The Use of Cannabinoids in Colitis: A Systematic Review and Meta-Analysis

Abstract

Background
Clinical trials investigating the use of cannabinoid drugs for the treatment of intestinal inflammation are anticipated secondary to preclinical literature demonstrating efficacy in reducing inflammation.

Methods
We systematically reviewed publications on the benefit of drugs targeting the endo-cannabinoid system in intestinal inflammation. We collated studies examining outcomes for meta-analysis from EMBASE, MEDLINE and Pubmed until March 2017. Quality was assessed according to mSTAIR and SRYCLE score.

Results
From 2008 papers, 51 publications examining the effect of cannabinoid compounds on murine colitis and 2 clinical studies were identified. Twenty-four compounds were assessed across 71 endpoints. Cannabidiol, a phytocannabinoid, was the most investigated drug. Macroscopic colitis severity (disease activity index [DAI]) and myeloperoxidase activity (MPO) were assessed throughout publications and were meta-analyzed using random effects models. Cannabinoids reduced DAI in comparison with the vehicle (standard mean difference [SMD] -1.36; 95% CI, -1.62 to-1.09; I² = 61%). FAAH inhibitor URB597 had the largest effect size (SMD -4.43; 95% CI, -6.32 to -2.55), followed by the synthetic drug AM1241 (SMD -3.11; 95% CI, -5.01 to -1.22) and the endocannabinoid anandamide (SMD -3.03; 95% CI, -4.89 to -1.17; I² not assessed). Cannabinoids reduced MPO in rodents compared to the vehicle; SMD -1.26; 95% CI, -1.54 to -0.97; I² = 48.1%. Cannabigerol had the largest effect size (SMD -6.20; 95% CI, -9.90 to -2.50), followed by the synthetic CB1 agonist ACEA (SMD -3.15; 95% CI, -4.75 to -1.55) and synthetic CB1/2 agonist WIN55,212-2 (SMD -1.74; 95% CI, -2.81 to -0.67; I² = 57%). We found no evidence of reporting bias. No significant difference was found between the prophylactic and therapeutic use of cannabinoid drugs.

Conclusions
There is abundant preclinical literature demonstrating the anti-inflammatory effects of cannabinoid drugs in inflammation of the gut. Larger randomised controlled-trials are warranted.

Full text

680
Inflamm Bowel Dis
• Volume 24, Number 4, April 2018
Review ARticle BAsic science
The Use of Cannabinoids in Colitis: A Systematic Review and
Meta-Analysis
Daniel G.Couch,* HenryMaudslay, BrettDoleman, PhD, Jonathan N.Lund, PhD, and
Saoirse E.O’Sullivan, PhD
Background: Clinical trials investigating the use of cannabinoid drugs for the treatment of intestinal inammation are anticipated secondary to
preclinical literature demonstrating efcacy in reducing inammation.
Methods: We systematically reviewed publications on the benet of drugs targeting the endo-cannabinoid system in intestinal inammation.
We collated studies examining outcomes for meta-analysis from EMBASE, MEDLINE and Pubmed until March 2017. Quality was assessed
according to mSTAIR and SRYCLE score.
Results: From 2008 papers, 51 publications examining the effect of cannabinoid compounds on murine colitis and 2 clinical studies were identi-
ed. Twenty-four compounds were assessed across 71 endpoints. Cannabidiol, a phytocannabinoid, was the most investigated drug. Macroscopic
colitis severity (disease activity index [DAI]) and myeloperoxidase activity (MPO) were assessed throughout publications and were meta-ana-
lyzed using random effects models. Cannabinoids reduced DAI in comparison with the vehicle (standard mean difference [SMD] -1.36; 95%
CI, -1.62 to-1.09; I2=61%). FAAH inhibitor URB597 had the largest effect size (SMD -4.43; 95% CI, -6.32 to -2.55), followed by the synthetic
drug AM1241 (SMD -3.11; 95% CI, -5.01 to -1.22) and the endocannabinoid anandamide (SMD -3.03; 95% CI, -4.89 to -1.17; I2 not assessed).
Cannabinoids reduced MPO in rodents compared to the vehicle; SMD -1.26; 95% CI, -1.54 to -0.97; I2=48.1%. Cannabigerol had the largest
effect size (SMD -6.20; 95% CI, -9.90 to -2.50), followed by the synthetic CB1 agonist ACEA (SMD -3.15; 95% CI, -4.75 to -1.55) and synthetic
CB1/2 agonist WIN55,212-2 (SMD -1.74; 95% CI, -2.81 to -0.67; I2=57%). We found no evidence of reporting bias. No signicant difference was
found between the prophylactic and therapeutic use of cannabinoid drugs.
Conclusions: There is abundant preclinical literature demonstrating the anti-inammatory effects of cannabinoid drugs in inammation of the
gut. Larger randomised controlled-trials are warranted.
Key words: cannabinoid, inflammation, gut, intestine, colitis
INTRODUCTION
Inammatory bowel disease (IBD) affects 200 per
100,000 adults in the United States and 400 per 100,000 in the
United Kingdom.1, 2 Major subtypes consist of Crohn’s dis-
ease and ulcerative colitis. A denitive clinical treatment for
these chronic relapsing diseases remains elusive, as currently
no therapy exists to reverse the clinical pathology without a
risk of signicant side effects. Corticosteroids, 5-ASA agents,
anti-TNFα antibodies, and other immunomodulatory drugs
have all been shown to induce signicant remission in IBD,
but are associated with bone marrow suppression, opportunis-
tic infection, infusion reactions, and malignancy secondary to
immunosuppression.3–5
The endocannabinoid system (ECS), consisting of mul-
tiple receptors and endogenous ligands, controls multiple
homeostatic processes including gastrointestinal motility, hun-
ger, perception of pain, and immunity.6–10 The targets of the
ECS consist of the classical CB1 and CB2 receptors, but also
the orphan GPR55 receptor, peroxisome proliferator-activated
receptors (PPARs), and transient receptor potential vanilloid
(TRPV) receptors. These targets are all found on the cells of
gut mucosa, submucosa, the enteric nervous system and the
immune system. Endocannabinoids, such as anandamide
(AEA) and 2-arachiodoylglycerol (2-AG), are intercellular
Received for publications October 12, 2017; Editorial Decision November 27,
2017.
School of Medicine, Royal Derby Hospital, University of Nottingham, Derby,
DE22 3DT United Kingdom.
© 2018 Crohn’s & Colitis Foundation. Published by Oxford University Press
on behalf of Crohn’s & Colitis Foundation.
Author Contributions:
DC, JL, and SO conceived and designed the study. DC and HM collected data.
DC, HM, BD, JL, and SO analyzed data. DC, JL, and SO were responsible for over-
all content of the article.
All authors drafted, revised, and approved the manuscript. The authors have no
conicts of interest to report. No funding was recived for this study.
*Address correspondence to: Daniel G.Couch, MB, ChB, School of Medicine,
Royal Derby Hospital, University of Nottingham, Derby. DE22 3DT United
Kingodm . E-mail: Couch27@gmail.com.
Abbreviations: 2-AG, 2-arachidonoyl glycerol; Ab-CBD, Abnormal can-
nabidiol; AEA, Anandamide; CBD, Cannabidiol; CBG, Cannabigerol; CBN,
Cannabinol; CI, Condence interval; CO, Croton oil; DAI, Disease activity
index; DNBS, Dinitrobenzene sulphonic acid; DSS, Dextran sulphate sodium; IC,
Intracolonic; IL-10, Interleukin-10; IV, Intravenous; MMJ, Medicinal cannabis;
MPO, Myeloperoxidase; OM, Oil of mustard; PEA, Palmitoylethanolamide; PO,
Oral; PPAR, Peroxisome Proliferator Activating Receptor; PR, Per rectum; SC,
Subcutaneous; SMD, Standard mean difference; THC, Δ9-Tetrahydrocannabinol;
TNBS, Trinitrobenzene sulphonic acid; TRPV1, Transient receptor potential vanil-
loid 1
doi: 10.1093/ibd/izy014
Published online 19 March 2018
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

681
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
lipid-signalling molecules derived on demand from membrane
precursors.11 They are metabolised by fatty acid amide hydro-
lase (FAAH), as well as N-acyl ethanolamine-hydrolysing acid
amidase (NAAA) in the case of AEA, and monoacylglycerol
lipase (MAGL) in the case of 2-AG.12–14 Palmitoylethanolamide
(PEA), also metabolised by NAAA, has been shown to activate
PPARα and may increase local concentrations of AEA or the
afnity of AEA to the CB1 receptor and, therefore, is included
as an atypical cannabinoid.15, 16 Phytocannabinoids include Δ-9
tetrahydrocannabinol (THC), cannabidiol (CBD), cannab-
igerol (CBG), cannibichromene (CBC), and up to 60 others and
are isolated from Cannabis Sativa.11 THC and CBD have found
place in clinical practice in the treatment of childhood epilepsy
and muscular spasticity in multiple sclerosis.17, 18 A growing
collection of synthetic cannabinoid agonists have been derived
possessing selective high afnity for the CB1, CB2, GPR55 and
TRPV1 receptors, and have been investigated pre-clinically for
roles in gut motility, satiety and immunity.8
Under inammatory conditions CB1, CB2, and both
PPARα and PPARγ expressions increase on the submucosa
and on adjacent immune cells, whereas GPR55 and TRPV1
expression decreases on the mucosa, but increases on enteric
nervous tissue.19–21 Levels of AEA, 2-AG, and PEA are upreg-
ulated in vitro, and also in animal in vivo and human ex-vivo
models of intestinal inammation.22–24 Early experimentation
in murine models demonstrated that cannabinoids prevent the
onset of experimental murine colitis or reduced its severity.25
Since these initial ndings, many reports—including clinical tri-
als—have now investigated the effect of cannabinoid ligands, or
the effect of blockade of their metabolising enzymes, on inam-
mation of thegut.
There is a signicant amount of promising preclinical
evidence for the use of cannabinoid agents in the treatment of
colitis. Within this study we aimed to gather all preclinical and
clinical evidence for the use of these drugs in colitis, and where
possible, perform meta-analyses across studies to assess the ef-
cacy of cannabinoids for further clinical trials. Where possible,
clinically relevant experimental endpoints were assessed.
METHODS
Search Strategy
All studies evaluating the effect of cannabinoid drugs
on inammation of the colon were searched from March
1980 to March 2017 by 2 independent researchers in Medline,
EMBASE, and Pubmed. Keywords included cannabidiol, tet-
rahydrocannabinol, anandamide, 2-AG, cannibichromene, can-
nabigerol, cannabinoid, cannabis sativa, colon, intestine, gut,
inammation, Crohn’s, ulcerative, and colitis. Names of syn-
thetic cannabinoid agents were also included. References from
included studies were searched by hand. Prespecied inclusion
and exclusion criteria were used to prevent bias. Experiments
must have been performed in the context of administration
of cannabinoid drugs to inammatory states of the colon in
humans or animals, either experimental or due to endogenous
disease (Crohn’s disease or ulcerative colitis). In vitro studies,
studies not examining the effect of cannabinoids in intestinal
inammation specically, or studies using cannabinoid antago-
nists as a primary agent were excluded. APRISMA checklist is
included in the appendix.
Data Acquisition
The mode of colitis induction in preclinical studies was
recorded in addition to the timing of cannabinoid application.
For the purposes of meta-analysis, data on the macroscopic or
histological disease scores (as listed in the disease activity index
[DAI]) and myeloperoxidase (MPO) activity were collected. If
the exact number of animals was not available, the lowest num-
ber of animals within the range given was used for the experi-
mental group, and the highest number used for the control/
vehicle group. Where studies reported the effects of more than
1 cannabinoid sharing a single control group for comparison,
control group numbers were equally distributed between com-
parisons to avoid unit of analysis issues. WebPlotDigitiser (ver-
sion 3.11) was used to extract values from gures in published
articles where no data values were given in the text.
Quality
Quality of included studies were assessed by 2 independ-
ent researchers to quantify risk of bias according to the 6-point
criteria developed by the Cochrane Collaboration risk of bias
tool.26 To assess the quality of preclinical studies, the STAIR
and Arrive preclinical assessment tools were adapted.27, 28 Each
of these items was awarded 1 point: randomization, assessor
blinding, results replicated in a second species, dose-response
experiments, results replicated in a second model of colitis,
n=5 or greater in each group, the use of clinically relevant end-
point to assess response of colitis, a denitive statement of ani-
mal numbers in each group, a statement regarding the housing
of animals, and a statement describing the location and timing
of animal experimentation (i.e. in animal housing or a separate
cage, time of day etc). The highest possible score was 10 points.
Data Analysis
Where possible, data were grouped into DAI and MPO
activity, and subdivided by species and compound. Data from
each group were analysed as forest plots using Cochrane Review
Manager Software (Review Manager 5.3, Copenhagen: The
Nordic Cochrane Centre, The Cochrane Collaboration, 2014),
and as funnel plots using Stat (Stat Corp.2009 Stat Statistical
Software: Release 11. College Station, TX, USA). Funnel plot
asymmetry was tested using the Egger linear regression test.
AP value of <0.05 was considered statistically signicant. As
differing studies measured MPO activity and DAI using var-
ious scales, we present effect estimates as standardized mean
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

682
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
differences (SMD) with 95% condence intervals (CI). We used
the following SMD values to assess results for clinical signif-
icance: <-0.5 small clinical signicance, -0.5 to -0.8 moderate
clinical signicance, and >-0.8 high clinical signicance. Due to
clinical heterogeneity between the various studies, a random-ef-
fects model was used. We assessed statistical heterogeneity
using the I2 statistic, with >50% regarded as evidence of statis-
tical heterogeneity. We assessed the quality of evidence using
the previously validated SYRCLE criteria, with studies graded
out of 10.29 Studies were weighted by sample size and statistical
signicance was set at a minimum of P<0.05.
RESULTS
Search Results and Study Characteristics
The search strategy returned 2008 results from which 199
relevant publications were identied. From these, 53 publica-
tions comprising 106 experiments examining 35 compounds
met the inclusion criteria (Fig.1, Tables 1 and 2). Thirty-four
studies were included in the meta-analysis.
Forty-three publications studied the effects of cannabi-
noids on experimental murine colitis, 5 in rats, and 3 in both
mice and rats. Two clinical trials examined the effect of a can-
nabinoid (THC and CBD) in Crohn’s disease. Within animal
publications, 43 used caustic agents (Di-nitrobenzine sulphonic
acid (DNBS), trinitrobenzene sulphonic acid (TNBS), oil of
mustard (OM), dextran sulphate sodium (DSS), and croton oil
(CO)) to induce colitis; 6 used intravenous or topical lipopol-
ysaccharide; 2 induced colonic inammation using surgical
arterial ligation or puncture of the colon; and 1 induced colitis
with interleukin-10 (IL-10) knock-down and DSS (Fig. 2A).
Across all publications, including clinical trials, 71 endpoints
were examined to evaluate the effect of cannabinoid drugs
on colitis. Forty-nine publications (89 experiments) examined
more than 1 endpoint. Of these endpoints, MPO and DAI were
the most consistently used (34 and 26 studies, respectively) and
were therefore selected for meta-analysis. Incidence of end-
points is given in Fig.2B.
The effect of 7 phytocannabinoids were studied across
18 publications; cannabinol (CBN), CBD, THC, CBC, CBG,
medicinal cannabis (MMJ), and abnormal CBD (Ab-CBD).
Four endocannabinoids were studied across 11 publica-
tions (PEA, ultramicronized PEA [uPEA], Arachidonyl-
2’-chloroethylamide [ACEA], and AEA); 15 synthetic
cannabinoid agonists were studied across 22 publications
(AM841, Adelmidrol, HU210, CP55940, WIN55,212-2,
AM1241, JHW015, JWH133, βCaryophyllene, O-1602,
HU308, αβ amyrin, CID16020046, compound 26, and
SAB378); and 9 compounds targeting the catabolism or trans-
port of endogenous cannabinoids were studied across 13 publi-
cations (ARN2508, PF-3845, compound 39, JZL184, AA5HT,
VDM11, URB597, AM9053, and AM3506). These compounds
are delineated by classin Table1. The degrees of positivity or
negativity of the outcomes of these studies are displayed in
Fig. 2C. Twenty-three studies investigated underlying recep-
tor mechanisms using knock-out (KO) animals or receptor
antagonists.
Of the 105 experiments comparing cannabinoids with the
vehicle or placebo, 67 (63.8%) favored cannabinoids, 34 (32.3%)
reported no difference, and 4 (3.8%) favored vehicle. Mice were
used in 89 experiments (68.5% of which favored cannabinoids),
while rats were used in 14 (71.4% favored cannabinoids). In 4
experiments, both mice and rats were used showing no differ-
ence between cannabinoids and the vehicle. In the 2 clinical
trials, no difference in primary outcome was found between
the use of THC cigarettes or oral CBD and placebo. Eleven
of 14 publications (78.6%) using synthetic CB2 receptor ago-
nists favored cannabinoid use over the vehicle, and a further 11
of 13 (84.6%) favored using FAAH inhibitors over the vehicle.
The outcome of all cannabinoids across publications is given
in Fig.2C.
Two clinical trials examining the effect of CBD and THC
in Crohn’s disease were found. Naftali etal (2013) conducted a
placebo-controlled study in Crohn’s disease patients, comparing
THC 115mg inhaled alone with placebo. Disease activity was
compared between groups by means of a validated question-
naire (Crohn’s disease activity index [CDAI]) after 8 weeks of
treatment. Anonsignicant reduction in clinical disease remis-
sion as dened by the authors was found at the end of the study
period; however, a secondary endpoint of reduction in overall
activity scores was found between groups (P=0.028). In a sec-
ond study, Naftali et al (2017) compared 10 mg of oral CBD
twice daily with placebo in Crohn’s disease, using CDAI in an
identical fashion. No reduction in disease activity was detected
Figure1. Record identication process.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

683
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
between groups. In both studies, the authors measured changes
in serum C-reactive protein (CRP). Within both experimental
and placebo groups, CRP levels were below 5 units per milliliter
at the end of the study periods. Clinically, CRP levels greater
than 5 units per milliliter are considered indicative of inam-
matory disease. Within both studies, the combination of CBD
and THC within a single study were not assessed.
Of the 104 experiments where timing of drug administra-
tion was stated, 37 administered cannabinoids therapeutically,
of which 62.2% favored cannabinoid treatment. Nineteen experi-
ments administered cannabinoids prophylactically, of which 52.6%
favored cannabinoid treatment. Forty-eight experiments adminis-
tered cannabinoids both prophylactically and therapeutically, of
which 75% favored cannabinoid treatment versus the vehicle.
Meta-analysis
Thirty-four studies reported the same endpoints of dis-
ease activity index or myeloperoxidase activity, allowing for
meta-analysis. Of the remaining studies, heterogeneity of end-
points prevented further meta-analysis.
Crohn’s Disease Activity Index(CDAI)
The use of 2 phytocannabinoids, THC or CBD, in 2
human studies were meta-analysed. Phytocannabinoid use
decreased severity scores in comparison with placebo (mean
difference [MD] -74.97; 95% CI, -229 to 0.79, I2=75% Fig.3).
THC alone had a signicant effect on reducing CDAI (MD
-154.00; 95% CI, -2.68.57 to -44.43), whereas CBD alone did
not (MD +4.00;, 95% CI -1.5.39 to +113.39).
TABLE1: Cannabinoid Drugs Found by Search Strategy
Cannabinoid Class Drug Description
Endocannabinoids AEA Anandamide
PEA Palmitoylethanolamide
uPEA Ultramicronised PEA
Phytocannabinoids Cannabis sativa Multiple compounds
CBC Cannibichromene
CBD Cannabidiol
CBG Cannabigerol
CBN Cannabinol
THC Tetrahydrocannabinol
Cannabinomimetics αβ Amyrin CB1 and CB2 agonist
ACEA Arachidonyl-2’-chloroethylamide
Adelmidrol PEA analogue
AM1241 CB2 full agonist, partial CB1 agonist
AM841 Peripherally restricted CB1 agonist
βCaryophyllene CB2 agonist
CID16020046 GPR55 inverse agonist
Compound 26 CB2 agonist
CP55,940 CB1 and CB2 agonist
HU210 THC analogue
HU308 CB2 agonist
JWH015 CB2 full agonist, weak CB1 agonist
JHW133 CB2 full agonist, weak CB1 agonist
O-1602 GPR18 and GPR55 agonist
SAB378 Peripherally restricted CB1 and CB2 agonist
WIN55,212-2 CB1 full agonist
Enzyme Inhibitors AA5HT FAAH inhibitor
AM3506 FAAH inhibitor
AM9053 NAAA inhibitor
ARN2508 FAAH inhibitor
compound 39 FAAH inhibitor
JZL184 MAGL inhibitor
PF-3845 FAAH inhibitor
URB597 FAAH inhibitor
Reuptake inhibitors VDM11 AEA reuptake inhibitor
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

684
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
(Continued )
TABLE2: Characteristics of Studies Included for Systematic Review
Study Species Model Compound Route/dosage
Time of Administration
Versus Inammation
Time of
Assessment Post
Inammation
Modied STAIR
score SRCYCLE Score
Capasso 200132 ICR mice CO PEA i.p 2.5–30mg/kg 20 minutes pre 4days 4 1
Izzo 20019ICR mice CO CP 55,940
Cannabinol
i.p.0.03–10nmol/m 4days post 20 minutes 3 0
i.p.10–3000nmol/m
Massa 200425 C57BL/6N mice DNBS SR141716 i.p.3mg/kg Pre, 24 and 48 hours post 3& 7days 4 2
HU210 i.p.0.05mg/kg
Mathison 200450 Spr-Dawley rats LPS ACEA i.p.1mg/kg 70 minutes post 120 minutes 5 0
JWH133 i.p.1mg/kg
D’Argenio 200622 C57/BJ mice
Wistar rats
DNBS VDM11 SC 5mg/kg Post 3& 7days 6 0
TNBS AA-5-HT SC 10mg/kg
Kimball 200651 CD-1 mice OM ACEA i.p.10mg/kg 24 hours pre 3days 3 1
JWH133 i.p.2.5mg/kg
Capasso 200852 ICR mice CO CBD i.p.5mg/kg 20 minutes pre Ach 4days 5 0
JWH015 i.p.10mg/kg
Engel 200853 AKR mice TNBS AEA i.p.5mg/kg 30 minutes pre 3days 3 1
Storr 200854 C57/BL mice TNBS URB597 i.p.5mg/kg 30 minutes pre or 24 hours
post
3days 4 1
VDM11 i.p.5mg/kg
Borelli 200946 ICR mice DNBS CBD i.p.1, 2, 5, 10mg/kg 24 hours post 3days 3 0
Li 200955 Rats Mice LPS HU210 100μg.kg 5 minutes 30 minutes 8 1
JWH133
AM630 100μg.kg
AM251 3mg/kg
Storr 200956 C57/BL mice TNBS JWH133 i.p.20mg/kg 30 minutes pre or 24 hours
post
1, 3, 5, 7days 7 1
DSS AM1241 i.p.10–20mg/kg
AM630 i.p.10mg/kg
Cassol Jr 201047 Wistar rats CLP CBD i.p.2.5, 4, 10mg/kg Simultaneous 9days 8 2
Cluny 201057 C57/BL mice DSS SAB378 i.p 0.1 or 1.0mg/kg 4days post 8days 5 1
TNBS AM251 i.p 1.0mg/kg
AM630 i.p 1.0mg/kg
WIN55,212-2 i.p 1, 2mg/kg
Kimball 201058 CD1 mice OM ACEA i.p.1mg/kg 30 minutes pre 28days 4 3
JWH133
i.p.1mg/kg
Jamontt 201045 Wistar rats TNBS THC i.p.5–20mg/kg 30 minutes pre 3days 5 1
CBD
i.p.5–20mg/kg
Alhouayek 201159 C57BL/6 mice TNBS JZL184 i.p.16mg/kg Pre onset 3days 2 1
Andrejak 201160 C57/BL mice TNBS Compound 39 i.p.5mg/kg 3days pre 3days 6 1
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

685
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
Study Species Model Compound Route/dosage
Time of Administration
Versus Inammation
Time of
Assessment Post
Inammation
Modied STAIR
score SRCYCLE Score
Bento 201161 CD1 mice DSS βCaryophyllene i.p.12.5, 25, 50mg/kg 3 -7days post 7days 4 1
Delipis 201149 OF1 mice LPS CBD i.p.10mg/kg 6 hours post 120 minutes 6 1
Lin 201143 C57/BL mice LPS CBD O-1602 i.p.10mg/kg 30 minutes pre 20 minutes 5 1
Spr-Dawley rats I.p.1mg/kg
Schicho 201162 C57/BL mice DSS O-1602 i.p.5mg/kg 30 minutes pre 7days 3 3
TNBS
Bashashati 201263 CD1 mice LPS AM3506 i.p.100 ug.kg 20 minutes pre 120 minutes 3 0
Izzo 201264 ICR mice CO CBC i.p.15mg/kg 20 minutes pre exam 4days 5 2
Lehmann 201265 Lewis rats LPS HU308 2.5mg/kg 15 minutes post 2–16 hours 4 0
CASP
Schicho 201242 C57/BL mice TNBS CBD i.p.10mg/kg 30 minutes pre onset 7days 4 0
PO 20mg/kg
PR 20mg/kg
Singh 201266 C57/BL mice IL-10 -/-DSS JWH133 i.p.2.5mg/kg Simultaneous 7–14days 5 1
Borrelli 201367 ICR mice DNBS CBG i.p.30mg/kg 3days pre 3days 5 1
Esposito 201433 CD-1 mice DSS PEA i.p.10mg/kg 2days post 7days 5 2
Li 201368 C57/BL mice DSS WIN55,212-2 i.p.5mg/kg Simultaneous 7days 4 1
Matos 201369 CD1 mice DSS αβ Amyrin PO 1, 3, 10mg/kg Pre and 3days post 7days 6 1
Naftali 201370 Clinical trial Crohn’s Cannabis sativa
extract (THC)
115mg inhaled N/A 8 weeks NA NA
Romano 201371 ICR mice DNBS CBC i.p 0.1–1.0mg/kg 24 hours post 3days 6 0
Wallace 201372 Wistar rats DNBS C.sativa (MMJ) IC 6mg/kg 30 minutes pre and 24
hours post
7days 4 1
AM630 PO 10mg/kg
Borelli 201573 ICR mice DNBS PEA i.p 1mg/kg 3days pre 3days 5 1
PO 1mg/kg
Capasso 201420 ICR mice OM PEA i.p.10mg/kg 30 minutes 3 and 7days 6 2
Fichna 201474 CD1 mice DSS AM841 i.p.0.01, 0.1,1mg/kg 15 minutes pre 3 and 7days 4 0
DNBS CB13 i.p.0.1mg/kg
Salaga 201475 C57/BL mice TNBS PF3845 i.p.10mg/kg 30 minutes 3 and 7days 2 0
DSS PO 5mg/kg
IC 5mg/kg
Sardinha 201476 C57/BL mice LPS HU308 IV 2.5mg/kg 15 minutes pre Simultaneous 6 0
AM630 IV 2.5mg/kg
URB597 i.p.0.6mg/kg
JZL184 i.p.16mg/kg
Alhouayek 201577 CD57/BL mice TNBS PEA i.p.10mg/kg Simultaneous and 5days
post
7days 4 1
DSS PF-3845 i.p.10mg/kg
AM9503 i.p.10mg/kg
TABLE2: Continued
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

686
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
Study Species Model Compound Route/dosage
Time of Administration
Versus Inammation
Time of
Assessment Post
Inammation
Modied STAIR
score SRCYCLE Score
El bakali 201578 C57/BL mice TNBS Compound 26 PO 10mg/kg 2days pre 7days 6 0
Impellizzeri
201579
CD1 mice DNBS uPEA i.p.10mg/kg 1 hour post 4days 9 2
Sasso 201580 CD1 mice TNBS ARN2508 PO 5mg/kg Simultaneous 7days 8 3
DSS
Stančić 201581 C57/BL mice DSS CID16020046 SC 20mg/kg 30 minutes 7days 6 1
TNBS
Cordaro 201682 CD1 mice DNBS Adelmidrol PO 10mg/kg 60 minutes post 4days 4 1
Feng 201683 C57/BL mice DSS WIN55,212-2 i.p.5mg/kg Simultaneous and 60
hours post
7days 5 1
Ke 201684 C57/BL mice DSS HU308 i.p.1mg/kg Simultaneous and daily 8days 4 2
Krohn 201640 CD1 mice TNBS Ab-CBD i.p.5mg/kg 45 minutes pre 4days 6 1
O-1918 i.p.5mg/kg
AM251 i.p.5mg/kg
AM630 i.p.5mg/kg
Pagano 201639 ICR mice DNBS CBD i.p.30mg/kg 24 hours post 3days 3 0
Pure CBD
CO PO 60mg/kg
Sarnelli 201685 CD1 mice DSS PEA i.p.2, 10mg/kg 2days post 7days 6 1
Lin 201786 C57/BL mice DSS HU210 i.p.0.05mg/kg 30 minutes pre 7days 5 1
Shamran 201787 C57/BL mice DSS FAAH-II i.p.5–40mg/kg 24 hours post 7days 6 1
Naftali 201735 Clinical trial Crohn’s CBD 10mg PO BD N/A 8 weeks NA NA
TABLE2: Continued
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

687
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
Disease Activity Index(DAI)
Thirty-four publications examined the effects of 25 canna-
binoid drugs across 68 experiments, within mouse and rat models
(total n= 948; n = 519 experimental vs 429 in control groups).
Cannabinoid drugs reduced DAI in comparison with the vehicle;
SMD -1.36; 95% CI, -1.62 to -1.09; I2=61% (Fig.4, Table 3).
On subgroup analysis, there was a signicant difference between
drug subtypes (P<0.001). DAI was signicantly reduced in mice
(SMD -1.49; 95% CI, -1.77 to -1.22; I2=61%). Seven experiments
within one publication examined the effects of cannabinoids on
rat colitis (THC and CBD, both conducted in a dose response
manner), but did not reach signicance at any concentration
(SMD -0.29; 95% CI, -0.77 to 0.20; I2=0%). SMD and con-
dence intervals for individual drugs on DAI are given in Table3.
FIGURE2. Positive, negative, and neutral outcomes of cannabinoid treatment across modes of inammation (A). Incidence of endpoints across
all experiments comparing cannabinoid treatment with control (B). The eect of cannabinoid drugs compared with control across all endpoints
expressed as primary drug investigated (C).
FIGURE3. Forest plot of the eects of cannabinoid treatment on Crohn’s Disease, assessed by reduction in CDAI in human studies.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

688
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
The largest effect size in DAI reduction was caused by
an enzyme inhibitor: the FAAH inhibitor URB597 (SMD
-4.43; 95% CI, -6.32 to -2.55). The largest effect size of DAI
reduction by an endocannabinoid was AEA (SMD -3.03; 95%
CI, -4.89 to -1.17); the largest effect size of DAI reduction
by a phytocannabinoid was CBD (SMD -0.56; 95% CI, -0.97
to -0.16; I2 = 29%), and the largest synthetic cannabinoid
effect size on DAI was AM1241 (SMD -3.11; 95% CI, -5.01
to -1.22). SMD and condence intervals of individual drugs
on DAI are given in Table4. Eighteen of 25 drugs had a large
effect size, 1 had a moderate effect size, and 6 had no signi-
cant effect onDAI.
Myeloperoxidase Activity(MPO)
Twenty-six publications investigated the effects of 21
cannabinoid drugs on MPO activity throughout 57 individ-
ual experiments (total n = 757, n = 419 in experimental vs
338 in control groups). Cannabinoid drugs reduced MPO in
comparison with the vehicle (SMD -1.26; 95% CI, -1.54 to
-0.97; I2=48.1%, Fig. 5, Table4). Overall, there was signi-
cant heterogeneity between studies, and there was signicant
subgroup difference (I2=48.1%; P<0.008). MPO was signif-
icantly reduced in mice and rats (SMD -1.28; 95% CI, -1.59 to
-0.98; I2=61%; and -1.06; 95% CI, -1.99 to -0.13; I2=56%,
respectively).
The largest effect size in MPO reduction was caused by
the phytocannabinoid CBG (SMD -6.20; 95% CI, -9.90 to -2.50;
I2 not assessed). The largest effect size by an endocannabinoid
was PEA (SMD -2.74; 95% CI, -4.42 to -1.06; I2= 85%). The
largest synthetic cannabinoid effect size on MPO was caused
by ACEA (SMD -3.15; 95% CI, -4.75 to -1.55; I2 not assessed).
And the largest effect size of any enzyme or transport inhib-
itor was AA5HT (SMD -2.27; 95% CI, -4.05 to -0.49; I2 not
assessed). SMD and condence intervals of individual drugs on
MPO activity are given in Table4. Thirteen of 21 cannabinoid
drugs had a large clinical effect, the remaining of which had no
signicant effect on MPO.
Time of Administration
From the 50 publications examining the effect of cannab-
inoids on murine colitis, 28 studies administered cannabinoid
agents either simultaneously with colitis onset or prophy-
lactically. Seventeen studies administered drugs between 15
minutes and 7days after the onset of colitis. Additionally, 7
studies compared the benet of prophylactic cannabinoid use
to therapeutic, but did not nd any difference in efcacy. To
investigate if the timing of drug treatment affected DAI or
MPO, we compared study size-weighted effect sizes (depend-
ent variable) with time of administration (covariate) using
meta-regression. We found that the timing of drug admin-
istration weakly predicted effect size in reducing DAI and
MPO, although this was of borderline statistical signicance
(P= 0.09, R2=11%, and P =0.055, R2=41%, respectively,
Fig.6A and B).
FIGURE4. Forest plot of the eects of cannabinoid treatment on
Disease Activity Score subdivided by drug type. Time of administration
in relation to onset of colitis is given where “p” represents prophylactic
administration, and “t” represents therapeutic administration.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

689
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
Quality and Risk ofBias
Of the 53 papers, 21 used randomization in their design,
7 reported blinding of assessment, 5 replicated their results in a
second species, and 14 replicated their ndings in a second model
of colitis. Fifty reported n ≥ 5 in control and experimental groups.
Fifteen publications reported specic numbers within groups.
All papers reported a clinically relevant endpoint. The median
study quality modied STAIR score was 5 out of 10 (mean
4.9, SD 2.29). Using meta-regression, higher quality scores pre-
dicted greater reductions in MPO activity (P=0.043, R2=65%,
Fig.6D), but not in DAI (P=0.98, R2=35%, Fig.6C).
The SYRCLE risk-of-bias score for each endpoint
showed a trend to larger reduction in DAI in studies with a
larger risk of bias (P=0.084, R2=69%, Fig.6E), but not MPO
(P=0.345, R2=8%, Fig.6F).
PublicationBias
Funnel plots comparing MPO activity and DAI were
constructed and analysed statistically for bias. The presence of
publication bias was not found in either group (MPO: Egger’s
statistic P =0.570, Fig.7A; DAI: Egger’s statistic P = 0.274,
Fig.7B).
DISCUSSION
The aim of this study was to determine the efcacy of can-
nabinoid drugs in reducing gut inammation to aid the design
of further clinical studies. We found 53 studies that examined
this effect using endocannabinoids, phytocannabinoids, syn-
thetic cannabinoids, and enzyme and reuptake inhibitors across
multiple models of murine and human colitis. In both qual-
itative assessment and meta-analysis, these controlled studies
TABLE3: The Eects of Cannabinoids on Disease Activity Score Caused by Experimental Colitis Grouped by Drug
No. of Studies No. of animals SMD [95% CI] P I2 (%) Clinical signicance
Endocannabinoids
PEA 6 118 -1.45 [-1.94, -0.96] <0.00001 25 High
AEA 1 12 -3.03 [-4.89, -1.17] 0.001 N/A High
Phytocannabinoids
CBD 12 181 -0.56 [-0.97, -0.16] 0.006 29 NS
THC 3 44 -0.53 [-1.24, 0.17] 0.14 0 NS
MMJ 1 30 -0.76 [-1.52, -0.00] 0.05 N/A Moderate
Cannabinomimetics
αβ Amyrin 4 28 -1.88 [-3.05, -0.72] 0.002 0 High
AM841 4 36 -1.87 [-3.57, -0.17] 0.03 66 High
βCaryophyllene 4 40 -1.52 [-2.32, -0.72] 0.0002 6 High
SAB378 4 56 0.28 [-0.38, 0.94] 0.41 28 NS
WIN55,212-2 4 60 -1.37 [-1.96, -0.78] <0.00001 0 High
CID16020046 2 16 -2.24 [-3.94, -0.54] 0.01 17 High
HU210 2 24 -2.89 [-6.24, 0.46] 0.09 81 NS
O-1602 2 28 -0.84 [-2.01, 0.33] 0.16 45 NS
ACEA 1 18 -0.87 [-1.85, 0.11] 0.08 N/A High
Adelmidrol 1 20 -1.85 [-2.94, -0.77] 0.0008 N/A High
AM1241 1 12 -3.11 [-5.01, -1.22] 0.001 N/A High
HU308 1 12 -0.73 [-1.92, 0.45] 0.23 N/A NS
Enzyme inhibitors
JWH133 4 53 -2.81 [-4.45, -1.17] 0.0008 71 High
PF3845 3 48 -2.21 [-3.11, -1.31] <0.00001 25 High
AA5HT 1 10 -2.16 [-3.90, -0.43] 0.01 N/A High
ARN2508 1 12 -2.66 [-4.38, -0.93] 0.002 N/A High
Compound 39 1 20 -1.47 [-2.48, -0.46] 0.004 N/A High
JZL184 1 22 -1.24 [-2.16, -0.31] 0.009 N/A High
URB597 1 18 -4.43 [-6.32, -2.55] <0.00001 N/A High
Transport inhibitors
VDM115 2 30 -3.06 [-4.21, -1.90] <0.00001 0 High
Total 68 948 -1.36 [-1.62, -1.09] <0.00001 61 High
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

690
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
demonstrate that the use of cannabinoid drugs are benecial
in reducing colonic inammation in rats and mice, with unclear
effects in human subjects.
In animal studies, cannabinoids were shown to reduce
inammation both qualitatively and at meta-analysis. Across
experiments included in this review, CB2 agonists, FAAH inhib-
itors, and CBD were the most widely studied and showed the
greatest therapeutic benet across all endpoints. Subgroup
analyses suggested that CBG caused the greatest reduction in
MPO activity scores, followed by synthetic CB1 agonist ACEA.
However, both agents were only studied within a single publi-
cation. In the MPO analysis, the most studied drug was CBD,
with 157 animals across 7 publications, demonstrating a sig-
nicant effect on MPO activity reduction. Similarly, within
DAI analysis, CBD was again the most-studied single drug,
including 181 animals across 6 publications. Although CBD
demonstrated a signicant effect on DAI reduction, the largest
reduction in DAI was caused by the FAAH antagonist URB597,
studied in 1 publication. There was statistical heterogeneity in
both MPO and DAI analyses, which was partially accounted
for by subgroup differences. At meta-regression, factors leading
to subgroup differences were quality, timing, and risk ofbias.
Receptor targets were explored in 23 publications using
receptor-specic agonists or antagonists and receptor knock-
down. In murine colitis, agonism of the CB1 or CB2 receptor
brought about reduction in inammation, and at subgroup
analysis, the use of the synthetic CB1/CB2 agonists acting
demonstrated the greatest reduction in disease scores and MPO
activity. In addition, agonism of the PPARα, GPR55, and
GPR18 receptors also reduced inammation of the colon. The
wide variation in the measured inammatory endpoints across
these studies prevented further meta-analysis. Interestingly, the
use of the peripherally restricted synthetic agonist SAB378,
which agonises both CB1 and CB2 receptors, had no signicant
effect on either MPO activity or DAI. This is in contrast to ex
vivo explant human colonic data, which demonstrated that can-
nabinoid agonism with AEA or CBD was benecial in colonic
mucosal inammation, which were peripherally restricted by
TABLE4: The Eects of Cannabinoids on Mpo Activity Caused by Experimental Colitis Grouped by Drug
No. of Studies No. of animals SMD [95% CI] P I2 (%) Clinical signicance
Endocannabinoids
PEA 7 94 -2.74 [-4.42, -1.06] 0.001 85 High
Phytocannabinoids
CBD 10 157 -1.03 [-1.40, -0.66] <0.00001 0 High
THC 3 29 -1.40 [-3.97, 1.17] 0.28 80 NS
CBC 1 10 -2.97 [-5.05, -0.89] 0.005 N/A High
CBG 1 10 -6.20 [-9.90, -2.50] 0.01 N/A High
Cannabinomimetics
βCaryophyllene 4 40 -1.26 [-2.48, -0.05] 0.04 60 High
AM841 4 48 -1.56 [-2.71, -0.41] 0.008 54 High
SAB378 4 42 -0.23 [-0.86, 0.39] 0.46 0 NS
WIN55,212-2 4 52 -1.74 [-2.81, -0.67] 0.001 57 High
αβ Amyrin 2 15 -0.38 [-1.48, 0.71] 0.5 0 NS
CID16020046 2 56 -1.04 [-1.61, -0.48] 0.0003 0 High
HU210 2 24 -0.63 [-1.48, 0.23] 0.15 2 NS
O-1602 2 20 -1.70 [-2.81, -0.60] 0.003 0 High
ACEA 1 16 -3.15 [-4.75, -1.55] 0.0001 N/A High
AM1241 1 10 -0.96 [-2.31, 0.39] 0.16 N/A NS
JWH133 1 16 -0.98 [-2.04, 0.07] 0.09 N/A NS
Ademidrol 1 20 -1.33 [-2.31, -0.34] 0.009 N/A High
Enzyme inhibitors
PF3745 3 46 -0.12 [-1.56, 1.32] 0.81 81 NS
AA5HT 1 10 -2.27 [-4.05, -0.49] 0.01 N/A High
URB597 1 16 -1.00 [-2.06, 0.06] 0.06 N/A NS
Transport inhibitors
VDM115 2 26 -1.91 [-3.72, -0.10] 0.04 59 High
Total 57 757 -1.26 [-1.54, -0.97] <0.00001 48.1 High
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

691
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
denition of the explant model.30, 31 Izzo et al9 found through
receptor antagonism that the effect of CBN in preventing
hypermobility caused by croton oil was mediated by CB1, but
not CB2. PEA was investigated by Capasso etal20, 32 using two
models of inammation-induced hypermotility. Using receptor
antagonists in both experiments, Capasso etal found that PEA,
in an OM model, acted through CB1, but not CB2 or PPARα.
In a CO model, PEA was still effective, but did not act through
CB1 or CB2. This suggests that the mechanism by which PEA
acts as an anti-inammatory agent was not mediated by a sin-
gle receptor, but by receptor codependence. ACEA was investi-
gated for receptor mechanism in 2 publications, both of which
found ACEA dependent on CB1. None of the reviewed studies
investigated a mechanism of action for AEA in gut inamma-
tion, however 1 ex vivo human study from Harvey etal found
that AEA prevented increased cytokine production in experi-
mentally inamed human mucosa and was dependent on CB2,
although the authors did not report antagonism of any other
receptor.31
The specic mechanism by which manipulation of the
cannabinoid system affects inammation is not clear. Esposito
et al33 demonstrated that PEA brought about anti-inamma-
tory effects on enteric glial cells acting at toll-like receptor 4,
suggesting that rather than acting at an epithelial mucosal level,
it acts at either an innate immune colonies or the enteric nerv-
ous system. This hypothesis has recently been evidenced by a
study demonstrating that both CBD and PEA do not act on
the immune response of epithelial cells, but are likely to require
the presence of other cell types, acting through down regulation
of NF-κβ.34 This is challenged by Cluny etal, who demonstrate
that peripherally restricted cannabinoids have a diminished
effect on inammation. Nevertheless, it is clear that the mech-
anism of action of cannabinoids does not simply lie at the epi-
thelial level, but is likely to reside within the gut-brainaxis.
From the clinical literature, we found 2 randomized
placebo-controlled studies examining the effect of phytocan-
nabinoids in humans. Our analysis found no overall effect
of THC or CBD on disease scores; however; there was large
statistical and clinical heterogeneity between these studies.
We found from meta-analysis that inhaled THC did have a
benecial effect on CDAI at 8 weeks, whereas CBD did not.
There may be several reasons for this heterogeneity: Firstly
in all groups, small cohort sizes were used which may have
overestimated positive or negative effects in both studies,
making meaningful conclusions difcult regarding the use of
CBD or THC in inammatory bowel disease. Secondly, within
the Naftali et al (2017) study, very low doses of CBD were
utilized compared to the use of CBD in other clinical trials,
which commonly used 600mg twice daily.35 Arecent trial in
drug-resistant epilepsy used 20mg.kg-1 daily for 4 weeks, with
a small number of participants experiencing side effects such
as vomiting and diarrhea.36 It is likely that in adult males such
FIGURE5. Forest plot of the eects of cannabinoid treatment on MPO
activity subdivided by drug type. Time of administration in relation to
onset of colitis is given, where “p” represents prophylactic administra-
tion and “t” represents therapeutic administration.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

692
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
FIGURE6. The eect of cannabinoid treatment on experimentally induced colitis determined by DAI (A) and MPO (B) predicted by timing of drug
administration in relation to colitis onset. The eect of study quality, determined by mSTAIR score, and SYRCLE score, on eect size in DAI (C, E) and
MPO (D, F). Study weights are represented by the diameter of the circle, with larger circles representing studies with largest weight in the analysis.
FIGURE7. Funnel plots evaluating for publication bias in (A) MPO activity and (B) DAI. Standard error of the standardized mean dierence (SE [SMD],
y axes) for each study is plotted against its eect size (SMD, x axes).
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

693
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
10mg doses had no clinical effect on Crohn’s disease, as insuf-
cient plasma concentrations may have been reached due to
the poor bioavailability of oral CBD. Amajor aw within the
Naftali etal 2013 trial is that sham cigarettes contained can-
nabis sativa owers in which active cannabinoids had been
removed. However, it is unlikely that other compounds pres-
ent in cannabis (such as terpenes), which are known to have an
anti-inammatory effect, had also been removed, which may
have introduced positive bias into the study.37 However, des-
pite these drawbacks, the Naftali etal 2013 trial demonstrated
a signicant reduction in pain and the use of steroid therapy,
with increased sleep and satisfaction levels with THC use com-
pared with placebo. Although not included in this analysis,
a study from Storr etal38 demonstrated that although can-
nabis use provided symptomatic relief from Crohn’s disease,
the risk of salvage surgery was increased within 6months of
use (OR: 5.03; 95% CI, 1.45–17.46). However, these ndings
have not yet been supported from randomized, blinding con-
trolled trials. We may suggest, therefore, that phytocannabi-
noid use may be a future therapy in intestinal inammation.
Although before rm conclusions are drawn, further clinical
studies examining their effects should be conducted at higher,
therapeutic dosages with adequately powered cohort sizes. As
MMJ use in inammatory bowel disease has been justied
because of its effects on appetite and diarrhea, studies may be
designed to examine these quality-of-life–affecting endpoints
directly.
We found that most of the existing cannabinoid-gut
research focuses on the therapeutic potential of CBD. This is
unsurprising as CBD is currently used clinically, is well toler-
ated, and has shown consistently positive results. Nine studies
found a positive, dose-dependent effect on local inammatory
cytokine expression, COX2 activation, MPO activity, enteric
glial cell activation, and caspase-3 production, with associated
improvements in macroscopic and histologic grades of inam-
mation.39–46 One study also showed that intraperitoneal CBD
administration decreased oxidative-stress scores of peripheral
lung and brain tissue following intestinal inammation,47 add-
ing to the existing evidence that CBD maintains the gut barrier
during inammation.48 Despite being the most-studied drug,
the mechanism by which CBD acts was not made clear by this
review. One study by De Fillipis etal44 found that hypermo-
tility caused by LPS administration in mice was reduced by
CBD through a CB1 dependent mechanism. Similarly, Capasso
et al (2008) found that CBD prevented croton oil-induced
hypermotility via CB1. de Fillipis et al (2011) demonstrated
that in human explant tissue S100B levels, as a marker of glial
cell activation, in vitro was decreased by CBD in a PPARγ
dependent mechanism (although other antagonists were not
investigated).49
The timing of cannabinoid administration correlated
with reduction in effect on colitis activity, although this did
not reach statistical signicance. There was a correlation
between the time of drug administration and effect size in
both DAI and MPO reduction, with earlier administration
of cannabinoid drugs producing a greater effect size. This
suggests that in clinical trials, cannabinoids may be used
prophylactically and therapeutically. There is promise, there-
fore, that compounds targeting the endocannabinoid system
may be able to not only prevent colonic inammation, but
also treat established intestinal inammatory conditions.
Because it is not clear if cannabinoids are more effective
when treating new-onset or established intestinal inamma-
tion, further study designs should investigate this endpoint
specically.
One important potential area for research is the combin-
ation of cannabinoid drugs with existing treatments for inam-
matory bowel disease. In clinical practice, it is common to treat
patients with acute severe Crohn’s disease and ulcerative colitis
with a combination of agents, such as antibiotic, anti-TNFα,
and corticosteroid therapy. One study compared the efcacy of
CBD and THC with that of sulphasalazine, a 5-ASA, a drug
commonly used in clinical practice.45 Although in this study
CBD and THC efcacy were comparable to that of sulphasala-
zine, the authors did not examine for the potential additive or
subtractive effect of these agents in the context of colitis.
The ndings of this study are limited by several factors
typically seen in meta-analyses and systematic reviews. We
found signicant heterogeneity between subgroups in both
DAI and MPO analyses, which suggested that 11% and 41%
of this was due to the difference in time of administration in
terms of changes in DAI and MPO, respectively. Additionally,
we found risk-of-bias study design to be high and median study
quality to be relatively low. Meta-regression demonstrated that
these factors signicantly correlated with study outcomes.
Although we did not analyze for differences between scoring
systems and mode of colitis, these factors may have also con-
tributed to heterogeneity and inuenced outcome. We sought
to overcome this variability between scoring systems with ran-
dom effects analysis. Additionally, within this review we have
examined the effect of cannabinoid drugs en mass, which may
have affected the overall outcome of meta-analyses. It is pos-
sible that some articles may not have been identied in initial
searches, or conference abstracts may have been missed from
the search period. Lastly, where control groups were compared
to multiple experimental groups within the same set of exper-
iments, variance and SMD may be exaggerated, leading to
furtherbias.
In conclusion, we have shown in this systematic review
and meta-analysis that cannabinoid drugs are benecial in
treating experimentally-induced murine models of colitis.
These positive ndings support the development of further
human clinical trials. Current literature converges on CBD,
and to avoid research bias, the effect of all cannabinoid drugs,
including the large number of phytocannabinoid drugs not yet
investigated, should also be investigated.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

694
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
REFERENCES
1. Kappelman MD, Rifas-Shiman SL, Kleinman K, etal. The prevalence and geo-
graphic dstribution of Crohn’s disease and ulcerative colitis in the United States.
Clin Gastroenterol Hepatol. 2007;5:1424–9.
2. Rubin GP, Hungin AP, Kelly PJ, etal. Inammatory bowel disease: epidemiology
and management in an English general practice population. Aliment Pharmacol
Ther. 2000;14:1553–9.
3. Sandborn WJ, Rutgeerts P, Feagan BG, etal. Colectomy rate comparison after
treatment of ulcerative colitis with placebo or iniximab. Gastroenterology.
2009;137:1250–60; quiz 1520.
4. Sutherland LR, Martin F, Bailey RJ, et al. A randomized, placebo-controlled,
double-blind trial of mesalamine in the maintenance of remission of Crohn’s
disease. The canadian mesalamine for remission of Crohn’s disease study group.
Gastroenterology. 1997;112:1069–77.
5. Herrinton LJ, Liu L, Weng X, etal. Role of thiopurine and anti-TNF ther-
apy in lymphoma in inammatory bowel disease. Am J Gastroenterol.
2011;106:2146–53.
6. Alhamoruni A, Wright KL, Larvin M, et al. Cannabinoids mediate oppos-
ing effects on inammation-induced intestinal permeability. Br J Pharmacol.
2012;165:2598–610.
7. Storr MA, Sharkey KA. The endocannabinoid system and gut-brain signalling.
Curr Opin Pharmacol. 2007;7:575–82.
8. Lee Y, Jo J, Chung HY, et al. Endocannabinoids in the gut. Am J Physiol
Gastrointest Liver Physiol. 2016; 1: 67–77.
9. Izzo AA, Fezza F, Capasso R, etal. Cannabinoid CB1-receptor mediated regula-
tion of gastrointestinal motility in mice in a model of intestinal inammation. Br
J Pharmacol. 2001;134:563–70.
10. Di Marzo V, Izzo AA. Endocannabinoid overactivity and intestinal inamma-
tion. Gut. 2006;55:1373–6.
11. Devane WA, Hanus L, Breuer A, etal. Isolation and structure of a brain constitu-
ent that binds to the cannabinoid receptor. Science. 1992;258:1946–9.
12. Freund TF, Katona I, Piomelli D. Role of endogenous cannabinoids in synaptic
signaling. Physiol Rev. 2003;83:1017–66.
13. Muccioli GG. Endocannabinoid biosynthesis and inactivation, from simple to
complex. Drug Discov Today. 2010;15:474–83.
14. Blankman JL, Simon GM, Cravatt BF. A comprehensive prole of brain
enzymes that hydrolyze the endocannabinoid 2-arachidonoylglycerol. Chem Biol.
2007;14:1347–56.
15. O’Sullivan SE. Cannabinoids go nuclear: evidence for activation of peroxisome
proliferator-activated receptors. Br J Pharmacol. 2007;152:576–82.
16. Smart D, Jonsson KO, Vandevoorde S, etal. ‘Entourage’ effects of N-acyl eth-
anolamines at human vanilloid receptors. Comparison of effects upon anan-
damide-induced vanilloid receptor activation and upon anandamide metabolism.
Br J Pharmacol. 2002;136:452–8.
17. Leocani L, Nuara A, Houdayer E, et al. Sativex(®) and clinical-neurophys-
iological measures of spasticity in progressive multiple sclerosis. J Neurol.
2015;262:2520–7.
18. Reddy DS, Golub VM. The pharmacological basis of cannabis therapy for epi-
lepsy. J Pharmacol Exp Ther. 2016;357:45–55.
19. Suárez J, Romero-Zerbo Y, Márquez L, et al. Ulcerative colitis impairs the acy-
lethanolamide-based anti-inammatory system reversal by 5-aminosalicylic acid
and glucocorticoids. Plos One. 2012;7:e37729.
20. Capasso R, Orlando P, Pagano E, etal. Palmitoylethanolamide normalizes intes-
tinal motility in a model of post-inammatory accelerated transit: involvement of
CB₁ receptors and TRPV1 channels. Br J Pharmacol. 2014;171:4026–37.
21. Akbar A, Yiangou Y, Facer P, etal. Expression of the TRPV1 receptor differs
in quiescent inammatory bowel disease with or without abdominal pain. Gut.
2010;59:767–74.
22. D’Argenio G, Valenti M, Scaglione G, et al. Up-regulation of anandamide lev-
els as an endogenous mechanism and a pharmacological strategy to limit colon
inammation. Faseb J. 2006;20:568–70.
23. D’Argenio G, Petrosino S, Gianfrani C, etal. Overactivity of the intestinal endo-
cannabinoid system in celiac disease and in methotrexate-treated rats. J Mol Med
(Berl). 2007;85:523–30.
24. Karwad MA, Macpherson T, Wang B, etal. Oleoylethanolamine and palmitoy-
lethanolamine modulate intestinal permeability in vitro via TRPV1 and pparα.
Faseb J. 2017;31:469–81.
25. Massa F, Marsicano G, Hermann H, etal. The endogenous cannabinoid system
protects against colonic inammation. J Clin Invest. 2004;113:1202–9.
26. Higgins JPT, Altman DG, Gøtzsche PC, etal. The Cochrane Collaboration’s tool
for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928.
27. Fisher M, Feuerstein G, Howells DW, etal.; STAIR Group. Update of the stroke
therapy academic industry roundtable preclinical recommendations. Stroke.
2009;40:2244–50.
28. Kilkenny C, Browne WJ, Cuthill IC, et al. Improving bioscience research
reporting: the ARRIVE guidelines for reporting animal research. Plos Biol.
2010;8:e1000412.
29. Hooijmans CR, Rovers MM, de Vries RB, etal. SYRCLE’S risk of bias tool for
animal studies. BMC Med Res Methodol. 2014;14:43.
30. Harvey BS, Sia TC, Wattchow DA, etal. Interleukin 17A evoked mucosal damage
is attenuated by cannabidiol and anandamide in a human colonic explant model.
Cytokine. 2014;65:236–44.
31. Nicotra LL, Vu M, Harvey BS, etal. Prostaglandin ethanolamides attenuate
damage in a human explant colitis model. Prostaglandins Other Lipid Mediat.
2013;100-101:22–9.
32. Capasso R, Izzo AA, Fezza F, etal. Inhibitory effect of palmitoylethanolamide
on gastrointestinal motility in mice. Br J Pharmacol. 2001;134:945–50.
33. Esposito G, Capoccia E, Turco F, etal. Palmitoylethanolamide improves colon
inammation through an enteric glia/toll like receptor 4-dependent PPAR-α acti-
vation. Gut. 2014;63:1300–12.
34. Couch DG, Tasker C, Theophilidou E, et al. Cannabidiol and palmitoyletha-
nolamide are anti-inammatory in the acutely inamed human colon. Clin Sci
(Lond). 2017;131:2611–26.
35. Naftali T, Mechulam R, Marii A, etal. Low-dose cannabidiol is safe but not
effective in the treatment for Crohn’s disease, a randomized controlled trial. Dig
Dis Sci. 2017;62:1615–20.
36. Devinsky O, Cross JH, Laux L, etal.; Cannabidiol in Dravet Syndrome Study
Group. Trial of cannabidiol for drug-resistant seizures in the dravet syndrome. N
Engl J Med. 2017;376:2011–20.
37. de las Heras B, Hortelano S. Molecular basis of the anti-inammatory effects of
terpenoids. Inamm Allergy Drug Targets. 2009;8:28–39.
38. Storr M, Devlin S, Kaplan GG, etal. Cannabis use provides symptom relief in
patients with inammatory bowel disease but is associated with worse disease
prognosis in patients with Crohn’s disease. Inamm Bowel Dis. 2014;20:472–80.
39. Pagano E, Capasso R, Piscitelli F, et al. An orally active cannabis extract with
high content in cannabidiol attenuates chemically-induced intestinal inamma-
tion and hypermotility in the mouse. Front Pharmacol. 2016;7:341.
40. Krohn RM, Parsons SA, Fichna J, etal. Abnormal cannabidiol attenuates experi-
mental colitis in mice, promotes wound healing and inhibits neutrophil recruit-
ment. J Inamm (Lond). 2016;13:21.
41. Harvey BS, Nicotra LL, Vu M, etal. Cannabinoid CB2 receptor activation atten-
uates cytokine-evoked mucosal damage in a human colonic explant model with-
out changing epithelial permeability. Cytokine. 2013;63:209–17.
42. Schicho R, Storr M. Topical and systemic cannabidiol improves trinitrobenzene
sulfonic acid colitis in mice. Pharmacology. 2012;89:149–55.
43. Lin XH, Yuece B, Li YY, etal. A novel CB receptor GPR55 and its ligands are
involved in regulation of gut movement in rodents. Neurogastroenterol Motil.
2011;23:862–e342.
44. de Filippis D, Iuvone T, d’amico A, etal. Effect of cannabidiol on sepsis-induced
motility disturbances in mice: involvement of CB receptors and fatty acid amide
hydrolase. Neurogastroenterol Motil. 2008;20:919–27.
45. Jamontt JM, Molleman A, Pertwee RG, etal. The effects of delta-tetrahydrocan-
nabinol and cannabidiol alone and in combination on damage, inammation and
in vitro motility disturbances in rat colitis. Br J Pharmacol. 2010;160:712–23.
46. Borrelli F, Aviello G, Romano B, etal. Cannabidiol, a safe and non-psychotropic
ingredient of the marijuana plant cannabis sativa, is protective in a murine model
of colitis. J Mol Med (Berl). 2009;87:1111–21.
47. Cassol-Jr OJ, Comim CM, Silva BR, etal. Treatment with cannabidiol reverses
oxidative stress parameters, cognitive impairment and mortality in rats submitted
to sepsis by cecal ligation and puncture. Brain Res. 2010;1348:128–38.
48. Alhamoruni A, Lee AC, Wright KL, etal. Pharmacological effects of cannabi-
noids on the caco-2 cell culture model of intestinal permeability. J Pharmacol Exp
Ther. 2010;335:92–102.
49. De Filippis D, Esposito G, Cirillo C, etal. Cannabidiol reduces intestinal inam-
mation through the control of neuroimmune axis. Plos One. 2011;6:e28159.
50. Mathison R, Ho W, Pittman QJ, etal. Effects of cannabinoid receptor-2 activa-
tion on accelerated gastrointestinal transit in lipopolysaccharide-treated rats. Br
J Pharmacol. 2004;142:1247–54.
51. Kimball ES, Schneider CR, Wallace NH, etal. Agonists of cannabinoid receptor
1 and 2 inhibit experimental colitis induced by oil of mustard and by dextran
sulfate sodium. Am J Physiol Gastrointest Liver Physiol. 2006;291:G364–G71.
52. Capasso R, Borrelli F, Aviello G, et al. Cannabidiol, extracted from cannabis
sativa, selectively inhibits inammatory hypermotility in mice. Br J Pharmacol.
2008;154:1001–8.
53. Engel MA, Kellermann CA, Rau T, et al. Ulcerative colitis in AKR mice is
attenuated by intraperitoneally administered anandamide. J Physiol Pharmacol.
2008;59:673–89.
54. Storr MA, Keenan CM, Emmerdinger D, etal. Targeting endocannabinoid deg-
radation protects against experimental colitis in mice: involvement of CB1 and
CB2 receptors. J Mol Med (Berl). 2008;86:925–36.
55. Li YY, Li YN, Ni JB, etal. Involvement of cannabinoid-1 and cannabinoid-2
receptors in septic ileus. Neurogastroenterol Motil. 2010;22:350–e88.
56. Storr MA, Keenan CM, Zhang H, et al. Activation of the cannabinoid 2
receptor (CB2) protects against experimental colitis. Inamm Bowel Dis.
2009;15:1678–85.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

695
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
57. Cluny NL, Keenan CM, Duncan M, et al. Naphthalen-1-yl-(4-
pentyloxynaphthalen-1-yl)methanone (SAB378), a peripherally restricted
cannabinoid CB1/CB2 receptor agonist, inhibits gastrointestinal motility
but has no effect on experimental colitis in mice. J Pharmacol Exp Ther.
2010;334:973–80.
58. Kimball ES, Wallace NH, Schneider CR, et al. Small intestinal cannabinoid
receptor changes following a single colonic insult with oil of mustard in mice.
Front Pharmacol. 2010;1:132.
59. Alhouayek M, Lambert DM, Delzenne NM, etal. Increasing endogenous 2-ara-
chidonoylglycerol levels counteracts colitis and related systemic inammation.
Faseb J. 2011;25:2711–21.
60. Andrz ejak V, Muccioli GG, Body-Malapel M, etal. New FAAH inhibitors based
on 3-carboxamido-5-aryl-isoxazole scaffold that protect against experimental col-
itis. Bioorg Med Chem. 2011;19:3777–86.
61. Bento AF, Marcon R, Dutra RC, etal. Β-caryophyllene inhibits dextran sulfate
sodium-induced colitis in mice through CB2 receptor activation and pparγ path-
wa y. Am J Pathol. 2011;178:1153–66.
62. Schicho R, Bashashati M, Bawa M, etal. The atypical cannabinoid O-1602 pro-
tects against experimental colitis and inhibits neutrophil recruitment. Inamm
Bowel Dis. 2011;17:1651–64.
63. Bashashati M, Storr MA, Nikas SP, et al. Inhibiting fatty acid amide hydrolase
normalizes endotoxin-induced enhanced gastrointestinal motility in mice. Br J
Pharmacol. 2012;165:1556–71.
64. Izzo AA, Capasso R, Aviello G, et al. Inhibitory effect of cannabichromene, a
major non-psychotropic cannabinoid extracted from cannabis sativa, on inam-
mation-induced hypermotility in mice. Br J Pharmacol. 2012;166:1444–60.
65. Lehmann C, Kianian M, Zhou J, etal. Cannabinoid receptor 2 activation reduces
intestinal leukocyte recruitment and systemic inammatory mediator release in
acute experimental sepsis. Crit Care. 2012;16:R47.
66. Singh UP, Singh NP, Singh B, etal. Cannabinoid receptor-2 (CB2) agonist ame-
liorates colitis in IL-10(-/-) mice by attenuating the activation of T cells and pro-
moting their apoptosis. Toxicol Appl Pharmacol. 2012;258:256–67.
67. Borrelli F, Fasolino I, Romano B, etal. Benecial effect of the non-psychotropic
plant cannabinoid cannabigerol on experimental inammatory bowel disease.
Biochem Pharmacol. 2013;85:1306–16.
68. Li YY, Yuece B, Cao HM, et al. Inhibition of p38/mk2 signaling pathway
improves the anti-inammatory effect of WIN55 on mouse experimental colitis.
Lab Invest. 2013;93:322–33.
69. Matos I, Bento AF, Marcon R, etal. Preventive and therapeutic oral adminis-
tration of the pentacyclic triterpene α,β-amyrin ameliorates dextran sulfate sodi-
um-induced colitis in mice: the relevance of cannabinoid system. Mol Immunol.
2013;54:482–92.
70. Naftali T, Bar-Lev Schleider L, Dotan I, et al. Cannabis induces a clinical
response in patients with Crohn’s disease: a prospective placebo-controlled study.
Clin Gastroenterol Hepatol. 2013;11:1276–80.e1.
71. Romano B, Borrelli F, Fasolino I, et al. The cannabinoid TRPA1 agonist can-
nabichromene inhibits nitric oxide production in macrophages and ameliorates
murine colitis. Br J Pharmacol. 2013;169:213–29.
72. Wallace JL, Flannigan KL, McKnight W, etal. Pro-resolution, protective and
anti-nociceptive effects of a cannabis extract in the rat gastrointestinal tract. J
Physiol Pharmacol. 2013;64:167–75.
73. Borrelli F, Romano B, Petrosino S, etal. Palmitoylethanolamide, a naturally
occurring lipid, is an orally effective intestinal anti-inammatory agent. Br J
Pharmacol. 2015;172:142–58.
74. Fichna J, Sałaga M, Stuart J, etal. Selective inhibition of FAAH produces anti-
diarrheal and antinociceptive effect mediated by endocannabinoids and cannabi-
noid-like fatty acid amides. Neurogastroenterol Motil. 2014;26:470–81.
75. Sałaga M, Mokrowiecka A, Zakrzewski PK, etal. Experimental colitis in mice is atten-
uated by changes in the levels of endocannabinoid metabolites induced by selective
inhibition of fatty acid amide hydrolase (FAAH). J Crohn’s Colitis. 2014;8:998–1009.
76. Sardinha J, Kelly ME, Zhou J, etal. Experimental cannabinoid 2 receptor-medi-
ated immune modulation in sepsis. Mediators Inamm. 2014;2014:978678.
77. Alhouayek M, Bottemanne P, Subramanian KV, et al. N-acylethanolamine-
hydrolyzing acid amidase inhibition increases colon N-palmitoylethanolamine
levels and counteracts murine colitis. Faseb J. 2015;29:650–61.
78. El Bakali J, Muccioli GG, Body-Malapel M, etal. Conformational restriction
leading to a selective CB2 cannabinoid receptor agonist orally active against col-
itis. ACS Med Chem Lett. 2015;6:198–203.
79. Impellizzeri D, Di Paola R, Cordaro M, etal. Adelmidrol, a palmitoylethanola-
mide analogue, as a new pharmacological treatment for the management of acute
and chronic inammation. Biochem Pharmacol. 2016;119:27–41.
80. Sasso O, Migliore M, Habrant D, etal. Multitarget fatty acid amide hydrolase/
cyclooxygenase blockade suppresses intestinal inammation and protects against
nonsteroidal anti-inammatory drug-dependent gastrointestinal damage. Faseb
J. 2015;29:2616–27.
81. Stančić A, Jandl K, Hasenöhrl C, etal. The GPR55 antagonist CID16020046 pro-
tects against intestinal inammation. Neurogastroenterol Motil. 2015;27:1432–45.
82. Cordaro M, Impellizzeri D, Gugliandolo E, etal. Adelmidrol, a palmitoyletha-
nolamide analogue, as a new pharmacological treatment for the management of
inammatory bowel disease. Mol Pharmacol. 2016;90:549–61.
83. Feng YJ, Li YY, Lin XH, etal. Anti-inammatory effect of cannabinoid agonist
WIN55, 212 on mouse experimental colitis is related to inhibition of p38mapk.
World J Gastroenterol. 2016;22:9515–24.
84. Ke P, Shao BZ, Xu ZQ, et al. Activation of cannabinoid receptor 2 ameliorates
DSS-induced colitis through inhibiting NLRP3 inammasome in macrophages.
Allen IC, editor. PLoS One. 2016;11:e0155076.
85. Sarnelli G, D’Alessandro A, Iuvone T, et al. Palmitoylethanolamide modulates
inammation-associated vascular endothelial growth factor (VEGF) signaling
via the akt/mtor pathway in a selective peroxisome proliferator-activated receptor
alpha (PPAR-α)-dependent manner. Plos One. 2016;11:e0156198.
86. Lin S, Li Y, Shen L, et al. The anti-inammatory effect and intestinal barrier
protection of HU210 differentially depend on TLR4 signaling in dextran sulfate
sodium-induced murine colitis. Dig Dis Sci. 2017;62:372–86.
87. Shamran H, Singh NP, Zumbrun EE, etal. Fatty acid amide hydrolase (FAAH)
blockade ameliorates experimental colitis by altering microrna expression and
suppressing inammation. Brain Behav Immun. 2017;59:10–20.
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

696
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018
Couch etal
APPENDIX: PRISMA Checklist
Section/topic # Checklist item
Reported on
page #
TITLE
Title 1 Identify the report as a systematic review, meta-analysis, or both. 1
ABSTRACT
Structured summary 2 Provide a structured summary including, as applicable: background; objectives; data sources;
study eligibility criteria, participants, and interventions; study appraisal and synthesis meth-
ods; results; limitations; conclusions and implications of key ndings; systematic review
registration number.
2
INTRODUCTION
Rationale 3 Describe the rationale for the review in the context of what is already known. 4
Objectives 4 Provide an explicit statement of questions being addressed with reference to participants, inter-
ventions, comparisons, outcomes, and study design (PICOS).
5
METHODS
Protocol and registration 5 Indicate if a review protocol exists, if and where it can be accessed (e.g., Web address), and, if
available, provide registration information including registration number.
6
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g.,
years considered, language, publication status) used as criteria for eligibility, giving rationale.
6
Information sources 7 Describe all information sources (e.g., databases with dates of coverage, contact with study
authors to identify additional studies) in the search and date last searched.
6
Search 8 Present full electronic search strategy for at least one database, including any limits used, such
that it could be repeated.
6
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in systematic review,
and, if applicable, included in the meta-analysis).
6
Data collection process 10 Describe method of data extraction from reports (e.g., piloted forms, independently, in dupli-
cate) and any processes for obtaining and conrming data from investigators.
6
Data items 11 List and dene all variables for which data were sought (e.g., PICOS, funding sources) and any
assumptions and simplications made.
6-7
Risk of bias in individual
studies
12 Describe methods used for assessing risk of bias of individual studies (including specication
of whether this was done at the study or outcome level), and how this information is to be
used in any data synthesis.
6-7
Summary measures 13 State the principal summary measures (e.g., risk ratio, difference in means). 7
Synthesis of results 14 Describe the methods of handling data and combining results of studies, if done, including
measures of consistency (e.g., I2) for each meta-analysis.
7
Risk of bias across studies 15 Specify any assessment of risk of bias that may affect the cumulative evidence (e.g., publication
bias, selective reporting within studies).
6
Additional analyses 16 Describe methods of additional analyses (e.g., sensitivity or subgroup analyses, meta-regres-
sion), if done, indicating which were pre-specied.
7
RESULTS
Study selection 17 Give numbers of studies screened, assessed for eligibility, and included in the review, with rea-
sons for exclusions at each stage, ideally with a ow diagram.
8+19
Study characteristics 18 For each study, present characteristics for which data were extracted (e.g., study size, PICOS,
follow-up period) and provide the citations.
8-11+28-29
Risk of bias within studies 19 Present data on risk of bias of each study and, if available, any outcome level assessment (see
item 12).
11
Results of individual
studies
20 For all outcomes considered (benets or harms), present, for each study: (a) simple summary
data for each intervention group (b) effect estimates and condence intervals, ideally with a
forest plot.
30-31
Synthesis of results 21 Present results of each meta-analysis done, including condence intervals and measures of
consistency.
10-11
Risk of bias across studies 22 Present results of any assessment of risk of bias across studies (see Item 15). 12
Additional analysis 23 Give results of additional analyses, if done (e.g., sensitivity or subgroup analyses, meta-regres-
sion [see Item16]).
11
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019

697
Inflamm Bowel Dis
•
Volume 24, Number 4, April 2018 The use of cannabinoids in colitis: a systematic review and meta-analysis
Section/topic # Checklist item
Reported on
page #
DISCUSSION
Summary of evidence 24 Summarize the main ndings including the strength of evidence for each main outcome; con-
sider their relevance to key groups (e.g., healthcare providers, users, and policy makers).
13
Limitations 25 Discuss limitations at study and outcome level (e.g., risk of bias), and at review-level (e.g., in-
complete retrieval of identied research, reporting bias).
17
Conclusions 26 Provide a general interpretation of the results in the context of other evidence, and implications
for future research.
13-17
FUNDING
Funding 27 Describe sources of funding for the systematic review and other support (e.g., supply of data);
role of funders for the systematic review.
1
From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement.
PLoS Med 67: e1000097. doi:10.1371/journal.pmed1000097.
APPENDIX: Continued
Downloaded from https://academic.oup.com/ibdjournal/article-abstract/24/4/680/4944355 by University of Nottingham user on 11 February 2019