Comparison of breast cancer mortality rates with those of ovarian and colorectal carcinoma

Abstract

Traditional survival curves cannot easily be used to predict outcome for an individual patient on a year-to-year basis. This difficulty is partly overcome by yearly mortality analysis. This method was employed to analyse long-term follow-up of three cancers: colorectal, ovarian and breast cancer. The study used prospectively collected cancer registry data from geographically defined regions in Scotland. Cohort sizes were 7196 patients with breast cancer, 3200 with colorectal cancer and 1866 with ovarian cancer. Follow-up extended to 23 years. Two distinct patterns of mortality emerged. Mortality rates for ovarian and colorectal cancer were initially high (41 and 21 per cent) but decreased rapidly; by 10 years patients had either died or were cured. The influence of stage diminished with follow-up. Breast cancer mortality was lower than that of colorectal or ovarian cancer, but remained raised in comparison to the general population throughout follow-up. The influence of breast cancer size reduced with follow-up, whereas that of nodal status persisted. Patients with breast cancer live at increased risk of death to the end of follow-up, supporting the concept of dormancy in breast cancer biology. This was not observed with colorectal or ovarian cancer.