Deciphering the molecular heterogeneity of soft tissue sarcoma by integrating multiomics and single cell sequence

Background

Soft tissue sarcoma is a highly malignant tumor with extensive heterogeneity across multiple omics. However, a comprehensive multi-omics subtyping system has not yet been established.

Methods

We integrated sarcoma multi-omics data, including clinical information, transcriptome expression profiles, DNA methylation, and somatic mutations. Using ten advanced clustering algorithms, we identified robust subtypes and validated the reproducibility of our analysis in two independent external datasets. We also identified subtype-specific treatment strategies and analyzed the differences in microenvironments between subtypes using single-cell data.

Results

Based on multi-omics subtyping, we identified two novel sarcoma molecular subtypes, named sarcoma multi-omics subtype 1 (SAMS1) and SAMS2. SAMS2 exhibited a poorer prognosis, with significantly activated Myc, glycolysis, and Wnt beta-catenin signaling pathways. SAMS2 was characterized by a lower abundance of immune cell infiltration and anti-tumor immunity deficiency, which owned a lower response rate to immunotherapy but was sensitive to certain targeted drugs, including pazopanib, axitinib, thapsigargin, and elesclomol. MK886 and NU1025 were identified as effective therapeutic targets for the SAMS2. In SAMS2-like tumor epithelial cells, HOXB13/COL16A1 and BASP1 regulated epithelial-mesenchymal transition. We found that WNT7B was highly expressed in STS and was associated with poor patient prognosis, suggesting its potential as a novel therapeutic target for STS patients.

Conclusion

The STS molecular subtyping system based on multi-omics data effectively distinguishes patients with poor prognosis. The subtyping results are robust and reliable, providing new insights for the precise diagnosis and treatment of these patients.