Clinical Oncology (1990) 2:71-75 © 1990 The Royal College of Radiologists
Clinical Oncology
Original Article Results of a Randomized, Double-Blind Comparative Study of Ondansetron and Metoclopramide in the Prevention of Nausea and Vomiting Following High-Dose Upper Abdominal Irradiation T. J. Priestman 1, J. T. Roberts 2, H. Lucraft 2, C. H. Collis 3, M. Adams 4, B. K. Upadhyaya 5 and S. Priestman 1 1Department of Radiotherapy, Queen Elizabeth Hospital, Birmingham, 2Department of Radiotherapy, Newcastle General Hospital, Newcastle upon Tyne, 3Department of Radiotherapy and Oncology, Royal Free Hospital, London, 4South Wales Radiotherapy and Oncology Service, Velindre Hospital, Cardiff and 5Glaxo Group Research Limited, Greenford
Abstract. Ondansetron is a 5-hydroxytryptamine 3receptor antagonist which has shown activity in the prevention of cytotoxic-induced emesis. Preliminary non-randomized studies also indicated efficacy in preventing sickness following radiotherapy. The present study was therefore undertaken to compare the efficacy and safety of ondansetron (8 mg tds orally) and metoclopramide (10 mg tds orally) in preventing sickness after single-exposure radiotherapy treatments of 8-10 Gy to the upper abdomen. Of 82 evaluable patients 38 received ondansetron and 44 metoclopramide. On the first day after irradiation vomiting or retching was prevented in all but one of the patients on ondansetron whereas metoclopramide achieved complete control of these symptoms in only 46% of subjects (P5 emetic episodes or alternative antiemetic therapy required). The response was then graded as 'success' (complete or major response) or 'failure' (all other categories). Proportions of 'successes' were compared between the two arms using the Mantel-Haenzel chi-squared test and the estimates of possibility of a success were derived together with 95% confidence intervals. The numbers of emetic episodes were analysed by non-parametric methods using Wilcoxon rank sums, as were the differences, from baseline, in daily nausea visual analogue scores. The grades of nausea on the verbal rating scale were analysed using an extended Mantel-Haenzel method (Mantel, 1963).
RESULTS
At the point of interim analysis 97 patients had been randomized and their characteristics are summarized in Table 1. There were no significant differences between the groups. One patient was aged 83 years but was retained in the analysis. Forty six patients received ondansetron and 51 metoclopramide. Prior to breaking the treatment code, however, eight patients who had received ondansetron and seven who had been given metoclopramide were withdrawn from the analysis. The reasons for withdrawal were as follows: seven patients (four on ondansetron, three on metoclopramide) were receiving steroids at a dose equiva-
Ondansetron in the Prevention of Nausea After Radiotherapy
73
Table 1. Patient characteristics Ondansetron
Metroclopramide
Number of patients 46 (23M; 23F) Median age (range) years 63 (33-77)
51 (25M; 26F) 65 (39-83)
Primary tumour sites Lung Breast Gastrointestinal Genitourinary Other
3 13 14 10 6
8 12 14 7 10
Site(s) of metastases Lumbar/thoracic spine Pelvis Liver Abdominal
22 4 8 12
26 10 9 6
lent to or greater than 10 mg of prednisolone daily at the time of their irradiation and it was felt that this might have influenced their emetic response; four patients (two in each group) were taking the analgesic Diconal at the time of their irradiation and this drug contains the antiemetic cyclizine; two patients on metoclopramide were found to be taking other antiemetics at the time of their radiotherapy; one patient on ondansetron vomited in the 24 hours prior to irradiation and one failed to return their diary card. The interim analysis was therefore based on 82 evaluable patients: 38 on ondansetron and 44 on metoclopramide. The control of vomiting and/or retching on day 1 and the outcome of the 'worst day' for days 2-3 and 1-3 is shown in Table 2. On Day 1, 37 of 38 patients receiving ondansetron had no emetic episodes and were categorized as complete responders, the remaining patient experienced a single episode of retching and was therefore classified as a major responder. In the metoclopramide group 20 of 44 (46%) had a complete response and 11 (25%) a major response; 13 experienced more than three emetic episodes and for the purpose of analysis were considered treatment failures. The difference between the two arms was statistically significant (P
Clinical Oncology
Original Article Results of a Randomized, Double-Blind Comparative Study of Ondansetron and Metoclopramide in the Prevention of Nausea and Vomiting Following High-Dose Upper Abdominal Irradiation T. J. Priestman 1, J. T. Roberts 2, H. Lucraft 2, C. H. Collis 3, M. Adams 4, B. K. Upadhyaya 5 and S. Priestman 1 1Department of Radiotherapy, Queen Elizabeth Hospital, Birmingham, 2Department of Radiotherapy, Newcastle General Hospital, Newcastle upon Tyne, 3Department of Radiotherapy and Oncology, Royal Free Hospital, London, 4South Wales Radiotherapy and Oncology Service, Velindre Hospital, Cardiff and 5Glaxo Group Research Limited, Greenford
Abstract. Ondansetron is a 5-hydroxytryptamine 3receptor antagonist which has shown activity in the prevention of cytotoxic-induced emesis. Preliminary non-randomized studies also indicated efficacy in preventing sickness following radiotherapy. The present study was therefore undertaken to compare the efficacy and safety of ondansetron (8 mg tds orally) and metoclopramide (10 mg tds orally) in preventing sickness after single-exposure radiotherapy treatments of 8-10 Gy to the upper abdomen. Of 82 evaluable patients 38 received ondansetron and 44 metoclopramide. On the first day after irradiation vomiting or retching was prevented in all but one of the patients on ondansetron whereas metoclopramide achieved complete control of these symptoms in only 46% of subjects (P5 emetic episodes or alternative antiemetic therapy required). The response was then graded as 'success' (complete or major response) or 'failure' (all other categories). Proportions of 'successes' were compared between the two arms using the Mantel-Haenzel chi-squared test and the estimates of possibility of a success were derived together with 95% confidence intervals. The numbers of emetic episodes were analysed by non-parametric methods using Wilcoxon rank sums, as were the differences, from baseline, in daily nausea visual analogue scores. The grades of nausea on the verbal rating scale were analysed using an extended Mantel-Haenzel method (Mantel, 1963).
RESULTS
At the point of interim analysis 97 patients had been randomized and their characteristics are summarized in Table 1. There were no significant differences between the groups. One patient was aged 83 years but was retained in the analysis. Forty six patients received ondansetron and 51 metoclopramide. Prior to breaking the treatment code, however, eight patients who had received ondansetron and seven who had been given metoclopramide were withdrawn from the analysis. The reasons for withdrawal were as follows: seven patients (four on ondansetron, three on metoclopramide) were receiving steroids at a dose equiva-
Ondansetron in the Prevention of Nausea After Radiotherapy
73
Table 1. Patient characteristics Ondansetron
Metroclopramide
Number of patients 46 (23M; 23F) Median age (range) years 63 (33-77)
51 (25M; 26F) 65 (39-83)
Primary tumour sites Lung Breast Gastrointestinal Genitourinary Other
3 13 14 10 6
8 12 14 7 10
Site(s) of metastases Lumbar/thoracic spine Pelvis Liver Abdominal
22 4 8 12
26 10 9 6
lent to or greater than 10 mg of prednisolone daily at the time of their irradiation and it was felt that this might have influenced their emetic response; four patients (two in each group) were taking the analgesic Diconal at the time of their irradiation and this drug contains the antiemetic cyclizine; two patients on metoclopramide were found to be taking other antiemetics at the time of their radiotherapy; one patient on ondansetron vomited in the 24 hours prior to irradiation and one failed to return their diary card. The interim analysis was therefore based on 82 evaluable patients: 38 on ondansetron and 44 on metoclopramide. The control of vomiting and/or retching on day 1 and the outcome of the 'worst day' for days 2-3 and 1-3 is shown in Table 2. On Day 1, 37 of 38 patients receiving ondansetron had no emetic episodes and were categorized as complete responders, the remaining patient experienced a single episode of retching and was therefore classified as a major responder. In the metoclopramide group 20 of 44 (46%) had a complete response and 11 (25%) a major response; 13 experienced more than three emetic episodes and for the purpose of analysis were considered treatment failures. The difference between the two arms was statistically significant (P
