Meta-analysis: low-dose aspirin (modestly) reduces risk of pre-eclampsia.

A large meta-analysis of patient-level data has found that low-dose aspirin given to pregnant woman at risk of pre-eclampsia modestly but consistently reduced their risk of adverse outcomes. Pre-eclampsia affects between 2% and 8% of pregnancies and can have severe adverse effects on both mother and baby. While the cause is not yet known, abnormalities of clotting and platelet function occur and suggest the potential for benefit from anti-platelet drugs: many studies have been carried out over the past two decades, but their results have not been clear-cut. Previous systematic reviews suggested a benefit overall, however there was still controversy. The authors of this meta-analysis aimed to used patient-level data from as many previous trials as possible in an attempt to clarify the situation.

They carried out a comprehensive literature search using a relevant secondary database that is regularly updated from other databases. Eligible studies were randomised controlled trials of women at risk of pre-eclampsia treated for primary prevention with one or more anti-platelet agents, against controls of placebo or no treatment. Where potentially eligible trials included both primary and secondary prevention arms, only patients in the primary prevention arm were included in the analysis. Variables for the analysis were pre-specified, and anonymised data for patients in all eligible trials was requested from the original study authors; this was re-coded if necessary, checked for consistency, corrected where necessary, and finally agreed with the original authors. Four primary outcomes were defined: pre-eclampsia, death in utero or before hospital discharge, delivery pre-term at less than 34 weeks gestation, and infant small for gestational age. These were combined as an additional composite outcome - 'pregnancy with serious adverse outcome' (pregnancy where the mother dies or develops pre-eclampsia or if any baby is preterm, small for gestational age, or does not survive to discharge from hospital). Results were analysed on an intention to treat basis, but analysis for each outcome was restricted to trials having at least 80% of the data available for that outcome. Analyses compared the effect of the anti-platelet agent against placebo or no treatment for each outcome.

There were 115 trials identified initially, of which 50 were excluded as having no comparison group or treating women with established eclampsia only and two were excluded because patients were not truly randomised. Of the 63 remaining, including a total of 38,026 women, data could not be obtained for 27 (accounting for about 10% of the total participants; trial authors not traceable n=7, refused n=1, original data lost or irretrievable n=17, or not supplied n=2). This left 36 trials involving 34,288 women for which data could be analysed. Of these, 31, including 32,217 women and their 32,819 babies, were primary prevention studies and were thus included in this analysis. In 27, accounting for the great majority of the data, aspirin was given alone (dose 50mg to 150mg daily). Three small trials used only other anti-platelet drugs, and three tested aspirin and dipyridamole in combination. Just over half the women included were in their first pregnancy, 70% were aged 20 to 35, and 90% had at least one risk factor. Overall, 8% developed pre-eclampsia.

Compared to control, treatment with an anti-platelet agent was associated with a small but robust reduction in the relative risk of both pre-eclampsia (RR 0.90, 95% CI 0.84 to 0.97, p=0.004) and preterm birth before 34 weeks' gestation (RR 0.90, 95% CI 0.83 to 0.98, p=0.011). They also reduced the risk of the composite outcome to a similar extent (RR 0.90, 95% CI 0.85 to 0.96, p=0.001). There were also similar reductions in the other two primary outcomes (baby small for age, stillborn, or died before discharge), however these were not statistically significant. Maternal outcomes were similar for the two groups, with no significant increase in bleeding in the anti-platelet group. Subgroup analyses did not reveal any subgroup in which there was particular benefit.

The authors conclude that the data shows that treatment with an anti-platelet drug, mainly aspirin, reduces the risk of pre-eclampsia and some adverse pregnancy outcomes by about 10%. It is not possible to determine from the available data whether any particular subgroup benefits, although as the risk reduction is relative the absolute benefit will depend on a woman's underlying risk. The trials recruited mainly women at low to moderate risk of pre-eclampsia, so the data for women at high risk is limited. The analysis does not suggest that aspirin treatment is associated with significant adverse effects, although because some of the data on post-partum haemorrhage is uncertain, this outcome needs to be treated with caution.

An accompanying Comment discusses the paper and its implications. They authors discuss the possible mechanism of the benefit, and suggest questions that still need answers. They agree with the trial authors that use will depend on the mother's pre-existing risk level, and should be after full discussion of the potential risks and benefits. There have been a number of media reports of this study, and UK experts caution that pregnant women should not start taking aspirin without discussion with their doctor.

Lancet, published early online 18 May 2007; DOI:10.1016/S0140-6736(07)60712-0 (link to abstract); Lancet, published early online 18 May 2007; DOI:10.1016/S0140-6736(07)60713-2 (Comment; link to full text, available to subscribers only); BBC News report.