Lebrun Laetitia, Gilis Nathalie, Dausort Manon, Gillard Chloé, Rusu Stefan, Slimani Karim, De Witte Olivier, Escande Fabienne, Lefranc Florence, D'Haene Nicky, Maurage Claude Alain, Salmon Isabelle
Department of Pathology, Université Libre de Bruxelles (ULB), Hôpital Universitaire de Bruxelles (HUB), CUB Hôpital Erasme, Erasme University Hospital, Rue Meylemeersch 90, 1070, Brussels, Belgium.
Department of Neurosurgery, Université Libre de Bruxelles (ULB), Hôpital Universitaire de Bruxelles (HUB), CUB Hôpital Erasme, Erasme University Hospital, 1070, Brussels, Belgium.
Sci Rep. 2025 Jan 22;15(1):2857. doi: 10.1038/s41598-025-87079-4.
Over the past decade, neuropathological diagnosis has undergone significant changes, integrating morphological features with molecular biomarkers. The molecular era has successfully refined neuropathological diagnostic accuracy; however, a substantial number of CNS tumor diagnoses remain challenging, particularly in children. DNA methylation classification has emerged as a powerful machine learning approach for clinical decision-making in CNS tumors. The aim of this study is to share our experience using DNA methylation classification in daily routine practice, illustrated through clinical cases. We employed a classification system to evaluate discrepancies between histo-molecular and DNA methylation diagnoses, with a specific focus on adult versus pediatric CNS tumors. In our study, we observed that 40% of cases fell into Class I, 47% into Class II, and 13% into Class III among the "matched cases" (≥ 0.84). In other words, DNA methylation classification confirmed morphological diagnoses in 63% of adult and 23% of pediatric cases. Refinement of diagnosis was particularly evident in the pediatric population (65% vs. 21% for the adult population, p = 0.006). Additionally, we discussed cases classified with low calibrated scores. In conclusion, our study confirms that DNA methylation classification provides significant added-value for CNS tumors diagnosis, particularly in pediatric cases.
在过去十年中，神经病理学诊断发生了重大变化，将形态学特征与分子生物标志物相结合。分子时代成功提高了神经病理学诊断的准确性；然而，相当数量的中枢神经系统肿瘤诊断仍然具有挑战性，尤其是在儿童中。DNA甲基化分类已成为中枢神经系统肿瘤临床决策的一种强大的机器学习方法。本研究的目的是分享我们在日常实践中使用DNA甲基化分类的经验，并通过临床病例进行说明。我们采用了一种分类系统来评估组织分子诊断和DNA甲基化诊断之间的差异，特别关注成人与儿童中枢神经系统肿瘤。在我们的研究中，我们观察到在“匹配病例”（≥0.84）中，40%的病例属于I类，47%属于II类，13%属于III类。换句话说，DNA甲基化分类在63%的成人病例和23%的儿童病例中证实了形态学诊断。诊断的细化在儿童群体中尤为明显（成人群体为65%对21%，p=0.006）。此外，我们还讨论了校准分数较低的病例。总之，我们的研究证实，DNA甲基化分类为中枢神经系统肿瘤诊断提供了显著的附加值，尤其是在儿童病例中。
