|
|
| |||||||||||||||||||||||||||||||
Background Although ondansetron was found to be effective as an antiemetic in numerous clinical trials of highly emetogenic combination-chemotherapy regimens that included cisplatin, its role in milder emetogenic regimens has not been fully defined. To address its use with a widely used but less emetogenic regimen, we performed a double-blind, randomized clinical trial comparing ondansetron with dexamethasone and metoclopramide in patients with breast cancer receiving chemotherapy with cyclophosphamide, methotrexate, and fluorouracil.
Methods A total of 165 women with breast cancer from 14 Canadian centers who were about to receive this chemotherapy for the first time were randomly assigned to receive either ondansetron (n = 85) or dexamethasone plus metoclopramide (n = 80), a widely used, standard antiemetic regimen. The patients recorded the incidence of nausea, emesis, and other side effects in diaries, and these data were compared in the two groups.
Results The patients who received dexamethasone and metoclopramide had significantly less nausea during the first 24 hours after chemotherapy was begun. Otherwise, there were no statistically significant differences in efficacy between the regimens. The incidence of drowsiness and increased appetite was higher in the group given dexamethasone and metoclopramide.
Conclusions For women with breast cancer who are being treated with cyclophosphamide, methotrexate, and fluorouracil, the efficacy of dexamethasone and metoclopramide in controlling nausea and vomiting equaled or exceeded that of ondansetron. (n Engl J Med 1993;328:1081-4.).
Methods
Ethical Considerations
The study was approved by the local human investigation committee of each participating center and was carried out in accordance with and with the approval of the Health Protection Branch of the Department of Health and Welfare, Government of Canada, with an assurance filed with and approved by the U.S. Department of Health and Human Services. The provisions of the Declaration of Helsinki were followed.
Patient Selection
Women were eligible for the study only if they had never received any form of cytotoxic chemotherapy and were about to begin treatment with the following regimen: intravenous fluorouracil at a dose of 600 mg per square meter of body-surface area and intravenous methotrexate at a dose of 40 mg per square meter on days 1 and 8 and oral cyclophosphamide at a dose of 100 mg per square meter daily on days 1 through 14. Chemotherapy could be given as an adjuvant to surgery or for metastatic disease. Other conditions for eligibility included a performance status of 0 to 2 as defined by the Eastern Cooperative Oncology Group (i.e., the patients were capable of only limited self-care and were confined to a bed or chair more than 50 percent of waking hours; this corresponds to a Karnofsky score of 30 to 40), an age of at least 18 years, availability for follow-up, and the provision of informed consent. Patients were excluded if they had severe concurrent illness, were clinically jaundiced or had laboratory evidence of hepatic dysfunction not attributable to metastatic involvement, or had a condition for which dexamethasone was medically contraindicated (e.g., active peptic ulcer or severe diabetes mellitus). Patients could not have received antiemetic therapy outside the study or have had nausea or vomiting within 24 hours before entry into the study; the concurrent use of non-study glucocorticoids, benzodiazepines, or both except for nighttime sedation was prohibited.
Randomization and Treatment
After eligibility criteria were ascertained and written informed consent was obtained, the patients were enrolled in the study and randomly assigned to receive either ondansetron (experimental treatment) or dexamethasone and metoclopramide (standard treatment). The entry of each patient was confirmed by telephoning the central randomization office at the National Cancer Institute of Canada Clinical Trials Group in Kingston, Ontario. Each patient was issued a study number corresponding to a previously prepared and blinded medication package stored in the center pharmacy. Each package contained either the experimental or control medication, ensuring that the patient and direct care givers remained unaware of the actual treatment selected.
Each patient assigned to the experimental treatment was given 8 mg of ondansetron orally one to two hours before intravenous chemotherapy was begun and an intravenous injection of placebo immediately before chemotherapy was begun. On leaving the treatment center, each patient was given a vial containing 22 8-mg capsules of ondansetron. These were taken every eight hours for the seven-day study period.
Each patient assigned to the standard treatment was given a 10-mg capsule of metoclopramide one to two hours before chemotherapy was begun and received 10 mg of dexamethasone intravenously immediately before intravenous chemotherapy was begun, to correspond with the placebo injection in the experimental group. Each patient in this group was given a vial of 10-mg capsules of metoclopramide that looked like the ondansetron capsules. These were also taken every eight hours for the duration of the study.
Oral cyclophosphamide was given either every eight hours or in a single daily dose at the discretion of the treating physician. The patients in both groups were given 10 mg of prochlorperazine to be taken orally or rectally if they had nausea or vomiting that was not controlled by the study drugs. The patients used diaries to record all medications taken in the seven days after chemotherapy was begun. All patients returned the vials and their diaries to the treatment center so that compliance and other data could be analyzed.
Statistical Analysis
The primary objective of this double-blind, parallel, multicenter, randomized clinical trial was to compare the antiemetic efficacy and side effects of ondansetron with those of dexamethasone and metoclopramide.
Pretreatment information recorded for all patients included height, weight, stage of disease (disease requiring adjuvant surgery or metastatic disease), complete blood counts, plasma electrolyte levels, and renal and liver function.
Study outcomes were recorded by patients after they had been instructed in the use of their diaries. The presence and severity of nausea and vomiting were recorded at base line and at 6-hour intervals for the first 24 hours and daily thereafter for the next six days. Each patient was asked to count the number of vomiting episodes during each interval (i.e., 0, 1 to 2, or 3 or more episodes). The duration and severity of nausea were rated for the same intervals on separate 100-mm visual-analogue scales. The degree of drowsiness and anxiety was also recorded on similar scales. The following basic descriptors were used for the various visual-analogue scales: "none of the time" and "all of the time" for the duration of nausea; "no nausea" and "extremely severe nausea" for the severity of nausea; "never drowsy" and "constantly drowsy" for drowsiness; and "no anxiety" and "severe anxiety" for anxiety. Food intake during the seven 24-hour periods was ranked on a four-point scale as ranging from "better than usual" to "liquids only." All patients were also instructed to record any other symptoms and potential adverse events. The diaries were collected at the clinic visit on day 8. This method of data collection has been a convenient and reliable means of evaluating nausea and vomiting as primary outcomes9,10.
The study was planned to include 160 patients to provide a power of more than 90 percent to detect at the 5 percent level a 10-mm difference (on a visual-analogue scale) in the mean severity of nausea over the seven-day period. This sample size also provided an 88 percent likelihood of detecting a 25 percent increase in the complete control of emesis from the anticipated 50 percent in the standard-treatment group. No interim analysis was planned or carried out.
The chi-square statistic was used to test the significance of the difference between treatment groups in the proportions of patients who had not vomited during seven days of the study.
Logistic regression was used to assess and control for the effects of important prognostic factors. Putative prognostic factors included age, Eastern Cooperative Oncology Group performance status (0 to 2), stage of disease (metastatic disease vs. disease requiring adjuvant surgery), center size (fewer than 20 patients enrolled vs. 20 or more patients enrolled), dose of chemotherapy regimen, and the schedule for the oral administration of cyclophosphamide (once daily vs. divided doses).
Multivariate categorical data analysis, as described by Landis et al.,11 was used to test whether the profile of the distribution of responses in the two treatment groups varied across time within a cumulative logit model.
The severity and duration of nausea, drowsiness, and anxiety were analyzed with the Wilcoxon rank-sum test for each interval. Stepwise multiple linear regression was used to assess and adjust for the effects of putative prognostic factors. The longitudinal structures of these outcomes were examined by multivariate analysis of variance. The parallelism of the profiles and the overall difference between the two treatment groups were examined with Wilks' lambda statistic. The method of Andrews et al.12 was used to determine an appropriate transformation to normality in these outcomes before multivariate analyses were performed.
Results
Between July 1989 and November 1990, a total of 165 patients from 14 Canadian centers were enrolled in the study. All randomized patients were eligible. One patient randomly assigned to receive ondansetron mistakenly did not receive oral cyclophosphamide and was therefore excluded from the efficacy analysis. Several patients did not complete their diaries. The data were analyzed both with and without the data on these patients and with different assumptions concerning their outcomes (e.g., that they either had or did not have treatment failure during the interval in question). The conclusions from all sets of analyses were similar. The results presented exclude patients for whom outcome data were not available at the specific interval examined. Overall, the data were at least 93 percent complete for each interval.
Of the 164 eligible patients enrolled, 84 were randomly assigned to receive ondansetron and 80 to receive dexamethasone and metoclopramide. The mean (±SD) ages of the groups were 53 ±13 and 51 ±11 years, respectively. There was almost no difference in the mean (±SD) heights (159.5 ±5.8 vs. 160.5 ±5.9 cm) and weights (67.6 ±12.1 vs. 67.3 ±11.8 kg) between the patients who received ondansetron and those who received dexamethasone and metoclopramide.
The proportion of patients who did not vomit during the seven-day study period is shown in Figure 1. The distributions of these proportions were analyzed with a cumulative logit model11 with a treatment effect and a time effect. Figure 1 suggests that patterns of freedom from vomiting did not vary over time between the two treatment groups, an impression supported by the fact that the test for the deviation from parallel profiles was not significant. As determined by the cumulative logit model, the overall mean proportions of patients who did not vomit during the seven days after chemotherapy was begun were not significantly different between the group given ondansetron (83 percent; 95 percent confidence interval, 74 to 89 percent) and the group given metoclopramide and dexamethasone (87 percent; 95 percent confidence interval, 79 to 92 percent). More than 66 percent of the patients did not vomit during the seven-day study. The apparent advantage of combination therapy with dexamethasone and metoclopramide over treatment with ondansetron alone was not statistically significant (71 percent of the patients given metoclopramide plus dexamethasone as compared with 61 percent of the patients given ondansetron; 95 percent confidence interval for the difference, -24 percent to +4 percent). Seventeen of the patients in the ondansetron group (20 percent) and 11 of the patients in the group given dexamethasone and metoclopramide (14 percent) required prochlorperazine (P = 0.30).
|
|
|
|
|
As compared with results obtained without the use of antiemetics,13 chemotherapy-induced emesis was well controlled both with ondansetron and with dexamethasone and metoclopramide. More than 70 percent of the patients did not vomit during the seven-day study period. When antiemetic therapy is not used, emesis may be an important problem13. The patients who received dexamethasone and metoclopramide, however, had more complete relief of nausea than those given ondansetron, a difference that was statistically significant during the first 24 hours of the study and of borderline significance (P = 0.08) for the seven-day period as a whole. Ondansetron was less effective in controlling the severity of nausea during the entire seven-day period (Figure 3), but was especially so during the first 24 hours after chemotherapy was begun (Figure 2), and it was about as effective as dexamethasone and metoclopramide in preventing emesis.
Dexamethasone has been claimed to enhance the antiemetic efficacy of ondansetron14. This suggests that chemotherapy-induced nausea and emesis are not caused by a simple mechanism at a single site (e.g., the serotonin-receptor site), but rather involve a variety of mechanisms. Such speculation could explain our observations, if dexamethasone was a better antinauseant than ondansetron early in the course of chemotherapy (the first 48 hours) and if metoclopramide was either unnecessary or approximately equivalent to ondansetron later in the course of chemotherapy.
The moderate emetogenic potential of chemotherapy with cyclophosphamide, methotrexate, and fluorouracil has been shown to be dose-related15. There was no significant difference between the treatment groups in the number of patients who did not vomit. There was a trend toward more drowsiness among the patients who received dexamethasone and metoclopramide. There was no detectable difference in the levels of anxiety between the two groups, nor was there any important difference in the ability to tolerate a clinical dose of any of the three chemotherapeutic drugs (Table 1).
In another study, ondansetron and dexamethasone were equivalent in patients whose chemotherapy included an intravenous anthracycline, cyclophosphamide, and etoposide (alone or in combination)16. In a retrospective analysis of our previous studies, we found that there were no apparent beneficial effects of treatment with serotonin antagonists in patients receiving chemotherapy in which cyclophosphamide was the primary emetogenic agent17. It appears that ondansetron (perhaps as representative of the serotonin-antagonist group of antiemetics) will require further clinical evaluation.
Supported by the National Cancer Institute of Canada and Glaxo Pharmaceuticals Canada.
We are indebted to the patients, data managers, nurses, physicians, pharmacists, and other medical personnel who participated; to the staff of the Central Office of the Clinical Trials Group of the National Cancer Institute of Canada and the Clinical Investigations Office of the Manitoba Cancer Foundation for their intellectual and technical support; to Lorna Geeves, Joan Powell, and Susan Ireland-Clarke for assistance in the preparation of the manuscript; and to Glaxo Pharmaceuticals for sponsorship.
Source Information
From the Department of Internal Medicine, University of Manitoba, the Manitoba Cancer Treatment and Research Foundation, Winnipeg (M.L.); Princess Margaret Hospital (D.W.), St. Michael's Hospital (E.W.), and Toronto Bayview Regional Cancer Center (K.I.P.), Toronto; Notre Dame Hospital (L.Y.), and Hotel Dieu de Montreal (J.L.), Montreal; British Columbia Cancer Agency, Victoria Clinic, Victoria (H.L.R., K.S.W.); Kingston Regional Cancer Center (W.S.L.) and the National Cancer Institute of Canada, Clinical Trials Group (M. Palmer, B.Z., J.L.P.), Kingston, Ont.; Ottawa Regional Cancer Center, Ottawa, Ont. (D.J.P.); and Laval Hospital, Ste.-Foy, Que. (M. Potvin).
Address reprint requests to Dr. Levitt at the Manitoba Cancer Treatment and Research Foundation, 100 Olivia St., Winnipeg, MB R3E 0V9, Canada.
References
| |||||||||||||||||||||||||||||||
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
