Should 5-Hydroxytryptamine-3 Receptor Antagonists Be Administered Beyond 24 Hours After Chemotherapy to Prevent Delayed Emesis Systematic Re
http://www.100md.com
《临床肿瘤学》
the Department of Public Health Sciences and Epidemiology, John A. Burns School of Medicine, University of Hawaii at Manoa, Honolulu, HI
Department of Clinical Pharmacology, Medical University of Vienna, Vienna, Austria
ABSTRACT
METHODS: This analysis is based on the Cancer Care Ontario Practice Guidelines Initiative meta-analysis of the efficacy of 5-HT3 antagonists. Results from the clinical trials covered in that meta-analysis were reanalyzed to provide estimates of absolute risk reductions (ARR) and numbers needed to treat (NNT) for 5-HT3 antagonists, as monotherapy or as adjunct treatment. Numbers of 5-HT3 antagonist unit doses per successfully treated patient were also calculated.
RESULTS: Five studies (comprising 1,716 assessable patients) compared a 5-HT3 antagonist with placebo; five studies (2,240 patients) compared a combination of a 5-HT3 antagonist and dexamethasone with dexamethasone monotherapy. ARR for monotherapy was only 8.2% (95% CI, 3.0% to 13.4%). On average, 74 5-HT3 antagonist doses must be administered to 12 patients (NNT, 12.2; 95% CI, 7.5 to 33.4) not receiving dexamethasone to protect one patient from delayed emesis. In those patients receiving dexamethasone as standard antiemetic treatment in the delayed phase, the addition of a 5-HT3 antagonist did not significantly improve control of delayed emesis as compared with dexamethasone monotherapy (ARR, 2.6%; 95% CI, –0.6% to 5.8%).
CONCLUSION: Neither clinical evidence nor considerations of cost effectiveness justify using 5-HT3 antagonists beyond 24 hours after chemotherapy for prevention of delayed emesis.
INTRODUCTION
Despite routine use in oncology practice, however, questions remain about the optimal use of these drugs. One area of controversy surrounds the question whether there is benefit in continuing administration of 5-HT3 antagonists beyond the first 24 hours after moderately to highly emetogenic chemotherapy, in the so-called delayed phase of CINV. This is an important issue, especially in light of frequently raised concerns about the cost of 5-HT3 antagonists.
The Cancer Care Ontario Practice Guidelines Initiative (CCOPGI) recently addressed this question.4 On the basis of a systematic review of available clinical trials results, they concluded that, "when 5-HT3 receptor antagonists are administered for more than 24 hours, the results of a meta-analysis indicate a small (4.1%) decrease in the absolute proportion of patients with delayed-onset emesis," and "the expense of these agents, the small benefit demonstrated in the meta-analysis, and the effectiveness of dexamethasone and possibly dopamine receptor antagonists, suggest that the first-line approach be limited to the initial 24 hours after chemotherapy."4
The evidence-based, systematic review performed by the CCOPGI provided a major contribution to clarifying the role of 5-HT3 antagonists in the prevention of delayed emesis. However, the CCOPGI synthesized clinical trials results without distinguishing between 5-HT3 antagonist monotherapy and combination therapy with a corticosteroid (dexamethasone). Thus, their assessment of the efficacy of 5-HT3 antagonists beyond 24 hours after chemotherapy is difficult to interpret and use in guiding clinical practice. In addition, although the CCOPGI alluded to the expense of these agents, it stopped short of relating the clinical efficacy results to drug acquisition cost.
In this article, we build on the most recent update of the CCOPGI systematic review (January 2003). We reanalyze results from the clinical trials covered in that meta-analysis to provide separate estimates of clinical efficacy of 5-HT3 antagonists administered beyond 24 hours, either as monotherapy, or as adjunct treatment to dexamethasone regimens (combination therapy). Subsequently, we compute the numbers of 5-HT3 antagonist doses per successfully treated patient (ie, per patient protected from delayed emesis by a 5-HT3 antagonist administered as monotherapy and as adjunct to dexamethasone).
These estimates will enable decision makers from different health care systems to compute drug acquisition cost per patient protected from delayed emesis by a 5-HT3 antagonist, and facilitate evidence-based decision making about antiemetic treatment choices.
METHODS
Definition of Terms and Clinical End Point
The present report focuses on the use of 5-HT3 antagonists for prevention of delayed emesis induced by moderately to highly emetogenic chemotherapy as defined in the CCOPGI, ASCO, and Multinational Association of Supportive Care in Cancer antiemetic guidelines.4,1,15
Antiemetic studies report results using various end points. We followed the CCOPGI approach and report the proportions of patients with complete control of delayed emesis. Delayed emesis is conventionally defined as at least one episode of vomiting that occurs more than 24 hours after the beginning of chemotherapy administration.1 This end point was recorded in all studies. Emesis (or vomiting) is considered to be a relevant CINV-related end point,1 and is used by the ASCO and MASCC antiemetic practice guidelines. In accordance with the CCOPGI meta-analysis methodology, we excluded nausea-related end points from the meta-analysis because individual trials reported different measures of nausea control.
Synthesizing the Evidence
For each study, we calculated the absolute risk reduction (ARR) and the number needed to treat (NNT).16 The NNT indicates the number of patients who need to receive a 5-HT3 antagonist beyond the first 24 hours after chemotherapy to prevent delayed emesis in one patient. The NNT is treatment, context, and comparator specific. All NNT comparisons used here are against placebo. This is also true for the combination-therapy trials; patients were randomly assigned to receive either dexamethasone plus a 5-HT3 antagonist treatment or dexamethasone plus placebo treatment (except for one trial that placed drugs into identical capsules to ensure that the oral treatment could not be identified14). The NNT with its 95% CI summarizes the benefit of an active treatment over control in a way that is meaningful for clinical decision making. It conveys both statistical and clinical significance to the prescriber and other health care decision makers.
To synthesize results from the individual studies, we computed the weighted, pooled ARR and NNT, with corresponding 95% CIs; this was done separately for the studies assessing 5-HT3 antagonist monotherapy and those that reported on 5-HT3 antagonist combination therapy (as an adjunct to dexamethasone). As in the CCOPGI meta-analysis, the fixed effects model was used for the meta-analysis because there were too few studies to estimate random effects.
Analysis of Drug Acquisition Cost
The NNT allows us to relate clinical efficacy results to drug acquisition cost because NNT can easily be converted into the number of drug doses (and thus drug cost that budget holders are expected to incur) required to prevent delayed emesis in one patient.
On the basis of the 5-HT3 antagonist regimens described in the trial reports, we computed for each clinical trial the number of 5-HT3 antagonist unit doses administered per patient beyond 24 hours after chemotherapy, per cycle, either as monotherapy or as an adjunct to dexamethasone. We define unit doses as the standard recommended single oral doses for ondansetron (8 mg), granisetron (1 mg), tropisetron (5 mg), and dolasetron (100 mg).1 We then multiplied the number of unit doses by the NNT from the same trial, to obtain the numbers of 5-HT3 antagonist unit doses a clinician or hospital needs to dispense, per one chemotherapy cycle, to protect one patient from delayed emesis. Finally, the mean number of unit doses was calculated, weighted by the number of assessable patients in the individual clinical trials.
To illustrate drug cost implications for at least one health care environment, we used a sample of US prices for the recommended dose of 8 mg oral ondansetron, and computed drug acquisition cost, per cycle, per patient treated, and per patient protected from delayed emesis, in the monotherapy or combination-therapy setting.
RESULTS
5-HT3 Receptor Antagonists as Monotherapy Beyond 24 Hours
Results from the five randomized trials are summarized in Table 1. 5-9 Only one study6 reported significant superiority of ondansetron over placebo in the control of delayed emesis. The remaining four trials failed to detect significant differences between patient groups receiving ondansetron or placebo.5,7-9 No other 5-HT3 antagonists were studied in this setting.
The meta-analysis synthesizing evidence from these studies shows that administering a 5-HT3 antagonist (ondansetron) beyond 24 hours results in a statistically significant ARR of 8.2% (95% CI, 3.0% to 13.4%). This suggests that 12 patients (NNT, 12.2; 95% CI, 7.5 to 33.4) have to be treated with ondansetron for more than 24 hours after chemotherapy to protect one patient from delayed emesis.
The rate of complete protection from delayed emesis obtained by pooling results from the clinical trials, and the ondansetron regimens used in these trials, suggest that a hospital needs to dispense, on average, 74.4 5-HT3 antagonist unit doses (95% CI, 45.7 to 203.7) per cycle to achieve protection from delayed emesis in one patient (Table 2). Assuming the price of $30.45 per tablet of oral ondansetron 8 mg (source, www.drugstore.com, as of June 17, 2004), this would translate into a drug acquisition cost per patient protected from delayed emesis of US $2,265 per cycle (95% CI, US $1,392 to US $6,203; Table 2).
5-HT3 Receptor Antagonists as Adjunct to Dexamethasone Beyond 24 Hours
The CCOPGI systematic review identified five studies that addressed this question, which were all published since 1997 (Table 1).10-14 These included two large-scale studies conducted by the National Cancer Institute of Canada Clinical Trials Group13,10 and a large-scale study by the Italian Group for Antiemetic Research.14 Each of these trials individually led to the conclusion that adding a 5-HT3 antagonist to dexamethasone beyond the first 24 hours after chemotherapy confers no significant benefit in the control of delayed emesis.
Likewise, our meta-analysis synthesizing these five studies failed to detect a significant difference in the control of delayed emesis between a combination of a 5-HT3 antagonist plus dexamethasone and dexamethasone monotherapy (ARR, 2.6%; 95% CI, –0.6% to 5.8%; NNT, 38.8; 95% CI, 17.3 to {infty}; P = not significant).
We used the meta-analysis results to relate the pooled estimate of clinical efficacy (or lack thereof) to the number of 5-HT3 antagonist doses and thus the 5-HT3 antagonist drug acquisition cost per patient protected from delayed emesis. We did so for illustrative purposes only; we were aware that 5-HT3 antagonists confer no significant benefit in prevention of delayed emesis when added to dexamethasone. Assuming the pooled rate of complete protection from delayed emesis (ignoring, for the moment, the lack of significance), and on the basis of the 5-HT3 antagonist regimens used in these trials, hospitals would need to dispense 422.9 doses of 5-HT3 antagonists (95% CI, 188.6 to {infty}), per cycle to protect one patient from delayed emesis. Assuming the price of $30.45 per tablet of oral ondansetron 8 mg (source, www.drugstore.com, as of June 17, 2004), this would translate into drug acquisition cost of 5-HT3 antagonists per patient protected from delayed emesis of US $12,877 per cycle (95% CI, US $5,743 to {infty}).
DISCUSSION
In the monotherapy setting, pooled results, comprising data from 1,716 assessable patients, show that prolonged administration of 5-HT3 antagonists is associated with a statistically significant reduction in the rate of delayed emesis. However, the difference is small and implies that about 12 patients have to be treated with a total of 74 5-HT3 antagonist doses per cycle to achieve protection from delayed emesis in one patient.
Although the effect is statistically significant, it must be emphasized that clinical trials have consistently shown superiority of administering a 5-HT3 antagonist plus dexamethasone combination over a 5-HT3 antagonist alone for control of delayed emesis.17 Moreover, studies of oral dexamethasone monotherapy for control of delayed emesis, conducted both before and after the availability of 5-HT3 antagonists, have shown efficacy of dexamethasone alone. In a study of patients receiving cisplatin, only 65% of patients receiving dexamethasone alone experienced delayed emesis, compared with 89% of patients receiving placebo.18 In a study of patients with moderately emetogenic chemotherapy, 57% had complete control and 30% had major control of delayed emesis among patients given maintenance oral dexamethasone, versus 33% and 15%, respectively, of patients who did not receive maintenance dexamethasone.19 Although prolonged use of dexamethasone presents some concern in clinical practice regarding potential side effects such as insomnia and gastrointestinal intolerance, particularly in the presence of comorbidities and polypharmacy,20 the overall risk-benefit ratio seems to be favorable,1,17 and wider use of dexamethasone in this setting has been advocated.1,21 Hence, our finding of a small, significant benefit of 5-HT3 antagonists relative to placebo has little relevance because available evidence does not support 5-HT3 antagonist monotherapy in this setting.
Moreover, our assessment of the efficacy of 5-HT3 antagonists in this setting may overestimate the true clinical benefit of these agents. First, we focused on control of delayed emesis as the clinical efficacy end point, excluding delayed nausea. However, nausea affects more chemotherapy-treated patients than emesis,1 and 5-HT3 antagonists were found to reduce nausea to a lesser extent than vomiting.22 Hence, for a combined end point of complete control of nausea and vomiting, which may be more relevant from a patient's perspective, 5-HT3 antagonists would be expected to be even less effective than for the end point used in this analysis. Second, continuing administration of 5-HT3 antagonists for more than 24 hours results in an increased incidence of adverse effects.7,11,14 Although not life threatening, these may further reduce the net benefit accrued from treatment with these agents.
With regard to using 5-HT3 antagonists as adjunct to dexamethasone beyond 24 hours after chemotherapy, none of the studies comparing a 5-HT3 antagonist plus dexamethasone combination with dexamethasone monotherapy found a significant improvement in the control of delayed emesis. Pooling results from these studies with a total of 2,240 assessable patients still failed to detect a significant improvement in control of delayed emesis. An additional randomized, non-placebo-controlled, open study of 149 patients comparing ondansetron plus dexamethasone with dexamethasone alone also failed to detect any significant improvement in the control of delayed emesis associated with adding ondansetron to dexamethasone in the delayed phase.23 In aggregate, these findings strongly argue against the use of 5-HT3 antagonists in this setting. Given the consistency of trial results and the total number of patients enrolled onto the trials, we would also argue that no additional research is needed to support this position.
In the absence of a statistically significant difference, and following strict methodology, clinical benefit cannot be quantified by way of NNT, and NNT figures are usually not calculated. However, several studies of daily clinical practice have shown that 5-HT3 antagonists, either alone or in combination with dexamethasone, are used for prevention of delayed emesis,2,3 and some antiemetic guidelines continue to recommend administering a 5-HT3 antagonist plus dexamethasone combination in the delayed phase.1,15 This may be motivated by the assumption that more studies might show significance, and patients should not be denied potentially beneficial treatment. To put these assumptions into perspective, we present the current best estimate of an NNT (with 95% CI; Table 1), emphasizing that this estimate is not based on a significant difference. The NNT figure of 39 illustrates that any potential benefit conferred by 5-HT3 antagonists in this setting is likely to be clinically irrelevant, even if it were shown to be significant.
In the face of limited resources, health care providers best serve their patients by making treatment choices based on evidence of clinical effectiveness and cost effectiveness. We therefore related clinical efficacy estimates to the drug acquisition cost of using 5-HT3 antagonists in the delayed phase.
Considerations of cost effectiveness dictate that costs are not calculated as per patient treated but per patient treated successfully. Hence, the point estimates of the number of unit doses required to protect one patient from delayed emesis in the monotherapy setting (74 unit doses per cycle; Table 2) or the combination-therapy setting (423 unit doses per cycle, not based on significant benefit) seem more relevant. Health care decision makers may find these numbers useful to calculate true drug acquisition costs based on local unit drug prices. The considerable cost of this treatment is illustrated by the examples based on a sample of US prices (Table 2).
In conclusion, our results show that neither clinical evidence nor considerations of cost effectiveness justify the use of 5-HT3 antagonists beyond 24 hours after chemotherapy for prevention of delayed emesis.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Gralla RJ, Osoba D, Kris MG, et al: ASCO: Recommendations for use of antiemetics—Evidence-based clinical practice guidelines. J Clin Oncol 17:2971-2994, 1999
2. Italian Group for Antiemetic Research: Transferability to clinical practice of the results of controlled clinical trials: The case of antiemetic prophylactic treatment for cancer chemotherapy-induced nausea and vomiting. Ann Oncol 9:759-765, 1998
3. Grunberg SM, Deuson RR, Mavros P, et al: Incidence of chemotherapy-induced nausea and emesis after modern antiemetics: Perception versus reality. Cancer 100:2261-2268, 2004
4. Cancer Care Ontario Practice Guideline Initiative: Use of 5-HT3 receptor antagonists in patients receiving moderately or highly emetogenic chemotherapy. Practice Guideline Report 12-3. http://www.cancercare.on.ca/pdf/full12_3.pdf
5. Gandara DR, Harvey WH, Monaghan GG, et al: The delayed-emesis syndrome from cisplatin: Phase III evaluation of ondansetron versus placebo. Semin Oncol 19:67-71, 1992 (suppl 10)
6. Kaizer L, Warr D, Hoskins P, et al: Effect of schedule and maintenance on the antiemetic efficacy of ondansetron combined with dexamethasone in acute and delayed nausea and emesis in patients receiving moderately emetogenic chemotherapy: A phase III trial by the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 12:1050-1057, 1994
7. Navari RM, Madajewicz S, Anderson N, et al: Oral ondansetron for the control of cisplatin-induced delayed emesis: A large, multicenter, double-blind, randomized comparative trial of ondansetron versus placebo. J Clin Oncol 13:2408-2416, 1995
8. Stewart A, McQuade B, Cronje JD, et al: Ondansetron compared with granisetron in the prophylaxis of cyclophosphamide-induced emesis in out-patients: A multicenter, double-blind, double-dummy, randomized, parallel-group study. Oncology 52:202-210, 1995
9. Olver I, Paska W, Depierre A, et al: A multicenter, double-blind study comparing placebo, ondansetron and ondansetron plus dexamethasone for the control of cisplatin-induced delayed emesis. Ann Oncol 7:945-952, 1996
10. Pater JL, Lofters WS, Zee B, et al: The role of the 5-HT3 antagonists ondansetron and dolasetron in the control of delayed onset nausea and vomiting in patients receiving moderately emetogenic chemotherapy. Ann Oncol 8:181-185, 1997
11. Sorbe BG, Berglind AM, Andersson H, et al: A study evaluating the efficacy and tolerability of tropisetron in combination with dexamethasone in the prevention of delayed platinum-induced nausea and emesis. Cancer 83:1022-1032, 1998
12. Goedhals L, Heron JF, Kleisbauer JP, et al: Control of delayed nausea and vomiting with granisetron plus dexamethasone or dexamethasone alone in patients receiving highly emetogenic chemotherapy: A double-blind, placebo-controlled, comparative study. Ann Oncol 9:661-666, 1998
13. Latreille J, Pater J, Johnston D, et al: Use of dexamethasone and granisetron in the control of delayed emesis for patients who receive highly emetogenic chemotherapy: National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 16:1174-1178, 1998
14. Italian Group for Antiemetic Research: Dexamethasone alone or in combination with ondansetron for the prevention of delayed nausea and vomiting induced by chemotherapy. N Engl J Med 342:1554-1559, 2000
15. MASCC: Prevention of chemotherapy- and radiology-induced emesis: Results of the Perugia Consensus Conference. Ann Oncol 6:215-220, 1998
16. Cook RJ, Sackett DL: The number needed to treat a clinically useful measure of treatment effect. BMJ 310:452-454, 1995
17. Ioannidis JPA, Hesketh PJ, Lau J: Contribution of dexamethasone to control of chemotherapy-induced nausea and vomiting: A meta-analysis of randomized evidence. J Clin Oncol 18:3409-3422, 2000
18. Kris MG, Grall RJ, Tyson LB, et al: Controlling delayed vomiting: Double-blind, randomized trial comparing placebo, dexamethasone alone, and metoclopramide plus dexamethasone in patients receiving cisplatin. J Clin Oncol 7:108-114, 1989
19. Koo WH, Ang PT: Role of maintenance dexamethasone in prophylaxis of delayed emesis caused by moderately emetogenic chemotherapy. Ann Oncol 7:71-74, 1996
20. Gridelli C, Maione P, Rossi A: Corticosteroids underemployment in delayed-induced nausea and emesis with poor adherence to American Society of Clinical Oncology Guidelines: Is this a reasonable clinical choice for the elderly J Clin Oncol 21:4066-4069, 2003
21. Mertens WC, Higby DJ, Brown D, et al: Improving the care of patients with regard to chemotherapy-induced nausea and emesis: The effect of feedback to clinicians on adherence to antiemetic prescribing guidelines. J Clin Oncol 21:1373-1378, 2003
22. Roscoe JA, Morrow GR, Hickok JT, et al: Nausea and vomiting remain a significant clinical problem: Trends over time in controlling chemotherapy-induced nausea and vomiting in 1413 patients treated in community clinical practices. J Pain Symptom Manage 20:113-121, 2000
23. Tsukada H, Hirose T, Yokoyama A, et al: Randomized comparison of ondansetron plus dexamethasone with dexamethasone alone for the control of delayed cisplatin-induced emesis. Eur J Cancer 37:2398-2404, 2001(Olga Geling, Hans-Georg E)
Department of Clinical Pharmacology, Medical University of Vienna, Vienna, Austria
ABSTRACT
METHODS: This analysis is based on the Cancer Care Ontario Practice Guidelines Initiative meta-analysis of the efficacy of 5-HT3 antagonists. Results from the clinical trials covered in that meta-analysis were reanalyzed to provide estimates of absolute risk reductions (ARR) and numbers needed to treat (NNT) for 5-HT3 antagonists, as monotherapy or as adjunct treatment. Numbers of 5-HT3 antagonist unit doses per successfully treated patient were also calculated.
RESULTS: Five studies (comprising 1,716 assessable patients) compared a 5-HT3 antagonist with placebo; five studies (2,240 patients) compared a combination of a 5-HT3 antagonist and dexamethasone with dexamethasone monotherapy. ARR for monotherapy was only 8.2% (95% CI, 3.0% to 13.4%). On average, 74 5-HT3 antagonist doses must be administered to 12 patients (NNT, 12.2; 95% CI, 7.5 to 33.4) not receiving dexamethasone to protect one patient from delayed emesis. In those patients receiving dexamethasone as standard antiemetic treatment in the delayed phase, the addition of a 5-HT3 antagonist did not significantly improve control of delayed emesis as compared with dexamethasone monotherapy (ARR, 2.6%; 95% CI, –0.6% to 5.8%).
CONCLUSION: Neither clinical evidence nor considerations of cost effectiveness justify using 5-HT3 antagonists beyond 24 hours after chemotherapy for prevention of delayed emesis.
INTRODUCTION
Despite routine use in oncology practice, however, questions remain about the optimal use of these drugs. One area of controversy surrounds the question whether there is benefit in continuing administration of 5-HT3 antagonists beyond the first 24 hours after moderately to highly emetogenic chemotherapy, in the so-called delayed phase of CINV. This is an important issue, especially in light of frequently raised concerns about the cost of 5-HT3 antagonists.
The Cancer Care Ontario Practice Guidelines Initiative (CCOPGI) recently addressed this question.4 On the basis of a systematic review of available clinical trials results, they concluded that, "when 5-HT3 receptor antagonists are administered for more than 24 hours, the results of a meta-analysis indicate a small (4.1%) decrease in the absolute proportion of patients with delayed-onset emesis," and "the expense of these agents, the small benefit demonstrated in the meta-analysis, and the effectiveness of dexamethasone and possibly dopamine receptor antagonists, suggest that the first-line approach be limited to the initial 24 hours after chemotherapy."4
The evidence-based, systematic review performed by the CCOPGI provided a major contribution to clarifying the role of 5-HT3 antagonists in the prevention of delayed emesis. However, the CCOPGI synthesized clinical trials results without distinguishing between 5-HT3 antagonist monotherapy and combination therapy with a corticosteroid (dexamethasone). Thus, their assessment of the efficacy of 5-HT3 antagonists beyond 24 hours after chemotherapy is difficult to interpret and use in guiding clinical practice. In addition, although the CCOPGI alluded to the expense of these agents, it stopped short of relating the clinical efficacy results to drug acquisition cost.
In this article, we build on the most recent update of the CCOPGI systematic review (January 2003). We reanalyze results from the clinical trials covered in that meta-analysis to provide separate estimates of clinical efficacy of 5-HT3 antagonists administered beyond 24 hours, either as monotherapy, or as adjunct treatment to dexamethasone regimens (combination therapy). Subsequently, we compute the numbers of 5-HT3 antagonist doses per successfully treated patient (ie, per patient protected from delayed emesis by a 5-HT3 antagonist administered as monotherapy and as adjunct to dexamethasone).
These estimates will enable decision makers from different health care systems to compute drug acquisition cost per patient protected from delayed emesis by a 5-HT3 antagonist, and facilitate evidence-based decision making about antiemetic treatment choices.
METHODS
Definition of Terms and Clinical End Point
The present report focuses on the use of 5-HT3 antagonists for prevention of delayed emesis induced by moderately to highly emetogenic chemotherapy as defined in the CCOPGI, ASCO, and Multinational Association of Supportive Care in Cancer antiemetic guidelines.4,1,15
Antiemetic studies report results using various end points. We followed the CCOPGI approach and report the proportions of patients with complete control of delayed emesis. Delayed emesis is conventionally defined as at least one episode of vomiting that occurs more than 24 hours after the beginning of chemotherapy administration.1 This end point was recorded in all studies. Emesis (or vomiting) is considered to be a relevant CINV-related end point,1 and is used by the ASCO and MASCC antiemetic practice guidelines. In accordance with the CCOPGI meta-analysis methodology, we excluded nausea-related end points from the meta-analysis because individual trials reported different measures of nausea control.
Synthesizing the Evidence
For each study, we calculated the absolute risk reduction (ARR) and the number needed to treat (NNT).16 The NNT indicates the number of patients who need to receive a 5-HT3 antagonist beyond the first 24 hours after chemotherapy to prevent delayed emesis in one patient. The NNT is treatment, context, and comparator specific. All NNT comparisons used here are against placebo. This is also true for the combination-therapy trials; patients were randomly assigned to receive either dexamethasone plus a 5-HT3 antagonist treatment or dexamethasone plus placebo treatment (except for one trial that placed drugs into identical capsules to ensure that the oral treatment could not be identified14). The NNT with its 95% CI summarizes the benefit of an active treatment over control in a way that is meaningful for clinical decision making. It conveys both statistical and clinical significance to the prescriber and other health care decision makers.
To synthesize results from the individual studies, we computed the weighted, pooled ARR and NNT, with corresponding 95% CIs; this was done separately for the studies assessing 5-HT3 antagonist monotherapy and those that reported on 5-HT3 antagonist combination therapy (as an adjunct to dexamethasone). As in the CCOPGI meta-analysis, the fixed effects model was used for the meta-analysis because there were too few studies to estimate random effects.
Analysis of Drug Acquisition Cost
The NNT allows us to relate clinical efficacy results to drug acquisition cost because NNT can easily be converted into the number of drug doses (and thus drug cost that budget holders are expected to incur) required to prevent delayed emesis in one patient.
On the basis of the 5-HT3 antagonist regimens described in the trial reports, we computed for each clinical trial the number of 5-HT3 antagonist unit doses administered per patient beyond 24 hours after chemotherapy, per cycle, either as monotherapy or as an adjunct to dexamethasone. We define unit doses as the standard recommended single oral doses for ondansetron (8 mg), granisetron (1 mg), tropisetron (5 mg), and dolasetron (100 mg).1 We then multiplied the number of unit doses by the NNT from the same trial, to obtain the numbers of 5-HT3 antagonist unit doses a clinician or hospital needs to dispense, per one chemotherapy cycle, to protect one patient from delayed emesis. Finally, the mean number of unit doses was calculated, weighted by the number of assessable patients in the individual clinical trials.
To illustrate drug cost implications for at least one health care environment, we used a sample of US prices for the recommended dose of 8 mg oral ondansetron, and computed drug acquisition cost, per cycle, per patient treated, and per patient protected from delayed emesis, in the monotherapy or combination-therapy setting.
RESULTS
5-HT3 Receptor Antagonists as Monotherapy Beyond 24 Hours
Results from the five randomized trials are summarized in Table 1. 5-9 Only one study6 reported significant superiority of ondansetron over placebo in the control of delayed emesis. The remaining four trials failed to detect significant differences between patient groups receiving ondansetron or placebo.5,7-9 No other 5-HT3 antagonists were studied in this setting.
The meta-analysis synthesizing evidence from these studies shows that administering a 5-HT3 antagonist (ondansetron) beyond 24 hours results in a statistically significant ARR of 8.2% (95% CI, 3.0% to 13.4%). This suggests that 12 patients (NNT, 12.2; 95% CI, 7.5 to 33.4) have to be treated with ondansetron for more than 24 hours after chemotherapy to protect one patient from delayed emesis.
The rate of complete protection from delayed emesis obtained by pooling results from the clinical trials, and the ondansetron regimens used in these trials, suggest that a hospital needs to dispense, on average, 74.4 5-HT3 antagonist unit doses (95% CI, 45.7 to 203.7) per cycle to achieve protection from delayed emesis in one patient (Table 2). Assuming the price of $30.45 per tablet of oral ondansetron 8 mg (source, www.drugstore.com, as of June 17, 2004), this would translate into a drug acquisition cost per patient protected from delayed emesis of US $2,265 per cycle (95% CI, US $1,392 to US $6,203; Table 2).
5-HT3 Receptor Antagonists as Adjunct to Dexamethasone Beyond 24 Hours
The CCOPGI systematic review identified five studies that addressed this question, which were all published since 1997 (Table 1).10-14 These included two large-scale studies conducted by the National Cancer Institute of Canada Clinical Trials Group13,10 and a large-scale study by the Italian Group for Antiemetic Research.14 Each of these trials individually led to the conclusion that adding a 5-HT3 antagonist to dexamethasone beyond the first 24 hours after chemotherapy confers no significant benefit in the control of delayed emesis.
Likewise, our meta-analysis synthesizing these five studies failed to detect a significant difference in the control of delayed emesis between a combination of a 5-HT3 antagonist plus dexamethasone and dexamethasone monotherapy (ARR, 2.6%; 95% CI, –0.6% to 5.8%; NNT, 38.8; 95% CI, 17.3 to {infty}; P = not significant).
We used the meta-analysis results to relate the pooled estimate of clinical efficacy (or lack thereof) to the number of 5-HT3 antagonist doses and thus the 5-HT3 antagonist drug acquisition cost per patient protected from delayed emesis. We did so for illustrative purposes only; we were aware that 5-HT3 antagonists confer no significant benefit in prevention of delayed emesis when added to dexamethasone. Assuming the pooled rate of complete protection from delayed emesis (ignoring, for the moment, the lack of significance), and on the basis of the 5-HT3 antagonist regimens used in these trials, hospitals would need to dispense 422.9 doses of 5-HT3 antagonists (95% CI, 188.6 to {infty}), per cycle to protect one patient from delayed emesis. Assuming the price of $30.45 per tablet of oral ondansetron 8 mg (source, www.drugstore.com, as of June 17, 2004), this would translate into drug acquisition cost of 5-HT3 antagonists per patient protected from delayed emesis of US $12,877 per cycle (95% CI, US $5,743 to {infty}).
DISCUSSION
In the monotherapy setting, pooled results, comprising data from 1,716 assessable patients, show that prolonged administration of 5-HT3 antagonists is associated with a statistically significant reduction in the rate of delayed emesis. However, the difference is small and implies that about 12 patients have to be treated with a total of 74 5-HT3 antagonist doses per cycle to achieve protection from delayed emesis in one patient.
Although the effect is statistically significant, it must be emphasized that clinical trials have consistently shown superiority of administering a 5-HT3 antagonist plus dexamethasone combination over a 5-HT3 antagonist alone for control of delayed emesis.17 Moreover, studies of oral dexamethasone monotherapy for control of delayed emesis, conducted both before and after the availability of 5-HT3 antagonists, have shown efficacy of dexamethasone alone. In a study of patients receiving cisplatin, only 65% of patients receiving dexamethasone alone experienced delayed emesis, compared with 89% of patients receiving placebo.18 In a study of patients with moderately emetogenic chemotherapy, 57% had complete control and 30% had major control of delayed emesis among patients given maintenance oral dexamethasone, versus 33% and 15%, respectively, of patients who did not receive maintenance dexamethasone.19 Although prolonged use of dexamethasone presents some concern in clinical practice regarding potential side effects such as insomnia and gastrointestinal intolerance, particularly in the presence of comorbidities and polypharmacy,20 the overall risk-benefit ratio seems to be favorable,1,17 and wider use of dexamethasone in this setting has been advocated.1,21 Hence, our finding of a small, significant benefit of 5-HT3 antagonists relative to placebo has little relevance because available evidence does not support 5-HT3 antagonist monotherapy in this setting.
Moreover, our assessment of the efficacy of 5-HT3 antagonists in this setting may overestimate the true clinical benefit of these agents. First, we focused on control of delayed emesis as the clinical efficacy end point, excluding delayed nausea. However, nausea affects more chemotherapy-treated patients than emesis,1 and 5-HT3 antagonists were found to reduce nausea to a lesser extent than vomiting.22 Hence, for a combined end point of complete control of nausea and vomiting, which may be more relevant from a patient's perspective, 5-HT3 antagonists would be expected to be even less effective than for the end point used in this analysis. Second, continuing administration of 5-HT3 antagonists for more than 24 hours results in an increased incidence of adverse effects.7,11,14 Although not life threatening, these may further reduce the net benefit accrued from treatment with these agents.
With regard to using 5-HT3 antagonists as adjunct to dexamethasone beyond 24 hours after chemotherapy, none of the studies comparing a 5-HT3 antagonist plus dexamethasone combination with dexamethasone monotherapy found a significant improvement in the control of delayed emesis. Pooling results from these studies with a total of 2,240 assessable patients still failed to detect a significant improvement in control of delayed emesis. An additional randomized, non-placebo-controlled, open study of 149 patients comparing ondansetron plus dexamethasone with dexamethasone alone also failed to detect any significant improvement in the control of delayed emesis associated with adding ondansetron to dexamethasone in the delayed phase.23 In aggregate, these findings strongly argue against the use of 5-HT3 antagonists in this setting. Given the consistency of trial results and the total number of patients enrolled onto the trials, we would also argue that no additional research is needed to support this position.
In the absence of a statistically significant difference, and following strict methodology, clinical benefit cannot be quantified by way of NNT, and NNT figures are usually not calculated. However, several studies of daily clinical practice have shown that 5-HT3 antagonists, either alone or in combination with dexamethasone, are used for prevention of delayed emesis,2,3 and some antiemetic guidelines continue to recommend administering a 5-HT3 antagonist plus dexamethasone combination in the delayed phase.1,15 This may be motivated by the assumption that more studies might show significance, and patients should not be denied potentially beneficial treatment. To put these assumptions into perspective, we present the current best estimate of an NNT (with 95% CI; Table 1), emphasizing that this estimate is not based on a significant difference. The NNT figure of 39 illustrates that any potential benefit conferred by 5-HT3 antagonists in this setting is likely to be clinically irrelevant, even if it were shown to be significant.
In the face of limited resources, health care providers best serve their patients by making treatment choices based on evidence of clinical effectiveness and cost effectiveness. We therefore related clinical efficacy estimates to the drug acquisition cost of using 5-HT3 antagonists in the delayed phase.
Considerations of cost effectiveness dictate that costs are not calculated as per patient treated but per patient treated successfully. Hence, the point estimates of the number of unit doses required to protect one patient from delayed emesis in the monotherapy setting (74 unit doses per cycle; Table 2) or the combination-therapy setting (423 unit doses per cycle, not based on significant benefit) seem more relevant. Health care decision makers may find these numbers useful to calculate true drug acquisition costs based on local unit drug prices. The considerable cost of this treatment is illustrated by the examples based on a sample of US prices (Table 2).
In conclusion, our results show that neither clinical evidence nor considerations of cost effectiveness justify the use of 5-HT3 antagonists beyond 24 hours after chemotherapy for prevention of delayed emesis.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Gralla RJ, Osoba D, Kris MG, et al: ASCO: Recommendations for use of antiemetics—Evidence-based clinical practice guidelines. J Clin Oncol 17:2971-2994, 1999
2. Italian Group for Antiemetic Research: Transferability to clinical practice of the results of controlled clinical trials: The case of antiemetic prophylactic treatment for cancer chemotherapy-induced nausea and vomiting. Ann Oncol 9:759-765, 1998
3. Grunberg SM, Deuson RR, Mavros P, et al: Incidence of chemotherapy-induced nausea and emesis after modern antiemetics: Perception versus reality. Cancer 100:2261-2268, 2004
4. Cancer Care Ontario Practice Guideline Initiative: Use of 5-HT3 receptor antagonists in patients receiving moderately or highly emetogenic chemotherapy. Practice Guideline Report 12-3. http://www.cancercare.on.ca/pdf/full12_3.pdf
5. Gandara DR, Harvey WH, Monaghan GG, et al: The delayed-emesis syndrome from cisplatin: Phase III evaluation of ondansetron versus placebo. Semin Oncol 19:67-71, 1992 (suppl 10)
6. Kaizer L, Warr D, Hoskins P, et al: Effect of schedule and maintenance on the antiemetic efficacy of ondansetron combined with dexamethasone in acute and delayed nausea and emesis in patients receiving moderately emetogenic chemotherapy: A phase III trial by the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 12:1050-1057, 1994
7. Navari RM, Madajewicz S, Anderson N, et al: Oral ondansetron for the control of cisplatin-induced delayed emesis: A large, multicenter, double-blind, randomized comparative trial of ondansetron versus placebo. J Clin Oncol 13:2408-2416, 1995
8. Stewart A, McQuade B, Cronje JD, et al: Ondansetron compared with granisetron in the prophylaxis of cyclophosphamide-induced emesis in out-patients: A multicenter, double-blind, double-dummy, randomized, parallel-group study. Oncology 52:202-210, 1995
9. Olver I, Paska W, Depierre A, et al: A multicenter, double-blind study comparing placebo, ondansetron and ondansetron plus dexamethasone for the control of cisplatin-induced delayed emesis. Ann Oncol 7:945-952, 1996
10. Pater JL, Lofters WS, Zee B, et al: The role of the 5-HT3 antagonists ondansetron and dolasetron in the control of delayed onset nausea and vomiting in patients receiving moderately emetogenic chemotherapy. Ann Oncol 8:181-185, 1997
11. Sorbe BG, Berglind AM, Andersson H, et al: A study evaluating the efficacy and tolerability of tropisetron in combination with dexamethasone in the prevention of delayed platinum-induced nausea and emesis. Cancer 83:1022-1032, 1998
12. Goedhals L, Heron JF, Kleisbauer JP, et al: Control of delayed nausea and vomiting with granisetron plus dexamethasone or dexamethasone alone in patients receiving highly emetogenic chemotherapy: A double-blind, placebo-controlled, comparative study. Ann Oncol 9:661-666, 1998
13. Latreille J, Pater J, Johnston D, et al: Use of dexamethasone and granisetron in the control of delayed emesis for patients who receive highly emetogenic chemotherapy: National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 16:1174-1178, 1998
14. Italian Group for Antiemetic Research: Dexamethasone alone or in combination with ondansetron for the prevention of delayed nausea and vomiting induced by chemotherapy. N Engl J Med 342:1554-1559, 2000
15. MASCC: Prevention of chemotherapy- and radiology-induced emesis: Results of the Perugia Consensus Conference. Ann Oncol 6:215-220, 1998
16. Cook RJ, Sackett DL: The number needed to treat a clinically useful measure of treatment effect. BMJ 310:452-454, 1995
17. Ioannidis JPA, Hesketh PJ, Lau J: Contribution of dexamethasone to control of chemotherapy-induced nausea and vomiting: A meta-analysis of randomized evidence. J Clin Oncol 18:3409-3422, 2000
18. Kris MG, Grall RJ, Tyson LB, et al: Controlling delayed vomiting: Double-blind, randomized trial comparing placebo, dexamethasone alone, and metoclopramide plus dexamethasone in patients receiving cisplatin. J Clin Oncol 7:108-114, 1989
19. Koo WH, Ang PT: Role of maintenance dexamethasone in prophylaxis of delayed emesis caused by moderately emetogenic chemotherapy. Ann Oncol 7:71-74, 1996
20. Gridelli C, Maione P, Rossi A: Corticosteroids underemployment in delayed-induced nausea and emesis with poor adherence to American Society of Clinical Oncology Guidelines: Is this a reasonable clinical choice for the elderly J Clin Oncol 21:4066-4069, 2003
21. Mertens WC, Higby DJ, Brown D, et al: Improving the care of patients with regard to chemotherapy-induced nausea and emesis: The effect of feedback to clinicians on adherence to antiemetic prescribing guidelines. J Clin Oncol 21:1373-1378, 2003
22. Roscoe JA, Morrow GR, Hickok JT, et al: Nausea and vomiting remain a significant clinical problem: Trends over time in controlling chemotherapy-induced nausea and vomiting in 1413 patients treated in community clinical practices. J Pain Symptom Manage 20:113-121, 2000
23. Tsukada H, Hirose T, Yokoyama A, et al: Randomized comparison of ondansetron plus dexamethasone with dexamethasone alone for the control of delayed cisplatin-induced emesis. Eur J Cancer 37:2398-2404, 2001(Olga Geling, Hans-Georg E)