Prenatal screening for chromosomal abnormalities using nuchal translucency (NT) assessment provides a prospective mother with the opportunity to obtain an accurate estimate of her individual and actual risk for having a fetus affected by trisomy 21. By being aware of her adjusted risk, she can exercise autonomy in order to make the most appropriate decisions regarding further diagnostic testing1, 2.
It is clear that in the field of prenatal screening, to be assigned a numerical risk rather than a positive or negative result is more meaningful for a pregnant woman and her doctor3. The policy to offer more widely such screening has had the further benefit of decreasing the number of prenatal invasive procedures offered on the basis of maternal age, following assignment of a low risk on ultrasound examination4. At the core of this ultrasound-based risk assessment is measurement of fetal NT, and the quantification of its thickness compared with the reference distribution of karyotypically normal fetuses. Rigorous trials, experiences gathered from various centers, careful training, assessment of competence and compulsory periodical audit of operators have allowed the widespread adoption of NT screening with effective results5, 6.
To date, however, the actual cause of the enlargement of the NT, that could be the key to distinguishing between karyotypically abnormal and normal fetuses, has not been clarified and remains a challenging enigma.
There has been continuing uncertainty surrounding our knowledge of the causal mechanism of increased NT for the last 10 years and no breakthrough has come from the most recent literature7–9. Despite this, the NT test has become widespread. The NT test is effective in terms of accuracy for screening; it works well when considered alongside some other additional markers in a ‘contingent screening’ policy, and some clever refinements to the risk calculations have been proposed (e.g. when a nuchal cord is present a mean between the NT thickness above and below the cord is proposed10).
Enlarged NT appears in fetuses affected by pathological conditions such as some chromosomal abnormalities, congenital heart defects, major extracardiac malformations and genetic syndromes11, 12. Structural heart defects leading to temporary cardiac failure, delay in the development of the heart with reduced compliance, venous congestion of the head and neck, altered composition of the extracellular matrix and delayed or abnormal lymphatic system development are among the putative causes of increased NT. A compelling hypothesis is that a common denominator of these conditions leading to the enlargement of NT may be the phenomenon of heterochronia, possibly affecting trisomy 21 fetuses by a generalized delay in developmental milestones.
Moreover, enlarged NT happens unexpectedly in some completely normal fetuses, and in cases of monochorionic diamniotic (and rarely monoamniotic) twins. In singleton pregnancies, enlarged NT (greater than the 95th centile) is encountered in about 5% of cases and in dichorionic pregnancies the frequency is about the same. However, in fetuses of monochorionic diamniotic pregnancies the frequency of an enlarged NT is significantly higher, about 15%, without epidemiological demonstration of a higher frequency of chromosomal abnormalities13, 14.
Enlarged NT in a monochorionic pregnancy
In a monochorionic pregnancy, in which fetuses share the same placenta, enlarged NT may affect one or both fetuses. First-trimester imbalance in the placental circulation between the two fetal compartments may be one of the causes of the NT enlargement, and this pathophysiological mechanism is unique to monochorionic pregnancy. When enlarged NT is found in both fetuses, the presence of an abnormal karyotype may be the underlying pathological condition, one which will affect both monozygotic twins in the same way. NT enlargement may also be attributable to the presence of fetal structural defects, either concordant or discordant, as is the case in singletons.
A recently emerging issue affecting first-trimester pregnancy assessment is the rare case of discordant NT in a monochorionic twin pair. Phenotypic and genotypic discordance has already been recognized in ‘non-identical’ monozygotic twin pairs. The global frequency of malformations seems to be higher in monozygotic twins than it is in singletons at birth, but some structural abnormalities are specific to monozygotic twins, resulting from the same mechanism that causes embryo splitting or from an unequal allocation of blastomeres. Monozygotic twins are usually discordant for major malformations such as spina bifida, anencephaly, holoprosencephaly, omphalocele and congenital heart disease15. Cardiac defects in monozygotics are most frequently of the ‘flow’ type, for example, ventricular septal defects and pulmonary stenosis16. The association between heart defects and enlarged NT is evident, and therefore in a monochorionic twin pair, the finding of a discordant NT may be a sign of cardiac disease. The positive predictive value of enlarged NT for fetal anomalies in monochorionic pregnancies, however, is affected by the greater frequency of enlarged NT in these pregnancies. A significant number of monozygotic twin pairs may differ in their genotype. Discrepant kayotype (heterokaryotypia) may result from a mitotic error arising either before twinning, resulting in a mosaic, or after splitting, resulting in the chromosomal abnormality only in one of the fetuses. Skewed X-chromosome inactivation, postzygotic dominant gene mutation, postzygotic recessive gene mutation, repeat expansion disorders and different imprinting processes between twins are some of the underlying causes of discordant genotype in monozygotic twins that result in phenotypic discordance17. Therefore, although the frequency of heterokaryotypia is very low, this possibility should be kept in mind when a discordant NT thickness is found18.
Regarding NT measurement, two methods have been proposed to quantify a discordant NT in a monochorionic pair. One option is to consider the NT enlarged if it is greater than the 95th centile of the normal range for CRL, derived from singleton fetuses. However, the validity of using a reference range for singletons is debatable since NT is greater than the 95th centile in about 15% of monochorionic pregnancies (compared with the 5% expected in singletons). There is evidence of fetuses of monochorionic pregnancies both affected by trisomy 21 being discordant for NT calculated by this method19.
Another method is to highlight a discordant NT by presenting the difference between the two NTs as a percentage. This method does not consider the relationship between NT and the length of the fetus (crown–rump length, CRL), which is implicit in the method of the 95th centile, but could be effective in detecting slight differences in NT between twins. A difference of greater than 20% between the two NTs of a monochorionic twin pair was found in about 25% of cases, and was associated with complications (intrauterine death or twin-to-twin transfusion syndrome, TTTS) in 30% of cases20. Furthermore, discordance in NT between fetuses was more frequent in cases that were later complicated by TTTS or had an unfavorable outcome20. A correlation between intertwin discordance in NT and adverse outcome of pregnancy was confirmed by other experiences, but a specific correlation between discordant NT and the development of TTTS was not21.
An indirect assessment of the fetal circulation in the first trimester may be provided by Doppler monitoring of flow patterns in the ductus venosus (DV). Singleton fetuses with enlarged NT show some alterations in DV flow in about 40% of cases, compared with in only 1% of cases when the NT is normal22. Reversal of flow in the DV during atrial contraction (a-wave) in at least one fetus has been found to be more frequent (18% of cases) in monochorionic pregnancies overall than in dichorionic pregnancies (8%)23. It has also been associated with a higher risk for development of TTTS, miscarriage and aneuploidy23, 24. Intracardiac flow evaluation, which has been proposed for the assessment of monochorionic pregnancy as early as in the second trimester, could be a promising tool for the first trimester too.
It should be borne in mind, however, that when a discordant NT is found in monochorionic pregnancies, in most cases there will not be complications.
Problems with NT screening in monochorionic pregnancy
While NT measurement is technically feasible in monochorionic pregnancies, its inclusion in algorithms for calculation of risk for aneuploidy may have some implications. It is well accepted that the background risk for trisomy 21 in a monozygotic twin pregnancy is the same as the maternal age-related risk for a singleton pregnancy, but data are not available for the distribution of NTs in trisomy 21 and karyotypically normal monochorionic fetuses to calculate reliable likelihood ratios. The influence of monochorionic placentation on NT distribution, with the higher frequency of enlarged NTs, should not be ignored.
Another problem is the discordant NT. Many questions arise, such as which of the two NTs should be considered for risk calculations if they are discordant in size? Should the rare phenomenon of heterokaryotypia be ignored? If the smaller NT is considered, this would lead to an underestimation of the risk, while the hazard of overestimation could be generated if the larger NT is considered. The impartial decision to make an arithmetical mean between the two NTs, and to calculate a risk for a hypothetical mean CRL19, while apparently working better than other methods, is difficult to accept.
Agreement between predicted risks and observed prevalence of trisomy 21 has not been assessed as it has in singletons25 in order to validate the use of such screening in monochorionic pregnancies. Moreover, differences among CRLs may affect biochemistry and result in difficulty for the estimation of risk26. The role of NT screening for the choice between amniocentesis and chorionic villus sampling when the NT is discordant is limited, due to the necessity of sampling each fetus (because of the risk of heterokaryotypia).
Other markers of trisomy 21, such as absent nasal bone, which, unlike NT, is not obviously influenced by placental flow imbalance, may be of value for monochorionic risk assessment27.
Information derived from the 11–14-week scan in the monochorionic pregnancy is invaluable as far as prognosis and outcome are concerned. It is possible, and we may be as yet unaware, that all the important cards in a monochorionic twin pregnancy are played in the first trimester. However, issues arising in monochorionic twins are specific and there are concerns about making generalizations for monochorionic twin pregnancies on the basis of singleton and dichorionic pregnancies regarding the use of NT screening for chromosomal abnormalities.