CADTH RAPID RESPONSE REPORT:

SUMMARY WITH CRITICAL APPRAISAL

Cannabinoids for
Behavioural Symptoms in
Adults with Dementia: A
Review of Clinical
Effectiveness and Guidelines

Service Line: Rapid Response Service

Version: 1.0
Publication Date: January 5, 2017
Report Length: 23 Pages
 Authors: Heidi Staples, Lorna Adcock

Cite As: Cannabinoids f or behav ioural sy mptoms in adults with dementia: a rev iew of clinical ef f ectiveness and guidelines. Ottawa: CADTH; 2018 Jan.
(CADTH rapid response report: summary with critical appraisal).

Acknowledgments:

ISSN: 1922-8147 (online)

Disclaimer: The inf ormation in this document is intended to help Canadian health care decision-makers, health care prof essionals, health sy stems leaders,
and policy -makers make well-inf ormed decisions and thereby improv e the quality of health care serv ices. While pat ients and others may access this document,
the document is made av ailable f or inf ormational purposes only and no representations or warranties are made with respect to its f itness f or any particular
purpose. The inf ormation in this document should not be used as a substitute f or prof essional medical adv ice or as a substitute f or the application of clinical
judgment in respect of the care of a particular patient or other prof essional judgment in any decision-making process. The Canadian Agency f or Drugs and
Technologies in Health (CADTH) does not endorse any inf ormation, drugs, therapies, treatments, products, processes, or serv ic es.

While care has been taken to ensure that the inf ormation prepared by CADTH in this document is accurate, complete, and up -to-date as at the applicable date
the material was f irst published by CADTH, CADTH does not make any guarantees to that ef f ect. CADTH does not guarantee and is not responsible f or the
quality , currency , propriety , accuracy, or reasonableness of any statements, information, or conclusions contained in any third-party materials used in preparing
this document. The v iews and opinions of third parties published in this document do not necessarily state or ref lect those o f CADTH.

CADTH is not responsible f or any errors, omissions, injury , loss, or damage arising f rom or relating to the use (or misuse) of any inf ormation, statements, or
conclusions contained in or implied by the contents of this document or any of the source materials.

This document may contain links to third-party websites. CADTH does not hav e control ov er the content of such sites. Use of third-party sites is gov erned by
the third-party website owners’ own terms and conditions set out f or such sites. CADTH does not make any guarantee with respect to a ny inf ormation
contained on such third-party sites and CADTH is not responsible f or any injury , loss, or damage suf f ered as a result of using such third -party sites. CADTH
has no responsibility f or the collection, use, and disclosure of personal inf ormation by third-party sites.

Subject to the af orementioned limitations, the v iews expressed herein are those of CADTH and do not necessarily represent the v iews of Canada’s f ederal,
prov incial, or territorial gov ernments or any third party supplier of inf ormation.

This document is prepared and intended f or use in the context of the Canadian health care sy stem. The use of this document ou tside of Canada is done so at
the user’s own risk.

This disclaimer and any questions or matters of any nature arising f rom or relating to the content or use (or misuse) of this document will be gov erned by and
interpreted in accordance with the laws of the Prov ince of Ontario and the laws of Canada applicable therein, and all proceed ings shall be subject to the
exclusiv e jurisdiction of the courts of the Prov ince of Ontario, Canada.

The copy right and other intellectual property rights in this document are owned by CADTH and its licensors. These rights are protected by the Canadian
Copyright Act and other national and international laws and agreements. Users are permitted to make copies of this document f or non-commercial purposes
only , prov ided it is not modif ied when reproduced and appropriate credit is giv en to CADTH and its licensors.

About CADTH: CADTH is an independent, not-f or-prof it organization responsible f or prov iding Canada’s health care decision-makers with objectiv e ev idence
to help make inf ormed decisions about the optimal use of drugs, medical dev ices, diagnostics, and procedures in our health ca re sy stem.

Funding: CADTH receiv es f unding f rom Canada’s f ederal, prov incial, and territorial gov ernments, with the exception of Quebec.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 2
 Context and Policy Issues
Cannabinoids are psychoactive constituents found naturally in both the cannabis
(marijuana) plant and in the human body as endocannabinoids.1 Cannabinoid receptors
(CB1, CB2) are mainly expressed in the central nervous system (CB1) and in immune cells
(CB2).1,2 When CB receptors are activated, a variety of neurotransmitters are inhibited such
as acetylcholine, dopamine, and glutamate,1 making the cannabinoid receptor an attractive
pharmacologic target. Effects on cognition, memory, motor activity, pain perception, and
energy balance have all been ascribed to CB1 activation while CB2 activation may play a
neuroprotective role through reduction of inflammation.2 Because of these pharmacologic
effects, there is interest in cannabinoids as a potential treatment for dementia and its
2,3
behavioural symptoms.

To date, two oral synthetic cannabinoids have been marketed in Canada: nabilone 4 and
dronabinol,5 the latter of which was discontinued in 2012. 5,6 Dronabinol had been indicated
for the treatment of AIDS-related anorexia and severe nausea and vomiting from cancer
chemotherapy.7 Nabilone is indicated for the treatment of severe nausea and vomiting from
cancer chemotherapy.8 Currently, there is no Health Canada approved indication for the
use of cannabinoids in dementia.

Antipsychotics have historically been used off-label for treating dementia-related

behavioural symptoms in adults residing in long-term care, but the modest benefit of
treatment was found to be outweighed by the harms – including an increased risk of death
– for most people.3 Thus, the use of antipsychotics has generally been reserved for short-
term treatment of the most severe cases of aggression. 9,10 Identifying safer, more effective
alternatives to antipsychotics is a priority. One such potential alternative under investigation
is the use of cannabinoids. The objective of this report was to review the evidence base for
the use of cannabinoids in the treatment of behavioural symptoms in adults with dementia.

Research Questions
1. What is the clinical effectiveness of cannabinoids for the treatment of behavioural
symptoms in adults with dementia?

2. What are the evidence-based guidelines regarding the use of cannabinoids for the
treatment of behavioural symptoms in adults with dementia?

Key Findings
A total of four publications met the inclusion criteria: two systematic reviews and two
randomized cross-over trials. No evidence-based guidelines were identified.

The systematic reviews collectively included eight unique studies, which primarily studied
treatment with dronabinol. Nabilone treatment was limited to a single case report.
Interpretation of the findings presented narratively by the systematic reviews is hampered
by sparse reporting. In particular, it is difficult to discern much about the patient population
studied due to a lack of reporting detail about patient characteristics, including age, sex,

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 3
 severity of dementia, co-morbidities, concomitant medications, and setting (i.e., community
versus long-term care). Although the studies of dronabinol treatment are consistent in
reporting a reduction in behavioural symptoms, the exposure to treatment tended to be
short, and almost half of the studies had no comparator. Moreover, adverse event reporting
was limited to three of the eight studies, further complicating risk-benefit determinations.
The two randomized cross -over trials, despite better reporting, contribute little to the
evidentiary base as they were small, exploratory safety sub-studies of short-term treatment
exposure to a form ulation of THC unavailable in Canada. Thus, there remains a gap in the
evidence on the use of cannabinoids in the treatment of dementia, which currently makes
evidence-informed decision-making challenging.

Methods
Literature Search Methods
A limited literature search was conducted on key resources including Ovid Medline,
Embase, PubMed, The Cochrane Library, University of York Centre for Reviews and
Dissemination (CRD) databases and a focused Internet search. No methodological filters
were applied to limit retrieval by publication type. The search was limited to English
language documents published between January 1, 2012 and November 29, 2017.

Selection Criteria and Methods

One reviewer screened citations and selected s tudies. In the first level of screening, titles
and abstracts were reviewed and potentially relevant articles were retrieved and assessed
for inclusion. The final selection of full-text articles was based on the inclusion criteria
presented in Table 1.

Table 1: Selection Criteria

Population Adults with dementia

Intervention Cannabinoids (e.g., nabilone)

Comparator Q1: Other cannabinoids, no treatment, placebo

Q2: Not applicable

Outcomes Q1: Clinical effectiveness (e.g., improvement or reduction in behavioural symptoms, especially agitation and
aggression), safety
Q2: Evidence-based guidelines

Study Designs HTA/systematic reviews/meta-analyses, randomized controlled trials, non-randomized studies, evidence-
based guidelines

Exclusion Criteria
Articles were excluded if they did not meet the selection criteria outlined in Table 1, they
were duplicate publications, or were published prior to 2012. Individual studies included in a
selected systematic review were also excluded.

Critical Appraisal of Individual Studies

The included systematic reviews were critically appraised using the AMSTAR 11 checklist
while the Downs and Black12 instrument was used for critically appraising the included

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 4
 randomized studies. Summary scores were not calculated for the included studies; rather, a
review of the strengths and limitations of each included study were described narratively.

Summary of Evidence
Quantity of Research Available
A total of 315 citations were identified in the literature search. Following screening of titles
and abstracts, 288 citations were excluded and 27 potentially relevant reports from the
electronic search were retrieved for full-text review. No potentially relevant publications
were retrieved from the grey literature search. Of these potentially relevant articles, 23
publications were excluded for various reasons, while 4 publications met the inclusion
criteria (two systematic reviews 13,14 and two randomized controlled trials 15,16 and were
included in this report. No evidence-based guidelines were identified. Appendix 1 describes
the PRISMA flowchart of the study selection.

Summary of Study Characteristics

Two systematic reviews and two randomized, controlled, cross -over trials were identified
from the literature search. They are summarized below and detailed in Appendix 2.

Study Design
The two systematic reviews were both narrative reviews of the evidence. Lim et al.,14 which
only included randomized controlled trial (RCT) designs, broadly examined the efficacy of
medical cannabinoids in neurodegenerative and psychiatric conditions, including dementia
13
and Alzheimer’s Disease (AD). Liu et al., on the other hand, included RCTs, observational
studies, and case reports , and restricted their review to agitation and aggression in
14
dementia and/or AD. Lim et al. identified a total of 24 studies, four of which were relevant
to this report: three in dementia and one in AD, published between 1997 and 2015, and
covering approximately 90 patients. All four were cross -over in design. Liu et al.13 identified
a total of six studies in dementia and/or AD published between 1997 and 2014, and
covering approximately 84 patients. There were two cross -over designs, one open-label
pilot, one placebo-controlled study, one ‘retrospective study’, and one case report. There
were two constituent studies (both cross-over designs) common to both systematic reviews.
(Appendix 5). In this report, the findings from these two overlapping studies are presented
with the Lim et al.14 review.

The two primary studies 15,16 were both randomized, repeated cross -over, double-blind,
placebo-controlled, 12-week Phase II safety sub-studies derived from the same main trial,17
a constituent trial in the systematic review by Lim et al.14 The trial by van den Elsen et al.16
(n=18) specifically examined mobility-related safety outcomes while the trial by Ahmed et
al.15 (n=10) was a dose escalation study that investigated the pharmacokinetics,
pharmacodynamics, and adverse events from two different doses of study drug. Study
patients were all community-dwelling.

Country of Origin
14 13
Of the included systematic reviews, one was from Singapore while the other was from
Canada. The two included primary studies were both 15,16 from the Netherlands.

Patient Population

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 5
 Scarce information about patient characteristics was reported in the systematic reviews.
Lim et al.14 report a mean age from three of four studies that ranged from 72.7 years to
78.4 years; there was no information provided on sex distribution, severity of illness, co -
morbidities, concomitant medications, or dwelling status. The information was similarly
scant in Liu et al.,13 in which mean age was not reported; however, the retrospective study
13
included in Liu et al. describes studying ‘inpatients’.

The patient populations in the sub-studies by van den Elsen et al,16 and Ahmed et al.15
were similar, owing to their common source population. 17 Patients were exclusively
community-dwelling, predominantly white (94% and 90%) men (83% and 70%), with a
mean age of 77.0 ± 6 years and 77.3 ± 5.6 years , respectively, and a diagnosis of
Alzheimer’s dementia (83% and 90%). Baseline mean Mini Mental State Examination
(MMSE) score was 19.1 ± 6.0 and 18.5 ± 6.0, respectively, indicating moderate cognitive
impairment18 for both samples. In van den Elsen et al.,16 cholinesterase inhibitors and
psychotropic medications were taken concomitantly by 61% and 28% of patients,
respectively. No information was provided on the prevalence of concomitant medications in
Ahmed et al.15.

Interventions and Comparators

In Lim et al.,14 interventions included dronabinol 2.5 mg compared with placebo (two
studies) and tetrahydrocannabinol (THC) 0.75 mg to 1.5 mg compared with placebo (two
studies). In Liu et al.,13 interventions included dronabinol 2.5 mg to 7.0 mg (five studies)
compared with placebo (three studies), melatonin (one study), or no control (two studies);
and nabilone 0.5 mg to 1 mg (one study) with no control comparison.

The randomized cross -over trial by van den Elsen et al.16 compared THC 1.5 mg twice daily
with placebo while that of Ahmed et al.15 compared two doses of THC (0.75 mg, 1.5 mg
twice daily) with placebo.

Outcomes
A variety of psychometric instruments were used to measure changes in behaviour. The
most commonly employed instruments in Lim et al.14 were the Cohen-Mansfield Agitation
Inventory (CMAI, three studies) and the Neuropsychiatric Inventory (NPI, three studies). In
Liu et al.,13 the NPI was used in three studies along with actigraphy (two studies). Additional
psychometric instruments that were used less frequently are detailed for each systematic
review in Table 2 of Appendix 2.

In the randomized cross -over trial by van den Elsen et al.,16 various mobility-related (e.g.,
gait, balance) assessments were conducted alongside the documentation of adverse
15
events. Since the trial by Ahmed et al. was principally an investigation of the
pharmacokinetic and pharmacodynamic parameters of the study drug, this Rapid Response
report presents the adverse event data that were collected during the trial.

Summary of Critical Appraisal

The critical appraisal of the two systematic reviews and two randomized, controlled, cross-
over trials are summarized below and detailed in Appendix 3.

Both Lim et al.14 and Liu et al.13 provided a statement of their research question, conducted
a comprehensive literature search – though did not pursue a supplemental grey literature
search – and reported sources of funding for the systematic review. Although both tea ms
14
employed at least two reviewers to conduct the review, Lim et al. did not report how these

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 6
 reviewers were involved in the study selection or data extraction process; rather, only
described their participation in the risk of bias assessment. Liu et al.13 included information
about the study selection process, but did not describe the data extraction process;
moreover, there was no risk of bias assessment performed. Both teams provided a list of
included studies, but the patient characteristics for the included studies were minimally
described in both cases so the overall composition of the individual study populations was
unclear. Neither Lim et al.14 nor Liu et al.13 appear to have registered their systematic
review protocol on PROSPERO.19 There was no statement of conflict of interest provided in
Lim et al.14 A statement of conflict of interest was provided in Liu et al.,13 in which two of the
five researchers declared having received financial support, including from the
pharmaceutical industry; the other three researchers declared no conflicts of interest.

The randomized, double-blind, placebo-controlled cross-over trials by van den Elsen et al.16
and Ahmed et al.15 were sub-studies derived from the same main trial (n=22) .17 Although
these two sub-studies shared a rigorous methodologic design, including double-blind,
placebo control, randomized sequence allocation, appropriate allocation concealment and
washout between study treatments, and used recognized international clinical criteria for
diagnosing dementia, the sample size for each trial was small (n=18 and n=10,
15
respectively). In the case of Ahmed et al., there was no information provided on how the
10 patients were selected for sub-study participation. However, it is likely that the 10
patients represent the initial ‘hospital admission’ cohort, who were originally recruited into
the main trial before it was determined that patients could be safely followed on an
outpatient basis.17 In the case of van den Elsen et al.,16 it would appear that all patients
from the main trial were potentially eligible, if they were able to complete mobility-related
assessments.16,17 Patient characteristics at baseline were provided in both trials, but in the
case of Ahmed et al.,15 there was no information provided on the distribution of co -
morbidities and concomitant medications; neither trial provided baseline information on
severity of dementia. History of prior exposure to cannabis or cannabinoids, a potential
confounder, was not reported, despite both trials taking place in The Netherlands, where
20
cannabis is widely available. Moreover, the potential harms of treatment may be
underestimated in both trials if the study patients were not naïve to cannabis or cannabinoid
due to the effects of tolerance from prior exposure. 21 Both trials were considered
exploratory, and in the case of van den Elsen 16 there was no adjustment for multiple
comparisons, thereby incurring the risk of a Type I error or fals e positive result.

Summary of Findings
What is the clinical effectiveness of cannab inoids for the treatment of b ehavioural
symptoms in adults with dementia?

Two systematic reviews and two randomized cross -over trials meeting the inclusion criteria
for this report were identified from the literature search to address the clinical effectiveness
of cannabinoids for the treatment of behavioural symptoms in adults with dementia.

The two systematic reviews 13,14 included a total of 10 studies, two of which overlapped
between the reviews, leaving eight unique studies. Neither systematic review pooled the
included studies for a meta-analysis; rather, a narrative summary was provided by each
review. Of these eight studies, five used dronabinol, two used THC, and one used nabilone.
All five dronabinol studies, including two which had no comparator, and the nabilone case
study reported improved behavioural outcomes, while the two studies on THC reported no
improvement. Adverse events were reported for three of eight studies.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 7
 The systematic review by Lim et al.14 included four studies (three in dementia and one in
AD, all cross-over designs) published between 1997 and 2015, and covering 90 patients
(range: 2 to 54). Limited information was provided on patient characteristics, except for age:
mean age ranged between 72.7 years to 78.4 years based on three studies. No information
was provided on the patients’ dwelling status (i.e., independent-living versus long-term care
residency). Interventions consisted of dronabinol 2.5 mg daily versus placebo (two studies)
and THC 0.75 mg to 1.5 mg two to three times daily versus placebo (two studies). The
findings from the four studies on the outcome of behaviour change were mixed: two studies
using dronabinol found reduced ‘disturbed behaviour’ and nighttime agitation, respectively,
while two studies using THC found no improvement in neuropsychiatric s ymptoms (NPS).
Adverse events (AEs) were reported in two of four studies. One study of dronabinol
reported common side effects of anxiety, emotional lability, tiredness, and somnolence
while one study of THC reported common side effects of dizziness and so mnolence. There
was no information provided for either study on the number or frequency of AEs. All four
studies were given an overall rating of ‘unclear’ for risk of bias assessment.

The systematic review by Liu et al.13 comprised six studies, including two studies described
by Lim et al.14 All four unique studies were in AD, published between 2006 and 2014, and
covered 71 patients (range: 1 to 40). Study designs included a case report, a ‘retrospective
study’, an open-label pilot study, and a placebo-controlled study. Information about patient
characteristics was limited. From one constituent study, there was a specific line description
of ‘inpatients’ (n=40), which suggested acute hospitalization or residency in a long-term
care institution as the setting. In another study of six patients, the diagnosis of ‘late-stage
dementia’ provided a sense of disease severity. Otherwise, it was difficult to appreciate the
nature of the study population from the systematic review. More detail was provided on
interventions, which consisted of dronabinol in three studies and nabilone in one study
(case report); one of the four studies included a comparison group (placebo, melatonin).
The findings from the four studies on the outcome of behavior change were positive overall:
the three dronabinol studies reported reduced motor agitation and aggressiveness (one
study) and reduced nocturnal motor activity (two studies); the case report of nabilone
reported reduced severity of agitation in a single patient with AD and behavioral
disturbances. One of the four studies (retrospective study) reported a total of 26 AEs, which
included sedation, delirium, urinary tract infection, and confusion. There was no information
provided on the frequency of AEs. Unlike Lim et al.,14 no risk of bias assessment was
performed on the included studies.

The two included Phase II, randomized, double-blind, placebo-controlled, repeated cross-
over trials 15,16 were sub-studies of a main trial (n=22) by van den Elsen et al.17 The
commercial formulation of THC used in the sub-studies is not available in Canada.

In the 12-week sub-study (n=18) by van den Elsen et al.,16 THC 1.5 mg twice daily was
compared with placebo on mobility-related safety outcomes . Study patients were
exclusively community-dwelling, predominantly white (94%) men (83%), with a mean age of
77.0 ± 6 years, and a diagnosis of Alzheimer’s dementia (83%). Baseline mean Mini Mental
State Examination (MMSE) score was 19.1 ± 6.0, indicating moderate cognitive
impairment.18 Cholinesterase inhibitors and psychotropic medications were taken
concomitantly by 61% and 28% of patients, respectively. Following THC administration,
increased body sway was observed compared with placebo on the outcomes of static
balance (eyes closed condition) and dynamic balance (during preferred speed walking) .
Increased stride length was observed for gait (during preferred speed walking).
Notwithstanding the lack of statistical adjustment for multiplicity, the clinical meaningfulness

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 8
 of these observed changes is uncertain, given the lack of information available for minimum
clinically important differences for several parameters. Moreover, based on previous work
15
by the same research group, the 1.5 mg THC dose used may have been too low to
produce meaningful differences in outcomes. AEs were reported for the original sample of
22 patients, four of whom did not participate in the mobility assessments. The overall
incidence of AEs was similar between THC and placebo phases (91 versus 93, P = 0.77).
Dizziness (10 versus 9 events), somnolence (2 versus 2 events), and balance disorders (1
versus 0) were recorded as mobility-related AEs. Falls were less frequent during the THC
than the placebo phase (2 versus 4).

In the 12-week sub-study (n=10) by Ahmed et al.,15 THC (0.75 mg to 1.5 mg twice daily)
was compared with placebo on pharmacokinetic, pharmacodynam ic, and safety outcomes.
Study patients were exclusively community-dwelling, predominantly white (90%) men
(70%), with a mean age of 77.3 ± 5.6 years, and a diagnosis of Alzheimer’s dementia
(90%). Baseline mean MMSE score was 18.5 ± 6.0, indicating moderate cognitive
impairment.18 Unlike van den Elsen et al.,16 no information was provided on the prevalence
of concomitant medications. Since pharmacokinetic and pharmacodynamic characteristics
were not outcomes of interest for this report, only adverse events are presented. A total of
98 AEs were reported for the study period: 43 for the THC phase versus 55 for the placebo
phase. The distribution of AEs was similar between THC and placebo phases, regardless
the level of THC exposure. No THC-related severe AEs were reported. THC treatment was
not associated with changes in physical, laboratory, or ECG findings. Of the 13 reported
AEs that were deemed possibly or probably related to study drug, six were considered
possibly related to THC: dizziness (one patient: 0.75 mg dose), fatigue (two patients: 0.75
mg dose, one patient; 1.5 mg dose, one patient), and agitation (three patients: 1.5 mg
dose).

What are the evidence-b ased guidelines regarding the use of cannab inoids for the
treatment of b ehavioural symptoms in adults with dementia?

No evidence-based guidelines on the use of cannabinoids for the treatment of behavioural

symptoms in adults with dementia were identified from the literature search.

Limitations
This review is limited by the lack of information available on the use of cannab inoids in
dementia, and on the use of nabilone, in particular. A single case report using nabilone was
13
described within the systematic review by Liu et al. The remainder of this report’s
evidence base resides with the use of dronabinol and THC. Although it is unclear from the
systematic review reporting, in particular, it appears that the long-term care setting is
understudied. No evidence-based clinical guidelines on the use of cannabinoids for the
treatment of behavioural symptoms in adults with dementia were identified from the
literature search.

Neither systematic review pooled the included studies for meta-analysis, presumably due to
excessive clinical and/or methodological heterogeneity; rather, a narrative summary was
provided. For each systematic review, details of the constituent studies were scant,
particularly with respect to patient characteristics, limiting the interpretation of the data.
Although the two included randomized cross -over trials included community-dwelling
patients with dementia who were aged in their late seventies , the trials were small (n = 18,
n = 10), of short duration (12 weeks), and exploratory in nature (i.e., underpowered), and
only studied safety, not efficacy, outcomes. The dosing of THC (up to 1.5 mg twice daily)

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 9
 used in each trial was suspected to be sub-therapeutic because of the lack of
pharmacodynamic effects observed in one of the trials. Moreover, the specific formulation
of THC used in the trials is not commercially available in Canada.

Conclusions and Implications for Decision or Policy Making

In this report, the clinical effectiveness of cannabinoids for the treatment of behavioural
symptoms in adults with dementia was examined within an evidence base consisting of two
systematic reviews of eight unique studies and two randomized cross -over trials.
Dronabinol – which was discontinued from the Canadian market in 2012 5 – was the most
commonly-studied drug and was associated with reductions in behavioural symptoms.
However, exposure to treatment tended to be short and the study population inadequately
described. Small sample size, short follow-up, and the absence of a comparator in some
studies, along with a general lack of adverse event reporting further complicate any risk-
benefit determination. Thus, there remains a gap in the evidence on the use of
cannabinoids in the treatment of dementia, which currently makes evidence -informed
decision-making challenging. Currently, nabilone is the only synthetic cannabinoid
marketed in Canada,4 but is in short supply.22

In addition to the evidence synthesized in this report, two citations of research in -progress
were identified. One is a protocol for a Cochrane systematic review on cannabinoids for the
treatment of dementia,2 which has an estimated completion date of July 2018.23 The other
is a registered (Canadian) randomized cross-over trial studying the safety and efficacy of
nabilone in Alzheimer’s Disease, which has an estimated completion date of January
2018.24

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 10
References

1. Hill KP. Medical marijuana for treatment of chronic pain and other medical and psychiatric
problems: A clinical review. JAMA [Internet]. 2015 [cited 2017 Nov 30];313(24):2474-83.
Available from: https://jamanetwork.com/journals/jama/fullarticle/2338266

2. Markovic D, Bosnjak D, Brkovic T, Jeric M, Rubic Z, Vuica VA, et al. Cannabinoids for the
treatment of dementia. Cochrane Database Syst Rev [Internet]. 2017 [cited 2017 Nov
30];2017(10). Available from:
http://onlinelibrary.w iley.com/doi/10.1002/14651858.CD012820/epdf
3. Weier M, Hall W. The use of cannabinoids in treating dementia. Curr Neurol Neurosci Rep.
2017 Aug;17(8):56, 2017.
4. Nabilone. Current status: marketed. 2014 [cited 2017 Dec 21]. In: Health Canada Drug
Product Database [Internet]. Ottawa: Health Canada. Available from: https://health-
products.canada.ca/dpd-bdpp/index-eng.jsp DIN:02393581.
5. Marinol (dronabinol): cancelled from market. 2012 [cited 2017 Dec 21]. In: Health Canada
Drug Product Database [Internet]. Ottawa: Health Canada. Available from: https://health-
products.canada.ca/dpd-bdpp/index-eng.jsp DIN:00611190.
6. Cannabinoids: drug comparison chart. In: RxFiles drug comparisons charts [Internet]. 11th
ed. Saskatoon (SK): RxFiles Academic Detailing Program; 2017. p. 167.

7. Marinol (dronabinol (delta-9-tetrahydrocannabinol;  THC)) 2.5 mg and 5 mg capsules

[product monograph] [Internet]. Saint-Laurent (QC): Abbott Laboratories, Limited; 2011
Jan 1. [cited 2018 Jan 4]. Available from: https://pdf.hres.ca/dpd_pm/00013378.PDF

8. CPS [Internet]. Ottawa: Canadian Pharmacists Association; c2017. Cesamet® (nabilone)

[product monograph]; 2009 Mar 17 [cited 2018 Jan 4]. Available from: http://www.e-
therapeutics.ca
9. Panza F, Solfrizzi V, Seripa D, Imbimbo BP, Santamato A, Lozupone M, et al. Progresses
in treating agitation: a major clinical challenge in Alzheimer's disease. Expert Opin
Pharmacother [Internet]. 2015 [cited 2017 Nov 30];16(17):2581-8. Available from:
http://www.tandfonline.com/doi/full/10.1517/14656566.2015.1092520
10. Press D, Alexander M. Management of neuropsychiatric symptoms of dementia. In: Post
TW, editor. UpToDate [Internet]. Waltham (MA): UpToDate; 2017 Oct 23 [cited 2017 Dec
20]. Available from: www.uptodate.com Subscription required.

11. Shea BJ, Grimshaw JM, Wells GA, Boers M, Andersson N, Hamel C, et al. Development
of AMSTAR: a measurement tool to assess the methodological quality of systematic
reviews [Internet]. BMC Med Res Methodol. 2007 Feb 15 [cited 2017 Dec 13];7:10.
Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1810543/pdf/1471-
2288-7-10.pdf
12. Downs SH, Black N. The feasibility of creating a checklist for the assessment of the
methodological quality both of randomised and non-randomised studies of health care
interventions. J Epidemiol Community Health [Internet]. 1998 Jun;52(6):377-84. Available
from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1756728/pdf/v052p00377.pdf
13. Liu CS, Chau SA, Ruthirakuhan M, Lanctot KL, Herrmann N. Cannabinoids for the
treatment of agitation and aggression in Alzheimer's disease. CNS Drugs. 2015;29(8):615-
23.

14. Lim K, See YM, Lee J. A systematic review of the effectiveness of medical cannabis for
psychiatric, movement and neurodegenerative disorders. Clin [Internet]. 2017 Nov 30
[cited 2017 Nov 30];Psychopharmacol. Neurosci.. 15(4):301 -12. Available from:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5678490/
15. Ahmed AI, van den Elsen GA, Colbers A, Kramers C, Burger DM, Van Der Marck MA, et
al. Safety, pharmacodynamics, and pharmacokinetics of multiple oral doses of delta -9-
tetrahydrocannabinol in older persons with dementia. Psychopharmacology (Berl )
[Internet]. 2015 Jul [cited 2017 Nov 30];232(14):2587-95. Available from:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4480847/

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 11
 16. van den Elsen GA, Tobben L, Ahmed AIA, Verkes RJ, Kramers C, Marijnissen RM, et al.
Effects of tetrahydrocannabinol on balance and gait in patients with dementia: a
randomised controlled crossover trial. J Psychopharmacol. 2017;31(2):184-91.

17. van den Elsen GA, Ahmed AI, Verkes RJ, Feuth T, Van Der Marck MA, Olde Rikkert MG.
Tetrahydrocannabinol in behavioral disturbances in dementia: a crossover randomized
controlled trial. Am J Geriatr Psychiatry. 2015 Dec;23(12):1214-24.

18. Guidelines & Protocol Advisory Committee. Standardized Mini-Mental State Examination
(SMMSE). In: BC guidelines.ca [Internet]. Victoria (BC): Government of British Columbia;
2014 [cited 2017 Dec 21]. Available from:
https://www2.gov.bc.ca/assets/gov/health/practitioner-pro/bc-guidelines/cogimp-
smmse.pdf
19. PROSPERO: international prospective register of systematic reviews [Internet]. York (GB):
University of York. 2017 [cited 2017 Dec 20]. Available from:
https://www.crd.york.ac.uk/prospero/
20. Toleration policy regarding soft drugs and coffee shops [Internet]. The Hague (NL):
Government of the Netherlands; 2017. [cited 2017 Dec 20]. Available from:
https://www.government.nl/topics/drugs/toleration-policy-regarding-soft-drugs-and-
coffee-shops
21. Wilkinson ST, Radhakrishnan R, D'souza DC. A systematic review of the evidence for
medical marijuana in psychiatric indications. J Clin Psychiatry. 2016 Aug;77(8):1050-64.

22. Reports for Nabilone [Internet]. Ottawa: Drug Shortages Canada; 2017. [cited 2017 Dec
21]. Available from: https://www.drugshortagescanada.ca/ingredient/2350

23. Markovic D, Bosnjak D, Brkovic T, Jeric M, Rubic Z. Cannabinoids for the treatment of
dementia [Cochrane protocol]. In: PROSPERO: international prospective register of
systematic reviews [Internet]. York (GB): University of York; 2018 [cited 2018 Jan 3].
Available from:
https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=83364
24. Sunnybrook Health Sciences Centre. Safety and efficacy of Nabilone in Alzheimer's
disease: a pilot study. 2015 Jan 30 [cited 2017 Dec 20; Last updae posted: 2017 Oct 5].
In: ClinicalTrials.gov [Internet]. Bethesda (M D): U.S. National Library of Medicine; 2000 - .
Available from: https://clinicaltrials.gov/ct2/show/NCT02351882 Identifier:
NCT02351882.

25. Klumpers LE, Beumer TL, van Hasselt JG, Lipplaa A, Karg er LB, Kleinloog HD, et al.
Novel Delta(9) -tetrahydrocannabinol formulation Namisol(R) has beneficial
pharmacokinetics and promising pharmacodynamic effects. Br J Clin Pharmacol [Internet].
2012 Jul [cited 2017 Dec 20];74(1):42-53. Available from:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3394127
26. van den Elsen GA, Ahmed AI, Lammers M, Kramers C, Verkes RJ, Van Der Marck MA, et
al. Efficacy and safety of medical cannabinoids in older subj ects: a systematic review.
Ageing Res Rev. 2014 Mar;14:56-64.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 12
Appendix 1: Selection of Included Studies

315 citations identified from electronic

literature search and screened

288 citations excluded

27 potentially relevant articles retrieved

for scrutiny (full text, if available)

No potentially relevant
reports retrieved from
other sources (grey
literature, hand search)

27 potentially relevant reports

23 reports excluded:
-irrelevant intervention (1)
-already included in at least one of the
selected systematic reviews (5)
-other (17)
-review articles (13)
-study design (i.e., case report,
case series) (2)
-editorial (1)
-research protocol (1)

4 reports included in review

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 13
Appendix 2: Characteristics of Included Publications
Table 2: Characteristics of Included Systematic Reviews
First Author, Types and Eligibility Population Intervention and Clinical
Publication Numbers of criteria Characteristics Comparator(s) Outcomes
Year, Country Primary Studies
Included,
Objective,
Sample size,
Duration,
Setting
14
Lim, 2017, 4/24 relevant RCTs comparing Limited narrative Dronabinol 2.5 mg Change in: CMAI;
Singapore studies (all 4 cross- cannabis (any synthesis on twice daily versus NPI; Barthel index;
over design, 3 DB; form, any route of patient placebo (2 studies); QoL-AD; CCGIC;
AD = 1, dementia = administration, for characteristics: THC 0.75 to 1.5 mg ZBI; Nonparametric
3; n=90* [range: 2 medical use) with Mean age = 72.7 to two to three times circadian rhythm
to 54]); of the 4 placebo or other 78.4 years** daily versus analysis; Lawton
relevant studies, 2 active treatments; placebo (2 studies) observed affect
were also included any age; male or scale
in the review by Liu female; with either
et al.13 a movement
disorder, or
Objective: To neurological (e.g.,
evaluate the AD, dementia) or
efficacy of medical psychiatric
cannabinoids condition
across a range of
neurodegenerative
disorders and
psychiatric
conditions

Duration and
setting of included
studies not
reported.

13
Liu, 2015, 6 studies total (2 RCTs, Limited narrative Dronabinol 2.5 mg Change in: CMAI,
Canada cross-over, 1 DB; 1 observational synthesis on to 7.0 mg daily (5 NPI, PAS,
open-label pilot; 1 studies, or case patient studies); versus actigraphy
placebo-controlled; studies evaluating characteristics: placebo (3 studies),
1 retrospective; 1 cannabinoids for “A significant melatonin (1
case report); n=84* the treatment of portion of all study), or no
[range: 1 to 40]; of agitation and/or patients had used control (2 studies)
the 6 studies, 2 aggression in or were using
were also included dementia or AD psychoactive Nabilone 0.5 to 1
in the review by medication to mg daily (1 study);
Lim et al.14 manage their no control
symptoms.” (p.616)
Objective: To Two of the six
evaluate the studies included
evidence for patients with
cannabinoids in the ‘probable’ AD

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 14
Table 2: Characteristics of Included Systematic Reviews
First Author, Types and Eligibility Population Intervention and Clinical
Publication Numbers of criteria Characteristics Comparator(s) Outcomes
Year, Country Primary Studies
Included,
Objective,
Sample size,
Duration,
Setting
treatment of dementia or AD.
agitation and
aggression in No information
dementia and/or provided on age,
AD sex distribution, or
co-morbidities.
Total sample size
could not be
determined from
the data; study
duration and
setting not
reported.

AD = Alzheimer’s disease; CCGIC = Caregiver Clinical Global Impression of Change; CMAI = Cohen-Mansfield Agitation Inventory; DB = double
blind; NLD = The Netherlands; NPI = Neuropsychiatric Inventory; PAS = Pittsburgh Agitation Scale; QoL-AD = Quality of Life in Alzheimer’s Disease
scale; RCT = randomized controlled trial; THC = tetrahydrocannabinol; ZBI = Zarit Burden Interview

*It is unclear whether the reported sample size represents all enrolled patients or the subset of those who received study drug or successfully completed the
protocol.

**Based on reporting by three studies. (Sex distribution w as not reported in three studies.)

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 15
Table 3: Characteristics of Included Primary Studies
First Author, Study Design, Eligibility Population Intervention and Clinical
Publication Objective, criteria Characteristics Comparator(s) Outcomes
Year, Country Sample Size,
Duration,
Setting
16 RCT (sub-study Community- Mean age = 77.0 ± THC 1.5 mg twice Changes in: static
van den Elsen,
from same main dwelling adults 6 years; 83% male; daily versus and dynamic
2017, NLD
trial 17 as Ahmed et aged ≥ 18 years;17 94% Caucasian; placebo balance (body
al.15 sub-study); diagnosis of dementia type: sway); gait
repeated cross - Alzheimer’s, Alzheimer’s (83%), (velocity, stride
over, DB, placebo- vascular, or mixed vascular (6%), length, double
controlled, Phase dementia per mixed (11%); mean support time, base
II; two sites NINCDS-ADRDA MMSE score = of support); mobility
or NINDS-AIREN 19.1 ± 6.0; tasks (stance, gait)
Objective: To criteria; NPI score concurrent
evaluate mobility- ≥ 10; able to walk ≥ medications: Adverse events
related safety 10 m and cholinesterase
outcomes of low- understand simple inhibitors (61%),
dose oral THC in instructions;16 psychotropic (28%:
patients with informal caregiver antidepressant
dementia at least once [17%],
weekly15 benzodiazepine
N=18* [17%],
Exclusions: current antipsychotic
12 weeks major psychiatric [11%]); baseline
disorder; any mean gait velocity:
Tertiary care; severe or unstable 91.8 ± 20.4 cm/s
geriatric specialty concomitant illness;
outpatient clinics frequent falls due
15
to orthostatic
Ahmed, 2015, RCT (sub-study Mean age = 77.3 ± THC 0.75 mg twice Adverse events
hypotension;
from same main history of alcohol or 5.6 years; 70% daily, 1.5 mg twice
NLD 17
trial as van den drug abuse; current male; 90% daily versus
Elsen et al.16 sub- Caucasian; placebo
use of tricyclic
study); dose dementia type:
antidepressants,
escalation study; Alzheimer’s (90%),
opioids, or CYP
repeated cross - 2C9, 2C19, or 3A4 mixed (10%); mean
over, DB, placebo- inhibitors MMSE score =
controlled, Phase 18.5 ± 6.0
II; two sites

Objective: To
evaluate safety,
PD, PK of multiple
low doses of THC
in older persons
with dementia

N=10

12 weeks

Tertiary care;
geriatric specialty

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 16
Table 3: Characteristics of Included Primary Studies
First Author, Study Design, Eligibility Population Intervention and Clinical
Publication Objective, criteria Characteristics Comparator(s) Outcomes
Year, Country Sample Size,
Duration,
Setting
outpatient clinics; in
addition, 4 x 3-day
hospital admissions
to facilitate blood
sampling

DB = double blind; cm/s = centimetres per second; MMSE = Mini Mental State Examination; NINCDS-ADRA = National Institute of Neurological Communicativ e Disorders
and Stroke – Alzheimer’s Disease and Related Disorders Association; NINDS-AIREN = National Institute of Neurological Disorders and Stroke – Association
Internationale pour la Recherche et l’Enseignement en Neurosciences; NLD = The Netherlands; NPI = Neuropsy chiatric Inv entory; PD = pharmacody namics; PK =
pharmacokinetics; RCT = randomized controlled trial; THC = tetrahy drocannabinol
Mean ± standard dev iation
17 16
*Main trial enrolled 22 patients; f our patients were excluded f rom v an den Elsen et al. due to either inability to understand instructions (n=2) or complete mobility
assessments (n=2).

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 17
Appendix 3: Critical Appraisal of Included Publications
Table 4: Strengths and Limitations of Systematic Reviews and Meta-Analyses using
AMSTAR 11
Strengths Limitations
14
Lim et al. 2017
 A description of the research question and inclusion  There is no indication that the systematic review
criteria was provided. protocol was registered on PROSPERO.
 Two reviewers assessed each included study for  There was no information provided on how studies
methodological quality; however, discrepancies were were selected or how discrepancies in study selection
not resolved by a third reviewer, rather by discussion were resolved. There is no information provided on the
amongst the two reviewers. personnel involved in data extraction.
 A fairly comprehensive literature search was performed  A list of included studies was provided, but the
including a hand search of references lists from accompanying description of study characteristics was
retrieved articles. so minimal as to be unhelpful for appreciating the study
 Internal validity was assessed and documented using population.
the Cochrane risk of bias tool.  There does not appear to have been any supplemental
 The results of the quality assessment were grey literature search conducted. A language filter was
appropriately applied both in the (narrative) analysis applied to select English language publications.
and conclusions.  A list of excluded studies was not provided.
 Sources of funding were reported for the systematic  Due to presumed clinical and/or methodological
review, but not for the constituent studies. heterogeneity, the authors did not pool the studies for
meta-analysis; instead, a narrative summary was
presented.
 Publication bias was not assessed.

13
Liu et al. 2015
 A statement of the research question was provided.  There is no indication that the systematic review
 Three reviewers participated in study selection; protocol was registered on PROSPERO.
however, there was no information provided on how  Inclusion criteria were minimal in detail.
consensus was achieved.  A list of included studies was provided, but the
 A fairly comprehensive literature search was accompanying description of study characteristics was
performed. so minimal as to be unhelpful for appreciating the study
 Sources of funding were reported for the systematic population.
review but not for the constituent studies. Two  It is unclear whether the three reviewers involved in
researchers disclosed prior funding from various study selection likewise participated in data extraction.
pharmaceutical industry companies, foundations, and There was no information provided on how
government agency sponsors. discrepancies were resolved.
 There does not appear to have been any supplemental
grey literature search conducted. No hand-search was
performed. A language filter was applied to select
English language publications.
 A list of excluded studies was not provided.
 There was no internal validity (i.e., risk of bias)
assessment performed.
 Due to presumed clinical and/or methodological
heterogeneity, the authors did not pool the studies for
meta-analysis; instead, a narrative summary was
presented.
 Publication bias was not assessed.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 18
Table 5: Strengths and Limitations of Randomized Controlled Trials using Downs & Black12
Strengths Limitations
16
van den Elsen et al. 2017
 Double-blind, quadruple-masked (i.e., patient, care  Small sample size (n=18); 4 patients from the original
provider, investigator, outcomes assessor),17 placebo- sample (n=22) were excluded from mobility
controlled, cross-over (patients served as their own assessments due to comprehension or logistical
control) RCT design, in which the sequence of issues. By reason of exclusion and in the absence of
treatment was randomized; allocation concealment was data, it is possible that these 4 patients may be at
adequate. higher risk for falls compared with the remaining 18
 A washout period of 4 days was included in the patients.
protocol. Based on the prior PK study of oral THC,25 the  Patients were all community-dwelling, so findings
length of this washout period appears adequate. cannot be viewed in the context of the long-term care
 Outcomes were adequately described. setting.
 Use of objective instruments , supported by some  No information provided on the following patient
validity data, for assessing balance and gait (i.e., characteristics:
accelerometry, electronic walkway with sensors) o Baseline severity of neuropsychiatric
 Internationally recognized clinical criteria specified for symptoms, such as agitation (e.g., NPI score)
diagnosis of dementia type o Prior exposure to cannabis or cannabinoids.
 Intervention (THC 1.5 mg) and comparator (placebo)  Exploratory safety analysis, which did not adjust for
adequately described. multiple comparisons, thereby incurring risk of Type I
 Adverse events were collected for all patients (n=22), error (i.e., false positive result)
regardless of whether they underwent mobility
assessments (n=18)
 Protocol for main trial registered on ClinicalTrials.gov17

15
Ahmed et al. 2015
 Double-blind, quadruple-masked (i.e., patient, care  Small sample size (n=10) and no information provided
provider, investigator, outcomes assessor),17 placebo- on how these patients were selected from the main
controlled, cross -over (patients served as their own trial 17(n=22).
control) RCT design, in which the sequence of  No information provided on the following patient
treatment was randomized; allocation concealment was characteristics :
adequate. o Baseline severity of neuropsychiatric
 A washout period of 4 days was included in the symptoms, such as agitation (e.g., NPI score)
protocol. Based on the prior PK study of oral THC, 25 the o Distribution of co-morbidities and concomitant
length of this washout period appears adequate. medications
 Outcomes were adequately described. o Prior exposure to cannabis or cannabinoids
 Internationally recognized clinical criteria specified for  Patients were all community-dwelling, so the findings
diagnosis of dementia type cannot be viewed in the context of the long-term care
 Interventions (THC 0.75 mg, 1.5 mg) and comparator setting.
(placebo) adequately described. Helpful schematic of  Possibility that the dosing studied (TDD = 1.5 mg, 3.0
cross-over protocol included. mg) was sub-therapeutic, given the greater number of
 All patients completed the study. adverse events in the placebo versus THC phases and
 Protocol for main trial registered on ClinicalTrials.gov17 the smaller than expected pharmacodynamic effects
observed, per the investigators.
 The THC formulation studied is not commercially
available in Canada.
 Exploratory safety analysis , which limits interpretation
of findings.

NPI = Neuropsy chiatric Inv entory ; PK = pharmacokinetic; TDD = total daily dose; THC = tetrahy drocannabinol

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 19
Appendix 4: Main Study Findings and Author’s Conclusions
Table 6: Summary of Findings of Included Studies
Main Study Findings Author’s Conclusion
Systematic reviews
14
Lim et al. 2017
This narrative review covered 4 trials, including 2, which were “Although results were inconsistent, there appears to be some
also covered by Liu et al.13 and are presented here: low quality evidence for cannabinoids for… agitation in
Alzheimer’s disease and dementia…
Alzheimer’s Disease:
However, concrete conclusions of its efficacy could not be made
One small DB, placebo-controlled, cross-over RCT (n=12,* 92% due to the unclear risk of bias presented in these trials…
male) of dronabinol in patients aged 65 to 82 years reported
decreased severity of ‘disturbed behaviour’ (CMAI, P = 0.05) Methodological issues such as inadequate description of
and decreased negative affect (P = 0.045). However, this trial allocation concealment and blinding, varying cannabinoid
had an ‘unclear’ rating for risk of bias and has been criticized by formulations and doses, and small sample sizes limit its potential
others 13,26 for methodologic flaws such as not including a clinical utility” (p.310)
washout period in its study, which likely led to the observed time
by treatment (‘carry-over’) effect.13 Moreover, ‘disturbed
13
behaviour’ was not the primary outcome measure of this trial.

Dementia:

The results of three small trials (n=78*), each with a rating of

‘unclear’ for risk of bias, were mixed on the outcomes of
nighttime agitation (improved with dronabinol in one trial) and
NPS (no improvement with THC capsules in two other trials).

Two of four studies did not report adverse events. The single
trial of dronabinol in AD reported common side effects of
anxiety, emotional lability, tiredness, and somnolence; however,
frequency distribution was not reported. Similarly, one of the two
THC trials reported common side effects of dizziness and
somnolence but without frequency distribution.
13
Liu et al. 2015
This narrative review covered 6 studies; however, 2 were also “The small number of studies in this review highlights the need
covered by Lim et al.14 including the single (positive) trial in AD for further randomized controlled trials to evaluate the safety and
and the (positive) dronabinol trial in dementia. These two efficacy, including the habituation and potential for abuse, of
overlapping trials are pres ented above. cannabinoids for the treatment of agitation and aggression in
severe dementia and AD.” (p. 621)
The results from the remaining 4 studies (n=71*) in AD, which
included a case report, a ‘retrospective study’, an open-label
pilot study, and a placebo-controlled study are presented here:

The case report described reduced severity of agitation with

nabilone. The retrospective study reported reduced motor
agitation (z = –4.4423, P < 0.0001) and aggressiveness (z =
–3.9102, P < 0.0001) with dronabinol. The open-label pilot study
of late-stage dementia reported reduced nocturnal motor activity
(P = 0.028), with an average decrease of 59% (range: 13% to
85%) compared with baseline during the first two days of
dronabinol treatment. The placebo-controlled study in patients

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 20
Table 6: Summary of Findings of Included Studies
Main Study Findings Author’s Conclusion
with ‘probable’ AD dementia reported a 16% reduction in
nocturnal motor activity (statistical significance not reported) with
dronabinol compared with baseline.

Three of four studies did not report adverse events. The

retrospective study recorded 26 AEs during dronabinol treatment
including sedation, delirium, urinary tract infection, and
confusion. No data on frequency were reported.

Randomized Controlled Trials

16
van den Elsen et al. 2017
This trial is a substudy (along with the substudy by Ahmed et “These first results suggest that low-dose oral THC is well
15 17
al. ) of the main trial by van den Elsen et al. tolerated by community-dwelling dementia patients concerning
mobility and risk of falling. This dose [1.5 mg twice daily] did not
Results presented for the subset of 18 patients who underwent show benefit** in the treatment of dementia-related NPS
mobility assessments: compared to placebo.” (p.189)

Static b alance (standing, eyes open or closed condition)

 Eyes open: No differences in body sway between THC and
placebo
 Eyes closed: Increased roll angle, pitch angle, pitch velocity
after THC versus placebo (0.32 [0.6]°/s, P = 0.05; 1.04
[1.5]°/s, P = 0.009; and 1.96 [3.3]°/s, P = 0.02, respectively)

Dynamic b alance (assessed during preferred speed walking)

 PSW alone: Increased pitch angle displacement after THC
versus placebo (1.18 [1.6]°, P = 0.005)
 PSW + cognitive dual task: No effect of THC on dynamic
balance

Gait (assessed during preferred speed walking)

 PSW alone: Increased stride length (4.3 [5.4] cm, P =
0.005) after THC versus placebo
 PSW + cognitive dual task: No effect of THC on gait

Adverse events
AEs were reported for the original sample of 22 patients
(including the 4 patients who did not participate in mobility
assessments).There was no difference between the THC and
placebo phases in the overall incidence of AEs (91 versus 93, P
= 0.77). Dizziness (10 versus 9 events), somnolence (2 versus 2
events), and balance disorders (1 versus 0) were recorded as
mobility-related AEs. Falls were less frequent during the THC
than the placebo phase (2 versus 4).
15
Ahmed et al. 2015
This trial is a substudy (along with the substudy by van den “Our data demonstrate that THC doses of 0.75 and 1.5 mg twice
16 17
Elsen et al. ) of the main trial by van den Elsen et al. daily are safe and well tolerated by older individuals with
dementia [based upon short-term use].” (p.2592)
Safety results presented for the subset of 10 patients, all of
whom completed the substudy.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 21
Table 6: Summary of Findings of Included Studies
Main Study Findings Author’s Conclusion
Adverse events
A total of 98 AEs were reported for the study period: 43 for the
THC phase versus 55 for the placebo phase. The distribution of
AEs was similar between THC and placebo phases, whether
THC exposure was 0.75 mg (21 versus 30, P = 0.290) or 1.5 mg
(22 versus 25, P = 0.435) THC twice daily. No THC-related
SAEs were reported. THC treatment was not associated with
changes in physical, laboratory, or ECG findings. Of the 13
reported AEs that were deemed possibly or probably related to
study drug, 6 were considered possibly related to THC:
dizziness (1 patient: 0.75 mg dose), fatigue (2 patients: 0.75 mg
dose, 1 patient; 1.5 mg dose, 1 patient), agitation (3 patients: 1.5
mg dose).

AD = Alzheimer’s Disease; AE = adv erse ev ent; CMAI = Cohen-Mansf ield Agitation Inv entory ; DB = double-blind; ECG = electrocardiography ; NPS = neuropsy chiatric
sy mptoms; PSW = pref erred speed walking; RCT = randomized controlled trial; s = second; SAE = serious adv erse ev ent; THC = tetrahy drocannabinol
*It is unclear whether the reported sample size represents all enrolled patients or the subset of those who receiv ed study drug or successf ully completed the protocol.
14
**Ef f icacy results from the main trial cited within this sub-study were published separately and were included in the sy stematic rev iew by Lim et al. summarized in this
report.

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 22
Appendix 5: Overlap between Included Systematic Reviews
Table 7: Overlap in Studies on Dementia or Alzheimer’s Disease between the
Included Systematic Reviews
Primary Study Systematic Review Citation
Citation 14 13
Lim et al. 2017 Liu et al. 2015
Volicer et al. 1997 ■ ■
Walther et al. 2011 ■ ■
a
van den Elsen et al. 2015 ■
b
van den Elsen et al. 2015 ■
Walther et al. 2006 ■
Mahlberg & Walther 2007 ■
Passmore 2008 ■
Woodward et al. 2014 ■

SUMMARY WITH CRITICAL APPRAIS AL Cannabinoids f or Behav ioural Sy mptoms in Adults with Dementia 23