Twin–twin transfusion syndrome complicates up to 15% of monochorionic pregnancies in the mid-trimester, and results in high perinatal mortality and morbidity if left untreated. Although some humoral factors play a role in the pathogenesis of the disease, an unequal placental sharing and the presence of placental vascular anastomoses at the chorionic plate, allowing blood volume shifts between the twins, are the anatomic prerequisite for this complication unique to monochorionic twins. Within monochorionic pregnancies, it is possible to identify a subgroup of twin pairs at high risk of developing twin–twin transfusion syndrome or selective intrauterine growth restriction during the course of pregnancy as early as in the first or early second trimester. If progressive amniotic fluid discrepancy advances to moderate twin–twin transfusion syndrome, and finally reaches the stages of severe twin–twin transfusion syndrome, accurate classification of the clinical picture and diagnosis of the individual fetal haemodynamic status are crucial for counselling parents on treatment options and possible outcomes. Clear criteria have been established for fetoscopic laser coagulation of placental vascular anastomoses, which is the treatment of choice for severe twin–twin transfusion syndrome, interrupting blood flow between the twins, and relieving uterine over-distension related to severe polyhydramnios.
Introduction
Monochorionic twinning is related to cleavage of a monozygotic blastomere after the third post-conceptional day. An earlier cleavage will result in monozygotic dichorionic twins. After day 9, the monochorionic twin pair will also share a common amniotic cavity (monoamniotic twins) and, after day 12, the cleavage may not be complete, resulting in conjoined twins . Around two-thirds of monozygotic twins are monochorionic, and 12% of these pregnancies will be complicated by twin–twin transfusion syndrome. Another 15% of the monochorionic twins will develop selective intrauterine growth restriction (sIUGR), and 5% will be complicated by TAPS, typically in the early third trimester. On the one hand, these complications are related to the presence of placental vascular anastomoses of monochorionic placentas, which allow shifts of blood from one twin to the other. On the other hand, the placental mass at disposition for each twin will cause growth restriction in one of them (sIUGR), if there is unequal sharing. In some cases, twin–twin transfusion syndrome (TTTS) is combined with severe growth restriction of the donor twin.
The presence of vascular anastomoses at the chorionic plate is almost universal to monochorionic placentas, most of them being arteriovenous in nature . As such, arteriovenous anastomoses are uni-directional (allowing blood flow from one twin to the other), with a feeding artery of one twin supplying a placental cotyledon by perforating the chorionic plate to the depth and a draining vein arising from the depth through the chorionic plate to carry oxygenated blood to the co-twin. During fetoscopy in cases of severe TTTS, and in colour dye injection studies of monochorionic placentas, arteriovenous anastomoses in both directions have almost always been observed . Arterio–arterial and veno–venous anastomoses, which run on the surface of the chorionic plate, one vessel directly continuing into the other, allow blood flow in both directions (bi-directional), with a low resistance, and are less common. In arterio–arterial anastomoses, the blood pressure of each twin causes the pulse waves to collide against each other with different beat frequencies, the net flow through these anastomoses thus depending on the haemodynamic state of the respective twin. Because of their low vascular resistance, huge inter-twin shifts of blood volume are possible through arterio–arterial anastomoses, on one hand allowing compensation of volume shifts occurring through arteriovenous anastomoses predominantly in one direction and, on the other hand, permitting an acute and massive blood shift towards an eventually hypotensive co-twin, with the consequence of striking hypovolaemia and the risk for hypoxic-ischaemic encephalopathy in the normotensive twin . Similar volume shifts are believed to occur through veno–venous anastomoses, although in a lower intensity. Such inter-twin vascular accidents, along with chronic transfusion through arteriovenous anastomoses, are believed to be responsible for the increased mortality and morbidity in monochorionic twin pairs . Because of this shared circulation, the development of monochorionic twins is dependent on each other and, in the event of intrauterine death (IUD) of one twin, the surviving twin has up to a 30–35% risk of consecutive IUD or neurologic sequelae . In the event of single fetal demise, the co-twin has a risk of IUD of 15% (3% in dichorionic twins) and a risk of neurologic injury of 26% (2% in dichorionic twins) and, in general, monochorionic twins are 4.81 times more likely to have neurodevelopmental morbidity .
Treatment of twin–twin transfusion syndrome
Mild-to-moderate cases of TTTS, characterised by discrepant amniotic fluid amounts and moderate differences between the twins in bladder fillings and fetal growth, may progress to any stage of severe TTTS; therefore, fortnightly ultrasound monitoring of these twin pairs and education of the mother of clinical symptoms of TTTS (e.g. rapid uterine growth, increasing discomfort owing to uterine over-distension, and early contractions) are mandatory. Nevertheless, most of these cases remain stable throughout the entire pregnancy, and laser therapy is not needed .
Fetoscopic laser coagulation of placental vascular anastomoses is the treatment of choice for severe TTTS from 16–26 weeks gestation , and is indicated for cases fulfilling the diagnostic sonographic criteria exposed earlier in this issue (See Sueters M., Oepkes D., Diagnosis of twin–twin transfusion syndrome, selective fetal growth restriction, twin anaemia polycythaemia sequence and twin reversed arterial perfusion sequence). Pregnancies fulfilling the Doppler criteria for stage III or echocardiographic criteria for stage IV (e.g. hydrops and advanced heart failure) that present in the late second trimester, as well as cases presenting with clear disease progression, may also undergo laser therapy after 26 weeks . Thus, with increasing experience, the treatment window for TTTS by fetoscopic laser coagulation has become wider from 15 to 29 weeks, although early procedures carry a higher risk for pregnancy loss.
In preparation for fetoscopic laser coagulation, apart from collecting data relevant for diagnosis and staging, a careful ultrasound scan provides the surgeon with a mental spatial map of the possible topography to be found at fetoscopy, by localising the cord insertion sites of the twins, the extent of the placenta, the localisation of the stuck twin, and the probable localisation of the vascular equator. A percutaneous access to the recipient’s amniotic cavity in local anaesthesia under ultrasound guidance, using 2-mm fetoscopes inserted through a previously inserted sheath, has become the standard technique . The insertion site on the maternal abdomen should be chosen opposite to the expected placental vascular equator, so that optimal access to the target vessels can be achieved. In cases with anterior placenta, the entry site has to be chosen through the lateral uterine wall, avoiding the placental edge, maternal uterine vessels, and neighbouring organs, and ideally the collapsed sac of the donor twin. More limited access to the vascular equator, and a tangential angle of view with an unfavourable angle of impact of laser energy on the target vessels, makes these cases more technically challenging. The introduction of 30° fetoscopes for anterior placentas, however, with integrated steering levers for bending the tip of the laser fibre towards the placenta, has shown results, which are comparable to posterior placentas . The insertion line of the inter-twin membrane on the chorionic plate and the cord insertion site of the recipient twin are important intra-amniotic landmarks that provide orientation to the vascular equator. Vessels crossing the line underneath the membrane have to be followed in both directions to check for anastomoses ( Figs. 1 and 2 ). Although the vascular equator will be directly visible from the recipients amniotic cavity ( Fig. 3 ), it may sometimes extend to an area underneath the insertion line of the inter-twin membrane, and some anastomoses may be located in the donor’s amniotic cavity ( Fig. 4 ). In these cases, laser coagulation has to be carried out through the membrane ( Fig. 5 a and b). Because arteries almost always cross over veins and carry-oxygen depleted blood, thus looking darker than light red veins, the direction of blood flow in the vessels can be established during fetoscopy. In a few cases with turbid amniotic fluid, an amnioinfusion with saline solution may improve visual conditions. With these factors in mind, the operator can clearly identify the anastomoses. A 400 μm laser fibre is advanced through to operative channel of the fetoscope, and laser energy is fired at a distance of about 1 cm from the vessels. A Neodymium–Yttrium–Aluminium garnet or diode lasers are commonly used, owing to their optimal energy absorbance in the spectrum of haemoglobin. Selective coagulation of the anastomoses at the vascular equator is the most common method, and some centres prefer a sequential coagulation of the arteriovenous anastomoses (from donor to recipient) first and thereafter the vascular anastomoses (from recipient to donor) . Recently, additional coagulation of the area between the selectively coagulated anastomoses, known as ‘Solomon technique’ ( Fig. 6 a and b), has gained interest, and results of survival rates, recurrence of TTTS, and the incidence of TAPS after laser (a complication seen in about 13% of procedures) are promising . With this technique, thin anastomoses, which may be difficult to detect in cases of turbid amniotic fluid or at the margin of an anterior placenta, may not be missed ( Fig. 7 ). The Solomon technique and a sequential coagulation, however, may lead to longer operating times, but this issue is currently under investigation.
The procedure is always finished with an amnioreduction of the polyhydramnios to a normal amount of amniotic fluid to relieve uterine over-distension. The average operating time is 30 mins, the risk for miscarriage or early delivery caused by preterm premature rupture of membranes (PPROM) is 7%, and delivery takes place at a median gestational age of 34 weeks . A cervical cerclage for a short cervix (<15 mm) directly after the laser procedure can be offered, and may be the only reason for regional anaesthesia . Data are controversial on the benefit of a cerclage after laser therapy; in a recent retrospective study, only a subgroup of patients with a cervix length between 16 and 20 mm showed a significant prolongation of pregnancy . A short cervix (<15 mm) was associated with preterm delivery, regardless of having received a preventive cerclage, thus suggesting an already substantial risk for early delivery. Additionally, pregnancy duration was not affected by the placement of a cerclage in women with a cervix length greater than 20 mm, compared with expectant management, thus suggesting that women whose cervix length is above this cut-off may not be at an increased risk for preterm birth. Therefore, further randomised-controlled studies are warranted to address this issue. In a recent retrospective, multicentre cohort study, Papanna et al. identified risk factors for preterm delivery after fetoscopic laser coagulation. A lower maternal age, history of previous praematurity, shortened cervical length, a larger cannula diameter, amnioinfusion (performed in 41% of cases), and iatrogenic preterm premature rupture of membranes (28%), were significantly associated with a lower gestational age at delivery.
As stage I cases do not necessarily progress to more advanced stages, and some of them even regress to almost balanced amniotic fluid amounts, the indication for fetoscopic laser coagulation in this group of women is controversial. Therefore, currently, a randomised-controlled trial comparing expectant management to primary laser is under way ( http://clinicaltrials.gov/ct2/show/NCT01220011 ). Women assigned to the expectant management arm of the study are followed weekly and, if stage progression or worsening of maternal obstetric parameters owing to massive polyhydramnios occurs, laser therapy is then immediately offered. Apart from exclusion criteria (e.g. short cervix, contractions, maternal discomfort, previous amniocentesis, fetal malformations, and PPROM), only 15% of TTTS cases present with stage I, and many women are referred to laser centres with geographical limitation for weekly follow up. Thus, recruitment seems time consuming.
Twin anaemia-polycythaemia sequence has been observed as a consequence of a few missed thin anastomoses after laser in up to 13% of the procedures. The presence of a few small arteriovenous anastomoses, allowing a slow transfusion, gradually leads to highly discordant haemoglobin levels between the twins, and has been related to TAPS . Antenatal diagnosis is made measuring the peak systolic velocities in the middle cerebral artery, when velocities exceed 1.5 multiples of media in the anaemic twin, together with decreased velocities (<1.0 multiples of media) in the polycythaemic co-twin. Postnatally, TAPS is diagnosed by neonatologists in twins, with marked haemoglobin concentration differences (<11 g/dl in the anaemic neonate and >20 g/dl in the polycythaemic neonate) ; however, some investigators prefer an inter-twin haemoglobin difference greater than 8 g/dl, and an elevated inter-twin reticulocyte count ratio greater than 1.7 . Around 5% of monochorionic twins present spontaneously with this complication in the early third trimester. The reduced incidence of arterio–arterial anastomoses to compensate for this chronic slow transfusion has also been observed in placental injection studies in placentas from TAPS . Even under optimal conditions of visualisation, small anastomoses may be missed. This is currently the rationale for applying the ‘Solomon technique’, a topic now under investigation . In selected TAPS cases, fetoscopic laser coagulation of the usually thin and marginal anastomoses, may represent a treatment option, as long as safe fetoscopic access is achievable . At the gestational age typical for TAPS, however, fetoscopy may be challenging, owing to turbid amniotic fluid and a limited access to the vascular equator because of fetal size and movements. Advanced gestational age at presentation may lead many centres to opt for early delivery. The use of intrauterine blood transfusion to the anaemic twin does not help the polycythaemic co-twin.
In a recent systematic review of TTTS cases treated with selective fetoscopic laser coagulation (SFLC), the rates of recurrence of TTTS range from 0% to 16% . Recurrence was defined in most studies as the reappearance of polyhydramnios and oligohydramnios, and was related to missed anastomoses. In most cases, a secondary intervention with SFLC was carried out, followed in frequency by amnioreduction and delivery. The lack of comparable data and information on perinatal outcomes precluded a more detailed analysis to guide recommendations for clinical management. In our experience, if recurrence shows a rapid progression, and a new entry port allows a different angle of view on the placenta, a secondary SFLC may be indicated. Otherwise, an amnioreduction may lead to advanced gestational ages.
Treatment of twin–twin transfusion syndrome
Mild-to-moderate cases of TTTS, characterised by discrepant amniotic fluid amounts and moderate differences between the twins in bladder fillings and fetal growth, may progress to any stage of severe TTTS; therefore, fortnightly ultrasound monitoring of these twin pairs and education of the mother of clinical symptoms of TTTS (e.g. rapid uterine growth, increasing discomfort owing to uterine over-distension, and early contractions) are mandatory. Nevertheless, most of these cases remain stable throughout the entire pregnancy, and laser therapy is not needed .
Fetoscopic laser coagulation of placental vascular anastomoses is the treatment of choice for severe TTTS from 16–26 weeks gestation , and is indicated for cases fulfilling the diagnostic sonographic criteria exposed earlier in this issue (See Sueters M., Oepkes D., Diagnosis of twin–twin transfusion syndrome, selective fetal growth restriction, twin anaemia polycythaemia sequence and twin reversed arterial perfusion sequence). Pregnancies fulfilling the Doppler criteria for stage III or echocardiographic criteria for stage IV (e.g. hydrops and advanced heart failure) that present in the late second trimester, as well as cases presenting with clear disease progression, may also undergo laser therapy after 26 weeks . Thus, with increasing experience, the treatment window for TTTS by fetoscopic laser coagulation has become wider from 15 to 29 weeks, although early procedures carry a higher risk for pregnancy loss.
In preparation for fetoscopic laser coagulation, apart from collecting data relevant for diagnosis and staging, a careful ultrasound scan provides the surgeon with a mental spatial map of the possible topography to be found at fetoscopy, by localising the cord insertion sites of the twins, the extent of the placenta, the localisation of the stuck twin, and the probable localisation of the vascular equator. A percutaneous access to the recipient’s amniotic cavity in local anaesthesia under ultrasound guidance, using 2-mm fetoscopes inserted through a previously inserted sheath, has become the standard technique . The insertion site on the maternal abdomen should be chosen opposite to the expected placental vascular equator, so that optimal access to the target vessels can be achieved. In cases with anterior placenta, the entry site has to be chosen through the lateral uterine wall, avoiding the placental edge, maternal uterine vessels, and neighbouring organs, and ideally the collapsed sac of the donor twin. More limited access to the vascular equator, and a tangential angle of view with an unfavourable angle of impact of laser energy on the target vessels, makes these cases more technically challenging. The introduction of 30° fetoscopes for anterior placentas, however, with integrated steering levers for bending the tip of the laser fibre towards the placenta, has shown results, which are comparable to posterior placentas . The insertion line of the inter-twin membrane on the chorionic plate and the cord insertion site of the recipient twin are important intra-amniotic landmarks that provide orientation to the vascular equator. Vessels crossing the line underneath the membrane have to be followed in both directions to check for anastomoses ( Figs. 1 and 2 ). Although the vascular equator will be directly visible from the recipients amniotic cavity ( Fig. 3 ), it may sometimes extend to an area underneath the insertion line of the inter-twin membrane, and some anastomoses may be located in the donor’s amniotic cavity ( Fig. 4 ). In these cases, laser coagulation has to be carried out through the membrane ( Fig. 5 a and b). Because arteries almost always cross over veins and carry-oxygen depleted blood, thus looking darker than light red veins, the direction of blood flow in the vessels can be established during fetoscopy. In a few cases with turbid amniotic fluid, an amnioinfusion with saline solution may improve visual conditions. With these factors in mind, the operator can clearly identify the anastomoses. A 400 μm laser fibre is advanced through to operative channel of the fetoscope, and laser energy is fired at a distance of about 1 cm from the vessels. A Neodymium–Yttrium–Aluminium garnet or diode lasers are commonly used, owing to their optimal energy absorbance in the spectrum of haemoglobin. Selective coagulation of the anastomoses at the vascular equator is the most common method, and some centres prefer a sequential coagulation of the arteriovenous anastomoses (from donor to recipient) first and thereafter the vascular anastomoses (from recipient to donor) . Recently, additional coagulation of the area between the selectively coagulated anastomoses, known as ‘Solomon technique’ ( Fig. 6 a and b), has gained interest, and results of survival rates, recurrence of TTTS, and the incidence of TAPS after laser (a complication seen in about 13% of procedures) are promising . With this technique, thin anastomoses, which may be difficult to detect in cases of turbid amniotic fluid or at the margin of an anterior placenta, may not be missed ( Fig. 7 ). The Solomon technique and a sequential coagulation, however, may lead to longer operating times, but this issue is currently under investigation.
