Prelabor rupture of membranes between 34 and 37 weeks: the intraamniotic

inflammatory response and neonatal outcomes

OBJECTIVE: We sought to determine the influence of microbial invasion of the

amniotic cavity (MIAC) and acute histologic chorioamnionitis (HCA) on the intensity of the
intraamniotic inflammatory response and neonatal morbidity in preterm prelabor rupture of
membranes (PPROM) between 34-37 weeks.
STUDY DESIGN: This study included 99 women with singleton pregnancies
complicated by PPROM between the gestational ages of 34-37 weeks. Amniocenteses were
performed at the time of admission, and MIAC and amniotic fluid interleukin-6
concentrations were determined. After delivery, the placenta was evaluated for the presence
of HCA.
RESULTS: Women with both MIAC and HCA had the highest intraamniotic
inflammatory response, which was mediated by interleukin-6 concentrations (both MIAC and
HCA: median 2164.0 pg/mL; HCA alone: median 654.8 pg/mL; MIAC alone: median 784.1
pg/mL; neither MIAC nor HCA: median 383.0 pg/mL;P<.0001) and the highest incidence of
newborns with early-onset sepsis (P.02).
CONCLUSION: Both MIAC and HCA affect the intensity of theintraamniotic
inflammatory response and the incidence of earlyonset sepsis following PPROM between 34-
37 weeks. The intensity of the intraamniotic inflammatory response should be considered in
the clinical management of PPROM between 34-37 weeks.

Key words: amniotic fluid, gestational age, histologic chorioamnionitis, interleukin-6,

microbial invasion of the amniotic cavity

Preterm prelabor rupture of mem-branes (PPROM) is defined as the leakage of

amniotic fluid <37 gestational weeks, which precedes the onset of reg-ular uterine activity.
PPROM accounts for approximately 30-40% of all preterm deliveries.
A substantial proportion of PPROM cases (40-60%) occur between 34-37 weeks.
Although newborns of mothers with PPROM between 34-37 weeks are often incorrectly
considered nearly physiologically mature, they represent a high-risk group with an increased
rate of morbidities and a higher rate of hospital readmission in the first month of life when
compared to full-term infants.
 Recent studies have evaluated the differences between expectant and active
management of women with PPROM between 34-37 weeks with respect to short-term
neonatal morbidity, espe-cially for early-onset sepsis. Although these studies were well
designed, they did not shed light on this problem, as they revealed no difference in the rate of
early-onset sepsis between active and expectant management of pregnancies with PPROM
between 34-37 weeks. Therefore, the optimal management of PPROM between 34-37 weeks
remains a clinical dilemma because there is lack of evidence to justify either expectant or
active management. This is in contrast to PPROM <34 weeks, a con-dition for which
expectant management is recommended.
Strong evidence indicates that in addition to gestational age (GA), the intraamniotic
inflammatory response is one of the most important determinants of the adverse neonatal
outcomes asso-ciated with PPROM. Our group recently showed that not only intra-amniotic
but also fetal inflammatory responses in PPROM between 24-37 gestational weeks are
highest when both microbial invasion of the amniotic cavity (MIAC) and acute histologic
cho-rioamnionitis (HCA) are present. Therefore, we hypothesize that PPROM pregnancies
between 34-37 weeks com-plicated by these conditions may have a higher intraamniotic
inflammatory response, resulting in worse short-term neonatal morbidity when compared to
PPROM between 34-37 weeks without MIAC and HCA. Furthermore, this topic is of
particular clinical relevance, as it may result in the optimization of the clinical management
of PPROM between 34-37 weeks.
Therefore, the purpose of this study was to evaluate the intensity of the intraamniotic
inflammatory response, as characterized by amniotic fluid inter-leukin (IL)-6 concentrations,
and sub-sequent neonatal morbidity based on the presence of MIAC and HCA in PPROM
pregnancies between 34-37 weeks.

MATERIALS AND METHODS

Sample collection
From May 2008 through March 2013, a prospective cohort study was conducted on
pregnant women at GA between 340 and 366 weeks who were admitted to the Department
of Obstet-rics and Gynecology, University Hospital in Hradec Kralove, Czech Republic.
Pregnant women with singleton preg-nancies, PPROM, and maternal age >18 years were
invited to participate in the study. Exclusion criteria included women with gestational
hypertension, preeclampsia, fetuses with an estimated weight <10th percentile, the presence
of either congenital or chromosomal fetal abnormalities, gestational or pre-gestational
diabetes, and signs of fetal hypoxia. The primary outcome of this study was the intensity of
the intra-amniotic inflammatory response, char-acterized by IL-6 concentrations in the
amniotic fluid. The secondary outcome was neonatal morbidity.
GA were established by first-trimester fetal biometry. In our department, PPROM
pregnancies between 34-37
weeks are actively managed. Induction of labor is initiated or a cesarean section is performed
within 24 hours after mem-brane rupture depending on the fetal status, the maternal serum
concentration of C-reactive protein, and cervicovaginal streptococcus beta colonization. All
these women receive antibiotics after hospital admission and amniocentesis.
PPROM was diagnosed visually using a sterile speculum examination to confirm the
pooling of amniotic fluid in the vagina and confirmed by a positive test for the presence of
insulin-like growth factorebinding protein-1 (ACTIM PROM test; Medix Biochemica, Kau-
niainen, Finland) in the vaginal fluid when necessary.
Ultrasound-guided transabdominal amniocentesis was performed upon admission, prior
to the administration of antibiotics. The amniotic fluid samples were immediately divided
into 3 poly-propylene tubes. The first and second tubes containing noncentrifuged sam-ples
were immediately transported to the microbiology laboratory, where the first tube was used
for polymerase chain reaction (PCR) testing for Ureaplasma species, Mycoplasma hominis,
and Chla-mydia trachomatis, and the second tube was used for aerobic and anaerobic
bacterial culture. The third tube was centrifuged for 15 minutes at 2000g to remove cells and
debris, divided into aliquots and stored at e70_C until analysis.
After delivery, the placentas were fixed in formalin, and tissue samples from the
placenta, umbilical cord, and placental membranes were routinely processed and embedded
in paraffin. Tissue sections were stained with hematoxylin-eosin for standard histo-logic
examination.
The study was approved by the insti-tutional review board committee (March 19, 2008;
no. 200804 SO1P). Written informed consent was obtained from all participants. All women
self-identified as Caucasian. Forty-three women in the study group were examined in one of
our previous reports on the intraamniotic inflammatory response in women with PPROM
throughout the whole range of preterm delivery.

Diagnosis of MIAC
MIAC was defined as a positive PCR for genital mycoplasmas (Ureaplasma parvum,
Ureaplasma urealyticum, and My-coplasma hominis) and/or Chlamydia trachomatis and/or as
the growth of any bacteria in the amniotic fluid except for coagulase-negative Staphylococcus
epi-dermidis, which was considered a skin contaminant.

Diagnosis of HCA
The degree of neutrophil infiltration was evaluated separately in the free mem-branes
(amnion and chorion-decidua), in the chorionic plate, and in the umbilical cord based on the
criteria provided by Salafia et al. A diagnosis of HCA was made based on histologic grades of
3-4 for the chorion-decidua (multiple or confluent foci of at least 5-20 neutrophils), 3-4 for
the chorionic plate (at least a few neutrophils present in the connective tissue or chorionic
plate), 1-4 for the umbilical cord (any neutrophils present in the umbilical cord), and/or 1-4
for the amnion (at least 1 focus of at least 5 neutrophils). His-tologic grades of 1-4 for the
umbilical cord were considered to indicate the presence of funisitis. Histopathological
examinations were performed by a single pathologist who was blinded to the clinical status of
the patient.

Measurement of intraamniotic inflammatory response

The intensity of the intraamniotic in-flammatory response was measured by determining the
IL-6 concentration in the amniotic fluid. IL-6 concentrations were assessed by the human IL-
6 Quantikine enzyme-linked immunosorbent assay (R&D Systems Inc, Min-neapolis, MN).
Amniotic fluid samples were diluted 10-fold. The sensitivity of the test was <0.70 pg/mL,
and the interassay and intraassay coefficients were <10%.
Diagnosis of severe neonatal morbidity
Maternal and perinatal medical records were reviewed by 2 investigators (M.K. and I.M.).
Data regarding morbidity and mortality were reviewed for all newborns. For the current
study, we defined severe neonatal morbidity as follows: the need for tracheal intubation,
respiratory distress syndrome (defined by the presence of _2 of the following criteria:
evidence of respiratory compro-mise, a persistent oxygen requirement for >24 hours,
administration of exogenous surfactant, and radiographic evidence of hyaline membrane
disease), transient tachypnea (any oxygen supplement re-quirement during the first 6 hours
that does not increase during the subsequent 18 hours as clinical conditions improve within 3-
6 hours, and chest radiographs either normal or indicating reduced translucency, infiltrates,
and hyperinsuf-flation of the lungs), intraventricular hemorrhage (diagnosed by cranial ultra-
sound examinations based on the criteria defined by Papile et al), necrotizing enterocolitis
(defined as the radiologic finding of either intramural gas or free intraabdominal gas),
retinopathy of pre-maturity (identified using retinoscopy), early and late-onset sepsis (defined
as any systemic bacterial infection docu-mented by a positive blood culture during the first 72
hours of life and the presence of symptoms of or strong clinical suspi-cion of sepsis [presence
of symptoms and elevated C-reactive protein and/or af-fected white blood cell count]
between 4-120 days of life), bronchopulmonary dysplasia (defined by the infants oxygen
requirement at 28 days of life), pneu-monia (diagnosed by abnormal findings on chest x-
rays), and neonatal death prior to hospital discharge.26

Statistical analysis
Demographic and clinical characteristics were compared using a nonparametric Jonckheere-
Terpstra test and the Mann-Whitney U test for continuous variables and presented as medians
(range). Categorical variables were compared using the Cochran-Armitage test for trends or
Fisher exact test and the results were presented as numbers (%). Receiver operating
characteristic curves were used to determine the best cutoff value of IL-6 concentration in the
amniotic fluid for the diagnosis of early-onset sepsis in newborns. A Spearman partial
correla-tion was used to adjust the data for GA. Differences were considered statistically
significant at P < .05. All P values were determined from 2-sided tests, and all statistical
analyses were carried out using SPSS 19.0 for Mac OS X (IBM Corp, Armonk, NY) and
GraphPad Prism 5.03

for Mac OS X (GraphPad Software, La Jolla, CA).

RESULTS
A total of 118 women with PPROM occurring at GA between 34-37 weeks were recruited for
the study. Of these 118 women, amniocentesis was not possible in 10 (8%) women, and the
re-sults of the histopathological assessment of the placenta were not available for 9 (7%)
women. The remaining 99 women were included in the study. The overall rate of MIAC was
23% (23/99), and HCA was found in 49% (49/99) of women. Funisitis was identified in 12%
(12/99) of women. Both MIAC and HCA were present in 12% (12/99) of women. HCA alone
and MIAC alone were present in 37% (37/99) and 11% (11/99) of women, respectively.
Almost 40% (39/99) of women exhibited neither MIAC nor HCA. No differences in maternal
and clinical characteristics were observed among these subgroups ( Table 1). No cases of
clinical cho-rioamnionitis were present in the study population. The most common bacteria
was Ureaplasma species, which was identified in 65% (15/23) of women. Polymicrobial
findings were found in 9% (2/23) of women (Ureaplasma species Mycoplasma hominis and
Ureaplasma species streptococcus alpha hemolytic). Table 2 lists all microorganisms
detected in the amniotic fluid.

Amniotic fluid IL-6 concentrations

We observed differences in the IL-6 concentrations in the amniotic fluid among the
subgroups based on the presence or absence of MIAC and HCA (P < .0001). Women with
both MIAC and HCA exhibited the highest median concentration of IL-6 among the sub-
groups (both MIAC and HCA: 2164.0 pg/mL; HCA alone: 654.8 pg/mL; MIAC alone: 784.1
pg/mL; and neither MIAC nor HCA: 383.0 pg/mL), and differences were observed in the IL-
6 concentrations in the amniotic fluid between women with both MIAC and HCA and other
women (MIAC and HCA vs HCA alone: P .005; MIAC and HCA vs MIAC alone: P .03;
MIAC and HCA vs neither MIAC nor HCA: P < .0001) ( Figure 1). Women without MIAC
and HCA had lower amniotic fluid IL-6 concentrations than women with HCA alone (P .
03) and those with MIAC alone (P .02), as shown in Figure 2. No differences in the
amniotic fluid IL-6 concentrations were observed between women with HCA alone and those
with MIAC alone (P .60) ( Figure 2).

Neonatal morbidity
Respiratory disorders (respiratory dis-tress syndrome or transient tachypnea of newborns) and
the need for tracheal intubation were observed in 14% (14/99) and 3% (3/99) of newborns,
respectively. Intraventricular hemorrhages of grades I-II were found in 6% (6/99) of new-
borns. No intraventricular hemorrhages of grades III or IV were observed. Pneumonia was
diagnosed in 2% (2/99) of newborns. Severe neonatal morbidity was observed in 25% (25/99)
of new
borns. This cohort of women had no evidence of bronchopulmonary dysplasia, retinopathy of
prematurity, necrotizing enterocolitis, or late-onset sepsis, and no neonatal death.
Selected neonatal morbidities are presented in Table 3 according to the presence or
absence of MIAC and/or HCA. Differences were observed among subgroups in early-onset
sepsis (P .02), which was observed in 4% (4/99) of newborns, and severe neonatal mor-
bidity (P .03). Three cases of early-onset sepsis occurred in the subgroup of women with
both MIAC and HCA. One case (culture proven) was in a woman with MIAC alone. Two
new-borns had culture-proven early-onset sepsis (Haemophilus influenzae and Streptococcus
pneumoniae). Differences in severe neonatal morbidity (P .003) were only observed among
the 3 GA subgroups (34-35, 35-36, and 36-37 weeks) ( Table 4).
The IL-6 concentrations in the amni-otic fluid are shown in Table 5 based on the
presence/absence of selected neonatal morbidities. The presence of early-onset sepsis was
associated with a higher IL-6 concentration (with early-onset sepsis: median 5303 pg/mL vs
without early-onset sepsis: median 575 pg/mL; P .001) by crude analysis and after
adjustment for GA (P <.0001). An amniotic fluid IL-6 con-centration of 3353 pg/mL was
found to be the best cutoff for the identifica-tion of early-onset sepsis (area under the receiver
operating characteristic curve: 0.98; sensitivity: 100%; speci-ficity: 98%; positive predictive
value: 58%; negative predictive value: 100%; odds ratio: 238; likelihood ratio: 32; P <.0001).

COMMENT
Principal findings of this study
First, the rates of MIAC and HCA in women with PPROM between 34-37 weeks were 23-
49%, respectively. Second, the intraamniotic inflammatory response was highest for women
with both MIAC and HCA. Third, no difference was observed in the intraamniotic inflam-
matory response between groups with MIAC alone and HCA alone, although these groups
had a higher intraamniotic inflammatory response than groups lacking both MIAC and HCA.
Fourth, the incidence of only early-onset sepsis and severe neonatal morbidity were affected
by MIAC and HCA in pregnancies with PPROM between 34-37 weeks. Fifth, the rate of
severe neonatal morbidity decreased with ad-vancing GA from 34-37 weeks. Finally, early-
onset sepsis was associated with higher IL-6 concentrations, indepen-dent of GA.

Meaning of the study

MIAC and HCA represent the most important conditions that determine the intraamniotic
inflammatory response and are major determinants of maternal and neonatal outcomes in
pregnancies complicated by PPROM between 34-37 weeks. Despite the difficulty of com-
paring the rates of MIAC among studies because these studies use different tech-niques to
detect bacteria, the rate of MIAC is not believed to be dramatically affected by GA in
PPROM. Our finding of MIAC in PPROM between 34-37 weeks is consistent with this, as
the rate of MIAC in this study did not differ
from previously published studies eval-uating MIAC in PPROM <34 weeks.
Nevertheless, the rate of MIAC can vary among different populations and among
techniques used for bacterial identification. Two pioneering studies by Romero et al showed
that the rates of MIAC in the US population based on cultiva-tion were 38-34% for patients
with pre-term and term rupture of membranes, respectively. Recently, DiGiulio et al 30
reported the rate of MIAC as 34% by the cultivation approach, 45% by PCR, and 50% by
combining both techniques.
It is difficult to compare the data on the incidence of HCA in this study with that from
previously published studies because of the different ways in which PPROM is managed
clinically and because of the classification of inflam-matory changes in the placenta and fetal
membranes. Nevertheless, the rate of HCA in PPROM between 34-37 weeks seems to be
lower than the incidence of HCA in PPROM before GA of 34 weeks as found in our previous
work
Our group recently reported that the presence of both MIAC and HCA was associated with
stronger intraamniotic and fetal inflammatory responses when compared to other PPROM
subgroups between GA of 24-37 weeks. Little information is available regarding the
intraamniotic inflammatory response in PPROM pregnancies between 34-37 weeks with
respect to the presence or absence of MIAC and/or HCA. In this study, we confirmed that the
highest intraamniotic inflammatory responses were observed for patients with both MIAC
and HCA. Moreover, we found that the group with HCA alone exhibited a higher
inflammatory response than the group without MIAC or HCA. One possible explanation for
this finding is that endogenous molecules called alar-mins that are released from the placenta
and fetal membranes can trigger the intraamniotic inflammatory response, leading to the
development of nonin-fectious HCA. Endogenous alarmins are produced and released in
response to tissue damage by necrotic cells. These endogenous molecules can elicit re-
sponses similar to those of exogenous microorganisms through the system of pattern
recognition receptors, and they have recently been identified in the am-niotic fluid. Moreover,
it was recently demonstrated that almost all HCA at term are noninfectious. Therefore, further
study is needed to clarify the role of alarmins for PPROM between 34-37 weeks.
The next interesting finding was that women with MIAC alone had a lower inflammatory
response than those with both MIAC and HCA and a higher response than those without both
MIAC and HCA. To determine the difference between the group with MIAC alone and the
group with both MIAC and HCA, we performed a secondary analysis. No differences were
identified with respect to the rate of genital mycoplasmas (64% for MIAC alone vs 83% for
MIAC with HCA; P .37, data not shown), the microbial load of genital mycoplasmas (1533
copies/mL median for MIAC alone vs 733 copies/mL for MIAC with HCA; P .78; data not
shown), the time
period between PPROM and delivery (16 hours median for MIAC alone vs 24 hours for
MIAC with HCA; P .16), or latency between amniocentesis and delivery (median 12 hours
for MIAC alone vs 14 hours for MIAC with HCA; P .51). Therefore, it remains unclear
why the group with MIAC alone elicited a lower inflammatory response than the group with
MIAC and HCA. We can only speculate as to whether the coor-dination of infectious and
noninfectious factors (microorganisms and alarmins together) is required for the group with
both MIAC and HCA to express the strongest inflammatory response. This speculation is
supported by a study by Romero et al that showed higher concentrations of alarmin in
PPROM pregnancies complicated by either the presence of bacteria in the amniotic fluid or
high amniotic fluid IL-6 concentra-tions. Nevertheless, the differences in the intraamniotic
inflammatory response between the group with MIAC alone and the group with MIAC and
HCA should be addressed in further research.
It is unclear whether MIAC and HCA in women with PPROM between 34-37 weeks
can affect short-term neonatal morbidity. The stratification into 4 subgroups based on MIAC
and HCA revealed that the presence of both MIAC and HCA was associated with the highest
rate of early-onset sepsis and severe neonatal morbidity. More-over, our results suggest that
pregnancies with PPROM between 34-37 weeks with amniotic fluid IL-6 concentrations
<3353 pg/mL are at a low risk of early-onset sepsis. This finding is clinically relevant,
especially in light of the fact that previous studies comparing active and expectant
management listed early-onset sepsis as a primary outcome.
We are aware that an invasive amnio-centesis must be performed to obtain amniotic fluid for
analysis, which is critical to determining the inflammatory status of the intraamniotic
compart-ment. Amniocentesis is also limited by the presence of severe oligohydramnios or
anhydramnios, and this seems to be a major clinical limitation. Therefore, these results should
be replicated for noninva-sive samples such as cervical and vaginal fluid.
A growing body of evidence indicates that neonatal morbidity in PPROM between 34-
37 weeks continuously declines with GA. Given this knowledge, we evaluated short-term
neonatal morbidity. The rates of severe neonatal morbidity decreased with advanced GA.
Moreover, trends were observed with respect to other aspects of neonatal morbidity.
However, the small sample size in the GA subgroups was the primary limitation that
prevented us from obtaining statis-tically significant results.

Strengths and weaknesses of the study

The primary strength of this study is the very short interval between amniocen-tesis and
delivery (median 10 hours) due to the active management of PPROM. Based on this finding,
we posit that the neutrophil infiltration of the placenta and fetal membrane at the time of de-
livery is very similar to the status at the time of the amniocentesis. One limita-tion of this
study is that a noncultivation technique (specific PCR) was used for the identification of
genital myco-plasmas and Chlamydia trachomatis, but not for the other bacteria.
In conclusion, the intraamniotic in-flammatory response is affected by MIAC and HCA
in PPROM between 34-37 weeks, and its intensity is highest when both of these conditions
are pre-sent. Moreover, this specific subgroup has the highest rate of early-onset sepsis.
Basing the management of PPROM be-tween 34-37 weeks on the intraamniotic
inflammatory response may offer a valuable clinical perspective because women with low
amniotic fluid IL-6 concentrations are at a low risk of early-onset sepsis and can thus be
treated expectantly.