Optimal Cut Point for Reduction in Long Episode Frequency to Predict Meaningful Change in Clinical Seizure Frequency
Arnold Gammaitoni1, Martha Morrell2, Jacqueline French3, Daniel Friedman4, Kathryn Davis5, Thomas Tcheng2, Cairn Seale2, Bradley Galer1, William Motley1
1Rapport Therapeutics, Inc., 2NeuroPace, 3New York University Comprehensive Epilepsy Center, 4New York University Grossman School of Medicine, NYU Langone Health, 5University of Pennsylvania
Objective:
To examine the optimal cut point for long episode (LE; prolonged abnormal electrographic activity) frequency reduction that predicts clinically meaningful clinical seizure (CS) frequency reduction in patients with a responsive neurostimulator (RNS® System, NeuroPace).
Background:
Novel proof-of-concept designs utilizing biomarkers with positive predictive ability of antiseizure medication (ASM) efficacy may require lower patient numbers for signal detection than traditional clinical trials. In drug-resistant focal epilepsy patients with an RNS® System, LE and CS frequency changes were correlated following a new ASM start.
Design/Methods:
Retrospective data from a long-term RNS® System study were analyzed. Analysis period: 8 weeks prior to (“baseline”) and 8 weeks following ASM start. Key inclusion criteria: no RNS® System setting changes during the analysis period; ≥8 LEs per 28d (baseline); LE and CS data available.
Outcomes, following ASM start vs baseline: median percentage change in CS and LE frequency per 28d, LE responder rate, and receiver operating characteristics (LE reduction required to predict clinically meaningful [≥50%] or profound [≥75%] CS reduction).
Results:
Overall, patients (n=45) experienced median 30% (LE) and 50% (CS) reductions following ASM initiation vs baseline. LE median reductions: clobazam (n=15), 50%; levetiracetam (n=4), 73.4%; lacosamide (n=26), 8.8%. CS median reductions: clobazam, 29.4%; levetiracetam, 78.6%; lacosamide, 51%. Responder rates (≥30% LE reduction): clobazam, 60%; levetiracetam, 100%; lacosamide, 39%. Overall, LE responders experienced a median 60% reduction in CS; ≥50% CS reduction was seen in 16/23 (70%) LE responders. LE reduction thresholds for predicting clinically meaningful and profound CS reductions were ≥30% (AUC=0.765) and ≥50% (AUC=0.735), respectively.
Conclusions:
Results suggest that a patient achieving ≥30% LE reduction following ASM initiation will experience clinically meaningful CS reduction 70% of the time. Higher LE response thresholds were associated with profound reductions in CS. LE responder analysis may serve as a biomarker in proof-of-concept studies with positive predictive ability of clinically meaningful efficacy.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.