Cannabidiol: Pharmacology and potential therapeutic role in epilepsy and other neuropsychiatric disorders

Use in epilepsy in the modern era

In the late 19th century, prominent English neurologists including Reynolds⁹ and Gowers¹⁰ used cannabis to treat epilepsy (see Text Box A). However, the use of cannabis for epilepsy remained very limited, and despite anecdotal successes, cannabis received scant or no mention from English-language epilepsy texts in the late 19th and early to mid-20th centuries.

Four controlled studies, mainly in the 1970s, examined the effect of CBD on seizures (Table 1, reviewed in¹¹). However, while two of the studies found limited improvements, all four suffered from methodological flaws, including small sample size and, in some cases, inadequate blinding.

One epidemiologic study of illicit drug use and new-onset seizures found that cannabis use appeared to be a protective factor against first seizures in men¹². The adjusted odds ratio (OR) was 0.42 for every cannabis use and 0.36 for cannabis use within 90 days of hospitalization. No effect was observed in women. The authors suggested that cannabis is protective of both provoked and unprovoked seizures, for men.

Whole cannabis or extracts

Preclinical studies, mainly in the 1970s, studied cannabis’s effects on seizure and epilepsy. In a rat maximal electroshock study (MES), cannabis resin (17% Δ⁹-THC content) was used with or without pharmacological modulation of monoamines and catecholamines (which did not independently affect seizure parameters) to suggest that modulation of serotonergic signaling contributed to the anticonvulsant effects of cannabis²⁵. However, because the non-Δ⁹-THC cannabinoid composition of the cannabis was unknown, the potential contributions of other cannabinoid or non-cannabinoid components were likewise unexplored.

In a dog model using a subconvulsant dose of penicillin (750,000 IU; i.v.), acute smoked cannabis (6 mg Δ⁹-THC via tracheotomy) caused muscular jerks, while repeated treatment produced epileptiform activity in occipital and frontal cortices that generalized to tonic-clonic seizures²⁶. Here, the authors suggested that Δ⁹-THC either reduced seizure threshold or increased BBB permeability, although results of a related study did not support the latter hypothesis²⁷.

Δ⁹-THC

Many early studies on the effects of specific cannabinoids in preclinical models of seizures focused on Δ⁹-THC and, later, synthetic CB₁ agonists. The results of these studies, which have been reviewed extensively elsewhere²⁸ and are summarized in Table 2, demonstrated mixed efficacy in acute seizure models in various species. In some models, Δ⁹-THC reduced seizure frequency or severity whereas in other studies there was no effect or even potentiation of convulsive effects. Similarly, synthetic CB₁ agonists have shown variable effects in seizure models. Finally, in some naïve, seizure susceptible rats and rabbits, Δ⁹-THC actually provoked epileptiform activity^{29; 30}. Finally, some studies found dose-limiting toxicity and tolerance to the anti-seizure effects with Δ⁹-THC administration. These findings suggest that Δ⁹-THC is not the sole cannabinoid responsible for the anti-seizure effects of cannabis, and thus activation of CB₁ receptors with Δ⁹-THC or synthetic agonists is unlikely to yield therapeutic benefit for patients with epilepsy.

Cannabidiol and related compounds

CBD is the only non-Δ⁹-THC phytocannabinoid to have been assessed in preclinical and clinical studies for anticonvulsant effects. In mice, CBD blocked MES-induced seizures in one study³¹ but had no effect on PTZ- or MES-induced seizures in another³². However, given the routes of administration used, the lack of efficacy in the latter study may reflect inadequate CBD levels, since several other reports (see below) have found CBD to be effective against both PTZ- and MES-induced seizures.

The anticonvulsant effects of CBD, Δ⁹-THC, and other cannabinoids were also compared using a variety of standard seizure models by Karler and Turkanis³³. Significant anticonvulsant effects against the MES test in mice were found for the following cannabinoids (approximate ED₅₀ values in parentheses): CBD (120 mg/kg), Δ⁹-THC (100 mg/kg), 11-OH-Δ⁹-THC (14 mg/kg), 8β- but not 8α-OH-Δ⁹-THC (100 mg/kg), Δ⁹-THC acid (200–400 mg/kg), Δ⁸-THC (80 mg/kg), cannabinol (CBN) (230 mg/kg), and Δ⁹-nor-9α- or Δ⁹-nor-9β-OH-hexahydro CBN (each 100 mg/kg). More recently, CBD has been shown to have anti-epileptiform and anticonvulsant effects in in vitro and in vivo models. In two different models of spontaneous epileptiform local field potentials (LFPs) in vitro, CBD decreased epileptiform LFP burst amplitude and duration. CBD also exerted anticonvulsant effects against PTZ-induced acute generalized seizures, pilocarpine-induced temporal lobe convulsions, and penicillin-induced partial seizures in Wistar-Kyoto rats^{34; 35}.

Despite CBD’s convincingly anticonvulsant profile in acute models of seizure, there is less preclinical evidence for CBD’s effects in animal models of chronic epilepsy. CBD exerted no effect on focal seizure with a secondary generalization produced by cobalt implantation³⁶, although Δ⁹-THC had a time-limited (~1 day) anticonvulsant effect. Model-specific effects were evident for CBD, which was effective in the MES and all of the GABA-inhibition-based models, but was ineffective against strychnine-induced convulsions³⁷. CBD has also been shown to increase the afterdischarge threshold and reduce afterdischarge amplitude, duration, and propagation in electrically kindled, limbic seizures in rats³⁸.

As mentioned previously, CBDV, the propyl variant of CBD, also has significant anticonvulsant properties. Using the same in vitro models of epileptiform activity described above³⁴, CBDV attenuated epileptiform LFPs and was anticonvulsant in the MES model in ICR mice and the PTZ model in adult Wistar-Kyoto rats. In the PTZ model, CBDV administered with sodium valproate or ethosuximide was well tolerated and retained its own additive anticonvulsant actions. It also retained efficacy when delivered orally. In contrast, while CBDV exerted less dramatic anticonvulsant effects against pilocarpine-induced seizures, it acted synergistically with phenobarbital to reduce seizure activity. CBDV exerts its effects via a CB₁-receptor-independent mechanism³⁹.

The mechanisms by which CBD and CVDV exert their anti-seizure effects are not fully known though several of the potential targets of cannabidiols described above may be involved. Via modulation of intracellular calcium through interactions with targets such as TRP channels⁴⁰, GPR55 or VDAC1⁴¹, CBD and related compounds may reduce neuronal excitability and neuronal transmission. Alternatively, cannabidiol’s anti-inflammatory effects, such as modulation of TNFα release⁴², or inhibition of adenosine reuptake⁴³ may also be involved in anti-ictogenesis. Careful pharmacological studies are needed to further delineate mechanisms.

Other phytocannabinoids

Of the plant cannabinoids that have been identified, few have been investigated beyond early screening for affinity or activity at CB receptors. Δ⁹-THCV, a propyl analog of Δ⁹-THC, is a neutral antagonist at CB₁ receptors⁴⁴. Δ⁹-THCV exerts some anti-epileptiform effects in vitro and very limited anticonvulsant effects in the PTZ model of generalized seizures⁴⁵. Synthetic CB₁-receptor antagonists/inverse agonists have also been investigated in some models of acute seizure and, while partial or full CB₁ agonism produces largely anticonvulsant effects, neutral antagonism has very limited effects upon seizure, and inverse agonism has either no effect or a limited proconvulsant effect (see Table 2). Finally, CBN exerted no effect upon chemically or electrically induced seizures in mice (³²; see above).

Distribution

The distribution of CBD is governed by its high lipophilicity (K_{octanol-water} ~6–7), and a high volume of distribution (~32 L/kg) has been estimated, with rapid distribution in the brain, adipose tissue, and other organs⁴⁶. CBD is also highly protein bound, and ~10% is bound to circulating red blood cells⁴⁷. Preferential distribution to fat raises the possibility of accumulation of depot in chronic administration, especially in patients with high adiposity.

Metabolism and elimination

Like most cannabinoids, CBD is metabolized extensively by the liver, where it is hydroxylated to 7-OH-CBD by P450 enzymes, predominantly by the CYP3A (2/4) and CYP2C (8/9/19) families of isozymes. This metabolite then undergoes significant further metabolism in the liver, and the resulting metabolites are excreted in the feces and, to a much lesser extent, in the urine. The terminal half-life of CBD in humans is estimated at 18–32 hours and, following single dose administration in chronic cannabis users, the clearance was 960–1560 ml/min⁴⁷.

Safety in humans

Multiple small studies of CBD safety in humans in both placebo-controlled and open trials have demonstrated that it is well tolerated across a wide dosage range. No significant central nervous system side effects, or effects on vital signs or mood, have been seen at doses of up to 1500 mg/day (p.o.) or 30 mg (i.v.) in both acute and chronic administration⁵⁰. Little safety data exists for long-term use in humans, though there have been many patient-years of exposure to nabiximols following approval in many European countries and Canada. There is some theoretical risk of immunosuppression, as CBD has been shown to suppress Interleukin 8 and 10 production and induce lymphocyte apoptosis in vitro^{51; 52}.

It should be noted that the above studies were performed in adults. The pharmacokinetics and toxicity of CBD in children is not well understood.

Drug-drug interactions

Little data exists regarding drug interactions with CBD in humans, though there are some theoretical concerns that could have implications for its use in people with epilepsy (PWE). CBD is a potent inhibitor of P450 isozymes, primarily CYP2C and CYP3A classes of isozymes, in vitro and in animal models⁵³. This is particularly important because many medications are substrates for CYP3A4. However, inhibition has typically not been observed at concentrations used in human studies⁵³.

Repeated administration of CBD may induce CYP2B isozymes (CYP2B1/6) in animal models, which may have implications for PWE as antiepileptic drugs (AEDs) such as valproate and clobazam are metabolized via these isozymes. Finally, because CBD is metabolized in a large part by CYP3A4, it is likely that common enzyme-inducing AEDs such as carbamazepine and phenytoin could reduce serum CBD levels.

Planned trials for CBD in Dravet and Lennox-Gastaut syndromes

Among children with treatment-resistant epilepsy, those suffering from early-onset and severe epilepsies such as Dravet syndrome (DS) and Lennox-Gastaut syndrome (LGS) suffer the greatest neurodevelopmental problems, including intellectual disability and autism. In DS, which most often results from mutations in the SCN1A gene, healthy, developmentally normal children present in the first year of life, usually around six months, with convulsive status epilepticus (SE) frequently triggered by fever. Further episodes of SE, hemiclonic or generalized, tend to recur and, after the first year of life, other seizure types develop, including focal dyscognitive seizures, absences, and myoclonic seizures⁵⁵. Seizures in DS are usually refractory to standard AEDs and, from the second year of life, affected children develop an epileptic encephalopathy resulting in cognitive, behavioral, and motor impairment. Outcome is generally poor, with intellectual disabilities and ongoing seizures in most patients.

Thus early and effective therapy for DS is crucial. More effective early control of epilepsy is associated with better developmental outcomes in children today than those who were treated 20 to 30 years ago. Currently, doctors know to avoid drugs that can worsen seizures (e.g., carbamazepine and lamotrigine) and to prescribe effective drugs (e.g., valproic acid, clobazam, topiramate, stiripentol) or dietary therapies (ketogenic or modified Atkins diet) earlier in the disease course. Stiripentol is the only compound for which a controlled trial has been performed in DS⁵⁶, and it has showed a high rate of responders (71% responders on STP versus 5% on placebo). Stiripentol was awarded Orphan Drug Designation for the treatment of DS by the European Medicine Agency in 2001 and by the FDA in 2008.

LGS is a rare but devastating childhood epilepsy syndrome that can result from diverse etiologies, including structural, metabolic, and many genetic disorders; in many cases the cause is unknown. LGS presents in children ages one to eight years; in most cases, onset is between the ages of three and five years. Most LGS patients experience multiple refractory seizures every day despite multiple AEDs and non-pharmacologic treatment including ketogenic diet, vagus nerve stimulation, and epilepsy surgery. The prognosis remains poor with current therapies. Morbidity is significant: Head injuries are common, so that patients often must wear helmets; some patients have even become wheelchair-bound as a result of violent drop attacks.

Effective treatments for both DS and LGS are needed. A recent U.S. survey of 19 parents, 12 of whom had children with DS, explored the use of CBD-enriched cannabis therapy⁵⁴. Of the 12 DS parental respondents, 5 (42%) reported a >80% reduction in seizure frequency. A single LGS parent responded and reported a >80% reduction in seizure frequency. Overall, parents reported improved alertness and lack of side effects apart from fatigue and drowsiness in some children. This may have been related to clinically significant levels of THC in some cannabis preparations used.

Patients with DS and LGS are potentially good candidates for a CBD trial given the need for more effective and better-tolerated therapies for these epilepsies, the high rate of seizure frequency, and the relative homogeneity of the specific syndromes. Several of the authors are currently initiating a study to determine the tolerability and optimal dose of CBD in children with DS and LGS. Inclusion criteria include a definite epilepsy syndrome diagnosis, ongoing seizures despite having tried two or more appropriate AEDs at therapeutic doses, and at least two seizures per week. To help improve the accuracy of seizure frequency reporting, seizures will be recorded with video-EEG to ensure that the seizure types documented by parents are confirmed by epileptologists. This is particularly important since these syndromes may include some seizure types that are difficult to identify (e.g., atypical absence) or quantify (e.g., eyelid myoclonias); these will not be used as countable seizure types in the planned studies. We will focus attention on the most disabling seizure types: tonic, atonic, and tonic-clonic seizures. Based on the information obtained from these dose tolerability studies, we will then plan subsequent randomized, placebo-controlled, double-blind studies in DS and LGS. The ultimate goal is to determine whether CBD is effective in treating these epilepsies, with the hope of improving seizure control and quality of life. While initial studies have been planned to focus on these severe childhood-onset epilepsies, there is no reason to believe based on available evidence that CBD would not be effective in other forms of treatment-resistant epilepsy.

Neonatal hypoxic-ischemic encephalopathy

Perinatal asphyxia resulting in newborn hypoxic-ischemic encephalopathy (NHIE) occurs in 2–9/1000 live births at term⁵⁷. Therapeutic hypothermia is the only available therapy for asphyxiated infants⁵⁷ but only provides neuroprotection in infants with mild NHIE. Cannabinoids are promising neuroprotective compounds; they close Ca²⁺ channels and prevent toxic intracellular Ca²⁺ buildup and reduce glutamate release⁵⁸. In addition, cannabinoids are antioxidants and anti-inflammatory, modulate toxic NO production, are vasodilators, and show neuroproliferative and remyelinating effects. Acute hypoxic or traumatic brain injury is associated with increased brain endocannabinoid levels⁵⁸.

In newborn rats, the CB receptor agonist WIN 55,212-2 reduces hypoxic-ischemic (HI) brain damage in vitro and in vivo by modulating excitotoxicity, nitric oxide toxicity, and inflammation, and enhances post-insult proliferation of neurons and oligodendrocytes⁵⁸. However, long-lasting deleterious effects of overactivating CB₁ receptors in the developing brain are a potential disadvantage of WIN 55,212-2. By contrast, CBD is an attractive alternative since it lacks CB₁-receptor activity⁵⁹. In the immature brain, CB₂ receptors are involved in CBD actions^{59; 60}. In forebrain slices from newborn mice deprived of oxygen and glucose, CBD reduced glutamate release, inducible nitric oxide synthase (iNOS) and COX-2 expression, cytokine production, and cell death⁵⁹. In newborn pigs, CBD reduced HI-induced injury to neurons and astrocytes; reduced cerebral hemodynamic impairment, brain edema, and seizures; and improved brain metabolic activity^{60; 61}. CBD restored motor and behavioral performance in the 72 hours after HI⁶¹. 5HT_1A and CB₂ receptors are involved in CBD neuroprotection at least in the first hours after HI⁶⁰.

In newborn rats, post-HI neuroprotection by CBD is sustained long term, so that CBD-treated asphyxiated newborn rats behave similarly to controls in motor and cognitive tests one month after HI⁶². CBD is also associated with cardiac, hemodynamic, and ventilatory benefits^60–62. Moreover, CBD is still neuroprotective when administered 12 hours after the HI insult in newborn mice and shows synergistic neuroprotective effects with hypothermia in newborn pigs. All these data make CBD a promising candidate for studies of the treatment of NHIE.

Cannabinoids for Psychiatric Symptoms

While epidemiological evidence identifies cannabis smoking as a risk factor for schizophrenia, several cannabinoid components of the plant are emerging as potential treatments for psychiatric symptoms.