Molecular Risk Stratification for Decision-Making Among Patients With Breast Cancer Eligible to Omit Radiation Therapy

Shoshana J Rosenzweig; Lillian A Boe; Hannah Y Wen; Boris Mueller; Diana Roth O'Brien; George E Naoum; Zeinab Abou Yehia; J Isabelle Choi; Emma Hahesy; Mahmoud El-Tamer; Simon N Powell; Atif J Khan; Lior Z Braunstein

doi:10.1016/j.ijrobp.2026.03.032

Molecular Risk Stratification for Decision-Making Among Patients With Breast Cancer Eligible to Omit Radiation Therapy

Int J Radiat Oncol Biol Phys. 2026 Mar 25:S0360-3016(26)00517-1. doi: 10.1016/j.ijrobp.2026.03.032. Online ahead of print.

Affiliations

¹ Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, New York.
² Department of Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York.
³ Department of Pathology, Memorial Sloan Kettering Cancer Center, New York, New York.
⁴ Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, New York.
⁵ Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, New York. Electronic address: BRAUNSTL@mskcc.org.

PMID: 41895478
DOI: 10.1016/j.ijrobp.2026.03.032

Abstract

Purpose: Radiation therapy (RT) following breast-conserving surgery reduces the risk of recurrence for early-stage breast cancer. A subset of patients with favorable clinical and pathologic characteristics can safely omit RT with a modest increase in recurrences, yet without survival implications. We sought to evaluate whether molecular risk profiling among such patients could further stratify them into (1) a higher-risk subgroup that may be inappropriate for RT omission and (2) an exceedingly favorable subgroup that might more strongly consider RT omission.

Methods and materials: Using a prospectively maintained institutional database, we evaluated patients aged ≥65 years who underwent breast-conserving surgery for early-stage, estrogen receptor-positive breast cancer (ie, patients putatively eligible for RT omission). Patients were stratified by clinicopathologic features and Oncotype DX Recurrence Score (RS; ≤25 vs >25) and were evaluated for oncologic outcomes and survival by receipt of RT.

Results: The overall cohort comprised 1587 patients: 92% had RS ≤25 (n = 1455) and 8% had RS >25 (n = 132). With a median follow-up of 74.4 months, the cumulative incidence of local recurrence was low throughout and did not significantly differ between those with RS ≤25 vs >25 (6-year rate 1.9% vs 0.8%; P = .5). Among those who did not undergo RT (n = 350), patients with RS >25 (n = 19) did not have a significantly higher incidence of local recurrence than patients with RS ≤25 (6-year rate 5.6% vs 4.2%; P = .5).

Conclusions: Among patients ≥65 years of age who underwent breast-conserving surgery for early-stage estrogen receptor-positive breast cancer, Oncotype DX RS did not further refine recurrence risk estimates for RT decision-making beyond the findings of landmark RT omission trials. Although the number of patients with Oncotype DX RS >25 who omitted RT was limited in this study, this subgroup did not exhibit a significantly higher risk than their low-molecular-risk counterparts, suggesting that risk score need not necessarily disqualify select patients who seek to forgo adjuvant RT. Prospective analyses with larger sample sizes will further elucidate the role of molecular assays in guiding RT decision-making across risk strata.