DISTURBANCES IN BIOTIN METABOLISM IN CHILDREN UNDERGOING LONG-TERM ANTICONVULSANT THERAPY

Background: In subjects undergoing long-term therapy with carbamazepin e and/or phenytoin, reduced plasma concentrations of biotin have been reported. However, the diagnostic value of plasma biotin is unclear, i n part because of the presence of significant plasma concentrations of biotin metabolites. Pathologic organic aciduria has also been reporte d with longterm anticonvulsant therapy, suggesting biotin deficiency, but no mechanism leading to deficiency has yet been determined. Method s: In the current study, we sought to determine whether biotin catabol ism was accelerated in children receiving longterm treatment with cert ain anticonvulsants and to assess biotin status as judged by urinary e xcretion of biotin and 3-hydroxyisovaleric acid, an organic acid that is an indicator of deficiency of a biotin-dependent enzyme. Seven chil dren treated with carbamazepine and/or phenytoin and six treated with phenobarbital provided untimed urine samples. Sixteen healthy children receiving no anticonvulsants served as controls. Biotin and biotin me tabolites were determined by highperformance liquid chromatography/avi din-binding assay. Urinary excretion of 3-hydroxyisovaleric acid was d etermined using gas chromatography/mass spectrometry. Results: Bisnorb iotin excretion was increased significantly in the carbamazepine/pheny toin group and in the phenobarbital group. Biotin sulfoxide excretion was significantly increased in the carbamazepine/phenytoin group but n ot in the phenobarbital group. 3-Hydroxyisovaleric acid excretion was increased significantly in the carbamazepine/phenytoin group. However, only one child (carbamazepine/phenytoin group) had a decreased urinar y excretion of biotin. Conclusion: These data provide evidence that lo ng-term administration of some anticonvulsants can accelerate biotin c atabolism. but the indicators of biotin status conflict.