Cannabinoids are a medically interesting group of molecules, and a great example of the potential benefits of pharmacognosy – the development of drugs from plants and other natural sources. Nature is a great laboratory, with millions of experiments over hundreds of millions of years. Plants mostly evolved substances as defense mechanisms, to be toxic to the creatures that might eat them. However toxins in measured low doses and in the right clinical context can be medicines. A toxin that causes diarrhea, for example, can be a treatment for constipation.
But of course just consuming plants is not a good idea. Most of the plants we eat for food were cultivated to have dramatically reduced protective chemicals. Animals evolved a taste for bitterness, which is a way of detecting these potentially harmful chemicals. So if our ancestors selected plants that were less bitter, they were also less toxic. I think most people know they shouldn’t walk into the woods or a field and eat a random plant. Chances are it will taste awful an they will get sick.
But through collective experience most cultures detected plants in their environment that can have potentially medicinal effects. Without modern scientific methods, however, only the most obvious effects were reliably detected. One such effect may be a reduction in pain, although pain is also highly susceptible to placebo effects. We are now in the middle of extreme hype over one plant’s potential medical benefits, including pain – cannabis. Two main constituents are phytocannabinoid tetrahydrocannabinol (THC) which is the primary psychoactive compound in cannabis, and cannabidiol (CBD).
SBM has already published several articles detailing the fact (going back to 2014) that the hype surrounding CBD and other cannabinoids as medicinals is outstripping the evidence. Historically putting hype before evidence usually doesn’t end well. We are seeing the creation of an industry and a market for CBD products before there is published scientific evidence confirming safety and efficacy. This was bound to happen in our loose regulatory environment. Experts were calling for lifting of restrictions on research, so we could find out if CBD and other products were useful. Promoters immediately took this to mean that it worked and started making money.
Preliminary evidence is now coming in so we can start to answer the many questions about efficacy. A recent systematic review and meta-analysis looks specifically at cannabinoids and acute pain. Here are the results:
Six trials (678 participants) were included examining oral (5 trials) and intramuscular (1 trial) cannabinoids. Overall, there was a small but statistically significant treatment effect favoring the use of cannabinoids over placebo (−0.90, 95% confidence interval [CI] −1.69 to −0.1, i2=65%, p=0.03). When stratified by route of administration, intramuscular cannabinoids were found to have a significant reduction in pain relative to placebo (−2.98, 95% CI −4.09 to −1.87, i2=0%, p<0.0001). No difference in effect was observed between oral cannabinoids and placebo (−0.21, 95% CI −0.64 to 0.22, i2=3%, p=0.34). Serious AEs were rare, and similar across the cannabinoid (14/374, 3.7%) and placebo groups (8/304, 2.6%).
For some reason, in their conclusions, the authors combine the evidence from oral and IM cannabinoids to say there is a small but statistically significant effect. I don’t think this makes sense, however, and is ultimately misleading. Rather, the most reliable outcome of this review is that oral cannabinoids have no effect beyond placebo for acute pain, with data from 5 trials. This is still a preliminary conclusion and we could argue for more research being necessary. But it is not encouraging that after 5 studies there is no effect from oral cannabinoids for acute pain. Effect sizes tend to decrease over research time, not increase, as protocols tighten.
While this is 5 studies of oral cannabinoids, they were all with different substances or doses. Two were with Nabilone at different doses, one with THC, one with AZD1940, and one with GW842166. Even combing this data is questionable (which is why I am not sure the meta-analysis was of any use). But reviewing the individual studies is useful.
None of the oral studies produced convincing evidence of efficacy, and together they are negative. One might argue that the studies were not powerful enough (total subjects 622) to detect a signal (pain reduction). But with pain reduction, if the signal is too small then that means the pain reduction is clinically insignificant (which is not functionally different from no pain reduction).
It’s possible that the cannabinoids do have a potentially useful effect but the bioavailability of the oral route is simply too low. The authors note this:
Specifically, oral absorption of cannabinoids is slow and variable with maximal plasma concentrations occurring 60–120 min postingestion but can be delayed upward of 6 h. Cannabinoids are subjective to significant first-pass liver metabolism, which further reduces the bioavailability.
Low oral bioavailability is what dooms most herbals from having a significant clinical effect. This is precisely why they are developed into pharmaceuticals in the first place – the active ingredients are identified, purified, and then tweaked chemically to reduce toxicity and increase bioavailability.
This may also be why the only study to show a significant effect was the intramuscular study, which used a form of artificial THC called levonantradol. When we take a closer look at this study, however, it is not impressive. First, it is small, with 56 subjects. Further, of the six studies reviewed it was found to have the highest risk of bias. This is especially important with pain studies because of the high potential for placebo effects. Further still, needle injections have a known greater placebo effect than taking pills. So basically we have one small preliminary study showing potential benefit for IM levonantrodol. This is the very definition of – too early to hype and requires further research.
So at present the data on cannabinoids in general for acute pain is not impressive. No oral agent shows positive preliminary data, and while further research is necessary this data does not predict a positive outcome. That leaves open the window, however, for developing pharmaceuticals based on these cannabinoids that have greater bioavailability and efficacy.
The IM route for levonantrodol has one preliminary study showing a significant effect, but such preliminary data usually does not survive later research (if we are playing the odds). But this is still encouraging preliminary results and deserves further research.
Research into cannabinoids is critical at this stage. First, we can use more options for the treatment of acute and chronic pain, especially in the face of an opioid epidemic. But we need to know that they actually work. If they don’t work, that is critical to know also, and hopefully that evidence would be used to blunt and reverse the exploding preliminary hype into this new class of drugs.