A prediction model for metachronous colorectal cancer: development and validation.

Background: Being able to estimate a colorectal cancer case's risk of metachronous colorectal cancer could enable risk-appropriate surveillance. The aim was to develop a risk prediction model to estimate individual 10-year risk of metachronous colorectal cancer following a colorectal cancer diagnosis.

Methods: A cohort of population-based colorectal cancer cases were recruited soon after their diagnosis between 1997 and 2012 from America, Canada, and Australia. Cox regression with the least absolute shrinkage and selection operator penalization was used to identify factors that predicted the risk of a new primary colorectal cancer diagnosed at least one year after the initial colorectal cancer. Potential predictors included demography, anthropometry, lifestyle factors, comorbidities, personal and family cancer history, medication use, and diagnosis age and pathological features of the first colorectal cancer. Internal validation through bootstrapping was used to evaluate the discrimination and calibration.

Results: 6,085 colorectal cancer cases were included. 138 (2.3%) were diagnosed with metachronous colorectal cancer over a median of 12 years (interquartile range 5 - 17 years). Metachronous colorectal cancer risk was predicted by body mass index, smoking, physical activity, family history of cancer and synchronous colorectal cancer, stage, grade, histological type and DNA mismatch repair status and diagnosis age of the first colorectal cancer. The model was valid with a c-statistic of 0.65 (95% CI: 0.63 - 0.68) and a calibration slope of 0.873 (standard deviation: 0.087).

Conclusions: Metachronous colorectal cancer can be predicted with reasonable accuracy by this prediction model that consists of clinical variables collected as part of routine practice.