Background: There is an unmet need to identify women diagnosed with DCIS with elevated risk after standard treatment for whom enhanced treatment strategies should be considered. A response type (RSt) signature has been assessed in women treated with breast conserving surgery (BCS) with or without radiotherapy (RT), or with mastectomy. Methods: Women diagnosed with pure DCIS and treated with BCS or, BCS with whole breast RT were consecutively collected in Sweden, 1986-2004 and the USA, 1999-2008. A third cohort treated with mastectomy was collected in Sweden from 1986-2004. Patients with FFPE tissue were included unless they had prior or simultaneous invasive breast cancer (IBC). The RSt and DCISionRT biosignatures were calculated using biomarkers (p16/INK4A, Ki-67, COX-2, PgR, HER2, FOXA1, SIAH2), scored by board certified pathologists. Pathology and clinical data were collected from medical records. Only women at elevated risk by DCISionRT (DS > 3) and negative margins were included in analysis. Total breast event risk included either ipsilateral DCIS or IBC. Multivariate Cox proportional hazards and survival analysis were used to assess cumulative incidence risk differences, hazard ratios, and 10-year risks adjusted for year of diagnosis. Results: Women with a poor RSt were all HER2 positive and remained at particularly elevated risk after BCS and RT (23% 10-yr risk, p < 0.0001, Table), with a risk profile similar to women treated with BCS without RT (RSt not evaluated). The distribution of age and size were not statistically different for good versus poor RSt, but poor RSt were more commonly Grade 3. In a small cohort of women treated with mastectomy, those women with a poor RSt were still at elevated risk after surgery, comparered to women with a good RSt (HR = 5.4, p = 0.014). The 10-year risk profile for women with a good RSt treated with BCS plus RT or mastectomy, was low (table). Conclusions: A novel biosignature identified women with a good and poor RSt, where the good RSt identified women with an apparent substantial benefit from RT, as well as women with a poor RSt who remained at particularly elevated risk after RT, and for whom enhanced treatment strategies should be considered. Additional validation studies are ongoing.

*Adjusted for Year of Diagnosis ≥ 1996