Single-twin demise can pose substantial risks for the surviving co-twin, including increased risk of fetal loss, preterm delivery, neurovascular injury, and end-organ damage. In this chapter, we summarise recently published research on the causes of single twin demise, the pathophysiology of injury to the surviving co-twin, and the evidence for current management strategies. The gestation at which single intrauterine fetal demise occurs, and the chorionicity of the multiple pregnancies, are the two most important factors when considering the risks to the surviving twin. Management should include fortnightly ultrasound scans for growth, umbilical artery Doppler studies, and liquor volume. In monochorionic twins, more complex Doppler assessment with middle cerebral artery Doppler velocimetry and a magnetic resonance imaging of the survivor’s brain at least 3 weeks after single intrauterine fetal demise occurs should be carried out to look for evidence of neurological morbidity. With no other obstetric complications, dichorionic pregnancies can be delivered at term. Monochorionic pregnancies are more difficult to manage, and are often delivered between 34 and 36 weeks.
Introduction
Most multiple pregnancies are twin pregnancies (98%), and are associated with a higher risk of perinatal mortality compared with singleton pregnancies . The overall incidence of single intrauterine fetal demise (sIUFD) after 20 weeks of pregnancy is variably reported between studies, but is estimated to be up to 6.2% of all twin pregnancies . This complication may occur at any trimester, with potentially profound consequences. Some evidence shows that, in many cases, this occurs in the first trimester, sometimes even before the first ultrasound scan examination (‘vanishing twin’), and this adds to defining the exact prevalence of this event. The introduction of National Institute of Health and Clinical Excellence guidelines relating to the care of multiple pregnancy in 2011 has placed emphasis on the requirement for first-trimester ultrasound examination to confirm viability, define chorionicity, and provide an opportunity for prenatal diagnosis . In sIUFD, not only fetal mortality potentially increases in the surviving co-twin, but also the risk of perinatal morbidity, especially neurologic injury and risks associated with spontaneous and iatrogenic preterm birth and its complications. One of the key influential variables associated with the risk of co-twin morbidity and mortality is chorionicity . It is chorionicity rather than zygosity that influences the outcome of this twin complication, because of the placental angioarchitecture of intertwin circulations in monochorionic twins. Loss rates of up to 30–50% have been associated with monochorionic, monoamniotic pregnancies .
Aetiology
Fetal and maternal factors contributing to sIUFD are listed in Table 1 . These include the presence of discordant structural congenital anomaly (with or without chromosomal differences) . Unequal placental ‘sharing’ and selective intrauterine growth restriction (sIUGR) are ‘sentinel associations’, and velamentous cord insertion seems to compound risk. In monochorionic, monoamniotic twins umbilical cord ‘entanglement’ has classically been associated with risk of co-twin demise. Recent epidemiological studies indicate that the risk of this complication is more prevalent between 18 and 24 weeks (than the third trimester), and is often associated with co-existent pathologies, such as ‘twin reverse arterial perfusion sequence . Chorionicity, as described earlier, has an important bearing on the rate and outcome of sIUFD with monochorionic (irrespective of amniocity) pregnancies having a significantly higher rate compared with dichorionic pregnancies . In a retrospective cohort study of twin pregnancies undergoing anatomic survey, McPherson et al. found that monochorionic twins had an increased risk of sIUFD (OR 1.69, 95% CI 1.04 to 2.75) and a double demise (OR 2.11, 95% CI 1.02 to 4.37). Of all the double demises in the cohort, 70% occurred before 24 weeks . The nearly universal ‘intertwin’ communicating placental circulation is responsible for many of the unique complications in monochorionic twins, including twin–twin transfusion syndrome (TTTS), twin anaemia–polycythaemia sequence (TAPS), and is the largest predisposing risk to monochorionic twins if sIUFD occurs. In TTTS, an unbalanced unidirectional arterio-venous shunt results in a net transfer of blood from one twin to another. The donor becomes hypovolaemic, which results in a decrease in the donors cardiac output and an increase in the peripheral vascular resistance. This induces tissue hypoxia, acidosis, and a rise in erythropoietin production . The recipient twin is hypervolaemic with polyhydramnios, and has a significant risk of cardiac dysfunction, whereas the donor twin is hypovolaemic with oligohydramnios. Both twins are at high risk of intrauterine death. In contrast to TTTS, which is essentially amniotic fluid discordance, TAPS is characterised by a severe haemoglobin discordance (at least 8g/dL) caused by chronic net transfusion over minuscule unidirectional anastomoses. It can be diagnosed antenatally by noting discordant middle cerebral artery peak systolic velocity measurements . The anaemic twin often has decreased amniotic fluid volume, whereas the recipient twin may have increased fluid, but the discordance is not severe as is required for the diagnosis of TTTS. Twin anaemia–polycythaemia sequence can occur spontaneously or iatrogenic after incomplete laser treatment for TTTS. Mortality associated with TAPS is much lower than TTTS; however, antenatal cerebral lesions may occur secondary to severe anaemia–polycythaemia.
| Fetal |
| Infection |
| Chromosomal anomaly |
| Structural anomaly |
| Cord anomaly (e.g. entanglement and velamentous) |
| Placental (e.g. twin–twin transfusion syndrome, twin anaemic polycythaemia sequence, and selective intrauterine growth retardation) |
| Maternal |
| Hypertensive disorders (i.e. Pre-Eclampsia) |
| Thrombophilia, abruption |
| Iatrogenic |
| Selective feticide |
| Complication after laser therapy for twin–twin transfusion syndrome |
Selective intrauterine growth restriction can occur in monochorionic and dichorionic twins, and is defined as a difference in estimated fetal weight of over 25% . The risk of sIUFD depends on estimated fetal growth discordance, fetal growth velocity, and presence of abnormal fetoplacental blood flow. The risk of fetal death in dichorionic pregnancies with sIUGR approximates that of IUGR singleton pregnancies ; therefore, it has been recommended that IUGR in dichorionic pregnancy is managed similarly to singleton pregnancies. Monochorionic twin pregnancy present a special situation, as sIUGR and twin-to-twin transfusion syndrome (TTTS) can co-exist in 50% of cases. In sIUGR, unequal placental sharing occurs, and is associated with 14% death of at least one twin . It has also been shown that the larger the placental share discordance the greater the risk of neonatal death. No neonatal deaths occurred when the mean placental share discordance was 15.1%, whereas the neonatal death rate increased to 23% when the mean placental share discordance was 63.2% . The prognosis in monochorionic pregnancies is further complicated by the type of sIUGR. Type II sIUGR (persistent absent or reversed end-diastolic flow in umbilical artery Doppler) has been associated with severe fetal deterioration rate in up to 90% . It has also been shown that sIUGR with umbilical artery Doppler has a long latency between the onset of absent end-diastolic flow and delivery in monochorionic pregnancies (21.7 days) compared with dichorionic pregnancies (14.4 days) . Presence of placental arterio-arterial anastomoses in the placenta may contribute to longer latency, maintaining nutrition and fetal growth along centile lines. Type III (intermittent absent or reversed end diastolic flow) has been reported to have an unexpected fetal demise rate of 15% , whereas type I (normal umbilical artery Doppler PI) carries the best prognosis, with intrauterine mortality rate of 2–4% .
Frequency of malformations in dizygotic twins is thought to be similar to that of singletons, whereas it has been reported to be two to three times higher in monozygotic twins, although the true incidence is difficult to assess as most studies failed to determine zygosity . An anomalous twin is at higher risk of sIUFD in both monochorionic and dichorionic twin pregnancies. Obviously, selective feticide is an option in case of discordant anomalies in twin pregnancies, although expectant management might be considered an alternative option in selected cases, as selective termination may increase the risk of miscarriage, particularly in monochorionic pregnancies.
Single intrauterine fetal demise is iatrogenic in the case of selective feticide and after laser therapy for TTTS. In dichorionic pregnancy, passage of substances from one twin into the circulation of the co-twin is unlikely. Selective feticide by ultrasound-guided injection of potassium chloride into the fetal heart, or less frequently into the umbilical cord, is now a well-established procedure. The procedure is associated with a second fetal loss rate of 7.5% . In monochorionic pregnancies, because of vascular anastomoses, the conventional technique of potassium chloride injection cannot be used . Double intrauterine demise may occur because of transfusion of potassium chloride to the co-twin or acute exsanguination of the co-twin in the fetoplacental unit of the dead twin. As such, ultrasound-guided bipolar cord coagulation and intra-fetal radiofrequency ablation are the preferred methods of producing complete circulatory confinement of the affected twin. After laser therapy for TTTS, survival of at least one twin is 80% , and donor twins seem to have lower survival rates (60%) than recipient twins (70%) .
Maternal factors, such as hypertensive disorders in pregnancy and diabetes, are also associated with higher rates of intrauterine death , but those are mostly double IUFD.
Aetiology
Fetal and maternal factors contributing to sIUFD are listed in Table 1 . These include the presence of discordant structural congenital anomaly (with or without chromosomal differences) . Unequal placental ‘sharing’ and selective intrauterine growth restriction (sIUGR) are ‘sentinel associations’, and velamentous cord insertion seems to compound risk. In monochorionic, monoamniotic twins umbilical cord ‘entanglement’ has classically been associated with risk of co-twin demise. Recent epidemiological studies indicate that the risk of this complication is more prevalent between 18 and 24 weeks (than the third trimester), and is often associated with co-existent pathologies, such as ‘twin reverse arterial perfusion sequence . Chorionicity, as described earlier, has an important bearing on the rate and outcome of sIUFD with monochorionic (irrespective of amniocity) pregnancies having a significantly higher rate compared with dichorionic pregnancies . In a retrospective cohort study of twin pregnancies undergoing anatomic survey, McPherson et al. found that monochorionic twins had an increased risk of sIUFD (OR 1.69, 95% CI 1.04 to 2.75) and a double demise (OR 2.11, 95% CI 1.02 to 4.37). Of all the double demises in the cohort, 70% occurred before 24 weeks . The nearly universal ‘intertwin’ communicating placental circulation is responsible for many of the unique complications in monochorionic twins, including twin–twin transfusion syndrome (TTTS), twin anaemia–polycythaemia sequence (TAPS), and is the largest predisposing risk to monochorionic twins if sIUFD occurs. In TTTS, an unbalanced unidirectional arterio-venous shunt results in a net transfer of blood from one twin to another. The donor becomes hypovolaemic, which results in a decrease in the donors cardiac output and an increase in the peripheral vascular resistance. This induces tissue hypoxia, acidosis, and a rise in erythropoietin production . The recipient twin is hypervolaemic with polyhydramnios, and has a significant risk of cardiac dysfunction, whereas the donor twin is hypovolaemic with oligohydramnios. Both twins are at high risk of intrauterine death. In contrast to TTTS, which is essentially amniotic fluid discordance, TAPS is characterised by a severe haemoglobin discordance (at least 8g/dL) caused by chronic net transfusion over minuscule unidirectional anastomoses. It can be diagnosed antenatally by noting discordant middle cerebral artery peak systolic velocity measurements . The anaemic twin often has decreased amniotic fluid volume, whereas the recipient twin may have increased fluid, but the discordance is not severe as is required for the diagnosis of TTTS. Twin anaemia–polycythaemia sequence can occur spontaneously or iatrogenic after incomplete laser treatment for TTTS. Mortality associated with TAPS is much lower than TTTS; however, antenatal cerebral lesions may occur secondary to severe anaemia–polycythaemia.
| Fetal |
| Infection |
| Chromosomal anomaly |
| Structural anomaly |
| Cord anomaly (e.g. entanglement and velamentous) |
| Placental (e.g. twin–twin transfusion syndrome, twin anaemic polycythaemia sequence, and selective intrauterine growth retardation) |
| Maternal |
| Hypertensive disorders (i.e. Pre-Eclampsia) |
| Thrombophilia, abruption |
| Iatrogenic |
| Selective feticide |
| Complication after laser therapy for twin–twin transfusion syndrome |
Selective intrauterine growth restriction can occur in monochorionic and dichorionic twins, and is defined as a difference in estimated fetal weight of over 25% . The risk of sIUFD depends on estimated fetal growth discordance, fetal growth velocity, and presence of abnormal fetoplacental blood flow. The risk of fetal death in dichorionic pregnancies with sIUGR approximates that of IUGR singleton pregnancies ; therefore, it has been recommended that IUGR in dichorionic pregnancy is managed similarly to singleton pregnancies. Monochorionic twin pregnancy present a special situation, as sIUGR and twin-to-twin transfusion syndrome (TTTS) can co-exist in 50% of cases. In sIUGR, unequal placental sharing occurs, and is associated with 14% death of at least one twin . It has also been shown that the larger the placental share discordance the greater the risk of neonatal death. No neonatal deaths occurred when the mean placental share discordance was 15.1%, whereas the neonatal death rate increased to 23% when the mean placental share discordance was 63.2% . The prognosis in monochorionic pregnancies is further complicated by the type of sIUGR. Type II sIUGR (persistent absent or reversed end-diastolic flow in umbilical artery Doppler) has been associated with severe fetal deterioration rate in up to 90% . It has also been shown that sIUGR with umbilical artery Doppler has a long latency between the onset of absent end-diastolic flow and delivery in monochorionic pregnancies (21.7 days) compared with dichorionic pregnancies (14.4 days) . Presence of placental arterio-arterial anastomoses in the placenta may contribute to longer latency, maintaining nutrition and fetal growth along centile lines. Type III (intermittent absent or reversed end diastolic flow) has been reported to have an unexpected fetal demise rate of 15% , whereas type I (normal umbilical artery Doppler PI) carries the best prognosis, with intrauterine mortality rate of 2–4% .
Frequency of malformations in dizygotic twins is thought to be similar to that of singletons, whereas it has been reported to be two to three times higher in monozygotic twins, although the true incidence is difficult to assess as most studies failed to determine zygosity . An anomalous twin is at higher risk of sIUFD in both monochorionic and dichorionic twin pregnancies. Obviously, selective feticide is an option in case of discordant anomalies in twin pregnancies, although expectant management might be considered an alternative option in selected cases, as selective termination may increase the risk of miscarriage, particularly in monochorionic pregnancies.
Single intrauterine fetal demise is iatrogenic in the case of selective feticide and after laser therapy for TTTS. In dichorionic pregnancy, passage of substances from one twin into the circulation of the co-twin is unlikely. Selective feticide by ultrasound-guided injection of potassium chloride into the fetal heart, or less frequently into the umbilical cord, is now a well-established procedure. The procedure is associated with a second fetal loss rate of 7.5% . In monochorionic pregnancies, because of vascular anastomoses, the conventional technique of potassium chloride injection cannot be used . Double intrauterine demise may occur because of transfusion of potassium chloride to the co-twin or acute exsanguination of the co-twin in the fetoplacental unit of the dead twin. As such, ultrasound-guided bipolar cord coagulation and intra-fetal radiofrequency ablation are the preferred methods of producing complete circulatory confinement of the affected twin. After laser therapy for TTTS, survival of at least one twin is 80% , and donor twins seem to have lower survival rates (60%) than recipient twins (70%) .
Maternal factors, such as hypertensive disorders in pregnancy and diabetes, are also associated with higher rates of intrauterine death , but those are mostly double IUFD.
Definition
Early first trimester or vanishing twin syndrome
Vanishing twin syndrome occurs when a diagnosis of a twin pregnancy is made ultrasonographically (usually in the early first trimester), followed by a repeat ultrasound scan weeks later where only one fetus can be identified. The true rate of vanishing twins is difficult to determine, but may be as high as 29% . Chorionicity is again an important factor, with prognosis for monochorionic twins being poor and associated with a high risk of progressing to double IUFD . Evidence though is conflicting and, in another study, no developmental delay was found when comparing surviving twins from a ‘vanishing twin’ pregnancy to that of singleton pregnancies up to 1 year of age .
Single-Twin demise over 14 weeks gestation
Single intrauterine fetal demise can also occur in second or third trimesters and, when preterm labour of the surviving twin does not ensue, this situation can leave obstetricians with difficult decisions of delivering a premature twin or conservative management with the risk of morbidity and mortality to the surviving twin as the pregnancy advances.
Determinants of the risks to the surviving co-twin
Timing of sIUFD and chorionicity are important factors in determining the outcome of a co-twin. Although it remains debatable, first-trimester loss is not known to result in an adverse outcome for the co-twin however; sIUFD in the second and third trimester can be associated with substantial risk in the co-twin . Preterm delivery is common in both monochorionic and diamniotic pregnancies, resulting in the complications of prematurity, including neonatal death, pulmonary hypoplasia, and necrotising enterocolitis. A recent systematic review by Hillman et al. showed no significant difference between the preterm delivery rate of monochorionic and dichorionic survivors (odds ratio 1.1, 95% CI 0.34 to 3.51; P = 0.9) ( Fig. 1 a ).
Chorionicity is another important consideration. Monochorionic pregnancies are at increased risk because of shared placental circulation. Hillman et al. showed that, after sIUFD, the rate of co-twin death in monochorionic pregnancies was 15% (95% CI 9.1 to 20.9) and dichorionic pregnancies was 3% (95% CI 0.3 to 5.7). Monochorionic twins had five times higher odds of co-twin demise after sIUFD compared with dichorionic twins (OR 5.24, 95% CI 1.75 to 15.7; P < 0.05) ( Fig. 1 b).
Pathophysiology
The exact mechanisms leading to the risk of morbidity and mortality of the co-twin after sIUFD are yet to be proven. There are two main theories: haemodynamic fluctuations between the twins, and transchorionic embolisation and coagulopathy. The former theory being the most supported. The remaining discussion will focus on monochorionic twin pregnancies, as these two theories are specific to monochorionicity.
Benirschke in 1961 first hypothesised that it was a passage of ‘thromboplastic material’ from the dead twin to its co-twin via placental vascular anastomoses, which in turn induced disseminated intravascular coagulation (DIC) in the co-twin. Thromboembolic material may be seen in surviving co-twins but, whether the thrombi have arisen from circulation within the dead twin or from haemodynamic changes within the survivor remains unclear. The resulting DIC can cause infarcts and cystic changes in the survivor’s renal, pulmonary, hepatic, splenic, and neurological systems . The resulting arteriolar occlusion causes end-organ damage; this has been shown both angiographically and from autopsy data . There are doubts about the time at which intracranial ultrasound anomalies have been found and whether DIC could have arisen this quickly or whether other causal factors are involved. For this reason, the thromboembolic mechanism has not been substantiated.
Acute exsanguination from the surviving fetus to the dead fetus along the placental anastomoses (a ‘back-bleed’) was first proposed by Fusi et al. This is now believed to be the favoured cause of co-twin morbidity and mortality after sIUFD in monochorionic twins. The surviving fetus subsequently suffers hypoperfusion, hypotension, and fetal anaemia, which in turn cause tissue hypoxia, acidosis and damage, particularly within the central nervous system. In this case report, the surviving twin sustained cerebral and renal lesions, was anaemic, but did not have coagulopathy. Nicolini et al. reported eight fetuses with sIUFD that underwent blood sampling either less than 24 h before sIUFD occurred (five cases) or less than 24 h after sIUFD (four cases). Four of the five pregnancies were not anaemic before the sIUFD (haematocrit 33–40%), and neither were their co-twins; however, all survivors sampled within 24 h after the death of their co-twin were anaemic (haematocrit 17–29%) . Similarly, Okamura et al. obtained fetal blood from five monochorionic twin survivors after sIUFD, and found that all five were anaemic, particularly when the intrauterine death had occurred within 24 h of sampling. One twin was sampled before and after death of the co-twin, and the haemoglobin concentration decreased from 15 to 10 g/dl. All five surviving fetuses sustained a cerebral injury ( Table 2 ). Bajoria et al. studied the outcome of twin pregnancies complicated by sIUFD in relation to vascular anatomy of the monochorionic placenta. It was found that, in twins without TTTS, the presence of superficial arterial–arterial anastomosis or venous–venous anastomosis had a higher incidence of intrauterine death, fetal anaemia, and neurological handicap. They hypothesised that these anastomoses allow a perimortem rapid transfer of blood from the live fetus to the dead fetus, causing subsequent neurological damage or fetal demise. This contradicts the thromboembolic theory, as the gradient is such that thromboembolic material could not have flowed from the dead fetus’s circulation to the survivor. On the contrary, a favourable outcome was observed with multiple bidirectional arterial–venous (AV) anastomoses. Because none of these twins had significant anaemia at birth, and all had normal neurological development, it was hypothesised that the AV/VA channels with oppositely directed blood flow allow the achievement of a relatively steady haemodynamic state .
