Serum Creatinine Versus Plasma Methotrexate Levels to Predict Toxicities in Children Receiving High-dose Methotrexate

Facilities for measuring methotrexate (MTX) levels are not available everywhere, potentially limiting administration of high-dose methotrexate (HDMTX). We hypothesized that serum creatinine alteration after HDMTX administration predicts MTX clearance. Overall, 122 cycles in 50 patients of non-Hodgkin lymphoma or acute lymphoblastic leukemia aged ≤18 years receiving HDMTX were enrolled prospectively. Plasma MTX levels were measured at 12, 24, 36, 48, 60, and 72 hours; serum creatinine was measured at baseline, 24, 48, and 72 hours. Correlation of plasma MTX levels with creatinine levels and changes in creatinine from baseline (Δ creatinine) were evaluated. Plasma MTX levels at 72 hours showed positive correlation with serum creatinine at 48 hours (P = .011) and 72 hours (P = .013) as also Δ creatinine at 48 hours (P = .042) and 72 hours (P = .045). However, cut-off value of either creatinine or Δ creatinine could not be established to reliably predict delayed MTX clearance. Greater than 50% Δ creatinine at 48 and 72 hours significantly predicted grade 3/4 leucopenia (P = .036 and P = .001, respectively) and thrombocytopenia (P = .012 and P = .009, respectively) but not mucositis (P = .827 and P = .910, respectively). Delayed MTX elimination did not predict any grade 3/4 toxicity. In spite of demonstration of significant correlation between serum creatinine and Δ creatinine with plasma MTX levels at 72 hours, cut-off value of either variable to predict MTX delay could not be established. Thus, either of these cannot be used as a surrogate for plasma MTX estimation. Interestingly, Δ creatinine effectively predicted hematological toxicities, which were not predicted by delayed MTX clearance.