2 Combinations of Antimetabolites and Ionizing Radiation

Home , Antimetabolite, Mercaptopurine, Pemetrexed

Combinations of Antimetabolites and Ionizing Radiation 19 2 Combinations of Antimetabolites and Ionizing Radiation

Hiroshi Harada, Keiko Shibuya, and Masahiro Hiraoka

CONTENTS lular enzymes and show cytotoxic effects by (a) substituting for a normal metabolite incorporated 2.1 Antimetabolites: Classes, Intracellular Metabolism, into key molecules, such as DNA and RNA, and (b) and Mechanisms of Action 19 2.1.1 Pyrimidine Analogs 19 occupying the catalytic site of a key enzyme and 2.1.1.1 5-Fluorouracil 19 competing with a normal metabolite. Consequently, 2.1.1.2 Gemcitabine 21 they interfere with DNA synthesis and proliferation 2.1.2 Folic Acid Analogs 22 of the cancer cell. Until recently, several kinds of 2.1.2.1 Methotrexate 22 antimetabolites have been developed and are cat- 2.1.2.2 Pemetrexed 23 2.1.3 6-Mercaptopurine as an Example for Purine egorized into three major groups, pyrimidine ana- Analogs 23 logs (see Fig. 2.1a), folic acid analogs (see Fig. 2.1b) 2.2 Mechanisms of Radiosensitization with and purine analogs (see Fig. 2.1c). The mechanisms Antimetabolites 24 of action of representative agents in each class are 2.2.1 dNTP Depletion 24 briefly described herein. 2.2.2 Cell Cycle Distribution 25 2.2.3 Checkpoint and p53 25 2.2.4 Apoptosis 26 2.3 Interactions of Antimetabolites with Radiation: 2.1.1 Preclinical to Clinical 26 Pyrimidine Analogs 2.3.1 5-Fluorouracil and its Prodrugs: Capecitabine, UFT, and S-1 26 2.3.1.1 5-Fluorouracil 26 2.1.1.1 2.3.1.2 Prodrugs of 5-FU: UFT; S-1; and Capecitabine 28 5-Fluorouracil 2.3.2 Gemcitabine 29 2.3.3 Pemetrexed 30 5-Fluorouracil (5-FU) is a nucleoside analog of uracil 2.4 Conclusion 30 in which a fluorine atom is inserted into the C-5 posi- References 31 tion in place of hydrogen (see Fig. 2.1a; Longley et al. 2003). It has been widely applied for the treatment of various kinds of cancers, particularly for colorectal and breast cancers. 5-FU requires metabolic acti- 2.1 vation to form its cytotoxic metabolites (see Fig. 2.2; Antimetabolites: Classes, Intracellular Malet-Martino and Martino 2002; Longley et Metabolism, and Mechanisms of Action al. 2003). 5-FU is converted to 5-fluorouridine-5’- monophosphate (5-FUMP) directly by orotate phos- Antimetabolites have antineoplastic activity, which phoribosyltransferase (OPRT) or indirectly via an is attributed to the fact that their structure is very intermediate metabolite, 5-fluorouridine (5-FUrd), similar to the normal metabolites required for cell by uridine phosphorylase and uridine kinase. 5- function and replication. After intracellular modi- FUMP is subsequently phosphorylated to 5-fluo- fication, the antimetabolites interact with intracel- rouridine-5’-diphosphate (5-FUDP) by pyrimidine monophosphate kinase and further to a cytotoxic metabolite, 5-fluorouridine-5’-triphosphate (5- H. Harada, MD FUTP) by pyrimidine diphosphate kinase. The 5- Shibuya K. , MD FUTP can mimic uridine-5’-triphosphate (UTP), M. Hiraoka, MD PhD Department of Radiation Oncology and Image-Applied and be recognized by RNA polymerases as the sub- Therapy, Graduate School of Medicine, Kyoto University, strate. This leads to the incorporation of 5-FU in Sakyo, Kyoto, 606-8507, Japan all classes of RNA, disrupting normal RNA func- 20 H. Harada et al. a Pyrimidine analog b Folic acid analog 5-Fluorouracil Gemcitabine hydrochloride Methotrexate

NH2 H2N N N H H N N O N N CH2N CONH C CH2CH2COOH NH2 O N CH3 COOH NH F O HOH2C HCl Pemetrexed H F O H N H H 2 OH F NH N O c Purine analog O O 6-mercaptopurine ONa N HN SH

N N O NaO N N H

Fig. 2.1a-c. Structures of antimetabolites. a Pyrimidine analogs, gemcitabine hydrochloride, and 5-ﬂ uorouracil. b Folic acid analogs, methotrexate and pemetrexed. c Purine analog, 6-mercaptopurine

TP Urd 5-FU 5-FdUrd facilitation UP 5-FU inhibition OPRT 5-FUrd OPRT TK

UK UMP 5-FUMP 5-FdUMP TS inhibition PMK PMK PMK

UDP 5-FUDP 5-FdUDP dUMP

PDK PDK PDK TS

UTP 5-FUTP 5-FdUTP dUTP dTMP

RNAP RNAP DNAP DNAP DNAP

Incorporation Incorporation Incorporation Incorporation into RNA into RNA into DNA into DNA

DNA replication RNA function RNA damage DNA damage DNA repair

Fig. 2.2. Metabolism and mechanism of action of 5-fl uorouracil. 5-fl uorouracil (5-FU), 5-fl uorouridine-5’-monophosphate (5-FUMP), orotate phosphoribosyltransferase (OPRT), 5-fl uorouridine (5-FUrd), uridine phosphorylase (UP), uridine kinase (UK), 5-fl uorouridine-5’-diphosphate (5-FUDP), pyrimidine monophosphate kinase (PMK), 5-fl uorouridine-5’-triphosphate (5-FUTP), pyrimidine diphosphate kinase (PDK), RNA polymerase (RNAP), uridine (Urd), Uridine-5’-monophosphate (UMP), uridine-5’-diphosphate (UDP), uridine-5’-triphosphate (UTP), 5-fl uoro-2’-deoxyuridine (5-FdUrd), thymidine kinase (TK), 5-fl uoro-2’-deoxyuridine-5’-monophosphate (5-FdUMP), 5-fl uoro-2’-deoxyuridine-5’-diphosphate (5-FdUDP), 5-fl uoro-2’- deoxyuridine-5’-triphosphate (5-FdUTP), thymidylate synthase (TS), thymidine-5’-monophosphate (dTMP), 2’-deoxyuridine- 5’-monophosphate (dUMP), thymidine-5’-triphosphate (dTTP), 2’-deoxyuridine-5’-triphosphate (dUTP), DNA polymerase (DNAP) Combinations of Antimetabolites and Ionizing Radiation 21 tion. The misincorporation inhibits the processing 2.1.1.2 of pre-ribosomal RNA to mature ribosomal RNA Gemcitabine (Kanamaru et al. 1986), the modification of tRNAs (Santi and Hardy 1987) and the splicing of mRNAs Gemcitabine (dFdCyd), 2’-deoxy-2’,2’-difluoro- (Doong and Dolnick 1988), leading to profound cytidine, is a nucleoside analog of deoxycytidine effects on cellular metabolism and viability. in which two fluorine atoms are inserted into the After the conversion of 5-FU to 5-fluoro-2’- deoxyribofuranosyl ring (see Fig. 2.1a). It shows deoxyuridine (5-FdUrd), the 5-FdUrd is sequentially broad-spectrum activity in the treatment of solid phosphorylated to cytotoxic 5-fluoro-2’-deoxyuri- malignancies, particularly for pancreatic (Kaye dine-5’-monophosphate (5-FdUMP), to 5-fluoro- 1994; Rothenberg et al. 1996) and non-small cell 2’-deoxyuridine-5’-diphosphate (5-FdUDP), and lung cancer (Gatzemeier et al. 1996; Crino et al. finally to cytotoxic 5-fluoro-2’-deoxyuridine-5’- 1999). In order for dFdCyd to produce its cytotoxic triphosphate (5-FdUTP). The 5-FdUMP and the effects, dFdCyd requires the following intracellular 5-FdUTP exert their antitumor effects through the modifications (see Fig. 2.3): The dFdCyd is firstly inhibition of thymidylate synthase (TS) (Santi et al. phosphorylated by deoxycytidine (dCyd) kinase 1974; Sommer and Santi 1974) and incorporation to the 5’-monophosphate of dFdCyd (dFdCMP; into DNA, respectively. TS is responsible for the pro- Heinemann et al. 1988). Subsequent phosphory- duction of thymidine-5’-monophosphate (dTMP) lations yield the active metabolites diphosphate from 2’-deoxyuridine-5’-monophosphate (dUMP) (dFdCDP) and triphosphate (dFdCTP) nucleotides through the transfer of a methyl group from 5,10- (Heinemann et al. 1988). These metabolites have methylenetetrahydrofolate (CH2THF). This reaction the potential to interfere with multiple steps of DNA is fully responsible for the production of de novo synthesis and are directly and indirectly responsible source of thymidylate, which is necessary for DNA for the cytotoxic properties of dFdCyd. replication and repair. The 5-FdUMP binds to the The dFdCTP form competes with dCTP for incor- nucleotide-binding site of TS, forms a stable ternary poration into DNA, and the incorporation of dFdCTP complex with TS and CH2THF, and interferes with the into DNA is strongly correlated with the inhibition access of normal substrate dUMP to the nucleotide- of further DNA synthesis (Huang et al. 1991). Once binding site, resulting in the inhibition of dTMP syn- the dFdCTP is incorporated into the end of the elon- thesis (Santi et al. 1974; Sommer and Santi 1974). gating DNA strand, only one more deoxynucleotide Depletion of dTMP directly results in the subsequent is added, and thereafter the DNA polymerases are depletion of thymidine-5’-triphosphate (dTTP) and unable to proceed, resulting in the termination of induces an imbalance of the deoxynucleotide pool the DNA synthesis. Moreover, proof-reading exonu- through various feedback mechanisms. As a result of cleases, such as DNA polymerase epsilon, are unable these actions, 5-FU inhibits DNA synthesis and repair, to remove the incorporated dFdCTP from the penul- leading to lethal DNA damage (Yoshioka et al. 1987; timate position and are unable to repair the growing Houghton et al. 1995). In addition, TS inhibition is DNA strands. This action is termed “masked chain accompanied by the accumulation of its substrate, termination” (Plunkett et al. 1995). dUMP, which might be subsequently metabolized to The dFdCDP form, on the other hand, can inhibit 2’-deoxyuridine-5’-triphosphate (dUTP; Mitrovski ribonucleotide reductase, which is responsible for et al. 1994; Aherne et al. 1996). Both the dUTP and a the reactions that synthesize the deoxynucleotides 5-FU metabolite, 5-FdUTP, can be misincorporated required for DNA synthesis and repair (Baker et into DNA. UracilDNAglycosylase (UDG) hydro- al. 1991). The inhibition of this enzyme causes a lyzes the (F)uracil-deoxyribose glycosyl bond of the reduction in the concentrations of the DNA precur- dUTP and the 5-FdUTP residues in DNA, generating sor pool (Heinemann et al. 1990). The reduction in a single strand break; however, since the depletion the dNTP (particularly dCTP) leads to the following of dTTP, which is caused by the TS inhibition, as “self-potentiation.” Firstly, the reduction enhances mentioned above, reduces the efficacy of the repair the incorporation of dFdCyd-derived nucleo- of such strand breaks, the misincorporated dUTP tides into DNA, because dFdCTP competes with and 5-FdUTP are hard to be removed, leading to cell dCTP for incorporation into DNA, as mentioned death. The misincorporation of 5-FU-metabolites above (Huang et al. 1991). Secondly, the reduction into DNA and the imbalance of intracellular nucleo- enhances the phosphorylation of the dFdCyd and tides results in various detrimental effects on DNA increases the intracellular dFdCDP and dFdCTP synthesis. because the dCyd kinase is negatively regulated 22 H. Harada et al.

dCyd dFdCyd dFdU facilitation Cyd / dCyd deaminase inhibition dCyd kinase

dCMP dFdCMP dFdUMP dCMP deaminase CDP

Ribonucleotide dCDP reductase dFdCDP

depletion of dCTP dCTP dFdCTP

DNAP DNAP

Incorporation Incorporation into DNA into DNA

Masked chain Termination of DNA Synthesis termination DNA synthesis

Fig. 2.3. Metabolism and mechanism of action of gemcitabine. Gemcitabine (dFdCyd), gemcitabine-5’-monophosphate (dFdCMP), gemcitabine-5’-diphosphate (dFdCDP), gemcitabine-5’-triphosphate (dFdCTP), 2’deoxycytidine (dCyd), 2’-deoxycytidine-5’-monophosphate (dCMP), 2’-deoxycytidine-5’-diphosphate (dCDP), 2’-deoxycytidine-5’-triphosphate (dCTP), cytidine-5’-diphosphate (CDP), DNA polymerase (DNAP), 2’-deoxy-2’,2’-diﬂ uorouridine (dFdU), dFdU-5’-monophosphate (dFdUMP) by the dCTP (Shewach et al. 1992a). Finally, the osteosarcoma, breast cancer, and head and neck reduction suppresses the deamination (inactiva- cancer (Chu et al. 1996). The MTX is taken up by tion) of dFdCyd and its derivatives because dCTP cells via the reduced folate carrier (RFC), such as is essential as a co-factor for the activity of dCMP- RFC1, and is then converted within the cells by folyl- deaminase (Hertel et al. 1990). As a result of these polyglutamate synthase to methotrexate polygluta- mechanisms, dFdCyd exhibits cell phase specific- mates (MTXPGs; see Fig. 2.4; Chabner et al. 1985). ity, primarily killing cells undergoing DNA synthe- The MTXPGs are retained longer in cells compared sis (S-phase) and also blocking the progression of with MTX (Jolivet et al. 1983) and compete with cells through the G1/S-phase boundary (Hertel et some cellular folate cofactors (e.g., 10-formyl-tet- al. 1990). Yet, additional observations suggest that rahydrofolate, 5,10-methylene-tetrahydrofolate and other effects are implicated in the interaction with 5,10-methenyl-tetrahydrofolate) for the interaction ionizing radiation also (Rosier et al. 2004). with dihydrofolate reductase (DHFR). This results in the inhibition of the activity of DHFR, thereby decreasing the amount of reduced folate, which is 2.1.2 the carbon donor for the purine ring formation in Folic Acid Analogs the critical pathways for DNA synthesis, DNA repair and cell replication, such as de novo purine synthe- 2.1.2.1 sis (DNPS) and thymidine synthesis (Allegra et Methotrexate al. 1985a,b). In addition, MTXPGs directly inhibit phosphoribosyl pyrophosphate amidotransferase Methotrexate (MTX), 4-amino 10-methyl folic acid (PRPPAT), glycinamide ribonucleotide formyltrans- (see Fig. 2.1b), is an analog of folic acid that has ferase (GARFT) and 5-aminoimidazole-4-carbox- been widely used for the treatment of childhood amide ribonucleotide formyltransferase (AICARFT), acute lymphoblastic leukemia (ALL) and a number which are also key enzymes in the DNPS pathway of other malignant diseases, such as lymphoma, (Segal et al. 1990; Kremer 1994). The DNPS inhi- Combinations of Antimetabolites and Ionizing Radiation 23

MTX Extracellular RFC1 facilitation inhibition Intracellular MTX

Folylpolyglutamate synthase

MTXPGs

DNPS DHFR TS DNA synthesis DNA repair PRPPAT cell replication GARFT DNPS AICARFT

Fig. 2.4. Metabolism and mechanism of action of methotrexate. Methotrexate (MTX), MTX polyglutamates (MTXPGs), dihydrofolate reductase (DHFR), phosphoribosyl pyrophosphate amidotransferase (PRPPAT), glycinamide ribonucleotide formyltransferase (GARFT), 5-aminoimidazole-4-carboxamide ribonucleotide formyltransferase (AICARFT), de novo purine synthesis (DNPS), thymidine synthase (TS). Pemetrexed has been reported to show antineoplastic activity through almost the same mechanism as MTX bition by both MTX and its metabolite results in MTX (see Fig. 2.4), leading to a subsequent decline purine depletion, leading to the inhibition of DNA in cell proliferation, which is particularly significant synthesis, decreased cell proliferation and signifi- in rapidly proliferating tumor cells. cant cytotoxicity.

2.1.2.2 2.1.3 Pemetrexed 6-Mercaptopurine as an Example for Purine Analogs Pemetrexed, L-glutamic acid, N-[4-[2-(2-amino- 4,7-dihydro-4-oxo-1H-pyrrolo[2,3-d]pyrimidin- 6-Mercaptopurine (6-MP) is a 6-thiopurine analog 5- yl)ethyl]benzoyl]-, disodium salt, heptahydrate of purine bases (see Fig. 2.1c), such as hypoxanthine (see Fig. 2.1b), is a multi-targeted folate analog that and guanine, and has been in clinical use for over exerts its antineoplastic activity in many kinds 30 years as an antileukemic agent. 6-MP is a prodrug of human malignancies, e.g., breast, pancreatic, that requires activation to exert its cytotoxic effect colorectal, head and neck, gynecological, and non- (see Fig. 2.5). It is first converted by hypoxanthine- small cell lung cancers (Hanauske et al. 2001), by guanine phosphoribosyl transferase (HGPRT) into disrupting folate-dependent metabolic processes. It 6-thioinosine monophosphate (6-TIMP; Lennard gains entry to cells via the reduced folate carrier, 1992). It is subsequently metabolized by a two-step and once inside a cell, it is intracellularly converted process involving inosine monophosphate dehydro- to polyglutamate forms, tri- and pentaglutamate, by genase (IMPDH) into 6-thioxanthine monophos- folylpolyglutamate synthetase. This modification phate (6-TXMP), and by guanosine monophosphate prolongs its intracellular retention (Mendelsohn synthetase into 6-thioguanosine 5’-monophosphate et al. 1999). The polyglutamate forms inhibit DHFR, (6-TGMP). 6-TGMP is further metabolized by a TS, and GARFT, all of which are folate-dependent series of kinases and reductases to deoxy-6-thio- enzymes responsible for the de novo biosynthesis of guanosine 5’-triphosphate (dGS). The resultant dGS thymidine and purine nucleotides (Shih et al. 1997). can be incorporated into DNA and trigger cell cycle As a result of these processes, pemetrexed inhibits arrest and apoptosis (Swann et al. 1996). In addi- DNA synthesis by almost the same mechanism as tion, thiopurine methyltransferase (TPMT) also can 24 H. Harada et al.

these activities and have reported significant tumor growth delay with the combination of antimetabolites and ionizing radiation in animal models; however, the mechanism of radiosensitization has yet to be fully understood. Initial studies have focused, as the mechanism, on the modulation of the metabolism of deoxynucleotides, the changes in cell cycle distribution, the role of the tumor suppressor protein p53, and the induction of apoptosis.

2.2.1 dNTP Depletion

Antimetabolite-induced perturbation of the intracellular dNTP pool and its radiosensitizing effect are summarized in Table 2.1. In the case of dFdCyd, it increases the radiosensitivity of cells even under conditions in which the drug alone shows no cytotoxicity (Shewach et al. 1994). The dFdCyd- mediated sensitization has been reported not to be dependent either on the intracellular concen- Fig. 2.5. Metabolism and mechanism of action of 6-Mercap- tration of dFdCTP or on the dFdCTP/dCTP ratio, topurine. 6-Mercaptopurine (6-MP), hypoxanthine-guanine but correlates to the dFdCDP-mediated decrease phosphoribosyl transferase (HGPRT), 6-thioinosine mono- in intracellular dATP pools (Shewach et al. 1994); phosphate (6-TIMP), inosine monophosphate dehydroge- therefore, it has been hypothesized that the deple- nase (IMPDH), 6-thioxanthine monophosphate (6-TXMP), guanosine monophosphate synthetase (GMS), 6-thioguano- tion of the dATP pool level by the dFdCyd treat- sine 5’-monophosphate (6-TGMP), deoxy-6-thioguanosine ment is a key factor in enhancing radiosensitivity 5’-triphosphate (dGS), DNA polumerase (DNAP), thiopurine (Shewach et al. 1994; Shewach and Lawrence methyltransferase (TPMT), S-methyl-thioinosine 5’-mono- 1995; Lawrence et al. 1996, 1997). In addition, phosphate (MeTIMP), phosphoribosyl pyrophosphate amido- it has been reported that a 5-FU-derivative, 5- transferase (PRPPAT) FdUrd, which depletes the intracellular TTP level but not dATP pools under the radiosensitizing condition, decreases the efficiency of repair of convert the 6-TIMP into S-methyl-thioinosine 5’- radiation-induced DNA damage (Bruso et al. 1990; monophosphate (MeTIMP). The resultant MeTIMP Heimburger et al. 1991). Likewise, a thymidine acts as a strong inhibitor of PRPPAT, which is respon- analog, 5-bromo-2’-deoxyuridine (BrdU), which is sible for de novo purine synthesis (Tay et al. 1969; known to be incorporated into DNA and decrease Allan and Bennett 1971); thus, 6-MP treatment the intracellular dCTP and TTP pools (Shewach results in purine depletion, leading to the inhibition et al. 1992b), increases the radiation-induced DNA of DNA synthesis and proliferation, and antileuke- damage, and moreover, decreases damage repair mic effect. (Iliakis et al. 1989; Ling and Ward 1990). Further- more, hydroxyurea, whose primary activity is the depletion of intracellular dNTP levels, can enhance the sensitivity of the cell to radiation (Sinclair 2.2 1968a). Taken together, perturbation of the bal- Mechanisms of Radiosensitization with ance of intracellular nucleosides seems to play an Antimetabolites important role in causing radiosensitization, at least in part. This hypothesis is consistent with the In addition to the cytotoxic effect of these antime- reports showing that imbalances of the intracellu- tabolites themselves, they have been reported to lar dNTP pool produce errors in DNA replication enhance the effect of radiation as potent radiosensi- (Kunz 1982; Bebenek et al. 1992; Martomo and tizers in vitro. Recent in vivo studies have confirmed Mathews 2002). Combinations of Antimetabolites and Ionizing Radiation 25

Table 2.1. Antimetabolite-induced imbalance of intracellular dNTP pool and its radiosensitizing effect

Antimetabolite Imbalance of dNTP Effect Reference dFdCyd dATP depletion Radiosensitization in vitro and in vivo Shewach et al. (1994), (gemcitabine) Shewach and Lawrence (1995), Lawrence et al. (1996), Lawrence et al. (1997) 5-FU TTP depletion Decrease in repair of DNA damage Bruso et al. (1990), (5-FdUrd) induced by radiation Heimberger et al. (1991) BrdU dCTP and TTP depletion Increase in radiation-induced DNA Iliakis et al. (1989), damage and decrease in damage repair Ling et al. (1990) Hydroxyurea dNTPs depletion Radiosensitization in vitro Sinclair et al. (1968a,b)

2.2.2 Radioresistance Cell Cycle Distribution DNA content per cell

The cell cycle for mammalian cells is composed of four sequential phases: mitosis (M-phase); gap 1 4 (G1); DNA synthetic phase (S-phase); and gap 2 (G2), followed again by M-phase (see Fig. 2.6). It is known 3 that cells are the most radioresistant in the S-phase, 2 while they are radiosensitive in the M- and G2- phases (Sinclair 1968b; Pawlik and Keyomarsi 2004). Moreover, cells at the G /S-boundary and in 1 1 content (per cell) DNA early S-phase are more radiosensitive than those in Radioresistance Relative late S-phase (Latz et al. 1998); therefore, cell cycle MSGMMSGG1 2 M modulation with an antimetabolite also seems to be Cell Cycle very important for the radiation-enhancing effect. The cell cycle effect of dFdCyd seems to be concen- Fig. 2.6. Cell cycle phase-dependent radiosensitivity of cells. tration dependent (Tolis et al. 1999; Cappella et al. The cell cycle for mammalian cells is composed of four sequential phases: mitosis (M-phase); gap 1 (G1), DNA syn- 2001). It was reported that low concentrations (IC50 thetic phase (S-phase) and gap 2 (G2), followed by M-phase values) of dFdCyd cause cell cycle arrest in early S- again. The DNA content in each cell phase (dashed curve) and phase (Merlin et al. 1998), and increasing concen- the relative radioresistance (solid curve) of the cell are shown trations of the drug results in a shift of this arrest to the early S-phase and moreover to the G1/S-boundary (Pauwels et al. 2003). The radiation-enhancing phenomenon still remains unclear, the treatment effect of dFdCyd, therefore, could be explained by may reduce the repair of radiation-induced cellular the accumulation of dFdCyd-treated cells in these damage and increase the probability of the fixation radiosensitive phases. Indeed, the percentage of of damage being lethal in S-phase cells. early S-phase cells was reported to correlate with the radiosensitizing effect (Pauwels et al. 2003); however, it is still unclear both why radiosensitization 2.2.3 occurs in cells at early S-phase and whether such a Checkpoint and p53 mechanism is responsible for the radiosensitizing property of the other antimetabolites, which exhibit In response to a stress signal, p53 protein is activated the same kind of cell phase specificity. by post-translational modifications and leads to the In addition, it has also been reported that both transcription of various genes, which determine dFdCyd and 5-FU selectively radiosensitize cells in whether the cell will continue to progress, arrest S-phase, which are relatively radioresistant, whereby in checkpoints and repair the DNA damage, induce the fluctuation of radiosensitivity within the cell cell senescence, or trigger apoptosis (Levine 1997; cycle is eliminated or at least reduced (Latz et al. Jin and Levine 2001). Cells expressing wild-type 1998). Although the mechanism causing the above p53 were reported to accumulate in the G1/S cell- 26 H. Harada et al. cycle checkpoint in response to many DNA-damag- to irradiation, the expression of proapoptotic fac- ing agents including irradiation (Linke et al. 1997; tors is further induced, leading to apoptosis. This Ostruszka and Shewach 2000). This arrest allows mechanism may be a reason why the maximum the cells not only to repair the DNA damage, but radiosensitization with dFdCyd and with 5-FU also to inhibit the replication of the cells with dam- occurs when cells are incubated with the drug aged genomic DNA (Sak et al. 2000). On the other before radiation. hand, cells expressing mutant p53 were reported to continue to progress into S-phase and G2-M after the treatment and irradiation (Ostruszka and Shewach 2000). This may affect the radiosensitiv- 2.3 ity of cells. Indeed, the p53 status and the inap- Interactions of Antimetabolites with propriate progression of cells into S-phase in the Radiation: Preclinical to Clinical presence of drugs, such as 5-FU and 5-FdUrd as well as dFdCyd, has been reported to affect radiation 2.3.1 sensitivity. This conclusion is firstly derived from 5-Fluorouracil and its Prodrugs: Capecitabine, the result that 5-FdUrd enhances the radiosensitiv- UFT, and S-1 ity of cells expressing G1/S cyclins in the presence of the drug, but not of cells expressing no activated 2.3.1.1 cyclins (Lawrence et al. 1996). It is further con- 5-Fluorouracil firmed by the report that 5-FdUrd greatly increases the radiosensitivity of a cell, which can progress Since 5-FU was first introduced in 1957, it has been into S-phase after irradiation because of the expres- playing an essential part in the treatment of a wide sion of mutant-type p53, but hardly shows a radio- range of solid tumors, such as epithelial malignan- sensitizing effect on a cell expressing wild-type p53 cies in the gastrointestinal tract, pancreatic cancer, (Naida et al. 1998). Likewise, a previous study sug- breast cancer, or head and neck cancer. Although gested that a cell line that has no ability to progress the single-agent response rates are not so high, through S-phase after dFdCyd and radiation is not 5-FU-containing treatment regimens produced a radiosensitized because of the expression of wild- highly significant survival improvement for sev- type p53 (Ostruszka and Shewach 2000). eral malignancies. Heidelberger et al. (1958) Yet, other reports have suggested that, regardless demonstrated very early that doses of radiation, of the p53 status, dFdCyd can enhance the radiosen- which were inhibitory but not curative for rodent sitivity of cells, as shown by Robinson and Shewach tumors, were made curative by combination with (2001), in whose study both wild-type p53-express- 5-FU. Subsequently, results in experimental ani- ing cells and mutant p53-expressing cells displayed mals (Vermund et al. 1961) and quantitative stud- high dATP depletion and S-phase accumulation ies showed that 5-FU could enhance cell killing after dFdCyd treatment, both of which we recog- by radiation ( Bagshaw 1961; Berry 1966). These nize as important factors in radiosensitization. That early observations led to a series of clinical investi- article suggests that p53 function alone might not gations from the 1960s to 1970s. Firstly, controlled determine the radiosensitivity in the presence of studies in gastric adenocarcinoma (Moertel et antimetabolites. al. 1969) and head and neck squamous carcinoma (Gollin et al. 1972) showed improved response and survival rates by combined therapy with radiation 2.2.4 and 5-FU; however, other studies could not achieve Apoptosis any remarkable results (Helsper and Sharp 1962; Hall et al. 1967; Stein and Kaufman 1968; Sato et Although antimetabolites are known to induce the al. 1970). These negative, or less impressive results, expression of proapoptotic effector proteins, the were considered to be caused not only by the organ level is not sufficient to induce cellular apoptosis. or tumor-specific characteristics, but also by some In the case that the intracellular checkpoint system factors which depend on the administration sched- is working, the damage of genomic DNA is repaired ule. During those periods, bolus 5-FU had been and the intracellular level of proapoptotic factors usually given with fractionated radiotherapy. In decreases during the cell cycle arrest; however, 1982, Byfield et al. (1982) demonstrated that cell during the above period, if the cells are exposed killing was maximized if the cells were continu- Combinations of Antimetabolites and Ionizing Radiation 27 ously exposed to 5-FU following irradiation. In of the clinical trials of the National Surgical Adju- 1994, a well-controlled study in patients with rectal vant Breast and Bowel Project (NSABP) showed that cancer [Mayo/North Center Cancer Treatment chemoradiation altered neither the incidence of dis- Group (NCCTG) 86-47-51’s protocol], comparing tant metastases nor survival, there was a reduction protracted venous infusion (PVI) with bolus injec- in the locoregional relapse (see Table 2.2; Wolmark tion during radiotherapy, showed that PVI resulted et al. 2000). To reduce the toxicities and to enhance in a significant improvement in both the time to the survival for colorectal cancer, optimizing the relapse and survival (O’connell et al. 1994). As regimen of chemoradiation is still under active 5-FU has cell-cycle specificity and a short plasma investigation. half-life, the prolonged exposure of cells to the For the treatment of esophageal cancer, the ran- agent would theoretically result in enhanced cell domized study, EST-1282, was undertaken by the killing. Presently, as a standard method, 5-FU is Eastern Cooperative Oncology Group (ECOG) to administered by PVI or by bolus infusion for con- determine whether the combined use of 5-FU, mito- secutive days during radiation. mycin C (MMC) and radiation therapy improved Table 2.2 shows the representative studies con- the survival of patients compared with radiation ducted to test the efficacy of chemoradiotherapy therapy alone. In that study, 5-FU was delivered with 5-FU for several cancers. by continuous infusion for 96 h, initiated on day 2 As for the patients with colorectal lesions, we need and day 28 during radiation therapy (Smith et al. to be very careful about gastrointestinal toxicities. 1998). The result was that patients treated with Miller et al. (2002) analyzed the NCCTG trial and chemoradiation had a longer median survival showed that patients who received 5-FU by PVI had than patients receiving radiation therapy alone. a higher risk of severe or life-threatening diarrhea The Radiation Therapy Oncology Group (RTOG) compared with those with bolus infusion during compared radiotherapy of 64 Gy in 32 fractions pelvic radiotherapy. Patients with locally advanced alone with chemoradiotherapy of 50 Gy in 25 frac- (T3 or T4) rectal cancer had been recommended to tions with cisplatin/5-FU, showing that combined receive radiotherapy and chemotherapy either pre- therapy significantly increased the overall survival operatively or postoperatively. Although the results (Cooper et al. 1999).

Table 2.2. Randomized trials of chemoradiation with 5-ﬂ uorouracil (5-FU)

Study Treatment No. of patients Median survival Survival Follow-up (months) (%) (years) Rectal cancer Krook et al. (1991) S+RT 100 35 7 (median) Mayo/NCCTG79-47-51 S+RT+5-FU 104 55 7 (median) O’Connell (1994) S+RT+bolus 5-FU 332 60 4 Mayo/NCCTG86-47-51 S+RT+infusional 5-FU 328 70 4 Wolmark et al. (2000) S+5-FU/LV 348 6265 5 NSABP R-02 S+5-FU or MOF/LV+RT 346 6265 5 Pancreatic cancer Moertel et al. (1981) RT 60 Gy 5.2 Gastrointestinal Tumor RT 40 Gy+5-FU 9.6 Study Group (GISTG) RT 60 Gy+5-FU 9.2 GISTG (1988) RT 54 Gy+5-FU 10.5 GISTG (1988) SMF 8 Esophageal cancer Cooper et al. (1985) RT64 Gy 62 9.3 0 (5 years) 8 RTOG 85-01 RT50 Gy+CDDP/5-FU 61 14.1 26 (5 years) 8 S surgery, RT radiotherapy, SMF STZ, mitomycin C and 5-FU, MOF 5-FU, semustine, and vincristine, LV leucovorin, CDDP cisplatin. 28 H. Harada et al.

2.3.1.2 1997). By the stabilization and prolongation of the Prodrugs of 5-FU: UFT; S-1; and Capecitabine effective half-life of FU, the opportunity for synergy with radiation had been considered to be greater. To A prodrug is defined as a pharmacologically inac- date, there has been a limited number of published tive compound that is converted into an active agent studies combining UFT with radiation therapy, but by a metabolic biotransformation. The prodrugs of recently a phase-I trial examined the use of the oral 5-FU are characterized by a pyrimidine ring with delivery of UFT/leucovorin with conventional RT a fluorine atom in position 5. The first generations of 45 Gy for pancreatic cancer. Results compared of prodrugs of 5-FU were represented by 5-fluoro- favorably with continuous-infusion regimens show- 2’-deoxyuridine (5-FdUrd), which was more effi- ing the potential benefit (Childs et al. 2000). In ciently metabolized by the liver than 5-FU. As for preoperative and postoperative rectal cancer ther- the second generation, ftorafur (FTO, 1-(2-tetrahy- apy, several studies are ongoing (Rosenthal et al. drofuryl)-5-fluorouracil, Tegaful or Futraful) and 2000). For unresectable non-small cell lung cancer, 5’-deoxy-5-fluorouridine (5’d5-FUrd, doxifluridine Ichinose et al. (2005) conducted a multi-institu- or Furtulon) were developed with the intention of tional phase-II study in which the combination of possible oral administration. They are designed chemotherapy of UFT plus cisplatin was given with to be well absorbed from the gastrointestinal tract concurrent radiotherapy, showing a satisfactory and enzymatically converted into 5-FU by hepatic response (overall response rate 81%) with no severe microsomal cytochrome P-450. toxicities. The third-generation compounds include those preferentially activated in the tumor: capecitabine, 2.3.1.2.2 and the dehydropyrimidine dehydrogenase (DPD) S-1 inhibitory compounds; UFT (FTO+uracil); and S-1 [FTO+5-chloro-2, 4-dihydroxypyrimidine (CDHP; An oral formulation of 5-FU with DPD inhibitors gimestat)+potassium oxonate (OXO; otastat)]. makes oral absorption very reliable and allows protracted exposure without the need for a venous 2.3.1.2.1 catheter infusion pump; however, the results of sev- UFT eral phase-I trials showed that the dose-limiting toxicity (DLT) was gastrointestinal reactions (diar- Although prolonged continuous exposure has been rhea, nausea, vomiting; Gonzalez Baron et al. demonstrated to have advantages over bolus admin- 1993; Muggia et al. 1996; Pazdur et al. 1998). The istration in rectal, head and neck, and esophageal pharmacology study showed that the occurrence of cancers, another dimension for the timing of FU in these toxic effects correlated significantly with the the clinic is illustrated by the animal models, which maximum plasma concentration and area under demonstrate daily cyclic patterns of mitosis and the concentration-vs-time curve (AUC 0–6 h) of enzyme activity. The activities of DPD, which inacti- 5-FU (Ho et al. 2000). The next strategy in the 1990s vates FU, have been shown to vary significantly over was to develop new kinds of drugs to alleviate the 24 h. Because the pharmacokinetics and sensitivity gastrointestinal toxicities, without reduction of the of 5-FU are determined by DPD, stabilization of the high plasma concentration of 5-FU. S-1 (or TS-1) is pharmacokinetics of 5-FU and the enhancement of a combination of FTO and two compounds, CDHP its efficacy have been attempted by means of DPD and OXO, which was developed in 1996 by Japanese inhibition. The UFT is a small molecule, a combina- groups. The CDHP and OXO were designed to act tion agent utilizing FTO and uracil in molar propor- as modulators of 5-FU. The CDHP is a reversible tions of 1:4. Uracil is a competitive and irrevers- and strong inhibitor of DPD. In vitro, CDHP is ible inhibitor of DPD. This combination ratio was 180 times more potent than uracil (Tatsumi et al. chosen based on preclinical models that suggested 1987). The OXO accumulates in the gastrointesti- tumor selectivity, which produces a constant reserve nal tissues and competitively inhibits the enzyme of 5-FU and its active metabolites and minimizes orotate phosphoribosyl transferase (OPRT), which production of inactive and potentially toxic metabo- converts 5-FU to 5-FUMP. The phosphorylation of lites (Fujii et al. 1979). Later, some other preclinical 5-FU within the digestive tract by OPRT has been studies demonstrated that this combination resulted considered the cause of gastrointestinal toxicities. in significant improvements in the tumor to normal In a preclinical study of Yoshida sarcoma-bearing tissue and tissue serum ratios of 5-FU (Taguchi rats, the administration of OXO with UFT mark- Combinations of Antimetabolites and Ionizing Radiation 29 edly reduced the injury of gastrointestinal tissues resistance (Kitazono et al. 1998). Takebayashi et and/or severe diarrhea without influencing the al. (1996) reported that higher levels of PD-ECGF/ antitumor effect of UFT (Shirasaka et al. 1993). TP expression in colorectal carcinomas were asso- The optimal molar ratio of the three constituents ciated with more aggressive malignant growth and (FTO/CDHP/OXO) in S-1 is 1:0.4:1. Several pre- unfavorable clinical outcome; thus, a tumor-specific clinical studies in experimental models of rodent enzyme is responsible for the local production of tumors or human xenografts demonstrated that S- the cytotoxic agent, 5-FU, from capecitabine, i.e., 1 significantly inhibited tumor growth with lower capecitabine should increase the concentration of gastrointestinal toxicities (Takechi et al. 1997; active FU in tumor site and decrease the concen- Fukushima et al. 1998; Cao et al. 1999). Clinically, tration in healthy normal tissues with a reduction S-1 has been approved for use in Japan for gastric of toxicity. Moreover, in human cancer xenografts, cancer, head and neck cancer, colorectal cancer, a number of chemotherapeutic agents, such as and non-small cell lung cancer, and in Korea for paclitaxel and docetaxel (Sawada et al. 1998), and advanced or recurrent gastric cancer, head and cyclophosphamide (Endo et al. 1999), and radia- neck cancer, and unresectable metastatic or recur- tion (Sawada et al. 1999), upregulate TP. Sawada rent colorectal cancer. Presently, several trials of et al. (1999) demonstrated that a single-dose local newly combined chemotherapy of S-1 plus cispla- irradiation of 5 Gy increased the TP levels by up to tin or CPT-11 with radiotherapy for head and neck 13-fold 9 days after irradiation. They also observed cancer, colorectal cancer, and non-small cell lung that whole-body irradiation upregulated TP in a cancer are ongoing. tumor, but it did not increase the enzyme level in the liver. Capecitabine has been approved in more 2.3.1.2.3 than 50 countries, e.g., in metastatic breast cancer Capecitabine (Blum 2001). On the basis of two large phase-III studies, capecitabine has been approved as a first- Capecitabine (N4-pentyloxycarbonyl-5’-deoxy-5- line alternative to IV FU/LV in advanced metastatic fluorovytidine) is an oral fluoropyrimidine car- colorectal cancer (Hoff et al. 2001; Van Cutsem et bonate that is converted to 5-FU through a cascade al. 2001) and recently, capecitabine has been used of three enzymes: carboxylesterase; cytidine (Cyd) in combination therapy with CPT-11 or oxaliplatin deaminase; and thymidine phosphorylase (TP). with promising results in phase-II studies. Phase- These enzymes have been known to have unique III studies comparing capecitabine plus oxaliplatin tissue localization patterns: carboxylesterase is with standard FU/LV/oxaliplatin (FOLFOX) are in almost exclusively located in high concentrations progress. The clinical benefits of combined therapy in the liver and hepatoma, but not other tumors with capecitabine and radiation are under investi- and normal tissues, Cyd deaminase is located in gation. The efficacy and safety of combined chemo- high concentration in the liver and various types of therapy with capecitabine lends support for its use solid tumors, and TP is more concentrated in vari- in chemoradiotherapy. ous types of tumor tissues than in normal tissues (Miwa et al. 1998). Oral capecitabine passes intact through the intestinal tract almost completely to 2.3.2 be first converted in the liver; thus, theoretically, Gemcitabine gastrointestinal injury should not occur with this compound. Schuller et al. (2000) demonstrated Gemcitabine exhibits cell phase specificity, primar- that in the resected samples of patients with colon ily killing cells undergoing DNA synthesis (S-phase) cancer, the concentration of 5-FU was on average and also blocking the progression of cells through 3.2 times higher than in adjacent healthy tissues. In the G1/S-phase boundary. Gemcitabine has recently 1995, TP was reported to be identical with platelet- been shown to be a potent radiosensitizer in preclin- derived endothelial cell growth factor (PD-ECGF; ical studies using human tumor cell lines, caused Haraguchi et al. 1994; Moghaddam et al. 1995), by perturbation of nucleotide metabolism, cell cycle which is expressed at higher levels in a wide variety modulation, or caspase activation and apoptosis (see of solid tumors compared with adjacent normal tis- section 2.1.1.2); however, the mechanisms have not sues. PD-ECGF/TP was suggested to be able to confer been fully elucidated. Even at very low concentra- resistance to apoptosis induced by hypoxia, and deg- tions, gemcitabine has been shown to be a power- radation products of thymidine are involved in this ful radiation sensitizer (Mason 1999). Lawrence 30 H. Harada et al. et al. (1996, 1997) showed that 100-nm gemcitabine, as active as previous, more toxic regimens. As for the which was noncytotoxic, radiosensitized HT29 cells combination with radiation, there are some preclini- up to 48 h after drug removal. During this period, cal studies to assess the radiosensitizing potential there was an increase in the S-phase population and of pemetrexed. Bischof et al. (2002) demonstrated dNTP pools remained depleted throughout the 72-h that pemetrexed enhanced the radiation-induced period after drug treatment. Actually, the most cell inactivation at moderately toxic exposures and common method of clinical administration is by over many hours after drug removal by in vitro short-term infusion (30–90 min), which is relevant survival assays. They also demonstrated the high to the minimalization of normal tissue toxicity, antiendothelial/antitumoral efficacy of the concur- while maintaining synergistic effects during radio- rent administration of irradiation, pemetrexed and therapy. Gemcitabine has shown promising clini- VEGFR inhibitor (SU5416) in vitro (Bischof et al. cal effectiveness against a range of solid tumors, 2004). Clinical trials of combinations of pemetrexed most importantly non-small cell lung cancer, breast and radiation with or without other chemothera- cancer, and pancreatic cancer; however, because of peutic agents for malignant mesothelioma and non- the radiosensitization of both tumor and normal small cell lung cancer are now ongoing. tissues, the therapeutic window is extremely narrow (Rube et al. 2004). Some clinical trials of concurrent radiotherapy with gemcitabine for non-small cell lung cancer or head and neck cancer revealed 2.4 severe pulmonary or esophageal toxicities. Many Conclusion investigators are now attempting to define the optimal dosing and schedule for concurrent gemcitabine Antimetabolite agents have remarkable antineo- and radiation therapy. plastic activities, and their radiosensitizing effects have been reported for several decades both preclin- ically and clinically. Although the mechanisms of 2.3.3 radiosensitization have not been fully understood, Pemetrexed several preclinical studies have focused on the modulation of the metabolism, the changes in cell Pemetrexed is a novel agent that inhibits several cycle distribution, the role of p53, and the induction enzymes of thymidylate and purine synthesis, dis- of apoptosis. Chemoradiation with antimetabolites rupting metabolic processes essential for cell rep- has become the standard therapy for the treatment lication. A large, randomized phase-III trial was of a wide range of solid tumors. Especially 5-FU has conducted to compare pemetrexed/cisplatin with been one of the most commonly used anticancer cisplatin in the treatment of malignant mesothe- drugs in combination with radiotherapy. Recently, lioma, and the combination of pemetrexed/cispla- oral 5-FU prodrugs are emerging in the clinical tin demonstrated superiority in the response rate, area, such as UFT, S-1, and capecitabine. They have survival, and quality of life (Hazarika et al. 2005). new characteristics related to unique metabolic pat- Now several clinical trials are ongoing for malignant terns or enzymatic inhibitors of degradation. These mesothelioma, non-small cell lung cancer, pancre- pharmacological features and oral formulation atic cancer, head and neck cancer, breast cancer, and allow protracted exposure of FU to the tumor tissues colorectal cancer. Among new combination thera- without the need for a central venous catheter or pies, those in progress include pemetrexed with infusion pump. The clinical trials have just started, carboplatin, oxaliplatin (Scagliotti et al. 2005), or but the preliminary results of the combination of gemcitabine (Monnerat et al. 2004) for lung cancer, oral 5-FU prodrugs and radiation are very encour- with gemcitabine for pancreatic cancer (Oettle et aging. Other newer antimetabolite agents, such as al. 2005), and with irinotecan for colorectal cancer gemcitabine or pemetrexed, have been developed (Hochster et al. 2005); however, a randomized and clinically investigated for several years. As phase-III study of pemetrexed plus gemcitabine vs for the combination with radiation, gemcitabine is gemcitabine in patients with advanced pancreatic being established in the key role of systemic therapy cancer showed that pemetrexed plus gemcitabine in locally advanced pancreatic cancer, and chemo- regimen was not superior in overall survival or radiation with gemcitabine is becoming one of the toxicities, but other studies demonstrated that the standard therapies, as an alternative to 5-FU and combinations had good tolerance and were at least radiation. Combinations of Antimetabolites and Ionizing Radiation 31

Currently, several combined chemotherapies of Buchsbaum DJ et al (2002) Treatment of pancreatic cancer these antimetabolites with other active agents, such xenografts with Erbitux (IMC-C225) anti-EGFR antibody, as cisplatin, CPT-11, or taxanes, are under investiga- gemcitabine, and radiation. Int J Radiat Oncol Biol Phys 54:1180–1193 tion. We need to determine the efficacy of chemora- Byfield JE et al (1982) Pharmacologic requirements for obtain- diation using these newly combined chemotherapy ing sensitization of human tumor cells in vitro to com- regimens, as compared with previous standard bined 5-Fluorouracil or ftorafur and X-rays. Int J Radiat chemoradiotherapies for each tumor. In the future, Oncol Biol Phys 8:1923–1933 the assessment of combinations with novel molecu- Cao S et al (1999) Persistent induction of apoptosis and sup- pression of mitosis as the basis for curative therapy with lar targeted therapies, such as, for example, those S-1, an oral 5-fluorouracil prodrug in a colorectal tumor studied in tumor xenografts treated with C225, gem- model. Clin Cancer Res 5:267–274 citabine, and radiation (Buchsbaum et al. 2002), is Cappella P et al (2001) Cell cycle effects of gemcitabine. Int J also warranted. Cancer 93:401–408 Chabner BA et al (1985) Polyglutamation of methotrexate. Is methotrexate a prodrug? J Clin Invest 76:907–912 Childs HA et al (2000) A phase I study of combined UFT plus leucovorin and radiotherapy for pancreatic cancer. Int J References Radiat Oncol Biol Phys 47:939–944 Chu E et al (1996) New concepts for the development and use Aherne GW et al (1996) Immunoreactive dUMP and TTP of antifolates. Stem Cells 14:41–46 pools as an index of thymidylate synthase inhibition; effect Cooper JS et al (1999) Chemoradiotherapy of locally advanced of tomudex (ZD1694) and a nonpolyglutamated quinazo- esophageal cancer: long-term follow-up of a prospective line antifolate (CB30900) in L1210 mouse leukaemia cells. randomized trial (RTOG 85-01). Radiation Therapy Oncol- Biochem Pharmacol 51:1293–1301 ogy Group. J Am Med Assoc 281:1623–1627 Allan PW, Bennett LL Jr (1971) 6-Methylthioguanylic acid, a Crino L et al (1999) Gemcitabine as second-line treatment for metabolite of 6-thioguanine. Biochem Pharmacol 20:847– advanced non-small-cell lung cancer: a phase II trial. J Clin 852 Oncol 17:2081–2085 Allegra CJ et al (1985a) Enhanced inhibition of thymidylate Doong SL, Dolnick BJ (1988) 5-Fluorouracil substitution synthase by methotrexate polyglutamates. J Biol Chem alters pre-mRNA splicing in vitro. J Biol Chem 263:4467– 260:9720–9726 4473 Allegra CJ et al (1985b) Inhibition of phosphoribosylamino- Endo M et al (1999) Induction of thymidine phosphorylase imidazolecarboxamide transformylase by methotrexate expression and enhancement of efficacy of capecitabine or and dihydrofolic acid polyglutamates. Proc Natl Acad Sci 5’-deoxy-5-fluorouridine by cyclophosphamide in mam- USA 82:4881–4885 mary tumor models. Int J Cancer 83:127–134 Bagshaw MA (1961) Possible role of potentiators in radia- Fujii S et al (1979) Effect of coadministration of uracil or cyto- tion therapy. Am J Roentgenol Radium Ther Nucl Med sine on the anti-tumor activity of clinical doses of 1-(2- 85:822–833 tetrahydrofuryl)-5-fluorouracil and level of 5-fluorouracil Baker CH et al (1991) 2’-Deoxy-2’-methylenecytidine and in rodents. Gann 70:209–214 2’-deoxy-2’,2’-difluorocytidine 5’-diphosphates: potent Fukushima M et al (1998) Anticancer activity and toxicity of mechanism-based inhibitors of ribonucleotide reductase. S-1, an oral combination of tegafur and two biochemical J Med Chem 34:1879–1884 modulators, compared with continuous i.v. infusion of 5- Bebenek K et al (1992) The effects of dNTP pool imbalances fluorouracil. Anticancer Drugs 9:817–823 on frameshift fidelity during DNA replication. J Biol Chem Gastrointestinal Tumor Study Group (1988) Treatment of 267:3589–3596 locally unresectable carcinoma of the pancreas: compari- Berry RJ (1966) Effects of some metabolic inhibitors on X-ray son of combined-modality therapy (chemotherapy plus dose-response curves for the survival of mammalian cells radiotherapy) to chemotherapy alone. Gastrointestinal in vitro, and on early recovery between fractionated X-ray Tumor Study Group. J Natl Cancer Inst 80:751–755 doses. Br J Radiol 39:458–463 Gatzemeier U et al (1996) Activity of gemcitabine in patients Bischof M et al (2002) Interaction of pemetrexed disodium with non-small cell lung cancer: a multicentre, extended (ALIMTA, multitargeted antifolate) and irradiation in phase II study. Eur J Cancer 32A:243–248 vitro. Int J Radiat Oncol Biol Phys 52:1381–1388 Gollin FF et al (1972) Combined therapy in advanced head and Bischof M et al (2004) Triple combination of irradiation, che- neck cancer: a randomized study. Am J Roentgenol Radium motherapy (pemetrexed), and VEGFR inhibition (SU5416) Ther Nucl Med 114:83–88 in human endothelial and tumor cells. Int J Radiat Oncol Gonzalez Baron M et al (1993) Phase I study of UFT plus leu- Biol Phys 60:1220–1232 covorin in advanced colorectal cancer: a double modula- Blum JL (2001) The role of capecitabine, an oral, enzymatically tion proposal. Anticancer Res 13:759–762 activated fluoropyrimidine, in the treatment of metastatic Hall TC et al (1967) A clinical pharmacologic study of che- breast cancer. Oncologist 6:56–64 motherapy and X-ray therapy in lung cancer. Am J Med Bruso CE et al (1990) Fluorodeoxyuridine-induced radiosensi- 43:186–193 tization and inhibition of DNA double strand break repair Hanauske AR et al (2001) Pemetrexed disodium: a novel in human colon cancer cells. Int J Radiat Oncol Biol Phys antifolate clinically active against multiple solid tumors. 19:1411–1417 Oncologist 6:363–373 32 H. Harada et al.

Haraguchi M et al (1994) Angiogenic activity of enzymes. Krook JE et al (1991) Effective surgical adjuvant therapy for Nature 368:198 high-risk rectal carcinoma. N Engl J Med 324:709–715 Hazarika M et al (2005) Pemetrexed in malignant pleural Kunz BA (1982) Genetic effects of deoxyribonucleotide pool mesothelioma. Clin Cancer Res 11:982–992 imbalances. Environ Mutagen 4:695–725 Heidelberger CL et al (1958) Studies on fluorinated pyrimidines. Latz D et al (1998) Radiosensitizing potential of gemcitabine II. Effects on transplanted tumors. Cancer Res 18:305–317 (2’,2’-difluoro-2’-deoxycytidine) within the cell cycle in Heimburger DK et al (1991) The effect of fluorodeoxyuridine vitro. Int J Radiat Oncol Biol Phys 41:875–882 on sublethal damage repair in human colon cancer cells. Lawrence TS et al (1996) Radiosensitization of pancreatic Int J Radiat Oncol Biol Phys 21:983–987 cancer cells by 2’,2’-difluoro-2’-deoxycytidine.” Int J Radiat Heinemann V et al (1988) Comparison of the cellular pharma- Oncol Biol Phys 34:867–872 cokinetics and toxicity of 2’,2’-difluorodeoxycytidine and Lawrence TS et al (1997) Delayed radiosensitization of human 1-beta-D-arabinofuranosylcytosine. Cancer Res 48:4024– colon carcinoma cells after a brief exposure to 2’,2’-diflu- 4031 oro-2’-deoxycytidine (Gemcitabine). Clin Cancer Res Heinemann V et al (1990) Inhibition of ribonucleotide reduc- 3:777–782 tion in CCRF-CEM cells by 2’,2’-difluorodeoxycytidine. Lennard L (1992) The clinical pharmacology of 6-mercapto- Mol Pharmacol 38:567–572 purine. Eur J Clin Pharmacol 43:329–339 Helsper JT, Sharp GS (1962) Combination therapy with 5- Levine AJ (1997) p53, the cellular gatekeeper for growth and fluorouracil and cobalt-60 for inoperable carcinoma of the division. Cell 88:323–331 lungs. Cancer Chemother Rep 20:103–106 Ling LL, Ward JF (1990) Radiosensitization of Chinese hamster Hertel LW et al (1990) Evaluation of the antitumor activity of V79 cells by bromodeoxyuridine substitution of thymi- gemcitabine (2’,2’-difluoro-2’-deoxycytidine). Cancer Res dine: enhancement of radiation-induced toxicity and DNA 50:4417–4422 strand break production by monofilar and bifilar substitu- Ho DH et al (2000) Oral uracil and Ftorafur plus leucovorin: tion. Radiat Res 121:76–83 pharmacokinetics and toxicity in patients with metastatic Linke SP et al (1997) p53 mediates permanent arrest over mul- cancer. Cancer Chemother Pharmacol 46:351–356 tiple cell cycles in response to gamma-irradiation. Cancer Hochster H et al (2005) Phase I/II dose-escalation study of Res 57:1171–1179 pemetrexed plus irinotecan in patients with advanced Longley DB et al (2003) 5-fluorouracil: mechanisms of action colorectal cancer. Clin Colorectal Cancer 5:257–262 and clinical strategies. Nat Rev Cancer 3:330–338 Hoff PM et al (2001) Comparison of oral capecitabine versus Malet-Martino M, Martino R (2002) Clinical studies of three intravenous fluorouracil plus leucovorin as first-line treat- oral prodrugs of 5-fluorouracil (capecitabine, UFT, S-1): a ment in 605 patients with metastatic colorectal cancer: review. Oncologist 7:288–323 results of a randomized phase III study. J Clin Oncol Martomo SA, Mathews CK (2002) Effects of biological DNA 19:2282–2292 precursor pool asymmetry upon accuracy of DNA replica- Houghton JA et al (1995) Ratio of 2’-deoxyadenosine-5’-tri- tion in vitro. Mutat Res 499:197–211 phosphate/thymidine-5’-triphosphate influences the com- Mason KA et al (1999) Maximizing therapeutic gain with gem- mitment of human colon carcinoma cells to thymineless citabine and fractionated radiation. Int J Radiat Oncol Biol death. Clin Cancer Res 1:723–730 Phys 44:1125–1135 Huang P et al (1991) Action of 2’,2’-difluorodeoxycytidine on Mendelsohn LG et al (1999) Enzyme inhibition, polyglutama- DNA synthesis. Cancer Res 51:6110–6117 tion, and the effect of LY231514 (MTA) on purine biosyn- Ichinose Y et al (2005) UFT plus gemcitabine combination thesis. Semin Oncol 26 (2 Suppl 6):42–47 chemotherapy in patients with advanced non-small-cell Merlin T et al (1998) Cell cycle arrest in ovarian cancer cell lung cancer: a multi-institutional phase II trial. Br J Cancer lines does not depend on p53 status upon treatment with 93:770–773 cytostatic drugs. Int J Oncol 13:1007–1016 Iliakis G et al (1989) Mechanism of radiosensitization by halo- Miller RC et al (2002) Acute diarrhea during adjuvant genated pyrimidines: effect of BrdU on radiation induction therapy for rectal cancer: a detailed analysis from a ran- of DNA and chromosome damage and its correlation with domized intergroup trial. Int J Radiat Oncol Biol Phys cell killing. Radiat Res 119:286–304 54:409–413 Jin S, Levine AJ (2001) The p53 functional circuit. J Cell Sci Mitrovski B et al (1994) Biochemical effects of folate-based 114:4139–4140 inhibitors of thymidylate synthase in MGH-U1 cells. Jolivet J et al (1983) The pharmacology and clinical use of Cancer Chemother Pharmacol 35:109–114 methotrexate. N Engl J Med 309:1094–1104 Miwa M et al (1998) Design of a novel oral fluoropyrimidine Kanamaru R et al (1986) The inhibitory effects of 5-fluoro- carbamate, capecitabine, which generates 5-fluorouracil uracil on the metabolism of preribosomal and ribosomal selectively in tumours by enzymes concentrated in human RNA in L-1210 cells in vitro. Cancer Chemother Pharmacol liver and cancer tissue. Eur J Cancer 34:1274–1281 17:43–46 Moertel CG et al (1969) Combined 5-fluorouracil and super- Kaye SB (1994) Gemcitabine: current status of phase I and II voltage radiation therapy of locally unresectable gastroin- trials. J Clin Oncol 12:1527–1531 testinal cancer. Lancet 2:865–867 Kitazono M et al (1998) Prevention of hypoxia-induced apop- Moertel CG et al (1981) Therapy of locally unresectable pan- tosis by the angiogenic factor thymidine phosphorylase. creatic carcinoma: a randomized comparison of high dose Biochem Biophys Res Commun 253:797–803 (6000 rads) radiation alone, moderate dose radiation (4000 Kremer JM (1994) The mechanism of action of methotrexate rads + 5-fluorouracil), and high dose radiation + 5-fluo- in rheumatoid arthritis: the search continues. J Rheumatol rouracil: The Gastrointestinal Tumor Study Group. Cancer 21:1–5 48:1705–1710 Combinations of Antimetabolites and Ionizing Radiation 33

Moghaddam A et al (1995) Thymidine phosphorylase is angio- Sawada N et al (1998) Induction of thymidine phosphory- genic and promotes tumor growth. Proc Natl Acad Sci USA lase activity and enhancement of capecitabine efficacy by 92:998–1002 taxol/taxotere in human cancer xenografts. Clin Cancer Res Monnerat C et al (2004) Phase II study of pemetrexed-gem- 4:1013–1019 citabine combination in patients with advanced-stage non- Sawada N et al (1999) X-ray irradiation induces thymidine small cell lung cancer. Clin Cancer Res 10:5439–5446 phosphorylase and enhances the efficacy of capecitabine Muggia FM et al (1996) Phase I and pharmacokinetic study (Xeloda) in human cancer xenografts. Clin Cancer Res of oral UFT, a combination of the 5-fluorouracil prodrug 5:2948–2953 tegafur and uracil. Clin Cancer Res 2:1461–1467 Scagliotti GV et al (2005) Pemetrexed combined with oxalipla- Naida JD et al (1998) The effect of activation of wild-type p53 tin or carboplatin as first-line treatment in advanced non- function on fluoropyrimidine-mediated radiosensitiza- small cell lung cancer: a multicenter, randomized, phase II tion. Int J Radiat Oncol Biol Phys 41:675–680 trial. Clin Cancer Res 11:690–696 O’Connell MJ et al (1994) Improving adjuvant therapy for Schuller J et al (2000) Preferential activation of capecitabine rectal cancer by combining protracted-infusion fluoroura- in tumor following oral administration to colorectal cil with radiation therapy after curative surgery. N Engl J cancer patients. Cancer Chemother Pharmacol 45:291– Med 331:502–507 297 Oettle H et al (2005) A phase III trial of pemetrexed plus Segal R et al (1990) Methotrexate: mechanism of action in gemcitabine versus gemcitabine in patients with unre- rheumatoid arthritis. Semin Arthritis Rheum 20:190–200 sectable or metastatic pancreatic cancer. Ann Oncol Shewach DS, Lawrence TS (1995) Radiosensitization of human 16:1639–1645 tumor cells by gemcitabine in vitro. Semin Oncol 22 (4 Ostruszka LJ, Shewach DS (2000) The role of cell cycle pro- Suppl 11):68–71 gression in radiosensitization by 2’,2’-difluoro-2’-deoxy- Shewach DS et al (1992a) Nucleotide specificity of human cytidine. Cancer Res 60:6080–6088 deoxycytidine kinase. Mol Pharmacol 42:518–524 Pauwels B et al (2003) Cell cycle effect of gemcitabine and its Shewach DS et al (1992b) Decrease in TTP pools mediated by role in the radiosensitizing mechanism in vitro. Int J Radiat 5-bromo-2’-deoxyuridine exposure in a human glioblas- Oncol Biol Phys 57:1075–1083 toma cell line. Biochem Pharmacol 43:1579–1585 Pawlik TM, Keyomarsi K (2004) Role of cell cycle in mediat- Shewach DS et al (1994) Metabolism of 2’,2’-difluoro-2’-deox- ing sensitivity to radiotherapy. Int J Radiat Oncol Biol Phys ycytidine and radiation sensitization of human colon car- 59:928–942 cinoma cells. Cancer Res 54:3218–3223 Pazdur R et al (1998) Phase I trial of uracil-tegafur (UFT) plus Shih C et al (1997) LY231514, a pyrrolo[2,3-d]pyrimidine- oral leucovorin: 28-day schedule. Cancer Invest 16:145– based antifolate that inhibits multiple folate-requiring 151 enzymes. Cancer Res 57:1116–1123 Plunkett W et al (1995) Preclinical characteristics of gem- Shirasaka T et al (1993) Inhibition by oxonic acid of gastroin- citabine. Anticancer Drugs 6 (Suppl 6):7–13 testinal toxicity of 5-fluorouracil without loss of its antitu- Robinson BW, Shewach DS (2001) Radiosensitization by gem- mor activity in rats. Cancer Res 53:4004–4009 citabine in p53 wild-type and mutant MCF-7 breast carci- Sinclair WK (1968a) The combined effect of hydroxyurea noma cell lines. Clin Cancer Res 7:2581–2589 and X-rays on Chinese hamster cells in vitro. Cancer Res Rosenthal DI et al (2000) Postoperative radiation therapy for 28:198–206 rectal cancer combined with UFT/leucovorin. Oncology Sinclair WK (1968b) Cyclic X-ray responses in mammalian (Williston Park) 14 (10 Suppl 9):59–62 cells in vitro. Radiat Res 33:620–643 Rosier JF et al (2004) Role of 2’-2’ difluorodeoxycytidine Smith TJ et al (1998) Combined chemoradiotherapy vs radio- (gemcitabine)-induced cell cycle dysregulation in radiotherapy alone for early stage squamous cell carcinoma of enhancement of human head and neck squamous cell car- the esophagus: a study of the Eastern Cooperative Oncol- cinomas. Radiother Oncol 70:55–61 ogy Group. Int J Radiat Oncol Biol Phys 42:269–276 Rothenberg ML et al (1996) A phase II trial of gemcitabine in Sommer H, Santi DV (1974) Purification and amino acid patients with 5-FU-refractory pancreas cancer. Ann Oncol analysis of an active site peptide from thymidylate syn- 7:347–353 thetase containing covalently bound 5-fluoro-2’-deoxyuri- Rube CE et al (2004) Increased expression of pro-inflamma- dylate and methylenetetrahydrofolate. Biochem Biophys tory cytokines as a cause of lung toxicity after combined Res Commun 57:689–695 treatment with gemcitabine and thoracic irradiation. Stein JJ, Kaufman J (1968) The treatment of carcinoma of Radiother Oncol 72:231–241 the bladder with special reference to the use of preopera- Sak A et al (2000) Protection of DNA from radiation-induced tive radiation therapy combined with 5-fluorouracil. Am J double-strand breaks: influence of replication and nuclear Roentgenol Radium Ther Nucl Med 102:519–529 proteins. Int J Radiat Biol 76:749–756 Swann PF et al (1996) Role of postreplicative DNA mismatch Santi DV, Hardy LW (1987) Catalytic mechanism and inhibi- repair in the cytotoxic action of thioguanine. Science tion of tRNA (uracil-5-) methyltransferase: evidence for 273:1109–1111 covalent catalysis. Biochemistry 26:8599–8606 Taguchi T (1997) Clinical application of biochemical modula- Santi DV, McHenry CS et al (1974) Mechanism of interaction tion in cancer chemotherapy: biochemical modulation for of thymidylate synthetase with 5-fluorodeoxyuridylate. 5-FU. Oncology 54 (Suppl 1):12–18 Biochemistry 13:471–481 Takebayashi Y et al (1996) Clinicopathologic and prognostic Sato Y et al (1970) Combined surgery, radiotherapy, and significance of an angiogenic factor, thymidine phosphor- regional chemotherapy in carcinoma of the paranasal ylase, in human colorectal carcinoma. J Natl Cancer Inst sinuses. Cancer 25:571–579 88:1110–1117 34 H. Harada et al.

Takechi T et al (1997) Antitumor activity and low intestinal metastatic colorectal cancer: results of a large phase III toxicity of S-1, a new formulation of oral tegafur, in experi- study. J Clin Oncol 19:4097–4106 mental tumor models in rats. Cancer Chemother Pharma- Vermund H et al (1961) Effects of combined roentgen irra- col 39:205–211 diation and chemotherapy on transplanted tumors in Tatsumi K et al (1987) Inhibitory effects of pyrimidine, barbi- mice. Am J Roentgenol Radium Ther Nucl Med 85:599– turic acid and pyridine derivatives on 5-fluorouracil degra- 607 dation in rat liver extracts. Jpn J Cancer Res 78:748–755 Wolmark N et al (2000) Randomized trial of postoperative Tay BS et al (1969) Inhibition of phosphoribosyl pyrophos- adjuvant chemotherapy with or without radiotherapy phate amidotransferase from Ehrlich ascites-tumour cells for carcinoma of the rectum: National Surgical Adjuvant by thiopurine nucleotides. Biochem Pharmacol 18:936–938 Breast and Bowel Project Protocol R-02. J Natl Cancer Inst Tolis C et al (1999) Cell cycle disturbances and apoptosis 92:388–396 induced by topotecan and gemcitabine on human lung Yoshioka A et al (1987) Deoxyribonucleoside triphosphate cancer cell lines. Eur J Cancer 35:796–807 imbalance. 5-Fluorodeoxyuridine-induced DNA double Van Cutsem E et al (2001) Oral capecitabine compared with strand breaks in mouse FM3A cells and the mechanism of intravenous fluorouracil plus leucovorin in patients with cell death. J Biol Chem 262:8235–8241