An investigation of the apparent breast cancer epidemic in France: screening and incidence trends in birth cohorts.

Background: Official descriptive data from France showed a strong increase in breast-cancer incidence between 1980 to 2005 without a corresponding change in breast-cancer mortality. This study quantifies the part of incidence increase due to secular changes in risk factor exposure and in overdiagnosis due to organised or opportunistic screening. Overdiagnosis was defined as non progressive tumours diagnosed as cancer at histology or progressive cancer that would remain asymptomatic until time of death for another cause.

Methods: Comparison between age-matched cohorts from 1980 to 2005. All women residing in France and born 1911-1915, 1926-1930 and 1941-1945 are included. Sources are official data sets and published French reports on screening by mammography, age and time specific breast-cancer incidence and mortality, hormone replacement therapy, alcohol and obesity. Outcome measures include breast-cancer incidence differences adjusted for changes in risk factor distributions between pairs of age-matched cohorts who had experienced different levels of screening intensity.

Results: There was an 8-fold increase in the number of mammography machines operating in France between 1980 and 2000. Opportunistic and organised screening increased over time. In comparison to age-matched cohorts born 15 years earlier, recent cohorts had adjusted incidence proportion over 11 years that were 76% higher [95% confidence limits (CL) 67%, 85%] for women aged 50 to 64 years and 23% higher [95% CL 15%, 31%] for women aged 65 to 79 years. Given that mortality did not change correspondingly, this increase in adjusted 11 year incidence proportion was considered as an estimate of overdiagnosis.

Conclusions: Breast cancer may be overdiagnosed because screening increases diagnosis of slowly progressing non-life threatening cancer and increases misdiagnosis among women without progressive cancer. We suggest that these effects could largely explain the reported "epidemic" of breast cancer in France. Better predictive classification of tumours is needed in order to avoid unnecessary cancer diagnoses and subsequent procedures.