Preeclampsia is a multisystem condition characterised by hypertension development (systolic greater than 140mmHg or diastolic greater than 90mmHg) after 20 weeks gestation with greater than 0.3g proteinuria in 24 hours, renal insufficiency, liver disease, neurological problems, haematological disturbances or fetal growth restriction.⁶

Severe preeclampsia refers to profound hypertension (greater than 160mmHg systolic or greater than 110mmHg diastolic) and extreme organ function derangements. These may involve the nervous system (eclampsia, headache, visual disturbances, hyper-reflexia and clonus); gastrointestinal system (epigastric or right upper quadrant pain, nausea and elevated liver transaminases); haematological system (thrombocytopaenia [less than 100 x 106/l], haemolysis and disseminated intravascular coagulation); or cardiorespiratory problems (pulmonary oedema); acute renal failure; and feto-placental compromise.⁴

Preeclampsia affects five to seven per cent of pregnancies, increases maternal and fetal morbidity, and contributes to 18 per cent of all maternal deaths.¹² Women at increased risk are primigravidas and those with pre-existing diabetes or hypertension, hydatidiform mole, or a family history of hypertension.¹⁴ Most morbidity and mortality is attributable to eclamptic complications.⁹ Hence, the benefit of preventing such events is appreciable. Magnesium sulphate is commonly used in preeclampsia to prevent seizures. Dosing regimes established in 1955 remain standard therapy today.¹⁴ However, there is still controversy surrounding which women should be treated and when, and how serum magnesium levels should guide clinical decision-making.

Pathophysiology of preeclampsia

The pathogenesis of preeclampsia is multifactorial and abnormal placentation appears to play a pivotal role. Recent hypotheses propose a two-stage model.²⁰ First, abnormal placentation involving maladapation of uterine spiral arteries and failed intervillous space remodelling leads to intermittent placental hypoxia and reoxygenation, predisposing to intrauterine growth retardation (IuGR).^12,14 A preeclamptic placenta may demonstrate degeneration, hyalinisation, calcification and necrosis.¹⁴

Abnormal placentation is followed by the release of factors into the maternal blood. This produces increased maternal serum soluble fms-like tyrosine kinase 1 (sFlt-1), decreased maternal vascular endothelial growth factor (VEGF), and altered soluble placental growth factor (PIGF).¹² Syncytiotrophoblast microfragments and necrotic trophoblastic material may also be released into the maternal circulation.²⁰

The response is then attenuated by numerous maternal factors including diabetes, diet, immunological factors and genetics. However, the relative importance of these is debatable.²⁰ The end result is maternal vascular changes, including vasoconstriction (producing organ hypoperfusion/ischaemia); breaching of collagen membranes (oedema and proteinuria); and eventual multiple organ dysfunction.¹⁴ Progression to eclampsia is thought to result from this cerebral vasospasm and oedema.

There is further dispute as to when this is initiated. Huppertz¹² suggests that the process begins well before clinical recognition. Early onset preeclampsia represents failed differentiation of all trophoblast cells, extreme fetal hypoxia and IuGR. Later, preeclampsia is associated with failed extravillous trophoblast differentiation leading to the syndrome of preeclampsia, but reduced fetal compromise compared to early onset disease.¹²

Mechanism of action of magnesium sulphate

The exact mechanism of magnesium sulphate in preeclampsia/eclampsia is unknown, however, several theories have been proposed. Firstly, magnesium induces vasodilation by calcium antagonism, decreasing myosin contractility, and promoting tunica media relaxation.¹¹ It may also act indirectly via the gestationally-dependent production of nitrous oxide, a potent vasodilator, and may inhibit endothelial platelet aggregation via prostaglandin I2.¹¹ Reducing cerebral vasospasm may also prevent eclampsia secondary to cerebral hypertension and oedema, and minimise cerebral ischaemia.¹¹

Further, magnesium sulphate may directly reduce cerebral oedema. Calcium antagonism reduces blood-brain-barrier permeability by inhibiting the contraction of cerebral endothelial cells, limiting pinocytosis and inhibiting astrocyte expression of aquaporin-4.¹¹ Magnesium may also antagonise NMDA (N-methyl-D-asparate) receptors, decreasing central glutamic stimulation and preventing seizure activity.¹¹ Furthermore, magnesium sulphate depresses neuromuscular junction transmission, which may reduce external manifestations of seizure activity.¹¹

Evidence for magnesium sulphate in preeclampsia

Several studies have investigated magnesium sulphate for seizure prevention in preeclampsia. Randomised control trials (RCTs) have investigated magnesium sulphate compared to placebo and other anticonvulsants. The largest RCT (n(test)=5055 n(control)=5055), the Magpie Trial³ investigated preeclamptic/eclamptic women treated with a standard loading and maintenance doses of magnesium sulphate compared to placebo. Women randomised to magnesium sulphate were significantly less likely to experience seizures, and women with eclampsia were less likely to experience recurrent seizures (RR=0.42 95% CI0.26-0.60). Smaller studies⁷ were insufficiently powered to demonstrate significant differences in seizure events. A Cochrane systematic review incorporating the Magpie Trial and several smaller studies of women with mild, moderate and severe preeclampsia has found that magnesium sulphate halves the risk of eclampsia.¹⁰

Magnesium sulphate regimes have also been associated with significantly fewer seizures and recurrent seizures compared to diazepam¹⁰, nimodipine⁵ and phenytoin¹⁰. However, these were small trials, with limited seizure event numbers.

Maternal outcomes

Common adverse effects of magnesium sulphate include nausea, vomiting, flushing, hypotension, muscle weakness, paralysis, diplopia, CNS depression and hyporeflexia.¹⁵ Areflexia occurs with total serum magnesium of 8 to 10mmol/l.¹¹ Life-threatening complications, including renal failure and respiratory paralysis, occur with higher serum magnesium levels (greater than 13mmol/l for respiratory paralysis). Coma, arrhythmias and cardiac arrest may ensue with still higher doses.¹¹ Overall incidence of magnesium toxicity in women undergoing therapy for preeclampsia is low.^10,3

Studies analysing mortality in women undergoing magnesium sulphate therapy for preeclampsia have failed to show significant differences in maternal mortality versus controls.^11,3 However,
there is a low overall mortality rate in these studies. The Magpie Trial³ also found no difference between test and control groups for overall maternal morbidity, respiratory depression/arrest, pneumonia, pulmonary oedema, cardiac arrest, renal failure, liver failure, coagulopathy and cerebrovascular events. Risk of placental abruption does not differ significantly with the administration of magnesium sulphate.^16,3,5

To minimise adverse maternal outcomes, magnesium sulphate therapy should be avoided in women with myasthenia gravis, and concurrent calcium channel blockers administration. Dosage should be modified according to renal function¹⁷ and monitoring should occur routinely as discussed below.

Perinatal outcomes

Crowther et al⁸ demonstrated no neonatal adverse effects when magnesium sulphate was given at 30 weeks as neuroprotection for preterm birth. There was also no significant increase in perinatal mortality when the magnesium sulphate was indicated for preeclampsia.^16,7,3 These was no difference in apgar scores lower than seven at five minutes, neonatal respiratory distress, intubation requirements, neonatal hypotonia, or length of special care nursery stay in neonates of women with magnesium sulphate treatment compared to controls.^3,5
Practical considerations in magnesium sulphate therapy Although a standard dosing regime for magnesium sulphate in preventing eclampsia exists, there is debate regarding the appropriate time and clinical situation for the instigating therapy, and how serum magnesium levels should guide decision-making.

Many propose that, given the potential risks of magnesium sulphate, its use is only justified in the presence of severe preeclampsia, not mild preeclampsia, in which baseline seizure risk is low.^21,19 A Cochrane review¹⁰ has further suggested that the number needed to treat (NNT) to prevent one seizure is double that for mild preeclampsia compared to severe preeclampsia (100 compared to 50). Additionally, magnesium does not appear to reduce the rate of progression of mild preeclampsia to severe preeclampsia¹³, or gestational hypertension without preeclamptic features to preeclampsia².

Others¹⁷ suggest that a NNT of 100, given the low cost and side effects of magnesium sulphate with appropriate monitoring, justifies its use in all preeclamptics. The Magpie Trial³ additionally demonstrated that magnesium sulphate consistently reduced relative risk of eclampsia regardless of preeclampsia severity upon treatment initiation. However, this trial relied on subjective clinical judgement to determine eligibility and severity. Additionally, 17 women had experienced seizures prior to recruitment. The use of more objective inclusion criteria would have enhanced the external validity of this study.

Timing of therapy initiation, in relation to gestational age and labour onset, was also not controlled in the Magpie Trial.³ Some women were treated antenatally, others 24 hours prior to labour or within 48 hours postpartum, depending on when symptoms were detected. However, a subanalysis of these groups suggested that effects of magnesium sulphate were independent of when therapy was initiated. This reinforces the importance of clinical monitoring in preeclampsia so that therapy can be instigated as new symptoms arise. Similarly, timing of therapy cessation is controversial. Studies have generally found that continuation for 24-48 hours postpartum, followed by appropriate antihypertensive use, is acceptable.¹⁷

Total serum magnesium is 0.65 to 1.11mmol/l in normal pregnancy, one-third to half of which is protein-bound.²² Therapeutic range advocated for magnesium sulphate treatment is 2.0 to 3.5mmol/l.¹¹ However, this estimate was based on a small, retrospective dataset17 and there are no large-scale trials to support using this range. The Magpie Trial did not measure serum magnesium levels to facilitate blinding.³ Furthermore, since magnesium sulphate exerts its effects through ionised magnesium, ionised magnesium levels could potentially provide more therapeutic relevance. Indeed, the two indices are poorly correlated.¹ Further research is needed to determine an appropriate monitoring protocol and improve magnesium sulphate safety and efficacy.

Standard regimes (4 to 6mg IV loading dose and 1 to 2mg hourly maintenance dose) do not always raise serum magnesium levels to therapeutic levels. One study concluded that 36.2 per cent of participants undergoing a standard regime achieved ‘subtherapeutic’ total serum magnesium levels.¹ Despite this, none of these patients developed seizures. However, this study may have been insufficiently powered to detect seizure events. Further, Aali et al¹ demonstrated that weight was inversely correlated with serum magnesium levels and suggest that per kilo dosing may be appropriate.

However, since side effects are roughly correlated with total serum magnesium levels, serum magnesium monitoring is important. This should be conducted with routine examination of tendon reflexes, respiratory rate and urine output to ensure that magnesium toxicity is detected and treated early.¹⁷

Conclusion

There is a long history of magnesium sulphate use in the management of severe preeclampsia and seizure prevention.⁹ There is good evidence to support magnesium sulphate therapy over placebo and other anticonvulsants. Magnesium sulphate does not appear to significantly increase maternal or perinatal morbidity or mortality. However, the exact mechanism of action is still unknown. Furthermore, although there is a recommended therapeutic serum magnesium range for preeclampsia therapy, it is well documented that standard regimes often fail to produce these levels. The majority of evidence is based on dosing regimes, rather than attaining these serum magnesium levels. Hence, following standard regimes, rather than titrating doses against serum magnesium levels, provides the best evidence-based clinical practice. Serum magnesium monitoring should still occur for early identification of magnesium toxicity.

Further trials investigating magnesium sulphate efficacy and the effects on both total and ionised magnesium levels are required to increase understanding and perfect clinical guidelines. However, given the current evidence supporting magnesium sulphate in preeclampsia, it may become increasingly difficult, and indeed unethical, to randomise preeclamptic women to placebo. Otherwise, comparing the efficacy and safety of standard regimes to doses titrated against serum magnesium levels would provide clinically useful information. However, these would have to be very large studies, given the relatively few seizure events recorded in women undergoing magnesium sulphate therapy.

Acknowledgement

The author would like to acknowledge and thank Dr Amelia Hare for her support and contribution to this article.