39P - Deciphering the connection of tumor genetics, tumor immune cell infiltration and prognosis in breast cancer

Date 04 May 2017
Event IMPAKT 2017
Session Welcome reception and Poster Walk
Topics Breast Cancer
Pathology/Molecular Biology
Presenter Jan Budczies
Authors J. Budczies, B.V. Sinn, B.M. Pfitzner, S. Villegas-Angel, B. Ingold-Heppner, S. Wienert, F. Klauschen, C. Denkert
  • Pathology, Charité Hospital, 10437 - Berlin/DE



With the advent of cancer immune therapies there is an urgent need to identify patients that profit from the new therapies and to overcome mechanisms of resistance.

We analyzed H&E sections together with mutation, mRNA expression and clinical data of the TCGA breast cancer cohort. Tumor infiltrating lymphocytes (TILs) were assessed in 879 tumors. Using specific mRNA makers (MCPcounter), the abundance of ten cell populations including T cells, B cells, NK cells, monocytic cells, myeloid dendritic cells and neutrophils was estimated for each of the tumors.

Among molecular subtypes, TILs were most abundant in TNBC. The content of T cells, B cells, NK cells and myeloid dendritic cells was highest in TNBC, while neutrophils were most abundant in luminal tumors. The former four cell types were positive prognostic markers of overall survival (OS) in TNBC, while neutrophils were a positive prognostic marker of OS in HR+/HER2- tumors. TILs correlated significantly with T cells (Spearman’s R=0.37), NK cells (R=0.34), monocytic cells (R=0.30), B cells (R=0.28), and myeloid dendritic cells (R=0.23). The top list of 22 mRNAs correlating most strongly with TILs (R>0.40) included the T cell markers CTLA4, ICOS and SRIPG and the chemokines CXCL9, CXCL10 and CXCL11.

Mutational load correlated positively with monocytic cells (R=0.14) and negatively with endothelial cells (R=- 0.25), neutrophils (R=-0.19) and fibroblasts (R=-0.13). MutSig 1 (age-related) correlated negatively with nine of the ten cell types including myeloid dendritic cells (R=- 0.27), endothelial cells (R=-0.25), neutrophils (R=-0.20), T cells (R = -0.18), B cells (R=-0.17) and NK cells (R=-0.17). MutSig 3 (defective DNA repair) correlated negatively with endothelial cells and neutrophils (both R=-014). Of the two APOBEC-related signatures, MutSig 2 did not correlate with any of the cell populations in contrast to MutSig13 that correlated positively with monocytic cells (R=0.19), cytotoxic lymphocytes (R=0.17), T cells (R=0.16), NK cells (R=0.15) and negatively with endothelial cells (R=-0.13).

The markers of immune cell microenvironment investigated here should be further evaluated in tissue cohorts from clinical trial testing immune therapies.

Clinical trial identification