Cannabis

Also called: marijuana, CBD oil, medical cannabis, THC.

Chapter 19 of The Cancer Evidence Handbook, Part Four. Last reviewed 2026-09-16. How this works.

When to call your team

If you are having treatment that lowers your immune system, contact your care team the same day if you notice:

Dried cannabis can carry fungal spores, and people whose immune systems are suppressed by treatment have died from the infection that follows. Heating it does not reliably remove the fungus, and vaporising avoids burning but does not make it sterile. These are reasons to make a phone call. They are not a diagnosis, and this page cannot tell you what is causing them. Telling your team you use cannabis is what lets them look for this.183

These are reasons to make contact, not a diagnosis. Whatever you are taking or thinking of taking, the people treating you can only weigh it if they know about it.

This block is about what to watch for, and it is not all this page holds: an interaction recorded against a treatment class and what has been recorded about harm are below. Nothing in this block is an all-clear for a question it does not mention.

Appetite and weight Grade C

Human evidence exists and does not settle it.

Harm has been recorded in people for this claim. The grade measures how strong the evidence for benefit is; it says nothing about the harm, and the two are read from different parts of this page.

Cannabis can make you want to eat more, but it has not been shown to stop the muscle loss that matters.

It increases appetite. It has not been shown to preserve muscle.

The harm

The fungal contamination risk described on this page applies to any inhaled use, whatever it is being used for.13

How the grade for appetite and weight was reached

The claim

The claim is that cannabis helps people eat during treatment, and that eating more protects them from wasting.

What it does

The appetite effect is real and long established. The second half of the claim is where it breaks: cachexia is not simple undereating, and appetite is not the same thing as muscle.7

The evidence

The register holds no primary citation for this claim. The chapter states that its position on appetite is written from the general position of the clinical literature and nothing on this page points at a paper for it.7

What has not been shown

No study in this record shows that cannabis preserves muscle or changes the course of cancer-related weight loss. Appetite and wasting are different problems.

This applies to: preserving muscle or changing weight outcomes.

Nobody has run a search to establish this, so it is the register’s reading of what exists rather than a gap somebody has looked for and confirmed.

What to ask

Is my appetite the problem here, or is something else causing the weight loss?

Cancer pain Grade C

Human evidence exists and does not settle it.

Harm has been recorded in people for this claim too; the grade says nothing about it.

Trials of cannabis for cancer pain have given mixed results, and several did not show what they set out to show.

Mixed randomised results, and several trials missed their primary endpoints.

The harm

The fungal contamination risk described on this page applies to any inhaled use.13

How the grade for cancer pain was reached

The claim

The claim is that cannabis relieves cancer pain, including pain that has not responded to other treatment.

What it does

Cannabinoid receptors are involved in how pain is processed, which makes the claim biologically reasonable. Reasonable is where a question starts.7

The evidence

The register holds no primary citation for this claim. The chapter records that its position on cancer pain is written from the general position of the clinical literature and that several randomised trials missed their primary endpoints, and that nothing on this page points at a paper for the position itself.7

What has not been shown

The randomised results in cancer pain do not agree with one another, and the trials that were designed to settle the question did not meet their primary endpoints.

This applies to: cancer pain.

Nobody has run a search to establish this, so it is the register’s reading of what exists rather than a gap somebody has looked for and confirmed.

What to ask

Is there anything for this pain that has better evidence behind it than cannabis does?

Anxiety and sleep Grade C

Human evidence exists and does not settle it.

Harm has been recorded in people for this claim too; the grade says nothing about it.

Many people say it helps them sleep and feel calmer, but there is little trial evidence in people with cancer.

Widely reported by people using it. Little cancer-specific trial evidence.

The harm

The fungal contamination risk described on this page applies to any inhaled use. There is also a licensed and well-evidenced programme for persistent insomnia, which has its own entry in this register.13

How the grade for anxiety and sleep was reached

The claim

The claim is that cannabis helps with the anxiety and broken sleep that come with a diagnosis and with treatment.

What it does

Whatever the mechanism, this is the use people most often report, and dismissing it would not be honest. What is missing is the trial evidence in this specific population.7

The evidence

The register holds no primary citation for this claim. The chapter records its position on sleep as written from the general position of the clinical literature, with nothing on this page pointing at a paper for it.7

What has not been shown

No randomised trial of cannabis for anxiety or sleep in people with cancer is recorded here.

This applies to: anxiety and sleep in people with cancer.

Nobody has run a search to establish this, so it is the register’s reading of what exists rather than a gap somebody has looked for and confirmed.

What to ask

Is there something for sleep or anxiety that has been tested in people having the treatment I am having?

Treating the tumour Grade D

No human evidence of benefit, or none that survived testing.

Harm has been recorded in people for this claim too; the grade says nothing about it.

The lab results people share come from levels no human body can reach, by smoking, eating or any other route.

The laboratory concentrations are not reachable by any route a person can use.

The harm

Dried cannabis can carry Aspergillus. A published case describes a man who died of disseminated aspergillosis after a marrow transplant, where the fungus cultured from his cannabis matched the organism from his lung biopsy. That is proof by culture matching rather than by coincidence of timing.8

Further cases are recorded in people with leukaemia and with bowel cancer, and cannabis smoke has been reported to impair the lung cells that clear inhaled organisms.2

One correction matters here, because the obvious advice is wrong. Vaporising avoids burning but does not guarantee sterilisation, and heating does not reliably remove fungal contamination. Guidance for people at risk of aspergillosis is that those who are immunocompromised should get individual specialist advice.3

How the grade for treating the tumour was reached

The claim

The claim is that cannabis, or CBD specifically, kills cancer cells. It is supported by real laboratory images shared very widely.

What it does

Cannabinoids do kill cancer cells in a dish. The question this book asks of every such finding is what concentration was needed, and whether a person can reach it. For cannabinoids the answer is that they cannot, by any route.7

The evidence

No trial has tested whether cannabis changes what a tumour does in a person.7

There is a finding pointing the other way that deserves more attention than it gets. A retrospective study of 140 people on nivolumab found a response rate of 37.5 percent in those not using cannabis against 15.9 percent in those who were, with cannabis the only significant factor for reduced response in the multivariate analysis. It was not associated with progression-free or overall survival.4

A prospective study from the same group reported a similar direction and was later corrected. A published critique of that work reports unverifiable statistics and arithmetic errors, and notes that the cannabis group was smaller, more often on second-line treatment, in poorer condition and more often lymphopenic at baseline. The register records the finding and the critique together, because that is the honest state of it.56

The studies this rests on

  • Cannabis and tumour response rate to nivolumab4

    An observational study, measured a clinical outcome, 140 participants.

    lower response rate among cannabis users; not associated with progression-free or overall survival.

    odds ratio 3.13, interval 1.24 to 8.1. reduced response rate, P=0.016

What has not been shown

No trial has tested whether cannabis or CBD changes what a cancer does in a person. The laboratory work is the whole of the evidence for the tumour claim.

This applies to: any tumour type, any endpoint.

Nobody has run a search to establish this, so it is the register’s reading of what exists rather than a gap somebody has looked for and confirmed.

What has never been measured

The concentrations used in the laboratory work are not recorded in this register yet, and neither is the concentration a person reaches. The gap between them is the argument, and it cannot be shown until both are cited.

This applies to: against the laboratory concentrations.

Nobody has run a search to establish this, so it is the register’s reading of what exists rather than a gap somebody has looked for and confirmed.

What to ask

Does using cannabis change how well my immunotherapy is likely to work?

Recorded against a cancer treatment

A retrospective study reported a lower response rate to one immunotherapy drug among people also using cannabis. The finding is contested and the critique of it is on this page. It is recorded because it points the opposite way to what most readers expect, and because immunotherapy is a treatment class rather than a single drug.46

Cannabis alongside checkpoint immunotherapy

A recorded interaction with Checkpoint immunotherapy.

Is there any reason my immunotherapy and cannabis should not be running at the same time?

Every recorded interaction, and how to turn it into a request for your pharmacist.

Also recorded for this entry

Does it reach the body?

The register records no human pharmacokinetic measurement for cannabinoids against the concentrations used in the laboratory work. The chapter note records that the laboratory concentrations themselves are not yet sourced, so neither side of the comparison can be shown here. What that question means, for every compound here.

Trials

The licensed cannabinoid medicines have their own entry. This entry is about the plant, which is not the subject of recruiting cancer trials.

References

  1. Sutton S, Lum BL, Torti FM. Possible risk of invasive pulmonary aspergillosis with marijuana use during chemotherapy for small cell lung cancer. Drug Intelligence and Clinical Pharmacy 1986;20:289-291.

    PubMed 3009125

    How this was reached: taken from the book's notes, chapter 19, note 3.

  2. Cescon DW, Page AV, Richardson S, Moore MJ, Boerner S, Gold WL. Invasive pulmonary aspergillosis associated with marijuana use in a man with colorectal cancer. Journal of Clinical Oncology 2008;26(13):2214-2215; and three related reports.

    No persistent identifier recorded.

    How this was reached: taken from the book's notes, chapter 19, note 4. The author names are abbreviated in the source that was read, and have not been filled in here.

  3. Guidance for people with or at risk of aspergillosis, National Aspergillosis Centre patient information service, updated 2026.

    No persistent identifier recorded.

    How this was reached: taken from the book's notes, chapter 19, note 5.

  4. Taha T, Meiri D, Talhamy S, Wollner M, Peer A, Bar-Sela G. Cannabis impacts tumor response rate to nivolumab in patients with advanced malignancies. The Oncologist 2019;24(4):549-554.

    doi:10.1634/theoncologist.2018-0383

    How this was reached: taken from the book's notes, chapter 19, note 6.

  5. Bar-Sela G, Cohen I, Campisi-Pinto S, et al. Cannabis consumption used by cancer patients during immunotherapy correlates with poor clinical outcome. Cancers 2020;12(9):2447, subsequently the subject of a published correction in Cancers 2022;14.

    No persistent identifier recorded.

    How this was reached: taken from the book's notes, chapter 19, note 7. The author names are abbreviated in the source that was read, and have not been filled in here.

  6. Piper BJ, Tian M, Saini P, et al. Immunotherapy and cannabis: a harmful drug interaction or reefer madness? Cancers 2024;16:1245.

    No persistent identifier recorded.

    How this was reached: taken from the book's notes, chapter 19, note 8. The author names are abbreviated in the source that was read, and have not been filled in here.

  7. The general position of the clinical literature on cancer pain, appetite, sleep, cannabinoid tumour concentrations and the CBD interaction profile.

    No persistent identifier recorded.

    How this was reached: taken from the book's notes, chapter 19, note 10.

  8. Fatal aspergillosis associated with smoking contaminated marijuana in a marrow transplant recipient.

    PubMed 3293934

    How this was reached: taken from the book's notes, chapter 19, note 3.