11 July 2006
A new paper by members of the Public Health Genetics Unit in Cambridge provides a meta‑analysis of the relative risks of developing colorectal cancer among individuals with a family history of the disease, produced by a systematic review of published literature on familial colorectal cancer risk [Butterworth et al. (2006) Eur J Cancer 42(2), 216-27].
As noted in the previous journal club piece, identification of individuals at high risk of colorectal cancer is important for offering appropriate surveillance measures and preventative interventions, to reduce the risk of developing serious disease. The first stage in locating people with inherited predisposition to colorectal cancer is by their family history, and Butterworth and colleagues found a substantial body of epidemiological data on the links between family history and disease risk in colorectal cancer. Using this information, they calculated pooled estimates of relative risk for different categories of family history, and converted these figures into measures of absolute, lifetime risk using life tables; a commentary in The Lancet proposes that absolute risk is a more easily comprehensible and hence useful concept for most people [Hakama M (2006) Lancet 368(9530), 101-103].
Risk estimates were affected by the age of the individual in question, but in brief, they found that the lifetime risk of colorectal cancer for a 50 year old with at least one affected first-degree relative was 3.4%, rising to 6.9% for those with at least two affected first-degree relatives, compared with a population level lifetime risk of 1.8%. That is, risk almost doubles if a close relative has had colorectal cancer and doubles again if at least two relatives have been affected. The authors note that although this study did not attempt to assess the relative contributions of genetic and shared environmental factors in contributing to familial risk, previous evidence has suggested that as yet unidentified hereditary factors are likely to play a significant role.
Comment: A well-constructed and performed meta-analysis such as that of Butterworth et al. provides a useful estimate of colorectal cancer risk based on family history that is much more reliable and robust than estimates produced by individual studies. This information, which includes risk curves showing an individual