303jumonline.

qxp:Layout 1

2/17/11

12:57 PM

Page 297

ORIGINAL RESEARCH

Monochorionic Diamniotic Twin

Pregnancy
Timing and Duration of Sonographic Surveillance for
Detection of Twin-Twin Transfusion Syndrome
Alissa Carver, MD, Sina Haeri, MD, MHSA, Julie Moldenhauer, MD, Honor M. Wolfe, MD,
William Goodnight, MD
ObjectiveTwin-twin transfusion syndrome complicates up to 15% of monochorionic
diamniotic gestations. Current recommendations for sonographic surveillance in monochorionic diamniotic pregnancies for detection of twin-twin transfusion syndrome vary.
Our objective was to determine an appropriate frequency of sonographic surveillance
to optimize detection of twin-twin transfusion syndrome in monochorionic diamniotic
gestations.
MethodsA retrospective cohort analysis of all nonanomalous monochorionic diamniotic twins delivered at the University of North Carolina over a 9-year period was
performed. The rate and gestational age of twin-twin transfusion syndrome onset were
calculated. The time to the diagnosis of twin-twin transfusion syndrome was evaluated
by a Kaplan-Meier survival curve; clinical factors at initial sonography were examined
for their use in prediction of twin-twin transfusion syndrome.
ResultsOf the 577 twin deliveries, 145 (25%) were monochorionic diamniotic and
included for analysis. The rate of twin-twin transfusion syndrome was 17.93% (n = 26).
The mean frequency of surveillance SD before diagnosis of twin-twin transfusion syndrome was 3.1 2.1 weeks. The mean gestational age at diagnosis of twin-twin transfusion syndrome was 21.3 3.4 weeks (range, 1529 weeks). Both a discordant
maximum vertical amniotic fluid pocket (>65% difference) and a discordant estimated
fetal weight (>25% difference) at initial sonography showed a significantly shorter time
to diagnosis of twin-twin transfusion syndrome (P < .0001).

Received July 27, 2010, from the Department of

Obstetrics and Gynecology (A.C.) and Division of
Maternal-Fetal Medicine (S.H., H.M.W., W.G.),
University of North Carolina at Chapel Hill School
of Medicine, Chapel Hill, North Carolina USA;
and Center for Fetal Diagnosis and Treatment,
Childrens Hospital of Philadelphia, Philadelphia,
Pennsylvania USA (J.M.). Revision requested
August 24, 2010. Revised manuscript accepted for
publication September 3, 2010.
This study was presented in part at the Society
for Maternal-Fetal Medicine 30th Annual Meeting, February 16, 2010; Chicago, Illinois.
Address correspondence to Sina Haeri, MD,
MHSA, Department of Obstetrics and Gynecology,
Division of Maternal-Fetal Medicine, University of
North Carolina School of Medicine, 3010 Old
Clinic Building, CB 7516, Chapel Hill, NC
27599-7516 USA.
E-mail: sinahaeri@gmail.com

ConclusionsEvaluation for twin-twin transfusion syndrome should begin in the second trimester. Weekly surveillance for those pregnancies with estimated fetal weight or
maximum vertical pocket discordance is recommended. For those with a concordant estimated fetal weight and maximum vertical pocket, sonographic evaluation every 2
weeks is warranted to 28 to 30 weeks. After that, development of twin-twin transfusion
syndrome is less likely, and a different paradigm of antenatal testing may be reasonable.
Key Words: monochorionic diamniotic gestations; sonography; surveillance, twin-twin
transfusion syndrome

win-twin transfusion syndrome affects 10% to 15% of all

monochorionic diamniotic twin gestations.1,2 Its etiology is
related to unbalanced vascular communication between the
two fetal circulatory systems, leading to a progression of biochemical and physiologic responses, resulting in substantial fetal and
neonatal morbidity and mortality.3 The sonographic diagnosis of
twin-twin transfusion syndrome requires discordance of the maximum vertical amniotic fluid pocket and is staged on the basis of the

2011 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2011; 30:297301 | 0278-4297/11/$3.50 | www.aium.org

303jumonline.qxp:Layout 1

2/17/11

12:57 PM

Page 298

Carver et alMonochorionic Diamniotic Twin Pregnancy Surveillance

presence of other findings, including abnormal umbilical

artery Doppler velocimetric findings, absence of the fetal
bladder, fetal hydrops, and fetal death.4 Previous studies
have suggested that early markers of twin-twin transfusion
syndrome may exist as well, including discordance in the
first-trimester crown-rump length and nuchal translucency;
however, these remain of limited sensitivity and specificity.57 Furthermore, the optimum frequency and timing
of sonographic surveillance in pregnancy for the detection
of twin-twin transfusion syndrome have yet to be established in the United States and vary from weekly to every
other week and from 16 weeks throughout the pregnancy.
Nonstandard monitoring creates the potential for unnecessary testing as well as a delay in diagnosis. To that
end, this study attempted to clarify the timing for sonographic monitoring of monochorionic diamniotic pregnancies. The primary objective of this study was to
determine the appropriate gestational age for frequent
sonographic evaluation necessary to identify twin-twin
transfusion syndrome in monochorionic diamniotic twin
pregnancies. The secondary objective was to examine the
association between amniotic fluid and fetal growth discordance at time of initial second-trimester sonography
with subsequent development of twin-twin transfusion
syndrome in an effort to suggest monochorionic diamniotic twin pregnancies that may require an increased testing interval.

Materials and Methods

A retrospective cohort analysis of all twin deliveries at the
University of North Carolina Hospitals between January
2000 and June 2009 was completed. We included all
monochorionic diamniotic twin pregnancies with prenatal
care at University of North Carolina. Patients with no
sonographic examinations in the second trimester and with
no available prenatal records were excluded from the study.
Approval by the University of North Carolina Institutional
Review Board was obtained before collection.
Maternal demographic data were abstracted from the
paper and electronic medical records by 3 investigators
using a standard data collection sheet. Gestational age was
based on the last menstrual period. Per our institutional protocol, in cases of an uncertain last menstrual period or where
there was a greater than 8% discordance between menstrual
and sonographic dating, the gestational age was assigned
on the basis of the earliest sonographic examination.
All patients received serial transabdominal sonographic examinations, performed at least every 2 weeks
(after 15 weeks gestation) or from the gestational age that

298

a monochorionic diamniotic twin pregnancy was confirmed (if presenting after 15 weeks gestation). In cases of
abnormal sonographic findings that raised the suspicion of
future development of twin-twin transfusion syndrome,
the frequency of sonographic surveillance was increased to
weekly or twice weekly, at the discretion of the attending
physician. Data including estimated fetal weight, maximum
vertical pocket, presence of the fetal bladder, and umbilical
artery Doppler findings were abstracted from each sonographic report using the R4 Acert system (version 4.10; R4
Acert, Strongsville, OH). All sonographic examinations
were performed by registered diagnostic medical sonographers and interpreted by board-certified or -eligible
maternal-fetal medicine specialists. Sonographic examinations were performed with HDI 5000 (Philips Healthcare,
Bothell, WA), Voluson Expert (GE Healthcare, Milwaukee, WI), and Voluson E8 (GE Healthcare) ultrasound
machines. Placental pathologic examination was used to
confirm chorionicity after birth in all cases.
Twin-twin transfusion syndrome was diagnosed by
the presence of oligohydramnios (maximum vertical
pocket <2 cm) and polyhydramnios (maximum vertical
pocket >8 cm at 20 weeks or >10 cm at >20 weeks) in
the absence of other causes. Twin-twin transfusion syndrome severity was assessed on the basis of the established
criteria of Quintero et al.4 Because the goal of this study
was to assess testing frequency to identify twin-twin transfusion syndrome, staging of twin-twin transfusion syndrome was not included in the analysis. Pregnancies in
which stage 1 or greater twin-twin transfusion syndrome
developed were grouped as twin-twin transfusion syndrome and served as cases. Pregnancies unaffected by
twin-twin transfusion syndrome were grouped as no twintwin transfusion syndrome and served as the referent
group. In pregnancies without twin-twin transfusion syndrome, for analysis in this study, at the time of the initial
second-trimester sonographic examination, discordant
amniotic fluid was defined as a 65% difference in the maximum vertical pocket [(larger smaller/larger) 100],
and discordant fetal weight was defined as a greater than
25% difference in estimated fetal weight. Our practice includes a substantial proportion of statewide referral patients and those with late intake to prenatal care; therefore,
this data set does not contain a meaningful proportion of
cases with first-trimester data including crown-rump
length.
Data were analyzed using SAS software (version 9.1;
SAS Institute, Inc, Cary, NC). Demographic characteristics between pregnancies complicated by twin-twin
transfusion syndrome and without were compared with

J Ultrasound Med 2011; 30:297301

303jumonline.qxp:Layout 1

2/17/11

12:57 PM

Page 299

Carver et alMonochorionic Diamniotic Twin Pregnancy Surveillance

Kruskal-Wallis and 2/Fisher exact tests as appropriate.

The rate and gestational age of onset of twin-twin transfusion syndrome were calculated and described. KaplanMeier survival curves were created to evaluate the time to
diagnosis of twin-twin transfusion syndrome. Delivered
patients and the diagnosis of twin-twin transfusion syndrome were treated as censored data. A log rank test was
used to compare the time to twin-twin transfusion syndrome based on estimated fetal weight discordance and
amniotic fluid (maximum vertical pocket) discordance at
time of the initial second-trimester sonographic evaluation.

Results
A total of 25,973 deliveries occurred in the study period,
including 577 twin pregnancies (2%). Of the 151 (26%)
confirmed monochorionic diamniotic twin pregnancies,
145 met inclusion criteria and were analyzed. Six cases
were excluded due to unavailability of prenatal records.
The maternal demographic and sonographic data are presented in Table 1. The rate of twin-twin transfusion syndrome was 17.93% (n = 26), with a mean gestational age
at diagnosis SD of 21.3 3.4 weeks (range, 1529
weeks), as confirmed by the Kaplan-Meier survival curve
(Figure 1). Of the cases of twin-twin transfusion syndrome,
50% were diagnosed by an estimated gestational age of
20 weeks 6 days, and 92.3% (24 of 26) were diagnosed
by 26 weeks 6 days. No further cases of twin-twin transfusion syndrome were identified after a gestational age of
28 completed weeks. The mean frequency of surveillance

SD before diagnosis of twin-twin transfusion syndrome

was 3.1 2.1 weeks.
We also examined the association between differences
in the estimated fetal weight and maximum vertical pocket
at the time of initial sonography and development of twintwin transfusion syndrome. Both a discordant estimated
fetal weight (>25% difference; Figure 2) and a discordant
maximum vertical pocket (>65% difference; Figure 3) at
initial sonography showed a significantly shorter time to
diagnosis of twin-twin transfusion syndrome (log rank test,
P < .0001). As expected, the mean gestational age at diagnosis of twin-twin transfusion syndrome was earlier with
estimated fetal weight discordance at initial sonography
without the diagnosis of twin-twin transfusion syndrome,
with a gestational age at diagnosis of 19.3 2.2 weeks with
estimated fetal weight discordance compared to 22.3 3.4
weeks without discordance (P = .05). Similarly, the mean
gestational age at diagnosis of twin-twin transfusion syndrome at initial sonography with amniotic fluid volume
discordance was 19.3 2.6 weeks compared to 22.2 3.3
weeks with concordant amniotic fluid volume concordance (P = .03). The presence of estimated fetal weight
and maximum vertical pocket discordance was demonstrative of a more rapid progression to the diagnosis of
twin-twin transfusion syndrome.

Discussion
Appropriately timed diagnosis of twin-twin transfusion
syndrome is crucial because a delay in diagnosis may re-

Table 1. Maternal and Sonographic Characteristics of the Groups With and Without Twin-Twin Transfusion Syndrome
Characteristic
Age, y
Race, n (%)
African American
Hispanic
White
Gestational age at delivery, wk
Sonographic features at time of first evaluation
Gestational age, wk
Maximum vertical pocket, twin A, cm
Maximum vertical pocket, twin B, cm
Maximum vertical pocket difference , %
Estimated fetal weight, twin A, g
Estimated fetal weight, twin B, g
Estimated fetal weight difference, %
Estimated fetal weight discordance (25%), %
Amniotic fluid volume discordance (65%), %

TTTS
(n = 26)

No TTTS
(n = 119)

28.2 6.6

28.4 6.2

.96

3 (12)
3 (12)
17 (65)
31.0 3.6

19 (16)
42 (36)
52 (44)
33.9 3.3

.57
.02
.05
.0002

18.8 3.4
6.1 4.1
4.0 3.3
59.6 31.5
330.4 193.2
293.8 164.5
25.2 10.5
34.6
30.8

20.3 5.2
4.5 1.4
4.6 1.4
18.3 18.1
488.9 472.2
483.0 462.5
9.3 7.6
4.2
1.68

.48
.02
.06
<.0001
.34
.10
<.0001
<.0001
<.0001

Data are mean SD unless otherwise stated. Kruskal-Wallis and 2 tests were used for group comparisons. TTTS indicates twin-twin transfusion syndrome.

J Ultrasound Med 2011; 30:297301

299

303jumonline.qxp:Layout 1

2/17/11

12:57 PM

Page 300

Carver et alMonochorionic Diamniotic Twin Pregnancy Surveillance

sult in a delay in treatment and increased perinatal morbidity. Most current recommendations for sonographic
surveillance in monochorionic diamniotic pregnancies for
detection of twin-twin transfusion syndrome concur with
the need for early monitoring (initiating sonography at 16
18 weeks estimated gestational age), whereas the timing
and frequency of follow-up are less well standardized in the
United States. Novel to our study was the evaluation of the
appropriate gestational age for sonographic surveillance
needed to identify twin-twin transfusion syndrome in
monochorionic diamniotic twin pregnancies, with atten-

tion to the potential for less frequent examinations at gestational ages with a reduced risk of twin-twin transfusion
syndrome. Our results lend further support to the recommendation that evaluations for twin-twin transfusion syndrome should begin in the second trimester. Evaluations
every 2 weeks are warranted to 28 to 30 weeks, except in
those pregnancies with estimated fetal weight and maximum vertical pocket discordance, in which weekly surveillance should be considered.
In our population, the range of twin-twin transfusion
syndrome diagnosis was 15 to 29 weeks gestation. This is
in concordance with separate reports by Lewi et al7 and
Sueters et al.8 In the former, Sueters et al8 prospectively followed 23 monochorionic diamniotic twins and reported
twin-twin transfusion syndrome development between 16
and 29 weeks gestation. Similar to our findings, they also
noted an association between amniotic fluid discordance
(in the absence of polyhydramnios and oligohydramnios)
and subsequent development of twin-twin transfusion syndrome.8 In the latter study, Lewi et al7 prospectively enrolled 208 monochorionic diamniotic twin pairs, of which
19 developed twin-twin transfusion syndrome between 16
and 28 weeks gestation. Our findings, along with those of
Sueters et al8 and Lewi et al,7 suggest that after 28 to 30
weeks gestation, development of twin-twin transfusion
syndrome is less likely. Therefore, the frequency of sonographic surveillance may be reduced in otherwise uncomplicated monochorionic diamniotic twin pregnancies.

Figure 2. Kaplan-Meier survival curve for the time to diagnosis of twintwin transfusion syndrome by estimated fetal weight (EFW) discordance
at the time of the first sonographic examination.

Figure 3. Kaplan-Meier survival curve for the time to diagnosis of twintwin transfusion syndrome by amniotic fluid volume (AFV) discordance
at the time of the first sonographic examination.

Figure 1. Kaplan-Meier survival curve for the time to diagnosis of twintwin transfusion syndrome (TTTS).

300

J Ultrasound Med 2011; 30:297301

303jumonline.qxp:Layout 1

2/17/11

12:57 PM

Page 301

Carver et alMonochorionic Diamniotic Twin Pregnancy Surveillance

There are certain fetal complications that have been noted

to occur later in pregnancy, such as intrauterine growth restriction and twin anemia/polycythemia. To that end, this
study does not recommend cessation of antenatal monitoring after 30 weeks but suggests that a different paradigm
rather than sonography every 2 weeks for the maximum
vertical pocket may be optimal after the risk of classic twintwin transfusion syndrome is reduced.
Our findings must be interpreted within the context of
the study design. This study was retrospective in design
and relied on existing sonographic data, and not all patients
had a sonographic examination in the first trimester. Ideally
it would have been helpful to evaluate the association between early crown-rump-length or nuchal translucency
and the subsequent development of twin-twin transfusion
syndrome. Furthermore, our twin-twin transfusion syndrome sample size was not large, which limited the extent
of our analysis. Finally, this data set did not include the initial and worst stages as well as the stage and type of intervention. Such an analysis, with the outcome of the time to
intervention or a specific stage, would further clarify the interval necessary for sonographic surveillance. However, the
goal of this study was to determine the optimal weeks during gestation at which frequent sonographic monitoring to
screen for the diagnosis of twin-twin transfusion syndrome
is necessary and not to address the interval to intervention.
Certainly, sonographic surveillance will be more aggressive once the diagnosis of twin-twin transfusion syndrome
is made. Nevertheless, these findings add to the existing
body of evidence and help in the formulation of standard
recommendations for twin-twin transfusion syndrome surveillance in monochorionic diamniotic twin pregnancies.
Despite these limitations, we conclude that evaluations
for twin-twin transfusion syndrome should begin in the second trimester, with biweekly evaluations to 28 to 30 weeks.
In those pregnancies with estimated fetal weight and maximum vertical pocket discordance, weekly surveillance is
warranted. Confirmation with larger sample sizes and incorporation of first-trimester crown-rump length measurements can further clarify an ideal fetal testing schedule.

4.

5.

6.

7.

8.

RA. Placental types and twin-twin transfusion syndrome. Am J Obstet Gynecol 2002; 187:489494.
Quintero RA, Morales WJ, Allen MH, Bornick PW, Johnson PK, Kruger
M. Staging of twin-twin transfusion syndrome. J Perinatol 1999; 19:550
555.
Sebire NJ, Souka A, Skentou H, Geerts L, Nicolaides KH. Early prediction
of severe twin-to-twin transfusion syndrome. Hum Reprod 2000;
15:20082010.
El Kateb A, Nasr B, Nassar M, Bernard JP, Ville Y. First-trimester ultrasound examination and the outcome of monochorionic twin pregnancies. Prenat Diagn 2007; 27:922925.
Lewi L, Gucciardo L, Huber A, et al. Clinical outcome and placental characteristics of monochorionic diamniotic twin pairs with early- and lateonset discordant growth. Am J Obstet Gynecol 2008; 199:511.e1511.e7.
Sueters M, Middeldorp JM, Lopriore E, Oepkes D, Kanhai HH, Vandenbussche FP. Timely diagnosis of twin-to-twin transfusion syndrome
in monochorionic twin pregnancies by biweekly sonography combined
with patient instruction to report onset of symptoms. Ultrasound Obstet
Gynecol 2006; 28:659664.

References
1.

2.

3.

Sebire NJ, Snijders RJ, Hughes K, Sepulveda W, Nicolaides KH. The hidden mortality of monochorionic twin pregnancies. Br J Obstet Gynaecol
1997; 104:12031207.
Lewi L, Jani J, Boes AS, et al. The natural history of monochorionic twins
and the role of prenatal ultrasound scan. Ultrasound Obstet Gynecol 2007;
30(special issue):401402.
Bermudez C, Becerra CH, Bornick PW, Allen MH, Arroyo J, Quintero

J Ultrasound Med 2011; 30:297301

301