At long last, there is a published veterinary clinical trial of a cannabis-based treatment! As I say in nearly every article I write, a single study neither definitively proves nor disproves even the specific hypothesis being studied, much less all of the other claims that might be made about the treatment being studied. Nevertheless, this is an important and welcome beginning to the process of a science-based evaluation of cannabis as a potential therapy for veterinary patients.
Gamble L-J, Boesch JM, Frye CW, Schwark WS, Mann S, Wolfe L, Brown H, Berthelsen ES and Wakshlag JJ (2018) Pharmacokinetics, Safety, and Clinical Efficacy of Cannabidiol Treatment in Osteoarthritic Dogs. Front. Vet. Sci. 5:165. doi: 10.3389/fvets.2018.00165
This study was conducted at the Cornell University College of Veterinary Medicine. It had two components. First, 4 beagles were given a commercial cannabis-based product containing predominantly cannabidiol (CBD), with low levels of THC and other compounds, in an olive oil base. Previous research has suggested that an oil-based formula is much more readily absorbed than other oral forms of CBD. The investigators measured the blood levels of CBD obtained at two doses (2mg/kg and 8mg/kg), and calculated the half-life of elimination, which provides an estimate of how long these blood levels may be maintained and, thus, how often it might be necessary to give a drug to achieve effective levels. These beagles were also observed for any obvious adverse effects.
The second component of the study was a randomized, double-blinded, placebo-controlled crossover clinical trial looking at the potential effects of the product on measures of pain and function in dogs with arthritis. 22 dogs were enrolled in the study, and 16 completed the trial and were included in the data analysis. Two validated owner-employed scales of pain and function were used, the Canine Brief Pain Inventory (CBPI) and the Hudson Visual Analogue Scale (Hudson). The dogs were also evaluated for pain, lameness, and weight-bearing by veterinarians. Some objective measures were used as well (force-plate measurements), but due to the fact that most dogs were lame in multiple legs, these data were unreliable and so not included in the analysis.
The design involved randomly assigning dogs to receive 4 weeks of either CBD or placebo treatment and evaluation at 2 and 4 weeks, then a washout period followed by a switch to the other treatment. Bloodwork was also done at the start of each treatment and at 2 and 4 weeks into the treatment for both placebo and CBD periods. An olive oil placebo with anise and peppermint oil was used as a placebo to mimic the appearance and odor of the CBD product (no measures were used, however, to determine if the owners and vets were effectively fooled by this technique). The subjects received the CBD, at the 2mg/kg dose, or the placebo every 12 hours.
The patients were mostly medium to large dogs (18-50kg), and about 2/3 were females. The subjects had arthritis in a variety of joints, and 9/16 were taking either carprofen or meloxicam during the study. The authors indicated that fish oil and glucosamine/chondroitin were also permitted but didn’t report whether any of the subjects were taking these agents.
This component of the study showed significant absorption of CBD. The higher dose, not surprisingly, resulted in blood levels more than 4 times those achieved with the lower dose. There is no reliable data, in dogs or any other species, for how much CBD one has to get into the blood to achieve a desirable effect or how much might cause harm, so this preliminary data is useful. It isn’t yet clear, however, what concentration should be seen as a therapeutic target or as an upper limit.
The elimination half-life was about 4 hours. This suggests that to maintain consistent blood levels, it might be necessary to dose this product 3-4 times per day. As the authors indicate, longer-term pharmacokinetic studies with a variety of dose and administration intervals are needed before the optimal amount and frequency can be known.
There were statistically significant and clinically meaningful decreases from baseline in CBPI and Hudson indices at 2 and 4 weeks during the CBD dosing, and there were no changes in these levels during the placebo period. Veterinary assessments more mixed, with decreases in pain measures during CBD administration but no apparent effect on lameness or weight-bearing (as opposed to NSAID use, which did reduce lameness score as well as pain)
Bloodwork showed few changes during the treatment and placebo periods. Creatinine and BUN increased during both periods, and this was associated with NSAID use, however the numbers always stayed within the normal range. The only clinical laboratory change that appeared meaningful was an increased in the liver enzyme alkaline phosphatase (ALK) in dogs receiving CBD. Previous research has shown that CBD affects the cytochrome enzyme system in the liver, and this kind of change in lab value is likely a reflection of this. The ALK elevation by itself does not suggest any harm to the liver, but the effects of CBD on the cytochrome system could lead to interactions with other drugs, which could present safety or efficacy concerns for these drugs.
The authors did not report any adverse effects in the study subjects, but they also did not report any mechanism for monitoring or evaluating undesirable effects, so it’s unclear whether any formal monitoring of adverse effects was done as part of the study.
All research studies have strengths and limitations which must be taken into account in order to determine how much confidence to place in the results and the extent to which the results can be applied to real-world patients, who may differ from the research subjects in various ways. The major strength of this study was a solid design incorporating most of the major tools for minimizing bias and other errors.
The subjects were randomly assigned to treatment order, and the crossover design further reduces potential differences between treatment and placebo groups that can confuse the results. Both owners and veterinarians were blinded to the treatment, though there was no specific measurement of whether the method used for blinding was successful. About half of the owners were “intimately involved in veterinary medical care,” presumably being vets or veterinary technicians/nurses. It is unclear how many of the owners had previous experience with medical or recreational cannabis and might have been able to distinguish the treatment product from the placebo.
Appropriate statistical methods were employed, and the outcome measurement tools (the CBPI and Hudson index) were established and validated measures of clinically relevant signs.
The authors themselves acknowledge that the sample size of 16 dogs is a limitation of this study. Extensive evidence shows that smaller studies are more likely to generate unreliable results which don’t accurately predict the effects of treatments in larger populations.
It is also concerning that 27% of the subjects initially enrolled dropped out of the study, since loss to followup greater that 20% is often considered a risk for bias in such studies. The authors report the reasons for dropout of 5 of the 6 patients not included, and there is no obvious pattern of differences in withdrawal between the study and control periods to suggest these cases would affect the final results.
The subjects were predominantly large breed and about 2/3 were female. This is reasonably representative of the population of dogs seen for treatment of arthritis in general practice, but it does limit the generalizability of the study results to small-breed dogs.
Both the pharmacokinetic and clinical trial components of the study were short (24 hours after single dose, and 4 weeks respectively). These are reasonable and practical starting points for clinical research, but longer-term studies are needed to better evaluate optimal dosing and the effects of CBD over the long term.
The outcome measurements used were subjective instruments, but they were both commonly used and formally validated tools, and the assessors were blinded, so this likely had little impact on the results. Meaningful improvement on objective measures would certainly be a desirable indication of the true impact of CBD on arthritis symptoms.
The fact that many of the owners were apparently veterinary professionals is also an interesting wrinkle to this study. Certainly, this is not representative of the general pet-owning population, and both the use and evaluation of CBD products might very well be different in the hands of owners without medical training. However, the blinding of the owners (if it was effective) should at least diminish any impact of this on the results.
There was no apparent placebo effect in this study, which is highly unusual for an arthritis study, especially one using subjective measurement instruments. The authors suggest one explanation for this may be that so many owners were veterinary professionals and might be less susceptible to placebo effects. This is a common but inaccurate notion, and studies of other arthritis treatments have shown that caregiver placebo effects impact veterinarians to nearly the same extent as pet owners. It’s not clear what, if any implication this lack of a placebo effect might have for the results of the study in general, but it is a little odd.
Also concerning is the lack of any reported adverse effects of the CBD. Human trials showing benefits from CBD and other cannabis-based treatments have consistently showed relatively high rates of minor side effects, as well as some that limit the tolerability of the treatment. This is actually a good sign since any real, truly effective medicine is going to have some risks and undesirable effects that have to be balanced against its benefits. Any study showing no adverse effects at all is a bit worrisome since it could be a sign the treatment is not truly active or that the study did not properly detect side effects that may emerge once the treatment is more widely used in the real world.
There are a couple of other interesting issues that arise in this study which are not, strictly speaking, limitations of the study itself but which might influence the application of the results to clinical practice.
The only treatments reported for study subjects other than the CBD and placebo were NSAIDs, either meloxicam or carprofen. The authors indicated that fish oil and glucosamine/chrondroitin were allowed, but they didn’t report whether or not any subjects were taking these. Glucosamine and chondroitin very likely have no meaningful clinical effects, so this would probably not have affected the results. There is some weak evidence that fish oil, however, might improve arthritis symptoms, so hopefully this was not an overlooked variable in this study.
NSAID use was associated with improvement in pain and in function, as expected. Whether or not the CBD had more or less impact given with or without concurrent NSAID use could not be determined from this small study. This is an important open question given that it seems many owners are choosing CBD as a first-line therapy prior to using NSAIDs due to beliefs about the safety and efficacy of the two treatments that are not always evidence based. One of the most common problems in human medicine is the unanticipated interactions between new drugs and other treatments that emerge when drugs go from carefully controlled studies to widespread use, and this will have to be kept in mind going forward with cannabis in veterinary patients.
The authors also mentioned that a significant reason for choosing the 2mg/kg dose was that the the cost of the product was prohibitive for larger dogs at higher doses. Cost is always a consideration in the use of veterinary treatments, since many owners are limited in what they are able and willing to spend on veterinary care. Unfortunately, the more work a company puts into developing and testing a product and ensuring good quality control, the more expensive that product is likely to be.
Given the abundance of untested and unregulated cannabis products on the market, several of which have already proven unreliable in quality, it is possible that owners may use the data for this product as justification for using CBD and then choose another, cheaper product that hasn’t undergone the same clinical testing or quality control. I have seen this happen in the past, when a probiotic product with better supporting research evidence than others on the market was often refused by clients, and ultimately withdrawn from the market, partly due to its higher cost.
Quality control has been a particular problem for cannabis-based remedies, with many failing government or independent testing of their contents and label claims. It is important for pet owners to realize that research evidence is often specific to the treatment tested, and it can’t always be generalized to other products that claim to be similar. While I don’t endorse any particular product over others in the absence of head-to-head comparison studies, which are almost never done, I would caution veterinarians and owners not to make the assumption that one encouraging arthritis study for one product validates the use of all the other products out there.
In the same vein, this study is very specific to arthritis in dogs. It says nothing whatsoever about the use of CBD for other conditions, including epilepsy, pain associated with other diseases, nausea, etc. These uses for CBD have to be validated in their own right. And other cannabis-based compounds also cannot be assumed to be safe and effective, for arthritis or anything else, based on this study of CBD. In medicine, the details matter a lot, and generalizing from even good studies such as this is a dangerous practice.
One small study of CBD for arthritis is not one giant leap for cannabis-based medicines in pets. Additional studies on the pharmacology and clinical effects, both positive and negative, need to be conducted, and rash generalizations from this study to other compounds, other conditions, other species, and other products is not justified.
That said, this is a good-quality study with limitations that are important but certainly not fatal. It is an excellent beginning to the science-based use of cannabis in dogs. Based on the existing evidence, including this study, I believe it is reasonable to consider oral CBD as a treatment for dogs with arthritis with the following caveats:
- Other treatments with better supporting evidence, such as calorie restriction for weight loss and NSAIDs, should be attempted first
- This product or others with verifiably similar formulation and quality control should be used
- Patients must be closely monitored for adverse effects and potential drug interactions which are almost certain to appear in larger, more diverse populations of patients than used in this study
As I have said before, the evidence for cannabis as a medical therapy is limited but encouraging and growing rapidly. The political and legal climate appears to be changing in ways that will hopefully make it easier to conduct the needed research and make use of cannabis products that can demonstrate safety and efficacy for veterinary patients. This study is unquestionably an important milestone, and I look forward to more and better evidence and ultimately the integration of cannabis, where justified, as yet another tool in the toolbox of science-based veterinary medicine.