Genome-Wide DNA Methylation Differences in Patients With Non-Functioning Pituitary Adenomas With or Without Postsurgical Intervention

Abstract Background: More than half of all surgically treated patients with non-functioning pituitary adenomas (NFPA) experience tumor progression during their follow-up, with some of these adenomas exhibiting locally aggressive behavior. Tumor progression is strongly associated with excess morbidity and mortality. Tumor control is, therefore, of vital importance for the patient outcome. There is no reliable method to predict tumor behavior after the initial surgical treatment. The aim of this study was therefore to study associations between DNA methylation patterns and tumor progression in patients with NFPA. Methods: In a case-controlled trial, 43 patients operated for NFPA with immunohistochemical characteristics of gonadotroph subtype were selected. Cases included patients who needed reintervention due to tumor progression (Reintervention group, n = 26), and controls had residual tumor without tumor progression during at least five years of follow-up (Radiologically stable group, n = 17). Genome-wide methylation analysis of the tumor tissue from each patient was performed on the Infinium MethylationEPIC BeadChip platform. Results: Principal component analysis (PCA) showed that clinical course was a strong contributor to the difference in DNA methylation profiles between the two groups. Also, unsupervised hierarchical clustering of the top 5,000 most variable CpG sites revealed three methylation profiles corresponding to, (1) the Reintervention group, (2) the Radiologically stable group, and (3) a mix of the two groups. Although the overall average methylation levels in the two groups did not differ (P = 0.16), 142,521 Differentially Methylated Positions (DMPs) were identified. Of these, 650 showed a difference in beta value methylation larger than 0.2. Interestingly, among the genes harboring the most significantly DMPs was GABRA1, and among the genes containing the largest number of DMPs were GATA2, NUP93, and LGALS1. In addition, the three most hypomethylated and the three most hypermethylated DMPs were all significantly (P <0.05) independently associated with tumor progression free survival (Log-rank test, data not shown). Conclusion: In this novel explorative study, we found a methylation pattern associated with postoperative tumor progression requiring reintervention in patients with NFPA. Among the differently methylated genes we identified, some have previously been described to be involved in pituitary adenoma differentiation or cancer development. Our study is a step towards deciphering epigenetic mechanisms involved in tumor progression and finding epigenetic signatures that might be used to predict postoperative tumor progression in NFPAs.