An oral combination of tetrahydrocannabinol and cannabidiol reduced agitation more than placebo in hospice-eligible people with dementia, according to topline results of the phase 2 LiBBY trial presented on 14 July 2026 at the Alzheimer's Association International Conference in London. On the primary measure, change in agitation on the Cohen-Mansfield Agitation Inventory at two weeks, the treated group showed a 6.27-point greater reduction than placebo, the investigators' institutions reported. LiBBY enrolled 120 participants with Alzheimer's disease or other dementias who were hospice-eligible and had clinically significant agitation; ten medical centres conducted study visits in participants' homes or places of residence. These are conference results, preliminary until a full peer-reviewed report appears.
The registration, NCT05644262, sets out what was actually tested. It describes a multicentre, randomised, double-blind, placebo-controlled phase 2 study of about 120 people aged 40 or over who are hospice care-eligible with agitation and dementia, randomised 50:50 to the study drug T2:C100 or placebo, with a 12-week double-blind period and an optional 12-week open-label extension. Entry criteria include current enrolment in outpatient or inpatient hospice care, stage 6d on the Functional Assessment Staging Test, a score of 12 or more on the Advanced Dementia Prognostic Tool as implemented by the Mitchell Index, and agreement not to use other cannabinoids. The trial began on 18 December 2023. Per the conference release, participants took 2 mg THC and 100 mg CBD twice daily through week one, then 4 mg THC and 200 mg CBD twice daily from weeks two to 12, in a digestible oil. The trial was funded mainly by the National Institute on Aging through the Alzheimer's Clinical Trial Consortium, with Alzheimer's Association support for the extension; the formulation was supplied by MediPharm Labs.
On a key secondary measure, clinician-rated global change in behaviour, 83.9 per cent of treated participants were rated improved at two weeks against 30.5 per cent on placebo, and 87.2 per cent against 23.6 per cent at 12 weeks. Overall adverse event rates were similar (46.7 per cent versus 42.4 per cent), while serious adverse events were more frequent in the treatment arm (23.3 per cent versus 11.9 per cent); investigators said none of the serious events were judged related to the study medication. Commenting on the presentation, Nathan Herrmann of the University of Toronto told Medscape Medical News that larger studies are needed to define safety and to assess outcomes such as falls and cognition, adding that "the true risks and benefits of cannabinoids in patients with dementia still require clarification".
For families and hospice teams, the supply question is unresolved. T2:C100 is an investigational formulation made for the trial and is not an approved medicine, so it cannot be prescribed or stocked outside research; when the study launched, its investigators noted there are no medicines approved by the US Food and Drug Administration for agitation at the end of life in dementia. MediPharm Labs said it is reviewing the full results and evaluating possible regulatory activities, and the MUSC-led team said it plans a larger study in a broader population. Co-lead investigator Brigid Reynolds of Georgetown University cautioned that "people should not assume that products available at dispensaries or online are equivalent" to the supervised formulation studied.