Twin-twin transfusion syndrome

OBJECTIVE: We sought to review the natural history, pathophysiology, diagnosis, and treatment
options for twin-twin transfusion syndrome (TTTS).
METHODS: A systematic review was performed using MEDLINE database, PubMed, EMBASE, and
Cochrane Library. The search was restricted to English-language articles published from 1966 through
July 2012. Priority was given to articles reporting original research, in particular randomized controlled
trials, although review articles and commentaries also were consulted. Abstracts of research presented
at symposia and scientific conferences were not considered adequate for inclusion in this document.
Evidence reports and guidelines published by organizations or institutions such as the National
Institutes of Health, Agency for Health Research and Quality, American College of Obstetricians and
Gynecologists, and Society for Maternal-Fetal Medicine were also reviewed, and additional studies
were located by reviewing bibliographies of identified articles. Consistent with US Preventive Task
Force guidelines, references were evaluated for quality based on the highest level of evidence, and
recommendations were graded accordingly.
RESULTS AND RECOMMENDATIONS: TTTS is a serious condition that can complicate 8-10%
of twin pregnancies with monochorionic diamniotic (MCDA) placentation. The diagnosis of TTTS
requires 2 criteria: (1) the presence of a MCDA pregnancy; and (2) the presence of oligohydramnios
(defined as a maximal vertical pocket of<2 cm) in one sac, and of polyhydramnios (a maximal vertical
pocket of>8 cm) in the other sac. The Quintero staging system appears to be a useful tool for describing
the severity of TTTS in a standardized fashion. Serial sonographic evaluation should be considered for
all twins with MCDA placentation, usually beginning at around 16 weeks and continuing about every
2 weeks until delivery. Screening for congenital heart disease is warranted in all monochorionic twins,
in particular those complicated by TTTS. Extensive counseling should be provided to patients with
pregnancies complicated by TTTS including natural history of the disease, as well as management
options and their risks and benefits. The natural history of stage I TTTS is that more than three-fourths
of cases remain stable or regress without invasive intervention, with perinatal survival of about 86%.
Therefore, many patients with stage I TTTS may often be managed expectantly. The natural history of
advanced (eg, stage ≥III) TTTS is bleak, with a reported perinatal loss rate of 70-100%, particularly
when it presents <26 weeks. Fetoscopic laser photocoagulation of placental anastomoses is considered
by most experts to be the best available approach for stages II, III, and IV TTTS in continuing
pregnancies at <26 weeks, but the metaanalysis data show no significant survival benefit, and the long-
term neurologic outcomes in the Eurofetus trial were not different than in nonlaser treated controls.
Even laser-treated TTTS is associated with a perinatal mortality rate of 30-50%, and a 5-20% chance
of long-term neurologic handicap. Steroids for fetal maturation should be considered at 24 0/7 to 33 6/7
weeks, particularly in pregnancies complicated by stage_III TTTS, and those undergoing invasive
interventions.
Question 1. How is the diagnosis of twin-twin transfusion syndrome made and how is it staged?
(Levels II and III)
Twin-twin transfusion syndrome (TTTS) is diagnosed prenatally by ultrasound. The diagnosis
requires 2 criteria: (1) the presence of a monochorionic diamniotic (MCDA) pregnancy; and (2) the
presence of oligohydramnios (defined as a maximal vertical pocket [MVP] of <2 cm) in one sac, and
of polyhydramnios (a MVP of >8 cm) in the other sac (Figure 1).1 MVP of 2 cm and 8 cm represent
the 5th and 95th percentiles for amniotic fluid measurements, respectively, and the presence of both is
used to define stage I TTTS.2 If there is a subjective difference in amniotic fluid in the 2 sacs that fails
to meet these criteria, progression to TTTS occurs in <15% of cases.3 Although growth discordance
(usually defined as >20%) and intrauterine growth restriction (IUGR) (estimated fetal weight <10% for
gestational age) often complicate TTTS, growth discordance itself or IUGR itself are not diagnostic
criteria.4 The differential diagnosis may include selective IUGR, or possibly an anomaly in 1 twin
causing amniotic fluid abnormality.5 Twin anemia- polycythemia sequence (TAPS) has been recently
described in MCDA gestations, and is defined as the presence of anemia in the donor and polycythemia
in the recipient, diagnosed antenatally by middle cerebral artery (MCA)–peak systolic velocity (PSV)
>1.5 multiples of median in the donor and MCA PSV <1.0 multiples of median in the recipient, in the
absence of oligohydramniospolyhydramnios. Further studies are required to determine the natural
history and possible management of TAPS.
TTTS can occur in a MCDA twin pair in triplet or higher-order pregnancies. The most commonly used
TTTS staging system was developed by Quintero et al2 in 1999, and is based on sonographic findings.
The TTTS Quintero staging system includes 5 stages, ranging from mild disease with isolated
discordant amniotic fluid volume to severe disease with demise of one or both twins (Table 1 and
Figures 2 and 3). This system has some prognostic significance and provides a method to compare
outcome data using different therapeutic interventions.2 Although the stages do not correlate perfectly
with perinatal survival, 7 it is relatively straightforward to apply, may improve communication between
patients and providers, and identifies the subset of cases most likely to benefit from treatment.8,9 Since
the development of the Quintero staging system, much has been learned about the changes in fetal
cardiovascular physiology that accompany disease progression (discussed below). Myocardial
performance abnormalities have been described, particularly in recipient twins, including those with
only stage I or II TTTS.10 Several groups of investigators have attempted to use assessment of fetal
cardiac function to either modify the Quintero TTTS stage11 or develop a new scoring system.12 While
this approach has some benefits, the models have not yet been prospectively validated. As a result, a
recent expert panel concluded that there were insufficient data to recommend modifying the Quintero
staging system or adopting a new system.8 Thus, despite debate over the merits of the Quintero system,
at this time it appears to be a useful tool for the diagnosis of TTTS, as well as for describing its severity,
in a standardized fashion.
FIGURE 1
Polyhydramnios-oligohydramnios sequence

Monochorionic diamniotic twins with twin-twin transfusion syndrome demonstrating polyhydramnios

in recipient’s sac (twin A) while donor (twin B) was stuck to anterior uterine wall due to marked
oligohydramnios.
FIGURE 2
Stage II twin-twin transfusion syndrome

Nonvisualization of fetal bladder (arrow) between umbilical arteries in donor twin.

FIGURE 3
Stage III twin-twin transfusion syndrome

Absent end-diastolic flow (arrows) in umbilical artery of donor twin.

TABLE 1
Staging of twin-twin transfusion syndrome2
Stage Ultrasound parameter Categorical criteria
I MVP of amniotic fluid MVP <2 cm in donor sac;
MVP >8 cm in recipient sac
II Fetal bladder Nonvisualization of fetal
bladder in donor twin over 60
min of observation (Figure 2)
III Umbilical artery, ductus Absent or reversed umbilical
venosus, and umbilical vein artery diastolic flow, reversed
Doppler waveforms ductus venosus a-wave flow,
pulsatile umbilical vein flow
(Figure 3)
IV Fetal hydrops Hydrops in one or both twins
V Absent fetal cardiac activity Fetal demise in one or both
twins
Question 2. How often does TTTS complicate monochorionic twins and what is its natural
history? (Levels II and III)
Approximately one-third of twins are monozygotic (MZ), and three-fourths of MZ twins are
MCDA. In general, only twin gestations with MCDA placentation are at significant risk for TTTS,
which complicates about 8-10% of MCDA pregnancies.13,14 TTTS is very uncommon in MZ twins
with dichorionic or monoamniotic placentation.15 Although most twins conceived with in vitro
fertilization (IVF) are dichorionic, it is important to remember that there is a 2- to 12-fold increase in
MZ twinning in embryos conceived with IVF, and TTTS can therefore occur for IVF MCDA
pregnancies.16,17 In current practice, the prevalence of TTTS is approximately 1-3 per 10,000 births.18
The presentation of TTTS is highly variable. Because pregnancies with TTTS often receive care at
referral centers, data about the stage of TTTS at initial presentation (ie, to nonreferral centers) are
lacking in the literature. Fetal therapy centers report that about 11-15% of their cases at referral were
Quintero stage I (probably underestimated as some referral centers did not report stage I TTTS cases),
20-40% were stage II, 38-60% were stage III, 6-7% were stage IV, and 2% were stage V.5,9 Although
TTTS may develop at any time in gestation, the majority of cases are diagnosed in the second trimester.
Stage I may progress to a nonvisualized fetal bladder in the donor (stage II) (Figure 2), and absent or
reversed end-diastolic flow in the umbilical artery of donor or recipient twins may subsequently develop
(stage III) (Figure 3), followed by hydrops (stage IV). However, TTTS often does not progress in a
predictable manner. Natural history data by stage are limited, especially for stages II-V, as staging was
initially proposed in 1999.2 This is because most natural history data were published before 1999, and
therefore was not stratified by stage (Table 2).19-21 Over three fourths of stage I TTTS cases remain
stable or regress without invasive interventions (Table 2).19-21 The natural history of advanced (eg,
stage ≥III) TTTS is bleak, with a reported perinatal loss rate of 70-100%, particularly when it presents
<26 weeks.22,23 It is estimated that TTTS accounts for up to 17% of the total perinatal mortality in
twins, and for about half of all perinatal deaths in MCDA twins.13,24 Without treatment, the loss of at
least 1 fetus is common, with demise of the remaining twin occurring in about 10% of cases of twin
demise, and neurologic handicap affecting 10-30% of cotwin remaining survivors. 25-27 Overall, single
twin survival rates in TTTS vary widely between 15-70%, depending on the gestational age at diagnosis
and severity of disease.22,26 The lack of a predictable natural history, and therefore the uncertain
prognosis for TTTS, pose a significant challenge to the clinician caring for MCDA twins.
TABLE 2
Natural history of stage I twin-twin transfusion syndrome19-21
Stage Incidence of progression to Incidence of resolution, Overall survival
higher stage regression to lower stage, or
stability
I 6/39 (15%) 33/39 (85%) 102/118 (86%)

Question 3. What is the underlying pathophysiology of TTTS? (Levels II and III)

The primary etiologic problem underlying TTTS is thought to lie within the architecture of the
placenta, as intertwin vascular connections within the placenta are critical for the development of TTTS.
Virtually all MCDA placentas have anastomoses that link the circulations of the twins, yet not all
MCDA twins develop TTTS. There are 3 main types of anastomoses in monochorionic placentas:
venovenous (VV), arterioarterial (AA), and arteriovenous (AV). AV anastomoses are found in 90-95%
of MCDA placentas, AA in 85-90%, and VV in 15-20%.28,29 Both AA and VV anastomoses are direct
superficial connections on the surface of the placenta with the potential for bidirectional flow (Figure
4). In AV anastomoses, while the vessels themselves are on the surface of the placenta, the actual
anastomotic connections occur in a cotyledon, deep within the placenta (Figure 4). AV anastomoses
can result in unidirectional flow from one twin to the other, and if uncompensated, may lead to an
imbalance of volume between the twins. Unlike AA and VV, which are direct vessel-to-vessel
connections, AV connections are linked through large capillary beds deep within the cotyledon. AV
anastomoses are usually multiple and overall balanced in both directions so that TTTS does not occur.
While the number of AV anastomoses from donor to recipient may be important, their size as well as
placental resistance likely influences the volume of intertwin transfusion that occurs.30 Placentas in
twins affected with TTTS are reportedly more likely to have VV, but less likely to have AA
anastomoses.28 It is thought that these bidirectional anastomoses may compensate for the unidirectional
flow through AV connections, thereby preventing the development of TTTS or decreasing its severity
when it does occur.31 Mortality is highest in the absence of AA and lowest when these anastomoses
are present (42% vs 15%).29 However, the presence of AA is not completely protective, as about 25-
30% of TTTS cases may also have these anastomoses.32 The imbalance of blood flow through the
placental anastomoses leads to volume depletion in the donor twin, with oliguria and oligohydramnios,
and to volume overload in the recipient twin, with polyuria and polyhydramnios. There also appear to
be additional factors beyond placental morphology, such as complex interactions of the renin-
angiotensin system in the twins,33-35 involved in the development of this disorder.
FIGURE 4
Selected anastomoses in monochorionic placentas

a-a, arterioarterial anastomosis; a-v, arteriovenous anastomosis; v-a, venous-arterial anastomosis.

Question 4. How should monochorionic twin pregnancies be monitored for the development of
TTTS? (Levels II and III)
All women with a twin pregnancy should be offered an ultrasound examination at 10-13 weeks
of gestation to assess viability, chorionicity, crown-rump length, and nuchal translucency. TTTS usually
presents in the second trimester, and is a dynamic condition that can remain stable throughout gestation,
occasionally regress spontaneously, progress slowly over a number of weeks, or develop quickly within
a period of days with rapid deterioration in the well-being of the twins. There have been no randomized
trials of the optimal frequency of ultrasound surveillance of MCDA pregnancies to detect TTTS.
Although twin pregnancies are often followed up with sonography every 4 weeks, sonography as often
as every 2 weeks has been proposed for monitoring of MCDA twins for the development of TTTS.36-
38 This is in part because, while stage I TTTS has been observed to remain stable or resolve in most
cases, when progression does occur it can happen quickly.39 However, studies that have
focusedonprogression of early-stage TTTS may not be applicable to the question of disease
development in apparently unaffected pregnancies.
Given the risk of progression from stage I or II to more advanced stages, and that TTTS usually
presents in the second trimester, serial sonographic evaluations about every 2 weeks, beginning usually
around 16 weeks of gestation, until delivery, should be considered for all twins with MCDA
placentation, until more data are available allowing better risk stratification37,38 (Figure 5).
Sonographic surveillance less often than every 2 weeks has been associated with a higher incidences of
latestage diagnosis of TTTS.40 This underscores the importance of establishing chorionicity in twin
pregnancies as early as possible.41 These serial sonographic evaluations to screen for TTTS should
include at least MVP of each sac, and the presence of the bladder in each fetus. Umbilical artery Doppler
flow assessment, especially if there is discordance in fluid or growth, is not unreasonable, but data on
the utility of this added screening parameter are limited. There is no evidence that monitoring for TAPS
with MCA PSV Doppler at any time, including >26 weeks, improves outcomes, so that this additional
screening cannot be recommended at this time.6
In addition to monitoring MCDA pregnancies for development of amniotic fluid abnormalities,
there are several second- and even first-trimester sonographic findings that have been associated with
TTTS. These findings are listed in Table 3.28,42-49 Before 14 weeks, MCDA twins can be evaluated
with nuchal translucency and crown lump length. Nuchal translucency abnormalities and crown-lump
length discrepancy have been associated with an increased risk of TTTS.28,29,38 If such findings
(Table 3) are encountered, it may be reasonable to perform more frequent surveillance (eg, weekly
instead of every 2 weeks) for TTTS. Velamentous placental cord insertion (Figure 6) has been found in
approximately one third of placentas with TTTS.28 Intertwin membrane folding (Figure 7) has been
associated with development of TTTS in more than a third of cases. The clinical utility of the
sonographic findings listed in Table 3 has not been prospectively evaluated, and several require Doppler
evaluation not typically performed in otherwise uncomplicated MCDA gestations. Thus, while they are
associated with TTTS and may potentially improve TTTS detection, they are not specifically
recommended as part of routine surveillance.

Question 5. Is there a role for fetal echocardiography in TTTS? (Levels II and III)
Screening for congenital heart disease with fetal echocardiography is warranted in all
monochorionic twins as the risk of cardiac anomalies is increased 9-fold in MCDA twins and up to 14-
fold in cases of TTTS, above the population prevalence of approximately 0.5%.54 Specifically, the
prevalence of congenital cardiac anomalies has been reported to be 2% in otherwise uncomplicated
MCDA gestations and 5% in cases of TTTS, particularly among recipient twins.55 Although many
cases are minor septal defects, an increase in right ventricular outflow tract obstruction has also been
reported.55 It is theorized that the abnormal placentation that occurs in monochorionic twins,
particularly in cases that develop TTTS, contributes to abnormal fetal heart formation.54
The functional cardiac abnormalities that complicate TTTS occur primarily in recipient twins.
Volume overload causes increased pulmonary and aortic velocities, cardiomegaly, and atrioventricular
valve regurgitation (Figure 8). Over time, recipient twins can develop progressive biventricular
hypertrophy and diastolic dysfunction as well as poor right ventricular systolic function that can lead to
functional right ventricular outflow tract obstruction and pulmonic stenosis (Figure 9). The
development of right ventricular outflow obstruction, observed in close to 10% of all recipient twins, is
likely multifactorial, a consequence of increased preload, afterload, and circulating factors such as
renin, angiotensin, endothelin, and atrial and brain natriuretic peptides. The cardiovascular response to
TTTS contributes to the poor outcome of recipient twins while recipients with normal cardiac function
have improved survival.60
A functional assessment of the fetal heart may be useful in identifying cases that would benefit
from therapy and in evaluating the response to treatment. The myocardial performance index or Tei
index, an index of global ventricular performance by Doppler velocimetry, is a measure of both systolic
and diastolic function,61 and has been used to monitor fetuses with TTTS.62 Donor twins with TTTS
tend to have normal cardiac function, whereas recipient twins may develop ventricular hypertrophy
(61%), atrioventricular valve regurgitation (21%), and abnormal right ventricular (50%) or left
ventricular (58%) function.11,58 Overall, two thirds of recipient twins show diastolic dysfunction, as
indicated by a prolonged ventricular isovolumetric relaxation time, which is associated with an
increased risk of fetal death.58
Although fetal cardiac findings are not officially part of the TTTS staging system, many centers
routinely perform fetal echocardiography in cases of TTTS and have observed worsening cardiac
function in advanced stages.11 However, cardiac dysfunction can also be detected in up to 10% of
apparently early-stage TTTS.11 It has been theorized that the early diagnosis of recipient twin
cardiomyopathy may identify those MCDA gestations that would benefit from early intervention. In
summary, scoring systems that include cardiac dysfunction have been developed, but their usefulness
to predict outcome in TTTS remains controversial. 63,64 Further evaluation of functional fetal
echocardiography as a tool for decision-making about intervention and management in TTTS is needed.
TABLE 3
First- and second-trimester sonographic findings associated with twin-twin transfusion
syndrome
First-trimester findings
Crown-rump length discordance
................................................................................................................................................................
Nuchal translucency >95th percentile or discordance >20% between twins
................................................................................................................................................................
Reversal or absence of ductus venosus A-wave
................................................................................................................................................................
Second-trimester findings
Abdominal circumference discordance
................................................................................................................................................................
Membrane folding
................................................................................................................................................................
Velamentous placental cord insertion (donor twin)
................................................................................................................................................................
Placental echogenicity (donor portion hyperechoic)
................................................................................................................................................................

Question 6. What management options are available for TTTS? (Levels I, II, and III)
The management options described for TTTS include expectant management, amnioreduction,
intentional septostomy of the intervening membrane, fetoscopic laser photocoagulation of placental
anastomoses, and selective reduction. The interventions that have been evaluated in randomized
controlled trials (RCTs) include intentional septostomy of the intervening membrane to equalize the
fluid in both sacs, amnioreduction of the excess fluid in the recipient’s sac, and laser ablation of
placental anastomoses. There have been 3 randomized trials designed to evaluate some of the different
treatment modalities for TTTS, all of which were terminated prior to recruitment of the planned subject
number after interim analyses, as discussed below.65-67 Despite the limitations and early termination
of these clinical trials, they represent the best available data upon which to judge the various treatments
for TTTS. Consultation with a maternal-fetal medicine specialist is recommended, particularly if the
patient is at a gestational age at which laser therapy is potentially an option. In evaluating the data,
considerations include the stage of TTTS, the details of the intervention, and the perinatal outcome. The
most important outcomes reported are overall perinatal mortality, survival of at least 1 twin, and, if
available, long-term outcomes of the babies, including neurologic outcome. Extensive counseling
should be provided to patients with pregnancies complicated by TTTS, including natural history of the
disease, as well as management options and their risks and benefits.
Expectant management involves no intervention. This natural history of TTTS, also called
conservative management, has limited outcome data according to stage, particularly for advanced
disease (Table 2). It is important that the limitations in the available data are discussed with the patient
with TTTS, and compared with available outcome data for interventions.
Amnioreduction involves the removal of amniotic fluid from the polyhydramniotic sac of the
recipient. It is usually done only when the MVP is >8 cm, with an aim to correct it to a MVP of <8 cm,
often to <5 cm or <6 cm.65-67 Usually an 18-65 or 2067-gauge needle is used. Some practitioners use
aspiration with syringes, while some use vacuum containers.66 Amnioreduction can be performed
either as a 1-time procedure, as at times this can resolve stage I or II TTTS, or serially, eg, every time
theMVP is >8 cm. It can be performed any time >14 weeks. Amnioreduction is hypothesized to reduce
the intraamniotic and placental intravascular pressures, potentially facilitating placental blood flow,
and/or to possibly reduce the incidence of preterm labor and birth related to polyhydramnios.
Amnioreduction may be used also >26 weeks, particularly in cases with maternal respiratory distress
or preterm contractions from polyhydramnios.68 Amnioreduction has been associated with average
survival rates of 50%, with large registries reporting 60-65% overall survival.69,70 However, serial
amnioreduction is often necessary, and repeated procedures increase the likelihood of complications
such as preterm premature rupture of themembranes,preterm labor, abruption, infection, and fetal
death.71 Another consideration is that any invasive procedure prior to fetoscopy may decrease the
feasibility and success of laser due to bleeding, chorioamnion separation, inadvertent septostomy, or
membrane rupture.
Septostomy involves intentionally puncturing with a needle the amniotic membranes between
the 2 MCDA sacs, theoretically allowing equilibration of amniotic fluid volume in the 2 sacs.66 In the
1 randomized trial in which it was evaluated, the intertwin membrane was purposefully perforated under
ultrasound guidance with a single puncture using a 22-gauge needle.66 This was usually introduced
through the donor’s twin gestational sac into the recipient twin’s amniotic cavity. If reaccumulation of
amniotic fluid in the donor twin sac was not seen in about 48 hours, a repeat septostomy was
undertaken.66 Intentional septostomy is mentioned only to note that it has generally been abandoned as
a treatment for TTTS. It is believed to offer no significant therapeutic advantage, and may lead to
disruption of the membrane and a functional monoamniotic situation. A randomized trial of
amnioreduction vs septostomy ended after an interim analysis found that the rate of survival of at least
1 twin was similar between the 2 groups, and that recruitment had been slower than anticipated 66
(Table 4). In all, 97% of the enrolled pregnancies had stages I-III TTTS, and results were not otherwise
reported by stage. In 40% of the septostomy cases, additional procedures were needed. No data on
neurologic outcome are available.66
Laser involves photocoagulating the vascular anastomoses crossing from one side of the
placenta to the other. This is usually performed by placing a sheath and passing an endoscope under
ultrasound guidance. Ultrasound is also used to map the vasculature to determine the placental
angioarchitecture. The primary theoretical advantage of laser coagulation is that it is designed to
interrupt the placental anastomoses that give rise to TTTS. The goal of laser ablation is to functionally
separate the placenta into 2 regions, each supplying one of the twins. This unlinking of the circulations
of the twins is often referred to as “dichorionization” of the monochorionic placenta. Adequate
visualization of the vascular equator that separates the cotyledons of one twin from the other is critical
for laser photocoagulation. Selective coagulation of AV as well as AA and VV anastomoses is preferred
over nonselective ablation of all vessels crossing the separating membrane as it appears to lead to fewer
procedure-related fetal losses.72 Sequential coagulation of the donor artery to recipient vein followed
by recipient artery to donor vein may theoretically allow some return of fluid from the recipient to the
donor prior to severing other connections.73,74 Criteria for laser have included MCDA pregnancies
between about 15-26 weeks with the recipient twin having MVP ≥8.0 cm at ≤20 weeks or ≥10.0 cm at
>20 weeks and a distended fetal bladder, and donor twin having MVP ≤2.0 cm in 1 trial,65 and MCDA
pregnancies at <24 weeks with the recipient twin having MVP >8 cm, and donor twin having MVP ≤2
cm and nonvisualized fetal bladder in the other.67 There is insufficient evidence to recommend
management in MCDA pairs with TTTS in higher-order multiple gestations, but laser has been
proposed as feasible and effective.75
 Selective reduction involves purposefully interrupting umbilical cord blood flow of 1 twin,
causing the death of this twin, with the purpose of improving the outcome of the other surviving twin.
Usually the cord occlusion is performed with radiofrequency ablation or cord coagulation, but other
procedures have been employed.76 Obviously this option can be associated with a maximum of 50%
overall survival, so, if ever considered, it is usually reserved for stages III or IV TTTS only.
FIGURE 6
Abnormal placental cord insertion

A, Velamentous or membranous placental cord insertion (PCI) (arrow) of monochorionic diamniotic

twin detected by color Doppler. B, Velamentous PCI confirmed on examination of placenta with
identification of anastomosis (arrows) passing beneath separating membrane and joining circulations
of twins.
FIGURE 7
Membrane folding

Membrane folding (arrow) suggestive of discordant amniotic fluid volume in monochorionic

diamniotic twin gestation.
FIGURE 8
Cardiac dysfunction in recipient twin

Color flow imaging demonstrating forward flow across atrioventricular valves in diastole and severe
tricuspid regurgitation (arrow) during systole in recipient twin.
FIGURE 9
Recipient twin cardiomyopathy
TABLE 4
Randomized trial of septostomy vs amnioreduction
variable Septostomy N=35 Amnioreduction N=36 P value
Mean gestational age at delivery, wk 30,7 29,5 .24
Survival of at least 1 twin at 28 d of 80% (28/35) 78 (28/36) .82
age
All perinatal deaths up to 28 d of age 30% (21/70) 36% (26/72) .40

Question 7. What are the management recommendations according to stage? (Levels I, II, and
III)
Stage I
There is no randomized trial specifically including stage I TTTS patients managed without
interventions, ie, expectantly or conservatively managed. Patients with stage I TTTS are often managed
expectantly, as over three-fourths of cases remain stable or regress spontaneously (Figure10).19-21
Because stage I TTTS progresses to more advanced TTTS in 10- 30% of cases, interventions have been
evaluated.
Stages I and IITTTShave been shown to regress following amnioreduction in up to 20-30% of
cases, a rate that is not significantly different than with expectant management, especially for stage
I.20,66
Laser has been studied for stage I TTTS in only 6 patients in the Eurofetus trial,65 and no
patients in the Eunice Kennedy Shriver National Institute of Child Health and Human Development
(NICHD) RCT.67 Only limited data exist from nonrandomized studies.8,9,20,39 In a metaanalysis of
stage I TTTS treated with laser photocoagulation, survival of both twins occurred in 45 of 60 twin pairs
(75%), with an 83% overall survival, rates that are similar to other management strategies including
expectant management, therefore providing no added benefit.9 In a review of the literature including
only stage I TTTS, the overall survival rates were 86% after expectant management, 77% after
amnioreduction, and 86% after laser therapy, leading the investigators to suggest that conservative
management in stage I TTTS is a reasonable option.20 The progression to higher stage was only 15%
for stage I after expectant management, and survival was similar if laser was employed as first- or
second-choice therapy in this review.20 Further studies are needed to determine the optimal
management of stage I TTTS.

Stages II, III, and IV

Currently, fetoscopic laser photocoagulation of placental anastomoses is considered by most
experts to be the best available approach for stages II, III, and IV TTTS in continuing pregnancies at
<26 weeks (Figure 10), but metaanalysis data show no survival benefit, and the long-term neurologic
outcomes in Eurofetus were not different than in nonlasertreated controls. There is no randomized trial
specifically including a group of TTTS patients with stages II, III, and IV, managed without
interventions, ie, expectantly. Data on natural history for stage ≥II are not available (Table 2).
Two randomized trials have evaluated the effectiveness of laser therapy in pregnancies
complicated by TTTS. In the first, called the Eurofetus trial, inclusion criteria were MCDA pregnancies
between 15 and 25 6/7 weeks with the recipient twin having MVP ≥8.0 cm at ≤20 weeks or ≥10.0 cm
at >20 weeks and a distended fetal bladder, and donor twin havingMVP ≤2.0 cm.Atotal of 142 women
were randomized from 3 centers in Europe (90% in France) to either selective laser photocoagulation
or serial amnioreduction. The trial was stopped after an interim analysis demonstrated laser to be
superior to amnioreduction with improved perinatal survival and fewer short-term neurologic
abnormalities. Over 90% of the patients randomized had either stage II or III TTTS (6 with stage I; only
2 with stage IV). The laser group also did have an initial amnioreduction at laser surgery. Eleven women
(16%) vs no women (0%) had voluntary termination of pregnancy after being randomized to
amnioreduction and laser, respectively. Selected results are shown in Table 5.
In the second trial, sponsored by the NICHD, inclusion criteria were MCDA pregnancies at <24
weeks with the recipient twin havingMVP >8 cm, and donor twin having MVP ≤2 cm and nonvisualized
empty fetal bladder. Stage I TTTS was therefore not included. A single diagnostic and therapeutic
qualifying amnioreduction was performed on all pregnancies. This trial was also terminated early due
to poor recruitment as well as increased neonatal mortality of recipient twins treated with laser therapy.
Ninety percent of the patients randomized had either stage II or III TTTS. Three US centers participated
(Children’s Hospital of Philadelphia; University of California, San Francisco; and Cincinnati Children’s
Hospital Medical Center). The laser group also had an initial amnioreduction at laser surgery. Selected
results are shown in Table 6. Infant outcome is available for this trial only up to 30 days of age. While
the survival of at least 1 twin was comparable to the Eurofetus trial for the laser groups (65% in NICHD
vs 76% Eurofetus), this outcome in the amnioreduction groups was better in the NICHD (75%)
compared to the Eurofetus study (56%). The better NICHD amnioreduction results may be due to the
standardized aggressive protocol used (performed every time the MVP was >8 cm). In contrast, the less
favorable NICHD laser results may have been due to the severity of TTTS cardiomyopathy, especially
in the recipients; the fact that there were more stage IV TTTS cases in NICHD (n = 4) than in Eurofetus
(n = 2); and that the upper gestational age for inclusion was also different inNICHD (<24 weeks) vs
Eurofetus (<26 weeks). Recipient twin mortality was significantly higher in the laser (70%) than the
amnioreduction (35%) group (Table 6). In a metaanalysis of these 2 trials, overall death was not
significantly different between laser and amnioreduction (risk ratio, 0.81; 95% confidence interval,
0.65– 1.01). These data on laser apply mostly to stage II and III TTTS, given the very limited number
of stage I or IV TTTS included in the 2 trials.
In summary, laser therapy has been associated with some perinatal benefits in 1 European
trial, which had some limitations, while no benefits were seen in another smaller US trial.
Like all invasive procedures, laser has been associated with complications, including preterm
premature rupture of the membranes, preterm delivery, amniotic fluid leakage into the maternal
peritoneal cavity, vaginal bleeding and/or abruption, and chorioamnionitis. Fetoscopy equipment is of
larger gauge than the spinal needles used for amnioreduction or septostomy and, as a result, the risks of
complications are up to 3-fold higher. In the Eurofetus trial, the overall risk for most complications was
about 3%.65 Maternal and perinatal risks can be particularly high in inexperienced hands. Despite these
risks, fetoscopic laser photocoagulation appears to be the optimal treatment for stage II-IV TTTS.
However, it is important to remember that even with laser therapy, intact survival of both twins with
TTTS is only about 50% (Table 7).
Expectant management and amnioreduction remain 2 options in cases of TTTS stage >I at <26
weeks of gestation, in which the patient does not have the ability to travel to a center that performs
fetoscopic laser photocoagulation.
In cases complicated by severe unequal placental sharing with marked discordant growth and
IUGR, major malformations affecting 1 twin, or evidence of brain injury either before or subsequent to
laser, selective reduction by cord occlusion76 or by termination of the entire pregnancy may be
reasonable management choices for the patient and her family <24 weeks’ gestation.

StageV
In cases of stage V TTTS, ie, death of 1 twin, no intervention has been evaluated in randomized
trials to try to ameliorate outcome. As stated above, in cases of death of 1 MCDA twin, the risks to the
cotwin included a 10% risk of death and 10-30% risk of neurologic complications (Figure 10). It may
be that the abnormal neurologic outcome in some survivors of TTTS is more correlated to whether or
not there was demise of a cotwin, than the actual modality used to treat the condition. It is well
recognized that death of 1 twin of a monochorionic pair can result in periventricular leukomalacia,
intraventricular hemorrhage, hydrocephaly, and porencephaly. Prior laser ablation appears to improve
neurologic outcomes in the survivor if there is a cotwin demise.
FIGURE 10
Algorithm for management of TTTS
MCDA pregnancy with MVP <2 cm in 1
sac and MVP <8 cm in other sac:
Diagnosis = TTTS

Do staging (Table 1): check fetal

bladder, UA Doppler

Stage I Stage II, III, IV Stage V

Counseling. Consider Counseling. Consider

referral to fetal center for Counsel regarding cotwin
expectant management,
laser treatment at 16-25 10% risk of death and 10-
with fetal bladder, UA
6/7 weeks; if unable or 30% risk of neurologic
Doppler, and hydrops
outside eligibility criteria, complications. Consider
ultrasonographic checks
consider amnioreduction expectant management.
at least once per week

TABLE 5
Randomized trial of laser photocoagulation vs amnioreduction (Eurofetus)

a Of women in amnioreduction group, 11 (16%) had voluntary termination of pregnancy between 21

25 wk; b Includes only children delivered in France and still alive at 6 mo of age.
TABLE 6
Randomized trial of laser photocoagulation vs amnioreduction (NICHD-sponsored)

Question 8. After in utero laser for TTTS, what is the expected survival and long-term outcome
of the twins? (Levels II and III)
In general, overall survival rates of 50-70% can be expected after fetoscopic laser for the
treatment of TTTS. Overall perinatal survival of fetuses with TTTS treated with laser was 56% in the
Eurofetus trial at 6 months of age,65 and 45% in the NICHD trial at 30 days67 (Tables 5 and 6,
respectively). The Eurofetus trial reported an 86% survival rate of at least 1 fetus for combined stage I
and II disease treated with laser, decreasing to 66% for combined stage III and IV. In recent
nonrandomized large series, summarizing >1000 cases of TTTS (about 86% with stages II and III)
treated with laser, the overall perinatal survival was about 65% (Table 7). Given publication bias, these
data probably represent the best current possible outcomes with this procedure.
Although the risk of membrane rupture may be as low as 10% in experienced centers, there
remains a 10-30% procedure-associated fetal loss with laser. Both double and single fetal demise are
common complications in advanced stages of TTTS treated with laser (Table 7). In a multicenter
observational study, fetal demise occurred in 24% of donors and in 17% of recipients after laser.
Survival of 1 or 2 fetuses after laser may depend on coexisting unequal placental sharing that may not
be visible before or even at the time of fetoscopy. Preoperative IUGR with absent or reversed end-
diastolic flow in the umbilical artery has a 20-40% increased risk of postoperative donor demise.
Recipient twin demise after laser is more common when the recipient has IUGR, reversed a-wave in
the ductus venosus, or hydrops. Improved recipient twin survival has been reported with the maternal
administration of nifedipine 24-48 hours prior to laser photocoagulation in cases of TTTS
cardiomyopathy, but more data are needed to suggest its use in this clinical situation. After successful
laser photocoagulation, the cardiac function of recipient twins tends to normalize in about 4 weeks.
Pulmonic valve abnormalities, affecting about20%of recipient twins with advanced TTTS, have also
been observed to improve after laser with less than a third of surviving twins having persistent pulmonic
valve defects requiring treatment after birth. Overall, 87% of postlaser recipient twins who survived
were reported to have normal echocardiograms at a median age just under 2 years.
Although procedure-related fetal loss is a recognized complication of fetoscopic laser
photocoagulation, survival with neurologic handicap is also a serious long-term sequela of TTTS, with
or without treatment. While the gestational age at delivery is a significant risk factor for adverse
neurologic outcome, initial studies suggested that neurologic outcomes may be better for those cases
managed with laser photocoagulation, compared to amnioreduction. Infants in the laser group of the
Eurofetus trial had a lower incidence of cystic periventricular leukomalacia and were more likely to be
free of neurologic complications at 6 months of age compared to those treated with amnioreduction
(Table 6). However, 6-year follow-up of 120 children from this trial found that laser therapy conferred
no significant benefit in terms of difference in major neurologic handicap among TTTS survivors
treated with laser vs amnioreduction.77 Another recent study also reported no difference in
neurodevelopmental outcome at 2 years of age among donors and recipients treated with laser or
amnioreduction, although they did observe a trend of increased major neurologic impairment in
survivors after amnioreduction compared to those treated with laser (9.5% vs 4.6%).
Overall, rates of long-term neurologic sequelae in laser-treated stage I TTTS are reported to be
about ≤3%, with rates of about 5-20% in survivors of any stage TTTS (Table 8). The risk of abnormal
neurodevelopment seems to be similar in donor and recipient survivors, and not drastically different
between those treated with laser or amnioreduction. Antenatally acquired severe brain lesions, including
cystic periventricular leukomalacia and grade-3 or -4 intraventricular hemorrhage, affect 10% of TTTS
compared to 2% of MCDA twins without TTTS (P = .02); this difference was seen to persist in findings
seen on cranial ultrasounds at the time of hospital discharge (14% vs 6%, P = .04). Other risk factors
for neurodevelopmental impairment in TTTS survivors are advanced gestational age at laser surgery,
low birth weight, and severe TTTS. Both ultrasound and magnetic resonance imaging (MRI) can be
used to evaluate abnormalities of the fetal brain. In general, fetal MRI to evaluate cortical development
and assess for ischemic injury is best in the third trimester. Following single twin demise in a MCDA
gestation, neurologic injury, when present in the surviving twin, may be detected by ultrasound in about
1-2 weeks, and by MRI as early as 1-2 days after the demise of the other twin. Routine neuroimaging
with MRI cannot yet be recommended given the limited data on benefit, although this has been
suggested by some authors for TTTS both prior to and after therapeutic interventions, or in cases
complicated by single twin demise. Follow-up studies of all survivors of TTTS are critical to determine
accurate long-term outcomes and stage-specific rates of neurologic handicap of these complicated
MCDA pregnancies.

In summary, even with the laser treatmentoption available,TTTSis still a severe condition in
terms of perinatal outcomes. Given the 30-50% chance of overall perinatal death and 5-20% chance of
neurologic handicap long-term, twin death or neurologic handicap is the outcome in up to two thirds of
laser-treated TTTS.
TABLE 8
Long-term neurologic outcome of laser-treated twin-twin transfusion syndrome survivors
Question 9. What antenatal monitoring should be suggested for pregnancies complicated by
TTTS? (Levels II and III)
There are no randomized trials to evaluate the effectiveness of antenatal monitoring for
pregnancies complicated by TTTS. Weekly monitoring of the umbilical artery Doppler flow andMVPof
amniotic fluid of each fetus may be considered. The evidence for effectiveness of serial (eg, weekly or
twice/wk) nonstress tests, biophysical profiles, and other antenatal testing modalities is insufficient to
make a recommendation, but these tests can be considered.
One reason for surveillance, even following laser therapy, is that not all anastomoses are ablated
at the time of laser. Residual anastomoses, either initially undetected, missed, or revascularized after
laser, have been observed in up to a third of cases. Placental casting has also demonstrated the presence
of deep, atypical AV anastomoses beneath the chorionic plate that would not be visible by fetoscopy.
Failure to coagulate all AV anastomoses can lead to persistent, recurrent or reversed TTTS. Persistent
or recurrent TTTS has been reported in 14% of cases postlaser and reversed TTTS, with the recipient
becoming anemic and the donor polycythemic, in 13% of cases. While TAPS can occur spontaneously
in a MCDA gestation, it is a known iatrogenic complication of laser. Screening by transvaginal
ultrasound for short cervical length in TTTS cases has also been proposed, as this is associated with
preterm birth, a known complication of TTTS. As there are no interventions shown to improve outcome
based on short transvaginal ultrasound cervical length in TTTS cases, this screening cannot be
recommended at this time.

Question 10. When should patients with TTTS be delivered? (Levels II and III)
MCDA pregnancies complicated by TTTS are at increased risk of several complications,
including but not limited to preterm birth, fetal demise, and cerebral injury. Because of the increased
risk of preterm birth, 1 course of steroids for fetal maturation should be considered at 24 to 33 6/7
weeks, particularly in pregnancies complicated by stage ≥III TTTS, and those undergoing invasive
interventions.
There are no clinical trials regarding optimal timing of delivery for TTTS pregnancies. This
depends on several factors, including disease stage and severity, progression, effect of interventions (if
any), and results of antenatal testing. Recommendations regarding timing of delivery with TTTS vary,
with some endorsing planned preterm delivery as early as 32-34 weeks, and others individualizing care
and allowing gestation to progress to 34-37 weeks, particularly in cases of mild disease (eg, stages I
and II) with reassuring surveillance.
The median gestational age at delivery in the major trials and case series of lasertreated TTTS
has been about 33-34 weeks (Table 7). Cases treated with laser generally have more advanced disease,
and they may be at risk for early delivery due to both TTTS and procedure-related complications.
However, prematurity has been identified as an independent risk factor for neurodevelopmental
impairment in the setting of TTTS. Given the spectrum of disease associated with TTTS, many variables
factor into decisions about timing of delivery, including disease stage, progression, response to
treatment, fetal growth, and results of antenatal surveillance. Delaying delivery until 34-36 weeks may
be reasonable even after successful laser ablation.

RECOMMENDATIONS

Levels II and III evidence, level B recommendation

1. The diagnosis of TTTS requires 2 criteria: (1) the presence of a MCDA pregnancy; and (2) the
presence of oligohydramnios (defined as a MVP of <2 cm) in one sac, and of polyhydramnios (a MVP
of >8 cm) in the other sac.

Levels II and III evidence, level B recommendation

2. The Quintero staging system appears to be a useful tool for describing the severity of TTTS in a
standardized fashion.

Levels II and III evidence, level B recommendation

3. Serial sonographic evaluations about every 2 weeks, beginning usually around 16 weeks of gestation,
until delivery, should be considered for all twins with MCDA placentation.
Levels II and III evidence, level B recommendation
4. Screening for congenital heart disease is warranted in all monochorionic twins, in particular those
complicated by TTTS.

Levels II and III evidence, level B recommendation

5. Extensive counseling should be provided to patients with pregnancies complicated by TTTS
including natural
history of the disease, as well as management options and their risks and benefits. Over three fourths of
stage I TTTS cases remain stable or regress without invasive interventions. The natural history of
advanced (eg, stage ≥III) TTTS is bleak, with a reported perinatal loss rate of 70-100%, particularly
when it presents <26 weeks. The management options available for TTTS include expectant
management, amnioreduction, intentional septostomy of the intervening membrane, fetoscopic laser
photocoagulation of placental anastomoses, selective reduction, and pregnancy termination.

Levels II and III evidence, level B recommendation

6. Patients with stage I TTTS may often be managed expectantly, as the natural history perinatal survival
rate is about 86%.

Levels I and II evidence, level B recommendation

7. Fetoscopic laser photocoagulation of placental anastomoses is considered by most experts to be the
best available approach for stages II, III, and IV TTTS in continuing pregnancies at <26 weeks, but the
metaanalysis data show no significant survival benefit, and the long-term neurologic outcomes in the
Eurofetus trial were not different than in nonlaser-treated controls. Lasertreated TTTS is still associated
with a 30-50% chance of overall perinatal death and a 5-20% chance of long-term neurologic handicap.

Levels I and II evidence, level B recommendation

8. Steroids for fetal maturation should be considered at 24 to 33 6/7 weeks, particularly in pregnancies
complicated by stage ≥III TTTS, and those undergoing invasive interventions.

Level III evidence, level C recommendation

9. Optimal timing of delivery for TTTS pregnancies depends on several factors, including disease stage
and severity, progression, effect of interventions (if any), and results of antenatal testing. Timing
delivery at around 34-36 weeks may be reasonable in selected cases.

Quality of evidence
The quality of evidence for each included article was evaluated according to the categories outlined
by the US Preventative Services taskforce:
I Properly powered and conducted RCT; well-conducted systematic review or metaanalysis of
homogeneous RCTs.
.........................................................................................................
II-1 Well-designed controlled trial without randomization.
.........................................................................................................
II-2 Well-designed cohort or case-control analytic study.
.........................................................................................................
II-3 Multiple time series with or without the intervention; dramatic results from uncontrolled
experiments.
.........................................................................................................
III Opinions of respected authorities, based on clinical experience; descriptive studies or case
reports; reports of expert committees.
Recommendations are graded in the following categories:
Level A
The recommendation is based on good and consistent scientific evidence.
Level B
 The recommendation is based on limited or inconsistent scientific evidence.
Level C
The recommendation is based on expert opinion or consensus.

This opinion was developed by the Publications Committee of the Society for Maternal-Fetal
Medicine with the assistance of Lynn L. Simpson, BSc, MSc, MD, and was approved by the Executive
Committee of the Society on September 20, 2012. Dr Simpson, and each member of the Publications
Committee (Vincenzo Berghella, MD [Chair], Sean Blackwell, MD [Vice-Chair], Brenna Anderson,
MD, Suneet P. Chauhan, MD, Joshua Copel, MD, Jodi Dashe, MD, Cynthia Gyamfi, MD, Donna
Johnson, MD, Sara Little, MD, Kate Menard, MD, Mary Norton, MD, George Saade, MD, Neil
Silverman, MD, Hyagriv Simhan, MD, Joanne Stone, MD, Alan Tita, MD, PhD, Michael Varner, MD,
Ms Deborah Gardner) have submitted a conflict of interest disclosure delineating personal, professional,
and/or business interests that might be perceived as a real or potential conflict of interest in relation to
this publication.