Effect of SLCO1B1 Polymorphisms on High-Dose Methotrexate Clearance in Children and Young Adults With Leukemia and Lymphoblastic Lymphoma

Abstract

High-dose (HD) methotrexate (MTX) is a critical component of treatment for hematologic malignancies in children and young adults. Therapeutic drug monitoring is necessary due to substantial interindividual variation in MTX clearance. Common function-altering polymorphisms in SLCO1B1 (encodes OATP1B1, which transports MTX) may contribute to clearance variability. We performed pharmacokinetic modeling using data for 106 children and young adults treated with HD MTX for hematologic malignancies; of 396 total courses of HD MTX, 360 consisted of 5 g/m² over 24 hours. We evaluated the contribution of clinical covariates and SLCO1B1 genotype (388A>G and 521T>C) to MTX clearance variability. Of the clinical covariates studied, patient weight improved the pharmacokinetic model most significantly (P < 0.001). The addition of the SLCO1B1 variants individually further improved the model (P < 0.05 for each). An interaction between these variants was suggested when both were included (P = 0.017). SLCO1B1 genotype should be considered in efforts to personalize HD MTX dosing.

Figure 1

Scatter plots for model‐predicted individual methotrexate (MTX) clearance (CL) by the base model. (a) Individual clearance vs. weight. (b) Individual CL vs. age. (c) Weight‐normalized individual CL vs. weight. (d) Weight‐normalized individual CL vs. age.

Figure 2

Population pharmacokinetic model diagnostic plots and predictive model checking for the final model, model 2. (a) Observed methotrexate (MTX) concentrations (in micromoles per liter (mcmol/L)) vs. population predicted concentrations. The gray diagonal line is a line of identity, whereas the black dashed line is a local polynomial regression fitting line for the data to show a trend. (b) Observed MTX concentrations vs. individual predicted concentrations. The gray diagonal line and the black dashed line are the same as in panel a. (c) Conditional weighted residuals (CWRES) vs. population predicted concentrations. The black dashed line is a local polynomial regression fitting line for the data to show a trend. (d) Predictive model check, where the lines represent the 5th (lower dashed line), 50th (middle solid line), and 95th (upper dashed line) percentiles of observed MTX concentrations; gray areas represent the 90th percentile region of predicted MTX concentrations from the final model using simulations. The main plot corresponds to the highlighted region of data in the inset.

Figure 3

Box plots for model‐predicted individual methotrexate (MTX) clearance (CL). Model‐predicted individual MTX CL over maximum four courses in pediatric patients with acute lymphoblastic leukemia receiving high‐dose MTX (5 g/m2). Model 1: by genotype for SLCO1B1 521T>C allele, left: wild‐type (TT), right: any variant (CC or CT) – (a) shows weight‐normalized clearance (CL divided by weight in kg) and (b) shows CL without weight normalization. Model 2: by genotype for SLCO1B1 521T>C and 388A>G, left: wild‐type for 521T>C (TT) separated by 388A>G genotype (wild‐type = AA), right: any variant for 521T>C (CC or CT) separated by 388A>G (wild‐type = AA) – (c) shows weight‐normalized clearance (CL divided by weight in kg) and (d) shows CL without weight normalization. Top and bottom of boxes represent 75th and 25th percentile of the model‐predicted individual MTX CL, respectively, whereas the thick horizontal line represents the median. The dashed vertical lines are the whiskers extending to the most extreme data point, which is no more than 1.5 times the interquartile range from the box.