Ductus Venosus Feasibility Study

Ann E Quinton. BAppSc(MRT),GradDipAppSc (Medical Ultrasonography), Senior Sonographer, Feto-Maternal Unit, Division of Women and Child's Health, Liverpool Hospital NSW

ASUM'99 - First Prize - The Giulia Franco Poster Award (Clinical or Technical Research) sponsored by Toshiba (Australia) Pty Ltd Medical Division.

Abstract
The ductus venosus is part of the compensatory regulatory system in the human fetus. The aim of this study was to determine if the ductus venosus could be identified and sampled in the fetus in a cross-sectional study involving 50 pregnant women. The feasibility of incorporating a ductus venosus pulsed Doppler measurement into an obstetric ultrasound examination in the clinical setting is discussed. An evaluation of the limiting factors when sampling and ultrasound techniques are discussed. The ductus venosus was successfully sampled in 49/50 (98%) of women with a mean time of 2.14 minutes (range 0.43-11.77). In comparison the umbilical artery was sampled in 50/50 participants with a mean time of 0.86 minutes (range 0.15-3.40). The time to sample the ductus venosus is statistically significant when compared to the umbilical artery (p< 0.0001). However in the clinical situation this difference would not be considered important. The results show the ductus venosus can be sampled in a varied cross-section of pregnant women. A transverse section of the fetal abdomen was the ultrasound imaging plane of choice, allowing the ductus venosus to be sampled in 98% of fetuses. The ductus venosus is a high velocity vessel, most easily identified by the aliasing produced by colour Doppler ultrasound.

Introduction
The Ductus Venosus is a small funnel shaped vessel that is found in the fetal liver, connecting the umbilical vein and the inferior vena cava. It is part of the compensatory regulatory system in the fetus. It preferentially directs oxygenated blood from the placenta to the fetal coronary arteries and brain by directing the blood via the right atria through the foramen ovale to the left side of the heart. Previous work done has supported the "animal study" hypothesis that the ductus venosus directs half of the oxygenated umbilical vein blood directly through the foramen ovale to the left atrium. Excessive mixing with the deoxygenated inferior vena cava blood entering the right atrium is therefore avoided (1).

As a fetus becomes compromised (one cause being placental vascular insufficiency) changes in the fetal arterial system are seen using pulsed Doppler ultrasound. Fetal hypoxaemia leads to decreased flow in diastole in the umbilical artery waveforms and increased middle cerebral artery flow (2). As the placenta shows increasing resistance to flow, the fetus attempts to compensate. The ductus venosus - left heart pathway is part of the compensatory mechanism. Doppler waveforms of the fetal venous system (umbilical vein, inferior vena cava and ductus venosus) can be used to monitor these attempts. It has been shown that the preferential flow of blood in the ductus venosus is maintained with peak velocities staying within the normal range, even with fetuses in advanced stages of intra-uterine growth retardation. However, haemodynamic compromise was found when the atrial component of the ductus venosus reduced or became reversed (3). A high perinatal mortality rate has been found in fetuses with abnormal venous Doppler waveforms, with seven of seven fetuses dying in utero in one study (3) and a mortality rate of five out of eight fetuses in another (4).

In 1997 Hecher K and Hackeloer BJ compared more established forms of fetal surveillance, a cardiotocogram (CTG) with fetal Doppler investigations (5). They reported changes in the ductus venosus preceded CTG changes when trying to determine the optimal time for delivering growth retarded fetuses. Indeed, they suggested that venous Doppler waveforms might be a better test of fetal well being than a CTG.

Aim
The aim of this project was to determine if the ductus venosus could be identified and successfully sampled in the fetuses of a cross-section of pregnant women, using grey-scale, colour and pulsed Doppler ultrasound. The feasibility of incorporating a ductus venosus pulsed Doppler measurement into an obstetric ultrasound examination in the clinical setting was assessed. Doppler studies of the ductus venosus may prove to be of assistance in high-risk pregnancies. It may help in prolonging pre-term pregnancies, in determining the optimal time for delivering the "at risk" fetus and decrease the false positive rate of "a compromised fetus." An evaluation of limiting factors was suggested and recommendations on ultrasound scanning techniques made. The technique suggested by DeVore GR and Horenstein J to sample the ductus venosus was used as a starting point (6). Several studies and personal experience have shown that the success rate of sampling the ductus venosus is not 100% within the normal time constraints of a routine scan (7,8).

Method
The study population was obtained by recruiting 50 pregnant patients between 20-40 weeks gestation referred to the Feto-Maternal Unit for an ultrasound assessment. A patient information sheet and consent form was given to the women before their ultrasound and any questions answered. The study was cross-sectional with the subjects recruited consecutively from the routine and high-risk antenatal population. Approval for this project was granted by the South Western Area Health Service Ethics Committee and the University of Sydney, Faculty of Health Sciences.

Gestational age was determined by an ultrasound performed before 20 weeks gestation. Maternal variables recorded were height, weight, age and ethnicity. Pregnancy details included gestational age, liquor volume, fetal position and activity. The ultrasound studies of the fetal circulation were performed using a Siemens Sonoline Elegra with a 3.5 MHz or 5MHz curved linear transducer. Spatial peak temporal average intensities were below the 100mW/cm2 recommended for fetal work. All recordings were done in the absence of fetal breathing.

Umbilical artery waveforms were recorded on all subjects and placental resistance was measured using the pulsatility index (PI). The PI was recorded as normal or abnormal. An abnormal PI was defined as being above the 95th percentile (9). The time taken to obtain an umbilical artery waveform was recorded. Before sampling the ductus venosus, fetal position was determined and recorded. Tracing the umbilical vein in the fetal abdomen either in a sagittal or transverse section using colour Doppler and grey scale imaging identified the ductus venosus. The colour Doppler velocity scale was adjusted so that no aliasing was demonstrated in the umbilical vein. As the ductus venosus is a narrow funnel shaped vessel with high velocity flow, it will demonstrate aliasing with these colour Doppler velocity settings (6). The pulsed Doppler sample gate was placed over the ductus venosus where it exited the umbilical vein and the pulsed Doppler was activated (figure 1). It has been reported that peak velocity waveforms of the ductus venosus isthmus are registered at this point (10). A spectral waveform of the ductus venosus was obtained and the angle of isonation was adjusted to less than 60 degrees. Velocity measurements and hard copy film records were made.

Figure 1: A pulsed Doppler waveform of the ductus venosus, which is sampled where it exits the umbilical vein.

A ductus venosus sample was judged to be satisfactory if three stable consecutive heart cycles were obtained and the angle of isonation was less than 60 degrees. Failure to obtain a sample was also recorded. The time was recorded by printing onto thermal paper film, the ultrasound screen, with its time keeping clock. This was done before the pulsed Doppler recordings were attempted and after they were obtained. The depth from the ultrasound transducer to ductus venosus was recorded. All variables were recorded on an information sheet and later entered into a database.

Results
All data were analysed using the statistical package SPSS (Statistical Package for the Social Sciences). Fifty participants were recruited for this cross-sectional study. The women's mean age was 28.5 years (range 19-42). Their height ranged from 147cm-178cm (mean 162.4) and their mean weight was 66.24kg (range 42-110). Gestational age at the time of sampling ranged from 20.71 weeks to 39.42 weeks (mean 30.22) - figure 2. Eighty two percent of the women were Caucasian, six percent Asian, six percent Mediterranean and six percent unknowns (total 100%).

Figure 2: Histogram showing normal distribution curve of gestational age at time of study.

A pulsed Doppler sample of the ductus venosus was successfully obtained in 49/50 participants (98%), in comparison to 100% success rate for the umbilical artery. The depth of the ductus venosus ranged from 4.5cm-15cm (mean 7.8). The mean time taken to sample the umbilical artery was 0.86 +/- 0.52 minutes (mean +/-SD). The range is shown in figure3. Ductus venosus mean sampling time was 2.14 +/- 1.98 minutes (mean +/- SD). The range is shown in figure 4. Statistically the time was significantly longer for the ductus venosus (p<0.0001: Mann Whitney U). The Mann Whitney U test was used, as the data was not normally distributed as shown in figures 3 and 4. The ductus venosus was sampled in 98% (48/49) of fetuses using a transverse section through the fetal abdomen and 2% (1/49) using a sagittal section. Normal umbilical artery Dopplers were recorded in 48/50 samples (96%) and abnormal umbilical artery Dopplers in 2/50 or 4%.

Figure 3: Histogram of time taken to sample umbilical artery. Plot is skewed to the right.

Figure 4: Histogram of time taken to sample ductus venosus. Plot is skewed to the right.

Fetal position was recorded as shown in Table 1.

Table 1. Fetal positions.

Liquor volume was recorded as normal, polyhydramnios or oligohydramnios. The volume was defined as normal if the amniotic fluid index (AFI) was between 5-25cm with a maximum pool >2cm and <10cm. Polyhydramnios was defined with an AFI>25cm and a maximum pool >10cm. Oligohydramnios was defined when there was no maximum pool >2cm and the AFI< 5cm. Eighty six percent (43/50) had normal liquor volume, 8% (4/50) had polyhydramnios and 6% (3/50) had oligohydramnios.

Discussion
This study demonstrated that by using a combination of grey scale, colour Doppler and pulsed Doppler ultrasound it is possible to identify the ductus venosus in-utero. This can be done across a range of women of different heights, weights and at different gestational ages. Umbilical artery Doppler studies are already a routine part of second and third trimester growth and biophysical profile ultrasounds. With a success rate of 98% for sampling the ductus venosus it is feasible to incorporate this measurement into an obstetric ultrasound. Even though the ductus venosus measurement took statistically significantly longer (p<0 .0001) in the clinical setting an average time of 2.14 minutes compared to 0.86minutes would not be considered important.

As the ductus venosus is such a small vessel, the easiest way to identify and sample it in the majority of cases was by the aliasing produced with colour Doppler (6). However when fetal position did not allow this, the ductus venosus could be found by moving the pulsed Doppler sample gate around the area where the ductus venosus arises from the umbilical vein and listening for its unique high velocity Doppler shift. The majority (98%) of ductus venosus samples was obtained in a transverse imaging plane of the fetal abdomen (figure 5). Only once was it more easily done in a para-sagittal section. As a transverse section of the fetal abdomen is routinely done when measuring the abdominal circumference, sampling the ductus venosus could become an extension of this and would be expected to decrease the additional time taken to obtain this measurement.

Figure 5: A transverse section of the fetal abdomen demonstrating the identifying aliasing in the ductus venosus.

The only limiting factor identified when sampling the ductus venosus was maternal obesity with a late second trimester fetus. Other more obese women were successfully scanned, however the gestational age, and therefore fetal size was greater. As with all types of obstetric ultrasound, maternal size and gestational age play a part. Fetal position did not prevent sampling. As most of the ductus venosus measurements were done during a quiet phase of activity, fetal movement and breathing did not have any effect on the ability to sample.

As the ductus venosus may be of assistance in high-risk pregnancies, measurements were done in pregnancies complicated with abnormal Dopplers, oligohydramnios or polyhydramnios. All high-risk pregnancies were successfully sampled. With increasing publications of changes in venous Doppler waveforms when abnormal umbilical artery Dopplers are found (2-6,11,12) it is noteworthy that the ductus venosus can be sampled in a varied cross-section of pregnant women within a reasonable period of time.

A description of the ductus venosus in the human fetus and its ultrasound appearance has been previously published (6,13). This study agrees with those findings and verifies the methods used to obtain the ductus venosus waveform.

A limitation of this study could be that the recordings were done by a single operator with experience in obtaining fetal ductus venosus Doppler ultrasound waveforms. This could explain the bias in the histograms in figures 3 and 4. A further study using different operators of varying experience may give a more normal distribution.

Conclusion
This study has shown that the ductus venosus can be sampled in a cross-sectional population of pregnant women in the clinical setting. A transverse cross-section of the fetal abdomen allowed the ductus venosus to be visualised and sampled successfully. Aliasing produced by the high velocity ductus venosus allowed for easy recognition. Incorporating the ductus venosus into fetal welfare studies may help in optimising the management of the at risk fetus.

Acknowledgements
I would like to acknowledge the support of Dr JS Smoleniec from Liverpool Hospital and Ms C-M Cook from Nepean Hospital for their support and encouragement.

References