2025 CPIC指南：基于TPMT和NUDT15基因型的硫嘌呤剂量（更新版）

DRUGS: THIOPURINES

Background

Three thiopurines are used clinically: azathioprine, mercaptopurine, and thioguanine. Although all three medications share many of the same pharmacologic effects, mercaptopurine and azathioprine are generally used for nonmalignant immunologic disorders, and mercaptopurine and thioguanine are used for hematological malignancies. Because azathiopurine is converted to mercaptopurine in cells, the two drugs can be considered to have identical interactions with TPMT and NUDT15. Recommendations for individuals with functional variants in one or both of these genes will be addressed in detail in the following sections.

# GENES: TPMT AND NUDT15

Background

TPMT. TPMT activity is inherited as a monogenic, autosomal codominant trait$^{6}$,$^{7}$ (Figure S1). Three TPMT single nucleotide variants (SNVs), which result in unstable proteins and enhanced TPMT protein degradation,$^{8}$ account for over 90% of low activity phenotypes in current literature and are the most common inactivating alleles studied to date: c.238G>C (TPMT *2), c.460G>A (*3B), and c.719A>G (*3C).$^{9}$11 Inherited TPMT deficiency is the primary genetic cause of thiopurine intolerance in individuals of European and African descent.

TPMT is expressed in red blood cells, liver, and hematopoietic stem cells. TPMT catabolizes mercopurine to an inactive methylmercopurine base, leaving less parent drug available for potential anabolism to pharmacologically active thioguanine nucleotides (TGNs, Figure 1). The intermediate metabolite of mercapopurine, thioinosine monophosphate (TIMP), is also a substrate for TPMT. MethylTIMP (MeTIMP) and its further phosphorylated metabolites (methylmercaptopurine nucleotides, MeMPNs, which include MeTIMP, MeTIDP, and MeTITP) have pharmacologic activity: they inhibit de novo purine synthesis and may contribute to some of the adverse effects of mercaptopurine, particularly hepatotoxicity.$^{12,13}$ Individuals who inherit two no-function TPMT alleles (homozygous or compound heterozygous TPMT-deficient individuals) are at high risk for life-threatening myelosuppression due to the generation of excessive TGNs if given standard doses of mercaptopurine (or azathioprine). There is also a growing body of evidence linking TPMT deficiency to the risk of hepatic sinusoidal obstructive syndrome in patients exposed to 6-thioguanine (6-TG) which appears to be highly dose dependent in acute lymphocytic leukemia (ALL) and inflammatory bowel disease (IBD) and influenced by the treatment regimen.$^{14,15}$

Thirty to sixty percent of TPMT heterozygotes for a no-function allele (or TPMT intermediate metabolizers) cannot tolerate full doses of mercaptopurine or azitopirine.$^{12,16}$ Good thiopurine tolerance in some TPMT intermediate metabolizers may be explained by the fact that they produce less MeMPNs than

do normal metabolizers.$^{17}$ The wide variability in thiopurine tolerance in IMs suggests that precise dosing in each patient cannot be determined by TPMT genotype alone.

Although TPMT has lower affinity for thioguanine than for mercaptopurine, TPMT significantly affects thioguanine pharmacokinetics and its cytotoxic effects.$^{18}$ Thioguanine is directly metabolized by TPMT to inactive methylthioguanine base, leaving less drug available for anabolism by hypoxanthine-guanine phosphoribosyltransferase (HPRT) and other enzymes to generate active TGN metabolites (Figure 1). In contrast to mercaptopurine, thioguanine is metabolized by HPRT into TGNs without involving intermediate metabolites such as TIMP (i.e. there are no methylTIMP produced from thioguanine). Within each TPMT phenotypic group, the initial recommended relative dosage decreases are similar for thioguanine, mercaptopurine, and azathologie. Complementary biochemical laboratory tests that directly measure TPMT activity can be helpful adjuncts to genotyping tests, which is further discussed in the “Other Considerations” in this article. However, at present, similar biochemical assays are not readily available for NUDT15 activity (Other Considerations and Supplement).

NUDT15. Through agnostic genome-wide association studies, variants in NUDT15 were identified that strongly influence thiopurine tolerance in patients with ALL$^{19}$ and IBD.$^{20}$ This hydrolase is expressed in most tissues. Decreased and no-function alleles in NUDT15 explain the majority of severe thiopurine-related myelosuppression in patients of Asian descent and are common in patients with Amerindian genetic ancestry. The NUDT15 poor metabolizer phenotype is observed at a frequency of about 1 in every 100 patients of East Asian descent (NUDT15 Frequency Table$^{21}$), which is more common than the TPMT poor metabolizer phenotype in populations of European descent (estimated to be 0.19%, TPMT Frequency Table$^{21}$). NUDT15 deficiency is also more prevalent in individuals who self-report as Hispanic, particularly those with high levels of Amerindian genetic ancestry.$^{19}$ Given the prevalence of NUDT15 variants across populations, particularly among non-European populations, inclusion of NUDT15 in genetic testing is of particular clinical importance to ensure equitable testing and treatment in admixed populations, such as that of the United States. As a nucleoside diphosphatase, NUDT15 catalyzes the conversion of cytotoxic metabolite thioguine triphosphate (TGTP) to the noncytotoxic methaqine monophosphate (TGMP). Defects in NUDT15-mediated degradation of TGTP result in higher concentrations of TGTP available for incorporation into DNA (DNA-TG, the primary antileukemic metabolite$^{22}$), eventually leading to excessive DNA damage and apoptosis. The pArg139Cys variant (NUDT15 *3; rs116855232; c.415C>T) was the first NUDT15 variant linked to thiopurine toxicity. It was shown that this amino acid change results in a nearly complete loss of enzymatic activity and protein stability in vitro. Patients carrying this allele showed excessive DNA-TG and severe myelosuppression.$^{23}$ In children with ALL, patients homozygous for the pArg139Cys variant allele tolerated only 8% of the standard dose of mercaptopurine, whereas the tolerated dose intensity values were 63% and 83.5%

![graph TD\n A1[6-Thioguanine(6TG)] --> A2[Thioxanthine]\n A1 --> A3[GDA]\n A1 --> A4[meTG]\n A1 --> A5[TGTP]\n A2 --> A6[XDH]\n A3 --> A6\n A4 --> A7[8-OHTG]\n A5 --> A8[TdGTP]\n A6 --> A7\n A6 --> A11[ABCC4]\n A7 --> A11\n A7 --> A12[AOX1]\n A8 --> A9[TdGDP]\n A9 --> A10[TMP]\n A9 --> A8\n A10 --> A18[MPR]\n A11 --> A13[Thiouric acid]\n A12 --> A13\n A13 --> A1\n A14[ABCC5] --> A15[TGMP]\n A15 --> A10\n A16[SLC29A2] --> A17[6TG]\n A17 --> A7\n A17 --> A18\n A18 --> A19[ITPA]\n A19 --> A20[TITP]\n A20 --> A21[HPRT1]\n A21 --> A22[IMPDH1]\n A22 --> A23[TMP]\n A23 --> A24[TXMP]\n A24 --> A25[GMPs]\n A25 --> A15\n A26[SLC29A1] --> A17\n A26 --> A4\n A27[SLC29A2] --> A4\n A27 --> A28[TPMT]\n A28 --> A29[NUDT15]\n A28 --> A30[NME1]\n A28 --> A31[NME2]\n A29 --> A30\n A29 --> A32[RRM1]\n A29 --> A33[RRM2]\n A30 --> A31\n A31 --> A32\n A32 --> A33\n A33 --> A34[TdGDP]\n A34 --> A35[TGDP]\n A34 --> A36[TdGTP]\n A35 --> A37[NUDT15]\n A36 --> A38[NME2]\n A37 --> A35\n A38 --> A39[RAC1]\n A39 --> A40[TGTP]\n A40 --> A41[RNA damage]\n A41 --> A42[apoptosis]\n A43[SLC29A2] --> A44[TPMT]\n A44 --> A45[meMP]\n A45 --> A46[ADK]\n A46 --> A47[meMPR]\n A47 --> A48[Hepatotoxicity]\n A49[SLC29A1] --> A44\n A49 --> A50[meTG]\n A51[SLC28A3] --> A52[TPMT]\n A52 --> A53[meTIMP]\n A53 --> A54[PPAT]\n A54 --> A55[PRA]\n A55 --> A56[Purine Synthesis]\n A57[SLC28A2] --> A58[PRPP]\n A58 --> A59[Ribose 5P]\n A58 --> A60[ADK]\n A60 --> A61[meMPR]\n A60 --> A62[TPMT]\n A62 --> A63[HPRT1]\n A62 --> A64[MPR]\n A63 --> A65[ITPA]\n A64 --> A66[NT5C2]\n A65 --> A66\n A66 --> A67[PRPP]\n A67 --> A68[IMPDH1]\n A68 --> A69[TMP]\n A69 --> A70[AdoMet]\n A69 --> A71[AdoHcy]\n A70 --> A72[NME2]\n A73[GSTA1] --> A74[GSTM1]\n A73 --> A75[GSTA2]\n A74 --> A76[Mercaptopurine]\n A75 --> A76\n A76 --> A77[XDH]\n A76 --> A78[SLC29A2]\n A76 --> A79[SLC29A1]\n A76 --> A80[SLC28A3]\n A76 --> A81[SLC28A2]\n A77 --> A79\n A77 --> A11\n A78 --> A81\n A79 --> A80\n A80 --> A81\n A81 --> A82[Mercaptopurine]\n A82 --> A83[TPMT]\n A83 --> A84[meMP]\n A84 --> A85[ADK]\n A85 --> A86[meMPR]\n A86 --> A87[Hepatotoxicity]\n A88[XDH] --> A89[Thiouric acid]\n A90[Thioxanthine] --> A89\n A91[6-Thioguanine(6TG)] --> A92[GDA]\n A92 --> A93[8-OHTG]\n A93 --> A94[AOX1]\n A94 --> A95[TGMP]\n A95 --> A96[TGDP]\n A96 --> A97[TGTP]\n A97 --> A98[RNA damage]\n A98 --> A99[apoptosis]\n A100[TGMP] --> A101[TdGMP]\n A101 --> A102[TdGDP]\n A101 --> A103[TdGTP]\n A103 --> A104[DNA damage]\n A104 --> A105[apoptosis]\n A106[AdoMet] --> A107[NME2]\n A108[AdoHcy] --> A109[NME1]\n A110[NUDT15] --> A111[TGDP]\n A111 --> A112[TGMP]\n A111 --> A113[TdGDP]\n A111 --> A114[TdGTP]\n A114 --> A115[DNA damage]\n A115 --> A106\n A116[PRPP] --> A117[PRA]\n A116 --> A118[P5P]\n A117 --> A119[Purine Synthesis]\n A118 --> A119\n A120[meTIMP] --> A121[PPAT]\n A120 --> A82\n A121 --> A122[PRA]\n A122 --> A123[Purine Synthesis]\n A124[RPRS1] --> A125[P5P]\n A124 --> A118\n](1/parseimages/figuree409abc6f25240d6a8d06dba65c130b2_0_0.png) for those heterozygous and wild-type for this SNP, respectively.$^{19}$ Similar findings were obtained in patients treated with mercaptopurine or azathioprine for IBD.$^{24}$ While most clinical studies focused on mercaptopurine and azathioprine due to their more common use, in vitro experiments using laboratory models indicated similar influence of NUDT15 on the cytotoxicity of thioguanine.$^{23}$ The variant p. Arg139Cys has been studied most extensively in patients receiving thiopurine therapy, thus providing the strongest evidence for clinical implementation. Additional variant alleles have been identified with varying prevalence among different ancestral groups and varying degrees of functional effects (NUDT15 Allele Functionality Table and Frequency Table$^{21}$). Most NUDT15 variants are rare and lack definitive clinical impact to rise to actionability, even though some clearly exhibited decreased NUDT15 activity in vitro. Nevertheless, the accumulation of preclinical and clinical data associating some rare variants with thiopurine toxicity motivated their re-assignment as clinically actionable in this update (e.g., NUDT15 * 4 and *9).$^{24,25}$

# Genetic test interpretation

Genetic testing analyzes the DNA sequence at specific nucleotides in the TPMT and NUDT15 genes (Supplement). Each named star (*) allele is defined by the genotype at one or more specific loci.

Allele function and resulting diplotype-to-phenotype assignments were derived using CPIC's established consensus-based framework for allele function classification.$^{26}$

The genotypes that constitute the haplotypes for TPMT and NUDT15 are described in the TPMT Allele Definition Table and NUDT15 Allele Definition Table.$^{21}$ The impact of these haplotypes on enzyme activity is described in the TPMT Allele Functionality Table and NUDT15 Allele Functionality Table.$^{21}$

Table 1 summarizes the assignment of the predicted TPMT and NUDT15 phenotypes, based on example star allele diplotypes, and these assignments are used to link genotypes with thiopurine prescribing recommendations. Flowcharts describing phenotype assignment based on TPMT and NUDT15 genotypes are available in Supplemental Figures S2 and S3, respectively. Of note, the phenotype of “possible intermediate metabolizer,” introduced in the previous guideline, describes an individual carrying one uncertain/unknown function allele PLUS one no-function allele, as this individual should be treated with “at least” the same precautions as would apply to an intermediate metabolizer. This current guideline provides specific dosing recommendations for the TPMT/NUDT15 “compound intermediate metabolizer” phenotype that describes an individual who is both a TPMT intermediate metabolizer and NUDT15 intermediate metabolizer (e.g., carrying one copy of no-function alleles in both genes) and thus requires greater dose reduction than intermediate metabolizers for a single gene. It should also be noted that we have now annotated some alleles as “decreased function” because the corresponding variant proteins retain residual enzymatic activity and their phenotype designation is added to the Allele Functionality Tables.$^{21}$

Although decreased and no-function TPMT and NUDT15 alleles have been identified in multiple populations (TPMT Frequency Table and NUDT15 Frequency Table$^{21}$). one of the

limitations inherent in a clinical genotyping test is that rare or previously undiscovered variants are typically not included. The use of sequencing-based methods may provide more complete information; however, rare alleles appear to account for less than 10% of the variation in TPMT activity.$^{11}$ Even when sequencing-based methods are used, rare variants may not be reported due to a lack of systematic guidelines to classify pharmacogenomic variants at the time of detection in a clinical laboratory setting.

# Available genetic test options

See Supplementary Material and the Genetic Testing Registry (https://www.ncbi.nlm.nih.gov/gtr/) for more information on available clinical testing options.

Incidental findings

There are no diseases or health traits that have been linked to variation in TPMT or NUDT15 in the absence of thiopurine treatment.$^{8,27}$

Linking genetic variability to variability in drug-related phenotypes

There is substantial evidence linking TPMT and NUDT15 genotypes with phenotypic variability in thiopurine response (see Tables S1 and S2). Pre-emptive dose adjustments based on TPMT genotype have reduced thiopurine-induced adverse effects without compromising desired antitumor and immunopasspressive therapeutic effects in several clinical settings (Table S1). Some of the clinical data upon which dosing recommendations are based (Tables 2–4) rely on measures of TPMT phenotype rather than genotype; however, because TPMT genotype is strongly linked to TPMT phenotype,$^{9}$⁻$^{11}$ these recommendations apply regardless of the method used to assess TPMT status. In addition, retrospective studies strongly indicate that patients with decreased or no-function NUDT15 alleles are at excessive risk of thiopurine toxicity if the standard dose is administered. This body of evidence, rather than randomized clinical trials, provides the basis for most of the dosing recommendations in Tables 2–4.

and set (Tables 2–4) rely on measures of TPMT phenotype rather than genotype; however, because TPMT genotype is strongly linked to TPMT phenotype,$^{9}$⁻$^{11}$ these recommendations apply regardless of the method used to assess TPMT status. In addition, retrospective studies strongly indicate that patients with decreased or no-function NUDT15 alleles are at excessive risk of thiopurine toxicity if the standard dose is administered. This body of evidence, rather than randomized clinical trials, provides the basis for most of the dosing recommendations in Tables 2–4.

Therapeutic recommendations

Thiopurines are used to treat malignant and nonmalignant conditions, and thus the approach to dosing adjustments and the propensity to initiate therapy at higher vs. lower starting doses based on TPMT/NUDT15 status may differ depending on the clinical indication. In addition, because the level of thiopurine tolerance is related to genetic ancestry,$^{19}$ the standard starting doses can vary by geographic regions.

Thiopurines have a unique role in the treatment of many malignancies (e.g., ALL). The standard starting doses of thiopurines are generally derived from trials that have predominantly included patients who are TPMT and NUDT15 normal metabolizers (who represent ~90% of the general populations) and received maximal tolerable doses by the standards of anticancer treatment. Hence, full doses should be given to those who are normal metabolizers for TPMT and NUDT15 (Tables 2–4). Thiopurines are also used in nonmalignant conditions for their

|Predicted phenotype|Genotypes|Examples of $diplotypes^{a}$| |---|---|---| |Assignment of predicted TPMT phenotypes based on genotypes||| |Normal metabolizer|An individual carrying two normal function alleles OR one normal function allele PLUS one decreased function allele|*1/*1 *1/*8| |Intermediate metabolizer|An individual carrying one normal function allele PLUS one no-function allele OR two decreased function alleles|*1/*2, *1/*3A, *1/*3B, *1/*3C, *1/*4 *8/*8| |Possible intermediate metabolizer|An individual carrying one uncertain/unknown function allele PLUS one no-function allele or one no-function allele PLUS one decreased function allele|*2/*9, *3A/*12 *3A/*8| |Poor metabolizer|An individual carrying two no-function alleles|*3A/*3A, *2/*3A, *3A/*3C, *3C/*4, *2/*3C, *3A/*4| |Indeterminate|An individual carrying one normal function allele PLUS one uncertain/unknown function allele or one decreased function allele PLUS one uncertain/unknown function allele or two uncertain/unknown function alleles|*1/*19 *8/*30 *19/*40| |Assignment of predicted NUDT15 phenotypes based on genotypes||| |Normal metabolizer|An individual carrying two normal function alleles OR one normal function allele PLUS one decreased function allele|*1/*1 *1/*5| |Intermediate metabolizer|An individual carrying one normal function allele PLUS one no-function allele OR an individual carrying two decreased function alleles|*1/*2, *1/*3 *5/*5| |Possible intermediate metabolizer|An individual carrying one uncertain/unknown function allele PLUS one no-function allele|*2/*15, *3/*21| |Poor metabolizer|An individual carrying two no-function alleles OR one no-function allele PLUS one decreased function allele|*2/*2, *2/*3, *2/*4 *3/*5| |Indeterminate|An individual carrying one normal function allele PLUS one uncertain/unknown function allele OR one decreased function allele PLUS one uncertain/unknown function allele OR two uncertain/unknown function alleles|*1/*12 *5/*20 *11/*14|

ªPlease refer to the TPMT and NUDT15 Diplotype-Phenotype Table online for a complete list. For allele function, please refer to the TPMT and NUDT15 Allele Functionality Table.$^{21}$

immune suppressive activity (e.g., for rheumatoid arthritis, systemic lupus erythematosus, vasculitis, IBDs, autoimmune dermatologic disorders, and myasthenia gravis). The dosing strategy for nonmalignancy differs in that the thiopurine (azathioprine or mercaptopurine) is often initiated at a relatively low dose followed by gradual increases to reach the maximal tolerable dose specified in the protocol. For these conditions, dose adjustments should be made based on close monitoring of clinical myeloppression (or disease-specific guidelines) and/or the measurement of thiopurine metabolites, such as TGNs or MeMPNs, especially in patients with TPMT deficiency.

Major changes from the 2018 guideline. In this update, we have shifted from providing gene-specific dosing recommendations to harmonizing guidance by drug (i.e., mercaptopurine, thioguanine, and azathioprine). Dosing recommendations are now provided for each drug by TPMT/NUDT15 phenotype, in their specific

indication (i.e., malignant or nonmalignant conditions). Figure 2 outlines the recommended course of action for each thiopurine drug if TPMT and/or NUDT15 genotypes are known.

We also introduced specific dosing recommendations for the compound-deficient phenotypes (i.e., intermediate or poor metabolizers for both TPMT and NUDT15) in this guideline update. There have been reports of patients with intermediate metabolizer status for both TPMT and NUDT15 (i.e., TPMT/NUDT15 compound intermediate metabolizers). These individuals tolerated lower doses of thiopurines compared to intermediate metabolizers for only TPMT or NUDT15,$^{28}$ indicating that the decreased activity in both enzymes results in additive effects on drug toxicity. The two genes are independent: the likelihood of an individual being an intermediate metabolizer for both genes depends upon the population frequencies for variant alleles in each gene (see Supplementary Tables). For example, given estimates of no-function alleles for NUDT15 of 11% and

|TPMT/NUDT15 phenotype|Implications for toxicity|Implications for phenotypic $measures^{a}$|Therapeutic $recommendations^{b}$|Classification of $recommend^{c}$| |---|---|---|---|---| |TPMT and NUDT15 normal metabolizer|Normal risk of thiopurine-related leukopenia, neutropenia and myelosuppression|TPMT NMs have lower erythrocyte concentrations of TGN metabolites and higher concentrations of MeMPNs compared to TPMT IMs and TPMT PMs. This is the 'normal' pattern.|Initiate therapy with standard starting dose of mercaptopurine (e.g., 75mg/m²/ day for normalignancy). During therapy, adjust the doses of myelosuppressive agents, as per standard clinical practice. It usually takes at least 2 weeks of stable dosing to reach steady state after each dose adjustment.|Strong| |TPMT normal metabolizer and NUDT15 (possible) intermediate metabolizer OR TPMT (possible) intermediate NUDT15 normal metabolizer|Increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites and low concentrations of MeMPNs compared to TPMT NMs when receiving standard dose.|Initiate therapy with decreased starting doses (30-80% of standard starting dose) if starting dose is ≥ 75mg/m²/day (for malignancy) or ≥ 1.5mg/kg/day (for normalignancy) If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust mercaptopurine doses based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2-4 weeks of stable dosing to reach steady state after each dose adjustment. If myelosuppression occurs, and the patient is on combination therapy, emphasis should be on reducing mercaptopurine over other agents.|Strong| |Any combination with TPMT poor metabolizer and/or NUDT15 poor $metabolizer^{d}$|Greatly increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression. Fatal toxicity possible without dose decrease|TPMT PMs have extremely high erythrocyte concentrations of TGN metabolites and no MeMPN compared to TPMT NMs.|For malignancy: initiate therapy with drastically reduced starting doses. Reduce starting dose by 10-fold and reduce frequency to thrice weekly instead of daily (e.g., 10mg/m²/day given 3days/week) During therapy, adjust mercaptopurine doses based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 4-6 weeks of stable dosing to reach steady state after each mercaptopurine over other agents. For nonalignancy: consider alternative nonhiopurine immunosuppressant therapy|Strong| |TPMT intermediate metabolizer and NUDT15 intermediate metabolizer (i.e., TPMT/ NUDT15 compound intermediate metabolizer)|Increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression. Higher risk of toxicity compared to single IM|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites and low concentrations of MeMPNs compared to TPMT NMs.|Initiate therapy with doses (20-50% of standard starting dose) if starting dose is ≥ 75mg/m²/day (for malignancy) or ≥ 1.5mg/kg/day (for nonalignancy) If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust mercaptopurine doses based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2-4 weeks of stable dosing to reach steady state after each dose adjustment. If myelosuppression occurs, and the patient is on combination therapy, emphasis should be on reducing mercaptopurine over other agents.|Strong|

IM, intermediate metabolizer; MeMPN, methylmercaptopurine nucleotide; NM, normal metabolizer; NUDT15, Nudix hydrolase 15; PM, poor metabolizer; TGN, thiquanine nucleotide; TPMT, thiopurine methyltransferase. ^This section primarily addresses the effects of TPMT genotype, as there are currently no clinically available pharmacological phenotypes associated with NUDT15 genotype. ^Standard starting doses may vary by patient characteristics and/or treatment regimens and according to local clinical practice. ^Rating scheme described in the Supplementary Material. ^This includes being NM, IM or possible IM for one gene and PM for the other gene, as well as being PM for both genes.

|TPMT/NUDT15 phenotype|Implications for toxicity|Implications for phenotypic $measures^{a}$|Therapeutic $recommendations^{b}$|Classification of $recommendation^{c}$| |---|---|---|---|---| |TPMT and NUDT15 normal metabolizer|Normal risk of thiopurine-related leukopenia, neutropenia, and myelosuppression|TPMT NMs have lower erythrocyte concentrations of TGN metabolites compared to TPMT IMs and TPMT PMs. This is the 'normal' pattern.|Initiate therapy with standard starting dose of thioguanine (e.g., 40mg/$m^{2}$/day for malignancy). During therapy, adjust the doses of myelosuppressive agents, as per standard clinical practice. It usually takes at least 2 weeks of stable dosing to reach steady state after each dose adjustment.|Strong| |TPMT normal metabolizer and NUDT15 (possible) intermediate metabolizer OR TPMT (possible) intermediate metabolizer and NUDT15 normal metabolizer|Increased risk of thiopurine-related leukopenia, neutropenia, and myelosuppression|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites compared to TPMT NMs.|Initiate therapy with decreased starting doses (30–80% of standard starting dose) if standard starting dose is ≥40mg/$m^{2}$/day. If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust the doses of thioguanine based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2–4 weeks of stable dosing to reach steady state after each dose adjustment. If myelosuppression occurs, and the patient is on combination therapy, emphasis should be on reducing thioguanine over other agents.|Moderate| |Any combination with TPMT poor metabolizer and/or NUDT15 poor $metabolizer^{d}$|Greatly increased risk of thiopurine-related leukopenia, neutropenia, and myelosuppression. Fatal toxicity possible without dose decrease|TPMT PMs have extremely high erythrocyte concentrations of TGN metabolites compared to TPMT NMs.|Initiate therapy with drastically reduced starting doses. Reduce daily dose by 10-fold and reduce frequency to thrice weekly instead of daily During therapy, adjust thioguanine doses based on degree of myelosuppression and disease-specific guidelines. It usually takes at least 4–6 weeks of stable dosing to reach steady state after each dose adjustment. If myelosuppression occurs, emphasis should be on reducing thioguanine over other agents.|Strong| |TPMT intermediate metabolizer and NUDT15 intermediate metabolizer (i.e., TPMT/NUDT15 compound intermediate metabolizer)|Increased risk of thiopurine-related leukopenia, neutropenia, and myelosuppression. Higher risk of toxicity compared to single IM|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites compared to TPMT NMs.|Initiate therapy with decreased starting doses (20–50% of standard starting dose) if standard starting dose$^{a}$ is ≥40mg/$m^{2}$/day. If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust the doses of thioguanine based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2–4 weeks of stable dosing to reach steady state after each dose adjustment. If myelosuppression occurs, and the patient is on combination therapy, emphasis should be on reducing thioguanine over other agents.|Moderate|

IM, intermediate metabolizer; NM, normal metabolizer; NUDT15, Nudix hydrolase 15; PM, poor metabolizer; TGN, thioguanine nucleotide; TPMT, thiopurine methyltransferase. $^{9}$This section primarily addresses the effects of TPMT genotype, as there are currently no clinically available pharmacological phenotypes associated with NUDT15 genotype. $^{6}$Standard starting doses may vary by patient characteristics and/or treatment regimens and according to local clinical practice. $^{6}$Rating scheme described in the Supplementary Material. $^{6}$This includes being NM, IM, or possible IM for one gene and PM for the other gene, as well as being PM for both genes.

of no-function alleles for TPMT of 2% in the global population, the frequency of carrying no-function alleles in both genes is approximately estimated at 0.2% overall. The growing body of clinical data on thiopurine tolerance in patients who are compound

intermediate metabolizers points to the need for specific dosage recommendation for this group of patients. Although rare, individuals who are intermediate metabolizers for one enzyme and poor metabolizers for the other, or poor metabolizers for both

|TPMT/NUDT15 phenotype|Implications for toxicity|Implications for phenotypic $measures^{a}$|Therapeutic $recommendations^{b}$|Classification $recommend^{c}$| |---|---|---|---|---| |TPMT and NUDT15 normal metabolizer|Normal risk of thiopurine-related leukopenia, neutropenia and myelosuppression|TPMT NMs have lower erythrocyte concentrations of TGN metabolites and higher concentrations of MeMPNs compared to TPMT IMs and TPMT PMs. This is the 'normal' pattern.|Initiate therapy with standard starting dose (e.g., 2 mg/kg/day for autoimmune diseases). During therapy, adjust doses of azathioprine based on disease-specific guidelines. It usually takes at least 2 weeks to reach steady state after each dose adjustment.|Strong| |TPMT normal metabolizer and NUDT15 (possible) intermediate metabolizer OR TPMT (possible) intermediate metabolizer and NUDT15 normal metabolizer|Increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites and low concentrations of MeMPNs compared to TPMT NMs.|Initiate therapy with reduced starting doses (30–80% of standard starting dose) if standard starting dose is ≥2 mg/kg/day. If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust the doses of azathioprine based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2–4 weeks of stable dosing to reach steady state after each dose adjustment.|Strong| |Any combination with TPMT poor metabolizer and/or NUDT15 poor $metabolizer^{d}$|Greatly increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression. Fatal toxicity possible without dose decrease|TPMT PMs have extremely high erythrocyte concentrations of TGN metabolites and no MeMPN compared to TPMT NMs.|Consider alternative nonthiopurine immunosuppressant therapy|Strong| |TPMT intermediate metabolizer and NUDT15 intermediate metabolizer (i.e., TPMT/NUDT15 compound intermediate metabolizer)|Increased risk of thiopurine-related leukopenia, neutropenia and myelosuppression Higher risk of toxicity compared to single IM|TPMT IMs have moderate to high erythrocyte concentrations of TGN metabolites and low concentrations of MeMPNs compared to TPMT NMs.|Initiate therapy with reduced starting doses (20–50% of standard starting dose) if standard starting dose$^{a}$ is ≥2 mg/kg/day. If starting dose is already below standard starting dose, dose reduction might not be necessary. During therapy, adjust the doses of azathioprine based on the degree of myelosuppression and disease-specific guidelines. It usually takes at least 2–4 weeks of stable dosing to reach steady state after each dose adjustment.|Moderate|

IM, intermediate metabolizer; MeMPN, methylmercaptopurine nucleotide; NM, normal metabolizer; NUDLT5, Nudix hydrolase 15; PM, poor metabolizer; TGN, thioguanine nucleotide; TPMT, thiopurine methyltransferase. $^{9}$This section primarily addresses the effects of TPMT genotype, as there are currently no clinically available pharmacological phenotypes associated with NUDT15 genotype. $^{8}$Standard starting doses may vary by patient characteristics and/or treatment regimens and according to local clinical practice. $^{6}$Rating scheme described in the Supplementary Material. $^{9}$This includes being NM, IM, or possible IM for one gene and PM for the other gene, as well as being PM for both genes.

enzymes, should be treated extremely cautiously, with decreased reduction in dosage at least as large as those recommended for poor metabolizers of a single gene.

Mercaptopurine therapeutic recommendations. Tolerated mercaptopurine dosage is correlated with the number of no-function alleles that the patient has in both TPMT and NUDT15.$^{9,10,19,20}$ The degree of mercaptopurine intolerance is largely comparable between carriers of TPMT vs. NUDT15 no-function alleles.$^{19}$

Genotype-guided prescribing recommendations apply primarily to starting doses of mercaptopurine. For TPMT and NUDT15 normal metabolizers, the starting dose does not need to be altered. Variability has been observed in the tolerated thiopurine dosages among TPMT or NUDT15 intermediate metabolizers, with a minority of individuals who do not seem to require significant dose reduction.$^{19,23,29-32}$ If starting doses of mercaptopurine are high (e.g., 75 mg/m$^{2}$ in ALL or 1.5 mg/kg/day for nonmalignant condition), lower than standard starting doses should be considered in TPMT or NUDT15 (possible) intermediate

![graph TD\n A1[TPMT and NUDT15 genotypes known] --> A2[Malignant condition Mercaptopurine and Thioguanine]\n A1 --> A9[Nonmalignant condition Mercaptopurine and Azathioprine]\n A2 --> A3[TPMT NM and NUDT15 NM]\n A2 --> A5[TPMT IM or NUDT15 IM⁺]\n A2 --> A7[TPMT IM and NUDT15 IM⁻]\n A2 --> A10[TPMT PM and/or NUDT15 PM]\n A3 --> A4[Use standard starting dose.]\n A5 --> A6[Consider dose reduction: 30-80% of standard starting dose.]\n A7 --> A8[Consider dose reduction: 20-50% of standard starting dose.]\n A10 --> A11[Consider dose reduction: 10% of standard starting dose, thrice weekly instead of daily.]\n A9 --> A12[TPMT NM and NUDT15 NM]\n A9 --> A14[TPMT IM or NUDT15 IM⁺]\n A9 --> A16[TPMT IM and NUDT15 IM⁻]\n A9 --> A17[TPMT PM and/or NUDT15 PM]\n A12 --> A13[Use standard starting dose.]\n A14 --> A15[Consider dose reduction: 30-80% of standard starting dose.]\n A16 --> A18[Consider dose reduction: 20-50% of standard starting dose.]\n A17 --> A19[Consider alternative immunosuppressant therapy.]\n](1/parseimages/figure7080b1bde3674b2cb0905d470a2374ea_0_0.png) metabolizers$^{29-34}$ (Table 2). On average, TPMT or NUDT15 IM patients can tolerate ~60% of the standard dose of thiopurine, even though it ranges from 30% to 80% across patients.$^{19}$ We acknowledge that our recommended dose range (30–80%) is broad but also believe it is necessary to account for the significant interindividual variability in thiopurine tolerance observed in the IM patient population. In contrast, a full dose of mercaptopurine poses a severe risk of prolonged hematopoietic toxicity in TPMT or NUDT15 poor metabolizers and pre-emptive dose reductions are strongly recommended.$^{16,35-37}$ Thus, markedly reduced doses (10-fold reduction and reduced frequency of drug administration) should be used in TPMT and/or NUDT15 poor metabolizers.$^{23,38,39}$ This approach has decreased the risk of acute toxicity without compromising relapse rate in ALL$^{40}$ or efficacy in IBD.$^{34,41}$ Of note, even at these markedly reduced dosages, erythrocyte TGN concentrations in TPMT poor metabolizers remain well above those tolerated and achieved by the majority of patients (who are TPMT normal metabolizers).$^{39,42}$ Patients with a TPMT/NUDT15 compound intermediate metabolizer phenotype are at increased risk of thiopurine-induced toxicity and require further reduction in starting dose compared to intermediate metabolizers for either enzyme.$^{28}$ In nonmalignant conditions, due to the availability of alternative agents, this guideline recommends alternative nonthiopurine agents for TPMT poor metabolizers and/or NUDT15 poor metabolizers rather than administering reduced doses of mercaptopurine.

Thioguanine therapeutic recommendations. Thioguanine is mainly used for the treatment of malignant diseases, e.g., acute myeloid leukemia and to a lesser extent ALL. The degree of intolerance is expected to be comparable between carriers of TPMT vs. NUDT15 no-function alleles, although there is limited evidence for thioguanine tolerance in NUDT15 variant carriers. The dosing recommendations are similar to those of mercaptopurine (Figure 2). For TPMT and NUDT15 normal metabolizers, the starting dose does not need to be altered (Table 3). In TPMT intermediate metabolizers, the erythrocyte concentration of TGNs is increased.$^{43}$ Starting dose may be adjusted if the protocol dose is considered as high (≥ 40 mg/m$^{2}$)

day); otherwise, dose reduction might not be necessary. TPMT/NUDT15 compound intermediate metabolizers will require further dose reduction compared to single gene intermediate metabolizers, although there are limited data available to confirm the lower tolerance to thioguanine in this group of patients. A full dose of thioguanine exposes TPMT and/or NUDT15 poor metabolizers to fatal toxicity due to extremely high erythrocyte concentration of TGNs and excessive levels of dose-normalized DNA-TG; drastic dose reduction is required to mitigate the risk of thioguanine-induced myelosuppression in this group of patients (Table 3).

Azathioprine therapeutic recommendations. Azathioprine interacts with TPMT and NUDT15 in a similar manner to mercaptopurine. The degree of intolerance is largely comparable between carriers of TPMT vs. NUDT15 no-function alleles.$^{44}$ Azathioprine is mainly prescribed for autoimmune diseases for which the maximal tolerated dose will depend on the clinical indication. For TPMT and NUDT15 normal metabolizers, the starting dose does not need to be altered (Table 4). If starting doses of azathioprine are high (e.g., ≥ 2 mg/kg/day as considered for IBD), lower than standard starting doses should be considered in TPMT and/or NUDT15 intermediate metabolizers. If the starting dose is already low, dose reduction might not be necessary.$^{45}$ TPMT/NUDT15 compound intermediate metabolizers will require further dose reduction compared to single gene intermediate metabolizers, although there are limited data available to confirm the lower tolerance to azathioprine in this group of patients. A high dose of azathioprine exposes TPMT and/or NUDT15 poor metabolizers to extremely high concentrations of TGNs and potentially fatal myelosuppression. Clinicians should consider alternative immunosuppressant therapy in TPMT and/or NUDT15 poor metabolizers rather than a decreased dose of azathioprine.

Pediatrics. Because thiopurines are essential for the treatment of both childhood malignancies (e.g., ALL) and immune suppressive disorders (e.g., IBD, mostly diagnosed in adults), much of the evidence supporting the original dosing recommendations

(summarized in Tables S1 and S2) was derived from studies including pediatric and adult patients. Thus, the guideline dosing recommendations are intended for use in both age groups.

Biogeographic groups. This guideline aims to provide compelling information on ancestry-related differences in the prevalence of no-function phenotypes of TPMT and NUDT15. Initial investigations into TPMT phenotype-genotype correlations predominantly focused on populations of European ancestry, largely due to the higher frequency of well-characterized no-function alleles within this group.$^{9}$ For NUDT15, early studies found that NUDT15's most common no-function allele (i.e., NUDT15 $^{3}$) was predominantly detected in individuals of Asian descent and of Amerindian genetic ancestry.$^{19,20}$ For both genes, the inclusion of individuals from diverse ancestry groups in pharmacogenetic studies has facilitated the detection of rare population-specific variants, for example, TPMT $^{8}$ in patients of African descent or NUDT15 $^{4}$ in patients of Amerindian descent (TPMT Frequency Table and NUDT15 Frequency Table$^{21}$). Nevertheless, given the robustness of the genotype-phenotype associations for both TPMT and NUDT15 across diverse populations, the impact of TPMT and NUDT15 genetic variation on thiopurine exposure and treatment outcomes is likely to be consistent across biogeographic groups. Therefore, the recommendations are not limited to any specific ancestry group.

Recommendations for incidental findings

Recommendations for incidental findings are not applicable.

Other considerations

If test results are available for only one gene (TPMT or NUDT15, but not both), prescribing recommendations based on that gene's results may be implemented, with the caveat that the other gene's results are missing and may have important implications, with up to 10-15% of patients having actionable variants in the non-tested gene. In addition, there may be other reasons underlying poor tolerance to thiopurines that are not related to TPMT or to NUDT15 genetic variation.

Complementary clinical laboratory tests are available that measure thiopurine metabolites in erythrocytes and are used to monitor and adjust therapy: TGNs (for mercaptopurine, azathioprine, and thioguanine) and MeMPNs for those on mercaptopurine or azathioprine (see Supplement for details on associations with TPMT). Clinical assays to measure erythrocyte TGN levels typically do not differentiate among the mono-, di-, and triphosphate forms of active TGNs and do not inform the potential toxicity among individuals with NUDT15 variants.$^{46,47}$ In some patients, elevated TPMT activity leads to excessive MeMPN levels,$^{13}$ increasing the risk of hepatotoxicity while reducing active TGN, although this relationship has not been definitively established$^{48,49}$. The potential impact of NUDT15 variants on MeMPN has not been well-characterized. There is evidence that intermediate and poor metabolizers for NUDT15 accumulate higher levels of DNA-TG in leukocytes than normal metabolizers given the same

thiopurine dosage, highlighting its potential to inform NUDT15 metabolizer status.$^{23}$

Monitoring of erythrocyte TGNs and/or MMPNs can inform clinicians on thiopurine exposure, particularly in case of unexpected clinical response/toxicity or suspected noncompliance, and help further optimize thiopurine dosing. Unlike measuring thiopurine metabolites in erythrocytes, DNA-TG in leukocytes is not yet implemented in clinical practice.

Implementation of this guideline. The guideline supplement contains resources that can be used within electronic health records (EHRs) to assist clinicians in applying genetic information to patient care for the purpose of drug therapy optimization (see Resources to incorporate pharmacogenetics into an electronic health record with clinical decision support sections of supplement).

POTENTIAL BENEFITS AND RISKS FOR THE PATIENT

The benefits of pre-emptive TPMT testing are that doses that are customized based on TPMT phenotype reduce the likelihood of acute myelosuppression without compromising disease control.$^{12,30,31,42}$ The risks would be that a proportion of TPMT intermediate metabolizers may spend a period of time at lower thiopurine doses than they can eventually tolerate, because only ~30–60% of TPMT intermediate metabolizers receiving conventional thiopurine doses experience severe myelosuppression.$^{12,32,42}$ However, because steady state is reached in 2–4 weeks, any period of “under-dosing” should be short, and using this approach, at least in ALL and in IBD, outcomes were not compromised.$^{12,30,31,33,42}$ Furthermore, metabolite monitoring may be used to optimize the thiopurine dose.

Similar benefits are expected with pre-emptive NUDT15 genotyping, especially for populations with a high prevalence of NUDT15 deficiency (such as for patients of Asian or Amerindian genetic ancestry), given that these variants have comparable effects as no-function alleles in TPMT. At least in ALL, leukemia cells with no-function NUDT15 alleles are also more sensitive to mercaptopurine,$^{23}$ and thus, in theory, NUDT15 genotype-guided dosing would not compromise the antileukemic efficacy of this drug. As with TPMT, thiopurine dose adjustments based on the NUDT15 genotype should be followed by close monitoring of tolerance. If the drug is well tolerated, the dose should be gradu-ally increased, especially if the initial dose reduction was signifi-ant (which is the case for intermediate metabolizers with a broad dose reduction range), to avoid under-dosing and ensure treatment efficacy.$^{41,50}$

A possible risk of genotyping is that the patient may have a rare variant that is not included in the test design, which would result in the patient being assigned a “wild-type” (1) genotype by default. Thus, the assigned normal metabolizer phenotype would not reflect the patient’s actual TPMT or NUDT15 activity. Use of a biochemical TPMT activity assay or pharmacologic measures of thiopurine metabolites may mitigate this risk by identifying the patient’s decreased activity, but such tests are not available or informative for NUDT15.

CAVEATS: APPROPRIATE USE AND/OR POTENTIAL MISUSE OF GENETIC TESTS

Most of the time, thiopurines are given orally daily for a period of at least several months. Genotype-based starting doses are just that – starting doses – and in most diseases, titration to the desired degree (or lack thereof) of myelosuppression is required. Other nongenetic variables also contribute to TPMT and NUDT15 enzyme activity and response to thiopurine therapy. Thus, clinicians must continue to evaluate markers of disease progression and/or of myelosuppression to adjust thiopurine doses up or down from the genotyped directed starting doses. One caveat is that secondary cancers have been associated with long-term exposure to thiopurines and eventually TPMT deficiency$^{51,52}$ and to a lesser extent with NUDT15 deficiency$^{53}$; whether capping doses of thiopurines in those with TPMT and/or NUDT15 deficiency will decrease this risk of secondary cancer is not known. In addition, some adverse reactions to thiopurines, such as flu-like syndrome, pancreatitis, and liver toxicity, are not related to TPMT nor NUDT15 deficiency.

SUPPORTING INFORMATION

Supplementary information accompanies this paper on the Clinical Pharmacology & Therapeutics website (www.cpt-journal.com).

ACKNOWLEDGMENTS

We acknowledge the critical input of members of the Clinical Pharmacogenetics Implementation Consortium (CPIC), funded by the National Institutes of Health. CPIC members are listed here: https://cpicpgx.org/members/ and the PharmVar TPMT and NUDT15 expert panels, especially Andrea Gaedigk, Ph.D., and the TPMT/NUDT15 Pharmacogenomics curation expert panels, especially Bailey Tibben, Ph.D.

FUNDING

This work was funded by the National Institutes of Health (NIH) for CPIC (U2H40G17037) and PharmGKB (U2H4G010615) and supported in part by the American Lebanese Syrian Associated Charities (ALSAC). This work was also supported by the NIH National Cancer Institute (R35GM141947 to JYY) and by the Robert Bosch Foundation, Stuttgart, Germany (to MS).

CONFLICT OF INTEREST

The authors declared no competing interests for this work.

DISCLAIMER

Clinical Pharmacogenetics Implementation Consortium (CPIC) guidelines reflect expert consensus based on clinical evidence and peer-reviewed literature available at the time they are written and are intended only to assist clinicians in decision-making, as well as to identify questions for further research. New evidence may have emerged since the time a guideline was submitted for publication. Guidelines are updated periodically on https://www.clinpgx.org/cpic/guidelines and it is the responsibility of the guideline user to consult this website for updates. Guidelines are limited in scope and are not applicable to interventions or diseases not specifically identified. Guidelines do not account for all individual variation among patients and cannot be considered inclusive of all proper methods of care or exclusive of other treatments. It remains the responsibility of the health care provider to determine the best course of treatment for the patient. Adherence to any guideline is voluntary, with the ultimate determination regarding its application to be solely made by the clinician and the patient. CPIC assumes no responsibility for any injury to persons or damage to property related to any use of CPIC's guidelines or for any errors or omissions.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

1. Relling, M.V. et al. Clinical Pharmacogenetics Implementation Consortium Guideline for thiopurine dosing based on TPMT and NUDT15 genotypes: 2018 update. Clin. Pharmacol. Ther. 105, 1095–1105 (2019). 2. Pratt, V.M. et al. TPMT and NUDT15 genotyping recommendations: a joint consensus recommendation of the Association for Molecular Pathology, Clinical Pharmacogenetics Implementation Consortium, College of American Pathologists, Dutch Pharmacogenetics Working Group of the Royal Dutch Pharmacists Association, European Society for Pharmacogenomics and Personalized Therapy, and Pharmacogenomics Knowledgebase. J. Mol. Diagn. 24, 1051–1063 (2022). 3. Feuerstein, J.D., Nguyen, G.C., Kupfer, S.S., Falck-Ytter, Y., Singh, S. & American Gastroenterological Association Institute Clinical Guidelines, C. American Gastroenterological Association Institute Guideline on therapeutic drug monitoring in inflammatory bowel disease. Gastroenterology 153, 827–834 (2017). 4. Lamb, C.A. et al. British Society of Gastroenterology consensus guidelines on the management of inflammatory bowel disease in adults. Gut 68, S1–S106 (2019). 5. National Comprehensive Cancer Network. NCCN Clinical Practice Guidelines in Oncology: Acute Lymphoblastic Leukemia (Version 2.2025). <https://www.nccn.org/professionals/physician_gls/pdf/all.pdf> (2025). 6. Weinshilboum, R.M. & Sladek, S.L. Mercaptopurine pharmacogenetics: monogenic inheritance of erythrocyte thiopurine methyltransferase activity. Am. J. Hum. Genet. 32, 651–662 (1980). 7. Weinshilboum, R. Inheritance and drug response. N. Engl. J. Med. 348, 529–537 (2003). 8. Evans, W.E. Pharmacogenetics of thiopurine S-methyltransferase and thiopurine therapy. Ther. Drug Monit. 26, 186–191 (2004).

11. Liu, C. et al. Gemewide approach validates thiopurine methyltransferase activity is a monogenic pharmacogenomic trait. Clin Pharmacol. Ther. 101, 373–381 (2017).

12. Relling, M.V. et al. Mercaptopurine therapy intolerance and heterozygosity at the thiopurine S-methyltransferase gene locus. J. Natl. Cancer Inst. 191, 2001–2008 (1999).

13. Nygaard, U., Toft, N. & Schmiegelow, K. Methylated metabolites of 6-mercaptopurine are associated with hepatotoxicity. Clin. Pharmacol. Ther. 75, 274–281 (2004).

14. Stanulla, M. et al. Hepatic sinusoidal obstruction syndrome and short-term application of 6-thioguanine in pediatric acute lymphoblastic leukemia. Leukemia 35, 2650–2657 (2021).

15. Toksvang, L.N. et al. DNA-TG and risk of sinusoidal obstruction syndrome in childhood acute lymphoblastic leukemia. Leukemia 36, 555–557 (2022).

16. Evans, W.E. et. Preponderance of thiopurine S-methyltransferase deficiency and heterozygosity among patients intolerant to mercaptopurine or azathioprine. J. Clin. Oncol. 19, 2293–2301 (2001).

17. Franca, R. et al. Impact of mercaptopurine metabolites on disease outcome in the AIEOP-BFM ALL 2009 protocol for acute lymphoblastic leukemia. Clin. Pharmacol. Ther. 114, 1082–1092 (2023).

18. Hartford, C. et al. Differential effects of targeted disruption of thiopurine methyltransferase on mercaptopurine and thioguanine pharmacodynamics. Cancer Res. 67, 4965–4972 (2007).

19. Yang, J.J. et al. Inherited NUDT15 variant is a genetic determinant of mercaptopurine intolerance in children with acute lymphoblastic leukemia. J. Clin. Oncol. 33, 1235–1242 (2015). 20, Yang, S.K. et al. A common missense variant in NUDT15 confers susceptibility to thiopurine-induced leukopenia. N. Naten. 46, 1017–1020 (2014). 21. CPIC. CPIC Guideline for Thiopurines and TPMT and NUDT15.<https://www.clinpgx.org/guideline/PA166251442>. Accessed July 9 2025. 22. Nielsen, S.N. et al. DNA-thioguanine nucleotide concentration and relapse-free survival during maintenance therapy of childhood acute lymphoblastic leukaemia (NOPHO ALL2008): a prospective substudy of a phase 3 trial. Lancet Oncol. 18, 515–524 (2017). 3. Moriyama, T. et al. NUDT15 polymorphisms alter thiopurine metabolism and hematopoietic toxicity. Nat. Genet. 48, 367–373 (2016). 24. Walker, G.J. et al. Association of Genetic Variants in NUDT15 with thiopurine-induced Myelosuppression in patients with inflammatory bowel disease. JAMA 321, 773–785 (2019). 25. Maillard, M. et al. Clinical Actionability of the NUDT15 *4 (p.R139H) allele and its association with Hispanic ethnicity. Clin. Pharmacol. Ther. 317, 724–731 (2025). 26. Tibben, B.M. et al. The Clinical Pharmacogenetics Implementation Consortium’s consensus-based framework for assigning allele function. Am. J. Hum. Genet. 112, 2842–2859 (2025). 27. Toksvang, L.N., Lee, S.H.R., Yang, J.J. & Schmiegelow, K. Maintenance therapy for acute lymphoblastic leukemia: basic science and clinical translations. Leukemia 36, 1749–1758 (2022). 28. Maillard, M. et al. Additive effects of TPMT and NUDT15 on thiopurine toxicity in children with acute lymphoblastic leukemia across multiethnic populations. J. Natl. Cancer Inst. 116, 702–710 (2024). 29. Lennard, L. & Lilleyman, J.S. Individualizing therapy with 6-mercaptopurine and 6-thioguanine related to the thiopurine methyltransferase genetic polymorphism. Ther. Drug Monit. 328–334 (1996). 30. Schmiegelow, K. et al. Thiopurine methyltransferase activity is related to the risk of relapse of childhood acute lymphoblastic leukemia: results from the NOPHO ALL-92 study. Leukemia 23, 557–564 (2009). 31. Schmiegelow, K. et al. Long-term results of NOPHO ALL-92 and ALL-2000 studies of childhood acute lymphoblastic leukemia. Leukemia 24, 345–354 (2010). 32. Stocco, G. et al. Genetic polymorphism of inosine triphosphate pyrophosphatase is a determinant of mercaptopurine metabolism and toxicity during treatment for acute lymphoblastic leukemia. Clin. Pharmacol. Ther. 85, 164–172 (2009). 33. Meggitt, S.J., Gray, J.C. & Reynolds, N.J. Azathioprine dosed by thiopurine methyltransferase activity for moderate-to-severe atopic eczema: a double-blind, randomised controlled trial. Lancet 367, 839–846 (2006). 34. Coenen, M.J. et al. Identification of patients with variants in TPMT and dose reduction reduces hematologic events during Thiopurine treatment of inflammatory bowel disease. Gastroenterology 149, 907–917.e7 (2015). 35. Zhu, Y. et al. Combination of common and novel rare NUDT15 variants improves predictive sensitivity of thiopurine-induced leukopenia in children with acute lymphoblastic leukemia. Haematologica 103, e293–e295 (2018). 36. Ailing, Z., Jing, Y., Jingli, L., Yun, X. & Xiaojian, Z. Further evidence that a variant of the gene NUDT15 may be an important predictor of azathioprine-induced toxicity in Chinese subjects: a case report. J. Clin. Pharm. Ther. 41, 572–574 (2016).

37. Yoshida, M. et al. Low NUDT15 expression levels due to biallelic NUDT15 variants and 6-mercaptopurine intolerance. Br. J. Haematol. 199, 270–276 (2022). 38. Sandborn, W. J. Rational dosing of azathioprine and 6-mercaptopurine. Gut 48, 591–592 (2001). 39. Evans, W.E., Homer, M., Chu, Y.Q., Kalwinsky, D. & Roberts, W.M. Altered mercaptopurine metabolism, toxic effects, and dosage requirement in a thiopurine methyltransferase-deficient child with acute lymphocytic leukemia. J. Pediatr. 119, 985–989 (1991). 40. Relling, M.V., Pui, C.H., Cheng, C. & Evans, W.E. Thiopurine methyltransferase in acute lymphoblastic leukemia. Blood 107, 843–844 (2006). 41. Chao, K. et al. Randomised clinical trial: dose optimising strategy by NUDT15 genotyping reduces leucopenia during thiopurine treatment of Crohn's disease. Aliment. Pharmacol. Ther. 54, 1124–1133 (2021). 42. Ford, L.T. & Berg, J.D. Thiopurine S-methyltransferase (TPMT) assessment prior to starting thiopurine drug treatment; a pharmacogenic test whose time has come. J. Clin. Pathol. 63, 288–295 (2010). 43. Lennard, L., Cartwright, C.S., Wade, R., Richards, S.M. & Vora, A. Thiopurine methyltransferase genotype-phenotype discordance and thiopurine active metabolite formation in childhood acute lymphoblastic leukemia. Br. J. Clin. Pharmacol. 76, 125–136 (2013). 44. Dickson, A.L. et al. TPMT and NUDT15 variants predict discontinuation of azathioprine for Myelotoxicity in patients with inflammatory disease: real-world clinical results. Clin. Pharmacol. Ther. 111, 263–271 (2022). 45. Makuuchi, M. et al. Real-world NUDT15 genotyping and thiopurine treatment optimization in inflammatory bowel disease: a multicenter study. J. Gastroenterol. 59, 468–482 (2024). 46. Lee, J.H. et al. Measurements of 6-thioguanine nucleotide levels with TPMT and NUDT15 genotyping in patients with Crohn's disease. PLoS One 12, e0188925 (2017). 47. Asada, A. et al. NUDT15 R139C-related thiopurine leukocytopenia is mediated by 6-thioguanine nucleotide-independent mechanism in Japanese patients with inflammatory bowel disease. J. Gastroenterol. 51, 22–29 (2016). 48. Jonason, D.E., Sievers, T., Trocke, L., Abraham, J.M. & Vaughn, B.P. Normal ranges of thiopurine methyltransferase activity do not affect thioguanine nucleotide concentrations with azathioprine therapy in inflammatory bowel disease. Crohns Colitis 360 2, otaa058 (2020). 49. van Egmond, R., Chin, P., Zhang, M., Sies, C.W. & Barclay, M.L. High TPMT enzyme activity does not explain drug resistance due to preferential 6-methylmercaptopurine production in patients on thiopurine treatment. Aliment. Pharmacol. Ther. 35, 1181–1189 (2012). 50. Zhou, Y. et al. Individualized use of 6-Mercaptopurine in Chinese children with ALL: a multicenter randomized controlled trial. Clin. Pharmacol. Ther. 115, 213–220 (2024). 51. Stensman, L.M., Kjeldsen, E., Nersting, J., Schmiegelow, K. & Hasle, H. Treatment-related myelodysplastic syndrome in a child with acute myeloid leukemia and TPMT heterozygosity. J. Pediatr. Hematol. Oncol. 37, e242–e244 (2015). 52. Yenson, P.R., Forrest, D., Schmiegelow, K. & Dalal, B.I. Azathioprine-associated acute myeloid leukemia in a patient with Crohn's disease and thiopurine S-methyltransferase deficiency. Am. J. Hematol. 83, 80–83 (2008). 53. Yoshida, M. et al. NUDT15 variants confer high incidence of second malignancies in children with acute lymphoblastic leukemia. Blood Adv. 5, 5420–5428 (2021).