DOI: 10.1002/pd.4634
ORIGINAL ARTICLE
Ultrasonographic prediction of birth weight discordance in twin pregnancies M. K. van de Waarsenburg1*, K. E. A. Hack1, R. J. Rijpma1, E. J. H. Mulder1, L. Pistorius2 and J. B. Derks1 1
Division of Perinatology and Gynecology, University Medical Centre Utrecht, Utrecht, The Netherlands Division of Perinatology and Gynecology, Tygerberg Hospital, Cape Town, South Africa *Correspondence to: M. K. van de Waarsenburg. E-mail: [email protected]
2
ABSTRACT Objective This study aimed to assess the accuracy of the ultrasonographic prediction of intertwin birth weight discordance based on crown–rump length, estimated fetal weight (EFW) and abdominal circumference. Method We retrospectively studied variables of prenatal surveillance of 281 twin pregnancies at the University Medical Centre Utrecht, between 2008 and 2011.
Results Fifteen per cent of twins had a birth weight discordance of ≥20%. One twin more commonly had intrauterine growth restriction in the birth weight discordance group compared with the concordantly grown group (69% vs 20%, respectively). Sonographically, EFW discrepancy and increased abdominal circumference ratio were significantly correlated with birth weight discordance. The negative predictive values were high (92% and 87%, respectively). The best prediction of weight discordance was given by EFW discordance at the last ultrasound before delivery. Conclusion Intertwin birth weight discordance is best predicted by an intertwin EFW discordance at the last ultrasound. If discordance is not suspected by ultrasound, the vast majority had no birth weight discordance. In case of birth weight discordance, the presence of at least one growth-restricted twin is very likely. However, because around 20% of concordantly grown twin pairs comprise at least one intrauterine growth restriction infant, one cannot rely on weight discordance alone. © 2015 John Wiley & Sons, Ltd.
Funding sources: None Conflicts of interest: None declared
INTRODUCTION Birth weight (BW) discordance of ≥20% occurs in about 16% of twin pregnancies, irrespective of chorionicity.1 Differences in weight can be explained by genetic factors (in case of dizygosity) or the influence of sex hormones in utero (in case of unlike-sex pairs)2 but are mostly due to unequal sharing of the monochorionic (MC) placenta or to a difference in size of the two dichorionic (DC) placentas.3 Furthermore, in MC placentas, the vascular anastomoses with subsequent intertwin transfusion may influence fetal growth. Growth discordance in twin gestations is considered to be a major risk for perinatal death and morbidity.4,5 It is associated with an increase of anomalies, intrauterine growth restriction (IUGR), preterm birth, infection, stillbirth, asphyxia and admission to a neonatal intensive care unit.6 Because BW discordance is associated with adverse outcome, prenatal recognition is important in order to plan appropriate prenatal monitoring, timing and mode of delivery and to counsel the prospective parents. Regular ultrasound assessments are used to measure fetal biometry in order to estimate the fetal weight and weight discordance. Assessing Prenatal Diagnosis 2015, 35, 906–912
the degree of intertwin weight discordance can be performed by calculating either the estimated fetal weight (EFW), from measurements of head circumference (HC), abdominal circumference (AC) and femur length (FL)7 or the ratio between both AC measurements within a pair8 or intertwin crown–rump length (CRL) discrepancy at the first-trimester ultrasound.9,10 The aim of this study was to assess the accuracy of the various sonographical estimations of size or weight discordance at various gestational ages (GAs) in the prediction of severe weight discordance at birth.
METHODS We retrospectively studied parameters of prenatal surveillance and outcome of 329 women with a twin pregnancy, who had their prenatal care and delivery between 2008 and 2011 at the University Medical Centre Utrecht, The Netherlands, a hospital with tertiary care and a neonatal intensive care unit. The twin pregnancies were routinely followed by experienced sonographers. © 2015 John Wiley & Sons, Ltd.
Prediction of discordance in twins
Cases with one or more of the following criteria were excluded, leaving 281 pregnancies to be studied. Monoamniotic twin pregnancies (n = 6), pregnancies with a selective feticide (n = 3), pregnancies complicated by congenital disorders (n = 24) or intrauterine fetal death (n = 18), pregnancies with a GA at birth of less than 22 weeks (n = 3) or fetuses with a BW less than 500 g (n = 6) were excluded as well as cases in which it was impossible to calculate a BW discordance (n = 4).11,12 Parameters of prenatal surveillance (biometry and Doppler assessments) were retrieved from an electronic ultrasound database (MOSOS-U version 11.0, Houten, The Netherlands). Twin gestations were monitored according to a protocol based on chorionicity. This protocol included a first-trimester determination of chorionicity,13,14 a detailed anomaly scan at 20 weeks’ GA and ultrasound assessment of growth and amniotic fluid volume at 20, 26, 30, 32, 34 and 36 weeks for DC twin gestations and fortnightly from 14 weeks onwards, including Doppler assessments of the umbilical artery (UA) and middle cerebral artery, for MC twin gestations. In DC twins, Doppler imaging of the UA was only performed in case of growth abnormalities of one or both twins. In case of abnormal Doppler findings, fetal surveillance was intensified, varying from daily to twice weekly evaluations that were performed during hospitalization or during visits to an outpatient clinic. Pulsatility index (PI) of the UA was determined from at least three reproducible waveforms. Doppler findings were considered abnormal if the mean PI calculated was more than 2 standard deviations (SD) above the mean, with or without absent or reversed end-diastolic flow.15 Oligohydramnios was defined as the deepest vertical pocket of amniotic fluid of less than 2 cm. A velamentous insertion of the umbilical cord is a condition in which the umbilical cord inserts into the chorioamniotic membranes rather than the placental mass. The EFW was calculated by a formula of Hadlock based on the HC, AC and FL measurements.7 Both CRL and EFW difference were calculated as the difference between the size or weight of the larger and smaller twins, divided by the size or weight of the larger twin. The AC ratio was calculated as the AC of the larger twin divided by the AC of the smaller twin. We considered severe size or weight discordance if the difference in CRL or EFW was 20% or more. Similarly, an AC ratio of at least 1.2 was used to indicate a significant difference between both fetuses. For comparison with literature, an AC ratio of 1.38 and a CRL discordance of 11%16 were also considered. Weight discordance at birth was calculated as the intertwin difference in BW expressed as a percentage of the heaviest twin; a value of ≥20% was defined as severe BW discordance. Early neonatal mortality was defined as the death of an infant during the first 7 days of postnatal life.17 Major neonatal morbidity was defined as the presence of any of the following complications: respiratory distress syndrome, intraventricular haemorrhage, periventricular leukomalacia, necrotizing enterocolitis or proven sepsis.18 Adverse outcome was defined as the occurrence of major neonatal morbidity and/or neonatal death. IUGR was defined as an EFW below the 10th percentile based on the last ultrasound before birth.19 Any Prenatal Diagnosis 2015, 35, 906–912
907
abnormal Doppler finding during the course of pregnancy was considered to be abnormal in these gestations. All data were stored in a database and analysed using the SPSS 20.0 statistical package (Chicago, IL, USA). Continuous variables were compared using independent Student’s t-tests if normally distributed or using Mann–Whitney U-tests if not. Statistical significance was based on two-sided tests, with p < 0.05 considered significant. Logistic regression was first performed in a univariate fashion to analyse predictive factors of BW discordance. We performed multivariate logistic regression with all parameters that were significant at a p-value of 0.60 were rated as substantial, and kappa values >0.80 are almost perfect. The correlation coefficient was shown by Spearman’s rho showing statistical dependence with a perfect correlation of +1 or 1. A receiver operating characteristic curve, plotting the sensitivity and false-positive rate of each diagnostic test, was constructed to determine the optimal sonographic prediction of BW discordance. The accuracy of these tests was measured by the area under the curve (AUC), with an AUC of 1 representing the perfect test and an AUC of 0.5 being a worthless test.
RESULTS Mean maternal age of the study population was 32.9 years. The median interval between the last ultrasound and delivery was 8 days (range 0–59 days). Median GA at delivery was 35+0 weeks (range 23+3–41+0 weeks) and three quarters of the twin pregnancies were DC. Median BW discordance was 10.6% (range: 0.0–60.6%). Out of the 281 included pregnancies, 42 twin pairs (14.9%) showed a BW discordance of ≥20%. Median EFW discordance at the last ultrasound was 7.3% (range 0.0–61.3%), median AC ratio at the last ultrasound was 1.0 (range 1.0–1.5) and median CRL discordance at first-trimester scan was 4.6% (range 0–40.2%). Baseline characteristics of the study population are shown in Table 1. At least one twin more commonly had IUGR in the growth-discordant group than in the appropriately grown twin group (69% vs 20%, respectively, p < 0.01), irrespective of the presence of Doppler abnormalities. IUGR of both infants within the twin pair occurred more frequently in twin pregnancies with discordant compared with concordant growth (16.7% vs 8.9%). This difference was however not statistically significant (p = 0.12). In twin pairs with severe BW discordance (≥20%), the majority also had significant EFW discordance (57.1% vs 6.3%, p < 0.01). There was also a significant difference in the occurrence of BW discordance when the AC ratio was ≥1.2 (16.7% vs 0.8%, p < 0.01). A velamentous umbilical cord insertion occurred significantly more often in severely discordant twins (23.1% vs 3.8%, p < 0.01). © 2015 John Wiley & Sons, Ltd.
M. K. van de Waarsenburg et al.
908
Table 1 Baseline characteristics of twin pairs with and without severe birth weight discordance Birth weight discordance < 20% (n = 239)
Variables
Birth weight discordance ≥ 20% (n = 42)
p-value
Univariate OR (95% CI)
Maternal Maternal age (years), mean (SD) Age ≥ 30 years, n (%)
33.0 (4.5) 185 (77.4%)
32.7 (4.7)
0.78
1.1 (0.4–1.7)
31 (73.8%)
0.61
0.8 (0.4-1.7)
Parity, median (range)
1 (0–6)
0 (0–4)
0.81
1.0 (0.7–1.5)
Nulliparity, n (%)
115 (48.1%)
22 (52.5%)
0.61
1.2 (0.6–2.3)
23.5 (18–52)
24.6 (20–56)
0.06
0.9 (0.9–1.0)
2 (11.8%)
0.53
1.7 (0.3–8.8)
176 (73.6%)
30 (71.4%)
0.77
1.1 (0.5–2.3)
63 (26.4%)
12 (28.6%)
3 (1.3%)
3 (7.1%)
0.02
6.0 (1.2–31.0)
2
BMI (kg/m ), median (range) BMI ≥ 35, n (%)
8 (7.3%)
Chorionicity, n (%) Dichorionic Monochorionic diamniotic Pre-existing hypertension, n (%) Gestational age at delivery (weeks), median (range)
36+6 (23+3–41+0)
34+5 (26+5–39+0)
0.16
1.1 (1.0–1.2)
4.4 (0.0–40.2)
5.1 (0.4–20.4)
0.13
0.9 (0.9–1.0)
0.25
3.8 (0.3–44.1)
Ultrasound CRL discordance (%), median (range) ≥20% discordance, n (%) ≥11% discordance, n (%) EFW discordance last US (%), median (range) EFW discordance last US ≥ 20%, n (%) AC ratio discordance last US, median (range)
2 (1.3%) 11 (7.1%) 6.2 (0.0–28.7) 15 (6.3%) 1.0 (1.0–1.32)
1 (4.8%) 4 (19.0%) 22.0 (1.1–61.3)
0.07
3.1 (0.9–10.8)
ORIGINAL ARTICLE
Ultrasonographic prediction of birth weight discordance in twin pregnancies M. K. van de Waarsenburg1*, K. E. A. Hack1, R. J. Rijpma1, E. J. H. Mulder1, L. Pistorius2 and J. B. Derks1 1
Division of Perinatology and Gynecology, University Medical Centre Utrecht, Utrecht, The Netherlands Division of Perinatology and Gynecology, Tygerberg Hospital, Cape Town, South Africa *Correspondence to: M. K. van de Waarsenburg. E-mail: [email protected]
2
ABSTRACT Objective This study aimed to assess the accuracy of the ultrasonographic prediction of intertwin birth weight discordance based on crown–rump length, estimated fetal weight (EFW) and abdominal circumference. Method We retrospectively studied variables of prenatal surveillance of 281 twin pregnancies at the University Medical Centre Utrecht, between 2008 and 2011.
Results Fifteen per cent of twins had a birth weight discordance of ≥20%. One twin more commonly had intrauterine growth restriction in the birth weight discordance group compared with the concordantly grown group (69% vs 20%, respectively). Sonographically, EFW discrepancy and increased abdominal circumference ratio were significantly correlated with birth weight discordance. The negative predictive values were high (92% and 87%, respectively). The best prediction of weight discordance was given by EFW discordance at the last ultrasound before delivery. Conclusion Intertwin birth weight discordance is best predicted by an intertwin EFW discordance at the last ultrasound. If discordance is not suspected by ultrasound, the vast majority had no birth weight discordance. In case of birth weight discordance, the presence of at least one growth-restricted twin is very likely. However, because around 20% of concordantly grown twin pairs comprise at least one intrauterine growth restriction infant, one cannot rely on weight discordance alone. © 2015 John Wiley & Sons, Ltd.
Funding sources: None Conflicts of interest: None declared
INTRODUCTION Birth weight (BW) discordance of ≥20% occurs in about 16% of twin pregnancies, irrespective of chorionicity.1 Differences in weight can be explained by genetic factors (in case of dizygosity) or the influence of sex hormones in utero (in case of unlike-sex pairs)2 but are mostly due to unequal sharing of the monochorionic (MC) placenta or to a difference in size of the two dichorionic (DC) placentas.3 Furthermore, in MC placentas, the vascular anastomoses with subsequent intertwin transfusion may influence fetal growth. Growth discordance in twin gestations is considered to be a major risk for perinatal death and morbidity.4,5 It is associated with an increase of anomalies, intrauterine growth restriction (IUGR), preterm birth, infection, stillbirth, asphyxia and admission to a neonatal intensive care unit.6 Because BW discordance is associated with adverse outcome, prenatal recognition is important in order to plan appropriate prenatal monitoring, timing and mode of delivery and to counsel the prospective parents. Regular ultrasound assessments are used to measure fetal biometry in order to estimate the fetal weight and weight discordance. Assessing Prenatal Diagnosis 2015, 35, 906–912
the degree of intertwin weight discordance can be performed by calculating either the estimated fetal weight (EFW), from measurements of head circumference (HC), abdominal circumference (AC) and femur length (FL)7 or the ratio between both AC measurements within a pair8 or intertwin crown–rump length (CRL) discrepancy at the first-trimester ultrasound.9,10 The aim of this study was to assess the accuracy of the various sonographical estimations of size or weight discordance at various gestational ages (GAs) in the prediction of severe weight discordance at birth.
METHODS We retrospectively studied parameters of prenatal surveillance and outcome of 329 women with a twin pregnancy, who had their prenatal care and delivery between 2008 and 2011 at the University Medical Centre Utrecht, The Netherlands, a hospital with tertiary care and a neonatal intensive care unit. The twin pregnancies were routinely followed by experienced sonographers. © 2015 John Wiley & Sons, Ltd.
Prediction of discordance in twins
Cases with one or more of the following criteria were excluded, leaving 281 pregnancies to be studied. Monoamniotic twin pregnancies (n = 6), pregnancies with a selective feticide (n = 3), pregnancies complicated by congenital disorders (n = 24) or intrauterine fetal death (n = 18), pregnancies with a GA at birth of less than 22 weeks (n = 3) or fetuses with a BW less than 500 g (n = 6) were excluded as well as cases in which it was impossible to calculate a BW discordance (n = 4).11,12 Parameters of prenatal surveillance (biometry and Doppler assessments) were retrieved from an electronic ultrasound database (MOSOS-U version 11.0, Houten, The Netherlands). Twin gestations were monitored according to a protocol based on chorionicity. This protocol included a first-trimester determination of chorionicity,13,14 a detailed anomaly scan at 20 weeks’ GA and ultrasound assessment of growth and amniotic fluid volume at 20, 26, 30, 32, 34 and 36 weeks for DC twin gestations and fortnightly from 14 weeks onwards, including Doppler assessments of the umbilical artery (UA) and middle cerebral artery, for MC twin gestations. In DC twins, Doppler imaging of the UA was only performed in case of growth abnormalities of one or both twins. In case of abnormal Doppler findings, fetal surveillance was intensified, varying from daily to twice weekly evaluations that were performed during hospitalization or during visits to an outpatient clinic. Pulsatility index (PI) of the UA was determined from at least three reproducible waveforms. Doppler findings were considered abnormal if the mean PI calculated was more than 2 standard deviations (SD) above the mean, with or without absent or reversed end-diastolic flow.15 Oligohydramnios was defined as the deepest vertical pocket of amniotic fluid of less than 2 cm. A velamentous insertion of the umbilical cord is a condition in which the umbilical cord inserts into the chorioamniotic membranes rather than the placental mass. The EFW was calculated by a formula of Hadlock based on the HC, AC and FL measurements.7 Both CRL and EFW difference were calculated as the difference between the size or weight of the larger and smaller twins, divided by the size or weight of the larger twin. The AC ratio was calculated as the AC of the larger twin divided by the AC of the smaller twin. We considered severe size or weight discordance if the difference in CRL or EFW was 20% or more. Similarly, an AC ratio of at least 1.2 was used to indicate a significant difference between both fetuses. For comparison with literature, an AC ratio of 1.38 and a CRL discordance of 11%16 were also considered. Weight discordance at birth was calculated as the intertwin difference in BW expressed as a percentage of the heaviest twin; a value of ≥20% was defined as severe BW discordance. Early neonatal mortality was defined as the death of an infant during the first 7 days of postnatal life.17 Major neonatal morbidity was defined as the presence of any of the following complications: respiratory distress syndrome, intraventricular haemorrhage, periventricular leukomalacia, necrotizing enterocolitis or proven sepsis.18 Adverse outcome was defined as the occurrence of major neonatal morbidity and/or neonatal death. IUGR was defined as an EFW below the 10th percentile based on the last ultrasound before birth.19 Any Prenatal Diagnosis 2015, 35, 906–912
907
abnormal Doppler finding during the course of pregnancy was considered to be abnormal in these gestations. All data were stored in a database and analysed using the SPSS 20.0 statistical package (Chicago, IL, USA). Continuous variables were compared using independent Student’s t-tests if normally distributed or using Mann–Whitney U-tests if not. Statistical significance was based on two-sided tests, with p < 0.05 considered significant. Logistic regression was first performed in a univariate fashion to analyse predictive factors of BW discordance. We performed multivariate logistic regression with all parameters that were significant at a p-value of 0.60 were rated as substantial, and kappa values >0.80 are almost perfect. The correlation coefficient was shown by Spearman’s rho showing statistical dependence with a perfect correlation of +1 or 1. A receiver operating characteristic curve, plotting the sensitivity and false-positive rate of each diagnostic test, was constructed to determine the optimal sonographic prediction of BW discordance. The accuracy of these tests was measured by the area under the curve (AUC), with an AUC of 1 representing the perfect test and an AUC of 0.5 being a worthless test.
RESULTS Mean maternal age of the study population was 32.9 years. The median interval between the last ultrasound and delivery was 8 days (range 0–59 days). Median GA at delivery was 35+0 weeks (range 23+3–41+0 weeks) and three quarters of the twin pregnancies were DC. Median BW discordance was 10.6% (range: 0.0–60.6%). Out of the 281 included pregnancies, 42 twin pairs (14.9%) showed a BW discordance of ≥20%. Median EFW discordance at the last ultrasound was 7.3% (range 0.0–61.3%), median AC ratio at the last ultrasound was 1.0 (range 1.0–1.5) and median CRL discordance at first-trimester scan was 4.6% (range 0–40.2%). Baseline characteristics of the study population are shown in Table 1. At least one twin more commonly had IUGR in the growth-discordant group than in the appropriately grown twin group (69% vs 20%, respectively, p < 0.01), irrespective of the presence of Doppler abnormalities. IUGR of both infants within the twin pair occurred more frequently in twin pregnancies with discordant compared with concordant growth (16.7% vs 8.9%). This difference was however not statistically significant (p = 0.12). In twin pairs with severe BW discordance (≥20%), the majority also had significant EFW discordance (57.1% vs 6.3%, p < 0.01). There was also a significant difference in the occurrence of BW discordance when the AC ratio was ≥1.2 (16.7% vs 0.8%, p < 0.01). A velamentous umbilical cord insertion occurred significantly more often in severely discordant twins (23.1% vs 3.8%, p < 0.01). © 2015 John Wiley & Sons, Ltd.
M. K. van de Waarsenburg et al.
908
Table 1 Baseline characteristics of twin pairs with and without severe birth weight discordance Birth weight discordance < 20% (n = 239)
Variables
Birth weight discordance ≥ 20% (n = 42)
p-value
Univariate OR (95% CI)
Maternal Maternal age (years), mean (SD) Age ≥ 30 years, n (%)
33.0 (4.5) 185 (77.4%)
32.7 (4.7)
0.78
1.1 (0.4–1.7)
31 (73.8%)
0.61
0.8 (0.4-1.7)
Parity, median (range)
1 (0–6)
0 (0–4)
0.81
1.0 (0.7–1.5)
Nulliparity, n (%)
115 (48.1%)
22 (52.5%)
0.61
1.2 (0.6–2.3)
23.5 (18–52)
24.6 (20–56)
0.06
0.9 (0.9–1.0)
2 (11.8%)
0.53
1.7 (0.3–8.8)
176 (73.6%)
30 (71.4%)
0.77
1.1 (0.5–2.3)
63 (26.4%)
12 (28.6%)
3 (1.3%)
3 (7.1%)
0.02
6.0 (1.2–31.0)
2
BMI (kg/m ), median (range) BMI ≥ 35, n (%)
8 (7.3%)
Chorionicity, n (%) Dichorionic Monochorionic diamniotic Pre-existing hypertension, n (%) Gestational age at delivery (weeks), median (range)
36+6 (23+3–41+0)
34+5 (26+5–39+0)
0.16
1.1 (1.0–1.2)
4.4 (0.0–40.2)
5.1 (0.4–20.4)
0.13
0.9 (0.9–1.0)
0.25
3.8 (0.3–44.1)
Ultrasound CRL discordance (%), median (range) ≥20% discordance, n (%) ≥11% discordance, n (%) EFW discordance last US (%), median (range) EFW discordance last US ≥ 20%, n (%) AC ratio discordance last US, median (range)
2 (1.3%) 11 (7.1%) 6.2 (0.0–28.7) 15 (6.3%) 1.0 (1.0–1.32)
1 (4.8%) 4 (19.0%) 22.0 (1.1–61.3)
0.07
3.1 (0.9–10.8)
