Purpose: To review the effect of intravenous magnesium in obstetrics on fetal/neonatal neuroprotection.
Methods: A systematic review of published studies.
Results: Five randomized trials and 4 meta-analyses have shown a significant 32% reduction of cerebral palsy when administering magnesium sulfate in case of preterm delivery. The pathophysiologic mechanism is not fully unraveled: modulation of the inflammatory process, both in the mother and the fetus, and downregulation of neuronal stimulation seem to be involved. After long-term high-dose intravenous administration of magnesium, maternal and neonatal adverse effects such as maternal and neonatal hypotonia and osteoporosis and specific fetal/neonatal cerebral lesions have been described. In case of administration for less than 48 h at 1 g/h and a loading dose of 4 g, these toxic amounts are not achieved. American, Canadian and Australian guidelines recommend the use of intravenous magnesium in any threatening delivery at less than 32 weeks. The "number needed to treat" to avoid 1 cerebral palsy is between 15 and 35.
Conclusions: Intravenous magnesium significantly reduces the risk for cerebral palsy in preterm birth. Open questions remain the optimal dosing schedule, whether or not repeating when delivery has been successfully postponed and a new episode of preterm labor occurs. Some concern has been raised on a too optimistic value for random error which might have led to over-optimistic conclusions in classic meta-analysis. Randomized trials comparing different doses and individual patient data meta-analysis might resolve these issues.