In singleton pregnancies at 11–13 weeks of gestation, about 4% of fetuses demonstrate reversed a-wave in the flow velocity waveform from the ductus venosus. This abnormal flow pattern is associated with increased risk for chromosomal abnormalities, cardiac defects, and fetal death.1 In a prospective study in singleton pregnancies, reversed a-wave was observed in 3.7% of 10,120 with normal outcome, 10.8% of the 185 resulting in miscarriage or fetal death, 62.1% of the 95 with abnormal karyotype, 25% of the 20 with major cardiac defect, and 4.3% of the 70 with noncardiac defect.1
Studies of ductus venosus flow in twin pregnancies have focused on the association of abnormal flow as an early ultrasound marker of severe twin–twin transfusion syndrome. In a series of 11 monochorionic twin pregnancies, severe twin–twin transfusion syndrome developed in the two cases with abnormal ductus venosus waveforms in the recipient fetus at 11–13 weeks of gestation and in none of the cases with normal ductus venosus flow patterns.2
The aim of this observational study of 179 monochorionic and 516 dichorionic twin pregnancies undergoing ultrasound examination at 11–13 weeks of gestation was to examine whether the value of ductus venosus flow in the prediction of adverse pregnancy in twins is similar to that in singleton pregnancies and the interaction between flow and chorionicity in the prediction of outcome.
MATERIALS AND METHODS
This was a prospective study in diamniotic twin pregnancies examined at 11 0/7 to 13 6/7 weeks of gestation as part of our policy of screening for chromosomal abnormalities by a combination of maternal age and fetal nuchal translucency thickness.3 The inclusion criteria were diamniotic twin pregnancies with two live fetuses at 11–13 weeks. Transabdominal ultrasound examination was performed to diagnose any major fetal defects and for measurement of the crown-rump length (CRL) and nuchal translucency thickness of each twin. Ductus venosus blood flow velocity waveforms also were obtained routinely by sonographers who had received the appropriate Fetal Medicine Foundation Certificate of Competence in this assessment, but the results of ductus flow were not used in the clinical management of the patients. The pregnancies were diagnosed as being dichorionic or monochorionic by the presence or absence, respectively, of placental tissue in the base of the inter-twin membrane (lambda sign).4 Gestational age was calculated from the last menstrual cycle unless there was a discordance of more than 6 days from the gestation estimated from the CRL of the fetus with the longest measurement. In each fetus, ductus venosus flow also was assessed.
The Doppler studies were undertaken during fetal quiescence. The magnification of the image was such that the fetal thorax and abdomen occupied the whole screen. A right ventral midsagittal view of the fetal trunk was obtained, and color flow mapping was used to demonstrate the umbilical vein, ductus venosus, and fetal heart. The pulsed Doppler sample was small (0.5–1.0 mm) to avoid contamination from the adjacent veins, and it was placed in the portion immediately above the umbilical sinus. The insonation angle was less than 30°, the filter was set at a low frequency (50–70 Hz) to allow visualization of the whole waveform, and the sweep speed was high (2–3 cm/s) so that the waveforms were spread widely, allowing better assessment of the a-wave. The waveforms were considered to be abnormal if the a-wave was reversed.5
Our policy for the follow-up of monochorionic twins includes ultrasound examinations at 16–18 weeks of gestation and monthly thereafter, unless there is evidence of twin–twin transfusion syndrome, in which case the frequency of examinations is increased as necessary. In cases of severe twin–twin transfusion syndrome, endoscopic laser coagulation of the communicating placental vessels is performed.6 The indications for such treatment are 1) ultrasound diagnosis of hydramnios in one twin and anhydramnios in the other and 2) absent or reversed end diastolic flow in either the umbilical artery or ductus venosus in one or both fetuses. In dichorionic twins, ultrasound examinations are performed routinely at 16–18 weeks and monthly thereafter.
Patients were asked to complete a questionnaire on maternal age, ethnic origin (white, African, Indian or Pakistani, Chinese or Japanese, and mixed), method of conception (natural or by in vitro fertilization [IVF]), and cigarette smoking during pregnancy (yes or no). Maternal weight and height were measured, and body mass index (BMI) was calculated as kg/m2. Written informed consent was obtained from each patient participating in the study, which was approved by the King’s College Hospital Research Ethics Committee.
Demographic details, ultrasound findings, fetal interventions, and pregnancy outcome were entered into a computer database. Pregnancy outcome was obtained from the maternity units or from the patients themselves.
In each pregnancy, the inter-twin discordance in nuchal translucency and CRL was calculated as the difference in each measurement between the two fetuses (largest-smallest) expressed as a percentage of the bigger measurement.
Comparisons between each of the pregnancy outcomes and the normal outcome group and between monochorionic and dichorionic pregnancies was by the Mann-Whitney U-test for continuous variables and the χ2 test and Fisher exact test for categorical variables. Kolmogorov-Smirnov test was used to test for normality of the distributions of maternal age and BMI. We used post hoc Bonferroni correction for the multiple comparisons, dividing the critical value for significance (.05) by the number of tests we had conducted, which was four for dichorionic and six for monochorionic twins.
Multiple logistic regression analysis in the whole population was used to determine the significant contributors to reversed a-wave in the fetal ductus venosus from maternal characteristics, chorionicity, different outcome groups, and inter-twin discordance in CRL and nuchal translucency. Similarly, multiple logistic regression analysis was performed to determine the significance of the contribution to each outcome group of reversed a-wave and inter-twin discordance in CRL and nuchal translucency and maternal characteristics.
The statistical software package SPSS 12.0 (SPSS Inc., Chicago, IL) was used for all data analyses.
During the study period (January 2006 to January 2008), we prospectively examined 733 diamniotic twin pregnancies with two live fetuses at 11 0/7 to 13 6/7 weeks of gestation and successfully obtained ductus flow waveforms from both fetuses in 728 (99.3%) cases. We excluded 33 (4.5%) cases for which there was a missing outcome. In the remaining 695 pregnancies, 516 were dichorionic and 179 were monochorionic. Measurement of fetal nuchal translucency was carried out in all fetuses. The median maternal age was 33.3 (interquartile range 29–36) years, the median gestational age was 89 (interquartile range 86–92) days, and the median BMI was 25.0 (interquartile range 22.5–28.3). In 555 (79.9%) cases, the mother was white, in 92 (13.2%) she was African, in 30 (4.3%) Indian or Pakistani, in 7 (1.0%) Chinese or Japanese, and in 11 (1.6%) of mixed ethnic origin. The maternal demographic characteristics in monochorionic and dichorionic twin pregnancies are shown in Table 1.
The prevalence of reversed a-wave in at least one of the fetuses was significantly higher in pregnancies complicated by miscarriage (28.6%, 95% confidence interval [CI] 13.6–50.2%, P=.005), fetal aneuploidy (70.0%, 95% CI 39.2–89.7%, P<.001), and twin–twin transfusion syndrome (38.5%, 95% CI 22.4–57.5%, P<.001) compared with the pregnancies with two healthy live births (7.7%, 95% CI 5.8–10.1%).
Pregnancy outcome was normal in 565 of the 619 (91.3%, 95% CI 88.8–93.3%) pregnancies with normal a-wave in both twins and in 47 of the 76 (61.8%, 95% CI 50.6–72.0%) of the pregnancies with reversed a-wave in at least one of the fetuses. Pregnancy outcome was normal in 33 of the 43 (76.7%, 95% CI 62.1–87.0%) dichorionic and in 14 of the 33 (42.4%, 95% CI 27.2–59.2%) monochorionic twins with reversed a-wave in at least one of the fetuses.
Of the 516 dichorionic twin pregnancies, there were 484 (93.8%, 95% CI 91.4–95.7%) that resulted in the live birth of two healthy babies, 10 (1.9%, 95% CI 0.9–3.5%) that resulted in miscarriage, eight (1.6%, 95% CI 0.7–3.0%) in which there was death of one fetus and live birth of the other, nine (1.7%, 95% CI 0.8–3.3%) in which one of the fetuses had a chromosomal abnormality (five with trisomy 21 and one each of trisomy 18, trisomy 13, triploidy, and partial deletion of chromosome 18) and the other twin was euploid, and five (1.0%, 95% CI 0.4–2.3%) in which one of the fetuses was healthy and the other had a major defect (one case each of Dandy Walker malformation with major ventricular septal defect, spina bifida, severe ventriculomegaly, transposition of the great arteries, and atrioventricular septal defect).
Reversed a-wave in the ductus venosus was observed in 47 (4.6%, 95% CI 3.4–6.0%) of the 1,032 fetuses and in at least one of the twins in 43 (8.3%, 95% CI 6.1–11.1%) of the pregnancies (Table 2). The prevalence of reversed a-wave was significantly higher in the pregnancies with chromosomal abnormalities than in those with normal outcome (66.7% compared with 6.8%).
Of the 179 monochorionic twin pregnancies, there were 128 (71.5%, 95% CI 64.3–78.0%) that resulted in the live birth of two healthy babies, 11 (6.2%, 95% CI 3.1–10.7%) that resulted in miscarriage, six (3.4%, 95% CI 1.2–7.2%) in which there was death of one fetus and either live birth of the other (n=5) or termination of pregnancy because the co-twin developed intraventricular and periventricular hemorrhage, one (0.6%, 95% CI 0.01–3.1%) in which the fetuses had trisomy 21, three (1.7%, 95% CI 0.4–4.8%) in which one of the fetuses was healthy and the other had a major defect (one case each of transposition of the great arteries, tetralogy of Fallot, and pulmonary atresia), 26 (14.5%, 95% CI 9.7–20.6%) that developed severe twin–twin transfusion syndrome (laser endoscopic surgery was carried out in 24 cases, and there was fetal death before surgery could be performed in two cases), and four (2.2% 95% CI 0.6–5.6%) in which there was a selective severe intrauterine growth restriction for which an endoscopic laser surgery was performed.
Reversed a-wave in the ductus venosus was observed in 45 (12.6%, 95% CI 9.3–16.5%) of the 358 fetuses and in at least one of the twins in 33 (18.4%, 95% CI 13.0–24.9%) of the pregnancies (Table 2). The prevalence of reversed a-wave was significantly higher in the pregnancies resulting in twin-to twin transfusion syndrome than in those with normal outcome (38.5% compared with 10.9%).
The prevalence of reversed a-wave in the ductus venosus was significantly higher in all monochorionic twins than in all dichorionic twins (18.4% compared with 8.3%, P<.001) but not in the pregnancies with normal outcome (10.9% compared with 6.8%, P=.171). Logistic regression analysis in the total population of 695 twin pregnancies demonstrated that, in addition to chorionicity, significant contributors to reversed a-wave in the ductus venosus were provided by inter-twin discordance in CRL and nuchal translucency, fetal aneuploidy, and development of severe twin–twin transfusion syndrome (Table 3).
Multiple logistic regression analysis to predict pregnancies with at least one fetal aneuploidy demonstrated significant contribution by reversed a-wave in at least one of the fetuses (odds ratio [OR] 11.27, 95% CI 2.32–54.73, P=.003) and inter-twin discordance in nuchal translucency (OR 1.07, 95% CI 1.03–1.11, P<.001) but not chorionicity (P=.413), maternal age (P=.354), ethnicity (P=.587), BMI (P=.979), conception by IVF (P=.830), smoking (P=.997), or inter-twin discordance in CRL (P=.056).
In the total of 695 pregnancies, there were 21 (3.0%, 95% CI 2.0–4.6%) miscarriages. Multiple logistic regression analysis demonstrated that significant contribution to miscarriage was provided by reversed a-wave in at least one of the fetuses (OR 3.38, 95% CI 1.16–9.83, P=.025), monochorionicity (OR 4.02, 95% CI 1.60–10.11, P=.003), inter-twin discordance in nuchal translucency (OR 1.04, 95% CI 1.00–1.07, P=.03), and maternal African ethnic group (OR 4.64, 95% CI 1.68–12.80, P=.003) but not inter-twin discordance in CRL (P=.074), maternal age (P=.094), BMI (P=.629), conception by IVF (P=.163), or smoking (P=.855).
In the total of 695 pregnancies, there were six (0.9%, 95% CI 0.4–1.9%) with major cardiac defects in one of the fetuses. The prevalence of reversed a-wave in the pregnancies with cardiac defects (50%, three of six) was not significantly higher than in the pregnancies with normal outcome (7.7%, 47 of 612) (P<.008). For this difference to have been significantly different with a power of 80% we would have needed to examine seven pregnancies with fetal cardiac defects and 748 with normal outcome.
In the total of 179 monochorionic pregnancies, there were 26 (14.5%, 95% CI 10.1–20.5%) that developed severe twin–twin transfusion syndrome. In the group with reversed a-wave in at least one of the fetuses, the prevalence of severe twin–twin transfusion syndrome was 30.3% (95% CI 17.3–47.5%, 10 of 33) compared with 11.0% (95% CI 6.8–17.2%, 16 of 146) in those with normal a-wave in both fetuses (P=.01). The mean inter-twin discordance in nuchal translucency was 19.6% (95% CI 10.9–28.3%) in the twin–twin transfusion syndrome group and 16.7% (95% CI 14.4–19.0%) in the non–twin–twin transfusion syndrome group (P=.780). Multiple logistic regression analysis demonstrated that significant contribution to severe twin–twin transfusion syndrome was provided by reversed a-wave in at least one of the fetuses (OR 5.09, 95% CI 1.94–13.37, P=.001) but not inter-twin discordance in nuchal translucency (P=.161) or CRL (P=.175), maternal age (P=.108), ethnicity (P=.999), BMI (P=.490), conception by IVF (P=.347), or smoking (P=.327).
The findings of this study in twin pregnancies that the prevalence of reversed a-wave in the ductus venosus at 11–13 weeks of gestation is more common in fetuses with aneuploidies and miscarriage are compatible with the data in singleton pregnancies.1 An additional factor affecting the prevalence of reversed flow in twins is chorionicity, being more common in monochorionic pregnancies and in particular those that subsequently develop severe twin–twin transfusion syndrome.
Studies in singleton pregnancies have shown that abnormal flow in the ductus venosus is a good marker for aneuploidies and improves the performance of first-trimester screening for trisomy 21 provided by a combination of maternal age, fetal nuchal translucency thickness, and maternal serum free ß–hCG and pregnancy-associated plasma protein A.1,7–9 In our study of twins, there was a similarly high prevalence of reversed a-wave in aneuploid compared with euploid fetuses as in singleton pregnancies,1 and assessment of the ductus venosus flow improved the prediction of aneuploidies provided by screening with fetal nuchal translucency alone. Nevertheless, this study of twins is too small, in comparison with that of singletons, to provide accurate prediction of the estimated improvement in the performance of screening with the inclusion of ductus venosus flow.
In our study of twins, as in our previous study in singletons,1 the rate of miscarriage in those with reversed a-wave and in those of African ethnicity was increased. The miscarriage rate was also higher in monochorionic than in dichorionic twins, which may be due to early twin–twin transfusion syndrome before 16 weeks, when endoscopic laser surgery can be performed. Another possible explanation for the association of early fetal death and miscarriage with reversed a-wave is that this abnormal flow pattern could be the consequence of impaired placentation leading to increased cardiac afterload.
The prevalence of reversed a-wave in fetuses from singleton pregnancies and dichorionic twins is similar, but in monochorionic twins it is substantially higher. In one third of monochorionic twins with reversed a-wave, there was subsequent development of severe twin–twin transfusion syndrome, and, therefore, the prevalence of abnormal flow was increased compared with dichorionic twins, even in those monochorionic twins that did not develop severe twin–twin transfusion syndrome. Thus, in monochorionic twins examined at 11–13 weeks, the rate of subsequent development of severe twin–twin transfusion syndrome is about 15%; this is increased to 30% if there is reversed a-wave in at least one of the fetuses and reduced to about 10% in those with normal a-wave in both fetuses. Consequently, ductus venosus flow at 11–13 weeks helps to modify the risk for subsequent development of severe twin–twin transfusion syndrome and the intensity of monitoring of such pregnancies. It is possible that, in the pregnancies we studied with abnormal Doppler and normal outcome, there was spontaneous resolution of twin–twin transfusion syndrome between the first and second trimesters, and, in contrast, in those with normal Doppler and subsequent twin–twin transfusion syndrome, the hemodynamic abnormality developed during the second trimester.
In twins, reversed a-wave in the ductus venosus at 11–13 weeks of gestation is associated with increased risk for aneuploidies, miscarriage, and development of severe twin–twin transfusion syndrome. However, in about 75% of dichorionic twins and 40% of monochorionic twins with reversed a-wave, the pregnancy outcome is normal.