Many people receiving cancer treatment use marijuana to manage symptoms, often making that decision outside a formal prescription. A new study from University of Kentucky researchers offers a closer look at why radiation therapy patients use cannabis and what they believe it does for them. Among the users surveyed, at least 80% reported improvements in symptoms and quality of life. Pain, nausea, anxiety, depression, poor appetite and sleep disturbance were among the reasons they turned to it.
Published September 22, 2026, in The Oncologist, the research surveyed 237 adults at an urban radiation oncology practice and a rural satellite clinic in Kentucky. Its finding is an important account of patients’ experiences during cancer care, especially in a region affected by high levels of opioid use. It is also a report of perceived benefit, rather than a clinical trial proving that marijuana caused the improvements. That distinction matters when patients and clinicians decide how medicinal cannabis fits alongside established symptom treatments.
What the Kentucky Researchers Studied
The team used an anonymous, tablet-based questionnaire completed without research staff overseeing patients’ answers. Adults receiving radiation therapy at the two oncology locations filled it out between July 2023 and February 2024. The survey asked about cannabis use before and after a cancer diagnosis, reasons for using it, symptoms, overall quality of life and differences between urban and rural participants. Researchers also set out to examine cannabis use alongside opioids and gabapentin, medicines commonly relevant to difficult cancer-related pain.
The design captured a range of patients in active radiation treatment instead of relying only on people enrolled in a cannabis program. It therefore offers insight into self-directed use that might otherwise be missed in medical records. Yet a questionnaire taken at one point in time cannot show how each person’s symptoms changed under a controlled treatment plan. Participants chose whether to use cannabis, which products to obtain and how to describe their experience. The researchers did not assign standardized THC or CBD doses, follow a placebo group or verify a drug’s effect on a tumor.
Cannabis Use Was Common, but the Percentages Need Context
Of the 237 respondents, 121—about 51%—said they had used cannabis medicinally without a prescription at some point in their lives. That lifetime figure should not be read as the share currently using marijuana during cancer treatment. Thirty-five participants, or 14.7% of the full survey group, reported recent use after diagnosis. Thirty, or 12.7%, reported recent use before diagnosis. Although the later figure was numerically higher, the difference did not meet the researchers’ threshold for statistical significance.
Those denominators are central to interpreting the headline result. “At least 80%” refers to improvement reported among cannabis users responding to the relevant survey questions; it does not mean that 80% of all 237 radiation patients improved with marijuana. Nor does the 51% lifetime-use figure mean half the group began using cannabis because of cancer. The study instead shows that past experience was widespread, a smaller group reported recent use, and patients who used cannabis for symptom relief often described a favorable experience. Keeping those groups distinct prevents a promising survey finding from becoming a misleading treatment claim.
Pain and Nausea Were Part of a Wider Symptom Burden
Pain and nausea featured prominently among the reasons participants used cannabis. For someone undergoing cancer treatment, either problem may have several sources: the disease itself, procedures, systemic treatments or other conditions. Persistent pain can disrupt movement and sleep; nausea can make eating and taking medication harder. The researchers also identified anxiety, depression, appetite loss and sleep disturbance among common motivations. Patients were responding to combinations of physical and emotional problems, not necessarily seeking relief from a single isolated symptom.
This breadth helps explain cannabis’s appeal in oncology, where symptoms often overlap. A person may report that sleep improves after pain eases, or that feeling less nauseated makes a meal possible. But the survey did not establish which cannabis formulation produced a given benefit, whether a particular symptom improved by a clinically measured amount, or how its performance compared with standard medicines. A report that patients used marijuana for pain and nausea is evidence of a treatment need and perceived response. It is not a head-to-head result showing marijuana works better than an antiemetic, analgesic or other supportive-care option.
What the Quality-of-Life Finding Means
At least four in five surveyed cannabis users said their symptoms and quality of life improved. Quality of life is particularly meaningful in cancer care because surviving treatment does not fully describe how a person is functioning day to day. The ability to rest, maintain appetite, spend time with family and tolerate ongoing therapy can matter as much to patients as an individual symptom rating. The Kentucky findings give those patient-reported experiences a place in the research discussion, including experiences with cannabis obtained and used without a conventional prescription.
The language of improvement still needs careful reading. Respondents recalled and evaluated their own experiences; investigators did not measure an objective before-and-after change caused by a controlled cannabis intervention. Expectations, changes in cancer therapy, other medicines and the ordinary fluctuation of symptoms may all influence how someone feels. People who found cannabis unhelpful may also have stopped using it, while those who felt benefit may have been more inclined to report continued use. The study’s signal is substantial within its users, but its design cannot determine how much of that signal cannabis itself produced.
An Urban-Rural Difference Raises New Questions
Recent cannabis users in the study were more likely to live in the urban setting than in the rural one. That difference is notable because patients in both locations received radiation oncology care through the same broader cancer-center system. Access to products, travel distance, local norms, cost, legal understanding and willingness to disclose use could all contribute to an urban-rural gap. The survey identifies the pattern, but it does not establish which of those factors caused it or whether people in the two clinics would respond differently to the same product.
The Kentucky setting also matters to the research question about pain medicines. In an area where opioid use has long been a public-health concern, patients and clinicians have reason to ask whether cannabis is being used with opioids or gabapentin, and whether it changes symptoms or medication needs. This study was designed to investigate such overlap, but the headline findings do not establish an opioid-sparing effect. It would be inaccurate to conclude from these survey results that patients reduced opioid doses, that combining medicines was safe for everyone or that cannabis can replace established cancer pain care. Those questions require more direct measurement.
How the Findings Fit With Oncology Guidance
The American Society of Clinical Oncology’s guideline recognizes that cannabis use among people with cancer has moved faster than the evidence needed to guide it. Its clearest supportive-care signal concerns nausea and vomiting from chemotherapy that persist despite appropriate standard antiemetic treatment; cannabinoids may help as an additional option in that setting. The guideline says the evidence remains uncertain for many other supportive-care outcomes and recommends against using cannabis as a treatment directed at the cancer itself outside a clinical trial. That assessment provides a useful framework for the Kentucky survey’s encouraging but uncontrolled reports.
The distinction between a cannabis product and an approved cannabinoid medicine is also important. Dronabinol and nabilone are FDA-approved for chemotherapy-related nausea and vomiting in patients who have not responded to standard antiemetic therapy. They have defined drug ingredients and dosing. Dispensary flower, edibles and oils can differ greatly in THC and CBD content, onset and duration. The Kentucky patients’ experiences with self-selected cannabis cannot be treated as a trial of those approved medicines, and findings for a standardized oral cannabinoid cannot automatically be extended to every retail marijuana product.
Safety Still Belongs in the Cancer-Care Conversation
Relief and risk can coexist. Cannabis and cannabinoid products may cause dizziness, drowsiness, changes in heart rate or blood pressure, and unwanted effects on thinking or mood. Those effects may matter more when a patient already has fatigue, dehydration or a risk of falling. They also complicate the use of other medicines that can cause sedation, including some pain and sleep drugs. Product potency and route of use shape the experience: an edible can take longer to have an effect and last longer than inhaled cannabis, making unintended overconsumption easier when a person expects immediate relief.
Cancer treatment adds questions that a general consumer survey cannot answer. Patients may receive several drugs at once, and researchers are still investigating how cannabis might interact with particular therapies, including immunotherapy. Smoking also exposes the lungs to combustion products, a concern that deserves special attention in someone with respiratory disease or treatment-related vulnerability. These possibilities do not erase the symptom relief reported by the Kentucky participants. They explain why the most useful clinical response is an open discussion about the exact product, dose, timing, other medicines and the symptom the patient hopes to manage.
What Would Stronger Evidence Look Like?
The next studies need to separate a patient’s reason for using cannabis from a demonstrated treatment effect. For nausea, investigators could compare a defined cannabinoid product with a placebo or another addition to standard antiemetic therapy, while measuring vomiting episodes and patient-reported distress. Pain trials would need to account for cancer type, cause of pain, existing analgesics and whether participants can reduce medication without losing control of symptoms. Quality of life should be measured over time with validated instruments rather than inferred from a single retrospective impression.
Researchers also need to identify who benefits and who does not. A balanced study would report adverse effects, discontinuation, product composition and treatment interactions alongside improvements. It would include patients from different geographic settings and follow them long enough to see whether early gains last. The Kentucky team’s anonymous tablet survey shows one workable way to hear directly from patients in both urban and rural oncology clinics. Its strongest contribution may be pointing researchers toward the symptoms and access questions that deserve rigorous testing next.
The Practical Meaning of the New Study
For patients, the results validate a real question: when pain, nausea, appetite problems or poor sleep continue despite treatment, could a carefully chosen cannabis-based option help? Many respondents who used marijuana believed that it did. Their reports deserve attention, especially when symptom burden is affecting daily life. At the same time, a favorable experience in a survey cannot predict whether another patient will benefit, which product would be appropriate or whether it can be combined safely with their current cancer regimen.
For clinicians, the finding is a reason to ask about cannabis plainly and without judgment. A patient using it outside the formal treatment plan may otherwise omit information that helps explain sedation, persistent symptoms or a possible drug interaction. The new study adds a useful view from Kentucky radiation clinics: cannabis use was common across the surveyed group’s histories, recent use varied by setting, and most users who assessed their response perceived improvement. It strengthens the case for better conversations and better trials, while leaving the question of proven effectiveness to studies designed to answer it.
