Maternal plasma cell-free DNA screening

Integration into clinical practice

Howard Cuckle and Peter Benn

Introduction

Maternal plasma cell-free (cf) DNA testing is more effective in screening for Down, Edwards, and Patau syndromes (trisomies 21, 18, and 13) compared to conventional methods that use multiple maternal serum markers and ultrasound examination. It can also be applied to sex chromosome abnormalities (SCAs), and to certain microdeletion syndromes. Nonetheless, there are practical issues that have limited widespread application: economics, uninterpretable tests (“no-calls”), patient choice, and implications for other services. These restraints are discussed.

Estimated performance of aneuploidy screening

Study type

The most recent meta-analysis includes data from 47 studies (1). There are two approaches to cfDNA evaluation. Retrospective in high-risk women with complete outcome information known; plasma samples are mostly drawn prior to invasive prenatal diagnosis (IPD). Alternatively, prospective on samples drawn in a conventional screening program.

There is no evidence that performance is correlated with maternal age or ultrasound markers, and biochemical markers are only weakly correlated with fetal fraction (FF, the relative proportion of cfDNA that is fetal in origin). Retrospective studies can therefore be assumed to be substantially unbiased. In contrast, prospective studies may overestimate the detection rate (DR) because of incomplete follow-up and “viability” bias resulting from the inclusion of detected nonviable cases. Not all those who test positive accept IPD, and some are lost to follow-up, potentially underestimating the false-positive rate (FPR) if such cases are excluded from the audit. For trisomies 13 and 18, FF is lower, and there is an excess of cases in no-call results (see later), which could bias both prospective and retrospective studies.

The first studies to be carried out were retrospective, and it might be argued that subsequent technical changes improved the performance. Studies included more late pregnancy samples when performance was better because of a larger FF. On balance, we choose to include only the retrospective studies, possibly providing a conservative estimate of performance (Table 35.1). Down syndrome DR was 99.3% (95% confidence interval 98.6%–99.7%) and FPR 0.11% (0.07%–0.19%). Edwards and Patau syndrome DRs are comparable to the “incidental” detection rates obtained through the conventional combined test using only a cut-off for Down syndrome (2). For SCAs, DRs with cfDNA are much higher than those for the combined test. cfDNA screening for all aneuploidies together will have an FPR approaching 0.8%, and this is substantially lower than the 5% rate for the combined test. The positive predictive value (PPV) for Down syndrome at birth is about one in two, which is much higher than 1 in 50 for the combined test (2).

The studies used in Table 35.1 largely exclude mosaic cases. This is particularly a problem for Turner syndrome since mosaicism is common in viable cases. Consequently, it is likely that a large proportion of the cases studied were pregnancies destined to spontaneously abort, and do not fully reflect the more clinically important surviving cases (3). Nevertheless, there is evidence that about half of the mosaic SCA cases are detected by cfDNA, representing a large increase in detection. Many of the mosaic cases may have a mild phenotype.

In conventional screening, twins discordant for aneuploidy have biochemical marker levels intermediate between concordant and unaffected twins, and therefore a reduced DR. A similar effect is seen in cfDNA testing, although it is offset to some extent by increased FF because of greater placental volume. There are 11 published studies of discordant twins (six cited in [1, 4–8]) with DR: Down syndrome 96.7% (57/59), Edwards syndrome 88.9% (8/9), and Patau syndrome 100% (2/2). The FPR was 0.05% (1/1952). This is better performance than the conventional combined test; in one study of dichorionic twins, the Down syndrome DR was 90% and FPR 5.9% (9).

cfDNA screening strategies for Down syndrome

Three types of strategies have been discussed where cfDNA testing is offered: “secondary” only to women with positive conventional screening results, “primary” to all women, or “contingent” to 10%–30% of women with the highest conventional test risks. Variants of contingent screening exclude those with very high risk, use additional conventional test markers, or exclude women of advanced maternal age.

Table 35.2 shows the model predicted performance of each strategy based on the conventional combined test. Secondary cfDNA will substantially reduce the number of IPDs compared with conventional screening resulting in fewer fetal losses and a small reduction in DR. Primary cfDNA maximizes DR while ensuring that relatively few have IPD. Contingent cfDNA vastly reduces the number of women requiring cfDNA while maintaining a DR much closer to primary cfDNA. Excluding those with very high risks has a slightly higher DR and, despite a much higher FPR, is considered worthwhile since about two-thirds of affected pregnancies are in this group and delay waiting for cfDNA results is avoided. All strategies have a much higher PPV than conventional tests (2).

Stay updated, free articles. Join our Telegram channel

Join