|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2004.08.186 on March 29 2004 © 2004 American Society of Clinical Oncology. Gene Expression Profiles and Molecular Markers To Predict Recurrence of Dukes' B Colon CancerFrom Veridex, LLC, a Johnson & Johnson Company, San Diego, CA; and the Departments of Medicine and Surgery, Washington University School of Medicine, St Louis, MO Address reprint requests to Yixin Wang, Veridex, LLC, a Johnson & Johnson Company, 3210 Merryfield Row, San Diego, CA 92121; e-mail: ywang44{at}vrxus.jnj.com
PURPOSE: The 5-year survival rate of patients with Dukes' B colon cancer is approximately 75%. Identification of the patients at high risk of recurrence in this group would allow better staging and more informed use of adjuvant chemotherapy. In this study, we used DNA chip technology to systematically identify new prognostic markers for tumor relapse in Dukes' B patients. PATIENTS AND METHODS: Using Affymetrix U133a GeneChip containing approximately 22,000 transcripts (Affymetrix, Santa Clara, CA), RNA samples from 74 patients with Dukes' B colon cancer were analyzed. Thirty-one patients developed tumor relapse in less than 3 years, whereas 43 patients remained disease-free for more than 3 years after surgery. Two supervised class prediction approaches were used to identify gene markers that can best discriminate between patients who would experience relapse and patients who would remain disease-free. A multivariate Cox model was built to predict recurrence. RESULTS: Gene expression profiling identified a 23-gene signature that predicts recurrence in Dukes'B patients. This signature was validated in 36 independent patients. The overall performance accuracy was 78%. Thirteen of 18 relapse patients and 15 of 18 disease-free patients were predicted correctly, giving an odds ratio of 13 (95% CI, 2.6 to 65; P = .003). The log-rank test indicated a significant difference in disease-free time between the predicted relapse and disease-free patients (P = .0001). CONCLUSION: The clinical value of these markers is that the patients at a high predicted risk of relapse (13-fold risk) could be upstaged to receive adjuvant therapy, similar to Dukes' C patients. Our data highlight the feasibility of a prognostic assay that could focus more intensive treatment for localized colon cancer.
Colorectal cancer is one of the common malignancies, with an estimated annual incidence of 135,000 cases and 55,000 deaths in the United States.1 Like other solid tumors, colon cancer is staged pathologically on the basis of the extent of primary organ involvement and the metastatic spread to lymph nodes or distant organs.2–5 The presence of positive lymph nodes in Dukes' C colon cancer disease predicts a 60% likelihood of recurrence within 5 years. Treatment of patients with Dukes' C colon cancer with a postsurgical course of chemotherapy reduces the recurrence rate to 40% to 50% and is now the standard of care for Dukes' C patients.6–8 Dukes' A and B patients present with no lymph node or distant metastases and have a better prognosis. Surgical resection is highly effective for localized disease, but a significant proportion of Dukes' B patients (25% to 30%) develop recurrence and die from the disease. The benefit of postsurgical chemotherapy in Dukes' B patients has been harder to determine and remains controversial. A recent meta-analysis of prospective randomized clinical trials of adjuvant chemotherapy in patients with Dukes' B disease has not shown a survival benefit of chemotherapy in these patients.9 However, in contrast to the above analysis, a review by the National Surgical Adjuvant Breast and Bowel Project of four consecutive adjuvant chemotherapy trials in Dukes' B and C colorectal cancer argues that adjuvant chemotherapy does improve survival for certain Dukes' B patients.10 There is clearly a need to identify predictive factors, in addition to nodal involvement, to guide identification of Dukes' B patients who are likely to experience relapse. This information would allow more informed planning for patients who are more likely to require and possibly benefit from adjuvant therapy. In light of the importance of this issue, there have been many attempts to find novel markers to identify patients with the potential for colon cancer progression.11–16 Unfortunately, few have been prospectively evaluated using rigorous techniques. As such, several promising markers, including microsatellite instability,17–20 tumor DNA ploidy,21–24 chromosomal deletions,25–28 and p5329 have yet to become widely adopted as the basis of prognostic assays. Another possible reason for the shortcomings of these marker candidates is that colon cancer progression is a function of multiple genetic events that may arise within the malignant epithelium or may be induced or modified by stromal events. Recognizing the potential complexity of disease progression, we chose to use a more comprehensive assessment of molecular events. DNA microarray-based gene expression profiling technology provides a strategy to search systematically with a combinatorial manner for molecular markers of cancer classification and outcome prediction.30 Furthermore, several reports in breast,31,32 lung,33,34 and lymphoma35,36 cancers suggest that the simultaneous analysis of a large number of genes may offer a powerful and complementary approach to clinical or pathologic examination. Here we report the results of a highly complex gene expression analysis of 74 patients with Dukes' B colon cancer. The result of this analysis includes a gene expression-based algorithm that can identify, with high confidence, patients who have an increased probability of recurrence from the currently homogenously classified Dukes' B group.
Patient Samples Frozen tumor specimens from 74 coded Dukes' B colon cancer patients were obtained from the Siteman Cancer Center, Washington University School of Medicine (St Louis, MO) and Clinomics (Pittsfield, MA) according to an institutional review board-approved protocol. Archived primary tumor and adjacent nonneoplastic colon tissue were collected at the time of surgery. The histopathology of each specimen was reviewed to confirm diagnosis and uniform involvement with tumor. Regions chosen for analysis contained no normal or benign colon epithelium. The total cell population was composed of at least 85% tumor cells. Uniform follow-up information was available from the Division of Colorectal Surgery Patient Database at Washington University and Clinomics. Postsurgery patient surveillance was carried out according to general practice for colon cancer patients, including physical examination, blood counts, liver function tests, serum carcinoembryonic antigen, and colonoscopy for all patients. Selected patients had abdominal computed tomography scan and chest x-ray. If tumor relapse was suspected, the patient underwent intensive work-up, including abdominal/pelvic computed tomography scan, chest x-ray, colonoscopy, and biopsy, when applicable. Time to recurrence or disease-free time was defined as the time period from the date of surgery to confirmed tumor relapse date for relapsed patients and from the date of surgery to the date of last follow-up for disease-free patients.
Gene Expression Analysis
Statistical Methods To identify gene markers that can best discriminate between the relapse and the disease-free patients, we used two supervised class prediction approaches to select markers from the 17,616 informative genes (Fig 1). In the first approach, we divided all 74 patients into a training set and a testing set with approximately equal numbers of patients (Fig 1A). The training set was used to select gene markers and to build a prognostic signature. The testing set was used for independent validation (Fig 1A). In the second approach, the patients were first placed into one of the two subgroups based on the result of the clustering analysis (Fig 1B). Each patient subgroup was then analyzed separately to select markers. The patients in a subgroup were divided into a training set and a testing set with approximately equal numbers of patients. The training set was used to select gene markers. The markers selected from each subgroup were combined to form a single signature to predict tumor recurrence for all patients as a whole. This signature was used in the combined testing set for independent validation (Fig 1B).
The following statistical methods were used to analyze the training set in marker selection and class prediction. First, univariate Cox proportional hazards regression was used to identify genes for which expression levels were correlated with patient disease-free time. A P value less than .02 in the estimated regression coefficients was used as the selection criteria. Secondly, t test was used to select genes that gave the best classification between the relapse and the disease-free patients (P < .01). To reduce the chance of distorting the P value caused by outlier patients, resampling 100 times on the patients was performed on t test to search for genes that have a greater than 80% confidence level. In brief, for each resampling, 80% of patients in the training set were randomly sampled, and t test was performed for each gene. After 100 iterations, only genes that gave significant P values (corrected P < .01) in more than 80 resamplings were kept for subsequent analysis. Genes found by both Cox model and t test were selected to build a signature for predicting outcome. Relapse hazard score was used to determine each patient's risk of recurrence. The score was defined as the linear combination of weighted expression, with the standardized Cox regression coefficient as the weight. In the case that the markers were from the patient subgroups, normalization to a target score of 100 was carried out to create the final score for each patient. Patients whose scores were equal to or greater than 100 were classified in the high risk of relapse group, whereas patients whose scores were less than 100 were predicted as the low risk of relapse group. The gene signature and the cutoff were validated in the testing set. Kaplan-Meier survival plots and log-rank tests were used to assess the differences in time to recurrence of the predicted high- and low-risk groups. All the statistical analyses were performed using S-Plus 6 software.
A and B are constants wi is the standardized Cox regression coefficient xi is the expression value in log2 scale
Patient and Tumor Characteristics Clinical and pathologic features of the patients and their tumors are listed in Table 1. All patients had information on age, sex, tumor-node-metastasis system stage, grade, tumor size, and tumor location. Seventy-three of the 74 patients had data on the number of lymph nodes that were examined, and 72 of the 74 patients had estimated tumor size information. The patient and tumor characteristics did not differ significantly between the relapse and nonrelapse patients. None of the patients received pre- or postoperative treatment. A minimum of 3 years of follow-up data was available for all the patients in the study.
Patient Subgroups Identified by Genetic Profiles Unsupervised hierarchical clustering analysis allowed us to cluster the 74 patients on the basis of the similarities of their expression profiles measured over 17,000 informative genes (Fig 2). In the dendrograms, the length and the subdivision of the branches display the relatedness of the colon tumors based on the expression of the genes. In addition, distinct patterns of genes were found in the clustering result. Examination of this result led us to identify two patient subgroups that have more than 600 differentially expressed genes between them (P < .00001). Cadherin 17 represented the top gene in the list and thus was selected as the indicator for patient subgroups. The larger subgroup, S1, and the smaller subgroup, S2, contained 54 and 20 patients, respectively (Fig 2). Interestingly, in the larger subgroup of the 54 patients, only 18 patients (33%) developed tumor relapse within 3 years, whereas in the smaller subgroup, 13 (65%) of the 20 patients had progressive disease (  2 P = .028). Thus using unsupervised clustering we can already distinguish, to some extent, between good-prognosis and poor-prognosis tumors.
To gain insight into the functions of genes differentially expressed between the two patient subgroups, we examined the dominant gene clusters from QT clustering analysis that had drastic differential expression between the two subgroups (Fig 2). The first gene cluster has a group of 39 downregulated genes in the smaller subgroup, represented by liver-intestine specific cadherin 17, fatty acid binding protein 1, caudal type homeo box transcription factors CDX1 and CDX2, mucin, and cadherin-like protein MUCDHL. The second gene cluster is represented by a group of 14 upregulated genes in the smaller patient subgroup, including serum-inducible kinase SNK, annexin A1, B-cell RAG-associated protein, calbindin 2, and tumor antigen L6. This result suggests that the smaller subgroup of the 20 patients represent less-differentiated tumors on the basis of their genetic profiles, which is associated with poor prognosis. Interestingly, poor differentiation was not evident when morphology of these tumors was examined, suggesting that molecular markers might be used to compensate tumor grading.
Gene Signature and Its Prognostic Value
To compare the 60-gene predictor and the 23-gene predictor, we produced the receiver operating characteristic curves for each, using the 36 independent patients in the testing set (Fig 3). The parameter that was used to assess the performance of a predictor was the area under the curve (AUC). The 60-gene predictor gave an AUC value of 0.50, whereas the 23-gene predictor generated an AUC of 0.74 (Fig 3). This result suggested that the 60-gene signature derived by the first approach of marker selection could not be validated, and the 23-gene signature derived from the second approach had a predictive power for patient outcome.
To summarize the validation result of the 23-gene prognostic signature, the 36 patients in the testing set included 27 patients from subgroup 1 and nine patients from subgroup 2. Furthermore, it consisted of 18 patients who developed tumor relapses within 3 years and 18 patients who remained disease-free for more than 3 years. The prediction resulted in 13 correct relapse classifications and 15 correct disease-free classifications. The overall performance accuracy was 78% (28 of 36 patients), with a sensitivity of 72% (13 of 18 patients) and a specificity of 83% (15 of 18 patients). This performance would indicate that the Dukes' B patients who have the relapse hazard score above the threshold of the prognostic signature have a 13-fold odds ratio (95% CI, 2.6 to 65; P = .003) to develop tumor relapse within 3 years compared with those who have the relapse hazard score below the threshold of the prognostic signature. Furthermore, the Kaplan-Meier analysis produced survival curves for the patient groups, and the log-rank test showed a significant difference in the time to recurrence between the group predicted with good prognosis and the group predicted with poor prognosis (P <. 0001; Fig 4.). In the multivariate Cox proportional hazards regression model, the estimated relative risk for tumor recurrence was 0.17 (95% CI, 0.06 to 0.51; P = .001), indicating that the 23-gene set represents a prognostic signature that is inversely associated with a higher risk of tumor recurrence (Table 3).
The functional annotation for these genes provides insight into the underlying biologic mechanism that leads to rapid metastases. Several genes are related to cell proliferation or tumor progression. For example, tyrosine 3 mono-oxygenase tryptophan 5-monooxygenase activation protein (YWHAH) belongs to 14-3-3 family of proteins that is responsible for G2 cell cycle control in response to DNA damage in human cells.38 RCC1 is another cell cycle gene involved in the regulation of onset of chromosome condensation.39 BTEB2 is a zinc finger transcription factor that has been implicated as a beta-catenin-independent Wnt-1–responsive gene.40 A few genes are involved in local immune responses. Immunoglobulin-like transcript 5 protein is a common inhibitory receptor for major histocompatibility complex I molecules.41,42 A unique member of the gelsolin/villin family capping protein, CAPG, is primarily expressed in macrophages.43 LAT is a highly tyrosine phosphorylated protein that links T-cell receptor to cellular activation.44 The results suggest that both tumor cell- and immune cell-expressed genes can be used as prognostic factors for tumor recurrence.
We report the successful prediction of outcome in Dukes' B colon cancer patients using a 23-gene signature derived from microarray gene expression data and classification methods. As expected, genes implicated in recurrence prediction do not show a striking functional clustering but include multiple features previously associated with colon cancer progression. The genes in the signature include those that regulate cell proliferation, cell signaling, and immune responses. Our study strongly indicates that colon cancer prognosis can be derived from gene expression profile of the primary tumor. The results also demonstrate the potential of DNA microarray-based recognition of gene expression patterns for the prediction of patient outcome in colon cancer. This is likely to have an impact on the current clinical practice for the eligibility of adjuvant chemotherapy on treatment of Dukes' B colon cancer patients. In the sample group of this study, a hierarchical clustering analysis of the gene expression profiles identified a subgroup of 20 tumors that consisted of disproportional relapse patients, predicting an unfavorable outcome for this subgroup. The percentage of the relapse patients in this group is two-fold higher compared with the rest of the patients. Interestingly, in these patients, we also detected that most of the samples represent less-differentiated tumors. Tumors in this cluster therefore seem to share biologic properties that allow cell de-differentiation and vigorous proliferative ability of the tumor. Although the outcome of this subgroup is poor, it is not clear whether this is due to an incorrect classification by pathologic examination for the studied samples or whether it represents a typical make-up of the Dukes' B tumors. We are currently analyzing independent sample groups that have similar follow-up data to define this subgroup more precisely. Genomic data will add substantial value to the clinical information, especially in a disease such as colon cancer, in which multiple, interacting biologic and environmental events define physiologic states and individual factors provide only limited information. Previous attempts to correlate selected characteristics of primary tumors with recurrence have proven unsuccessful in colon cancer. The ability to use systematic profiles of gene expression in predicting outcome could provide important information on both previously known and unknown biologic attributes in tumor characteristics. As an initial example, our study here focuses on Dukes' B tumors in which the analysis defines the 23-gene signature for prediction of recurrence. We have used both unsupervised and supervised classification methods in this analysis. Our selection of multivariate patterns in gene expression data from the primary tumors and examination of the value of such patterns in prediction of independent testing samples resulted in a predictive accuracy of 78% in the independent samples. This represented an odds ratio of 13 (95% CI, 2.6 to 65; P = .003) that is much higher than the characterized prognostic factors in colon cancer, such as lymph node status. The validation of the 23-gene signature in the testing samples illustrates the important role of statistical methods in genomics-based biomedical research. Statistical analysis helps us not only define such patterns and their relevance to individual patient cases, but also determines the exact predictive and prognostic values of these markers with appropriate validation analysis. Improved predictions of disease courses such as recurrence will profoundly affect clinical decisions. Results of several studies show that certain lymph node-negative tumors behave like node-positive tumors in colon cancer.45–49 However, following the current clinical guidelines, few of these lymph node-negative patients are offered adjuvant treatment for colon cancer. Because 25% to 30% of the patients would develop tumor relapse, the prognosis signature would provide a powerful tool to select the patients who are at high risk and ensure that they receive adjuvant treatment. This ability to identify the patients who need intensive clinical intervention could lead to an improvement in cancer survival. Our study is the first example of using a genomics approach to systematically search for molecular markers for colon cancer prognosis. Additional studies will be required to define the full utility of the markers and their independence from pathologic staging. If validated, the recommendation of adjuvant chemotherapy in resected colon cancer may be guided in the future by this prognostic marker for Dukes' B disease. The markers can also be tested with resected tumors other than Dukes' B or with biopsy samples to potentially impact both adjuvant and neoadjuvant therapy. Finally, genes that are overexpressed in tumors with a poor-prognosis profile are potential targets for the rational development of new cancer drugs. Identification of such targets may improve the efficiency of developing therapeutics for colon and other tumor types as well as better understanding of the biology that is related to metastasis and tumor growth.
The authors indicated no potential conflicts of interest.
We thank Shyunshyun Lee and Jack Yu for technical assistance and Mitch Raponi and Steve Rosen for critical reading of the manuscript. We also thank Mark Watson, MD, PhD, director of the Siteman Cancer Center Tissue Procurement Core.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Greenlee RT, Hill-Harmon MB, Murray T, et al: Cancer Statistics, 2001. CA Cancer J Clin 51:15–36, 2001 2. Compton CC, Fielding LP, Burgart LJ, et al: Prognostic factors in colorectal cancer: College of American Pathologists Consensus Statement 1999. Arch Pathol Lab Med 124:979–994, 2000[Medline] 3. Steinberg SM, Barwick KW, Stablein DM: Importance of tumor pathology and morphology in patients with surgically resected colon cancer: Findings from the Gastrointestinal Tumor Study Group. Cancer 58:1340–1345, 1986[CrossRef][Medline] 4. Wolters U, Stutzer H, Keller HW, et al: Colorectal cancer: A multivariate analysis of prognostic factors. Eur J Surg Oncol 22:592–597, 1996[CrossRef][Medline] 5. Guerra A, Borda F, Javier Jimenez F, et al: Multivariate analysis of prognostic factors in resected colorectal cancer: A new prognostic index. Eur J Gastroenterol Hepatol 10:51–58, 1998[Medline] 6. Saltz LB, Kelsen DP: Adjuvant treatment of colorectal cancer. Annu Rev Med 48:191–202, 1997[CrossRef][Medline]
7. Moertel CG, Fleming TR, Macdonald JS, et al: Fluorouracil plus levamisole as effective adjuvant therapy after resection of stage III colon carcinoma: A final report. Ann Intern Med 122:321–326, 1995
8. Wolmark N, Rockette H, Mamounas E, et al: Clinical trial to assess the relative efficacy of fluorouracil and leucovorin, fluorouracil and levamisole, and fluorouracil, leucovorin, and levamisole in patients with Dukes' B and C carcinoma of the colon: Results from National Surgical Adjuvant Breast and Bowel Project C-04. J Clin Oncol 17:3553–3559, 1999
9. International Multicenter 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
10. Mamounas E, Wieand S, Wolmark N, et al: Comparative efficacy of adjuvant chemotherapy in patients with Dukes' B versus Dukes' C colon cancer: Results from four National Surgical Adjuvant Breast and Bowel Project adjuvant studies (C-01, C-02, C-03, and C-04). J Clin Oncol 17:1349–1355, 1999 11. Jass JR: Towards a molecular classification of colorectal cancer. Int J Colorectal Dis 14:194–200, 1999[CrossRef][Medline] 12. Gennari L, Doci R, Rossetti C: Prognostic factors in colorectal cancer. Hepatogastroenterology 47:310–314, 2000[Medline] 13. Ratto C, Sofo L, Ippoliti M, et al: Prognostic factors in colorectal cancer: Literature review for clinical application. Dis Colon Rectum 41:1033–1049, 1998[CrossRef][Medline]
14. Watanabe T, Wu TT, Catalano PJ, et al: Molecular predictors of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 344:1196–1206, 2001 15. Barratt PL, Seymour MT, Stenning SP, et al: UKCCCR AXIS trial collaborators: Adjuvant X-ray and Fluorouracil Infusion Study—DNA markers predicting benefit from adjuvant fluorouracil in patients with colon cancer: A molecular study. Lancet 360:1381–1391, 2002[CrossRef][Medline]
16. Noura S, Yamamoto H, Ohnishi T, et al: Comparative detection of lymph node micrometastases of stage II colorectal cancer by reverse transcriptase polymerase chain reaction and immunohistochemistry. J Clin Oncol 20:4232–4241, 2002 17. Hawkins N, Norrie M, Cheong K, et al: CpG island methylation in sporadic colorectal cancers and its relationship to microsatellite instability. Gastroenterology 122:1376–1387, 2002[CrossRef][Medline]
18. Halling KC, French AJ, McDonnell SK, et al: Microsatellite instability and 8p allelic imbalance in stage B2 and C colorectal cancers. J Natl Cancer Inst 91:1295–1303, 1999 19. Wang C, van Rijnsoever M, Grieu F, et al: Prognostic significance of microsatellite instability and Ki-ras mutation type in stage II colorectal cancer. Oncology 64:259–265, 2003[CrossRef][Medline]
20. Gryfe R, Kim H, Hsieh ET, et al: Tumor microsatellite instability and clinical outcome in young patients with colorectal cancer. N Engl J Med 342:69–77, 2000 21. Witzig TE, Loprinzi CL, Gonchoroff NJ, et al: DNA ploidy and cell kinetic measurements as predictors of recurrence and survival in stages B2 and C colorectal adenocarcinoma. Cancer 68:879–888, 1991[CrossRef][Medline] 22. Martinez-Lopez E, Abad A, Font A, et al: Allelic loss on chromosome 18q as a prognostic marker in stage II colorectal cancer. Gastroenterology 114:1180–1187, 1998[CrossRef][Medline] 23. Zhou W, Goodman SN, Galizia G, et al: Counting alleles to predict recurrence of early-stage colorectal cancers. Lancet 359:219–225, 2002[CrossRef][Medline] 24. Chapman MA, Hardcastle JD, Armitage NC: Five-year prospective study of DNA tumor ploidy and colorectal cancer survival. Cancer 76:383–387, 1995[CrossRef][Medline] 25. Zhou W, Galizia G, Lieto E, et al: Counting alleles reveals a connection between chromosome 18q loss and vascular invasion. Nat Biotechnol 19:78–81, 2001[CrossRef][Medline] 26. Ogunbiyi OA, Goodfellow PJ, Herfarth K, et al: Confirmation that chromosome 18q allelic loss in colon cancer is a prognostic indicator. J Clin Oncol 16:427–433, 1998[Abstract]
27. Shibata D, Reale MA, Lavin P, et al: The DCC protein and prognosis in colorectal cancer. N Engl J Med 335:1727–1732, 1996
28. Sun XF, Rutten S, Zhang H, et al: Expression of the deleted in colorectal cancer gene is related to prognosis in DNA diploid and low proliferative colorectal adenocarcinoma. J Clin Oncol 17:1745–1750, 1999 29. McLeod HL, Murray GI: Tumor markers of prognosis in colorectal cancer. Br J Cancer 79:191–203, 1999[Medline] 30. Frederiksen CM, Knudsen S, Laurberg S, et al: Classification of Dukes' B and C colorectal cancers using expression arrays. J Cancer Res Clin Oncol 129:263–271, 2003[Medline] 31. Van't Veer LJ, Dai H, van de Vijver MJ, et al: Gene expression profiling predicts clinical outcome of breast cancer. Nature 415:530–536, 2002[CrossRef][Medline]
32. Van de Vijver MJ, Yudong DH, Van't Veer LJ, et al: A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 347:1999–2009, 2002 33. Beer DG, Kardia SLR, Huang C, et al: Gene expression profiles predict survival of patients with lung adenocarcinoma. Nat Med 8:816–824, 2002[Medline]
34. Bhattacharjee A, Richards WG, Staunton J, et al: Classification of human lung carcinoma by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc Natl Acad Sci U S A 98:13790–13795, 2001 35. Shipp MA, Ross KN, Tamayo P, et al: Diffuse large B-cell lymphoma outcome prediction by gene-expression profiling and supervised machine learning. Nat Med 8:68–74, 2002[CrossRef][Medline]
36. Rosenwald A, Wright G, Chan WC, et al: The use of molecular profiling to predict survival after chemotherapy for diffuse larger B-cell lymphoma. N Engl J Med 346:1937–1947, 2002 37. Lipshutz RJ, Fodor SP, Gingeras TR, et al: High density synthetic oligonucleotide arrays. Nat Genet 21:20–24, 1999[CrossRef][Medline] 38. Yu T, Robb VA, Singh V, et al: The 4.1/ezrin/radixin/moesin domain of the DAL-1 protein 4.1B tumor suppressor interacts with 14-3-3 proteins. Biochem J 365:783–789, 2002[CrossRef][Medline]
39. Ohtsubo M, Kai R, Furuno N, et al: Isolation of the active cDNA of the human cell cycle gene involved in the regulation of onset of chromosome condensation. Genes Dev 1:585–593, 1987
40. Ziemer LT, Pennica D, Levine AJ, et al: Identification of a mouse homolog of the human BTEB2 transcription factor as a beta-catenin-independent Wnt-1-responsive gene. Mol Cell Biol 21:562–574, 2001 41. Hakansson L, Adell G, Boeryd B, et al: Infiltration of mononuclear inflammatory cells into primary colorectal carcinomas: An immunohistological analysis. Br J Cancer 75:374–380, 1997[Medline]
42. Moller P, Momburg F, Koretz K, et al: Influence of major histocompatibility complex class I and II antigens on survival in colorectal carcinoma. Cancer Res 51:729–736, 1991
43. Dabiri GA, Young CL, Rosenbloom J, et al: Molecular cloning of human macrophage capping protein cDNA. J Biol Chem 267:16545–16552, 1992 44. Zhang W, Sloan-Lancaster J, Kitchen J, et al: Influence of major histocompatibility complex class I and II antigens on survival in colorectal carcinoma. Cell 92:83–92, 1998[CrossRef][Medline]
45. Liefers GJ, Cleton-Jansen AM, van de Velde CJ, et al: Micrometastases and survival in stage II colorectal cancer. N Engl J Med 339:223–228, 1998
46. Feezor RJ, Copeland EM III, Hochwald SN: Significance of micrometastases in colorectal cancer. Ann Surg Oncol 9:944–953, 2002 47. Steinberg SM, Barkin JS, Kaplan RS, et al: Prognostic indicators of colon tumors: The Gastrointestinal Tumor Study Group experience. Cancer 57:1866–1870, 1986[CrossRef][Medline] 48. Markowitz SD, Dawson DM, Willis J, et al: Focus on colon cancer. Cancer Cell 1:233–236, 2002[CrossRef][Medline] 49. Caplin S, Cerottini JP, Bosman FT, et al: For patients with Dukes' B (TNM stage II) colorectal carcinoma, examination of six or fewer lymph nodes is related to poor prognosis. Cancer 83:666–672, 1998[CrossRef][Medline] Submitted August 28, 2003; accepted December 16, 2003.
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
|
TOP
ABSTRACT
INTRODUCTION
