Seizure-free outcome in randomized add-on trials of the new antiepileptic drugs

Purpose: The goal of this study is to (1) provide clinically useful, previously unpublished comparative analyses of seizure-freedom rates for newer antiepileptic drugs (AEDs), and (2) recommend a standard for data presentation and analysis. Methods: Data were reviewed from placebo-controlled adjunctive trials in refractory patients of gabapentin (GPN), lamotrigine (LTG), topiramate (TOP), tiagabine (TGB), oxcar-bazepine (OXC), levetiracetam (LEV), zonisamide (ZNS), and pregabalin (PGB). Seizure-freedom analyses in these publications, if included at all, consistently included both patients who completed the trial, and those who dropped out prior to completion (last observation carried forward, LOCF). This has the potential to increase reported seizure-free outcomes. Pharmaceutical companies were contacted for the provision of unpublished seizure-free data in the patients who completed the entire study. Results: In most cases, LOCF analysis produced a higher rate of seizure freedom compared to completer analysis. A total of 0%-1.1% of the LOCF population was seizure-free in the GPN trials (completer data not available). For the remaining AEDs, seizure-freedom results in the LOCF versus completer populations were: 0.7% versus 0.8% (LTG trial); 12% versus 2.6% (OXC trial); 3.6%-6.4% versus 3.9%-7.1% (LEV trial); 3.7%,-7.9% versus 1.3%-1.4% (PGB trial); and 6.0% versus 3.0% (ZNS trial, minus titration period). Conclusions: By employing LOCF, a clinically unrealistic picture of seizure-free rates may be reported. Access to completer data is informative, as it includes only those patients who were able to tolerate the drug at doses that produced seizure freedom. Ideally, data from both ITT and completer analyses should be made available.