|
|
Objective Magnesium sulfate may prevent eclampsia by reducing cerebral vasoconstriction and ischemia. Nimodipine is a calcium-channel blocker with specific cerebral vasodilator activity. Our objective was to determine whether nimodipine is more effective than magnesium sulfate for seizure prophylaxis in women with severe preeclampsia.
Methods We conducted an unblinded, multicenter trial in which 1650 women with severe preeclampsia were randomly assigned to receive either nimodipine (60 mg orally every 4 hours) or intravenous magnesium sulfate (given according to the institutional protocol) from enrollment until 24 hours post partum. High blood pressure was controlled with intravenous hydralazine as needed. The primary outcome measure was the development of eclampsia, as defined by a witnessed tonic–clonic seizure.
Results Demographic and clinical characteristics were similar in the two groups. The women who received nimodipine were more likely to have a seizure than those who received magnesium sulfate (21 of 819 [2.6 percent] vs. 7 of 831 [0.8 percent], P=0.01). The adjusted risk ratio for eclampsia associated with nimodipine, as compared with magnesium sulfate, was 3.2 (95 percent confidence interval, 1.1 to 9.1). The antepartum seizure rates did not differ significantly between groups, but the nimodipine group had a higher rate of postpartum seizures (9 of 819 [1.1 percent] vs. 0 of 831, P=0.01). There were no significant differences in neonatal outcome between the two groups. More women in the magnesium sulfate group than in the nimodipine group needed hydralazine to control blood pressure (54.3 percent vs. 45.7 percent, P<0.001).
Conclusions Magnesium sulfate is more effective than nimodipine for prophylaxis against seizures in women with severe preeclampsia.
Since magnesium sulfate has a cerebral vasodilator effect,13 one explanation for its antieclamptic action is that it decreases ischemia by reducing cerebral vasospasm. We hypothesized that if eclampsia is caused by cerebral ischemia, nimodipine, a calcium-channel blocker that is a specific cerebral vasodilator,14,15 would be an ideal alternative drug. Potential advantages of nimodipine also include its oral route of administration, minimal toxicity, and antihypertensive effect. Therefore, we compared the efficacy of nimodipine with that of magnesium sulfate for the prevention of seizures in patients with severe preeclampsia.
Methods
We conducted an unblinded, randomized, controlled trial between 1995 and 2000. We enrolled women with severe preeclampsia at 14 sites in eight countries (see the Appendix). The protocol was approved by the local review board at each site. All patients provided written informed consent. Patients with severe preeclampsia in whom the decision had been made to effect delivery were eligible provided they had not already received magnesium sulfate and did not have a preexisting seizure disorder. The diagnostic criteria16 are described in detail in the Supplementary Appendix, available with the full text of this article at http://www.nejm.org. Severe preeclampsia was defined as elevated blood pressure (at least 140/90 mm Hg) with proteinuria (a dipstick reading of 1+ or more) in association with one or more of the following: headache, clonus, visual disturbances, epigastric pain or pain in the right upper quadrant, oliguria (urinary output of less than 500 ml per 24 hours), pulmonary edema, elevated liver aminotransferase levels (an alanine aminotransferase or aspartate aminotransferase level of more than 40 U per liter), elevated creatinine level (at least 1.5 mg per deciliter [133 µmol per liter]), hemolysis, thrombocytopenia, intrauterine growth restriction, or oligohydramnios. A blood pressure of 160/110 mm Hg or higher with proteinuria in the absence of any of the other features was also classified as severe preeclampsia. Patients were randomly assigned according to center (Epistat Services) in blocks of six, with the use of sealed opaque envelopes, to receive either nimodipine (60 mg orally every four hours) or magnesium sulfate according to the institutional protocol (a 6-g loading dose followed by an intravenous infusion of 2 g per hour, or a 4-g loading dose followed by an infusion of 1 g per hour). Treatment was continued for up to 24 hours ante partum (after which the protocol called for cesarean section unless delivery was imminent) and for 24 hours post partum. Measurement of blood levels of magnesium sulfate was not required. Blood pressure was measured with the use of automated monitors or manual sphygmomanometry (Korotkoff phase V). Urinary protein was measured with use of a semiquantitative method (dipstick) or, if available, a 24-hour urine collection.
The study was not blinded, because of logistic and economic constraints. The primary outcome measure (eclampsia) was binary, objective, and not subject to observer or measurement bias. Patients were treated according to the usual routine of each hospital, with no interference by the research group. Blood pressure was controlled with intravenous hydralazine for women whose systolic blood pressure was at least 160 mm Hg, whose diastolic blood pressure was at least 110 mm Hg, or whose overall blood pressure was at least 160/110 mm Hg for more than 20 minutes after the initiation of treatment.
The primary outcome measure was eclampsia (defined as a witnessed tonic–clonic convulsion) at any time from the initiation of study-drug administration until 24 hours post partum. Secondary outcome measures included blood-pressure control, complications and adverse drug effects, complications of labor and delivery, and indicators of fetal and neonatal condition. Postpartum hemorrhage (as recorded in the chart) was defined as estimated blood loss in the 24 hours after delivery of more than 500 ml in the case of vaginal delivery and of more than 1000 ml in the case of cesarean delivery.
Women in the nimodipine group who had a seizure received magnesium sulfate. Women who had a seizure while receiving magnesium sulfate continued to receive this drug according to the institutional regimen.
We initially planned to study 1000 patients per group, with interim analyses conducted every six months by an independent trial monitor and adjusted for multiple analyses with use of the Lan–DeMets method. The study was terminated early (August 2000) when a planned interim analysis showed a significantly higher rate of seizure in the nimodipine group.
All statistical analyses were performed with the use of SAS software (version 8, SAS Institute). The Wilk–Shapiro test was used to check the normality of distribution. The two groups were compared with use of analysis of variance, the two-sample t-test, Wilcoxon's rank–sum test, Fisher's exact test, the chi-square test, or logistic regression, as appropriate. Univariate analyses were conducted to identify demographic and clinical variables associated with the occurrence of seizures. Statistically significant variables were identified with the use of Fisher's exact test and univariate logistic regression. Variables that were significantly associated with seizure in the univariate analyses were included as covariates in a multiple logistic-regression analysis, along with other selected variables that were considered to warrant evaluation.
The type of anesthesia was classified as none or local infiltration only, regional (epidural or spinal), and general. The type of anesthesia was excluded from the overall analysis of seizures and was analyzed only in the group of patients who had a postpartum seizure (these patients were all in the nimodipine group). The antepartum effect of anesthesia could not be analyzed, because of uncertainty about the timing of the administration of anesthesia (before or after the seizure).
The odds ratios calculated from logistic regression were considered to provide a reasonable estimate of the risk ratio.17 In comparisons that contained a zero frequency cell, the logit estimator used for the risk ratio incorporated a correction of 0.5 in every cell and a confidence interval could not be obtained. All significance tests were two-tailed, with an 
level of 0.05.
Bayer provided the nimodipine. This sponsor was not involved with the design of the study, the interpretation of the data, or the writing of this report. The investigators had full access to the data.
Results
We enrolled 1750 patients from July 1995 until November 2000. The Appendix shows the number of patients enrolled and the number who received study drug at each site. Data were available for 1650 of 1750 patients (94.3 percent): 819 of the 1650 were given nimodipine, and 831 were given magnesium sulfate. A total of 99 patients were enrolled, were assigned a study number, and underwent randomization but did not receive either drug. In most cases this was because they gave birth before the drug could be administered or because of logistic issues that prevented timely administration of the study drug. Hence, data sheets were not completed on these patients. One patient in the magnesium sulfate group was withdrawn from the study because the induction of labor was stopped and conservative management instituted. The median antepartum duration of treatment was 8.8 hours in the magnesium sulfate group and 8.1 hours in the nimodipine group.
Base-line demographic and clinical characteristics were generally similar in the two groups (Table 1); the only exception was that base-line systolic blood pressure was slightly higher in the magnesium sulfate group (P=0.04). No significant differences between groups were found in any of the recorded laboratory variables: hematocrit, platelet count, alanine aminotransferase level, aspartate aminotransferase level, uric acid level, creatinine level, blood urea nitrogen level, sodium level, or potassium level (data not shown).
|
|
|
|
There was no significant difference between the two groups in the mean arterial pressure recorded before an antepartum seizure (mean [±SD], 124±17 mm Hg among 12 patients in the nimodipine group and 115±11 mm Hg among 7 patients in the magnesium sulfate group; P=0.22). In addition, the 19 women who had an antepartum seizure did not have a significantly higher mean arterial pressure at admission than those who did not have an antepartum seizure (131±20 mm Hg vs. 125±16 mm Hg, P=0.35). Base-line blood pressures and overall blood-pressure readings were similar in the two groups (data not shown), although the magnesium sulfate group more often required hydralazine. Among the patients whose initial mean arterial pressure exceeded 126 mm Hg, fewer in the nimodipine group than in the magnesium sulfate group required hydralazine for blood-pressure control (57.8 percent vs. 66.7 percent, P=0.04). Patients in the nimodipine group who required hydralazine were more likely to have eclampsia than patients in the magnesium sulfate group who required hydralazine (4.0 percent vs. 1.1 percent, P=0.01). Among patients who did not receive hydralazine, the rate of eclampsia was 1.4 percent in the nimodipine group and 0.5 percent in the magnesium sulfate group (P=0.30). Patients who required hydralazine had significantly higher blood-pressure values on admission than those who did not require hydralazine (data not shown). The increased risk of seizure associated with nimodipine persisted after adjustment for hydralazine use and systolic blood pressure at admission (Table 4).
The mean arterial pressure in the nimodipine group was reduced by a mean of 8.2 percent in the first hour after drug administration, and this reduction was maintained at three hours (8.3 percent). There was a 4.2 percent decrease in mean arterial pressure during the first hour in the magnesium sulfate group, and a 7.2 percent decrease within three hours.
Discussion
This study showed that parenterally administered magnesium sulfate is significantly better than oral nimodipine at preventing eclamptic seizures in women with severe preeclampsia. This difference was especially apparent in the postpartum period.
Although we compared magnesium sulfate with nimodipine, the recent MAGPIE trial12 compared magnesium sulfate with placebo in women with mild or severe preeclampsia and also demonstrated a beneficial effect of magnesium sulfate. In that study, there was also a lower maternal mortality rate in the magnesium sulfate group, although the difference was not significant. On the basis of these findings, there is now a considerable effort to promote the prophylactic use of magnesium sulfate in countries where its use was previously not the norm. Despite this, neither the cause of eclampsia nor the mechanism of action of magnesium sulfate as a seizure prophylactic is known.
It has been hypothesized that cerebral vasospasm and resultant ischemia are the predominant cause of eclampsia. If this were true then a specific cerebral vasodilator would be more effective at relieving vasospasm (and thus a better drug to prevent eclampsia) than magnesium sulfate. Our findings do not support this hypothesis.
A recent report18 on the cerebral hemodynamic changes in patients with preeclampsia may explain our findings, since it suggests that elevated cerebral perfusion pressure, rather than decreased cerebral blood flow, is the primary cause of injury. Increased cerebral perfusion pressure is believed to result in "cerebral barotrauma" and vasogenic (and, rarely, cytotoxic) edema. Nimodipine has been shown to increase cerebral perfusion pressure in patients with preeclampsia, whereas magnesium sulfate decreases it.19 Our finding that nimodipine is less effective than magnesium sulfate in preventing seizures suggests that seizures in patients with eclampsia that are not the result of obvious hemorrhage are more likely to be due to overperfusion (hypertensive encephalopathy)18,20,21,22,23,24 than ischemia. The higher rate of seizure in the nimodipine group may be explained on the basis of the nimodipine-induced reduction in protective vasoconstriction, which initiated or worsened overperfusion. This effect may be amplified in the postpartum period when the levels of constrictor substances released by the placenta25 are reduced. We observed a higher rate of seizure post partum, but not ante partum, in the nimodipine group than in the magnesium sulfate group.
The increased risk of seizure associated with elevated cerebral perfusion pressure may be compounded by the use of anesthetic agents or potent vasodilator drugs that interfere with cerebral autoregulation26 or by clinically significant volume expansion (such as through iatrogenic infusion of fluid or autologous uterine transfusion at the time of uterine contraction post partum). We found that women who required hydralazine for blood-pressure control were more likely to have a convulsion when they were given nimodipine than when they were given magnesium sulfate. Although the number of patients studied was small and the implications of this finding are not clear, in these patients the combination of a potent specific cerebral vasodilator (nimodipine) and a drug that prevents functional cerebral autoregulation (hydralazine)26 may have caused hypertensive encephalopathy.
Magnesium sulfate has been traditionally thought to have a transient hypotensive effect.27 In a third of women whose mean arterial pressure exceeded 126 mm Hg at admission, no drug other than magnesium sulfate was required for blood-pressure control. Although this difference may simply represent the effect of bed rest, a change in position, or the use of anesthesia, this observation warrants further study. The fact that no patient in this study died, had an intracranial hemorrhage, or had permanent renal or liver failure suggests that the degree of blood-pressure reduction associated with the administration of magnesium sulfate can be regarded as safe.
Although our analysis suggested that magnesium sulfate was associated with an increased risk of postpartum hemorrhage and respiratory difficulty, we do not believe that these are major risks of this therapy. These findings came to light as the result of post hoc analyses and were not anticipated a priori. The average estimated blood loss was similar in the two groups at the time of delivery, and we did not collect data on the need for blood transfusion. Respiratory difficulty was not objectively measured, and we relied on the subjective assessment of the delivery team. Neither of these two complications was reported as being associated with magnesium sulfate in the MAGPIE trial.12
We also found that chronic hypertension, a systolic blood pressure of more than 180 mm Hg at admission, the need for additional hydralazine, low blood sodium levels, and a young age were independent risk factors for eclampsia in women with severe preeclampsia. Although underlying hypertension, extremely elevated blood pressure, and young age have been previously reported as risk factors for eclampsia,28 the association of eclampsia with low sodium levels is less well known. The findings that regional and general anesthesia were associated with a higher risk of seizure than either no anesthesia or local anesthesia among patients taking nimodipine raise the possibility that these medications have an additive cerebral excitatory interrelation that magnesium sulfate does not. Given that these secondary findings were not the result of a priori hypotheses and may be the result of confounding factors, they should be regarded as hypothesis-generating observations.
Our study demonstrates that magnesium sulfate is more effective than nimodipine in preventing eclampsia in women with severe preeclampsia and reinforces the use of magnesium sulfate as the standard of care against which potential alternatives should be measured. Finally, the lack of effectiveness of nimodipine, a cerebral vasodilator, supports the hypothesis that eclampsia may be caused by cerebral overperfusion rather than decreased cerebral blood flow.
Supported by a grant from the Orphan Products Division of the Food and Drug Administration and by a grant from Bayer.
Source Information
From the Departments of Obstetrics and Gynecology, University of Utah Health Sciences Center, Salt Lake City (M.A.B.); Intermountain Health Care Utah Valley Regional Medical Center, Provo, Utah (M.A.B.); the University of Cape Town, Cape Town, South Africa (J.A.); the University of Texas Medical Branch, Galveston (G.R.S.); and the University of Texas School of Public Health, Houston (J.C.A.).
Address reprint requests to Dr. Belfort at 1034 N. 500 West, Provo, UT 84604, or at uvmbelfo{at}ihc.com.
References
The members of the Nimodipine Study Group were as follows (the numbers in parentheses are the numbers of patients who received at least one dose of study drug and the total numbers enrolled): John Anthony, University of Cape Town, Cape Town, South Africa (806/882); Maria Nobrega de Oliveira, Maternidade Escola J. Cicco, Natal, Brazil (282/303); Michael A. Belfort, Baylor College of Medicine, Houston (158/159); Omur Taskin, Inonu University School of Medicine, Malatya, Turkey (79/79); Jack Moodley, Nelson Mandela Medical School, Durban, South Africa (56/56); Jorge Carillo, Hospital Padre Hurtado, Santiago, Chile (51/52); Mario Festin, University of the Philippines, Manila (53/53); Liliana Voto, Fundacion Miguel Marguiles, Buenos Aires, Argentina (39/39); David Adair, Louisiana State University Medical Center, Shreveport (34/34); Baha Sibai, University of Tennessee, Memphis (33/34); Bjorn Uys, Frere Hospital, East London, South Africa (22/22); Jose Garrido Calderon, Esquina J.J. Perez, Santo Domingo, Dominican Republic (15/15); James W. Van Hook, University of Texas Medical Branch, Galveston (12/12); and Claudia Maria Vilas Freire, Maternidade Odete Valadares, Bello Horizonte, Brazil (10/10).
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Prevention of Eclampsia
Weinstein L., Bonicalzi V., Canavero S., Belfort M. A., Saade G. R.
Extract |
Full Text |
PDF
N Engl J Med 2003;
348:2154-2155, May 22, 2003.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
