Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Comparison of clinical efficacy of serotonin receptor antagonists in highly emetogenic chemotherapy

This is not the most recent version

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

The primary objective of this review is to investigate the clinical efficacy of different serotonin receptor antagonists (5‐HT3 RAs) in the control of acute and delayed emesis induced by highly emetogenic chemotherapy.

The secondary objectives are to examine eligible studies for information on adverse events and to assess if there are important differences in the adverse events caused by the different anti‐emetic agents.

Background

Nausea and vomiting are considered among the most distressing side effects associated with chemotherapy in cancer patients. However, the magnitude of the problem is not fully appreciated by physicians and nurses, as illustrated by the underestimation of the predicted rate of nausea and vomiting in a survey, conducted by Grunberg 2003, of a group of women with breast cancer treated with chemotherapy.

Severe nausea and vomiting can result in metabolic disturbances secondary to dehydration and malnutrition. This could in turn interfere with the clinical course of the disease and with the patients acceptance of chemotherapy. The patient´s perception of adequate control of nausea and vomiting is therefore an essential part of the management plan of cancer, both for improving quality of life as well as supporting full adherence with the chemotherapy program (Hesketh 1999).

Vomiting is a partially understood complex mechanism based on the interaction between humoral factors, afferent fibers, as well as inhibition and excitation of somatic and visceral musculature (Ettinger 1995).

The vomiting centre in the brain medulla which is responsible for the coordination of emesis is deemed to be activated by afferent impulses from the chemoreceptor trigger zone (CTZ), the vestibular apparatus, the mid brain, the limbic system and the pharynx or gastrointestinal tract (Siegel 1981).

Following activation of the vomiting centre, efferent impulses to the salivation center, abdominal muscles, respiratory center and cranial nerves lead to vomiting (Craig 1987). The understanding of the physiology of emesis has been greatly accelerated by the identification of specific neurotransmitter receptors that play a pivotal role in drug‐induced emesis. Several critical receptors which trigger the act of vomiting when stimulated by chemotherapeutic agents, their metabolites or released neurotransmitters have been identified in the central nervous system (CNS) and in the gastrointestinal (GI) tract (Borison 1983; Tack 2000).

Activation in the brain medulla of either the vomiting centre or the CTZ, or both, in the area postrema (located outside the brain and sensitive to both chemical stimuli from the blood and the cerebrospinal fluid (CSF)), is mediated through dopamine, opiate, histamine, acetylcoline, neurokinin (NK‐1) or serotonin receptors (Ettinger 1995). Acute emesis caused by cytotoxic drugs is associated with an increase in the concentration of serotonin in the intestine and the brainstem. However, the current hypothesis about chemotherapy‐induced nausea and vomiting (CINV) is that chemotherapy‐induced serotonin release from enterochromaffin cells in the gut results either in direct interaction with CNS serotonin receptors or in the stimulation of 5HT3 receptors on afferent vagal fibers in the gut. This then generates impulses to the centre of vomiting in the lateral reticular formation in the medulla (ASHP 1999; Cubeddu 1990a). The validity of this pathogenic model is supported by the fact that anti‐emetic agents which bind to CTZ and peripheral receptors are the most effective in the prevention of CINV (Gralla 1991). Delayed nausea and vomiting is less well understood and may involve mechanisms other than those mediated through serotonin receptors (Kris 1994; Rudd 1994).

The first receptors targeted by drug research have been dopamine receptors which are found in high concentration in the CTZ. Phenothiazines were the first chemical substances known to antagonize dopamine receptors. They showed some weak anti‐emetic activity, associated with hypotension and other important side effects (Moertel 1963). The introduction of benzamide metoclopramide represented a substantial improvement in the prevention of CINV (Gralla 1981). However, despite the high affinity with dopamine receptors resulting in their full saturation, this chemical agent was found to be consistently efficacious at elevated doses only, raising uncertainty about its mechanism of action (Strum 1982).

The characterization of the serotonin receptors and the demonstration of a moderate affinity of metoclopramide for the binding of the correspondent type 3 (5‐HT3) suggested that the prevention of emesis by this drug at high doses could be ascribed to a blockade of the serotonin receptors (Fozard 1978). The development of highly selective antagonists for the 5‐HT3 receptors marked a substantial improvement in the control of CINV because of the better therapeutic index compared with metoclopramide and the absence of the extrapyramidal reactions adverse mediated by dopamine receptors binding (Cubeddu 1990a). The first selective 5‐HT3 RA, Ondansetron, was approved in 1991, followed by Granisetron in 1993, Tropisetron in 1994, Dolasetron in 1997, and Palonosetron in 2003. To date, the 5‐HT3 RAs, either alone or in combination with a corticosteroid, represent the "anti‐emetic gold standard" for moderate to highly emetogenic cancer chemotherapy (ASCO 2006; ESMO 2005).

The various compounds are characterized by different chemical structure but have the same general mechanism of action: blockade of serotonin receptors (Koeller 2002). However, selectivity of 5‐HT3 receptors binding is somewhat different among the 5‐HT3 RAs. For example, Granisetron strongly and very selectively binds to 5‐HT3 receptors, whereas Ondansetron displays 20% of unselective non‐5HT3 binding (Blower 2003). Good bio‐availability in the range of 50 to 80% is common to all these drugs with no substantial differences in absorption between intravenous and oral administration (Balfour 1997). The terminal half‐life of the 5‐HT3 RAs differs. Palonosetron has the longest (40 hours) compared with the shortest for Ondansetron (three to five hours). Hepatic metabolism of the 5‐HT3 RAs is mediated through drug oxidation by the different isoenzymes of the citochrome P450 and particularly from CYP1A2, CYP2D6 and CYP3A4 (Davis 2001).

Granisetron is the only drug of this class not metabolised by the CYP2D6 pathway but primarily via the CYP3A family (Bloomer 1994). This difference might be important for individual patients, considering that the CYP2D6 enzyme is characterized by genetically polymorphic variants with altered function. However, from a pharmacokinetic perspective, slow or fast metabolic pathways of the 5‐HT3 RAs can prevail in the individual patient, resulting in either lower or enhanced plasma concentration of the drug (Blower 2002; Kaiser 2002).

Compared with Granisetron, Dolasetron, Palonosetron and Tropisetron, which do not require dose adjustment in patients with hepatic impairment (Palmer 1994; Hoechst 1999; Rhoda 1993; TML 2004) a dose reduction of Ondansetron is recommended in patients with severe hepatic failure because of significantly reduced clearance of the drug (Blake 1993; Figg 1996).

Given that these differences might be relevant for the care of individual patients and despite two previous systematic reviews comparing Granisetron with Ondansetron in the setting of highly emetogenic chemotherapy, (Del Giglio 2000; Mendarte 2000), a systematic review assessing the clinical efficacy of the various 5‐HT3 RAs is warranted.

Objectives

The primary objective of this review is to investigate the clinical efficacy of different serotonin receptor antagonists (5‐HT3 RAs) in the control of acute and delayed emesis induced by highly emetogenic chemotherapy.

The secondary objectives are to examine eligible studies for information on adverse events and to assess if there are important differences in the adverse events caused by the different anti‐emetic agents.

Methods

Criteria for considering studies for this review

Types of studies

We will only include randomised trials where randomization is explicit and appropriate with at least 20 patients per treatment group and data on at least one of the outcome variables. Cross‐over randomised studies will be included if information for the first phase of the study are available. As there is a likelihood of a carry‐over effect, cross‐over studies will be included only if they report first period data. Randomised trials with inadequate allocation concealment, such as allocation generated by alternation, use of case record numbers, dates of birth or day of the week; or by any other procedure which is transparent before allocation will be excluded.

Studies of nausea and vomiting associated with moderately emetogenic chemotherapy, radiotherapy, autologous or allogeneic bone marrow transplantation or surgery will be excluded.

Types of participants

Adult cancer patients undergoing highly emetogenic chemotherapy. This is defined as cytotoxic drug for which emesis is expected to occur in more than 90% of chemotherapy administrations in the absence of antiemetic prophylaxis and will be based on the Antineoplastic Agents Emetic Risk Classification adopted by the American Society of Clinical Oncology (ASCO 2006).

Types of interventions

Any 5‐HT3 RA compared with any other drugs of this class according to the following combinations:

  • 5‐HT3 RAs as single anti‐emetic treatment in both study groups;

  • 5‐HT3 RAs in combination with corticosteroids in both study groups;

  • 5‐HT3 RAs as single anti‐emetic treatment in both study groups with different dosing and dosing schedules ‐ The inclusion of trials employing different dosages of drugs is based on theoretical considerations about serotonin receptors saturation associated with this class of drugs (Blower 2002). However, as long as the minimum efficacious dose of the drug is administered, it is currently believed that greater doses will not result in greater efficacy. Concerning the route of administration, both oral and intravenous are currently considered equivalent in term of efficacy (Gralla 1998; Perez 1998).

Types of outcome measures

  • Vomiting: proportion of participants with complete absence of vomiting or retching, either documented by direct observation or according to patient diaries.

  • Nausea: proportion of patients with complete absence of nausea, with or without mild nausea as documented in patient diaries.

  • Total control of nausea and vomiting: proportion of participants with complete absence of vomiting and absence of nausea, with or without mild nausea.

  • Adverse events: proportion of patients experiencing minor or severe adverse events, or both.

  • Acute is defined as occurring within 24 hours of chemotherapy.

  • Delayed is defined as occurring after 24 hours but within seven days of chemotherapy. Where data are given on a daily basis, the rate for the worst of these days will be used.

  • At least three days of follow‐up will be required for the inclusion of trials designed to evaluate delayed nausea and vomiting. Data for acute and delayed nausea and vomiting will be recorded separately.

Search methods for identification of studies

Databases and registers of clinical trials will be searched. We will hand‐search relevant journals, conference proceedings and bibliographies from retrieved trials, meta‐analyses, narrative reviews. Drug companies and authors of eligible trials will be contacted for further details.

Electronic searches

The search strategy will be based on the guidelines of the Cochrane Handbook for Reviews of the Effects of Interventions and the Cochrane Pain, Palliative Care and Supportive Care (PaPas) Review Group (CRG).

The following electronic databases will be searched: The Cochrane Central Register of Controlled Trials (CENTRAL), The Cochrane Pain, Palliative and Supportive Care (PaPas) Group Register, PubMed, EMBASE and LILACS. We will also search the following clinical trials registers:

  • The meta‐register of Controlled Trials‐International database of ongoing and completed trials (www.controlled‐trials.com);

  • The International Standard Randomised Controlled Trial Number (ISRCTN) Register (www.controlled‐trials.com);

  • The National Research Register‐database of ongoing and recently completed research projects funded by, or of interest to, the UK's National Health Service (www.doh.gov.uk);

  • The National Cancer Institute Clinical Trials service (www.cancer.gov);

  • Information on Clinical Trials and Human Research Studies database containing information about US federally and privately supported clinical research in human volunteers. (http://clinicaltrials.nci.nih.gov);

  • The Clinical Trials Centre of the National Health and Medical Research Council of Australia website (www.ctc.usyd.edu.au/);

  • The Clinical Trials Register of Trials Central (www.trialscentral.org/index.html).

Our searches will begin from the year 1990, because the first commercial 5‐HT3 RA, Ondansetron, was approved in 1991 (Cubeddu 1990b).
No language restriction will be applied.
PubMed will be searched through a methodological filter for Randomized Trials (Publication Type) and phases one and two of the Cochrane Sensitivity Search Strategy for Randomized Trials (Clarke 2003; Robinson 2002) ‐ please see Appendix 1.
EMBASE will be searched using a combination of free text terms (i.e. random*, crossover*, cross‐over*, factorial*).
LILACS: A methodological filter for clinical trials will be applied for searching the database.

Definition and combination of search terms

The search will be built on three concepts: cancer, chemotherapy and 5‐HT3 RAs antiemetic agents. The intervention‐related concept (i.e. the anti‐emetic therapy) and the condition‐related concept (i.e. the cancer chemotherapy) will be combined. The intervention‐related concept will be searched using Boolean term OR to combine the different 5‐HT3 RAs. The part of the search strategy terms for the condition‐related concept will be built up by combining terms for neoplasms and cancer chemotherapy with the OR term. As the review is focused on highly emetogenic chemotherapy regimens, studies using Cisplatin (the most representative drug of this category), will be searched for specifically.

The search terms to be used are based in part on searches employed in other narrative and systematic reviews, as well the citations in relevant papers already identified.
The search will be based on the relevant controlled‐vocabulary terms for the specific database, along with free text words and phrases.

Intervention‐related terms

To ensure maximum sensitivity, we will use the names of the different 5‐HT3 RAs and surrogate terms for anti‐emetics.
As noted above, this will be combined with the OR term. However, to increase specificity, for reports of trials that directly compare at least two 5HT3‐RAs which is the focus of this review, we will also run a search in which pairs of 5HT3‐RAs are combined with the AND term and these pairs are combined with the OR term.

Searching other resources

In addition, the following websites of large cooperative research groups in cancer and pharmaceutical companies will be searched:

  • The European Organization for Research and Treatment of Cancer (EORTC) website (www.eortc.be/);

  • The Eastern Cooperative Oncology Group website (ecog.dfci.harvard.edu/);

  • The Southwest Oncology Group website (swog.org/)

  • The Glaxo Wellcome Register of Clinical Trials (www.gsk.com)

Before hand‐searching potential relevant journals, the Cochrane's Master List of Journals (http://www.cochrane.us/cochraneemainpage.asp) will be checked to determine which journals or conferences proceedings have already been hand‐searched within The Cochrane Collaboration.

The conference proceedings of the American Society of Haematology (1992 to 2006), the American Society of Clinical Oncology (ASCO) (1995 to 2006), and the European Society of Medical Oncology (ESMO) (1995 to 2006) will be hand‐searched. We will check the citations in retrieved papers, other meta‐analyses and narrative reviews for studies of potential relevance.

Searching for unpublished studies and grey literature will be performed according to the following information sources:

  • the System for Information on grey literature in Europe (SIGLE) ‐ database covering European grey literature since 1995;

  • index to Theses produced by British and Irish Universities Dissertation;

  • Dissertation Abstract (DA) Online‐database covering more than 1,000,000 doctoral dissertations and master theses, including:

‐ DAI ‐ Dissertation Abstract International,
‐ MAI ‐ Master Abstract International,
‐ ADD ‐ American Doctoral Dissertation,
‐ CDI ‐ Comprehensive Dissertation Index.

Manufacturers of the following commercially available 5‐HT3 RAs will be contacted for unpublished trials or eligible studies that have been missed for other reasons:

  • Roche,

  • Glaxo‐Smith‐Kline,

  • Novartis,

  • where deemed necessary, principal clinical investigators will be contacted for retrieving unpublished papers.

Data collection and analysis

Two review authors will work independently to identify potentially relevant trials from the records retrieved in the searches of the bibliographic databases. Using the full text of each study, two authors will independently classify trials for inclusion in the review as eligible or not eligible, according to the review's inclusion criteria. Controversies will be resolved by consensus between the two authors. Two review authors will independently assess the methodological quality of the included trials according to the following domains: concealment of allocation, double‐blinding, intention to treat analysis and loss to follow‐up.

Each study will also be assessed using the zero to five point scale described by Jadad 1996, as summarised below:

  1. was the study described as randomised? (1 = yes; 0 = no)

  2. was the study described as double‐blind? (1 = yes; 0 = no)

  3. were withdrawals and dropouts described? (1 = yes; 0 = no)

  4. was the method of randomisation well described and appropriate? (1 = yes; 0 = no); deduct one point if inappropriate

  5. was the double blinding well described and appropriate? (1 = yes; 0 = no); deduct one point if inappropriate.

Following validation of a pilot version of the data extraction form using a small samples of studies, two review authors will independently extract data from eligible trials. One review author will enter the data into The Cochrane Collaboration software program (Review Manager) and a second author will check these data.

Data extraction will include the following items:

  • participants: age range; gender (as a number and a percentage); performance status (number and percentage); type of cancer (number and percentage); setting of study; previous chemotherapy; type of chemotherapy; setting in‐patient or out‐patient;

  • intervention: antiemetic drugs (5‐HT3 RAs), concomitant corticosteroids (yes/no); dose and schedule of antiemetic drugs;

  • outcomes: definition of nausea and vomiting in the trial; proportion of patients with no acute vomiting or acute nausea, or delayed vomiting or delayed nausea; proportion of patients with complete absence of acute nausea and vomiting or complete absence of delayed nausea and vomiting. Where data are given on a daily basis, the rate for the worst day will be chosen. Proportion of patients with non severe adverse events, i.e. headache, constipation, diarrhoea, dizziness, ECG modifications. Proportion of patients with severe adverse events;

  • design: objective of the study; type of study (single‐ /multi‐centre, parallel/cross‐over, open/blind); description of randomisation; description of concealment of allocation; description of blinding of treatment; intention to treat analysis (yes/no); information of patients excluded after randomisation (yes/no); number of patients randomised; number of valuable patients; description of reasons for exclusions (yes/no); country where the study was performed; funding (source).

Statistical considerations

  • Where appropriate, trials will be combined using RevMan 4.2. An intention‐to‐treat analysis will be performed where possible.

  • For dichotomous variables, risk ratio (RR) with 95% confidence intervals (CI) will be calculated for individual studies. All similar studies will be pooled using fixed‐effect RR and 95% CIs.

  • For pooled effects, heterogeneity will be tested using Cochrane's Q homogeneity test (P < 0.05 will be considered statistically significant). The I2 test will also be used. Sources of heterogeneity will be investigated.

  • Results will also be expressed as numbers needed to treat to benefit (NNTs) for nausea and vomiting and numbers needed to treat to harm (NNHs) for mild and serious adverse drug reactions.

Sub‐group analyses ‐
Whenever possible heterogeneity will be examined using the following subgroups:
1) trials based on different 5‐HT3 RAs doses
2) trials based on different doses of cisplatin
3) trials based on cisplatin > 50 mg/m2 single or multiday chemotherapy regimens versus other chemotherapy regimens

Sensitivity analyses

Sensitivity analysis will be performed to examine the influence of the following:

  • methodological quality of the included trials;

  • random effects versus fixed effects model for meta‐analyses;

  • risk ratio and risk differences versus odds ratio;

  • ITT analysis with imputed missing data according to best case/worst case scenario for the different treatment groups in every trial;

  • exclusion of studies including no chemotherapy naïve patients;

  • exclusion of trials using discordant routes of administration for pairs of serotonin receptor antagonists;

  • exclusion of trials incorporating mild nausea in the outcome.