Deceptive morphologic and epigenetic heterogeneity in diffuse intrinsic pontine glioma

// Marianna Bugiani 1, * , Sophie E.M. Veldhuijzen van Zanten 2, 3, * , Viola Caretti 4 , Pepijn Schellen 3 , Eleonora Aronica 5 , David P. Noske 3, 6 , William P. Vandertop 3, 7 , Gertjan J.L. Kaspers 2, 3, 8 , Dannis G. van Vuurden 2, 3 , Pieter Wesseling 1, 8 and Esther Hulleman 2, 3 1 VU University Medical Center, Department of Pathology, Amsterdam, The Netherlands 2 VU University Medical Center, Department of Paediatrics, Division of Oncology/Haematology, Amsterdam, The Netherlands 3 Brain Tumor Center Amsterdam, Cancer Center Amsterdam, VU University Medical Center and Academic Medical Center, Amsterdam, The Netherlands 4 Stanford University School of Medicine, Department of Neurology, Neurosurgery and Pediatrics, Stanford, California, USA 5 Academic Medical Center, Department of Pathology, Amsterdam, The Netherlands 6 VU University Medical Center, Department of Neurosurgery, Amsterdam, The Netherlands 7 VU University Medical Center and Academic Medical Center, Neurosurgical Center Amsterdam, Amsterdam, The Netherlands 8 Prinses Máxima Center for Pediatric Oncology and University Medical Center Utrecht, Utrecht, The Netherlands * These authors have contributed equally to this work Correspondence to: Sophie E.M. Veldhuijzen van Zanten, email: s.veldhuijzen@vumc.nl Keywords: diffuse intrinsic pontine glioma, histone 3, H3 K27M, trimethylation, intratumoral heterogeneity Received: April 15, 2017     Accepted: June 12, 2017     Published: July 31, 2017 ABSTRACT Historically, the diagnosis of diffuse intrinsic pontine glioma (DIPG) was based on typical imaging findings and clinical characteristics instead of pathology. However, the discovery of mutations in histone H3 variants, and the availability of tumor material for molecular analysis, has led to a paradigm shift in DIPG research and clinical practice. Using data from whole-brain autopsies in a series of nine DIPG patients with known histone mutational status, we here aim to review histopathological characteristics with special focus on intratumoral heterogeneity (ITH) and histone 3 K27 trimethylation (H3 K27me3). All DIPGs showed marked histologic ITH, with 56% even showing focal areas resembling a WHO grade I phenotype. As expected, H3 K27me3 immunoreactivity was lost in the tumors that were H3 K27M-mutated (seven patients; 67% H3.3, 11% H3.1). Strikingly, the H3K27 wildtype tumors (two patients; 22%) also contained H3 K27me3-immunonegative areas. Our study underscores the importance of the choice of the biopsy site, as ITH in DIPGs could theoretically lead to erroneous histological diagnoses with small biopsies. New in this respect is our finding that a substantial number of otherwise typical DIPGs has areas resembling WHO grade I tumors (esp. pilocytic astrocytoma, subependymoma). Furthermore, our study shows that negative H3 K27me3 immunohistochemistry in a DIPG does not imply a H3 K27-mutant tumor.