html xmlns=”http://www.w3.org/1999/xhtml” xmlns:mml=”http://www.w3.org/1998/Math/MathML” xmlns:epub=”http://www.idpf.org/2007/ops”>
Introduction
Preterm prelabour rupture of membranes (pPROM) may be defined as the spontaneous rupture of fetal membranes at least an hour prior to the onset of labour, in a viable fetus (≥23 weeks), and before 37 completed weeks of gestation. It complicates 2–3% of all pregnancies, thus affecting some 14 000 pregnancies in the UK and 140 000 in the USA each year. Preterm prelabour rupture of membranes accounts for 30–40% of preterm deliveries, and is an independent risk factor for neonatal morbidity and mortality from prematurity, sepsis and pulmonary hypoplasia. Infants born after prolonged periods of pPROM have an excess risk of long-term neurological deficits and pulmonary disease. Subclinical intrauterine infection is a major aetiological factor in the pathogenesis of pPROM. Studies of transabdominal amniocentesis following pPROM show that the frequency of positive culture for infection of the amniotic fluid is 25–40%. The risk of a positive culture is inversely related to the gestational age at which pPROM occurred. Women with intrauterine infection have shorter latency than non-infected women, and infants born with sepsis have a four-fold increase in mortality compared to infants born without sepsis [1]. These findings fuelled interest in the use of antibiotics to prevent pPROM in at-risk women, increase latency after pPROM and as prophylaxis against neonatal morbidity such as oxygen dependency, intraventricular haemorrhage, necrotizing enterocolitis, neonatal sepsis and mortality. In one study, women with a prior history of pPROM had a 13.5% risk of subsequent preterm birth due to pPROM compared to 4.1% risk among their peers without such a history (RR 3.3, p < 0.01) [2]. These women also had a 14-fold higher risk of pPROM at less than 28 weeks in the subsequent pregnancy (1.8% versus 0.13%, p < 0.01), raising the question of whether antibiotics may reduce this risk of pPROM. McGregor and colleagues showed that in women who are at increased risk of pPROM, prophylactic antibiotics significantly reduced the incidence of pPROM in the subsequent pregnancy [3].
Aetiology and Pathophysiology
The tensile strength of the membranes is resident mostly in the amnion, an avascular structure consisting of five distinct histological layers. The epithelial cells of the amnion layer secrete types 3 and 4 collagen, and non-collagenenous glycoproteins including fibronectins [4]. The chorion, on the other hand, provides the feto-maternal interface through the interaction between the cytotrophoblasts and the maternal decidua. Our understanding of the cellular and molecular factors that govern the structural integrity of fetal membranes and their regulation has increased in recent years and it is now evident that only a proportion of cases of pPROM are attributable to infection. Evidence has also accumulated that other pathologic processes unrelated to infection may play a role including choriodecidual fusion defects [5], fetal growth dysregulation [6], activation of membrane apoptosis [7], up-regulation of matrix metalloproteinases, and inhibition of tissue inhibitors of matrix metalloproteinases [8], nutritional factors [9] and smoking [10]. A detailed review of these factors and processes is outside the scope of this chapter, but these observations suggest that routine antibiotics for pPROM may be a simplistic response to a very complex problem.
Diagnosis and Initial Assessment
The diagnosis of pPROM is made by clinical suspicion, maternal history, speculum examination and simple bedside tests. Although it is stated that maternal history has an accuracy of 90% for the diagnosis of pPROM [11], the management and prognosis of a pregnancy complicated by pPROM is so drastically different from the one without that it is critically important to confirm the diagnosis before initiating interventions for pPROM. The presence of a pool of amniotic fluid in the posterior fornix on speculum examination is confirmatory of pPROM. Whether this fluid needs to be tested for confirmation is debatable. There may be a role for nitrazine pH paper testing when there is a very small amount of fluid in the posterior fornix and the observer is uncertain. However, the nitrazine test is not specific, and has a false-positive rate of 17% [11]. The pH of vaginal fluid changes towards the alkaline range in the presence of bacterial vaginosis, other vaginitis, contamination with cervical secretion, blood, semen or urine. In doubtful cases, the clinician should visualize the external cervical os with a Cusco’s speculum and ask the patient to cough gently. Amniotic fluid may be observed trickling down through the os. An extended pad test is also useful. A panel of newer tests has been evaluated for pPROM, including fetal fibronectins and insulin-like growth factor binding protein-1 (IGFBP-1) in cervico-vaginal secretions. These have sensitivities of 94% and 75% respectively, and specificities of 97% [12,13]. Spontaneous rupture of membranes during the second and early third trimesters is usually associated with a near total loss of the entire amniotic fluid pool. Therefore, ultrasound scan evidence of marked oligohydramnios or anhydramnios is highly suggestive in women suspected of a diagnosis of pPROM [14]. Such an ultrasound examination should also evaluate the fetal presentation, growth and estimated fetal weight.
A digital vaginal examination should be avoided following a diagnosis of pPROM unless there is a strong suspicion of labour or imminent delivery. If one or two digital vaginal examinations were done after pPROM, this should not constitute an indication to abandon conservative management or pursue an immediate induction of labour for fear of increased risk of feto-maternal infectious morbidity. Studies have shown that two or fewer digital examinations were associated with a shorter latency period but no increase in fetal or maternal infectious morbidity [15,16].
Current State of the Management of pPROM
In many units, women with a diagnosis of pPROM are admitted into hospital and managed conservatively until 37 completed weeks of gestation in an attempt to increase fetal maturity. Conservative management may include Trendelenburg (head-down tilt) position, four-hourly measurements of maternal temperature, heart rate, respiratory rate and blood pressure, daily cardiotocographs (CTG), weekly, twice or thrice weekly maternal white cell counts, C-reactive protein measurements and culture of vaginal swabs, all in an effort to detect intrauterine infection or chorioamnionitis. However, chorioamnionitis is a fetal disease, not maternal, and maternal inflammatory markers are raised in only 10–15% of cases of proven histological chorioamnionitis [17], suggesting that maternal markers are not sufficiently sensitive to guide clinical decisions. The intrauterine compartment is sequestrated and not necessarily contiguous with the maternal systemic circulation. Furthermore, the bacterial species that commonly participate in chorioamnionitis are frequently subpathogenic, non-pyrogenic and are often not detected on routine microbiological culture methods. Prognostically, it is fetal rather than maternal host response to infection that is predictive of adverse neonatal outcome. However, the search for evidence of fetal inflammation by amniocentesis and/or cordocentesis is not routinely done, and its role in reducing the risk of neonatal complications has not been evaluated.
Appropriate Setting for Management
The value of inpatient management beyond five days is dubious. The majority of women with infection-driven pPROM will deliver within this time frame. Outpatient care with instruction for the patient to take her own temperature and be reviewed once or twice weekly in hospital is reasonable. She should, however, be discouraged from vaginal intercourse, protected or not, and any form of intravaginal cleansing. Immersion in bath water is not associated with an increase in infectious morbidity. A one-off vaginal swab for pathogens and specifically for group B Streptococcus (GBS) will suffice. There is little or no correlation between the organisms that cause amniotic fluid infection/chorioamnionitis and those isolated from the lower genital tract. Nursing women with pPROM in a head-down tilt position is unnecessary, and may encourage a stagnant pool of amniotic fluid and cervico-vaginal secretions, potentially encouraging bacterial growth and multiplication. Although fetal CTG is recommended as part of the surveillance, there are no specific or reliable CTG patterns that are predictive of fetal inflammation or sepsis until very late, and neurological injuries may occur without significant CTG changes. Clinical chorioamnionitis defined as maternal fever (temperature ≥38 °C) and the presence of any two or more of maternal or fetal tachycardia, uterine tenderness, foul-smelling or purulent vaginal discharge, maternal leukocytosis or raised C-reactive protein is poorly predictive of histologic chorioamnionitis, which has been shown to be more predictive of abnormal neonatal outcomes including periventricular echodensity/echolucency, ventriculomegaly, intraventricular haemorrhage and seizures [18].
Biophysical profile scores have been proposed and in some centres used for the prediction of intrauterine infection. There is conflicting evidence that abnormal biophysical profile scores or Doppler studies of the placenta or fetal circulation provided accurate distinction between infected and non-infected fetuses [14,19,20].
Timing of Delivery
The practice of expectant management until 37 completed weeks of gestation to improve fetal maturity is historical and based on the assumption that the prolongation of pregnancy automatically translated to better perinatal outcome. The risks of acute cord compression, cord prolapse and ascending infection are not insignificant. It is well established that the risks of prematurity-related morbidity, including respiratory distress syndrome, necrotizing enterocolitis and high-grade intraventricular haemorrhage and mortality, diminish significantly beyond 32–34 weeks, and that neonatal survival at ≥34 weeks in tertiary units is similar to 37–40 weeks. The increase in survival per additional week of conservative management is less than 1%. Furthermore, studies evaluating the risk–benefit analysis of induction of labour at 34 weeks found similar vaginal delivery rates, and no increase in the risk of obstetric intervention such as instrumental vaginal or caesarean delivery [21]. On the other hand, these studies documented an excess incidence of ascending infection, neonatal sepsis, cord prolapse or compression, fetal demise and longer hospital stay in cases managed conservatively [21,22]. They concluded that the benefits of delivery at 34 weeks’ gestation outweigh the risks of conservative management without increasing obstetric intervention. This balance, however, is in favour of expectant management prior to 32 completed weeks. Delivery before 32 weeks’ gestation is associated with a significant risk of gestational age-related morbidity and mortality. Therefore, unless there are concerns regarding fetal well-being, clinical and/or biochemical evidence of infection, women with pPROM remote from term (≤32–34 weeks) should be managed conservatively to prolong gestation and reduce the risk of gestational age-dependent morbidity and mortality in the newborn. Even with conservative management, 70–80% of women with preterm prelabour rupture of membranes deliver within one week of membrane rupture, leaving a smaller subset of fetuses to remain and mature in utero. The potential benefit has to be balanced against the risk of amnionitis, abruption, umbilical cord compression or prolapse and fetal demise.
Antibiotic Therapy to Prolong Latency and Prevent Neonatal Morbidity After pPROM
Evidence from meta-analysis of 22 randomized trials including 6872 mother–infant pairs suggests that antibiotic therapy following pPROM significantly reduced chorioamnionitis (RR 0.66, 95%CI 0.46–0.96), delayed preterm birth within 48 hours (RR 0.71, 95%CI 0.58–0.87), oxygen therapy (RR 0.88, 95% CI 0.81–0.96) and abnormal cerebral imaging prior to discharge from the hospital (RR 0.81, 95% CI 0.68–0.98) [23]. Co-amoxiclav was associated with an increased risk of neonatal necrotizing enterocolitis (RR 4.72, 95% CI 1.57–14.23) and should be avoided. The antibiotic of choice is still unclear. There was no significant reduction in perinatal mortality or evidence of longer-term benefit in childhood. However, the advantages on short-term morbidities were thought to justify recommendation of routine antibiotics in women with pPROM. A follow-up study evaluated the children’s health at seven years of age and found that antibiotics had little effect [24], but increased the risk of functional impairment among those who were exposed to erythromycin with intact membranes [25]. Questions have persisted regarding the methodology and interpretation of the primary studies, and the choice of antibiotic. Published meta-analyses are dominated by the larger, multi-arm, multi-centre, placebo-controlled ORACLE (Overview of the Role of Antibiotics in Curtailing Labour and Early delivery) trial [26]. Although the role of bacterial vaginosis in spontaneous preterm delivery and pPROM was known at the time of the ORACLE trial, none of the antibiotics used (co-amoxiclav and erythromycin) was effective against bacterial vaginosis organisms. Other methodological flaws in the design, analyses and interpretation were outlined in an accompanying lead commentary [27].
Widespread use of broad-spectrum antibacterial drugs may lead to the emergence of antibiotic-resistant organisms. There had been an increase in erythromycin-resistant organisms even in the years leading up to the publication of the ORACLE trial [28–30]. This trend was attributed to increased use of erythromycin for prophylaxis against GBS for women who were allergic to penicillin, and is likely to escalate with its increased use for pPROM. Broad-spectrum antibiotics may eliminate protective commensal flora, especially in the gut, encourage antimicrobial resistance and the emergence of unusual and more pathogenic species. Furthermore, early-life exposures are recognized as an important factor in the immunological health of children. It has been suggested that the significant rise in childhood allergy in developed countries may be related to abnormal initial gut colonization of infants as a result of obstetric and neonatal practices, including antibiotic exposure [31]. Routine antibiotic therapy presumes an infectious aetiology and is based on the belief that antibiotics somehow prevent or reverse fetal damage. This is not necessarily the case and indeed the reverse may well be true [32]. A significant number of mother–fetus pairs whose pPROM were unrelated to infection are exposed to broad-spectrum antibiotics and no studies have yet examined the risk versus benefit equation of mass antibiotic therapy for pPROM.
Tocolysis
Tocolysis may be applied prophylactically to prevent the onset of uterine contractions/labour in women with pPROM, or used therapeutically to abolish uterine contractions/labour after pPROM. Randomized trials of prophylactic [33,34] and therapeutic tocolysis [35,36], including case control studies [37], failed to show prolongation of pregnancy or reduction in perinatal morbidity or mortality. Therefore, the use of tocolysis should be individualized and restricted to those situations where there is no clinical or biochemical evidence of infection and a course of corticosteroids needed to be completed, or to allow the transfer of the woman to a tertiary centre. The choice of a tocolytic agent in this setting is a matter for local guidelines. The efficacies of the available agents on the market are broadly similar. The β-adrenoceptor agonists have become less popular because of their marked cardiovascular side-effects.
Antenatal Corticosteroids Administration
Initial concerns that corticosteroids may increase the risks of chorioamnionitis, postpartum endometritis and neonatal sepsis are not supported by current evidence. A meta-analysis of 15 randomized controlled trials (RCTs) including over 1400 women with pPROM showed that antenatal steroids did not increase the risk of maternal (RR 0.86; 95% CI 0.61–1.20) or neonatal infectious morbidity (RR 1.05; 95% CI 0.66–1.68) [38]. On the other hand, it showed that antenatal steroids reduced the risks of respiratory distress syndrome (RR 0.56; 95% CI 0.46–0.70) intraventricular haemorrhage (RR 0.47; 95% CI 0.31–0.70), and necrotizing enterocolitis (RR 0.21; 95% CI 0.05–0.82) [38]. Taken together, these substantial reductions in the risk of major gestational age-related morbidity translate to a significant reduction in neonatal mortality. Data on whether the same dose of corticosteroid translated to similar magnitudes of benefit for twins and other higher-order pregnancies with pPROM, because of greater maternal volume of distribution, are scanty. In diabetic women, the administration of corticosteroids may result in the loss of glycaemic control. This risk has to be carefully balanced with the potential benefit. Therefore, the value of steroid administration is questionable for women with pPROM near term – for example ≥32–34 weeks – or in whom fetal lung maturation can be demonstrated. Below 32 completed weeks of gestation, consideration should be given for sliding scale insulin/glucose infusion for 48–96 hours during the course of antenatal prophylactic steroid. Intramuscular betamethasone 12 mg 24 hours apart is preferable to dexamethasone 12 mg 12 hours apart. Betamethasone has a stronger evidence base and, unlike dexamethasone, it is not associated with necrotizing enterocolitis.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
