|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2004.02.945 on March 29 2004 © 2004 American Society of Clinical Oncology.
Of What Value Genomics in Colorectal Cancer? Opportunities and ChallengesDepartment of Oncology, Centre for Cancer Research, Queen's University Belfast, Belfast, Northern Ireland Colorectal cancer (CRC) affects more than 160,000 patients each year in the United States, and approximately 60,000 patients die from this disease.1 Although it is a highly treatable and often curable disease when localized to the bowel, the prognosis for recurrent metastatic disease remains poor and most often is the ultimate cause of death. In the postoperative setting, there is clear evidence that adjuvant chemotherapy significantly improves outcome in patients with CRC.2,3 Chemotherapeutic drugs such as the fluoropyrimidines, irinotecan and oxaliplatin are now used as part of standard care and have been shown to significantly improve survival in patients with widespread metastatic disease.4-6 However, the management of patients with potentially curative, earlier stage disease (stage II and III) remains an area of debate, as the combined 5-year survival for these patients is 60%. Indeed, only one third of the 40% of patients at risk for relapse after primary surgical treatment derive any benefit from chemotherapeutic treatment. This means that giving adjuvant chemotherapy to stage III patients exposes more than 50% to treatment that may be unnecessary. In addition, the benefit of adjuvant chemotherapy in patients with stage II disease still remains very controversial, with a recent meta-analysis of prospective randomized clinical trials of adjuvant therapy not showing a significant survival benefit.7 Therefore, many more patients undergo chemotherapy for their disease who may not derive any significant clinical benefit. The identification of patients at high risk of relapse has traditionally depended on pathologic features, such as depth of bowel wall penetration, lymph node involvement, and adherence to, or invasion of, adjacent organs. These anatomic characteristics thus create populations with a similar clinical stage of disease, but with wide variability in cure rates.8,9 However, anatomic staging systems have been too imprecise to provide detailed predictions of individual patient outcome. Recent advances in our understanding of the molecular biology of CRC have led to the identification of other potential prognostic and predictive factors. Molecular diagnostic markers such as microsatellite instability, loss of heterozygosity, and thymidylate synthase level already identify patients with increased risk of relapse.10-12 However, cancer treatment and colorectal cancer clinical trials are still being based on the availability of new active compounds, rather than on an integrated approach that uses the genetic makeup of the tumor and the genotype of the patient. More recently, the application of microarray technology to the cancer field has created the potential to obtain large-scale gene expression profiles or genetic fingerprints of colorectal cancer. This has the potential to impact very significantly on the diagnostic capability and prognostic classification of tumors, which may aid in the prediction of response of individual patients to defined chemotherapeutic regimens.13 In this issue of the Journal of Clinical Oncology, Wang et al present the results of an important study using gene expression profiling to define a set of prognostic molecular markers in patients with stage II CRC.14 They harvested RNA from 74 patients with stage II CRC and performed gene expression profiling using an oligonucleotide microarray gene chip containing approximately 22,000 transcripts. To identify those genes that may identify patients with a high risk of relapse, the investigators used two supervised class prediction approaches to select markers from the 17,616 informative genes. The first involved dividing all 74 patients into a training set and a testing set, with equal numbers in each. The training set was used to select gene markers and to build a prognostic signature, and the testing set was used for independent validation. Using this approach, 60 genes were selected from the 38 patients in the training set, and a Cox regression model was then used to predict patient recurrence. The second step was to place patients into one of two subgroups based on the result of clustering analysis. Each patient subgroup was divided into a training and testing set and analyzed separately to select markers. In total, 23 genes were selected from both subgroups, and then used in the training set. The investigators compared the 60-gene and 23-gene predictor subsets and found that only the 23-gene signature profile, derived using the second approach, was predictive. This gene signature was validated in 36 independent patients with an overall accuracy of 78%. Thirteen of 18 patients with relapse and 15 of 18 disease-free patients were correctly predicted, representing an odds ratio of 13, which is considerably higher than the more established prognostic factors currently in clinical practice. This study is the first example in CRC of a genomics approach systematically identifying molecular markers for colon cancer prognosis. It clearly demonstrates the potential of DNA microarray technology and describes how recognition of specific gene expression patterns may help in the prediction of patient outcome in this disease. Moreover, it points toward the redefinition of the staging of colorectal cancer based on genetic signature or molecular phenotype, rather than solely on the anatomic and pathologic staging of the primary tumor. The authors conclude that this molecularly-defined poor prognostic subgroup of stage II patients may represent those to whom therapy should be recommended. One must be cautious in this interpretation, however, as the ability to predict a poor clinical prognosis does not necessarily indicate a benefit from an intervention. However, it does provide the opportunity to define a group who are likely to do well with surgery alone and who would not necessarily require any adjuvant treatment. Our understanding of CRC biology has progressed rapidly over the last 15 years, associated with an increasing number of drugs becoming available to treat this disease. A major challenge will be to make the most effective use of this increasing arsenal of chemotherapeutic and novel targeted agents. This must involve genomic approaches, such as those identified in this study, as they provide the potential to radically improve response and to spare patients the harmful side effects of chemotherapeutic treatment from which some are unlikely to benefit. Although this high throughput array technology holds exciting promise for the future, the ability to fulfill that potential will depend on the characterization of tissue and blood samples, collected from well characterized and uniformly treated patient populations, using validated and standardized assays. Moreover, the clinical value of these predictive and prognostic gene sets in colorectal cancer must ultimately be validated in prospective clinical trials. Unless the next generation of clinical studies is designed to test the therapeutic as well as prognostic value of defined gene sets, the true clinical potential of this new technology and new therapeutic agents may not be realized. Author's Disclosures of Potential Conflicts of Interest The author indicated no potential conflicts of interest. REFERENCES
1. Jemal A, Thomas A, Murray T, et al: Cancer statistics 2002. Ca Cancer J Clin 52:23-47, 2002 2. Van Cutsem E, Dicato M, Wils J: Recent advances in the management of colorectal cancer. Eur J Cancer 37:2302-2309, 2001
3. Sobrero AF, Aschele C, Bertino JR: Fluorouracil in colorectal cancer: A tale of two drugs—Implications for biochemical modulation. J Clin Oncol 15:368-381, 1997 4. Doulliard JY, Cunningham D, Roth AD, et al: Irinotecan combined with fluorouracil compared with fluorouracil alone as first line treatment for colorectal cancer: A multicentred randomised trial. Lancet 355:1041-1047, 2000[CrossRef][Medline]
5. Saltz LB, Cox JV, Blanke C, et al: Irinotecan Study Group. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. N Engl J Med 343:905-914, 2000
6. Goldberg R, Sargeant DJ, Roscoe F, et al: A randomised controlled trial of fluorouracil plus leucovorin, irinotecan and oxaliplatin combinations in patients with previously untreated metastatic colorectal cancer. J Clin Oncol 22:23-30, 2004
7. International multi-centered pooled analysis of B2 colon cancer trials (Impact B2) investigators: Efficacy of adjuvant fluorouracil and folinic acid in B2 colon cancer. J Clin Oncol 17:1356-1363, 1999 8. Fisher ER, Sass R, Palekar A, et al: Duke's Classification revisited. Cancer 64:2354-2360, 1989[CrossRef][Medline] 9. Astler VB, Cioller FA: The prognostic significance of direct extension of carcinoma of the colon and rectum. Ann Surg 139:846-852, 1954[Medline] 10. McLeod HL, Murray GI: Tumor markers of prognosis in colorectal cancer. Br J Cancer 79:191-203, 1999[Medline]
11. Watanabe T, Wu TT, Catalano PJ, et al: Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 344:196-1206, 2001
12. Allegra CJ, Paik S, Colangelo LS, et al: Prognostic value of thymidylate synthase Ki-67 and p53 in patients with Dukes’ B and C colon cancer and NCI-NSABP collaborative study. J Clin Oncol 21:241-250, 2003 13. Longley DB, Harkin DP, Johnston PG: 5-Fluorouracil: Mechanisms of action and clinical strategies. Nat Rev Cancer 3:330-338, 2003[CrossRef][Medline]
14. Wang Y, Jatkoe T, Zhang Y, et al: Gene expression profiles and molecular markers to predict recurrence of Dukes B colon cancer. J Clin Oncol 22:1564-1571, 2004
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2004 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
