3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1573
ORIGINAL RESEARCH
Does Early Second-Trimester Sonography Predict Adverse Perinatal Outcomes in Monochorionic Diamniotic Twin Pregnancies? M. Baraa Allaf, MD, Winston A. Campbell, MD, Anthony M. Vintzileos, MD, Sina Haeri, MD, Pouya Javadian, MD, Amir A. Shamshirsaz, MD, Paul Ogburn, MD, Reinaldo Figueroa, MD, Joseph Wax, MD, Glenn Markenson, MD, Martin R. Chavez, MD, Samadh F. Ravangard, DO, Rodrigo Ruano, MD, PhD, Haleh Sangi-Haghpeykar, PhD, Bahram Salmanian, MD, Marjorie Meyer, MD, Jeffery Johnson, MD, Ali Ozhand, MD, PhD, Sarah Davis, MD, Adam Borgida, MD, Michael A. Belfort, MD, PhD, Alireza A. Shamshirsaz, MD Received November 6, 2013, from the Department of Obstetrics and Gynecology, Stony Brook Medicine, Stony Brook, New York USA (M.B.A., P.O., R.F.); Department of Obstetrics and Gynecology, Winthrop University Hospital, Mineola, New York USA (M.B.A., A.M.V., M.R.C.); Department of Obstetrics and Gynecology, University of Connecticut, Farmington, Connecticut USA (W.A.C., S.F.R., Al.A.S.); Department of Obstetrics and Gynecology, Baylor College of Medicine and Texas Children’s Hospital, Houston, Texas USA (S.H., P.J., R.R., H.S.-H., B.S., M.A.B., Al.A.S.); Department of Obstetrics and Gynecology, George Washington University, Washington, DC USA (Am.A.S.); Sidra Medical and Research Center, Doha, Qatar (P.O.); Hartford Hospital, Hartford, Connecticut USA (R.F., A.B.); Department of Obstetrics and Gynecology, Maine Medical Center, Portland, Maine USA (J.W., G.M.); Department of Obstetrics and Gynecology, Baystate Medical Center, Springfield, Massachusetts USA (S.F.R.); Department of Obstetrics and Gynecology, University of Illinois, Chicago, Illinois, USA (S.F.R.); Department of Obstetrics and Gynecology, University of Vermont, Burlington, Vermont USA (M.M., S.D.); Department of Obstetrics and Gynecology, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire USA (J.J.); and Department of Preventive Medicine, University of Southern California, Los Angeles, California USA (A.O.). Revision requested December 30, 2013. Revised manuscript accepted for publication January 10, 2014. Address correspondence to Alireza A. Shamshirsaz, MD, Department of Obstetrics and Gynecology, Baylor College of Medicine Texas Children’s Fetal Center, Texas Children’s Hospital, 6651 Main St, Suite F1020, Houston, TX 77030-2354 USA. E-mail:
[email protected],
[email protected] Abbreviations
AUC, area under the curve; CI, confidence interval; IUFD, intrauterine fetal death; TTTS, twin-twin transfusion syndrome
Objectives—To determine whether intertwin discordant abdominal circumference, femur length, head circumference, and estimated fetal weight sonographic measurements in early second-trimester monochorionic diamniotic twins predict adverse obstetric and neonatal outcomes. Methods—We conducted a multicenter retrospective cohort study involving 9 regional perinatal centers in the United States. We examined the records of all monochorionic diamniotic twin pregnancies with two live fetuses at the 16- to 18-week sonographic examination who had serial follow-up sonography until delivery. The intertwin discordance in abdominal circumference, femur length, head circumference, and estimated fetal weight was calculated as the difference between the two fetuses, expressed as a percentage of the larger using the 16- to 18-week sonographic measurements. An adverse composite obstetric outcome was defined as the occurrence of 1 or more of the following in either fetus: intrauterine growth restriction, twin-twin transfusion syndrome, intrauterine fetal death, abnormal growth discordance (≥20% difference), and very preterm birth at or before 28 weeks. An adverse composite neonatal outcome was defined as the occurrence of 1 or more of the following: respiratory distress syndrome, any stage of intraventricular hemorrhage, 5-minute Apgar score less than 7, necrotizing enterocolitis, culture-proven earlyonset sepsis, and neonatal death. Receiver operating characteristic and logistic regression– with–generalized estimating equation analyses were constructed. Results—Among the 177 monochorionic diamniotic twin pregnancies analyzed, intertwin abdominal circumference and estimated fetal weight discordances were only predictive of adverse composite obstetric outcomes (areas under the curve, 79% and 80%, respectively). Receiver operating characteristic curves showed that intertwin discordances in abdominal circumference, femur length, head circumference, and estimated fetal weight were not acceptable predictors of twin-twin transfusion syndrome or adverse neonatal outcomes. Conclusions—In our cohort, only second-trimester abdominal circumference and estimated fetal weight discordances in monochorionic diamniotic twin pregnancies were predictive of adverse composite obstetric outcomes. Twin-twin transfusion syndrome and adverse neonatal outcomes were not predicted by any of the intertwin discordances measured. Key Words—monochorionic twins; nuchal translucency; obstetric ultrasound; second trimester; twin-twin transfusion syndrome
doi:10.7863/ultra.33.9.1573
©2014 by the American Institute of Ultrasound in Medicine | J Ultrasound Med 2014; 33:1573–1578 | 0278-4297 | www.aium.org
3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1574
Allaf et al—Early Second-Trimester Sonography in Monochorionic Diamniotic Twin Pregnancies
W
eight discordance is defined as a 15% to 25% difference in actual weight between twins by the American College of Obstetricians and Gynecologists.1 In monochorionic diamniotic twin gestations, the placental mass is often not equally shared, which may lead to fetal growth impairment and severely discordant birth weights.2,3 Severely discordant growth, which is commonly observed in twin-twin transfusion syndrome (TTTS), contributes substantially to mortality and morbidity in monochorionic diamniotic twins.4,5 The data regarding first-trimester sonography for predicting adverse perinatal outcomes in monochorionic diamniotic twin gestations predominantly involves biometric discordances.6–11 There are recent reports on early second-trimester sonographic examinations for predicting adverse perinatal outcomes.3,12,13 To our knowledge, only 2 studies reported early second-trimester biometrics with regard to monochorionic diamniotic twin pregnancy outcomes.3,12 The ability to predict adverse obstetric and neonatal outcomes early in the pregnancy is invaluable, as it will afford providers a chance to offer early intervention and address any modifiable risk factors, which may improve outcomes. Consequently, our aim in this study was to determine whether there is any relationship between discordance in biometric measurements (intertwin discordant abdominal circumference, femur length, head circumference, and estimated fetal weight) in monochorionic diamniotic twin gestations and adverse obstetric and neonatal outcomes.
Materials and Methods Between January 2007 and June 2011, this multicenter, retrospective study was conducted at 9 regional perinatal centers in the United States (Baylor College of Medicine, Houston, TX; Baystate Medical Center, Springfield, MA; Dartmouth-Hitchcock Medical Center, Lebanon, NH; Hartford Hospital, Hartford, CT; Maine Medical Center, Portland, ME; Stony Brook University, Stony Brook, NY; University of Connecticut Health Center, Farmington, CT; University of Vermont, Burlington, VT; and Winthrop University Hospital, Mineola, NY). Approval for this study was obtained from the local Institutional Review Board at each participating site before data collection. The obstetric sonographic data from each institution were reviewed for all monochorionic diamniotic twin pregnancies with two live fetuses at the 16- to 18-week sonographic examination that had a documented first trimester sonographic examination at 11 weeks to 13 weeks 6 days as previously reported.8
1574
Maternal-fetal medicine specialists or registered diagnostic medical sonographers were responsible for all sonographic examinations. In addition, all centers had accreditation by the American Institute of Ultrasound in Medicine at the time of the sonographic examinations. The abdominal circumference, femur length, head circumference, and estimated fetal weight measured at 16 to 18 weeks were used to calculate the intertwin difference between the two fetuses and expressed as a percentage of the larger measurement. We excluded all cases with a known chromosomal abnormality or major congenital malformation, pregnancies whose initial second-trimester examinations were later than 18 weeks’ gestation, and pregnancies that did not have follow-up sonographic examinations. Twin-twin transfusion syndrome was defined according to the sonographic criteria and staging of Quintero et al.14 Intrauterine growth restriction was defined as estimated fetal weight below the 10th percentile, and abnormal growth discordance was set at a difference of 20% or greater. Very preterm birth was defined as delivery at or before 28 weeks’ gestation. Chorionicity was confirmed on the basis of a review of the first-trimester sonographic reports using previously published criteria15,16 and placental pathologic reports. An adverse composite obstetric outcome was defined as the occurrence of 1 or more of the following in either fetus: intrauterine growth restriction, TTTS according to the classification of Quintero et al,14 intrauterine fetal death (IUFD), abnormal growth discordance of 20% or greater on follow-up sonography after 18 weeks, and very preterm birth at 28 weeks or earlier. An adverse composite neonatal outcome was defined as the occurrence of 1 or more of the following: respiratory distress syndrome, any stage of intraventricular hemorrhage, 5-minute Apgar score less than 7, necrotizing enterocolitis, culture-proven early-onset sepsis, and neonatal death. Data analysis focused on the assessment of sonographic biometric (abdominal circumference, femur length, head circumference, and estimated fetal weight) discordance to predict an adverse composite obstetric or neonatal outcome. To adjust for clustering in the analysis of neonatal and live-birth outcomes, a logistic regression analysis was performed via generalized estimating equation models using SAS software (SAS Institute Inc, Cary, NC). Sensitivity and specificity values for various cutoff levels were calculated and graphically displayed on receiver operating characteristic curves. Power calculation for the study was based on asymptotic theory and comparison of the areas under the curve (AUCs) of various parameters to minimally acceptable values.17 The available sample sizes
J Ultrasound Med 2014; 33:1573–1578
3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1575
Allaf et al—Early Second-Trimester Sonography in Monochorionic Diamniotic Twin Pregnancies
for the study’s primary outcomes were 54 adverse composite obstetric outcomes (31%) and 123 controls (69%). With these sample sizes and a minimally acceptable AUC of 0.60, our study had at least 80% power, with α set at 5%, to detect a difference of 0.2 (AUC ≥0.8).
Results Among the 254 neonates from the 177 monochorionic diamniotic twin pregnancies included in this study, we excluded 81 from the neonatal analysis because of incomplete data and 19 (from 14 pregnancies) because of IUFD. We found no significant differences in patient demographic data among all of the centers as previously published.8 A total of 154 of the 177 twin pregnancies (87%) underwent cesarean delivery. The mean gestational age ± SD at the time of delivery was 34.5 ± 3.9 weeks. As we previously described, 54 of 177 pregnancies (31%) were complicated by adverse obstetric outcomes. Of those 54 cases, 21 (12%) had intrauterine growth restriction; 19 (11%) had TTTS; 14 (8%) had IUFD; abnormal growth discor-
dance of 20% or greater was seen in 14 (8%); and very preterm delivery at 28 weeks or earlier was seen in 10 (6%).8 Sixty-nine of 254 neonates (27%) had adverse neonatal outcomes. Respiratory distress syndrome was the most common neonatal complication (59 of 254 [23%]), and neonatal death had the lowest incidence (4 of 254 [1.5%]).8 Receiver operating characteristic curves for the prediction of adverse composite obstetric outcomes by discordance in abdominal circumference, head circumference, femur length, and estimated fetal weight are shown in Figure 1. The AUCs were as follows: abdominal circumference, 0.79 (95% confidence interval [CI], 0.70–0.88); head circumference, 0.70 (95% CI, 0.60–0.81); femur length, 0.69 (95% CI, 0.66–0.84); and estimated fetal weight, 0.80 (95% CI, 0.74-0.91). The AUCs for prediction of adverse composite neonatal outcomes by discordance were as follows: abdominal circumference, 0.49 (95% CI, 0.37–0.57); head circumference, 0.54 (95% CI, 0.48–0.63); femur length, 0.56 (95% CI, 0.42–0.62); and estimated fetal weight, 0.53 (95% CI, 0.41–0.67; Figure 2). The AUCs for
Figure 1. Receiver operating characteristic curves for prediction of adverse composite obstetric outcomes by discordance in abdominal circumference (AC), head circumference (HC), femur length (FL), and estimated fetal weight (EFW).
Figure 2. Receiver operating characteristic curves for prediction of adverse composite neonatal outcomes by discordance in abdominal circumference, head circumference, femur length, and estimated fetal weight. Abbreviations are as in Figure 1.
J Ultrasound Med 2014; 33:1573–1578
1575
3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1576
Allaf et al—Early Second-Trimester Sonography in Monochorionic Diamniotic Twin Pregnancies
predicting subsequent TTTS development as a single complication by discordance were as follows: abdominal circumference, 0.65 (95% CI, 0.50–0.75); head circumference, 0.61 (95% CI, 0.46–0.76); femur length, 0.62 (95% CI, 0.43–0.62); and estimated fetal weight, 0.66 (95% CI, 0.58–0.81; Figure 3). The mean degree of discordance for head circumference was 0.047 ± 0.08 (median [range], 0.03 [0–0.67]). The mean degrees of discordance for abdominal circumference and femur length were 0.06 ± 0.09 (median [range], 0.04 [0–0.90]) and 0.06 ± 0.08 (median [range], 0.04 [0–0.90]), respectively. Figure 4 shows the AUCs and 95% CIs for discordance in abdominal circumference, femur length, head circumference, and estimated fetal weight in forest plots for adverse composite obstetric outcomes, composite neonatal outcomes, and TTTS.
Discussion We previously published a report concerning intertwin discordant measurements (nuchal translucency and crown-rump length) on first-trimester sonography for predicting adverse outcomes in monochorionic diamniotic pregnancies.8 Our results showed no predictive value of first-trimester sonographic measurement discordance for adverse obstetric or neonatal outcomes. In this study, we investigated the potential of early second-trimester fetal sonographic measurements to predict adverse obstetric and neonatal outcomes in monochorionic diamniotic twin gestations. Intertwin abdominal circumference and estimated fetal weight discordances were the only parameters that were predictive of adverse composite obstetric outcomes, AUCs of 79% and 80%, respectively. Although our study with 254 neonates from 177 monochorionic diamni-
otic twin pregnancies is one of the largest studies concerning monochorionic diamniotic twins in the second trimester, the exclusion of 81 neonates from the neonatal analysis because of incomplete data and 19 (from 14 pregnancies) because of IUFD may justify the fact that we found no notable biometric discordance predictive with respect to predicting neonatal outcomes and TTTS. In our study, TTTS was defined according to the sonographic criteria and staging of Quintero et al.14 Maximum vertical pools of 2 and 8 cm represent the 5th and 95th percentiles for amniotic fluid measurements, respectively, and the presence of both is used to define stage 1 TTTS. When the difference in amniotic fluid in the two sacs fails to meet these criteria, progression to TTTS occurs in less than 15% of cases.13 The higher risk of perinatal morbidity and mortality in monochorionic diamniotic twins has led to increased monitoring of these pregnancies,12 with many researchers examining the utility of various prediction models. For example, in a prospective cohort of 202 twin pregnancies, Lewi et al3 estimated the prediction of complicated fetal outcomes with identification of risk factors in the first and early second trimesters. They postulated that early second-trimester sonography allows classification of monochorionic twins into high- and low-risk pregnancies and identification of those who are likely to have birth weight discordance.3 Similarly, our findings support the role of estimated fetal weight and abdominal circumference discordances in identifying pregnancies at risk of adverse obstetric outcomes. Fox et al12 demonstrated the potential utility of secondtrimester sonography for identifying “at risk” pregnancies. They analyzed the data of 306 twin pregnancies for dichorionic (n = 271) and monochorionic (n = 35) placentation
Figure 3. Receiver operating characteristic curves for prediction of TTTS as a single complication by discordance in abdominal circumference, head circumference, femur length, and estimated fetal weight. Abbreviations are as in Figure 1.
1576
J Ultrasound Med 2014; 33:1573–1578
3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1577
Allaf et al—Early Second-Trimester Sonography in Monochorionic Diamniotic Twin Pregnancies
separately. In dichorionic pregnancies, they found that increased second-trimester weight discordance and secondtrimester small-for-gestational-age status were both associated with low birth weights, low birth weight percentiles, and increased birth weight discordance. However, in monochorionic pregnancies (n = 35), many of the differences were present but were not statistically significant because they were underpowered for this subgroup analysis. However, our results suggest that in a population of unselected monochorionic diamniotic pregnancies, intertwin abdominal circumference and estimated fetal weight discordances are effective early markers for prediction of adverse obstetrics outcome. Lewi et al3 found that intertwin abdominal circumference discordance at 16 weeks can predict TTTS as a separate complication. However, we found no acceptable predictors of TTTS with regard to receiver operating characteristic curves for abdominal circumference, femur length, head circumference, and estimated fetal weight. This finding could have been due to possible loss of severe cases of TTTS cases presenting as IUFD during the first trimester.
In our study, we excluded 81 neonates from the neonatal analysis because of incomplete data and 19 because of IUFD. All of these factors could have resulted in a lack of association of neonatal complications with regard to receiver operating characteristic curves for abdominal circumference, femur length, head circumference, and estimated fetal weight. Despite the limitations of study, it is one of the largest studied monochorionic diamniotic twin cohorts with respect to the utility of second-trimester sonography in predicting adverse obstetric and neonatal outcomes. Our results indicate that early second-trimester abdominal circumference and estimated fetal weight discordances may be associated with an increased risk of subsequent adverse obstetric outcomes. This information may be helpful for early detection of complications and may offer an opportunity for timely intervention that could improve outcomes. Further research is necessary to identify markers associated with adverse outcomes and determine the ideal surveillance strategy for monochorionic diamniotic twin pregnancies.
Figure 4. Forest plots for adverse composite obstetric outcomes, composite neonatal outcomes, and TTTS with AUCs and 95% CIs for discordance in abdominal circumference, femur length, estimated fetal weight, and head circumference. Abbreviations are as in Figure 1.
J Ultrasound Med 2014; 33:1573–1578
1577
3309jum1531-1668online_Layout 1 8/21/14 8:33 AM Page 1578
Allaf et al—Early Second-Trimester Sonography in Monochorionic Diamniotic Twin Pregnancies
References 1.
2.
3.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
1578
American College of Obstetricians and Gynecologists. ACOG Practice Bulletin No. 56: multiple gestation: complicated twin, triplet and higher order multifetal pregnancy. Obstet Gynecol2004; 104:869–883 (reaffirmed 2009). Fick AL, Feldstein VA, Norton ME, Wassel Fyr C, Caughey AB, Machin GA. Unequal placental sharing and birth weight discordance in monochorionic diamniotic twins. Am J Obstet Gynecol 2006; 195:178–183. Lewi L, Cannie M, Blickstein I, et al. Placental sharing, birthweight discordance and vascular anastomoses in monochorionic diamniotic twin placentas. Am J Obstet Gynecol 2007; 197:587.e1–587.e8. Adegbite AL, Castille S, Ward S, Bajoria R. Neuromorbidity in preterm twins in relation to chorionicity and discordant birth weight. Am J Obstet Gynecol 2004; 190:156–163. Gratacós E, Carreras E, Becker J, et al. Prevalence of neurological damage in monochorionic twins with selective intrauterine growth restriction and intermittent absent or reversed end-diastolic umbilical artery flow. Ultrasound Obstet Gynecol 2004; 24:159–163. Fratelli N, Prefumo F, Fichera A, Valcamonico A, Marella D, Frusca T. Nuchal translucency thickness and crown rump length discordance for the prediction of outcome in monochorionic diamniotic pregnancies. Early Hum Dev 2011; 87:27–30. Kagan KO, Gazzoni A, Sepulveda-Gonzalez G, Sotiriadis A, Nicolaides KH. Discordance in nuchal translucency thickness in the prediction of severe twin-to-twin transfusion syndrome. Ultrasound Obstet Gynecol 2007; 29:527–532. Allaf MB, Vintzileos AM, Chavez MR, et al. First-trimester sonographic prediction of obstetric and neonatal outcomes in monochorionic diamniotic twin pregnancies. J Ultrasound Med 2014; 33:135–140. Nicolaides KH, Heath V, Cicero S. Increased fetal nuchal translucency at 11–14 weeks. Prenat Diagn 2002; 22:308–315. Johansen ML, Oldenburg A, Rosthøj S, Maxild JC, Rode L, Tabor A. Crown-rump length discordance in first trimester: a predictor of adverse outcome in twin pregnancies? Ultrasound Obstet Gynecol 2014; 43:245– 246. O’Connor C, McAuliffe FM, Breathnach FM, et al; Perinatal Ireland Research Consortium. Prediction of outcome in twin pregnancy with first and early second trimester ultrasound. J Matern Fetal Neonatal Med 2013; 26:1030–1035. Fox NS, Saltzman DH, Schwartz R, Roman AS, Klauser CK, Rebarber A. Second-trimester estimated fetal weight and discordance in twin pregnancies: association with fetal growth restriction. J Ultrasound Med 2011; 30:1095–1101. Society for Maternal-Fetal Medicine Publication Committee. SMFM Clinical Guideline: twin-twin transfusion syndrome. Am J Obstet Gynecol 2013; 208:3–18. Quintero RA, Morales WJ, Allen MH, Bornick PW, Johnson PK, Kruger M. Staging of twin-twin transfusion syndrome. J Perinatol 1999; 19:550– 555.
15. Dias T, Arcangeli T, Bhide A, Napolitano R, Mahsud-Dornan S, Thilaganathan B. First-trimester ultrasound determination of chorionicity in twin pregnancy. Ultrasound Obstet Gynecol 2011; 38:530–532. 16. Sepulveda W, Sebire NJ, Nicolaides KH. The lambda sign in twin pregnancies. Ultrasound Obstet Gynecol 1996; 8:429. 17. Pepe MS. Sample size based on the AUC. In: The Statistical Evaluation of Medical Tests for Classification and Prediction. Oxford, England: Oxford University Press; 2003:224–227.
J Ultrasound Med 2014; 33:1573–1578