Small biopsies of primary and metastatic pancreatic cancers recapitulate tumoral and stromal heterogeneity

Background Pancreatic ductal adenocarcinoma (PDAC) has a dismal prognosis with most patients presenting with advanced disease. With molecular subtypes of PDAC gaining clinical relevance, the ability to characterise tumor heterogeneity is of great interest. Single cell RNA (scRNA) sequencing provides the opportunity to understand multistep cancer and stroma progression at high resolution, but it has been deemed unfeasible in small biopsies like core and endoscopic ultrasound-guided find needle aspirations (EUS-FNA) so far.

Material and methods After tissue dissociation, small biopsies of 4 primary tumors and 4 metastases were processed by a droplet-based scRNA technology (10X). A total of 31,130 cells were analyzed using Seurat to comprehensively characterize epithelial and stromal cells amongst lesion types.

Results Epithelial cells (46.17%) were most commonly represented, followed by T cells (28.52%), and myeloid cells (12.79%), whereas natural killer (NK) cells (5.19%), endothelial cells (1.31%), fibroblasts (1.28%), and dendritic cells (DC, 0.51%) were less frequently found. Subgroup analysis revealed that cell phenotypes were largely equally represented across primary and metastatic lesions. This suggests that the composition of the tumor microenvironment across disease stages is not significantly different. By applying previously identified molecular subtypes of PDAC including Bailey, Collison, and Moffit gene sets, we were able to itemize the contributions of cell type to transcriptomic classifications. This revealed that the Moffit molecular subtypes of classical and basal cells was more specific with identifying cells of cancer origin. Trajectory inference of stromal cells revealed the evolution of an immunosuppressive environment, while epithelial cells demonstrate divergence of basal-like subtype with enrichment of interferon response and metastatic related pathways including epithelial-mesenchymal transition and cell migration.

Conclusion Our work demonstrates the feasibility of single cell sequencing from real-world small biopsies. We reveal substantial cellular heterogeneity and uncover molecular subtypes and cell type specific pathways enriched during metastatic progression at single-cell resolution.

Publication History

Article published online:
08 September 2020