Preterm premature rupture of membranes, chorioamnion inflammatory scores and neonatal respiratory outcome

Authors


V Zanardo, Department of Pediatrics, Padua University School of Medicine, Via Giustiniani, 3, 35128 Padua, Italy. Email zanardo@pediatria.unipd.it

Abstract

Objective  To evaluate whether histological chorioamnionitis (HCA), in the setting of preterm premature rupture of membranes (PPROM), affects infant respiratory outcome.

Design  A prospective histological study on 287 consecutive placentas was performed in preterm infants (≤32+6 weeks gestation), categorised into four groups: according to the presence or absence of HCA, in the setting or in absence of PPROM.

Setting  Neonatal intensive care unit, Department of Pediatrics, Padua University, Padua, Italy from January 2001 to December 2006.

Results  Among the 287 NICU admitted preterm infants, 68/287 (23.6%) presented with HCA, 16/68 (23.5%) with a coexisting fetal inflammatory response, and 74/287 (25.7%) with PPROM. HCA was associated with a greater frequency of vaginal delivery (P < 0.0001), lower gestational age (P < 0.0001) and lower birth weight (P < 0.01). HCA had no effect on fetal lung maturation, however, it was a significant risk factor for CLD (RR; 95% CI 2.08; 1.30–3.33). HCA and the fetal inflammatory response were also significant risk factors for PPROM (RR; 95% CI 2.07; 1.42–3.03 and 2.64; 1.71–4.09 respectively). Conversely, HCA in the setting of PPROM failed to reveal any RDS protection or subtype CLD risk. Multivariate analysis demonstrated significant independent effects of presence of maternal HCA (P = 0.04), gestational age (P < 0.0001) and interaction HCA-gestational age (P = 0.04) on CLD development, regardless of the presence of fetal HCA or fetal HCA-gestational age interaction, PPROM or PPROM-gestational age interaction.

Conclusions  Histological chorioamnionitis is a significant PPROM and CLD risk factor, but it fails to provide any protection from RDS. HCA in the setting of PPROM also failed to reveal any RDS protection or subtype CLD risk.

Introduction

The relationship between in uterus infection or inflammation, defined as chorioamnionitis, and the subsequent respiratory outcome of preterm infants is not yet fully understood.1 The term chorioamnionitis refers to both the clinical syndrome of intrauterine infection and the histological process responsible for it. Most important, chorioamnionitis is often clinically silent. The pathologist diagnoses it after birth as a continuum, which refers to both the origin of inflammatory cells, maternal versus fetal and the severity of the inflammation, with different sites and severity scores causing different fetal lung effects.2–4 In addition, the presence of chorioamnionitis may itself cause rupture of the membranes.5 This is the reason why PPROM and chorioamnionitis frequently coexist.6,7

Preterm premature rupture of membranes (PPROM), defined as spontaneous rupture of the membranes before 37 weeks of gestation, is the major identifiable cause of preterm delivery,8,9 occurring approximately in one-third of all preterm births.8 Multifactorial in nature,5 PPROM, however, is most strongly associated with the presence of intrauterine infection/inflammation.10 This is relevant, considering that a fetal inflammatory response syndrome, with a complex series of events, which include an inflammatory cascade, increased surfactant production and alterations to lung structure11,12 are present in nearly one-half of fetuses with PPROM or preterm labour and intact membranes.13,14 Furthermore, several studies suggested that PPROM confers protection against RDS.15–18 However, a large review did not confirm this protective effect.19 In this scenario, the fetal lung significance of chorioamnion inflammation presence or absence in the setting of PPROM remains to be determined.20 Critical parameters for considering this relationship include potential confounders such as the impact of histologically confirmed staging and grading chorioamnionitis on PPROM3,21 and the incidence of related acute and chronic respiratory diseases in this population.

In this study, we focused these issues by examining chorioamnion inflammation as a potential clinical risk factor for fetal lung inflammation and maturation in a recent cohort of LBW infants in presence or absence of PPROM, given the current standards of antenatal steroids and surfactant use.

Materials and methods

A prospective histological study on placentas was performed in all neonates <32 completed weeks delivered at the level III Maternity Hospital of the Institute of Gynecology and Reproductive Science of Padua University and consecutively admitted to the NICU at the Department of Pediatrics of Padua University from January 2001 to December 2006, and identified from a prospectively managed database. The Institutional Review Board approval was obtained. All maternal, obstetric, and neonatal records were reviewed. The placentas were regularly submitted for histopathological diagnosis in all cases of preterm delivery. All neonates included in the analysis had placental pathology performed. An experienced pathologist in placental pathology examined and sampled the placentas, and issued the report according to the Guidelines of the College of American Pathologists.22

Placental histological examination included a minimum of three cross-sections of the umbilical cord taken from fetal and placental side of the umbilical cord, three membrane rolls and a sample of the chorionic plate. The inflammatory findings at histology were recorded and accounted for the worst area scored, according to the classification of Redline23 and included the presence of acute maternal histological chorioamnionitis (HCA) (Stage 1: subchorionitis/chorionitis; Stage 2: chorioamnionitis; Stage 3: necrotising chorioamniuonitis) with or without a coexisting fetal inflammatory response (Stage 1: chorionic vasculitis/umbilical phlebitis; Stage 2: umbilical arteritis; Stage 3: umbilical perivasculitis), graded for the degree of severity/as follows: 1—rare neutrophils [<10/high-powered field (HPF)], 2—intermediate neutrophils (11–30/HPF) and 3—abundant neutrophils (>30/HPF).

All the included women had amniorrhexis clearly identified by physical examination (documentation of two of the following: decreased amniotic fluid index, or positive nitrazine, ferning or pooling), and accurate dating confirmed by ultrasonography. Women were excluded if they had any of the following: >32 weeks gestation, fetal anomalities and intrauterine fetal demise on presentation. Labour was not a criterion for exclusion, nor was tocolysis. In the course of their admission, evaluations of cervical and vaginal cultures were routinely obtained. All women were uniformly managed by the antenatal administration of glucocorticoids (betametasone 12 mg intramuscularly 24 hours apart for a total of two doses) and antibiotics. The antimicrobial drug chosen was left to the discretion of the attending obstetrician. Women were monitored using daily non-stress testing and twice weekly using biophysical profiles. Tocolysis was employed with ritodrine, if regular uterine contractions were present for a 48-hour period after admission, to gain fetal exposure to steroids. Cesarean section was performed when fetal distress or maternal or obstetric indications were present. All neonates were evaluated and managed by a certified neonatologist. Apgar scores were determined by a neonatologist in the delivery room. The policy of the unit during this period was to administer surfactant treatment to infants (by laryngeal mask airway or intubated) for RDS.24 Neonatal morbidity included RDS and CLD. RDS was defined by the need for supplemental oxygen and ventilation support and by the characteristic radiographic findings. CLD was defined in premature infants dependent on supplemental oxygen at 36 weeks corrected postnatal age.25 Demographic and clinical data hypothesised a priori to have some impact on adverse perinatal outcome were specifically abstracted, and these are listed in Tables 1 and 2.

Table 1.   Clinical profile and respiratory outcome of preterm infants according to the presence or absence of HCA, in the setting or in absence of PPROM
n (%)HCAP ; RR 95% CIPPROMP ; RR 95% CI
PresentAbsentPresentAbsent
  1. HCA, histological chorioamnionitis; PPROM, preterm premature rupture of membranes; RDS, respiratory distress syndrome; CLD, chronic lung disease; RR, relative risk; CI, confidence interval.

  2. Numbers are stated as n (%) or mean ± SD. RR; 95% CI; statistical significance: P < 0.05.

  3. *Fischer’s test: **Wilcoxon’s test.

Total number68 (23.6)219 (76.3) 74 (25.7)213 (74.2) 
Prenatal steroids49 (72.0)127 (57.9.) 62 (83.7)167 (78.4) 
Vaginal delivery28 (41.1)11 (5.0)0.000120 (27.0)19 (8.9)0.0001
Gestational age (week)26.9 ± 2.429.5 ± 2.40.0001**28.2 ± 2.729.0 ± 2.40.006**
Birth weight (g)1004 ± 3571190 ± 4170.011115 ± 4181143 ± 409 
Apgar
1 min6 ± 2.26 ± 3.5 6.6 ± 2.36.5 + 2.1  
5 min6 ± 3.18 ± 1.40.058.8 ± 0.68 ± 1.5 
RDS48 (70.5)137 (62.5) 41 (55.4)144 (67.6) 
Surfactant (n)36 (52.9)100 (45.6) 23 (31.0)113 (53.0) 
CLD: O2 36 week20 (29.4)35 (15.9)0.02*; 2.06; 1.30–3.3316 (21.6)39 (18.3) 
Death5 (7.3)15 (6.8) 5 (6.7)15 (7.0) 
Table 2.   Clinical profile and respiratory outcome of preterm infants with histological chorioamnionitis in the setting or in absence of PPROM
n (%)PPROM 74No-PPROM 213P ; RR 95% CI
  1. HCA, histological chorioamnionitis; PPROM, preterm premature rupture of membranes; RDS, respiratory distress syndrome; CLD, chronic lung disease; RR, relative risk; CI, confidence interval.

  2. Numbers are stated as n (%) or mean ± SD. RR; 95% CI; statistical significance: P < 0.05.

HCA29 (39.1)39 (18.3)2.07; 1.42–3.03
Prenatal steroids26 (89.6)30 (76.9) 
Gestational age (week)26.8 ± 2.327 ± 2.60.01
Vaginal delivery12 (41.2)16 (41.0) 
Birth weight (g)986 ± 3761018 ± 3460.01
Staging
I1123 
II96 
III910 
RDS19 (65.5)29 (74.3) 
CLD6 (20.6)14 (35.8) 
Fetal HCA10 (13.5)6 (2.8)2.64; 1.71–4.09
Staging
I21 
II63 
III22 
RDS7 (70.0)5 (83.3) 
CLD2 (20.0)1 (16.6) 

Anthropometrical and clinical data of mothers and premature infants were expressed as mean ± SD. Differences between groups were assessed by Student’s t test, Fisher’s test and Wilcoxon’s test as appropriate. The relative risk (RR, 95% confidence interval, CI) was performed in the comparison between the cases with or without HCA in the setting or in the absence of PPROM. A logistic regression model was created to control the independent effects of explanatory variables (maternal and fetal HCA presence/absence, gestational age, interaction maternal and fetal HCA-gestational age, PPROM presence/absence and interaction PPROM-gestational age) on CLD development.These interactions were included to check for possible continuous-by-class effects, because of the interactions HCA-gestational age and PPROM-gestational age. A P-value <0.05 was considered significant.

Results

Among the NICU admitted preterm infants, 68/287 (23.6%) presented with HCA, 16/68 (23.5%) with a coexisting fetal inflammatory response and 74/287 (25.7%) with PPROM (Table 1).

Histological chorioamnionitis was associated with a greater frequency of vaginal delivery [28 (41.1%) versus 11 (5.0%); P = 0.0001], lower gestational age [26.9 ± 2.4 weeks versus 29.5 ± 2.4 weeks; P < 0.0001], lower birth weight (1004.7 ± 357 g versus 1190 ± 417 g; P = 0.01). HCA had no effect with respect to fetal lung maturation, however, was a significant risk factor for CLD [20/68 (29.4%) versus 35/219 (15.9%), RR, 95% CI 2.08; 1.30–3.33] (Table 1).

Preterm premature rupture of membranes infants were comparable with no-PPROM infants in all selected demographics and clinical variables, and in the neonatal respiratory outcome. However, PPROM was associated with a greater frequency of vaginal delivery [20 (27.0%) versus 19 (8.9%); P = 0.0001] and lower gestational age [28.2 ± 2.7 weeks versus 29 ± 2.4 weeks; P = 0.006] respectively (Table 1).

The present findings also indicate that maternal HCA and coexisting fetal inflammatory response were significant risk factors for PPROM (29/74 (39.1%) versus 39/213 (18.3%) and 10/74 (13.5%) versus 6/213 (2.8%); RR; 95% CI 2.07; 1.42–3.03 and 2.64; 1.71–4.09 respectively).

Premature infants with HCA in the setting of PPROM presented lower gestational age [26.8 ± 2.3 weeks versus 27 ± 2.6 weeks, P < 0.01] and lower birth weight [986 ± 376 g versus 1018 ± 346 g, P < 0.01] respectively. Conversely, HCA in the setting of PPROM failed to reveal any RDS protection or subtype CLD risk (Table 2).

Multivariate analysis by a logistic regression model showed significant independent effects of presence of maternal HCA (P = 0.04), gestational age (P < 0.0001) and interaction maternal HCA-gestational age (P = 0.04) on CLD development; presence of fetal HCA (P = 0.28) and interaction of fetal HCA-gestational age (P = 0.41), presence of PPROM (P = 0.38) and interaction of PPROM-gestational age (P = 0.35) had no significant effects.

Discussion

Chorioamnion inflammation is frequently diagnosed in preterm delivered placentas after indicated delivery or spontaneous preterm labour, either with or without intact membranes, and has been thought responsible for an inflammatory cascade, which induces the preterm parturition syndrome and the fetal inflammatory response syndrome.13,14,18

In this study, we found an association between maternal HCA, coexisting fetal inflammatory response and PPROM, in preterm infants <32 weeks gestation. In addition, HCA proved to be a significant risk factor for CLD development, while it was ineffective in respect to fetal lung maturation. Conversely, HCA in the setting of PPROM failed to reveal any protection towards RDS or any additive subtype risk for CLD. These findings are not surprising given that, while obstetricians have historically attributed fetal membrane rupture to physical stress7 the prenatal inflammation is actually believed to have a crucial role in the pathogenesis of PPROM. Our histopathological findings confirmed a pathogenetic role of chorioamnionitis in PPROM,5 the frequent coexistence of the two,10 as well as the relationship between chorioamnionitis occurrence and chronic lung damage. Previously described risk factors for CLD, not significant in this study, included also PPROM.4 Nevertheless, the available epidemiological data do not unanimously support the ‘early-protection, late-damage’ scenario purportedly initiated by the chorioamnionitis.2 In keeping with an increased risk for RDS in severe chorioamnionitis,26 postmortem data from human fetuses suggest that HCA is associated with a prominent inflammatory response in the fetal lung,5 similar to the pattern observed in infants who die with RDS.27 In particular, the role of chorioamnionitis in CLD has been controversial. Although most studies have found a positive association, methodological questions have been raised regarding comparability of birth weights, the definitions of both chorioamnionitis and CLD, the time period of study and geographical variations.4 Further analyses considering clinical and HCA subdivided by origin of inflammatory cells (maternal versus fetal) and severity of inflammation, failed to reveal any subtype that was positively correlated with CLD.21 In one additional study, funisitis with arterial inflammation was associated with CLD, while funisitis without arteritis was not.28 Ohyama29 showed that sub chronic chorioamninitis, especially when associated with amnion necrosis, is a unique prognostic indicator of CLD. Others found significant correlation between necrotising funisitis and CLD.30 Possible explanations for these differential effects include differentiated rates of co-morbidities, geographical and racial differences in cytokine response or organ susceptibility to cytokines, and different postnatal responses, which arise according to different in uterus environments. Our data, which show significant differences in delivery route, support this last explanation.

The present findings, given the current standards of antenatal steroids, surfactant use, and the significant low gestational age and birth weight of the preterm infant with HCA, confirm that HCA is per se a significant risk factor for PPROM and CLD. Conversely, HCA in the setting of PPROM failed to reveal any protection on RDS or any additional subtype risk on CLD.

There are certain limitations in our work. Firstly, only the relationship between HCA, in the setting or in absence of PPROM, and neonatal respiratory outcome was investigated. Even if the women had chorioamnionitis, they may be dissimilar in some important ways (i.e. women who had clinical chorioamnionitis and developed PPROM, women with PPROM tocolyzed for up to 48 hours who developed histological and/or clinical chorioamnionitis, women treated with antibiotics or with ritodrine) and thus such information might have been important for the data analysis. Having multiple cytokine measures taken over perinatal time from neonates with different HCA scores would allow for the verification of the assumption underlying this analysis and the other related adverse long-term respiratory sequelae of the prematurity.31 Secondly, we might lack sufficient power to detect the unanticipated PPROM disparity in the effect of HCA on CLD.

However, the study has several strengths. We used reproducible, reliable histological methods, already validated in previous outcome studies in LBW infants4,27 as well as accurate assessment of neonatal history from a single institution, prospectively managed by a database. Moreover, the fact that 23% of the placentas in the study contained HCA, 39% in the setting of PPROM, provides a large enough group to be confident with the results and the measurable presence of RDS and CLD in NICU admitted premature infants.

In conclusion, the present findings suggest that HCA is a significant risk factor for PPROM occurrence and for CLD development, while it is ineffective with respect to fetal lung maturation. They also indicate that HCA in the setting of PPROM failed to reveal any RDS protection or subtype risk for CLD.

Disclosure of interests

No conflict of interest.

Contribution to authorship

The authors below are responsible for the reported research, have participated in the concept and design, analysis and interpretation of data, drafting or revising of the manuscript, and have approved the manuscript as submitted.

Details of ethics approval

The Institutional Review Board approval was obtained. Regione del Veneto Azienda ospedaliera di Padova Prot. N.1136P; 14 April 2006.

Funding

No funding.

Ancillary