# Novel and replicated clinical and genetic risk factors for toxicity from high-dose methotrexate in pediatric acute lymphoblastic leukemia

## Metadata
**Authors:** Mark Zobeck, M Brooke Bernhardt, Kala Y Kamdar, Karen R Rabin, Philip J Lupo, Michael E Scheurer
**Journal:** Pharmacotherapy
**Date:** 2023 Feb 27
**DOI:** [10.1002/phar.2779](https://doi.org/10.1002/phar.2779)
**PMID:** 36764694
**PMCID:** PMC10085626
**URL:** https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10085626/
**PDF:** [https://pmc.ncbi.nlm.nih.gov/articles/PMC10085626/pdf/nihms-1884162.pdf](https://pmc.ncbi.nlm.nih.gov/articles/PMC10085626/pdf/nihms-1884162.pdf)

## Abstract

Methotrexate (MTX) is a key component of treatment for high-risk pediatric acute lymphoblastic leukemia (ALL) but may cause acute kidney injury and prolonged hospitalization due to delayed clearance. We conducted a retrospective cohort study of pediatric patients with ALL who received 4000–5000 mg/m2 of MTX to identify clinical and genetic factors that may predict which children are at risk for creatinine increase and prolonged MTX clearance. We performed germline genotyping to determine genetic ancestry and allele status for 49 single nucleotide polymorphisms (SNPs) identified from the literature as related to MTX disposition. Bayesian hierarchical ordinal regression models for creatinine increase and for prolonged MTX clearance were developed. Hispanic ethnicity, body mass index (BMI) < 3%, BMI between 85–95%, and Native American genetic ancestry were found to be associated with an increased risk for creatinine elevation. Older age, Black race, and use of the intensive monitoring protocol were associated with a decreased risk for creatinine elevation. Older age, B- compared to T-ALL, and the minor alleles of rs2838958/SLC19A1 and rs7317112/ABCC4 were associated with an increased risk for delayed clearance. Black race, MTX dose reduction, and the minor allele of rs2306283/SLCO1B1 were found to be associated with a decreased risk for delayed clearance. These predictors of MTX toxicities may allow for more precise individualized toxicity risk prediction.

Keywords: Methotrexate, leukemia, nephrotoxicity, pharmacogenomics, pediatrics

**Keywords:**Keywords: Methotrexate, leukemia, nephrotoxicity, pharmacogenomics, pediatrics

## Introduction

Methotrexate (MTX) is a key component of treatment for pediatric acute lymphoblastic leukemia (ALL). Many effective ALL regimens include multiple administrations of MTX in doses that may cause life-threatening adverse effects such as gastrointestinal inflammation, hepatitis, kidney injury, and myelosuppression. These toxic side effects may increase the risk of infections, organ dysfunction, treatment delays, and poor treatment outcomes. MTX doses ≥ 1000 mg/m^2^2 are commonly given in the hospital with aggressive supportive care and therapeutic drug monitoring to decrease these risks. Many patients do not clear the drug as expected, which results in prolonged hospitalizations and increased health care costs.

Due to the frequency and severity of MTX-associated toxicity, much work has been done to describe the individual sociodemographic,^1,2^[1](#R1)1,[2](#R2)2 clinical ^3,4^[3](#R3)3,[4](#R4)4 and germline genetic ^5–7^[5](#R5)5–[7](#R7)7 risk factors for toxicity. Further, many toxicity risk-prediction models of MTX have been developed from these data.^8,9^[8](#R8)8,[9](#R9)9 Despite these efforts, clinically useful models for predicting toxicity are still nascent, and considerable interindividual variability remains unaccounted for in the pharmacokinetic models, suggesting further research into risk factors for toxicity is needed to improve the precision of predictions. Racial and ethnic subpopulations with differential toxicity rates from MTX, such as Hispanic patients, ^10,11^[10](#R10)10,[11](#R11)11 are not well represented in prior analyses, introducing potential bias in model predictions.

To continue to improve our knowledge of sociodemographic, clinical, and germline genetic predictors of MTX toxicity and broaden our understanding of the effects of toxicity among at-risk populations, we conducted a retrospective cohort study of children with ALL treated with high-dose MTX at Texas Children’s Hospital (TCH), a quaternary care hospital in Houston, Texas that serves a large multi-ethnic population.

## Methods

### The Clinical Cohort

Every patient with ALL treated at TCH who received high-dose MTX, defined as 5000 mg/m^2^2 for children ≥ 1 year old and 4000 mg/m^2^2 for children < 1 year, from September 2010 to December 2017, was extracted from the electronic medical record. Demographic and clinical information, including age, sex, body mass index (BMI) before each dose, leukemia type, and MTX administration records, was extracted from the electronic medical record and processed for analysis. The charts of approximately 10% of patients were manually reviewed after data processing to ensure accuracy. Data about self-reported race and ethnicity were collected using the National Institutes of Health standard self-report form or through chart review. BMI was measured as a percentile-for-age on the World Health Organization growth charts. Leukemia types included B-ALL, T-ALL, and infant ALL. The analysis of the Clinical Cohort was conducted under an Institutional Review Board -approved retrospective chart review with a waiver of consent.

MTX was administered per the patient’s treatment protocol. Most patients received an intravenous bolus of 10% of the total dose over 30 minutes and then the remaining amount over 23.5 hours. MTX levels were obtained at 24 hours, 42 hours, and 48 hours from the start of the infusion and every 12 hours after the 48-hour level if predefined clearance thresholds were not met. Some patients received MTX according to an institutional Intensive Monitoring Protocol (IMP),^12^[12](#R12)12 where multiple levels in the first 24 hours were obtained and the MTX infusion rate was adjusted according to the IMP. Clinicians may also have administered MTX with a dose reduction due to past toxicity. Per the treatment protocol, the usual and customary supportive care (hydration, alkalinization, leucovorin rescue, antiemetics) was provided.

### The Germline Genomic Cohort

Genotype data were available for a subset of patients from the Clinical Cohort under an IRB-approved protocol requiring families’ consent and assent when appropriate. DNA was extracted from blood samples using the PerkinElmer (PerkinElmer, Inc., Waltham, MA, USA) Prepito instrument. Each singe nucleotide polymorphism (SNP) was genotyped using the Illumina (Illumina, Inc., San Diego, CA, USA) Infinium Global Screening Array according to manufacturer protocols. We estimated genetic ancestry using STRUCTURE version 2.3.4 software by assuming an admixture model with European, African, East Asian, or Native American underlying ancestral subpopulations using HapMap Phase 3 reference populations.^13^[13](#R13)13

### SNP selection

We searched the Pharmacogenomics Knowledge Base (PharmGKB) and reviewed the literature to identify candidate SNPs related to MTX pharmacokinetics and pharmacodynamics.^14^[14](#R14)14 We queried PharmGKB in September 2018 and found 38 SNPs specifically associated with toxicity in ALL patients. We identified 11 SNPs with evidence of activity in Hispanic populations from two recent genome-wide association studies ([Table S1](#SD1)Table S1 for complete SNP list; [Table S2](#SD1)Table S2 for genotype coding^15,16^[15](#R15)15,[16](#R16)16).

### Primary outcomes

The primary outcomes of interest were creatinine increase, measured as a ratio of peak to baseline (pre-infusion) creatinine for each MTX dose, and clearance time, measured in 6-to-12-hour increments 24 hours from the start of MTX infusion. For this study, 48 hours was the earliest time that MTX clearance could be achieved, but all patients had to stay in the hospital through hour 72 for intravenous fluids. If a patient had not cleared MTX by hour 72, they remained inpatient and continued to receive fluids and leucovorin until clearance was achieved.

### Statistical Analysis

Descriptive statistics were produced for all variables. We classified creatinine increases according to the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0 for descriptive purposes.^17^[17](#R17)17 Directed Acyclic Graphs (DAGs) were developed for each clinical and genetic predictor for the outcomes of creatinine increase and time to MTX clearance ([Figure S1](#SD1)Figure S1). They were analyzed to identify covariates to include in a model for each outcome ([Table S3](#SD1)Table S3 for the adjustment set of each predictor). Multivariable, hierarchical Bayesian ordinal regression models were developed to estimate the effects of interest for each outcome ([Appendix B](#SD1)Appendix B for further details). To compare relative predictive abilities for models with different variables, the differences in expected logarithmic pointwise predictive density estimated via leave-one-out cross validation (ELPD-LOO) was calculated.^18^[18](#R18)18 Analyses were performed using R version 4.1.0.^19^[19](#R19)19 The Bayesian models were developed and visualized using the packages “brms,” “ggplot2,” “tidybayes,” and “bayesplot.”^20–23^[20](#R20)20–[23](#R23)23

## Results

There were 375 patients in the clinical cohort who were treated for ALL between September 2010 and December 2017. Thirty-seven percent of these patients were female, 58% were self-reported Hispanic ethnicity, 78% had a diagnosis of B- ALL, and 4% had infant ALL ([Table 1](#T1)Table 1). The clinical cohort received a total of 1,308 doses of MTX. Of these, 112 (8.6%) were administered with a dose reduction of 20–80%, and 133 (10.1%) were administered according to the IMP. Further dose characteristics are presented in [Table 2](#T2)Table 2.

### Table 1.

Patient Characteristics

| Variable | Value (Total Patients = 375) |
| -------- | ---------------------------- |
|  |  |
| Age (years), Median (IQR) | - | 9.7 (4.8, 13.7) |
|  |  |
| Sex, N (%) | Female | 143 (38%) |
|  |  |
| Male | 232 (62%) |
|  |  |
| Ethnicity, N (%) | Non-Hispanic | 157 (42%) |
|  |  |
| Hispanic | 218 (58%) |
|  |  |
| Race, N (%) | Asian | 25 (6.8%) |
|  |  |
| Black | 37 (10%) |
|  |  |
| Native American | 13 (3.5%) |
|  |  |
| White | 293 (80%) |
|  |  |
| Unknown | 7 (%) |
|  |  |
| BMI at first dose, N (%) | Less than 3% | 34 (9.1%) |
|  |  |
| 3–10% | 34 (9.1%) |
|  |  |
| 10–85% | 195 (52%) |
|  |  |
| 85–95% | 57 (15.4%) |
|  |  |
| Greater than 95% | 52 (14.1%) |
|  |  |
| Disease type, N (%) | B ALL | 293 (78%) |
|  |  |
| T ALL | 67 (18%) |
|  |  |
| Infant ALL | 15 (4.0%) |
|  |  |
| Protocol Dose (mg/m2), N (%) | 4000 | 16 (4.3%) |
|  |  |
| 5000 | 359 (96%) |
### Table 2.

Dose Characteristics

| Characteristic | N (%)(Total Doses = 1 308) |
| -------------- | -------------------------- |
|  |  |
| Dose Number | 1 | 367 (28%) |
|  |  |
| 2 | 325 (25%) |
|  |  |
| 3 | 291 (22%) |
|  |  |
| 4 | 275 (21%) |
|  |  |
| ≥5 | 50 (3.8%) |
|  |  |
| Dose Reduction | 0% | 1 196 (91%) |
|  |  |
| 20% | 34 (2.6%) |
|  |  |
| 25% | 61 (4.7%) |
|  |  |
| 40% | 4 (0.3%) |
|  |  |
| 50% | 9 (0.7%) |
|  |  |
| 60% | 2 (0.2%) |
|  |  |
| 80% | 2 (0.2%) |
|  |  |
| Intensive Monitoring Protocol | Not used | 1 175 (90%) |
|  |  |
| Used | 133 (10%) |
For all diagnosis types, 117 doses (8.9%) resulted in a CTCAE grade 2, five in grade 3, and one in grade 4 creatinine elevation ([Table 3](#T3)Table 3). Clearance was delayed beyond 72 hours from the start of MTX infusion in 36.1% of doses (472/1308). Among the patients with B- and T-ALL, this resulted in 550 extra hospital days for the cohort, 15.7% (550/3495) of the total observed hospital days (patients with infant ALL were excluded because they remain inpatient after receiving MTX). For patients who experienced delayed clearance, the median number of extra hospital days was 2, with an interquartile range of 1 – 4 days and a maximum of 16 days.

### Table 3.

Clinical Outcome for methotrexate doses

| Variable | Clinical Classification | Category | N (%) |
| -------- | ----------------------- | -------- | ----- |
|  |  |  |  |
| Creatinine increase from baseline(Total doses = 1 308) | No toxicity | < 25% | 965 (74%) |
|  |  |  |  |
| Clinically actionable elevation | ≥ 25% to < 50% | 220 (17%) |  |
|  |  |  |  |
| CTCAE1 Grade 2 | ≥ 50% to < 75% | 79 (6.0%) |  |
|  |  |  |  |
| ≥ 75% to < 100% | 12 (0.9%) |  |  |
|  |  |  |  |
| ≥ 100% to < 200% | 26 (2.0%) |  |  |
|  |  |  |  |
| CTCAE Grade 3–4 | ≥ 200% to < 300% | 2 (0.2%) |  |
|  |  |  |  |
| ≥ 300% | 4 (0.3%) |  |  |
|  |  |  |  |
| Methotrexate clearance times(Total doses = 1 308) | Ideal clearance time | 48-h | 836 (64%) |
|  |  |  |  |
| No extra hospital days | > 48-h to = 72-h | 139 (11%) |  |
|  |  |  |  |
| Extra hospital days | >72-h to = 96-h | 190 (15%) |  |
|  |  |  |  |
| >96-h to = 120-h | 67 (5.1%) |  |  |
|  |  |  |  |
| > 120-h | 76 (5.8%) |  |  |
In multivariable regression models for creatinine increase and nephrotoxicity ([Table 4](#T4)Table 4), Hispanic ethnicity, BMI < 3%, and BMI between 85–95% were associated with an increased risk for creatinine increase. Older age, Black race, and use of the IMP were associated with a decreased risk for creatinine increase. Older age and a diagnosis of B- compared to T-ALL were associated with an increased risk for delayed clearance. Black race and MTX dose reductions were associated with a decreased risk for delayed clearance ([Table 4](#T4)Table 4).

### Table 4.

Results of multivariable1 Bayesian hierarchical ordinal regression models for increased creatinine and for delayed MTX clearance

|  | Variable | Reference Level | Increased creatinine | Delayed clearance |
| - | -------- | --------------- | -------------------- | ----------------- |
|  |  |  |  |  |
| OR2 | 95% CI2 | OR | 95% CI |  |
|  |  |  |  |  |
| Sex | Female | Male | 1.23 | 0.92 – 1.66 | 1.20 | 0.90 – 1.64 |
|  |  |  |  |  |
| BMI | < 3% | BMI 10–85% | 1.81 | 1.08 – 2.99 4 | 1.16 | 0.70 – 1.92 |
|  |  |  |  |  |
| 3–10% | 0.95 | 0.55 – 1.61 | 1.08 | 0.66 – 1.78 |
|  |  |  |  |  |
| 85–95% | 1.54 | 1.01 – 2.37 | 1.07 | 0.73 – 1.59 |
|  |  |  |  |  |
| > 95% | 0.89 | 0.56 – 1.40 | 1.02 | 0.68 – 1.56 |
|  |  |  |  |  |
| Increasing Age | Age (standardized)3 | - | 0.80 | 0.68 – 0.95 | 1.61 | 1.37 – 1.89 |
|  |  |  |  |  |
| Race | Black race | White | 0.64 | 0.40 – 0.99 | 0.64 | 0.40 – 1.065 |
|  |  |  |  |  |
| Asian | 0.83 | 0.47 – 1.41 | 0.74 | 0.43 – 1.28 |
|  |  |  |  |  |
| Native American | 0.72 | 0.32 – 1.53 | 0.63 | 0.29 – 1.35 |
|  |  |  |  |  |
| Unknown | 0.81 | 0.30 – 2.07 | 0.28 | 0.08 – 0.86 |
|  |  |  |  |  |
| Ethnicity | Hispanic ethnicity | Non-Hispanic | 1.42 | 1.03 – 1.92 | 1.14 | 0.86 – 1.52 |
|  |  |  |  |  |
| Dose number | Dose 2 | Dose 1 | 0.81 | 0.58 – 1.14 | 0.83 | 0.61 – 1.34 |
|  |  |  |  |  |
| Dose 3 | 1.31 | 0.81 – 1.60 | 0.85 | 0.62 – 1.17 |
|  |  |  |  |  |
| Dose 4 | 0.63 | 0.44 – 0.92 | 0.96 | 0.70 – 1.34 |
|  |  |  |  |  |
| Dose 5 | 1.06 | 0.48 – 2.32 | 1.55 | 0.82 – 3.02 |
|  |  |  |  |  |
| Disease type | T-ALL | B-ALL | 0.98 | 0.65 – 1.48 | 0.63 | 0.42 – 0.92 |
|  |  |  |  |  |
| Infant ALL | 1.42 | 0.71 – 2.85 | 1.46 | 0.67 – 3.16 |
|  |  |  |  |  |
| Administration | Intensive Monitoring Protocol | Routine Monitoring | 0.42 | 0.24 – 0.76 | 0.72 | 0.45 – 1.12 |
|  |  |  |  |  |
| Dosing | Dose Reduction | Normal dosing | 0.82 | 0.48 – 1.40 | 0.65 | 0.40 – 1.05 |
Germline genetic information was available for 154 patients in the clinical cohort who received 594 total doses of MTX (demographic and dose characteristics are available in [Tables S4](#SD1)Tables S4 and [S5](#SD1)S5). On multivariable regression with genetic admixture estimates for Asian, African, and native American (NA) genetic ancestry, NA ancestry had an odds ratio (OR) for increased creatinine clearance of 1.08 (95% compatibility interval [CI] 1.02–1.15) for every 10% increase in NA genetic ancestry. No other genetic ancestry variable was associated with increased creatinine or delayed clearance (full results in [Table S6](#SD1)Table S6).

The 49 SNPs with previous evidence of association with MTX toxicity in the literature were assessed using multivariable ordinal regression, which included adjustments for genetic ancestry, dose reduction, and use of the IMP. [Figure 1](#F1)Figure 1 depicts the 95% CIs for the coefficients of each SNP for the outcomes of creatinine increase and delayed clearance. [Table 5](#T5)Table 5 highlights the SNPs that were likely or clearly associated with the outcomes. [Figures 2](#F2)Figures 2 depicts the outcomes by genotype for the SNPs with the clearest evidence of association. SNPs in the genes *GGH*GGH, *SLC19A1,*SLC19A1, and *MTHFD1*MTHFD1 demonstrated likely associations with creatinine increase ([Figure 2](#F2)Figure 2, left panel). SNPs in the genes *SLC19A1, ABCC4,*SLC19A1, ABCC4, and *SLCO1B1*SLCO1B1 were strongly associated with delayed MTX clearance ([Figure 2](#F2)Figure 2, right panel). SNPs in the genes *PACSIN2*PACSIN2 and *CCND1*CCND1 were likely associated with delayed clearance.

### Figure 1. Coefficient results from the multivariable, hierarchical Bayesian ordinal regression models for each SNP.

![Figure 1.](https://cdn.ncbi.nlm.nih.gov/pmc/blobs/2627/10085626/c3edd2d49d2f/nihms-1884162-f0001.jpg)

[View larger image](https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=10085626_nihms-1884162-f0001.jpg)

The association with creatinine increase is in the left panel and delayed MTX clearance in the right. Values are on the log-odds scale. Median coefficient value (circle), 50% compatibility interval (thick line segment), and 95% compatibilty interval (thin line segment) are shown. MTX, methotrexate.

### Table 5.

Results of the Bayesian hierarchical logistic regression model of risk alleles for creatinine increase from baseline and prolonged MTX clearance.1

| SNP | OR | 95% CI | Minor Allele | Reference Allele | Associated Gene | Previous Toxicity Association2 |
| --- | -- | ------ | ------------ | ---------------- | --------------- | ------------------------------ |
| Creatinine increase |  |  |  |  |  |  |
| rs2838958 | 1.35 | 0.99 – 1.81 | A | G | SLC19A1 | Patients with the AA genotype may have a worse response to methotrexate as compared to patients with the AG and GG genotype. |
| rs2236225 | 0.77 | 0.58 – 1.00 | A | G | MTHFD1 | Although conflicting evidence exists, the AA genotype was associated with decreased event-free survival and decreased risk for hepatotoxicity compared to the GG genotype. |
| rs11545078 | 0.63 | 0.34 – 1.04 | A | G | GGH | Patients with the AA genotype may have increased accumulation of active methotrexate metabolites and increased risk for thrombocytopenia as compared to patients with the GG genotype. |
| Prolonged methotrexate clearance |  |  |  |  |  |  |
| rs2838958 | 1.73 | 1.24 – 2.38 | A | G | SLC19A1 | Patients with the AA genotype may have a worse response to methotrexate as compared to patients with the AG and GG genotype. |
| rs7317112 | 1.70 | 1.18 – 2.43 | G | A | ABCC4 | Patients with the AA genotype may have an increased risk for mucositis compared to patients with the AG or GG genotype. |
| rs9344 | 1.34 | 0.97 – 1.83 | A | G | CCND1 | Although conflicting evidence exists, patients with the AA genotype may have decreased risk for toxicity compared to patients with the AG or GG genotypes. |
| rs2413739 | 0.77 | 0.54 – 1.05 | T | C | PACSIN2 | Patients with the TT genotype may have an increased risk for GI toxicity compared to patients with the CT or TT genotype. |
| rs2306283 | 0.72 | 0.53 – 0.97 | G | A | SLCO1B1 | Patients with the GG genotype may have decreased exposure to MTX compared to AA genotype. |
### Figure 2. Distribution of values for creatinine increase from baseline (left) and MTX clearance hours (right) by genotype for the three SNPs with the largest association.

![Figure 2.](https://cdn.ncbi.nlm.nih.gov/pmc/blobs/2627/10085626/43feb6092e5d/nihms-1884162-f0002.jpg)

[View larger image](https://www.ncbi.nlm.nih.gov/core/lw/2.0/html/tileshop_pmc/tileshop_pmc_inline.html?title=Click%20on%20image%20to%20zoom&p=PMC3&id=10085626_nihms-1884162-f0002.jpg)

The grey shaded area shows the density of distribution of creatinine increase values. The points underneath the shaded area represent individual dose values. The right shoulder of the box represents the 75% for the data values. The right tip of the thin line segment represents 1.5 times the interquartile range beyond the 75%. Thick vertical lines signify that all values contained in the box and whiskers fall on the ≤ 25% creatinine increase category or the 48-hour clearance time for MTX for the left and right plots, respectively. MTX, methotrexate; SNPs, single nucleotide polymorphisms.

A model for delayed clearance was fit using the NA genetic ancestry variable, five best-performing SNPs ([Table 5](#T5)Table 5), and clinical variables (dose number, disease type, age, use of IMP, and dose reductions; [Table S7](#SD1)Table S7). The combined clinical and genetic model outperformed models that included only clinical information and only genetic information when ELPD was compared via leave-one-out cross-validation (difference of −8.6 [standard error (SE) 4.8] for combined model vs. clinical-only model; −9.2 [SE 4.0] for combined vs. SNP-only model). This is evidence that the combined model may have superior out-of-sample predictive abilities for delayed clearance. The combined genetic and clinical model for creatinine increase, including the top three performing SNPs ([Table 5](#T5)Table 5), showed less pronounced differences in leave-one-out performance compared to clinical-only (−3.2 [SE 2.3]) and genetic-only (−5.8 [SE 4.2]). The smaller differences suggested similar out-of-sample predictive abilities between the models for creatinine increase.

## Discussion

This study provided a comprehensive assessment of the clinical features and associated risk factors for nephrotoxicity in a pediatric ALL cohort treated with over 1300 doses of intravenous MTX. We found that although more than 9% of doses resulted in a CTCAE grade 2 or higher creatinine elevation, only 0.5% of doses resulted in grade ≥3 toxicity, with five patients suffering CTCAE grade 3 toxicities (serum creatinine > 3x baseline) and one grade 4 (> 6x baseline). This rate is similar to previously published estimates of severe acute kidney injury (AKI) in pediatric ALL patients treated with high-dose MTX ranging from 0.5 to 1%.^24,25^[24](#R24)24,[25](#R25)25

We found that 25% of MTX courses required extra hospitalization days, resulting in 550 additional hospital days that comprised 15.7% of the total hospital time in the dataset. As this estimate assumes that patients would be discharged at the time of clearance and ignores extra hospital time due to complications from MTX, it may be seen as a lower bound on the total extra hospital time due to MTX infusions. This estimate also assumes that the earliest a patient who clears MTX at 48 hours may be discharged is at 72 hours, consistent with institutional practice during the study interval. This is an important measure of the cumulative burden of MTX toxicity and represents an important target for reduction as researchers seek to reduce not only clinical toxicities of pediatric ALL treatment but financial toxicity as well.^26,27^[26](#R26)26,[27](#R27)27

Of the clinical risk factors, BMI <3% (OR 1.81 [95%CI 1.08–2.99]) and BMI between 85–95% (OR 1.54 [95%CI 1.01–2.37]) were associated with increased risk for creatinine elevation. It is noteworthy that BMI >95% did not demonstrate an association with creatinine increase, which obscures a clear biological interpretation of the finding for the BMI 85–95% category. A large study from St. Jude Children’s Research Hospital of 621 children with ALL failed to show an association between BMI and MTX clearance, intracellular accumulation of the active metabolite MTX polyglutamate (MTX-PG), survival or relapse, although AKI as an outcome was not included in this study.^28^[28](#R28)28 Several other studies have demonstrated that decreased serum total protein or albumin are risk factors for MTX toxicity, which may plausibly mediate an effect of low weight on MTX-induced toxicity.^29–35^[29](#R29)29–[35](#R35)35 Malnourished patients with BMI<3% may be at risk for altered pharmacokinetics due to changes in the level of plasma proteins, or alterations in hepatic or renal function. This may affect drug disposition and/or may influence the perceived toxicity of the drug. However, the pharmacokinetics of chemotherapy in malnourished pediatric patients are not understood and are vastly understudied. As a protein-bound drug, the pharmacokinetics of MTX may be significantly altered in malnourished patients, placing them at risk for increased toxicities. Given that over 80% of the children with cancer live in low- and middle-income countries with significant malnutrition rates, malnutrition likely represents an important and modifiable risk factor for MTX-induced toxicity.

Increasing patient age was associated with a decreased risk of creatinine increase (OR 0.80 [95%CI 0.68–0.95]) but an increased risk for delayed clearance (OR 1.61 [95%CI 1.37–1.89]). We have previously demonstrated that increasing age was associated with increased risk for receiving Carboxypeptidase G_2_2 (CPDG 2; glucarpidase), a rescue medicine for severe nephrotoxicity or high serum MTX levels.^11^[11](#R11)11 Other studies have shown that increasing age is associated with decreased MTX clearance^36,37^[36](#R36)36,[37](#R37)37 but increased risk for AKI^29,36^[29](#R29)29,[36](#R36)36 and liver toxicity^36^[36](#R36)36. Contrary to this, our study showed a lower risk of creatinine increase for increasing age. Supplementary investigations into the interaction of creatinine change and age to predict MTX clearance time demonstrated a strong association with the interaction (OR 4.31 [95% CI 2.61–6.90]), implying that older children are at higher risk for delayed clearance for a given change in creatinine compared to younger children. More work is needed to understand how age interacts with each outcome.

The IMP was associated with a decreased risk of increased creatinine (OR 0.42 [95% CI 0.24–0.76]), but the effect on MTX clearance time was equivocal (OR 0.72 [95% CI 0.45–1.12]). The IMP was evaluated in a trial at our institution in 22 pediatric patients with leukemia with a previous history of nephrotoxicity due to MTX. None of the 54 doses of MTX resulted in a CTCAE grade 2 or higher kidney injury.^12^[12](#R12)12 Our results in the current report are similar in that 4 of the 133 doses (3%) resulted in a CTCAE grade 2 or higher toxicity. These results provide essential support for the claim that the IMP is effective at reducing kidney injury in routine clinical practice. A fixed-dose reduction is another common approach following severe MTX toxicity. In our sample, 85% of dose reductions were either 20% or 25% reductions in the original dose. This practice showed a likely reduced risk for delayed clearance (OR 0.65 [95% CI 0.40–1.05]) but no clear relationship with risk for increased creatinine.

Hispanic ethnicity was associated with an increased risk for creatinine increase (OR 1.42 [95% CI 1.03–1.92]). We have previously shown in this sample that Hispanic patients were at much higher risk for requiring CPDG_2_2 than non-Hispanic patients.^11^[11](#R11)11 Hispanic ethnicity has also been associated with higher 24-hour MTX levels and a higher incidence of MTX-induced neurotoxicity.^1,38^[1](#R1)1,[38](#R38)38 Less evidence is available about racial differences in MTX toxicity, although individuals who identify as Black or Hispanic have long been found to have inferior survival outcomes.^39^[39](#R39)39 Black compared to White race was associated with a decreased risk for creatinine increase (OR 0.64 [95% CI 0.4–0.99]) and a likely decreased risk for delayed clearance (OR 0.64 [95% CI 0.40–1.06]). Although many factors contribute to these disparities, these differences in racial and ethnic groups suggest the possibility that genetic factors are an etiologic factor in toxicities. Of the genetic ancestry variables we evaluated, NA was shown to be associated with an increased risk for creatinine increase (OR 1.08 [95% CI 1.02–1.15] for every 10% increase in NA genetic ancestry). In our study, NA ancestry correlated with self-described Hispanic ethnicity (Pearson correlation 0.84). These findings are consistent with our results for Hispanic ethnicity and risk for creatinine increase ([Table 4](#T4)Table 4) and our previous results that Hispanic ethnicity and NA ancestry are associated with an increased risk for CPDG_2_2 use_._.^11^[11](#R11)11 These findings may indicate that pharmacogenomic alterations that cluster in the Hispanic population confer increased risk for treatment-induced toxicity and poor outcomes.

We identified one SNP associated with delayed MTX clearance and likely associated with increased creatinine. Each additional A allele of rs2838958, an intronic variant of *SLC19A1,*SLC19A1, demonstrated an OR of 1.73 (95% CI 1.24–2.38) for prolonged MTX clearance and 1.35 (95% CI 0.99–1.81) for increased creatinine ([Table 5](#T5)Table 5 and [Figures 2](#F2)Figures 2). *SLC19A1*SLC19A1 is a gene that encodes the reduced folate carrier protein 1 (RFC1) that, among other functions, actively transports MTX into leukemia and nonmalignant cells such as the intestinal epithelium, hepatocytes, and red blood cells.^40^[40](#R40)40 Reduced expression of RFC1 has been associated with decreased MTX-PG accumulation in leukemia cells, and various SNPs have been associated with plasma MTX levels and gastrointestinal, hepatic, and bone marrow toxicities.^41–43^[41](#R41)41–[43](#R43)43 The AA genotype was associated with delayed MTX clearance, decreased event-free survival, and increased risk of minimal residual diseasein Chinese pediatric ALL patients.^7,44^[7](#R7)7,[44](#R44)44 Notably, this SNP was the only locus that was likely or very clearly associated with both outcomes in our study. This evidence suggests an active role of this locus in producing clinically-relevant aberrations in MTX disposition.

The G allele of rs7317112, an intronic variant of *ABCC4,*ABCC4, was the other SNP with a clear association with prolonged MTX clearance (OR 1.70 [95% CI 1.18–2.43]). It failed to show evidence of association with increased creatinine. *ABCC4*ABCC4 codes for an organic anion efflux transporter protein for MTX and has specifically been identified in renal cells to facilitate MTX elimination.^45,46^[45](#R45)45,[46](#R46)46 We previously identified this SNP as an independent risk factor for requiring CPDG_2_2 in this cohort of patients (OR 3.10 [95% CI 1.12–6.75]).^11^[11](#R11)11 An indication for CPDG_2_2 is severely elevated plasma MTX levels. Thus, a proposed mechanism may be an interplay between impaired MTX clearance causing elevated serum MTX and CPDG_2_2 administration. A previous study reported a possible association between the GG genotype and delayed clearance (significant on univariable analysis only), and another study reported that the AA genotype (the reference genotype in our study) was associated with risk for mucositis.^5,47^[5](#R5)5,[47](#R47)47 Further investigation in external cohorts is needed to validate these findings.

The G allele of rs2306283, which causes a missense mutation in *SLCO1B1,*SLCO1B1, demonstrated a decreased risk for prolonged clearance (OR 0.72 [95% CI 0.53–0.97]). *SLC01B1*SLC01B1 codes for a membrane-bound organic anion transporter that facilitates the intracellular influx of MTX into hepatocytes.^46^[46](#R46)46 One research group previously demonstrated that the G allele is associated with increased transport activity *in vitro*in vitro, which correlated to increased MTX clearance clinically.^6,48^[6](#R6)6,[48](#R48)48 They found evidence of effect after adjusting for rs4149056, a related SNP in *SLCO1B1*SLCO1B1 that had been shown to impair the function of the transporter. By including this SNP in the model, we also identified a slightly larger and more precise effect size (results in [Table 5](#T5)Table 5 compared to OR 0.75 [95% CI 0.53–1.02] without rs4149056), lending support to the previously described importance of these SNPs.

These genetic associations may assist clinical providers in identifying patients at risk for toxicity who may benefit from more therapy modification or more intensive supportive care. We showed preliminary results that models that integrated genetic and clinical information performed superiorly in predicting delayed clearance compared to models with only clinical or only genetic predictors ([table S7](#SD1)table S7). Smaller and less convincing improvements in predictive accuracy were obtained by integrating clinical and genetic information with the models for increased creatinine. As researchers attempt to improve the precision and individual risk-prediction models, this is promising evidence that combining clinical and genetic predictors may improve predictions.

Our study must be considered in light of certain limitations. As a retrospective cohort study, the conclusions may be subject to confounding despite DAG-informed covariate adjustment due to the ever-present risk of residual confounding. Certain desirable variables, such as the presence of baseline kidney injury, intravenous hydration, urine pH values, were unavailable for the analysis; however, we believe by including the dose number, the IMP, and dose reductions (which often accompany changing hydration practices) in the model, we were able to account for much of the dose-to-dose variability in the outcomes that changing intravenous hydration practices or other treatment modifications may have caused. The functional form of the relationship between the genotypes and the number of alleles is unknown in most cases. We chose to use an additive model, but other functional forms are possible. Only single-gene effects were evaluated, so we cannot comment on gene-gene, haplotype, and gene-environment effects. Finally, our study looked at only two complications of MTX treatment and did not evaluate other complications such as mucositis, myelosuppression, and neurotoxicity.

There are also several strengths of this study. We used Bayesian methods with skeptical priors derived from simulations to control the risk for false discovery. We used multilevel models to account for the varying effects of repeated observations in each subject. We used ordinal models that increased the power of the analysis compared to methods such as logistic regression that dichotomize the outcome. All effect estimates were adjusted according to DAGs that encoded plausible causal relationships between covariates. Finally, the large number of Hispanic patients in the sample allowed for a detailed analysis of an underrepresented ethnic group in MTX pharmacogenomic literature.

Our study demonstrates that creatinine increase and delayed MTX clearance remain prevalent complications of high-dose MTX therapy, a key component of curative ALL treatment. High-grade nephrotoxicity is rare, but excess hospital days due to delayed clearance comprise a large number of the total estimated hospital days associated with MTX administration. We identified both new and previously described clinical and demographic predictors of increased creatinine or delayed clearance. We identified new and previously described clinical and genetic predictors of toxicity. The predictors demonstrated improved predictive accuracy for delayed clearance, suggesting that these features may help to improve personalized toxicity risk prediction for creatinine increase and delayed clearance due to MTX. Additional research is needed to improve, calibrate, and validate the models for clinical use.

## Supplementary Material

## Acknowledgments

This work was funded by the T32 GM07526-41 Medical Genetics Research Fellowship Program, the St. Baldrick’s Foundation Consortium Grant (number: 522277) with support from the Micaela’s Army Foundation, the Adolescent and Childhood Cancer Epidemiology and Susceptibility Service (ACCESS) funded by the Cancer Prevention and Research Institute of Texas (RP160771), the Leukemia & Lymphoma Society Translational Research Program Grant (6314-11), and NIH grant 1P20CA262733-01 (Principal Investigators: Karen R. Rabin and Philip L. Lupo).

## Footnotes

## Associated Data

*This section collects any data citations, data availability statements, or supplementary materials included in this article.*This section collects any data citations, data availability statements, or supplementary materials included in this article.

### Supplementary Materials

### Supplementary Materials

## References

1. Taylor OA, Brown AL, Brackett J, Dreyer ZE, Moore IK, Mitby P, et al. Disparities in Neurotoxicity Risk and Outcomes among Pediatric Acute Lymphoblastic Leukemia Patients. Clin Cancer Res. 2018. Oct 15;24(20):5012–7.  [DOI](https://doi.org/10.1158/1078-0432.CCR-18-0939) | [PMC free article](/articles/PMC6191323/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/30206159/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Clin%20Cancer%20Res&title=Disparities%20in%20Neurotoxicity%20Risk%20and%20Outcomes%20among%20Pediatric%20Acute%20Lymphoblastic%20Leukemia%20Patients&author=OA%20Taylor&author=AL%20Brown&author=J%20Brackett&author=ZE%20Dreyer&author=IK%20Moore&volume=24&issue=20&publication_year=2018&pages=5012-7&pmid=30206159&doi=10.1158/1078-0432.CCR-18-0939&)

2. Siryk A, Grunwald H, Brahim A, Gentile F, Shenderov F, Ballestas C. Impact of Ethnicity on Toxicities Associated with Dose Escalating Methotrexate in Pediatric Patients with Acute Lymphoblastic Leukemia. Blood. 2019. Nov 13;134(Supplement_1):5074–5074.  [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Blood&title=Impact%20of%20Ethnicity%20on%20Toxicities%20Associated%20with%20Dose%20Escalating%20Methotrexate%20in%20Pediatric%20Patients%20with%20Acute%20Lymphoblastic%20Leukemia&author=A%20Siryk&author=H%20Grunwald&author=A%20Brahim&author=F%20Gentile&author=F%20Shenderov&volume=134&issue=1&publication_year=2019&pages=5074-5074&)

3. Widemann BC, Adamson PC. Understanding and Managing Methotrexate Nephrotoxicity. The Oncologist. 2006;11(6):694–703.  [DOI](https://doi.org/10.1634/theoncologist.11-6-694) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/16794248/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=The%20Oncologist&title=Understanding%20and%20Managing%20Methotrexate%20Nephrotoxicity&author=BC%20Widemann&author=PC%20Adamson&volume=11&issue=6&publication_year=2006&pages=694-703&pmid=16794248&doi=10.1634/theoncologist.11-6-694&)

4. Howard SC, McCormick J, Pui CH, Buddington RK, Harvey RD. Preventing and Managing Toxicities of High-Dose Methotrexate. The Oncologist. 2016. Dec;21(12):1471–82.  [DOI](https://doi.org/10.1634/theoncologist.2015-0164) | [PMC free article](/articles/PMC5153332/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/27496039/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=The%20Oncologist&title=Preventing%20and%20Managing%20Toxicities%20of%20High-Dose%20Methotrexate&author=SC%20Howard&author=J%20McCormick&author=CH%20Pui&author=RK%20Buddington&author=RD%20Harvey&volume=21&issue=12&publication_year=2016&pages=1471-82&pmid=27496039&doi=10.1634/theoncologist.2015-0164&)

5. Lopez-Lopez E, Ballesteros J, Piñan MA, Sanchez de Toledo J, Garcia de Andoin N, Garcia-Miguel P, et al. Polymorphisms in the methotrexate transport pathway: a new tool for MTX plasma level prediction in pediatric acute lymphoblastic leukemia. Pharmacogenet Genomics. 2013. Feb;23(2):53–61.  [DOI](https://doi.org/10.1097/FPC.0b013e32835c3b24) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/23222202/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pharmacogenet%20Genomics&title=Polymorphisms%20in%20the%20methotrexate%20transport%20pathway:%20a%20new%20tool%20for%20MTX%20plasma%20level%20prediction%20in%20pediatric%20acute%20lymphoblastic%20leukemia&author=E%20Lopez-Lopez&author=J%20Ballesteros&author=MA%20Pi%C3%B1an&author=J%20Sanchez%20de%20Toledo&author=N%20Garcia%20de%20Andoin&volume=23&issue=2&publication_year=2013&pages=53-61&pmid=23222202&doi=10.1097/FPC.0b013e32835c3b24&)

6. Ramsey LB, Panetta JC, Smith C, Yang W, Fan Y, Winick NJ, et al. Genome-wide study of methotrexate clearance replicates SLCO1B1. Blood. 2013. Feb 7;121(6):898–904.  [DOI](https://doi.org/10.1182/blood-2012-08-452839) | [PMC free article](/articles/PMC3567337/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/23233662/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Blood&title=Genome-wide%20study%20of%20methotrexate%20clearance%20replicates%20SLCO1B1&author=LB%20Ramsey&author=JC%20Panetta&author=C%20Smith&author=W%20Yang&author=Y%20Fan&volume=121&issue=6&publication_year=2013&pages=898-904&pmid=23233662&doi=10.1182/blood-2012-08-452839&)

7. Liu SG, Gao C, Zhang RD, Zhao XX, Cui L, Li WJ, et al. Polymorphisms in methotrexate transporters and their relationship to plasma methotrexate levels, toxicity of high-dose methotrexate, and outcome of pediatric acute lymphoblastic leukemia. Oncotarget. 2017. May 10;8(23):37761–72.  [DOI](https://doi.org/10.18632/oncotarget.17781) | [PMC free article](/articles/PMC5514947/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/28525903/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Oncotarget&title=Polymorphisms%20in%20methotrexate%20transporters%20and%20their%20relationship%20to%20plasma%20methotrexate%20levels,%20toxicity%20of%20high-dose%20methotrexate,%20and%20outcome%20of%20pediatric%20acute%20lymphoblastic%20leukemia&author=SG%20Liu&author=C%20Gao&author=RD%20Zhang&author=XX%20Zhao&author=L%20Cui&volume=8&issue=23&publication_year=2017&pages=37761-72&pmid=28525903&doi=10.18632/oncotarget.17781&)

8. Taylor ZL, Mizuno T, Punt N, Baskaran B, Sainz AN, Shuman W, et al. MTXPK.org: A clinical decision support tool evaluating high-dose methotrexate pharmacokinetics to inform post-infusion care and use of glucarpidase. medRxiv. 2020. May 6;2020.04.08.20056713.  [DOI](https://doi.org/10.1002/cpt.1957) | [PMC free article](/articles/PMC7484917/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/32558929/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=medRxiv&title=MTXPK.org:%20A%20clinical%20decision%20support%20tool%20evaluating%20high-dose%20methotrexate%20pharmacokinetics%20to%20inform%20post-infusion%20care%20and%20use%20of%20glucarpidase&author=ZL%20Taylor&author=T%20Mizuno&author=N%20Punt&author=B%20Baskaran&author=AN%20Sainz&publication_year=2020&pmid=32558929&doi=10.1002/cpt.1957&)

9. Medellin-Garibay SE, Hernández-Villa N, Correa-González LC, Morales-Barragán MN, Valero-Rivera KP, Reséndiz-Galván JE, et al. Population pharmacokinetics of methotrexate in Mexican pediatric patients with acute lymphoblastic leukemia. Cancer Chemother Pharmacol. 2020. Jan 1;85(1):21–31.  [DOI](https://doi.org/10.1007/s00280-019-03977-1) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/31673826/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer%20Chemother%20Pharmacol&title=Population%20pharmacokinetics%20of%20methotrexate%20in%20Mexican%20pediatric%20patients%20with%20acute%20lymphoblastic%20leukemia&author=SE%20Medellin-Garibay&author=N%20Hern%C3%A1ndez-Villa&author=LC%20Correa-Gonz%C3%A1lez&author=MN%20Morales-Barrag%C3%A1n&author=KP%20Valero-Rivera&volume=85&issue=1&publication_year=2020&pages=21-31&pmid=31673826&doi=10.1007/s00280-019-03977-1&)

10. Mullikin D, Ranch D, Khalfe Y, Lucari B, Zobeck MC, Assanasen C, et al. Hispanic ethnicity is associated with prolonged clearance of high dose methotrexate and severe nephrotoxicity in children and adolescents with acute lymphoblastic leukemia. Leuk Lymphoma. 2020. Jul 1;0(0):1–4.  [DOI](https://doi.org/10.1080/10428194.2020.1783445) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/32605405/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Leuk%20Lymphoma&title=Hispanic%20ethnicity%20is%20associated%20with%20prolonged%20clearance%20of%20high%20dose%20methotrexate%20and%20severe%20nephrotoxicity%20in%20children%20and%20adolescents%20with%20acute%20lymphoblastic%20leukemia&author=D%20Mullikin&author=D%20Ranch&author=Y%20Khalfe&author=B%20Lucari&author=MC%20Zobeck&volume=0&issue=0&publication_year=2020&pages=1-4&pmid=32605405&doi=10.1080/10428194.2020.1783445&)

11. Zobeck MC, Bernhardt MB, Kamdar KY, Rabin KR, Lupo PJ, Scheurer ME. Novel risk factors for glucarpidase use in pediatric acute lymphoblastic leukemia: Hispanic ethnicity, age, and the ABCC4 gene. Pediatr Blood Cancer. 2021;68(8):e29036.  [DOI](https://doi.org/10.1002/pbc.29036) | [PMC free article](/articles/PMC8238882/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/33788417/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pediatr%20Blood%20Cancer&title=Novel%20risk%20factors%20for%20glucarpidase%20use%20in%20pediatric%20acute%20lymphoblastic%20leukemia:%20Hispanic%20ethnicity,%20age,%20and%20the%20ABCC4%20gene&author=MC%20Zobeck&author=MB%20Bernhardt&author=KY%20Kamdar&author=KR%20Rabin&author=PJ%20Lupo&volume=68&issue=8&publication_year=2021&pmid=33788417&doi=10.1002/pbc.29036&)

12. Foster JH, Thompson PA, Bernhardt MB, Margolin JF, Hilsenbeck SG, Jo E, et al. A prospective study of a simple algorithm to individually dose high-dose methotrexate for children with leukemia at risk for methotrexate toxicities. Cancer Chemother Pharmacol. 2019. Feb 1;83(2):349–60.  [DOI](https://doi.org/10.1007/s00280-018-3733-2) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/30488179/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer%20Chemother%20Pharmacol&title=A%20prospective%20study%20of%20a%20simple%20algorithm%20to%20individually%20dose%20high-dose%20methotrexate%20for%20children%20with%20leukemia%20at%20risk%20for%20methotrexate%20toxicities&author=JH%20Foster&author=PA%20Thompson&author=MB%20Bernhardt&author=JF%20Margolin&author=SG%20Hilsenbeck&volume=83&issue=2&publication_year=2019&pages=349-60&pmid=30488179&doi=10.1007/s00280-018-3733-2&)

13. Mao X, Bigham AW, Mei R, Gutierrez G, Weiss KM, Brutsaert TD, et al. A Genomewide Admixture Mapping Panel for Hispanic/Latino Populations. Am J Hum Genet. 2007. Jun 1;80(6):1171–8.  [DOI](https://doi.org/10.1086/518564) | [PMC free article](/articles/PMC1867104/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/17503334/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Am%20J%20Hum%20Genet&title=A%20Genomewide%20Admixture%20Mapping%20Panel%20for%20Hispanic/Latino%20Populations&author=X%20Mao&author=AW%20Bigham&author=R%20Mei&author=G%20Gutierrez&author=KM%20Weiss&volume=80&issue=6&publication_year=2007&pages=1171-8&pmid=17503334&doi=10.1086/518564&)

14. Whirl‐Carrillo M, McDonagh EM, Hebert JM, Gong L, Sangkuhl K, Thorn CF, et al. Pharmacogenomics knowledge for personalized medicine. Clin Pharmacol Ther. 2012;92(4):414–7.  [DOI](https://doi.org/10.1038/clpt.2012.96) | [PMC free article](/articles/PMC3660037/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/22992668/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Clin%20Pharmacol%20Ther&title=Pharmacogenomics%20knowledge%20for%20personalized%20medicine&author=M%20Whirl%E2%80%90Carrillo&author=EM%20McDonagh&author=JM%20Hebert&author=L%20Gong&author=K%20Sangkuhl&volume=92&issue=4&publication_year=2012&pages=414-7&pmid=22992668&doi=10.1038/clpt.2012.96&)

15. Chittoor G, Haack K, Mehta NR, Laston S, Cole SA, Comuzzie AG, et al. Genetic variation underlying renal uric acid excretion in Hispanic children: the Viva La Familia Study. BMC Med Genet. 2017. Jan 17;18(1):6.  [DOI](https://doi.org/10.1186/s12881-016-0366-3) | [PMC free article](/articles/PMC5240212/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/28095793/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=BMC%20Med%20Genet&title=Genetic%20variation%20underlying%20renal%20uric%20acid%20excretion%20in%20Hispanic%20children:%20the%20Viva%20La%20Familia%20Study&author=G%20Chittoor&author=K%20Haack&author=NR%20Mehta&author=S%20Laston&author=SA%20Cole&volume=18&issue=1&publication_year=2017&pages=6&pmid=28095793&doi=10.1186/s12881-016-0366-3&)

16. Voruganti VS, Kent JWJ, Debnath S, Cole SA, Haack K, Goring HH, et al. Genome-wide association analysis confirms and extends the association of SLC2A9 with serum uric acid levels to Mexican Americans. Front Genet [Internet]. 2013. [cited 2020 Apr 16];4. Available from: 10.3389/fgene.2013.00279/full  [DOI](https://doi.org/10.3389/fgene.2013.00279/full) | [PMC free article](/articles/PMC3863993/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/24379826/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Front%20Genet%20%5BInternet%5D&title=Genome-wide%20association%20analysis%20confirms%20and%20extends%20the%20association%20of%20SLC2A9%20with%20serum%20uric%20acid%20levels%20to%20Mexican%20Americans&author=VS%20Voruganti&author=JWJ%20Kent&author=S%20Debnath&author=SA%20Cole&author=K%20Haack&publication_year=2013&pmid=24379826&doi=10.3389/fgene.2013.00279/full&)

17. Common Terminology Criteria for Adverse Events (CTCAE). 2017;155.  [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Common%20Terminology%20Criteria%20for%20Adverse%20Events%20(CTCAE)&volume=155&publication_year=2017&)

18. Vehtari A, Gelman A, Gabry J. Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC. Stat Comput. 2017. Sep 1;27(5):1413–32.  [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Stat%20Comput&title=Practical%20Bayesian%20model%20evaluation%20using%20leave-one-out%20cross-validation%20and%20WAIC&author=A%20Vehtari&author=A%20Gelman&author=J%20Gabry&volume=27&issue=5&publication_year=2017&pages=1413-32&)

19. R Core Team. R: A Language and Environment for Statistical Computing [Internet]. Vienna, Austria: R Foundation for Statistical Computing; 2018. Available from: https://www.R-project.org/  [https://www.R-project.org/](https://www.R-project.org/) | [Google Scholar](https://scholar.google.com/scholar_lookup?title=R:%20A%20Language%20and%20Environment%20for%20Statistical%20Computing%20%5BInternet%5D&publication_year=2018&)

20. Bürkner PC. brms: An R Package for Bayesian Multilevel Models Using Stan. J Stat Softw. 2017;80(1):1–28.  [Google Scholar](https://scholar.google.com/scholar_lookup?journal=J%20Stat%20Softw&title=brms:%20An%20R%20Package%20for%20Bayesian%20Multilevel%20Models%20Using%20Stan&author=PC%20B%C3%BCrkner&volume=80&issue=1&publication_year=2017&pages=1-28&)

21. Wickham H ggplot2: Elegant Graphics for Data Analysis [Internet]. Springer-Verlag; New York; 2016. Available from: https://ggplot2.tidyverse.org  [https://ggplot2.tidyverse.org](https://ggplot2.tidyverse.org) | [Google Scholar](https://scholar.google.com/scholar_lookup?title=ggplot2:%20Elegant%20Graphics%20for%20Data%20Analysis%20%5BInternet%5D&author=H%20Wickham&publication_year=2016&)

22. Kay M tidybayes: Tidy Data and Geoms for Bayesian Models [Internet]. 2020. Available from: http://mjskay.github.io/tidybayes/  [http://mjskay.github.io/tidybayes/](http://mjskay.github.io/tidybayes/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=tidybayes:%20Tidy%20Data%20and%20Geoms%20for%20Bayesian%20Models%20%5BInternet%5D&author=M%20Kay&publication_year=2020&)

23. Gabry J, Mahr T. bayesplot: Plotting for Bayesian Models [Internet]. 2020. Available from: https://mc-stan.org/bayesplot  [https://mc-stan.org/bayesplot](https://mc-stan.org/bayesplot) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=bayesplot:%20Plotting%20for%20Bayesian%20Models%20%5BInternet%5D&author=J%20Gabry&author=T%20Mahr&publication_year=2020&)

24. Christensen AM, Pauley JL, Molinelli AR, Panetta JC, Ward DA, Stewart CF, et al. Resumption of high-dose methotrexate after acute kidney injury and glucarpidase use in pediatric oncology patients. Cancer. 2012;118(17):4321–30.  [DOI](https://doi.org/10.1002/cncr.27378) | [PMC free article](/articles/PMC3713608/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/22252903/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer&title=Resumption%20of%20high-dose%20methotrexate%20after%20acute%20kidney%20injury%20and%20glucarpidase%20use%20in%20pediatric%20oncology%20patients&author=AM%20Christensen&author=JL%20Pauley&author=AR%20Molinelli&author=JC%20Panetta&author=DA%20Ward&volume=118&issue=17&publication_year=2012&pages=4321-30&pmid=22252903&doi=10.1002/cncr.27378&)

25. Svahn T, Mellgren K, Harila-Saari A, Åsberg A, Kanerva J, Jónsson Ó, et al. Delayed elimination of high-dose methotrexate and use of carboxypeptidase G2 in pediatric patients during treatment for acute lymphoblastic leukemia: Svahn et al. Pediatr Blood Cancer. 2017. Jul;64(7):e26395.  [DOI](https://doi.org/10.1002/pbc.26395) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/27966809/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pediatr%20Blood%20Cancer&title=Delayed%20elimination%20of%20high-dose%20methotrexate%20and%20use%20of%20carboxypeptidase%20G2%20in%20pediatric%20patients%20during%20treatment%20for%20acute%20lymphoblastic%20leukemia:%20Svahn%20et%20al&author=T%20Svahn&author=K%20Mellgren&author=A%20Harila-Saari&author=A%20%C3%85sberg&author=J%20Kanerva&volume=64&issue=7&publication_year=2017&pmid=27966809&doi=10.1002/pbc.26395&)

26. Roser K, Erdmann F, Michel G, Winther JF, Mader L. The impact of childhood cancer on parents’ socio-economic situation—A systematic review. Psychooncology. 2019;28(6):1207–26.  [DOI](https://doi.org/10.1002/pon.5088) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/30970149/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Psychooncology&title=The%20impact%20of%20childhood%20cancer%20on%20parents%E2%80%99%20socio-economic%20situation%E2%80%94A%20systematic%20review&author=K%20Roser&author=F%20Erdmann&author=G%20Michel&author=JF%20Winther&author=L%20Mader&volume=28&issue=6&publication_year=2019&pages=1207-26&pmid=30970149&doi=10.1002/pon.5088&)

27. Kelada L, Wakefield CE, Vetsch J, Schofield D, Sansom-Daly UM, Hetherington K, et al. Financial toxicity of childhood cancer and changes to parents’ employment after treatment completion. Pediatr Blood Cancer. 2020;67(7):e28345.  [DOI](https://doi.org/10.1002/pbc.28345) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/32333636/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pediatr%20Blood%20Cancer&title=Financial%20toxicity%20of%20childhood%20cancer%20and%20changes%20to%20parents%E2%80%99%20employment%20after%20treatment%20completion&author=L%20Kelada&author=CE%20Wakefield&author=J%20Vetsch&author=D%20Schofield&author=UM%20Sansom-Daly&volume=67&issue=7&publication_year=2020&pmid=32333636&doi=10.1002/pbc.28345&)

28. Hijiya N, Panetta JC, Zhou Y, Kyzer EP, Howard SC, Jeha S, et al. Body mass index does not influence pharmacokinetics or outcome of treatment in children with acute lymphoblastic leukemia. Blood. 2006. Dec 15;108(13):3997–4002.  [DOI](https://doi.org/10.1182/blood-2006-05-024414) | [PMC free article](/articles/PMC1895448/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/16917005/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Blood&title=Body%20mass%20index%20does%20not%20influence%20pharmacokinetics%20or%20outcome%20of%20treatment%20in%20children%20with%20acute%20lymphoblastic%20leukemia&author=N%20Hijiya&author=JC%20Panetta&author=Y%20Zhou&author=EP%20Kyzer&author=SC%20Howard&volume=108&issue=13&publication_year=2006&pages=3997-4002&pmid=16917005&doi=10.1182/blood-2006-05-024414&)

29. Cheng DH, Lu H, Liu TT, Zou XQ, Pang HM. Identification of Risk Factors in High-Dose Methotrexate-Induced Acute Kidney Injury in Childhood Acute Lymphoblastic Leukemia. Chemotherapy. 2018;63(2):100–6.  [DOI](https://doi.org/10.1159/000486823) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/29672292/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Chemotherapy&title=Identification%20of%20Risk%20Factors%20in%20High-Dose%20Methotrexate-Induced%20Acute%20Kidney%20Injury%20in%20Childhood%20Acute%20Lymphoblastic%20Leukemia&author=DH%20Cheng&author=H%20Lu&author=TT%20Liu&author=XQ%20Zou&author=HM%20Pang&volume=63&issue=2&publication_year=2018&pages=100-6&pmid=29672292&doi=10.1159/000486823&)

30. Grossie VB Jr, Ho DHW, Loo TL. Effect of Malnutrition on Methotrexate Toxicity and Tissue Levels of Dihydrofolate Reductase in the Rat 1, 2, 3. Cancer Treat Rep. 1982;66(1):85–9.  [PubMed](https://pubmed.ncbi.nlm.nih.gov/7053272/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer%20Treat%20Rep&title=Effect%20of%20Malnutrition%20on%20Methotrexate%20Toxicity%20and%20Tissue%20Levels%20of%20Dihydrofolate%20Reductase%20in%20the%20Rat%201,%202,%203&author=VB%20Grossie&author=DHW%20Ho&author=TL%20Loo&volume=66&issue=1&publication_year=1982&pages=85-9&pmid=7053272&)

31. Mihranian MH, Wang YM, Daly JM. Effects of nutritional depletion and repletion on plasma methotrexate pharmacokinetics. Cancer. 1984;54(10):2268–71.  [DOI](https://doi.org/10.1002/1097-0142(19841115)54:10<2268::aid-cncr2820541034>3.0.co;2-2) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/6488146/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer&title=Effects%20of%20nutritional%20depletion%20and%20repletion%20on%20plasma%20methotrexate%20pharmacokinetics&author=MH%20Mihranian&author=YM%20Wang&author=JM%20Daly&volume=54&issue=10&publication_year=1984&pages=2268-71&pmid=6488146&doi=10.1002/1097-0142(19841115)54:10<2268::aid-cncr2820541034>3.0.co;2-2&)

32. Murry DJ, Riva L, Poplack DG. Impact of nutrition on pharmacokinetics of anti-neoplastic agents. Int J Cancer. 1998;78(S11):48–51.  [DOI](https://doi.org/10.1002/(sici)1097-0215(1998)78:11+<48::aid-ijc14>3.0.co;2-3) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/9876478/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Int%20J%20Cancer&title=Impact%20of%20nutrition%20on%20pharmacokinetics%20of%20anti-neoplastic%20agents&author=DJ%20Murry&author=L%20Riva&author=DG%20Poplack&volume=78&issue=S11&publication_year=1998&pages=48-51&pmid=9876478&doi=10.1002/(sici)1097-0215(1998)78:11+<48::aid-ijc14>3.0.co;2-3&)

33. Wiczer T, Dotson E, Tuten A, Phillips G, Maddocks K. Evaluation of incidence and risk factors for high-dose methotrexate-induced nephrotoxicity. J Oncol Pharm Pract. 2016. Jun 1;22(3):430–6.  [DOI](https://doi.org/10.1177/1078155215594417) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/26152702/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=J%20Oncol%20Pharm%20Pract&title=Evaluation%20of%20incidence%20and%20risk%20factors%20for%20high-dose%20methotrexate-induced%20nephrotoxicity&author=T%20Wiczer&author=E%20Dotson&author=A%20Tuten&author=G%20Phillips&author=K%20Maddocks&volume=22&issue=3&publication_year=2016&pages=430-6&pmid=26152702&doi=10.1177/1078155215594417&)

34. Amitai I, Rozovski U, El‐Saleh R, Shimony S, Shepshelovich D, Rozen‐Zvi B, et al. Risk factors for high-dose methotrexate associated acute kidney injury in patients with hematological malignancies. Hematol Oncol [Internet]. [cited 2020 Sep 25];n/a(n/a). Available from: 10.1002/hon.2759  [DOI](https://doi.org/10.1002/hon.2759) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/32506594/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Hematol%20Oncol%20%5BInternet%5D&title=Risk%20factors%20for%20high-dose%20methotrexate%20associated%20acute%20kidney%20injury%20in%20patients%20with%20hematological%20malignancies&author=I%20Amitai&author=U%20Rozovski&author=R%20El%E2%80%90Saleh&author=S%20Shimony&author=D%20Shepshelovich&pmid=32506594&doi=10.1002/hon.2759&)

35. Reiss SN, Buie LW, Adel N, Goldman DA, Devlin SM, Douer D. Hypoalbuminemia is significantly associated with increased clearance time of high dose methotrexate in patients being treated for lymphoma or leukemia. Ann Hematol. 2016. Dec;95(12):2009–15.  [DOI](https://doi.org/10.1007/s00277-016-2795-7) | [PMC free article](/articles/PMC5572815/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/27542957/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Ann%20Hematol&title=Hypoalbuminemia%20is%20significantly%20associated%20with%20increased%20clearance%20time%20of%20high%20dose%20methotrexate%20in%20patients%20being%20treated%20for%20lymphoma%20or%20leukemia&author=SN%20Reiss&author=LW%20Buie&author=N%20Adel&author=DA%20Goldman&author=SM%20Devlin&volume=95&issue=12&publication_year=2016&pages=2009-15&pmid=27542957&doi=10.1007/s00277-016-2795-7&)

36. Csordas K, Hegyi M, Eipel OT, Muller J, Erdelyi DJ, Kovacs GT. Comparison of pharmacokinetics and toxicity after high-dose methotrexate treatments in children with acute lymphoblastic leukemia: Anticancer Drugs. 2013. Feb;24(2):189–97.  [DOI](https://doi.org/10.1097/CAD.0b013e32835b8662) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/23187460/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Comparison%20of%20pharmacokinetics%20and%20toxicity%20after%20high-dose%20methotrexate%20treatments%20in%20children%20with%20acute%20lymphoblastic%20leukemia:%20Anticancer%20Drugs&author=K%20Csordas&author=M%20Hegyi&author=OT%20Eipel&author=J%20Muller&author=DJ%20Erdelyi&volume=24&issue=2&publication_year=2013&pages=189-97&pmid=23187460&doi=10.1097/CAD.0b013e32835b8662&)

37. Aumente D, Santos Buelga D, Lukas JC, Gomez P, Torres A, Garc??a MJ. Population Pharmacokinetics of High-Dose Methotrexate in Children with Acute Lymphoblastic Leukaemia: Clin Pharmacokinet. 2006;45(12):1227–38.  [DOI](https://doi.org/10.2165/00003088-200645120-00007) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/17112298/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Population%20Pharmacokinetics%20of%20High-Dose%20Methotrexate%20in%20Children%20with%20Acute%20Lymphoblastic%20Leukaemia:%20Clin%20Pharmacokinet&title=Garc??a%20MJ&author=D%20Aumente&author=D%20Santos%20Buelga&author=JC%20Lukas&author=P%20Gomez&author=A%20Torres&volume=45&issue=12&publication_year=2006&pages=1227-38&pmid=17112298&doi=10.2165/00003088-200645120-00007&)

38. Salzer WL, Winick NJ, Wacker P, Lu X, Devidas M, Shuster JJ, et al. Plasma Methotrexate, Red Blood Cell Methotrexate, and Red Blood Cell Folate Values and Outcome in Children With Precursor B-acute Lymphoblastic Leukemia: A Report From the Childrenʼs Oncology Group. J Pediatr Hematol Oncol. 2012. Jan;34(1):e1–7.  [DOI](https://doi.org/10.1097/MPH.0b013e31820ee239) | [PMC free article](/articles/PMC3771527/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/21364468/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=J%20Pediatr%20Hematol%20Oncol&title=Plasma%20Methotrexate,%20Red%20Blood%20Cell%20Methotrexate,%20and%20Red%20Blood%20Cell%20Folate%20Values%20and%20Outcome%20in%20Children%20With%20Precursor%20B-acute%20Lymphoblastic%20Leukemia:%20A%20Report%20From%20the%20Children%CA%BCs%20Oncology%20Group&author=WL%20Salzer&author=NJ%20Winick&author=P%20Wacker&author=X%20Lu&author=M%20Devidas&volume=34&issue=1&publication_year=2012&pages=e1-7&pmid=21364468&doi=10.1097/MPH.0b013e31820ee239&)

39. Kadan-Lottick NS, Ness KK, Bhatia S, Gurney JG. Survival Variability by Race and Ethnicity in Childhood Acute Lymphoblastic Leukemia. JAMA. 2003. Oct 15;290(15):2008–14.  [DOI](https://doi.org/10.1001/jama.290.15.2008) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/14559954/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=JAMA&title=Survival%20Variability%20by%20Race%20and%20Ethnicity%20in%20Childhood%20Acute%20Lymphoblastic%20Leukemia&author=NS%20Kadan-Lottick&author=KK%20Ness&author=S%20Bhatia&author=JG%20Gurney&volume=290&issue=15&publication_year=2003&pages=2008-14&pmid=14559954&doi=10.1001/jama.290.15.2008&)

40. Yee SW, Gong L, Badagnani I, Giacomini KM, Klein TE, Altman RB. SLC19A1 Pharmacogenomics Summary. Pharmacogenet Genomics. 2010. Nov;20(11):708–15.  [DOI](https://doi.org/10.1097/FPC.0b013e32833eca92) | [PMC free article](/articles/PMC2956130/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/20811316/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pharmacogenet%20Genomics&title=SLC19A1%20Pharmacogenomics%20Summary&author=SW%20Yee&author=L%20Gong&author=I%20Badagnani&author=KM%20Giacomini&author=TE%20Klein&volume=20&issue=11&publication_year=2010&pages=708-15&pmid=20811316&doi=10.1097/FPC.0b013e32833eca92&)

41. Kager L, Cheok M, Yang W, Zaza G, Cheng Q, Panetta JC, et al. Folate pathway gene expression differs in subtypes of acute lymphoblastic leukemia and influences methotrexate pharmacodynamics. J Clin Invest. 2005. Jan 3;115(1):110–7.  [DOI](https://doi.org/10.1172/JCI22477) | [PMC free article](/articles/PMC539195/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/15630450/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=J%20Clin%20Invest&title=Folate%20pathway%20gene%20expression%20differs%20in%20subtypes%20of%20acute%20lymphoblastic%20leukemia%20and%20influences%20methotrexate%20pharmacodynamics&author=L%20Kager&author=M%20Cheok&author=W%20Yang&author=G%20Zaza&author=Q%20Cheng&volume=115&issue=1&publication_year=2005&pages=110-7&pmid=15630450&doi=10.1172/JCI22477&)

42. Kishi S, Cheng C, French D, Pei D, Das S, Cook EH, et al. Ancestry and pharmacogenetics of antileukemic drug toxicity. Blood. 2007. Jan 30;109(10):4151–7.  [DOI](https://doi.org/10.1182/blood-2006-10-054528) | [PMC free article](/articles/PMC1885506/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/17264302/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Blood&title=Ancestry%20and%20pharmacogenetics%20of%20antileukemic%20drug%20toxicity&author=S%20Kishi&author=C%20Cheng&author=D%20French&author=D%20Pei&author=S%20Das&volume=109&issue=10&publication_year=2007&pages=4151-7&pmid=17264302&doi=10.1182/blood-2006-10-054528&)

43. Gregers J, Christensen IJ, Dalhoff K, Lausen B, Schroeder H, Rosthoej S, et al. The association of reduced folate carrier 80G>A polymorphism to outcome in childhood acute lymphoblastic leukemia interacts with chromosome 21 copy number. Blood. 2010. Jun 10;115(23):4671–7.  [DOI](https://doi.org/10.1182/blood-2010-01-256958) | [PMC free article](/articles/PMC2890175/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/20335220/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Blood&title=The%20association%20of%20reduced%20folate%20carrier%2080G>A%20polymorphism%20to%20outcome%20in%20childhood%20acute%20lymphoblastic%20leukemia%20interacts%20with%20chromosome%2021%20copy%20number&author=J%20Gregers&author=IJ%20Christensen&author=K%20Dalhoff&author=B%20Lausen&author=H%20Schroeder&volume=115&issue=23&publication_year=2010&pages=4671-7&pmid=20335220&doi=10.1182/blood-2010-01-256958&)

44. Zhan M, Chen Z, Ding C, Qu Q, Wang G, Liu S, et al. Risk prediction for delayed clearance of high-dose methotrexate in pediatric hematological malignancies by machine learning. Int J Hematol. 2021. Oct;114(4):483–93.  [DOI](https://doi.org/10.1007/s12185-021-03184-w) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/34170480/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Int%20J%20Hematol&title=Risk%20prediction%20for%20delayed%20clearance%20of%20high-dose%20methotrexate%20in%20pediatric%20hematological%20malignancies%20by%20machine%20learning&author=M%20Zhan&author=Z%20Chen&author=C%20Ding&author=Q%20Qu&author=G%20Wang&volume=114&issue=4&publication_year=2021&pages=483-93&pmid=34170480&doi=10.1007/s12185-021-03184-w&)

45. Chen ZS, Lee K, Walther S, Raftogianis RB, Kuwano M, Zeng H, et al. Analysis of Methotrexate and Folate Transport by Multidrug Resistance Protein 4 (ABCC4): MRP4 Is a Component of the Methotrexate Efflux System. Cancer Res. 2002. Jun 1;62(11):3144–50.  [PubMed](https://pubmed.ncbi.nlm.nih.gov/12036927/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Cancer%20Res&title=Analysis%20of%20Methotrexate%20and%20Folate%20Transport%20by%20Multidrug%20Resistance%20Protein%204%20(ABCC4):%20MRP4%20Is%20a%20Component%20of%20the%20Methotrexate%20Efflux%20System&author=ZS%20Chen&author=K%20Lee&author=S%20Walther&author=RB%20Raftogianis&author=M%20Kuwano&volume=62&issue=11&publication_year=2002&pages=3144-50&pmid=12036927&)

46. Mikkelsen TS, Thorn CF, Yang JJ, Ulrich CM, French D, Zaza G, et al. PharmGKB summary: methotrexate pathway. Pharmacogenet Genomics. 2011. Oct;21(10):679–86.  [DOI](https://doi.org/10.1097/FPC.0b013e328343dd93) | [PMC free article](/articles/PMC3139712/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/21317831/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pharmacogenet%20Genomics&title=PharmGKB%20summary:%20methotrexate%20pathway&author=TS%20Mikkelsen&author=CF%20Thorn&author=JJ%20Yang&author=CM%20Ulrich&author=D%20French&volume=21&issue=10&publication_year=2011&pages=679-86&pmid=21317831&doi=10.1097/FPC.0b013e328343dd93&)

47. den Hoed MAH, Lopez-Lopez E, te Winkel ML, Tissing W, de Rooij JDE, Gutierrez-Camino A, et al. Genetic and metabolic determinants of methotrexate-induced mucositis in pediatric acute lymphoblastic leukemia. Pharmacogenomics J. 2015. Jun;15(3):248–54.  [DOI](https://doi.org/10.1038/tpj.2014.63) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/25348617/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Pharmacogenomics%20J&title=Genetic%20and%20metabolic%20determinants%20of%20methotrexate-induced%20mucositis%20in%20pediatric%20acute%20lymphoblastic%20leukemia&author=MAH%20den%20Hoed&author=E%20Lopez-Lopez&author=ML%20te%20Winkel&author=W%20Tissing&author=JDE%20de%20Rooij&volume=15&issue=3&publication_year=2015&pages=248-54&pmid=25348617&doi=10.1038/tpj.2014.63&)

48. Ramsey LB, Bruun GH, Yang W, Treviño LR, Vattathil S, Scheet P, et al. Rare versus common variants in pharmacogenetics: SLCO1B1 variation and methotrexate disposition. Genome Res. 2012. Jan;22(1):1–8.  [DOI](https://doi.org/10.1101/gr.129668.111) | [PMC free article](/articles/PMC3246196/) | [PubMed](https://pubmed.ncbi.nlm.nih.gov/22147369/) | [Google Scholar](https://scholar.google.com/scholar_lookup?journal=Genome%20Res&title=Rare%20versus%20common%20variants%20in%20pharmacogenetics:%20SLCO1B1%20variation%20and%20methotrexate%20disposition&author=LB%20Ramsey&author=GH%20Bruun&author=W%20Yang&author=LR%20Trevi%C3%B1o&author=S%20Vattathil&volume=22&issue=1&publication_year=2012&pages=1-8&pmid=22147369&doi=10.1101/gr.129668.111&)