Review article

Therapeutic use of Cannabis sativa on

chemotherapy-induced nausea and vomiting among
cancer patients: systematic review and meta-analysis
F.C. MACHADO ROCHA, msc, phd student, Department of Psychiatry, Paulista School of Medicine, Federal
University of São Paulo, São Paulo, S.C. STÉFANO, phd, psychologist, Department of Psychiatry, Paulista
School of Medicine, Federal University of São Paulo, São Paulo, R. DE CÁSSIA HAIEK, msc, psychologist,
Department of Psychiatry, Paulista School of Medicine, Federal University of São Paulo, São Paulo,
L.M.Q. ROSA OLIVEIRA, student of medicine, Paulista School of Medicine, Federal University of São Paulo,
São Paulo, & D.X. DA SILVEIRA, phd, professor, Department of Psychiatry, Paulista School of Medicine,
Federal University of São Paulo, São Paulo, Brazil

MACHADO ROCHA F.C., STÉFANO S.C., DE CÁSSIA HAIEK R., ROSA OLIVEIRA L.M.Q. & DA SILVEIRA
D.X. (2008) European Journal of Cancer Care 17, 431–443
Therapeutic use of Cannabis sativa on chemotherapy-induced nausea and vomiting among cancer patients:
systematic review and meta-analysis

This paper aims to evaluate the anti-emetic efficacy of cannabinoids in cancer patients receiving chemotherapy
using a systematic review of literature searched within electronic databases such as PUBMED, EMBASE,
PSYCINFO, LILACS, and ‘The Cochrane Collaboration Controlled Trials Register’. Studies chosen were
randomized clinical trials comprising all publications of each database until December 2006. From 12 749
initially identified papers, 30 fulfilled the inclusion criteria for this review, with demonstration of superiority
of the anti-emetic efficacy of cannabinoids compared with conventional drugs and placebo. The adverse effects
were more intense and occurred more often among patients who used cannabinoids. Five meta-analyses were
carried out: (1) dronabinol versus placebo [n = 185; relative risk (RR) = 0.47; confidence interval (CI) = 0.19–
1.16]; (2) Dronabinol versus neuroleptics [n = 325; RR = 0.67; CI = 0.47–0.96; number needed to treat
(NNT) = 3.4]; (3) nabilone versus neuroleptics (n = 277; RR = 0.88; CI = 0.72–1.08); (4) levonantradol versus
neuroleptics (n = 194; RR = 0.94; CI = 0.75–1.18); and (5) patients’ preference for cannabis or other drugs
(n = 1138; RR = 0.33; CI = 0.24–0.44; NNT = 1.8). The superiority of the anti-emetic efficacy of cannabinoids
was demonstrated through meta-analysis.

Keywords: cancer, cannabis, chemotherapy, meta-analysis, randomized clinical trial, systematic review.

IN T R O D U C T I O N listed on the American pharmacopoeia until 1944

(Bonnie & Whitebread 1974), when it was removed due to
Marijuana has been used by throughout human
political pressure to ban its use in the US (Walsh et al.
history for many purposes (Karniol 2000). It was
2003).
Although marijuana has not returned to the American
Correspondence address: Francisco Carlos Machado Rocha, Rua Joaquim
de Almeida, número 81, Bairro: Mirandópolis, São Paulo, SP, CEP 04050- pharmacopoeia, in 1986 the Food and Drug Administra-
010, Brazil (e-mail: [email protected]). tion authorised the use of its active element, delta-9-
Accepted 2 November 2007 tetrahydrocannabinol (THC), for medical purposes (Walsh
DOI: 10.1111/j.1365-2354.2008.00917.x et al. 2003) to treat nausea and vomiting side effects in
European Journal of Cancer Care, 2008, 17, 431–443 patients receiving chemotherapy.

© 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
MACHADO ROCHA et al.

Cannabinoids interact with various neurotransmitters ME THODS

and neuromodulators, such as gamma-aminobutyric
Systematic review
acid (GABA), histamine, serotonin, dopamine, glutamate,
norepinephrine, prostaglandins and opioid peptides (Gro- Criteria for inclusion in this review
tenhermen 2002b).
Type of study All randomized clinical trials about the
Aside from these neurotransmitters, there seems to exist
subject published in the literature were objects for this
in the brain a ‘cannabinoid system’ (Karniol 2000). This
study.
system probably interacts with the classic neurotransmis-
sion systems to produce the pharmacological actions of Type of participant People with any type of cancer receiv-
Cannabis sativa (Pertwee 1990; Adams & Martin 1996; ing chemotherapeutic treatment, irrespective of gender,
Mallet & Beninger 1996; Felder & Glass 1998). The exist- age and place of treatment. The chemotherapeutic schemes
ence of a cannabinoid neurotransmission system in the included those of low, moderate and high emetic potential.
central nervous system, the function of which is still
unknown, opens a wide potential for the discovery of Type of intervention Pharmacological interventions
therapeutic drugs with action on it (Karniol 2000). based on substances derived from C. sativa and/or smoked
Nausea and vomiting, which can be ‘acute’, ‘retarded’ cannabis, irrespective of the time of intervention and of
or ‘anticipatory’ reactions (Fiori & Gralla 1984), are the the association with other types of therapy for nausea and
chemotherapy side effects considered by patients as the vomiting in cancer patients receiving chemotherapy.
most stressful (Barowski 1984). Up to three-fourths of all
cancer patients experience chemotherapy-related emesis Search strategy for study identification
(Schwartzberg 2007). Chemotherapy-induced nausea and
Searches were made on the electronic databases
vomiting also have the potential to cause depression,
MEDLINE (PUBMED), EMBASE, PSYCINFO, LILACS and
anxiety and a feeling of helplessness (Wilcox et al. 1982;
‘The Cochrane Collaboration Controlled Trials Register’.
Dodds 1985).
The bibliographic search strategy comprised the initial
Today, there are three synthetic cannabinoid drugs that
period of the databases until December 2006. The first
have been evaluated in clinical trials for the treatment of
authors of the selected studies were contacted, and the
nausea and vomiting in patients receiving chemotherapy:
bibliographies and references of these papers were also
delta-9-THC, nabilone and levonantradol (Walsh et al.
examined. There was no language restriction, but only
2003).
complete papers published in peer-reviewed journals were
Up until now, two medications, Marinol (dronabinol,
considered. Data related to other clinical settings (e.g.
delta-9-THC) and Cesamet (nabilone), have been approved
radiotherapy) were not considered.
to be prescribed for nausea and vomiting associated with
The search expression was based on the following
chemotherapy in cancer patients. Marinol has also been
Medical Subject Heading terms and categories: ‘therapeu-
approved for use in cases of anorexia and cachexia in AIDS
tics’, ‘drug therapy’, ‘chemical and pharmacologic
patients (Grotenhermen 2002a).
phenomena’, ‘neoplasms’, ‘antineoplastic and immuno-
This study describes a systematic research for evalua-
suppressive agents’, ‘marijuana abuse’, ‘Cannabis’, ‘ran-
tion of cannabis as a therapeutic agent for treating
domized controlled trials’, ‘double-blind method’, ‘single-
chemotherapy-induced nausea and vomiting in cancer
blind method’, ‘clinical trials’, ‘placebos’, ‘research
patients.
design’, ‘comparative study’, ‘evaluation studies’,
‘follow-up studies’, ‘prospective studies’ and ‘random allo-
cation’. Adjustments were made to the terms used accord-
O B JE C T I V E ing to the electronic database consulted.

This review aims to evaluate, through a systematic litera- Study description

ture review, interventions using C. sativa in the treat-
ment of nausea and vomiting in patients with any type of The characteristics of the included studies are shown in
cancer receiving chemotherapy, tested in randomized Table 1.
clinical trials and compared with any type of control
Methodological quality of the included studies
group.
The methodological quality evaluation of the clinical
studies is considered of vital importance for conducting

432 © 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Table 1 Characteristics of the included studies

Interventions and Chemotherapic Methodologic

Study Methods Participants therapeutic schedules Outcomes agents quality

Ahmedzai et al. Randomized, cross-over and Patients with lung cancer Nabilone 2 mg ¥ 2 (27) vs Nausea, vomiting, anorexia, adverse Cyclophosphamide, adriamycin, B
(1983) double-blind (small cell bronchial prochlorperazine 10 mg ¥ 3 (30) effects and preference etoposide, methotrexate, vincristin
carcinoma)
Chan et al. (1987) Randomized, cross-over and Children with various Nabilone 1–4 mg (30) vs prochlorperazine Vomiting, adverse effects and Doxorubicin, cyclophosphamide B
double-blind paediatric malignancies 5–20 mg (30) preference fluorouracil, methotrexate,

© 2008 The Authors

vincristin, etoposide; cisplatin
was not used
Chang et al. (1979) Randomized, double-blind and Patients with osteogenic Dronabinol 10 mg/m2 ¥ 4 (15) vs placebo Nausea, vomiting, food intake, adverse High dose of methotrexate B
paired (dronabinol-placebo or sarcoma (15) effects and THC plasma avaliation
placebo-dronabinol)
Chang et al. (1981) Randomized, double-blind and Patients with sarcoma Dronabinol 10 mg/m2 ¥ 4 (8) vs placebo Nausea, vomiting, adverse effects and Combination of doxorubucin and B
paired (dronabinol-placebo or (8) THC plasma avaliation cyclophosphamide
placebo-dronabinol)
Colls et al. (1980) Randomized, double-blind, Patients with solid Dronabinol 12 mg/m2 ¥ 2 (35) vs Nausea, vomiting and adverse effects Cyclophosphamide, mustine, and B
cross-over and multicentric tumours metoclopramide 4, 5 mg/m2 ¥ 1 others
intravenous (35) vs thiethylperazine 6,
6 mg/m2 ¥ 3 (35)
Crawford and Randomized, double-blind, Patients with Nabilone 1 mg ¥ 5 (37) vs metoclopramide Nausea, vomiting, adverse effects and Cisplatin, cyclophosphamide, B
Buckman (1986) cross-over adenocarcinoma of the 1 mg/kg ¥ 5 (39) preference adriamycin, methotrexate,
ovary or germ cell vincristin, bleomycin
tumours

Journal compilation © 2008 Blackwell Publishing Ltd

Dalzell et al. (1986) Randomized, double-blind, Children with various Nabilone 1–3 mg (18) vs domperidone Nausea, vomiting, adverse effects and Cisplatin, cyclophosphamide, B
cross-over tumours 15–45 mg (18) preference vincristin
Einhorn et al. (1981) Randomized, double-blind, Adults with various Nabilone 2 mg ¥ 4 (80) vs Nausea, vomiting, appetite, adverse Cisplatin, cyclophosphamide, B
cross-over tumours prochlorperazine 10 mg ¥ 4 (80) effects and preference bleomycin, vincristin
Fritak et al. (1979) Randomized, double-blind Patients with Dronabinol 15 mg ¥ 2 (38) vs Nausea, vomiting and adverse effects Vincristin, doxorubicin, A
and paralell gastrointestinal prochlorperazine 10 mg ¥ 2 (41) vs fluorouracil, and others
tumours placebo (37)
George et al. (1983) Randomized, double-blind, Women with advanced Nabilone 1 mg ¥ 3 (18) vs Nausea, vomiting, adverse effects Cisplatin, cyclophosphamide, B
double-placebo, and gynaecological cancer chlorpromazine 12,5 mg ¥ 1–2 and preference adriamycin
cross-over intramuscular (IM) (18)
Gralla et al. (1984) Randomized, double-blind Patients with various Dronabinol 10 mg/m2 ¥ 5 (15) Nausea, vomiting, adverse effects High dose cisplatin B
and paired (the patients were tumours vs metoclopramide 2 mg/kg ¥ 5 and ‘high’
exposed to both drugs) endovenous (15)
Herman et al. (1979) Randomized, double-blind and Patients with various Nabilone 2 mg ¥ 3–4 (113) vs Nausea, vomiting, adverse effects Cisplatin, cyclophosphamide, A
cross-over tumours prochlorperazine 10 mg ¥ 3–4 (113) and preference vinblastin, bleomycin, and
others
Hutcheon et al. Randomized, paralell and Patients with various Levonantradol 0,5 mg ¥ 3 IM (27) vs Nausea, vomiting, appetite and Cisplatin, cyclophosphamide, B
(1983) blind tumours levonantradol 0,75 mg ¥ 3 IM (28) vs adverse effects fluorouracil, vincristin
Levonantradol 1,0 mg ¥ 3 IM (26) vs
chlorpromazine 25 mg ¥ 3 IM (27)
a (0.5 mg) Randomized, paralell and Patients with various Levonantradol 0,5 mg ¥ 3 IM (27) vs Nausea, vomiting, appetite and Cisplatin, cyclophosphamide, B
blind tumours chlorpromazine 25 mg ¥ 3 IM (27) adverse effects fluorouracil, vincristin
b (0.75 mg) Randomized, paralell and Patients with various Levonantradol 0,75 mg ¥3 IM (28) vs Nausea, vomiting, appetite and Cisplatin, cyclophosphamide, B
blind tumours chlorpromazine 25 mg ¥ 3 (27) adverse effects fluorouracil, vincristin
Johansson et al. Randomized, double-blind and Patients with various Nabilone 2 mg ¥ 2 (18) vs prochlorperazine Nausea, vomiting, appetite, adverse Cisplatin (50 mg/m2), adriamycin B
(1982) cross-over tumours 10 mg ¥ 2 (18) effects and preference (40 mg/m2), cyclophosphamide
(500 mg/m2), vinblastin, vincristin
Jones et al. (1982) Randomized, double-blind Patients with various tumours Nabilone 2 mg ¥ 2 (24) vs placebo (24) Nausea, vomiting, adverse effects and Cisplatin, adriamycin, and others B
and cross-over (breast, lymphoma, ovary, preference
lung e others)
Kluin-Neleman et al. Randomized, double-blind Patients with Hodgkin Dronabinol 10 mg/m2 ¥2 (11) Nausea, vomiting, adverse effects and MOPP B
(1979) and cross-over and non Hodgkin vs placebo (11) THC plasma avaliation
lymphomas

433
Therapeutic use of Cannabis sativa on chemotherapy side effects
 434
Table 1 Continued

Interventions and Chemotherapic Methodologic

Study Methods Participants therapeutic schedules Outcomes agents quality

Lane et al. (1991) Randomized, double-blind, Patients with various Dronabinol 10 mg ¥4 (17) vs Nausea, vomiting and adverse effects Cyclophosphamide, doxorubicin, B
MACHADO ROCHA et al.

paralell and multicentric tumours (breast, prochlorperazine 10 mg ¥ 4 (20) vs 5-fluorouracil, vincristin,

colon, lymphoma, dronabinol 10 mg ¥ 4 mais etoposide
lung and others) prochlorperazine 10 mg ¥ 4 (17)
Levitt (1982) Randomized, double-blind and Patients with various Nabilone 2 mg ¥ 2 (36) vs placebo (36) Nausea, vomiting, appetite, adverse Cisplatin, adriamycin, B
cross-over tumours (lung, ovary, effects and preference cyclophosphamide, fluorouracil,
breast, and others) methotrexate, vincristin, and
others
McCabe et al. (1988) Randomized, comparative Patients with various Dronabinol 15 mg/m2 ¥6 (36) vs Nausea, vomiting, adverse effects and Cyclophosphamide, doxorubicin, B
and cross-over tumours prochlorperazine 10 mg ¥ 6 (36) preference fluorouracil, vincristin, and others
Neidhart et al. (1981) Randomized, double-blind and Patients with various Dronabinol 10 mg ¥ 4 (average) (37) vs Nausea, vomiting, adverse effects and Cisplatin, adriamycin, and others A
cross-over tumours haloperidol 2 mg ¥ 5 (average) (36) preference
Niederle et al. (1986) Randomized and cross-over Patients with Nabilone 2 mg ¥ 2 (20) vs alizapride Nausea, vomiting, appetite, adverse Low dose cisplatin and adriamycin B
nonseminomatous 150 mg ¥ 3 (20) effects and preference
testicular cancer
Niiranen and Randomized, double-blind and Patients with lung cancer Nabilone 1 mg ¥ 2 (24) vs prochlorperazine Nausea, vomiting, appetite, adverse Cisplatin, cyclophosphamide, A
Mattson (1985) cross-over 7,5 mg ¥ 2 (24) effects and preference adriamycin, vincristin,
vindesine e etoposide
Orr and McKernan Randomized, double-blind and Patients with various Dronabinol 7 mg/m2 ¥ 4 (55) vs Nausea, vomiting and adverse effects Cyclophosphamide, doxorubicin, B
(1980) cross-over tumours prochlorperazine 7 mg/m2 ¥ 4 (55) vs fluorouracil, and others
placebo (55)
Pomeroy et al. (1986) Randomized, double-blind and Patients with various Nabilone 1 mg ¥ 3 (19) vs domperidone Nausea, vomiting, appetite and Cisplatin in 70%, adriamycin in B
paralell tumours 20 mg ¥ 3 (19) adverse effects 19%, and others
Sallan et al. (1975) Randomized, double-blind and Patients with various Dronabinol 15 mg or 10 mg/m2 ¥ 3 (20) vs Nausea, vomiting, appetite and adverse Various agents B
paired (the patients were tumours placebo (20) effects
exposed to both drugs)
Sallan et al. (1980) Randomized, double-blind and Patients with various Dronabinol 15 mg or 10 mg/m2 ¥ 3 (73) vs Nausea, vomiting, appetite, development Cisplatin, cyclophosphamide, B
cross-over tumours prochlorperazine 10 mg ¥ 3 (73) of ‘high’ and preference methotrexate, and others
Sheidler et al. (1984) Randomized, double-blind and Patients with various kinds of Levonantradol 1,0 mg ¥ 3 IM (16) vs Nausea, vomiting, adverse effects and High doses cisplatin, A
cross-over cancer (solid tumours and prochlorperazine 10 mg ¥ 3 IM (16) preference cyclophosphamide and/or
haematologic malignancies) adriamycin
Steele et al. (1980) Randomized, double-blind and Patients with various Nabilone 2 mg ¥ 2 (37) vs prochlorperazine Nausea, vomiting, adverse effects and High doses cisplatin, low doses B
cross-over tumours 10 mg ¥ 2 (37) preference cisplatin, mechlorethamine,
streptozotocin, actinomycin D,
or DTIC
Ungerleider et al. Randomized, double-blind and Patients with various Dronabinol 7,5–12,5 mg ¥3 (181) vs Nausea, vomiting, Nausea, vomiting, Various agents: 66% with high A
(1982) paired (the patients were tumours (carcinomas, prochlorperazine 10 mg ¥ 3 (172) appetite, mood, anxiety, emetic potencial; 27% with
exposed to both drugs) sarcomas, lymphomas, and concentration, activity, interaction, moderate; and 7% with
others) adverse effects and preference low emetic potencial
Wada et al. (1982) Randomized, double-blind and Patients with various Nabilone 2 mg ¥ 2 (84) vs placebo (84) Nausea, vomiting, adverse effects and Cisplatin, adriamycin B
cross-over tumours preference

© 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Therapeutic use of Cannabis sativa on chemotherapy side effects

systematic reviews and for searching the best available types of chemotherapeutic agents. Many studies used
evidence of the therapeutic effect of an intervention. The some form of standard design (mostly the ‘cross-over’
critical evaluation of each clinical trial to be included in a ones); however, since the studies were reviewed over a
systematic review is vital to limit potential biases or sys- long period of time, there was considerable variation in
tematic errors, to help possible comparisons to be made, their designs.
and to serve as a guide to the interpretation of the final In many studies, the dose of anti-emetic medication was
findings (Mulrow & Oxman 1997). This is made through a adjusted during the research, either to increase its efficacy
careful analysis of the random distribution processes as or to reduce the side effects. There were also studies
well as through verification of how individuals who left wherein the protocol allowed the administration of an
the study before it ended were treated by the statistical anti-emetic other than the studied drugs to patients who
analysis (treatment intention analysis). In a meta-analysis, required them or who presented with unbearable nausea
the double-blind follow-up and the treatment intention and vomiting.
analysis are fundamental to reduce the so-called confusing Of the 30 studies included in the systematic review, 17
factors of results. were excluded from the meta-analysis on the anti-emetic
Considering the importance of the methodological efficacy due to a number of reasons (see Table 2).
quality evaluation of the included studies, the criteria Finally, it was possible to include in this meta-analysis
used were those described by Mulrow and Oxman (1997) data related to 13 randomized clinical trials on the use of
and the Jadad Scale (Jadad et al. 1996). Two independent cannabis for treating nausea and vomiting in cancer
reviewers evaluated the quality of the included studies patients receiving chemotherapy (total anti-emetic effi-
(FCMR and SCS). cacy). Eighteen clinical trials were included for the
The trials included in this review were rated as quality outcome ‘preference for one of the study drugs’.
A or B according to the randomization procedure of allo- All studies included in this meta-analysis, except three
cation concealment, following the Cochrane Collabora- (Niederle et al. 1986, which used alizapride; Hutcheon
tion Manual for methodological quality evaluation et al. 1983, which used chlorpromazine; and Dalzell
(Mulrow & Oxman 1997). They are as follows: et al. 1986, which used domperidone) compared cannabis
with prochlorperazine, a neuroleptic, as the control drug.
A ‘Low risk of bias’. Adequate allocation concealment
Thirty-one papers were excluded for failing to meet the
(e.g. central computer-generated randomisation).
study criteria (Appendix 1).
B ‘Moderate bias risk’. Unclear or doubtful allocation
The category shown in Figure 1 comprises two studies.
concealment.
In terms of anti-emetic efficacy, there was not a statisti-
C ‘High bias risk’. Inadequate allocation concealment
cally significant difference in favour of dronabinol [n =
(e.g. the use of alternate numbers, date of birth, etc.).
185; relative risk (RR) = 0.47; confidence interval (CI) =
0.19–1.16; P = 0.10].
R ES U L T S
The category shown in Figure 2 comprises five studies.
Using the search strategy, we identified 12 749 papers. In terms of anti-emetic efficacy, there was a statistically
Their titles were scanned to exclude papers that did not significant difference in favour of dronabinol [n = 325;
satisfy the objectives of this review. A total of 735 abstracts RR = 0.67; CI = 0.47–0.96; P = 0.03; number needed to
were evaluated in detail. Most of them were excluded treat (NNT) = 3.4].
because they did not satisfy the objectives of this review. The category shown in Figure 3 comprised six studies.
Finally, 96 complete papers were analysed. Thirty random- In terms of anti-emetic efficacy, there was not a statisti-
ized clinical trials using C. sativa to treat chemotherapy- cally significant difference in favour of nabilone (n = 277;
induced nausea and vomiting were identified. RR = 0.88; CI = 0.72–1.08; P = 0.21).
Most studies used a ‘cross-over’ design, although Fritak The category shown in Figure 4 comprised two studies.
et al. (1979), Hutcheon et al. (1983), Pomeroy et al. (1986) One of them allowed three comparisons: three different
and Lane et al. (1991) used ‘parallel’ design studies. doses of levonantradol (0.5 mg, 0.75 mg and 1.0 mg) were
In the individual studies, the size of most samples was compared with a neuroleptic (Hutcheon et al. 1983). In
small: 17 studies had fewer than 50 patients, seven terms of anti-emetic efficacy, there was not a statistically
studies had between 50 and 100 patients, and only six significant difference in favour of levonantradol (n = 194;
studies had 100 patients or more. In total, the studies RR = 0.94; CI = 0.75–1.18; P = 0.60).
included 1 719 patients who had different types of The category shown in Figure 5 comprised 18 studies. In
cancer, were of different ages and receiving different terms of preference for one of the drugs, there was a

© 2008 The Authors 435

Journal compilation © 2008 Blackwell Publishing Ltd
MACHADO ROCHA et al.

Table 2. Studies excluded from the meta-analysis (outcome: total anti-emetic efficacy)
Number of studies Reasons Articles
2 studies The studies exposed dichotomic data on the total anti-emetic Neidhart et al. (1981); Sallan et al. (1975)
efficacy via number of chemotherapy sequences, not
number of patients.
3 studies The studies exposed dichotomic data on the partial anti-emetic Jones et al. (1982); Levitt (1982); Wada et al.
efficacy, not total. (1982)
1 study The period analysed was 5 days, while the other meta-analysis Herman et al. (1979)
studies evaluated the acute anti-emetic efficacy during a
period of up to 24 h.
2 studies Did not present dichotomic data on the total anti-emetic Chang et al. (1979); *1981
efficacy;
Compared equal outcomes, the different variable being the
chemotherapeutic drug used.
9 studies Failed to present dichotomic data on the total anti-emetic Pomeroy et al. (1986); George et al. (1983);
efficacy. Colls (1980); Crawford and Buckman (1986);
Einhorn et al. (1981); Steele et al. (1980);
Gralla et al. (1984); Ungerleider et al. (1982);
Kluin-Neleman et al. (1979).

Review: Revisão Sistemática da Literatura Sobre o Uso Terapêutico da Cannabis no Tratamento dos Efeitos Colaterais de Náusea e V ômito em Pacientes com Câncer Submetidos à Quimioterapia
Comparison: 03 Eficácia: dronabinol (delta-9-THC) versus placebo
Outcome: 01 Pacientes que apresentaram náusea e/ou vômitos no período de até 24 horas depois da quimioterapia.

Study dronabinol (THC) placebo RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 95% CI

Frytak 1979 22/38 30/37 51.44 0.71 [0.52, 0.98]

Orr 1980 15/55 50/55 48.56 0.30 [0.19, 0.47]

Total (95% CI) 93 92 100.00 0.47 [0.19, 1.16]

Total events: 37 (dronabinol (THC)), 80 (placebo)
Test for heterogeneity: Chi² = 11.37, df = 1 (P = 0.0007), I² = 91.2%
Test for overall effect: Z = 1.63 (P = 0.10)

0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Figure 1. Dronabinol (delta-9-tetrahydrocannabinol) versus placebo.

Review: Revisão Sistemática da Literatura Sobre o Uso Terapêutico da Cannabis no Tratamento dos Efeitos Colaterais de Náusea e V ômito em Pacientes com Câncer Submetidos à Quimioterapia
Comparison: 04 Eficácia: dronabinol (delta-9-THC) versus neurolépticos (5 estudos com Prochlorperazine)
Outcome: 01 Pacientes que apresentaram náusea e/ou vômitos no período de até 24 horas depois da quimioterapia

Study dronabinol (THC) neurolépticos RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 95% CI

Frytak 1979 22/38 24/41 20.46 0.99 [0.68, 1.44]

Orr 1980 15/55 47/55 18.68 0.32 [0.20, 0.50]
Sallan 1980 9/15 11/12 18.63 0.65 [0.42, 1.02]
McCabe 1988 27/36 36/36 24.67 0.75 [0.62, 0.91]
Lane 1991 10/17 14/20 17.57 0.84 [0.51, 1.37]

Total (95% CI) 161 164 100.00 0.67 [0.47, 0.96]

Total events: 83 (dronabinol (THC)), 132 (neurolépticos)
Test for heterogeneity: Chi² = 18.65, df = 4 (P = 0.0009), I² = 78.6%
Test for overall effect: Z = 2.20 (P = 0.03)

0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Figure 2. Dronabinol (delta-9-tetrahydrocannabinol) versus neuroleptics.

statistically significant difference in favour of the Can- there is no systematic error (bias) due to paper omission
nabis components (n = 1138; RR = 0.33; CI = 0.24–0.44; generated by languages other than English, multiplicity
P < 0.00001; NNT = 1.8). of issues generated by a single study, poor methodology,
Figure 6 shows the ‘Funnel Plot’ of the risk difference inaccurate analysis or fraud. Also, the absence of perfect
versus the sample size. It shows some measure of symme- symmetry suggests clinical and methodological heteroge-
try and normal distribution (Gaussian), suggesting that neity inherent to the execution of the trials by different

436 © 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Therapeutic use of Cannabis sativa on chemotherapy side effects

Review: Revisão Sistemática da Literatura Sobre o Uso Terapêutico da Cannabis no Tratamento dos Efeitos Colaterais de Náusea e V ômito em Pacientes com Câncer Submetidos à Quimioterapia
Comparison: 02 Eficácia: Nabilone versus neurolépticos (4 estudos com Prochlorperazine, 1 com Alizapride e 1 com Domperido
Outcome: 01 Pacientes que apresentaram náusea e/ou vômitos no período de até 24 horas depois da quimioterapia

Study Nabilone Neurolépticos RR (random) Weight RR (random)

or sub-category n/N n/ N 95% CI % 95% CI

Johansson 1982 15/18 18/18 25.03 0.83 [0.68, 1.02]

Ahmedzai 1983 8/27 19/30 7.48 0.47 [0.25, 0.89]
Niiranen 1985 21/24 19/24 21.93 1.11 [0.86, 1.43]
Dalzell 1986 18/18 18/18 Not estimable
Niederle 1986 14/20 18/20 18.04 0.78 [0.56, 1.07]
Chan 1987 27/30 27/30 27.52 1.00 [0.84, 1.18]

Total (95% CI) 137 140 100.00 0.88 [0.72, 1.08]

Total events: 103 (Nabilone), 119 (Neurolépticos)
Test for heterogeneity: Chi² = 11.04, df = 4 (P = 0.03), I² = 63.8%
Test for overall effect: Z = 1.24 (P = 0.21)

0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Figure 3. Nabilone versus neuroleptics.

Review: Revisão Sistemática da Literatura Sobre o Uso Terapêutico da Cannabis no Tratamento dos Efeitos Colaterais de Náusea e V ômito em Pacientes com Câncer Submetidos à Quimioterapia
Comparison: 05 Eficácia: Levonantradol versus Neurolépticos (1 estudo com Prochlorperazine e 1 estudo com Chlorpromazine)
Outcome: 01 Pacientes que apresentaram náusea e/ou vômitos no período de até 24 horas depois da quimioterapia

Study Levonandradol Neurolépticos RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 95% CI

Hutcheon 1983 13/26 18/27 17.33 0.75 [0.47, 1.20]

Hutcheon a (0,5 mg) 13/27 18/27 17.02 0.72 [0.45, 1.16]
Hutcheon b (0,75 mg) 20/28 18/27 25.28 1.07 [0.75, 1.53]
Sheidler 1984 15/16 14/16 40.38 1.07 [0.86, 1.34]

Total (95% CI) 97 97 100.00 0.94 [0.75, 1.18]

Total events: 61 (Levonandradol), 68 (Neurolépticos)
Test for heterogeneity: Chi² = 4.78, df = 3 (P = 0.19), I² = 37.3%
Test for overall effect: Z = 0.52 (P = 0.60)

0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Figure 4. Levonantradol versus neuroleptics.

Review: Revisão Sistemática da Literatura Sobre o Uso Terapêutico da Cannabis no Tratamento dos Efeitos Colaterais de Náusea e Vômito em Pacientes com Câncer Submetidos à Quimioterapia
Comparison: 01 Preferência: Cannabis versus qualquer controle
Outcome: 01 Número de pacientes, por grupo de tratamento, que não preferiu Cannabis ou Controle

Study Cannabis Controle RR (random) Weight RR (random)

or sub-category n/N n/N 95% CI % 9 5 % CI

Herman 1979 18/103 85/103 8.10 0.21 [0.14, 0.33]

Sallan 1980 5/25 20/25 5.67 0.25 [0.11, 0.56]
Steele 1980 10/33 23/33 7.22 0.43 [0.25, 0.76]
Einhorn 1981 17/77 60/77 8.06 0.28 [0.18, 0.44]
Neidhart 1981 6/13 7/13 5.88 0.86 [0.40, 1.86]
Johansson 1982 3/16 13/16 4.42 0.23 [0.08, 0.66]
Jones 1982 2/18 16/18 3.35 0.13 [0.03, 0.47]
Levitt 1982 3/31 28/31 4.26 0.11 [0.04, 0.32]
Wada 1982 20/84 64/84 8.28 0.31 [0.21, 0.47]
Ahmedzai 1983 3/19 16/19 4.37 0.19 [0.07, 0.54]
George 1983 5/15 10/15 5.72 0.50 [0.22, 1.11]
Sheidler 1984 5/12 7/12 5.58 0.71 [0.31, 1.63]
Niiranen 1985 6/22 16/22 6.15 0.38 [0.18, 0.78]
Crawford 1986 10/22 12/22 7.01 0.83 [0.46, 1.51]
Dalzell 1986 1/13 12/13 1.98 0.08 [0.01, 0.55]
Niederle 1986 7/17 10/17 6.37 0.70 [0.35, 1.40]
Chan 1987 5/25 20/25 5.67 0.25 [0.11, 0.56]
McCabe 1988 1/24 23/24 1.93 0.04 [0.01, 0.30]

Total (95% CI) 569 569 100.00 0.33 [0.24, 0.44]

Total events: 127 (Cannabis), 442 (Controle)
Test for heterogeneity: Chi² = 48.64, df = 17 (P < 0.0001), I² = 65.0%
Test for overall effect: Z = 7.36 (P < 0.00001)

0.1 0.2 0.5 1 2 5 10

Favours treatment Favours control

Figure 5. Preference for cannabis or control.

© 2008 The Authors 437

Journal compilation © 2008 Blackwell Publishing Ltd
MACHADO ROCHA et al.

Preference who used cannabinoids than in those who used control

250 drugs.
The ‘intention to treat’ analysis was carried out for the
200 majority of the studies. The side-effects analysis included
the patients who abandoned the study before completion
Sample Size

150 and therefore were not evaluated for anti-emetic efficacy.

N It was not possible to establish a dose–response relation-
100 ship for there was insufficient data quality in the original
papers on this aspect. In some studies, the dose was
50 adjusted during the study itself, either to attain a possible
efficacy enhancement or to reduce the unwanted side
0 effects.
-1 -0,8 -0,6 -0,4 -0,2 0 The relationship described between the cannabinoid
Risk Difference plasma concentration and its therapeutic effect was not
Figure 6. ‘Funnel plot’ of the risk difference versus sample size. clearly examined in the studies. In one study, the anti-
emetic efficacy was related to the THC plasma concentra-
tion. In another study, there was no correlation between
authors, in different geographical regions and dealing with the THC plasma levels and the efficacy or the side effects.
methodological and clinical peculiarities. In terms of medication safety, the systematic review
showed that the cannabinoids were toxic for some
patients even when the drugs were given orally and their
D I S C U SSI O N
use restricted (for 24 h). Some side effects occurred almost
Using the statistical model of ‘random effect’, the meta- exclusively in patients who were exposed to the cannab-
analysis shows that dronabinol cannabinoid had a better inoid agents: 5% presented with paranoid delusions, 6%
acute anti-emetic efficacy (remission) than conventional presented with hallucinations, and almost 13% presented
anti-emetic drugs on cancer patients treated with poten- with dysphoria and/or depression. The number of patients
tially emesis-inducing chemotherapeutic agents. Cannab- who left the study due to the occurrence of cannabinoid
inoids nabilone and levonantradol did not have superior side effects is the main parameter of the eventual toxicity
acute anti-emetic efficacy when compared with the related to this substance.
conventional anti-emetics in the studies included in During this review, it was observed that although the
this meta-analysis (statistically significant difference). patients showed a higher number of collateral effects as
However, they had a clinically significant difference well as a higher symptom intensity during the treatment
towards the intervention. These results must be consid- with cannabinoids, most dropouts were not due to pos-
ered with caution due to the small number of studies and sible cannabinoid toxicity. These dropouts were respon-
the small patient sample of each study. sible for almost 30% of the nearly 400 dropouts in all the
Compared with placebo, dronabinol was not more effec- studies included. The other reasons for dropping out were
tive in the total remission of nausea and/or vomiting condition evolution, change of the chemotherapeutic
(random statistical model), although due to ethical con- strategy during the study, death due to cancer, protocol
siderations, placebo should not be used for patients receiv- violation, use of concomitant anti-emetic medication,
ing chemotherapy. inadequate data and low efficacy using two drugs.
In terms of partial improvement of nausea and/or vom- However, some side effects such as ‘high’ sensation,
iting, the random studies in this systematic review of data sleepiness, sedation and euphoria, which were more fre-
on the frequency of vomiting episodes and severity of quent when cannabinoids were used, would be potentially
nausea show that cannabinoids apparently had a better ‘beneficial’ for most patients; in other words, they would
anti-emetic efficacy than conventional drugs when used be pleasant during the chemotherapeutic treatment
in cancer patients who underwent chemotherapy using (Tramèr et al. 2001).
potentially emetic-inducing agents. However, the absence In the double-blind and cross-over studies included in
of data on the standard deviation in most studies makes this review, most patients preferred the cannabis-based
this an arguable conclusion. treatment when asked about their preferred drug. This
On the other hand, side effects (described below) preference was significant in relation to the control drugs
occurred more frequently and more intensely in patients (prochlorperazine, chlorpromazine, domperidone, halo-

438 © 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Therapeutic use of Cannabis sativa on chemotherapy side effects

peridol, alizapride, metoclopramide, placebo) used in the circumstances (Moore et al. 1998), and this result tendency
studies. From this preference, it can be hypothesized that may have been repeated in this analysis.
because nausea and vomiting during chemotherapy have Today, two anti-emetics that are prescribed demon-
such important impact and cause such discomfort to the strate good efficacy in reducing acute emesis: selective
patients, the patients prefer the cannabis side effects antagonists for 5-hydroxytryptamine (5-HT) receptor and
instead of the conventional anti-emetic medications that protachykinin (NK1) receptor. The latter retards vomiting
have lower efficacy. caused by chemotherapy with high emesis-inducing
In this review, it is important to consider some limita- potential (Olver 2004).
tions because some analyses potentially overrate the can- During the 1990s, 5-HT3 receptor antagonist combined
nabinoids’ efficacy and underrate their damage. with dexamethasone became the gold standard in acute
According to the Cochrane Manual, the studies were of emesis prophylaxis caused by chemotherapy (MASCC
acceptable quality. Of all selected studies, 70% presented a 1998). However, if there is failure to respond or there is an
proper mask method description. Most cross-over studies increase in emesis, this cannot be corrected by an increase
used a ‘double dummy’ design. The psychological effect of in dosage or frequency of administration. It seems that
smoking a C. sativa cigarette was not an analysed factor other receptor mechanisms may be involved (Tattersall
since cannabis was given via orally ingested capsules. et al. 1994; Herstedt 1996). Besides, in cases of delayed
However, the cannabinoids presented specific collateral emesis (from the second day onwards), the above combi-
effects, which were not presented by the control drugs, nation rarely obtains 50% of the desired effect (Kris et al.
and these factors presented a high incidence. In a study on 1985; Olver et al. 1996).
orally given nabilone, many patients identified which Nowadays, the anti-emetic indications for chemo-
drug they had received because of the collateral effects therapy with high emesis-inducing potential are 5-HT3
experienced. In a series of 100 blinded treatment interven- receptor antagonists, dexamethasone and aprepitant
tions using THC and placebo, the nurses identified the during the acute emetic phase, and aprepitant and dexam-
active drugs in 85% of the cases and the patients in 95% ethasone (for two more days) during the delayed emesis
of the cases (Tramèr et al. 2001). Such high values allow phase (Olver 2004). According to Walsh et al. (2003), can-
us to hypothesise that there was some bias on the part of nabinoids are fourth-line agents to be considered when
the observer in these studies. dealing with nausea and vomiting.
Some studies selected groups of patients who had not Some agents consider that when compared with modern
responded to the anti-emetic treatment with conven- anti-emetics, cannabinoids are ‘only’ modestly effective
tional drugs in previous chemotherapy cycles. This could and because of this more research on cannabinoids would
have introduced among the patients a bias in favour of be indispensable.
cannabis and against the drugs they knew they were However, cannabinoids seem to act through different
refractory to. mechanisms and can be effective for people who respond in
Some studies selected groups of patients with a previous an unsatisfactory way to the anti-emetic drugs used today.
history of smoking cannabis. In a study by Vinciguerra There are at least two types of cannabinoid receptors,
et al. (1988), it was claimed that young people with previ- CB1 and CB2, to which potent and selective antagonists
ous exposure to marijuana were predisposed to better anti- have been developed. The blockage of CB1 cannabinoid
emetic efficacy. However, it is not clear whether this receptors induces vomiting, suggesting the existence of an
factor alone was a bias. There are a few studies similar to endogenous cannabinoid system within the emetic cir-
this situation, and in one of them, the patients who had no cuits. This also suggests that the delta-9-THC anti-emetic
previous history of cannabis use demonstrated better effi- activity would be due to the stimulation of the CB1 recep-
cacy when compared with the other group (Ungerleider tor (Darmani 2001).
et al. 1982). Besides, delta-9-THC and its synthetic analogues (CP
The sample size can also be a source of criticism of the 55, 940 and WIN 55, 212–2) were able to prevent the
results of the studies. Of the 30 studies included in this inducing of this condition. However, it is not yet known
review, 13 had more than 50 patients included, and only six whether the cannabinoid receptor antagonist can override
had more than 100 patients. However, of the studies with or oppose the delta-9-THC capability of preventing vom-
numbers of patients available for analysis of anti-emetic iting caused by cisplatin chemotherapeutic agents.
efficacy, only nine had more than 50 patients analysed, and Ferrari et al. (1999)reported that a number of cannab-
only four included more than 100 patients. Small samples inoids (delta-9-THC, delta-8-THC, 7-hydroxy-delta-9-
have already shown an overrating of the effect under other THC, nabilone, HU 210) seem to be effective in

© 2008 The Authors 439

Journal compilation © 2008 Blackwell Publishing Ltd
MACHADO ROCHA et al.

preventing vomiting induced by cisplatin or apomorfin in 5 The number of dropouts from studies due to unbearable
cats and pigeons (Darmani 2001). collateral effects was significantly higher for patients
No clinical trials in humans comparing the action of who used cannabinoids. These dropouts were respon-
cannabinoids with modern anti-emetics for nausea and/or sible for approximately one-third of the dropouts for all
vomiting in cancer patients receiving chemotherapy were studies included in the systematic review.
found. 6 Most dropouts occurred due to other causes than the
Today, cannabinoids alone would not be used as first-line collateral effects of the cannabinoids.
medication for treating nausea and vomiting, but they 7 Patients showed a clear preference for cannabinoids as
seem promising because they have an auxiliary anti-emetic anti-emetic medication when receiving chemotherapy.
mechanism. The drugs’ efficacy can be higher when com- 8. Possible use of cannabinoids to treat chemotherapy-
bined with other anti-emetic drugs than when each drug is induced nausea and vomiting.
used alone. Because the cannabinoids’ mechanism is dif- 9. This study demonstrates the need for further work
ferent from other medications, they can benefit refractory to evaluate the use of cannabinoids and modern
patients or be used as auxiliaries to enhance the effect of anti-emetics.
existent anti-emetic medications if it is confirmed that the
synergy among cannabinoids, 5-HT3 receptor antagonist
P OTE NTI A L I NTE RE STS C ONFL I C T
and dexamethasone is similar to the synergy observed
among cannabinoids and prochlorperazine. This paper has no conflict of interests.
Thus, smaller doses of cannabinoids in combination
with modern anti-emetic medications might eventually
A C KNOWL E DGE ME NTS
not only enhance the anti-emetic efficacy, but also reduce
the cannabinoids’ collateral effects observed in this review. Coordination of Improvement of People of Higher Level
Recent findings on cannabinoids and endocannabinoids (CAPES) and Cochrane Center of Brazil.
receptors have opened a new era in research on their
physiological and pharmacological uses, as well as
research on the cannabinoids’ molecular bases (Walsh RE FE RE NC E S
et al. 2003). This information could help develop new Abrahamov A., Abrahamov A. & Mechoulam R. (1995) An effi-
synthetic cannabinoid antagonists for possible therapeutic cient new cannabinoid anti-emetic in pediatric oncology. Life
Sciences 56, 2097–2102.
use and could separate the desired effects from the
Adams I.B. & Martin B.R. (1996) Cannabis: pharmacology and
unwanted ones (Pertwee 1999a,b). toxicology in animals and humans. Addiction 91, 1585–1614.
Ahmedzai S., Carlyle D.L., Calder I.T. & Moran F. (1983) Anti-
CO N C L U SI O N S emetic efficacy and toxicity of nabilone, a synthetic cannab-
inoid, in lung cancer chemotherapy. British Journal of Cancer
48, 657–663.
1 The cannabinoid dronabinol had an anti-emetic effi- Anti-emetic Subcommitee of the Multinational Association of
cacy superior to neuroleptics for cancer patients receiv- Supportive Care in Cancer (MASCC) (1998) Prevention of
chemotherapy- and radiotherapy-induced emesis: results of the
ing chemotherapy.
Perugia Consensus Conference. Annals of Oncology 9, 1022–
2 Although there was not a statistically significant dif- 1029.
ference between the cannabinoid dronabinol and Barowski M.T. (1984) Advances in anti-emetic therapy. Cancer
placebo for cancer patients receiving chemotherapy, a Treatment Reviews 11, 237–256.
Bonnie R.J. & Whitebread C.H., I.I. (1974) The Marijuana Con-
clinically significant difference in favour of dronabinol viction: A History of Marijuana Prohibition in the United
was observed. States. University Press of Virginia, Charlottsville, VA, USA.
3 Although there was not a statistically significant dif- Chan H.S., Correia J.A. & MacLeod S.M. (1987) Nabilone versus
prochlorperazine for control of cancer chemotherapy-induced
ference between the cannabinoid nabilone and neuro- emesis in children: a double-blind, crossover trial. Pediatrics
leptics in cancer patients receiving chemotherapy, a 79, 946–952.
clinically significant difference in favour of nabilone Chang A.E., Shiling D.J., Stillman R.C., Goldberg N.H., Seipp
C.A., Barofsky I. et al. (1979) Delta-9-tetrahydrocannabinol as
was observed.
an antiemetic in cancer patients receiving high-dose methotr-
4 Although there was not a statistically significant dif- exate. A prospective, randomized evaluation. Annals of Inter-
ference between the cannabinoid levonantradol and nal Medicine 91, 819–824.
neuroleptics in cancer patients receiving chemo- Chang A.E., Shiling D.J., Stillman R.C., Goldberg N.H., Seipp
C.A., Barofsky I. et al. (1981) A prospective evaluation of delta-
therapy, a clinically significant difference in favour of 9-tetrahydrocannabinol as an antiemetic in patients receiving
levonantradol was observed. adriamycin and cytoxan chemotherapy. Cancer 47, 1746–1751.

440 © 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Therapeutic use of Cannabis sativa on chemotherapy side effects

Citron M.L., Herman T.S., Vreeland K., Krasnow S.H., Fossieck Felder C.C. & Glass M. (1998) Cannabinoid receptors and their
B.E., JrHarwood S., Franklin R. & Cohen M.H. (1985) endogenous agonists. Annual Review of Pharmacology and
Antiemetic efficacy of levonantradol compared to delta-9- Toxicology 38, 179–200.
tetrahydrocannabinol for chemotherapy-induced nausea and Ferrari F., Ottani A. & Giuliani D. (1999) Cannabimimetic activ-
vomiting. Cancer Treatment Report 69, 109–112. ity in rats and pigeons of HU 210, a potent antiemetic drug.
Colls B.M. (1980) Cannabis and cancer chemotherapy. Lancet i, Pharmacological, Biochemistry and Behavior 62, 75–80.
1187–1188. Fiori J.J. & Gralla R.J. (1984) Pharmacologic treatment of
Colls B.M., Ferry D.G., Gray A.J., Harvey V.J. & McQueen E.G. chemotherapy-induced nausea and vomiting. Cancer Investi-
(1980) The antiemetic activity of tetrahydrocannabinol versus gation 2, 351–361.
metoclopramide and thiethylperazine in patients undergoing Fritak S., Moertel C.G.O., Fallon J.R., Rubin J. & Creagan E.T.
cancer chemotherapy. The New Zealand Medical Journal 91, (1979) Delta-9-tetrahydrocannabinol as an antiemetic for
449–451. patients receiving cancer chemotherapy. A comparison with
Cone L.A., Greene D.S. & Helm N.A. (1982) Use of nabilone in prochlorperazine and a placebo. Annals of Internal Medicine
the treatment of chemotherapy-induced vomiting in an 91, 825–830.
outpatient setting. Cancer Treatment Reviews 9 (Suppl. B), Garb S., Beers A.L., Bograd M., McMahon R.T., Mangalik A.,
63–70. Ashmann A.C. & Lenine S. (1980) Two-pronged study of tet-
Crawford S.M. & Buckman R. (1986) Nabilone and metoclopra- rahydrocannabinol (THC) prevention of vomiting from cancer
mide in the treatment of nausea and vomiting due to cisplati- chemotherapy. IRCS Medical Science 8, 203–204.
nun: a double blind study. Medical Oncology and Tumor George M., Pejovic M.H., Thuaire M., Kramar A. & Wolff J.P.
Pharmacotherapy 3, 39–42. (1983) Randomised comparative trial of a new anti-emetic:
Cronin C.M., Sallan S.E., Gelber R., Lucas V.S. & Laszlo J. (1981) nabilone, in cancer patients treated with cisplatin. Biomedi-
Anti-emetic effect of intramuscular levonantradol in patients cine and Pharmacotherapy 37, 24–27.
receiving anticancer chemotherapy. Journal of Clinical Phar- Gilbert C.J., Ohly K.V., Rosner G. & Peters W.P. (1995) Ran-
macology 21, 43S–50S. domised, double-blind comparison of a prochlorperazine-based
Cunningham D., Forrest G.J., Soukop M., Gilchrest N.L. & versus a metoclopramide-based anti-emetic regimen in patients
Calder I.T. (1985) Nabilone and prochlorperazine: a useful com- undergoing autologous bone marrow transplantation. Cancer
bination for emesis induced by cytotoxic drugs. British Medical 76, 2330–2337.
Journal 291, 864–865. Gralla R.J., Tyson L.B., Bordin L.A., Clark R.A., Kelsen D.P., Kris
Cunningham D., Bradley C.J., Forrest G.J., Hutcheon A.W., M.G. et al. (1984) Anti-emetic therapy: a review of recent
Adams L., Sneddon M., Harding M., Herr D.J., Soukop M. & studies and a report of a random assignment trial comparing
Kaye S.B. (1988) A randomised trial of oral nabilone and metoclopramide with delta-9-tetrahydrocannabinol. Cancer
prochlorperazine compared to intravenous metoclopramide Treatment Report 68, 163–172.
and dexamethasone in the treatment of nausea and vomiting Grotenhermen F. (2002a) Review of therapeutic effects. In: Phar-
induced by chemotherapy regimens containing cisplatin or cis- macology, Toxicology and Therapeutic Potential (eds Groten-
platin analogues. European Journal of Cancer Clinical Oncol- hermen F. & Russo E.), pp. 55–65. The Haworth Press, New
ogy 24, 685–689. York, USA.
Dalzell A.M., Bartlett H. & Lilleyman J.S. (1986) Nabilone: an Grotenhermen F. (2002b) Effects of cannabis and the cannab-
alternative anti-emetic for cancer chemotherapy. Archives of inoids. In: Pharmacology, Toxicology and Therapeutic Poten-
Disease in Childhood 61, 502–505. tial (eds Grotenhermen F. & Russo E.), pp. 55–65. Xxxxxx, New
Darmani N.A. (2001) Delta(9)-tetrahydrocannabinol and syn- York, USA.
thetic cannabinoids prevent emesis produced by the cannab- Heim M.E., Romer W. & Queisser W. (1981) Clinical experience
inoid CB (1) receptor antagonist/inverse agonist SR 141716A. with levonantradol hydrochloride in the prevention of cancer
Neuropsychopharmacology 24, 198–203. chemotherapy-induced nausea and vomiting. Journal of Clini-
Devine M.L., Dow G.J., Greenberg B.R., Holsten D.W., Icaza L., cal Pharmacology 21, 86S–89S.
Jue P.Y., Meyers F.H., O’Brien E., Roberts C.M., Rocchio G.L., Herman T.S., Einhorn L.H., Jones S.E., Nagy C., Chester A.B.,
Stanton W. & Wesson D.R. (1987) Adverse reactions to delta- Dean J.C. et al. (1979) Superiority of nabilone over prochlorpera-
9-tetrahydrocannabinol given as an antiemetic in a multicenter zine as an antiemetic in patients receiving cancer chemo-
study. Clinical Pharmacology 6, 319–322. therapy. New England Journal of Medicine 300, 1295–1297.
Diasio R.B., Ettinger D.S. & Satterwhite B.E. (1981) Oral levo- Herstedt J. (1996) New perspective in antiemetic treatment.
nantradol in the treatment of chemotherapy-induced emesis: Support Care Cancer 4, 416–419.
preliminary observations. Journal of Clinical Pharmacology 21, Higi M., Niederle N., Bremer K., Schmitt G., Schmidt C.G. &
81S–5S. Seeber S. (1982) Levonantradol bei der Behandlung von
Dodds L.J. (1985) The control of cancer chemotherapy-induced zytostatika-bedingter Übelkeit und Erbrechen. Deutsche
nausea and vomiting. Journal of Clinical Hospital Pharmacol- Medizinische Wochenschrift 107, 1232–1234.
ogy 10, 143–166. Hutcheon A.W., Palmer J.B., Soukop M., Cunningham D.,
Einhorn L. (1982) Nabilone: an effective anti-emetic agent in McArdle C., Welsh J. et al. (1983) A randomised multicentre
patients receiving cancer chemotherapy. Cancer Treatment single blind comparison of a cannabinoid anti-emetic (levo-
Reviews 9, 55–61. nantradol) with chlorpromazine in patients receiving their first
Einhorn L.H., Nagy C., Furnas B. & Williams S.D. (1981) cytotoxic chemotherapy. European Journal of Cancer Clinical
Nabilone: an effective antiemetic in patients receiving cancer Oncology 19, 1087–1090.
chemotherapy. Journal of Clinical Pharmacology 21, 64– Jadad A., Moore A., Carrol D., Jenkinson C., Reynolds D.J.M.,
69S. Gavaghan D.J. et al. (1996) Assessing the quality of reports of
Ekert H., Waters D., Jurk I.H., Mobilia J. & Loughnan P. (1979) randomised clinical trials: is blinding necessary? Controlled
Amelioration of cancer chemotherapy-induced nausea and Clinical Trials 17, 1–12.
vomiting by delta-9-tetrahydrocannabinol. Medical Journal of Johansson R., Kikku P. & Groenroos M. (1982) A double-blind,
Australia 2, 657–659. controlled trial of nabilone vs prochlorperazine for refractory

© 2008 The Authors 441

Journal compilation © 2008 Blackwell Publishing Ltd
MACHADO ROCHA et al.

emesis induced by cancer chemotherapy. Cancer Treatment Niamatali C., Fallon S.D. & Egan E.L. (1984) Nabilone in the
Reviews 9, 25–33. management of prochlorperazine resistant cancer chemo-
Jones S.E., Durant J.R., Greco F.A. & Robertone A. (1982) A therapy induced emesis. Irish Medical Journal 77, 276–277.
multi-institutional phase III study of nabilone vs placebo in Niederle N., Schutte J. & Schmidt C.G. (1986) Crossover com-
chemotherapy-induced nausea and vomiting. Cancer Treat- parison of the antiemetic efficacy of nabilone and alizapride in
ment Reviews 9, 45–48. patients with nonseminomatous testicular cancer receiving
Karniol I.G. (2000) Cannabis sativa e derivados. In: Dependência cisplatin therapy. Klinische Wochenschrift 64, 362–365.
de Drogas, (eds Seibel S.D. & Toscano A. Jr), pp. 131–142. Niiranen A. & Mattson K. (1987) Antiemetic efficacy of nabilone
Atheneu, São Paulo, Brasil. and dexamethasone: a randomized study of patients with lung
Kenny J.B. & Wilkinson P.M. (1982) Levonantradol effectiveness cancer receiving chemotherapy. American Journal of Clinical
in cancer patients resistant to conventional anti-emetics. Clini- Oncology 10, 325–329.
cal Oncology 8, 335–339. Olver I. (2004) Two decades of progress in treating chemotherapy
Kluin-Neleman J.C., Neleman F.A., Meuwissen O.J.A.T. & Maes induced emesis. Cancer Forum 28, 140–143.
R.A.A. (1979) Delta-9-tetrahydrocannabinol (THC) as an anti- Olver I., Pasta W., Derpierre A. et al. (1996) A multicentre,
emetic for patients receiving cancer chemotherapy; a double double-blind study comparing placebo, ondansetron and
blind cross-over trial against placebo. Veterinary and Human ondansetron plus dexamethasone for the control of cisplatin-
Toxicology 21, 228–240. induced delayed emesis. Annals of Oncology 7, 945–952.
Kris M.G., Gralla R.J., Clark R.A. et al. (1985) Incidence, course, Orr L.E. & McKernan J.F. (1981) Antiemetic effect of
and severity of delayed nausea and vomiting following D9-tetrahydrocannabinol in chemotherapy-associated nausea
the administration of high-dose cisplatin. Journal of Clinical and emesis as compared to placebo and compazine. Journal of
Oncology 3, 1379–1384. Clinical Pharmacology 21, 76S–80S.
Lane M., Smith F.E., Sullivan R.A. & Plasse T.F. (1990) Dronab- Pertwee R.G. (1990) The central neuropharmacology of psycho-
inol and prochlorperazine alone and in combination as anti- tropic cannabinoids. In: Psychotropic Drugs of Abuse. (ed
emetic agents for cancer chemotherapy. American Journal of Balfour D.J.K.), pp. 355–429. Pergamon Press, New York, USA.
Clinical Oncology 13, 480–484. Pertwee R.G. (1999a) Evidence for the presence of CB1 cannab-
Lane M., Vogel C.L., Ferguson J., Krasnow S., Saiers J.L., Hamm J. inoid receptors on peripheral neurons and for the existence of
et al. (1991) Dronabinol and prochlorperazine in combination neuronal non-CB1 cannabinoid receptors. Life Sciences 65,
for treatment of cancer chemotherapy-induced nausea and vom- 597–605.
iting. Journal of Pain and Symptom Management 6, 352–359. Pertwee R.G. (1999b) Pharmacology of cannabinoid receptor
Laszlo J., Lucas V.S., Hanson D.C., Cronin C.M. & Sallan S.E. ligants. Current Medicinal Chemistry 6, 635–664.
(1981) Levonantradol for chemotherapy-induced emesis: phase Pomeroy M., Fennely J.J. & Towers M. (1986) Prospective ran-
I–II oral administration. Journal of Clinical Pharmacology 21, domised double-blind trial of nabilone versus domperidone in
51S–56S. the treatment of cytotoxic-induced emesis. Cancer Chemo-
Levitt M. (1982) Nabilone vs placebo in the treatment of therapy and Pharmacology 17, 285–288.
chemotherapy-induced nausea and vomiting in cancer patients. Rose M. (1980) Cannabis: a medical question? Lancet 1, 703.
Cancer Treatment Reviews 9, 49–53. Sallan S.E., Zinberg N.E., Frei E., 3rd (1975) Anti-emetic effect
Levitt M., Wilson W., Bowman D., Kemel S., Krepart G., Marks V., of delta-9-tetrahydrocannabinol in patients receiving cancer
Schipper H., Thomson G., Weinerman B. & Weinerman R. chemotherapy. New England Journal of Medicine 293, 795–
(1981) Physiologic observations in a controlled clinical trial 797.
of the antiemetic effectiveness of 5, 10, and 15 mg of D9- Sallan S.E., Cronin C., Zelen M. & Zinberg N.E. (1980) Anti-
tetrahydrocannabinol in cancer chemotherapy. Ophthalmologic emetics in patients receiving chemotherapy for cancer: a
implications. Journal of Clinical Pharmacology 21, 103S– 109S. randomized comparison of delta-9-tetrahydrocannabinol and
Lucas V.S. & Laszlo J. (1980) D9-tetrahydrocannabinol for refrac- prochlorperazine. New England Journal of Medicine 302, 135–
tory vomiting induced by cancer chemotherapy. Journal of the 138.
American Medical Association 243, 1241–1243. Schwartzberg L.S. (2007) Chemotherapy-induced nausea and
Lucraft H.H. & Palmer M.K. (1982) Randomised clinical trial vomiting: clinician and patient perspectives. Journal of Support
of levonantradol and chlorpromazine in the prevention of Oncology 5, 5–12.
radiotherapy-induced vomiting. Clinical Radiology 33, 621– Sheidler V.R., Ettinger D.S., Diasio R.B., Enterline J.P. & Brown
622. M.D. (1984) Double-blind multiple-dose cross-over study of the
Mallet P.E. & Beninger R.J. (1996) The endogenous cannabinoid antiemetic effect of intramuscular levonantradol compared to
receptor agonist anandamide impairs memory in rats. Behav- prochlorperazine. Journal of Clinical Pharmacology 24, 155–
ioral Pharmacology 7, 276–284. 159.
McCabe M., Smith F.P., Macdonald J.S., Woolley P.V. & Goldberg Stambaugh J.E., McAdams J. & Vreeland F. (1984) Dose ranging
D. (1988) Efficacy of tetrahydrocannabinol in patients refrac- evaluation of the antiemetic efficacy of intramuscular levo-
tory to standard antiemetic therapy. Investigational New nantradol in cancer subjects with chemotherapy-induced
Drugs 6, 243–246. emesis. Journal of Clinical Pharmacology 24, 480–485.
Moore R.A., Tramèr M.R., Carroll D., Wiffen P.J. & McQuay H.J. Steele N., Gralla R.J., Braun D.W., Jr & Young C.W. (1980)
(1998) Quantitative systematic review of topically-applied non- Double-blind comparison of the anti-emetic effects of nabilone
steroidal anti-inflamatory drugs. British Medical Journal 316, and prochlorperazine on chemotherapy-induced emesis.
333–338. Cancer Treatment Report 64, 219–224.
Mulrow C.S., Oxman A.D., eds. (1997) Cochrane Collaboration Sweet D.L., Miller N.J., Weddington W., Senay E. & Sushelsky L.
Handbook [updated September 1997]. In: The Cochrane (1981) D9-tetrahydrocannabinol as an antiemetic for patients
Library, Update Software, Oxford, UK. receiving cancer chemotherapy. A pilot study. Journal of Clini-
Neidhart J.A., Gagen M.M., Wilson H.E. & Young D.C. (1981) cal Pharmacology 21, 70S–75S.
Comparative trial of the antiemetic effects of THC and halo- Tattersall F.D., Rycroft W., Hill R.G. & Hargreaves R.J. (1994)
peridol. Journal of Clinical Pharmacology 21, 38–42S. Enatioselective inhibition of apomorphine-induced emesis in

442 © 2008 The Authors

Journal compilation © 2008 Blackwell Publishing Ltd
 Therapeutic use of Cannabis sativa on chemotherapy side effects

the ferret by the neurokinin1 receptor antagonist CP-99,994. Vinciguerra V., Moore T. & Brennan E. (1988) Inhalation mari-
Neuropharmacology 33, 259–260. juana as an antiemetic for cancer chemotherapy. New York
Tramèr M.R., Carroll D., Campbell F.A., Reynolds D.J.M., Moore State Journal of Medicine 88, 525–527.
R.A. & McQuay H.J. (2001) Cannabinoids for control of Wada J.K., Bogdon D.L., Gunnel J.C., Hum G.J., Gota C.H. &
chemotherapy-induced nausea and vomiting: quantitative sys- Rieth T.E. (1982) Double-blind, randomised, crossover trial of
tematic review. British Medical Journal 323, 1–8. nabilone vs. placebo in cancer chemotherapy. Cancer Treat-
Tyson L.B., Gralla R.J., Clark R.A., Kris M.G., Bordin L.A., Bosl ment Reviews 9, 39–44.
G.J. (1985) Phase 1 trial of levonantradol in chemotherapy- Walsh D., Nelson K.A. & Mahmoud F.A. (2003) Established and
induced emesis. American Journal of Clinical Oncology 8, potential therapeutic applications of cannabinoids in oncology.
528–532. Support Care Cancer 11, 137–143.
Ungerleider J.T., Andrysiak T., Fairbanks L., Goodnight J., Sarna Welsh J., Stuart F., Sangster G., Milstead R., Kaye S., Cash H.,
G. & Jamison K. (1982) Cannabis and cancer therapy: a com- Charlton D. & Calman K. (1983) Oral levonantradol in the
parison of delta-9-THC and prochlorperazine. Cancer 50, 636– control of cancer chemotherapy-induced emesis. Cancer Che-
645. motherapy and Pharmacology 11, 66–67.
Ungerleider J.T., Fairbanks L.A., Andrysiak T., Sarna G., Good- Wilcox P.M., Fetting J.H., Nettesheim K.M. et al. (1982) Antici-
night J. & Jamison K. (1985) THC or compazine for the cancer patory vomiting in women receiving cyclophosphamide, meth-
chemotherapy patient – the UCLA study. Part II: patient drug otrexate, and 5-FU (CMF) adjuvant chemotherapy for breast
preference. American Journal of Clinical Oncology 8, 142–147. carcinoma. Cancer Treatment Report 66, 1601–1604.

A PP E N D I X 1

Characteristics of the excluded studies:

• Six papers compared different doses of the same cannabinoid drug: Diasio et al. (1981); Laszlo et al. (1981); Higi et al.
(1982); Welsh et al. (1983); Stambaugh et al. (1984); and Tyson et al. (1985).
• One study compared two cannabinoid drugs: Citron et al. (1985).
• Twenty studies were excluded because they were not the object of the present review (non-randomized, open-label,
absence of comparison between cannabinoids and control drugs, correspondence, lack of relevant data, data on
physiological measurements only): Ekert et al. (1979), Colls (1980); Garb et al. (1980); Lucas and Laszlo (1980); Rose
(1980); Cronin et al. (1981); Heim et al. (1981); Levitt et al. (1981); Sweet et al. (1981); Cone et al. (1982); Kenny and
Wilkinson (1982); Lucraft and Palmer (1982); Niamatali et al. (1984); Cunningham et al. (1985, 1988); Devine et al.
(1987); Niiranen and Mattson (1987); Vinciguerra et al. (1988); Abrahamov et al. (1995);Gilbert et al. (1995).
• Four studies were duplicated completely or partially: Orr and McKernan (1981); Einhorn (1982); Ungerleider et al.
(1985); and Lane et al. (1990).

© 2008 The Authors 443

Journal compilation © 2008 Blackwell Publishing Ltd