Leanne Tite takes you through a randomised double blind placebo controlled crossover trial

Abstract

Objective—To evaluate the effect of the oral synthetic δ-9-tetrahydrocannabinol dronabinol on central neuropathic pain in patients with multiple sclerosis.

Design—Randomised double blind placebo controlled crossover trial.Setting—Outpatient clinic, University Hospital of Aarhus, Denmark.

Participants—24 patients aged between 23 and 55 years with multiple sclerosis and central pain.

Intervention—Orally administered dronabinol at a maximum dose of 10 mg daily or corresponding placebo for three weeks (15-21 days), separated by a three week washout period.

Main outcome measure—Median spontaneous pain intensity (numerical rating scale) in the last week of treatment.

Results—Median spontaneous pain intensity was significantly lower during dronabinol treatment than during placebo treatment (4.0 (25th to 75th centiles 2.3 to 6.0) v 5.0 (4.0 to 6.4), P=0.02), and median pain relief score (numerical rating scale) was higher (3.0 (0 to 6.7) v >0 (0 to 2.3), P=0.035). The number needed to treat for 50% pain relief was 3.5 (95% confidence interval 1.9 to 24.8). On the SF-36 quality of life scale, the two items bodily pain and mental health indicated benefits from active treatment compared with placebo. The number of patients with adverse events was higher during active treatment, especially in the first week of treatment. The functional ability of the multiple sclerosis patients did not change.

Conclusions—Dronabinol has a modest but clinically relevant analgesic effect on central pain in patients with multiple sclerosis. Adverse events, including dizziness, were more frequent with dronabinol than with placebo during the first week of treatment.

This month's paper is Svendsen K B, Jensen T S, Bach F W. Does the cannabinoid dronabinol reduce central pain in multiple sclerosis? Randomised double blind placebo controlled crossover trial. BMJ 2004;329:253. You can read it by going to studentbmj.com and clicking on the link.

Why do the study?

Multiple sclerosis is a chronic progressive disorder in which the lining of nerve fibres (myelin) in the central nervous system is attacked by the body's immune system. The resulting scar tissue and absence of myelin eventually leads to neuronal dysfunction in sufferers causing a range of symptoms. Damage to nerves in the central nervous system can also lead to sensations of pain in different parts of the body. In fact acute or chronic pain can occur in 30-80% of people with multiple sclerosis. Some recent animal studies have shown that cannabinoids (chemical compounds present in marijuana and cannabis) can reduce sensitivity to pain, including pain caused by neuronal dysfunction. These studies have led to further research providing some evidence that cannabinoids might have a similar effect in humans. So far research with humans has been sparse, however, and has focused on pain caused by other disorders or more generalsymptoms of multiple sclerosis. So the aim of the current study was to show that cannabinoids are effective in reducing pain in people with multiple sclerosis specifically caused by damage to nerves in the central nervous system.

And how can any study possibly need to be so complicated? It might appear at first as though the authors are trying to sound a bit clever, but in fact each of these words describes a useful control that has been incorporated into the study design to make the study's findings more trustworthy. When you break it all into smaller pieces and look at each in turn, it all starts to look pleasingly straightforward. So where do we start? As with any research, a good place is always with the objectives of the study itself. The investigators' aim was to show that dronabinol (the cannabinoid in question) has pain alleviating properties for people with multiple sclerosis. Essentially, they were attempting to demonstrate a cause-effect relation between the drug and the patients' pain. To do this, the researchers needed to show that both parts of the cause-effect relation are true: not only that pain is reduced following consumption of dronabinol (the effect—this is the easy bit to show) but also that the effect is, beyond doubt, due only to the dronabinol (the cause—this is the more difficult part). To do this, the investigators must rule out any other possible explanation for the effect of pain reduction after consumption of the dronabinol. Of course, in any experiment you can't literally rule out all possible explanations for an event that you have observed (that would be impossible) so instead you do it logically: by comparing two groups of people with multiple sclerosis, whose treatment is identical in every way except for the presence of dronabinol. The group that doesn't get dronabinol is the control group, and they are instead given a placebo, a pill that is identical to the dronabinol pill except that it does not contain dronabinol. Then because the study groups are identical but for one factor (whether or not they were given dronabinol) any difference in pain experience must logically be due to the dronabinol.

But groups of humans are never identical, so how can you make sure your experimental and control groups are truly identical? Well of course you can't, so instead you randomly allocate patients to experimental versus control groups so that any differences in the patients between the different groups are also randomly distributed and will not systematically affect the outcome of the study in either group. To make doubly sure that the effects of the dronabinol and the placebo were not influenced by existing “individual differences” between the patients in each group, a crossover method was used, which means that patients swapped treatments partway through the study (from dronabinal or placebo), so that all patients were at some point both control and experimental participants. This means that the effects of each treatment could be compared in the same patient. A final control in the study is the use of blinding. This means that those involved in the study are unaware of who is receiving what treatment when. In a double blind study, neither the researchers nor the patients know which treatment they are receiving at any one time otherwise both might subconsciously alter their behaviour in ways that “fit” with the experimental objectives. This is especially true when the outcome measures involve subjective data, as in the self reported data collected about patients' own assessment of their pain in this study.

The investigators recruited participants using a postal questionnaire and from the local outpatients multiple sclerosis clinic. Between these two methods, they hoped to contact everyone in the local area (Aarhus County, Denmark) who had been diagnosed with the condition. This is a good approach to sampling because if everyone contacted then takes part, the sample can be said to be representative of the population under study (that is, everyone with MS in the area of Aarhus, Denmark). Getting a representative sample is important if you want to be able to generalise from your study participants to the whole of the population of interest. However, in this study the sample was effectively self selecting because only those who chose to answer the postal questionnaire could be included in the study. This is a problem because it means that an element of bias has been introduced into the sample. For example, it might be that those most ill with multiple sclerosis were not able to respond to the questionnaire and become involved in the study, leaving the effect of the treatment on this group unknown. However, in any study involving humans there is always an element of self selection. The important point is that these kind of unavoidable drawbacks are acknowledged when evaluating the data later on.

Once all eligible patients had been recruited to the study, a baseline measure of their pain was taken in a one week period before any treatment was started. This is because all patients' level of pain will naturally vary so an initial measure is required to see how much each patient's pain was improved by treatment. You will also notice a washout period was allowed partway through thestudy. This is a period before the crossover when thesecond treatment (that is, placebo or dronabinol) is started to make sure that any effects of the first treatment do not carry over and influence the effects of the second treatment. The investigators also double checked for carry over effects during statistical analysis by looking for unusual patterns in the data. During the study, patients recorded their pain twice every day in a diary using a 10 point rating scale. Interestingly, none of the 24 participants dropped out at any point during the study. This is an important point, especially for trials of treatments for an illness. Often, patients drop out of a trial if the treatment is not working, if it is difficult to stick to, or if the adverse effects are intolerable. Naturally, these patients' data are not then included in the study, and so only those for whom the treatment worked are leftin the study, making the treatment look better than it really is.

The data show that treatment with dronabinol reduced self reported pain by around 21% in all patients, suggesting that cannabinoids can indeed reduce central nervous pain in people with multiple sclerosis. When thinking about the effectiveness of treatments, the effect size—that is, how much of an effect the treatment has—is just as important as whether or not the treatment works. Here, the reduction in pain of 21% is about the same as that of other drug treatments for multiple sclerosis, meaning that dronabinol might be a useful alternative for patients who do not respond to other treatments. The adverse effects of a treatment are also important to consider when looking at the usefulness of a drug. In this study some adverse effects were reported, but none were so problematic as to prevent any patient from carrying on with the treatment.

The objective of this study, to show that dronabinol has pain alleviating properties for people with multiple sclerosis, is not ground breaking research—there is already a growing body of evidence indicating that cannabinoids have such properties for humans. However, not all research has to be ground breaking to be good. A well designed and well controlled study that moves a treatment from “possibly effective” to “probably effective” is just as valuable as a study that discovers a new idea. And it is the thoroughness of the design of this study that makes it good. One drawback though is the small sample size involved (just 24 participants): larger samples reduce the impact of random events on data and so make the findings more accurate and more trustworthy. However, the use of lots of different controls in the study helps to overcome this problem. Of course, there are still lots of unanswered questions about the use of dronabinol for treating pain in MS. For example, what is the optimum dosage? What are the long term effects, both adverse and medicinal? Are there certain groups of people who cannot use the drug or for whom it is not effective? All these questions can be answered in further research, in larger clinical trials which might bring the reality of dronabinol treatment to more people with multiple sclerosis, a reality that has been brought one step closer thanks to this well designed study.

Guidelines: We intend publishing within 24 hours all responses that contribute substantially to the topic under discussion. To avoid points that have already been made please read other people's responses before posting your own. All responses will be eligible for publication in the paper studentBMJ, providing that current appointment and place of work are given. Name and email address are required to send a response and will be published with the response.

Please note: When writing a respnse to a response, please enter the original title of the article and the author of the response you are responding to.