Authors :
Presenting Author: Caitlin Grzeskowiak, PhD – Epilepsy Foundation
Gary Cutter, PhD – University of Alabama; Gabriel Biondo, N/A – Notre Dame; Ojas Potnis, B.S. – Texas A & M; Rachel Sukonnik, N/A – University of Delaware; Brandy Fureman, PhD – Epilepsy Foundation; Ruben Kuzniecky, MD – Zucker Hofstra School of Medicine; Daniel Lowenstein, MD – UCSF; Jacqueline French, MD – NYU Comprehensive Epilepsy Center
Rationale: Does focal treatment resistant epilepsy (FTRE) improve or worsen over time? This is an important question, as many open label long-term extension studies of new therapies suggest improvement over time invokes a possible disease modifying effect. The Human Epilepsy Project 2 (HEP2) is a prospective, observational, multicenter US study with the goal of identifying treatment response in participants with FTRE who were treated with multiple interventions at their physicians’ direction. Here, we report on the trends in seizure (sz) frequency for participants over the course of the study.
Methods:
All subjects had focal epilepsy, were 16–65 y/o at enrollment, failed ≥4 anti-seizure medications (ASMs) (at least two due to failure of seizure control), and receiving ≥1 ASM. Participants were recruited from 10 US epilepsy centers and followed up to 36 months. Seizure frequency data were collected via daily electronic seizure diaries, monthly coordinator check-ins, and medical records. Analyses of monthly seizure counts included repeated measure assessments by month of f/u, which assume data are missing at random. Thus, to prevent confounding by dropouts, we compared the first half of the f/u to the second half for each participant. We examined all data in a single model with the response variable as monthly seizure counts and the independent variable month of f/u. To assess effects of those with differing amounts of f/u, we created three cohorts of f/u times: < 12, 12-24 and >24 months of diary entries and used the repeated measures models on each cohort to examine the slopes of the sz counts in time within cohort.
Results:
A total of 128 subjects provided seizure frequency data over time. Participant average age was 40 ± 12 years, 57.4% female,71.3% white. Average age at epilepsy dx was 19.6 ± 13.6 years. Overall, 55% of the participants had a medicine added during the observation period. A total of 68.2% of participants experienced seizure frequency reduction in the second half of their f/u compared to the first half. Table 1 shows the overall linear regression results from the repeated measures models overall and those run separately by cohort. All the slopes are negative showing continued improvement with time. The mean seizure count averaged over the first three months showed a mean of 14.0, 16.3 and 29.5 for cohort 1, 2, and 3 respectively. The models would predict a reduction in Cohort 1 from a three month baseline average of 14.0 of -9.5 seizures at 12 months. In cohort 2, the drop would be from 16.5 by -3.9 and Cohort 3 from 29.5 down by -15.1. Fig. 1 shows the trend in the LS Means sz counts estimated from the model over the projected months of f/u, although as the time increases the model overestimates the reductions.
Conclusions:
Our data, from a prospective trial, suggests improvement of seizures over time in FTRE. The data cannot distinguish between “natural history” and effects of ongoing active management. In either case, these data support the need for cautious interpretation of open-label studies that report improvement trends over time. Future analyses of our data will address impact of specific medication interventions during the trial.
Funding: Funded by UCB, Neurelis, & SK Life Sciences.