Time to Event Trial Endpoints Best Measured Efficacy When Seizure Counting Started During Maintenance Excluding Titration
Abstract number :
1.285
Submission category :
7. Anti-seizure Medications / 7B. Clinical Trials
Year :
2023
Submission ID :
352
Source :
www.aesnet.org
Presentation date :
12/2/2023 12:00:00 AM
Published date :
Authors :
Presenting Author: Neo Kok, BS – University of Michigan
Advith Reddy, BS – University of Michigan; William Stacey, MD, PhD, FAES – University of Michigan; John Stern, MD – UCLA; Page Pennell, MD, FAES – University of Pittsburgh; Jacqueline French, MD – NYU; Wesley Kerr, MD, PhD – University of Michigan
Rationale: Placebo-controlled randomized clinical trials (RCTs) evaluate the efficacy of novel treatments for epilepsy. Time to event endpoints, like time to prerandomization seizure count (T-PSC), may measure this efficacy in less time, which may improve recruitment, cost, and participant risk by reducing exposure to placebo or ineffective treatment. While T-PSC has been demonstrated to be effective, the inclusion of titration in the T-PSC endpoint may underestimate treatment efficacy and exclude participants prior to reaching the assigned maintenance dose.
Methods: We applied the T-PSC design to six double-blind RCTs of adjunctive antiseizure medications (ASMs) for epilepsy including levetiracetam, brivaracetam, lacosamide, and lamotrigine for either focal or generalized-onset epilepsy. For each trial and in a meta-analysis across trials, we re-calculated the primary efficacy endpoints of median percent reduction in seizure frequency and 50% responder rate using data from the full-length trial, as well as before T-PSC when the T-PSC endpoint started during titration as compared to maintenance only. We compared the individual-participant level correspondence of efficacy when T-PSC started during titration or maintenance to the full-length trial using Spearman’s Rho and Cohen’s Kappa.
Results:
Inclusion of titration as compared to maintenance only in T-PSC did not significantly underestimate the effect size of the group-level primary efficacy outcomes (MPR titration 41% reduction in effect size with titration vs 10% reduction with maintenance only each compared to the full-length trial, p=0.12; 50RR 27% reduction with titration vs 21% reduction with maintenance, p=0.72). The T-PSC design with either start point did not significantly underestimate the effect size of the primary efficacy outcomes (MPR p=0.21, 50RR p=0.06) The correspondence of individual-level treatment efficacy was reduced when T-PSC started with titration as compared to maintenance only (Cohen’s Kappa: Titration 77%, Maintenance Only 84%, p< 0.05; Spearman’s Rho: Titration 72%, Maintenance Only 84%, p< 0.01). There was an elevated rate of participants who had greater than 50% response in the full-length trial but had less than 50% response prior to T-PSC when T-PSC started with titration, compared to maintenance only (False Negative Rate: Titration 29%, Maintenance Only 14%).
Anti-seizure Medications