Characterization of tumor immune microenvironment in breast cancer using multiplexed imaging
Leeat Keren1, Marc Bosse1, Diana Marquez1, Robert West1, Sean C. Bendall1, Michael Angelo1*
1Department of Pathology, Stanford University, California, USA
Invasive breast cancer (IBC) is a leading cause of cancer death in women in the US. Triple-negative breast cancer (TNBC) is an aggressive subtype of IBC, which lacks the receptors that commonly fuel breast cancer growth, and is therefore unresponsive to targeted therapy. To date, there is limited understanding of the molecular mechanisms underlying TNBC and no means to predict patient prognosis. Cancer development is a complex process that involves multiple cell types, each defined by coexpression of multiple proteins, and depends on the interplay between individual cells in the tumor, microenvironment, and immune system. This complexity is not captured with current diagnostic and analysis methods, which primariy rely on immunohistochemical assays that evaluate the expression of only one or two proteins. We apply a recently-developed technology, Multiplexed ion beam imaging (MIBI), for profiling TNBC. MIBI enables to obtain single-cell expression levels for ~50 proteins simultaneously within the spatial context and morphological features of the tissue. In MIBI, antibodies are labeled with metals, which can be detected by secondary ion mass spectrometry. The method allows to detect dozens of proteins simultaneously with high sensitivity across a large dynamic range. Importantly, MIBI is compatible with archived specimens, which allows profiling human samples with documented long-term clinical outcomes. Our analysis reveals considerable heterogeneity in morphology and expression between patients for both tumor cells and the microenvironment, including stromal cells, vasculature and extracellular matrix. We find large differences in size and composition of immune cell infiltrates into the tumors, which correlate with both immune- and tumor-expressed proteins. High-dimensional imaging allows to combine single-cell data on morphology, expression and spatial location to uncover tumor complexity and heterogeneity.