UK Medical Cannabis Registry: An Updated Analysis of Cannabis-Based Medicinal Products for Multiple Sclerosis

Abstract

Introduction

Multiple sclerosis (MS) is a neurodegenerative disease presenting with a wide range of motor, sensory, and psychiatric symptoms. Although nabiximols is licensed for MS-induced spasticity, cannabis-based medicinal products (CBMPs) have also displayed promising therapeutic potential for managing pain, sleep, and anxiety. Therefore, further evaluation of CBMP treatment for MS is warranted. This study aimed to assess the efficacy and tolerability of CBMP treatment in patients with MS by investigating changes in MS-specific and general health-related patient-reported outcome measures and adverse events.

Methods

This was a prospective case series including patients with MS enrolled on the UK Medical Cannabis Registry. Changes in MS Quality of Life-54 (MSQOL-54), Generalised Anxiety Disorder-7 (GAD-7), Single-Item Sleep Quality Scale (SQS), and EQ-5D-5L scores were assessed from baseline up to 24 months. The prevalence and severity of all adverse events were also assessed.

Results

This study included 203 patients, of whom 47.29% (n = 96) were female and 80.79% (n = 164) had prior cannabis exposure. Improvements in the MSQOL-54 subscales: change in health, energy, health distress, pain, physical function, and physical role limitations, along with improvements in SQS and EQ-5D-5L scores, were seen at all follow-up times compared to baseline (p < 0.050). A total of 278 adverse events were reported by 26 patients (12.81%). Most adverse events were mild (n = 91, 32.73%) or moderate (n = 138, 49.64%) in severity, with fatigue (n = 27, 13.30%) and spasticity (n = 17, 8.37%) being the most common.

Conclusion

CBMP treatment over 24 months was associated with improvements in health-related quality of life and was well tolerated in patients with MS. Future randomised controlled trials with more representative study populations are needed to establish causal relationships.

Introduction

Multiple sclerosis (MS) is an autoimmune neurodegenerative disease of the central nervous system (CNS). The pathogenesis is characterised by demyelination, neuronal loss, and gliosis secondary to autoreactive T cells crossing the blood-brain barrier [1, 2]. Approximately 2.8 million people globally were estimated to have MS in 2020, with females being twice as likely to be affected as males [3]. MS is categorised into four main subtypes, each reflecting different patterns of disease onset and progression: relapsing-remitting, primary progressive, secondary progressive, and progressive-relapsing [4]. Due to variability in CNS region involvement and disease progression among individuals, MS is associated with a broad spectrum of clinical symptoms. Patients may be affected by motor symptoms, sensory symptoms, psychiatric symptoms, as well as bladder and bowel dysfunction [5, 6]. Furthermore, chronic pain, mostly neuropathic or musculoskeletal in origin, has been reported in 50–75% of patients with MS [7, 8].

Management of MS typically involves disease-modifying therapies to reduce disease progression, as well as treatments to address ongoing symptoms [9]. Baclofen and gabapentin are currently recommended as first-line therapies to manage spasticity, while antidepressants such as amitriptyline and duloxetine are commonly prescribed for neuropathic pain [10, 11]. However, despite the availability of these pharmacological therapies, patients with MS display worse outcomes across multiple health domains compared to the general population [12]. Moreover, although 30% of medications used in symptomatic treatment of MS aim to reduce pain, the efficacy and tolerability of currently recommended therapies are limited [8, 13]. Therefore, this highlights a critical need to explore alternative management strategies to relieve MS-associated symptoms and improve health-related quality of life (HRQoL) in this patient population.

There is increasing evidence for the involvement of the endocannabinoid system (ECS) in modulating inflammatory and neurodegenerative processes [14–18]. Cannabinoid type 1 (CB1) receptors are primarily located in the central and peripheral nervous systems [19–21]. CB1 receptor activation inhibits ascending pathways of the spinothalamic tract, reducing nociceptive signalling [22, 23]. It also regulates pain transmission by inhibiting γ-aminobutyric acid (GABA) release, which stimulates descending inhibitory pathways in the periaqueductal gray and raphe nucleus [22, 23]. Cannabinoid type 2 (CB2) receptors, largely expressed in immune, haematopoietic, and glial cells, work closely with CB1 receptors to regulate pain signalling and reduce the release of proinflammatory cytokines [23, 24].

The cannabis plant contains over 144 phytocannabinoids, which are shown to have a wide range of biological effects [25]. In particular, (−)-trans-Δ⁹-tetrahydrocannabinol (THC) and cannabidiol (CBD) have been widely studied for their therapeutic properties [25, 26]. THC acts as a partial agonist of CB1 and CB2 receptors [27]. CBD competitively binds to fatty acid binding proteins, reducing anandamide transport and breakdown by the enzyme fatty acid amide hydrolase [28].

Through interactions with the ECS, THC and CBD have displayed analgesic, muscle relaxant, neuroprotective, and anti-inflammatory properties in preclinical and clinical studies [22, 29–34]. In the context of MS, this is through direct and indirect interaction with components of the ECS within the peripheral nervous system and CNS. Cannabinoid analgesic effects are proposed to arise from the activation of cannabinoid receptors at peripheral, spinal, and supraspinal levels, resulting in an inhibition of nociceptive transmission, reduction in neurotransmitter release, and activation of inhibitory descending pathways [22]. Decreased excitatory neurotransmitter release and promotion of inhibitory signals by cannabinoids in the spinal cord and basal ganglia are also thought to contribute to their antispasmodic effect [35]. Moreover, inhibition of voltage-gated calcium channels through cannabinoid receptor activation reduces calcium influx and the excitotoxic effects of excessive glutamate, which is suggested to contribute to the neuroprotective and anti-inflammatory effects of cannabinoids [33, 36]. Therefore, cannabis-based medicinal products (CBMPs) containing these phytocannabinoids show promise for managing MS symptoms.

Nabiximols, an oromucosal spray containing THC and CBD in a 2.7:2.5 mg ratio, is licensed for treatment-resistant MS-induced spasticity [37–39]. Multiple randomised controlled trials (RCTs) have found nabiximols results in greater improvements in MS-induced spasticity than placebo [30, 40, 41]. Moreover, sustained reductions in spasticity have been observed with long-term nabiximols use beyond 1 year [42]. Despite its favourable efficacy in some patients, the overall response rate and the magnitude of response are limited. In a double-blind RCT, only 40% of participants using nabiximols achieved an improvement of 30% or more in spasticity severity, and just 17.5% experienced an improvement of 50% or more [30]. In another RCT, only 47.6% of patients using nabiximols experienced a ≥20% improvement in spasticity intensity [43]. This highlights a crucial unmet need for alternative treatment options among patients who do not respond to nabiximols. Moreover, over 80% of patients using nabiximols have experienced an adverse event in some studies, with CNS symptoms such as dizziness and impaired balance being commonly reported [30, 44]. These adverse effects may be related to the THC content, with THC being the main psychoactive component of cannabis [45, 46]. As an oromucosal spray, many oral adverse events have also been reported following nabiximols use, including oral pain, oral mucosal disorder and tooth discolouration [42]. Considering these limitations, further investigation into unlicensed CBMP therapies, including different THC:CBD dose ratios, routes of administration and excipients, is warranted.

Furthermore, with cannabinoids showing improvements in chronic neuropathic and inflammatory pain in animal models, an expanding body of research is investigating the role of CBMPs in conditions associated with chronic pain [47–52]. In an RCT involving 66 patients with MS-related neuropathic pain, 5 weeks of adjunctive treatment with nabiximols resulted in almost twice the mean reduction in pain from baseline compared to placebo [53]. This was further supported in another study of 339 patients, where 14 weeks of nabiximols use was associated with greater reductions in pain compared to placebo [54]. Moreover, numerous studies in patients with MS have explored pain as a secondary outcome and found improvements [55, 56]. However, the quality of evidence for CBMP use for MS-associated pain is low, with most studies being limited by short follow-up times [57]. These concerns have also been echoed in the National Institute for Health and Care Excellence’s review for CBMP use in chronic pain [58]. As pain remains a significant and often refractory symptom in MS, there is a need for higher quality studies assessing long-term CBMP use for MS-associated pain [8, 13, 57].

The UK Medical Cannabis Registry (UKMCR) provides real-world data on patients prescribed CBMPs for a range of conditions [59]. A previous analysis of the UKMCR in patients with MS found CBMP use over 6 months to be associated with improvements in many symptoms, including pain, cognitive function, as well as improvements in HRQoL, anxiety and sleep [60]. This study looks to provide an updated analysis of the UKMCR, including a larger sample size and longer follow-up times to further assess CBMP use in MS. This study had the primary aim to assess changes in patient-reported outcome measures (PROMs) in patients prescribed CBMPs for MS. The secondary aim was to assess the prevalence of adverse events in these patients.

Methods

Study Design

This was a case series looking to assess CBMP treatment in patients with MS. The UKMCR prospectively collects data on patients prescribed CBMPs in the UK and Crown Dependencies [59]. Managed by Curaleaf Clinic, this registry began data collection in 2019, requiring all enrolled patients to have been prescribed CBMPs and provided informed consent [59]. The Central Bristol Research Ethics Committee has provided ethical approval to the UKMCR (Ref: 22/SW/0145).

This study included patients enrolled on the UKMCR with the primary CBMP indication of MS. Patients enrolled on the registry for fewer than 2 years (as of the data extraction date: January 6, 2025) and those who had completed fewer than one baseline PROM were excluded.

Data Collection

Data collection methods for the UKMCR have been outlined in previous registry articles [59, 60].

Patient Demographics and CBMP Prescriptions

Patient demographic data, comorbidities, and prior drug history were recorded during the initial consultation. Demographic data consisted of age, sex, occupation, and body mass index (BMI). Alongside CBMP indications and patient comorbidities, the Charlson Comorbidity Index (CCI) was also recorded. The CCI is a validated scoring system widely used to predict long-term mortality [61]. Drug and alcohol data included tobacco smoking status, typical weekly alcohol consumption (units) and cannabis status prior to starting CBMP treatment. Lifetime cannabis consumption, in gram years, was quantified through the multiplication of mean daily cannabis consumption (in grams) by the years of cannabis use [59].

All CBMP prescriptions were issued by specialists following a multidisciplinary meeting. These prescriptions were directly linked to the UKMCR to record data on CBMP type, route of administration, and daily THC and CBD doses. Patients were also advised to stop consuming any external cannabis sources.

Patient-Reported Outcome Measures

Patients were asked to complete PROMs using an online platform at baseline and at 1, 3, 6, 12, 18, and 24 months. The PROMs assessed were Multiple Sclerosis Quality of Life-54 (MSQOL-54), EQ-5D-5L, Generalised Anxiety Disorder-7 (GAD-7), Single-Item Sleep Quality Scale (SQS), and Patient Global Impression of Change (PGIC).

The MSQOL-54 assesses MS-specific and general HRQoL. It consists of 12 subscales: “physical function,” “role limitations-physical,” “role limitations-emotional,” “pain,” “emotional well-being,” “energy,” “health perceptions,” “social function,” “cognitive function,” “health distress,” “sexual function,” and “overall quality of life.” Alongside these 12 subscales, there are also two single-item measures, “satisfaction with sexual function” and “change in health,” and two summary composite scores for “physical health” and “mental health.” Scoring for each component ranges from 0 to 100, with higher values suggestive of an improved HRQoL [62].

EQ-5D-5L is also a measure of HRQoL, encompassing five domains: “mobility,” “self-care,” “usual activities,” “pain/discomfort,” and “anxiety/depression.” The scoring for each domain ranges from “0 – no problems” to “5 – extreme problems.” This is subsequently converted into country-specific index values, where a value of 1 represents perfect health, whereas values below 0 represent HRQoL worse than death [63].

GAD-7 assesses the severity of anxiety symptoms experienced across the past 2 weeks. The severity of anxiety symptoms is classified as follows: 0–4: minimal, 5–9: mild, 10–14: moderate, and 15–21: severe [64]. A minimal clinically important difference (MCID) in GAD-7 has been reported as a decrease of 4 points or greater [65].

SQS is a single-item measure of sleep quality over the past week. The rating ranges from “0 – terrible” to “10 – excellent” [66]. An MCID in SQS is reported as an increase of 2.6 points or more [66].

PGIC is used to assess the overall patient perception of change in response to a treatment. Scoring for PGIC ranges from “1 – no change” to “7 – a great deal better” [67].

Adverse Events

Patients can report adverse events as they occur or alongside PROM data collection. Alternatively, adverse events can also be documented by clinicians during consultations. All adverse events are recorded according to the Common Terminology Criteria for Adverse Events version 4.0 [68].

Statistical Analysis

Data on patient demographics, CBMP prescriptions, and adverse events were reported using descriptive statistics. This included the mean ± standard deviation and median (interquartile range) for parametric and non-parametric data, respectively.

Following the central limit theorem, changes in all PROMs (except PGIC) were assessed using repeated measures analysis of variance (ANOVA), followed by post hoc pairwise comparisons with Bonferroni correction. Due to the absence of a baseline PGIC value, descriptive statistics were used to report PGIC scores. Multiple imputation with five imputations was used to handle any missing PROM data [69].

Univariable and multivariable logistic regression models were used to assess the effect of various factors on the likelihood of achieving an improvement in EQ-5D-5L index value or MCID in GAD-7 and SQS scores at 24 months. The odds ratio (OR) and 95% confidence intervals (CI) were reported for these analyses. Statistical analysis was performed using R statistical software (R Core Team, version 4.4.3), with a p < 0.050 considered statistically significant.

Results

As of January 6, 2025, 34,563 patients were enrolled onto the UKMCR. After the implementation of the inclusion and exclusion criteria, 203 patients were included in this study. Of these, 52.71% (n = 107) and 47.29% (n = 96) were males and females, respectively (Table 1). The mean age of study participants was 47.93 ± 10.72 years, and the mean BMI was 26.56 ± 6.59 kg/m². The most common employment category was unemployed (n = 100, 49.26%). The median CCI was 0.00 (0.00–6.00).

Table 1.

Demographic data of study participants at baseline (n = 203)

Baseline demographic	n (%)/mean±SD/median [IQR]
Age, years	47.93±10.72
Sex
Male	107 (52.71)
Female	96 (47.29)
BMI, kg/m2	26.56±6.59
Occupation
Clerical support workers	6 (2.96)
Craft and related trades workers	5 (2.46)
Elementary occupations	4 (1.97)
Managers	17 (8.37)
Plant and machine operators, and assemblers	2 (0.99)
Professional	26 (12.81)
Service and sales workers	7 (3.45)
Skilled agricultural, forestry and fishery workers	1 (0.49)
Technicians and associate professionals	6 (2.96)
Other occupations	23 (11.33)
Unemployed	100 (49.26)
Not recorded	6 (2.96)
CCI	0.00 [0.00–6.00]
Tobacco consumption
Current smoker	51 (25.12)
Ex-smoker	106 (52.22)
Never smoked	46 (22.66)
Weekly alcohol consumption, units	0.00 (0.00–4.00)
Cannabis status
Current user	114 (56.16)
Ex-user	50 (24.63)
Never used	39 (19.21)
Lifetime cannabis consumption, g years	5.00 (2.00–20.00)

All percentages, means, medians, SD, and IQR values have been reported to 2 decimal places. SD, standard deviation; IQR, interquartile range.

Most patients were ex-smokers of tobacco (n = 106, 52.22%), and the median weekly alcohol consumption was 0.00 (0.00–4.00) units. 56.16% (n = 114) of patients were already using cannabis before starting CBMP treatment, while 24.63% (n = 50) were ex-users and 19.21% (n = 39) were cannabis naïve. The lifetime cannabis consumption among prior cannabis users was 5.00 (2.00–20.00) g years.

CBMP Prescriptions

Information regarding CBMP preparations as well as CBD and THC doses can be seen in Table 2. The most prescribed CBMP preparation was a combination of dried flower and oil, prescribed to 44.33% of patients (n = 90) at 24 months. Prescriptions of dried flower alone increased from 14.29% (n = 29) at baseline to 23.15% (n = 47) at 24 months. Conversely, prescriptions of oil alone decreased from 39.90% (n = 81) to 23.15% (n = 47) over the same time period. The median daily CBD dose rose from 20.00 (4.50–21.00) mg/day at baseline to 26.05 (15.38–55.00) mg/day and 31.00 (16.50–65.88) mg/day at 12 and 24 months, respectively. A rise in the median daily THC dose was also seen from 19.60 (2.00–22.00) mg/day at baseline to 123.50 (24.50–233.75) mg/day and 148.50 (33.00–253.25) mg/day at 12 and 24 months, respectively.

Table 2.

Route of administration and median daily CBD/THC dose of cannabis-based medicinal products prescribed to study participants at baseline and at 1, 3, 6, 12, 18, and 24 months (n = 203)

​	n (%)
	baseline	1 month	3 months	6 months	12 months	18 months	24 months
Route of administration
Dried flower only	29 (14.29)	29 (14.29)	35 (17.24)	42 (20.69)	49 (24.14)	47 (23.15)	47 (23.15)
Oil only	81 (39.90)	77 (37.93)	60 (29.56)	55 (27.09)	47 (23.15)	45 (22.17)	47 (23.15)
Dried flower and oil	93 (45.81)	97 (47.78)	105 (51.72)	100 (49.26)	98 (48.28)	97 (47.78)	90 (44.33)
Dried flower and pastille	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)	1 (0.49)	5 (2.46)	9 (4.43)
Dried flower, oil, and pastille	0 (0.00)	0 (0.00)	0 (0.00)	0 (0.00)	1 (0.49)	3 (1.48)	3 (1.48)
Othera	0 (0.00)	0 (0.00)	1 (0.49)	2 (0.99)	3 (1.48)	4 (1.97)	6 (2.96)
None prescribed	0 (0.00)	0 (0.00)	2 (0.99)	4 (1.97)	4 (1.97)	2 (0.99)	1 (0.49)

Median daily CBD/THC dose, mg/day	Median [IQR]
	baseline	1 month	3 months	6 months	12 months	18 months	24 months
CBD	20.00 [4.50–21.00]	20.00 [20.00–26.05]	25.50 [16.00–45.40]	25.50 [16.50–50.78]	26.05 [15.38–55.00]	30.50 [16.50–61.00]	31.00 [16.50–65.88]
THC	19.60 [2.00–22.00]	110.00 [10.20–121.00]	114.00 [12.45–198.80]	121.00 [15.10–223.25]	123.50 [24.50–233.75]	148.50 [32.61–244.00]	148.50 [33.00–253.25]

Data are expressed as number (%). All percentages have been reported to 2 decimal places. CBD, cannabidiol; THC, (−)-trans-Δ⁹-tetrahydrocannabinol; IQR, interquartile range.

^aOther CBMP combinations combined to prevent inadvertent reidentification.

Patient-Reported Outcome Measures

Mean scores for the components of MSQOL-54 can be seen in Table 3. Improvements in mean scores were seen at all follow-up times compared to baseline for the following subscales: change in health, energy, health distress, pain, physical function, and role limitations – physical (p < 0.050). Moreover, improvements in the mental health and physical health composite scores were seen at all follow-up times compared to baseline (p < 0.050). Compared to baseline, the mean scores for the emotional well-being subscale improved till 3 months (p < 0.010), while the mean scores for the cognitive and social function subscales improved till 12 months (p < 0.010). Improvements in the overall quality of life scale were seen at 12 and 18 months, compared to baseline (p < 0.001), but not at 24 months (p = 1.000).

Table 3.

Mean ± standard deviation scores of individual Multiple Sclerosis Quality of Life-54 (MSQOL-54) components at baseline and at 1, 3, 6, 12, 18, and 24 months of follow-up (n = 203)

Scale	Follow-up, months	Mean score±standard deviation	p value
Change in health	Baseline	33.25±25.90	Reference
	1	46.67±31.83	<0.001***
	3	50.25±33.28	<0.001***
	6	48.89±34.94	<0.001***
	12	52.46±37.15	<0.001***
	18	54.93±38.30	<0.001***
	24	43.60±39.12	0.018*
Cognitive function	Baseline	48.30±27.62	Reference
	1	56.80±27.58	<0.001***
	3	55.96±31.00	0.003**
	6	57.09±34.59	0.005**
	12	59.06±34.61	<0.001***
	18	55.49±35.80	0.188
	24	54.83±37.00	0.455
Emotional wellbeing	Baseline	57.85±21.27	Reference
	1	64.93±23.35	<0.001***
	3	64.61±24.03	0.002**
	6	62.36±28.70	0.637
	12	62.78±31.00	0.635
	18	61.87±25.52	0.906
	24	62.94±29.53	0.769
Energy	Baseline	25.44±18.84	Reference
	1	33.79±23.10	<0.001***
	3	34.64±24.53	<0.001***
	6	37.85±25.38	<0.001***
	12	36.02±27.09	<0.001***
	18	37.93±28.87	<0.001***
	24	35.43±31.43	0.002**
Health distress	Baseline	36.75±27.49	Reference
	1	46.40±28.94	<0.001***
	3	48.52±31.67	<0.001***
	6	45.52±34.56	0.004**
	12	46.72±34.87	0.007**
	18	52.59±36.31	<0.001***
	24	50.69±38.24	<0.001***
Health perceptions	Baseline	24.93±18.40	Reference
	1	30.02±22.26	0.001***
	3	30.76±23.65	0.003**
	6	30.49±25.87	0.022*
	12	30.89±27.70	0.050
	18	30.57±27.69	0.101
	24	32.61±29.30	0.019*
Pain	Baseline	33.98±24.40	Reference
	1	45.77±27.85	<0.001***
	3	49.21±28.00	<0.001***
	6	47.91±30.44	<0.001***
	12	50.37±34.17	<0.001***
	18	47.81±31.93	<0.001***
	24	46.18±33.54	<0.001***
Physical function	Baseline	26.55±25.47	Reference
	1	33.94±31.08	<0.001***
	3	32.41±30.57	0.004**
	6	41.08±36.73	<0.001***
	12	37.93±37.78	<0.001***
	18	39.66±39.46	<0.001***
	24	36.58±38.01	0.005**
Role limitations – emotional	Baseline	40.23±42.56	Reference
	1	53.04±44.32	0.002**
	3	51.40±43.02	0.010*
	6	48.77±45.75	0.248
	12	54.19±46.41	0.012*
	18	51.07±46.82	0.186
	24	47.13±47.56	1.000
Role limitations – physical	Baseline	15.64±28.66	Reference
	1	34.36±39.94	<0.001***
	3	31.40±37.26	<0.001***
	6	37.56±43.26	<0.001***
	12	36.45±43.24	<0.001***
	18	39.16±44.29	<0.001***
	24	33.50±41.90	<0.001***
Sexual function	Baseline	58.13±33.20	Reference
	1	63.22±33.70	0.114
	3	60.31±36.61	1.000
	6	55.79±38.56	1.000
	12	61.74±39.20	1.000
	18	57.68±42.16	1.000
	24	63.10±42.03	1.000
Sexual function satisfaction	Baseline	44.21±35.45	Reference
	1	49.51±33.92	0.309
	3	50.86±35.65	0.085
	6	44.95±38.65	1.000
	12	46.43±40.22	1.000
	18	46.67±41.97	1.000
	24	47.66±42.84	1.000
Social function	Baseline	41.79±24.17	Reference
	1	53.41±27.56	<0.001***
	3	52.96±27.39	<0.001***
	6	51.11±32.63	0.001**
	12	53.94±32.75	<0.001***
	18	49.06±36.18	0.157
	24	46.88±35.61	0.905
Overall quality of life	Baseline	48.80±17.08	Reference
	1	53.56±20.76	0.070
	3	53.56±20.34	0.100
	6	53.85±22.06	0.066
	12	57.90±24.50	<0.001***
	18	59.70±23.95	<0.001***
	24	52.66±29.07	1.000
Mental health composite	Baseline	47.64±20.91	Reference
	1	56.13±23.49	<0.001***
	3	54.97±24.30	<0.001***
	6	54.48±21.44	<0.001***
	12	57.06±20.95	<0.001***
	18	57.56±24.32	<0.001***
	24	55.87±29.17	0.011*
Physical health composite	Baseline	31.10±17.14	Reference
	1	40.63±20.62	<0.001***
	3	40.55±17.69	<0.001***
	6	41.70±26.61	<0.001***
	12	42.65±27.47	<0.001***
	18	39.26±25.54	<0.001***
	24	41.80±26.70	<0.001***

Mean and standard deviation values have been reported to 2 decimal places, and p values have been reported to 3 decimal places.

Table 4 shows the mean scores of other general health-related PROMs. The mean SQS scores improved at all follow-up times compared to baseline (p < 0.010), while improvements in GAD-7 scores were only seen at 1, 3, 6, and 18 months of follow-up (p < 0.010). The proportion of patients achieving an MCID in SQS and GAD-7 scores at 24 months was 47.29% (n = 96) and 37.44% (n = 76), respectively. Improvements in EQ-5D-5L index values were seen at all follow-up times compared to baseline (p < 0.050). When assessing the individual domains of the EQ-5D-5L, improvements were seen in the pain and discomfort subscale up to 24 months of follow-up (p < 0.001), while improvements in the mobility and usual activities domain were seen only up to 3 and 6 months of follow-up (p < 0.010), respectively. The PGIC scores were 4.98 ± 1.64, 4.71 ± 2.19 and 5.08 ± 2.04 at 1, 12 and 24 months, respectively.

Table 4.

Mean ± standard deviation scores of general health related PROM scores at baseline and at 1, 3, 6, 12, 18, and 24 months of follow-up (n = 203)

Scale	Follow-up, months	Mean score±standard deviation	p value
GAD-7	Baseline	7.50±6.03	Reference
	1	5.26±5.18	<0.001***
	3	5.47±5.37	<0.001***
	6	5.14±6.09	<0.001***
	12	6.84±7.15	1.000
	18	5.35±6.55	0.003**
	24	6.44±6.99	1.000
SQS	Baseline	4.55±2.71	Reference
	1	6.37±2.68	<0.001***
	3	6.20±2.58	<0.001***
	6	6.16±3.14	<0.001***
	12	5.82±3.48	<0.001***
	18	5.74±3.73	0.005**
	24	6.85±3.59	<0.001***
EQ-5D-5L Index Value	Baseline	0.28±0.35	Reference
	1	0.45±0.34	<0.001***
	3	0.44±0.34	<0.001***
	6	0.37±0.42	0.009**
	12	0.43±0.45	<0.001***
	18	0.40±0.45	0.027*
	24	0.42±0.43	0.002**
EQ-5D-5L Mobility	Baseline	3.27±1.16	Reference
	1	2.97±1.26	<0.001***
	3	2.94±1.29	<0.001***
	6	3.04±1.40	0.208
	12	3.12±1.52	1.000
	18	3.01±1.59	0.450
	24	2.95±1.63	0.199
EQ-5D-5L Self Care	Baseline	2.49±1.19	Reference
	1	2.35±1.23	0.740
	3	2.28±1.30	0.358
	6	2.31±1.33	1.000
	12	2.58±1.55	1.000
	18	2.65±1.54	1.000
	24	2.55±1.56	1.000
EQ-5D-5L Usual Activities	Baseline	3.10±1.07	Reference
	1	2.73±1.17	<0.001***
	3	2.79±1.30	0.003**
	6	2.64±1.33	<0.001***
	12	2.81±1.48	0.132
	18	2.89±1.52	1.000
	24	2.85±1.49	0.549
EQ-5D-5L Pain and Discomfort	Baseline	3.58±0.96	Reference
	1	2.84±1.11	<0.001***
	3	2.69±1.05	<0.001***
	6	2.60±1.16	<0.001***
	12	2.70±1.34	<0.001***
	18	2.88±1.44	<0.001***
	24	2.72±1.33	<0.001***
EQ-5D-5L Anxiety and Depression	Baseline	2.41±1.14	Reference
	1	2.25±1.21	1.00
	3	2.21±1.20	0.50
	6	2.14±1.31	0.21
	12	1.97±1.21	<0.001***
	18	2.30±1.36	1.00
	24	2.17±1.32	0.60
PGIC	Baseline	–	–
	1	4.98±1.64	–
	3	5.20±1.57	–
	6	5.23±1.72	–
	12	4.71±2.19	–
	18	5.01±2.14	–
	24	5.08±2.04	–

Mean and standard deviation values have been reported to 2 decimal places, and p values have been reported to 3 decimal places. PROM, patient-reported outcome measure; GAD-7, Generalised Anxiety Disorder-7; SQS, Single-Item Sleep Quality Scale; PGIC, Patient Global Impression of Change.

Adverse Events

A total of 278 (136.94%) adverse events were reported by 26 (12.81%) patients. Most adverse events were mild (n = 91, 32.73%) and moderate (n = 138, 49.64%) in severity, as shown in Table 5. The percentage of severe and life-threatening/disabling adverse events was 16.91% (n = 47) and 0.72% (n = 2), respectively. The most reported adverse events were fatigue (n = 27, 13.30%), spasticity (n = 17, 8.37%), generalised muscle weakness (n = 16, 7.88%), and somnolence (n = 16, 7.88%).

Table 5.

Adverse events reported by study participants

Adverse event	n				Total, n (%)
	mild	moderate	severe	life-threatening/disabling
Fatigue	6	11	10	0	27 (13.30)
Spasticity	4	7	5	1	17 (8.37)
Generalised muscle weakness	0	9	7	0	16 (7.88)
Somnolence	0	12	4	0	16 (7.88)
Concentration impairment	7	8	0	0	15 (7.39)
Lethargy	5	10	0	0	15 (7.39)
Ataxia	3	8	2	0	13 (6.40)
Blurred vision	5	7	0	0	12 (5.91)
Constipation	9	3	0	0	12 (5.91)
Dizziness	2	7	1	0	10 (4.93)
Dry mouth	7	3	0	0	10 (4.93)
Headache	4	4	2	0	10 (4.93)
Vertigo	6	4	0	0	10 (4.93)
Abdominal pain	3	3	3	0	9 (4.43)
Insomnia	2	5	2	0	9 (4.43)
Cognitive disturbance	2	6	0	0	8 (3.94)
Confusion	1	5	1	0	7 (3.45)
Dyspepsia	4	3	0	0	7 (3.45)
Nausea	5	2	0	0	7 (3.45)
Tremor	4	3	0	0	7 (3.45)
Amnesia	3	1	0	0	4 (1.97)
Diarrhoea	1	0	3	0	4 (1.97)
Lung infection	0	2	2	0	4 (1.97)
Urinary tract infection	0	2	2	0	4 (1.97)
Fall	0	3	0	0	3 (1.48)
Pharyngitis	0	3	0	0	3 (1.48)
Anorexia	2	0	0	0	2 (0.99)
Delirium	1	1	0	0	2 (0.99)
Dysgeusia	0	1	1	0	2 (0.99)
Fever	2	0	0	0	2 (0.99)
Rash	1	1	0	0	2 (0.99)
Other (single occurrences)	2	4	2	1	9 (4.43)
Total	91	138	47	2	278 (136.94)

“Other” refers to adverse events reported only once and aggregated to avoid inadvertent reidentification. All percentages are reported to 2 decimal places.

Univariable and Multivariable Analysis

EQ-5D-5L Index Value

Univariable analysis revealed that those with a BMI less than 20.00 kg/m² had a lower likelihood of showing an improvement in the EQ-5D-5L index values at 24 months (OR = 0.32, 95% CI = 0.12–0.86, p = 0.026) (online suppl. Table 1; for all online suppl. material, see https://doi.org/10.1159/000549178). This was also seen in multivariable analysis (OR = 0.24, 95% CI = 0.08–0.71, p = 0.012). Multivariable analysis also showed that females had a higher likelihood of reporting an improvement in the EQ-5D-5L index value (OR = 2.60, 95% CI = 1.29–5.43, p = 0.009) (online suppl. Table 2).

Generalised Anxiety Disorder-7

Upon univariable analysis, baseline SQS scores below 4 (OR = 2.82, 95% CI = 1.36–6.13, p = 0.007) and baseline GAD-7 scores above 4 (scores 5–9: OR = 29.00, 95% CI = 8.10–185.96, p < 0.001; scores 10–14: OR = 53.67, 95% CI = 13.93–358.10, p < 0.001; scores 15–21: OR = 110.00, 95% CI = 25.82–788.32, p < 0.001) had higher odds of achieving an MCID in GAD-7 scores at 24 months. However, patients aged over 60 years were less likely to report an MCID in this PROM scale (OR = 0.25, 95% CI = 0.07–0.73, p = 0.016) (online suppl. Table 3). On multivariable analysis, baseline GAD-7 scores above 4 continued to be associated with a higher likelihood of achieving an MCID in GAD-7 scores (scores 5–9: OR = 47.06, 95% CI = 11.30–334.81, p < 0.001; scores 10–14: OR = 86.44, 95% CI = 17.96–685.64, p < 0.001; scores 15–21: OR = 181.75, 95% CI = 31.65–1669.28, p < 0.001). Also, a CBD dose of between 16.50 and 30.99 mg/day had higher odds of reporting an MCID in this PROM scale (OR = 4.55, 95% CI = 1.27–17.97, p = 0.024) (online suppl. Table 4).

Single-Item Sleep Quality Scale

Patients with baseline SQS scores below 7 (scores 0–3: OR = 13.58, 95% CI = 5.59–38.53, p < 0.001; scores 4–6: OR = 10.46, 95% CI = 4.12–30.67, p < 0.001) or baseline GAD-7 scores above 14 (OR = 3.57, 95% CI = 1.47–9.13, p = 0.006) had a higher likelihood of achieving an MCID in SQS scores at 24 months in univariable analysis (online suppl. Table 5). On multivariable analysis, baseline SQS scores below 7 (scores 0–3: OR = 14.25, 95% CI = 5.29–44.60, p < 0.001; scores 4–6: OR = 10.19, 95% CI = 3.75–31.78, p < 0.001) continued to be associated with higher odds of an MCID in SQS scores (online suppl. Table 6).

Adverse Events

Univariable analysis showed that patients who had prior cannabis use before starting CBMP treatment were less likely to report an adverse event (current users: OR = 0.34, 95% CI = 0.13–0.88, p = 0.024; ex-user: OR = 0.25, 95% CI = 0.06–0.83, p = 0.031). On the other hand, age over 60 years (OR = 12.86, 95% CI = 2.02–251.09, p = 0.022) and a CBD dose over 30.99 mg/day (dose 31.00–65.87 mg/day: OR = 5.17, 95% CI = 1.22–35.50, p = 0.044; dose 65.88 mg/day and over: OR = 6.46, 95% CI = 1.59–43.57, p = 0.020) had higher odds of experiencing an adverse event (online suppl. Table 7). On multivariable analysis, an age between 41 and 60 years (age 41–50 years: OR = 14.99, 95% CI = 2.00–337.43, p = 0.026; age 51–60 years: OR = 40.57, 95% CI = 4.31–1,072.45, p = 0.005) and a CBD dose over 30.99 mg/day (dose 31.00–65.87 mg/day: OR = 19.56, 95% CI = 2.58–429.44, p = 0.013; dose 65.88 mg/day and over: OR = 29.45, 95% CI = 3.73–673.31, p = 0.006) were associated with a greater likelihood of experiencing an adverse event (online suppl. Table 8).

Discussion

This was an observational study of the UKMCR, investigating the association between CBMP prescriptions and patient-reported symptoms in patients with MS. Improvements were seen across multiple components of the MSQOL-54, including pain, energy, physical and mental wellbeing (p < 0.050). Positive changes in general HRQoL measures, including SQS and EQ-5D-5L, were also seen (p < 0.050). Treatment with CBMPs was well tolerated, with adverse events being reported by 12.81% of patients, and most adverse events were mild or moderate in severity.

Initiation of CBMPs was associated with improvements across a range of MS-related symptoms, including sustained reductions in pain across 24 months. This has been supported in a 5-week RCT, where nabiximols led to greater reductions in pain severity than placebo [53]. Similarly, in the SAVANT trial, 12-week treatment with nabiximols also led to improvements in MS-related pain [41]. While the existing literature has largely focused on an oromucosal spray as the route of administration [41, 53, 54], in the present study, most patients used a combination of vaporised dried flower and sublingual oils. The proportion of patients using sublingual oils decreased over time. While this may partly be due to inhaled preparations being less expensive per dose of active pharmaceutical ingredient in this private setting, the pharmacokinetics of the different modes of administration should also be considered. Inhaled CBMP preparations have a faster onset of action, potentially more suited in managing acute episodes of pain [70, 71]. In contrast, the longer duration of action of sublingual preparations may be beneficial for managing chronic ongoing pain [71]. Patients with MS may experience a diverse spectrum of pain arising from neuropathic or musculoskeletal aetiologies, which can also be either acute or chronic in nature [7, 8]. Therefore, future RCTs should aim to investigate multiple CBMPs across a range of formulations and doses of cannabinoids to optimise pain management and enable treatment to be tailored according to the characteristics of the pain.

However, conflicting findings have also been reported. In an RCT by Wade et al. involving 160 patients with MS, reductions in pain were reported in both the CBMP and placebo groups, resulting in no significant difference between the two groups [40]. Similarly, a large proportion of placebo responders were also seen in an RCT by Langford et al. investigating CBMP use for neuropathic pain relief in patients with MS [54]. These findings highlight the strong influence of the placebo effect when assessing pain severity [72]. This is particularly important to consider given the subjective nature of pain, typically measured through patient-reported questionnaires, which may give rise to expectation bias. Furthermore, the placebo effect may even be amplified in this setting due to increased expectancy due to positive media attention on the therapeutic potential of CBMPs [72]. These factors strengthen the need for further double-blind studies with placebo groups to better evaluate the true effect of CBMPs in MS-related pain. Furthermore, the subjective nature of pain, coupled with the considerable pain mechanism heterogeneity seen in MS, underscores the need for a more comprehensive assessment of pain in these patients [7, 8]. This could be mediated through the use of multidimensional pain scales, for example, the McGill Pain Questionnaire, which assesses the location, characteristics and severity of the pain experienced, along with changes in the affective response to pain [73]. Despite the drawback of being more time-consuming, the use of multi-dimensional pain scoring systems could provide a more detailed evaluation of the impact of CBMPs on pain in patients with MS.

Improvements in HRQoL were demonstrated using multiple scales in this study, including the physical and mental health composite scores of MSQOL-54 and EQ-5D-5L index values. Another observational study assessing nabiximols use in patients with MS also reported improvements in both MSQOL-54 composite scores across 3 months [74]. Furthermore, improvements in HRQoL have also been seen in prior UKMCR studies [60, 75]. However, in the current study, an improvement in the overall quality of life subscale was only seen at 12 and 18 months. This may suggest that although early symptomatic relief may be exhibited with CBMP use, changes in overall HRQoL may take a longer time to manifest, given the wide range of symptoms seen in MS. Several studies of patients with MS have reported no improvements in HRQoL measures following CBMP treatment, despite reporting symptomatic relief [41, 54, 76]. A potential reason for this could be due to the heterogeneity in MS disease progression, with individuals potentially experiencing either progression or remission of MS [4, 77]. Future studies should aim to monitor disease progression during the follow-up period and ideally have an adequate control arm with balanced MS severity to avoid confounding.

Sleep disturbances are common in MS secondary to direct CNS effects, in addition to the effects of symptoms, such as pain [78]. CBMP treatment was associated with improvements in sleep quality, evidenced by increases in the mean SQS scores at all follow-up times up to 24 months. This builds on the findings of previous studies supporting CBMP efficacy in improving sleep. In a retrospective medical record review, 40% of patients with MS reported an improvement in sleep following CBMP treatment [79]. Moreover, in a double-blind RCT, nabiximols use in MS led to reduced sleep disruption compared to placebo [43]. Zajicek et al. [55] also reported greater improvements in sleep with CBMP use over 12 months. The ECS plays a key role in regulating the sleep-wake cycle [80], and therefore, CBMPs may offer a promising option for improving sleep in individuals with MS.

Mental health conditions, including anxiety, are prevalent among individuals with MS [5]. Several studies have suggested CBMPs have anxiolytic properties [81, 82], which supports the reductions in GAD-7 scores seen in this study at 1, 3, 6 and 18 months. This is thought to be mediated through interactions with 5-hydroxytryptamine 1A receptors and transient receptor potential vanilloid-1 channels [83]. However, improvements in the EQ-5D-5L Anxiety and Depression domain were only seen at 12 months. Careful CBMP dosing is crucial when assessing anxiety as the anxiolytic properties of CBMPs are thought to be dose-dependent [84]. Notably, THC has been shown to have biphasic effects, with higher THC doses exhibiting anxiogenic properties [84, 85]. Therefore, future studies involving patients with MS should comprehensively assess changes in mental health symptoms, including anxiety, and explore varying THC:CBD dose ratios.

CBMPs were generally well tolerated by patients in this study, with most adverse events being mild or moderate in severity. Although the prevalence of adverse events was higher in this study than in previous UKMCR analyses [60, 86, 87], this is expected given the longer follow-up time of 24 months, leading to the accumulation of adverse events over time. Interestingly, prior studies in patients with MS evaluating currently approved symptomatic and disease-modifying treatments have also frequently reported high rates of side effects as well as a broad spectrum of side effects [88–91]. For example, in a cross-sectional study investigating the use of disease-modifying medications in 191 patients with MS, a total of 484 adverse drug reactions were reported [88]. The multifaceted, complex pathophysiology of MS, involving multiple CNS regions, may partly contribute to the high rate of adverse events seen when managing this condition [1, 2]. Furthermore, polypharmacy is common in these patients, which may further amplify the incidence of adverse events due to interactions between various drug treatments [92]. With such heterogeneity in how the disease manifests across patients, variability in treatment response is to be expected [4, 6]. Therefore, a personalised medicine approach may be beneficial to reduce the incidence of side effects and potentially enhance treatment adherence and effectiveness [93].

The most common adverse events were fatigue, spasticity, and generalised muscle weakness. However, it is important to note that these are also symptoms of MS [94, 95]. As this study did not assess whether reported adverse events were treatment-related, it is unclear whether these occurred secondary to CBMP treatment or reflected a deterioration of MS. In clinical trials of nabiximols, oromucosal side effects have been common [42], yet this was not found in the present study. This is likely due to differences in routes of administration and the excipient used in medium-chain triglyceride oils compared to ethanol and propylene glycol use in nabiximols [39]. This further amplifies the need to investigate alternative routes of CBMP administration for MS beyond nabiximols.

When interpreting findings, the limitations of this study must also be noted. Firstly, the absence of a placebo group coupled with the observational study design meant that any conclusions regarding causality could not be made. Furthermore, although MS disproportionately affects females [96, 97], they comprised less than half of the study population. This is important to consider because the progression of MS has been shown to vary based on sex [98, 99]. Moreover, some studies have reported sex-related differences in CBMP effects [100, 101]. Therefore, the underrepresentation of females in this study limits the generalisability of findings to the wider MS patient population. Also, over 80% of patients had previously used cannabis before starting CBMP treatment. As there have been concerns regarding the development of tolerance with extended cannabis use [102, 103], the examination of the effects in cannabis-naïve populations is important. Future studies should therefore aim to recruit more female and cannabis-naïve patients when assessing CBMP efficacy in MS. Another limitation of this study was that the type of MS was not recorded, and disease progression was not monitored over time. Both these factors could substantially impact changes in MS symptoms and HRQoL, making it challenging to evaluate the efficacy and tolerability of CBMPs in these patients. Moreover, self-reported questionnaires are prone to recall and social desirability bias, which may also skew findings.

In conclusion, this observational study found CBMP treatment was associated with improvements in many HRQoL measures, including pain and sleep in patients with MS. Also, CBMP use over 2 years was generally well tolerated. However, study limitations meant causal relationships could not be established. Therefore, there is a need for long-term RCTs with more clinically representative study populations to better assess CBMP treatment for MS.

Statement of Ethics

Ethical approval provided by the Southwest-Central Bristol Research Ethics Committee (Reference No.: 22/SW/0145). All participants completed written, informed consent prior to enrolment in the registry.

Conflict of Interest Statement

Yashvi Shah has no conflicts of interest. Simon Erridge is Research Director at Curaleaf Clinic. Evonne Clarke is the Patient Care Director at Curaleaf Clinic. Katy McLachlan is Chief Pharmacist at the Curaleaf Clinic. Ross Coomber is the Operations Director at Curaleaf Clinic. James Rucker is a consultant psychiatrist at Curaleaf Clinic. James Rucker is funded by a fellowship (CS-2017-17-007) from the National Institute for Health Research (NIHR). Mark Weatherall is a consultant Neurologist at Curaleaf Clinic. Mikael Hans Sodergren is the Chief Medical Officer at Curaleaf International. The authors have no other relevant affiliations or financial involvement with any organisation or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

Yashvi Shah, Simon Erridge, Evonne Clarke, Katy McLachlan, Ross Coomber, James Rucker, and Mark Weatherall performed material preparation and data collection. Yashvi Shah, Simon Erridge, and Mikael Hans Sodergren analysed the data and wrote the first draft of the manuscript. All authors contributed to the study’s conception and design, commented on previous versions of the manuscript, and read and approved the final manuscript.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

The data that support the findings of this study are not publicly available to preserve the privacy of individuals’ pseudonymised data but are available from the Data Management Committee of the UK Medical Cannabis Registry upon reasonable request. Further enquiries can be directed to the corresponding author.