ESTRO 2021 Abstract Book

S156

ESTRO 2021

Results From October 2018 to December 2020, 55 patients(p) have been enrolled. At time of this analysis, 36p (50% TN and 50% HER2+) have completed primary radiochemotherapy and surgery. With a median follow up of 15.7 months (7-37), 23p (64%) reached pCR (Miller- Payne G5) whereas response in 31p (86%) was classified as grade G4-G5 (Table 1). Table 1. Phenotype/Pathological response grade G3 G4 G5 total TN 3 (17%) 2 (11%) 13 ( 72% ) 18 Her-2+ 2 (11%) 6 (33%) 10 ( 56% ) 18 total 5 (14%) 8 (22%) 23 (64%) 36 We performed a comparative analysis with Chi-Square test. According to molecular subtype, 13p TN (72%) reached pCR vs. 10p HER2+ (56%), p = 0.052. We also found that age <52 years-old and ki 67 < 55 were related to achieving pCR (p: 0.018 and p: 0.052 respectively) Presence of tumor-infiltrating lymphocytes (TILs) was associated with response to neoadjuvant chemotherapy among patients with TN/HER2+ breast cancer. TILS<10% associated pCR in 9/10p (90%) vs. 6/6p (50%) when TILS≥10% (p = 0.011) Clinical tolerance was acceptable without any acute or late toxicities >G3. Surgical wound complication was observed in 1p (3%). Conclusion Preoperative concomitant radiochemotherapy is a feasible, well tolerated and promising approach for TN/HER2+ breast cancer patients, although further follow-up is necessary to confirm the clinical relevance of observed pCR rates. PH-0224 Association between the genomic immune response and locoregional control in ER- breast cancer C. Liveringhouse 1 , C. Wilson 2 , M. Mills 3 , R. Thapa 2 , B. Fridley 2 , T. Robinson 3 , R. Diaz 3 1 H. Lee Moffitt Cancer Center, Department of Radiation Oncology, Tampa, USA; 2 H. Lee Moffitt Cancer Center and Research Institute, Department of Biostatistics and Bioinformatics, Tampa, USA; 3 H. Lee Moffitt Cancer Center and Research Institute, Department of Radiation Oncology, Tampa, USA Purpose or Objective Prior data suggest an immunosuppressive tumor microenvironment associated with estrogen receptor negative (ER-) primary breast tumors is associated with inferior locoregional control (LRC). Improved understanding of the mechanisms of immune silencing and activation is crucial to improving LRC for this disease. Materials and Methods We retrospectively identified 151 patients with non-metastatic ER- breast tumors who did not receive neoadjuvant systemic therapy and who had available genomic profiling. Gene expression was assessed with Affymetrix Rosetta/Merck Human RSTA 2.0 microarray chips. Populations of tumor infiltrating immune cells were inferred from gene expression data via the CIBERSORT deconvolution algorithm. High vs low expression/proportion was defined as above vs below the median. Correlation between continuous variables was assessed with the Pearson correlation coefficient. Time to event analyses were performed with Kaplan Meier and Cox proportional hazards models. Results The median age at diagnosis was 57 years, 97% of tumors were T1-2, and 30% had positive axillary lymph nodes. All patients were treated surgically with mastectomy (57%) or lumpectomy (43%); 82% received adjuvant chemotherapy, and 54% received adjuvant radiation. High proportion (≥median) of activated memory CD4+ T helper cells (Th), CD8+ T cells, activated natural killer (NK) cells, and memory B cells were associated with superior LRC (p<0.05). A high ratio of Th to T regulatory cells (Treg) (p=0.05) and M1 to M2 macrophages (p=0.043) were also associated with improved LRC. High expression of structural major histocompatibility II (MHC-II) genes such as CD74, HLA-DQA1, HLA-DPA1, HLA-DOA was associated with improved LRC (p<0.05). Expression of MHC-II genes were positively correlated with expression of downstream adaptive immune response genes including IFN gamma and IFN-gamma responsive chemokines (CXCL9-11) and signal transducers (STAT1), Th response (LAG3, CD3), B-cell function (MS4A1), and cytotoxic CD8+ T- and NK cell function (GZMA, GZMK, PRF1, NKG7). Expression of HLA-DOA was positively correlated with an increased ratio of Th to Treg, and M1 macrophages to M2 macrophages. There was no significant correlation between any MHC-II gene expression and proportion of tumor infiltrating dendritic cells, suggestive of tumor MHC-II expression. There was no significant interaction between radiotherapy and MHC-II expression or Th to Treg ratio on Cox proportional hazards models. Conclusion Stimulation of a robust adaptive immune response via antigen presentation by MHC-II is associated with improved local control for ER- breast cancer. These data lead to the hypothesis that breast tumors themselves may activate this response by self-presentation of MHC-II molecules in an IFN-gamma dependent manner. Histological tissue evaluation is needed to confirm the presence of breast tumor MHC-II presentation. Similar immune biomarkers may be useful to predict which patients might benefit from immunomodulatory therapy.