Thymidylate Synthase Inhibitors

Vol. 2, 227-243, February 1996 Clinical Cancer Research 227

Minireview Thymidylate Synthase Inhibitors

Nikolaos Touroutoglou and Richard Pazdur 1 Introduction Division of Medicine, The University of Texas M. D. Anderson TS 2 inhibitors are a new class of compounds that target TS, Cancer Center, Houston, Texas 77030 an enzyme in the folate metabolic pathways necessary for thymidylate synthesis, specifically. By inhibiting TS, these agents decrease de novo thymidylate synthesis, which is necessary for Abstract DNA synthesis and repair. Contrary to prior empiric approaches Folate analogues that inhibit thymidylate synthase (TS) to anticancer drug discovery, the development of new TS inhib- selectively were developed based on TS and folate molecular itors was based on the targeted and specific design of new structures and properties. The structure-activity relation- molecules and relied on extensive knowledge of the structure, ship, preclinical and clinical development, and issues of conformation, and properties of TS and the mode of action of potential importance in the future success of these TS inhib- folate analogues. These compounds are being tested currently in itors are reviewed herein. Properties of these new com- preclinical and clinical settings, and their role in clinical practice pounds depend mainly on the use of the reduced folate as single agents or in combination regimens remains to be carrier (RFC) proteins for cellular entry and on intracellu- defined. This review will provide a background of the mecha- lar polyglutamation, which augments TS inhibitory function nism of TS inhibition, the development and current status of and cellular retention. CB3717 [Zeneca (formerly ICI Phar- folate analogues that inhibit TS selectively, the mechanisms of maceuticals), Macclesfield, United Kingdom], the first selec- resistance to TS inhibition, and the current approaches to aug- tive TS inhibitor developed, demonstrated antineoplastic ment the efficacy of these new compounds. activity in Phase I trials, but its development was abandoned due to nephrotoxicity. ZD1694 (Tomudex; Zeneca), a water- TS soluble, nonnephrotoxic quinazoline, demonstrated activity A critical step in the de novo pathway of DNA synthesis is in colorectal, breast, and pancreatic cancer. Phase III trials the production of the pyrimidine nucleotide TMP from dUMP. of Tomudex in advanced colorectal cancer were completed This reaction is catalyzed by the enzyme TS using the folate recently. LY231514 (Eli Lilly Research Labs, Indianapolis, cosubstrate CH2THF (Fig. 1) and is the only source of de novo cellular thymidylate. IN) uses the RFC and polyglutamation to exert its selective TS (Fig. 1) is a two-substrate enzyme that accepts both TS inhibitory action. Phase I trials of this compound have dUMP and CHzTHF to form a ternary complex through which been completed. ZD9331 (Zeneca), currently in preclinical reductive methylation of dUMP to TMP occurs. In this process, studies, was designed to obviate the use of the RFC for tetrahydrofolate is oxidized to dihydrofolate. This reaction pro- cellular entry. 1843U89 (Glaxo-Wellcome, Research Trian- vides the de novo cellular thymidylate nucleotides necessary for gle Park, NC), currently in Phase I trial, does not require the DNA synthesis and repair. TS inhibition results in a thymine- RFC; polyglutamation of this potent TS inhibitor leads to its less state, which is toxic to actively dividing cells. This cyto- higher cellular retention without augmenting its TS inhibi- toxicity is effected likely by DNA fragmentation that results tory activity. Currently in clinical trials, AG337 (p.o. and i.v. from dTTP depletion, which increases misincorporation of forms) and AG331 (both by Agouron, La Jolla, CA) are dUTP (1). Cells containing the salvage enzyme TK, which uses lipophilic potent TS inhibitors with action independent of circulating thymidine, can circumvent this cytotoxicity. How- the RFC and polyglutamation. Multiple mechanisms ever, in normal tissues and in some tumor cells, the circulating through which cells may overcome TS inhibition have been thymidine concentrations may not be sufficient to achieve a described. Combining TS inhibitors with other agents that significant effect (2). affect TS, interfere with TS gene and mRNA regulation, or The enzyme involved in regeneration of tetrahydrofolate inhibit salvage mechanisms of thymidylate depletion may from dihydrofolate is DHFR (Fig. 1). Folate analogues (antifo- potentially enable greater clinical utility of this class of lates), such as MTX, inhibit DHFR, deplete tetrahydrofolate, compounds. and cause a marked decrease in pyrimidine and purine synthesis (3-5). Mechanisms of resistance to MTX include reduced drug transport across cell membranes (6-8), reduced drug polyglu-

2 The abbreviations used are: TS, thymidylate synthase; CH2THF, 5,10- methylene tetrahydrofolate; DHFR, dihydrofolate reductase; MTX, methotrexate; 5-FU, 5-fluorouracil; FdUrd, fluorodeoxyuridine; IdUrd, Received 5/31/95; revised 9/11/95; accepted 9/27/95. iododeoxyuridine; RFC, reduced folate carrier; FPGS, folylpolygluta- 1 To whom requests for reprints should be addressed, at Division of mate synthase; FdUMP, fluoro-dUMP; TK, thymidine kinase; DLT, Medicine, Box 92, The University of Texas M. D. Anderson Cancer dose-limiting toxicity; MTD, maximum tolerated dose; PR, partial re- Center, 1515 Holcombe Boulevard, Houston, TX 77030. Phone: (713) sponse; LV, leucovorin; ICso, 50% inhibitory concentration; AZT, 792-3634; Fax: (713) 745-1827. zidovudine.

MTX

MTX (Glun) I=,,, MTX

I I "~ " ,~ I

...... t u 'I I Thymidylate Synthase I methylene I,DMFRI FH4 Glun) N5-'1° FH2 (Glun) FH4 (Glun) FH4 (Glun) A ! t- I °~ 1 dUMP dTMP

de novo Pyrimidine DNA Synthesis Synthesis

...... L..2... ~ ~ ~-iv ~ _ ..:L ~ :..~~ V N 5"I°methenyl _.., IcYCLO I N 10 formyl I G.APJAICAR I "= ~ FH4 (Glun) FH4 (Glun) -'" "'- FH4 (Glun) -__....__"--

E o Purines -r T N5 meth'¢l FH4 -'~ N5 formyl FH4

Fig. 1 Folate metabolic pathways. Inner square, role of TS in DNA synthesis. Dashed arrows, inhibitory effect. FH4, tetrahydrofolate; FH2, dihydrofolate; (Glu.), polyglutamate; MS, methionine synthase; CYCLO, cyclo,5,10,methenyl-FH4-cyclohydroxylase; GAR, glycinamide ribotide transformylase; A1CAR, aminoimidazole carboxamide ribotide transformylase.

tamation (polyglutamation confers intracellular drug retention), the intracellular level of CHaTHF, resulting in the stabilization gene mutation resulting in altered forms of DHFR to which of the ternary complex and thus a higher degree and duration of MTX cannot bind effectively, and an increase in cellular DHFR TS inhibition (13, 14). via DHFR gene amplification (9). The development of antifo- Selective TS inhibition is of interest in drug development lates that inhibit TS (rather than DHFR) selectively may cir- for a variety of reasons. Immunological quantitation using cumvent MTX tumor resistance, especially that resulting from monoclonal antibodies against TS (15, 16) revealed an inverse increased levels of DHFR. association between the treatment response to 5-FU and the TS inhibition can be achieved with either fluoropyrimi- cellular expression of TS; thus, TS overexpression of tumor dines or folate analogue compounds. The first compounds to cells can be a marker and prognostic indicator for 5-FU resis- have clinically significant TS-inhibiting activity were the flu- tance. An association between the degree of TS inhibition and oropyrimidines 5-FU and FdUrd; these are metabolized to the clinical response to chemotherapy in colorectal and breast 5-fluoro-2-deoxyuridylate (the most efficacious analogue of de- cancers has been noted (17, 18). Moreover, expression of TS oxyuridylate) and form a ternary complex together with TS and was found to predict for disease-free survival and survival in CH2THF, resulting in potent TS inhibition. These drugs also patients with rectal cancer independently (19). Folate analogues may become incorporated into RNA or DNA via fluoro-UTP are less susceptible to metabolic degradation than fluoropyrimi- (10, 11) or 5-fluoro-dUTP (12), respectively. These events may dines and are not incorporated into RNA or DNA (3). Finally, obscure quantitation of TS inhibition. Despite this, potentiation tetrahydrofolate is a relatively large molecule providing a vari- of the cytotoxic effects of 5-FU and FdUrd was predicted and ety of sites amenable to manipulation in the design of novel achieved successfully by the addition of folinic acid to increase TS-specific inhibitors.

HN--N I II

Ptefldlne Ring p-AmlnobenzolcAcid GlutamylResidues O COOH 0 10 II I . .NH__~ ,~--C- NH-- CH N CONH H 3 N I

1 COOH ZD9331

" " Tetrahydrofolate 0 COOH ?. .-(. .,,,"~ CHe~"~,. '~ l/ I H." .'T -"X ~", ]", / II I c CH H,N/~N~ IcflH CB 3717 c~oo.

NH~ ~ i(~ CIOOH COOH I - /NH"('. ,)-C- NH-- CH o / s \~ I HN~CHr~% /~'C -NH-iH CH' I II I c,, H.N,,~N,,~ ZD 1694 CH. Memotrexam COOH COOH /NH2

COOH

--c~ so~--. i"oo. j 1843U89 CH2 SO,- N O H~C \ / CH~ AG-331 Fig. 2 Chemical structures of tetrahydrofolate, MTX, 1843U89, ZD9331, CB3717, ZDl694, AG337, and AG331.

To reach and inhibit target enzymes such as DHFR or TS, tiation of TS inhibition that may exceed by 100-fold that of the antifolates enter the cell via the RFC protein transport system on monoglutamated form (23, 24). the cell membrane. Efficient or poor cellular antifolate uptake is Cellular uptake of antifolates via the RFC and the ability of determined by the molecular structure and polarity of the anti- antifolates to undergo polyglutamation have major roles in TS folate, the presence of reduced folates or antifolates (which may inhibition. A compound that has weak TS inhibitory activity in compete for the use of RFC), the intracellular folate depletion, enzyme assays may undergo rapid cellular uptake by the RFC and RFC deficiency (20). Once antifolates are inside the cell, and extensive polyglutamation by FPGS in cell cultures or in glutamate residues are added by the enzyme FPGS in a ~-car- vivo, resulting in potent TS inhibition. Alternatively, a com- boxyl linkage (Fig. 2). Polyglutamates with as many as seven or pound that is a potent TS inhibitor in enzyme assays may show eight glutamyl residues have been identified (21). These add only weak activity in vivo if it does not enter the cell readily or more negative charges to the antifolate molecule and increase is an extremely poor substrate for FPGS. Folate analogues may solubility (22), resulting in the agent's marked reduction in be regarded as prodrugs with polyglutamation as an activation efflux from cells and its prolonged retention in tissues having step (24). high FPGS activity, such as the liver. Formation of ternary TS Structure. Iterative protein crystallographic analysis complexes with dUMP and folate protects the enzyme from enabled the study of Escherichia coli TS and, subsequently, carboxypeptidase inactivation. Moreover, polyglutamation in- human TS in both the native form (25, 26) and the complex with creases the affinity of folate analogues for folate-dependent dUMP and folate cofactor or a folate analogue (26-31). The enzymes (but not for DHFR) markedly and results in a poten- primary sequence of the E. coli enzyme has high homology

(46% identity) to the human TS sequence (25), and in the human not bind TS in its absence. The significance of these findings enzyme, the active site residues are conserved, with 75% iden- remains to be determined. tity (29). The tetraglutamated forms of ZD1694 [Tomudex; Zeneca TS is a symmetric, obligate dimer of chemically identical (formerly ICI Pharmaceuticals) Macclesfield, United Kingdom] units of about 35 kilodaltons each (25). Each of the two subunits or CB3717 (Zeneca) bind to both active site regions, whereas folds into a three-layer domain anchored by a large, six- CH2THF and 1843U89 prefer one of the two sites (35). The stranded, mixed [3 sheet. A [3 sandwich is created by the back- number of binding sites for dUMP, dGMP, and GMP was only side of one sheet in juxtaposition against the corresponding face one per TS dimer in the presence of 1843U89, with the second of the equivalent sheet in the second protomer. This back-to- site being extremely weak even in the presence of folylpolyglu- back apposition of the two [3 sheets forms the interface between tamates (36). These results suggest that each of the two TS the two subunits. Although the left-handedness of the [3 sheet subunits binds folate substrates or analogues differently, and twist in each monomer is normal, the twist of one sheet relative that fact has a major role in the inhibition kinetics of each folate to the other is unexpectedly right-handed, in contrast to that of analogue. However, such studies are difficult to perform; inter- other proteins of similar structure (25, 32). Each [3 sheet is pretation of the data has in the past produced conflicting results; distorted by a series of stacked [3 bulges, dividing each sheet and this issue has been a subject of considerable debate (27, 36). into two smaller classical [3 sheets orthogonal to one another. TS Catalytic Activity. Methylene tetrahydrofolate is the The stacked [3 bulges are stabilized by hydrogen bonding and source of the methyl group to be transferred to dUMP. It is a one-carbon donor as well as the subsequent reductant of the account for the unusual right-handed twist of the opposing [3 transferred methylene group. The details of this process have sheets. An extended structure is created that anchors the phos- been deduced by studying the ternary complex of TS with phate of bound dUMP and controls the precise orientation of the FdUMP and methylene tetrahydrofolate, a stable structural an- catalytically essential sulfhydryl group on a conserved cysteine alogue of a true catalytic intermediate. Comparison of the three- residue in the active site (32). dimensional structure with that of the apoenzyme provided The unique TS portion of the ternary complex crystal form insight into the conformational changes that accompany ternary is a dimer. However, in unliganded TS crystal structures of complex formation and their significance in the reaction process humans, E. coli, Lactobacillus casei, and T 4 bacteriophage, this (26). In a different approach, the ternary complex of TS with unit is a monomer. Thus, structural differences between mono- dUMP and CB3717 (a selective TS inhibitor discussed below) mers can be observed only in the ternary complex crystal form. has been used (31). The monomers in this form are similar to one another but differ The above studies suggest that TS may activate dUMP by from the unliganded structures in mobility, reflecting mechanis- the formation of a covalent bond between the sulfhydryl group tic nonequivalence of active sites in TS (27). of the conserved cysteine residue at the TS active site and C-6 The substrate-binding site is a funnel-shape cleft extending of dUMP (see Fig. 3). This binding is induced by the folate (37) into the interior of each subunit. Components of each monomer and converts C-5 of dUMP to a powerful nucleophile (38). contribute to each of two active sites (25), which are separated Monoglutamylated folates bind to TS in an ordered fashion in widely on either side of the molecule. This fact accounts for the which dUMP binds first. However, the intracellular pool of inability to dissociate the native enzyme into catalytically active folates consists mainly of polyglutamylated molecules with af- monomers. Moreover, it suggests a way for the active sites to finities for TS that may exceed 100-fold (39). This high binding communicate in the process of ligand binding (25, 33). There is affinity of the polyglutamates results in a random binding order no conclusive evidence for cooperativity between monomers. of ligands to TS. Moreover, folate analogues that, as a result of However, ligand binding is associated with large conformational their structures, have a high affinity for TS even in the mono- changes and involves most of the domains of the protein. Thus, glutamated form can bind to the enzyme before dUMP (31). The the liganded state at both active sites will play a role in ligand binding of the polyglutamyl moiety to TS involves only a few affinity (34). Binding studies using equilibrium dialysis and localized interactions, such as hydrogen bonds and van der spectroscopy suggest that the two active sites may not react Waals forces, and seems to depend on more diffuse electrostatic equivalently, although both active sites can be filled through forces, which permit flexibility and freedom of motion (30). covalent attachment of an inhibitor and folate. Binding studies Only the first glutamate residue near the p-aminobenzoic acid of FdUMP suggest that both active sites contribute to this ring of the folate is fixed rigidly by its interaction with TS. The process in an asymmetric fashion, wherein one site is preferred additional glutamate molecules are progressively more disor- over the other and, in some cases, only one active site is used at dered. The polyglutamyl chain thus originates within the con- a time (33). In the presence of folates or folate analogues, fines of the active site cavity and stretches out along the surface binding of FdUMP to both active sites is enhanced greatly (35, of TS exposed to solvent (30). This chain acts mainly to con- 36). However, the novel folate analogue and TS inhibitor strain the motion of the cofactor but has no significant effect on 1843U89 (Glaxo-Wellcome, Research Triangle Park, NC) pro- the conformation of the ternary complex compared with that of motes the formation of ternary complexes with TS and dUMP as the monoglutamate. This action leads to augmented binding well as with selective purine nucleotides (36). There has been no affinity of the cofactor for TS, whereas the Vmax/Km value is left prior evidence that purine nucleotides bind to TS in the presence unaffected (40). or absence of folates or other folate analogues. Surprisingly, Ternary complex formation is associated with a large con- 1843U89 not only enhanced the binding of FdUMP or dUMP to formational change involving four residues at the carboxyl TS but also enabled that of dGMP, GMP, and IMP, which did terminus of the protein that close down on the distal side of the

H H "11 N~-'N~,/-N H 2

--% ~ o % _,x,Ng.~. ~ CH 2

HN O5- I 0 [ H",'S- Enz H .i~~~.~ rlbou O S-Enz I PABA~'flng~ .~ rlbose HN~ ~N H Pterin ring Fig. 3 Proposed chemical mechanism for TS. Reproduced with . ~N~ ~---N permission from The FASEB Journal (34). H."o O~',,,,...,~6N~--..': "~H H H 2 ?~--~ i4~folate

" o"~iI I~ (Jrib°se cys-198enzyme ~ :~ N~::n- z ".i~ii"Base

rlbose1

I rlbose I I I rlbose rlbose

quinazoline ring of the inhibitor, capping the active site and Development of TS Inhibitors sequestering the bound ligands from bulk solvent. These CB3717 changes are due mainly to the binding of the folate and not to Extensive analysis of TS inhibition at the molecular level dUMP (26, 31); they involve segmental accommodation of prompted the planning, synthesis, and development of an anti- several helices, [3 strands, and loops that move as units against folate compound, CB3717 (Nl°-propargyl-5 - 8-dideazafolic the [3 sheet interface between monomers. After CHzTHF binds, acid; Zeneca; Fig. 2). its imidazolidine ring opens, and a covalent bond is formed Preclinical Data. In Vitro Studies. CB3717 is a 2- between its N-5 methylene group and C-5 of dUMP (27). The amino-4-hydroxy quinazoline carrying a propargyl group steady state intermediate breaks down with a hydride transfer (-CHz-C-~CH) on N 1° that increases the affinity of this antifo- from C-6 of the cofactor to the transferred methylene group on late for the catalytic site of TS greatly (41). In vitro studies with dUMP, thus forming dTMP and releasing dihydrofolate (Fig. 3). isolated murine (L 1210) and human (W l-L2) TS demonstrated This reaction process has been studied also in the presence of FdUMP or folate analogues (26, 28). In the presence of the that CB3717 was the tightest binding, monoglutamate, folate- fluorinated dUMP analogue FdUMP, a stable covalent complex based TS inhibitor at the time of its discovery, with a K i of 2.7 forms between FdUMP, TS, and folate, and a hydride shift nM (41, 42). CB3717 inhibits TS, competing with CHzTHF for cannot occur; this results in enzyme inactivation. The fluoride of the same site on the enzyme, and has affinity for DHFR in FdUMP cannot be extracted by TS to force collapse of the enzyme assays. However, cell culture studies and clinical trials ternary complex that leads to the reaction products. With the indicated that the latter finding was not significant, because the folate analogue CB3717, the formation of a covalent bond inhibition of TS was far more potent than that of DHFR (41, 43). between TS and dUMP is still induced. However, CB3717 CB3717 does not use the RFC for cellular entry but undergoes cannot form a covalent bond with dUMP, as is the case with extensive polyglutamation by FPGS. Maximal potentiation of folate and dUMP, and a methylene group transfer cannot occur. TS inhibition (114-fold) was achieved by the triglutamate form Details of these processes are discussed extensively elsewhere of CB3717 (43). The cytotoxicity was completely reversible (26, 27, 34). with the addition of thymidine to the culture medium but was

only partially reversible with folinic acid, a finding corroborating focused on the synthesis of water-soluble analogues. This led to prior data showing potent TS inhibition to be the sole cytotoxic the development of quinazoline compounds, in which the ben- action (41, 44, 45). CB3717 caused a reduction in intracellular zene ring of the p-aminobenzoate residue (see Fig. 2) has been thymine nucleotide pools and an elevation of dUMP levels. replaced with various ring structures, including pyridine, thio- Moreover, CB3717 did not induce significant overproduction of phene, or thiazole (24, 54). After extensive biological in vitro DHFR in tumor cells, and cells resistant to MTX as a result of and in vivo evaluation of such analogues, the Nl°-methylthio- raised DHFR levels were not cross-resistant to CB3717. Ac- phene analogue ZD1694 (Tomudex; Zeneca; Fig. 2) was se- cordingly, CB3717 was a potent growth inhibitor in both MTX- lected for further development (54). This new quinazoline folate sensitive and -resistant cell lines (41). analogue inhibits TS selectively. Similar to CB3717, ZD 1694 Animal Studies. L1210 leukemia-bearing mice adminis- results in decreased TMP production, which leads to inhibition tered CB3717 achieved a long-term survival rate of 90% with- of DNA synthesis, resulting in cell death. Because Phase III out significant toxic effects, whereas optimal doses of MTX trials of this agent have been completed, its properties and elicited a long-term survival rate of only 20% (41). A daily × 5 development are discussed in detail below. dose schedule of CB3717 produced dose-related increases in Biochemistry and in Vitro Studies. ZD1694, or N-(5- mean survival time. CB3717 caused elevated deoxyuridine and [N-(3,4-dihydro-2 methyl-4-oxoquinazolin-6-ylmethyl)-N-meth- depleted thymidine levels in plasma (45). Biliary excretion was yltk; 1amino]-2-thenoyl)-L-glutamic acid, is a yellow-brown to dark the major route of drug elimination. In mice, the DLT was renal, green granular powder that is water soluble at pH 7. ZD1694 with drug accumulation and retention in the kidneys (46). Pre- behaves as a mixed noncompetitive TS inhibitor with respect to cipitation within the renal tubule resulting from the poor solu- tetrahydrofolate (42). As expected of a pure TS inhibitor and as bility of CB3717 in normal pH urine may account for its noted previously with CB3717, the addition of thymidine prevents nephrotoxicity (42, 47). the in vitro cytotoxic action of ZD1694 (55). In contrast to CB3717, Clinical Studies. Phase I Studies. Because of the vari- ZD1694 uses the RFC for cellular entry effectively, and thus, able incidence of side effects, the MTD of CB3717 was difficult higher intracellular levels are available rapidly for polyglutamation to determine with precision (48). The DLT of a 1-h infusion was by FPGS (42). ZD1694 is an excellent substrate for FPGS, with an nephrotoxicity. In >50% of patients, significant reductions in affinity 30 times higher than that of CB3717 (24, 42). In cell glomerular filtration rates occurred at doses >400 mg/m 2, al- cultures, 95% of the drug is present intracellularly in the polyglu- though milder reductions were observed at lower dose levels tamated form within 4 h of ZD1694 exposure, resulting in more (48). Malaise, associated with elevated liver enzymes, resulted potent and prolonged TS inhibition (24, 56). Because of its effec- in weakness, lethargy, arthralgia, and occasional fever. Only tive use of the RFC and FPGS, this analogue is 500-fold more mild hematological side effects were observed. A subsequent active in inhibiting cell growth than CB3717, despite being 20 Phase I trial found renal toxicity to be severe and unpredictable times less potent as a TS inhibitor in enzyme assays (Ki, 60 nM; (49); another trial demonstrated that urine alkalinization de- Ref. 42). ZD1694 also inhibits DHFR via polyglutamation but is creased the renal toxic effects but noted that profound malaise --100-fold more specific for TS than for DHFR (56). Because the was present at dose levels as low as 150 mg/m 2 (50). In these activity of ZD1694 is inhibited completely by thymidine (42), TS trials, the drug demonstrated antitumor activity in ovarian, inhibition is responsible for limiting the cell growth rate. Folinic breast, and lung cancer and mesothelioma. acid antagonizes this growth inhibition in tumor cell lines in a Phase II Studies. A Phase II study showed the activity of competitive fashion. In vitro data suggest that folinic acid competes CB3717 in advanced breast cancer, but severe renal failure with ZD1694 for both cellular uptake via the RFC and intracellular occurred in 17% of patients (51); urine alkalinization was not polyglutamation. used during this study. Other trials described the activity of Activity. ZD1694 alone was 1800 times more cytotoxic CB3717 in cisplatin-refractory ovarian cancer (resolution of to MGH-U1 bladder cancer cells after a 24-h exposure than was malignant ascites in 2 and a decrease of CA-125 levels in 1 of 5-FU plus LV (57). Progress through the cell cycle was delayed 8 patients; Ref. 50) and hepatocellular cancer (6 patients with in a dose-dependent manner in HCT-8 ileocecal human carci- decreases in tumor sizes and >50% fall in serum a-fetoprotein noma cells exposed to ZD1694 (58), and further studies showed levels, classified as partial remissions, and 1 patient with stable its significant activity in cisplatin-resistant (59) and ileocecal disease among 14 chemotherapy-naive patients; Ref. 52). cancer cells (60). The cell cycle specificity of ZD 1694-induced However, because severe renal toxic effects were unpre- DNA damage was analyzed in HCT-8 cells. Commitment to dictable, and profound malaise was noted even at lower doses, DNA replication was a prerequisite for damage, observed only further development of CB3717 was abandoned. Nevertheless, in the early to middle S-phase (61). this drug served as a prototype for studies of TS inhibition and Animal Pharmacology. Experiments in mice bearing a was the first compound to be developed rationally to inhibit TS variant lymphoma-L5178Y tumor line showed that a single dose selectively. of 10 mg/kg ZD1694 cured the animals, whereas CB3717 ex- hibited poor activity at the MTD of 200 mg/kg (62). The ZD1694 (Tomudex) rationale for a single-dose schedule is the rapid cellular transport Available preclinical and clinical data indicated that a less of ZD1694 and its metabolism to polyglutamates. ZD1694 did toxic analogue of CB3717 would be useful in cancer treatment. not demonstrate renal and hepatic toxic effects at doses of 500 Because the renal and hepatic toxic effects of CB3717 appeared mg/kg, as observed with CB3717 at 100 mg/kg (63). Human due to its poor aqueous solubility (46, 53, 54), further research tumor xenograft studies showed ZD1694 to be superior to

CB3717, 5-FU, and MTX with respect to antitumor activity and min ZD1694 infusion every 3 weeks at doses ranging from 0.1 therapeutic margin (64). to 3.5 mg/m 2. Pharmacokinetic studies revealed interpatient Animal Toxicology. Murine toxic reactions consisted of variability but little intrapatient variability with repeated dosing. leukopenia, thrombocytopenia, and damage to the small intes- The plasma concentration-versus-time profile for ZD1694 tinal mucosa. Unlike the renal and hepatic toxic effects of showed a triphasic decline following infusion and a prolonged CB3717, the hematological and gastrointestinal toxic effects of t,/2y (50-100 h), as observed in the rat and dog (71). ZD1694 could be prevented by synchronous thymidine admin- In this trial, the DLTs for ZD1694 were gastrointestinal, istration, suggesting that these effects result from the antipro- hematological, and malaise or asthenia at an MTD of 3.5 mg/m 2. liferative activity of ZD 1694 (42) and not from other physical or No correlation was observed between the pharmacokinetic pa- chemical properties. Moreover, folinic acid administration con- rameters and the increases in liver enzymes. However, the long ferred some improvement in neutrophil and platelet counts as t,/2y of the drug may be responsible for prolonged hematopoietic well as in gastrointestinal toxic reactions in ZD1694-treated stem cell inhibition or gastrointestinal toxic effects. There was BALB/c mice (65, 66), suggesting possible "rescue" activity of no statistically significant difference in the pharmacokinetic thymidine or folinic acid in patients with severe toxic reactions parameters (plasma concentration-versus-time profiles, peak to ZD 1694. plasma concentration, and terminal phase tv2) or toxic effects In dogs treated with the daily × 5 schedules of ZD1694 between patients with normal versus mild to moderate hepatic (0.5 and 1 mg/mZ/day), myelosuppression occurred 4 days after dysfunction (bilirubin, 1.25-3 times the upper limit of normal; the last dose and had resolved by day 8. At the same dose aspartate aminotransferase or alanine aminotransferase, 3-10 schedules, reversible histopathological changes in the epididy- times the upper limit of normal). No dose reduction is thus mis, testis, and thymus were observed. These changes included recommended for patients with hepatic dysfunction. Contrary to the maturation arrest of spermatogenesis, absence of or reduced this, patients with mild to moderate renal impairment (defined as spermatozoa, and exfoliated seminiferous epithelial cells in the creatinine clearance <65 ml/min) had statistically significantly testis and epididymis. Thymic lymphocytosis and atrophy were higher pharmacokinetic parameters than did patients with nor- noted (67). In genetic toxicity assays, ZD1694 is not mutagenic mal renal function. The drug manufacturer recommends that but causes DNA strand breaks through its antimetabolic activity. patients with renal impairment receive reduced doses of this ZD1694 (25 mg/m 2) caused a reduction in canine fetal body drug. Among the 28 patients in this trial who received ZD1694 weight and an increase in postimplantation fetal loss. This dose doses of 3.0 and 3.5 mg/m 2, 43% developed asthenia or malaise, level was also associated with an increased incidence of major fetal soft-tissue and skeletal abnormalities; therefore, ZD1694 is 32% had diarrhea, 32% had grade 3 or 4 reversible elevations of contraindicated in pregnant women (67). liver transaminase levels, and 21% had grade 3 or 4 leukopenia. Animal Pharmacokinetics. After the i.v. administration Four patients died of complications related to drug exposure at of ZD1694 to rats and dogs, the major and probably only dose levels of 3.0 and 3.5 mg/m2; 3 of these 4 had grade 4 remaining circulatory compound was unchanged drug. Biliary diarrhea. Death was attributed to disease progression in 1 pa- excretion predominates in the rat, whereas renal and biliary tient, to cardiac decompensation in another who had a prior routes of elimination are equal in the dog. Rapid drug clearance myocardial infarction, to dehydration in the third patient, and to was noted in both the rat and dog (t,/:, 0.6 h). The drug dispo- pulmonary hemorrhage associated with grade 4 thrombocytope- sition was triphasic, with short initial phases and a relatively nia in the fourth patient (71-74). Three objective PRs were slow ~/phase in both species (t,/2"y, 22-56 and 14-29 h in the rat reported, one each in ovarian cancer, breast cancer, and adeno- and dog, respectively). The long terminal phase is probably due carcinoma of unknown origin. All patients had received prior to the slow return of ZD1694 to the plasma from a deep tissue chemotherapy (71-74). Because patients had poor tolerance of compartment or to the release of polyglutamate-containing RBC 3.5 mg/m / in this Phase I trial, the recommended Phase II dose from the marrow into the circulation. The latter has been ob- was 3.0 mg/m 2. served at 2-4 weeks in humans. Significant ZD1694 accumu- A U.S. Phase I trial of ZD1694 conducted at the National lation was not observed with repeated daily administration (63, Cancer Institute included 50 patients; 76% had colorectal cancer 68). treated previously with 5-FU-based regimens, and the others had Polyglutamates were detected in liver, kidney, and intesti- lung cancer, sarcoma, or hepatoma. ZD1694 was administered nal tissue from mice injected with 5 mg/kg [5-3H]ZD1694 and as a 15-rain infusion every 3 weeks at dose levels between 0.6 accounted for at least 75% of the total drug concentration. and 4.5 mg/m 2 (75, 76). In this trial, the DLTs for the drug, first Polyglutamated forms were retained in tissues even when most seen at 2.8 mg/m 2, were grade 3 fatigue and malaise that lasted of the drug had been cleared from the plasma (69), supporting for several days and increased in severity or duration with the use of the infrequent-dose schedule implemented subse- repeated dosing. The only other DLT was neutropenia. In con- quently in clinical trials. trast to the European Phase I trial results, in this U.S. trial, the Clinical Studies. Phase I Studies. The first Phase I trial MTD was 4.5 mg/m 2, gastrointestinal and hematological toxic of ZD1694 began in Europe in February 1991 and included 61 effects occurred less frequently, and no treatment-related deaths patients with solid tumors (colorectal, ovarian, head and neck, were reported. No complete responses or PRs were observed in upper gastrointestinal, and others), most of whom had received this trial, but two minor responses were reported (one in a prior chemotherapy (70, 71). After single doses of ZD1694, patient with colon cancer and one in another with sarcoma). polyglutamation resulted in prolonged TS inhibition in mouse Based on results of this U.S. trial, the recommended ZD1694 tumors (69); therefore, patients were treated with a single 15- dose for Phase II studies was 4.0 mg/m 2 (75, 76).

Phase H Studies. Various completed international clini- in association with myelosuppression. Moreover, folinic acid cal trials (67, 77-83) together accrued 439 patients with eight may ameliorate severe, established, ZD1694-induced toxic re- different tumor types, including 177 patients with colorectal actions (66). cancer, 46 with breast cancer, 33 with hepatocellular cancer, 31 Based on its antitumor activity and safety profile in colo- with ovarian cancer, 41 with non-small cell lung cancer, 74 with rectal cancer, ZD1694 seemed to be a promising candidate for gastric or pancreatic cancer, and 21 with small cell lung cancer. comparative Phase III trials. Preclinical data in colon cancer cell In all Phase II studies, patients received ZD1694 at 3 mg/m 2. lines suggested that ZD1694 may not follow the resistance Based on the recommendation of the National Cancer Institute patterns of 5-FU and thus could prove useful against 5-FU- Phase I trial and on the observed lack of toxicity with 3.0 resistant clones (64). The objective response rate of 26% ob- mg/m 2, 19 patients with non-small cell lung cancer received 4.0 tained with 3 mg/m 2 ZD 1694 compares well with response rates mg/m 2. The toxicity reported with the higher dose was compa- recently reported in a meta-analysis of data from advanced rable to the 3.0-mg/m 2 dose (67). colorectal cancer patients treated with 5-FU plus LV (85). The The most frequently observed adverse events or laboratory incidence of severe toxicity at this dose of ZD1694 also com- abnormalities in Phase II studies were asthenia, diarrhea, nausea pares favorably with North Central Cancer Treatment Group and vomiting, leukopenia, and elevations in liver transaminase data for 5-FU plus folinic acid (86). Contrary to 5-FU, which is levels. Twelve percent of patients experienced severe asthenia extensively metabolized, ZD 1694 is eliminated primarily by the or malaise. Overall, severe toxic effects (WHO grade 3 or 4) kidneys as an intact drug, and its clearance can be predicted by were observed as anemia in 6%, neutropenia in 14%, elevated simple studies of renal function. Moreover, ZD1694 is admin- transaminase levels in 18%, and diarrhea in 8%. Grade 3 or 4 istered easily as a single i.v. dose every 3 weeks, whereas 5-FU nausea and vomiting were seen in 12% but were controlled plus LV is often given as a weekly or 5-day i.v. regimen. easily with antiemetics. Nonspecific cutaneous effects were Phase III Studies. A European Phase III trial comparing noted in 14% as maculopapular rashes. Five patients died of 3.0 mg/m 2 ZD1694 with 5-FU plus LV in 437 patients with severe drug-related hematological suppression and sepsis com- advanced or recurrent colorectal cancer was completed recently; bined with severe gastrointestinal toxic reactions. preliminary data (with a median follow-up of 5.3 months) show In Phase II trials, objective response rates were noted in response rates of 20 and 13% for Tomudex and 5-FU, respec- 6.5% of patients with platinum-resistant ovarian cancer, 9% tively. In addition, 9.4% of patients who received Tomudex and with chemotherapy-naive, non-small cell lung cancer, 25% with 3% of those in the 5-FU plus LV group had a 40-50% reduction advanced breast cancer who had received no prior chemother- in measurable lesions. Mucositis and WHO grade 3 and 4 apy, and 26% with advanced colorectal cancer who had received neutropenia were lower in the Tomudex group (P < 0.001). 3 no prior chemotherapy (67, 77-80). The North American Phase III trial opened to accrual in March The Phase II trial of ZD1694 (3.0 mg/m 2 every 3 weeks) in 1994 and compared 3 mg/m 2 ZD1694 with 5-FU plus LV in 22 patients with previously untreated advanced pancreatic can- advanced colorectal cancer. Initially, this trial also included a cer reported PRs in 14%, all of whom continued to respond for 4-mg/m 2 ZD 1694 arm, but unexpected severe toxic effects and 4-5 months (81). The ZD1694 Phase II trial in advanced colo- concern about the safety profile led to early closing of that dose rectal cancer involved 177 patients, >80% of whom had hepatic arm. The trial continued as a two-arm study and closed to metastases (80, 84) and only 5% of whom had received adjuvant accrual on June 30, 1995. Results from this study are pending. chemotherapy. All received 3 mg/m 2 ZD1694, which was ad- Tomudex was approved in the United Kingdom for treatment of ministered without dose reduction or delay in 82% of the pa- advanced coiorectal cancer in August 1995. tients (67). Almost 50% of these 177 patients received five courses of treatment, and 20% received more than six cycles. Newer TS Inhibitory Agents Objective responses were noted in 26%, with an average re- In an effort to deal with the various mechanisms of tumor sponse duration of 6-7 months. In an additional 26%, however, resistance and to increase the activity of TS inhibition, new signs of antitumor activity (minor tumor response or decrease in compounds have been synthesized (see Fig. 2 and Table 1). tumor marker levels) were noted. Grade 3 or 4 toxic effects included asthenia in 12% and ZD9331 nausea or vomiting in 10%. Grade 4 neutropenia occurred in Antifolates that are potent TS inhibitors but do not require 3%. In 4% of patients, lower extremity cellulitis that began extensive polyglutamation to exert adequate cytotoxicity are frequently in one leg and subsequently involved both legs was presently under development; such agents can counteract re- observed. This problem responded to antibiotic treatment, al- duced polyglutamation in tumors that express low or defective though microorganisms were not cultured from the areas of FPGS (resulting in less potent TS binding and increased efflux inflammation in any of the cases. Most patients who developed of the monoglutamate from the cell). Such an antifolate is cellulitis had history of diabetes, deep venous thrombosis, or ZD9331 (Zeneca; Fig. 2), a water-soluble quinazoline TS inhib- concurrent neutropenia. Two patients died of severe diarrhea, itor that, like ZD1694, uses RFC for transport and enters cells mucositis, and hematological toxic effects; one of these had readily but, unlike it, is not a substrate for FPGS as a result of received prior radiation therapy for non-Hodgkin's lymphoma. An additional patient died; however, he did not receive aggres- sive treatment for diarrhea and dehydration (67). These trials stressed the need for early hospitalization and vigorous support of patients with severe gastrointestinal toxic effects, especially 3D. Cunningham, personal communication.

Table 1 TS inhibitors currently under development ZD1694 Compound (Tomudex) ZD9331 1843U89 AG331 AG337 Refs. 42, 73, 84 87-89 36, 91, 94 99, 101, 109 100, 104, 108

Manufacturer Zeneca Zeneca, Glaxo-Wellcome Agouron Agouron

Use of RFC Yes Yes No No No

Polyglutamation Yes No Yes (see comment) No No

Current status Phase III trial vs. Preclinical trials Phase I trials Phase I trials Phase II trials 5-FU and LV in colorectal cancer completed Comments One-step polyglutamation to Good p.o. bioavailability diglutamate only; increases intracellular retention, not potency of TS inhibition

a ~-tetrazole moiety that inhibits ~ polyglutamation. ZD9331 antifolates, increasing the number of glutamate residues of has a K i of 0.4 nM against isolated TS. The cell growth ICso of 1843U89 enhances the TS binding and inhibitory actions of this ZD9331 is 25 nM in L1210 cells and 7 nM in W1L2 cells. These compound only slightly. However, metabolism to the digluta- values increase to >4000 nM in the presence of thymidine, mate form increases 1843U89 intracellular accumulation and confirming the TS locus as the site of action of ZD9331. As retention markedly (92). expected, this activity of the antifolate is retained against the Experiments with the human TK-deficient GC3TK cell L1210:RD1694 cell line, which cannot polyglutamate antifo- line, used to circumvent problems associated with murine RFC lates (87). transport and the high circulating thymidine levels that normally Pharmacokinetic studies after i.v. bolus injection of counteract the action of the drug in mice, indicate that 1843U89 ZD9331 in mice showed a t'/:~t of 4-6 h (in contrast to 100 h for has marked in vivo antitumor activity (91). Further studies ZD1694; Ref. 88). Because ZD9331 is not retained in cells by showed its activity in human colonic WiDr adenocarcinoma polyglutamation, a prolonged infusion (3 mg/kg over 24 h) is multicellular tumor spheroids (93) and in xenografts of human required to exert a curative effect on L5178Y TK-/- tumors. ovarian SKOV-3 and colonic DLD-1 cancer in CD-1 athymic The new compound has a therapeutic ratio of -8 compared with mice (90, 94). The first Phase I study with this agent is currently 1.5 for ZD1694. Compared with antitumor doses of ZD1694, underway. which produced toxic effects to the bone marrow and small intestine, antitumor doses of ZD9331 produced only hematolog- Computer-assisted Drug Discovery ical toxicity. Murine experiments showed the activity of Historically, anticancer drug discovery relied on screening ZD9331 against a panel of human tumor xenografts, including natural or synthetic compounds using receptor- or cell-based ovarian, colorectal, gastric, and small cell lung cancers (89). assays and on developing antimetabolites based on key bio- Currently, this compound is in a preclinical stage of develop- chemical pathways and chemical formulas of involved mole- ment. cules (95). These compounds are then modified to improve their activity and toxicity profiles. X-ray crystallographic analysis of 1843U89 molecular structures achieves resolution that enables construc- 1843U89 is a novel benzoquinazoline folate analogue (Fig. tion of the three-dimensional structure of a protein, ligand, or 2) in which the pteridine nucleus of the folate molecule is receptor-ligand complex. Studies of the binding of inhibitors to replaced by a lipophilic benzoquinazoline moiety. This com- their targets and of receptor-ligand interactions provide infor- pound lacks any structural feature corresponding to the p-ami- mation that then may be used in the de novo drug design. The nobenzoyl glutamate moiety of folates (90). 1843U89 is an structure of a receptor-ligand complex can be used in computer- extremely potent, noncompetitive TS inhibitor in enzyme assays assisted graphics and molecular modeling for the design process (Ki, 90 pM), with kinetics suggesting binding to TS at a site of the ideal ligand (29). This process uses computer-predicted independent of the substrate-binding site. Growth inhibition can interaction energies between a protein surface of a known struc- be reversed by thymidine alone but not by folinic acid, indicat- ture and a functional group probe (e.g., hydroxyl, amino, or ing that TS is its exclusive site of action (91). 1843U89 enters methyl; Ref. 96). These methods are changing the process of the cells via the RFC and, although an excellent substrate for drug discovery and development. FPGS, is metabolized only one step to a diglutamate form. The AG337 and AG331. Using the above approach for drug monoglutamated form is as potent a TS inhibitor in vitro as the discovery, TS was a target for the computer-assisted, structure- polyglutamated derivatives of ZD1694 (91). In contrast to other based design of inhibitors using high-resolution, X-ray crystal-

lography and molecular mechanics to analyze the E. coli TS observed in that trial (102, 103, 105). These observations and structure (29, 97-99) and to investigate its interaction with murine data on prolonged infusion schedules suggest that pro- inhibitory ligands, including 5-FU and a classic glutamate- longed infusions are necessary to optimize antiproliferative ef- containing folic acid analogue. With computer programming, fects. molecules were then selected that would occupy the maximum Phase I studies of AG337 in a 5-day, continuous-infusion space at the TS-active site and use electrostatic forces or hydro- schedule were completed recently in the United States and gen bonding optimally to stabilize the enzyme-inhibitor com- United Kingdom (103, 105, 106). In these trials, the t,/2 of the plex. The newly synthesized compounds are lipophilic, lack a drug was <2 h. Drug clearance decreased from 190 ml/min/m2 negatively charged glutamate side chain, and are extremely for the lowest dose to 85 ml/min/m2 for the highest dose, potent TS inhibitors (ICsos in the nanomolar range). These indicating the saturable elimination and nonlinear pharmacoki- properties imply that these drugs can enter cells readily without netics of this agent (107). Myelosuppression and mucositis were the RFC and thus have activity in tumor clones deficient in this the DLTs; radiographic and biochemical evidence of antitumor carder protein. Drug potency is not dependent on polyglutama- activity also was noted. Based on these trials, the recommended tion. These drugs may exert effective TS inhibition on FPGS- Phase II dose was 1000 mg/m2/day (108). Phase II trials in deficient cells; however, the prolonged intracellular retention of malignant tumors of the head and neck, lung, pancreas, prostate, the new compounds, conferred normally by polyglutamation liver, and colon were initiated at multiple clinical centers in the and causing prolonged TS inhibition, is lost. United States in October 1994 and are currently ongoing. The new compounds, AG337 and AG331 (Agouron, La AG331. Treatment of L1210 cells with 4-10 IXM AG331 Jolla, CA; Fig. 2), are potent TS inhibitors with purified human in vitro elicits a >85% reduction in TS activity. AG331 was TS Kis of 16 tXM (100) and 2 nM (99), respectively. Both active against leukemia and colon cancer cell lines, and its compounds have minor structural similarity to the natural folate antitumor activity was observed in vivo against the i.p.-im- cofactor, and both inhibit the growth of a variety of tumor cell planted murine L5178Y lymphoma (102). In animal studies, lines; the IC5o for L1210 in vitro ranged from 0.18 to 3.5 IxU plasma AG331 concentrations of >20 ~M were tolerated with- (101). The addition of thymidine reversed both growth inhibi- out irreversible toxic effects. The toxic effects seen most fre- tion and cell cycle arrest in the early S-phase, findings consistent quently in preliminary animal investigations are mild to mod- with TS as the primary site of action (100-102). Moreover, both erate grades of local vein irritation, transient weight loss, and AG337 and AG331 were active in vitro against multidrug- histamine release resulting in flushing or swelling (102). resistant cell lines (102). AG331 is currently being evaluated in Phase I trials. A AG337. Animal Pharmacokinetics of AG33Z AG337 had excellent p.o. bioavailability (96). p.o. absorption of AG337 study of AG331 administered as a 10-min i.v. infusion to 18 was rapid, with peak plasma levels as measured by high-perfor- patients with refractory tumors has been completed (102). The mance liquid chromatography achieved in < 1 h. Comparison of tl/2 of the drug was found to be -20 h; with each dose increase, the areas under the curve after administration of i.v. or p.o. a steady increase in the area under the curve and a dose- AG337 at 75 mg/m 2 demonstrated the drug to have p.o. bio- dependent decrease in drug clearance were noted, indicating the availabilities of 100% in mice and >80% in dogs. saturable elimination kinetics of AG331 (88). Mild to moderate Animal Pharmacology and Toxicology. p.o. AG337 nausea and vomiting and mild flushing were observed at 130 demonstrated 100% curative activity against the i.p.- or i.m.- and 225 mg/m 2, respectively, and may be attributed to histamine implanted L5178Y/TK murine lymphoma and produced long release (102). The MTD has not yet been reached, and thus far, growth delays in the GC3 M/TK human colon carcinoma i.p. objective tumor responses have not been observed in the ongo- xenograft. The murine antitumor activity of AG337 was elicited ing Phase I trial of AG331 (109). at dose levels that were well tolerated, even on a repeated-dose Recent preclinical data on AG331 in H35 rat and HEPG2 schedule; high dose levels also were tolerated in dogs (100). human hepatoma cells revealed that at higher doses of this Five-day infusion schedules of AG337 in mice were well toler- antifolate, a loss of the TS inhibitory effect occurred, yet cyto- ated and induced substantial growth retardation in 90% of toxic effects were preserved. This suggests an alternative mode tumors, with a median delay of 14 days (103). These data of action for AG331 and the need to investigate the possibility corroborate in vitro AG337 studies demonstrating that a 24-h that in these hepatoma cells, this drug undergoes biotransfor- exposure to this agent was not sufficient to induce a significant mation into other, as yet unidentified, toxic metabolites (110). rate of cell death but that a 72- to 120-h exposure produced concentration-dependent cytotoxicity (102). Resistance to TS Inhibition Clinical Studies of p.o. AG33Z Preliminary data from an Multiple mechanisms of resistance to the new selective TS ongoing Phase I trial demonstrate that p.o. AG337 has a bio- inhibitors have been described, with many of the studies using availability of >80%, with peak levels achieved within 1 h and ZD 1694 as their prototype. The active transport of ZD 1694 via the drug still detectable at 6 h. Mild nausea but no myelosup- the RFC and subsequent polyglutamation by FPGS, resulting in pression was observed at 360 mg/m 2 (104). increased TS affinity and intracellular polyglutamate retention, Clinical Studies of i.v. AG33Z In a Phase I study of are responsible for the potent TS inhibition and antitumor ac- AG337 administered as a 24-h infusion, plasma deoxyuridine tivity of the drug in vitro and in vivo. Decreased uptake via the levels increased to 60-390% of pretreatment levels (an event RFC, failure to polyglutamate ZD1694, and elevation of intra- that suggests significant TS inhibition) but normalized rapidly cellular TS levels by TS gene amplification have been described after the infusion ended. No tumor antiproliferative effects were in studies of ZD1694-resistant cell lines (111, 112) and are

similar to observed mechanisms of resistance to 5-FU (113) or cells treated with 1843U89 or AG331 and LY231514 demon- MTX (114-116). Some issues have been addressed in develop- strated that it is possible to achieve higher rates of TS protein ing new folate analogues that are more potent TS inhibitors than synthesis and resistance by the same mechanism described for ZD1694. LY231514 (Eli Lilly Research Labs, Indianapolis, IN), ZD 1694 (127). These in vitro data obtained using synchronized with a K i of 3.4 riM, also uses the RFC and polyglutamation to cells suggest that TS induction may develop in such cell popu- inhibit selectively TS (117, 118). A Phase I trial of this drug has lations during treatment with most, if not all, TS inhibitors. been completed. Neutropenia was the DLT, and nonhematologi- Accumulation of TS protein may result in clinical drug resis- cal toxicities included mild fatigue, anorexia, and nausea. No tance in patients treated with 5-FU or folate analogues, and these antitumor responses were observed (119). Other drug develop- data may explain in part the low response rates to these drugs in ment strategies led to compounds such as ZD9331 (which does the clinical setting. not require polyglutamation; Refs. 87-89), 1843U89 (which TS Interaction with Cancer-related Genes. Normal or requires only one-step polyglutamation; Refs. 36, 91), and abnormal expression of certain genes, including p53 and bcl-2, AG331 (99, 109) and AG337 (which require neither transport by can affect the ultimate response of tumors to cytotoxic drugs the RFC nor polyglutamation to exert their cytotoxic action; (128-132). TS may affect these genes specifically. In vitro, Refs. 108, 120). translation of p53 mRNA was repressed specifically with the TS Induction. Induction of TS leading to increased TS addition of exogenous human TS in human colon cancer H630 activity levels as a result of TS protein overexpression has been cells, and TS was demonstrated to complex with p53 and c-myc noted as a resistance mechanism to 5-FU (18). Gene amplifica- mRNAs (133, 134). The role of this mRNA-protein interaction, tion is the primary mechanism by which DHFR levels increase although not yet clear, suggests the possibility of a regulatory after cell exposure to MTX, resulting in MTX resistance, and role of TS in p53 and c-myc cellular expression. In addition, this has been described also with TS in cell lines resistant to c-myc protein decreased and p53 and Rb proteins increased after ZD1694 (112). In addition, overexpression of TS protein as a exposure of HCT-8 cells to ZD1694 (135). Although prelimi- result of deregulation of TS mRNA translation during treatment nary, these data are intriguing, because such interactions may with a TS inhibitor is another mechanism of TS induction. have a critical role in retarding cell progression through the Translation of human TS mRNA is controlled by its own S-phase and in initiating apoptosis after drug exposure. If the end product (TS) in a negative autoregulatory manner (121, cytotoxic activity of TS inhibitors is mediated ultimately by the 122). TS protein inhibits TS mRNA by binding of the protein to induction of apoptosis, then mutations in the genes encoding specific TS mRNA sequences critical for the initiation of trans- these proteins or interference with their activity could decrease lation. TS protein can regulate its own mRNA translation. This the expected cytotoxicity. Studies of bcl-2 expression in human TS-specific process does not affect translation of unrelated lymphoma cells demonstrated that this is possible in vitro. mRNA transcripts. The intracellular level of TS protein is thus Following a 36-h exposure to cytostatic drugs, including FdUrd, critical in the regulation of TS mRNA translation. This model cells transfected with a vector alone underwent apoptosis could explain in vitro, in vivo, and clinical observations describ- readily. In contrast, cells transfected with a bcl-2 vector and ing TS level increases within cancer cells following acute ex- expressing this protein showed markedly decreased levels of posures to 5-FU (18, 123-125). FdUMP, dUMP, and selective apoptosis (8 versus 67%) independent of TS levels and TS TS inhibitors (ZD1694) may bind to TS protein, producing a enzymatic activity that were unaffected by bcl-2 expression conformational change that inhibits its binding to TS mRNA. (136). Preliminary reports from studies of p53 function in pa- Translation of TS mRNA then proceeds uncontrolled, resulting tients with advanced colorectal cancer suggest that patients with in TS accumulation and drug resistance (122, 126). Coadmin- point mutations in p53 may not respond to chemotherapy with istration of a translational (but not a transcriptional) inhibitor 5-FU and LV (137). Taken together, these data emphasize the overcame TS protein accumulation. In addition, TS protein increasing importance of a molecular understanding of intrinsic accumulation induced by ZD1694 was not reversed by thymi- insensitivity to TS inhibitors as well as of TS interaction with dine coadministration despite a complete reversal of cytotoxic- cell cycle regulatory genes. ity. This explains observations of increases in TS levels of up to Pretreatment TS Levels. Pretreatment TS protein levels 40-fold in normal breast epithelial cells and up to 10-fold in and TS gene expression predicted for responses of metastatic human breast cancer cells after ZD1694 exposure, whereas TS colorectal and primary gastric cancer to treatment with a com- mRNA levels remain unchanged (126). In these synchronized bination of 5-FU and LV and a combination of cisplatin, 5-FU, cells, TS protein induced normally during the S-phase and and LV, respectively (15). Tumors with high levels of TS decreasing subsequently during G~ was found to be expressed protein (measured by Western blot using the TS 106 monoclonal constitutively throughout the entire cell cycle after ZD1694 antibody and densitometry scanning) and TS gene expression treatment. No overexpression of TS mRNA could be demon- (quantitated by reverse transcription PCR and expressed as a strated. Once accumulated by ZD1694, TS was not induced in TS:[3-actin mRNA ratio) showed poor responses to therapy, the S-phase, nor did it decrease in G2, and it remained expressed whereas tumors with low TS protein and gene expression had throughout the cell cycle. Thus, TS accumulation was potenti- more responsive disease. These data are preliminary, are based ated by TS expression throughout the entire cycle after ZD 1694 on small numbers, and will require validation in larger clinical treatment. trials. However, they do suggest that, in the future, clinicians These effects were attributed to a loss of inhibition of TS may be able to predict responses to 5-FU-containing regimens mRNA translation, whereas TS gene transcription remained based on pretreatment TS protein levels or TS gene expression. unchanged. Subsequent studies of human manamary epithelial Together with the data from TS induction after exposure to TS

inhibitors before and after therapy, quantitation of TS levels pathways. Pretreatment of HCT-8 colon carcinoma cells with may help in further defining of the role of TS regulation in drug ZD1694 was reported to enhance the cytotoxicity of 5-FU by resistance. These correlations will need to be investigated for potentiating the efficacy of TS inhibition (142). Following cell applicability with the newer TS inhibitors. treatment with ZD1694, a strong TS inhibition leads to an The role of each of the numerous and diverse mechanisms altered balance of dUMP and dTTP pools and allows for an of TS regulation and resistance to TS inhibition remains to be increased incorporation of 5-FU into RNA (2.5-3.0-fold that of defined. It is not yet clear which factors predispose cells to cells not exposed to ZD1694). This event has an immediate develop a certain resistance mechanism or which mechanism(s) effect on short-lived mRNAs (such as that of TK), leading to may at least in part account for the relatively low response rates cell death, and this approach potentiated the effect of ZD1694 seen with ZD1694. However, the data presented above suggest on sensitive cell lines markedly (143). Plans are in development that TS inhibition alone may not be sufficient to result in to test this sequence-dependent synergistic combination in the clinically significant cytotoxicity. A limited number of strate- clinic.4 gies (currently in preclinical development) attempting to im- Other strategies have focused simply on potentiating TS prove the efficacy of TS inhibitors are presented briefly below. inhibition or using agents synergistic with TS inhibitors. Be- Moreover, elucidation of intrinsic tumor properties pertinent to altered signal transduction and metastatic potential (including cause folate analogues bind to TS at a site different from that of tumor-induced angiogenesis) may lead to new therapeutic strat- FdUMP, combining 5-FU with selective TS inhibitors could egies. potentiate the action of these agents. TS inhibition may be Improving Drug Efficacy. To counteract TS induction, enhanced by using both a dUMP and a folate analogue to the development of an antisense RNA that targets the TS mRNA stabilize the TS-cofactor-substrate ternary complex. When sequences responsible for TS translation has been suggested AG331 or polyglutamated ZD1694 was combined with 5-FU, (126). In another approach using gene therapy, W1L2:R179 enhanced FdUMP binding to TS was observed. The ternary cells resistant to TS inhibitors through a 30-fold increase in TS complex consisting of TS, FdUMP, and AG337 or ZD1694 was activity over that of their parental W1L2 cell line were trans- stable for at least 2 h. This led to potentiation of the growth- fected with the TS ribozyme using an EBV-based vector. This inhibitory effect of 5-FU, AG337, or ZD1694 on colon cancer hammerhead ribozyme cleaves the CUC sequences in a triple, cell lines (144). tandemly repeated sequence located in the 5' untranslated re- In agreement with the above studies, a Phase I trial of gion of TS mRNA. Thus, the transfected cells became sensitive Tomudex in combination with 5-FU, folinic acid, and IFN-a for to TS inhibitors (138). IFN-~/was shown to enhance the sensi- recurrent adult solid tumors is currently underway (145). tivity of colon cancer ceils to 5-FU by interfering with acute TS induction associated with 5-FU exposure and thus may prove to Conclusion have a role in dealing with this problem (123). Whether these Inhibition of TS results in a decrease of the de novo interventions will prove clinically useful in efforts to overcome the increased TS mRNA translation noted with TS inhibition thymidylate synthesis that is necessary for DNA synthesis and requires further investigation. Alternatively, a 40-fold increase repair. The development of ZD1694 is an example of a lengthy, in TS levels in normal cells as mentioned above could render targeted approach to the design of new cancer chemotherapeutic them theoretically more resistant to the TS inhibitor, thus re- agents; ZD1694, a potent, selective TS inhibitor, has demon- suiting in a host-protective effect, offering the possibility for strated antitumor effects in both preclinical and Phase II clinical cell-selective treatment strategies. trials and is undergoing Phase III evaluation currently in the Combining the inhibition of de novo and salvage pathways United States. More detailed data from the European Phase III of thymidylate synthesis can provide a synergistic effect. The trial in colorectal cancer will be available for critical review activity of salvage enzymes can be markedly higher than that of soon. The outcome of these studies may contribute to more the rate-limiting enzymes of de novo biosynthesis (139). AZT convenient and less toxic alternatives to the currently available inhibits TK, an enzyme that is up-regulated and involved in chemotherapy for colorectal cancer. Newer antifolate com- salvage pathways of TS inhibition, effectively. Although AZT pounds with diverse structures and properties designed and alone did not inhibit TS, studies combining ZD1694 with AZT developed by computer-assisted techniques are also undergoing in MGH-U1 human bladder cancer cells showed the combina- preclinical and clinical testing. However, given the limitations tion to be more cytotoxic (140). Up-regulation of nucleoside posed by the variety of mechanisms through which cells bypass transporter may lead to increased uptake of preformed deoxy- TS inhibition, it is questionable whether each of these drugs will thymidine or its analogue IdUrd in vitro (141). IdUrd alone did be more successful than single-agent 5-FU or MTX. Combining not inhibit TS activity in MGH-U1 and HCT-8 cells. DNA TS inhibitors with other cytotoxic agents, agents that interfere incorporation of this nucleoside analogue increased with admin- with TS gene and mRNA regulation, or agents that inhibit istration of ZD1694 due to up-regulation of nucleoside trans- salvage mechanisms following thymidylate depletion may pro- porter. IdUrd did not increase the TS inhibitory activity of vide greater clinical utility of these new compounds. ZD1694, but its DNA incorporation resulted in an increase in DNA single-strand breaks that paralleled cytotoxicity. More- over, the therapeutic index was greater for the combination than for either agent alone (141). Sequential combinations of 5-FU and ZD1694 also have been tested for inhibition of salvage 4j. R. Bertino, personal communication.

Acknowledgments 19. Johnston, P. G., Fisher, E. R., Rockette, H. E., Fisher, B., Wolmark, N., Drake, J. C., Chabner, B. A., and Allegra, C. The role of thymidylate We thank Diane Rivera for her assistance in preparing the manu- synthase expression in prognosis and outcome of adjuvant chemother- script and Y. Ishida for her assistance with the illustrations. apy in patients with rectal cancer. J. Clin. Oncol., 12: 2640-2647, 1994. 20. Jansen, G., Westerhof, G. R., Jarmuszewski, M. J., Kathmann, I., References Rijksen, G., and Schornagel, J. H. Methotrexate transport in variant 1. Ayusawa, D., Arai, H., Wataya, Y., and Sento, T. A specialized form human ccrf-cem leukemia cells with elevated levels of the reduced of chromosomal DNA degradation induced by thymidylate stress in folate carrier. Selective effect on carrier-mediated transport of physio- mouse FM3A cells. Mutat. Res., 200: 221-230, 1988. logical concentrations of reduced folates. J. Biol. Chem., 265: 18272- 18277, 1990. 2. Grem, J. L. Fluorinated pyrimidines. In: B. A. Chabner and J. M. Collins (eds.), Cancer Chemotherapy: Principles and Practice, pp. 180- 21. Balinska, M., Galivan, J., and Coward, J. K. Effiux of methotrexate 224. Philadelphia: J. B. Lippincott Company, 1990. and its polyglutamate derivatives from hepatic cells in vitro. Cancer Res., 41: 2751-2756, 1981. 3. Harrap, K. R., Jackman, A. L., Newell, D. R., Taylor, G. A., Hughes, 22. Hoffbrand, A. V., Tripp, E., Houlihan, C. M., and Scott, J. M. L. R., and Calvert, A. H. Thymidylate synthase: a target for anticancer Studies on the uptake of synthetic conjugated folates by human marrow drug design. Adv. Enzyme Regul., 29: 161-179, 1989. cells. Blood., 42: 141-146, 1973. 4. Roth, B., and Cheng, C. C. Recent progress in the medicinal chem- 23. Calabresi, P., and Chabner, B. A. Antineoplastic agents. In: A. istry of 2,4-diaminopyridines. Prog. Med. Chem., 19: 269-331, 1982. Goodman Gilman, T. W. Rall, A. S. Nies, and P. Taylor (eds.), Good- 5. Sirotnak, F. M., Burchall, J. J., Ensminer, W. D., and Montgomery, man and Gilman's The Pharmacologic Basis of Therapeutics, pp. 1223- J. A. Folate Antagonists as Therapeutic Agents, Vol. 1. Cambridge: 1227. New York: Pergamon Press, 1990. Academic Press, 1984. 24. Jackman, A. L., Marsham, P. R., Moran, R. G., Kimbell, R., 6. Fischer, G. A. Defective transport of amethopterin (methotrexate) as O'Connor, B. M., Hughes, L. R., and Calvert, A. H. Thymidylate a mechanism of resistance to antimetabolites in L5178Y cells. Biochem. synthase inhibitors: the in vitro activity of a series of heterocyclic Pharmacol., 11: 1233-1234, 1962. benzoyl ring modified 2-desamino-2-methyl-Nl°-substituted-5,8-didea - 7. Harrap, K. R., Hill, B. T., Furness, M. E., and Hart, L. I. Sites of zafolates. Adv. Enzyme Regul., 31: 13-27, 1991. action of amethopterin in intrinsic and acquired drug resistance. Ann. 25. Hardy, L. W., Finer-Moore, J. S., Montfort, W. R., Jones, M. O., NY Acad. Sci., 186: 312-324, 1971. Santi, D. V., and Stroud, R. M. Atomic structure of thymidylate syn- 8. Kessel, D., Hall, T. C., and Roberts, D. Modes of uptake of metho- thase: target for rational drug design. Science (Washington DC), 235: trexate by normal and leukemic human leukocytes in vitro and their 448-455, 1987. relation to drug response. Cancer Res. 28: 564-570, 1968. 26. Matthews, D. A., Villafranca, J. E., Janson, C. A., Smith, W. W., 9. Bertino, J. R., and Romanini, A. Chemotherapeutic agents: folate Welsh, K., and Freer, S. Stereochemical mechanism of action for antagonists. In: J. F. Holland, E. Frei III, R. C. Bast, Jr., D. W. Dufe, D. thymidylate synthase based on the X-ray structure of the covalent Morton, and R. Weichselbaum (eds.), Cancer Medicine, Vol. 1, pp. inhibitory ternary complex with 5-fluoro-2'-deoxyuridylate and 5,10- 698-711. Philadelphia: Lea & Febiger, 1993. methylenetetrahydrofolate. J. Mol. Biol., 214: 937-948, 1990. 27. Montfort, W. R., Perry, K. M., Fauman, E. C., Finer-Moore, J. S., 10. Heidelberger, C. Fluorinated pyrimidines. Prog. Nucleic Acid Res., Maley, G. F., Hardy, L., Maley, F., and Stroud, R. M. Structure, multiple 4: 1-50, 1965. site binding, and segmental accommodation in thymidylate synthase on 11. Heidelberger, C. Chemical carcinogenesis, chemotherapy: cancer's binding dUMP and an anti-folate. Biochemistry, 29: 6964-6977, 1990. continuing core challenges--G. H. A. Clowes Memorial Lecture. Can- 28. Matthews, D. A., Appelt, K., Oatley, S. J., and Xuong, N. H. Crystal 30: cer Res., 1549-1569, 1970. structure of Escherichia coli thymidylate synthase containing bound 12. Schuetz, J. D., Wallace, H. J., and Diasui, R. B. 5-Fluorouracil 5-fluoro-2'-deoxyuridylate and 10-propargyl-5,8-dideazafotate. J. Mol. incorporation into DNA of CF- 1 mouse bone marrow cells as a possible Biol., 214: 923-936, 1990. mechanism of toxicity. Cancer Res., 44: 1358-1363, 1984. 29. Appelt, K., Bacquet, F. J., Bartlett, C. A., Booth, C. L. J., and Freer, 13. Peters, J. G., van Groeningen, C. J., Van der Wilt, C., Meijer, S., S. T. Design of enzyme inhibitors using iterative protein crystallo- Smid, K., Laurensse, E., and Pinedo, H. M. Time course of inhibition of graphic analysis. J. Med. Chem., 34: 1925-1934, 1991. thymidylate synthase in patients treated with fluorouracil and leucov- 30. Kamb, A., Finer-Moore, J., Calvert, H., and Stroud, R. Structural orin. Semin. Oncol., 19 (2 Suppl. 3): 26-35, 1992. basis for recognition of polyglutamyl folates by thymidylate synthase. 14. Keyomarsi, K., and Moran, R. G. Folinic acid augmentation of the Biochemistry, 31: 9883-9890, 1992. effects of fluoropyrimidines on murine and human leukemic cells. 31. Kamb, A., Finer-Moore, J. S., and Stroud, R. Cofactor triggers the Cancer Res., 46: 5229-5235, 1986. conformational change in thymidylate synthase: implications for an 15. Johnston, P. G., Lenz, H. J., Leichman, C. G., Danenberg, K., ordered binding mechanism. Biochemistry, 31: 12876-12884, 1992. Allegra, C. J., Danenberg, P., and Leichman, L. Thymidylate synthase 32. Matthews, D. A., Appelt, K., and Oatley, S. J. Stacked [3-bulges in gene and protein expression correlate and are associated with response thymidylate synthase account for a novel fight-handed rotation between to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res., opposing [3 sheets. J. Mol. Biol., 205: 449-454, 1989. 55: 1407-1412, 1995. 33. Danenberg, K. D., and Danenberg, P. V. Evidence for a sequential 16. Johnston, P. G., Drake, J. C., Trepel, J., and Allegra, C. J. The interaction of the subunits of thymidylate synthetase. J Biol Chem., 254: immunological quantitation of thymidylate synthase using the monoclo- 4335-4348, 1979. nal antibody TS106 in 5-FU-sensitive and -resistant human cancer cell 34. Stroud, R. M., and Finer-Moore, J. S. Stereochemistry of a multi- lines. Cancer Res., 52: 4302-4312, 1992. step/bipartite methyl transfer reaction: thymidylate synthase. FASEB J., 17. Peters, G. J., van der Wilt, C. L., van Groeningen, C. J., Smid, K., 7: 671-677, 1993. Meijer, S., and Pinedo, H. M. Thymidylate synthase inhibition after 35. Dev, I. K., Dallas, W. S., Ferone, R., Hanlon, M., McKee, D. D., administration of fluorouracil with or without leucovorin in colon cancer and Yates, B. B. Mode of binding of folate analogs to thymidylate patients: implications for treatment with fluorouracil. J. Clin. Oncol., synthase. Evidence of two asymmetric but interactive substrate binding 12: 2035-2042, 1994. sites. J. Biol. Chem., 269: 1873-1882, 1994. 18. Swain, S. M., Lippman, M. E., Egan, E. F., Drake, J. C., Steinberg, 36. Weichsel, A., Montfort, W., Ciesla, J., and Maley, F. Promotion of S. M., and Allegra, C. J. Fluorouracil and high-dose leucovorin in purine nucleotide binding to thymidylate synthase by a potent folate previously treated patients with metastatic breast cancer. J. Clin. Oncol., analogue inhibitor, 1843U89. Proc. Natl. Acad. Sci. USA, 92: 3493- 7: 890-899, 1989. 3497, 1995.

37. Pogolotti, A. L. J., Danenberg, P. V., and Santi, D. V. Kinetics and 54. Marsham, P. R., Hughes, L. R., Jackman, A. L., Hayter, A. J., mechanism of interaction of 10-propargyl-5,8-dideazafolate with thy- Oldfield, J., Wardleworth, J. M., Bishop, J. A. M., O'Connor, B. M., and midylate synthase. J. Med. Chem., 29: 478-482, 1986. Calvert, A. H. Quinazoline antifolate thymidylate synthase inhibitors: 38. Santi, D. V., McHenry, C. S., Raines, R. T., and Ivanetich, K. M. heterocyclic benzoyl ring modifications. J. Med. Chem., 34: 1594- Kinetics and thermodynamics of the interaction of 5-fluoro-2'-deoxyuri- 1605, 1991. dylate with thymidylate synthase. Biochemistry, 26: 8606-8613, 1987. 55. Jackman, A. L., Taylor, G. A., Bishop, J. A., O'Connor, B. M., 39. Sikora, E., Newell, D. R., Jackman, A. J., Simmons, A. J., Jones, Bisset, G., Hughes, L. R., Moran, R. G., and Calvert, A. Biochemical T. R., and Calvert, A. H. Formation and retention and biological activity aspects of a new thymidylate synthase inhibitor, D1694. Proc. Am. of Nm-propargyl-5,8-dideazafolic acid (CB3717) polyglutamates in Assoc. Cancer Res., 31: 342, 1990. L1210 cells in vitro. Biochem. Pharmacol., 37: 4047-4054, 1988. 56. Jackman, A. L., Taylor, G. A., Gibson, W., Kimbell, R., Brown, M., 40. Lu, Y-Z., Aiello, P. D., and Matthews, R. G. Studies on the Calvert, A. H., Judson, I. R., and Hughes, L. R. ICI D1694, a quinazo- polyglutamate specificity of thymidylate synthase from fetal pig liver. line antifolate thymidylate synthase inhibitor that is a potent inhibitor of Biochemistry, 23: 6870-6876, 1984. L 1210 tumor cell growth in vitro and in vivo: a new agent for clinical study. Cancer Res., 51: 5579-5586, 1991. 41. Jones, T. R., Calvert, A. H., Jackman, A. L., Brown, S. J., Jones, M., and Harrap, K. R. A potent antitumour quinazoline inhibitor of thymi- 57. Erlichman, C., and Mitrovski, B. Biochemical effects of D1694 dylate synthase: synthesis, biological properties and therapeutic results inhibition of thymidylate synthase in MGH-UI bladder cancer cells. in mice. Eur. J. Cancer, 17: 11-19, 1981. Proc. Am. Assoc. Cancer Res., 33: 507, 1992. 42. Jackman, A. L., Jodrell, D. I., Gibson, W., and Stephens, T. C. ICI 58. A_rredondo, M. A., Guimaraes, M. A., Yin, M. B., Slocum, H. K., D1694, an inhibitor of thymidylate synthase for clinical study. In: R. A. and Rustum, Y. M. Prolonged cell cycle phase perturbation induced by Harkness, G. B. Elion, and N. Zollner (eds.), Purine and Pyrimidine the quinazoline antifolate thymidylate synthase inhibitor ICI D1694. Metabolism in Man, Vol. VII, Part A, pp. 19-23. New York: Plenum Proc. Am. Assoc. Cancer Res., 33: 428, 1992. Press, 1991. 59. Kelland, L. R., Abel, G., Brown, M., and Jackman, A. L. Cytotox- 43. Calvert, A. H., Newell, D. R., Jackman, A. L., Gumbrell, L. A., icity of ICI D 1694 and cisplatin in sensitive and acquired resistant Sikora, E., Grzelakowska-Sztabert, B., Bishop, J. A. M., Judson, I. R., tumor cell lines. Br. J. Cancer, 65 (Suppl. 16): 64, 1992. Harland, S., and Harrap, K. R. Recent preclinical and clinical studies 60. Zhang, Z. G., Yin, M. B., and Rustum, Y. M. Cytotoxicity, DNA with the thymidylate synthase inhibitor Nl°-propargyl-5,8-dideazafolic strand breaks, thymidylate synthase inhibition and methylenetetrahydro- Acid (CB 3717). Natl. Cancer Inst. Monogr., 5: 213-218, 1987. folate pool decrease induced by ICI D 1694. Proc. Am. Assoc. Cancer 44. Jackson, R. C., Jackman, A. L., and Calvert, A. H. Biochemical Res., 33: 410, 1992. effects of a quinazoline inhibitor of thymidylate synthetase, CB3717, on 61. Matsui, S., Lu, K., Wrzosek, C., and Rustum, Y. Double-strand human lymphoblastoid cells. Biochem. Pharmacol., 32: 3783-3790, DNA damage induced by thymidylate synthase (TS) inhibitor, ZD1694, 1983. in human colon carcinoma cell, HCT-8, requires DNA replication. Proc. 45. Jackman, A. L., Taylor, G. A., Calvert, A. H., and Harrap, K. R. Am. Assoc. Cancer Res., 36: 378, 1995. Modulation of antimetabolite effects: effects of thymidine on the effi- 62. Stephens, T. C., Calvete, J. A., Janes, D., Waterman, S. E., Val- cacy of the quinazoline-based thymidylate synthetase inhibitor, CB caccia, B. E., Hughes, L. R., and Calvert, A. H. Antitumor activity of a 3717. Biochem. Pharmacol., 33: 3269-3275, 1984. new thymidylate synthase (TS) inhibitor, D1694. Proc. Am. Assoc. 46. Newell, D. R., Alison, D. L., Calvert, A. H., Harrap, K. R., Jarman, Cancer Res., 31: 342, 1990. M. J., Jones, T. R., Manteuffel-Cymborowska, M., and O'Connor P. 63. Jodrell, D. I., Newell, D. R., Calvete, J. A., Stephens, T. C., and Pharmacokinetics of the thymidylate synthase inhibitor Nl°-propargyl - Calvert, A. H. Pharmacokinetic and toxicity studies with the novel 5,8-dideazafolic acid (CB3717) in the mouse. Cancer Treat. Rep., 70: quinazoline thymidylate synthase (TS) inhibitor, D1694. Proc. Am. 971-979, 1986. Assoc. Cancer Res., 31: 341, 1990. 47. Jodrell, D. I., Newell, D. R., Morgan, S. E., Clinton, S., Bensted, 64. Stephens, T. C., Valcaccia, B. E., Sheader, M. L., Hughes, L., and J. P. M., Hughes, L. R., and Calvert, A. H. The renal effects of Jackman, A. L. The thymidylate synthase (TS) inhibitor ICI D 1694 is Nl°propargyl-5,8-dideazafolic acid (CB3717) and a non-nephrotoxic superior to CB3717, 5-fluorouracil (5-FU) and methotrexate (MTX) analogue, ICI D1694, in mice. Br. J. Cancer, 64: 833-838, 1991. against a panel of human tumor xenografts. Proc. Am. Assoc. Cancer 48. Calvert, A. H., Alison, D. L., Harland, S. J., Robinson, B. A., Res., 32: 328, 1991. Jackman, A. L., Jones, T. R., Newell, D. R., Siddik, Z. H., Wiltshaw, E., 65. Jackman, A. L., Farrugia, D. C., Clarke, S. J., Aherne, G. W., Boyle, McElwain, T., Smith, I., and Harrap, K. A Phase I evaluation of the F. T., Seymour, L., Azab, M., and Kennealey, G. Delayed rescue of quinazoline antifolate thymidylate synthase inhibitor, Nl°-propargyl - ZD1694 toxicity in BALB/c mice with thymidine (dThd) or leucovorin 5,8-dideazafolic acid, CB 3717. J. Clin. Oncol., 4: 1245-1252, 1986. (LV). Proc. Am. Assoc. Cancer Res., 36: 377, 1995. 49. Vest, S., Bork, E., and Hansen, H. H. A Phase I evaluation of 66. Aherne, G. W., Farrugia, D. C., Ward, E., Sutcliffe, F., and Jack- Nm-propargyl 5,8-dideazafolic acid. Eur. J. Cancer & Clin. Oncol., 24: man, A. L. ZD1694 (Tomudex) and polyglutamate levels in mouse 201-204, 1988. plasma and tissues measured by radioimmunoassay (RIA) and the effect 50. Sessa, C., Zucchetti, M., Ginier, M., Willems, Y., D'Incalci, M., of leucovorin (LV). Proc. Am. Assoc. Cancer Res., 36: 376, 1995. and Cavalli, F. Phase I study of the antifolate Nm-propargyl-5,8-didea - 67. The International Medical Function. Tomudex (ZD1694), a Thymi- zafolic acid, CB3717. Eur. J. Cancer & Clin. Oncol., 24: 769-775, dylate Synthase Inhibitor: Clinical Investigator's Manual. Macclesfield, 1988. United Kingdom: Zeneca Pharmaceuticals, 1994. 51. Cantwell, B. M. J., Macaulay, V., Harris, A. L., Kaye, S. B., Smith, 68. Jodrell, D. I., Newell, D. R., Gibson, W., Hughes, L. R., and I. E., Milsted, R. A., and Calvert, A. H. Phase II study of the antifolate Calvert, A. H. The pharmacokinetics of the quinazoline antifolate ICI D Nm-propargyl-5,8-dideazafolic acid (CB3717) in advanced breast can- 1694 in mice and rats. Cancer Chemother. Pharmacol., 28: 331-338, cer. Eur. J. Cancer & Clin. Oncol., 24: 733-736, 1988. 1991. 52. Bassendine, M. F., Curtin, N. J., Loose, H., Harris, A. L., and 69. Jackman, A. L., and Gibson, W. Polyglutamation of the thymidylate James, O. F. Induction of remission in hepatocellular carcinoma with a synthase (TS) inhibitor, ZD1694 (Tomudex), in normal mouse tissues. new thymidylate synthase inhibitor CB 3717: a Phase II study. J. Proc. Am. Assoc. Cancer Res., 36: 377, 1995. Hepatol., 4: 349-356, 1987. 70. Judson, I., Clarke, S., Ward, J., Sutcliffe, F., Planting, A., Spiers, J., 53. Newell, D. R., Siddik, Z. H., Calvert, A. H., Jackman, A. L., Alison, Smith, R., Verweij, J., and Judson, I. Pharmacokinetics of ICI D1694 D. L., McGhee, G. K., and Harrap, K. R. Pharmacokinetics and toxicity following a 15-minute infusion in patients with advanced cancer. Proc. studies with CB3717. Proc. Am. Assoc. Cancer Res., 23: 181, 1982. Am. Soc. Clin. Oncol. Annu. Meet., 11: 117, 1992.

71. Clarke, S. J., Ward, J., Tierney, K., Walker, M., Sutcliffe, F., non-polyglutamatable quinazoline thymidylate synthase (TS) inhibitor. Planting, A., Verweij, J., Spiers, J., Smith, R., and Judson, I. Phase I and Proc. Am. Assoc. Cancer Res., 35: 301, 1994. pharmacokinetic study of ICI D1694: a folate-based thymidylate syn- 88. Walton, M. I., Gibson, W., Aherne, L., Lawrence, N., Stephens, T., thase (TS) inhibitor. Ann. Oncol., 3 (Suppl. 1): 135, 1992. Smith, M., and Jackman, A. Pharmacokinetics of the potent non-poly- 72. Clarke, S. J., Ward, J., Planting, A., Spiers, J., Smith, R., Verweij, glutamatable thymidylate synthase inhibitors CB30900 and ZD9331 in J., and Judson, I. Phase I trial of ICI D 1694: a novel thymidylate mice. Proc. Am. Assoc. Cancer Res., 35: 301, 1994. synthase inhibitor. Proc. Am. Assoc. Cancer Res., 33: 406, 1992. 89. Stephens, T. C., Smith, M. N., McClockey, M. L., and Boyle, F. T. 73. Clarke, S. J., Jackman, A. L., and Judson, I. R. The toxicity of ICI ZD9331, a novel non-polyglutamated thymidylate synthase (TS) inhib- D1694 in man and mouse. Adv. Exp. Med. Biol., 338: 601-604, 1993. itor: in vivo antitumor efficacy and toxicity to normal murine tissues. 74. Clarke, S. J., Ward, J., De Boer, M., Planting, A., Verweij, J., Proc. Am. Assoc. Cancer Res., 35: 305, 1994. Sutcliffe, F., Azab, M., and Judson, I. Phase I study of the new 90. Pendergast, W., Dickerson, S. H., Johnson, J. V., Dev, I. K., Ferone, thymidylate synthase inhibitor "Tomudex" (ZD1694) in patients with R., Duch, D. S., Hall, W. R., Humphreys, J., Kelly, I. M., and Wilson, advanced malignancy. Ann. Oncol., 5 (Suppl. 5): 132, 1994. D. C. Benzoquinazoline inhibitors of thymidylate synthase: enzyme 75. Sorensen, J. M., Jordan, D., Grem, J. L., Hamilton, J. M., Arbuck, inhibitory activity and cytotoxicity of some 3-amino and 3-methylben- S. G., Johnson, P., Kohler, D. R., Goldspiel, B. R., and Allegra, C. J. zo[f]quinazolin-l(2H)-ones. J. Med. Chem., 36: 2279-2291, 1993. Phase I trial of D1694, a pure thymidylate synthase inhibitor. Proc. Am. 91. Duch, D. S., Banks, S., Dev, I. K., Dickerson, S. H., Ferone, R., Soc. Clin. Oncol., 12: 158, 1993. Heath, L. S., Humphreys, J., Knick, V., Pendergast, W., Singer, S., 76. Sorensen, J. M., Jordan, E., Grem, J. L., Arbuck, S. G., Chen, A. P., Smith, G., Waters, K., and Wilson, R. Biochemical and cellular phar- Hamilton, J. M., Johnson, P., Kohler, D., Goldspiel, B., and Allegra, C. macology of 1849U89, a novel benzoquinazoline inhibitor of thymidy- Phase I trial of ZD1694 ("Tomudex") a direct inhibitor of thymidylate late synthase. Cancer Res., 53: 810-818, 1993. synthase. Ann. Oncol., 5 (Suppl. 5): 132, 1994. 92. Ferone, R., Hanlon, M. H., Waters, K. A., and Dev, I. K. Influence of intracellular polyglutamation on the cytotoxicity of the thymidylate 77. Gore, M., Earl, H., Cassidy, J., Tattersal, M., Mansi, J., and Azab, synthase inhibitor 1843U89. Proc. Am. Assoc. Cancer Res., 34: 274, M. Phase II study of "Tomudex" (ZD1694), a new thymidylate syn- 1993. thase inhibitor with antitumour activity in breast cancer. Ann. Oncol., 5 (Suppl. 5): 132, 1994. 93. Banks, S. D., Waters, K. A., Barrett, L. L., Dickerson, S., Pender- gast, W., and Smith, G. K. Destruction of WiDr multicellular tumor 78. Smith, I. E., Spielmann, M., Bonneterre, J., Namer, M., Green, M., spheroids with the novel thymidylate synthase inhibitor 1843U89. Can- Wander, H. E., Toussaint, C., and Azab, M. "Tomudex" (ZD1694), a cer Chemother. Pharmacol., 36: 455-459, 1994. new thymidylate synthase inhibitor ZD1694 in advanced breast cancer. 94. Wilson, H. R., Heath, L. S., Knick, V. C., Kszalka, G. W., and Ann. Oncol., 5 (Suppl. 5): 132, 1994. Ferone, R. In vivo antitumor activity of 1843U89, a new antifolate 79. Smith, I. E., Spielmann, M., Namer, M., Bonneterre, J., Wander, thymidylate synthase inhibitor. Proc. Am. Assoc. Cancer Res., 33: 407, H. E., Green, M., Lowery, K., and Azab, M. Antitumour activity of a 1992. new thymidylate synthase inhibitor ZD1694 in advanced breast cancer. 95. Boyd, M. R., and Paull, K. D. Some practical considerations and Breast Cancer Res. Treat., 27: 149, 1993. applications of the National Cancer Institute in vitro anticancer drug 80. Zalcberg, J., Cunningham, D., Banc Custsem, E., Francois, J. E., discovery screen. Drug Dev. Res., 34: 91-109, 1995. Schornagel, J. H., Adenis, A., Green, M., Starkhammer, H., and Azab, 96. Jackson, R. C., Johnston, A. L., Shetty, B. V., Varney, M. D., M. Good antitumour activity of the new thymidylate synthase inhibitor and Webber, S. E. Molecular design of thymidylate synthase inhibitors. "Tomudex" (ZD1694) in colorectal cancer. Ann. Oncol., 5 (Suppl. 5): Proc. Am. Assoc. Cancer Res., 34: 566-567, 1993. 133, 1994. 97. Varney, M. D., Marzone, G. H. P., Palmer, C. L., Reddy, M. R., 81. Pazdur, R., Casper, E. S., Meropol, N. J., Fuchs, C., and Kennealey, Tatlock, J., Romines, W. H., Wu, B. W., and Musick, L. Crystal- G. T. Phase II trial of "Tomudex" (ZD1694), a thymidylate synthase structure-based design and synthesis of benz[od]indole-containing in- inhibitor in advanced pancreatic cancer. Proc. Am. Soc. Clin. Oncol. 13: hibitors of thymidylate synthase. J. Med. Chem., 35: 663-676, 1992. 207, 1994. 98. Reich, S. H., Fuhry, M. A. M., Nguyen, D., Pint, M. J., Welsh, 82. Gore, M., Earl, H., Cassidy, J., Tattersal, M., Mansi, J., and Azab, K. M., Webber, S., Janson, C. A., Jordan, S. R., Matthews, D. A., Smith, M. Phase II study of Tomudex (ZD1694) in refractory ovarian cancer W., and Bartlett, C. Design and synthesis of novel 6,7-imidazotetrahy- Ann. Oncol., 5 (Suppl. 5): 133, 1994. droquinoline inhibitors of thymidylate synthase using iterative protein 83. Burris, H. I., Von Hoff, D., Bowen, K., Heaven, R., Rinaldi, D., crystal structure analysis. J. Med. Chem., 35: 847-858, 1992. Eckardt, J., Campbell, L., Robert, F., Patton, S., and Kennealey, G. A 99. Webber, S., Shetty, B., Johnston, A., Welsh, K., Varney, M., Deal, Phase II trial of ZD1694, a novel thymidylate synthase inhibitor, in J., Morse, C., and Soda, K. In vitro properties and antitumor activity of patients with advanced non-small cell lung cancer. Ann. Oncol., 5 AG-331, a novel lipophilic thymidylate synthase inhibitor. Proc. Am. (Suppl. 5): 133, 1994. Assoc. Cancer Res., 33: 413, 1992. 84. Cunningham, D., Zalcberg, J., Francois, E., Van Cutsem, J. H., 100. Webber, S., Johnston, A., Shetty, B., Webber, S. E., Welsh, K., Ellis, P., and Azab, M. Tomudex (ZD1694), a new thymidylate synthase Hilliard, J., Kosa, M., Morse, C., and Soda, K. Preclinical studies on inhibitor with good antitumor activity in advanced colorectal cancer AG337, a novel lipophilic thymidylate synthase inhibitor. Proc. Am. (ACC). Proc. Am. Soc. Clin. Oncol., 13: 199, 1994. Assoc. Cancer Res., 34: 273, 1993. 85. Piedbois, P., Buyse, M., Rustum, Y., Machover, D., Erlichman, C., 101. Boritzki, T. J., Bartlett, C. A., Huffaker, T., Morse, C. A., Webber, Carlson, R. W., Valone, F., Labianca, R., Doroshow, J. H., and Petrelli, S., Welsh, K., and Jackson, R. C. Biochemical properties of new N. for the Advanced Colorectal Cancer Meta-Analysis Project. Modu- lipophilic thymidylate synthase inhibitors designed from the X-ray lation of fluorouracil by leucovorin in patients with advanced colorectal structure of the E. coli enzyme. Proc. Am. Assoc. Cancer Res., 33: 33, cancer: evidence in terms of response rate. J. Clin. Oncol., 10: 896-903, 1992. 1992. 102. Clendeninn, N. J., Peterkin, J. J., Webber, S., Shetty, B. V., Koda, 86. Poon, M. A., O'Connel, J. H., Wieland, H. S., Krook, J. E., R. T., Leichman, L., Leichman, C. G., Jeffers, S., Muggia, F., and Gerstner, J. B., Tschetter, L. K., Levitt, R., Kardinal, C. G., and O'Dwyer, P. AG-331, a "non-classical" lipophilic thymidylate syn- Maillard, J. A. Biochemical modulation of fluorouracil with leucovorin: thase inhibitor for the treatment of solid tumors Ann. Oncol., 5 (Suppl. confirmatory evidence of improved therapeutic efficacy in advanced 5): 133, 1994. colorectal cancer. J. Clin. Oncol., 9: 1967-1972, 1991. 103. Calvete, J. A., Balmanno, K., Taylor, G. A., Raft, I., Newell, D. R., 87. Jackman, A. L., Aherne, G., Kimbell, R., Brunton, L., Hardcastle, Lind, M. J., Calvert, A. H., Webber, S., and Clendeninn, N. J. Preclinical A., Wardleworth, J. W., Stephens, T. C., and Boyle, F. ZD9331, a and clinical studies of prolonged administration of the novel thymidylate

synthase inhibitor, AG337. Proc. Am. Assoc. Cancer Res., 35: 306, LY231514, in patients with advanced solid tumors. Proc. Am. Soc. Clin. 1994. Oncol., 13: 430, 1994. 104. Calvete, J. A., Balmanno, K., Raft, I., Newell, D. R., Taylor, G. A., 120. Taylor, G. A., Raft, I., Balmanno, K., Calvete, J. A., Newell, D. R., Boddy, A. V., Lind, M. J., Bailey, N. B., Calvert, A. H., Webber, S., and Webber, S., Jackson, R. C., Gumbrell, L., Chapman, F., and Oakey, A. Clendeninn, N. J. Preclinical and clinical studies of the novel thymidy- Preclinical and early clinical studies with the lipophilic thymidylate late synthase inhibitor AG337, given by oral administration. Proc. Am. synthase inhibitor, AG337. Br. J. Cancer., 67 (Suppl. XX): 17, 1993. Assoc. Cancer Res., 36: 380, 1995. 121. Chu, E., Voeller, D., Koeller, D. M., Drake, J. C., Takimoto, C. H., 105. Raft, I., Taylor, G. A., Balmanno, K., Calvete, J. A., Newell, D. R., Maley, G., Maley, F., and Allegra, C. J. Identification of an RNA Lind, M. J., and Calvert, A. H. A Phase I study of the novel antifolate binding site for human thymidylate synthase. Proc. Natl. Acad. Sci. 3,4-dihydro-2-amino-6-methyl-4-oxo-5-(4-pyridylthio)-quinazolone di- USA, 90: 517-521, 1993. hydrochloride (AG337) given by a 24 h intravenous continuous infu- 122. Chu, E., Koeller, D. M., Casy, L. J., Drake, C. J., Chabner, B. A., sion. Proc. Am. Assoc. Cancer Res., 35: 306, 1994. Elwood, P. C., Zinn, S., and Allegra, C. J. Autoregulation of human 106. Johnson, P. Agouron Initiates Phase II Efficacy Trials of Anti- thymidylate synthase messenger RNA translation by thymidylate syn- Tumor Drug AG337. La Jolla, CA: Agouron Pharmaceuticals, Inc., thase. Proc. Natl. Acad. Sci. USA, 88: 8977-8981, 1991. 1994. 123. Chu, E., Zinn, S., Boarman, D., and Allegra, C. J. The interaction 107. Boddy, A. V., Calvete, J. A., and Raft, I. Non-linear pharmacoki- of interferon-~ and 5-fluorouracil in the H630 human colon carcinoma netics in cancer patients of AG337, a rationally-designed thymidylate cell line. Cancer Res., 50: 5834-5840, 1990. synthase (TS) inhibitor. Proc. Am. Assoc. Cancer Res., 36: 236, 1995. 124. Spears, C. P., Shahinian, A. H., Moran, R. G., Heidelberger, C., 108. Raft, I., Taylor, G. A., Calvete, J. A., Balmanno, K., Boddy, A. V., and Corbett, T. H. In vivo kinetics of thymidylate synthetase inhibition Bailey, N. B., Lind, M. J., Newell, D., Calvert, A. H., Johnston, A., and of 5-fluorouracil-sensitive and resistant murine colon adenocarcinomas. Clendeninn, N. J. A Phase I clinical study of the novel antifolate AG337 Cancer Res., 42: 450-456, 1982. given by a 5-day continuous infusion. Proc. Am. Assoc. Cancer Res., 125. Keyomarsi, D., and Moran, R. G. Mechanism of the cytotoxic 36: 240, 1995. synergism of fluoropyrimidines and folinic acid in mouse leukemic 109. Koda, R. T., Leichman, L., Webber, S., Shetty, B. V., Leichman, cells. J. Biol. Chem., 263: 14402-14409, 1988. C. G., Jeffers, S., and Muggia, F. M. Pharmacokinetics of AG-331 following single escalating i.v. doses in patients with refractory solid 126. Keyomarsi, K., Samet, J., Molnar, G., and Pardee, A. B. The tumors. Proc. Am. Assoc. Cancer Res., 35: 432, 1994. thymidylate synthase inhibitor ICI D 1694 overcomes translational detainment of the enzyme. J. Biol. Chem., 268:15132-15149, 1993. 110. O'Connor, B. M., Webber, S., Jackson, R. C., Galivan, J., and Rhee, M. S. Biological activity of a novel rationally designed lipophilic 127. Keyomarsi, K., Molnar, G., and Pardee, A. B. Loss of translational thymidylate synthase inhibitor. Cancer Chemother. Pharmacol., 34: control of thymidylate synthase (TS) by antifolates results in the accu- 225-229, 1994. mulation of the TS protein. Proc. Am. Assoc. Cancer Res., 34: 276, 111. Jackman, A. L., Kelland, L. R., Brown, M., Gibson, W., Kimbell, 1993. R., Aherne, W., and Judson, I. R. ICI D 1694 resistant cell lines. Proc. 128. Lowe, S. W., Ruley, H. E., Jacks, T., and Housman, D. E. p53- Am. Assoc. Cancer Res., 33: 406, 1992. dependent apoptosis modulates the cytotoxicity of anticancer agents. 112. Freemantle, S. J., Lunec, J., Jackman, A. L., Kelland, L. R., Cell, 74: 957-967, 1993. Aheme, G. W., and Calvert, A. H. Molecular characterisation of three 129. Lowe, S. W., Bodis, S., McClathcey, A., Remington, L., Ruley, cell lines selected for resistance to folate-based thymidylate synthase H. E., Fisher, D. E., Housman, D. E., and Jacks, T. p53 status and the inhibitors. Br. J. Cancer, 67 (Suppl. 20): 79, 1993. efficacy of cancer therapy in vivo. Science (Washington DC), 266: 113. Berger, S. H., Jenh, C. H., Johnson, L. F., and Berger, F. G. 807-810, 1994. Thymidylate synthase overproduction and gene amplification in fluoro- 130. Oltvai, Z. N., and Korsmeyer, S. J. Checkpoints of dueling dimers deoxyuridine-resistant human cells. Mol. Pharmacol., 28: 461-467, foil death wishes. Cell, 79: 189-192, 1994. 1987. 131. Starzynska, T., Bromley, M., Ghosh, A., and Stern, P. L. Prog- 114. Cowan, K. H., and Jolivet, J. A methotrexate-resistant human nostic significance of p53 overexpression in gastric and colorectal breast cancer cell line with multiple defects including diminished for- carcinoma. Br. J. Cancer, 66: 558-562, 1992. mation of methotrexate polyglutamates. J. Biol. Chem., 259: 10793- 132. Thompson, C. B. Apoptosis in the pathogenesis and treatment of 1O8O0, 1984. disease. Science (Washington DC), 267." 1456-1462, 1995. 115. Rosowsky, A., Wright, J. E., Cucchi, C. A., Lippke, J. A., Tantra- 133. Chu, E., Takechi, T., Jones, K. L., Voeller, D. M., Copur, S. M., vahi, R., Ervin, T. J., and Frei, E. III. Phenotypic heterogeneity in Maley, G. F., Maley, F., Segal, S., and Allegra, C. J. Thymidylate cultured human ~aead and neck squamous cell carcinoma lines with synthase binds to c-myc RNA in human colon cancer cells and in vitro. low-level methotrexate resistance. Cancer Res., 45: 6205-6212, 1985. Mol. Cell. Pharmacol., 15: 179-185, 1995. 116. Pizzorno, G., Mini, E., Coronnello, M., McGuire, J. J., Moroson, J. B. A., Cashmore, A. R., Dreyer, R. N., Lin, J. T., Mazzei, T., Periti, 134. Chu, E., Copur, S., Jones, K. L., Khleif, S., Voeller, D., Maley, P., and Bertino, J. Impaired polyglutamylation of methotrexate as a G. F., Maley, F., and Allegra, C. J. Thymidylate synthase regulates the cause of resistance in CCRF-CEM cells after short-term, high dose translation of p53 mRNA. Proc. Am. Assoc. Cancer Res., 36: 563, 1995. treatment with this drug. Cancer Res., 48: 2149-2155, 1988. 135. Arredondo, M. A., Yin, M. B., Lu, K., Schrber, C., Slocum, H., 117. Shih, C., Grindey, G. B., Barnett, C. J., Pearce, H. L., Engelhardt, and Rustum, Y. M. Decreased c-myc protein accompanied by increased J. A., Todd, G. C., Rinzel, S. M., Worzalla, J. F., Gossett, L. S., and p53 and Rb is associated with thymidylate synthase inhibition and cell Everson, T. P. Structure-activity relationship studies of novel pyrrol- cycle retardation induced by D1694. Proc. Am. Assoc. Cancer Res., 35: opyrimidine antifolate LY231514. Proc. Am. Assoc. Cancer Res., 33: 304, 1994. 411, 1992. 136. Fisher, T. C., Milner, A. E., Gregory, C. D., Jackman, A. L., 118. Grindey, G. B., Shih, C., Barnett, C. J., Pearce, H. L., Engelhardt, Aherne, G. W., Hartley, J. A., Dive, C., and Hickman, J. A. bcl-2 J. A., Todd, G. C., Rinzel, S. M., Worzalla, J. F., Gossett, L. S., and modulation of apoptosis induced by anticancer drugs: resistance to Everson, T. P. LY231514, a novel pyrrolopyrimidine antifolate that thymidylate stress is independent of classical resistance pathways. Can- inhibits thymidylate synthase (TS). Proc. Am. Assoc. Cancer Res., 33: cer Res., 53: 3321-3326, 1993. 411, 1992. 137. Lenz, H. J., Ju, J. F., Danenberg, K. D., Leichman, L., Leichman, 119. Rinaldi, D. A., Burris, H. A., Dorr, F. A., Nelson, J., Fields, S. M., C., Haag, C., and Danenberg, P. V. TS levels and p53 function predict Kuhn, J. G., Eckardt, J., Lu, P., Woodworth, J. R., and Corso, S. W. A for response to 5-fluorouracil/leucovorin in patients with advanced Phase I evaluation of the novel thymidylate synthase inhibitor, colon cancer. Proc. Am. Soc. Clin. Oncol., 14: 202, 1995.

138. Kobayashi, H., Takemura, Y., Miyachi, H., Skelton, L., and Jack- edly enhances 5-fluorouracil cytotoxicity. Proc. Am. Assoc. Cancer man, A. L. The effect of hammerhead ribozyme against thymidylate Res., 35: 330, 1994. synthase on the cytotoxicity of thymidylate synthase inhibitor. Proc. 143. Izzo, J., Zielinski, Z., Chang, Y. M., and Bertino, J. R. Molecular Am. Assoc. Cancer Res., 36: 342, 1995. mechanisms of the synergistic sequential administration of D1694 (To- 139. Weber, G., and Prajda, N. Targeted and non-targeted actions of mudex) followed by FUra in colon carcinoma cells. Proc. Am. Assoc. anti-cancer drugs. Adv. Enzyme Regul., 34: 71-89, 1994. Cancer Res., 36: 381, 1995. 140. Pressacco, J., and Erlichman, C. Combination studies with 3'- 144. Van der Wilt, C. L., Pinedo, H. M., Kuiper, C. M., Smid, K., and azido-deoxythymidine (AZT) plus ICI D1694. Cytotoxic and biochem- Peters, G. J. Biochemical basis for the the combined antiproliferative ical effects. Biochem. Pharmacol., 46: 1989-1997, 1993. effect of AG337 or ZD1694 and 5-fluorouracil. Proc. Am. Assoc. 141. Pressacco, J., Hedley, D. W., and Erlichman, C. ICI D1694 and Cancer Res., 36: 379, 1995. idoxuridine: a synergistic antitumor combination. Cancer Res., 54: 145. National Cancer Institute PDQ. Phase I study of Tomudex in 3772-3778, 1994. combination with 5-FU/CF/IFN-A for recurrent adult solid tumors in- 142. Chang, Y. M., Zielinski, Z., Izzo, J., Porubcin, M., and Bertino, cluding gastrointestinal, breast, lung, and ovarian carcinomas. Curr. J. R. Pretreatment of colon carcinoma cells to D1694 (Tomudex) mark- Clin. Trials Oncol., 2: 749, 1995.

N Touroutoglou and R Pazdur

Clin Cancer Res 1996;2:227-243.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/2/2/227

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/2/2/227. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.