Abstract IA008: Mouse-INtraDuctal (MIND): An in vivo model for the discovery of epithelial/stromal cross talks that drive DCIS invasive and metastatic progression

Abstract

Abstract A large fraction of human DCIS (>50%) may not need the multimodality treatment options currently offered to all patients. More importantly, while we may be overtreating many, we cannot identify those most at risk for invasion/metastasis. Revealing the cellular and molecular mechanisms by which some DCIS remain indolent while others advance to invasive and metastatic breast cancers is currently a clinical unmet need. To address this gap, we developed the Mouse-INtraDuctal (MIND) model, by which patient-derived (PDX) DCIS epithelial cells are injected intraductally and allowed to progress naturally in mice. Single cell RNA-sequencing (scRNA-seq) was utilized to profile the DCIS epithelial and stroma cells in progressors vs. non-progressors. To distinguish between stromal (diploid) cells and tumor (aneuploid) cells, we calculated CNA profiles from RNA using CopyKAT. Cell-type specific differential gene expression analysis of DCIS epithelial cells and microenvironment cell types in progressors and non-progressors was performed. We also predicted putative ligand/receptor interactions between the tumor cells and cell types in the microenvironment by CellPhoneDB. Among 37 PDX DCIS MIND models followed for a median of 9 months, 20 (54%) grafted into 95 glands, showed in vivo invasive progression (progressed) while 17 (46%), injected into 107 glands, remained non-invasive (non-progressed). ScRNA-seq was performed on 13 DCIS samples including 10 progressors and 3 non-progressors. The resulting data identified genes significantly upregulated (>log2 fold=1, FDR p<0.05) in the progressors. Notable genes included NEAT1, EIF4EBP1, SCGB2A2, TFF1 and TFF3 that were upregulated in the progressors. NEAT1, the core structural component of the paraspeckles, is frequently overexpressed in human cancers and its expression is correlated with worse survival in cancer patients. NEAT1 drives tumor progression by regulating genes involved in cellular growth, migration, invasion, metastasis, EMT, stemness, radio- and chemoresistance, supporting its role as a potential biomarker and therapeutic target. TFF1/TFF3 mRNAs show increased expression in metastatic breast cancers. EIF4EBP1 is located on chrom 8p11-p12 which is frequently amplified in breast cancer and is associated with poor clinical prognosis. CellPhoneDB identified expression of several receptor/ligand interactions including CD74/MIF involved in epithelial/stromal and stromal/stromal cross talks that may play a role in DCIS invasive progression. Conclusions: Future studies will validate our findings using patient DCIS samples with known long-term outcome and in vivo MIND models to further refine risk associated biomarkers for invasion/metastasis and to identify more effective preventive therapies. Citation Format: Fariba Behbod. Mouse-INtraDuctal (MIND): An in vivo model for the discovery of epithelial/stromal cross talks that drive DCIS invasive and metastatic progression [abstract]. In: Proceedings of the AACR Special Conference on Rethinking DCIS: An Opportunity for Prevention?; 2022 Sep 8-11; Philadelphia, PA. Philadelphia (PA): AACR; Can Prev Res 2022;15(12 Suppl_1): Abstract nr IA008.