Abstracts

Rationale:
Third-generation antiepileptic drugs (AEDs) are generally believed to have better safety profiles than first- or second-generation AEDs. Brivaracetam (BRV), one of the newest third-generation AEDs, is a structural analog of levetiracetam (LEV), which has been shown to be associated with psychiatric adverse events. No studies have directly compared the psychiatric effects of eslicarbazepine acetate (ESL) and BRV or LEV in adults. Network meta-analysis (NMA) compares the effects of multiple treatments, including those that lack head-to-head trials, by combining direct and indirect evidence. The objective of this study was to compare ESL and BRV via NMA to assess differences in total psychiatric treatment-emergent adverse event (TEAE) endpoints.
Method:
A systematic literature review was conducted to identify randomized clinical trials and pooled studies of ESL or BRV, which reported psychiatric adverse events in patients diagnosed with focal seizures. Pooled analyses with trials that lacked a placebo arm, which precluded an indirect comparison with the other pooled analyses, and data for ESL 400 mg, which is not an approved maintenance dose, were excluded. The two pooled analyses in adult patients who reported overall and dose-specific psychiatric TEAEs for ESL (3 adjunctive trials) and BRV (6 adjunctive trials) were included in the NMA. Two scenarios were considered: psychiatric TEAE comparing ESL and BRV overall (Scenario 1) and psychiatric TEAE comparing ESL and BRV at different dosages (Scenario 2). Fixed effects Bayesian models were conducted to estimate odds ratios (ORs) with 95% credible intervals (CrI) for each direct and indirect comparison. Number needed to harm (NNH) for treatment vs. placebo was also calculated.
Results:
In Scenario 1, the odds of psychiatric TEAEs were lower for ESL compared to BRV (OR: 0.74, 95% CrI: 0.45-1.23). NNH was higher with ESL compared to BRV (ESL: 172.28, BRV: 32.39). In Scenario 2, the odds of psychiatric TEAEs were lower for ESL 1200 mg and ESL 800 mg compared to BRV 50 mg (ESL 1200 mg OR: 0.69, 95% CrI: 0.37-1.27; ESL 800 mg OR: 0.75, 95% CrI: 0.41-1.38), BRV 100 mg (ESL 1200 mg OR: 0.65, 95% CrI: 0.35-1.17; ESL 800 mg OR: 0.71, 95% CrI: 0.39-1.27), BRV 150 mg (ESL 1200 mg OR: 0.92, 95% CrI: 0.46-1.87; ESL 800 mg OR: 1.00, 95% CrI: 0.51-2.03), and BRV 200 mg (ESL 1200 mg OR: 0.75, 95% CrI: 0.39-1.46; ESL 800 mg OR: 0.82, 95% CrI: 0.43-1.59). NNH was higher with ESL compared to BRV (ESL 800 mg: 100.33, ESL 1200 mg: 628.27, BRV 50 mg: 28.54, BRV 100 mg: 24.33, BRV 150 mg: 112.80, BRV 200 mg: 37.93). Underlying simplifying assumptions of the NMA technique might have led to wide observed CrIs (Please see Tables).
Conclusion:
Using the NMA technique described, ESL was associated with a profile suggesting a lesser likelihood of psychiatric TEAEs, relative to BRV; however, these differences were not statistically significant.
Funding:
:Sunovion Pharmaceuticals Inc.