<scp>DNA</scp> methylation as a predictor of pituitary neuroendocrine tumour behaviour: A systematic review

Pituitary neuroendocrine tumours (PitNETs) range from slow-growing to highly aggressive tumours; however, traditional prognostic markers often fail to predict clinical outcomes reliably. DNA methylation has recently emerged as a promising biomarker for assessing tumour behaviour. This systematic review evaluates its predictive value in PitNETs. To systematically assess the clinical applicability of DNA methylation profiles in predicting behaviour of PitNETs. Systematic review. A comprehensive search was conducted in Medline, Embase, Web of Science, and Cochrane CENTRAL on December 13, 2024, with an update on October 17, 2025. The search included studies on adult PitNET patients, specifically examining tumour behaviour in relation to DNA methylation. Excluded were studies that focused on cell-free DNA, investigated a single gene with no established relevance to tumour behaviour, or assessed tumour size only. Data were extracted from 20 eligible studies by four independent reviewers. The risk of bias was assessed using the QUIPS tool. Due to methodological differences across studies, the findings were summarised narratively. Twelve studies investigated tumour invasiveness, two examined tumour aggressiveness and five examined PitNET regrowth, recurrence and re-intervention. The majority of studies concentrated on non-functioning PitNETs and used Illumina arrays or PCR-based methods. These analyses identified several differentially methylated genes linked to invasiveness (e.g., PHYHD1, WNT4, STAT6, CDH1, CDH13), aggressive behaviour (e.g., AIP, PDCD1, LINE-1), and tumour regrowth (e.g., TERT, FAM90A1, ING2). DNA methylation profiling shows potential for predicting PitNET behaviour, but methodological inconsistencies limit its clinical application. Standardized methods and prospective validation are needed for clinical integration.