Radiation Therapy De-escalation in Breast Cancer Patients With Pathologic Complete Response: An Integrated Analysis Using Real-World and Simulated Data

Jian Wang; Jing Ning; Simona F Shaitelman; Heather Y Lin; Isabelle Bedrosian; Yu Shen

doi:10.1016/j.ijrobp.2026.05.017

Radiation Therapy De-escalation in Breast Cancer Patients With Pathologic Complete Response: An Integrated Analysis Using Real-World and Simulated Data

Int J Radiat Oncol Biol Phys. 2026 May 21:S0360-3016(26)00733-9. doi: 10.1016/j.ijrobp.2026.05.017. Online ahead of print.

Authors

Jian Wang¹, Jing Ning¹, Simona F Shaitelman², Heather Y Lin¹, Isabelle Bedrosian³, Yu Shen⁴

Affiliations

¹ Departments of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, Texas.
² Breast Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas.
³ Breast Surgical Oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas. Electronic address: ibedrosian@mdanderson.org.
⁴ Departments of Biostatistics, The University of Texas MD Anderson Cancer Center, Houston, Texas. Electronic address: yshen@mdanderson.org.

Abstract

Purpose: Advances in neoadjuvant chemotherapy (NAC) have markedly increased pathologic complete response (pCR) rates in breast cancer, prompting re-evaluation of the need for adjuvant radiation therapy (RT) in patients with excellent treatment response. The NSABP B-51 trial showed no survival benefit from regional nodal irradiation in patients achieving nodal pCR after NAC; whole-breast RT was still routinely delivered irrespective of breast response. Thus, the potential to safely omit RT in patients achieving breast pCR remains uncertain.

Methods and materials: We conducted an integrated analysis combining real-world evidence, clinical trial data, and simulated extensions to evaluate long-term overall survival (OS) with versus without adjuvant RT in patients achieving pCR (defined as ypT0 ypN0) after NAC and breast-conserving surgery. Model inputs were derived from published trials, observational cohorts, and cancer registry data. We assessed OS across a range of local-regional recurrence (LRR) risks and time-to-recurrence distributions reflective of contemporary clinical practice.

Results: RT was associated with small OS gains across stages. For stage I disease, 5-year LRR rates of 3.0% with RT versus 7.5% without RT corresponded to 20-year OS of 85.6% and 84.0%, respectively (absolute difference 1.6%), requiring >12,000 patients for 80% power. For stage II to III disease (5-year LRR 3.0% vs 10.5%), 20-year OS was 81.6% vs 78.3%, respectively (difference 3.3%), requiring ∼3800 patients for adequate power. Treatment effects on OS were consistent across age groups (≥65 years) and in the triple-negative subtype.

Conclusions: Among patients achieving pCR after NAC, the projected OS benefit from adjuvant RT appears small. This supports efforts toward more selective RT use and suggests that RT de-escalation warrants further investigation in subgroups with favorable biology and excellent treatment response. These findings reinforce the need for response-adapted decision frameworks and the incorporation of pCR into future RT guidelines. As treatment becomes increasingly personalized, integrating pathologic response into RT decision-making may reduce treatment burden without compromising outcomes. Prospective randomized trials are needed to validate these findings.