Multifetal Pregnancy
BS. Nguyễn Hồng Anh
Multietal pregnancies may result rom two or more ertilization events, rom a single ertilization ollowed by a splitting o the zygote, or both. In part because o inertility therapy, the twin birth rate in the United States was 3.2 percent in 2019. For higher-order multietal births, the number peaked in 1998
at 0.2 percent o all births. Subsequent eorts in the reproductive endocrinology community to curb this rate have led to
declines. Specically, the rate o triplets or more declined by 55
percent rom 1998 to 2019 (Martin, 2021).
Multietal gestations remain problematic or both the mother
and her etuses. For the etus, multietal births accounted or
3 percent o all live births in 2013 but or 15 percent o all
inant deaths. Specically, the inant mortality rate or twins
was more than our times the rate or singletons, and or triplets
it was 12-old higher (Matthews, 2015). Neonates rom multi-
etal gestations also make up a disproportionate percentage o
very-low-birthweight newborns in the United States (Martin,
2021). Tese two outcomes in singletons and twins rom Parkland Hospital are shown in Table 48-1. Risks or congenital
malormations and preterm birth are greater with multietal
gestation and discussed later.
Te mother may also experience higher morbidity and mortality rates, and these rise with the number o etuses (Society
or Maternal-Fetal Medicine, 2019). In one study o more than
44,000 multietal pregnancies, the risks or preeclampsia, postpartum hemorrhage, and maternal death were twoold higher
than these rates in singleton gestations (Walker, 2004). Rates
o placenta previa and placenta accreta spectrum are increased
(Miller, 2021a). Moreover, compared with rates or singletons
the peripartum hysterectomy rate was threeold greater or
twins and 24-old higher or triplets or quadruplets in one
study (Francois, 2005).
MECHANISMS OF MULTIFETAL GESTATIONS
■ Dizygotic versus Monozygotic Twinning
Multietal gestations are oten described by zygosity, amnionicity, and chorionicity, which are the number o zygotes, amnions,
and chorions, respectively. win etuses most oten result rom
ertilization o two separate ova, which yields dizygotic or raternal twins. Less oten, twins arise rom a single ertilized ovum
that then divides to create monozygotic or identical twins.
Dizygotic twins are not in a strict sense true twins because
they result rom the maturation and ertilization o two ova
during a single ovulatory cycle. Genetically, these twins are like
any other pair o siblings. wins o opposite sex are almost
always dizygotic. Rarely, somatic mutations or chromosome
aberration in one monozygotic twin can yield diering gender
karyotype or phenotype. For example, postzygotic loss o the Y
chromosome in one 46,XY twin creates a phenotypically emale
twin with 45,X urner syndrome.
Monozygotic twins, although they have the same genetic
heritage, also are usually not identical. For example, the division o one ertilized zygote into two does not necessarily yield
equal sharing o protoplasmic material. In addition, monozygotic twins may be discordant or genetic mutations because
o a postzygotic mutation, or they may have the same genetic
disease but with marked variability in its expression. In emale
etuses, skewed lyonization can produce dierential expression o X-linked traits or diseases. Last, sesquizygosity is a rare
event o dispermic ertilization that leads to monozygotic twins.
Some cases may have sex discordance (Gabbett, 2019).
■ Monozygotic Twinning
Mechanisms underlying monozygotic twinning are poorly
understood. One association is assisted reproductive technology (AR) and in-vitro ertilization (IVF). Te monozygotic
twinning incidence is twoold greater in pregnancies conceived
using blastocyst transer compared with transer o a later-stage
embryo. Te predisposition to splitting may stem rom culture
media eects (Busnelli, 2019).
Te outcome o the monozygotic twinning process depends
on when division occurs as depicted in Figure 48-1. I zygotes
divide within the rst 72 hours ater ertilization, two embryos,
two amnions, and two chorions develop, and a monozygotic,
dichorionic, diamnionic twin gestation evolves. wo distinct
placentas or a single, used placenta may develop. I division
occurs between days 4 through 8, a monozygotic, monochorionic, diamnionic twin pregnancy results. By 8 days, the chorion and the amnion have already dierentiated, and division
results in two embryos within a common amnionic sac, which
is a monozygotic, monochorionic, monoamnionic twin gestation.
Conjoined twins result i division initiates later.
Rarely, monochorionic twins may be dizygotic. Mechanisms
are unclear. In one review o 31 cases, nearly 80 percent were
conceived ater AR procedures, and 90 percent were associated with chimerism (Peters, 2017).
■ Factors Affecting Twinning
Dizygotic twinning is much more common than monozygotic
splitting o a single oocyte, and its incidence is positively inuenced by inertility treatment and by maternal age, race, heredity, and size. By contrast, the requency o monozygotic twin
births is relatively constant worldwide—approximately one set
per 250 births. Tis incidence is generally independent o most
demographic actors, except AR.
O inertility therapies, ovulation induction with clomiphene
citrate or with ollicle-stimulating hormone (FSH) plus human
chorionic gonadotropin (hCG) remarkably enhances the likelihood o multiple concurrent ovulations. Moreover, with IVF,
the greater the number o embryos that are transerred, the
greater the risk o twins or more. In 2017, AR contributed to
1.9 percent o all newborns in the United States and to 14.7 percent o all neonates in multietal gestations (Sunderam, 2020).
Te American Society or Reproductive Medicine (2017)
revised their guidelines regarding the number o embryos or
blastocysts transerred during IVF to reduce the incidence o
multietal pregnancies. For example, women younger than 35
years are encouraged to receive a single-embryo transer, regardless o embryo stage. Tese practices have eectively lowered
multietal rates (Martin, 2021).
Advancing maternal age and delayed childbearing are other
important risk actors (Adashi, 2018; Otta, 2016). One explanation is FSH levels, which rise with age and lead to greater
ovarian stimulation (Beemsterboer, 2006). Higher levels o
FSH have been linked with dizygotic twinning (Lambalk,
1998). Another explanation may be a higher use o AR in
older women. Paternal age has also been linked to twinning,
but its eect is elt to be small (Abel, 2012).
Diferent races and ethnic groups vary in their requency o
multietal births. In the United States in 2019, the twin birth
rate was 4.1 percent in black women, 3.3 percent in whites,
and 2.5 in Hispanics (Martin, 2021). In one rural community
in Nigeria, twinning occurred once in every 20 births (Knox,
1960). Tese marked dierences in twinning requency may
stem rom racial variations in FSH levels (Nylander, 1973).
Heredity is another actor, and maternal inuence appears
to supersede that o the ather. In a study o 4000 genealogical records, women who themselves were a dizygotic twin gave
birth to twins at a rate o one set per 58 births (White, 1964).
Women who were not a twin, but whose husbands were a dizygotic twin, gave birth to twins at a rate o one set per 116
pregnancies. Genome-wide association studies have identied
potentially contributory genes, and two are related to FSH
(Mbarek, 2016). However, the contribution o these variants
to the overall incidence o twinning is likely small (Hoekstra,
2008).
Maternal size is another risk actor. Nylander (1971) showed
an increasing gradient in the twinning rate in taller, heavier
women. Tese had a twinning rate 25 to 30 percent higher
than short, nutritionally deprived women. Likewise, another
study ound an association between greater maternal weight
and dizygotic twinning (Reddy, 2005). Evidence acquired during and ater World War II suggests that twinning correlates
more with nutrition than with body size. Widespread undernourishment in Europe during those years was associated with
a marked all in the dizygotic twinning rate (Bulmer, 1959).
■ Superfecundation and Superfetation
Superecundation is ertilization o two ova within the same
menstrual cycle but not at the same coitus nor necessarily by
sperm rom the same male. Te latter leads to heteropaternity
(Silver, 2021).
In superetation, an interval as long as or longer than a menstrual cycle intervenes between ertilizations. Superetation is
not known to occur spontaneously and is likely due to AR
(Lantieri, 2010). Pseudo-superetation oten results rom markedly unequal growth o twins with the same gestational age.
DIAGNOSIS OF MULTIFETAL GESTATION
■ Clinical Evaluation
Early diagnosis can help with management o the associated
risks posed by twins. Accurate undal height measurement can
be an initial tool. With multiples, uterine size is typically larger
during the second trimester than that expected or a singleton. In one study, undal heights obtained between 20 and 30
weeks’ gestation averaged 5 cm greater in twins than in singletons (Rouse, 1993).
Palpation o etal parts to diagnose twins beore the third
trimester is difcult. Even then, obesity or hydramnios can hinder assessment. Palpating two etal heads strongly supports the
diagnosis. Moreover, a hand-held Doppler ultrasonic unit may
isolate two etal heartbeats i their rates are clearly distinct rom
each other and rom the mother.
Overall, however, clinical criteria alone to diagnose multietal
gestations is unreliable. In the Routine Antenatal Diagnostic Imaging with Ultrasound (RADIUS) trial, or 37 percent o women
who did not have a screening ultrasound examination, their twin
pregnancy was not diagnosed until 26 weeks’ gestation. In 13 percent o unscanned women, their multietal gestation was identied
only during their admission or delivery (LeFevre, 1993).
■ Sonography
Sonographic examination should detect practically all sets o
twins. Further, it should aim to determine etal number, estimated
gestational age, chorionicity, and amnionicity. Importantly, the
risk or many twin-specic complications varies in relation to
these (Table 48-2) (Hack, 2008; Manning, 1995).
With sonography, separate gestational sacs, i present, can
be identied early in twin pregnancy (Fig. 48-2). Subsequently,
each etal head should be seen in two perpendicular planes to
avoid mistaking a etal trunk or a second etal head. Ideally,
two etal heads or two abdomens are seen in the same image
plane to avoid scanning the same etus twice and interpreting
it as twins. Higher-order multietal gestations are more challenging to evaluate. Even in the rst trimester, identiying the
actual number o etuses and their position can be difcult.
Tis determination is especially important i selective reduction
is considered (p. 858).
In determining chorionicity, sonography’s accuracy diminishes as gestational age advances. It has a 98-percent accuracy
in the rst trimester but may be incorrect in up to 10 percent
o second-trimester examinations. Overall, chorionicity can
be correctly determined with sonography beore 24 weeks in
approximately 95 percent o cases (Emery, 2015; Lee, 2006).
Te sonographic eatures that are used to determine chorionicity vary according to gestational age. Early in the rst
trimester, the number o chorions equates to the number o
gestational sacs. A thick band o chorion that separates two
gestational sacs signals a dichorionic pregnancy, whereas monochorionic twins have a single gestational sac. I the gestation is
monochorionic diamnionic, it may be difcult to visualize the
thin intervening amnion beore 8 weeks’ gestation. I the intervening membrane is difcult to visualize, the number o yolk
sacs usually correlates with the number o amnions. However,
the number o yolk sacs as a predictor o amnionicity may not
always be accurate (Shen, 2006). Although uncommonly seen
early, cord entanglement identies a monoamnionic gestation.
At 10 to 14 weeks’ gestation, sonographic assessment o
chorionicity may be determined using our eatures. Tese are
the number o placental masses, presence o an intervening
membrane dividing the sacs, thickness o that membrane, and
etal gender (Emery, 2015; Khalil, 2016). First, two separate
placentas suggest dichorionicity. Te converse is not necessarily true, such as cases with a single used placental mass. Second, identication o a thick dividing membrane—generally
≥2 mm—supports a presumed diagnosis o dichorionicity. In
a dichorionic pregnancy, this visualized membrane is composed
o our layers—two amnions and two chorions. Also, the twin
peak sign—also called lambda or delta sign—is seen by examining the point o origin o the dividing membrane on the placental surace. Te peak appears as a triangular projection o
placental tissue extending a short distance between the layers o
the dividing membrane (Fig. 48-3).
In contrast, monochorionic pregnancies have a dividing
membrane that is so thin (generally <2 mm) that it may not
be seen until the second trimester. Te relationship between
the membranes and placenta without apparent extension o
placenta between the dividing membranes is called the T sign
(Fig. 48-4). Evaluation o the dividing membrane can establish
chorionicity in more than 99 percent o pregnancies in the rst
trimester (Maruotti, 2016; Miller, 2012). Lack o a dividing
membrane signals a monochorionic monoamnionic gestation.
Last, twins with diering gender indicates a dichorionic
(and dizygotic) gestation. Rare exceptions were described earlier (p. 839).
■ Other Diagnostic Aids
Magnetic resonance (MR) imaging may help delineate complications in monochorionic twins, including conjoined twins
(Hu, 2006). Abdominal radiography can be used i etal number in a higher-order multietal gestation is uncertain. However, radiographs generally have limited utility and lead to an
incorrect diagnosis i etuses move during the exposure.
No biochemical test reliably identies multietal gestations.
Serum levels o β-hCG and o maternal serum alpha-etoprotein (MSAFP) are usually higher, but ranges may overlap with
those o singletons.
■ Placental Examination
Careul visual examination o the placenta and membranes ater
delivery can establish zygosity and chorionicity in many cases.
First, the placenta is gently delivered to preserve the attachment o the amnion and chorion. With one common amnionic
sac or with juxtaposed amnions not separated by chorion, the
etuses are monozygotic (see Fig. 48-1). I adjacent amnions
are separated by chorion, the etuses could be either dizygotic
or monozygotic, but dizygosity is more common (Fig. 48-5).
I the neonates are o the same sex, blood typing o cord blood
samples may be helpul. Dierent genders or blood types
reects dizygosity, although the same gender or blood type in
each etus does not conrm monozygosity. Postnatal zygosity
genetic testing is available, and the benets and ethics have
been debated (Brown, 2015; Craig, 2015).
MATERNAL PHYSIOLOGICAL ADAPTATIONS
Te physiological burdens o pregnancy and likelihood o serious maternal complications are typically greater with multietal
gestations than with singleton ones. In the rst trimester and
with its higher serum β-hCG levels, multietal gestations oten
cause nausea and vomiting. In women carrying more than one
etus, blood volume expansion is greater and averages 50 to
60 percent compared with 40 to 50 percent in those with a
singleton (Pritchard, 1965). Tis augmented hypervolemia
teleologically osets blood loss with vaginal delivery o twins,
which is twice that with a single etus. Although red cell mass
also accrues, it does so proportionately less in twin pregnancies. Combined with greater iron and olate requirements, this
predisposes to anemia.
Women carrying twins also have a typical pattern o blood
pressure change. One study assessed serial blood pressures in
more than 13,000 singleton and twin pregnancies (MacDonaldWallis, 2012). As early as 8 weeks’ gestation, the diastolic blood
pressure in women with twins was lower than that with singleton pregnancies. It generally rose by a greater degree at term. An
earlier study demonstrated that this later rise was at least 15 mm
Hg in 95 percent o women with twins compared with only 54
percent o women with a singleton (Campbell, 1986).
Hypervolemia along with decreased vascular resistance has
an impressive eect on cardiac unction. In one study o 119
women with twins, cardiac output rose another 20 percent
above that in women with a singleton (Kametas, 2003). Similarly, a study o serial maternal echocardiography examinations
ound a greater elevation in cardiac output in 20 women with
uncomplicated twin pregnancies (Kuleva, 2011). Both studies
ound that the augmented cardiac output was predominantly
due to greater stroke volume rather than higher heart rate.
Vascular resistance was signicantly lower in twin gestations
throughout pregnancy compared with singleton ones. In a
study o 30 uncomplicated twin pregnancies, this same group
o investigators identied progressive diastolic dysunction
rom the rst to third trimester. Te dysunction subsequently
normalized ater delivery (Ghi, 2015).
Uterine growth in a multietal gestation is substantively
greater than in a singleton one. Te uterus and its nonetal
contents may achieve a volume o 10 L or more and weigh
in excess o 20 pounds. Especially with monozygotic twins,
excessive amounts o amnionic uid may rapidly accumulate.
In these circumstances, maternal abdominal viscera and lungs
can be appreciably compressed and displaced by the expanding
uterus. As a result, the size and weight o the large uterus may
preclude more than a sedentary existence or these women.
Rarely, maternal renal unction can become seriously
impaired, most likely as the consequence o obstructive uropathy (Jena, 1996). Hydramnios is a common associate, and therapeutic amniocentesis may provide relie or the mother, may
improve obstructive uropathy, and possibly may lower rates o
preterm labor or rupture o membranes. Unortunately, hydramnios oten develops remote rom term and rapidly reaccumulates.
PREGNANCY COMPLICATIONS
Several complications, discussed next, complicate multietal pregnancy. Others are preterm birth and discordant etal
growth, which are discussed in later sections.
■ Spontaneous Abortion and Vanishing Fetus
Among twins, monochorionic twins have signicantly higher
early etal loss rates than dichorionic pairs (D’Antonio, 2013)
wins achieved through AR may be at greater risk or abortion compared with those conceived spontaneously (Szymusik,
2012). However, among AR-conceived gestations, twins have
lower miscarriage rates than singletons (Matias, 2007).
In some cases, only one etus is spontaneously lost. As a
result, the incidence o twins in the rst trimester is much
greater than the incidence o twins at birth. Estimates suggest
that while 1 in 8 pregnancies begin multietal, only 1 in 80
births are multietal (Corsello, 2010). Sonography studies in
the rst trimester have shown that one twin dies and “vanishes”
beore the second trimester in 10 to 40 percent o all twin pregnancies (Brady, 2013; Harris, 2020). Te incidence is higher
ollowing AR conception.
A vanishing etus is more common in higher-order multiples. In one study o spontaneous reduction in 709 multietal
pregnancies beore 12 weeks’ gestation, one or more embryos
died in 36 percent o twin pregnancies, in 53 percent o triplet pregnancies, and in 65 percent o quadruplet pregnancies
(Dickey, 2002). Interestingly, in one study, ultimate pregnancy
duration and birthweight were inversely related to initial gestational sac number regardless o the nal number o etuses
at delivery (Seong, 2020). Tis eect was most pronounced in
twins who started as quadruplets. Evidence or adverse immediate and long-term eects o twin spontaneous reduction on the
remaining pregnancy is conicting (Harris, 2020; McNamara,
2016).
Spontaneous reduction o a twin gestation may aect prenatal screening results. In one study o AR-conceived gestations,
56 twin pregnancies with a single early demise and 897 singleton gestations were compared (Gjerris, 2009). First-trimester
serum marker concentrations did not dier between the groups
i the embryonic loss was identied beore 9 weeks’ gestation. I
diagnosed ater 9 weeks, the marker levels were higher and less
precise. With a vanishing twin, rst-trimester maternal serum
levels o the pregnancy-associated plasma protein A (PAPP-A)
can be elevated. Second-trimester MSAFP and dimeric inhibin
A levels also can be higher (Huang, 2015). Early loss o one
twin may also aect noninvasive prenatal testing using cell-ree
DNA (cDNA) (Chap. 17, p. 336). In one report, 15 percent
o the alse-positive results were attributed to this eect (Futch,
2013). Te development o single-nucleotide polymorphism
technology or cDNA testing holds promise in better identiying these cases (Curnow, 2015). Tus, a spontaneously reduced
abortus is ideally identied to help avoid conusion with results
rom aneuploidy and neural-tube deect screening.
■ Congenital Malformations
Te incidence o these is appreciably higher in multietal gestations compared with singleton ones. In one study, the congenital malormation rate was 406 per 10,000 twins compared
with 238 per 10,000 singletons (Glinianaia, 2008). Tis rate
in monochorionic twins was almost twice that o dichorionic
twin gestations. One large study between 1998 and 2010 ound
that twins had a 73-percent greater risk o congenital heart disease than singletons. Te occurrence risk and concordance were
substantially higher among monochorionic twins (Best, 2015;
Gijtenbeek, 2019).
From a 30-year European registry o multietal births, structural anomaly rates rose steadily rom 2.2 percent in 1987 to
3.3 percent in 2007 (Boyle, 2013). Yet, during this time, the
proportion o dizygotic twins grew by 30 percent, whereas the
proportion o monozygotic twins remained stable. Tis higher
risk o congenital malormations in dizygotic twins over time
correlated with increased availability o AR. An increase in
birth deect rates related to AR has been reported (Wen,
2020). However, i data are adjusted or maternal age or duration o subertility, the risk o congenital anomalies does not
appear to be increased by AR (Zhu, 2006).
■ Low Birthweight
Multietal gestations are more likely to be low birthweight
than singleton pregnancies due to restricted etal growth and
preterm delivery. Birthweights in twins closely paralleled those
o singletons until 28 to 30 weeks’ gestation. Tereater, twin
birthweights progressively lagged (Fig. 48-6). Beginning at 35
to 36 weeks’ gestation, twin birthweights clearly diverge rom
those o singletons. Tus, abnormal growth should be diagnosed only when etal size is less than expected or multietal
gestation. Accordingly, twin and triplet growth curves have been
developed (Kim, 2010; Odibo, 2013; Vora, 2006). o conrm
suitable growth in dichorionic pairs, we perorm sonography
every 4 weeks, starting at 16 to 20 weeks. Monochorionic twins
are imaged every 2 weeks or twin-twin transusion syndrome
(p. 848). o identiy suspected etal-growth restriction, we use
the standards o birthweight or twin gestations stratied by
placental chorionicity (Ananth, 1998).
■ Hypertension
Compared with mothers o singletons, those with multietal
gestations are more likely to develop a pregnancy-associated
hypertensive disorder, and the incidence urther rises with
advancing etal number (Day, 2005). In one analysis, 14 percent o parturients with twins developed a hypertensive disorder
o pregnancy (Aviram, 2021). No specic zygosity coners a
greater rate o these disorders (Lučovnik, 2016). However, with
twins and gestational diabetes, the preeclampsia incidence is
increased (Dave, 2021). Similar to singleton pregnancies with
hypertensive disorders, etal-growth restriction is a potential
outcome (Proctor, 2019) (Chap. 40, p. 702).
Data suggest that rising etal number and levels o antiangiogenic actor play roles in preeclampsia pathogenesis in multiples (Chap. 40, p. 694). Nonhypertensive women with twins
compared with those with singletons have higher levels o the
antiangiogenic actors soluble ms-like tyrosine kinase 1 (sFlt-1)
and soluble endoglin and lower levels o proangiogenic placental
growth actor (PlGF) (Faupel-Badger, 2015). In women with
twins and preeclampsia in one study, both sFlt-1 levels and
sFlt-1/PlGF ratios rose and PlGF concentrations declined compared with normotensive women with twins (Rana, 2012). Aspirin therapy is recommended to help prevent preeclampsia and
is discussed with prenatal care (p. 853).
■ Longterm Infant Development
In cohort studies evaluating twin and singletons, academic
perormance is similar and is unaected by AR-assisted conception (Christensen, 2006; Spangmose, 2017). For mono- or
dichorionic pairs born preterm, neurodevelopment at 2 years
was reported to be similar (osello, 2021).
However, among normal-birthweight neonates, the cerebral
palsy (CP) risk is twoold higher among multietal pregnancies compared with singletons. In term and preterm neonates
rom 12 large CP registries, the overall CP rate was 1.6 per
1000 singleton live births and 7 per 1000 live births rom multietal pregnancies (Perra, 2021). In this and another registry
study, the higher CP rate in multietal gestations was attributed mainly to prematurity (Sellier, 2021) (Chap. 34, p. 620).
Fetal-growth restriction, congenital anomalies, twin-twin trans-
usion syndrome, and etal demise o a co-twin are other potential contributors.
UNIQUE FETAL COMPLICATIONS
Several unique complications arise in multietal pregnancies.
Tese are described in twins but also can be ound in higherorder multietal gestations. Most etal complications attributed
to the twinning process itsel are seen with monozygotic twins.
Teir pathogenesis is best understood ater reviewing the possibilities shown in Figure 48-1.
■ Monoamnionic Twins
Tese twins make up less than 1 percent o all twin pregnancies
and 4 percent o monochorionic pairs (Chitrit, 2021; Sebire,
2000). However, they have high rates o many signicant complications.
Fetal loss rates in monoamnionic twins are substantial.
Among etuses alive beore 16 weeks’ gestation, less than hal
survive to the neonatal period. Fetal abnormalities and spontaneous miscarriage contribute to most losses (Preumo, 2015).
A high perinatal death rate is attributable to preterm birth,
twin-twin transusion syndrome, cord entanglement, birthweight discordance, and congenital anomalies (Buca, 2020;
Saccone, 2020).
Congenital anomaly rates in monoamnionic twins reach
18 to 28 percent (Post, 2015). Concordance o anomalies is
ound in only approximately one quarter o cases. Because o
the greater risk o cardiac anomalies, etal echocardiography
is indicated. O note, monoamnionic twins are by denition
monozygotic and thus presumed to be genetically identical,
except in rare cases (Zwijnenburg, 2010). Interestingly, the
risk or Down syndrome in each etus o the monozygotic pair
is similar to or lower than the risk in maternal age-matched
singletons (Sparks, 2016).
Twin-twin transusion syndrome rates in monoamnionic
twins are lower than in monochorionic diamnionic pregnancies
(Murgano, 2020). Tis may be due to the near universal presence in monoamnionic twins o arterioarterial anastomoses,
which are presumed to be protective (Hack, 2009). Nonetheless, twin-twin transusion syndrome surveillance is recommended and described in that section.
Umbilical cord entanglement is a requent event (Fig. 48-7).
Although color-ow Doppler sonography can help diagnose
entanglement, actors that lead to pathological umbilical vessel
constriction are unknown. Fetal death rom cord entanglement
is thus unpredictable, and monitoring or this is relatively ine-
ective. Moreover, studies o inpatient or outpatient etal surveillance o monoamnionic twin pregnancies show conicting
data. Heyborne and coworkers (2005) reported no stillbirths in
43 women admitted at 26 to 27 weeks’ gestation or daily etal
surveillance. Conversely, 13 etuses died in 44 women managed
as outpatients. In the MONOMONO Working Group (2019)
study o 195 women with these twins, the etal death rate was
3.3 percent in 75 mothers admitted at 24 to 29 weeks’ gestation compared with 10.8 percent in the 120 gravidas monitored outpatient. Notably, no etal deaths occurred ater 32
weeks’ gestation in either group. Last, Van Mieghem and associates (2014) ound the risk o “potentially preventable death”
was not signicantly dierent in women with inpatient (2.1
percent) or outpatient (4.7 percent) etal surveillance.
In the United States, mothers with monoamnionic twins are
oten admitted at 24 to 28 weeks’ gestation to begin 1 hour o
daily etal heart rate monitoring (American College o Obstetricians and Gynecologists, 2021b). Optimal surveillance is
unclear and may include nonstress testing or biophysical prole
assessment (Chap. 20, p. 386). Betamethasone, discussed later,
is considered to promote pulmonary maturation (p. 855). I
etal testing remains reassuring and no other intervening indications arise, cesarean delivery is perormed at 320/7 to 340/7
weeks’ gestation. o help manage these many care items, the
Society or Maternal-Fetal Medicine (2020) oers a prenatal
checklist.
■ Unique and Aberrant Twinning
Several aberrations in monozygotic twinning result in a spectrum o etal malormations. Tese are traditionally ascribed
to incomplete splitting o an embryo into two separate twins.
However, they theoretically may result rom early secondary
usion o two separate embryos. Tese separated embryos are
either symmetrical or asymmetrical. Te spectrum o asymmetrical twinning includes external parasitic twins, etus-in-etu,
and twin reversed-arterial-perusion (RAP) sequence, which
is described later (p. 850).
Conjoined Twins
Joining o the twins may begin at either pole and produce characteristic orms depending on which body parts are joined or
shared (Fig. 48-8) (Spencer, 2000). Te requency o conjoined
twins has a prevalence o 1.5 in 100,000 births, and thoracopagus is the most common type (Mutchinick, 2011).
Conjoined twins can be identied using sonography in
the rst trimester (Chen, 2011). Tis provides an opportunity or parents to decide whether to continue the pregnancy.
During sonographic interrogation, etal poles are seen to be
closely associated and do not change relative position rom one
another (Fig. 48-9). Other clues are more than three vessels
in the umbilical cord, ewer limbs than expected, spine hyper-
exion, bid etal pole, and increased nuchal thickness. Treedimensional ultrasound, color Doppler, and MR imaging are
valuable adjuncts to clariy shared organs (Baken, 2013).
Postnatal surgical separation may be successul i essential
organs are not shared. Conjoined twins may have discordant
structural anomalies that urther complicate decisions about
whether to continue the pregnancy. Consultation with a pediatric surgeon oten assists with parental decisions.
Viable-aged conjoined twins should be delivered by cesarean.
For pregnancy termination, however, vaginal delivery is possible
because the union is oten pliable. Still, dystocia is common,
and vaginal delivery may be traumatic to the uterus or cervix.
External Parasitic Twins and Fetus-in-fetu
Attached to a relatively normal twin, an external parasitic twin
is a grossly deective etus or merely etal parts. It usually consists o externally attached supernumerary limbs, oten with
some viscera. Classically, however, a unctional heart or brain
is absent. Parasitic twins are believed to derive rom a dead
deective twin, whose surviving tissue attaches to and receives
vascular support rom the normal co-twin (Spencer, 2001). In
one study, parasitic twins accounted or 4 percent o all conjoined twins (Mutchinick, 2011).
With etus-in-etu, one embryo may enold early within its
co-twin and mainly intraabdominally. Normal development
o this rare parasitic twin usually arrests in the rst trimester.
Tus, normal spatial arrangement and many organs are lacking.
Classically, vertebral or axial bones are ound in the etiorm
mass, whereas a heart and brain are absent. Tese masses are
thought to be a monozygotic, monochorionic diamnionic twin
gestation and are typically supported by large parasitic vessels
to the host (McNamara, 2016).
■ Monochorionic Twins and
Vascular Anastomoses
All monochorionic placentas likely share some anastomotic
connections. With rare exceptions, anastomoses between twins
are unique to monochorionic twin placentas. However, the
number, size, and direction o these seemingly haphazard connections vary markedly (Fig. 48-10). In one analysis o more
than 200 monochorionic placentas, the median number o
anastomoses was eight (Zhao, 2013).
Anastomoses may be artery-to-artery, vein-to-vein, or arteryto-vein communications and are located on the chorioni
surace o the placenta. In contrast to these supercial connections, deep artery-to-vein communications can extend through
the capillary bed o a given villus (Fig. 48-11). Tese deep arteriovenous anastomoses create a common villous compartment
or “third circulation” that has been identied in approximately
hal o monochorionic twin placentas.
Whether these anastomoses are dangerous to either twin
depends on the degree to which they are hemodynamically
balanced. In those with signicant pressure or ow gradients,
a shunt will develop between etuses. Tis chronic etoetal
transusion may result in several clinical syndromes that include
twin-twin transusion syndrome (TTTS), twin anemia–polycythemia sequence (TAPS), and twin reversed-arterial-perusion
(TRAP) sequence.
Twin-Twin Transfusion Syndrome
In this syndrome, blood is transused rom a donor twin to its
recipient sibling such that the donor may eventually become
anemic and its growth may be restricted. In contrast, the recipient becomes polycythemic and may develop circulatory overload
with heart ailure maniest as hydrops. Classically, the donor
twin is smaller and pale, and its recipient sibling is larger and
has volume excess. As a result, the recipient neonate may suer
hyperviscosity and occlusive complications. Polycythemia in the
recipient twin may also lead to severe hyperbilirubinemia and
kernicterus (Chap. 33, p. 606). S complicates 10 to 15
percent o monochorionic twins (Marwan, 2019).
Chronic S results rom unidirectional ow through
deep arteriovenous anastomoses. Deoxygenated blood rom a
donor placental artery is pumped into a cotyledon shared by
the recipient (see Fig. 48-11). Once oxygen exchange is completed in the chorionic villus, the oxygenated blood leaves the
cotyledon via a placental vein o the recipient twin. Unless compensated—typically through supercial arterioarterial anastomoses—this unidirectional ow leads to volume depletion in
the donor and volume excess in the recipient (Lewi, 2013).
S typically presents in midpregnancy when the donor
etus becomes oliguric rom hypovolemia and decreased
renal perusion (Society or Maternal-Fetal Medicine, 2013).
Tis etus develops oligohydramnios, and the recipient etus
develops severe hydramnios, presumably due to increased urine
production rom hypervolemia. Virtual absence o amnionic
uid in the donor sac prevents etal motion, giving rise to the
descriptive term stuck twin or polyhydramnios–oligohydramnios
syndrome—“poly–oli.” Tis amnionic uid imbalance is associated with growth restriction, contractures, and pulmonary
hypoplasia in the donor twin, and premature rupture o the
membranes and heart ailure in the recipient.
Fetal Brain Damage. Cerebral palsy, microcephaly, porencephaly, and multicystic encephalomalacia are serious complications associated with placental vascular anastomoses. Te exact
pathogenesis o neurological damage is not ully understood
but is likely caused by ischemic necrosis leading to cavitary
brain lesions. In the donor twin, ischemia results rom hypotension, anemia, or both. In the recipient, ischemia develops
rom blood pressure instability and episodes o proound hypotension (Lopriore, 2011). Cerebral lesions may also be associated with preterm delivery (Chap. 34, p. 618). In one review
o 315 liveborn etuses rom pregnancies with S, cerebral
abnormalities were ound in 8 percent (Quarello, 2007).
I one twin o an aected pregnancy dies, blood is acutely
transused rom high-pressure vessels o the living twin through
anastomoses to low-resistance vessels o the dead twin. Tis
leads rapidly to hypovolemia and possible ischemic antenatal
brain damage in the survivor (Fusi, 1990, 1991). A less likely
cause is emboli o thromboplastic material originating rom the
dead etus.
Te acuity o hypotension ollowing the death o one twin
with S makes successul intervention or the survivor nearly
impossible. Even with delivery immediately ater a co-twin
demise is recognized, the hypotension that occurs at the moment
o death has likely already caused irreversible brain damage
(Langer, 1997; Wada, 1998). As such, immediate delivery is not
considered benecial in the absence o another indication.
Diagnosis. S is diagnosed based on two sonographic criteria (Society or Maternal-Fetal Medicine, 2013). First, a monochorionic pregnancy is identied. Second, hydramnios dened
by a largest vertical pocket >8 cm in one sac and oligohydramnios dened by a largest vertical pocket <2 cm in the other
twin is ound. Growth discordance or growth restriction may
be ound with S, but these per se are not considered diagnostic criteria.
Organizations that include the American College o Obstetricians and Gynecologists (2021b), Society or Maternal–Fetal
Medicine (2013), and North American Fetal Terapy Network
recommend sonographic surveillance o pregnancies at risk or
S (Emery, 2015). Tese examinations begin at approximately 16 weeks’ gestation, and studies are perormed every 2
weeks. In one study o 108 monochorionic twin pairs, a sonographic evaluation interval >2 weeks was associated with a
higher Quintero stage at the time o S diagnosis ( Torson,
2011).
Once identied, S is typically classied by the Quintero (1999) staging system (Fig. 48-12):
• Stage I—discordant amnionic uid volumes as described in
the earlier paragraph, but urine is still visible sonographically
within the bladder o the donor twin
• Stage II—criteria of stage I, but urine is not visible within the
donor bladder
• Stage III—criteria of stage II and abnormal Doppler studies
o the umbilical artery, ductus venosus, or umbilical vein
• Stage IV—ascites or frank hydrops in either twin
• Stage V—demise of either fetus.
In addition to these criteria, cardiac unction o the recipient
twin also declines with S (Wohlmuth, 2018). Although
etal echocardiographic ndings are not part o the Quintero
staging system, many centers routinely perorm etal echocardiography or S. Teoretically, earlier diagnosis o etal
cardiomyopathy may identiy pregnancies that would benet
rom early intervention (Votava-Smith, 2015). One measure,
the myocardial perormance index (MPI), is a Doppler index o
each ventricle’s unction. Scoring systems that include cardiac
unction have been developed, but their useulness to predict
outcomes remains controversial (Miller, 2021b; Society or
Maternal-Fetal Medicine, 2013).
Management and Prognosis. Te prognosis or multietal gestations complicated by S is related to Quintero stage and
gestational age at presentation. Some stage I cases remain stable
or regress without intervention. However, 60 percent progress
(Emery, 2016; Stirnemann, 2021). Conversely, outcomes in
those identied at stage III or higher are much worse, and the
perinatal loss rate is 70 to 100 percent without intervention
(Miller, 2021b; Society or Maternal-Fetal Medicine, 2013).
Several therapies available or S include laser ablation o
vascular placental anastomoses, amnioreduction, and selective
eticide. Similar to amniocentesis (Chap. 17, p. 344), amnioreduction is needle drainage o excess amnionic uid.
Tese techniques have been evaluated in randomized trials.
Te Euroetus trial included 142 women with severe S
diagnosed beore 26 weeks’ gestation. Participants were randomly assigned to laser ablation o vascular anastomoses or to
serial amnioreduction (Senat, 2004). Laser ablation yielded a
76-percent rate o survival to age 6 months or at least one
twin compared with a 51-percent rate with amnioreduction.
Analyses o most randomized studies conrm better neonatal
outcomes with laser therapy compared with selective amnioreduction (Marwan, 2019; Roberts, 2008; Rossi, 2008, 2009).
However, evaluation o twins rom the Euroetus trial through
6 years o age did not demonstrate an additional survival benet beyond 6 months or improved neurological outcomes in
those treated with laser (Salomon, 2010). At this time, laser
ablation o anastomoses is preerred or severe S (stages
II–IV). Optimal therapy or stage I disease is controversial, and
laser ablation or expectant surveillance is an option (Emery,
2015; Stirnemann, 2021).
O techniques, an ablation method may laser individual
anastomoses or may ablate the entire the vascular equator. Tis
equator is the border between twin vasculature on the chorionic
surace. Equatorial ablation, which is reerred to as a Solomon
technique, reduces the likelihood o S recurrence. Tis
method also lowers rates o APS, which can be a laser-therapy
sequelae and described next (Slaghekke, 2014).
Ater laser therapy, close ongoing surveillance is necessary.
Weekly ultrasound and Doppler studies are recommended
(Marwan, 2019; Miller 2021b). Tese evaluations monitor
etal growth, amnionic uid volumes, placental unction, and
anemia. Delivery timing is usually inuenced by S recurrence, etal-growth restriction, or by abnormal Doppler velocimetry values, which reect poor placental unction (Chap. 47,
p. 830).
Another treatment option, selective etal reduction, has generally been considered i severe amnionic uid and growth disturbances develop beore 20 weeks. In such cases, both etuses
typically will die without intervention. Any substance injected
into one twin may aect the other twin because o shared circulations. Tus, or the etus chosen or reduction, eticidal
methods aim to occlude the umbilical vein or umbilical cord.
Radiorequency ablation, etoscopic ligation, or coagulation
with laser, monopolar, or bipolar energy are options (Challis,
1999; Chang, 2009; Parra-Cordero, 2016). Even ater these
procedures, however, risks to the remaining etus are still appreciable (Rossi, 2009). Early termination o the entire pregnancy
is yet another option.
Twin Anemia–Polycythemia Sequence
Chronic etoetal transusion underlies this orm, which is
characterized by signicant hemoglobin dierences between
donor and recipient twins. Similar to S, the donor twin
in spontaneous APS is anemic and usually smaller than the
recipient twin, which is polycythemic. However, APS lacks
the discrepancies in amnionic uid volumes typical o S.
Spontaneous APS can develop at any gestational age and
complicates 1 to 6 percent o monochorionic twins (Marwan,
2019). Iatrogenic APS develops in up to 13 percent o pregnancies ater laser ablation o the placenta and usually develops
within 5 weeks o a procedure (Lewi, 2013; ollenaar, 2021).
In iatrogenic APS, the ormer S recipient twin usually
becomes anemic, whereas the ormer donor becomes polycythemic (ollenaar, 2016).
Sonographically measuring blood ow velocity in the etal
middle cerebral artery (MCA) can accurately identiy etal
anemia (Chap. 14, 262). Antenatally, APS is diagnosed by
discordant MCA peak systolic velocity (PSV) values between
twins. Specically, an MCA-PSV value that is >1.5 multiples
o the median (MoM) in the donor twin and <1.0 MoM in
the recipient twin (Society or Maternal-Fetal Medicine, 2013).
Others have suggested alternative threshold values (Khalil,
2020; ollenaar, 2019).
A staging system or APS has been proposed, and higher
stage is associated with increased perinatal mortality rate
( Slaghekke, 2010; ollenaar, 2021). Screening or spontaneous APS is controversial, as improvements in perinatal outcomes have not been demonstrated (Khalil, 2016; Society or
Maternal-Fetal Medicine, 2013).
Management options include expectant care, delivery, laser
surgery, intrauterine transusion, selective eticide, and pregnancy termination. A clinical trial evaluating etoscopic laser
therapy or APS is currently ongoing. Antenatal surveillance
mirrors that just described or S. Again, delivery timing is
usually inuenced by worsening etal growth or by abnormal
Doppler velocimetry values. Postnatal treatment oten requires
blood transusion or the donor twin and partial exchange
transusion o the recipient twin.
Twin Reversed-arterial-perfusion Sequence
Also known as acardiac twinning, this rare, serious complication o monochorionic multietal gestation has an estimated
incidence o 1 case in 35,000 births. With classic RAP
sequence, the twin pair is a normally ormed donor twin that
shows eatures o heart ailure and a grossly malormed recipient twin that lacks a heart (acardius) and other structures. In
one theory, the RAP sequence is caused by a large artery-toartery placental shunt, oten also accompanied by a vein-tovein shunt (Fig. 48-13). Within the single, shared placenta,
arterial perusion pressure o the donor twin exceeds that in the
recipient twin. Te recipient thus receives reversed blood ow
containing deoxygenated arterial blood rom its co-twin (Lewi,
2013). Tis “used” arterial blood reaches the recipient twin
through its umbilical arteries and preerentially goes to its iliac
vessels. Tus, only the lower body is perused, and disrupted
growth and development o the upper body ollows. In these
cases, ailed head growth is called acardius acephalus; a partially
developed head with identiable limbs is called acardius myelacephalus; and ailure o any recognizable structure to orm is
acardius amorphous, which is shown in Figure 48-14. Because
o this vascular connection, the normal donor twin must not
only support its own circulation but also must pump blood to
and through the underdeveloped acardiac recipient. Tis may
lead to cardiomegaly and high-output heart ailure in the donor
twin (Fox, 2007; Marwan, 2019).
In the past, the donor-twin mortality rate exceeded 50 percent. Tis stemmed largely rom complications o prematurity
or rom a prolonged high-output state leading to cardiac ailure
(Dashe, 2001). Risk is directly related to size o the acardiac
twin. When the acardiac twin is large, treatment is generally
oered. Radiorequency ablation (RFA) is the preerred modality o therapy, and contemporary reports now suggest improved
perinatal outcomes. Te North American Fetal Terapy Network reviewed their experiences with 98 cases rom 1998 to
2008 in which RFA o the umbilical cord was perormed.
Median gestational age at delivery was 37 weeks’ gestation, and
80 percent o donor neonates survived (Lee, 2013).
■ Hydatidiform Mole with Coexisting
Normal Fetus
Tis rare gestation contains one normal etus, and its co-twin
is a complete molar pregnancy. It must be dierentiated rom
a partial molar pregnancy, which is a singleton, triploid etus
and its placenta composed o molar tissue (Fig. 13-4, p. 238).
Diagnosis in the rst hal o pregnancy is common. Sonographically, a normal-appearing twin is accompanied by its
co-twin, which is a large placenta containing multiple small
anechoic cysts. Oten, these pregnancies are terminated, but
pregnancy continuation is increasingly adopted. Te live birth
rate rom one review was 50 percent (Zilberman, 2020). Te
risk o gestational trophoblastic neoplasia (GN), which is a
malignant sequelae o hydatidiorm mole, is similar whether
the pregnancy is terminated or not (Massardier, 2009; Sebire,
2002). Given the limited number o cases, robust data or
rm recommendations are lacking. Discussed in Chapter 13
(p. 237), complications o expectant management include
vaginal bleeding, hyperemesis gravidarum, thyrotoxicosis, and
early-onset preeclampsia. Many o these complications result
in preterm birth with its attendant adverse perinatal outcomes.
Logically, close antepartum and postpartum surveillance is
needed or those continuing the pregnancy. Postpartum GN
surveillance is essential and described in Chapter 13 (p. 240).
DISCORDANT GROWTH
Fetal size inequality develops in approximately 15 percent o
twin gestations (Miller, 2012). Generally, as the weight dierence within a twin pair grows, the perinatal mortality rate rises
proportionately. Earlier discordancy and monochorionicity
pose increased mortality risks or the smaller twin. Specically,
with discordant growth identied at or beore 20 weeks’ gestation in studies, 8 to 15 percent o the growth-restricted etuses
die (Couck, 2020; Curado, 2020; D’Antonio, 2018).
■ Pathogenesis
Te etiology o growth discordance in monochorionic
twins likely diers rom that in dichorionic twins. First, in
monochorionic twins, the single placenta is not always equally
shared, and this leads to higher discordant growth rates than
in dichorionic pairs (Fig. 48-15). In cases with S, discordancy in monochorionic twins is usually attributed to placental
anastomoses that cause a perusion imbalance between twins.
Reduced perusion o the donor twin can diminish placental
and etal growth (Lewi, 2013). Last, monochorionic twins at
times can be discordant in size because they are discordant or
structural anomalies.
In dichorionic twins, discordancy may result rom various
actors. Dizygotic etuses may have dierent genetic growth
potential, especially i they are o opposite genders. Second,
because the placentas are separate and require more implantation space, one placenta might have a suboptimal implantation
site. Additionally, umbilical cord abnormalities such as velamentous insertion, marginal insertion, or vasa previa may play
a role (Chap. 6, p. 114). One study showed that the incidence
o severe discordancy is twice as great in triplets as it is in twins
(Bagchi, 2006). Tis supports the view that in-utero crowding promotes multietal growth restriction. Placental pathology
may be contributory. In one study o 668 twin placentas, a
strong relationship between histological placental abnormalities
and birthweight discordancy was observed in dichorionic but
not monochorionic twin pregnancies (Kent, 2012).
■ Diagnosis
Antenatal size discordancy between twins can be best determined sonographically. Crown-rump length dierences are not
reliable predictors or birthweight discordance (Miller, 2012).
Tus, most discordancy surveillance begins ater the rst trimester. One common method uses sonographic etal biometry
to compute an estimated weight or each twin (Chap. 15, p.
274). Percent discordancy is then calculated as the weight o
the larger twin minus the weight o the smaller twin, divided by
the weight o the larger twin. Te American College o Obstetricians and Gynecologists (2021b) denes discordance as an
estimated etal weight dierence >20 percent.
■ Management
Te risk or adverse perinatal outcomes in the setting o growth
discordance remains controversial. Some suggest that twin gestations with discordant growth between two etuses that are
appropriately grown or gestational age are not at increased risk
or adverse outcomes (American College o Obstetricians and
Gynecologists, 2021b; Appleton, 2007). Others have shown an
increased risk or adverse etal outcomes but not or poor neonatal outcomes (Amaru, 2004; D’Antonio, 2018). However,
accumulated data suggest that weight discordancy exceeding
25 to 30 percent most accurately predicts an adverse perinatal outcome (Cohen, 2001; Chen, 2019). At Parkland Hospital, twin weight-discordancy values rom 1370 delivered twin
pairs were stratied by percentage increments (Hollier, 1999).
Te incidence o respiratory distress syndrome, intraventricular hemorrhage, seizures, periventricular leukomalacia, sepsis,
and necrotizing enterocolitis rose directly with the percentage
o weight discordancy. Rates o these conditions grew substantially i discordancy exceeded 25 percent. Te relative risk o
etal death increased signicantly to 5.6 i weight discordancy
was >30 percent and rose to 18.9 i >40 percent.
Nonstress testing and biophysical prole assessment have
all been recommended in management o twin growth discordancy. I signicant discordancy is identied in a monochorionic twin pair, umbilical artery Doppler studies in the smaller
etus may help guide management (Gratacós, 2007). Data are
limited to establish the optimal timing o delivery o twins with
size discordancy alone. At advanced gestational ages, delivery
can be pursued.
Selective Fetal-growth Restriction
Restricted growth o one twin etus is termed selective etalgrowth restriction (sFGR) and usually develops late in the second
and early third trimester. Some diagnose sFGR i the abdominal
circumerence (AC) measurement dierence exceeds 20 mm or
i etal-growth discordance is >20 percent (Khalil, 2019). I
sFGR is diagnosed, weekly testing o etal well-being, evaluation o amnionic uid volume, and umbilical artery Doppler
velocimetry are undertaken. Investigators have correlated Doppler results with placental ndings and with the degree o sFGR
to predict etal outcome (Gratacós, 2012). Tese correlations
have yielded categories o sFGR.
• Type I shows positive end-diastolic ow, a smaller degree of
weight discordance, and a relatively benign clinical course.
• Type II displays persistently absent end-diastolic ow in the
smaller twin and carries a high risk o deterioration and demise.
• Type III has intermittently absent or reversed end-diastolic
ow. Because o large artery-to-artery anastomoses associated
with the placentas in this last category, type III is associated
with a lower risk o deterioration than type II. In all evaluated
cases, unequally shared placenta was noted to some degree.
Serial evaluation o etal growth is perormed every 3 weeks
(Chap 47, p. 829). At Parkland Hospital, daily inpatient surveillance is undertaken in women with etal-growth restriction
in one twin or with twin discordancy exceeding 25 percent in
the setting o a twin with restricted growth.
FETAL DEMISE
At any time during multietal pregnancy, one or more etuses
may die, either simultaneously or sequentially. Most causes
stem rom complications o etal anomaly or chorionicity.
Related to the latter, compared with dichorionic twins, monochorionic pairs suer higher rates o sFGR, S, or APS
rom unequal vascular anastomoses, and monoamnionic pairs
can die rom cord entanglement.
When death is early, it may maniest as a vanishing twin (p.
843). For the survivor, the risk o death ater the rst trimester
is not increased, and the pregnancy requires no additional surveillance or this specic indication. In a slightly more advanced
gestation, the dead etus may be compressed appreciably—etus
compressus, or it may be attened remarkably through desiccation—
etus papyraceus (Fig. 48-16).
In gestations past 20 weeks, the stillbirth rate in twins in the
United States was 1.4 percent o live births and exceeded the
0.6-percent rate in singletons (MacDorman, 2015). Ater a single
intrauterine etal demise (sIUFD) in these later gestations, risks to
the remaining co-twin include death, preterm birth, and neurological injury (D’Antonio, 2017; Mackie, 2019). With the last, acute
hypovolemia rom volume shits within placental anastomoses
occurs immediately, and neurological injury is not preventable.
Related to co-twin death, monochorionic diamnionic twins
with an sIUFD were 16 times more likely to experience death
o the co-twin than were dichorionic twins with an sIUFD in
one large series (Morikawa, 2012). In most studies, the rate
o co-twin demise ater sIUFD declines ater 26 to 28 weeks’
gestation and with advancing gestational age (Mackie, 2019;
McPherson, 2012; Southwest Tames Obstetric Research
Collaborative, 2012; Wood, 2014).
Te rate o preterm birth ollowing sIUFD is increased but
is similar between mono- and dichorionic twins. Advancing
rates are seen at later gestational ages (D’Antonio, 2017; Ong,
2006).
Te neurological prognosis or a surviving co-twin is inuenced by chorionicity. In one review o diamnionic twin
pregnancies complicated by sIUFD beore 34 weeks, neurodevelopmental morbidity was 29 percent in monochorionic
twins and was threeold higher than in dichorionic gestations
(Mackie, 2019). In other series with sIUFD ater 34 weeks,
the likelihood o neurological decits was essentially the same
between monochorionic and dichorionic twin pregnancies
(Hillman, 2011).
Because o this signicant neurological risk, i one etus o
a monochorionic twin gestation dies ater the rst trimester
but beore viability, pregnancy termination can be considered
(Blickstein, 2013). Ater viability, ultimate delivery timing
balances the risk o prematurity against co-twin demise. Early
management emphasizes maternal and co-twin saety and gestation prolongation (American College o Obstetricians and
Gynecologists, 2021b). Antenatal corticosteroids or survivor
lung maturity can be considered. For sIUFD occurring ater 34
weeks’ gestation, delivery is reasonable (Spong, 2011).
Last, the death o one etus could theoretically trigger coagulation deects in the mother. Only a ew cases o maternal coagulopathy ater a single etal death in a twin pregnancy have been
reported. Tis is probably because the surviving twin is usually
delivered within a ew weeks o the demise (Eddib, 2006).
PRENATAL CARE
At Parkland Hospital clinics, women with multietal gestations
are seen every 2 weeks beginning at 22 weeks’ gestation. One
imperative is preterm delivery prevention, and a digital cervical examination is perormed at each visit to screen or cervical
shortening or dilation.
Te maternal diet should provide additional requirements
or calories, protein, minerals, vitamins, and essential atty acids.
Te Institute o Medicine (IOM) (2009) recommends a 37- to
54-lb weight gain or women with twins and a normal body
mass index. Te daily recommended augmented caloric intake
or women with twins is 40 to 45 kcal/kg/d. Diets contain 20
percent protein, 40 percent carbohydrate, and 40 percent at
divided into three meals and three snacks daily. Based on upperintake levels rom the Food and Nutrition Board o the IOM,
one review endorsed supplementation o calcium, magnesium,
zinc, and vitamins C, D, and E (Goodnight, 2009).
Discussed in Chapter 40 (p. 705), low-dose aspirin prophylaxis is recommended in women at high risk o preeclampsia,
which includes those with a multietal pregnancy (LeFevre,
2014). An 81-mg oral daily dose is initiated between 12 and
28 weeks’ gestation and is continued until delivery (American
College o Obstetricians and Gynecologists, 2020b).
Prenatal screening or etal aneuploidy is carried out as
described in Chapter 17 (p. 334). Tus, the combined test
or secondary screening is done and interpreted with value
thresholds set or multietal pregnancies. Noninvasive prenatal
screening with cDNA also is acceptable (American College o
Obstetricians and Gynecologists, 2020d; Judah, 2021; Khalil,
2021). Interpretation caveats or those with a vanishing twin
were described earlier (p. 844).
o screen or structural anomalies, a midpregnancy anatomic sonography survey is perormed. In addition, monochorionic twins undergo echocardiography because o an increased
risk or cardiac anomalies (Bahtiyar, 2007). As noted earlier
(p. 844), etal growth is assessed with sonography. Amnionic
uid volume is quantied using the deepest vertical pocket in
each sac. A measurement <2 cm is considered oligohydramnios, and a measurement >8 cm is considered hydramnios
(Duryea, 2017; Hernandez, 2012).
win pregnancies are at increased risk or stillbirth compared to singleton gestations. Te intrauterine etal death rate
is two- to threeold higher in monochorionic pregnancies
compared with dichorionic gestations (Lee, 2008; Morikawa,
2012). Cheong-See and colleagues (2016) reported a higher
stillbirth rate in monochorionic gestations compared with
dichorionic gestations beyond 34 weeks’ gestation. Outpatient
weekly nonstress test or biophysical prole may help lower the
rate o intrauterine death in twin gestations (Booker, 2015;
Burgess, 2014). Weekly antenatal surveillance is undertaken at
36 weeks’ gestation or uncomplicated dichorionic pregnancies
and at 32 weeks’ gestation or uncomplicated monochorionic
gestations (American College o Obstetricians and Gynecologists, 2021a). However, all antenatal testing schemes have high
alse-positive rates. In cases o abnormal testing in one twin and
normal results in another, iatrogenic preterm delivery remains a
major concern. Tereore, patients are careully counseled prior
to initiating antenatal testing in twins.
PRETERM BIRTH
Gestational length shortens and preterm birth risk rises with
accruing etal number. In 2019 in the United States, the
preterm birth rate in twins and triplets was 60 and 98 percent,
respectively (Martin, 2021). One review showed that approximately 60 percent o preterm births in twins are indicated, a
third result rom spontaneous labor, and 10 percent ollow
preterm prelabor rupture o membranes (PPROM) (Chauhan,
2010).
O risks, preterm birth rates vary with chorionicity and are
higher with monochorionic compared with dichorionic twins.
Other implicated actors are prior preterm birth, adolescence,
nulliparity, obesity, and diabetes (Marleen, 2018, 2021).
■ Prediction
One goal o multietal prenatal care is to identiy women likely
to experience preterm delivery. o and associates (2006) sonographically measured cervical length in 1163 twin pregnancies
at 22 to 24 weeks’ gestation. Rates o preterm delivery beore
32 weeks were 66 percent in those with cervical lengths o 10
mm; 24 percent or lengths o 20 mm; and only 1 percent or
lengths o 40 mm. In one review, a cervical length <20 mm
was most accurate or predicting birth beore 34 weeks’ gestation. Te specicity was 97 percent, and positive likelihood
ratio was 9.0 (Conde-Agudelo, 2010). Notably, a closed
internal os by digital examination predicted postponed delivery
equally well as the combination o a normal sonographically
measured cervical length and negative etal bronectin test
result ( McMahon, 2002).
Unortunately, cervical length assessment in twin pregnancies has not led to improved outcomes (Gordon, 2016). Te
Society or Maternal–Fetal Medicine (2016) recommends
against routine cervical length screening in multietal gestations.
Fetal bronectin levels also may predict preterm birth (Marleen, 2020). Again, neonatal outcomes are not improved in
twins, and this screening tool is not recommended (American
College o Obstetricians and Gynecologists, 2021b).
■ Prevention
Bed Rest
In general, most strategies or preterm birth prevention are
ineective or singleton and multietal pregnancies ( American
College o Obstetricians and Gynecologists, 2021b). O
options, bed rest with or without hospitalization does not
prolong multietal pregnancy. In one metaanalysis, the practice did not reduce the risk o preterm birth (da Silva Lopes,
2017). At Parkland Hospital, elective hospitalization was
compared with outpatient management, and no advantages
were ound (Andrews, 1991). Importantly, however, almost
hal o women managed as outpatients required admission or
specic indications such as hypertension or threatened preterm delivery.
Limited physical activity, early work leave, more requent
health-care visits, serial sonographic examinations, and structured maternal education regarding preterm delivery risks have
been advocated to reduce preterm birth rates in women with
multiple etuses. However, little evidence suggests that these
practices substantially change this outcome.
Prophylactic Tocolysis
Tis has not been studied extensively in multietal pregnancies.
In one review o prophylactic oral beta-mimetic therapy that
included 374 twin pregnancies, treatment did not reduce the
rate o twins delivering beore 37 or beore 34 weeks’ gestation
(Yamasmit, 2015). In light o the Food and Drug Administration warning against the use o oral terbutaline because o
maternal side eects, the prophylactic use o beta-mimetic
drugs in multietal gestations seems unwarranted.
Progesterone Therapy
17-alpha-hydroxyprogesterone caproate (17-OHPC) injections are
not eective or multietal gestations (Caritis, 2009; Combs,
2011; Rouse, 2007). Moreover, women carrying twins and
having a cervical length <36 mm or a length <25 mm did not
benet despite their greater risk or preterm birth (Durnwald,
2010; Senat, 2013).
Vaginal micronized progesterone in various ormulations in
randomized studies has also proved ineective or preterm
birth prevention beore 34 weeks in a general twin population
(Norman, 2009; Rode, 2011). For women with a short cervix,
subgroup analysis and a metaanalysis show conicting results
regarding benets (D’Antonio, 2021; Klein, 2011; Rehal,
2021). At Parkland Hospital, current management o multietal
gestations does not typically include any progesterone therapy.
Cervical Cerclage
Prophylactic cerclage does not improve perinatal outcome in
women with twin pregnancies. Studies have included women
with and without a short cervix (D’Antonio, 2021; Jarde, 2017;
Newman, 2002). Physical examination–indicated cerclage in
women with a second-trimester twin gestation and a dilated
cervix may be benecial. Roman and coworkers (2016, 2020)
reported that women undergoing rescue cerclage plus prophylactic antibiotic and indomethacin administration beore 24
weeks’ gestation had lower preterm birth and perinatal mortality rates than those without a cerclage.
Pessary
Te silicone Arabin vaginal pessary encircles and theoretically
compresses the cervix, alters the inclination o the cervical
canal, and relieves direct pressure on the internal cervical os.
In a study o its use in women with a short cervix between 18
and 22 weeks’ gestation, a subgroup analysis o 23 women with
twins showed a signicant reduction in the delivery rate beore
32 weeks compared with the rate in 23 control pregnancies
(Arabin, 2003). In another randomized trial o twins, women
treated with a cervical pessary had signicantly ewer births
beore 34 weeks (Goya, 2016).
Other randomized studies comparing pessary against expectant care are less avorable. In one trial with 1180 uncomplicated twin pregnancies, pessary use ailed to alter the preterm
birth rate (Nicolaides, 2016). Another trial o 813 unselected
women had similar ndings. However, delivery rates beore
32 weeks were lower—29 versus 14 percent—in a subset o
women with a cervical length <38 mm (Liem, 2013). Instead,
other trials show no benet rom pessary use in this subgroup
(Nicolaides, 2016; Norman, 2021). At this time, pessary use is
not recommended by the American College o Obstetricians
and Gynecologists (2021b).
■ Treatment
Tocolytic therapy to help halt preterm labor in multietal
pregnancy does not measurably improve neonatal outcomes
( Canadian Preterm Labor Investigators Group, 1992; Yamasmit,
2015). Moreover, tocolytic therapy in women with a multietal
pregnancy entails higher maternal risk than in singleton pregnancy. Tis stems in part rom augmented pregnancy-induced
hypervolemia, which raises cardiac demands and increases the
susceptibility to iatrogenic pulmonary edema (Chap. 50, p.
883). In one study, all women with a multietal gestation treated
with a beta-mimetic drug or preterm labor had more cardiovascular complications—43 versus 4 percent—than women with
singletons (Gabriel, 1994). In a retrospective analysis, niedipine
tocolysis in 58 singleton and 32 twin pregnancies led to higher
incidences o side eects such as maternal tachycardia in women
with twins—19 versus 9 percent (Derbent, 2011).
Antenatal corticosteroids or etal lung maturation have
not been well studied in multietal gestations. O studies,
many compare outcomes among singletons and twins whose
mothers either did or did not receive this therapy. In one
large retrospective study evaluating betamethasone, neonatal morbidity rates were reduced in both twin and singleton
treated groups (Melamed, 2016). Another similar comparison ound reduced rates o periventricular hemorrhage rates
in treated groups but not improved respiratory distress syndrome (RDS) rates (Gonçalves-Ferri, 2021). Similarly, in
another study, short-term neurological outcome was better in
treated twin and singletons, but rates o mortality, RDS, and
cerebral palsy were not reduced (Ushida, 2020). However, in
one study solely with twins, the rate o RDS or o composite
neonatal morbidity was not lower with corticosteroid therapy
(Viteri, 2016). Tese authors posited pharmacokinetics in
twin gestation as a potential explanation. Currently, recommended use o agents is the same or multietal and singleton
gestations (American College o Obstetricians and Gynecologists, 2021b). Tis therapy is described ully in Chapter 45
(p. 802).
■ Preterm Prelabor Membrane Rupture
Te requency o PPROM rises with increasing plurality. In
a population-based study o more than 290,000 live births,
the proportion o preterm birth complicated by PPROM was
13 percent in singletons. Tis rate was 17, 20, 20, and 100
percent in twins, triplets, quadruplets, and higher-order multiples, respectively (Pakrashi, 2013). For both twin and singleton pregnancies, most studies show comparable neonatal risk
with expectant management, which is outlined in Chapter 45
(p. 798) (Kibel, 2017). However, the time between births,
oten termed latency, is shorter with twins. With PPROM
ater 24 weeks’ gestation, the median number o days to subsequent delivery was 4 days or twins compared with 7 days
or singletons in one study (Madden, 2021). Ater 30 weeks’
gestation, this latency or twins is signicantly shortened
(Mercer, 1993).
■ Delayed Delivery of Second Twin
Rarely, ater preterm birth o one etus, it may be advantageous or undelivered etus(es) to remain in utero. Tis may
be especially so at periviable gestational ages. However, this
advantage must be balanced against substantial maternal risk.
In one systematic review, nearly 40 percent o mothers su-
ered complications with delayed delivery (Cheung, 2020).
Tese include inection, postpartum hemorrhage, and placental abruption.
From population-based data o delayed deliveries, the
median latency duration was 6 days (range 6 to 107 days) in
200 twin pregnancies (Zhang, 2004). In all studies, retained
etuses had better overall survival rates than the rst-born neonate. In one series o 38 twins, the mean latency was 19 days.
O rst-born twins born at <25 weeks’ gestation, none survived, but 50 percent o their co-twins did. Beyond 25 weeks,
survival rates were 65 or rst-born and 95 percent or secondborn twins (Arabin, 2009). In their series o 28 twin pairs,
cerclage, antibiotic therapy, and tocolysis did not improve
outcomes (Fayad 2003).
I delayed delivery is attempted, counseling should include
the potential or serious maternal complications. Te range
o gestational age or which benets outweigh the risks or
delayed delivery is likely narrow, and gestations o 22 to 24
weeks would seem the most probable to benet (Oyelese,
2005). In our experience, good candidates or delayed delivery
are rare.
LABOR AND DELIVERY
■ Delivery Timing
Several actors aect this timing and include gestational age,
etal growth, maternal complications, lung maturity, and stillbirth risk. Te substantially greater stillbirth rate in monochorionic monoamnionic twins was discussed earlier (p. 845).
In dichorionic diamnionic twins, stillbirth rates rose to 10.6
deaths per 1000 pregnancies at 386/7 weeks’ gestation in a systematic review o more than 30,000 twin pairs. Te peak rate
in monochorionic diamnionic twins was 9.6 deaths per 1000
pregnancies at 376/7 weeks (Cheong-See, 2016).
In uncomplicated twins, stillbirth risk is balanced against
the morbidity o prematurity. Pulmonary maturation is usually
synchronous in twins and occurs several weeks earlier than in
singletons (Leveno, 1984). However, in some cases, pulmonary
unction may dier. With sFGR, the smaller, stressed twin has
lower rates o RDS but higher rates o bronchopulmonary dysplasia (Groene, 2021).
Te American College o Obstetricians and Gynecologists (2021b) recommends delivery at 380/7 to 386/7 weeks’
gestation or uncomplicated dichorionic diamnionic twin
pregnancies. Uncomplicated monochorionic diamnionic
twin pregnancies can be delivered between 340/7 and 376/7
weeks. For monochorionic monoamnionic twin pregnancies,
delivery is recommended at 320/7 to 340/7 weeks. At Parkland
Hospital, we generally ollow these recommendations and do
not routinely deliver monochorionic diamnionic twin pregnancies beore 37 weeks unless another obstetrical indication
develops.
■ Preparations
A litany o complications may be encountered during labor and
delivery o multiple etuses. Rates o uterine contractile dys-
unction, abnormal etal presentation, umbilical cord prolapse,
placenta previa, placental abruption, emergent operative delivery, and postpartum hemorrhage rates rom uterine atony are
higher. Moreover, second twins at term have worse composite
neonatal outcomes compared with outcomes o their co-twin
regardless o delivery method (Smith, 2007; Torngren-Jerneck,
2001). All o these must be anticipated, and thus certain precautions and special arrangements are prudent. Tese include
the ollowing:
1. An appropriately trained obstetrical attendant should
remain with the mother throughout labor. Continuous
electronic monitoring is preerable. I membranes are ruptured and the cervix dilated, the presenting etus is monitored internally
2. An intravenous inusion system capable o delivering uid
rapidly is established. In the absence o hemorrhage, lactated
Ringer or an aqueous dextrose solution is inused at a rate o
60 to 125 mL/hr.
3. Blood or transusion is readily available i needed.
4. An obstetrician skilled in intrauterine identication o etal
parts and in intrauterine manipulation o a etus should be
present.
5. A sonography machine is readily available to evaluate the
presentation and position o the etuses during labor and to
image the remaining etus ater delivery o the rst.
6. An anesthesia team is immediately available in the event that
emergent cesarean delivery is necessary or that intrauterine
manipulation is required or vaginal delivery.
7. For each etus, at least one attendant who is skilled in resuscitation and care o newborns and who has been appropriately inormed o the case should be immediately available.
8. Te delivery area should provide adequate space or the
nursing, obstetrical, anesthesia, and pediatric team members
to work eectively. Equipment must be on site to provide
emergent anesthesia, operative intervention, and maternal
and neonatal resuscitation.
■ Analgesia and Anesthesia
For multietal pregnancies, decisions regarding analgesia and
anesthesia must actor planned route o delivery. Other potential problems may stem rom preterm labor, preeclampsia,
desultory labor, need or intrauterine manipulation, and postpartum uterine atony and hemorrhage.
Labor epidural analgesia is ideal because it provides excellent pain relie and can be rapidly extended cephalad i internal podalic version or cesarean delivery is required. I general
anesthesia becomes necessary or intrauterine manipulation
during vaginal birth, uterine relaxation can be accomplished
rapidly with a halogenated inhalation agent (Chap. 25, p.
481). Some clinicians use intravenous or sublingual nitroglycerin or intravenous terbutaline to achieve uterine relaxation. I used, these agents are usually best administered by
the anesthesia team.
■ Fetal Presentation
For labor and delivery, the etal presentations are best determined sonographically. I active labor is conrmed, the decision or vaginal or cesarean delivery is reached, and etal
presentation is a major actor. Among the possible presentation
combinations, those most common at admission or delivery
are cephalic-cephalic, cephalic-breech, and cephalic-transverse.
At Parkland Hospital between 2008 and 2013, 71 percent o
twin pregnancies had a cephalic presentation o the rst etus
at the time o labor and delivery admission. During parturition, etal presentation o a second twin can be unstable. For
them, compound or ootling breech presentations and ace
or brow attitudes are relatively common and even more so i
etuses are small or numerous, amnionic uid is excessive, or
maternal parity is high. Cord prolapse also is requent in these
circumstances.
■ Delivery Route
Cephalic First Twin
With cephalic–cephalic presentation, general consensus supports
consideration o vaginal birth in a laboring woman near term
(American College o Obstetricians and Gynecologists, 2021b;
D’Alton, 2010). From studies, planned cesarean delivery does
not improve neonatal outcome when both twins are cephalic
(de Castro, 2016; Schmitz, 2017). With trials o labor, vaginal
delivery rates approximate 80 percent.
With cephalic–noncephalic presentation, the optimal delivery route remains controversial. Patient selection and provider
expertise with vaginal breech delivery is crucial and described in
Chapter 28 (p. 527). As a result, one common option is cesarean delivery o both twins. Less oten, ater spontaneous vaginal
delivery o a rst twin, intrapartum external cephalic version o
the second twin can be perormed. In case series, this practice,
compared with internal podalic version, was associated with
higher rates o intrapartum cesarean delivery and etal distress
(Gocke, 1989; Smith 1997). With internal podalic version, a
hand placed into the uterus grasps etal eet to deliver the etus by
breech extraction (Fig. 48-17). Last and least desirable is vaginal
delivery o the rst but cesarean delivery o the second twin due
to intrapartum complications. Tese are umbilical cord prolapse,
placental abruption, contracting cervix, or nonreassuring etal
heart rate. Most but not all studies report the worst composite
etal outcomes or this scenario (Alexander, 2008; Rossi, 2011).
For cephalic–noncephalic twins, each with birthweights
>1500 g and gestational ages >32 weeks, several reports attest
to the saety o vaginal delivery. In one study o 5915 pregnancies with a cephalic rst twin and either cephalic or noncephalic
second twin, 25 percent planned or cesarean delivery (Schmitz,
2017). Te other 75 percent planned a trial o vaginal delivery, which was successul in 80 percent. Interestingly, perinatal
mortality and morbidity rates were signicantly higher in the
planned cesarean delivery group delivered at <37 weeks—
5.2 versus 3.0 percent, respectively. Additionally, or those
delivered vaginally, neonatal outcomes were similar or second
twins that were cephalic or noncephalic (Schmitz, 2018). Te
win Birth Study had similar inclusion criteria and randomly
assigned women to planned vaginal or planned cesarean delivery. Perinatal outcomes were similar in both groups. For those
delivered vaginally, neonatal outcomes were similar or second
twins that were cephalic or noncephalic (Barrett, 2013).
For neonates weighing <1500 g compared with those weighing more, comparable or even better etal outcomes with vaginal delivery compared with cesarean delivery have been reported
(Mol, 2020; Sentilhes, 2015). However, rates o urgent operations or the second twin may be higher (Hiersch, 2021).
Other investigators advocate cesarean delivery or both
etuses o a cephalic–noncephalic twin pair (Armson, 2006;
Homann, 2012). Yang and coworkers (2005a,b) studied
15,185 cephalic–noncephalic pairs. Te risks o asphyxiarelated neonatal deaths and morbidity were higher in the group
in which both twins were delivered vaginally compared with
the group in which both twins were delivered surgically.
Breech First Twin
Problems with the rst twin presenting as a breech are similar to
those encountered with a singleton breech etus. First, the etal
body can be small, and delivery o the extremities and trunk
through an inadequately eaced and dilated cervix can leave
the relatively larger head trapped above the cervix. Tis is more
likely when disproportion between the head and body is signi-
cant. Examples are preterm or growth-restricted etuses or those
with macrocephaly rom hydrocephaly. Second, umbilical cord
prolapse is an ever-present risk. Last, twin etuses may become
locked together during delivery i the rst presents breech and
the second cephalic. As the breech o the rst twin descends
through the birth canal, the chin locks between the neck and
chin o the second cephalic-presenting co-twin. Tis phenomenon is rare, and Cohen and coworkers (1965) described it only
once in 817 twin gestations.
I these problems are anticipated or identied, cesarean
delivery is oten preerred with a viable-sized etus. Even without these problems, many obstetricians perorm cesarean delivery i the rst twin presents as breech, and this is our practice.
However, data support the saety o vaginal delivery or twins
older than 32 weeks and weighing >1500 g (Blickstein, 2000).
In one study, cesarean delivery was planned in 1169 such pairs.
Vaginal delivery was planned in 298 pairs and was successul in
64 percent. Cesarean delivery o the second twin was done in
1 percent. Neonatal mortality or morbidity measures did not
dier between delivery groups (Korb, 2020).
■ Labor Augmentation or Induction
In general, active labor with twins progresses more slowly in
both nulliparas and multiparas compared with that in singletons (Hochler, 2021). Second-stage labor o the rst twin also
is longer (Levin, 2021).
In women with twins who meet all criteria or oxytocin
administration, labor augmentation is suitable (Chap. 26, p.
492). For labor induction, studies have ound that oxytocin
alone or in combination with cervical ripening can saely be
used in twin gestations (Hamou, 2016; Ko, 2014). Compared
with prelabor cesarean delivery, others have ound higher
maternal morbidity rates with labor induction, and ultimate
cesarean delivery rates approximate 40 percent (Dougan, 2020;
Grossman, 2021). In an analysis o twin births in the United
States, induction rates o twin pregnancies have declined rom
nearly 14 percent in 1999 to 10 percent in 2008 (Lee, 2011).
Generally, at Parkland Hospital we do not augment or induce
labor in women with a multietal gestation. Concerns include
risks or uterine rupture rom an overdistended uterus and postpartum hemorrhage. In suitable candidates with a strong desire
or vaginal birth, amniotomy induction has been one option.
■ Vaginal Delivery
Ater delivery o the rstborn, one clamp is placed near the
neonate, and another is placed nearer the placenta. Until the
last etus is delivered, each cord must remain clamped to prevent etal hypovolemia and anemia caused by blood leaving
the placenta via anastomoses and then through an unclamped
cord. Cord blood is generally not collected until ater delivery o all etuses. Ater the second neonate is delivered, two
plastic clamps are placed on the placenta’s cord to dierentiate it rom the rst. In higher-order deliveries, color-tagged
or alphabetically labeled clamps can be simpler than adding
additional clamps. Tis same practice holds or cesarean delivery. At this time, evidence is insufcient to recommend or or
against delayed umbilical cord clamping in multietal gestations
(American College o Obstetricians and Gynecologists, 2020a).
Following vaginal delivery o the rst twin, the presenting
part o the second twin, its size, and its relationship to the birth
canal should be quickly ascertained by combined abdominal,
vaginal, and, at times, intrauterine examination. Sonography is
a valuable aid. I the etal head or the breech is xed in the birth
canal, moderate undal pressure is applied and membranes are
ruptured. Immediately aterward, digital examination o the
cervix is repeated to exclude cord prolapse. Labor is allowed to
resume. I contractions do not begin within approximately 10
minutes, dilute oxytocin may be used to stimulate contractions.
Te preerred interval between delivery o the rst and second twins is requently cited as <30 minutes. In some studies,
longer intertwin intervals are associated with poorer outcome
o the second twin (Leung, 2002; Stein, 2008). Others have
correlated etal heart rate tracing abnormalities during the
intertwin interval rather than its length with poorer outcome
(Algeri, 2019).
I the occiput or breech presents immediately over the pelvic
inlet, but is not xed in the birth canal, the presenting part can
oten be guided into the pelvis by one hand in the vagina, while
a second hand on the uterine undus exerts moderate pressure
caudally. A presenting shoulder may be gently converted into a
cephalic presentation. Alternatively, with abdominal manipulation, an assistant can guide the presenting part into the pelvis.
Sonography can aid guidance and allow heart rate monitoring.
I the occiput or breech is not over the pelvic inlet and cannot be so positioned by gentle pressure or i appreciable uterine
bleeding develops, delivery o the second twin can be problematic. o obtain a avorable outcome, an obstetrician skilled in
intrauterine etal manipulation and anesthesia personnel skilled
in providing anesthesia to eectively relax the uterus or vaginal
delivery o a noncephalic second twin are essential (American
College o Obstetricians and Gynecologists, 2021b). o take
maximum advantage o the dilated cervix beore the uterus contracts and the cervix retracts, delay should be avoided. Prompt
cesarean delivery o the second etus is preerred i no one present is skilled in the perormance o internal podalic version or
i anesthesia that will provide eective uterine relaxation is not
immediately available.
■ Trial of Labor after Cesarean Delivery
Te main concern with trial o labor ater cesarean delivery
(OLAC) is uterine rupture rom a distended uterus. One
metaanalysis ound rates o rupture and o successul vaginal
birth were comparable with those or OLAC in singleton gestations (Kabiri, 2019). Other studies also support the saety o
OLAC or selected women with twins (Cahill, 2005; Ford,
2006). In one assessment o 186 women undergoing OLAC,
two thirds delivered both twins vaginally. O ailed attempts,
45 percent underwent cesarean or delivery o the second twin
(Varner, 2005). According to the American College o Obstetricians and Gynecologists (2019), evidence currently does not
suggest an increased risk o uterine rupture, and women with
twins and one previous cesarean delivery with a low transverse
incision may be considered OLAC candidates. At Parkland
Hospital, we recommend repeat cesarean delivery.
■ Cesarean Delivery Technique
Several unusual intraoperative problems can arise during cesarean delivery o twins or higher-order multiples. Supine hypotension is common, and thus gravidas are positioned in a let
lateral tilt to deect uterine weight o the aorta. A low transverse hysterotomy is preerable i the incision can be made large
enough to allow atraumatic delivery o all etuses. Piper orceps
can be used i a second twin is presenting breech. In some cases,
a vertical hysterotomy beginning as low as possible in the lower
uterine segment may be advantageous. For example, i a etus
is transverse with its back down and the arms are inadvertently
delivered rst, it is much easier and saer to extend a vertical
uterine incision upward than to extend a transverse incision
laterally or to make a “” incision vertically.
■ Triplet or Higherorder Gestation
Fetal heart rate monitoring during labor with triplet pregnancies is challenging. A scalp electrode can be attached to the
presenting etus, but it is difcult to ensure that the other two
etuses are each being monitored separately. With vaginal delivery, the rst neonate is usually born with little or no manipulation. Subsequent etuses, however, are delivered according to
the presenting part. Tis oten requires complicated obstetrical maneuvers such as total breech extraction with or without
internal podalic version or even cesarean delivery. Associated
with malposition o etuses is an increased incidence o cord
prolapse. Moreover, reduced placental perusion and hemorrhage rom separating placentas are more likely during delivery.
Data rom the Consortium on Sae Labor ound that only 17
percent o women with triplet gestation attempting vaginal
birth were actually delivered vaginally. In these births, composite neonatal morbidity rates were increased (Lappen, 2016).
In 7000 triplet pregnancies, vaginal delivery also was associated with a higher perinatal mortality rate (Vintzeleos, 2005).
For all these reasons, we deliver most pregnancies complicated
by three or more etuses by cesarean delivery. Vaginal delivery is reserved or those circumstances in which survival is not
expected because etuses are markedly immature or abnormal
or maternal complications make cesarean delivery hazardous
to the mother. Centers or Disease Control and Prevention
(2009) national data show that 94 percent o triplets are delivered by cesarean.
Other cases series describe more positive outcomes (Grobman, 1998; Peress, 2019). In sum, the American College o
Obstetricians and Gynecologists (2021b) notes vaginal birth
may be considered or uncomplicated triplet pregnancies, i
the rst etus is cephalic and attendants are experienced with
multietal vaginal birth. As with twins, candidate etuses should
weigh >1500 g.
REDUCING FETAL NUMBER
For multietal gestations, multietal pregnancy reduction (MPR)
aims to lower etal number to improve survival rates o the
remaining etuses. Instead, with selective reduction, early pregnancy intervention ocuses on a etus with an anomaly or serious health risk. With selective termination, the indications are
analogous to those or selective reduction, but interventions
are perormed at a later gestational age (American College o
Obstetricians and Gynecologists, 2020c).
Selective reduction or MPR is typically done in the late rst
trimester. Tis gestational age is chosen because most spontaneous abortions have already occurred, the remaining etuses are
large enough to be evaluated sonographically, the amount o
devitalized etal tissue remaining ater the procedure is small,
and the risk o aborting the entire pregnancy as a result o the
procedure is low.
Skill in sonographically guiding needles through the mother’s abdomen and uterus is required. For reductions o etuses
with their own chorion, potassium chloride is used and is
injected into the etal heart or thorax. Entering or traversing the
sacs o etuses picked or retention is avoided. For monochorionic etuses and potentially shared vasculature, cord-occlusion
is used, and with radiorequency ablation (RFA), the needle is
inserted into the intraetal portion o the umbilical cord.
Beore reduction procedures, discussion should include the
morbidity and mortality rates associated with expectant care,
entire pregnancy termination, or selective etal reduction (American College o Obstetricians and Gynecologists, 2020c). With
these options, specic risks include (1) loss o all remaining
etuses, (2) abortion or retention o the wrong etus, (3) damage
without death to a etus, (4) preterm labor, (5) etuses with discordant or growth restriction, and (6) maternal complications.
With reduction procedures, uncommon potential complications are inection, hemorrhage, or disseminated intravascular
coagulopathy because o retained products o conception.
■ Multifetal Pregnancy Reduction
In most cases o MPR, higher-order gestations are reduced
by one or more etuses. With triplets, reduction to twins or
a singleton lowers the rate o preterm birth beore 34 weeks
compared with expectant management. Miscarriage rates are
not higher (Anthoulakis, 2017; Morlando, 2015). With even
higher-order multiples, spontaneous loss and preterm birth
rates also decline ater MPR (Evans, 2014; Liu, 2020).
win gestations also may be reduced to a singleton pregnancy. Maternal comorbidities or concerns or monochorionic
twin complications, described earlier, are requent indications
(Rao, 2021; Vieira, 2019).
■ Selective Reduction or Selective Termination
I multiple etuses are discordant or anatomical or genetic
anomalies, elimination o the abnormal etus is an option.
Other indications are severe S, APS, RAP, or sFGR.
Because abnormalities are oten not ully delineated until the
second trimester, selective termination is perormed later in
gestation than selective reduction and entails greater risk. Tis
procedure is thereore usually not perormed unless the abnormality is severe but not lethal. In some cases, termination is
considered because the abnormal etus may jeopardize the normal one
Nhận xét
Đăng nhận xét