Relative performance of brivaracetam as adjunctive treatment of focal seizures in adults: a network meta-analysis

Abstract

Objective: To estimate the relative efficacy, safety and tolerability of adjunctive brivaracetam and other antiepileptic drugs (AEDs) using a Bayesian network meta-analysis (NMA) approach.

Methods: A systematic literature review (SLR) identified randomized controlled trials of AEDs treating focal (partial-onset) seizures for ≥8 weeks and assessed them for inclusion in the NMA. Bayesian random-effects NMA was performed for several outcomes. All interventions within the licensed dose range were included in the network of evidence.

Results: The SLR identified 82 studies; 65 were included in the NMA. These studies had baseline mean age 33.1–38.0 years, mean duration of epilepsy 18.7–23.0 years and median seizure frequency/28 days 8.1–11.8. All AEDs had significantly higher odds than placebo of achieving ≥50% responder rates (odds ratios 1.83–3.58) and all AEDs had a trend of higher odds than placebo of achieving seizure freedom (odds ratios 1.36–5.73), most being statistically significant. Tolerability outcomes were comparable between AEDs; most AEDs had higher odds than placebo of treatment-emergent adverse events leading to discontinuation, serious AEs, nausea, fatigue, dizziness and somnolence.

Conclusions: This NMA would appear to show relative equivalence in efficacy, safety and tolerability outcomes of the included AEDs. However, patient heterogeneity within trials and in clinical practice should be considered when interpreting these results. While NMAs are based on the best available evidence the authors suggest that, due to the inability of NMAs to capture unmeasured confounding factors and population heterogeneity, NMAs must not be the sole basis for comparative treatment recommendations.

Keywords:

• Anticonvulsant
• Bayesian analysis
• epilepsy
• mixed treatment meta-analysis
• network meta-analysis

Acknowledgements

The authors thank Neha Sharan of BresMed India for her contribution to the systematic literature review, and thank Ritesh Bisen, Preety Rajora, Surabhi Gupta and Neha Jiandani of BresMed India for independent reviewing and extraction of studies in the systematic literature review. The authors would like to acknowledge Sally Cotterill PhD, of QXV Comms, an Ashfield Company, part of UDG Healthcare plc, and Carole Nadin, of Fleetwith, for medical writing support that was funded by UCB Pharma.