Medical Cannabis in France: What the First Major Study Reveals About Its Actual Impact

Four years after the launch of the national pilot program, a study commissioned by the ANSM finally provides concrete data on the effect of medical cannabis on treatments and the care pathways of French patients. The results, published in September 2025 by the AHeaD team at the Inserm Research Center at the University of Bordeaux and obtained by the Union of Industrialists for the Promotion of Hemp Extracts (UIVEC) following a CADA request, are encouraging.

A first-of-its-kind experiment, real-world data

France's Pilot Program for Medical Cannabis began on March 26, 2021 and officially ended on December 31, 2024. His original goal The goal was to test the feasibility of the prescription and dispensing process, and, secondarily, to collect the first French data on the efficacy and safety of cannabis in a medical setting.

It was in this context that the ANSM commissioned the Bordeaux research team to conduct the U.Cannabis study, the final report of which was published on September 15, 2025. The methodology is based on data matching the ReCann register, which listed the patients included in the trial, with those from the National Health Data System (SNDS), the massive database of the French national health insurance system. In practical terms, this made it possible to track, month by month, changes in each patient’s medication use and healthcare utilization over a period ranging from one year before the start of medical cannabis treatment to six months afterward.

A total of 1,971 patients were included in the final analysis, following data cleaning and verification of matching. The median age was 56 years, and 54% of the patients were women. The vast majority of the population was treated for two conditions: refractory neuropathic pain (56.1% of patients) and painful spasticity associated with multiple sclerosis or other central nervous system disorders (17.6%). The other indications—palliative care, refractory symptoms in oncology, and refractory epilepsy—each accounted for less than 11% of the cohort. Medical cannabis is prescribed to these patients as a last resort, that is, after other treatments have failed.

Lower Use of Pain Relievers, Especially Among Patients with Neuropathic Pain

The most robust finding of the study concerns refractory neuropathic pain, by far the best-documented indication due to the size of the patient cohorts. Before starting medical cannabis, approximately 73% of patients consistently used pain relievers every month. As soon as treatment began, this proportion dropped significantly: on average, 40 out of every 1,000 patients stopped using pain relievers of any type.

More specifically, it is the so-called “atypical” pain relievers—gabapentinoids (pregabalin, gabapentin), tricyclic antidepressants, and certain dual-action antidepressants—that have seen the sharpest declines. This comes as no surprise: these medications are, in fact, the first-line treatment for neuropathic pain, and their decline suggests that medical cannabis is serving as a partial substitute. The use of strong opioids is also declining, which is potentially one of the most interesting findings of the study, given that opioid dependence is a growing public health concern in Europe.

Furthermore, the cumulative doses administered also decrease after treatment initiation, not just the number of patients taking the medication. In other words, those who continue to take analgesics are taking smaller doses. This dual trend—fewer users and reduced doses—supports the interpretation that there is a real effect, even if it is modest.

Another notable finding in this indication: the use of anxiolytic benzodiazepines decreases significantly after the start of cannabinoid treatment, which may reflect an improvement in anxiety associated with chronic pain.

Results vary depending on the indications

In palliative care settings and cases of refractory symptoms in oncology, the results follow the same pattern, but with much greater variations, which can be explained in part by the specific dynamics of these patient populations. Patients in palliative care experience an escalation in medication use in the months leading up to the start of treatment, followed by a sharp decline afterward: the high mortality rate in this group (only 88 of 214 patients were still being followed at 6 months) makes interpreting the data more complex.

For painful spasticity associated with MS, the picture is less favorable: while the number of patients taking non-opioid analgesics decreases after initiation of treatment, the cumulative doses administered increase for nearly all drug classes. The study offers a technical explanation—a possible artifact of the segmented regression method—but this finding warrants attention. It is possible that medical cannabis is used in addition to existing treatments for this condition, rather than as a substitute for them.

Finally, refractory epilepsy remains the great unknown: the sample size (149 patients) is too small for the statistical analyses to be conclusive. Visually, there appears to be a downward trend in the use of anticonvulsant benzodiazepines after initiation of treatment, a finding consistent with the known pharmacological interaction between CBD and Clobazam, but it does not reach the significance threshold.

Treatment adherence that raises questions

One of the study’s most significant cross-sectional findings concerns treatment adherence: at six months, only 54.8% of the enrolled patients were still using medical cannabis. This figure is likely an underestimate, as the authors themselves point out gaps in the ReCann registry.

But it reflects a reality familiar to clinicians: adverse effects (primarily neuropsychiatric and gastrointestinal) and the perceived lack of efficacy lead a significant proportion of patients to discontinue treatment within the first few months. Treatment persistence is particularly low in palliative care settings (21.5% at 6 months), which can be attributed to deaths.

What These Results Really Mean

The authors are cautious. The U.Cannabis study is not a randomized clinical trial: it lacks a control group, and it is therefore impossible to rule out the possibility that some of the observed trends are related to the natural course of the conditions or to other factors. Sample sizes remain limited for several indications, and the six-month follow-up period is short for chronic diseases.

That said, the results are consistent with the available international literature. The most recent meta-analyses published in the BMJ and the Cochrane Library ¹, conclude that medical cannabis has a small but real effect on chronic pain and the associated symptoms. The French study confirms that this signal also exists under real-world conditions—with actual prescriptions, actual patients, and an actual healthcare system.

For France, this data comes at a pivotal moment. The pilot program has ended. The widespread adoption of medical cannabis has been expected for many months. The U.Cannabis study does not settle this debate but provides a strong argument for those advocating a transition toward permanent, regulated access: even if it's slight, the effect is there.

¹

• Wang et al., BMJ, 2021 – a meta-analysis on medical cannabis for chronic cancer-related and non-cancer-related pain
• Mücke et al., Cochrane, 2018 – A review of cannabis-based medications for chronic neuropathic pain in adults
• Solmi et al., BMJ, 2023 - umbrella review on the benefits and risks of cannabis (meta-analyses of randomized trials and observational studies)