44: Multiple pregnancies and births

Multiple pregnancies and births

Isaac Blickstein and Oren Barak

Obstetrics and Gynecology, Kaplan Medical Center, Rehovot, The Hadassah‐Hebrew University School of Medicine, Jerusalem, Israel

The natural pattern of plurality exhibits the relative rare birth of twins (about 1 per 80–100 births) and the extremely rare occurrence of high‐order multiple pregnancies. The rarity of high‐order multiple pregnancies can be appreciated by the quasi‐mathematical Hellin–Zellany rule for twins, triplets, and quadruplets [1]. According to this law, if the frequency of twins in a population is 1/N, then the frequency of triplets will be 1/N2 and that of quadruplets 1/N3.

The Hellin–Zellany appears to be quite accurate as long as a population remains homogenous and enjoys natural conceptions. Otherwise, it became clear that deviations from the rule often exist because of racial differences in the frequency of dizygotic (DZ) twinning. The next significant deviation appeared after the emergence of effective infertility treatment when physician‐made (iatrogenic) multiple pregnancies are now seen in almost all countries, with frequencies approaching 50% of twins and more than 75% of high‐order multiple pregnancies. The contribution of infertility treatment to perinatal medicine can be appreciated from data of the Israel Neonatal Network. The data indicate that among infants weighing less than 1500 g, 10% of singletons were conceived by assisted reproduction compared with 60% of twins and 90% of triplets [2]. More recently, Tul et al. [3] found that the incidence of twins after assisted reproduction (Assisted Reproductive Technologies (ARTs)) born at <32 weeks increased 27‐fold from 1987 to 2010 and has not reduced from its peak incidence over the last decade.


Most spontaneous human conceptions (>99.2%) emerge from a single zygote (i.e. monozygotic, (MZ)), whereas in the remaining cases, more than one ovum is ovulated and fertilized, resulting in polyzygotic conceptions (dizygotic (DZ), trizygotic, etc.). This phenomenon appears to occur more often in taller, older, parous, heavier, and black women. Although direct and indirect evidence point to a genetic predisposition of DZ twining, the exact mechanism whereby the ovary is naturally hyperstimulated is basically unknown. In contrast, all infertility treatments are associated with ovarian stimulation and polyovulation.

The vast majority of MZ conceptions result in singleton birth. In only a small fraction of the cases (0.4% of all natural conceptions) the zygote splits to form an MZ twin gestation. The only factor known to increase the frequency of MZ twins is assisted reproduction [4]; however, the true incidence of zygotic splitting following ART is unknown. In a large study of single‐embryo transfers, a sixfold increase in zygotic splitting was found, and this incidence was not influenced by using fresh versus frozen‐thawed embryos or by performing embryo transfers during a spontaneous versus an induced cycle [4]. Regardless, the mechanism of spontaneous zygotic splitting is unclear. A recent hypothesis suggests that the potential to undergo splitting might be an inherent characteristic of the oocyte [5].

From a clinical point of view, the placental arrangement (i.e. chorionicity and amnionicity) are more important than zygosity. DZ twins have two placentas (separate or fused), each with its chorion and amnion, forming the so‐called dichorionic (DC) placenta. Placentation of the MZs, are assumed to depend on the stage of embryonic development at which the split occurs. Early splits (about one third) result in DC placentas, whereas later splits result in monochorionic (MC) placentas. Moreover, if the amnion has not yet differentiated, the MC placenta includes two amniotic sacs: the monochorionic–diamniotic (MCDA) placenta (about two‐thirds of the cases). If the split occurs later than eight days after fertilization, a monochorionic–monoamniotic (MCMA) placenta develops. Finally, even later splits result in all varieties of conjoined twins. This construct of events appearing in every textbook, is unproven, as evidenced by the recent controversy regarding the quasi‐accepted relationship between the timing of zygotic splitting and placentation [6, 7].

In any case, because MZs with a DC placenta cannot be differentiated clinically from same‐sex DZ twins (half of the DZs) who also have a DC placenta, zygosity can be determined with certainty only in the DC‐unlike‐sex twins (all must be DZs) and in twins with an MC placenta (all must be MZs). Simple calculation reveals that we are blind to zygosity in about 45% of the cases, and zygosity determination, if required, must be performed by DNA testing. Importantly, nothing should be said about zygosity to parents of same‐sex twins with a DC placenta.

Maternal consequences

The significant changes in women’s role in western societies witnessed after World War II, facilitated by effective contraception, allowed ample time to achieve education and a career, but resulted in increased maternal age at first delivery. Because age and fecundity are inversely related, infertility treatment to achieve a pregnancy often becomes inevitable. Because all infertility treatments carry an increased risk of multiple gestations, the end result of these socio‐medical trends is an increased age of the cohort of mothers of multiples. US data clearly illustrate that the increase in maternal age is more prominent in high‐order multiple pregnancies than in twins and in twins than in singletons, with a net result of multiples being more often delivered to older mothers in whom chronic disease conditions have already accumulated [8, 9].

Older maternal age frequently combines with the inevitable overwhelmed maternal homeostasis. Consider the fact that the average singleton, twin, and triplet has a similar birthweight until 28 weeks (around 1000 g). Thus, by 28 weeks, the mother of twins and the mother of triplets has accumulated twice and three times the fetal mass of singletons, respectively. This excess of fetal mass must come from either existing maternal resources or from supplemental energy. It is thus clear that during the third trimester all maternal systems in a multiple pregnancy are overwhelmed and some may be only a step away from clinical insufficiency.

Two examples vividly demonstrate the situation. The first is the increased frequency of clinically significant anemia during twin gestation as a result of either depleted maternal iron stores or from inadequate iron supplementation [10]. A second example relates to the increased cardiac output [11, 12]. Kuleva et al. [12] demonstrated a significantly higher increasing cardiac output throughout pregnancy in multiple as compared to a singleton gestation. It has been estimated that in the worst‐case scenario (e.g. preterm labor due to infection in a multiple pregnancy) the cardiac output may exceed 10 l min−1 (two to three times the normal value). It is therefore understandable why cardiac function so easily turns into dysfunction. Ghi et al. [13] showed that in uncomplicated twin gestations, significant changes in maternal systolic and diastolic function occur from the first to the third trimester. However, whereas diastolic parameters normalize after pregnancy, a relative systolic dysfunction persists after delivery.

Regardless of the altered maternal physiology in a multiple pregnancy, some maternal disease conditions are more frequent in these gestations. Foremost are hypertensive disorders which are two to three times more frequent [14] and their most dangerous complication – eclampsia – is six times more frequent among mothers of multiple gestations [15]. Moreover, pre‐eclamptic toxemia (PET) occurs earlier in multiples than in singletons and often occurs in a more severe form [16]. Because triplets and other high‐order multiples were rare in the past, scant data exist on hypertensive disorders in high‐order multiple pregnancies. Along with the current epidemic dimensions of multiple gestations it has been shown that the risk of hypertensive disorders is plurality dependent, whereby the risk in triplets is higher than that in twins, and the risk in twins is higher than that in singletons [17]. This may suggest that hyperplacentation is an important reason for the higher incidence of pre‐eclampsia in multiples. It is unknown why PET is more frequent in multiples. One potential explanation comes from a recent population‐based study [18] that reiterated the exceptionally important association between the high pre‐gravid body mass index (BMI) (rather than weight gain) and pre‐eclampsia. Another theory suggested a higher incidence of hypertensive disorders in DZ twins (more “immunogenic” difference) than in MZ twins. A recent study, confirmed previous smaller studies that appear to disprove this theory [19].

In contrast to the clear association to pre‐eclampsia, data are still conflicting about the relationship of multiple pregnancy and gestational diabetes. It appears that most tests to detect glucose intolerance showed a diabetogenic effect of multiple gestations, without a significantly increased rates of gestational diabetes. However, as with pre‐eclampsia [17], it was shown that the risk of gestational diabetes is plurality dependent [20] pointing, at least in a teleological way, to hyperplacentosis as a potential common denominator for both gestational diabetes and hypertension. Yet, another common denominator – pregravid obesity – seems to equally important [18].

Whereas it is clear how hypertensive disorders influence a multiple pregnancy, the effect of gestational diabetes is less robust. Fox et al. [21] suggested that it is not clear that glycemic control in twin pregnancy is improving outcome and was in fact associated with an increased risk of small for gestational age (SGA) infants. A study by Simões et al. [22] found also that pre‐gravid obesity appears to predispose women to gestational diabetes. They showed that twins from the gestational diabetes group had more respiratory distress syndrome and had a threefold, but not significantly, increased perinatal mortality rate. Birth weight characteristics were similar in both groups.

Mothers of multiples are at considerably greater risk of preterm labor and delivery. Many prophylactic measures, including progestatives, cervical sutures (cerclage), beta‐sympathominetics, bed rest, and hospitalization, failed to significantly reduce this complication. Nevertheless, expecting mothers of multiples are frequently asked to leave work and to conduct a more sedentary lifestyle.

Table 44.1 lists the most common maternal complications during multiple gestations.

Table 44.1 Maternal complications more frequently seen in multiple pregnancies

Hypertensive diseases

  • Pre‐eclamptic toxemia

  • Hemolysis, elevated liver enzymes, low platelets (HELLPs) syndrome

  • Acute fatty liver

  • Pregnancy‐induced hypertension

  • Chronic hypertension

  • Eclampsia
Gestational diabetes mellitus (?)
Premature contractions and labor

  • Complications associated with tocolysis
Delivery‐associated complications

  • Cesarean section

  • Operative delivery

  • Premature rupture of membranes

  • Postpartum endometritis

  • Placental abruption

Fetal–neonatal consequences

In all mammals, an inverse relationship exists between litter size and both gestational age and birthweight. In the human, the average gestational age at birth is around 40 weeks for singletons, 35.3 weeks for twins, 31.9 weeks for triplets, and 29.5 weeks for quadruplets [23]. Although multiple pregnancies display many specific complications, the consequences of preterm birth are by far the most common and most important in terms of morbidity and mortality.


Multiples are notorious for an increased risk of malformations. However, the increased risk is mainly related to MZ twinning whereas the malformation rates of each of the DZ twins is similar to that of singletons [24]. Nonetheless, the mother of DZ twins has an increased risk that one of the twins will be affected. In contrast, the higher malformation rate among MZs is explained by the hypothesis of a common teratogen: the one that causes the split of the zygote might be also responsible for the malformation.

Malformations among multiples are grouped into four types (Table 44.2) [24]. The first type includes malformations that are more frequent among multiples, especially those of the central nervous and the cardiovascular systems. The second type involves malformations specific to MZ twinning such as twin reverse arterial perfusion (TRAP) sequence and the various forms of conjoined twins. The third type relates to consequences of placental malformations, in particular the MC placenta, resulting in the twin–twin transfusion syndrome (TTTS ), selective intrauterine growth restriction (sIUGR), and twin anemia–polycythemia sequence (TAPS). Finally, the fourth type involves skeletal abnormalities such as clubfoot that are caused by intrauterine fetal crowding.

Table 44.2 Categories of structural defects in twins

Category Defect
Malformations more common in twins than in singletons Neural tube defects

  • Congenital heart disease
  • Esophageal and anorectal atresias
  • Intersex
  • Genitourinary tract anomalies
  • Amniotic band syndrome
Malformations unique to monozygotic twins TRAP sequence
Conjoined twins
Twin embolization syndrome
Placental malformations Single umbilical artery
Twin–twin transfusion syndrome
Velamentous cord insertion
Selective growth restriction
Twin anemia–polycythemia sequence
Deformations due to intrauterine crowding Skeletal (postural) abnormalities; i.e. clubfoot, dolichocephalus

Some malformations can have a major impact on the normal twin. For instance, in the TRAP sequence, the circulation of the severely anomalous twin is entirely supported by the normal (pump) twin. Sooner or later, this cardiac overload will lead to cardiac insufficiency in the normal twin. Another example is the case in TTTS whereby both twins are usually completely normal, but the anomalous transplacental shunt of blood can cause serious morbidity in both twins. The most striking example is the case of single fetal demise in MC twins, whereby the surviving fetus dies in utero soon after the death of the first twin. Alternatively, the surviving twin can be seriously damaged (see later). A final example is the presence of an anencephalic twin which may be surrounded by severe polyhydramnios, increasing the risk of preterm birth. In such discordant lethal malformations, the risk of reducing the anencephalic twin should be weighed against the risk of endangering the normal fetus by the procedure‐related preterm birth.

In contrast to structural malformations, chromosomal anomalies are not more frequent among multiples. For example, each member of the multiple gestation has the same maternal‐age‐dependent risk for trisomy 21. However, as with the probability calculations for structural anomalies, the risk for a mother that one of her twins will have trisomy 21 is greater than that of a mother of a singleton. Roughly, the risk for the mother that one of her twins will have trisomy 21 is 5/3 the risk of a mother of a singleton of the same age [25].

Because multiples are commonly seen in older mothers and invasive cytogenetic procedures (amniocentesis or chorionic villus sampling) carry a higher risk of pregnancy loss when performed in multiples, there is a genuine need for non‐invasive maternal screening of aneuploidy to minimize the need for invasive procedures in these premium pregnancies. Regrettably, screening tests like the PAPP‐A and inhibin and the triple test (second trimester maternal serum human chorionic gonadotropin (hCG) or free beta‐hCG, alpha‐fetoprotein, and unconjugated estriol) have a significantly lower prediction for trisomy 21 in multiples compared with singletons. An advance in this area is the implementation of nuchal translucency thickness measurement with or without biochemical markers in screening for aneuploidy [26]. Sarno et al. [27] examined the role of cell‐free DNA testing in twin pregnancies and showed that the fetal fraction is lower (except for MZ twins where it is rather higher) and the failure rate is higher compared to singletons. The authors maintained that, at present, the data were too small for a fair assessment of performance of screening for trisomy 21, but it may be similar to that in singleton pregnancies.

Most structural anomalies can be detected by comprehensive sonographic and echocardiographic scans as well as Doppler velocimetry are able to detect many structural and functional cardiovascular anomalies. When a malformed twin is found the question of selective reduction of the anomalous twin might be discussed with the parents. In multichorionic multiples, reduction is accomplished by ultrasound‐guided intracardiac injection of potassium chloride. However, because of the risk to the survivor in MC sets, highly invasive procedures are used to interrupt the umbilical circulation of the anomalous twin. In some instances, for example in the TRAP sequence, intrafetal radiofrequency might be employed to reduce the malformed twin.

All invasive procedures (amniocentesis, chorionic villus sampling, and the reduction methods) are associated with the risk of 5–10% of membrane rupture and loss of the entire pregnancy. When an invasive procedure is considered during the second trimester, the risk of extremely preterm birth of the normal twin is apparent. In some countries without an upper limit of gestational age for fetal reduction, invasive diagnostic methods might be deferred until 32 weeks, thus minimizing the risk of procedure‐related preterm birth.

Embryonic and fetal demise

With the advent of sonography, it was clear that more twin pregnancies are generated than born. The early loss of one twin was eventually called vanishing twin syndrome (VTS) to denote the disappearance of an embryonic structure during the first trimester [28]. This spontaneous reduction appears to be the natural equivalent of intentional multifetal pregnancy (numerical) reduction.

Logically, the true frequency of VTS is unless sonography is performed at an early stage. One estimate of VTS frequency comes from iatrogenic conceptions: Spontaneous reduction of one or more gestational sacs or embryos occurred before the 12th week of gestation in 36% of twin, 53% of triplet, and 65% of quadruplet pregnancies [29]

Only gold members can continue reading. Log In or Register to continue

Jul 19, 2020 | Posted by in GYNECOLOGY | Comments Off on 44: Multiple pregnancies and births
Premium Wordpress Themes by UFO Themes