Abstract
Adult spinal pilocytic astrocytomas (sPA) pose a challenge due to their rarity and limited understanding of clinical progression and biology. Using epigenetic methodologies and a database of diffusely infiltrative intramedullary spinal cord gliomas (IMSCG), this study characterizes clinical and molecular aspects of adult intramedullary pilocytic astrocytomas. We analyzed patients who underwent surgery for diffusely infiltrating IMSCG at four German university medical centers. Clinical data were retrieved from digital records. To identify molecular patterns and provide diagnosis, we used the 850K DNA methylation array, with classification using the Heidelberg classifier. T-distributed stochastic neighbor embedding (t-SNE) identified clustering patterns. The study was approved by the ethics committee (PV4904). One-Hundred and thirty-three patients (n=133) with IMSCG were included. Of the 53 IMSCG methylation analysis samples, 18 tumors were identified as low-grade gliomas/pilocytic astrocytomas. The average age at surgery was 31 years. Near gross total tumor debulking was achieved in 32% of cases, while 26% underwent biopsy, with remaining cases being subtotal resections (42%) with no major complications. 32% of patients experienced decline in postoperative McCormick scores. The methylation array identified molecular matches in only 15% of patients using the 11.4 classifier and 23% using the latest 12.8 classifier. In the t-SNE plot, the 18 spinal PA formed a distinct cluster next to other pilocytic astrocytomas without clear overlap with existing tumor sample clusters. This differs from spinal H3K27-mutant diffuse midline gliomas, which could all be assigned molecular diagnosis with a matching score of at least 0.98. The histological and epigenetic profiling of PA demonstrates the challenges in making accurate diagnoses. Using this multicenter study, we present an epigenetic clustering pattern distinct from cerebral pilocytic astrocytomas’ profiles, indicating a distinct pathobiology.