Drugs policy: medicinal cannabis Contents

3Current evidence base

24.The Chief Medical Officer’s (CMO’s) report reviewed the evidence on the therapeutic benefits of CBPMs. She found that there is evidence, though contested, for the use of CBPMs in the following conditions: chronic pain in adults; chemotherapy induced nausea and vomiting; multiple sclerosis spasticity syndromes; and intractable epilepsy. The report also found areas where there is limited or no evidence that CBPMs are effective.

25.The CMO’s report contains a number of caveats and nuances because of conflicting conclusions about the evidence. For example, the CMO’s review records that the US National Academies of Sciences, Engineering and Medicines do not think there is sufficient evidence for the use of CBPMs in intractable epilepsy, whereas the Australian and Irish studies did find sufficient evidence.23

26.Based on her report, the CMO recommended that cannabis-based medicinal products be moved to Schedule 2 of the Misuse of Drugs Regulations 2001, which allows research to be carried out into these products, and enables them to be prescribed by specialist doctors. The CMO’s report sets the background to the dilemma for those clinicians who are not confident to prescribe these products off-licence.

Research into CBPMs

27.A more appropriate summary of the of the evidence base of CBPMs within each discipline is available in the clinical guidance produced by the Royal College of Physicians, Association of British Neurologists and British Paediatric Neurology Association.24 The evidence has not led to more products being licensed as the evidence is not robust enough to demonstrate safety and efficacy for licencing purposes and companies have not used the existing evidence to make successful licensing applications.

28.For licensing purposes, companies who carry out the research present the findings of the research as evidence for the benefits and safety of the product, after which the medicines regulator will decide whether or not it is safe and efficacious enough to be licensed. GW Pharmaceuticals outlined what a licensed product ensures:

Approval from a medicines regulator ensures the safety, quality and efficacy of a medicine. This will provide prescribers with a robust evidence base (on which its approval is based) to inform clinical decisions and includes: safety data to protect human exposure; strong efficacy and safety data from clinical trials in the target patient populations; and a medicine of a consistently high quality which has been produced in an audited manufacturing plant, to reliable manufacturing and quality standards, with clear guidance on labelling and dosing.25

Licensing medicinal cannabis products would bring forward important information for clinicians to consider when prescribing.

29.Products that are licensed are also required to undergo assessment for cost-effectiveness by NICE if they are to be routinely prescribed by the NHS. If a product is assessed to be cost-effective, NICE will recommend it, after which these products will usually be available on NHS prescription. The research into and evidence for the efficacy of a product are important factors in NICE’s appraisal system.

30.One of the debates we heard throughout our inquiry was what type of evidence should be accepted to demonstrate the efficacy and safety of CBPMs. The CMO told us that “the only way we can get it licensed is through doing randomised controlled trials.”26 There is a clear hierarchy of evidence when it comes to licensing. It is accepted that randomised controlled trials are one of the best forms of evidence from a clinical trial and the “gold standard” in evidence.27

Figure 1: Pyramid of hierarchy of scientific evidence.28

C:\Users\ahmedmb\AppData\Local\Microsoft\Windows\INetCache\Content.MSO\4C29DFE8.tmp

SR: Systematic review

MA: Meta-analysis

31.A double-blind randomised controlled trial is a study in which a number of similar people are randomly assigned to groups to test a specific treatment. One group has the intervention being tested, while the other group takes a dummy treatment (placebo). The groups are assessed at specific times and any differences are recorded statistically. In such a trial, the researchers and the patients do not know who is on what treatment, therefore reducing bias. Randomised controlled trials are also important in understanding the place of a new treatment alongside existing treatments. An open-label trial is that where the researchers and the patients are aware of the treatment being given. Randomised controlled trials are necessary for licensing purposes as they are objective measures of safety and efficacy.29 While other methods of evidence in the graphic above can provide valuable information for clinicians making a judgement, these forms of evidence do not suffice for a licence.

32.Other witnesses took a different view and told us that they felt randomised controlled trials were not the most appropriate method of approaching CBPMs. Peter Carroll, Campaign Director at End Our Pain, told us:

We have to take a broader view of the evidence, because there is a point where multiple anecdotal stories build up to a pattern of evidence, and it seems absurd to me that we have to wait three, four or five years for trials to be produced when there are real-life cases now.30

Anecdotes or isolated stories of success do not amount to evidence that can be used to license a product. However, anecdotes may be useful in identifying areas where more thorough research could be carried out.31

33.However, we also heard that anecdotal and open-label trials do not provide a rigorous enough evidence base. Professor O’Callaghan told us:

The problem with open-label, non-randomised, non-blinded studies is that they almost invariably overestimate efficacy. That is why, when we are licensing medicines, we demand randomised controlled trials as the level of evidence we need for efficacy. There are severe biases that could be at play in open-label studies that could distort the results.32

34.Anecdotes do not allow researchers to assess the true impact of the products and their place compared to existing treatments. The most efficient way of ensuring that patients in the future have access to CBPMs is for them to be subject to research and to be licensed.

35.One of the explanations provided for the lack of randomised controlled trials in CBPMs is that it is inherently difficult to carry out such trials into medicinal cannabis products. Professor Mike Barnes, Chair of the Medicinal Cannabis Clinicians Society, said:

Cannabis is a plant and it does not lend itself very well to the standard pharmaceutical approach. It is not a single molecule that we can compare against a placebo. There are over 2,500 varieties of cannabis, each with a different structure of cannabinoids and terpenes, each with subtle differences. Which one would you pick for the standard pharmaceutical model?33

36.We heard that trials do not necessarily have to have a large number of patients for the trial to produce good quality evidence.34 We further heard from the CMO that if a drug performs very well in a trial, the trial does not have to continue. She said:

Modern, good trials have a blinded data management committee who regularly review the data, and if it is clear that it is harmful, or that it is an advantage, you break it and you stop there, and then you, of course, ethically provide the drug to the rest of the people who were in the trial and not taking it. I would think that was the right way to do the trial and the best way to get to the right answer.35

If medicinal cannabis products are trialled and found to be clearly efficacious, the trial can be concluded quickly. This should provide encouragement to those who believe they can demonstrate clear advantages of medicinal cannabis to conduct trials.

37.We also heard that the side effects of cannabis are generally well known as cannabis has been used for a sustained period. Professor Mike Barnes told us:

[ … ] cannabis is rather different in the sense that it has rather a lot of experience over thousands of years. It is not something that has come on to the street in the last few decades. In this country today, about 3 million people take cannabis, of which 1 million take it for medical purposes—1 million people. That is an accumulated great experience, so we know an awful lot about potential side-effects. If some awful side-effect was going to emerge, it probably would have emerged by now.36

38.Whilst illegal street cannabis may have been used for many years, medicinal grade cannabis-based products have not been used for a substantial period of time. Street cannabis is generally high in tetrahydrocannabinol (THC) and of inconsistent quality.37 Street cannabis high in THC is known to damage the developing brain.38 While street cannabis may have a long history, medicinal grade cannabis has a relatively short history of being used to target specific medical conditions and side effects should be evaluated in the context of medicinal use.

39.The history of medicine contains many examples of products and procedures assumed to be so beneficial that they should be introduced without rigorous research, only for them to be later withdrawn because they turned out to be harmful.

40.The CMO pointed to an example: there had been a widespread assumption that steroids would help reduce brain swelling (oedema) and reduce harm after head injury. Doctors regularly gave steroids to head trauma patients before a randomised controlled trial was eventually conducted “and the outcomes were worse if you give steroids.”39

41.There is ongoing controversy around the evidence base for CBPMs. One argument put forward during our inquiry was that there is plenty of research and evidence across the world which the UK should take account of. Professor Mike Barnes told us:

At the moment, there are 128 trials of cannabis ongoing worldwide. We should not forget that other jurisdictions—sensible jurisdictions, if I can use that word—Canada, Australia, Germany and other European countries, such as Denmark, have introduced cannabis legislation and allowed doctors to prescribe, and they are prescribing with more freedom than we are here. I think we are a little bit obsessed with UK-based evidence. We need to take into account global evidence.40

42.We welcome the large number of trials being undertaken. Whilst the CMO’s review and the professional guidance did take account of the existing international evidence, we endorse the call for further research to be prioritised. It is the quality of the trial and the evidence that is being presented by the company which influences licensing decisions rather than the country in which it is conducted. The Government should work with other countries to facilitate and encourage research that demonstrates the safety and efficacy of medicinal cannabis.

43.The British Paediatric Neurology Association told us that there is no randomised controlled trial data within childhood epilepsy that demonstrates that the addition of THC confers any added medical benefit and there is no adequate safety data concerning products that contain THC.41 The Association of British Neurologists said that there is good evidence for the use of cannabidiol (CBD) in two complex epilepsy syndromes - Dravet and Lennon-Gastaut syndrome.42 The Royal College of Physicians told us that there is also good evidence that CBPMs are effective in preventing chemotherapy-induced nausea and vomiting, but they have a high side effect profile and there are more efficacious agents available.43 In the treatment of chronic pain, Professor Andrew Goddard, President of the Royal College of Physicians, told us there is a “weak” suggestion of its efficacy. He told us:

You would need to treat 24 people in order to see benefit in one person. When it comes to the harms of those drugs, you only need to treat six people to see significant harms.44

44.The vast majority of evidence that we have received from clinicians told us that there is a weak evidence base for CBPMs in general.45 We cannot ignore the conclusions that clinicians have come to. If clinicians feel that the evidence base is not strong enough for them to prescribe the product, the focus must be on generating sufficient evidence.

45.A further issue with the current evidence base for CBPMs is that there is limited evidence on how they interact with other drugs. Professor Helen Cross told us that “even a small amount of cannabidiol may interact with liver-metabolised drugs.”46 Dr Imran Malik from the Royal College of Psychiatrists also explained:

For example, day in, day out, I see patients who have epilepsy but have mood disorders associated with that. When we are to prescribe whatever form of medicinal cannabis we get approval for, we do not have any drug interaction information available to us. In real life, they are on multiple drugs, so they would be on antidepressants, antipsychotics or a number of other physical health-related medications. As a clinician, when I stick my neck out to prescribe something that does not have a licence, I still do not know what the interactions are.47

It is explicit in the GMC’s guidance Good Medical Practice 2013 that in providing clinical care a doctor must “check that the care or treatment you provide for each patient is compatible with any other treatments the patient is receiving, including (where possible) self-prescribed over-the-counter medications”.48 Until data on how CBPMs interact with other drugs is available, it would contradict the GMC’s guidance if clinicians were encouraged to prescribe CBPMs without being confident about the way the prescribed products interact with other drugs.

Current evidence base: conclusions

46.It is clear there is not a firm evidence base for those CBPMs which were rescheduled in November 2018. It is critical that the Government and industry should work to further the evidence base in order to improve the confidence of doctors who can begin to prescribe CBPMs and also allow the products to be licensed. Specialist centres can play an important role as they are a valuable resource in bringing together expert clinicians and patients with complex conditions. We support the case for specialist centres to be able to lead on building the evidence base.

47.We are very mindful of the plight of children living with severe and intractable epilepsy, who are already on powerful drugs which may not be licensed in the age groups or for the specific conditions where they are being used. We heard about the specific case of Jorja Emerson, who is receiving medication that her father told us was not appropriate for her age at the time it was first prescribed.49 We are deeply sympathetic towards the struggle of patients and their families who see others being treated with CBPMs whilst not being able to obtain it themselves.

48.There needs to be a sense of urgency on the part of the Government, industry and clinicians in responding to children with severe and intractable epilepsy. We are aware of research proposals submitted and being prepared.50 These proposals need to be prioritised so that all children with these conditions can access clinical trials of CBPMs through specialist centres. Additional trials have a dual benefit of furthering the evidence base and allowing patients in the trials to access therapies which may treat their condition. We encourage paediatric neurologists and other specialists to help families access appropriate clinical trials. Patients and their families are remarkably well informed with regard to the publicly available information and published research in this subject area.51 Clinicians should take advantage of their knowledge and keenness to be involved in furthering the evidence base.

49.While we fully support the proposed Randomised Control Trials and calls for them to be started as a matter of urgency, we believe that other means of gathering evidence must also be investigated. RCTs can take up to 4 years to complete. Furthermore, the nature of RCTs is that some patients may be required to take a placebo, which, in some cases, could mean existing patients being taken off their current treatment with CBPMs which are giving them relief from their symptoms. It is unlikely that a patient or parent would be willing to accept this. Nor should they have to.

50.The parents and clinicians who supported them made an impassioned argument for observational trials to be conducted alongside RCTs. We call on the National Institute for Health Research to engage fully with these parents and clinicians to discuss their proposal and explore all ways to improve the evidence base.

51.It is important that clinicians and researchers retain the confidence of patients and their families. Faith in the role and importance of research will be undermined if clinicians are unwilling to participate and to make sure that children can take part in clinical trials. The rarity of some of the conditions that medicinal cannabis is used to treat makes it additionally important that all those who could take part are facilitated to do so. Clinicians and relevant bodies should endeavour to keep patients and their families informed of the importance of research trials in ensuring their safety, whilst making every effort to update patients about trials they can participate in.

Barriers to research

52.As mentioned previously, we welcome the Government’s decision to move medicinal cannabis from Schedule 1 to Schedule 2. This will help facilitate much needed research. Products in Schedule 1 are judged to have no therapeutic benefit and used mainly in research under a Home Office licence. Moving CBPMs to Schedule 2 allows the product to be more available for researchers and prescribers, but still under special requirements. We also welcome the National Institute of Health Research’s (NIHR) call for research proposals. The call for proposals was initiated alongside the rescheduling. Successful applications could be expected to start trials in Summer 2020.52

53.Whilst we welcome the call, we call on those involved to prioritise and aim to expedite this process for clinical trials into intractable childhood epilepsy. There needs to be a greater sense of urgency in the response to the plight of those children living with intractable epilepsy and to make sure that their families can all access clinical trials of CBPM. The evidence gathered by our inquiry highlighted a number of additional barriers facing research in this area that need to be recognised if research is to be carried out widely and quickly.

54.There are a number of conditions where research is needed to establish the place of CBPMs. These include, but are not limited to: epilepsy; chronic pain; chemotherapy induced nausea and vomiting; multiple sclerosis; fibromyalgia; Tourette syndrome; anxiety and posttraumatic stress disorder; and palliative care. The Chief Medical Officer’s review identifies these, along with other conditions where the value of medicinal cannabis has been explored.

55.One of the barriers to research is that industry is not always prepared to supply products for research trials to those clinicians or organisations proposing to carry out those trials. We heard throughout our inquiry that some pharmaceutical companies were not willing to provide their product for trial.53 Professor Helen Cross told us of her experience of being in discussion with one particular company where it has taken six months to persuade them to supply, and the company have been known to pull out of trials early.54

56.Professor Helen Cross suggested a possible reason why some companies do not submit their products to randomised controlled trials:

I think there is a plan for some of them to engage in trials but there is also a belief that there is lots of money, and that they do not need to do the trials because it is just going to be prescribed and therefore it is going to be okay. There are one or two companies that are discussing with us about doing the trials, but they cost a lot of money, and are they going to get their licence at the end? If it is going to be prescribed anyway, they may or may not need to. If they are based in countries where it is all legalised, they may not feel that they need to do that.55

57.We appreciate that it is not all pharmaceutical companies who bring this attitude, but it is unacceptable that some are behaving this way. Industry’s lack of engagement is one of the reasons why there is a lack of robust international evidence. The British Paediatric Neurology Association suggested that:

The manufacturers of the CDMPs may not be willing to provide their product for analysis in robust RCTs. There is evidence from other countries that manufacturers have been reluctant to facilitate RCTs in the past and this may be why there are no RCTs of these products (with the exception of Epidiolex) in the scientific/medical literature.56

58.A further barrier to research is that pharmaceutical companies do not want to carry out the necessary research themselves to achieve a licence. In part, this reflects the difficulties in obtaining a patent for CBPMs as the chemical compound can be difficult to patent. However, there are few incentives for drug companies to go through the licensing process, especially if any products they produced could be rapidly produced and sold by other companies. Companies may also not want to carry out research because randomised controlled trials can be expensive to carry out. The National Institute for Health Research (NIHR) call for research proposals demonstrates that public money is being invested into research. This is against the norm as it is private companies who stand to profit from conducting research and achieving a licence. The Chief Medical Officer told us:

Randomised controlled trials are the only way to get these drugs licensed and they would normally be funded by the industry. It is time that the industry started to say what they are going to do about funding trials to get the licences so that patients can have access. This cannot be just left to the public sector.57

As suggested above, it may be that industry does not want to invest the money into research itself as they believe their products will soon be prescribed anyway.

59.We heard throughout our inquiry that there are a vast range of CBPMs. It would not be feasible to attempt research into every single product. Professor Goddard told us:

As I said earlier, we need to focus on a very small number of products. If you are trying to choose from 50-odd different types of cannabis, it will get confusing and will take much longer to get some answers. You look at some pure CBD, CBD with a little bit of THC, and CBD with a bit more THC; then you focus on specific areas. That might be, for example, looking at pain in patients with fibromyalgia or patients with multiple sclerosis. If you try to do too many things, it is going to take much longer and we will not get a clear answer, so we have to be very focused.58

By focusing research on certain products, there is a greater likelihood of these products rapidly gathering an evidence base, thereby improving the availability for patients.

60.It is unfair on patients and their families if they are asked to wait years for research to be conducted and for clinicians to prescribe. Families are seeing CBPMs being used in others and feel that they are being denied potentially efficacious treatment. This sentiment is driving families abroad to source these products themselves. Peter Carroll told us:

Dame Sally Davies herself said that we have to wait maybe three or four years before that kind of high-quality, gold-standard data is with us. How do you explain that to the parent sitting behind me now who is—and you may disapprove of this—sourcing a full plant extract cannabis, bringing it into the country and treating her child with it, and the child has improved dramatically?59

We should not be treating patients or their families who are resorting to bringing medication here from abroad because they cannot obtain it on prescription here as if they are committing a criminal offence. Neither should patients have their medication confiscated, as happened recently to the mother of Teagan Appleby. We are pleased that following the outcry in Parliament and beyond, the medication was subsequently restored to Teagan’s family. This cruel practice must not happen again.

61.We were told of examples where the use of CBPMs has benefitted a patient. Peter Carroll said:

I showed them [BPNA] the story of Alfie Dingley, who had 150 seizures a week, each potentially life threatening, and now goes 300 days without a single seizure, rides a bike and goes to school. Does he have to wait for a randomised controlled trial?60

Conclusions and recommendations

62.The current evidence base for the safety and efficacy of CBPMs is not extensive or robust but there are compelling examples in some particularly distressing and dangerous conditions such as intractable childhood epilepsy which highlight the urgent case to clarify their place in treatment. Good quality research is required and circumventing this is unhelpful in the long run to future patients.

63.Individual cases highlight the potential benefit which CBPMs might bring, but individual cases do not amount to a strong enough evidence base for licences, or give clinicians and families the information on the best combinations of CBMPs and their place in treatment or allow an informed discussion of potential risks. Robust randomised controlled trials must be carried out as soon as possible. For highly effective treatments we heard that randomised controlled trials can demonstrate this with smaller numbers of patients and in a shorter time frame than is required for treatments with less marked benefits.61

64.We do not agree that randomised controlled trials should be set aside for cannabis. There are well rehearsed dangers in using anecdotal evidence. The Government and relevant organisations should focus on expediting and encouraging clinical trials. Carrying out clinical trials is the safest and most effective way of ensuring that patients gain access to the most appropriate medication.

65.There are a number of barriers to research into CBPMs. Industry’s lack of willingness to provide their product and to carry out research themselves is a great concern. We also recognise that the chemical compounds and how patients with the same condition can react differently can make research challenging but not insurmountable.

66.We reiterate concerns expressed in our previous report, Brexit: medicines, medical devices and substances of human origin, about the future of collaborative research between EU member states. Now that the UK has changed the scheduling of cannabis to facilitate research it would be helpful to be able to participate in multi-centre pan European trials. The Government needs to set out how it will ensure that Brexit does not jeopardise opportunities for patients to participate in international clinical trials and post marketing surveillance.

67.The Department of Health and Social Care should investigate those instances where pharmaceutical companies do not provide their medicinal cannabis product for research and take appropriate action where necessary. The Department should not be afraid to ‘name and shame’ companies who are not doing all they can to make their products available for research. The Department should also set out a plan to encourage industry to take a more active role in research itself and should present this plan in response to this report.

68.We welcome the broad call for research proposals into medicinal cannabis products by the National Institute of Health Research (NIHR). The Department of Health and Social Care and the NIHR should encourage and focus research into those specific conditions where the Chief Medical Officer’s report found good evidence for the use of cannabis based medicinal products.

69.The National Institute of Health Research should make resources immediately available for a programme of clinical trials for the treatment of intractable epilepsy. This will allow many more patients to access treatments in specialist centres. These trials should be facilitated as a matter of urgency. Families of children suffering from these distressing and life-threatening conditions should not have to travel abroad to seek treatment, but we will fail future patients if we do not establish the evidence base for the place of medicinal cannabis in treatment.

70.The Department of Health and Social Care should set out in its response to this report how it will work with research organisations here in the UK and internationally to ensure that research is being co-ordinated and encouraged in the most appropriate areas. Government should also set out how it will ensure that the future of European multi-centre clinical trials and the post marketing surveillance that protects patient safety are not put at risk by Brexit.

71.The Department of Health and Social Care should look at how medicinal cannabis is made available to patients in other EU member states such as the Netherlands and see whether lessons might be learnt which could be helpful.


23 Q5

25 GW Pharmaceuticals (DMC0090)

28 GW Pharmaceuticals (DMC0090)

35 Q7

41 British Paediatric Neurology Association (DMC0087)

42 Association of British Neurologists (DMC0017)

43 Royal College of Physicians (DMC0034)

45 Association of British Neurologists (DMC0017) Faculty of Pain Medicine of the Royal College of Anaesthetists (DMC0032) Royal College of Physicians (DMC0034) Dr Ruth Williams (DMC0039) Cambridge University Hospitals NHS Trust (DMC0053) Dr Sophia Monica (DMC0063) Professor J Helen Cross (DMC0065)

48 General Medical Council (DMC0071)

49 Mr Robin Emerson (DMC0098)

50 Professor J Helen Cross (DMC0065)

51 Mr Robin Emerson (DMC0098)

52 Department of Health and Social Care (DMC0020)

53 British Paediatric Neurology Association (DMC0087)

56 British Paediatric Neurology Association (DMC0087)

57 Q3




Published: 3 July 2019