Gene expression profiling of colorectal cancer and metastases divides tumours according to their clinicopathological stage.

Gene expression profiling of matched colorectal carcinomas and metastases could reveal key molecular events involved in tumour progression and metastasis. Expression profiles have been created from 25 colorectal carcinomas (CRCs, pT1-4), corresponding normal colonic mucosa, and 14 liver metastases using cDNA arrays containing 1176 cancer-related genes (Clontech). Hierarchical clustering clearly distinguished carcinomas from non-cancerous tissues, separated tumours into high-stage (pT4 and extensive lymph node or distant metastases) and low-stage (< or =pT3) groups, and correlated with the histopathological classification in 87% (33/38 cases). Most primary tumours and matched liver metastases clustered on terminal branches of the dendrogram. Statistical analysis (Mann-Whitney U-test) revealed 40 tumour-specific genes (29 up-regulated, 11 down-regulated) which allowed identification of malignant tissue samples by cluster analysis. A specific expression signature in matching metastases was not found, but a set of 23 classifier genes with statistically significant expression patterns in high- and low-stage tumours was identified. These genes may represent important targets in colorectal carcinogenesis and might provide useful clinicopathological tools in the management of colorectal cancer.