Thiopurine Smethyltransferase

Another approach to optimize pharmacotherapy of leukemia with thiopurines is phenotyping and/or genotyping of TPMT. TPMT is a ubiquitously expressed cytoso-lic methyltransferase that catalyzes the S-methylation of aromatic and heterocyclic sulfhydryl compounds (130, 131). Neither the biological function of TPMT nor its endogenous substrates are known. In thiopurine-treated patients, the activity of TPMT determines the balance between 6-TGN and 6-MMPR.

As already described above, high 6-TGN levels in combination with low 6-MMPR suggest low TPMT activity, while low 6-TGN levels in combination with high 6-MMPR suggest high TPMT activity. In 1980, Weinshilboum and Sladek first reported on a trimodal frequency distribution of TPMT activity in 298 randomly selected blood donors (132). In subsequent segregation analysis of selected families, the same authors showed that 66% of the observed total variance in RBC TPMT activity was mono-genetically inherited (133). In Weinshilboum and Sladek's study, 88.6% of the randomly selected study subjects were homozygous for the trait of high TPMT activity (TPMTh/TPMTh), 11.1% were heterozygous (TPMTH/TPMTL), and one out of 298 was homozygous for the trait with low or undetectable TPMT activity (TPMTL/ TPMTl) (132).

This distribution of TPMT activity in Caucasian populations has been subsequently confirmed by other studies (134,135). However, TPMT activity was also shown to differ between different ethnic populations (136). For example, African Americans were demonstrated to have lower enzyme activity in comparison with Americans of Caucasian descent. Other factors that have been variably associated with TPMT activity include age (higher in children, especially neonates), gender (slightly higher in men), smoking (higher in smokers), and thiopurine treatment (137,138,139,140).

The increased TPMT activity during thiopurine administration in children with ALL decreased to levels in the normal range again after treatment was stopped (141). Of importance, three additional issues with regard to TPMT activity and thiopurine therapy in leukemic patients have to be addressed here. First, children with leukemia frequently may require RBC transfusions due to bone marrow insufficiency inherent with the disease. Transfused donor RBCs may bias TPMT activity measurements, and therefore TPMT phenotyping should be performed earliest after 8 weeks upon receiving RBC tranfusions. Second, in patients with leukemia, TPMT activity may also be biased by an older age of RBC. In the course of leukemia, RBC production decreases, leading to relatively older RBC populations with lower TPMT activity. Third, physicians should be aware that some drugs that may be co-administered may have a potential influence on TPMT activity. Such interactions with TPMT activity have been described in ex vivo experiments for aminosalicylates and furosemide (142, 143,144).

TPMT activity can be analyzed by different phenotyping assays including the ra-diochemical method developed by Weinshilboum and more recent non-radioactive HPLC methods using either 6-MP or 6-TG as substrates ( 145, 146, 147, 148, 149,150,151,152,153). The radiochemical and HPLC assays have been shown to lead to comparable results with 6-MP as a substrate (146,147,148,149,150). However, when using 6-TG, TPMT activity was measured at 30% higher levels (152).

6-TG as a substrate offers the advantage of more sensitive and specific quantitation of TPMT activity through measurements of the highly fluorescent TPMT product 6-MTG (140,151). In RBC, TPMT activity correlates well with TPMT activity in lymphoblasts (154). TPMT activity shows large intra- and inter-individual variations. Using an HPLC assay and 6-TG as a substrate, the most comprehensive observation so far on individuals of similar ethnic origin (1214 healthy German Caucasians), identified cut-off values of > 22, 3-22, and < 2nmol/(g-1 hemoglobine x h-1), respectively, to distinguish TPMT wild-type (TPMTH/TPMTH), heterozygous (TPMTH/TPMTL), and homozygous mutant individuals (TPMTL/TPMTL) from each other (140).

Within individuals characterized as TPMTH/TPMTH, a subgroup of ultra-rapid me-tabolizers can be described by a cut-off TPMT activity level of > 50nmol/(g-1 hemoglobine x h-1). However, reported values in the literature may differ depending on the method used for measuring TPMT activity and the population under investigation. In analogy with RBC 6-TGN levels, the comparison of TPMT activity between independent studies is complicated by different units of measurement [e.g., nmol/(mL-1 RBC x h-1) or nmol/(g-1 hemoglobine x h1] (146,147,148,149,150,151).

Weinshilboum and Sladek's observation of TPMT activity being monogenetically inherited was based on phenotypic measurement of TPMT and was subsequently confirmed by cloning and characterizing the gene coding for TPMT (155). The TPMT gene has been localized to chromosome 6p22.3 and encodes a 245 amino acid protein with a predicted molecular mass of 35 kD (155). The first report of the TPMT gene described a 34 Kb gene with 10 exons and a start codon in exon 3 (156). However, in the literature, exon 2 has only been reported once in human cDNA, and two later studies described a smaller-sized gene of 25 kb and 27 kb, respectively, containing only 9 exons (157, 158). To date, 24 mutant alleles (TPMT*2, *3A, *3B, *3C, *3D, *4, *5, *6, *7, *8, *9, *10, *11, *12, *13, *14, *15, *16, *17, *18, *19, *20, *21, *22) responsible for variation in TPMT enzyme activity have been described (Table 1) (159,160,161,162,163,164,165,166,167,168,169,170,171). Most of these alleles are characterized by one or more single nucleotide polymorphisms (SNPs) in the coding sequence of the TPMT gene (non-synonymous SNP), resulting in decrease or loss of enzyme activity. Moreover two variants that influences mRNA splicing (TPMT*4 and *15) are described.

The distribution of TPMT mutant alleles differs significantly among ethnic populations. In the Caucasian population, TPMT*3A, *2, and *3C are the most frequently observed variant alleles which, altogether, account for more than 95% of variant alleles (Table 2) (64,170). In Asian and African populations, TPMT*3C is the most frequent variant allele (171,172). Comprehensive work was done to investigate the functional consequences of TPMT variants and recently alleles including 13 non-synonymous TPMT SNPs were transiently expressed in COS-1 cells and enzyme activity and protein quantity were determined (173).

Previously, laboratory experiments demonstrated that the expression of TPMT*3A, *3B, or *3C transfected into COS-1 or yeast cells results in a decrease of enzyme activity and protein expression (156,160). Loss of activity in these experiments was highest for TPMT*3A, followed by TPMT*3B and *3C. Although there were changes in substrate kinetics, the functional effects resulted primarily from alterations in level of enzyme

Table 1

Phenotypically Relevant Variant TPMT Alleles Reported in the Literature

Table 1

Phenotypically Relevant Variant TPMT Alleles Reported in the Literature

Alleles

Exon

Mutation

Amino acid change

Reference

*1S

7

474T>C

210a

*2

5

238 G>C

Ala80Pro

159

*3A

7

460 G> A

Ala154Thr

156,160

10

719A>G

Tyr240Cys

*3B

7

460 G> A

Ala154Thr

156,160

*3C

10

719A>G

Tyr240Cys

156,160

*3D

5

292 G>T

Glu98X

161

7

460 G> A

Ala154Thr

10

719A>G

Tyr240Cys

*4

G> A transition that disrupts the intron/exon acceptor splice junction at the final 3' nucleotide of intron 9

161

*5

4

146T>C

Leu49Ser

161

*6

8

538A>T

Tyr180Phe

161

*7

10

681T>G

His227Gln

162

*8

10

644 G> A

Arg215His

163

*9

5

356A>C

Lys119Thr

140

*10

7

430 G>C

Gly144Arg

164

*11

6

395 G>A

Cys132Tyr

165

*12

6

374C>T

Ser125Leu

166

*13

3

83A>T

Glu28Val

166

*14

3

1A>G

Met1Val

167

*15

G>A transition that disrupts the intron/exon acceptor splice junction at the final 3' nucleotide of intron 7

167

*16

7

488 G> A

Arg163His

140

*17

3

124C>G

Gln42Glu

140

*18

4

211 G>A

Gly71Arg

140

*19

5

365A>C

Lys122Glu

168

*20

10

712A>G

Lys238Glu

169

*21

4

205C>G

Leu69Val

169

*22

7

488 G>C

Arg163Pro

169

a Frequently observed silent change, reported allele frequency 0.215.

a Frequently observed silent change, reported allele frequency 0.215.

protein. Using pulse chase experiments performed with cultured mammalian cells as well as experiments performed with rabbit reticulocyte lysate, it was demonstrated for example that the common TPMT*3A variant allozyme was degraded much more rapidly than the TPMT wild-type enzyme through a ubiquitin-proteasome-mediated process

Table 2

Population Frequencies of Clinically Relevant TPMT Variant Alleles in Different Ethnic

Groups (64,170,171,172)

Population TPMT Allele Frequencies (%)

Table 2

Population Frequencies of Clinically Relevant TPMT Variant Alleles in Different Ethnic

Groups (64,170,171,172)

Population TPMT Allele Frequencies (%)

*2

*3A

*3C

Caucasian

0.00-0.50

2.50-5.70

0.00-3.80

South American

0.30-2.20

1.50-3.60

0.00-2.54

African

0.00

0.00

7.60-10.10

African American

0.40

0.80

2.40

Asian

0.00

0.00-1.00

0.00-3.00

(174, 175, 176). Moreover, it was shown that chaperone proteins, especially hsp90, are involved in targeting TPMT*3A (176) and very recently the assumption that the TPMT*3A polymorphisms might result in misfolding and protein aggregation with ag-gresome formation was elucidated (177).

Besides the TPMT*2, *3A, *3B, and *3C alleles, most other phenotypically relevant variant alleles have only been reported in single patients. In addition, several intronic mutations and mutations outside of the open reading frame have been described (140,178). Furthermore, a variable number tandem repeat (VNTR) within a GC-rich area in the 5'-flanking region of the TPMT gene has been reported to modulate levels of TPMT activity (179,180). TPMT VNTR length was described to vary between three and nine repeats (VNTR*V3 to *V9) with VNTR*V4 and *V5 being the most frequent of these17-bp or 18-bp repeats. However, neither strong nor consistent associations between the number of tandem repeats and TPMT activity can be demonstrated yet (181,182,183).

TPMT genotyping is complicated by a processed pseudogene located on chromosome 18q21.1 and shows 96% homology to the TPMT coding sequence (184). Because this pseudogene is due to genomic integration of TPMT mRNA upon reverse transcription, it is important to choose primers for TPMT genotyping that extend throughout the exon boundaries (178). TPMT genotyping can be performed by different assays including restriction fragment length polymorphism analysis after polymerase chain reaction (RFLP-PCR), denaturing HPLC, real-time PCR analysis, molecular haplotyping, a multiplex amplification refractory mutation system (ARMS) strategy, arrayed primer extension (APEX), pyrosequencing, DNA microarray, and so on (185,186,187,188,189,190,191,192,193).

The RFLP-PCR method is especially prone to analytical pitfalls due to potentially incomplete restriction enzyme digestion and requires strict quality control measures (194). Of importance, the TPMT*3A allele is conventionally analyzed by separate genotyping of the two involved SNPs, 460 G>A and 719A>G, and therefore it cannot distinguish if the two variant nucleotides are present in cis or trans (on two different alleles). Thus, this approach is inherent with the risk of misclassifying an individual compound heterozygous for TPMT*3B (460 G>A) and TPMT*3C (719A>G), and therefore is deficient in TPMT activity, as heterozygous for a TPMT*3A allele. Although the TPMT*3B allele is rare, misclassification can be overcome by applying techniques for TPMT genotyping that take this problem into account (e.g., molecular haplotyping by long-range genomic PCR to cover the approximately 9 kb difference between the two SNPs and intramolecular ligation) (190).

TPMT genotype can be used as a surrogate marker of TPMT activity. In several independent studies, TPMT genotype showed excellent concordance with TPMT phenotype. For example, Yates and colleagues analyzed the TPMT phenotype in 282 unrelated Caucasian Americans (170). Subsequently, all individuals phenotypically deficient or with intermediate TPMT activity and a randomly selected sample of individuals with high TPMT activity were TPMT genotyped (TPMT*2 and *3A, *3B, *3C).

In this study, 21 patients had a heterozygous TPMT phenotype. With a frequency of 85%, TPMT*3A was the most prevalent variant allele, followed by TPMT*2 and TPMT*3C with about 5% each. All 6 patients who phenotypically displayed TPMT deficiency had two mutant alleles; 20 of the 21 patients with intermediate TPMT activity had one variant allele; and all of the selected 21 patients with high activity did not carry one of the tested TPMT variant alleles. Thus, the major inactivating TPMT variants can be detected reliably by a PCR-based method and demonstrated an excellent concordance with TPMT phenotype.

Coulthard and colleagues analyzed the relationship between TPMT phenotype measured directly in the target of drug action, the leukemic cell, and TPMT genotype (195). They demonstrated that the TPMT activity in lymphoblasts from 38 children and adults homozygous for TPMT*1 was significantly higher than that in the five genotypically heterozygotes. Of interest, in the same study, a comparison of activity in blasts from AML and ALL showed a higher level in AML, and therefore suggests that factors other than genotype may also have an influence on TPMT expression.

The most comprehensive analysis of TPMT phenotype versus genotype published to date was conducted by Schaeffeler et al. (140). In their study, RBC TPMT activity and genotype (TPMT*2 and *3 alleles) were analyzed in 1214 healthy Caucasian blood donors. Discordant cases between phenotype and genotype were systematically sequenced. The frequencies of the mutant alleles were 4.4% for TPMT*3A, 0.4% for TPMT*3C, and 0.2% for TPMT*2. All seven TPMT-deficient subjects identified by Schaeffeler and colleagues were homozygous or compound heterozygous carriers for these alleles.

In 17 individuals with intermediate TPMT activity discordant to TPMT genotype, four novel genetic variants were identified (TPMT*9, *16, *17, and *18) leading to amino acid changes. Taking these additionally discovered variants into consideration, the overall concordance rate between TPMT phenotype and genotype was 98.4%. Specificity, sensitivity, and the positive and negative predictive power of the genotyping test were estimated to be higher than 90%. Thus, solid data support that TPMT phenotype can be predicted by molecular diagnostics. This information is of particular importance to leukemia patients, as phenotypic assessment of TPMT activity in these patients may be rendered unreliable by prior RBC transfusion.

Most studies in leukemia patients have used RBC as a surrogate tissue for phenotyping TPMT activity, and several of them have shown that childhood leukemia patients with TPMTl/TPMTl phenotypes are at high risk of developing severe hematologic toxicity after treatment with standard doses of thiopurines (196,197,198). Thus, dose adjustment at least in TPMT deficient patients is required with an initial dose reduction to 10%-15% of the standard dose of 6-MP as exemplarily shown for ALL patients (199,200) as well as patients with Crohn's disease (201). However, it was also demonstrated that TPMT phenotype or genotype influences the effectiveness of therapy.

Low TPMT activity has been associated with higher 6-TGN levels and improved survival, while high TPMT activity has been associated with lower 6-TGN concentrations and an increased relapse risk (123,127,141). The most comprehensive studies on thiop-urines and their metabolism in the field of childhood ALL have been conducted since the early 1990s at St. Jude Childrens' Research Hospital (67,68,69,79,198,202,203,204,205). As a consequence of their findings on the prognostic importance of 6-MP dose intensity in childhood ALL (e.g., avoidance of thiopurine treatment interruption resulted in fewer relapses), Relling and colleagues at St. Jude used measurements of TPMT activity and thiopurine metabolites, as well as clinical tolerance to maintenance therapy, to guide treatment. They selectively decreased the dose of 6-MP without modification of concurrent chemotherapeutic co-medication in patients with low or intermediate TPMT activity or increased doses in patients demonstrating persistently high absolute leukocyte counts.

Following such a strategy in the St. Jude study, Total XIIIB, TPMT genotype was no longer predictive of hematologic relapse (5-year cumulative incidences of 13.2% vs. 6.7% among patients homozygous for TPMT*1 vs. TPMT genotypes conferring low enzyme activity, respectively) (205). This important finding of Relling and colleagues at St. Jude points to the potential of individualized thiopurine dosing strategies for improving outcome in childhood ALL. In the above-mentioned NOPHO ALL-92 study on treatment of children with precursor B-cell ALL, 6-MP and methotrexate maintenance therapy was adjusted by leukocyte counts, RBC 6-TGN, and methotrexate levels (pharmacology group), or by leukocytes only (control group) (128).

Unexpectedly, girls in the pharmacology group had a significantly increased relapse risk (19% vs. 5% in the control group) because of an increased relapse hazard during the first year off therapy. TPMT activity was the strongest predictor of risk of relapse for girls in the pharmacology group, and girls who relapsed off therapy had higher TPMT activity than those who did not relapse, although this was not the case for girls relapsing on therapy.

Schmiegelow and colleagues speculated that dose escalation of 6-MP may lead to increased intracellular levels of methylated metabolites, such as 6-MMPR, which inhibit purine de novo synthesis in leukemic lymphoblasts of patients with high TPMT activity and therefore, lead to a delay of these cells in S phase under maintenance treatment and potential regrowth after treatment discontinuation.

With relevance to the administration of thiopurines in the early course of childhood ALL, the BFM Study Group reported on the association of TPMT genotype and minimal residual disease (MRD) in 810 children with childhood ALL enrolled into their trial ALL-BFM 2000 (206). In this trial, DNA-based MRD analysis after induction and after consolidation treatment was used for risk-adapted treatment stratification. A 4-week cycle of 6-MP was applied in-between these two MRD measurements. In patients homozygous for the TPMT*1 allele or those heterozygous for a variant TPMT allele, MRD levels on treatment day 33 were equally distributed between the groups. However, when MRD levels were assessed on treatment day 78, after administration of consolidation treatment, including a 4-week cycle of 6-MP at a dose of 60mg/m2/d, significant differences with regard to clearance of MRD were observed between TPMT wild-type and heterozygous patients. For heterozygous patients, this distribution translated into a 2.9-fold reduction in risk of having measurable MRD after induction consolidation treatment (relative risk = 0.34; 95% CI = 0.13 - 0.86; p = 0.02). This point estimate did not significantly change in multivariate analysis including variables known to be associated with treatment response. Besides supporting an important role for 6-MP dose intensity, the BFM data demonstrate a substantial impact of TPMT genotype on MRD after administration of 6-MP in the early course of childhood ALL and may provide a rationale for genotype-based adaptation of 6-MP dosing in the early course of childhood ALL, provided that the described observations translate into an improved long-term outcome.

With regard to long-term toxicity, some reports in the literature have suggested a relationship of secondary malignancies after treatment of childhood ALL with TPMTh/TPMTl and TPMTL/TPMTL phenotypes. In a study at St. Jude Children's Research Hospital, SJCRH Total XIIIHR, patients with lower TPMT activity showed a trend towards a higher incidence of AML associated with application of the topoiso-merase II inhibitor etoposide (203).

A second study conducted at St. Jude, Total Therapy Study XII, reported on a higher incidence of brain tumors in childhood ALL patients with lower TPMT activity who had received cranial radiotherapy concurrent with 6-MP treatment in the initial maintenance phase (204). Six out of 52 patients receiving cranial radiotherapy with concurrent 6-MP treatment developed a brain tumor. Of these 6 patients, 4 had RBC 6-TGN levels above the 70th percentile for the entire cohort of 52 patients, and 3 patients were heterozygous for TPMT.

The 8-year cumulative incidence of brain tumors among children with low TPMT activity was 42.9% (standard error 20.6) versus 8.3% (standard error 4.7) in TPMT wild-type patients. In the Scandinavian NOPHO ALL-92 trial, Thomsen and colleagues reported on a significantly higher risk of therapy-associated AML or myelodysplastic syndrome in patients with low TPMT activity and high erythrocyte 6-TGN levels and/or 6-MMPR (207).

In contrast, successful TPMT genotyping of 72 patients out of a total of 115 patients with subsequent secondary malignant neoplasms after treatment for childhood ALL on seven consecutive BFM protocols (ALL-BFM 79, 81, 83, 86, 90, 95, and 2000) did not reveal a higher frequency of TPMT alleles associated with lower TPMT activity among these patients (208). Also, in stratified analyses by entities of secondary malignant neoplasms, no significant associations with TPMT alleles conferring lower enzyme activity have been observed.

The main contrast between the St. Jude Children's Research Hospital protocol and NOPHO protocols in comparison to BFM protocols for treatment of childhood ALL are that 6-MP starting doses in maintenance are lower in the latter ones (50 vs. 75 mg/m2/d). In addition, on BFM protocols the dose of 6-MP given concurrent with high-dose methotrexate is significantly lower (25 vs. 75 mg/m2/d).

Whereas on the NOPHO ALL-92 protocol patients did not regularly receive cranial radiotherapy, 6-MP application concurrent with cranial irradation during early maintenance was lower on BFM compared to St. Jude protocols (50 vs. 75mg/m2/d). Other differences of potential importance to this issue, which only apply to the comparison of

BFM and St. Jude protocols, are that on BFM protocols no topoisomerase II inhibitors are given in close association with thiopurines and, finally, that on BFM protocols no intrathecal triple therapy (methotrexate, cytarabine, and a glucocorticoid) are given concurrent with cranial radiotherapy and 6-MP. Another important toxicity issue associated with TPMT status relates to the above-described VOD-like symptoms of the liver in childhood ALL patients treated with 6-TG on the British MRC ALL97 trial (209). In this trial, TPMT activity was significantly lower in children in whom VOD developed while no differences in RBC 6-TGN levels were described. This information in association with ongoing research efforts will help to develop a better understanding of 6-TG-associated liver toxicity and may help to identify those individuals upfront who should not be administered 6-TG.

Kicking the Habit

Kicking the Habit

Stop Thinking About How You're Going To Quit Smoking When You Can Instantly Stomp It In Less Than 30 Days With A Proven Set of Techniques. When your

Get My Free Ebook


Post a comment