Double-Blind, Randomized, Placebo-Controlled, Crossover Study of Oral Cannabidiol and Tetrahydrocannabinol for Essential Tremor

Standard Protocol Approvals, Registrations, and Patient Consents

The overall goal of this pilot trial (registered at clinicaltrials.gov: NCT03805750) was to determine whether TILRAY TN-CT120 LM is a viable therapy for ET. The study was approved by the Institutional Review Board of the University of California San Diego (UCSD) (Project # 180414) and all study procedures took place at UCSD. Participants were enrolled from January 22, 2019 to February 3, 2020. All participants provided written informed consent during a study screening visit. Since cannabis is a classified as a Schedule 1 agent by the United States Drug Enforcement Agency (DEA), we applied for approval to conduct the trial, along with import/export permits from the DEA and Health Canada, respectively, to receive the investigational medical product from Tilray Canada. Additional review and approval were received from the Research Advisory Panel of California.

Data Availability

The data that support the findings of this study are available by written request from the corresponding author (KL).

Screening

Participants with a diagnosis of ET were recruited from the UCSD Movement Disorders Center and the general community. After providing informed consent and being screened for eligibility criteria (see below), participants underwent a complete neurological examination to confirm their diagnosis of ET and exclude other neurological disorders. Cognitive function was measured by the Montreal Cognitive Assessment (MoCA; range 0–30, lower is worse) [6]. Participants who were taking antitremor medications were required to have been on a stable dose of their tremor medications for a minimum of 6 weeks prior to the screening visit and were encouraged to remain on these doses for the remainder of the study period. Participants were asked about their other tremor medication doses at each follow-up visit to confirm stable doses.

Eligibility Criteria

Inclusion criteria were: 1) ET diagnosed by a movement disorders neurologist (confirmed at screening); 2) age 21 years and older; 3) taking a stable dose of anti-tremor medication (e.g., propranolol, topiramate) for at least 6 weeks prior to screening; and 4) a minimum score of 2 on the Essential Tremor Rating Assessment Scale (TETRAS) [7] upper limb tremor assessment (tremor amplitude of at least 1–3 cm). Exclusion criteria were: 1) clinically relevant non-ET related abnormal findings on neurological exam; 2) Presence of rest tremor; 3) diagnosis of mild cognitive impairment or dementia; 4) currently pregnant or nursing; 5) women of childbearing potential unable or unwilling to use an effective form of contraception during course of the trial; 6) use of medications known to interact with cannabinoid-type agents; 7) history of alcohol use disorder (based on DSM-5 criteria [8]) or substance abuse; 8) exposure in the past 21 days to primidone or within the past seven days to benzodiazepines, ketoconazole, ritonavir, clarithromycin, rifampin, carbamazepine, St. John’s Wort, digoxin, or other agents known to be strong inducers or inhibitors of CYP3 A, CYP2C19, and CYP2C9; 9) unwillingness to abstain from consumption of grapefruit-containing products that are known to strongly inhibit CYP3 A4; 10) taking concomitant medications that are highly protein-bound with a narrow therapeutic index (e.g., warfarin, cyclosporine, and amphotericin B); 11) unwillingness to take a cannabis-derived agent; 12) allergy or sensitivity to sorbitol, xylitol, stevia, other natural sweeteners, or cannabis; 13) active use of cannabis or a cannabis-derived product (within 4 weeks of screening) or unwillingness to abstain from recreational use during the study; 14) history of clinically relevant Axis I psychiatric disorder(s) (e.g., mania, bipolar depressive disorder, schizophrenia, schizoaffective disorder, or other major psychiatric disorder) [8]; 15) active or prior history of suicidal ideation and/or behavior; 16) clinically relevant coagulopathy, immunologic, gastrointestinal, respiratory, cardiovascular (e.g., uncontrolled hypertension, myocardial infarction in the last 18 months, bundle branch block, congestive heart disease), or endocrine (e.g., uncontrolled diabetes mellitus, hyperthyroidism) disorder; 17) current or chronic infection; or 18) renal insufficiency (estimated glomerular filtration rate < 60 mL/min/1.73 m²).

Randomization

After completing the screening visit, participants were assigned in a ratio of 1:1 to two possible sequences (order of the treatment: placebo ➔ THC/CBD and THC/CBD ➔ placebo) based on a pre-set randomization list that was developed by the research pharmacy. THC/CBD and placebo were both provided by the manufacturer in matching capsules, preventing participants from discriminating which agent they were taking. The participants and all study team members were blinded to which agent was administered. Participants were asked which treatment they thought they were receiving at each assessment visit and during the telephone follow-up visits to assess potential unblinding.

Study Design

This was a phase 1b/II, double-blind, placebo-controlled, crossover trial of THC and CBD for ET. Each participant had a baseline assessment, and eight follow-up assessments as schematized in Figure 1. Each study period included one-week titration, two-week treatment, and one-week tapering. There was a three-week washout period between the two study arms.

Figure 1

Dose Titration, and Treatment

The study drug dose titration is shown in Figure 2. The same titration schedule was used for TN-CT120 LM and placebo. Participants remained on the highest tolerated dose throughout the treatment arm. The starting dose, titration schedule, and target dose were selected based on safety, tolerability, and efficacy data reported in previous studies using cannabinoids to treat pain [91011]. Adverse effects were inventoried at each study visit. If the participant or investigator felt the higher study drug dose was causing troublesome adverse effects, the dosage could be lowered. During the study design phase, regulators recommended to study cannabis in a ‘real world’ environment, knowing that dietary fat enhances systemic exposure of oral THC and CBD [1213], and food intake increases the bioavailability of CBD [14]. Participants were advised to schedule their dose at a similar time every morning after breakfast to maintain similar systemic exposure throughout the study. Regular reminders were provided to participants to ensure they were taking the study drug post-meals. During the mornings of the assessment days, participants were provided with a standardized FDA high-fat, high-calorie meal (1000 calories, 50% fat) prior to taking the study drug to evaluate real-world food effects.

Figure 2

Dose Taper and Washout

After completing a two-week course on the highest tolerated dose of THC/CBD (up to 15/300 mg/day) or placebo, participants remained on this dose and returned for assessment and completion of research measures at visits 4 and 8. Participants were then tapered gradually over one week, similarly to the initial upward dose titration. Participants completed a three-week washout period prior to being crossed over to the alternate arm to allow return to baseline. Washout duration was selected taking the elimination half-life of THC/CBD into consideration [15].

Assessments

For this digital outcome measure, each participant was asked to trace an Archimedes spiral displayed on a tablet PC screen (IBM Thinkpad X61) twice with each unsupported hand, proceeding from inside the figure to outside. This technique provides an objective measure of tremor and may enable detection of efficacy with fewer participants [16]. Participants with severe tremor who have difficulty even attempting the task are not appropriate for this test, though no such participants were excluded in this study. The tremor amplitude measured by the digital spiral drawing was chosen as the primary outcome. We have previously validated this method using an algorithm to derive the maximum and mean tremor amplitude from the spiral drawings in this cohort with ET [17]. Our prior work demonstrated that this digital measure has a high correlation with manual measures of tremor amplitude and human ratings of the TETRAS score for spiral drawing severity, with excellent test-retest reliability. At each visit, digitized spiral drawings and accelerometry were assessed at baseline (time 0) and at six time points after taking the study drug (15, 25, 60, 100, 200, and 230 minutes).

Tremor severity was rated using the TETRAS [7]. To avoid examiner placebo effects, participant evaluations on this rating were videotaped for subsequent assessments by two blinded raters at the completion of the study. This was a secondary outcome measure. The Global Impression of Change (CGI) scale [18] was used to rate the degree of change from baseline. All adverse events were assessed and recorded using the Patient Reported Outcomes of the Common Terminology Criteria for Adverse Events (PRO-CTCAE) version 1.0 [19]. Suicidality was assessed by trained study personnel using the Columbia-Suicide Severity Rating Scale (C-SSRS) [20]. Blinding assessment was performed at visits 4 and 8; participants and clinicians were asked their opinion of whether participants were in the treatment or placebo arm. Detailed descriptions of the secondary assessments can be found in the Supplementary Appendix.

Pharmacokinetic Analyses of Cannabinoids in Plasma

Plasma concentrations of cannabinoids (CBD, THC, and their metabolites, including 7-OH-CBD, 11-OH-THC, 7-COOH-CBD, CBG, CBN, THC-COOH, and THC-V), were measured using published methods [21]. Plasma concentrations were obtained pre-dose and 50–240 minutes post-dose at weeks 1 and 2 (visits 2, 3, 4/6, 7, 8) in both the active drug and placebo arms. Plasma concentration-time data of CBD, THC, 7-OH-CBD, 11-OH-THC, 7-COOH-CBD, CBG, CBN, THC-COOH, and THC-V were analyzed by standard non-compartmental methods (Phoenix WinNonlin version 8.4). Maximum and minimum concentrations (C_max, C_min) along with corresponding time points (T_max, T_min) were observed directly. Area under the concentration time curve (AUC) was estimated by the trapezoidal rule. The correlations between pharmacokinetic (PK) exposures with demographic factors (age, body size) were examined, along with planned analysis of any relationship between exposure and the resulting efficacy and/or safety response.

Outcome Measures

The primary study outcome was digitized spiral (SPR) amplitude, including maximum, mean, and tremor power features. Maximum amplitude represents the largest deviation from an ideal spiral in millimeters, while mean amplitude represents the average of all tremor amplitudes within a single spiral. Tremor power represents a unitless measure of the intensity of tremor within a peak tremor frequency for a particular spiral. The spiral algorithm calculates rhythmicity by measuring a narrow frequency band at a peak frequency. Secondary outcome measures included the TETRAS [7], CGI [18], PRO-CTCAE [19], plasma concentrations of THC/CBD and their metabolites, and the C-SSRS [20]. Vital signs and ECGs were assessed since the safety of cannabinoids in older adults are not well characterized and cardiovascular effects, including increased heart rate and decreased blood pressure, have been reported in healthy individuals. Serious adverse events such cardiac arrhythmias and acute myocardial infarctions have also been reported [22].

Sample Size Estimation

A sample size of 14 participants was selected for this exploratory proof-of-concept study as a practical and feasible choice to gather preliminary data and insights, given the absence of prior information in this population to guide a formal power calculation. However, due to multiple constraints (e.g., expiration of IMP, resource/funding limitations) that were exacerbated by the COVID-19 pandemic, we were unable to reach the target sample size.

Data Analysis

Baseline Characteristics

A total of 13 participants were screened and seven were enrolled; six were ineligible (reasons for exclusion included diagnoses of Parkinson disease (n = 2), primary writing tremor (n = 1), and functional tremor (n = 1)) (Figure 3). Table 1 shows the baseline demographic and clinical characteristics of participants enrolled. Study participants were all White men (100%) in their late 60s (67.0 ± 11.3 year), who had positive family history of ET (100%) and mean ET duration of around 40 years. Only one of the seven participants (14.3%) was taking other medications for tremor at the time of enrollment. This participant, along with four others, had previously tried and failed medications for ET. Participants were asked about prior substance use during screening; four participants reported prior cannabis use years earlier.

Figure 3

Treatment effects

Seven participants were randomized (N = 7; four in the placebo to THC/CBD sequence and three in the THC/CBD to placebo sequence). Two participants (both in the placebo to THC/CBD sequence) withdrew from the study during the second period while taking the study drug; one had a serious adverse event, and the other was unable to tolerate the lowest dose of one capsule. The mean (SD) study baseline SPR measures (maximum, mean, and tremor power) were 7.16 mm (4.40), 2.18 mm (1.17), and 0.051 (mm/s)²/Hz, respectively.

Primary Analysis

We found no evidence to support greater tremor reduction in the THC/CBD arm compared to placebo (maximum amplitude: mean difference Δ = –1.02, effect size d = –0.43, p = 0.13; mean amplitude: Δ = 0.083, d = 0.025, p = 0.37; amplitude power: Δ = 0.0004, d = –0.026), p = 0.92) (Table 2). Also, average change from baseline to time = 60, 100, 200, and 230 minutes post-dose was not significantly different between the arms (maximum amplitude: χ² = 2.81, df = 1, adjusted p = 0.15; mean amplitude: χ² = 0.77, df = 1, adjusted p = 0.38; amplitude power: χ² = 2.71, df = 1, adjusted p = 0.15) (Table 3). In addition, changes in the primary outcomes over time (from baseline to 230 minutes after dose) did not significantly differ between placebo and THC/CBD treatments (Table 3). Figure 4 depicts mean amplitude measures with 95% confidence interval (CI) for each time point.

Figure 4

Similar results were obtained for analyses stratified by right and left hands, dominant and non-dominant hands, and worse and not-worse hands (data not shown). Results showed that all treatment arm by time interaction effects were not significant, which indicates that the mean response profiles from the time zero (pre-dose) to 230 minutes after dose did not differ between placebo and THC/CBD treatments.

Secondary Endpoints

Blinding Assessment

Both the clinicians and the participants were 91.7% accurate in correctly determining the treatment arm.

Adverse Events

Seven enrolled participants reported a total of 101 adverse events over the course of the study. Among these, 28 (29.8%) were reported during the placebo condition, 66 (70.2%) were reported during the THC/CBD condition, and seven were reported during the taper period (Supplementary Table 1). There were no Adverse Effects of Special Interest.

One serious adverse event (SAE) was recorded. On the day of the event, the participant boarded a flight at 6am and took the prescribed three capsules of IMP containing THC/CBD at 6:30am but had not eaten as required by the protocol. At 7am, the participant ate breakfast and then began to “feel high” and lightheaded, while his spouse noted he appeared pale. At 8:30am, the participant went to the bathroom on the plane and within minutes of returning to his seat (~8:45am) his spouse described a spell consisting of his head and eyes rolling back, then tonic-clonic posturing that lasted approximately 1–2 minutes, followed by emesis. The participant had no prior history of seizure. Medical personnel on the flight assessed vital signs that were reportedly normal though no documentation was available. The participant remained pale and felt exhausted for a period of 20 minutes before returning to baseline. Upon flight landing, he was taken to a local emergency department for additional evaluation. Upon arrival, blood pressure readings could not be obtained due to hypotension. Laboratory and imaging studies were normal with bradycardia (heart rate 43 beats per minute) noted on ECG. The participant received intravenous fluids and was discharged hours later with a diagnosis of syncope. He had no further sequelae, though he discontinued further use of the study drug and was withdrawn from the study. The event was also reviewed by the study data safety monitoring board, which concurred with the suspected diagnosis of convulsive syncope that had a reasonable possibility of being study related.

Pharmacokinetic Results

The pharmacokinetic results from visits 2, 3, 4, 6, 7, and 8 are shown in Supplementary Table 2 for six participants; data was missing for participant P02, whose samples were collected but were not found during processing. The median Tmax for THC and CBD was 4.0 hours (range = 1.5–4.0 hours) and 3.0 hours (range: 1.5–4.0 hours), respectively. The mean adjusted Cmax for CBD was 0.31 ng/mL/mg (SD = 0.24) and for THC was 0.32 ng/mL/mg (SD = 0.26). Concentrations of CBD, THC, and their metabolites varied among and within individuals, with only one participant, P07, showing a consistent concentration of the predominant metabolite, 7-COOH-CBD. Several metabolites, including CBN, CBG, and THC-V, were below the limit of quantitation in plasma. While we observed a consistent washout in THC, low concentrations of CBD and its metabolites (7-OH-CBD and 7-COOH-CBD) were apparent for some participants in the drug –> placebo arm, with 7-OH-CBD detectable in some participants through visit 8, although there was a ~98% reduction in levels compared to the drug arm.