Page 4 - The Use of Cannabis and Cannabinoids in Treating Symptoms of Multiple Sclerosis: a Systematic Review of Reviews

8 Page 4 of 12 Curr Neurol Neurosci Rep (2018) 18:8

system [21]. To enable an assessment of the evidence Five reviews were graded as 1+ in the SIGN grading sys-
contained in the reviews, we also rated the individual studies tem. This represents ‘well-conducted meta-analyses, system-
included in the review according to the GRADE criteria [22]. atic reviews, or randomised control trials with a low risk of
Where reviews reported an assessment quality metric for each bias’. Six reviews had a SIGN grading of 1− comprising ‘me-
study, this was considered in the assessment. As per the ta-analyses, systematic reviews, or randomised control trials
GRADE rating, randomised controlled trials (RCTs) were with a high risk of bias’. Quality as rated with the AMSTAR
considered high quality evidence, downgraded RCTs (for rea- scale ranged from 2 to 10 out of a possible 11, with a mean
sons of bias, sample size or other issues around design) were score of 6. The review covered studies published between
considered moderate quality evidence, double-downgraded 1981 and 2013, and the reviews themselves were published
RCTs (e.g. downgraded two levels from high to low quality between 2006 and 2016. The reviewed studies assessed a
because of multiple concerns with study design or bias) or range of cannabinoids including tetrahydrocannabinol
observational studies were considered low quality evidence, (THC), cannabidiol (CBD), THC:CBD formulations, pharma-
and triple-downgraded RCTs, downgraded observational ceutical cannabinoids (dronabinol and nabilone), smoked
studies, case series or case studies were considered very low Cannabis sativa plant material and oral cannabinoid extracts
quality evidence (see Supplementary Table 4). (Table 2).
Evidence was examined on eight pre-specified outcomes.
Data Collection Details by outcome domain are provided below. A summary
of the review evidence on the eight outcomes is presented in
Data were extracted from reviews using a standardised data Table 2.
extraction tool implemented in a custom-built Microsoft
Quality of the Evidence Contained in the Reviews
Access database. The data extracted included details about
the interventions, populations, study methods and outcomes
of interest. Data extraction tools were piloted and reviewed by Overall, 32 published reports were identified from the 11 sys-
the study authors before the results of the extraction were tematic reviews. Of these, four provided very low quality ev-
finalised. idence, 17 provided low quality evidence, 9 provided moder-
ate quality evidence and two publications from one larger
RCT (> 300 people) judged to have a low risk of bias provided
Data Synthesis
high quality evidence (see Appendix 3).
Review findings were synthesised to highlight when multiple
Disability and Disability Progression
reviews arrived at the same or different conclusions and to
describe the strength of the evidence in each case. We synthe-
Six reviews reported data relating to disability or disease pro-
sised findings by generating a set of statements that represent-
gression using different scales and outcome measures [24, 25,
ed the findings according to their quality and the similarity in
29, 30••, 31, 33] (see Supplementary Table 5). Overall, the
review conclusions.
effects of cannabinoids on disability and disease progression
were mixed. Reviews did not report conclusions on this out-
come, or focus on disability and disease progression as the
Results primary outcome.

Results are presented grouped by cannabinoid types. Where Pain
reviews did not identify studies that reported on outcomes
measure for a specific cannabinoid product, the gaps in the Seven reviews reported on a range of cannabinoids for the
evidence are also indicated in Table 2. treatment of pain in patients with multiple sclerosis [24, 26,
28•, 29, 30••, 31, 33] (see Supplementary Table 6). Although
Description of Reviews the effects were mixed, reviews presented evidence that most
cannabinoids reduced pain on at least some measures.
Eleven reviews met the eligibility criteria (see Fig. 1; Table 1). Two reviews of medium quality (AMSTAR score 4 and 5
Two Cochrane reviews were identified that focused on ataxia out of 11) concluded that there was an evidence that THC and
and tremor [23] and spasticity [24]. The remaining nine sys- THC:CBD/nabiximols were efficacious or probably effica-
tematic reviews focused on multiple sclerosis [25, 26], move- cious in reducing pain or painful spasm in multiple sclerosis
ment disorders more broadly [27, 28•, 29], or included studies [28•, 29]. Some reviews concluded that there was insufficient
of multiple sclerosis as part of more comprehensive reviews of evidence or mixed findings [24, 26, 33]. One review cited a
the therapeutic uses of cannabinoids [30••, 31–33]. non-significant meta-analysis of 3 studies (565 participants)