Objective
The objective of the study was to compare neonatal morbidity and long-term neurodevelopmental outcome between very preterm infants with placental underperfusion and very preterm infants with histological chorioamnionitis.
Study Design
We measured the mental and motor development at age 2 and 7 years in 51 very preterm infants with placental underperfusion and 21 very preterm infants with histological chorioamnionitis.
Results
At 2 years, very preterm infants with placental underperfusion had poorer mental development than very preterm infants with histological chorioamnionitis (mean [SD] 90.8 [18.3] vs 104.1 [17.2], adjusted d = 1.12, P = .001). Motor development was not different between both groups (92.8 [17.2] vs 96.8 [8.7], adjusted d = 0.52, P = .12). At 7 years, large, although nonsignificant, effects were found for better mental and motor development and fewer behavioral problems in infants with histological chorioamnionitis.
Conclusion
Placental pathology contributes to variance in mental development at 2 years and should be taken into account when evaluating neurodevelopmental outcome of very preterm infants.
Normal placental functioning is of major importance in fetal development by its critical role in protection of the fetus and exchanging nutrients between mother and child. Very preterm birth (less than 32 weeks of gestation) is strongly associated with pathology of the placenta. Placental dysfunction exposes the fetus to an unfavorable intrauterine environment and may contribute to both preterm delivery and damage to the developing brain.
The 2 most common types of placental pathology in preterm delivery are maternal vascular underperfusion leading to fetal hypoxia (placental underperfusion) and pathology associated with acute bacterial intrauterine infection (histological chorioamnionitis). Other placental lesions associated with preterm birth are chronic villitis, fetal thrombotic vasculopathy, or massive villous fibrin deposition. The relation between placental underperfusion and preterm birth is possibly mediated by an interplay between fetal nutritional status, activity of the hypothalamic-pituitary-adrenal axis and increased production of corticotrophin-releasing hormone, inducing the onset of labor. In histological chorioamnionitis, the local release of cytokines possibly stimulates prostaglandin release by fetal membranes and uterine deciduas, both contributing to preterm labor.
Besides contributing to preterm birth, placental underperfusion and histological chorioamnionitis may also directly affect brain development. Placental underperfusion may lead to fetal hypoxia, which contributes to the breakdown of the blood-brain barrier and triggers glutamate excitotoxicity. Free radicals in combination with a developmental lack of antioxidant enzymes in oligodendrocytes may further explain the impact of hypoxia on the premature brain. Although cerebral blood flow increases in response to fetal hypoxia, this brain-sparing effect is probably insufficient to maintain adequate cerebral oxidative metabolism. In histological chorioamnionitis, inflammatory cytokines may be neurotoxic and may also increase the permeability of the blood-brain barrier. They interfere with normal myelination of the brain directly by damaging myelin and indirectly by causing damage to myelin producing cells. Cytokines may also have a systemic effect causing hypotension and disseminated intravascular coagulation, propelling the development of brain damage.
Neurodevelopmental outcome in very preterm infants varies widely. Better prediction of outcome can improve selection of infants for early therapeutic interventions. Differences in pathological mechanisms in placental underperfusion and histological chorioamnionitis may contribute to variation in neonatal morbidity and outcome. Several studies have addressed the relationship between placental pathology and short-term neurodevelopmental outcome in very preterm infants. However, studies concerning long-term outcome remain scarce. Therefore, the differences in outcome following these 2 etiological causes of preterm delivery remain unclear.
The current study aimed to compare neonatal morbidity and the short-term and long-term neurodevelopmental outcome of very preterm infants with placental underperfusion and very preterm infants with histological chorioamnionitis.
Materials and Methods
Sample
The infants of this study are part of a larger cohort of a randomized controlled trial on postnatal glutamine supplementation in very preterm infants (gestational age [GA] <32 weeks or birthweight [BW] <1500 g [range, 24.4–33.3 weeks, n = 102]). To disentangle the effect of placental underperfusion and histological chorioamnionitis, we excluded infants who had both types of pathology (ie, lesions of placental underperfusion and lesions of histological chorioamnionitis [n = 11]). Postnatal glutamine supplementation did not influence placenta pathology and did not influence neurodevelopmental outcome at 2 years. The different types of placental pathology were equally present in both treatment groups (data not shown). The medical ethical review board of our institute approved the study protocol. Written informed consent was obtained from all parents.
Placental histological examination
Placentas were macroscopically examined and preserved in 10% buffered formalin for 24 hours immediately after delivery.
A minimum of 2 cassettes containing full-thickness sections from the center of the normal-appearing placental disk were taken. Furthermore, sections were taken from the placental membranes and the placental and fetal ends of the umbilical cord. Placenta dimensions were obtained and the placentas were weighed after the removal of the umbilical cord and membranes. The microscopic examination was undertaken by an experienced perinatal pathologist (J.P.v.d.V.), who was unaware of the clinical outcome of the infants.
Histologically, placental underperfusion was scored in the presence of characteristics associated with maternal vascular underperfusion as described by Redline et al : increased syncytial knots, villous agglutination, distal villous hypoplasia, and intervillous fibrin. Grade 1 placental underperfusion was scored when less than 30% of the villous parenchyma showed 2 or more signs of maternal underperfusion, in the presence of no more than 1 nonmarginal villous infarct. Grade 2 placental underperfusion was scored when placental weight was below the 10th percentile for gestation, in the presence of either 2 or more signs of maternal underperfusion in greater than 30% of the villous parenchyma, or multiple nonmarginal villous infarcts.
Histological chorioamnionitis was subdivided into a group with only a maternal response and a group with an additional fetal response. A maternal inflammatory response was defined as the presence of neutrophils in the placental membranes or the chorionic plate. A fetal inflammatory response was diagnosed when neutrophils were present in the umbilical cord vessel walls or the fetal chorionic plate vessels.
Perinatal characteristics
Perinatal characteristics were recorded in the initial study. In this study, we analyzed perinatal characteristics including maternal age, gravidity, parity, gestational age, BW, BW z -scores, head circumference z -scores, Apgar score at 5 minutes, and umbilical cord pH. BW and head circumference z-scores were determined according to Usher and McLean.
Mortality and neonatal morbidity
We assessed the incidence of serious neonatal infections as previously described. Furthermore, we assessed the occurrence of intraventricular hemorrhage (IVH), periventricular leukomalacia (PVL), Bell stage II or III necrotizing enterocolitis (NEC), retinopathy of prematurity (ROP), bronchopulmonary dysplasia (BPD), age at discharge, and mortality.
Short-term and long-term neurodevelopmental outcome
Short-term neurodevelopmental outcome was assessed at the corrected age of 2 years by an experienced pediatric psychologist using the Bayley Scales of Infant Development II (BSID-II), consisting of Mental Development Index (MDI) and Psychomotor Development Index (PDI).
At 7 years, 4 subtests of the Wechsler Intelligence Scale for Children-III (WISC-III) were used to assess performal IQ (PIQ) and verbal IQ (VIQ). Motor development was measured with the Movement Assessment Battery for Children (MABC), evaluating manual dexterity, ball skills, and balance; higher scores indicate poorer motor performance. Behavioral problems were measured by the parent-rated Children Behavior Checklist and the parent-rated Disruptive Behavior Disorders questionnaire. Higher scores indicate poorer functioning on all subscales.
Statistical analysis
Data were analyzed using analysis of variance for continuous data and logistic regression for dichotomous data. Data that were nonnormally distributed were corrected using a van der Waerden transformation. Effect size Cohen’s d were calculated. Cohen’s d is defined by the difference between 2 means divided by the pooled SD for those means, and effect sizes of 0.2, 0.5, and 0.8 are considered a small, medium, and large effect, respectively.
The effect of severity of placental pathology (maternal vs fetal histological chorioamnionitis, grade 1 vs grade 2 placental underperfusion) was determined for outcome measures, which were different between infants with histological chorioamnionitis and infants with placental underperfusion. To account for possible bias in selection because of deceased and very disabled infants in our cohort, we performed additional logistic regression analysis using a dichotomous outcome measure, with a cutoff point of –1 SD, including deceased and severely disabled infants in the –1 SD group. Stability of neurodevelopmental outcome was evaluated by calculating correlations and the odds ratio between outcome measurements at 2 and 7 years.
To further clarify the relations between placental pathology and long-term neurodevelopment, analyses were performed with and without adjustment for possible confounders. Possible confounders were defined as factors significantly different between both groups and known to be related to outcome. We did not adjust for head circumference z -score because of the high correlation with BW z -score (r = 0.88). Furthermore, we did not adjust for placental weight because placental weight was part of our definition of placental underperfusion. Therefore, adjusting for placental weight would remove part of the variance associated with placental underperfusion. Adjustments were made using linear or logistic regression analysis. To exclude the possibility that glutamine supplementation influenced the results, we tested the interaction effects between glutamine supplementation, placenta pathology, and outcome. Data were analyzed using SPSS 17.0 (SPSS Inc, Chicago, IL), and a P < .05 was considered significant.
Results
Subjects
Twenty-one infants with histological chorioamnionitis and 51 infants with placental underperfusion were included in our analysis. There were no significant interaction effects among glutamine intervention, placental pathology, and outcome (all P > .15).
Perinatal characteristics
Infants with histological chorioamnionitis had a lower gestational age than infants with placental underperfusion ( P < .001, d = –1.11) but had higher BW z -scores ( d = .79, P = .005) ( Table 1 ). Head circumference z -score was higher in infants with histological chorioamnionitis than in infants with placental underperfusion ( d = .85, P = .003). Infants with histological chorioamnionitis were less often born with cesarean delivery (odds ratio [OR], 0.19; 95% confidence interval [CI], 0.06–0.57). Other perinatal characteristics were not different among both groups.
Characteristics | Chorioamnionitis (n = 21) | Placental underperfusion (n = 51) | P value |
---|---|---|---|
Birthweight, g a | 1129.7 (317.8) | 1111.9 (322.2) | .83 |
Gestational age, d a | 193.7 (13.3) | 206.7 (9.8) | < .001 |
Birthweight z -score a | 0.07 (1.2) | –1.12 (1.78) | .005 |
Head circumference z -score a | 0.57 (1.0) | –0.45 (1.35) | .003 |
Male, n (%) | 10 (47.6) | 27 (52.9) | .68 |
Maternal age, y a | 29.1 (6.3) | 30.6 (4.7) | .29 |
Multiple birth, n (%) | 3 (14.3) | 10 (19.6) | .64 |
CRIB score a | 3.71 (2.9) | 3.71 (3.0) | .99 |
Apgar score at 5 min <7, n (%) | 5 (23.8) | 15 (29.4) | .63 |
Umbilical cord pH <7.10, arterial, n (%) | 0 (0) | 4 (7.8) | .19 |
Cesarean delivery, n (%) | 7 (33.0) | 37 (72.6) | .01 |
Placental weight, g a , b | 285.06 (122.33) | 214.98 (86.71) | .02 |
Preeclampsia, eclampsia, HELLP, n (%) | 0 (0) | 18 (35.29) | .02 |
Clinical abruptio placentae, n (%) | 0 (0) | 4 (7.8) | .19 |
Antenatal corticosteroid, n (%) | 16 (76.0) | 43 (84.3) | .42 |
Mechanical ventilation, d a , c | 16.4 (17.8) | 9.0 (13.0) | .08 |
Age at discharge, d a , c | 84.6 (33.4) | 67.9 (35.3) | .10 |
Social economical status a | 3.1 (0.8) | 3.3 (0.7) | .38 |
a Data are given as mean ± SD;
Mortality and neonatal morbidity
Incidence of ROP was higher in infants with histological chorioamnionitis (29%) than in infants with placental underperfusion (7%) (OR, 5.64; P =.03; 95% CI, 1.19–26.83) ( Table 2 ). After adjustments for GA, BW z -score, and mode of delivery, incidence of ROP was not different between both groups ( P = .75), indicating that the difference in incidence of ROP is largely explained by differences in these characteristics. The incidence of mortality, IVH, sepsis, pneumonia, 1 or more infections, and BPD was not different among both groups. Due to the low incidence, meningitis (n = 0), NEC (n = 4), and PVL (n = 4) were not analyzed.
Outcome measure | Chorioamnionitis (n = 21) | Placental underperfusion (n = 51) | OR | P value | Adjusted OR a | Adjusted P value a |
---|---|---|---|---|---|---|
Mortality, n (%) | 4 (19) | 5 (10) | 2.20 | .29 | 2.35 | .41 |
IVH, n (%) | 9 (43) | 12 (24) | 2.44 | .11 | 1.31 | .69 |
Sepsis, n (%) | 10 (48) | 32 (63) | 0.54 | .24 | 0.36 | .13 |
Pneumonia, n (%) | 3 (14) | 7 (14) | 1.05 | .95 | 1.09 | .94 |
≥1 positive culture, n (%) b | 12 (57) | 35 (69) | 0.61 | .35 | 0.25 | .05 |
ROP, stage 1 or greater, n (%) c | 5 (29) | 3 (7) | 5.64 | .03 | 1.59 | .75 |
BPD, stage 1 or greater, n (%) d | 8 (47) | 17 (37) | 1.57 | .43 | 0.65 | .66 |
a Adjusted for gestational age, birthweight z-score, cesarean delivery;
b Blood culture or tracheal aspirate;
c n = 18 for chorioamnionitis; n = 47 for placental underperfusion;
d n = 17 for chorioamnioinitis; n = 47 for placental underperfusion.
Neurodevelopmental outcome at age 2 years
At 2 years, 57 infants (79.2%) were available for follow up (n = 43 for placental underperfusion and n = 14 for histological chorioamnionitis). Nine infants (12.5%) were deceased and 6 (8.3%) were lost to follow-up. Infants with placental underperfusion had significantly lower MDI scores compared with infants with histological chorioamnionitis ( d = .75, P = .02) ( Table 3 ). The difference in PDI scores was small ( d = .18) and nonsignificant ( P = .59). Adjustment for gestational age, BW z- score, and mode of delivery did not change the detrimental effect of placental underperfusion on MDI scores ( d = 1.12, P < .01).
Variable | Chorioamnionitis (n = 14) | Placental underperfusion (n = 43) | P value | Cohen’s d | Adjusted P value a | Adjusted d a |
---|---|---|---|---|---|---|
MDI score b | 104.1 (17.2) | 90.8 (18.3) | .02 | .75 | .001 | 1.12 |
PDI score b | 96.8 (8.7) | 92.8 (17.2) | .41 | .23 | .12 | .52 |
Chorioamnionitis (n = 18) | Placental underperfusion (n = 48) | P value | OR | Adjusted P value a | Adjusted OR a | |
---|---|---|---|---|---|---|
MDI <85 or death c | 5 (27.7) | 20 (41.6) | .30 | .54 | .19 | .32 |
PDI <85 or death c | 6 (33.3) | 16 (33.3) | .99 | 1.00 | .70 | .73 |
a Adjusted for gestational age, birthweight z -score, cesarean delivery;
b Data are given as mean (SD);
c Including deceased and disabled infants, data are given as n (%).