Over recent years, there has been great interest amongst both health professionals and lay people in the use of cannabis for a range of indications. This has been supported by multiple reports of benefit, especially via social media and anecdotal reports from the community. This has led to the legalisation of cannabis for medical use in many countries despite a lack of evidence of benefit in most cases.

The most common reasons given by cancer patients for the non-recreational use of cannabinoids are sleep quality, pain, mood changes, lack of appetite and digestive problems. ¹

The indications for which cannabis has been approved for medicinal use in Australia include childhood epilepsy in two rare neurological conditions, muscle spasms from multiple sclerosis (MS), chemotherapy-induced nausea and vomiting (CINV), chronic non-cancer pain and palliative care. ² The laws dictating medicinal cannabis use in the United States vary from state to state but common indications in addition to the above include post-traumatic stress disorder, inflammatory bowel disease, glaucoma and symptoms related to HIV/AIDS. The regulatory systems in place across a range of countries including Israel and Canada have been discussed in relation to the UK experience. ³

The challenge for researchers has been to obtain evidence to support use in a medical context and to define therapeutic indications.

Over the last decade, much research has been undertaken and a multitude of systematic reviews have been published. A consistent pattern has emerged. Those reviews that include uncontrolled studies, case series and observational studies report significant benefits in a range of conditions, including pain, nausea, sleep, appetite, anxiety.^4,5 In the context of the worldwide opioid epidemic, it has also been suggested that medicinal cannabis is an effective opioid sparer. ⁶

Conversely, those reviews that include quality randomised controlled trials (RCTs) only, tend to report either no or minimal benefit in these areas. For example, a systematic review of RCTs involving cancer patients reported no effect on pain, sleep or opioid consumption. ⁷ Barakji et al., ⁸ in their systematic review and meta-analysis of placebo-controlled trials involving over 7000 patients with any type of pain (acute, chronic, cancer or non-cancer), showed that cannabinoids did not reduce acute or chronic pain nor improve quality of life. Chronic pain and sleep quality improved but with effect sizes below that considered to be clinically relevant. Nor was there evidence of benefit in the exploratory outcomes assessed, that is, morphine consumption, physical function or depressive symptoms. ⁸

The inability of many reviews to make firm conclusions is complicated by the fact that there are a wide range of medicinal cannabis products, used over a range of doses and routes with varying measures in different population groups. This heterogeneity has led to problems when attempting to undertake meta-analyses or show benefits. ⁹

These points are well illustrated when considering pain relief. Pain is one of the most common reasons given for use of cannabis, with multiple anecdotal reports of benefit, especially in those people suffering from chronic pain. As mentioned above, significant improvements in pain are reported in those reviews that include uncontrolled studies. ⁴ A recent Cochrane review however, looking specifically at cancer pain, reported with moderate certainty evidence that tetrahydrocannabinol (THC) containing products were ineffective in relieving moderate to severe opioid refractory cancer pain. ⁷ When trials of both cancer and non-cancer pain situations are included, Barakji et al. ⁸ report that cannabinoids can reduce chronic pain and improve sleep quality. However, the benefits were small, with effect sizes of questionable importance. Giossi et al., ¹⁰ reviewed the evidence of benefit in chronic primary pain, conditions such as fibromyalgia, irritable bowel syndrome and chronic migraine, that is, pain conditions that would be expected to have a degree of emotional overlay. Surprisingly, the authors were able to show very little evidence of benefit apart from long-term use in fibromyalgia. ¹⁰

Umbrella reviews compile evidence from multiple existing reviews and possibly provide the strongest evidence. They pool evidence from different sources of evidence and consider convergence of results from different study designs. ¹¹ Convincing or converging evidence from a recent review states that cannabidiol (CBD) ‘could be considered a potential beneficial treatment for epilepsy’ and that cannabis-based medicines could also be considered for chronic pain across different conditions such as MS, for nausea and vomiting in mixed conditions and sleep in cancer. These recommendations come with caveats, however, around risk-to-benefit ratios and comparative efficacy and safety with existing treatment options. The authors also postulate a role for cannabis in palliative care but give no justification for this. Similar guidance has been released by Ontario Health. ¹² The key messages here include a statement that ‘there is no evidence to recommend cannabis or cannabinoids as initial therapies for any health condition, including cancer-related symptoms’. They also state that consideration may be given to using cannabis when standard therapies have failed in people with cancer receiving palliative care, with no supporting evidence and despite emerging evidence to the contrary. ¹²

Most medical societies and special interest groups take a very conservative approach when discussing the use of cannabis for pain and agree that there is currently insufficient evidence to recommend cannabis as a standard treatment for pain.^12–14

The evidence of benefit for other symptoms is even more sparse. The Multinational Association of Supportive Care in Cancer guideline committee found insufficient evidence to recommend cannabinoids for cancer-related nausea, anorexia-cachexia or taste disturbance. ¹⁵ Similarly, they found no high-quality evidence to support the use of cannabis for those with psychological symptoms (anxiety, depression, insomnia). ¹⁶

Very few studies to date have been restricted specifically to those patients receiving palliative care. The most recent review of cannabis in palliative care included patients with a range of life-limiting diseases. Positive treatment effects were noted for a number of symptoms including pain, nausea, appetite, sleep and fatigue. Of the 52 studies included, 32 were non-randomised. As a consequence, the quality of evidence was low or very low and no meta-analysis was possible because of the range of cannabinoid products and the heterogeneity of study outcomes. ¹⁷ Most patients receiving palliative care have cancer. Recent guidelines from the American Society of Clinical Oncology state that the ability of cannabinoids to improve supportive care outcomes ‘remains uncertain’. ¹⁸

Rather than trying to demonstrate a benefit for individual symptoms, our group is attempting to show whether CBD or THC/CBD combinations can reduce symptom burden over and above that achieved by palliative care alone in patients with advanced cancer.^19,20 Rather than concentrating on individual symptoms, for which it has been very difficult to demonstrate benefit, we have chosen to take a more holistic view by using a total symptom score as our primary outcome measure. Synthetic CBD alone proved no better than placebo in this context. ¹⁹ The results of the subsequent trial using a one-to-one combination of THC and CBD in the same model are currently being analysed. ²⁰ As there remains no evidence-based guidance on the best combination of cannabinoids nor the ideal dose, our ongoing research is exploring these issues.

Many reviews point to the increased risk of adverse events associated with cannabis and that in many cases, the harmful effects outweigh potential benefits. ⁸ Although CBD appears to be relatively well tolerated, ¹⁹ THC has significant psychoactive effects. The most common adverse effects reported from the use of CBD/THC combination products include sedation, nausea/vomiting, dizziness, euphoria and gastrointestinal effects. ²¹ These have the potential to increase workplace risks, with decreased alertness and reaction times, reduced ability to safely operate machinery. The deleterious effect of cannabis on driving ability is well described. ²² Convincing or converging evidence supports avoidance of cannabis during adolescence and early adulthood, in people with mental health disorders, and in pregnancy. ¹¹ Smoking cannabis is associated with the same deleterious effects as tobacco. ²³ Increasing cannabis use disorder prevalence among patients with chronic pain is becoming a public health matter. ²⁴ Recently, concern has been raised regarding the deleterious effect of CBD on liver function. ²⁵

Despite the above, public opinion remains very positive and is supported by media reporting bias. This under reports the substantial placebo effect, ²⁶ the adverse effects, ²¹ potential deleterious effects in those with cancer ²⁷ and the potential for abuse, especially in young people. ²⁸ Few of the in vitro claims of benefit (e.g. anti-tumour and anti-inflammatory effects have been translated into clinical benefit).^29,30 This is consistent with the demonstrated publication bias towards positive outcomes. ²⁶

In summary, outside a few narrow indications, the place of medicinal cannabis as a ‘novel therapy’ for symptom management in palliative care remains unclear. ‘Medicinal cannabis’ can easily be sourced for a wide range of indications via cannabis clinics that are now commonplace in most cities. This is an expensive option for consumers, however, as very few cannabinoid products are registered and receive government subsidies. Most national agencies have only approved products for spasticity associated with MS, anorexia in patients with AIDS, CINV and CBD for resistant epilepsy in children with Dravet and Lennox-Gastaut syndromes. Some clinicians have reported being put under considerable pressure to prescribe cannabis by patients ³¹ and as described above, some reviews have recommended the use of cannabis in ‘palliative medicine’ (implying nothing to lose), but with no evidence to support this. ¹¹ It has been suggested that cannabis should (only) be used when ‘all else has failed’. The author’s view is that the treatments and interventions offered to patients receiving palliative care should be subject to the same level of scientific rigour as treatments offered to other patients. Therefore, with the present state of knowledge, cannabis should only be prescribed within a trial setting or in closely monitored observational settings.

Why has it been so hard to demonstrate a place for medicinal cannabis in palliative care? One problem may be that most studies to date have used synthetic or specific cannabinoids whereas the ‘entourage effect’ seems important. ³² That is ‘whole flower’ products containing a combination of different cannabinoids, plus plant terpenes and flavonoids. Whole plant products containing high dose THC, ingested by smoking or vapourising for rapid effect seem most popular. ³³ Similarly, standard outcome measures (e.g. numerical rating scales to measure pain) may not be appropriate for trials of cannabis. To this end, our group is trying to develop a ‘happiness scale’ that captures the feel-good factor that so many people report when taking cannabis. Should this endpoint prove more appropriate for trials for this type of medication, the true role of cannabinoids in palliative care may become clearer.