341Oncotype DX ® Recurrence Score ® test as well as T-stage and mismatch repair sta-
tus, should be considered during the treatment decision-making process.
Interest is focused on DNA methylation, an epigenetic mechanism that is
involved in everything from imprinting to X-chromosome inactivation, for deter-
mining prognosis of CRC. The results of an analysis of the methylation patterns
using pyrosequencing in CRC samples taken from two independent prospective
cohorts suggest that decreased methylation in regions of the genome called long
interspersed nucleotide element-1 (LINE-1) elements is independently associated
with poor survival outcomes (Ogino et al. 2008 ). A 30 % decrease in LINE-1 meth-
ylation doubled the risk of CRC-specifi c mortality, and the lower the methylation
level, the worse the patient outcomes. Methylation changes associated with mortal-
ity may refl ect genomic instability, transcriptional dysregulation, and the activation
of oncogenes, infl ammation, or oxidative stress. Although follow-up studies are still
needed, there are good prospects of clinical application of the results.
Another study has identifi ed a 50-gene signature in early-stage colon cancer that
predicts cancer recurrence (Garman et al. 2008 ). The investigators compiled gene
expression data from publicly available datasets, assessing the expression patterns
in 52 samples taken from individuals with known survival outcomes. This signature
included RAS and TNF family genes previously implicated in carcinogenesis as
well as genes in several pathways linked to metastasis. The team validated nine of
the top 10 differentially expressed genes using RT-PCR. Along with its prognostic
implications, preliminary results suggest that the signature, which was validated in
two independent patient groups, may also provide clues for treating colon cancer.
Examination of gene expression in early-stage CRC revealed certain patterns that
seem to put some patients at higher risk for recurrence. The signature could detect
recurrence with more than 90 % accuracy regardless of the early colon cancer’s
tumor, node, metastasis, or TNM stage. Identifi cation of these patients may enable
targeted and proactive treatment to prevent this recurrence. The investigators also
tested whether the gene signature was useful for guiding individuals’ treatment and
identifying new drugs. Using the Broad Institute’s Connectivity Map, they assessed
the gene expression profi les of cells treated with a range of drugs to look for profi les
resembling the cancer recurrence signature. Their research suggests that at least
four drugs may infl uence the genes involved in the recurrence signature. Subsequent
experiments indicated that cell lines with the high recurrence risk signature are sen-
sitive to at least two of these compounds: the COX2 inhibitor celecoxib and the
PI3K inhibitor LY-294002. That, in turn, suggests it may be useful to test the treat-
ments in those with the high-risk signature in order to identify patients who may
benefi t from such treatments rather than standard chemotherapy. This will individu-
alize the treatment plans for patients with CRC and improve survival. Clinical trials
are planned to test this.
Identifi cation of genetic factors underlying drug response in CRC still remains a
promising area for improving management of CRC patients. Genetic variations
identifi ed in genes encoding thymidylate synthase, dihydropyrimidine dehydroge-
nase, glutathione S-transferase pi, and uridine diphosphate glucosyltransferase 1A1
seem to be promising predictors of drug effi cacy and/or toxicity in CRC (Fogli and
Personalized Management of Cancers of Various Organs