Abstract

Background Osteosarcoma (OS) is the most common primary malignant bone tumor in adolescents, characterized by high heterogeneity and poor prognosis. Efferocytosis, the clearance of apoptotic cells, has been implicated in tumor progression and immune evasion, but its role in OS remains unclear. Methods We integrated TARGET-OS as a training cohort with three GEO datasets for validation. Efferocytosis pathways were quantified by ssGSEA, and WGCNA was applied to identify associated gene modules. Candidate genes were screened using univariate Cox regression, and a prognostic signature was developed with machine learning models and validated across cohorts. Functional enrichment, immune infiltration, and immunotherapy prediction analyses were performed. scRNA-seq from six OS patients and spatial transcriptomic profiling were further used to characterize the cellular distribution and communication of efferocytosis-related genes. The functions of MAGEA11 in OS were explored both in vivo and in vitro. Results The Brown module showed the strongest association with efferocytosis pathways. The StepCox + Ridge model achieved robust prognostic performance and stratified patients into risk groups with significantly different survival. Enrichment analysis revealed upregulated genes related to endothelial and nitric oxide pathways, while downregulated genes were linked to immune signaling and extracellular matrix remodeling. High-risk patients exhibited elevated M2 macrophages, altered checkpoint expression, and greater predicted sensitivity to immunotherapy. At the single-cell level, efferocytosis activity was enriched in OS cells, with MAGEA11 showing the highest expression. High-risk tumors displayed stronger intercellular signaling, particularly from OS cells and CAFs to macrophages and endothelial cells. Spatial transcriptomics confirmed enrichment of efferocytosis at tumor and interface regions, correlating positively with stromal and myeloid cells and negatively with T cells