High-grade astrocytoma with piloid features: a single-institution case series and literature review

High-grade astrocytoma with piloid features (HGAP) is a recently described primary brain tumor and the first requiring a specific methylation pattern for diagnosis, as its histologic features are often compatible with other tumors such as glioblastoma (GBM). Characterized by molecular alterations in CDKN2A/B, NF1, BRAF, FGFR1, and ATRX, they may be located anywhere in the CNS but show a predilection for the posterior fossa. Reports are limited to retrospective case series, and the standard of care is not yet established. We performed a retrospective review of electronic medical records of all patients with HGAP at our institution. Records were queried for demographic, radiographic, clinical, surgical, pathologic, and outcome data. Eighteen patients were included with a median 17.1 months follow-up. Of these, 12 (63.2%) were women with a mean age of 43 years (range 24-67). The most common tumor locations were the cerebellum (8 patients, 42.1%) and thalamus (6 patients, 31.6%). On imaging, tumors were most commonly homogeneously contrast-enhancing (10 patients, 52.6%) or rim enhancing with central necrosis (5 patients, 26.3%). Ten patients (52.6%) underwent biopsy, while nine (47.4%) underwent resection, of which four (44.4%) underwent gross total resection. Adjuvant therapy included radiation in 16 patients (88.9%) and systemic treatment in 16 patients (88.9%). The initial systemic treatment was temozolomide in 14 patients (77.8%). One patient received upfront trametinib (a MEK1 inhibitor), and one patient received upfront dabrafenib (a BRAF inhibitor). At last follow up, 11 patients (57.9%) had progressive disease. Median progression-free survival (PFS) was 5.4 months (range 1.6-28.2 months), and median overall survival (OS) had not been reached. HGAP is a newly described rare glial tumor without an established standard of care. Its aggressive behavior and targetable mutations warrant further investigation regarding predictors of outcome for this entity.