Article Text
Abstract
The neuro-protective effect of antenatal magnesium sulfate on very preterm infants has been demonstrated in good-quality randomised controlled trials and meta-analyses. Magnesium administered prior to preterm delivery crosses over to the foetal circulation and acts via several pathways to reduce perinatal neuronal damage. Meta-analysis of the trial data indicates that antenatal magnesium sulfate reduces the risk of cerebral palsy by one-third, and results in one fewer case in every 50 women treated. Treatment is associated with discomfort and flushing in some women, but maternal side-effects are mostly transient and manageable. Magnesium sulfate has also been found to be without any serious adverse consequences in newborn infants. Consensus recommendations and guidelines have been developed and implemented internationally, and endorsed by the UK Royal College of Obstetricians and Gynaecologists. However, magnesium sulfate for neuro-protection of very preterm infants has not yet become established widely in UK practice. Paediatricians, neonatologists and advocacy groups for preterm infants and their families could contribute to raising awareness and engage in dissemination activities and implementation initiatives to develop local protocols for adoption of this safe, effective and cost-effective intervention to reduce the burden of cerebral palsy in children born very preterm.
- Evidence Based Medicine
- Multidisciplinary team-care
- Neurodevelopment
- Neonatology
- Neurodisability
Statistics from Altmetric.com
Over the past five decades, the development and adoption of key evidence-based care practices and interventions have improved important outcomes for preterm infants substantially. In particular, the use of antenatal corticosteroids and postnatal surfactant replacement has reduced dramatically early mortality due to respiratory failure in very preterm (<32 weeks’ gestation) neonates. However, major long-term neurological morbidities occur in surviving infants. Developing, assessing and implementing strategies to reduce the incidence and severity of neuro-disability, particularly cerebral palsy, the most common physical disability in childhood, is a priority in perinatal care.1–4
Cerebral palsy
Cerebral palsy, a group of non-progressive, activity-limiting disorders of movement or posture, occurs in around 2.1 infants per 1000 live births. The prevalence of cerebral palsy is inversely related to gestation at birth (table 1). One-quarter of all cases of cerebral palsy occur in infants born before 34 weeks’ gestation.5 ,6
Most preterm-born children with cerebral palsy are able to walk independently.7 ,8 However, the severity of cerebral palsy increases with decreasing gestational age at birth. More than one-third of infants with cerebral palsy born before 26 weeks’ gestation are non-ambulant.9 Co-existing neurological or developmental impairment is common and these co-morbidities often limit daily activities, quality of life and independence to a greater degree than the motor impairment itself (table 2).7
Aetiology of cerebral palsy in preterm infants
The causes of antenatal and perinatal brain injury resulting in cerebral palsy in preterm infants remain to be fully defined but are likely to include damage arising from hypoxic and reperfusion injuries, as well as infectious and inflammatory processes. Several mediators and pathogenic pathways contribute to neuronal death, including the release of excitatory amino acids, pro-inflammatory cytokines and oxygen-free radicals. A major research challenge is to develop methods to limit or prevent perinatal neuronal damage and reduce the severity or prevalence of cerebral palsy and associated co-morbidities following preterm birth.4
Magnesium
Antenatal magnesium sulfate given to women at imminent risk of very preterm delivery is currently the intervention supported most strongly by a robust body of evidence as being effective in preventing cerebral palsy in newborn infants.10 Magnesium is transferred across the placenta to the foetal circulation within 1 h of maternal intravenous administration.11 In the cell, magnesium is a key mediator in pathways implicated in apoptosis secondary to inflammation and hypoxic–ischaemic injury. The putative neuro-protective effects of magnesium include competitively reducing intracellular calcium entry, blocking glutamate and other excitatory neurotransmitter receptors and modulating the actions of pro-inflammatory cytokines and oxygen free radicals (box 1).12 ,13 Magnesium is also vasoactive and has potential haemodynamic benefits for the foetus and newborn infant, including stabilising blood pressure and cerebral arterial perfusion.11 ,15
Magnesium-mediated cellular processes14
Intermediary metabolism
glycolysis
oxidative phosphorylation
Membrane integrity
Neuromuscular excitability
nerve conduction
muscle contraction
Protein synthesis
Nucleic acid aggregation
Magnesium sulfate in peripartum care
Tocolysis
Magnesium tocolysis (intravenous magnesium sulfate) has been assessed in clinical trials since the 1970s.16 The recent Cochrane review of magnesium sulfate for preventing preterm birth in threatened preterm labour identified 37 eligible randomised controlled trials (in which a total of 3571 women participated) but concluded that magnesium sulfate is “ineffective at delaying birth or preventing preterm birth, [and] has no apparent advantages for a range of neonatal and maternal outcomes as a tocolytic agent”.17 Despite this conclusive evidence, magnesium tocolysis remains widely used in obstetric care in several countries including the USA.
Pre-eclampsia and eclampsia
Magnesium is well established as an effective intervention for treating women with pre-eclampsia. Meta-analysis of data from six trials (11 444 women with pre-eclampsia) demonstrates that magnesium sulfate versus placebo or no anticonvulsant reduces the risk of eclampsia by more than half, without affecting the incidence of stillbirth or neonatal death.18 There is also strong trial evidence that magnesium sulfate is more effective that other anticonvulsants (such as diazepam or phenytoin) in treating women with eclampsia.19 Following the dissemination of these findings, and their incorporation into national and international policy statements and guidelines, magnesium sulfate has become the drug of choice for the peripartum care of women with pre-eclampsia and eclampsia.20
Neuro-protection for preterm infants
Observational data from studies examining the use of magnesium sulfate for tocolysis or for treating pre-eclampsia first indicated the potential neuro-protective effects for preterm infants. Cohort and case-control studies reported strong associations between exposure to antenatal magnesium sulfate and protection from cerebral palsy in very low birth weight infants.21 ,22 These hypothesis-generating studies prompted clinical investigators internationally to undertake randomised controlled trials to provide unbiased evidence about the neuro-protective effects of magnesium in this population.23
Randomised controlled trials
Over the past 20 years, investigators have undertaken five randomised controlled trials (with 6145 participating infants in total) that have assessed the effect of antenatal magnesium sulfate on mortality and serious morbidity, including cerebral palsy, in preterm infants (table 3).24–28
Four trials sought to assess primarily the neuro-protective effects of magnesium sulfate on newborn infants.24–28 The Magpie Trial assessed the effect of magnesium sulfate treatment of pre-eclampsia on neurological outcomes for newborn infants.26 The methodological quality of the trials was good, with an overall low risk of bias. The three largest trials were ‘double-blinded’ and neurological outcomes were assessed by an investigator who was unaware of the treatment allocation in all of the trials.
The Cochrane review of these trials concluded that antenatal magnesium sulfate given to women at imminent risk of preterm delivery substantially reduced the risk of cerebral palsy in the child.23 Meta-analysis indicates that the risk of cerebral palsy is almost one-third lower in infants who received antenatal magnesium sulfate:
Risk ratio 0.68 (95% CI 0.54 to 0.87)
Risk difference −0.02 (95% CI −0.03 to −0.01).
The trials showed comparable results and the direction of effect on cerebral palsy was consistent. The meta-analysis was not statistically heterogeneous, indicating that the pooled estimate of effect is robust (figure 1). This effect is consistent with one fewer case of cerebral palsy in every 50 women treated with magnesium sulfate (95% CI 33 to 100).
More uncertainty exists about the effect of antenatal magnesium sulfate on longer term neurodevelopment, including functional, cognitive and behavioural outcomes. To date, only two trials have reported outcomes for children at school age, and their findings need to be interpreted cautiously because more than 20% of participants have not been available for assessment.29 ,30 Meta-analyses of data from school age assessments of participants in the large randomised controlled trials are needed to provide precise estimates of the effect on long-term outcomes.
Side effects
Magnesium sulfate infusion is associated with several maternal unpleasant side effects. Women commonly experience a feeling of warmth, facial flushing and discomfort at the intravenous cannula site. In the trials, women receiving magnesium sulfate were three times more likely than controls to decline to continue receiving the intervention because of these effects, but this was uncommon (8% of participants). At high serum levels, magnesium can cause muscle weakness and more serious side effects including respiratory depression are possible, but this is rare and is mainly associated with excessive doses.32
In infants, the effect of magnesium blockage of calcium entry into cells may theoretically be hypotonia, respiratory depression and apnoea requiring respiratory support. However, analyses of data from BEAM (Beneficial Effects of Antenatal Magnesium Sulfate), the largest trial of magnesium for neuro-protection, did not find a difference in the rates of respiratory depression at birth. More recent data from large cohort studies have not demonstrated an increased need for intensive delivery room resuscitation in very preterm infants antenatally exposed to magnesium sulfate.32 ,33
Costs
Magnesium sulfate is inexpensive but other cost implications for prophylactic use include the requirement for administration via an infusion pump as well as the need for staff to accompany and monitor women receiving magnesium sulfate. Formal analyses indicate that antenatal magnesium sulfate for neuro-protection of very preterm infants is a highly cost-effective intervention in the prevention of cerebral palsy and improving the quality of life.34
Individual data meta-analyses
The ‘trial level’ meta-analyses are not able to answer precisely further key questions about the use of magnesium sulfate for neuro-protection and an international collaboration of the trial teams and methodologists is currently undertaking an individual participant data meta-analysis (participant level data supplied by all trials) to determine whether or how pre-specified participant and treatment characteristics affect important outcomes (box 2).35 These refined analyses may further inform translation and application of the policy into practice.
Pre-specified factors for individual participant data meta-analyses35
Primary cause of preterm birth
spontaneous preterm labour
pregnancy-induced hypertension
antepartum haemorrhage
Presence or absence of ruptured amniotic membranes
Primary indication for antenatal magnesium sulfate
neuro-protection
pre-eclampsia
tocolysis
Gestational age
Plurality (singleton or multiple)
Time from treatment to delivery
Mode of administration and dose
loading
maintenance regimen
repeat administration
Are more trial data needed?
Some commentators have also argued that the existing data should be applied cautiously because of concerns that the meta-analyses demonstrating beneficial effects are imprecise.36 ,37 Consequently, investigators in Scandinavia are now undertaking a new large randomised controlled trial in which 1240 women at risk of delivery before 33 weeks’ gestation are being invited to participate. This trial is not designed as ‘stand-alone’, but uses an innovative and efficient design and sample size calculation based on integrating the data into the existing meta-analysis.38
International guidelines
In 2010, authorities in the USA and Australasia introduced guidelines recommending the use of magnesium sulfate for neuro-protection of very preterm infants, a process expedited perhaps because most of the primary trials and systematic reviews were developed and delivered in those countries.39 ,40 More recently, several other national bodies, including those in Canada and the Republic of Ireland, have developed similar practice guidelines and recommendations.
Implementation into practice has been supported by quality improvement programmes and by integration into benchmarking and audit cycles.41 The American College of Obstetricians and Gynaecologists guideline does not specify a gestational age threshold for use but recommends that services ‘develop specific guidelines around the issues of inclusion criteria, dosage, concurrent tocolysis and monitoring in accordance with one of the larger trials’. The Australasian guidelines recommend that magnesium sulfate for neuro-protection is restricted to women at imminent risk of delivery before 30 weeks’ gestation, arguing that uncertainty on the effects for infants born after longer gestations still remains. Currently, the Australasian investigators are undertaking a new trial to assess whether magnesium sulfate versus placebo given to women with threatened, imminent delivery between 30 and 34 weeks’ gestation affects rates of mortality and neuro-disability, including cerebral palsy.42
UK recommendations and practice
Magnesium sulfate for neuro-protection is a less well-established policy and practice in the UK than in North America and Australasia. The UK Royal College of Obstetrics and Gynaecologists (RCOG) acknowledges in a Scientific Impact Paper that the supporting evidence base is strong, that maternal side effects are generally mild and manageable, and that obstetricians and midwives are already familiar with magnesium sulfate for management of pre-eclampsia.43 The RCOG suggests several possible factors that may be contributing to the delay in adopting this intervention in the UK. These include concerns about side effects in infants, the large number needed to treat for benefit (especially as compared with maternal administration of corticosteroids to prevent respiratory distress syndrome) and uncertainty about the applicability of data from trials where prevention of cerebral palsy was not the pre-specified primary outcome. This last point may reflect a misunderstanding of the nature of the systematic review process, where a published protocol with a priori outcome definitions reduces the impact of selective reporting biases.
Another element that may be contributing to delay in adoption relates to the uncertainty about the diagnosis of preterm labour. If magnesium sulfate is given to every woman in suspected preterm labour, many would not actually progress to preterm delivery and would have received treatment unnecessarily. Although this scenario is the same as that encountered when considering antenatal corticosteroid treatment for women with threatened preterm delivery, the difference with magnesium sulfate is the need for an intravenous infusion, and midwife supervision and monitoring in a labour ward setting.
The RCOG endorses the Australasian recommendations, and suggests that these are appropriate sources for developing local protocols (table 4). However, this advice has not yet translated into an RCOG formal ‘Green-top’ guideline which would lead to an expectation of the practice being adopted.
Implementation- policy into practice
The history of the adoption of antenatal corticosteroids for women with threatened preterm delivery has lessons for the adoption of magnesium sulfate for neuro-protection for very preterm infants.44 The largest trial of antenatal corticosteroids for women at risk of preterm birth (1070 infants) was published in 1972. This and further trials undertaken during the 1970s and 1980s provided consistent evidence of benefit for preterm infants (figure 2). Despite this, most women with impending preterm delivery during that era did not receive antenatal corticosteroids and further trials continued to be approved, funded and undertaken. It was not until the mid-1990s that clear recommendations by national bodies drove changes in practice and widespread adoption of antenatal corticosteroids for women with impending preterm delivery. It is likely that many thousands of preventable deaths occurred worldwide while neonatologists and obstetricians delayed implementing the findings of the early trials.45
The adoption of the consensus guidelines on the use of magnesium sulfate for neuro-protection of very preterm infants in Australasia has been supported by quality improvement initiatives integrated within clinical networks and regional and national audit and benchmarking processes. This facilitates identification of local barriers and facilitators to implementation, and use of specific strategies tailored to the needs of individual centres.46 Such embedded approaches are needed within the UK, with consultation and collaboration between clinicians (midwives, obstetricians, paediatricians), and service-user groups to ensure the success of any implementation strategy. Neonatologists and neonatal nurses and infant and family-advocacy groups, have central roles in raising awareness in local, regional and national fora to advance discussion and reduce delays in the introduction of this intervention. On an individual case basis, neonatologists and paediatricians can reinforce the message to midwifery and obstetric colleagues when coordinating perinatal care for a woman with threatened preterm delivery: “You've administered antenatal steroids, thanks, have you considered magnesium sulphate?”
References
Footnotes
Contributors All authors contributed to this review.
Funding SO and WM are funded by the National Institute for Health Research (NIHR) under its Programme Grants for Applied Research funding scheme (RP-PG-0609-10107) and an NIHR Cochrane Programme Grant (13/89/12).
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
Linked Articles
- Fantoms