Age-specific interval breast cancers in New South Wales and meta-analysis of studies of women aged 40–49 years

Abstract

Objective: To compare interval cancers in the 40–49 year age group with other age groups in New South Wales and with published trials and service studies. Setting: New South Wales data were derived from the population-based biennial mammography screening program, which achieved state-wide coverage in 1995. Women aged 40–49 years screened during 1995–1998 were included. Methods: Bilateral two-view mammography with reading by two radiologists was employed for biennial screening examinations. Interval cancers were detected by the screening program and by linkage with the state-wide cancer registry. Incidence of interval cancer based on the date of diagnosis was estimated as a proportion of the expected underlying breast cancer incidence for first- and second-year interval cancers. Sensitivity estimates were also calculated. Comparative data for the 40–49 year age group were derived from the published literature for meta-analyses of trial and service studies. Results: Interval cancer rates for New South Wales decreased with increasing age, with the highest proportional incidence in the 40–49 year age group for first year (56%, 95% confidence interval [CI] 50–62%) and second-year (86%, 95% CI 82–90%) interval cancers. Proportional incidence for women aged 50–69 years for first- and second-year interval cancers was 31% (95%CI 29–33%) and 50% (95% CI 47–52%) respectively. Sensitivity estimates for the program increased significantly with age, with lowest sensitivity estimates evident for women 40–49 years. In women aged 40–49 years the meta-analysed proportional incidence rate for randomised trials of screening for first- and second-year interval cancers was 42% (95% CI 21–62%) and 63% (95% CI 55–71%) respectively, while for service studies it was 44% (95% CI 31–58%) and 72% (95% CI 51–92%). Proportional incidence in the New South Wales program for women aged 40–49 years was not significantly different to the meta-analysed proportional incidences for trial and service studies in the first year, or for service studies in the second year. Conclusion: Proportional incidence of interval cancer was significantly higher in women aged 40–49 years in New South Wales relative to older age groups, but did not differ significantly from service studies of women in a similar age group. The effectiveness of mammography screening for this age group needs to be examined in view of the comparatively high rate of interval cancers.