The Emerging Epigenetic Landscape of Pituitary Tumors: Implications for Current Classification Schema

Introduction: The vast majority of pituitary tumors are benign; however, a fraction of tumors are invasive and have malignant potential. The cellular pathways that drive transformation in pituitary neoplasms are poorly characterized. Importantly, there have been few genetic alterations identified in pituitary tumors. We were consequently motivated to review the epigenetic landscape of pituitary tumors given that novel techniques in epigenetics, the study of alterations in gene expression, provide a new dimension to characterize these tumors, and their findings for pituitary tumors may hold implications for prognostication and management.

Methods: We reviewed primary epigenetic studies of pituitary tumors with a focus on histone modification, DNA methylation, and transcript modification.

Results: High levels of methylation may be associated with clinically aggressive behavior in pituitary adenomas. DNA methyltransferase overexpression has been detected in pituitary tumors, especially in high grade macroadenomas. Three large-scale methylome studies have been conducted thus far, demonstrating that methylation differences at CpG sites in promoter regions were found to distinguish several types of adenomas from normal pituitary tissue. However, the lack of correlation between hypermethylation and gene expression in these studies suggests that methylation alone does not necessarily correspond to changes in gene expression.

Histone deacetylases and methyltransferases modify histone tails to alter chromatin states. Histone modifications, particularly methylation at histone 3 (H3), have been linked to increased p53 expression and longer progression-free survival in pituitary adenoma, thus acting like a tumor suppressor. H3 methylation may be correlated with methylation at certain CpG sites, which raises the question of the temporality between gene-level and chromatin-level epigenetic changes in pathophysiology. Sirtuins are histone deacetylating enzymes that are expressed at higher values in GH-expressing compared with nonfunctional adenomas. Sirtuin expression also correlates inversely with size, but not aggressiveness or proliferation, in somatotrophs. Chromatin structure may also be modified by histone citrullination; upregulations in citrullinating enzymes may be an early pathogenic marker of prolactinomas in particular.

The downstream effect of these epigenetic modifications is to modify gene expression. Relative to normal pituitary tissue, pituitary tumors have differentially expressed mRNA transcripts, lncRNA transcripts, and miRNA transcripts. Numerous genes involved with cell growth and signaling show altered methylation status for pituitary adenomas, including cell cycle regulators, components of signal transduction pathways, apoptotic regulators, and pituitary developmental signals.

Discussion: The limited clinical predictive capacity of the current pituitary tumor classification system suggests that there are likely tumor subclasses that remain to be discovered. Ongoing epigenetic studies could provide a basis for adding methylation and/or acetylation screening to standard pituitary tumor workups to further inform clinical decision-making. A novel classification system that integrates histological, radiographic, genetic, and epigenetic information is likely to reflect the complexity of tumor biology and accurately characterize tumors.