PATHOLOGICAL FEATURES FOR COLORECTAL CANCERS
Colorectal cancers with MSI-high features characteristically have increased mucin secretion, differentiation, and lymphocytic infiltrate. This is seen both in sporadic Tumours and those arising in Lynch syndrome. There is now an evidence that sporadic MSI-H cancers develop in adenomas with a somatic BRAF mutation and DNA methylation, particularly of MLH1, while those in Lynch syndrome arise in adenomas with somatic mutations in APC, beta-catenin, and/or KRAS. This also translates into morphological differences, Lynch syndrome-associated cancers showing more Tumours budding (dedifferentiation), and sporadic Tumours being more heterogeneous and displaying mucin secretion (Jass 2004; McGivern et al. 2004).
Molecular genetic studies of colon cancer have provided strong evidence for a multistep pathogenesis. This work started out with a simple linear model involving at least five genes (Fearon and Vogelstein 1990). Although this model is too simplistic, it has provided a framework for many other studies.
APC mutations are thought to occur early in the process and may cause spindle aberrations via the interaction of APC with the kinetochore at mitosis through connections with the microtubules, leading to chromosome abnormalities (Powell 2002). Originally, it was thought that the critical factor is the accumulation of mutations rather than the particular order in which they occur, but more recent work, focusing on the idea of cell-type-specific “gatekeepers,” suggests that there are genes which act as rate-limiting steps:
it is likely that such genes are altered early in the carcinogenetic process. For example, in most CRC, biallelic APC mutations (or mutation plus loss of heterozygosity) occur at an early stage (Powell et al. 1992). These early alterations (such as the loss of sequences C-terminal to the beta-catenin regulatory domain (Sidransky 1997)) could lead to a screening test for early CRC based on the detection of such mutations in stool. A stool test for APC mutations was developed (Traverso et al. 2002) but has not been taken up commercially, and it is not in widespread use. Other assays are based on multiple genes: for example, stool analysis of three gene markers (TP53, BAT26, and KRAS) detected 71 % of CRC patients and 92 % of those whose Tumours actually had an alteration in these genes (Dong et al. 2001). Despite these and other encouraging findings, commercial molecular-based stool tests remain unadopted in the clinical setting.