Journal Club Papers
Paper 1: Short Duration of Antenatal Corticosteroid Exposure and Outcomes in Extremely Preterm Infants.
Chawla et al.
JAMA Network Open. (21 Feb 2025). DOI:10.1001/jamanetworkopen.2024.61312.
Contribution from Dr Hannah Gordon.
Antenatal corticosteroids reduce the risk of neonatal morbidity and mortality among infants born extremely preterm (22+0– 27+ 6 weeks’ gestation). The greatest benefit is conferred following two doses of 11.4mg betamethasone 24 hours apart; but in the context of unexpected, precipitate or urgently indicated preterm birth, two doses are not always feasible. What is the benefit to infants birthed before the second dose can be administered? Chawla et al have attempted to fill this knowledge gap.
This study retrospectively analysed data for infants born 22-27+6 weeks’ gestation across 15 tertiary centres across the United States. They included infants who had received no steroids or only ONE dose of betamethasone, and excluded infants >1500g, or where steroid exposure had been for >24 hours.
The primary outcome was survival to hospital discharge. The secondary outcome was survival at 36 weeks’ postmenstrual age free from a composite of major neonatal complications including: severe intracranial haemorrhage, cystic periventricular leukomalacia, surgical necrotising enterocolitis, severe bronchopulmonary dysplasia, and severe retinopathy of prematurity.
Findings:
· The study included 1806 infants; 1331 (73.7%) who received one dose of steroids and 475 (26.3%) receiving no steroids.
· They found a positive association between time from first steroid exposure and survival to hospital discharge.
· For each hour between first steroid exposure and birth, there was a 1% increase in survival to hospital discharge (adjusted relative risk (aRR) 1.01 [95% confidence interval (CI) 1.00–1.01]).
· There was also a 1% increase in survival free from major neonatal morbidity for every hour between first exposure to steroids and birth (aRR 1.01 [95% CI: 1.01–1.02]).
Results were adjusted for variables considered to influence survival, including gestational age, sex, mode of birth, multiple pregnancy, socioeconomic status and small for gestational age.
While 1% per hour does not seem much, it should be noted that death is a very hard endpoint.
Examining pregnancy and birth characteristics by time from first dose to birth, they found those with a shorter time to birth (1.4h) were more likely to have chorioamnionitis, antepartum haemorrhage, vaginal breech birth and need advanced resuscitation and surfactant. Those who birthed >9.5 hours after first steroid dose were more likely to have hypertension in pregnancy and a small for gestational age fetus. Thus, it is likely that the timing of steroid administration varies according to preterm birth type (spontaneous vs iatrogenic). And it is also possible that- if a longer time is available from steroid exposure to birth - other unmeasured interventions may have helped improve neonatal outcomes, such as: more time for evaluation, maternal stabilisation or administration of medications such as antibiotics or Magnesium Sulphate.
Despite these concerns, this paper provides important reassurance that every hour counts for the benefit of antenatal corticosteroids among infants facing extremely preterm birth.
Paper 2: Intrapartum sildenafil to improve perinatal outcomes: A randomized clinical trial.
Kumar et al.
JAMA. 09 June 2025. DOI: 10.1001/jama.2025.7710.
Hypoxic-ischaemic injury during labour occurs when the infant does not receive enough oxygen, often due to inadequate placental reperfusion time between contractions and/ or underlying placental insufficiency. Hypoxic ischemic injury is an important cause of intrapartum stillbirth, neonatal death and newborn encephalopathy - particularly in low- resource settings. A treatment that prevents hypoxic- ischemia during labour would be hugely beneficially globally.
Sildenafil citrate is a phosphodiesterase inhibitor and a potent vasodilator. Previous large trials showed administering sildenafil in pregnancies complicated by severe fetal growth restriction revealed no benefit, and may even caused fetal harm (potential increased risk of neonatal pulmonary hypertension). Hence, while it use cannot be supported for this indication, a previous publication by this team suggested short term use during labour could reduce the risk of operative birth for fetal distress. This is plausible – it might open up vessels to increase maternal blood supply to the placenta and baby.
To test this, Kumar et al., undertook a pragmatic, multicentre, placebo-controlled, double-blind RCT conducted at 13 hospitals in Australia.
They enrolled women with a singleton or dichorionic twin pregnancy and were planning a vaginal birth at term (≥37 weeks’ gestation).
They were randomised after the onset of labour (either spontaneous or induced) to either 50mg sildenafil or placebo orally every eight hours during labour, up to three doses.
The primary outcome was a composite of ten perinatal outcomes, including: intrapartum stillbirth, neonatal death <28 days, 5-minute Apgar <4, acidosis at birth, hypoxic-ischaemic encephalopathy, neonatal seizures, neonatal respiratory support >4 hours, NICU admission >48 hours, persistent pulmonary hypertension of the newborn, or meconium aspiration syndrome. There were several secondary and tertiary outcomes.
The expected rate of primary outcome was 7%, and the trial was powered for a 35% reduction.
Findings
1626 women were randomised to sildenafil and 1631 to placebo, with 1625 infants from each group available for follow up. Most (58%) of women were nulliparous, and induction of labour occurred in approximately 85% of cases.
· The primary outcome occurred in 83 infants in the sildenafil group (5.1%) vs 84 infants in the placebo group (5.2%). After adjusting for maternal clustering due to twin pregnancies, there was no significant difference between the groups (Relative risk (RR) 1.02, 95% CI 0.75-1.37).
· There were no cases of intrapartum stillbirth or neonatal death, and no difference between the sildenafil and placebo groups for any of the components of the primary outcome.
· There were also no differences in any of the secondary or tertiary outcomes, including emergency operative birth, length of labour, length of hospital stay, maternal ICU admission, cord prolapse, intrapartum bleeding, abruption, or chorioamnionitis.
This large, well-conducted trial disappointingly found no difference in the risk of adverse perinatal outcomes or emergency operative birth among women randomised to sildenafil in labour at term, compared with placebo.
There are a few caveats. The primary outcome occurred less commonly than expected (5% vs 7%) meaning the trial was underpowered. Secondly, pragmatic multicentre trials such as this will inevitably have some heterogeneity between sites: the primary outcome varied between 1.8-9%. But overall, the hopeful story of sildenafil improving outcomes in pregnancy or birth seems to be over.
This trial completed recruitment of over 3000 women in just over 33 months- which demonstrates the power of a working together to run a collaborative perinatal trial network in Australia (and congratulations to the Aussie team for a superb piece of research 👏 ).
Paper 3: Induction of labour versus standard care to prevent shoulder dystocia in fetuses suspected to be large for gestational age in the UK (the Big Baby trial): a multicentre, open-label, randomised controlled trial.
Gardossi et al.
The Lancet. 2025. https://doi.org/10.1016/ S0140-6736(25)00162-X
Shoulder dystocia can lead to serious maternal and neonatal complications including brachial plexus injury, fractures, and neonatal death. Macrosomic fetuses are at higher risk of shoulder dystocia. Therefore induction of labour has been proposed for large-for-gestational-age fetuses at the end of pregnancy to reduce this risk.
The most recent Cochrane review concludes inducting labour reduces the risk of shoulder dystocia by 40% (RR 0.60, 95% CI 0.37–0.98), and fracture by 80% (RR 0.20, 95% CI 0.05-0.79), albeit with no significant effect on brachial plexus injury. However, this meta-analysis included <1200 births, of which more than 800 came from the Boulvain (et al) trial. In that trial, women with a suspected large fetus (confirmed on ultrasound to be >95th centile) were scheduled for induction between 37+0 and 38+6 weeks. The Boulvain et al trial leaves uncertainty about whether induction is effective at a later gestation. This is important given the renewed focus on preventing early term birth- and/ or in fetuses >90th percentile.
To address this, Gardosi et al. conducted an open-label, multicentre RCT across 106 centres including pregnancies with a fetus estimated to be >90th percentile in weight (based on ultrasound performed from 35 – 38 weeks’, using GROW customised charts). Women with gestational or preexisting diabetes requiring drug treatment were excluded.
Women were randomised to induction of labour between 38+0 – 38+4 weeks’ gestation or standard care. The primary outcome was incidence of shoulder dystocia ‘requiring additional obstetric manoeuvres’, determined by 2 independent panellists.
Findings:
· The trial aimed to recruit 4,000 participants but was stopped at 2,893 as the incidence of shoulder dystocia was lower than expected. It was calculated that the new sample size needed was an infeasible 12,844!! Thus, the trial was underpowered.
· Analysing the already recruited 2893 participants; 1147 were induced vs 1146 in the standard care group.
· For the primary outcome, induction of labour did not significantly reduce shoulder dystocia; 2.3% (33/1145) induction group vs 3.1% (44/1439) standard care, relative risk 0.75 (95%CI 0.51, 1.09).
· However, induction did not result in the anticipated large differences in gestation or birthweight between the two arms that the team had hoped for. With most in the standard care group delivering only about week later (mean difference in gestation -6 days), the mean difference in birthweight was only 163.6grams. Additionally, only 40% of infants were born >90th percentile in both groups (likely due to technical limitations of trying to diagnose large for gestational age at the 36 week ultrasound).
· These smaller than expected differences were thought to be due to significant cross over between the groups, with 24% of women in the standard care group induced or birthed <38+4 days.
· Some crossover was expected and the authors had pre-planned a per protocol analysis to overcome this. In this follow ‘per protocol’ analysis, anyone deviating from the randomisation assigned to them were excluded.
· Per protocol analysis (ie outcomes for those who actually birthed according to their allocated groups): induction of labour significantly reduced the risk of shoulder dystocia: 2.3% (27/1180) induction group vs 3.7% (40/1074) standard care group; relative risk 0.62 (95%CI 0.41, 0.92).
· Unsurprisingly, the per protocol analysis resulted in larger differences in gestation at birth (-8.1 days) and birthweight (-213.3grams) between the groups.
· Among secondary maternal outcomes there was a significant reduction in both elective and emergency caesarean section and postpartum haemorrhage (≥ 500ml) in the induction group.
· For secondary neonatal outcomes there were no differences in intrapartum birth injury (fractures, brachial plexus injury, or both); RR 1.95 (0.36 to 10.65), or prematurity associated problems (use of phototherapy, respiratory support, or both); RR 1.16 (0.87 to 1.56). However, the trial was significantly underpowered to detect these.
The Big Baby trial is the largest study to date investigating induction of labour among pregnancies suspected to be LGA. The number of participants was greater than all patients in the Cochrane review combined. Despite this, it remained underpowered for the primary outcome, due to the lower-than-expected incidence of shoulder dystocia in the control group. This was partly because 25% of patients in the control group delivered before 38+4 weeks: this may reflect secular trends in current practice where babies suspected to be LGA are birthed earlier.
When confined to participants who were managed as per protocol, significant differences were observed.
Another takeaway is the modest positive predictive value of ultrasound in identifying babies expected to be big: only 40% of those suspected to be large for gestational age on US were found to be >90th centile at birth. This means finding the appropriate target group for timed birth is challenging.
So where do the results of this heroic trial leave us? It seems unlikely the trial will be repeated, which means that- despite the uncertainty- we need to make decisions based on this information we’ve got.
We suspect the findings are likely to entrench ‘a priori’ views of clinicians. Those who do not believe induction is useful will note the primary outcome is negative. Those who believe it is useful may note the per protocol analysis (which is a secondary analysis) is positive.
For those in equipoise, perhaps the middle ground might be induction at around 39 weeks, and at a higher threshold of EFW >95th centile.