A pineal mass in a 39‐year‐old woman

Access at https://isn-slidearchive.org/?col=ISN&fol=Archive&file=BPA-24-09-CIR-276.R1.svs. A 39-year-old woman, with no significant medical history, presented progressively worsening behavioral problems. Brain MRI revealed a nodular lesion in the pineal region, with a necrotic center, heterogeneous peripheral enhancement, and an ependymal spread to the floor of the third ventricle and posterior part of the bulb (Figure 1). Germ cell markers (HCG, AFP) in CSF were negative. Because of the development of obstructive hydrocephalus, a ventriculocisternostomy with endoscopic ventricular biopsy (three biopsies from 1 to 5 mm) was performed, followed by a resection 1 month later. The surgical specimen was composed of fragments with a total weight of 3 g (Box 1). Histologically, the lesion had a predominantly piloid morphology with some radial perivascular arrangements on an abundant myxoid background. Most cells showed an eosinophilic cytoplasm and bi- or multipolar extensions as well as round to angular nuclei with homogeneous chromatin. Floret-like multinucleated cells were present. In certain places, the glial cells were arranged in cords, giving the tumor a chordoid morphology. Cytonuclear atypia ranged from mild to pronounced. Mitotic index was 4 mitoses per 10 fields. Endothelial vascular proliferation and necrotic areas were present. No Rosenthal fibers or granular bodies were found. Tumor cells were immunopositive for GFAP and OLIG2 as well as focally for synaptophysin, but were negative for CD34, TTF1, EMA, NeuN, and Chromogranin A. ATRX, H3K27me3, INI1, and BRG1 expression was retained (Figure 2). No mutation or gene rearrangement was detected by next-generation sequencing (NGS), RNA sequencing, or droplet digital PCR (FGFR1) in panel genes, including MAPK pathway genes (BRAF, KRAS, HRAS, and FGFR1-3). The DNA methylation array data, obtained using EPICv2 Beadchips, was analyzed using the DKFZ brain classifier v12.8 (https://www.molecularneuropathology.org/) and the resulting methylation profiles matched (calibrated score 0.99) with the methylation class of high-grade astrocytoma with piloid features (HGAP). The profiles were then clustered within the HGAP class using Uniform Manifold Approximation and Projection (UMAP). Copy number analysis showed homozygous deletion of the CDKN2A/CDKN2B loci (Figure 2). High-grade astrocytoma with piloid features (HGAP). Gliomas account for about 14%–22% of all pineal region tumors. They are mainly circumscribed low-grade astrocytomas, especially in children, and diffuse high-grade gliomas. The 2021 WHO classification has expanded the scope of circumscribed gliomas to include HGAP. This novel and rare glioma, mainly occurring in the adult cerebellum, has not been previously reported in the pineal region. This particular tumor, with a distinctive DNA methylation profile essential to the diagnosis, without specific histological or molecular hallmarks, may mimic other malignant or benign gliomas developing in the pineal region [1]. Before DNA methylation profiling, our case could only be classified as a glioma, NOS, with a predominant piloid and a focal chordoid appearance on a myxoid background. Gliomas usually described in the pineal region were considered. Among them, a diffuse midline glioma was rapidly excluded because of the presence of H3K27 trimethylation and the absence of H3K27M mutation. A chordoid glioma, a well-circumscribed glial tumor and usually occurring in the anterior part of the third ventricle, was also eliminated based on the immunonegativity of TTF1. The morphological diagnosis of a desmoplastic myxoid tumor (DMT), SMARCB1 mutant [2], of the pineal region was also considered. DMT is a novel entity recently added to the pineal tumor category in the 2021 WHO classification. DMT is formed by cords of small cells in a myxoid and collagenous background. The mitotic count is generally low. This GFAP-negative tumor is characterized by a loss of expression of SMARCB1/INI1 and classically expresses CD34 and EMA. These immunohistochemical criteria were absent in our case, allowing us to rule out this diagnosis. The histology and immunophenotype of our case remained compatible with a pilocytic astrocytoma, in particular in its pilomyxoid form. It was only by studying the methylation profile that the diagnosis of HGAP could be made. Although HGAP has been reported mainly in the cerebellum, it can theoretically occur throughout the central nervous system. In addition, the histological features of HGAP may vary from low- to high-grade glioma-like and, therefore, without further molecular testing, there is no clear definition of the tumor. Molecularly, alterations in the MAPK pathway, homozygous deletion of CDKN2A/CDKN2B loci, and alteration of ATRX are commonly found [1] and may point towards the diagnosis of an HGAP but are inconstant. Except for NF1 and ATRX mutational status, the main molecular alterations have been investigated in our case. Our case was highly challenging because of the conservation of ATRX in immunohistochemistry and the absence of detected alteration in the MAPK pathway. Appropriate identification of HGAP is crucial as this entity is associated with a poor clinical outcome despite a potential low-grade appearance [3]. In conclusion, this case we have described widens the spectrum of gliomas in the pineal region. Epigenetic characterization was critical in this case to reach a definitive diagnosis of HGAP. This case highlights the importance of advanced molecular studies, including whole-genome methylation profiling, in the evaluation of brain tumors. Charlotte Paoli: Writing—original draft (equal). Thibault Fabas: Data curation (equal); formal analysis (equal). Lydiane Mondot: Data curation (equal); formal analysis (equal); writing—review and editing (equal). Catherine Godfraind: Data curation (equal); formal analysis (equal); writing—review and editing (equal). Bérengère Dadone-Montaudié: Data curation (equal); formal analysis (equal); writing—review and editing (equal). Fanny Burel-Vandenbos: Conceptualization (lead); supervision (lead); writing—review and editing (lead). The authors thank Ms. Jessica Zufelt for her help with the English language and Dr. Julien Boyer for his help in umap presentation. The authors thank the Conseil Departemental des Alpes-Maritimes 06 for its support in the creation of the epigenetic platform in the Department of Pathology and Molecular Oncology of the University Hospital of Nice. The authors declare no conflict of interest. Each patient whose case is reviewed in the French national RENOCLIP-LOC network is informed and gives consent. Data are available upon request to authors.