Premature Rupture Membranes and tPTL: a Personalised Approach (PROMPT)

Overview

Preterm birth (less than 37 weeks) affect approximately 8% of babies in the UK and is the worldwide leading cause of death in children under the age of 5. Subclinical infection affects approximately 50% of women giving birth before 32 weeks. Infection contributes to significant neonatal morbidity and mortality. Antibiotics such as erythromycin is currently used to treat women who present with preterm rupture of membranes. While this has shown short-term improvement in neonatal morbidity, it has not had any impact in reducing the perinatal mortality and also little effect on the health of the children at age seven. Some antibiotics such as co-amoxiclav has not shown to be effective in delaying delivery and some studies have shown that antibiotics increases rather than reduces the risk of cerebral palsy. Many women do not display signs of infection and the underlying bacteria is multifactorial (bacterial vaginosis, trichomoniasis, gonorrhoea, Chlamydia, ureaplasma, Group B streptococcal and E. Coli) and remains a diagnostic challenge. The only available clinical approach is to test the sample of amniotic fluid for bacteria and small case series have shown prolongation of pregnancy when accurately targeted antibiotic treatment is used. This research aims to prove that targeted antibiotic therapy results in a greater prolongation of pregnancy than standard management for women with preterm prelabour rupture of membranes (PPROM) and/or threatened preterm labour (tPTL). Women will be randomised to standard care versus BioFire directed antibiotic treatment in addition to standard care. Investigators will use the BioFire point of care testing to identify the presence of infection and identify with anti-microbial resistance genes the bacteria possess to guide the antibiotic treatment. To be certain that the presence of infection is detected the investigators will use PCR to test the amniotic fluid for IL-6 and white cell count.

Overall 10-15% of women delivering before 37 weeks have intraamniotic infection and of those delivering before 32 weeks have 50% chance. Consequently, since infection is one of the most important causes of spontaneous PTL it is possible that antibiotic treatment may delay delivery and reduce the risk of fetal damage and consequently long-term disability. However, most cases do not display typical symptoms or signs of infection, the only way to make the diagnosis reliably is through performing an amniocentesis. The risk of this procedure is low in the 3rd trimester and should be considered particularly in those in PTL at less than 32 weeks, when the risk of infection is high. Studies in non-human primate models of intra-amniotic infection support for this approach, showing that specific antibiotic therapy can prevent infection-induced PTL and prolong pregnancy with improved neonatal outcomes. Human studies have been less positive, with antibiotics failing to prolong pregnancy in PTL or eradicate infection in PPROM. The Oracle trial assessed the ability of two antibiotics, erythromycin and co-amoxiclav, to reduce the risk of PTD in women with PPROM using a composite primary outcome of neonatal death, chronic lung disease, or major cerebral abnormality on ultrasonography before discharge from hospital. Erythromycin, but not co-amoxiclav, reduced the risk of the primary outcome. Two recent papers have challenged this negative view of the use of antibiotics for intraamniotic infection. In the first, amniocentesis was performed to identify infection or inflammation and broad-spectrum antibiotics administered to women with a singleton pregnancy presenting with tPTL at between 20 and 34 weeks. Infection was thought to be present in at least 60% of cases as judged by the presence of a raised WCC, although only 20% had positive cultures using standard microbiological approaches. 100% had an elevated IL-6 level confirming inflammation. Repeat amniocentesis confirmed resolution of inflammation in all cases and overall average prolongation was 11.4 days. In a control group not given antibiotics average prolongation was 3 days. In 3 cases infection was shown to be cleared by antibiotic treatment. In the second report, women undergoing rescue cerclage had an amniocentesis and those found to have evidence of infection or inflammation treated with broad spectrum antibiotics. A repeat amniocentesis demonstrated resolution of infection/inflammation with antibiotic treatment. Investigators plan to recruit women to a multicenter randomised trial of standard management vs. amniocentesis with the use of BioFire filmArray point of care testing to identify the presence of bacteria and identify common anti-microbial resistance genes the bacteria possess to guide the selection of an additional antibiotic agent. The BioFire filmArray clinical application is currently being studied in patients with pneumonia in the INHALE study. Investigators plan to use the BioFire filmArray in a maternity setting and will use the information generated by the BioFire filmArray platform to give targeted antibiotics in the amniocentesis arm in addition to routine erythromycin with the aim of prolonging latency to delivery and of improving neonatal outcome.