Did the Government Confirm Cannabis Cures Cancer? What NCI Said

No. In 2013 a press release from a medical marijuana advocacy group claimed that the US government had "confirmed" cannabis prevents or cures certain cancers. What the National Cancer Institute (NCI) had actually published was a review of research, and the most striking findings in it came from cells in lab dishes and from animals, not from people with cancer. NCI's current guidance is clear: the US Food and Drug Administration has not approved cannabis or any cannabinoid as a treatment for cancer.

That does not make the original research meaningless. It does mean the headline went much further than the evidence. Below is what the 2013 claim was based on, what the science shows today, and where cannabis-based medicines genuinely help people living with cancer.

Where the "government confirms" claim came from

The story started as a release distributed through GlobeNewswire in March 2013 by Advocates for the Disabled and Seriously Ill, a patient group associated with the Los Angeles lawyer Matthew Pappas. Pappas was representing patients in disputes with cities that were trying to close medical marijuana dispensaries, and the release was written to support that fight.

Its key move was to quote NCI's Physician Data Query (PDQ) summary on cannabis. The PDQ is a regularly updated digest of published studies, written for doctors and patients. It summarises research; it is not a government ruling that a treatment works. By describing that summary as a federal report confirming cancer-fighting effects, the release turned a literature review into an endorsement it never contained.

Pappas framed the federal stance as contradictory. In the release he called the government's position on patients using doctor-recommended cannabis "hypocritical considering the benefits reported by its own National Cancer Institute." That was a legal and political argument, and it is fair to report it as such. It was not a medical finding.

The release also mentioned Bill Rosendahl, a Los Angeles city councilmember who had been diagnosed with stage four cancer in 2012 and who reportedly spoke out against a citywide dispensary ban. Rosendahl later saw his cancer go into remission in 2013; it returned, and he died in March 2016.

What the laboratory studies actually showed

The PDQ summary does describe a long list of preclinical results, and they are genuinely interesting to researchers:

• Breast cancer cells: cannabidiol (CBD) triggered programmed cell death in breast cancer cell lines, including types that respond to estrogen and types that do not, while having little effect on normal breast cells.
• Lung cancer: THC and related compounds slowed the growth of lung cancer cells in culture and in mice; in one mouse model THC cut tumour growth by around 60%.
• Colon cancer: in mice exposed to a chemical that causes colon tumours, CBD reduced the number of precancerous and cancerous lesions.
• Brain tumours: cannabinoids caused glioma cells to die in culture and shrank gliomas in rats and mice.

These are the findings the 2013 release leaned on. The problem is the leap from a dish or a mouse to a patient. Most compounds that kill cancer cells in the lab never become useful medicines, because the doses do not translate, the drug does not reach the tumour, the side effects are too severe, or the effect simply disappears in a whole human body. The same NCI summary also reports a less convenient result: in mice with normal immune systems, THC suppressed immunity and in some models made tumours grow faster.

The human evidence on cannabis and cancer

Studies in people are few and small. The best known is a Spanish pilot study published in 2006, in which THC was delivered directly into the tumours of nine patients with recurrent glioblastoma. The procedure appeared safe and there were signs of reduced tumour cell growth in some samples, but the study had no control group and could not show that anyone lived longer. More recently, small trials have tested a cannabis extract spray alongside the chemotherapy drug temozolomide for recurrent glioblastoma. These are early-stage studies, not proof of a cure.

NCI's patient summary, last reviewed in 2023, notes that no ongoing clinical trials of cannabis as a cancer treatment in people were found in its searches. Larger reviews of medical cannabis reach the same broad conclusion: the evidence does not support using cannabis to treat cancer itself.

This matters in a very practical way. People who have delayed or replaced proven treatment with cannabis oil have, in documented cases, lost the window in which their cancer could be treated. If you or someone close to you is considering cannabis during cancer care, talk to a doctor or the oncology team first, partly because cannabinoids can interact with other medicines.

Smoking cannabis and lung cancer risk

The 2013 release also cited studies that found no clear link between smoking marijuana and lung cancer, including a Los Angeles case-control study of more than 600 lung cancer patients. Those studies were real, and that research did not show the strong association seen with tobacco.

But "no clear link found" is not the same as "safe". Cannabis smoke contains many of the same harmful substances as tobacco smoke. NCI's summary describes a cohort study in which men who smoked both tobacco and cannabis had a higher lung cancer risk, while a population study found that light cannabis use was not linked to higher risk. Heavy, long-term smoking of anything burned remains hard on the lungs, and researchers still regard the question as open.

Where cannabinoids do help people with cancer

The more solid story is about symptoms, not cures. Two drugs based on cannabinoids, dronabinol and nabilone, are FDA-approved for nausea and vomiting caused by chemotherapy when standard anti-nausea drugs have not worked. In trials they performed about as well as, or better than, older alternatives, though they can cause dizziness and drowsiness.

For other symptoms the results are mixed:

• Pain: some trials found that inhaled cannabis or a cannabinoid mouth spray eased nerve pain or cancer pain for some patients; others found little added benefit over standard painkillers.
• Appetite: in people with advanced cancer, dronabinol did not improve appetite or weight better than an established drug, megestrol.
• Sleep: a few small studies reported better sleep, while a 144-patient trial of oral CBD found no benefit over placebo for cancer symptoms overall.

The 2013 release also cited insomnia and claimed cannabis compared favourably with sleeping pills. Some patients do report better sleep, but the evidence is limited, and cannabis carries its own side effects: impaired memory and coordination, dependence in some users, and, at high doses or in vulnerable people, anxiety or psychosis. It is also unsafe to drive while impaired.

The legal and political backdrop

Much of the 2013 story was about politics. At the time, the White House Office of National Drug Control Policy argued that marijuana had no accepted medical use, while states such as California allowed it for patients. Advocates like Pappas saw the NCI research as proof that federal policy ignored science.

A lot has changed since. Most US states now have medical cannabis programmes, and since 2014 a budget amendment has barred the Justice Department from spending money to prosecute people following state medical cannabis laws. Under federal law, however, cannabis has remained a Schedule I substance for most of that period, and a process to move it to a less restrictive schedule began in 2024. Meanwhile, the FDA has approved a purified CBD medicine for two rare forms of epilepsy, showing that cannabinoids can become mainstream drugs when trials support them.

Common myths, checked

"The government admitted cannabis cures cancer." False. NCI summarised lab and animal research; it has never endorsed cannabis as a cancer treatment.

"If it kills cancer cells in a dish, it works in people." Not necessarily. Bleach kills cancer cells in a dish too. Human trials are the only way to know whether a treatment helps patients.

"Cannabis is completely harmless." No medicine is. Cannabis does not cause fatal overdose in the way opioids can, but it has real side effects and drug interactions.

"Cannabinoids have no medical use at all." Also false. Approved cannabinoid drugs help with chemotherapy nausea, and research on pain and other symptoms continues.

The fair summary is that cannabis is a promising subject for cancer research and a useful tool for some symptoms, but it is not a cancer cure, and nothing NCI published in 2013 said otherwise.

Sources: National Cancer Institute, Cannabis and Cannabinoids (PDQ, patient version); National Cancer Institute, Cannabis and Cannabinoids (PDQ, health professional version); Wikipedia, Medical cannabis; Wikipedia, Bill Rosendahl; original press release by Advocates for the Disabled and Seriously Ill via GlobeNewswire (March 2013).

Related:

• Cannabis Oil and Cancer: Why Universal Cure Claims Are Unsupported