Pharmacogenetic determinants of outcome in acute lymphoblastic leukaemia

Present day paediatric co-operative group acute lymphoblastic leukaemia (ALL) protocols cure approximately 80% of patients, a result achieved largely through the use of risk-stratified therapies that employ multiple chemotherapy agents. These risk-based therapies utilize host and leukaemia traits to select the most appropriate therapy. However, these risk-stratified approaches predict therapy response imperfectly and an important fraction of patients experience relapse or therapy-related toxicity. Pharmacogenetics, the study of genetic variations in drug-processing genes and individual responses to drugs, may enable the improved identification of patients at higher risk for either disease relapse or chemotherapy-associated side effects. While the impact of genetic variation in the thiopurine-S-methyltransferase gene on ALL treatment outcome and toxicity has been extensively studied, the role of other polymorphisms remains less well known. This review summarizes current research on the impact of genetic variation in drug-processing genes in paediatric ALL and reviews important methodological and statistical issues presently challenging the field of pharmacogenetics. Pediatric co-operative groups worldwide cure nearly 80% of children with acute lymphoblastic leukaemia (ALL) (Pui et al, 2001). The cure of ALL involves up to 3 years of rotating combinations of six to 10 essential cytotoxic drugs with different mechanisms of action and in varying doses, schedules and routes (Table I). Successful treatment that began in the 1970s gave rise in the 1980s to a model that predicts the risk of disease recurrence and adjusts the therapy based on one host characteristic, age, multiple disease characteristics including white blood cell count (WBC), karyotype, DNA index, immunophenotype and the complex characteristic of early response to therapy (Sather, 1986; Smith et al, 1996; Felix et al, 2000). Although this model has proved a valid and practical predictor of outcomes of populations of patients, it cannot identify roughly half of the population of non-responding patients, and it fails to detect the individual patient who will experience severe toxicity at drug doses tolerated by the majority (Donadieu et al, 1998). Pharmacogenetics is the study of the inherited variations in drug-processing genes and individual responses to drugs (Weinshilboum, 2003). The pharmacogenetic ‘unit of measure’ may be simple nucleotide polymorphisms (SNPs) or larger blocks of heritable variation across a gene, known as haplotypes. In the prevailing model of ALL where outcomes depend on interactions between host, disease and treatment, pharmacogenetics occupies the area between host and treatment (Fig 1). SNPs in the genes encoding the proteins that either mediate disposition of the essential drugs or their effects can modify risk relapse and toxicity beyond what host and disease predict alone (Relling & Dervieux, 2001). ‘s may also mediate the risk of development of ALL (Sinnett et al, 2000; Krajinovic et al, 2001). Here we review studies that have examined SNPs as determinants of efficacy and toxicity in patients with ALL treated according to recent protocols with standard chemotherapeutic agents. A schematic conceptual framework for these agents and polymorphisms included in this review is presented in Fig 2 and a summary of studies is presented in Table I. We also consider select laboratory, epidemiological and statistical issues that presently challenge the field. Host, disease and therapy interactions. Genetic and other factors influencing chemotherapy efficacy and toxicity. Investigations of thiopurine metabolism offer the first and most compelling example of clinical relevance of pharmacogenetics in paediatric oncology (Fig 3). 6-Mercaptopurine (6-MP) or less often 6-thioguanine (6-TG) forms the backbone of ALL maintenance therapy. Hypoxanthine phosphoribosyltransferase (HPRT) catalyses the first step in the conversion of 6-MP to 6-thioinosine 5′-monophosphate, which in turn is metabolized, to 6-TG nucleotides (TGNs) 6-thioguanasine mono, di and triphosphate, all of which are active, as shown in Fig 3 (Krynetski et al, 1995). Incorporation of TGNs into DNA and RNA is toxic to both normal and malignant cells. Thiopurine-S-methlytransferase (TPMT) on chromosome 6p22.3 catalyses the conversion of active TGNs into substantially less active s-methylated bases. Reduced TPMT activity leads to toxic levels of intracellular TGNs (Chabner & Longo, 2001). 6-Thioguanine and 9-mercaptopurine metabolism. Weinshilboum and Sladek (1980) described the monogenetic inheritance of TPMT activity based on measurement of the concentrations of TGNs in erythrocytes from 298 patients and correctly estimated that one in 300 individuals have homozygous inheritance of two alleles conferring reduced activity and 11% have inherited one such allele. Molecular methods to detect TPMT polymorphisms and TPMT activity were developed by 1997 (Krynetski et al, 1995; Yates et al, 1997). Three of eight known SNPs account for 98% of the clinically relevant abnormal TPMT SNPs. The TMPT*2 allele contains a G238C transversion mutation that reduces S-methylation 100-fold in yeast. The TMPT*3A allele can have two mutations, G460A and A719G, that reduce TPMT protein levels 200-fold in yeast, leaving no detectable activity (Yates et al, 1997). While the ratio of the different mutations varies among ethnic groups, the proportion of individuals with wild type or mutant TPMT is relatively constant (McLeod et al, 2000). A number of studies have proved that, given standard doses of 6-MP, ALL patients homozygous for TMPT mutant alleles experience severe or fatal myelotoxicity and increased relapse because of long delays in therapy (McLeod et al, 2000). Homozygous carriers tolerate 10% or less of protocol dose; reductions are necessary in a minority of heterozygotes (McLeod et al, 2000). Studies from St Jude Children's Research Hospital (SJCRH) show that when patients are treated pharmacologically according to phenotype or genotype, carriers of variant TMPT alleles experience outcomes as good as, or better than, those with wild-type TPMT (Relling et al, 1999a; McLeod et al, 2000). In their studies, the outcome of patients with wild-type TPMT correlates more with dose of thiopurines than with the TGNs. Even with dose reductions, patients with TPMT polymorphisms maintain and tolerate intracellular concentrations of red blood cell (RBC) TGNs that are higher than the median of those with wild-type TPMT, at least from a haematological perspective (McLeod et al, 2000). Unfortunately, those with TPMT polymorphisms have a significantly increased risk of developing secondary brain tumours when treated with cranial irradiation. Six of 52 (12·8%) irradiated patients developed tumours compared with none of 101 patients who received no irradiation (P = 0·0008) (Relling et al, 1999b). Of the six patients with tumours, four had RBC TGN concentrations above the 70th percentile and three had TPMT mutant alleles, so that the excess risk for those with polymorphisms was 4·7-fold (P = 0·0077). Relling et al (1998) reported that patients who developed secondary AML on SJCHR total therapy XIIIHR had decreased levels of TMPT activity compared with patients who did not develop secondary AML. This difference approached, but did not reach, statistical significance, with a P value of 0·01. Although tests for TMPT enzymatic function and genotype are commercially available, the high cost, rarity of homozygous mutation, general awareness concerning TMPT polymorphism and toxicity, and modest specificity of the TPMT genotype information, limit the universal testing in all patients before they receive thiopurines. The standard assay for TMPT function has been measurement of TGNs in erythrocytes (RBC TGNs). Clinical trials with patients receiving oral 6-MP and methotrexate (MTX) in the UK, Canada and Denmark demonstrate that those with lower than median RBC TGN concentrations experience more relapses (Lennard & Lilleyman, 1987; Lennard et al, 1987; Koren et al, 1990; Schmiegelow et al, 1995). In reports from the UK and Denmark, girls experienced a lower relapse rate that correlated with RBC TGNs (Lennard et al, 1987; Lennard & Lilleyman, 1989; Schmiegelow et al, 1995). Because their studies of the early 1990s showed that in addition to sex, concentrations of TGNs and in erythrocytes were significantly with in a study et al, compared with to and 6-MP to receive doses of 6-MP and achieved higher TPMT activity P = but of those pharmacologically experienced relapse compared with of those treated by protocol (P = et al, 2003). In in studies among patients treated with thiopurines and or and pharmacologically by genotype or outcomes correlated more with 6-MP dose than with TGN in both However, those with higher TMPT levels received more 6-MP and to experience more et al, 2001). Relling et al that higher doses or a of the impact of RBC TGNs. also when 6-TG is given in of among children to 6-MP or of those given developed disease et al, 2003). These are and are on the is a for the conversion of Lennard et al the activity of and in of children receiving 6-MP for did not with RBC TGNs. may be (Lennard et al, a role in ALL therapy across co-operative group trials and of metabolism (Chabner & Longo, 2001). and other in the This of the reduced also the and SNPs four of the in methotrexate metabolism have been in relapse or toxicity in ALL patients (Fig and to chromosome and contains that catalyses the conversion of to in the polymorphisms have been into Of these and polymorphisms have been with a decreased activity of and increased levels of et al, 1995; et al, et al, 1998). reports have a of these polymorphisms on leukaemia risk et al, et al, 2001). These polymorphisms in by the variant allele in approximately of and of & 2000). The variant allele is in of and allele are in & 2003). studies have genotype in one study examined et al, a ALL relapse risk et al, and three toxicity risk et al, et al, et al, 2003). et al reported that from six patients with the variant allele had increased in to These that patients with the variant allele have a decreased risk of However, the number of patients the impact of the polymorphism on relapse Krajinovic et al the role of and polymorphisms in patients treated on protocols and These three protocols high risk and patients as by the risk et al, 1996; et al, 2000). The patients of patients treated according to or with these three protocols at the had of in and maintenance therapy et al, 2003). were as leukaemia relapse or from While polymorphism alone was on the an increased risk of with a ratio of the of the with the allele a in with a of These are with the that multiple polymorphisms in the may to treatment response in paediatric However, the are with on in patients with the variant allele. studies in larger will be to the role of polymorphisms in leukaemia relapse Of the three studies of genotype and ALL toxicity, one an increased risk of toxicity et al, and two no et al, et al, 2003). et al described an increased risk of in receiving maintenance ALL therapy with doses from was as and other in reductions or therapy The compared homozygous with a of heterozygotes and wild-type no toxicity was significantly with genotype, or was the of wild type to variant for TPMT, known to modify ALL treatment toxicity risk et al, was on et al the of polymorphism and and In this of patients, no impact of polymorphisms was on either toxicity. In a et al reported the risk of and in patients treated with at The polymorphism was not with either toxicity. However, patients with had higher levels that approached, but did not reach, statistical (P = of no impact of genotype on or patients with the variant allele had increased levels at leukaemia While studies are studies individual with to detect modest risk are The reduced is the of into et al, reports have in to et al, et al, and with reduced is a of methotrexate of are in with more than two of chromosome and in correlates with chromosome number in but not et al, The of in patients with are to their to is in a at chromosome that one two polymorphisms and polymorphisms in the have been Of the polymorphism has been with an risk of et al, and an increased risk of leukaemia relapse et al, et al patients treated with showed no between in the patient for the variant allele as described to and contains that are in a et al, catalyses the conversion of to one polymorphism with has been into Of the other known the has been in paediatric The wild-type allele contains two of a the variant allele contains three The allele has been with decreased protein et al, and has been with a and development et al, et al, the allele be to be with an increased risk of ALL studies have examined the impact of this polymorphism in paediatric Krajinovic et al described an increased risk of relapse with the et al no In the patients described in the Krajinovic et al showed that patients homozygous for the had a in relapse risk = This increased risk when the group was to patients in with years of of therapy = and also et al the polymorphism in patients with relapse and patients in on the ALL and ALL all had either cell or ALL ALL were standard or and received doses with a dose of for four doses, and had years of of therapy. an ratio for relapse of for patients homozygous for the allele. These may be to the of patients and as well as issues in genotype may also a Although the homozygous variant allele was in both of wild type and heterozygotes were significantly P (Table in the wild type and may from allele in or et al, allele reported by Krajinovic et al were in those reported by et al were the allele were the development of paediatric allele in a of patients with ALL not to a decreased number of alleles be et al reported the with a than number of alleles These in genotype the of genotype and While both groups a assay et al, of for or interactions between the allele and and have also been reported et al, Krajinovic et al, 2003). These are in the above and and that also known as or to and contains in a is a the among and et al, four polymorphisms have been into are for the polymorphism with a variant allele of approximately with the studies, Krajinovic et al reported on the role of the and ALL relapse the variant allele was with an increased risk of = Although an increased risk of relapse was not on the of variant allele with the variant allele was with a significantly increased of = Although the use of beyond ALL and are pharmacogenetic studies concerning SNPs in and intracellular metabolism in of the are with in cell from but these are for the most than inherited mutations et al, 2003). The studies in SNPs in to disposition and outcomes in of the of genes through the and and agents in the of children with In an of and genetic polymorphisms of and in disposition of in children with of and the polymorphism was with increased in et al, The that the between genotype and phenotype may by of SNPs in and with disposition and with outcomes in haematological and and cell in ALL et al, et al, et al, 2003). the complex on and on chromosome is in and of and activity of the et al, The polymorphism of the effects protein and and the effects of polymorphism of the more allele and the less Homozygous is with high protein and protein may with the of to et al, In patients with ALL treated on and protocols et al did not of polymorphism with The study was to of patients with ALL and with disease because of response to et al, and polymorphisms were In the group of patients, a of genotype with was not In was with relapse in patients with ALL in but not patients with was with increased risk of relapse in and of was with response in but not with While this study is the of patients and the of the the to the that may risk of relapse and response to et al, et al the by including as above and a A polymorphism at in the first of also the as two or more variant alleles and risk as or one patients treated according to among patients with patients was and among the with The of with factors was (P = In a study of of patients with ALL and et al no difference in of the among patients and and no of with In their of SNPs that response to treatment of ALL Krajinovic et al their that variant had as a result of of activity through of the and toxicity of et al, that this also the of The most gene concerning of host from is the gene a of the of mutation is in cytotoxic chemotherapy and through gene The have been extensively et al, 2001). are at least SNPs at in et al, 2003). studies on SNPs no impact on the and of drugs has et al examined the polymorphism in cell and blood from patients with ALL and no between and and or of the cell or of is that from mutations is more important than The such as and a role in the metabolism of chemotherapy agents to paediatric ALL therapy & Relling & Dervieux, 2001). The also as the an group to a chemotherapeutic This addition is a necessary to the of more and by two reports have the role of in leukaemia relapse risk et al, et al, and three reports have examined the role of these polymorphisms in therapy-related toxicity et al, et al, et al, 2003). Krajinovic et al six polymorphisms in and in patients treated on and protocols between and Of patients experienced an as relapse or from disease or of therapy. was with a of The of at least one allele was with an increased risk of leukaemia = The that is not to the chemotherapy agents on the and Three reports have the role of The four and These genes to chromosome and are in to their alleles in and substantially by the variant alleles and have been in paediatric are to reviews of gene function and polymorphisms et al, et al, et al, 2003). recent studies and in a of patients from the Children's risk ALL et al, et al, 2003). This included patients with relapse and patients in patients had years of of therapy and toxicity were the The that with decreased function be with an increased risk of relapse but a decreased risk of toxicity. In this the and alleles had no impact on relapse However, with the in the and alleles were with a decreased risk of of However, when for multiple this was no other studies have examined the of secondary and et al, et al, Felix et al (1998) reported a = of the allele on secondary leukaemia risk in a group of and secondary leukaemia This ratio was for a as the allele more in than In this of patients ALL patients but of the secondary leukaemia In to Felix et al et al reported no between either or alleles and secondary leukaemia risk in a group of and by also did not show a between the variant alleles and This had approximately to detect the described by Felix et al with variant allele of 10% in the reported variant allele However, the study increased above 80% as the variant allele to the group by such as and the the chemotherapy agents in ALL therapy & Of most research has on the The is of eight and & The and genes are of these and to and The most polymorphisms gene of and as well as two nucleotide polymorphisms in of these in ethnic are to more reviews of gene and polymorphism & et al, 2000; & 2000). one study has a of polymorphisms on relapse risk et al, three other the studies have not a impact et al, et al, Krajinovic et al, et al reported a study of of and patients from the and ALL and were on Homozygous of and were for leukaemia = and = as was the genotype, = an of the risk of relapse with no one genotype, or two or three a of the addition of genotype, P = Three studies with a total of patients, and polymorphism and risk of one of these also The by et al the fraction of these patients, with total patients, of experienced relapse and experienced These patients were from patients on protocols and In this both the and homozygous a risk for relapse of that was not no difference in by either the patients had a lower relapse rate than all patients on the four et al reported on and and in of patients in ALL protocols total total and total in the by et al for had a significantly lower risk of relapse than all between and relapse risk was a that the genotype was with a decreased risk of relapse P = in a study group described in the Krajinovic et al reported no impact of and polymorphisms on ALL relapse the relatively of patients for and it that these polymorphisms not and relapse risk on is given the more modest of interactions between polymorphisms and other as well as of polymorphisms in ALL toxicity, may be are two on polymorphisms in the gene and an increased risk of chemotherapy-associated et al, et al, 2001). The gene is a of the one of two gene and is at 2003). polymorphisms have been in both gene et al, These the polymorphism and the these reports the of the polymorphism in six patients with et al reported that this polymorphism was not in patients on treatment While by these reports on the role of these polymorphisms in toxicity. to and is of six catalyses the two of to and both and in 1995; 2000). two polymorphisms have been into Of these two the in a to at has been with an increased risk of relapse in one study with an increased of relapse risk of et al, studies have an increased risk of therapy-related leukaemia with the variant allele et al, et al, one study has been reported et al, et al a excess of the variant allele in secondary leukaemia patients when compared with allele allele were from reports and a a of allele was not et al reported an increased of for the homozygous variant genotype in a group of patients with secondary AML and patients with AML. patients were from co-operative group AML While were not it is that the patient population was to patients of In the described in the et al patients with secondary leukaemia and patients in ALL therapy. between genotype and secondary leukaemia was reported in by In the for secondary leukaemia was no between and was is one on polymorphisms in cell proteins and one on DNA utilize the patient group at the of et al, 2003). et al reported that patients homozygous variant for the polymorphism had a decreased P = This in was more in patients with the P Krajinovic et al polymorphisms in DNA and of these polymorphisms was significantly with While is general that homozygous and polymorphisms of TMPT to toxicity and relapse in children with is no other polymorphism that is with outcomes among the studies The for among studies treatment methods and Pharmacogenetics is a and the laboratory, epidemiological and statistical of the studies use of of which were years at of often have of the to beyond individual with of DNA are to allele that may result in et al, are to et al, 2001). These issues and the of DNA early in therapy on patients in co-operative group trials will for and The of DNA in clinical trials and of may than genotype of genetic variation in relevant genes et al, 2003). A may a more and of genetic between research on developing across will be to the appropriate statistical and of a issues are also in these the of the recent studies, all studies are and use a of patients with through no of the the population is not of the clinical ALL therapy may in the trials to the study population and between studies treatment may modest and between trials often may result in studies that often than and of all patients on co-operative group trials will but not these the statistical issues in pharmacogenetics the most studies with the of multiple on the appropriate for multiple However, a minority of studies this the research more of this a or may The appropriate of multiple a is to the multiple methods have well described in and interactions the of paediatric pharmacogenetics studies et al, 1996; & 2000). This will more complex as the of DNA on clinical trials more and of are different methods of interactions. These but are not a and et al, et al, & and 2001). While a of genetic variation may more than individual the high number of in a gene may than the statistical day pharmacogenetics studies are to polymorphisms in with known In a ALL proteins are in drug intracellular interactions and (Weinshilboum, 2003). ALL therapy not involves multiple chemotherapeutic agents but also and and patient to these is of an with outcome may be protocol an that is a of relapse in one study may as in these the pharmacogenetics of ALL is one of the most of because of on therapy that is and of We cannot the host and we cannot the Pharmacogenetics may simple to the therapy. was by the and by the in Clinical

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.