A genetic study of liver metastasis form primary colorectal cancers and uveal melanomas.

Cancer is a multifactorial disease, with the development of metastases being a major cause of morbidity and mortality. Several molecular pathways are thought to be
involved in the tumourigenesis of colorectal cancer and possibly metastasis. The aim of this research was to explore aspects of these pathways and relate genetic changes to clinicopathological variables and outcome, with the goal of ascertaining novel genetic abnormalities predictive of metastasis. A second cancer, uveal melanoma was also studied as a comparison, as these cancers invariably metastasise to the liver, in which abnormalities of chromosomes 3 and 8 have already been shown to be predictive of liver metastasis and
hence a poor prognosis. Fluorescent in situ hybridisation (FISH) analysis of fresh-frozen uveal melanomas using alpha-centromeric probes for chromosomes 3 and 8 confirmed a significant association between genetic imbalance and the presence of liver metastasis and reduced survival. Difficulties were encountered with using FISH for the analysis of chromosomal abnormalities in formalin-fixed paraffin-embedded samples of colorectal
cancers and their liver metastases. Therefore, comparative genomic hybridisation (CGH) was utilised for the genome-wide analysis for regions of chromosomal amplification and loss. The technique was partially successful; however problems were encountered in
obtaining analysable target metaphase slides, this was partly overcome by the manufacture of target slides in-house. The analysis of microsatellite instability (MSI) using polymerase chain reaction (PCR) was more successful with the presence of MSI being associated with the presence of solitary metastases, but interestingly not with an improved prognosis. Finally, the collection of a single paired fresh-frozen sample of a colorectal cancer and its metastasis enabled all the techniques to be applied, thus showing that prospective collection and analysis is not only feasible but would allow the clinical significance of genetic abnormalities to be assessed more accurately.