Introduction

Acute lymphoblastic leukemia (ALL) is the eleventh most common cancer worldwide. In Jordan, ALL is the most common cancer in childhood accounting for almost 30% of childhood cancers [1, 2]. Treatment of ALL usually consists of 3 phases: (a) induction phase, which usually includes glucocorticoids, vincristine, l-asparaginase and anthracyclines; (b) consolidation phase, which lasts for 8 weeks, where high doses of methotrexate (MTX) along with 6-mercaptopurine (6-MP) are given, and (c) maintenance phase, which lasts for 2–3 years [3]. In recent years, cure rates of ALL have increased to approximately 80% due to the use of multi-chemotherapeutic agents regimen, However, treatment-related toxicities still present a challenge for clinicians and may lead to interruption or discontinuation of treatment [4, 5]. One of the chemotherapeutic agents used that leads to treatment-related toxicity which in some cases could be life-threatening is MTX [6].

MTX is an anti-metabolite, specifically, a folate antagonist. Its direct target is dihydrofolate reductase (DHFR) enzyme which is responsible for the conversion of dihydrofolate into tetrahydrofolate that plays a crucial role in cell growth and cellular metabolism [7]. DHFR enzyme is encoded by the DHFR gene that is located on chromosome no. 5 and spans 30 kb, inhibition of DHFR by MTX raises plasma homocysteine levels by 50%, which reflects a low folate status, this is particularly observed in patients with 19 bp deletion in intron 1 of DHFR [8].

Inhibition of DHFR also affects methylene tetrahydrofolate reductase enzyme (MTHFR) which converts 5,10 methylene tetrahydrofolate into 5 methylene tetrahydrofolate; the latter being the main circulatory form of folate which acts as a co-substrate for the conversion of homocysteine to methionine. Afterwards, methionine is converted to S-adenosylmethionine which serves as a methyl donor for many biological compounds such as nucleic acids and proteins [9]. More commonly studied polymorphisms of MTHFR gene are C677T and A1298C which are both known missense polymorphisms associated with decreased enzyme activity and have been linked to MTX toxicity not only in ALL but also in rheumatoid arthritis, osteosarcoma, and lymphoma [10,11,12]. C677T transition results in substitution of alanine to valine while A1298C transversion leads to the substitution of alanine to glutamate [13, 14].

Thymidylate synthase enzyme is another target of the anti-folate, MTX. It is responsible for the initial step in DNA synthesis, in which deoxythymidine monophosphate, a precursor needed for DNA synthesis, is produced from deoxyuridine monophosphate, resulting in the oxidation of methylene tetrahydrofolate to dihydrofolate [15]. This enzyme is coded by the gene thymidylate synthase (TYMS), located on chromosome 18p11.32 and is composed of six introns with sizes ranging between 507 and 6271 bp, and seven exons with sizes ranging between 72 and 250 bp [16]. One of the polymorphisms seen in TYMS is the 6-bp deletion at nucleotide 1494 of the TYMS mRNA which has been associated with decreased mRNA stability in vitro and lower TYMS expression in vivo [17]. Another gene involved in the folate pathway is the dihydropyrimidine dehydrogenase (DPYD) which is located on the short arm of chromosome number one (1p22), and is known to be highly polymorphic with approximately 40 different polymorphisms, of these, are T85C and T496C [18]. DPYD T496C is a missense mutation, in which valine substitutes methionine in the protein product at position 166, while in DPYD T85C cysteine is substituted for arginine [19].

As multiple genes affect MTX treatment outcomes in patients with ALL, the aim of this study was to assess the relationship between folate pathway gene polymorphisms and high-dose MTX (HDMTX) toxicity in children with ALL.

Materials and methods

Study population

This is a cross-sectional observational study with toxicity data of 50 patients collected retrospectively while toxicity data of the remaining 14 patients were collected prospectively over a period of 8 weeks of consolidation. It involved 64 children with ALL recruited from the Royal Medical Services (RMS)/Amman that were diagnosed between the year 2013 and 2017 and treated according to modified St. Jude total XIII protocol [20]. Patients were eligible for recruitment if they fulfilled the following criteria: (1) younger than 18 years, (2) diagnosed with ALL as their primary disease by an oncologist, and (3) lacking other malignancies. Patients were recruited between March, 2017 and December, 2017. The research was approved by the ethics committee in RMS (institutional review board, TF3/1/IRB/1762 on February 14th, 2017) in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration (IRB no. 1762, 14/2/2017). Informed consent was obtained from all participants and their parents or legal guardians before inclusion into the study.

Treatment plan

Patients were stratified according to the modified St. Jude protocol into two groups: low and high risk, where patients in the high-risk group received 5 g/m2 MTX and patients in the low-risk group received 2.5 g/m2 during the consolidation phase. The consolidation phase consisted of 8 weeks; during this period MTX was given 4 times with 2-week intervals in addition to daily doses of 6-MP 50 mg/m2.

Data collection

All patients’ guardians in this study were interviewed, and a detailed data collection sheet was filled. Information gathered from the patients or their caregivers included age, gender, clinical history of the patient and his or her family history. The date of starting induction, consolidation, and maintenance treatments, immunologic subtype, risk classification, the total dose of MTX, body weight, and body surface area (BSA) were collected from the patients’ medical files without further confirmation.

At the beginning and after each of the four cycles of MTX therapy, the following parameters were collected when available: total dose of MTX, measured MTX concentrations at 48 h after MTX infusion, kidney function test, including; blood urea nitrogen (BUN) and serum creatinine (Scr), liver function test, including transaminase enzymes, and blood test, including WBCs, absolute neutrophil count (ANC), platelets, and hemoglobin values. Other toxicity parameters such as a delay in MTX cycle, incidence of infections, and number of “packed RBCs given” were also recorded.

Genotyping

Three milliliters of peripheral venous blood were collected from patients in K3EDTA-coated tubes and DNA extraction was performed by Promega Wizard® DNA purification kit (Promega Corporation, USA) according to manufacturer’s instructions. Polymerase chain reaction (PCR) technique was conducted using a thermal cycler (Bio-Rad®, USA).

DHFR 19 bp deletion, MTHFR C677T, MTHFR A1298C, DPYD T496C, DPYD T85C and TYMS 1494del6 polymorphisms were analyzed using PCR amplification followed by sequencing [21]. Primers used in the amplification process are shown in Table 1. All primers were synthesized by Princess Haya Biotechnology Center (Irbid, Jordan). Sequences were analyzed using the sequencing analysis software (Chromas Lite, version 2.1.1).

Table 1 Primers used in this study
Full size table

Statistical analysis

Data were analyzed using SPSS software version 22 (SPSS® Inc, Chicago, USA). Descriptive data were expressed as count and percentage for categorical variables or as mean ± SD for continuous data. Lowest values of hematological toxicities during the consolidation phase were graded according to the National Cancer Institute criteria version 3 [22]. Presence of hepatic toxicity was considered significant when aspartate transaminase (AST) or alanine transaminase (ALT) levels were three times the upper normal limit. Published research represents toxicities related to MTX utilizing different methods, and we will adopt two of these methods. First method used is the worst toxicity during the four MTX cycles [23], whilst the second one is total MTX courses taken by the patients [24]. Discrete variables were compared using Chi-square or Fisher exact tests as appropriate. Common odds ratios (OR) and 95% confidence intervals (95% CI) were calculated as a measurement of association among discrete variables. All tests were two-tailed and p value < 0.05 was considered to be statistically significant.

Haplotype analysis

The interaction between genetic polymorphisms at the two loci was assessed by evaluating the combined genotype effects and haplotype analysis. Haplotype frequencies were calculated using Multiallelic Interallelic Disequilibrium Analysis Software [(MIDAS®), University of Southampton, Highfield, Southampton, UK] [25] and linkage disequilibrium was represented by Lewontin’s coefficient (D′).

Results

Patient demographics

This study included 64 children with ALL aged 2–16 years (mean 7.7 ± 4 years); 31 boys and 33 girls. The most common immunophenotype was B-lineage (N = 55, 85.9%); however, few carried the T-lineage (N = 9, 14.1%). Most of the T-lineage carriers were observed in boys (N = 8, 89%). Patients’ characteristics and treatment related data are summarized in Table 2. We analyzed the distribution of genotypes according to different patient characteristics (gender, age, risk status) to ensure internal validity and there was no significant association between the studied polymorphisms and patients’ demographics (Supplementary Tables 1, 2, 3).

Table 2 Demographics and clinical data
Full size table

MTX-related toxicities

Hematologic toxicities were manifested the most, as 40 (62.5%) patients developed grade 3–4 neutropenia and 55 (85.9%) were neutropenic. Similarly, all patients had leucopenia with 20 (31.3%) patients had grade 3 /4. A fewer number of patients developed grade 3–4 thrombocytopenia [7 (11.1%)], and grade 3 anemia [5 (8.1%)]. Regarding non-hematologic toxicities, 22 (40%) and 11 (19%) patients had AST and ALT levels three times higher than the upper normal limit, respectively. Data reflecting renal toxicity (Scr and BUN) were also collected but no reading was above normal values.

Genotype, allele, diplotype and haplotype frequencies

Genotype and allele frequencies are described in Table 3. The minor allele frequencies (MAF) for DHFR 19 bp deletion, MTHFR C677T, MTHFR A1298C, DPYD T85C, DPYD T496C and TYMS, were 0.336, 0.414, 0.313, 0.198, 0.103, and 0.46, respectively. The most common genotype in MTHFR C677T, MTHFR A1298C and TYMS 1494del6 was heterozygous [N = 29 (45.3%), N = 30 (46.9%), and 29 (45.3%), respectively]. On the other hand, the most common genotype in DHFR 19 bp deletion, DPYD T85C, DPYD T496C and TYMS 1494del6 was homozygous dominant [N = 32 (50%), N = 43 (68.3%) and N = 50 (79.4%), respectively].

Table 3 Genotype and allele frequencies
Full size table

The most common MTHFR diplotype was CT/AC [N = 19 (29.7%)] followed by TT/AA [N = 11 (17.2%)] while the least common MTHFR diplotype was TT/CC [N = 0 (0%)] followed by TT/AC and CT/CC with equal percentages (N = 1; 1.6%). The diplotype TT/CC was not encountered in this study. There was a strong linkage disequilibrium (D′ = 0.8, p = 0.0003) between C677T and A1298C, where the distance between the two single nucleotide polymorphisms (SNPs) is 1092 bp. All four haplotypes were observed, with T_A haplotype having the highest frequency (N = 50, 39%) and T_C haplotype having the lowest frequency (N = 3, 2.3%), which is consistent with the diplotype analysis. MTHFR haplotype frequencies are summarized in Table 4.

Table 4 Haplotype analysis of MTHFR C677T and A1298C
Full size table

All four haplotypes of DPYD were observed, with T_T haplotype having the highest frequency (N = 95; 75.4%) and the C_C haplotype having the lowest frequency (N = 6; 4.8%). Relatively moderate strength linkage disequilibrium was observed (D′ = 0.37, p = 0.04). It should be noted that the distance between the two DPYD SNPs is 183,794 bp, still there was significant linkage disequilibrium. There were only 6 diplotypes of DPYD because of the lack of a homozygous minor genotype of DPYD T496C. The most common DPYD diplotype was TT/TT (N = 38; 59.4%) and the least common diplotype was CC/TT (N = 2; 3.2%) and there was no CC/CC (0; 0%). Diplotype analysis was consistent with haplotype data. DPYD haplotypes are summarized in Table 5.

Table 5 Haplotype analysis of DPYD T85C and T496C
Full size table

Association of polymorphisms with MTX toxicity

Analysis of the effect of genetic polymorphisms on different MTX toxicities during the consolidation phase showed that a significant association exists between DHFR 19 bp deletion, MTHFR C677T and leucopenia (Table 6). Carriers of the homozygous dominant genotype of both genes are at an increased risk of developing the more serious grades of leucopenia. Carriers of insertion/insertion (Ins/Ins) genotype of DHFR are 5.4 times more likely to develop grade 3 and grade 4 leucopenia compared to minor allele carriers [Ins/Ins: 50% vs. Ins/Del + Del/Del: 15.6%; OR (95% CI) = 5.4 (1.6–17.8); p = 0.006] while carriers of one C allele of MTHFR C677T are 4.5 times more likely to develop grade 2 and higher leucopenia compared to TT carriers [CC + CT: 84% vs.TT: 53.8%; OR (95% CI) = 4.5 (1.2–17); p = 0.03] (Table 6). Regarding MTHFR A1298C, carriers of the minor allele C are at an increased risk of developing neutropenia [AC + CC: 97% vs. AA: 78.6%, OR (95% CI) = 6.1 (1.3–29.5); p = 0.04]. Furthermore, carriers of the deletion allele of TYMS 1494del6 are at risk of developing neutropenia compared to the insertion genotype carriers [Del/Del + Del/Ins: 94% vs. Ins/Ins: 71.4%, respectively. OR (95% CI) = 6 (1.2–31.1); p = 0.04].

Table 6 Association between folate gene polymorphisms and MTX toxicity
Full size table

When the results were analyzed in light of “total MTX courses” carriers of the dominant allele and dominant genotype of MTHFR C677T needed more frequent MTX cycles and showed grade 3 and grade 4 neutropenia [CC + CT: 41.6% vs. TT: 21.4%; OR (95% CI) = 2.6 (1–6.8); p = 0.054 and CC: 43.3% vs. TT: 21.4%; OR (95%): 2.5 (1.01–6.5); p = 0.046]. This observation comes in line with the findings of the first method in assessing toxicity. Similarly, carriers of the MTHFR 1298A dominant allele are at a higher risk of leucopenia [AA: 15 (21.4%) vs. AC: 8 (9.3%); OR (95% CI): 2.7 (1.05–6.7); p = 0.042 and AA vs. AC + CC: 11 (11.3%); OR (95% CI): 2.1 (0.91–5.0); p = 0.076].

Furthermore, a significant association was found between carriers of DHFR 19 bp insertion/insertion genotype and grade 3 and higher leucopenia compared to patients carrying the minor allele [Ins/Ins: 21.2% vs. Ins/Del + Del/Del: 10.3%; OR (95% CI): 2.3 (0.98–5.6); p = 0.052]. Additionally, there was a significant association between TYMS 1494 deletion allele carriers and the worst degrees of leucopenia as they were 2.8 times more likely to develop grade 3 and 4 leucopenia compared to the minor genotype carriers [Ins/Del + Del/Del: 24 (18.6%) vs. Ins/Ins = 2 (5%); OR (95% CI) = 2.8 (1.1–7.5); p = 0.044].

On the other hand, no significant association was found between DPYD genetic polymorphisms and MTX toxicities irrespective of assessment methods. Moreover, no significant association was found between any of the studied genetic polymorphisms and non-hematologic toxicities (hepatic, renal, incidence of infections, MTX level > 1 µl and delay in MTX cycle). Data on association of (MTHFR C677T and A1298C), (DHFR 19 bp deletion), (DPYD T85C and T496C), and (TYMS 1494del6) with severe (grade3/grade4) toxicity are detailed in Supplementary Tables 4–7.

Discussion

In this study, we evaluated the influence of genetic polymorphisms in the folate genes on HDMTX toxicity in children with ALL. Study findings showed that out of all the toxicities studied, hematologic toxicities were affected the most. All patients suffered from leucopenia and 85.9% had neutropenia indicating that WBCs are affected the most among all hematologic toxicities. The least observed hematologic toxicity was anemia, with 5 (8.1%) patients having grade 3 anemia. The severity of myelosuppression is affected by the life span of cells. Because RBCs survive 120 days, clinically significant anemia is least expected. Platelets survive around 10 days, and granulocytes survive approximately 6–8 h [26]. This explains the variability in severity of the observed hematologic toxicities. Similar trends were observed previously [23, 24]. Regarding genetic polymorphisms, we found that the MTHFR 677C>T polymorphism was associated with leucopenia. Specifically, ALL patients harboring the dominant allele exhibited a substantial increase in the incidence of worst degree leucopenia. This was consistent with the “total MTX courses” method, as carriers of the dominant allele were at an increased risk of neutropenia compared to TT carriers. As for MTHFR A1298C, we found an association between this polymorphism and neutropenia as well as leucopenia. This was the opposite with respect to MTHFR C677T. This can be explained by the strong linkage disequilibrium existing between the minor 677T allele and the dominant1298A allele (39%) (Table 4). Dominant alleles with genes involved in the folate pathway were associated with the worst outcome of adverse drug reactions (leucopenia and neutropenia), in contrast with a number of studies where the minor allele was the one associated [11, 23, 27,28,29,30,31]. On the other hand, similar trends were reported by four studies in German, Uruguayan, Canadian and Mexican children with ALL. In Germany, 34 children with ALL were included in a study that had interesting findings: the major genotype of C677T was associated with an increased risk of developing grade 3 and 4 leucopenia. Also, it was concluded that carriers of 1298CC had an increased incidence of anemia [24]. Similarly, 41 Uruguayan patients were studied, where carriers of the 677TT genotype had a protective effect against hematologic toxicities decreasing the risk by four times [32]. Additionally, the Canadian study showed that carriers of the minor allele of MTHFR C677T had lower rates of grade 3 leucopenia [33]. The Mexican studied the effect of MTHFR on the outcome of ALL in 109 patients and concluded that carriers of the dominant genotype 677CC had three times higher risk of developing mucositis compared to minor allele 677T [34].

The conflicting results reported to date are likely due to differences in MTX dosing, dose adjustments in the case of toxicity, and schedules. Recruited patients in different studies were treated by different protocols (St Jude’s Children Research Hospital protocols and their modifications [23, 27, 34]; Berlin–Frankfurt–Muenster protocols and their modifications [28, 29, 31]; Associazione Italiana Ematologia Oncologia Pediatrica [30]; SHOP [35]; HYPERCVAD–MA [35]; CODOX–M–IVAC [35]; MTX–ARAC [35]; GIMEMA ALL [36]; or multiple protocols [11, 24, 30,31,32,33, 35, 36]). Incomplete ascertainment of toxicity, especially in retrospective studies, is one of the reasons for conflicting findings (prospective studies [27, 28, 31,32,33, 35, 36] vs. retrospective studies [11, 23, 24, 30, 34]). Additionally, some studies evaluated toxicities during maintenance phase [28, 35] while others were during consolidation phase [11, 23, 24, 27, 30,31,32,33,34, 36].

A further confounding factor is the age of participants and cancer type. Although the majority of studies were conducted on pediatric subjects, some were evaluated in adults [32, 36]. Some studies included patients with malignant lymphoma [11] or non-Hodgkin lymphoma [32] besides ALL. Moreover, modest and heterogeneous sample sizes were recruited in different studies, which might impact their statistical power to detect clinically meaningful differences. Small sample sizes were reported (34 [24]; 40 [27, 28]; 41 [32, 36]; 47 [29]), while larger sizes were reported as well (109 [34]; 122 [36]; 126 [37]; 127 [23]; 151 [30]; 186 [33]; 286 [31]). Researchers utilized different methods in genotyping which contributes to the conflicting findings. Most of studies adopted polymerase chain reaction/restriction length polymorphisms [23, 27, 28, 30,31,32, 34, 36]. Other methods included TaqMan SNP genotyping assays [11], GeneProof [38] and in situ hybridization with subsequent enzymatic color reaction [24].

To the best of our knowledge, this is the second study that finds an association between DHFR 19 bp deletion and MTX toxicity in children with ALL. Carriers of the dominant genotype of DHFR are more likely to develop the worst degrees of leucopenia compared to the minor allele carriers. The same was observed when “total MTX courses” method was considered. On the other hand, Giletti et al. found no association between DHFR 19 bp deletion and MTX hematologic toxicity [32]. This lack of association was also observed in a study done on 122 adult Italian patients [36]. A recently published Spanish study revealed that carriers of at least one insertion allele had a lower median level of WBCs [35]. Regarding TYMS 1494del6, we found considerable results as patients with the deletion allele were at 6 times higher risk of developing neutropenia compared to patients with the minor genotype and 4.5 more likely to develop grade 3 and 4 leucopenia when “total MTX courses” method was used. On the other hand, another published study indicated a lack of association [23]. Concerning DPYD polymorphisms, none of the toxicities evaluated revealed association. None of the previously published studies (N = 13) evaluated DPYD genetic polymorphisms in children with ALL treated with HDMTX [27, 30, 39, 40].

We decided to utilize two different methods to assess toxicity. Out of 13 different studies reviewed, only two studies evaluated toxicities by “total MTX courses” methods. Generally, we observed the overall concordance between the two methods. However, the interpretations of current findings should be taken with caution because of a number of limitations;. one of them is the magnitude of missing data. It should be noted that such shortcoming was observed in previous studies [11, 23, 24]. Also the nature of the study should be taken into account as the data were collected retrospectively.

In conclusion, this pharmacogenetic study is the first to be carried out on a sample of Jordanian children with ALL. Results showed that the dominant alleles in three genes (DHFR, MTHFR, TYMS) were associated with hematologic toxicity, but none of them was associated with MTX level, hepatotoxicity or renal function.