Author Interviews, Cancer Research, Colon Cancer / 12.06.2015
Women and Men Have Similar Colon Cancer Recurrence-Free Death Risk
MedicalResearch.com Interview with:
Prof. Catherine Quantin
Teaching Hospital, Department of Biostatistics and Medical Informatics France;
Dijon University Hospital, Clinical Investigation Center,
Clinical Epidemiology/Clinical Trials Unit, Dijon, France and
Dr Michal Abrahamowicz Ph.D
Department of Epidemiology, Biostatistics and Occupational Health
McGill University, Montreal, Canada
Medical Research: What is the background for this study?
Response: One difficulty, common to prognostic studies of cancer, concerns the need to separate the effects of prognostic factors on different clinical endpoints, such as disease recurrence vs recurrence-free death. Some published prognostic studies used a Cox regression model that included recurrence as a time-dependent covariate, to assess the impact of recurrence on mortality, and to adjust for recurrence when estimating the effects of other prognostic factors on mortality. However, the Cox model is limited to the assessment of the effects of covariates on a single endpoint, such as death. This limitation is overcome by multi-state models, that make it possible to model alternative pathways of disease progression and to assess the impact of prognostic factors on both recurrence-free death vs death after recurrence, and recurrence followed by death.
Another difficulty, is that the cause of death is not available or not accurately coded. Yet, some patients are likely to die of causes not related to the disease of primary interest, especially in cancers with longer survival and in those that affect older subjects. The effects of prognostic factors estimated with Cox model, or classic multi-state models, are not able to discriminate between their effects on the mortality due to cancer of primary interest vs natural mortality. However, age is a very strong predictor of overall mortality, but is not systematically associated with higher cancer-specific mortality.
To deal with this difficulty, many prognostic studies use relative survival methods.
The general idea is to use the mortality tables for the relevant general population to estimate survival corrected for the expected natural mortality, due to other causes of death. (more…)