Abstract
BACKGROUND: Chordomas are rare bone malignancies originating from remnants of the notochord, primarily occurring at the base of the skull or sacrum. While typically slow growing with low-grade histology, chordomas can be locally aggressive and highly recurrent. MATERIALS AND METHODS: We retrospectively identified 29 patients with sacral chordomas and utilized archival tissue to uncover novel insights into chordoma molecular heterogeneity, using a combination of DNA methylation profiling, spatial transcriptomics and multiplexed immunofluorescence. RESULTS: Unsupervised clustering of methylation probes revealed two distinct epigenetic clusters, Cluster1 and Cluster2. Cluster2 was enriched for recurrent tumors and exhibited significant hypomethylation across most differentially methylated probes in intergenic and open sea regions. Multivariate linear regression analysis indicated that exon boundary and intergenic regions primarily drove these methylation changes. Digital Spatial Profiler (DSP) analysis identified differential gene expression among 9489 genes between tumor and stromal regions across 120 regions of interest, revealing 824 significantly differentially expressed genes. Tumor regions showed reduction of expression of major histocompatibility complex (MHC) class II genes and up-regulation of MHC class I genes compared with stroma. Multiplexed immunofluorescence revealed stromal enrichment of CD3+ and CD14+ cell populations, which was also associated with distinct survival groups. Stromal regions exhibited significant populations of immune-activated T cells and expression of immune checkpoint factors T-cell immunoglobulin and mucin-domain containing protein 3 and programmed cell death protein 1 (PD-1) in T-cell subsets. Stromal myeloid cells showed increased CD47 and PD-1 expression. CONCLUSIONS: Our study identified distinct epigenetic profiles in sacral chordomas, which were associated with recurrence, and revealed expression of checkpoint markers TIM3, CD47 and PD-1, warranting further investigation through functional validation.