Time-varying risks of second events following a DCIS diagnosis in the population-based Vermont DCIS cohort.

Abstract

Long-term disease-free survival patterns following surgical, radiation, and endocrine therapy treatments for ductal carcinoma in situ (DCIS) are not well characterized in general US practice. We identified 1252 women diagnosed with DCIS in Vermont during 1994-2012 using data from the Vermont Breast Cancer Surveillance System, a statewide registry of breast imaging and pathology records. Poisson regression and Cox regression with time-varying hazards were used to evaluate disease-free survival among self-selected treatment groups. With 7.8years median follow-up, 192 cases experienced a second breast cancer diagnosis. For women treated with breast-conserving surgery (BCS) alone, the annual rate of second events decreased from 3.1% (95% CI 2.2-4.2%) during follow-up years 1-5 to 1.7% (95% CI 0.7-3.5%) after 10years. In contrast, the annual rate of second events among women treated with BCS plus adjuvant radiation therapy increased from 1.8% (95% CI 1.1-2.6%) during years 1-5 to 2.8% (95% CI 1.6-4.7%) after 10years (P < 0.05 for difference in trend compared to BCS alone). Annual rates of second events also increased over time among women treated with BCS plus adjuvant radiation and endocrine therapy (P = 0.01 for difference in trend compared to BCS alone). The rate of contralateral events increased after 10years for all groups with adjuvant treatments. The rate of second events did not vary over time among women who underwent ipsilateral mastectomy (P = 0.62). Long-term risk of a second event after DCIS varies over time in a manner dependent on initial treatment.