Abstract P4-06-08: The spatial localization of immune cells predicts prognosis and response to therapy in inflammatory breast cancer

Abstract

Abstract Introduction The mechanisms contributing to the aggressive biology of inflammatory breast cancer (IBC) are still under investigation. Our lab reported a 79-gene signature that is shaped by specific immune response programs and discriminates between IBC and non-IBC (nIBC). Furthermore, the presence of cytotoxic CD8+ immune cells is associated with a better prognosis in proliferative subtypes of breast cancer. However, not only the presence of CD8+ cells, but also the interaction with other immune cells plays a role in the functional immune response. In this study we assessed the spatial associations between immune cells in IBC. Methodology Affymetrix gene expression data of 105 IBC patients were analyzed using CIBERSORT and xCell modules to narrow down the number of stainings for the immunophenotyping. To analyze the composition of the immune infiltrate, we used five validated antibodies: CD79α (B-cell lineage), CD8 (cytotoxic T-cells), FOXP3 (Tregs), CD163 (Tumor associated macrophages, TAMs) and the SP142 PDL1 antibody. A standard H&E stained section was used to mark the tumor area on pretreatment biopsy sections. Subsequently, 5 slides were stained according to a validated protocol, scanned and evaluated using VISIOPHARM® software that makes virtual multiplexing possible after the alignment of the scanned images. Using both point pattern analysis and the Morisita–Horn index (MHI), developed for ecological studies, we assessed the co-localization of the different types of immune cells. Currently, we report the result of our validation cohort (30 patient samples). Results Most of our IBC patients presented with a hormone receptor positive carcinoma (64.7%). Almost a quarter of the patients (23.3%) with initially localized disease achieved complete pathological response (pCR) after neo-adjuvant chemotherapy (NACT). For every staining we report the median relative marker area (RMA), MHI for colocalization with CD8 applying a square tessellation of 100 μm and the number of cells in a radius of 30 μm around CD8+ cells (direct cell-cell contact) in table 1. RMAMHI # X+ cells (30 μm)CD80.33 % CD1630.12 %0.7213.13CD79α0.04 %0.6522.52FOXP30.01 %0.7011.37PDL1 0.7460.70Table 1: Median spatial properties. The RMA of CD8 predicted pCR after NACT (RMA= 1.07% pCR vs 0.35% no pCR, P=0.04), but was not prognostic for OS (P= 0.445). PDL1 positivity predicted neither pCR nor OS in this cohort. However, OS of patients with more PDL1+ cells ( > 0.703) in close contact with the CD8+ cells was significantly shorter (5y OS: 50% vs 68%, P= 0.03). Interestingly, the colocalization of CD8+ cells with TAMs (MHI= 0.69 no pCR vs. 0.75 pCR, P= 0.04) and CD79α+ B-cells (MHI= 0.63 no pCR vs. 0.69 pCR, P= 0.04) was also associated with pCR after NACT, while the number of CD163+ or CD79α+ cells was not. Conclusion In this study we described the dynamic interplay between cancer and immune cells. In a validation cohort of 30 patient samples we showed that the colocalization of TAMs or B-cells with cytotoxic T cells was associated with pCR after NACT. Furthermore, patients with more PDL1+ cells around CD8+ cells (r= 30 μm) had a worse prognosis while solely the number of PDL1+ or CD8+ cells was not prognostic. By December we will present data on 179 IBC patients. Citation Format: Van Berckelaer C, Colpaert C, Rypens C, Marien KM, Waumans Y, Kockx M, Vermeulen P, Dirix L, Van Laere S, Van Dam P. The spatial localization of immune cells predicts prognosis and response to therapy in inflammatory breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-06-08.