Genetic Factors Influencing Pyrimidine- Antagonist Chemotherapy

Home , Antimetabolite, Capecitabine, Cytarabine, Dihydropyrimidinase, Doxifluridine, DPYS

Genetic factors influencing Pyrimidine- antagonist chemotherapy

JG Maring1 ABSTRACT 2 Pyrimidine antagonists, for example, 5-fluorouracil (5-FU), cytarabine (ara-C) HJM Groen and gemcitabine (dFdC), are widely used in chemotherapy regimes for 2 FM Wachters colorectal, breast, head and neck, non-small-cell lung cancer, pancreatic DRA Uges3 cancer and leukaemias. Extensive metabolism is a prerequisite for conversion EGE de Vries4 of these pyrimidine prodrugs into active compounds. Interindividual variation in the activity of metabolising enzymes can affect the extent of 1Department of Pharmacy, Diaconessen Hospital prodrug activation and, as a result, act on the efficacy of chemotherapy Meppel & Bethesda Hospital Hoogeveen, Meppel, treatment. Genetic factors at least partly explain interindividual variation in 2 The Netherlands; Department of Pulmonary antitumour efficacy and toxicity of pyrimidine antagonists. In this review, Diseases, University of Groningen & University Medical Center Groningen, Groningen, The proteins relevant for the efficacy and toxicity of pyrimidine antagonists will Netherlands; 3Department of Pharmacy, be summarised. In addition, the role of germline polymorphisms, tumour- University of Groningen & University Medical specific somatic mutations and protein expression levels in the metabolic Center Groningen, Groningen, The Netherlands; pathways and clinical pharmacology of these drugs are described. Germline 4Department of Medical Oncology, University of Groningen & University Medical Center polymorphisms of uridine monophosphate kinase (UMPK), orotate phos- Groningen, Groningen, The Netherlands phoribosyl transferase (OPRT), thymidylate synthase (TS), dihydropyrimidine dehydrogenase (DPD) and methylene tetrahydrofolate reductase (MTHFR) Correspondence: and gene expression levels of OPRT, UMPK, TS, DPD, uridine phosphorylase, Dr JG Maring, Department of Pharmacy, uridine kinase, thymidine phosphorylase, thymidine kinase, deoxyuridine Diaconessen Hospital Meppel & Bethesda Hospital Hoogeveen, Hoogeveenseweg 38, triphosphate nucleotide hydrolase are discussed in relation to 5-FU efficacy. 7943KA Meppel, The Netherlands. Cytidine deaminase (CDD) and 50-nucleotidase (5NT) gene polymorphisms Tel: þ 31 522 234900 and CDD, 5NT, deoxycytidine kinase and MRP5 gene expression levels and Fax: þ 31 522 243900 their potential relation to dFdC and ara-C cytotoxicity are reviewed. E-mail: [email protected] The Pharmacogenomics Journal (2005) 5, 226–243. doi:10.1038/sj.tpj.6500320

Keywords: pyrimidine antagonists; 5-fluorouracil; gemcitabine; cytarabine; polymorphisms; mutations; gene expression

INTRODUCTION Pyrimidine antagonists belong to the group of antimetabolite anticancer drugs and show structural resemblance with naturally occurring nucleotides (see Figure 1). Their action is accomplished through incorporation as false precursor in DNA or RNA or through inhibition of proteins involved in nucleotide metabolism. The most commonly used pyrimidine antagonists are 5-fluorouracil (5-FU), gemcitabine (dFdC) and cytarabine (ara-C). Newer oral variants of 5-FU are capecitabine and tegafur. 5-FU and its analogues are used, for example, in the treatment of colorectal, breast and head and neck cancer,1–3 whereas dFdC is especially prescribed for non-small-cell lung cancer and pancreatic cancer.4,5 Ara- C is used in the treatment of leukaemia.6 All pyrimidine antagonists are prodrugs Received: 17 December 2004 Revised: 3 May 2005 and intracellular conversion into cytotoxic nucleosides and nucleotides is Accepted: 5 May 2005 needed to produce cytotoxic metabolites. Proteins, involved in pyrimidine Pyrimidine antagonist pharmacogenetics JG Maring et al 227

Figure 1 Overview of the chemical structures of the naturally occurring pyrimidines cytosine and uracil and the synthetic pyrimidine analogues cytarabine, dFdC, capecitabine, 5-Fu and tegafur.

metabolism, handle these synthetic drugs, as if they were activity results from inhibition of thymidylate synthase (TS) naturally occurring substrates. The extensive metabolism of by FdUMP, as well as from incorporation of 5-FU metabolites pyrimidine antagonists implies that the intracellular con- into RNA and DNA. Only a small part of the 5-FU dose is centrations of cytotoxic metabolites, thus indirectly the activated via these routes, as in humans 80–90% of the potential antitumour effects, largely depend on intracellular administered dose is degraded to 5,6 dihydrofluorouracil metabolic enzyme activity. (DHFU) by dihydropyrimidine dehydrogenase (DPD). DHFU The aim of this review is to summarise pharmacogenomic can be further degraded to fluoro-b-ureidopropionate data regarding proteins related to the efficacy and toxicity of (FUPA) by dihydropyrimidinase (DPYS) and subsequently pyrimidine antagonists and to identify potential predictive to fluoro-b-alanine (FBAL) by b-ureidopropionase (BUP1). 5- and/or prognostic genetic factors for toxicity and treatment FU degradation occurs in all tissues, including tumour outcome. The impact of germline polymorphisms as well as tissue, but is highest in the liver.9 tumour-specific somatic mutations and protein expression levels on the clinical pharmacology and metabolic pathways of these drugs will be discussed. Capecitabine and tegafur Capecitabine is an oral prodrug of 5-FU. After absorption Metabolic Pathways of Pyrimidine Antagonists from the gut, capecitabine is converted into 50-deoxy-5- 5-Fluorouracil fluorocytidine (50-dFCR) by carboxyl-esterase in the liver, The anabolic conversion of 5-FU into nucleotides is essential and subsequently further converted into 50-deoxy-5-fluoro- for its action. Several enzymes of the pyrimidine metabolic uridine (50-dFUR) by cytidine deaminase (CDD). Finally, 5- pathway are required for the conversion of 5-FU to FU results from bioconversion of 50-dFUR by TP10 (see Figure nucleotides.7 Cytotoxic nucleotides can be formed by three 2). Tegafur is another oral 5-FU prodrug, that is converted routes, as illustrated in Figure 2: (1) conversion of 5-FU to 5- into 5-FU by cytochrome P450 (CYP450) enzymes in the fluoro-uridine-monophosphate (FUMP) by orotate phos- liver. CYP2A6 is the main CYP450 enzyme involved in phoribosyl transferase (OPRT); (2) sequential conversion of tegafur activation, but CYP1A2 and CYP2C8 also play a role 5-FU to FUMP by uridine phosphorylase (UP) and uridine (see Figure 2). Tegafur is combined with uracil in a molar kinase; (3) sequential conversion of 5-FU to 5-fluoro-deoxy- proportion of 1 : 4 available in the commercial preparation uridine-monophosphate (FdUMP) by thymidine phospho- UFTs. Uracil is a competitive substrate for DPD and its role rylase (TP) and thymidine kinase (TK).8 The antitumour in UFTs is to diminish 5-FU catabolism by DPD.

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 228

Figure 2 Metabolism of 5-FU and 5-FU analogues. For explanation of symbols and metabolic routes, see text.

dFdC and ara-C ara-C, which halts polymerase progression at the analogue insertion site. The masked termination of dFdC makes the The metabolic pathways of dFdC and ara-C are almost inserted analogue more resistant to removal from DNA.12 alike11,12 (see Figure 3). Membrane transport of both dFdC Other differences with ara-C are the faster membrane and ara-C is mediated by equilibrative nucleoside transpor- transport velocity of dFdC, the greater effectiveness of dFdC ters. Subsequently, dFdC is phosphorylated into dFdC phosphorylation by dCK and the longer intracellular monophosphate (dFdCMP) by deoxycytidine kinase (dCK). retention of dFdCTP. These factors may, at least in part, The same enzyme is responsible for the intracellular explain the different spectra of antitumour activity of both phosphorylation of ara-C into cytarabine monophosphate drugs. Only a small part of the dFdC and ara-C dose is (ara-CMP). dCK is the rate-limiting enzyme in the biotrans- responsible for the cytotoxic effects, since more than 90% of formation of both dFdC and ara-C. Inactivation of dFdCMP the dose is inactivated by the enzyme CDD into dFdU and and ara-CMP can occur through dephosphorylation by 50- uracil-arabinoside (ara-U), respectively. nucleotidase (5NT). Monophosphates, escaping from dephosphorylation are available for further phosphorylation Germline Polymorphisms and Fluoropyrimidine Efficacy into di- and triphospates by dCMP kinase and nucleoside 5-Fluorouracil diphosphate kinase, respectively. dFdC diphosphate (dFdCDP) is a potent inhibitor of the enzyme ribonucleotide Genetic polymorphisms of enzymes involved in the meta- reductase (RNR), which will lead to depletion of deoxycy- bolic activation pathway of 5-FU have been described for the tidine diphosphate (dCDP) and deoxycytidine triphosphate enzymes uridine monophosphate kinase (UMPK) and (dCTP) in the cell. This may favour the incorporation of orotate phosphorylase transferase (OPRT). Three allelic dFdCTP into DNA. Moreover, dFdC has the unique property variants of UMPK have been recognised in the human that, after incorporation of dFdC monophosphate in DNA, population: UMPK1, UMPK2 and UMPK3.14 The UMPK1 one more deoxynucleotide molecule can be inserted. This allele is associated with about three times the catalytic stops DNA polymerase.13 This pattern is distinct from that of activity of the UMPK2 allele. Therefore, UMPK2 homo-

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 229

Figure 3 Metabolism of dFdC and cytarabine. For explanation of symbols and metabolic routes, see text.

zygotes are relatively deficient of total UMPK enzyme. The Africans and Caucasians (about 50%), but the frequency of allele frequency of UMPK1 is about 95–97% and that of homozygous TSER*3 was reported to be significantly higher UMPK2 3–5%, in both Caucasians and Asians.15–17 UMPK3 is in Asians (67%) compared to Caucasians (38%).24 The rarely seen. The consequences of this polymorphism for 5- TSER*4 and TSER*9 alleles are found in higher frequencies FU chemotherapy have not yet been studied. in African populations (2–7%) compared to Caucasians In the OPRT gene, a G213A mutation in exon 3 and a (0–1%).25 440G mutation in exon 6 have been observed, with allele The consequences of the TSER genotype for fluoropy- frequencies of 26 and 27%, respectively.18 Both mutations rimidine chemosensitivity have been studied in both do not significantly compromise in vitro OPRT activity, and tumour cell lines and clinical trials. In two studies with therefore it seems unlikely that they affect 5-FU-antitumour tumour cell lines, the TSER polymorphism had no effect on efficacy. 5-FU chemosensitivity.26,27 Contrary to in vitro data, in 65 Additionally to the 5-FU-activating enzymes, far more patients with rectal cancer, tumour downstaging after information is available regarding polymorphisms of target preoperative 5-FU-based chemoradiation was observed in enzyme TS. For TS, at least five genotypes have been only 22% of homozygotes for triple tandem repeat identified, characterised by variable numbers of a 28-bp sequences compared to 60% of patients who were hetero- tandem repeat sequence in the DNA promoter enhancer zygous or homozygous for double tandem repeats.28 In region (TSER). So far, alleles containing 2 (TSER*2), 3 another small retrospective study of 24 patients with (TSER*3), 4 (TSER*4), 5 (TSER*5) and 9 (TSER*9) copies of metastatic colorectal cancer, three out of four homozygous the tandem repeat have been identified.19 The double and TSER*2 patients responded on capecitabine, whereas only triple repeats are the predominant alleles. The triple tandem one out of 12 of heterozygotes and two out of eight genotype is associated with higher TS mRNA and TS protein homozygous TSER*3 patients responded.29 Furthermore, in levels compared to the double tandem genotype.20–23 The a study of 221 patients with Dukes C colorectal cancer, effect of higher numbers of repeats is not yet clear. The allele patients homozygous for TSER*3 gained no survival benefit frequency of TSER*3 appears almost similar in Asians, from 5-FU treatment, whereas survival was increased in

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 230

heterozygotes and patients homozygous for TSER*2.30 tetrahydrofolate. 5,10-Methyltetrahydrofolate is an essential However, the TSER genotype did not seem to be an cofactor in the biosynthesis of dTMP (see Figure 2). The efficacious marker for tumour sensitivity to 5-FU-based oral dissociation of FdUMP from the ternary complex with TS adjuvant chemotherapy in 135 Japanese colorectal pa- and 5,10-methyltetrahydrofolate is suppressed when levels tients.31 Surprisingly, in a small study of 54 colorectal of 5,10- methyltetrahydrofolate are increased.39 So far, two cancer patients, not the TSER*2 but the TSER*3 genotype MTHFR polymorphisms have been recognised. A common correlated with an increased disease-free survival after C677T transition in exon 4 of the MTHFR gene results in a adjuvant 5-FU chemotherapy.32 thermolabile enzyme variant with lower specific activity.40 Several factors can possibly explain these inconclusive The second polymorphism concerns a codon 1298 A to C data. Firstly, within the second 28 bp tandem repeat transition in exon 7, which is also associated with decreased sequence, a G/C polymorphism was recently discovered. enzymatic activity.41 In vitro transfection of mutant 677T Functional analysis revealed that the TSER*3/G genotype MTHFR cDNA in colon and breast cancer cells increased the has three to four times greater efficiency of translation than chemosensitivity of these cells to 5-FU, suggesting at least a other polymorphic sequences.33,34 In a study of the G/C modulating role of this genotype on cellular fluoropyrimi- polymorphism in 258 primary colorectal tumours, it was dine sensitivity.42 The geographic and ethnic distribution of found that all low-expression genotypes were associated the 677T genotype was studied by Wilcken et al.43 The with longer survival after adjuvant fluoropyrimidine che- homozygote T genotype was found to be particularly motherapy, whereas no benefit was seen for high-expression common in northern China (20%), southern Italy (26%) genotypes.33 and Mexico (32%). In two studies with, respectively, 45 and Secondly, a 6 bp deletion genotype (TS1494del6) in the 30 51 colorectal cancer patients, the 677T genotype appeared untranslated region of the TS gene has been associated with to affect the folate pool.44,45 In the latter study, no effect was reduced TS mRNA stability.35,36 The frequency of this seen of the 677T genotype on overall survival after oral 5- genotype was found to be 41% in non-Hispanic whites, FU-based chemotherapy.45 Contrary to this, in a study of 43 26% in Hispanic whites, 52% in African-Americans and 76% metastatic colorectal cancer patients receiving fluoropyri- in Singapore Chinese. The relative instability of TSmRNA midine-based chemotherapy, the presence of one or two may have its effect on TS protein activity and thus indirectly mutant alleles was associated with increased tumour on fluoropyrimidine chemosensitivity. However, this poly- response.46 In a recent study of 98 dissiminated colorectal morphism was not directly related to in vitro chemosensi- cancer patients, the C677T mutation was also associated tivity in tumour cell lines27 or disease-free survival after with higher response rates on 5-FU/folinic acid chemothe- adjuvant 5-FU chemotherapy in a small study of 54 colo- rapy.47 In the same study, the A1298C polymorphism was rectal cancer patients.32 studied, but this was not related to the response rate. The Another cause of conflicting data is the loss of hetero- exact roles of the MTHFR polymorphisms have to be zygosity (LOH), that is frequently observed for TS.37 The TS elucidated in larger prospective clinical trials. gene is located on chromosome 18p11.32, a region most Besides the above-mentioned polymorphisms in the 5-FU frequently lost in colorectal cancer. In a small study of 30 anabolic route, genetic variations in 5-FU catabolic enzymes stage IV colorectal cancer patients, LOH was observed in 17 can also have a profound effect on 5-FU toxicity. At least the out of 22 heterozygotes.38 The presence of a TSER*3 allele phenomenon of inherited DPD deficiency is an important (homozygous TSER*3, heterozygous and TSER*3/loss) was issue. DPD deficiency is caused by molecular defects in the associated with lower survival on fluoropyrimidine che- DPD gene that result in complete or partial loss of DPD motherapy. LOH in tumour tissue is a significant factor, to activity.48 This can cause extreme 5-FU toxicity. So far, 33 be taken into account when TS genotypes are to be different mutations have been identified in the dihydropy- considered for outcome prediction. It is important to bear rimidine dehydrogenase (DPYD) gene. At least 11 of these 33 in mind that LOH limits the use of germline DNA from for mutations, including one polymorphism, were detected in example, blood for predictive genotyping. patients suffering from severe 5-FU-associated toxicity.48–50 Finally, it must be stressed that many studies of the TS In patients who are deficient for DPD, 5-FU clearance is polymorphism were underpowered to detect relations of the dramatically reduced and standard doses of 5-FU cause different genotypes with clinical outcome. Hence, the exact excessive toxicity in these patients.48,51,52 The frequency of consequences of the triple tandem repeat genotypes for 5-FU DPD deficiency has been estimated to be as high as 2–3% in chemosensitivity are not yet clear. Sufficiently powered Caucasians based on measurements of DPD activity in clinical trials and additional mechanistic studies may be peripheral mononuclear cells in patients and healthy needed to elucidate the cause of so far unexpected out- volunteers 53,54 This percentage is probably high enough comes. to justify screening on DPD deficiency prior to 5-FU-based Directly linked to the 5-FU-mediated inhibition of TS is chemotherapy. Instead of screening for specific mutations in the presence of intracellular folate. A polymorphism that the DPYD gene, Mattison et al55 developed a simple uracil may influence the efficacy of 5-FU and its analogues by breath test for DPD phenotyping, based on the release of 13 13 influencing folate pools is that of the methylene tetrahy- CO2 from 2- C uracil in the presence of intact DPD. drofolate reductase (MTHFR) gene. MTHFR catalyses the Expired air was collected 5–180 min after oral ingestion of conversion of 5,10-methyltetrahydrofolate to 5-methyl- 6 mg/kg 2-13C uracil. Partially deficient DPD breath profiles

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 231

were well differentiated from normal profiles. Oral loading tumour samples of patients with acquired 5-FU resistance.63 with uracil, prior to chemotherapy, with subsequent Low OPRT as well as UP and UK activity has been associated measurement of the uracil clearance in plasma, might also with 5-FU resistance in tumour cell lines.64–67 The role of give a good prediction of the patient’s DPD status. Such tests OPRT has also been investigated in a few clinical studies. are currently developed, but not yet available.56 High OPRT activity was associated with good in vitro Apart from DPD, it was recently discovered that deficiency sensitivity to 5-FU, measured by Collagen Gel Droplet in dihydropyrimidinase (DHP), the second enzyme in 5-FU Embedded Culture Drug Sensitivity test, Fluoresceine Dia- catabolism, can also result in severe 5-FU toxicity.57 At least cetate Assay or Histoculture Drug Response Assay, in human six mutations have been identified in the the DHP encoding colorectal cancer tissues.68,69 Furthermore, in 37 patients gene. Missense mutation G883A in exon 5 has been treated with oral tegafur-uracil for disseminated colorectal associated with severe 5-FU toxicity.58 This mutation has cancer, responding tumours had higher expressions of OPRT been shown to result in mutant DHP enzyme without than nonresponding tumours.70 In the same study, there residual activity. was no difference in UP mRNA between responding and nonresponding tumours. The role of UP in 5-FU sensitivity Capecitabine and tegafur is probably marginal, since transfection of tumour cells with UP cDNA did not affect fluoropyrimidine sensitivity.71 More Polymorphisms in capacitabine-metabolising enzymes have research is needed to establish the exact role of OPRT in 5- only been reported for CDD. Since this polymorphism may FU sensitivity. also be relevant for dFdC and ara-C metabolism, it will be Most research of fluoropyrimidine activation pathways discussed in the next paragraph. Furthermore, mutations in has been focused on the role of TP. Results regarding the role CYP2A6 have recently been identified to affect tegafur of tumoral TP expression appear to be contradictory. In 59 metabolism. Two mutant alleles, CYP2A6*4C and patients, high as well as low TP mRNA expression levels have CYP2A6*11, were detected in a patient with largely reduced been correlated with response to 5-FU therapy. An increase tegafur clearance. The prevalence of these mutations and its in both relapse-free survival and overall survival was impact on clinical outcome are not yet clear. suggested in TP-positive compared to TP-negative breast tumours in a small study of 109 patients treated with Gemcitabine and ara-C adjuvant cyclophosphamide, methotrexate and 5-FU (CMF).72 Furthermore, in 38 metastatic colorectal cancer There are at least two polymorphisms in the metabolising patients treated with 5-FU and leucovorin (LV), TP mRNA pathway, that might be relevant for dFdC and ara-C toxicity levels in nonresponding tumours were higher than in and efficacy. Firstly, cloning of human CDD revealed two responding patients.73 In 28 patients treated for advanced protein variants (CDD1 and CDD2) with different in vitro gastric cancer with 5-FU plus pirarubicin and cisplatin, high deamination rates of ara-C.60 Theoretically, patients who tumour tissue TP expression was associated with response on deaminate dFdC or ara-C more efficiently are shorter chemotherapy.74 In another 126 advanced gastric cancer exposed to the parent drugs, while faster metabolism of patients, TP overexpression was associated with increased capecitabine results in increased 5-FU levels. In vitro patient survival after fluoropyrimidine chemotherapy, but transfection of human bladder cancer cells with CDD2 correlated with unfavourable prognosis in patients not cDNA did increase CDD activity, and indeed made the cells treated with fluoropyrimidines.75 Consequently, it is hard more sensitive to 50dFUR and capecitabine, but resistant to to draw conclusions from current, in general small trials. dFdC.61 This warrants further investigation of the possible Taking a close look at 5-FU metabolism, one might expect role of the CDD genotype in fluoropyrimidine chemosensi- that cells with higher TP levels would be more sensitive to 5- tivity. FU, due to higher FdUMP levels resulting from increased 5- Secondly, 50-NT deficiency is an autosomal recessive FU activation (see Figure 2). However, TP also functions as condition, known to cause haemolytic anaemia. Several an angiogenesis factor since TP- and platelet-derived mutations causing 50-NT deficiency have been identified.62 endothelial cell growth factor (ECGF1) have been recognised On theoretical grounds, in patients with 50-NT deficiency, as the same protein. High TP gene expression might treatment with dFdC or ara-C may result in increased therefore be associated with a more aggressive and malig- toxicity although this has not yet been published. An nant tumour phenotype.73 Transfection of cancer cells with overview of polymorphic genes and related enzymes with TP cDNA has been performed in several in vitro studies to possible relevance for fluoropyrimidine chemotherapy is investigate the effect of TP overexpression on chemosensi- presented in Table 1. tivity. A clear correlation was found between TP activity and in vitro sensitivity to the 5-FU analogue 50-dFUR, and to a Impact of Somatic Mutations And Gene Expression Levels: lesser extent to 5-FU itself.76–81 Marchetti et al76 also studied 5-FU and its Metabolic Enzymes the effect of TP overexpression on in vivo tumour growth in a 5-Fluorouracil rat model. The impact on tumour growth turned out to be Data on expression levels of enzymes involved in the relatively modest and only involved the initial stages of metabolic activation of 5-FU in relation to chemosensitivity tumour growth. These results suggest that tumour growth are sparse. Low UMPK levels were observed in 29 colorectal and chemosensitivity are independently related to TP

www.nature.com/tpj h hraoeoisJournal Pharmacogenomics The 232

Table 1 Overview of proteins may affect drug treatment outcome due to genetic polymorphisms

Enzyme Polymorphism Allele frequency Drug Patient data Consequences

Dihydropyrimidine 11 mutations found in IVS14G-A mutation 5-FU and analogues Miscellaneous cancer Increased toxicity of 5-FU dehydrogenase (DPD) partial DPD-deficient most frequent types patients 0.5–1%

Dihydropyrimidinase G883A missense mutation o1% 5-FU and analogues Colorectal cancer Increased toxicity of 5-FU yiiieatgns pharmacogenetics antagonist Pyrimidine (DHP) in exon 5

Methylene C667T in exon 4 45% in northern China 5-FU and analogues Colorectal cancer Possibly increased efficacy tetrahydrofolate reductase 51% in southern Italy of 5-FU (MTHFR) 56% in Mexico A1298C in exon 7 Unknown Colorectal cancer Unknown

Orotate phosphoribosyl G213A in exon 3 26% 5-FU and analogues Colorectal cancer Possibly increased efficacy Maring JG transferase (OPRT) of 5-FU 440G in exon 6 27% tal et Thymidylate synthase (TS) TSER*3 increased TS 50% in Asians and 5-FU and analogues Colorectal cancer Possibly decreased activity Caucasians efficacy of 5-FU

TSER*4/TSER*9 2–7% in Africans Not available Unknown 0–1% in Caucasians

1494del6 26–41% in Caucasians Not available Unknown 52% in African-Americans 76% in Singapore Chinese

Uridine monophosphate UMPK1 variant 95–97% 5-FU and analogues Colorectal cancer Unknown kinase (UMPK) UMPK2 variant 3–5% UMPK3 variant o1%

Cytochrome P450 2A6 CYP2A6*4C Unknown Tegafur Colorectal cancer Decreased metabolism of (CYP2A6) tegafur into 5-FU CYP2A6*11 Unknown

Cytidine deaminase CDD1 variant 70% Capecitabine, Not available Unknown (CDD) CDD2 variant 30% gemcitabine, ara-C

50-nucleotidase (5NT) Several mutations Unknown Gemcitabine and ara-C Leukaemia Unknown Pyrimidine antagonist pharmacogenetics JG Maring et al 233

expression. A major point of interest in interpreting the lines demonstrated an inverse correlation between both clinical data in relation to in vitro data is the role of DPD mRNA expression and activity with 5-FU response.85 infiltrating stromal fibroblasts and macrophages in TP DPD mRNA levels were also related to primary 5-FU expression. In most tumour types, TP expression is much resistance in seven human gastrointestinal cell lines.86 higher in infiltrating cells compared to tumour cells.82 This Furthermore, high DPD activity and DPD mRNA levels were may largely trouble the interpretation of data of in vivo PCR correlated with low sensitivity to 5-FU in three human as well as immunohistochemistry (IHC) studies. Conse- gastric, two colon, one breast and one pancreatic carcinoma quently, this aspect particularly has to be taken into account xenografts in the nude mice model.87 However, data of in study designs to validate TP as a predictive marker. presently available clinical studies are less unequivocal. An Data on the role of TK in 5-FU chemosensitivity are also overview of clinical relevant studies is presented in Table 2. contradictory. Chemosensitivity was unchanged in TK- Most studies have been performed in colorectal cancer,68,88–91 deficient colon cancer cells compared to parent cells, some in gastric,92–94 head and neck,95–97 and non-small-cell suggesting only a marginal role of TK in 5-FU sensitivity.83 lung cancer,98–100 and few in breast and bladder can- In another study, TK overexpression has been associated cer.101,102 A number of studies have indicated that there is with 5-FU resistance in gastric tumour cell lines,66 but other no strong link between DPD mRNA levels and DPD protein groups found a reduction of TK activity in chemoresistant activity levels. It has been reported that DPD mRNA, but not colon cancer cell lines.65,67 Intracellular availability of DPD activity, is reduced in colorectal tumours compared deoxyribose-1-phosphate might account for these differ- with normal mucosa, although considerable overlap ences, since the activation of 5-FU by TP and TK is fully exists in measured mRNA levels.103–110 Reduced mRNA dependent on the availability of this co-substrate. If cellular levels have also been reported in ovarian cancer.111 levels of this co-substrate are too low, hardly any 5-FU can be DPD activity appears to be increased in breast cancer (see metabolised via this route. Only one patient study is Table 3).101,112 This suggests that DPD is not only regulated available regarding tumour TK activity in relation to at transcriptional and translational, but also at the post- treatment efficacy. In this study, tumour TK activity could transcriptional, level. This hampers the use of DPD mRNA as not be related to the efficacy of second-line 5-FU (poly)- predictive marker for 5-FU efficacy. Inhomogeneity of chemotherapy in 121 advanced breast cancer patients who tumour tissue samples may, in part, also account for failed on first-line tamoxifen treatment.84 Thus, TK is conflicting results. probably not a strong factor related to 5-FU sensitivity. Summarising, it can be concluded that data on tumour A route that has been extensively evaluated in relation to expression levels of 5-FU metabolising enzymes in 5-FU chemoresistance is the intracellular biodegradation of relation to chemosensitivity are sparse and that studies 5-FU by DPD. In vitro experiments in human tumour cell show conflicting results. This may at least partly be due to

Table 2 Overview of studies regarding intratumoral DPD expression levels in relation to chmeosensitivity

Cancer Patients Chemotherapy regimen Assay Outcome Reference

Colorectal 33 5-FU/LV mRNA Low DPD related to response 88 100 5-FU ELISA No correlations found 89 348 5-FU mRNA No correlations found 90 54 5-FU Activity No correlations found 68 309 5-FU mRNA Low DPD related to longer survival 91

Gastric 81 5-FU Activity High DPD related to poor survival 92 22 DFUR ELISA Low DPD related to DFUR sensitivity 93 38 5-FU/MTX IHC No correlations found 94

Head and neck 62 5-FU Activity No correlations found 95 82 5-FU/LV/cisplatin Activity No correlations found 96 109 UFT IHC Low DPD correlated with UFT response 97

NSCLC 54 UFT IHC Low DPD correlated with better prognosis 98 68 5-FU IHC High DPD correlated with poor survival 99 60 5-FU Activity No correlations found 100

Breast 191 5-FU Activity High DPD related to poor DFS 101

Bladder 74 5-FU Activity Low DPD related to 5-FU sensitivity 102

MTX ¼ methotrexate;; IHC ¼ immunohistochemistry.

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 234

Table 3 DPD expression and activity in different tissue types

Tumour type Samples Parameter Tumour vs Normal Reference

Gastrointestinal 51, matched mRNA Lower 103 Activity No difference

41, matched mRNA Lower 104 Activity No difference

10 tumours+7 metastases, matched mRNA Lower 105

15 tumours+10 metastases, 25 nonmatched controls Protein Lower 106

36, matched Protein No difference 107

43, matched mRNA Lower 108 Protein No difference Activity No difference

36, matched Protein Lower 109

40, matched mRNA Lower 110 Protein Lower Activity No difference

Ovarian 85 tumour+27 nonmatched controls mRNA Lower 111

Breast 26, matched Activity Higher 112

49, matched Activity Higher 101

small numbers and therefore lack of statistical power in studied in relation to 5-FU efficacy. A large amount of the studies. preclinical in vitro data suggest a correlation between TS activity and sensitivity to 5-FU. In a panel of 19 nonselected Capecitabine/tegafur breast, digestive tract and head and neck cancer cell lines, as well as in a panel of 13 nonselected human colon cancer cell In the animal model, high TP/DPD ratios have been lines, the TS activity was inversely correlated with 5-FU associated with antitumour efficacy of capecitabine and its sensitivity.85,118 Several in vitro studies in tumour cell lines metabolite 50dFUR in human tumour xenografts.113 In vivo, also suggest TS overexpression as a mechanism of resistance high TP/DPD ratios suggested a better clinical outcome after after repeated exposure to 5-FU.86,119,120 Chemosensitivity adjuvant treatment with 50dFUR in a study of 88 colorectal to 5-FU was also increased after transfection of human colon cancer patients, and in 17 metastatic gastric cancer cancer cells with antisense TS cDNA.121 However, conflict- patients.114,115 TP expression was examined by IHC in 650 ing data have been observed in clinical studies evaluating breast tumours of patients with early breast cancer receiving the prognostic role of mRNA, protein and activity levels of adjuvant 50dFUR.116 Eight-year follow-up data showed that TS in relation to tumour response and clinical outcome to 5- high TP expression in the tumour was a favourable FU-based chemotherapy. An overview of relevant clinical prognostic indicator. Thus, TP expression seems to be a studies is presented in Table 4. Most studies regarding TS predictive factor for 50dFUR efficacy. Since the activation of expression in relation to chemosensitivity have been tegafur depends on several CYP450 enzymes, interindividual performed in patients with disseminated colorectal can- variations in enzyme expression may affect the rate of cer,68,91,122–134 some in gastric cancer94,135–138 and few tegafur metabolism. Indeed, the relative contribution of studies in head and neck cancer,95,96,139 non-small-cell lung each enzyme is known to differ among patients, but the cancer99,100 pancreatic and breast cancer.84,140,141 clinical consequences of this phenomenon are unclear.117 Several factors may account for the difficult interpretation of combined clinical data. Firstly, it has been reported that TS Impact of Somatic Mutations and Gene Expression Levels: protein can downregulate its own translation, whereas its 5-FU and its Target Enzymes transcription is regulated by E2F, a cell cycle checkpoint Contrary to enzymes related to the 5-FU metabolic activa- regulator. Together, this results in low TS levels in stationary tion pathway, the target enzyme TS has been extensively phase cells. Although cells with a low TS might theoretically

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 235

Table 4 Overview of studies regarding intratumoral TS expression levels in relation to chemosensitivity

Cancer Patients Chemotherapy regimen Assay Outcome Reference

Colorectal 47 5-FU/LV bolus Activity High TS related to non-response Peters et al122 46 5-FU/LV protracted mRNA High TS related to non-response Leichman et al123 36 5-FU/LV mRNA + IHC TS106 Low TS related to response Lenz et al124 41 5-FU/LV bolus IHC polyclonal Low TS related to response Cascinu et al125 52 5-FU/LV protracted IHC TS106 High TS related to non-response Wong et al126 48 5-FU/LV/MTX IHC polyclonal Low TS related to response Aschele et al127 108 5-FU/LV/MTX IHC TS106 Low TS related to response Paradiso et al128 27 5-FU/LV/MTX IHC polyclonal Low meta-TS related to response Aschele et al129 29 5-FU/LV/MITO or MITOX mRNA Low TS related to response Kornmann et al130 50 5-FU/LV/oxaliplatin mRNA High TS related to non-response Shirota et al131 134 5-FU/IFN alfa-2b IHC polyclonal No correlations found Findlay et al132 124 5-FU/LV; 5-FU/MTX IHC Low TS related to response 5FU/LV Aschele et al133 309 5FU/LV MRNA High TS related to longer survival Kornmann et al91 54 5-FU Activity Low TS related to 5-FU sensitivity Fujii et al68 54 5-FU/LV Activity Low TS related to 5-FU response Noordhuis et al134

Gastric 65 5-FU/LV/cisplatin mRNA Low TS related to longer survival Lenz et al135 38 5-FU/LV/cisplatin mRNA High TS related to non-response Metzger et al136 30 5-FU/LV/cisplatin IHC TS106 High TS related to non-response Yeh et al137 39 5-FU/cisplatin IHC Low TS related to response Boku et al138 38 5-FU/MTX IHC No correlations found Tahara et al94

Head and neck 70 5-FU/LV/cisplatin/MTX/IFN IHC TS106 High TS related to non-response Johnston et al139 82 5-FU/LV/cisplatin TS activity No correlations found Etienne et al96 52 5-FU TS activity No correlations found Etienne et al95

NSCLC 60 in vitro 5-FU sensitivity FdUMP binding TS not related to 5-FU sensitivity Higashiyama et al100 68 5-FU or UFTs IHC polyclonal Low TS related to longer survival Huang et al99

Breast 75 CAF regimen mRNA No correlations found Lizard-Nacol et al140 121 CAF or CMF regimen TS activity High TS related to better treatment Foekens et al84 Efficacy after failure on tamoxifen

Pancreatic 73 5-FU based TS mRNA High TS related to 5-FU response Hu et al141

MTX ¼ methotrexate; MITO ¼ mitomycine; MITOX ¼ mitoxantrone; IHC ¼ immunohistochemistry; TS106 ¼ monoclonal antibody against TS; IFN ¼ interferon. be more sensitive to 5-FU, the low proliferation rate prevents Acknowledging the importance of TS expression in colo- induction of DNA damage and 5-FU antitumour efficacy.142 rectal cancer, another determinant for 5-FU cytotoxicity may Secondly, complex interactions between oncogenes and TS be the enzyme deoxyuridine triphosphate (dUTP) nucleo- have been demonstrated, such as binding of TS protein to p53 tidohydrolase (dUTPase), which is the key regulator of dUTP and c-myc RNA and inhibition of TS promotor activity by pools. There are at least two isoforms of dUTPase, a nuclear wild-type p53. The inhibition of TS by 5-FU or antifolates may and a mitochondrial form, encoded by the same gene. In also decrease the TS-protein-mediated downregulation of TS tumour cell lines, increased dUTPase levels accounted for mRNA translation, resulting in increased TS protein synthesis. resistance to the 5-FU metabolite 50dFUR.144,145 In a small The clinical implications of these complex indirect and direct retrospective study in tumours of 20 metastatic colorectal interactions between oncogenes and TS are as yet unclear.143 cancer patients, high nuclear dUTPase protein expression was Finally, the assay type may have influenced the outcome associated with poor tumour response on 5-FU therapy.146 of some studies, since IHC as well as rt-PCR were used to However, larger studies are needed to establish the role of measure TS expression. IHC has been performed using dUTPase in 5-FU chemoresistance. different types of antibodies, including monoclonal TS106 as well as polyclonal antibodies. Differences in binding Impact of Somatic Mutations and Gene Expression Levels: specificity and selectivity may have had a decisive impact on dFdC and ara-C study outcome. Furthermore, the issue of infiltrating cells Resistance to nucleoside analogues can be due to a number disturbing tumour tissue homogeneity and thus troubling of factors, affecting drug metabolism, including increased PCR data, as mentioned earlier when interpreting TP data, deamination, loss of expression of activating kinases and may also hold for TS. increased activity of nucleotidases. In vitro, gene transfer of

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 236

CDD cDNA into murine fibroblast cells was shown to induce tumour contain amplified c-myc and wild-type p53 might cellular resistance to ara-C through increased deamina- benefit from 5-FU-based chemotherapy.181 Interestingly, tion.147 In patients with acute myeloid leukaemia (AML), there is an increasing evidence of interactions between TS pretreatment CDD activity did correlate with response on and p53. P53 gene mutation and protein overexpression are ara-C induction treatment in two studies, both involving 36 associated with increased TS mRNA levels and TS cytoplas- patients.148,149 Ara-C resistance has also been associated matic protein in colorectal tumours,124,128,182 It has been with low dCK activity both in vitro150,151 and in 21 AML demonstrated that wild-type p53 is able to bind TS mRNA, patients.152 In turn, increased dCK activity was associated which results in a feedback regulation of TS protein with increased activation of both compounds to cytotoxic synthesis.183 In cells with mutated p53, this feedback nucleoside triphosphate derivates. However, mutational mechanism fails and, as previously mentioned, overexpres- inactivation of dCK is not thought to confer resistance to sion of TS is associated with poor clinical outcome. ara-C in AML patients because mutations in the dCK gene are rarely found in refractory or relapsed AML pa- Impact of Current data on Clinical Practice tients.153,154 Alternatively, inactivation of dCK by the Summarising, we conclude that, in tumour material, none formation of alternatively spliced dCK transcripts has been of the enzymes discussed above can currently function as demonstrated in seven out of 12 patients with resistant AML predictive marker by itself. So far, only small studies have compared to one out of 10 patients with sensitive AML.155 been performed and drawing conclusions from combined This mechanism is probably a cause of therapy failure in data of these often statistically underpowered studies is patients with resistant AML.156 Finally, high 5NT expression fairly impossible due to differences in applied techniques or in blast cells was shown to be an independent prognostic patient groups. Clinical practice affecting studies are not factor for poor outcome in 108 AML patients after ara-C- available yet. Good predictive genetic markers or marker sets containing regimens.157 The balance between dCK and 5NT are lacking. The specificity of a predictive marker needs to be might predict drug toxicity, since strongly increased high, as it will be used to discriminate between treat- 5NT activity combined with reduced dCK activity was ment options. Probably, panels of genes will be needed observed in a human leukaemic cell line, resistant to ara-C to obtain an adequate specificity. To find such gene and dFdC.158 panels, far more studies are needed. Looking at current Apart from resistance to nucleoside analogues caused by data, a combination of the expression of DPD, TS, TP and aberrant drug metabolism, altered drug transport may cause OPRT, all four crucial enzymes in the metabolism of 5-FU, at decreased chemosensitivity. Recently, overexpression of the least appears to be a promising approach in colorectal human multidrug resistance protein 5 (MRP5), an ABC cancer68–70,184–186 (Table 5). transporter known to transport nucleotide monophosphates, has been associated with resistance to dFdC in Future Perspectives: The Polygenetic Approach vitro.159 MRP5-mediated efflux of dFdC metabolites may Despite the here summarised potential genes and proteins lower the accumulation of dFdC in cells. So far, other interfering with pyrimidine antagonist metabolism, only chemoresistance-related transport mechanisms have not few factors indeed have been proven to affect chemotherapy been identified for pyrimidine antagonists. efficacy and/or toxicity. Most consistent data are available regarding the role of TS, DPD and p53 in 5-FU chemother- Other Proteins Related to Pyrimidine Antagonist apy, and that of TP in capecitabine chemotherapy (see Chemosensitivity Tables 1 and 5). The impact of germline DPD deficiency on As mentioned in the previous sections, expression levels of 5-FU pharmacokinetics and the development of severe life- specific proteins, directly involved in drug metabolism of threatening toxicity is obvious. Whether or not to screen pyrimidine antagonists, may affect treatment efficacy. patients on DPD deficiency before starting chemotherapy is, However, expression of genes related to cell growth and however, an issue of debate.187–189 Since genetic abberations the apoptotic pathway may also affect chemosensitivity. in the DPYD gene explain less than half of all cases of This interaction of apoptosis regulator proteins with extreme 5-FU-related toxicity, costly screening will only be chemosensitivity is an intriguing issue, but an extensive partially preventive. Therefore, a solid cost–benefit analysis, analysis of current literature on this subject falls outside the preferentially embedded in a large prospective clinical trial, scope of this review. The most extensively studied genes will be needed to establish the added value of DPYD with regard to pyrimidine antagonist chemotherapy are bcl- genotyping. Screening (genotypically or phenotypically) 2, bax, c-myc and p53.160–168 Here, we will only briefly might become standard clinical practice as soon as a rapid, report on the putative role of p53 with respect to 5-FU sensitive and cheap test becomes available. efficacy, as this relationship has been most extensively Regarding 5-FU efficacy, some attempts have been made studied. Mutations in the p53 gene and p53 overexpression recently to identify putative markers of response in tumour have been associated with 5-FU chemoresistance both in material for use in patient treatment guidance.96,138,183,190 vitro169–171 and in vivo in colorectal,124,172–175 head and However, despite some promising studies, the overall data are neck,176,177 and breast cancer.178 However, results in some difficult to interpret and not always in line with previous other studies are less unequivocal.179,180 Recently, it has results in larger patient groups. This might, at least in part, be been suggested that only patients whose primary colorectal due to the complexity in transcription and translation of

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 237

Table 5 Overview of proteins that potentially may affect drug treatment outcome due to altered gene expression

Protein Drug Patient data Consequences

Dihydropyrimidine 5-FU Miscellaneous cancer High DPD activity in tumour tissue associated dehydrogenase (DPD) types with reduced 5-FU efficacy dUTP nucleotidohydrolase 5-FU and Colorectal cancer High dUTPase activity may be associated (dUTPase) analogues with reduced 5-FU efficacy

Thymidine kinase (TK) 5-FU and analogues Breast cancer Unknown, inconclusive data

Thymine phosphorylase 5-FU Colorectal cancer Unknown, inconclusive data (TP) ¼ endothelial cell growth Gastric cancer factor (ECGF1) Breast cancer

Thymidylate synthase (TS) 5-FU and Colorectal cancer High TS expression associated with reduced analogues 5-FU efficacy

Cytidine deaminase (CDD) Ara-C Acute myeloid leukaemia High CDD expression associated with (gemcitabine) reduced efficacy of ara-C

Deoxycytidine kinase (dCK) Ara-C Acute myeloid leukaemia Low dCK expression associated with (gemcitabine) resistance to ara-C

50-nucleotidase (5NT) Ara-C Acute myeloid leukaemia High 5NT expression associated with (gemcitabine) reduced efficacy of ara-C

some of the genes discussed above (eg TS). Additional of candidate genes to focus on in future research has been mechanistic studies are needed to explore this issue more in summarised in this review. On the other hand, further detail.Furthermore,asmentionedbefore,manyavailable developments in microarray techniques and proteomics studies are insufficiently powered to reach statistical signifi- may eventually lead to the identification of a subset of cance. Therefore, incorporation of pharmacogenomic issues relevant genes for a certain drug in a certain tumour type. in future phase II and III clinical trials should be encouraged. Few studies in cancer cell lines and cancer xenografts have Another cause for inconsistency in current data may be already shown the possibilities of using cDNA microarray found in the applied scientific methodology. A crucial factor profiling for identification of novel genes involved in in the search for predictive markers is the availability of a response or resistance to chemotherapy.192–195 Microarray standardised, accurate, precise and robust test. Method analysis of 2400 genes revealed five novel 5-FU-inducible validation, such as performed for the TS106 antibody for transcriptional target MCF-7 breast cancer cells (SSAT, annexin TS IHC, is a prerequisite for standardised testing.191 Subse- II,MAT-8,thymosinb-10 and chaperonin-10).193 In another quently, these quality-controlled diagnostics should be study, analysis of 9216 genes in a panel of 30 colon carcinoma tested not only retrospectively, but also in prospective cell lines revealed 420 genes that were correlated with 5-FU multi-centre (and multi-laboratory) clinical trials. Unfortu- response.194 Genes involved in DNA replication and repair nately, such tests are not available yet. (including MLH1, PCNA, replication factor C, nucleosome Also causing inconsistency in the outcome of monoge- assemby protein 1, origin recognition complexes and topoi- netic studies may be the fact that the impact of other genes somerase II) and protein processing and targeting (including related to response and survival is ignored in the final several chaperones, lectin mannose binding 1, heat shock calculations. Since cancer is a genetically heterogeneous 70 kDa protein 8, nucleophosmin and hypoxia upregulated 1) disease, the monogenetic approach is likely too simple. This were significantly enriched for expression on the 430 gene implies that far more research is needed to gain more insight list. Recently, Wang et al analysed gene expression in five pairs into the exact role of genetic polymorphisms, and gene of 5-FU-resistant and parental cancer cell lines with a cDNA expression levels on pyrimidine antagonist chemotherapy. microarray approach.195 They identified nuclear factor kappa The interaction between TS and p53, combined with the B (NFKappaB) p65 as one of the genes related to 5-FU impact of several TS polymorphisms, illustrates the com- resistance and demonstrated that transfection of NFKappaB in plexity of pharmacogenomic research. MCF-7 cells induced 5-FU resistance. The two most important approaches in current and near These studies show the potential power of these techni- future research to find predictive marker sets will be the ques for identifying predictive markers for drug sensitivity, candidate gene and the microarray strategies. A broad range as well as novel targets to overcome drug resistance.

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 238

In the near future, identification of relevant germline polymorphisms, combined with relevant mRNA expression MTHFR methylene tetrahydrofolate reductase MTX methotrexate levels in tumour tissue, might permit more effective, 5NT 50-nucleotidase individualised cancer chemotherapy. Ideally, a process of OPRT orotate phosphoribosyl transferase ‘chemotyping’, that is, choosing the right chemotherapy RNR ribonucleotide reductase regimen in the right dose, based on selected genotypical and SNP single-nucleotide polymorphism TK thymidine kinase phenotypical information, should precede drug prescribing. TP thymidine phosphorylase The interdisciplinary Pharmacogenetics Anticancer Agents TS thymidine synthase Research (PAAR) group is an example of an excellent UDPK uridine diphosphate kinase initiative to coordinate research on these issues.196 Even- UK uridine kinase UMPK uridine monophosphate kinase tually, well-controlled prospective clinical trials with ade- UP uridine phosphorylase quate sample size and statistical power will be needed to USF upstream stimulatory factor demonstrate the surplus value of such new concepts above current practice. REFERENCES DUALITY OF INTEREST 1 Harari PM. Why has induction chemotherapy of advanced head and None declared. neck cancer become a United States community standard of practice? J Clin Oncol 1997; 15: 2050–2055. 2 Macdonald JS, Astrow AB. Adjuvant therapy of colon cancer. Semin Oncol 2001; 28: 30–40. ABBREVIATIONS 3 Cochrane Review. Multi-agent chemotherapy for early breast cancer. AML acute myeloid leukaemia Cochrane Database Syst Rev 2002; CD000487. ara-C cytarabine 4 Wiernik PH. Current status of and future prospects for the medical ara-CMP cytarabine monophosphate management of adenocarcinoma of the exocrine pancreas. J Clin ara-U uracil-arabinoside Gastroenterol 2000; 30: 357–363. BSA body surface area 5 Sorenson S, Glimelius B, Nygren P, SBU group. Swedish Council of BUP1 b-ureidopropionase Technology Assessment in Health Care. A systematic overview of CDD cytidine deaminase chemotherapy effects in non-small cell lung cancer. Acta Oncol 2001; CMF cyclophosphamide, methotrexate and 5-fluorouracil 40: 327–339. CMP deoxycytidine monophosphate 6 Bishop JF. Approaches to induction therapy with adult acute myeloid CYP450 cytochrome P450 leukaemia. Acta Haematol 1998; 99: 133–137. CYP1A2 cytochrome P450 1A2 7 Pinedo HM, Peters GF. Fluorouracil: biochemistry and pharmacology. CYP2A6 cytochrome P450 2A6 J Clin Oncol 1988; 6: 1653–1664. CYP2C8 cytochrome P450 2C8 8 Diasio RB, Johnson MR. The role of pharmacogenetics and pharma- dCDP deoxycytidine diphosphate cogenomics in cancer chemotherapy with fluorouracil. Pharmacology dCK deoxycytidine kinase 2000; 61: 199–203. dCTP deoxycytidine triphosphate 9 Ho DH, Townsend L, Luna MA, Bodey GP. Distribution and inhibition dFCR 50-deoxy-5-fluorocytidine of dihydrouracil dehydrogenase activities in human tissues using dFdC gemcitabine 5-fluorouracil as a substrate. Anticancer Res 1986; 6: 781–784. dFdCDP gemcitabine diphosphate 10 Kuhn JG. Fluorouracil and the new oral fluorinated pyrimidines. Ann dFdCMP gemcitabine monophosphate Pharmacother 2001; 35: 217–227. dFdCTP gemcitabine triphosphate 11 Kufe DW, Spriggs DR. Biochemical and cellular pharmacology of dFdU 20-deoxy-20,20-difluorouridine cytosine arabinoside. Semin Oncol 1985; 12: 34–48. dFdUMP 20-deoxy-20,20-difluorouridine monophosphate 12 Plunkett W, Huang P, Xu YZ, Heinemann V, Grunewald R, Gandhi V. dFUR 50-deoxy-5-fluorouridine DFdC: metabolism, mechanisms of action, and self-potentiation. Sem DPD dihydropyrimidine dehydrogenase Oncol 1995; 22: 3–10. DPYD dihydropyrimidine dehydrogenase gene 13 Plunkett W, Huang P, Gandhi V. Preclinical characteristics of dFdC. DPYS dihydropyrimidinase Anticancer Drugs 1995; 6:7–13. dTMP deoxy-thymidine monophosphate 14 Giblett ER, Anderson JE, Chen SH, Teng YS, Cohen F. Uridine dUMP deoxyuridine monophosphate monophosphate kinase: a new genetic polymorphism with possible dUTP deoxy-uridine triphosphate clinical implications. Am J Hum Genet 1974; 26: 627–635. dUTPase dUTP nucleotidohydrolase 15 Komatsu N, Kimura Y, Kido A, Oya M. Polymorphism of uridine ECGF endothelial cell growth factor monophosphate kinase: population study in Japanese and phenotyp- FdUDP 5-fluoro-deoxy-uridine-diphosphate ing in blood stains. Int J Legal Med 1990; 104: 13–16. FdUMP 5-fluoro-deoxy-uridine monophosphate 16 Halasa J, Schlesinger D, Manczak M. UMPK polymorphism in the Polish FdUTP 5-fluoro-deoxy-uridine-triphosphate population. Arch Immunol Ther Exp 1985; 33: 621–624. 5-FU 5-fluorouracil 17 Gallango ML, Suinaga R. Uridine monophosphate polymorphism in FUDP 5-fluoro-uridine-diphosphate two Venezuelan populations. Am J Hum Genet 1978; 30: 215–218. FUDR 5-fluoro-deoxy-uridine 18 Suchi M, Mizuno H, Kawai Y, Tsuboi T, Sumi S, Okajima K et al. FUMP 5-fluoro-uridine-monophosphate Molecular cloning of the human UMP synthase gene and character- FUTP 5-fluoro-uridine-triphosphate ization of point mutations in two hereditary orotic aciduria families. Am IFN interferon J Hum Genet 1997; 60: 525–539. IHC immunohistochemistry 19 Marsh S, McLeod HL. Thymidylate synthase pharmacogenetics in LV leucovorin colorectal cancer. Clin Colorect Cancer 2001; 1: 175–178. MITO mitomycine 20 Kawakami K, Omura K, Kanehira E, Watanabe Y. Polymorphic tandem MITOX mitoxantrone repeats in the thymidylate synthase gene is associated with its protein MRP multi-drug resistance protein expression in human gastrointestinal cancers. Anticancer Res 1999; 19: 3249–3252.

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 239

21 Marsh S, McKay JA, Cassidy J, McLeod HL. Polymorphism in the on chromosome 18 affects tumor response and survival in individuals thymidylate synthase promoter enhancer region in colorectal cancer. heterozygous for a 28-bp polymorphism in the TS gene. Clin Cancer Int J Oncol 2001; 19: 383–386. Res 2004; 10: 433–439. 22 Kawakami K, Salonga D, Park JM, Danenberg KD, Uetake H, Brabender 39 Danenberg PV, Danenberg KD. Effect of 5, 10-methylenetetrahydro- J et al. Different lengths of a polymorphic repeat sequence in the folate on the dissociation of 5-fluoro-20-deoxyuridylate from thymidy- thymidylate synthase gene affect translational efficiency but not its late synthetase: evidence for an ordered mechanism. Biochemistry gene expression. Clin Cancer Res 2001; 7: 4096–4101. 1978; 17: 4018–4024. 23 Merkelbach-Bruse S, Hans V, Mathiak M, Sanguedolce R, Alessandro R, 40 Stern LL, Bagley PJ, Rosenberg IH, Selhub J. Conversion of 5-formylhy- Ruschoff J et al. Associations between polymorphisms in the drofolic acid to 5-methylhydrofolic acid in unimpaired folate-adequate thymidylate synthase gene, the expression of thymidylate synthase persons homozygous for the C677T mutation in the methylene mRNA and the microsatellite instability phenotype of colorectal tetrahydrofolate reductase gene. J Nutr 2000; 130: 2238–2242. cancer. Oncol Rep 2004; 11: 839–843. 41 Weisberg I, Tran P, Christensen B, Sibani S, Rozen R. A second genetic 24 Marsh S, Colli-Duguid ES, Li T, Liu X, McLeod HL. Ethnic variation in polymorphism in methylenetetrahydrofolate reductase (MTHFR) asso- the thymidylate synthase enhancer region polymorphism among ciated with decreased enzyme activity. Mol Genet Metab 1998; 64: Caucasians and Asian populations. Genomics 1999; 58: 310–312. 169–172. 25 Marsh S, Ameyaw MM, Githang’a J, Indalo A, Ofori-Adjei D, McLeod 42 Sohn KJ, Croxford R, Yates Z, Lucock M, Kim YI. Effect of the HL. Novel thymidylate synthase enhancer region alleles in African methylenetetrahydrofolate reductase C677T polymorphism on che- populations. Hum Mutat 2000; 16: 528. mosensitivity of colon and breast cancer cells to 5-fluorouracil and 26 Etienne MC, Ilc K, Formento JL, Laurent-Puig P, Formento P, methotrexate. J Natl Cancer Inst 2004; 96: 134–144. Cheradame S et al. Thymidylate synthase and methylenetetrahydro- 43 Wilcken B, Bamforth F, Li Z, Zhu H, Ritvanen A, Redlund M, Stoll C folate reductase gene polymorphisms: relationships with 5-fluorouracil et al. Geographical and ethnic variation of the 677C4T allele of 5,10 sensitivity. Br J Cancer 2004; 90: 526–534. methylenetetrahydrofolate reductase (MTHFR): findings from over 27 Mauritz R, Beumer IJ, Marsh S, McLeod HL, Smid K, van Groeningen CJ 7000 newborns from 16 areas world wide. J Med Genet 2003; 40: et al. Polymorphism of the thymidylate synthase gene and thymidylate 619–625. synthase levels in colon cancer cell lines and different tissues of 44 Wisotzkey JD, Toman J, Bell T, Monk JS, Jones D. Polymorphisms colorectal cancer patients. Nucleosides Nucleotides Nucleic Acids 2004; and stage III colon cancer: response to therapy. Mol Diagn 2000; 23: 1381–1384. 5: 5–6. 28 Villafranca E, Okruzhnov Y, Dominguez MA, Garcia-Foncillas J, 45 Kawakami K, Omura K, Kanehira E, Watanabe G. Methylene Azinovic I, Martinez E et al. Polymorphisms of the repeated sequences tetrahydrofolate reductase polymorphism is associated with in the enhancer region of the thymidylate synthase gene promoter folate pool in gastrointestinal tissue. Anticancer Res 2001; 21: may predict downstaging after preoperative chemoradiation in rectal 285–289. cancer. J Clin Oncol 2001; 19: 1779–1786. 46 Cohen V, Panet-Raymond V, Sabbaghian N, Morin I, Batist G, Rozen R. 29 Park DJ, Stoehlmacher J, Zhang W, Tsao-Wei D, Groschen S, Lenz HJ. Methylenetetrahydrofolate reductase polymorphism in advanced Thymidylate synthase gene polymorphism predicts response to colorectal cancer: a novel genomic predictor of clinical response capecitabine in advanced colorectal cancer. Int J Colorectal Dis 2002; to fluoropyrimidine-based chemotherapy. Clin Cancer Res 2003; 9: 17: 46–49. 1611–1615. 30 Iacopetta B, Grieu F, Joseph D, Elsaleh H. A polymorphism in the 47 Etienne MC, Formento JL, Chazal M, Francoual M, Magne N, enhancer region of the thymidylate synthase promoter influences the Formento P et al. Methylenetetrahydrofolate reductase gene survival of colorectal cancer patients treated with 5-fluorouracil. Br J polymorphisms and response to fluorouracil-based treatment in Cancer 2001; 85: 827–830. advanced colorectal cancer patients. Pharmacogenetics 2004; 14: 31 Tsuji T, Hidaka S, Sawai T, Nakagoe T, Yano H, Haseba M et al. 785–792. Polymorphism in the thymidylate synthase promoter enhancer region 48 Van Kuilenburg ABP, Haasjes J, Richel DJ, Zoetekouw L, Van Lenthe H, is not an efficacious marker for tumor sensitivity to 5-fluorouracil-based Waterham HR et al. Clinical implications of dihydropyrimidine oral adjuvant chemotherapy in colorectal cancer. Clin Cancer Res 2003; dehydrogenase (DPD) deficiency in patients with severe 5-fluorouracil 9: 3700–3704. associated toxicity. Identification of new mutations in the DPD gene. 32 Jonker DJ, Maroun JA, Le Francois B, Dahrouge S, Birnboim HC. Clin Cancer Res 2000; 6: 4705–4712. Correlation between 2 thymidylate synthase gene polymorphisms 49 Collie-Duguid ES, Etienne MC, Milano G, McLeod HL. Known variant and clinical outcomes in colorectal cancer patients receiving DPYD alleles do not explain DPD deficiency in cancer patients. adjuvant chemotherapy. Proc Am Soc Clin Oncol 2002; (abstract Pharmacogenetics 2000; 10: 217–223. 638); 21: 160a. 50 Van Kuilenburg ABP, De Abreu RA, Van Gennip AH. Pharmacogenetic 33 Kawakami K, Watanabe G. Identification and functional analysis of and clinical aspects of dihydropyrimidine dehydrogenase deficiency. single nucleotide polymorphism in the tandem repeat sequence of Ann Clin Biochem 2003; 40: 41–45. thymidylate synthase gene. Cancer Res 2003; 63: 6004–6007. 51 Diasio RB, Beavers TL, Carpenter JT. Familial deficiency of dihydropyr- 34 Mandola MV, Stoehlmacher J, Muller-Weeks S, Cesarone G, Yu MC, imidine dehydrogenase. Biochemical basis for familial pyrimidinemia Lenz HJ et al. A novel single nucleotide polymorphism within the 50 and severe 5-fluorouracil-induced toxicity. J Clin Invest 1988; 81: tandem repeat polymorphism of the thymidylate synthase gene 47–51. abolishes USF-1 binding and alters transcriptional activity. Cancer Res 52 Maring JG, van Kuilenburg ABP, Haasjes J, Piersma H, Groen HJM, Uges 2003; 63: 2898–2904. DRA et al. Reduced 5-FU clearance in a patient with low DPD activity 35 Ulrich CM, Bigler J, Velicer CM, Greene EA, Farin FM, Potter JD. due to heterozygosity for a mutant allele of the DPYD gene. Br J Cancer Searching expressed sequence tag databases: discovery and confirma- 2002; 86: 1028–1103. tion of a common polymorphism in the thymidylate synthase gene. 53 Lu Z, Zhang R, Diasio RB. Dihydropyrimidine dehydrogenase Cancer Epidemiol Biomarkers Prev 2000; 9: 1381–1385. activity in human peripheral blood mononuclear cells and liver: 36 Mandola MV, Stoehlmacher J, Zhang W, Groshen S, Yu MC, Iqbal S et population characteristics, newly identified deficient patients, and al. A 6 bp polymorphism in the thymidylate synthase gene causes clinical implication in 5-fluorouracil chemotherapy. Cancer Res 1993; message instability and is associated with decreased intratumoral TS 53: 5433–5438. mRNA levels. Pharmacogenetics 2004; 14: 319–327. 54 Etienne MC, Lagrange JL, Dassonville O, Fleming R, Thyss A, Renee N 37 Kawakami K, Ishida Y, Danenberg KD, Omura K, Watanabe G, et al. Population study of dihydropyrimidine dehydrogenase in cancer Danenberg PV. Functional polymorphism of the thymidylate synthase patients. J Clin Oncol 1994; 12: 2248–2253. gene in colorectal cancer accompanied by frequent loss of hetero- 55 Mattison LK, Ezzeldin H, Carpenter M, Modak A, Johnson MR, Diasio zygosity. Jpn J Cancer Res 2002; 93: 1221–1229. RB. Rapid identification of dihydropyrimidine dehydrogenase defi- 38 Uchida K, Hayashi K, Kawakami K, Schneider S, Yochim JM, Kuramochi ciency by using a novel 2-13C-uracil breath test. Clin Cancer Res 2004; H et al. Loss of heterozygosity at the thymidylate synthase (TS) locus 10: 2652–2658.

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 240

56 Maring JG, Oonk BN, De Vries EGE, Hospers GAP. Plasma pharmaco- patients with advanced and recurrent gastric cancer. Anticancer Res kinetics of uracil after an oral uracil challenge dose for dihydropyr- 2001; 21: 2149–2153. imidine dehydrogenase (DPD) phenotyping. Proc Am Soc Clin Oncol 75 Saito H, Tsujitani S, Oka S, Kondo A, Ikeguchi M, Maeta M et al. The 2004 (abstract 2120); 23: 157. expression of thymidine phosphorylase correlates with angiogenesis 57 Van Kuilenburg ABP, Meinsma R, Zonnenberg BA, Zoetekouw L, Baas and the efficacy of chemotherapy using fluorouracil derivatives in F, Madsuda K et al. Dihydropyrimidinase deficiency and severe 5- advanced gastric carcinoma. Br J Cancer 1999; 81: 484–489. fluorouracil toxicity. Clin Cancer Res 2003; 9: 4363–4367. 76 Marchetti S, Chazal M, Dubreuil A, Fischel JL, Etienne MC, Milano G. 58 Hamajima N, Kouwaki M, Vreken P, Matsuda K, Sumi S, Imaeda M et Impact of thymidine phosphorylase surexpression on fluoropyrimidine al. Dihydropyrimidinase deficiency: structural organization, chromo- activity and on tumour angiogenesis. Br J Cancer 2001; 85: 439–445. somal localization, and mutation analysis of the human dihydropyr- 77 Morita T, Matsuzaki A, Tokue A. Enhancement of sensitivity to imidinase gene. Am J Hum Genet 1998; 63: 717–726. capecitabine in human renal carcinoma cells transfected with 59 Daigo S, Takahashi Y, Fujieda M, Ariyoshi N, Yamazaki H, Koizumi W thymidine phosphorylase cDNA. Int J Cancer 2001; 92: 451–456. et al. A novel mutant allele of the CYP2A6 gene (CYP2A6*11) found 78 de Bruin M, van Capel T, Van der Born K, Kruyt FA, Fukushima M, in a cancer patient who showed poor metabolic phenotype towards Hoekman K et al. Role of platelet-derived endothelial cell growth tegafur. Pharmacogenetics 2002; 12: 299–306. factor/thymidine phosphorylase in fluoropyrimidine sensitivity. Br J 60 Kirch HC, Schroder JK, Hoppe H, Esche H, Seeber S, Schutte J. Cancer 2003; 88: 957–964. Recombinant gene products of two natural variants of the human 79 Evrard A, Cuq P, Robert B, Vian L, Pelegrin A, Cano JP. Enhancement of cytidine deaminase gene confer different deamination rates of 5-fluorouracil cytotoxicity by human thymidine-phosphorylase expres- cytarabine in vitro. Exp Hematol 1998; 26: 421–425. sion in cancer cells: in vitro and in vivo study. Int J Cancer 1999; 80: 61 Morita T, Matsuzaki A, Kurokawa S, Tokue A. Forced expression of 465–470. cytidine deaminase confers sensitivity to capecitabine. Oncology 2003; 80 Evrard A, Cuq P, Ciccolini J, Vian L, Cano JP. Increased cytotoxicity and 65: 267–274. bystander effect of 5-fluorouracil and 5-deoxy-5-fluorouridine in 62 Marinaki AM, Escuredo E, Duley JA, Simmonds HA, Amici A, Naponelli human colorectal cancer cells transfected with thymidine phosphory- V et al. Genetic basis of hemolytic anemia caused by 50-nucleotidase lase. Br J Cancer 1999; 80: 1726–1733. deficiency. Blood 2001; 97: 3327–3332. 81 Schwartz EL, Baptiste N, Wadler S, Makower D. Thymidine phosphor- 63 Banerjee D, Gorlick R, Sowers R, Miles JS, Rode W, Kemeny N ylase mediates the sensitivity of human colon carcinoma cells to 5- et al. Lack of uridine monophosphate kinase (UMPK) expression fluorouracil. J Biol Chem 1995; 270: 19073–19077. in tumour samples from colorectal cancer patients clinically resistant 82 van Triest B, Pinedo HM, Blaauwgeers JL, van Diest PJ, Schoenmakers to 5-fluorouracil. Proc Am Soc Clin Oncol 2002 (abstract 350); PS, Voorn DA et al. Prognostic role of thymidylate synthase, thymidine 21: 88a. phosphorylase/platelet-derived endothelial cell growth factor, and 64 Inaba M, Mitsuhashi J, Sawada H, Miike N, Naoe Y, Daimon A et al. proliferation markers in colorectal cancer. Clin Cancer Res 2000; 6: Reduced activity of anabolizing enzymes in 5-fluorouracil-resistant 1063–1072. human stomach cancer cells. Jpn J Cancer Res 1996; 87: 212–220. 83 Radparvar S, Houghton PJ, Germain G, Pennington J, Rahman A, 65 Inaba M, Naoe Y, Mitsuhashi J. Mechanisms of 5-fluorouracil resistance Houghton JA. Cellular pharmacology of 5-fluorouracil in a human in human colon cancer DLD-1 cells. Biol Pharm Bull 1998; 21: 569–573. colon adenocarcinoma cell line selected for thymidine kinase 66 Chung YM, Park SH, Park JK, Kim YT, Kang YK, Yoo YD. Establishment deficiency. Biochem Pharmacol 1990; 39: 1759–1765. and characterization of 5-fluorouracil-resistant gastric cancer cells. 84 Foekens JA, Romain S, Look MP, Martin PM, Klijn JG. Thymidine kinase Cancer Lett 2000; 159: 95–101. and thymidine synthase in advanced breast cancer: response to 67 Murakami Y, Kazuno H, Emura T, Tsujimoto H, Suzuki N, Fukushima tamoxifen and chemotherapy. Cancer Res 2001; 61: 1421–1425. M. Different mechanisms of acquired resistance to fluorinated 85 Beck A, Etienne MC, Cheradame S, Fischel JL, Formento P, Renee N pyrimidines in human colorectal cancer cells. Int J Oncol 2000; 17: et al. A role for dihydropyrimidine dehydrogenase and thymidylate 277–283. synthase in tumour sensitivity to fluorouracil. Eur J Cancer 1994; 30A: 68 Fujii R, Seshimo A, Kameoka S. Relationships between the expression 1517–1522. of thymidylate synthase, dihydropyrimidine dehydrogenase, and 86 Kirihara Y, Yamamoto W, Toge T, Nishiyama M. Dihydropyrimidine orotate phosphoribosyltransferase and cell proliferative activity and dehydrogenase, multidrug resistance-associated protein, and thymi- 5-fluorouracil sensitivity in colorectal carcinoma. Int J Clin Oncol 2003; dylate synthase gene expression levels can predict 5-fluorouracil 8: 72–78. resistance in human gastrointestinal cancer cells. Int J Oncol 1999; 69 Isshi K, Sakuyama T, Gen T, Nakamura Y, Kuroda T, Katuyama T et al. 14: 551–556. Predicting 5-FU sensitivity using human colorectal cancer specimens: 87 Ishikawa Y, Kubota T, Otani Y, Watanabe M, Teramoto T, Kumai K comparison of tumor dihydropyrimidine dehydrogenase and orotate et al. Dihydropyrimidine dehydrogenase activity and messenger RNA phosphoribosyl transferase activities with in vitro chemosensitivity to level may be related to the antitumor effect of 5-fluorouracil on human 5-FU. Int J Clin Oncol 2002; 7: 335–342. tumor xenografts in nude mice. Clin Cancer Res 1999; 5: 883–889. 70 Ichikawa W, Uetake H, Shirota Y, Yamada H, Takahashi T, Nihei Z et al. 88 Salonga D, Danenberg KD, Johnson M, Metzger R, Groshen S, Tsao- Both gene expression for orotate phosphoribosyltransferase and its Wei DD et al. Colorectal tumors responding to 5-fluorouracil have low ratio to dihydropyrimidine dehydrogenase influence outcome follow- gene expression levels of dihydropyrimidine dehydrogenase, thymi- ing fluoropyrimidine-based chemotherapy for metastatic colorectal dylate synthase, and thymidine phosphorylase. Clin Cancer Res 2000; cancer. Br J Cancer 2003; 89: 1486–1492. 6: 1322–1327. 71 Cuq P, Rouquet C, Evrard A, Ciccolini J, Vian L, Cano JP. 89 Ikeguchi M, Makino M, Kaibara N. Thymidine phosphorylase and Fluoropyrimidine sensitivity of human MCF-7 breast cancer cells stably dihydropyrimidine dehydrogenase activity in colorectal carcinoma and transfected with human uridine phosphorylase. Br J Cancer 2001; 84: patients prognosis. Langenbecks Arch Surg 2002; 387: 240–245. 1677–1680. 90 Kornmann M, Link KH, Galuba I, Ott K, Schwabe W, Hausler P et al. 72 Fox SB, Engels K, Comley M, Whitehouse RM, Turley H, Gatter KC Association of time to recurrence with thymidylate synthase and et al. Relationship of elevated tumour thymidine phosphorylase in dihydropyrimidine dehydrogenase mRNA expression in stage II and III node-positive breast carcinomas to the effects of adjuvant CMF. Ann colorectal cancer. J Gastrointest Surg 2002; 6: 331–337. Oncol 1997; 8: 271–275. 91 Kornmann M, Schwabe W, Sander S, Kron M, Strater J, Polat S et al. 73 Metzger R, Danenberg K, Leichman CG, Salonga D, Schwartz EL, Thymidylate synthase and dihydropyrimidine dehydrogenase mRNA Wadler S et al. High basal level gene expression of thymidine expression levels: predictors for survival in colorectal cancer patients phosphorylase (platelet-derived endothelial cell growth factor) in receiving adjuvant 5-fluorouracil. Clin Cancer Res 2003; 9: 4116–4124. colorectal tumors is associated with nonreponse to 5-fluorouracil. Clin 92 Terashima M, Irinoda T, Fujiwara H, Nakaya T, Takagane A, Abe K et al. Cancer Res 1998; 4: 2371–2376. Roles of thymidylate synthase and dihydropyrimidine dehydrogenase 74 Kikuyama S, Inada T, Shimizu K, Miyakita M, Ogata Y. P53, bcl-2 and in tumor progression and sensitivity to 5-fluorouracil in human gastric thymidine phosphorylase as predictive markers of chemotherapy in cancer. Anticancer Res 2002; 22: 761–768.

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 241

93 Terashima M, Fujiwara H, Takagane A, Abe K, Araya M, Irinoda T et al. dihydropyrimidine dehydrogenase (DPD) enzyme activities and DPD Role of thymidine phosphorylase and dihydropyrimidine dehydrogen- messenger RNA levels. Clin Cancer Res 2001; 7: 2783–2787. ase in tumour progression and sensitivity to doxifluridine in gastric 111 Fujiwaki R, Hata K, Nakayama K, Fukumoto M, Miyazaki K. Gene cancer patients. Eur J Cancer 2002; 38: 2375–2381. expression for dihydropyrimidine dehydrogenase and thymidine 94 Tahara M, Ochiai A, Fujimoto J, Boku N, Yasui W, Ohtsu A et al. phosphorylase influences outcome in epithelial ovarian cancer. J Clin Expression of thymidylate synthase, thymidine phosphorylase, dihy- Oncol 2000; 18: 3946–3951. dropyrimidine dehydrogenase, E2F-1, Bak, Bcl-X, and Bcl-2, and 112 Anan K, Mitsuyama S, Tamae K, Suehara N, Nishihara K, Ogawa Y et clinical outcomes for gastric cancer patients treated with bolus al. Increased dihydropyrimidine dehydrogenase activity in breast 5-fluorouracil. Oncol Rep 2004; 11:9–15. cancer. J Surg Oncol 2003; 82: 174–179. 95 Etienne MC, Cheradame S, Fischel JL, Formento P, Dassonville O, 113 Ishikawa T, Sekiguchi F, Fukase Y, Sawada N, Ishitsuka H. Positive Renee N et al. Response to fluorouracil therapy in cancer patients: the correlation between the efficacy of capecitabine and doxifluridine and role of tumoral dihydropyrimidine dehydrogenase activity. J Clin Oncol the ratio of thymidine phosphorylase to dihydropyrimidine dehydro- 1995; 13: 1663–1670. genase activities in tumors in human cancer xenografts. Cancer Res 96 Etienne MC, Pivot X, Formento JL, Bensadoun RJ, Formento P, 1998; 58: 685–690. Dassonville O et al. A multifactorial approach including tumoural 114 Nishimura G, Terada I, Kobayashi T, Ninomiya I, Katigawa H, Fushida S epidermal growth factor receptor, p53, thymidylate synthase and et al. Thymidine phosphorylase and dihydropyrimidine dehydrogen- dihydropyrimidine dehydrogenase to predict treatment outcome in ase levels in primary colorectal cancer show a relationship to clinical head and neck cancer patients receiving 5-fluorouracil. Br J Cancer effects of 50-deoxy-5-fluorouridine as adjuvant chemotherapy. Oncol 1999; 79: 1864–1869. Rep 2002; 9: 479–482. 97 Kawasaki G, Yoshitomi I, Yanamoto S, Mizuno A. Thymidylate synthase 115 Nishina T, Hyodo I, Moriwaki T, Endo S, Hirasaki S, Doi T et al. The and dihydropyrimidine dehydrogenase expression in oral squamous ratio of thymidine phosphorylase (TP) to dihydropyrimidine dehy- cell carcinoma: an immunohistochemical and clinicopathologic study. drogenase (DPD) levels in tumour tissues is predictive of response to Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002; 94: 717–723. 50-deoxy-5-fluorouridine (an intermediate metabolite of capecitabine): 98 Nakagawa T, Tanaka F, Takata T, Matsuoka K, Miyahara R, Otake Y et a prospective study in patients with metastatic gastric cancer. Proc Am al. Predictive value of dihydropyrimidine dehydrogenase expression in Soc Clin Oncol 2002 (abstract 602); 21: 151a. tumor tissue, regarding the efficacy of postoperatively administered 116 Tominaga T, Toi M, Ohashi Y, Abe O, on behalf of the 5 0 -BC study UFT (Tegafur + Uracil) in patients with p-stage I nonsmall-cell lung group. Prognostic and predictive value of thymidine phosphorylase cancer. J Surg Oncol 2002; 81: 87–92. activity in early-stage breast cancer patients. Clin Breast Cancer 2002; 99 Huang CL, Yokomise H, Kobayashi S, Fukushima M, Hitomi S, Wada H. 3: 55–64. Intratumoral expression of thymidylate synthase and dihydropyrimi- 117 Komatsu T, Yamazaki H, Shimada N, Nakajima M, Yokoi T. Roles of dine dehydrogenase in non-small cell lung cancer patients treated with cytochromes P450 1A2, 2A6 and 2C8 in 5-fluorouracil formation from 5-FU based chemotherapy. Int J Oncol 2000; 17: 47–54. tegafur, an anticancer prodrug, in human liver microsomes. Drug 100 Higashiyama M, Kodama K, Yokouchi H, Takami K, Fukushima M, Metab Dispos 2000; 28: 1457–1463. Minamigawa K et al. Thymidylate synthase and dihydropyrimidine 118 van Triest B, Pinedo HM, van Hensbergen Y, Smid K, Telleman F, dehydrogenase activities in non-small cell lung cancer tissues: Schoenmakers PS et al. Thymidylate synthase level as the main relationship with in vitro sensitivity to 5-fluorouracil. Lung Cancer predictive parameter for sensitivity to 5-fluorouracil, but not for folate- 2001; 34: 407–416. based thymidylate synthase inhibitors, in 13 nonselected colon cancer 101 Horiguchi J, Yoshida T, Koibuchi Y, Iijima K, Ninomiya J, Takei H et al. cell lines. Clin Cancer Res 1999; 5: 643–654. DPD activity and immunohistochemical DPD expression in human 119 Fukushima M, Fujioka A, Uchida J, Nakagawa F, Takechi T. breast cancer. Oncol Rep 2004; 11: 65–72. Thymidylate synthase (TS) and ribonucleotide reductase (RNR) may 102 Mizutani Y, Wada H, Fukushima M, Yoshida O, Ukimura O, Kawauchi be involved in acquired resistance to 5-fluorouracil (5-FU) in human A et al. The significance of dihydropyrimidine dehydrogenase (DPD) cancer xenografts in vivo. Eur J Cancer 2001; 37: 1681–1687. activity in bladder cancer. Eur J Cancer 2001; 37: 569–575. 120 Copur S, Aiba K, Drake JC, Allegra CJ, Chu E. Thymidylate synthase 103 Uetake H, Ichikawa W, Takechi T, Fukushima M, Nihei Z, Sugihara K. gene amplification in human colon cancer cell lines resistant to Relationship between intratumoral dihydropyrimidine dehydrogenase 5-fluorouracil. Biochem Pharmacol 1995; 49: 1419–1426. activity and gene expression in human colorectal cancer. Clin Cancer 121 Matsuoka K, Tsukuda K, Suda M, Kobayashi K, Ota T, Okita A et al. The Res 1999; 5: 2836–2839. transfection of thymidylate synthase antisense suppresses oncogenic 104 Ishikawa Y, Kubota T, Otani Y, Watanabe M, Teramoto T, Kumai K et properties of a human colon cancer cell line and augments the al. Thymidylate synthetase and dihydropyrimidine dehydrogenase antitumor effect of fluorouracil. Int J Oncol 2004; 24: 217–222. levels in gastric cancer. Anticancer Res 1999; 19: 5635–5640. 122 Peters GJ, Van der Wilt CL, Van Groeningen CJ, Smid K, Meijer S, 105 Johnston SJ, Ridge SA, Cassidy J, McLeod HL. Regulation of Pinedo HM. Thymidylate synthase inhibition after administration of dihydropyrimidine dehydrogenase in colorectal cancer. Clin Cancer fluorouracil with or without leucovorin in colon cancer patients: Res 1999; 5: 2566–2570. implications for treatment with fluorouracil. J Clin Oncol 1994; 12: 106 Collie-Duguid ES, Johnston SJ, Boyce L, Smith N, Cowieson A, Cassidy 2035–2042. J et al. Thymidine phosphorylase and dihydropyrimidine dehydrogen- 123 Leichman CG, Lenz HJ, Leichman L, Danenberg K, Baranda J, Groshen ase protein expression in colorectal cancer. Int J Cancer 2001; 94: S et al. Quantitation of intratumoral thymidylate synthase expression 297–301. predicts for disseminated colorectal cancer response and resistance to 107 Hiroyasu S, Shiraishi M, Samura H, Tokashiki H, Shimoji H, Isa T et al. protracted-infusion fluorouracil and weekly leucovorin. J Clin Oncol Clinical relevance of the concentrations of both pyrimidine nucleoside 1997; 15: 3223–3229. phosphorylase (PyNPase) and dihydropyrimidine dehydrogenase 124 Lenz HJ, Hayashi K, Salonga D, Danenberg KD, Danenberg PV, (DPD) in colorectal cancer. Jpn J Clin Oncol 2001; 31: 65–68. Metzger R et al. P53 point mutations and thymidylate synthase 108 Miyamoto S, Ochiai A, Boku N, Ohtsu A, Tahara M, Yoshida S et al. messenger RNA levels in disseminated colorectal cancer: an analysis of Discrepancies between the gene expression, protein expression, and response and survival. Clin Cancer Res 1998; 4: 1243–1250. enzymatic activity of thymidylate synthase and dihydropyrimidine 125 Cascinu S, Aschele C, Barni S, Debernardis D, Baldo C, Tunesi G et al. dehydrogenase in human gastrointestinal cancers and adjacent Thymidylate synthase protein expression in advanced colon cancer: normal mucosa. Int J Oncol 2001; 18: 705–713. correlation with the site of metastasis and the clinical response to 109 Hotta T, Taniguchi K, Kobayashi Y, Johata K, Sahara M, Naka T et al. leucovorin-modulated bolus 5-fluorouracil. Clin Cancer Res 1999; 5: Comparative analysis of thymidine phosphorylase and dihydropyrimi- 1996–1999. dine dehydrogenase expression in gastric and colorectal cancers. 126 Wong NACS, Brett L, Stewart M, Leitch A, Longley DB, Dunlop MG Oncol Rep 2004; 11: 1045–1051. et al. Nuclear thymidylate synthase expression, p53 expression 110 Tanaka-Nozaki M, Onda M, Tanaka N, Kato S. Variations in and 5FU response in colorectal carcinoma. Br J Cancer 2001; 85: 5-fluorouracil concentrations of colorectal tissues as compared with 1937–1943.

www.nature.com/tpj Pyrimidine antagonist pharmacogenetics JG Maring et al 242

127 Aschele C, Debernardis D, Casazza S, Antonelli G, Tunesi G, Baldo C tumour cells by 5-fluorodeoxyuridine: implications for mechanisms et al. Immunohistochemical quantitation of thymidylate synthase of resistance and cytotoxicity. Proc Natl Acad Sci USA 1992; 89: expression in colorectal cancer metastases predicts for clinical 10474–10478. outcome to fluorouracil based chemotherapy. J Clin Oncol 1999; 17: 145 Canman CE, Radany EH, Parsels LA, Davis MA, Lawrence TS, Maybaum 1760–1770. J. Induction of resistance to fluorodeoxyuridine cytotoxicity and DNA 128 Paradiso A, Simone G, Petroni S, Leone B, Vallejo C, Lacava J et al. damage in human tumour cells by expression of Escherichia coli Thymidylate synthase and p53 primary tumour expression as deoxyuridinetriphosphatase. Cancer Res 1994; 54: 2296–2298. predictive factors for advanced colorectal cancer patients. Br J Cancer 146 Ladner RD, Lynch FJ, Groshen S, Xiong YP, Sherrod A, Caradonna SJ et 2000; 82: 560–567. al. dUPT nucleotidohydrolase isoform expression in normal and 129 Aschele C, Debernardis D, Tunesi G, Maley F, Sobrero A. Thymidylate neoplasmic tissues: association with survival and response to synthase protein expression in primary colorectal cancer compared 5-fluorouracil in colorectal cancer. Cancer Res 2000; 60: 3493–3503. with the corresponding distant metastases and relationship with 147 Schroder JK, Kirch C, Flasshove M, Kalweit H, Seidelmann M, Hilger R clinical response to 5-fluorouracil. Clin Cancer Res 2000; 6: 4797–4802. et al. Constitutive overexpression of the cytidine deaminase gene 130 Kornmann M, Link KH, Lenz HJ, Pillasch J, Metzger R, Butzer U et al. confers resistance to cytosine arabinoside in vitro. Leukemia 1996; 10: Thymidylate synthase is a predictor for response and resistance in 1919–1924. hepatic artery infusion therapy. Cancer Lett 1997; 118: 29–35. 148 Jahns-streubel G, Reuter C, Auf der Landwehr U, Unterhalt M, Schleyer 131 Shirota Y, Stoehlmacher J, Brabender J, Xiong YP, Uetake H, E, Wormann B et al. Activity of thymidine kinase and of polymerase Danenberg KD et al. EECC1 and thymidylate synthase mRNA levels alpha as well as activity and gene expression of deoxycytidine predict survival for colorectal cancer patients receiving combination deaminase in leukemic blasts are correlated with clinical response in oxaliplatin and fluorouracil chemotherapy. J Clin Oncol 2001; 19: the setting of granulocyte-macrophage colony-stimulating factor- 4298–4304. based priming before and during TAD-9 induction therapy in acute 132 Findley MP, Cunningham D, Morgan G, Clinton S, Hardcastle A, myeloid leukemia. Blood 1998; 90: 1968–1976. Aherne GW. Lack of correlation between thymidylate synthase levels in 149 Schroder JK, Kirch C, Seeber S, Schutte J. Structural and functional primary colorectal tumours and subsequent response to chemother- analysis of the cytidine deaminase gene in patients with acute myeloid apy. Br J Cancer 1997; 75: 903–909. leukaemia. Br J Haematol 1998; 103: 1096–1103. 133 Aschele C, Debernardis D, Bandelloni R, Cascinu S, Catalano V, 150 Owens JK, Shewach DS, Ullman B, Mitchell BS. Resistance to 1-beta-D- Giordani P et al. Thymidylate synthase protein expression in colorectal arabinofuranosylcytosine in human T-lymphoblasts mediated by cancer metastases predicts for clinical outcome to leucovorin- mutations within the deoxycytidine kinase gene. Cancer Res 1992; modulated bolus or infusional 5-fluorouracil but not methotrexate- 52: 2389–2393. modulated bolus 5-fluorouracil. Ann Oncol 2002; 13: 1882–1892. 151 Kobayashi T, Kakihara T, Uchiyama M, Fukuda T, Kishi K, Shibata A. 134 Noordhuis P, Holwerda U, Van Der Wilt CL, Van Groeningen CJ, Smid Low expression of the deoxycytidine kinase (dCK) gene in a 1-beta-D- K, Meijer S et al. 5-Fluorouracil incorporation into RNA and DNA in arabinofuranosylcytosine-resistant human leukemic cell line KY-Ra. relation to thymidylate synthase inhibition of human colorectal Leuk Lymphoma 1994; 15: 503–505. cancers. Ann Oncol 2004; 15: 1025–1032. 152 Colly LP, Peters WG, Richel D, Arentsen-Honders MW, Starrenburg 135 Lenz HJ, Leichman CG, Danenberg KD, Danenberg PV, Groshen S, CW, Willemze R. Deoxycytidine kinase and deoxycytidine deaminase Cohen H et al. Thymidylate synthase mRNA level in adenocarcinoma values correspond closely to clinical response to cytosine arabinoside of the stomach: a predictor for primary tumour response and overall remission induction therapy in patients with acute myelogenous survival. J Clin Oncol 1996; 14: 176–178. leukemia. Semin Oncol 1987; 14: 257–261. 136 Metzger R, Leichman CG, Danenberg KD, Danenberg PV, Lenz HJ, 153 Flasshove M, Strumberg D, Ayscue L, Mitchell BS, Tirier C, Heit W et al. Hayashi K et al. ERCC1 mRNA levels complement thymidylate synthase Structural analysis of the deoxycytidine kinase gene in patients with mRNA levels in predicting response and survival for gastric cancer acute myeloid leukemia and resistance to cytosine arabinoside. patients receiving combination cisplatin and fluorouracil chemother- Leukemia 1994; 8: 780–785. apy. J Clin Oncol 1998; 16: 309–316. 154 Van den Heuvel-Eibrink MM, Wiemer EAC, Kuijpers M, Pieters R, 137 Yeh KH, Shun CT, Chen CL, Lin JT, Lee WJ, Lee PH et al. High Sonneveld P. Absence of mutations in the deoxycytidine kinase (dCK) expression of thymidylate synthase is associated with the drug gene in patients with relapsed and/or refractory acute myeloid resistance of gastric carcinoma to high dose 5-fluorouracil-based leukemia (AML). Leukemia 2001; 15: 855–867. systemic chemotherapy. Cancer 1998; 82: 1626–1631. 155 Veuger MJ, Honders MW, Landegent JE, Willemze R, Barge RM. 138 Boku N, Chin K, Hosokawa K, Ohtsu A, Tajiri H, Yoshida S et al. High incidence of alternatively spliced forms of deoxycytidine kinase Biological markers as a predictor for response and prognosis of in patients with resistant acute myeloid leukemia. Blood 2000; 96: unresectable gastric cancer patients treated with 5-fluorouracil and cis- 1517–1524. platinum. Clin Cancer Res 1998; 4: 1469–1474. 156 Veuger MJT, Heemskerk MHM, Honders W, Willemze R, Berge RMY. 139 Johnston PG, Mick R, Recant W, Behan KA, Dolan ME, Ratain MJ et al. Functional role of alternatively spliced deoxycytidine kinase in Thymidylate synthase expression and response to neoadjuvant sensitivity to cytarabine of acute myeloid leukemic cells. Blood 2002; chemotherapy in patients with advanced head and neck cancer. 99: 1373–1380. J Natl Cancer Inst 1997; 89: 308–313. 157 Galmarini CM, Graham K, Thomas X, Calvo F, Rousselot P, El Jafaari A 140 Lizard-Nacol S, Genne P, Coudert B, Riedinger JM, Arnal M, Sancy C et et al. Expression of high Km 50-nucleotidase in leukemic blasts is an al. MRD1 and thymidylate synthase (TS) gene expressions in advanced independent prognostic factor in adults with acute myeloid leukemia. breast cancer: relationship to drug exposure, p53 mutations, and Blood 2001; 98: 1922–1926. clinical outcome in the patients. Anticancer Res 1999; 19: 3575–3582. 158 Dumontet C, Fabianowska-Majewska K, Mantincic D, Callet Bauchu E, 141 Hu YC, Komorowski RA, Graewin S, Hostetter G, Kallioniemi OP, Pitt Tigaud I, Gandhi V et al. Common resistance mechanisms to HA et al. Thymidylate synthase expression predicts the response to 5- deoxynucleoside analogues in variants of the human erythroleukaemic fluorouracil-based adjuvant therapy in pancreatic cancer. Clin Cancer line K562. Br J Haematol 1999; 106: 78–85. Res 2003; 9: 4165–4171. 159 Davidson JD, Ma L, Iverson PW, Lesoon A, Jin S, Horwitz L et al. Human 142 Peters GJ, Backus HH, Freemantle S, van Triest B, Codacci-Pisanelli G, multidrug resistance protein 5 (MRP5) confers resistance to dFdC. Proc van der Wilt CL et al. Induction of thymidylate synthase as a 5- AACR 2002 (abstract 3868); 43: 780. fluorouracil resistance mechanism. Biochim Biophys Acta 2002; 1587: 160 Mirjolet JF, Barberi-Heyob M, Didelot C, Peyrat JP, Abecassis J, 194–205. Millon R et al. Bcl-2/Bax protein ratio predicts 5-fluorouracil sensitivity 143 Van Triest B, Pinedo HM, Giaccone G, Peters GJ. Downstream independently of p53 status. Br J Cancer 2000; 83: 1380–1386. molecular determinants of response to 5-fluorouracil and antifolate 161 Yang QF, Sakurai T, Yoshimura G, Shan L, Suzuma T, Tamaki T et al. thymidylate synthase inhibitors. Ann Oncol 2000; 11: 385–391. Expression of Bcl-2 but not Bax or p53 correlates with in vitro 144 Canman CE, Tang HY, Normolle DP, Lawrence TS, Maybaum J. resistance to a series of anticancer drugs in breast carcinoma. Breast Variations in patterns of DNA damage induced in human colorectal Cancer Res Treat 2000; 61: 211–216.

The Pharmacogenomics Journal Pyrimidine antagonist pharmacogenetics JG Maring et al 243

162 Keith FJ, Bradbury DA, Zhu YM, Russell NH. Inhibition of bcl-2 with 180 Arango D, Corner GA, Wadler S, Catalano PJ Augenlicht LH. C-myc/ antisense oligonucleotides induces apoptosis and increases the p53 interaction determines sensitivity of human colon carcinoma cells sensitivity of AML blasts to Ara-C. Leukemia 1995; 9: 131–138. to 5-fluorouracil in vitro and in vivo. Cancer Res 2001; 61: 4910–4915. 163 Ahmed N, Sammons J, Hassan HT. Bcl-2 protein in human myeloid 181 Lenz HJ, Danenberg KD, Leichman CG, Florentine B, Johnston PG, leukaemia cells and its down-regulation during chemotherapy- Groshen S et al. P53 and thymidylate synthase expression in untreated induced apoptosis. Oncol Rep 1999; 6: 403–407. stage II colon cancer: association with recurrence, survival and site. 164 Balkham SE, Sargent JM, Elgie AW, Williamson CJ, Taylor CG. Clin Cancer Res 1998; 4: 1227–1234. Comparison of BCL-2 and BAX protein expression with in vitro 182 Peters GJ, Van Triest B, Backus HHJ, Kuiper CM, Van der Wilt CL, sensitivity to ARA-C and 6TG in AML. Adv Exp Med Biol 1999; 457: Pinedo HM. Molecular downstream events and induction of thymi- 335–340. dylate synthase in mutant and wild-type p53 colon cancer cell lines 165 Bold RJ, Chandra J, McConkey DJ. DFdC-induced programmed cell after treatment with 5-fluorouracil and the thymidylate synthase death (apoptosis) of human pancreatic carcinoma is determined by inhibitor raltitrexed. Eur J Cancer 2000; 36: 916–924. Bcl-2 content. Ann Surg Oncol 1999; 6: 279–285. 183 McKay RA, Lloret C, Murray GI, Johnston PG, Bicknell R, Ahmed FY et 166 Van Slooten HJ, Clahsen PC, van Dierendonck JH, Duval C, Pallud C, al. Application of the enrichment approach to identify putative Mandard AM et al. Expression of Bcl-2 in node-negative breast cancer markers of reponse to 5-fluorouracil therapy in advanced colorectal is associated with various prognostic factors, but does not predict carcinomas. Int J Oncol 2000; 17: 153–158. response to one course of perioperative chemotherapy. Br J Cancer 184 Ichikawa W, Uetake H, Shirota Y, Yamada H, Nishi N, Nihei Z et al. 1996; 74: 78–85. Combination of dihydropyrimidine dehydrogenase and thymidylate 167 Schneider HJ, Sampson SA, Cunningham D, Norman AR, Andreyev HJ, synthase gene expressions in primary tumors as predictive parameters Tilsed JV et al. Bcl-2 expression and response to chemotherapy in for the efficacy of fluoropyrimidine-based chemotherapy for meta- colorectal adenocarcinomas. Br J Cancer 1997; 75: 427–431. static colorectal cancer. Clin Cancer Res 2003; 9: 786–791. 168 Nabeya Y, Loganzo Jr F, Maslak P, Lai L, de Oliveira AR, Schwartz GK 185 Ichikawa W, Takahashi T, Suto K, Hirayama R. Gene expressions for et al. The mutational status of p53 protein in gastric and esophageal thymidylate synthase (TS), orotate phosphoribosyl transferase (OPRT), adenocarcinoma cell lines predicts sensitivity to chemotherapeutic and thymidine phosphorylase (TP), not dihydropyridimine dehydro- agents. Int J Cancer 1995; 64: 37–46. genase (DPD), influence outcome of patients treated with S-1 for 169 Zheng M, Wang H, Zhang H, Ou Q, Shen B, Li N et al. The influence of gastric cancer. Proc Am Soc Clin Oncol 2004 (abstract 4050); 23: 325. the p53 gene on the in vitro chemosensitivity of colorectal cancer cells. 186 Smorenburg CH, Peters GJ, Van Groeningen CJ, Noordhuis P, Van Riel J Cancer Res Clin Oncol 1999; 125: 357–360. JM, Pinedo HM et al. First line tailored chemotherapy in advanced 170 Petak I, Tillman DM, Houghton JA. P53 dependence of Fas induction colorectal cancer with 5-fluorouracil/leucovorin or oxaliplatin/irinote- and acute apoptosis in response to 5-fluorouracil-leucovorin in human can chosen by the expression of thymidylate synthase (TS) and colon carcinoma cell lines. Clin Cancer Res 2000; 6: 4432–4441. dihydropyrimidine dehydrogenase (DPD). Proc Am Soc Clin Oncol 171 Benhattar J, Cerottini JP, Saraga E, Metthez G, Givel JC. P53 mutations 2004 (abstract 3624); 23: 276. as a possible predictor of response to chemotherapy in metastatic 187 Raida M, Schwabe W, Hausler P, Van Kuilenburg ABP, Van Gennip AH, colorectal carcinomas. Int J Cancer 1996; 69: 190–192. Behnke D et al. Prevalence of a common point mutation in the 172 Ahnen DJ, Feigl P, Quan G, Fenoglio-Preiser C, Lovato LC, Bunn PA dihydropyrimidine dehydrogenase (DPD) gene within the 50-splice et al. Ki-ras mutation and p53 overexpression predict the clinical donor site of intron 14 in patients with severe 5-fluorouracil (5-FU)- behavior of colorectal cancer: a Southwest Oncology Group Study. related toxicity compared with controls. Clin Cancer Res 2001; 7: Cancer Res 1998; 58: 1149–1158. 2832–2839. 173 Backus HHJ, Van Riel JMGH, Van Groeningen CJ, Vos W, Dukers DF, 188 Innocenti F, Ratain MJ. Correspondence regarding: Raida M et al. Bloemena E et al. Rb, mcl-1 and p53 expression correlate with clinical Prevalence of a common point mutation in the dihydropyrimidine outcome in patients with liver metastases from colorectal cancer. Ann dehydrogenase (DPD) gene within the 50-splice donor site of intron 14 Oncol 2001; 12: 779–785. in patients with severe 5-fluorouracil (5-FU)-related toxicity compared 174 Liang JT, Huang KC, Cheng YM, Hsu HC, Cheng AL, Hsu CH et al. P53 with controls. Clin Cancer Res 2002; 8: 1314. overexpression predicts poor chemosensitivity to high-dose 5-fluoro- 189 Behnke D, Raida M, Kliche KO, Pichlmeier U. Reply to the letter by uracil plus leucovorin chemotherapy for stage IV colorectal cancers Innocenti and Ratain. Clin Cancer Res 2002; 8: 1315–1316. after palliative bowel resection. Int J Cancer 2002; 97: 451–457. 190 Leichman CG, Lyman GH, Remedios P, Leichman L, Litwin A, Loud P 175 Cabelguenne A, Blons H, de Waziers I, Carnot F, Houllier AM, Soussi T et al. Molecular biologic correlates with continuous infusion 5-FU et al. P53 alterations predict tumour response to neoadjuvant (CIFU) or capecitabine (C) in disseminated colorectal cancer (CRC). chemotherapy in head and neck squamous cell carcinoma: a Proc Am Soc Clin Oncol 2002 (abstract 1720); 21: 431a. prospective series. J Clin Oncol 2000; 18: 1465–1473. 191 Johnston PG, Liang CM, Henry S, Chabner BA, Allegra CJ. Production 176 Temam S, Flahault A, Perie S, Monceaux G, Coulet F, Callard P and characterization of monoclonal antibodies that localize human et al. P53 gene status as a predictor of tumour response to induc- thymidylate synthase in the cytoplasm of human cells and tissue. tion chemotherapy of patients with locoregionally advanced squa- Cancer Res 1991; 51: 6668–6676. mous cell carcinomas of the head and neck. J Clin Oncol 2000; 18: 192 Zembutsu H, Ohnishi Y, Tsunoda T, Furukawa Y, Katagiri T, Ueyama Y 385–394. et al. Genome-wide cDNA microarray screening to correlate gene 177 Kandioler-Eckersberger D, Ludwig C, Rudas M, Kappel S, Janschek E, expression profiles with sensitivity of 85 human cancer xenografts to Wenzel C et al. TP53 mutation and p53 overexpression for prediction anticancer drugs. Cancer Res 2002; 62: 518–527. of response to neoadjuvant treatment in breast cancer patients. Clin 193 Maxwell PJ, Longley DB, Latif T, Boyer J, Allen W, Lynch M et al. Cancer Res 2000; 6: 50–56. Identification of 5-fluorouracil-inducible target genes using cDNA 178 Elsaleh H, Powell B, Soontrapornchai P, Joseph D, Goria F, Spry N et al. microarray profiling. Cancer Res 2003; 63: 4602–4606. P53 gene mutation, microsatellite instability and adjuvant chemother- 194 Mariadason JM, Arango D, Shi Q, Wilson AJ, Corner GA, Nicholas C apy: impact on survival of 388 patients with Dukes C carcinoma. et al. Gene expression profiling-based prediction of response of colon Oncology 2000; 58: 52–59. carcinoma cells to 5-fluorouracil and camptothecin. Cancer Res 2003; 179 Rosty C, Chazal M, Etienne MC, Letoublon C, Bourgeon A, Delpero JR 63: 8791–8812. et al. Determination of microsatellite instability, p53 and K-RAS 195 Wang W, Cassidy J, O’Brien V, Ryan KM, Collie-Duguid E. Mechanistic mutations in hepatic metastases from patients with colorectal cancer: and predictive profiling of 5-fluorouracil resistance in human cancer relationship with response to 5-fluorouracil and survival. Int J Cancer cells. Cancer Res 2004; 64: 8167–8176. 2001; 95: 162–167. 196 Website: http://www.pharmacogenetics.org.