Cannabidiol Lacks Direct Effect on Cortical Excitability: A Randomized, Double Blind, Placebo Controlled, 3‐Way Crossover Trial
Centre for Human Drug Research Leiden The Netherlands
Leiden University Medical Centre Leiden The Netherlands
* Correspondence: Geert Jan Groeneveld (ggroeneveld@chdr.nl)Abstract
Cannabidiol (CBD) is approved as an adjunctive treatment of seizures associated with Dravet syndrome, Lennox–Gastaut Syndrome, and tuberous sclerosis. Its therapeutic and adverse effects are thought to arise, at least partly, from a pharmacokinetic interaction with clobazam, another anti‐seizure medication (ASM). The goal of this study was to evaluate the intrinsic anti‐epileptic and sedative properties of CBD. A randomized, double‐blind, placebo‐controlled, 3‐way crossover trial was conducted in 25 healthy males. On each visit, single doses of 30 mg CBD, 700 mg CBD, or placebo were administered orally. The effects of CBD on cortical excitability were measured using transcranial magnetic stimulation (TMS) combined with electromyography (EMG) and electroencephalography (EEG). Sedative properties were assessed using a validated CNS test battery. Pharmacokinetic sampling was performed. Data were analyzed using a mixed‐effects model. CBD did not have significant effects on single pulse and paired pulse TMS‐EMG parameters, compared to placebo. Some significant clusters were seen on paired pulse TMS‐EEG at 3 hours post‐dose for 30 mg CBD, and at 3 and 5 hours post‐dose for 700 mg CBD. CBD did not have significant effects on any tests assessing its sedative properties. These results suggest that CBD may lack intrinsic anti‐epileptic and sedative properties and that its effects could be primarily a product of interactions with other drugs, notably clobazam.
Boxed Text
- WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?
CBD reduces seizure frequency in Dravet syndrome, Lennox–Gastaut syndrome, and tuberous sclerosis. However, its intrinsic anti‐epileptic action remains debated, as a pharmacokinetic interaction with clobazam is suspected to (at least in part) explain its efficacy and adverse effects.
- WHAT QUESTION DID THIS STUDY ADDRESS?
This study investigated whether CBD modulated human cortical excitability or produced measurable CNS effects in healthy volunteers at over‐the‐counter (30 mg) or therapeutic (700 mg) doses.
- WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?
CBD did not alter cortical excitability, as measured with TMS‐EMG and single‐pulse TMS‐EEG. Although CBD‐induced effects on paired‐pulse TMS‐EEG were detected, these lack a clear physiological interpretation. Compared with the pronounced signatures typically observed after single doses of established antiseizure medicines, these findings suggest that CBD alone does not meaningfully decrease cortical excitability. Furthermore, CBD did not affect vigilance, eye movements, balance, memory, or subjective state.
- HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?
This study provides further support to the idea that the anti‐epileptic and adverse effects of CBD may be caused in large part by its interaction with clobazam, rather than by intrinsic pharmacological actions of CBD itself.
Untitled section
Cannabidiol (CBD) is a non‐intoxicating constituent of the cannabis plant. 1 CBD is available as an over‐the‐counter (OTC) health supplement and is increasingly used to reduce anxiety and to improve sleep and mood, 2 despite a lack of evidence for such use. 3 CBD is approved by the Federal Drug Administration (FDA) in the US and the European Medicines Agency (EMA) as an adjunctive treatment of seizures associated with Dravet syndrome (DS), Lennox‐Gastaut Syndrome (LGS), and tuberous sclerosis (TSC). 4 , 5 , 6
Although CBD has been demonstrated to reduce seizure frequency in the three rare epileptic syndromes, 1 there is no consensus on the underlying mechanism through which CBD exerts its anti‐epileptic effects. 7 Several hypotheses have been proposed, including CBD‐induced modulation of intracellular Ca2+ (either via antagonism of the G protein‐coupled receptor 55 (GPR55), 8 , 9 or via desensitization of the vanilloid receptor 1 (TRPV1) 8 ), inhibition of adenosine reuptake, 8 modulation of voltage‐gated sodium channels (Nav) 10 , 11 , 12 and modulation of the GABAA chloride current. 13
At least in part, however, the therapeutic effects of CBD are thought to arise from an interaction with clobazam, another antiseizure medication (ASM), rather than from any intrinsic anti‐epileptic action of CBD itself. 7 , 14 In the registration trials in DS and LGS patients, 56% of the patients assigned to CBD treatment used clobazam concomitantly, 7 and substantially greater reductions in seizure frequency were observed in these patients. 15 A pharmacokinetic (PK) interaction between the two drugs seems the most obvious explanation: clinical studies show that CBD, a CYP3A4 and CYP2C19 inhibitor, increases concentrations of N‐desmethylclobazam, the active metabolite of clobazam, in the order of 2.6 to 6‐fold. 16 , 17 , 18 , 19 The adverse effects of CBD appear to be similarly driven by its interaction with clobazam. A meta‐analysis of the data from the DS, LGS, and TSC trials found the odds of adverse events to be almost four times higher in patients receiving CBD concomitantly with clobazam, with primarily the incidence of somnolence and sedation being increased, 1 both of which are typical adverse effects associated with clobazam. 20
The magnitude of intrinsic anti‐epileptic effects of CBD remains unclear due to the widespread concomitant clobazam use in the clinical trials in DS, LGS, and TSC patients. 7 , 21 Illustrative of such uncertainty is the decision by the EMA to only approve CBD for treatment of DS and LGS (although not TSC), when combined with clobazam. 6 Our own research group has even argued that, based on clinical trial simulations, the PK interaction with clobazam sufficiently explains the full antiseizure effect of CBD in LGS, and that CBD might lack intrinsic anti‐epileptic effects altogether. 14 , 22
The main goal of this study was to evaluate the intrinsic anti‐epileptic activity of CBD by measuring its effects on cortical excitability in healthy volunteers. Transcranial magnetic stimulation (TMS) is a technique that allows for non‐invasive stimulation of the motor cortex. 23 Responses to such stimulation can be measured with electromyography (EMG) over contralateral muscles, allowing for assessment of neuronal excitability along the cortico‐spinal tract, or using electroencephalography (EEG), which measures cortical excitability more directly. 24 Cortical excitability is a relevant measure of anti‐seizure drug effects, as it is increased across different epilepsy types 25 and has been shown to normalize in patients who became seizure free after successful treatment with ASMs. 26 Demonstrating drug‐induced changes in cortical excitability in healthy volunteers likely has high translational value to patient populations, as various approved, widely prescribed, efficacious ASMs have been found to reduce cortical excitability in healthy volunteers. 23 , 27 Accordingly, TMS‐EMG and TMS‐EEG are increasingly utilized in early‐phase clinical trials of novel ASMs to identify potential anti‐seizure effects in healthy volunteers. 24 , 28
A secondary aim of this study was to assess the sedative properties of CBD using a validated test battery for assessment of central nervous system (CNS) drug effects at CBD doses representative of both OTC and anti‐seizure use.
METHODS
Participants and study design
This study had a double‐blind, randomized, placebo‐controlled, 3‐way cross‐over design. Each participant had three visits to the clinical research unit, during which the effects of two doses of CBD were compared to placebo. There was a 2‐week washout period between each visit. The study was conducted at the Centre for Human Drug Research (CHDR) in Leiden, the Netherlands, between December 2023 and March 2024. The study was approved by the Medical Ethics Committee of Stichting Beoordeling Ethiek Biomedisch Onderzoek (Assen, the Netherlands) and conducted in compliance with the International Conference on Harmonization Good Clinical Practice guidelines. The study was registered under the European Union Clinical Trials Information System number 2023‐508311‐23‐00.
Each participant provided written informed consent before study participation. Healthy males, aged 18–55 years, were selected following a medical screening according to protocol‐specific inclusion and exclusion criteria (Supplemental Material S1 ). Females were excluded due to potential menstrual cycle‐related confounding effects on cortical excitability. 29 Participants with an increased risk of TMS‐related complications based on the TMS safety questionnaire 30 were excluded.
Study drugs
The active formulation used in this study was Clinican®, a 10% CBD almond oil solution, with an orange flavoring liquid added to disguise the taste. The placebo formulation consisted of almond oil with an orange flavoring liquid. On each visit, participants received single oral doses of either 30 mg CBD (as 0.3 mL 10% CBD oil +7 mL placebo), 700 mg CBD (as 7 mL 10% CBD oil +0.3 mL placebo), or placebo (as 7.3 mL placebo).
The low dose of 30 mg CBD was deemed representative of OTC use. 31 The high dose of 700 mg CBD corresponded to the high end of the approved range for treatment of DS and LGS (twice daily dosing up to 10 mg/kg), 6 assuming a 70 kg individual.
Fasting was required for at least 4 hours prior to every scheduled visit. Shortly after arrival, participants received a semi‐standardized light breakfast as described previously. 32 Participants remained fasted for at least 2 hours before and 1.5 hours after study drug administration, except for a small biscuit (~50 kCal) eaten directly after dosing to remove the taste of the study drug.
Pharmacodynamic assessments
TMS
TMS measurements were done pre‐dose for baseline and at 3 and 5 hours following dosing. TMS was performed in accordance with current guidelines, 33 using a MagPro R30 with MagOption stimulator and an MCF‐B65 butterfly coil (2 × 75 mm) (MagVenture GmbH, Hueckelhoven, Germany). The motor hotspot of the dominant abductor digiti minimi (ADM) muscle was stimulated, as assessed by the Edinburgh Handedness Questionnaire. 34 The TMS coil was positioned tangentially to the skull and laterally at a 45° angle with the midline. After determining the resting motor threshold (rMT) 35 at the start of each measurement, a single pulse protocol was applied immediately followed by a paired pulse protocol. The single pulse protocol consisted of 75 single pulses at an intensity of 120% the rMT. The paired pulse protocol consisted of 75 pairs of pulses in randomized order with inter‐stimulus intervals (ISI) of 2, 15, and 100 ms. Conditioning and test pulses were delivered at an intensity of 120% rMT, except for ISIs of 2 and 15 ms, where 80% of rMT was used for the conditioning pulses. A random interval of approximately 4 seconds (range 3.5–4.5 seconds) was kept between single pulses and pairs of paired pulses. Adapted noise was applied to minimize auditory evoked potential generation. 36 EEG was continuously recorded during TMS using a 40‐channel recording system (Refa‐40, TMSi B.V., the Netherlands). Electrodes were placed according to the international 10–20 system (32‐lead cap, ANT Waveguard), but replacing electrodes placed at the earlobes (i.e., A1 and A2) with electrodes placed at the mastoids (i.e., M1 and M2). The electrode impedance was kept below 5 kΩ and the ground electrode was placed between electrodes Fz and Fpz. EEG signals were recorded from 32 electrodes with a sample frequency of 2048 Hz. ADM muscle activity was continuously recorded using Ag/AgCl electrodes placed in a belly‐tendon montage.
The following TMS‐EMG parameters were extracted: mean single pulse peak‐to‐peak MEP amplitude (μV); paired pulse long intra‐cortical inhibition at ISI 100 ms (LICI100), defined as the percentage ratio of the mean MEP amplitude after the test pulses and the mean MEP amplitude after the conditioning pulses; and paired pulse short intra‐cortical inhibition at ISI 2 ms (SICI2) and intracortical facilitation at ISI 15 ms (ICF15), defined as the percentage ratio of the mean MEP amplitude after the test pulses and the mean amplitude of the unconditioned single pulse MEPs. Details of the TMS‐EEG data synthesis are provided in the Supplemental Material S1 .
NeuroCart® CNS test battery
The CNS test battery was done twice pre‐dose for baseline and repeated 2, 4, and 6 hours following dosing. Saccadic eye movements were employed as a sensitive measure of sedation. 37 Smooth pursuit eye movements and the adaptive tracking test assessed visuomotor coordination and vigilance. 38 , 39 The body sway task measured postural stability. 40 Visual Analogue Scales (VAS) according to Bond and Lader assessed study participants’ subjective state using a series of horizontal bipolar scales ranging from 0 to 100, where values of 0 and 100 represented opposing subjective states and a value of 50 represented the neutral state. 41 Subjective psychedelic effects were evaluated using the VAS according to Bowdle on a scale from 0 to 100 mm. 42 , 43 Working memory was evaluated using a computerized version of the N‐Back test 44 and the Visual Verbal Learning Test (VVLT) evaluated episodic memory and learning behavior. Further test procedure descriptions are provided in the Supplemental Material S1 .
Pharmacokinetic assessments
Venous blood samples were collected pre‐dose and at 2, 3, 4, 5, and 6 hours post‐dose. Plasma CBD concentrations were determined using a validated assay with a lower limit of quantification of 0.75 ng/mL by the iC42 Laboratory (Department of Anesthesiology, University of Colorado, Aurora, Colorado). 45 Further details regarding the bioanalytical method are provided in the Supplemental Material S1 .
PK parameters were calculated using the PKNCA package (version 0.9.5) in R v4.0.3 (R Foundation for Statistical Computing/R Development Core Team, Vienna, Austria, 2019). All PK parameter calculations were based on actual sampling time. For the calculation of the PK parameters, concentrations below the limit of quantification were dropped (treated as missing). AUC was calculated using the log‐linear trapezoidal rule.
Sample size, randomization and blinding
The sample size calculation was based on TMS data obtained previously, and specifically the MEP amplitude. 27 A sample size of 24 had a power of 0.80 to detect a difference in means of −300 μV, assuming a SD of differences of 500 μV, using a paired t‐test with a 0.05 two‐sided significance level. This effect size of 300 μV was deemed a relevant magnitude, as we have shown previously that known effective ASMs (valproic acid, levetiracetam and lorazepam) produce effect sizes in this range. 27 The sample size calculation was performed with SAS v9.4 (SAS Institute Inc., Cary NC).
Study staff and subjects remained blinded until the database lock. The balanced Williams design randomization code was generated using SAS v9.4 by a study‐independent statistician. Blinded study staff assigned the randomization numbers to the participants sequentially after medical screening.
Statistical analysis
All pharmacodynamic parameters except for TMS‐EEG were analyzed with a mixed effects model with treatment, period, time, and treatment by time as fixed factors, and participant, participant by treatment, and participant by time as random factors, and the average baseline value as covariate. For the VAS Bowdle parameters (VAS “Internal Perception,” VAS “External Perception,” VAS” Feeling High”), a constant value of 2 mm was added to each measurement to allow log‐transformation and satisfy the model’s normality assumption for residuals. The analysis results of VAS “Feeling High” were subsequently back‐transformed, reporting the estimated difference as a percentage change. Similarly, the MEP amplitude, all paired pulse TMS‐EMG parameters, and body sway were also log‐transformed for analysis, with the estimated difference presented as percentage change after back‐transformation. The general treatment effect and specific contrasts were reported with the estimated difference and 95% confidence intervals, the least square mean estimates, and the corresponding P‐values. All calculations were performed using SAS for Windows v9.4 (SAS Institute, Inc., Cary, NC). No adjustments for multiple comparisons were employed in accordance with the exploratory nature of this study. 46
The effects of CBD and placebo on single and paired pulse TEPs were compared using cluster‐based permutation analysis, a nonparametric method suited for multi‐dimensional TMS‐EEG data. 47 Dependent samples t‐tests were used for electrode and time comparisons, clustering t‐values with P < 0.05. Significant clusters were determined using permutation testing (1,500 permutations), and results were reported if less than 5% of the summed t‐values obtained by permutation were larger than the cluster value observed in the original data. P‐values of significant clusters are reported. In addition to analyzing the entire period between 0 and 300 ms, the same analysis was applied to specific time periods of interest (TOIs) around the characteristic TEP components (N15: 0–20 ms; P30: 20–40 ms; N45: 40–55 ms; P55: 55–80 ms; N100: 80–130 ms; P180: 130–230 ms).
RESULTS
Participants and demographics
A total of 41 male participants were screened, of which 25 were enrolled in the study and dosed at least once. A summary of the baseline demographics is provided in Table S1 . Of the 25 dosed participants, one was excluded prior to completion due to an event of superficial thrombophlebitis (considered unrelated to the study drug by the investigator) and was replaced; 24 participants completed the trial per protocol (Figure 1 ).
Pharmacodynamic outcomes
Single doses of 30 or 700 mg CBD had no significant effects, when compared to placebo, on the single pulse TMS‐EMG parameters (peak‐to‐peak MEP amplitude (Figure 2 ) and rMT) and paired pulse TMS‐EMG parameters (LICI100, SICI2 and ICF15) (Table 1 ).
| Parameter | Estimated difference (95% CI) p‐value | LS Means | |||
|---|---|---|---|---|---|
| CBD 30 mg vs. Placebo | CBD 700 mg vs. Placebo | Placebo | CBD 30 mg | CBD 700 mg | |
| Single pulse TMS‐EMG | |||||
| Peak‐to‐peak amplitude (μV) | 1.9% (−17.6%, 26.1%) P = 0.86 | −2.9% (−21.9%, 20.7%) P = 0.78 | 740.27 | 754.56 | 718.54 |
| Resting motor threshold (%MSO) | 0.7 (−0.6, 2.1) P = 0.28 | −1.2 (−2.6, 0.1) P = 0.08 | 58.0 | 58.8 | 56.8 |
| Paired pulse TMS‐EMGa | |||||
| Short intracortical inhibition 2 ms (%) | −0.1% (−20.9%, 26.2%) P = 0.99 | 6.5% (−15.7%, 34.5%) P = 0.59 | 35.25 | 35.21 | 37.52 |
| Intracortical facilitation 15 ms (%) | −7.2% (−27.2%, 18.1%) P = 0.53 | −14.9% (−33.2%,8.3%) P = 0.18 | 131.34 | 121.84 | 111.72 |
| Long intracortical inhibition 100 ms (%) | 9.5% (−25.2%, 60.1%) P = 0.63 | −3.6% (−34.1%, 41.2%) P = 0.85 | 3.40 | 3.73 | 3.28 |
| Saccadic peak velocity (degrees/s) | −0.11 (−8.66, 8.43) P = 0.98 | −1.74 (−10.39, 6.91) P = 0.69 | 489.74 | 489.63 | 488.00 |
| Smooth pursuit eye movement (%) | −0.01 (−1.93, 1.90) P = 0.99 | −1.20 (−3.15, 0.74) P = 0.22 | 51.03 | 51.02 | 49.83 |
| Adaptive tracking (%) | 0.01 (−1.16, 1.18) P = 0.98 | −0.59 (−1.77, 0.59) P = 0.31 | 31.63 | 31.65 | 31.04 |
| Body sway (mm) | −3.5% (−15.0%, 9.6%) P = 0.58 | −3.5% (−15.3%,9.9%) P = 0.58 | 254.87 | 246.04 | 245.94 |
| VAS Bond and Lader | |||||
| VAS “Alertness” (mm) | −0.6 (−1.9, 0.7) P = 0.34 | −0.9 (−2.2, 0.4) P = 0.16 | 49.9 | 49.2 | 48.9 |
| VAS “Calmness” (mm) | 1.5 (−0.5, 3.5) P = 0.13 | 0.8 (−1.3, 2.8) P = 0.45 | 52.1 | 53.6 | 52.9 |
| VAS “Mood” (mm) | −0.1 (−1.3, 1.1) P = 0.89 | −0.9 (−2.1, 0.4) P = 0.18 | 51.9 | 51.8 | 51.0 |
| VAS Bowdle | |||||
| VAS “External perception” (log(mm)) | 0.0083 (−0.0052, 0.0217) P = 0.22 | 0.0045 (−0.0093, 0.0182) P = 0.52 | 0.3336 | 0.3418 | 0.3380 |
| VAS “Internal perception” (log(mm)) | −0.0016 (−0.0120, 0.0087) P = 0.74 | −0.0022 (−0.0129, 0.0084) P = 0.67 | 0.3266 | 0.3250 | 0.3244 |
| VAS “Feeling High” (mm) | −4.7% (−11.2%, 2.3%) P = 0.18 | 1.6% (−5.4%, 9.2%) P = 0.64 | 2.16 | 2.06 | 2.20 |
| N‐Back | |||||
| Reaction time 0‐back (ms) | 8.79 (−11.37, 28.94) P = 0.38 | 10.45 (−9.36, 30.27) P = 0.29 | 383.31 | 392.09 | 393.76 |
| Reaction time 1‐back (ms) | 18.85 (−2.07, 39.76) P = 0.08 | 15.95 (−5.32, 37.22) P = 0.14 | 415.99 | 434.84 | 431.94 |
| Reaction time 2‐back (ms) | 16.24 (−17.24, 49.71) P = 0.33 | −12.55 (−46.43, 21.34) P = 0.46 | 513.96 | 530.20 | 501.42 |
| Ratio correct 0‐back | 0.001 (−0.017, 0.020) P = 0.88 | −0.004 (−0.023, 0.015) P = 0.67 | 0.964 | 0.965 | 0.960 |
| Ratio correct 1‐back | 0.019 (−0.015, 0.054) P = 0.26 | 0.022 (−0.013, 0.057) P = 0.20 | 0.924 | 0.943 | 0.946 |
| Ratio correct 2‐back | 0.014 (−0.030, 0.057) P = 0.53 | 0.019 (−0.026, 0.063) P = 0.40 | 0.895 | 0.908 | 0.913 |
| Visual Verbal Learning Test | |||||
| Immediate recall trial 3 (N correct) | −0.6 (−2.0, 0.9) P = 0.42 | 0.3 (−1.1, 1.8) P = 0.65 | 17.6 | 17.0 | 18.0 |
| Delayed recall (N correct) | −0.0 (−1.4, 1.4) P = 0.99 | −0.6 (−2.1, 0.8) P = 0.37 | 13.2 | 13.2 | 12.6 |
| Delayed recognition (reaction time correct) (ms) | 25.6 (−21.9, 73.2) P = 0.28 | −0.5 (−48.7, 47.7) P = 0.98 | 805.7 | 831.3 | 805.2 |
| Delayed recognition (N correct) | 0.2 (−1.1, 1.6) P = 0.73 | 0.1 (−1.3, 1.5) P = 0.85 | 24.2 | 24.5 | 24.3 |
For single pulse TMS‐EEG, single doses of 30 mg CBD significantly decreased the N15 TEP component (i.e., less negative) ( P = 0.02) compared to placebo in an ipsilateral centroparietal cluster at the 3 h post‐dose timepoint (Figure 3 ).
For paired pulse TMS‐EEG (ISI 100 ms), single doses of 700 mg CBD significantly decreased the N45 (i.e., less negative) ( P = 0.01) and increased the P60 TEP component (i.e., more positive) ( P = 0.03) compared to placebo in a contralateral centroparietal cluster at the 3 hour post‐dose timepoint. Similarly, at the 5 hour post‐dose timepoint, 700 mg CBD significantly increased the P30 (i.e., more positive) ( P = 0.04) and decreased the N45 (i.e., less negative) ( P = 0.04) compared to placebo in a contralateral fronto‐centroparietal cluster at ISI 100 ms (Figure 4 ).
Single doses of 30 or 700 mg CBD had no significant effects when compared to placebo on the CNS test battery parameters (saccadic and smooth pursuit eye movements, adaptive tracking test performance, postural stability, VAS “Alertness,” VAS “Mood,” VAS “Calmness,” VAS “Internal Perception,” VAS “External Perception,” “Feeling High,” and n‐Back and VVLT test performance) (Table 1 ).
Pharmacokinetic outcomes
After administration of 30 mg CBD, the mean ± SD AUClast was 20.3 ± 8.4 hour ng/mL and the mean ± SD C max was 8.8 ± 4.2 ng/mL (Table 2 ). Following the administration of 700 mg CBD, the mean ± SD AUClast was 931 ± 413 hour ng/mL and the mean ± SD C max was 395 ± 203 ng/mL (Table 2 ). The median (min, max) T max for both dose levels was 3 (2, 4) hours. PK parameters increased more than dose‐proportionally (Table 2 ). The concentration‐time profiles of cannabidiol are displayed in Figure 5 .
| Treatment | Parameter | Unit | N | Mean | SD | CV% | Geometric mean | Geometric CV% | Median | Min | Max |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 30 mg CBD | AUClast | h*ng/mL | 25 | 20.3 | 8.4 | 41.3 | 18.7 | 46.3 | 18.6 | 6.2 | 42.5 |
| C max | ng/mL | 25 | 8.8 | 4.2 | 47.4 | 7.9 | 49.9 | 8.43 | 2.9 | 19.6 | |
| t max | h | 25 | 3 | 2 | 4 | ||||||
| 700 mg CBD | AUClast | h*ng/mL | 24 | 931 | 413.0 | 44.4 | 858 | 42.1 | 775 | 436 | 2,213 |
| C max | ng/mL | 24 | 395 | 202.8 | 51.3 | 352 | 51.8 | 330 | 160 | 866 | |
| t max | h | 24 | 3 | 2 | 4 |
DISCUSSION
This study aimed to evaluate the intrinsic anti‐epileptic properties of CBD by measuring its effects on cortical excitability with TMS‐EMG and TMS‐EEG, as well as CBD effects on vigilance, visuomotor coordination, and memory.
Effects of CBD on TMS‐EMG measures
No significant treatment effects of CBD were found on any TMS‐EMG measure of cortical excitability, whereas literature reports consistent effects of various ASMs on TMS‐EMG parameters 23 ; our research group has previously demonstrated such effects following single doses of lorazepam, valproic acid, and levetiracetam in a similar population and experimental setup to the current study. 27 Therefore, the lack of significant changes following CBD administration suggests that CBD lacks an intrinsic anti‐epileptic effect. More specific inferences regarding the mechanism of action can be made, as it is known that voltage‐gated sodium channel blockers (e.g., carbamazepine and lamotrigine) increase the rMT, positive allosteric modulators of the GABAA receptor (i.e., benzodiazepines) decrease the MEP amplitude and ICF, and increase SICI, and the specific GABAB receptor agonist baclofen increases LICI. 23 Levetiracetam, an ASM with an unclarified mechanism of action, decreased the MEP amplitude. 27 The absence of such findings in this study suggests that the putative anti‐epileptic effect of CBD is not mediated by voltage‐gated sodium channels, GABAA receptors, or GABAB receptors, and has a different mechanism from levetiracetam.
Effects of CBD on TMS‐EEG measures
CBD had significant treatment effects on TMS‐EEG at both dose levels. The significant clusters at the 700 mg dose level are found both 3 and 5 hours post‐dose and have the characteristics of a drug‐induced effect. The significant clusters are consistent across the two timepoints, as they are found exclusively after paired‐pulsed stimulation with an ISI of 100 ms, in a largely (although not perfectly) similar topographical distribution, and in (partially) overlapping temporal regions of interest (in both cases including the N45 peak). The magnitude of the effect is greater at 3 hours post‐dose, which is consistent with the observed pharmacokinetic T max of CBD. Conversely, the significant cluster after single pulse stimulation at the 30 mg CBD dose level is more likely to represent a chance finding, as it is not reproduced after administration of 700 mg CBD. It could also be a stimulation artifact, which is more common in the TEP components closer to the test stimulus. In any case, the finding at 30 mg CBD appears irrelevant to the evaluation of anti‐epileptic effects of CBD, since much higher doses are required to reduce seizure frequency in patients.
Interpretation of the significant clusters at the 700 mg CBD dose level is challenging. There is a general lack of understanding of how paired‐pulse TMS‐EEG findings correspond to the underlying physiological processes. It is unclear whether the observed significant changes in the paired‐pulse TEPs signify a reduction or an increase in cortical excitability. One earlier study utilizing TMS‐EEG found that baclofen, a GABAB receptor agonist, enhanced the N45 peak in a contralateral topographic region after paired‐pulse stimulation at an ISI of 100 ms, 48 which is similar to the enhancement of the N45 peak we observed with 700 mg CBD. However, it would be premature to conclude that CBD therefore is a GABAB receptor agonist. The same study reported effects on the N100 and P180 peaks (paired pulse), as well as single‐pulse TMS‐EEG – which are all absent in our study. Furthermore, baclofen affects the LICI and has well‐documented sedative effects, which we would expect to measure using TMS‐EMG and the CNS test battery respectively, if CBD indeed acted as a GABAB receptor agonist.
One speculative explanation is that TMS‐EEG, a method that measures cortical electrical activity directly and exclusively, was the only method sensitive enough to detect the effects of CBD. In contrast, TMS‐EMG measures a peripheral response to central stimulation, and the CNS test battery offers even fewer direct tests of CNS function. However, we deem it unlikely that CBD, a drug that markedly reduces seizure frequency in hard‐to‐treat epileptic syndromes, would have such extremely subtle effects in comparison to other proven efficacious ASMs.
Effects of CBD on the CNS test battery
The CNS test battery results certainly support the common characterization of CBD as a “non‐intoxicating” cannabinoid. 49 CBD did not differ significantly from placebo on any measure of vigilance, sedation, visuo‐motor coordination, postural stability, subjective drug effects, and working and episodic memory. This makes it highly unlikely that somnolence and lethargy, two frequently observed adverse effects of CBD, are attributable to CBD itself, and supports the hypothesis that their incidence is primarily driven by the interaction with clobazam or other ASMs. To our knowledge, this study is the first to investigate CBD effects on postural stability and eye movements; the absence of significant treatment effects is in line with the general non‐impairing character of CBD.
Strengths and limitations
The main strength of our study lies in the use of TMS‐EMG and TMS‐EEG, which are validated tools for the assessment of drug effects on cortical excitability. Whereas the assessment of anti‐epileptic effects of CBD as a monotherapy in epilepsy patients may not be feasible for ethical and practical reasons, the use of TMS‐EMG and TMS‐EEG made such an assessment possible in healthy individuals instead. Further strengths of the study include a cross‐over design, which allowed for within‐subject comparisons, adequate power, and a relevant dose selection, with the high dose in line with doses used for the treatment of epileptic syndromes (on a weight‐adjusted basis) and the low dose representative of OTC CBD use for self‐care. The validated CNS test battery assessed pharmacodynamic effects that had never been assessed for CBD previously.
Our study, however, is not without limitations. Most importantly, changes in cortical excitability in healthy (male) volunteers are a surrogate marker for anti‐epileptic drug effects, and not the actual outcome measure of interest—which is seizure frequency reduction in patients. Although increased cortical excitability due to reduced cortical inhibition appears typical for most types of epilepsy, 23 a paradoxically reduced cortical excitability has been found in a population of LGS patients. 50 Therefore, a translation of our study results to patient populations should happen with cautio; although CBD did not clearly reduce cortical excitability in this study, it does not rule out that CBD could have an intrinsic anti‐epileptic effect. However, the results of this study make this less likely, as most anti‐epileptic drugs typically show clear and characteristic effects on cortical excitability in healthy participants, whereas CBD does not. 23
It is possible that CBD does in fact reduce cortical excitability, but this study erroneously failed to show this. A potential reason could be the administration of single doses of CBD, whereas reductions of cortical excitability may require a chronic treatment regimen to manifest. Another possibility is that CBD reduces cortical excitability in patients, just not in healthy individuals. However, we consider these possibilities less likely since single doses of typical ASMs (levetiracetam, valproic acid, lorazepam) clearly reduced cortical excitability in healthy individuals in a similar study set‐up. 27 Clinical studies with a longer treatment duration could provide clarity on these questions.
Although the findings of this manuscript cast doubt on the intrinsic anti‐epileptic effects of CBD, it is important to mention the evidence to the contrary. Meta‐analyses that pooled patients not using clobazam from registration studies in LGS and DS found significant seizure reduction compared to placebo—although with a considerably smaller effect size compared to clobazam users. 1 , 15 , 21 However, such subgroup meta‐analyses are acknowledged to have important methodological limitations, that is, their post hoc nature, lack of randomization for clobazam use, and the pooling across different seizure syndromes. 21 The more recent registration study in patients with TSC found that CBD significantly reduced seizure frequency while including a lower—although still sizeable—proportion of clobazam users (approx. 25%, vs. 56% in DS/LGS trials). 51 Nevertheless, it remains unclear whether the overall treatment effect for CBD would differ significantly from placebo if clobazam users were excluded from analysis. Uncontrolled retrospective studies found no association between concomitant clobazam use and seizure frequencies, but such study designs cannot be relied upon to provide definitive conclusions. 52 , 53 Ultimately, the conclusive proof that CBD is an anti‐seizure medication in the absence of concomitant clobazam can only come from randomized controlled trials in defined populations that are adequately powered to answer this specific question.
Pharmacokinetics
This study reached clinically meaningful CBD exposures, as measured plasma concentrations at the 700 mg dose level were in line with findings in adult epileptic patients, who responded to treatment with CBD. 54 The CBD plasma concentrations increased more than dose proportionally, and the C max and the AUC at the 700 mg dose level were approximately double what would be expected based on the concentrations measured at 30 mg CBD and assuming a linear dose‐exposure relationship. Whereas previous publications have established the dose‐proportionality of Epidyolex® in the therapeutic range (5–10 mg/kg twice daily), 16 our finding suggests a less than dose proportional exposure at the sub‐therapeutic dose of 30 mg (0.43 mg/kg, assuming a 70 kg individual). Alternatively, the non‐linearity in absorption could be specific to the formulation used in this study. A speculative, although hypothetically plausible, explanation is the possibility of auto‐inhibition of CBD metabolism, as CBD is both an inhibitor and substrate of the CYP2C19 and CYP3A4 enzymes. 55 , 56
CONCLUSION
This study found no evidence that CBD reduces cortical excitability. There were no significant treatment effects on TMS‐EMG and single pulse TMS‐EEG measures of cortical excitability, whereas other ASMs are known to produce typical and mechanism‐specific changes on these measures. CBD does seem to affect cortical excitability as measured with paired pulse TMS‐EEG – although further interpretation is challenging because of the limited current understanding of how paired pulse TMS‐EEG measures relate to the underlying physiology. The absence of an obvious reduction in cortical excitability casts doubts on the extent, or possibly, the very existence of intrinsic anti‐epileptic effects of CBD. Such doubts are further exacerbated by the unclarified mechanism of action, a lack of clinical trials demonstrating efficacy in absence of concomitant ASMs, and the prominent pharmacokinetic interaction with clobazam. The absence of sedative effects on a sensitive battery of CNS tests suggests that typical adverse effects associated with CBD may also be primarily a product of its interactions with other drugs.
In conclusion, although the addition of CBD reduces seizure frequency in DS, LGS, and TCS patients, it remains to be demonstrated that CBD has actual anti‐seizure properties by itself; adequately powered, randomized, controlled trials specifically designed to address this question are needed.
FUNDING
No funding was received for this work.
CONFLICT OF INTEREST
The authors declared no competing interests for this work.