Correspondence
Antiemetic Effects of Ondansetron and Metopimazine
N Engl J Med 1993; 329:1356-1357October 28, 1993
- Article
To the Editor:
The conclusion of Herrstedt et al. (April 15 issue)1 that “ondansetron plus metopimazine is a highly effective and safe antiemetic regimen that is markedly superior to treatment with ondansetron alone in patients receiving moderately emetogenic chemotherapy” is misleading.
The primary end point of antiemetic treatment is complete protection from vomiting and nausea, not the mean number of emetic episodes and the mean score of nausea. These are generally evaluated as secondary end points.
Considering the data on complete protection from vomiting as marginals of a two-by-two table, we found that the two-sided McNemar test indicated no significance for any possible structure of the frequencies of association (we adopted this procedure because the data obtained for each treatment in the two cycles of chemotherapy were not shown). Therefore, the reduction in the mean number of emetic episodes in patients receiving ondansetron plus metopimazine could be due only to patients who had emesis.
The target population consisted of patients with emesis who were pretreated with antiemetics. Therefore, the results cannot be generalized to patients at lower risk, such as patients with cancer who are not pretreated and those who have had no emesis in previous cycles of chemotherapy.
The total number of the most frequently reported adverse events was higher with ondansetron plus metopimazine than with ondansetron alone (42 vs. 27).
Chemotherapy-induced emesis is highly variable2. Useful information from small studies, such as that of Herrstedt et al., can be obtained only if the regimens tested differ markedly in efficacy as regards major therapeutic end points (i.e., complete protection from emesis) or toxicity. This almost always necessitates validation in larger, controlled studies.
2 ReferencesFausto Roila, M.D.
Policlinico HospitalEnzo Ballatori, Ph.D.
Albano Del Favero, M.D.
University of Perugia, 06122 Perugia, Italy1
Herrstedt J ,Sigsgaard T ,Boesgaard M ,Jensen TP ,Dombernowsky P . Ondansetron plus metopimazine compared with ondansetron alone in patients receiving moderately emetogenic chemotherapy. N Engl J Med 1993;328:1076-1080
Full Text | Web of Science | Medline2
Tonato M ,Roila F ,Del Favero A . Methodology of antiemetic trials: a review. Ann Oncol 1991;2:107-114
Web of Science | Medline
Author/Editor Response
The authors reply:
To the Editor: Contrary to the view of Roila et al., it is not generally accepted that the only primary end point of antiemetic trials is the frequency of complete protection from nausea and vomiting. According to a recent review by Gralla et al.,1 “The results can be expressed in different ways. First, using the mean or median number of vomiting episodes per patient, with the range of episodes given for the group. Another related expression of efficacy is to use response categories.” Others prefer to use the number of emetic episodes instead of response categories as the primary end point,2 as did Roila et al. in 19873.
The use of response categories is complicated by the fact that there is no established definition of complete response. In studies with ondansetron the definition recently adopted by Roila et al. is used -- i.e., complete response means no emetic episodes, including nausea. In studies with granisetron complete response is defined as the absence of emetic episodes and of nausea or only mild nausea,4 and in studies with tropisetron complete response is defined as the absence of emetic episodes and of nausea5. These differences complicate comparisons that involve response rates.
The statement of Roila et al. that the two-sided McNemar test indicates no significance for any possible structure of frequencies of association is incorrect. On day 1 (acute emesis) the difference only approached significance (P = 0.058, chi2 = 3.57 with 1 df), but on days 2 to 5 (delayed emesis) the number of patients with a complete response was as follows: 14 patients with both treatments, 1 patient with ondansetron only, and 8 patients with ondansetron plus metopimazine only, meaning that significantly more patients receiving the combination had complete responses, defined as no emetic episodes (P = 0.020, chi2 = 5.44 with 1 df; or, with Yates' correction, P = 0.046, chi2 = 4.00 with 1 df).
A total of 35 adverse events were reported with ondansetron alone, and 49 with the combination. This difference is due primarily to the fact that patients reported significantly more mild constipation when receiving the combination. Not only the frequency but also the severity of adverse events is important. Only four patients graded an adverse event as severe -- and all when receiving ondansetron alone.
As two of the inclusion criteria in our study were that patients should have received previous chemotherapy and should have had at least one vomiting episode during their most recent cycle of chemotherapy, it is obvious that our results cannot be generalized to patients who have not previously received chemotherapy. New findings, whether from smaller crossover or larger parallel trials, need of course to be validated by other groups.
5 ReferencesJorn Herrstedt, M.D.
Tine Sigsgaard, M.D.
Per Dombernowsky, M.D.
University of Copenhagen, DK-2730 Herlev, Denmark1
Gralla RJ ,Clark RA ,Kris MG ,Tyson LB . Methodology in anti-emetic trials. Eur J Cancer 1991;27:Suppl 1:S5-S8
CrossRef | Web of Science | Medline2
Pater JL ,Willan AR . Methodologic issues in trials of antiemetics. J Clin Oncol 1984;2:484-487
Web of Science | Medline3
Roila F ,Tonato M ,Basurto C , et al. Antiemetic activity of high doses of metoclopramide combined with methylprednisolone versus metoclopramide alone in cisplatin-treated cancer patients: a randomized double-blind trial of the Italian Oncology Group for Clinical Research. J Clin Oncol 1987;5:141-149
Web of Science | Medline4
Chevallier B . The control of acute cisplatin-induced emesis -- a comparative study of granisetron and a combination regimen of high-dose metoclopramide and dexamethasone. Br J Cancer 1993;68:176-180
CrossRef | Web of Science | Medline5
Bleiberg H ,Van Belle S ,Paridaens R ,De Wasch G ,Dirix LY ,Tjean M . Compassionate use of a 5-HT3-receptor antagonist, tropisetron, in patients refractory to standard antiemetic treatment. Drugs 1992;43:Suppl 3:27-32
CrossRef | Medline
