Visual analytics identifies key miRNAs for differentiating peripancreatic paraganglioma and pancreatic neuroendocrine tumors.

Abstract

Paragangliomas (PGL), a type of neuroendocrine tumor, pose a significant diagnostic challenge due to their potential for unpredictable locations and asymptomatic presentation. Misdiagnosis of peripancreatic PGLs, particularly as pancreatic neuroendocrine tumors (PANNETs), is a pressing issue as it can negatively impact both pre- and post-treatment decision-making. The aim of our study was to identify microRNA markers for the reliable differential diagnosis of peripancreatic PGLs and PANNETs, addressing a crucial unmet need in the field and advancing the standard of care for these patients. Morphing projections tool was used to analyze miRNA data from PGL and PANNET tumors present in the TCGA database. The findings were validated using two additional databases: GSE29742 and GSE73367. Our research uncovered substantial differences in the miRNA expression profiles of PGL and PANNET, leading to the identification of 6 key miRNAs (miR-10b-3p, miR-10b-5p, and the miRNA families miR-200c/141 and miR-194/192) that can effectively differentiate between the two types of tumors. These miRNA levels hold potential as biomarkers for improved diagnosis, offering a solution to the diagnostic challenge posed by these tumors and potentially improving the standard of care for patients.