Yu Y, Carey M, Pollett W, et al. - Through a patient cohort that involved 738 incident cases diagnosed between 1999 and 2003, the researchers evaluated the long-term survival features and prognostic correlations and potential time-varying impacts of clinico-demographic variables and two molecular markers (microsatellite instability (MSI) and BRAF Val600Glu mutation) in a population-based patient cohort followed up to ~ 19 years. Nearly 61% and 26% of patients died from any cause and from colorectal cancer, and ~ 10% and ~ 20% encountered recurrences and distant metastases, respectively, during the follow-up. Stage IV disease and postdiagnostic recurrence or metastasis were firmly related to the risk of death from colorectal cancer. If a patient had outlived the first 6 years with no disease-related event (i.e., recurrence, metastasis, or death from colorectal cancer), their risks became very minimal following this time period. Discrete sets of markers were correlated with various outcome measures. Throughout the follow-up, the impacts of variables were consistent in some cases. For example, consistently over the follow-up, the MSI-high tumor phenotype and elderly age at diagnosis prognosticated longer metastasis-free survival times. The impacts of the variables differed with time, in some other cases. For example, in people who received this treatment following 5.5 years postdiagnosis, adjuvant radiotherapy treatment was correlated with the risen risk of metastasis, but not before that. In conclusion, the long-term survival features of a prospective cohort of colorectal cancer individuals, associations between baseline variables and a complete set of individual outcomes over a long time, and the time-varying impacts of a group of variables were reported in this study. Moreover, in colorectal cancer, the results exhibited advance our understanding of the long-term prognostic features and were expected to encourage more studies and clinical care approaches.