You are on page 1of 7

Article

Assessment of Umbilical Cord Coiling


During the Routine Fetal Sonographic
Anatomic Survey in the Second Trimester
Mladen Predanic, MD, MSc, Sriram C. Perni, MD,
Stephen T. Chasen, MD, Rebecca N. Baergen, MD,
Frank A. Chervenak, MD

Objective. The purpose of this study was to evaluate the sonographic accuracy to determine the
umbilical coiling index (UCI) during the routine fetal anatomic survey in the second trimester.
Methods. In 300 consecutive women with singleton pregnancies and absence of gross fetal anomalies who had a routine second-trimester fetal anatomic survey, a distance between 2 pairs of coils was
measured from the longitudinal images of the umbilical cord, and the antenatal UCI (aUCI) was calculated. The aUCI was compared with true UCI results obtained after birth. Results. Two hundred thirty-six patients had adequate sonographic umbilical cord images, and all required demographic,
antenatal, and labor data collection to meet the inclusion criteria. A statistically significant correlation
between aUCI and true UCI was found (P < .0001; r = 0.643). The mean aUCI was 0.402 (80% confidence interval, 0.382), and the true UCI at birth was 0.203 (80% confidence interval, 0.176). The
sonographic evaluation showed 12.3% and 8.9% of hypocoiled and hypercoiled cords, whereas evaluation at birth found 10.6% and 8.1% hypocoiled and hypercoiled umbilical cords, respectively. The
sensitivity values of sonography to predict hypocoiling and hypercoiling at birth were 78.9% and
25.4%, respectively. Conclusions. A sonographic evaluation of umbilical cord coiling in the second
trimester correlates with the true UCI at birth, although the sensitivity in predicting coiling patterns as
hypocoiled and hypercoiled cords is less accurate. A difference between the aUCI and matched UCI at
birth could be explained by a sonographic error in the sampling of different umbilical cord segments
with discordant coiling patterns or the possibility of a dynamically evolving UCI with advancing gestational age. Key words: coiling index; sonography; umbilical coiling.

Abbreviations
aUCI, antenatal umbilical coiling index; UCI, umbilical
coiling index

Received August 31, 2004, from the Division of


Maternal-Fetal Medicine, Department of Obstetrics
and Gynecology (M.P., S.C.P., S.T.C., F.A.C.), and
Department of Pathology and Laboratory
Medicine (R.N.B.), Weill Medical College of Cornell
University, New York, New York USA. Revision
requested September 13, 2004. Revised manuscript
accepted for publication September 21, 2004.
Address correspondence and reprint requests to
Mladen Predanic, MD, MSc, Division of MaternalFetal Medicine, Department of Obstetrics and
Gynecology, Weill Medical College of Cornell
University, 525 E 68th St, Suite M-704, New York,
NY 10021 USA.
E-mail: mlp2001@med.cornell.edu

ormal umbilical cord coiling is approximately


1 coil/5 cm of umbilical cord length or 0.20 to
0.24 coils/cm.14 The latter number represents
umbilical coiling quantified via the umbilical
coiling index (UCI).4 The UCI is calculated by dividing the
total number of coils by the total length of the cord immediately after delivery.4 Most published studies define
hypocoiled (undercoiled) or hypercoiled (overcoiled)
umbilical cords as below the 10th and above the 90th
percentiles, respectively.2,46 In one of the largest studies,
of 1329 umbilical cords, a total of 13% hypocoiled and
21% hypercoiled umbilical cords were found.1 The presence of hypocoiled and hypercoiled cords was associated
with fetal death (21% and 37%, respectively), fetal intoler-

2005 by the American Institute of Ultrasound in Medicine J Ultrasound Med 2005; 24:185191 0278-4297/04/$3.50

Assessment of Umbilical Cord Coiling

ance of labor (15% and 14%, respectively), and


intrauterine growth restriction (29% and 10%,
respectively).1 These findings were confirmed by
others,24 and adverse fetal outcomes were
attributed to abnormal coiling that likely predisposed umbilical cord vessels to thrombosis, constriction, or both.1
In an attempt to prognosticate adverse fetal outcomes, a sonographic evaluation of umbilical
cord coiling found a significant correlation
between antenatal UCI (aUCI) and postnatal UCI
measurements, but these studies were limited to
the third trimester or the immediate postpartum
period.57 However, an attempt to establish the
correlation between the aUCI obtained in the second trimester and the UCI at delivery showed less
compelling results. Under the assumption that
the umbilical coiling is fully developed by the end
of the first trimester and does not change thereafter but, rather, that the cord lengthens between
established coils,8 the true UCI should be predictable from the sonographic assessment in the
second trimester.9 Although the aUCI was measured easily and reliably on second-trimester
sonography, it did not accurately reflect the true
UCI at term.9 A discordance between the aUCI
and UCI was attributed to the presence of
mixed coiling patterns or possibly to the evolution of UCI at latter gestation.
The objectives of this study were to evaluate the
sonographic accuracy during the routine fetal
anatomic survey in the second trimester for
determination of umbilical cord coiling and to
establish the relationship between the aUCI and
true UCI obtained at birth.

Materials and Methods


This prospective cross-sectional study was performed to assess the relationship between antenatal and true umbilical cord coiling at birth.
Three hundred thirty-one women had a routine
fetal anatomic sonographic survey at 18 to 23
weeks gestational age performed between
August 1, 2003, and September 15, 2003. Three
hundred consecutive women with singleton
pregnancies, absence of gross fetal anomalies,
and planned delivery at our institution were
included in the study. All sonographic examinations were performed with a 6-MHz transabdominal transducer with multihertz and
harmonic capability (Acuson Sequoia 512;
Siemens Medical Solutions, Mountain View, CA).
186

Sonographic images including longitudinal


views of the umbilical cord that were obtained at
maximal magnification were stored on digital
imaging media (KinetDx version 2.8.3; Siemens
Medical Solutions) and were available for retrospective assessment and remeasuring of all fetal
parameters as needed. An evaluation of the
umbilical cord and recording of longitudinal
cord images for assessment of cord coiling added
an additional 2 to 3 minutes to the routine fetal
anatomic survey. The gestational age of the pregnancy was determined in relation to the estimated date of confinement, as defined by 280 days
from the last menstrual period that was either
less than 7 days discrepant from first-trimester
sonography or 14 days different from secondtrimester sonography. Otherwise, the estimated
date of confinement from the earliest sonographic examination was used. Each patient was
included only once. Patients medical records
were reviewed for maternal demographic characteristics, antepartum complications, and birth
data.
The exclusion criteria were multifetal pregnancy, inadequate demographic, antenatal, and
labor data, and an inadequate or inappropriate
longitudinal image of the umbilical cord to allow
for an accurate aUCI measurement, as well as the
presence of a single-artery umbilical cord.
The images of the umbilical cord were retrieved
from the digital media within 24 hours of the
sonographic examination, and the distance
between the coils was measured from the inner
edge of an arterial or venous wall to the outer
edge of the next coil along 1 side of the umbilical
cord (Figure 1). The distance between the adjacent coils was measured with the built-in calipers
(straight line) or by tracing the distance along
the cord side if the cord was substantially curved.
The aUCI was calculated as a reciprocal value of
the distance between a pair of coils (aUCI = 1/distance in centimeters). All sonographic measurements were performed by a single author (M.P.),
whereas a true UCI was evaluated within the 24
hours of delivery by another author (R.N.B.), who
was blinded to the aUCI findings. The umbilical
cord attachment to the placenta, including any
cord segment that had been removed, was examined, and the UCI was determined by dividing the
number of umbilical cord twists by the total cord
length. Because all antenatal UCI measurements
were performed by a single author (M.P.), only an
evaluation of intraobserver variation was perJ Ultrasound Med 2005; 24:185191

Predanic et al

formed. Intraobserver variation was assessed on


50 sets of 3 repeated aUCI measurements.
A statistical analysis was performed with Prism
version 3.02 software (GraphPad Software, Inc,
San Diego, CA). The Spearman rank correlation
test for a nongaussian sample and polynomial
regression least squares correlation were used
appropriately, whereas arithmetic mean, SD,
median, and range of values for recorded variables presented maternal demographic data. An
interclass reliability analysis was performed to
assess intraobserver variance. The Institutional
Review Board of the Weill Medical College of
Cornell University evaluated and approved the
study.

Results
During the study period, a total of 331 pregnancies had fetal anatomic scans performed
between 18 and 23 weeks gestation. Thirty-one
patients were excluded on the basis of multifetal
pregnancy (21 patients), fetal anomalies (7
patients), and the presence of a 2-vessel umbilical cord (3 patients). From the remaining 300
consecutive patients who were initially included
in the study, only 236 had adequate sonographic
umbilical cord images and all demographic,
antenatal, and labor data to meet the inclusion
criteria. The mean gestational age SD was 20.4
0.9 weeks, and the mean maternal age was 31.8
5.1 years (range, 1447 years), with median parity of 0 (range, 05). Most patients gave birth at
Figure 1. Longitudinal (axial) sonogram of the umbilical cord
with the normal UCI (normally coiled). A distance of 22.6 mm
was measured between the 2 pairs of coils from the inner edge
of the artery to the outer edge of the same artery at the adjacent umbilical twist along the ipsilateral cord side.

J Ultrasound Med 2005; 24:185191

term, and the mean gestational age at delivery


was 38.9 1.6 weeks (range, 3042 weeks).
Normal vaginal delivery was achieved in 149
patients (63.1%); 87 patients (36.9%) had surgical
delivery; 76 (32.2%) had cesarean delivery; and
11 (4.7%) had forceps- or vacuum-assisted vaginal delivery. The mean neonatal birth weight was
3376 499.8 g (range, 11954650 g). The mean
Apgar scores were 8.6 0.9 (range, 19) and 9.0
0.3 (range, 610) at 1 and 5 minutes, respectively.
A comparison between the aUCI and true UCI
showed a statistically significant correlation
(Spearman correlation, P < .0001; r = 0.643). The
polynomial least squares regression equation
between the aUCI and UCI was y = 0.06215 +
(0.4367 x) (0.1851 x2), where y and x were
the UCI and aUCI, respectively. Figure 2 represents a polynomial second-order regression
curve with the mean, 10th, and 90th percentile
curves superimposed over individual aUCI
measurements.
From the aforementioned scattergram, it
appears that the aUCI value was approximately
twice that of the UCI measurement at birth. If
translated into the distance between the coils, a
distance between a pair of coils on the sonographic examination performed during the second trimester was approximately half the
distance between the coils at delivery. Table 1
shows the mean, 10th, and 90th percentile values
for the UCI and distance between the coils
obtained during the sonographic examination
(aUCI) and after delivery (true UCI).
The distribution of the hypocoiled and hypercoiled umbilical cords on sonography followed a
Figure 2. Antenatal UCI plotted against true UCI obtained at
birth, with superimposed lines representing mean (50th), 10th,
and 90th percentiles.

187

Assessment of Umbilical Cord Coiling

Table 1. Umbilical Coiling Index and Paired Distance Between the


Coils Measured at the Fetal Anatomic Sonographic Examination
and After Delivery
Parameter

90th percentile
Mean
10th percentile

Sonographic Examination
aUCI
Distance, cm

0.592
0.402
0.21

1.67
2.49
4.80

UCI

Delivery
Distance, cm

0.291
0.203
0.115

3.44
4.93
8.69

Distance indicates the distance between pairs of coils.

gaussian pattern, with approximately similar


proportions of abnormally coiled cords at both
tails of the distribution (Figure 3). A slightly larger number of umbilical cords, but not statistically significant, had been labeled as hypocoiled by
sonographic assessment (12.3%) than were truly
obtained at birth (10.6%), whereas similar results
were obtained for hypercoiled cords regardless of
the mode of assessment (8.9% by sonography
and 8.1% at birth). Hypocoiled umbilical cords
showed a typical pattern in which the vein was
almost straight or with minimal coiling, whereas
the arteries twisted around the vein at the large
distance (Figure 4). In contrast, hypercoiling was
distinctive for the vein and arteries twisting as a
single unit (Figure 5).
Although most of the cords had a coiling pattern
in which the arteries twisted over the umbilical
vein, in a very small proportion of umbilical cords,
an unusual twisting pattern was observed. In 4.2%
of cases (10 cords), substantial coiling of the vein
around the straight or minimally coiled arteries
was observed (Figure 6). In 6 of 10 umbilical cords,
coiling was excessive and was labeled as a hypercoiled cord with mean aUCI and UCI values of
0.60 and 0.26, respectively (range, 0.270.93 for the
aUCI and 0.170.35 for the UCI).

The intraobserver reproducibility of aUCI


sonographic measurements appeared to be
highly reliable, with an average measure intraclass correlation of 0.9948 (95% confidence
interval, 0.97550.9907; reliability coefficient,
.9953). Regardless of reliable aUCI reproducibility, when the abnormally coiled umbilical cords
(defined by 10th and 90th percentile cutoff values) were stratified according to the coiling
assessed at birth, the accuracy of sonography for
predicting hypocoiled or hypercoiled cords was
less compelling (Table 2). The sensitivity values
of sonography for detecting hypocoiling and
hypercoiling at birth were 78.9% (15 of 19) and
25.4%, respectively.

Discussion
The umbilical cord is the major fetomaternal
unit that provides communication between the
placenta and the fetus. However, it is a part of the
fetal anatomy and may be prone to compression,
tension, or torsion, with subsequent interruption
of blood flow. It is thought that coiling provides a
protective effect to these forces, therefore securing an interrupted blood supply to the fetus. The
true etiology of umbilical coiling is unclear, but it
is thought to result from fetal movement as well
as unequal vascular growth.10,11 It appears that
the umbilical coiling pattern is established in the
first trimester, although the presence of a mixed
coiling pattern and even reversal of the coiling
direction in third trimester have been shown.9
We observed a significant correlation between
the aUCI and true UCI at term. The provided
Figure 4. Sonogram of a hypocoiled umbilical cord with a
straight vein and sparsely coiled arteries.

Figure 3. Distribution of the normal and abnormal coiling patterns obtained by the sonographic evaluation and after delivery.

188

J Ultrasound Med 2005; 24:185191

Predanic et al

equation, or roughly halving the aUCI value,


could be used to predict the true UCI at birth. If
translated into the distance between coils, the
distance obtained by sonography at approximately 20 weeks gestation would be half the distance at term. This observation is likely related to
the length of the cord and numbers of coils. If a
certain number of coils is present in the 50- to
60-cm-long umbilical cord at a term birth, it is
thought that the same number of coils would be
present at 20 weeks gestation, when a cord
length would likely be half (30 cm) the cord
length at term. This simple comparison suggests
a rough correlation of the aUCI versus UCI with
a ratio 2:1. These findings are not in agreement

with previously published mean aUCI values of


0.62 0.2 found at 13 to 28 weeks gestational
age.9 A discrepancy in aUCI mean values
between our study and a study by Qin et al9 could
be 2-fold. First, different gestational ages at
which the sonographic examination was performed and, second, a difference of how calipers
were placed to measure the distance between
the pair of coils could explain this discrepancy. If
the cord length is dependant on the gestational
age and independent of the number of coils, an
earlier gestational age will be associated with
shorter cords but the same number of coils compared with a more advanced gestation with a
longer cord and, therefore, lower aUCI values.
Indeed, in our study, the mean gestation was 20.4
weeks (range, 1823 weeks), whereas Qin et al9
reported mean gestation of 18.8 weeks (range,
1328 weeks). In addition, we think that the distance between the pair of coils has to be measured in the same manner; calipers should be
placed along the ipsilateral side of the cord without crossing or measuring the distance in the
middle of the cord (Figure 7). Even more, if the
cord is substantially curved, we recommend
tracing a distance along the outer edge of the
cord, similar to the measurement of cervical
length in curved cervices. This would provide a
means to standardize the antenatal sonographic
measurement of the umbilical cord coiling.
In consideration of the abnormal versus normal coiling distribution in our study, we
observed UCI 10th and 90th percentiles that
were in agreement with previously published

Figure 6. Venous type of umbilical cord coiling in which the vein


coils around the arteries at a higher rate, whereas the arteries
appear as straight vessels.

Figure 7. Significantly different measurements, representing a


distance between 2 pairs of coils, obtained when calipers were
placed at different levels.

Figure 5. Sonogram of a hypercoiled cord in which the umbilical vein and adjacent arteries are twisted as a single unit rather
than arteries twisting at a higher rate around the umbilical vein,
such as in Figure 1.

J Ultrasound Med 2005; 24:185191

189

Assessment of Umbilical Cord Coiling

Table 2. Tabulation of Umbilical Coiling Assessed by Sonography


and Coiling at Birth to Assess the Sonographic Accuracy for
Predicting True Coiling
aUCI

UCI at Birth
Hypocoiled Normally Coiled Hypercoiled

>90th percentile
10th90th percentiles
<10th percentile
Total

0
4
15
19

15
163
14
192

6
19
0
25

Total

21
186
29
236

Cutoff values for 10th and 90th percentiles are presented in Table 1.

results by Strong at al.12 This agreement confirms


that our sample population was appropriate;
therefore, sonographic evaluation of coiling in
our patients should represent true secondtrimester coiling. In comparison with the true
UCI distribution and percentage of hypocoiled
and hypercoiled cords, antenatal sonographic
coiling assessment in the second trimester
showed similar percentages of abnormal coiling,
although a larger number of hypocoiled cords
were noted on sonography (12.3% versus 10.6%
at birth). This could be explained by an error of
sonography in overestimation of the distance
between the umbilical cords, by the fact that
concordant coiling is present in 70% of cords
only,13 or by a sampling error in sonographic
measurements.
As we mentioned before, theoretically, sonography should show true coiling in the second
trimester because of the optimal umbilical cord
visualization and the fact that the true UCI in
our study correlated to previous published
results.1,9,12 Regardless of the significant correlation between the aUCI and UCI, it appears that
sonography in the second trimester is not as
accurate in predicting hypocoiled or hypercoiled cords at birth as expected. Sensitivity values of sonography for detecting hypercoiling
and hypocoiling of 17.3% and 9.1% were reported previously.9 This difference was attributed to
the development of mixed coiling patterns during the second and third trimesters in approximately 25% of cases that were not recognized on
an antenatal sonographic examination. We
obtained somewhat better results, with sensitivity of 78.9% and 25.4% for predicting hypocoiled
and hypercoiled umbilical cords, respectively.
According to our results, a sonographic evaluation is more sensitive for detecting true
hypocoiled rather than hypercoiled umbilical
190

cords. It is also possible that the error in measurement is more common if hypercoiling is present. In addition, poor sensitivity of sonography
for accurately describing cord coiling could be
found in differences between coiling patterns
at the different cord segments. Substantially
increased coiling was found at the fetal end of the
umbilical cord when compared with the middle
segment (free loop) and placental end.13 At the
same time, concordant hypocoiling and hypercoiling were found in only 4.4% and 6.3% of cases,
respectively, when the fetal versus placental cord
ends were compared. Nevertheless, the mean
sonographic UCI obtained in the middle of the
cord (free loop) would be the approximate arithmetic mean of the UCI at the fetal and placental
ends.9
In addition, we observed small proportions of
umbilical cords (4.2%) with twisting patterns in
which substantial coiling of the vein around the
straight or minimally coiled arteries was noted.
These cords were mainly labeled as hypercoiled
(6 of 10) and were likely a consequence of different growth rates between the vein and umbilical
arteries (the presumed case of false knots).
In conclusion, prenatal sonographic evaluation
of umbilical cord coiling during the secondtrimester fetal anatomic survey is accurate
enough to predict true coiling at birth if the aUCI
is obtained in an adequate and standardized
manner. The distance between the 2 pairs of coils
should be measured by placement of both
calipers along the ipsilateral side of the longitudinal umbilical cord image. The true accuracy of
the abnormal coiling determined by sonography
in the second trimester of gestation and its relationship with adverse pregnancy outcome are to
be determined in future studies.

References
1.

Machin GA, Ackerman J, Gilbert-Barness E.


Abnormal umbilical cord coiling is associated with
adverse perinatal outcomes. Pediatr Dev Pathol
2000; 3:462471.

2.

Ezimokhai M, Rizk DE, Thomas L. Abnormal vascular


coiling of the umbilical cord in gestational diabetes
mellitus. Arch Physiol Biochem 2001; 109:209214.

3.

Ercal T, Lacin S, Altunyurt S, Saygili U, Cinar O,


Mumcu A. Umbilical coiling index: is it a marker for
the foetus at risk? Br J Clin Pract 1996; 50:254256.

J Ultrasound Med 2005; 24:185191

Predanic et al

4.

Strong TH Jr, Jarles DL, Vega JS, Feldman DB. The


umbilical coiling index. Am J Obstet Gynecol 1994;
170:2932.

5.

Rana J, Ebert GA, Kappy KA. Adverse perinatal outcome in patients with an abnormal umbilical coiling
index. Obstet Gynecol 1995; 85:573577.

6.

Ezimokhai M, Rizk DE, Thomas L. Maternal risk factors for abnormal vascular coiling of the umbilical
cord. Am J Perinatol 2000; 17:441445.

7.

Degani S, Lewingsky RM, Berger H, Spiegel D.


Sonographic estimation of umbilical coiling index
and correlation with Doppler flow characteristics.
Obstet Gynecol 1995; 86:990993.

8.

Edmonds HW, Washington DC. The spiral twist of


the normal umbilical cord in twins and in singletons.
Am J Obstet Gynecol 1954; 67:102120.

9.

Qin Y, Lau TK, Rogers MS. Second-trimester ultrasonographic assessment of the umbilical coiling
index. Ultrasound Obstet Gynecol 2002; 20:458
463.

10. Finberg HJ. Avoiding ambiguity in the sonographic


determination of the direction of umbilical cord
twist. J Ultrasound Med 1992; 11:185189.
11. Chaurasia BD, Agarwal BM. Helical structure of the
human umbilical cord. Acta Anat (Basel) 1979;
103:226230.
12. Strong TH Jr, Elliott JP, Radin TG. Non-coiled umbilical blood vessels: a new marker for the fetus at risk.
Obstet Gynecol 1993; 81:409411.
13. Blickstein I, Varon Y, Varon E. Implications of differences in coiling indices at different segments of the
umbilical cord. Gynecol Obstet Invest 2001;
52:203206.

J Ultrasound Med 2005; 24:185191

191

You might also like