Abstracts

RATIONALE: FDA regulatory requirements for approval of antiepileptic drugs for use in monotherapy include demonstration of superiority over a comparator. This has led to inclusion of either a placebo or [dsquote]pseudo-placebo[dsquote] arm in several monotherapy studies. The FDA recognizes historical control as a valid control group, if multiple rigorous studies demonstrate a stable and highly reproducible behavior under similar conditions. Nine studies have been performed in which patients were randomized to either a [dsquote]pseudoplacebo[dsquote] or active study drug, and then withdrawn towards monotherapy. Similar methodology was used for all studies. We performed a meta-analysis of all pseudo-placebo arms to determine whether they could serve as a historical control for future monotherapy trials.
METHODS: Review of the literature and direct communication with pharmaceutical companies revealed 9 (7 published, 2 unpublished) randomized controlled trials in which patients with refractory partial seizures were randomized to either active compound or [dsquote]pseudo-placebo[dsquote], before (8 trials) or after (1 trial) being withdrawn towards monotherapy. All trials used similar inclusion criteria. Study drugs were gabapentin, tiagabine, topiramate, felbamate and oxcarbazepine. The [dsquote]pseudoplacebo[dsquote] arm consisted of either low dose of the study drug or 15 mg/kg of valproic acid. The outcome measure chosen for this analysis was percent of patients exiting the trial, estimated from Kaplan-Meier curves. Patients exited the trial if they experienced seizure worsening as identified by predetermined exit criteria. Four nearly identical exit criteria were used in 8/9 trials. The 9th trial used only 3 exit criteria, and only one exactly matched a criterion used in the other studies. This trial was also an outlier in terms of results.
RESULTS: A total of 398 pts were enrolled in the pseudoplacebo cohorts for the 8 trials using almost identical exit criteria (range of 22-94 patients/cohort). Exit rates ranged from 81-100%, and confidence interval lower limits for exit rates are all above 68%.
CONCLUSIONS: These results indicate that the behavior of patients randomized to pseudoplacebo, and subsequently withdrawn towards monotherapy is predictable and reproducible when a standard set of exit criteria is used. Over 68% will worsen and meet exit criteria. Choice of pseudoplacebo does not appear to influence exit rate. These results may enable the use of historical controls for future monotherapy studies, obviating the need for a placebo/pseudoplacebo arm.
Objective: At the end of this presentation the participant will be able to discuss the natural history of withdrawal to monotherapy in patients randomized to pseudoplacebo
(Disclosure: Salary - Wang/PledgerJohnson and Johnson PRD, L.L.C., Grant - French: Novartis, Consulting - French: Novartis, Glaxo-Wellcome, Ortho-Mcneil, Abbott, Stock - Wang, The Johnson & Johnson, Honoraria - French: Novartis, Glaxo-Wellcome, Ortho-Mcneil, Abbott)