Twin pregnancies are high risk and require regular antenatal surveillance and discussion regarding time and mode of delivery, especially in uncomplicated twins.

The rate of multiple births varies from country to country, with twin pregnancies accounting for just over three per cent of all births in Australia during 2008. Twin pregnancies are associated with complications for both the mother and the fetus. From the maternal perspective, there are increased risks of hyperemesis, preterm delivery, premature rupture of membranes, gestational diabetes mellitus and pre-eclampsia. The twins themselves face increased risks of perinatal mortality, predominantly related to prematurity, as well as complications related to chorionicity/amnionicity and complications occurring during birth.

The rates of delivery before 32 weeks are approximately two per cent for singletons compared to 12.5 per cent for twins. The risk of infant death is even greater: 29.8/1000 for twins and 59.6/1000 for triplets, compared with 6/1000 for singletons. Rates of cerebral palsy have also been estimated to be four to eight times higher in twins when compared to singletons.¹

Twins can be classified as either monozygotic (originating from the fertilisation and subsequent division of one egg) or dizygotic (originating from the fertilisation and development of two eggs). Twins can be further classified by their chorionicity: dizygotic twins are always dichorionic, diamniotic (DCDA). In relation to monozygotic pregnancies, DCDA twins result if the fertilised egg splits in the first three days after fertilisation. Monochorionic diamniotic (MCDA) twins result if the split occurs at days between days four and eight, and monochorionic monoamniotic (MCMA) twins result if the split occurs after the eighth day post-fertilisation. Conjoined twins occur with division 13 days or later; this is extremely rare.

DCDA twin pregnancies are usually monitored by ultrasound scans every four weeks after the morphology scan and more frequently if there is evidence of discordant growth. Given consistent evidence of increasing risk in twin pregnancies extending past 38–39 weeks, planning elective delivery at 38 weeks in well-dated, uncomplicated dichorionic twin pregnancies is a rational management approach.^2,3 The timing of twin delivery is, however, highly debated. According to findings of multiple population-based studies worldwide, the lowest risk for perinatal mortality and morbidity for twin births appears to be between 36 and 38 weeks gestation. This debate unfortunately could not be answered by the randomised trial by Dodd and colleagues, but showed a trend for lower risk in the group delivered electively at 37 weeks group.⁷

MCDA twin pregnancies are considered high-risk pregnancies owing to the twins sharing a single placenta. The incidence of monochorionic twinning is approximately one in 400 and is fairly constant. MCDA twin pregnancies are characterised by placental vascular anastomoses and resulting inter-fetal transfusion. MCDA twins are considered high risk by virtue of their three-to-five-fold increase in perinatal morbidity and mortality compared to dichorionic twin pregnancies.

Twin to twin transfusion syndrome (TTTS) complicates about 15 per cent of MCDA twin pregnancies, with discordant intrauterine growth restriction complicating an additional 25 per cent. Furthermore, in the event of intrauterine fetal death (IUFD) of one twin, there is a 20–25 per cent risk of death or neurological damage in the surviving twin from acute inter-twin transfusion. These complications are the result of significant neurological damage secondary to hypotension occurring in the live twin at the time of demise of the other.¹

Frequency of ultrasound surveillance for MCDA pregnancies varies in different centres, although the Society for Maternal and Fetal Medicine has published recommendations for the diagnosis and management of TTTS.^2,3 All women with MCDA pregnancies are offered routine first-trimester screening, cervical length screening at later visits and most centres recommend fortnightly ultrasound from 16 weeks onwards. Early nuchal translucency discordance, with a difference of 20 per cent or more between nuchal thicknesses of the two fetuses, imparts a higher risk of TTTS. Regular ultrasound scans are undertaken to identify problems such as: TTTS; twin reversed arterial perfusion (TRAP); early-onset fetal growth restriction; the presence of congenital malformations (especially cardiac abnormalities)⁴; selective fetal growth restriction (defined as a difference in weight of 25 per cent or more between twins); and twin anaemia and polycythaemia sequence (TAPS). When such findings are evident, appropriate referral to a tertiary centre with a maternal-fetal medicine unit is recommended.

TAPS is diagnosed by performing fetal middle cerebral artery peak systolic velocity (MCA PSV) in both fetuses: MCA levels above 1.5 multiple of the median (MoM) (anaemia) in one twin and less than 0.8 MoM (polycythemia) in the other are diagnostic. In these twins, there is usually a large inter-twin discordance of haemoglobin levels without an amniotic fluid discordance. This is seen in both uncomplicated MCDA twins and in MCDA twins treated with laser photocoagulation. The vessels involved are usually very small and deep, hence treatment with laser photocoagulation may not correct the process. Antenatal treatment remains a difficult decision in these twins and it is vital to alert the treating neonatal team when TAPS is suspected before delivery.

Thanks to frequent monitoring and standardised criteria, early recognition and management of complications in MCDA pregnancies has improved in recent times and enhanced efforts are now directed towards the monitoring of uncomplicated monochorionic twins. There has been ongoing debate on the timing and mode of delivery of uncomplicated MCDA twin pregnancies. This is important because of an increased incidence of unexplained stillbirth in uncomplicated MCDA twins and acute intrapartum TTTS. Most complicated MCDA twins are delivered before 36 weeks gestation. For otherwise uncomplicated MCDA twins, the risk of stillbirth rises steadily with gestational age. This was demonstrated by a recent meta-analysis in which the rate of stillbirth per 1000 uncomplicated MCDA twin pregnancies increased steadily from 32 to 37 weeks. When compared to uncomplicated dichorionic pregnancies, the odds ratio for stillbirth per pregnancy at 32, 34 and 36 weeks gestation were found to be significantly higher. This information should be weighed against the risk of prematurity when planning timing of delivery in uncomplicated MCDA twins.⁵

MCMA twin pregnancies occur rarely, representing about one per cent of cases of monozygotic twins. It is generally believed that monoamniotic twins are at high risk of fetal death and neonatal morbidity, secondary to umbilical cord entanglement, prematurity and congenital anomalies. They undergo ultrasound surveillance fortnightly and debate continues as to whether inpatient management improves perinatal outcomes. It is estimated that the risk of sudden, unexpected stillbirth in monoamniotic twins remains somewhere between five per cent and ten per cent after 32 weeks gestation. Because of this continuing risk, elective delivery after the administration of antenatal corticosteroid therapy is recommended by 32 weeks gestation.

One of the important causes of fetal death in MCMA twins is cord entanglement. Medical amnio-reduction with Sulindac has been studied in a small number of patients with the intention of stabilising fetal lie and, hopefully, preventing cord accidents.⁶ However, a recent systematic review of cord entanglement in MCMA twin pregnancies revealed that the increased risk of neonatal morbidity and mortality in these pregnancies was owing to multiple factors such as TTTS, prematurity and congenital anomalies, not only cord entanglement.¹

Intrauterine death of one fetus in a multiple gestation during the second or third trimester complicates between 0.5 per cent and 6.8 per cent of twin pregnancies and can have severe sequelae for the surviving fetus. Monochorionic twins are at a several-fold increased risk for a single fetal death compared with dichorionic twins. The aetiology of IUFD in a multiple pregnancy may be owing to genetic or anatomic anomalies, placental insufficiency or cord abnormalities, such as a velamentous insertion. In monochorionic pregnancies, 25–35 per cent of IUFDs are associated with TTTS.

Single IUFD in a multiple gestation can cause multicystic encephalomalacia and/or multi-organ damage in monochorionic pregnancies, preterm labour and delivery in both dichorionic and monochorionic twins; as well as maternal consumptive coagulopathy.¹ Aside from the chorionicity of a twin pregnancy, the gestational age at which a single IUFD occurs is important. In MCDA and MCMA twin pregnancies, single IUFD causes significant neurological sequelae when the death happens in the second trimester. Owing to a 20 per cent risk of neurological sequelae in the survivor if delivery occurs at this gestation, immediate delivery is not recommended. Follow-up ultrasound scan and fetal MRI two to three weeks after the event helps in counselling and appropriate management. The goal is to optimise the outcome for the survivor, while avoiding prematurity and its potential adverse sequelae.

Infants from a twin pregnancy are at a higher risk of death in the peripartum period than are infants from a singleton pregnancy. Some of this increased risk is owing to a higher risk of preterm birth. In addition, the second-born twin has an increased risk of a poor perinatal outcome compared with the firstborn twin. A policy of planned vaginal birth for women with a twin pregnancy in a hospital setting is associated with a 30–40 per cent rate of emergency caesarean section. Among those twins in which the first twin is born vaginally, there is still a risk of emergency section for the birth of the second twin. It is possible that some of the adverse outcomes may be avoided by appropriately timed delivery by caesarean section.⁸ However, a recent randomised twin birth study has shown that there is no difference in the neonatal mortality or serious morbidity between delivery by caesarean section and vaginal delivery. The study had a strict protocol that was adhered to and among the group randomised to vaginal delivery only 56.2 per cent delivered vaginally and 39.6 per cent had caesarean section.⁹

Women with multiple pregnancies need introduction to support systems in the antenatal period. Mothers of multiple births face higher rates of postpartum depression and twin births may be associated with longer term parental divorce.¹⁰