Pre-eclampsia has a strong association with adverse outcomes: high incidences at term, and low incidences of morbidity; low incidences remote from term, high incidences of morbidity. The ultimate cure for pre-eclampsia is delivery of the placenta, but gestational age at delivery has an effect on perinatal outcome. We review aspects of two contrasting management approaches, and their indication areas are described (i.e. a more interventionist approach of stabilization (with antihypertensive drugs if needed and corticosteroids for acceleration of fetal lung maturation) and delivery, compared with a more temporizing approach, that treats the symptoms until delivery is necessary. In general, the temporizing approach has been shown to be safe, but beyond 37 weeks gestational age, there is evidence for better outcomes after early delivery. Below that gestational age, trials are awaited that help determine the balance in individual cases, depending on severity of maternal disease and fetal condition.
Introduction
Hypertensive disorders of pregnancy (and pre-eclampsia specifically) remain a leading cause of maternal mortality internationally. The ultimate cure for pre-eclampsia is delivery of the placenta, although women may worsen (or new complications may develop) in the immediate postpartum period.
The maternal syndrome of pre-eclampsia is associated with significant maternal morbidity, especially when the onset is below 32 weeks’ gestation. Potentially life-threatening conditions are rare but significant, and include placental abruption, eclampsia or other encephalopathies, stroke, pulmonary oedema, liver rupture, and renal insufficiency.
Fetal and neonatal consequences include stillbirth and complications of iatrogenic very preterm delivery. As such, they are strongly associated with gestational age at onset of pre-eclampsia, as well as severity of the maternal syndrome, which results in iatrogenic premature delivery and fetal growth restriction. Severe neonatal morbidity associated with very preterm delivery (i.e. retinopathy of prematurity, necrotising enterocolitis, respiratory distress syndrome, infectious morbidity and brain injury) may cause neonatal death or long-term morbidity (i.e. blindness, chronic lung disease, cerebral palsy or more subtle neurodevelopmental disabilities). In addition, intrauterine fetal growth restriction (IUGR) is commonly associated with early onset pre-eclampsia (before 34 weeks’). This IUGR is associated with compensatory physiological adaptation of blood flow (brain sparing) and, can ultimately lead to fetal death without delivery.
On the basis of these observations, two contrasting management approaches have been developed for pre-eclampsia of early onset: a more interventionist approach of stabilisation (with antihypertensive treatment if needed and corticosteroids for acceleration of fetal lung maturation) and delivery, compared with a more temporising approach. Although such temporising management has long been called ‘expectant management’, this is a misnomer. ‘Expectant’ suggests a passive attitude from maternity-care providers. In reality, however, temporising management is active, requiring close maternal and fetal surveillance to determine the optimal timing of delivery, in addition to antihypertensive treatment (if needed) and antenatal corticosteroids used in an interventionist approach.
The relative maternal and perinatal benefits and risks of the interventionist compared with the temporising approaches are yet to be established. It must be assumed that prolongation of pregnancy associated with temporising care exposes the woman to further risk of complications, over and above that associated with the postpartum period (during which women often deteriorate). On the other hand, it is probable that temporising management reduces neonatal morbidity and mortality.
This review focuses on patient selection and the elements that together constitute temporising management of pre-eclampsia ( Fig. 1 ) across gestational ages. It is based on available evidence from published meta-analyses (including Cochrane reviews), guidelines, and other relevant studies identified through a structured search of Medline until October 2010. The search was broad and the key words used among others (text word, MESH-term) were ‘pre-eclampsia or haemolysis elevated liver enzymes low platelets (HELLP) syndrome or fetal growth restriction’ and ‘management’; limits for English language, no limits for gestational age.
Acute symptom management
Maternal manifestations of pre-eclampsia show impressive variation between and within patients. Maternal expression ranges from isolated gestational hypertension, to pre-eclampsia, classically defined as gestational hypertension with proteinuria, with or without end-organ involvement such as eclampsia or the HELLP syndrome (haemolysis elevated liver enzymes low platelets). This broad definition of pre-eclampsia (which includes gestational hypertension without proteinuria but with end-organ dysfunction) has been advocated to ensure that women are optimally managed and that clinical mimickers of pre-eclampsia (e.g. vasculitis) are recognised. In a randomised-controlled trial on temporising management in early gestational ages, it was shown that, during further pregnancy, the natural course of the disease is to develop other disorders within the hypertensive disorders spectrum, such as isolated gestational hypertension or normotensive IUGR. In term pre-eclampsia, such within-patient clinical variability seems to be less common, possibly as a result of intervention (with delivery), differences in underlying pathophysiology, or both.
An initial period of maternal and fetal evaluation and treatment is undertaken for all women irrespective of whether ultimate care is driven by an interventionist or temporising approach (discussed below). Centralisation of care is justifiable for patients with either severe maternal disease or low anticipated birth weight, given the high probability of iatrogenic premature delivery and the possibility of maternal complications requiring intensive care facilities. At every evaluation during the monitoring period, fetal benefits and maternal risks of further prolongation of pregnancy have to be weighed against the benefits and risks of immediate delivery ( Fig. 1 ).
Acute symptom management
Maternal manifestations of pre-eclampsia show impressive variation between and within patients. Maternal expression ranges from isolated gestational hypertension, to pre-eclampsia, classically defined as gestational hypertension with proteinuria, with or without end-organ involvement such as eclampsia or the HELLP syndrome (haemolysis elevated liver enzymes low platelets). This broad definition of pre-eclampsia (which includes gestational hypertension without proteinuria but with end-organ dysfunction) has been advocated to ensure that women are optimally managed and that clinical mimickers of pre-eclampsia (e.g. vasculitis) are recognised. In a randomised-controlled trial on temporising management in early gestational ages, it was shown that, during further pregnancy, the natural course of the disease is to develop other disorders within the hypertensive disorders spectrum, such as isolated gestational hypertension or normotensive IUGR. In term pre-eclampsia, such within-patient clinical variability seems to be less common, possibly as a result of intervention (with delivery), differences in underlying pathophysiology, or both.
An initial period of maternal and fetal evaluation and treatment is undertaken for all women irrespective of whether ultimate care is driven by an interventionist or temporising approach (discussed below). Centralisation of care is justifiable for patients with either severe maternal disease or low anticipated birth weight, given the high probability of iatrogenic premature delivery and the possibility of maternal complications requiring intensive care facilities. At every evaluation during the monitoring period, fetal benefits and maternal risks of further prolongation of pregnancy have to be weighed against the benefits and risks of immediate delivery ( Fig. 1 ).
Treatment of severe hypertension
Severe hypertension (i.e. systolic blood pressure ≥ 160–170 mm Hg, diastolic blood pressure ≥ 110 mm Hg, or both) during pregnancy poses a serious threat to women and their babies. A direct statistical relationship exists between height of (systolic) blood pressure risk and risks of maternal complications, particularly stroke. The need for antihypertensive treatment is universally agreed at these blood pressure levels. A wide range of short-acting antihypertensive drugs is available for this purpose. The most common options are labetalol, hydralazine, and calcium channel blockers, particularly nifedipine capsules. No agent has been shown to be clearly better than another. This is discussed in more detail in the chapter by Eastabrook et al. in this issue of Best Practice and Research Clinical Obstetrics and Gynaecology on the origins and end-organ consquence of pre-eclampsia; however, in summary, until better evidence is available, the choice of antihypertensive drugs should depend on the clinician’s experience and familiarity with a particular drug, and the anticipated adverse drug effects.
Treatment of non-severe hypertension
Non-severe hypertension during pregnancy (i.e. systolic blood pressure 140 to 160–170 mmHg, diastolic blood pressure 90–110 mmHg, or both) is common. In contrast to severe hypertension discussed above, the indications for antihypertensive therapy and target ranges for antihypertensive therapy vary widely across centres and between published guidelines. At the heart of this controversy are differences in anticipated effects. Antihypertensive drugs are often used in the belief that lowering blood pressure will prevent progression to more severe disease, and thereby decrease the risk of adverse maternal and perinatal outcomes. Pooled data suggest that early treatment (with any antihypertensive drug) may indeed decrease the incidence of more elevated blood pressure and proteinuria, but there is no evidence that true adverse maternal or perinatal outcomes are reduced in frequency; in fact, there is some evidence that fetal growth may be adversely affected. The ongoing CHIPS trial (Control of Hypertension In Pregnancy Study) will determine the effects on serious maternal and perinatal outcomes of ‘less tight’ versus ‘tight’ control of non-severe hypertension.
Prevention of seizures of eclampsia
Eclampsia, the occurrence of one or more convulsions (seizures), in association with pre-eclampsia, is a rare but serious complication, and can be decreased in frequency by anticonvulsant therapy. In pooled data, it was unequivocally shown that the use of magnesium sulphate for 24–48 h is the single most effective anticonvulsant drug, more than halving the risk of eclampsia compared with placebo or no therapy (relative risk [RR] 0.41, 95% confidence interval [CI] 0.29 to 0.58). Mild side-effects (predominantly flushing) were more common with magnesium sulphate (24% v 5%; RR 5.26, 95% CI 4.59 to 6.03; number needed to treat [NNT] for harm 6, 95% CI 6 to 5). No overall difference in the risk of stillbirth or neonatal death was found (RR 1.04, 95% CI 0.93 to 1.15).
Worthy of discussion are the indications for magnesium sulphate, extracted from the data of the authoritative MAGPIE trial that dominates the data in the Cochrane review. The NNT to prevent one case of eclampsia varied substantially between subgroups within the MAGPIE trial: 109 (95% CI 72 to 225) for women with ‘mild’ pre-eclampsia, 91 (95% CI 63 to 143) for the whole study population (40% of whom had severe pre-eclampsia), 63 (95% CI 38 to 181) for women with severe pre-eclampsia, and 36 (95% CI 21 to 125) for women deemed by their physicians to have ‘imminent’ eclampsia. As such, it has been argued that prophylactic treatment with magnesium sulphate is indicated in severe pre-eclampsia and ‘imminent’ eclampsia, but not necessarily in ‘mild’ pre-eclampsia.
Corticosteroids for haemolysis elevated liver enzymes low platelets syndrome
It has been suggested from observational studies that steroid treatment in HELLP syndrome may improve disordered maternal haematological and biochemical features. The controversy about steroid use for this purpose is based on uncertainty about whether steroids improve clinical maternal and perinatal outcomes. In a 2010 Cochrane review , 11 trials (550 women) were identified that compared corticosteroids with placebo or no treatment. No significant difference was found in the risk of maternal death (RR 0.95, 95% CI 0.28 to 3.21), maternal death or severe maternal morbidity (RR 0.27, 95% CI 0.03 to 2.12), or perinatal or infant death (RR 0.64, 95% CI 0.21 to 1.97). The only clear effect of treatment on individual outcomes was improved platelet count. Thus, no clear evidence was found of any effect of corticosteroids on substantive clinical outcomes, and no support for the routine use of steroids for the management of HELLP.
Corticosteroids for fetal lung maturation
After the publication of the randomised trial of Liggins and Howie in 1972, the introduction and widespread use of antenatal corticosteroids in severely preterm fetuses resulted in a drastic decrease in neonatal mortality and morbidity. International guidelines now recognise the use of single course antenatal corticosteroids as standard care in pregnancies at imminent risk of preterm delivery at less than 34 weeks. Further information is required about the optimal dose-to-delivery interval, ideal corticosteroid, effects in multiple pregnancies, and confirmation of long-term effects into adulthood. In particular, there are concerns about the negative effect on neuronal and cerebral growth given that antenatal corticosteroids may accelerate the maturation of the fetal brain.
Although recommendations for antenatal corticosteroids are based on gestational age criteria alone, the studies underlying this recommendation usually excluded pregnancies complicated by pre-eclampsia or fetal growth restriction. This decision was based, in part, on data showing that growth-restricted fetuses experience an acceleration of lung maturation related to intra-uterine stress, and may not benefit from antenatal corticosteroids, although this has been debated.
Debate is ongoing about the relative benefits and risks of repeat doses of antenatal corticosteroids; conflicting evidence on short-term benefits were seen in two large recent trials; data on outcomes in later childhood are pending. At present, pending more data, it would seem prudent to limit corticosteroid use to a single course, given the uncertainties about short-term benefits and long-term risks.
In summary, it is recommended that single-dose corticosteroids for fetal lung maturation are given to all women at risk of delivery at less than 34 weeks, including those with pre-eclampsia, growth-restricted fetuses, or both. More data are required before repeated courses of corticosteroids can be recommended.
Pain
Maternal pain, such as headache, chest pain, or right-upper quadrant discomfort are considered to be non-specific markers of potential subclinical end-organ dysfunction of pre-eclampsia and, as such, they may be the best predictors for imminent maternal pathology. Their presence should prompt close maternal and fetal surveillance to detect the development of actual end-organ dysfunction. Under close observation, analgesic use is acceptable, and use of prophylactic magnesium sulphate for eclampsia prevention has been discussed above.
Plasma volume expansion
Reports that diminished circulating plasma volume, reduced cardiac output and end-organ vasoconstriction are key features of pre-eclampsia or IUGR received interest as early as the 1970s. It was hypothesised that a lower blood-pressure target combined with plasma-volume expansion might reduce the maternal and perinatal risks associated with pre-eclampsia. Numerous observational cohort studies documented beneficial effects for women with pre-eclampsia and fetuses with IUGR. However, as recently as 2005, no randomised trials had been published on the relevant clinical end points.
The one relevant randomised-controlled trial from The Netherlands enrolled 216 women with severe disease between 24 and 34 weeks (mean 30 weeks), who were deemed eligible for temporising management. The underlying hypertensive disorders included the full spectrum of severe disease: HELLP syndrome, severe pre-eclampsia, eclampsia or severe IUGR with gestational hypertension. Plasma volume expansion was associated with lower haemoglobin levels and higher urine output, but not with a difference in either the primary outcome (a neonatal neurological examination score at term age) or secondary outcomes (of perinatal mortality, neonatal morbidity and maternal morbidity). If anything, the 95% confidence interval was more consistent with higher maternal morbidity (including pulmonary oedema), a shorter prolongation of pregnancy, more oxygen use for neonates, and more neonatal morbidity in the plasma volume expansion group. No improvement was seen in umbilical artery Doppler profile, or fetal growth after plasma volume expansion.
In addition, these data did not support use of plasma volume expansion for treatment of maternal oliguria or prevention of hypotension at the start of intravenous vasodilatory antihypertensive medication. In women who were managed without plasma volume expansion, intravenous antihypertensive medication was given without previous volume loading; hypotensive episodes were rarely observed, and fetal distress necessitating delivery associated with hypotension was not observed.
Despite the strong hypothetical case for plasma volume expansion in severe early onset hypertensive disorders of pregnancy, plasma volume expansion has now been shown not to be beneficial; in fact, it is probably harmful. It is assumed that the volume loading is quickly counteracted by strong homeostatic mechanisms. These regulatory mechanisms are still intact, despite having a different set point. Increases in atrial natriuretic peptide, as reported in other studies, as well as increases in urine output, probably antagonise the haemodilution.
Oliguria
Oliguria is commonly seen in women with severe pre-eclampsia, particularly after caesarean section. Oliguria is usually transient, lasting for hours to days. However, it creates clinical anxiety because it may be a marker of renal insufficiency and progressive renal failure. As such, close follow-up of renal function is required, by measuring serum creatinine looking for relative changes (and not just a value that is ‘abnormal’). Responding to transient oliguria with volume loading is not a rational strategy. Intravenous fluids may increase renal output but they do not change renal function; they are only indicated with persistent oliguria and rising creatinine to ensure that there is not a component of pre-renal failure. The use of diuretics for transient oliguria is also not rational, given the presence of intravascular underfilling rather than overload. Use of diuretics in frank oliguric renal failure makes fluid management easier; it does not change the renal function per se . Diuretics are only indicated in the presence of pulmonary oedema.
Timing of delivery
The ultimate cure for pre-eclampsia is delivery of the placenta, although women may worsen (or new complications may appear) in the immediate postpartum period, particularly the first 5 days.
When considering the relative benefits and risks of the interventionist and temporising approaches to pre-eclampsia management, a number of assumptions are made. It is assumed that antenatal maternal risk will be decreased by an interventionist approach; postpartum maternal risk will be unaltered by an interventionist approach; and that close maternal and fetal surveillance will minimise, but not eliminate, maternal and fetal risks, as these adverse outcomes are often unpredictable and rapidly progressive.
Prediction of the maternal complications of pre-eclampsia is an area of active investigation, among which is the recent effort to create a clinically validated and useful prediction model (PIERS). Sensitive and validated prediction models may aid in the patient selection of women who are most appropriate for temporising care. At present, the observational literature (39 studies, 4650 women) suggests that only 40% of women are deemed eligible for temporising care.
It should be emphasised that most studies reporting on temporising management of pre-eclampsia were conducted at academic centres. Under these situations, all women were managed according to local protocol, involving intensive monitoring of maternal and fetal well-being, and cared for by experienced doctors and nurses in a high-risk pregnancy care setting. As such, temporising management should be restricted to a tertiary-care setting. Many of these women will already require transfer to such a setting because of their gestational age and the need for high level neonatal intensive care should they deliver.
Over 40 studies on temporising management of pre-eclampsia have been published, although most have been observational in nature. We will review the evidence according to gestational age at presentation, as it is the most important predictor of adverse outcomes for the baby. However, the choice between an interventionist stabilise-and-deliver approach and a temporising approach will also be influenced by the availability of medical staff, medicolegal pressure, and individual patient characteristics and desires. The criteria for which women should undergo delivery varies, although variability primarily relates to maternal indications for delivery. Some clinicians deliver at the onset of symptomatic pre-eclampsia, whereas others deliver only for treatment-resistant hypertension, pulmonary oedema, recurrent HELLP syndrome or other serious maternal morbidity, or when further prolongation of pregnancy may be evidently hazardous for the mother. Fetal indications for delivery are consistently non-reassuring fetal heart rate tracings (i.e. decelerations or prolonged poor variability). Consensus is lacking on whether IUGR, abnormal umbilical artery or central venous (ductus venosus) Dopplers should prompt delivery.
Gestational age before 24 weeks
Pre-eclampsia, with onset before 24 weeks’ gestation is a rare but severe condition in pregnancy with few data to inform management. In the South African population, gestational age below 24 weeks uniformly resulted in fetal demise. In a US tertiary care centre, gestational age below 24 weeks was associated with a grim fetal prognosis in those opting for temporising management. Despite obvious advances in neonatal intensive care, these older studies are consistent with more recent data from New Zealand, the USA and The Netherlands. Maternal morbidity associated with such early onset disease is significant, with up to 65% of women experiencing at least one major maternal complication, consistent with other cohort studies.
Given the maternal risks and the absence of obvious perinatal benefits at under 24 weeks’, temporising management cannot be recommended as a routine treatment option. After extensive counselling, termination of pregnancies should be considered as an option.
Gestational age between 24 and 34 weeks
Several large retrospective cohort studies from different settings (The Netherlands, South-Africa, France, Egypt and USA) have reported on the results from temporising management strategies. Combined with data from two randomised-controlled trials, a picture arises of a severe maternal condition with significant, but in general reversible, morbidity, as shown in Table 1 . Interpretation of these data are complicated by the lack of a comparison group (for all but the two trials), between-study variability in diagnostic criteria for severe pre-eclampsia, and differences in inclusion criteria related to underlying pre-existing medical conditions, IUGR, HELLP syndrome, or eclampsia. Variability was also found in entry to the study before or after administration of steroids for fetal lung maturation, and variation in local criteria for fetal viability. For the latter reason, a similar table for neonatal outcomes has not been provided.
| Author | Year published | Number of patients | Maternal disease at inclusion | Gestational age at inclusion (weeks) | Prolongation (days) b | Severe maternal morbidity excl HELLP (%) c | HELLP (%) |
|---|---|---|---|---|---|---|---|
| Odendaal et al. (RCT) | 1990 | 18 | PET | 28–34 | 7.1 | 27 | NR |
| Sibai et al. (RCT) | 1994 | 49 | PET | 28–32 | 15.4 | 4 | NR |
| Fenakel et al. | 1991 | 49 | PET | 26–36 | 12.5 | NR | 0 |
| Moodley et al. | 1993 | 38 | PET | <32 | 12.3 | 16 | NR |
| Olah et al. | 1993 | 28 | PET | 24–32 | 9.5 | 4 | 8 |
| Visser et al. | 1994 | 114 | HELLP/PET | 25–34 | 10 | 13 | 8 |
| Visser et al. | 1995 | 254 | HELLP/PET | 20–32 | 14 | 9 | 9 |
| Van Pampus et al. | 1998 | 82 | HELLP/PET | 25–34 | 6 | 7 | NR |
| Hall et al. | 2000 | 340 | PET | 24–34 | 11 | 22 | 5 |
| Vigil-deGracia et al. | 2003 | 129 | HELLP/PET | 24–34 | 8.5 | 12 | 9 |
| Haddad et al. | 2004 | 239 | PET | 24–34 | 5 | 16 | 14 |
| Oettle et al. | 2005 | 131 | PET | 24–34 | 11.6 | 30 | 5 |
| Ganzevoort et al. | 2005 | 216 | HELLP/PET | 24–34 | 8.7 | 13 | 18 |
| Van Runnard Heimel et al. | 2006 | 31 | HELLP | 24–30 | 7.5 | 10 | 61 |
| Bombrys et al. | 2008 | 32 | PET | 24–27 | 8 | 22 | 16 |
| Bombrys et al. | 2009 | 66 | PET | 27–34 | 5 | 23 | 8 |
| Abdel-Hady et al. | 2010 | 211 | PET | 24–34 | 12 | 13 | 8 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
