Abstracts

Rationale: Extended-release formulations of antiepileptic drugs (AEDs) have been shown to reduce side effects compared to conventional-release forms. Previous research has demonstrated that compliance decreases with the number of doses taken daily. Evidence of improved compliance, although intuitive and seemingly practical, has not been documented for Divalproex ER in any prospective trial to date. With increasing availability of generic AEDs, hospitals, managed care and government institutions have mandated switching patients from standard-release (SR) and extended release (ER) Divalproex to immediate release (IR) VPA to lower costs. The purpose of this study is to prospectively document whether compliance and dose schedule precision is better for ER compared to IR VPA.Methods: The study population consists of 20 non-driving patients over 16 years of age with relatively well-controlled epilepsy on SR Divalproex. Half were randomly assigned to receive IR VPA tid or qid in daily doses equivalent to their baseline dose of SR, and the other half were assigned to recieve ER in bioequivalent daily doses to their baseline dose. Compliance, defined as number of times study dose was actually taken compared to prescribed, was monitored with a computer-chip bottlecap (MEMS, AARDEX). Precision, defined as actual time of day study doses were taken, was compared to prescribed times using MEMS caps. The study period was 42 days, with a 7 day practice phase using MEMS with baseline SR dosing. Patients who could not comply at least 75% during baseline were not enrolled. Routine laboratories and VPA levels were assessed at days 0 and 42 to document metabolic variations.Results: The initial half of the enrollees demonstrated better compliance with ER than IR. Patients on IR complained of difficulty remembering doses more commonly than GI distress. Dosing precision was similar in both groups. Over 50 patients were screened to enroll the initial half of the population. The most common reason for refusing the study was unwillingness for randomization to tid. The second most common reason was transportation. Enrollment will soon conclude and final results and statistical analysis will be presented at the AES.Conclusions: This study documents better compliance with ER divalproex than with IR VPA. Dosing precision is not necessarily improved by changing to once-daily formulations. Less variable peak to trough variation, however, may render imprecision potentially more forgiving with ER than with IR VPA. The high number of patients refusing enrollment underscores that patients have become accustomed to bid or qd dosing of AEDs and dislike more frequest regimens. Formularies requiring such regressive changes adversely affect epilepsy patients' overall desired quality of life, despite apparent financial advantages. More study is needed to assess how improved compliance affects overall seizure control, especially in a less well-controlled epilepsy population.