Abstract Background: Internal mammary node (IMN) metastasis is an adverse prognostic factor in breast cancer. While surgical dissection of IMNs is rarely performed due to its invasiveness and lack of proven survival benefit, internal mammary node irradiation (IMNI) remains the standard approach for managing clinically positive IMNs. With the increasing use of neoadjuvant therapy (NAT), a critical question emerges—whether a boost to the region of IMN metastasis is still necessary when an imaging based complete nodal response was observed. This study aims to evaluate whether omitting a postoperative IMN boost is oncologically safe in patients with baseline imaging-positive IMNs who achieve an IMN clinical complete response (icCR) following NAT. Methods: This retrospective cohort study included 1326 female breast cancer patients treated between 2018 and 2021 at our center. Inclusion criteria were: (1) histological confirmed invasive breast cancer with M0; (2) received standard NAT; (3) underwent breast-conserving or modified radical mastectomy; (4) received postoperative radiotherapy with IMNI; (5) underwent pre- and post-NAT breast magnetic resonance imaging (MRI). IMN metastasis was defined by MRI features including short-axis diameter ≥5 mm and suspicious morphological features, such as loss of hilum or high-signal in diffusion weighted imaging. IcCR was defined as complete disappearance of all previously positive IMNs on post-NAT MRI. All imaging was independently reviewed by two radiation oncologists, with discrepancies resolved by consensus. The postoperative IMNI regimen delivered 50 Gy in 25 fractions to IMN, with a subset of patients receiving a simultaneous integrated boost to 60 Gy to the pre-NAT involved internal mammary node site. Prognosis was evaluated by disease-free survival (DFS), local recurrence-free survival (LRFS), distant metastasis-free survival (DMFS), and overall survival (OS). Multivariable analysis was performed using Firth-corrected Cox regression. Sensitivity analyses including standard Cox and Bayesian Cox models with informative priors were utilized to validate the robustness of findings. Interaction between icCR status and IMN boost was also evaluated. Results: Of 1326 patients, 208 (15.7%) presented with imaging-positive IMN at baseline. The median follow-up was 53 months. IMN-positive patients had significantly worse 3-year DFS than IMN-negative patients (85.3% vs. 94.9%, HR=3.48, P0.001). Among IMN-positive patients, 132 (63.5%) achieved icCR. Neither icCR (HR=0.63, 95%CI: 0.30-1.33, P=0.226) nor IMN boost (HR=0.71, 95%CI: 0.35-1.46, P=0.350) were independent predictors of DFS. However, a significant interaction between icCR status and IMN boost was observed (HR for interaction=2.79, 95%CI: 1.02-7.61, P for interaction=0.045), confirmed by Bayesian analysis (posterior HR=1.85, 95%CrI: 1.15-3.80, posterior P= 0.978). Subgroup analysis showed that an IMN boost did not improve DFS in icCR patients (HR=1.06, 95%CI: 0.41-2.78, P=0.899), but significantly reduced risk by 70% in non-icCR patients (HR=0.30, 95%CI: 0.10-0.88, P=0.029). For other endpoints, no significant interactions between icCR and IMN boost were found, although a trend was noted for DMFS (P for interaction = 0.078), and the analyses may have been underpowered due to a limited number of events. Conclusion: Achieving icCR post-NAT is not an independent prognostic factor but serves as a predictive marker for IMN boost benefit. A postoperative IMN boost confers substantial DFS benefit in patients with residual disease but may be safely omitted in those achieving icCR. These findings support a risk-adapted radiation strategy and warrant validation in prospective and multi-institutional cohorts. Citation Format: Y. PengY. XiangJ. TieS. ZhangC. ShiC. GuoW. Wang. Magnetic resonance imaging based internal mammary node response after neoadjuvant therapy to guide postoperative internal mammary node boost radiotherapy in breast cancer abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr RF1-03.
Peng et al. (Tue,) studied this question.