The present report is in line with pre-clinical and preliminary clinical data suggesting that CBD (and perhaps Δ9-THC) may be effective for some patients with epilepsy. However, the adverse events in both children, including Δ9-THC intoxication and increased seizures, highlight the need for accurate dosing standards and quality assurance of cannabis-derived medications. The use of edible products makes it difficult to titrate doses and to establish the actual amount of cannabinoids across different products or different batches of a same product. Moreover, when taken orally, Δ9-THC leads to the synthesis of much larger amounts of 11-OH-THC compared to when it is ingested through smoking. Since 11-OH-THC is also active, it is possible that its psychopharmacological effects may combine with those of Δ9-THC to produce stronger effects in the CNS (Benjamin and Fossler, 2016). This is of particular concern when we consider that the pharmacological potency of Δ9-THC is much higher than that of CBD, and thus the amount of Δ9-THC required to produce an effect (as well as side-effects) at a given intensity is much lower than the amount of CBD. Our patients, for instance, who received 208 mg and 250 mg of CBD-enriched extract with 4.03% and 3.1% of Δ9-THC, would ingest 8.3 mg and 7.5 mg of Δ9-THC, which can be considered a very high dose (Benjamin and Fossler, 2016), especially for children.

Surveys with caregivers and patients have also examined the effects of CBD-enriched extracts in epilepsy. One of such surveys involving parents of children with severe epilepsies in a Facebook group reported improvements in 16 out of 19 patients treated with CBD/Δ9-THC-enriched extracts, and two patients were reported to have become seizure-free (Porter and Jacobson, 2013).

A retrospective case series of 75 children with refractory epilepsy taking oral cannabis extracts reported that 25 (33%) patients presented >50% reduction in seizure frequency, whereas 44% had adverse events including increased seizures (13%) and rare events such as developmental regression, abnormal movements, status epilepticus requiring intubation, and death (Press et al., 2015). More recently, a retrospective study described the successful effects of CBD-enriched medical cannabis in 74 children (1–18 years old) with refractory epilepsy from five Israeli pediatric epilepsy clinics (Tzadok et al., 2016). CBD-enriched treatment significantly reduced the frequency of seizures in most of the children (66/74, 89%) and only five (7%) patients reported seizure exacerbation that led to the discontinuation of treatment with CBD-enriched extracts. Despite this, minor and infrequent side effects were reported.

There are anecdotal reports describing the antiepileptic effects of crude cannabis (Friedman and Devinsky, 2015), but evidence of seizure exacerbation and lack of effects is also available (Tofighi and Lee, 2012; Hamerle et al., 2014). Since C. sativa contains many other substances, with different cannabinoid concentrations and ratios, there is no sufficient information to allow for definitive conclusions (Rosenberg et al., 2015). Moreover, the pharmacokinetic interactions of CBD and Δ9-THC (and of other cannabinoids) raise safety concerns, as both compounds can inhibit cytochrome P450 enzymes (CYP2C and CYP3A4) that act in the metabolism of many of the anticonvulsants commonly used with CBD (Geffrey et al., 2015).

Some therapeutic trials have tested the effects of isolated cannabinoids in the treatment of epilepsy. In an early prospective, placebo-controlled three-month trial in adults with treatment-resistant epilepsy, Mechoulam and Carlini (1978) showed that two out of four patients treated with purified CBD (200 mg/day) became seizure free and one presented partial improvement. Later, a prospective, placebo-controlled trial in adolescents and adults with treatment-resistant convulsive seizures treated for 8–18 weeks reported that four out of eight subjects in the purified CBD (200–300 mg/day) arm became seizure free (Cunha et al., 1980). More recently, in an open-label trial (Devinsky et al., 2016), patients (N = 162, 1–30 years-old) with severe childhood-onset refractory epilepsy were enrolled in an expanded access program. Participants were given purified oral CBD (from 2 to 5 mg/kg/day to a maximum up-titrated dose of 25 mg/kg/day) during 12 weeks. The authors reported that CBD reduced seizure frequency and presented an adequate safety profile.

The pharmacological mechanisms underlying the antiepileptic action of CBD are not yet well understood, as CBD interferes with different neurotransmitter systems in a number of ways. For instance, there is evidence that CBD inhibits the reuptake and metabolism of anandamide, increases hippocampal neurogenesis, interacts with 5HT1A and TRPV1 receptors, and presents antioxidant and neuroprotective effects which may, at least in part, help to explain its anticonvulsant profile (Reif et al., 2007; Zuardi, 2008; Campos and Guimarães, 2008; Wolf et al., 2010). Moreover, the analysis of endocannabinoid signaling elements and related proteins in lymphocytes of patients with Dravet syndrome showed changes in the alpha-1H unit of the voltage-dependent calcium channel and an up-regulation of CB2 receptors, associated with an activation of lymphocytes and changes in inflammation-related genes (Rubio et al., 2016). These changes have been described in inflammatory diseases and epilepsy and may support a potential dysregulation of the endocannabinoid system in the SNC.

We cannot exclude the possibility that the case presentations may have been impacted by the side effects and pharmacokinetic interactions of the other medications the children were taking, or even by other chemical substances that could be present in the cannabis extracts initially used. However, neither the cannabinoids nor the other medications are known to induce mydriasis or eye redness and the patients presented improvements with CBD.

The two cases reported here highlight the need for cannabis-derived compounds produced according to good laboratory and manufacturing practices (GMP/GLP) that could lead to the development of medicines consistently produced and controlled by international regulatory standards. The use of handcrafted, uncontrolled cannabis-based medicines by children and adolescents raises special concerns because of the well-known harmful long-term effects of Δ9-THC on the developing brain (Volkow et al., 2014). The potentially toxic effects of Δ9-THC and other cannabis constituents, including cognitive impairment and chronic psychiatric disturbances, have not yet been well studied in younger patients who might be more vulnerable than adults to potential long-term adverse effects (Volkow et al., 2014). Therefore, randomized clinical trials using high-quality and reliable cannabis-derived substances without other impurities (e.g., solvents) are required to ascertain the safety and efficacy of cannabinoids as medicines.