Five-year annual incidence and clinico-molecular features of breast cancer after the last negative screening mammography at age 68-69.

The European Commission Initiative on Breast Cancer recommendation for triennial screening of women aged 70–74 is based on very weak evidence. A cohort of Italian women who had their last biennial screening mammography at age 68–69 was followed up for 5 years, assumed to represent the interval to another hypothetical screening mammography, in order to determine the annual proportional incidence of interval breast cancer. The cohort included 118,370 women. They had their last mammography between 1997 and 2008. Incident breast cancers were identified by record-linking the cohort with the regional breast cancer registry. The expected incidence in the age range 65–74 was estimated with an age-period-cohort model. The number of interval cancers was divided by the expected number to obtain their proportional incidence. Overall, there were 298,658 woman-years at risk with 371 interval cancers versus 988.8 expected. In the first, second, third, fourth, and fifth interval year, the proportional incidence was 0.09 (95% confidence interval, 0.06–0.13), 0.32 (0.25–0.39), 0.60 (0.49–0.73), 0.75 (0.60–0.92), and 0.81 (0.60–1.07), respectively. Between the second and the fifth year, tumour stage and molecular subtype did not change significantly. Though not supported by these findings, the proposal of triennial screening for women aged 70–74 merits further research, because the 95% confidence interval of the third-year proportional incidence of interval cancer included 0.50—the maximum limit considered acceptable for women aged 50–69. • The third-year incidence of breast cancer relative to the expected one was 0.60 (95% confidence interval, 0.49–0.73). • Between the second and the fifth year, tumour stage and molecular subtype did not change significantly (p >0.10). • The proposal of a 3-year screening interval at age 70–74 merits further evaluation.