Effects of genetic polymorphisms on methotrexate levels and toxicity in Chinese patients with acute lymphoblastic leukemia

Methotrexate (MTX) has an antitumor effect when used for the treatment of acute lymphoblastic leukemia (ALL). This study aims at evaluating the associations between 14 polymorphisms of six genes involved in MTX metabolism with serum MTX concentration and toxicity accompanying high-dose MTX. Polymorphisms in 183 Chinese patients with ALL were analyzed using TaqMan single nucleotide polymorphism genotyping assay. The serum MTX concentration was determined using homogeneous enzyme immunoassay. MTX-related toxicities were also evaluated. Renal toxicity was significantly associated with higher serum MTX concentrations at 24, 48, and 72 hours, and MTX elimination delay (P = 0.001, P < 0.001, P < 0.001, and P < 0.001, respectively), whereas SLCO1B1 rs4149056 was associated with serum MTX concentrations at 48 and 72 hours, and MTX elimination delay in candidate polymorphisms (P = 0.014, P = 0.019, and P = 0.007, respectively). SLC19A1 rs2838958 and rs3788200 were associated with serum MTX concentrations at 24 hours (P = 0.016, P = 0.043, respectively). MTRR rs1801394 was associated with serum MTX concentrations at 72 hours (P = 0.045). Neutropenia was related to SLC19A1 rs4149056 (odds ratio [OR]: 3.172, 95% confidence interval [CI]: 1.310–7.681, P = 0.011). Hepatotoxicity was associated with ABCC2 rs2273697 (OR: 3.494, 95% CI: 1.236–9.873, P = 0.018) and MTRR rs1801394 (OR: 0.231, 95% CI: 0.084–0.632, P = 0.004). Polymorphisms of SLCO1B1, SLC19A1, ABCC2, and MTRR genes help predict higher risk of increased MTX levels or MTX-related toxicities in adult ALL patients.


INTRODUCTION
Acute lymphoblastic leukemia (ALL) is a hematological malignancy characterized by blast lymphocyte proliferation in the bone marrow, peripheral blood, and extramedullary sites. 1 Compared with the high-cure rate of childhood ALL, the outcome of adult ALL is poor.
Methotrexate (MTX) is an essential antitumor drug for ALL and a folate antagonist that can competitively inhibit dihydrofolate reductase, which converts dihydrofolate into tetrahydrofolate. The lack of tetrahydrofolate impairs purine and thymidine synthesis, which in turn impairs DNA replication and eventually results in cell death. 2 Additionally, MTX inhibits other enzymes involved in folate metabolism, including methylenetetrahydrofolate reductase (MTHFR) and methionine synthase reductase (MTRR). MTX enters the cells by binding to solute carrier organic anion transporter family member 1B1 (SLCO1B1) or solute carrier 19A1 (SLC19A1), while being exported from the cells through various ATP-binding cassette (ABC) transporters (ABCC1-5 and ABCG1-2) (Fig. 1). [3][4][5] The MTX distribution, bioavailability, and clearance in the body show marked differences, independent of disease-type, and administration routes. 6 However, an interpatient variability in response to MTX is associated with variations in genes encoding enzymes and membrane transporter proteins involved in the MTX metabolism. Patients with ALL receiving high-dose MTX (HD-MTX) frequently suffer from mucositis, gastrointestinal toxicity, hematologic toxicity, hepatoxicity, and nephrotoxicity. 7,8 MTX-related toxicities limit MTX dosage and result in treatment delay. Thus, the potential tolerance of patients with ALL to MTX should be determined at the pretreatment stage.
Previous studies have investigated the association between single nucleotide polymorphisms (SNPs) of genes in the MTX/ folate metabolic pathway and MTX levels and MTX-related toxicity; however, the results are contradictory (see Supplemental Table 1, http://links.lww.com/BS/A51). The conflicting outcomes may be caused by small or varying populations, different treatment regimens and nonuniform toxicity standards. Moreover, most previous studies were conducted in children with ALL. Therefore, the current study aimed at exploring the association between 14 SNPs of six genes involved in MTX metabolism with serum MTX concentrations and HD-MTX-related toxicities in adolescent and adult Chinese patients with ALL.

Patient and clinical data
A total of 183 patients with ALL who received HD-MTX treatment and were ≥14 years old were included in this retrospective study. All patients were newly diagnosed and treated in the Leukemia Department of the Institute of Hematology & Blood Diseases Hospital, Chinese Academy of Medical Sciences & Peking Union Medical College (Tianjin, China) from September 2015 to December 2019. The study was approved by the Ethics Committee of the hospital (No. HG2021038-EC-1). Written informed consent was obtained from all patients or their guardians.
According to the ChiCTR-TNC-09000397 protocol, the MTX dose was 2 g/m 2 in patients with Philadelphia chromosome-positive ALL (Ph + ALL) and 3 g/m 2 in those with Philadelphia chromosome-negative ALL (Ph − ALL). About onethird of the total MTX dose was infused within 1 hour, the remainder administered as a 23-hour continuous intravenous infusion. Leucovorin rescue therapy was started at a dose of 7.5% of the MTX dose and 36 hours after the start of MTX administration. The total leucovorin delivered was divided into 6 to 8 doses at 6-hour intervals (the first dose was 50 mg) until the serum MTX concentration was <0.3 μmol/L. The serum MTX concentration was measured daily at 24, 48, and 72 hours after the start of MTX administration and was determined by homogeneous enzyme immunoassay (Siemens Viva-E Drug Test System, Erlangen, Bavaria, Germany) using the MTX detection kit (Siemens, Newark, DE, USA) in the hospital's Clinical Pharmacy Laboratory, with a safe value of 0.3 μmol/L. MTX elimination delay was defined as MTX concentration of >1.0 μmol/L at 48 hours and >0.3 μmol/L at 72 hours.
MTX-related toxicities, including hematologic, hepatic and renal, were evaluated according to the Common Terminology Criteria for adverse events version 5.0 (CTCAE v.5.0). 9 The highest toxicity grade was observed and registered in each patient within 2 weeks of starting the MTX medication. Toxicities were assessed based on clinical parameters, such as white blood cell, absolute neutrophil and platelet counts, and hemoglobin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, and creatinine levels. An increase or decrease in one or more grades of these items from their initial values was considered as toxicity.

Statistical analysis
The SPSS statistical software version 25.0 (IBM, Armonk, NY, USA) was employed for statistical analysis. Hardy-Weinberg equilibrium (HWE) of genotype frequencies was assessed in 183 patients using the chi-squared test. Categorical variables (such as MTX elimination delay) were assessed by chi-squared or Fisher's exact test. Continuous variables (such as the serum MTX concentration) were expressed as median and interquartile range (IQR). The Mann-Whitney U test was carried out to analyze the associations between serum MTX concentration and SNPs and MTX-related toxicity. The association between MTX-related toxicity and elimination delay was evaluated using Fisher's exact test. The association between SNPs and MTX elimination delay, as well as MTX-related toxicity, was assessed using logistic regression analysis and presented as odds ratio (ORs) and 95% confidence interval (CI). In logistic regression analysis, gender, age, ethnicity, immunophenotype, MTX dose, body surface area, and tyrosine kinase inhibitors (TKI) were included as covariates. A P value of <0.05 was considered statistically significant.

Patient characteristics and genotyping
The characteristics of the 183 patients with ALL are summarized in Table 1 Table 2. Genotype frequencies of 12 SNPs were in the HWE, except for ABCC2 rs2273697 and MTHFR rs1801133. Table 3 summarizes MTX-related toxicities. Leukopenia was the most common with a frequency of 78.14%, followed by neutropenia (67.76%) and thrombocytopenia (57.92%). The frequency of anemia and increased creatinine level was 56.28% and 25.68%, respectively.

MTX concentration, MTX elimination delay, and MTXrelated toxicity
The associations among MTX-related toxicity, serum MTX concentrations, and MTX elimination delay were evaluated using the Mann-Whitney U test or Fisher's exact test. The results are shown in Table 3. Patients with increased creatinine levels had higher serum MTX concentrations at 24, 48, and 72 hours compared with those without increased levels (P = 0.001, P < 0.001, and P < 0.001, respectively). Additionally, the frequency of renal toxicity was significantly higher in patients with than those without MTX elimination delay (P < 0.001).

MTX concentration, MTX elimination delay, and SNPs
Next, we explored the associations of SNPs with serum MTX concentration and MTX elimination delay using the Mann-Whitney U test and logistic regression analysis (see Supplemental  Table 3, http://links.lww.com/BS/A52). Since dasatinib may affect MTX clearance, 10 TKI was included as a covariate in the logistic regression analysis. The serum MTX concentrations of patients with SLCO1B1 rs4149056 TT genotype at 48 and 72 hours were significantly higher than those with the TC or CC genotype (P = 0.014 and P = 0.019, respectively; Fig. 2A and 2B). Patients with SLC19A1 rs2838958 GA or AA genotype had higher serum MTX concentrations at 24 hours than those with GG genotype (P = 0.016; Fig. 2C). Additionally, patients with SLC19A1 rs3788200 AG or GG genotype had higher serum MTX concentrations at 24 hours than those with AA genotype (P = 0.043; Fig. 2D). Serum MTX levels at 48 hours were higher in patients with MTRR rs1801394 AG or GG genotype (P = 0.045; Fig. 2E). However, no significantly statistical difference was detected in the MTX serum concentrations at 72 hours between patients with MTRR rs1801394 G allele (AG or GG genotype) and AA genotype (P = 0.072; Fig. 2F). Moreover, MTX elimination delay was less in patients with SLC19A1 rs4149056 TC or CC genotype (OR 0.319, 95% CI: 0.138-0.736, P = 0.007).

MTX-related toxicities and SNPs
We analyzed the effects of SNPs on MTX-related toxicity using logistic regression analysis (see Supplemental

DISCUSSION
MTX is a key antitumor drug widely used for the treatment of ALL. The response to MTX greatly varies among patients receiving HD-MTX treatment. In the current retrospective study, we analyzed the association of candidate SNPs with pharmacokinetics and toxicities of HD-MTX in 183 Chinese patients with ALL.
First, we analyzed the associations between serum MTX concentrations, MTX elimination delay, and MTX-related toxicities. As in previous reports, 11,12 the present study showed that renal toxicity was significantly associated with high MTX concentrations and high incidence of MTX elimination delay. Earlier research has demonstrated that hyperbilirubinemia was associated with high MTX concentrations. 12 However, effects of MTX delayed elimination and high MTX concentrations on hematological profiles and liver function were not found in our study.
In general, we support the use of the MTX concentration as an objective and efficacious marker of MTX-related toxicities. SLCO1B1 transporter is encoded by the SLCO1B1 gene, which transports MTX into the cells. 4 Genetic variants of SLCO1B1 have been reported to affect MTX pharmacokinetics. Two variants in SLCO1B1, rs4149081 and rs11045879, have been identified as being associated with MTX-related toxicity and MTX clearance in a genome-wide analysis study. 13 Ramsey et al. confirmed that the SLCO1B1 variants, rs4149056 and rs2306283, were predictors of MTX clearance in patients with ALL. 14,15 In the present study, associations between SLCO1B1 rs4149056 and MTX serum level and elimination delay were established, a result similar to that of den Hoed et al. 16 Additionally, an increased risk of neutropenia was associated with rs4149056 variant (TC or CC genotype). An earlier study in children with juvenile idiopathic arthritis suggested that rs4149056 variant (TC or CC genotype) was significantly associated with MTX gastrointestinal toxicity and TT genotype was a risk factor for developing hepatotoxicity. 17 However, associations of rs2306283, rs4149081, and rs11045879 with MTX metabolism were not established. Differences in these findings indicate that genetic polymorphisms may affect SLCO1B1's function through different SNPs simultaneously. Moreover, Ramsey et al. reported that dasatinib delayed MTX clearance through the mediation of transporter SLCO1B1. 10 There have also been recent case reports of delayed MTX clearance caused by concomitant use of imatinib. 18,19 The TKI mechanisms affecting MTX metabolism need to be studied further.
MTX can also be transported into cells through the cell membrane by binding to the membrane protein encoded by SLC19A1 (also known as RFC1). 3 Therefore, the influence of SLC19A1 polymorphisms on MTX metabolism remains controversial. In the current study, both SLC19A1 rs2838958 and SLC19A1 rs3788200 mutant alleles were significantly associated with higher serum MTX levels. These findings were not supported by previous studies. 20,21 A study conducted in Slovenia suggested that SLC19A1 rs2838958 AA genotype results in a significant level of mucositis; however, no association with leukopenia was reported. 22 Furthermore, our results indicated that SNPs of SLC19A1 have no effect on MTX toxicity. These inconsistent results may be explained by the differences in study populations and treatment regimens.
The ABC transporter family is one of the largest membrane transporters mediating the export of MTX from cells. 5 ABCC2 polymorphisms, especially rs717620, have been confirmed as contributing to the variability of MTX kinetics. 23 However, the influence of ABCC2 rs717620 on MTX levels, MTX clearance, and hematological toxicity was at variance with the results of several clinical studies. [24][25][26][27] The effects of ABCC2 rs717620 on MTX metabolism were not observed in our study. Neither MTX serum level nor MTX elimination delay was associated with ABCC2 rs3740065, as reported by Liu et al. 26 Furthermore, we did not find any association between ABCC2 rs3740065 and MTX toxicity. Conversely, Lopez-Lopez et al, in a cohort of 151 children with ALL, suggested that the ABCC2 rs3740065 G allele was significantly associated with high MTX concentration at 72 hours and an elevated risk of MTX-related toxicity. 20 Although no genetic variant of ABCC2 rs2273697 was found to have an impact on serum MTX concentration and MTX elimination delay as previously reported, 23,28 patients with the ABCC2 rs2273697 A allele had a higher risk of developing hepatoxicity (increased AST). Genetic variability in ABCC4 rs9302061 and rs7317112 did not influence MTX metabolism, a result consistent with that of a previous study. 20 However, den Hoed et al. 16 carried out a study on ABCC4 rs7317112 in Dutch children with ALL and indicated that the AA genotype was related to mucositis and high MTX plasma levels at 48 h.
Furthermore, we demonstrated that MTRR rs1801394 G allele (AG or GG genotype) was associated with high serum MTX concentrations. Nonetheless, a study from the Netherlands found that MTRR rs1801394 AA genotype was associated with higher MTX levels. 16 MTRR rs1801394 polymorphism had no impact on MTX elimination delay and MTX toxicity in the current study. Associations of MTRR rs1801394 with plasma MTX level and MTX-induced toxicity were not observed in Turkish children with ALL. 29 Furthermore, the association of this polymorphism with MTX-related toxicity was not detected in a study of 162 Chinese patients with rheumatoid arthritis treated with MTX. 30 Further studies are warranted to determine the effects of polymorphism on MTX metabolism.
The decreased activity of MTHFR enzyme may affect MTX efficacy and toxicity. Two variants, rs1801131 and rs1801133, in MTHFR are the most extensively investigated polymorphisms in genetic variation studies of patients with ALL. 4 Both MTHFR rs1801131 AC or CC genotype and AA genotype were identified as risk factors for MTX-related toxicities in several studies. 7,8 The majority of studies have revealed that MTHFR rs1801133 CT or TT genotype had a risk of high MTX plasma concentration, MTX elimination delay and MTX-related toxicity. 8,31 Several other studies have suggested that the MTHFR rs1801133 CC genotype had a high risk of MTX-related toxicity. 7,32, 33 Han et al claimed that MTHFR rs1801133 T allele and rs1801131 AA genotype were risk factors for hematopoietic toxicity in a study of 157 adult Chinese patients with hematological malignancies. 34 Several previous studies had shown neither rs1801131 nor rs1801133 polymorphisms in MTHFR were significantly associated with MTX-related toxicity and delayed MTX clearance in pediatric patients with ALL. 29,35,36 Unfortunately, associations of MTHFR rs1801131 and rs1801133 polymorphisms with MTX metabolism are not yet established. Given the cumulative evidence, whether SNPs of MTHFR can be used as predictors of MTX concentration and MTX-related toxicity remains controversial.
The results of studies on the association between genetic polymorphisms and MTX metabolism are inconsistent, which could be attributed to small sample sizes, different toxicity criteria, varied treatment protocols, and racial differences. The advantages of the current study include unified diagnostic criteria, treatment regimens, and quantifiable and objective toxicity data from a single center. Nonetheless, the retrospective study was limited to MTX-related toxicities, which included hematologic toxicity, hepatotoxicity, and renal dysfunction but did not analyze other common drug side effects, such as oral mucositis and vomiting or diarrhea. The current findings require further substantiation.

CONCLUSION
The current results indicated that MTX serum levels forecast MTX-related toxicity. Genetic polymorphisms involved in the MTX metabolism pathways, such as SLCO1B1 rs4149056, SLC19A1 rs2838958, SLC19A1 rs3788200, and MTRR rs1801394, were associated with MTX serum levels in adult ALL. Moreover, genetic polymorphisms, such as SLC19A1 rs4149056, ABCC2 rs2273697, and MTRR rs1801394, were related to MTX-related toxicities.