Abstracts

Rationale: Between September 1993 and August 1994, its first year of commercial availability, treatment with felbamate (FBM) was associated with 33 cases of aplastic anemia. Studies have suggested that a reactive FBM metabolite (atropaldehyde) may have caused marrow toxicity,1 and a urine screening test has recently been developed to identify patients at risk. Other studies suggest that the presence of a rare HLA allele in the human MHC also may be a marker for patients at risk. The use of HLA typing to predict risk for the development of serious adverse effects has been reported among patients treated with clozapine.2 Methods: The Felbamate Patient Registry, with enrollment of more than 1000 patients, has been ongoing since July 1997 with the goal of gathering data that would allow the safest possible use of felbamate in patients with epilepsy. For patients enrolled in the Registry, stead-state urine specimens are analyzed to determine an acid monocarbamate to mercapturate ratio and, blood samples are taken for HLA typing. Results: In felbamate (FBM) metabolism, reactive atropaldehyde is normally conjugated with glutathione and then excreted in the urine as mercapturates. Patients with a 2 to 1 ratio of acid monocarbamate to mercapturate in the urine are considered within the normal range, whereas patients with a ratio four or more standard deviations (SD) above a 2 to 1 ratio may be considered at risk. The reproducibility of this ratio has been demonstrated in the more than 1000 patients currently in the Registry. Twelve (12) patients have had a SD greater than 4; one with a SD greater than 17 had FBM discontinued. HLA typing to date suggests that genes in the HLA locus are non-randomly associated with epilepsy genes in the HLA-F region of chromosome 6. Conclusion: The Felbamate Patient Registry continues to accumulate data and has provided valuable feedback to physicians treating patients on FBM. 1Thompson CD et al. Epilepsia 1999;40:769-776; 2Yunis JJ et al. Drug Safety 1992;7(suppl 1):7-9