Tolerability and Efficacy of a 10:25 Preparation of Δ9-Tetrahydrocannabinol and Cannabidiol for Treatment of Chronic Back or Neck Pain: A Multiple-Dose Escalation Study

Chronic back pain is one of the leading causes of disease burden, causing moderate interference with daily activities for almost two in five sufferers [1]. It affects between 1 and 16% of the population [1, 2]. Current treatment of chronic back pain includes non-pharmacological approaches (exercise, rehabilitation, acupuncture, mindfulness, etc.) [3], and in patients with inadequate response, pharmacological treatment is used [3]. Paracetamol and nonsteroid anti-inflammatory drugs (NSAIDs) are first-line therapies.

However, given the large variability in the presentation and outcomes of patients with chronic back and neck pain [4], treatment can be challenging, particularly when simple analgesia is ineffective. Given this, in addition to limitations on opioid prescribing resulting from concerns about their long-term effectiveness and the limited evidence supporting their effectiveness, there has been increasing interest in finding alternative treatments for patients with chronic pain [5, 6].

One such option is medicinal cannabis. Evidence for medicinal cannabis in chronic non-cancer pain is limited, but it suggests that cannabis is more likely than placebo to produce between 30% and 50% reductions in pain scores and is more likely than placebo to produce significantly greater reduction in pain intensity rating, although the differences are small [7, 8]. The summary of evidence of THC:cannabidiol (CBD) extract, for which there was only one study, provided a moderate level of evidence of 30% reduction in pain scores and a low level of evidence for change in pain scores [7]. Efficacy results are heterogenous, likely due to the different combinations and formulations included in the analysis. Sublingual or orobuccal preparations are useful as they may provide rapid absorption and avoid a first-pass effect [9, 11].

In this study, Cybis^® 10:25, an oromucosal-administered THC and CBD combination, was assessed. Cybis^® 10:25 contains a 10:25 ratio of THC:CBD in medium-chain triglycerides. The CBD is present in order to counteract the psychoactive effects of THC and may also contribute to the analgesic effect [12]. This was the first study of Cybis^® 10:25 in humans. The overall aim was to demonstrate the safety and tolerability of Cybis^® 10:25 in participants with moderate to severe chronic back or neck pain that was unresponsive to over-the-counter (OTC) non-opioid analgesics.

This was a non-randomized, single-arm, open-label study of Cybis^® 10:25 in participants with chronic back or neck pain in which participants received escalating doses of Cybis^® 10:25. This study was designed with reference to the IMPACCT consensus guidelines [13] and the European Medicines Agency guidelines [14]. This study was reviewed and approved by the Bellberry Human Research Ethics Committee (approval number 2021-03-305), Clinical Trials Registration (Clinicaltrials.gov NCT04976738).

This study was conducted between October 2021 and June 2022 at two community sites in Sydney, Australia. Adult (≥18 years of age and <75 years of age) men or women who had a history of back or neck pain of at least 3 months duration were invited to participate. They had an average back or neck pain score of between five and nine on a 10-point visual analogue scale; had to have failed to achieve self-reported satisfactory pain relief using OTC paracetamol or NSAIDs; and were willing to cease all current pain medications. In addition, they agreed to cease driving a car or operating heavy machinery from the date of the first dose, until 7 days after the last dose of Cybis^® 10:25. All participants were required to provide written informed consent.

Potential participants were excluded if they had a history of cannabis use disorder or other substance abuse disorders, were currently using cannabis or other medicinal cannabis products, or had significant other illness. Participants were not permitted to use any pain medications beyond the study intervention for at least 3 days prior to commencing study intervention and throughout the trial period, with the exception of rescue medication (two tablets, each tablet containing 500 mg paracetamol with 150 mg ibuprofen, every 6 h as clinically necessary, with a maximum of six tablets in 24 h).

Participants took multiple ascending doses of Cybis^® 10:25, containing 10 mg/mL of D9-THC, 25 mg/mL of CBD, formulated in medium-chain triglyceride oil. Administration occurred oromucosally. On days 1–7, participants received 0.5 mL once-daily (od); on days 8–14, they received 0.5 mL twice-daily (bd), on days 15–21, they received 1.0 mL bd; and on days, 22–28 they received 1.5 mL bd. Participants were followed up again on day 35 (7 days after the last dose of Cybis^® 10:25).

The primary objective of this trial was to assess the safety and tolerability of Cybis^® 10:25 at low (0.5 mL od), medium-low (0.5 mL bd), medium (1.0 mL bd), and high (1.5 mL bd) doses in participants with chronic back or neck pain refractory to OTC non-opioid analgesics. Secondary objectives were to investigate the pharmacokinetics of Cybis^® 10:25; to investigate the dose-response relationship between Cybis^® 10:25 and pain response as measured by the change from baseline in Numerical Pain Rating Scale (NPRS) and Brief Pain Inventory Short Form (BPI-SF); to investigate the dose-response relationship between Cybis^® 10:25 and mood as measured by the change from baseline in Depression Anxiety Stress Scale (DASS), change from baseline in Medical Outcomes Study Sleep Scale (MOS-SS), and change from baseline in Self-Assessment of Treatment (SAT-II) scale; and to investigate rescue medication use in participants prescribed Cybis^® 10:25 for chronic back and neck pain.

To assess pharmacokinetics, blood samples were drawn at Day 1 prior to dosing (0 [visit window −15 min]) and at 0.25 h ± 5 min, 0.5 h ± 5 min, 1.0 h ± 5 min, 1.5 h ± 5 min, 2.0 h ± 5 min, 2.5 h ± 5 min, 3.0 h ± 5 min, 3.5 h ± 5 min, 4.0 h ± 5 min, 5.0 h ± 10 min, 6.0 h ± 10 min, 9.0 h ± 30 min, 12.0 h ± 30 min, and on Day 2, 24.0 h ± 1 h post dosing. During the pharmacokinetic sampling, a light lunch was permitted after the 4 h sample, and a light snack was permitted 4 h later. A light dinner was permitted. Alcohol was not permitted throughout the study.

The sample size calculation was based on the secondary objective of investigating the dose-response relationship in terms of change from baseline in NPRS. Assuming a non-linear change in NPRS at each dose, then using a one-sided alpha of 0.05, and assuming a standard deviation of 2.2, and three dose levels with an assumed maximum change in NPRS of 2.1 (the minimum clinically important difference for the NPRS is a two-point change [15]), then 19 participants per dose level were required for 80% power. Allowing for 20% drop out, then 24 participants per dose level were planned to be recruited. The sample size was calculated using a linear contrast test.

Analyses were based on the full analysis set (FAS), that is, all participants who enrolled and received at least one dose of Cybis^® 10:25. An additional FAS was developed (FAS by dose) which included all participants enrolled who received at least one dose of Cybis^® 10:25 but placed participants in dose cohorts only if they received that dose. The latter was used for comparisons by dose. Safety analysis was conducted in the FAS population.

Adverse events were summarized by dose cohort. Pharmacokinetic parameters were estimated using non-compartmental methods. The dose-response relationship between Cybis^® 10:25 and efficacy outcomes (NPRS, BPI, DASS, MOS-SS, SAT-II) was assessed using mixed models repeated measures with baseline as a covariate in the model. Sensitivity analyses assessed the impact of rescue medication, by including in the analysis only those participants without rescue medication use.

There was no adjustment for multiple comparisons, and p < 0.05 was considered statistically significant. All analyses were conducted using Stata MP v17 for Mac (StataCorp, Texas Station, USA). Although this was a non-randomized study, this study has been reported in accordance with CONSORT [16] and the pain-specific supplement to CONSORT [17].

Forty people were screened for this study, of whom 28 were included in the study (Fig. 1). The median age of participants was 63.3 years (range 31.3–74.7 years), and half were female (Table 1). Participants had experienced back or neck pain for a median of 10 years (range 7 months–53 years). Only 4 (14%) had previous experience of cannabis, while 26 (93%) had prior analgesic (opioid, paracetamol, NSAID, and other analgesic) use.

The percentage adherence to investigational medicine product (IMP) was high (median 100%) for all four doses, although there was some variability, particularly at the highest dose (adherence range 28.6–100%). The proportion of participants who were compliant with 80% or more of doses was higher in lower doses (93% for 0.5 mL od to 67% in 1.5 mL bd). Of the original cohort (n = 28), 20 (71%) participants reached the highest dose cohort.

The pharmacokinetics of single doses of 0.5 mL Cybis^® 10:25 showed significant interpatient variability (Table 2); briefly, THC C_max was 1.84 ng/mL CBD C_max was 2.49 ng/mL. T_max occurred at 9.58 h (THC) and 11.78 h (CBD). There were dose-dependent increases in median trough concentrations of Cybis^® 10:25.

There were significant reductions in pain in patients administered all doses of Cybis^® 10:25 compared to baseline, and the linear effect of dose was significant (p < 0.001, Table 3). Clinically significant reductions in NPRS scores (>2 point reduction in NPRS) were observed for 1.0 mL bd and 1.5 mL bd doses (both p < 0.001, Fig. 2). The mean and median NPRS reductions were 2.3 and 2.0 points, respectively, for 1.0 mL bd dose, and 2.7 and 3.0 points, respectively, for 1.5 mL bd dose. This equated with a mean 28.8% (median 28.6%) reduction for the 1.0 mL dose and a mean 34.1% (median 37.5%) reduction for the 1.5 mL dose. This pattern remained when excluding any participants who had been administered rescue medication (see Fig. 2b). Similarly, there were dose-dependent improvements in pain levels as assessed by the BPI. Significant improvements from baseline were observed in Pain Now and Pain Average for all doses (Table 3). The proportion of participants who experienced pain other than minimal headache, sprain, or toothache reduced from 81% at baseline to 53% during the 1.5 mL bd dose. There was also a reduction in pain interference in all domains including general activity, mood, walking ability, normal work, relationships, sleep, and enjoyment of life by dose (see Table 4).

Approximately two-thirds of participants (61%) required rescue medication at some point during the study. Rescue medication was more commonly required by participants during treatment with lower doses of Cybis^® 10:25 (58% required rescue medication at 0.5 mL od dose, 48% at 0.5 mL bd dose, 33% at 1.0 mL bd dose, and 32% at 1.5 mL bd dose). The median number of tablets of rescue medication (500 mg paracetamol + 150 mg ibuprofen) taken during the 0.5 mL od dose was 2 (range 0–37), while at the 0.5 mL bd dose and higher, the median number of rescue medication doses taken was zero.

Cybis^® 10:25 appeared to reduce levels of depressive symptoms from baseline in a dose-dependent manner (p = 0.0182), with significant reductions in DASS depression scores for baseline observed at 1.0 mL bd and 1.5 mL bd doses (Table 3). Numerical reductions in anxiety compared to baseline were observed for 0.5 mL od, 0.5 mL bd, 1.0 mL bd, and 1.5 mL bd doses; however, this did not appear to be dose dependent (p = 0.0505). Reductions in stress from baseline were dose dependent (p = 0.0053), with reductions significant for all doses (p < 0.01).

Participants self-reported numerical increases in sleep disturbance but improvements in sleep adequacy with increasing doses of Cybis 10:25. However, significant changes only occurred compared to baseline for 0.5 mL od, 0.5 mL bd, and 1.5 mL bd doses (Table 3). This contrasts to the results from the BPI, which suggested there may have been improvements in sleep, although these did not appear to be dose dependent (see Table 4).

Although not statistically significant (p = 0.12), there were greater proportions of participants who reported that treatment improved pain “quite a lot” or “very much” at higher dose levels (0.5 mL od 4%, 0.5 mL bd 17%, 1.0 mL bd 32%, and 1.5 mL bd 45%). Seven days after the cessation of treatment, most felt that Cybis^® 10:25 was either somewhat better (35%) or very much better (39%) than their previous treatments.

Overall Cybis^® 10:25 was well tolerated. Most participants (82%) experienced at least one adverse event, typically of mild or moderate severity. More participants experienced adverse events at lower dose levels than at higher doses. Vomiting, dizziness, and headache did not appear to be dose-dependent. Related adverse events are presented by severity in Table 5. One severe adverse event was reported: nephrolithiasis, which occurred in one participant 7 days after the last dose of Cybis^® 10:25. Three adverse events led to dose withdrawal: constipation, anxiety, and syncope. The episode of syncope was reported as the only serious adverse event. It occurred in a participant administered 1.0 mL bd. There were no clinically significant changes in systolic or diastolic blood pressure, QTc interval, or clinically important hematological and clinical chemistry values. There were no deaths and no pregnancies reported.

Chronic pain, particularly chronic back and neck pain, remains a significant challenge for patients and clinicians alike [18]. Given its prevalence and impact on mood and function, finding new treatment options is important [19]. A recent review has highlighted the need for “safer and more effective pharmaceutical options to provide pain relief and improvement in function while minimizing the risks of currently available analgesics” [19]. Medicinal cannabis may provide one such therapeutic option; however, to date, there is limited high-quality clinical trial evidence on the use of medicinal cannabis in chronic pain. Our study was conducted to address this gap. This was the first clinical study of Cybis^® 10:25 in people with chronic back or neck pain. Our study has shown that doses of up to 1.5 mL bd are well tolerated, with clinically significant reductions in pain for 1.0 mL bd and 1.5 mL bd doses. This represents, to our knowledge, the first published dose-response study of a THC:CBD combination.

We found significant heterogeneity in pharmacokinetics following single dose of 0.5 mL Cybis^® 10:25. THC and CBD are highly lipophilic with poor oral bioavailability. There are several possible reasons for interpatient differences in pharmacokinetic parameters, including differences in body fat, rapid distribution into tissues, potential swallowing of the product, or differential absorption in different parts of the oral mucosa [20]. Another potential contributing factor is the age of the included patients with long disease duration who were previously treated with other analgesic agents. Maximal concentrations of CBD (C_max) were higher than those reported following doses of CBD in a different CBD:THC formulation (5 mg CBD +5.4 mg THC): median 2.49 ng/mL in the present study versus 0.39 ng/mL in the 5 mg CBD +5.4 mg THC study [10]. However, THC levels were similar (2.49 ng/mL in the present study compared to 2.50–3.75 n/mL in the 10 mg CBD +10.8 mg THC) [10]. Peak concentrations occurred later in the present study compared to other studies [10, 20]. This delayed peak was influenced heavily by a number of participants with bimodal peak concentrations or with delayed peak concentrations.

A similar finding has been reported previously [20]. However, in that study, the time to maximal concentration was much shorter (median CBD t_max was 4.5 h and THC t_max was also 4.5 h following doses of 16.2 mg THC +15 mg CBD) [20].

Our study reported a mean NPRS pain reduction of 28.8% (median 28.6%) following 1.0 mL bd doses and a mean 34.1% (median 37.5%) reduction following 1.5 mL bd doses. This is similar to that published in systematic reviews of reductions of pain of between 30% and 50% following use of medicinal cannabis in the chronic non-cancer pain setting [7]. The NPRS results were echoed by the BPI results, along with dose-dependent reductions in depressive symptoms and stress but not anxiety. There is preliminary evidence supporting the efficacy of CBD as an anxiolytic and antidepressant and in improving sleep disturbance [21]. Improvement in BPI interference scores was encouraging, given that this is a surrogate for function. The sleep results in the present study were contradictory, with improvements in pain interference with sleep reported in the BPI but paradoxical increases in self-reported sleep disturbance with the MOS-SS. The reason for this is unclear but may be related to the study being underpowered to detect a difference in this endpoint, or differences in the underlying sleep construct each survey is designed to assess [22].

Encouragingly, the efficacy observed was in the setting of typically mild adverse events, as has been reported previously [23, 24]. As with other studies [23, 25], the most common adverse events were headache, dizziness, fatigue, paresthesia, vomiting, nausea, and anxiety. It was of clinical interest that the majority of adverse events occurred during the lower doses of Cybis^® 10:25, which suggests that there is a level of tolerance observed to treatment with CBD:THC combinations. Whether inclusion of CBD in CBD:THC combinations reduces THC-related toxicity requires additional clinical research.

This study has several limitations. First, as it was designed as a first-in-human study, it is not randomized, and therefore the relative efficacy to placebo or active control has not yet been established. Second, dosing was unblinded. Assessment of the efficacy of Cybis^® 10:25 compared to placebo or active control is planned in future larger randomized controlled studies.

The authors thank Belinda Butcher, BSc (Hons), MBiostat, PhD, AStat, CMPP, for providing medical writing assistance on this manuscript in accordance with GPP 2022 [26].

This study protocol was reviewed and approved by the Bellberry Human Research Ethics Committee, approval number 2021-03-305. Written informed consent was obtained from all participants.

The authors declare the following conflicts of interest: Paul Glare and Richard Chye are medical advisors to Cymra Life Sciences Ltd. Mark Bloch, Mark Arya, and Andrew Moore’s institutions received research funding from Cymra Life Sciences Ltd. to perform this study. John Montgomery is an employee of Cymra Life Sciences Ltd.

This study was funded by Cymra Life Sciences Ltd., the developer of Cybis^® 10:25. Medical writing services were funded by Cymra Life Sciences Ltd.

P.G., R.C., and J.M. were involved in the design of the study. M.B., M.A., and A.M. implemented the study at their institutions. All authors were involved in the drafting and review of the manuscript and approved its submission.

The datasets that support the findings of this study are not publicly available due to ethical restrictions. Data requests can be made to the corresponding author which, subject to ethics approval, will not be unreasonably withheld.