Demographics and Clinical Factors Associated with Persistence on Cannabidiol in US Patients with Lennox-Gastaut Syndrome (LGS), Dravet Syndrome (DS), Tuberous Sclerosis Complex (TSC), or Other Refractory Epilepsies
Leah Burn1, Vicki Osborne2, Arthur Sillah1, Timothy Saurer1, Michael Faithe1, Maggie McCarter1, Sanket Shah1, Timothy Barnes3, Arunima Sachdev3, Anuj Gupta3, Mohankumar Kurukumbi4
1Jazz Pharmaceuticals, Inc., 2Jazz Pharmaceuticals, UK Ltd., 3Optum Life Sciences VES, 4Inova Fairfax Hospital
Objective:
To evaluate sociodemographic and clinical factors associated with persistence on plant-derived, highly purified CBD (Epidiolex® 100 mg/mL oral solution) among patients with LGS, DS, TSC, or other refractory epilepsies.
Background:
CBD is approved in the US for treating seizures associated with LGS, DS, or TSC in patients aged ≥1 year. Real-world studies report long-term CBD persistence in 70–77% of patients, but patient characteristics influencing persistence remain unclear.
Design/Methods:
This retrospective cohort study used the US Optum® Market Clarity database. Patients with epilepsy and ≥1 CBD prescription between 5/15/2018 and 6/30/2024 were eligible. Baseline period was 12 months before index date (date of first pharmacy claim for CBD after diagnosis). Follow-up period was variable post-index. Patients with missing/invalid demographic data or evidence of baseline CBD use were excluded. Persistence was based on time to discontinuation (>60-day supply gap). Multivariable Cox proportional hazard model(s) identified factors associated with persistence. To account for multiple comparisons across factors, Bonferroni correction was applied as a sensitivity analysis.
Results:
The analytical cohort included 7815 patients (LGS, n=4649; DS, n=558; TSC, n=185; other refractory epilepsies, n=2423). Pre-Bonferroni corrections, lower persistence was associated with higher comorbidity burden, epilepsy-related surgeries, and anxiety, and higher persistence with clobazam use, polypharmacy, enrollment in the northeastern/southern US, patient age 3–17 years, and Hispanic ethnicity. Post-Bonferroni corrections, several associations remained significant: baseline anxiety was associated with lower persistence (hazard ratio [HR]=1.14; P=0.039), while clobazam use (HR=0.87; P<0.001), enrollment in the northeast (HR=0.77; P<0.001), and diagnosis of LGS, DS, or TSC vs other refractory epilepsies (HR=0.77, 0.78, 0.67, respectively; all P<0.001) were associated with higher persistence.
Conclusions:
Persistence on CBD may vary by diagnosis, comorbidities, concomitant medication use, and region of enrollment. Identification of these factors may support clinicians in recognizing patients at risk for discontinuation and support treatment continuity.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.