[Recent developments of pharmacogenomics in the treatment of colorectal cancers]

Colorectal cancer (CCR), which is one of the most common causes of cancer, has benefited from the major advances in the understanding of the intracellular signaling pathways implicated in the initiation, growing and local and metastasis dissemination of tumor, which have occurred during the 20 past years. The pharmacogenomics approach, especially the determination of the genetic polymorphisms, tries to find prognosis and predictive biomarkers permitting to identify patients who could benefit from a particular treatment or those exhibiting higher risks of toxicity. Among the numerous biomarkers, which have been studied, few are currently in use in clinical practice. The phenotyping of DPD and UGT1A1 activities, and to a lesser extent, its genotyping, appears as the most useful tool in terms of prediction of toxicities induced by two major drugs: 5-FU and irinotecan. For oxaliplatin, the determination of the polymorphisms of reparases and detoxification systems such as GSTpi seems interesting, but its exact place should be more defined. It is in the field of targeted therapies that the pharmacogenomics approach seems to be the more relevant. KRAS mutation is a dramatic example of single nucleotide polymorphism, which is able to identify a priori patients that could receive or not an anti-EGFR monoclonal antibody such as cetuximab or panitumumab. It is obvious that pre-clinical identification of molecular biomarkers predictive of the sensitivity of the drug targets, which subsequently implicate the selection of patients and the rational evaluation of responses, will be the cornerstone of any clinical trials concerning targeted therapies. Besides the determination of drug target polymorphisms, it is also important to consider those related to the distribution and metabolism. In this area, the determination of enzymatic activities should recover its place besides the genomic profiling.