Objectives – To evaluate the impact of different dosages of topiramate (TPM) add-on to stable antiepileptic therapy for refractory partial epilepsy in adults.
Material and methods – Pooled intention-to-treat analysis of six similarly designed double-blind, placebo-controlled trials, including 481 patients treated with doses of TPM 200, 400, 600 and 800 mg/day, and 265 patients receiving placebo.
Results – Seizures were reduced by ≥50% from baseline in 41% of TPM-treated patients and 15% of placebo-treated patients (P < 0.001); 5 and 0.8%, respectively, were seizure-free (P < 0.003). TPM was significantly better than placebo regardless of gender, age, baseline seizure rate as well as number and type of concomitant antiepileptic drugs. Efficacy was statistically significant in favour of TPM at all dose levels: at least 50% seizure reduction was achieved in 40% of patients with 200 mg, 41% with 400 mg, 44% with 600 mg and 41% with 800 mg TPM when compared with 15% with placebo (P ≤ 0.001 for each dosage arm vs placebo). The median reduction in monthly seizure frequency was 38%, 42%, 45% and 38% vs 8%, respectively (P ≤ 0.001). Moreover, response to TPM was significantly superior to placebo at each of the dose levels tested for most of the baseline variables. The total percentage of withdrawals increased with the dosage, and the withdrawals caused by adverse events increased from 3% with placebo to 7% with 200 mg TPM (not significant vs placebo), 15% with 400 mg TPM (P = 0.08), 16% with 600 mg TPM (P = 0.002) and 15% with 800 mg TPM (P = 0.003).
Conclusion – The efficacy of TPM add-on in partial epilepsy is consistent across efficacy endpoints and independent of study population characteristics. The response at 200 mg TPM is similar to the response at higher doses, but as drop-outs caused by adverse events are more frequent above the 200 mg dose, this pooled analysis supports that 200 mg daily is a good target dose for add-on therapy in most patients with partial epilepsy, showing an excellent balance between efficacy and tolerability.