Letters

Cannabinoids in pain management

Study was bound to conclude that cannabinoids had limited efficacy

EDITORCampbell et al's paper on whether cannabinoids are effective and safe in the management of pain purports to be qualitative and systematic,¹ but it is neither. Because it focused on two clinically questionable synthetic cannabinoids and oral delta-9-tetrahydrocannabinol (THC) without providing any focus on the synergistic components of herbal cannabis, and examined only certain facets of the broad topic of pain, it ensured that a conclusion of limited efficacy was reached. That is not news.

What is surprising, in contrast, is that the authors chose to broaden the alleged impact of their limited investigation to relegate the use of cannabis and cannabinoids to a back seat in future analgesic applications. This contention is not supported by their limited data.

I see nothing published about pioneering British doctors and their clinical successes with cannabis extracts in a myriad of painful conditions between 1840 and 1940.^2-4 I see virtually nothing of modern scientific studies showing the multifactorial benefits of cannabis on a range of neurotransmitter systems, which I have reviewed.⁵ No mention is made of bureaucratic and political obstructions to clinical research into cannabis; one cannot show results when the requisite studies are not permitted. Thus until recently we have been left with an overwhelming (but ignored) body of anecdotal evidence from patients and their doctors.

What is truly newsworthy here is that the BMJ has ignored peer review and editorial standards in a scandalous manner. The popular media have seized the opportunity, and in the process valuable laboratory and clinical research, and their funding, in analgesia and pain control have been severely compromised. Great shame accrues to the journal as a result. Instead of probity we have propaganda.

Ethan Russo, clinical assistant professor, University of Washington School of Medicine. 
Montana Neurobehavioral Specialists, 900 North Orange Street, Missoula, MT 59802, USA erusso@blackfoot.net

Competing interests: Professor Russo has been a scientific adviser to GW Pharmaceuticals (a manufacturer of cannabis-based medicine extracts), which has reimbursed expenses for travel with regard to visits and clinical research. He is also the editor in chief of Journal of Cannabis Therapeutics.

Few well controlled trials of cannabis exist for systematic review

EDITORCampbell et al gave themselves an impossible task with their systematic review.¹ Anyone who has reviewed the scientific literature on the medical uses of cannabis rapidly finds that there is a dearth of well controlled clinical trials.² A meta-analysis of the use of cannabis in treating pain is therefore likely to find little of substance to comment on.

Unfortunately, this did not deter the authors from coming to a series of emphatic but ill founded conclusions. I hope that these will not be taken as the last word on the topic: large and well controlled clinical trials are about to start in the United Kingdom³ and a wealth of animal data support a role for cannabinoids in pain modulation.⁴

Leslie Iversen, visiting professor. 
Department of Pharmacology, University of Oxford, Oxford OX3 9DU les.iversen@pharm.ox.ac.uk

Competing interests: None declared.

Spasticity is not the same as pain

EDITORThe systematic review on cannabinoids in the treatment of pain¹ referenced a paper that I coauthored on the efficacy of a cannabinoid (delta-9-tetrahydrocannabinol (THC)) in spasticity.² My confusion in reading the review was the implication that that paper had anything to do with pain.

Spasticity and pain are distinctly different entities. Although pain may accompany symptoms of spasticity such as flexor or tonic spasms, the assessments of spasticity do not usually include the types of measure seen with analgesics. Our results have been confirmed by other researchers, and antispastic activity has been documented for marijuana,³ THC,⁴ and nabilone.⁵

More importantly, all studies that have been published have shown an antispastic effect for the cannabinoids. Currently, considerable research effort is under way to evaluate cannabinoids in multiple sclerosis. This effort is undermined when review articles are cited in the media as evidence that cannabinoids are either ineffective or unsafe.

Denis J Petro, neurologist. 
Arlington, VA 22209, USA djpmsmd@aol.com

Competing interests: None declared.

Cannabinoid receptor agonists will soon find their place in modern medicine

EDITORI am unsure whether the methods applied in the systematic reviews by Campbell et al and Tramèr et al are able to answer the questions of today's interest. ^{1 2} If you pool the data from older studies of pain you will miss most of the interesting information, particularly differences in efficacy for different painful conditions, differences between the cannabinoids, and interindividual differences with regard to side effects.

Cannabinoids are weak analgesics compared with the opiates.³ The question of interest is not so much whether they are potent analgesics compared with codeine but, rather, which painful conditions they are effective in. By stating that cannabinoids may have potential in neuropathic pains, particularly with spastic components, even Kalso hints at a need for such a differentiated assessment.⁴

The same is true for side effects. Levonantradol has not been brought on to the market, because it has a higher rate of side effects than tetrahydrocannabinol (THC). Today an interesting question might be, by which strategies could psychotropic side effects be reduced? Interindividual variation in side effects is high. Some patients may profit more because they tolerate relatively high doses without perceiving any unpleasant effects. Will we learn which patients have a favourable risk:benefit ratio?

With regard to antiemetic efficacy, I agree that modern serotonin receptor antagonists are effective to treat nausea and vomiting in cancer chemotherapy, but sometimes they fail and sometimes cannabinoids seem to be superior.⁵

The study by Maurer et al of 1990 cited by Campbell et al refers to another important aspectthe synergistic use of several pharmacological effects of cannabinoids, in this case the analgesic and antispastic effects in spinal cord injury. Results of research showing that cannabinoids reduce opioid induced emesis and act synergistically with opioids against pain point to a possible combination of analgesic and antiemetic effects of cannabinoids.

I believe that cannabinoid receptor agonists will find their place in modern medicine within the next few years. It will be interesting to see which indications they will be approved for and whether they will be limited to synthetic derivatives from drug companies engaged in cannabinoid research.

Franjo Grotenhermen, doctor. 
nova-Institut, D-50354 Hürth, Germany franjo.grotenhermen@nova-institut.de

Competing interests: The author is chairman of the International Association for Cannabis as Medicine, Cologne.

1.	Campbell FA, Tramèr MR, Carroll D, Reynolds DJM, Moore RA, McQuay HJ. Are cannabinoids an effective and safe treatment option in the management of pain? A qualitative systematic review. BMJ 2001; 323: 13-16[Abstract/Full Text]. (7 July.)
2.	Tramèr MR, Carroll D, Campbell FA, Reynolds DJM, Moore RA, McQuay HJ. Cannabinoids for control of chemotherapy induced nausea and vomiting: quantitative systematic review. BMJ 2001; 323: 16-21[Abstract/Full Text]. (7 July.)
3.	Grotenhermen F. Cannabis in der Schmerztherapie-ein neues Adjuvans? [Cannabis in pain therapya new adjuvant?] Forschung und Praxis, Wissenschaftsjournal der Ärztezeitung 1999; 276: 22-26. (In German.)
4.	Kalso E. Cannabinoids for pain and nausea. BMJ 2001; 323: 2-3[Full Text]. (7 July.)
5.	Gonzalez-Rosales F, Walsh D. Intractable nausea and vomiting due to gastrointestinal mucosal metastases relieved by tetrahydrocannabinol (dronabinol). J Pain Symptom Manage 1997; 14: 311-314[Medline].

Authors' reply

EDITORSystematic reviews tell us about research that has already been done. What we should learn from them is the research agenda for the future. For clinical trials of interventions we know that certain study architectures, particularly those that control selection bias and observer bias by randomisation and masking, go a long way to ensuring that those biases are minimised.

So why did we not include in our reviews, as Russo comments, "any focus on the synergistic components of herbal cannabis"? Because, despite including cannabis and marijuana in our search strategies, using various spellings, we found none. We did find information in the form of randomised controlled trials of tetrahydrocannabinol (THC) and synthetic cannabinoids. If there was good evidence on the efficacy and harm of herbal cannabis in the form of randomised controlled trials then we missed it. It is unlikely that such evidence exists.

If the BMJ ignored the peer review process it was not obvious to us. The route to publication was long and occasionally tortuous, with considerable argument with editors and peer reviewers. Like most authors, we believe that we could have been treated better, but the BMJ can be cleared of the slur that it shirked its responsibilities.

For acute pain, our attitude to any drug with dismal efficacy and a high rate of serious and potentially serious adverse events would be the same as it was for cannabis in these trials. We already have better drugs. For more difficult problems, such as painful spasms and neuropathic pain, none of us would want to overlook any possibility. For nausea and vomiting our attitude was similarly cautious. Where serious and common adverse events occur, this will limit the use of cannabinoids. Circumstances will sometimes dictate otherwise, as Grotenhermen points out.

We are surprised that these reviews were not done before fresh trials were funded. If they had been, the quality of the debate and of the decision making would have been higher. Large controlled studies are to be welcomed for some clinical problems, but their design should take into account the pitfalls of preceding trials lest the same mistakes are repeated. The ethics of starting a trial without doing a systematic review are questionable. The question that the trial seeks to solve is critical.

Previous trials do not answer important questions about relief of painful spasm. They do, however, suggest little future for existing cannabinoids in acute pain management and emesis.

Fiona A Campbell, consultant in anaesthetics and pain management. 
Pain Management Centre, Queen's Medical Centre, Nottingham NG7 2UH fiona.campbell@mail.qmcuk-tr.trent.nhs.co.uk

Andrew Moore, consultant biochemist. 
Henry J McQuay, professor of pain relief. 
Dawn Carroll, senior research nurse. 
Pain Research, Nuffield Department of Anaesthetics, The Churchill, Oxford Radcliffe Hospital, Oxford OX3 7LJ

Martin R Tramèr, staff anaesthetist. 
Division d'Anesthésiologie, Département APSIC, Hôpitaux Universitaires, CH-1211 Genève 14, Switzerland

D John M Reynolds, consultant clinical pharmacologist. 
Department of Clinical Pharmacology, Radcliffe Infirmary, Oxford OX2 6HE

Competing interests: None declared.