A posterior fossa mass in a 4‐year‐old female

Access at https://isn-slidearchive.org/?col=ISN&fol=Archive&file=BPA-25-03-CIR-104_2.svs. A 4-year-old female presented to the emergency department with persistent headaches, vomiting, and balance difficulties for 1 month. Neurological examination revealed significant ataxia and anisocoria. A posterior fossa heterogeneous mass with calcifications centered in the fourth ventricle was found, measuring approximately 5.0 × 4.7 × 4.2 cm (Figure 1). It was associated with severe supratentorial hydrocephalus. There was no evidence of spinal dissemination. She underwent placement of a right frontal external ventricular drain followed by suboccipital craniotomy and resection of the mass, with subsequent chemoradiation. Histologic sections revealed a high-grade neoplasm with complex morphology. Some areas showed definite embryonal morphology with densely packed small cells with high nuclear/cytoplasmic ratio and brisk mitotic activity (Box 1, Figure 2A). The majority of the remaining tumor showed a mixture of haphazardly arranged spindled cell bundles with abundant eosinophilic cytoplasm and striations, typical of skeletal muscle differentiation (Figure 2B), and relatively mature neurons within a neuropil matrix, indicative of advanced neuronal differentiation (Figure 2C). By immunohistochemistry (IHC), synaptophysin was positive in the embryonal component and areas with neuronal differentiation. Positive YAP1 stain was present and limited to the tumor cells with feature of skeletal muscle (Figure 2D). GAB1 stain was negative (Figure 2E), and beta-catenin expression was cytoplasmic (Figure 2F). Desmin, myogenin, and MyoD1 showed positivity in regions of skeletal muscle differentiation (Figure 2G,H). GFAP staining was largely confined to reactive astrocytes, while OLIG2 was negative. INI1 and BRG1 expressions were retained in the tumor cells. The IHC profile confirmed the morphological impression of medullomyoblastoma [1]. Chromosomal microarray and a Neuro-Oncology targeted next generation sequencing (NGS) panel were performed. The tumor showed partial gain of 1q and gain of chromosome 8, and harbored KBTBD4 (p.R313_W315delinsGSATMR) and SMARCA4 (p.R1189Q) pathogenic mutations. Two variants of uncertain significance (VUS) on PDGFRA (p.R293H) and NTRK1 (p.V321M) were also present. These findings were most consistent with a non-WNT/non-SHH group 3/ group 4 medulloblastoma, favoring subgroup 2 [2]. Whole-genome methylation analysis performed at the Pathology Laboratory, National Cancer Institute using the Bethesda version 2 classifier yielded a “no match” result but suggested that the tumor belonged to the medulloblastoma family (Confidence Score [CS] = 0.909). Dimensionality reduction using UMAP (uniform manifold approximation and projection) positioned the tumor within a region enriched with medulloblastomas group 3/group 4, subgroup 2 (Figure 2I). Using the St. Jude methylation classifier, the tumor matched to group 3/group 4 medulloblastoma, subgroup 2, with high confidence (CS = 0.99). Medulloblastoma, non-WNT/non-SHH, CNS WHO grade 4, with extensive skeletal muscle and neuronal differentiation (medullomyoblastoma). Medulloblastoma with myogenic differentiation (medullomyoblastoma), first described by Marinesco and Goldstein in 1933, is an exceptionally rare subtype of medulloblastoma (<1%). Until 2007, it was considered as a distinct variant in the World Health Organization (WHO) Classification of CNS Tumors [3]. However, it is now recognized as a differentiation pattern of medulloblastoma that can occur in any medulloblastoma subtype [2]. Due to its rarity, its clinical behavior and genetic alterations remain poorly understood. Limited reports suggest that medullomyoblastoma follows an aggressive clinical course, while long-term survival has been reported in other cases, highlighting the need for further studies to clarify its prognostic implications. The diagnosis of medullomyoblastoma is based on morphologic and immunohistochemical features. Some cases as the one presented here show areas classic of medulloblastoma alternating with areas demonstrating skeletal muscle differentiation. In other cases, however, skeletal muscle and neuronal differentiation may be very limited and/or only appreciated using IHC for desmin, myogenin, and MyoD1. An unusual YAP1/GAB1/beta-catenin staining pattern in medulloblastoma should prompt further testing for potential myogenic and/or melanocytic differentiation. Additionally, although this case carried a SMARCA4 mutation, the retention of INI1 and BRG1 expression effectively ruled out atypical teratoid/rhabdoid tumor. The molecular classification of medullomyoblastoma remains challenging, as these tumors do not align neatly with the current medulloblastoma molecular subgroups based on the 2021 WHO Classification. Some evidence suggests that medullomyoblastomas are distributed across the SHH, WNT, and group 3/group 4 molecular subgroups, indicating their molecular heterogeneity. Our case exhibited an IHC and molecular profile of a non-WNT/non-SHH group 3/group 4 medulloblastoma and appeared to align with subgroup 2, a finding supported by DNA methylation profiling in one of the two classifiers, likely reflecting variability in the classifiers' development and classification thresholds for this rare and heterogeneous tumor. In summary, while molecular classification plays a pivotal role in CNS tumor diagnoses, this case highlights the importance of histopathologic and molecular correlation. As medulloblastoma with myogenic differentiation does not seem to harbor a defined molecular profile, an integrated approach combining morphology, IHC, and molecular analysis is crucial for an accurate diagnosis. Yi Zhu and Vy Huynh contributed to the study conception and design. All authors contributed to material preparation, data collection and analysis, and have reviewed and approved the final manuscript. The diagnosis of this case was made in collaboration with Caterina Giannini, MD, PhD, for which we are grateful. We thank Tiffany A. Mainella, MS, for her assistance with digital slide scanning. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.