According to our survey, almost half of the participants have used or are using CBD for treatment of epilepsy in children and adolescents. The rising CBD prescription rate within the last years found in our survey corelates with a growing interest in medical cannabis in general and CBD in particular (1). However, the clinical evidence to support its use is scarce and the individual experience is mostly limited to a few patients per practitioner. Interestingly, even practitioners outside of specialized epilepsy centers und neuropediatric departments are using CBD in their patients. Given the lack of individual in depth experience in treating children and adolescents with CBD, a more centralized treatment approach may be suggested and treatment data should be prospectively collected.

CBD has been proven to be effective in acute and chronic seizure models in rodents (2, 3). However, sound clinical data to prove the efficacy of CBD, depending on the epilepsy type, are limited (15). Results of two randomized controlled trials on CBD as add-on anticonvulsant in patients with Dravet and Lennox-Gastaut syndrome have been published recently. In both trials CBD was efficient to reduce convulsive or drop seizures frequency respectively (10, 11).

Open label (8, 9), observational (16) and retrospective studies (5–7) indicate that other forms of intractable epilepsies may also respond to CBD. However, these observations are limited for various reasons (1). Consistently, structural or genetic epilepsies other than Dravet and Lennox-Gastaut syndrome were not considered to be treatable with CBD by most respondents of our survey. Unsurprisingly, given the limited data, survey participants would rather treat only patients with proven pharmacoresistancy, when other treatment options are exhausted or do not appear appropriate.

Randomized controlled trials primarily focus on seizure control as a main outcome criteria, implicating seizure control to be the primary treatment goal (10, 11). However, parental surveys and case reports suggest that from the patient's perspective, other treatment effects of CBD, e.g. improved sleep or behavior may be regarded as an equally beneficial treatment goal (5, 7, 9). Aforementioned aspects may impair the quality of life as much as the seizures themselves (17) and may be responsive to a CBD treatment (18). Interestingly therefore, in our participants opinion, improving the patient's quality of life was even more important than seizure reduction when initiating a CBD treatment. However, this may be in part due to the limited practitioner's expectation concerning seizure control in the respective patient population and we cannot provide information about decision making based on individual patient data. Nevertheless, increasing patient's quality of life appears to be a major motive for initiating a CBD treatment and should be included as outcome criteria in future CBD trials (19).

The two major neuroactive components of the cannabis plant out of more than hundred different cannabinoids are delta9-tetrahydrocannabinol (THC) and CBD (20). The latter cannabinoid is used by most participants. While the approval of the European Medicines Agency (EMA) is pending for a purified CBD pharmaceutical, purified CBD is currently only available as an individual pharmacy-dispensed preparation in some European countries or even prohibited in others. This may be the rationale to use over-the-counter preparations as reported by some participants, however, the reasons therefore are not elucidated by our survey. CBD enriched hemp oils are mixtures of cannabis extracts, presumably with a higher CBD content, and are freely available as dietary supplement. However, CBD concentrations in those preparations are not standardized. Quality controls revealed low CBD concentrations, potentially impacting anticonvulsive efficacy of the preparation, and revealed increased concentrations of THC in a large proportion of hemp oil samples (21). Therefore, when intending to initiate a CBD based treatment, there is currently no alternative available for individually dispended purified CBD preparations.

Preclinical data on THC prove anticonvulsant (22), as well as proconvulsive (23) and adverse effects on neurocognitive functioning (24). Clinical data show adverse structural and functional effects resulting from long term THC use (25, 26). Nevertheless, the use of THC containing preparations is still reported in our survey. Unlike CBD, the combination of THC and CBD (e.g. Sativex®) and dronabinol, a synthetic THC preparation (e.g. Marinol®), are EMA approved pharmaceuticals. This may facilitate its utilization and the approval of reimbursement from health care providers. Since clinical data concerning the safety and efficacy of THC or hemp oils for treatment of childhood epilepsy are not available, these preparations cannot be recommended.

A wide CBD dose range from < 10 mg/kg to 50 mg/kg body weight per day was reported by our survey participants. The dose that proved to be effective in two randomized controlled trials was 20 mg/kg, but open label trials suggest that individual patients required much higher doses up to 50 mg/kg for a response (8). However, dosing may depend on several factors, e.g. epileptic etiology, seizure types, age of the patient and comedications (7, 8, 27). Given the lack of data, an upper permissible dose cannot be defined and CBD doses up to 50 mg/kg were found to be safe (8, 28). Long term data of the use of CBD are equally limited (29). Therefore, there is no defined timeframe that qualifies as CBD treatment failure and this is reflected by the participant's divergent opinions about when to discontinue a CBD treatment.

The majority of participants in our survey would reduce clobazam when initiating a CBD treatment. While cytochrome P450 inhibiting CBD has shown relevant interactions with other anticonvulsant drugs in preclinical studies (30–32), the only clinically relevant interaction consists of increasing plasma concentrations of the active clobazam metabolite N-desmethyl-clobazam (8, 33–35). This interaction is claimed to be partially accountable for the anticonvulsant effect of CBD but also for increased sedation (8, 27, 33). Elevated aminotransferase levels during CBD treatment are reported to result almost exclusively from a concomitant valproate therapy (10, 33, 34). Nevertheless, liver enzymes are routinely measured by most participants regardless of comedication. Purified CBD preparations increase serum levels in a dose-dependent manner (33). However, pharmacokinetics of different CBD preparations or different routes of application and correlations between CBD serum level and anti-seizure effects or side effects have not been fully established (36, 37). Accordingly, almost none of the participants measure CBD serum levels.

There are several surveys of patients' or parents' opinions regarding CBD treatment for epilepsy (38–41), but only one that is including health care professionals. In that survey 52% of epileptologists and general neurologists would advise against using medical marijuana even in severe cases of epilepsy (14). In contrast, according to our survey the vast majority is using or would consider using CBD under certain circumstances, and only a minority would refrain from using it (n = 25) CBD. This may indicate a change of attitude concerning a cannabinoid-based treatment for epilepsy among practitioners treating children and adolescents for epilepsy.

According to our survey, most participants receive inquiries about CBD treatment on a regular basis. Interestingly, only about 40% would recommend CBD treatment actively whereas the majority discusses this treatment option only on request by patients or their parents. In our experience, families searching for alternative treatment options are burdened by a long history of seizures, side effects and complications. In an Australian nationwide survey on medicinal cannabis use for epilepsy, a significant proportion of children and adults with epilepsy were commencing cannabis-based products without medical supervision, even resulting in unsupervised reduction of concomitant antiepileptic treatment (41). Therefore, according to our survey some participants would rather commence a CBD treatment even if not convinced about its efficacy than having patients using it without medical supervision. However, decision to commence CBD treatment is mainly made on individual case-by-case basis and most patients of our participants are treated outside of clinical trials.

Interestingly, a substantial number of answers about their country specific availability and regulations concerning reimbursement of CBD were contradictory. This is in line with the results of an US survey about the use of medical cannabis in cancer patients. Only 5% of pediatric oncologists knew their state-specific regulations in this regard (42). This may indicate a further need for providing adequate information to health care providers given the striking differences between European countries in regulation, availability and covering of costs of medical cannabinoids in general and CBD in particular (12).

Given the widely differing practice concerning indications and limitations, choice of preparation, dosing and monitoring revealed by our survey, official guidelines for the use of CBD for epilepsy treatment appear to be advisable to harmonize and potentially improve its use.

There are several limitations concerning our survey. Since this was an open-access survey we cannot generate a response rate. We relied onto the participants to reply truthfully and thoroughly. Numbers and percentages of CBD prescribers may be overestimated by a participation bias, that may be indicated by a substantial variation of responses between countries. Furthermore, we were not able to relate the numbers of participants to the total number of physicians that are treating children and adolescents with CBD in participating countries. Therefore, we cannot to draw a representative picture for the extent of CBD use in participating countries. These aspects need to be considered when interpreting our findings.

Nevertheless, we presented a broad overview of certain aspects of CBD use by European experts of childhood epilepsy and highlighted several limitations for its use in clinical practice.