A large Australian longitudinal cohort registry demonstrates sustained safety and efficacy of oral medicinal cannabis for at least two years.

Following regulatory access to the medical prescription of Good Manufacturing Product (GMP)-grade medicinal cannabis (MC) in November 2016, up to 100,000 Australians are now actively taking regularly prescribed MC [ 1 ]. Australians can access a prescription for MC from their treating physician for a wide range of clinical conditions via a Special Access Scheme [ 1 , 2 ]. Recreational cannabis (RC) remains illegal in nearly all States and Territories. The continuing proscription of RC and initial cannabis negative urinary screen of our cohort provides a unique environment to evaluate oral MC where efficacy and safety can be assessed by Real World Data (RWD), with likely less conflation of privately consumed unregulated and indeterminate dosed RC or inhaled MC. The number of medicinal cannabis producers and products has rapidly increased in Australia, with at least 375 available MC products and brands, varying in ranges of ratios, profiles, concentrations, excipients, and delivery systems. The Therapeutic Goods Administration (TGA), Australia’s therapeutics goods regulator, has grouped MC products into five categories reflecting the varying concentrations and ratios of the two major cannabinoids, Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD): 1. CBD, 2. CBD-dominant, 3. Balanced, 4. THC-dominant and 5. THC [ 3 ]. The TGA regulates MC production standards through the Therapeutic Goods (Standard for Medicinal Cannabis) (TGO 93) Order 2017 and the Office of Drug Control. This standard provides appropriate regulatory controls to ensure quality, stability, and safety. THC is a ’Controlled Drug’ under Schedule 8 (S8) of the Poisons Standard. CBD products are Schedule 4 and must be prescribed by a registered medical practitioner and contain at least 98% CBD and 2% or less of other minor cannabinoids including THC. Randomised controlled trials (RCTs) are the gold standard for assessing efficacy of pharmaceuticals but have been challenging in the area of MC because of the various formulations (oral, inhaled, topical), varying concentration and ratios of cannabinoids, and the generic nature of MC. This has limited research in RCTs [ 4 ]. RWD is a mechanism for bridging the evidentiary gap and can help to inform design of RCTs on clinical indications and doses of cannabinoids. RWD studies, by definition, have broader inclusion criteria, which can provide additional and unexpected insights into the safety and efficacy of MC for those who are either ineligible or not represented in RCTs [ 5 ]. Recent observational studies and RWD analysis of registries in the UK and Canada have reported on the safety of MC and shown improvements in outcomes such as pain, sleep, anxiety, and quality of life (QoL). Most notably, these studies have nearly all included smaller numbers of patients (fewer than 200) and have reported on shorter outcomes at 6 or 12 month follow up periods. In addition, they have little or no information on doses or ratios of THC and CBD [ 6 – 9 ]. A recent study by Schleider et al. (2022) using registry data of 10,000 patients from Israel’s largest clinic, observed high safety, decreases in pain levels and improvements in QOL on 4,166 patients that reported at a 6-month follow-up; however, this study included a range of MC formulations—smoking, vaporised or sublingual oil [ 10 ]. MC registries in other jurisdictions, where RC and inhaled products are included, have been required to estimate dose and exposure to these major cannabinoids based on patient reported usage [ 8 , 11 , 12 ], making it difficult to determine optimal dose. The Australian Emyria Clinical e-Registry (AECeR) describes the longitudinal monitoring of cannabis naïve patients, who commence on a defined dose of oral MC utilising the available range of TGA regulated MC products in Australia. Oil-based plant-derived oral MC has been the dominant prescribed product in Australia, however inhaled “flower” products have recently increased, accounting for up to 40% of MC prescriptions [ 1 ]. MC products in Australia attract no government subsidy and little private subsidisation, with Department of Veteran Affairs subsidies available only for specific approved conditions [ 13 ]. Nevertheless, the out-of-pocket cost for MC is decreasing and is approximately $AUD 2–4,000 per year [ 1 ]. We present up-to-date data from the AECeR, which commenced in December 2018 and has monitored nearly 4,000 patients taking prescribed oral MC for up to 24 months.

Of 6,523 patients enrolled for assessment at Emerald Clinics between Dec 2018 and April 2022, 3,961 patients completed initial assessments and questionnaires for prescription of oral MC. Patient demographics at baseline are reported in Table 2 , showing an even distribution of gender, a mean age of 56.1 years (range 2–96 years, SD 19.28). Of the 57.9% who reported education level, only 8.8% did not complete secondary schooling, and 53.4% of patients were not part of the labour force due largely to being retirees. This also includes children and those electing not to work. Notes a Includes those still attending primary or high school as well as anyone where the highest level of educational achievement could not be determined. b Includes individuals looking for full-time or part-time work, students, retirees, and those not working by choice (such as homemakers), in addition to anyone unable to work due to a medical or health condition. Table 3 shows the medications used by participating patients recorded at baseline. Note that the level of non-prescribed or recreational cannabinoids was low (0.7%), further highlighting not only the continued prohibition of RC in Australia but also the key inclusion requirement of returning a THC-negative urinary result before commencing treatment (with the only exception being under compassionate grounds, which corresponds to an exceedingly small number of patients to date, i.e., 3). 46.2% of patients were taking opioids, whilst 41.8% were taking antidepressants and 33.7% benzodiazepines. “Other” includes medication for concomitant chronic disease such as diabetes, hypertension or chronic lung disease. Notes a Corresponds to (medicinal) cannabis use by individuals at baseline prior to commencing treatment. The primary diagnosis of participating patients is noted in Table 4 and indicates the majority (71.9%) of patients were prescribed oral MC for: chronic pain related conditions, mental health disorders (15.4%). Other conditions include neurodegenerative diseases, irritable bowel syndrome and chronic fatigue syndrome. Notes a Includes back pain and complex regional pain syndrome. b Includes alcohol use disorder, Alzheimer’s disease, anorexia and wasting, attention deficit hyperactivity disorder, autism, behavioural disorder, chemotherapy-induced nausea and vomiting, chronic fatigue syndrome, dementia, endometriosis, essential tremor, hereditary spastic paraplegia, inflammatory bowel disease, irritable bowel syndrome, motor neuron disease, obsessive compulsive disorder, panic disorder and benzodiazepine dependence, refractory nausea and vomiting, spasticity, tinnitus, and Tourette syndrome. Table 5 describes the number of oral MC products prescribed with 11,951 (81.2%) prescriptions for patients taking one oral MC product mostly containing a “Balanced” ratio of TCH/CBD 7,400 (50.3%) and 4,570 (31.1%) taking "CBD-only". For those taking more than one product the majority were taking "CBD-only" during the day and a “THC-dominant" product at night (data not shown). The overall median daily dose was: THC 10.0 mg, CBD 22.5mg. Fig 1 demonstrates prescribing patterns over time for THC/CBD dose and ratios. Total cannabinoid dose rose to 87.9mg at 6 months and remained stable over the next 2 years. Historically, prescribing ratios have changed over two years. Balanced product was predominant in 2019 (90%) down in 2022 to 33%, and a rise of CBD-only product from <10% in 2019 to 45% of all prescribed products in 2022. Fig 2 shows scores from the RAND 36-Item Short Form Health Survey (SF-36) for routine monitoring and assessment of well-being and care outcomes. Note that the eight subscale scores are divided into the respective Physical (Panel A) and Mental health (Panel B) component domains. Overall, the figures indicate a statistically significant and sustained (p<0.001) improvement across all measured parameters over the two year sample window (see Table 6 for the associated p -values). Notes a Compares the respective scores between baseline and 3 months after. b Compares the respective scores between baseline and 12 months after. c Compares the respective scores between baseline and 24 months after Fig 3 then displays scores from the Brief Pain Inventory–Short Form (Panel A), the Depression Anxiety Stress Scale (Panel B) and the Insomnia Severity Index questionnaires (Panel C) used for routine assessment of well-being, sleep and care outcomes. Here as well, the corresponding figures suggest both a sustained and statistically meaningful (p<0.001) improvement in all measured parameters over the sample period. Next, Fig 4 demonstrates the clinician reported outcomes based on the respective Clinical Global Impression (CGI) subscale scores. Most notably, these measures are suggestive of a statistically relevant and persistent clinician perceived improvement of the average patient’s global functioning after initiating MC (p<0.001). This finding is perhaps particularly remarkable when considering the substantial efficacy of this health intervention, even after accounting for potential adverse effects experienced by patients (as captured by the Efficacy Index subscale of the CGI measure). Finally, Table 6 compares mean scores between baseline and the respective follow-up windows at 3, 12, and 24 months after commencing treatment. The results confirm the findings in Figs 2 – 4 –that is, we observe statistically important differences in scores at least at the 0.1 percent level of statistical significance for each follow-up window relative to the corresponding baseline for up to two years. Moreover, the mean number of medications each patient was taking initially (6.3: 95%CI 6.1–6.4) nearly halved at 3 months (3.3: 95%CI 3.2–3.4, p<0.001) continued to be significantly reduced at 12 months (5.1: 95%CI 4.9–5.2, p<0.001), but by 24 months was not significantly different (6.4: 95%CI 6.1–6.7), p = 0.5). Of the 3,961 patients included in the analysis, 1,477 patients (37.3 percent) reported experiencing at least one adverse event deemed by the treating doctor to be possibly, likely or definitely related to the oral MC treatment. Table 7 represents an overview of the most frequent types of treatment related adverse events (TRAEs) across the different levels of clinician-assessed severity. Sedation/sleepiness and dry mouth are the two most commonly reported TRAEs, with the majority (68.2 percent and 79.9 percent, respectively) assessed as “mild”. There were 77 severe TRAEs (<2%) requiring a dose adjustment or cessation of oral MC treatment without lasting sequelae. Two isolated TRAEs (hallucination and mania) were considered serious, which is defined as an important medical event requiring hospitalisation or lifesaving intervention.

This large Australian longitudinal cohort registry of cannabis naïve, complex chronic disease patients treated with oral MC for over 24 consecutive months, demonstrates safety of oral generic medicinal cannabis, and demonstrated oral MC improves patient and clinician reported impact of pain, sleep and well-being. The AECeR addresses some of the limitations inherent to RWD and previously published cannabis registries. The detailed data curation and rigour of a very large bespoke registry, with a heterogeneous complex cohort, over an extended period of time, with high retention rates, provides greater reassurance about efficacy and safety of oral MC. It also provides detailed information on oral doses of THC and CBD to inform future studies. Further sub analyses with regard to specific clinical indications and patient reported outcomes are planned and future matched cohort or appropriately powered randomised controlled studies should be considered.

The AECeR is the first Australian national web-based medicinal cannabis treatment electronic registry. It commenced in December 2018 and is privately owned by Emyria Ltd. This study reviews the use of medical cannabis for more than 2 years, in the largest cohort review of oral MC to date. Note that the AECeR is a continuous ongoing enrolment registry, with the number of patients commencing treatment and being tracked across follow-ups increasing over the sample period from December 2018 to April 2022. This means that there are fewer patient baseline data at earlier time points than more recent time points, as the number of enrolled patients increased. Importantly, retention rates were maintained at nearly 70% at 12 months. Patients who attended the national Emerald Clinics Network and were enrolled in the registry between December 2018 to April 2022 are all included in this analysis. All patients included in the registry have undergone a comprehensive assessment by a multidisciplinary team. Baseline data were prospectively entered by patients and clinicians including demographic and routine clinical information, comorbidities, concomitant medication, alcohol and other drug use, and symptom-related data. Clinician and patient standardised validated questionnaires were completed and reviewed to assess the degree of impairment of physical and mental health function, daily activities, quality of life, adverse events, dosing and additional information required for personalised patient care. A urine screening test for THC was conducted at baseline. Presence of urinary-THC was an exclusion for AECeR except in compassionate use (e.g., palliative care). Moreover, pregnancy and breast feeding, serious cardiac disease, and serious mental health conditions (including past history of psychosis and suicidality) were all precluded from any prescriptions of oral MC and thus, were also omitted. The standard practice database was of clinical trial grade, with all staff handling patient data having completed the International Conference on Harmonisation–Good Clinical Practice and data privacy training. Participants were prescribed only oral oils or capsules, available on the Special Access Scheme–Category B (SAS-B), for which a Certificate of Analysis demonstrating compliance and stability was available. This ensures that all products were within expiry and contained the prescribed active ingredients and excipients. All registered patients (or legal guardians of those without capacity to give consent, including minors) gave written informed consent and agreed to the use of their de-identified data for research purposes. Review by a Human Research Committee was not required, as all assessments were conducted as part of routine clinical care in line with the Special Access Scheme requirements. This publication involves the use of existing collections of data or records that contain only non-identifiable data about human beings. Australian research is guided by the Australian National Health and Medical Research Council’s “National Statement on Ethical Conduct in Human Research (2007, updated 2018)” [ 14 ], which permits non-Human Research Ethics Committee (HREC) pathways for research that is deemed to be of low or negligible risk. Consistent with this provision, and in following internal research review Emyria additionally consulted an independent ethics committee chair and an independent ethics consultant, who agreed that the collection and use of de-identified patient data for this registry protects the rights safety and well-being without risk to the individuals. The patient cohort described had regular, approximately two monthly, clinical monitoring visits. Data collection was repeated, reviewed, and monitored for adherence and validity. The patients’ physical and mental health status was assessed through clinical assessment and validated surveys completed by the patient and health care professionals. The surveys included: Clinical Global Impression (CGI), Brief Pain Inventory (BPI), Depression Anxiety and Stress Scale 21 (DASS-21), Insomnia Severity Index (ISI) and the RAND 36 Item Short Form Survey (SF-36). Where clinically appropriate, additional questionnaires were also completed, such as the Cannabis Use Disorders Identification Test (CUDIT), Douleur Neuropathique en 4 (DN4), IBS Symptom Severity Score (IBS-SSS) for Irritable Bowel Syndrome (IBS) and the Autistic Behaviour Checklist (ABC) for Autism Spectrum Disorder (ASD). These instruments were selected to ensure quality-assured, validated and standardised documentation of all treatment-relevant data for the routine care of patients. All patients provided written informed consent for health data collection and use, and agreed to the prescription and regular monitoring of unregistered prescription medication/s. Health data included information related to adverse effects, concomitant medications and Australian regulatory restrictions or exclusions for the use of MC. Restrictions on driving, use of heavy machinery and certain vocational activities were adhered to as per current Australian regulations. Current ‘zero tolerance’ drug driving legal frameworks in Australia criminalise the presence of THC in bodily fluids irrespective of impairment. Emerald Clinical Network is independent from and has no affiliation with MC producers. All clinicians were independent contractors who choose individually for whom, when and what MC product to prescribe for referred patients. Patients who were eligible were prescribed oral MC products, which were dispensed at independent pharmacies. MC products in Australia are grouped by the TGA into five categories based on the proportion of CBD content (or THC, in the case of THC-dominant and THC only categories) compared with the total cannabinoid content rather than total milligrams (mg) per volume. Table 1 provides an overview of the different MC categories by percentage. Note that throughout our analysis, we convert the total dose values to daily oral dosages of THC and CBD in mg/day to facilitate readability. Patients underwent a two-week carefully monitored deliberate dose titration and were monitored at least every 8-weeks for up to 12 months and then 12-weekly. Treatment Related Adverse Events (TRAEs) were collected at each subsequent prescription visit. This continuously accumulating registry has approximately 120–150 new enrolments every month (or approximately 1800 new enrolments per year). The RAND Short Form Health Survey (SF-36) is a patient self-report 36 item quality of life questionnaire [ 15 ] used for the routine monitoring and assessment of well-being and care outcomes. Questions include items related to physical functioning, bodily pain, role limitations due to physical health, personal or emotional problems, emotional well-being, social functioning, energy/fatigue and general health perceptions. The Brief Pain Inventory–Short Form (BPI-SF) is a validated self-report participant questionnaire [ 16 ] which assesses the severity of pain and its impact on daily functions. Assessment areas include severity of pain, impact of pain on daily function, location of pain, pain medications and amount of pain relief in the past 24 hours or the past week. The BPI-SF assesses pain scores by Numeric Rating Scale, with responses ranging from 0–10, with 0 = no pain, to 10 = pain as bad as you can imagine. The Depression Anxiety and Stress Scale-21 (DASS-21) is a validated symptom scale designed to measure the state of depression, anxiety and stress [ 17 ]. The DASS-21 asks patients to rate 21 statements from a 4-point score of 0–3 according to the following: 0 = it did not apply to me at all–“Never”, 1 = Applied to me to some degree, or some of the time–“Sometimes”, 2 = Applied to me to a considerable degree, or a good part of time–“Often”, and 3 = Applied to me very much, or most of the time–“Almost always”. The Insomnia Severity Index (ISI) is a validated patient self-report 7-item questionnaire that assesses the nature, severity and impact of insomnia [ 18 ]. The dimensions evaluated include severity of sleep onset, sleep maintenance, early morning awakening, sleep dissatisfaction, interference of sleep, difficulties with daytime functioning, noticeability of sleep problems by others and distress caused by the sleep difficulties. The ISI uses a 5-point Likert scale of 0–4 to rate each item (0 = no problem; 4 = very severe problem), yielding a total score ranging from 0 to 28. The total score is interpreted as follows: absence of insomnia (0–7); sub-threshold insomnia (8–14); moderate insomnia (15–21); and severe insomnia (22–28). The Clinical Global Impression Scale (CGI) was developed to provide a clinician’s global assessment of the patient’s functioning before and after commencement of medication [ 19 ]. Using a 7-point scale to assess the patient change since initiation of treatment (1 = very much improved 2 = much improved; 3 = minimally improved; 4 = no change; 5 = minimally worse; 6 = much worse; 7 = very much worse) This takes into account the experienced clinician’s knowledge of the patient’s history, circumstances, symptoms, behaviour, and function as well as the 16-point Efficacy Index which considers both medication efficacy and safety. In addition to the validated questionnaires, clinicians also collected information on all adverse events reported by patients at each visit. Adverse event terminology was described by the patient and coded retrospectively using Medical Dictionary for Regulatory Activities (MedDRA) terminology [ 20 ]. All adverse events were graded as mild (not causing discomfort, no intervention or change to prescribed dose/product required), moderate (can cause discomfort, intervention and/or change to prescribed dose/product required), severe (causing severe discomfort, cessation of prescribed product required) or serious (medical emergency, life threatening, disabling, causing hospitalisation or death). Clinicians used medical judgement to determine if the adverse event was likely, possibly, unlikely or not at all related to the prescribed medicinal cannabis product. All adverse events classified as “likely” or “possibly” related to the prescribed product are considered to be Treatment Related Adverse Events (TRAEs). Data were extracted from the AECeR registry in April 2022. Descriptive statistics presenting the respective numbers and proportions of patients–as well as means and standard deviations (SD), where applicable–were used to describe the demographic, clinical, medication use, and adverse events of patients included in the registry. To assess the patterns over time regarding patient and clinician reported outcomes, each outcome measure was plotted at baseline (before going on treatment) as well as select follow-up windows–that is, 3, 6, 12, 18, and 24 months after commencing MC. As this is continuous RWD registry data there are fewer participants at the end time points than the beginning time points. Retention rates at those intervals were 100, 98, 68, 45, and 35 percent, respectively. In addition, and to test whether the observed differences when undergoing this health intervention are not only quantitatively meaningful but also statistically important on conventional levels of statistical significance, independent t-tests were performed between baseline and the corresponding follow-up scores. Note that statistical significance was tested at the p = 0.05 level throughout. Finally, all analyses were performed using R 4.2.1.

(ZIP) Click here for additional data file.