Integrative Analysis on Malignant Pheochromocytoma or Paraganglioma in TCGA and EMBL-EBI

Abstract

Introduction: Methods of predicting malignant behavior of pheochromocytomas (PCCs) or paragangliomas (PGLs) are needed. However, there are few reliable histopathologic criteria to predict malignant behavior in PCC/PGLs. Recent genomic analysis of The Cancer Genome Atlas (TCGA) and European Molecular Biology Laboratory-European Bioinformatics Institute (EMBL-EBI) has provided genetic information enabling more accurate differentiation of disease entities. TCGA and EMBL-EBI could be utilized for the genetic analysis of malignant PCC/PGLs. Therefore, the objective of this study was to determine genomic expression differences and mutational differences of malignant PCC/PGL using data from TCGA and EMBL-EBI.

Methods: TCGA had data of multigenomic analysis for 184 PCC/PGL samples while EBML-EBI had data for 202 PCC/PGL samples. Clinical information, mutation status, and mRNA expression dataset were downloaded from TCGA and EMBL-EBI. Following inclusion/exclusion criteria, 59 of 184 PCC/PGL samples in TCGA and 171 of 202 PCC/PGL samples in EMBL-EBI were selected. Of 59 samples in TCGA, 12 were malignant and 47 were benign. Of 171 samples in EMBL-EBI, 19 were malignant and 152 were benign. Data of mRNA expression and mutations were compared between these two groups.

Results: Thirty up/down-regulated pathways in malignant PCC/PGLs were related to cancer signaling, metabolic alteration, prominent mitosis, and junctional dissociation. Twenty-one up-regulated genes and 11 down-regulated genes were significantly enriched in functional annotation pathways. Malignancy was more often in germline mutations of SDHB or VHL. Four candidate mutational genes (ATRX, SETD2, MUC16, and PAX1) were proposed as genes susceptible to malignancy. Overall survival rates were significantly correlated with the presence or absence of these somatic mutations.

Conclusions: TCGA and EMBL-EBI showed differences in mRNA expression and mutations between malignant and benign PCC/PGLs. Improved recognition of prognostic factors can help us achieve proper diagnosis and provide appropriate treatment for PCC/PGLs.