Osteosarcoma

PurposeIdentifying discrete subgroups associated with treatment response and resistance in localized Ewing sarcoma (EWS) remains a challenge. The primary objective of the Children's Oncology Group (COG) biology study AEWS18B1-Q was to molecularly characterize patients with localized EWS on prospective modern-day trials.Patients and methodsWe analyzed clinical and molecular features from patients with localized EWS enrolled on frontline COG trials. All patients had available formalin-fixed paraffin-embedded (FFPE) tissue, frozen tissue, or whole-genome-amplified material. Sequencing was performed for identification of canonical fusions, recurrent copy number alterations (CNAs), and alterations in TP53 and STAG2. Available tissue was analyzed for loss of STAG2 protein expression. Molecular features were evaluated for their association with cumulative incidence of relapse in univariate and multivariable analyses.ResultsThree hundred fifty-one patients had sufficient tissue, which in most cases was extracted from two FFPE slides. EWS canonical fusions were identified in 282 patients (80.3%). Pathogenic mutations in TP53 and STAG2 were identified in 5.1% and 7.6% of patients, respectively. A total of 63.1% of patients were found to have recurrent CNAs. In univariate analysis, there was an increased cumulative incidence of relapse in patients with TP53 mutation (5-year cumulative incidence of relapse 43%, 95% CI [17% to 67%] v 22%, 95% CI [17% to 27%]; Gray's test P = .039), STAG2 mutation (53%, 95% CI [29% to 73%] v 21%, 95% CI [16% to 26%]; P v 16%, 95% CI [9% to 24%]; P = .005). In a multivariable analysis, STAG2 mutation was the only molecular biomarker that remained prognostic.ConclusionThis is a prospective validation of the molecular prognostic features of patients with localized EWS receiving standard-of-care therapy on therapeutic clinical trials. Building on previous work, patients with STAG2 mutations were at high risk of relapse.

Osteosarcoma (OS) is a relatively rare bone malignancy that primarily affects children and young adults and is associated with significant morbidity and mortality. Cisplatin is a mainstay of treatment, but its efficacy is limited by off-target toxicities. Immunotherapy is not effective due to a poor antigenic tumor microenvironment. Here, we address these challenges by using manufactured M1 macrophage-derived vesicles (MVs) loaded with cisplatin. Human blood and mouse RAW 264.7 M1 macrophages were used to prepare empty (E-MVs) and cisplatin-loaded MVs (C-MVs). Human OS cell lines were used in vitro and in a tibia xenograft mouse model to evaluate the anti-cancer and immune-stimulating abilities of MVs. C-MVs had lower IC50s but equivalent DNA damage in OS cell lines when compared with free cisplatin. E-MVs and C-MVs were observed to accumulate in the tumor in OS tumor-bearing mice. C-MVs significantly reduced tumor burden and prolonged survival in a mouse model of OS. Animals dosed with free cisplatin experienced weight loss and renal and hepatic toxicity, while equivalent doses of C-MVs did not cause these effects. In addition, both E-MVs and C-MVs showed immunomodulation of the tumor microenvironment with a significant increase in the M1/M2 macrophages ratio (7-fold and 22-fold, respectively) and increased levels of TNF-α in serum (1.8-fold and 2.1-fold, respectively) compared to control mice. Collectively, these experiments support further development of C-MVs for the treatment of OS.

BackgroundOsteosarcoma is a highly aggressive bone malignancy characterized by frequent metastasis and therapy resistance. Although mitophagy and pyruvate metabolism are increasingly recognized as critical metabolic regulators, their interaction in osteosarcoma remains poorly understood. The autophagy-related protein GABARAP, central to mitochondrial quality control, has not been systematically evaluated in osteosarcoma.MethodsSingle-cell RNA sequencing (scRNA-seq) datasets (GSE162454, GSE237070) were analyzed to delineate cellular heterogeneity and malignant states, with prognostic clusters identified by Scissor and inferCNV. Tumor microenvironment (TME) composition and intercellular signaling were profiled using CellChat. Pathway enrichment and multi-omics integration across TARGET, GSE21257, and GSE32981 highlighted mitophagy-pyruvate coupling, which were further validated by spatial transcriptomics and in vitro functional assays.ResultsWe mapped the osteosarcoma ecosystem and identified two malignant subpopulations, Ost_1 and Cho_2 (Mal_Ost/Cho), exhibiting high genomic instability, stemness, and poor prognosis. The osteosarcoma TME displayed profound immune remodeling, characterized by infiltration of T/NK cells alongside enrichment of immunosuppressive Tregs and M2-polarized macrophages. Enhanced MIF-mediated signaling between Mal_Ost/Cho and T/NK compartments suggested a key mechanism of immune evasion. Both malignant subtypes demonstrated coordinated activation of mitophagy and pyruvate metabolism, sustaining metabolic adaptation and tumor progression. Multi-omics integration pinpointed GABARAP as a central hub regulating this mitophagy-metabolism axis, spatially enriched within metabolic hotspots and immunosuppressive niches. Functionally, GABARAP depletion disrupted mitophagy flux, mitochondrial integrity, and energy production, thereby impairing osteosarcoma cell proliferation and migration.ConclusionThese findings reveal that GABARAP links mitophagy-driven metabolic adaptation with immune evasion, representing a key regulator and potential therapeutic target in osteosarcoma.

Tumor cell plasticity drives metastasis and therapy resistance, yet its regulation by oncoprotein dosage dynamics remains poorly understood. In Ewing sarcoma (EwS), variations in EWS::FLI1 (EF) fusion oncoprotein activity have been associated with epithelial-mesenchymal plasticity (EMP). Using degron technology, we precisely modulated endogenous EF in EwS cells and linked phenotypic states to distinct oncoprotein dosages. Strikingly, modest EF depletion promoted a pro-metastatic phenotype that diminished upon near-complete EF loss, revealing a paradoxical effect of submaximal EF inhibition. Nascent RNA-sequencing uncovered distinct gene clusters with heterogenous transcriptional responses to graded EF loss. Genes most sensitive to subtle EF depletion harbored GGAA microsatellites within EF-bound enhancers, while chromatin profiling uncovered candidate cofactors regulating EF-repressed EMP programs. Transient EF depletion followed by rapid restoration, modelling oncoprotein fluctuations, caused persistent dysregulation of genes functionally linked to enhanced extravasation and metastatic burden in preclinical models. This study highlights the therapeutic challenge of incomplete EF elimination, serving a paradigm in which oncoprotein dosage dynamics act as non-genetic drivers of disease progression and reveal novel vulnerabilities of advanced disease.