The Scientists Who Discovered Cannabis Could Kill Cancer Cells Explain Why That Doesn't Mean Cannabis Cures Cancer
Anticancer mechanisms of cannabinoids
Bottom Line
The same research team that pioneered cannabinoid anticancer research (Guzman 2000, 2003, 2006) published a comprehensive review documenting strong preclinical evidence for cannabinoid anticancer effects — while explicitly warning that this evidence does not support using cannabis as a cancer treatment outside of clinical trials.
Why It Matters
This review carries unique authority because the authors are the same team that pioneered cannabinoid anticancer research — the group behind the landmark 2000 Nature Medicine study, the 2003 Nature Reviews Cancer review, and the 2006 human pilot trial. When the researchers who discovered these effects say "the evidence isn't sufficient for clinical use," it carries more weight than the same statement from any outside observer. The paper serves as both a comprehensive scientific resource and a responsible public health message.
The Backstory
In 2016, three scientists at Complutense University of Madrid published a review paper in Current Oncology titled "Anticancer mechanisms of cannabinoids." If you read only the title, you'd think it was another entry in the growing list of papers documenting how cannabinoids kill cancer cells.
But this paper carried a message that no title could capture: the researchers who had spent two decades building the scientific case for cannabinoid anticancer effects were now, very deliberately, telling the world to slow down.
Guillermo Velasco, Cristina Sanchez, and Manuel Guzman are not outside skeptics. They are the group. Guzman led the 2000 Nature Medicine study that first showed THC could shrink brain tumors in rats. He conducted the 2006 human pilot trial. He wrote the 2003 Nature Reviews Cancer paper that put cannabinoid anticancer research on the map. Velasco and Sanchez have been his collaborators through all of it.
When these researchers say the evidence does not support using cannabis as a cancer treatment, they are not dismissing the science. They are defending it from its own fans.
What the Evidence Actually Shows
The review documents genuine, reproducible anticancer effects across multiple cancer types in preclinical models. This is not disputed science. The mechanisms are well-characterized:
This evidence is real. It is reproduced across independent laboratories and multiple tumor types. It involves defined molecular pathways, not hand-waving. By any standard of preclinical science, the cannabinoid anticancer hypothesis has earned serious investigation.
What the Evidence Does Not Show
Here is where the review's importance lies. The authors document, with equal clarity, the reasons this preclinical evidence cannot be extrapolated to a clinical recommendation:
One human trial exists. As of 2016, the only human data came from Guzman's own 2006 pilot — nine glioblastoma patients receiving intracranial THC. That's it. Nine patients, no control group, in one of the rarest brain cancers. Every other data point is from petri dishes and animals.
Resistance mechanisms exist. Some cancer cells overexpress a protein called midkine (MDK), which activates ALK signaling and blocks the cannabinoid-induced autophagy that leads to cell death. Not all cancer cells die when exposed to cannabinoids — even in a petri dish. Claims of universal anticancer efficacy are contradicted by the group's own data.
Dose matters in dangerous ways. At submicromolar concentrations, THC can paradoxically promote cancer cell proliferation through ADAM17/EGFR transactivation. The dose that kills cancer cells and the dose that feeds them are separated by a narrow window. Smoking cannabis does not deliver controlled doses to tumor tissue.
Immunity is complicated. THC can suppress anti-tumor immune responses by shifting Th1 to Th2 cytokine profiles. This means cannabinoids might simultaneously kill cancer cells directly while weakening the immune system's ability to kill them. The net effect in a living person — with a complex immune system interacting with a genetically heterogeneous tumor — is unknown.
The Misinformation Problem
Myth vs. Reality
The researchers who study cannabinoids and cancer believe cannabis cures cancer.
The researchers who pioneered cannabinoid anticancer research — Guzman, Velasco, Sanchez — have consistently and explicitly stated that the evidence does not support using cannabis as a cancer treatment. They advocate for rigorous clinical trials. They warn against replacing proven treatments. Their caution is all the more significant because they have more professional incentive than anyone to overstate their findings.
The Evidence
A 2019 analysis found cannabis cancer cure stories received 4.26 million social media engagements versus 36,000 for accurate debunking. Online searches for 'cannabis cancer' grew at 10x the rate of standard therapy searches. Patients who abandon conventional treatment for alternatives face 2-5x higher mortality risk.
Velasco et al. (2016); Lien & Bhatt (2019), Cureus
The review was published at a critical moment. By 2016, the "cannabis cures cancer" narrative had reached mass saturation on social media. Facebook posts claiming cannabis killed cancer were shared millions of times. YouTube videos featuring individual testimonials — survivorship bias presented as evidence — accumulated tens of millions of views. Patients were delaying or abandoning proven treatments.
The misinformation was not built on nothing. It was built on studies like Guzman's 2000 paper. Real science, published in real journals, demonstrating real anticancer effects — then stripped of every caveat, every limitation, every "in rats" and "in a petri dish," and presented as if smoking a joint could cure your tumor.
The Velasco 2016 review was, in part, an attempt by the researchers whose work was being weaponized to set the record straight.
The Gap Between Lab and Clinic
Why does something that works in a dish fail in a person? The review addresses this directly:
The Responsible Position
The authors conclude by advocating for rigorous clinical trials — not for self-treatment, not for replacing chemotherapy with cannabis oil, not for the unfounded optimism that characterizes most public discourse on this topic. Their position:
“These observations have already contributed to the foundation for the development of the first clinical studies that will analyze the safety and potential clinical benefit of cannabinoids as anticancer agents.”
— Velasco, Sanchez & Guzman
Department of Biochemistry and Molecular Biology, Complutense University of Madrid
The authors advocating for clinical trials — not self-medication — as the appropriate next step for their own research program
This is science working as it should. Researchers generate a hypothesis, test it in models, publish findings with appropriate caveats, and advocate for the next level of evidence. What makes this review necessary is that the public conversation skipped every step between "cannabinoids kill cancer cells in a dish" and "cannabis cures cancer" — and the researchers felt compelled to intervene.
Since this review was published, the Twelves 2021 trial provided the first randomized clinical data: 83% versus 44% one-year survival for THC+CBD plus chemotherapy in 21 glioblastoma patients. The ARISTOCRAT Phase II trial is now recruiting 230+ patients. The clinical evidence is growing — slowly, rigorously, the way it should. But it is still preliminary, and the fundamental message of this review has not changed.
Key Takeaways
Cite this study
Velasco, Guillermo; Sánchez, Cristina; Guzmán, Manuel. (2016). Anticancer mechanisms of cannabinoids. Current Oncology, 23(Suppl 2), S23-S32. https://doi.org/10.3747/co.23.3080