, . 17:4: 319–322 (1997)

UMBILICAL DOPPLER VELOCIMETRY INFETUSES WITH TRISOMY 18 AT 10–18 WEEKS’

GESTATION

. . , . , . , . , . *, . † .

Prenatal Diagnosis Unit, Department of Obstetrics and Gynaecology; *Biochemistry Service, Hospital Clinic;†Biostatistics Unit, University of Barcelona, Barcelona, Spain

Received 14 February 1996Revised 29 July 1996

Accepted 13 August 1996

SUMMARY

The aim of our study was to obtain measurements of the umbilical artery pulsatility index (PI) in pregnanciesbefore invasive procedures for prenatal diagnosis, in order to investigate its potential prognostic value in predictingtrisomy 18. We performed a prospective study including 1785 consecutive women from 10 to 18 weeks with singletonpregnancies undergoing chorionic villus sampling (n=559) or genetic amniocentesis (n=1226) in our unit. Dopplermeasurements were performed transvaginally (tenth to 13th week of gestation) or transabdominally (14th to 18thweek of gestation) immediately before the invasive procedure. In 7 out of 10 fetuses subsequently diagnosed astrisomy 18, the PI was above the 95th centile, providing a detection rate of 70 per cent, a specificity of 95·1 per cent,a positive predictive value of 7·7 per cent, and a negative predictive value of 99·8 per cent. When the 90th percentilewas assayed as a cut-off, the efficacy of PI as a marker of trisomy 18 yielded a sensitivity of 90 per cent and aspecificity of 90·4 per cent, with a positive predictive value of 5·2 per cent and a negative predictive value of 99·9 percent. We suggest that although the use of a single PI measurement for screening purposes needs to be confirmed byfurther investigation, trisomy 18 fetuses show an abnormal increase in umbilical PI in the first half of pregnancy, andits relation to the early onset of fetal growth retardation needs to be further explored.? 1997 by John Wiley & Sons,Ltd.

: Doppler velocimetry, trisomy 18, prenatal diagnosis

INTRODUCTION

It is well known that pregnancies with abnormalumbilical Doppler velocimetry, particularly whenabsent or reversed flow velocities are detected, areassociated with growth-retarded infants and smallplacentae, and that genetic disorders may affectfetal growth and placentation (Giles et al., 1985;Kaufmann et al., 1979; Jauniaux et al., 1992).Among the more common autosomal aneu-ploidies, trisomy 18 fetuses have been reported asthe ones in which fetal growth is more severelyaffected (Drugan et al., 1992; Khun et al., 1995).Preliminary data also suggest that trisomy 18fetuses may have reversed end-diastolic umbilical

artery velocities in early pregnancy (Martinezet al., 1996a).On this basis, the aim of our study was to obtain

measurements of the umbilical artery pulsatilityindex (PI) in pregnancies before invasive proce-dures for prenatal diagnosis, in order to investi-gate its potential prognostic value in predictingtrisomy 18.

MATERIAL AND METHODS

We performed a prospective study including1785 consecutive women with singleton preg-nancies undergoing chorionic villus sampling(CVS) (n=559) or genetic amniocentesis (n=1226)in our institution. The indications for the invasiveprocedure were the risk of fetal aneuploidybecause of advanced maternal age (>37 years), a

Correspondence to: Josep M. Martinez Crespo, Avda Repú-blica Argentina 263, Atico 2a, Barcelona 08023, Spain.

CCC 0197–3851/97/040319–04 $17.50? 1997 by John Wiley & Sons, Ltd.

positive biochemical screening in maternal serumfor Down’s syndrome (>1/270), previous offspringwith chromosomal defects, and risk of Mendeliandisorders. Gestational age ranged from 10 to 18weeks on the day of the procedure, as determinedby ultrasonographic measurement of the crown–rump length or the biparietal diameter.Doppler measurements were performed trans-

vaginally (10th to 13th week of gestation) ortransabdominally (14th to 18th week of gestation)immediately before the invasive procedure. A com-mercially available Hitachi EUB 515-A real-timeultrasound scanner, with colour Doppler, wasused. The vaginal probe used a frequency of6·5 MHz and the abdominal probe a frequency of3·5 MHz. The cut-off level of the high-pass filterwas set at 50 Hz, and energy output levels werelower than 50 mW/cm2 spatial peak-temporalaverage. The colour Doppler maximal velocitysetting was adjusted so that the umbilical arterywas homogeneous in colour, did not demonstratealiasing, and filled the lumen of the artery. Thestudy was approved by the hospital ethical com-mittee. For calculating the umbilical artery PI, theDoppler sample volume was placed on a freefloating loop of the umbilical cord. Velocity wave-forms were recorded three times and the peaksystolic velocity, end-diastolic velocity (whenpresent), and mean maximum velocity weremeasured. Five to seven consecutive waveformswere analysed each time and the results averaged,using a conventional microcomputerized programlinked to the equipment. Hard copies of eachrecording were encoded on a Sony printer device.This technique, as well as its reproducibility, hasbeen previously reported and validated (Martinezet al., 1995).Standard statistical analysis with SPSS-WIN for

PC software was used to evaluate the efficacy, interms of sensitivity, specificity, and predictivevalues, for umbilical PI as a marker of trisomy 18,analysed for each week of gestation.

RESULTS

Satisfactory velocity waveform recordings wereobtained in all cases. Maternal age showed abimodal distribution depending on the procedureto be performed (CVS: 37·9 years, SD 4·2; amnio-centesis: 34·9 years, SD 4·1). A similar distributionwas found for gestational age (CVS: 11·2 weeks,SD 0·9; amniocentesis: 16·1, SD 0·9).

In order to assess the efficacy of umbilical veloci-metry in trisomy 18 fetuses, we retrospectivelyexcluded other chromosomal abnormalities,including 25 cases of trisomy 21, and fetal lossesafter the procedure. Therefore, a chromosomallynormal group of 1698 pregnancies continuingbeyond the 22nd week of gestation was defined asthe control group. Complete cytogenetic and ultra-sonic follow-up to 4 weeks post-procedure wasavailable in all cases.Ten fetuses with trisomy 18 were diagnosed,

and in 7 out of 10 the PI was over the 95thpercentile, yielding a detection rate of 70 percent, whereas 1615 out of 1698 chromosomallynormal fetuses had a PI below the 95th percen-tile, showing a specificity of 95·1 per cent, apositive predictive value of 7·7 per cent (7/90),and a negative predictive value of 99·8 per cent(1615/1618). When we used the 90th percentile asa cut-off, the efficacy of PI as a marker oftrisomy 18 produced a sensitivity of 90 per cent(9/10) and a specificity of 90·4 per cent (1536/1698), with a positive predictive value of 5·2 percent (9/171) and a negative predictive value of99·9 per cent (1536/1537). Furthermore, areversed end-diastolic flow was detected in onlytwo of the 1785 cases, both affected by trisomy18.Table I and Fig. 1 show umbilical artery PI

values in the fetuses affected by trisomy 18. TableII summarizes these data and shows the efficacy ofPI, using the 90th and 95th percentiles as twodifferent cut-offs.

Table I—Umbilical artery pulsatility index (PI) andpercentile in which it is situated in cases of trisomy 18

CaseNo.

Ga(weeks) PI Centile

1 10 3·64 >952 10 2·77 <90*3 10 3·56 >954 11 3·08 >955 12 2·92 >956 12 2·72 90–957 12 3·60 >958 14 2·46 >959 15 2·08 >9510 16 1·90 90–95

*Not detected.GA=gestational age.

320 . . .

DISCUSSION

Umbilical artery Doppler velocimetry measure-ments provide indirect and reliable information onthe perfusion of the fetoplacental circulation andhave been shown to be reproducible even duringthe first half of the pregnancy (Huisman et al.,1993; Martinez et al., 1995). Flow velocity wave-forms from the umbilical artery are characterizedby the absence of end-diastolic velocities up to the13th week. From the 14th week onwards, end-diastolic blood flow in the umbilical artery alwaysbegins to appear in uncomplicated pregnancies(Fisk et al., 1988), representing the downstreamvascular impedance to blood flow at the level ofthe placenta. As a result of angiogenesis, thenumber of villi and vessels increases with gesta-tional age, whereas placental vascular resistancegradually decreases, with a relative increase inthe diastolic component of the umbilical flow(Giles et al., 1985; Jauniaux et al., 1992).Trisomy 18 pregnancies have been associated

with fetal growth retardation, even during the first

trimester, whereas this is not so in fetuses withother trisomies (Khun et al., 1995). During thethird trimester of pregnancy, the finding of anabsent or reversed end-diastolic velocity in theumbilical artery has proven to be a good predictorof an adverse fetal outcome in ‘at-risk’ populationsand has been associated with a higher incidence ofchromosomal abnormalities (Rizzo et al., 1994;Nicolaides et al., 1988). Although the use ofDoppler in the first half of pregnancy is not wellestablished, and remains in the research field, itmay be a valuable tool to investigate certainfirst-trimester abnormalities. In a previous study of924 singleton pregnancies undergoing fetal karyo-typing at 10–18 weeks’ gestation, we foundan abnormal upwards trend in umbilical artery PIin trisomic fetuses at this stage of pregnancy(Martinez et al., 1996b).We can only speculate on the underlying cause

of the abnormal Doppler velocimetric values intrisomy 18 fetuses. Fetuses with chromosomalanomalies often have signs of early functional andmorphological developmental disorders, reflectedby a high rate of miscarriage and an early delay inintrauterine growth (Drugan et al., 1992; Khunet al., 1995; Fitzsimmons et al., 1990; Benacerrafet al., 1987).The abnormal Doppler pattern in such chromo-

somally abnormal pregnancies can be related toinadequate placentation and dislocation of thetrophoblastic shell, which supports the finding thatembryonic and placental development are closelyrelated in early pregnancy (Giles et al., 1985;Jauniaux et al., 1992). Previous findings in thesecond half of pregnancy suggest that pregnanciescomplicated by growth retardation have a signifi-cant reduction in the number of arterioles in thetertiary stem villi (Bracero et al., 1989; Locci et al.,1993; Jauniaux et al., 1992). In such cases, patho-logical Doppler patterns, such as absent or

Fig. 1—Umbilical artery pulsatility index in trisomy 18 fetuses(-)

Table II—Efficacy of umbilical artery PI in detecting trisomy 18, using the 90th and 95thpercentiles as two different cut-offs

Sensitivity(%)

Specificity(%)

PPV(%)

NPV(%)

A 7/10 (70) 1615/1698 (95·1) 7/90 (7·8) 1615/1618 (99·8)B 9/10 (90) 1536/1698 (90·4) 9/171 (5·2) 1536/1537 (99·9)

A: umbilical artery PI§95th percentile.B: umbilical artery PI§90th percentile.

321 18

reversed diastolic blood flow, are more frequentlyencounterd and could explain our findings ofelevated umbilical artery PI in fetuses affected bytrisomy 18 earlier in pregnancy.The use of a single measurement for screening

purposes needs further investigation, and becausethis study was performed in women at high risk ofaneuploidy, these results cannot be extrapolated toa low-risk population. Sensitivity and specificityare stable properties of the test and thus shouldremain constant among different populations butthe predictive values are a function of prevalenceand may be considerably altered. Therefore,prospective studies in an unselected population arerequired.

REFERENCES

Benacerraf, B.R., Gelman, R., Frigoletto, F.D. (1987).Sonographic identification of second-trimesterfetuses with Down’s syndrome, N. Engl. J. Med., 317,1371–1376.

Bracero, L.A., Beneck, D., Kirshenbaum, N., Peiffer,M., Stalter, P., Schulman, H. (1989). Doppler veloci-metry and placental disease, Am. J. Obstet. Gynecol.,161, 388–393.

Drugan, A., Johnson, M.P., Isada, N.B., Holzgreve, W.,Zador, I., Dombrowski, M.P., Sokol, R.J., Hallak,M., Evans, M.I. (1992). The smaller than expectedfirst-trimester fetus is at increased risk for chromo-some anomalies, Am. J. Obstet. Gynecol., 167, 1525–1528.

Fisk, N.M., Maclachlan, N., Ellis, C., Tannirandorn,Y., Tonge, H.M., Rodeck, C.H. (1988). Absentend-diastolic flow in first trimester umbilical artery,Lancet, ii, 1256–1257.

Fitzsimmons, J., Droste, S., Shepard, T.H., Pascoe-Manson, J., Fantel, A. (1990). Growth failure insecond-trimester fetuses with trisomy 21, Teratology,42, 337–345.

Giles, W.B., Trudinger, B.J., Baird, P.J. (1985). Fetalumbilical artery velocity waveforms and placentalresistance: pathological correlation, Br. J. Obstet.Gynaecol., 87, 780–785.

Huisman, T.W.A., Stewart, P.A., Stijnen, T.,Wladimiroff, J.W. (1993). Doppler flow velocity wave-forms in late first- and second-trimester fetuses: repro-ducibility of waveform recordings, Ultrasound Obstet.Gynecol., 3, 260–263.

Jauniaux, E., Jurkovic, D., Campbell, S., Hustin, J.(1992). Doppler ultrasonographic features of thedeveloping placental circulation: correlation with ana-tomic findings, Am. J. Obstet. Gynecol., 166, 585–587.

Kaufmann, P., Sen, D.K., Schweikhart, G. (1979).Classification of human placental villi, Cell TissueRes., 200, 409–423.

Khun, P., Brizot, M.L., Pandya, P.P., Snijders, R.J.,Nicolaides, K.H. (1995). Crown–rump length inchromosomally abnormal fetuses at 10 to 13 weeks’gestation, Am. J. Obstet. Gynecol., 172, 32–35.

Locci, M., Nazzaro, G., De Placido, G., Nazzaro, A.,Colacurci, N., Montagnani, S., Montemagno, U.(1993). Correlation of Doppler and placentalimmunohistochemical features in normal and intra-uterine growth-retarded fetuses, Ultrasound Obstet.Gynecol., 3, 240–245.

Martinez, J.M., Comas, C., Ojuel, J., Borrell, A.,Puerto, B., Sentis, J., Fortuny, A. (1995). The influ-ence of the site of Doppler recording on umbilicalartery pulsatility index during first trimester, Ultra-sound Obstet. Gynecol., 5, 325–327.

Martinez, J.M., Comas, C., Borrell, A., Puerto, B.,Antolin, E., Ojuel, J., Fortuny, A. (1996a). Reversedend-diastolic umbilical artery velocity in two cases oftrisomy 18 at 10 weeks; gestation. Ultrasound Obstet.Gynecol., 7, 447–449.

Martinez, J.M., Comas, C., Ojuel, J., Puerto, B., Borrell,A., Fortuny, A. (1996b). Umbilical artery pulsatilityindex in early pregnancies with chromosomeanomalies, Br. J. Obstet. Gynaecol, 103, 330–334.

Nicolaides, K.H., Bilardo, C.M., Soothill, P.W.,Campbell, S. (1988). Absence of end-diastolic fre-quencies in umbilical artery: a sign of fetal hypoxiaand acidosis, Br. Med. J., 297, 1026–1027.

Rizzo, G., Pietropoli, A., Capponi, A., Arduini, D.,Romanini, C. (1994). Chromosomal abnormalities infetuses with absent end-diastolic velocity in umbilicalartery: analysis of risk factors for an abnormal karyo-type, Am. J. Obstet. Gynecol., 171, 827–831.

322 . . .