Objective
We sought to evaluate the association between placental histological abnormalities and birthweight discordance and growth restriction in twin pregnancies.
Study design
We performed a multicenter, prospective study of twin pregnancies. Placentas were examined for evidence of infarction, retroplacental hemorrhage, chorangioma, subchorial fibrin, or abnormal villus maturation. Association of placental lesions with chorionicity, birthweight discordance, and growth restriction were assessed.
Results
In all, 668 twin pairs were studied, 21.1% monochorionic and 78.9% dichorionic. Histological abnormalities were more frequent in placentas of smaller twins of birthweight discordant pairs ( P = .02) and in placentas of small for gestational age infants ( P = .0001) when compared to controls. The association of placental abnormalities with both birthweight discordance and small for gestational age was significant for dichorionic twins ( P = .01 and .0001, respectively). No such association was seen in monochorionic twins.
Conclusion
In a large, prospective, multicenter study, we observed a strong relationship between abnormalities of placental histology and birthweight discordance and growth restriction in dichorionic, but not monochorionic, twin pregnancies.
Rates of twin pregnancies are rising as assisted reproductive techniques have become ever more advanced. Twin pregnancies have significantly higher rates of perinatal morbidity and mortality than singleton pregnancies. While this is, to a large extent, related to the high rate of preterm delivery in these pregnancies, the excess of fetal growth abnormalities is also significant. Independently of gestational age at delivery, twins with significant birthweight discordance have poorer perinatal outcomes.
The etiology of birthweight discordance in twins has been extensively investigated. In monochorionic twins, differences are largely attributed to twin-to-twin transfusion syndrome, inequalities in distribution of placental mass between the 2 fetuses, and abnormalities in cord insertion site. In dichorionic twins there may be a difference in genetic growth potential in certain cases, but frequently growth discordance is a pathological entity leading to adverse neonatal outcomes.
Placental pathological examination at both a gross and microscopic level is useful in informing our knowledge of the etiology of abnormal growth in both singleton and twin pregnancies, with a variety of placental pathological lesions implicated in intrauterine growth restriction.
In a large prospective cohort of twin pregnancies we evaluated the association of placental pathology with twin growth restriction.
Materials and Methods
The Evaluation of Sonographic Predictors of Restricted Growth in Twins (ESPRiT) Study was a prospective, multicenter, observational study of twin pregnancies carried out by the Perinatal Ireland Research Consortium at 8 tertiary-level obstetric units in Ireland from May 2007 through October 2009. The ESPRiT Study was set up with the primary aim of establishing a level of birthweight discordance in twin pregnancies that would serve as an independent predictor of adverse perinatal outcome. The study had a number of prespecified secondary analyses including the evaluation of the role of placental pathology in the etiology of birthweight discordance and restricted growth in twin pregnancies. Institutional review board approval was obtained in each center and participants gave written informed consent, including consent to placental histological examination after delivery.
Inclusion criteria for the study were twin pregnancies enrolled <22 weeks’ gestation, with both twins alive at the time of enrollment and intact membranes. Monoamniotic twins were excluded, as were cases where one or both twins had a major structural or a chromosomal abnormality.
Following enrollment all subjects underwent intensive sonographic surveillance with regular assessment of biometric parameters as well as multivessel Doppler studies. Delivery outcome data were collected including mode of delivery, gestational age, and birthweight. Perinatal morbidity outcomes that were assessed were death, respiratory distress syndrome, hypoxic-ischemic encephalopathy, intraventricular hemorrhage, periventricular leukomalacia, necrotizing enterocolitis, and sepsis.
Following delivery placentas were collected and labeled as A (1 cord clamp) or B (2 cord clamps) according to birth order. Placental examination was carried out in the pathology department of the delivery hospital according to a defined study protocol, as outlined below. Pathologists were not blinded to clinical outcomes. Where there was no pathologist available to carry out the placental examination locally in accordance with the study protocol, the placental examination was carried out in the pathology department of the Rotunda Hospital, Dublin, the coordinating hospital for the study.
Formalin fixation was carried out as per local practices in the delivery hospital. Placentas undergoing fixation were immersed in 10% phosphate-buffered formalin for a minimum of 24 hours.
Gross placental examination was performed to evaluate umbilical cord vessel number and umbilical cord insertion site. Evaluation of placental chorionicity was performed by examination of the intertwin membrane. Both gross and histological examinations of the intertwin membrane were performed and each twin pregnancy was recorded as dichorionic or monochorionic.
The membranes were sampled at the rupture site and at a further random site and examined histologically. A cross section of each umbilical cord from the placental end and the fetal end was submitted for microscopy. Each placental disc was examined macroscopically and multiple sections from each disc were submitted for histological assessment. When examining the monochorionic placentas the vascular equator was identified and the placental mass on either side of this was considered separately and the findings were assigned to the corresponding twin. The following placental pathological lesions were recorded: placental infarction, chorangioma, subchorial fibrin deposition, and retroplacental hematoma. All lesions identified on gross examination were sampled for histological assessment. The size of the lesion was recorded and its location relative to twin 1 or twin 2. The presence of histological abnormalities in placental villus maturation was noted for either twin in each twin pair. The composite outcome of placental abnormalities used in the final analysis was the presence of ≥1 of the placental lesions described.
Birthweight discordance was calculated for each twin pair by expressing the absolute difference in birthweight as a percentage of the birthweight of the larger twin. For the purposes of this analysis a discordance level of ≥20% was deemed significant. Twins were classified as appropriate for gestational age (AGA) or small for gestational age (SGA) by plotting birthweights on twin-specific birthweight centiles. SGA was defined as birthweight <5th centile for gestational age.
The frequency of placental pathological lesions was compared for monochorionic and dichorionic twin pregnancies. The composite placental pathology was then assessed as a factor in birthweight discordance and growth restriction. Frequency of occurrence was compared between smaller twins of birthweight discordant pairs and the larger co-twins and concordant controls. The relative frequency was also analyzed between twins with birthweight <5th centile for gestation and those appropriately grown. Analyses were stratified by chorionicity.
Statistical analyses were performed using SAS software (version 9.1; SAS Institute, Cary, NC). Relative frequencies were compared using χ 2 test. Paired student t test was used to analyze continuous variables. A P value of < .05 was considered statistically significant.
Results
Of 1001 twin pairs recruited to the ESPRiT Study, 66.7% (n = 668) had complete placental pathological examination data available for analyses. Of these, 21.1% (n = 141) were monochorionic and 78.9% (n = 527) dichorionic. Table 1 illustrates the clinical characteristics of the cohort. Monochorionic twins were delivered at an earlier mean gestational age and were on average 302 g lighter than their dichorionic counterparts. The relative frequency of both birthweight discordance of ≥20% and birthweight <5th centile was not statistically significantly different between the 2 groups. A composite measure of adverse perinatal outcome, which included any of the morbidity measures described above or perinatal death, was more frequent in monochorionic twins.
| Variable | Monochorionic | Dichorionic | P value |
|---|---|---|---|
| Total, n (%) | 141 (21.1) | 527 (78.9) | |
| Mean BW, g | 2201 (632) | 2504 (567) | < .0001 |
| Mean GA at delivery, wk | 34.7 (2.9) | 36.3 (2.4) | < .0001 |
| BW discordance >20% | 26 (18.4) | 88 (16.7) | 0.6 |
| BW <5th centile | 16 (5.7) | 76 (7.2) | 0.4 |
| Composite perinatal morbidity | 83 (29.4) | 150 (14.23) | < .0001 |
Overall 34.7% (n = 464/1336) of twins in the study group had a placenta that demonstrated ≥1 of the placental pathological lesions assessed. Lesions were more frequently seen in placentas of monochorionic than dichorionic twins ( P = .009) ( Table 2 ). When the individual histological abnormalities were categorized all abnormalities other than chorangioma were significantly more common in monochorionic placentas ( Table 2 ).
| Variable | Monochorionic | Dichorionic | P value |
|---|---|---|---|
| Infarction | 34 (18.7) | 88 (8.3) | .0001 |
| Chorangioma | 0 (0) | 7 (0.7) | .6 |
| Subchorial fibrin | 32 (17.6) | 113 (10.7) | .012 |
| Retroplacental hematoma | 17 (9.3) | 53 (5.0) | .03 |
| Abnormal villus maturation | 53 (29.1) | 151 (14.3) | .0001 |
| Composite placental outcome | 117 (41.5) | 347 (32.9) | .009 |
The relationship between birthweight discordance and placental abnormalities was then analyzed. Overall 17.1% (n = 114) of the cohort had >20% difference in birthweight between the smaller and larger twin. The results for the smaller discordant twins (n = 114) were compared to the combined group of larger co-twins and twins with concordant birthweight (n = 1222). In all, 44.7% of the smaller twins had abnormal findings at placental examination. This was significantly more frequent than the comparison group of larger twins and twins with concordant birthweight (33.8%, P = .02) ( Figure 1 ).
