomes of interest are initially selected and hybridized to a specially designed array and analyzed by the degree of fluorescence. Juneau and associates showed that microarray-based targeted analysis resulted in high sensitivity and specificity comparable to next-generation sequencing. The proposed advantages of microarray-based noninvasive prenatal screening are decreased variability that will allow testing samples with lower fetal fraction and improved turnaround time and reduced cost owing to individual hybridization instead of sample multiplexing. ^,

A third approach focuses on evaluating single nucleotide polymorphisms. By measuring these polymorphic loci, the number and identity of each allele from each sequence read are determined. The allelic information from the mother (and from the father, if available) is used to model a set of hypotheses corresponding to different genetic inheritance patterns and crossover locations for every possible copy number count. Bayesian statistics then assign a probability to each hypothesis (whether the findings are most consistent with monosomy, disomy, or trisomy), and a maximum likelihood analysis is performed to calculate that the probability of that hypothesis is correct. Because the single nucleotide polymorphism–based method is dependent on analysis of maternal genotype information, it cannot be used in pregnancies resulting from egg donors. The SNP-based approach can identify triploidy. This approach may prove to be useful in the detection of vanishing twins and determination of zygosity.

Cell-Free DNA Screening Performance Characteristics

Cell-free DNA screening has the highest sensitivity and the lowest false-positive rate of all the Down syndrome screening tests currently available. The detection rates based on samples for which a result is returned are approximately 99.7% for trisomy 21, 97.9% for trisomy 18, and 99.0% for trisomy 13 at a combined false-positive rate of 0.13% ( Table 30.11 ). Although Down syndrome screening using cell-free fetal DNA is better than biochemistry, with or without NT assessment, it is still not diagnostic. All screen-positive cases should be confirmed by invasive testing, and false-negative results do occur.

Cell-free DNA screening was initially studied in high-risk populations consisting of women at an increased risk for having a fetus with trisomy 21 based on advanced maternal age (≥35 years old at delivery), a positive conventional aneuploidy screen, a history of a prior pregnancy with a trisomy, a parental balanced robertsonian translocation with increased risk of fetal trisomy 13 or trisomy 21, or an ultrasound abnormality associated with fetal aneuploidy. ^, Subsequent studies demonstrated that the detection rates and false-positive rates are similar in low-risk and high-risk populations. ^{, ,} The detection rates appear to be similar using all three techniques.

Although the sensitivity and false-positive rates associated with cell-free DNA screening are similar in low-risk and high-risk populations, the PPV for trisomy 21, trisomy 13, and trisomy 18 is higher in older women owing to the increased prevalence of these conditions. It is helpful to provide a PPV when counseling women who have positive cell-free DNA test results. Although not all laboratory reports provide this information, patient-specific PPVs may be obtained using an online PPV calculator ( https://www.perinatalquality.org/Vendors/NSGC/NIPT/ ). The PPV associated with cell-free DNA screening is considerably higher in women of all ages compared with standard aneuploidy screening with biochemistry with or without NT. In a study of 1914 women with a mean age of 29.6 years who had standard aneuploidy screening during the first or second trimester with or without NT and cell-free DNA screening, the PPVs for standard screening versus cell-free DNA testing were 4.2% versus 45.5% for trisomy 21 and 8.3% versus 40.0% for trisomy 18.