Four fluorescence polarization immunoassays for therapeutic drug monitoring evaluated.

We evaluated four fluorescence polarization immunoassays--those for phenytoin, procainamide, N-acetylprocainamide (NAPA), and quinidine--from Roche Diagnostic Systems, done in the Cobas Bio FP centrifugal analyzer. The assays for phenytoin, NAPA, and quinidine demonstrated a linear response over the expected ranges of concentrations, and analytical recovery of test drug added to drug-free sera was greater than 90%. However, recovery in the procainamide assay was poor (69-82%) for samples containing greater than 8 mg/L, owing to nonlinearity. Results of method-comparison studies of the four assays paralleled the recovery studies, although the quinidine assay demonstrated a bias (1.0 mg/L higher) when compared with EMIT (Syva). The precision of the phenytoin assay was acceptable at all concentrations tested (total CV less than 7.0%). Imprecision of the other assays was significant at certain concentrations: total CV greater than 10.0% at subtherapeutic values for NAPA and quinidine, and greater than 9.0% for low (2.4 mg/L) and moderate concentrations (9.6 mg/L) of procainamide. Interferences were not significant for hemolyzed, icteric, or lipemic specimens or for specimens with added drug metabolites. The calibration curves for all four assays had good stability (greater than 60 days).