5-hydroxytryptamine type-3 receptor antagonists for chemotherapy-induced and radiotherapy-induced nausea and emesis

Can we safely reduce the dose of administered agents?


  • Matti Aapro M.D.,

    Corresponding author
    1. Clinique de Genolier, Genolier, Switzerland
    • Clinique de Genolier, 1 Rte du Muids, CH-1272 Genolier, Switzerland
    Search for more papers by this author
    • Fax: (011) 44 223669131

    • Dr. Aapro has acted as a consultant for Roche Laboratories, Helsinn, Novartis Pharma, and Aventis, all of which are producers of hydroxytryptamine type-3 receptor antagonists.

  • Peter Blower Ph.D., D.Sc.

    1. Poole House, Great Yeldham, Essex, United Kingdom
    Search for more papers by this author
    • Dr. Blower has acted as a consultant to GlaxoSmithKline, Novartis Pharma, and Hoffman-La Roche Laboratories and is a stockholder for GlaxoSmithKline.



Nausea and emesis as a consequence of chemotherapy or radiotherapy can have an adverse effect on patients' quality of life during cancer treatment and may last for > 5 days after administration. Guidelines suggest that, used at appropriate doses, the 5-hydroxytryptamine type-3 (5-HT3) receptor antagonists—which are considered the antiemetic “gold standard” when they are administered in combination with corticosteroids—demonstrate equivalent efficacy and safety. However, due to financial considerations, these agents often are used at lower doses than recommended.


A literature review of relevant publications pertaining to the control of chemotherapy-induced nausea and emesis and dosing issues of the 5-HT3 receptor antagonists was undertaken to provide a comprehensive review of dosing issues relevant to the 5-HT3 receptor antagonists.


The issue of “down dosing” was particularly pertinent because of the nature of the 5-HT3 receptor antagonist dose-response curve: A steep dose-response profile within a narrow dose range suggests that antiemetic control will be lost suddenly after dose deescalation. However, the array of predisposing and confounding patient factors indicates that it is unlikely that a loss of antiemetic control will be apparent across a population; rather, individuals will experience loss of control as the dose is reduced below threshold. Of the 4 5-HT3 receptor antagonists currently licensed in the United States (granisetron, ondansetron, dolasetron, and palonosetron), ondansetron is used sometimes at lower than optimal doses, and there is evidence to suggest that even the approved oral dose of dolasetron may be suboptimal.


Suboptimal dosing not only will be detrimental to patients' quality of life but, ultimately, will prove counterproductive in terms of hospital resources, and it will add to the already significant socioeconomic burden associated with cancer therapy. Therefore, the dose of antiemetic agent administered should be sufficiently high to ensure good emesis control across the whole patient population. Cancer 2005. © 2005 American Cancer Society.

Nausea and emesis resulting from cytotoxic drug therapy or radiotherapy may cause considerable physical and emotional distress to cancer patients1 and may last for > 5 days after the trigger stimulus.2, 3 Chemotherapy-induced nausea and vomiting (CINV) and radiotherapy-induced nausea and vomiting (RINV) can be classified according to the time of onset and duration of symptoms. Symptoms that occur within the first 24 hours of chemotherapy or radiotherapy, defined classically as acute nausea and emesis, can be classified further into early-acute (within 12 hours) and late-acute (12–24 hours), whereas symptoms that occur > 24 hours after therapy are described as delayed emesis. A proportion of patients who experience inadequate control of nausea and emesis during treatment cycles (refractory or breakthrough emesis) are prone to developing symptoms of anticipatory emesis,4 which may incapacitate patients for up to 24 hours prior to treatment.

Antiemetic therapeutic guidelines, as compiled by a number of organizations and institutions, have set out to guide prescribing physicians when they are making important treatment decisions for each patient. The most recent of these are the revised guidelines from the Multinational Association of Supportive Care in Cancer (MASCC) 2004 consensus conference, which have been published on the MASCC website (http://www.mascc.org).2

The objective of these guidelines is to prevent the symptoms of nausea and emesis, and antiemetic prophylaxis is now standard practice to prevent CINV and RINV evoked by moderately and highly emetogenic therapies.2, 5–7 The prevention of emesis allows more patients to resume activities of daily living immediately after treatment. This inevitably has important implications for patients' quality of life and their psychological well being during therapy and also, ultimately, may affect outcomes after treatment. However, despite antiemetic prophylaxis, a few patients still may suffer with these potentially debilitating side effects. Indeed, some patients even may be willing to compromise their potentially life-saving treatment to avoid these complications.8

The 5-Hydroxytryptamine Type-3 Receptor Antagonists

Nausea and emesis have been rated consistently by patients and nursing staff as two of the most debilitating consequences of antineoplastic therapy,9–12 and it has been shown that they have a negative impact on patient quality of life after cancer therapy.13 Significantly, however, since the introduction of the 5-hydroxytryptamine type (5-HT3) receptor antagonists, emesis no longer is rated highest9 but has been replaced by nausea as the most distressing consequence of chemotherapy or radiotherapy reported by patients.10, 11

The 5-HT3 receptor antagonists currently are considered the “gold standard” of antiemetic therapy administered in clinical practice in combination with corticosteroids.5–7 Antiemetic guidelines suggest that, at appropriate doses, the 5-HT3 receptor antagonists demonstrate equivalent efficacy and safety, despite warnings about cardiovascular side effects in the prescribing information for dolasetron, tropisetron, and palonosetron.14–16 However, due to financial considerations, these agents often are used at lower doses than recommended—doses that may be suboptimal or at the threshold for efficacy. Consequently, fewer patients may achieve adequate antiemetic control, necessitating the use of rescue medication, lengthening hospital stays, and risking the development of anticipatory nausea and emesis, all of which will have an impact on the overall cost of therapy. Obviously, these complications will add to the significant burden associated with patients' diagnosis, treatment, and recovery from cancer.

Four 5-HT3 receptor antagonists are licensed in the United States: granisetron, dolasetron, ondansetron, and palonosetron. Tropisetron also is licensed in Europe. It has been shown that each of these is effective in reducing nausea and emesis associated with cancer chemotherapy, although limited data are available for palonosetron, which has become available only recently and currently is licensed only in the United States. Other agents, including tropisetron, are available in other countries. Among these currently available 5-HT3 receptor antagonists, ondansetron reportedly is used most often at suboptimal doses.17

The Dangers of Suboptimal Dosing

The doses of 5-HT3 receptor antagonists currently recommended in the MASCC guidelines are shown in Table 1.2 The issue of “down-dosing”—that is, the administration of these drugs at lower doses than recommended by the package insert—is particularly pertinent, because the 5-HT3 receptor antagonists have a very steep dose-response curve within a narrow dose range.18 This implies that control of emesis will be lost suddenly as doses are reduced (Fig. 1).19 However, because many predisposing and confounding factors exist between patients,20 it is unlikely that a gradual loss in emesis control will be apparent across the whole population. Rather, individuals will experience increased episodes of nausea and emesis as the dose is reduced below threshold. Thus, down-dosing may have potentially serious consequences for particular patients.

Table 1. Dose of the Different 5-Hydroxytryptamine Type 3 Receptor Antagonists, as Recommended by the Multinational Association of Supportive Care in Cancera
  • a

    See the report of the Multinational Association of Supportive Care in Cancer Antiemetic Guideline Committee, 2004.2

  • b

    The U.S. Food and Drug Administration-approved dose is 32 mg.

  • c

    Randomized studies have tested the schedule of 8 mg twice daily.

  • d

    The 1-mg dose is preferred by some panelists based on the results from a small, randomized study in moderately emetogenic chemotherapy and from a Phase II study in highly emetogenic chemotherapy.

 Intravenous8 mg or 0.15 mg/kgb
 Oral16 mgc
 Intravenous1 mg or 0.01 mg/kg
 Oral2 mg (or 1 mgd)
 Intravenous100 mg or 1.8 mg/kg
 Oral100 mg
 Intravenous5 mg
 Oral5 mg
Palonosetron (intravenous)0.25 mg
Figure 1.

This schematic shows the dose-response curve of the 5-hydroxytryptamine type-3 (5-HT3) receptor antagonists. Arrows depict the position of the recommended dose on the curve and the consequence of reducing the dose on the control of emesis. Reproduced with permission from Cancer Today 2002;Suppl 1:1–15.19

Patient Susceptibility

The susceptibility of individual patients to CINV and RINV is dependent on a variety of predisposing factors, including female gender, age, low alcohol intake, high anxiety levels, and a history of motion sickness (Table 2).6, 20–23 However, the relevance of an individual's risk factor profile frequently is underestimated in clinical practice. In early clinical trials, patient susceptibility all too often was overlooked during the search to delineate the emetogenicity of the treatment regimens versus optimal antiemetic dose. For example, a decade ago, patients with cancer who received moderately emetogenic chemotherapy regimens, such as cyclophosphamide, hydroxydaunomycin, oncovin, and prednisone (CHOP) or 5-fluorouracil, doxorubicin, and cyclophosphamide (FAC), were offered older, conventional antiemetics; patients who received chemotherapy regimens that were considered highly emetogenic (e.g., cisplatin > 50 mg/m2) were offered the newer 5-HT3 receptor antagonists. However, the patients who received highly emetogenic therapies often were male patients with lung cancer24 who smoked and drank heavily and, thus, were protected in part from emesis—not solely because they received 5-HT3 receptor antagonists but because of what we now know probably was allosteric and direct modulation of the 5-HT3 receptor by alcohol and nicotine, respectively, which reduced the risk of nausea and emesis in such patients.25–27 This differs from the patients who received CHOP/FAC regimens, who were likely to be female patients with breast cancer who had a high propensity for nausea and emesis (that is, young and female with low alcohol consumption).20

Table 2. Confounding Factors that Influence a Patient's Propensity for Developing Chemotherapy-Induced Nausea and Emesis and Radiotherapy-Induced Nausea and Emesisa
  • CINV: chemotherapy-induced nausea and emesis; RINV: radiotherapy-induced nausea and emesis; NSAIDs: nonsteroidal antiinflammatory drugs.

  • a

    Data from ASHP, 19996; Doherty, 199920; Xu et al., 199321; Koffman et al., 200322; and Schnell, 2003.23

Factors that may increase patients' susceptibility to developing CINV/RINV
 Prior exposure to CINV/RINV
 Evidence of anticipatory nausea and emesis
 History of motion sickness
 History of moderate/severe sickness during pregnancy
 History of emesis, e.g., frequent “food poisoning,” bilious attacks, enteric infections
 Recent exposure to anesthetics, especially nitrous oxide and halogenated, volatile anesthetics
 Recent abdominal, ophthalmic, or facial/cranial surgery
 Concomitant use of nauseogenic drugs, e.g. morphine and other narcotic analgesics, cardiac glycosides, NSAIDs and other antiinflammatory agents
 Raised plasma urea levels associated with hepatic or renal pathology
 Female gender
 Pediatric patients
 Asian ethnicity
 Afro-Caribbean ethnicity
Factors that may decrease patients' susceptibility to developing CINV/RINV
 Alcohol consumption
 Old age
 Cannabinoid use

Some subsequent studies have compared the efficacy of 5-HT3 receptor antagonists with older, conventional antiemetics and found that women were protected significantly better from acute nausea and emesis with granisetron28 or ondansetron,29 weakening the hypothesis that antiemetic regimens should be determined solely by the predicted emetogenicity of the chemotherapy agents received. Rather, a major consideration for all consulting physicians ought to be individual patient characteristics.

Differences in patients' susceptibility for developing nausea and emesis, as discussed above, now can be explained in part by our current understanding of the allosteric modulation of the 5-HT3 receptor. The 5-HT3 receptor belongs to the ligand-gated ion channel superfamily, and the receptor complex possesses pharmacologically distinct recognition sites by which the receptor can be modulated.25, 26, 30–32 In addition to alcohol,26, 33 it is believed that nicotine is able to modulate the 5-HT3 receptor complex, because it displays homologous topologic organization to the nicotinic acetylcholine receptor,34 and estradiol can stabilize the 5-HT3 receptor in both the resting state and the open state.35 Thus, alcohol and nicotine—two commonly used stimulants—are among a host of predictable factors that can affect a patient's propensity for CINV and RINV. General anesthetics also can affect 5-HT3 receptor-mediated susceptibility to nausea and emesis, which has particular relevance for the assessment of antiemetics in animal models.36

In addition, growing interest in pharmacogenetics has identified certain ethnic populations that are at risk from genetic polymorphism of hepatic cytochrome P450 (CYP) metabolic enzymes.37 This translates into a clinical risk for the development or treatment of CINV when patients are either ultrarapid or poor metabolizers for the CYP2D6 isoenzyme.38 The presence of different metabolizer types differs within ethnic populations; for example, ultrarapid metabolizers are present as 2–4% of the population in Northern European countries, 4–5% of the population in North America, and up to 29% of the population in Ethiopia.39 Because ondansetron, dolasetron, tropisetron, and palonosetron are metabolized in part by CYP2D6,40 diminished antiemetic efficacy—leading to a greater frequency of nausea and emesis—or an increased potential for drug-drug interactions may occur as a consequence of ultrarapid or poor metabolism, respectively.16, 38, 41 Therefore, any reduced efficacy observed in ultrarapid metabolizers is likely to be exacerbated if patients in this population receive suboptimal doses of a 5-HT3 receptor antagonist that is metabolized through this isoenzyme. Although the pharmacokinetics of the new 5-HT3 receptor antagonist, palonosetron, reportedly are unaffected by poor or extensive metabolizers of the enzyme,16, 42 only three patients with each phenotype were assessed, no control population was analyzed, and the effect of the ultrarapid phenotype (which is expected to result in increased clearance and decreased drug efficacy) was not assessed. Further data also are required to assess the metabolism of palonosetron in different ethnic groups. Granisetron, unlike other agents in the class, is not metabolized by CYP2D640 and is discussed below.

Preclinical Studies

Preclinical studies using animal models of emesis have identified the dose-response profile of the 5-HT3 receptor antagonists. However, the results of many in vivo studies may be confounded by the use of animals that have recovered insufficiently from anaesthesia from surgical procedures to insert cannula for the intravenous (IV) administration of chemotherapy agents.36 Anesthetics themselves can either sensitize (e.g., halothane, ketamine, nitrous oxide) or suppress (e.g., propofol) emesis in response to emetic stimuli, and a number of studies investigating the antiemetic efficacy of ondansetron may have failed to provide the animals with an adequate recovery period (> 3 days), thus potentially overestimating antiemetic efficacy. Indeed, early preclinical studies in ferrets supported a linear dose-response curve for ondansetron, similar to granisetron, whereas subsequent investigations in a series of well controlled studies identified the true form of the ondansetron dose-response curve as nonlinear.43 This emphasizes the need for caution when comparing discrete preclinical studies using different methodologies, anesthetic agents, and animal models, and it underlines the importance of adequate controls.


It has been shown that granisetron is effective against highly or moderately emetogenic chemotherapy over a range of doses, with a minimum effective dose of 10 μg/kg IV or 1 mg orally once daily.17, 44–52 Although doses < 10 μg/kg IV result in significantly fewer patients achieving a complete response (no emesis and no use of rescue therapy) or a major response (≤ 2 emesis episodes) (2 μg/kg IV44, 45 or 5 μg/kg IV46), these doses are not used routinely in clinical practice. In Europe, the registered dose has been 3 mg or 40 μg/kg, based on analysis of early published data, and currently is 2 mg orally, 1–3 mg IV, or 3 mg intramuscularly. However, the United States Food and Drug Administration deemed that the 10 μg/kg dose was fully effective, because the slight increase in response rates with the higher dose was not statistically significant. It is noteworthy that those studies did not include corticosteroids.

Nevertheless, in a recent review, the dose-responsiveness of granisetron was examined for the treatment of CINV, and the results suggest that there is some benefit to the higher dose (40 μg/kg) in certain patient groups.47 However, none of the studies cited demonstrated a significant difference in primary efficacy variables between granisetron at 10 μg/kg and at 40 μg/kg. For example, in the study by Navari and colleagues,46 a complete response (no emesis and no use of rescue therapy) was observed in 18%, 41%, 40%, and 47% of patients who received cisplatin-based chemotherapy (81–120 mg/m2) who also received granisetron at doses of 5 μg/kg, 10 μg/kg, 20 μg/kg, or 40 μg/kg, respectively. The 6% improvement in the complete response rate (20 of 49 patients vs. 22 of 47 patients) after granisetron 40 μg/kg, compared with granisetron 10 μg/kg, was not statistically significant.

There is a greater likelihood of suboptimal dosing after oral versus IV administration, because a once-daily 2.0-mg oral dose is administered to patients as 2 1.0-mg tablets, which increases the potential that patients may receive or take only 1 of these tablets. Two studies have examined the dose-responsiveness of oral granisetron48, 49; however, those studies examined the efficacy of granisetron administered twice daily, which does not readily equate to the potentially suboptimal dosing profile of granisetron 1 mg once daily. Nevertheless, patients undergoing moderately emetogenic chemotherapy experienced good antiemetic control with granisetron 0.5 mg twice daily (total daily dose, 1.0 mg).48, 49 In the first study, 53.6% of patients who received granisetron 0.5 mg twice daily experienced a complete response (no emesis, no worse than mild nausea, no rescue medication, and no withdrawal during the treatment period) compared with 58.8% and 54.5% of patients who received granisetron 1.0 mg twice daily and 2.0 mg twice daily (P value not significant).48 In the second study, proportionally more patients experienced a complete response, with ≈ 70% of patients remaining emesis-free after granisetron 0.5 mg twice daily49 compared with ≈ 80% of patients who received granisetron 1.0 mg twice daily, which differed significantly from the lower dose (P < 0.009). The complete response rate with granisetron 2.0 mg twice daily was similar to the rate achieved with granisetron 0.5 mg twice daily.


Studies that compared the efficacy of 2 IV doses of dolasetron, 1.8 mg/kg and 2.4 mg/kg, have demonstrated equivalent efficacy.53, 54 However, the results from some studies have suggested that the approved oral dose of dolasetron may be suboptimal.55–57 There is a significant dose-response effect across 4 oral doses of dolasetron (25 mg, 50 mg, 100 mg, and 200 mg), with the highest antiemetic efficacy demonstrated at 200 mg. Eighty-three percent of patients who underwent moderately emetogenic chemotherapy experienced a complete response (no emetic episodes and no rescue medication) after they received dolasetron at a dose of 200 mg, whereas 73%, 71%, and 45% of patients were complete responders after they received dolasetron at doses of 100 mg, 50 mg, and 25 mg, respectively (P < 0.001; linear trend with dose).55 Similarly, there was a significant linear trend for patients who experienced a complete plus major response (0–2 emetic episodes and no rescue medication) after they received dolasetron 25–200 mg (P < 0.001). However, statistical analyses did not investigate differences in efficacy between individual doses.

One study showed that the antiemetic efficacy of oral and IV dolasetron 100 mg also was less effective than standard doses of granisetron or ondansetron for the prevention of nausea and emesis associated with highly emetogenic chemotherapy regimens.56 In that study, patients who received ondansetron or granisetron were grouped together into a “standard” treatment group and were given doses of antiemetic according to the high-dose chemotherapy/radiotherapy conditioning schedule they received before they underwent hematopoietic stem cell transplantation. For example, ondansetron was administered at a dose of 10 mg IV every 6 hours, 8 mg orally every 6 hours, or 8 mg orally twice daily; and granisetron was administered at a dose of 2 mg orally once daily or 1 mg orally twice daily, depending on the chemotherapy or radiotherapy regimen administered. In contrast, dolasetron was administered as a single fixed dose (100 mg orally) for all treatment regimens. Only 65% of patients who received dolasetron achieved a major response (1–2 emesis episodes with any level of nausea or no emesis with severe nausea) or a complete response (no emesis and mild-to-moderate nausea), compared with 87% of patients who received granisetron or ondansetron (P < 0.05). The study did not distinguish between patients who received either chemotherapy alone or in combination with radiotherapy, although an equal number of patients in each group (n = 15 patients) received concomitant total body irradiation.56 Dolasetron-treated patients were free from emesis on 144 days of conditioning, compared with granisetron-treated and ondansetron-treated patients, who had 159 emesis-free conditioning days (P < 0.05). In addition, nausea occurred earlier in dolasetron-treated patients (on Days − 4, − 2, and − 1; P < 0.05).56 Therefore, further reduction from the recommended dose of dolasetron made in an attempt to enhance cost effectiveness potentially may have disastrous consequences on the outcome of emesis control for individual patients.


At clinically effective doses, tropisetron did not demonstrate an appreciable dose-response effect on control of emesis in patients who were receiving highly emetogenic chemotherapy.58 Total control (no emesis and no nausea) was achieved in 71%, 51%, 61%, and 58% of patients who received IV tropisetron at doses of 5 mg, 10 mg, 20 mg, and 40 mg, respectively. However, a comparison of IV tropisetron at doses of 2 mg and 5 mg indicated that the lower dose was below the threshold for therapeutic efficacy in patients who received highly emetogenic58 or moderately emetogenic59 chemotherapy. Seventy-three percent of patients who received tropisetron 5 mg IV experienced total control (no emesis) after moderately emetogenic chemotherapy, compared with only 55% of patients who received tropisetron 2 mg IV (P = 0.02). Similarly, better control of acute nausea (≤ 15 minutes) was experienced by patients after tropisetron 5 mg IV compared with tropisetron 2 mg IV.59

In addition, comparative studies have demonstrated that, like dolasetron, even the approved dose of tropisetron (5 mg IV or orally) may be suboptimal in patients who are receiving highly emetogenic chemotherapy.58, 60, 61 For example, significantly fewer patients who were treated with tropisetron 5 mg IV achieved a complete response (no nausea or emesis or only mild nausea) plus a major response (1 emesis episode or no emesis with moderate-to-severe nausea) after high-dose cisplatin chemotherapy compared with patients who received ondansetron 24 mg IV (P = 0.021).60 Therefore, a reduction in the dose of tropisetron below 5 mg IV also may be detrimental to patients in terms of emesis control and the effect on their quality of life.


It has been shown that ondansetron is effective for preventing emesis in patients who are undergoing emetogenic chemotherapy. Although few, large, well controlled, dose-finding studies have been conducted with ondansetron,62–66 available evidence from Phase III and later studies suggests that the agent is most effective when administered as a single high dose (32 mg IV) or as a multiple dosing regimen (0.15 mg/kg IV 3 times daily).

However, ondansetron often is administered at doses as low as 8 mg once daily for both moderately and highly emetogenic chemotherapy in association with corticosteroids.17 Some studies demonstrate good efficacy with ondansetron 8 mg62 or equal efficacy between ondansetron 32 mg IV and 8 mg IV once daily.63 Compared with ondansetron 32 mg IV (51%) and granisetron 3 mg IV (56%), 59% of patients who received ondansetron 8 mg IV achieved a complete response (no emetic episodes) after cisplatin-based chemotherapy (≥ 50 mg/m2).63 However, efficacy should never be compromised over cost, and some patients do not respond well to this low dose.64, 65, 67–69

Patients who received ondansetron 8 mg IV once daily after moderately emetogenic chemotherapy experienced a significantly inferior antiemetic response for all primary and secondary variables compared with patients who received ondansetron 32 mg IV once daily (P < 0.05).64 Similarly, in patients who were undergoing highly emetogenic chemotherapy, significantly fewer patients who received ondansetron 8 mg IV once daily than patients who received ondansetron 32 mg IV once daily experienced a complete response (no emetic episodes; 35% vs. 48%; P = 0.048). In addition, more patients receiving the lower dose had emetic episodes (P = 0.015), and the failure rate was higher (> 5 emetic episodes and required rescue medication; 39% vs. 20%; P = 0.018).64 In line with this finding, a recent retrospective analysis of 224 women with breast carcinoma who received cyclophosphamide-containing regimens in the United States between 1998 and 2002 found that significantly fewer patients who received ondansetron 8 mg IV experienced total control of emesis compared with patients who received either ondansetron 32 mg IV or granisetron 10 μg/kg IV or 1 mg orally (42.6% vs. 64.5% and 67.5%, respectively; P < 0.01).70 Patients who received ondansetron 8 mg IV also required significantly more rescue antiemetics than patients who received granisetron (44.1% vs. 23.8%; P < 0.05). The view that a single dose of ondansetron 8 mg IV provides suboptimal antiemetic efficacy has been supported further by a recent meta-analysis in which a pooled analysis of noncisplatin-based studies comparing ondansetron 8 mg IV once daily with granisetron 3 mg IV significantly favored granisetron (P = 0.041).71

Some investigators believe that concomitant administration of corticosteroids with ondansetron will compensate for these variations in efficacy of different doses of ondansetron. However, similarly poor efficacy results have been obtained after ondansetron 8 mg IV in combination with dexamethasone.65 More patients who received ondansetron 32 mg IV plus dexamethasone 8 mg IV on the first day of high-dose cisplatin chemotherapy experienced a complete response (no emesis) compared with patients who received ondansetron 8 mg IV plus dexamethasone 8 mg IV (P = 0.0001). In addition, 17% of patients who received low-dose ondansetron were regarded as failures (> 5 emesis episodes) compared with only 5% of patients who received ondansetron 32 mg.

Even studies in which ondansetron 8 mg IV was administered twice daily have demonstrated inferiority compared with both ondansetron 24 mg IV and 32 mg IV once daily for the control of cisplatin-induced nausea.72 Only 36% of patients experienced complete control of nausea (no nausea, no rescue medication, and no withdrawal) after they received ondansetron 8 mg IV twice daily compared with 56% (P = 0.001) and 50% (P = 0.019) of patients who received 24 mg or 32 mg, respectively. Low-dose ondansetron—below the recommended dose—like dolasetron and tropisetron, therefore, may have severe consequences in terms of patients' quality of life and as potential cost implications incurred by the use of rescue medications and prolonged hospital stays or readmittance to hospital after discharge.


The new 5-HT3 receptor antagonist palonosetron only recently (in 2003) received marketing approval in the United States for the prophylaxis of CINV; thus, limited clinical data are available for the agent. The results of a Phase II study investigating the efficacy of single doses of palonosetron 0.3–90 μg/kg IV have suggested that the agent does not demonstrate a marked dose-response effect on control of emesis in patients receiving cisplatin-based chemotherapy.73 A complete response (no emesis, no retching, no rescue therapy) was achieved in 24%, 46%, 40%, 50%, and 46% of patients who received IV palonosetron at doses of 0.3–1.0 μg/kg, 3.0 μg/kg, 10.0 μg/kg, 30.0 μg/kg, and 90.0 μg/kg, respectively. No significant differences in efficacy were observed between doses of 3 μg/kg and 90 μg/kg, although all of these doses showed a trend toward superiority compared with the lowest dose (0.3–1.0 μg/kg).

Two Phase III studies conducted with palonosetron at doses of 0.25 mg IV and 0.75 mg IV have shown that the agent is superior to both ondansetron 32 mg IV74 and dolasetron 100 mg IV75 in the prevention of emesis induced by moderately emetogenic chemotherapy. It is noteworthy that, compared with ondansetron, the approved dose of palonosetron (0.25 mg) was more effective than the higher dose of 0.75 mg: Significantly more patients achieved a complete response with palonosetron 0.25 mg than with ondansetron 32 mg in the acute phase (0–24 hours: 81.0% vs. 68.6%, respectively), in the delayed phase (24–120 hours: 74.1% vs. 55.1%), and overall (69.3% vs. 50.3%; P < 0.025 for all comparisons).74 The complete response rates achieved with palonosetron 0.75 mg in the acute phase (73.5%), in the delayed phase (64.6%), and overall (58.7%) did not differ significantly from the rates achieved with ondansetron.

The apparently lower efficacy of high doses (> 0.25 mg) of palonosetron merits further investigation. If this bell-shaped dose-response curve is a genuine finding, then one possible explanation is the conversion of the single marketed enantiomer of palonosetron into multiple enantiomeric forms, because the other three enantiomers of palonosetron are less potent.76 It is known that enantiomeric conversion occurs in some drugs with a long in vivo plasma half-life, and palonosetron could be tested in vitro by incubating in plasma at body temperature for 12 hours and assaying for the presence of other enantiomers.

In the second study, the complete response rates achieved overall and in the delayed phase (24–120 hours) with palonosetron at doses of 0.25 mg and 0.75 mg were significantly higher compared with the response rates achieved with dolasetron 100 mg after moderately emetogenic chemotherapy (46.0% and 47.1% vs. 34%, respectively; and 54.0% and 56.6% vs. 38.7%; P < 0.025 for both comparisons).75 The palonosetron response rates also were higher numerically in the acute phase (0–24 hours: 63.0% and 57.1% vs. 52.9%), although the differences were not significant. Although palonosetron was significantly more effective than dolasetron only in the delayed phase, control of delayed emesis is correlated with effective control of acute-onset emesis, lending further support to the suggestion that the 100-mg dose of dolasetron may be suboptimal. The long-acting efficacy of palonosetron indicated in these studies suggests a role for serotonin beyond Day 1 of chemotherapy and warrants further investigation.

A potential problem with palonosetron may result from the use of a fixed dose (mg) rather than a μg/kg dose, because some larger patients may receive an inadequate dose. However, further studies are required to assess fully patients' response to therapy and whether, in clinical practice, this agent is used at doses lower than those recommended.

Treating Patients with Refractory Emesis

Some studies have shown that patients who have previously failed ondansetron can be treated successfully with granisetron in subsequent cycles (Fig. 2).77–79 For example, 58% and 29% of patients who failed previous treatment with ondansetron achieved complete control of nausea and emesis after cyclophosphamide-based chemotherapy (≥ 500 mg/m2) and cisplatin-based chemotherapy (≥ 50 mg/m2), respectively.77 However, the initial high failure rate with ondansetron in these studies may be due to the fact that this agent was administered at a low dose (for example, 8 mg IV once daily plus dexamethasone77).80 Therefore, it is not surprising that, in subsequent chemotherapy cycles with granisetron at a dose of 3 mg IV plus dexamethasone (a combination with proven antiemetic efficacy), more patients experienced complete protection from emesis compared with patients who were re-treated with ondansetron 8 mg IV. Moreover, it is unlikely that patients achieved a better response with granisetron purely due to use of a high dose (3 mg), because a number of studies have demonstrated an equivalent emetic response after granisetron, at a dose of 10 μg/kg or 40 μg/kg.44, 46

Figure 2.

These charts illustrate the proportion of ondansetron-refractory patients who achieved complete emetic control when they were switched to intravenous granisetron 3 mg (a) compared with intravenous ondansetron 8 mg plus intravenous dexamethasone 10 mg (based on data from de Wit et al., 200177) and (b) over successive chemotherapy cycles (based on data from Carmichael et al., 1998.78).

A further rationale that recently was proposed for the efficacy of granisetron in patients who were refractory to ondansetron therapy may be the issue of genetic polymorphism of the CYP2D6 enzyme.81 Recent data suggest that those patients who are ultrarapid metabolizers of CYP2D6 substrates are at risk of developing severe acute CINV after moderately to highly emetogenic chemotherapy regimens, despite treatment with ondansetron.39 Ondansetron is metabolized partially by CYP2D6 in addition to CYP3A4, CYP1A1, and CYP1A2; whereas granisetron is metabolized principally by CYP3A4.40 Therefore, ultrarapid metabolizers of CYP2D6 may clear CYP2D6 substrates quickly, lowering the serum concentration of ondansetron, for example, to below threshold. Indeed, patients who were treated with ondansetron who were identified as ultrarapid metabolizers of this isoenzyme experienced a significantly higher mean number of emesis episodes compared with other patient groups (0–4 hours: P < 0.001; 5–24 hours: P < 0.03) after chemotherapy.39 Not surprisingly, nausea also was rated as more severe in these patients.

Pharmacodynamic Properties of the 5-HT3 Receptor Antagonists

There is appreciable variation in the pharmacodynamic properties of the 5-HT3 receptor antagonists. However, although data investigating receptor-binding characteristics of granisetron, ondansetron, tropisetron, and palonosetron to 5-HT3 receptors are forthcoming, the binding characteristics of hydrodolasetron (the active metabolite of dolasetron) have not been published.

Granisetron and tropisetron display insurmountable antagonism at 5-HT3 receptors; that is, once they are bound to the receptor, these agents are not able to be displaced by the addition of further 5-HT.82 This is in contrast to ondansetron, which can be displaced by exogenous 5-HT and, as such, demonstrates competitive antagonism.82 It is believed that the insurmountable antagonism of granisetron underlies its 24-hour duration of action83; indeed, significant antiemetic activity still is evident at 48 hours postdose and does not return to normal until 72 hours after the single dose. In healthy volunteers, cutaneous skin-flare data demonstrate statistically significant receptor antagonist activity of granisetron 24 hours after administration of a single 40 μg/kg IV injection,83, 84 which contrasts with its plasma half-life in volunteers (5 hours).85 By comparison, ondansetron has a short plasma half-life in healthy volunteers (3.5 hours),86 and antagonist activity at the 5-HT3 receptor no longer is detectable 24 hours after the administration of 8-mg or 16-mg IV doses.87

The variation in the pharmacodynamics of ondansetron and granisetron also may explain in part the findings of 2 recent noncisplatin-based studies, which reported the possible inferiority of ondansetron 8 mg IV versus ondansetron 32 mg or granisetron 10 μg/kg, 1 mg, or 3 mg IV.70, 71 Whereas the peak incidence of cisplatin-induced emesis occurs around 4 hours after administration, emesis after other chemotherapy agents (e.g., cyclophosphamide and carboplatin) may peak some 12–16 hours after treatment.88 Therefore, patients who received treatment with agents that have a shorter duration of action are more likely to experience breakthrough symptoms. Granisetron and ondansetron also may differ in their functional activity; unlike ondansetron, in which the mode of action is through the blockade of 5-HT3 receptors on abdominal vagal nerves, granisetron antagonizes both 5-HT3 autoreceptors on enterochromaffin cells89 and 5-HT3 receptors on abdominal vagal afferents.90

It has been suggested that palonosetron shows insurmountable antagonism of 5-HT3 receptors, although this is based on in vitro experiments on guinea pig ileum,76 which responds in a unique way to 5-HT through both 5-HT3 receptors and 5-HT4 receptors.89 Although 5-methoxytryptamine was used to desensitize 5HT4 receptors, no evidence was presented to support the claim that a 5-HT4 effect was not present functionally.76 The apparently insurmountable block, therefore, may have been the effect of an intact, 5HT4-mediated effect or a slow antagonist-receptor dissociation.

The effect of 5-HT on vagal afferent nerves is 5-HT3-mediated and does not involve 5-HT4 receptors. In this tissue, granisetron and tropisetron showed insurmountable block of 5-HT3 receptors (rabbit and rat), whereas ondansetron was a competitive antagonist.82, 91 Data on the effects of palonosetron on the vagus nerve are not available. Therefore, although palonosetron binding data do exist, the complexity of the 5-HT contractile response in the guinea pig ileum requires additional in vitro studies to be conclusive.


The current review has demonstrated that the clinical objective of preventing emesis will be achieved only when the 5-HT3 receptor antagonists are administered at adequate and proven doses. Antiemetic guidelines produced by various societies (e.g., MASCC,2 American Society for Clinical Oncology [ASCO],5 American Society for Health-System Pharmacists [ASHP]6) concur that these agents should be given in combination with a corticosteroid to patients who are at high or moderate risk of developing emesis. These guidelines also provide a dosing guide for the use of 5-HT3 receptor antagonists, although they are at variance with each other and with the recommendations laid out by the manufacturers. For example, ASCO and MASCC guidelines recommend the use of dolasetron 100 mg orally or IV, whereas ASHP guidelines recommend 100–200 mg orally for cisplatin-induced emesis. Therefore, ASHP recognizes that the recommended dose of dolasetron (100 mg)14 may be suboptimal and has been shown to be inferior to standard doses of granisetron or ondansetron.56

These guidelines also state that single dosing regimens are as effective as multiple daily dosing and, as such, recommend once-daily dosing for all agents. However, experience suggests that ondansetron 8 mg once daily may be a suboptimal dose, particularly in patients receiving chemotherapies who have a late-acute onset profile of emesis.58, 70 These patients may benefit from multiple daily dosing schedules recommended by the manufacturer (0.15 mg/kg 3 times daily)86 or the single high dose (32.0 mg)86 that has been used in some comparative studies.64, 70

Some investigators have documented the importance of tailoring antiemetic therapy with the emetogenicity of the chemotherapy regimen.92 Although this is a valid consideration, ondansetron 8 mg IV once daily often is used as an antiemetic regimen (and is recommended by the antiemetic guidelines)2, 5, 6 for moderately emetogenic chemotherapy, for example, with cyclophosphamide.92 The low dose of ondansetron, coupled with its short half-life in cancer patients (4 hours),86 the competitive nature of its antagonism, which means it may be displaced readily by endogenous serotonin,82 and the late-acute onset of emesis after cyclophosphamide20 make this combination potentially detrimental in terms of patients' control of nausea and emesis. Therefore, beyond the emetogenic rating of the chemotherapy, it is important to consider the susceptibility of each patient, the duration of action and the nature of the antagonism of the antiemetic agent compared with the onset and duration of emesis evoked by the cancer treatment (Fig. 3).19, 83–85

Figure 3.

(A) The duration of receptor blockade of intravenous ondansetron versus granisetron expressed as a percentage of maximum inhibition of 5-hydroxytryptamine (5-HT)-induced skin flare compared with (B) the onset and duration of early and late acute-onset chemotherapy regimens. An asterisk indicates P < 0.05. NS: not significant. The graph in A was calculated based on data from Upward et al., 199083; Cooper et al., 198884; Roche Laboratories, 200185; and Sweetland et al., 1992.87 The graph in B was reproduced with permission from Cancer Today 2002;Suppl 1:1–15.19

Granisetron is not administered routinely at doses lower than recommended; consequently, patients are able to achieve adequate control of nausea and emesis after cytotoxic drug therapy. This is in contrast with other agents, which often are used at suboptimal doses—this reduction in dosage may lead to more patients experiencing uncontrolled emesis. Specifically, even at its recommended oral dose, some studies have shown that the efficacy of dolasetron is suboptimal,55–57 and comparative studies with tropisetron have also demonstrated that the recommended dose may result in fewer patients achieving adequate control of nausea and emesis.58, 60, 61 Clinical experience with palonosetron is sparse, although the dangers of suboptimal and supraoptimal dosing may apply to this agent.

Suboptimal dosing with any of the 5-HT3 receptor antagonists ultimately will prove counterproductive in terms of hospital resources—demands on nursing staff, use of rescue medication, lengthened hospital stays, and the possible readmittance of patients with breakthrough emesis postdischarge—and will add to the already significant socioeconomic burden associated with cancer therapy. In addition, although we are now able to predict and explain many of the factors that contribute to a patient's propensity for developing the debilitating symptoms of nausea and emesis, our comprehension of all issues still requires further study and clarification. Therefore, the dose of antiemetic agent administered needs to be sufficiently high to ensure good emesis control in the whole population.