What are cfDNA and cfDNA aneuploidy screenings?

CfDNA consists of small (<200 base pairs) fragments of DNA that are free floating in the plasma. During pregnancy, after 10 weeks of gestation, approximately 10-15% of the total cfDNA in the maternal plasma is of placental origin (ie, derived from trophoblast) and can be used therefore to test for fetal disorders. Cell-free DNA screening (also referred to as cfDNA testing, noninvasive prenatal testing, and noninvasive prenatal screening) is a test that uses next-generation sequencing of cfDNA in maternal plasma combined with bioinformatic algorithms to determine the probability of certain fetal chromosomal conditions in pregnancy.

Although such testing has been demonstrated to be possible for a variety of genetic conditions (which include blood type, autosomal single gene disorders, and determination of fetal sex to assess risk for X-linked diseases), the large majority of clinical tests are done to test for fetal chromosomal disorders. Initially focused only on Down syndrome, the laboratories that perform the testing all subsequently have added assessment for trisomies 13 and 18 and the sex chromosomes. In addition, some laboratories provide testing for additional trisomies and some microdeletion syndromes. It is likely that the range of available conditions that can be tested for with the use of cfDNA screening will continue to increase over time.

How is cfDNA analysis performed?

CfDNA screening was made possible by 2 developments: advances in next-generation sequencing after completion of the Human Genome Project in 2001 and the discovery that cfDNA of placental origin is present in the maternal circulation and can be analyzed from a plasma sample. Different laboratories use somewhat different platforms, but the common theme of next-generation sequencing is increased automation, which allows faster and cheaper sequencing than earlier methods.

Some cfDNA tests use an approach known as massively parallel shotgun sequencing , in which all of the cfDNA that is extracted from a maternal sample is sequenced. Each DNA fragment is localized to its chromosome of origin, and the number of fragments that arises from each chromosome is counted. Additional material that arises from a given chromosome increases the chance that the fetus carries an extra copy of that chromosome. Another method also counts chromosome fragments but includes those from a more limited subset of the total genome; this is more efficient than massively parallel shotgun sequencing and thereby allows for somewhat lower costs. A third approach involves the analysis of many thousands of single nucleotide polymorphisms to determine the genotype and relative copy number of each chromosome of interest. Each platform has limitations and specific maternal or fetal criteria that must be met for optimal test performance.

In addition to differences in how the sequencing and laboratory analyses are performed by the different platforms, there are also differences in the bioinformatics analysis and interpretation. This interpretation and the presentation of results are important and complex parts of the comparison of the overall test characteristics.

How is cfDNA analysis performed?

CfDNA screening was made possible by 2 developments: advances in next-generation sequencing after completion of the Human Genome Project in 2001 and the discovery that cfDNA of placental origin is present in the maternal circulation and can be analyzed from a plasma sample. Different laboratories use somewhat different platforms, but the common theme of next-generation sequencing is increased automation, which allows faster and cheaper sequencing than earlier methods.

Some cfDNA tests use an approach known as massively parallel shotgun sequencing , in which all of the cfDNA that is extracted from a maternal sample is sequenced. Each DNA fragment is localized to its chromosome of origin, and the number of fragments that arises from each chromosome is counted. Additional material that arises from a given chromosome increases the chance that the fetus carries an extra copy of that chromosome. Another method also counts chromosome fragments but includes those from a more limited subset of the total genome; this is more efficient than massively parallel shotgun sequencing and thereby allows for somewhat lower costs. A third approach involves the analysis of many thousands of single nucleotide polymorphisms to determine the genotype and relative copy number of each chromosome of interest. Each platform has limitations and specific maternal or fetal criteria that must be met for optimal test performance.

In addition to differences in how the sequencing and laboratory analyses are performed by the different platforms, there are also differences in the bioinformatics analysis and interpretation. This interpretation and the presentation of results are important and complex parts of the comparison of the overall test characteristics.

How accurate is cfDNA aneuploidy screening?

CfDNA screens for trisomies 13, 18, and 21 and sex chromosomal abnormalities; the accuracy of screening for each of these conditions varies somewhat by condition and platform used. The ability of cfDNA to identify the presence (or absence) of a chromosomal aneuploidy depends on a number of factors, including, the amount of fetal DNA that is present, the a priori chance that a chromosome abnormality is present (that is, the woman’s risk based on maternal age or results of other screening), and other factors such as the presence of a multifetal gestation or a nonviable second embryo/fetus or the presence of placental mosaicism.

Some laboratories report the probability of aneuploidy; most commonly, this is stated to be >99% in patients who are at increased risk and <1/10,000 in patients who are at low risk. Such results suggest a degree of certainty that is near diagnostic; however, this is a population statistic and applies only to the entire population of women who were screened and not to an individual’s result. It is important for providers to recognize that a positive result for any of these aneuploidies confers a chance that the fetus is affected, which is usually far <99%, particularly in lower risk patients.

To determine how likely it is that a positive result indicates an affected fetus, the positive predictive value (PPV) should be assessed. PPV is the proportion of positive results that are true positives and is dependent not only on the sensitivity and specificity of the test but also is highly dependent on the prevalence of the condition. When testing for rare conditions (such as aneuploidy in younger women), the PPV is much lower than when testing for more common conditions (such as trisomy 21 in older women). Therefore, more false-positive results are expected in women who are at low risk or when screening is done for very rare conditions. The PPV for trisomy 21 has been reported as varying from 45% in low-risk patients to ≥96% in the highest risk patients. In one study of diagnostic testing after abnormal cfDNA screens, aneuploidy was confirmed in 93% of trisomy 21 cases, in 64% of trisomy 18 cases, in 44% of trisomy 13 cases, and in 38% of sex chromosomal abnormalities.

How should a cfDNA “positive test” be interpreted?

Laboratory reports vary in how they report findings that are “positive” based on cfDNA analysis. Aneuploidy risk is generally reported as “positive” or “detected” or as a probability of >99% in patients who are at increased risk. After a positive test, patients should be referred for posttest counseling to a maternal-fetal medicine subspecialist, geneticist, and/or genetic counselor. Such counseling should include a discussion of the predictive value of cfDNA as a screening test (including the possibility that the result is a false positive) and the offer of diagnostic testing (chorionic villous sampling or amniocentesis) for confirmation, particularly in women who are considering pregnancy termination.

In part because of the manner in which cfDNA results are presented and how the tests are being marketed, there is some confusion by both providers and patients regarding the possibility of false-positive results. Often, it is assumed mistakenly that this testing is diagnostic. In a study that described outcomes of cfDNA testing from commercial testing of >30,000 women, approximately 6% of women with a positive cfDNA test result proceeded to pregnancy termination without confirmatory diagnostic testing. It is important that obstetric providers understand and appropriately interpret the results of cfDNA screening and accurately convey this information to their patients as part of pretest counseling ( Table 1 ).

What are the implications of a failed cfDNA test result and how should such cases be treated?

Because of the fact that screening requires a minimum amount of cfDNA, often referred to as the “fetal fraction (FF),” there is a risk for test failure because of low FF. In addition, some tests do not provide a result because of difficulties in interpretation of sequencing data. Although FF is relatively constant from 10-22 weeks of gestation, it is lower at <10 weeks of gestation and less likely to provide a result. Overall, the chance of test failure is reported at 0.9-8.1% and varies in part by whether the laboratory measures FF and requires a minimum concentration. Low FF and failed results have been associated with fetal aneuploidy. In one recent study, 8% of the patients overall had failed testing; this increased to 16% in cases with fetal aneuploidy. In this study, the odds ratio for aneuploidy was 9.2 in cases with failed tests.

Given this association, women with failed cfDNA screens should be counseled that they are at increased risk for trisomy, particularly trisomies 13 and 18, and triploidy. It is therefore appropriate to offer the option of diagnostic testing in these cases, given the increased risk. A repeat cfDNA screen will be successful in 50-80% of cases. Whether the patient chooses to attempt cfDNA screening again may depend in part on gestational age; a patient at a more advanced gestation may not wish to delay obtaining definitive information, given the increased risk.

What are other limitations of cfDNA aneuploidy screening?

Low FF has been associated with maternal obesity; in 1 study, cfDNA aneuploidy screening failed to provide a result in 20% of women >250 lb and 50% of women >350 lb. Therefore, in obese or morbidly obese women, cfDNA aneuploidy screening may not be the best screening option.

At present, there are limited data on the use of cfDNA aneuploidy screening in multifetal gestations; most published studies have included a small number of aneuploid fetuses. In these series, it has been noted that the failure rate is higher and the detection rate may be lower, although the number of cases is very small. With regard to a nonviable cotwin, it is recognized that a high percentage of fetal losses are aneuploid, which is also true with a dead twin. The presence of a second gestational sac has been associated with false-positive cfDNA results; therefore, this test is not a good option for women with a “vanishing twin” or empty second sac. At this time, the data are too limited to recommend routine cfDNA aneuploidy screening in women with multifetal gestations.

What evaluation is appropriate for women with a false-positive cfDNA aneuploidy screening result?

Given that the cfDNA present in maternal plasma is a mixture of maternal and placental DNA, a number of biologic phenomena can cause a false-positive, or discordant, cfDNA result. Many cases are thought to result from confined placental mosaicism or a cotwin death or a vanishing twin. Cases of false-positive results for sex chromosomal aneuploidy have been reported in which pregnant women were found to have a sex chromosomal abnormality themselves, often in mosaic form. This has led to the discussion of the possible benefit of karyotyping of women who have a false-positive cfDNA screening result to rule out a mosaic chromosomal abnormality in the mother.

A few cases of maternal malignancies with chromosomal abnormalities within the tumor have been reported in patients with false-positive cfDNA screening results. These reports have raised the question about the benefit of further evaluation for maternal malignancy in women with false-positive results. Although such cases of malignancy are of interest in considering the underlying biologic evidence of cfDNA, at this time the clinical utility and yield of an extensive work up (eg, multiple diagnostic imaging studies to search for undiagnosed maternal malignancy) are unknown.