Screening for aneuploidy in the first trimester by assessment of blood flow in the ductus venosus

Authors


*Dr B. Thilaganathan, Fetal Medicine Unit, Academic Department of Obstetrics and Gynaecology, St George's Hospital Medical School, Cranmer Terrace, London SW17 ORE, UK.

Abstract

Objective To assess the role of ductus venosus Doppler assessment in screening for fetal aneuploidy in pregnancies at 11–14 weeks of gestation.

Design Prospective observational study.

Setting In a tertiary referral fetal medicine unit.

Population Two hundred fifty-six consecutive pregnancies between 11 and 14 weeks of gestation referred to our unit.

Methods Nuchal translucency was measured and colour Doppler imaging was used to assess normal (forward) or abnormal (absence/reversed) ductus venosus flow during atrial contraction.

Main outcome measures Fetal karyotype, structural abnormalities and pregnancy outcomes.

Results The nuchal translucency measurement was ≥95th centile in 90 (35.2%) pregnancies, and 123 (48.0%) had a nuchal translucency-related risk of ≥1:300. There were 46 chromosomally abnormal pregnancies and six euploid pregnancies with adverse outcome. The prevalence of abnormal ductus venosus Doppler waveforms in normal pregnancies was 4.49% (95% CI 1.76–8.76%). The sensitivities of abnormal nuchal translucency measurement alone, ductus venosus velocimetry alone or nuchal translucency and ductus venosus combined for Down's Syndrome were 80.4%, 58.7% and 93.5%, respectively. The likelihood ratios for aneuploidy with abnormal nuchal translucency measurement alone, ductus venosus velocimetry alone or nuchal translucency and ductus venosus combined were 3.33, 9.83 and 3.48, respectively.

Conclusion There is clear association between abnormal flow in the ductus venosus and fetal aneuploidy. The use of ductus venosus velocimetry in combination with nuchal translucency is better than either test alone, since it increases the sensitivity in the detection of Down's Syndrome to 94% and decreases the likelihood ratio of a negative test to 0.08.

Introduction

Measurement of nuchal translucency is well established as an effective non-invasive method of screening for chromosomal abnormalities1. First trimester nuchal translucency screening can identify about 77% of aneuploidies, for a false positive rate of 5%2. In addition, nuchal translucency measurements greater than 3 mm are associated with an increased likelihood of major structural defects and genetic syndromes3–7. More recently, Doppler studies of the ductus venosus have been included in nuchal translucency screening programmes in an attempt to improve the effectiveness of screening for fetal aneuploidy8–13. However, the sensitivity of abnormal ductus venosus Doppler indices for aneuploidy in these studies has ranged from 55% to 91%10–13. The aim of this study was to assess the role of first trimester Doppler assessment of the ductus venosus in screening for fetal aneuploidy.

Methods

This was a prospective observational study carried out over 18 months. All women referred to the unit for first trimester assessment and karyotyping were invited to participate in the study. Indications for referral included increased risk for aneuploidy by nuchal translucency, maternal request for fetal karyotyping, fetal abnormality in a previous pregnancy and suspected fetal abnormality in the current pregnancy. Only women who consented to participate in the study were recruited. The study had approval from the local research ethics committee.

Ultrasound examinations were performed by four observers (all the authors) with at least two years of ultrasound experience, who were trained for at least three months in the technique of Doppler assessment of the ductus venosus. An ATL HDI 5000 ultrasound machine (Letchworth, UK) was used, with a 7 MHz curvilinear transabdominal probe. The high-pass filter was set to minimum and the pulse repetitive frequency was 2.5 kHz. The maximum achievable thermal and mechanical indices were 1.2 and 1.0, respectively, in keeping with current safety guidelines14. The exposure of the scan was limited to a maximum of 5 minutes and if satisfactory waveforms were not formed within that time, the Doppler investigation was abandoned. The size of the sampling gate was set to 1 mm.

Localisation of the ductus venosus was facilitated using colour and pulsed Doppler imaging. Flow velocity waveforms from the ductus venosus were obtained in the midsagittal plane of the fetal trunk (with the angle of insonation always being less than 30°), immediately distal to the portal sinus and proximal to the infundibulum of the inferior vena cava. This enabled avoidance of contamination from the intrahepatic portion of the umbilical vein, the left hepatic vein and inferior vena cava. The blood flow velocity waveforms were classified categorically as being normal (forward flow) or abnormal (absent or reversed flow) with regard to the pattern of blood flow during atrial contraction (Fig. 1). As previous studies showed poor reliability of quantitative Doppler indices in the ductus venosus, absolute Doppler values were not derived and thus only the categoric assessment of flow in the ductus venosus during atrial contraction was taken into account in the present study15. The Doppler examination in the ductus venosus was repeated a minimum of three times to confirm the assessment of flow in the ductus venosus during atrial contraction.

Figure 1.

(a) Forward (normal) flow in the ductus venosus during atrial contraction in a fetus at 13 weeks of gestation. (b) Reversed (abnormal) flow in the ductus venosus during atrial contraction in a fetus at 12 weeks of gestation.

During the ultrasound examination, a full structural survey was undertaken, the nuchal translucency was measured and the risk of Down's syndrome adjusted by the nuchal translucency was calculated. Fetal karyotyping was performed by chorionic villus sampling when indicated. In pregnancies with a nuchal translucency measurement >99th centile and normal karyotype, fetal echocardiography was performed at 12–16 weeks of gestation and a detailed ultrasound scan at 20 weeks of gestation. Ascertainment of structural defects was based on the ultrasound findings, examination at birth, postmortem examinations in cases of intrauterine death or termination of pregnancy and telephone interviews with the parents.

Results

During the study, 349 women underwent chorionic villus sampling, of whom 260 women agreed to be recruited for first trimester Doppler assessment of flow in the ductus venosus. The median maternal age was 35 years (range 15 to 42 years), the median crown–rump length was 65 mm (range 45 to 88 mm) and the median gestational age was 12 + 3 weeks (range 11 + 0 to 13 + 6 weeks). In four pregnancies (1.5%), blood flow in the ductus venosus could not be assessed within the 5 minutes allocated for this part of the investigation.

Nuchal translucency was ≥95th centile in 90 (35.2%) pregnancies, and 123 (48.0%) had a risk of Down's syndrome ≥1:300, adjusted for maternal age and nuchal translucency. There were 46 chromosomally abnormal pregnancies (Table 1), of which 9 had a nuchal translucency measurement <95th centile (Table 2). There were six adverse outcomes in the chromosomally normal fetuses (Table 3), where five pregnancies had increased nuchal translucency and two had abnormal blood flow in the ductus venosus. The frequencies of nuchal translucency ≥95th centile and abnormal flow in the ductus venosus are shown in Fig. 2.

Table 1.  The outcomes of pregnancy.
 Number of fetusesOutcome
  1. TOP = termination of pregnancy; LB = live birth; IUD = intrauterine death.

Normal outcome204 
Aneuploidies46 
Trisomy 2130TOP ×28, LB ×2
Trisomy 185TOP ×5
Trisomy 134TOP ×4
Klinefelter's Syndrome2LB ×2
Turner's Syndrome3TOP ×3
Triple X1LB ×1
Mosaic trisomy 81TOP ×1
Other adverse outcomes6 
Multiple congenital abnormalities2TOP ×2
Major cardiac defect1TOP ×1
Twin-to-twin transfusion syndrome1IUD ×1
Unexplained IUD2IUD ×2
Table 2.  Indication for prenatal diagnosis in chromosomally abnormal fetuses with normal nuchal translucency (NT) measurements.
CaseMA (years)GA (weeks + days)CRL (mm)NT (mm)Age related riskNT riskDV flowIndication for prenatal diagnosisAnomaly or karyotype
  1. MA = maternal age; GA = gestational age; CRL = crown rump length; ASD = atrial septal defect; AVSD = atrioventricular septal defect; T = trisomy.

13511 + 4550.81:4421:1395reversedmaternal requestASD/mosaic T8
23112 + 2631.61:5251:3692reversedholoprosencephalyT13
32911 + 5521.31:6341:4740reversedholoprosencephalyT13
42912 + 4672.11:7211:1904reversedomphalocoele, diaphragmatic herniaT18
53212 + 5581.71:4391:1395reversedmonochorionic twin sibling with increased NTT21
63913 + 0691.31:691:413absentadvanced maternal ageT21
74711 + 4492.11:81:17positiveincreased NTT21
84513 + 1661.91:121:48positiveincreased NTKlinefelter's Syndrome
94812 + 5691.81:51:24positiveAVSD on 20 week scanT21
Table 3.  Ductus venosus (DV) blood flow in chromosomally normal fetuses with adverse outcome.
CaseMA (years)GA (weeks + days)CRL (mm)NT (mm)NT riskDV flowAnomalyOutcome
  1. MA = maternal age; GA = gestational age; CRL = crown rump length; NT = nuchal translucency; TOP = termination of pregnancy; IUD = intrauterine death.

13212 + 2654.51:9positivehypoplastic left heart syndromeTOP at 14 weeks
22711 + 5527.21:15positivethoracic dystrophyTOP at 16 weeks
33012 + 1593.61:35positivemonoamniotic twinsIUD at 17 weeks
44212 + 3641.51:303positiveunexplained intrauterine deathIUD at 21 weeks
53413 + 2674.11:9reversedagenesis of corpus callosumTOP at 22 weeks
62912 + 3697.51:9reversedunexplained intrauterine deathIUD at 15 weeks
Figure 2.

Flow chart summarising the ductus venosus flow (DV) in pregnancies with normal (<95th centile) and increased (≥95th centile) nuchal translucency (NT) measurements in the study population.

The frequency of abnormal blood flow in the ductus venosus in pregnancies with normal nuchal translucency measurement (<95th centile) and normal karyotype was 4.49% (7/156). The sensitivities of abnormal nuchal translucency alone, ductus venosus velocimetry alone and nuchal translucency and ductus venosus combined for Down's Syndrome were 80.4%, 58.7% and 93.5%, respectively (Tables 4 and 5). The likelihood ratios for aneuploidy with abnormal nuchal translucency alone, ductus venosus velocimetry alone or nuchal translucency and ductus venosus combined were 3.33, 9.83 and 3.48, respectively (Table 5). There were too few adverse pregnancy outcomes to calculate the likelihood ratio for this outcome.

Table 4.  Efficiency of nuchal translucency (NT) alone, ductus venosus (DV) velocimetry alone and combined NT and DV velocimetry in screening for Down's Syndrome.
OutcomeNT alone
+ve−veTotal
  1. Outcome +ve = abnormal karyotype.

+ve37946
−ve48156204
Total85165250
OutcomeDV velocimetry alone
+ve−veTotal
+ve271946
−ve12192204
Total39211250
OutcomeCombined NT and DV velocimetry
+ve−veTotal
+ve43346
−ve55149204
Total98152250
Table 5.  Efficiency of nuchal translucency (NT) alone, ductus venosus (DV) velocimetry alone and combined NT and DV velocimetry in screening for Down's Syndrome. Assuming a population prevalence of 1:400 for Down's Syndrome, the odds of being affected given a positive result (OAPR), or posterior odds, are expressed as a risk.
 Sensitivity (%)Specificity (%)PPV (%)NPV (%)LR +veLR −veOAPR
  1. PPV = positive predictive value; NPV = negative predictive value; LR +ve = likelihood ratio of a positive test; LR −ve = likelihood ratio of a negative test.

NT807644953.330.261:122
DV599469919.830.441:42
NT + DV947344983.480.081:115

Discussion

This study shows that Doppler assessment of the ductus venosus is feasible in early pregnancy, and that there is a clear association between abnormal blood flow in the ductus venosus and aneuploidy. Absent or reversed flow during atrial contraction was observed in 59% of the chromosomally abnormal fetuses, and abnormal flow in the ductus venosus increased the risk for fetal aneuploidy 10-fold. More importantly, the combination of increased nuchal translucency and abnormal blood flow in the ductus venosus has a sensitivity of 94% for fetal aneuploidy.

The pregnancies in the study were selected, having been referred to our department for assessment of the risk of fetal aneuploidy by a combination of maternal age and nuchal translucency. Consequently, 48% of the pregnancies had an increased risk of aneuploidy. Doppler assessment of the ductus venosus is likely to be used in selected, high risk pregnancies, due to the specialised nature of the assessment. In view of the poor reliability of quantitative Doppler measurements of blood flow in the ductus venosus are unreliable15, and so we assessed blood flow categorically (normal, absent or reversed). Despite using the smallest sample gate16, overlap of signals from neighbouring vessels such as the umbilical vein or inferior vena cava may confound the interpretation of the results. We attempted to avoid this problem by repeated Doppler assessments of blood flow in the ductus venosus15.

In this study, the sensitivity of abnormal flow in the ductus venosus for fetal aneuploidy was 59%, which is similar to the 55% and 65% found in two recent studies12,13. Abnormal blood flow in the ductus venosus alone increases the risk of fetal aneuploidy 10-fold, and combining blood flow in the ductus venosus and nuchal translucency increases the sensitivity to 94%. However, the majority of aneuploid fetuses with abnormal flow in the ductus venosus had either increased nuchal translucency (21/27) or major structural abnormalities (3/27) detectable by an ultrasound scan in the first trimester. The association between abnormal flow in the ductus venosus and increased nuchal translucency suggests that these are not independent factors. Our findings are supported by those of Bilardo et al.12, who showed that sensitivity for fetal aneuploidy detected by abnormal blood flow in the ductus venosus was 65% and that there was an association between nuchal translucency and the pulsatility index in the ductus venosus. However, these authors included high pulsatility index in the ductus venosus in their definition of abnormal blood flow in the ductus venosus giving them a higher false positive rate for abnormal flow (17%) in the normal population, compared with our study (4%). Although Antolin et al.13 state that they did not identify an association between increased nuchal translucency and abnormal blood flow in the ductus venosus, they had only 14 aneuploid fetuses between 10 and 14 weeks of gestation in their study. Furthermore, 8 of the 12 fetuses with nuchal translucency ≥95th centile also had the pulsatility index in ductus venosus >95th centile, making it likely that the lack of a statistical association is due to the relatively small number of pregnancies studied.

Of the six pregnancies with normal karyotype that resulted in adverse outcomes, five had increased nuchal translucency and two had abnormal blood flow in the ductus venosus. These findings are at variance with a previous study, which suggested that 8 of the 10 pregnancies with adverse outcomes had abnormal flow in the ductus venosus12. The relatively small number of cases evaluated in these studies makes it difficult to assess the true role of blood flow in the ductus venosus in predicting adverse outcome. The value of abnormal flow in the ductus venosus in the detection of major congenital heart defects has is yet to be established; there was only one case of a major congenital heart defect in our study. The sensitivity of increased nuchal translucency in detecting major congenital heart disease is lower than previously assumed, making it necessary to carry out much larger studies to evaluate blood flow in the ductus venosus as a screening test for congenital heart disease17–19.

Conclusion

There is a clear association between abnormal blood flow in the ductus venosus and fetal aneuploidy. Abnormal flow increases the risk for fetal aneuploidy by 10-fold. Ductus venosus velocimetry in combination with increased nuchal translucency is better than either test alone, since it increases the sensitivity in the detection of Down's Syndrome to 94% and decreases the likelihood ratio of a negative test to 0.08. Hence, in a population with a background prevalence of 1:400 for Down's Syndrome, the risk of Down's Syndrome with a positive combined nuchal translucency and ductus venosus test is 1:115, whereas with a negative result the risk is 1:7981. The specialised nature of Doppler velocimetry of the ductus venosus performed in the first trimester may hinder acceptance of this test in screening.

Acknowledgements

Dr Eleni Mavrides was supported by the Special Trustees of St George's Hospital Medical School (Grant No. QK 00.6.58). The authors would like to thank the staff in the Fetal Medicine Unit at St George's Hospital in their help with the collection of data.

Ancillary