Medical Cannabis in Children and Adolescents in Switzerland: a Retrospective Study


 Background: Evidence on use and efficacy of medical cannabis for children is limited. We examined clinical and epidemiological characteristics of medical cannabis treatment and caregiver-reported effects in children and adolescents in Switzerland.Methods: We collected clinical data from children and adolescents (<18 years) who received Δ9-tetrahydrocannabinol (THC), cannabidiol (CBD), or a combination of the two between 2008 and 2019 in Switzerland. Results: Out of 205 contacted families, 90 agreed to participate. The median age at the first prescription was 11.5 years (interquartile range (IQR) 6-16), and 32 patients were female (36%). Fifty-one (57%) patients received CBD and 39 (43%) THC. Patients were more likely to receive THC therapy if one of the following symptoms or signs were present: spasticity, pain, lack of weight gain, vomiting, or nausea, whereas seizures were the dominant indication for CBD therapy. Improvements were reported in 59 (66%) study participants. The largest treatment effects were reported for pain, spasticity, and frequency of seizures in participants treated with THC, and for those treated with pure CBD, the frequency of seizures. However, 43% of caregivers reported treatment interruption, mainly because of lack of improvement (56%), side effects (46%), the need for a gastric tube (44%), and cost considerations (23%).Conclusions: The effects of medical cannabis in children and adolescents with chronic conditions are unknown, but the caregiver-reported data reported here may justify trials of medical cannabis with standardized concentrations of THC or CBD to assess its efficacy in the young.

Synthetic or natural cannabis-containing preparations with more than 1% THC are regulated as narcotic drugs in Switzerland (15). In 2011, a revision of the Swiss Law on Narcotics and Psychotropic Substances (Narcotics Law) authorized the Federal O ce of Public Health to issue exceptional licenses for the medical use of substances containing more than 1% THC (16). In contrast, pure CBD-containing preparations do not require exceptional authorizations for the prescription to patients (16,17). At present, several oral THC-containing preparations are

Study design and data collection
This is a retrospective observational study that collected data from patients' families. We obtained a list of all children and adolescents under 18 years of age who received medical THC-or CBD-containing preparations between 2008 and June 2019 from the 'Bahnhof Apotheke Langnau AG' pharmacy in Langnau, Switzerland. The children and adolescents who received medical cannabis preparations from the pharmacy in Langnau came from all regions in Switzerland. Until end of April 2019, the pharmacy in Langnau is one of two pharmacies in Switzerland authorized by the authorities to provide medical cannabis, as prescribed by the treating physician. The pharmacy is a pioneer and has a leading role in the production of medical cannabis. The patient list included details on the date of birth, sex, medical diagnosis, date of prescriptions, the preparations including the concentration of CBD or THC, and the patients' and the prescribing physicians' contact details.
We developed and piloted a standardized questionnaire in collaboration with an advisory panel that consisted of two pharmacists and three clinicians prescribing medical cannabis. We piloted the German and French version of the questionnaire with caregivers for completeness and understandability. The questionnaire had three sections: (i) basic information such as sex, age, diagnosis, medications other than medical cannabis, and non-pharmacological therapies; (ii) details on the medical cannabis therapy including the symptoms triggering therapy, type of medical cannabis, initial and current or last dosage, side effects, treatment interruptions, and treatment effects and (iii) costs of cannabis therapy, including coverage of the expenses by health insurance or out of pocket. We assessed treatment effects using a Likert scale with options ranging from 'much less', 'less', 'no change', 'more', to 'much more'. The survey was provided in paper form or electronically in a REDCap application (19).
We sent all children and adolescents' caregivers an invitation letter with information on the study, the informed consent form, and the questionnaire with a prepaid return envelope. We sent non-responders another questionnaire 4-6 weeks later. In the event of continued nonresponse, we contacted families by phone. Depending on the caregivers' preference, either a caregiver or the adolescent (≥14years old) or both lled in the questionnaire.

De nitions
We assigned children and adolescents to two groups: those treated with a pure CBD preparation, and those treated with a THC-containing preparation (with or without low concentrations of CBD). Patients treated with both THC and pure CBD were analysed in the THC group. The medical cannabis preparations were standardized for THC or CBD

Statistical analysis
We used descriptive statistics to characterize patients and treatment effects and assessed differences between groups using chi-square, Fisher's exact, or Wilcoxon rank-sum tests. We assessed changes in the dosage of products using the paired t-test. All analyses were done in Stata (version 15.1, College Station, TX, USA).

Ethics
The Cantonal Ethics Committee Bern (2019-00049), Switzerland approved this study. Written informed consent was obtained from the caregivers of each child or adolescent younger than 14 years. Among older adolescents, either the adolescent or the caregivers gave informed consent.

Results
The patient list of the pharmacy included data on 205 children or adolescents who were treated with THC or CBD containing preparations between February 2008 and June 2019. The rst prescription to a child or adolescent was in 2013. Initially, we received 77 responses of the contacted caregivers. After the reminders, 90 caregivers (43.9%) agreed to participate in the study. Figure 1 shows the recruitment into the study. We compared the characteristics of the 90 participating children or adolescents to the 115 patients who did not participate. There were no differences in age and the type of medical cannabis prescribed. Compared to non-responders, participants were more likely to be male, to have more than one ICD-10 diagnosis, to have a disease of the nervous system or an endocrine, nutritional and metabolic disease, and have multiple prescriptions of medical cannabis products. According to the ICD-10 classi cation, the most common diseases among the 205 children or adolescents were diseases of the nervous system (120; 59%), mental and behavioural disorders (26; 13%), cancer (15; 7%), congenital malformations, deformations and chromosomal abnormalities (15; 7%), and endocrine, nutritional, and metabolic diseases (11; 5%). For 34 (17%) the diagnosis was missing (Table S1).
Fifty-one participants (57%) were treated with pure CBD and 39 (43%) with a THC preparation. Six patients who received both THC and pure CBD were included in the THC group. THC was more commonly prescribed to participants with cancer (p=0.03), whereas pure CBD was more frequently prescribed to participants with epilepsy (p<0.001). Participants were more likely to receive THC therapy if one of the following symptoms or signs were present: spasticity, pain, lack of weight gain, loss of appetite, vomiting, or nausea, whereas seizures were the dominant indication for pure CBD therapy. The daily dosage of medical cannabis preparations increased over time for both THC and pure CBD preparations ( Figure 2).

Treatment interruption and side effects
During medical cannabis treatment, 39 of the 90 participants (43%) reported a treatment interruption ( Table 2). Among them, 22 stopped treatment de nitively, and 17 resumed treatment (six continued with the same preparation and dosage, seven continued with the same preparation but a different dosage, four continued with another preparation). The median time from treatment initiation to treatment interruption was eight weeks (IQR 3-32 weeks). The reasons given for the treatment interruption included lack of improvement in 22 patients (56%), side effects in 18 (46%), the need for a gastric tube in 17 cases (44%) preventing the continuation of treatment, and cost considerations in 9 patients (23%, Table 2). Side effects were observed in 25 of the 90 participants (28%) and similar in the THC and CBD group. The three most common side effects were tiredness, sedation, and dry mouth ( Table 2).

Awareness, prescription, and cost modalities
Caregivers learned about medical cannabis therapy through the media (44; 49%), their family doctor or medical specialist (37; 41%), and friends or family members (21; 23%). In most cases (82; 91%) specialistsneuropaediatricians, neurologists, oncologists, or palliative care specialists-prescribed the preparation. The cost of the rst prescription was reimbursed by the invalidity insurance (an insurance covering some chronic diseases such as epilepsy or cerebral palsy), in 50 participants (56%), and the health insurance covered the cost for ten patients (11% , Table 3). For 27 participants (30%), caregivers paid out of their pocket. This situation persisted during the treatment. The monthly cost was below 300 USD for 24 participants (27%), between 301 to 600 USD for 22 (24%), more than 600 USD for 25 (28%). The cost was unknown for the remaining 19 children or adolescents. Many caregivers were concerned about the high cost of medical cannabis preparations. Table S2 gives further details about costs.

Treatment effects of medical cannabis preparations
In 59 of 90 participants (66%), the treatment with medical cannabis was reported to be successful by the caregivers ( Table 2). Participants treated with products containing THC most frequently reported a reduction of pain, spasticity, seizures, and a reduction in the number of drugs taken. Participants treated with pure CBD containing products most frequently reported a reduction in the frequency of seizures. Irrespective of treatment with THC or CBD, caregivers felt that their children or adolescents were more relaxed, more satis ed, and in a better general condition than before the therapy with medical cannabis (Figure 3, Table S3). The use of other therapies like physiotherapy, osteopathy, speech, or occupational therapy did not change, regardless of whether the patients received THC or CBD (Figure 3).

Discussion
The use of medical cannabis for the treatment of a variety of diseases among children and adolescents is increasing. We found that THC was most frequently used to treat pain, spasticity, seizures, lack of weight gain, and nausea, while pure CBD was used to treat pain, seizures, and sleep disorders. The largest treatment effects were reported for pain, spasticity, and frequency of seizure in participants treated with THC, and for those treated with pure CBD, the frequency of seizures. Irrespective of treatment with THC or CBD, or a preparation containing both, the majority of caregivers felt that their children and adolescents were more relaxed, more satis ed, and had increased quality of life with medical cannabis treatment. The treatment was reported to be successful in twothirds of participants. Treatment interruption was frequent, mainly due to of lack of improvement, side effects, or cost considerations.
In our study, 80% of our participants suffered from a disease of the nervous system. Among them, the most common diagnosis was epilepsy, for which 55% of caregivers reported a reduction in seizures and 6% reported seizure aggravation while on medical cannabis. In a similar study examining treatment with cannabidiol-enriched cannabis in children with treatment-resistant epilepsy, 84% of caregivers reported a reduction in the frequency of their children's seizures (22). Another retrospective study based on clinical records showed that children with intractable epilepsy treated with cannabis oil reported a reduction in the frequency of seizure in almost 90% of children, along with aggravation of seizures in 7% (23). Beyond observational studies, randomized controlled trials (RCT) were conducted in children with treatment-resistant epilepsy to assess the e cacy and tolerability of CBD compared to placebo. These studies con rmed that CBD reduces seizure frequency (8-11). A recent systematic review and meta-analysis summarized results from intervention studies and concluded that CBD is more effective than placebo for treatment-resistant epilepsy, regardless of the aetiology of the epileptic syndrome (24).
The second most common indication for the use of medical cannabis was spasticity. Among our participants, 49% -mainly those with cerebral palsy -reported a reduction of spasticity. A case series of 16 children, adolescents or young adults with resistant spasticity in palliative care who received 2.5% dronabinol showed a reduction spasticity (25). An intervention study in children with complex motor disorders who were treated either with cannabis oil with a 20:1 or 6:1 CBD to THC formulation showed similar improvements in both groups in spasticity, sleep di culties, pain severity, and quality of life (26). However, a recent multicentre RCT (2020) in 72 children or adolescents with cerebral palsy or another central nervous system injury after birth found no signi cant difference in the spasticity between a Cannabis extract (Nabiximols) and placebo groups after 12 weeks of treatment (3). Today the evidence that medical cannabis has an impact on spasticity in children is weak. In adults with multiple sclerosis or paraplegia, a systematic review of RCTs showed some evidence supporting the e cacy of medical cannabis in spasticity (6, [27][28][29]. The use of medical cannabis in children and adolescents poses risks. In our study, we found that medical cannabis was prescribed for a variety of conditions, even though the evidence is weak for many conditions. Although there is some evidence in adults supporting the e cacy of medical cannabis for some diseases, we should not extrapolate results from adults to children. Limited evidence exists for the effective use of different cannabis derivatives, the dosage, and indications. Indeed, the high rate of treatment interruption or stop (43%) seen in our study was driven by side effects and lack of improvement. Thus, clinicians should closely monitor children and adolescents on medical cannabis for e cacy and adverse effects, and they should be experienced in the treatment of the underlying disease. Little is currently known about both the interaction of medical cannabis with other drugs and long-term outcomes that may be particularly consequential for children. Treatment guidelines are needed to inform decision making by clinicians and caregivers. Although such guidelines are available in a few countries (30-32), they currently do not exist in Switzerland and many other countries. Another reason for stopping treatment was the costs for medical cannabis which have to be paid out of the caregivers' pocket. The production of quality-controlled medical cannabis is laborious. The costs for other medical and non-medical treatments in these children, which will generally be covered by health insurance, might be substantially higher (33). This is one of the rst studies reporting on prescription practices, dosages, and treatment outcomes in a representative sample of children and adolescents using medical cannabis in a real-life setting. Two strengths particularly distinguish our study from other studies. Participants were recruited from patients under 18 years of age who received medical cannabis over the entire period since the beginning of its use in Switzerland. The pharmacy in Langnau is the largest professional distributor of medical cannabis in Switzerland. Also, even though the dosage of medical cannabis preparations varied, the concentrations were standardized. Other publications and recommendations often lack such standardization, which hampers comparisons across studies complicated.
The most important limitation of our study was the lack of a comparison group. Another limitation derives from the fact that the caregivers reported the data. This is common in paediatrics, where outcomes are often reported by caregivers, especially in children with disabilities. A further limitation is that of potential bias arising from group differences between those who participated and those who did not. Participant health and the situations of participating families may have been better than those of non-participants. How this could have in uenced our ndings is unknown.

Conclusions
In Switzerland, medical cannabis was mainly prescribed to children and adolescents with neurological diagnoses, particularly in participants with epilepsy. Although evidence supporting e cacy is lacking, medical cannabis was prescribed to children and adolescents for a variety of other conditions. For two-thirds of participants treated with standardized THC or CBD preparations, the caregiver reported an improvement in their condition and well-being. However, the remaining stopped because of lack of effectiveness or side effects. Medical cannabis could be a promising and useful therapy to improve the condition of those children and adolescents. However, we need further RCTs with standardized THC and CBD preparations to assess the e cacy of medical cannabis in different diseases and long-term effects in the young.

Competing interests
MF is owner of the Bahnhof Apotheke Langnau AG, which produces and distributes medical cannabis-based preparations to patients upon medical prescription. DE works as pharmacist in the same pharmacy and is a board member of the Swiss Society of Cannabis in Medicine. However MF and DE had no role in study design and data analysis. All other authors declare that they have no con icts of interest.

Code availability
The collected data and the datasets used for this study are available from the corresponding author on reasonable request.

Authors' contributions
KZ conceptualized and designed the study, developed the questionnaire, coordinated the study, collected, entered, and analysed the data. KZ drafted the initial manuscript, and reviewed and revised the manuscript. LF Table 1 Basic characteristics of the 90 included participants.    Changes of the prescribed medical cannabis preparation from initial to current dosage by patients for whom initial and current dosage was available.