Epigenetic Profiling Reveals Molecular Markers Associated with Metastatic Pituitary Adenoma

Introduction: Pituitary adenomas (also called pituitary neuroendocrine tumors, PitNET) occur in approximately 10% of the population and, while generally benign, often cause hormonal disturbances and exert pressure on adjacent neural structures. Although many adenomas respond well to standard treatments, some exhibit treatment resistance or even progress to metastasis. Currently, no consistent histological or molecular markers exist to accurately predict clinical outcomes in patients with pituitary adenomas. Recent advances in epigenetic analysis have proven effective in classifying and understanding many central nervous system tumors, offering a promising approach to explore the pathophysiology of pituitary adenomas.

Methods: We conducted epigenetic profiling of 228 pituitary adenomas, comprised of 156 (140 indolent and 16 metastatic) from a multi-institutional, international cohort, and 70 adenomas from publicly available datasets, utilizing methylation array analysis. Following data preprocessing, unsupervised consensus clustering was applied to the discovery cohort of 140 indolent adenomas and validation cohort of the 70 publicly available adenomas. Epigenetic profiles of metastatic carcinomas were subsequently compared to those of indolent adenomas using dimensionality reduction, differentially methylated probe (DMP), and enrichment analyses.

Results: Methylation profiling reliably separated adenomas into groups using unsupervised clustering, first by transcription factor subtype and then by hormone subtype, mirroring the current World Health Organization classification schema for pituitary adenomas. Comparison of epigenetic profiles of primary tissue from ACTH adenomas that later metastasized (the largest subgroup of metastatic adenomas in our study) versus primary tissue from indolent ACTH adenomas revealed over 5000 DMPs, with most promoter-region DMPs being hypermethylated in metastatic adenomas. Enrichment analysis revealed hypermethylation of numerous tumor suppressor genes (ex. EDNRB, FOXP1, RUNX3) and biological processes typically dysregulated in malignancy (ex. DNA damage repair, telomere maintenance, DNA damage response).

Conclusion : Epigenetic analysis of pituitary adenomas effectively distinguishes adenomas based on their hormonal and transcription factor subtypes. Additionally, epigenetic differences between indolent and metastatic adenomas are detectable even at initial tumor resection, with hypermethylated genes and pathways in metastatic lesions highlighting potential therapeutic targets for aggressive adenoma behavior.