Abstracts

In a double-blind trial, TPM was at least as effective as 600 mg/day CBZ and 1250 mg/day VPA monotherapy in patients with newly diagnosed epilepsy. (Privitera 2003) We report post-hoc analyses of relative effectiveness in partial-onset (POS) and primary generalized (GEN) seizures. Patients [ge]6 yrs were eligible if epilepsy had been diagnosed [lt]3 mos before screening. To individualize therapy, investigators selected CBZ or VPA as preferred agent. Patients were randomized to double-blind treatment with TPM (100 or 200 mg/day) or investigator[rsquo]s choice of CBZ or VPA. Dose reductions to manage adverse events (AEs) were not allowed. Patients exited if unable to achieve assigned dose or if therapy change required. Primary endpoint was time to exit; secondary efficacy endpoints were time to first seizure and patients seizure-free for last 6 months of treatment. TPM data pooled for comparison with CBZ and VPA. In post-hoc analyses, patients stratified by baseline seizure type: POS (simple partial, complex partial, secondarily generalized) or GEN (tonic-clonic, tonic, clonic). The intent-to-treat population was 613 patients (CBZ, N=126 patients; VPA, N=78; TPM, N=409). 382 patients had POS (TPM, N=258; CBZ, N=92; VPA, N=32); 211 had GEN (TPM, N=142; CBZ, N=31; VPA, N=38). TPM did not differ significantly from CBZ and VPA in time to exit among patients with POS (P=0.803) or GEN (P=0.720). Retention rates were 1-3% higher in both POS and GEN for TPM vs. CBZ/VPA at all time points (90, 180, 270, 365 days). The 95% confidence interval around these differences showed that retention rates with TPM could be 11-14% higher vs. CBZ/VPA in patients with POS and 15-19% higher in patients with GEN. At the opposite limit of the 95% confidence interval, retention rates could favor CBZ/VPA by 8-10% in POS and 9-12% in GEN. The seizure-free rate was slightly lower (5%) for TPM vs. CBZ/VPA in POS but higher (6%) in GEN. From the 95% confidence interval, the seizure-free rate could favor CBZ/VPA by 15.5% in POS and 8.5% in GEN. When the difference favored TPM, the seizure-free rate could be 5.5% higher in POS and 20% higher in GEN. Retention times and time to first seizure were generally shorter in both POS and GEN in VPA- vs. TPM- or CBZ-treated patients. Overall, differences between TPM and CBZ and VPA were not statistically significant. Discontinuation rates due to AEs were TPM 100, 19%; TPM 200, 28%; CBZ, 25%; VPA, 23%. AEs differed according to their known profiles. Regardless of seizure type, TPM is at least as effective as CBZ and VPA in newly diagnosed epilepsy. (Supported by Johnson [amp] Johnson Pharmaceutical Research [amp] Development, LLC)