INTRODUCTION — In utero exposure to magnesium sulfate before early preterm birth appears to decrease the incidence and severity of cerebral palsy. Cerebral palsy is the leading cause of neurologic impairment in young children, and preterm birth and low birth weight are the most important risk factors for developing the disease. (See "Cerebral palsy: Epidemiology, etiology, and prevention".)
The use of magnesium sulfate before birth for neuroprotection will be reviewed here. Use of magnesium sulfate for tocolysis or for seizure prevention in preeclampsia/eclampsia is discussed separately. (See "Inhibition of acute preterm labor", section on 'Magnesium sulfate' and "Preeclampsia: Intrapartum and postpartum management and long-term prognosis", section on 'Regimen'.)
●Stabilization of cerebral circulation by stabilizing blood pressure and normalizing cerebral blood flow.
●Prevention of excitatory injury by stabilization of neuronal membranes and blockade of excitatory neurotransmitters, such as glutamate.
●Protection against oxidative injury via antioxidant effects.
●Protection against inflammatory injury via anti-inflammatory effects.
Although maternal administration of magnesium sulfate has not been associated with a decreased risk of fetal brain injuries that are believed to cause cerebral palsy (eg, severe intraventricular hemorrhage or cystic white matter injury [3-6]), it has been associated with a reduction in cerebellar hemorrhage in preterm infants .
EVIDENCE OF EFFICACY FROM RANDOMIZED TRIALS AND META-ANALYSES — Evidence of the neuroprotective effects of magnesium sulfate is based on data from randomized trials and meta-analyses of these trials. There is significant heterogeneity among the three major trials designed specifically to assess the neuroprotective benefits of magnesium sulfate: the Australasian Collaborative Trial of Magnesium Sulfate (ACTOMgSO4) , the Beneficial Effects of Antenatal Magnesium Sulfate (BEAM) trial , and the PREMAG trial [5,8]. These trials are summarized in the table (table 1). It is important to emphasize that the combined outcome of "cerebral palsy or death" is important methodologically, as cerebral palsy and death are competing outcomes (ie, death precludes subsequent diagnosis of cerebral palsy).
Death and/or cerebral palsy — Multiple meta-analyses have evaluated the neuroprotective effects of magnesium sulfate administered to patients at risk of preterm birth and have consistently found a reduction in cerebral palsy in offspring [9-13]. The Cochrane meta-analysis of magnesium sulphate for women at risk of preterm birth for neuroprotection of the fetus is a representative example of this evidence . Five randomized trials were included: The three trials described above were specifically designed to evaluate the neuroprotective effect of magnesium sulfate [3-5,8], a fourth trial had both a tocolytic arm and a neuroprotective arm , and a fifth trial, Magpie, was designed to assess the efficacy of magnesium sulfate for prevention of seizures in patients with preeclampsia . These trials included 6145 infants, of whom 1493 were born preterm.
For subgroup analysis, studies were classified by their primary study intent: neuroprotection [3,4,6,8] or "other," which included seizure prophylaxis (Magpie trial ) in preeclampsia or tocolysis of preterm labor . For the purpose of the review, the trial with tocolytic and neuroprotective arms was treated as two separate trials, with the subgroup of patients in the neuroprotection arm considered along with the other trials designed specifically to assess the neuroprotective role of magnesium sulfate.
Major findings were:
●Cerebral palsy – "Any cerebral palsy" was significantly reduced in the overall exposed population (relative risk [RR] 0.68, 95% CI 0.54-0.87). The absolute risk of cerebral palsy was 3.4 percent for fetuses exposed to antenatal magnesium sulphate therapy versus 5 percent for unexposed fetuses, giving an absolute risk reduction of 1.6 percent. The number of patients that would need to be treated to prevent one child from developing cerebral palsy was 63 (95% CI 43-87).
In the neuroprotection trials subgroup, both "any cerebral palsy" and "moderate/severe cerebral palsy" were significantly reduced (RR 0.71, 95% CI 0.55-0.91 and RR 0.64, 95% CI 0.44-0.92, respectively).
●Gross motor dysfunction – Substantial gross motor dysfunction was significantly reduced overall (RR 0.61, 95% CI 0.44-0.85) and in the neuroprotection trials subgroup (RR 0.60, 95% CI 0.43-0.83).
●Death or cerebral palsy – "Death or cerebral palsy" was not significantly reduced in the overall exposed population (RR 0.94, 95% CI 0.78-1.12) but was reduced in the neuroprotection trials subgroup (RR 0.85, 95% CI 0.74-0.98).
●Stillbirth or pediatric death – "Stillbirth or pediatric death" was not significantly reduced overall (RR 1.04, 95% CI 0.92-1.17) or in the neuroprotection trials subgroup (RR 0.95, 95% CI 0.80-1.12).
Interpretation — Given the concern about competing outcomes, the finding that magnesium sulfate had no significant effect on the combined outcome of "death or cerebral palsy" in the overall group of trials merits discussion. It is possible that the nonsignificant difference of the combined outcome is due to an increased risk of fetal or infant death in a subgroup of magnesium exposed fetuses, as suggested by two trials [4,6]. For this reason, some clinicians do not administer magnesium sulfate for neuroprotection, despite the protective effect against cerebral palsy. On the other hand, the combined outcome of "death or cerebral palsy" was reduced significantly when only trials designed specifically to assess the neuroprotective effect of magnesium sulfate were analyzed; we believe this provides significant reassurance that the decrease in occurrence and severity of cerebral palsy is not achieved as a result of an increased risk of death. This conclusion is supported by 2019 systematic review including 197 studies (40 randomized and 138 nonrandomized, 19 case reports) that evaluated the outcome of infants exposed in utero to magnesium sulphate and found that magnesium sulfate given for neuroprotection was not associated with an increased risk of perinatal death or other adverse infant outcomes .
Other short-term outcomes — Short-term pediatric outcomes other than death and cerebral palsy have also been evaluated by meta-analysis, which found similar risks of blindness, deafness, and developmental delay in magnesium exposed and nonexposed pregnancies . Importantly, magnesium exposure did not decrease the frequency of adverse effects, such as Apgar score less than seven at five minutes, intraventricular hemorrhage, periventricular leukomalacia, neonatal seizures, or need for ongoing respiratory support. The effect of gestational age at randomization (<30 versus <32 to 34 weeks) did not alter the findings significantly. (See 'Lower and upper gestational age' below.)
Outcomes at school-age — Only one of the randomized trials discussed above, ACTOMgSO4, reported on long-term outcomes after magnesium sulfate prophylaxis and did not find a long-term benefit. Data were available for 669 children (77 percent of the original study) at a mean age of 8.4 years, corrected for preterm birth . Rates of cerebral palsy, abnormal motor function, and other neurologic, cognitive, and behavioral outcomes at school age did not differ significantly between magnesium exposed and nonexposed pregnancies, but magnesium exposure was associated with a trend towards improved survival (RR 0.80, 95% CI 0.62-1.03).
The lack of benefit at school age in this trial does not negate the benefit of magnesium sulfate therapy in decreasing cerebral palsy risk, given results of the meta-analyses of randomized trials discussed above. There are significant differences in patient population and study protocols among the three large randomized trials, and thus long-term outcome data from other trials are needed before making definitive conclusions about long-term outcomes.
Background — The following discussion reflects the authors' approach to use of magnesium sulfate for neuroprotection before preterm birth. This approach is based on the evidence cited above (see 'Evidence of efficacy from randomized trials and meta-analyses' above) and in the sections below and is generally consistent with American College of Obstetricians and Gynecologists guidance .
The precise dose, duration, and timing of administration of magnesium sulfate to maximize the neuroprotective benefit has not been established . Others have taken a slightly different approach, including administering magnesium sulfate as early as 23 weeks of gestation, using a 6 g loading dose with a 2 g/hour maintenance dose, stopping treatment after 12 hours, and retreating patients who meet criteria for treatment and have been off therapy for more than six hours.
Candidates for treatment — Pregnancies at high risk of preterm birth within 24 hours are appropriate candidates of magnesium sulfate neuroprotection. This includes pregnancies with recent preterm prelabor rupture of membranes, preterm labor with intact membranes, or planned medically or obstetrically indicated preterm delivery.
Patients enrolled in the various trials of magnesium sulfate for neuroprotection represented the full range of obstetric indications for preterm birth, but differed significantly in the proportion of subjects comprising each indication. For example, in the Australasian Collaborative Trial of Magnesium Sulfate (ACTOMgSO4)  and PREMAG  trials, 63 to 88 percent of patients were in preterm labor. In contrast, in the Beneficial Effects of Antenatal Magnesium Sulfate (BEAM) trial , the majority (85 percent) of patients had preterm prelabor rupture of membranes. The significant heterogeneity among subjects makes it impossible to determine whether fetuses born preterm due to a specific subtype of preterm birth are more likely to benefit from magnesium sulfate administration compared with other subtypes.
Contraindications and cautions — Use of magnesium sulfate is contraindicated in patients with:
●Myasthenia gravis, since it can precipitate a severe myasthenic crisis [19,20]
It should be used cautiously in patients with:
●Myocardial compromise or cardiac conduction defects, because of its anti-inotropic effects
●Impaired renal function (defined as a serum creatinine greater than 1.0 mg/dL [88.4 micromol/L]), since magnesium is eliminated by the kidneys. These patients will have an exaggerated rise in serum magnesium concentration and may develop magnesium toxicity at the usual maintenance infusion doses. Maintenance infusion dosing must be adjusted or eliminated in such patients, but a standard loading dose is given since their volume of distribution is not altered. If an adjusted maintenance dose is given, magnesium levels should be monitored and the dose readjusted or eliminated, as appropriate, to avoid toxicity. Extreme caution is advised when treating patients with reduced GFR (<30 mL/min).
Lower and upper gestational age
●Lower limit – For patients at the lower limit of viability and at high risk for birth within 24 hours (see "Periviable birth (limit of viability)"), we confer with the neonatology team and jointly counsel the mother about possible management strategies. If they opt for neonatal interventions at this gestational age, we administer magnesium sulfate for neuroprotection.
None of the randomized trials of magnesium sulfate for neuroprotection included pregnancies at <24 weeks of gestation, although a prospective observational study included pregnancies as early as 22+0 weeks .
●Upper limit — We limit use of magnesium sulfate for neuroprotection to pregnancies <32 weeks of gestation because the majority of data from randomized trials was derived from pregnancies <32 weeks.
The upper limit of gestational age for the neuroprotective effect of magnesium sulfate has not been well-studied. In a meta-analysis that stratified by the gestational age at randomization: <32 to 34 weeks (5235 fetuses) versus <30 weeks (3107 fetuses), benefits were similar for both gestational age ranges . The numbers needed to prevent one case of cerebral palsy in the <32 to 34 weeks group and the <30 weeks group were 56 and 46 pregnancies, respectively.
Dose — Our preferred dose is magnesium sulfate 4 g intravenously over 20 minutes followed by maintenance dose of 1 g/hour.
This appears to be the lowest effective dosing strategy, and we believe it is likely to have a more favorable side effect and safety profile than the alternative higher dose regimen (6 g loading, 2 g/hour maintenance) used in one of the seminal trials (table 1). In addition, it seems biologically plausible that the neuroprotective effects of magnesium sulfate are secondary to residual concentrations of the drug in the neonate's circulation, despite findings from one of the other seminal trials that omitted the maintenance dose.
Data are limited regarding optimal maternal loading and maintenance doses to confer neonatal benefit and avoid potential harm. Because it is theoretically possible that magnesium sulfate may have both neuroprotective and toxic fetal effects depending on dose/exposure , this is a crucial area for future investigation.
Timing — Magnesium sulfate should be initiated within 24 hours before birth, with the following caveats:
●Every effort should be made to reserve therapy for pregnancies at high risk of birth within 24 hours and avoid use in those in which the diagnosis of preterm labor is questionable (threatened preterm labor) or preterm prelabor rupture of membranes has occurred without preterm labor; however, these assessments are subjective.
●When a preterm cesarean birth is planned before labor, we administer the loading dose and then initiate maintenance therapy. In our practice, we aim for 6 to 12 hours of maintenance therapy prior to a scheduled cesarean birth.
●If emergency or expeditious delivery is indicated because of maternal or fetal status, delivery should not be delayed to administer magnesium sulfate.
●If induction of labor is likely to take longer than 24 hours, it is reasonable to delay administration until cervical ripening is achieved and the likelihood of birth is more proximate since we do not administer magnesium sulfate for more than 24 hours or retreat. (See 'Duration' below and 'Retreatment' below.)
The importance of appropriate patient selection and timing of the single course of therapy was suggested by a study that reported exposure to magnesium within the 12 hours before delivery was associated with a reduced odds of cerebral palsy compared with exposure that was terminated >12 hours before delivery .
Duration — Magnesium sulfate is discontinued when the infant is born.
●Minimum duration – Predicting the time of vaginal birth is challenging and available data are insufficient to inform the minimal duration of maintenance therapy needed to achieve a benefit. Therefore, we believe providers should have a low threshold for initiating therapy in patients who are in advanced labor, even though some guidelines suggest not beginning therapy if birth is expected within four hours.
●Maximum duration – We limit the magnesium sulfate infusion to a maximum of 24 hours, even if the birth has not occurred, as this was the maximum duration of therapy in seminal trials (table 1). The upper limit of safe and effective exposure in this setting (ie, neuroprotection) is not well defined.
A secondary analysis of the BEAM trial noted no difference in cerebral palsy or neonatal death rates among newborns with in utero duration of exposure less than 12 hours, 12 to 18 hours, or greater than 18 hours .
Retreatment — The authors do not administer more than one course of magnesium sulfate for neuroprotection as there are limited data regarding any benefit in pregnancies that do not deliver after the initial course.
Of the trials designed to assess the neuroprotective benefits of magnesium sulfate, only the BEAM trial  allowed retreatment. These investigators readministered the full dose of magnesium sulfate if delivery was again considered imminent, the pregnancy was <34 weeks of gestation, and the initial magnesium infusion had been discontinued for more than six hours. Based on this trial, some clinicians, including the section editor of this topic, offer retreatment using the BEAM criteria. Although magnesium sulfate is not costly and there is no known risk of harm from retreatment, a prudent approach to decision making may be to individualize retreatment based on factors such as the gestational age at the time for initial treatment, time since initial administration, and indication(s) for/urgency of delivery.
Side effects — Given the widespread use of magnesium sulfate for prevention of eclampsia, most providers are familiar with magnesium toxicity and the common maternal, fetal, and neonatal side effects. (See "Preeclampsia: Intrapartum and postpartum management and long-term prognosis", section on 'Signs of magnesium toxicity'.)
In a meta-analysis of magnesium sulphate for neuroprotection of the fetus in pregnancies at risk of preterm birth, patients in the magnesium group often experienced side effects that led to cessation of therapy . However, the risk of a severe adverse maternal outcome, including death, cardiac arrest, or respiratory arrest, is very low.
Monitoring — Urine output and deep tendon reflexes should be closely monitored in all patients. The maintenance phase of treatment should be continued only if a patellar reflex is present (loss of reflexes being the first manifestation of symptomatic hypermagnesemia), respirations exceed 12 per minute, and the urine output exceeds 100 mL per four hours. (See "Preeclampsia: Intrapartum and postpartum management and long-term prognosis", section on 'Regimen'.)
Antenatal corticosteroids — A course of antenatal corticosteroids is routinely administered to pregnancies <34 weeks of gestation at risk for delivery within the next seven days because this therapy reduces the rate of respiratory distress syndrome, intraventricular hemorrhage, necrotizing enterocolitis, and neonatal death. (See "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery", section on '22+0 to 22+6 weeks' and "Antenatal corticosteroid therapy for reduction of neonatal respiratory morbidity and mortality from preterm delivery", section on '23+0 to 33+6 weeks'.)
Data regarding the combination of antenatal corticosteroids and magnesium sulfate in extremely preterm gestations are limited but are reassuring that improved survival is not associated with increased neurodevelopmental impairment. In pregnancies at 22+0 to 26+6 weeks of gestation, a prospective study reported that the combination of antenatal corticosteroids and magnesium sulfate was associated with a 30 to 50 percent reduction in the composite outcome of death or severe neurodevelopmental impairment at 18 to 26 months corrected age compared with use of either therapy alone or neither therapy . These findings in children born extremely preterm, who are at high risk for death and neurodevelopmental impairment, underscore the importance of administering both therapies to minimize both mortality and morbidity.
Choice of tocolytic for patients in preterm labor — Magnesium sulfate should not be chosen for tocolysis based solely upon fetal neuroprotective effects. If tocolysis is indicated, the most effective agent with the most favorable side effect profile should be administered, and this would not be magnesium sulfate. These data, and choice and use of tocolytics for inhibition of preterm labor, are reviewed separately. (See "Inhibition of acute preterm labor".)
Maternal side effects are increased when magnesium sulfate is administered concomitantly with beta agonists or calcium channel blockers. We acknowledge that the data on the combined use of calcium channel blockers and magnesium sulfate are sparse; however, case reports suggest an increased risk of symptomatic hypocalcemia, hypotension, and cardiac suppression [9,10,14]. Thus, in patients <32 weeks of gestation who are candidates for tocolysis, we use indomethacin for labor inhibition when magnesium sulfate is being administered for fetal neuroprotection.
SUMMARY AND RECOMMENDATIONS
●Treatment – In pregnancies at high risk of preterm birth within 24 hours, we suggest administration of magnesium sulfate (Grade 2B). In utero exposure to magnesium sulfate before early preterm birth reduces the incidence of cerebral palsy and severe motor dysfunction in offspring (table 1). (See 'Evidence of efficacy from randomized trials and meta-analyses' above.)
•Upper and lower gestational age – We limit magnesium sulfate therapy for neuroprotection to pregnancies that have reached a gestational age at which the mother opts for neonatal interventions but are less than 32 weeks of gestation. (See 'Lower and upper gestational age' above.)
•Dose and duration – We administer a 4 g intravenous loading dose over 20 minutes followed by a 1 g/hour infusion. (See 'Dose' above.)
The infusion is discontinued at birth or by 24 hours after initiation of the infusion if birth has not occurred. (See 'Duration' above.)
●Tocolysis – If tocolysis is indicated because of preterm labor, the most effective agent with the most favorable side effect profile should be chosen. In general, we use indomethacin in pregnancies <32 weeks of gestation. (See 'Choice of tocolytic for patients in preterm labor' above.)
●Antenatal corticosteroids – A course of antenatal corticosteroids should also be administered according to standard guidelines. (See 'Antenatal corticosteroids' above.)