Recurrent genomic alterations with impact on survival in colorectal cancer identified by genome-wide array comparative genomic hybridization

Gastroenterology. 2006 Dec;131(6):1913-24. doi: 10.1053/j.gastro.2006.10.021. Epub 2006 Oct 15.

Abstract

Background & aims: Although genetic aspects of tumorigenesis in colorectal cancer (CRC) have been well studied, reliable biomarkers predicting prognosis are scarce. We aimed to identify recurrently altered genomic regions (RAR) in CRC with high resolution, to investigate their implications on survival and to explore novel cancer-related genes in prognosis-associated RARs.

Methods: A 1-Mb resolution microarray-based comparative genomic hybridization (array CGH) was applied to 59 CRCs. RARs, defined as genomic alterations, detected in more than 10 cases were identified and analyzed for their association with survival. Expression levels of genes in prognosis-associated RARs were examined by real-time quantitative polymerase chain reaction.

Results: Twenty-seven RARs were identified. Eleven high-level amplifications and 2 homozygous deletions also were detected, but they were not as common as RARs. Multivariate analysis revealed RAR-L1 (loss on 1p36; hazard ratio = 8.15, P = .002) and RAR-L20 (loss on 21q22; hazard ratio = 3.53, P = .034) are independent indicators of poor prognosis. Expression of CAMTA1, located in RAR-L1, was reduced frequently in CRCs, and low CAMTA1 expression was associated significantly with poor prognosis, which indicates that CAMTA1 may play a role as a tumor suppressor in CRC. Five pairs of RARs were correlated significantly to each other and 3 pairs share genes involved in the same biological functions, suggesting possible collaborative roles in tumorigenesis.

Conclusions: We identified recurrent genomic changes in 59 CRCs. RARs could be more important in sporadic tumors where the effect of genomic changes on tumorigenesis is relatively smaller than in familial cancer. Our results and analysis strategy will be helpful to elucidate pathogenesis of CRCs or to develop biomarkers for predicting prognosis.

Publication types

  • Comparative Study
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Biomarkers, Tumor / genetics
  • Colorectal Neoplasms / genetics*
  • Colorectal Neoplasms / mortality*
  • Colorectal Neoplasms / pathology
  • DNA, Neoplasm / genetics*
  • Female
  • Gene Dosage / genetics
  • Gene Expression Profiling
  • Genome, Human / genetics
  • Humans
  • Male
  • Middle Aged
  • Multivariate Analysis
  • Nucleic Acid Hybridization / genetics*
  • Prognosis
  • Survival Analysis

Substances

  • Biomarkers, Tumor
  • DNA, Neoplasm