REVIEWS

CANCER PHARMACOGENETICS: POLYMORPHISMS, PATHWAYS AND BEYOND

Cornelia M. Ulrich*, Kim Robien* and Howard L. McLeod‡ Inherited genetic variations can affect a patient’s response to chemotherapeutic agents given for cancer. Pharmacogenetics aims to use knowledge of these variations to ‘tailor’ therapy for improved response and reduced . Most research so far has focused on single polymorphisms. A more comprehensive approach to predict treatment response will be to consider in entire biological and pharmacological pathways. Of particular relevance to cancer is folate metabolism, which is the target of methotrexate and 5-. Furthermore, efforts have begun to construct pathways of that have pharmacological relevance for individual chemotherapeutic agents. Together, these pathway strategies offer a higher likelihood of achieving the promise of genetically guided cancer therapy.

THERAPEUTIC INDEX With increasingly comprehensive information available results in significant changes in the abil- The ratio of the median lethal from the Human Project, pharmacogenetics ity to metabolize . However, it is possible that the to the median effective dose for a — the science of incorporating information on inherited combination of several polymorphisms in components given . Used to describe genetic variability into predicting treatment response — of a ‘biological’ pathway or ‘pharmacological’ pathway the dose range over which a has a therapeutic effect without is flourishing. A better understanding of pharmacoge- might significantly influence therapeutic response. FOLATE unacceptable toxicity. netic associations is especially important in cancer metabolism has been the target for several key chemo- chemotherapy, as many chemotherapeutic agents given therapeutic agents, and is an example of a biological POLYMORPHISM for cancer are characterized by a narrow THERAPEUTIC pathway with pharmacogenetic relevance2. Variation within a (often at INDEX1. POLYMORPHISMS in both the individual’s genome, as This review will illustrate the relative contribution of a single nucleotide) where two 1 or more alleles exist at a well as the tumour genome, will affect drug response — genetic variation in individual candidate genes within a frequency of at least 1% in the tumours are expected to be of the same genetic makeup biological pathway to patient toxicity and therapeutic general population. with respect to specific polymorphic sites as somatic outcome, and show how genetic variability in both tissue, unless new have occurred or the site is biological and pharmacological pathways can affect treat- subject to chromosomal loss.Yet, drug-related toxicity ment response. In addition, novel strategies for pharma- almost exclusively depends on the genotype of non- cogenetic discovery will be described as we try to move *Fred Hutchinson Cancer tumour tissue1. So, inherited polymorphisms will have a beyond the limitations of our current understanding of Research Center, Cancer key role with respect to toxicity, a crucial dose-limiting the genetic mechanisms that regulate drug activity. Prevention Research factor in most cancer chemotherapy regimens. Program, Seattle, Washington 98109,USA. Research that has used single polymorphisms as Single-gene/single-variant pharmacogenetics ‡Washington University, markers of variation in treatment response has been The classical approach to identifying a genetic basis for Department of , promising, but has lead to some conflicting results. extreme drug toxicity or aberrant drug effect has served St Louis, Missouri Many of the variant forms of drug-metabolizing us well during the initial phase of pharmacogenetic 63130, USA. Correspondence to C.M.U. enzymes show only a small deviation from wild-type research; a clinical phenotype — such as altered phar- e-mail: [email protected] enzyme activity, and are part of polygenic metabolic or macokinetics or extreme toxicity — has been used to doi:10.1038/nrc1233 pharmacological pathways. It is quite rare that a single identify PROBANDS, followed by evaluation of genetic

912 | DECEMBER 2003 | VOLUME 3 www.nature.com/reviews/cancer REVIEWS

Summary However, many of the genetic variants that are associated with extreme drug toxicity are uncom- • The field of pharmacogenetics attempts to use genetic information to predict an mon, and therefore explain only a small proportion individual’s drug response. It is especially important in cancer chemotherapy given the of the population variance that is seen in drug narrow of these drugs. response. In addition, it is recognized that most drug • So far, pharmacogenetic research has largely focused on the effect of single candidate effects are polygenic in nature. Genetic polymor- polymorphisms. However, many of the genetic variants that are associated with phisms have been identified in 93% of all known extreme drug toxicity are rare and explain only a small portion of the variation seen in genes, with two coding-region single-nucleotide drug response. polymorphisms (SNPs) observed in most genes that • Understanding the interactions of genetic variants within a biological or have been evaluated so far17,18. The initial SNP map pharmacological pathway will allow for an improved ability to predict drug response. from the Human Genome Project discovered 1.42 • Folate metabolism — a target of antifolate chemotherapeutic agents and million variants, with the public databases, which are thymidylate-synthase inhibitors — is a biological pathway of substantial interest to growing rapidly, now containing over 5 million pharmacogenetic researchers. human SNPs. Research on the impact of common • Pharmacological pathways are being constructed for the systematic evaluation of the genetic variants on drug response is just beginning, genes that regulate variation in the toxicity and of anticancer agents. and will constitute a considerable research effort over • Mouse models show promise in identifying key enzymes in pharmacogenetic pathways the next years. and will allow study of genetic variation in these pathways. The biological-pathway approach Most enzymes function in complex networks with sev- eral regulatory mechanisms. So, it is unlikely that any polymorphisms, for example in the drug-metabolizing one variant with modest effects on enzyme function will enzyme that is responsible for the degradation of the affect disease or treatment outcomes, whereas the com- drug. These genetic polymorphisms include nucleotide bination of several variants within the same pathway repeats, deletions, insertions and mutations that influence might result in significant disturbances. Alternatively, gene expression and/or function3. several genetic variants might cancel each other out and Genetic variants within a specific candidate gene diminish differences in the response that is associated provide the mechanistic basis for many of the early with a single polymorphic allele. An example of a bio- examples in pharmacogenetics. For instance, muta- logical pathway with relevance to cancer chemotherapy tions or deletions in the CYTOCHROME P450 enzyme is folate metabolism. CYP2D6 occur in the general population and about 10% have been associated with a poor-metabolizer Folate metabolism and antifolates. Folate acts as a phenotype4. This is important for codeine and related donor for methyl groups, such as in the conversion of pain , as the activation of codeine to homocysteine to methionine — a precursor for the morphine is dependent on the catalytic function of universal methyl donor S-adenosyl-methionine — CYP2D6 (REF. 5). Patients with low CYP2D6 activity and in the synthesis of purines and pyrimidines19

FOLATE do not receive pain relief from codeine or other ana- (FIG. 1). Therefore, folate has a key role in normal cell 5 One of the B vitamins. The logues, because they are not able to form morphine . growth and replication and has been an attractive primary role of this B vitamin is As another example, polymorphisms in thiopurine target for chemotherapeutic agents. Antifolate agents as a carrier of methyl groups, methyltransferase and dihydropyrimidine dehydro- are drugs that target specific enzymes in folate especially for purine, pyrimidine genase have been associated with altered (for example, dihydrofolate reductase) and methionine synthesis. metabolism and increased risk of severe toxicity and are used for the treatment of various cancers, PROBAND from the anticancer agents 6-mercaptopurine and including haematological, colorectal, breast and An individual with the condition 5-fluorouracil, respectively 6–10. In addition, a variant pancreatic malignancies. of interest, who serves as the number of dinucleotide-repeat sequences (5–8 Methotrexate and 5-fluorouracil are folate-pathway starting point for exploration of ′ a family pedigree for the genes repeats) in the promoter for uridine 5 -diphosphate- inhibitors that have been available for 50 years and con- that are responsible for the glucuronosyltransferase 1A1 (UGT1A1) has an influ- tinue to be used in various treatment regimens for cancer condition. ence on in vitro and in vivo glucuronidation of and autoimmune diseases. Their specific mechanisms of SN-38, the active metabolite of irinotecan11. Patients action have been reviewed previously20,21. Folate ana- CYTOCHROME-P450 ENZYMES with seven repeat sequences have a fourfold relative logues are structurally similar to folate but are able to A family of haem-containing intracellular oxidizing enzymes risk of experiencing severe toxicity after treatment inhibit the action of various enzymes in folate metabo- that are responsible for the first with irinotecan, including grade III/IV diarrhoea and lism (in the case of methotrexate, primarily dihydrofolate phase of metabolism of many neutropaenia, compared with patients with six repeat reductase)21–24. 5-Fluorouracil is a fluoropyrimidine that, drugs and other ingested . sequences12,13. These examples of altered drug metab- on activation to the nucleotide form, develops a stable

THYMIDYLATE-SYNTHASE olism from genetic polymorphisms have recently been complex with thymidylate synthase, which inhibits the 20,25  INHIBITORS complemented with studies that associate tumour activity of the enzyme . Capecitabine (Xeloda ) is an A category of chemotherapeutic response or patient survival with polymorphisms in orally administered fluoropyrimidine that is subsequently agents, including 5-fluorouracil, membrane transporters (such as ABCB1 in patients activated to fluorouracil26. which tightly bind and inhibit with acute myelogenous leukaemia14) and excision The therapeutic effectiveness and cytotoxicity of the activity of thymidylate synthase, and therefore prevent repair enzymes (such as ERCC2 or XRCC1 in patients these drugs are largely attributed to the role of folate DNA replication. with colorectal cancer15,16). in nucleotide synthesis — both folate antagonists and

NATURE REVIEWS | CANCER VOLUME 3 | DECEMBER 2003 | 913 REVIEWS

Regulatory enzyme C677T: TT genotype associated with greater toxicity of MTX and possibly 5-FU A1298C: Common polymorphism Cell membrane affecting protein function

Pyrimidine synthesis MTHFR 5-FU dUMP

hFR dTMP Serum 5, 10-Methylene THF Thymidylate folate synthase 5-Methyl THF

SHMT RFC Drug target TSER polymorphism: the 2rpt/2rpt THF Drug transporter genotype responds better to 5-FU treatment, but with greater toxicity. G80A: AA genotype associated Possible association with survival with higher MTX serum levels, 10-Formyl THF among MTX-treated patients and worse prognosis among Methionine MTX-treated children synthase

Homocysteine GART DHF AICARFT GAR Methionine AICAR CBS Purine synthesis

Cystathionine SAH SAM DHFR

Cysteine CH3XX

DNA methylation Methotrexate

Figure 1 | Folate metabolism and related pathways. This simplified figure illustrates the interconnectedness of folate metabolism and proteins for which functional polymorphisms have been identified. Polymorphisms have been found that are associated with pharmacogenetic outcomes in three key proteins in these pathways: the drug transporter protein reduced folate carrier (RFC); the regulatory enzyme 5,10-methylenetetrahydrofolate reductase (MTHFR); and the drug target thymidylate synthase. Key enzymes are denoted as ovals, substrates as rectangles. Red ovals denote enzymes with genetic polymorphisms that have been investigated in pharmacogenetic studies. Orange ovals denote enzymes for which functional genetic polymorphisms have been described. 5-FU, 5-fluorouracil; AICAR, 5-aminoimidazole-4-carboxamine ribonucleotide; AICARFT, AICAR formyltransferase; CBS, cystathionine-β-synthase; DHF, dihydrofolate; DHFR, DHF reductase; dTMP, deoxythymidine monophosphate; dUMP, deoxyuridine monophosphate; GAR, glycinamide ribonucleotide; GART, phosphoribosylglycinamide formyltransferase; hFR, human folate ; MTX, methotrexate; SAH, S-adenosylhomocysteine; SAM, S-adenosylmethionine; SHMT, serine hydroxymethyltransferase; THF, tetrahydrofolate; X, various substrates for methylation.

MUCOSITIS Inflammation, irritation and ulceration of the mucosal membranes, especially of the THYMIDYLATE-SYNTHASE INHIBITORS reduce proliferation of Folate metabolism and genetic variants. Recently, it has oral cavity and gastrointestinal neoplastic cells. In situations of folate deficiency, become evident that there is large inter-individual tract, that occurs as a result of insufficient conversion of deoxyuridine monophos- variability in the mechanisms to maintain homeostasis cytotoxic chemotherapy or phate to deoxythymidine monophosphate can result within folate metabolism because of genetic polymor- radiation therapy. in uracil misincorporation into DNA strands during phisms in key enzymes (reviewed in REF. 2). Genetic

MYELOSUPPRESSION replication, subsequently resulting in double-strand variants in thymidylate synthase, 5,10-methylenete- Suppression of blood-cell breaks27–30. Activated metabolites of fluoropyrimidines trahydrofolate reductase (MTHFR) and the reduced formation in the bone marrow. inhibit the activity of thymidylate synthase and are folate carrier (RFC) have been investigated with also incorporated into both DNA and RNA, which respect to clinical outcomes (FIG. 1; TABLE 1). TRANSCRIPTIONAL-ENHANCER ′ ELEMENT results in further DNA damage and interference with The 5 -untranslated region of the thymidylate- 20 A region of DNA that might be RNA synthesis and function . Since folate is needed synthase gene functions as a cis-acting TRANSCRIPTIONAL- several thousand base pairs for DNA synthesis and repair, rapidly dividing ENHANCER ELEMENT and contains polymorphic 28-base- upstream from a given gene’s normal tissues such as the gastrointestinal tract and pair tandem repeats33,34. The presence of a triple repeat promoter region, but can bind haematopoietic cells are the main sites of toxicity, with (3rpt) results in approximately 2–4-fold greater gene with gene-regulatory proteins 35,36 and increase the rate of common side effects including oral and intestinal expression compared with a double repeat (2rpt) . transcription of that gene. MUCOSITIS and MYELOSUPPRESSION20,21,31,32. The triple-repeat homozygote (3rpt/3rpt) genotype is

914 | DECEMBER 2003 | VOLUME 3 www.nature.com/reviews/cancer REVIEWS

associated with lower plasma folate concentrations 3rpt allele might protect against the development of and, among individuals with low dietary folate intake, acute lymphocytic leukaemia38, and is associated with increased plasma homocysteine concentrations37.The a reduced risk of colorectal adenomas in the presence

Table 1| Polymorphisms in folate metabolism and response to cancer drugs Polymorphism Study population Findings References TSER 28-base-pair repeats Tumour tissue from 50 patients with 2rpt/2rpt individuals (n = 11) had 36 metastatic colorectal cancer receiving significantly greater toxicity 5-fluorouracil 65 patients with rectal cancer 3rpt/3rpt individuals (n = 27) had a 77 receiving pre-operative 5-fluorouracil lower probability of tumour downstaging compared with 2rpt/2rpt individuals (n = 13) Tumour tissue from 221 patients with No association between TSER polymorphism 59 Dukes’ C colorectal cancer receiving and survival. Similar effects of 5-fluorouracil either surgery alone or surgery and treatment seen among each of the TSER 5-fluorouracil genotypes 24 patients with metastatic colorectal 3rpt/3rpt homozygous individuals had 58 cancer receiving 5-fluorouracil decreased median survival 205 children with acute lymphoblastic 3rpt/3rpt homozygous children (n = 50) had 68 leukaemia treated with methotrexate significantly shorter event-free survival times compared with those with the other TSER genotypes 24 patients with metastatic colorectal Individuals with the 2rpt/2rpt genotype (n = 4) 78 cancer receiving capecitabine had a greater response rate compared with individuals with the 2rpt/3rpt (n = 12) or homozygous 3rpt/3rpt genotypes (n = 8) MTHFR C677T 51 patients with stage III colon No differences in survival based on genotype 61 cancer receiving 5-fluorouracil and 677TT (n = 4) leucovorin 6 patients with early breast cancer 5 of the 6 patients who experienced grade IV 66 receiving cyclophosphamide, toxicity (NCI-CTC) had the 677TT genotype methotrexate and 5-fluorouracil 220 patients with CML following Patients with 677TT genotype (n = 36) 62 haematopoietic-cell transplant receiving methotrexate experienced increased oral mucositis and somewhat slower platelet recovery 61 patients with acute leukaemia Increased methotrexate toxicity among 64 receiving maintenance chemotherapy patients with the 677TT genotype (n = 15) including methotrexate 6 children with acute lymphoblastic Lymphocytes from the two children with the 67 leukemia who had had methotrexate- 677TT genotype (n = 2) exhibited greater related in vitro methotrexate sensitivity compared with those with the 677CC (n = 2) or 677CT genotypes (n = 2) 43 patients with ovarian cancer 677TT individuals (n = 13) had a significantly 65 receiving either methotrexate alone higher risk of experiencing grades III–IV toxicity or in combination with carboplatin (WHO criteria), and had significantly higher plasma homocysteine levels after methotrexate treatment 43 patients with metastatic colorectal Individuals with the 677TT genotype (n = 17) 70 cancer receiving fluoropyrimidine were significantly more likely to respond to chemotherapy treatment than with those with the 677CT (n = 21) or 677CC genotypes (n = 5) Reduced folate carrier G80A 204 children with acute lymphoblastic Patients with the 80AA genotype (n = 45) 69 leukaemia treated with methotrexate had significantly shorter event-free survival compared with those with the 80GG genotype (n = 61); patients with the 80AA genotype had higher plasma methotrexate levels than those with other genotypes, suggesting decreased cellular uptake of methotrexate CML, chronic myeloid leukaemia; MTHFR, 5,10-methylenetetrahydrofolate reductase; NCI-CTC, National Cancer Institute Common Toxicity Criteria; rpt, repeat; TSER, thymidylate-synthase enhancer region; WHO, World Health Organization.

NATURE REVIEWS | CANCER VOLUME 3 | DECEMBER 2003 | 915 REVIEWS

of high folate intake or a MTHFR 677TT genotype the distributions that were seen in other patients at the (see below)39. A second common polymorphism in centre where the study was carried out or to those seen thymidylate synthase — a 6bp deletion at base pair in other studies36,57. 1494 in the 3′-untranslated region40 — seems to be Our group studied 24 patients who were treated with associated with decreased messenger RNA levels of the 5-fluorouracil and leucovorin chemotherapy for metasta- enzyme in colorectal tumours41. tic colorectal cancer as part of a pilot study 58. Tumour MTHFR is a central regulatory enzyme in folate response, as measured after every four cycles of metabolism that diverts folate metabolites from chemotherapy by computed tomography, was more likely pyrimidine synthesis towards methionine synthesis. A among those who had the TSER 2rpt/2rpt genotype. common functional polymorphism of the gene encod- Further investigations that are based on this pilot study ing this enzyme occurs at C677T (alanine222valine)42. are underway. The 677TT genotype produces an enzyme with only Another study compared Australian patients with 30% of the activity of the wild-type (677CC) enzyme colorectal cancer who did not receive chemotherapy in vitro and is associated with increased plasma con- with patients treated with a cycle of 5-fluorouracil- centrations of homocysteine. Several diseases have been based adjuvant chemotherapy once a month for six associated with the MTHFR C677T polymorphism, months 59. The survival benefits that were associated including acute lymphocytic leukaemia43,44, colorectal with 5-fluorouracil-based chemotherapy were similar polyps and colon cancer45–49, neural-tube defects50,51 among patients with the TSER 3rpt/3rpt genotype and, possibly, cardiovascular disease52; increased risks (HAZARD RATIO (HR) of the effect of chemotherapy on are usually seen among individuals with a low folate survival = 0.62) compared with those with the 2rpt/2rpt status. A second common polymorphism in the or 3rpt/2rpt genotype (HR = 0.52) The analyses did not MTHFR gene (occuring at A1298C; Glut429Ala)53,54 account for other factors that might affect survival, such also results in somewhat decreased MTHFR activity as the age of the patients or the treatment regimens that and has been associated with a reduced risk of certain were used. forms of acute adult and paediatric leukaemias43,44 and Overall, findings regarding the TSER variant indicate colon cancer55. that genotypes that are associated with lower thymidy- Furthermore, a polymorphism in the RFC gene late-synthase expression confer an increased sensitivity (G80A, argenine27histidine) seems to be associated with towards 5-fluorouracil therapy, although possibly at the a higher affinity for folate56. Among 169 healthy individ- expense of greater toxicity. However, most of these stud- uals who were stratified by MTHFR C677T genotype, ies had limited sample sizes, which needs to be taken the variant A allele in RFC was consistently and linearly into consideration when interpreting the results. associated with higher plasma folate concentrations56. Another commonly investigated genetic variant is the C677T polymorphism in MTHFR (discussed Response to antifolates and thymidylate-synthase above), which alters the activity of the key regulatory inhibitors. The variability in receptors, drug targets MTHFR enzyme in folate metabolism60. A study among and other key regulatory enzymes within folate 51 patients with stage III colon cancer who were treated metabolism that is described above seems to trigger a with 5-fluorouracil did not show a difference in survival differential response to antifolate chemotherapeutic by MTHFR genotype61. However, the study included agents and thymidylate-synthase inhibitors. Several only four individuals with the MTHFR 677TT geno- recent studies show how polymorphisms in the folate types and lacked adjustment for other factors that are pathway affect the treatment response to drugs that potentially related to survival. target this biological pathway (TABLE 1). Our group investigated the toxicity of methotrexate LOSS OF HETEROZYGOSITY The thymidylate-synthase promoter-enhancer in 220 patients with chronic myelogenous leukaemia The loss of one allele in a region (TSER) polymorphism has been the most com- who were undergoing haematopoietic-cell transplan- tumour cell for a gene in which monly investigated, usually in studies that evaluate tation62. These patients all received methotrexate the individual is normally heterozygous. response to or toxicity from 5-fluorouracil treatment. as part of a standardized regimen with cyclosporine Among 50 patients with metastatic colorectal cancer for the prevention of GRAFT-VERSUS-HOST DISEASE63.The HAZARD RATIO who underwent therapy with 5-fluorouracil as a com- MTHFR 677TT genotype — which corresponds to The comparison of the risk of ponent of several different regimens, patients who lower enzyme activity — was associated with 36% experiencing an event at any were homozygous for the TSER-double-repeat geno- higher oral mucositis scores compared with the 677CC given point in time (hazard) among those with a particular type (2rpt/2rpt; associated with lower thymidylate- genotype; patients with 677CT showed intermediately risk factor with those without synthase expression) were more likely to respond to increased mucositis, consistent with the residual the risk factor. 5-fluorouracil therapy, but were also more likely to in vitro enzyme activity of the heterozygous genotype. experience significant toxicity 36. It is of concern that Furthermore, recovery of platelet counts — a sign of GRAFT-VERSUS-HOST DISEASE A potentially life-threatening genotyping in this study was undertaken with DNA haematopoietic-cell engraftment after transplantation condition, which might occur that had been collected from tumour tissue rather than — was 20–40% slower in patients with the MTHFR following transplantation of from somatic cells, which are the sites for drug toxicity. 677TT genotype62. solid organs or haematopoietic LOSS OF HETEROZYGOSITY can occur in tumour cells on Two studies from Italy support the concept of cells, in which donor chromosome 18 — where TSER is found — and the increased methotrexate toxicity among patients with a T lymphocytes recognize host cells as foreign and attack TSER-genotype frequencies among the patients who MTHFR 677TT genotype. Among 61 patients with acute host tissues. were evaluated for clinical outcomes were dissimilar to leukaemia who received maintenance chemotherapy

916 | DECEMBER 2003 | VOLUME 3 www.nature.com/reviews/cancer REVIEWS

The interconnectedness of folate metabolism and the CPT-11 need to consider the entire pathway rather than individ- 3435C>T associated ual polymorphisms is further illustrated by three recent with higher drug investigations68–70. The findings of these studies strongly indicate that polymorphisms in proteins that are not direct drug targets of the chemotherapeutic agent used — such as the MTHFR protein — can affect treatment response to other drugs that target folate metabolism. Cohen et al. reported that the MTHFR genotype CPT-11 ABCB1 seems to alter response to 5-fluorouracil therapy; CYP3A4 APC individuals with the MTHFR 677TT genotype (n = 5) CPT-11 were significantly more likely to respond to treatment CYP3A5 70 CES1 NPC than those with 677CT or 677CC genotypes . Furthermore, the TSER genotype seems to affect CES2 CES1 response among children who are treated with 68 UGT1A1*28 allele methotrexate . Among 205 children with acute lym- CES2 associated with fourfold phoblastic leukaemia who were treated with risk of severe toxicity methotrexate induction and maintenance chemother- SN-38 apy, the risk of having ‘an event’ (such as relapse or UGT1A1 death due to disease) varied by TSER genotype. SN-38 SN-38G Compared with children with a TSER 2rpt/2rpt geno- type, those with a 3rpt/3rpt genotype were at increased risk (multivariate ODDS RATIO (OR) = 4.8). ABCB1 Within the same patient population, children with the RFC 80AA genotype had a significantly shorter event- TOPI ABCG2 ABCC1 free survival than those with the 80GG genotype, indicating that the cellular uptake of methotrexate ADPRT 69 ABCC2 XRCC1 might have a role in the response . These studies point towards the need to consider the entire folate κ TDP1 NF B1 metabolism as a biological pathway when evaluating associations between genetic polymorphisms and CDC45L Cell death treatment outcomes, rather then the current focus on SN-38 individual components in isolation.

Figure 2 | Possible pathways of irinotecan metabolism. Irinotecan (CPT-11) can be The pharmacological-pathway approach converted into the active metabolite SN-38 by carboxylesterases (CES) outside or inside the cell. A second approach to understanding pathways that CPT-11 and SN-38 are both substrates for the ATP-binding cassette (ABC) transport proteins — are key to pharmacogenetics research involves the P-glycoprotein (ABCB), ABCC and ABCG — which transport the drug out of the cell. Alternatively, construction of pharmacological pathways. It is clear CPT-11 and SN-38 can be inactivated by cytochrome P450 enzymes (CYP) or uridine that although the single-candidate-gene approach has diphosphate glycosyltransferase (UGT), respectively. If SN-38 persists, it binds to its target topoisomerase I (TOPI), interfering with DNA synthesis and repair processes, culminating in cell provided the foundation for cancer pharmacogenet- death. ADPRT, ADP-ribosyltransferase; APC, inactive metabolite of SN-38; CDC45L, cell-division ics, evaluation of a gene in isolation as part of its cycle 45L; NPC, inactive metabolite of SN-38; SN-38G, SN-38 glucuronide; TDP, tyrosyl-DNA normal biological pathway does not often provide suf- phosphodiesterase; XRCC1, X-ray-repair cross-complementing defective-1. ficient information about variation in expected thera- peutic response to justify dose modification. A logical next step is the construction of pathways of genes that

ODDS RATIO including methotrexate, increased methotrexate intol- are specifically of importance for the regulation of A comparison of the risk of erance was observed among patients with the 677TT drug activity, with the use of a composite view of experiencing a particular genotype64. Similarly, patients with ovarian cancer and these pathways as a tool for individualized therapy. outcome (such as a disease or the 677TT genotype who received methotrexate expe- For drugs that mimic natural molecules, such as the adverse drug reaction) between rienced significantly higher plasma homocysteine levels antifolates, antipyrimidines, anti-oestrogens and oth- two groups of people – one group with a particular risk than those with other genotypes, as well as greater ers, there are pre-existing biological pathways that factor (such as a genotype or a toxicity 65. Lastly, two case reports66,67 provide further define the disposition and activity of the agents. nutrient), and the other group indication for greater methotrexate sensitivity and, However, many anticancer agents do not follow an without the risk factor being possibly, significant toxic responses among those with existing biological pathway, but, rather, interact with a studied. a MTHFR 677TT genotype. pharmacological pathway that consists of a diverse TOPOISOMERASE I Most of these studies are based on small sample array of genes that would not have direct interplay in An enzyme that creates a sizes, which precludes robust conclusions and might normal . transient single-strand break in result in publication bias. Nevertheless, the consistent The construction of these pharmacological pathways the phosphodiester backbone of findings of these studies regarding a link between entails the use of knowledge regarding drug absorption, DNA, allowing the strand of DNA to uncoil and rotate freely decreased MTHFR activity (677TT genotype) and , activation and other metabolic functions. during replication. increased toxicity are compelling. This is illustrated in FIG. 2, for the TOPOISOMERASE I

NATURE REVIEWS | CANCER VOLUME 3 | DECEMBER 2003 | 917 REVIEWS

Table 2 | Various approaches to analyzing pharmacogenetic variation Approach Advantages Limitations Candidate polymorphism • Quick and relatively inexpensive • Only helpful in the clinical setting if the analysis • Uses knowledge regarding drug polymorphic genotype corresponds to a and distinct phenotype (high sensitivity/specificity) • Assumes extensive previous knowledge • Hypothesis driven regarding what gene is likely to be important — such as one that determines the excreted proportion of a drug • Very limited approach Pathway approach • Accounts for associations between • Requires a solid biochemical understanding proteins in a metabolic pathway — pharmacological or biological — of the • Uses knowledge regarding drug respective pathway metabolism and other meaningful • Need to understand how the drug is biological associations (such as metabolized and its mechanisms of action regulatory function and homeostasis) • Data analysis is more complex than evaluation • More likely to explain inter-individual • Requires large study sizes variation in drug response • Hypothesis driven Genomics or proteomics • Does not require a hypothesis • Does not use information on known biological (microarray technology) • Provides a complete gene- or protein- or pharmacological associations expression profile (tumour or individual) • Optimal data management and data analysis • Provides information on associations techniques are not well defined at this point that have not previously been suspected • Expensive (expression arrays), not fully • Provides a large amount of data developed (proteomics) or not currently • Probably useful for prediction of practical in a clinical setting tumour response • Not clear how useful for prediction of toxicity

inhibitor irinotecan. This figure shows the complexity of definitive. One under-used tool is mouse , in factors that influence an individual drug, with a need for which ex vivo and in vivo strategies can be performed activation of irinotecan to SN-38; transporters causing using data analysis approaches to discover genes that efflux of both parent and metabolite; P450-mediated are associated with drug activity71. The laboratory inactivation of parent drug; glucuronidation of the mouse and humans share a high degree of similarity in active metabolite; variation in the cellular therapeutic anatomy, and genome content. By using target; and then a series of death genes that mediate the well-defined inbred mouse strains, breeding strategies ultimate fate of the cell. It is no surprise that this level have been used to identify regions of the genome that of complexity dilutes the contribution of any one are associated with a particular phenotype — such as component of the pathway. drug effect — to localize the gene that contributes to The key issue now is the comprehensive evaluation the phenotypic variation and to discover the genetic of pharmacological pathways to identify the compo- variants that provide the mechanistic basis for the nents that are rate-limiting to toxicity and antitumour effect. A recent example of the power of mouse effect. This will allow more focused polygenic studies to genomics in the context of cancer therapy was per- be conducted in clinical trials to provide validation of formed to identify the gene(s) that are associated with their worth in clinical decision-making. susceptibility to pulmonary fibrosis after bleomycin chemotherapy. By conducting a genome-wide scan in Pharmacogenetic-discovery strategies offspring that were produced from breeding a ‘sensi- The recognition that drug effect is under the control tive’ mouse strain with a ‘resistant’ strain, two regions of several genes or defined drug pathways has led to a of the genome were associated with bleomycin-related need for pharmacogenetic-discovery experiments lung fibrosis72. The first locus was on mouse chromo- without the constraints of predetermined candidate some 17 and is a region that has been associated with genes. Pharmacogenetic discovery in the field of can- pulmonary fibrosis from other causes (for example, cer chemotherapy is severely limited by the inability radiation, ozone and particle exposure) and contains to perform classical human genetic screens. The use the candidate gene tumour-necrosis factor-α. The of familial genetic strategies to identify putative genes second locus on mouse chromosome 11 contains for predicting toxicity or drug response is not possible, bleomycin hydrolase, an enzyme that detoxifies because members of a given family do not acquire the bleomycin, and a glycoprotein Sparc, which showed same malignancy in the same window of time. It is also expression differences between susceptible and resis- unethical to give cytotoxic chemotherapy to normal tant mouse strains72. Lessons learned from the past 25 volunteer individuals, which prevents the acquisition years of mouse genetics have provided a rich resource of large data sets of unrelated individuals on which to of genomic information from the commonly available perform genome-wide screens. inbred strains and the knowledge of statistical genetics Microarray studies will reveal some of the genes to rapidly apply this approach to pharmacogenetic that are associated with drug effect, but these studies questions. These pharmacogenetic-discovery strategies often include an insufficient number of samples to be can be used to explore genetic variability in relevant

918 | DECEMBER 2003 | VOLUME 3 www.nature.com/reviews/cancer REVIEWS

pathways. Concerted efforts are ongoing to take approaches’, which use microarray technology to advantage of the power of mouse genetics, including describe gene or protein expression in its entirety will knockout and transgenic techniques, to identify phar- certainly be useful in refining cancer diagnoses and, in macological pathways of relevance to the prediction of turn, predicting tumour response to specific drug effect in individuals with cancer. drugs73–76. However, these methods might be less help- ful for predicting potential toxicity, since gene and Outlook protein expression in susceptible tissues are likely to Considering the narrow therapeutic index for many differ once a drug is administered. Furthermore, cancer chemotherapeutic agents, the ability to better genomic or proteomic methods do not take variations predict response and potentially life-threatening toxic in protein function or stability into account, which, in effects is highly important to the clinician. We expect the case of folate metabolism, have been shown to be that, in the future, pharmacogenetic research using associated with clinical outcome. the pharmacological-pathway approach and research Knowledge regarding existing pathways, both bio- using the genomic or proteomic approaches will logical and pharmacological, has been assembled over complement each other (TABLE 2). The use of single several decades and can be used for targeted pathway candidate genes has been useful as part of initial approaches. We propose that future pharmacogenetic investigations but it will most likely never provide the research considers entire pathways, their regulatory sensitivity and specificity that is needed for tailored mechanisms and homeostasis, and evaluates genetic treatment decisions. Recently developed ‘-omic variability in a given pathway comprehensively.

1. Relling, M. V. & Dervieux, T. Pharmacogenetics and cancer 15. Park, D. J. et al. A Xeroderma pigmentosum group D gene normal and tumorous target tissues. J. Clin. Invest. 52, therapy. Nature Rev. Cancer 1, 99–108 (2001). polymorphism predicts clinical outcome to platinum-based 1804–1811 (1973). 2. Ulrich, C. M., Robien, K. & Sparks, R. Pharmacogenetics chemotherapy in patients with advanced colorectal cancer. 32. Sirotnak, F. M. & Moccio, D. M. Pharmacokinetic basis for and folate metabolism: a promising direction. Cancer Res. 61, 8654–8658 (2001). differences in methotrexate sensitivity of normal 3, 299–313 (2002). 16. Stoehlmacher, J. et al. A polymorphism of the XRCC1 gene proliferative tissues in the mouse. Cancer Res. 40, 3. Watters, J. W. & McLeod, H. L. Cancer pharmacogenomics: predicts for response to platinum based treatment in 1230–1234 (1980). current and future applications. Biochim.Biophys. Acta 2, advanced colorectal cancer. Anticancer Res. 21, 33. Kaneda, S. et al. Role in translation of a triple tandemly 99–111 (2003). 3075–3079 (2001). repeated sequence in the 5’-untranslated region of human 4. Rogers, J. F., Nafziger, A. N. & Bertino, J. S. Jr. 17. Marsh, S., Kwok, P. & McLeod, H. L. SNP databases and thymidylate synthase mRNA. Nucleic Acids Res. 15, Pharmacogenetics affects dosing, efficacy, and toxicity of pharmacogenetics: great start, but a long way to go. Hum. 1259–1270 (1987). cytochrome P450-metabolized drugs. Am. J. Med. 113, Mutat. 20, 174–179 (2002). 34. Marsh, S. et al. Novel thymidylate synthase enhancer region 746–750 (2002). 18. Sachidanandam, R. et al. A map of human genome alleles in African populations. Hum. Mutat. 16, 528 (2000). 5. Sindrup, S. H. & Brosen, K. The pharmacogenetics of sequence variation containing 1.42 million single nucleotide 35. Horie, N., Aiba, H., Oguro, K., Hojo, H. & Takeishi, K. codeine hypoalgesia. Pharmacogenetics 5, 335–346 polymorphisms. Nature. 409, 928–933 (2001). Functional analysis and DNA polymorphism of the tandemly (1995). 19. Wagner, C. Biochemical role of folate in cellular metabolism repeated sequences in the 5’-terminal regulatory region of 6. Relling, M. V. et al. Mercaptopurine therapy intolerance (ed. Bailey, L.) (Marcel Dekker, New York, 1995). the human gene for thymidylate synthase. Cell Struct. Funct. and heterozygosity at the thiopurine S-methyltransferase 20. Longley, D. B., Harkin, D. P. & Johnston, P. G. 5-Fluorouracil: 20, 191–197 (1995). gene locus. J. Natl Cancer Inst. 91, 2001–2008 mechanisms of action and clinical strategies. Nature Rev. 36. Pullarkat, S. T. et al. Thymidylate synthase gene (1999). Cancer 3, 330–338 (2003). polymorphism determines response and toxicity of 5-FU Lowering the dose of mercaptopurine allowed 21. Chu, E. & Allegra, C. in Cancer Chemotherapy and chemotherapy. Pharmacogenomics J. 1, 65–70 (2001). children with the thiopurine S-methyltransferase Biotherapy (eds. Chabner, B. & Longo, D.) 109–148 37. Trinh, B. N., Ong, C.-N., Coetzee, G. A., Yu, M. C. & Laird, P. W. enzyme deficiency to complete the course of (Lippincott–Raven Publishers, Philadelphia, 1996). Thymidylate synthase: a novel genetic determinant of chemotherapy and maintain sufficient erythrocyte 22. Allegra, C. J. et al. Enhanced inhibition of thymidylate plasma homocysteine and folate levels. Hum. Genet. 111, thioguanine nucleotide concentrations. synthase by methotrexate polyglutamates. J.Biol. Chem. 299–302 (2002). 7. Evans, W. E. et al. Preponderance of thiopurine 260, 9720–9726 (1985). 38. Skibola, C. F. et al. Polymorphisms in the thymidylate S-methyltransferase deficiency and heterozygosity among 23. Baggott, J. E., Vaughn, W. H. & Hudson, B. B. Inhibition of synthase and serine hydroxymethyltransferase genes and patients intolerant to mercaptopurine or azathioprine. J. Clin. 5-aminoimidazole-4-carboxamide ribotide transformylase, risk of adult acute lymphocytic leukemia. Blood 99, Oncol. 19, 2293–2301 (2001). adenosine deaminase and 5’-adenylate deaminase by 3786–3791 (2002). 8. McLeod, H. L. & Siva, C. The thiopurine S-methyltransferase polyglutamates of methotrexate and oxidized folates and by 39. Ulrich, C. M., Bigler, J., Bostick, R., Fosdick, L. & Potter, J. D. gene locus: implications for clinical pharmacogenomics. 5-aminoimidazole-4-carboxamide riboside and ribotide. Thymidylate synthase promoter polymorphism, interaction Pharmacogenomics. 3, 89–98 (2002). Biochem. J. 236, 193–200 (1986). with folate intake, and risk of colorectal adenomas. Cancer 9. Milano, G. & McLeod, H. L. Can dihydropyrimidine 24. Chu, E., Drake, J. C., Boarman, D., Baram, J. & Allegra, C. J. Res. 62, 3361–3364 (2002). dehydrogenase impact 5-fluorouracil-based treatment? Eur. Mechanism of thymidylate synthase inhibition by 40. Ulrich, C. et al. Searching expressed sequence tag J. Cancer 36, 37–42 (2000). methotrexate in human neoplastic cell lines and normal databases: discovery and confirmation of a common 10. Mattison, L. K., Soong, R. & Diasio, R. B. Implications of human myeloid progenitor cells. J. Biol. Chem. 265, polymorphism in the thymidylate synthase gene. Cancer dihydropyrimidine dehydrogenase on 5-fluorouracil 8470–8478 (1990). Epidemiol. Biomarkers Prev. 9, 1381–1385 (2000). pharmacogenetics and pharmacogenomics. 25. Grem, J. in Cancer chemotherapy and biotherapy (eds. 41. Lenz, H.-J. et al. A 6 base-pair deletion in the 3′UTR of the Pharmacogenomics 3, 485–492 (2002). Chabner, B. & Longo, D.) 149–211 (Lippincott–Raven thymidylate synthase (TS) gene predicts TS mRNA 11. Iyer, L. et al. Phenotype–genotype correlation of in vitro Publishers, Philadelphia, 1996). expression in colorectal tumors. A possible candidate gene SN-38 (active metabolite of irinotecan) and bilirubin 26. McGavin, J. K. & Goa, K. L. Capecitabine: a review of its use for colorectal cancer risk. Proc. Am. Assoc. Cancer Res. 43, glucuronidation in human liver tissue with UGT1A1 in the treatment of advanced or metastatic colorectal 660 (2002). promoter polymorphism. Clin. Pharm. Ther. 65, 576–582 cancer. Drugs 61, 2309–2326 (2001). 42. Frosst, P. et al. A candidate genetic risk factor for vascular (1999). 27. Blount, B. C. et al. Folate deficiency causes uracil disease: a common in methylenetetrahydrofolate Individuals who were homozygous for the UGT1A1 misincorporation into human DNA and chromosome reductase. Nature Genet. 10, 111–113 (1995). variant were found to have significantly lower levels of breakage: implications for cancer and neuronal damage. 43. Skibola, C. F. et al. Polymorphisms in the glucuronidation following irinotecan administration. Proc. Natl Acad.Sci. USA 94, 3290–3295 (1997). methylenetetrahydrofolate reductase gene are associated grade III and grade IV diarrhoea and neutropaenia 28. Duthie, S. J. & McMillan, P. Uracil misincorporation in human with susceptibility to acute leukemia in adults. Proc. Natl only occurred in patients who were UGT1A1 DNA detected using single cell gel electrophoresis. Acad. Sci. USA 96, 12810–12815 (1999). heterozygous or a homozygous variant. Carcinogenesis 18, 1709–1714 (1997). 44. Wiemels, J. L. et al. Methylenetetrahydrofolate reductase 12. Ando, Y. et al. Polymorphisms of UDP-glucuronosyl 29. Duthie, S. J. & Hawdon, A. DNA instability (strand breakage, (MTHFR) polymorphisms and risk of molecularly defined transferase gene and irinotecan toxicity: a pharmacogenetic uracil misincorporation, and defective repair) is increased by subtypes of childhood acute leukemia. Proc. Natl Acad. Sci. analysis. Cancer Res. 60, 6921–6926 (2000). folic acid depletion in human lymphocytes in vitro. FASEB J. USA 98, 4004–4009 (2001). 13. Iyer, L. et al. UGT1A1*28 polymorphism as a determinant of 12, 1491–1497 (1998). 45. Ma, J. et al. Methylenetetrahydrofolate reductase irinotecan disposition and toxicity. Pharmacogenomics J. 2, 30. Lorico, A. et al. Accumulation of DNA strand breaks in cells polymorphism, dietary interactions, and risk of colorectal 43–47 (2002). exposed to methotrexate or N10-propargyl-5,8-dideazafolic cancer. Cancer Res. 57, 1098–1102 (1997). 14. Illmer, T. et al. MDR1 gene polymorphisms affect therapy acid. Cancer Res. 48, 2036–2041 (1988). 46. Chen, J. et al. A methylenetetrahydrofolate reductase outcome in acute myeloid leukemia patients. Cancer Res. 31. Chabner, B. A. & Young, R. C. Threshold methotrexate polymorphism and the risk of colorectal cancer. Cancer Res. 62, 4955–4962 (2002). concentration for in vivo inhibition of DNA synthesis in 56, 4862–4864 (1996).

NATURE REVIEWS | CANCER VOLUME 3 | DECEMBER 2003 | 919 REVIEWS

47. Ulrich, C. M. et al. Colorectal adenomas and the C677T Increased severity of oral mucositis and slower This report outlines the use of microarray analysis to MTHFR polymorphism: evidence for gene–environment platelet recovery was observed among individuals classify patients with breast cancer as having a gene- interaction? Cancer Epidemiol. Biomarkers Prev. 8, 659–668 with the homozygous variant MTHFR 677TT expression profile that is consistent with either poor (1999). genotype compared with individuals with the wild- or good prognosis. Multivariate Cox regression 48. Slattery, M. L., Potter, J. D., Samowitz, W., Schaffer, D. & type (MTHFR 677CC) or heterozygous (MTHFR analysis found the gene-expression profile to be a Leppert, M. Methylenetetrahydrofolate reductase, diet, and 677CT) genotype following haematopoietic-cell strong predictor of disease outcome. risk of colon cancer. Cancer Epidemiol. Biomarkers Prev. 8, transplantation for chronic myelogenous 75. van ‘t Veer, L. J. et al. Gene expression profiling predicts 513–518 (1999). leukaemia. clinical outcome of breast cancer. Nature 415, 530–536 49. Le Marchand, L. et al. B-vitamin intake, metabolic genes, 63. Storb, R. et al. Methotrexate and cyclosporine compared (2002). and colorectal cancer risk (United States). Cancer Causes with cyclosporine alone for prophylaxis of acute graft versus 76. Rothenberg, M. L., Carbone, D. P. & Johnson, D. H. Improving Control 13, 239–248 (2002). host disease after marrow transplantation for leukemia. the evaluation of new cancer treatments: challenges and 50. Shaw, G. M., Rozen, R., Finnell, R. H., Wasserman, C. R. & N. Engl. J. Med. 314, 729–735 (1986). opportunities. Nature Rev. Cancer 3, 303–309 (2003). Lammer, E. J. Maternal vitamin use, genetic variation of 64. Chiusolo, P. et al. Preponderance of methylenetetra- 77. Villafranca, E. et al. Polymorphisms of the repeated infant methylenetetrahydrofolate reductase, and risk for hydrofolate reductase C677T homozygosity among sequences in the enhancer region of the thymidylate spina bifida. Am. J. Epidemiol. 148, 30–37 (1998). leukemia patients intolerant to methotrexate. Ann. Oncol. synthase gene promoter may predict downstaging after 51. Ou, C. Y. et al. 5,10 Methylenetetrahydrofolate reductase 13, 1915–1918 (2002). preoperative chemoradiation in rectal cancer. J. Clin. Oncol. genetic polymorphism as a risk factor for neural tube 65. Toffoli, G. et al. Effect of methylenetetrahydrofolate 19, 1779–1786 (2001). defects. Am. J. Med. Genet. 63, 610–614 (1996). reductase 677C→T polymorphism on toxicity and Patients with rectal cancer receiving pre-operative 52. Kluijtmans, L. A. et al. Thermolabile homocysteine plasma level after chronic methotrexate 5-fluorouracil who were homozygous for the triple methylenetetrahydrofolate reductase in coronary artery treatment of ovarian cancer patients. Int. J. Cancer 103, 28bp enhancer-region repeat of the thymidylate- disease. Circulation 96, 2573–2577 (1997). 294–299 (2003). synthase gene were found to have a lower 53. van der Put, N. M. et al. A second common mutation in the 66. Toffoli, G., Veronesi, A., Boiocchi, M. & Crivellari, D. MTHFR probability of tumour downstaging and lower methylenetetrahydrofolate reductase gene: an additional risk gene polymorphism and severe toxicity during adjuvant disease-free survival rates when compared with factor for neural-tube defects? Am. J. Hum. Genet. 62, treatment of early breast cancer with cyclophosphamide, patients who were homozygous for the double 1044–1051 (1998). methotrexate, and fluorouracil (CMF). Ann. Oncol. 11, repeat or heterozygous 2/3 repeat. 54. Weisberg, I., Tran, P., Christensen, B., Sibani, S. & Rozen, R. 373–374 (2000). 78. Park, D. J. et al. Thymidylate synthase gene polymorphism A second genetic polymorphism in 67. Taub, J. W. et al. Polymorphisms in methylenetetra- predicts response to capecitabine in advanced colorectal methylenetetrahydrofolate reductase (MTHFR) associated hydrofolate reductase and methotrexate sensitivity in cancer. Int. J. Colorectal Dis. 17, 46–49 (2002). with decreased enzyme activity. Mol. Genet. Metab. 64, childhood acute lymphoblastic leukemia. Leukemia 16, 169–172 (1998). 764–765 (2002). Acknowledgements 55. Keku, T. et al. 5,10-Methylenetetrahydrofolate reductase 68. Krajinovic, M., Costea, I. & Chiasson, S. Polymorphism of Support for K. R. was provided by a training grant from the National codon 677 and 1298 polymorphisms and colon cancer in the thymidylate synthase gene and outcome of acute Cancer Institute. H. M. is supported in part by the National Institute African Americans and whites. Cancer Epidemiol. lymphoblastic leukaemia. Lancet 359, 1033–1034 of Health Pharmacogenetics Research Network. Biomarkers Prev. 11, 1611–1621 (2002). (2002). 56. Chango, A. et al. A polymorphism (80G->A) in the reduced 69. Laverdiere, C., Chiasson, S., Costea, I., Moghrabi, A. & Competing interests statement folate carrier gene and its associations with folate status and Krajinovic, M. Polymorphism G80A in the reduced folate The authors declare that they have no competing financial interests. homocysteinemia. Mol. Genet. Metab. 70, 310–315 (2000). carrier gene and its relationship to methotrexate plasma 57. Marsh, S., Collie-Duguid, E. S., Li, T., Liu, X. & McLeod, H. L. levels and outcome of childhood acute lymphoblastic Ethnic variation in the thymidylate synthase enhancer region leukemia. Blood 100, 3832–3834 (2002). Online links polymorphism among Caucasian and Asian populations. 70. Cohen, V. et al. Methylenetetrahydrofolate reductase Genomics 58, 310–312 (1999). polymorphism in advanced colorectal cancer: a novel DATABASES 58. Marsh, S., McKay, J. A., Cassidy, J. & McLeod, H. L. genomic predictor of clinical response to fluoropyrimidine- The following terms in this article are linked online to: Polymorphism in the thymidylate synthase promoter based chemotherapy. Clin. Cancer Res. 9, 1611–1615 Cancer.gov: http://cancer.gov/ enhancer region in colorectal cancer. Int. J. Oncol. 19, (2003). acute lymphocytic leukaemia | acute myelogenous leukaemia | 383–386 (2001). 71. Watters, J. W. & McLeod, H. L. Using genome-wide colorectal cancer 59. Iacopetta, B., Grieu, F., Joseph, D. & Elsaleh, H. A mapping in the mouse to identify genes that influence LocusLink: http://www.ncbi.nlm.nih.gov/LocusLink/ polymorphism in the enhancer region of the thymidylate drug response. Trends Pharmacol. Sci. 24, 55–58 ABCB1 | CYP2D6 | ERCC2 | MTHFR | UGT1A1 | XRCC1 synthase promoter influences the survival of colorectal (2003). cancer patients treated with 5-fluorouracil. Br. J. Cancer 85, 72. Haston, C. K. et al. Bleomycin hydrolase and a genetic locus FURTHER INFORMATION 827–830 (2001). within the MHC affect risk for pulmonary fibrosis in mice. Pharmacogenetics and Pharmacogenomics Knowledge 60. Ueland, P. M., Hustad, S., Schneede, J., Refsum, H. & Hum. Mol. Genet. 11, 1855–1863 (2002). Base: http://www.pharmgkb.org/do/serve?id=home.welcome Vollset, S. E. Biological and clinical implications of the This report outlines how mouse models were used to Pharmacogenetics Research Network, Washington MTHFR C677T polymorphism. Trends Pharmacol. Sci. 22, conduct a genome-wide screen for identification of University: http://pharmacogenetics.wustl.edu 195–201 (2001). susceptibility loci for bleomycin-induced pulmonary Kyoto Encyclopedia of Genes and : 61. Wisotzkey, J. K., Toman, J., Bell, T., Monk, J. S. & Jones, fibrosis. http://www.genome.ad.jp/kegg/ D. MTHFR (C677T) polymorphisms and stage III colon 73. Marton, M. J. et al. Drug target validation and identification Pharmacogenomics glossary, Cambridge Healthtech cancer: response to therapy. Mol. Diagn. 4, 95–99 (1999). of secondary drug target effects using DNA microarrays. Institute: http://www.genomicglossaries.com/content/ 62. Ulrich, C. M. et al. Pharmacogenetics of methotrexate: Nature Med. 4, 1293–1301 (1998). pharmacogenomics.asp toxicity among marrow transplantation patients varies with 74. van de Vijver, M. J. et al. A gene-expression signature as a The Mouse Models of Human Cancers Consortium: the methylenetetrahydrofolate reductase polymorphism. predictor of survival in breast cancer. N. Engl. J. Med. 347, http://emice.nci.nih.gov Blood 98, 231–234 (2001). 1999–2009 (2002). Access to this interactive links box is free online

920 | DECEMBER 2003 | VOLUME 3 www.nature.com/reviews/cancer