Mapping inter- and intra-tumor heterogeneity in Ductal Carcinoma in situ and invasive breast cancer using integrative multi-omic profiling

Document Type


Publication Date



oregon; chiles


Background: Molecular profiling of ductal carcinoma in situ (DCIS) has shown some prognostic utility in the clinic. However, there is still an incomplete understanding of the diversity of molecular mechanisms by which DCIS progresses to invasive breast cancer (IBC).Here, we utilize integrative multi-omic profiling of co-occurring DCIS and IBC in humans as a model for mapping the relationship between tumor mutations, global gene expression and morphological changes in DCIS and IBC. Methods: We performed targeted panel DNA sequencing (Tempus xT), whole transcriptome profiling and hyperplex immunofluorescence (Lunaphore COMET) on a cohort of 39 pairs of co-occurring DCIS and IBC lesions. We performed an integrative analysis of the above multi-modal data to uncover similar and dissimilar molecular trajectories in DCIS/IBC pairs. DAVID, KEGG and Reactome databases were utilized to analyze differential pathway activity between DCIS and IBC lesions. Results: IBC samples significantly overexpressed pathways associated with cycling cells and proliferation. For DCIS, many of the enriched pathways were related to CYP enzymes and xenobiotic metabolism. We also assessed enrichment of specific transcription factor binding sites (TFBS) with the associated differentially expressed genes. The top site enriched in IBC was a promoter regulatory element of unknown function (M120 motif KRCTCNNNNMANAGC) as well as sites regulated by heat shock transcription factor 4 (HSF4). Top TFBS enriched in DCIS included MEF2, HMEF2, AR and CEBPE. The top pathways associated with IBC using DAVID were related to extracellular matrix (ECM) organization and regulation. Interestingly, the top pathways associated with DCIS also included several pathways associated with ECM organization. On Reactome pathways diagrams for ECM formation distinct components of ECM formation and organization were enriched in IBC vs DCIS lesions, with invadopodia formation components and proteoglycans specifically enriched in IBC and integrin/laminin signaling enriched in DCIS. Among the 39 cases profiled, we observed distinct patterns of DCIS/IBC pairs with highly similar mutational profiles indicative of derivation from a shared progenitor as well as pairs with little-to-no molecular concordance that likely arose independently. Even within mutationally similar DCIS/IBC pairs there were different degrees of gene expression concordance. We subdivided mutationally similar pairs into two groups based on gene expression similarity (Euclidean Distance) and identified 115 genes that differed between these groups (q< 0.05), with the dissimilar DCIS lesions hallmarked by upregulation of a variety of pathways related to B-cell function, complement cascade and cell cycle progression. Specific mutations also were related to conserved transcriptional programs in these lesions, with MCL-1 amplified tumors showing a common gene expression signature of 357 altered genes (q< 0.05) with key differences in extracellular matrix organization and adaptive immune response pathways. Dissimilarly, tumors with CKS1B copy number alterations associated with 627 differentially expressed genes (q< 0.05) with the top targets related to endothelial function and angiogenesis. Conclusions: Integrative multi-modal analysis helps us better understand tumor progression at multiple levels of biomolecular dysregulation and shows the complex interplay among genome alterations, gene expression reprogramming and tissue morphology. Such approaches may yield future biomarker signatures that better encapsulate the diverse mechanisms by which DCIS lesions evolve.

Clinical Institute


Clinical Institute

Women & Children




Kaplan H, Dowdell A, Ben-Shimol R, Crabtree A, Berry A, Robinson F, Carney C, Bifulco C, Piening B. SABCS Annual Meeting; December 5-9; San Antonio, TX. 2023: PO3-24-01.

This document is currently not available here.