Clinically Meaningful Reduction in Drop Seizures in Patients with Lennox-Gastaut Syndrome Treated with Cannabidiol: Post Hoc Analysis of Phase 3 Clinical Trials.

Abstract

Background and objective: In clinical trials of patients with Lennox-Gastaut syndrome (LGS), a ≥ 50% reduction in drop seizure frequency is generally accepted as a key endpoint. However, smaller reductions (< 50%) may yet be impactful for patients in real-world settings. This exploratory analysis evaluated the threshold for a clinically important response in drop seizures that is associated with the Caregiver Global Impression of Change (CGIC) scale score in patients with LGS treated with cannabidiol (CBD) oral solution and assessed the suitability of CGIC as an anchor for meaningful change.

Methods: This exploratory post hoc analysis included patients with LGS (N = 215, age 2-55 years) receiving CBD (Epidiolex^® [USA]/Epidyolex^® [EU]; 100 mg/mL oral solution) in two phase 3 randomized placebo-controlled trials (NCT02224690, April-October 2015, and NCT02224560, June- December 2015). Reduction in drop seizures (involving sudden loss of muscle tone) was anchored to CGIC scores of "slightly improved" or better or "much improved" or better, to determine the threshold at which seizure reduction can be considered clinically meaningful to patients. Spearman's correlation indicated suitability of anchors (absolute value ≥ 0.30 deemed appropriate).

Results: In the 215 patients receiving CBD with a CGIC score recorded, CGIC was "slightly improved" or better in 60% of patients, and "much improved" or better in 31% after 14 weeks of treatment. With a CGIC rating of "slightly improved" or better, the best threshold for a clinically important response in drop seizure reduction was - 30.6% (57.7% of patients). Mean and median percentage reductions in drop seizures were - 46.9% and - 58.6%, respectively. Using "much improved" or better, the best threshold was - 49.6% (40.5% of patients). Mean and median percentage reductions in drop seizures were - 57.6% and - 66.0%, respectively. Spearman's correlation was 0.47.

Conclusion: Anchoring to CGIC of "slightly improved" or better, the threshold for a clinically meaningful reduction in drop seizure frequency was 31%, suggesting that a 50% cutoff may overlook patients with meaningful improvements in their overall condition, as perceived by their caregivers. CGIC scores, although potentially less nuanced than other standardized clinical assessments, were appropriate anchors to determine thresholds. This exploratory analysis may help contextualize clinical trial data to better understand potential patient benefit attained by reductions in drop seizure frequency observed in real-world settings that are < 50%.

Clinical trials registration numbers: NCT02224560 and NCT02224690.