Using global DNA-methylation and metabolome profiles to classify variants of unknown significance (VUS) in phaeochromocytoma and paraganglioma tumours

Abstract

Abstract Background Phaeochromocytomas and paragangliomas (PPGLs) are rare but strongly heritable neuroendocrine tumours. Most patients undergo genetic testing and approximately 40% will have a germline causal mutation. In 10% of patients, genetic analysis identifies uncharacterised mutations, or variants of unknown significance (VUSs). Tests to distinguish between benign and pathogenic mutations would aid clinicians' and patients' decision-making. One test could exploit the biology of a subset of PPGLs. Mutations in genes controlling the Tricarboxylic Acid cycle cause a deranged metabolome, accumulation of oncometabolites and global DNA-hypermethylation. Data show that this stable chemical modification will be greatly elevated in pathogenic over benign mutations, and be present alongside a driver metabolome. Methods Clinical characteristics were compiled for 3 patient groups: causal germline mutations, no identified mutation (sporadic) and VUS. Archival Formalin Fixed Paraffin Embedded (FFPE) material was identified locally and also provided by the international ENS@T collaboration. Tissue analysis was optimised on mouse FFPE tissue. Laser microdisection allowed collection of precise tissue areas. DNA and metabolite extraction was optimised before analysis was undertaken by Liquid Chromatography Mass Spectrometry (LC-MS). Results 93 patients were recruited locally. Clinical characteristics were collected for 25 sporadic and 56 germline cases, including 14 VUS cases. Optimisation experiments in mouse tissue show that discrete areas can be extracted by laser microdissection. LC-MS identifies cytosine, 5'-methylcytosine and 5'-hydroxymethylcytosine in 1 μg of DNA at biologically relevant ratios. TCA cycle metabolites were detected using LC-MS: succinate, fumarate and malate were found in the metabolite extract from the same amount of tissue. Conclusions This study utilises DNA hypermethylation and metabolomics for clinical application. It represents a large repository of PPGL VUS samples nationally. Global DNA-methylation levels and metabolite profiles are shown to be quantifiable from existing FFPE patient samples, using reduced amounts of patient material. Legal entity responsible for the study Guy's and St Thomas' NHS Trust, London, UK; King's College London, UK. Funding National Institute of Health Research, King's College London, Isaac Shapera Research Trust, The Bernice Bibby Research Trust. Disclosure All authors have declared no conflicts of interest.