|
A correction has been published: N Engl J Med 1996;335(25):1922.
|
| |||||||||||||||||||||||||||||||||||||||||||||||
Background Artemisinin (qinghaosu) and its derivatives are rapidly effective antimalarial drugs derived from a Chinese plant. Preliminary studies suggest that these drugs may be more effective than quinine in the treatment of severe malaria. We studied artemether in Vietnam, where Plasmodium falciparum has reduced sensitivity to quinine.
Methods We conducted a randomized, double-blind trial in 560 adults with severe falciparum malaria. Two hundred seventy-six received intramuscular quinine dihydrochloride (20 mg per kilogram of body weight followed by 10 mg per kilogram every eight hours), and 284 received intramuscular artemether (4 mg per kilogram followed by 2 mg per kilogram every eight hours). Both drugs were given for a minimum of 72 hours.
Results There were 36 deaths in the artemether group (13 percent) and 47 in the quinine group (17 percent; P = 0.16; relative risk of death in the patients given artemether, 0.74; 95 percent confidence interval, 0.5 to 1.11). The parasites were cleared more quickly from the blood in the artemether group (mean, 72 vs. 90 hours; P<0.001); however, in this group fever resolved more slowly (127 vs. 90 hours, P<0.001), the time to recovery from coma was longer (66 vs. 48 hours, P = 0.003), and the hospitalization was longer (288 vs. 240 hours, P = 0.005). Quinine treatment was associated with a higher risk of hypoglycemia (relative risk, 2.7; 95 percent confidence interval, 1.7 to 4.4; P<0.001), but there were no other serious side effects in either group.
Conclusions Artemether is a satisfactory alternative to quinine for the treatment of severe malaria in adults.
Methods
This study was conducted in a special research ward at the Center for Tropical Diseases in Ho Chi Minh City, Vietnam, and was approved by the center's ethics and scientific-review committee. A full-time study team of specially trained physicians and nurses provided medical care. All laboratory measurements made during the acute phase of illness were conducted on site. A three-month pilot study was conducted before the main study to familiarize staff members with the clinical and laboratory procedures.
Objectives
The study was designed to detect a difference in mortality of 33 percent with 95 percent confidence and 80 percent power. The mortality rate for severe malaria was approximately 30 percent before the study. Except where indicated the secondary end points were defined prospectively.
Drugs
Patients were randomly assigned to receive artemether (50 mg per milliliter) or quinine dihydrochloride (250 mg per milliliter) for a minimum of 72 hours. The drugs were issued in packs of 10 identical 3-ml ampules by the Kunming Pharmaceutical Company (Kunming, People's Republic of China). Randomly selected ampules were checked for sterility and potency. The initial dose of both drugs was 0.08 ml per kilogram of body weight (4 mg of artemether per kilogram and 20 mg of quinine salt per kilogram), half of which was injected into the anterior thigh of each leg. The maintenance dose was 0.04 ml per kilogram for each drug (2 mg of artemether per kilogram and 10 mg of quinine salt per kilogram) and was given every eight hours.
When the patient could take oral medication reliably, a second independent randomization was performed. Patients received either a single oral dose of 15 mg of mefloquine per kilogram (Lariam, Hoffmann–LaRoche, Basel, Switzerland) or oral quinine sulfate (Government Pharmaceutical Organization of Thailand, Bangkok) at a dose of 10 mg per kilogram three times daily for up to four days (i.e., for a total of seven days of antimalarial treatment). Treatment with quinine was followed by two tablets of pyrimethamine–sulfadoxine (Fansidar, Hoffmann–LaRoche).
Randomization and Blinding
The drugs for each patient were placed in a coded sealed envelope, and the envelopes were randomized in blocks of 20. Once a patient was enrolled in the study the envelope was opened. Subsequent analysis of efficacy was on an intention-to-treat basis. Although the ampules, drug volumes, and administration schedules of artemether and quinine were identical, the viscosity and color of the two drugs were slightly different. To maintain blinding, a separate team of nurses, who were not otherwise involved with the care of the study patients, drew up and gave the injections. The drugs were kept in an opaque packet in a locked cabinet during the study.
Clinical Procedures
Patients were included in the study if they (or an accompanying relative) gave informed consent, had asexual forms of P. falciparum on a peripheral-blood smear, were older than 14 years, were not in the first trimester of pregnancy, were not intravenous drug users, had received less than 3 g of quinine or two doses of artemisinin or a derivative in the previous 48 hours, and had one or more of the following: a score on the Glasgow Coma Scale of less than 11 (indicating cerebral malaria); anemia (hematocrit, <20 percent), with a parasite count exceeding 100,000 per cubic millimeter on a peripheral-blood smear; jaundice (serum bilirubin, >2.5 mg per deciliter [50 µmol per liter]), with a parasite count of more than 100,000 per cubic millimeter on a peripheral-blood smear; renal impairment (urine output, <400 ml per 24 hours; and serum creatinine, >3 mg per deciliter [250 µmol per liter]); hypoglycemia (blood glucose, <40 mg per deciliter [2.2 mmol per liter]); hyperparasitemia (>10 percent parasitemia); and systolic blood pressure below 80 mm Hg with cool extremities (indicating shock).
Each patient underwent a full clinical examination that included a detailed neurologic assessment. A complete blood count, count and estimation of the life-cycle stage of the parasites, biochemical analyses, measurements of plasma glucose and lactate, and blood cultures were done, and plasma was stored for measurement of cytokines, plasma quinine, and coagulation indexes. Arterial-blood gases and pH were measured on admission beginning in April 1992 (when 234 patients had been enrolled). Blood was obtained by a finger-prick for hematocrit measurements and blood smears every 4 hours for the first 24 hours and every 6 hours until three consecutive smears were negative for asexual stages of P. falciparum. The degree of parasitemia was determined on the basis of the number of parasitized red cells per 1000 red cells (thin film) or the number of parasites per 400 leukocytes (thick film).
Management and Clinical Monitoring
Patients were cared for according to standard recommendations.1 All patients were given isotonic saline to restore fluid balance, and fluid balance was maintained with saline or 5 percent dextrose in water. When necessary, a central venous catheter was inserted and the central venous pressure maintained at 5 cm of water. Blood was transfused if the hematocrit fell below 20 percent. Hypoglycemia was corrected with a bolus injection of 50 ml of 30 percent dextrose in water and a subsequent maintenance infusion of 5 to 10 percent dextrose in water. Detailed clinical and nursing observations were recorded a minimum of every 4 hours for the first 24 hours and every 6 hours thereafter.
In all patients blood glucose, lactate, and cytokine levels were measured 4, 8, 12, and 24 hours after admission. Beginning in June 1992 (after the enrollment of 259 patients) an electrocardiogram with a rhythm strip (paper speed, 50 mm per second) was obtained before treatment, 12 hours after treatment was begun, 4 hours after the last parenteral dose of antimalarial agent, and at discharge. Standard intervals were recorded, and the QT interval was corrected with the use of Bazett's formula (QT/
RR). A diagnostic lumbar puncture was performed if the score on the Glasgow Coma Scale was below 14. Opening cerebrospinal fluid pressures were measured, the cerebrospinal fluid was analyzed microscopically, and levels of protein, glucose, and lactate were determined. Peritoneal dialysis was started in patients with established renal failure. Beginning in September 1993 (after the enrollment of 427 patients) hemofiltration was also available. There were no facilities for ventilation. Acetaminophen was given for a temperature above 39°C, and intravenous diazepam, intramuscular phenobarbital, and if necessary, intravenous phenytoin were given for convulsions. Antibiotics with no clinical antimalarial activity (i.e., not tetracyclines, macrolides, trimethoprim–sulfamethoxazole, or chloramphenicol) were prescribed for suspected cases of bacterial sepsis.
On discharge a detailed neurologic examination, electrocardiography, and beginning in November 1993 (after the enrollment of 441 patients), audiography (frequency range, 250 to 8000 Hz) were performed. A full autopsy was performed on patients who died of malaria if permission could be obtained from relatives. All information was recorded in the patients' records and then triple-entered in a computer data base.
Statistical Analysis
The study was reviewed continuously by an outside monitor. Categorical data were analyzed by Fisher's exact test, and continuous data by the Kruskal–Wallis test with the statistical programs Statview v.4.1 (Abacus Concepts, Berkeley, Calif.) and Stata v.4 (Stata, College Station, Tex.). The lengths of time to recovery were compared in the two groups by survival analysis with the Peto and Peto modification of the Wilcoxon test.13 Cox regression analysis was used to determine the contribution of different variables to recovery.
Results
Between May 1991 and January 1996, 561 patients were enrolled in the study. Only one patient was excluded from the analysis, because a review of the admission blood smear did not confirm the presence of malarial parasites. He died of an intracranial hemorrhage. There was a steady decline in the number of patients recruited during the study that paralleled the overall decline in the incidence of malaria in Vietnam. After reviewing the results of an interim analysis conducted in October 1994 (after 500 patients had been recruited), we decided to continue the study for one more year.
The characteristics of the patients on admission are shown in Table 1. There were no significant differences in any of the major variables between the two treatment groups. The median total dose of antimalarial agent given parenterally, including that given before the study, was 120 mg of quinine per kilogram (range, 20 to 246) and 20 mg of artemether per kilogram (range, 4 to 44).
|
The overall mortality rate was 15 percent (83 of 560 patients). The difference in the mortality rate between the artemether group and the quinine group was not significant (13 percent vs. 17 percent, P = 0.16) (Table 2). Four patients had neurologic sequelae (three in the artemether group and one in the quinine group), and thus there was also no significant difference in the combined number of deaths and cases of neurologic sequelae between the groups. In a multiple logistic-regression model that incorporated factors identified at the outset as associated with outcome, treatment with artemether was associated with a lower mortality (P=0.028) (Table 3). Among patients with cerebral malaria (score on the Glasgow Coma Scale, <11), the overall mortality rate was 16 percent (45 of 290 patients): 15 percent in the artemether group (21 of 142) and 16 percent in the quinine group (24 of 148, P = 0.75).
|
|
The cause of death was often multifactorial. Of the 83 patients who died, 59 had acute renal failure (dialysis had been started in 30 of these patients), 56 had intractable shock, 35 had a terminal respiratory arrest with continued pulse, 25 had serious gastrointestinal bleeding, and 12 had pulmonary edema. A full autopsy was performed on 50 of the patients who died. Serial measurements of arterial-blood gas were begun in 1992 and showed that 42 of the 50 (84 percent) had metabolic acidosis.
Recovery
Artemether treatment was associated with quicker clearance of parasites from the peripheral blood but slower resolution of fever, slower recovery from coma, and longer hospitalizations (P<0.005 for all comparisons) (Figure 1A, Figure 1B, Figure 1C, Figure 1D and Table 4). We used a Cox proportional-hazards model to assess the contribution of different variables evaluated at admission (score on the Glasgow Coma Scale, creatinine and lactate values, log parasite count, and the presence of jaundice, hypoglycemia, and shock) to the time to recovery from coma (defined as the time to reach a score of 15 on the Glasgow Coma Scale). In patients with cerebral malaria, quinine treatment remained significantly associated with a more rapid recovery than artemether treatment (hazard ratio, 1.39; 95 percent confidence interval, 1.06 to 1.81; P = 0.017), and this difference became evident after 48 hours (Figure 1A, Figure 1B, Figure 1C, and Figure 1D). The markers of disease severity at admission in patients with impaired consciousness that lasted more than 48 hours were similar in the two groups. In a similar model, the differences in the time to resolution of fever between the two groups could not be explained by the occurrence of supervening bacterial infections.
|
|
There was no significant difference between the artemether group and the quinine group in the fall in the hematocrit from base-line values (median reduction, 27 percent vs. 30 percent; range, 0 to 71 vs. 0 to 68; P = 0.48), blood-transfusion requirements (Table 2), or the hematocrit values at discharge (22 percent vs. 24 percent; range, 10 to 40 vs. 13 to 41; P = 0.18). However, patients in the artemether group had significantly lower mean reticulocyte counts one week after treatment (2.3 percent vs. 5.6 percent; range, 0.1 to 16.1 vs. 0.0 to 28; P<0.001). Blackwater (red or black urine caused by hemoglobinuria but not hematuria) developed in 11 patients: 7 in the artemether group and 4 in the quinine group (2 percent vs. 1 percent; relative risk, 2.7; 95 percent confidence interval, 0.86 to 8.3; P = 0.11). The overall incidence of blackwater in patients who had received quinine either before or during the trial was 4.9 percent (21 of 431; 95 percent confidence interval, 3.0 to 7.4 percent).
Adverse Effects
Quinine was associated with hypoglycemia: hypoglycemia developed in 25 percent of the quinine-treated patients as compared with 11 percent of the artemether recipients (relative risk, 2.3; 95 percent confidence interval, 1.6 to 3.4; P<0.001). There was an increased incidence of culture-negative pyuria in the artemether group (P = 0.03), but the incidence of urinary tract infections was not significantly different between groups. There were no other definite systemic adverse effects. There were no significant electrocardiographic abnormalities in the patients in whom recordings were made, although in 60 of the quinine recipients (45 percent), as compared with 38 of the artemether recipients (25 percent), the corrected QT interval was prolonged by more than 0.5 second (relative risk, 1.8; 95 percent confidence interval, 1.3 to 2.5; P = 0.001). Only 12 patients in the quinine group (9 percent) and 11 in the artemether group (7 percent) had a corrected QT interval that was prolonged by more than 25 percent (P = 0.67). Prolongation of the corrected QT interval was not associated with any other clinical finding, including the development of shock or the duration of coma. There were no differences between the two treatment groups in auditory acuity on discharge from the hospital.
Oral Treatment
Overall, 247 patients were randomly assigned to receive oral mefloquine and 224 to receive oral quinine; 89 patients did not receive oral antimalarial agents because they had either died (74) or had received seven days of parenteral treatment (15). Nine patients died after receiving oral antimalarial agents (four given mefloquine and five given quinine). The oral antimalarial agents had no effect on recovery times or parasite clearance as assessed by Cox regression analysis.
Discussion
Our results show that artemether is a safe and effective treatment for severe falciparum malaria in Vietnamese adults. The mortality rate among artemether-treated patients was 26 percent lower than that among quinine recipients, but the 95 percent confidence interval ranged from a 50 percent reduction to an 11 percent increase in mortality. However, in our prospectively designed multivariate analysis, which took into account other factors that contribute to outcome, treatment with artemether was associated with a significantly lower mortality rate than treatment with quinine. This suggests that artemether is at least as good as, and may be better than, quinine for severe chloroquine-resistant malaria. In a setting of relative quinine resistance,14 similar to that in Vietnam,15 artemether proved considerably superior to quinine in smaller, open trials of severe malaria conducted in Burma and Thailand.9,10,11 The large differences in mortality observed in these studies were not substantiated in our large and detailed double-blind, randomized trial. This cannot be ascribed to differences in clinical severity or drug administration, since the clinical and laboratory features of severe malaria in our trial were similar to those reported previously and the doses of artemether were higher than those usually recommended.
Although treatment with artemether resulted in a more rapid reduction in the level of parasitemia (as reported previously3), the overall times to defervescence, recovery of consciousness in patients with cerebral malaria, and discharge from the hospital were longer in artemether-treated patients. Survival analysis indicated a significant divergence in the times to recovery from coma among patients who remained unconscious for more than two days. In these patients, treatment with artemether was associated with slower recovery. There are three possible explanations for this unexpected finding: the play of chance, neurotoxicity,16,17 or a consequence of the beneficial effect of artemether — patients who would have died if they had received quinine survived because they received artemether and took longer to recover. Although there was no other evidence of neurotoxicity, an acute reversible drug effect cannot be excluded, particularly since the difference in the times to recovery from coma became evident after 48 hours, after a total of 14 mg of artemether per kilogram had been given. In support of the third explanation, the times to recovery from coma were strongly associated with other measures of disease severity, and these measures were similar in all patients whose recovery from coma took longer than 48 hours.
In uncomplicated malaria, the artemisinin derivatives consistently shorten all aspects of recovery, whereas in severe malaria, in which the processes that cause organ dysfunction and death may already be largely irreversible,1 they do not. Pharmacokinetic factors may contribute to this difference. Parenteral artemether is dissolved in groundnut oil and administered by intramuscular injection. As compared with orally administered artemether, intramuscularly administered artemether is absorbed slowly and incompletely. This delay in achieving therapeutic blood levels may offset any intrinsic pharmacodynamic advantage of artemether. In general, the water-soluble artesunate has resulted in the most rapid therapeutic responses.3 Artesunate is given intravenously and is also rapidly bioavailable after intramuscular or oral administration.18 Artesunate may be more effective than artemether in severe malaria. By contrast, absorption of intramuscular quinine in severe malaria is regular, and its plasma-concentration profiles are similar to those seen after intravenous administration.19,20
The intramuscular administration of quinine is painful and causes local tissue damage, which sometimes results in sterile abscesses21,22 and, occasionally, tetanus.23 In this study, both treatment regimens were well tolerated and only nine quinine-treated patients (3.2 percent) had abscess formation, difficulty walking because of local pain, or both, despite the use of a concentrated solution (250 mg per milliliter, with a pH of 2). Thus, the risks of serious local reactions from intramuscular quinine are low, provided a scrupulous aseptic technique of injection is used. Quinine causes hypotension if given too quickly by intravenous injection, and it also prolongs ventricular repolarization, but substantial (>25 percent) prolongation of the corrected QT interval occurred in only 12 patients (9 percent), and no serious toxic effects were associated with the use of a loading dose of quinine despite the fact that the majority of patients had been treated previously with quinine.24 Indeed, there were no serious cardiovascular or nervous system effects with either drug.
Quinine is a potent stimulator of the secretion of insulin by pancreatic beta cells25 and was associated with an increased risk of hypoglycemia. Overall, our results confirm that intramuscular administration of quinine is an acceptable alternative to intravenous administration and that the principal adverse effect of quinine in severe malaria is hypoglycemia.25 In rodents, dogs, and primates, artemether (and the related compound, arteether) induced an unusual and selective pattern of damage to certain brain-stem nuclei.16,17 This has been the main concern overlying the further development of these compounds. With the possible exception of prolonged recovery from coma in patients given artemether, we did not detect any evidence of permanent damage to the central nervous system despite the use of maintenance doses that were three times higher than those now usually recommended. The auditory nuclei are among the most sensitive to damage by these compounds, but there was no evidence of residual hearing impairment.
Artemether is an effective alternative to quinine for severe malaria. It is simple to administer, equivalent in overall cost to quinine, and has no apparent local or serious systemic adverse effects. It is one of a family of new antimalarial agents that are active against quinine-resistant parasites. These new drugs should not be used in an uncontrolled or unregulated way, or resistance to them will develop.
Supported by the Wellcome Trust of Great Britain.
We are indebted to the director and staff of the Center for Tropical Diseases for their support of the study, in particular Drs. N.T. Dung, T.T.M. Trang, B.M. Cuong, and D. Bethell and the nurses and technicians on the malaria ward; to Julie Simpson for statistical advice; and to Rhône–Poulenc Rorer for kindly donating the quinine and artemether.
Source Information
From the Wellcome Trust Clinical Research Unit, Center for Tropical Diseases, Ho Chi Minh City, Vietnam (T.T.H., N.H.P., N.T.H.M., T.T.H.C., P.P.L., D.X.S., L.V.C., H.V.); and the Centre for Tropical Medicine, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, United Kingdom (N.P.J.D., D.W., T.E.A.P., N.J.W.).
Address reprint requests to Dr. White at Wellcome Trust Clinical Research Unit, Center for Tropical Diseases, Cho Quan Hospital, Ho Chi Minh City, Vietnam.
References
| |||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Artemether in Severe Malaria
Gachot B., Doyon F., Hill C., Fruchter O., Newmark A., van Hensbroek M. B., Greenwood B., Kwiatkowski D., Hien T. T., Day N. P.J., White N. J.
Extract |
Full Text
N Engl J Med 1996;
335:1922-1924, Dec 19, 1996.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
