Cannabis as a treatment for cancer symptoms is heavily supported by existing research, a large-scale review has found. What’s more, existing research also supports the potential of cannabis to directly fight the disease.
These findings represent a significant contribution to debates within the medical field and concerning marijuana policies, which have long hampered the study of such questions—though the methodology has attracted criticism.
A team led by Ryan Castle looked at 10,000 peer-reviewed studies concerning cannabis, cancer and treatment. Their findings, published in the journal Frontiers in Oncology on April 14, indicated “a significant consensus supporting the use of medical cannabis” in cancer treatment.
Overall, they found a statistical correlation in studies supporting cannabis as treatment that was 31.8 times larger than those not in support. Cannabis was shown to have anti-inflammatory properties, to manage symptoms like pain, nausea and poor appetite, and to function as an anti-carcinogenic.
“The most surprising thing was how uncontroversial the science was, given how controversial the policy debate is.”
Castle, the director of research at the Whole Health Oncology Institute, told Filter that going in, his team expected the scientific literature to be more closely divided.
“But what we found was a much more striking level of agreement on the fact it is beneficial,” he said. “Not just with what we call palliative effects, how the person feels and their symptoms, but that it is directly reducing the disease burden, improving cancer symptoms and outcomes, cancer recovery as well as associated diseases based on inflammatory effects. The most surprising thing was how uncontroversial the science was, given how controversial the policy debate is.”
“The consistency of positive sentiments across a wide range of studies suggests that cannabis should be re-evaluated within the medical community as a treatment option,” the review authors wrote. “The findings have implications for public health research, clinical practice, and discussions surrounding the legal status of medical cannabis.”
The authors explained how the existing literature includes a variety of different study designs—like randomized trials, observational and case report studies—focused on distinct areas such as care during chemotherapy or cancer-fighting properties. Cannabis is complicated to study, as myriad strains and formulations contain many different compounds—with THC and CBD the most familiar and most researched.
Normally, conducting a meta-analysis on this scale would be likely to take years. But these researchers used a machine learning technique called “sentiment analysis,” which scans huge chunks of text data to find trends in the writing that are “positive,” “negative” or “neutral.”
There’s a clear advantage in speeding up a daunting task, although the technique is still in its early stages when applied to medical literature. Determining whether a finding is positive or negative is not always clear-cut, especially when researchers commonly seek to use objective, non-emotive language. The correlations the technique identifies are not completely precise or exhaustive, and cannot predict effects or efficacy at the individual scale.
As such, some cancer researchers are skeptical. Chris Twelves, a clinical cancer professor at the University of Leeds in the United Kingdom, co-published his own literature review of cannabis-for-cancer studies in 2024. It looked at a smaller quantity of research that Twelves emphasized is of higher quality and reliability: 13 systematic reviews and five primary studies. His results were far more ambiguous, stating “Whether cannabis and/or cannabinoids can improve other supportive care outcomes remains uncertain,” and recommending against doctors using cannabis for cancer care outside of clinical trials.
“It it is misleading to call this a ‘meta-analysis,’ which has a specific meaning when applied to new cancer treatments,” Twelves told Filter of the new review. “Using it here infers the authors are presenting clinical trials whereas what they report is a disparate collection of disparate data, very little of it from clinical trials.”
“I’m not familiar with ‘sentiment analysis,'” he added. “Investigators invariably see positive aspects to their work so relying on them to ‘mark their homework’ isn’t a rigorous approach to evaluating data. The authors comment on others ‘cherry picking’ data on cannabinoids but in research, as when shopping, sorting the good from the rotten is important.”
In their keyword analysis of existing research, the review authors found over 25,000 unique occurrences of keywords suggesting support for the efficacy or benefits of medical cannabis, comprising 64.5 percent of the total. That’s more than twice as many as the 30.6 percent of keyword occurrences that indicated the evidence did not support medical cannabis. A much smaller 4.7 percent of keyword occurrences were found to indicate unclear or indeterminate results.
“When you look at the findings for cancer dynamics and the interaction with cancer directly, we found huge one-sided positive associations with medical cannabis.”
Some of the findings for specific topics were much more lopsided. With “cancer dynamics,” for instance—concerning tumor size, growth and remission—language in support of cannabis for cancer had a 32.4-times higher correlation than that not in support.
“When you look at the findings for cancer dynamics and the interaction with cancer directly, like tumor reduction and the size of the cancer present,” Castle said, “we found huge one-sided positive associations with medical cannabis. The vast majority of studies that looked at specifically the size of tumors with medical cannabis, found it was having a positive effect.”
In the review, the authors described cannabis as having been found to have a “robust association” with reduced tumor growth and a healing effect on cancerous processes.
Across the board, there was also strong support for the anti-inflammatory effects of cannabis—a key finding because it concerns a risk factor for many diseases. Cannabis was strongly supported as an adjunct therapy for chemotherapy, for appetite management and for pain relief, too.
Overall, the authors wrote, “These findings revealed a significant trend suggesting support of cannabis’ therapeutic potential, particularly in managing cancer-related symptoms and possibly exerting direct anti-carcinogenic effects. Across all categories examined—health metrics, cancer treatments, and cancer dynamics—there is a consistent consensus that supports the potential of medical cannabis.”
“The growing consensus around cannabis’ therapeutic benefits also highlights the need for medical professionals to stay informed about the latest research,” the authors added, “as cannabis continues to evolve from a controversial substance to a scientifically validated treatment option.”
Twelves’ view of the existing research is that the evidence is mixed on cannabis’s efficacy for treating appetite, pain and mood problems associated with cancer. And as for its use in actually fighting cancerous tumors, he said there’s almost no solid clinical data, and the existing case reports are of poor quality. “I agree that there is a need for more high quality clinical trials of cannabinoids in cancer,” he said, “but we also need a balanced approach to analyzing what data we have.”
“What we really need is for cannabis to be rescheduled.”
The evolution the authors of the new review describe has important policy implications, of course, as the US federal government has failed to enact any major marijuana policy reform. The Biden administration initiated a process to move cannabis from Schedule I to the less-restricted Schedule III of the Controlled Substances Act, but it wasn’t completed. The Trump administration’s Drug Enforcement Administration, under an anti-marijuana acting administrator, has put the process on hold.
The review authors highlighted this situation, as well as calling for further research “to explore the full therapeutic potential of cannabis and address knowledge gaps.”
“What we really need is for cannabis to be rescheduled,” Castle said. Under Schedule I, “it is considered too dangerous and lacking any and all medical benefits, and can’t be studied in human trials except under extremely limited circumstances. That’s made it difficult to gather the information we would need to be able to have these double-blind trials.”
“Rescheduling to Schedule III would allow researchers like us to work directly with cannabis and patients without worrying about federal interference,” he continued, “and that would allow a critical mass of clinical trials that can get into the details of what cannabis is capable of, what are its benefits and limitations.”
*Update, April 28: This article has been edited to reflected the comments from Chris Twelves, received after publication.
Photograph by Dank Depot via Flickr/Creative Commons 2.0
Show Comments