Abstract 5307: Imaging mass cytometry enables identification of distinct tissue phenotypes in highly autofluorescent lung and colon cancer tissues, producing consistent data across serial sections

• Published in: Cancer research (Chicago, Ill.) Ill.), 2023-04, Vol.83 (7_Supplement), p.5307-5307
• Main Authors: Kala, Smriti, O’ Neill, Brenna, Pemberton, James, Mansfield, James, Schneider, Fabian, Hupple, Clinton
• Format: Article
• Language: English
• ISSN: 1538-7445; 1538-7445
• DOI: 10.1158/1538-7445.AM2023-5307

Background: Successful implementation of immunotherapy requires a deep understanding of the spatial interactions between various cell types in the tumor microenvironment (TME). Most fixed tissues are autofluorescent and staining with cyclic immunofluorescence methods often produces data that are difficult to analyze due to the challenges of background subtraction. Imaging Mass Cytometry™ (IMC™) is a powerful tool for high-plex imaging, utilizing CyTOF® technology to simultaneously assess 40-plus protein markers at subcellular resolution without spectral overlap or background autofluorescence. This study demonstrates a tissue phenotyping workflow in highly autofluorescent lung and colon cancer tissues using high-plex IMC, which produces reliable data that can be easily analyzed. Methods: Serial sections of lung and colon cancer tissues were stained with a 30-marker panel comprised of structural, tumor, stroma, immune cell, and immune activation markers as well as the IMC Cell Segmentation Kit, for improved nucleus and plasma membrane demarcation. The data analysis pipeline used Phenoplex™ (Visiopharm®) software for straightforward, accurate, and quantifiable phenotyping. The analysis workflow consisted of tissue segmentation (to define tumor, stroma, extracellular matrix, and regions of necrosis), nuclear detection using a deep-learning algorithm pre-trained on IMC DNA channels, cell segmentation, a threshold-based cellular phenotyping step, and spatial analyses. Statistical analyses of the reproducibility between serial sections were performed using a paired t-test. Results: In this work, we have shown that analysis of IMC images from highly autofluorescent lung and colon cancer tissues can uncover tissue phenotypic signatures of the TME through the determination of distinct immune cell types found in the vicinity of cancerous cells. Moreover, cell segmentation generated cell counts that were highly consistent across the serial sections, with only 2.7% variability. This demonstrates the power of IMC in generating robust data across adjacent serial sections, which can be easily analyzed without the need to train the analysis software for background subtraction. Conclusions: Overall, this work demonstrates that by avoiding autofluorescence, IMC can generate high-quality, reproducible data, consistent across serial sections. The images can be easily and accurately analyzed in a streamlined way using the Phenoplex software, thus empowering IMC users to be conf