Analysis of the implementation of DNA methylation assay for routine CNS tumors diagnosis

Abstract

Introduction: In the last years, DNA methylation array emerged as a new tool for CNS tumor diagnosis. Objectives: To analyze the value of methylation assay for routine CNS tumor diagnosis. Patients and Methods: We performed methylation array in 182 CNS tumors (100 gliomas (GL), 27 medulloblastomas (MDB), 20 ependymomas (EP), 17 neuronal/ glioneuronal tumors (N/GNT), 5 meningiomas, 7 embryonal tumors (ET), and 6 sarcomas) from 2019 for subtyping GL, EP, MDB, meningiomas, and sarcomas, defining the diagnosis in cases without precise histological features or confirming the histological diagnosis. Results: Scores were >0.75 in 147 cases (81%). The histological diagnosis was maintained and allowed subtyping in 127 cases (70%) (77 GL, 27 MDB, 12 EP, etc.), defined differential diagnoses in 16 cases (9%) (3 sarcomas, 5 ET, 3 EP, etc.), and changed the diagnosis in 4 cases (2%). Scores were ≤0.75 in 35 cases (19%), however, in 7/14 adult diffuse GL, CNV analysis (EGFR amplification, and Chr.10q and CDKN2A/B deletions) led to the diagnosis of the cases as IDH-wt glioblastomas based on the WHO classification. In 8/35 cases displaying scores ≤0.75, RNA and/or DNA sequencing was performed and led to the diagnosis of 4 tumors (1 sarcoma, 2 DMG H3K27-altered, and 1 pilocytic astrocytoma). Thus, in 24 cases (13%) diagnosis could not be achieved; among these cases, 10 were N/GNT. Conclusion: The methylation array is a valuable tool to diagnose most CNS tumors because a single technique offers both the diagnosis and molecular data, mainly for adult diffuse GL, MDB, and EP. Nowadays, the main limitations of the method are the high amount of DNA necessary for performing the study, economic issues for developing countries, and the absence in the classifier of uncommon tumors, such as low-grade N/GNT, and some of the new tumors of the 2021 WHO classification. The growth of the case base will improve the identification skills of the system, probably perfecting its diagnostic capabilities.