Abstract PD6-03: Spatio-molecular dissection of the breast cancer metastatic microenvironment
Daniel L Abravanel,Johanna Klughammer,Timothy Blosser,Yury Goltsev,Sizun Jiang,Yunjao Bai,Evan Murray,Shahar Alon,Yi Cui,Daniel R Goodwin,Anubhav Sinha,Ofir Cohen,Michal Slyper,Orr Ashenberg,Danielle Dionne,Judit Jané-Valbuena,Caroline BM Porter,Asa Segerstolpe,Julia Waldman,Sébastien Vigneau,Karla Helvie,Allison Frangieh,Laura DelloStritto,Miraj Patel,Jingyi We,Kathleen Pfaff,Nicole Cullen,Ana Lako,Madison Turner,Isaac Wakiro,Sara Napolitano,Abhay Kanodia,Rebecca Ortiz,Colin MacKichan,Stephanie Inga,Judy Chen,Aaron R Thorner,Asaf Rotem,Scott Rodig,Fei Chen,Edward S Boyden,Garry P Nolan,Xiaowei Zhuang,Orit Rozenblatt-Rosen,Bruce E Johnson,Aviv Regev,Nikhil Wagle
DOI: https://doi.org/10.1158/1538-7445.sabcs21-pd6-03
IF: 11.2
2022-02-15
Cancer Research
Abstract:Abstract Metastatic breast cancer (MBC) remains incurable due to inevitable development of therapeutic resistance. Although tumor cell intrinsic mechanisms of resistance in MBC are beginning to be elucidated by bulk sequencing studies, the roles of the tumor microenvironment and intratumor heterogeneity in therapeutic resistance remain underexplored due to both technological barriers and limited availability of samples. To comprehensively capture these characteristics we have adapted a research biopsy protocol to collect tissue for an array of single-cell and spatio-molecular assays whose performance we have optimized for MBC, including single-cell and single-nucleus RNA sequencing, Slide-Seq, Multiplexed Error-Robust FISH (MERFISH), Expansion Sequencing (ExSEQ), Co-detection by Indexing (CODEX) and Multiplexed Ion Beam Imaging (MIBI). To date, we have successfully performed single-cell or single-nucleus RNAseq in 67 MBC biopsies and generated detailed accompanying clinical annotations for each. These samples provide a representation of the clinicopathological diversity of MBC including different breast cancer subtypes (44 HR+/HER2-, 3 HR-/HER2+, 3 HR+/HER2+, 16 TNBC, 1 unknown), common anatomic sites of metastasis (37 liver, 9 axilla, 7 breast, 5 bone, 3 chest wall, 3 neck, 1 brain, 1 lung, 1 skin), metastatic presentations (53 recurrent, 14 de novo) and histologic subtypes in the breast (45 IDC, 7 ILC, 6 mixed, 3 DCIS, 1 mucinous, 5 unknown/NA). Following optimization, both single-cell and single-nucleus RNA seq perform well in these MBC biopsies recovering all expected cell types including the malignant, stromal (e.g. fibroblasts, endothelial cells), myeloid (e.g. monocytes, macrophages) and lymphoid compartments (e.g. T cells, B cells, NK cells) as well as relevant oncogenic programs (e.g. cell cycle programs in all compartments; EMT-like and ER signaling programs in the malignant compartment, immune checkpoint programs in the lymphoid compartment; and fibroblast activation and vascular homeostasis programs in the stromal compartment). In addition to differences between the two techniques, these data demonstrate substantial intratumor heterogeneity in cell type composition. For example in liver biopsies the average number of cells per sample compartment by single nucleus RNA-seq was 6745 malignant (56%, SD 4216), 4637 stromal (41%, SD 3727), 1196 lymphoid (8%, SD 1617) and 874 myeloid (6%, SD 852); in breast biopsies the average number of cells per compartment by single nucleus RNA-seq was 6421 malignant (70%, SD 3497), 1628 stromal (24%, SD 117), 333 lymphoid (4%, SD 170) and 213 myeloid (3%, SD 117). Additionally, we find both inter- and intra-tumor heterogeneity in expression patterns and programs including, for example, expression of ER, PR and HER2 within clinical receptor subtypes (log normalized counts for ER expression in tumor cells by single cell RNA-seq: HR+/HER2- 0.921 (SD 0.714); HR+/HER2+ 0.768 (SD 0.624); HR-/HER2+ 0.018 (SD 0.122); and HR-/HER2- 0.005 (SD 0.066). For a subset of 13 biopsies we are also completing the spatiomolecular characterization methods on serial sections of a single adjacent biopsy. This unique experimental setup was designed to enable efficient comparison and integration of these assays. In spite of differences between experimental techniques and readouts, cell typing can be approached by annotation transfer from matching single cell or single nucleus RNAseq data, enabling exploratory analyses including evaluation of spatial phenotypes and cell type colocalization. Overall, these single cell and spatial data afford a comprehensive atlas including cell types, cell states/programs, cell interactions and spatial organization in MBC lesions. Future analyses will include serial biopsies over time and integration of clinicopathologic data including therapeutic response and resistance. Citation Format: Daniel L Abravanel, Johanna Klughammer, Timothy Blosser, Yury Goltsev, Sizun Jiang, Yunjao Bai, Evan Murray, Shahar Alon, Yi Cui, Daniel R Goodwin, Anubhav Sinha, Ofir Cohen, Michal Slyper, Orr Ashenberg, Danielle Dionne, Judit Jané-Valbuena, Caroline BM Porter, Asa Segerstolpe, Julia Waldman, Sébastien Vigneau, Karla Helvie, Allison Frangieh, Laura DelloStritto, Miraj Patel, Jingyi We, Kathleen Pfaff, Nicole Cullen, Ana Lako, Madison Turner, Isaac Wakiro, Sara Napolitano, Abhay Kanodia, Rebecca Ortiz, Colin MacKichan, Stephanie Inga, Judy Chen, Aaron R Thorner, Asaf Rotem, Scott Rodig, Fei Chen, Edward S Boyden, Garry P Nolan, Xiaowei Zhuang, Orit Rozenblatt-Rosen, Bruce E Johnson, Aviv Regev, Nikhil Wagle. Spatio-molecular dissection of the breast cancer metastatic microenvironment [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr PD6-03.
oncology