For many experts, it was a victory for evidence-based medicine when the 2019 National Institute for Health and Care Excellence (NICE) guidelines failed to recommend cannabis for treating pain. But, for cannabis advocates, it robbed hundreds of thousands of people of a potentially life-changing therapy.
This stand-off continues today and at its core lies the question of what type of evidence counts when it comes to making clinical decisions.
Advocates say that evidence of humans consuming cannabis for therapeutic purposes for millenia and the millions of people currently taking it for pain relief — both legally and illegally — are testament to its value.
“Cannabis isn’t a wonder drug. There are people on the internet that say it cures everything known to man — of course it doesn’t,” says Mike Barnes, a former neurologist who founded and is now president of the UK’s Medical Cannabis Clinicians Society.
“It’s just another medicine. It’s not suitable for everybody. It doesn’t help everybody. But it is very, very useful.”
Barnes, who also founded and now chairs the Cannabis Industry Council, believes that when NICE and other major medical societies advised against using cannabis as an analgesic, they “ignored the real-world evidence — of which there is a great deal — that demonstrates that cannabis is very useful for pain”.
However, many dedicated pain researchers have a different view. They do not dismiss cannabis entirely, but what troubles them is that, when it comes to using it for pain relief, there is simply no evidence of the type that modern medicine routinely requires to approve a treatment.
Such data come predominantly from well-conducted randomised controlled trials (RCTs) — evidence-based medicine’s gold standard. But in the case of cannabis, Andrew Rice, president elect of the International Association for the Study of Pain (IASP) and a pain clinician and researcher at Imperial College London, says: “The trials that have been done don’t show an analgesic effect for pain.”
The few that have are of too low a quality to take seriously, Rice says, which points to a wider problem: good quality trials — RCTs especially — are few and far between.
In 2018, the IASP — concerned by snowballing medical cannabis use across the world — commissioned a major review of the literature, examining RCTs, systematic reviews, preclinical science and safety.
One caveat that emerged was the massive variety of cannabis-related products tested as putative analgesics. ‘Cannabis-based medicine’ (CBM) is an umbrella term covering everything from unprocessed plant material, resins and oils — all containing vast mixtures of bioactive compounds, which vary from strain to strain — through to preparations of single cannabinoids (mainly THC and CBD), onto more tightly controlled pharmaceutical products containing different but known quantities of cannabinoids. Consequently, few studies looked at exactly the same intervention.
Emma Fisher, a psychologist at the University of Bath, led the analysis of RCTs — the conclusion of which did not so much focus on study outcomes but on methods. Using widely deployed assessment criteria, all 36 identified RCTs were judged to provide either low- or very low-quality evidence.
The issues were multiple, Fisher says, including small sample sizes, problems with randomisation or blinding, endpoints that were poorly defined and not preregistered, refusals to share data and high risks of biased reporting. Her review did not conclude that cannabis is ineffective for pain, but rather that we do not know if it is.
“We can’t have much confidence in the effects that are found, which goes both ways,” says Fisher. “When we’re seeing that there is an effect, we’re not very confident. And where there isn’t an effect, we’re also not confident.”
Positive RCT data are precisely what regulators ordinarily demand before greenlighting a medicine. But UK and US regulators have never approved cannabis for pain.
In the UK, CBM was legalised by the Home Office in 2018. In the United States, 38 state governments have similarly legalised medicinal cannabis, but the Food and Drug Agency has never approved its use for pain[4,5].
“This is a really odd situation,” says Rice. “It’s the first time — we’re aware of — that what is effectively a medicine has bypassed conventional regulatory systems. And we can’t find any way that this has been justified.”
It is a situation that creates challenges for frontline doctors.
Alan Fayaz, a pain specialist at University College London Hospitals NHS Foundation Trust, runs NHS clinics for people whose pain has resisted all previous treatments. He also runs equivalent clinics privately — where, occasionally, he prescribes cannabis oils (administered sublingually) which, he says, sometimes, help his patients immensely.
Working amidst a swirling mass of differing opinions, Fayaz feels that two polarised groups have reached something of an impasse. Speaking on behalf of the British Pain Society (BPS), he says: “I think we would want more compelling evidence. We are at loggerheads because there is a paucity of data.”
Opening the door
The Home Office’s decision to legalise CBM came in response to high-profile campaigns to allow 6-year-old Alfie Dingley and 12-year-old Billy Caldwell access to it for their intractable epilepsy. Such a condition is mercifully very rare, but between a third and half of UK adults suffer from some form of chronic pain, and therefore the promotion of CBM as an analgesic made pain a focal point in terms of how widespread its use might become.
Rice says that, initially, it looked like such use would be supported. But then major and specialist medical bodies — including the Royal College of Physicians, the BPS and, following its review, the IASP — issued cautious position statements, saying the evidence base for treating pain with CBM was weak and more research was needed[6–8].
Most consequential was NICE’s 2019 recommendation that CBM not be made available on the NHS for pain — making this a matter for private medicine.
However, following the legalisation of CBM, a coalition of doctors, commercial entities and patient advocates blossomed, and these stakeholders are now well networked. The Medical Cannabis Clinicians Society that Barnes founded is an organisation supporting and educating around 400 clinician members interested in CBM. And the Cannabis Industry Council represents the UK cannabis business, of which members include companies selling or growing cannabis, as well as patient and legal groups.
Barnes is also chief medical officer of Drug Science, a charity founded in 2010 to inform and campaign for drug law reform. Drug Science currently runs a project called T21, which collates evidence from people’s experiences of CBM. According to estimates from T21, 23 private clinics currently provide around 25,000 people with cannabis — with approximately half of those taking it for chronic pain (see Figure).
This number is tiny compared with the roughly 5.3 million Americans legally accessing medical marijuana in 2021. It is also, as Barnes stresses — citing a 2019 YouGov survey sponsored by the Centre for Medicinal Cannabis — a very small fraction of the estimated 1.4 million people in the UK who use street cannabis to self-medicate. The most common reason given for cannabis use in the survey was pain relief.
Barnes says the private sector is growing and that, whereas CBM was initially only available from private clinics whose sole function was to prescribe these drugs, some private pain clinics now routinely use them.
Looking forward, Barnes says: “The numbers will go up once people are going back to their GPs, going back to their consultants, and most of them will say, ‘Look, this is helping me.’ I think that will eventually trickle into the system.”
“I don’t think we’re looking for a big bang,” he says, “I think we’re looking for slow drip, drip, drip.”
That said, a reversal of NICE’s recommendation could constitute a big bang and Drug Science’s T21 project states that its explicit goals are increasing access to CBM and getting the NHS to fund it — with pain sitting at the top of its list of relevant conditions.
Trials and tribulations
Barnes says the T21 project’s ongoing observational trial has now recruited about 4,000 patients. Initial data from 75 participants, published in 2021, indicated a significant self-reported improvement in health from baseline in participants with a range of conditions.
The trouble is, Barnes can already see an upcoming issue with the trial, saying, “some would dismiss it because it’s observational”.
Indeed, Rice is immediately sceptical, viewing such studies as a poor substitute for RCTs. “Real-world data is very easy to get,” he says, “But then when you go and scrutinise it: what do you mean by real-world data? What are you collecting? What are your controls?”
Among issues confounding observational studies is the fact that they typically involve a self-selecting group of patients, who have actively sought out cannabis prescriptions from doctors who strongly support the use of these drugs. These patients are also paying hundreds of pounds per month to take the drug. And it is robustly established that, in clinical trials of potential analgesic drugs, the placebo effect is strikingly large[12,13].
Rice is pessimistic that the high-quality trials he and others desperately want to see will ever emerge. Because medicinal cannabis has been so widely legalised without having to pass through standard regulatory procedures, he fears there is simply no incentive to spend the “somewhere between US$10m and US$40m per trial” that it costs to do the type of large-scale RCT normally required to convince regulators.
And while Barnes argues that the distinctive aroma of cannabis makes blinded trials impossible, this is unlikely to convince researchers that there is no need for better randomised trials.
Besides this specific issue, Fayaz describes two general challenges with analgesia trials. One is having to rely on subjective self-reports rather than a hard, physically-measured endpoint and the other is the hugely variable causes of pain. “As a result,” Fayaz says, “the evidence base for most treatments of pain is very poor.”
These points are roundly acknowledged by the pain research community. But this community is also highly sensitive to the costs of bypassing rigorous testing.
“The lesson from history,” says Andrew Moore, a pain and evidence-based medicine researcher, who recently retired from the University of Oxford, “from opioids and a pile of other therapies — is where there is no evidence, eventually something bad happens.”
Barnes calls comparisons with the opioid crisis “complete nonsense”.
“I don’t think that the evidence base is weak; I think it’s very robust,” he says, adding: “You do not die from a cannabis overdose.”
But while the risk of fatal overdoses is essentially zero, researchers who want cannabis treated like any other medicine still insist that the risks it does pose — be they cannabis use disorder or elevated psychosis risk — must be formally assessed and weighed against the benefits it offers.
When the IASP reviewed studies of CBM’s potential harms, it again lamented the lack of high-quality data. Noting that the risk profiles of different CBMs — from street cannabis to carefully controlled pharmaceutical products — will vary, it called for “a need for caution and more robust harms evaluation in future studies”.
But another parallel with opioids is equally troubling to pain researchers. And that is commercial involvement.
Moore led the IASP’s analysis of systematic reviews. And, like medicinal cannabis RCTs, when these were assessed using standard quality metrics, “almost all of them are critically low”, he says.
“The interesting thing about this particular area,” Moore says, “is that there are more systematic reviews than trials. And there are other things as well. The more critically low quality you seem to go; the more people write ‘There’s great evidence that cannabinoids are good for pain.’ They say these drugs are effective and safe based on crap or non-existent data. We call it ‘spin’. There’s an awful lot of spin out there.”
Moore is concerned that many reviews are written by companies and their associates. “There’s an awful lot of industry money coming in… and there’s a lot of interest in having something to sell,” he says.
Perhaps, ironically, one argument used to support the use of medicinal cannabis for pain, is that it may help people currently taking prescription opioids to reduce their opioid use. However, Deepika Slawek, a pain specialist and addiction researcher at Albert Einstein College of Medicine in New York, says the data on this are “really conflicting”.
An early observation that opioid-related deaths fell in states that had legalised medical cannabis did not hold up in a later study[16,17]. “Maybe that’s because ecological studies are not the best way of looking at this, and we really need RCTs,” Slawek points out. However, a recent, retrospective, observational study showed a reduction in opioid prescriptions written for people using CBM in New York state.
“I think it’s still an open question,” says Slawek, who is now commencing two prospective observational trials to further examine the relationship[19,20].
On the front line
Ultimately, the IASP’s review of cannabis, cannabinoids and pain treatment led to it publishing suggestions for how to address knowledge gaps in this field.
Its recommendations to include measures of quality of life and metrics related to living with pain in trials resonate with Fayaz. “Is it appropriate for us just to be looking at the intensity of pain and not looking at the impact of treatments on function, wellbeing, sleep and mood?” he asks. “If cannabis can have an effect on those things, maybe there is a broader role for it.”
He is clear that, at the dosages he uses, patients are not stoned. But he says: “Anecdotally speaking, what I hear isn’t always that the pain is going away. The language is more about ‘I’m less bothered by it’; ‘I can do more’.”
What Fayaz wants most from research is guidance as to when he might most profitably explore CBM. This, he thinks, might emerge from attending more to individual cases and categories. “Let’s narrow it down and see if it’s the pain condition that dictates that the cannabis works or is it the patient characteristics?” he says. “Try and work that out. I think that is the future.”
In New York, Slawek estimates around 70% of patients seeking medicinal cannabis at the Montefiore Hospital in the Bronx — where she practices — want it for refractory pain. First- and second-line treatments have not worked, and doctors will no longer prescribe them opioids.
“The thing that’s frustrating from a clinical and patient perspective,” she says, “is that we don’t have a lot of options for managing those patients. So, many patients will even try unregulated cannabis before they get to me for the management of their pain.”
Slawek too has seen cannabis benefit her patients. But, like Fayaz, she wants a better evidence base to guide her. Drugs that have passed through regular approval routes come with robust data-driven guidelines, indicating when and for whom they are useful. With cannabis, she says, “we’re working backwards”.
Faced with a complex mix of patients using both CBM she has prescribed as well as street cannabis, she says her job is not to put her fingers in her ears and pretend this is not happening. “We can listen to our patients to hear what they’re telling us about what symptoms they’re experiencing, what relief they’re experiencing from cannabis,” she says. “Then test it in an actual scientific manner to get real answers about the benefits and the risks.”
What Slawek says of New York is likely true of everywhere: “Cannabis is out there — and it’s not going to go away.”
- 1Cannabis-based medicinal products. National Institute for Health and Clinical Excellence. 2019.https://www.nice.org.uk/guidance/ng144/chapter/Recommendations (accessed 3 May 2023).
- 2Fisher E, Moore RA, Fogarty AE, et al. Cannabinoids, cannabis, and cannabis-based medicine for pain management. Pain. 2020;Publish Ahead of Print. doi:10.1097/j.pain.0000000000001929
- 3Rescheduling of cannabis-based products for medicinal use in humans (accessible version). Gov.uk. 2018.https://www.gov.uk/government/publications/circular-0182018-rescheduling-of-cannabis-based-products-for-medicinal-use-in-humans/rescheduling-of-cannabis-based-products-for-medicinal-use-in-humans-accessible-version (accessed 3 May 2023).
- 4State-by-State Medical Marijuana Laws. ProCon.org. 2022.https://medicalmarijuana.procon.org/legal-medical-marijuana-states-and-dc/ (accessed 3 May 2023).
- 5FDA and Cannabis: Research and Drug Approval Process. US Food and Drug Administration. 2023.https://www.fda.gov/news-events/public-health-focus/fda-and-cannabis-research-and-drug-approval-process (accessed 3 May 2023).
- 6Recommendations on cannabis-based products for medicinal use. Royal College of Physicians. 2018.https://www.rcplondon.ac.uk/projects/outputs/recommendations-cannabis-based-products-medicinal-use (accessed 3 May 2023).
- 7BPS Position Statement on the medicinal use of cannabinoids in pain management. The British Pain Society. 2018.https://www.britishpainsociety.org/static/uploads/resources/files/BPS_Position_Statement_on_the_medicinal_use_of_cannabinoids_in_pain_management.pdf (accessed 3 May 2023).
- 8IASP Position Statement on the Use of Cannabinoids to Treat Pain. International Association for the Study of Pain. 2021.https://www.iasp-pain.org/publications/iasp-news/iasp-position-statement-on-the-use-of-cannabinoids-to-treat-pain/ (accessed 3 May 2023).
- 9Number of legal medical marijuana patients in the U.S. as of 2021, by state. Statista. https://www.statista.com/statistics/585154/us-legal-medical-marijuana-patients-state/ (accessed 3 May 2023).
- 10Couch D. LEFT BEHIND: THE SCALE OF ILLEGAL CANNABIS USE FOR MEDICINAL INTENT IN THE UK. Centre for Medicinal Cannabis. 2020.https://thecmcuk.org/wp-content/uploads/2022/02/Left-Behind012020.pdf (accessed 3 May 2023).
- 11Sakal C, Lynskey M, Schlag AK, et al. Developing a real-world evidence base for prescribed cannabis in the United Kingdom: preliminary findings from Project Twenty21. Psychopharmacology. 2021;239:1147–55. doi:10.1007/s00213-021-05855-2
- 12Colloca L. The Placebo Effect in Pain Therapies. Annu. Rev. Pharmacol. Toxicol. 2019;59:191–211. doi:10.1146/annurev-pharmtox-010818-021542
- 13Turner J, Deyo R, Loeser J, et al. The importance of placebo effects in pain treatment and research. JAMA 1994;271:1609–14.https://www.ncbi.nlm.nih.gov/pubmed/7880221
- 14Mohiuddin M, Blyth FM, Degenhardt L, et al. General risks of harm with cannabinoids, cannabis, and cannabis-based medicine possibly relevant to patients receiving these for pain management. Pain. 2020;Publish Ahead of Print. doi:10.1097/j.pain.0000000000002000
- 15Moore RA, Fisher E, Finn DP, et al. Cannabinoids, cannabis, and cannabis-based medicines for pain management. Pain. 2020;Publish Ahead of Print. doi:10.1097/j.pain.0000000000001941
- 16Bachhuber MA, Saloner B, Cunningham CO, et al. Medical Cannabis Laws and Opioid Analgesic Overdose Mortality in the United States, 1999-2010. JAMA Intern Med. 2014;174:1668. doi:10.1001/jamainternmed.2014.4005
- 17Wendelboe AM, Mathew R, Chongsuwat T, et al. Is There Less Opioid Abuse in States Where Marijuana Has Been Decriminalized, Either for Medicinal or Recreational Use? A Clin-IQ. Journal of Patient-Centered Research and Reviews. 2019;6:267–73. doi:10.17294/2330-0698.1704
- 18Nguyen T, Li Y, Greene D, et al. Changes in Prescribed Opioid Dosages Among Patients Receiving Medical Cannabis for Chronic Pain, New York State, 2017-2019. JAMA Netw Open. 2023;6:e2254573. doi:10.1001/jamanetworkopen.2022.54573
- 19Albert Einstein College of Medicine. Does Medical Cannabis Reduce Opioid Use in Adults With Pain (ReLeaf-E). ClinicalTrials.gov. 2020.https://clinicaltrials.gov/ct2/show/NCT04308148?term=Arnsten%2C+Julia&draw=2&rank=2 (accessed 3 May 2023).
- 20Albert Einstein College of Medicine. MEMO-Medical Marijuana and Opioids Study. ClinicalTrials.gov. 2017.https://clinicaltrials.gov/ct2/show/NCT03268551?term=MEMO&recrs=d&age=1&draw=2 (accessed 3 May 2023).
- 21Haroutounian S, Arendt-Nielsen L, Belton J, et al. IASP Presidential Task Force on Cannabis and Cannabinoid Analgesia. Pain. 2021;Publish Ahead of Print. doi:10.1097/j.pain.0000000000002266