Selection of the wavelength range and spectrophotometric determination of leucovorin and methotrexate in human serum by a net analyte signal based method

Leucovorin (LV) and methotrexate (MTX) were determined in human blood serum samples by using a model based in the net analytical signal concept. The calibration method used is a variation of the original hybrid linear analysis (HLA) method, developed by Goicoechea and Olivieri (HLA/GO). The calibration set was composed by nine serum samples with different amounts of LV and MTX in the range of 0–10 μg ml−1. The selection of the optimum wavelength range involved the calculation of the net analyte signal regression plot for each test sample, in conjunction with the calculation of the minimum error indicator. Relative errors of prediction (REP, %) of 3.0 and 5.3% were calculated for LV and MTX, respectively. Only two factors were necessary to optimize the proposed HLA/GO model. Sensitivity, selectivity, analytical sensitivity and limit of detection of the proposed procedure were calculated. Detection limits of 0.34 and 0.93 μg ml−1 for LV and MTX were determined. The proposed model was tested in the analysis of serum samples, without previous separation steps, obtaining recovery values between 96 and 99%, and between 92 and 103% for LV and MTX, respectively.