Spatial transcriptomics of multifocal ileal neuroendocrine tumors reveals tumor heterogeneity based on tumor microenvironment and new biomarkers

Akitada Yogo; Naoki Akanuma; Grace E. Kim; Chrissie Thiriwell; Netta Makinen; Matthew Meyerson; Eric K Nakakura

doi:10.1530/endoabs.108.B2

B2

Prev Next

Section Contents Cite

Endocrine Abstracts (2024) 108 B2 | DOI: 10.1530/endoabs.108.B2

NANETS2024 17th Annual Multidisciplinary NET Medical Symposium NANETS 2024 Basic Science (15 abstracts)

Spatial transcriptomics of multifocal ileal neuroendocrine tumors reveals tumor heterogeneity based on tumor microenvironment and new biomarkers

Akitada Yogo ^1,2 , Naoki Akanuma ³ , Grace E. Kim ³ , Chrissie Thiriwell ⁴ , Netta Mäkinen ^5,6 , Matthew Meyerson ^5,6,7 & Eric K Nakakura ^1,2

Author affiliations

¹Division of Surgical Oncology, Section of Hepatopancreaticobiliary Surgery, Department of Surgery, University of California, San Francisco, San Francisco, CA; ²Helen Diller Family Comprehensive Cancer Center, San Francisco, CA; ³Department of Pathology, University of California, San Francisco, CA; ⁴Bristol Medical School, University of Bristol, Bristol, UK; ⁵Department of Medical Oncology, Dana-Farber Cancer Institute, 450 Brookline Avenue, Boston, MA; ⁶Cancer Program, Broad Institute of Harvard and MIT, Cambridge, MA; ⁷Departments of Genetics and Medicine, Harvard Medical School, Boston, MA

Background: Ileal neuroendocrine tumors (i-NETs) are characterized by a high incidence of multiple primary tumors (>30-40%) and production of serotonin/other hormones. Recent whole genome sequencing analyses revealed an absence of shared somatic variations among synchronous primary tumors, so the mechanisms underlying multifocal tumor development are not known. In this study, we evaluated gene expression patterns of multifocal i-NETs with spatial resolution in order to develop new hypotheses about tumorigenesis focusing on the tumor microenvironment.

Methods: FFPE blocks of surgically resected specimens from 4 patients with multifocal i-NETs were used. Tissue microarrays were constructed from 72 cores (18 one-mm cores per patient). Spatial gene expression libraries were constructed using Visium v1 (10x Genomics). A total of 8,295 spots were analyzed: a median of 3,102 genes, 5,944 UMI counts per spot, and a total of 16,778 genes were detected in each capture area. R packages including Seurat, clusterProfiler, and monocle3 were used for data analysis.

Results: Spatial transcriptomics analysis reliably captured spatial information of malignant and non-malignant cells in distinct tissue compartments within the ileum and regional lymph node/mesenteric masses. Unsupervised clustering demonstrated differences of the i-NET in various microenvironments--mucosa, submucosa, muscularis propria, and intranodal/perinodal regions of the lymph node/mesenteric masses. In all 4 patients, gene expressions of tumor cells in the mucosa were similar among multifocal tumors while tumor cells in other microenvironments clustered separately. Trajectory analysis was consistent with the supposition that tumors in the mucosa are likely the origin of i-NETs found in the other microenvironments. Tumor cells in all microenvironments exhibited characteristic gene expression patterns of serotonin receptors (HTR1B, HTR1D, and HTR7), ghrelin receptor (GHSR), GIP receptor (GIPR). Over-representation or enrichment analysis found that tumor cells in the mucosa or those with potential for metastasis exhibited overexpression of specific gene sets, including genes related to lipid metabolism.

Conclusions: This is the first spatial transriptomics analysis of multifocal i-NETs revealing similarities among tumors located in the mucosa and distinct clustering of tumors situated in other microenvironments. The finding that tumor heterogeneity in multifocal i-NETs varies depending on the microenvironment has not been previously described. The spatial data also reveal known, as well as, new biomarkers of i-NETs. In particular, we identified serotonin and other hormone receptors—some of which may be tumor specific--suggesting possible endocrine/paracrine signaling in i-NETs.

ABSTRACT ID28490