I analyzed clinical trials of cannabis from the federal clinicaltrials.gov database, which revealed trends important for the future of medical cannabis.
As I mentioned in a previous article, the first observation of the medicinal properties of cannabis in the Western medical literature was in 1843. However, it took many more decades for controlled clinical trials of new medicines to become standard.
Now we are in the era of evidence-based medicine and medical cannabis is being studied in a wider range of diseases than ever before. Since most clinical trials are captured in a US federal government database, I decided to take a look at this important public dataset.
What I found is both a recent history of the clinical study of medical cannabis, but also trends that point to where this field is going in the future.
Clinical trials of medical cannabis on clinicaltrials.gov
All data was obtained from clinicaltrials.gov. This site is a registry of clinical trials from all over the world.
ClinicalTrials.gov was created as a result of the Food and Drug Administration Modernization Act of 1997 (FDAMA). FDAMA required the U.S. Department of Health and Human Services (HHS), through NIH, to establish a registry of clinical trials information for both federally and privately funded trials conducted under investigational new drug applications to test the effectiveness of experimental drugs for serious or life-threatening diseases or conditions.
I can’t guarantee that every clinical study of cannabis is in this registry since not everyone is required to report their trial to clinicaltrials.gov. You can read more about the specific rules regarding who is required to register their trial and report results.
I did an Advanced Search only for interventional studies containing “cannabis” as a keyword. An interventional study is where the investigators administer a drug (for example) to see what it does. Also, clinicaltrials.gov automatically includes acronyms of cannabis such as marijuana, etc.
This search resulted in 435 clinical trials of cannabis. I further selected clinical trials with the following characteristics:
- The intervention included cannabis, a cannabis extract, or THC (I excluded synthetic cannabinoids)
- The trial was performed in patients (as opposed to certain studies done in healthy volunteers)
- The trial had a primary or secondary objective of therapeutic intent (as opposed to a pure safety or pharmacokinetic study)
Overview of clinical trials of medical cannabis
After applying the above filters, 112 clinical trials remained that fit all criteria. The majority of studies obtained in the original search were not assessing cannabis as a potential medicine, but studying things like dependence or another non-therapeutic aspect of cannabis. This dataset of 112 clinical trials of medical cannabis was used for all subsequent analysis.
All clinical trials were categorized into one of 7 categories of diseases: chronic pain (non-neuropathic), neuropathic pain, neurologic, neurodegenerative, psychiatric, cancer and other. The above plot shows the start time and end time (when available) of the trials in my dataset. Clinical studies go back to 2001 since clinicaltrials.gov was made public in 2000.
Disease areas of clinical trials of cannabis
I next assessed the disease categories and individual diseases that were studied. The number of clinical trials of each disease is shown below. The most commonly studied individual disease (by far) was multiple sclerosis, with over 20 clinical trials. The variety of different diseases studied is impressive. I cannot think of another drug class outside of cannabinoids that has been studied across such a diverse set of diseases.
The above diseases were categorized mainly on the patient population. However, the disease alone does not tell you what the primary objective of the clinical trial was. For example, in a population of cancer patients, a study may be looking at progression of the tumor itself, but more likely is assessing the impact of cannabis on pain, appetite, anxiety, or chemotherapy-induced nausea and vomiting. Therefore, I reanalyzed the data to look at what symptom was primarily assessed.
This analysis shows that pain is by far the most common symptom assessed in clinical trials of medical cannabis, representing 42% of all trials. Spasticity (17%), nausea & vomiting (6%), and anorexia (6%) were also fairly common symptoms assessed in clinical trials. 29% of trials were grouped into the “Other” category. Many of these were measures of disease severity that are specific to individual diseases,.
Trends over time in clinical trials of cannabis
The total number of clinical trials of medical cannabis starting each year appears to be consistent since the year 2001. Each year, 7 new clinical trials of medical cannabis are started on average.
Now lets turn our attention to trends in individual therapeutic areas. Although the total number of clinical trials initiated hasn’t changed much in 15 years, there are distinct trends in each disease area.
The neurologic area (dominated by multiple sclerosis trials, but also including cerebral palsy, cervical dystonia, familial dysautonomia, and fibromyalgia) has decrease by two thirds. The neuropathic pain area (including peripheral neuropathy from HIV and diabetes, brachial plexus and spinal injury, and phantom limb pain) has also sharply decreased.
The likely reason for the decrease in clinical trial starts in multiple sclerosis and neuropathic pain is likely due to the results obtained in the earlier trials. Many of these trials were sponsored by GW Pharma for Sativex, their cannabinoid extract oral spray. This product has now gained regulatory approval spasticity due to multiple sclerosis and symptomatic relief of neuropathic pain in some countries. Having gained approval in these diseases, their research focus has moved to other areas.
All other disease areas have seen an increase in the start of medical cannabis trials. In fact, if we exclude multiple sclerosis, cannabis clinical trial starts for all other diseases over the last 5 years have increased by 41% compared to a decade earlier.
The most dramatic increases in clinical trial starts occurred in two diseases areas. The first is neurodegenerative diseases (composed of Alzheimer’s Disease, Amyotrophic Lateral Sclerosis, Huntington’s Disease, Parkinson’s Disease, and Dementia), which increased from a single trial start in 2001-2005 to 5 trial starts in 2011-2015. The second area is the “Other” category, which includes such diverse diseases as tinnitus, sleep apnea, sickle cell, and irritable bowel syndrome. The Other category has increased from 2 trial starts in 2001-2005 to 8 trial starts in 2011-2015. These trends indicate that the diseases being considered for medical use of cannabis are much more diverse than they were a decade ago.
Phases of clinical trials of medical cannabis
In a typical drug development scenario, most trials are considered to be Phase 1 (small pilot study), with fewer Phase 2 (looking at preliminary efficacy/safety) and even fewer Phase 3/4 (confirmatory efficacy) trials. The reason is that with a new drug there are often issues with safety or pharmacokinetics that cause drug programs to be terminated during Phase 1. A lack of efficacy or a safety issue may cause a program to be terminated in Phase 2. Only a fraction of drugs look promising enough at the end of Phase 2 to justify the investment of a more expensive Phase 3 trial.
Below, I plot the number of clinical trials of medical cannabis by research phase. If a trial was classified as spanning two phases, I counted it as the later phase. The majority of clinical trials were classified as Phase 2 or Phase 3/4 studies, with only 6 trials classified as Phase 1 studies. A specific phase was not listed for 9 trials.
The Phase 3/4 trials were dominated by studies of Sativex. Out of 46 Phase 3/4 trials, 29 of them (63%) used Sativex.
Why is the number of Phase 1 trials of medical cannabis so low? The main factor is that the safety and pharmacokinetic properties of cannabis, cannabis extracts and synthetic THC have already been well established. There is little need to repeat a Phase 1 study and in most cases, investigators can go straight to a Phase 2 to look for signals of efficacy. However, a study may still be considered Phase 1 if it is a small pilot study or safety needs to be established in a sensitive patient population.
Next, I looked at how the phases of clinical trials of medical cannabis have changed over time.
As seen above, there was a 33% drop in Phase 3/4 study starts from 18 trials in 2001-2005 to 12 trials in 2011-2015. This drop was mostly due to the large number of trials initiated for Sativex in multiple sclerosis and neuropathic pain in the earlier time range.
Interestingly, the number of trials considered to be Phase 1 has been steady increasing. There was only 1 trial prior to 2010, but then 2 from 2010-2015, and already 2 more in 2016. This may represent the expansion of cannabis in new patient populations. The Phase 1 studies included use in palliative care, schizophrenia, tinnitus, lung cancer, gliomas, and a new drug combination for HIV wasting syndrome.
Funding of clinical trials of medical cannabis
I next assessed how clinical trials of medical cannabis are funded. On clinicaltrials.gov, the options for funding include Industry, NIH, US federal government, and Other (as well as any combination of these for joint funding). for this analysis, I combined the NIH (US National Institute of Health) and US federal government, since the NIH is part of the US federal government.
Funding of clinical trials of medical cannabis was dominated by the “Other” category, which may include both non-US federal governments as well as private non-profit organizations. These entities funded 55 clinical studies. Industry-sponsored trials were second, with a total of 35 trials of medical cannabis. In third place were trials funded by a mix of industry and other sources. Despite having the largest budget for biomedical research of any country, trials funded by the US federal government were a small minority. Only 2 trials were independently funded by the US federal government and an additional 7 trials were sponsored jointly by the US federal government and another organization. Altogether, trials funded by the US federal government represent 8% of all clinical trials of medical cannabis.
Since clinical trials can vary widely in size and cost between a Phase 1 and a Phase 3 trial, I assessed the total number of patients enrolled in trials by the funding source of the trial.
Although the “Other” category funded the largest number of clinical trials of medical cannabis, industry-sponsored trials enrolled the highest total number of patients. Clinical trials sponsored solely by industry enrolled 6910 patients, more than twice as many as in trials funded by the “Other” category. Clinical trials funded solely or partly by industry enrolled 68% of all patients in trials of medical cannabis. This reflects that industry has sponsored the majority of large Phase 3 studies that seek to confirm efficacy in treating a disease. This trend is typical in drug development.
Funding of trials by the US federal government using this metric is low. Of all patients enrolled in a clinical trial of medical cannabis, only 5% were enrolled in a trial funded solely or in part by the US federal government.
Next, I looked at how funding has changed over time, using the total number of patients enrolled.
Since GW Pharma started a large number of Phase 3 trials in the 2001-2005 period, it is not surprising that enrollment in industry-sponsored trials has decreased since that time. However, the majority of patients still come from industry-sponsored trials. In the period from 2011-2015, there were more patients enrolled in industry-sponsored trials than trials from all other funding sources combined.
Interestingly, enrollment in trials funded by the “Other” category increased by 96% over the period of a decade. This may represent an increased investment in cannabis trials by non-US governments.
Using a dataset from clinicaltrials.gov, I have shown that the number of clinical trials of medical cannabis starting each year has been fairly constant for the last 15 years. However, the characteristics of these trials has substantially changed. Trials started in the 2001-2005 period were focused almost exclusively on two disease areas – neuropathic pain and multiple sclerosis. Many were large Phase 3 trials funded by industry.
Clinical trials of medical cannabis have shifted towards assessing efficacy in a much wider number of therapeutic areas and diseases. There has been an increase in smaller Phase 1 pilot studies. Clinical trials are funded more often from non-industry sources, although the US federal government has not contributed to this.
UPDATE: See key information from this analysis of clinical trials of cannabis in an infographic.
I will continue to analyze trends in the science of cannabis. Please add your email to stay up to date!
Last modified: April 12, 2017