Incidence and Risk Factors of Interval and Screen-Detected Breast Cancer

Abstract

ImportanceMammographic screening is the only proven method for early detection and mortality reduction of breast cancer (BC). However, many patients are missed at prior screening; thus, they receive their diagnosis between the interval of screening rounds, called interval cancer (IntCa). Some IntCas are fast growing between screening rounds.ObjectiveTo investigate the incidence and proportion of IntCa and screen-detected breast cancer (ScrCa) and identify factors associated with IntCa.Design, Setting, and ParticipantsThis population-based cohort study was conducted from January 1989 to March 2020, with follow-up until 2020 and a mean (SD) follow-up of 13 (8.3) years. The statistical analysis was performed from February 2023 to June 2024. It included cancer-free women (N = 527 144) residing in Stockholm, Sweden, who were invited to undergo mammography screening (aged 40-74 years) during 1989 to 2020. An additional cohort of women were included who were participating in the Karolinska Mammography Project for Risk Prediction of Breast Cancer study and had mammography data available.ExposuresFamily cancer history (defined from the Swedish Multi-Generation Register and Cancer Register), mammographic density, and various demographic, reproductive, and other factors (multiple Swedish registers).Main Outcomes and MeasuresIncidence of ScrCa and IntCa (defined from the Swedish Cancer Register in conjunction with individual screening histories).ResultsA total of 29 049 women (5.5%) received a diagnosis of BC, of whom 10 631 (2.0%) had ScrCa and 4369 (0.8%) IntCa. ScrCa and IntCa incidences increased during the period. The proportion of IntCa among screened patients with BC was around 30%, which decreased with older age. Factors associated with increased risk of IntCa included older age at first childbirth, higher education level, hormone replacement therapy, and higher mammographic density. Risk estimates of family cancer history on IntCa were family history of BC (hazard ratio [HR], 1.85; 95% CI, 1.72-1.99), family history of IntCa (HR, 2.92; 95% CI, 2.39-3.55), and hereditary breast and ovarian cancers (HR, 1.45; 95% CI, 1.36-1.54), with risk further elevated with the number of relatives who received a diagnosis when younger than the median age. Women with IntCa were more likely to have estrogen receptor (ER)–negative cancers than women with ScrCa (22% vs 11%), and having family history of ER-negative BC was associated with 3-fold risk for ER-negative IntCa.Conclusions and RelevanceThe results of this cohort study suggest that IntCa rates have not decreased with age-based screening, and implementing risk-based screening considering IntCa-specific risk factors is necessary for improving outcomes.