Cannabinoids Reduce Tumors in Glioblastoma, Review Finds

A systematic review and meta-analysis published on May 14, 2026 in the journal Pharmaceuticals found statistically significant antitumor effects for cannabinoids in animal models of glioblastoma and breast cancer — providing what the authors describe as the most comprehensive preclinical synthesis of the topic to date.

The study¹, led by researchers from Semmelweis University in Budapest and the University of Pécs in Hungary, screened more than 27,000 records from PubMed, Embase, and the Cochrane Central Register, identifying 189 eligible studies and including 52 in the quantitative meta-analysis. It covered cannabinoid effects across glioblastoma, breast, lung, prostate, colon, pancreatic, and liver cancer models, pooling both cell-line (in vitro) and animal (in vivo) experiments.

The strongest result, and the one with the most reproducible pattern across studies, was glioblastoma. In animal xenograft models — experiments in which human tumor cells are implanted in mice — cannabinoid monotherapy reduced tumor volume by an average of 980.58 cubic millimeters compared to vehicle controls, a statistically significant finding that held across cannabinoid types, including CBD-rich preparations, THC-containing formulations, and THC:CBD combinations. The review’s authors describe glioblastoma as “the tumor type with the most reproducible evidence base.” In cell-line experiments, cannabinoids also significantly reduced glioblastoma cell viability, though with wide variability across studies.

What the numbers show across tumor types

In breast cancer animal models, cannabinoid monotherapy produced a statistically significant average tumor volume reduction of 402.64 mm³. The effect grew substantially when CBD was paired with the chemotherapy drug doxorubicin: that combination reduced tumor volume by 1,295 mm³ compared to doxorubicin alone — the largest single effect the review identified across all combination-therapy experiments, and the finding the authors cite as the clearest case for cannabinoids as chemotherapy adjuncts.

Lung and prostate cancer data showed significant reductions in specific contexts. Lung cancer animal models produced a pooled tumor volume reduction of 562 mm³ against vehicle controls. In prostate cancer, THC:CBD combinations achieved a significant reduction of 1,136 mm³, though CBD alone produced no significant effect in that tumor type, and some prostate combination data trended antagonistically.

The weakest findings came from digestive cancers. Colon cancer models showed a trend toward tumor reduction that did not reach statistical significance, with wide variability across studies. Pancreatic and hepatocellular carcinoma data showed inconsistent or null responses. That divergence matters for reading these results accurately: the glioblastoma and breast cancer signals do not extend to these tumor types on this evidence base.

CBD’s consistent profile and THC’s variability

Across tumor types, CBD showed the most consistent antitumor profile. The authors describe its activity as broad and relatively reliable across cancer models, with the additional practical advantage of a clinical safety record from approved indications — making it relatively easier to move toward combination trials than THC or synthetic compounds. Earlier reviews have looked at similar questions from different angles: MyCannabis covered CBD’s anticancer activity across canine tumor models, and a broader look at what current science says about CBD and cancer cells is worth reading alongside this review.

THC showed a more complicated picture across tumor types. In glioblastoma, THC-containing formulations reduced tumor volume significantly. In breast cancer models, THC-rich products alone trended toward tumor growth rather than reduction, though adding THC to chemotherapy substantially reversed that pattern. In colon cancer, THC trended toward increased tumor volume. The variability fits with mechanistic research showing that THC’s activity at cannabinoid receptors can generate immunosuppressive signaling that sometimes offsets its direct cytotoxic effects, in ways that appear tumor-type and context dependent.

The review also references early clinical data on glioblastoma. A phase Ib randomized trial — cited by both the review’s authors and in American Society of Clinical Oncology guidance — found that a pharmaceutical 1:1 THC:CBD preparation combined with temozolomide produced higher one-year survival rates than placebo in 21 patients with recurrent glioblastoma. The trial was not sized to confirm efficacy and the difference in progression-free survival at six months was not significant. But it offers directional clinical support for glioblastoma as the most compelling tumor type for further cannabinoid research.

The preclinical-to-clinical gap

Every data point in this analysis comes from cell cultures or animal models. The animal experiments relied almost entirely on xenograft designs, which by definition use immunocompromised animals and cannot fully replicate how the immune system interacts with tumors in humans. That limitation matters particularly for cannabinoids, which interact with the endocannabinoid system in immune cells throughout the body.

Risk of bias was rated predominantly high across included studies — a finding the authors attribute largely to incomplete reporting of randomization and blinding procedures, rather than confirmed methodological failures. It’s a meaningful distinction, but it doesn’t resolve the concern. Heterogeneity across the pooled analyses ranged from 34% to 100%, meaning individual study results varied widely. The authors are explicit: these findings should be treated as hypothesis-generating, not as evidence of uniform efficacy.

The review calls for biomarker-driven trial design — profiling cannabinoid receptor expression and downstream signaling in tumors to identify which patients and cancer subtypes are most likely to respond — as the key step toward translating the preclinical signals into clinical knowledge. A companion review in Current Oncology Reports by the same research group, published in 2025, found that cannabinoids reduce chemotherapy-related nausea and pain in cancer patients, a more established clinical picture that could inform how adjunct trials are structured.

The glioblastoma data, by this review’s account, is the most consistent in the preclinical cannabinoid oncology literature. Whether it translates into survival benefit for patients is a question that requires trials at a scale that does not yet exist.

References:

^{1. Creangă-Murariu, I., Rezuș, I.-I., Karami, R., Makolli, A., Chifu, C., Rancz, A., Sipos, Z., Ferdinandy, P., Papp, R., Teutsch, B., Tamba, B.-I., Hegyi, P., & Bunduc, S. (2026). Antitumor Activity of Cannabinoids and Their Interaction with Chemotherapy: A Systematic Review and Meta-Analysis of Preclinical Evidence. Pharmaceuticals, 19(5), 768. https://doi.org/10.3390/ph19050768}