Early diagnosis of fetal cardiac anomalies by transvaginal echocardiography in a high risk population

Free Communications

F C 0 1 : C A R D I A C 1

F01The accuracy and indications of fetal echocardiographyin high risk patients: a 4-year clinical experience in anobstetric unit

O. Gomez, B. Puerto, A. Borrell, M. Palacio & V. Cararach

Objective: To evaluate the accuracy and assess the indications of fetalechocardiography to diagnose structural heart disease in our high riskpopulation, by an obstetric team.Methodology: Over a 4-year period, a transabdominal echocardiogra-phy was performed in 912 high risk pregnancies for congenital heartdisease at 18–38 weeks of gestation. In the last 140 cases, an earlyexamination combining either the transvaginal or the transabdominalroute at 13–16 weeks was also carried out. Follow up was obtainedfrom neonatal examinations and autopsies. The main indications forreferral were: pregestational diabetes mellitus, family risk, increasednuchal translucency, suspected cardiac anomaly at screening ultra-sound, women at high risk of chromosomal abnormality reluctant toinvasive test, suspected arrhythmia, single umbilical artery, exposureto teratogens, and pregnancies affected by a chromosomal abnormality.Results: There were 79 (8.6%) major congenital heart defects, andmost of them were yielded in the group of screening by ultrasound.Seventy of them (88.6%) were diagnosed correctly as abnormal prena-tally. Atrioventricular septal defects (21 cases) and hypoplastic left ven-tricle syndrome (12 cases) were the most prevalent heart anomalies inthe fetus. Isolated septal defects and coarctation of the aorta were themost difficult lesions to detect, being the ones with false positive andnegative diagnosis. Another four cases of non-significant (<3 mm, nosurgery nor medical treatment required) ventricular septal defects wereoverlooked during gestation, but closed by the sixth month of life.Fifteen apparently normal cases were lost to follow up.Conclusion: Our obstetric team results, in accordance with others pre-viously published, show a good effectiveness to diagnose congenitalheart anomalies. Most major detectable cardiac anomalies occur inthe group without previously-known risk.

F02Correlation between fetal cardiac diagnosis by obstetric andpediatric cardiologist sonographers and comparison withpostnatal findings

M. Meyer-Wittkopf, B. Schonfeld, S. G. Cooper & G. F. Sholler

Fetal echocardiography is optimized by a team strategy of imaging byobstetricians/high risk obstetric scanners (O) and fetal pediatric cardi-ologists (FPC). We retrospectively examined 1037 studies (1995–1999), and identified 249 cases of major congenital heart disease.The O diagnosis was compared to the FPC diagnosis and postnataldiagnosis. The rate of complete accurate diagnosis for O and FPC diag-nosis were 59% (17% false positive, 41% negative), respectively.Major differences in diagnosis or detail were found in 79 patients afterFPC was completed and in 35/79 (44%) this was judged to have poten-tial significant impact on management and prognosis counseling. Thecomplementary roles of O and FPC remain important. FPC can contri-bute with additional detail in some cases which may significantlyimpact on counselling and planning.

F03Factors influencing the prenatal detection of structuralcongenital heart diseases

S.-F. Wong, F.-Y. Chan, R. Cincotta, A. Lee-Tannock & C. Ward

Background: Congenital heart disease is the commonest congenitalanomaly, but the reported antenatal detection rate varies widely from4.5 to 36%.

Purpose: To assess the factors influencing the antenatal detection rateof structural congenital heart diseases.Methods: A retrospective study was conducted at a major obstetrichospital in Australia between 1st January 1996 and 30th June 1999.The medical records of all fetuses and infants born with congenitalheart diseases, except isolated patent ductus arteriosus, were reviewed.Only pregnancies that had antenatal ultrasound scan assessments formorphological surveys were included. The antenatal detection ratefor structural congenital heart diseases was determined. The followingfactors that might influence the detection rate, were assessed: complex-ity of the lesions; experience of the sonologists (performance in tertiaryvs. non-tertiary institutions); identification of other structural or chro-mosomal anomalies; and maternal body mass index.Results: The incidence of structural congenital heart disease in this ser-ies, excluding cases referred from other hospitals, was 7% (179/25529). One-hundred and fifty-three fetuses and babies with structuralcongenital cardiac diseases were included. The overall detection ratefor congenital heart diseases in this series was 39.9%. Detection ratesfor isolated septal defects were poor (13.2%). The detection rates weresignificantly higher for complex lesions (54%), for lesions with conco-mitant septal defects (47.6%), and for lesions with abnormal fourchamber views (62.9%). The detection rate was also higher if the scanwas performed in the tertiary institution (59%), and if there were otherchromosomal or structural anomalies. Maternal body mass index didnot affect the detection rate in the current series. Stepwise logisticregression analysis showed that three independent variables affectingthe detection rate were complexity of the cardiac lesion, experienceof the operator (tertiary scan), and the detection of other chromosomalanomalies.Conclusion: The public needs to be educated on the limitations ofantenatal screening for congenital heart anomalies, especially on iso-lated septal defects. To improve future detection rates, further empha-sis on the training and supervision of the sonologists needs to bestressed.

F04Role of routine fetal echocardiography in clinical practice:its impact on prenatal diagnosis of congenital heart diseasein low risk population

A. Chinnaiya & N. B. Mohammed

Purpose: Congenital heart disease occurs in approximately 8 of 1000live births. Its prenatal diagnosis is important for proper perinataland neonatal management. Basing referral for fetal echocardiographyon presence of risk factors, excludes about 85% of fetuses withsevere detectable heart defects from screening. The present studyaims to examine the non-selected low risk population to determinethe diagnostic accuracy of fetal echocardiography in detecting cardiacanomalies.Methods: We included 39 808 low risk pregnant women who receivedantenatal care at the National University Hospital, Singapore betweenJanuary 1986 and December 1994 and underwent routine second tri-mester ultrasound imaging to exclude fetal anomalies at the AntenatalDiagnostic Center. Ultrasonographers trained to do congenitalabnormality scan including echocardiography (level IV scanning), per-formed examinations that was checked by obstetricians speciallytrained in obstetric ultrasonography. The duration of each scan wasapproximately 30 min. Equipment used were Hitachi 40,400,565A,Alokha 650 and Acuson 128XP. All subjects were followed to delivery.Postnatally, the neonates were examined by a neonatalogist for theconfirmation of presence or absence of congenital anomalies. Patientswho elected to terminate their pregnancies had their fetuses sent forpostmortem assessment.

Ultrasound Obstet Gynecol 2001; 18 (Suppl. 1): 11–31

11

Results: A total of 294 cases of congenital heart defects were identifiedprenatally by fetal echocardiography. Atrial and ventricular septaldefects were the commonest anomalies identified (154), followed byleft hypoplastic heart (23). In 36 subjects, abnormal cardiac sono-graphic findings were not confirmed postnatally on examination. Onthe other hand, there were 44 cases of cardiovascular anomalies thatwere missed by ultrasound prenatally. We obtained sensitivity of85.4% for detection of congenital heart disease and a specificity of99.9% to rule out any such anomalies. Our positive predictive ratewas 87.7% and a negative predictive rate was 99.9%.Conclusion: We recommend routine screening by fetal echocardiogra-phy of all pregnancies for prenatal detection of cardiac anomalies, toincrease the sensitivity and cost-effectiveness of antenatal obstetricultrasound screening and hence, to improve the perinatal managementof cardiac anomalies.

F05Prenatal diagnosis of congenital heart disease in a lowrisk obstetrics population in Singapore

G. S. H. Yeo, B. H. Woo & Y. L. Yan

Objective: To compile information of the type of Congenital HeartDisease (CHD), the concomitant presence of extra-cardiac structuraldefect and chromosomal abnormality in our center to assist patientcounseling. The true positive rate (TPR) and positive predictive value(PPV) of the prenatal screening of various CHD is calculated.Methods: A single center, prospective, cross-sectional observationalstudy of screening scans on low risk patients. A search of all the caseswith CHDs from June 1996 to May 1999 yield a total of 226 cases ofconfirmed CHDs identified by postmortem examinations, surgery orneonatal 2D echocardiography.Results: The commonest defect is ventricular septal defect (VSD) and/or atrial septal defect (ASD) (75 cases), followed by Fallot’s tetralogy/double outlet right ventricle (DORV) (31), pulmonary stenosis/atresia(14), hypoplastic left heart (12), tubular hypoplasia (9), atrio–ventricu-lar septal defect (AVSD) (8) and Ebstein anomaly (8).

The TPR differ for individual defect, ranging from 100% for hypo-plastic left heart, hypoplastic right heart and Ebstein anomaly to0% for truncus arteriosus and tubular hypoplasia. The PPV vary forthe individual defects, ranging from 100% for Ebstein anomaly, coarc-tation of aorta, pulmonary stenosis, atrial septal defect and singleventricle to 0% for truncus arteriosus.

The presence of a prenatally diagnosed CHD is associated with anincreased risk of chromosomal abnormalities from 85.7% for AVSDand 57.1% for coarctation of aorta to 14.3% for Ebstein anomaly.Extra-cardiacdefects are common in fetuses with chromosomal disorder.Conclusion: CHD is the commonest defect in the fetus and should beroutinely screened for. Availability of de facto information in a diag-nostic center is essential to help the mothers to come to an informeddecision. Karyotyping should be routinely offered. Extra-cardiaclesions are common associations and should be carefully looked for.An acceptable positive predictive value is particularly important inpost-screening counseling.

F06Should the three vessels view be the part of routinescanning for congenital heart defects?

E. V. Ioudina & M. V. Medvedev

Objective: To estimate the effectiveness of the three vessels view inprenatal diagnosis of congenital heart defects (CHD).

Methods: A total of 44 904 pregnant women of low- and high risk forCHD were examined from 16 to 39 weeks and followed throughouttheir gestation between January 1994 and December 2000.Results: CHD were diagnosed in 249 cases, 169 (67.9%) were classi-fied as major cardiac anomalies. The overall prevalence of CHD was5.5%. Out of 169 cases 130 (76.9%) were detected prenatally includ-ing 46 isolated CHD. The detection rate of isolated CHD was 63.9%(46/72). From 1994 to 1996 four chamber view was the only one obli-gatory plan of heart examination. For the next three years, we used fourchamber view in combination with identification of outflow tracts. In1999 the three vessels view became the standard test in screening pro-gram for CHD. The detection rate of CHD increased from 54.2% overthe first period of time to 80.4% in 1997–98 and to 87.9% at the end ofthe study. In 2000, this index run up to 89.4%. The part of missedmain vessels malformations decreased from 39% in 1994 to 20% in2000.Conclusion: The three vessels view is an easily obtained scan of thefetal heart. This plane can simplify prenatal cardiac examination andappreciably increase the detection rate of CHD. The three vessels viewshould be an obligatory part of routine scanning for CHD.

F07Early diagnosis of fetal cardiac anomalies bytransvaginal echocardiography in a high riskpopulation

C. Comas, C. Mortera, M. Torrents & J. M. Carrera

Objective: To examine the accuracy of early transvaginal fetalechocardiography (TV-echo) performed in a high risk population.Methods: A series of 70 high risk pregnancies were screened by TV-echo at 12–17 weeks gestation between March 2000 and May 2001.A total of 74 fetal heart examinations were performed, including fourtwin pregnancies. For each fetus, visualization of the four chamberview, the origin of the great arteries, aortic and arterial ductal archeswas attempted in a segmental approach. B-mode and color/pulsedDoppler flow imaging were used in all cases, while M-mode was occa-sionally performed. The average duration of complete heart examina-tion was approximately 30 min. All the examinations were performedby the same experienced operator. Reliability was assessed by conven-tional transabdominal echocardiography at 20–22 weeks, by postnatalfollow up in the first three months of life, and/or by autopsy in cases oftermination of pregnancy.Results: The rate of successful visualization of the fetal heart was90.5% (67/74). In 12 out of 74 fetuses the final diagnosis was abnor-mal (16.2%), including 10 structural heart abnormalities and twocases of heart dysfunction. In 8 out of 10 cases with structural heartdefects, the diagnosis was suspected at early TV-echo. In three cases,although the main diagnosis was correct, details of cardiac defects weremissed at early TV-echo. Among the heart defect group, five cases hadan abnormal karyotype (41.7%) and nine cases showed extra-cardiacanomalies (75%). There were two false negative cases at early TV echo(two cases of hypoplastic left ventricle). There were no false positivediagnoses. In our series, the sensitivity to detect congenital heartdefects was 83% by early TV-echo and 80% by conventional trans-abdominal echocardiography.Conclusions: Early high-frequency TV-echo can provide a comprehen-sive assessment of the fetal heart. The high rate of successful visualiza-tion of the fetal heart provides a reliable diagnosis of major cardiacdefects at this early stage of pregnancy. However, at present thisexamination should be restricted to high risk pregnancies.

11th World Congress on Ultrasound in Obstetrics and Gynecology FC01: Cardiac 1

12

F C 0 2 : D O P P L E R

F08Regional variation in power Doppler perfusionmeasurements within normal placenta

A. W. Welsh & N. M. Fisk

Background: Power Doppler ultrasound is well suited to the evalua-tion of placental perfusion, with its sensitivity to low flow, lack of alias-ing and relative angle independence. Previously published work fromour group has proposed the use of power Doppler evaluation of abnor-mal placentation to monitor progress after significant placental abrup-tion. Commercially available software (CQ Analysis, Kinetic ImagingLtd., UK) is able to segment color flow information from compoundcolor/greyscale images and express flow information in relation to ves-sels (mean color velocity or energy) or tissue (percentage integratedvelocity or energy). Care must be taken to either standardize powerDoppler values to local regions of 100% blood flow or alternativelyto compare regions at the same depth and machine settings sat the sameexamination.Methods: Ten patients with uncomplicated pregnancies wererecruited for additional power Doppler ultrasound examination. Allwere in the third trimester with anterior placentas. The placental loca-tion was initially mapped, and then separate regions were chosen,approximating to different placental quadrants. With constantmachine settings for each case, and depth being constant, multiple clipswere stored from each region. After image segmentation, all frameswere examined individually to exclude any with movement or flashartefact. The percentage of ICE was measured in each frame: firstly,carefully tracing the fetal placental vessels away from the surface andsecondly, using an oval shaped region of interest as defined by CQAnalysis. From these two regions, two different examinations wereperformed, averaging over the whole clip to remove cardiac cycleeffects, and looking manually at systolic frames only.Results: A total of 8156 frames were analyzed, of these 872 selected assystolic. The variation in regional perfusion assessed with power Dop-pler from normal placenta ranged from 19 to 132%. There were nostatistically significant differences in reproducibility for the techniquesfor region selection and for region selection.Conclusions: In this study, an attempt was made to show the highdegree of variation in normal perfusion within the human placentaas a function of space. Unlike other organs where power Doppler per-fusion evaluation may be linked to specific planes or landmarks, this isnot technically possible with the placenta. An exception to this may bewithin areas of pathological placentation that are followed prospec-tively. However, caution should be exercised in the use of powerDoppler for evaluation of normal human placenta perfusion.Acknowledgements: Alec Welsh was supported by a Research Train-ing Fellowship from RCOG/WellBeing, the BUPA Research Founda-tion Clinical Scientist Award 2000 and the 2001 Bernhard Baron/RCOG Travelling Scholarship.

F09Analysis of the hemodynamics of the renal artery of normalfetuses by the color Doppler method

T. Iura, N. Higashiyama, S. Miyazaki, K. Ooshima & S. Makinoda

Objectives: Urination interval or the pulsatility index (PI) and resis-tance index (RI) of the renal artery have often been measured to assessrenal function of fetuses in cases such as oligoamnion. However, PI andRI do not reflect dynamic changes of the renal artery not only in fetuseswith oligoamnion, etc. but also in normal fetuses. For this reason, wehave analyzed the blood flow of the renal artery in normal fetuses usingnew indexes to clarify its hemodynamics.Study methods: The healthy fetuses of 33 pregnant women partici-pated in this study with informed consent of mothers in their 20–40th week of pregnancy. The blood flow of the renal arteries wasexamined at the renal hilum using a color Doppler apparatus usingthe indexes of PI, RI, maximum velocity (Vmax), mean velocity time(VMT), sectional area of the renal artery (AREA) and renal blood flow

(RBF). The gestational age of the fetuses were divided into five groupsof 20–24, 25–28, 29–32, 33–36, and 37–40 weeks through longitudi-nal study. Statistical analysis was performed using Fisher’s protectedleast significant difference.Results: RI did not show marked changes throughout all gestationalage. In contrast, the Vmax and RBF of the renal artery for the 20–24 weeks group were 21.98 cm/s and 0.049 L/min, respectively(mean), and those for the 25–28 weeks group were significantly higherat 27.68 cm/s and 0.003 L/min, respectively (P < 0.05 and 0.05,respectively). The same applied for VMT and AREA. The 95% confi-dence interval of the Vmax of the right renal artery was 23.2–20.8,28.7–26.6, 33.7–31.4, 34.6–32.8 and 40.7–37.8 for the 20–24,25–28, 29–32, 33–36 and 37–40 weeks group, respectively. Thus, wehave determined the 95% confidence interval for the Vmax, VMT,AREA and RBF of the renal artery of fetuses in different stages ofdevelopment.Conclusions: The present study demonstrated that Vmax, VMT, AREAand RBF were more useful than either PI or RI in assessing the hemo-dynamics of the renal artery of fetuses in different stages of develop-ment. Moreover, we have clarified the normal range of the Vmax,VMT, AREA and RBF for the renal artery of fetuses after the 20thweek of gestation. These results are very useful to assess the renalfunction of fetuses.

F10The anatomy and morphology of the ductus venosusin the human fetus: relevance to clinical practice

E. Mavrides, G. Moscoso, J. Carvalho, S. Campbell& B. Thilaganathan

Background: Doppler of the fetal ductus venosus (DV) is becomingincreasingly commonplace in clinical practice. However, the true ana-tomical relationships of the DV are yet to be clearly defined. Further-more, there is conflicting evidence regarding the structure of the DV,especially with regard to the presence of a DV inlet sphincter. Theaim of the present study is to examine the anatomy and morphologyof the DV.Design: A prospective study on 24 fetuses obtained from medical ter-minations between 13 and 17 weeks’ gestation. Following microdissec-tion of the liver, specimens were prepared for study of the anatomicalrelations, scanning electron microscopy or histology.Results: A venous confluence at the terminal end of the umbilical vein,termed the portal sinus, was identified connected to the right and leftintrahepatic portal veins. The DV, a branchless hourglass shaped ves-sel, originated from the portal sinus ascended steeply in the direction ofthe diaphragm. The endothelial surface of the DV showed longitudinalcorrugations and at the level of the inlet, there was ‘shelf’ separatingthe DV from the portal sinus. Histological studies revealed elastinfibers, six to seven cell layers thick at the DV inlet and a single layerof smooth muscle extending from the shelf to the DV outlet. A largenumber of nerve fibers were present in the adventia of the DV inlet.Conclusions: This study clearly delineates the anatomical relations ofthe fetal ductus venosus. The morphological findings refute the pre-sence of a DV sphincter, but provide an insight into the physiologyof the human fetal DV. In clinical practice, the consistent use of thesuggested appropriate terminology would allow collection of compar-able data between units and enable operators to be confident of whichvessels they are sampling by Doppler ultrasound.

F11Use of fetal umbilical to middle cerebral Dopplervelocimetry ratio in the presence of normal umbilicalDoppler findings as a predictor for adverse perinataloutcome in term fetuses

W. W. K. To & C. K. M. Mok

Background: Doppler velocimetry studies of the fetal umbilical artery(UA) in the antenatal period have been shown to be correlated with

23–28 October 2001, Melbourne, Australia FC02: Doppler

13

significant intrauterine growth restriction, subsequent intrapartumfetal distress requiring intervention, and other major adverse perinataloutcome. The addition of middle cerebral Doppler (MCA) studies andexpression of the two parameters as a ratio has been shown to predictintrauterine growth restriction (IUGR) sensitively. However, an abnor-mal ratio was not shown to be associated with increased risk of majorperinatal morbidity or mortality or subsequent neurological damage.This study aims at evaluating the correlation between the UA/MCADoppler ratio and adverse perinatal outcome in a group of term fetuseswith suspected IUGR when the umbilical Doppler findings were stilllargely within the normal range.Methods: A retrospective analysis of all pregnancies undergoing ultra-sound and Doppler studies for evaluation of suspected IUGR after34 weeks over a period of 6 months in a general obstetric unit thatcatered for an annual delivery of around 3500 deliveries per year.The data from the ultrasound and Doppler evaluation within 1 weekfrom delivery were analyzed. Those with absent or reversed end diasto-lic flow, or a systolic/diastolic (S/D) ratio over 3.5 in UA Doppler stu-dies were excluded from analysis. The S/D ratio of UA and MCADoppler findings were then expressed as a ratio, and these werecorrelated with the perinatal events and immediate neonatal outcomeparameters.Results: A total of 123 babies with complete data were included in theanalysis, of which 10 were from five twin pregnancies. Seventy-two(58.5%) were confirmed to be growth restricted with birth weightbelow the 10th percentile for gestation. Sixteen (13%) had an UA/MCA Doppler ratio under 0.75. The incidence of IUGR was signifi-cantly higher in this group with abnormal Doppler ratio (87.5%) ascompared to the normal ratio group (55%) (P < 0.05). Similarly, theabnormal Doppler ratio group had higher incidence of significant oli-gohydramnios (75% vs. 17%, P < 0.001), induction of labor (50% vs.20%, P < 0.01), total cesarean section (62.5% vs. 15.8%, P < 0.001),and cesarean section for fetal distress (43% vs. 11.2%, P < 0.001)than the normal ratio group. The incidence of admission to NICU,meconium stained liquor or intrapartum cardiotocogram abnormalitiesdid not differ.Conclusion: In the presence of normal UA Doppler findings, earlychanges of cerebral vasodilation leading to a relatively low UA/MCADoppler ratio are associated with a higher incidence of obstetric inter-vention and perinatal morbidity. The MCA Doppler studies should bevaluable in the evaluation of suspected growth restriction even whenUA Doppler findings are largely normal.

F12Quantification of fetal power Doppler: determinationof the intravascular standardization point

A. W. Welsh, J. Rubin, B. Fowlkes & N. M. Fisk

Background: With its sensitivity to low flow, lack of aliasing and rela-tive angle independence, power (amplitude) Doppler ultrasound is wellsuited to assessment of tissue perfusion. However, the influence ofmachine settings, depth and other factors related to attenuation meansthat perfusion measurements taken from deep tissue without standar-dization to 100% flow are relatively meaningless. Previously publishedadult research has proposed a technique to estimate a true value for100% amplitude from a cumulative power distribution across a largelocal vessel. This standardization value allows measurement of frac-tional moving vascular volume as an ultrasound index of perfusion.Methods: Fetal renal and aortic blood flow was evaluated using thepower Doppler mode of the Acuson SequoiaTM. Color flow informa-tion was digitally segmented to a linear 8-bit scale and analyzed usingCQ Analysis software as previously reported by our group. Usingimages from the fetal aorta with and without the kidney, of both seg-mented and unsegmented images, an attempt was made to plot thecumulative power distribution in eight fetuses (27–37 weeks’ gesta-tion). Fetal aortic amplitude information was then examined using acombination of CQ Analysis, Scion Image and Excel software. Peakamplitude was examined between different ultrasound clips in a singlefetus and in 50 frames through the cardiac cycle in 10 fetuses. Spatialvariation in peak aortic amplitude was examined in 50 frames from

four fetuses with varying angle of insonation from 738 to 888. Temporalvariation was examined in 50 frames from a single fetus.Results: In the human fetus, there is no peak vascular value or corre-sponding ‘knee point’ attributable to the effects of rouleaux. In con-trast, the peak aortic value is relatively constant for 100% flow, bothbetween multiple imaging clips and throughout the cardiac cycle in the10 fetuses. Spatial variation for the 100% flow value increased from anaverage coefficient of variation of 6.5% at 798 to 28% at 898.Temporal variation was 8% over 50 frames.Conclusions: The human fetus has a reduced degree of rouleaux for-mation in part due to an altered form of fibrinogen and reduced con-centration of plasma proteins, so the peak amplitude measurement inthe center of large vessels represents true 100% flow. This finding isconfirmed by the lack of cardiac cycle effects on the amplitude profile.The central value of larger blood vessels, such as the fetal aorta, if mea-sured at an appropriate angle of insonation, may provide a relativelyconstant standardization value for vessels or regions at a similar depth.From these, true measurements of fractional moving vascular volumemay be derived.Acknowledgements: Alec Welsh was supported by a Research Train-ing Fellowship from RCOG/WellBeing, the BUPA Research Founda-tion Clinical Scientist Award 2000 and the 2001 Bernhard Baron/RCOG Travelling Scholarship.

F13Preliminary experiences using quantitative color velocityimaging (CVI): a simple means to measure fetal volumeflow?

B. Tutschek, T. Zimmermann, K. Wagner, T. Reihs & W. Henrich

Purpose: Color velocity imaging (CVI) quantifies blood flow usingtime domain ultrasonography and M-mode vessel diameter measure-ments. It has been validated in vitro and correlates well with Dopplerbased velocity measurements. The purpose of this study was to applyan easy to use CVI software algorithm in fetal studies.Methods: A high resolution ultrasound system (ATL HDI 5000, 7–4and 5–2 MHz curved probes) was used. Analysis of ultrasound datawas performed using dedicated software (HDILab, version 1.91;CVI_Q plug-in version 2.0). The content of the cineloop, usually withseveral cardiac cycles, is stored in digital format, concurrent with anongoing exam. Volume flow, peak velocity and vessel diameter(derived from color M-mode) were measured off-line.Results: Data acquisition is quick and simple. The proximal portion ofthe intrahepatic umbilical vein with its straight course offers good sam-pling conditions. The ductus venosus can also readily be measured.Off-line analysis does not require specific hardware and takes less than3 min per vessels. There is significant intraobserver variablity, but gen-erally CVI volumes flow measurements agreed with published dataderived from invasive or Doppler based methods.Conclusions: CVI volume flow measurement using this system is rapidand easy. The software also delineates the functional lumen diameter(based on color flow information) and calculates blood flow velocities,enabling correlation with Doppler measurements. The calculations arebased on the assumption of certain standard conditions (laminar flowand circular cross section, e.g. in adult carotid artery), but restrictionsin fetal vessels (e.g. tortuous vessels of the free umbilical cord withregions of non-laminar flow, fetal and maternal movements) apply.

F14Ultrasound time domain and color M-mode analysis ofuterine blood flow in twin pregnancy: prediction ofuteroplacental complications, placental insufficiencyand intrauterine growth restriction

M. O. Thompson, K. S. Vines & K. Harrington

Objectives: A preliminary assessment of the usefulness of a non-Doppler technique of uterine artery blood flow measurement in twinpregnancy, for the prediction of preeclampsia, utero-placental insuffi-ciency and intrauterine growth restriction. To compare the resultsobtained with the contemporary Doppler method.

11th World Congress on Ultrasound in Obstetrics and Gynecology FC02: Doppler

14

Methods: A prospective study of 32 twin pregnancies presenting con-secutively for a routine second trimester ultrasound scan. Maternaldemographic and physical characteristics were recorded in all cases.Using previously described methods, blood flow studies were then con-ducted using the two different techniques; pulsed color Doppler wasused to measure blood flow velocity waveform resistance indices, whilea time domain correlation technique with color M-mode imaging wasused to measure blood volume flow in both uterine arteries. The pla-cental site and morphology, uterine artery waveform characteristics,and presence or absence of diastolic notches were recorded. The ultra-sound fetal growth measurements were plotted on reference charts toassess individual fetal growth patterns, and the obstetric outcomeswere recorded for all the pregnancies. Statistical analysis was per-formed to assess the predictive value of each method of blood flowmeasurement for the relevant obstetric complications.

Exclusions: One mother with sickle cell trait (Hb A þ S).Results: An abnormal uterine artery waveform (high resistance withpersistence of uterine artery notching) was of limited value in the pre-diction of pre-eclampsia and IUGR in twin pregnancies. There wasclose correlation between a significant reduction in uterine artery bloodvolume flow and early onset intrauterine growth restriction requiringdelivery before 34 weeks gestation. Uterine artery blood volume flowmay be less influenced by the placental site compared with traditionallyused resistance and pulsatility indices.Conclusions: Uterine artery notching appears to be of limited value inthe direct prediction of growth complications in twins. The volumeflow method correctly identified all the pregnancies complicated byearly onset intrauterine growth restriction. Both methods may wellprove to be complementary in the antenatal management of twin preg-nancy although further studies will be required.

F C 0 3 : A N E U P L O I D Y

F15Wide iliac angle and short limbs. The best marker forTrisomy 21?

G. Pritchard

Aim:

� To create a standard protocol for measuring the posterior angle ofthe iliac crests.

� To construct a score based upon the iliac angle and fetal proximallong bone lengths.

� To compare results between fetuses affected by Trisomy 21 andthose unaffected.

Methods: Fetal morphology and biometry including biparietal dia-meter, humeral length and femur length were assessed between 15and 23 weeks gestation according to standard protocol. A symmetricalimage of the transverse plane of the iliac wings demonstrating thewidest angle and the longest length of the iliac bones was recorded.This view was best obtained with the ultrasound beam perpendicularto the long axis of the fetal spine and perpendicular to the skin line,imaged from posteriorly. It can also be obtained from an anterioraspect. The iliac angle was measured from the internal angle at theintersection of tangential lines drawn along the posterior iliac crests.Results: A total of 2998 fetuses, 30 with Trisomy 21 and 2968 unaf-fected fetuses (902 had amniocentesis) were assessed. There was a sig-nificant difference (16.788 (95% CI 16.694; 16.846)) between themean iliac angle for Trisomy 21 fetuses (94.33 � 7.28) and unaffectedfetuses (77.55 � 8.868). The total of 908 was selected as cut-off. The 22of 30 (83%) Trisomy 21 fetuses and 228 of 2968 (7.7%) unaffectedhad iliac angle greater than or equal to 908 (sensitivity 83%, specificity92.3%). A previous study in our unit had shown that short proximallong bones occurred in 41 of 71 (58%) Trisomy 21 fetuses and 609 of9022 (6.7%) of a low risk unaffected group (unpublished data). Iliacangle greater than or equal to 908 scored one. Short proximal longbones also scored one. A positive index was a score of 2. The 14 of30 (46.7%) Trisomy 21 fetuses and 29 of 902 (3.2%) known un-affected fetuses had a score of 2. The likelihood ratio for positivelimb–iliac index is 14 (46.7/3.2).Conclusions: In screening for Trisomy 21, the prevalence and likeli-hood ratio of fetal anomaly (29%, 96), nuchal skin fold (39%, 11),echogenic bowel (20%, 2), renal pelvic dilatation (20%, 4.2), echo-genic intracardiac focus (35%, 2.6) and short long bones (58%, 1.9)are known for our population. A positive iliac angle/limb length scorehas a prevalence of 47% and likelihood ratio of 14. In our population,this combination surpasses the performance of other markers.

F16Early antenatal diagnosis of cardiac failure usingDoppler ultrasound of umbilical artery

C. G. V. Murta, A. F. Moron, M. A. P. Avila & F. A. P. Vasques

Background: During the last 2 years, few prenatal ultrasonographicstudies have documented a strong association between abnormal fetal

venous return and chromosomal abnormalities. Articles in the Englishlanguage medical literature were identified through MEDLINE. Thereis only one report of reverse flow pattern in the umbilical vein in severeintrauterine growth restriction at 29 weeks’ gestation. The present caseis the earliest documentation of this alteration in fetal venous return.

A 30-year-old primigravida was referred to our ultrasonographyclinic at 12 weeks of gestation. The fetal crown-rump length(57 mm) was not compatible with menstrual dates. The ultrasoundexamination showed an increased nuchal translucency (9.1 mm), bilat-eral pyelectasis, hiperechoic bowel, echogenic intracardic focus andventricular septal cardiac defect. The fetal heart rate was 147 beats/min. The Doppler examination revealed high retrograde flow in theinferior vena cava, reverse flow in the ductus venosus during atrial con-traction and reverse flow in the umbilical vein. Cytogenetic analysis bychorionic villus sampling diagnosed trisomy 9. Ten days later, fetaldemise was observed. A non-mosaic trisomy of chromosome 9 wasconfirmed on fetal cells’ analysis. At autopsy, the characteristics find-ings were of complete trisomy 9 and the multiple abnormalitiesdetected by ultrasonography were confirmed.Conclusion: Doppler blood evaluation of the venous return to theheart might provide relevant information on cardiac function. Thereverse blood flow in umbilical vein may indicate an impending fetaldeath. We speculate that the association of abnormal nuchal translu-cency thickness with alterations of the fetus venous return might be acomplementary tool to assess the risk for fetal aneuploidy in the firsttrimester of pregnancy. In addition, these Doppler velocimetry altera-tions can be an alarm sign for investigating cardiac abnormalities.

F17Ductus venosus Doppler velocimetry for the detectionof Down syndrome: new criterion of risk

A. F. Moron & M. A. P. Avila

Objective: To test the hypothesis the application of ductus venosusDoppler velocimetry may serve as a screening tool between 10 and14 weeks gestation for the detection of fetuses with Down syndromeand estimate new criterion of risk.Patients and methods: A total of 491 fetuses were studied consecu-tively. In 132 cases, a cytogenetic study was performed on materialobtained from a biopsy of the chorionic villus, and in 359 cases thepostnatal pheno-type was used as a basis for the result. In additionto the routine ultrasonographic examination, all the fetuses were sub-mitted to measurement of the nuchal translucency thickness. For statis-tical analysis the Student’s t-test and ANOVA were used. The sensibility,specificity, positive and negative predictive values, false-positiveprobability and likelihood ratio was calculated.Results: There were 21 cases of Down syndrome. Of these 21 fetuses,ductus venosus blood flow during atrial contraction was either absent(n ¼ 3) or reversed (n ¼ 17) in 95.2%. In the chromosomally normalfetuses (n ¼ 270), only eight had abnormal Doppler profiles in the duc-tus venosus (specificity ¼ 98.2%, positive and negative predictive

23–28 October 2001, Melbourne, Australia FC02: Doppler

15

values ¼ 71.4 and 99.8%, respectively, and positive and negative like-lihood ratio was 56 and 0.1, respectively).Conclusion: Our preliminary results suggest that the presence ofDown syndrome may be strongly suspected when there is reverse orabsent flow in the ductus venosus Doppler velocimetry during atrialcontraction. We speculate the possibility of new criterion to calculatethe new risk of Down syndrome was calculated from the fetal medicinefoundation risk, multiplied by a likelihood rate of Doppler of ductusvenosus, using at least the values of 50 for positive likelihood rateand 0.1 for negative likelihood rate.

F18First trimester ductus venosus Doppler screeningin pregnancies with increased nuchaltranslucency risk

E. Mavrides, S. Sairam, B. Hollis, J. Carvalho& B. Thilaganathan

Objective: To evaluate the role of ductus venosus (DV) Doppler inpregnancies with an increased nuchal translucency (NT) adjusted riskfor the prediction of fetal outcome.Methods: Doppler assessment of the DV was performed in 124 fetuseswith a first trimester NT-adjusted risk of >1:300 for Down syndrome.The pregnancy outcome, including karyo-type and structural abnorm-alities, were ascertained for all pregnancies.Results: The DV Doppler was abnormal in 27 (21.8%) pregnancieswith an increased first trimester NT-adjusted risk. The outcome wasnormal in 69 pregnancies (55.6%). Thirty-nine pregnancies (39.5%)were aneuploid and further six pregnancies had adverse outcomes(three unexplained intrauterine deaths, two structural abnormalitiesand one major cardiac defect). The sensitivity of abnormal DV in thisselected population was 51.3% for aneuploidy and 33.3% for adverseoutcome. An abnormal DV Doppler increases the risk for aneuploidyby 11.8 (95% CI: 3.3–42.2).Conclusions: The finding of an abnormal DV Doppler signal increasesthe risk for aneuploidy 12-fold in a pregnancy with an increased NT-adjusted risk. Although DV Doppler can be used in addition to NTscreening for aneuploidy, the sensitivity of DV Doppler is less thanpreviously reported.

F19Integrated first and second trimester screeningfor Down syndrome

Y. H. Lam, C. P. Lee, S. Y. Sin, R. Tang, H. S. Wong, S. F. Wong,M. H. Y. Tang & H. H. N. Woo

Purposes: It is uncertain whether first trimester nuchal translucency ismore effective than the well established second trimester serum screen-ing for fetal Down syndrome or whether their combination works best.We report our data on a large multicenter non-interventional trial inwhich all subjects underwent both first and second trimester screeningto assess their relative efficacy.Methods: All women who attended the obstetric clinic before15 weeks were recruited. An ultrasound examination was performedat 10–14 weeks to measure the nuchal translucency. The nuchal mea-surements were not acted upon unless the fetus showed gross featuresof hydrops fetalis. All women had serum alpha-fetoprotein and humanchorionic gonadotrophin assay at 15–20 weeks. The Down syndromerisk assigned by serum screening was disclosed and amniocentesis wasoffered if this assigned risk was at or >1:250 or if the women were35 years and older. Efficacy of different combinations of screeningwere compared.Results: Between January 1997 and August 2000, 17590 women wererecruited (19% was 35 years and older). After excluding subjects whomiscarried, defaulted the serum test and other reasons, 16237 pregnan-cies were included in the final analysis. Of these, 35 pregnancies wereaffected by Down syndrome (2.2 cases per 1000 pregnancies). At afalse-positive rate of 5%, the detection rate of Down syndrome bynuchal translucency alone; nuchal translucency þ age; serum humanchorionic gonadotrophin, alpha-fetoprotein þ age and nuchal trans-lucency þ human chorionic gonadotrophin þ alpha-fetoprotein þ age

were 61, 69, 73 and 86%, respectively. Comparison of areas under thereceiver-operating characteristic curves showed statistically significantdifferences between these methods of screening.Conclusions: Integration of nuchal translucency and second trimesterserum alpha-fetoprotein and human chorionic gonadotrophin assayyielded the best screening efficacy for Down syndrome.

F20The affect of image size on nuchal translucencymeasurement: a report from the FaST study

A. G. Edwards, S. F. Mulvey & E. M. Wallace

Purpose: To establish whether image size affects nuchal translucencymeasurement.Methods: A random sample of participants in the FaST study wereselected to have additional measurements performed to determinewhether image size affects the nuchal translucency measurement. TheFaST study is a collaborative research project funded by the Depart-ment of Human Services, Victoria Government, and Monash Univer-sity. This study was established to assess the relative merits of first andsecond trimester screening for Down syndrome and to thereby deter-mine the optimum screening strategy. Nuchal translucency measure-ments were performed using an ATL HDI3000 with C5-2 probewith the fetus in a sagittal section and the amnion excluded. UsingHD zoom, the image size was adjusted so that the fetus occupied 60,100 or 200% (head and neck view) of the monitor. Three measure-ments were taken at each image size, in random order, according tothe methods described by the fetal medicine foundation. The ultra-sound machine operator was blinded from the measurement valuesduring the examination. Statistical analysis utilizing STATA 6.0 wasperformed using analysis of variance after log transformation of thenuchal translucency data to produce a normal distribution.Results: Forty-eight women were scanned with nine nuchal translu-cency measurements performed in each case. All scans were transab-dominal. The CRL ranged from 44 to 80 mm and BPD from 15 to28 mm. The mean NT was 1.5, 1.3, and 1.1 mm at 60, 100 and200% image sizes, respectively. Analysis of variance with Bonferronimultiple-comparison test revealed statistically significant differencesbetween 60 and 200% (P < 0.0001) and 100 and 200% (P ¼ 0.025)images, but not between the 60 and 100% images (P ¼ 0.057).Conclusion: Image size appears to make a significant difference to thenuchal translucency measurement with an inverse relationship betweenthe fetal image size and the nuchal translucency recorded.

F21Quality assessment in prospective nuchal translucencyscreening for Down syndrome in the Copenhagen firsttrimester study

K. Wojdemann, M. Christiansen, K. Sundberg, A.-C. Shalmi& A. Tabor

Purpose: To develop and apply a quality control system in a Downsyndrome (DS) screening study using nuchal translucency (NT) asinterventional marker.Methods: In a prospective DS screening study, the NT was measuredin 9236 pregnancies. For quality assessment two models were intro-duced: (I) image-scoring evaluation of the NT measurements and (II)establishment of the distribution of NT multiple of the medians(MoMs) over time and influence of an intervention.Results: The observer variability in the image-scoring evaluation washigh with a kappa value of 0.48 in the overall validation. A revisedmodel showed better interobserver agreement with a kappa value of0.58; regarding the single criteria the differences were however, stillunsatisfactory, i.e. we found highly significant differences for the cri-teria ‘position of the fetus’ (P ¼ 0.0026) and ‘magnification of theimage’ (P ¼ 0.0001). Regarding the distribution of the NT MoMs,the median NT MoM (in groups of 50 fetuses each) stabilized after ashort learning phase (representing the practical part of the sonogra-pher’s certification). The intergroup standard deviation decreasedfrom initial 0.100–0.060 after the learning phase, and 0.046 after theintervention.

11th World Congress on Ultrasound in Obstetrics and Gynecology FC03: Aneuploidy

16

Conclusions: When well-trained certified examiners perform NTscreening, continuous evaluation of the distribution of the NT MoMseems to be a good method to assess the quality for a center and mayalso be used to identify individual examiners deviating from the mean

performance. Intervention towards the examiners seems to decreasevariation, which should theoretically decrease the false positive rate.The image-scoring methods we introduced cannot be recommendedfor quality control in an NT screening program.

F C 0 3 a : O B S T E T R I C S M I S C E L L A N E O U S

F22The importance of funneling to the level of a cervicalcerclage

J. M. O’Brien, A. L. Hill & J. R. Barton

Purpose: Our purpose was to evaluate the utility of ultrasound surveil-lance after cerclage placement and to propose a rationale for cervicalsonography in this setting.Methods: A retrospective case series was evaluated. Women under-going cervical cerclage by a maternal–fetal medicine specialist regard-less of indication and delivering between January 1999 and April 2001were included. Transvaginal assessment of cervical length and thedegree of cervical funneling after cerclage were compared to preopera-tive values and to outcomes. Funneling to the cerclage was defined asmembranes prolapsing down the endocervical canal until they reachedthe plane of the cerclage.Results: A total of 53 cases were available for analysis. Cervical cerc-lage resulted in a significant increase in cervical length as the meanlength increased from 2.1 � 1.2 cm to 2.9 � 0.8 cm after the proce-dure, P < 0.001, but this measure was not correlated with gestationalage at delivery. Funneling to the level of the cerclage, however, wasassociated with an earlier gestational age at delivery 31.3 � 5.6 weeksvs. 36.8 � 2.8 weeks for those cases without this finding, P < 0.001.This observation was statistically significant irrespective of the indica-tion for the procedure and was noted by 28 weeks’ gestation in allpatients followed. The incidence of premature rupture of the mem-branes was also significantly greater postcerclage in women with des-cent of the membranes to the cerclage (52%) vs. those without thisfinding (3%, P ¼ 0.002).Conclusions: Funneling to the cerclage has predictive value for earlierpreterm delivery in patients who have undergone cervical cerclage.Serial sonography up to 28 weeks’ gestation may be useful in identify-ing patients at higher risk for premature rupture of the membranes.

F23Prediction of premature labor using AIWA’s scoring

H. Yoshitake, Y. Koyama, J. Fukuda & K. Uehira

Purpose: Clinical investigation into the correlation between AIWA’sscore and premature labor.Patients: Eight hundred and twenty-nine (829) mothers delivered sin-gle baby at AIWA Maternity Hospital in 2000. The number of primi-para (P) was 394, while that of multipara (M) was 435. We eventuallyanalyzed 587 deliveries in 829, because the resting 238 deliveries werenot fully followed up before delivery and the resting four were under-taken therapeutic premature termination. These 587 mothers had atleast one time of AIWA’s scoring in both 14–24 weeks of gestationand 25–32 weeks of gestation.Methods: We have been using the AIWA’s score to detect and preventpremature labor for 4 years, which consists of findings of transvaginalultrasonography (cervical length and funneling of internal os of theuterus), and internal examination (looseness of external os of theuterus and consistency of the cervix). The full score is 12, which meansextremely high risk of premature labor. On the other hand, the scorezero means no risk of premature labor. The patients, whose score over8, should go into hospital to have cervical cerclage. and if multiparahas histories of premature deliveries and/or threatened prematurelabor, P-score (1–3) is added to AIWA’s score according to the degreeof the risk. Five hundred and eighty-seven (587) deliveries were dividedin four groups, including (1) over 1 week hospitalization (2) havingcervical cerclage (3) premature deliveries (4) normal range deliveries

(including OPD therapy and <1 week hospitalization). The distribu-tion of AIWA’s score was analyzed separately between nullipara andmultipara in 14–19, 20–24, 25–28, and 29–32 weeks of gestation,respectively. In hospitalized patients, the most previous scores beforehospitalization was analyzed. And, indications for prophylactic cervi-cal cerclage before 27 weeks of gestation were: (1) nullipara and#8807; 8 (AIWA’s score alone), multipara and #8807; 8 (AIWA’sscore þ P-score) (2) few uterine contraction with negative fetal fibro-nectin (3) agreements of patients and her family with informed con-sents. The patients with all three indications had prophylacticcervical cerclages.Conclusions: (1) The rate of prophylactic cervical cerclage was 4.5%(nullipara: 1.6%, multipara: 6.6%) in year 2000. This rate was higherthan those of many other hospitals. The rate of premature labor afterthe prophylactic cervical cerclages was 15%, however, all cases of pre-mature labor occurred after 36 weeks of gestation. Eighty-eight per-cent (88%) of the operation was performed in 23–27 weeks, whileonly 12% was operated before 22 weeks of gestation. Seventy-threepercent of cases, which had prophylactic cervical cerclages, had beenexcellent course during pregnancy without over 1-month hospitaliza-tion and without re-hospitalization. (2) In the hospitalized case with-out cervical cerclage for premature labor, total hospitalized days hadpositive correlation with AIWA’s score (P < 0.05). (3) The rate of pre-mature labor and deliveries were 2.8% in nullipara before 37 weeksand 0.7% before 36 weeks, which declined than those in many otherhospitals reported before. and there was no delivery before 33 weeks ofgestation. (4) The success rates of prediction of premature labor withAIWA’s score and #8807; 8 (P), were 48.6% (PPV) and 92.7% (NPV).Those rates seemed better than those of any other hospital reported inthe past. (5) According to these four evidences, we can use AIWA’sscore, which is reproducible and reliable, in order to predict andprevent premature delivery.

F24Ultrasonographic evaluation of changes in placentallocation with the advance in gestation

M. Luria & Y. Z. Diamant

Purpose: To evaluate the change in location of the placenta duringpregnancy and the actual increase in the distance between the internaloss and lower placental margins in relation to the advance in gestation.Methods: A total of 37 pregnant women were scanned three times dur-ing pregnancy: the first time between 14 and 22 weeks’ gestation, thesecond between 23 and 29 weeks’ gestation and the last between 30and 34 week’s gestation. Examinations were performed by two quali-fied examiners. Placenta and cervix were allocated and distancebetween the lower placental margin and the interval oss was measured.Examinations were either transvaginal or transabdominal, dependingon patients’ cooperation.Results: Gradual increase in the distance between the placental marginand the internal oss was found. Mean value of 50.8 mm at 14–22 weeks increased to 111.6 mm at 30–34 weeks. Significant statisticaldifference was found between first, second and third measurements.Effect of maternal age on these changes was evaluated too and wasfound not to affect this phenomenon. Borderline results were foundregarding the effect of parity.Conclusions: Low-lying placenta in early pregnancy is a known phe-nomenon. Little is known about quantitative change in placental loca-tion with advance in gestation. In the present study, we tried toquantitate the change. We found gradual, consistent and significantincrease in the distance between placental margin and internal oss

23–28 October 2001, Melbourne, Australia FC03: Aneuploidy

17

along gestation. This change increased with increase in parity in spiteof the fact that parity is a risk factor for placenta previa. The presentquantitative evaluation of placental location can serve as a predictor oflow-lying placenta.

F25Behavioral organization of the fetus in early pregnancy

G. R. McCartney & P. G. Hepper

Purpose: The behavior of the human fetus consists of complexdynamic relationships. To date, previous methods of measuring fetalbehavior have failed to capture the complexity of behavior. However,as behavior is a reflection of neural functioning it is only whenquantitative assessment of this complexity has been undertakenthat underlying neural organization can be elucidated. In order toreveal fundamental features of neurobehavioral organization anddevelopment, the present study provided a quantitative descriptionof the complex spatial (sequential) and temporal (synchronousand asynchronous) patterning of behavior in the human fetus in earlypregnancy.Methods: Twenty fetuses from normal singleton pregnancies werescanned at two weekly intervals for a period of 1 h from 9 to 17 weeks’gestation. An ethogram of the timing and sequencing of right and leftarm and leg movements was constructed from each videotaped session.A description of sequential and synchronous relationships betweenlimb movements was determined using stochastic modeling. The tem-poral patterning of activity in arm, leg and all limb movements wasexamined by quantitatively determining bout criterion using survivoranalysis.Results: Fetuses exhibited within and between limb synchronous,sequential and temporal patterning from the appearance of single limbmovements indicating that initial patterning may be genetically hard-wired. Self-organization of behavior was apparent in increasing syn-chronous, sequential and temporal patterning with advancing gesta-tion. These changes coincided with rapid increases in neurulation,synaptogenesis and myelinization in early pregnancy. Different sequen-tial, synchronous and temporal organization was found in arm and legmovements indicating differential neural control of co-ordination inthese limbs. Furthermore, lateralization and gender differences werefound to be important features of behavioral organization.Conclusions: The present study has revealed that quantitative assess-ment can define and measure behavioral complexity and this can elu-cidate fundamental neurobehavioral organization in early pregnancy.

F26Third trimester measurements

S. L. C. Westerway, R. Heard & J. Morris

Third trimester pregnancy ultrasound for assessment of fetal size/growth is an important examination, which should be undertaken with

care. This project looked at inter and intra sonographer measurementreproducibility on women in the final 4 weeks of pregnancy. Measure-ments taken included the biparietal diameter, occipito-frontal dia-meter, head and abdominal circumference, femur and humeruslengths, umbilical artery resistive index, amniotic fluid index and cer-vical length. Images were captured on an ATL 5000 and transferred toan MO disc. Copies of the disc were distributed to 20 MIA practicesthat had access to either ATL 5000 or 3000 systems. The disc imageswere downloaded and all sonographers, including students, measuredthe fetal parameters as prompted, and recorded the results. Statisticalanalysis was performed using SPSS analysis of variance model ICC 2.1,which enabled comparisons for inter/intra sonographer variations.Paired t-test was used to calculate the mean differences seen in repeatmeasurements by the individual sonographers. The results reflectedthe degree of experience of the sonographers, particularly with imagesat difficult angles. The study highlights the importance of standardizedmeasuring protocols and image planes, particularly for patientsundergoing ultrasound examinations at different practices.

F27Tissue characterization with gray level histogram widthand assessment of fetal lung immaturity

K. Maeda, M. Utsu, P. E. Kihaile, M. Serizawa, N. Yamamoto& T. Ito

Purpose: The aims are to clarify the reproducibility of gray level histo-gram width (GLHW) in ultrasound machines and to demonstratetissue character of fetal organs and tissues in B-mode image.Methods: The GLHW is calculated manually and automatically withthe ROM provided by Aloka, and the values are compared among sixAloka, each one of Toshiba, Yokogawa and Kretz 503D wuth use ofRMI phantoms. Clinically, the placenta, fetal brain periventricularechodensity (PVE), fetal lung and liver, those of preterm fetuses withthe refernece to stable microbubbles of amniotic fluid, and meconiumstained amniotic fluid are studied.Results: No influence of clinical gain control, image depth, STC, 1 cm2

or more ROI or 1000 or more pixels is confirmed. GLHW is large inhigh image contrast, but it is easily compensated. Some own values arecorrected and standardized. GLHW is reproducible among machines.Clinically, GLHW is large in grade III placenta, the value is large infetal brain PVE and it is large in meconium stained fluid. Normal fetallung GLHW is low and lower than fetal liver before 30 weeks of preg-nancy, fetal GLHW is 31% or less in the group of five or less stablemicrobubbles with 90% sensitivity and 80% specificity. In gynecology,GLHW is high and coefficient of variation is low in multiple ROIs ofovarian malignancy. GLHW is large in hyperplastic endometrium.Adult organs and tissues are of interest particularly in hepatic diseases.Conclusion: GLHW is reproducible in ultrasound machines, and itis useful in tissue characterization. It is possible to assess fetal lungimmaturity with GLHW.

F C 0 4 : C A R D I A C 2

F28Evaluation of normal and abnormal fetal heart anatomyusing 3D echocardiography

M. Meyer-Wittkopf, J. Vaughan & G. F. Sholler

We applied 3D echocardiography to the diagnosis or exclusion of con-genital heart disease (CHD) using rate gating with Doppler. Gated 3Dvolume data sets enabled diagnostically acceptable visualization of allcardiac structures in 16 of 29 fetuses without CHD and in 7 of 22fetuses with CHD, and significant elements of anatomy in the remain-der. In fetuses without CHD a dynamic 3D perspective of valve mor-phology and ventricular wall motion of the four- and five-chamberview projection was available in 19/29 cases. A high quality 3D recon-struction of the site and spatial orientation of VSD could be obtained in

9/13 patients with CHD. The 2D imaging remained the principal diag-nostic modality in all cases with additonal structural detail obtained by3D imaging in only two fetuses with CHD. The 3D imaging of the fetalheart is feasible and may provide additional information of clinicalvalue in a small number of cases with CHD when compared with 2Dimaging.

F29The diagnostic value of fetal heart biometry in earlypregnancy

U. Germer, C. Berg & U. Gembruch

Purpose: To evaluate the value of early cardiac biometry for thedetection of congenital heart defects (CHD) using our normative data.

11th World Congress on Ultrasound in Obstetrics and Gynecology FC03a: Obstetrics Miscellaneous

18

Methods: Cases were selected from all cases with CHD diagnosedbetween 10 and 17 weeks of gestation between 1999 and 2000.Detailed fetal two-dimensional and color-coded Doppler echocardio-graphy was performed. The transversal heart diameter, both ventricu-lar dimensions, heart area and circumference, thoracic diameter,thoracic circumference and area and the diameters of the pulmonarytrunk and ascending aorta were measured.Results: A total of 31 cases of CHD between 10 and 17 weeks of gesta-tion were diagnosed. Eight fetuses presented with insufficient biometricmeasurements. In the remaining 23 fetuses different complex abnorm-alities were present. In three fetuses with an atrioventricular septaldefect and two with a ventricular septal defect fetal heart biometryshowed normal values. In 18/23 fetuses with CHD fetal heart biometrywas partly abnormal. Nuchal translucency (NT) thickness measure-ments were performed before 14 weeks of gestation and 10/12 fetusespresented with an increased NT. Both fetuses with normal NT showedan abnormal fetal heart biometry. Venous Doppler evaluation was per-formed in 22 cases, 12 fetuses demonstrated an increased pulsatility inthe ductus venosus and 7 with a pulsatile flow in the umbilical veinin addition. In 7/10 fetuses with normal venous Doppler fetal heartbiometry was partly abnormal.Conclusion: Early fetal heart biometry may be helpful in the prenataldiagnosis of some major CHD by demonstration of discordant dia-meters of the fetal heart and great arteries. At 10–14 weeks of gesta-tion, nuchal translucency thickness measurements and biometricmeasurements of the fetal heart seem to be complementary methodsfor the detection of cardiac defects.

F30Three-dimensional quantitative echocardiographicassessment of ventricular volume in healthy human fetuseswith and in fetuses with congenital heart disease

M. Meyer-Wittkopf, S. Schmidt, S. G. Cooper & G. F. Sholler

We examined the feasibility and appearance of 3D fetal echo assess-ment of ventricular volumes in 29 normal fetuses and 22 fetuses withcongenital heart disease (gestation 18–35 weeks). Gated 3D volumesets could be obtained in 51 of 57 cases. Normal right and left ventri-cular volume, stroke volume data, and RV/LV ratios were obtained,together with normal changes with gestation. In patients with congeni-tal heart disease, those with a biventricular heart (index lesion teratol-ogy of Fallot) demonstrated normal combined ventricular volumes andright/left volume ratios for gestation age. Fetuses with single ventricu-lar heart with intact septum (hypoplastic left or right heart) demon-strated a small but significant reduction in combined ventricular sizeand stroke volume for gestation age.

F31Fetal cardiac function in fetuses with intracardiacechogenic foci

S. Degani, Z. Leibovitz, I. Shapiro & G. Ohel

Objective: To evaluate cardiac dimensions and function in euploidfetuses with intracardiac echogenic foci.Study design: Forty-eight fetuses with a single cardiac echogenicfocus situated in the left ventricle, had echocardiography performedat 22–24 weeks of gestation. Fifty normal fetuses at 22–24 weeksgestation served as controls. The 2-D and M-mode directed fetal echo-cardiography were used to exclude cardiac anomalies and measureright and left ventricular free walls and interventricular septal thicknessand ventricular systolic and diastolic dimensions. Cardiac size wasexpressed as a ratio of ventricular wall thickness/biparietal diameter,and function was expressed as ventricular shortening fraction. Dopplerfetal echocardiography measurements included pulmonary and aorticmaximum systolic velocities and time to peak velocities as indices ofventricular systolic function and the ratio between E-wave and A-waveat the level of atrioventricular valves as indices of ventricular diastolicfunction.Results: Early ventricular filling/active atrial filling ratios (E/A ratios)were significantly lower in fetuses with intracardiac echogenic focithan in control fetuses. No significant differences were found in cardiac

dimensions, ventricular shortening fraction and Doppler systolicindices.Conclusion: Euploid fetuses with intracardiac echogenic foci showlow E/A ratio values in mid-trimester echocardiography. This findingmight indicate cardiac diastolic dysfunction.

F32Prenatal diagnosis of cardiosplenic syndromes by fetalechocardiography: a 10-year experience

C. T. Berg, A. A. Geipel, A. A. Baschat, T. Kohl, M. Krapp,U. Germer & U. Gembruch

Purpose: To assess the accuracy of fetal echocardiography in the diag-nosis of cardiosplenic syndromes.Methods: Retrospective survey of 27 fetuses with the pre- or postnataldiagnosis of cardiosplenic syndromes in our database over a period of10 years.Results: In 24 out of 27 fetuses, the diagnosis of cardiosplenic syn-drome was confirmed postpartum. In the remainder, two cases of situsinversus totalis with heart block and multiple cardiac malformations,but no evidence of heterotaxy were diagnosed at autopsy or after birth.In one fetus with situs ambiguus, dextrocardia, bradyarrhythmia andtransposition of great arteries (TGA) no autopsy was performed aftertermination of pregnancy. In 18 fetuses, left atrial isomerism wereprenatally diagnosed and confirmed at autopsy or postpartum. Mainechocardiographic features were azygos continuation of interruptedinferior caval vein (15), complete atrioventricular septal defect(CAVSD) (11), persistent bradyarrhythmia (11), pulmonary stenosis(PS) (9), ventricular septal defect (VSD) (8), fetal hydrops (8) and dex-trocardia (5). Other associated cardiac malformations were l-TGA (2),double outlet right ventricle (DORV) (3), hypoplastic left ventricle (3),and partial/total anomalous venous connection (1/1). Ten pregnancieswere terminated, two fetuses demised in utero, six children survived. Insix fetuses, right atrial isomerism was prenatally diagnosed and con-firmed postpartum. Echocardiographic features were CAVSD (4), dex-trocardia (4), PS (2), l-TGA (1), azygos continuation of interruptedinferior caval vein (1), fetal hydrops (1), DORV (1) and bradyarrhyth-mia with cardiomegaly (1). One pregnancy was terminated, two fetusesdemised within 10 days postpartum, three children survived.Conclusion: A combination of CAVSD, first and early second trime-ster persistent bradyarrhythmia and/or azygos continuation of inter-rupted inferior caval vein, and/or visceral situs ambiguus shouldsuggest specific diagnosis of cardiosplenic syndrome. The correct diag-nosis made in utero allows for maternal counseling and transfer to anappropriate facility prior to delivery for corrective surgery.

F33Fetal congenital heart block

C. Pennell, K. P. Reid & J. E. Dickson

Purpose: To ascertain the incidence and outcome of all cases ofprenatally diagnosed congenital heart block (CHB) referred to ourinstitution during the decade 1990–2000.Methods: Cases were identified from the fetal medicine database per-mitting review of the clinical presentation, ultrasound findings, mater-nal immunological status, subsequent management and pregnancyoutcome.Results: Eleven cases of CHB were identified during the period ofreview (100% ascertained for Western Australia) with a diagnosis atmedian gestation of 20 weeks (16.8–32). Fetal cardiac anatomy wasnormal in 64% of cases. Fetal hydrops was present at diagnosis in27% with a further 33% subsequently developing hydrops. Only18% of fetuses had major structural cardiac anomalies. The mediangestation at delivery was 35 weeks (25–38) and there was a female gen-der bias (7 : 4). The mortality rate was 64% with three pregnancy ter-minations, one second trimester intrauterine fetal death, one neonataldeath and two infant deaths. Maternal immunological abnormalitieswere common with 36% of the women having clinical autoimmuneconditions. Seventy-eight percent of women had SSA and/or SSB anti-bodies. Two of the six infants that survived the neonatal period hadpersistent abnormal autoantibodies during their first year of life. One

23–28 October 2001, Melbourne, Australia FC04: Cardiac 2

19

infant with SSA and SSB died at one year of age from left ventricularfailure believed to be of autoimmune origin. This infant was one oftriplets, with only the female triplet being affected. Another fetuswith SSA and SSB had thrombocytopenia at delivery. Postmortemassessment of five fetuses confirmed focal calcification at the atrioven-tricular junction associated with a lymphocytic infiltrate and subendo-cardial fibrosis. Eight subsequent pregnancies have occurred and nonewere complicated by CHB. Three women underwent prophylacticimmunosuppression.Conclusions: Fetal congenital heart block is a rare condition with astrong association with autoimmune disease and poor outcome. Therecurrence risk in this series is lower than previously reported.

F34Congenital heart disease and the spectrum of karyotypicabnormalities in a prenatal population

A. E. Quinton, J. S. Smoleniec, J. Clarke & R. F. Ogle

Aim: To evaluate the frequency and type of chromosome anomalies infetuses diagnosed by ultrasound with congenital heart disease (CHD).Methods: The study is a retrospective descriptive case series withthe nonrandom sample taken from all the pregnant women who had

an ultrasound at the Feto-Maternal Unit. There were 69 cases ofcongenital heart defects (CHD) collected from a data base search.Results: Abnormal chromosomes were present in 47% of cases. Fif-teen percent had trisomy 18, 13% trisomy 21 and 5% trisomy 13. Tri-ploidy, monosomy X and 22q11 deletion occurred with a frequency of3% each. Unusual chromosome anomalies were trisomy 16 mosaic andtrisomy 7p, unbalanced translocation 7 and 11. Chromosome analysiswas unavailable in 9% of cases. Maternal age was significantly differ-ent for those with CHD and an abnormal karyotype (P ¼ 0.019, t-test).CHD and the presence of other fetal anomalies was significantly asso-ciated with an abnormal karyotype (Fisher’s exact test, P ¼ 0.029).Hypoplastic left heart syndrome was the most common heart defectfound with 44% of fetuses having an abnormal karyotype. Fifty per-cent of fetuses with an atrioventricular septal defect had an abnormalkaryotype. Hypoplastic left heart was found in 33% of trisomy 18fetuses and an atrioventricular septal defect in 50% of trisomy 21cases. These results show a poor outcome for fetuses diagnosed in uterowith CHD, with an overall mortality rate of 72%.Conclusion: The frequency and type of abnormal chromosomes foundin fetuses diagnosed with CHD agrees with other reports in the litera-ture. The risk of aneuploidy with CHD is influenced by maternal age.Detection of CHD is improved when other fetal anomalies are present.

F C 0 5 : 3 D

F35Effects of US on maternal–fetal bonding: 2D vs. 3D

D. H. Pretorius, K. M. Uyan, R. Newton, A. Hull,G. James & T. Nelson

Objective: To compare the effects of two-dimensional ultrasound(2DUS) vs. three-dimensional ultrasound (3DUS) imaging on thematernal–fetal bonding process.Methods: A total of 100 normal pregnant subjects were randomlyselected retrospectively. Ultrasounds were performed between 1998and 2000. Fifty mothers who had 2DUS and 50 who had 3DUS wererandomly selected to screen. We designed a postpartum survey consist-ing of 10 questions regarding parental–fetal bonding for telephonicinterview.Results: A randomized sampling of data was performed to equalizecells for some study questions. Mothers who received 3DUS sharedtheir ultrasound photographs with significantly more number ofpeople (mean, 31.72) than did mothers receiving 2DUS (mean,16.02) (t ¼ � 2.810, d.f. 1, 46, P ¼ 0.007). Subjects screened with3DUS had a greater tendency to form a mental image of the babypost-screening (82%) than 2DUS subjects (39%). The 3DUS parentswere more likely to receive comments on similarities/differences withthe actual baby than parents having 2DUS studies (x ¼ 4.109 d.f. 1,P ¼ 0.043). Further, 70% of the mothers receiving 3DUS felt they‘knew’ the baby right after birth because of the images they sawcompared to only 56% of 2DUS moms.Conclusion: The 3DUS images appears to have a more positive influ-ence on the perceptions of mothers compared to 2DUS images. Speci-fically, mothers who had 3DUS showed their images to twice as manypeople as those having 2DUS studies. 3DUS may have the potentialbenefit to strengthen the support system for families expecting theirnewborns.

F36Three-dimensional obstetric ultrasound (3DUS) is perceivedby lay non-pregnant adults to be more valuable thantwo-dimensional ultrasound (2DUS)

A. D. Hull, D. H. Pretorius, R. Newton, S. A. Asfoor & G. James

Purpose: To determine the impact of 3DUS on non-pregnant layadults.Methods: A 15-minute audiovisual presentation on 3DUS and 2DUSwas shown to 120 students It included a brief explanation of the

technology and both 2D and 3D images of a normal fetus at 30 weeksgestation. A number of 87 subjects (64 female, 23 male, age21 � 0.2 years) completed a self-administered questionnaire after thepresentation.Results: Ninety percent of the subjects had never heard of 3DUS priorto the presentation. All were able to appreciate the difference between2DUS and 3DUS. Seventy-eight percent said that they would request3DUS in a future pregnancy. The commonest reasons given for thischoice were better detection of fetal anomalies and visualization ofgreater anatomical detail compared to 2DUS. Ninety-four percent ofsubjects believed that 3DUS was likely to be a significant tool for med-ical use. Seventy-five percent of women and 61% of men felt that 3DUSwould affect parental fetal attachment.Conclusion: Lay subjects readily appreciate the differences between2D and 3DUS. Most believed that 3DUS was superior to 2DUS inassessing the fetus, afforded a diagnostic advantage to physicians andwould have beneficial effects on parental attachment. These findingshave important implications when considering consumer demand fornew imaging modalities in obstetrics.

F37Sonographers and sonologists hold favorable views of therole of three-dimensional ultrasound (3DUS) in obstetricsfollowing a formal lecture on the subject

A. D. Hull, D. H. Pretorius, R. Newton & G. James

Purpose: To assess awareness and attitudes towards the use of 3DUSin obstetrics in a group of medical professionals attending a nationalimaging conference.Methods: Around 200 subjects attended a 45-min lecture on 3DUSand obstetrics at a major national meeting. The lecture included slidesand multimedia content. Eighty-eight individuals (68 female and 20male) completed a self-administered questionnaire regarding theirimpressions of the subject following the talk.Results: No subject had personally undergone 3DUS, however, 78%of women said that they would like to have a 3DUS in a future preg-nancy. Sixty-five percent of men said they would encourage their part-ner to have a 3DUS in a future pregnancy. Of those who favored 3DUSin a future pregnancy, 54% cited better visualization of fetal anatomyas their main reason for this choice, 12% felt that 3DUS affordedadvantages over 2DUS in anomaly detection. Only 5% felt that3DUS would be more reassuring than 2DUS to them personally.Despite this, 80% of all subjects felt that 3DUS would have a major

11th World Congress on Ultrasound in Obstetrics and Gynecology FC04: Cardiac 2

20

role in reassurance for lay women carrying a normal fetus in the future.Seventy-two felt that 3DUS should be more widely available for prena-tal diagnosis than at present. 3DUS was felt to offer advantages over2DUS for anomaly detection by 38% of subjects. The multiplanar cap-ability of 3DUS, production of ‘realistic images’ and ready recognitionof fetal structures were cited as reasons for this view.Conclusion: 3DUS used in an obstetric setting appears to have alreadygained significant levels of acceptance amongst this audience of ima-ging professionals. Most believe that 3DUS should be more widelyavailable for the evaluation of both normal and abnormal fetuses.The value of 3DUS in the production of readily understandable imagesof a normal fetus in the reassurance of pregnant women was widelyendorsed.

F38Freehand three-dimensional ultrasound in the diagnosisof fetal anomalies

F. A. Guerra, J. Caro, J. Mansilla, R. Gutierrez, R. Aguilar & A. Isla

Purpose: To determine the clinical impact and additional informationof freehand three-dimensional ultrasound (3DUS) in the diagnosis offetal anomalies compared to two-dimensional ultrasound (2DUS).Methods: We studied a total of 146 anomalies found in 95 fetuses. Inall of them the fetal anomaly was first assessed by 2DUS, and then byfreehand 3DUS. According to the fetal anomaly, fetal or maternal con-ditions, one or more of the following modes of data capture wereemployed: volume capture (n ¼ 42 anomalies), static capture (n ¼ 21anomalies), and free style capture (n ¼ 83 anomalies), then the mostsuitable screen presentation was used. Clinical impact, fetal outcomeand additional information were determined in all cases.Results: Freehand 3DUS images provided additional information onlyin three anomalies (2%) compared with 2DUS, were equivalent in 132anomalies (90%), and were disadvantageous in 11 anomalies (8%) Inmost cases, when the anomalies were well visualized by 2DUS, the3DUS images added no further information. In none of the casesthe 3DUS changed the clinical management. Furthermore, all filesobtained by 3DUS were more advantageous in term of post-processingcapabilities, manipulation of images by multiples physician and teach-ing activities compared with 2DUS.Conclusions: The different freehand 3DUS modalities were diagnosticequivalent to 2DUS in the majorities of the clinical cases and has noadditional clinical impact on patient management. The most importantadvantages were related to the post-processing capabilities.

F39Abstract withdrawn

F40Three-dimensional sonographic assessment of fetalbehavior in the early second trimester of pregnancy

A. Kuno, T. Yanagihara, U. Hanaoka, H. Tanaka& T. Hata

Objective: The purpose of our study was to evaluate the fetalbehavioral pattern in the early second trimester of pregnancy byuse of a specially developed abdominal real-time three-dimensionalsonography.Methods: Real-time three-dimensional sonographic examinationswere performed on 11 normal pregnant women from 14 to 18 weeksof gestation. This imaging system provides continuous three-dimensional sonographic images every 1–2 s. The fetal movementwas monitored for 60 min in each patient, and head, mouth, arm,trunk, and leg movements were evaluated. All fetal behavioral patternswere observed during the period studied.Results: Active phase (time with fetal movements) was 59.4%, andresting phase 40.6%. The most active fetal behavioral pattern wasan arm movement, whereas the least was a mouth movement. More-over, each fetal movement was synchronized and harmonized withother fetal movements.Conclusions: Real-time three-dimensional sonography provides anovel means on the evaluation of fetal behavior in the early secondtrimester of pregnancy. These results suggest that real-time three-dimensional sonography become an important modality in futureearly fetal behavior research and in evaluation of fetal well-being inthe second trimester of pregnancy.

F41The three-dimensional ultrasound diagnosis of acrania inearly pregnancy

T.-Y. Hsu

Objective: To compare the application of two- and three-dimensional(2D and 3D) ultrasound for the antenatal diagnosis of acrania in theearly pregnancy.Materials and methods: Two- and three-diminsional (2D and 3D)ultrasound in early pregnancy were studies prospectively between April1999 and March 2001. A volume scan (Voluson 530D Kretz Technik,Austria) were performed after the conventional 2D (Acuson 128 XP,mt. view, CA) ultrasound examination were complete. The ability tovisualize acrania and weeks gestation of diagnosis were comparedbetween the two techniques. All relevant English-language clinicalresearch articles retrieved were reviewed and compared.Results: Five fetuses with acrania in early pregnancy was detected, ourcases were diagnosed earlier than previous reports (13.4 � 1.67 weeksvs. 17 � 1.83 weeks). The serum level of a fetoprotein were elevated infour cases. Only two fetuses had chromosome study and all werenormal.Conclusion: Our experience of the use of 3D transvaginal ultra-sound, a more accurate version of the fetal brain and face can beobtained. We believe that with technical improvements 3D, sonogra-phy permits more detailed evaluation than is possible with current2D sonography.

23–28 October 2001, Melbourne, Australia FC05: 3D

21

F C 0 6 : F E T A L D I A G N O S I S

F42Chorionic villus sampling for prenatal diagnosis in Zagreb:experience with 3250 cases

M. Podobnik, M. Podgajski, Z. Duic & S. Ciglar

Background: This study comprises data from 3250 women allocatedto transcervical CVS (TC-CVS), transabdominal CVS (TA-CVS)and transvaginal CVS (TV-CVS). We have assessed the efficacy oftransabdominal CVS compared with transcervical CVS and transvagi-nal CVS, and examined factors that have been implicated in causingspontaneous abortion.Methods: Over an 11-year period, 3250 women underwent chorionicvillus sampling at 9–12 weeks of gestation, by transcervical (500patients) transabdominal (2650 patients) and transvaginal technique(100 patients). Transvaginal color Doppler was used to investigatethe uteroplacental and fetal vessels in 500 pregnancies between 11and 12 weeks of gestation (400 TA-CVS and 100 TC-CVS) beforeand after CVS procedures.Results: Five hundred (15.4%) patients underwent TC-CVS and spon-taneous abortion after TC-CVS occurred in five cases (1.0%). 2650(81.5%) patients underwent TA-CVS using a 20 gauge spinal needleand abortion after TA-CVS occurred in 11 cases (0.4%). One hundred(3.1%) patients underwent TV-CVS using a 20-gage needle and abor-tion after TV-CVS occurred in three cases (3.0%). The spontaneousabortion rate after CVS procedures was 0.7%. Some 121 (4.6%) casesshowed chromosomal aberration. The spontaneous abortion rate waslower among cases allocated to TA-CVS after 11 weeks of gestation.There were no significant differences in mean pulsatility indices (PI)between maternal, fetal and intraplacental circulation, before and afterCVS procedures.Conclusions: Transabdominal CVS has lower fetal loss rate thantranscervical and transvaginal CVS.

F43Chorionic villus sampling after the first trimester

M. Podobnik, M. Podgajski, Z. Duic & S. Ciglar

Background: The purpose of this article was to evaluate the associa-tion between late CVS (placental biopsy) and complications betweensampling and delivery.Methods: Late chorionic villus sampling under ultrasound guidancewas carried out in 3000 (85.7%) cases in the second trimester and500 (14.3%) cases in the third trimester of pregnancy. Out of 3500 lateCVS, 710 (20.3%) were performed because of suspicious ultrasono-graphic findings. In the 500 patients between 13 and 16 weeks of gesta-tion color, Doppler was used to investigate the uteroplacental and fetalvessels before and after late chorionic villus sampling.Results: In 30 patients (0.85%), complications between sampling anddelivery were found. There were only 11 (0.31%) spontaneous abor-tions four to six weeks after late chorionic villus sampling. We found166 (4.74%) chromosomal abnormalities. In the group with suspiciousultrasonic findings (710 cases), we found significant oligohydramniosin 314 (44.2%) and significant polyhydramnios in 155 (21.8%), and110 (15.5%) had chromosomal abnormalities. Among the 166 patientswith chromosomal abnormalities, ultrasonographic findings in 90(54.2%) were detected after 20th week of pregnancy. There were nosignificant differences in mean pulsatility indices between uteroplacen-tal and fetal vessels before and after late chorionic villus sampling. Pre-liminary data for 25 trisomic fetuses indicate an abnormally increasedumbilical PI and abnormally decreased middle cerebral artery PI.Conclusions: Late CVS is a safe method of prenatal diagnosis for high-risk couples and does not significantly affect the outcome of pregnancy.

F44Does chorionic villus sampling (CVS) influence fetal growth?

T. Larsen & T. H. Nguyen

Purpose: To examine the relationship between CVS and birth weight.

Methods: A cohort of singleton pregnancies with CVS carried out inDenmark (n ¼ 19737) was compared to the total population of single-ton pregnancies without AC (amniocentesis) or CVS (n ¼ 873254) forinfants born in the years between 1980 and 1996. Exclusion criteriawere multiple pregnancies, infants with malformations, chromosomalabnormalities, stillbirths, and death within the first living year.The pregnancies with CVS were grouped according to gestationalage (8–12th week of gestation) at the time of CVS, and mean birthweight was calculated for each group. Maternal age, parity and pre-vious abortions were taken into account when comparing mean birthweights. To exclude social status as a confounder, the overall meanbirth weight (pregnancies without CVS or AC) was also compared tothe mean birth weight of a subgroup of infants (also without CVS orAC) born to women who undergo CVS in some other pregnancy.Results: The mean birth weight of the CVS cohort was found to besignificantly higher than that of the total population and inverselyrelated to the gestational age at CVS.Conclusion: The invasive procedure seems to stimulate growth. Noknown potential cofounders can explain the higher birth weight afterpregnancies with CVS.

F45An audit of invasive prenatal diagnosis for Bart’shydrops

T. Y. T. Tan & G. S. H. Yeo

Objective: To audit invasive prenatal diagnosis performed forpregnancies at risk of Bart’s hydrops fetalis in KK hospital.Methodology: This is a retrospective study in a single tertiary center inSingapore. All cases of invasive prenatal diagnosis performed for par-ents whose pregnancies are at risk of Bart’s hydrops fetalis between 16September 1993–28 November 2000 were traced from the KK invasiveprocedures register.Results: A total of 63 invasive prenatal procedures were done (59chorionic villus sampling and four cordocentesis) for a total of 44 cou-ples. The majority of the carriers had SEA deletions (84 or 95.5%);Thai deletions (2), compound beta and SEA alpha thalassaemia-1 traits(1), and HbH disease (1) accounted for the rest. The number of inva-sive prenatal procedures has increased steadily through the years from3 in 1994 to 13 in 2000. The major indications for the invasive proce-dures were past history of Bart’s hydrops (30) and both parents beingknown alpha thalassaemia carriers (27). Both parents were identifiedas carriers at �12 weeks in the 1st pregnancy in 12 couples (27.3%),beyond that time but before the development of hydrops in any preg-nancies in 13 couples (29.5%), and after the development of hydrops in19 couples (43.2%). Eighteen Bart’s hydrops were diagnosed and mostof them were terminated, either by vacuum aspiration within the first14 weeks (8) or by prostaglandins beyond the first trimester (8). Theiatrogenic fetal loss rate for the invasive procedures was 0% in thisaudit.Conclusion: Further measures are necessary to increase the proportionof at-risk couples who are identified as carriers before the first trimesterin the first pregnancy to allow early and safe prenatal diagnosis andearly terminations of affected pregnancies.

F46Is nuchal translucency measurement useful as a screeningtest for major cardiac abnormalities?

A. Lee-Tannock, R. B. Cincotta, F. Y. Chan & F. Hawthorne

Aim: To determine if measurement of nuchal translucency is a usefulpredictor of major cardiac abnormality in a low risk population.Methods: A total of 1888 fetuses (1849 patients) were scanned in ourdepartment as part of a nuchal translucency screening programmebetween 1 January 1999 and 31 December 2000. Those pregnancieswith abnormal karyotypes were excluded from this study. Follow upwas obtained and cases of major isolated cardiac anomalies wereidentified.

11th World Congress on Ultrasound in Obstetrics and Gynecology FC06: Fetal Diagnosis

22

Results: There were 35 fetuses that had normal chromosomes and anuchal translucency measurement greater than the 95th centile. Nineof these had a nuchal translucency measurement greater than3.5 mm, and 12 of these had a nuchal translucency measurementgreater than 4.5 mm. In our study population, there were six fetusesdiagnosed with an isolated major cardiac defect (3.2/1000 births).All of these fetuses had normal nuchal translucency measurements lessthan the 95th centile and none came from the group with increasednuchal translucency. In one case, the diagnosis was made at 12 weeks,three were diagnosed at the 18–20 weeks morphology scan and twowere diagnosed after birth.Discussion: Several studies have demonstrated a link between anincreased nuchal translucency measurement and cardiac defects infetuses with normal chromosomes. Hyett suggests that increasednuchal translucency at 10–14 weeks will detect 55% of major cardiacabnormalities. According to the risk estimates from this paper, we esti-mate that there should have been 1.5 cases of cardiac anomalies in thelow risk group whilst there should have been two cardiac anomalies inthe fetuses above the 95th centile. We found six major cardiac anoma-lies in the low risk group and none in the higher risk group. However,the difference between the estimated and observed number of cases wasnot statistically significant due to the small size of our screening study.We would recommend larger studies to determine the efficacy ofnuchal translucency as a screening tool for detection of cardiac anoma-lies. Detailed cardiac imaging at the 18–20 weeks morphology scanshould continue to be performed on all patients regardless of thenuchal translucency measurement in the first trimester.

F47Determination of fetal gender by first trimesterultrasonography

A. A. Yamasaki & G. Braia

Objective: To assess the accuracy of first trimester ultrasonography indetermining fetal sex by orientation of genital tubercle in relation tospine in a midsagittal plane.Material and methods: Between May 2000 and 2001, 110 patientswere studied from 11 to 14 weeks of gestation (average 12.1 weeks)at first trimester ultrasonography. The genital region was examinedin a midsagittal plane and the fetal gender was assigned as maleif the angle of the genital tubercle to a horizontal line through thelumbosacral skin surface was greater than 308 and female when the

genital tubercle was parallel or convergent (<308). All final resultswere checked after birth.Results: Among 148 patients, only in 16 (10.8%) fetal sex was notdetermined. In 132 patients (89.2%), genital tubercle was identifiedand sex was determined. There were 106 (80.3%) hits and 26 mistakes(19.7%).Conclusion: The diagnosis of fetal sex at first trimester has an impor-tant role in case of X-linked anomalies so invasive procedures can beavoided if fetuses are female. Our results are similar to others in litera-ture. The accuracy of gender determination by ultrasonography was80.3%. Most common mistake was to assign male fetuses as female.We believe that this error is due to the angle variation of the genitaltubercle during the exam.

F48The role of magnetic resonance imaging (MRI) in theprenatal diagnosis and evaluation of fetal tumors

S. Yagel, S. Nadjari & M. Gomori

Objective: Fetal tumors are rare abnormalities, problematic to diag-nose accurately by ultrasonography alone. In many cases precise diag-nosis depends on advanced tissue characterization, difficult toaccomplish with ultrasound. The recent advent of rapid MRI has madethis modality more amenable to prenatal diagnosis. Our aim was toinvestigate the role of this modality in the diagnosis fetal tumors.Patients and methods: Thirteen cases with suspected fetal tumorsunderwent thorough early and/or late ultrasound examination, fol-lowed by rapid MRI. Postnatal MRI examination or postmortem(PM) pathology was performed as indicated.Results: Ultrasound examination follow by MRI revealed the follow-ing cases: brain tumor, 2; tuberous sclerosis, 1; intracerebral hemor-rhage, 1; brain cyst, 1; cystic adenomatoid malformation CCAM, 2;pericardial teratoma, 1; intra thoracic hygroma, 1; sub diaphragmaticlung sequestration, 2; hemorrhagic ovarian cyst, 1; liver hemangioma,1. Postnatal or PM examination confirmed prenatal rapid MRI diagno-sis in 12/13 cases, while ultrasound diagnosis was confirmed in 7/13cases. In 10/13 cases MRI examination was more accurate than ultra-sound examination and significantly improved our prenatal evaluationof the findings, patient counseling and treatment.Conclusion: Rapid MRI plays an important role in the precise diagno-sis of fetal tumors especially when tissue characterization is essentialfor the diagnosis.

F C 0 7 : S C R E E N I N G F O R F E T A L A N O M A L Y

F49Accuracy of antenatal ultrasound diagnosis assessedby Gold Standard – postmortem

D. M. Amaraddio, F. Y. Chan, M. Stone, D. Cooper& D. Payton

Background: Antenatal ultrasound is an accepted and widespreadmethod for prenatal diagnosis. It has been shown that the performanceof ultrasound in tertiary referral centers is better than in general units.Even so, some anomalies could still be missed in tertiary centers. Auditsbased on neonatal outcomes may not be completely accurate, as someanomalies may not be apparent yet. Postmortem is considered to be theGold Standard for all identifiable fetal anomalies.Objective: To evaluate the accuracy of antenatal ultrasound diagnosisin a tertiary referral center, using postmortem findings as the GoldStandard.Methods: The departmental ultrasound database was searched from1993 to 1996. All fetuses who had a fetal or perinatal death were iden-tified. This was matched with the database in the department of pathol-ogy. Charts of fetuses who matched the entry criteria (antenatalultrasound performed in the department, and postmortem available)

were reviewed. The postmortem findings were compared to theultrasound report.Results: One hundred and ninety-two fetuses with perinatal deathwere identified from the ultrasound database. One hundred and fourreceived a postmortem. In 64 cases (70.3%), there was full concor-dance between postmortem and ultrasound findings. In 23 cases(25.3%), the ultrasound findings were confirmed, but additionalanomalies were detected at postmortem. In three cases (3.3%), therewere minor variations between the ultrasound and postmortem find-ings. There was no false positive diagnosis, and the only major anom-aly not detected antenatally was one case with placenta accreta. Of allorgan systems classified, best performance was in neural tube defects,where full concordance was found in 95.6% of cases. Most of thevariations between ultrasound and postmortem were in fetuses withmultiple anomalies. None of these variations would have affectedobstetric management.Conclusion: In 96.7% of the fetuses identified to have congenitalanomalies in this series, the findings were confirmed in postmortem.Postmortem contributed further information in 28.6% of cases. Whileantenatal ultrasound in this tertiary setting is shown to be accurate,postmortem is still important for future counseling.

23–28 October 2001, Melbourne, Australia FC06: Fetal Diagnosis

23

F50Routine ultrasound screening in 12169 pregnancies

F.-Y. Chan, S. Wong, R. Cincotta, A. Parry, A. Welsh,M. Stone & D. Cooper

Purpose: To assess the detection rate for congenital malformationsin a routine ultrasound screening program in a tertiary Australianinstitution.Methods: The findings of routine ultrasound screening performed inthe Mater Mother’s Hospital, Brisbane over a 6-year period was pro-spectively collected. Pregnancy and neonatal outcomes were obtainedfrom the obstetric and neonatal database. The presence of major con-genital anomalies, as classified according to the Australian NationalPerinatal Statistics Unit, was checked against the hospital’s perinataldatabase.Results: Over the 6-year period, a total of 26649 obstetric ultrasoundscans were performed in the unit, 12169 (45.7%) were routine ultra-sound scans. This is the largest Australian cohort study on routineultrasound screening. A total of 75 major congenital anomalies weredetected during the antenatal period. The number of major congenitalanomalies in the cohort was 131, giving a sensitivity of 57.3%. Thesensitivities according to different organ systems were: central nervoussystem (78%); urogenital tract (83%); gastro-intestinal tract (38%);cardiovascular system (27%); skeletal system (50%); respiratory sys-tem (67%) and others (63%). These results were significantly betterthan that from the RADIUS study (P < 0.001), and comparable tothose from other tertiary referral centers.Conclusion: The sensitivity for detection of congenital anomalies byroutine ultrasound screening at Mater Mothers’ Hospital is compar-able to most published series from tertiary institutions, and signif-icantly better than that reported in the RADIUS study.

F51Feasibility of a national Web-based ultrasound learningportfolio for OB/Gyn residents: the KOALA program

K. F. K. Fung, A. Gagliari, M. Walker, C. Lacasse, W. Faught& M. F. K. Fung

Purpose: To establish the feasibility of a national ultrasound learningportfolio in resident education in OB/Gyn in order to create a needsassessment for educational programming, ensure students are meetingeducational objectives and promote self-directed learning.Methods: The Internet-based KOALA learning portfolio was intro-duced into 15 of 16 Canadian medical schools in 1997 to track residentlearning encounters in OB/Gyn ultrasound. Data entered over 3-yearperiod analyzed using SPSS 10.0 to identify: (1) categories of ultra-sound cases entered; (2) critical incidents of learning; (3) domain andstimulus of questions posed; and (4) educational resources used forlearning.Results: Between 1 July 1997 and 30 June 2000, 3368 ultrasoundencounters were voluntarily recorded by 50 residents at 11 university-affiliated residency training programs. Twelve residents (25%)were categorized as high-volume, having entered ¼ 100 cases. Learn-ing encounters were recorded according to the following themes:routine second trimester dating scans; diagnostic procedures; amnio-centesis, CVS, cordocentesis; assessment of fetal well-being; targetedanomaly screen; routine gynecological exam. Critical incidents oflearning were identified in 25% of ultrasound cases (837/3368). Over-all, the majority of questions asked by residents following their ultra-sound learning encounter were of a cognitive nature (53.3%), and weresignificantly associated with identification of a critical incidence oflearning (P < 0.001). Overall, medline was the educational resourcemost likely to be used case-based learning when critical incidents wereidentified (P < 0.001). Questions of high volume users were most oftenstimulated by reflections and self-assessment (81.5%) when comparedto low-volume users (61.3%). Textbooks were the educationalresource most often used by high-volume users (73.9%) to answerquestions concerning their educational encounters. High-volume userswere more likely to report a change in their practice as a result of learn-ing acquired through the use of KOALA systems (P < 0.001). Barriersidentified during implementation included lack of computer Internet

access in the clinical environment, suboptimal faculty developmentinitiatives, and failure to integrate the learning portfolio as a compo-nent of the resident evaluation process.Conclusions: (1) Implementation of a Web-based learning portfolio isfeasible, allowing collection of data from geographically dispersedareas and programs; (2) voluntary use of the KOALA system wasacceptable when the system was made available to OB/Gyn residents;and (3) use of Internet-based learning portfolio such as KOALA pre-sents a unique opportunity for ultrasound training programs and spe-cial society such as ISUOG to assess the educational needs of trainingphysician learners.

F52MRI and US imaging in fetal anomalies

R. Ximenes, D. S. Ximenes, A. Ximenes, J. Sjzenfeld, S. Ajzen,G. D’Hippolitto, S. M. Goldman & A. C. S. Carvalho

Purpose: The use of MRI in diagnose fetal abnormalities and thecorrelation with ultrasound.Methods: Twenty-five patents in which were diagnosed fetal abnorm-alities (CNS, thorax and renal) were submitted to MRI. The stage ofpregnancy at the time of MR imaging ranged 16–34 weeks gestation.The half-Fourier single shot rapid acquisition with relaxation enhance-ment (RARE) technique was used to obtain T2-weighted images, thissequence has excellent contrast resolution of fetal tissues. It is a singleselection acquisition technique limits the artifacts related to maternaland fetal motion.Results: Fetus with CNS, thorax and renal abnormalities were sub-mitted to MRI study. The CNS anomalies were vein of Galen aneur-ysm, archnoid cyst, Dandy–Walker complex, megacysterna magna,espina bifida. Thorax anomalies were adenomatoid cystic malforma-tion. The renal anomalies were obstructive urophaty, renal displasia.The images of the ultrasound scans and the MRI images were analyzedindividually and after compared each other to understand the limitsand benefits of each technique.Conclusion: Fast MR imaging techniques provide excellent resolutionfor imaging the fetal anatomies without the need for maternal or fetalsedation, and should be used as a complementary tool in the diagnosisof fetal anomalies.

F53Accuracy of assessment of fetal anomalies usingcompressed fetal ultrasound video clips

L. M. Begg, F. Y. Chan, G. Edie, R. Hockey & R. Wootton

Background: Ultrasound is an accepted method of diagnosis andassessment of fetal anomalies, but the performance of ultrasound var-ies widely between tertiary and general units. Telemedicine involvesthe delivery of health care across a distance by combining telecommu-nications and medical expertise, and bridges the health care gapbetween the city and the regional centers, where access to subspecialistcare is usually limited. Our group has shown that real-time teleultra-sound consultation is not only technically feasible, but also offersdiagnostic and management modifications.

Current real-time telemedicine relies on digital (ISDN) lines andhence is costly and still limited to larger centres. If the Internet couldbe used as a method of downloading ultrasound images taken in thefield quickly and effectively, this would bring tertiary consultation toeven extremely remote centers. Before ultrasound video clips could betransferred across the Internet efficiently, they need to be compressedinto smaller files. Our group has previously determined that the max-imum compression ratio that was considered clinically acceptable(without compromising image quality) was 1:50 to 1:100.Aim: To assess the ability of experienced ultrasonologists to makeaccurate diagnoses based on compressed ultrasound clips of fetalanomalies, as well as their confidence in making such diagnoses.Methods: Nine different clips of fetal anomalies and two different nor-mal controls were compressed using commercially available software.The original uncompressed file sizes ranged from 55.6 to 622.7 MB.Final file sizes ranged from 1.6 to 10 MB (i.e. compression ratios of1:10 to 1:130). Four experienced tertiary ultrasound specialists blinded

11th World Congress on Ultrasound in Obstetrics and Gynecology FC07: Screening for Fetal Anomaly

24

to the compression factor evaluated the images. The examiners wereasked to make a diagnosis from a selection of 48 possible choices. Theywere also asked to rate their confidence in making the diagnosis (basedon the quality of transmission) and image clarity on a scale of 1–7.Data was analyzed using Duncan’s multiple range test for variables,with alpha ¼ 0.05.Results: A total of 71 versions of the 11 video clips were played at ran-dom to the observers. Intra- and interobserver variations were not sig-nificant. All of the observers were certain of their diagnoses (meanscore above 4), although the certainty scores for cardiac anomaliesand for normal fetuses were lower than the scores for other anomaliesOf the 71 video clips, 66 were correctly diagnosed (93%). Two werepartially correct (2.8%), and three were incorrect (4.2%). All theincorrect diagnoses were cardiac anomalies.Conclusion: Video clips of fetal ultrasound can be digitized, com-pressed and displayed on the computer without clinically or statisti-cally significant loss of diagnostic certainty or image clarity. Forcompressions up to 1:100, the diagnostic accuracy was very high,although cardiac anomalies seem to pose the most difficulties. Whilethe potential for consultation via the Internet is enormous, largerclinical trials need to be performed to confirm these results.

F54Transmission of real-time fetal ultrasound images using theIP network vs. ISDN lines

F.-Y. Chan, A. Taylor, B. Soong, P. Timothy, A. Lee-Tannock,L. Begg, R. Cincotta & R. Wootton

Background: Our group has shown that real-time teleultrasound con-sultation is not only technically feasible, but also offers diagnostic andmanagement modifications in the clinical setting [1,2]. It helps tobridge the health care gap between the city and the country, whereaccess to subspecialists is usually difficult. Majority of telemedicinework around the world is currently performed with ISDN connection,which is still relatively costly, and is limited to larger centers whereISDN access is available. Rapid evolution of technology has enabledInternet technology to be available to most corners of the world, andthe cost of using IP networks is rapidly reducing.Purpose: To compare the quality of real-time ultrasound imagestransmitted via the IP network vs. ISDN lines.Methods: ISDN and/or IP (broadband wireless local multipoint distri-bution service, AAPT) connections between three Queensland Hospitalsites were established. A standard ultrasound videotape containingfetal images from three different trimesters of pregnancy was playedto four experienced observers at random using either ISDN or IP con-nections at various bandwidths. The observers were blinded to the net-work or bandwidths used, and evaluated the quality of the images witha standard seven-point scoring sheet on 30 fetal anatomic landmarks.Results: A total of 12 evaluations were performed for various combi-nations of bandwidths and networks. The intra- and interobserver

variations were not significant. The most significant variableaffecting the mean score was the bandwidth used. The mean scorefor transmission using IP was about 0.5 points lower than that usingISDN linkage across all the different bandwidths, but the differencesdid not reach statistical difference (P ¼ 0.06).Conclusion: IP connection is a reasonable alternative to ISDN linkagefor real-time teleultrasound transmission. It has enormous potential toexpand telemedicine applications to more remote sites at lower cost,and further explorations as to its reliability and limitations should beperformed.References

1 Chan FY, Whitehall J, Hayes L et al. Minimum requirements forremote real-time fetal tele-ultrasound consultation. J TelemedicineTelecare 1999; 5 (3): 171–6.

2 Chan FY, Soong B, Lessing K et al. Clinical value real-time tertiaryfetal ultrasound consultation telemedicine – preliminary evaluation.Telemedicine J 2000; 6 (2): 237–42.

F55The outcome of antenatally diagnosed talipesequinovarus

S. P. Bakalis, S. Sairam, M. Patton, K. Harrington,K. Nicolaides & B. Thilaganathan

Aim: Congenital talipes equinovarus is a relatively poorly studied con-genital anomaly with very little information available for parents. Theaim of this study is to investigate the natural history and postnatal out-come of fetal talipes detected at 18–23 weeks in a large, unselectedobstetric population.Patients and methods: This is a prospective study of routine antenatalultrasound screening in an unselected obstetric population, carried outon patients booking at three major London hospitals. Congenitaltalipes was classified as postural, tetralogic (associated with other con-genital abnormalities) and idiopathic. Outcomes measured includedtermination of pregnancy, stillbirth, aneuploidy, fetal structuralabnormality and neonatal death.Results: The incidence of congenital talipes was 0.11% (111/103,228)in our study population. Of these, 64.9% were live born, 31.5% wereterminated, and 3.6% were intrauterine deaths. Talipes was bilateral in60% of the cases, with an overall male to female ratio of 1.5:1. Therewere significant differences in the prevalence of tetralogic, idiopathicand postural talipes in the unilateral vs. bilateral talipes groups. Theoutcome of pregnancy was good in 97% of fetuses with a diagnosisof idiopathic and in 36% of those with tetralogic talipes. Fetuses witha diagnosis of postural talipes were postnatally found to have otherabnormalities in 7% of cases.Conclusion: The outcome of antenatally detected congenital talipesis significantly worse than previously reported. The outcome ofpregnancy depends on the laterality of the talipes and the presenceof associated features.

F C 0 8 : F E T A L H E A D A N D C H E S T

F56Antenatal diagnosis of the absent inferiorcerebellar vermis – is there any clinicalsignificance?

J. S. M. Wong, B. H. Woo & G. S. H. Yeo

Objective: The clinical significance of antenatal ultrasound diagnosisof the absent inferior cerebellar vermis (AICV), whether isolated orassociated with other ultrasound findings, has not been well-under-stood with limited research previously. Because of its association withsome syndromes, e.g. Dandy–Walker or Walker–Warburg syndromes,we aim to study the outcome for the purpose of counseling of parentswhen this is diagnosed antenatally.

Methods: A single-center, prospective observational study of screen-ing scans of a low risk population was conducted from 1 October1996 to 30 September 2000. Cases of AICV diagnosed were followedup postnatally for up to 4 years with cranial ultrasound and develop-mental assessments.Results: There were 56 746 screening scans on 64 225 babies deliveredduring this period. Twenty-eight cases of AICV were diagnosedbetween 19 and 34 weeks. Six had isolated AICV and all occurred inwomen less than 35 years old. All had normal outcome and develop-mental assessment on follow up. The other 22 were associated withother ultrasound findings like echogenic bowel, structural heart defectsor Dandy–Walker malformation. Four of them occurred in mothers35 years and older with three (75%) having abnormal karyotype. Five

23–28 October 2001, Melbourne, Australia FC07: Screening for Fetal Anomaly

25

occurred in mothers aged 30–34 years with two (40%) having abnor-mal karyotype. Of the remaining 13 in those less than 30 years, five(38%) had abnormal karyotype. These include trisomy 13 and 18,Turner and Goldenhar syndrome.Conclusion: The antenatal diagnosis of the isolated AICV is associatedwith low risk of chromosomal abnormalities and thus, does not requirefurther karyotypic analysis. However, if other ultrasound findings arealso seen, an abnormal karyotype can be expected in up to 38% of thepatients (in our series), and even more (75%) if the maternal age is35 years and above. Karyotyping should be offered in these cases.

F57Transvaginal 2D/3D US and MRI of the fetal CNS

R. K. Pooh & K. Pooh

Introduction: Transvaginal sonography of the fetal brain produce aseries of oblique sections from the same ultrasound window. Transva-ginal 3D automatic volume acquisition requires only 1.84–6.07 s forscanning without a shift and angle-change of the transducer. Multipla-nar image analysis and volumetric system are useful for objective eva-luation of the fetal brain. Recent advances in fast MRI technology hasremarkably improved the T2-weighted image resolution withoutmotion artifacts, despite a short acquisition time. We used both tech-nologies for the fetal brain assessment.Patients and methods: Between May 2000 and April 2001, 11 caseswith fetal brain abnormalities at 20–37 weeks of gestation wereexamined by transvaginal 2D/3D sonography and MRI. Logiq 700/500 (GE MS) and Voluson 530D (Medison), and 1.5 T MR unitVISART (Toshiba) with 2D half-Fourier fast SE sequence were usedin this study. Obtained US images and MRI were compared for brainevaluation.Results: From our study, TVS is better in depicting the corpus callo-sum in the sagittal section, while MRI with 2D half-Fourier fast SEis superior in brainstem detection. In evaluation of the ventricular sys-tem, cerebrum and cerebellum, the same information was obtainedfrom both methods. The same prenatal diagnoses were obtained fromUS and MR data in 10 cases. In one case with partial agenesis of corpuscallosum at 20 weeks of gestation, US image was more demonstrativethan MRI.Conclusions: Prenatal neuroimaging has been approaching postnatalimaging. Both transvaginal sonography and MRI are useful for evalua-tion of the fetal CNS. Transvaginal 2D/3D sonography should be usedfor routine assessment of fetal CNS, and MRI device should be used formore detailed evaluation including brainstem assessment. Practically,MRI is useful regardless of fetal head position, while US is not influ-enced by fetal movement.

F58Gender-adjusted multiples of the median in amniotic fluidalpha-fetoprotein-screening for neural-tube defects halvefalse-positive rate without affecting detection rate

P. Kozlowski, A. J. Knippel & R. Stressig

Introduction: Previous studies have shown that fetal gender had influ-ence on various pregnancy complications and prenatal diagnostic bio-chemical markers. We have evaluated, whether elevation of amnioticfluid alpha-fetoprotein (AF-AFP) is associated with fetal sex andwhether a sex-related difference can help to identify pregnancies withAFP-associated malformations or poor pregnancy outcome.Material and methods: From our database, we obtained 6461 single-ton gestations with absolute values and corresponding multiples of themedian (MoM) for AF-AFP for the period database April 1997–March1999. In all instances high-resolution ultrasonography was performedbefore amniocentesis to measure biparietal diameter. Patients with AF-AFP �2 MoM were identified, details of pregnancy outcome wereobtained and compared to matched-pair-controls having AF-AFP <2MoM.Results: A total of 262 patients showed AF-AFP levels �2 MoM. Out-come information was available in 232 cases (88.6%) including allcases with fetal anomalies recorded at ultrasonographic examination.Of these fetuses significantly more had male gender (147 male fetuses

vs. 85 female, P < 0005). Having a positive screen result in the risk ofAFP-associated malformations was significantly higher for femalefetuses (25 female fetuses (29.4%) vs. 22 male fetuses (15%) withAFP-associated malformations, P < 0025). Pregnancies with falsepositive AF-AFP had a significantly higher risk for poor pregnancy out-come compared with pregnancies having normal AF-AFP (11 fetallosses (59%) from 185 vs. two fetal losses (0.9%) from 232,P < 0025) but fetal gender had no significant influence.Conclusions: Measuring AF-AFP routinely in amniotic fluid obtainedfor cytogenetic evaluation is a cheap method to identify fetuses, onwhich high-resolution ultrasonography in a level 3 center should befocused. Adjusting the cut-off MoM to 2.5 for male and to 2.0 forfemale fetuses halves the false-positive rate from 3.4 to 1.7 withoutaffecting the detection rate of 95% in our population. Larger studiesare required to determine suitable sex-adjusted cut-off values.

F59Isolated oral clefts (OC) detected by routineultrasonography – is it associated with chromosomalabnormalities?

H. Y. Wee, B. H. Woo & G. S. H. Yeo

Purpose: To determine if chromosomal abnormalities are associatedwith isolated OC (cleft-lip or palate) detected by sonographers trainedto perform level two fetal structural surveys in a low risk obstetricpopulation.Methods: A prospective observational study was conducted onwomen undergoing screening ultrasonography between 18 and34 weeks from 1 October 1996 to 30 September 2000 in a singlematernity unit. Obstetric and neonatal charts were reviewed in all caseswhere OC in isolation were detected by ultrasonography. Antenatallymissed cases of isolated OC were also analyzed.Results: There were 64 225 deliveries in this period. Follow up wascomplete for 40/42 cases of isolated OC detected by ultrasonography.Unilateral, bilateral and central OC accounted for 30, 12, 0 cases,respectively. Karyotyping was performed in 11 cases; nine cases hadnormal chromosomes, one had 46XY, 16qhþ (same as mother’s)and another had 46XX, t (2; 10) (q33; p12.2) (parental karyotype nor-mal). In both cases with chromosomal aberrations, isolated unilateralcleft-lip and palate were confirmed at delivery. Termination of preg-nancy was performed for the latter; whether the balanced translocationwould have contributed to an adverse outcome remains unknown. Sixnewborns had associated abnormalities at delivery: moderate VSD,mild pulmonary stenosis, bilateral clubfeet, Peter’s anomaly and twocases of polydactyly. Except for one case of polydactyly which hadno karyotyping, the other five had normal chromosomes. Karyotypingwas deemed unnecessary by the neonatologist for 27 newbornswith confirmed isolated OC. One newborn had no OC (PPV 39/40 ¼ 97.5%). After excluding all cases where associated structuralabnormalities were noted at birth, the sensitivity of detecting isolatedOC confirmed at delivery was 33/40 (82%).Conclusions: Isolated OC detected by routine ultrasonography in aunit with high sensitivity is associated with a 1/40 (2.5%) risk of a sig-nificant chromosomal abnormality. However, this abnormality mayhave no adverse effect. This is useful for counseling couples consideringprenatal karyotyping.

F60Prenatally diagnosed fetal thoracic malformations:a 5-year experience

C. E. Portmann & F. Y. Chan

Background: A 5-year retrospective audit of cases referred to theMater Mothers Hospital Fetal Medicine Unit with ultrasonographi-cally identified thoracic anomalies was undertaken with specific refer-ence to the distribution of pathologies, antenatal progress andoutcomes of affected fetuses.Results: Twenty-one fetuses with non-cardiac, intrathoracic anoma-lies were identified and managed over a 5-year period. Twelve demon-strated cystic mass lesions, seven echogenic lung masses were identified

11th World Congress on Ultrasound in Obstetrics and Gynecology FC08: Fetal Head and Chest

26

and one each of primary mediastinal shift (later found to be hypoplas-tic lung) and arterio–venous malformation of the lung. All lesions wereunilateral. Antenatal diagnosis of congenital cystic adenomatoid mal-formations of the lung (CCAML) was made in the 12 cases of macro-cystic lung lesions and proved correct in 10–1 was found at resection tobe a bronchoplumonary sequestration (BPS) with associated AV mal-formation and chondromatous hamartoma, and one awaits follow up.Of the seven echogenic lung lesions, the antenatal diagnosis provedcorrect in only the two cases of BPS. Neonatal diagnoses included con-genital diaphragmatic hernia (1), hybrid BPS/CCAM (1), resolvingbronchial obstruction (2) and one loss to follow up with normal chestX-ray. Complete resolution of the lesion occurred in five cases (mostlyechogenic lesions). Polyhydramnios was found in only one case andresolved spontaneously – there were no cases of hydrops or fetal death.Neonatal outcomes included respiratory distress at birth (2), respira-tory distress within 28 days (another two cases) and small for gesta-tional age at birth (2/21). Thoracic CT scans were performed orawaited in all cases barring one loss to follow up with normal chestX-ray. Some abnormalities were noted in all cases in which CT scanhas been performed to date (15), including those cases with normalneonatal chest X-ray and apparent antenatal resolution of ultrasoundfindings.Conclusion: While correct antenatal diagnosis can be made in manycases of cystic lung lesions, echogenic lung lesions may be difficult todiagnose correctly and adjunctive investigation such as MRI may provebeneficial. Postnatal CT scan should be considered as a routine for allantenatally diagnosed intrathoracic lesions regardless of apparentresolution of the lesion on antenatal ultrasound or apparently normalneonatal chest X-ray.

F61Relationship between fetal pulmonary artery blood flowvelocity waveform and fetal lung length in pregnanciescomplicated by oligohydramnios

M. Wojcieszyn, A. Lukasik, K. Swadzba, M. Stokowacka& R. Poreba

Objectives: The objective of this study is to evaluate weather there is acorrelation between Doppler velocity waveform parameters obtainedfrom fetal pulmonary artery and fetal lung length considered as oneof the lung hypoplasia risk parameters in pregnancies complicated byoligohydramnios.Methods: A total of 35 singleton pregnancies between 26 and39 weeks complicated by oligohydramnios due to premature ruptureof the membranes, placental insufficiency or idiopathic etiology and69 normal pregnancies were examined by real time ultrasonography,color and pulse Doppler. Pulse Doppler signal was obtained from thedistal part of right or left pulmonary artery after visualization withcolor Doppler at the level of four chamber view. The systolic to diasto-lic ratio, resistance index and pulsatility index were used to quantifyvelocity waveforms. The right lung of the fetus was measured longitud-inally from the apex next to clavicle to the base at the top of dia-phragm. The results of fetal lung length measurements werecompared with namogram given by Roberts.

Results: Statistically significant difference in distal pulmonary arterypulsatility index between normal pregnancies (1.74 � 0.23) and preg-nancies complicated by oligohydramnios (2.72 � 0.56) with decreasedvalue of fetal lung length (below fifth percentile) was observed. Thesensitivity, specificity, positive and negative predictive value of 95,91, 93, 91%, respectively, were obtained. In the group with oligohy-dramnios and normal values of fetal lung length, the difference in pul-satility index in comparison with normal pregnancies was notsignificant. No differences in systolic to diastolic ratio and resistanceindex between examined groups were found.Conclusions: Strong relationship between increased pulsatility in-dex value obtained from distal part of pulmonary artery and decreasedvalue of fetal lung length measurement in pregnancies complicatedby oligohydramnios was observed. The measurement of pulmonaryartery pulsatility index enriched with fetal lung length may giveadditional information useful in predicting pulmonary hypoplasia risk.

F62Congenital diaphragmatic hernia: prenatal evaluationand outcome

L. Masini, M. De Santis, G. Noia, B. Milej, A. F. Cavaliere,C. Romagnoli, L. Perrelli & A. Caruso

Objectives:

� To describe the incidence of congenital diaphragmatic hernia(CDH).

� To describe the hernia type (right, left, eventration), the gesta-tional age at diagnosis, the associated anomalies and chromosomaldisorders.

� To describe the perinatal outcome and the pre- and postoperativemortality rate.

Material and methods: A total of 51 fetuses and babies with congeni-tal diaphragmatic hernia or eventration (referred to collectively asCDH) were studied at Dept. Obstetrics and Gynecology, UniversitaCattolica di Roma, in the 20-year period from 1981 to 2001.Results: Prenatal diagnosis was made in 48 cases (17 were diagnosedbefore 26 weeks, 31 between 27 and 40 weeks), two cases were diag-nosed in the neonatal period and one case at the postmortem analysis.The most common hernia was left-sided (38 cases); six cases wereright-sided; four cases had eventration; four cases had complex anom-aly with exomphalos and ectopia cordis (suspicious Cantrell’s pental-ogy). Twenty-three fetuses had associated anomalies, including fourchromosomal disorders (trisomy 18 in two cases, trisomy 13 in oneand unbalanced translocation in one). Polyhydramnios was presentin 20 fetuses.Perinatal outcome: Two patients had termination of pregnancy inother institutions; 44 patients delivered (36 cesarean section and 8vaginal delivery); 2 pregnancies are still ongoing and 3 patients (earlydiagnosis) have no follow up. Out of the 44 live births, 16 died beforesurgery and 28 underwent surgical repair. Fifteen early postoperativedeaths occurred.Conclusions: The overall survival in babies with CDH is very poor,despite multidisciplinary management, mainly for associated anoma-lies. A careful ultrasound survey may suggest which fetuses have anisolated diaphragmatic defect and a better prognosis.

F C 0 9 : F E T A L A B D O M E N

F63The postnatal significance of mild fetal pyelectasis

T. M. Chudleigh & L. S. Chitty

Purpose: To determine the incidence of significant renal pathology infetuses with mild pyelectasis in the second trimester, and to defineprognostic criteria with which to identify those fetuses requiringdetailed postnatal follow up.Methods: A prospective, observational, multicenter study of unse-lected fetuses identified during routine second trimester anomaly

screening. Mild pyelectasis was defined as an antero-posterior diameterof 5–10 mm. All fetuses with unilateral or bilateral pyelectasis identi-fied between 16 and 25 completed weeks of gestation were recruited.Results: Four hundred and twenty-five fetuses with mild pyelectasiswere recruited from a population of 101600 pregnancies. Dilatationwas bilateral in 270 and unilateral in 155 cases. Multicystic or horse-shoe kidney was subsequently diagnosed in four cases. Prenatal pyelec-tasis improved in 90 cases, remained unchanged in 124 casesand deteriorated in 177 cases. One prenatal scan only was performedin 30 cases. Neonatal renal findings were abnormal in 185 cases.

23–28 October 2001, Melbourne, Australia FC08: Fetal Head and Chest

27

Pathological findings were present in 48 cases of which 16 resulted insurgery. The primary diagnosis was obstruction in 15 cases, duplexkidney in 11 and upper tract dilatation in 22 cases leading to surgeryin eight, five and three cases, respectively. Significantly, more cases ofpathology were present in fetuses where the dilatation increased to10 mm or greater in the third trimester than in those where it remainedunchanged.Conclusions: Postnatal upper urinary tract dilatation is demonstratedin 43.9% of fetuses with mild second trimester pyelectasis. Urinarytract pathology is present in 11.4% of such fetuses and 3.8% requiresurgery. All cases with persistent pyelectasis, particularly those withprogressive dilatation, should be investigated postnatally.

F64Fetal hydronephrosis – natural history and predictorsof postnatal outcome

Y. K. Sohn, S. Y. Han, D. W. Chun, J. S. Park, J. K. Jun,B. H. Yoon, H. C. Syn & H. P. Lee

Objectives: The aim of this study was to evaluate the natural history ofantenatally detected hydronephrosis and to assess the relationshipbetween the severity of fetal hydronephrosis and postnatal outcome.Materials and methods: During the study period (1992 through2001), we identified 70 cases of fetal hydronephrosis and reviewedtheir antenatal and postnatal follow up record retrospectively. Theinclusion criteria was renal pelvic antero-posterior (AP) diameter4 mm or greater before 33 weeks of gestation and 7 mm or greater ator after 33 weeks of gestation. Sonographically diagnosed fetal simplerenal cyst, multicystic dysplastic kidney, or polycystic kidney wereexcluded.Results: Thirty cases (43.4%) showed hydronephrosis without definiterenal pathology after postnatal evaluation. Thirty-one cases (44.9%)had anatomical abnormalities and the most common renal pathologywas ureteropelvic junction obstruction (18 of 31, 58%). Eight cases(11.5%) had regression of hydronephrosis before or at birth. One casewas lost in follow up. Twenty cases (28.9%) required surgical interven-tion, 32 cases (46.3%) needed medical intervention, and 17 cases(24.6%) required neither intervention nor long-term follow up. Allwith renal pelvic AP diameter of 15 mm or greater prior to deliveryneeded either medical or surgical intervention postnatally. With cutoff value of 15 mm or greater, the sensitivity and specificity in predict-ing surgical intervention were 90, and 75%, respectively. The fetuswith renal pelvic AP diameter of 15 mm or less prior to delivery hadless risk for surgical intervention than 15 mm or greater (Odds ratio0.046, 95% CI 0.009–0.224). Gestational age at detection of fetalhydronephrosis had no statistical significance in predicting outcome.Conclusion: We concluded that severity of fetal hydronephrosis couldaffect a postnatal management strategy, either medical or surgicalintervention, which is useful in prospective counseling.

F65Neonatal renal ultrasound – a 5-year reviewin an obstetric hospital

C. M. Acton & M. Pahuja

Aim: To review a consecutive series of neonatal renal scans presentingto the imaging department in an obstetric hospital, record the abnorm-alities seen and assess the accuracy of the antenatal diagnosis with sub-sequent outcome.Material and methods: In the 5 years (1996–2000), there were 788neonates referred for renal ultrasound scans. Many of the infantshad a micturating cystourethrogram (MCU) at the time of presenta-tion. The commonest indication in 92% of neonates was fetal pyelec-tasis reported on the antenatal scan. The hospital records of theseinfants were reviewed with further details obtained from cliniciansand surgeons and subsequent investigations.Results: The results were classified according to the findings. Therewere a total of 788 neonates who had a renal scan. There were 592(75%) in whom the scan was considered normal and of these 49 hadan MCU and 17 had associated vesico-ureteric reflux (VUR) grades1–4. There were 125 (16%) infants with persisting mild-moderate

pyelectasis/hydronephrosis and of these 61 resolved or were lost to fol-low up and 64 persisted at follow up scans variably timed 3–15 monthslater. Most of these had an MCU and 26 had associated VUR grade1–5, not always on the side of the pyelectasis. Miscellaneous abnorm-alities were seen in 71 (9%) neonates, multicysticdysplastic kidney 25,pelviureteric junction obstruction (PUJ) 7, solitary kidney 6, crossedrenal ectopia 5, uncomplicated duplex kidney 5, ureterocoele 5, twowith posterior urethral valves and one with vesico-ureteric junctionobstruction (VUJ). There were four infants with significant VUR, threewith thinned renal parenchyma and one with a small kidney. Therewere three with autosomal recessive polycystic kidney disease (Potter’stype II), one with small dysplastic kidneys.Conclusion: In this review, there were 113 neonates (14%) found tohave significant renal anomaly or VUR. Many of the miscellaneousfetal renal abnormalities have a characteristic and identifying appear-ance on the antenatal ultrasound. The majority of neonatal renal scansare referred because of mild/moderate fetal pyelectasis and many ofthese resolve over a variable time. Persisting moderate pyelocaliectasisis not a sensitive sign of VUR, nor in differentiating early PUJ obstruc-tion from the normal variant that subsequently resolves. Follow upultrasound may confirm resolution or increasing hydronephrosis.Nuclear medicine studies will compare function and drainage of thesystems in patients with PUJ obstruction.

F66Functional evaluation of normal kidney bythree-dimensional (3D) ultrasonography – the hourlyfetal urine production rate

J. K. Jun, S. M. Jin, E. M. Ko, J. S. Park, B. H. Yoon, S. Y. Moon,H. C. Syn & H. P. Lee

Background and objective: Fetal urine is the main source of amnioticfluid, which is very important marker of fetal wellbeing. Single deepestpocket measurement or amniotic fluid index method are widely usedto evaluate amniotic fluid volume. They are well correlated withreal amniotic fluid volume in cases of normal amniotic fluid volume.However, they are not in extreme cases, which are clinically moreimportant. Direct measurement of fetal urine production may beanother option to evaluate amniotic fluid volume and fetal kidneyfunction. Moreover, it is well known that 3D USG is superior to 2DUSG especially in volume measurement of irregular shape object. Weundertook this study to measure the hourly fetal urine production rate(HFUPR) with 3D US in normal pregnancies and evaluate its clinicalimplication and feasibility.Material and methods: From 20 weeks of gestation to term, 157 preg-nant women were enrolled. We used Voluson 730 to measure 3D blad-der volume. Vocal (virtual organ computer aided analysis) softwarewas used to measure the volume. Initially, we measured the volumeof fetal urinary bladder and observed whether the bladder was empty-ing or not in 5 and 10 min after initial measurement. We excluded 35cases whose bladder was found empty within 5 min. If distendedbladder was still found, we calculate the HFUPR. We constructed abest-fit graph of the HFUPR according to gestational age (GA) inweeks.Results: We constructed the following quadratic equation and plotteda graph on the basis of our data.

HFUPR ¼ 241.4 � 20.2 � GA þ 0.43 � GA � GA(r ¼ 0.849, F ¼ 153.9, P < 0.0001)

We also showed 10 and 90% values of each gestational week.Conclusion: We demonstrated that serial 3D USG assessment of fetalbladder volume could be used as an alternative method to evaluateamniotic fluid volume.

F67Increased risk for prenatal asphyxia and death in fetuseswith duodenal obstruction

A. Brantberg, H.-G. K. Blaas, K. A. Salvesen & S. H. Eik-Nes

Purpose: The aim of this study was to describe the outcome of a cohortof fetuses with duodenal obstruction diagnosed prenatally. Most

11th World Congress on Ultrasound in Obstetrics and Gynecology FC09: Fetal Abdomen

28

previous studies are based on results from newborns and have shown agood outcome with a survival rate up to 95–100%.Methods: Fetuses with prenatally diagnosed duodenal obstructionwere registered and prospectively evaluated at the NCFM fromJanuary 1985–December 2000.Results: Duodenal obstruction was found in 29 fetuses at mean29 weeks and 2 days. Polyhydramnios was found in 24 (83%). Six(21%) had trisomy 21. Associated anomalies, including trisomy 21,were found in 18 cases (62%). Four pregnancies with normal karyo-type ended with intrauterine fetal death at 31–35 gestational weeks.One of these had associated anomalies. In all cases, death occurredunexpectedly. Mean gestational age at delivery was 35 weeks and5 days. Eighteen (64%) delivered prematurely. The mean birth weightwas 2562 g. Four infants (14%) were small for gestational age. Threeinfants died postnatally; all three had associated anomalies. Twoinfants suffered from severe neurological impairment with cerebralpalsy and epilepsy, and two suffered from psycho-motoric impairmentsuggesting that these four might have had intrauterine asphyxia.Another fetus had sustained intrauterine bradycardia with subsequentemergency CS.Conclusions: The present study of 29 prenatally diagnosed duodenalobstructions showed a relatively poor outcome with four intrauterinedeaths, three postnatal deaths and four infants with neurologicalsequele. The neurological sequele indicate possible intrauterineasphyxia. We suggest that there may also be a common reason forthe cases of intrauterine fetal death as well as the case with bradycar-dia. We consider the possibility that it could be caused by bradycardia/asystolia following a vagal overactivity due to distension of the uppergastrointestinal tract.

F68Gastroschisis – can prenatal ultrasonography predictneonatal outcome?

J. R. Peter, F. Y. Chan & R. Hockey

Background: Gastroschisis carries significant neonatal morbidity, andoccasionally mortality. Previous studies looking at ways to prognosti-cate this condition by antenatal ultrasound have shown conflictingresults. Small numbers limited the power of most previous studies.The current study is one of the largest so far reported.Purpose: To evaluate the usefulness of prenatal ultrasonograpicparameters to predict neonatal outcome for gastroschisis.Methods: The charts, photos, and videotapes of all babies with gastro-schisis who were assessed during pregnancy at the Mater Mothers’Hospital, Brisbane over an 8.5-year period (January 1993–May2001) were reviewed. Bad neonatal outcome was defined as death,severe bowel complications (atresia, perforation, necrotic segments),need for multiple bowel operations, or a combination of these. Variousultrasonographic parameters assessed included: gestation at firstdiagnosis, maximum bowel diameter, maximum bowel thickness,

presence of other anomalies, evidence of growth restriction, andpolyhydramnios.Results: Forty-five patients met the entry criteria during the study per-iod. Six neonates died, 12 had severe bowel complications, and 14 hada combination of the bad outcomes. Apart from polyhydramnios, noneof the other ultrasound parameters were significantly associated withthe bad outcomes. Polyhydramnios was most significantly associatedwith severe bowel complications (P < 0.001, Fisher’s exact test), withlikelihood ratio of 11.7 (P ¼ 0.001).Conclusion: The current study show that most standard ultrasoundparameters assessed antenatally in fetal gastroschisis were not signifi-cantly associated with neonatal outcome, except for polyhydramnios,which is strongly predictive of severe bowel complications in the neo-natal period. This information is important both for parents and forhealth professionals managing pregnancies with fetal gastroschisis.

F69Fetal ovarian cysts: development and neonatal outcome

C. Mittermayer, B. Ulm, D. Dorfler-Grassauer, E. Horcher,J. Deutinger & G. Bernaschek

Introduction: The aim of the study was to evaluate the outcome offetal ovarian cysts in relation to their size and ultrasonic appearance.Methods: We retrospectively analyzed pre- and postnatal charts of 61infants with a prenatal diagnosis of ovarian cysts between 1991 and2000. When completely anechoic and with a thin wall, the cyst wasdefined uncomplicated. Cysts, which presented themselves echogenic,with a fluid-debris level, a retraction clot, or a septation were definedcomplicated.Results: In a total of 61 fetuses, 65 ovarian cysts were detected bytransabdominal ultrasound: 35 (57%) cysts on the left side; 22(36%) on the right side and four fetuses (7%) had bilateral cysts. Threepatients with uncomplicated cysts were lost to follow up and one fetuswith bilateral cysts died in the 27th week of gestation. In 17 cysts,treatment was necessary. Fourteen cysts (all complicated) were oper-ated after delivery because of persistence or enlargement. The histolo-gical results were either follicular or theca lutein cysts in 12 cases, onelymphangioma and one teratoma. Two cysts were aspirated in uteroand one after delivery. In the remaining 40 fetuses, 43 cysts where onlycontrolled by ultrasound. Eight cysts regressed before delivery and 35cysts after delivery independent of their sonographic appearance. Themean diameter of cysts that required treatment was significantly differ-ent from the mean diameter of cysts that resolved spontaneously(6.8 � SD 2.4 cm vs. 3.3 � SD 0.8 cm; P < 0.01).Conclusion: Complicated cysts which do not regress should be treatedeither by laparotomy or laparoscopically after delivery. Uncomplicatedcysts which exceed 5 cm could be treated by in utero aspiration oraspiration after delivery to avoid further complications. Cysts smallerthan 5 cm, presenting the tendency to regress, should be left untouchedindependent of their sonographic appearance.

F C 1 0 : F E T A L T R E A T M E N T

F70Prediction of outcome of twin twin transfusion syndromeat presentation

R. B. Cincotta, F. Y. Chan, A. Lee-Tannock& D. Amaraddio

Aim: The aim of this study was to determine what factors are impor-tant in prediction of outcome in twin twin transfusion syndrome(TTTS) at presentation.Methods: A retrospective review was performed of all cases of TTTSmanaged at the Mater Mother’s Hospital, Brisbane, Australia from1994 until 2001. TTTS was diagnosed when there was coexisting oli-gohydramnios (MVP < 2 cm) and polyhydramnios (MVP > 8 cm) in amonochorionic pregnancy.

Results: There were 50 cases of TTTS during this time (100 fetuses).Overall, there was a 61% survival rate, with 14% of the fetuses havingan NND, 23% were stillborn and there was one termination of preg-nancy. Both babies survived in 55% of pregnancies, one survived in16% and both twins were lost in 29%. The average gestation at diag-nosis was 21.5 weeks (range 16–29) and the average gestation at deliv-ery was 28.5 weeks (range 16–36). Sixty percent of pregnancies had amean of 2.1 amnioreduction performed, and 2002 � 976 mL wasremoved. The poorest prognostic factor was the presence of hydropsin the recipient twin. Hydrops at presentation resulted in only a 36%survival for either twin in the affected pregnancy. Hydrops resolvedspontaneously in two out of 11 cases and these were the only hydropictwins to survive. At the time of presentation, abnormal Dopplers in therecipient twin were associated with significantly poorer survival than

23–28 October 2001, Melbourne, Australia FC09: Fetal Abdomen

29

for normal Dopplers (27% vs. 64% P ¼ 0.02) but in contrast to otherstudies there was no significant difference seen in outcomes for the pre-sence of abnormal Dopplers in the donor twin. Other individual factorsdid not prove to be good at predicting outcome.Conclusions: The poorest prognostic indicators for survival in TTTSwere the presence of hydrops and abnormal umbilical artery andvenous Dopplers in the recipient. Research into identifying good prog-nostic markers may identify subgroups of TTTS that may benefitfrom more aggressive therapies in order that we might further improveoutcomes.

F71Selective photocoagulation of communicating vesselsin monochorionic twins with selective growthretardation

R. A. Quintero, P. W. Bornick, M. Allen & W. J. Morales

Objective: The purpose of this study was to assess the outcome ofmonochorionic (MC) twins with selective intrauterine growth retarda-tion (SIUGR) treated with selective laser photocoagulation of thecommunicating vessels (SLPCV).Study design: Patients at less than 26 weeks with MC SIUGR pregnan-cies were eligible for the study. SIUGR was defined as <10th percentilefor gestational age. Absent or reverse end-diastolic velocity in theumbilical artery of the SIUGR twin was required for eligibility afterJanuary 2000.Results: Thirty patients met criteria for the study: 17 were managedexpectantly (group I), two underwent umbilical-cord ligation of theSIUGR twin, and 11 underwent SLPCV (group II). Survival rates forat least one fetus were no different between groups I and II (14/17 or82.3% vs. 8/11 or 72.3%, P ¼ 0.4). However, concomitant demise ofthe co-twin occurred in 4/7 patients and iatrogenic premature deliveryfor deterioration of the SIUGR twin was necessary in two patients ingroup, resulting in significant neonatal morbidity. Of the live-bornbabies, neurological handicap was present in 3/22 (13.6%) vs. 0/12in groups I and II, respectively (P < 0.0001).Conclusions: SIUGR in MC twins can be effectively treated withSLPCV. By unlinking the circulations between the fetuses, the preg-nancy is rendered ‘functionally’ dichorionic, improving pregnancymanagement and resulting in decreased neonatal morbidity. Thisapproach constitutes a new valuable alternative in the managementof MC twin pregnancies with SIUGR and an extension of the use ofSLPCV from TTTS to MC SIUGR. A randomized clinical trial ofexpectant management vs. SLPCV for MC SIUGR can be considered.

F72Long-term neurological outcome of survivors of twintwin transfusion syndrome (TTTS)

R. B. Cincotta, P. H. Gray, Y. M. Rogers & F. Y. Chan

Aim: The aim of this study was to examine the long-term neurologicaloutcome of survivors of TTTS.Methods: Survivors of TTTS were followed up in the multidisciplinaryGrowth and Development Unit in the Mater Children’s Hospital. Adetailed neurological examination and a neurosensory motor develop-mental assessment was performed. Cerebral palsy (CP) was diagnosedon the basis of this examination according to standard criteria. AGriffith’s developmental assessment was performed by a psychologist.A general quotient (GQ) more than two standard deviations belowthe mean (GQ < 76) represented global developmental delay(GDD).Results: There were 39 cases of TTTS delivered at the Mater Mother’sHospital, from January 1994 until June 2000. There were 48 survivinginfants (62%), 17 stillbirths (22%), and 13 NND (17%). The meangestation of delivery was 28.5 weeks (range 19–36). Fifty-nine percentof cases were treated with amnioreduction whilst the others were deliv-ered prior to treatment. Of the 48 survivors, detailed neurologicalassessment from 1 to 4 years was available for 35 (73%). An additionalfour twins were known to have normal long-term outcomes but formalneuro-developmental assessment had not been performed and theothers had been lost to follow up. Of those with complete assessment,

29 (83%) were normal. There were three (8.6%) cases with CP andthree (8.6%) with GDD. Of the three cases with CP, one had a co-twinthat had a FDIU, one had a co-twin with GDD and the other had anormal co-twin. The other two cases of GDD had normal co-twins.Conclusion: Survivors of TTTS have a significant risk of major neuro-developmental impairment such as cerebral palsy and global develop-mental delay. Further research is required to determine what factorscontribute to this risk and to develop treatments to minimise theincidence and severity of these handicaps.

F73Accuracy in estimating fetal urinary bladder volumeusing a modified ultrasound technique

M. A. Fagerquist, U. O. Fagerquist, H. Steyskal,A. Oden & S. G. Blomberg

Purpose: The fetal urine production rate can be used as non-invasivetest of fetal well-being. However, the accuracy of measuring urine pro-duction rate by ultrasound depends on the accuracy of volume estima-tion of the certain bladder image. The aim of this study was to assessthe accuracy of measurement of volumes of the bladder using a mod-ified ultrasound technique and digitising the bladder images. Also, itwas possible to compare these results with the corresponding findingsassessed by conventional ultrasound technique.Methods: A swaying movement of the ultrasound transducer causescharacteristic changes on the screen, which can be utilized for identify-ing whether or not the ellipsoid configuration of fetal bladder iscorrectly sectioned along its midline. In this study, 34 unselectedpregnant women participated and the gestational ages of the fetuseswere between 24 and 40 weeks. The assumption was made of a circularappearance of the bladder in the transversal plane and then the calcu-lation of volume was simplified. We distinguished between differentkinds of errors, i.e. technical limitations due to the particular ultra-sound machine, errors due to the operators measurement in theselected image of the bladder and errors due to the selection procedureof appropriate image.Results: The technical limitations and intraoperators errors measuringthe volume of the selected bladder image (the ‘frozen’ error) occupiedonly a minor part of the total variability, whereas the major part con-sisted of errors was owing to the selection of scan, the ‘freezing’ proce-dure. Concerning this dominating part the standard deviation of themeasurement errors of volume estimation using the modified techniquewas SD ¼ 0.37892 þ 0.04119 � volume, e.g. for 5–40 mL was 11.7–5.1%, respectively. Corresponding standard deviation without thismodified technique was SD ¼ 0.36516 þ 0.09978 � volume, e.g. SDfor 5–40 mL was 17.3–10.9%, respectively.Conclusion: When estimating the volume of fetal urinary bladder themajor part of measurement errors due to the selection of appropriatebladder image. Using a modified technique this error can be reduced.

F74In utero release of constricting limb bands: successfultiming of operation with Doppler flow assessments

F.-Y. Chan, R. Cincotta & P. Borzi

Background: Constricting limb bands in utero can lead to vascularcompromise, resulting in limb amputations and/or significant neuralcompression sequele. In utero release has been described to improveoutcome. When constriction bands were identified early in pregnancy,the question as to when operation should be performed remainscontroversial.Purpose: We report two cases, where the vascular status of the limbswas serially assessed and operation performed when Doppler flowassessment showed significant compromise.Methods: Case 1 had insertion of fetal vesico-amniotic shunt per-formed at 17 weeks for bladder outlet obstruction. While the urinarybypass was successful, the free end of the shunt was found to havewrapped round the left fetal thigh. With advancing gestation, the shuntwas shown to progressively indent on the thigh, and Doppler flowassessment showed 37% reduction in flow below the constriction by29 weeks. Case 2 was found to have constricting amniotic bands to

11th World Congress on Ultrasound in Obstetrics and Gynecology FC10: Fetal Treatment

30

the left fetal thigh and right leg at 19 weeks. Serial assessments showedno evidence of worsening until 29 weeks, when Doppler flow showedreduction in flow by 50 and 34%, respectively, on the left and rightlegs. After extensive counseling, fetoscopic release of the constrictionswere performed for the two cases at 30 and 32 weeks, respectively.Results: Both operations were performed with two port uterineaccesses (3 mm) under general anaesthesia. The constrictions wereidentified and divided under direct vision. In case 1, the cut end ofthe shunt retracted into the fetal bladder. Recurrence of bladder outletobstruction with hydronephrosis and anhydramnios necessitated deliv-ery at 32 weeks. Anterior and posterior urethral valves were resected at6 months. Follow up at 2 years of age showed no renal impairment.Case 2 delivered at 36 weeks. Significant limb indentations could beidentified at the constriction sites for both fetuses after birth. Case 1has no neurological deficit of the limbs while case 2 need furthersurgery for talipes and has minor functional impairments.

Conclusions: In utero release of constricting limb bands may reducelong-term sequele. Serial Doppler flow assessments can assist in theappropriate timing of such procedures.

F75Separation of siamese twins at the head (craniopagus):the Brisbane experience

S. F. Campbell, K. G. Jamieson & S. J. Sinnott

Siamese twins have intrigued both the physician and layperson formany centuries. Craniopagus twins (joined at the head) are exceedinglyrare with an incidence of 1/5 million births. Most clinicians never see acase of craniopagus, and those who do rarely see more than one. Theauthor would like to present two cases of craniopagus twins born andseparated recently in Brisbane, Australia. The problems associatedwith separation and outcomes will be discussed.

23–28 October 2001, Melbourne, Australia FC10: Fetal Treatment

31