Can the degree of retrograde diastolic flow in abnormal umbilical artery flow velocity waveforms predict pregnancy outcome?

Can the degree of retrograde diastolic flow in abnormal umbilical artery flow velocity waveforms predict pregnancy outcome?
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  ORIGINAL PAPER 229 Ultrasound Obstet Gynecol 2002; 19 : 229–234 BlackwellScienceLtd Can the degree of retrograde diastolic flow in abnormal umbilical artery flow velocity waveforms predict pregnancy outcome?  J. BRODSZKI * , E. HERNANDEZ-ANDRADE * , S. GUDMUNDSSON†, M. DUBIEL‡, G. P. MANDRUZZATO§, R. LAURINI¶ and K. MAR S ÁL * Departments of Obstetrics and Gynecology, University Hospitals of * Lund and † Malmö, Sweden, ‡ Department of Perinatology, University School of Medical Sciences, University Hospital Poznan, Poland, § Department of Obstetrics and Gynecology, Instituto per l’Infanzia Burlo Garofolo, Trieste, Italy and ¶ Institut Universitaire de Pathologie, Lausanne, Switzerland  KEYWORDS : Perinatal outcome, Reverse diastolic flow, Umbilical artery ABSTRACT Objective Reverse end-diastolic flow is the most patholo- gical type of the umbilical artery flow velocity waveform. Weaimed to investigate whether additional prognostic informationcan be obtained from umbilical artery waveforms in cases withreverse end-diastolic flow. Subjects and methods Umbilical artery Doppler velocitywaveforms from 44 fetuses with reverse end-diastolic flowwere analyzed and the following parameters measured:the highest amplitude and the area below the maximumvelocity curve of forward and reverse flow (A, B and C, D,respectively) and the duration of forward and reverseflow (Tc and Td, respectively). Ratios A/B, C/D and Tc/Td were calculated. The cut-off values for A/B, C/D and Tc/Td with the best predictive values for perinatal death wereestablished with the help of receiver operating charac-teristics curves. The three curves were compared with eachother.  Results Of the three ratios, A/B and C/D had the best capacityto predict perinatal death. Both ratios had acceptable sens-itivities, specificities and positive predictive values. In thisregard, A/B and C/D were comparable. The cut-off values forA/B and C/D were 4.3 and 4.52, respectively. Survivors had significantly higher A/B and C/D ratios than non-survivors(  P = 0.0001 and 0.0003, respectively). Significantly morefetuses with A/B or C/D below the established cut-off valueshad pulsations in the venous system (  P < 0.05). In fetuseswith a gestational age ≤  210 gestational days the survival ratewas significantly higher in those with A/B or C/D above thecut-off values (  P = 0.03 and 0.003, respectively). Conclusions The A/B or C/D ratio can be used for quanti-fication of the reverse end-diastolic flow waveforms in theumbilical artery and may offer additional information to theevaluation of fetal condition. INTRODUCTION Doppler velocimetry of the umbilical artery provides a non-invasive tool to assess fetoplacental hemodynamics and has,since its introduction into clinical practice, gained an importantrole in monitoring high-risk pregnancies with a placental causeof pregnancy complication, i.e. intrauterine growth restriction(IUGR) 1 . Abnormal Doppler velocimetry in the umbilical arteryis initially characterized by an increased pulsatility index. Anelevated pulsatility index suggests an increase in uteroplacentalresistance and precedes the total absence of end-diastolic flow(AED flow) 2  or even reversal of end-diastolic flow (RED flow).Numerous studies have demonstrated that AED and REDflows represent an ominous sign of late fetal compromise withincreased perinatal mortality and morbidity 3–9 . Although bothAED and RED flows represent extreme pathology as regardsthe umbilical artery flow velocity waveform, clinical featuresand perinatal outcome are largely different if AED flow or REDflow is observed. In the latter case the prevalence of long-termneurological damage and handicap in survivors is very high 4,10,11 .In such fetuses, active management with prompt delivery,depending on the local limits of viability, has been advocatedin order to prevent long-term sequelae 4,9,12 . Zelop et al  . 8 , onthe other hand, reported that gestational age is the dominantdeterminant of perinatal outcome in cases with RED flow atgestational ages below 28 weeks and suggested that interven-tion on behalf of the fetus should be undertaken with caution.In their opinion, efforts should be made to optimize theintrauterine environment rather than resorting to delivery.The timing of delivery from initial diagnosis of RED flow(immediately, within hours or within days) at early gestationalages remains a problem where complications of prematuritymust be weighed against the benefit of avoiding death in utero .We aimed to investigate whether additional information canbe obtained from the umbilical artery Doppler waveforms incases with RED flow which would be of help in clinical decision-making and predicting outcome of pregnancy. Correspondence: Dr J. Brodszki, Department of Obstetrics and Gynecology, University Hospital Lund, SE-221 85 Lund, Sweden (e-mail: Accepted 16-1-01  Abnormal umbilical artery velocity waveformsBrodszki et al. 230 Ultrasound in Obstetrics and Gynecology MATERIALS AND METHODS Umbilical artery Doppler velocity waveforms from 56 fetuseswith RED flow were collected from four European centers:the Departments of Obstetrics and Gynecology in Malmö(Sweden), Lund (Sweden), Trieste (Italy) and Poznan (Poland).All cases that occurred at the centers during a specific timeperiod were taken into consideration. Two fetuses withchromosomal and structural abnormalities were excluded.Of the remaining 54 waveforms, 44 records were consideredsatisfactory for analysis: four from Lund, 21 from Malmö,four from Poznan and 15 from Trieste. The criteria for con-sidering a Doppler velocity waveform as eligible were absenceof fetal breathing movements and presence of at least fiveuniform beats on the registered waveform.Doppler measurements were performed transabdominallyin a free loop of the umbilical cord. All centers but one (Lund)used the same type of equipment: an Acuson 128 XP (Acuson,Mountain View, CA, USA) duplex scanner fitted with 3.5-and 5.0-MHz transducers with power and color Doppleroptions. The high pass filter was set at 125 Hz. In Lund, thescanner used was an HDI 5000 (ATL, Bothell, WA, USA) witha 5.0-MHz transducer.Each waveform record was scanned with Adobe PhotoshopProgram 4.0 (Adobe Systems Inc., San Jose, CA, USA) andthen imported into the NIH Image 1.55 analysis program(NIH, Bethesda, MD, USA) in digitalized form. In each record,five beats were analyzed and the following parameters meas-ured: the highest amplitude and the area of forward and reverseflow (A, B and C, D, respectively) and the duration of forwardand reverse flow (Tc and Td, respectively) (Figure 1). RatiosA/B, C/D and Tc/Td were calculated and the mean valuesfor each record obtained. All offline measurements wereperformed by one of the authors (J.B.).The cut-off values for A/B, C/D and Tc/Td with the bestsensitivity, specificity and positive and negative predictivevalues (PPV and NPV) for perinatal death were established withthe help of receiver operating characteristics (ROC) curves(Figure 2). The three ROC curves were compared with eachother. The odds ratio (OR) with 95% confidence interval(CI) was calculated for all waveform ratios and perinatal death.Of the 44 fetuses, 19 died in utero  and seven in the neonatalperiod. Thirty-five fetuses were growth-restricted (data notavailable in three cases). The mean gestational age at deliverywas 225 (range, 179–273) days in survivors and 195 (range,159–238) days in non-survivors. Twenty-four fetuses weredelivered by Cesarean section and 20 were delivered vaginally.Twenty-two fetuses were delivered within 24 h from diagnosisof retrograde blood flow in the umbilical artery. In theremaining 22 cases the mean interval from diagnosis of REDflow to delivery was 7 (range, 2–26) days. Primary indicationsfor delivery were fetal distress (defined as presence of repetitive late decelerations), severe IUGR, intrauterine fetaldeath, abruptio placentae and pre-eclampsia refractory toantihypertensive treatment. In most cases several of thesecomplications were present. In eight cases the criteria fordelivery included the finding of RED flow.Complete clinical data on the mothers and postnatal data(for survivors) were available in 41/44 and 16/18 cases,respectively. Additional Doppler examinations consisted of assessment of blood flow in the umbilical vein in the free loopof the umbilical cord and assessment of resistance in the uterinearteries, and were performed in 25/44 cases. Survivors and non-survivors were compared with regard to the mean birth weight,weight deviation, gestational age, placental weight, pH inumbilical artery and vein, velocimetry variables, presence of bilateral notching in the uterine arteries and presence of pulsations in the fetal venous system. Pulsations were definedas a reduction in blood velocity of > 15% from the basal mean.Survival rate was compared for two subgroups accordingto gestational age ( ≤  210 days and > 210 days). Survivorswere divided into two subgroups according to the values of umbilical artery waveform parameters and compared forpostnatal complications and clinical characteristics of themothers.Continuous variables were evaluated using the Mann–Whitney U  -test. Categorical data were analyzed using Fisher’s exacttest or the χ 2  test, where appropriate. A significant differencewas considered present if P < 0.05. RESULTS The descriptive variables were compared between survivorsand non-survivors (Table 1). Non-survivors had a significantlylower gestational age at delivery ( P = 0.0005) and were smaller( P < 0.0001). There was a significant difference in the degreeof growth restriction between the two groups ( P = 0.0008).Placental weight in non-survivors was significantly lower( P = 0.0068). There were no differences in presence of notchingin the uterine arteries or in infant gender. The seven fetusesthat died postnatally had lower Apgar scores at 1 and 5 min( P < 0.0001) and a lower pH in the umbilical vein ( P = 0.04). ABCDTcTd Figure 1 Variables of the umbilical artery waveform with retrograde diastolic flow. A and B: the highest amplitude of the forward and reverse flow, respectively; C and D: area of the forward and reverse flow, respectively; Tc and Td: duration of the forward and reverse flow, respectively.  Ultrasound in Obstetrics and Gynecology 231 Abnormal umbilical artery velocity waveformsBrodszki et al. Mothers of survivors and non-survivors were compared forclinical characteristics: smoking and presence of IUGR, pre-eclampsia in combination with IUGR, notching in uterine arteriesand diabetes, with no differences being found (Table 2).The cut-off point for the C/D ratio with the best values of sensitivity, specificity, and PPV for prediction of perinataldeath was 4.52 according to the ROC curve (Figure 2). Theoptimal cut-off points for A/B and Tc/Td were 4.30 and 1.56,respectively. The values of sensitivity, specificity and PPV aregiven in Table 3.Survivors had significantly higher A/B and C/D ratios thannon-survivors ( P = 0.0003 and 0.0001, respectively) (Table 4).The Tc/Td ratio did not differ between the two groups.Of the three ROC curves, the C/D curve had the largestarea under the ROC curve, thus the best predicting capacity.When the three ROC curves were compared statistically, therewas a difference between curves C/D and Tc/Td ( P = 0.009),but no differences were found between curves A/B and Tc/Td,or A/B and C/D. Of the three ratios (A/B, C/D and Tc/Td)C/D had the best capacity to predict perinatal death. The riskof perinatal death was 13.6 times higher in fetuses with aratio of C/D ≤  4.52 (95% CI, 2.99–61.59).Fetuses with a gestational age ≤  210 gestational days(Group I) or > 210 gestational days (Group II) were separatelycompared for survival rate in relation to the C/D ratio. InGroup I, the survival rate of fetuses with a C/D ratio > 4.52was significantly higher than in fetuses with a C/D ratio ≤  4.52 ( P = 0.003). In Group II there was no significantdifference in the survival rate ( P = 0.29).Survivors were divided into two subgroups based on theC/D ratio ( ≤  4.52 and > 4.52) and compared for postnatalcomplications (Table 5). There were no significant differences.Significantly more fetuses with C/D ≤  4.52 had pulsations inthe venous system ( P < 0.05). DISCUSSION Reverse end-diastolic flow represents the extreme end of thespectrum of changes that occur in the process of increasingresistance to flow in the placenta and progressive deteriorationof the compromised fetus. Abnormal umbilical artery wave-forms are most often seen in pregnancies with IUGR and/orpre-eclampsia, i.e. in pregnancy complications caused byabnormal placentation.The initial increase in impedance to umbilical flow isfollowed by arterial redistribution in the fetal circulation 13–15 .A further increase in placental resistance may lead to reversalof end-diastolic velocities in the umbilical artery and finallyto fetal death. As long as the fetus has normal venous flow, thissuggests continuing fetal circulatory compensation. Abnormalvenous flow, on the other hand, indicates a breakdown of compensatory hemodynamic mechanisms 16 . An increase inplacental resistance and the occurrence of arterial redistributionthrough peripheral vasoconstriction leads to an increase inthe right ventricular afterload, i.e. an increase in end-diastolicpressure. This in turn may lead to abnormal flow patternsin the ductus venosus and successively to umbilical venouspulsations 17 . An abnormal ductus venosus blood velocity wave-form with decreased end-systolic and end-diastolic velocities    S  e  n  s   i   t   i  v   i   t  y   (   %   ) 100806040200010080604020 False-positive rate (%)C/D = 4.52(b)(a)(c)    S  e  n  s   i   t   i  v   i   t  y   (   %   ) 100806040200100806040200010080604020 False-positive rate (%)A/B = 4.3    S  e  n  s   i   t   i  v   i   t  y   (   %   ) 010080604020 False-positive rate (%)Tc/Td = 1.56 Figure 2 Receiver operating characteristics curves for ratios A/B (a), C/D (b) and Tc/Td (c) with regard to the prediction of perinatal death. Arrows denote the optimal cut-off points.  Abnormal umbilical artery velocity waveformsBrodszki et al. 232 Ultrasound in Obstetrics and Gynecology is more frequently recorded in high-risk pregnancies, buthas not proved to be a sensitive indicator of adverse perinataloutcome 18 . However, it has been demonstrated that absentor reverse flow in the ductus venosus during atrial contrac-tion indicates a poor prognosis 17,19 .Pulsations in the umbilical venous flow are usually relatedto adverse outcome of pregnancy 20 . An association betweenpulsatile flow in the umbilical vein and increased neonatalmortality has been shown in a study of 37 fetuses with AEDflow: neonatal mortality in the group with pulsations in theumbilical vein was 63% compared to 19% in fetuses withoutpulsations 21 . Whether umbilical vein pulsations are observedonly intra-abdominally or both in the intra-abdominal partand in the free loop of the umbilical cord has implicationson the outcome. Based on a prospective study in 83 high-riskpregnancies, Hofstaetter et al. 22  concluded that pulsationsrecorded only in the intra-abdominal part of the umbilical veinmight be an earlier indicator of worsening fetal condition, thanare pulsations extending into the cord. Consequently, fetuseswith the former type of pulsations have a more favorableoutcome. Our observation that, with one exception, all non-survivors had pulsations in the venous system, specifically Tae 1 Outcome in survivors vs. non-survivors CharacteristicSurvivorsNon-survivors Pn Mean (SD)MedianRange n Mean (SD)MedianRange Gestational age at delivery (days)18225 (26)233179–27326195 (21)196159–238 0.0005Birth weight (g)181590 (998)1285604–403023638 (289)570195–1520< 0.0001Birth-weight deviation* (%)18–27 (26)–27–62 to 2223–50 (19)–54–73 to 17 0.0008Small-for-gestational age†13———22——— 0.048Placental weight (g)15320 (161)240150–67014196 (116)165100–510 0.0068pH umbilical artery97.22 (0.11)7.277.07–7.3647.11 (0.16)7.146.91–7.25—pH umbilical vein127.29 (0.09)7.297.11–7.4237.13 (0.13)7.17.02–7.27—Apgar at 1 min18—71–1026—00–7Intrauterine death————19————*Deviation from the expected gestational age related birth weight. †Data missing in three non-survivors. SD, standard deviation. Tae 2 Cinica caracteristics o te moters CharacteristicSurvivors (  n  (%))Non-survivors (  n  (%)) Smoker1/17 (6)1/20 (5)IUGR9/18 (50)10/23 (43)Pre-eclampsia + IUGR4/18 (22)12/25 (48)Diabetes3/18 (17)0/25 (0)Bilateral notch in uterine artery velocity waveforms*7/16 (44)10/13(77)*Uterine artery Doppler velocimetry not performed in two survivors. IUGR, intrauterine growth restriction. Table 3 Prediction of perinatal death by waveform variables according to optimal cut-off points determined by receiver operating characteristics (ROC) curves VariableRatioA/BC/DTc/Td  Sensitivity (%)697358Specificity (%)838367Positive predictive value (%)868679Area under ROC curve (%)828472A/B, highest amplitude of forward flow/highest amplitude of reverse flow; C/D, area of forward flow/area of reverse flow; Tc/Td, duration of forward flow/duration of reverse flow. For explanation of velocimetry variables see Figure 1. Tae 4 Dopper veocimetry variae ratios in survivors vs. non-survivors RatioSurvivors(  n  = 18)Non-survivors(  n  = 26) PA/BMean (SD)6.0 (2.19)3.7 (1.4)0.0003Median5.43.2Range2.7–10.01.8–7.6C/DMean (SD)8.7 (5.0)4.2 (2.3)0.0001Median7.63.1Range3.3–19.41.5–10.2Tc/TdMean (SD)2.1 (0.9)1.4 (0.1)NSMedian1.71.3Range1.1–4.10.8–2.4A/B, highest amplitude of forward flow/highest amplitude of reverse flow; C/D, area of forward flow/area of reverse flow; Tc/Td, duration of forward flow/duration of reverse flow. For explanation of velocimetry variables see Figure 1. Tae 5 Neonata moriity in survivors accoring to te C/D ratio o umbilical artery velocity waveform C/D ≤  4.52(  n  )C/D > 4.52 (  n  ) Respiratory distress36Cerebral hemorrhage12Necrotizing enterocolitis10Hypoglycemia25Retinopathy of prematurity01Neonatal seizures11Thrombocytopenia13Total313Data missing in two survivors. Some of the neonates had more than one complication.  Ultrasound in Obstetrics and Gynecology 233 Abnormal umbilical artery velocity waveformsBrodszki et al. in the umbilical vein in the free loop of the umbilical cord,supports their conclusion.The finding of umbilical artery RED flow in the late third-trimester fetuses will without doubt lead to immediate delivery.The problematic fetuses are the late second-trimester and earlythird-trimester fetuses, in which complications of prematuritymust be weighed against the benefit of avoiding fetal death in utero . Each day gained is of importance for the prognosisof the fetus.Reverse flow alone does not seem to be the most sensitiveindicator of imminent fetal death in these cases, since varioustime spans from diagnosis of reverse flow to delivery withsurvival have been reported in the literature. Zelop et al. 8 reported a time interval of 3.6 ±  1.5 days with a perinatalmortality of 33.3%, Brar and Platt 3  reported 4.2 ±  1.4 daysand 50% perinatal mortality and Wang et al. 9  reported 10 ±  8.2 hand 50% neonatal mortality. The mean gestational age atdelivery was 29.1 ±  0.6, 30.1 ±  2.5 and 31.8 ±  3.2 gestationalweeks, respectively. Kurkinen-Raty et al. 5  reported a medianof 1 (range, 1–21) day with a 33.3% perinatal mortality anda median gestational age of 31.2 ±  2.4 gestational weeks.Zelop et al. 8  had a maximum time from diagnosis to deliverywith survival of 18 days. In contrast, the group of Wang et al. 9  delivered within hours, yet had the highest neonatalmortality (50%). One may speculate that these fetuses weremore compromised than those of Zelop et al.  In none of these studies was venous Doppler examination performed.The time interval between the occurrence of AED frequenciesand antepartum late heart rate decelerations was evaluatedin a study by Arduini et al  . 21 . The time interval variedconsiderably among fetuses, and was mainly determined bygestational age, the presence of venous pulsations and maternalhypertension.As in previous reports 4,8 , we could confirm that the non-survivors were significantly younger, smaller and had a moresevere degree of growth restriction than the survivors. Whensurvivors and non-survivors in our study were comparedfor maternal disease, such as pre-eclampsia or diabetes, or thepresence of bilateral notching in uterine arteries, no differenceswere found. Neither were there differences found betweenthe two groups regarding smoking habits. Gestational age,degree of growth restriction and venous pulsations seemed todetermine the outcome in fetuses with RED flow.There were only three survivors with a C/D ratio ≤  4.52(Figure 3). Two were fetuses of diabetic mothers. In the thirdcase the mother felt fewer fetal movements in gestational week34+6. These three survivors were much older and largerthan all the other fetuses with a C/D ratio ≤  4.52. Similarly,there were three survivors with an A/B ratio ≤  4.3 (Figure 4).Two of these survivors were the same as in the group with aC/D ratio ≤  4.52: one of the two fetuses of the diabetic mothersand the fetus with reduced fetal movements. The third survivorwas a fetus of a mother who developed pre-eclampsia.Alterations of the venous flow velocity waveforms mightbe in a closer temporal relationship to immediate intrauterinefetal jeopardy than retrograde flow itself in the late second-trimester fetus. Retrograde flow is without doubt an ominoussign, but so far no quantification of ‘severity’ of retrogradeflow has been used. It would seem obvious that the larger theretrograde flow in relation to the forward flow, the morecompromised the fetus will be. Doppler examination of thefetal venous flow, i.e. in the umbilical vein intra-abdominallyand in the free loop of the umbilical cord, in combinationwith the quantification of the retrograde flow waveformsmight be helpful in assessing the immediate fetal jeopardyand in the timing of the delivery. The C/D ratio or the A/B ratiocan be used for quantification of the retrograde flow waveforms.Both have comparable values of sensitivity, specificity andPPV. From a practical point of view, the A/B ratio is easier toobtain than the C/D ratio and does not require scanning of thewaveforms and importing into a separate data program, althoughstatistically, the C/D ratio had the better predicting capacity.    C   /   D 20181614121086420280140160180200220260240Gestational age (days) Figure 3 C/D ratio plotted against gestational age.  , neonatal death;  , survived;  , intrauterine death.    A   /   B 1110987654321280140160180200220260240Gestational age (days) Figure 4 A/B ratio plotted against gestational age.  , neonatal death;  , survived;  , intrauterine death.
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!