Endocrine Abstracts

Background: Pancreatic neuroendocrine tumors (PNETs) are a rare, understudied form of cancer with few curative options, and their occurrence is rising. With approximately 4,000 new cases diagnosed per year in the US, and a 5-year relative survival rate of 54%, it’s imperative to understand the molecular mechanisms governing PNET carcinogenesis. To date, most studies used whole-tumor sequencing to characterize PNETs, which hampers the discovery of their microenvironment, cellular composition, and cell of origin. In this study, we used single-cell assays on cryopreserved primary PNETs that vary in grade, stage, and metastasis status to characterize the chromatin and transcriptomes of single cells obtained from resected tumors.

Methods: Assay for Transposase Accessible Chromatin sequencing (ATAC-seq) is a robust method to probe areas of accessible chromatin which likely function as regulatory genomic regions called enhancers. We used a nuclei isolation protocol designed to recover nuclei from cryopreserved tissue for single-cell assays. We performed single-nucleus ATAC-seq and RNA-seq using a 10x genomics multiomics platform. We then used a suite of bioinformatic tools to perform clustering, marker gene identification, and molecular pathway analysis.

Results: We identified several distinct cell populations within PNETs, including cancer associated fibroblasts, immune cells, and cancer cell populations. Our preliminary analysis suggests there is little to no heterogeneity within individual tumor cells. In contrast, each tumor is significantly different than others with little shared genes. When we compared the tumor samples to normal pancreas cell profiles, we found that the most aggressive PNETs tend to gain acinar or duct-like identity as they progress in grade.

Conclusions: To date, the mechanisms underlying of PNET progression are unknown and there is little known about intra-PNET heterogeneity. The current targeted therapy approved for metastatic PNETs is Lu177-DOTATATE. However, it is affected by tumor heterogeneity and ineffective against metastatic high-grade tumors. Our results provide guidance for development of targeted therapies. Moreover, our study provides new molecular markers that vary between PNETs, allowing to further subtype PNETs. Moving forward, our group aims to use the data from this study to guide the development of an in-vitro PNET tumor progression model.

Abstract ID 21437