Clinical Significance of Prognostic and Predictive Markers in Colorectal
Total Page:16
File Type:pdf, Size:1020Kb
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 multiple genetic abnormalities are Clinical significance of prognostic accumulated over time during the pro- gression from adenoma to adenocarci- and predictive markers in colorectal noma (Figure 1).1 Mutations in genes such as the Kirsten-ras (K-ras), aden- cancer omatous polyposis coli (APC), deleted in colon cancer (DCC) and the p53 DB Longley, U McDermott and PG Johnston tumour suppressor gene are the most common genetic alterations found in Department of Oncology, Cancer Research Centre, Queen’s University Belfast, Belfast, sporadic CRC. Approximately 5% of Northern Ireland CRC is inherited and familial aden- omatous polyposis (FAP) and heredi- tary non-polyposis colon cancer The Pharmacogenomics Journal (2002) 2, fication and microsatellite instability (HNPCC) are the most well-charac- 209–216. doi:10.1038/sj.tpj.6500124 (MSI). In addition, the expression of terized syndromes.2 FAP is caused by individual genes can be assessed at the mutations in the APC gene, whereas INTRODUCTION levels of mRNA and protein using the 90% of HNPCC cases are caused by Colorectal cancer (CRC) is the second techniques of real-time reverse tran- mutations in genes involved in mis- most common cause of cancer death scription PCR and immunohistochem- match repair (MMR) leading to in the US. Approximately 130 000 new istry (IHC) respectively. The ultimate microsatellite instability (MSI). In cases of CRC are diagnosed in the US goal of this research is the tailoring of addition, 10–15% of sporadic cases of each year with an annual mortality of treatment to the molecular pheno- CRC have mutations in MMR genes. approximately 56 000 people. In the types of tumour and patient. This post-operative setting, there is clear review highlights potentially Loss of heterozygozity (LOH) evidence that offering adjuvant ther- important prognostic and predictive More than 75% of CRCs have lost a apy significantly improves survival in markers in CRC. Due to space restric- portion of chromosome 17p or 18q, or stage III CRC, however, the manage- tions many worthy contributions both. The 17p segment contains p53, ment of stage II disease is an area of could not be cited. which is regarded as one of the most controversy and debate. The 5-year important tumour suppressor genes. It survival for stage II patients is approxi- Colorectal Cancer Biology has a number of key functions such as mately 75%, indicating that the The genetic model of CRC proposed DNA damage repair, initiation of cell majority are cured by surgery alone, by Vogelstein and others suggests that cycle checkpoints and initiation of and many patients who would have been cured by surgery undergo chemo- therapy unnecessarily. Furthermore, Table 1 Prognostic molecular markers in colorectal cancer some patients who are not rec- ommended to receive adjuvant treat- Marker Measurement technique Comments ment may benefit from chemotherapy. Recent studies have started to define p53(LOH 17p) IHC, DNA sequencing p53 overexpression measured by IHC generally prognostic molecular markers that associated with worse prognosis, however, not a identify those patients at risk of consistent observation. p53 mutation by sequencing associated with shorter survival time. relapse after surgery and who may LOH 18q RFLP, PCR, Southern Blot Associated with worse prognosis. Locus contains benefit most from chemotherapy 3 potential tumour suppressor genes: DCC, (Table 1). Furthermore, molecular mar- Smad2, Smad4. kers are being identified that predict LOH 8p RFLP, PCR, Southern Blot Associated with worse prognosis. Identity and whether a tumour will respond to a function of genes at this locus not determined. particular chemotherapy. In addition, MSI PCR MSI-H tumours associated with survival some markers may have therapeutic advantage and less aggressive clinical course. value by predicting the degree of sys- K-ras SSCP, DNA sequencing Some studies have found an association between K-ras mutation and survival, but others have not. temic toxicity for a particular treat- The particular type of mutation may be ment. For such studies tumour DNA important. can be analysed to identify genetic TS IHC, RT-PCR, genotyping High TS expression and TSER*3/TSER*3 genotype changes such as point mutations, loss associated with worse clinical outcome. of heterozygosity (LOH), gene ampli- Molecular markers for colorectal cancer DB Longley et al 210 Chromosome: 5q 12p 18q 17p Alteration: loss/mutation Mutation loss loss/mutation Gene: APC K-ras DCC, Smad 2/4 p53 Other alterations DNA methylation changes Normal Hyperprolif Early Intermediate Late Carcinoma Metastasis epithelium epithelium Adenoma Adenoma Adenoma Figure 1 A genetic model for colorectal carcinogenesis (adapted from Fearon and Vogelstein1). apoptosis.3 p53 has been the most whereas those without p53 over- cer and appear to have tumour sup- extensively studied molecular marker expression did.13 However, other stud- pressor functions.20 in patients with CRC and mutations ies have found no correlation between Loss of material from the short arm have been found in 40–60% of cases.2 p53 status and response to 5FU-based of chromosome 8 is found in approxi- Missense mutations of the p53 gene chemotherapy.14 mately 50% of cases of CRC. The intro- result in production of a protein with Allelic loss on chromosome 18q is duction of chromosome 8p into colon a longer half-life, therefore, detection found in approximately 70% of CRCs cancer cell lines has been shown to of increased p53 expression by IHC and has been associated with worse decrease their tumourigenecity and has been used as a surrogate for p53 prognosis in stage II and III invasiveness.21 Furthermore, LOH at mutation in many studies. However, tumours.15,16 Jen et al found that stage 8p has been shown to correlate with this assumption has been shown to be II patients who retained both 18q the presence of micrometastases.22 valid in only 60–80% of cases in stud- alleles have an excellent 5-year sur- Halling et al found that in stage II and ies that have compared p53 expression vival of 93% compared to 54% of III CRCs 8p allelic imbalance was an measured by IHC with p53 mutations those who had lost one allele.16 This independent prognostic indicator for analysed by DNA sequencing.4 Despite data suggests that patients with stage decreased survival and time to recur- numerous investigations into the II disease who have a deletion on chro- rence.23 This was the first large study prognostic value of p53 over- mosome 18q have a more aggressive exploring the role of 8p allelic imbal- expression in patients with CRC, the clinical course and may represent a ance in prognosis of CRC. issue still remains controversial. Many subset of stage II patients who would studies have been statistically benefit from adjuvant chemotherapy. Microsatellite instability (MSI) underpowered, and the issue is further Furthermore, a recent study by Watan- MSI is caused by mutations in MMR complicated by the use of different abe et al found that retention of both genes such as hMSH2, hMLH1 and antibodies which have varying levels 18q alleles predicted for favourable hMSH6 resulting in failure of the DNA of sensitivity and capacity for outcome after adjuvant 5FU-based mismatch repair system to correct detecting mutant forms of p53. Studies chemotherapy in stage III CRC.17 The errors that occur during replication. of at least 100 patients with CRC using 18q segment contains three candidate Numerous studies have shown a sur- p53 monoclonal antibodies have gen- tumour suppressor genes: DCC, Smad vival advantage and a less aggressive erally found that p53 overexpression 2 and Smad 4. The product of the DCC clinical course for patients with high was associated with worse prognosis.5,6 gene is a cell adhesion molecule.18 and frequency MSI.23,24 Furthermore, MSI- This association has not been a con- it has been postulated that decreased high tumours have been reported to sistent observation as some studies DCC expression may lead to altered metastasise less often than microsatel- have found the opposite,7 or no corre- adhesion and contribute to enhanced lite stable tumours.24 The MSI pheno- lation.8,9 However, the majority of tumour growth and metastatic spread type also appears to be associated with DNA sequencing studies have shown a of colorectal tumours. Smad proteins excellent survival in patients who significantly shorter survival time in are transcription factors involved in receive adjuvant 5FU-based chemo- patients with mutant p53.10 Data from the transforming growth factor  therapy.25 A recent study found that studies in cell line models have sug- (TGF-) pathway.19 These proteins are 61% of stage III CRC patients with gested that p53 status influences 5FU involved in signalling from TGF- MSI-high tumours had a mutation in chemosensitivity.11,12 This has been receptors and regulate transcription of the gene for the type II receptor for borne out in several clinical studies, target genes including c-myc, CBFA1, TGF-1.17 Furthermore, these patients for example Ahnen et al found that FLRF and furin. Targeted disruption of had a 74% 5-year survival following patients with stage III CRC whose Smad genes in mice has revealed their adjuvant 5FU-based chemotherapy tumours overexpressed p53 did not importance in embryonic develop- compared to 46% in patients with derive significant survival benefit from ment. Furthermore, Smad 2 and Smad MSI-high tumours that lacked the adjuvant 5FU-based treatment, 4 have been implicated in human can- mutation. Thus, mutation of the gene The Pharmacogenomics Journal Molecular markers for colorectal cancer DB Longley et al 211 for the type II receptor for TGF-1 may activated by addition of a farnesyl used in the treatment of metastatic predict the response of MSI-high group, therefore, inhibitors of farnesyl CRC, and recent studies have begun to tumours to 5FU-based chemotherapy.