The identification of people at increased genetic risk of colorectal cancer and the
provision of appropriate clinical screening represents one approach to the prevention of
colorectal cancer in the Scottish population. This thesis aims to contribute to current
knowledge regarding the available tools for identifying those at increased genetic risk
in a population, namely genetic testing and family history assessment.
Key issues relating to the use of family history in this context were addressed through
the analysis of a unique data set, comprising family history information reported by a
colorectal cancer case or control subject at interview and the results of record linkage
of this data to the Scottish Cancer Registry. Retrospective family history case-control
analysis showed that individuals with an affected first-degree relative were at an
increased risk of developing colorectal cancer (ORimh 2.14, 95% CI = 1.11, 4.14).
Prevalence of such a family history in control subjects was 9.4% (95% CI = 4.9, 13.9).
Substantial under-reporting of family history was evident, with sensitivity of interview
as a means of determining a history of colorectal cancer in a first-degree relative being
approximately 0.55 for both cases and controls. These studies illustrate the potential
advantages of targeting people with a family history, but also highlight some of the
limitations of such an approach.
The genetic epidemiology of the mismatch repair genes hMLH 1 and hMSH2 and their
association with colorectal cancer was considered in a systematic literature review.
Although conventional epidemiological studies are lacking, there is compelling
evidence to implicate mutations in these genes in the aetiology of a sub-set of
colorectal cancers, with penetrance of approximately 80% in males and 40% in
females. A total of 550 different published gene variants were identified, and this high
degree of heterogeneity was illustrated in a unique database. This review indicates that
carriers of mismatch repair gene mutations merit particular consideration in the context
of colorectal cancer prevention through targeting people at increased genetic risk.
Accordingly, the challenge of identifying asymptomatic mismatch repair gene mutation
carriers in Scotland was addressed through the development of a computer model of
cascade genetic testing, a strategy in which a mutation is identified in one family
member and systematically traced through a pedigree. The model predicts that
application of cascade genetic testing to colorectal cancer cases < 55 years of age over
a twenty- year period would involve testing 7142 patients and 849 relatives of known
carriers, and would identify 321.2 (95%CI = 305.3, 337.1) asymptomatic mutation
carriers, representing approximately 27% of the estimated 1209 carriers in Scotland.
Model outcomes were highly sensitive to the prevalence and penetrance of mutations,
and the participation rates of those offered testing. Overall, outcomes from this
computer model suggest that cascade genetic testing is potentially a useful means of
identifying asymptomatic mismatch repair gene mutation carriers in Scotland. Followup
work should ensure that it is also of practical importance as a tool for planning
research and health policy.
Identification and screening of mismatch repair gene mutation carriers is an important
approach to colorectal cancer prevention, but is only relevant to a minority of people at
increased genetic risk. Hence, despite inherent limitations, family history remains a
crucial tool for genetic risk assessment in a population. An integrated approach to the
prevention of colorectal cancer through targeting people at increased genetic risk can
potentially provide substantial health benefits to a sub-group of the population, and thus
contribute to the overall prevention of colorectal cancer in Scotland.