Abstract
Objective —
To assess the effect of screening for breast cancer in women aged 40–49 in the Swedish two county trial, in terms of mortality reduction, advanced cancer reduction, mode of detection, and the histology of tumours detected.
Setting —
The Swedish two county trial of screening for breast cancer, in which 77080 women aged 40–74 (19 844 aged 40–49) were randomly allocated to receive regular invitation to mammographic screening for breast cancer, and 55985 women aged 40–74 (15 604 aged 40–49) were allocated to an unscreened control group.
Methods —
The screening interval in the younger age group was two years and in the older age group about three years. Statistical analysis of mortality and incidence rates was performed by Poisson regression. Relative survival was estimated using proportional hazards regression.
Results —
The relative mortality for the group invited to screening compared with the control group was 0·87 (95% confidence interval 0·54 to 1·41) in the 40–49 age group, in close agreement with the relative incidence of advanced cancers. For Kopparberg county the relative mortality was 0·73 (95% CI 0·37 to 1·41) and for Ostergotland 1·02 (95% CI 0·52 to 1·99). The lesser effect in the 40–49 group as a whole was largely due to a higher rate of interval cancers in this age group, and the occurrence in Ostergotland of a higher number of cancers after randomisation, but before screening started, and in women who refused screening. The higher interval cancer rate was consistent with the higher proportion of ductal grade 3 and medullary cancers in women aged 40–49 at diagnosis.
Conclusions —
A major difficulty in screening women aged 40–49 is the rapid progression of a subset of tumours arising in this age group. Shortening the screening interval from two years would be necessary to achieve a higher mortality reduction.