Multifetal Pregnancies: Epidemiology, Clinical Characteristics, and Management
Lucas Otaño
César Meller
Horacio Aiello
Introduction
A multifetal gestation is a high-risk pregnancy because it is associated with an increased rate of maternal, perinatal, and infant morbidity and mortality compared to singleton pregnancies.1,2 Multifetal pregnancies account for 2% to 3% of all live births but are responsible for around 25% of the neonatal intensive care unit admissions. One of the main problems is preterm birth and its short- and long-term consequences.1
Epidemiology
The twin birth rate has significantly increased in most regions of the world during the last decades. In the United States, it rose around 75% in 35 years, from 19/1000 births in 1980 to 334/1000 births in 2016.3 The main reasons for this are delayed childbirth, advanced maternal age at conception, and the widespread use of assisted reproduction techniques (ART).
The incidence of monozygotic twins is around 1 in 250 pregnancies, and it is quite constant across the world,4 showing a slight increase in cases of ART. The incidence of dizygotic twins varies with race (it is more frequent in African ethnicity, intermediate in white European, and lower in East Asian), with maternal age (it increases with maternal age), and with family history (specially with maternal history), and it is highly influenced by the use of ART.4
ART plays an important role in twins, especially in dizygotic twins. In 2016, among infants conceived by ART, 30.4% were twins and 1.1% were triplets and higher order infants, with the consequent increase in adverse maternal and perinatal outcomes.5
Classification
Multifetal pregnancies can be classified according to the number of fetuses, zygosity, or chorionicity. According to the number of fetuses, multiple pregnancies can be twins, triplets, quadruplets, etc. In clinical practice, most are twins. According to zygosity, multiple pregnancies can be classified as dizygotic or monozygotic. Dizygotic twins result from the fertilization of two oocytes by two sperms that produce two genetically different embryos. Monozygotic twins result from the fertilization of one oocyte and one sperm and the early splitting of the resulting embryo in two genetically identical twins.6
According to chorionicity and amnionicity, twin pregnancies can be dichorionic (DC) or monochorionic (MC). In DC twins, each fetus has its own placenta and its own amniotic sac (diamniotic [DA]), whereas in MC, both fetuses shared the same placenta and can be DA (˜99%) or monoamniotic (MA; ˜1%).7
The relationship between zygosity, chorionicity, and amnionicity is described in Algorithm 5.1. Around 70% of twins are dizygotic, and all of them are DCDA. Monozygotic twins are the result of the cleavage of a single inner cell mass. If this cleavage occurs before the fourth day after conception, the twin pregnancy will be DCDA (˜30% of all monozygotic twins). If the cleavage occurs after the third day, the pregnancy will be MC (˜70% of monozygotic twins). In 99% of those MC twins, the cleavage of the inner cell mass will take place between the fourth and eighth day, and it will result in MCDA twins. Only in 1% of MC pregnancies the cleavage will occur after the eighth day resulting in MCMA
twins. In rare cases, the twinning process will take place after the 12th day producing an incomplete inner cell mass cleavage that results in conjoined twins.7
twins. In rare cases, the twinning process will take place after the 12th day producing an incomplete inner cell mass cleavage that results in conjoined twins.7
From an obstetrical standpoint, the most important distinction is if the twin pregnancy is DC or MC. MC twins have a significantly higher morbidity and mortality, and they also have complications exclusive for MC placentation. Knowledge of the chorionicity is crucial for prenatal screening and diagnosis of fetal anomalies, for choosing the optimal strategy of surveillance, for the appropriate management of complications, and for perinatal decision-making. In clinical practice, it is no longer acceptable to talk about “twin pregnancy”—it must be classified as MC or DC.
Ultrasound Diagnosis and Surveillance
Reliable diagnosis of multiple pregnancy is performed by ultrasound, ideally during the first trimester.1,8 Ultrasonography has become an integral clinical tool for appropriate screening, diagnosis, surveillance, and optimization of care in multifetal pregnancies.9
The ultrasound assessment of multiple pregnancies includes
Determining gestational age (dating of the pregnancy)
Determining chorionicity and amnionicity
Twin labeling
Ultrasound surveillance
Detection and management of complications
Determining Gestational Age
As in singletons, gestational age of multifetal pregnancies conceived spontaneously should be dated ideally during the first trimester, based on the fetus with the larger crown-rump length (CRL) at the 11- to 14-week scan. In cases of twins conceived by ART, dating should be based on the embryonic age from fertilization.8 For spontaneous twin pregnancies with first consultation after 14 weeks, gestational age is estimated based on the fetus with the larger head circumference.8
Determining Chorionicity and Amnionicity
Diagnosis of chorionicity is the most important goal for prenatal care and management of a twin pregnancy. Chorionicity is determined by ultrasound, and it is based on different features depending on the gestational age. It is usually best performed during the first trimester (sensitivity ≥98%), and it becomes more difficult during the second and third trimesters.10 Algorithm 5.2A and B and Figure 5.1 summarize the most important features for chorionicity and amnionicity assessment.
Algorithm 5.2 A, Assessment of chorionicity and amnionicity before 11 weeks of gestation. B, Assessment of chorionicity and amnionicity between 11 and 14 weeks of gestation. |
First Trimester
Before 10 weeks of gestation, determining chorionicity is usually easy. The first structures that can be seen at 5 to 6 weeks are the gestational sacs. In a panoramic view of the uterus, if one gestational sac with two embryos is observed (Figure 5.2A-D), the diagnosis of MC twin pregnancy is established.
Conversely, if two gestational sacs are identified with one embryo in each sac (Figure 5.2C and D), DC twin pregnancy is diagnosed. An ultrasound image showing the chorionicity should be documented and kept in the patient’s records for future reference.8
Conversely, if two gestational sacs are identified with one embryo in each sac (Figure 5.2C and D), DC twin pregnancy is diagnosed. An ultrasound image showing the chorionicity should be documented and kept in the patient’s records for future reference.8
Figure 5.1 Ultrasound assessment of chorionicity in dichorionic (A, top) and monochorionic (B, bottom) pregnancies. |
All DC twin pregnancies are DA. However, if an MC twin pregnancy is diagnosed, the next step is establishing amnionicity. Of note, more than 99% of MC twins are DA. To establish amnionicity in MC twins before 10 weeks, the amniotic membranes have to be identified. In early pregnancies, the amnion is seen as a very thin membrane that surrounds the embryo. From 8 to 9 weeks onward, a thin intertwin membrane dividing both amniotic sacs can be observed (Figure 5.3).
If the amniotic membranes cannot be identified, amnionicity can be inferred through the assessment of the yolk sacs. Usually, the number of yolk sacs correlates with the number of amniotic cavities. Therefore, if two yolk sacs are observed, it is more likely to be an MCDA (Figure 5.4). However, this concept has been challenged because several exceptions have been reported11,12 (Figure 5.5A). If monoamnionicity is suspected, confirmation can be obtained by the observing entanglement of the cords by color Doppler (Figure 5.5B).
At the 11- to 14-week scan, determining chorionicity and amnionicity is based on the assessment of the number of placental masses, and the characteristics of the membranes at the placental insertion sites. When two chorionic masses are identified, a DC twin pregnancy is diagnosed (Figure 5.6). If a single mass is observed, it could mean either an MC twin or a DC twin with fused placentas (Figure 5.7). In this circumstance, chorionicity is established by exploring the intertwin membranes at the site of the placental insertion. In DC twins, the intertwin membrane is thick and easily observed, and the placental insertion is thicker and shows a typical feature known as the “λ” (lambda) sign (Figure 5.8) produced due to the presence of chorionic tissue between membranes. Membrane insertion in MC twins shows a very thin membrane that
ends in a “T” form (Figures 5.7B and 5.8). Use of lambda sign predicts chorionicity with a high degree of accuracy before 14 weeks.13 Moreover, the combination of number of placentas, “T” sign, and “lambda” sign have shown more than 99% of sensitivity and specificity.14
ends in a “T” form (Figures 5.7B and 5.8). Use of lambda sign predicts chorionicity with a high degree of accuracy before 14 weeks.13 Moreover, the combination of number of placentas, “T” sign, and “lambda” sign have shown more than 99% of sensitivity and specificity.14
Figure 5.2 A and B, Monochorionic pregnancy with two embryos inside the same gestational sac. C and D, Dichorionic twin pregnancy with each embryo inside its own gestational sac (arrows). |
Other ultrasound features related to chorionicity are the thickness and the number of layers of the intertwin membrane (Figure 5.8). In MC twins, the membrane is made of two layers of amnion, is thin, and is sometimes even difficult to identify by ultrasound. In DC twins, the membranes have four layers—two chorionic and two amniotic—that are thicker and usually easily identifiable by ultrasound.
If chorionicity cannot be determined by abdominal ultrasound, a transvaginal approach should be attempted. If chorionicity cannot be determined during a first-trimester scan in a routine setting, a second opinion should be sought from a specialist. If MCMA is diagnosed, the patient should be referred to a tertiary center.8
Occasionally, there are pitfalls in determining chorionicity.15 For instance, a single bilobulated
placenta or different phenotypic sex in heterokaryotypic fetuses could be a source of a misdiagnosis.
placenta or different phenotypic sex in heterokaryotypic fetuses could be a source of a misdiagnosis.
Second and Third Trimesters
The performance in determining chorionicity decreases with increasing gestational age. Sometimes it becomes extremely difficult to establish a definite diagnosis. It has to be assessed using the number of placental masses and concordance or discordance of twins’ sex. If chorionicity cannot be established (DA with a single placental mass, twins of the same sex, and lack of first-trimester images), the pregnancy should be managed as MC.8
Twin Labeling
When referring to a twin pregnancy, it is desirable to differentiate each twin throughout the pregnancy. Thus, labeling the fetuses is a major process in the prenatal care of multifetal pregnancies. Labeling includes a description of each twin’s location (left-right, superior-inferior, anterior-posterior) and any other information that could help discriminate both twins: discordant sex, cord insertion sites, any discordant ultrasound marker, or anomaly.Usually, the lower twin is labeled as “A” or “1” and the other as “B” or “2.” For example, “twin A, at right side of the uterus, female, and with an intracardiac echogenic focus.” However, if during the pregnancy one twin changes position, it is useful to keep the original nomenclature.8
A consistent labeling is important for a better assessment and follow-up of complicated twin pregnancies. For example, in the first trimester when interpreting an abnormal genetic test, during the second and third trimesters when monitoring an abnormal condition (ie, intrauterine growth restriction [IUGR], a structural defect, etc). or for cases in which one twin is affected by a condition that requires special management (ie, congenital heart defect, diaphragmatic hernia, fetal anemia, etc).
Ultrasound Surveillance of Twin Pregnancies
The role of ultrasound in the follow-up of twin pregnancies has been mostly standardized during
the last decade, with several clinical guidelines published.1,8,16
the last decade, with several clinical guidelines published.1,8,16
Ultrasound assessment of twin pregnancies includes a first-trimester scan between 11 and 14 weeks. During the second and third trimesters, the ultrasound surveillance of an uncomplicated twin pregnancy varies according to chorionicity.
Ultrasound Examination at 11 to 13 6/7 Weeks and Screening for Chromosomal Anomalies
Ideally, every pregnant woman should undergo an ultrasound examination between 11 and 13 6/7 weeks. This is the optimal time for dating the pregnancy, establishing chorionicity and amnionicity, estimating aneuploidy risk, and screening for major structural anomalies (Figures 5.1A and B).
In dizygotic twins, the aneuploidy risk for each fetus is similar to single pregnancies. Thus, the probability of identifying an aneuploid fetus is two times the risk of a single pregnancy.1 On the other hand, in monozygotic twins, where fetuses are genetically identical, the likelihood of finding an aneuploidy is similar to single gestations.6 However, when an aneuploidy is diagnosed, usually both fetuses are affected. In rare cases, monozygotic twins have discordant karyotypes, which is known as heterokaryotypic twins.6,15
However, in clinical practice, the distinction between dizygosity and monozygosity is not always feasible. When assessing twins by ultrasound, it is important to bear in mind a few basic concepts: all MC twins are monozygotic, while all DC with discordant sex are dizygotic. DC twins with concordant sex are mostly dizygotic, but around 20% are monozygotic (Algorithm 5.1).
Current recommendations for first-trimester combined screening are as follows: in DC twins, an individual risk for each fetus is estimated according to nuchal translucency (NT) measurement and eventually according to other secondary markers such as nasal bone, ductus venosus (DV), or mitral valve regurgitation. In MC twins (always monozygotic), a unique risk is estimated for both fetuses based on the average NT of each fetus.8
Regarding the use of biochemical markers, such as free beta-hCG and PAPP-A, their performance is lower compared to singleton, but they still increase the detection rate for aneuploidies in twins. In higher order multiple pregnancies—three or more fetuses—biochemical markers are not useful.1,6
The risk for structural anomalies is also higher than in singletons, especially in monozygotic twins. Screening for structural anomalies is traditionally performed in the second trimester, but with the progressive improvement in imaging quality and standardization of anatomical assessment, a significant number of major anomalies can be detected at 12 to 13 weeks.17
Surveillance During Second and Third Trimesters
The ultrasound surveillance in second and third trimesters will depend on chorionicity.
DC Twins
MC Twins
In MC twins, after the first-trimester scan, it is recommended to conduct ultrasound surveillance every 2 weeks starting at 16 weeks. Basic ultrasound measurments include amniotic fluid assessment and estimated fetal weight (EFW), with a detailed scan at 22 to 22 weeks.1,8,16 The purpose of closer surveillance in MC twins is that early detection of severe complications, including twin-to-twin transfusion syndrome (TTTS) and selective intrauterine growth restriction (sIUGR), can occur. Around 10% to 15% of MC twins will develop a TTTS and ˜15% will have sIUGR.
The minimal contents in the ultrasound surveillance in MC twins vary among the different guidelines, but there is consensus that (Figure 5.1):
Every MC has to be assessed every 2 weeks, beginning at 16 weeks until delivery.
Every scan must include the maximum vertical pocket (MVP) of amniotic fluid of each sac for early diagnosis of TTTS and the EFW of each fetus for detection of sIUGR or discordant growth.
Every MC twin pregnancy should have a detailed scan at 20 to 22 weeks, including a cardiac screening assessment.
Women with uncomplicated MC twins should be delivered around 36 to 37 weeks—34 to 37 6/7 weeks—according to the 2014 American College of Obstetricians and Gynecologists (ACOG) Practice Bulletin.
There are some other aspects that differ among guidelines, like the use of Doppler and CL measurement. Regarding the use of Doppler in uncomplicated twins, there is no consensus on when, how often, and which vessels should be assessed. The measurement of CL is also suggested in the second trimester in the International Society of Ultrasound in Obstetrics and Gynecology (ISUOG) guidelines,8 but it is not recommended in the ACOG Practice Bulletin1 or in the Royal College of Obstetricians and Gynaecologists (RCOG) guidelines.16
Complications in Multifetal Pregnancies
Complications in multiple pregnancies, either maternal or fetal, are more frequent than in singletons. In addition, the higher the number of fetuses, the higher the complication rate.
Maternal Complications
Maternal complications are more common in patients with twins than in singletons1 and include a higher incidence of gestational hypertension and preeclampsia, gestational diabetes mellitus, acute fatty liver of pregnancy, intrahepatic cholestasis of pregnancy, hyperemesis, anemia, urinary tract infections, postpartum hemorrhage, cesarean delivery, and postpartum depression.1,18 Moreover, the risk of severe acute maternal morbidity and near-miss cases is four to five times higher than in singletons.2 Of note, neither zygosity nor chorionicity seem to influence the prevalence of maternal complications.19,20,21
In 2018, the American College of Obstetricians and Gynecologists and Society for Maternal-Fetal Medicine issued an opinion on “low-dose use aspirin during pregnancy”22 and classified multifetal gestation as a risk factor for preeclampsia and, therefore, a candidate for prophylaxis with low-dose aspirin between 12 and 28 weeks of gestation (ideally starting before 16 weeks).22
Fetal Complications
There is increased incidence of
Vanishing twin
Prematurity
IUGR
Fetal malformations
Fetal death
Placental anomalies
In addition, the prevalence and severity of complications are also related to chorionicty and amnionicity. Fetal complications are more frequent in MC twins than in DC twins, and there are complications that are exclusive to MC twins.
In DC twins, fetal circulations are completely independent, even in those with fused placental masses. On the other hand, almost all MC twins have placental vascular anastomosis between fetal circulations (Figures 5.9 and 5.10). This difference in placental angioarchitecture is the main factor that accounts for the excess of morbidity and mortality in MC twins. The risk of mortality in MC twins is two times the risk of that in DC twins and four times the risk of singletons. In the same way, MC twins have four to five times higher risk of
neurologic morbidity than DC twins and 25 to 30 times the risk of singletons.23
neurologic morbidity than DC twins and 25 to 30 times the risk of singletons.23
Figure 5.10 Monochorionic twin pregnancy with several placental anastomoses. Note a great arterioarterial anastomosis (▲) between the two cords. |
Vanishing Twin
Early miscarriage of one twin in the first trimester is a phenomenon that can occur in any twin pregnancy. It has been extensively evaluated in pregnancies conceived by in vitro fertilization, where a prevalence of 15% to 36% has been reported. The clinical outcome of the surviving twin has shown some controversial results, but Romanski et al demonstrated that there is no difference in perinatal outcomes in pregnancies that progress to at least 24 weeks of gestation.24
Preterm Birth
Women with multifetal gestations are 5 to 13 times more likely to give birth preterm than singletons,1,3 and prematurity is the main source of perinatal morbidity and mortality.18 Around 50% to 60% deliver before 37 weeks, around 20% before 34 weeks, and around 12% before 32 weeks.1,3 Spontaneous as well as iatrogenic preterm births are increased in multifetal pregnancies. Uterine overdistension, preterm rupture of membranes, and preterm labor may be responsible for the increased risk in spontaneous prematurity, while the higher incidence of maternal and fetal complications increase the risk of iatrogenic preterm delivery. Given the magnitude of the problem, developing accurate screening tools and effective prevention interventions are major areas of research.
The most common screening test for preterm birth in both singletons and twins is cervical length measured at transvaginal ultrasound. In twins, ISUOG Practice Guidelines recommend one measurement in midgestation8; other guidelines do not include cervical length in the screening protocols,1,16,25 whereas other groups recommend performing serial measurements.26 Unfortunately, all strategies have shown only a modest performance in screening for preterm birth.18,27
Multiple interventions have been evaluated to reduce preterm birth in twins, but none have been proven to be effective.1,18,28 However, this is an area where practice is likely to change in view of the emerging evidence.27 Potential interventions that have been studied include the following:
Hospitalization or bed rest: This does not prolong pregnancy and may be harmful as it increases the risk of venous thromboembolism and maternal stress. However, a 2017 Cochrane review29 concluded that there is a need for large-scale, multicenter, randomized controlled trials before definitive conclusions can be drawn.
Progesterone: Guidelines do not recommend the use of progesterone in twins.1,28,30 However, a 2017 meta-analysis of individual patient data showed that vaginal progesterone decreases preterm birth <33 weeks (RR 0.69, 95% CI 0.51-0.93) and neonatal morbidity and mortality in women with
a twin gestation and a cervix < 25 mm in midtrimester.31 Further research is required before definite conclusions can be drawn. Importantly, evidence suggests that there are no long-term adverse outcomes caused for in utero exposure.30,31
Cerclage: Guidelines do not recommend the use of cerclage in twins.1,28 However, a 2019 meta-analysis found that cerclage reduced preterm birth in twins with cervical length of <15 mm or dilated cervix of >10 mm.32 Importantly, there was a difference in the effects between randomized controlled trials (tending to indicate that cerclage placement was not beneficial) and the cohort studies (showing encouraging results). Therefore, as with progesterone, further research is needed.
Tocolytic therapy: As in singletons, prophylactic therapy is not recommended in twins.1 A Cochrane review33 showed that oral betamimetics did not reduce preterm birth in asymptomatic women with twin pregnancies, and guidelines advice against its use.28 On the other hand, there is consensus that one course of tocolysis for up to 48 hours is acceptable, depending on gestational age, to allow a course of corticosteroids and magnesium sulfate in the setting of acute preterm labor.1,18 Finally, the indications for antenatal corticosteroids up to 34 weeks of gestation and magnesium sulfate for fetal neuroprotection are similar to those in singletons.
Corticosteroids: These should be administered for pregnant women who are at risk of preterm delivery within 7 days regardless of fetal number, starting at 23 to 24 weeks of gestation and up to 34 weeks.1 Serial courses (more than two) are not recommended, but a single repeat course may be considered in patients <34 weeks if the first course was administered more than 7 to 14 days previously.1 Even though some guidelines included a single course of corticosteroids in the late preterm period (34-36 6/7 weeks) in singletons that have not received a previous course, there is still no evidence for this approach in twins.34 Importantly, routine prophylactic administration of corticosteroids to all twin pregnancies should be avoided and may be harmful.18,35
Magnesium sulfate: This should be considered for fetal neuroprotection when women present between 23 to 24 weeks and 30 to 33 6/7 weeks (depending on different guidelines) with imminent preterm birth, defined as a high likelihood of birth because of active labor with cervical dilatation ≥4 cm or planned preterm birth for fetal or maternal indications, regardless of fetal number.1,36,37,38
Intrauterine Growth Restriction
The prevalence of IUGR is increased in multifetal gestations, and it is a major cause for fetal morbidity and mortality, not only for the affected twin but also for the normal co-twin. Thus, fetal growth assessment is one of the main features to be monitored in multifetal pregnancies’ surveillance.
IUGR management in multifetal pregnancies is different than in singletons. The clinical decision-making is highly influenced by chorionicity and the presence of the normal co-twin. IUGR in MC twins will be discussed in detail below.
Discordant Anomalies
The prevalence of congenital anomalies is 1.3-fold higher in twins than in singletons.39 Therefore, screening for fetal anomalies is an issue of utmost importance.1,8 Although dizygotic twins are almost always discordant for anomalies, monozygotic twins show a concordance rate of structural anomalies of 20%, and if present, chromosomal anomalies and single-gene disorders usually affect both fetuses.39
There are some discordant anomalies unique to monozygotic twins, like the twin reversal arterial perfusion (TRAP) sequence or acardiac fetus (see below). Among dizygotic twins, there is also a unique condition: the coexistence of a complete hydatidiform mole and a normal fetus. This is a very high-risk condition both for the mother and the fetus that needs to be managed in a tertiary center by a multidisciplinary team that includes a gynecologic oncologist.40
Discordant anomalies should be managed on an individual basis after a comprehensive counseling according to the type of anomaly, gestational age at diagnosis, chorionicity, and parental preferences. Expectant management, termination of pregnancy, fetal therapy, and selective feticide are among the different therapeutic alternatives. Decision-making in these situations is usually complex because any decision may also affect the normal fetus.
Fetal Death
The fetal death rate is increased in multifetal pregnancies. In a population-based study, the rates of spontaneous loss of fetuses before week 22 were 0.9%, 2.4%, and 20.8% for DC, MCDA, and MCMA
twins, respectively.41 The risk of intrauterine death after 22 weeks was also higher for MC than DC twins. The odds of pregnancies resulting in at least one live born infant were 98.2% in DC, 92.3% in MCDA, and 66.7% in MCMA.41
twins, respectively.41 The risk of intrauterine death after 22 weeks was also higher for MC than DC twins. The odds of pregnancies resulting in at least one live born infant were 98.2% in DC, 92.3% in MCDA, and 66.7% in MCMA.41
There are significant differences in the clinical implications of a fetal death or imminent fetal death of one twin between MC and DC pregnancies due to the presence of placental vascular anastomoses in MC twins (see “Single fetal demise in MC twins”). Although severe hemodynamic changes in one twin can adversely affect the co-twin in MC pregnancies, in DC twins, imminent death of one twin has no direct effect on the co-twin’s health (Figure 5.11).
Placental Anomalies
Twin pregnancies are associated with an increase prevalence of placenta previa, vasa previa, and velamentous cord insertions.42,43,44 Thus, it is important to carefully assess the placentas’ and cord insertions during the ultrasound examination.