Efficacy and safety of paediatric medicinal cannabis use: A scoping review

Abstract

Introduction

The use of medicinal cannabis in the paediatric age group is increasing despite the lack of evidence for its efficacy or safety.

Objective

To map the available evidence on the efficacy and safety of medicinal cannabis in children and adolescents.

Methods

We conducted a scoping review and searched six electronic databases and grey literature. A study was eligible for inclusion when it investigated the efficacy or safety of medicinal cannabis for any condition, more than half of the participants were 0 to 18 years old, and had any study design except single case reports.

Results

We included 36 studies in our final analysis, 32 of which investigated the efficacy or safety of cannabis in treatment-resistant epilepsy. The remaining 4 studies examined patients with cancer, dysautonomia, Epidermolysis Bullosa, and motor disorders.

Conclusions

There is a lack of evidence on the efficacy and safety of medicinal cannabis in most paediatric conditions.

As medicinal cannabis is legalized in an increasing number of jurisdictions, more adolescents and caregivers of children and adolescents are turning to cannabis as a treatment option as they learn about it from other caregivers, the media, or social media (1). For clinicians, it can be challenging to address the efficacy and safety of medicinal cannabis use in children and adolescents due to the lack of evidence. A recent systematic review examined the evidence for medicinal cannabinoids in paediatrics, but some evidence may have been missed by the focused search utilized, and as this is a fast-moving field, there is already substantially more evidence than when the review was published in 2017. As we are not aware of any other well-designed broad reviews for efficacy and safety of medicinal cannabis in paediatrics, we determined another knowledge synthesis was warranted (2). Our aim in this scoping review is to map the available evidence regarding the efficacy and safety of medicinal cannabis in the paediatric population.

METHODS

Search methods

We searched Ovid MEDLINE, Embase, CINAHL, PsycInfo, Web of Science Core Collection, and Google Scholar until September 19, 2018. See Supplementary Appendix 1 for our full MEDLINE search strategy. To locate grey literature sources, we searched clinical trial registries (ClinicalTrials.gov, WHO International Clinical Trials Platform, EU Clinical Trial Register, OpenTrials), theses and dissertations (Networked Digital Library of Theses and Dissertations, ProQuest Dissertations and Theses Global, Open Access Theses and Dissertations), and conference proceedings from key conferences identified by the study team. We also searched the websites Charlotte’s Web, GW Pharmaceuticals, and the 36 producers currently licensed in Canada to sell or produce cannabis oil. Although including grey literature materials has the potential to introduce bias, we addressed this by looking for articles that were clear in their methodology (e.g., blinded randomized trials). The references of all included studies were hand-searched and Web of Science was used to search the citing articles of each study. We did not limit by publication date or language and used a paediatric search filter to limit to a paediatric population (3).

Study inclusion

For inclusion, a study must have (1) investigated either the efficacy and/or safety of medicinal cannabis in any condition; (2) a study population where at least 50% were between 0 and 18 years or the paediatric data were reported separately; and (3) any study design except single case reports. When paediatric data were reported separately, we excluded adults from our analysis. We excluded semisynthetic/synthetic tetrahydrocannabinoids because of their differing chemical profile. Additionally, their efficacy and safety has already been well established as antiemetics for chemotherapy-induced nausea and vomiting elsewhere (4,5). We focused this scoping review instead on emerging treatments using medicinal cannabis.

Study selection

All search results were imported into Endnote^© and duplicates were removed. The titles and abstracts, followed by the full text, of the articles were screened independently and in duplicate by two team members using Rayyan (6). In both the screening and review phases, where the two team members did not agree, the team met to come to a consensus decision to include or exclude. Data were extracted independently and in duplicate by two team members from the included full text articles using a piloted data extraction form. We extracted data on the sample size and age of the paediatric participants, condition(s) treated by cannabis, study design, preparation of the cannabis, outcome measures, major findings, and adverse events reported. We further categorized studies by the type of cannabis product investigated: (1) pharmaceutical grade products that are produced through a method of extraction in which a highly purified form of cannabinoid is produced with minimal by-products under government drug production guidelines (e.g., Epidiolex); (2) nonpharmaceutical standardized extracts that follow Good Manufacturing Practices within a natural products regulatory framework, such as Health Canada or the US Food and Drug Administration, to produce extracts with known cannabidiol (CBD) and tetrahydrocannabinol (THC) concentrations (e.g., CanniMed Ltd.); (3) nonstandard products with CBD and THC concentrations that are not reliably and/or transparently tested and likely contain multiple bioactive compounds (e.g., ‘artisanal’ or homemade products); or (4) unclear, where studies did not, or were not able to, report the type of cannabis product. For the purposes of this review, we will use medicinal cannabis as an umbrella term to describe all products. To report our scoping review, we used the PRISMA guidelines (7).

RESULTS

Our database and grey literature search retrieved 10,993 results. The removal of duplicates resulted in 8,717 references, of which 8,325 were excluded for not meeting the inclusion criteria. The resulting 392 full-text articles were reviewed and 356 were excluded. This left 36 studies for the final analysis. See Figure 1 for a PRISMA flowchart of the study selection process and reasons for exclusion during the full-text review.

Figure 1.

PRIMSA flow chart of study selection.

Study characteristics

The included studies were published between 2009 and 2018, with 26 (72%) of the studies published in 2017 or 2018. The study designs were 14 open-label, 7 cross-sectional surveys, 6 retrospective chart reviews, 5 randomized controlled trials (RCTs), 3 case series, and 1 semistructured interview. Thirty-two of the studies investigated cannabis use in epilepsy, while neuromotor disorders, Epidermolysis Bullosa, dysautonomia, and cancer each had 1 study. Eighteen of the studies used pharmaceutical grade CBD (14 of these used an extract containing predominately cannabidiol from a single producer); 4 studies used nonpharmaceutical standardized extracts; 14 studies used nonstandardized products or were unclear as to the chemical content of cannabis. A supplementary table summarizes the characteristics of the included studies (Supplementary File).

Drug-resistant epilepsy

There have been 32 studies to date investigating the efficacy of medicinal cannabis for seizures in drug-resistant epilepsies in children and adolescents. These studies range from case studies to RCTs.

Four RCTs showed that the use of CBD resulted in a greater reduction in seizures compared to placebo (8–11). Thiele et al. found that in patients with drug-resistant epilepsy in Lennox-Gastaut syndrome, median drop seizure frequency decreased by 21.1% compared to placebo (8). Devinsky et al. found that the use of CBD reduced the median amount of seizures by up to 21.6% compared to placebo (9). A RCT comparing doses of CBD in Dravet syndrome found children experienced more adverse drug reactions than placebo, but all doses were well tolerated (11). Common adverse drug reactions seen across these RCTs included somnolence, pyrexia, decreased appetite, diarrhea, vomiting, and upper-respiratory tract infection. See Table 1 for the most common adverse drug reactions. All RCTs investigated pharmaceutical grade CBD in doses ranging from 5 to 20 mg/kg/day.

Table 1.

Common adverse events in studies with >10 children with drug-resistant epilepsy

Adverse event	Frequency (%) n=2,051
Somnolence	384 (19.42%*)
Diarrhea	303 (14.77%)
Decreased appetite	213 (10.39%)
Increase in seizure frequency/new seizures	207 (10.09%)
Infections	184 (8.97%)
Fatigue	137 (6.93%*)
Vomiting	127 (6.19%)
Pyrexia	99 (4.83%)
Status epilepticus	74 (3.61%)
Abnormal liver function tests	67 (3.27%)
Increased appetite	62 (3.02%)
Weight gain	48 (2.34%)

*1,977 patients included in this calculation due to limited reporting.

Thirteen of the 32 studies were open-label trials that reported improvements in seizure frequencies, seizure durations, and clinical questionnaire ratings. Decreases in seizure frequencies and durations were found regarding all seizure types and specifically for motor seizures in studies investigating pharmaceutical grade CBD in doses ranging from 5 to 25 mg/kg/day and CBD/THC (20:1) with 4 to 32 mg/kg/day of CBD (12–21). Chen et al. found that the use of CBD in addition to other antiepileptic drugs improved ratings in the Global Impression of Change assessment from both caregivers and physicians (20). Hausman-Kedem et al. found that quality of life was also improved in 100% of patients with drug-resistant epilepsy in Sturge-Weber syndrome (22). One study used nonpharmaceutical standardized extracts of CBD/THC (50:1) with a dose of 7 to 16 mg/kg/day of CBD and found a reduction in motor seizures in 70.6% of children (23). The most observed adverse events amongst these studies were the following: somnolence, fatigue, diarrhea, anorexia, vomiting, respiratory infection, decreased appetite, and movement disorders. Gaston et al. identified that CBD interacts with the metabolism of certain antiepileptic drugs (24).

Six cross-sectional surveys collected caregiver reports of treatment with CBD (25–30). Four studies showed either a ≥50% or >50% reduction in seizures in 40 to 68% of patients (26–29). Two studies investigated nonpharmaceutical standardized extracts at doses of less than 1 to 11 mg/kg/day, one study included both nonpharmaceutical standardized and nonstandardized products at a dose of 2.9 to 7.7 mg/kg/day although some doses were unknown. Two studies investigated nonstandardized products of CBD/THC with an unknown daily dose and less than 0.5 to 28 mg/kg CBD, respectively and in two studies, the preparation was unclear and the dose was not reported. Two studies also noted improvements in neurocognitive and behavioural symptoms such as sleep, mood, and language (28,29). Common adverse events reported were increased or decreased appetite, sleep irregularities, drowsiness, and fatigue.

Six retrospective chart reviews supported findings from cross-sectional surveys as well, of which one investigated pharmaceutical grade CBD with a dose of 3 to 22 mg/kg/day, one investigated nonpharmaceutical standardized extracts of CBD/THC (20:1) at doses between 1 and 20 mg/kg/day, two investigated nonstandardized products at unknown doses, and two had unclear preparations at unknown doses. CBD led to a >50% reduction in seizure frequency in 4 to 51% of the patients (31–36). Multiple reports noted improvements in neurocognitive and behavioural symptoms (32–35). Somnolence, gastrointestinal symptoms, and irritability were common adverse reactions reported in these six studies.

Two qualitative semistructured interviews reported decreases in seizure frequency and length with nonstandardized and unclear products at unknown doses (1,37). Similarly, a case report reported on two children with treatment refractory epilepsy with pharmaceutical CBD treatment at doses of 6 or 12 mg/kg/day replacing nonpharmaceutical standardized extracts of CBD/THC (38). The parents reported neurocognitive behavioural benefits of the treatment and improvements in general behaviour, speech, understanding, and attention. No adverse events were found.

Other conditions

The remaining four studies each investigate a different condition. Li and Rassekh report the cases of two children with cancer who presented with hypotension as a result of parent administration of a nonstandardized, homemade CBD/THC oil of an unknown concentration and dose (39). The parental indications included anticancer effect and/or symptom relief. In an open-label study, Palmieri et al. report improvements in quality of life in 12 girls with dysautonomia after HPV vaccination when treated with unknown doses of nonpharmaceutical standardized CBD enriched hemp oil (40). The study reported adverse events of sleepiness and confusion in one patient. Chelliah et al. report three cases of parental initiated use of topical nonstandardized CBD oil in children with Epidermolysis Bullosa at an unknown dose (41). The parents reported fewer blisters, shorter healing time and less pain. In a RCT, Libzon et al. reported that nonpharmaceutical standardized CBD:THC oil (20:1 or 6:1) with mean doses between 3-73-5.53 mg/kg/day led to improvements in spasticity, gross motor function, and quality of life in children with motor disorders due to cerebral palsy, neuro-genetic syndromes, and traumatic brain injury (42). Adverse events included worsening of seizures, behavioural changes, and somnolence.

DISCUSSION

Despite a limited amount of studies available at this time, trials of medicinal cannabis in a paediatric population demonstrate benefits in seizure control for children with drug-resistant epilepsy. However, this scoping review has identified possible safety concerns. In the studies to date, discontinuation of treatment due to adverse events ranged from 0.17 to 20.00%. The RCTs had lower rates, with discontinuation due to adverse events between 1.82 and 7.02% in the CBD groups and 0.47 and 5.55% in the placebo groups. Common short-term adverse events reported in these studies included pyrexia, vomiting, diarrhea, and dizziness (8–13,15,16,19,21,23,36). Longer-term adverse events were also observed in the studies, including weight gain or loss, changes in appetite, fatigue, somnolence, changes in mood, and increase in seizure frequency (8–15,17,19–21,23,25–29,32–35,37,42). While decreased appetite, anorexia, diarrhea, drowsiness, fatigue, movement disorders, somnolence, and vomiting were common adverse events amongst all product types, some adverse events reported were specific to the pharmaceutical and nonpharmaceutical preparations. Notably, pharmaceutical preparations led to increase in seizures, status epilepticus, and weight loss (14,15,40), while nonpharmaceutical preparations resulted in depression and mood changes, increased appetite, weight gain, and memory loss (21,23,25–29,33,37). A recent meta-analysis also found that patients treated with CBD for drug-resistant epilepsy had fewer adverse drug reactions with nonpharmaceutical standardized extracts than with pharmaceutical grade products, possibly due to lower doses (43). In addition, CBD may interact with anticonvulsants such as clobazam by altering cytochrome-dependent metabolism, leading to increased side effects including loss of appetite and diarrhea, yet at the same time potentially enabling lower anticonvulsant doses with positive outcomes (10,14,36). These potential drug–drug interactions require careful monitoring when using CBD. Many caregivers of children with drug-resistant epilepsy wish to trial medicinal cannabis to manage their child’s symptoms. Although evidence exists that pharmaceutical, nonpharmaceutical, and nonstandardized extracts of cannabinoids may be efficacious, clinicians should be aware of the adverse events that have been reported with the use of medicinal cannabis.

This scoping review demonstrates that there is evidence to support the use of medicinal cannabis in Dravet and Lennox-Gastaut Syndrome. Emerging evidence has been presented to support its use in other syndromes co-morbid with drug-resistant epilepsy such as CDKL5 deficiency, Aicardi-Goutierre syndrome, Dupq15, Doose syndrome, Sturge-Webber and Tuberous Sclerosis. For indications other than drug-resistant epilepsy, there is currently a lack of evidence for the use of medicinal cannabis. Our review found no studies in paediatrics regarding pain, despite pain being a well-studied area in adults (44,45). Our review highlights the need for well-designed studies looking at dosing, efficacy, as well as short- and long-term safety outcomes for other paediatric indications. The availability of pharmaceutical grade products, the legalization of cannabis in many jurisdictions, and the increasing number of publications using high-quality trials designs, means that there will be opportunities to create a more substantial evidence base for nonepilepsy conditions in the near future.

A previously published systematic review also highlighted the evidence for drug-resistant epilepsy and the lack of evidence for other conditions (2). Our review searched more sources, including Web of Science, Embase, and grey literature and we used a broader set of search terms and used a combination of controlled vocabulary and keywords when available. While there are many articles found in both searches, ours found several more not previously identified due to our differing study design. Excluding synthetic cannabinoids and single case reports, Wong et al. identified 9 studies, while we identified 36. Sixteen of our 36 studies were published after May 2017 when Wong et al. completed their searches of databases, but an additional 11 studies may not have been identified by their focused search or may have been excluded due to differing inclusion criteria. Nine of our studies investigated drug-resistant epilepsy and their inclusion in our review provides a more complete evidence base for this condition (15,17,18,21,26,27,32,36,38). The remaining two studies are in cancer and dysautonomia, which is the only evidence available in both conditions (39,40). Additionally, our review discussed the safety and possible adverse events that can be a result of different product preparations of medicinal cannabis.

Limitations of this review include that our database search was in September 2018, in the context of a rapidly evolving literature, alongside real-world changes in legislation and increasing public use of cannabis (46,47). In this situation, continuous updating of reviews will be required. Additionally, we were not able to locate the full text of some identified studies that as a result were excluded from our review. Nevertheless, this scoping review captures the evidence supporting the use of CBD in drug-resistant epilepsy. It also highlights the danger of extrapolating to other conditions, given the profound gaps in evidence for the use of medicinal cannabis in other paediatric conditions.