Interference of 7-hydroxymethotrexate with the determination of methotrexate in plasma samples from children with acute lymphoblastic leukemia employing routine clinical assays

The accuracy of two clinical assays, the enzyme-multiplied immunoassay (EMIT) and fluorescence polarization immunoassay (FPIA2), universally employed for measurement of plasma levels of methotrexate (MTX) in children administered a high dose of this drug for treatment of acute lymphoblastic leukemia was evaluated here. Because of its superior specificity, sensitivity, and precision, high performance liquid chromatography (HPLC) was selected as the reference method with which the other two procedures were compared using approximately 420 different plasma samples for method comparison. 7-Hydroxymethotrexate (7-OHMTX), the major plasma metabolite of MTX, that can be detected in plasma at relatively high concentrations for long periods following infusion of a high dose of MTX, was also quantitated by HPLC. Forty-two and 66 h after infusion, the plasma level of MTX was overestimated in 2% and 3% of the samples by the FPIA2 procedure in 5% and 31% by the EMIT assay. The overall correlation coefficients (r(2)) for the values obtained by FPIA2 or EMIT versus those based on HPLC were 0.989 and 0.663, respectively. The presence of 7-OHMTX exerted a highly significant influence (p = 0.0007 as determined by the unpaired t-test) on MTX measurement by the EMIT assay. We conclude that the rapid automated procedures routinely used at present and in particular EMIT, suffer from cross-reactivity with metabolites of MTX. Thus, the relatively high percentage of samples in which the level of MTX is overestimated at check-points by EMIT may result in longer periods of hospitalization, higher costs and prolonged administration of elevated doses of rescue leucovorin with an increased risk for relapse. (C) 2004 Elsevier B.V. All rights reserved.