An oral cannabis-based oil containing cannabidiol (CBD) and Δ9-tetrahydrocannabinol (THC) failed to significantly reduce chronic pain or other nonmotor symptoms in patients with Parkinson disease during a nine-week phase 2 trial, according to findings published in Movement Disorders.
Testing a New Formulation
Lead author Gabriel T. Kubota, MD, PhD, and colleagues at the University of São Paulo Hospital conducted a parallel-group, double-blind study at outpatient clinics in Brazil. The research included adults 25 and older diagnosed with Parkinson disease who reported pain lasting three months or longer.
Of 140 individuals screened, 101 were randomized 1:1 to receive the cannabis-based oil or a matched placebo, with 87 completing the protocol. Participants underwent a three-week titration phase starting at two drops twice daily, guided by tolerability, before reaching a maintenance dose across a six-week period.
The final formulation contained 96 mg/mL CBD and 2.1 mg/mL THC. The primary outcome measured the difference in the Parkinson’s Disease Pain Classification System (PD-PCS) score at the final visit on day 63.
Null Results Across Measures
Researchers observed no significant difference in PD-PCS scores between the groups at the final assessment. The study was halted early after an interim futility analysis indicated results were unlikely to show a meaningful benefit.
Effect sizes at day 21 and day 63 were small and not clinically significant. Secondary pain endpoints, including the Brief Pain Inventory, showed no significant difference between groups, nor did the proportion of participants achieving a 30% reduction in pain intensity.
Across a battery of nonmotor symptom assessments—including the MDS-Non-Motor Rating Scale, autonomic dysfunction scales, and cognitive tests—no between-group differences emerged. Motor outcomes measured with the MDS-Unified Parkinson’s Disease Rating Scale Part III, the New Freezing of Gait Questionnaire, and the Timed Up-and-Go scale also remained unchanged.
Despite the lack of therapeutic effect, the cannabis formulation was generally safe and well-tolerated. Both groups experienced mild adverse events, with irritability occurring more frequently in the CBD/THC group. Two serious adverse events, one case each of Guillain-Barré syndrome and cryptococcal meningitis, were deemed unrelated to treatment.
Related: Aged care news and consultations now available
For patients managing daily discomfort in a disease where options are limited, this result is a stark reminder of how complex pain management can be. It shifts the focus from the promise of cannabis to the reality of clinical trial design, where a high placebo response rate can obscure a true treatment effect, making it difficult to distinguish between a failed drug and a misunderstood therapeutic window.
Context and Limitations
Pain affects an estimated 40% to 85% of patients with Parkinson disease and is among the most disabling and undertreated nonmotor symptoms. Current pharmacologic options are limited; neither duloxetine nor oxycodone-naloxone has demonstrated convincing benefit in controlled trials.
Despite this, cannabinoids have attracted growing clinical interest. Preclinical analgesic data, web-based patient surveys, open-label observational studies, and a phase Ib safety trial established tolerability, but no prior phase 2 or 3 randomized trial had tested a cannabinoid formulation for PD-related pain as the primary end point.
The authors suggest the negative findings may be partially attributable to the low THC dose employed. At 0.96 mg/day, the concentration was substantially lower than formulations associated with analgesic benefit in other pain conditions, including a phase 3 trial of cannabis-based treatments for chronic low back pain that used 32.5 mg THC.
They also noted a pronounced placebo response, with more than one-third of placebo recipients reporting meaningful pain reduction, which may have diminished the ability to detect a treatment signal. Notably, no cognitive safety signal emerged in the active treatment group, contrasting with trends observed in prior trials using higher THC concentrations.
The study’s nine-week follow-up period limits conclusions about long-term safety and efficacy. Plasma cannabinoid concentrations were not measured, and participants received no standardized instructions on meal timing relative to dosing, an important variable given that bioavailability of both CBD and THC increases with high-fat food intake.
Given the complexity of pain management in Parkinson disease, further research is needed to explore the potential benefits of cannabinoids in this context, including the optimal dosing and formulation of these compounds.