RESUMEN
Colorectal cancer staging describes disease spread by assessing tumour penetration of the bowel wall and the presence of lymph node involvement and distant metastases. Although useful for stratifying patients into risk groups that indicate prognosis as well as treatment, conventional staging frequently fails to predict the biological behaviour of individual tumours within the same stage. Accordingly, up to a third of patients undergoing apparently curative surgery develop systemic disease and die from cancer-related causes. This heterogeneity of clinical outcome is mirrored by similarly extensive molecular diversity, where progression to systemic disease is associated with a range of genetic and epigenetic alterations. This has led to the idea that tumour stratification based on molecular taxonomy might increase the prognostic accuracy for the individual patient. However, despite the description of many putative individual and multiple molecular biomarkers, the data generated by molecular diagnostic assays have not proved to be either robust or consistently reproducible. Experimental bias exists throughout the entire process, starting with sample collection and ending with data interpretation. There is an urgent need to integrate clinical, biochemical, genetic and molecular data, and findings must be validated in prospective, well-controlled clinical studies of large numbers of diverse patients across multiple institutions. Only then can the implementation of molecular staging assays into clinical practice be considered.