- Top of page
- Material and methods
Sonography plays a major role in the management of twin pregnancies. Early diagnosis enables optimization of antenatal care, including serial ultrasound examinations to detect those obstetric complications unique to multiple pregnancies.
Congenital anomalies may be the result of the teratogenic insult that causes the twinning. For any given defect in a twin pregnancy, the pregnancy may be concordant or discordant, although the majority of structural defects are discordant, regardless of zygosity. Discordance in non-identical (dizygotic) twins is usually due to differences in genetic predisposition, whereas in identical (monozygotic twins) twins, it may be a consequence of the underlying stimulus to zygote splitting, variation in gene expression, or abnormal placentation1.
Chorionicity denotes the type of placentation. Monochorionic twins are always monozygotic, while dichorionic twins can be monozygotic as well as dizygotic2. Chorionicity can be determined before 14 + 6 weeks with the lambda sign3. Vascular complications in monochorionic twins are well known and may cause different disorders4. The most important is twin–twin transfusion syndrome (TTTS), detected in 4–35%5–7 of twins with monochorionic placentation. Other complications include twin reversed arterial perfusion, congenital heart disease due to vascular instability, fetal demise, and long-term neonatal and pediatric morbidity secondary to vascular insults in fetal brain, heart and kidneys. There are now two main treatment options for TTTS: serial amnion drainage, a symptomatic treatment to reduce the risk of preterm labor, and fetoscopic laser therapy combined with amnion drainage. Both therapies can improve the outcome and so it is important to recognize TTTS as early as possible8.
Structural anomalies have been reported to occur more often in monozygotic twins compared with dizygotic twins and singletons4, 9–13, with relative risks of congenital anomalies in twins compared with singletons of 1.17 (95% CI, 1.04–1.3) for dizygotic twins11 and 1.25 (95% CI, 1.21–1.28) for monozygotic twins14. The risk of malformations in monozygotic twins is higher than that in dizygotic twins, with a relative risk of congenital anomalies in monozygotic compared with dizygotic twins of 1.4–2.710, 11, 15.
Detailed sonographic evaluation may diagnose congenital abnormalities and the particular complications associated with twin pregnancy. Detection of fetal abnormalities is dependent on several factors, such as time of screening, skill and experience of the sonographers, and technical equipment; the resolution of ultrasound equipment has changed dramatically over the last decade, which could influence the detection rate of abnormalities. The available data, concerning differences in risk of abnormalities in monozygotic compared with dizygotic twins, are based on only a few retrospective studies.
The aim of this study, therefore, was to evaluate screening with serial ultrasound examinations from weeks 12 to 23 in twins stratified according to chorionicity. The antenatal detection rate of structural as well as chromosomal abnormalities and the outcome of screening for TTTS among monochorionic twins were estimated. Secondly, we assessed whether the frequencies of abnormalities differed between monozygotic and dizygotic twins, and between twins conceived naturally and those conceived by assisted reproduction.
Material and methods
- Top of page
- Material and methods
From November 1999 to May 2003, women with twin pregnancies were enrolled into a prospective multicenter observational study at five university fetal medicine referral centers (four in Denmark and one in Sweden). The women were invited to join the study when a twin pregnancy was diagnosed before 14 + 6 weeks of gestation, estimated from the crown–rump length or biparietal diameter of the largest twin. Exclusion criteria were maternal age below 18 years, and a lack of fluency in Danish or Swedish. The local scientific ethics committee and the Data Protection Agency approved the study protocol. Written informed consent was obtained from all participants.
During the study period 524 pregnant women were included in the study. Twenty-nine women were excluded for different reasons (wish to drop out or failure to attend the booked appointment). Among the remaining 495 pregnancies, 46% were conceived naturally, and 54% were assisted by either in-vitro fertilization, intracytoplasmic sperm injection, egg donation or intrauterine insemination with or without ovarian stimulation.
At the time of inclusion a transabdominal or transvaginal scan was performed to determine chorionicity. The pregnancy was classified as dichorionic if two separate placentae or the intertwin membrane (twin peak or lambda sign) were seen. Monochorionic pregnancies were diagnosed on the basis of a single placenta and no lambda sign3.
A nuchal translucency thickness (NT) scan was performed if the gestational age was below 13 + 6 weeks at the time of inclusion and the woman had not undergone chorionic villus sampling according to the regulations of the Danish National Board of Health (1994). Risk assessment was performed according to the guidelines of The Fetal Medicine Foundation (FMF)16, 17. If the estimated risk of trisomy 21 based on maternal age and NT measurement exceeded 1 : 250, the woman was offered an invasive test. When the NT measurement was above the 99th percentile, further investigation was carried out according to the departments' guidelines, which in most cases included early fetal echocardiography in week 14.
All pregnancies had an anomaly scan including biometry in week 19, fetal echocardiography in week 21 performed by specialists in fetal echocardiography, and a cervical assessment scan in week 23 including fetal biometry. Fetuses were classified into groups according to their structural abnormalities, following Wald et al.18 (Table 1); Table 2 gives the classification for the most common fetal malformations.
Table 1. Classification of fetuses into groups according to their structural abnormalities, following Wald et al.18
|A||Abnormalities associated with serious disability for which termination of pregnancy is ‘justifiable’|
|B||Abnormalities for which termination of pregnancy avoids continuing with an unproductive pregnancy|
|C||Abnormalities for which in-utero treatment reduces morbidity|
|D||Abnormalities for which immediate postnatal treatment reduces morbidity|
|P||Those for which there is a possible benefit in antenatal identification but no clear evidence that this is so|
|PM||Those in which there is an indirect marker for another disorder (e.g. a trisomy)|
|O||Those for which there is no benefit in antenatal identification|
Table 2. The most common fetal malformations classified according to Wald et al.18
|Central nervous system|| |
| Spina bifida||A,B (1‰)|
| Anencephaly||B (0.5‰)|
| Holoprosencephaly||A,B (0.1‰)|
| Dandy–Walker syndrome||A,B (≤ 0.1‰)|
| Hydracephaly without spina bifida||PM (1‰)|
| Agenesis of corpus callosum||PM (0.1‰)|
| Congenital diaphragmatic hernia||PM (0.5‰)|
| Congenital cystic adenomatoid malformation||PM (0.1‰)|
| Cystic hygroma with NT ≥ 6 mm||PM (0.1‰)|
| Cleft lip/palate||PM (1.5‰)|
| Gastroschisis||P (0.1‰)|
| Omphalocele||PM (0.5‰)|
| Bilateral renal agenesis||B (0.1‰)|
| Hydronephrosis/multicystic dysplastic kidneys||PM (1.0‰)|
|Heart defect: severe||A,B,D (2‰)|
| Hypoplastic left/right heart syndrome|
| Ebstein's anomaly|
| Single atrium|
| Truncus arteriosus|
| Total anomalous pulmonary drainage|
| Ectopia cordis|
| Complex heart disease|
| Atrioventricular defects|
| Transposition of the great vessels|
|Heart defect: moderate||PM (2‰)|
| Tetralogy of Fallot|
| Coarctation of the aorta|
| Pulmonary stenosis|
| Double outlet ventricle|
|Heart defect: mild||PM (3.5‰)|
| Ventricular septal defect|
| Atrial septal defect|
| Aortic stenosis|
| Osteogenesis imperfecta||B (≤ 0.1‰)|
| Thanatophoric dysplasia||B (≤ 0.1‰)|
| Limb reduction||PM (0.5‰)|
| Meckel–Gruber syndrome||B (≤ 0.1‰)|
The monochorionic pregnancies were seen every second week from 12 to 23 weeks of gestation in order to rule out early TTTS, which was classified according to the Quintero stages19 (Stage 1: abnormal amniotic fluid level with deepest vertical pool of at least 8.0 cm in the amniotic cavity of the recipient and not more than 2 cm in the amniotic cavity of the donor twin; Stage 2: collapsed bladder in donor twin; Stage 3: abnormal Doppler flow in either twin; Stage 4: hydrops in either twin).
To assess zygosity, we isolated DNA from all twin pairs, extracted from cord blood by the salting-out method20. If the procedure failed, a filter blood spot was obtained after birth and a blood spot (θ = 3 mm) was punched out. This was soaked in 20 µL 0.2 M sodium hydroxide. The tubes were incubated at 75 °C for 5 min, neutralized by the addition of 10 µL 40 mM Tris, pH 7.5, and a sample of 5 µL was used for polymerase chain reaction (PCR) amplification of eight highly polymorphic microsatellites: TNFa, D21S11, D21S1412, D18S535, D13S258, D18S386, D21S1411 and D13S631. Primer sequences can be obtained from the GDB Human GenomeDatabase (http://gdbwww.gdb.org/). PCR was performed using a GenAmp PCR system 2600 (Applied BioSystems, Foster City, CA, USA), with standard conditions for all reactions, and 25 cycles of: 45 s at 94 °C, 45 s at 56 °C and 45 s at 74 °C. PCR products were analyzed on an ABI PRISM® 310 Genetic Analyzer using GENESCAN software (ABI, PE BioSystems, Foster City) and a GeneScan™ Size standard (PE BioSystems) as a size marker. All the markers had a heterozygosity level of greater than 0.80. If twins shared alleles for all eight markers they were classified as being monozygotic.
Information about fetal outcome was retrieved from the obstetric files. Participants were also contacted by mail or telephone 8 months or more after delivery. When it was not possible to reach the family, the infants' personal registration numbers were checked for admittance to any Danish or Swedish hospital in the National Hospital Registry; in Denmark and Sweden all hospital admissions are reported. In case of hospitalization, discharge reports were sought.
All frequencies were compared by χ2 test and Student's t-test. The relative risk, the ratio of the proportions of cases having an abnormal outcome in the two groups, was determined.
- Top of page
- Material and methods
The ultrasound scan at inclusion assessed 28% (64) of the twins conceived naturally and 4% (10) of the twins conceived by assisted reproduction as monochorionic diamniotic, and 1% of all the twins to be monoamniotic. In four pregnancies classified by sonography as monochorionic, the zygosity test showed dizygotic twins. These cases were reclassified as dichorionic.
Table 3 shows the distribution of monozygotic and dizygotic twins among naturally conceived twins and among twins conceived by assisted reproduction. Of the 945 infants delivered, 741 were dizygotic (78%), 190 were monozygotic, and zygosity was unknown in 14 (1.5%); 523 (55%) infants were conceived by assisted reproduction.
Table 3. Distribution of chorionicity and zygosity among twin pregnancies conceived naturally or by assisted reproduction
|Conception||Dizygotic (n) (n = 393)||Monozygotic (n)(n = 102)||Total|
Sixty eight percent of the 495 pregnant women had an NT scan. A chromosomal abnormality was diagnosed in six fetuses (0.6%), all from dichorionic pregnancies; five of these cases were diagnosed antenatally. They all had a risk above the 95th centile based on maternal age and NT measurement. There were four cases of trisomy 21, all of which underwent selective termination. One fetus had a 47,XXX karyotype and the parents decided to continue the pregnancy. The discordance in NT was between 2.0 and 7.0 mm in the cases with trisomy 21, and it was only 0.3 mm for the case with triple X. One case of trisomy 21 was diagnosed postnatally in a twin from a dichorionic pregnancy in which no NT scan had been performed.
Twenty-five cases (2.6%) with malformations were diagnosed in the 945 fetuses/infants, 24 with structural malformations, and one with sequelae after intrauterine crowding without other malformations. Seven (29%) of the cases with structural malformations were diagnosed antenatally, and 17 (71%) were diagnosed after birth. No malformation was detected beyond 9 weeks after birth. Seven cases (33%) were classified as severe, belonging to categories A–D (Tables 1 and 2), and six of these seven cases were diagnosed antenatally (Table 4). The incidence of structural malfomations was 3.2% among monozygotic and 2.2% among dizygotic twins; this difference was not significant (P = 0.59), nor was the difference between twins conceived naturally and those conceived by assisted reproduction (P = 0.68) (Table 5). Cardiac abnormalities accounted for 68% (17/25) of all abnormalities. The incidence of all cardiac malformations was 1.8%, with five (0.5%) major, four (0.4%) moderate and eight (0.8%) minor abnormalities. There was no significant difference between the incidence of cardiac malformations in monozygotic and that in dizygotic twins. Table 4 shows data on detection and outcome for each pregnancy with fetal structural or chromosomal abnormalities.
Table 4. Data and outcome of pregnancies in twins with chromosomal or structural abnormalities, including abnormalities due to vascular anastomosis
|Type of abnormality/Case||Chorionicity||Zygosity||Means of conception||Maternal age (years)||Nuchal scan results (mm)||Abnormality||Time of prenatal diagnosis||Outcome|
|Twin 1||Twin 2|
|Central nervous system|
| Case 1§||DC||DZ||IUI||29.0||> 83||NM||> 83||NM||Anencephaly¶||Nuchal scan||NND|
| Case 2||DC||DZ||Natural||28.0||80||1.3||69||1.3||Cerebellar atrophia||ND||Survived|
| Case 3||DC||MZ||Natural||24.6||60||1.7||60||1.6||Cleft lip/palate||ND||Survived|
| Case 4||DC||Unalike sex||IUI||32.3||45||0.6||45||1.9||Cleft lip/palate||ND||Survived|
| Case 5*||DC||DZ||ICSI||34.4||46||1.1||61||2.1||Obstructive uropathy||ND||NND|
| Case 6||DC||Unknown||Natural||29.9|| ||NM||NM|| ||Bilateral renal agenesis¶||Nuchal scan||Induced abortion|
| Case 7§||DC||Unknown||IVF||31.5||> 83||NM||> 83||NM||One kidney||ND||Survived|
| Case 8†§||DC||Unknown||IVF||41.5||> 83||< 1.0||> 83||7.9||HLHS + DS¶||Nuchal scan||Selective termination|
| Case 9||DC||Unalike sex||Egg donation||43.3||74||2.1||75||1.2||TGA¶||Anomaly scan||Survived|
| Case 10||DC||DZ||IVF||35.7||78||1.8||80||1.6||HLHS¶||Repeat anomaly scan||Survived|
| Case 11‡§||DC||Unalike sex||Natural||23.3||> 83||< 1.0||> 83||2.3||Atrioventricular defects + DS¶||ND||Survived|
| Case 12||DC||MZ||Natural||33.4||62||1.4||62||1.2||HLHS¶||Anomaly scan||NND|
| Case 13||MC DA||MZ||Egg donation||41.8||62||1.3||63||1.1||Coarctation of aorta||ND||Survived|
| Case 14||MC DA||MZ||Natural||28.3||75||1.6||68||1.5||Coarctation of aorta||ND||Survived|
| Case 15||DC||Unalike sex||Natural||29.2||55||8.9||58||1.6||DORV||ND||Survived|
| Case 16§||DC||DZ||ICSI||27.3||83||NM||76||NM||Coarctation of aorta||Anomaly scan||Survived|
| Case 17||DC||DZ||ICSI||33.7||66||1.7||61||2.2||Atrial septal defect||ND||Survived|
| Case 18§||DC||DZ||IVF||35.6||> 83||1.9||> 83||1.3||Ventricular septal defect||ND||Survived|
| Case 19||DC||DZ||ICSI||37.7||71||1.6||83||NM||Ventricular septal defect||ND||Survived|
| Case 20||DC||MZ||Natural||31.8||61||NM||61||NM||Aorta stenosis||ND||Survived|
| Case 21||DC||Unalike sex||IVF||31.9||55||NM||53||NM||Ventricular septal defect||ND||Survived|
| Case 5*||DC||DZ||ICSI||34.4||46||1.1||61||2.1||Atrial septal defect||ND||NND|
| Case 22||DC||Unalike sex||Natural||27.5||61||1.5||62||0.8||Atrial septal defect||ND||Survived|
| Case 23||MC DA||MZ||Natural||34.6||70||< 1.0||73||< 1.0||Ventricular septal defect||ND||Survived|
| Case 24||DC||DZ||IUI||28.5||53||1.3||58||1.4||Collapse of lumbar spine||ND||Survived|
|Sequel: intrauterine crowding|| |
| Case 5*||DC||DZ||ICSI||34.4||46||1.1||61||2.1||Club foot||ND||NND|
| Case 25||MC DA||MZ||Natural||30.9||61||< 1.0||62||1.6||Club foot||ND||Survived|
|Sequel: vascular anastomoses|| |
| Case 26||MC DA||Unknown||Natural||36.1||53||< 1.0||50||1.4||Aplasia cutis||ND||Spontaneous reduction|
| Case 27||DC||Unknown||IUI||38.7||68||< 1.0||61||2.1||Trisomy 21||Nuchal scan||Selective termination|
| Case 28||DC||Unknown||ICSI||42.2||74||2.1||75||1.8||47,XXX||Nuchal scan||Survived|
| Case 8†§||DC||Unknown||IVF||41.5||> 83||< 1.0||> 83||7.9||Trisomy 21||Nuchal scan + early fetal echo||Selective termination|
| Case 29||DC||Unknown||Natural||35.8||57||3.7||57||< 1.0||Trisomy 21||Nuchal scan||Selective termination|
| Case 11‡§||DC||Unalike sex||Natural||23.3||> 83||< 1.0||> 83||2.3||Trisomy 21||ND||Survived|
| Case 30||DC||Unknown||IVF||40.6||55||2.5||56||< 1.0||Trisomy 21||Nuchal scan||Selective termination|
Table 5. Incidence of fetal structural and chromosomal abnormalities stratified by means of conception and zygosity
| ||Means of conception||Zygosity||Total|
|Pregnancies (n)||266||229||393||102|| ||495|
|Fetuses initially (n)||532||458||786||204|| ||990|
|Fetal losses before birth (n)||9||36|| ||45||45|
|Structural abnormalities (n (%))|| |
| Central nervous system||1 (0.2)||1 (0.2)||2 (0.3)|| ||2 (0.2)|
| Digestive tract||1 (0.2)||1 (0.2)||1 (0.1)||1 (0.5)|| ||2 (0.2)|
| Genitourinary tract||2 (0.4)||1 (0.2)||1 (0.1)|| ||2||3 (0.3)|
| Cardiac||10 (1.9)||7 (1.7)||11 (1.5)||5 (2.6)||1||17 (1.8)|
| Major||3 (0.6)*||2 (0.5)||3 (0.4)||1 (0.5)||1||5 (0.5)|
| Moderate||2 (0.4)||2 (0.5)||2 (0.3)||2 (1.1)|| ||4 (0.4)|
| Minor||5 (1.0)||3 (0.7)||6 (0.8)||2 (1.1)|| ||8 (0.8)|
| Skeletal||0||1 (0.2)||1 (0.1)|| ||1 (0.1)|
| Total||14 (2.7)||11 (2.6)||16 (2.2)||6 (3.2)||3||25 (2.6)|
| P|| |
| ||P = 0.68||P = 0.59|| |
|Chromosomal abnormalities||4 (0.8)*||2 (0.5)||1 (0.1)|| ||5||6 (0.6)|
Among the major cardiac malformations, all but one were detected during pregnancy: three cases of hypoplastic left heart syndrome (HLHS) and one case with transposition of the great arteries. One fetus had an NT value above the 95th percentile, but a normal karyotype. One of the cases with HLHS had trisomy 21, and the heart defect was diagnosed during early fetal echocardiography. The other two major cardiac malformations were diagnosed at the routine anomaly scan, as was the case with transposition of the great arteries. The case undetected prenatally had an atrioventricular septal defect and trisomy 21 (the case mentioned above, in which an NT scan was not performed). Among the moderate cardiac malformations, one case with coarctation of the aorta was detected prenatally due to an abnormal four-chamber view at the anomaly scan. Two other cases with coarctation and one with double outlet ventricle were diagnosed in the neonatal period. None of the minor cardiac abnormalities was detected by ultrasound (Table 4). The combined performance of the NT scan, early fetal echocardiography and the 19-week anomaly scan to detect cardiac abnormalities could be estimated. The fetal echocardiography performed by specialists in week 21 confirmed the diagnoses but did not identify any additional cardiac malformations. The prenatal detection rate for major cardiac abnormalities was 80%, while that for cardiac defects overall was only 29%.
There were two cases of central nervous system malformations, both in dichorionic dizygotic twins. One fetus had anencephaly; the cotwin was delivered at 36 weeks and was doing well at the time of writing. The other case had cerebellar atrophia. This was not detected during pregnancy, but from week 16 the fetus had signs of early growth restriction and there was a discrepancy between the biometry of the twins. There were two cases with cleft-lip and palate, which were not detected antenatally; one was monozygotic, and the other was dizygotic. Three fetuses had malformations in the genitourinary system. One had bilateral renal agenesis diagnosed at the NT scan, and the other two cases were not diagnosed prenatally; these belonged to Group O (Table 1). One dizygotic dichorionic twin had a collapse of the lumbar spine, which was not observed prenatally. There were two cases of club foot, possibly due to intrauterine crowding; one was monozygotic, and the other was dizygotic. The prenatal detection rate for all abnormalities was 86% for cases belonging to Groups A–D (Table 1), but only 36% when all cases were considered.
Complications due to vascular anastomoses in monochorionic twins were seen in one pregnancy in which one twin apparently vanished between 13 and 16 weeks' gestation. The pregnancy continued and the cotwin was delivered at term with widespread skin aplasia. This child was doing well at the time of writing, apart from the serious consequences of the skin disease.
Twin–twin transfusion syndrome
Signs of TTTS were found in 17 cases (23%), 15 of which were diamniotic and two of which were monoamniotic pregnancies. None of the NT measurements in these fetuses was above the 95th centile (Figure 1). Among the twins with signs of TTTS, 43% had a discordance in the NT measurement of more than 0.5 mm compared with 45% among twins without signs of TTTS. Figure 2 shows the discordance in NT measurement for monochorionic twins with and without signs of TTTS; there was no difference in discordance rate between these two groups. Six (Cases 3, 4, 5, 6, 8 and 9, Table 6) of the diamniotic pregnancies were lost spontaneously between the NT scan and week 24, and all had signs of severe TTTS at the time of fetal demise. In one case (Case 7) the parents chose a termination of pregnancy, although laser coagulation was offered.
Figure 1. Nuchal translucency thickness (NT) measurements and crown–rump length (CRL) in a series of 671 twins having a nuchal scan, showing individuals with chromosomal anomalies (▪), those with both chromosomal and major cardiac anomalies () and those with major (□), moderate (×) and minor (▵) cardiac abnormalities, as well as cases with twin–twin transfusion syndrome (TTTS) that miscarried (○) and that were delivered (●). The line indicates the 95th centile according to The Fetal Medicine Foundation.
Download figure to PowerPoint
Figure 2. The discordance in nuchal translucency thickness (NT) measurements and crown–rump length (CRL) of the largest twin in monochorionic twins with (▵; n = 14) and without (▪; n = 33) twin–twin transfusion syndrome.
Download figure to PowerPoint
Table 6. Data concerning twins with twin–twin transfusion syndrome (TTTS)
|Case||Amnion.||Nuchal scan results (mm)||TTTS||Birth-weight data||Obstetric and fetal outcome|
|Twin 1||Twin 2||Actual weight (g)||Weight deviation (%)*|
|CRL||NT||CRL||NT||Twin 1||Twin 2||Twin 1||Twin 2||GA at miscarriage/ delivery (weeks)||Outcome|
|1||MA||78||1.7||77||1.3||Stage 1|| ||14 + 1||Miscarriage|
|2||MA||48||3.9||46||1.3||Stage 3|| ||12 + 3||TOP; cord entanglement|
|3||DA||67||1.3||68||1.6||Stage 2|| ||17 + 1||Miscarriage|
|4||DA||31||NM||31||NM||Stage 3|| ||18 + 4||Miscarriage|
|5||DA||56||2||59||1.2||Stage 1|| ||15 + 3||Miscarriage|
|6||DA||66||1.8||45||0.5||Stage 4 at week 13|| ||14 + 5||Miscarriage|
|7||DA||48||1.2||49||0||Stage 3|| ||17 + 4||TOP; laser surgery offered|
|8||DA||70||1.7||72||1.1||Stage 2|| ||22 + 6||Miscarriage|
|9||DA||71||< 1.0||72||< 1.0||Stage 2|| ||18 + 5||Miscarriage|
|10||DA||77||2.3||78||1.1||Stage 1 at week 17||1530||1280||6.1||− 11.2||29 + 1||Minor neurological disorders|
|11||DA||65||1.3||65||1.3||Stage 1 at week 23||1270||995||− 9.5||− 29.1||29 + 1||Minor neurological disorders|
|12||DA||79||1.7||83||0.9||Stage 1 at week 17||2550|| ||− 22.4|| ||38 + 3||Spontaneous reduction after laser surgery; survivor doing well|
|13||DA||71||1.9||73||2.3||No signs at week 23; Stage 1 at week 25||1760||1740||− 7.4||− 8.4||31 + 4||Both doing well|
|14||DA||> 83||NM||> 83||NM||No signs at week 23; Stage 3 at week 25|| ||1027|| ||− 16.9||27 + 6||Spontaneous reduction after laser surgery|
|15||DA||53||< 1.0||50||1.4|| ||3710|| ||− 0.3|| ||40 + 4||Spontaneous reduction between 12 and 17 weeks; skin aplasia in the survivor|
|16||DA||58||1.3||58||1.5||Stage 1 at week 23||2325||2105||− 8.2||− 16.9||34 + 5||Both doing well|
|17||DA||> 83||NM||> 83||NM||No signs at 23 weeks||1815|| ||− 25.7|| ||34 + 2||Spontaneous reduction between 34 and 35 weeks; acute TTTS; the liveborn died after 2 months|
A spontaneous reduction (Case 15) occurred in one pregnancy between 12 and 17 weeks. The cotwin was delivered at term but had severe skin aplasia. Incipient signs (Stage 1) of TTTS were seen in two cases (Cases 10 and 11), in weeks 17 and 23, respectively. They were both delivered at 29 weeks and developed minor neurological sequelae. Two other cases (Cases 13 and 14) had discordance in amniotic fluid volume, but were not classified as TTTS Stage 1 according to the definition. They both developed TTTS in week 25 and one of them (Case 14) had endoscopic laser surgery performed. In one case (Case 17), acute TTTS was observed at 34 weeks. One twin died before delivery, and the other died 2 months after birth. Endoscopic laser surgery was performed in two cases (Cases 12 and 14). Both had a spontaneous reduction the following day.
Among the 74 monochorionic twin pregnancies, four were monoamniotic. In one case (Case 2, Table 6), cord entanglement and severe hydrops were observed at the NT scan. The woman had an induced abortion. Another monoamniotic pregnancy (Case 1) miscarried with incipient signs of TTTS, while the last two women delivered liveborn twins. All four children were doing well at the time of writing.
- Top of page
- Material and methods
This study demonstrates that an ultrasound monitoring program including an NT scan and assessment of chorionicity before 14 + 6 weeks, and a 19-week anomaly scan including a four-chamber view, detected six of seven structural anomalies justifying a termination of pregnancy (Table 1) and four of five autosomal aneuploidies (Table 4). We found a 2.6% incidence of structural malformations, which is twice the prevalence of 1.3% in newborns as estimated in a meta-analysis of 19 studies comprising 180 000 pregnancies18.
Cardiac malformations were the most common malformations, as has been shown previously9, 13, 21. The incidence of major cardiac abnormalities in our study was 0.5%, twice the birth prevalence of 0.2% found in a meta-analysis based on 14 studies18 and 1.4-fold higher than that found in a cohort study of 7339 unselected singleton pregnant women22. Our result was in accordance with the result from the national twin survey of births in England and Wales9. The relative risk for cardiac malformations among 76 000 twins compared with singletons was found to be 1.6 (95% CI, 1.36–1.84). It was not the aim of our study to evaluate the effectiveness of the NT scan, especially not as a screening test for cardiac abnormalities. The detection rate for major cardiac abnormalities using an NT cut-off of ≥ 2.5 mm was, however, only 20% (1 in 5). This result is in accordance with the results of Mavrides et al.22, but is much lower than those of a previous study23. Our prenatal ultrasound detection rate of major cardiac defects was 80% (4/5), but overall the detection rate for cardiac defects was only about 30%. Others have found a detection rate for major cardiac anomalies of 46 (range, 0–100)%, and for moderate and mild cardiac anomalies, 20 (range, 0–78)% and 16 (range, 0–100)%, respectively18. Interestingly, the cardiac abnormalities found prenatally were diagnosed either at the NT scan, early fetal echocardiography or at the 19-week anomaly scan, whereas fetal echocardiographic screening at 21 weeks by specialists did not add any further cases. While the small number of cases in this study does not allow any conclusion to be drawn concerning the need for routine fetal echocardiography in twins at 21–22 weeks, the relatively low incidence of cardiac defects in the twin population does not seem to justify routine 21-week fetal echocardiography.
We did not find a significant difference in rates of abnormalities between monozygotic and dizygotic pregnancies. The relative risk was 1.5 (95% CI, 0.6–3.7) between monozygotic and dizygotic twins. This result is comparable to a study in Aberdeen including 657 women with twin pregnancies in which the relative risk was 1.4 (95% CI, 0.8–2.5)15 (Table 7). The relative risk in 508 women with twin pregnancies between monozygotic and dizygotic twins was 1.4 (95% CI, 0.9–2.0) in the National Collaborative Perinatal Project from 12 institutions in the USA. The twins were followed to the age of 7 years11. In a study in Taipei10, the relative risk of congenital abnormalities was 2.7 (95% CI, 1.1–7.0) between 482 monozygotic twins and 252 dizygotic twins, but the distribution between monozygotic and dizygotic twins was very different in our study compared with all the other studies.
Table 7. Rates of congenital abnormalities in monozygotic (MZ) and dizygotic (DZ) twin pregnancies
|Reference||Description||Number||Incidence of major malformations||RR||95% CI|
|Chen et al.10||Four hospitals 1985–1989 in Taipei.|| ||MZ: 26/964 (2.7%)|| |
| Zygosity with blood cell antigens.||482 MZ||252 DZ||DZ: 5/504 (1%)||2.7||1.1–7.0|
|Myrianthopoulos11||NCPP cohort study from 12 institutions in USA. Twins followed from the beginning|| || || |
| of pregnancy until the infants were 7 years old.|| ||MZ: 40/373 (11%)|| |
| Zygosity with blood cell antigens.||373 MZ||617 DZ||DZ: 48/617 (7.8%)||1.4||0.9–2.0|
|Corney et al.15||Retrospective survey from Aberdeen.|| ||MZ: 20/380 (5.3%)|| |
| Zygosity with blood cell antigens.||380 MZ||712 DZ||DZ: 26/712 (3.7%)||1.4||0.8–2.5|
|This study||Prospective study from Scandinavia. Zygosity with|| ||MZ: 6/190 (3.2%)|| |
| DNA.||190 MZ||741 DZ||DZ: 16/741 (2.2%)||2||0.6–3.7|
In the literature, the detection rate of anomalies varies as a result of differences in postnatal ascertainment, definition of anomalies and operator capability4. It is therefore preferable to have specific prevalences and/or relative risks in order to compare data between studies. Our dataset, as well as those of the other studies shown in Table 7, may have been too small to detect the relatively modest differences in rates of abnormalities between monozygotic and dizygotic pregnancies. In order to have a relative risk of around 2, using an expected ratio between monochorionic and dichorionic twins of 0.25, a level of significance of 0.05 and a power of 0.90, the calculated sample size should have been 479 monozygotic and 1867 dizygotic twins. The difference in prevalence in our population was apparently even smaller, which would have further increased the magnitude of the required sample size.
An explanation for this could be that previous studies differentiated only between monozygotic and dizygotic twins and not between monochorionic and dichorionic pregnancies. Thus, the difference in prevalence of structural malformations between monozygotic and dizygotic twins in our study was small compared with the results of others because we classified sequelae from TTTS (those vascular disruptive sequences culminating in TTTS, including infarction of the intestine or skin resulting in widespread skin aplasia or intestinal atresia, or infarction of the brain, kidneys, liver and lungs) separately and not as structural malformations. Previously, there has been less focus on the difference between malformations and sequelae24.
We found signs of TTTS in 23% of the monochorionic pregnancies. In half of these cases fetal demise of both twins occurred before 24 weeks. In spite of an intensive monitoring program, the outcome for these affected pregnancies was very poor, with two terminations of pregnancy, seven miscarriages, two spontaneous reductions and two selective terminations. We did not find an increased NT measurement in the twins developing TTTS in contrast to the findings of Sebire et al.7, 25. Our results were independent of whether we used the 95th centiles from The FMF or delta NT calculated from the median values in our population. We could not find any correlation between NT discordance and TTTS. This result is in accordance with a recently published cohort study of 50 monochorionic twins followed prospectively with NT measurement and ductus venosus flow evaluation26. Recently, Senat et al.8 in a randomized, controlled trial compared the safety and efficacy of laser surgery and amnioreduction in TTTS diagnosed between 15 and 26 weeks8; endoscopic laser treatment resulted in a higher likelihood ratio of survival of at least one twin, and a smaller risk of neurological complications in the infant at 6 months of age. Whereas previous data19 had shown a benefit of laser treatment only in fetuses with TTTS Stages 3 and 4, Senat et al.8 also found a better outcome for fetuses with TTTS Stages 1 and 2. Moreover, they showed that irrespective of the treatment modality, the prognosis was better when the treatment was started in the early stages of the TTTS8. Our data suggest that, if we had managed early-stage cases more proactively, we might have improved the outcome for some of the fetuses affected by TTTS. When signs of TTTS are diagnosed in a twin pregnancy, the parents have the option of continuing the entire pregnancy, terminating the entire pregnancy or choosing endoscopic laser treatment.
In conclusion, we found that a sonographic evaluation before 13 + 6 weeks is very important in twins. As in singleton pregnancies, this examination can detect those twin fetuses at increased risk of chromosomal abnormality, but furthermore it can detect the high-risk group of monochorionic twins that needs intensive counseling and monitoring. A supplementary anomaly scan for all types of twins in week 19–20 can detect more than 80% of major structural abnormalities. The incidence of TTTS was 23% in monochorionic pregnancies from 13 weeks and the outcome of these pregnancies was very poor. In the future, we should focus on twins with monochorionic placentation, with intensive ultrasonic evaluation from weeks 14–23 in order to detect and treat early the development of TTTS and diminish fetal loss and infant mortality rates among twins.