EJC Paediatric Oncology (Dec 2024)

The impact of methylome analysis on the diagnosis and treatment of CNS tumours in children and adolescents: A population-based study in Greece

  • Maria Filippidou,
  • Stavros Glentis,
  • Ilona Binenbaum,
  • Martin Sill,
  • Kleoniki Roka,
  • Antonia Vlachou,
  • Georgia Avgerinou,
  • Jonas Ecker,
  • Florian Selt,
  • Martin Hasselblatt,
  • Mirjam Blattner-Johnson,
  • Kathrin Schramm,
  • Clio Trougkou,
  • Dimitrios Doganis,
  • Nikolaos Katzilakis,
  • Vita Ridola,
  • Evgenia Papakonstantinou,
  • Vassilios Papadakis,
  • Emmanouel Hatzipantelis,
  • Eleftheria Kokkinou,
  • Roser Pons,
  • Christina Kanaka-Gantenbein,
  • Dominik Sturm,
  • Steffen Hirsch,
  • Nicola Dikow,
  • Kristian W. Pajtler,
  • Cornelis M. van Tilburg,
  • Michael C. Frühwald,
  • Till Milde,
  • Olaf Witt,
  • David T.W. Jones,
  • Andreas Von Deimling,
  • Felix Sahm,
  • Kalliopi Stefanaki,
  • Stefan M. Pfister,
  • Antonis Kattamis

Journal volume & issue
Vol. 4
p. 100198

Abstract

Read online

Background: The recently published WHO classification of central nervous system (CNS) tumours recognizes DNA methylation profiling as a desirable and, for some diagnoses, essential diagnostic tool adjunctive to conventional histopathology. DNA methylation profiling is not routinely available in many countries, including Greece. Methods: In this collaborative study, we report the DNA methylation results in a series of children and adolescents with CNS tumours in Greece (2018–2023). In total, 130 tumour samples were analyzed using the latest applicable version of the Heidelberg brain tumour classifier. Results: Upon initial analysis, 80 % (104/130) achieved calibrated scores (Cs) ≥ 0.9 and matched an established methylation class family/subclass. Among them, methylation results confirmed (90/104, 86.5 %), refined (50/104, 48 %) or changed (10/104, 9.6 %) the histological diagnosis. Only four results were regarded as non-contributing (4/104, 3.9 %). Twenty-six tumour samples received Cs < 0.9. Despite low scores, methylation results supported the initial diagnosis with lower confidence in 38.5 % (10/26) and established the diagnosis in two tumours with non-conclusive histopathology. Additional t-distributed stochastic neighbour embedding (t-SNE) analysis allowed the possible classification of twelve tumours. Nine more samples reached high Cs using the newer brain tumour classifiers, since available. Samples co-tested in Greece demonstrated excellent test reproducibility, supporting the analysis' local implementation. Methylome profiling impacted the clinical management of 40 % of patients, modifying stratification, prognosis, or treatment approach. Conclusions: This study supports the need to integrate methylome analysis into routine diagnostics in our country and highlights the importance of collaboration between European pediatric oncology centres.

Keywords