The multiomic tumor microenvironment landscape of invasive lobular carcinoma of the breast

• Published in: Journal of clinical oncology Vol. 42; no. 16_suppl; p. 1044
• Main Authors: Mouabbi, Jason A, Meric-Bernstam, Funda, Nasrazadani, Azadeh, Nelson, Blessie Elizabeth, Salem, Basim, Kushnarev, Vladimir, Tkachuk, Arina, Sotova, Maria, Polyakova, Margarita, Baranov, Oleg, Turova, Polina, Chernyshov, Konstantin, Kotlov, Nikita, Brown, Jessica H., Clayton, Patrick, Sarachakov, Alexander, Postovalova, Ekaterina, Bagaev, Alexander, Tripathy, Debashish, Layman, Rachel M.
• Format: Journal Article
• Language: English
• Published: 01.06.2024
• ISSN: 0732-183X; 1527-7755
• DOI: 10.1200/JCO.2024.42.16_suppl.1044

1044 Background: The tumor microenvironment (TME) in invasive lobular carcinoma (ILC) has been shown to play a critical role in tumor progression and therapeutic response. However, dissecting the complexity and heterogeneity of the TME requires advanced profiling techniques that overcome the limitations of traditional histopathological methods. This study leverages single-cell spatial phenotyping using multiplex immunofluorescence (MxIF) integrated with RNA sequencing (RNA-seq) to evaluate the TME landscape in ILC. Methods: MxIF was performed by processing 2,139,084 cells from FFPE whole-slide tissues of treatment naïve, de novo metastatic ILC patients treated at MD Anderson Cancer Center with the PhenoCycler system. Prepared slides were stained with a comprehensive panel of 27 biomarkers to delineate immune cells, immune checkpoints, tissue structure, and predictive biomarkers (i.e., ER, PR, and HER2). Bulk RNA-seq was performed, employing the Kassandra cell deconvolution algorithm (1) for advanced profiling, enabling a multi-omic correlation using Spearman’s rank correlation coefficient of spatial phenotypes with transcriptomic data (single cell expression data and immune gene expression signatures). Signature enrichment scores were calculated using the ssGSEA algorithm (2). Transcriptome-based TME subtyping was performed using functional gene signatures as described in ref (3). Results: Using MxIF, 14 primary cell populations and 25 distinct immune and stromal cell subtypes were delineated within the TME. RNA-seq analysis enabled us to categorize the TME into three major groups using RNA-seq analysis: Immune Enriched Non-Fibrotic (IE; 37% of samples), Highly Vascularized (HV; 30% of samples), and Immune-Enriched Fibrotic (IE-F; 23% of samples). Each type demonstrated unique spatial phenotypes as revealed by MxIF, corresponding to the RNA-seq derived TME subtypes. MxIF analysis of the IE group displayed a robust immune presence with minimal fibrosis. Conversely, in the IE-F group there was significant fibrosis alongside dense immune cell infiltrates as evidenced by a large fraction of fibroblasts and CD45+ per mm2. CD31+ endothelial cells were predominant in the HV group, indicating an extensive vascular network. Conclusions: By integrating RNA-seq and MxIF, we present a detailed multiomic analysis that uncovers the intricate heterogeneity of the TME in ILC. Further evaluation of the TME with this approach can lead to the development of personalized therapeutic strategies, emphasizing the value of comprehensive TME profiling in understanding and treating ILC. 1. Zyatsev et al., Cancer Cell, 2022. 2. Barbie et al, Nature, 2009. 3. Barbie et al, Nature, 2009.