Ultrasound features of dichorionic versus monochorionic twins.
The most recent and largest study to date on the accuracy of amnionicity and chorionicity determination using lambda and T signs and number of placental masses of 613 pregnancies, reported a sensitivity of 100% and specificity of 99.8% when ultrasound was undertaken between 11 and 14 weeks gestation[4]. In the second trimester, the lambda sign gradually disappears limiting its use, and therefore if amnionicity and chorionicity have not been determined in the first trimester, fetal sex determination may help but is limited as a good number of dichorionic twins may be same sex. Although measuring intertwin membrane thickness has been described (membrane thickness of >1.5 mm used to diagnose dichorionicity), it is not as reliable as using lambda and T sign, technically more challenging and affected by inter and intraobserver variability in the second and third trimesters[5].
Monoamniotic twins can be diagnosed erroneously in the second trimester when there is severe TTTS. Therefore, if not diagnosed in the first trimester and monoamniocity suspected, the sonographer needs to look carefully to exclude a collapsed amniotic membrane around a donor fetus with an absent bladder, and it is also worthwhile assessing whether the cord insertions are close together or not.
Importance of labeling fetuses
It is known that labeling twins by assigning numbers (twin 1 and twin 2) and allocating the label “twin 1” to the fetus closest to the cervix in early pregnancy, does not accurately determine which will be the leading twin as pregnancy progresses or indeed birth order. This is particularly true for laterally orientated twins, i.e., left and right twins where 8.5% change presenting order between the first and last scans, and 20.3% delivered by cesarean versus 5.9% delivered vaginally change birth order (i.e., the twin labelled “twin 2” delivers first)[6]. Correct labeling according to orientation in relation to the mother as lateral maternal left and maternal right or vertical upper and lower, is better than assigning a fetus number as it enables consistency with longitudinal biometric assessment, accuracy when interpreting screening results and undertaking invasive diagnostic tests where necessary, and avoids misconception about birth order ensuring the parents and pediatric team are aware of the possibility of peripartum switch (i.e., possible change in birth order). This is particularly important if one fetus has an abnormality that is not outwardly obvious, e.g., cardiac abnormality.
Screening for aneuploidy
Down’s syndrome and other aneuploidy screening in multiple pregnancy is complicated because i) there is a higher risk of aneuploidy, ii) the sensitivity (i.e., detection rate (DR)) of screening tests is probably lower compared with singleton screening, iii) the false-positive rate (FPR) is higher, iv) the likelihood of being offered invasive diagnostic testing is higher as is the risk of complications of invasive diagnostic testing, and v) in the event of an affected fetus, the options are complex including selective reduction and risks to the surviving normal fetus or fetuses. The published literature on first-trimester screening in multiple pregnancy is of poor quality and there is limited published literature on second-trimester screening in multiple pregnancy.
The UK NICE guideline reviewed nine studies that evaluated first-trimester screening[3], three evaluated combined screening – nuchal translucency (NT), maternal age, other maternal factors, and serum screening using beta-human chorionic gonadotrophin and pregnancy-associated plasma protein-A, three evaluated NT and maternal age and six evaluated NT alone. Two of these studies included triplets but did not report separate data for twins versus triplets. One only evaluated monochorionic twins. For twins, all methods have high sensitivities but combined screening overall performs best and should be offered.
For dichorionic twin pregnancies, each fetus has an individual risk of Down’s syndrome or other aneuploidy and therefore the risk should be calculated for each fetus separately, i.e., fetus specific risk. It should be noted that about 10% of dichorionic twins are monozygotic but clinically it is not possible to determine if this is the case.
For monohorionic twins each fetus has the same risk of being affected with Down’s syndrome or other aneuploidy (unless there is heterokaryotypia, which is extremely rare) and the overall risk is the same as in a singleton pregnancy. Therefore, the NT measurement should be averaged and used to calculate a pregnancy specific risk.
For triplets or higher order multiple pregnancies, there are no normograms for serum screening, and therefore NT and maternal age is the only available screening (Table 24.1).
| Type of multiple pregnancy | First Trimester | Second Trimester |
|---|---|---|
| Dichorionic twins | Combined NT test Calculate fetus specific risk | Serum screening Calculate pregnancy specific risk |
| Monochorionic twins | Combine NT test Calculate pregnancy specific risk by using average NT | Serum screening Calculate pregnancy specific risk |
| Triplets or higher order | NT and maternal age alone | No available test |
| Vanishing twin | Combined NT test but note not as accurate | Serum screening but note not as accurate |
NT, nuchal translucency.
It is also important to note that if there has been a ‘vanishing twin’, i.e., the pregnancy started as multiple but then naturally reduced to singleton in the first trimester, serum markers may be higher than in singleton pregnancy and the results of screening using serum markers not as accurate. Women need to be fully informed about the higher risks with screening and need to be aware that decision making and options are complex if the screening test is positive. This requires experienced professionals providing information and counseling before the screening test, and indeed afterwards if the result is positive. Furthermore, if the test is positive and the woman opts for invasive diagnostic testing, this should be performed by a specialist who has the expertise to subsequently perform selective termination of pregnancy if required[7].
If first-trimester screening is not possible (e.g., the woman presents too late), there is limited published evidence to enable recommendations but the NICE guideline[3] recommends offering second-trimester serum screening for twins. With the triple test, the DR is 63% but FPR high at 10.8% (which may be improved by using the quadruple test)[8]. For triplet or higher order pregnancy, there are no second-trimester screening options.
Recently, noninvasive prenatal testing has become commercially available for screening for trisomies 21, 18 and 13; monosomy X and sex determination with high levels of accuracy and low false-positive and false-negative rates. Many women are opting for this test and it is predicted that other forms of screening will become obsolete in the not too distant future in singleton pregnancy. In mulitiple pregnancy, however, there is little robust published evidence about accuracy and false-positive or negative rates and although available commercially for multiple pregnancies, further research is needed to quantify these issues. Of course even if this form of screening becomes the norm in the future, in multiple pregnancy where chromosomal abnormality is likely to be discordant (particularly if there is dizygosity or polyzygosity), detailed ultrasound and, in cases where it is not obvious (i.e., no visible structural features aiding identification of the affected fetus), invasive testing will be required to identify and confirm diagnosis of an affected fetus.
Screening for fetal anomalies
Structural abnormalities, particularly cardiac abnormalities, are more common in twin and higher order pregnancies. This is mainly because of the higher incidence of abnormalities in monozygotic twins (owing to the unusual nature of the cleavage of the conceptus) compared with dizygotic twins[9,10].
The management of these pregnancies where one fetus has an abnormality is complex. Timely diagnosis enables more choices, time to prepare, optimizing fetal surveillance depending on the anomaly, involvement of other specialists (e.g., genetics team, pediatric surgeons) and appropriate birth planning (place, timing and mode), including access to intrauterine therapy where it is possible.
There is limited published evidence about screening for structural abnormalities in twin or higher order pregnancies. Logic suggests that the scan will take longer and that visualization at scan may be limited depending on fetal lies, but there is little reason to expect midtrimester ultrasound to be significantly less or more effective in multiple pregnancy. The limited evidence suggests DRs for twin pregnancy is similar to published data for singletons[9,11]. Therefore, routine anomaly screening by ultrasound between 18 weeks and 20+6 weeks gestation as in singleton pregnancy is recommended.
Abnormalities specific to monozygotic twins are midline, such as holoprosencephaly and neural tube defects, and cardiac abnormalities. Therefore, the value of fetal echocardiography in addition to routine anatomy scan is questioned. As not all monozygotic twins are monochorionic, this policy would need to be applied to all twins irrespective of chorionicity unless one were to undertake fetal sexing and exclude discordant sex twins, which can complicate matters (as couples may not want to know sex of the babies). A Scandanavian study of twin pregnancies[11] where women had a package of scans (NT scan, anomaly scan at 19 weeks, fetal echocardiography at 21 weeks and a cervical length at 23 weeks), found that 0.5% of the fetuses had cardiac anomalies, 80% of which were detected at the 19-week anomaly scan (i.e., before fetal echocardiography), and therefore concluded that formal fetal echocardiography is not justified.
Screening for FGR
Fetuses of multiple pregnancies are at increased risk of being small for gestational age (SGA) and, if there is placental dysfunction, intrauterine growth restricted (IUGR). Both SGA and IUGR fetuses and babies have poorer perinatal outcomes and therefore identifying growth problems is important.
Symphysis-fundal height measurement is not effective in identifying growth problems in twin pregnancy, and serial ultrasound scans are required to identify both small babies but also a significant size difference between fetuses.
The problem with interpreting the published literature to inform the best parameters to use is that criteria for abnormality and definitions of SGA or IUGR or growth discordance are variable and one is often not comparing like with like. The NICE guideline presents a review of 26 studies of ultrasound parameters in twin pregnancies, including various fetal biometric measurements, estimated fetal weight (EFW) based on formulae of ultrasound parameters, Doppler ultrasound of the umbilical cord and composite screening strategies[3]. Acknowledging that most of the evidence is low or very low quality, the conclusions were: i) any single fetal biometric parameter is a poor a predictor of IUGR or birthweight discordance; ii) an EFW ≤10th centile is a moderately useful predictor of IUGR, defined as birthweight ≤10th centile; iii) the best cut-off for intertwin birthweight discordance is an EFW difference of 25% or more; iv) the best EFW is derived when applying a formula that includes at least two biometric parameters; v) the best predictor of IUGR or discordance between twins is an ultrasound carried out within 28 days of birth; vi) there is no strong evidence supporting the routine use of umbilical artery (UA) Doppler for the prediction of IUGR or birthweight discordance; and vii) there is no strong evidence that any composite screening strategy detects IUGR in twin pregnancy. No studies addressed the value of amniotic fluid volume assessment or middle cerebral artery (MCA) Doppler examination. No studies addressed timing and frequency of scanning. There was no evidence to guide management of triplet pregnancies, but it seems logical to apply the conclusions to triplets and higher order pregnancies. On the basis of the detailed review, the recommendation is that EFW discordance should be calculated using two biometric parameters from 20 weeks’ gestation, scans should be undertaken at intervals of <28 days, a ≥25% EFW discordance should be considered significant, and UA Doppler should not be used to monitor for IUGR or birthweight differences in twin and triplet pregnancies[3].
More recently, a large UK cohort study of 2161 twin pregnancies (302 monochorionic and 1859 dichorionic twin pregnancies) has shown that EFW discordance is accurate in predicting birthweight discordance, both EFW and birthweight discordance are good predictors of adverse outcome and that the optimal cut-off for the prediction of perinatal mortality irrespective of chorionicity or individual fetal size is an EFW discordance of ≥25%[12].
Note: the formula for calculating EFW discordance percentage is:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
