Objective
We sought to determine whether twin gestations with an anomalous fetus are at increased risk of preterm delivery or intrauterine growth restriction (IUGR) compared to twins with 2 normal fetuses.
Study Design
This was a retrospective cohort of twins undergoing ultrasound 15-22 weeks’ gestation. Groups were defined by the presence of 1 twin with a major anomaly (discordant) or by twins with no major anomalies (normal). The primary outcomes were preterm delivery (<37 weeks) and IUGR (<10th percentile).
Results
Of 1977 twin pregnancies, 66 had a twin with a major anomaly. Preterm delivery occurred in 42 (63.6%) discordant twins, compared to 1271 (66.5%) normal twins (risk ratio, 1.0; 95% confidence interval, 0.8–1.2). IUGR was diagnosed in 15 (22.7%) normal co-twins, compared to 406 (21.3%) presenting twins in normal twins (risk ratio, 1.1; 95% confidence interval, 0.7–1.7).
Conclusion
Twins discordant for major anomalies are not at increased risk of preterm delivery or IUGR compared to twins with no major anomalies.
Twin gestations are at increased risk for preterm delivery, intrauterine growth restriction (IUGR), and structural anomalies. The presence of an anomaly in 1 twin not only impacts the survival and outcome of the affected twin, but potentially of the normal co-twin as well. Structural anomalies are known to increase the risk of preterm delivery and IUGR in singleton gestations. Prior studies have demonstrated that preterm delivery is increased with cleft lip/palate, genitourinary anomalies, congenital diaphragmatic hernia, neural tube defects, heart defects, gastroschisis, and omphalocele. The concern therefore is that the presence of 1 anomalous fetus will adversely impact the outcome of the normal co-twin by exacerbating the risk of preterm delivery and growth restriction. This possibility presents both provider and parents with the dilemma of how to manage a pregnancy where twins are discordant for structural anomalies.
Prior studies are conflicting as to whether the presence of 1 anomalous fetus impacts the risk of preterm delivery or growth restriction. We therefore sought to estimate the risk of preterm delivery and growth restriction of the normal co-twin in pregnancies complicated by twins discordant for structural anomalies.
Materials and Methods
This is a retrospective cohort study of all patients who underwent routine second-trimester (15-22 weeks) ultrasound for anatomic survey at a single, tertiary care center. Institutional review board approval was obtained. Data were collected prospectively by dedicated nurses from 1990 through 2008. Each patient undergoing ultrasound in our center receives a standardized handout requesting information regarding pregnancy complications, delivery complications, and neonatal outcomes, to be filled out and returned after delivery. The coordinator called the patient, and in cases where the patient could not be reached, the physician, if the form was not returned within 4 weeks of the delivery date.
Patients were included in this study if they carried a twin gestation; singleton gestations, intrauterine fetal demise, and higher order multiple gestations were excluded. Twin pregnancies complicated by monoamnionicity (due to its rarity), twin-twin transfusion syndrome, structural anomalies in both the fetuses, and pregnancies resulting in selective reduction were also excluded. Gestational age was determined by last menstrual period if known and concordant with ultrasound (within 7 days of first-trimester ultrasound or 14 days of second-trimester ultrasound) or by the earliest ultrasound when the last menstrual period was unknown or discordant. Chorionicity was determined at the earliest ultrasound available. First-trimester diagnosis of chorionicity was based on the number of gestational sacs, amnions, and yolk sacs present or the presence of a lambda sign. Second-trimester determination of chorionicity was based on gender discordance, presence of 2 placental masses, and characteristics of the intertwin dividing membrane (twin peak sign, T-sign, thickness of membrane). Final diagnosis of chorionicity was determined by an attending dedicated to obstetric ultrasound.
An anomaly was defined as a defect in the structure of an organ that resulted from a specific primary abnormality of development. Major anomalies were considered defects of organogenesis ; examples of major anomalies include neural tube defects, congenital heart defects, abdominal wall defects, and renal agenesis. Markers of aneuploidy, such as nuchal thickness, echogenic intracardiac focus, and hypoplastic nasal bone, were not considered structural anomalies. The exposure group was defined as a twin pregnancy with at least 1 major anomaly in 1 fetus and an anatomically normal co-twin. The unexposed group was defined as a twin pregnancy where neither twin had a major anomaly.
The primary outcomes of this study were the incidence of preterm delivery and IUGR. Preterm delivery was defined as delivery <37 weeks and 0 days’ gestation. Because complications of prematurity are more common and severe at earlier gestations, secondary analyses of preterm delivery <34 and <28 weeks were performed. IUGR was defined as birthweight <10th percentile for gestational week at delivery based on the Alexander growth standard. To compare growth restriction between groups, the normal co-twin of the exposed group was compared to the presenting twin of the unexposed group. Additionally, secondary analyses of IUGR defined as birthweight <5th percentile for gestational age were performed. Some suggestion has been made that twin outcomes vary by presentation ; therefore, the IUGR analysis was also performed controlling for presentation and comparing the normal co-twin to the nonpresenting twin.
Because prior studies examining the impact of fetal anomalies on the risk of preterm delivery demonstrate a differential effect based on the type of anomaly, a subanalysis was performed to examine the impact of types of anomalies on preterm delivery. Anomalies were considered as high risk when a prior study associated it with an odds ratio >2. High-risk anomalies considered in the subanalysis were central nervous system malformations, urogenital malformations, heart defects, and body wall defects.
The incidence and types of anomalies in the exposed group were described. Subjects with 1 anomalous twin were compared to subjects with 2 normal twins using descriptive and univariate statistics using unpaired Student t test or Mann-Whitney U test, as appropriate, for continuous variables and χ 2 test or Fisher exact test, as appropriate, for dichotomous variables. Continuous variables were tested for normality visually and with the Kolmogorov-Smirnov test. Stratified analyses were used to identify potentially confounding variables in the outcome-exposure association. Multivariable logistic regression models were then developed to better estimate the effect the presence of an anomalous co-twin on preterm delivery and IUGR. Potential confounders were added to the model based on biologic plausibility (chorionicity, race, hypertensive disorders, diabetes) and statistical significance in the unadjusted analysis. Covariates were retained in the model if they changed the effect size around the primary covariate by >10%. The statistical analysis was performed using STATA, version 10, Special Edition (StataCorp, College Station, TX).
Results
Over an 18-year period, 2445 twin pregnancies were identified at our institution, 97 of which were complicated by twins discordant for major anomalies. After excluding miscarriages (n = 34), monoamniotic twins (n = 15), pregnancies where both twins were anomalous (n = 2), twin-twin transfusion syndrome (n = 45), and pregnancies with incomplete outcomes data (n = 372), 1977 pregnancies remained for analysis. After applying exclusion criteria, 27 cases of twins discordant for anomalies were excluded (5 for undergoing termination, the remainder for missing data). Of the included cases, 66 cases of twins discordant for major anomalies were identified. Twins discordant for major anomalies and normal twin pairs were similar with respect to maternal age, race, chorionicity, tobacco use, preeclampsia, and diabetes ( Table 1 ).
