Interference in immunoassay measurements of total and free phenytoin in uremic patients: a reappraisal.

Accumulation of phenytoin metabolites in uremia has been shown to interfere with some immunoassay methods for phenytoin measurement. We evaluated the effects of uremia (serum creatinine > 13 mg/L) on free and total phenytoin concentrations measured by the Abbott TDx fluorescence polarization immunoassay, and the DuPont aca and Syva EMIT homogeneous enzyme-multiplied immunoassay methods, using HPLC as the comparison method. In uremic patients, the TDx assay showed both fixed and proportional bias in comparison with the HPLC for both total phenytoin concentration (TDx = 1.24 x HPLC + 1.9 mg/L) and free phenytoin concentration (TDx = 1.52 x HPLC + 0.24 mg/L). The total bias was as great as 100% for both total and free phenytoin. Cross-reactive substances other than 5-(p-hydroxyphenyl)-5-phenylhydantoin (HPPH) and HPPH-glucuronide appeared to be responsible. In contrast, there was minimal interference with aca and EMIT assays.