Determining Percent EGFR Activation in Tumor Cells in Intact FFPE Tissue Sections.
Clifford Hoyt, Kristin Lane, Gregory Innocenti, Randall Wetzel. Cambridge Research and Instrumentation, Inc., Woburn, MA; Cell Signaling Technology, Danvers, MA
Background: Many areas in oncology research depend on revealing signaling pathway activity in FFPE tissue sections. However, techniques such as immunohistochemistry, micro-array detection, and analysis of sample lysates provide averages from volumes of tissue, including many cells not of interest. These methods blur out key proteomic information that resides at the cellular level, relating to the signaling states of individual cells. The purpose of this study is to demonstrate effective, practical and reliable cytometric analysis of signaling pathway proteins in intact tissue sections, using EGFR expression and activation as examples. In lung cancer, EGFR expression and/or activation have been linked to therapeutic response as well as angiogenesis and metastasis.
Design: The approach integrates multiplexed immuno-fluorescence labeling technology that is robust and specific, automated multispectral high-throughput image acquisition to capture and distinguish multiple labels (VectraTM), and image analysis algorithms to differentiate relevant tissue regions (e.g., malignant and normal epithelia, stroma, necrosis, etc.) and segment cellular compartments. We designed a pilot study using human tumor xenografts to determine whether per-cell co-expression subtypes can be reliably determined. We describe methodology and present results using amplified and unamplified non-small cell lung carcinoma cell lines (H1975/L858R, H3255/L858R, H1650/E746_A750del, and HCC827/E746_A750del). A multi-label immunofluorescence kit from Cell Signaling Technology was used, labeling total EGFR, pEGFR and pERK, plus a DAPI counterstain.
Results: Analysis was performed on sections from multiple xenografts of from each cell line. Vectra's pattern-recognition-guided high power field selection was used to acquire ten 20x multispectral images of tumor regions. Machine-learning-based automated image analysis was performed to locate cancer cells, segment subcellular compartments, and extract IF signals on a per-cell basis from each field. Percentage of EGFR+ cells that were actively signaling (pEGFR+) was calculated for each xenograft. Percent expression levels concur with molecular phenotype.
Conclusions: Together, the multispectral platform and image analysis software, with multiplex panels of activation state-specific antibodies, reveal molecular subtypes, which may lead to new targeted strategies for oncology research and potentially clinical care, to diagnose progression of disease, guide therapy selection, and monitor response.
Tuesday, March 1, 2011 1:00 PM
Poster Session IV # 263, Tuesday Afternoon