Clinical significance of prognostic and predictive markers in ... - Nature

The Pharmacogenomics Journal (2002) 2, 209–216

 2002 Nature Publishing Group All rights reserved 1470-269X/02 $25.00 www.nature.com/tpj

CLINICAL IMPLICATION

Clinical significance of prognostic and predictive markers in colorectal cancer DB Longley, U McDermott and PG Johnston Department of Oncology, Cancer Research Centre, Queen’s University Belfast, Belfast, Northern Ireland The Pharmacogenomics Journal (2002) 2, 209–216. doi:10.1038/sj.tpj.6500124 INTRODUCTION Colorectal cancer (CRC) is the second most common cause of cancer death in the US. Approximately 130 000 new cases of CRC are diagnosed in the US each year with an annual mortality of approximately 56 000 people. In the post-operative setting, there is clear evidence that offering adjuvant therapy significantly improves survival in stage III CRC, however, the management of stage II disease is an area of controversy and debate. The 5-year survival for stage II patients is approximately 75%, indicating that the majority are cured by surgery alone, and many patients who would have been cured by surgery undergo chemotherapy unnecessarily. Furthermore, some patients who are not recommended to receive adjuvant treatment may benefit from chemotherapy. Recent studies have started to define prognostic molecular markers that identify those patients at risk of relapse after surgery and who may benefit most from chemotherapy (Table 1). Furthermore, molecular markers are being identified that predict whether a tumour will respond to a particular chemotherapy. In addition, some markers may have therapeutic value by predicting the degree of systemic toxicity for a particular treatment. For such studies tumour DNA can be analysed to identify genetic changes such as point mutations, loss of heterozygosity (LOH), gene ampli-

fication and microsatellite instability (MSI). In addition, the expression of individual genes can be assessed at the levels of mRNA and protein using the techniques of real-time reverse transcription PCR and immunohistochemistry (IHC) respectively. The ultimate goal of this research is the tailoring of treatment to the molecular phenotypes of tumour and patient. This review highlights potentially important prognostic and predictive markers in CRC. Due to space restrictions many worthy contributions could not be cited. Colorectal Cancer Biology The genetic model of CRC proposed by Vogelstein and others suggests that

Table 1

multiple genetic abnormalities are accumulated over time during the progression from adenoma to adenocarcinoma (Figure 1).1 Mutations in genes such as the Kirsten-ras (K-ras), adenomatous polyposis coli (APC), deleted in colon cancer (DCC) and the p53 tumour suppressor gene are the most common genetic alterations found in sporadic CRC. Approximately 5% of CRC is inherited and familial adenomatous polyposis (FAP) and hereditary non-polyposis colon cancer (HNPCC) are the most well-characterized syndromes.2 FAP is caused by mutations in the APC gene, whereas 90% of HNPCC cases are caused by mutations in genes involved in mismatch repair (MMR) leading to microsatellite instability (MSI). In addition, 10–15% of sporadic cases of CRC have mutations in MMR genes. Loss of heterozygozity (LOH) More than 75% of CRCs have lost a portion of chromosome 17p or 18q, or both. The 17p segment contains p53, which is regarded as one of the most important tumour suppressor genes. It has a number of key functions such as DNA damage repair, initiation of cell cycle checkpoints and initiation of

Prognostic molecular markers in colorectal cancer

Marker

Measurement technique

p53(LOH 17p)

IHC, DNA sequencing

LOH 18q

RFLP, PCR, Southern Blot

LOH 8p

RFLP, PCR, Southern Blot

MSI

PCR

K-ras

SSCP, DNA sequencing

TS

IHC, RT-PCR, genotyping

Comments p53 overexpression measured by IHC generally associated with worse prognosis, however, not a consistent observation. p53 mutation by sequencing associated with shorter survival time. Associated with worse prognosis. Locus contains 3 potential tumour suppressor genes: DCC, Smad2, Smad4. Associated with worse prognosis. Identity and function of genes at this locus not determined. MSI-H tumours associated with survival advantage and less aggressive clinical course. Some studies have found an association between K-ras mutation and survival, but others have not. The particular type of mutation may be important. High TS expression and TSER*3/TSER*3 genotype associated with worse clinical outcome.

Molecular markers for colorectal cancer DB Longley et al

210

Chromosome: 5q Alteration: loss/mutation Gene: APC

12p Mutation K-ras

18q loss DCC, Smad 2/4

17p loss/mutation p53

Other alterations

DNA methylation changes

Normal epithelium

Hyperprolif epithelium

Early Adenoma

Intermediate Adenoma

Late Adenoma

Carcinoma

Metastasis

Figure 1 A genetic model for colorectal carcinogenesis (adapted from Fearon and Vogelstein1).

apoptosis.3 p53 has been the most extensively studied molecular marker in patients with CRC and mutations have been found in 40–60% of cases.2 Missense mutations of the p53 gene result in production of a protein with a longer half-life, therefore, detection of increased p53 expression by IHC has been used as a surrogate for p53 mutation in many studies. However, this assumption has been shown to be valid in only 60–80% of cases in studies that have compared p53 expression measured by IHC with p53 mutations analysed by DNA sequencing.4 Despite numerous investigations into the prognostic value of p53 overexpression in patients with CRC, the issue still remains controversial. Many studies have been statistically underpowered, and the issue is further complicated by the use of different antibodies which have varying levels of sensitivity and capacity for detecting mutant forms of p53. Studies of at least 100 patients with CRC using p53 monoclonal antibodies have generally found that p53 overexpression was associated with worse prognosis.5,6 This association has not been a consistent observation as some studies have found the opposite,7 or no correlation.8,9 However, the majority of DNA sequencing studies have shown a significantly shorter survival time in patients with mutant p53.10 Data from studies in cell line models have suggested that p53 status influences 5FU chemosensitivity.11,12 This has been borne out in several clinical studies, for example Ahnen et al found that patients with stage III CRC whose tumours overexpressed p53 did not derive significant survival benefit from adjuvant 5FU-based treatment,

The Pharmacogenomics Journal

whereas those without p53 overexpression did.13 However, other studies have found no correlation between p53 status and response to 5FU-based chemotherapy.14 Allelic loss on chromosome 18q is found in approximately 70% of CRCs and has been associated with worse prognosis in stage II and III tumours.15,16 Jen et al found that stage II patients who retained both 18q alleles have an excellent 5-year survival of 93% compared to 54% of those who had lost one allele.16 This data suggests that patients with stage II disease who have a deletion on chromosome 18q have a more aggressive clinical course and may represent a subset of stage II patients who would benefit from adjuvant chemotherapy. Furthermore, a recent study by Watanabe et al found that retention of both 18q alleles predicted for favourable outcome after adjuvant 5FU-based chemotherapy in stage III CRC.17 The 18q segment contains three candidate tumour suppressor genes: DCC, Smad 2 and Smad 4. The product of the DCC gene is a cell adhesion molecule.18 and it has been postulated that decreased DCC expression may lead to altered adhesion and contribute to enhanced tumour growth and metastatic spread of colorectal tumours. Smad proteins are transcription factors involved in the transforming growth factor ␤ (TGF-␤) pathway.19 These proteins are involved in signalling from TGF-␤ receptors and regulate transcription of target genes including c-myc, CBFA1, FLRF and furin. Targeted disruption of Smad genes in mice has revealed their importance in embryonic development. Furthermore, Smad 2 and Smad 4 have been implicated in human can-

cer and appear to have tumour suppressor functions.20 Loss of material from the short arm of chromosome 8 is found in approximately 50% of cases of CRC. The introduction of chromosome 8p into colon cancer cell lines has been shown to decrease their tumourigenecity and invasiveness.21 Furthermore, LOH at 8p has been shown to correlate with the presence of micrometastases.22 Halling et al found that in stage II and III CRCs 8p allelic imbalance was an independent prognostic indicator for decreased survival and time to recurrence.23 This was the first large study exploring the role of 8p allelic imbalance in prognosis of CRC. Microsatellite instability (MSI) MSI is caused by mutations in MMR genes such as hMSH2, hMLH1 and hMSH6 resulting in failure of the DNA mismatch repair system to correct errors that occur during replication. Numerous studies have shown a survival advantage and a less aggressive clinical course for patients with high frequency MSI.23,24 Furthermore, MSIhigh tumours have been reported to metastasise less often than microsatellite stable tumours.24 The MSI phenotype also appears to be associated with excellent survival in patients who receive adjuvant 5FU-based chemotherapy.25 A recent study found that 61% of stage III CRC patients with MSI-high tumours had a mutation in the gene for the type II receptor for TGF␤-1.17 Furthermore, these patients had a 74% 5-year survival following adjuvant 5FU-based chemotherapy compared to 46% in patients with MSI-high tumours that lacked the mutation. Thus, mutation of the gene


211

for the type II receptor for TGF␤-1 may predict the response of MSI-high tumours to 5FU-based chemotherapy. It is possible that immunohistochemical analysis of hMLH1 and hMSH2 can be used as a surrogate for MSI.

activated by addition of a farnesyl group, therefore, inhibitors of farnesyl transferase may be useful in tumours with K-ras mutations.28 C-myc is a transcription factor involved in regulating cell cycle progression, the expression of which is elevated in approximately 70% of CRCs.29 Mutational inactivation of the APC gene (found in the majority of CRCs) has been implicated in the upregulation of c-myc in CRCs. A recent study by Arango et al found that patients whose tumours had both amplified c-myc and wild-type p53 responded well to 5FU-based chemotherapy, whereas patients with unamplified c-myc and wild-type p53 did not benefit.30 These results suggest a predictive role for c-myc.

Oncogenes K-ras gene mutations are believed to be among the earliest events in CRC tumourigenesis and are present in approximately 50% of cancers.1 K-ras point mutations (most commonly at codons 12, 13 and 61) result in constitutive activation of downstream signal transduction pathways such as the mitogen activated protein kinase (MAPK) pathway. Although some studies have found an association between K-ras mutation and prognosis,13 others have failed to do so.26 It has recently been reported that particular types of K-ras mutations may be useful markers for identifying patients likely to benefit from 5FUbased adjuvant chemotherapy.27 K-ras is also of interest as a potential chemotherapeutic target. Ras proteins are

Chemotherapeutic Agents 5-Fluorouracil (5FU) has been the mainstay of CRC therapy for over 30 years, therefore most predictive data has been reported for this agent. CPT11 (irinotecan) and oxaliplatin are relatively new agents that are currently

used in the treatment of metastatic CRC, and recent studies have begun to identify predictive markers for these agents. Novel therapies such as the multitargeted antifolate (MTA) and ZD-1839 (Iressa), which inhibits the epidermal growth factor receptor (EGFR), are currently in clinical trials. The mechanism of action of both the established and the newer chemotherapeutic agents are summarised in Figure 2. 5-Fluorouracil (5FU) 5FU is converted intracellularly to fluorodeoxyuridine monophosphate (FdUMP), which binds covalently to the nucleotide synthetic enzyme thymidylate synthase (TS) in the presence of the reduced folate co-factor 5,10methylenetetrahydrofolate (CH2THF) (Figure 2).31 The resulting inhibition of TS causes nucleotide pool imbalances resulting in disruption of DNA synthesis and repair and ultimately causing DNA strand breaks.32 In addition, another 5-FU metabolite FUTP can be misincorporated into nuclear and

EGFR 5FU

DPD

ZD-1839

Degradation FUMP

TP K-ras

RNA damage

LV

DHFR

GARFT

CH2THF

MTA

RTX

FdUMP

TS

Farnesyl transferase

FTI

dTMP

ERCC1

Oxaliplatin DNA damage

CPT-11

SN-38

Topo 1

p53

Figure 2 Molecular targets of established and novel chemotherapeutic agents. Potential predictive markers are boxed. The abbreviations used are: TS, thymidylate synthase; TP, thymidine phosphorylase; DPD, dihydropyrimidine dehydrogenase; 5FU, 5-fluorouracil; FdUMP, fluorodeoxyuridine monophosphate; FUMP, fluorouridine monophosphate; dTMP, deoxythymidine monophosphate; LV, leucovorin; CH2THF, 5,10-methylene tetrahydrofolate; RTX, raltitrexed; MTA, multitargeted antifolate; DHFR, dihydrofolate reductase; GARFT, glycinamide ribonucleotide formyl transferase; ERCC1, excision repair cross complementing 1 gene product; EGFR, epidermal growth factor receptor; Topo I, DNA topoisomerase I; and FTI, farnesyl transferase inhibitor.

www.nature.com/tpj


212

cytoplasmic RNA, disrupting normal RNA processing and function.32 Response rates for 5FU monotherapy as a first line treatment for advanced CRC are 10–20% with median survival of less than 12 months.33 When leucovorin (LV) is administered with 5FU, the intracellular reduced folate pools are expanded and FdUMP-TS-CH2THF complex formation is increased, resulting in more effective TS inhibition. 5FU/LV combination therapy has been demonstrated to improve survival and quality of life in CRC patients.34 Pre-clinical studies have demonstrated that TS expression is a key determinant of 5FU sensitivity,35 and multiple clinical investigations have demonstrated an improved response to 5FU-based therapy in patients with low TS expression in their tumours.31,36 In addition, several studies have demonstrated that patients with high TS expression in their tumours have a significantly worse clinical outcome than those with low TS irrespective of response to treatment, therefore TS may also be a valuable prognostic marker.37,38 The TS gene promoter is polymorphic with either two (TSER*2) or three (TSER*3) tandem repeats of 28 base pairs.39 In vitro studies have demonstrated that three copies of the tandem repeat generated approximately 2.6-fold higher TS mRNA expression than two copies. Recently, preliminary studies have suggested that stage III CRC patients with the TSER*3/TSER*3 genotype do not receive the survival benefit from adjuvant 5FU treatment observed in patients with the TSER*2/TSER*2 or TSER*2/TSER*3 genotypes.39. A recent study found that methotrexate treated patients with acute lymphoblastic leukaemia who had the TSER*3/TSER*3 genotype had a poorer outlook than those with the other genotypes, therefore TSER genotyping may be therapeutically useful in a range of cancers.40 Recent studies have found that TS levels measured in primary tumours do not correspond with those measured in the corresponding metastasis.41,42 For example, in matched samples of primary and metastatic


tumours from the same patient, TS mRNA expression in lymph node and pulmonary metastases was significantly higher than in the corresponding primary tumour, whereas liver metastases expressed lower levels of TS than primary tumours.41 These results may explain why TS expression in primary tumour samples may not predict response of metastases to treatment.43 Thymidine phosphorylase (TP) reversibly converts 5FU to its active metabolite FdUMP. Cell culture and xenograft model systems indicated that transfection of TP into cancer cells increased their sensitivity to 5FU, presumably due to greater metabolic activation of 5FU to FdUMP.44 However, retrospective analysis of TP mRNA expression in 38 colorectal tumours indicated that tumours with the highest TP levels do not respond to 5FU-based chemotherapy.45 Indeed, high TP expression has been found to be an indicator of worse prognosis in CRC.46 These findings may be due to the role of TP as an angiogenic factor (initially identified as platelet-derived endothelial cell growth factor). Thus, high TP expression may be a marker for other molecular changes associated with the development of a more invasive and malignant tumour phenotype with increased resistance to chemotherapy. Dihydropyrimidine dehydrogenase (DPD) is the rate limiting enzyme in 5FU catabolism. Normally over 80% of administered 5FU is broken down by DPD in the liver, however, 3–5% of the normal population are deficient in DPD and experience severe toxicity when treated with 5FU.47 High levels of DPD mRNA expression in colorectal tumours have been shown to correlate with resistance to 5FU.48 This study also indicated that DPD, TS and TP are independent predictive markers of response to 5FU. Furthermore, use of all three markers increased the ability to predict outcome to 5FU-based chemotherapy to 100%, as all of the tumours that responded expressed TS, TP and DPD below their respective cutoff values. To prevent DPD-mediated degradation of 5FU both in liver and tumour, efforts have been made to modulate 5FU therapy with DPD

inhibitors. One such DPD inhibitor, eniluracil, has now entered clinical trials. Capecitabine is an oral fluoropyrimidine carbamate that avoids degradation by DPD in the liver.33 Capecitabine is absorbed unchanged through the gastrointestinal wall and is converted in the liver to 5⬘-deoxy-5fluorouridine (5⬘FDUR) by carboxylesterase and cytidine deaminase. Importantly, 5⬘FDUR is converted to 5FU by TP, which is more active in tumour than in normal tissue, resulting in tumour-selective generation of 5FU. Thus, unlike 5FU, capecitabine may be more active in tumours expressing high levels of TP, although high tumour TS and DPD expression would still be expected to predict resistance to this agent. Capecitabine has already shown considerable activity in xenograft models.49 Antifolates Raltitrexed (RTX) was the first TS-specific inhibitor to be approved for clinical use. It is a quinazoline analog of folic acid that is transported into cells by the reduced folate carrier (RFC) and is then polyglutamated by folylpolyglutamate synthase (FPGS).50 Cellular retention and TS binding affinity are both increased by polyglutamation. A randomised multicentre trial found that total objective response rates for RTX and 5FU/LV were similar in patients with advanced CRC.51 In cell culture model systems the major resistance mechanisms to RTX are increased levels of TS, decreased levels of polyglutamation and impaired transport by RFC.52,53 In contrast to 5FU/LV, RTX would be expected to have activity in tumours expressing elevated levels of DPD and/or TP mRNA. A novel multitargeted antifolate (MTA) that inhibits TS, the folate metabolic enzyme dihydrofolate reductase (DHFR) and the purine synthetic enzyme glycinamide ribonucleotide formyl transferase (GARFT) is in phase III clinical trials and has shown activity in a range of cancers (Figure 2).54 We have previously demonstrated that response to MTA in vitro is highly sensitive to TS expression levels.52 In vitro data from our laboratory suggests that inactivation of p53


213

increases resistance to RTX and MTA,12 however, this has yet to be assessed in patients. CPT-11 (Irinotecan) CPT-11 is a pro-drug that is converted in vivo by carboxylesterase into SN-38, which acts as a potent inhibitor of DNA topoisomerase I (Topo I).55 Topo I relaxes supercoiled double-stranded DNA in front of DNA replication forks by breaking one strand of the DNA double helix, passing the other strand through the cleavage site and religating the cleaved DNA. SN-38 interferes with this reaction by reversibly stabilizing the covalent complex formed between Topo I and the cleaved DNA, thereby inhibiting religation of Topo Ilinked single-stranded DNA breaks. It has been proposed that collision of the DNA replication fork with the SN-38 stabilized DNA-Topo I complex during S phase generates DNA double-strand breaks. In clinical studies CPT-11 has demonstrated single-agent activity in advanced CRC with response rates of 20-25%.56 In combination with 5FU, CPT-11 has produced response rates of 49% for advanced CRC.57 Furthermore, CPT-11 has demonstrated significant activity in tumours that are refractory to 5FU.58 The predictive value of Topo I for response to CPT-11 has not yet been demonstrated. Furthermore, most pre-clinical studies have failed to find a strong correlation between levels of Topo I protein or gene expression and CPT-11 sensitivity in cell line models.59 In vitro data suggest that MSI-high tumour cells are more responsive to CPT-11, whereas p53 status may not be a major factor for resistance,60 however, data from clinical studies are lacking. Platinum compounds Oxaliplatin is a platinum-based chemotherapeutic agent carrying a bulky 1,2-diamino-cyclohexane ring used in the treatment of advanced CRC.61 The bulkier carrier group is thought to cause platinum-DNA adducts that are more cytotoxic than those formed by other platinum agents such as cisplatin and carboplatin. Indeed, MMR deficiency and enhanced ability to bypass the site of DNA damage have

been demonstrated to cause resistance to cisplatin but not to oxaliplatin.62 Oxaliplatin has modest activity as a single agent, but is very active in combination with 5FU/LV, with response rates of approximately 50% seen in first-line treatment of metastatic CRC.63 Excision repair cross complementing (ERCC) proteins are components of the nucleotide excision repair system. High ERCC1 gene expression has been shown to correlate with poor survival and response of gastric cancer patients treated with 5FU and cisplatin.64 More recently, the same group has reported that survival of patients with metastatic CRC following 5FU/oxaliplatin chemotherapy was significantly better in patients expressing low levels of ERCC1 mRNA compared to those expressing high levels.65 It appears from this study that ERCC1 is an independent predictive marker of response to 5FU/oxaliplatin chemotherapy. Other proteins involved in nucleotide excision repair such as ERCC2 and XPF may also affect the ability of tumour cells to remove DNAplatinum adducts and should be assessed in future studies. We have found that p53 null cancer cells are more resistant to oxaliplatin and cisplatin (unpublished data), suggesting that p53 status may be an important predictive factor for these drugs. Novel therapies Overexpression and/or mutation of epidermal growth factor receptor (EGFR) have been detected in a number of human cancers and are associated with aggressive disease and poor prognosis.66 Several tyrosine kinase inhibitors have been developed to target EGFR, the most advanced of which ZD-1839 (Iressa) is in phase III clinical trials. ZD-1839 competitively inhibits ATP binding to EGFR and has been shown to have significant antitumour activity against a number of human tumour xenograft models when combined with a range of cytotoxic compounds.66 Vascular endothelial growth factor (VEGF) is the major angiogenic growth factor in CRC and is associated with worse prognosis.67 An antibody treatment to target VEGF is in phase I clinical

trials.68 Farnesyl transferase inhibitors (FTIs) that target post-translational modification of ras have been developed and may be of use in colorectal tumours expressing mutant Kras.28 Targeted chemotherapy using cytotoxic radicals conjugated to analogues of peptide hormones have been designed to selectively deliver high doses of cytotoxic agents to tumour cells that express peptide hormone receptors.69 One such cytotoxic analogue of somatostatin (SST) consisting of 2-pyrrolinodoxorubicin linked covalently to an SST octapeptide has recently demonstrated activity in colorectal cell lines expressing the SST receptor regardless of the p53 status of the cell line.70 These results suggest that patients with colorectal cancers that express the SST receptor may benefit from treatment with cytotoxic SST analogues. COX-2 overexpression has been inversely correlated with patient survival.71 In a colon cancer cell line exhibiting high COX-2 expression, selective COX-2 inhibition was cytotoxic and suppressed invasiveness,72 suggesting that COX-2 inhibitors may have a role to play as chemotherapeutic agents. The development of bioassays that measure the expression of the above target molecules could potentially permit the identification of patients most likely to benefit from these new therapies. CONCLUSIONS The aims of studying molecular markers are to determine prognosis and identify patients who are most likely to benefit from chemotherapy. Analysis of prognostic markers such as LOH at chromosome 18q may help to identify patients as either, likely to be cured by surgery alone, or require adjuvant treatment.16 The study using TS, TP and DPD expression to predict response to 5FU-based chemotherapy demonstrates that combining information from more than one independent predictive marker has the potential to identify a high percentage of responding patients.48 In the future, simultaneous analysis of a set of key markers could be used for tumour evaluation, possibly using DNA microarray technology. If such testing

www.nature.com/tpj


214

predicted resistance to a particular drug, alternative therapies could be used. It is important to note that in the metastatic disease setting, TS expression in the primary tumour has been found not to correlate with expression at metastatic sites. This may be a general phenomenon as a similar finding has recently been reported for Bax.73 Clearly, the possibility of differential expression of a marker in primary and metastatic sites must be considered when attempting to predict disease response to chemotherapy. Many of the new therapies that are being developed have targeted mechanisms of action and therefore have readily identifiable potential predictive markers (eg EGFR for ZD-1839). DNA microarray and proteomic technologies have the potential to identify many new candidate markers and to facilitate evaluation of their clinical significance. Meta-analyses may help to resolve problems that arise when different studies produce conflicting results such as the studies into the prognostic and predictive value of p53. Ultimately, the usefulness of a marker will come from a demonstration that application of the information it provides improves outcome in a prospective trial. DUALITY OF INTEREST None declared. Correspondence should be sent to PG Johnston, Cancer Research Centre, Queen’s University Belfast, University Floor, Belfast City Hospital, Lisburn Road, Belfast BT9 7AB, Northern Ireland. Tel: 44-2890-263911 Fax: 44-2890-263744 E-mail: oncology얀qub.ac.uk

REFERENCES 1 Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990; 61: 759–767. 2 Johnston PG, Allegra CJ. Colorectal cancer biology: clinical implications. Semin Oncol 1995; 22: 418–432. 3 Lane DP. Cancer p53, guardian of the genome. Nature 1992; 358: 15–16. 4 Sjogren S, Inganas M, Norberg T, Lindgren A, Nordgren H, Holmberg L et al. The p53 gene in breast cancer: prognostic value of complementary DNA sequencing versus immunohistochemistry. J Natl Cancer Inst 1996; 88: 173–182.


5 Kaklamanis L, Savage A, Whitehouse R, Doussis-Anagnostopoulou I, Biddolph S, Tsiotos P et al. Bcl-2 protein expression: association with p53 and prognosis in colorectal cancer. Br J Cancer 1998; 77: 1864–1869. 6 Bosari S, Viale G, Bossi P, Maggioni M, Coggi G, Murray JJ et al. Cytoplasmic accumulation of p53 protein: an independent prognostic indicator in colorectal adenocarcinomas. J Natl Cancer Inst 1994; 86: 681–687. 7 Soong R, Grieu F, Robbins P, Dix B, Chen D, Parsons R et al. p53 alterations are associated with improved prognosis in distal colonic carcinomas. Clin Cancer Res 1997; 3: 1405–1411. 8 Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG et al. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001; 344: 1196–1206. 9 Allegra CJ, Parr AL, Wold LE, Mahoney MR, Sargent DJ, Johnston PG et al. Investigation of the prognostic and predictive value of thymidylate synthase, p53 and Ki-67 in patients with locally advanced colon cancer. J Clin Oncol 2002; 20: 1735-1743. 10 Kressner U, Inganas M, Byding S, Blikstad I, Pahlman L, Glimelius B et al. Prognostic value of p53 genetic changes in colorectal cancer. J Clin Oncol 1999; 17: 593–599. 11 Bunz F, Hwang PM, Torrance C, Waldman T, Zhang Y, Dillehay L et al. Disruption of p53 in human cancer cells alters the responses to therapeutic agents. J Clin Invest 1999; 104: 263–269. 12 Longley DB, Boyer J, Allen WL, Latif T, Ferguson PR, Maxwell PJ et al. The role of thymidylate synthase induction in modulating p53 regulated gene expression in response to 5-fluorouracil and antifolates. Cancer Res 2002; 62: 2644–2649. 13 Ahnen DJ, Feigl P, Quan G, Fenoglio-Preiser C, Lovato LC, Bunn PA Jr et al. Ki-ras mutation and p53 overexpression predict the clinical behavior of colorectal cancer: a Southwest Oncology Group study. Cancer Res 1998; 58: 1149–1458. 14 Elsaleh H, Powell B, Soontrapornchai P, Joseph D, Goria F, Spry N et al. p53 gene mutation, microsatellite instability and adjuvant chemotherapy: impact on survival of 388 patients with Dukes’ C colon carcinoma. Oncology 2000; 58: 52–59. 15 O’Connell MJ, Schaid DJ, Ganju V, Cunningham J, Kovach JS, Thibodeau SN. Current status of adjuvant chemotherapy for colorectal cancer. Can molecular markers play a role in predicting prognosis? Cancer 1992; 70: 1732–1739. 16 Jen J, Kim H, Piantadosi S, Liu ZF, Levitt RC, Sistonen P et al. Allelic loss of chromosome 18q and prognosis in colorectal cancer. N Engl J Med 1994; 331: 213–221. 17 Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG et al. Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001; 344: 1196–1206. 18 Keino-Masu K, Masu M, Hinck L, Leonardo ED, Chan SS, Culotti JG et al. Deleted in

19

20

21

22

23

24

25

26

27

28

29

30

31

32

Colorectal Cancer (DCC) encodes a netrin receptor. Cell 1996; 87: 175–185. Attisano L, Lee-Hoeflich ST. The Smads. Genome Biology 2001; 2: reviews 3010.1– 3010.8. Duff EK, Clarke AR. Smad4 (DPC4)—a potent tumour suppressor? Br J Cancer 1998; 78: 1615–1619. Tanaka K, Kikuchi-Yanoshita R, Muraoka M, Konishi M, Oshimura M, Miyaki M. Suppression of tumorigenicity and invasiveness of colon carcinoma cells by introduction of normal chromosome 8p12-pter. Oncogene 1996; 12: 405–410. Kelemen PR, Yaremko ML, Kim AH, Montag A, Michelassi F, Westbrook CA. Loss of heterozygosity in 8p is associated with microinvasion in colorectal carcinoma. Genes Chromosomes Cancer 1994; 11: 195–198. Halling KC, French AJ, McDonnell SK, Burgart LJ, Schaid DJ, Peterson BJ et al. Microsatellite instability and 8p allelic imbalance in stage B2 and C colorectal cancers. J Natl Cancer Inst 1999; 91: 1295– 1303. Gryfe R, Kim H, Hsieh ET, Aronson MD, Holowaty EJ, Bull SB et al. Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 2000; 342: 69–77. Elsaleh H, Powell B, McCaul K, Grieu F, Grant R, Joseph D et al. P53 alteration and microsatellite instability have predictive value for survival benefit from chemotherapy in stage III colorectal carcinoma. Clin Cancer Res 2001; 7: 1343–1349. Kressner U, Bjorheim J, Westring S, Wahlberg SS, Pahlman L, Glimelius B et al. Ki-ras mutations and prognosis in colorectal cancer. Eur J Cancer 1998; 34: 518–521. Gnanasampanthan G, Elsaleh H, McCaul K, Iacopetta B. Ki-ras mutation type and the survival benefit from adjuvant chemotherapy in Dukes’ C colorectal cancer. J Pathol 2001; 195: 543–548. Johnston SRD. Farnesyl transferase inhibitors: a novel targeted therapy for cancer. Oncology 2001; 2: 18–26. Erisman MD, Rothberg PG, Diehl RE, Morse CC, Spandorfer JM, Astrin SM. Deregulation of c-myc gene expression in human colon carcinoma is not accompanied by amplification or rearrangement of the gene. Mol Cell Biol 1985; 5: 1969–1976. Arango D, Corner GA, Wadler S, Catalano PJ, Augenlicht LH. c-myc/p53 interaction determines sensitivity of human colon carcinoma cells to 5-fluorouracil in vitro and in vivo. Cancer Res 2001; 61: 4910–4915. Johnston PG, Lenz HJ, Leichman CG, Danenberg KD, Allegra CJ, Danenberg PV et al. Thymidylate synthase gene and protein expression correlate and are associated with response to 5-fluorouracil in human colorectal and gastric tumors. Cancer Res 1995; 55: 1407–1412. Peters GJ, Kohne CH. Fluoropyrimidines as antifolate drugs. In: Jackman AL (ed). Antifolate Drugs in Cancer Therapy, Humana Press: Totowa New Jersey, 1999, pp 101– 145.


215

33 Johnston PG, Kaye S. Capecitabine: a novel agent for the treatment of solid tumors. Anticancer Drugs 2001; 12: 639–646. 34 Poon MA, O’Connell MJ, Moertel CG, Wieand HS, Cullinan SA, Everson LK et al. Biochemical modulation of fluorouracil: evidence of significant improvement of survival and quality of life in patients with advanced colorectal carcinoma. J Clin Oncol 1989; 7: 1407–1418. 35 Johnston PG, Drake JC, Trepel J, Allegra CJ. Immunological quantitation of thymidylate synthase using the monoclonal antibody TS 106 in 5-fluorouracil-sensitive and -resistant human cancer cell lines. Cancer Res 1992; 52: 4306–4312. 36 Lenz HJ, Hayashi K, Salonga D, Danenberg KD, Danenberg PV, Metzger R et al. p53 point mutations and thymidylate synthase messenger RNA levels in disseminated colorectal cancer: an analysis of response and survival. Clin Cancer Res 1998; 4: 1243– 1250. 37 Edler D, Gilmelius B, Hallstrom M, Jakobsen A, Johnston PG, Magnusson I et al. Thymidylate synthase expression in colorectal cancer: a prognostic and predictive marker of benefit from adjuvant fluorouracil-based chemotherapy. J Clin Oncol 2002; 20: 1721–1728. 38 Edler D, Kressner U, Ragnhammar P, Johnston PG, Magnusson I, Glimelius B et al. Immunohistochemically detected thymidylate synthase in colorectal cancer: an independent prognostic factor of survival. Clin Cancer Res 2000; 6: 488–492. 39 Marsh S, McLeod HL. Thymidylate synthase pharmacogenomics in colorectal cancer. Clinical Colorectal Cancer 2001; 1: 175–178. 40 Krajinovic M, Costea I, Chiasson S. Polymorphism of the thymidylate synthase gene and outcome of acute lymphoblastic leukaemia. Lancet 2002; 359: 1033–1034. 41 Yamada H, Ichikawa W, Uetake H, Shirota Y, Nihei Z, Sugihara K et al. Thymidylate synthase gene expression in primary colorectal cancer and metastatic sites. Clinical Colorectal Cancer 2001; 1: 169–173. 42 Gorlick R, Metzger R, Danenberg KD, Salonga D, Miles JS, Longo GS, et al. Higher levels of thymidylate synthase gene expression are observed in pulmonary as compared with hepatic metastases of colorectal adenocarcinoma. J Clin Oncol 1998; 16: 1465–1469. 43 Findlay MP, Cunningham D, Morgan G, Clinton S, Hardcastle A, Aherne GW. Lack of correlation between thymidylate synthase levels in primary colorectal tumours and subsequent response to chemotherapy. Br J Cancer 1997; 75: 903–909. 44 Evrard A, Cuq P, Ciccolini J, Vian L, Cano JP. Increased cytotoxicity and bystander effect of 5-fluorouracil and 5-deoxy-5-fluorouridine in human colorectal cancer cells transfected with thymidine phosphorylase. Br J Cancer 1999; 80: 1726–1733. 45 Metzger R, Danenberg K, Leichman CG, Salonga D, Schwartz EL, Wadler S et al. High basal level gene expression of thymidine phosphorylase (platelet-derived endothelial cell growth factor) in colorectal

46

47

48

49

50

51

52

53

54

55

56

57

tumors is associated with nonresponse to 5fluorouracil. Clin Cancer Res 1998; 4: 2371–2376. Takebayashi Y, Akiyama S, Akiba S, Yamada K, Miyadera K, Sumizawa T et al. Clinicopathologic and prognostic significance of an angiogenic factor, thymidine phosphorylase, in human colorectal carcinoma. J Natl Cancer Inst 1996; 88: 1110–1117. Lyss AP, Lilenbaum RC, Harris BE, Diasio RB. Severe 5-fluorouracil toxicity in a patient with decreased dihydropyrimidine dehydrogenase activity. Cancer Invest 1993; 11: 239–240. Salonga D, Danenberg KD, Johnson M, Metzger R, Groshen S, Tsao-Wei DD et al. Colorectal tumors responding to 5-fluorouracil have low gene expression levels of dihydropyrimidine dehydrogenase, thymidylate synthase, and thymidine phosphorylase. Clin Cancer Res 2000; 6: 1322–1327. Ishikawa T, Utoh M, Sawada N, Nishida M, Fukase Y, Sekiguchi F et al. Tumor selective delivery of 5-fluorouracil by capecitabine, a new oral fluoropyrimidine carbamate, in human cancer xenografts. Biochem Pharmacol 1998; 55: 1091–1097. Hughes LR, Stephens TC, Boyle FT, Jackman AL. Raltitrexed (Tomudex) a highly polyglutamatable antifolate thymidylate synthase inhibitor. In: Jackman AL (ed). Antifolate Drugs in Cancer Therapy. Humana Press: Totowa New Jersey, 1999, pp 147–165. Pazdur R, Vincent M. Raltitrexed (Tomudex) vs 5-fluorouracil leucovorin (5FU + LV) in patients with advanced colorectal cancer (ACC): results of a randomised multicenter North American trial. Proc Am Soc Clin Oncol 1997; 16: 801. Longley DB, Ferguson PR, Boyer J, Latif T, Lynch M, Maxwell P et al. Characterization of a thymidylate synthase (TS)-inducible cell line: a model system for studying sensitivity to TS- and non-TS-targeted chemotherapies. Clin Cancer Res 2001; 7: 3533– 3539. Drake JC, Allegra CJ, Moran RG, Johnston PG. Resistance to tomudex (ZD1694): multifactorial in human breast and colon carcinoma cell lines. Biochem Pharmacol 1996; 51: 1349–1355. John W, Picus J, Blanke C, Clark J, Schulman L, Rowinsky E et al. Activity of MTA in patients with advanced colorectal cancer— results from a phase II study. Proc Am Soc Clin Oncol 1997; 17: 1067. Voigt W, Matsui S, Yin MB, Burhans WC, Minderman H, Rustum YM. TopoisomeraseI inhibitor SN-38 can induce DNA damage and chromosomal aberrations independent from DNA synthesis. Anticancer Res 1998; 18: 3499–3505. Pitot HC, Wender DB, O’Connell MJ, Schroeder G, Goldberg RM, Rubin J et al. Phase II trial of irinotecan in patients with metastatic colorectal carcinoma. J Clin Oncol 1997; 15: 2910–2919. Douillard JY, Cunningham D, Roth AD, Navarro M, James RD, Karasek P et al. Irinotecan combined with fluorouracil compared with fluorouracil alone as first-line treatment for metastatic colorectal cancer:

58

59

60

61

62

63

64

65

66

67

68

a multicentre randomised trial. Lancet 2000; 355: 1041–1047. Cunningham D, Pyrhonen S, James RD, Punt CJ, Hickish TF, Heikkila R et al. Randomised trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 1998; 352: 1413–1418. Goldwasser F, Bae I, Valenti M, Torres K, Pommier Y. Topoisomerase I-related parameters and camptothecin activity in the colon carcinoma cell lines from the National Cancer Institute anticancer screen. Cancer Res 1995; 55: 2116–2121. Jacob S, Aguado M, Fallik D, Praz F. The role of the DNA mismatch repair system in the cytotoxicity of the topoisomerase inhibitors camptothecin and etoposide to human colorectal cancer cells. Cancer Res 2001; 61: 6555–6562. Fink D, Nebel S, Aebi S, Zheng H, Cenni B, Nehme A 2et al. The role of DNA mismatch repair in platinum drug resistance. Cancer Res 1996; 56: 4881–4886. Raymond E, Faivre S, Woynarowski JM, Chaney SG. Oxaliplatin: mechanism of action and antineoplastic activity. Semin Oncol 1998; 25: 4–12. Giacchetti S, Perpoint B, Zidani R, Le Bail N, Faggiuolo R, Focan C et al. Phase III multicenter randomized trial of oxaliplatin added to chronomodulated fluorouracil-leucovorin as first-line treatment of metastatic colorectal cancer. J Clin Oncol 2000; 18: 136–147. Metzger R, Leichman CG, Danenberg KD, Danenberg PV, Lenz HJ, Hayashi K et al. ERCC1 mRNA levels complement thymidylate synthase mRNA levels in predicting response and survival for gastric cancer patients receiving combination cisplatin and fluorouracil chemotherapy. J Clin Oncol 1998; 16: 309–316. Shirota Y, Stoehlmacher J, Brabender J, Xiong YP, Uetake H, Danenberg KD et al. ERCC1 and thymidylate synthase mRNA levels predict survival for colorectal cancer patients receiving combination oxaliplatin and fluorouracil chemotherapy. J Clin Oncol 2001; 19: 4298–4304. Ciardiello F, Caputo R, Bianco R, Damiano V, Pomatico G, De Placido S et al. Antitumor effect and potentiation of cytotoxic drugs activity in human cancer cells by ZD1839 (Iressa), an epidermal growth factor receptor–selective tyrosine kinase inhibitor. Clin Cancer Res 2000; 6: 2053–2063. Cascinu S, Graziano F, Valentini M, Catalano V, Giordani P, Staccioli MP et al. Vascular endothelial growth factor expression, Sphase fraction and thymidylate synthase quantitation in node-positive colon cancer: relationships with tumor recurrence and resistance to adjuvant chemotherapy. Ann Oncol 2001; 12: 239–244. Gordon MS, Margolin K, Talpaz M, Sledge GW Jr, Holmgren E, Benjamin R et al. Phase I safety and pharmacokinetic study of recombinant human anti-vascular endothelial growth factor in patients with

www.nature.com/tpj


216

advanced cancer. J Clin Oncol 2001; 19: 843–850. 69 Schally AV, Nagy A. Chemotherapy targeted to hormone receptors on tumors. Eur J Endocrinol 1999; 141: 1–14. 70 Szepeshazi K, Schally AV, Halmos G, Armatis P, Hebert F, Sun B et al. Targeted cytotoxic somatostatin analogue AN-238 inhibits somatostatin receptor-positive experimental colon cancers independently


71

72

of their p53 status. Cancer Res 2002; 62: 781–788. Sheehan KM, Sheahan K, O’Donoghue DP, MacSweeney F, Conroy RM, Fitzgerald DJ et al. The relationship between cyclooxygenase-2 expression and colorectal cancer. JAMA 1999; 282: 1254–1257. Chen WS, Wei SJ, Liu JM, Hsiao M, Kou-Lin J, Yang WK. Tumor invasiveness and liver metastasis of colon cancer cells correlated

with cyclooxygenase-2 (COX-2) expression and inhibited by a COX-2-selective inhibitor, etodolac. Int J Cancer 2001; 91: 894– 899. 73 Jansson A, Sun XF. Bax expression decreases significantly from primary tumor to metastasis in colorectal cancer. J Clin Oncol 2002; 20: 811–816.