Relationship Between the Free and Total Methotrexate Plasma Concentration in Children and Application to Predict the Toxicity of HD-MTX

Abstract

High-dose methotrexate (HD-MTX) can be highly effective as well as extremely toxic. Many drug molecules can bind to plasma proteins to different extents in vivo, whereas only the free drug can reach the site of action to exert a pharmacological effect and cause toxicity. However, free MTX concentrations in plasma have not been reported. Traditional analyses of free drugs are both cumbersome and inaccurate. We collected 92 plasma samples from 52 children diagnosed with ALL or NHL or other lymphomas that were treated with HD-MTX. The hollow fiber centrifugal ultrafiltration (HFCF-UF) was used to prepare plasma samples for analysis of the free MTX concentration. Protein precipitation was employed to measure the total MTX concentration. The HFCF-UF is a simple method involving a step of ordinary centrifugation; the validation parameters for the methodological results were satisfactory and fell within the acceptance criteria. A linearity coefficient r 2 of 0.910 was obtained for the correlation between the free and total MTX plasma concentrations in 92 plasma samples. However, the free and total MTX concentrations was only weakly correlated in 16 clinical plasma specimens with total MTX concentrations >2 μmol L−1 (r 2 = 0.760). Both the free and total MTX concentrations at 42 h were negatively correlated with the creatinine clearance (CCr) level (P = 0.023, r = −0.236 for total MTX and P = 0.020, r = −0.241for free MTX, respectively). The free MTX concentration could not be accurately estimated from the total MTX concentration for patients with high MTX levels which are conditions under which toxic reactions are more likely to occur. High plasma MTX levels could become a predictor of the occurrence of MTX nephrotoxicity to draw people s attention. The proposed HFCF-UF method is a simple and accurate way to evaluate efficacy and toxicity in clinical therapeutic drug monitoring.