THC/CBD in Advanced Cancer: Pain Improvement Did Not Reduce Overall Symptom Burden
A placebo-controlled advanced-cancer trial found a small pain benefit from 1:1 THC/CBD oil, but no significant improvement in its primary measure of overall symptom burden. It also found greater toxicity and more withdrawals in the cannabis arm.
The primary question covered more than pain
The trial randomized 144 patients receiving palliative care to THC/CBD oil or matched placebo, with dose adjustment according to tolerability and perceived benefit. The active formulation contained 10 mg of each cannabinoid per milliliter. Treatment included two weeks of dose selection and two further weeks of data collection.
The main endpoint was the total symptom distress score from the Edmonton Symptom Assessment System at day 14. That total combines symptoms such as pain, nausea, tiredness, drowsiness, anxiety and wellbeing. Improvement in one item need not produce an advantage in the total.
Both groups improved on the main outcome
The mean total-score change was −6.30 with THC/CBD and −6.98 with placebo, with no significant difference between groups. Baseline-adjusted analysis was also nonsignificant. A clinically meaningful total-score reduction was not more frequent with active treatment.
Pain, a secondary outcome, improved more in the cannabis arm: mean change −1.42 versus −0.46, p=0.04. Conversely, the wellbeing item favored placebo. A balanced account should preserve this mixed pattern rather than convert the pain result into overall symptom superiority.
Attrition affects later impressions of benefit
By day 28, 39 of 72 patients in the cannabis arm and 22 of 72 in the placebo arm had exited. Reasons included deterioration, adverse effects and lack of perceived benefit. Later global-improvement ratings among those remaining may therefore describe a selected group who tolerated or valued treatment.
More active-arm patients were taking opioids at baseline, although doses among opioid users were similar. The investigators performed sensitivity analyses and did not find that this imbalance changed the main conclusion. Nonetheless, concurrent treatment and substantial missing follow-up data remain important context.
What can a treatment claim reasonably say?
The evidence supports a modest secondary pain signal alongside a null primary symptom result and increased adverse effects. It did not demonstrate an opioid-sparing effect. Psychoactive effects could also complicate interpretation of subjective benefit despite matching and blinding procedures.
This was a short trial of a specific oil in advanced cancer. It does not prove that no patient can benefit, or that another formulation must produce the same result. Equally, it does not justify claiming that this combination broadly improves palliative symptoms. Treatment benefit, tolerability and functional effects each need their own evidence.
Related analysis: Equal THC Doses, Different Plasma Profiles: A Human Formulation Trial.
Primary research
Janet R. Hardy and colleagues. Medicinal cannabis for symptom control in advanced cancer: a double-blind, placebo-controlled, randomised clinical trial of 1:1 tetrahydrocannabinol and cannabidiol. Supportive Care in Cancer (2025). DOI: 10.1007/s00520-025-09763-5.
