Abstracts

Rationale: A previous study of perampanel investigated health care resource utilization as a proxy for clinical effectiveness using real-world data. While the results demonstrated a reduction in resource utilization following initiation of perampanel, an alternative explanation suggested “active management” was at work (i.e., any change in therapy decreases utilization when compared to no change). To evaluate the specific influence of medication, hospitalization rates following initiation of perampanel were compared to those of lacosamide, another branded antiepileptic drug (AED) commonly prescribed in the treatment of epilepsy. Methods: Symphony Health, a nationally representative medical and pharmacy claims database, was used in this study. Patients were identified if they had filled a prescription of either perampanel or lacosamide during the period 07/01/14 to 06/30/16. The index date was the date of the first fill of the medication, and patients were selected if they were at least 12 years of age or older, had continuous observations for the year prior to and following this date, if they had at least two diagnoses of epilepsy or non-febrile convulsions, and if they had at least one additional medication fill following the index date. Subsequently, using propensity scores, patients were matched 1:1 on age, gender, index year, region, Charlson Comorbidity Index (CCI), number of previous AEDs, and evidence of previous hospitalizations. The outcome variables of interest were all-cause and epilepsy-related hospitalizations in the one year following medication initiation. Results: After matching, there were 1,717 patients in each cohort and most of the baseline demographics were balanced. A higher percentage of patients with 3 or more AEDs at baseline, however, was observed in the perampanel cohort (60.5%), compared to the lacosamide cohort (57.7%), p<.001. The results showed that initiation on perampanel resulted in a 9.6% reduction in all-cause hospitalizations, compared to a 5.8% reduction for those initiated on lacosamide, p<.05.  Epilepsy-related hospitalizations also decreased, 9.9% vs. 8.3% for perampanel and lacosamide, respectively, p<.05. Among those who had any hospitalization in the baseline period, patients treated with perampanel demonstrated a 59.9% reduction in the all-cause hospitalizations vs. a 48.6% reduction for those treated with lacosamide, p<.05. A similar result was seen for epilepsy-related hospitalizations, a reduction of 65.0% vs. 58.9%, respectively, p<.05. Conclusions: In patients diagnosed with epilepsy, treatment with perampanel, compared to treatment with lacosamide, was associated with a significantly greater reduction in all-cause and epilepsy-related hospitalizations.  Even when considering the effects of active management, these results suggest that the choice of adjunctive medication may lead to differences in health care utilization. Funding: This study was funded by Eisai Inc.