Novel multi-omic biosignature for identification of early-stage invasive breast cancer patients who do and do not benefit from adjuvant radiation treatment after breast conservation surgery.

Abstract

e12566 Background: Adjuvant radiation therapy (RT) has been a standard component of breast conserving therapy (BCT), demonstrating a reduction in local regional recurrence (LRR) and an improvement in breast cancer mortality. In addition, endocrine therapy (ET) has been shown to have both ipsilateral and contralateral risk reduction benefits as well as mortality benefits. Given the benefits of RT and ET are not uniform for all patients, locoregional therapy for early stage- hormone receptor positive (HR+) breast cancer continues to evolve with a focus on individualized risk stratification; genomic and multiomic approaches are increasingly being considered to support informed decision making for patients, allowing clinicians to optimize treatment and identify patients who will have a significant benefit from RT and/or ET and those who will not. Methods: A group of 782 patients (T1/2, N0/1, HR+, HER2-) who underwent breast conservation surgery (BCS) between 1986-2022 were identified from a multi-institutional cohort of which 630 were 50 years or older and node negative. A multi-omic biosignature combining proteomic and genomic biomarkers (AidaBreast, PreludeDx, Laguna Hills, CA) was developed and cross-validated to calculate an individualized Decision Score (DS) on a 10-point scale and corresponding 10-year LRR risk for prognosis. The association between the DS, 10-yr LRR rate, and RT and ET benefit was assessed using multivariable (MVA) Cox proportional hazards. Results: The median age was 66 years old with 74% (465/630) of patients receiving RT and 37% (234/630) receiving ET, and 27% (n=171) receiving both ET and RT. Median follow up was 10 years. 39% (n=246) of patients had a Low Risk Decision Scores (DS≤4) and on MVA had no significant reduction in LRR risk with RT (HR= 0.92; p=.90) or ET (HR= 1.5; p=.65). The corresponding 10-year LRR risks were ≤5% independent of treatment with ET or RT. In contrast, patients with high Decision Scores (DS>4) (61%, n=384) had a significant reduction in LRR risk with RT (HR= 0.5, p=0.03) and had a trend toward reduced LRR rate with ET of (HR=0.5; p=0.16) with corresponding 10-year LRR risks of 20% (95%CI 10%, 28%) vs 12% (95%CI 7%,16%) without vs with RT. On MVA, neither age, grade, nor size were associated with LRR risk (p≥.4) after adjusting for DS. Conclusions: The multi-omic biosignature (AidaBreast, PreludeDX, Laguna Hills, CA) stratifies early-stage HR+ HER2-negative invasive breast cancer patients into those with little to no LRR benefit from RT or ET over 10-years versus those who have a meaningful reduction in LRR risk with adjuvant RT.