Pregnancies that are complicated by severe preeclampsia at <34 weeks’ gestation are associated with high rates of maternal and perinatal mortality and morbidities worldwide. The rates of these complications are substantially higher in developing countries with limited health resources. In Western countries, the rate of severe preeclampsia at <34 weeks’ gestation is 0.3%. These rates are higher in women with a history of preterm preeclampsia and multifetal gestation and in those with preexisting medical disorders such as hypertension, renal disease, pregestational diabetes mellitus, or connective tissue disease. Historically, women with severe preeclampsia at <34 weeks’ gestation have had delivery initiated on diagnosis or after completion of a course of corticosteroids for fetal lung maturity because the clinical course of such pregnancies is often characterized by progressive deterioration if delivery if not accomplished.

The main objective of expectant management of severe preeclampsia must always be the safety of the mother and fetus. Although delivery is always appropriate for the mother, it might not be best for a premature fetus. Management with delivery after corticosteroid administration leads to high neonatal mortality and morbidity rates, whereas attempts to prolong pregnancy with expectant management may result in fetal death, increased rates of fetal growth restriction, and increased maternal morbidities. Since 1990, 2 randomized trials that included only 133 women and several observational studies that included approximately 1900 women evaluated the benefits and risks of expectant management of severe preeclampsia at <34 weeks’ gestation. The results of these studies were the subject of 3 recent reviews that suggested that expectant treatment in a select group of women with severe preeclampsia from 24-33 weeks of gestation in a suitable hospital is relatively safe and improves neonatal outcome.

The MEXPRE Latin Study is the first multicenter trial that was designed to compare the benefits and risks of delivery after the completion of corticosteroids vs expectant treatment in 267 women with severe preeclampsia from 28-33 weeks’ gestation. The trial was conducted at 8 tertiary hospitals in Latin America: 133 women were assigned randomly to delivery, and 134 women were assigned randomly to expectant treatment. The primary outcome was perinatal death. Secondary outcomes were a composite of adverse maternal and neonatal outcomes. The sample size was calculated with the assumption of a 30% reduction in the rate of the primary outcome from 15% in the delivery group to 10.5% in the expectant treatment group. Women in the delivery group received corticosteroid therapy followed by delivery 24-72 hours later. Women in the expectant treatment group received corticosteroid therapy followed by delivery only for specific maternal/fetal indications or until 34 weeks’ gestation. Three women were excluded after randomization in the expectant treatment group, which left 264 women for analysis. The average pregnancy prolongation in the expectant group was 10.3+ to 8 days, which was significantly higher than the average of 2.2+ to 0.8 days in the delivery group ( P < .0001). Despite this difference in pregnancy prolongation, there were no significant differences between groups in either perinatal deaths (9.4% in delivery group vs 8.7% in expectant group), composite neonatal morbidities (56.4% vs 55.5%, respectively), or composite maternal morbidities (20.3% vs 25.3% respectively). In addition, women who were assigned randomly to the expectant group had higher rates of small-for-gestational-age infants (21.7% vs 9.4%; P = .005) and higher rates of abruptio placentae (7.6% vs 1.5%; P = .01). On the basis of the results of this trial, the authors suggest that severe preeclampsia at 28-33 weeks’ gestation should be managed with delivery after corticosteroid administration.

The results of the trial are not surprising because pregnancy outcome during expectant management depends on gestational age at onset and the experience of the clinical center in treating such patients. This trial was conducted at several hospitals in 6 countries in Latin American with different expertise in treating such patients. There was no standardized protocol for what is considered an “uncontrollable blood pressure” with the use of oral antihypertensive medications. As a result, this indication was the reason for delivery in 40% of women in the expectant treatment group. One would expect this rate to be approximately 6% if maximum doses of combined antihypertensive medications are used. In addition, 42% of the patients were enrolled at 32-33 6/7 weeks’ gestation; an estimated fetal weight of <10th percentile was considered an indication for delivery. This will explain the shorter duration of pregnancy prolongation in this trial (average, 10 days) compared with the average of 15 days reported by Sibai et al.

The results of the trial warrant other comments. First, the trial did not have adequate sample size to answer the question of whether expectant management is associated with lower rate of perinatal mortality (the primary outcome) because the rate of this outcome in the delivery group was substantially lower than assumed in the sample size calculations. This is not surprising considering that 42% were enrolled at 32-33 weeks’ gestation, but the perinatal mortality rate was only 2%. In addition, in a previous retrospective study of expectant management in similar patients that was reported from Panama by Vigil-De Gracia et al, the reported perinatal mortality rate in 89 women with gestational age of 29-33 weeks’ gestation was only 3%. Second, the rate of neonatal respiratory distress syndrome (RDS) in the expectant treatment group was very high (46%) and was similar to that in the delivery group (52%), despite the fact that these infants were exposed to corticosteroids and were delivered at an average gestational age of 32 weeks. One would expect the rate of RDS to decrease in the expectant group, because this group had a longer latency period. Also, the rate of RDS in neonates with similar gestational age in the study by Vigil-De Gracia et al in Panama was only 17%. This finding casts doubt about the definition and/or the diagnostic criteria that were used for RDS among the various included hospitals. Nevertheless, the results of this trial suggest that severe preeclampsia at <34 weeks’ gestation in Latin American countries and other countries with limited health resources should be managed with delivery after corticosteroid therapy. On the other hand, expectant treatment in a select group of women with severe preeclampsia from 24-33 0/7 weeks’ gestation after corticosteroid therapy in developed countries improves perinatal and neonatal outcomes but requires intensive in-hospital maternal and fetal surveillance.

Finally, current guidelines for expectant management of severe preeclampsia at <34 weeks’ gestation are based on the results from 1 randomized trial in the United States with only 95 patients, 1 randomized trial from South Africa with only 58 patients, and several observational studies. Several retrospective and observational studies have reported increased rates of maternal morbidities with such management. Therefore, there is an urgent need for randomized trials with adequate sample size to address the safety of this management in developed countries. Alternatively, there is a need to develop a biomarker to identify those women who will not benefit from such treatment or to develop a targeted novel therapy to treat the underlying pathophysiologic condition in an attempt to prevent adverse maternal outcomes and prolong gestation beyond what is expected with current management.