Page 60 - 20dynamics of cancer
P. 60
MULTISTAGE PROGRESSION 45
LOH. In the absence of LOH, two separate mutations to APC are needed,
whereas only one mutation to β-catenin is needed. This may explain
why there is a rise in the ratio of β-catenin to APC initiating mutations
in HNPCC.
The morphological sequence in HNPCC follows the classical pathway.
In the classical pathway, the adenoma to carcinoma ratio is about 30:1.
By contrast, HNPCC patients have an adenoma to carcinoma ratio of
about 1:1 (Jass et al. 2002b). This suggests much faster progression from
adenoma to carcinoma in HNPCC, probably driven by the high somatic
mutation rate in MSI cells.
The spectrum of later mutations in HNPCC differs from later muta-
tions in the classical pathway (Jass et al. 2002b). HNPCC tumors have
less LOH. The K-RAS mutation frequency is about the same, but HNPCC
may have fewer p53 mutations, and more mutations in various growth-
related genes with repetitive sequences, including TGFβ-RII, IGF-II, and
BAX.
In another study, Rajagopalan et al. (2002) found that 61 percent of
330 colorectal tumors had either a BRAF or K-RAS mutation, but a tu-
mor never had mutations in both genes. Mutually exclusive mutation
of these genes supports the suggestion that they have similar effects
in tumorigenesis (Storm and Rapp 1993). The ratio of BRAF to K-RAS
mutations was significantly higher in MMR deficient cancers compared
to MMR proficient cancers. This difference in mutation frequency again
supports the idea that particular aberrations in DNA repair affect the
mutation spectrum of tumors, although the functional changes caused
by different mutations may sometimes be similar.
HYPERMETHYLATION
Some colorectal cancers accumulate changes in gene expression by
hypermethylation of promoter regions, which can suppress transcrip-
tion. Commonly hypermethylated genes in colorectal cancers include
p14, p16, hMLH1, MGMPT, and HPP1 (Jass et al. 2002a; Issa 2004).
Jass et al. (2002b) proposed multiple pathways to cancer via hyper-
methylation, accounting for up to 40 percent of all colorectal cancers.
These arguments are, at present, based on limited sample sizes. But the
existing data do hint at interesting hypotheses about alternative path-
ways.