Aim To compare the effectiveness of ondansetron and granisetron in the prevention of nausea and vomiting caused by cisplatin.
Design Double-blind crossover study with random allocation using a Latin square design in sets of four. Patients were given either ondansetron 8mg plus IV bolus dexamethasone 8mg or granisetron 3mg by IV infusion plus IV bolus dexamethasone 8mg immediately prior to chemotherapy and oral dexamethasone 4mg three times a day on days 2 to 4.
Subjects and setting 21 patients attending Airedale general hospital for treatment with highly emetogenic chemotherapy.
Outcome measures Nausea and vomiting scores on days 1 to 7 post chemotherapy. The criterion for success of prevention of nausea and vomiting was that patients would suffer from no more than mild nausea on day 1.
Results 40 courses of ondansetron and 49 courses of granisetron were studied. In 90% of courses patients suffered from no nausea or mild nausea on day 1 and in 77.5% of courses patients suffered from no vomiting on day 1 when given ondansetron. In comparison, in 94% of courses patients suffered from no nausea or mild nausea on day 1 and in 88% of courses patients suffered from no vomiting on day 1 when given granisetron. The mean nausea score for days 1 to 7 for patients treated with ondansetron was always more than for those treated with granisetron. The biggest difference was on day 2 when 35% (95% CI 20.6–51.7) of patients treated with ondansetron suffered from moderate to severe nausea compared with 22% (95% CI 20.6–51.7) of patients treated with granisetron. The mean number of episodes of vomiting on days 1 to 7 for patients treated with ondansetron was nearly always more than for those treated with granisetron. The biggest difference was on day 2 when 50% (95% CI 33.8–66.2) of patients treated with ondansetron as the antiemetic suffered from 2 or more episodes of vomiting compared with 18% (95% CI 8.8–32) of patients treated with granisetron.
Conclusions Granisetron is more effective than ondansetron when used in combination with dexamethasone in the prevention of acute and delayed vomiting caused by highly emetogenic chemotherapy. It is significantly better on day 2 when the majority of patients suffer most.
Recent developments in our understanding of the pathophysiology of emesis and the availability of a wide selection of drugs should allow good control of acute emesis in patients undergoing chemotherapy. Effective antiemetic cover is especially important with the first cycle of treatment to avoid problems with anticipatory vomiting on subsequent courses.1 There is, however, considerable room for improvement, particularly with regard to the problem of delayed emesis, which remains a significant cause of treatment-related morbidity in cancer patients despite the use of various antiemetic drugs including 5-hydroxytryptamine type 3 (5HT3) antagonists. There are currently three licensed in many countries - ondansetron, granisetron and tropisetron - and there are others in an advanced state of clinical evaluation.2 Although their chemical structures are similar, they show some differences in potency, pharmacokinetic and pharmacodynamic characteristics. A variety of dosage schedules of administration have been suggested and possible differences in efficacy among these compounds cannot be excluded.3
Various studies have been published comparing the 5HT3 receptor antagonists, but convincing data on clinically significant differences are still lacking for many of them, particularly in relation to delayed emesis.4 Ondansetron and granisetron are both claimed to be highly effective in the prevention of acute nausea and vomiting caused by chemotherapy.5 Granisetron has a longer half-life than ondansetron and is claimed to be more effective in the prevention of delayed nausea and vomiting associated with chemotherapy, but this is usually in combination with dexamethasone. The effectiveness of ondansetron is improved when used in combination with dexamethasone.6 Dexamethasone is known to be useful in the control of delayed emesis when used in conjunction with other antiemetics.7,8
The Italian Group for Antiemetic Research carried out a study comparing intravenous ondansetron 8mg versus granisetron 3mg. Dexamethasone was added to both treatments, 20mg by IV infusion 45 minutes before the cisplatin and 8mg IM twice a day on days 2 and 3 and 4mg twice a day on day 4. The group concluded from this study that ondansetron 8mg and granisetron 3mg combined with dexamethasone in this schedule showed similar efficacy and tolerability in the prevention of cisplatin-induced emesis.4 The use of oral dexamethasone for the prevention of delayed emesis was not investigated in this study.
Combinations of antiemetics may be more effective than a single drug but further studies are needed to establish the most effective combinations and whether particular combinations are more suitable for specific chemotherapy regimens or specific patients.
The optimum dose of ondansetron has been investigated in several studies. A single dose of 32mg IV was found to be no different to six-hourly doses of 8mg over a 24-hour period. However, there were some differences in the hour-to-hour control of nausea and vomiting.9
Krzakowski et al compared the effectiveness of IV ondansetron 8mg in combination with IV dexamethasone 20mg with oral ondansetron 24mg in combination with oral dexamethasone 12mg in once daily regimens administered to patients receiving cisplatin 50mg/m2 or greater. Complete control of emesis was achieved in 85 per cent of patients in the oral group and 83 per cent in the IV group. No nausea was reported in 70 per cent of patients in the oral group and 68 per cent in the IV group.10
The Italian Group for Antiemetic Research also carried out a double-blind, dose-finding study of four intravenous doses of dexamethasone in combination with ondansetron 8mg IV in the prevention of cisplatin-induced emesis. Complete protection from vomiting was significantly superior in patients who received 20mg compared with those who received 4mg and 8mg of dexamethasone (P>0.005) and was superior, but not significantly, compared with those who received 12mg. Complete protection from nausea was superior, but not significantly, in patients who received 20mg of dexamethasone.11
A double-blind, placebo-controlled comparative study by Goedhals concluded that granisetron plus dexamethasone did not appear to confer any benefit over the use of dexamethasone alone in controlling delayed nausea and vomiting.12
Similarly, in a study comparing the use of tropisetron in combination with dexamethasone or dexamethasone alone in the control of cisplatin-induced delayed emesis, no significant differences were recorded regarding the control of delayed emesis.13
The Italian Group for Antiemetic Research identified the dose of cisplatin, the dependence effect and the interaction between antiemetic treatment and the results of acute emesis control as important determinants for the occurrence of delayed emesis. The group was unable to demonstrate that age, sex, Karnofsky performance status, treatment modality or type of cancer were important factors influencing the occurrence of delayed emesis.14
Different cytotoxic drugs have different emetogenic potential and comparison of the efficacy of antiemetic drugs must be made in the light of similar emetic challenge.15 Other factors that influence the severity of emesis must also be considered in the design of the trial. Female patients are more likely to experience emesis. Age can be a determinant of the side effects seen with different antiemetics.16 Previous treatment with chemotherapy may also influence the severity of nausea and vomiting.1
There is some question as to whether 5HT3 antagonists maintain their efficacy over repeated cycles of treatment with cisplatin. Tzekova et al assessed the antiemetic efficacy of granisetron in repeated cycles of chemotherapy with platinum derivatives. The study included 50 patients who received between two and five cycles of chemotherapy with cisplatin or carboplatin. Total control of emesis was achieved in 60 per cent of patients after the first cycle of chemotherapy, and this percentage did not change significantly over the first five cycles of chemotherapy.17 However, this study did not look at delayed emesis and patients were not given dexamethasone with the granisetron. De Wit et al assessed the sustainment of efficacy of granisetron in combination with dexamethasone in 125 patients scheduled to receive 70mg/m2 either alone or in combination with other drugs. Patients received IV dexamethasone and granisetron on day 1 and oral dexamethasone and granisetron on days 1 to 6. Irrespective of the type of analysis used, the antiemetic efficacy of granisetron/dexamethasone decreased over cycles. The initial complete acute emesis control decreased from 66 per cent to 30 per cent. For delayed emesis, the initial complete protection rate of 52 per cent decreased to 21 per cent. In addition they observed that the protection failure in the delayed emesis period adversely influenced the acute emesis protection in the next cycle.18
Improved antiemetic control makes cancer chemotherapy more acceptable and allows more chemotherapy to be given in the outpatients setting and admissions to be shorter. This would be more convenient and more economical than treating patients in hospital. With less vomiting there is less risk from dehydration and consequent renal failure.
These considerations prompted us to perform a study to compare the effectiveness of ondansetron 8mg with granisetron 3mg in combination with intravenous dexamethasone on day 1 and oral dexamethasone on days 2 to 4 in controlling acute and delayed emesis caused by highly emetogenic chemotherapy.
Following a submission to the local research ethics committee at Airedale general hospital, it was agreed to investigate the comparative effectiveness of ondansetron and granisetron in the prevention of acute and delayed nausea and vomiting associated with highly emetogenic chemotherapy.
Study protocol The study protocol was as follows:
Inclusion criteria Patients attending Airedale general hospital over an 18-month period for treatment with highly emetogenic chemotherapy, including cisplatin.
Exclusion criteria Hypersensitivity to on-dansetron, granisetron or related substances.
Consent Informed consent was obtained from all patients who were then randomly allocated to receive either ondansetron or granisetron for each treatment period within a course of chemotherapy, as part of a set protocol for the prevention of nausea and vomiting.
Design of randomisation Many comparisons employ crossover studies whereby drug A and drug B are crossed over within the same patient on the first two cycles of chemotherapy. However, patients failing to achieve adequate antiemetic control on the first cycle may possibly be subject to so-called period interaction and are likely to experience a similar pattern of emesis with the second cycle.1 For these reasons and because of the difficulty in matching patients for disease, treatment and other factors, it was decided to use a double-blind, crossover study with random allocation using a Latin square design in sets of four. This would also address the possible problem of lack of sustainment of antiemetic efficacy after repeated cycles of cisplatin. Possible combinations using this design are shown in Table 1.
Treatment courses 5 to 8 would follow a similar pattern to this.
As only relatively small numbers of patients were likely to be involved, each patient would receive in their first four treatment courses at least two courses of ondansetron and two of granisetron, rather than all one or the other. This would ensure that information would be obtained on both ondansetron and granisetron. The same antiemetic would not always be used for the first treatment course. It was intended that 20 patients would be entered into the study and each patient would receive at least six courses of chemotherapy and possibly up to 10. It was, therefore, anticipated that approximately 120 treatment courses would be studied. Using this method of double-blind, crossover design, patients act as their own controls to rule out the effect of individual variation and prescriber influence on patients' expectations.
Table 1: Possible course combinations |
||||
| Course 1 | Course 2 | Course 3 | Course 4 | |
| Patient 1 | Ondansetron | Ondansetron | Granisetron | Granisetron |
| Patient 2 | Ondansetron | Granisetron | Granisetron | Ondansetron |
| Patient 3 | Ondansetron | Granisetron | Ondansetron | Granisetron |
| Patient 4 | Granisetron | Ondansetron | Ondansetron | Granisetron |
| Patient 5 | Granisetron | Ondansetron | Granisetron | Ondansetron |
| Patient 6 | Granisetron | Granisetron | Ondansetron | Ondansetron |
Treatment protocol Antiemetic therapy to be given immediately prior to the chemotherapy on day 1 was: ondansetron 8mg in 100ml sodium chloride 0.9 per cent given as an intravenous infusion over 10 minutes, plus dexamethasone 8mg IV bolus, or granisetron 3mg in 100ml sodium chloride 0.9 per cent given as an intravenous infusion over 10 minutes, plus dexamethasone 8mg IV bolus. On days 2 to 4 therapy was dexamethasone 4mg three times a day orally and/or metoclopramide 10 or 20mg orally.
Data collection It is important to make comparisons between drugs at therapeutic doses and to assess both the acute and delayed phases of emesis. Using a standard data collection form, information was collected for each course of chemotherapy for each patient. Information regarding the severity of nausea and the number of episodes of retching or vomiting recorded. The layout of the form was explained to each of the patients at the beginning of each course of chemotherapy (day 1). Patients were asked to record on the form, by ticking the relevant boxes, details of the severity of nausea and the number of episodes of retching and vomiting on days 1 to 7. The form was then returned to the investigators in the stamped, addressed envelope provided. Data were also collected from patients' notes and through interview.
The severity of nausea and the number of episodes of retching or vomiting for each 24-hour period was scored as outlined in Table 2.
It was agreed that if at any time during the study it became apparent that either treatment was vastly superior the study would be terminated.
Table 2: Scoring for severity of nausea and the number of episodes of retching or vomiting for each 24h period |
|||
| Severity of nausea | Score | Number of episodes | Score |
| None | 0 | None | 0 |
| Mild | 1 | 1 | 1 |
| Moderate | 2 | 2 | 2 |
| Severe | 3 | 3 | 3 |
| 4 | 4 | ||
| More than 4 | 5 | ||
Twenty-one patients (12 women, nine men) were entered into the study. Results were not collected from five patients for the following reasons:
Data on control of emesis were collected for 16 patients (10 women, six men) who had received at least one treatment with each of ondansetron and granisetron.
The mean age of patients included in the results was 56 (range 37–74). All patients received cisplatin mean dose 74mg/m2 (range 59–100mg/m2). A total of 40 courses of ondansetron and 49 courses of granisetron were studied.
The criterion for success of prevention of nausea and vomiting with ondansetron and granisetron would be that patients would suffer no more than mild nausea on day 1.
In 90 per cent of courses, patients suffered from no nausea or mild nausea on day 1 and in 77.5 per cent of courses patients suffered from no vomiting on day 1 when given ondansetron as the antiemetic.
In comparison, in 94 per cent of courses, patients suffered no nausea or mild nausea on day 1 and in 88 per cent of courses patients suffered from no vomiting on day 1 when given granisetron as the antiemetic.
Figure 1 shows the mean nausea scores on days 1 to 7 post chemotherapy for the group as a whole when treated with ondansetron or granisetron as the antiemetic prior to chemotherapy. The mean nausea score on day 1 was 0.65 with ondansetron and 0.44 with granisetron for the group of patients as a whole. The mean number of episodes of vomiting on day 1 was 0.68 with ondansetron and 0.43 with granisetron.
The mean nausea score for days 1 to 7 for patients treated with granisetron was always lower than for those treated with ondansetron. The biggest difference was on day 2. On further analysis of the results for day 2, it was found that the number of courses where patients suffered from more than mild nausea (score 2 to 3) were greater when the patient had been given ondansetron rather than granisetron. When treated with ondansetron as the antiemetic 35 per cent (95 per cent CI 20.6–51.7), patients suffered from moderate to severe nausea. When treated with granisetron 22 per cent (95 per cent CI 11.8–36.6) of patients suffered from moderate to severe nausea.
Figure 2 shows the mean number of episodes of vomiting on days 1 to 7 for the group as a whole when patients were given either ondansetron of granisetron prior to chemotherapy. Patients given granisetron suffered a weekly mean of 4.48 episodes of vomiting compared with those treated with ondansetron who suffered a weekly mean of 7.39. On most days, patients given ondansetron suffered more episodes of vomiting than those given granisetron. The biggest difference was on day 2.
When prescribed ondansetron as the antiemetic, 50 per cent (95 per cent CI 33.8– 66.2) of patients suffered from two or more episodes of vomiting on day 2 whereas 18 per cent (95 per cent CI 8.8–32) of patients given granisetron as the antiemetic suffered from two or more episodes of vomiting on day 2. The criterion for successful prevention of nausea and vomiting for ondansetron and granisetron would be that patients would suffer no more than mild nausea on day 1.
The mean nausea and vomiting scores for each of the days of the week post chemotherapy for courses 1+2, 3+4, and 5+6 were compared. They showed that the scores varied but did not seem to be affected by the number of previous courses.
Figure 1: Mean nausea score on days 1 to 7 post chemotherapy in patients given ondansetron ( n ) or granisetron ( n ) |
Figure 2: Mean number of vomiting episodes on days 1 to 7 post chemotherapy in patients given ondansetron ( n ) or granisetron ( n ) |
The incidence of acute emesis and nausea in the first 24 hours after chemotherapy is greater than 90 per cent in patients receiving highly emetogenic chemotherapy such as cisplatin, unless treated with antiemetic agents. Good control of these symptoms following chemotherapy is an important prognostic factor for later control of delayed emesis and nausea and of emesis and nausea experienced in subsequent cycles of chemotherapy. It is, therefore, important that emesis and nausea should be well controlled on the first cycle of chemotherapy, as this may have a significant effect on a patient's quality of life and willingness to complete the course of treatment. Furthermore, uncontrolled emesis frequently results in patients whose nutritional status is poor. This study has demonstrated the efficacy of both IV ondansetron and graniseton in combination with IV dexamethasone on day 1 and oral dexamethasone on days 2 to 4 in controlling both acute and delayed nausea and vomiting. The highest incidence of nausea and vomiting occurred on day 2, when granisetron was significantly better than ondansetron in controlling vomiting. The crossover design of this study ensured that any period effect or effect of age, sex, or condition being treated were taken into account, because patients acted as their own controls. The Italian Group for Antiemetic Research carried out a study comparing intravenous ondansetron 8mg versus granisetron 3mg in prevention of cisplatin induced acute and delayed nausea and vomiting in 973 patients. The group concluded from that study that ondansetron and granisetron showed similar efficacy and tolerability.10 Dexamethasone was added to both treatments, 20mg by IV infusion 45 minutes before the cisplatin and 8mg IM twice a day on days 2 and 3 and 4mg twice a day on day 4. The use of oral dexamethasone for the prevention of delayed emesis was not investigated in this study, which only looked at one course of treatment in each patient.
In our study dexamethasone was given intravenously at a dose of 8mg on day 1 and 4mg three times a day orally on days 2 to 4. Each patient acted as his or her own control and all patients received at least one course each of ondansetron and granisetron. The total number of courses was typical for patients receiving this type of chemotherapy.
The difference in outcome in our study compared with the Italian work is likely to reflect differences in the steroid dose which masked the differing efficacy of the 5HT3 antagonists.
Granisetron is more effective than ondansetron when used in combination with dexamethasone in the prevention of acute and delayed vomiting caused by highly emetogenic chemotherapy. It is significantly better on day 2 when the majority of patients suffer the most. With the trend towards day-case chemotherapy, it is more desirable to use effective once daily control. Day 2 is usually the day when patients go home and the medical staff may be unaware of the problems on this day unless there is close questioning of patients on their return. Patients receiving highly emetogenic chemotherapy should receive a combination of granisetron injection and intravenous and oral dexamethasone as defined in the protocol to prevent unnecessary suffering. It would be useful to repeat this study in the same design on a larger number of patients.
Ms Stewart is clinical economy pharmacist, Dr Crawford is consultant oncologist and Professor Taylor is director of pharmacy and clinical professions at Airedale general hospital. Correspondence to Ms Stewart at Pharmacy Department, Airedale General Hospital, Keighley, West Yorkshire BD20 6TD.