Multiple Gestation: Clinical Characteristics and Management

Diagnosis of Multiple Gestation

Positive sonographic diagnosis of multiple gestation can be made by visualizing multiple gestational sacs with yolk sacs by 5 weeks’ gestation and multiple embryos with cardiac activity by 6 weeks. If two gestational sacs are seen on early ultrasound studies, the chance of delivering twins is 57%; this increases to 87% if two embryonic poles with cardiac activity are visualized. If three gestational sacs are seen on early ultrasound, the chance of delivering triplets is 20%, increasing to 68% if three embryonic poles with cardiac activity are visualized. In addition to twins, the early sonographic visualization of two intrauterine fluid collections may represent a singleton in a bicornuate uterus, a singleton with a subchorionic hemorrhage, or a “vanishing twin,” in which an initially obvious second gestational sac or embryo does not continue after the first trimester.

Ultrasound evaluation is fundamental for the diagnosis of multifetal gestation. The Routine Antenatal Diagnostic Imaging with Ultrasound Study (RADIUS) of over 15,000 pregnant women reported that 38% of twin pregnancies remained unrecognized until after 26 weeks’ gestation in women who did not have a routine second-trimester ultrasound examination, and 13% of twins were not diagnosed until delivery. The Helsinki Ultrasound Trial reported similar findings: approximately 25% of twin pregnancies were not identified until after 21 weeks’ gestation.

Chorionicity

Because 20% of twins are monochorionic and such pregnancies are associated with a higher perinatal mortality risk that may be influenced by obstetric care, accurate determination of chorionicity is essential for clinical management. In most women, sonographic assessment can accurately determine chorionicity. Sonographic determination of chorionicity should be sought for all multiple gestations and is best performed in the first trimester. Before 8 weeks’ gestation, visualizing clearly separate gestational sacs, each surrounded by a thick echogenic ring, is suggestive of dichorionicity. If separate echogenic rings are not visible, monochorionicity is likely. In such situations, counting the number of yolk sacs may assist in establishing amnionicity. Two fetal poles with two yolk sacs in a monochorionic gestation suggests diamnionicity, whereas the presence of two fetal poles with only one yolk sac suggests a monoamniotic gestation. However, the specificity of this finding for monoamnionicity is uncertain. The sensitivity of first- and second-trimester ultrasound for predicting monochorionicity is approximately 90%; the specificity falls from 99% for first-trimester sonography to 95% in the second trimester ( Table 37.2 ). Later in gestation, if the fetuses are discordant for sex or two distinct placentas are seen, a dichorionic gestation can be confirmed with confidence ( Fig. 37.1 ). In the absence of these findings, monochorionicity is possible, and other sonographic features should be assessed.

The visualization of only one placental mass has a positive predictive value for monochorionicity of only 42%, because many dichorionic gestations can develop apparent fusion of separate placentas as pregnancy progresses. Counting the number of layers in the dividing membrane, near its insertion into the placenta, is 100% predictive of dichorionicity but is not as reliable in predicting monochorionicity. When this method is used, it is assumed that the placentation is monochorionic if only two layers are present, with the presence of three or four layers suggesting dichorionicity. Although there is no consensus on the definition of a thin or thick dividing membrane, the use of a membrane thickness cutoff value of 2 mm has also been reported to correctly assign chorionicity in more than 90% of cases, but the reproducibility of this measurement has been questioned. Given that the difference in membrane thickness is less obvious later in pregnancy, accounting for the reduction in sensitivity when ultrasound is taken in the second trimester, first-trimester visualization and diagnosis are optimal. Visualization of a triangular projection of placenta between the layers of the dividing membrane (known as the twin-peak or lambda sign) is also useful in diagnosis of dichorionicity, but its absence is not as reliable for predicting monochorionicity. Although each of these sonographic features individually has a poor positive predictive value for monochorionicity, use of a composite sonographic approach (i.e., one placenta, sex concordance, thin dividing membrane, and absence of the twin-peak sign) may yield a positive predictive value for monochorionicity of 92%.

The use of transvaginal sonography in the first trimester, together with this composite approach, produces correct assignment of chorionicity and amnionicity in almost 100% of cases. If the initial ultrasound examination is not performed until the second trimester, its precision in assigning chorionicity declines. Although sensitivity is not perfect, specificity for monochorionicity is almost 100% when this approach is used in the first trimester, falling to 95% in the second trimester. As stated earlier, because different types of twin pregnancies require different counseling and follow-up, it is important to provide a comprehensive ultrasound evaluation, ideally in the first trimester, by a trained professional with experience in the diagnosis of multifetal gestations.

Detection of Fetal Anomalies

Detailed sonographic survey of fetal anatomy is indicated in multifetal pregnancies, because the risk for congenital anomalies is increased. The accuracy of ultrasonography for detecting congenital fetal anomalies in multiple gestations has not been adequately studied in large series. Smaller, single-center series have tried to establish the predictive value of prenatal ultrasound for the detection of anomalies in multiple gestations. In a series of 24 anomalous fetuses in twin gestations, serial ultrasonography at a specialist center achieved an 88% detection rate, with 100% specificity, for the prenatal diagnosis of anomalies. An 83% rate of detecting fetuses with Down syndrome in twin pregnancies was achieved by combining risks derived from maternal age and nuchal translucency thickness measurement at 10 to 14 weeks’ gestation, with a 5% false-positive rate. The finding of increased nuchal translucency in one fetus of a monochorionic pair may also presage the development of twin-twin transfusion syndrome (TTTS).

Evaluation of Fetal Growth

Serial ultrasonography is the most accurate method to assess fetal growth in cases of multiple gestation. Intrauterine growth of twins is similar to that of singletons until 30 to 32 weeks’ gestation, when the abdominal circumference measurements of twins begin to lag behind those of singletons. The landmark study by Grantz and colleagues was a prospective cohort of 171 women with twin gestations. Its objective was to define the trajectory of fetal growth in dichorionic twins empirically using longitudinal two-dimensional ultrasonography and to compare the fetal growth trajectories for dichorionic twins with those based on a growth standard that was developed for singletons. The patients were recruited from eight US sites from 2012 to 2013. After an initial sonogram at 11 0/7 to 13 6/7 weeks’ gestation during which dichorionicity was confirmed, women were assigned randomly to one of two serial ultrasonography schedules. The comparatively asymmetric growth pattern in twin gestations, initially evident at 32 weeks’ gestation, is consistent with the concept that the intrauterine environment becomes constrained in its ability to sustain growth in twin fetuses. Near term, nearly 40% of twins would be classified as small for gestational age based on a singleton growth standard.

Composite assessments of fetal weight appear to be superior to individual biometric parameters (e.g., abdominal circumference, femur length) for predicting growth discordance. Although individual growth curves for twin and triplet gestations have been described, singleton fetal weight standards are still commonly used to assess growth in multiple gestation. Because growth restriction is a dynamic process and sibling fetuses are immediately available for comparison, we consider it reasonable to assess growth in multiple gestation with serial evaluations, based on singleton growth curves, using as many biometric parameters as possible and comparing sibling estimated fetal weights for discordance.

Similarly, until recently, there was uncertainty regarding the degree of intertwin growth discordance that was considered clinically significant. The ESPRIT trial prospectively followed 1028 unselected twin pregnancies with detailed serial sonographic assessment. Perinatal morbidity increased only when the degree of growth discordance exceeded 18%, a threshold that surprisingly was the same regardless of chorionicity. Growth discordance greater than 20% has also been shown to be an important predictor for adverse perinatal outcomes, even when individual fetal sizes are appropriate for gestational age.

It is unclear whether adverse outcomes seen with significant weight discordance are related to continuation of pregnancy in a “potentially hostile” intrauterine environment or to iatrogenic prematurity. We use significant weight discordance as an indication for close fetal surveillance rather than an indication for immediate delivery. Decisions regarding delivery are then made based on the results of tests of fetal well-being, together with gestational age, rather than solely based on significant weight discordance.

Measurement of Cervical Length

Several strategies have been proposed to prevent preterm deliveries in twins: tocolytics, bed rest, hospitalization, home uterine activity monitoring, cerclage, and, most recently, progesterone. Unfortunately, none has proven effective in multiple gestations. Whether to perform universal cervical length screening in twins remains controversial. A randomized controlled trial by Gordon and associates of 125 twin pairs without prior preterm birth (<28 weeks) showed that routine second-trimester transvaginal ultrasound assessment of cervical length was not associated with improved outcomes when incorporated into the standard management of otherwise low-risk twin pregnancies. However, ultrasound surveillance of cervical length in multiple gestations can identify those at increased risk for preterm delivery. The largest study of cervical length assessment in this setting followed 1163 twin pregnancies after cervical length measurement at 22 to 24 weeks’ gestation. The median cervical length was 35 mm, and in only 8% of twin pregnancies was the cervix less than 20 mm long. A recent meta-analysis by Lim and coworkers concluded that second-trimester cervical length is a strong predictor of preterm birth in twins. However, the authors cautioned that, in the absence of effective preventive strategies, routine cervical length measurement in this population may not be warranted.

The technique of sonographic measurement of cervical length in multiple gestations does not differ from that described for singleton gestations. The optimal interval at which to perform sonographic assessments of cervical length during multifetal gestation is unclear. Our practice has been to measure cervical length every 2 weeks from 16 to 24 weeks’ gestation in the cases of multiple gestation deemed to be at highest risk for preterm delivery (e.g., higher order multiple gestations, maternal history of a preterm singleton birth). For all other multiple gestations, it may be reasonable to perform sonographic assessment of cervical length at the time of sonography for fetal anatomy or growth.

Confirmation of Fetal Well-Being

Ultrasonography is useful to confirm fetal well-being in multiple gestations. The nonstress test (NST) in multiple gestation is discussed in Chapter 32 . The biophysical profile may also be of benefit in multiple gestation if an NST is abnormal or when an NST is impractical to perform, as it may be in cases of higher order multiple gestation. There is no evidence that routine biophysical profile testing in the absence of specific additional high-risk factors has any benefit in multiple gestations. In a study of 539 twin pregnancies by Booker and coworkers, the use of the sonographic biophysical profile for routine antenatal surveillance had a low false-positive rate, with a very low incidence of fetal death. The authors concluded that the sonographic biophysical profile should be considered as a primary mode for antenatal surveillance in twin pregnancies, with a reflex NST for an abnormal biophysical score, although efficacy remains uncertain.

Doppler velocimetry may also be used to evaluate fetal well-being in multiple gestations. Umbilical artery systolic-to-diastolic ratios are similar in singleton and twin gestations. A decrease in this ratio may occur before the sonographic detection of growth restriction. Normal Doppler velocimetry indices of other fetal vessels, such as the middle cerebral artery and the descending aorta, are similar for singleton, twin, and triplet fetuses. We employ Doppler velocimetry of the umbilical artery whenever multiple gestation is complicated by significant growth restriction or discordance. Monitoring monochorionic-diamniotic pregnancies usually begins at 16 to 18 weeks’ gestation by assessment of amniotic fluid volume (AFV) and fetal bladder in both twins for early detection of TTTS. Furthermore, it may be reasonable to begin measurement of middle cerebral artery peak systolic velocity in both fetuses of a monochorionic-diamniotic twins at 26 to 28 weeks for early detection of twin anemia-polycythemia sequence (TAPS). There are inadequate data to determine the optimal frequency of such monitoring, but measurement once a week seems reasonable, with more frequent monitoring if abnormalities are detected (e.g., discordant AFVs that do not yet meet criteria for stage I TTTS).

Sonographic measurement of AFV is an important tool to evaluate fetal well-being. In a dye-dilution study of diamniotic twin pregnancies, the AFV in each amniotic sac was noted to be independent of the volume in the neighboring sac and was similar to singleton fluid volumes. There is no agreement on the optimal sonographic method to assess AFV in multiple gestations. Methods in use include the following ^, :

• A single overall amniotic fluid index without reference to the dividing membrane

• Individual amniotic fluid indices for each sac

• Largest two-diameter pocket in each sac

• Single deepest pocket in each sac

• Subjective assessment of the relative distribution of fluid between sacs

No one method, however, has been shown to be optimal for predicting perinatal outcome in multiple gestation. A systematic review by Ippolito and colleagues of over 13,000 twins found sonographic measurements of AFV in twin pregnancies to be reproducible and not dependent on sonographer experience. However, sonographic estimation of AFV (using the amniotic fluid index, two-diameter pocket, or single deepest pocket) has high specificity but poor sensitivity for detecting abnormal AFV (oligohydramnios or polyhydramnios). These discrepancies among methods for AFV assessment in identifying abnormal fluid volumes suggest that caution should be exercised when interpreting data characterizing AFV in normal and complicated pregnancies. Our own practice, however, is to focus on deepest vertical pocket when quantifying AFV in multiple gestations.

Risks of Chromosomal Abnormalities

Monozygotic twins are thought to have the same Down syndrome risk per pregnancy as maternal age–matched singletons, and dizygotic twin pregnancies are thought to have twice the risk of at least one affected fetus as maternal age–matched singleton pregnancies. After correction for different rates of dizygosity based on maternal age and race, it has been suggested that invasive prenatal diagnosis should be offered to women in the United States with twin gestations who are 31 years of age or older ( Table 37.3 ). In a triplet gestation, the same calculus applies: The chance that a 28-year-old woman will have at least one fetus with Down syndrome is similar to that of a 35-year-old woman with a single fetus. These alterations in risk assessment for chromosomal abnormalities are important because invasive prenatal diagnosis may be considered at an even earlier maternal age for gravidas with higher order multiple gestations.

All women, regardless of age, should be made aware of the relative advantages and disadvantages of screening versus diagnostic tests for fetal aneuploidy. Women with multiple gestations who are concerned about aneuploidy risk should be counseled and supported to make their own personal decision about choosing a screening test, a diagnostic test, or no testing at all. As per American College of Obstetricians and Gynecologists (ACOG) guidelines, invasive genetic testing by amniocentesis or chorionic villus sampling should be offered to any woman with a multiple gestation who desires definitive genetic testing.

First-Trimester Screening for Aneuploidy

With the availability of late first-trimester MFPR procedures in higher order multiple gestations, interest in first-trimester screening tests for fetal abnormalities has increased. First-trimester screening now encompasses noninvasive prenatal testing (NIPT) in the form of cell-free fetal DNA or analyte screening.

Noninvasive Prenatal Screening

In singleton pregnancies, cell-free DNA analysis of maternal blood provides effective screening for trisomies 21, 18, and 13. It is the most sensitive and specific test for screening for the common aneuploidies. ACOG now recommends offering cell-free DNA analysis to all pregnant women regardless of maternal age or baseline risk. However, ACOG recognizes that no method of aneuploidy screening is as accurate in multiples as it is in singletons.

Noninvasive prenatal screening for Down syndrome using cell-free fetal DNA may be challenging with multiple gestations because cell-free fetal DNA in the maternal circulation is derived from each fetus. Because it is impossible to determine which twin is abnormal based on cell-free fetal DNA analysis alone, results are reported for the entire pregnancy, and invasive testing is required to distinguish which twin, if any, is affected. An additional challenge in twin pregnancy is that the amount of cell-free fetal DNA contributed by each twin is lower than in singleton pregnancies and may be quite different for the two fetuses in cases of dizygotic twins. Nonetheless, there is mounting evidence that screening test performance of cell-free fetal DNA in multiples is similar to that reported in singleton pregnancies and is superior to that of the first-trimester combined test or second-trimester analyte screening.

Gill and colleagues reported a systematic review and meta-analysis of cell-free fetal DNA in 1272 twin pregnancies. They found the detection rate for trisomy 21 was 99% and the false positive rate was less than 0.1%, with the reported number of twin pregnancies with trisomies 18 and 13 being too small for accurate assessment of the predictive performance of the cell-free fetal DNA test. This was confirmed in an updated meta-analysis and cohort study by the same group. As cell-free fetal DNA is obtained via a serum sample, it does not negate the need for a first-trimester ultrasound, which is invaluable in confirming chorionicity, assessing viability, and evaluating the nuchal translucency measurement, which is a marker for both genetic and structural abnormalities.

Cell-free fetal DNA in twin pregnancy is associated with a higher failure rate or “no call result” rate than in singleton pregnancies. This is due to the use of an estimation of fetal fraction for each twin of at least 4% that is used to minimize the risk of a false-negative result. Galeva and colleagues sought to determine predictors of failure in obtaining a result with first testing in twins. They found that increasing maternal weight, race, parity, earlier gestational age, mode of conception, dichorionicity, and serum-free beta hCG and PAPPA levels likely related to a small placental mass were independent predictors of cell-free fetal DNA test failure. When an insufficient fetal DNA result is first obtained, most women will subsequently obtain a result on the second sampling, although those with a higher pretest probability of failure should be informed as such. The decision to repeat the test or move to invasive genetic testing should be individualized.

First-Trimester Combined Screening

First-trimester combined screening for aneuploidy with nuchal translucency, free β-hCG, and pregnancy-associated plasma protein A (PAPP-A) was previously considered the standard of care in the United States for singleton pregnancies. However, data remain insufficient to generate recommendations for the widespread use of this combined screening for multiple gestations. First-trimester free β-hCG and PAPP-A levels are about twice as high in twin pregnancies as in singleton pregnancies. Chorionicity and the use of assisted reproductive techniques have a significant impact on first-trimester maternal serum marker levels, making their interpretation in the setting of multiple gestations more complex. PAPP-A and free β-hCG levels are significantly lower in monochorionic than in dichorionic twin pregnancies, and PAPP-A levels are lower in pregnancies from IVF than in those spontaneously conceived. Data supporting combined screening in twin pregnancies are limited, with insufficient numbers of affected pregnancies to allow robust data interpretation. In one study of 448 twin gestations, this form of combined screening was estimated to have an 88% detection rate for Down syndrome, with a 7.3% screen-positive rate. In another series of 24 multiple gestations and 79 singleton control subjects, the distribution of nuchal translucency measurements, including 95th percentile values, was similar in all cases, implying that this form of screening could be implemented using established normative data from singleton populations. Care needs to be taken when interpreting nuchal translucency measurements in monochorionic twins, because an increased measurement in only one fetus may represent early TTTS rather than elevated aneuploidy risk. One of the main benefits of nuchal translucency measurement in multiple gestations is that, if MFPR is being planned, this measurement can be useful in deciding which fetus or fetuses to target for reduction.