The aim was to compare mortality in a consecutive series of women with
breast cancer to that of an age-matched cohort of the Scottish popula
tion. An analysis of outcome has been performed in a retrospective ser
ies of 3933 women with breast cancer diagnosed between 1954 and 1964 a
nd followed for between 30 and 40 years (median 32 years). The ratio o
f observed to expected mortality (O:E) was calculated by quinquennia f
or the group as a whole and for subsets defined by clinical prognostic
criteria. O:E mortality remained significantly greater than unity for
at least 25 years following diagnosis and treatment, indicating a fai
lure to demonstrate cure of the disease in a statistical sense, i.e. b
y the elimination of the excess hazard of death in the study group. Cl
inical prognostic factors, such as tumour size, palpability of axillar
y nodes and their combination in clinical stages, significantly discri
minate for survival for the first 10 years only. These data confirm ot
her long-term follow-up studies of women with breast cancer. For pract
ical purposes, the disease should be considered to be incurable. The i
mplications of this concept for the development of clinical research a
re discussed briefly.